text
stringlengths 66
377k
|
|---|
Serum gamma-glutamyl transferase is a predictor of mortality in patients with acute myocardial infarction
Gamma-glutamyl transferase (GGT) is involved in the pathogenesis of atherosclerosis and has been associated with adverse cardiovascular outcomes in patients with ischemic heart disease. However, the association between GGT and long-term mortality has not been studied in patients with acute myocardial infarction (AMI).A total of 2239 AMI patients for whom serum GGT values were available and who underwent percutaneous coronary intervention (PCI) were enrolled in the COREA-AMI (CardiOvascular Risk and idEntificAtion of potential high-risk population in Korean patients with AMI) registry. Patients with acute liver injury were excluded. Patients were classified into 2 groups according to normal (n = 1983) or elevated (n = 256) levels of serum GGT. The primary clinical outcome was all-cause mortality. The secondary outcome was cardiac death and recurrent non-fatal myocardial infarction (MI).The median follow-up period was 3.7 years, and both groups had similar characteristics. Patients with elevated GGT had significantly higher all-cause mortality compared to patients with normal GGT (21.9% vs. 14.4%, P = .001). The multivariate Cox proportional hazards model showed that elevated serum GGT level was independently correlated with mortality (hazard ratio 2.12 [1.44-3.11]; P < .001). Although elevated serum GGT was independently associated with long-term mortality after 30 days after PCI, there was no association within 30 days after PCI. Elevated GGT was also associated with death of cardiac causes with statistical significance. In the subgroup analysis, stronger associations were observed in the young and female patients and in patients who had ST-segment elevation MI and preserved left ventricular ejection fraction at the first echocardiography after the indexed PCI.Elevated serum GGT is an independent predictor of long-term mortality in AMI patients.Abbreviations: 2xULN = two-times of upper limit of normal, ALP = alkaline phosphatase, ALT = alanine aminotransferase, AMI = acute myocardial infarction, AST = aspartate aminotransferase, AUC = areas under the curve, BMI = body mass index, CAD = coronary artery disease, CI = confidence interval, ECG = electrocardiogram, GGT = gamma-glutamyl transferase, HbAc1 = glycated hemoglobin, HDL-C = High-density lipoprotein cholesterol, HR = hazard ratio, hsCRP = high-sensitivity C-reactive protein, IQR = interquartile range, LDL-C = low-density lipoprotein cholesterol, LVEF = left ventricular ejection fraction, MI = myocardial infarction, NSTEMI = Non ST-segment elevation myocardial infarction, PCI = percutaneous coronary intervention, ROC = receiver-operating characteristic, STEMI = ST-segment elevation myocardial infarction, TIMI = thrombolysis in myocardial infarction.
# Introduction
The enzyme gamma-glutamyl transferase (GGT) is present in the serum and on the surface of various cell membranes. GGT is considered a marker of liver or biliary tract diseases and alcohol consumption. However, GGT has recently been identified as a novel indicator of the development and prognosis of cardiovascular diseases. Although the exact mechanism has not been elucidated, the abundance of GGT in atheroma and its function in blood vessels may play a role. GGT catalyzes the first step in the extracellular degradation of glutathione. During the process of GGT-mediated glutathione degradation, low-density lipoprotein (LDL) is oxidized and accumulates in the arterial wall; this process is involved in the pathogenesis of atherosclerosis. [bib_ref] Gamma-glutamyl transpeptidase-dependent iron reduction and LDL oxidation-a potential mechanism in atherosclerosis, Paolicchi [/bib_ref] Moreover, degradation of the antioxidant glutathione results in formation of peroxide free radicals and, consequently, oxidative stress. Thus, the combination of abundant oxidized LDL and GGT in atherosclerotic plaques causes oxidative stress in the endothelium, which can affect plaque evolution and rupture. [bib_ref] Gamma-glutamyltransferase, atherosclerosis, and cardiovascular disease: triggering oxidative stress within the plaque, Emdin [/bib_ref] Furthermore, many studies have reported an association between serum GGT levels and various established cardiovascular disease risk factors, such as hypertension, diabetes, metabolic syndrome, and coronary artery disease (CAD). [bib_ref] The significance of serum gamma-glutamyltransferase in cardiovascular diseases, Pompella [/bib_ref] [bib_ref] Gamma-glutamyltransferase levels and risk of metabolic syndrome: a meta-analysis of prospective cohort..., Liu [/bib_ref] [bib_ref] Serum gamma-glutamyltransferase: independent predictor of risk of diabetes, hypertension, metabolic syndrome, and..., Onat [/bib_ref] [bib_ref] Gamma glutamyl transferase and metabolic syndrome, cardiovascular disease, and mortality risk: the..., Lee [/bib_ref] In addition, increased GGT levels in established CAD patients have been associated with an increase in secondary events, including myocardial infarction (MI), stroke, and cardiovascular death. [bib_ref] Gamma-glutamyltransferase as a risk factor for cardiovascular disease mortality: an epidemiological investigation..., Ruttmann [/bib_ref] [bib_ref] Serum gamma-glutamyltransferase predicts non-fatal myocardial infarction and fatal coronary heart disease among..., Lee [/bib_ref] However, studies of the association between serum GGT levels and long-term clinical outcomes in patients with acute MI (AMI) have included only a few patients with ST-segment elevation MI (STEMI) and yielded inconsistent results. [bib_ref] Effect of serum gamma-glutamyl transferase levels on myocardial perfusion and long-term prognosis..., Ozcan [/bib_ref] [bib_ref] The relationship between gammaglutamyl transferase levels and the clinical outcomes in patients..., Gul [/bib_ref] Therefore, we investigated whether higher serum GGT levels can predict short-term and long-term mortality in patients with AMI.
# Methods
This study used data from the CardiOvascular Risk and idEntificAtion of potential high-risk population in Korean patients with AMI (COREA-AMI) registry, which was designed to evaluate real-world outcomes in "all-comers" with AMI. The COREA-AMI, a large, observational registry included clinical, angiographic, short-term and long-term outcome data for AMI patients who underwent percutaneous coronary intervention (PCI) at 9 major cardiac centers in Korea between January 2004 and December 2009. [bib_ref] Benefit of beta-blocker treatment for patients with acute myocardial infarction and preserved..., Choo [/bib_ref] Initially, our study sample included a total of 4748 patients. Among these, 2281 patients who had serum GGT values were available were enrolled. To avoid the confounding effects of unknown underlying active liver disease on the prognosis, patients who had a serum alanine aminotransferase (ALT) level >3 times the upper limit of normal (ULN) and the ALT greater than the level of aspartate aminotransferase (AST) were excluded (n = 28). [bib_ref] Alanine aminotransferase: a clinical and regulatory tool for detecting liver injury-past, present,..., Senior [/bib_ref] To avoid unreasonable deviation of GGT level, the data were trimmed with exclusion of extreme values. The value of 5 patients (who were excluded because of lower than measurable range) was recorded near to zero; these cases were deleted for the possibility of data collection errors. We screened the accessible electrical medical records of the patients from the highest GGT level. The number of patients with GGT level over 2Â ULN), 3Â ULN, and 5Â ULN were 47 (2.1%), 14 (0.6%), and 9 (0.4%), respectively. We excluded the only 9 patients of over 5Â ULN because 3 of them had obvious hepatobiliary problems; one died from GB cancer, another had recurrent cholangitis, and the other had pancreatic disease. We could not find clear reason of GGT elevation among the others. Finally, a total 2239 of patients were included in this analysis [fig_ref] Figure 1: Study flow chart [/fig_ref].
AMI was diagnosed based on characteristic clinical symptoms, serial changes on electrocardiograms (ECGs) consistent with infarction, and increased cardiac enzyme values. The diagnosis was confirmed by coronary angiography in all patients. We excluded patients who were not indicated for PCI based on coronary angiography to strengthen the homogeneity of the study population. All patients received standard medical treatment during PCI and hospitalization. The study protocol was approved by the institutional review board at each participating center and is in accordance with the Declaration of Helsinki. All patients provided written informed consent at the time of admission to enrollment in the registry and the use of their clinical data in future retrospective analyses.
We recorded demographic data, cardiovascular risk factors, and laboratory data for all patients. Cardiovascular risk factors included smoking status, previously diagnosed diabetes mellitus, hypertension, chronic kidney disease, and history of familial CAD. Data on other risk factors were reported by the patients themselves or extracted from medical records. Blood samples were drawn within 24 hours of the initial visit and used for a standard battery of hematological and biochemical tests. Serum GGT levels were measured using the enzymatic colorimetric test at 37°C, and L-g-glutamyl-3-carboxy-4-nitroanilide was used as the substrate at each cardiac center under identical conditions. [bib_ref] The relationship between gammaglutamyl transferase levels and the clinical outcomes in patients..., Gul [/bib_ref] Patients were categorized into 2 groups based on elevated or normal serum GGT levels compared to the upper limit of the clinical reference range. The normal reference range was 9 to 85 U/L for males and 5 to 55 U/L for females. All procedures were performed according to current standard guidelines. The specific drug-eluting stent used in the procedures was chosen by the operator. The operator assessed the type of lesion according to American College of Cardiology/American Heart Association guidelines. After the procedure, aspirin was prescribed indefinitely, and clopidogrel was prescribed for at least 6 months. Immediate post-procedural and in-hospital events were recorded. Patient follow-up was conducted during office visits or through telephone interviews at 1, 6, and 12 months and annually thereafter. Echocardiography was performed within 3 days of the PCI, and a quantitative assessment of the left ventricular systolic function was performed using the modified biplane Simpson method to calculate the left ventricular ejection fraction (LVEF).
The primary objective of this study was to evaluate the association between GGT level and all-cause mortality during clinical follow-up post intervention. The secondary objectives were to evaluate the association between high GGT levels and cardiac death and recurrent non-fatal MI. Cardiac death was defined as death from CAD, heart failure, or arrhythmia, and death was attributed to cardiac events unless non-cardiac death could be clearly identified. [bib_ref] Clinical end points in coronary stent trials: a case for standardized definitions, Cutlip [/bib_ref] Recurrent MI was defined as the presence of recurrent symptoms and new ECG changes that considered to be MI or cardiac markers that were at least twice the normal limit. Medical records were thoroughly reviewed by an independent research nurse. Telephone interviews were conducted to collect data on the occurrence of adverse events following PCI. Clinical outcomes of interest were confirmed by source documents and centrally adjudicated by a local events committee at the Cardiovascular Center of Seoul St. Mary's Hospital and an independent group of clinicians who were unaware of patient status. To verify the accuracy of mortality data, we matched our data to official national data collected by the National Statistical Office from death certificates, which previous studies have shown to be reliable. [bib_ref] Benefit of beta-blocker treatment for patients with acute myocardial infarction and preserved..., Choo [/bib_ref] We classified patients into two groups according to normal or high GGT levels and used these 2 categories in the subsequent analyses. Differences between groups of continuous variables were evaluated using an independent t-test or the Mann-Whitney U test. Differences in discrete variables were analyzed using a chisquare or Fisher's exact test and expressed as counts and percentages. Landmark analyses were performed to evaluate the impact of high serum GGT on short-term and long-term mortality. The landmark method of survival analysis uses a fixed time after PCI. In this study, the cut-off for early mortality was the 30th day after PCI. We constructed Kaplan-Meier curves to the end points for patients with normal GGT or high GGT, and differences between the groups were assessed by the log-rank test. Cox proportional hazard models were applied to calculate estimated hazard ratios (HRs) for each end-point. We selected covariates that differed significantly between the groups at baseline and that previous studies have related to GGT level or cardiovascular outcomes after PCI. [bib_ref] Gamma-glutamyltransferase to determine cardiovascular risk: shifting the paradigm forward, Turgut [/bib_ref] [bib_ref] The prognostic value of serum gamma glutamyl transferase activity on admission in..., Akpek [/bib_ref] The HRs were adjusted for important covariates that had significant effects (p < 0.05) on clinical outcomes in the univariate analysis. All analyses were two-tailed, and clinical significance was defined as p < 0.05. The same process was used for subgroup analyses to evaluate differences according to age, gender, STEMI or Non ST-segment elevation MI (NSTEMI), body mass index (BMI), high or low levels of LDL cholesterol (LDL-C), high or low levels of highdensity lipoprotein (HDL) cholesterol (HDL-C), hypertriglyceridemia (triglyceride >150), glycated hemoglobin(HbA1c), and LVEF ≥50% at the first echocardiography after the indexed PCI.
To analyze the association between GGT levels and the mortality in our study group of AMI patients, we computed receiver-operating characteristic (ROC) curves, tested for equality of the areas under the curves (AUCs), and calculated 95% confidence intervals (CIs) for GGT. Statistical analyses were performed using the statistical package SPSS V.20.0 (SPSS Inc., Chicago, IL) and MedCalc V.12.7 (MedCalc Software, Mariakerke, Belgium).
# Results
Patient GGT levels were non-normally distributed, and a high GGT level, defined as above the normal range, was observed in 256 patients (11.4%). The median GGT level was 31 ( interquartile range [IQR] 20-55; mean 46 ± 44.6) U/L in males and 21 (IQR 13-34; mean 29.3 ± 26.4) U/L in females, and the percentages of male and female patients with high GGT levels were 11.3% and 11.7%, respectively. The baseline characteristics of the GGT groups are summarized in . At baseline, patients in the high serum GGT group had more conventional cardiovascular risk factors compared to the normal serum GGT group. Patients in the high GGT group were younger than those Baseline characteristics of normal GGT group and high GGT group.
## Overall (n = 2239)
Normal GGT (n = 1983) High GGT (n = 256) P in the normal serum GGT group (mean age of 59.7 ± 13.6 vs. 62.5 ± 12.6 years, respectively). Compared with the normal GGT group, more patients in the high GGT group were obese (BMI ≥25 kg/m 2 ; 36.7% vs. 43.1%, P = .048). High-sensitive Creactive protein (hsCRP), uric acid, and serum triglyceride were positively associated with high serum GGT levels, and the difference of them was also significant. No reflow phenomenon after PCI was also more frequently observed in high GGT group (7.8% vs. 4.2%, P = .009). Biomarkers associated with liver disease, including AST, ALT, and alkaline phosphatase (ALP), were higher in the high GGT group. No other differences were observed between the two groups. A total of 341 deaths (15.2%) were recorded during a median follow-up time of 3.7 years (IQR: 2.4-5.0 years). The number of cardiac death and noncardiac death of high GGT versus normal GGT group was 31 (12.1%) vs. 153 (7.7%) and 19 (7.4%) vs.123 (6.2%), respectively. The proportion of unrevealed cause of death was 2.3% (n = 6) vs. 0.9% (n = 18). All-cause mortality during the entire follow-up period was significantly higher in the high GGT group than the normal GGT group (21.9% vs. 14.4%, P = .001 by the log-rank test). Early mortality at day 30 following PCI (5.9% vs. 3.2%, P = .03) and late mortality from day 30 to the end of follow-up (17.1% vs. 11.5%, P = .013) were also higher in the high GGT group than the normal GGT group. The Kaplan-Meier curves for all-cause mortality are presented in [fig_ref] Figure 2: Kaplan-Meier survival curves [/fig_ref]. Compared with the normal GGT group, the high GGT group had an age and sex adjusted HR for death of 1.97 (P < .0001, Model 1). Additional adjustment for differences at baseline (Model 2: BMI ≥25 kg/m 2 , ALT, AST, ALP, uric acid, hsCRP, hypertriglyceridemia (triglyceride >150), and presence of no-reflow phenomenon after PCI) and other cardiovascular risk factors (Model 3: hypertension, diabetes, chronic kidney disease, current smoking, high LDL-C (LDL ≥100 mg/dL), Killip class ≥ II at admission and final TIMI flow <III after PCI) could not attenuated this relationship. Elevated GGT remained an independent risk factor for all-cause death (multivariableadjusted HR 2.12, P < .001) [fig_ref] Table 2: Hazard ratios and 95% confidence intervals for mortality according to the 2... [/fig_ref]. In landmark analysis, the effect of high GGT group on the long-term mortality after 30 days after indexed PCI was consistent with the result of overall period death (adjusted HR 1.81, P = .009), but there was no difference in early 30 days mortality between 2 groups. When we analyzed cardiac death separately, the association of GGT level was consistent with the former result (multivariable-adjusted HR 1.86 P = .037) . The total number of patients who suffered nonfatal MI was 43, the difference between high and normal GGT group was insignificant (n = 9 [3.5%] vs. n = 34 [1.8%], P = .107).
The results of the subgroup analysis are presented in . The association between serum GGT level and all-cause mortality differed among the specific groups. After stratifying by sex, an association between serum GGT levels and all-cause death was observed in both female and male patients. However, the association was stronger in female patients than in male patients (P for interaction, <.001). When stratified by the presentation of MI depending on the ST-segment change, the association between serum GGT levels and all-cause mortality was only observed among STEMI patients (P for interaction, <0.001). Also, the association between high serum GGT levels and allcause mortality was significant in who had preserved LVEF at admission (LVEF ≥50%, P for interaction, .023). The association between high GGT levels and all-cause mortality was stronger in patients under the age of 65 years (P for interaction, .019) and in those had high HDL-C (P for interaction, .002), but the HR value of each subgroup showed just a slight difference. Additionally, the lower the HbA1c level of subgroups, the association tended to be stronger, but there were too many missing values (n = 1090) to show statistical significance (P for trend, <.452). The difference between obese and nonobese patients was not clear. Whether the value of GGT was more meaningful in low LDL-C (<100 mg/dL, HR = 2.13, P = .003) was not obvious (P for interaction = .31). The baseline difference of male and female subgroup is presented in . Male and female AMI patients have very different cardiovascular risk factors, respectively. Men tend to be obese and have more history of current smoking. Male AMI patients' uric acid level is higher than women's and HDL-C is lower. Female AMI patients are much older and have more hypertension and diabetes. Estimated glomerular filtration rate and hemoglobin are significantly lower in these patients, and total and LDL-C level is higher. Furthermore, more females presented as worse Killip class and NSTEMI. As a result, the prognosis of female AMI patients was shown much worse than male patients (mortality 22.4% vs.12.3%). On this basis, we repeated the same survival analysis on each sex group.is a summary of the association of high GGT and mortality in according to sex. After stratifying by sex, a stronger association between serum GGT level and all-cause death was observed in female patient. Especially in long-term mortality after 30th day, the association was observed only in females. In the aspect of cardiac death, the association was significant only in females, but the difference between 2 sex groups turned to be uncertain after multivariable adjustment.
ROC curve analyses were done to find out the cutoff value. It demonstrated that cutoff values of the serum GGT for predicting all-cause mortality could not be determined (AUC = 0.501, . Subgroup analysis. The association between serum GGT level and all-cause mortality differed among the specific groups according to age, sex, final diagnosis, LDL cholesterol, HDL cholesterol, body mass index, hypertriglyceride, ejection fraction and HbA1c. HRs were adjusted for age, sex, hypertension, diabetes, chronic kidney disease, current smoking, BMI ≥25 kg/m 2 , AST, ALT, ALP, uric acid, hsCRP, hypertriglyceridemia (triglyceride >150 mg/dL), high LDL cholesterol (LDL ≥100 mg/dL), Killip class ≥II at admssion, presence of no-reflow phenomenon during PCI, and TIMI flow <III after PCI. ALP = alkaline phosphatase, ALT = alanine aminotransferase, AMI = acute myocardial infarction, AST = aspartate aminotransferase, BMI = body mass index, GGT = gamma-glutamyl transferase, HDL = high-density lipoprotein, HR = hazard ratio, hsCRP = high-sensitivity C-reactive protein, LDL = low-density lipoprotein, PCI = percutaneous coronary intervention, TIMI = thrombolysis in myocardial infarction. The baseline characteristics of male and female AMI patients. In female group, we could find statistical significant value (AUC = 0.58, P = .003), but AUC showed very low discrimination accuracy. The cutoff value of GGT in female group was 58 U/L with 21.5% sensitivity and 91.6% specificity. It was very close to the minimum value of in female high GGT group in our study (57 U/L) [fig_ref] Figure 5: ROC curve analyses of serum GGT level and all-cause mortality in female... [/fig_ref].
# Discussion
In this study, we observed a significant association of serum GGT levels with all-cause mortality and identified serum GGT as a reliable, independent predictor of long-term mortality and cardiac mortality following AMI over a median follow-up period of 3.7 years. Interestingly, high GGT exhibited worse clinical outcomes in the female and in patients with STEMI and normal LVEF. These results are consistent with those of previous studies. In addition, this study is stronger than previous studies because it involved a larger AMI sample in which all subjects received PCI with drug eluting stent and had a longer follow-up period. These findings may be explained by changes in lipid metabolism in the setting of AMI, in association with an acute-phase reaction and inflammatory response. [bib_ref] Investigators of National Registry of Myocardial Infarction b. Relationship between serum low-density..., Reddy [/bib_ref] The patterns of alteration in the composition of lipid profile were wide in various previous reports, but a consensus about active inflammatory changes after AMI was developed. As previously explained, GGT is an enzyme associated with lipid metabolism and oxidative stress; therefore, the level of serum GGT at admission may reflect the degree of baseline change in lipid metabolism and inflammation after AMI, which may have implications for prognosis. In a small cohort study suggest the alteration of lipid metabolism might be result of secondary liver failure because of extensive and severe MI, [bib_ref] HDL cholestrol and the acute phase reaction following myocardial infacrtion and acute..., Fahie-Wilson [/bib_ref] but the impact of high GGT on mortality was not altered after adjustment of liver enzyme and lipid profiles. It suggests the independent role of GGT on inflammation process apart from the level of lipids and liver enzymes.
The differences in baseline characteristics between the 2 GGT groups observed in this study are also consistent with explanation. The inflammatory markers hsCRP [bib_ref] GGT levels in type II diabetic patients with acute coronary syndrome (does..., Emiroglu [/bib_ref] and uric acid [bib_ref] The prognostic role of gammaglutamyltransferase activity in non-diabetic ST-elevation myocardial infarction, Lazzeri [/bib_ref] were significantly higher in the high GGT group. Uric acid is considered a metabolic product of inflammation and is elevated in various inflammatory conditions such as CAD and heart failure. [bib_ref] Elevated serum uric acid in nondiabetic people mark pro-inflammatory state and HDL..., Onat [/bib_ref] [bib_ref] Serum gamma-glutamyltransferase and the risk of heart failure in men and women..., Wang [/bib_ref] In general, LDL-C is the most important determinants of atherosclerotic events in the lipid profile. [bib_ref] Investigators of National Registry of Myocardial Infarction b. Relationship between serum low-density..., Reddy [/bib_ref] It is interesting that there were some trends toward lower LDL-C levels in patients with high GGT. In our multivariable analysis, the high LDL-C tended to be preventive for mortality (≥100 mg/dL; HR = 0.81 P = .139) or LDL (as continuous variable, mg/dL; HR 1.0, P = .309), but it was not statistically significant. The use of statins could be a confounding factor, but unfortunately there was no . Univariate-and multivariate-adjusted time-to-death curves of high gamma-glutamyl transferase (GGT) on normal GGT group about all-cause mortality (early within 30 days and long term) and cardiac mortality in according to female and male. HRs were adjusted for age, hypertension, diabetes, chronic kidney disease, current smoking, body mass index ≥25 kg/m 2 , aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, uric acid, high-sensitivity Creactive protein, hypertriglyceridemia (triglyceride >150 mg/dL), high low-density lipoprotein (LDL) cholesterol (LDL ≥100 mg/dL), Killip class ≥ II at admssion, presence of no-reflow phenomenon during percutaneous coronary intervention (PCI) and thrombolysis in myocardial infarction flow <III after PCI. CI = confidence interval, HR = hazard ratio. information about either the usage of statin at the time of blood sampling or the level of LDL-C at the time of discharge in our registry. These might be one of the limitations of retrospective data from real world practice. Other lipid profiles including HDL-C and triglyceride have been previously reported in association with GGT, but the results were conflicting as our study. Subgroup analyses revealed that the association between serum GGT level and all-cause mortality were more significant in STEMI patients. This association may reflect difference in nature of coronary obstruction in STEMI and NSTEMI. STEMI is primarily attributed to acute plaque rupture owing to inflammation, suggesting a role for GGT. This discrepancy may also be because of the inclusion of an insufficient number of NSTEMI patients.
Adverse outcomes may also be attributed to the association between serum GGT level and metabolic syndrome. Obesity and hypertriglyceridemia were more frequently observed in the high GGT group. In this study, high GGT was more significant in patients who were younger than 65 years or who had normal left ventricular function. Such patients are usually expected to have a favorable prognosis but may be negligent in their long-term health care. In this regard, GGT levels may serve as a guide for these patients because they may receive the greatest benefit from being aware of metabolic disturbances. Although there was no statistical significance, the lower the HbA1c level of subgroups, the association tended to be stronger with GGT that could be an implication in the same vein.
To determine whether male and female patients with AMI differed with baseline risk factors and treatment, we repeated analysis of baseline characteristics after re-grouping. Age at the time of MI showed big difference as previous studies, [bib_ref] Sex differences in long-term mortality after myocardial infarction: a systematic review, Bucholz [/bib_ref] women were much older than men. So, the greater risk of mortality and morbidity towing to age may attenuate the other risk difference of females. But in our study, after adjustment for age the meaning of high GGT was not changed. Our study reported a higher percentage of hypertension, diabetes mellitus and higher killip classes in women, whereas current smoking, obesity, familial history of CAD, and presentation as STEMI tended to be more in men. The distribution of risk factors was different between them and these findings are consistent with previous reports, [bib_ref] Sex differences in long-term mortality after myocardial infarction: a systematic review, Bucholz [/bib_ref] then we can find out that we should assess differently and separately of each sex group. Interestingly, there was significant difference of discharge medication between 2 groups. In females, statins were less prescribed in spite of higher total and LDL-C, and angiotensin-converting enzyme-inhibitors, angiotensin type II receptor blockers, and beta-blockers were not prescribed either to more females those tended to present lower LVEF. These factors may explain the higher long-term mortality of women with AMI.
A previous STEMI study suggested that the differences in microvascular reperfusion after PCI between men and women may attribute to higher risks in female. [bib_ref] Gender differences in epicardial and tissurelevel reperfusion in patients undergoing primary angioplasty..., Pu [/bib_ref] In the study, TIMI myocardial perfusion grade and incomplete ST segment resolution were used as a parameter of microvascular dysfunction, and females had lower TIMI grade and more incomplete ST segment resolution. In our study, the percentage of lower TIMI flow grade after PCI and presence of no-reflow phenomenon during PCI were compared; no significant difference was observed between both sex groups. But presence of no-reflow was significantly different between high GGT and normal GGT group. Furthermore, no-reflow and TIMI flow after PCI were also important covariables with clinical significance in multivariable analysis of allcause mortality in association with GGT. This might be an implication of the association of GGT and myocardial microvascular dysfunction, but more concrete investigation will be needed.
We made 3 interesting observations concerning the association between elevated serum GGT levels and clinical outcome in patients with AMI. First, elevated serum GGT levels are a useful marker for easily and reliably predicting long-term clinical outcomes in patients with AMI. GGT levels are reportedly higher in patients with CAD than in the general population, and GGT levels are higher in patients with NSTEMI and STEMI than in patients with unstable angina. [bib_ref] GGT levels in type II diabetic patients with acute coronary syndrome (does..., Emiroglu [/bib_ref] In our AMI population, the proportion of patients with high GGT levels who did not have overt liver disease exceeded 11%, and the results were statistically significant after adjusting for liver markers such as AST, ALT, and ALP. Therefore, this method is not only acceptable but also likely easy to apply in clinical practice. Second, elevated GGT levels were more strongly associated with clinical outcome in female patients than in male patients. No previous studies have examined sex differences in the association between GGT level and outcomes in MI. [bib_ref] Sex differences in long-term mortality after myocardial infarction: a systematic review, Bucholz [/bib_ref] Studies of the association of GGT and other cardiovascular risk factors have observed heterogeneity in sex differences. In 2 studies of coronary calcification and hypertension in Korea, GGT level was an independent predictor in men but not women. [bib_ref] Association of coronary artery calcification and serum gamma-glutamyl transferase in Korean, Lee [/bib_ref] [bib_ref] Gender differences in the association between serum gamma-glutamyltransferase and blood pressure change:..., Ha [/bib_ref] However, one study of the association between GGT and vascular events observed a significant positive association between GGT levels and cardiovascular disease in women. [bib_ref] Gamma-glutamyltransferase is associated with incident vascular events independently of alcohol intake: analysis..., Fraser [/bib_ref] The sex differences observed in our study may be because of the differences in alcohol consumption between men and women. More Korean men drink alcohol than women. [bib_ref] Gender differences in the association between serum gamma-glutamyltransferase and blood pressure change:..., Ha [/bib_ref] As an alternative explanation, BMI differed significantly between the normal and high GGT groups and between men and women. There was a striking 34% to 39% reduction in mortality in patients with BMI ≥25 kg/m 2 at [fig_ref] Table 2: Hazard ratios and 95% confidence intervals for mortality according to the 2... [/fig_ref] and the patients with BMI ≥25 kg/m 2 were more in men than women. It is mostly understood that patients with overweight or mild obesity (BMI of 25-30 kg/m 2 ) might have relatively better outcomes, so this might affect the result consequently. Severe obesity (BMI >30 kg/m 2 ) is commonly considered as a predictor of worse outcomes, and the mortality of overt obese (BMI >30 kg/m 2 ; n = 97, 11.3%) group tended to be higher than that of overweight (BMI of 25-30 kg/m 2 ; n= 731, 9.7%) group in our AMI patients, but it was not statistically significant (HR = 1.21, Log rank, P = .557).
Third, our study population is unique in that we included both STEMI and NSTEMI patients and all patients underwent PCI with DES. Previous studies have included either STEMI patients or non-ST segment elevation-ACS patients. According to current guidelines on long-term management following ACS, the treatment strategy depends primarily on whether the clinical outcome is MI. [bib_ref] ESC Guidelines for the management of acute coronary syndromes in patients presenting..., Hamm [/bib_ref] [bib_ref] ACCF/AHA guideline for the management of ST-elevation myocardial infarction: executive summary: a..., O'gara [/bib_ref] Therefore, the overall outcome following AMI is also crucial. The all-cause mortality rates in our study population and in the high GGT level group were slightly higher and much higher, respectively, than those in the HORIZONS AMI trial in patients with STEMI (approximately 6%-7%). [bib_ref] Heparin plus a glycoprotein IIb/IIIa inhibitor versus bivalirudin monotherapy and paclitaxel-eluting stents..., Stone [/bib_ref] When the early mortality rate was assessed, it was similar to the result that obtained using the long-term outcome. This result suggests that the initial GGT level might reflect not only the acute phase of inflammation, but also the chronic systemic metabolic status of an individual, which can affect the long-term outcome.
# Limitations
This study has several limitations. First, this was an observational study and may be subject to bias and confounding. Specifically, information on alcohol consumption, which can influence the www.md-journal.com level of GGT, was not included in this study and 52% of patients did not have a GGT serum level recorded. Moreover, the exclusion of liver disease in patients was not confirmed by imaging or serology for viral hepatitis. [bib_ref] Gamma-glutamyltransferase is associated with incident vascular events independently of alcohol intake: analysis..., Fraser [/bib_ref] [bib_ref] Patterns of alcohol consumption and myocardial infarction risk: observations from 52 countries..., Leong [/bib_ref] Isolated GGT elevation has not been significantly associated with adverse outcomes in liver disease; therefore, the relevance of GGT elevation in MI should not to be attenuated. Second, there were a limited number of adverse events, and we were unable to specify the various subtypes of adverse cardiovascular events. Third, although many possible hypotheses of the mechanisms of GGT have existed, no clear evidence was proven. Even there is no report on the time course of GGT in the setting of AMI like AST, LDH, or CPK. Recently, various subtypes of GGTs with specific functions have been studied. [bib_ref] A high performance gel filtration chromatography method for gamma-glutamyltransferase fraction analysis, Franzini [/bib_ref] [bib_ref] Correlates and reference limits of plasma gamma-glutamyltransferase fractions from the Framingham Heart..., Franzini [/bib_ref] These studies might be a key to understand the mechanism of the association between GGT and cardiovascular disease. In addition, a large prospective study with serial GGT values and associated variables will contribute to a detailed understanding of the pathogenesis of cardiovascular disease.
# Conclusion
In summary, an elevated GGT level is an independent predictor of adverse long-term prognosis and increase of cardiac mortality in patients with AMI. Stronger associations were observed in the young and female patients and in patients who had STEMI and initially preserved LVEF after the indexed PCI.
[fig] Figure 1: Study flow chart. Inclusion and exclusion criteria of study population. ALT = alanine aminotransferase, AMI = acute myocardial infarction, AST = aspartate aminotransferase, GGT = gamma-glutamyl transferase. [/fig]
[fig] 060 ACE-: Is = angiotensin-converting enzyme inhibitors, ALP = alkaline phosphatase, ALT = alanine aminotransferase, ARBs = angiotensin type II receptor blockers, AST = aspartate aminotransferase, BMI = body mass index, CAD = coronary artery disease, eGFR = estimated glomerular filtration rate, GGT = gamma-glutamyl transferase, HDL = high-density lipoprotein, hsCRP = high-sensitivity C-reactive protein, LDL = low-density lipoprotein, LVEF = left ventricular ejection fraction, MDRD = Modification of Diet in Renal Disease equation, MI = myocardial infarction, STEMI = ST segment elevation myocardial infarction, TIMI = thrombolysis in myocardial infarction. [/fig]
[fig] Figure 2: Kaplan-Meier survival curves. All-cause mortality of patients with acute myocardial infarction according to normal or high serum GGT level. GGT = gamma-glutamyl transferase. [/fig]
[fig] P: ACE-Is = angiotensin-converting enzyme inhibitors, AL = alkaline phosphatase, ALT = alanine aminotransferase, ARBs = angiotensin type II receptor blockers, AST = aspartate aminotransferase, BMI = body mass index, CAD = coronary artery disease, eGFR = estimated glomerular filtration rate, GGT = gamma-glutamyl transferase, HDL = high-density lipoprotein, hsCR = high-sensitivity Creactive protein, LDL = low-density lipoprotein, LVEF = left ventricular ejection fraction, MDRD = Modification of Diet in Renal Disease equation, MI = myocardial infarction, STEMI = ST segment elevation myocardial infarction, TIMI = thrombolysis in myocardial infarction. = .94). [/fig]
[fig] Figure 4: Univariate-and multivariate-adjusted time-to-death curves of high gamma-glutamyl transferase (GGT) on normal GGT group about all-cause mortality (early within 30 days and long term) and cardiac mortality in according to female and male. HRs were adjusted for age, hypertension, diabetes, chronic kidney disease, current smoking, body mass index ≥25 kg/m 2 , aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, uric acid, high-sensitivity Creactive protein, hypertriglyceridemia (triglyceride >150 mg/dL), high low-density lipoprotein (LDL) cholesterol (LDL ≥100 mg/dL), Killip class ≥ II at admssion, presence of no-reflow phenomenon during percutaneous coronary intervention (PCI) and thrombolysis in myocardial infarction flow <III after PCI. CI = confidence interval, HR = hazard ratio. [/fig]
[fig] Figure 5: ROC curve analyses of serum GGT level and all-cause mortality in female acute myocardial infarction patients. AUC = area under the curve, GGT = gamma-glutamyl transferase, ROC = receiver-operating characteristic. [/fig]
[table] Table 2: Hazard ratios and 95% confidence intervals for mortality according to the 2 groups of serum GGT. BMI = body mass index, CI = confidence interval, GGT = gamma-glutamyl transferase, hsCRP = high-sensitivity C-reactive protein, PCI = percutaneous coronary untervention, TIMI = thrombolysis in Myocardial Infarction. The parameters which showed statistical significance after multivariate analysis are demonstrated. * Model 1 was adjusted for age and sex; unadjusted hazard ratios of these variables are included for reference. † Model 2 was adjusted for model 1 plus BMI ≥25 kg/m 2 , AST, ALT, ALP, uric acid, hsCRP, hypertriglyceridemia (triglyceride>150 mg/dL) and presence of no-reflow phenomenon during PCI. ‡ Model 3 was adjusted for model 2 plus hypertension, diabetes, chronic kidney disease, current smoking, LDL ≥100 mg/dL, Killip class ≥II at admssion and TIMI flow <III after PCI. [/table]
|
Giant cell granuloma of the maxilla. Global management, review of literature and case report
Giant cell granuloma is a relatively rare benign entity but can be locally aggressive. Histologically characterized by intense proliferation of multinucleated giant cells and fibroblasts. Affects bone supported tissues. Definitive diagnosis is given by biopsy. Clinically manifest as a mass or nodule of reddish color and fleshy, occasionally ulcerated surface. They can range from asymptomatic to destructive lesions that grow quickly. It is a lesion to be considered in the differential diagnosis of osteolytic lesions affecting the maxilla or jaw. Its management passed from conservative treatment with intralesional infiltration of corticosteroids, calcitonin or interferon, to the surgical resection and reconstruction, for example with microvascular free flaps.
# Introduction
Giant cell granuloma (GCGs) was first described by Jaffe in 1953 (1) as "giant cell reparative granuloma", but currently does not refer to him as reparative, because of its locally destructive. It is classified as peripheral if it affects the extremities and central if it develops in the midline (being the least common type). It is a relatively rare entity. Accounted for 7% of the maxillary tumors (his preferred location it´s the incisor region, and more frequently in the jaw than the maxilla) (2). It is more common in children and young adults, with a slight predominance in females (2). As etiological factors (3) have been related several factors, especially local irritants (such as extractions or poorly fitting dentures) and hormonal (4) (in fact, when we diagnose a GCGs, we should discard the coexistence of primary hyperparathyroidism, because classic brown tumors features of this disease are virtually indistinguishable from the histology of GCGs). Another theory relates to its origin it is an intraosseous vascular lesions similar to angiomas of soft tissues (5). In any case usually affects bone supported tissues.
## Case report
Female patient 16 years old with no history of interest who presents a lesion of several months, in the gingiva of the second quadrant, whose size, according to the patient, has not increased in recent weeks [fig_ref] Figure 1: a Initial clinical view of the lesion [/fig_ref]. It is clinically asymptomatic. On physical examination, the lesion is reddish, soft and fleshy. It causes a significant and bulging prominence, both lobby and palatal and mobility of the pieces 22 and 23. The analytical requested in its health center is completely normal. Orthopantomography shows the bone defect that coincides with lesion, pieces 18, 28, 38 and 48 included and root fragments of pieces 16 and 36, and periapical lesions in relation to pieces 36 and 37. When patient is attended in consultation, a facial, axial and coronal computed tomography (CT)-scan and 3D reconstructions are requested [fig_ref] Figure 1: a Initial clinical view of the lesion [/fig_ref] , which shows the defect displayed on the Orthopantomography. In Hematoxylin-eosin staining appears intense proliferation of fibroblasts and multinucleated giant cells, and it is reported as "giant cell granuloma." Based on the age of the patient, in the absence of clinical and benign nature of the lesion, we opted for conservative treatment by six cycles of intralesional injection of triamcinolone, with the further implementation of regular radiological controls. However, due to the persistence and escalation of the lesion [fig_ref] Figure 2: Appearance of the lesion after conservative treatment with intralesional triamcinolone [/fig_ref] , we completed treatment with resection of the granuloma and reconstruction of the defect with microvascular fibula free flap with skin paddle associated by anastomosis of the peroneal vessels to the facial vessels [fig_ref] Figure 3: a Adaptation of microvascular fibula free flap skin paddle associated to the... [/fig_ref]. In a second procedure, two months after reconstruction, the flap was defatted [fig_ref] Figure 3: a Adaptation of microvascular fibula free flap skin paddle associated to the... [/fig_ref].
# Discussion
GCGs clinically manifests as a mass or nodule of reddish color (although it can sometimes be bluish) and occasionally ulcerated fleshy surface. Its range can be from asymptomatic, small and slow-growing to large and destructive lesions that grow rapidly. Imaging testing are essential, because show the true extent of GCGs and their behavior in the tissue in which it sits. Although, as a first approximation we can make use of the Ortopantomography, it is often necessary to perform CT, sometimes even three-dimensional reconstructions. Definitive diagnosis is determined by biopsy. Thus, histologically characterized (4) by the intense proliferation of multinucleated giant cells and fibroblasts, a dense vascular stroma with hemosiderin deposits. The differential diagnosis is made with pyogenic granuloma, gingival fibroma, fibrosarcoma and distant metastases of tumors, because they can resemble clinically GCGs.
As for handling, typically resorted to surgery, from simple curettage to resection of the lesion. After the surgical treatment has a high rate of recurrence (between 13-49%) (6). Currently calls for a conservative therapeutic, that prevents or diminishes the aesthetic and functional sequelae caused by standard treatments. So basically include three lines of conservative treatment. The most used is the intralesional injection of corticosteroids, with different guidelines and protocols, of which the most widespread in the literature is the weekly injection of triamcinolone (7) associated with local anesthetic for a period six weeks. Other conservative options are the use of calcitonin (8) (usually in the form of nasal spray) or interferon-α (INF-α) (4,8), both in pattern of several months, depending on the success in reducing the granuloma. Occasionally, due to the high recurrence rate or incomplete disappearance of GCGs with conservative treatment is necessary to supplement the treatment with surgical removal of the same, and if possible, reconstruction of the defect created, for example with buccal fat pad flap (9) or microvascular free flaps, such as the fibula. Whatever the method chosen, it is essential to clinical and radiological control of the patient, through revisions and imaging testing (mainly CT). We can conclude that GCGs is a relatively rare entity of benign histologic nature but can have a very aggressive local behavior. A definitive diagnosis is reached by biop-sy of the lesion. It should be present in the differential diagnosis in all expansive osteolytic lesions affecting the jaw or maxilla, such as pyogenic granuloma, gingival fibroma, fibrosarcoma and distant metastases of tumors. Its treatment includes from removal of it to the simple intralesional injection of corticosteroids or the use of calcitonin or INF-α, always with clinical and radiological control due to its high rate of recurrence. Given the failure of conservative treatment, management be should an aggressive resection and reconstruction, when it is possible. An adequate solution is microvascular fibula free flap.
[fig] Figure 1: a Initial clinical view of the lesion. b View of the defect created by the lesion in 3D reconstruction CT-scan. [/fig]
[fig] Figure 2: Appearance of the lesion after conservative treatment with intralesional triamcinolone. [/fig]
[fig] Figure 3: a Adaptation of microvascular fibula free flap skin paddle associated to the resultant defect after excision. b Appearance of the flap once thinned two months after the reconstructive surgery. [/fig]
|
The association between sedentary behaviour and risk of anxiety: a systematic review
Background: Previous research has linked sedentary behaviour (SB) to adverse physical health outcomes in adults and youth. Although evidence for the relationship between SB and mental health outcomes (e.g., depression) is emerging, little is known regarding risk of anxiety.Methods: A systematic search for original research investigating the association between SB and risk of anxiety was performed using numerous electronic databases. A total of nine observational studies (seven cross-sectional and two longitudinal) were identified. Methodological quality of studies was assessed and a best-evidence synthesis was conducted. Results: One cross-sectional study demonstrated a strong methodological quality, five cross-sectional studies demonstrated a moderate methodological quality and three studies (two cross-sectional one longitudinal) received a weak methodological quality rating. Overall, there was moderate evidence for a positive relationship between total SB and anxiety risk as well as for a positive relationship between sitting time and anxiety risk. There was inconsistent evidence for the relationship between screen time, television viewing time, computer use, and anxiety risk. Conclusion: Limited evidence is available on the association between SB and risk of anxiety. However, our findings suggest a positive association (i.e. anxiety risk increases as SB time increases) may exist (particularly between sitting time and risk of anxiety). Further high-quality longitudinal/interventional research is needed to confirm findings and determine the direction of these relationships.
# Background
Sedentary behaviour (i.e. activities which require minimal body movement resulting in low energy expenditure similar to resting level [1.0 to 1.5 metabolic equivalent (METs)] [bib_ref] Age-related change in physical activity in adolescent girls, Pate [/bib_ref] includes behaviours such as sitting for a range of purposes (e.g. work, travel), and screen-based activities such as computer use, electronic gaming, and television viewing. Time spent in sedentary behaviour has emerged as a potentially important indicator of health in adult populations [bib_ref] Physical activity and sedentary behavior: a review of longitudinal studies of weight..., Must [/bib_ref] , independent of achieving sufficient physical activity. Among the general adult population some evidence suggests sedentary behaviour is associated with increased risk of developing various chronic diseases (e.g. overweight and obesity [bib_ref] Television watching and other sedentary behaviors in relation to risk of obesity..., Hu [/bib_ref] , cardiovascular disease [bib_ref] Sitting time and mortality from all causes, cardiovascular disease, and cancer, Katzmarzyk [/bib_ref] , osteoporosis [bib_ref] Health benefits of physical activity: the evidence, Warburton [/bib_ref] , type two diabetes [bib_ref] Television watching and other sedentary behaviors in relation to risk of obesity..., Hu [/bib_ref] , and various cancers [bib_ref] Reducing sedentary behavior in minority girls via a theory-based, tailored classroom media..., Spruijt-Metz [/bib_ref]. However, reviews of prospective studies suggest that there is yet insufficient or no evidence to conclude relationships between sedentary behaviour and certain health outcomes (e.g. adulthood weight gain, cardiovascular disease risk, and some cancers) [bib_ref] Sedentary behaviors and health outcomes among adults: a systematic review of prospective..., Proper [/bib_ref] [bib_ref] Sedentary behaviors and subsequent health outcomes in adults a systematic review of..., Thorp [/bib_ref]. A range of health consequences of sedentary behaviour in children and adolescents have also been identified (e.g. unfavourable body composition, decreased fitness, lowered scores for self-esteem, and pro-social behaviour and decreased academic achievement [bib_ref] Systematic review of sedentary behaviour and health indicators in school-aged children and..., Tremblay [/bib_ref] , sleep problems, musculoskeletal pain, depression, and poor psychological well-being [bib_ref] The health indicators associated with screen-based sedentary behavior among adolescent girls: a..., Costigan [/bib_ref]. However, like that of the adult literature, reviews of prospective studies in young people suggest that there still remains insufficient evidence to conclude associations between sedentary behaviour and some health indicators (e.g. body weight, blood pressure, bone mass).
The association between sedentary behaviour and mental health issues such as depression [bib_ref] Sedentary behavior and depression among adults: a review, Teychenne [/bib_ref] [bib_ref] Physical activity, sedentary behavior and depression among disadvantaged women, Teychenne [/bib_ref] and self-esteem [bib_ref] Associations between sedentary behavior and self-esteem in adolescent girls from schools in..., Nihill [/bib_ref] have been explored within some population groups (e.g., women, adolescents), and on the balance it has been suggested that a positive relationship exists between most sedentary behaviours and depression [bib_ref] Sedentary behavior and depression among adults: a review, Teychenne [/bib_ref] and self-esteem [bib_ref] Systematic review of sedentary behaviour and health indicators in school-aged children and..., Tremblay [/bib_ref]. However little is known regarding the relationship between sedentary behaviour and other mental illnesses such as anxiety.
Anxiety is a mental illness that affects approximately 14 % of Australians adultsand 15 % of [bib_ref] EBMH notebook: generalised anxiety disorder, Gale [/bib_ref] [bib_ref] Anxiety disorders and comorbid medical illness, Roy-Byrne [/bib_ref] [bib_ref] Metabolic and physiologic responses to video game play in 7-to 10-year-old boys, Wang [/bib_ref] [bib_ref] Impact of singular excessive computer game and television exposure on sleep patterns..., Dworak [/bib_ref] [bib_ref] Higher levels of psychological distress are associated with a higher risk of..., Mommersteeg [/bib_ref] [bib_ref] Physical activity and mental health in children and adolescents: a review of..., Biddle [/bib_ref] year olds, with global estimates indicating anxiety to affect over 27 million people. Anxiety is characterised by excessive and persistent (yet often unrealistic) worry which can inhibit one's ability to carry out activities of daily living [bib_ref] EBMH notebook: generalised anxiety disorder, Gale [/bib_ref]. Physiological symptoms can include: pounding heart; difficulty breathing; upset stomach; muscle tension; sweating; and, feeling faint, or shaky [bib_ref] EBMH notebook: generalised anxiety disorder, Gale [/bib_ref]. The illness has been shown to be linked to other serious diseases such as increased risk of cardiovascular disease and cancer [bib_ref] Anxiety disorders and comorbid medical illness, Roy-Byrne [/bib_ref] with estimates that anxiety and depression together contribute to 10 % of the total burden of disease in Australian women. Considering anxiety has such a large impact on society (e.g. healthcare costs, quality of life, and life expectancy) and is highly prevalent across the life span, it is important to understand the behavioural factors that may be linked to it.
It has been hypothesised that sedentary behaviour may lead to anxiety through biological pathways. For example, engaging in screen-based entertainment, such as video gaming, has been shown to increase the arousal of the central nervous system (CNS) [bib_ref] Metabolic and physiologic responses to video game play in 7-to 10-year-old boys, Wang [/bib_ref] , which could potentially lead to increased levels of anxiety. Additionally, screen-based sedentary behaviours have been linked to disrupted sleeping patterns which may also cause elevated levels of anxiety [bib_ref] Impact of singular excessive computer game and television exposure on sleep patterns..., Dworak [/bib_ref]. Given the plausibility of these short-term effects of sedentary behaviours on mood (i.e. distraction, CNS arousal, sleep), it is likely that the cumulative effects of these behaviours may further result in longer-term impacts on anxiety risk. Alternatively, the link between sedentary behaviour and anxiety could be explained by poor metabolic health. For example, research has shown sedentary behaviour to be linked to an increased risk of type 2 diabetes [bib_ref] Television watching and other sedentary behaviors in relation to risk of obesity..., Hu [/bib_ref] , a disease which is subsequently associated with poor mental health [bib_ref] Higher levels of psychological distress are associated with a higher risk of..., Mommersteeg [/bib_ref]. On the other hand, it could be hypothesised that those with mental illness (e.g. anxiety) tend to engage in sedentary behaviour more than those without the illness.
Although research has indicated an inverse association between physical activity and anxiety may exist [bib_ref] Physical activity and mental health in children and adolescents: a review of..., Biddle [/bib_ref] [bib_ref] Anxiety outcomes after physical activity interventions: meta-analysis findings, Conn [/bib_ref] [bib_ref] Relationship between physical activity and depression and anxiety symptoms: a population study, Mello [/bib_ref] , currently the evidence regarding the association between sedentary behaviour and anxiety is unclear. Thus, the aim of this systematic review is to investigate the association between sedentary behaviour and risk of anxiety across the lifespan.
# Methods
## Search strategy
A structured electronic search (employing PRISMA reporting guidelines) of publication years from 1990 (through November 2014) was conducted, since the 1990's saw an increase in sedentary behaviour levels of the population with the widespread use of online technology. Databases included: CINAHL complete, Medline/Medline complete, PsychARTICLES, Psychology, and Behavioral Sciences Collection, PsychINFO, and SPORTDiscus. The following search strings were used: (anxiety OR anxious OR mental health OR mental illness) AND (sedentary behav* OR sitting OR TV OR television OR computer OR screen). These strings were further limited to peerreviewed publications written in English. First, title, and abstracts of articles identified in the search process were assessed for suitability. Second, full-text articles were retrieved, and assessed for inclusion. Third, reference lists from retrieved full-text articles were searched. Additional records were identified through other sources (i.e. Authors own bibliographic library).
## Study selection criteria
For the purpose of this review, risk of anxiety was defined as either the likelihood of developing or experiencing an anxiety disorder or non-clinical anxiety symptoms. Studies were considered eligible if they: (1) examined 'healthy' children, adolescents, or adults (i.e. those who were not patients suffering from underlying chronic physical conditions that may confound results); (2) examined risk of anxiety specifically; (3) assessed screen-based sedentary behaviour or sitting time; (4) and involved a cross-sectional, longitudinal, or experimental study design. However, only intervention studies which primarily show the relation between sitting and anxiety (i.e. not just the effect of an intervention on anxiety) were eligible to be included. Conference abstracts, dissertations, theses, and articles published in nonpeer-reviewed journals were not included for review.
## Data extraction
Key study characteristics of the identified studies were extracted including: the country of origin, size/source of study population, study design, domain (e.g. leisure time sitting, occupational sitting, total sitting), measures used, indicator of sedentary behaviour (e.g. computer use, television viewing, screen time, sitting), and study results in terms of the association between sedentary behaviour and risk of anxiety.
# Methodological quality
A modified version of an eight-component rating scalewas used to determine methodological quality of each study. The tool assesses eight methodological components of research studies including: selection bias (e.g. response rate, representativeness), study design (e.g. cohort, RCT, etc.), confounders (e.g. controlling for confounders such as age, socio-economic position etc.), blinding (e.g. awareness of group allocation, etc.), data collection methods (e.g. valid, reliable), withdrawals, and dropouts (e.g. percent providing full data), intervention integrity (e.g. percent receiving intervention) and analyses (e.g. appropriateness of study design). In regards to the observational studies that were assessed, the tool was modified so that those studies were not scored on 1) the blinding component or 2) other interventionspecific criteria within any other components. Thus, since only observational studies were included in this review, a maximum number of six components were scored. Each of the components was given an overall quality score of weak, moderate, or strong. If a component was not described, authors of those studies were attempted to be contacted to provide this information. In the instance that authors of those studies did not respond, the (undescribed) component was given a weak rating. Once each component was rated, an overall study rating of weak (if ≥2 of the components were scored weak), moderate (if <3 components were scored strong with no more than one weak score), or strong (if ≥3 components were scored strong) was given to each study. Two reviewers (SAC and KP) independently assessed the methodological quality of studies meeting the inclusion criteria. Scoring discrepancies were resolved via consensus and inter-rater reliability was calculated using percentage agreement.
## Best-evidence synthesis
In order to draw conclusions on the association between sedentary behaviour and anxiety risk with regards to the methodological quality of studies, a best-evidence synthesis was conducted, which was based on previous systematic reviews in the area of sedentary behaviour and health outcomes [bib_ref] Sedentary behaviors and health outcomes among adults: a systematic review of prospective..., Proper [/bib_ref]. As such, three levels of evidence considered:
# Results
Literature searching yielded 983 studies (see [fig_ref] Figure 1: Flow of study selection through the phases of the review [/fig_ref]. A total of 71 duplicates were removed and thus 912 studies were screened by title. After further screening of abstracts (n = 177) and full papers (n = 42) a total of nine studies were included in the review (see [fig_ref] 1: Strong evidence [/fig_ref]. Most studies employed a cross-sectional study design (7/9) and two a longitudinal design. Samples sizes ranged from 189 [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] to 13,470 [bib_ref] Associations between sport and screen-entertainment with mental health problems in 5-year-old children, Griffiths [/bib_ref]. Descriptive characteristics of the nine studies are outlined in [fig_ref] 1: Strong evidence [/fig_ref]. A total of two studies included children/adolescents [bib_ref] Associations between sport and screen-entertainment with mental health problems in 5-year-old children, Griffiths [/bib_ref] [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] , seven included adults [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] [bib_ref] Cross-sectional associations between sitting at work and psychological distress: Reducing sitting time..., Kilpatrick [/bib_ref] [bib_ref] Physical activity, sedentary index, and mental disorders in the SUN Cohort Study, Sanchez-Villegas [/bib_ref] [bib_ref] Associations of sedentary behavior and physical activity with psychological distress: a cross-sectional..., Sloan [/bib_ref] [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] and one longitudinal study examined adolescents with follow-up conducted during adulthood [bib_ref] Adolescent predictors of objectively measured physical activity and sedentary behaviour at age..., Uijtdewilligen [/bib_ref]. Most studies (7/9) measured sedentary behaviour using self-report methods such as the Global Physical Activity Questionnaire [bib_ref] Associations of sedentary behavior and physical activity with psychological distress: a cross-sectional..., Sloan [/bib_ref] or a modified version of the International Physical Activity Questionnaire [bib_ref] Cross-sectional associations between sitting at work and psychological distress: Reducing sitting time..., Kilpatrick [/bib_ref]. Of those studies that utilised self-report measures, two measured screen-based entertainment (i.e. television/computer use/electronic game use) [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] [bib_ref] Physical activity, sedentary index, and mental disorders in the SUN Cohort Study, Sanchez-Villegas [/bib_ref] , two examined sitting (either occupational sitting [bib_ref] Cross-sectional associations between sitting at work and psychological distress: Reducing sitting time..., Kilpatrick [/bib_ref] or overall sitting [bib_ref] Associations of sedentary behavior and physical activity with psychological distress: a cross-sectional..., Sloan [/bib_ref] , one examined television viewing [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] , one compared television viewing, computer use, and overall screen time [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] , whilst one study compared six different forms of sedentary behaviour (total sitting, sitting at the computer, sitting watching television, transport-related sitting, work-related sitting, leisure-time sitting) [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref].
Only one study utilised objective measures of sedentary behaviour (i.e. overall sitting time) using accelerometers [bib_ref] Adolescent predictors of objectively measured physical activity and sedentary behaviour at age..., Uijtdewilligen [/bib_ref] , and one study used parent (proxy) report of screen-based entertainment [bib_ref] Associations between sport and screen-entertainment with mental health problems in 5-year-old children, Griffiths [/bib_ref]. Risk of anxiety (i.e. anxiety symptoms) was measured using various selfreport measures including the Screen for Child Anxiety Related Emotional Disorders (SCARED) [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] , the Behavioral Risk Factor Surveillance Survey [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] , the Kessler Psychological Distress Scale (K10) [bib_ref] Cross-sectional associations between sitting at work and psychological distress: Reducing sitting time..., Kilpatrick [/bib_ref] , the General Health Questionnaire (GHQ-12) [bib_ref] Associations of sedentary behavior and physical activity with psychological distress: a cross-sectional..., Sloan [/bib_ref] , the depression, anxiety and stress scale [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] and the Achievement Motivation Test (AMT) [bib_ref] Adolescent predictors of objectively measured physical activity and sedentary behaviour at age..., Uijtdewilligen [/bib_ref]. Interview methods were used in one study which utilised the Composite International Diagnostic Interview (CIDI) to determine presence of anxiety disorder [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] , whilst a parent (proxy) report measure of the Strengths and Difficulties Questionnaire (SDI) (which specifically examined the Emotional Symptoms subscale , an indicator of anxiety) was utilised in one other study [bib_ref] Associations between sport and screen-entertainment with mental health problems in 5-year-old children, Griffiths [/bib_ref] [fig_ref] 1: Strong evidence [/fig_ref].
# Methodological quality
Methodological quality scores are provided in [fig_ref] 1: Strong evidence [/fig_ref]. Initial agreement between reviewers was 90 % (K = 0.82) on the items. Overall, one cross-sectional study [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] demonstrated a strong methodological quality, five cross-sectional studies [bib_ref] Associations between sport and screen-entertainment with mental health problems in 5-year-old children, Griffiths [/bib_ref] [bib_ref] Cross-sectional associations between sitting at work and psychological distress: Reducing sitting time..., Kilpatrick [/bib_ref] [bib_ref] Physical activity, sedentary index, and mental disorders in the SUN Cohort Study, Sanchez-Villegas [/bib_ref] [bib_ref] Associations of sedentary behavior and physical activity with psychological distress: a cross-sectional..., Sloan [/bib_ref] [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] demonstrated a moderate methodological quality and three studies (two cross-sectional [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] , one longitudinal [bib_ref] Adolescent predictors of objectively measured physical activity and sedentary behaviour at age..., Uijtdewilligen [/bib_ref] received a weak methodological quality rating. All of the studies were missing essential information regarding the methodological quality. For example, only three studies included reliable and valid measures for both sedentary behaviour and anxiety [bib_ref] Cross-sectional associations between sitting at work and psychological distress: Reducing sitting time..., Kilpatrick [/bib_ref] [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] [bib_ref] Adolescent predictors of objectively measured physical activity and sedentary behaviour at age..., Uijtdewilligen [/bib_ref] , and only two studies demonstrated a high retention rate (i.e. 80-100 %) [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] [bib_ref] Physical activity, sedentary index, and mental disorders in the SUN Cohort Study, Sanchez-Villegas [/bib_ref]. Further detail of the scoring of methodological quality of each study is provided in (see Additional file 1: .
## Main findings
Of the 9 studies included in this review, five (four crosssectional [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] [bib_ref] Cross-sectional associations between sitting at work and psychological distress: Reducing sitting time..., Kilpatrick [/bib_ref] [bib_ref] Associations of sedentary behavior and physical activity with psychological distress: a cross-sectional..., Sloan [/bib_ref] , one longitudinal [bib_ref] Adolescent predictors of objectively measured physical activity and sedentary behaviour at age..., Uijtdewilligen [/bib_ref] found a positive association between sedentary behaviour and risk of anxiety (i.e. increased sedentary behaviour was linked to increased risk of anxiety). One longitudinal study [bib_ref] Physical activity, sedentary index, and mental disorders in the SUN Cohort Study, Sanchez-Villegas [/bib_ref] found no association between sedentary behaviour and risk of anxiety. One cross-sectional study found both inverse and null associations between sedentary behaviour and risk of anxiety [bib_ref] Associations between sport and screen-entertainment with mental health problems in 5-year-old children, Griffiths [/bib_ref] (dependent on the target group examined), whilst two cross-sectional studies [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] found both positive associations and null associations (dependent on the sedentary behaviour examined). Since one strong- [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] , three moderate- [bib_ref] Cross-sectional associations between sitting at work and psychological distress: Reducing sitting time..., Kilpatrick [/bib_ref] [bib_ref] Associations of sedentary behavior and physical activity with psychological distress: a cross-sectional..., Sloan [/bib_ref] [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] and three weak-quality studies [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] [bib_ref] Adolescent predictors of objectively measured physical activity and sedentary behaviour at age..., Uijtdewilligen [/bib_ref] demonstrated at least one positive relationship between sedentary behaviour and anxiety risk, based on the bestevidence synthesis, there was moderate evidence for the overall relationship between sedentary behaviour and anxiety risk.
## Sitting time
A total of five studies (four cross-sectional, one longitudinal) examined the association between overall sitting time and risk of anxiety [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] [bib_ref] Cross-sectional associations between sitting at work and psychological distress: Reducing sitting time..., Kilpatrick [/bib_ref] [bib_ref] Associations of sedentary behavior and physical activity with psychological distress: a cross-sectional..., Sloan [/bib_ref] [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] [bib_ref] Adolescent predictors of objectively measured physical activity and sedentary behaviour at age..., Uijtdewilligen [/bib_ref] in adults, with all but one [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] finding a positive association between sitting time and anxiety risk. Although the longitudinal study by Uijtdewilligen and colleagues [bib_ref] Adolescent predictors of objectively measured physical activity and sedentary behaviour at age..., Uijtdewilligen [/bib_ref] demonstrated this positive association, the direction of this relationship indicated that anxiety symptoms during adolescence were predictive of sitting time in adulthood. Furthermore, Rebar et al. found that although transportrelated and overall sitting time were associated with higher risk of anxiety, sitting time undertaken for work purposes or during leisure time were not associated with anxiety risk [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref]. Based on the consistent findings of the three moderate- [bib_ref] Cross-sectional associations between sitting at work and psychological distress: Reducing sitting time..., Kilpatrick [/bib_ref] [bib_ref] Associations of sedentary behavior and physical activity with psychological distress: a cross-sectional..., Sloan [/bib_ref] [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] and one weak-quality [bib_ref] Adolescent predictors of objectively measured physical activity and sedentary behaviour at age..., Uijtdewilligen [/bib_ref] studies, the best-evidence synthesis demonstrated there was moderate evidence for the positive relationship between sitting time and anxiety risk.
Screen-time (i.e. combined TV, computer and/or electronic games)
A total of four studies (three cross-sectional, one longitudinal) examined the relationship between screen-time and risk of anxiety [bib_ref] Associations between sport and screen-entertainment with mental health problems in 5-year-old children, Griffiths [/bib_ref] [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] [bib_ref] Physical activity, sedentary index, and mental disorders in the SUN Cohort Study, Sanchez-Villegas [/bib_ref]. Of those, two crosssectional studies (one in a sample of adolescence and one in adults) showed a positive association between screen-time (i.e. combined TV and computer use) and anxiety risk [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref]. More specifically, Cao and colleagues [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] showed that high school students who spent more than 2-hours a day engaged in screen-based behaviours were 36 % more likely to experience anxiety symptoms than those who engaged in less than 2-hours a day.
In contrast, one cross-sectional study [bib_ref] Associations between sport and screen-entertainment with mental health problems in 5-year-old children, Griffiths [/bib_ref] showed an inverse association between screen-time (i.e. combined TV, computer, and electronic games use) and anxiety risk amongst 5-year old girls (results were not significant for boys). Those findings suggested that girls who spent less than 2-hours in screen-based entertainment were more likely to suffer symptoms related to anxiety. However, Sanchez-Villegas et al. [bib_ref] Physical activity, sedentary index, and mental disorders in the SUN Cohort Study, Sanchez-Villegas [/bib_ref] found no association between self-reported TV/computer use time and anxiety (defined as previous or current diagnoses of anxiety from a health professional). Based on the inconsistent findings of the one strong- [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] , two moderate- [bib_ref] Associations between sport and screen-entertainment with mental health problems in 5-year-old children, Griffiths [/bib_ref] [bib_ref] Physical activity, sedentary index, and mental disorders in the SUN Cohort Study, Sanchez-Villegas [/bib_ref] and one weak-quality [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] studies, there was insufficient evidence for the relationship between screen time and anxiety risk.
## Television viewing
Three cross-sectional studies amongst adults examined the relationship between television viewing and risk of anxiety [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref]. Two of those studies showed positive associations [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] , suggesting that television viewing was associated with an increased likelihood of anxiety symptoms. However, based on the inconsistent findings of the one moderate- [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] and two weak-quality [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] studies, there was insufficient evidence for the relationship between television viewing and anxiety risk.
## Computer use
Two cross-sectional studies investigated the association between computer use and risk of anxiety in adults [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref]. Although the moderate quality study by showed a positive association between sitting at the computer and anxiety symptoms [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] , the weak quality study by de Wit et al. showed no significant associations between the two factors [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref]. However, that study included only a measure of computer use during leisuretime and thus it is not clear as to whether these results would remain similar for computer use undertaken for work purposes. Based on these inconsistent findings there was insufficient evidence for the relationship between computer use and anxiety risk.
# Discussion
This is the first review to examine evidence regarding the association between sedentary behaviour and risk of anxiety. It is important to better understand this relationship as this information may help to inform the development of lifestyle change strategies for reducing the risk of anxiety in different population groups. It is clear from this review that the current body of evidence exploring the relationship between sedentary behaviour and risk of anxiety is limited, with only nine studies currently been published. On the balance, however, most studies (78 %) found at least one positive association between sedentary behaviour and anxiety risk [bib_ref] Correlates of television viewing among african american and caucasian women, Granner [/bib_ref] [bib_ref] Screen time, physical activity and mental health among urban adolescents in China, Cao [/bib_ref] [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] [bib_ref] Cross-sectional associations between sitting at work and psychological distress: Reducing sitting time..., Kilpatrick [/bib_ref] [bib_ref] Associations of sedentary behavior and physical activity with psychological distress: a cross-sectional..., Sloan [/bib_ref] [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] [bib_ref] Adolescent predictors of objectively measured physical activity and sedentary behaviour at age..., Uijtdewilligen [/bib_ref]. In other words, there is moderate evidence suggesting that engaging in overall sedentary behaviour was linked to an increased risk of anxiety. These findings are similar to those found in previous reviews that have assessed the relationship between sedentary behaviour and other specific mental health outcomes such as depression [bib_ref] Sedentary behavior and depression among adults: a review, Teychenne [/bib_ref]. However, in our review, based on the bestevidence synthesis and when considering the different types of sedentary behaviour separately, moderate evidence was found for the positive relationship between sitting time and anxiety risk, whilst inconsistent evidence was found for the relationship between screen time, television viewing time, computer use, and anxiety risk. There is currently limited insight into the underlying mechanisms that may explain this positive relationship between sedentary behaviour and anxiety risk. As already discussed, plausible biological pathways may include central nervous system arousal [bib_ref] Metabolic and physiologic responses to video game play in 7-to 10-year-old boys, Wang [/bib_ref] , sleep disturbances [bib_ref] Impact of singular excessive computer game and television exposure on sleep patterns..., Dworak [/bib_ref] or poor metabolic health [bib_ref] Higher levels of psychological distress are associated with a higher risk of..., Mommersteeg [/bib_ref] resulting from engagement in sedentary behaviour. Furthermore, drawing from previously suggested hypotheses that have been used to explain the link between sedentary behaviour and other mental disorders (i.e. depression), it could be the displacement of physical activity when engaging in sedentary behaviour that explains the relationship with increased anxiety risk, since physical activity has been shown to be beneficial in reducing anxiety in both children/adolescents [bib_ref] Exercise in prevention and treatment of anxiety and depression among children and..., Larun [/bib_ref] and adults [bib_ref] Physical activity, exercise, depression and anxiety disorders, Strohle [/bib_ref]. Alternatively, the link may be explained by a social withdrawal theory which posits that engaging in prolonged sedentary behaviours, such as television viewing, may lead to social solitude and withdrawing from interpersonal relationships which has been linked to increased feelings of social anxiety [bib_ref] The developmental psychopathology of anxiet, Rubin [/bib_ref]. On the other hand, it may be that those suffering anxiety symptoms are more inclined to engage in sedentary behaviours as a means of coping with anxiety, as has been suggested in previous research amongst adolescence with social physique anxiety [bib_ref] Social physique anxiety in adolescence: an exploration of influences, coping strategies, and..., Sabiston [/bib_ref].
In contrast, one cross-sectional study included in our review [bib_ref] Associations between sport and screen-entertainment with mental health problems in 5-year-old children, Griffiths [/bib_ref] showed that sedentary behaviour was inversely associated with risk of anxiety in girls (i.e. those who spent less than 2-hours in screen-based entertainment were more likely to suffer symptoms related to anxiety). These findings may suggest that screen-based entertainment could be beneficial for relieving/managing anxiety symptoms in children. Alternatively, due to the cross-sectional nature of the study, they may suggest that children with higher levels of anxiety may be less likely to engage in screen-based entertainment and perhaps more likely to engage in other non-screen based sedentary activities such as reading/studying. There is a small body of literature that suggests some forms of sedentary behaviour may in fact have a positive impact on mental health, specifically depressive symptoms [bib_ref] Internet paradox revisited, Kraut [/bib_ref] [bib_ref] Users divided? Exploring the gender gap in Internet use, Shaw [/bib_ref] , however, further intervention and prospective studies are required to determine the direction of the relationship between sedentary behaviour and anxiety symptoms.
All studies included in this review were limited by several methodological weaknesses. For example, most studies (7/9) employed a cross-sectional study design and therefore causality and/or direction of relationships were unable to be determined. Secondly, self, or proxyreport measures of sedentary behaviour were utilised in most (n = 8) studies with such measures increasing the likelihood of recall problems and provision of socially desirable responses and thus to overcome these limitations further research involving objective measures of sedentary behaviour (e.g. accelerometers, posture monitors [i.e. activPALs]) is recommended. However, since the relationship between sedentary behaviour and anxiety risk may be dependent on the domain/type of sedentary behaviour (an aspect of sedentary behaviour that is not able to be measured using such monitors), a combination of both objective and subjective (e.g. self-report surveys) methods is warranted. This further highlights the need for the development of valid objective measures of sedentary behaviour which assess not only the dose (e.g. frequency, duration) but also the domain (e.g. leisure, work, transport), and context (TV viewing, computer use, tablet/smart phone use) in which these behaviours occur. Thirdly, only two studies [bib_ref] Are sedentary television watching and computer use behaviors associated with anxiety and..., De Wit [/bib_ref] [bib_ref] Associations of overall sitting time and sitting time in different contexts with..., Rebar [/bib_ref] compared different types of sedentary behaviours and their relationship with anxiety risk and therefore we were unable to clearly determine which specific sedentary behaviours may be linked to anxiety. Further, since few studies used the same method to define/assess sedentary behaviour, and every study included a different measure of anxiety symptoms, clear-cut conclusions were difficult to determine. Finally, given that anxiety, and depression are often co-morbid disorders [bib_ref] Comorbid depression and anxiety spectrum disorders, Gorman [/bib_ref] , it is difficult to disentangle the relationship sedentary behaviour has with anxiety alone compared to comorbidity associations.
This review enhances the understanding of where the field is at in terms of sedentary behaviour and anxiety research. Although only limited evidence is currently available on the association between sedentary behaviour and risk of anxiety, on the balance this evidence suggests a positive association may exist between overall sedentary behaviour (e.g. sitting time specifically) and anxiety risk, whilst inconsistent evidence remains for other types of sedentary behaviours (e.g. computer use, television viewing, screen time), and their link with anxiety risk. This review further highlights the need for more highquality longitudinal and intervention research to confirm and disentangle cross-sectional research findings.
[fig] Figure 1: Flow of study selection through the phases of the review [/fig]
[table] 1: Strong evidence: defined as consistent (i.e. at least 75 % of studies show results in same direction) results in ≥2 high quality studies 2. Moderate evidence: defined as consistent results in one high quality study and at least one weak quality study; or consistent results in ≥2 weak quality studies 3. Insufficient evidence: defined as having only one available study; or inconsistent results in ≥2 studies. [/table]
|
Cross-sectional examination of 24-hour movement behaviours among 3- and 4-year-old children in urban and rural settings in low-income, middle-income and high-income countries: the SUNRISE study protocol
# Introduction
The early years (defined as <5 years) are arguably the most critical period in life for developing important physical, motor, social and cognitive skills. [bib_ref] Systematic review of the relationships between physical activity and health indicators in..., Carson [/bib_ref] As a sensitive period of brain development, this period in life provides a window of opportunity where developmental plasticity can be exploited to positively influence the trajectory of a child's life in each of the above developmental areas, and to reduce health inequities. [bib_ref] Bridging research in evolution and human health, Lea [/bib_ref] However, it is also a period for which, despite advances in technology, many gaps still exist in the evidence base. Consider a young child's physical and motor development in the context of how they move throughout a typical day. It is a combination of sleep, sitting, standing and different intensities of physical activity, the latter mostly in the form of play and other activities of daily living. Little is known about how these behaviours-individually and in combination-influence one another and how they relate to healthy growth and development. [bib_ref] Systematic review of the relationships between sedentary behaviour and health indicators in..., Poitras [/bib_ref] The dearth of information is even more pronounced in low-income and middle-income countries (LMICs), [bib_ref] The whole day matters: understanding 24-hour movement guideline adherence and relationships with..., Rollo [/bib_ref] many of which are in a period of rapid urbanisation that may further impact the healthy development of these behaviours among this age group. The WHO has identified the prevention of obesity in young children as one of its key priorities for the 21st century [bib_ref] Systematic review of the relationships between sedentary behaviour and health indicators in..., Poitras [/bib_ref] ; movement behaviours play a key role in this priority. A specific recommendation from the WHO Report of the Commission on Ending Childhood Obesity (rec 4.12) 3 was to develop international guidelines for movement behaviours for the early years (<5 years). This is especially important for LMICs, where awareness of the importance of healthy levels of these behaviours in the early years is low, and benchmarks to determine their prevalence are lacking. In response to this recommendation and the growing evidence on the relationship between physical activity, 1 sedentary behaviour [bib_ref] Systematic review of the relationships between sedentary behaviour and health indicators in..., Poitras [/bib_ref] and sleep duration [bib_ref] Systematic review of the relationships between sleep duration and health indicators in..., Chaput [/bib_ref] individually, and in combination, [bib_ref] The whole day matters: understanding 24-hour movement guideline adherence and relationships with..., Rollo [/bib_ref] and health indicators in the early years, the WHO released the first guidelines for physical activity, sedentary behaviours and sleep for children under 5 years of age, in April 2019.These guidelines are based on an integrated movement behaviour paradigm [bib_ref] Introducing 24-hour movement guidelines for the early years: a new paradigm gaining..., Tremblay [/bib_ref] and provide recommendations for each of the three movement behaviours (physical activity, sedentary behaviour and sleep) across a 24-hour period, aligning with national recommendations from countries such as Canada [bib_ref] Canadian 24-hour movement guidelines for children and youth: an integration of physical..., Tremblay [/bib_ref] and Australia. [bib_ref] A collaborative approach to adopting/adapting guidelines -The Australian 24-Hour Movement Guidelines for..., Okely [/bib_ref] Systematic reviews of studies investigating the relationship between movement behaviours and health indicators have reported that the overwhelming majority were conducted in high-income countries (HICs), even multicountry studies, with very few conducted in LMICs and virtually none comparing HICs with LMICs. It is not known how urbanisation and economic development, particularly in LMICs, is associated with young children's movement behaviours. [bib_ref] Height and body-mass index trajectories of school-aged children and adolescents from 1985..., Ncd-Risc [/bib_ref] Globally, it is estimated that 70% of people will live in cities by 2050 and that most of these people will be children or adolescents. [bib_ref] Cities alive: designing for urban Childhoods, Voce [/bib_ref] The majority of this increase in urbanisation will occur in LMICs, and there is concern over how this will be managed. Key challenges particularly in urban environments, such as traffic, pollution, crime, social fears, employment demands, inequitable access to adequate and healthy foods and urban sprawl, may all negatively affect movement behaviours, making it easier for young children to adopt unhealthy levels of these behaviours. In LMICs, there is less likely to be the infrastructure, health and social services, education, economic or policy support to ensure that these core challenges are addressed. [bib_ref] Cities alive: designing for urban Childhoods, Voce [/bib_ref] Rapid urbanisation and high population densities reduce green spaces and public playgrounds, worsen air and noise pollution, increase motorised transportation and reduce walkability, which make outdoor physical activity less safe for children. Sleep duration and quality might also be affected in settings with high population densities, because children often share bedrooms and beds and are more exposed to electronic media in these spaces. [bib_ref] Impact of television on the quality of sleep in preschool children, Brockmann [/bib_ref] Many families transitioning to cities also desire for their children to have technology-centric lifestyles often seen in HICs, resulting in children engaging in long periods of sedentary screen time that probably replaces active play and increases exposure to marketing of unhealthy foods. [bib_ref] A future for the world's children? A WHO-UNICEF-Lancet Commission, Clark [/bib_ref] Not only does this place children at possible risk of overweight and obesity but healthy child development may be compromised due to lack of play and sleep opportunities.
A systematic review 4 of studies examining adherence to 24-Hour Movement Guidelines among preschoolers (ages 3 and 4 years), reported that of the nine studies, all from HICs, between 5%-24% met all three of the daily movement guidelines for this age group: (1) at least 180 min of physical activity, of which at least 60 min is energetic play, (2) no more than 1 hour of Strengths and limitations of this study ► A generic parent questionnaire may not be sensitive enough to identify the contextual nuances necessary to understand the patterns, prevalence and correlates of movement behaviours. ► Reliance on data management platforms that require reliable internet connectivity is a challenge in some regions.
## Open access
sedentary screen time and (3) 10-13 hours of good quality sleep. Prior to the conception of the SUNRISE study there were only two known studies from LMICs that had examined compliance with the 2019 WHO guidelines. A Brazilian study reported that physical activity compliance was the highest at 43%, with 35% and 15% meeting sleep and sedentary screen time recommendations respectively, while only 3% of preschoolers met all three recommendations. [bib_ref] Adherence to 24-hour movement guidelines in low-income Brazilian preschoolers and associations with..., De Lucena Martins [/bib_ref] A study among Chinese kindergarteners found the proportion of children who met the physical activity, sedentary screen time and sleep guidelines were 65%, 88% and 29%, respectively, with only 15% meeting all three guidelines and 2.7% not meeting any of the guidelines. [bib_ref] Proportion of kindergarten children meeting the who guidelines on physical activity, sedentary..., Guan [/bib_ref] For these reasons, the SUNRISE International Study of Movement Behaviours in the Early Years (https:// sunrisestudy. com) was designed to collect data on the movement behaviours of preschool-aged children, the factors that are related to these behaviours and how the movement behaviours are associated with obesity and other health and developmental outcomes in a large-scale international sample of urban and rural communities. The primary aim of the SUNRISE study is to determine the proportion of 3-and 4-year-old children sampled in participating countries who meet the WHO Global guidelines for physical activity, sedentary and sleep behaviour.Further, the study aims to determine if these proportions differ by gender, parental education level, urban/rural location and among countries of differing human and economic development.
As a secondary aim, associations between 24-hour movement behaviours and health and development outcomes will be examined. These outcomes include overweight and obesity, gross and fine motor skills and executive function. These secondary aims have been chosen because significant associations have also been found between the composition of 24-hour movement behaviours and indicators of adiposity and bone and skeletal health among preschoolers. [bib_ref] The whole day matters: understanding 24-hour movement guideline adherence and relationships with..., Rollo [/bib_ref] There is also emerging evidence from a number of countries that the prevalence of developmental delay in important domains such as motor and physical development are high, around 15% for gross motor skills and up to 32% for fine motor skills. [bib_ref] Relationships between motor skill delay and executive functions in vulnerable children aged..., Okely [/bib_ref] [bib_ref] High prevalence of developmental delay among children under three years of age..., Wei [/bib_ref] [bib_ref] Prevalence and risk factors of gross motor delay in Pre-schoolers from low-income..., Veldman [/bib_ref] [bib_ref] Screening for developmental delay among children attending a rural community welfare clinic..., Bello [/bib_ref] In addition the study seeks to determine potential correlates of 24-hour movement behaviours using a social ecological model, to examine inter-relationships between individuals and the social (eg, family, safety, noise, Early Childhood Education and Care (ECEC)), physical (eg, urban/rural, air quality, outdoor play space) and policy (ECEC policies on food, play, screen time) environment.An overarching goal of SUNRISE is the co-creation of new knowledge and building an international network of researchers interested in 24-hour movement behaviour of young children. This paper aims to describe the protocol of the SUNRISE study, detailing how the different iterations of the pilot study phases have contributed to finalising the methods and measures to be used in the SUNRISE main study.
# Methods and analysis
Project leadership and management SUNRISE is guided by an international Leadership Group (see online supplemental table 1) comprising a member from each UN Sustainable Development Goal region (sub-Saharan Africa, Northern Africa and Western Asia, Europe and Northern America, Central and Southern Asia, Eastern and South-Eastern Asia, Oceania, Australia and New Zealand, Latin America and the Caribbean). The group is gender balanced and includes early/midcareer and late-career researchers from low, middle and high-income countries, along with a project statistician and quality assurance expert, WHO representative and an external advisor. The Leadership Group is responsible for developing the protocols for the study, including but not limited to ethical considerations, sampling units and recruitment, governance, budget, data management, training of staff, quality assurance, communications and publications. Video conference meetings are held bimonthly, with the Leadership Executive, made up of three members of the Leadership Group meeting during the alternate months. The SUNRISE Coordinating Centre based at the University of Wollongong (UOW), Australia, is responsible for the overall administration and of the study. This centre comprises a data manager, project, research and equipment officers, a programmer and postdoctoral fellows.
Sample selection SUNRISE is an international cross-sectional study that aims to recruit approximately 1000 healthy, gender balanced children aged 3 and 4 years from each participating country, with equal numbers (500 each) from urban and rural communities representing low-income, middle-income and high-income countries from each major geographical region of the world (see . Countries have been recruited from each of the four levels of the World Bank income classifications (low, middle, high and very high). There has been an effort to spread countries geographically and by income status. Asia and Africa are highly-represented in the sample as 90% of the anticipated increase in the global urban population will occur in these regions over the next 30 years.Recruitment of countries into the SUNRISE study has occurred on a continuous basis. Initially countries were recruited by members of the SUNRISE Leadership Group through existing collaborations. The Leadership Group met in Hong Kong in August 2017 to develop the first iteration of the study protocol. A SUNRISE workshop was held at the 2018 International Society of Behavioral Nutrition and Physical Activity meeting in Hong Kong to introduce the study to the international research community and further recruitment occurred at subsequent conferences, Open access through word of mouth and professional and research student networks. Participating countries have therefore either been actively invited or have expressed interest in participating. In each country a partnering institution and local chief investigator (CI) are identified who take responsibility for all aspects of conducting the study at the local level. Given the vastly differing contexts in which SUNRISE is being conducted, the local CI and research team in each country determine the most appropriate approach to recruiting children of consenting parents/caregivers into the study. This occurs using a convenience cluster sampling approach either through ECEC services, or from the community at a village level. This has occurred in some LMIC rural locations where ECEC services are limited. The sampling frame for each country comprises a geographical area within reasonable travel distance from the CIs institution. If recruitment is conducted through ECEC services, then the primary sampling unit is the ECEC service. In countries where children will be exclusively recruited from both public and private ECEC centres, a maximum of 20 children per centre will be recruited to ensure there is variability within the sample. [bib_ref] The International study of childhood obesity, lifestyle and the environment (ISCOLE): design..., Katzmarzyk [/bib_ref] This will result in a minimum of 50 public and/or private centres participating in the main study. Services will not be randomly selected but countries do need to ensure that the sample is broadly representative of the country in terms of sex, socioeconomic backgrounds of parents [bib_ref] The International study of childhood obesity, lifestyle and the environment (ISCOLE): design..., Katzmarzyk [/bib_ref] and urban/rural residence. [bib_ref] The prospective urban rural epidemiology (pure) study: examining the impact of societal..., Teo [/bib_ref] All children in the selected ECECs or villages who are within the age range and who can wear an accelerometer are eligible to participate. There are no other exclusion criteria.
## Pilot study
To be eligible for participation in the SUNRISE main study each country is required to conduct a pilot study to assess the feasibility and acceptability of the study protocol, particularly recruitment and data collection methods. For the pilot study, 100 children aged 3 and 4 years are recruited equally from urban and rural settings (50 children from each). Each country is advised to use the national definition of the country to determine what is an urban and rural area. These are typically official definitions. This sample size has been deemed sufficient to test the feasibility of the protocol in each setting. In addition to completing all measures proposed for the main study, the pilot study protocol also includes a survey providing information about the settings in which children spend time and the research team's ability to recruit the required sample in an urban and rural area. Several focus groups are conducted in each country with key stakeholders such as parents and ECEC service staff to discuss the acceptability of the measures and the study protocol. Findings from the focus groups (see table 2) have been used to modify the protocol and refine the measures. Results from the pilot studies are also being used for subsequent funding applications for the SUNRISE main study at a country level as well as collectively.
## Patient and public involvement
Since the aim of the pilot work is in part to test and refine the study protocol based on local contexts, data collected have been analysed and used to refine the methods and measures over time. Consequently, there have been three distinct pilot phases for this study. A total of 8 countries participated in Phase I from March 2018 to July 2019, 12 countries participated in Phase II from August 2019 to September 2020 and 21 countries are currently involved in Phase III of the pilot study (see [fig_ref] Figure 1: Map of participating countries [/fig_ref] for map of participating countries). [fig_ref] Table 3: Pilot study phases Countries Eight reports of minor skin irritation following use... [/fig_ref] highlights the main changes that have been made to the protocol across the three phases. Changes were made based on feedback that was obtained through the focus groups with parents and ECEC service staff as well as feedback from local research teams.
The COVID-19 pandemic has resulted in delays for several countries conducting the pilot study. Consequently, there will be some countries still trialling the protocol, while others will commence with the main study. Any additional countries wishing to participate in SUNRISE will use the Phase III protocol for their pilot study and this will also be used in the main study. Additional questions have been added to the protocol to capture the influence of COVID-19 on the movement behaviours in the ECEC setting (see online supplemental table 2).
## Training and capacity building
Face-to-face training sessions organised for all country CIs and key local staff are conducted by the SUNRISE Coordinating Centre in advance of data collection in each country. All staff working on the study receive 2 days of training in all steps of the data collection process with particular emphasis on the administration of all assessments in this age group. Data collectors are certified by trained experts as competent to make the required measurements. This involves the watching of online training videos, observation by trainers of data collectors practicing the assessments on children, viewing protocol videos and successfully completing a test on the assessment protocol as part of the face-to-face training by experts. The CIs from each country are mentored throughout the entire research and surveillance process (see below).
Since August 2020, due to ongoing COVID-19-related travel restrictions, the training sessions have been adapted for online delivery. Training sessions for staff in have been held for the SUNRISE teams in Singapore, Russia and Nigeria. To facilitate this, videos have been developed to demonstrate the assessment protocols and practice assessments are conducted over Zoom or recorded and sent to the coordinating centre for feedback. These are publicly available on the SUNRISE YouTube channel (https://www. youtube. com/ channel/ UCUg mfAG HO1q W7HV 73vDVSLw).
## Open access
## Measures
The measures described below are those used in Phase III of the protocol development.
## Primary outcome
Twenty-four-hour movement behaviours (physical activity, sedentary time and sleep) are assessed using accelerometry 27 (see [fig_ref] Table 3: Pilot study phases Countries Eight reports of minor skin irritation following use... [/fig_ref] for the types of accelerometers trialled). For the main study, the ActiGraph (GT3X, GT3X+) accelerometer will be used. This waist worn device is the most widely used and extensively validated accelerometer for physical activity, sedentary behaviour 28 and sleep assessment. [bib_ref] ActiGraph GT3X+ and Actical wrist and hip worn Accelerometers for sleep and..., Smith [/bib_ref] A pool of accelerometers is made available for use across countries. Children are asked to wear the device continuously (including sleeping and while engaging in water-based activities such as bathing and swimming) for a minimum of 5 days to get three full days (3×24-hour period) of data. This will provide data on total physical activity, moderate-intensity to vigorous-intensity physical activity, total sitting time and total lying time.
In addition, some components of the 24-hour movement behaviours are reported by the primary caregiver on behalf of the participating child. These questions were originally developed based on the recommendations for each behaviour guideline. [bib_ref] A collaborative approach to adopting/adapting guidelines -The Australian 24-Hour Movement Guidelines for..., Okely [/bib_ref] The brief survey, which takes about 15-20 min to complete, asks the caregiver about the child's physical activity (only asked in Phase I), bed, wake and nap times (from which sleep-time is calculated), use of electronic media and restrained sitting (to calculate sedentary time) (see online supplemental table 3).
## Secondary outcomes
A range of health and development outcomes will be obtained to assess associations with 24-hour movement behaviours: During the SUNRISE training, data collectors are advised to administer the assessments tasks to suit each centre programme/schedule, data collector's schedule as required; data collectors trained on device safety and participation ethics.
## Parent feedback on questionnaire
Questions were understood and the support of data collectors and translation into local language assisted comprehension; challenges around estimating time spent in PA during time when kids are at the centre; estimation difficult due to seasonality of PA.
Translation of questions into local language as feasible; provide time ranges as response options; adding seasons/time of year to questionnaire; providing link to video instructions on accelerometer use for parents.
## Children's overall feedback on participation
Parents and ECEC staff both reported that children enjoyed participating in the study.
Confirms the acceptability and appropriateness of the measures and data collection procedures.
## Feedback on accelerometers
Responses to the accelerometers varied.
Most found the monitors interesting and children were excited and proud to wear them. Challenges included placement of activPAL monitor on thigh, discomfort, irritation, difficulties dressing, bathing and sleeping, several reports of rashes and wear time compliance.
Use of the activPAL (placed on thigh) was stopped for Phase II and ActiGraph (belt around waist) was used exclusively. For Phase III all children will wear the ActiGraph on the hip.
Feedback on motor skills tasks Motor skills were seen as age-appropriate and informative for educators about children's strengths and weaknesses. Children enjoyed the tasks.
Confirms the acceptability and appropriateness of the measures and data collection procedures.
## Feedback on executive function tasks
The iPad games were generally perceived as fun and age appropriate. Variability among children sustaining interest due to considerable differences in exposure to devices between countries.
Choice of iPad tasks to be appropriate for most children regardless of previous device exposure. Data collectors trained to support children to feel comfortable in using unfamiliar tools.
ECEC, Early Childhood Education and Care; PA, physical activity.
## Open access
Anthropometry-height and weight are measured to the nearest millimetre and kilogram using a portable stadiometer and an electronic, calibrated scale following standardised procedures as per WHO protocols. Gross and fine motor skills will be measured via validated activities from the National Institute of Health (NIH) Toolbox. [bib_ref] Motor assessment using the NIH Toolbox, Reuben [/bib_ref] All measures are scored as outlined by the NIH Toolbox protocol. Five tests are included.
Gross motor skills ► Children perform a standing long jump to determine lower body explosive strength and mobility. A line is marked on the floor and the child stands with their toes just behind the line. The child then jumps with two feet together as far as they can and lands on two feet. The child is given one practice and two test trials. A measurement is taken from the front of the line and the heel of the foot that is closest to the line. The distance is recorded to the nearest centimetre. The average of the distance values recorded is used. ► A supine-timed up and go (STuG) test is used to assess mobility and posture. A line is marked 3 m from a wall (using tape or chalk). A large target (circle or X) is marked on the wall at the child's eye level. The child lies supine (on their back) with their feet (heels) on the line. On 'go' the child is required to get up as quickly as possible, run and touch the target and run back across the 3 m line. The child has one practice and is then given two trials. Timing is started when the assessor says 'Go' and stopped as soon as the child's torso crosses the line. The average time taken to complete the task is used. ► A one-legged standing balance test is used to measure posture and balance. The child stands on one leg, with the arms held freely at the side of the body for up to 30 s. The child must keep the standing leg fixed but may keep the free leg in any position as long as it is off the floor. Hooking the free leg around the standing leg is not permitted. Swaying is allowed, and the arms may move from the sides but may not hold the free leg.
Timing is started when the free leg leaves the floor.
Timing is stopped if the child moves the standing leg or hooks the free leg around the other leg, or touches the free leg with their hands. If the child maintains balance for 30 s the assessment is stopped. The test is then repeated on the other leg. The length of time that the child is balancing on each leg is recorded. The average time spent balancing is used. ► A handgrip dynamometer (TKK5825, Grip-A, Takei, Tokyo) is used to assess upper extremity strength.
The test measures the capacity of the hand and arm muscles to produce the tension and power necessary for maintaining posture, initiating movement or controlling movement during conditions of loading the musculoskeletal system. The child is required to squeeze the grip dynamometer continuously with full force with their right hand for at least 3 s without letting his/her arms touch their body. The test is then repeated with the left hand. The maximum measure attained is recorded. The timer begins as soon as the assessor says 'Go' and is stopped as soon as the final peg is placed back in the well in the pegboard. The 9-hole pegboard test and grip strength test are based on the assessments in the NIH Toolbox Motor Battery (ages 3-6). [bib_ref] Motor assessment using the NIH Toolbox, Reuben [/bib_ref] The one-legged standing balance test and STuG were recommended as assessments for static balance and mobility (general mobility and locomotion), respectively. This recommendation was made The actiGraph monitor will be used to objectively assess all movement behaviours (ie, sleep, sedentary time and physical activity). Monitor placement will be on the hip. Although the wear compliance rate is not as high as anticipated and there are challenges regarding night-time wear, easy removal of the device and analysis of the sleep data, strategies to resolve these matters are being worked on with the country teams. Parent proxy-report of their child's physical activity was dropped as many parents are not aware of the amount of physical activity their child participates in when at an ECEC service. Physical activity data are exclusively collected via accelerometry.
## Open access
The parent report question on total sleep time was broken down to capture, bed time, wake time and naps as we have found that the main time when children remove the accelerometer is when they go to bed so it provides us with some data on sleep for these children. Also, parent-reported information tends to report the time their child went to bed and got up which is different to the accelerometer which reports the time they went to sleep and woke up. Further parents also report on sedentary screen time as this cannot be assessed by accelerometry. Elements from the non-proprietary NIH Toolbox were deemed more culturally feasible and will replace the ASQ to assess motor skills. Food habits and eating behaviour questions added to parent/caregiver questionnaire to provide additional context for interpreting anthropometric data. Centre questions added to gain more insight into the influences on movement behaviours in the child care centre setting.
ASQ, Ages and Stages Questionaire; ECEC, Early Childhood Education and Care; EYT, Early Years Toolbox; NIH, National Institute of Health.
## Table 3 continued
Open access by the Motor Domain Group in their proposal for assessment of motor function for mobility and posture for the National Children's Study (USA).For assessment of mobility this Group recommended the vertical jump for age 5 onwards. On consultation with the lead author, it was recommended that a standing long jump be used for children under the age of 5. The standing long jump is widely used in motor development assessment batteries for this age group. Executive function Inhibition and working memory, two key indicators of cognitive function, will be measured using the validated Early Years Toolbox (EYT). [bib_ref] An early years toolbox for assessing early executive function, language, self-regulation, and..., Howard [/bib_ref] Two brief, game-like assessment from the toolbox will be administered via an iPad and scored according to the EYT protocol. ► The Go/No-Go task is an assessment of 'inhibition'-the ability to control behavioural urges and impulses. Children are presented with fish and sharks and are instructed to tap the iPad screen whenever they see a Fish ('catch the fish') and refrain from responding when a Shark appears ('avoid the sharks'). ► The Mr Ant task is an assessment of 'visual-spatial working memory'-the amount of visual information that can be concurrently coordinated in the mind. Children are presented with an image of a cartoon character-Mr Ant-who has a number of coloured dots placed in different spatial locations on his body. After a predetermined amount of time, these dots disappear, and the child is then asked to recall the locations of the dots by tapping the corresponding locations.
## Parent questionnaire
In addition to the parent questionnaire on the child's 24-hour movement behaviour, standardised validated questions are asked of parents to assess potential correlates of 24-hour movement behaviour using a social ecological model, to examine inter-relationships between individuals, social, physical and physical environment.This includes questions on the child's dietary diversity,eating behaviour and food insecurity 37 at the family level. Further, sociodemographic information based on a modified version of the WHO STEPS survey 38 is recorded (see online supplemental [fig_ref] Table 3: Pilot study phases Countries Eight reports of minor skin irritation following use... [/fig_ref]. The questions assessing children's movement behaviours were based on the recommendations made by the surveillance subcommittee of the Guideline Development Group for the Australian 24-hour movement behaviour guidelines for the early years in 2017. The committee provided guidance on how to assess the proportion of children meeting the recommendations for each of the movement behaviours. The questionnaire in its entirety has not yet been validated.
## Ecec service questionnaire
In settings where recruitment occurs at the ECEC service level, influences on 24-hour movement behaviour are assessed via a staff questionnaire. Questions cover the services' policies around food provision, the physical activity habits of the children and their sleep behaviour while at the ECEC service, to assess centre-level policy correlates of 24-hour movement behaviours (see online supplemental table 2).
All instruments were assessed for their cultural relevance and appropriateness, and in countries where English is not a first language, the instruments were translated by a native speaker and back-translated to ensure accuracy. [fig_ref] Table 4: Response rates from pilot studies as at article submission date Country [/fig_ref] reports the initial results from the pilot studies including response rates for the main measures.
Outdoor air quality Pollution levels will be measured at each participating ECEC service or at the village level during data collection using the Plume Labs application. The Plume mobile application, provides real-time pollution level data provided by environmental monitoring networks run by local and national governments around the world. Specifically, the application reports on particulate matter 2.5 (PM2.5), PM10 levels and give and overall air quality index. The purpose of this is to determine whether air quality is related to children's outside physical activity time and intensity during the day.
## Data collection procedures
All data are collected by local data collectors in each country. A detailed protocol manual sets out the data collection instruments and procedures. To minimise errors caused by entering data from hard copy to electronic format, and to ensure data transfer efficiency, data are collected and managed using REDCap electronic data capture tools hosted at the UOW, Australia. REDCap (Research Electronic Data Capture) is a secure, web-based application designed to support data capture for research studies, providing: (1) an intuitive interface for validated data entry; (2) audit trails for tracking data manipulation and export procedures; (3) automated export procedures for seamless data downloads to common statistical packages; and (4) procedures for importing data from external sources. One exception is the parent questionnaire, which is either completed by parents on paper or where literacy poses challenges, via interview, with the data entered directly into REDCap by the data collector. The Coordinating Centre loans iPads to participating countries with the REDCap projects and Early Years Toolbox games preloaded onto them.
Initial contact with individual ECEC service or village is made by a member of the local research team in each country. Dates for each service/village visit are then provided to all data collectors and local protocols for entering and exiting each service/village are followed.
As there are several measures to be taken with the children, the following order has been trialled and is recommended to ensure a smooth data collection process, giving children sufficient variability between tasks to sustain their engagement: (1) Executive function Open access task (Mr Ant), (2) attach ActiGraph accelerometer, (3) measure height and weight, (4) assess gross motor skills, (5) assess fine motor skills and (6) executive function task 2 (GoNoGo). All measures are conducted on 1 day and take approximately 20 min to complete. It is not anticipated that participation in any measure will compromise the results of any subsequent measures. Data in REDCap can be collected offline. To commence, the participant is allocated their unique study identification number. The consent form must then be completed which allows the remaining data to be collected and entered as per the suggested order. The data are saved as the data collector progresses through the various screens. The accelerometer monitor identification number is also entered to enable linking with the accelerometry data. Once all data have been collected and the iPad is able to be connected to the internet all data are sent to the project's REDCap server. From the server it is transferred to the UOW Coordinating Centre's databases for checking, and analysis. Comprehensive data guidelines that set out the principles, protocols, methods and procedure governing the management, access, use and dissemination of the data have been developed for the study.
## Open access
# Data analysis plan
Data on participant attributes such as demographic and anthropometric and movement behaviours will be summarised separately for boys and girls, and across urban and rural settings, as counts and percentages for categorical variables and means and SD for continuous variables. Given that the primary aim of the SUNRISE study is to determine the proportion of 3-and 4-year-old children sampled in participating countries who meet the WHO Global guidelines for physical activity, sedentary and sleep behaviour, prevalence estimates analysis will be conducted. We will also report the proportions who meet any combinations of the guidelines, including those who do not meet any recommendations. Linear and generalised linear models will be used for association analysis to assess whether associations exist (i) between meeting all (or any of the) guidelines and factors, such as sex (primary aims) and (ii) between health and development outcomes and 24-hour movement behaviours (secondary aims). The models will be covariate-adjusted, and treat ECEC service/villages within site and children within ECEC service/villages as well as ECEC service/ villages within countries as random effects for all analyses. Statistical significance will be defined as p<0.05 with appropriate adjustments for multi-testing.
Main study sample size calculation Data obtained from SUNRISE pilot studies from 17 countries, showed that the proportions of children meeting all components of the WHO Global guidelines varied across countries from 2.3% to 42.7% with a mean country proportion of 21.0%. These proportions also differed within many countries when comparing rural with urban areas. The mean absolute difference in the proportion between rural and urban was 9.6%.
The power calculation is based on achieving 80% power and a 5% significance level for each country and assuming a rate of 21% of meeting all three guidelines and detecting a difference of 9.6% in either direction (two-sided) between urban and rural and based on an equal allocation to both rural and urban settings. This provides an effect size of 0.23 (small effect),and results in a sample size of n=558 per country.
When assuming a response rate of 76% (based on average response rate in the 16 pilot studies), the required sample size increases to n=734. The sampling of the main study will be partly based on cluster sampling (usually childcare centres). The intra-class correlation (ICC) was estimated for each country, however, due to low pilot sample sizes in each country; the countryspecific ICC estimates vary widely and are unreliable. Instead, we combined countries and estimated the ICC after controlling for country-specific effects. The resulting ICC estimate was zero. However, a zero ICC might be too optimistic (resulting in smaller sample size) for the main study, possibly resulting in too low power, if the ICC is indeed positive in the main study. To be conservative, we used an ICC estimate of 0.022 from the PATH study. [bib_ref] Adherence to 24-hour movement guidelines for the early years and associations with..., Cliff [/bib_ref] Taking this value and assuming data to be collected on 20 children (on average) per child care centre when 25 are recruited (on average), the sample size of n=735 is further increased to approximately n=1000. This sample size is also sufficiently large to have a margin of error of at most 5% for a 95% CI for the proportion of meeting all three guidelines for each country and for both rural and urban populations when centred around 21%.
## Ethics and dissemination ethics approval
The overarching protocols for the pilot study (2018/044) and the main study (2019/378) have been approved by the UOW, Australia, Human Research Ethics Committee. The protocol is also approved in each country by the ethics committees at the participating institutions of each CI, as applicable (see online supplemental table 4). In addition, approval is sought from any government departments or organisations who oversee relevant settings in the different countries. Participant information sheets, consent forms and questionnaires are translated and back translated, as necessary, into the local language of each country following approved procedures of the local institutions. Parental consent is obtained for all participating children.
Quality assurance of data collection, storage and management This component of the project has been supported by the Australian Health Services Research Institute at the UOW. A comprehensive Research Data Management Plan that sets out the principles and protocols governing the management, access, use and dissemination of the data has been developed. The data are securely stored in databases on Open access password-protected servers at the UOW. All participating countries retain ownership of their data. The terms of collaboration are detailed in a Collaboration Agreement that is entered into between the UOW and the participating institutions of each CI. This collaboration agreement also details intellectual property issues, data use, confidentiality, privacy, loaning of equipment and the roles and responsibilities of each party. The agreement is reviewed by each country's legal representative and executed by both the participating country's authorised representative and the UOW authorised representative, prior to training in each country.
## Global impact of sunrise
As the first-known international study of movement behaviours in the early years, SUNRISE will make a significant contribution to our understanding of the physical activity, sedentary behaviour and sleep patterns of preschool aged children. Further, SUNRISE will investigate if these behaviours differ by gender, parental education level, urban/rural location and among countries of differing human and economic development, and how these compare to current global movement behaviour guidelines. The study will also enable analysis of associations between movement behaviours and health and developmental indicators and provide evidence on potential correlates of movement behaviours among young children.
A considerable strength of the study is the breadth of the resulting data set. SUNRISE data will be collected in urban and rural settings in 39 countries, from every WHO region, with up to 60% of participating countries being of low-income or middle-income, resulting in a truly unique data set. A further strength of the study is the extensive pilot testing of the protocol (as described in this article) which has resulted in a scientifically robust, culturally appropriate, feasible, standardised protocol. The strong data management and capacity building elements incorporated into the design of the study are a further strength.
In conclusion, the results to be obtained by SUNRISE, particularly on the correlates of movement behaviours, have the potential to inform the development of future interventions to improve the movement behaviours in culturally specific ways across a diverse range of settings. The collaborative international network of researchers and practitioners brought together by this study will be instrumental in driving this important agenda further and will no doubt make considerable contributions to ensuring that young children reach their full developmental potential.
[fig] Fine motor skills The 9 -: hole pegboard test (PAT-A8515, Sammons Preston, Illinois, USA) is used to assess dexterity or manipulation. Dexterity is a central component of hand function and relates to both the speed and accuracy of hand movements. A child is timed picking up nine pegs one at a time and inserting them into the pegboard (31.1 cm × 26.0 cm × 4.3 cm). The test is then repeated with the left hand. [/fig]
[fig] Figure 1: Map of participating countries. [/fig]
[fig] Author affiliations 1: Early Start, School of Health and Society, Faculty of the Arts, Social Sciences and Humanities, University of Wollongong, Wollongong, New South Wales, Australia 2 Physical Activity for Health, University of Strathclyde, Glasgow, UK 3 Children's Hospital of Eastern Ontario Research Institute, Ottawa, Ontario, Canada 4 MRC-Wits DPHRU, University of the Witwatersrand, Johannesburg- [/fig]
[table] Open access Table 1: SUNRISE participating country characteristics Shatin, N.T. Hong Kong Urban: Hong Kong Island, Kowloon and the New Territories Ho Chi Minh City, Vietnam Urban: District Tân Bình and District 1, Ho Chi Minh city Rural: District Bình Chánh and District Nhà Bè, Ho Chi Minh city Hong Kim Tang Thanh Van Kim *Obtained from the World Bank. Data-World Bank Country Lending Groups, 2020. AFRO, African region; EMRO, Eastern Mediterranean region; EURO, European region; PAHO, Pan-American region; SEARO, South-east Asian region; WPRO, Western Pacific region. [/table]
[table] Table 2: Focus group and interview findings [/table]
[table] Table 3: Pilot study phases Countries Eight reports of minor skin irritation following use of the activPAL were reported in Canada, Bangladesh and Australia. Two other studies have also documented minor cases of skin irritation. [/table]
[table] Table 4: Response rates from pilot studies as at article submission date Country [/table]
|
Amphiphilic aminoglycosides: Modifications that revive old natural product antibiotics
Widely-used Streptomyces-derived antibacterial aminoglycosides have encountered challenges because of antibiotic resistance and toxicity. Today, they are largely relegated to medicinal topical applications. However, chemical modification to amphiphilic aminoglycosides can revive their efficacy against bacterial pathogens and expand their targets to other pathogenic microbes and disorders associated with hyperactive connexin hemichannels. For example, amphiphilic versions of neomycin and neamine are not subject to resistance and have expanded antibacterial spectra, and amphiphilic kanamycins are effective antifungals and have promising therapeutic uses as connexin hemichannel inhibitors. With further research and discoveries aimed at improved formulations and delivery, amphiphilic aminoglycosides may achieve new horizons in pharmacopeia and agriculture for Streptomyces aminoglycosides beyond just serving as topical antibacterials.
# Introduction
Naturally occurring aminoglycosides (AGs) isolated primarily from Streptomyces sp. are important broad spectrum antibacterial agents that have been used clinically for decades; [bib_ref] Aminoglycosides: A practical review, Gonzalez [/bib_ref]. Besides streptomycin, neomycin and kanamycin have been two of the most studied and employed classes of AG antibiotics. Neomycin belongs to a group of AGs containing a 4,5-disubstituted 2-deoxystreptamine core (ring II), while kanamycin contains a 4,6-disubstituted 2-deoxystreptamine core. Other AG groups include gentamicin [bib_ref] Gentamicin, a new antibiotic complex from micromonospora, Weinstein [/bib_ref] and sisomicin [bib_ref] Antibiotic 6640, a new micromonospora-produced aminoglycoside antibiotic, Weinstein [/bib_ref] , which structurally resemble kanamycin antibiotics but are produced by Micromonospora sp. Structurally, gentamicin and sisomicin usually contain deoxygenation at ring I, and have a different ring III aminosugar from kanamycin. The emergence of antibiotic resistant bacteria has significantly hampered the use of naturally occurring AGs.
During the 1970s, several semisynthetic AGs with improved activity against resistant bacteria were developed, such as amikacin [bib_ref] BB-K8, a new semisynthetic aminoglycoside antibiotic, Kawaguchi [/bib_ref] , netilmicin, isepamicin [bib_ref] The syntheses and biological properties of l-n-(s-4-amino-2-hydroxybutyryl)-gentamicin b and 1-n-(s-3-amino-2-hydroxypropionyl)-gentamicin B, Nagabhushan [/bib_ref] , arbekacin, and most recently plazomicin;. However, these new AGs were limited when confronted with bacterial resistance arising from overexpression of AGmodifying enzymes. Over a hundred different AG-modifying enzymes are known to inactivate AGs by introducing structural modifications at various AG sites [bib_ref] Aminoglycoside modifying enzymes, Ramirez [/bib_ref] [bib_ref] Mechanisms of resistance to aminoglycoside antibiotics: Overview and perspectives, Garneau-Tsodikova [/bib_ref]. Some display substrate promiscuity and modify both neomycin and kanamycin classes [bib_ref] Substrate promiscuity of an aminoglycoside antibiotic resistance enzyme via target mimicry, Fong [/bib_ref]. In addition, ototoxicity and nephrotoxicity often associated with AGs persist despite efforts devoted to lowering these side effects [bib_ref] New developments in aminoglycoside therapy and ototoxicity, Xie [/bib_ref]. Finally, as a result of the prevalence of AGmodifying enzymes, more structural motifs are needed at specific positions of AGs which adds to the challenges and cost of synthesizing new AGs. As an example, an expensive AG, sisomicin, was used as starting material for the development of plazomicin to circumvent part of the problem of regioselective chemical modification, which even at a relatively concise synthesis resulted in a cost >$3,000 US dollars per 100 mg.
An alternative approach is to revive and repurpose AGs using cost-effective AGs like neomycin or kanamycin as feedstocks for the synthesis of amphiphilic AGs (AAGs) [bib_ref] Antifungal amphiphilic aminoglycosides, Chang [/bib_ref] [bib_ref] Antibiotic hybrids: The next generation of agents and adjuvants against gram-negative pathogens?, Domalaon [/bib_ref]. AGs are hydrophilic, and the attachment of hydrophobic groups creates AAGs. In general, AAGs have two distinct features that make them different from AGs: broader antimicrobial profile and altered mode of action. AGs are active against aerobic Gram-positive and Gram-negative bacteria but inactive against anaerobic bacteria and fungi. In contrast, AAGs have antibacterialaerobic and facultative anaerobic-and antifungal activities. Traditional AGs display antibacterial activity by binding to 16S rRNA thereby interfering with protein synthesis, whereas AAGs exert bioactivity by primarily targeting microbial membranes [bib_ref] Investigation of antibacterial mode of action for traditional and amphiphilic aminoglycosides, Udumula [/bib_ref]. The latter bioactivity allows AAGs to be active against bacteria that are resistant to AGs, and also against fungi.
In contrast to the semisynthesis of AGs, where regioselective incorporation of structural motifs is essential, the synthesis of AAGs can be performed at the sites that are most amenable to modification. This drastically alleviates the burden of chemical modification and the cost of production. The most cost-effective feedstocks of AGs are neomycin B and kanamycin A, and this review focuses on the applications and biological activities of AAGs derived from neomycin B, kanamycin A and neamine, a derivative from neomycin B. Particular attention is directed toward synthetic approaches as they dictate the costs of production and the feasibility of offering marketable products.
## Amphiphilic neamines and neomycins
Several groups have reported the synthesis and antibacterial investigations of amphiphilic neomycins; [bib_ref] Design, Synthesis, and Antibacterial Activities of Neomycin-Lipid Conjugates: Polycationic Lipids with Potent..., Bera [/bib_ref] [bib_ref] Antibacterial activities of aminoglycoside antibiotics-derived cationic amphiphiles. Polyol-modified neomycin B-, kanamycin A-,..., Bera [/bib_ref] [bib_ref] Surprising alteration of antibacterial activity of 5 -modified neomycin against resistant bacteria, Zhang [/bib_ref] [bib_ref] Synthesis and Combinational Antibacterial Study of 5"-Modified Neomycin, Zhang [/bib_ref]. The lead amphiphilic neomycins show broad spectrum and unusual biological activities, especially against bacteria resistant to AGs. For example, 5"-aminoneomycin with attached linear acyl groups displays prominent antibacterial activities against a panel of bacteria, including methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococci (VRE). The activity against VRE is of particular interest since facultative enterococci are intrinsically resistant against traditional AGs such as neomycin, amikacin, gentamicin, tobramycin and kanamycin due to the lack of AG-uptake mechanisms. Therefore, the unusual activities of AAGs, especially against VRE, strongly imply that the attachment of hydrophobic groups results in alteration of the traditional antibacterial mode of action.
The antibacterial minimum inhibitory concentrations (MICs) of amphiphilic neomycins are summarized in Tables 1, 2. In general, the amphiphilic neomycin derivatives in [fig_ref] TABLE 1: MICs [/fig_ref] are less active against Gram-negative bacteria. There is no clear structure-activity relationship (SAR) for the derivatives with aromatic motifs attached (1b,c, 2g,h), but the activity of 1c against MRSA is of particular interest [fig_ref] TABLE 1: MICs [/fig_ref]. For the derivatives with linear alkyl chains attached (1a, 2a-f), increasing the hydrophobicity (lipophilicity) appears to enhance antibacterial activity, as 2e (C16, hexadecyl) is the most active. Nevertheless, extending the alkyl chain further to C18 (octadecyl) decreases the activity slightly. The most significant findings are the unusual activities of 2e against MRSA (Tables 1, 2) and VRE. Traditional AGs are inactive or much less active against these two pathogenic bacterial strains. Amphiphilic neomycins with hydrophobic groups attached at the N-6 position have also been investigated [bib_ref] Paromomycin and neomycin B derived cationic lipids: Synthesis. and transfection studies, Mével [/bib_ref] , and trends of biological activities similar to those of other amphiphilic neomycins were observed.
Neamine is obtained by acid-hydrolysis of neomycin. Modification to amphiphilic neamine usually begins with the protection or masking of amino groups, for example tritylation followed with alkylation of hydroxyl groups and global deprotection; [bib_ref] Di-Nmethylation of anti-gram-positive aminoglycoside-derived membrane disruptors improves antimicrobial potency and broadens spectrum..., Benhamou [/bib_ref] [bib_ref] Bacterial lipid membranes as promising targets to fight antimicrobial resistance, molecular foundations..., Mingeot-Leclercq [/bib_ref] [bib_ref] New Broad-Spectrum Antibacterial Amphiphilic Aminoglycosides Active against Resistant Bacteria: From Neamine Derivatives..., Zimmermann [/bib_ref]. In Structures of natural AGs. Structures of semisynthetic aminoglycosides.
general, amphiphilic neamines have been noted for their unique activities especially against Pseudomonas sp [bib_ref] Bacterial lipid membranes as promising targets to fight antimicrobial resistance, molecular foundations..., Mingeot-Leclercq [/bib_ref] [bib_ref] New Broad-Spectrum Antibacterial Amphiphilic Aminoglycosides Active against Resistant Bacteria: From Neamine Derivatives..., Zimmermann [/bib_ref]. This is significant as Pseudomonas sp. are known for their unusual membrane compositions and efflux resistance that compromise the activity of many antibiotics. Additionally, amphiphilic neamines with di-N-methylation have improved activities against Gramnegative bacteria [bib_ref] Di-Nmethylation of anti-gram-positive aminoglycoside-derived membrane disruptors improves antimicrobial potency and broadens spectrum..., Benhamou [/bib_ref].
Amphiphilic paromomycins have also been synthesized and employed as inhibitors against AG-modifying enzymes [bib_ref] Inhibition of aminoglycoside-deactivating enzymes aph(3 )-iiia and aac(6 )-ii by amphiphilic paromomycin..., Szychowski [/bib_ref] and antibacterials toward Grampositive pathogens [bib_ref] Siteselective displacement of tobramycin hydroxyls for preparation of antimicrobial cationic amphiphiles, Berkov-Zrihen [/bib_ref]. AAGs derived from streptomycin, gentamicin and sisomicin have not been reported, probably due to cost considerations and the challenges of site-selective modifications.
Since AAGs have been shown to have mode of antibacterial action different from tradidional AGs, several groups have engaged in investigations on their mode of action. Using atomic force microscopy, Mingeot-Leclercq and co-workers have reported that amphiphilic neamine derivatives bind to the lipopolysaccharides of P. aeruginosa and induce membrane depolarization (Mingeot-Leclercq and Décout, 2016). More evidence supporting this mode of action comes from studies using the fluorogenic dye SYTOX [bib_ref] Investigation of antibacterial mode of action for traditional and amphiphilic aminoglycosides, Udumula [/bib_ref]. SYTOX cannot penetrate the intact bacterial membrane, but enters cells through pores formed in the bacterial membrane and binds nucleic acids which results in strong fluorescence. Following exposure to 6e, Escherichia coli and S. aureus strains showed high levels of green fluorescence, whereas S. aureus cells treated with neomycin and SYTOX did not, as pore formation with traditional AGs is not expected.
It is interesting to note the similarities between the antibacterial activities of AAGs and lipopeptides, another class of amphiphilic compounds. Addition of acyl chains to peptides alters membrane lipid composition and packing. Structures of representative amphiphilic neomycins.
The antimicrobial activity of lipopeptides is associated with membrane modifications, including changes in its curvature, surface charge and dipole potential, as well response to changes in ionic strength [bib_ref] Role of Lipid Composition, Physicochemical Interactions, and Membrane Mechanics in the Molecular..., Balleza [/bib_ref] [bib_ref] Antimicrobial peptides: Mechanism of action and lipid-mediated synergistic interactions within membranes, Juhi [/bib_ref]. Although details of the mode of action of AAG are still being deciphered, it is likely that they will resemble those of antimicrobial lipopeptides.
Among the reported investigations of amphiphilic neomycin and neamine, several studies found these compounds less toxic based on hemolysis experiments [bib_ref] Synthesis and Combinational Antibacterial Study of 5"-Modified Neomycin, Zhang [/bib_ref] [bib_ref] Di-Nmethylation of anti-gram-positive aminoglycoside-derived membrane disruptors improves antimicrobial potency and broadens spectrum..., Benhamou [/bib_ref]. Compounds 2a, 2e and 2f were found to be weakly antifungal with MICs ranging from 62.5 to 125 µg/mL against the wheat and barley pathogen Fusarium graminearum. However, little is known regarding the bioactivity of these compounds beyond their antibacterial activity.
## Amphiphilic kanamycins
Early successes in the synthesis of amphiphilic kanamycins (AK) employed kanamycin B as the starting material [bib_ref] New derivatives of kanamycin B obtained by modifications and substitutions in position..., Schepdael [/bib_ref]. Following the protection of Boc groups, the 6"-OH (the only primary hydroxyl group) can be selectively modified using a nucleophilic substitution approach. Derivatives with various functional groups, such as thioether, alkoxyl, alkylamino, alkylamido and esters, have been synthesized and evaluated for their antibacterial activities. Although these compounds manifest good antibacterial activity against AG susceptible strains, they do not display significant antibacterial activity against AG-resistant bacteria such as Streptococcus faecalis and P. aeruginosa [bib_ref] New derivatives of kanamycin B obtained by modifications and substitutions in position..., Schepdael [/bib_ref].
Tobramycin, a kanamycin class AG, has been converted to 6"-modified amphiphilic derivatives using similar methods; [bib_ref] 6 -Thioether tobramycin analogues: Towards selective targeting of bacterial membranes, Herzog [/bib_ref]. From the selected MICs, a long linear alkyl chain (C14 and C16) seems to improve antibacterial activity [fig_ref] TABLE 3: MICs of 6"-modified amphiphilic tobramycin derivatives [/fig_ref]. The most active compounds,8a and 8b, inhibit growth of VRE and tobramycin-resistant E. coli. On the other hand, oxidation of alkylthio to sulfones or sulfoxides slightly decreases antibacterial activities.
Other AK designs have been reported, including kanamycin A with a 6 -alkyltriazole group [bib_ref] Evaluation of amphiphilic aminoglycoside-peptide triazole conjugates as antibacterial agents, Bera [/bib_ref] , 4 ,6dialkylthio tobramycin (Berkov-Zrihen et al., 2013b) and 6 -chelostolcarbonyl kanamycin A. [bib_ref] Design, Synthesis, and Antibacterial Activities of Neomycin-Lipid Conjugates: Polycationic Lipids with Potent..., Bera [/bib_ref] [bib_ref] Antibacterial activities of aminoglycoside antibiotics-derived cationic amphiphiles. Polyol-modified neomycin B-, kanamycin A-,..., Bera [/bib_ref] Synthesis of amphiphilic neamine.
In general, these compounds manifest broad spectrum antibacterial activities at low micromolar concentrations. In 2006, it was reported that AGs such as neomycin, paromomycin, ribostamycin and streptomycin possess antifungal activities against Phytophthora and Pythium species. Inspired by this article, screening of libraries of neomycin, kanamycin and pyranmycin derivatives led to the identification of AK FG08 as the most potent antifungal AG tested ; [bib_ref] Antibacterial to antifungal conversion of neamine aminoglycosides through alkyl modification. Strategy for..., Chang [/bib_ref]. Interestingly, FG08 was inactive against bacteria, a discovery pointing to a new and significant bioactivity of AAGs and AKs with potential for broader applications in medicine and agriculture. Furthermore, the antifungal activity of FG08 involved formation of pores in the plasma membrane [bib_ref] Membrane lipid-modulated mechanism of action and non-cytotoxicity of novel fungicide aminoglycoside FG08, Shrestha [/bib_ref]. Therefore, Synthesis of AKs. Synthesis of amphiphilic tobramycin. attachment of hydrophobic groups to make AAGs not only expanded the antibacterial target species spectrum, but also uncovered a "switch" that adds antifungal activity to the AG antimicrobial profile.
Although FG08 showed promising fungicidal activity against crop disease caused by Fusarium head blight (FHB), its complex synthesis makes it unsuitable for scale-up production, and therefore, practical applications for FG08 in medicine and agriculture are limited. Based on the SAR investigation of FG08 and related compounds , a simplified protocol for the production of a 2nd generation antifungal AK (K20) was explored [bib_ref] Antifungal amphiphilic aminoglycoside K20: Bioactivities and mechanism of action, Shrestha [/bib_ref]. In addition to showing broad spectrum antifungal activity, K20 can be synthesized in 200-300 g scale batch [fig_ref] TABLE 4: MICs [/fig_ref]. Furthermore, K20 also exerts strong synergism with commercially available crop fungicides [bib_ref] Antifungal amphiphilic aminoglycoside K20: Bioactivities and mechanism of action, Shrestha [/bib_ref] [bib_ref] In vitro antifungal synergy between amphiphilic aminoglycoside K20 and azoles against Candida..., Shrestha [/bib_ref] and these K20 combinations demonstrated efficacy in wheat field trials against FHB . In short, a combination of (2 )).
## Figure 7
Structures of FG08 and K20.
K20 and crop fungicides can reduce FHB and production of the toxin deoxynivalenol that is associated with FHB. These findings suggest the possible use of K20 in combination with established crop fungicides at lower concentrations, thus reducing costs, toxicities and environmental impacts of crop protection strategies. Newer and improved antifungal AK designs have been reported since the discovery of K20 [bib_ref] In vitro antifungal synergy between amphiphilic aminoglycoside K20 and azoles against Candida..., Shrestha [/bib_ref] [bib_ref] Serum prevents interactions between antimicrobial amphiphilic aminoglycosides and plasma membranes, Logviniuk [/bib_ref]. A one-step synthesis of AK was developed to better compete with the commercially available crop fungicides and therapeutic antifungal agents [fig_ref] FIGURE 8 One: -step synthesis of AKs 11a and 11b [/fig_ref] ; [bib_ref] Scalable and cost-effective tosylation-mediated synthesis of antifungal and fungal diagnostic 6 -Modified..., Subedi [/bib_ref]. Among the compounds synthesized, the lead compounds 11a and 11b show strong antifungal activities, including activities against clinical strains of C. auris and Aspergilla fumigatus [fig_ref] TABLE 5 Summarized: MICs [/fig_ref]. In pharmacokinetic studies in a mouse model, 11b did not show signs of toxicity (e.g., body weight loss) up to the highest doses tested (52 mg/kg). The cost of 11a and 11b production is lower than those of K20, making them more competitive than the widely used azole-based antifungal agents. As a group, the antifungal AKs represent a significant breakthrough for the classical AGs as they allow for the repurposing of large stockpiles of kanamycin sulfate, which is otherwise clinically obsolete as an antibacterial agent.
## Amphiphilic kanamycins as connexin hemichannel inhibitors
Large stockpiles of AGs, mainly neomycin sulfate and kanamycin sulfate are available worldwide, but their clinical use is limited due to the prevalence of bacterial resistance. Thus, efforts have been devoted to repurposing AGs for other applications. These include AGs as therapeutic options for spinal muscular atrophy (SMA) [bib_ref] Novel aminoglycosides increase SMN levels in spinal muscular atrophy fibroblasts, Mattis [/bib_ref] and disorders associated with hyperactivity of connexin hemichannels (HCs) [bib_ref] Inhibition by commercial aminoglycosides of human connexin hemichannels expressed in bacteria, Fiori [/bib_ref]. Following these examples, AKs have also been investigated as connexin HC inhibitors with the goal of identifying leads that do not have antibacterial activity, show reduced cytotoxicity, and target HCs formed by specific connexin isoforms.
Connexins are transmembrane proteins that constitute the key building blocks of connexin HCs. Gap-junction channels (GJCs) formed by the head-to-head docking of two HCs, one from each of two adjacent cells, play diverse and pivotal roles in development, intercellular communication, and regulation [bib_ref] Gap junctions: Basic structure and function, Meşe [/bib_ref] [bib_ref] Gap junctions, Nielsen [/bib_ref]. Abnormal increases and decreases in the activity of HCs and GJCs have been linked to numerous genetic and acquired disorders such as deafness [bib_ref] Connexin-26 mutations in deafness and skin disease, Lee [/bib_ref] , heart arrhythmias and infarcts [bib_ref] Hemichannels in cardiomyocytes open transiently during ischemia and contribute to reperfusion injury..., Shintani-Ishida [/bib_ref] [bib_ref] Connexin 43 is an emerging therapeutic target in ischemia/reperfusion injury, cardioprotection and..., Schulz [/bib_ref] [bib_ref] Connexins in cardiovascular and neurovascular health and disease: Pharmacological implications, Leybaert [/bib_ref] , and cancer [bib_ref] Gap junctions and cancer: Communicating for 50 years, Naus [/bib_ref] [bib_ref] Connexin's Connection in Breast Cancer Growth and Progression, Banerjee [/bib_ref]. Connexin HCs are mostly closed, but abnormally increased HC activity in inherited or acquired disorders results in cell damage. However, there are more than twenty different connexin isoforms, which have different functional properties and regulation, and are expressed differentially in various cells, tissues and organs. Thus, desirable inhibitors should be selective not only toward connexin HCs but also the HCs of interest.
Among the connexin isoforms, connexin 43 (Cx43) and connexin 26 (Cx26) have attracted particular attention. Cx43 is expressed in many organs, including heart, brain, liver, kidney, skin, and myometrium. Mutations of Cx43 have been reported to associate with acute and chronic diseases occurred in these organs [bib_ref] Cardiac connexins, mutations and arrhythmias, Delmar [/bib_ref] [bib_ref] Connexin 43: A new therapeutic target against chronic kidney disease, Prakoura [/bib_ref] [bib_ref] Suppression of connexin 43 phosphorylation promotes astrocyte survival and vascular regeneration in..., Slavi [/bib_ref] [bib_ref] The Role of Connexin-43 in the Inflammatory Process: A New Potential Therapy..., Xu [/bib_ref] Amphiphilic kanamycin as connexin inhibitor. [bib_ref] Inhibition of connexin 43 attenuates oxidative stress and apoptosis in human umbilical..., Ma [/bib_ref] , whereas Cx26 mutations can lead to deafness and/or skin disorders [bib_ref] Connexin-26 mutations in deafness and skin disease, Lee [/bib_ref]. Following the recent discovery that natural AGs inhibit connexin HCs without affecting GJCs [bib_ref] Extracellular gentamicin reduces the activity of connexin hemichannels and interferes with purinergic..., Figueroa [/bib_ref] , various classes of AKs were screened against Cx43 and Cx26 HCs. These AKs manifested inhibitory effects, and further SAR investigations revealed that several 6 -and 6 -modified AKs display inhibitory activities that can be tuned toward Cx43 HCs; [bib_ref] Amphiphilic aminoglycosides with increased selectivity for inhibition of connexin 43 (Cx43) hemichannels, Subedi [/bib_ref] [bib_ref] Advances in the development of connexin hemichannel inhibitors selective toward Cx43, Subedi [/bib_ref]. The results showed that compounds 12a and 13c were most selective toward Cx43 when compared to Cx26 HCs, with calculated selectivities (Cx26 IC 50 /Cx43 IC 50 ) of ∼7 and ∼8, respectively [fig_ref] TABLE 6: IC 50 of selected 6 -and 6 -modified AKs toward Cx26 and... [/fig_ref]. These selectivities are ∼30-fold higher than that of kanamycin A (0.24). In addition, the AKs selected had no antibacterial activity and were not cytotoxic to HeLa cells at concentrations of up to 100 µM.
## Conclusions and future perspectives
In developed countries, AGs are often considered as a "last resort" for the treatment of serious bacterial infections. Semisynthetic approaches have been effective to modify naturally occurring AGs, yielding new AGs with revived antibacterial activity against AG-resistant bacteria. However, due to the rampage of bacteria equipped with diverse mechanisms of resistance such as AG-modifying enzymes, efflux pumps, decreases in membrane permeability and RNA target modifications, the semisynthetic approach still faces major challenges. The obvious problem is the cost of production. To combat the constantly evolving resistance mechanisms and complicated synthetic efforts, rare or expensive natural AGs have been employed as starting material. As a result, the end products are economically costly. Furthermore, these semisynthetic AGs still act by the same mechanisms as traditional AGs, binding to the 16S rRNA and interfering with protein synthesis, and thus, resistance evolves readily and soon after development of these new AGs. Finally, these AGs are likely to be less effective against bacteria such as Enterococcus or Pseudomonas sp., which are intrinsically resistant to AGs or equipped with flexible efflux resistance mechanisms, respectively.
AAGs offer an alternative solution to combat resistant pathogens. They can be prepared from cost effective AG feedstocks such as neomycin and kanamycin sulfates, which are mass produced by fermentation but have few clinical uses due to the resistance problem. Since AG-modifying enzymes are cellular enzymes, AAGs that target the bacterial membrane can be bactericidal without entering the cells, circumventing bacterial resistance conferred by AG-modifying enzymes. These two features, cost effective synthesis and bypassing a predominant bacterial mode of AG resistance, enable the potential uses of AAGs as effective antibacterials and for AGbased therapeutic uses previously limited by AG resistance. Additionally, AAGs exert no or much reduced cytotoxicity when compared to traditional AGs.
Highlights about AAGs in this review are that: (1) amphiphilic neomycins and neamine demonstrate potent antibacterial activities; (2) AKs have antibacterial as well as antifungal activities. The antifungal activity of AKs is of particular interest because traditional AGs are generally inactive against fungi and AKs employ novel membrane pore forming modes of action; (3) there are strong synergies between AKs and commercially used antifungal agents; (4) methods of cost-effective synthesis of lead AKs have been developed; and (5) AKs are connexin HC inhibitors. As such, AKs are potential therapeutics and experimental tools for deciphering the functions of HCs in human physiology and pathophysiology. AKs offer potential selective inhibition of HCs such as those formed by Cx43, and do not affect GJCs, thus paving the way for therapeutic use against disorders specifically due to hyperactive HCs.
Despite progress on the development and testing of AAGs, demonstration of efficacy in vivo is a crucial step toward the systemic applications of these compounds. To date, most of the reported AAGs have been studied only in vitro and a recent study showed that serum could prevent interactions between antimicrobial AAGs and plasma membranes [bib_ref] Serum prevents interactions between antimicrobial amphiphilic aminoglycosides and plasma membranes, Logviniuk [/bib_ref]. This report suggests limited use of AAGs as non-systemic drugs or for topical application. However, progress on formulations and carrier/delivery vehicles for AAGs or AAG modifications to reduce binding to serum proteins could improve clinical efficacy.
[fig] FIGURE 8 One: -step synthesis of AKs 11a and 11b. [/fig]
[table] TABLE 1: MICs (µg/mL) of 5 -modified neomycin AAGs against MRSA [/table]
[table] TABLE 3: MICs of 6"-modified amphiphilic tobramycin derivatives (in µg/mL)(Herzog et al., 2012). [/table]
[table] TABLE 4: MICs (µg/mL) of K20 and kanamycin A against bacteria and fungi a . [/table]
[table] TABLE 5 Summarized: MICs (µg/mL) of 11a and 11b b . (A) MICs of 11a and 11b against fungi (Subedi et al., 2020b) [/table]
[table] TABLE 6: IC 50 of selected 6 -and 6 -modified AKs toward Cx26 and Cx43 HCs. [/table]
|
Outcomes of Patients with Intestinal Failure after the Development and Implementation of a Multidisciplinary Team
Aim. A multidisciplinary team was created in our institution to manage patients with intestinal failure (INFANT: INtestinal FailureAdvanced Nutrition Team). We aimed to evaluate the impact of the implementation of the team on the outcomes of this patient population. Methods. Retrospective chart review of patients with intestinal failure over a 6-year period was performed. Outcomes of patients followed up by INFANT (2010INFANT ( -2012 were compared to a historical cohort (2007)(2008)(2009). Results. Twenty-eight patients with intestinal failure were followed up by INFANT while the historical cohort was formed by 27 patients. There was no difference between the groups regarding remaining length of small and large bowel, presence of ICV, or number of infants who reached full enteral feeds. Patients followed up by INFANT took longer to attain full enteral feeds and had longer duration of PN, probably reflecting more complex cases. Overall mortality (14.8%/7.1%) was lower than other centers, probably illustrating our population of "early" intestinal failure patients. Conclusions. Our data demonstrates that the creation and implementation of a multidisciplinary program in a tertiary center without an intestinal and liver transplant program can lead to improvement in many aspects of their care.
# Introduction
Intestinal failure (IF) is a comprehensive term that describes a state of malabsorption in which the intestine is unable to maintain energy, fluid, electrolyte, or micronutrient needs, leading to inadequate growth and development [bib_ref] Short bowel syndrome: how short is too short?, Goday [/bib_ref] [bib_ref] Short bowel syndrome: epidemiology and etiology, Wales [/bib_ref]. This can result from intestinal obstruction, dysmotility, surgical resection, or a congenital defect. The most common cause of IF is Short Bowel Syndrome (SBS) which is associated with a high incidence of morbidity and mortality [bib_ref] Neonatal short bowel syndrome: population-based estimates of incidence and mortality rates, Wales [/bib_ref] [bib_ref] Short bowel syndrome and intestinal failure: consensus definitions and overview, O'keefe [/bib_ref].
Large pediatric centers, with well-established intestinal transplant programs, have reported improvements in communication and coordination between services [bib_ref] Neonatal short bowel syndrome outcomes after the establishment of the first Canadian..., Diamond [/bib_ref] and improved outcomes including decreased mortality [bib_ref] Improved survival in a multidisciplinary short bowel syndrome program, Modi [/bib_ref] [bib_ref] Improved outcomes in paediatric intestinal failure with aggressive prevention of liver disease, Sigalet [/bib_ref] by using a multidisciplinary approach for children with IF, which allows for better coordination of care [bib_ref] An integrated approach to intestinal failure: results of a new program with..., Fishbein [/bib_ref] [bib_ref] The experience of a regional pediatric intestinal failure program: successful outcomes from..., Javid [/bib_ref] [bib_ref] Establishment of an intestinal rehabilitation program in an international tertiary care center, Matarese [/bib_ref] [bib_ref] Role of an intestinal rehabilitation program in the treatment of advanced intestinal..., Torres [/bib_ref]. There are only a few reports of those multidisciplinary groups and Pediatric Hospital that are not transplant centers in Canada [bib_ref] Improved outcomes in paediatric intestinal failure with aggressive prevention of liver disease, Sigalet [/bib_ref] [bib_ref] Implementation of a multidisciplinary team approach and fish oil emulsion administration in..., Sant' Anna [/bib_ref].
Our institution, the Montreal Children's Hospital (MCH), is a tertiary hospital that, although not performing liver/small bowel transplant, has a very complex population of patients. In December 2009, a multidisciplinary team (INFANT: INtestinal Failure and Advanced Nutrition Team) was developed at the MCH with the goal of improving the coordination of the complex care required by this population.
The aim of our study was to assess the impact of implementing our multidisciplinary team on the clinical outcomes of patients with IF diagnosed and followed up at our institution.
# Methods
We undertook a retrospective review of the outcomes of intestinal failure patients at our center following the implementation of this multidisciplinary team and compared those to a matched cohort followed up prior to the creation of the team.
## Intestinal failure team development.
The INFANT is composed of professionals from gastroenterology, neonatology, general surgery, nursing, nutrition, pharmacy, social work, and occupational therapy. The primary goal of this multidisciplinary team is to coordinate the highly complex care of patients diagnosed with intestinal failure. Patients are referred to INFANT when the treating neonatologist and surgeon expect intestinal failure to develop or when neonates already experience difficulty progressing enteral feeds.
The group meets monthly to establish guidelines and discuss aspects of the care of the inpatient population. Additional weekly rounds are performed by gastroenterologists and daily visits from the treating team (neonatology, surgery, or general pediatricians). Discharged patients are followed up weekly by a gastroenterologist and a nutritionist from INFANT until clinical condition is stable enough to allow longer intervals between follow-up visits.
Protocols such as ethanol lock therapy (ELT) for prevention of catheter related blood stream infections (CRBSI) and fish oil based emulsions to reverse/stabilize total parenteral nutrition (TPN) cholestasis were put in place by INFANT.
The ELT protocol was created with the purpose of providing guidelines for the safe administration of ELT for the prevention of CRBSI as well as providing teaching guidelines for nurses to educate and train families and caregivers in the safe administration of this therapy in the home setting. The ELT requires injection of each lumen of the central venous catheter (CVC) with 70% ethanol for a minimum of 4 hours and maximum of 24 hours at a frequency of three times per week. The following criteria had to be fulfilled in order for a patient to qualify for this therapy: diagnosis of intestinal failure, weight greater than 5 kg if patient has a single lumen central venous catheter (CVC) or greater than 9 kg if patient has a double lumen CVC or subcutaneous port, 3-month corrected age, PN cycled off for a minimum of 4 consecutive hours, history of two CRBSI within a defined period of time (6 months), serum ethanol level less than 2.5 mmol/L before initiation of therapy, patent CVC lumen before initiation of therapy, silicone-based CVC, and parent/caregiver consent to the use of ethanol locks.
The protocol of fish oil based emulsions to reverse/stabilize TPN cholestasis was created with the purpose of standardizing the use of fish oil based lipid emulsions at the Montreal Children's Hospital. Up until the creation of INFANT, patients diagnosed with TPN cholestasis were started on fish oil based lipid emulsions based on the decision of the treating neonatologist. The protocol established criteria for the use of fish oil emulsion, which consisted of serum direct bilirubin >50 mol/L and anticipated need for TPN for >4 weeks. Infants with liver disease secondary to cystic fibrosis, inborn errors of metabolism, and infectious hepatitis were excluded. Once consent was obtained from parents and the treatment was approved by Health Canada, a dose of 1 g/kg of fish oil based lipid emulsion was started in conjunction with 1 g/kg of conventional lipid emulsion. If there was no improvement within 4 weeks, conventional intralipid was discontinued. Fish oil based lipid emulsion was discontinued once the infant reached full enteral feeds, as well as discontinuation of parenteral nutrition.
Suggested approaches to the diagnosis and management of CRBSI and small bowel bacterial overgrowth (SBBO) were also created by INFANT and disseminated to the treating team at our institution.
Regarding diagnosis and initial management of CRBSI and management of the catheter, suggested approaches included the collection of central and peripheral blood cultures for aerobes, anaerobes, and fungi which should always be drawn in febrile or unwell TPN patients with CVC, regardless of the presence of a clinical focus. Initial antibiotic therapy should target at least coagulase negative staphylococcus, as well as gram negative rods. In the case of CRBSI involving fungi, the infected catheter should be removed within 24 hours. In the case of bacterial infections, catheter salvage measures should be undertaken including antibioticlock during periods of TPN.
Finally the suggested approach to the diagnosis and treatment of SBBO included sampling of small bowel fluid (if access is available) to obtain specific bacterial counts, with identification and sensitivities in order to guide therapy. If flora was unknown and symptoms were mild to moderate therapy was started with metronidazole 10 mg/kg/dose two to three times per day in a cycled manner, one week out of every two to four weeks. If it is clinically obvious that patient improved while on treatment but does not tolerate being off antibiotics, then alternating between two to three agents was advisable (metronidazole, gentamicin, amoxicillin-clavulanate, and cephalexin).
## Patient population.
Patients diagnosed with SBS, from 3 years before to 3 years after the creation of the multidisciplinary team, were eligible for inclusion (December 1, 2006 to November 30, 2012).
We used the Canadian Association of Pediatric Surgeons (CAPS) definition of SBS, namely, the need for parenteral nutrition (PN) for more than 42 days after bowel resection or a residual small bowel length of less than 25% expected for gestational age [bib_ref] Neonatal short bowel syndrome outcomes after the establishment of the first Canadian..., Diamond [/bib_ref].
A search was performed through our institution's medical records for patients born between December 1, 2006, and December 15, 2012, diagnosed with necrotizing enterocolitis (NEC), volvulus, gastroschisis, Hirschsprung's disease (HD), intestinal atresia, small bowel perforation, dysmotility, gastroparesis, gastric necrosis, or meconium ileus. The resulting group of 334 patients was then divided into two cohorts based on year of birth; the cut point corresponding to the date INFANT was created.
Of
## Data collection.
A retrospective chart review was conducted to obtain information on (a) demographics and clinical characteristics, (b) nutrition, and (c) morbidity and mortality, as described below.
The following information on patient demographics and clinical characteristics was abstracted: gestational age, birth weight, sex, primary diagnosis, etiology of SBS, type of surgery performed, presence of stoma, age at the time of surgery, small bowel and colon length remaining after surgery, presence of ileocecal valve (ICV), length of Neonatal Intensive Care Unit (NICU) stay and hospitalization, and length of follow-up by INFANT.
Nutritional data was obtained regarding duration of parenteral nutrition (PN) dependence, time to reach full enteral feeds if applicable, percentage of calories from PN if not fully enterally fed, home PN, gastrostomy tube insertion, and type of central venous line (CVL) if patient is on home PN.
Data on morbidity and mortality included number of septic episodes and etiology of sepsis, presence of cholestasis, maximum direct bilirubin, presence of liver failure, liver and bowel transplantation, and death and cause of death. The management strategies used to wean TPN in both groups were decided by the neonatologist and pediatric gastroenterologist in charge of that patient.
Operating room measurements of remaining small and large bowel were not available for all patients. In order to standardize the measurements, the percentages of small bowel and colon remaining after surgery were calculated by subtracting the length of intestine resected (data from pathology report available for all patients) from the average length of intestine for the patient according to the corrected gestational age at the time of the surgery [bib_ref] Normal intestinal length in preterm infants, Touloukian [/bib_ref] [bib_ref] Establishing norms for intestinal length in children, Struijs [/bib_ref].
Length of follow-up by INFANT was calculated as having time 0 being the date when the gastroenterologist saw the patient for the first time, until the last outpatient visit (when follow-up ended before the completion of the study) or the date when data collection was completed in the event that follow-up was still ongoing. Data on length of NICU and hospital stay, PN dependence, number and etiology of septic episodes, and cholestasis were obtained from electronic health record information.
Cholestasis was defined by a conjugated bilirubin level greater than 34 mol/L [bib_ref] Establishing norms for intestinal length in children, Struijs [/bib_ref] and severe cholestasis was defined as conjugated bilirubin level greater than 50 mol/L. Liver failure was defined as a combination of a level of direct bilirubin exceeding 200 mmol/L for a minimum of 2 weeks, an international normalized ratio greater than 1.5, albumin level lower than 20 g/L, thrombocytopenia with counts lower than 100.000, signs of portal hypertension, or bridging fibrosis seen on liver biopsy [bib_ref] Neonatal short bowel syndrome outcomes after the establishment of the first Canadian..., Diamond [/bib_ref].
## Data analysis and statistics.
All data was uploaded into the REDCap database system and exported into SPSS for analysis. Means of continuous variables were compared using -test and proportions using Chi square, with a two-sided alpha value of 0.05. Patients from the pre-INFANT cohort were compared with patients from the INFANT group in all analyses.
# Results
## Demographics and clinical
Characteristics. The demographic and clinical characteristics of the pre-INFANT and INFANT groups were very similar [fig_ref] Table 1: Demographics and clinical characteristics of infants with SBS [/fig_ref]. The etiology of SBS was also similar in both groups except for the incidence of NEC that was higher in the pre-INFANT cohort (74.1% versus 39.3%; = 0.009). Some patients had more than one etiology (i.e., gastroschisis and NEC). There was no significant difference in terms of number of patients on home PN (3.7% versus 7.1%, = 0.574) but the duration of PN was longer in the INFANT cohort (107.9 versus 171.6 days, = 0.006). Gastrostomy feeding tubes were more frequently used in the INFANT cohort (22.5% versus 46.4%, = 0.059) although this did not reach statistical significance [fig_ref] Table 2: Nutritional outcomes and surgical interventions of patients with SBS [/fig_ref].
## Morbidity and mortality.
With regard to infectious complications there was no difference in the number of patients with at least one septic episode (18 versus 24, = 0.096) as well as number of septic episodes per patient (2.83 in both groups). The presence of cholestasis was similar in both groups (85.2% versus 82.1%, = 0.76) whereas peak direct bilirubin was significantly lower in the INFANT cohort (181.6 versus 116.6, = 0.026). No patients were diagnosed with liver failure in either cohort. One patient in each group was assessed for bowel transplant at another center, but there were no patients transplanted at the time of data collection. There were no bowel lengthening procedures in any of the groups.
Overall mortality was lower after creation of the INFANT team (14.8% versus 7.1%, = 0.362) but this was not statistically significant. The combined outcome of sepsis, severe cholestasis, and mortality did not differ between the groups [fig_ref] Table 3: Morbidity and mortality of infants with SBS [/fig_ref].
# Discussion
The management of infants with SBS remains a major challenge. Great advancements have been made over the last few decades in the treatment of this condition. The prognosis of SBS has significantly changed with the development and subsequent refinement of parenteral nutrition, use of lipid reduction strategies and fish oil based lipid preparations, improvement of the care of central catheters, and intestinal lengthening procedures among other techniques [bib_ref] Short bowel syndrome, Donohoe [/bib_ref] [bib_ref] A multidisciplinary approach to the treatment of intestinal failure, Sudan [/bib_ref]. Despite these advancements, this patient population still suffers tremendous morbidity and mortality. Significant intellectual, emotional, and financial investments are still necessary [bib_ref] Interdisciplinary management of infantile short bowel syndrome: resource consumption, growth, and nutrition, Olieman [/bib_ref].
Our study was performed with the goal of establishing whether the development and implementation of a multidisciplinary team, to follow up patients with intestinal failure, would improve their outcomes in a tertiary hospital that does not have a transplant program.
Analysis of patient demographics and clinical characteristics demonstrates that both groups had similar etiologies except for a higher number of patients with NEC in pre-INFANT cohort (74.1% versus 39.3%; = 0.009). For reasons that are not clear, during the period of December 2006 to November 2009, there were twice as many patients diagnosed with NEC at the Montreal Children's Hospital when compared to the following 3 years.
There were a higher number of patients on home PN in the INFANT cohort (7.1% versus 3.7%; = 0.574) as well as an overall longer duration of TPN (171.6 days versus 107.9 days, = 0.006). One possible explanation could be that patients with more severe disease had improved survival after the establishment of our intestinal rehabilitation program, which can also provide intensive outpatient care. Since the number of patients in each group is relatively small, the survival of a small number of patients dependent on TPN could explain these findings.
Despite the longer duration on TPN, the INFANT cohort had a similar number of patients with at least one septic episode (85.7% versus 66.7%, = 0.096) as well as number of septic episodes per patient (mean of 2.83 episodes per patient in both groups). This could possibly be attributed to improved infection control measures or reflect the fact that patients on home PN usually see their infection rates go down compared to when they are in hospital.
As it has been well described in the literature, the use of long term TPN may lead to parenteral nutrition associated liver disease (PNALD), a condition that if left untreated may progress to cirrhosis [bib_ref] Short bowel syndrome: epidemiology and etiology, Wales [/bib_ref]. Those patients with liver failure may be left with small bowel and liver transplant as their only therapeutic option. In our study a comparable number of patients in both groups had cholestasis, but peak direct bilirubin was significantly lower in the INFANT cohort (116.6±82.8 versus 181.6 ± 124.1; = 0.026). Following the creation of INFANT, there were changes in management strategies including the early introduction of enteral feeds, standardization of the use of new omega-3 based lipid emulsions, cycling of TPN, and systematic use of prophylaxis for small bowel bacterial overgrowth, which could explain the improvement in peak bilirubin.
There was a trend towards increased use of gastrostomy tubes for enteral feeding in the INFANT group, which almost reached statistical significance (46.4% versus 22.2%, = 0.059). Gastrostomy tubes are often preferred to nasogastric tube feeding in patients requiring long term enteral feeding access [bib_ref] Pediatric enteral nutrition, Axelrod [/bib_ref]. INFANT advocates for early insertion of gastrostomy tubes, a practice which was not as systematic at our institution prior to the establishment of this team [bib_ref] A retrospective review of enteral nutrition support practices at a tertiary pediatric..., Ricciuto [/bib_ref].
Referral centers with established transplant programs report mortality rates around 30% [bib_ref] Neonatal short bowel syndrome: population-based estimates of incidence and mortality rates, Wales [/bib_ref] [bib_ref] Neonatal short bowel syndrome outcomes after the establishment of the first Canadian..., Diamond [/bib_ref] [bib_ref] Improved survival in a multidisciplinary short bowel syndrome program, Modi [/bib_ref] [bib_ref] Coordinated interdisciplinary management of pediatric intestinal failure: a 2-year review, Koehler [/bib_ref] prior to the establishment of multidisciplinary teams for the care of patients with intestinal failure. At our institution the mortality rate before INFANT was 14.8%, perhaps indicative of a less critically ill population. Despite dealing with a different population in terms of severity of disease (nontransplant center), the multidisciplinary approach to the care of patients with intestinal failure at our institution led to a 50% reduction in mortality (14.8% versus 7.1%, = 0.362). This result would need to be confirmed in a larger number of patients, as our sample size was too small to reach statistical significance or yield precise point estimate of mortality. A recent systematic review of the impact of multidisciplinary intestinal rehabilitation programs on the outcome of pediatric patients with intestinal failure documented a reduction in septic episodes and an increase in overall patient survival [bib_ref] The impact of multi-disciplinary intestinal rehabilitation programs on the outcome of pediatric..., Stanger [/bib_ref].
This data demonstrates that the creation and implementation of a multidisciplinary program in a tertiary center without an intestinal and liver transplant program can lead to improvement in many aspects of their care.
We are currently collecting prospective data on the patients followed up by INFANT in order to evaluate the long term benefits of an intestinal failure team in our population and to use this information to optimize the care and long term outcomes of these complex patients.
# Ethical approval
This retrospective cohort study of patients diagnosed with IF, using historical controls, received Research Ethics' Board approval.
[fig] Figure 1, Figure 1 182: patients born between December 1, 2006, and November 30, 2009, with one of the diagnoses listed above, 155 did not fulfill criteria for SBS, leaving 27 subjects to form the pre-INFANT cohort. Of the 152 neonates born between December 1, 2009, and December 15, 2012, with one of the above-mentioned conditions, 28 were referred to INFANT due to a suspected or confirmed diagnosis of intestinal failure and constituted the INFANT cohort (Flow chart of patient selection. INFANT = INtestinal Failure and Advanced Nutrition Team; TPN = total parenteral nutrition. [/fig]
[table] Table 1: Demographics and clinical characteristics of infants with SBS. CGA = corrected gestational age; ICV = ileocecal valve; NICU = Neonatal Intensive Care Unit; SB = small bowel; INFANT = INtestinal Failure and Advanced Nutrition Team. Results are presented as mean ± SD or (%); [ ] = of patients for that specific variable. [/table]
[table] Table 2: Nutritional outcomes and surgical interventions of patients with SBS. TPN = total parenteral nutrition; CVL = central venous line; results are presented as mean ± SD or (%); [ ] = of patients for that specific variable. [/table]
[table] Table 3: Morbidity and mortality of infants with SBS. [/table]
|
Does the pre-operative buccal soft tissue level at teeth or gingival phenotype dictate the aesthetic outcome after flapless immediate implant placement and provisionalization? Analysis of a prospective clinical case series
Background: Immediate implant placement (IIP) often is related to mid-buccal recession in literature. To draw conclusions about the behavior of the soft tissues following IIP, pre-operative aesthetic measurements have to be taken into account. The aim of analysis of these prospective clinical case series data was to elucidate whether the pre-operative buccal soft tissue level (STL) or gingival phenotype influence the 1-year pink aesthetic outcome after performing flapless immediate implant placement and provisionalization (FIIPP) maxillary incisor cases. Materials and methods: In 97 patients, a maxillary incisor was replaced performing FIIPP. STL and phenotype were analyzed on light-photographs made pre-operatively (T 0 ), direct post-operatively (T 1 ), after placement of the permanent crown (T 2 ), and 1 year post-operatively (T 3 ). To investigate if a pre-operative buccal soft tissue deficiency or excess influenced the total pink esthetic score (total-PES) per patient at T 3 , PES-3 was modified by adding a minus ("−") or plus ("+") in case of a STL-deficiency or excess, respectively.Results: Pre-operatively, 40% of the cases showed a mid-buccal recession (STL-deficiency), 19% STL-excess, while in 41% an equal level in comparison with the contra-lateral tooth was observed (STL-neutral). One year post-operatively, 79% (31/39) of the recession cases showed soft tissue gain, while STL-excess cases showed the highest rate of soft tissue reduction (94%; 17/18). This resulted in a decrease of soft tissue recessions and excesses (to 26% and 4%, respectively), and an increase of ideal STL (PES-3-score 2) to 70%. The 1-year aesthetic outcome was not statistically different (p = 0.577) between patients with a pre-operative soft tissue recession (mean T 3 total-PES = 12.18) or STL excess (mean T 3 total-PES = 11.94). Of the total population, 71 patients with a thin, and 26 with a thick phenotype were evaluated. No statistical difference (p = 0.08) was present in aesthetic outcome between the two phenotypes (thin mean T 3 total-PES = 12.30, thick mean T 3 total-PES = 11.65). Conclusion: Regardless of phenotype, preoperative soft tissue recession, or excess, comparable high aesthetic outcomes were achieved 1 year post-operatively.
Trial registration: Ethical approval was obtained and registered on 20 October 2015 (NTR5583/NL4170).
Keywords: Immediate implant placement and provisionalization, Immediate restoration, Aesthetic outcome, Modified pink esthetic score, Soft tissue level, Mid-buccal recession, Pink aesthetic score, Flapless implant surgery, IIP, IIPP
# Background
The risk of a poor pink aesthetic outcome of immediately placed implants often is related to mid-buccal recession in literature [bib_ref] Clinical and esthetic outcomes of implants placed in postextraction sites, Chen [/bib_ref] [bib_ref] A systematic review on the frequency of advanced recession following single immediate..., Cosyn [/bib_ref] [bib_ref] Esthetic outcomes following immediate and early implant placement in the anterior maxilla..., Chen [/bib_ref]. Surgical and restorative approaches, as also implant position, are pivotal in achieving an optimal aesthetic outcome, particularly on the point of the mid-buccal soft tissue level.
Hardly any of the studies on the field of the aesthetic outcome after implant therapy are comparable with each other considering heterogeneity of treatment protocols, implant position, materials, and aesthetic scores used. To evaluate the soft tissues around dental implants, in 2005, the Pink Esthetic Score (PES) was introduced by Fürhauser et al.; soft tissues are judged using seven criteria which each can be scored by 0, 1, or 2. Therefore, the total pink aesthetic score (total-PES) ranges from 0 to 14 (total-PES [bib_ref] Clinical and esthetic outcomes of implants placed in postextraction sites, Chen [/bib_ref] [bib_ref] A systematic review on the frequency of advanced recession following single immediate..., Cosyn [/bib_ref] [bib_ref] Esthetic outcomes following immediate and early implant placement in the anterior maxilla..., Chen [/bib_ref] [bib_ref] Soft and hard tissue assessment of immediate implant placement: a case series, Juodzbalys [/bib_ref] [bib_ref] Immediate implant placement postextraction without flap elevation, Chen [/bib_ref] [bib_ref] Immediate single-tooth implants in the anterior maxilla: 3-year results of a case..., Cosyn [/bib_ref] [bib_ref] Immediate implant placement and provisionalization after long-axis root fracture and complete loss..., Noelken [/bib_ref] [bib_ref] Immediate and conventional single implant treatment in the anterior maxilla: 1-year results..., Raes [/bib_ref] [bib_ref] Immediate nonfunctional loading of NobelPerfect implants in the anterior dental arch in..., Noelken [/bib_ref] [bib_ref] Soft tissue preservation and pink aesthetics around single immediate implant restorations: a..., Cosyn [/bib_ref] [bib_ref] Immediate implant placement: the fate of the buccal crest. A retrospective Cone..., Groenendijk [/bib_ref] [bib_ref] A long-term prospective cohort study on immediately restored single implants installed in..., Raes [/bib_ref]. This method is applied in many researches [bib_ref] Soft and hard tissue assessment of immediate implant placement: a case series, Juodzbalys [/bib_ref] [bib_ref] Immediate implant placement postextraction without flap elevation, Chen [/bib_ref] [bib_ref] Immediate single-tooth implants in the anterior maxilla: 3-year results of a case..., Cosyn [/bib_ref] [bib_ref] Immediate implant placement and provisionalization after long-axis root fracture and complete loss..., Noelken [/bib_ref] [bib_ref] Immediate and conventional single implant treatment in the anterior maxilla: 1-year results..., Raes [/bib_ref] [bib_ref] Immediate nonfunctional loading of NobelPerfect implants in the anterior dental arch in..., Noelken [/bib_ref] [bib_ref] Soft tissue preservation and pink aesthetics around single immediate implant restorations: a..., Cosyn [/bib_ref] [bib_ref] Immediate implant placement: the fate of the buccal crest. A retrospective Cone..., Groenendijk [/bib_ref] [bib_ref] A long-term prospective cohort study on immediately restored single implants installed in..., Raes [/bib_ref] [bib_ref] Immediate implant placement and provisionalization: aesthetic outcome 1 year after implant placement...., Groenendijk [/bib_ref] and generally preferred over the simplified-PES, introduced by Belser et al. [bib_ref] Outcome evaluation of early placed maxillary anterior single-tooth implants using objective esthetic..., Belser [/bib_ref]. This latter was launched in favor of 'ease of use and understanding' by varying the total-PES score between 1 and 10 (total-PES 1-10 ). For this goal, important information is sacrificed by merging the original criteria "alveolar deficiency," "soft tissue color," and "soft tissue texture."
To draw conclusions about the behavior of the soft tissues following IIP, pre-operative aesthetic measurements have to be taken into account. In our prospective cases series, we reported a high over-all aesthetic outcome (total-PES = 12.1) 1 year after FIIPP [bib_ref] Immediate implant placement and provisionalization: aesthetic outcome 1 year after implant placement...., Groenendijk [/bib_ref]. This analysis elucidates whether the pre-operative buccal soft tissue level (STL) or gingival phenotype influences the 1-year pink aesthetic outcome after performing FIIPP in single tooth maxillary incisor cases.
# Material and methods
In a prospective clinical case series, 100 consecutive patients were treated with flapless immediate implant placement and provisionalization (FIIPP), due to a failing maxillary incisor. The CARE reporting guidelines were used. Inclusion criteria were (1) presence of one failing single maxillary incisor in between two neighboring healthy teeth, (2) sufficient occlusal support, (3) absence of periodontal disease, (4) absence of bruxism, (5) existence of an adequate bone height at the apical area of the socket (at least 5 mm) to allow primary implant stability. Intact sockets, as also sockets with a peri-apical bone defect or a crestal bone defect not exceeding 5 mm, were included. Reasons for extraction comprised crown or root fracture, root resorption, caries, and persisting endodontic pathology. Exclusion criteria were (1) smoking habits exceeding more than 10 units a day, (2) pregnancy, (3) bone diseases or a history of irradiation, (4) ASA III or higher. Both surgical and restorative procedures were performed following a standardized protocol [bib_ref] Immediate implant placement: the fate of the buccal crest. A retrospective Cone..., Groenendijk [/bib_ref] [bib_ref] Immediate implant placement and provisionalization: aesthetic outcome 1 year after implant placement...., Groenendijk [/bib_ref].
## Pink aesthetic outcome and gingival phenotype
Both the implant and contra-lateral site were photographed in a standardized wayat different time points; pre-operatively (T 0 ), 7-14 days post-operatively (T 1 ), direct after placement of the permanent crown (T 2 ), and 1 year post-operatively (T 3 ). On each time point, two light photographs were taken: one perpendicular to the mid-buccal of the tooth arch, and one perpendicular to the implant site. Before examination, the light photographs were placed in a digital format. Evaluation of the pink aesthetic outcome was executed as described by Fürhauser et al., by two blinded examiners, who were not involved in the patient treatments. The same was true for the phenotype analysis. The inter-examiner reliability showed an intra-class correlation coefficient (ICC) of 0.979 for the PES.
## Stl measurements
For these measurements, light photographs perpendicular to the tooth arch at T 0 , T 1 , T 2 , and T 3 were used and placed into a digital format. Reference lines were drawn through, gingival margin of the contra-lateral incisor, incisal edge of contra-lateral incisor, and distal from the central and lateral incisors. The gingival margin of the failing tooth at T 0 was drawn in blue as a reference at different time points.
In order to enlighten if the pre-operative buccal STL influenced the total-PES per patient at T 3 , the PES-3 of the total pink aesthetic score (PES)was modified. The original PES-3 index by only describes a discrepancy in the STL. Whether this is positive (excess) or negative (deficiency) remains unclear. For instance, a site can show an excess of soft tissue of 1-2 mm pre-operatively and show a deficiency of 1-2 mm 1 year post-operatively; however, in both situations, the PES-3 score is 1. Our proposal is to change the PES-3 into the modified PES-3 (mPES-3); a minus ("−") is added to the score when a STL deficiency is observed, and a plus ("+") if an excess of soft tissue is present. The exact method is presented in [fig_ref] Table 1: Modified PES-3 [/fig_ref]. In case of STLdeficiency, a "minus" sign to PES-3-score was added, and in case of a STL-excess, a "plus" sign. As reference, always the contra-lateral reference tooth was used.
# Statistical methods
Total-PES at different time points T 0 , T 1 , T 2 , and T 3 of patients with a pre-operative soft tissue recession or excess, as well as biotype were compared and tested on significant difference using Levene's test for equality of variances and t test for equality of means. Statistical significance was defined as p = 0.05.
# Results
Of the 100 included patients, 97 were available for evaluation; 1 was excluded because the implant site was traumatized; the already installed implant was replaced by a new one. Another patient withdrew because of relocation. Of the third patient, the light photos were missing at T 0 . The remaining 97 patients consisted of 56 females and 41 males with a mean age of 46 years (range 17-80 years). Unfortunately, eight light photos were missing at T 1 , and seven at T 2 .
## Stl measurements
Examples of 3 cases per pre-operative STL and their modified PES-3 scores (mPES-3) at T 0 /T 3 are shown in [fig_ref] Figure 1: Three examples of cases with comparable pre-operative soft tissue level [/fig_ref]. Distribution of cases per STL-group (mPES-3 = 0 −, 1−, 2, 1+, 0+) per timepoint is shown in [fig_ref] Figure 2: Distribution of cases per STL-group [/fig_ref].
Pre-operatively, 40% of the cases showed a mid-buccal recession (STL-deficiency), 19% STL-excess, while in 41% an equal level in comparison with the contra-lateral tooth was observed (STL-neutral). One year postoperatively, 79% (31/39) of the recession cases showed soft tissue gain, while STL-excess cases showed the highest rate of soft tissue reduction (94%; 17/18). This resulted in a decrease of soft tissue recessions and excesses (to 26% and 4%, respectively), and an increase of ideal STL (PES-3 = 2) to 70%.
Comparing 1-year pink aesthetic outcome of cases with a baseline STL-deficiency (mean T 3 total-PES = 12.18) or STL-excess (mean T 3 total-PES = 11.94 ), there was no statistical difference (p = 0.577) found [fig_ref] Table 2: Comparison of mean total-PES for all points in time, based on surplus,... [/fig_ref]. Thus, whether a pre-operative STL-deficiency or STLexcess was present, it did not affect the pink aesthetic outcome 1 year post-operatively.
## Gingival phenotype
Of the total population, 71 patients with a thin phenotype and 26 with a thick phenotype were found [fig_ref] Table 3: Effect of biotype on pink aesthetic outcome 1 year post-operatively [/fig_ref]. There was no statistical difference (p = 0.079) in aesthetic outcome found between patients with a thin (mean T 3 total-PES = 12.30) or thick phenotype (mean T 3 total-PES = 11.65).
# Discussion
A difference in pink aesthetic outcome was expected between cases showing a pre-operative STL-excess, a preoperative STL-neutral, or pre-operative STL-deficiency. Especially for the latter, a lower total-PES was expected. However, the contrary appeared true; after performing FIIPP, also for the STL-deficiency cases a not significant different high total-PES score was noted. As such, the pre-operative STL at teeth did not influence the overall pink aesthetic outcome 1 year post-operatively.
In cases with a pre-operative excess of soft tissue comparing to the contra-lateral incisor, reduction of soft tissue height is required. In 15 cases, the desired STL reduction was reached to level up with the contra-lateral tooth. This was obtained by less filling of the socket with bone substitute and/or by lateral pressure of the abutment/crown onto the soft tissue during placement. It has to be taken into account that the soft tissues after socket-grafting by application of a bone substitute are less resilient. That is probably the reason that in one, case the surplus of soft tissue remained. In the two recession cases of the "excess" group (from mPES-3 T 0 = 1+ to mPES-3 T 3 = 1−), probably too much lateral pressure of the supra-structure onto the tissues caused a undesired soft tissue deficiency. Despite of this, in 100% of the cases, the same or better PES-3 score was achieved. Appliance of the mPES-3 clarified what really happened with the STL. Cases starting with a STL-neutral seemed to be most challenging to treat. In this group, the pre-operative STL is already optimal, the soft tissues have only to be preserved, however, not over-or under-contoured. A slight change of the soft tissue will result in a lesser aesthetic outcome. The post-operative recessions probably are a result of bone-substitute leakage due to post-operative bleeding, or placement of a too bulky permanent abutment causing pressure onto the surrounding hard-and soft tissues.
Surprisingly, STL gain occurred in cases were preoperatively a STL-deficiency (recession) was present. This is in confirmation with the findings of Noelken et al. [bib_ref] Clinical and esthetic outcome with immediate insertion and provisionalization with and without..., Noelken [/bib_ref]. They performed immediate implant placement and provisionalization (IIPP) on a single maxillary tooth with a pre-operative recession in 26 patients, of which 13 were treated with a connective tissue graft, and 13 without such graft. After a mean follow-up period of 45 months, recessions were significantly reduced in both groups. In another study, in which IIPP in intact sockets and defect sockets was compared, similar data on the field of total-PES score were presented [bib_ref] Gingival recession behavior with immediate implant placement in the anterior maxilla with..., Pohl [/bib_ref]. In both intact extraction sockets, as well as in alveoli with buccal bone defects, IIPP rendered similar outcomes with concern STL-deficiency, green (mPES = 2) is optimal STL-neutral, light (mPES = 1+) and dark blue (mPES = 0+) concern STL-excess regard to total-PES, height of the buccal gingival margin, and peri-implant bone level after 1 year. Although their treatment protocol is different, these authors corroborate that recession sites can improve after IIPP. Multiple other authors already stressed that, due to performing ridge preservation, a gain in alveolar height was observed [bib_ref] Ridge preservation with freeze-dried bone allograft and a collagen membrane compared to..., Iasella [/bib_ref] [bib_ref] Comparison of an allograft in an experimental putty carrier and bovinederived xenograft..., Vance [/bib_ref]. This implicates that in our study, ridge preservation was not disturbed by immediate implant placement. It is unclear how our results align with other IIPP or early and delayed placement protocol studies. In 2015, Schropp and Isodor [bib_ref] Papilla dimension and soft tissue level after early vs. delayed placement of..., Schropp [/bib_ref] published that 1 to 1.5 years after performing early or delayed implant placement, less than 60% of the cases showed an appropriate crown length, thus STL-level. In addition, they stated that early placed implants tended to be superior to delayed-placed implants with respect to STL level. So cautiously, it can be stated that performing FIIPP conform this standardized protocol [bib_ref] Immediate implant placement: the fate of the buccal crest. A retrospective Cone..., Groenendijk [/bib_ref] [bib_ref] Immediate implant placement and provisionalization: aesthetic outcome 1 year after implant placement...., Groenendijk [/bib_ref] and shows better results at the point of soft tissue level than early or delayed protocols. Further research is necessary to confirm or decline these assumptions.
A thin gingival phenotype did not affect the pink aesthetic outcome. An explanation may be that by implants' palatal positioning, a thick hard tissue crest is created.
A shortcoming of this study was that the STL was not measured in millimeter, but that a classification was used to monitor the STL, such as the PES. Unexpected was the high number of cases with a pre-operative STLdeficiency; indeed, it was possible to gain soft tissue without applying a soft tissue graft. Within the limitations of this prospective case series, we may conclude that patients with a small pre-operative STL-deficiency or STL-excess showed the same high aesthetic outcome (total-PES) as compared to cases with a pre-operative STL-neutral, 1-year post-operatively. Pre-operative midbuccal recession, as well as STL-excess cases, tended to improve, while the cases with a neutral pre-operative STL were the most difficult to obtain.
[fig] Figure 1: Three examples of cases with comparable pre-operative soft tissue level (STL) and their STL at T 3 ; STL-deficiency > 2 mm (mPES = 0−), STLdeficiency 1-2 mm (mPES = 1−), STL-neutral (mPES = 2), STL-excess (mPES = 1+), and STL-excess (mPES = 0+) [/fig]
[fig] Figure 2: Distribution of cases per STL-group (mPES-3 = 0−, 1−, 2, 1+, 0+) per time point in percentages. Red (mPES = 0−) and orange (mPES = 1−) [/fig]
[table] Table 1: Modified PES-3 (mPES-3); in case of a deficiency, a minus sign, and in case of a surplus, a plus sign is added behind the original PES-3 score [/table]
[table] Table 2: Comparison of mean total-PES for all points in time, based on surplus, or deficiency of STL at T 0 . Equal variances were assumed [/table]
[table] Table 3: Effect of biotype on pink aesthetic outcome 1 year post-operatively [/table]
|
Effects of Genetic Variants in the Nicotine Metabolism Pathway on Smoking Cessation
Background. We aimed to investigate the associations of various genetic variants in the nicotine metabolism pathway with smoking cessation (SC) in the Chinese Han population. Method. A case-control study was conducted where 363 successful smoking quitters were referred to as cases, and 345 failed smoking quitters were referred to as controls. A total of 42 genetic variants in 10 genes were selectedand genotyped. e weighted gene score was applied to analyze the whole gene effect. Logistic regression was used to explore associations of each genetic variant and gene score with smoking cessation. Results. Our study found that the variants CYP2A6 * 4, rs11726322, rs12233719, and rs3100 were associated with a higher probability of quitting smoking, while rs3760657 was associated with a lower probability of quitting smoking. Moreover, the gene scores of CYP2D6, FMO3, UGT2B10, UGT1A9, UGT2B7, and UGT2B15 were shown to exert a positive effect, while the gene score of CYP2B6 was detected to exert a negative effect on successful smoking cessation. Conclusion. is study revealed that genetic variants in the nicotine metabolic pathway were associated with smoking cessation in the Chinese Han population.
# Introduction
e usage of tobacco remains the leading cause of preventable death worldwide. It is also associated with an increased risk of various diseases, including respiratory problems, cardiovascular disorders, and cancers [bib_ref] Nicotine addiction, Benowitz [/bib_ref] [bib_ref] Impact of smoking and smoking cessation on cardiovascular events and mortality among..., Mons [/bib_ref] [bib_ref] Quantified relations between exposure to tobacco smoking and bladder cancer risk: a..., Van Osch [/bib_ref]. In humans, considerable individual variation is observed in the rate of nicotine metabolism, which contributes to differences in various smoking behaviors [bib_ref] CYP2A6-and CYP2A13-catalyzed metabolism of the nicotine d 5′(1′) Iminium ion, Von Weymarn [/bib_ref] [bib_ref] Elimination of cotinine from body fluids: disposition in smokers and nonsmokers, Haley [/bib_ref] [bib_ref] Disposition of nicotine and eight metabolites in smokers and nonsmokers: Identification in..., Kyerematen [/bib_ref]. Compared to the slow metabolizers, a larger amount of cigarette consumption is associated with rapid metabolism of nicotine [bib_ref] Known and novel sources of variability in the nicotine metabolite ratio in..., Chenoweth [/bib_ref] [bib_ref] e relationship between the nicotine metabolite ratio and three self-report measures of..., Schnoll [/bib_ref] [bib_ref] Relationship between cyp2a6 and chrna5-chrna3-chrnb4 variation and smoking behaviors and lung cancer..., Wassenaar [/bib_ref] , which leads to a higher degree of nicotine dependence [bib_ref] e relationship between the nicotine metabolite ratio and three self-report measures of..., Schnoll [/bib_ref] [bib_ref] e role of cyp2a6 in the emergence of nicotine dependence in adolescents, Audrain-Mcgovern [/bib_ref] [bib_ref] Rapid nicotine clearance is associated with greater reward and heart rate increases..., Sofuoglu [/bib_ref] , increased risk of lung cancer [bib_ref] Relationship between cyp2a6 and chrna5-chrna3-chrnb4 variation and smoking behaviors and lung cancer..., Wassenaar [/bib_ref] [bib_ref] Association between cancer risk and drug-metabolizing enzyme gene (cyp2a6, cyp2a13, cyp4b1, sult1a1,..., Tamaki [/bib_ref] [bib_ref] Genetic determinants of cytochrome p450 2a6 activity and biomarkers of tobacco smoke..., Yuan [/bib_ref] , and a lower probability of success in quitting smoking [bib_ref] Cyp2a6 slow nicotine metabolism is associated with increased quitting by adolescent smokers, Chenoweth [/bib_ref] [bib_ref] e nicotine metabolite ratio is associated with early smoking abstinence even after..., Chenoweth [/bib_ref] [bib_ref] Nicotine metabolic rate predicts successful smoking cessation with transdermal nicotine: a validation..., Schnoll [/bib_ref] [bib_ref] e nicotine metabolite ratio in pregnancy measured bytrans-3′-hydroxycotinine to cotinine ratio: characteristics..., Vaz [/bib_ref]. e metabolism of nicotine involves multiple polymorphic catalytic enzymes, whose genetic variabilities were reported to influence nicotine metabolism [bib_ref] Genetic influences on individual differences in nicotine glucuronidation, Lessov-Schlaggar [/bib_ref] [bib_ref] Ethnic variation in cyp2a6 and association of genetically slow nicotine metabolism and..., Schoedel [/bib_ref] [bib_ref] Genetic polymorphisms in human cyp2a6 gene causing impaired nicotine metabolism, Yoshida [/bib_ref] [bib_ref] e contribution of common ugt2b10 and cyp2a6 alleles to variation in nicotine..., Bloom [/bib_ref] [bib_ref] Nicotine metabolism and smoking: ethnic differences in the role of p450 2a6, Murphy [/bib_ref]. Considering the importance of nicotine metabolism in SC, we wondered whether genetic variants in the nicotine metabolic pathway contributed to the individual variations in SC. Several studies [bib_ref] Cyp2a6 slow nicotine metabolism is associated with increased quitting by adolescent smokers, Chenoweth [/bib_ref] [bib_ref] Cyp2a6 genetic variation and potential consequences, Xu [/bib_ref] have reported the relationship between gene variants in genes encoding phase I drug metabolic enzymes, especially CYP2A6, and SC. However, to our knowledge, data on the relationship between genetic variants encoding phase II drug metabolic enzymes and smoking behaviors have been lacking so far.
Hence, our study aims to investigate the associations between the genetic variants in the whole nicotine metabolic pathway and SC in the Chinese Han population.
# Methods
# Related definitions.
According to the WHO data, some of the definitions are as follows: smokers: who have/ had smoked 100 or more cigarettes (or the equivalent amount of tobacco) during their lifetime; successful smoking quitters: those who comply with the criterion of smokers but have not smoked at all continuously for two years or more during the survey period [bib_ref] Trends in smoking and quitting in China from 1993 to 2003: national..., Qian [/bib_ref] ; failed smoking quitters: those who comply with the criterion of smokers and had quit smoking but relapsed or did not smoke at all continuously for less than two years during the survey period.
## Subjects.
is community-based study was conducted among people of 17 villages belonging to three counties (Pingyin, Ju'nan, and Liangshan) of Shandong, China, between April and May 2013. Male participants aged 18 years or more who spontaneously quit smoking anytime were interviewed face to face by well-trained investigators. ey were asked to complete a questionnaire designed based on the Global Adult Tobacco Survey (GATS) Core Questionnaire with optional questions. All subjects recruited were of Han Chinese ethnicity. A total of 708 blood samples, including 363 for successful smoking quitters and 345 for failed smoking quitters, were collected successfully and genotyped. Baseline characteristics of all the subjects are provided in [fig_ref] Table 1: e characteristics of the subjects [/fig_ref].
In this study, a community-based case-control analysis was performed, where successful smoking quitters were referred to as cases, and failed smoking quitters were referred to as controls. is study was approved by the Ethics Review Committee of Shandong University, where all subjects provided informed consent.
## Selection of the genetic variants.
e genetic variants in this study were selected based on the following criteria:
(1) Identification of
## Cyp2a6 genotyping.
e presence of CYP2A6 * * 4 (whole gene deletion) was detected by a two-step allelicspecific PCR assay [bib_ref] Association between cancer risk and drug-metabolizing enzyme gene (cyp2a6, cyp2a13, cyp4b1, sult1a1,..., Tamaki [/bib_ref] [bib_ref] Characterisation and pcr-based detection of a cyp2a6 gene deletion found at a..., Oscarson [/bib_ref]. e first PCR reaction (PCR I) was performed using the primers 2Aex7F and 2A6R1 [fig_ref] Table 3: Primers used for amplification of CYP2A6 * 4 [/fig_ref]. All primers were all synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). e total reaction mixture (50 μl) included 4 μl of genomic DNA, 1.5 μl of each primer (10 μM), 25 μl of 2 × Taq PCR Master Mix (BBI), and 18 μl of ddH 2 O. e PCR program was as follows: initial denaturation step at 94°C for 4 min, followed by 35 cycles of denaturation at 94°C for 30 s, annealing at 56°C for 30 s, extension at 72°C for 2.5 min, and then a final extension at 72°C for 10 min.
Next, the allele-specific PCR reaction was performed (PCR II), which involved a PCR mixture containing 2 μL of PCR I product, 2 μL of primer 2A6ex8F (10 μM) or primer 2A7ex8F (10 μM), 2 μL of primer 2A6R2, 25 μL of 2 × Taq PCR Master Mix (BBI), and 19 μL of ddH 2 O, making up a total volume of 50 μL. e amplification was performed as per the following program: initial denaturation at 94°C for 4 min, followed by 30 cycles of denaturation at 94°C for 30 s, annealing at 52°C for 45 s, extension at 72°C for 1.5 min, and then a final extension at 72°C for 10 min. e amplified PCR products were analyzed on a 1.5% agarose gel (BBI) and stained with ethidium bromide.
As per the definition by Tamaki et al. [bib_ref] Association between cancer risk and drug-metabolizing enzyme gene (cyp2a6, cyp2a13, cyp4b1, sult1a1,..., Tamaki [/bib_ref] , the CYP2Aspecific 1,181-bp product amplified using the 2A6ex8F/ 2A6R2 primer pair alone indicated the presence of wildtype CYP2A6 (CYP2A6 non * 4/non * 4). Similarly, the product amplified using the primer pair 2A7ex8F/2A6R2 alone indicated a CYP2A6 deletion (CYP2A6 * 4/ * 4 * ). However, if one individual sample showed product amplification in both reactions, it indicated heterozygosity (CYP2A6 non * 4/ * 4).
2.6. Construction of the Weighted Gene Score. To increase the power of the test and explore the effect of the whole genes on SC, the weighted gene score of each gene was calculated as the sum of each genotype multiplied by its weight, which was then divided by the sum of the weights [bib_ref] Use of allele scores as instrumental variables for mendelian randomization, Burgess [/bib_ref] [bib_ref] Risk prediction of prevalent diabetes in a swiss population using a weighted..., Lin [/bib_ref]. e risk score was calculated as per the following equation:
[formula] weighted gene score � w 1 × SNP 1 + w 2 × SNP 2 + · · · + w k × SNP k w 1 + w 2 + · · · + w k ,(1) [/formula]
where SNP i has a value of 0, 1, or 2 according to the number of minor alleles for SNP; W k is the weight of SNP calculated using logistic regression, where smoking cessation is the dependent variable, and each variant is the independent variable; K is the number of SNPs in each gene [bib_ref] Use of allele scores as instrumental variables for mendelian randomization, Burgess [/bib_ref] [bib_ref] Using multiple genetic variants as instrumental variables for modifiable risk factors, Palmer [/bib_ref]. Since we analyzed only one genetic variant in the CYP2A6 gene (CYP2A6 * 4), we did not calculate the gene score of CYP2A6.
# Statistical analysis.
To describe the demographic characteristics of participants, the frequency and percentage of categorical variables were calculated along with the mean and standard deviation of metric variables. Additionally, Pearson's χ 2 test was used to compare the differences between the categorical variables of the case and control groups, while the Hardy-Weinberg Equilibrium (HWE) was used to compare differences among the control groups. If the variance between the groups was 2 Genetics Research homogeneous, a one-way analysis of variance was used to test the differences in means of metric variables between the groups. Otherwise, the Kruskal-Wallis equality-ofpopulations rank test was used. e associations between the genetic variants and SC were evaluated using the odds ratios (ORs) and 95% confidence intervals (CIs), which were first calculated using univariate logistic regression, and then by multiple logistic regression analysis with adjustments being done for age, occupation, education level, marital status, age of smoking onset and pack-year. e association studies were analyzed among four genetic models, including codominant, additive, dominant, and recessive models. Furthermore, the Akaike information criterion (AIC) was utilized to determine the best genetic model for each SNP.
e complete gene effect on SC was analyzed using the logistic regression, where odds ratios (ORs) and multivariate-adjusted ORs (adjusted for age, occupation, education level, marital status, age of smoking onset, and packyear) were calculated for each gene score and smoking cessation.
To estimate the value of power, we used a range of minor allele frequencies (MAF) in genetic variants, including 0.1, 0.2, 0.3, 0.4, and 0.5. We assumed an odds ratio (OR) of 1.5, the population risk of 11.7%, and the type I error rate (α) of 0.05. Our results showed that a total sample size of 708 subjects, including 363 cases and 345 controls, provided a power of more than 80% for 0.2, 0.3, 0.4, and 0.5 of MAF and more than 60% power for 0.1 of MAF under the additive inheritance model. While power calculation was performed using the program QUANTO 1.2.4, the other statistical analyses were carried out using the STATA/SE version 15.1 (Stata Corporation, College Station, TX, USA). All reported probabilities (P value) were two-sided, and a P value less than 0.05 was considered statistically significant.
# Results
## Participant characteristics.
e demographic characteristics of both the groups, including 363 successful smoking quitters (cases) and 345 failed smoking quitters (controls), are summarized in [fig_ref] Table 1: e characteristics of the subjects [/fig_ref]. Compared to the failed smoking quitters, the successful smoking quitters showed significantly higher age and shorter smoking duration (P < 0.001). Moreover, the distribution of the marital status between the case and control groups showed significant differences (P < 0.05). [fig_ref] Table 2: Genetic variants investigated in this study [/fig_ref] lists the minor allele frequency (MAF) for each genetic variant in all the subjects, along with the P value of the control subjects from the HWE test.presents the genotyping assay results obtained from the two-step allelic-specific PCR assay. e multiple logistic regression analysis, adjusted for the potential confounders, showed significant associations of CYP2A6 * 4, rs11726322, rs12233719, and rs3100 with the increased possibility of SC, as well as rs3760657 with the reduced possibility of SC. Based on the Akaike information criterion (AIC), the corresponding optimal recessive model was CYP2A6 * 4, rs11726322, and rs12233719, while the optimal dominant model was rs3760657 and the additive model was rs3100. However, no significant relationship was found between other genetic variants and SC among the four genetic models.
## Associations of genetic variants in the nicotine metabolic pathway with smoking cessation.
## Association between the gene scores and smoking
Cessation.
e results of the association between the gene scores and SC are presented in [fig_ref] Table 4: Association between the gene scores and smoking cessation [/fig_ref]. Positive significant associations were observed between the gene scores of CYP2D6, FMO3, UGT2B10, UGT1A9, UGT2B7, and UGT2B15 and SC. However, a negative association was found between CYP2B6 gene score and SC. e results
# Discussion
Smoking behavior is a complex trait with a multigenic etiology, which is influenced by both environmental and genetic factors. A genetic influence with heritability has been identified in smoking cessation (SC), which is estimated at 50-58% [bib_ref] Genetic architecture of smoking behavior: a study of Finnish adult twins, Broms [/bib_ref] [bib_ref] Genetic correlations between smoking initiation and smoking behaviors in a twin sample, Hardie [/bib_ref] [bib_ref] Latent class typology of nicotine withdrawal: genetic contributions and association with failed..., Xian [/bib_ref] [bib_ref] e heritability of failed smoking cessation and nicotine withdrawal in twins who..., Xian [/bib_ref]. In this study, we investigated the associations between various genetic variants in the nicotine metabolic pathway and SC in the Chinese Han population. Our results showed that CYP2A6 * 4, rs11726322, rs12233719, and rs3100 were associated with a higher probability while rs3760657 was associated with a lower probability of quitting smoking. Moreover, the gene scores of CYP2D6, FMO3, UGT2B10, UGT1A9, UGT2B7, and UGT2B15 were shown to exert a positive effect on successful SC, while the gene score of CYP2B6 was detected to exert a negative effect. Nicotine is primarily metabolized by the following three pathways: cytochrome P450 (CYPs)-catalyzed C-oxidation, UDP-glucuronosyltransferases (UGTs)-catalyzed glucuronidation, and flavin-containing monooxygenase 3 (FMO3)-catalyzed N-oxidation [bib_ref] Nicotine chemistry, metabolism, kinetics and biomarkers, Benowitz [/bib_ref]. In smokers, 70-80% of nicotine is converted to cotinine before metabolization to other metabolites [bib_ref] Nicotine metabolic profile in man: comparison of cigarette smoking and transdermal nicotine, Benowitz [/bib_ref]. e C-oxidation of nicotine to cotinine occurs via a two-step mechanism, where nicotine is first oxidized to the nicotine-Δ-1′ (5′)-iminium ion mediated by CYPs, which is followed by its conversion to cotinine by cytosolic aldehyde oxidase 1 (AOX1) [bib_ref] Nicotine metabolism, human drug metabolism polymorphisms, and smoking behaviour, Tricker [/bib_ref]. Subsequently, cotinine is further oxidized by CYPs to trans-3′-hydroxycotinine (3HC), which accounts for 27-40% of the nicotine dose recovered in urine. is is the main nicotine metabolite detected in the urine of smokers [bib_ref] Nicotine metabolism, human drug metabolism polymorphisms, and smoking behaviour, Tricker [/bib_ref] [bib_ref] Metabolism and disposition kinetics of nicotine, Hukkanen [/bib_ref]. CYP2A6 is the major CYP enzyme involved in the C-oxidation of nicotine. However, in some individuals, other CYPs, including CYP2B6 and CYP2D6, also contribute minorly [bib_ref] A major role for cyp2a6 in nicotine c-oxidation by human liver microsomes, Messina [/bib_ref] [bib_ref] Nicotine metabolism and cyp2d6 phenotype in smokers, Caporaso [/bib_ref] [bib_ref] Roles of cyp2a6 and cyp2b6 in nicotine c-oxidation by human liver microsomes, Yamazaki [/bib_ref]. e Genetics Research gene encoding the CYP2A6 enzyme is highly polymorphic and results in extensive interindividual variations in the CYP2A6 enzyme activity, affecting the rate of metabolism of nicotine [bib_ref] Cyp2a6 genetic variation and potential consequences, Xu [/bib_ref] [bib_ref] Cyp2a6 * 6, a novel polymorphism in cytochrome p450 2a6, has a..., Kitagawa [/bib_ref]. Nakajima et al. [bib_ref] Deficient cotinine formation from nicotine is attributed to the whole deletion of..., Nakajima [/bib_ref] first reported that the poor metabolism of nicotine to cotinine was attributed to the whole deletion of the CYP2A6 gene (CYP2A6 * 4) in humans. Compared to other ethnic populations, the CYP2A6 * 4 was reported to have a high allelic frequency in Asians (11-24%) [bib_ref] Ethnic variation in cyp2a6 and association of genetically slow nicotine metabolism and..., Schoedel [/bib_ref] [bib_ref] Characterisation and pcr-based detection of a cyp2a6 gene deletion found at a..., Oscarson [/bib_ref] [bib_ref] Interindividual variability in nicotine metabolism: C-oxidation and glucuronidation, Nakajima [/bib_ref] [bib_ref] Relationship between interindividual differences in nicotine metabolism and cyp2a6 genetic polymorphism in..., Nakajima [/bib_ref]. In this study, CYP2A6 * 4 was correlated with a higher probability of quitting smoking. However, the deviation from HWE in the control group for CYP2A6 * 4 was observed in our study; our result may be biased and needs to be further confirmed.
For in vitro C-oxidation of nicotine, CYP2B6 is responsible for an approximately 10% catalytic efficiency of the CYP2A6 enzyme [bib_ref] Nicotine dependence pharmacogenetics: role of genetic variation in nicotine-metabolizing enzymes, Ray [/bib_ref]. While CYP2A6 is primarily expressed in the liver, CYP2B6 is expressed at higher levels in the brain. Higher brain activity for CYP2B6 may cause altered sensitivity to centrally acting drugs, which may contribute to influencing the localized metabolism of nicotine in the brains of human smokers [bib_ref] Smoking, alcoholism and genetic polymorphisms alter cyp2b6 levels in human brain, Miksys [/bib_ref].
Approximately 4-7% of absorbed nicotine is excreted in the urine as nicotine N′-oxide [bib_ref] Evidence for urinary excretion of glucuronide conjugates of nicotine, cotinine, and trans-3′-hydroxycotinine..., Byrd [/bib_ref] [bib_ref] Molecular genetics of nicotine metabolism, Mwenifumbo [/bib_ref] , which is converted by flavin-containing monooxygenase 3 (FMO3) [bib_ref] Interindividual variability in nicotine metabolism: C-oxidation and glucuronidation, Nakajima [/bib_ref] [bib_ref] Metabolism of nicotine by human liver microsomes: stereoselective formation of trans-nicotine n′-oxide, Cashman [/bib_ref]. A common decrease in the function of FMO3 alleles exerts modest effects on N-oxidation activity caused by a slight modulation of protein levels and/or function and is more likely to contribute to general population variation in FMO3 [bib_ref] Population distribution of human flavin-containing monooxygenase form 3: gene polymorphisms, Cashman [/bib_ref]. In 2014, Chenoweth et al. [bib_ref] Variation in p450 oxidoreductase (por) a503v and flavin-containing monooxygenase (fmo)-3 e158k is..., Chenoweth [/bib_ref] showed that FMO3 E158K (rs2266782) modestly influenced the systemic nicotine metabolism within the CYP2A6 subgroups. However, in this study, we did not observe any significant relationship between FMO3 alleles and SC.
Nicotine, cotinine, and trans-3′-hydroxycotinine undergo further phase II detoxification reactions catalyzed by the UDP-glucuronosyltransferase (UGTs) family of enzymes via the conjugation with glucuronic acid [bib_ref] Trans-3′-hydroxycotinine o-and n-glucuronidations in human liver microsomes, Yamanaka [/bib_ref]. In smokers, 3-5% of absorbed nicotine is excreted in the urine as nicotine N-glucuronide, 16-17% as cotinine N-glucuronide, as well as an additional 8-9% as trans-3′-hydroxycotinine-O-glucuronide [bib_ref] Evidence for urinary excretion of glucuronide conjugates of nicotine, cotinine, and trans-3′-hydroxycotinine..., Byrd [/bib_ref] [bib_ref] Molecular genetics of nicotine metabolism, Mwenifumbo [/bib_ref]. e UDP-glucuronosyltransferases (UGTs) include a large family of conjugation enzymes, detoxifying a wide variety of both endogenous and exogenous substrates [bib_ref] Identification and validation of the microrna response elements in the 3′-untranslated region..., Papageorgiou [/bib_ref]. Kuehl and Murphy [bib_ref] N-glucuronidation of nicotine and cotinine by human liver microsomes and heterologously expressed..., Kuehl [/bib_ref] reported that UGT1A4 and UGT1A9, the UGTs isoforms, were responsible for nicotine and cotinine N-glucuronidation. Although glucuronidation of trans-3′-hydroxycotinine to Oglucuronide occurs mainly via UGT2B7 and UGT2B10 [bib_ref] Glucuronidation of trans-3′-hydroxycotinine by ugt2b17 and ugt2b10, Chen [/bib_ref] , it is partly glucuronidated by UGT1A9 and UGT2B15 [bib_ref] Characterization of nicotine and cotinine n-glucuronidations in human liver microsomes, Nakajima [/bib_ref]. A change in the gene encoding enzymes (i.e., point mutation, deletion, and gene conversion) responsible for chemical metabolism may lead to overproduction, underproduction, malfunction, or absence of the protein, finally resulting in alterations in the functioning of the enzyme [bib_ref] Nicotine metabolism, human drug metabolism polymorphisms, and smoking behaviour, Tricker [/bib_ref]. Many UGT gene variants were reported to influence the glucuronidation of tobacco-related compounds, including UGT1A4 [bib_ref] Variation of hepatic glucuronidation: novel functional polymorphisms of the udp-glucuronosyltransferase ugt1a4, Ehmer [/bib_ref] [bib_ref] Udpglucuronosyltransferase 1a4 polymorphisms in a Japanese population and kinetics of clozapine glucuronidation, Mori [/bib_ref] , UGT2B7 [bib_ref] Ugt1a and ugt2b genetic variation alters nicotine and nitrosamine glucuronidation in european..., Wassenaar [/bib_ref] , and UGT2B10 [bib_ref] Ugt1a and ugt2b genetic variation alters nicotine and nitrosamine glucuronidation in european..., Wassenaar [/bib_ref] [bib_ref] Glucuronidation genotypes and nicotine metabolic phenotypes: importance of functional ugt2b10 and ugt2b17..., Chen [/bib_ref].
Several studies have reported a significant impact of nicotine metabolism on the probability of success in quitting smoking [bib_ref] Cyp2a6 slow nicotine metabolism is associated with increased quitting by adolescent smokers, Chenoweth [/bib_ref] [bib_ref] e nicotine metabolite ratio is associated with early smoking abstinence even after..., Chenoweth [/bib_ref] [bib_ref] Nicotine metabolic rate predicts successful smoking cessation with transdermal nicotine: a validation..., Schnoll [/bib_ref] [bib_ref] e nicotine metabolite ratio in pregnancy measured bytrans-3′-hydroxycotinine to cotinine ratio: characteristics..., Vaz [/bib_ref]. Evidence suggests that smokers adapt their smoking behavior to maintain desired levels of nicotine in the body. e level of nicotine in the body is determined according to the intake of nicotine and the rate of metabolism by the liver. Based on the above-mentioned associations between genetic polymorphisms and enzymatic activity, we infer that genetic variants in the nicotine metabolic pathway may contribute to the individual variability in SC.
In our study, we not only analyzed single variants but also used gene scores to analyze the whole effect of the genes in the nicotine metabolic pathway on SC. A gene score is important to model multifactorial polygenic traits, particularly when the gene score consists of many common variants with small effects [bib_ref] Use of allele scores as instrumental variables for mendelian randomization, Burgess [/bib_ref]. e gene score may explain a considerable proportion of variations in the risk factor, even when none of the variants can explain it individually, thus, making it a very popular method for genetic association studies [bib_ref] Use of allele scores as instrumental variables for mendelian randomization, Burgess [/bib_ref].
One of the strengths of our study is the exploration of the relationship between the gene variants encoding phase II drug metabolic enzymes, including UGT1A9, UGT2B10, and UGT2B7, and smoking behavior. To our knowledge, data on such a relationship is rarely available so far. Most researchers have focused on genes encoding phase I drug metabolic enzymes, especially CYP2A6. Our study reveals that rs11726322 of UGT2B10, rs12233719 of UGT2B7, and rs3100 of UGT2B15 may influence the susceptibility of SC, which has not been reported in previous publications. Additionally, we employed the method of gene score for the association analysis, which increased the power and simplicity of the test.
However, we acknowledge some limitations in our study. Firstly, since smoking behaviors have a complex etiology generated by many factors, some other uninvestigated variables may alter the results. Secondly, this was a communitybased case-control study and may have some biases, including selection and recall bias, which may distort the results.
# Conclusions
is study revealed that five single variants (CYP2A6 * 4, rs11726322, rs12233719, rs3100, and rs3760657) and gene scores of CYP2B6, CYP2D6, FMO3, UGT2B10, UGT1A9, UGT2B7, and UGT2B15 might influence the susceptibility of SC in the Chinese Han population. Further examinations using larger sample size and other ethnic groups may be required to confirm our findings.
## Data availability
All data generated or analyzed during this study are included within the article and its supplementary information files.
Ethical Approval e study procedure conformed to the ethical guidelines of the Declaration of Helsinki, and approval for the study was [bib_ref] Elimination of cotinine from body fluids: disposition in smokers and nonsmokers, Haley [/bib_ref] Genetics Research obtained from the Ethics Review Committee of Shandong University.
## Consent
A written informed consent was obtained from the patients enrolled.
## Conflicts of interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.
[fig] Figure 1: (a) First-step PCR results (1,961-bp) (b) second-step PCR results (1,181-bp). Lanes 1 and 2 are wildtype CYP2A6 (non * 4/ non * 4); lanes 3 and 4 are heterozygous deletions (non * 4/ * 4); lanes 5 and 6 are the CYP2A6 deletion genotype ( * 4/ * 4). [/fig]
[table] Table 1: e characteristics of the subjects.Values are expressed as mean ± SD or frequency (%). Age group: young � age <45, middle � 45≤ age <65, old � age ≥65. Education: low � below the primary school, middle � primary and middle school, high � high school or more. Pack-year � CPD/20 × number of years smoked. CPD: cigarettes per day. [/table]
[table] Table 2: Genetic variants investigated in this study. [/table]
[table] Table 3: Primers used for amplification of CYP2A6 * 4 (gene deletion). [/table]
[table] Table 4: Association between the gene scores and smoking cessation. [/table]
|
Repetition Enhancement for Frequency-Modulated but Not Unmodulated Sounds: A Human MEG Study
Background: Decoding of frequency-modulated (FM) sounds is essential for phoneme identification. This study investigates selectivity to FM direction in the human auditory system. Methodology/Principal Findings: Magnetoencephalography was recorded in 10 adults during a two-tone adaptation paradigm with a 200-ms interstimulus-interval. Stimuli were pairs of either same or different frequency modulation direction. To control that FM repetition effects cannot be accounted for by their on-and offset properties, we additionally assessed responses to pairs of unmodulated tones with either same or different frequency composition. For the FM sweeps, N1m event-related magnetic field components were found at 103 and 130 ms after onset of the first (S1) and second stimulus (S2), respectively. This was followed by a sustained component starting at about 200 ms after S2. The sustained response was significantly stronger for stimulation with the same compared to different FM direction. This effect was not observed for the non-modulated control stimuli.Conclusions/Significance: Low-level processing of FM sounds was characterized by repetition enhancement to stimulus pairs with same versus different FM directions. This effect was FM-specific; it did not occur for unmodulated tones. The present findings may reflect specific interactions between frequency separation and temporal distance in the processing of consecutive FM sweeps.Citation: Heinemann LV, Rahm B, Kaiser J, Gaese BH, Altmann CF (2010) Repetition Enhancement for Frequency-Modulated but Not Unmodulated Sounds: A Human MEG Study. PLoS ONE 5(12): e15548.
# Introduction
To identify complex acoustic stimuli such as speech sounds, the auditory system has to process different components of the sound pattern in a fast and precise way. Recent magnetoencephalography (MEG), functional magnetic resonance imaging (fMRI) and psychophysical studies have investigated the processing of complex sounds such as animal vocalizations or human speech sounds [bib_ref] Vowel sound extraction in anterior superior temporal cortex, Obleser [/bib_ref] [bib_ref] The perception of FM sweeps by Chinese and English listeners, Luo [/bib_ref] [bib_ref] Temporal dynamics of adaptation to natural sounds in the human auditory cortex, Altmann [/bib_ref]. These vocalizations vary in a number of properties, for example in terms of amplitude, frequency, and modulation rate [bib_ref] Relationships between human auditory cortical structure and function, Hall [/bib_ref]. Among these features, fast frequency modulations play a crucial role both in human speech and in animal vocalizations [bib_ref] Anterior forebrain neurons develop selectivity by an intermediate stage of birdsong learning, Solis [/bib_ref] [bib_ref] Plasticity in primary auditory cortex of monkeys with altered vocal production, Cheung [/bib_ref]. In human speech, successful decoding of frequency variations and FM sweeps is essential for phoneme identification [bib_ref] On the role of formant transitions in vowel recognition, Lindblom [/bib_ref] [bib_ref] Role of formant transitions in the voiced-voiceless distinction for stops, Stevens [/bib_ref] [bib_ref] Separate influences of acoustic AM and FM sensitivity on the phonological decoding..., Witton [/bib_ref]. Conversely, early deficits of FM processing have been proposed to affect reading skills [bib_ref] Sensitivity to dynamic auditory and visual stimuli predicts nonword reading ability in..., Witton [/bib_ref] and reduced neuronal responses to FM stimuli in a mismatch negativity paradigm in adults have been found to correlate with reading abilities [bib_ref] Auditory event-related potentials differ in dyslexics even when auditory psychophysical performance is..., Stoodley [/bib_ref]. A major source for understanding the cerebral implementation of FM processing are electrophysiological studies in monkeys, cats and bats. Neuronal responses to FM sweeps in primary auditory cortex have been classified according to direction selectivity and modulation rate. While most neurons respond to a broad range of modulation rates and to both upward and downward FM sweeps [bib_ref] Functional organization of squirrel monkey primary auditory cortex: responses to frequency-modulation sweeps, Godey [/bib_ref] [bib_ref] Frequency-modulation encoding in the primary auditory cortex of the awake owl monkey, Atencio [/bib_ref] , selectivity for the direction of FM sweeps could be found along the tonotopic gradient in the monkey auditory cortex. Low-frequency neurons appeared to prefer upward and high-frequency neurons downward FM sweeps [bib_ref] Frequency-modulation encoding in the primary auditory cortex of the awake owl monkey, Atencio [/bib_ref] [bib_ref] Sensitivity of neurons in cat primary auditory cortex to tones and frequency-modulated..., Heil [/bib_ref].
Earlier human psychophysical studies have shown adaptation to FM direction after repeated exposure to short FM sweeps [bib_ref] Evidence for direction-specific channels in the processing of frequency modulation, Gardner [/bib_ref] suggesting dedicated channels for FM direction coding. As these adaptation effects should be observable also at the level of mass neuronal signals, we applied a two-tone adaptation paradigm to examine neuronal computation processes in the human auditory cortex. This paradigm is based upon the neurophysiological finding that stimulus repetition reduces neural activity for several seconds. This method allows identifying the stimulus selectivity and the time course of adaptation effects in certain cortical areas. Adaptation experiments have been applied both in the visual [bib_ref] fMR-adaptation: a tool for studying the functional properties of human cortical neurons, Grill-Spector [/bib_ref] and auditory system [bib_ref] Temporal dynamics of adaptation to natural sounds in the human auditory cortex, Altmann [/bib_ref] [bib_ref] Processing of sound sequences in macaque auditory cortex: response enhancement, Brosch [/bib_ref] [bib_ref] Human posterior auditory cortex gates novel sounds to consciousness, Jaaskelainen [/bib_ref] [bib_ref] Probing category selectivity for environmental sounds in the human auditory brain, Doehrmann [/bib_ref]. Besides neuronal adaptation, however, neuronal enhancement has also been observed especially in the auditory system. Two-tone experiments have revealed distinct parameter combinations leading to an enhanced neuronal response to the succeeding stimulus [bib_ref] Processing of sound sequences in macaque auditory cortex: response enhancement, Brosch [/bib_ref] [bib_ref] Sequence sensitivity of neurons in cat primary auditory cortex, Brosch [/bib_ref] [bib_ref] Long-lasting modulation by stimulus context in primate auditory cortex, Bartlett [/bib_ref].
The present study assessed whether FM direction selectivity in the human cortex can be described with MEG repetition effects. We conducted an experiment which consisted of two parts: in experiment 1 pairs of frequency-modulated sweeps were presented, while unmodulated tone pairs were used in experiment 2. FM pairs consisted of sweeps with either the same or the opposite direction of frequency modulation. In either case, stimuli had the same frequency composition. We hypothesized that selectivity for FM sweep direction in the auditory cortex would lead to differential responses for stimuli with different FM directions compared to same FM directions. This difference was expected to emerge across the N1m and the P2m components as former studies had found repetition effects in both of these components e.g. [bib_ref] Temporal dynamics of adaptation to natural sounds in the human auditory cortex, Altmann [/bib_ref] [bib_ref] Auditory evoked fields covary with perceptual grouping, Loveless [/bib_ref]. In experiment 2, pairs of identical vs. different unmodulated tones were used in order to examine the alternative explanation of whether repetition effects observed for the FM sweeps may be accounted for by their on-and offset parameters instead of being sweep-related.
Surprisingly, we did not observe a reduction of the neuromagnetic signal for same FM directions. In contrast, a repetition enhancement effect was observed for the sustained response starting at about 200 ms after S2. This response was significantly stronger for stimulation with the same compared to different FM direction.
# Results
## Experiment 1: fm sweeps
In experiment 1, we tested repetition effects for stimulus pairs with same or different FM direction. To quantify the MEG responses to the frequency-modulated sweeps, we calculated the GFP across all subjects for left-and right-hemisphere sensors. As shown in , two MEG signal components followed stimulation with S1 and S2. S1 elicited an N1m component peaking at about 113 ms after stimulus onset. For S2 the N1m component was followed by a sustained response which returned to baseline about 500 ms after S2 onset (see . Mean peak amplitudes of the N1m across left-hemisphere sensors amounted to about 39 fT (sd: 8.5 fT) (mean peak latency: 102 ms, sd: 17 ms) in response to S1, whereas over the right hemisphere peak amplitudes to S1 reached 53 fT (sd: 22.7 fT) (mean peak latency: 105 ms, sd: 16 ms). Across left-hemisphere sensors N1m peak amplitudes in response to S2 amounted to about 54 fT (sd: 24.7 fT) (mean peak latency: 132 ms, sd: 17 ms), whereas across the right-hemisphere sensors they reached 68 fT (sd: 29 fT) (mean peak latency: 130 ms, sd: 15 ms). As we were interested in repetition effects, we analysed the N1m in response to the second stimulus. N1m peak amplitudes were calculated for each subject using a time window of 50 ms (100-150 ms after S2 onset). Employing a repeated measurement analysis of variance (ANOVA) with the factor hemisphere (right/left) and repetition (same/different), no differences across hemispheres (F1,9 = 4.02, P = 0.08) and for repetition (F1,9 = 0.35, P = 0.52) or interactions were observed. Mean N1m peak latencies in response to the second tone presentation (same vs. different) also did not differ between hemispheres or conditions. To compare the peak latencies in response to S1 and S2 we applied an ANOVA with the factors hemisphere (right/left), position (S1/S2) and repetition (same/ different). Mean peak latencies in response to the second stimulus were significantly longer. We found a significant main effect for position (F1,9 = 31.91, P = 0.00) but no effects for hemisphere (F1,9 = 1.81, P = 0.21) or repetition (F1,9 = 1.05, P = 0.33).
To investigate repetition effects at the source level, two regional symmetric sources were fitted to each subjects' average evoked magnetic field (average across all conditions) across a time interval of 100-150 ms after S2 onset. This source model was used to explain the magnetic signal in each condition and to calculate source waveforms for each individual subject and each condition. Averaged Talairach coordinates were: (x, y, z) = 242, 217, 13 mm (sd: 7, 11, 6). As shown in , 'same' and 'different' conditions were combined and compared by applying a bootstrapping procedure. Based on this bootstrapping statistics significant differences across right-hemisphere sensors (p,0.001, uncorrected for multiple comparisons) appeared at about 150 ms and lasted until 350 ms after S2 onset. Different results were found for left-hemisphere sensors where significant differences (p,0.001, uncorrected for multiple comparisons) only appeared at 200-300 ms after the onset of S2.These analyses revealed repetition enhancement for same compared to different FM sweeps. This effect was more pronounced in the right-hemisphere. This may suggest that the right-hemisphere plays a special role in this process.
As no differences between 'same' and 'different' conditions were observed for the N1m component but were evident for the later sustained response at 200-300 ms after the second stimulus presentation, we fitted two symmetric sources to this sustained activity. The averaged Talairach coordinates were: (x, y, z) = 241, 215, 14 mm (sd: 9, 11, 5). These two models do not differ significantly (F1,9 = 0.81, P = 0.39). Thus, the N1m source model would also show some validity for the later sustained activity.
## Experiment 2: unmodulated tones
To control whether repetition effects for same versus different FM directions could be accounted for by the on-and offset parameters of the employed stimuli, we conducted experiment 2 with a similar design as experiment 1, but using unmodulated tones. As shown in , GFP of the evoked magnetic fields elicited by S1 presentation showed an N1m component starting at about 77 ms after S1 onset. S2 was also followed by an N1m component which returned to baseline at about 200 ms after S2 onset. Peak amplitudes of the N1m in response to the first unmodulated tones across left and right-hemisphere sensors amounted to about 55.7 fT (sd: 15.5) with a mean peak latency of 112 ms, sd: 15 ms and 74.8 fT (sd: 38.1) (mean peak latency: 114 ms (sd: 15 ms)), respectively. N1m peak amplitudes in response to S2 amounted to about 62 fT (sd: 22) with a mean peak latency of 124 ms (sd: 20 ms) and 80 fT (sd: 22.6) (mean peak latency: 122 ms, sd: 16 ms) for left-and right-hemisphere sensors, respectively. A repeated measurement ANOVA was conducted for mean peak amplitudes with the factors hemisphere (right/left) and position (S1/S2) and repetition (same/different). No significant differences between left and right-hemisphere (F1,9 = 4.73, P = 0.06), or in response to the first or the second stimulus (F1,9 = 1.58, P = 0.24) and for repetition (F1,9 = 0.24, P = 0.63) could be found. Analysing peak latencies, we found significantly longer latencies in response to the second tone (F1,9 = 7.39, P = 0.02), but no significant difference between right and lefthemisphere (F1,9 = 1.02, P = 0.76) and no repetition effect (F1,9 = 0.00, P = 0.94).
To model the evoked magnetic signals of the unmodulated tones, all conditions were averaged and two symmetric regional sources were fitted. For each subject this model (mean approximated Talairach coordinates: (x, y, z) = 236 -16, 10 (sd: 11, 13, 7)) was used to compare the repetition effect for unmodulated tones with the repetition of FM sweeps. 'Same' and 'different' conditions in experiment 2 were combined and the difference waveform was calculated. reveals that in contrast to the difference waveforms computed for the FM sweeps in experiment 1, no significant differences were observed after S2 onset. The analysis of the unmodulated tones showed neither enhancement nor inhibition effects in response to S2. This was true for both hemispheres. Similar to experiment 1, the source waveforms showed significant differences between the right and the lefthemisphere suggesting a stronger involvement of the righthemisphere in the present type of auditory processing.
To compare repetition effects for the modulated tones with those for the unmodulated tones we calculated the difference waveforms between modulated and unmodulated conditions for each subject (see . These difference waveforms were compared using a bootstrapping procedure. For left-hemisphere sensors significant differences (p,0.001, uncorrected for multiple comparisons) appeared at 200-300 ms and 400-500 ms after the onset of S2. Across right-hemisphere sensors significant differences (p,0.001, uncorrected for multiple comparisons) between modulated and unmodulated tones appeared at about 40-80 ms and at 130-550 ms after S2 onset.
# Discussion
This study used a two-tone paradigm to investigate selectivity for FM sweep direction in the human auditory cortex. Repetitive stimulation of selective neural populations has been proposed to result in suppression effects due to their refractoriness [bib_ref] The slow response of the human cortex to auditory stimuli: recovery process, Davis [/bib_ref] [bib_ref] The N1 wave of the human electric and magnetic response to sound:..., Naatanen [/bib_ref]. In consideration of previous reports that have described FMdirection-selective neurons in auditory belt cortex in non-human primates [bib_ref] Processing of frequency-modulated sounds in the lateral auditory belt cortex of the..., Tian [/bib_ref] , we had expected MEG response decrements after repeated stimulation with the same FM tone. Surprisingly, we found an enhancement of the sustained magnetic field response after FM sweep repetition. These repetition enhancement effects occurred for same compared with different FM directions in experiment 1 but not for the comparison of unmodulated tones with same versus different frequency in experiment 2. The MEG signal enhancement was observed during a time interval between 200-300 ms after S2 onset over the left hemisphere and between 150-350 ms after the second sound presentation across righthemisphere sensors. Furthermore we observed stronger responses in right-hemisphere components about 40-80 ms after S2 onset for difference waveforms of modulated tones compared to unmodulated tones. In the auditory system both neuronal response decrements and enhancements in response to repeated stimuli have been reported. In particular, response enhancement has been found in the macaque auditory cortex when a tone was preceded by another tone at a short interval of 70-300 ms [bib_ref] Processing of sound sequences in macaque auditory cortex: response enhancement, Brosch [/bib_ref] [bib_ref] On the prediction of sweep rate and directional selectivity for FM sounds..., Brimijoin [/bib_ref]. Furthermore, ERP response enhancements have been shown in humans in response to sequences of 1000-Hz tone pairs at random stimulus-onset asynchronies (SOA) between 100 and 1000 ms [bib_ref] Facilitation of the N1 peak of the auditory ERP at short stimulus..., Budd [/bib_ref]. The strongest enhancements of the N1 peak were found for the shorter SOAs of 100-300 ms. However, several studies have found response decrements to repeated tones in terms of psychophysical detection thresholds [bib_ref] Evidence for a reappraisal of the psychophysical selective adaptation paradigm, Moody [/bib_ref] , the N1 event-related potential [bib_ref] Decrement of the N1 auditory event-related potential with stimulus repetition: habituation vs...., Budd [/bib_ref] and at a later stage for the P2m component acquired with MEG [bib_ref] Temporal dynamics of adaptation to natural sounds in the human auditory cortex, Altmann [/bib_ref].
Trying to determine the principles or even mechanisms underlying response enhancement to repeated presentation of identical FM stimuli (i.e with the same modulation direction) is difficult as even the representation of single FM tones has only been studied in very few investigations using electrophysiological or MEG techniques [bib_ref] Human auditory cortex responses to rising versus falling glides, Pardo [/bib_ref] [bib_ref] Tonotopic organization of the human auditory cortex probed with frequency-modulated tones, Weisz [/bib_ref]. This makes it hard to provide an substantial explanation that rests on previous empirical findings and goes beyond speculation.
Single-neuron studies have located the strongest interactions between consecutive tones to the cortical level [bib_ref] Tone-sequence analysis in the auditory cortex of awake macaque monkeys, Brosch [/bib_ref] and a number of behavior-lesion studies in animals have suggested that tonesequence analysis critically relies on the auditory cortex e.g. [bib_ref] Neural representation of sound patterns in the auditory cortex of monkeys, Brosch [/bib_ref]. While repetitive stimulation mostly leads to response attenuation [bib_ref] Interstimulus interval dependence of the auditory vertex response and its magnetic counterpart:..., Hari [/bib_ref] , several cases of response facilitation have been reported as well, mainly in long sequences with randomly varying interstimulus intervals (ISIs). Such facilitation was found for the N1m component and seemed to depend to a great extent on the repetition rate [bib_ref] Evoked responses of human auditory cortex may be enhanced by preceding stimuli, Loveless [/bib_ref]. The ISI used in the present study (300 ms between stimulus onsets) is at the long end of the range where facilitatory interactions between auditory stimuli were found in these studies [bib_ref] The enhancement of the N1 wave elicited by sensory stimuli presented at..., Wang [/bib_ref].
We suggest that the MEG signal enhancement effect for repetitions of identical FM sweeps may result from the interaction of two parameters, frequency separation and temporal distance between repetitions. In central auditory neurons, changes in activity across successive presentations of two sounds have been found to depend on the similarity of their frequencies: tones of similar frequency were attenuated while responses to tones with clearly deviating frequency were facilitated (see [bib_ref] Tone-sequence analysis in the auditory cortex of awake macaque monkeys, Brosch [/bib_ref] , . The repeated FM stimuli in this study might have mimicked the twotone pattern necessary for facilitation in such a representation. Frequencies that were present at the end of the first stimulus were followed by maximally different frequencies at the onset of the second tone. That is, at a relatively short temporal distance, frequency separation was high, and may have resulted in enhancement effects. Also, similar frequencies were only present at a long temporal distance. In contrast, in stimulus pairs with the opposite sweep direction, the end of the first and the onset of the second tone were composed of identical frequencies, and stronger frequency separation was present only at a longer temporal distance. Thus, in non-repeated sweeps, frequency-wise responses may have primarily been attenuated. As a result, the observed pattern of relative response enhancement for repeated vs. nonrepeated sweeps may have resulted.
Based on the same assumptions one might expect an enhancement for nonrepeated unmodulated tones (high/low; low/high conditions). There are however two differences to the FM sweeps that may explain why we did not observe these effects in our data. First, unmodulated tones may activate suppression along their temporal extent as, in contrast to FM sweeps, their frequency composition does not change over time. Second, even in the 'different tone' condition, only one of four frequencies that each complex tone was made up from was actually changed. Thus, in the complex tones used in this study, suppression due to similarity may have outweighed the potential enhancement effects of frequency separation. Therefore, an interaction of frequency separation and temporal distance may explain both our results in FM and unmodulated tones. It is an interaction model as the effects of frequency separation strongly depend on temporal relations.
While this scenario can account for our results, it remains speculative. However there is some support for it from the literature. As described recently, repetitive FM components in a steady-state tone can lead to increased MEG activity at a . Grand average regional source waveforms. The regional sources in response to FM sweeps of same or different modulation direction have mean Talairach coordinates of (x, y, z) = 242, 216, 13 mm (sd: 7, 11, 6) (upper graphs). An example of the position for the regional sources repetition rate around 4 Hz that is well compatible to the ISI used here [bib_ref] Spatiotemporal reconstruction of the auditory steady-state response to frequency modulation using magnetoencephalography, Millman [/bib_ref]. In addition, the structure of spectral response areas as noted above [bib_ref] Tone-sequence analysis in the auditory cortex of awake macaque monkeys, Brosch [/bib_ref] is directly related to the mechanisms underlying FM selectivity, as they were investigated in cortical neurons [bib_ref] Neural mechanisms underlying selectivity for the rate and direction of frequency-modulated sweeps..., Razak [/bib_ref]. In summary, there is some support for frequency separation and temporal distance as necessary components leading to the specific enhancement in the same FM configuration.
In addition, we found a stronger response for the difference waveforms of the FM sweeps compared to the unmodulated tones. This difference appeared across the right hemisphere at 40-80 ms after stimulus presentation. One possible explanation for this finding could be the special role of the right hemisphere in the processing of frequency-modulated sounds. Both lesion studies in animals e.g. [bib_ref] Right auditory cortex lesion in Mongolian gerbils impairs discrimination of rising and..., Wetzel [/bib_ref] and studies in epileptic patients [bib_ref] Functional imaging of the auditory system: the use of positron emission tomography, Johnsrude [/bib_ref] have demonstrated significant decreases in direction selectivity of FM tones when the right hemisphere is affected but not when the left hemisphere is lesioned. Also imaging studies have shown a stronger right-hemisphere involvement in an FM direction discrimination task [bib_ref] Auditory lexical decision, categorical perception, and FM direction discrimination differentially engage left..., Poeppel [/bib_ref] [bib_ref] Contralateral White Noise Selectively Changes Right Human Auditory Cortex Activity Caused by..., Behne [/bib_ref]. This specialization of the right auditory cortex could account for the stronger response to modulated than unmodulated sounds.
In summary our results suggest enhancement effects for tone pairs with similar FM direction and short ISIs (200 ms) but not for unmodulated tones. We hypothesize that this effect results from an interaction between frequency separation and temporal distance between the consecutively presented sounds. To corroborate this hypothesis, further studies are needed that systematically manipulate stimulus timing and frequency differences and test whether MEG enhancement effects are linked to behavioral facilitation.
# Materials and methods
## Subjects
Participants were 10 healthy, right-handed adults (4 males, mean age 27). All subjects had normal hearing abilities as determined by self-report and reported no history of otological, neurological or psychiatric disease. Each subject gave written informed consent to participate in the study. The study was performed in accordance with the ethical standards laid down in the declaration of Helsinki of 1964. It was approved by the local ethics committee of the Goethe University Medical Faculty.
## Stimuli
Stimuli were created using MATLAB (The MathWorks, R2007a), with a sampling frequency of 44.1 kHz and a duration of 100 ms. All stimuli were shaped by rising and falling 5 ms ramps and were presented via insert earphones (E-A-R-tone 3A, Aearo Corporation, Indianapolis, USA) binaurally at a comfortable intensity of approximately 80-85 dB(A). In experiment 1, 20 different upward and 20 downward logarithmic frequencymodulated sweeps with a modulation rate of 10 octaves per second were used (1 octave per 100 ms). The sounds consisted of four sinusoidal components with frequencies which were separated by one octave. The different FM sweeps started at different frequencies, with the lowest rising FM sweep starting at 187.5, 375, 750 and 1500 Hz rising to 375, 750, 1500 and 3000 Hz, respectively. Each FM sweep differed in 1/20 steps of a octave from the next higher and lower FM sweep, respectively. Thus, the calculated for the FM sweeps for one subject is illustrated in the center of figure 2. Grand average regional source waveforms in response to same and different unmodulated tones (mean Talairach coordinates of the sources: (x, y, z) = 236. 216, 10 (sd: [bib_ref] Auditory event-related potentials differ in dyslexics even when auditory psychophysical performance is..., Stoodley [/bib_ref] [bib_ref] Frequency-modulation encoding in the primary auditory cortex of the awake owl monkey, Atencio [/bib_ref] [bib_ref] On the role of formant transitions in vowel recognition, Lindblom [/bib_ref] are shown in the lower graphs. Difference waveforms are shown in blue and red for the left and right-hemisphere, respectively. The confidence range, obtained with the bootstrapping procedure, is plotted in grey. Significant differences are found mainly across right-hemisphere sensors in response to the second FM sweep at 150-350 ms after S2 onset (p,0.001, uncorrected for multiple comparisons). There were no significant differences in response to the unmodulated tones. Dotted grey lines in each of the four graphs indicate the beginning and ending of the stimuli. doi:10.1371/journal.pone.0015548.g002 . Source difference waveforms between modulated and unmodulated tones. Difference waveforms from the source models for the differences between same-different modulated and same-different unmodulated tones are shown in blue and red for the left and righthemisphere, respectively. The confidence range, obtained with the bootstrapping procedure, is plotted in grey. Significant differences are found across right-and left-hemisphere sensors in response to the second stimulus at 150-300 ms after S2 onset (p,0.001, uncorrected for multiple comparisons). In addition, the same-different waveforms for modulated and unmodulated tones are depicted in purple and green, respectively. Dotted grey lines in each of the four graphs indicate the beginning and ending of the stimuli. doi:10.1371/journal.pone.0015548.g003 highest rising FM sweep started at 362, 724, 1449 and 2898 Hz and rose to 724, 1449, 2898 and 5796 Hz, respectively. We applied a logarithmic Gaussian filter with a mean of 1050 Hz and a standard deviation of 0.59 octaves (mean-sd: 698 Hz; mean+sd: 1581 Hz) for a smooth fade-out of the lowest and highest frequency complexes. These stimuli were thus a short, continuous, and glissando-like version of the Shepard's illusion [bib_ref] Circularity in judgments of relative pitch, Shepard [/bib_ref] with the advantage that the frequency transgression from S1 to S2 was similar for sweep pairs with same and different FM directions. The experimental conditions in experiment 1 consisted of FM sweep pairs. The two sound stimuli of each pair always covered the same frequency range, that is the second stimulus of a pair was either identical to the first (''same'' conditions) or a time-reversed version of the first stimulus (''different'' conditions). In the ''upward same'' condition, an FM sweep ascending in frequency was presented twice. Similarly, in the ''downward same'' condition a descending FM sweep was presented twice. In the ''different up/down'' conditions, an ascending/descending S1 was followed by a timereversed version of S1 (see [fig_ref] Figure 4: Schematic illustration of the stimuli used in both parts of the experiment... [/fig_ref].
In experiment 2, 40 unmodulated sounds consisting of four sinusoidal components that contained frequencies separated by one octave. As shown in [fig_ref] Figure 4: Schematic illustration of the stimuli used in both parts of the experiment... [/fig_ref] these frequencies were consistent with the start and end frequencies of the complex FM sounds. Two different types of stimulus pairs were presented for this experimental part: in the ''same'' conditions, the same complex stimulus was presented twice, either ''low/low'' or ''high/high''. In the different conditions, S1 was followed by a complex tone S2 without the lowest but an additional higher component of S1 (low/high) or without the highest but an additional lower component (high/low). The added component was always separated by 1 octave from the lowest/highest component of S1 (see [fig_ref] Figure 4: Schematic illustration of the stimuli used in both parts of the experiment... [/fig_ref].
## Procedure
The MEG experiment consisted of two parts. In experiment 1 (three runs) we investigated the processing of FM tones, while in experiment 2 (two runs) unmodulated tones were employed. Each run consisted of 361 trials and had a duration of six minutes. Subjects thus completed a randomized sequence of five runs which were separated by short breaks. Trials in both parts consisted of a two-tone paradigm following the same structure: the first stimulus (S1, 100 ms) was followed by an inter-stimulus interval (ISI, 200 ms) and the second stimulus (S2, 100 ms) [fig_ref] Figure 4: Schematic illustration of the stimuli used in both parts of the experiment... [/fig_ref]. Trials were separated by a silent inter-trial interval of 600-900 ms. Subjects were instructed to watch a silent movie during both parts of the experiment. Experiment 1 served to investigate frequency direction selectivity employing pairs of FM sounds of either same or different frequency modulation direction. We presented either a) two identical upward FM sweeps, b) two identical downward FM sounds ('same' conditions), c) an upward followed by a downward FM sound (i.e., the identical sound played in reverse order) or d) a downward followed by an upward FM sound ('different' conditions) [fig_ref] Figure 4: Schematic illustration of the stimuli used in both parts of the experiment... [/fig_ref]. All conditions were randomized across trials.
In experiment 2, pairs of unmodulated tones were presented instead of FM sweeps. The stimuli were paired according to the following conditions: a) two identical unmodulated 'high' tones, b) two identical unmodulated 'low' tones ('same' conditions), c) an unmodulated 'high' tone followed by a unmodulated 'low' tone and d) a unmodulated 'low' tone followed by a unmodulated 'high' tone ('different' conditions) [fig_ref] Figure 4: Schematic illustration of the stimuli used in both parts of the experiment... [/fig_ref].
## Meg acquisition and data analysis
The neuromagnetic signals were recorded using a whole-head MEG system (CTF-MEG, VSM MedTech Inc., Coquitlam, Canada) with 275 magnetic gradiometers with an average distance between the sensors of 2.2 cm. The signals were recorded at a sampling rate of 600 Hz. Before MEG recordings, three head position indicator coils were placed at the nasion and the preauricular points and the head position was determined at the beginning and the end of each recording to ensure that head movements did not exceed 0.5 cm.
For both parts of the experiment, the MEG signals for the four conditions in experiment 1 and the two conditions in experiment 2 were averaged separately. The averaging epoch ranged from 500 ms before S1 to 1000 ms after S1 onset. A prestimulus period of 100 ms before S1 served as baseline. The data were low-pass filtered with a cutoff at 30 Hz before averaging. Epochs which contained signal variations larger than 3.5 pT were excluded from averaging. This procedure left on average 94% of the trials for further analysis. Averaged event-related fields (ERF) of all sensors for the 'same' and 'different' conditions are shown in [fig_ref] Figure 5: Grand-averaged event-related fields [/fig_ref]. For a first inspection, data were averaged across subjects and combined by calculating the global field power (GFP) [bib_ref] Reference-free identification of components of checkerboard-evoked multichannel potential fields, Lehmann [/bib_ref]. The GFP was calculated by the root mean square over all right-and left-hemisphere sensors and time points for each condition. Singlesubject and grand average auditory evoked potentials were computed with the BESA 5.2 software package (MEGIS software, Grä felfing, Germany).
Source locations and time courses of source activities were computed for the average of all four 'same' and 'different' conditions for each subject separately. In experiment 1 two symmetric regional sources located in the superior temporal lobes were used to model the evoked magnetic field. The two regional sources were calculated for the N1m of the second stimulus across a time range of 100-150 ms after S2 onset. These sources were used to model the evoked magnetic fields in each condition. To test the quality of this model, goodness-of-fit (GOF) values were calculated for each condition and subject separately. Mean GOF values amounted to 85,8% (sd: 7.3) for the 'up/up' condition, 78.7% (sd: 5.3) for the 'down/down' condition, 79.7% (sd: 5.6) for the 'up/down' and 82.6% (sd: 8.8) for the 'down/up' condition. A similar data analysis was performed for the unmodulated tones used in experiment 2. Two symmetric regional sources were calculated for the averaged conditions for the N1m of S2 (time range 100-150 ms). These sources modeled 81.1% (sd: 5.9) of the evoked magnetic field of the 'high/high' condition, 76.4% (sd: 17.8) of the 'low/low' condition, and 77.1% (sd: 10.3) and 79.2% (sd: 14.4) of the 'high/low' and 'low/high' conditions, respectively.
To compare conditions, source waveforms were calculated for each subject and condition. For each subject the difference between the source waveforms of the 'same' and 'different' conditions was calculated. The rationale for comparing the 'same' and 'different' conditions rather than S1 and S2 was to overcome the ERF distortion due to the short ISI. S1 and S2 were temporally closely adjacent, resulting in a strong influence of S1 on S2. However, the ''same'' and ''different'' conditions share the same stimulus history up to S2 and thus a comparison should be unbiased by previous stimulation (for a discussion of event-related potential (ERP) distortions due to adjacent stimuli and possible solutions see [bib_ref] Distortion of ERP averages due to overlap from temporally adjacent ERPs: analysis..., Woldorff [/bib_ref]. To test for significant differences, a nonparametric bootstrapping procedurebased on 1000 iterations was applied.
[fig] Figure 1 g001, Figure 2: Evoked magnetic responses plotted as global field power (GFP). (A) GFP evoked by the frequency-modulated sweeps averaged across the left-and right-hemisphere sensors. (B) GFP evoked by unmodulated tones over the left-and the right-hemisphere sensors. doi:10.1371/journal.pone.0015548. [/fig]
[fig] Figure 4: Schematic illustration of the stimuli used in both parts of the experiment and experimental conditions. (A) All stimuli consisted of four harmonic components indicated by the black bars, and were filtered with a Gaussian band-pass as symbolized by the Gaussian curve at the ordinate. The lighter gray shades of the lower and higher components indicate reduced sound intensity due to the filtering, see section on Stimuli for further details. The central sketch depicts an ascending FM sweep as used in experiment 1, the left and right sketches show the nonmodulated stimuli used in experiment 2. (B) experimental conditions and procedure for the frequency-modulated sweeps, and (C) unmodulated tones. doi:10.1371/journal.pone.0015548.g004 [/fig]
[fig] Figure 5: Grand-averaged event-related fields (ERF). ERFs for all subjects and sensors for the 'same' (on the left) and 'different' (on the right) conditions are shown. In parts (A) and (B) of the figure the raw data are shown for the frequency-modulated tones in experiment 1, and (C) and (D) show the ERF for the non modulated-tones in experiment 2 of the study. Dotted grey lines in each of the four graphs indicate the beginning and ending of the stimuli. doi:10.1371/journal.pone.0015548.g005 [/fig]
|
Bottom-up assembly of viral replication cycles
Bottom-up synthetic biology provides new means to understand living matter by constructing minimal life-like systems. This principle can also be applied to study infectious diseases. Here we summarize approaches and ethical considerations for the bottom-up assembly of viral replication cycles.
## Bottom-up engineering of life-like systems
Bottom-up synthetic biology has opened up new avenues for the construction of artificial systems that recreate the structure and function of living cells. New fundamental knowledge about the mechanisms underlaying life, such as the physical principles of cell division 1 , cellular motility 2 , cell communication [bib_ref] Vesicle induced receptor sequestration: mechanisms behind extracellular vesicle-based protein signaling, Staufer [/bib_ref] and morphogenesis 4 has been achieved by the application of in vitro reconstitution approaches. In the present context, we define bottom-up synthetic biology as a field that strives to construct minimal life-like materials by bottom-up reconstitution of cellular phenomena in vitro. Towards this, biological and artificial building blocks are applied to create molecular structures that feature functions inherent to life. Historically, and driven by its emergence from the physical sciences, the field has focused on the construction of synthetic cells that recapitulate the constitutional characteristics of natural cells (e.g. metabolism [bib_ref] Cell fuelling and metabolic energy conservation in synthetic cells, Sikkema [/bib_ref] , division 1 , evolution 7 , information processing 8 , compartmentalization [bib_ref] Bottom-up assembly of target-specific cytotoxic synthetic cells, Bücher [/bib_ref]. In its essence, bottom-up synthetic biology redefines the algorithm of conventional biological research, that follows an observe-describe-understand concept, by studying living systems in a design-build-understand manner. This learning-by-building strategy is empowered by reducing the molecular complexity of cellular phenomena, allowing for a step-wise deconstruction of life´s most fundamental processes. Along this line of research, several new tools have been developed to engineer artificial cells, for instance microfluidic technologies to build and manipulate lipid vesicles serving as synthetic cell scaffolds, and advanced genetic engineering strategies to create synthetic cells that exchange information with their environment [bib_ref] Synthetic cells with self-activating optogenetic proteins communicate with natural cells, Adir [/bib_ref] [bib_ref] Microfluidic production and characterization of biofunctionalized giant unilamellar vesicles for targeted intracellular..., Staufer [/bib_ref]. In this regard, the ever-increasing molecular toolbox for in vitro reconstitution of cellular phenomena has greatly accelerated progress in the field 11 , bringing the ultimate goal, crafting of a truly living synthetic system, within reach.
Apart from research efforts directed towards unrevealing the fundamental principles of cellular life forms, several studies have provided an initial demonstration of how the reductionistic designbuild-understand approach can be applied to biomedical research objectives [bib_ref] Can bottom-up synthetic biology generate advanced drug-delivery systems?, Lussier [/bib_ref] [bib_ref] Artificial cells drive neural differentiation, Toparlak [/bib_ref] [bib_ref] Bottom-up assembly of targetspecific cytotoxic synthetic cells, Hernandez Bücher [/bib_ref] [bib_ref] Polymer-based porous microcapsules as bacterial traps, Luo [/bib_ref]. In recent years, studies based on minimal biological systems have brought forward synthetic cells designed for therapeutic purposes. These include synthetic cells capable of controlled production and release of insulin upon exposure to increased glucose concentrations in vivo [bib_ref] Synthetic beta cells for fusion-mediated dynamic insulin secretion, Chen [/bib_ref] and synthetic cells that autonomously produce therapeutic proteins within tumors [bib_ref] Synthetic cells synthesize therapeutic proteins inside tumors, Krinsky [/bib_ref]. These advances showcase the potential of bottom-up synthetic biology in biomedical applications and inform research that could empower completely new therapeutic agents. Moreover, this also demonstrates how artificial biological systems can be applied to decipher molecular mechanisms underlying disease states. Importantly, such engineering strategies are based on a fundamental principle of synthetic biology: modularity. Researchers in bottom-up synthetic biology, design individual cellular modules (e.g. compartments [bib_ref] Bottom-up assembly of functional intracellular synthetic organelles by droplet-based microfluidics, Staufer [/bib_ref] , cytoskeletons [bib_ref] Engineering light-responsive contractile actomyosin networks with DNA nanotechnology, Jahnke [/bib_ref] or phase-separated organelles [bib_ref] Programming molecular self-assembly of intrinsically disordered proteins containing sequences of low complexity, Simon [/bib_ref] that can be combined in almost a plug-and-play manner. This enables breaking down biological questions into addressable subsets while allowing for the sequential deconstruction of individual molecular mechanisms. As such, studies based on a bottom-up strategy are especially powerful to deconvolute molecular mechanisms of disease processes that occur in a sequential and modular manner. The reconstitution of viral infection cycles is a prime example of a systematic dissection by bottom-up synthetic biology principles.
In fact, earlier studies have demonstrated how viruses, or parts thereof, can be assembled bottom-up and this ambitious approach has been recently transferred into the realm of SARS-CoV-2 research [bib_ref] New approaches for bottom-up assembly of tobacco mosaic virus-derived nucleoprotein tubes on..., Azucena [/bib_ref] [bib_ref] DNA-directed patterning for versatile validation and characterization of a lipid-based nanoparticle model..., Kozminsky [/bib_ref]. Intriguingly, in vitro reconstitution of the SARS-CoV-2 viral envelope and integration of the spike glycoprotein on lipid vehicles revealed how inflammatory fatty acids are exploited to trigger a molecular switch that couples local inflammation states to SARS-CoV-2 infectivity [bib_ref] Synthetic virions reveal fatty acid-coupled adaptive immunogenicity of SARS-CoV-2 spike glycoprotein, Staufer [/bib_ref] [bib_ref] Free fatty acid binding pocket in the locked structure of SARS-CoV-2 spike..., Toelzer [/bib_ref] [bib_ref] Structural insights in cell-type specific evolution of intra-host diversity by SARS-CoV-2, Gupta [/bib_ref]. These efforts are not only driven by the implementation of synthetic biology in biomedicine but also mark the beginning of a first bottom-up assembled viral infection cycle. Bottom-up construction of fully artificial, custom-designed viral replication systems holds great promise to systematically dissect and decipher the sequential process during viral infection, replication, propagation, and transmission. This approach could be particularly valuable for SARS-CoV-2, the causative agent of the ongoing global pandemic, as resolving the sequential process associated with intracellular viral replication, could complement the ongoing elucidation of immunological aspects in COVID-19 and uncover novel mechanisms that can be potentially targeted to interfere with viral propagation. Based on recent progress in synthetic cell engineering, researchers in the field now dispose of a wide variety of molecular tools that could be combined and adapted for bottom-up engineering of a synthetic viral replication cycle recapitulating in a defined and sequential, and therefore controllable way, the corresponding process of the live virus.
## A modular approach towards engineering of synthetic viral infection cycles
Towards the bottom-up assembly of a compartmentalized virus-like structure capable of intracellular replication, certain essential features of the virions of choice need to be considered for the design process: Virions have evolved as highly efficient carriers of genetic information that have mastered the reprogramming of host cells for viral propagation. As inherent infectious agents, viruses exploit the molecular machinery of host cells to produce and propagate viral particles. The viral infection process itself is modular, and therefore ideally suited to be tackled by synthetic biology approaches [bib_ref] Modular evolution of coronavirus genomes, Vakulenko [/bib_ref] [bib_ref] Modular organization of SARS coronavirus nucleocapsid protein, Chang [/bib_ref] [bib_ref] Insights to SARS-CoV-2 life cycle, pathophysiology, and rationalized treatments that target COVID-19..., Trougakos [/bib_ref] [fig_ref] Figure 1 |: Viral replication cycle [/fig_ref]. Generally, the individual elements associated with intracellular viral replication cycles vary between virus species but typically include six defined sequential modules:
1. A module that allows identification of and binding to target cells by molecular recognition patterns (e.g. the receptor-binding domain (RBD) in the SARS-CoV-2 spike glycoprotein).
## 2.
A structural module and co-stimulatory factors initiate the uptake of the virions by the host cell (e.g. the viral envelope of the SARS-CoV-2 virion). 3. A module that facilitates the release of viral genetic information into the cell cytoplasm or nucleus (e.g. the fusion domain of the SARS-CoV-2 spike glycoprotein). 4. A module that initiates the engagement of the host cell transcription-translation machinery in order to synthesize additional genomic copies and viral capsids. 5. A module that regulates the reprogramming of the host cell to facilitate viral capsid assembly (e.g. the formation of subcellular viral factories by the ER-Golgi intermediate compartment in SARS-CoV-2-infected cells). 6. A module that allows for the controlled exocytosis of correctly assembled virions. Importantly, viruses have evolved towards a precise molecular orchestration of the individual modules and virions have been optimized by evolution to compress all required functionality into a single structure characterized by the lowest achievable complexity. This evolution-driven reduction of complexity is paralleled in bottom-up synthetic biology research, underscoring the value of this approach to studying viral replication. For instance, in the case of SARS-CoV-2, several functionalities (i.e. modules) have been merged into a single molecule, the spike glycoprotein that realizes receptor recognition, cellular uptake, and release of genetic information. Similarly, the SARS-CoV-2 genome encompasses multiple overlapping open reading frames, providing a highly efficient compression of genetic information. In this context, an important boundary condition for a bottom-up assembled viral replication cycle is the need for the genetic inscription of all structural and functional modules. This genetic information will need to be readable for the cellular machinery and allow for efficient production of new synthetic virion particles, mimicking exactly the process of a live virus. This also restricts the usage of non-natural building blocks, such as synthetic lipids, as these will need to be supplemented and processed by the host cells. At the same time, however, such exogenic functional elements might rationally be incorporated into the design in order to limit the propagation of synthetic viruses (such as nutrient limitation in bacterial strain culture) for biosafety purposes. Such biorthogonal switches would not only allow to cap off the synthetic replication cycle but also tune the temporal dynamics of the system when interfaced with living cells. In this regard, two fundamentally different scenarios and engineering strategies will need to be developed and combined. Firstly, a strategy for in vitro assembly of the initial synthetic viral particles. If necessary, this process can be performed under non-physiological conditions and supported by molecular engineering technologies such as microfluidics or organic synthesis of nanoparticles. The central purpose of these firstgeneration synthetic virions will be the initial transduction of host cells (including modules 1-3). Secondly, the synthetic virions generated will need to be capable of producing second-generation virions (modules 4-6), assembled in the intracellular environment, limiting the applicable synthesis tools, and building blocks to those provided by the cellular machinery.
A synthetic cycle should be designed under consideration of the natural mechanisms for viral replication to enable drawing biomedically relevant conclusions. However, the reductionistic approach pursued will require simplification and abstraction of certain modules, providing more controllable and quantitative means for virology as compared to studies based on natural viruses. Achieving a true integration of all required functional modules, within the boundaries of a particle of the lowest possible complexity, is potentially the most ambitious task for the bottom-up assembly of a synthetic viral replication cycle. For this, individual modules need to be designed with compatible chemistry and maintain functionality in the intracellular and extracellular environment. Although the unification of individual viral functionalities into a single particle structure is a challenge in itself, several technologies that allow engineering the synthetic replicates of the modules detailed above have been presented:
## Module 1
The main scaffolding element of membrane-enclosed virions are lipids that form from the host cell into a unilamellar lipid membrane around the densely packed virus genome. As compartmentalization is a central objective of bottom-up synthetic biology, a variety of technologies have been previously developed to precisely form and manipulate viral envelope-like lipid vesicles. For instance, liposome technologies have been adopted to construct artificial mimetics of several virus species [bib_ref] All the small things: how virus-like particles and liposomes modulate allergic immune..., Kratzer [/bib_ref]. Moreover, there are biorthogonal methods available for the controlled incorporation of recombinant viral proteins into these vesicles [bib_ref] Hemagglutinin of influenza virus partitions into the nonraft domain of model membranes, Nikolaus [/bib_ref]. Liposomes have not only been functionalized with class I viral fusion proteins, but also with other targeting moieties that can direct liposome interactions with the cells of choice [bib_ref] Bottom-up assembly of biomedical relevant fully synthetic extracellular vesicles, Staufer [/bib_ref]. Therefore, an array of technologies is already available for the controlled assembly of synthetic virus scaffolding elements with cell-specific tropism.
## Module 2
For intracellular uptake, viruses exploit a combination of effects based on molecular self-assembly and receptor-mediated endocytosis [bib_ref] Forces during cellular uptake of viruses and nanoparticles at the ventral side, Wiegand [/bib_ref]. Several studies have demonstrated that virus-like particles with a similar size to natural viruses can induce uptake by target cells via passive, entropy-driven membrane bending and receptor clustering [bib_ref] Mechanics of receptor-mediated endocytosis, Gao [/bib_ref]. Therefore, bottom-up assembled viruses will need to be optimized for receptor avidity by tuning the ligand density on their surface. In this way, efficient intracellular uptake could be rapidly achieved without the need for further complexity-increasing building blocks [bib_ref] Forces during cellular uptake of viruses and nanoparticles at the ventral side, Wiegand [/bib_ref].
## Module 3
As genetic information-carrying vehicles, virions are highly efficient in releasing their genetic cargo into cells. In fact, viruses have served as tools for drug delivery, aiming to achieve efficient integration of genomic elements into cells for therapeutic purposes. So far, even the most sophisticated artificial drug delivery platforms do not achieve the transduction efficiency and targeting-specificity of natural viruses. Therefore, most gene-therapy approaches are currently still based on viral vector systems. However, several molecular designs for the intracellular release of cargo from liposomal carriers have been adopted from viral blueprints. For instance, synthetic peptides mimicking the fusogenic properties of viral fusion proteins have been engineered on virus-like particles to achieve endosomal release of the cargo [bib_ref] A display of pH-sensitive fusogenic GALA peptide facilitates endosomal escape from a..., Nishimura [/bib_ref]. Other approaches exploit the intrinsic fusogenic abilities of viral envelope proteins, although recombinant versions of such proteins tend to display instability in solution. To overcome the stability limitations, hemagglutinin-like mechanisms have been mimicked by DNA nanotechnology, which allows proximity ligation of membranes [bib_ref] Hemifusion and fusion of giant vesicles induced by reduction of inter-membrane distance, Heuvingh [/bib_ref]. However, fusogenic peptides provide a clear advantage over synthetic compounds as they can be genetically encoded for synthetic viral replication cycles. The peptides could potentially be tethered to the synthetic virus envelope via palmitoylation, inspired by the anchoring mechanism of the influenza fusion protein haemagglutinin. Altogether, by taking inspiration from drug delivery technologies, fusogenic modules for synthetic virus replication are readily available and a controlled release of the genetic viral cargo can be achieved for synthetic viral replication cycles.
## Module 4
With the advancement of genetic engineering methods, viral genomes can be readily produced by recombinant DNA technology. Large viral genomes, such as SARS-CoV, horsepox, and baculoviruses, have been designed and sequentially assembled for gene therapy purposes [bib_ref] MultiBac: baculovirus-mediated multigene DNA cargo delivery in insect and mammalian cells, Gupta [/bib_ref] [bib_ref] Synthetic recombinant bat SARS-like coronavirus is infectious in cultured cells and in..., Becker [/bib_ref] [bib_ref] Construction of an infectious horsepox virus vaccine from chemically synthesized DNA fragments, Noyce [/bib_ref]. Moreover, in the context of SARS-CoV-2, strategies to reverse engineer complete viral genomes from subgenomic fragments have been presented [bib_ref] Rapid reconstruction of SARS-CoV-2 using a synthetic genomics platform, Thao [/bib_ref]. Based on this, tailored chemically synthetized viral genomes could be incorporated into virion-like particles, although the mere size of viral genomes will represent an additional challenge that needs to be tackled by advanced encapsulation technologies (such as those based on microfluidics 41 ) for initial liposome assembly and programmed intracellular packing (see module 5). Likely, strategies to reduce the immunogenicity and increase the stability of artificial RNA genomes will be included, for example by incorporating pseudouridine nucleotides [bib_ref] The Critical contribution of pseudouridine to mRNA COVID-19 vaccines, Morais [/bib_ref]. By incorporation of appropriate translational elements including ribosome binding sites and regulatory sequences, the intracellular transcription-translation machinery could be highjacked to produce second-generation genome copies and structural virion proteins. The proposed approach mainly aims to bottom-up assemble viral phenotypes (e.g. virions) that incorporate functional but artificial copies of viral genomes. The genomic element could either encompass the full genetic information found in the virus to be mimicked, or only retain those elements strictly necessary to achieve viral replication. Ideally, a viral particle could be designed that is able to replicate within cells without requiring genetic elements, although basing such a complex process only on self-assembly appears challenging. In order to minimize the genetic payload, single ORFs could be incorporated or a full genome with overlapping ORFs, synthesized by recombinant technologies, could be utilized.
## Module 5
From a molecular system engineering perspective, a module that achieves reprogramming of the cellular machinery and formation of intracellular viral factories that produce correctly assembled virions is presumably the most challenging. In this regard, exploiting the intrinsic mechanisms that would lead to the formation of viral factories, cellular replication networks, and lipid droplets is the most feasible approach as active engineering approaches could be circumvented [bib_ref] Lipid droplets fuel SARS-CoV-2 replication and production of inflammatory mediators, Dias [/bib_ref] [bib_ref] Ultrastructural modifications induced by SARS-CoV-2 in Vero cells: a kinetic analysis of..., Eymieux [/bib_ref] [bib_ref] Protease-resistant cell meshworks: an indication of membrane nanotube-based syncytia formation, Staufer [/bib_ref]. Recent data suggest that intrinsically disordered domains in viral proteins (e.g. non-structural proteins 1 and 2) and host proteins initiate or sustain the formation of phase-separated organelles that serve as viral factories promoting the self-assembly of SARS-CoV-2 virion fragments [bib_ref] Deconstructing virus condensation, Lopez [/bib_ref]. Exploiting the natural function of nonstructural SARS-CoV-2 proteins to engineer a synthetic replication cycle is a promising strategy as it may allow simultaneously blocking the translation of host mRNAs and enabling the formation of virion factories [bib_ref] The key features of SARS-CoV-2 leader and NSP1 required for viral escape..., Bujanic [/bib_ref]. However, while designing a replication module on the basis of the cellular machinery allows one to closely model natural virus replication, this strategy compromises the quantitative and controllable characteristics of the overall system. For the initial benchmarking of this module, the bottom-up approach could be combined with a more top-down focused strategy based on in vitro transcription/ translation systems from cell lysates. In this way, large-scale screenings in protocell systems could allow for targeted optimization of the conditions required for viral factory formation.
## Module 6
Cellular products are subjected to constant degradation and recycling at homeostatic conditions. However, newly assembled lipid-enveloped viral particles are able to escape from this autophagic process by promoting the directed sorting of virions into multivesicular bodies for side-directed exocytosis. Intracellular sorting of cargo for vesicle packing and endocytosis is based on a complex but coordinated interplay between supramolecular machines (e.g., the endosomal sorting complex required for transport, ESCRT) that localize specific cargos along the Golgi-ER-membrane axis. This machinery is highjacked by viruses for assembly and release of virions for future infection cycles [bib_ref] The regulation of Endosomal Sorting Complex Required for Transport and accessory proteins..., Ahmed [/bib_ref]. In order to engineer this process in a controlled, tunable, and quantitative manner, modules inspired by extracellular vesicle (EV) engineering technologies could be applied. EVs have been linked to the evolution and maturation of viruses and have also been shown to play a critical role in SARS-CoV-2 infection [bib_ref] Extracellular vesicles and viruses: are they close relatives?, Hoen [/bib_ref] [bib_ref] Circulating exosomes are strongly involved in SARS-CoV-2 infection, Barberis [/bib_ref]. Efforts directed towards the engineering of artificial EVs and specific loading of transgenic cargo into these cell-derived liposomes resulted in a rich toolbox of genetic engineering strategies that allow for programmable sorting of proteins and nucleic acids into multivesicular bodies for control packing into EVs. Specific genetically encoded EV sorting peptides derived from endogenous EV proteins (e.g., lysosomeassociated membrane protein 2b, platelet-derived growth factor receptor, or tetraspanins) could be fused to proteins of interest in order to induce directed loading into virions during a synthetic viral replication cycle [bib_ref] Designer exosomes: a new platform for biotechnology therapeutics, Jafari [/bib_ref]. For sorting of nucleic acid components, methods based on the phase separation characteristics of EV sorting proteins (e.g., YBX1) could be repurposed to load RNA genomes into synthetic virions. Furthermore, viral genomes could be tagged with specific EVsorting sequences that by nature promote the inclusion of RNAs into EVs. Specific loading strategies of this kind could be quantitatively tuned by regulating the sumoylation of the sorting proteins [bib_ref] Sumoylated hnRNPA2B1 controls the sorting of miRNAs into exosomes through binding to..., Villarroya-Beltri [/bib_ref]. The proposed engineering strategy would allow to the exploitation of the natural relationship between EVs and enveloped viruses, providing a potentially efficient instrument to regulate synthetic virion exocytosis.
## Ethical considerations of synthetic viral replication cycles
Technological progress is driven by ethical goals and technology aims to improve the quality of life for all people. Engineers developed codes of conduct that document this ethical orientation. According to such written ethos, engineers should respect principles of sustainability, safety, health, and welfare of the public. The research focused on bottom-up engineering of synthetic life-like systems, alike conventional engineering, is not value-free [bib_ref] Can science make sense of life?, Wollenberg [/bib_ref]. The suggested modularization approach, to create a synthetic virus replication cycle, is linked to the hope to mitigate some crucial risks associated with traditional virus research and is ultimately motivated by developing research tools for public benefits, such as new vaccines and gene therapy technologies.
In the 2010s, research on viruses attracted public attention due to genetic engineering efforts on influenza A virus subtypes (H1N1 / H5N1) [bib_ref] Airborne transmission of influenza A/H5N1 virus between ferrets, Herfst [/bib_ref] [bib_ref] Experimental adaptation of an influenza H5 HA confers respiratory droplet transmission to..., Imai [/bib_ref]. Inspired by a top-down engineering strategy, research activities of concern included gain-of-function experiments that manipulated infectious agents, enhancing or generating potential for pandemic development. There was public concern that these engineered viruses would spread accidentally or be misused for terrorist or military purposes. Subsequently, several statements by scientific committees and research policy decisions emerged. A moratorium in the US stopped public funding for such gain-of-function research on the influenza virus, severe acute respiratory syndrome (SARS) virus, and the Middle East respiratory syndrome (MERS) virus from 2014 to 2017 [bib_ref] US suspends risky disease research, Reardon [/bib_ref] [bib_ref] Ban on gain-of-function studies ends, Burki [/bib_ref]. This triggered a controversial discussion on ethical guidelines for the publication of security-relevant information [bib_ref] Gain-of-function research: ethical analysis, Selgelid [/bib_ref]. The recent SARS-CoV-2 pandemic has brought these concerns back to the attention of the scientific community and alerted the wider public [bib_ref] A SARS-like cluster of circulating bat coronaviruses shows potential for human emergence, Menachery [/bib_ref] [bib_ref] Engineered bat virus stirs debate over risky research, Butler [/bib_ref].
In contrast to previous top-down genetic engineering approaches, the bottom-up engineering of synthetic viral replication cycles promises to minimize key risks, as a bottom-up design implicates a reduction of complexity, and synthetic viral replication cycles ideally only encompass properties that are absolutely necessary. This risk minimization by design would facilitate a high level of control over the engineered system. Along this line, several approaches have been developed to study the molecular mechanisms underlying viral infection, host immune response, and evolution of viruses. Some of the most extensively used tools include pseudoviruses and genetic engineering of viral genomes. Pseudoviruses are viral particles that do not display a relevant health threat to humans (e.g. based on lentiviruses) but incorporate molecular features of the virus under study. Pseudoviruses are especially valuable for analysis of host humoral immunity, e.g. in antibodybased neutralization studies. Moreover, they have been applied for several decades and a large body of experimental protocols and biosafety/biosecurity assessments exist. However, these systems do not allow to recapitulate of the biophysical properties of a target virus in full, a significant disadvantage that limits the translation of the obtained findings. Further, genetic engineering of viruses, although in parts challenging, has proven to be a powerful approach to studying viral infections, as it allows the assessment of viral genome organization and gene regulation. Moreover, with the help of viral genome engineering, gain-of-function (GoF) experiments can be performed by introducing recombinant DNA fragments and genes into viral genomes. However, as GoF experiments on human pathogenic viruses display a significant health threat, frequently revised regulatory mechanisms are in place to assess research work with unpredictable but most likely dangerous outcomes. In this context, the bottom-up assembly of viral replication cycles can be classified as a complementary technology. While the concept can be viewed as a GoF approach to introduce functionality into assembles of molecules instead of viral genomes, it mitigates the risk of uncontrolled viral replication by providing a high degree of molecular control. The complementary nature of the proposed approach allows to create molecular systems that recreate specific functional and structural features of viruses that cannot be obtained with conventional tools. For instance, it allows us to define and tune the density of fusion proteins present on the enveloping membrane, highlighting that in the bottom-up approach safety is mostly provided by degrees of control. However, as a novel approach, and in contrast to more established tools, bottom-up assembly of viral replication cycles lacks a significant amount of empiric knowledge that could be applied for risk assessment.
Research on bottom-up assembled viruses must deal with risks, either resulting from the research activity itself (biosafety issues) or regarding the use and misuse of the produced knowledge (biosecurity issues).
## Biosafety
In most expectations, artificial viral replication cycles appear to be harmless as long as they do not interact with natural systems. In the case of incubation with living cells, risk mitigation tools, such as the use of obligatory exogenous agents for synthetic virion replication, could be established to prevent the uncontrolled iterative replication of the designed systems outside of laboratory conditions. With such "safety bars", foreseeable risks could be minimized to a greater extent. However, findings and biological insights reached with synthetic replication cycles are scope-limited, unless they are integrated into experimental setups with higher biological relevance (e.g. cells and animals). Conditions that are "more natural" and equipped with fewer safety mechanisms could increase the significance and impact of the research but need to be strictly balanced with the associated risk factors. In this context, research activities on potentially harmful and infectious agents must be conducted under appropriate biosafety precautions (e.g. laboratories with different biosafety levels). Different molecular safety bars could be designed to warrant the safe employment of each of the proposed modules. Specifically, the use of artificial promoters 70 to drive gene expression in module 4, could be applied to specifically induce expression of the artificial genomes specifically within target cells. Such a genetically inscribed safety bar would display high resilience as an inherent design feature of the system. However, over several experimental cycles, genetic recombination could compromise the stability of such a genetic tool. Another safety bar to regulate and limit uncontrolled iterations between modules 6 and 1 could include encoding fusion incompetent release proteins. In this way, mature virions could be produced by the full completion of one viral cycle but unintended replication could be limited. It could be expected that such a safety bar shows high persistence in the designed system, as the spontaneous evolution of a fusion-competent form would require a series of mutations that would not be selected for under culture conditions. Further safety bars might include the deletion of glycosylation sites in viral proteins and other mechanisms of viral immune evasion. In this way, rapid clearance by the immune system would increase the biosafety level. Eventually, in order to maximize biosafety, a combination of several safety bars should be applied to compensate for the possible failure of single mechanisms in individual modules. This could be complemented by routine quality checks in experimental setups that include long-time frame experimental analysis, e.g. by sequencing the genomes present in a culture repeatedly. Therefore, we propose the following risk assessment scheme to categorize biosafety aspects in synthetic virus research:
- Low risk (low complexity): "safety bars" + research on isolated synthetic virus or individual modules, to be performed under biosafety level 1 condition. - Medium risk (medium complexity): "safety bars" + synthetic virus is inserted into natural systems or a combination of two or more consecutive modules, to be performed under biosafety level 2 conditions. - Higher risk (high complexity): "safety bars" + synthetic virus is inserted into natural systems with a set of modules that would allow full replication, to be performed under biosafety level 3 conditions.
Matching risk levels with laboratory biosafety levels not only offers elevated safety standards for research by increasing the protective measures (e.g. personal protective equipment and advanced safety instrumentation) but also provides extended documentation and surveillance protocols. Moreover, it offers a practical solution since BSL standards are well established across research institutions and associated regulations are in place. Nonetheless, as research risks might increase by interfacing synthetic viral replication cycles with other engineering strategies of limited predictive value (e.g. top-down genetic engineering of host cells), the proposed scheme only serves as initial benchmarking. The molecular design strategy can reduce the risk, but this does not imply that this approach is risk-free and that the security bar mechanism in place should not lead to misjudgment of the potential hazards. Therefore, such emerging bottom-up technologies should also be managed according to established safety regulations. In some instances, however, the current regulatory mechanisms might not prove sufficient, e.g. in the case that specific hazards are still unknown today. In such cases, institutions and involved stakeholders are required to commit to and adapt existing measures in order to mitigate associated risks. Ideally, the actors involved should develop a moral awareness that claims to oversee the consequences of their actions to the best of their knowledge. This moral awareness should be encouraged by scientific institutions and advisory comities. It also refers to fundamental principles such as sustainability, justice, peace, and human rights. Research funders are also in a position to impose risk-minimizing conditions. In addition, there are national and international regulatory measures to protect the public 72 .
## Biosecurity
Regarding biosecurity issues, Dual Use Research of Concern (DURC) [bib_ref] Dual use in the 21st century: emerging risks and global governance, Ienca [/bib_ref] needs to be carefully considered within the bottom-up framework. For instance, targeted optimization of viral replication cycles is possible and could be legitimized for synthetic virus approaches. The suggested modularization approach, to create synthetic virus replication cycles, aims to mitigate some crucial risks associated with conventional research on natural viruses and is ultimately motivated by developing research tools for public benefits. However, with good reason, "engineering of viruses and viral delivery systems" is listed by the Fast Track Action Subcommittee on Critical and Emerging Technologies of the White House. The committee lists those "advanced technologies that are potentially significant to U.S. national security". This classification refers to the precautionary principle, which includes, but is not limited to, the question if technology has intolerable biosecurity risks [bib_ref] The Precautionary Attitude: Asking Preliminary Questions, Wolff [/bib_ref]. Synthetic virus research implies technological knowledge that involves serious potential risks, some of which are known but most of which are yet unknown. In particular, emerging technologies in the field of virus research are partially associated with dystopian and catastrophic concerns. The following remarks argue that specified risk assessment is needed based on the significant uncertainties regarding the proposed research concepts on synthetic viral replication cycles. We ask for risk assessment, which includes iterative monitoring mechanisms and integrates public consultations and concerns. Such a monitoring mechanism should evaluate biosafety and biosecurity issues of synthetic viruses while in progress and inform media, political actors, and members of political representative bodies. The aim must be to identify and mitigate risks from the very beginning.
Similar to the National Science Advisory Board for Biosecurity (NSABB, USA) the German Commission for Biosecurity (ZKBS) assesses current developments in the field of synthetic biology and makes its findings available to ministries, the parliament, and the public. The commission summarizes that current research on synthetic biology does not pose any risks to biosafety/biosecurity, either in Germany or worldwide, other than those already assessed with the help of the German Genetic Engineering Act and other international regulations for "conventional" genetic modifications. However, this appraisal could change if it were to become possible to create synthetic viruses, which feature functional replication systems. The possibility of misuse (pathogens with pandemic potential) of this technology has led to a variety of security measures that could contain new regulations by the government. In Germany, there was a controversial debate on legal regulation measures, which was also reflected in the German Bundestag. Eventually, it was decided to retain the governance within the scientific community, informed by expert committees in the involved research institutions. In accordance with the Recommendations for Handling Security-Relevant Research by the German Research Foundation (DFG) and the German National Academy of Sciences Leopoldina, the risk assessment comprises the following modules: Risk analysis that aims to minimize identified risks, documentation and communication risks, evaluation of proposed publications, and training of the staff involved. A central module was the establishment of local commissions for ethics in security-related research (KEFs) at German research institutions. The central task of these commissions is to identify possible risks, such as dual-use problems, at a very early stage and to propose suitable measures to minimize them. In individual cases, this examination may also lead to the decision, that "a highrisk project only being carried out at a later point in time, following a research moratorium, or perhaps not at all, even when the project is not prohibited by law". The Joint Committee on the Handling of Security-Relevant Research by the DFG and the Leopoldina advises the local KEFs (see for example the "Key questions for the ethical assessment of security-relevant research" 82 ), documents the safety-relevant risks identified, functions as a platform to share experience in preventing risks and oversees the publication of the results. Importantly, this also applies to synthetic virus research. The experts involved should steadily review synthetic virus technologies in order to immediately report security-related hazards. The focus on biosecurity issues should not distract from the fact that there is also a need for governance in the field of biosafety [bib_ref] Advances in synthetic biology and biosafety governance, Li [/bib_ref].
The actors involved also include scientists as authors of peerreviewed publications and scientific publishers who may disclose security-related knowledge. However, biosecurity issues have yet not been the focus of published protocols or research results. The publication of the de novo assembly of SARS-CoV-2 genomes triggered discussions about necessary guidelines regarding security-relevant publications [bib_ref] Engineering SARS-CoV-2 using a reverse genetic system, Xie [/bib_ref]. Members of the Engineering Biology Research Consortium (EBRC) recommend that the review and publication process needs to implement control mechanism to deal more seriously with safety and security issues: "The outcomes of the author, reviewer, and editor surveys should be used as a basis for discussion on minimizing publication risks. In some cases, additional safety and/or security experts may need to be engaged, and it may be valuable for authors to discuss security concerns that may result from publication with research institutions, funders, and (rarely) appropriate governmental officials" [bib_ref] Making security viral: shifting engineering biology culture and publishing, Mackelprang [/bib_ref].
What counts for security-relevant aspects of emerging technologies in general, applies to GoF research of pathogens in particular: It must establish an iterative, easily revisable risk assessment and risk management strategy. From a molecular perspective, the approach proposed here is also a GoF approach, which needs ethical and legal expertise from the beginning. On an international level, several guidelines and rules have been proposed and implemented [bib_ref] Comparison of International Guidance for Biosafety Regarding Work Conducted at Biosafety Level..., Johnso [/bib_ref]. At national levels, mechanisms for regulation have been adopted, for example, the HHS P3CO framework in the USand existing NIH guidelines are applicable, at least in part, to the research outlined in this comment. Of note, these regulations are not considered strict enough for a larger scientific community and the Research Service of the US Congress listed several legal possibilities to regulate even more restrictively. The extent to which GoF research at the molecular level should also be subject to stricter requirements should be part of an ongoing debate.
Apart from an unintended spread of engineered pathogenic viruses, one of the most serious biosecurity risks is the misuse of this technology. Mindful of biowarfare and bioterrorism, it must be carefully considered whether safety-relevant research results can be published in full. It must be prevented that synthetic viruses that are created through published knowledge to misuse these pathogens as weapons. This also concerns the filing of patents. Against the background of potential benefits for medical care, such preventive measures should not disproportionately hinder scientific exchange and technological progress [bib_ref] The precautionary principle in the context of multiple risks, Rushton [/bib_ref] [bib_ref] The precautionary principle and risk-risk tradeoffs, Hansen [/bib_ref]. These mechanisms are part of an permanent monitoring process that characterizes a responsible scientific culture within a "web of prevention" [bib_ref] The challenge of framing for efforts to mitigate the risks of "dual..., Husbands [/bib_ref]. Responsible scientists have to acknowledge legal requirements, even though responsible science goes beyond legal regulations alone. Every scientist is also morally obliged to identify and mitigate risks in a reflexive and adaptive system of soft law regulation and self-governance. This process could also make new national and international legal requirements necessary if the creation of artificial viruses will be feasible. Furthermore, the role of intelligence services is part of the web of prevention, which needs to be highlighted to detect threats that affect public health, the economy, and (national) security [bib_ref] Technical Aspects of Biosecurity: Screening Guidance, Attribution, and Traceability, Appleton [/bib_ref]. In this case, careful trade-offs must be made between freedom of research and safety obligations to the public. Of note, virus research will need to be increasingly understood in the global context that takes place under different national legal regimes 98 . The exchange of securityrelevant data and research results can collide with existing export regulations 99,100 . While a moratorium or even a ban on synthetic virus research as the ultima ratio is conceivable, it is rather unlikely that such measures could be effectively enforced due to the number of researchers involved in different countries. Nevertheless, efforts must be intensified to establish international ethical and legal standards to guide security-related research [bib_ref] A global forum on synthetic biology: the need for international engagement, Dixon [/bib_ref].
Possible misuse of the developed technologies could also arise from a do-it-yourself (DIY) research mentality. Such efforts largely evade regulatory and legal surveillance common to established research institutions. If successfully adopted, the proposed modular approach could reduce the tacit knowledge required to conduct experimental work on bottom-up assembled viral replication cycles. This in turn could result in biosecurity-relevant threats that could be exploited for the development of bioweapons. Codes of conduct and voluntary commitments already exist in the DIY communities, which are the results of intensive reflection and debate [bib_ref] Addressing Emerging Synthetic Biology Threats: The Role of Education and Outreach in..., Novossiolova [/bib_ref]. The same legal requirements apply to DIY laboratories as to academic or industrial laboratories [bib_ref] Biosafety in DIY-bio laboratories: from hype to policy: Discussions about regulating DIY..., Sundaram [/bib_ref]. Therefore, also in the context of DIY research on bottom-up assembled viral replication cycles, increased awareness of security-relevant aspects should be encouraged and, if necessary, new legal measures should be implemented.
Scientific data sharing and open-science efforts are particularly strong in the bioengineering field, wherefore publication bans would represent a remarkable restriction. Patents could also hinder scientific progress by restricting access to key technical information and licensing-based pay-walls [bib_ref] Can patents deter innovation? The anticommons in biomedical research, Heller [/bib_ref] [bib_ref] The anticommons at 20: concerns for research continue, Contreras [/bib_ref]. Bottom-up assembled infection cycles may lead to patentable innovations that can be used for biomedical applications such as vaccine development. Especially regarding crises of a global scale, like the COVID-19 pandemic, it may be appropriate to tailor IP policies in terms of public health requirements. In this view, various tools have been developed, such as patent pools or equitable licensing strategies of universities that are applicable to the bottom-up assembly of viruses [bib_ref] University patenting and licensing practices in the United Kingdom during the first..., Keestra [/bib_ref] [bib_ref] Equitable Licenses in University-Industry Technology Transfer, Godt [/bib_ref]. Therefore, the ethical dimension of synthetic virion research involves the assessment of benefits and risks concerning patenting and licensing strategies.
Altogether, bottom-up engineering of viral replication promises improved risk management as the involved levels of complexity and associated risks can be controlled precisely. However, in cooperation with legal and administrative authorities, politicians, media, and the public, it is necessary to assess whether the existing laws are sufficient or whether new possibilities for technological misuse require new legal rules. Research that deals responsibly with the challenges outlined can only be guaranteed if the interconnected fields of action are clearly recognized. To attribute responsibility to the actors involved, it must be clear who performs what kind of research activity under which safety conditions and legal frameworks. Of crucial importance for the acceptance of synthetic viruses, research will be risk management within ethics committees and other involved bodies from the scientific community as well as transparent communication. However, it is not possible to predict to which extent political decision-makers, NGOs, and media coverage will come to the same conclusions in their risk assessment, highlighting the importance of public trust in this discourse.
# Conclusions
In summary, for all of the presented modules, several biomimetic engineering strategies have been presented here that can be adopted for the design and bottom-up construction of synthetic viral replication cycles. However, the challenge becomes formidable when aiming to unify all solutions into a single structure, as current approaches to link individual modules are not compatible. For instance, while large viral genomes can be chemically synthesized, host intrinsic mechanisms that direct RNAs into vesicles are most efficient for sorting small RNA sequences such as microRNAs and mRNAs. Therefore, finding practical solutions to engineer a continuous and self-sustaining replication cycle will require creative approaches that might take further inspiration from artificial molecular engineering tools such as DNA nanotechnology or biorthogonal chemistry. At the same time, this will also need to be balanced with the need to mirror natural viral replication as close as possible to foster new application-focused insights into viral replication and the treatment of associated diseases.
Once these challenges are successfully overcome, a powerful new tool for virology could be created. In fact, virus biology has been revolutionized by the application of semi-synthetic virions earlier. For instance, pseudoviruses have enabled unpreceded quantitative insights into the mechanism underlying viral cell entry, immune evasion, and evolution in many viral diseases including SARS-CoV-2 [bib_ref] Establishment of a well-characterized SARS-CoV-2 lentiviral pseudovirus neutralization assay using 293T cells..., Neerukonda [/bib_ref] [bib_ref] Assessing the application of a pseudovirus system for emerging SARS-CoV-2 and re-emerging..., Huang [/bib_ref]. A fully synthetic viral replication cycle could additionally allow studying fundamental questions, such as the importance of the self-assembly process during virion maturation. Moreover, the physiochemical interaction between virions and host cells could be studied with a high spatial and temporal resolution by the use of artificial particles that closely mimic the biophysical properties of virions. With direct relevance for studies on SARS-CoV-2, the implementation of synthetic viral replication cycles could empower the identification of new functional mechanisms that might be targeted by dedicated antiviral treatments. Such mechanisms could include processes associated with phase separation during virus factory assembly, quantitative knowledge of which, and the role of cellular regulation, remain largely elusive. Furthermore, synthetic replication cycles could mature into a programmable tool to study mechanisms of cellular defense against virusinduced metabolic reprogramming and activation of immune evasive mechanisms. Ultimately, bottom-up assembly of viral replication cycles can bring forward a new generation of techniques that allow deconstructing the temporal dynamics underlying viral infection in mechanistic, molecular detail.
[fig] Figure 1 |: Viral replication cycle. The SARS-CoV-2 infection cycle is depicted in a schematic view, highlighting the sequential and modular nature of the replication process. For each step of the viral replication cycle, viruses have evolved specific molecular modules that are efficiently compressed into single virion particles. [/fig]
|
Medial Malleolar Insufficiency Fracture of the Ankle in an Elderly Patient with Osteoporosis
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.Insufficiency fracture is a type of stress fracture, which is the result of normal stresses on abnormal bone. Postmenopausal osteoporosis is the most common cause of insufficiency fractures. An early diagnosis is best made with a bone scan or magnetic resonance imaging, as radiographs may initially appear normal. Insufficiency fractures of the lower leg and ankle are less common. Furthermore, reports of medial malleolar insufficiency fracture without any history of trauma in elderly patients are extremely rare. Thus, we report a case with a medial malleolar insufficiency fracture of the ankle in an elderly patient with osteoporosis. This case shows that we should be aware of the possibility of encountering an uncommon medial malleolar insufficiency fracture as a cause of pain in the ankle region of an elderly patient with osteoporosis.
# Introduction
Stress fractures can be divided into fatigue and insufficiency fractures. [bib_ref] Fatigue, insufficiency, and pathologic fractures, Pentecost [/bib_ref] A fatigue-type stress fracture occurs when abnormal or repeated stress is applied to normal bone, whereas an insufficiency-type fracture is produced by normal stress on abnormal bone such as when there is impaired bone structure. Insufficiency fracture patients are older, typically postmenopausal women with osteoporotic bones. [bib_ref] Insufficiency stress fractures of the foot and ankle in postmenopausal women, Kaye [/bib_ref] The risk factors are the same as those of osteoporosis, and also include abnormal bone conditions which affect bone turnover such as Paget's disease, osteomalacia, diabetes mellitus, hyperparathyroidism, chronic renal failure, highdose glucocorticoid therapy, and rheumatoid arthritis. [bib_ref] Stress fractures: current concepts, Daffner [/bib_ref] The most common sites of insufficiency fractures are the vertebrae, sacrum, pelvic rami, and femoral neck. [bib_ref] Osteoporotic insufficiency fractures of the pelvis simulating a malignancy in an elderly..., Maheshwari [/bib_ref] Less often, the lower leg and foot are involved, with the common sites being the distal tibia, calcaneus, metatarsal, and talus. [bib_ref] Insufficiency fractures of bilateral distal tibias associated with chronic liver disease, Tsuchie [/bib_ref] [bib_ref] Bilateral undisplaced insufficiency neck of femur fractures associated with shortterm steroid use:..., Gurdezi [/bib_ref] [bib_ref] Atraumatic bilateral insufficiency fractures of the talar neck in a rheumatoid patient, Bischoff [/bib_ref] However, reports of a medial malleolar insufficiency fracture without a history of trauma in elderly patients is extremely rare and it has not been reported in Korea up until now. Thus, we report a case with medial malleolar insufficiency fracture of the ankle in an elderly patient with osteoporosis that was treated with percutaneous screw fixation after there was no symptom improvement with conservative treatment.
## Case
A 65-year-old female patient presented with left ankle pain of 2 months' duration. There was no history of trauma or bisphosphonate use for osteoporosis and the pain was activity-related and relieved partially by rest or elevation. In the past medical history, no special diseases such as diabetes, hypertension, and cerebrovascular diseases were present and she was a non-smoker and a non-drinker. The physical examination showed mild localized tenderness in the medial aspect of the ankle and no swelling. The range of motion of the ankle was full, and stability was intact. She was 157 cm in height, had a body weight of 58 kg, and her body mass index was 23.5 kg/m 2 . The bone mineral density (BMD) was measured utilizing peripheral quantitative computed tomography (P-QCT; Somatom sensation 16, Simens, Erlangen, Germany). On assessment of the results, severe osteoporosis was considered because the BMD and T-score of the lumbar vertebrae (L1-4) were 40.4 mg/ cm 3 and -4.31 . At the time of visiting, a plain radiograph of the left ankle showed a fracture that was not clearly depicted but showed as a faint vertical sclerotic line in the medial malleolus of the ankle joint . She was treated with restriction of activity and non steroidal antiinflammatory drugs (NSAIDs). After 3 weeks of treatment, she still had pain in the left ankle. Magnetic resonance imaging (MRI) was performed. The MRI of the left ankle revealed an irregular vertical low signal line with surrounding marrow edema at the junction of the medial malleolus and the tibial plafond which did not extend to the medial malleolar cortex of the tibia . We diagnosed a medial malleolar insufficiency fracture of the ankle.
An operation was performed because of progressive pain and difficulty of ambulation during the follow-up, leading to early recovery and normal activities. The patient was placed in the supine position under spinal anesthesia. The fracture was internally fixated with a percutaneous cannulated screw under c-arm guidanceand at the same time an arthroscopic procedure was performed. The arthroscopic examination revealed no displacement of the medial fracture line within the joint but synovitis was observed. Arthroscopic debridement was performed. After surgery, she was kept in a non-weight bearing cast for 5 weeks. At the fifth week, she was allowed to perform ankle range of motion exercises and partial weight-bearing walking. As of the 4-month follow-up, there was bony union and she had recovered to full activity with the treatment including osteoporosis medications. The patient was informed of this report.
# Discussion
Insufficiency-type stress fractures affect abnormal bone that is exposed to stresses that normally would not be sufficient to cause mechanical failure and thus they are often observed in postmenopausal osteoporotic women. [bib_ref] Insufficiency stress fractures of the foot and ankle in postmenopausal women, Kaye [/bib_ref] The cause of insufficiency fractures is often multifactorial, and typically the main factor is osteoporosis followed by osteopenia. Osteopenia decreases bone elasticity enough to allow fractures to occur from normal activity; however, according to Breer et al. [bib_ref] Stress fractures in elderly patients, Breer [/bib_ref] , the stress fracture incidence in elderly patients seems to be multifactorial and not based on osteoporotic changes alone. Also, balanced calcium and vitamin D metabolism seems to be of paramount importance in the prevention of insufficiency fractures in elderly patients. [bib_ref] Stress fractures in elderly patients, Breer [/bib_ref] The reason why osteoporosis leads to insufficiency fractures remains unclear, but in this case, severe osteoporosis was considered a cause of the insufficiency fracture. Alth ough the defined mechanisms are unclear, insufficiency fractures mainly occur in the medial malleolus of the ankle when a physiological torsional load is applied to the bone, or a muscular contracture, or a misalignment force such as occurs with a sprained ankle impacts bone that has inadequate mechanical strength.
The diagnosis of an insufficiency fracture is primarily based on the history and physical examination. The leading clinical symptoms in patients with an insufficiency fracture is pain, and this may be worse with loading and increased physical activity. [bib_ref] Medial malleolar stress fracture in athletes: diagnosis and operative treatment, Lempainen [/bib_ref] Radiographs are still the most widely used imaging methods for identification of insufficiency fractures, but the sensitivity is limited. MRI is the diagnostic agent of choice due to its higher sensitivity and specificity, and it allows the differentiation of benign versus malignant fractures. [bib_ref] Imaging of insufficiency fractures, Krestan [/bib_ref] The typical MRI appearance of an insufficiency fracture is a low marrow signal and prominent enhancement on T1-weighed images and a high marrow signal with extension into adjacent soft tissue on T2weighed images. An early diagnosis is best made with a bone scan or an MRI, as radiographs may initially appear normal. In this case, the clinical signs were mild pain in the medial aspect of the ankle, which was confirmed by an MRI. [bib_ref] MRI and CT of insufficiency fractures of the pelvis and the proximal..., Cabarrus [/bib_ref] [bib_ref] Insufficiency fracture of the body of the calcaneus in elderly patients with..., Ito [/bib_ref] The treatment of an insufficiency fracture depends on the clinical presentation and whether the fracture is complete or incomplete. Treatment options are conservative treatment, a splint or a cast, and an operation such as internal fixation. Also, treatment of the underlying cause of bone weakness is essential. [bib_ref] Insufficiency stress fractures of the foot and ankle in postmenopausal women, Kaye [/bib_ref] [bib_ref] Stress fractures: current concepts, Daffner [/bib_ref] We experienced a rare case of medial malleolar insufficiency fracture of the ankle that was related to severe osteoporosis in an elderly patient. Medial malleolar insufficiency fracture of the ankle is easily overlooked unless it is suspected. When ankle pain becomes suddenly exacerbated in postmenopausal elderly patients, insufficiency fractures should be considered, and if they are present, diagnosed with a bone scan or an MRI.
[fig] Figure 1, Figure 2: Bone mineral density of the lumbar spine by peripheral quantitative computed tomography (P-QCT). (A) Initial radiograph of the left ankle shows a fracture was a faint vertical sclerotic line in the medial malleolus of the ankle joint. (B) Coronal T2-weighted and (C) proton density weighted magnetic resonance image shows an irregular low signal line and surrounding bone marrow edema at the junction of the medial malleolus and the tibial plafond which did not extend to the medial malleolar cortex of the tibia. [/fig]
[fig] Figure 3: (A) Postoperative radiograph of the left ankle shows good fixation with a cannulated screw. (B) Radiograph at 4 months after the operation shows bony union. A B http://dx.doi.org/10.11005/jbm.2013.20.2.119 [/fig]
|
Ultra-broadband terahertz absorption by exciting the orthogonal diffraction in dumbbell-shaped gratings
Metamaterials, artificial electromagnetic media consisting of periodical subwavelength metal-based micro-structures, were widely suggested for the absorption of terahertz (THz) waves. However, they have been suffered from the absorption of THz waves just in the single-frequency owing to its resonance features. Here, in this paper, we propose a simple periodical structure, composed of two 90 degree crossed dumbbell-shaped doped-silicon grating arrays, to demonstrate broadband THz wave absorption. Our theoretical and experimental results illustrate that THz waves can be efficiently absorbed more than 95% ranging from 0.92 THz to 2.4 THz. Such an ultra-wideband polarization-independent THz absorber is realized mainly based on the mechanisms of the anti-reflection effect together with the [61, 0]-order and [0, 61]-order grating diffractions. The application of our investigation can be extend to THz couplers, filters, imaging, and so on. E lectromagnetic metamaterials, made of densely artificial arranged subwavelength resonance cells, have produced exotic effects such as invisible cloaking, negative refraction, Fano resonance and superlens 1-8 . Recently, metamaterial-based perfect absorbers have attracted a great deal of interests in worldwide due to a host of potential applications including imaging, detecting, and sensing 9-11 . The first single-frequency microwave metamaterial perfect absorber (MPA) was proposed by N. I. Landy et al., and it was composed of two metallic layers separated with a loss dielectric spacer to couple the electric and magnetic fields, respectively 12 . H. Tao et al., also designed a ring resonator (SRR)/loss dielectric spacer/metal-wire sandwich 'fishnet' MPA to realize a single-frequency and polarization-dependent THz absorber with absorbance of 70% 13 . Then, polarizationindependent single-frequency THz absorber was obtained based on the interference effects by using a symmetrical metal-based ERR and a mirror underneath the ERR 14 . Here, the SRR is designed to change the phase of the transmitted and the reflected waves. The loss dielectric spacer and the metal-substrate are applied to absorb the incident waves and avoid the transmission of the incident waves, respectively. At certain frequency, all of the reflected waves are destructively coherent with each other, and thus no incident wave is reflected, resulting in a single-frequency of MPA. According to the destructive coherent effects, dual-band, triple-band, and multi-band THz absorbers were demonstrated by embedding much more ERR with different sizes or shapes 15-20 . Broadband THz absorbers were also realized by introducing multi-layered (much more than two layers) gradually varied ERR structure 21-23 . In addition, many similar structured absorbers have been widely investigated in microwave and optical frequencies 24-31 . But, all of these multi-band and broadband THz absorbers are suffered from the difficulties of either alignment or fabrication (complex structure), which hinders their practical application.As a practical solution, R. Kakimi et. al., proposed a single-layered photonic-crystal slab, which can be easily fabricated with doped silicon, to capture of THz waves 32 . The distributed Bragg reflections in this structure create photonic bandgaps, which can be utilized for wave confinement. So, when they operate the crystals at the photonic band edges (leaky-mode regime at the photonic band edges), the in-plane guide wave and the freespace wave incident from the out-of-plane direction can be strongly coupled with each other, resulting in the guide-mode resonance (in-plane resonant modes). In addition, a mirror was introduced underneath the photonic-crystal slab to induce Fabry-Pérot resonance. Therefore, a 50 GHz (with absorptivity $ 90%) bandwidth of THz absorber was realized by combining the guided-mode resonance and the Fabry-Pérot resonance. Recently, M. Pu et. al., have theoretically reported another simple structure i.e., doped silicon with square-shaped grating arrays, to realize a much wider bandwidth of THz absorption in virtue of the anti-reflection effects and the first-OPEN SUBJECT AREAS: TERAHERTZ OPTICS OPTICAL PHYSICS
order diffraction [bib_ref] Engineering heavily doped silicon for broadband absorber in the terahertz regime, Pu [/bib_ref] , and the experimental demonstration was realized in our previous work [bib_ref] A polarization independent broadband Terahertz absorber, Shi [/bib_ref]. However, the square-shaped grating array just utilize the [61, 0]-order grating diffraction, which suppresses the bandwidth of the absorber. Here, in this paper, different from the above broadband absorbers based on the destructive coherent effects, our ultra-broadband THz absorber is realized by using the antireflection effects, especially the [61, 0]-order and the [0, 61]-order grating diffractions. We achieve such an ultra-broadband THz absorber by designing a single-layered periodical structure consisting of two 90 degree crossed dumbbell-shaped doped-silicon grating. In such a single-layered dumbbell-shaped doped-silicon grating, the smaller horizontal gap between the horizontal dumbbells and the bigger horizontal gap between the vertical dumbbells are designed to excite the [61, 0]-order and the [0, 61]-order grating diffractions, simultaneously. And, the absorption bandwidth is further enhanced in such a single-layered grating rather than exciting the second-order grating diffraction (to enhance the absorption bandwidth) in a double-layered grating, as theoretical discussed in Ref. [bib_ref] Engineering heavily doped silicon for broadband absorber in the terahertz regime, Pu [/bib_ref]. Our numerical simulations and especially the experimental testing results demonstrate a polarization-independent THz absorber with nearly 1.5 THz absorption bandwidth (with absorptivity $ 95%). Here, we want to emphasis that the absorption bandwidth defined in this paper is not the full width at half maximum (FWHM). As defined in Ref. [bib_ref] Capture of a terahertz wave in a photonic crystal slab, Kakimi [/bib_ref] , our absorption bandwidth means the corresponding bandwidth for absorptivity $ 95%. To the best of our knowledge, such an absorption bandwidth with absorptivity $ 95% (especially, our experimental results) is much broader than all of the previous reported THz absorbers.
# Results and discussion
The schematic and the fundamental principal of such an ultra-broadband THz absorber are shown inOur model system has a square lattice with 90 degree rotational symmetry, for which the absorption is insensitive to the polarization of the incident THz waves. Each unit cell consists of two 90 degree crossed dumbbell-shaped doped-silicon strips. The corresponding geometric parameters are p 5 96 mm, l 5 27 mm, s 5 17 mm, and w 5 25 mm. The thickness (h) of each dumbbell-shaped strip (pattern) and substrate are 38 mm and 462 6 10 mm, respectively. The permittivity of both the dumbbell-shaped doped-silicon grating and the substrate can be described by Durde model as follows:
[formula] e~e ? { v 2 p v(vzi=t) ,ð1Þ [/formula]
where e ' 5 11.7, t is the carrier relaxation time, and v p is the plasmon frequency. In this paper, the resistivity of the boron-doped silicon is 0.54 V cm, and thus the corresponding t is 0.571 ps, and v p is 19.1 THz. The fundamental principal of our model system is shown in. In the low frequency regime, the dumbbellshaped doped-silicon grating is equivalent to an effective medium coating on the substrate, and the absorption of THz waves is mainly attributed to the anti-reflection effect between the incident and the reflected THz waves. However, in high frequency regime, these dumbbell-shaped strip arrays can be considered as two 90 degree crossed grating (the horizontal stripe and the vertical stripe grating array). Therefore, the [61, 0]-order and [0, 61]-order grating diffractions are separately contributed to the absorption of THz waves. By choosing the proper structure parameters (as shown above), the anti-reflection effect and the grating diffractions can be jointed with each other, resulting in a wide-bandwidth of THz absorber. Here, we want to emphasis that such a crossed dumbbell-shaped doped-silicon grating array is designed through the parameters optimization based on numerical simulations. Although a cross-shaped dopedsilicon grating can also be designed as a broadband THz absorber, the corresponding bandwidth is less than the crossed dumbbellshaped doped-silicon grating (The numerical simulations are not shown here). [fig_ref] Figure 2 |: SEM of the fabricated periodical array slab [/fig_ref] shows the scanning electron microscope (SEM) images of the dumbbell-shaped doped-silicon grating. The testing sample is fabricated by using the traditional photolithography and the inductively coupled plasma (ICP) etching on a 0.54 V?cm p-type silicon wafer (with thickness of 500 6 10 mm). Based on a serial of technical processing, the dumbbell-shaped doped-silicon grating with the thickness of 38 mm is formed on the surface of the doped-silicon (see the methods).
The calculated and measured absorption spectra for the TE (transverse electric) and TM (transverse magnetic) incident THz waves are illustrated in system with frequency resolution of 4.58 GHz. For TE incident THz wave, as shown in [fig_ref] Figure 3 |: The simulated [/fig_ref] of the red line, the peak absorbance is 99.3%, and the bandwidth with absorptivity $ 95% is about ,1.5 THz. The measured result, as depicted in [fig_ref] Figure 3 |: The simulated [/fig_ref] of the blue line, also demonstrates an over 1.5 THz absorption bandwidth (absorptivity $ 95%). Both the calculated and the measured spectra show good agreement, except for a slight difference in resonance frequency. This discrepancy can be considered as a result of structural difference between the fabricated sample, and the calculated model. Comparing Figs. 3(a) and 3(b), we can find that for both TE and TM incident THz waves, the absorption spectra are nearly consistent with each other for the normal incidence. It means that such an ultra-broadband THz absorber is insensitive to the polarization of the incident waves, due to the symmetrical periodical structure. Such an absorption bandwidth for both TE and TM incident THz waves is much broader than that of the case in doped-silicon photonic crystal slab (Ref. 32). From Figs. 3(a) and 3(b), we can also find that three absorption peaks at 1.265 THz, 1.72 THz, and 2.155 THz appear in the spectra. Such a broadband THz absorber is caused closely combined with these three peaks. The physical mechanism of these three peaks can be explained qualitatively as follows: the left peak is induced by the anti-reflection effect, while the right two peaks are caused mainly by the grating diffractions. When the carrier density is n 5 1.6 3 10 16 cm 23 (with the resistivity of 0.54 V?cm), the quality factor of our designed periodical structure is very low, as shown in [fig_ref] Figure 3 |: The simulated [/fig_ref] (inset). Therefore, these three peaks can be broadened and combined into each other, resulting in an ultra-broadband THz absorber, as mentioned in Ref. [bib_ref] Capture of a terahertz wave in a photonic crystal slab, Kakimi [/bib_ref]. For f 5 1.265 THz, the wavelength of the incident THz wave is larger than the period of the structure, so the grating can be considered as an effective medium. According to the effective medium theory, the effective index of the two-dimensional grating array can be calculated as 35
[formula] n eff~½ nz2n _ 2D z2n 2D =5 ð2Þ with n~(1{f 2 )n 1 zf 2 n 2 n _ 2D~ffi ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (1{f )e 1 zf e \ p n 2D~ffi ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi e1e == f e1z(1{f )e == q e \~e 1 e2 f e1z(1{f )e2 e ==~( 1{f )e 1 zf e 2 8 > > > > > > > < > > > > > > > : , [/formula]
and f~S si S 0 (here, S si is the surface area of the two crossed dumbbellshaped doped-silicon strips, and S 0 5 p * p), n 1 5 1 and n 2 is the index of doped-silicon. In order to prove the anti-reflection effect at low frequency regime (f 5 1.265 THz), we calculate the reflection spectrum of a doped-silicon substrate coating with such effective medium (calculated from Eq. (2)) upon its upper-surface, as shown in [fig_ref] Figure 3 |: The simulated [/fig_ref]. Obviously, it is a typical reflection spectrum, and a dip nearly at f 5 1.265 THz with small reflection which is caused by the anti-reflection effects between the reflected waves. Here, the effective medium can be regarded as anti-reflection coating layer above the substrate, and thus, the absorption peak at f 5 1.265 THz is caused by the anti-reflection effects. From the field distribution shown in www.nature.com/scientificreports [fig_ref] Figure 4 |: The electric field distribution at the interface between the periodical arrays and... [/fig_ref] (a 1 ), it can be found that, the incident electric field is almost localized in the two crossed dumbbell-shaped doped-silicon arrays and less of incident wave is transmitted into the substrate (see [fig_ref] Figure 4 |: The electric field distribution at the interface between the periodical arrays and... [/fig_ref] (a 2 )), demonstrating that in low frequency regime, the grating arrays are equivalent to an effective anti-reflection coating, and the incident wave is mainly absorbed in this effective medium (grating layer). However, in the high-frequency regime, the reflection spectrum of [fig_ref] Figure 3 |: The simulated [/fig_ref] still shows a high reflection which is mismatched with the absorption characterizes in [fig_ref] Figure 3 |: The simulated [/fig_ref]. That is to say, the effective medium theory can't completely apply to the high-frequency regime.
In high frequency regime, the period of the structure is larger than the wavelength in the silicon but still smaller than that in free space (l/n , p , l), and the periodical structure can be mainly considered as grating arrays [bib_ref] Engineering heavily doped silicon for broadband absorber in the terahertz regime, Pu [/bib_ref]. Taking TE (electric field is parallel to x-axis) incident THz waves for example, the incident electric field will 'see' two crossed strip arrays with different widths along x-axis. So, two kinds of grating (horizontal stripe and vertical stripe gratings) diffractions can be realized by using these crossed dumbbell-shaped doped-silicon grating arrays. Based on two-dimensional rigorous coupled-wave analysis (2D-RCWA) method, we calculate the grating diffractions of the grating, as shown in [fig_ref] Figure 3 |: The simulated [/fig_ref] Furthermore, from the field distribution of [fig_ref] Figure 4 |: The electric field distribution at the interface between the periodical arrays and... [/fig_ref] , we can conclude that the peak at f 5 1.72 THz is mainly due to the [61, 0]-order field diffraction through the smaller air gap between the horizontal dumbbell-shaped doped-silicon strips (see Figs. 4(b 1 ) and 4(b 2 )), and the [0, 0]-order field diffraction through the vertical dumbbell-shaped doped-silicon strips. The peak at f 5 2.155 THz is mainly due to the [0, 61]-order field diffractions through the bigger horizontal air gap between the vertical dumbbell-shaped doped-silicon strips, and the [0, 0]-order field diffraction through the horizontal dumbbellshaped doped-silicon strips (see Figs. 4(c 1 ) and 4(c 2 )). Comparing with Figs. 4(a 2 ), 4(b 2 ), and 4(c 2 ), significant field distributions appear in the substrate (as shown in [fig_ref] Figure 4 |: The electric field distribution at the interface between the periodical arrays and... [/fig_ref] (b 2 ) and 4(c 2 )), which means that in the high-frequency regime, the periodical arrays are considered as grating arrays and much of the incident waves is absorbed in substrate through the grating diffraction.
Although the broadband absorption in high-frequency regime is mainly caused by the grating diffractions as discussed above, it is not clear about the distinct mechanism of the absorption due to the grating diffractions. [fig_ref] Figure 5 |: Reflection spectra and transmission spectra for the doped silicon slab with pattern [/fig_ref] and 5(b) illustrate the calculated and the measured transmission spectra together with the reflection spectra, respectively. For the doped silicon slab with/without pattern (grating array), the transmission efficiency is nearly zero. However, the reflection efficiency is quite different from each other, and the reflection efficiency for the doped silicon slab without pattern is significantly higher than the case of the doped silicon slab with pattern. It means that the pattern on the top of the doped silicon slab plays a key role in reducing the reflection of THz wave. And, the absorption efficiencies induced by the grating diffractions are about 23% and 27% for f 5 1.72 THz and f 5 2.155 THz, respectively [fig_ref] Figure 5 |: Reflection spectra and transmission spectra for the doped silicon slab with pattern [/fig_ref]. Now, we investigate the influence of the pattern (grating array) thickness and the period on the absorption spectra. [fig_ref] Figure 6 |: The dependences of the absorption spectra on [/fig_ref] shows the absorption spectra with different pattern thickness. By increasing the pattern thickness, all of the three peaks are shifted to the lower frequency (redshift). For the left peak, it is caused by the antireflection effects with the corresponding anti-reflection condition of nh 5 l/4 (where n is the index of the doped-silicon and h is the thickness of the pattern). Therefore, the left peak is shifted to the low frequency with increasing of the pattern thickness (h). However, the other two peaks at high-frequency regime, mainly caused by the zero-order and the first-order grating diffractions, also appear the redshift effects for the increasing of the pattern thickness. This may be understood as follows: First, the grating diffraction frequencies are independent on the thickness of the pattern. Second, the redshift of the two peaks at high-frequency regime is caused by the Fabry-Pérot resonance in the pattern. That is to say, when the zero-order and the first-order diffracted waves transmit to the interface between the pattern and the substrate, a partial of diffracted THz wave is reflected upward due to the impedance mismatching. Furthermore, the reflected THz wave is mainly reflected in the interface between the pattern and the free space, leading to the Fabry-Pérot resonance in the pattern (Here, partial of the THz waves on Fabry-Pérot resonance can also be considered as separating original from the grating diffractions.). Such Fabry-Pérot resonance effect in the pattern is depended on the thickness of the pattern, and it appears the redshift effects with increasing the pattern thickness. Therefore, the two peaks in the highfrequency regime show the redshift effects. When increasing the period of the grating (the thickness and the structure size of the pattern is fixed), the left peak shows the blueshift effects while the resonance frequency of the other two peaks at high-frequency regime are nearly fixed, as shown in [fig_ref] Figure 6 |: The dependences of the absorption spectra on [/fig_ref]. The effective index of the grating is decline by increasing the period of the grating, and thus, the left peak at low frequency regime, caused by the anti-reflection effect, appears the blueshift effects (according to the anti-reflection condition of nh 5 l/4). At high-frequency regime, the other two peaks are nearly fixed at the original resonance frequency (with the increasing of the pattern period) due to the competitive balance between the blueshift of the Fabry-Pérot resonance and the redshift of the grating diffractions. So, by adjusting the period and the thickness of the pattern, a broad-band THz absorber can be realized.
For practical applications, the incident THz wave isn't always radiating normally into the THz absorbing device. Therefore, we also study the absorption characterizes of this THz absorber at various incident angles for TE and TM polarizations shown in [fig_ref] Figure 7 |: The absorption spectra with different incidence angles for [/fig_ref]. For TE polarization [fig_ref] Figure 7 |: The absorption spectra with different incidence angles for [/fig_ref] , the maximum absorption remains above 90% with bandwidth of 1.5 THz even for incident angle as large as 45u. The absorption efficiency of TM polarization is larger than 95% with bandwidth of 1.5 THz with incident angle up to 45u. All of these results verify that such a THz absorber is insensitive to incident directions (with incident angle # 45u).
In summary, a novel broadband THz absorber has been proposed and experimentally demonstrated. In such a device, THz waves could be efficiently absorbed over 95% with bandwidth of 1.5 THz by etching two 90 degree crossed dumbbell-shaped grating on borondoped silicon, which was much broader than all of the previous reported THz absorbers. Anti-reflection effect and especially the [61, 0]-order and [0, 61]-order grating diffractions were applied to qualitatively analyze the performance of the designed THz absorbers. The calculated results were in good agreement with the measured ones in experiments. The designed THz wave absorber device was insensitive to the polarization states of incident waves due to the symmetry structure. Our investigation may have wide applications extended to THz filters, couplers, detectors, modulators, and switches.
# Methods
Numerical simulation. The absorption spectra and electric-field distributions are calculated by using a commercial electromagnetic simulator of CST Studio SuiteH 2012. We use frequency-domain solver to acquire transmission spectra. Unit cell boundary condition was used to simulate the THz absorber. The electric-field The complex dielectric constant of p-type doped-silicon was determined by virtual of Drude model. In order to analyze the absorption characteristics of the periodical array at high-frequency region, two-dimensional rigorous coupled wave analysis (2D-RCWA) was used to calculate the corresponding diffraction orders, based on a statevariables representation.
Fabrication and Testing system. The two 90 degree crossed dumbbell-shaped grating was fabricated on a p-type doped-silicon wafer by using the traditional photolithography and inductively coupled plasma (ICP) etching. The resistivity of the boron-doped silicon is 0.54 V cm. AZP4620 image reversal photoresist layer with thickness of 1 mm was spin-coated and patterned on the silicon substrate based on the standard photolithography. Then, etching for a proper time, the two crossed dumbbell-shaped doped-silicon grating is formed on the surface of the doped silicon.
The absorption characterizes of the sampler was tested based on a THz timedomain spectroscopy (THz-TDS) system. We use fiber laser to pump and detect the THz waves. The emitter and the detector are the LTG-GaAs-based photoconductive antenna. For the transmission spectra, two off-axis parabolic mirrors are used to focus the terahertz waves on the sample, and the other two off-axis parabolic mirrors collect and focus the transmit THz wave on the detector. For the reflection spectra, a 50/50 terahertz beam splitter is added into the system to gather the reflected wave from the sample. The absorption spectra are obtained based on A 5 1 2 T 2 R (T is the absorbance, while R is reflectance).
[fig] Figure 1 |: 3(a) and 3(b), respectively. The numerical calculations of the absorption spectra are carried out by using the commercial microwave software CST Microwave StudioH. The measured results are tested via a THz time domain spectroscopy (THz-TDS) Schematics illustrating the (a) periodical structure and (b) unit cell of the THz wave absorber. (c) The fundamental principal of THz wave absorption for broadband operation. [/fig]
[fig] Figure 2 |: SEM of the fabricated periodical array slab. [/fig]
[fig] Figure 3 |: The simulated (red line) and measured (blue line) results of the polarization-independent broadband THz absorber with (a) TE incident THz wave, and (b) TM incident THz wave. (c) The calculated reflection spectrum in the case of that the periodical array on top of the substrate is equivalent as an effective medium. (d) The diffraction efficiency (DE) of different diffraction orders. The inset in (a) is the quality factor of such periodical structure. [/fig]
[fig] Figure 4 |: The electric field distribution at the interface between the periodical arrays and the substrate for f 5 1.265 THz (a), f 5 1.688 THz (b) and f 5 2.135 THz (c), respectively. The inset in (c) is the electric field distribution at y 5 38 mm. [/fig]
[fig] Figure 5 |: Reflection spectra and transmission spectra for the doped silicon slab with pattern (red and blue dash dot lines) and without pattern (red and blue solid lines): (a) calculated results, and (b) measured results. (c) The difference of the absorption between the doped silicon slab with pattern and the doped silicon slab without pattern: red line is the simulation result, and the blue line is the experimental result. [/fig]
[fig] Figure 6 |: The dependences of the absorption spectra on: (a) grating depth h; (c) grating period p. [/fig]
[fig] Figure 7 |: The absorption spectra with different incidence angles for (a) TE polarized THz wave, and (b) TM polarized THz wave. www.nature.com/scientificreports SCIENTIFIC REPORTS | 5 : 8901 | DOI: 10.1038/srep08901 distributions were obtained based on the time-domain solver at certain frequencies. [/fig]
|
Somatic VHL Mutation in a Patient With MEN1-Associated Metastatic Pancreatic Neuroendocrine Tumor Responding to Sunitinib Treatment: A Case Report
Multiple endocrine neoplasia type 1 (MEN1) and von Hippel-Lindau (VHL) are autosomal-dominant diseases caused by germline mutations in tumor-suppressor genes. A patient with a germline MEN1 mutation and a somatic VHL mutation in the tumor has not been reported. Herein, we report on a patient with MEN1 and a metastatic nonfunctioning pancreatic neuroendocrine tumor (PNET) with a somatic VHL mutation. This patient underwent a pancreaticoduodenectomy for a grade 2 PNET obstructing her pancreatic duct. The patient developed liver and regional lymph node metastases as well as growth of a PNET in the remnant pancreas. As part of a clinical trial for mutation-targeted therapy, a biopsy of the metastatic tumor was obtained. The clinical diagnosis, confirmed by OncoVAR-NET and molecular profiling analysis, revealed MEN1 with a germline deletion in exon 2 and a c.402 deletion C, p.Phe134LeufsX51. In addition, a somatic mutation in the VHL gene-a nonsense mutation, c.529A.T, p.Arg177Ter-was identified by hybrid capture sequencing. The mutations were confirmed by Sanger sequencing. Comparative genomic hybridization showed loss of heterozygosity in both the MEN1 and VHL genes. The patient was treated with sunitinib and had a partial response to treatment. This case illustrates not only that a second hit occurs in tumor suppressor genes but that somatic mutations are also possible in additional tumor suppressor genes. This suggests that targeted therapy selection should include analysis of somatic mutations even when the susceptibility gene is known.
fifth decade of life, biochemical manifestations are present in 98% of patients with MEN1, and clinical manifestations are present in 80% [bib_ref] Clinical practice guidelines for multiple endocrine neoplasia type 1 (MEN1), Thakker [/bib_ref]. The penetrance of pancreatic tumors is 30% to 70% [bib_ref] Clinical practice guidelines for multiple endocrine neoplasia type 1 (MEN1), Thakker [/bib_ref]. Common manifestations seen in MEN1 include primary hyperparathyroidism, pituitary adenomas, and neuroendocrine tumors of the gastro-entero-pancreatic tract [bib_ref] Clinical practice guidelines for multiple endocrine neoplasia type 1 (MEN1), Thakker [/bib_ref]. MEN1 is diagnosed on the basis of one of the following three criteria: (1) occurrence of two or more MEN1-associated tumors (pituitary, parathyroid, or gastro-entero-pancreatic tumors), [bib_ref] Clinical practice guidelines for multiple endocrine neoplasia type 1 (MEN1), Thakker [/bib_ref] occurrence of one of the MEN1-associated tumors in a first-degree relative of a patient with a clinical diagnosis of one of the MEN1 manifestations, or (3) the presence of a germline MEN1 mutation [bib_ref] Clinical practice guidelines for multiple endocrine neoplasia type 1 (MEN1), Thakker [/bib_ref]. The MEN1 gene locus is on chromosome 11q13 and encodes the menin protein [bib_ref] Hereditary pancreatic endocrine tumours, Alexakis [/bib_ref]. As in other tumor suppressor-inherited cancer syndromes, a multistep model of "two hits" occurs, causing inactivating mutations in MEN1.
von Hippel-Lindau (VHL) disease is an autosomal-dominant disorder caused by a germline mutation in the tumor-suppressor gene VHL. It is located on chromosome 3p25.5 [bib_ref] Differential genetic alterations in von Hippel-Lindau syndrome-associated and sporadic pheochromocytomas, Bender [/bib_ref] [bib_ref] Endocrine manifestations of von Hippel-Lindau disease, Cassol [/bib_ref]. The incidence of VHL is 1 in 36,000 live births with a penetrance of 95% at 65 years of age [bib_ref] Endocrine manifestations of von Hippel-Lindau disease, Cassol [/bib_ref]. The clinical manifestation of VHL includes retinal hemangioblastomas, central nervous system hemangioblastomas, renal cystic and solid tumors, endolymphatic sac tumors, pheochromocytomas, cystic pancreatic lesions, and PNETs. PNETs are present in 11% to 17% of patients with VHL [bib_ref] Endocrine manifestations of von Hippel-Lindau disease, Cassol [/bib_ref] [bib_ref] Evaluation and management of pancreatic lesions in patients with von Hippel-Lindau disease, Keutgen [/bib_ref]. Although both MEN1 and VHL may manifest with PNETs, no germline MEN1 mutation with a coexisting somatic VHL mutation has been reported in the literature.
In sporadic PNETs, rates of somatic MEN1 mutations have been reported as 13% to 44%, whereas rates of somatic VHL mutations have been reported as ,1% [bib_ref] Role of disease-causing genes in sporadic pancreatic endocrine tumors: MEN1 and VHL, Moore [/bib_ref] [bib_ref] Mutations and allelic deletions of the MEN1 gene are associated with a..., Görtz [/bib_ref]. With advances in next-generation sequencing, it is becoming possible to efficiently analyze driver mutations in cancer for precision medicine with targeted therapy. In patients with low-and intermediategrade PNETs, there are various treatment options, including somatostatin analogues, sunitinib, everolimus, and peptide receptor radionuclide therapy. Sunitinib is a tyrosine kinase inhibitor that is active against multiple tyrosine kinase receptors, including VEGFR, PDGFR, KIT, RET, and FMS-like tyrosine kinase 3 (FLT3) [bib_ref] Mutation-targeted therapy with sunitinib or everolimus in patients with advanced low-grade or..., Neychev [/bib_ref]. Everolimus is a mammalian target of the rapamycin (mTOR) signaling pathway inhibitor and a derivative of rapamycin. It inhibits the PI3K/mTOR-raptor pathway, which regulates tumor cell proliferation and angiogenesis [bib_ref] Mutation-targeted therapy with sunitinib or everolimus in patients with advanced low-grade or..., Neychev [/bib_ref]. Both sunitinib and everolimus have shown significantly improved progressionfree survival in patients with progressive metastatic neuroendocrine tumors, including PNETs, and both drugs are approved by the US Food and Drug Administration for this indication [bib_ref] Daily oral everolimus activity in patients with metastatic pancreatic neuroendocrine tumors after..., Yao [/bib_ref] [bib_ref] Sunitinib malate for the treatment of pancreatic neuroendocrine tumors, Raymond [/bib_ref]. However, it is unclear which patients with metastatic PNETs benefit from which agent.
Thus, we are conducting a phase II trial of mutation-targeted therapy with sunitinib or everolimus in patients with advanced low-or intermediate-grade neuroendocrine tumors of the gastrointestinal tract and pancreas with or without cytoreductive surgery (NCT02315625) to test whether the tumor genotype determines response to treatment [bib_ref] Mutation-targeted therapy with sunitinib or everolimus in patients with advanced low-grade or..., Neychev [/bib_ref]. As part of our clinical trial, tumor biopsy and mutational analysis by next-generation sequencing is performed. Patients with somatic or germline mutations in MEN1/PDGFR/KIT/FLT3 or wild-type tumor are treated with sunitinib [bib_ref] Mutation-targeted therapy with sunitinib or everolimus in patients with advanced low-grade or..., Neychev [/bib_ref]. Patients with somatic or germline mutations in NF1/PTEN/P13K/AKT/mTOR/VHL are treated with everolimus. When there is disease progression, subjects cross over to treatment with the other medication. Treatment is discontinued in the case of disease progression after crossover, unacceptable toxicity, or withdrawal of consent [bib_ref] Mutation-targeted therapy with sunitinib or everolimus in patients with advanced low-grade or..., Neychev [/bib_ref]. Herein, we report on a patient who enrolled in this trial with a germline MEN1 mutation and metastatic PNET who was also found to have a somatic VHL mutation that responded to sunitinib therapy.
## Case description
A 45-year-old woman with MEN1 presented with abdominal pain, distention, bloating, and diarrhea. Her medical history was notable for primary hyperparathyroidism, which required a 3.5-gland parathyroidectomy, a nonfunctioning pituitary adenoma (for which she had transphenoidal resection because of growth and associated mass effect on the internal carotid artery), and Zollinger-Ellison syndrome, for which she was treated with pantoprazole twice daily. She also had a history of well-differentiated metastatic pulmonary neuroendocrine tumors, for which she had a video-assisted thoracoscopic resection. Her family history was notable for a father, a brother, two paternal uncles, one paternal aunt, a paternal grandmother, and a son all diagnosed with MEN1. In June 2010, she presented with symptoms and biochemical evidence of pancreatitis.
A workup revealed an obstructing complex mass in the head of the pancreas and lesions in the tail of the pancreas and duodenum []. She underwent a pancreaticoduodenectomy and had an uncomplicated postoperative course. Pathology showed duodenal and retroperitoneal gastrinomas as well as a PNET. Lesions were well differentiated with a WHO grade 1 tumor and a pT2N0MX stage. Immunohistochemical staining of the tumor was positive for synaptophysin and chromogranin [ [fig_ref] Figure 2: Tumor immunohistochemistry [/fig_ref] and 2(B)]. Immunohistochemistry was also performed for VHL (ab140989; Abcam, Cambridge, MA) and menin (ab2605; Abcam, Cambridge, MA) [ [fig_ref] Figure 2: Tumor immunohistochemistry [/fig_ref] and 2(D)]. Although there was positive staining in the menin protein, menin is a nuclear protein, and localization to the cellular membrane, as shown in [fig_ref] Figure 2: Tumor immunohistochemistry [/fig_ref] , is evidence of an abnormal protein [bib_ref] Unusual clinical and pathological presentation of a neuroendocrine tumor in a patient..., Sakurai [/bib_ref]. VHL staining appears predominantly cytoplasmic in normal tissue [bib_ref] Von Hippel-Lindau gene alterations in sporadic benign and malignant pheochromocytomas, Dannenberg [/bib_ref]. As shown in [fig_ref] Figure 2: Tumor immunohistochemistry [/fig_ref] , the neoplastic cells showed both nuclear and cytoplasmic staining.
Follow-up imaging in June 2014 showed growth of the tumor in the body and tail of the pancreas, mesenteric lymph node, and three arterial-phase enhancing liver lesions []. The patient was treated with sunitinib and had response to treatment by November 2016]. Of note, between the initiation of therapy and the response to treatment, our patient received no other therapy for her neuroendocrine tumors outside of sunitinib, and she remains on the same medication at the same dose.
# Discussion
MEN1 and VHL are autosomal-dominant diseases with known susceptibility genes that cause PNETs. In both of these inherited syndromes, a second hit (somatic mutation) occurs in addition to the germline, resulting in multiorgan disease manifestation. Although secondary somatic mutations may occur in inherited tumor-suppressor cancer syndromes, a coexisting germline mutation in the tumor-suppressor gene with a somatic mutation in another tumor-suppressor gene responsible for an inherited syndrome has not been reported. Our patient presented with a known germline mutation in the MEN1 gene and a family history of MEN1. She developed a metastatic PNET. She was enrolled in a clinical trial to assess whether the tumor genotyping was associated with a response to sunitinib and everolimus. Her previous clinical diagnosis of MEN1 was supported by her family history as well as by tumor genotyping using tumor tissue from her previous operative intervention, confirming the presence of a c.402 deletion C, p.Phe134LeufsX51 in MEN1 in exon 2 with LOH in the MEN1 locus in the tumor tissue; thus, she was started on sunitinib therapy. In addition, she had a coexisting somatic nonsense mutation in VHL gene c.529A.T, p.Arg177Ter. A germline MEN1 mutation with a somatic mutation in the tumor suppressor VHL has not been observed and may be a tumor genotype that is associated with aggressive PNETs.
Somatic VHL mutations in PNETs may be associated with metastatic progression [bib_ref] VHL inactivation is an important pathway for the development of malignant sporadic..., Schmitt [/bib_ref] [bib_ref] Deletion at 3p25.3-p23 is frequently encountered in endocrine pancreatic tumours and is..., Barghorn [/bib_ref] [bib_ref] High frequency loss of heterozygosity in von Hippel-Lindau (VHL)-associated and sporadic pancreatic..., Lott [/bib_ref] [bib_ref] Different molecular profiles characterize well-differentiated endocrine tumors and poorly differentiated endocrine carcinomas..., Furlan [/bib_ref]. In a study by Hessman et al. [bib_ref] Genetic alterations on 3p, 11q13, and 18q in nonfamilial and MEN 1-associated..., Hessman [/bib_ref] , LOH at 3p was observed in 45% of nonfamilial PNETs and in 36% of MEN1-associated PNETs. They also found that 92% of malignant tumors had allelic loss at 3p and 11q13 [bib_ref] Genetic alterations on 3p, 11q13, and 18q in nonfamilial and MEN 1-associated..., Hessman [/bib_ref]. Possible VHL mutations associated with welldifferentiated PNETs were found in one of 60 cases, according to a review of LOH studies in PNETs and gastrointestinal neuroendocrine tumors performed by Furlan et al. [bib_ref] Different molecular profiles characterize well-differentiated endocrine tumors and poorly differentiated endocrine carcinomas..., Furlan [/bib_ref]. Lott et al. [bib_ref] High frequency loss of heterozygosity in von Hippel-Lindau (VHL)-associated and sporadic pancreatic..., Lott [/bib_ref] found involvement of chromosome 3p in sporadic pancreatic islet cell tumorigenesis, with associated progression to malignancy. Chung et al. [bib_ref] Localization of putative tumor suppressor loci by genome-wide allelotyping in human pancreatic..., Chung [/bib_ref] demonstrated that both 3p and 11q (associated with VHL and MEN1, respectively) were commonly deleted loci in PNETs. Therefore, the combination of germline MEN1 mutation and somatic VHL mutation may be associated with more aggressive disease. Furthermore, even in patients with a known germline mutation in either VHL or MEN1, tumor genotyping may reveal a somatic mutation that would allow the selection of the most appropriate agent to target the altered pathway(s).
In the case described, sunitinib therapy was selected because the patient had MEN1 and preclinical studies showed response to sunitinib therapy in transgenic mouse models of MEN1-associated PNETs [bib_ref] Brivanib, a dual FGF/VEGF inhibitor, is active both first and second line..., Allen [/bib_ref]. The presence of a somatic mutation in VHL is associated with activation of the mTOR pathway; thus, everolimus treatment could be effective. However, VHL is associated with major tumor angiogenesis, which may explain why the patient had a response to sunitinib therapy [bib_ref] Use of the tyrosine kinase inhibitor sunitinib in a patient with von..., Jimenez [/bib_ref] [bib_ref] Molecular targeted therapy for carcinoid and islet-cell carcinoma, Yao [/bib_ref].
The case described illustrates that therapy selection based on tumor genotype in practice needs to be tested in clinical trials. Although "precision" treatment is selected according to the driver mutation(s) present and the expected pathway to be activated, response to treatment may be determined by the presence of coexisting mutations. Thus, in the current clinical trial in which the patient is enrolled, subjects who do not respond to one agent on the basis of the tumor genotype or presence of germline mutation are crossed over to treatment with the other agent to determine the important tumor genotype that predicts response to a selected agent.
[fig] Figure 1: (A) Computed tomography scan (axial section) with contrast showing mass in the head of the pancreas (arrow) that abuts the superior mesenteric vein and the second/third portion of the duodenum. [/fig]
[fig] Figure 2: Tumor immunohistochemistry. (A) Immunohistochemistry for synaptophysin (320 magnification, brown color). (B) Immunohistochemistry for chromogranin (320 magnification, brown color). (C) Immunohistochemistry for menin (320 magnification, brown color) with an inset of normal pancreatic tissue (340 magnification, brown color). (D) Immunohistochemistry for VHL (320 magnification, brown color) with an inset of normal pancreatic tissue (340 magnification, brown color). No other genetic alterations in genes associated with PNETs, such as PTEN, PIK3CA, AKT1, MTOR, TSC1, TSC2, NF1, TP53, FLT3, PDGFRA, ATM, KIT, and ATRX, were identified. Sanger sequencing confirmed the MEN1 and VHL mutations noted in the original molecular profiling [ [/fig]
[fig] Figure 3: (A) Computed tomography scans (axial section) of the abdomen with contrast showing enhancing mass along the dorsal aspect of the pancreas before and after treatment, decreasing from 3.1 cm at the greatest diameter to 2.5 cm, as well as a decrease in representative hepatic lesions (arrows). (B) Octreotide scan showing liver and peripancreatic uptake consistent with metastatic tumors (arrows). (C) DNA sequencing showing the same MEN1 germline and somatic mutations. (D) VHL somatic nonsense mutation, c. 529A.T. (E) Comparative genomic hybridization array showing LOH in chromosomes 3 and 11. (F) Sanger sequencing confirming the presence of c.402 deletion C, p.Phe134LeufsX51 MEN1 and c.529A.T, p.Arg177Ter VHL mutations. [/fig]
|
Antiretroviral therapy affects the z-score index of deviant cortical EEG rhythms in naïve HIV individuals
Objective: Here we tested the effect of combined antiretroviral therapy (cART) on deviant electroencephalographic (EEG) source activity in treatment-naïve HIV individuals. Methods: Resting state eyes-closed EEG data were recorded before and after 5 months of cART in 48 male HIV subjects, who were naïve at the study start. The EEG data were also recorded in 59 age-and sex-matched healthy subjects as a control group. Frequency bands of interest included delta, theta, alpha1, alpha2 and alpha3, based on alpha frequency peak specific to each individual. They also included beta1 (13-20 Hz) and beta2 (20-30 Hz). Low-resolution brain electromagnetic tomography (LORETA) estimated EEG cortical source activity in frontal, central, temporal, parietal, and occipital regions. Results: Before the therapy, the HIV group showed greater parietal delta source activity and lower spatially diffuse alpha source activity compared to the control group. Thus, the ratio of parietal delta and alpha3 source activity served as an EEG marker. The z-score showed a statistically deviant EEG marker (EEG+) in 50% of the HIV individuals before therapy (p b 0.05). After 5 months of cART, delta source activity decreased, and alpha3 source activity increased in the HIV subjects with EEG+ (about 50% of them showed a normalized EEG marker). Conclusions: This procedure detected a deviant EEG marker before therapy and its post-therapy normalization in naïve HIV single individuals. Significance: The parietal delta/alpha3 EEG marker may be used to monitor cART effects on brain function in such individuals.
# Introduction
Infection with human immunodeficiency virus (HIV) causes neuropathological changes as well as neurological and neuropsychological symptoms in most (i.e. N50%) subjects over time [bib_ref] Neurological manifestations of AIDS, Mcarthur [/bib_ref] [bib_ref] Altered peripheral nerve conduction in HIV patients, Chavanet [/bib_ref] [bib_ref] Peripheral neuropathies in HIV infections -a prospective study of 56 cases, Gastaut [/bib_ref] [bib_ref] CSF and HIV, Chalmers [/bib_ref] [bib_ref] HIV-1 gp120 modulates the immunological function and expression of accessory and co-stimulatory..., Williams [/bib_ref] [bib_ref] Monocyte maturation, HIV susceptibility, and transmigration across the blood brain barrier are..., Williams [/bib_ref] [bib_ref] Memory loss in persons with HIV/AIDS: assessment and strategies for coping, Selnes [/bib_ref] [bib_ref] The Neuropathology of HIV/AIDS, Anthony [/bib_ref] [bib_ref] Virological response to salvage therapy in HIV-infected persons carrying the reverse transcriptase..., Antinori [/bib_ref] [bib_ref] Leukocyte transmigration across the blood-brain barrier: perspectives on neuroAIDS, Roberts [/bib_ref]. Fortunately, combined antiretroviral therapy (cART) mitigates these effects of HIV on brain function [bib_ref] HIV-associated neurocognitive disease continues in the antiretroviral era, Clifford [/bib_ref] , thanks to the attenuation of viral load (VL) and increased CD4 cell counts [bib_ref] A controlled trial of two nucleoside analogues plus indinavir in persons with..., Hammer [/bib_ref] [bib_ref] Continued CD4 cell count increases in HIV-infected adults experiencing 4 years of..., Hunt [/bib_ref] [bib_ref] Monocyte maturation, HIV susceptibility, and transmigration across the blood brain barrier are..., Williams [/bib_ref]. However, the prevalence of HIV patients with neurological and neuropsychological symptoms remains relatively high [bib_ref] Dynamics of cognitive change in impaired HIV-positive patients initiating antiretroviral therapy, Cysique [/bib_ref] [bib_ref] Evaluation of HIV RNA and markers of immune activation as predictors of..., Sevigny [/bib_ref]. One of the reasons for this situation is the lack of biomarkers suitable for monitoring the brain function effects of HIV and cART, especially for the personalized clinical management of HIV and therapy monitoring on an individual-, rather than population-level.
Resting state electroencephalographic (EEG) rhythms are a readily observed and quantified functional feature of the brain, and represent promising candidates for disease biomarkers. EEG rhythms reflect neurophysiological mechanisms of cortical neural synchronization related to the fluctuation of cortical arousal during changes in states of vigilance [bib_ref] Resting state cortical rhythms in mild cognitive impairment and Alzheimer's disease: electroencephalographic..., Babiloni [/bib_ref]. It can be hypothesized that quantitative EEG biomarkers can contribute to a preliminary instrumental assessment of HIV subjects, as well as to understanding the neurophysiological mechanisms underlying HIV's effects on brain function. In this line, preceding studies have shown the topography of resting state eyes-closed EEG rhythms in groups of HIV individuals. Compared to control groups of healthy subjects, HIV groups exhibited decreased posterior alpha (8-13 Hz) power density [bib_ref] Prospective associations between lateralised brain function and immune status in HIV infection:..., Gruzelier [/bib_ref]. This effect preceded cognitive and neurological impairment, was associated with changes in psychiatric status, and was normalized by cART . In addition, deviant theta (4-7 Hz) and alpha power density was associated with mood and immune status (i.e. CD4 counts) in a group of asymptomatic HIV subjects [bib_ref] Prospective associations between lateralised brain function and immune status in HIV infection:..., Gruzelier [/bib_ref].
To overcome partially head volume conduction effects blurring the localization of the cortical generators of resting state scalp EEG rhythms, our research group has used a popular freeware called low-resolution brain electromagnetic tomography (LORETA; [bib_ref] LORETA (low resolution brain electromagnetic tomography): new authentic 3D functional images of..., Pascual-Marqui [/bib_ref]. Results were quite promising at the group level. Compared to healthy subjects, naïve HIV subjects were characterized by greater activity of central and parietal delta sources (b4 Hz), as well as lower alpha (8-12 Hz) sources in widespread cortical regions [bib_ref] Cortical sources of resting-state EEG rhythms are abnormal in naïve HIV subjects, Babiloni [/bib_ref]. These differences in alpha sources were less marked in HIV subjects experiencing cART for at least 12 months . Furthermore, they showed normal delta source activity . This beneficial effect of cART on brain function was confirmed by a longitudinal study in which EEG rhythms were recorded before and after 5 months of cART in naïve HIV subjects [bib_ref] Antiretroviral therapy effects on sources of cortical rhythms in HIV subjects: Responders..., Babiloni [/bib_ref].
More recently, we have tested a simple statistical procedure based on the computation of z-score (p b 0.05, one-tailed) to identify treatment-naïve HIV male individuals having deviant activity of cortical (LORETA) sources of resting state eyes-closed delta and alpha rhythms [bib_ref] Abnormal cortical sources of resting state electroencephalographic rhythms in single treatment-naïve HIV..., Babiloni [/bib_ref]. The ratio of the activity between parietal delta and high-frequency alpha sources served as an EEG marker of interest. The z-score of this EEG marker allowed the identification of a relatively high percentage (about 40-50%) of treatment-naïve HIV individuals with a statistical difference of EEG activity (p b 0.05, one-tailed). These results suggested that this EEG marker might enrich the instrumental assessment of HIV effect on brain function in treatment-naïve HIV male individuals. However, the study could not clarify the extent to which the mentioned z-score procedure is sensitive to the impact of cART on naïve HIV individuals (no follow-up). Therefore, we sought to address this unanswered question in the present study. We hypothesized that the naïve HIV individuals with z-scores indicating that normal EEG source activity before the therapy would show a global stability of this activity after 5 months of cART. In other words, we expected their EEG rhythms to remain relatively unchanged and within the previously determined levels of normalcy even after treatment. In contrast, those individuals with z-scores indicating deviant EEG source activity before treatment were expected to show beneficial effects of cART on this activity (i.e. a normalization of EEG source activity) at the follow-up.
# Methods
## Subjects
This study included EEG and clinical data of 48 naïve HIV male subjects (mean age 39.4 years ± 1.6 standard error, SE), recruited at University S. Andrea Hospital and Tor Vergata Hospital of Rome (Italy). It also included EEG and clinical data of an age-matched control group of 59 cognitively normal male subjects (Healthy; mean age 39 years ± 2.2 SE), selected from a university archive to obtain the best matching of age and gender between the two groups.
All experiments were performed with the written informed consent of each participant and with the approval by the local ethical committee, in line with the Code of Ethics of the World Medical Association (Declaration of Helsinki) and the standards established by the Author's Institutional Review Board.
## Diagnostic criteria
All participants received laboratory exams for the confirmation of the HIV serostatus. These exams included HIV RNA viral load, Complete Blood Count (CBC), Treponema screening, CD4 lymphocyte count and percent, toxoplasmosis and cytomegalovirus antibody titers, HBV-HCV screening, serum protein, renal and liver function, and albumin. Toxicological analyses on urine samples controlled for the use of cocaine, opiates, amphetamine, and marijuana. No naïve HIV subjects showed CD4 counts compatible with a diagnosis of full-blown AIDS. A Computerized Diagnostic Interview Schedule interview (i.e. Version IV, CDIS-IV) controlled for DSM-IV Axis I and II disorders. The naïve HIV subjects completed questionnaires or brief interviews that assessed their medical history, medication use, parental psychopathology, psychiatric symptoms, demographics, drug and alcohol use, and cognitive status. Furthermore, neuropsychological tests assessed memory, language, executive function/attention, and visuo-construction abilities. Memory was evaluated by the Prose Memory Test delayed recall of a story [bib_ref] Standardizzazione e taratura italiana di test neuropsicologici, Spinnler [/bib_ref]. Language was assessed by 1-minute verbal fluency for letters [bib_ref] Tre test clinici di ricerca e produzione lessicale. Taratura su soggetti normali, Novelli [/bib_ref] and 1-minute verbal fluency for fruits, animals or car trades [bib_ref] Tre test clinici di ricerca e produzione lessicale. Taratura su soggetti normali, Novelli [/bib_ref]. Executive function and attention were evaluated by the Trail Making Test part A and B [bib_ref] Validity of the Trail Making Test as an indication of organic brain..., Reitan [/bib_ref]. Finally, Mini-Mental State Examination (MMSE) tested global basic cognitive functions [bib_ref] Mini-mental state". A practical method for grading the cognitive state of patients..., Folstein [/bib_ref]. reports information about personal and clinical characteristics of the healthy and naïve HIV subjects of the present study. An independent t-test evaluated the presence or absence of statistically significant differences between the two groups (i.e. Healthy and naïve HIV) for age, education, and MMSE score (p b 0.05). Results showed a statistically significant difference in the MMSE score (p = 0.0005), which was higher in the Healthy group than in the naïve HIV group. Furthermore, there was a higher education in the Healthy group than in the naïve HIV group (p = 0.01). In contrast, no statistically significant difference was found for age (p N 0.9). Therefore, education was used as a covariate in the subsequent statistical comparisons between the Healthy and the naïve HIV groups.
Exclusion criteria included mental retardation, intelligent quotient (IQ) score lower than 70, seizures, acute illness, and neurosurgery. These criteria also included history of head injury with loss of consciousness [bib_ref] ASPD blunts the effects of HIV and antiretroviral treatment on event-related brain..., Bauer [/bib_ref] and major neurological and psychiatric disorders such as schizophrenia or bipolar disorder (Diagnostic and Statistical Manual of Mental Disorders fourth edition, DSM-IV). Naïve HIV subjects were excluded for medical disorders such as chronic obstructive pulmonary disease, hypertension, hepatic encephalopathy, ocular disorders, Type 1 diabetes, and cirrhosis. They were also excluded in Means (±standard error, SE) of the personal and/or clinical features of healthy and treatment-naïve HIV male individuals of the present study. The third column reports the results of t-tests (p b 0.05, one-tailed) comparing these mean values between the two groups (i.e. Healthy vs. treatment-naïve HIV case of a positive urine toxicology or breathalyzer tests. They were also excluded for a recent (past year) dependence upon alcohol, cocaine or opiates. All healthy subjects underwent cognitive screening as well as physical and neurological examinations. They were neither affected by chronic systemic illnesses, major neurological or psychiatric diseases nor assumed psychoactive drugs. None of them had a major depression based on a clinical exam.
## Eeg recordings
Resting state eyes-closed EEG data were collected by 0.3-70 Hz bandpass, 19 electrodes placed according to and cephalic reference. To monitor eye movements, the horizontal and vertical electrooculograms (EOGs, 0.3-70 Hz bandpass) were also collected. EEG and EOG data were recorded for 5 min in continuous mode with 256 Hz sampling rate. In the naïve HIV subjects, these data were recorded at baseline (pre-treatment; T0) and after 5 months of cART (T5).
The recordings were carried out in the late morning. To control the state of vigilance, an expert continuously monitored subject's behavior and EEG-EOG signals. The subject was verbally alerted in the case of behavioral or EEG drowsiness. Specifically, the experimenter monitored the appearance of "tonic" theta rhythms, K-complexes and sleep spindles (behavior in the control subjects). Of note, a low percentage of subjects (much less than 10%) required such verbal prompt.
The 5 min duration of the EEG recording allowed the comparison of the present results with several previous EEG studies in neurological patients using recording periods shorter or equal to 5 min [bib_ref] Mapping distributed sources of cortical rhythms in mild Alzheimers disease: A multi-centric..., Babiloni [/bib_ref] [bib_ref] Sources of cortical rhythms change as a function of cognitive impairment in..., Babiloni [/bib_ref] [bib_ref] Resting EEG sources correlate with attentional span in mild cognitive impairment and..., Babiloni [/bib_ref] [bib_ref] Cortical sources of resting-state EEG rhythms are abnormal in naïve HIV subjects, Babiloni [/bib_ref] [bib_ref] Antiretroviral therapy effects on sources of cortical rhythms in HIV subjects: Responders..., Babiloni [/bib_ref] [bib_ref] Regional correlations between the EEG and oxygen metabolism in dementia of Alzheimer's..., Buchan [/bib_ref] [bib_ref] EEG power spectrum differences in early and late onset forms of Alzheimer's..., Pucci [/bib_ref] [bib_ref] Quantitative EEG changes in Alzheimer patients during long-term donepezil therapy, Rodriguez [/bib_ref] [bib_ref] EEG power changes are related to regional cerebral glucose metabolism in vascular..., Szelies [/bib_ref] or about 1 min [bib_ref] Dementia of the Alzheimer type: effects on the spontaneous EEG described by..., Dierks [/bib_ref] [bib_ref] Spatial pattern of cerebral glucose metabolism (PET) correlates with localization of intracerebral..., Dierks [/bib_ref] ; longer EEG epochs would have reduced data variability but increased risks for dropping vigilance and arousal.
## Preliminary eeg data analysis
The collected EEG data were off-line fragmented in consecutive periods (epochs) of 2 s. An automatic computerized procedure preliminarily identified the EEG epochs with ocular, muscular, and other types of artifacts (including verbal prompts). EEG epochs with sporadic blinking artifacts (threshold of ±50 μV) were much less than 10% of the total. These epochs were corrected by an autoregressive method [bib_ref] Computerized processing of EEG-EOG-EMG artifacts for multicentirc studies in EEG oscillations and..., Moretti [/bib_ref]. Afterwards, two independent experimenters (i.e. G.N and S.C.), blind to the diagnosis, manually confirmed the EEG epochs accepted for a further analysis. The EEG epochs with an incomplete removal of the blinking or other artifacts were manually rejected. Of note, a control analysis showed no difference (p N 0.05) in the percentage of the eyes blinks between the Healthy and the naïve HIV group (exact Fisher test, p N 0.05).
## Spectral analysis of the eeg data
A spectrum analysis (FFT, Welch technique, Hanning windowing function with no phase shift) computed power density of EEG rhythms with frequency resolution of 0.5 Hz. Absolute EEG power density values were normalized by the following procedure: 1) mean EEG power density was computed for all frequency bins and all electrodes; and 2) absolute power density values for all frequency bins at each electrode were divided by the mean power density computed at 1). The EEG frequency bands of interest were determined based on the identification of EEG power density peak in the range of alpha rhythms (e.g. 7-13 Hz) in any subject, the so-called individual alpha frequency peak (IAF peak; [bib_ref] Induced alpha band power changes in the human EEG and attention, Klimesch [/bib_ref]. In this line, the EEG frequency bands of interest were defined as follows: (i) delta, IAF −8 Hz to IAF −6 Hz; (ii) theta, IAF − 6 Hz to IAF − 4 Hz; (iii) alpha 1, IAF − 4 Hz to IAF − 2 Hz; (iv) alpha 2, IAF − 2 Hz to IAF; and (v) alpha 3, IAF to IAF +2 Hz. Of note, the mean IAF peak was 10.1 Hz (±0.1 SE) in naïve HIV group and 10.2 (± 0.1 SE) in the Healthy group. No statistical difference in the IAF peak was observed between the two groups (Student's t-test one-tailed, p N 0.05). For higher EEG band frequencies, standard fixed frequency bands were selected, as a beta power density peak was not clearly detectable in all subjects. Specifically, they were defined as beta 1 (13-20 Hz) and beta 2 (20-30 Hz).
## Cortical (loreta) source of eeg rhythms
The LORETA software (http://www.unizh.ch/keyinst/NewLORETA/ LORETA01.htm) was used for the estimation of the activity (i.e. neural current density) of cortical sources of the EEG rhythms [bib_ref] LORETA (low resolution brain electromagnetic tomography): new authentic 3D functional images of..., Pascual-Marqui [/bib_ref] [bib_ref] Low resolution brain electromagnetic tomography (LORETA) functional imaging in acute, neuroleptic-naïve, first-episode,..., Pascual-Marqui [/bib_ref]. LORETA is a functional imaging technique belonging to a family of linear inverse solution procedures like minimum norm solution, weighted resolution optimization or weighted minimum norm solution, which model 3D distributions of EEG sources [bib_ref] Constraining EEG-MEG source imaging with statistical neuroanatomy, Valdès [/bib_ref] [bib_ref] Low resolution brain electromagnetic tomography (LORETA) functional imaging in acute, neuroleptic-naïve, first-episode,..., Pascual-Marqui [/bib_ref] [bib_ref] Systemic regularization of linear inverse solutions of the EEG source localization problem, Phillips [/bib_ref] [bib_ref] A self-coherence enhancement algorithm and its application to enhancing three-dimensional source estimation..., Yao [/bib_ref]. Of note, LORETA has been successfully used in recent EEG studies on neurological subjects by the present experimental set up of the present study [bib_ref] Spatial pattern of cerebral glucose metabolism (PET) correlates with localization of intracerebral..., Dierks [/bib_ref] [bib_ref] Mapping distributed sources of cortical rhythms in mild Alzheimers disease: A multi-centric..., Babiloni [/bib_ref] [bib_ref] Sources of cortical rhythms change as a function of cognitive impairment in..., Babiloni [/bib_ref] [bib_ref] Resting EEG sources correlate with attentional span in mild cognitive impairment and..., Babiloni [/bib_ref] [bib_ref] Cortical sources of resting-state EEG rhythms are abnormal in naïve HIV subjects, Babiloni [/bib_ref] [bib_ref] Cortical sources of resting state EEG rhythms in "experienced" HIV subjects under..., Babiloni [/bib_ref] [bib_ref] Antiretroviral therapy effects on sources of cortical rhythms in HIV subjects: Responders..., Babiloni [/bib_ref]. Results of these studies suggest that LORETA provides reliable estimates of distributed cortical source activity of resting state EEG rhythms collected from 19 electrodes placed according to 10-20 montage [bib_ref] Effect of the 5-HT(1A) partial agonist buspirone on regional brain electrical activity..., Anderer [/bib_ref] [bib_ref] Non-invasive localization of P300 sources in normal aging and age-associated memory impairment, Anderer [/bib_ref] [bib_ref] Brain regions activated during an auditory discrimination task in insomniac postmenopausal patients..., Anderer [/bib_ref] [bib_ref] Reduced event-related current density in the anterior cingulate cortex in schizophrenia, Mulert [/bib_ref] [bib_ref] P300 and LORETA: comparison of normal subjects and schizophrenic patients, Winterer [/bib_ref] [bib_ref] Mapping distributed sources of cortical rhythms in mild Alzheimers disease: A multi-centric..., Babiloni [/bib_ref] [bib_ref] Sources of cortical rhythms change as a function of cognitive impairment in..., Babiloni [/bib_ref] [bib_ref] Resting EEG sources correlate with attentional span in mild cognitive impairment and..., Babiloni [/bib_ref] [bib_ref] Cortical sources of resting-state EEG rhythms are abnormal in naïve HIV subjects, Babiloni [/bib_ref] [bib_ref] Cortical sources of resting state EEG rhythms in "experienced" HIV subjects under..., Babiloni [/bib_ref] [bib_ref] Antiretroviral therapy effects on sources of cortical rhythms in HIV subjects: Responders..., Babiloni [/bib_ref] [bib_ref] Neurocortical electrical activity tomography in chronic schizophrenics, Veiga [/bib_ref]. Indeed, these resting state EEG rhythms are generated by largely distributed cortical sources whose EEG activity can be recorded by 10-20 system without spatial aliasing. Overall, the present results suggest that topographically widespread cortical sources of resting state delta and alpha rhythms reflect neurophysiologic abnormalities in naïve HIV subjects, at least at a group level.
LORETA provides 3D solutions to the EEG inverse problem (i.e. LORETA solutions). The LORETA solutions estimate the neural current density into a head model formed by 3 spherical compartments representing the electrical properties of the scalp, skull, and brain. The brain compartment is co-registered to the Talairach probability brain atlas. It includes cortical gray matter and hippocampus, digitated in the Brain Imaging Center of Montreal Neurological Institute. The brain compartment is constituted by 2394 voxels with 7 mm of spatial resolution. Any voxel contains an equivalent current dipole, which is fixed as position and orientation. The LORETA solutions estimate the current intensity of all equivalent current dipoles of the brain compartment to explain scalp EEG power density at scalp electrodes [bib_ref] LORETA (low resolution brain electromagnetic tomography): new authentic 3D functional images of..., Pascual-Marqui [/bib_ref].
Solutions of the EEG inverse problem are under-determined and illconditioned, when the number of spatial samples (electrodes) is lower than that of the unknown samples (current density at each voxel). For this reason, the cortical LORETA solutions (i.e. estimates of dipole current density at any voxel) predicting scalp EEG spectral power density were regularized to estimate maximally-smoothed rather than circumscribed EEG source activity [bib_ref] LORETA (low resolution brain electromagnetic tomography): new authentic 3D functional images of..., Pascual-Marqui [/bib_ref] [bib_ref] Low resolution brain electromagnetic tomography (LORETA) functional imaging in acute, neuroleptic-naïve, first-episode,..., Pascual-Marqui [/bib_ref]. Furthermore, these solutions were normalized averaging any estimated dipole current density at each voxel and frequency bin by the mean of the dipole current density computed across all frequencies (0.5-45 Hz) and voxels of the brain volume. This procedure of normalization typically fits EEG power density into a Gaussian distribution and reduces its inter-subject variability [bib_ref] Quantitative EEG: I. tecniques and problems of frequency analysis and topographic mapping, Nuwer [/bib_ref] [bib_ref] Regional differences in brain electrical activity in dementia: use of spectral power..., Leuchter [/bib_ref]. After this normalization, the LORETA solutions lost the original physical dimension and were represented by arbitrary units. In this scale, the value "1" was equal to the mean of the dipole current density at all frequencies (0.5-45 Hz) and voxels of the brain volume.
In line with the intrinsic low-spatial resolution of LORETA solutions, we developed a MATLAB software to average all LORETA solutions of Brodmann areas (BAs) belonging to a given cortical macro-region of interest (ROI). Specifically, we considered frontal, central, parietal, occipital, temporal, and limbic ROIs in the LORETA brain compartment. [fig_ref] Table 2: Brodmann areas included in the cortical regions of interest [/fig_ref] reports the list of BAs relative to any ROI used in the present study.
Finally, a main advantage of LORETA source estimation is that the respective contribution of several cortical sources of scalp EEG rhythms can be approximately disentangled. For example, EEG rhythms recorded at scalp parietal electrodes are the summation of EEG activity generated by parietal, occipital, and temporal source activity. LORETA uses a mathematical model of head taking into account the volume conductor effects and may disentangle the contribution of the difference cortical sources of EEG rhythms recorded at scalp parietal electrodes.
## Statistical analyses
Sixth main statistical sessions were performed by commercial tool STATISTICA 10 (StatSoft Inc., www.statsoft.com).
The first statistical session tested the hypothesis that the activity of EEG cortical (LORETA) sources did differ between the Healthy and naïve HIV groups. To this aim, an ANOVA used regional normalized LORETA solutions at T0 as a dependent variable (p b 0.05). The ANOVA factors were Group (Healthy, HIV; independent variable), Band (delta, theta, alpha 1, alpha 2, alpha3, beta 1, beta 2), and ROI (central, frontal, parietal, occipital, temporal, limbic). Mauchly's test evaluated the sphericity assumption. The degrees of freedom were corrected by Greenhouse-Geisser procedure when appropriate. The education was used as a covariate. Duncan test was used for post-hoc comparisons (p b 0.05 one-tailed). The hypothesis would be confirmed by the following two statistical results: (i) a statistical interaction effect including the factor Group (p b 0.05); and (ii) a post-hoc test indicating statistically significant differences between the normalized LORETA solutions with the pattern Healthy ≠ naïve HIV (Duncan test, p b 0.05, one-tailed).
The second statistical session tested the hypothesis that the z-score of the ratio between parietal delta and alpha 3 LORETA source activity (i.e. the "EEG marker") at T0 allows the identification of a relatively high percentage (about 40-50%) of naïve HIV individuals with statistical abnormality of EEG activity. To this aim, z-score was calculated for the EEG marker as follows:
[formula] z ¼ x−μ σ ; [/formula]
where "x" is the value of the EEG marker in a given naïve HIV subject; "μ" is the mean of the EEG marker in the reference Healthy group; and "σ" is the standard deviation of the EEG marker in that control group. Overall, z-score indicates the probability that any given naïve HIV subject belongs to the Healthy group in terms of the EEG marker. The statistical threshold was set at p b 0.05, one-tailed. This means that a given naïve HIV subject was considered as having a statistically deviant EEG marker ("EEG+") when z-score value of that EEG marker was higher than the threshold of p b 0.05 (one-tailed) with reference to the Healthy group. If this is not the case, naïve HIV individuals were denoted as "EEG−" (i.e. statistically normal EEG marker). The third statistical session tested the hypothesis that EEG+ status is relevant for the assessment of naïve HIV subjects. To this aim, the MMSE score, CD4 count, and VL at T0 were compared between the sub-group of naïve HIV subjects with EEG + and the sub-group of those with EEG− (Student test, p b 0.05 one-tailed). The working hypothesis would be confirmed by t-values indicating lower MMSE score, lower CD4 count, and higher VL in the sub-group of naïve HIV subjects with EEG + than in the sub-group of those with EEG − (p b 0.05 one-tailed).
The fourth statistical session tested the hypothesis that 5 months of cART induces an improvement of global cognitive status and/or biological markers in the two sub-groups of naïve HIV subjects (i.e. EEG+ and EEG−). To this aim, MMSE score, CD4 count, and VL in the two HIV subgroups were compared between T0 and T5 (Student test, p b 0.05 onetailed). The hypothesis would be confirmed by t-values indicating higher MMSE score, lower CD count, and higher VL at T5 than at T0 (p b 0.05 one-tailed; nonparametric test for the MMSE score).
The fifth statistical session tested the hypothesis that 5 months of cART induce a recovery of EEG cortical (LORETA) source activity only in the sub-group of naïve HIV subjects with EEG +. To this aim, two ANOVAs (one for the sub-group of naïve HIV subjects with EEG+ and one for the sub-group of naïve HIV subjects with EEG−) were computed using regional normalized LORETA solutions as a dependent variable (p b 0.05). The ANOVA factors were Time (T0, T5), Band (delta, theta, alpha 1, alpha 2, alpha 3, beta 1, and beta 2), and ROI (central, frontal, parietal, occipital, temporal, and limbic). Mauchly's test evaluated the sphericity assumption. Correction of the degrees of freedom was made with the Greenhouse-Geisser procedure when appropriate. Duncan test was used for post-hoc comparisons (p b 0.05). The hypothesis would be confirmed by the following three results: (i) a statistical interaction effect including the factor Time (p b 0.05) for the sub-group of naïve HIV subjects with EEG+; (ii) a post-hoc test indicating statistically significant differences of the regional normalized LORETA solutions with the pattern T0 ≠ T5 (Duncan test, p b 0.05, one-tailed) for the sub-group of naïve HIV subjects with EEG+; and (iii) no statistical interaction effect including the factors Time for the sub-group of naïve HIV subjects with EEG−.
The sixth statistical session tested the hypothesis that the activity of EEG cortical (LORETA) sources is related to biological markers of the HIV. To address this issue, we performed a correlation analysis by Pearson test between normalized LORETA solutions and CD4 count in the two sub-groups of − naïve HIV subjects (i.e. EEG + and EEG −; p b 0.05) at T0. The hypothesis would be confirmed by a statistical correlation (p b 0.05) in the sub-group of naïve HIV subjects with EEG+, and no statistical correlation (p N 0.05) in the sub-group of naïve HIV subjects with EEG−.
# Results
3.1. Cortical (LORETA) sources of EEG rhythms estimated before the therapy (T0): the Healthy vs. the naïve HIV group maps the grand average of the normalized LORETA solutions (i.e., normalized dipole current density at cortical voxels) modeling the distributed EEG cortical sources at delta, theta, alpha 1, alpha 2, alpha 3, beta 1, and beta 2 bands in the Healthy group and in the naïve HIV group before. These data refer to the period before the beginning of the cART (T0). The Healthy group showed alpha 2 and alpha 3 sources with the highest activity in the occipital regions. Delta, theta, and alpha 1 sources pointed to moderate activity when compared to the alpha 2 and alpha 3 source activity. Finally, the beta 1 and beta 2 sources had the lowest activity. When compared to the Healthy group, the naïve HIV group was characterized by higher parietal delta source activity and lower posterior alpha 2 and alpha 3 source activity.
The ANOVA on the normalized LORETA solutions pointed to a statistically significant interaction (F(30, 3120) = 14.649, p b 0.0001) among the factors Group (HIV and Healthy; independent variable), Band (beta 1, beta 2, alpha 1, alpha 2, alpha 3, theta, and delta), and ROI (frontal, central, temporal, parietal, occipital, and limbic). shows the normalized LORETA solutions for this statistical ANOVA interaction. These solutions modeled the activity of the cortical sources of EEG rhythms at the various frequency bands. In the figure, the source activity across the frequency bands has the shape of EEG relative power spectra.
Notably, profile and magnitude of the source activity differed in the naïve HIV group compared to the Healthy group. The highest values of the source activity were observed in posterior cortical regions at highfrequency alpha sub-bands (i.e. alpha 2 and 3). Duncan planned post-hoc testing showed that the parietal delta source activity was higher in the naïve HIV group than in the control Healthy group (p b 0.005). Furthermore, compared to the control Healthy group, the naïve HIV group showed a lower alpha 1 source activity in the limbic region (p b 0.05), as well as lower alpha 2 and alpha 3 source activity in parietal, occipital, temporal, and limbic regions (p b 0.00005).
The present results confirm that the parietal cortical sources of resting state eyes-closed delta and high-frequency alpha rhythms are . Grand average of the normalized low-resolution brain electromagnetic tomography (LORETA) solutions (i.e. normalized dipole current density at the cortical voxels) modeling the activity of distributed electroencephalographic (EEG) cortical sources at delta, theta, alpha 1, alpha 2, alpha 3, beta 1, and beta 2 bands in the Healthy and naïve HIV groups before the therapy (T0). The normalization of the LORETA solutions was obtained by computing the ratio between the LORETA current density at each voxel with the mean of the current density values averaged across all frequencies (0.5-45 Hz) and 2394 voxels of the brain source space. The color scale ranges from 0 to the maximum value of the normalized current density estimated at alpha 2 frequency band. . Mean regional normalized LORETA solutions relative to a statistical ANOVA interaction (F(30, 3120) = 14.649, p b 0.0001) among the factors Group (HIV and Healthy; independent variable), Band (delta, theta, alpha 1, alpha 2, alpha 3, beta 1, beta 2), and ROI (central, frontal, parietal, occipital, temporal, limbic). suitable candidates for the computation of a valid EEG marker for a neurophysiological assessment of naïve HIV patients before the beginning of the cART. represents the individual values of the ratio between parietal delta and alpha-3 source activity (i.e. the EEG marker) in all healthy and naïve HIV subjects before the therapy (i.e. T0). As expected, the EEG marker showed higher values (as a trend towards EEG abnormalities) in the HIV group than the Healthy group. plots the z-score of the EEG marker (i.e. the ratio between parietal delta and alpha-3 source activity) in the naïve HIV subjects before the therapy (T0). In the figure, the dashed horizontal separation line indicates the threshold of the statistical difference (p b 0.05 one-tailed) that splits all naïve HIV individuals into the following two sub-groups: the sub-group of the naïve HIV subjects with a statistically deviant EEG marker (i.e. EEG+, above the horizontal separation line) and the sub-group of those with a normal EEG marker (i.e. EEG −, below the horizontal separation line). Interestingly, there was a relatively high percentage (i.e. 50%) of the naïve HIV subjects with z-score values indicating a statistically deviant EEG marker (i.e. EEG+). draws the scatterplot between the z-score at T0 and T5 for all naïve HIV individuals. In this figure, the diagonal separation line divided the naïve HIV individuals showing a reduction (above the diagonal line those with an improvement of the EEG marker at T5) vs. an increase (down the diagonal line those with a worsening of the EEG marker at T5) in the z-score after 5 months of cART. Note the symmetrical and homogenous distribution of the naïve HIV individuals around the diagonal line until z-score values of about 2.5, suggesting a Gaussian distribution of the cART effects on the EEG marker. Instead, there was a prevalence of the naïve HIV individuals down the diagonal line with the z-score values higher than about 2.5, pointing to a beneficial effect of the cART on the EEG marker in the majority of those naïve HIV individuals. reports the values of a statistical analysis (p b 0.05, onetailed) computed to evaluate the differences between the two subgroups of the naïve HIV subjects (i.e. EEG + and EEG −) for each dependent variable of interest, namely the age, education, MMSE score, IAF, CD4 count, VL, Trial making test A, Trial making test B, Trial making test B-A, and Verbal fluency for letter and category. Mann-Whitney U test was used for the MMSE score, and t-test for the remaining variables (p b 0.05). Results showed that compared with the naïve HIV sub-group with EEG−, the naïve HIV sub-group with EEG+ exhibited a lower global cognitive status, as revealed by the MMSE score, and abnormal serological indexes such as lower CD4 count and higher VL (p b 0.05). The other variables showed no statistical difference (p N 0.05).reports the values of t-tests comparing the IAF, CD4 count, and VL between T0 (i.e. before the therapy) and T5 (i.e. after 5 months of therapy) in the naïve HIV sub-group with EEG + and that with EEG −. For the MMSE score, that table reports the results of the Wilcoxon test (p b 0.05). From T0 to T5, the two HIV sub-groups showed an improvement in the serological indexes. Namely, there was a higher CD4 count and a lower VL (p b 0.05) at T5 than T0. . A) Individual values of the EEG marker in all naïve HIV (before the therapy, T0) and healthy subjects. This EEG marker was defined as the ratio between parietal delta and alpha-3 source activity. The EEG source activity was estimated by the LORETA software. Specifically, this software estimated dipole current density in regions of interest of an average model of the human cerebral cortex (see the Methods). B) z-Score of the EEG marker in all naïve HIV individuals. The z-score values were computed using a control Healthy group as a normative reference. Dashed horizontal separation lines indicate the threshold of a statistical difference (p b 0.05, one-tailed) that splits all naïve HIV subjects in two sub-groups at T0: the sub-group of the naïve HIV individuals with a statistically deviant EEG marker (i.e. EEG +, above the line) and the sub-group of those with a normal EEG marker (i.e. EEG−, below the line). C) Scatterplot between the z-score at T0 and T5 in all naïve HIV subjects. In the scatterplot, the naïve HIV individuals with EEG− and EEG+ are represented as yellow and red circles, respectively.
## Effects of cart on clinical markers and cortical (loreta) sources of eeg rhythms
## Table 3
Means (±SE) of the demographic, clinical features, and neuropsychological data in a subgroup of naïve HIV individuals with a statistically deviant EEG marker (i.e. EEG+) and another sub-group of naïve HIV individuals with a normal EEG marker (i.e. EEG−). All data refer were collected before (T0) the beginning of a combined antiretroviral therapy (cART). The third column reports the results of a statistical analysis (p b 0.05, one-tailed) comparing the reported mean values between the two sub-groups (i.e. EEG− vs. EEG+). illustrates the grand average of the normalized LORETA solutions (i.e., normalized dipole current density at cortical voxels) modeling the EEG source activity at delta, theta, alpha 1, alpha 2, alpha 3, beta 1, and beta 2 bands in the two HIV sub-groups (i.e. EEG + and EEG −). These LORETA solutions referred to both T0 and T5. The HIV sub-group with EEG+ was characterized by a decrement of the delta source activity from T0 to T5 in widespread cortical regions. In parallel, there was a slight increase of alpha source activity from T0 to T5 in posterior regions. On the contrary, the HIV sub-group with EEG− showed similar EEG source activity at T0 and T5. This effect was true at all frequency bands. Of note, the alpha source activity was quite higher in the HIV sub-group with EEG− than in that with EEG+. This difference was expected based on the criterion (the z-score of the EEG marker) used for the formation of the two sub-groups of HIV subjects. It is irrelevant for the following statistical comparisons due to the "within" design of the planned ANOVAs in the two sub-groups of HIV subjects.
In the sub-group with EEG + before the therapy (N = 24), the ANOVA on the regional normalized LORETA solutions showed a statistically significant interaction (F(30, 690) = 3.187, p b 0.0001; see bottom) among the factors Time (T0, T5), Band (delta, theta, alpha 1, alpha 2, alpha 3, beta 1, beta 2), and ROI (frontal, central, temporal, limbic, parietal, occipital). Duncan planned post-hoc testing showed that compared with T0, T5 was associated with increased delta source activity in central, frontal, parietal, occipital, temporal, and limbic regions (p b 0.005). It was also related to increased limbic alpha 1 source activity (p b 0.05). Furthermore, there was an increase of source activity in parietal, occipital, and limbic alpha 2 and alpha 3 regions (p b 0.05). The ANOVA also showed a statistically significant interaction (F(6, 138) = 4.1882, p b 0.0005) between the factors Time and Band. Duncan planned post-hoc testing showed that compared to T0, T5 was associated with an increased delta activity in widespread sources (i.e. regardless the factor ROI) in the HIV subgroup with EEG+ (p b 0.005).
An exploratory correlation analysis (Pearson test, p b 0.05) was performed between the some EEG variables of interest (i.e. global delta, posterior alpha-3 source activity, and parietal delta/alpha-3 source activity) and the serological variables (i.e. viral load and CD4 count) before the therapy, when there was no interference of the cART. Results showed a statistically significant correlation only between the global delta source activity and the CD4 count (r = −0.51; p = 0.01), possibly due to a high inter-subject variability in the alpha source activity in the HIV subjects .
In the sub-group with EEG− (N = 24), the ANOVA on the regional normalized LORETA solutions showed no statistically significant effect of the cART (p N 0.05; see . To further control about the lack of a relationship between EEG source activity and HIV in these subjects, an exploratory correlation analysis (Pearson test, p b 0.05) was performed between some EEG variables of interest (i.e. global delta, posterior high-frequency alpha, and parietal delta/alpha-3 source activity) and the serological variables (i.e. viral load and CD4 count) before the therapy. Results showed no statistically significant correlation between these variables (p N 0.05).
In , some diagrams plot the z-score of the EEG marker (i.e. parietal delta/alpha-3 source activity) in the naïve HIV individuals (circles) of the two EEG+ and EEG− sub-groups before (T0) and after 5 months of cART (T5). Overall, most naïve HIV individuals with EEG− showed a quite stable z-score from T0 to T5 while several naïve HIV people with EEG+ unveiled a reduced z-score from T0 to T5 . In , other diagrams draw the scatterplots between the z-score at T0 and T5 in the naïve HIV individuals with EEG+ and EEG−. The naïve HIV persons with EEG− showed a symmetrical distribution around the diagonal separation line , indicating a Gaussian impact of the cART on the z-score. In contrast, most of the naïve HIV individuals with EEG+ were located down the diagonal separation line , pointing to an apparent reduction in the z-score at T5 (i.e. trend towards the normalization of the EEG marker). Other diagrams of the illustrate the scatterplots between the z-score at T0 and the difference in the z-score between T5 and T0 (i.e. T5 minus T0) as an index of the change due to the therapy. The horizontal separation lines divide the naïve HIV . Grand average of the normalized LORETA solutions modeling the distributed EEG cortical sources at delta, theta, alpha 1, alpha 2, alpha 3, beta 1, and beta 2 bands in the two subgroups of treatment-naïve HIV individuals (i.e. EEG+ and EEG−) before (T0) and after 5 months of cART (T5). Color scale ranges from 0 to the maximum value of the normalized current density estimated at alpha 3 frequency band.
individuals showing a reduction (down the horizontal separation line) vs. an increase (above the horizontal line) in the z-score after 5 months of cART. Note that the naïve HIV individuals with EEG− were symmetrically and homogenously distributed around the horizontal separation line, indicating a Gaussian distribution of the cART effect on the z-score. In contrast, most of the naïve HIV individuals with EEG+ were located down the horizontal separation line, pointing to a prominent reduction in the z-score at T5 (i.e. trend towards the normalization of the EEG marker).
The statistical analyses confirmed the above readouts. A Fisher-test showed that the T5 minus T0 variance was greater in the naïve HIV individuals with EEG + than those with EEG + .
Furthermore, a t-test unveiled that the T5 minus T0 values were more negative (greater z-score reduction) in the former than the latter HIV group (p b 0.05). Moreover, 49% (N = 11) of the naïve HIV individuals with EEG + (N = 24) changed to EEG − at T5 bottom-right quadrant) while only 8% (N = 2) of those with EEG − (N = 24) did . These converging findings suggest that 5 months of cART reduced the z-score (normalized the EEG marker) much more in the naïve EEG individuals with EEG+ than those with EEG − did. Indeed, the naïve EEG individuals with EEG + but not EEG− were supposed to be sensitive to the therapy.
The above results showed that there were two naïve HIV sub-groups of particular interest, namely the EEG + subjects who recovered to . Top: Mean regional normalized LORETA solutions relative to a statistically non-significant ANOVA interaction (p N 0.9) among the factors Time (T0, T5), Band (delta, theta, alpha 1, alpha 2, alpha 3, beta 1, beta 2), and ROI for the sub-group of treatment-naïve HIV subjects with a normal EEG marker (i.e. EEG−). Bottom: Mean regional normalized LORETA solutions relative to a statistically significant ANOVA interaction (F(30, 3120) = 14.649, p b 0.0001) among the factors Time, Band, and ROI for the sub-group of treatment-naïve HIV subjects with a statistically deviant EEG marker (i.e. EEG+).
EEG− (N = 11 out of 24) and those who remained EEG+ (N = 13 out of 24). How did these HIV sub-groups contribute to the difference T5 minus T0 in the CD4 count and VL in the global HIV group? To address this issue, two control t-tests were performed (p b 0.05). The respective dependent variables were the difference T5 minus T0 in the CD4 count and VL. For the VL, there was a greater difference T5 minus T0 (i.e. decrease) in the EEG + recovered to EEG − than the EEG + stable (p = 0.01). This finding suggests that most of the VL reduction in the whole HIV group was driven by the EEG+ subjects recovered to EEG −. For the CD4 count, no statistical effect was observed (p = 0.3), indicating that the recovered CD4 count in the whole HIV group was driven by both HIV sub-groups, namely the subjects with EEG+ recovered to EEG− and those with a stable EEG+. This statistical design was not used for the naïve EEG − subjects as practically all remained EEG− at T5 (i.e. N = 22 out of 24).
## Control analyses
To cross-validate LORETA cortical source estimates, we performed a control analysis at T0 by computing topographic maps of EEG power density for the most representative bands of interest (i.e. delta, alpha 2 and alpha 3) for the Healthy and the naïve HIV groups, separately . In line with normalization procedure used for LORETA source estimation, EEG power density at each scalp electrode was normalized to EEG power density averaged across all frequencies (0.5-45 Hz) and across all electrodes (N = 19). plots the statistical t maps of scalp EEG power density for delta, alpha 2 and alpha 3, respectively. These maps showed the statistical differences of the EEG power density between the Healthy and the naïve HIV group. Although these results globally confirmed the differences between the two groups based on the comparison of the corresponding LORETA cortical source estimates, the spatial resolution of EEG power density mapped at scalp level appeared to be lower than that at source level.
# Discussion
Previous evidence indicated that HIV patients showed a beneficial cART effect on brain function [bib_ref] HIV-associated neurocognitive disease continues in the antiretroviral era, Clifford [/bib_ref] as reduction in the VL and an increase in the CD4 cell count [bib_ref] A controlled trial of two nucleoside analogues plus indinavir in persons with..., Hammer [/bib_ref] [bib_ref] Continued CD4 cell count increases in HIV-infected adults experiencing 4 years of..., Hunt [/bib_ref] [bib_ref] Monocyte maturation, HIV susceptibility, and transmigration across the blood brain barrier are..., Williams [/bib_ref] [bib_ref] Cortical sources of resting state EEG rhythms in "experienced" HIV subjects under..., Babiloni [/bib_ref]. Furthermore, we demonstrated that before the cART, about . Scatterplot between the global delta LORETA current density and the CD4 count before the therapy in the naïve HIV subjects with a statistically deviant EEG marker (i.e. EEG+). . A) z-Score of the EEG marker in all naïve HIV subjects with EEG− at baseline (T0) and after 5 months of cART (T5). B) z-Score of the EEG marker in all naïve HIV subjects with EEG+ at T0 and T5. C) Scatterplot of the z-score at T0 and T5 in all naïve HIV subjects with EEG−. D) Scatterplot of the z-score at T0 and T5 in all naïve HIV subjects with EEG+. E) Scatterplot of the z-score at T0 and T5 minus T0 (T5-T0) in all naïve HIV subjects with EEG−. F) Scatterplot of the z-score at T0 and T5 minus T0 (T5-T0) in all naïve HIV subjects with EEG+. 50% of naïve HIV male single individuals were characterized by an abnormal activity of parietal cortical sources of resting state delta and high-frequency alpha rhythms [bib_ref] Abnormal cortical sources of resting state electroencephalographic rhythms in single treatment-naïve HIV..., Babiloni [/bib_ref]. Such abnormality was defined as the z-score of the ratio between the parietal delta and high-frequency alpha source activity in any individual naïve HIV male with respect to a reference group of healthy subjects. This ratio, then, was proposed as an EEG marker for the neurophysiological assessment of naïve HIV male subjects before the therapy [bib_ref] Abnormal cortical sources of resting state electroencephalographic rhythms in single treatment-naïve HIV..., Babiloni [/bib_ref].
Results of the present study confirmed those reported in a reference previous research in HIV individuals [bib_ref] Abnormal cortical sources of resting state electroencephalographic rhythms in single treatment-naïve HIV..., Babiloni [/bib_ref]. Compared with the Healthy group, the naïve HIV group pointed to a decrease in the low-and high-frequency (i.e. alpha 2 and alpha 3) source activity in several cortical regions (i.e. parietal, occipital, temporal and limbic regions). In contrast, the increase in the delta (i.e. parietal) and the decrease in the low-frequency alpha (i.e. limbic) source activity were regionally circumscribed. Of note 50% (N = 24) of the present naïve HIV subjects pointed to a z-score indicating a deviant EEG marker (i.e. EEG+) before the therapy.
At this early stage of the research, we can just speculate about the clinical neurophysiological meaning of these effects. Resting state eyes-closed alpha rhythms in healthy humans are prominent in widespread posterior cortical regions, especially in visuospatial and somatomotor areas [bib_ref] Event-related EEG/MEG synchronization and desynchronization: basic principles, Pfurtscheller [/bib_ref]. These rhythms are generated because of the cortical neural synchronization in distributed pyramidal neurons, which regulates low brain arousal and vigilance. This effect of synchronization actively inhibits the cerebral cortex, and may depend on the efficiency of the signal transmission and functional connectivity within cerebral neural networks involving the basal forebrain, ascending neurotransmitter systems, thalamus, and cerebral cortex [bib_ref] Event-related EEG/MEG synchronization and desynchronization: basic principles, Pfurtscheller [/bib_ref]. The functional meaning of low-and high-frequency alpha rhythms is supposed to be different. Low-frequency alpha rhythms (i.e. alpha 1 and alpha 2) may subserve subject's global arousal and alertness while high-frequency alpha rhythms (i.e. alpha 3) may reflect the event-related oscillation within peculiar neural systems engaged in the elaboration of semantic or sensorimotor information [bib_ref] Induced alpha band power changes in the human EEG and attention, Klimesch [/bib_ref] [bib_ref] EEG alpha and theta oscillations reflect cognitive and memory performance: a review..., Klimesch [/bib_ref]. Keeping in mind this theoretical premise, the present results hint that the naïve HIV subjects with EEG+ were characterized by a widespread alteration in the neural systems oscillating in broad posterior cortical regions that underpin several cognitive functions such as attention, sensorimotor, and semantic information processing. Instead, a prominent decrease in the lower-frequency alpha rhythm (i.e. alpha 1) was localized in the limbic region, and it might be associated with the regulation of global brain arousal and alertness. Limbic region includes anterior and posterior cingulate cortices and medial prefrontal cortex (BA 31, 32 and 33), which are part of the so-called default mode network (DMN). Previous evidence has shown that DMN operates to maintain a cortical idling and plays a causal role in the enhancement of dominant low-frequency alpha rhythms in quiet wakefulness [bib_ref] Resting-state modulation of α rhythms by interference with angular gyrus activity, Capotosto [/bib_ref].
As mentioned above, the present finding showed a noticeable increase in the delta source activity localized in the parietal region. This effect might be associated with vascular lesions in the periventricular white matter, which were found to affect neurotransmission from subcortical to parietal regions in patients with cerebrovascular lesions and cognitive deficits [bib_ref] Effect of white matter disease on functional connections in the aging brain, Leuchter [/bib_ref]. Future research should evaluate if naïve HIV subjects with EEG+ suffer from this kind of vascular lesions. In previous studies, HIV patients, especially those with cognitive deficits, were characterized by a white matter hyperintensity in MRIs as a reflection of similar cerebrovascular lesions [bib_ref] Neuroimaging in the brain in HIV-1-infected patients, Thurnher [/bib_ref].
As a novelty of the present study, here we tested the hypothesis that naïve HIV individuals with z-score indicating a deviant EEG marker (i.e. EEG+), but not HIV patients with normal EEG marker (EEG+), would show a normalization of the EEG source activity after 5 months of cART. This novel hypothesis was grounded on the concept that z-scores of an adequate EEG marker would be sensitive not only to HIV's detrimental effects on brain function, but also to the beneficial effects of cART. Indeed, the present naïve HIV subjects exhibited beneficial effects after 5 months of cART in both delta and high-frequency alpha source activity, although the major effect was observed on the former. Namely, the delta source activity decreased, while the high-frequency alpha source activity increased.
Importantly, as a further novelty, we report that almost half (49%) of these HIV individuals with EEG+ showed a statistical normalization of the proposed EEG marker (z-score) after 5 months of cART. Notably, no statistical difference was found before and after the therapy in the HIV patients with EEG−. Overall, these findings extend previous evidence suggesting that the beneficial cART effects on brain function are significant only in HIV patients with deviant EEG marker before the therapy.
An interesting result of the present study was the cART pronounced effect on delta source activity in the naïve HIV individuals with EEG+ status prior to therapy. A tentative explanation can be based on the physiological meaning of resting state delta rhythms in human subjects. These rhythms are negligible in the physiological condition of relaxed wakefulness, in contrast to dominant amplitude of posterior cortical alpha rhythms [bib_ref] Sources of cortical rhythms change as a function of cognitive impairment in..., Babiloni [/bib_ref] [bib_ref] Memory processes, brain oscillations and EEG synchronization, Klimesch [/bib_ref] [bib_ref] EEG alpha and theta oscillations reflect cognitive and memory performance: a review..., Klimesch [/bib_ref] [bib_ref] Event-related desynchronization in the alpha band and the processing of semantic information, Klimesch [/bib_ref] [bib_ref] Induced alpha band power changes in the human EEG and attention, Klimesch [/bib_ref] [bib_ref] Event-related EEG/MEG synchronization and desynchronization: basic principles, Pfurtscheller [/bib_ref]. Nevertheless, it has been proposed that cortical association areas physiologically . Spatial distribution of EEG power density for most representative EEG bands (i.e. delta, alpha 2 and alpha 3) at baseline (T0) and for the Healthy and the naïve HIV. . Topographical maps of statistical values at electrode (sensor) level for frequency bands that presented at baseline most statistical significant differences between Healthy and naïve HIV groups (i.e. delta, alpha 2 and alpha 3). generate delta rhythms as a correlate of integrative information processing across widely spatially distributed neural assemblies [bib_ref] Experimental connections between EEG data and the global wave theory, Nuñez [/bib_ref]. In this line, delta rhythms have been correlated positively with cortical metabolism in normal subjects, as revealed by fluorodeoxyglucose positron emission tomography [bib_ref] Neural networks for generation and suppression of alpha rhythm: a PET study, Sadato [/bib_ref] [bib_ref] Correlation of PET and qEEG in normal subjects, Alper [/bib_ref]. On the other hand, cortical delta rhythms have shown an abnormal high power in patients with cognitive impairment such as Alzheimer's disease, cerebrovascular disease, and Parkinson's disease [bib_ref] Mapping distributed sources of cortical rhythms in mild Alzheimers disease: A multi-centric..., Babiloni [/bib_ref] [bib_ref] Sources of cortical rhythms change as a function of cognitive impairment in..., Babiloni [/bib_ref] [bib_ref] Cortical sources of resting state electroencephalographic rhythms in Parkinson's disease related dementia..., Babiloni [/bib_ref]. In those with mild cognitive impairment and Alzheimer's disease, these abnormal delta rhythms were associated with white matter vascular lesions [bib_ref] Resting state cortical electroencephalographic rhythms and white matter vascular lesions in subjects..., Babiloni [/bib_ref] , atrophy of hippocampus and cortical gray matter [bib_ref] Global functional coupling of resting EEG rhythms is related to white-matter lesions..., Babiloni [/bib_ref] [bib_ref] Resting state cortical electroencephalographic rhythms are related to gray matter volume in..., Babiloni [/bib_ref] , and cortical hypometabolism [bib_ref] Quantitative EEG and perfusional single photon emission computed tomography correlation during long-term..., Rodriguez [/bib_ref]. Taking together these studies and the present data, we posit that such deviant activity of widespread delta cortical sources reflects both the HIV impact on brain function and the beneficial effects of cART. The underlying clinical neurophysiological mechanisms have yet to be fully elucidated. Neuronal-glia and inflammatory markers due to HIV might be associated with cerebrovascular, metabolic, and structural changes, as recently shown in vivo in humans by fluorodeoxyglucose positron emission tomography [bib_ref] Role of nuclear medicine in neuroHIV: PET, SPECT, and beyond, Sathekge [/bib_ref] as well as spectroscopic [bib_ref] Neuronal-glia markers by Magnetic Resonance Spectroscopy in HIV Before and After Combination..., Sailasuta [/bib_ref] , structural [bib_ref] Structural gray matter change early in male patients with HIV, Li [/bib_ref] , and functional [bib_ref] Aprr. Lower cognitive reserve in the aging human immunodeficiency virus-infected brain, Chang [/bib_ref] magnetic resonance imaging (MRI). In this line of reasoning, these changes might partially derail neural synchronization mechanisms and induce a partial "disconnection mode" of the cerebral cortex .
Keeping in mind the above data and considerations, one might argue that brain atrophy may have an unpredictable impact on the EEG cortical source estimation with a normal average brain model as that implemented in the LORETA freeware. Indeed, a fine EEG cortical source estimation should use a brain source model adapted to the individual cerebral atrophy. Furthermore, it should use a high spatial sampling of the potential distribution over the scalp (i.e. 64-128 electrodes). However, we think that the present methodological approach was adequate for an exploratory study for the following reasons. In the naïve HIV subjects of this study, the magnitude of the brain atrophy is expected to be low, although it can derange the cortical neural synchronization mechanisms generating rsEEG rhythms. In this line, the global cognitive performance, as revealed by the MMSE score, was significantly lower in the HIV group than the matched control group. However, the HIV group exhibited a global cognitive performance within the inferior normal limit, on average. This quasi-normal cognitive status in the HIV group may be associated with an initial phase of the HIV neuroinvasion. At that initial phase, the effects of the HIV neuroinvasion may be not so disruptive to produce marked white matter vascular lesions and grey matter atrophy [bib_ref] Neuroimaging in the brain in HIV-1-infected patients, Thurnher [/bib_ref] [bib_ref] Cerebrospinal fluid HIV escape associated with progressive neurologic dysfunction in patients on..., Peluso [/bib_ref] [bib_ref] Imaging of the brain in patients with human immunodeficiency virus infection, Gottumukkala [/bib_ref] to alter significantly the macroscopic volume of the cortical sources generating rsEEG rhythms. The impact of the brain atrophy may be furtherly mitigated from the expected widespread distribution of the rsEEG potential distributions. Indeed, several independent research groups have shown that rsEEG rhythms are generated by largely distributed cortical sources whose EEG activity can be recorded by 10-20 montage system without spatial aliasing. For this reason, several independent research groups used LORETA cortical source estimation from EEG data collected by a relatively low spatial sampling with 19 electrodes of 10-20 system [bib_ref] Effect of the 5-HT(1A) partial agonist buspirone on regional brain electrical activity..., Anderer [/bib_ref] [bib_ref] Non-invasive localization of P300 sources in normal aging and age-associated memory impairment, Anderer [/bib_ref] [bib_ref] Brain regions activated during an auditory discrimination task in insomniac postmenopausal patients..., Anderer [/bib_ref] [bib_ref] Mapping distributed sources of cortical rhythms in mild Alzheimers disease: A multi-centric..., Babiloni [/bib_ref] [bib_ref] Sources of cortical rhythms in adults during physiological aging: a multi-centric EEG..., Babiloni [/bib_ref] [bib_ref] Sources of cortical rhythms change as a function of cognitive impairment in..., Babiloni [/bib_ref] [bib_ref] Genotype (cystatin C) and EEG phenotype in Alzheimer disease and mild cognitive..., Babiloni [/bib_ref] [bib_ref] Apolipoprotein E and alpha brain rhythms in mild cognitive impairment: A multicentric..., Babiloni [/bib_ref] [bib_ref] Reduced event-related current density in the anterior cingulate cortex in schizophrenia, Mulert [/bib_ref] [bib_ref] Neurocortical electrical activity tomography in chronic schizophrenics, Veiga [/bib_ref] [bib_ref] P300 and LORETA: comparison of normal subjects and schizophrenic patients, Winterer [/bib_ref]. According to the theory of the signal sampling, a temporal or spatial sampling should be higher than the spatial and temporal (frequency) information content of the signal to be recorded, to avoid the so-called aliasing (i.e. a distortion in the signal reconstruction). The present methodology may fit this theory. It can be assumed that rsEEG rhythms have a low spatial information content due to a widespread cortical neural synchronization underpinning a low brain arousal and vigilance in a condition of quiet wakefulness. In the present study, we corroborated this assumption with the results of a control analysis performed to cross-validate the LORETA cortical source estimation. In that analysis, the spectral variables were computed from the rsEEG rhythms at the sensor (electrode) level. The results confirmed those obtained at the cortical source level. Overall, topographically widespread cortical sources of delta and alpha rhythms reflected neurophysiologic abnormalities in HIV patients, at least at the group level. It was concluded that the use of a normal brain source model and a traditional 10-20 electrode montage was adequate for this exploratory study, due to the general low spatial information content in the rsEEG rhythms and the expected low brain atrophy in the present naïve HIV subjects. However, future confirmatory studies should use MRI-based realistic head models in naïve HIV subjects for an EEG cortical source estimation performed from rsEEG rhythms recorded using 64-128 scalp electrodes.
## Specificity and clinical significance of the present results
An important remaining issue at present regards the specificity of cART impact on the parietal delta/alpha 3 EEG marker. The current study design prevents a definitive conclusion, as we did not have access to repeated EEG recordings in HIV subjects not receiving cART and control healthy subjects. However, two findings support a preliminary positive answer to this question. Indeed, the naïve HIV individuals showing a normal EEG marker before the therapy (i.e. EEG−) did not exhibit altered EEG source activity after 5 months of cART. Furthermore, such EEG marker was not correlated with the CD4 count. These findings suggest that the z-score procedure can detect not only the cases in which HIV produces deviant EEG source activity but also the cases in which HIV does not produce these effects. These findings support the idea that the present procedure can unveil the pathological effects of HIV on the neural synchronization mechanisms of cerebral cortex generating resting state EEG rhythms.
Another issue of the present results is about the clinical significance of the cART impact on the parietal delta/alpha 3 EEG marker. The present results suggest that the z-score of the EEG marker may enrich the instrumental multifaceted assessment of naïve HIV individuals. This marker should not be considered surrogate index of cognitive impairment in naïve HIV individuals, as the neuropsychological and neurophysiological dimensions of the subject's assessment are complementary rather than overlapping and redundant. These two dimensions explore different levels of brain functioning, namely neurophysiological mechanisms and cognitive performance. In this theoretical framework, the z-score of the EEG marker estimated before and after the cART may be useful to monitor the effect of the therapy on brain function in naïve HIV individuals. This neurophysiological assessment can provide additional useful information for the personalized clinical management of the patient. Let us consider the following example with two naïve HIV subjects after 5 months of cART: Both subjects were EEG+ before the therapy and had similar values in MMSE, autonomy in the daily life, and cognitive reserve as indexed by education years and intellectual occupations across the lifespan. One subject's EEG marker status changed from EEG + to EEG − at the follow-up, while the other subject remained EEG+ at follow-up. The evaluation of the EEG marker would suggest that, compared to the HIV subject who recovered to EEG − status after therapy, the HIV subject who remained in the EEG + status should receive more neuroprotective therapeutic resources and clinical attention to address his cART-resistant neurophysiological "frailty".
Finally, other issues were the recording of the EEG rhythms at only two time-points (pre-treatment and follow-up), with a relatively short follow-up (5 months), and an undetermined lengthiness of the HIV infection at the date of the pre-treatment EEG recording. It would be important to demonstrate a fine-grain time-course of the EEG marker in association with clinical evolution of the infection and cognitive status, especially in the naïve HIV subjects with pre-treatment EEG+. In these individuals, it can be hypothesized that the HIV infection was longer when compared to those with pre-treatment EEG −. Unfortunately, we could not test that hypothesis in the present study. We just knew the time of the serological diagnosis of HIV, which does not reliably reflect the time of the HIV infection. In several cases, that diagnosis is incidental. Furthermore, the event of the HIV contagion could not necessarily correspond to the episode told by the subject.
Keeping in mind the above considerations, future research should aim at testing the EEG marker with serial follow-ups across years, especially in the naïve HIV subjects with pre-treatment EEG+. Furthermore, we encourage the quest for the individual clinical cases in whom the event of the HIV contagion can be ascertained for the application of the present methodological approach.
# Conclusions
In the present study, we computed the z-score (p b 0.05, one-tailed) of an EEG marker to monitor the effects of 5 months of cART on brain function in naïve HIV (male) individuals with respect to a reference group of healthy subjects. The EEG marker was defined as the ratio between parietal delta and high-frequency alpha source activity. Before the therapy, the z-score of the EEG marker was statistically deviant (EEG +) in 50% of the naive HIV individuals. In this HIV sub-group with EEG +, delta source activity decreased, while high-frequency alpha source activity increased after 5 months of therapy. Furthermore, N40% of them showed a normalized EEG marker (EEG−) at that followup. These results suggest that this statistical procedure was able to identify treatment-naïve HIV single individuals with a deviant EEG source activity before the cART. Furthermore, it was able to monitor the normalization of this EEG source activity after 5 months of therapy in many of them.
[table] Table 2: Brodmann areas included in the cortical regions of interest (ROIs) of the present study. LORETA Brodmann areas into the regions of interest (ROIs) [/table]
|
Mesenchymal Stem Cell-Derived Extracellular Vesicles for Therapeutic Use and in Bioengineering Applications
Citation: McLaughlin, C.; Datta, P.; Singh, Y.P.; Lo, A.; Horchler, S.; Elcheva, I.A.; Ozbolat, I.T.; Ravnic, D.J.; Koduru, S.V. Mesenchymal Stem Cell-Derived Extracellular Vesicles for Therapeutic Use and in Bioengineering Applications. Cells 2022, 11, 3366. https://doi.
# Introduction
Extracellular vesicles (EVs) are extremely small, acellular, lipid-membrane enclosed vesicles that are secreted by cells into the extracellular space to modulate cellular communication and important physiological processes, including angiogenesis [bib_ref] Shedding light on the cell biology of extracellular vesicles, Van Niel [/bib_ref] [bib_ref] Extracellular Vesicles in Angiogenesis, Todorova [/bib_ref]. These heterogeneous groups of particles are released from a variety of cells during normal, stressed, and diseased conditions to optimize the local milieu for the circumstance [bib_ref] Extracellular vesicles: Exosomes, microvesicles, and friends, Raposo [/bib_ref]. EV size can range from 30 nm to more than 2 µm [bib_ref] Extracellular Vesicle Heterogeneity: Subpopulations, Isolation Techniques, and Diverse Functions in Cancer Progression, Willms [/bib_ref] [bib_ref] The potential roles of stem cell-derived extracellular vesicles as a therapeutic tool, Huang [/bib_ref] and they are mainly categorized into three major classes based on their biogenesis: exosomes (small EVs or sEVs; 30 nm to 150 nm in size), microvesicles (MVs; 150 nm to 1 µm in size) [bib_ref] Shedding light on the cell biology of extracellular vesicles, Van Niel [/bib_ref] [bib_ref] Classification, Functions, and Clinical Relevance of Extracellular Vesicles, Van Der Pol [/bib_ref] , and apoptotic bodies (500 nm to 2 µm) [bib_ref] Apoptotic Bodies: Particular Extracellular Vesicles Involved in Intercellular Communication, Battistelli [/bib_ref] [fig_ref] Figure 1: Main types of extracellular vesicles and their size, source, function, their biomarkers,... [/fig_ref].
EVs function by carrying and releasing a defined cargo payload, which can include various proteins (cytokines, chemokines, growth factors, interleukins (ILs)), transcription factors, and nucleic acids (DNA, mRNAs, microRNAs, piRNAs, LncRNAs, sn/snoRNAs, mtRNAs and circRNAs; [fig_ref] Figure 2: Schematic diagram of various sources of MSCs and their EVs that carrying... [/fig_ref] [bib_ref] Biological properties of extracellular vesicles and their physiological functions, Yáñez-Mó [/bib_ref] [bib_ref] Role of mesenchymal stem cell derived extracellular vesicles in autoimmunity: A systematic..., Wang [/bib_ref]. EVs are now also considered to play a vital role in cell-to-cell communication and may mediate many physiologic processes, such immunomodulation and angiogenesis. For example, during tissue development, remodeling, and wound repair, proangiogenic EVs can be released from endothelial cells, various It is now known that mesenchymal stem cells (MSC) also secrete a unique EV pool in response to the underlying microenvironment. Recent literature supports MSCs' function through a complex paracrine mechanism, mediated in part by EVs, which facilitate locoregional cell-to-cell communication [bib_ref] Mesenchymal stem cell-derived extracellular vesicles affect disease outcomes via transfer of microRNAs, Qiu [/bib_ref] [bib_ref] The extracellular vesicles-derived from mesenchymal stromal cells: A new therapeutic option in..., Nooshabadi [/bib_ref]. These MSC-derived heterogeneous EVs possess regenerative potential that can be developed for acellular therapeutics for various diseases [fig_ref] Figure 3: Schematic diagram of EV sources and their use in therapeutic and biomarker... [/fig_ref] [bib_ref] The potential roles of stem cell-derived extracellular vesicles as a therapeutic tool, Huang [/bib_ref] [bib_ref] Extracellular Vesicles in Skin Wound Healing, Narauskaitė [/bib_ref] [bib_ref] Extracellular Vesicles as Biomarkers and Therapeutic Tools: From Pre-Clinical to Clinical Applications, Ciferri [/bib_ref]. MSCs are adult multipotent progenitor cells found in various tissues, such as bone marrow, umbilical cord, and adipose tissue [bib_ref] Role of mesenchymal stem cell derived extracellular vesicles in autoimmunity: A systematic..., Wang [/bib_ref]. However, it is still unclear how this affects the corresponding EV functionality.
In this review, we provide a broad overview of what is known about EVs from specific MSC sources, limited instances of clinical translation, and how they have been applied to emerging bioengineering research. It is now known that mesenchymal stem cells (MSC) also secrete a unique EV pool in response to the underlying microenvironment. Recent literature supports MSCs' function through a complex paracrine mechanism, mediated in part by EVs, which facilitate locoregional cell-to-cell communication [bib_ref] Mesenchymal stem cell-derived extracellular vesicles affect disease outcomes via transfer of microRNAs, Qiu [/bib_ref] [bib_ref] The extracellular vesicles-derived from mesenchymal stromal cells: A new therapeutic option in..., Nooshabadi [/bib_ref]. These MSC-derived heterogeneous EVs possess regenerative potential that can be developed for acellular therapeutics for various diseases [fig_ref] Figure 3: Schematic diagram of EV sources and their use in therapeutic and biomarker... [/fig_ref] [bib_ref] The potential roles of stem cell-derived extracellular vesicles as a therapeutic tool, Huang [/bib_ref] [bib_ref] Extracellular Vesicles in Skin Wound Healing, Narauskaitė [/bib_ref] [bib_ref] Extracellular Vesicles as Biomarkers and Therapeutic Tools: From Pre-Clinical to Clinical Applications, Ciferri [/bib_ref]. MSCs are adult multipotent progenitor cells found in various tissues, such as bone marrow, umbilical cord, and adipose tissue [bib_ref] Role of mesenchymal stem cell derived extracellular vesicles in autoimmunity: A systematic..., Wang [/bib_ref]. However, it is still unclear how this affects the corresponding EV functionality.
## Source of stem cell extracellular vesicles (figure 4)
## Bone marrow mesenchymal stem cell-derived evs
Stem cells derived from bone marrow (BM) have long been utilized clinically to treat various hematopoietic malignancies. However, it has now been shown that even their EVs alone may hold therapeutic application. For example, BM-MSC-EVs have been found to be released under conditions of hypoxia to promote neoangiogenesis in vitro and in vivo [bib_ref] Extracellular vesicles derived from human bone marrow mesenchymal stem cells promote angiogenesis..., Bian [/bib_ref]. This is consistent with the finding that they increase cancer cell survival under stress and support breast tumor growth in vivo [bib_ref] Extracellular vesicles from bone marrow mesenchymal stem/stromal cells transport tumor regulatory microRNA,..., Vallabhaneni [/bib_ref]. There is evidence that these Reprinted from [bib_ref] Extracellular Vesicles as Biomarkers and Therapeutic Tools: From Pre-Clinical to Clinical Applications, Ciferri [/bib_ref].
In this review, we provide a broad overview of what is known about EVs from specific MSC sources, limited instances of clinical translation, and how they have been applied to emerging bioengineering research. [fig_ref] Figure 4: Schematic diagram of overall MSC isolation and their clinical applications in various... [/fig_ref]
## Source of stem cell extracellular vesicles
## Bone marrow mesenchymal stem cell-derived evs
Stem cells derived from bone marrow (BM) have long been utilized clinically to treat various hematopoietic malignancies. However, it has now been shown that even their EVs alone may hold therapeutic application. For example, BM-MSC-EVs have been found to be released under conditions of hypoxia to promote neoangiogenesis in vitro and in vivo [bib_ref] Extracellular vesicles derived from human bone marrow mesenchymal stem cells promote angiogenesis..., Bian [/bib_ref]. This is consistent with the finding that they increase cancer cell survival under stress and support breast tumor growth in vivo [bib_ref] Extracellular vesicles from bone marrow mesenchymal stem/stromal cells transport tumor regulatory microRNA,..., Vallabhaneni [/bib_ref]. There is evidence that these EVs regulate osteoblast activity and differentiation in vitro [bib_ref] Bone marrow stromal/stem cell-derived extracellular vesicles regulate osteoblast activity and differentiation in..., Qin [/bib_ref] and can even promote bone regeneration in vivo [bib_ref] Bone marrow stromal/stem cell-derived extracellular vesicles regulate osteoblast activity and differentiation in..., Qin [/bib_ref]. Others have also shown that BM-MSC-EVs can also enhance tendon healing, likely by modulating macrophage phenotypes and creating an anti-inflammatory environment [bib_ref] Extracellular vesicles from bone marrow-derived multipotent mesenchymal stromal cells regulate inflammation and..., Shi [/bib_ref]. BM-MSC-EVs are also thought to have a regenerative role on renal cells [bib_ref] Renal Regenerative Potential of Different Extracellular Vesicle Populations Derived from Bone Marrow..., Bruno [/bib_ref] and may play a role in attenuating renal fibrosis [bib_ref] Extracellular vesicles produced by bone marrow mesenchymal stem cells attenuate renal fibrosis,..., Shi [/bib_ref].
EVs regulate osteoblast activity and differentiation in vitro [bib_ref] Bone marrow stromal/stem cell-derived extracellular vesicles regulate osteoblast activity and differentiation in..., Qin [/bib_ref] and can even promote bone regeneration in vivo [bib_ref] Bone marrow stromal/stem cell-derived extracellular vesicles regulate osteoblast activity and differentiation in..., Qin [/bib_ref]. Others have also shown that BM-MSC-EVs can also enhance tendon healing, likely by modulating macrophage phenotypes and creating an anti-inflammatory environment [bib_ref] Extracellular vesicles from bone marrow-derived multipotent mesenchymal stromal cells regulate inflammation and..., Shi [/bib_ref]. BM-MSC-EVs are also thought to have a regenerative role on renal cells [bib_ref] Renal Regenerative Potential of Different Extracellular Vesicle Populations Derived from Bone Marrow..., Bruno [/bib_ref] and may play a role in attenuating renal fibrosis [bib_ref] Extracellular vesicles produced by bone marrow mesenchymal stem cells attenuate renal fibrosis,..., Shi [/bib_ref].
## Adipose mesenchymal stem cell-derived evs
Adipose-derived stem cell (ADSC) EVs are another subset found to have multiple biological effects, including angiogenesis, cell survival, apoptosis regulation, inflammation modulation, and tissue regeneration [bib_ref] Adipose-derived stem cell extracellular vesicles: A systematic review, Wong [/bib_ref]. Although ADSC-EV research is sparse, there is much excitement in the field. This is because ADSC retrieval is technically simple and yields abundant quantities. Furthermore, ADSC-EVs have a higher proliferative rate and lower senescence compared with other types of MSC-EV sourcing [bib_ref] Adipose-derived stem cell extracellular vesicles: A systematic review, Wong [/bib_ref]. Research studies have shown that ADSC-EVs may have some utility in in the treatment of neurologic diseases by decreasing pathologic proteins in Alzheimer's and Huntington's disease [bib_ref] Exosomes from adipose-derived stem cells ameliorate phenotype of Huntington's disease in vitro..., Lee [/bib_ref] [bib_ref] The exosome of adipose-derived stem cells reduces β-amyloid pathology and apoptosis of..., Lee [/bib_ref] and normalizing protein levels in Amyotrophic lateral sclerosis [bib_ref] Adipose-derived stem cell exosomes alleviate pathology of amyotrophic lateral sclerosis In Vitro, Lee [/bib_ref]. Other studies have shown a benefit in increasing the amount of muscle and peripheral nerve fibers [bib_ref] Therapeutic Potential of Human Adipose-Derived Stem Cell Exosomes in Stress Urinary Incontinence-An..., Ni [/bib_ref] [bib_ref] Exosomes derived from mesenchymal stem cells exert therapeutic effect in a rat..., Li [/bib_ref] , potentially playing a role in hepatic regeneration.
## Dental pulp stem cells
EVs have also been isolated from dental pulp stem cells (DPSCs). As an MSC source, DPSCs have an excellent proliferative capacity, can produce abundant amounts of EVs, and are thought to have advantages for osteogenic differentiation and decreased cell apoptosis and senescence [bib_ref] Transplantation of Hepatocyte Growth Factor-Modified Dental Pulp Stem Cells Prevents Bone Loss..., Kong [/bib_ref] [bib_ref] Comparative Analysis of Human Mesenchymal Stem Cells from Umbilical Cord, Dental Pulp,..., Ren [/bib_ref]. Dental pulp can be obtained from wisdom teeth, deciduous teeth, supernumerary teeth, and impacted teeth. DPSC-EVs have been shown to mitigate hematopoietic damage after radiation [bib_ref] Dental Pulp Stem Cell-Derived Extracellular Vesicles Mitigate Haematopoietic Damage after Radiation, Kong [/bib_ref] and may play a role in retinal neuroprotection and treatment [bib_ref] Extracellular vesicle therapy for retinal diseases, Mead [/bib_ref].
## Adipose mesenchymal stem cell-derived evs
Adipose-derived stem cell (ADSC) EVs are another subset found to have multiple biological effects, including angiogenesis, cell survival, apoptosis regulation, inflammation modulation, and tissue regeneration [bib_ref] Adipose-derived stem cell extracellular vesicles: A systematic review, Wong [/bib_ref]. Although ADSC-EV research is sparse, there is much excitement in the field. This is because ADSC retrieval is technically simple and yields abundant quantities. Furthermore, ADSC-EVs have a higher proliferative rate and lower senescence compared with other types of MSC-EV sourcing [bib_ref] Adipose-derived stem cell extracellular vesicles: A systematic review, Wong [/bib_ref]. Research studies have shown that ADSC-EVs may have some utility in in the treatment of neurologic diseases by decreasing pathologic proteins in Alzheimer's and Huntington's disease [bib_ref] Exosomes from adipose-derived stem cells ameliorate phenotype of Huntington's disease in vitro..., Lee [/bib_ref] [bib_ref] The exosome of adipose-derived stem cells reduces β-amyloid pathology and apoptosis of..., Lee [/bib_ref] and normalizing protein levels in Amyotrophic lateral sclerosis [bib_ref] Adipose-derived stem cell exosomes alleviate pathology of amyotrophic lateral sclerosis In Vitro, Lee [/bib_ref]. Other studies have shown a benefit in increasing the amount of muscle and peripheral nerve fibers [bib_ref] Therapeutic Potential of Human Adipose-Derived Stem Cell Exosomes in Stress Urinary Incontinence-An..., Ni [/bib_ref] [bib_ref] Exosomes derived from mesenchymal stem cells exert therapeutic effect in a rat..., Li [/bib_ref] , potentially playing a role in hepatic regeneration.
## Dental pulp stem cells
EVs have also been isolated from dental pulp stem cells (DPSCs). As an MSC source, DPSCs have an excellent proliferative capacity, can produce abundant amounts of EVs, and are thought to have advantages for osteogenic differentiation and decreased cell apoptosis and senescence [bib_ref] Transplantation of Hepatocyte Growth Factor-Modified Dental Pulp Stem Cells Prevents Bone Loss..., Kong [/bib_ref] [bib_ref] Comparative Analysis of Human Mesenchymal Stem Cells from Umbilical Cord, Dental Pulp,..., Ren [/bib_ref]. Dental pulp can be obtained from wisdom teeth, deciduous teeth, supernumerary teeth, and impacted teeth. DPSC-EVs have been shown to mitigate hematopoietic damage after radiation [bib_ref] Dental Pulp Stem Cell-Derived Extracellular Vesicles Mitigate Haematopoietic Damage after Radiation, Kong [/bib_ref] and may play a role in retinal neuroprotection and treatment [bib_ref] Extracellular vesicle therapy for retinal diseases, Mead [/bib_ref].
## Umbilical cord mesenchymal stem cell-derived evs
EVs have been isolated from stem cells of the human umbilical cord (hUC) as well. EV-MSC-hUCs (EV: extracellular vesicle; MSC: mesenchymal stem cells; hUC: human umbilical cord) have been shown to have protective effects in cerebral ischemia/reperfusion injury [bib_ref] Mechanism of Human Umbilical Cord Mesenchymal Stem Cells Derived-Extracellular Vesicle in Cerebral..., Wang [/bib_ref] and may aid in improving alveolarization and angiogenesis in lung injury [bib_ref] Human Umbilical Cord Mesenchymal Stem Cell-Derived Small Extracellular Vesicles Alleviate Lung Injury..., You [/bib_ref]. Similar to other MSC-EVs, they can promote cancer cell growth in some instances (specifically lung adenocarcinoma) [bib_ref] Human umbilical cord mesenchymal stem cell-derived extracellular vesicles promote lung adenocarcinoma growth..., Dong [/bib_ref] and have antiproliferative effects in other instances (en-dometrial) [bib_ref] Human Umbilical Cord Mesenchymal Stem Cell-Derived Extracellular Vesicles Inhibit Endometrial Cancer Cell..., Li [/bib_ref]. Other studies note the potential of hUC-MSC-EVs (EV: extracellular vesicle; MSC: mesenchymal stem cells; hUC: human umbilical cord) in improving nerve regeneration after transection [bib_ref] Extracellular vesicles from human umbilical cord mesenchymal stem cells improve nerve regeneration..., Ma [/bib_ref]. EVs also specifically isolated from Wharton's Jelly MSCs have also been found to have a diverse array of functions. For example, those that have been found to suppress CD4-expressing T cells [bib_ref] Extracellular Vesicles from Wharton's Jelly Mesenchymal Stem Cells Suppress CD4 Expressing T..., Crain [/bib_ref] are renally and neurally protective and can decrease oxidative stress [bib_ref] Mesenchymal Stromal Cell-Derived Extracellular Vesicles Protect Against Acute Kidney Injury Through Anti-Oxidation..., Zhang [/bib_ref] [bib_ref] Extracellular vesicles derived from human Wharton's jelly mesenchymal stem cells protect hippocampal..., Bodart-Santos [/bib_ref]. They also may play a role in stimulating muscle regeneration [bib_ref] Muscle functional recovery is driven by extracellular vesicles combined with muscle extracellular..., Magarotto [/bib_ref].
## Skeletal muscle stem cells
EVs isolated from skeletal muscle cells have also shown to efficiently promote myogenesis and muscle regeneration [bib_ref] Extracellular Vesicles from Skeletal Muscle Cells Efficiently Promote Myogenesis in Induced Pluripotent..., Baci [/bib_ref]. These sEVs contain numerous signals, including myogenic growth factors related to muscle development such as insulin-like growth factor, fibroblast growth factor 2, and platelet-derived growth factor [bib_ref] Exosomes from differentiating human skeletal muscle cells trigger myogenesis of stem cells..., Choi [/bib_ref]. In vivo studies have demonstrated the effects of these sEVs on reducing damaged areas.
## Cargo: evs carry a wide range of materials for molecular functions
EVs effectively serve as a signature of tissue type which may be reflected in their plentiful cargo assortment of bioactive proteins, lipids, and coding and noncoding RNAs. Release of this cargo allows for a means of intercellular communication and modification of recipient cells, including functional genetic material exchange [bib_ref] Extracellular vesicles: Exosomes, microvesicles, and friends, Raposo [/bib_ref] [bib_ref] Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic..., Valadi [/bib_ref]. Several databases have been created to organize the carrying cargo of EVs, including Vesiclepedia, ExoCarta [bib_ref] ExoCarta: A Web-Based Compendium of Exosomal Cargo, Keerthikumar [/bib_ref] , EVpedia, exoR BASE [bib_ref] A database of circRNA, lncRNA and mRNA in human blood exosomes, Li [/bib_ref] , and EVmiRNA [bib_ref] EVmiRNA: A database of miRNA profiling in extracellular vesicles, Liu [/bib_ref].
## Rna
The most well-studied and plentiful content of EVs is mRNA and microRNA (miRNA). miRNAs are 19-24-nucleotide noncoding RNA segments that serve as post-transcriptional regulators of up to 30% of mammalian genes [bib_ref] MicroRNAs: Genomics, biogenesis, mechanism, and function, Bartel [/bib_ref] [bib_ref] The widespread regulation of microRNA biogenesis, function and decay, Krol [/bib_ref]. It has been shown that miRNAs are selectively sorted into EVs, secreted through a ceramide-dependent pathway, and taken up by target cells identified by surface receptors and adhesion molecules that facilitate fusion and endocytosis [bib_ref] Mesenchymal stem cell-derived extracellular vesicles affect disease outcomes via transfer of microRNAs, Qiu [/bib_ref]. EVs have the ability to horizontally transfer genetic information to recipient cells, resulting in epigenetic modification [bib_ref] Endothelial progenitor cell-derived microvesicles activate an angiogenic program in endothelial cells by..., Deregibus [/bib_ref]. This finding has enlightened realms of diagnostics, as the genetic material in the pathophysiologic EV could serve as a signature, as well as unlock doors for therapeutics to alter pathologic cells with EVs as the vector. Moreover, EV mRNA and miRNA have demonstrated functionality in altering gene expression in recipient cells [bib_ref] Endothelial progenitor cell-derived microvesicles activate an angiogenic program in endothelial cells by..., Deregibus [/bib_ref] [bib_ref] Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic..., Valadi [/bib_ref] [bib_ref] Embryonic stem cell-derived microvesicles reprogram hematopoietic progenitors: Evidence for horizontal transfer of..., Ratajczak [/bib_ref].
## Proteins
Proteins carried in EVs are typically derived from the cytosol, plasma membrane, or endosome; proteins from specific organelles are typically spared from inclusion in the EV [bib_ref] Biogenesis, secretion, and intercellular interactions of exosomes and other extracellular vesicles, Colombo [/bib_ref]. Protein composition in EVs is not dependent on protein quantity in the cell of origin, but rather relies on a regimented protein-sorting mechanism during EV production, though the exact mechanism remains poorly understood [bib_ref] Proteomics, transcriptomics and lipidomics of exosomes and ectosomes, Choi [/bib_ref].
## Lipids
Lipid composition of EVs has been published with less frequency compared to protein and nucleic acid composition. Lipid composition of EVs depends on the cell of origin, as their lipid bilayer largely resembles the donor cell plasma membrane, including phospholipids, sphingomyelin, ganglioside, GM3, and cholesterol [bib_ref] Proteomics, transcriptomics and lipidomics of exosomes and ectosomes, Choi [/bib_ref]. Compared to donor cells, EVs tend to be enriched with phosphatidylserine, desaturated phosphatidylethanolamine, desaturated phosphatidylcholine, sphingomyelin, and cholesterol to provide structural support to the vesicle [bib_ref] Proteomics, transcriptomics and lipidomics of exosomes and ectosomes, Choi [/bib_ref]. While thought to be relatively inert, there have been increasing publications on biologic activity of lipid cargo, namely sphingomyelin-mediated tumor angiogenesis [bib_ref] Extracellular membrane vesicles from tumor cells promote angiogenesis via sphingomyelin, Kim [/bib_ref] and prostaglandin-mediated signaling pathways [bib_ref] Exosomes account for vesicle-mediated transcellular transport of activatable phospholipases and prostaglandins, Subra [/bib_ref].
## Mesenchymal stem cell evs for clinical use
MSC-EVs have been studied extensively in preclinical applications across organ systems [bib_ref] Extracellular Vesicles as Biomarkers and Therapeutic Tools: From Pre-Clinical to Clinical Applications, Ciferri [/bib_ref] [bib_ref] Mesenchymal Stem Cell Derived Extracellular Vesicles for Tissue Engineering and, Tsiapalis [/bib_ref]. They share the advantages of MSCs in terms of derivation from multiple tissue types and potential restoration of a wide variety of damaged tissue and pathophysiologic processes. However, the excitement surrounding MSC-EVs stems from their theoretical advantages over MSCs that have thus far panned out in the literature. First, their intercellular facilitations are paracrine in nature, suggesting that MSC-EVs can mediate the therapeutic effects of MSCs without requiring cellular integration [bib_ref] Exosomes and Their Therapeutic Potentials of Stem Cells, Han [/bib_ref]. Second, MSC-EVs do not have the ability to self-replicate, ameliorating the fear of uncontrolled cellular replication that accompanies the use of MSCs. While there is yet to be a recognized and standardized isolation procedure for MSC-EVs [bib_ref] The Need for Internationally Recognised Nomenclature and Stringent Purification Criteria, Simpson [/bib_ref] , the culture process is less invasive, their manufacturing is higher-yield, and the EVs are more stable in that their luminal contents are protected from degradation by their membranes [bib_ref] Exosomes in bodily fluids are a highly stable resource of disease biomarkers, Boukouris [/bib_ref]. In addition to pragmatic benefits, multiple studies have demonstrated a superior outcome when comparing MSC-EVs with MSCs [bib_ref] Induced Pluripotent Stem Cell (iPSC)-Derived Extracellular Vesicles Are Safer and More Effective..., Adamiak [/bib_ref] [bib_ref] Comparison of the therapeutic effect of amniotic fluid stem cells and their..., Zavatti [/bib_ref].
Due to their heterogeneity and involvement in a vast amount of physiologic and pathophysiologic processes, MSC-EVs have been investigated in a multitude of tissue types [bib_ref] Mesenchymal Stem Cell Derived Extracellular Vesicles for Tissue Engineering and, Tsiapalis [/bib_ref]. At the center of their potential lies their ability to mediate immune responses, hemostasis, and angiogenesis [bib_ref] Mesenchymal stem cells release exosomes that transfer miRNAs to endothelial cells and..., Gong [/bib_ref] , particularly in tissue types where endogenous regeneration is ineffective. Indeed, evidence spans across clinical disciplines, demonstrating that EVs mitigate destruction following myocardial ischemia, protect against renal injury, recover lung and liver injuries, and have regenerative potential against nerve injury, largely mediated by their anti-inflammatory, antiapoptotic effects, and proangiogenetic ability [bib_ref] Extracellular Vesicles: Evolving Factors in Stem Cell Biology, Nawaz [/bib_ref].
## Lung
## Lung injury
MSC-EVs carry specialized cargo that can be used to treat lung diseases. MSC-EVs contains pro/anti-inflammatory properties, reduce oxidative stress, and are remodeled in a variety of in vivo inflammatory lung disease models [fig_ref] Figure 5: MSC-EVs for potential therapeutics for lung diseases [/fig_ref] [bib_ref] Therapeutic Use of Extracellular Vesicles for Acute and Chronic Lung Disease, Worthington [/bib_ref]. Acute and chronic respiratory disorders are leading cause of the deaths worldwide and chronic lung disease with lowest curative option over a half billion people [bib_ref] Prevalence and attributable health burden of chronic respiratory diseases, 1990-2017: A systematic..., Soriano [/bib_ref]. Alternative treatments developed in clinics to fight this disease use MSC-EVs, which possess regenerative capacity [bib_ref] Effects of Mesenchymal Stromal Cell-Derived Extracellular Vesicles in Lung Diseases: Current Status..., Guo [/bib_ref]. Several studies have shown that they enhance acute and chronic conditions in preclinical settings [bib_ref] Mesenchymal stromal cell extracellular vesicles as therapy for acute and chronic respiratory..., Tieu [/bib_ref] [bib_ref] Therapeutic use of mesenchymal stem cell-derived extracellular vesicles in acute lung injury, Lee [/bib_ref] [bib_ref] Mesenchymal stem cell derived secretome and extracellular vesicles for acute lung injury..., Monsel [/bib_ref].
## Respiratory disease (covid-19)
The recent outbreak of coronavirus disease 2019 (COVID-19) has infected more than 267 million people, with more than 5 million deaths [bib_ref] Anti-SARS-CoV-2 effect of extracellular vesicles released from mesenchymal stem cells, Chutipongtanate [/bib_ref] ; this devastating illness has made us look for alternative treatments. MSC-derived EVs have the therapeutic potential to treat inflammatory and other diseases [fig_ref] Figure 6: Schematic diagram of cytokine storm produced by SARS-CoV2 infection [/fig_ref] [bib_ref] Mesenchymal stem cell-based therapy and exosomes in COVID-19: Current trends and prospects, Abdelgawad [/bib_ref]. There have been recent studies of MSCderived EVs for therapy to mitigate cytokine storms in COVID-19 and promote tissue repair in severely ill patients, as these EVs have immunomodulatory and regenerational capabilities. Chutipongtanate et al. showed that MSC-derived EVs induced COVID-19-infected lung epithelial cells to suppress viral replication and mitigate the production of infectious virions [bib_ref] Anti-SARS-CoV-2 effect of extracellular vesicles released from mesenchymal stem cells, Chutipongtanate [/bib_ref]. The SARS-CoV virus infects human cells by recognizing the angiotensinconverting enzyme (ACE2) receptor of host cells [bib_ref] The Potential Role of Extracellular Vesicles in COVID-19, Yan [/bib_ref]. There are anti-inflammatory drugs and a recombinant IL antagonist that can be used to treat this infection; however, these treatments help to improve recovery but not restore lung damage. MSC-EVs from the bone marrow, adipose, and umbilical cord showed improved recovery and survival; several clinical trials are in progress now [fig_ref] Table 1: List of registered clinical trials using MSC-EVs [/fig_ref].
## Respiratory disease (covid-19)
The recent outbreak of coronavirus disease 2019 (COVID-19) has infected more than 267 million people, with more than 5 million deaths [bib_ref] Anti-SARS-CoV-2 effect of extracellular vesicles released from mesenchymal stem cells, Chutipongtanate [/bib_ref] ; this devastating illness has made us look for alternative treatments. MSC-derived EVs have the therapeutic potential to treat inflammatory and other diseases [fig_ref] Figure 6: Schematic diagram of cytokine storm produced by SARS-CoV2 infection [/fig_ref] [bib_ref] Mesenchymal stem cell-based therapy and exosomes in COVID-19: Current trends and prospects, Abdelgawad [/bib_ref]. There have been recent studies of MSC-derived EVs for therapy to mitigate cytokine storms in COVID-19 and promote tissue repair in severely ill patients, as these EVs have immunomodulatory and regenerational capabilities. Chutipongtanate et al. showed that MSC-derived EVs induced COVID-19-infected lung epithelial cells to suppress viral replication and mitigate the production of infectious virions [bib_ref] Anti-SARS-CoV-2 effect of extracellular vesicles released from mesenchymal stem cells, Chutipongtanate [/bib_ref]. The SARS-CoV virus infects human cells by recognizing the angiotensin-converting enzyme (ACE2) receptor of host cells [bib_ref] The Potential Role of Extracellular Vesicles in COVID-19, Yan [/bib_ref]. There are anti-inflammatory drugs and a recombinant IL antagonist that can be used to treat this infection; however, these treatments help to improve recovery but not restore lung damage. MSC-EVs from the bone marrow, adipose, and umbilical cord showed improved recovery and survival; several clinical trials are in progress now [fig_ref] Table 1: List of registered clinical trials using MSC-EVs [/fig_ref].
## Cardiac
Treatment surrounding acute and chronic cardiac ischemia is currently focused on decreased oxygen demand and increasing supply. Regeneration of endogenous cardiomyocytes is an ineffective process and is thought to generate from preexisting cardiomyocytes [bib_ref] Mammalian heart renewal by pre-existing cardiomyocytes, Senyo [/bib_ref]. As such, once a scar is formed, local destruction and dilation results in irreversible heart failure. Due to the poor ability of endogenous cardiac myocytes to regenerate following ischemic insult, the study of MSC-EVs in cardiac regeneration is particularly promising in regard to its possible clinical benefit following myocardial ischemia. The first study demonstrating decreased cardiac infarct size was published in 2010, wherein Lai et al. utilized a mouse model of ischemia-reperfusion injury, revealing paracrine mediation of angiogenesis and decreased apoptosis through 50-100 nm sEVs [bib_ref] Exosome secreted by MSC reduces myocardial ischemia/reperfusion injury, Lai [/bib_ref]. This group went on to reveal increased ATP levels and decreased oxidative stress through the activation of the PI3k/Akt pathway via MSC-EVs in vivo [bib_ref] Mesenchymal stem cell-derived exosomes increase ATP levels, decrease oxidative stress and activate..., Arslan [/bib_ref]. Since these discoveries, EVs derived from bone marrow- [bib_ref] Extracellular vesicles derived from human bone marrow mesenchymal stem cells promote angiogenesis..., Bian [/bib_ref] , embryonic- [bib_ref] Embryonic Stem Cell-Derived Exosomes Promote Endogenous Repair Mechanisms and Enhance Cardiac Function..., Khan [/bib_ref] , umbilical cord- [bib_ref] Human umbilical cord mesenchymal stem cell derived exosomes encapsulated in functional peptide..., Han [/bib_ref] , and amniotic fluid [bib_ref] Reactivating endogenous mechanisms of cardiac regeneration via paracrine boosting using the human..., Balbi [/bib_ref] MSCs, as well as cardiac progenitor cells [bib_ref] Cardiac progenitor-derived exosomes protect ischemic myocardium from acute ischemia/reperfusion injury, Chen [/bib_ref] [bib_ref] Extracellular vesicles from human embryonic stem cell-derived cardiovascular progenitor cells promote cardiac..., Wu [/bib_ref] , have demonstrated cardioprotective effects in the context of ischemia. Inhibition of cardiomyocyte apoptosis was found to be partially mediated by the microvesicles' transfer of mi-R221 [bib_ref] Cardiomyocyte Protection by GATA-4 Gene Engineered Mesenchymal Stem Cells Is Partially Mediated..., Yu [/bib_ref]. sEVs from CD34 + stem cells and cardiac progenitor cells increased endothelial cell viability, migration, and angiogenesis [bib_ref] Exosomes From Human CD34 + Stem Cells Mediate Their Proangiogenic Paracrine Activity, Sahoo [/bib_ref] [bib_ref] Cardiomyocyte progenitor cell-derived exosomes stimulate migration of endothelial cells, Vrijsen [/bib_ref]. Further studies on the exact mediators of these effects demonstrate that sEV microRNA-21-5p increased cardiac contractility and calcium handling, likely via PI3K signaling [bib_ref] Exosomal microRNA-21-5p Mediates Mesenchymal Stem Cell Paracrine Effects on Human Cardiac Tissue..., Mayourian [/bib_ref].
## Renal
The first work describing the emergence of EVs in renal protection was published over a decade ago [bib_ref] Mesenchymal Stem Cell-Derived Microvesicles Protect Against Acute Tubular Injury, Bruno [/bib_ref] , in which BM-MSC-EVs conferred resistance of tubular cells to apoptosis and induction of tubular cell proliferation. Since this discovery, there has been an increasing number of publications investigating the role of EVs in tubular cell recovery in the context of both acute and chronic kidney injury. The therapeutic modalities for kidney injury are particularly intriguing, given that the current treatments for acute or chronic kidney injury, renal replacement, and dialysis, respectively, are supportive and invasive measures that do not address the underlying insult.
Independent of mechanism or inciting agent causing renal insult, MSC-EVs demonstrate amelioration of acute renal injury in rat and mice models through their proliferative, proangiogenic, antifibrotic, antiapoptotic, and anti-inflammatory effects [bib_ref] Stem Cell-Derived Extracellular Vesicles and Kidney Regeneration, Grange [/bib_ref]. These effects are mediated through the horizontal transfer of mRNA [bib_ref] Mesenchymal Stem Cell-Derived Microvesicles Protect Against Acute Tubular Injury, Bruno [/bib_ref] [bib_ref] Microvesicles Derived from Mesenchymal Stem Cells Enhance Survival in a Lethal Model..., Bruno [/bib_ref] , including human IGF-1 receptor mRNA [bib_ref] Transfer of Growth Factor Receptor mRNA Via Exosomes Unravels the Regenerative Effect..., Tomasoni [/bib_ref] , and miRNA [bib_ref] Transfer of microRNA-486-5p from human endothelial colony forming cell-derived exosomes reduces ischemic..., Viñas [/bib_ref] [bib_ref] Mesenchymal Stromal Cells Derived Extracellular Vesicles Ameliorate Acute Renal Ischemia Reperfusion Injury..., Gu [/bib_ref] from MSC-EVs to injured renal cells. There is also early evidence suggesting that this mode of communication is reciprocal, in that EVs from injured renal cells can communicate with stem cells in the setting of injury [bib_ref] Genetic communication by extracellular vesicles is an important mechanism underlying stem cell-based..., Zhao [/bib_ref].
The prediction and early detection of acute kidney injury (AKI) are challenging [bib_ref] Predicting acute kidney injury: Current status and future challenges, Pozzoli [/bib_ref] and current treatments focus on preventing the initial insult and slowing disease progression. Although the kidney has regenerative capacity after acute injury [bib_ref] Renal regeneration after acute kidney injury, Coelho [/bib_ref] , the mortality rate among patients with severe AKI remains high, and only a single episode of AKI predisposes the kidney to the development of chronic kidney disease (CKD) [bib_ref] Stem Cell-Derived Extracellular Vesicles and Kidney Regeneration, Grange [/bib_ref] [bib_ref] Renal Injury Repair: How About the Role of Stem Cells, Li [/bib_ref]. Renal replacement therapy can be used for treatment in the maintenance phase of AKI, but an increase in its dosage does not improve outcomes; thus, the study of MSC-EVs as a treatment modality is necessary. In 2008, for the first time, human BM-MSC was successfully used to prevent AKI induced by cisplatin in mice [bib_ref] Human Bone Marrow Mesenchymal Stem Cells Accelerate Recovery of Acute Renal Injury..., Morigi [/bib_ref]. Mice injected with BM-MSC showed preserved tubular epithelial and peritubular microvessel integrity, a significant reduction in apoptotic cells, and a prolonged overall survival [bib_ref] Human Bone Marrow Mesenchymal Stem Cells Accelerate Recovery of Acute Renal Injury..., Morigi [/bib_ref]. In a similar model of cisplatin-induced renal failure, indistinguishable levels of protection were seen after intravenous and intraperitoneal injections of BM-and ADSC-MSCs in adult mice. Isolated MSCs were cultured in conditioned medium and numerous epithelial colonies were found in all tubular regions, suggesting that endogenous tubular cells from all kidney compartments have the ability to proliferate and migrate [bib_ref] Stromal cells protect against acute tubular injury via an endocrine effect, Bi [/bib_ref]. Two recently published meta-analyses pooled the previously published preclinical work on EVs in AKI. The first included 45 studies focused on EVs derived exclusively from MSCs, determining that MSC-EVs produced an increased therapeutic effect on renal recovery compared to MSCconditioned medium [bib_ref] Extracellular vesicles derived from mesenchymal stromal cells may possess increased therapeutic potential..., Zhang [/bib_ref]. The second metanalysis analyzed 31 studies, including a variety of EV sources, and found stem cell-derived EVs to be equally effective compared to stem cells in the treatment of AKI. Importantly, the analysis also demonstrated that EV source and delivery dose to have independent effects on the efficacy of EVs [bib_ref] Extracellular vesicles for acute kidney injury in preclinical rodent models: A meta-analysis, Liu [/bib_ref].
EVs also show emerging promise in chronic kidney disease (CKD) given that their antifibrotic and anti-inflammatory effects that could mitigate ongoing damage. In a rat model of diabetic nephropathy, urine-derived stem cell sEVs prevented the progression of diabetic kidney injury via a reduction in microalbumin secretion, decreased caspase-3 overexpression, and the prevention of podocyte and tubular cell apoptosis [bib_ref] Exosomes secreted by human urine-derived stem cells could prevent kidney complications from..., Jiang [/bib_ref]. Again, in a rat model of diabetic-induced nephropathy, serial injections of MSC-EVs significantly improved renal functional parameters, including serum creatinine (Cr) and blood urea nitrogen (BUN), and slowed down the progression of fibrosis histologically [bib_ref] Stem cell-derived extracellular vesicles inhibit and revert fibrosis progression in a mouse..., Grange [/bib_ref]. These findings translated clinically as depicted by the first clinical trial, which investigated two doses of hUCMSC-derived EVs in 20 patients with stage III and IV CKD. The results were promising, demonstrating significant improvement in serum Cr and BUN in the experimental group compared to the placebo group. Additionally, participants treated with MSC-EVs demonstrated decreased serum concentration of proinflammatory TNF-a and increased concentration of anti-inflammatory TGF-b1 and IL-10, supporting the hypothesis of immune regulation in mediating improvement of renal function [bib_ref] Umbilical cord mesenchymal stem cells derived extracellular vesicles can safely ameliorate the..., Nassar [/bib_ref].
Another related but distinct clinical implication of MSC-EVs is their emerging role in solid organ transplant conditioning. In organ donation after cardiac death (DCD), organs are transplanted from a deceased donor to a living donor. During the transport process, organs are perfused with a solution to improve viability, but does not prevent ischemic injury. Compared to rat DCD kidney cold-perfused with standard protocols, those supplemented with solutions supplemented with MSC-EVs demonstrated upregulation of cell energy metabolism and ion membrane transport and decreased effluent material associated with global ischemic damage [bib_ref] Perfusion of isolated rat kidney with Mesenchymal Stromal Cells/Extracellular Vesicles prevents ischaemic..., Gregorini [/bib_ref]. Taken together, supplementing ischemic organs in preparation for transplant with MSC-EVs may preserve viability and decrease reperfusion injury.
## Cartilage
Osteoarthritis (OA) is the most prevalent chronic joint disease that irreversibly destroys the cartilage matrix. The precise molecular mechanisms involved in this degradation and the development of OA are poorly understood [bib_ref] Osteoarthritis Pathogenesis: A Review of Molecular Mechanisms, Xia [/bib_ref]. Current treatments are mainly directed toward relieving the symptoms of OA and include the use of nonsteroid antiinflammatory drugs (NSAID) and pain medications [bib_ref] Efficacy and safety of topical NSAIDs in the management of osteoarthritis: Evidence..., Rannou [/bib_ref]. While topical and oral NSAIDs have a moderate effect on pain relief, these therapies have many side effects and they cannot stop or reverse the ongoing cartilage degeneration [bib_ref] Efficacy and safety of topical NSAIDs in the management of osteoarthritis: Evidence..., Rannou [/bib_ref] [bib_ref] Stem cell-based therapies for osteoarthritis: Challenges and opportunities, Diekman [/bib_ref]. The search for new beneficial interventions to decelerate the progression of OA or stop cartilage degeneration except for total joint replacement surgery is an imperative issue [bib_ref] Osteoarthritis Pathogenesis: A Review of Molecular Mechanisms, Xia [/bib_ref].
Studies demonstrating the chondroprotective and anti-inflammatory effects of MSC-EVs are particularly intriguing in the context of OA, the most prevalent rheumatic disease characterized by degradation of cartilage [bib_ref] Mesenchymal stem cell-derived exosomes: A new therapeutic approach to osteoarthritis?, Mianehsaz [/bib_ref]. Historically, osteoarthritis is treated symptomatically with systemic anti-inflammatories or immobilization, or surgically through the removal of the affected joint or joint fusion as to eliminate the source of pain and disability. Promising results in clinical trials demonstrate MSCs ability to regenerate cartilage. A year following the intervention, there was no stem cell DNA detected in regenerative tissue, suggesting the role of MSCs as modulators with paracrine mechanisms, spurring further interest in MSC-EVs and sEVs [bib_ref] Mesenchymal stem cell-derived exosomes: A new therapeutic approach to osteoarthritis?, Mianehsaz [/bib_ref] [bib_ref] Allogeneic Mesenchymal Stem Cells Stimulate Cartilage Regeneration and Are Safe for Single-Stage..., De Windt [/bib_ref].
Preclinical studies demonstrated that human and murine bone marrow-derived [bib_ref] Mesenchymal Stromal/stem Cell-derived Extracellular Vesicles Promote Human Cartilage Regeneration In Vitro, Vonk [/bib_ref] and human embryonic stem cell-derived MSC-EVs and sEVs [bib_ref] Exosomes derived from human embryonic mesenchymal stem cells promote osteochondral regeneration, Zhang [/bib_ref] promote cartilage regeneration. Human BM-MSC-EVs diminished the TNFa upregulation of COX2 when cocultured with OA chondrocytes and promoted cartilage regeneration [bib_ref] Mesenchymal Stromal/stem Cell-derived Extracellular Vesicles Promote Human Cartilage Regeneration In Vitro, Vonk [/bib_ref]. In OA-like chondrocytes, murine BMMSCs reinstated the expression of chondrocyte markers and decreased catabolic and inflammatory markers in vitro, amounting to chondroprotective effects in vivo. Zhang et al. published multiple studies utilizing human embryonic MSC sEVs in OA rat models. They found exosome CD-73 mediation of the AKT and ERK signaling pathways to manifest increased cellular proliferation and infiltration, matrix synthesis, and cartilage regeneration in murine knee and hip OA chondrocytes and distal femur osteochondral defects [bib_ref] Exosomes derived from human embryonic mesenchymal stem cells promote osteochondral regeneration, Zhang [/bib_ref] [bib_ref] MSC exosomes mediate cartilage repair by enhancing proliferation, attenuating apoptosis and modulating..., Zhang [/bib_ref]. These findings were further elucidated in murine temporomandibular joint OA, in which there was exosome-mediated repair of matrix expression, subchondral bone architecture, and joint restoration, again through AKT and ERK signaling [bib_ref] MSC exosomes alleviate temporomandibular joint osteoarthritis by attenuating inflammation and restoring matrix..., Zhang [/bib_ref]. Intra-articular injection of ESC-MSC sEVs into surgically destabilized mouse knees decreased cartilage destruction and matrix degradation through an increase in type II collagen synthesis [bib_ref] Exosomes from embryonic mesenchymal stem cells alleviate osteoarthritis through balancing synthesis and..., Wang [/bib_ref]. MSC-EVs have been shown to protect cartilage and bone degradation by increasing chondrocyte marker expression such as type II collagen and aggrecan, decreasing catabolic markers such as matrix metalloproteinase-13 (MMP-13) and ADAMTS5, decreasing inflammatory markers, and protecting chondrocytes from apoptosis [bib_ref] Mesenchymal stem cell-derived exosomes: A new therapeutic approach to osteoarthritis?, Mianehsaz [/bib_ref]. Tofino-Vian et al. supported this finding by treating OA chondrocytes with AD-MSC-EVs, which showed a decrease in MMP activity and MMP-13 expression, while increasing cytokine IL-10 and collagen expression [bib_ref] Microvesicles from Human Adipose Tissue-Derived Mesenchymal Stem Cells as a New Protective..., Tofiño-Vian [/bib_ref]. These regenerative and immunoregulatory properties make BM-MSC-EVs a favorable choice as an ideal therapy for OA. Given these promising preclinical results, the first clinical trial has been initiated, investigating osteochondral explants from arthroplasty patients treated with adipose-derived MSC-EVs in order to validate a cell-free approach in an ex vivo OA model (NCT04223622).
## Wound healing
Despite advances in wound management, almost 50% of chronic wounds are resistant to treatment [bib_ref] Stem cells in cutaneous wound healing, Cha [/bib_ref]. MSC-EVs have anti-inflammatory, antiaging, and wound-healing effects, as shown in numerous studies performed in vitro and in vivo models. EVs could have a massive effect on treatment of skin and wound healing. It is accepted that the basic mechanism of wound healing is inflammation.
The effect of menstrual blood-derived MSC-EVs on wound healing in diabetic mice caused a decrease in inflammation via induced macrophage polarization from a proinflammatory M1 to M2, an increase in neoangiogenesis through vascular endothelial growth factor A upregulation, and activation of the NF-kB signaling pathway [bib_ref] Promising effects of exosomes isolated from menstrual blood-derived mesenchymal stem cell on..., Dalirfardouei [/bib_ref]. In addition, the mice had a significant increase in wound closure and possibly less scar formation [bib_ref] Promising effects of exosomes isolated from menstrual blood-derived mesenchymal stem cell on..., Dalirfardouei [/bib_ref]. This finding was supported by the use of MSCs preconditioned with LPS in a cutaneous wound model in diabetic rats, which showed a well-timed transition from M1 to M2, with the resolution of chronic inflammation and curative effects on wound healing [bib_ref] LPS-preconditioned mesenchymal stromal cells modify macrophage polarization for resolution of chronic inflammation..., Ti [/bib_ref]. MSC-EVs can affect T cells, promoting an anti-inflammatory state. In a study by Du et al., peripheral blood mononuclear cells (PBMCs) were isolated from asthmatic patients and cultured with BM-MSCs [bib_ref] Mesenchymal stem cell exosomes promote immunosuppression of regulatory T cells in asthma, Du [/bib_ref]. The MSC-EVs showed an upregulation of IL-10 and TGF-beta 1 in PBMCs, leading to proliferation and an increase in the immunosuppression capacity of regulatory T cells, which play an important role in the development of asthma [bib_ref] Mesenchymal stem cell exosomes promote immunosuppression of regulatory T cells in asthma, Du [/bib_ref]. Wound-healing effects were seen in a diabetic foot ulcer model in rats where ADSC-EVs increased granulation tissue formation, blood vessel density, and growth factor expression, which accelerates cutaneous wound healing by decreasing the ulcerated area and relative fibrosis [bib_ref] Exosomes from adipose-derived stem cells overexpressing Nrf2 accelerate cutaneous wound healing by..., Li [/bib_ref]. Topical administration of a gel containing MSC-EVs accelerated skin wound closure in rats [bib_ref] Extracellular Vesicles from Adipose-Derived Mesenchymal Stem/Stromal Cells Accelerate Migration and Activate AKT..., Ferreira [/bib_ref]. This group went on to show that exposure to ADSC-MSC-EVs causes increased migration and proliferation of fibroblasts and keratinocytes, as well as activation of the AKT signaling pathway to help promote wound healing [bib_ref] Extracellular Vesicles from Adipose-Derived Mesenchymal Stem/Stromal Cells Accelerate Migration and Activate AKT..., Ferreira [/bib_ref]. MSC-EVs have been used to induce hair growth [bib_ref] Hair growth promoting effects of adipose tissue-derived stem cells, Won [/bib_ref] , help with epidermal barrier repair [bib_ref] Exosomes from Human Adipose Tissue-Derived Mesenchymal Stem Cells Promote Epidermal Barrier Repair..., Shin [/bib_ref] , and have beneficiary effects on the healing of irradiated wounds in mice.
In mammalian soft tissue, the physiologic process of wound healing occurs in three overlapping phases with the primary goal of reestablishing a functional skin barrier for protection from the environment. The first inflammatory phase involves hemostasis through activation of the clotting cascade, infiltration of monocyte-derived macrophages, and subsequent cytokine release to signal fibroblast infiltration to carry out the next phase. During the second proliferative phase, granulation tissue is formed via fibroblast creation of extracellular matrix that serves as the scaffold for re-epithelialization and endothelial cell mediation of angiogenesis. The final remodeling phase consists of collagen reorganization from type III to type I; wound contraction mediated by myofibroblasts. Wounds that are unable to progress normally through this dynamic process develop into chronic wounds if the cycle is inadequate, or excessive scars if the cycle is uncontrolled. Inflammation is involved in all phases of wound healing and has a direct effect on scar formation and pathology. Several preclinical studies investigated the role MSC-EVs and sEVs throughout the stages of wound healing to aid in the healing of chronic wounds and decreasing the formation of scar tissue.
In an in vitro model of wound healing, it was found that BM-MSC sEVs caused dosedependent proliferation of fibroblasts in normal and chronic wounds and increased tube formation of endothelial cells through the activation of AKT, ERK, and STAT3 pathways and induction of growth factors [bib_ref] Mesenchymal Stem Cell Exosomes Induce Proliferation and Migration of Normal and Chronic..., Shabbir [/bib_ref]. Multiple studies focused on EVs and sEVs sourced from ADSCs. ADSC-derived sEVs demonstrated dose-dependent increases in fibroblast cell proliferation and migration as well as collagen synthesis via the PI3K/AKT pathway [bib_ref] Extracellular Vesicles from Adipose-Derived Mesenchymal Stem/Stromal Cells Accelerate Migration and Activate AKT..., Ferreira [/bib_ref] [bib_ref] Cell-free therapy based on adipose tissue stem cell-derived exosomes promotes wound healing..., Zhang [/bib_ref]. Further, for full-thickness incisions in a mouse model, ADSC sEVs expedited healing time and reduced scar formation [bib_ref] Cell-free therapy based on adipose tissue stem cell-derived exosomes promotes wound healing..., Zhang [/bib_ref]. The reduction in scar formation mediated by ADSC-derived sEVs may be through the regulation of collagen type III and I ratios and reduction of fibroblast differentiation into myofibroblasts. Additionally, one study found a long noncoding RNA MALAT1 (metastasis-associated lung adenocarcinoma transcript-1) in ADSC-derived sEVs to be responsible for fibroblast cell migration in the context of ischemic wounds [bib_ref] Human Adipose-Derived Stem Cell Conditioned Media and Exosomes Containing MALAT1 Promote Human..., Cooper [/bib_ref]. MSCs can affect the natural process of aging. Hu et al. looked at the antiaging efficacy of human dermal fibroblast (HDF) EVs derived from 3D spheroids in a mouse photo aging model [bib_ref] Needle-Free Injection of Exosomes Derived from Human Dermal Fibroblast Spheroids Ameliorates Skin..., Hu [/bib_ref]. An increase in procollagen type I expression, a decrease in MMP-1 expression via downregulation of TNF-alpha, and upregulation of TGF-beta were seen showing the anti-skin-aging properties of MSCs [bib_ref] Needle-Free Injection of Exosomes Derived from Human Dermal Fibroblast Spheroids Ameliorates Skin..., Hu [/bib_ref]. Nicotinamide adenine dinucleotide (NAD+) declines with age and plays a role in many important diseases of aging [bib_ref] NAD+ and sirtuins in aging and disease, Imai [/bib_ref]. Yoshida et al. interestingly found that extracellular nicotinamide phosphoribosyltransferase (eNAMPT) is carried in EVs and enhances NAD+ biosynthesis [bib_ref] Extracellular Vesicle-Contained eNAMPT Delays Aging and Extends Lifespan in Mice, Yoshida [/bib_ref]. eNAMPT-containing EVs were isolated from young mice and injected into aged mice, resulting in enhanced wheel-running activity and an extended life span in the older mice [bib_ref] Extracellular Vesicle-Contained eNAMPT Delays Aging and Extends Lifespan in Mice, Yoshida [/bib_ref].
Investigation of pluripotent stem cell-derived MSC sEVs on wound sites in a rat model revealed that MSC sEVs expedited re-epithelization, decreased scar size, promoted collagen formation, and increased the formation and maturation of blood vessels in the wound. This was hypothesized to be due to the stimulation of HDFs and increase in collagen and elastin secretion in associated in vitro models [bib_ref] Exosomes released from human induced pluripotent stem cells-derived MSCs facilitate cutaneous wound..., Zhang [/bib_ref]. Further, human umbilical cord MSC-derived exosome accelerates re-epithelization through WNT4 protein activation of ß-catenin signaling in a rat skin second-degree burn model [bib_ref] HucMSC-Exosome Mediated-Wnt4 Signaling Is Required for Cutaneous Wound Healing, Zhang [/bib_ref]. The effect of sEVs and EVs in clinical wound healing is emerging, with the first clinical trials investigating nanoparticles from Wharton's Jelly MSC-conditioned media on chronic ulcerative wounds completing recruitment (NCT04134676) and another investigating the safety and efficacy of MSC-EVs in chronic dystrophic epidermolysis bullosa wounds (NCT04173650). Taken together, nanoparticles demonstrate significant promise in preclinical studies in their ability to optimize the characteristics of fibroblasts-the workhorses of the wound-healing cycle-to aid in the healing of chronic wounds and decrease scar burden.
## Evs in bioengineering applications
Bioengineering is a diverse field that draws input from both the biological and engineering sciences to create new diagnostic and therapeutic modalities for patient care. Often, this relies heavily on advances in materials science and additive manufacturing technologies.
# Materials
In tissue regeneration, the application of an appropriate engineering strategy for developing three-dimensional (3D) constructs possessing the ability to deliver bioactive agents at the site of action with adequate temporo-spatial control is crucial for restoring the structural and functional integrity of damaged tissues/organs. A number of natural polymeric materials, such as silk, collagen, fibrin, gelatin, and hyaluronic acid, have shown promise for fabrication of such constructs due to their structural versatility, low immunogenicity, biocompatibility, and biodegradability in physiological conditions [bib_ref] Natural and Synthetic Bioinks for 3D Bioprinting, Khoeini [/bib_ref]. Such polymeric hydrogels serve to retain the encapsulated drugs, particles, or cells for targeted therapeutic applications. Amongst them, encapsulated particles can further control the degradation of drugs while reducing cellular integration. Important parameters that are required to be optimized include cellular adhesion, material characteristics, surface topology, and charge. Moreover, porous scaffolds can enhance cell infiltration and proliferation. Thus, the combination of scaffold technology with EVs can greatly enhance their clinical applicability [bib_ref] Technological Advances of 3D Scaffold-Based Stem Cell/Exosome Therapy in Tissues and Organs...., Gu [/bib_ref].
Hydrogels are a backbone in tissue engineering and represent a platform technology that has found wide clinical relevance, especially in reconstructive surgery and wound healing. However, hydrogels have also been used as an EV delivery system to mitigate the otherwise rapid clearance resulting from systemic administration. showed that such a strategy could be used to improve hepatic regeneration in chronic liver failure by sustaining MSC-EV release [bib_ref] Hydrogel-Mediated Sustained Systemic Delivery of Mesenchymal Stem Cell-Derived Extracellular Vesicles Improves Hepatic..., Mardpour [/bib_ref]. Unfortunately, hydrogels alone do not have the desired mechanical characteristics for all intended uses and often need to be combined with other materials. Silk also has longstanding clinical use, primarily as surgical suture. However, in order to improve scaffold physical characteristics, it has often been used as a composite material with other hydrogels. Now, it has been shown that these composite materials can further be loaded with EVs to impart specific therapeutic purposes. For example, silk composite hydrogels, when loaded with MSC-EVs, promoted wound healing and angiogenesis while reducing inflammation in a mouse wound-healing model. While flat scaffolds are quite common for wound healing, silk composite scaffolds can also be shaped for more specific applications. Vorp et al. showed that when tubed scaffolds were used as aortic interposition grafts in rats, the EV-loaded cohort led to improved endothelium formation and smooth muscle proliferation, culminating in increased patency after eight weeks in vivo [bib_ref] Extracellular vesicles enhance the remodeling of cell-free silk vascular scaffolds in rat..., Cunnane [/bib_ref]. The continued development of manufacturing technologies will only serve to increase available options for scaffold shaping and biologic localization.
## Bioprinting with evs
Bioprinting is an enabling biofabrication technique that allows for the precise deposition of bioactives as per user-defined architectures and compositions [bib_ref] Intra-Operative Bioprinting of Hard, Soft, and Hard/Soft Composite Tissues for Craniomaxillofacial Reconstruction, Moncal [/bib_ref]. Bioprinting methods have now been explored for developing scaffolds for delivery of EVs. For example, Diomede et al. and Pizzicannella et al. employed human gingival MSC-EVs complexes to develop 3D polylactide (PLA) scaffolds, which can release EVs at the desired levels [bib_ref] Three-dimensional printed PLA scaffold and human gingival stem cell-derived extracellular vesicles: A..., Diomede [/bib_ref] [bib_ref] 3D Printing PLA/Gingival Stem Cells/ EVs Upregulate miR-2861 and -210 during Osteoangiogenesis..., Pizzicannella [/bib_ref]. To achieve the higher release of EVs, polyethylene imine was coated on PLA scaffolds. Subsequently, the authors demonstrated the positive effect of improving PLA-EV adhesion by this mechanism on bone regeneration. Higher osteoinductivity was confirmed by biocompatibility and differentiation assays as well as using next-generation sequencing for transcriptomic analysis. Further, in vivo investigations of these scaffolds on cortical calvaria defects of rodents have indicated enhanced osteogenicity and bone tissue growth [bib_ref] Three-dimensional printed PLA scaffold and human gingival stem cell-derived extracellular vesicles: A..., Diomede [/bib_ref]. In a further advancement to the study, 3D-bioprinted PLA supplemented with MSC-EVs has shown increased OPN, RUNX2, COL1A1, and VEGF-A expressions in cells cultured on these scaffolds [bib_ref] 3D Printing PLA/Gingival Stem Cells/ EVs Upregulate miR-2861 and -210 during Osteoangiogenesis..., Pizzicannella [/bib_ref]. Additionally, the expression of miR-2861 and miR-210 was also found to be increased. PLA-EV adhesion by this mechanism on bone regeneration. Higher osteoinductivity was confirmed by biocompatibility and differentiation assays as well as using nextgeneration sequencing for transcriptomic analysis. Further, in vivo investigations of these scaffolds on cortical calvaria defects of rodents have indicated enhanced osteogenicity and bone tissue growth [bib_ref] Three-dimensional printed PLA scaffold and human gingival stem cell-derived extracellular vesicles: A..., Diomede [/bib_ref]. In a further advancement to the study, 3D-bioprinted PLA supplemented with MSC-EVs has shown increased OPN, RUNX2, COL1A1, and VEGF-A expressions in cells cultured on these scaffolds [bib_ref] 3D Printing PLA/Gingival Stem Cells/ EVs Upregulate miR-2861 and -210 during Osteoangiogenesis..., Pizzicannella [/bib_ref]. Additionally, the expression of miR-2861 and miR-210 was also found to be increased. [bib_ref] 3D Printing PLA/Gingival Stem Cells/ EVs Upregulate miR-2861 and -210 during Osteoangiogenesis..., Pizzicannella [/bib_ref]. (C) Three-dimensional coprinting of PCL with an alginate-based hydrogel encapsulating lyosecretome (i). (ii) SEM morphological and structural characterizations of 3D-printed PCL scaffolds loaded with lyosecretome. Presence (red arrows) or absence (blue arrows) of material deposition on PCL fibers is indicated. (iii) Scaffold geometry and dimensions of coprinting parallelepiped-and cylindrical-shaped scaffolds with a "soft heart" of lyosecretome-laden alginate (i.e., bioink) (adapted and reproduced from [bib_ref] 3D Bioprinted Scaffolds Containing Mesenchymal Stem/Stromal Lyosecretome: Next Generation Controlled Release Device..., Bari [/bib_ref]. (D) Osteoblastogenesis by extracellular vesicle-associated bone morphogenetic protein 2. (i) Representative bright-field images of ALP assay (C2C12 cells) and mineralization assay (MC3T3 cells); (ii) histological evaluation in vivo. After six weeks of grafting, samples were stained with von Kossa silver staining or Methylene blue and acid fuchsin images for the 3D-PLA/hGMSCs/Evs sample. Black arrows indicated blood vessels. Scale bar: 50 µm; (iii) 3D-MicroCT analysis showing 3D volume rendering and virtual transverse sectioning of 3D-PLA/hGMSCs/EVs (adapted and reproduced from [bib_ref] 3D Printing PLA/Gingival Stem Cells/ EVs Upregulate miR-2861 and -210 during Osteoangiogenesis..., Pizzicannella [/bib_ref]. (C) Three-dimensional coprinting of PCL with an alginate-based hydrogel encapsulating lyosecretome (i). (ii) SEM morphological and structural characterizations of 3D-printed PCL scaffolds loaded with lyosecretome. Presence (red arrows) or absence (blue arrows) of material deposition on PCL fibers is indicated. (iii) Scaffold geometry and dimensions of coprinting parallelepipedand cylindrical-shaped scaffolds with a "soft heart" of lyosecretome-laden alginate (i.e., bioink) (adapted and reproduced from [bib_ref] 3D Bioprinted Scaffolds Containing Mesenchymal Stem/Stromal Lyosecretome: Next Generation Controlled Release Device..., Bari [/bib_ref]. (D) Osteoblastogenesis by extracellular vesicle-associated bone morphogenetic protein 2. (i) Representative bright-field images of ALP assay (C2C12 cells) and mineralization assay (MC3T3 cells); (ii) ALP staining of C2C12s post-72 h seeding on bioprinted patterns with indicated 20 overprints (Ops); (iii) histological images showing H&E and Masson's trichrome staining of BMP2-EV-bioprinted implants (* indicates bone tissue); (iv) representative µCT 3D reconstructions of mouse leg scans containing either native EV or BMP2-EVs bioprinted implants. Arrow points to heterotopic ossification (adapted and reproduced from [bib_ref] Cell trafficking and regulation of osteoblastogenesis by extracellular vesicle associated bone morphogenetic..., Yerneni [/bib_ref].
Combination of 3D printing and freeze-dried formulation (lyosecretome) methods has been recently employed for fabrication of MSC secretome consisting of cytokines, proteins, and EVs. Two different strategies have been applied for fabrication of polycaprolactone (PCL)-based 3D-/printed scaffolds: in one approach, PCL was coprinted with MSC secretome incorporated in alginate hydrogels; and in another method, MSC secretome was adsorbed on prefabricated scaffolds. The absorption method allowed for a burst release of EVs from scaffolds, whereas slow release of EVs from encapsulating lyosecretome was observed in the coprinting method. The kinetic studies confirmed the diffusion release of EVs from secretome from both loading methods. The release of EVs was also found to depend on concentration of polymer, crosslinking modes, scaffold geometries, etc. Such prototypes have been tested for bone tissue regeneration [bib_ref] 3D Bioprinted Scaffolds Containing Mesenchymal Stem/Stromal Lyosecretome: Next Generation Controlled Release Device..., Bari [/bib_ref].
In another study, bone morphogenetic protein-2 (BMP-2)-loaded EVs were assessed through radiolabel-based assays [bib_ref] Cell trafficking and regulation of osteoblastogenesis by extracellular vesicle associated bone morphogenetic..., Yerneni [/bib_ref]. Initially, it was observed that in vitro radiolabeled study of liquid-phase BMP2-loaded EVs demonstrated their cellular transport and osteogenic activities. While cell trafficking with free BMP2 confirmed the association of BMP2 within EV cargo is a natural process, in the natural BMP2-EV complex it was found that BMP2 is attached unstably on the EV surface. Therefore, a solid-phase bioprinted eBMP2-EV niche was developed on collagen-rich matrix for both in vitro and in vivo assessment. Studies showed that EV incorporated with BMP-2 induced in vitro and in vivo osteogenic differentiation in picogram and nanogram level doses, respectively. The study suggested the role of BMP-2-based cell signaling for internalization of BMP-2-loaded EVs, which differ from the established growth factor-based mechanism. EVs isolated from human umbilical vein endothelial cells can be used as bioadditives for developing unique bioinks for developing 3D-bioprinting constructs.
Gelatin methacrylamide (GelMA) bioink has also been used for fabricating bioprinted EV-based scaffolds. In this process, gelatin was reacted with methacrylic anhydride and bioinks were separately loaded in EVs for different culture conditions. Bioinks alginate and calcium chloride crosslinkers were allowed to flow through a single-syringe pump at different flow rates. Multilayer scaffolds with aligned fibers were fabricated by keeping 50 µm distance in between. Several parameters of this bioprinting model were studied, such as the effect of EVs on the bioink viscosity, concentration of the crosslinker on bioprintability, scaffolds with different fiber orientations, fiber diameters on bioprinting speed, etc., for developing hierarchical vasculature constructs. Further, the distribution of EVs on the bioprinted matrix was tested at different magnifications. Higher-magnification images showed remarkable differences in surface properties where matrices with EVs showed rough surface topology while EV-free matrices consisted of smooth surface characteristics. The integrity of the EV membrane was investigated through isolating EVs from scaffolds through enzymatic digestion. Results showed that the compatibility of bioprinting with EV encapsulation as the EV membrane was not damaged during bioprinting and the size of the EV remained unchanged. Such EV-based bioinks show immense potential for supporting the transplantation and revascularization of ischemic regions, and can also meet the challenge of microvessel integration in tissue-engineered constructs. However, in-depth standardization is still required for clinical-grade EV production. In vivo implantation of such 3D-bioprinted scaffolds supported the in situ vascularization through the formation of microvessels. Results indicate the possibilities of therapeutic application of this technology in critical clinical needs, especially for ischemic tissue revascularization. Towards exosome-based therapeutics, Chen et al. used desktop-stereolithography technology to fabricate a 3D-printed cartilage extracellular matrix (ECM)/GelMA/exosome scaffold with radially oriented channels. The results showed the system restored cartilage mitochondrial dysfunction, enhanced chondrocyte migration, and macrophage M2 polarization along with osteochondral defect repair in a rabbit model [bib_ref] Desktop-stereolithography 3D printing of a radially oriented extracellular matrix/mesenchymal stem cell exosome..., Chen [/bib_ref]. To overcome the limit of EV production, a recent study reported a scalable method for high-quantity EV manufacturing. Herein, MSCs were densely encapsulated in micrometer-scale hydrogel fibers by coaxial bioprinting, and the results indicated that the developed system augmented particles by~1009-fold compared to conventional 2D culture, while preserving their proangiogenic properties [bib_ref] A scalable coaxial bioprinting technology for mesenchymal stem cell microfiber fabrication and..., Chen [/bib_ref]. Addressing scalable EV production, bioreactor systems have also been successfully used. Patel et al. reported a 3D-printed scaffold-perfusion bioreactor system with enhanced response of dynamic culture on EV production from endothelial cells (ECs) [bib_ref] Enhanced extracellular vesicle production and ethanol-mediated vascularization bioactivity via a 3D-printed scaffold-perfusion..., Patel [/bib_ref].
radially oriented channels. The results showed the system restored cartilage mitochondrial dysfunction, enhanced chondrocyte migration, and macrophage M2 polarization along with osteochondral defect repair in a rabbit model [bib_ref] Desktop-stereolithography 3D printing of a radially oriented extracellular matrix/mesenchymal stem cell exosome..., Chen [/bib_ref]. To overcome the limit of EV production, a recent study reported a scalable method for high-quantity EV manufacturing. Herein, MSCs were densely encapsulated in micrometer-scale hydrogel fibers by coaxial bioprinting, and the results indicated that the developed system augmented particles by ~1009-fold compared to conventional 2D culture, while preserving their proangiogenic properties [bib_ref] A scalable coaxial bioprinting technology for mesenchymal stem cell microfiber fabrication and..., Chen [/bib_ref]. Addressing scalable EV production, bioreactor systems have also been successfully used. Patel et al. reported a 3D-printed scaffold-perfusion bioreactor system with enhanced response of dynamic culture on EV production from endothelial cells (ECs) [bib_ref] Enhanced extracellular vesicle production and ethanol-mediated vascularization bioactivity via a 3D-printed scaffold-perfusion..., Patel [/bib_ref].. (B) Three-dimensional-printed extracellular matrix/mesenchymal stem cell sEV scaffold for osteochondral defect repair. (i) In vivo fluorescence imaging of sEVs in PBS and in the 3D-printed scaffold. Both groups of sEVs were labeled with Vybrant DiO (green) dye before implantation; (ii) macroscopic evaluation of the osteochondral defect regions at 12 weeks; (iii) MRI imaging of repaired knees at 12 weeks after surgery (yellow arrow indicates the repaired sites); (iv) µCT reconstruction imaging of repaired knees at 12 weeks after surgery (adapted and reproduced from. (B) Three-dimensionalprinted extracellular matrix/mesenchymal stem cell sEV scaffold for osteochondral defect repair. (i) In vivo fluorescence imaging of sEVs in PBS and in the 3D-printed scaffold. Both groups of sEVs were labeled with Vybrant DiO (green) dye before implantation; (ii) macroscopic evaluation of the osteochondral defect regions at 12 weeks; (iii) MRI imaging of repaired knees at 12 weeks after surgery (yellow arrow indicates the repaired sites); (iv and v) µCT reconstruction imaging of repaired knees at 12 weeks after surgery (adapted and reproduced from [bib_ref] Desktop-stereolithography 3D printing of a radially oriented extracellular matrix/mesenchymal stem cell exosome..., Chen [/bib_ref]. (C) Three-dimensional coaxial bioprinting for EV production. (i) Scalable production of highly enriched EVs using 3D cell fiber culture. MSCs and alginate solution were extruded through the inner and outer streams of the coaxial printer head, respectively, to form a dense 3D cell fiber that promotes cell survival and stemness maintenance; (ii) TEM of EVs isolated from 3D cell fiber culture showing a typical 'cup-shaped' morphology; (iii) evaluation of endothelial cell tube formation with EVs using a Matrigel matrix. Morphology of the endothelial network stimulated with 3D (2.8 × 109 particles mL −1 ) EVs (adapted and reproduced from [bib_ref] A scalable coaxial bioprinting technology for mesenchymal stem cell microfiber fabrication and..., Chen [/bib_ref].
The microfluidic-based "deterministic lateral displacement (DLD)" technique has been widely developed through 3D-printing or additive-manufacturing processes, in which arrays of micro/nanoposts are incorporated in microfluidic channels. Such systems facilitate the transport of suspended particles, and flow parameters can be modulated by altering the variables of geometric design. A "two-photon "direct laser writing (DLW)" method has been shown to be promising for manufacturing of DLD arrays, which can efficiently transport the submicron particles. The experimental result on testing of 860 nm fluorescent particles using DLD arrays fabricated by DLW-printing aided in understanding the hydrodynamic barrier of fluorescent particles along the angled arrayed of microposts. The lateral displacement of the study noted 15.3 ± 8.6 µm over a 500 µm channel length. Results indicate the promising application of 3D-printed DLD systems for the manipulation of submicron particles [bib_ref] Direct Laser Writing for Deterministic Lateral Displacement of Submicron Particles, Alsharhan [/bib_ref] that can be used for efficient bioprinting of EVs.
## Clinical translation
MSC-EVs are highly versatile and potent biological structures in terms of regenerative and anti-inflammatory capabilities for the creation of effective cell-free therapeutics. They exhibit targetable biodistribution, improved stability and handling, and decreased toxicity with no thrombogenic or tumorigenic hazards following intravascular delivery. These properties allow for their effective clinical applications. Huang et al. investigated the differences in protein from MSC-EVs and suggested that the isolation methods might introduce variations in the protein composition in sEVs, which directly affects their biological function [bib_ref] Identification and characterization of hADSC-derived exosome proteins from different isolation methods, Huang [/bib_ref]. Tieu et al. reviewed and evaluated in vivo preclinical studies using MSC-derived EVs as an intervention for any animal disease model [bib_ref] An Analysis of Mesenchymal Stem Cell-Derived Extracellular Vesicles for Preclinical Use, Tieu [/bib_ref]. The results indicated that despite the variations, 81% of studies indicated benefits for all evaluated outcomes, and 97% of studies reported improvements for at least one outcome. In 72% of investigations, EVs' involvement improved outcomes, organ function, and survival compared to unmodified MSC-EVs. Additionally, advantages of this cell-free therapy are seen in diseases affecting all organ systems with fundamentally different pathophysiology, implying that MSC-EVs may be used to treat a range of disorders with little risk of side effects. To eliminate any source of variability and animal-related contaminations, the use of xeno-and EV-free culture medium is advised for clinical applications.
To scale up EV production, 3D culture in bioreactors, such as multilayered cell culture flasks, hollow fiber bioreactors, stirred-tank bioreactors, and spheroidal aggregates of MSCs has been investigated. Particularly, the stirred-tank and hollow fiber bioreactors, which are closed, scalable, and offer a high surface-to-volume ratio for MSC growth, have shown promising results [bib_ref] Large-Scale Automated Hollow-Fiber Bioreactor Expansion of Umbilical Cord-Derived Human Mesenchymal Stromal Cells..., Vymetalova [/bib_ref] [bib_ref] Generation and testing of clinical-grade exosomes for pancreatic cancer, Mendt [/bib_ref]. The results have demonstrated that EV production in bioreactors is at least 40 times higher than in 2D culture methods [bib_ref] Efficient production and enhanced tumor delivery of engineered extracellular vesicles, Watson [/bib_ref]. For good therapeutic effect and for toxicity, biodistribution, or pharmacokinetic studies, highly purified and homogenous EV fractions are a requisite that majorly depend on the isolation method. An ideal isolation technique should be scalable, cost-effective, compatible with a highthroughput production process, and preferably be a closed system. Several methods have been investigated over time. The technique of ultracentrifugation (UC) has traditionally been regarded as the most common method for isolating MSC-EV from either cells or biological fluids. However, associated drawbacks with UC include impairing the integrity of EVs, being relatively time-consuming, necessitating processing samples in large volumes, and requiring specialized lab equipment. Nevertheless, sequential centrifugation steps have been used for the large-scale production of clinical-grade MSC-EVs [bib_ref] Generation and testing of clinical-grade exosomes for pancreatic cancer, Mendt [/bib_ref]. Further, the impurities and other protein contaminants in the isolated materials make it challenging to determine whether the biological activity is due to EVs or the co-isolated soluble mediators. Monguió-Tortajada et al. used size-exclusion chromatography (SEC) for isolating welldefined EVs similar to those of parental MSCs, highlighting the importance of well-purified preparations of MSC EVs to achieve immunosuppressive effects. Additionally, other techniques involving affinity chromatography solutions (heparin affinity binding [bib_ref] Heparin affinity purification of extracellular vesicles, Balaj [/bib_ref] or shiga or cholera toxin binding [bib_ref] MSC secretes at least 3 EV types each with a unique permutation..., Lai [/bib_ref] may be appropriate for purely analytical purposes; however, these may not entirely meet the current good manufacturing practice (GMP) regulations and the demand for large-scale purification.
For the industrial and clinical use of EVs, it is necessary to implement the regulatory issues for EV production and clinical uses. Lerner et al. in an ISEV (International Society for Extracellular Vesicles) position paper explored how to categorize EVs according to the anticipated active ingredients in biological medicines and pharmaceuticals along with discussing the legal problems in EV-based therapeutics [bib_ref] Applying extracellular vesicles based therapeutics in clinical trials-an ISEV position paper, Lener [/bib_ref]. Compliance with the regulatory frameworks is pivotal for the approval of EV-based therapies and their large-scale implementation. In compliance, several companies have already developed various EV-based therapies for a range of diseases and therapeutic targets, and MSC-EVs account for about 40% of these products [bib_ref] Manufacturing of Human Extracellular Vesicle-Based Therapeutics for Clinical Use, Gimona [/bib_ref]. Large-scale processes are being implemented using selectively closed systems for improved safety to assure reliable production procedures. These processes are mostly bioreactor technologies for EV production and ultrafiltration technologies for EV purification. For the manufacturing of MSC-EV or secretome products, several GMP-compliant procedures have been created [bib_ref] Reproducibility of GMP-compliant production of therapeutic stressed peripheral blood mononuclear cell-derived secretomes,..., Laggner [/bib_ref]. The key problem associated with the industrialization of EV-based medicines for regenerative medicine is to develop novel manufacturing techniques under GMP for EV scalable production and isolation. Different production methods will generate different products; thus, it is essential to develop a standardized operating procedure (SOPs) using standardized and reproducible assays to produce a defined and stable EV product. Overall, although most of the preclinical trials involving MSC-EV-based therapeutics are yet to be translated into clinical trials, the results from several studies are promising for future clinical applications.
# Conclusions
EV-based therapies possess immense potential in the field of regenerative medicine in the future because of their bioactive cargo and therapeutic potential for a variety of diseases. Existing studies provide solid evidence that EVs generated by MSCs of different origins demonstrate paracrine activity like that of their parental cells. However, several challenges must be considered and overcome before their clinical utility is achieved, including isolation and purification producing homogenous EVs; scaling up of EV production, optimizing, and setup of GMP-compliant procedures; standardized SOPs for better reproducibility; and optimizing the safety, immunogenicity, or optimal doses of EVs. Despite the challenges, based on the growing amount of evidence, it could be said that when refined and thoroughly characterized in terms of their bioactive cargo and mechanisms of action, MSC-EVs will be a suitable alternative to cell-based therapies as allogenic treatments for regenerative medicine applications.
[fig] Figure 1: Main types of extracellular vesicles and their size, source, function, their biomarkers, and carrying cargo. [/fig]
[fig] Cells 2022 ,: 11, x FOR PEER REVIEW 3 of 27 [/fig]
[fig] Figure 2, Figure 2: Schematic diagram of various sources of MSCs and their EVs that carrying cargo. Schematic diagram of various sources of MSCs and their EVs that carrying cargo. [/fig]
[fig] Figure 2: Schematic diagram of various sources of MSCs and their EVs that carrying cargo. [/fig]
[fig] Figure 3: Schematic diagram of EV sources and their use in therapeutic and biomarker applications. Reprinted from[18]. [/fig]
[fig] Figure 4: Schematic diagram of overall MSC isolation and their clinical applications in various organs and treatments (literature described in manuscript indicated in red arrows). [/fig]
[fig] Figure 5: MSC-EVs for potential therapeutics for lung diseases. Reprinted from[69]. [/fig]
[fig] Figure 6: Schematic diagram of cytokine storm produced by SARS-CoV2 infection. Reprinted from[77]. [/fig]
[fig] Figure 7: (A) Three-dimensional-printed poly(lactide) (PLA) scaffold and human gingival stem cell-derived (hGMSCs) extracellular vesicles (EVs). (i,ii) Topographic AFM image showing round morphology of EVs and polyethyleneimine (PEI)-engineered EVs (PEI-Evs); (iii) the images at low magnification showed construct integrated smoothly with the host; (iv) high-magnification images showing the new bone formation stained with acid fuchsin in both samples grafted at 6 weeks postsurgery (adapted and reproduced from [148]). (B) Three-dimensional-printed poly(lactide) (PLA)/Gingival Stem Cells/ Evs. (i) Representative µCT picture and SEM acquisition of 3D-printed PLA (3D-PLA); (ii) histological evaluation in vivo. After six weeks of grafting, samples were stained with von Kossa silver staining or Methylene blue and acid fuchsin images for the 3D-PLA/hGMSCs/Evs sample. Black arrows indicated blood vessels. Scale bar: 50 µm; (iii) 3D-MicroCT analysis showing 3D volume rendering and virtual transverse sectioning of 3D-PLA/hGMSCs/EVs (adapted and reproduced from [/fig]
[fig] Figure 8: (A) Three-dimensional-bioprinted endothelial-derived extracellular vesicles. (i) SEM images of 3D-bioprinted hydrogels (GelMA with EVs). Yellow arrows show vesicles; (ii) representative images of serum-free M200 medium in hypoxia (SM Hypoxia) EV sample explanted neovascular networks after EDTA decalcification treatment; (iii) immunofluorescence image of vWFpositive vessels showing branching and angiogenesis (adapted and reproduced from [/fig]
[fig] Author: Contributions: Major contribution towards: writing manuscript: C.M.; writing manuscript on bioprinting: P.D.; revised manuscript, Figures 7 and 8, bioprinting section: Y.P.S.; writing manuscript: A.L.; editing manuscript: S.H.; revised whole manuscript, rewriting, editing: I.A.E.; writing and correcting bioprinting section: I.T.O.; writing and editing manuscript: D.J.R.; Designed review article, wrote manuscript, preparing figures and supervised: S.V.K. All authors have read and agreed to the published version of the manuscript. [/fig]
[fig] Funding: Department of Surgery at Penn State College of Medicine, Hershey PA 17033 USA. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. [/fig]
[table] Table 1: List of registered clinical trials using MSC-EVs. (Source: Clinicaltrials.gov; accessed on 18 March 2022). [/table]
|
Abernethy malformation: a case report
Background: Abernethy malformation is a very rare congenital vascular malformation defined by diversion of portal blood away from liver. It is commonly associated with multiple congenital anomalies. We present a case of Abernethy malformation, without associated congenital anomalies from India.Case presentation: A 5-year-old female child presented with short history of jaundice. A provisional diagnosis of acute viral hepatitis was made in view of clinical presentation and local endemicity of viral hepatitis A. Persistence of jaundice on follow up after 4 weeks led to detailed investigations. Ultrasound and doppler study of abdomen revealed drainage of portal vein into inferior vena cava. CT angiography was performed which confirmed the diagnosis of Type 1 b Abernethy malformation without associated major anomalies. We discuss the common clinical presentations, associated anomalies, diagnostic workup and treatment options of this disorder.Conclusion:The treatment of the patients with congenital porto-systemic shunts depends on the site of the shunt, associated congenital anomalies and the extent of liver damage but the prognosis depends on the complications irrespective of anatomical type. However, the extent of associated abnormalities should not deter paediatricians to refer patients for treatment. Whenever possible closure of the shunt should be advised for cure or to prevent complications. Only symptomatic type I patients with absence of possibility to close the shunt may require liver transplant. Long-term follow-up is indicated for all patients.
# Background
John Abernethy first reported congenital absence of portal vein and congenital mesenterico-caval shunt in 1793 [bib_ref] Account of two instances of uncommon formation in the viscera of the..., Abernethy [/bib_ref]. Abernethy malformation is defined as congenital diversion of portal blood away from the liver by either end-to-side or side-to-side shunt [bib_ref] Congenital extrahepatic portocaval shunts-the Abernethy malformation, Howard [/bib_ref]. Morgan and Superina [bib_ref] Congenital absence of the portal vein: two cases and a proposed classification..., Morgan [/bib_ref] classified congenital extrahepatic portosystemic shunt (CEPSh) into two types (See [fig_ref] Table 1: Morgen and Superina[3] classification of congenital extra hepatic portosystemic shunt [/fig_ref]. Till the year 2008 only forty cases of CEPSh type I and twenty-two of CEPSh type II, had been described in literature [bib_ref] Congenital extrahepatic portosystemic shunts, Murray [/bib_ref] [bib_ref] Congenital veno-venous malformations of the liver: widely variable clinical presentations, Witters [/bib_ref]. This is despite the tremendous advances in the diagnostic medical imaging, thus underpinning the rarity of these cases. Only one paediatric case has been reported from India [bib_ref] Successful living donor liver transplant in a child with Abernethy malformation with..., Singhal [/bib_ref]. In this article we describe a patient diagnosed to have type I b Abernethy malformation after prolonged jaundice.
## Case presentation
We present a case of 5 year old girl who presented to paediatric outpatient clinic of R.D. Gardi Medical College and C.R. Gardi Hospital, Surasa, Ujjain, India with complains of decreased appetite since past four to five days, yellowish discoloration of sclera noticed since past two days and deep yellow coloured urine since past two days. Patient had no history of clay coloured stools, fever, vomiting, haematemesis, malena or history suggestive of hepatic encephalopathy. The perinatal history was uneventful with no history of umbilical sepsis. The child had no significant past medical history including that of jaundice. Immunization of the child was appropriate for age. At initial examination her anthropometric measurements showed a weight of 15 Kg and a height of 110 cm. The weight-for-age was between 3 rd to 15 th percentiles; height-for-age between 50 th to 85 th percentiles and weight-for-height below 3 rd percentile, according to the WHO Reference Growth Standards 2007 (www.who.int/childgrowth/standard/chart_catalogue ). The anthropometric findings were suggestive of acute on chronic malnutrition with stunting. The dietary evaluation revealed a diet inadequate in calories (by one third), proteins (by one fourth) and iron (by half) for her age and sex. The vitals at presentation were heart rate of 120 per minute, respiratory rate of 20 per minute and blood pressure of 98/ 60 mm of Hg in right in right upper limb in supine position. Four-limb blood pressure measurement was also normal. Icterus was noticed. Examination of ears, nose and throat showed no abnormality. No skin rash was present. No clubbing, cyanosis or oedema of arms or legs was noted. Cardiovascular examination was normal. The lungs were clear on percussion and auscultation. The abdomen was soft with no organomegaly or clinical ascites. A provisional diagnosis of acute viral hepatitis was made based on history, clinical examination and local endemicity of acute viral hepatitis caused by hepatitis A or hepatitis E. Only blood investigation advised was serum bilirubin, which showed total bilirubin 9.29 mg/dl with direct bilirubin of 7.38 mg/dl and indirect bilirubin of 2.41 mg/dl. Patient was treated symptomatically for viral hepatitis and asked to follow up if any danger signs of acute viral hepatitis appear or after 4 weeks. No serological confirmation of viral hepatitis was deemed necessary in view of acute viral hepatitis caused by hepatitis A virus being an endemic disease in the geographic region.
Patient followed four weeks later with persistence of yellowish discoloration of sclera but had no other symptoms. Her clinical examination was unremarkable. A liver profile, complete blood counts and abdominal ultrasonography (USG) were ordered. The complete blood counts were within normal limits and did not reveal any evidence of hypersplenism. The liver profile showed total bilirubin 3. The portal vein drained into inferior vena cava. A CT angiography was advised following the above abdominal ultrasound report. The contrast enhanced CT scan of upper abdomen with MIP to include of whole thickness of portal vein revealed that the superior mesenteric vein and the splenic vein join to form a confluence that drained directly in the inferior vena cava. The portal vein drained in the inferior vena cava at the level of head of pancreas caudal to the intrahepatic portion [fig_ref] Figure 1: Contrast enhanced CT scan of upper abdomen with MIP to include of... [/fig_ref]. A dilated inferior vena cava is seen in [fig_ref] Figure 2: CT scan abdomen [/fig_ref]. The CT scan also showed a mass lesion in right lobe of liver probably an adenoma or focal nodular hyperplasia [fig_ref] Figure 2: CT scan abdomen [/fig_ref]. In view of the above features and absence of portal vein radicals in liver parenchyma a diagnosis of Abernethy malformation type 1 b was made. To complete the work up a fasting blood sugar and serum ammonia was done which were normal. (Blood sugar 71 mg/L and serum ammonia 20 μgm/dL (normal values 10-80 μgm/dL). A skeletal survey did not show any skeletal abnormalities. An echocardiogram with colour doppler did not show any associated congenital cardiac defects.
## Consent
Written informed consent was obtained from the patient's father for publication of this case report and accompanying images.
# Discussion
Due to the endemicity of acute viral hepatitis caused by hepatitis A and hepatitis E most clinicians do not investigate a typical "viral hepatitis" case. In the present case the jaundice can be explained only by the viral hepatitis alone. But persistence of jaundice led us to investigate the present case. Thus, the possibility of missing a diagnosis in resource poor setting is high where both the diagnostic facilities and follow up of patients is often inadequate. However, one more basic indicator pointing towards a chronic disease were the anthropometric measurements. The importance of diagnostic imaging and follow-up of disease course is illustrated in this case. We now discuss various possible clinical presentations and associated congenital anomalies of the Abernethy malformations followed by diagnostic workup and treatment options. Congenital intra-hepatic portosystemic shunts can present in the early neonatal period with growth restriction [bib_ref] Congenital intrahepatic portosystemic shunt: prenatal diagnosis and possible influence on fetal growth, Delle Chiaie [/bib_ref] , galactosemia [bib_ref] Hypergalactosaemia in a newborn: self-limiting intrahepatic portosystemic venous shunt, Gitzelmann [/bib_ref] [bib_ref] Clinical features and outcome of eight infants with intrahepatic porto-venous shunts detected..., Ono [/bib_ref] [bib_ref] Differential diagnosis of neonatal mild hypergalactosaemia detected by mass screening: clinical significance..., Nishimura [/bib_ref] , neonatal cholestasis [bib_ref] Elevated plasma bile acids in hypergalactosaemic neonates: a diagnostic clue to portosystemic..., Sakura [/bib_ref] , and hepatic encephalopathy [bib_ref] Clinical features of congenital portosystemic shunt in children, Kim [/bib_ref]. Cases of CEPSh have a variable presentation. In countries having neonatal screening programmes, some cases can be diagnosed by neonatal screening tests due to presence of galactosemia [bib_ref] Hypergalactosaemia in a newborn: self-limiting intrahepatic portosystemic venous shunt, Gitzelmann [/bib_ref] [bib_ref] Clinical features and outcome of eight infants with intrahepatic porto-venous shunts detected..., Ono [/bib_ref] [bib_ref] Differential diagnosis of neonatal mild hypergalactosaemia detected by mass screening: clinical significance..., Nishimura [/bib_ref]. Many patients are diagnosed due to associated defects like heart disease, which are present in up-to 60% of the patients [bib_ref] Congenital extrahepatic portosystemic shunts, Murray [/bib_ref] [bib_ref] Clinical features of congenital portosystemic shunt in children, Kim [/bib_ref] [bib_ref] Portopulmonary hypertension and hepatopulmonary syndrome, Hoeper [/bib_ref] [bib_ref] Cardiac anomalies in the setting of the Abernethy malformation of the portal..., Ratnasamy [/bib_ref]. Symptoms of secondary complications like hypoglycemia, hyperammonemia, encephalopathy and cardiac failure can be transient and can resolve spontaneously [bib_ref] Clinical features of congenital portosystemic shunt in children, Kim [/bib_ref]. Subclinical course is more common and some patients might not have any symptoms throughout life [bib_ref] Clinical features of congenital portosystemic shunt in children, Kim [/bib_ref] [bib_ref] Portopulmonary hypertension and hepatopulmonary syndrome, Hoeper [/bib_ref] [bib_ref] The role of operative intervention in management of congenital extrahepatic portosystemic shunt, Sanada [/bib_ref]. Later, other lesions like nodular regenerative hyperplasia, partial nodular transformation, hepatoblastoma, hepatocellular carcinoma and adenoma can develop [bib_ref] Congenital extrahepatic portosystemic shunts, Murray [/bib_ref].
The two possible explanations of development of these neoplasm are: a) diversion of hepatotrophic substances like insulin and glucagon away from liver leading to altered development, function and regenerative capacity of the liver and b) increased arterial hepatic flow [bib_ref] Congenital extrahepatic portosystemic shunts, Murray [/bib_ref]. Our patient does not have any evidence of hepatic neoplasm (normal alfa feto protein) at present after more than two year follow up, however in view of the documented increased risk of hepatocellular carcinoma and adenoma we have advised a regular radiological and serological surveillance at six monthly interval.
Despite presence of porto-systemic shunt, patients with CEPSh usually do not present with encephalopathy.
Hyperammonemia may be present without encephalopathy especially at younger age. Clinical encephalopathy is more common at older age. The possible explanations are a) an increased sensitivity of the aging brain to toxic effects of ammonia and other toxic metabolites b) extent of shunt determined by the portal/systemic shunt ratio. A shunt ratio of more than 60% may predict the age of onset of hepatic encephalopathy [bib_ref] Successful living donor liver transplant in a child with Abernethy malformation with..., Singhal [/bib_ref] [bib_ref] Intrahepatic portal-hepatic venous anastomosis: a portal-systemic shunt with neurological repercussions, Crespin [/bib_ref]. In our case, the ammonia concentration was within normal range and patient has had no clinical evidence of encephalopathy till date.
Varying degree of dyspnoea on exertion could be a presentation of portopulmonary hypertension or hepatopulmonary syndrome. Hepatopulmonary syndrome presents with hypoxemia and or orthopnea with clubbing; due to abnormal intrapulmonary vascular dilatations [bib_ref] Portopulmonary hypertension and hepatopulmonary syndrome, Hoeper [/bib_ref] [bib_ref] Complications of congenital portosystemic shunts in children: therapeutic options and outcomes, Franchi-Abella [/bib_ref]. In our patient there is no evidence of hepatopulmonary syndrome, till date clinically as well as on echocardiography.
Congenital hepatic shunt can also present with hypoglycaemia. This might be due to combined effect of defective uptake of glucose and hyperinsulinemia due to reduced hepatic degradation of normal quantity of secreted insulin [bib_ref] Glucose intolerance and poststimulatory hypoglycemia secondary to a probably congenital intrahepatic portacaval..., Duprey [/bib_ref]. A work up of common associated anomalies was done but revealed no anomaly. These include cardiac defects (60%), biliary atresia (20%), polysplenia (20%), situs inversus (10%) and malrotation (10%) [bib_ref] Cardiac anomalies in the setting of the Abernethy malformation of the portal..., Ratnasamy [/bib_ref] [bib_ref] Congenital absence of the portal vein and role of liver transplantation in..., Shinkai [/bib_ref].
A diagnosis of Abernethy malformation is made by non-invasive cross-sectional imaging techniques such as ultrasound, CT or MRI, which show the shunt and any intrahepatic portal vein branches [bib_ref] Complications of congenital portosystemic shunts in children: therapeutic options and outcomes, Franchi-Abella [/bib_ref].
The planning of treatment is dependent on the type of shunt as classified by Morgan and Superina [bib_ref] Congenital absence of the portal vein: two cases and a proposed classification..., Morgan [/bib_ref] and needs to be tailored to the individual patient in accordance with preoperative evaluation [bib_ref] The role of operative intervention in management of congenital extrahepatic portosystemic shunt, Sanada [/bib_ref]. Usually in patients with CEPSh type I, occlusion of shunt is not performed, as it is the only drainage route for the mesenteric venous blood. But recently published experience by several authors [bib_ref] The role of operative intervention in management of congenital extrahepatic portosystemic shunt, Sanada [/bib_ref] [bib_ref] Complications of congenital portosystemic shunts in children: therapeutic options and outcomes, Franchi-Abella [/bib_ref] point that many patients with CEPSh type I malformations might have small portal vein radicals which cannot be seen on ultrasonography but could be visualized on shunt angiography. The balloon occlusion test of the fistula can also be done [bib_ref] The role of operative intervention in management of congenital extrahepatic portosystemic shunt, Sanada [/bib_ref] [bib_ref] Complications of congenital portosystemic shunts in children: therapeutic options and outcomes, Franchi-Abella [/bib_ref]. This test helps to decide on a single stage or a two-staged shunt closure procedure. A two-step procedure allows the portal branches to enlarge slowly and can avoid acute severe portal hypertension [bib_ref] The role of operative intervention in management of congenital extrahepatic portosystemic shunt, Sanada [/bib_ref] [bib_ref] Complications of congenital portosystemic shunts in children: therapeutic options and outcomes, Franchi-Abella [/bib_ref]. Extremely hypoplastic or undetectable portal veins will require banding of the fistula before closure [bib_ref] Complications of congenital portosystemic shunts in children: therapeutic options and outcomes, Franchi-Abella [/bib_ref]. Shunt closure results in restoration of intrahepatic portal flow in most patients [bib_ref] Complications of congenital portosystemic shunts in children: therapeutic options and outcomes, Franchi-Abella [/bib_ref]. Clinical improvement in form of regression of benign liver masses, and regression or stabilization of pulmonary, cardiac, neurological, and renal complications is seen in patients post shunt [bib_ref] Complications of congenital portosystemic shunts in children: therapeutic options and outcomes, Franchi-Abella [/bib_ref]. CEPSh typeI patients also need clinical, biochemical and imaging follow-up [bib_ref] Complications of congenital portosystemic shunts in children: therapeutic options and outcomes, Franchi-Abella [/bib_ref] ; as is done in our patient. For patients with CEPSh type I developing complications like encephalopathy or neoplasms, till recently liver transplant was the only treatment option [bib_ref] Successful living donor liver transplant in a child with Abernethy malformation with..., Singhal [/bib_ref] [bib_ref] The role of operative intervention in management of congenital extrahepatic portosystemic shunt, Sanada [/bib_ref] [bib_ref] Congenital absence of the portal vein and role of liver transplantation in..., Shinkai [/bib_ref] [bib_ref] Value of carbon dioxide wedged venography and transvenous liver biopsy in the..., Collard [/bib_ref] , but transplant should be reserved for exceptionally complex anatomy where closure of the shunt is not possible.
Patients with CEPSh type II with hepatic encephalopathy can benefit by early shunt occlusion surgery [bib_ref] The role of operative intervention in management of congenital extrahepatic portosystemic shunt, Sanada [/bib_ref] [bib_ref] Complications of congenital portosystemic shunts in children: therapeutic options and outcomes, Franchi-Abella [/bib_ref].
Reconstruction of the portal vein should be done early to avoid mesenteric venous congestion [bib_ref] Value of carbon dioxide wedged venography and transvenous liver biopsy in the..., Collard [/bib_ref]. Shunt surgery when possible is the treatment of choice for CEPSh type I and type II. Our patient is asymptomatic at present but since the patient has nodular lesion in the liver she should be considered "symptomatic". In view of this our patient has been counselled for need for shunt surgery.
# Conclusions
Persistence of jaundice should be investigated at all ages. Ultrasound is useful tool for screening of congenital anomalies associated with liver. In resource poor settings the diagnosis of portosystemic shunts can be missed or significantly delayed due to lack of resources need for the diagnosis, including availability of trained ultra-sonologists. The prognosis of the patients with congenital portosystemic shunts depends on the site of the shunt as determined by Morgan and Superina classification, the associated congenital anomalies and the extent of liver disease. Many patients will benefit from shunt surgery. The extent of associated abnormalities should not deter paediatricians to refer patients for treatment. A long-term follow-up is indicated for all asymptomatic patients of Abernethy malformation.
[fig] 4: mg/dl with direct bilirubin of 3.23 mg/dl and indirect bilirubin of 0.22 mg/dl; aspartate transaminase (AST) 98.2 IU/L (normal values 15-46 IU/L); alanine transaminase (ALT) 109.3 IU/L (normal values 13-69 IU/L); alkaline phosphatase 269.3(normal values 38-126 IU/L); prothombin time (PT) 14 with control of 16; INR 1.2. Abdominal ultrasound showed course ecotexture of the liver with anomalous portal vein. The size of portal vein was 8 mm. [/fig]
[fig] Figure 1: Contrast enhanced CT scan of upper abdomen with MIP to include of whole thickness of portal vein shows aberrant opening of portal vein into inferior vena cava (marked by white solid arrow in the figure). Note dilated inferior vena cava cranial to confluence in the image. [/fig]
[fig] Figure 2: CT scan abdomen: Mass lesion in right lobe of liver (marked by white solid arrow). [/fig]
[table] Table 1: Morgen and Superina[3] classification of congenital extra hepatic portosystemic shunt [/table]
|
Orthotics and taping in the management of vertebral fractures in people with osteoporosis: a systematic review
Objective: To establish the current evidence base for the use of orthotics and taping for people with osteoporotic vertebral fracture (OVF). Design: Systematic review of quantitative and qualitative studies.Eligibility criteria for selecting studies: All study designs were considered if they reported in English and evaluated the impact of using an external support, such as a spinal brace, orthosis or postural tape, with adults with OVF. All outcomes were considered.Results: Nine studies were included comprising two parallel-group randomised controlled trials, four randomised cross-over trials, two before-after (single arm) studies and a parallel group observational study. No qualitative studies were identified. A wide range of outcomes assessing impairments, activities and participation were assessed but the findings were mixed. The quality of studies was limited.Conclusions: The current evidence for using orthotic devices or taping for people with OVF is inconsistent and of limited quality and therefore careful consideration should be taken by clinicians before prescribing them in practice.
# Introduction
Symptomatic osteoporotic vertebral fractures (OVF) affect 1 in 6 women and 1 in 12 men during their lifetime.In Europe the annual incidence of OVF is 10.7 and 5.7 per thousand in women and men, respectively.In the USA, 700 000 people are reported to have an OVF each year,however, these figures may underestimate the actual size of the problem as most OVF go unreported and, therefore, undiagnosed.With the prevalence of osteoporosis predicted to increase by 2021, a rise in associated fractures is also likely.OVF have personal, societal and economic costs.Patients experience severe pain in the acute phase but also up to 2 years postfracture. Increased disability and depression, and a reduced quality of life are also reported.Up to a fifth will suffer a subsequent vertebral fracture within a yearand there is an increased risk of mortality.The use of primary care services has been found to be 14 times that of the general population,with 8% of people with OVF requiring hospitalisation.In the USA, the management OVF has an annual estimated cost of $13.8 billion.Guidelines for the non-medical and nonsurgical management of OVF are conflicting. The American Academy of Orthopaedic Surgeons (AAOS)report inconclusive evidence in relation to rehabilitative interventions whereas the Scottish Intercollegiate Guidelines Network (SIGN)recommend electrical field therapy and supervised exercise programmes. There are a variety of external support devices available including rigid thoracolumbar spinal orthoses or hyperextension braces that are often used in the management of OVF, however, they are not recommended in current guidelines.Bracing is reported to assist healing when worn for up to 3 months but when worn for an extended period can result in muscle atrophy, postural muscle weaknessand skin problems.This said, these devices are used by the people with OVF and clinicians continue to prescribe them, although practice varies. In view of recently published studies, this review aimed to identify and update the current evidence base for the use of orthotics and taping for people with OVF.
# Methods
We used the preferred reporting items for systematic reviews and meta-analyses (PRISMA) as a guideline for reporting this study. A predefined protocol was registered with PROSPERO (CRD 42015020893).
## Identification and selection of studies
We included primary studies that used quantitative or qualitative methods evaluating the impact of using an external support, such as a spinal brace, orthosis or postural tape, with adults with OVF. We were interested in outcomes relating to the WHO International Classification of Functioning, Disability and Health (ICF) domains of body structure and function, activities and participation. We were also interested in the experiences and perceptions of users of the external support.
We excluded studies that involved traumatic vertebral fractures, non-vertebral fractures and those involving children, reviews and opinion papers, studies published only as an abstract and those where full text was not available in English. We also excluded controlled studies where the intervention also included surgical, pharmacological and rehabilitation interventions, except where these were provided to intervention and comparator participants. For non-controlled studies, only those where the evaluation related to the orthotic device/tape were included.
The search strategy (see online supplementary file) was applied from 1980 to April 2015 to Medline, Medline-In Process, EMBASE, AMED, CINAHL, PEDro, TRIP, EThOS, ProQuest Dissertations and Theses and Cochrane (CDSR, DARE, CMR, HTA, EED). In addition we searched clinical trials databases, including Cochrane Central, UK Clinical Research Network portfolio, Controlled Clinical Trials register and the Australian and New Zealand Clinical Trials register. We also undertook forward and backward citation checking.
The search and screening process was managed using Endnote. Two reviewers independently screened titles and abstracts. Full-text papers were then screened against the eligibility criteria independently by two reviewers. Discrepancies were discussed and agreed.
## Assessment of study characteristics
Data were extracted using a prepiloted form by one reviewer and checked by a second. The data included study characteristics (study design, selection criteria, setting and sample size), funding sources, ethical approval, population (age, gender, time since OVF), intervention and comparator characteristics (nature of intervention, duration of wear, concomitant interventions and adherence), outcomes, time points of follow-up and findings.
Study quality was assessed using the Cochrane Risk of Bias tooland was extracted by one reviewer and checked by a second. The data included random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting and other sources of bias. Each item was rated as low risk, high risk or unclear risk of bias and reported for individual studies and across the studies.
## Data synthesis
A narrative approach was used to synthesise the study findings due to methodological (study design, outcomes) and clinical ( participant and intervention characteristics) heterogeneity. This approach enabled exploration of relationships within the data.Risk of bias was summarised for individual studies and across studies.
# Results
Our initial search resulted in 667 citations (figure 1) of which 84 were assessed against the selection criteria; the remainder did not meet the inclusion criteria. Nine studies were included in this review
## Characteristics of studies
The studies included 468 participants of which 404 (86%) were women. Where reported, participants had a mean (SD) age of 72.1 (7.9) years and had sustained an OVF between 3 days and 2 years previously. Six studies did not report time since fracture. The number of participants in each study varied from 13 to 108 (Mean=52). Studies took place in Europe,Asia 23-26 and Australia.The majority of participants were ambulatory and community-dwelling. One study was undertaken with inpatients.Most studies took place in an outpatient setting with the setting being unclear in one study.Eight studies evaluated six different orthotic devices including rigid supports (TLSO, 3 point orthosis, plaster corset); semi-rigid supports (Spinomed and Spinomed Active); and flexible supports (soft brace) and one examined the effect of postural taping. Four studies evaluated two different types of orthotics. Controls included no brace, an alternative orthotic device (soft brace) or placebo (hypoallergenic tape). The devices were worn for between 15 min and 24 h a day. Adherence to wearing the orthotics was assessed by self-report in three studies although adherence was not defined. Additional data regarding individual study characteristics are reported in table 1.
Two parallel-group randomised controlled trials, four randomised cross-over trials,22 24 27 two before-after (single arm) studiesand a parallel group observational studywere included. Three studies had three study arms. No qualitative studies were identified. Of the four studies using crossover design, two did not report the findings from the first treatment period. The two studies undertaken by Pfeifer et al reported that almost all participants that received the intervention during the first period refused to stop using the device for the crossover. Therefore, to maintain clarity when comparing the different intervention groups in these two studies, only the findings from the first period were considered in this review.
## Study quality
The results of the risk of bias assessment of study quality are presented inand table 2. The reporting of studies was poor with no studies using ConSORT guidelines.Only one study reported the randomisation allocation processes. None had previously registered their studies or published a protocol. Four studies did not report ethical approvals and three did not report funding sources. One study was termed a 'pilot' although the aims were stated to determine efficacy.Four studies undertook sample size calculations. For the four studies using a cross-over design, additional quality assessments were made in relation to appropriate design, randomised treatment order, carry-over effect and unbiased dataand these are recorded inunder 'Other'.
## Effect of the intervention
The effectiveness of orthotics and taping are reported in terms of impairments, activities and participation.
Follow-up varied from immediate, short term (<1 month),medium term (1-3 months)and longer term (>3 months).Impairments The majority of outcomes related to impairments such as pain, postural stability, back strength, angle of kyphosis 21-23 and fracture union.Pain: Two studies targeted those with acute fractures using TLSO or soft brace and found no benefit, whereas those studies with those with longer term fractures reported mixed findings. Pfeifer et al reported the Spinomed device reduced pain measured using Milner's rating scale (1=low pain to 4=very severe pain). When compared with no brace they found an Absolute Difference=−1.6 (95% CIs −2.1 to −1.1), with similar findings from an earlier study.Valentin et al,however, did not find improvements in pain on a 0-10 scale (with higher scores indicting worse pain) at 3 months when using Spinomed (median difference (range)=−1 (−4.7 to 1.7); p=0.06). Li et al 25 reported that Spinomed was no better than a soft brace after 3 weeks using a 0-10 pain scale (mean pain (SD) 4.0 (2.0) vs 4.5 (2.1)).
Postural stability: A range of different methods were used to assess postural stability such as a force plate,computerised dynamic posturography 24 and a sway metre. Each assessment method reported multiple complex components of postural stability but there were no consistent findings within or across the studies. One cross-over studyexamined the immediate effect of the brace or no brace on balance performance but did not appear to consider potential carry-over effects nor report the findings from the first time period so it remains unclear as to the actual effect on postural stability of the orthosis.
Back strength: Three studies assessed the use of Spinomed on isometric back strength. Pfeifer et al reported a mean increase (SD) of 180 (152) Newtons when wearing the device and with an absolute difference of 182 (95% CI 125.1 to 238.9) Newtons compared with wearing no brace.
Angle of kyphosis: Five studies reported angle of kyphosis using a range of techniques, including an inclinometer, 27 radiographs 23 25 and three-dimensional photomorphometry. Postural taping was found to have immediate improvements in thoracic kyphosis when compared with placebo or no tape (Mean angle in degrees (SD) 55.3 (13.5); 57.2 (13.8); and, 58.2 (12.3) respectively; p=0.024).Li et al 25 only assessed this outcome on 10/51 participants and it was unclear as to how they were selected. Two studies were unclear as to whether the findings represented an improvement or deterioration in kyphosis. Fracture union: Murata et al 23 reported at 2 months 54.7% of participants had fracture 'settling'; with 88.7% settled at 6 months, however, the study did not have a control group and was, therefore, not possible to establish what benefit wearing the TLSO gave over 'normal' healing. The reporting of adverse events was poor with only one study explicitly identifying them as an outcome of interest.Talic et al 20 found that plaster corsets resulted in an increased risk with four patients (16%) developing pressure sores, with no adverse events related to using a three-point orthosis.
## Activities
The impact of orthotics on activities was evaluated using the Oswestry Disability Index (ODI),the Functional Independence measure, Elderly Mobility Scale and Modified Functional Ambulation Category,limitations in everyday life and walking ability component of the Japanese Orthopaedic Association Back Pain Evaluation Questionnaire ( JOABPEQ).Pfeifer et al reported reduced disability associated with using Spinomed when compared with no brace (absolute difference −2.3 (95% CI −1.7 to −2.9), 21 although it is unclear whether this is superior to a soft brace.Kim et al 26 found no between group differences in ODI for those with an acute OVF when comparing TLSO (mean difference −1.88; 95% CI −7.02 to 9.38) or soft bracing (mean difference 2.41; −7.86 to 9.27) with no brace.
## Participation
Participation, in relation to quality of life and well-being, was evaluated in four studies. The SF36 domains were not improved by using a TLSO or soft brace when compared with no brace (Mean Physical component score (PCS) 32, 35 and 30 respectively; p=0.716)or Spinomed (median within-group difference in Physical component score=6.5, range −9.2 to 13.3; p=0.07)whereas the Hobi well-being scale did improve after wearing Spinomed (absolute differ-ence=12.7; 95% CI 9.7 to 15.7).There was no indication that either outcome was moderated by population characteristics.
# Discussion
This review aimed to establish the effectiveness of orthotics or taping in the management of OVF. We found that the nine included studies were of limited quality and reported varied populations, interventions and outcomes. For example, no studies reported whether outcome assessors were blinded to allocation, thus the potential for detection bias is unknown. A previous review of the non-surgical management of OVF included three RCTs evaluating bracingreported medium term pain relief and reduced disability; however, studies were considered to be of low/very low quality. We found little consideration of any potential adverse events associated with bracing or taping with this population. It has been suggested that adherence to wearing orthotics is poor 31 but despite several included studies stating adherence data was collected it was never defined and rarely reported. In addition, there is a complete absence of qualitative research involving this population and their experiences of wearing these devices which would be an essential component of any future development and evaluation. This said, in two cross-over studies, the fact that the participants refused to stop wearing the orthotics at the point of crossover would suggest positive experiences. We found no evidence to counter the recommendations of the American Academy of Orthopaedic Surgeons 13 that indicated there was inconclusive evidence to support the using of bracing in the management of OVF, and that the quality of studies in this clinical area remains limited. Our findings also concur with a recent systematic review of taping that reported it was no better than placebo or no taping in terms of pain or disability for people with back pain.In terms of strengths, our study used contemporary methods for undertaking a systematic review and registered the protocol prospectively on PROSPERO. We searched an extensive range of databases, including grey literature. One limitation was the exclusion of non-English full-text papers, however, we are confident from our extensive search (that was not restricted to English) and screening process that any potential papers would have failed to meet other selection criteria and therefore would not have been included in the review. It could also be suggested that including non-randomised and non-controlled studies is a limitation. However, the purpose of our review was to establish the current evidence base, and not establish effectiveness, and therefore we considered it important to include all study designs. In addition, the use of the risk of bias tool enabled a judgement on the overall quality of the included studies, and we found that even the included randomised controlled trials were not without limitations.
# Conclusions
There is inconclusive evidence that TLSO or soft braces reduce pain or disability. Although the use of the Spinomed appears to have some benefit in terms of increasing back strength and reducing disability, it does not necessarily offer better outcomes when compared with other devices, such as soft braces. The quality of studies examining the effectiveness of orthotics or taping for the management of OVF is generally limited and therefore we would err on the side of caution when considering their use in clinical practice. Overall, there is limited evidence for the use of orthotics or taping either in the acute or long-term management of those with OVF. Further studies using high-quality methods and reporting are required to determine whether taping or orthotics are effective and cost-effectiveness.
Contributions VAG, AJH and AB developed the study protocol. AH and AB performed the searches. VG and AH undertook all screening and all authors were involved in data extraction. VG undertook the synthesis. All authors contributed to the final manuscript.
Funding This review was supported by a research grant from AGILE (Chartered Physiotherapists working with Older People) and also by the National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care South West Peninsula. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement No additional data are available.
Open Access This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http:// creativecommons.org/licenses/by/4.0/ |
Risk of adverse pregnancy and perinatal outcomes after high technology infertility treatment: a comprehensive systematic review
In the literature, there is growing evidence that subfertile patients who conceived after infertility treatments have an increased risk of pregnancy and perinatal complications and this is particularly true for patients who conceived through use of high technology infertility treatments. Moreover, high technology infertility treatments include many concomitant clinical and biological risk factors. This review aims to summarize in a systematic fashion the current evidence regarding the relative effect of the different procedures for high technology infertility treatments on the risk of adverse pregnancy and perinatal outcome. A literature search up to August 2016 was performed in IBSS, SocINDEX, Institute for Scientific Information, PubMed, Web of Science and Google Scholar and an evidence-based hierarchy was used to determine which articles to include and analyze. Data on prepregnancy maternal factors, low technology interventions, specific procedures for male factor, ovarian tissue/ovary and uterus transplantation, and chromosomal abnormalities and malformations of the offspring were excluded. The available evidences were analyzed assessing the level and the quality of evidence according to the Oxford Centre for Evidence-Based Medicine guidelines and the Grading of Recommendations Assessment, Development, and Evaluation system, respectively. Current review highlights that every single procedure of high technology infertility treatments can play a crucial role in increasing the risk of pregnancy and perinatal complications. Due to the suboptimal level and quality of the current evidence, further well-designed studies are needed.
# Background
Throughout the years, it has always been clear to scientists that the primary endpoint in reproductive medicine was the healthy baby and that all other endpoints would be considered only a surrogate [bib_ref] Surrogate end-points or primary outcomes in clinical trials in women with polycystic..., Legro [/bib_ref] [bib_ref] Live birth is the correct outcome for clinical trials evaluating therapy for..., Barnhart [/bib_ref]. The published infertility clinical trials have rarely reported clear data about the possible harm of the medical, surgical and biological procedures for enhancing fertility [bib_ref] Assisted reproductive technologies and perinatal morbidity: interrogating the association, Barnhart [/bib_ref] [bib_ref] Improving the Reporting of Clinical Trials of Infertility Treatments (IMPRINT): modifying the..., Bhattacharya [/bib_ref] , as well as giving very little relevance to long-term effects of those procedures on maternal and offspring health. Only 4.8 % and 5.7 % of randomized controlled trials (RCTs) on infertility treatments reported on perinatal and maternal outcome [bib_ref] How are neonatal and maternal outcomes reported in randomised controlled trials (RCTs)..., Braakhekke [/bib_ref]. This is probably because obstetrical and infant care are delivered by other providers and patients are lost to follow-up.
Notwithstanding these limitations, more and more data available in the literature seem to demonstrate that pregnancy following infertility treatments are at higher risk of adverse pregnancy and perinatal outcomes when compared with those after natural conception (NC) independently from scientific approach [bib_ref] Assisted reproductive technology and perinatal outcomes: conventional versus discordant-sibling design, Dhalwani [/bib_ref] , and this is particularly true for pregnancies achieved thanks to high technology infertility treatments [bib_ref] How are neonatal and maternal outcomes reported in randomised controlled trials (RCTs)..., Braakhekke [/bib_ref]. The majority of this risk is a "pure" iatrogenic risk due to the high rates of multiple births, i.e. 41.1 % of the United States (US) infants conceived with assisted reproductive technologies (ART) were born as multiple-birth infants compared with only 3.5 % of infants among the general birth population [bib_ref] Assisted Reproductive Technology Surveillance-United States, Sunderam [/bib_ref]. The rate of multiple deliveries following ART represents about 18.7 % of total multiple-birth infants [bib_ref] Assisted Reproductive Technology Surveillance-United States, Sunderam [/bib_ref]. However, because also singleton infants conceived with ART are at higher risk of preterm birth (PTB) and low birthweight (LBW) [bib_ref] Assisted Reproductive Technology Surveillance-United States, Sunderam [/bib_ref] , other determinants cannot be excluded. Patients' characteristics including the infertility state [bib_ref] Infertility and the risk of adverse pregnancy outcomes: a systematic review and..., Messerlian [/bib_ref] [bib_ref] Why do singletons conceived after assisted reproduction technology have adverse perinatal outcome?..., Pinborg [/bib_ref] and many preconception risk factors for subfertility [bib_ref] Risk of poor neonatal outcome at term after medically assisted reproduction: a..., Ensing [/bib_ref] [bib_ref] Pregnancy complications in spontaneous and assisted conceptions of women with infertility and..., Palomba [/bib_ref] [bib_ref] Effects of in vitro fertilization and maternal characteristics on perinatal outcomes: a..., Seggers [/bib_ref] can largely increase the absolute and relative risk of obstetric morbidity.
Although systematic reviews with and/or without meta-analysis have been published on specific topics, at the moment no comprehensive review is available in the literature, discussing the impact on maternal and perinatal outcome of each element and/or clinical/biological choice which comprise the high technology infertility treatments. Based on these considerations, the aim of the current document was to comprehensively review in a systematic fashion the hitherto published evidences regarding the effects of high technology infertility treatments on the obstetric risk of patients with female and couple infertility. The effects on the risk of chromosomal abnormalities and malformations of the offspring was not a study aim.
# Materials and methods
The methodology used for the current systematic review consisted of searching all available articles for each specific issue to explore the relationship between high technology infertility treatments and pregnancy and perinatal complications. High technology infertility treatments were considered as all interventions for fertility enhancement including manipulation of female gametes. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement [bib_ref] Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement, Moher [/bib_ref] was followed but after comprehensive search all the authors agreed to prepare the document in a narrative fashion in consideration of the multifaceted aspects to discuss.
Multiple strategies were used to search and identify relevant demographic, epidemiological, clinical and experimental studies. Sociological online libraries (IBSS, SocINDEX), Institute for Scientific Information, PubMed, Web of Science and Google Scholar were consulted. Only articles written in English were considered. The search was conducted independently by two authors (S.P. and S.S.). The literature available up to August 2016 was captured, including all available studies which reported data about the relationship between each fertility technique and the related obstetric and perinatal complications, matching every intervention with every potential obstetric disorder and perinatal health impact, as shown in [fig_ref] Table 1: Key words used to study the relationship between specific procedure of ARTs... [/fig_ref]. Additional journal articles were identified from the bibliography of the studies included. At study design, all the authors agreed to exclude from final analysis data of pregnancy and perinatal complications related to: 1. prepregnancy maternal factors; 2. low technology interventions (intervention aimed to enhance fertility without any manipulation of female gametes, ie. lifestyle intervention programs, insulin sensitizing drugs, ovulation inductors, macro-and microsupplements, intrauterine insemination, etc.); 3. specific high technology infertility treatments for male factor [including intracytoplasmic sperm injection (ICSI), specific techniques for sperm selection and/or retrieval, etc.]; 4. ovarian tissue/ovary and uterus transplantation; 5. chromosomal abnormalities and malformations of the offspring. The choice to consider a study relevant in order to be included in the current review was arbitrarily taken by each author, even if an evidence-based hierarchy was used. Exploratory studies on mechanisms of action and/or pathogenesis of any complication were included only in absence of available clinical data. Data on the efficacy of each procedure were reported only as necessary for the study aim. Any disagreement or uncertainty was resolved by discussion to reach a consensus.
The available evidence about the relationship between high technology infertility treatments and adverse obstetric and perinatal outcomes was analyzed assessing the level of evidence according to the Oxford Centre for Evidence-Based Medicine (OCEM)-Levels of Evidence 2011 guidelines [bib_ref] Evidence-based mechanistic reasoning, Howick [/bib_ref] and the quality of evidence according to the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system [bib_ref] GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables, Guyatt [/bib_ref]. [fig_ref] Figure 1: PRISMA 2009 [13] flow diagram [/fig_ref] is the flow diagram of the systematic review including the numbers of studies screened, assessed for eligibility, and included in the review, with reasons for exclusions at each stage [bib_ref] Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement, Moher [/bib_ref]. [fig_ref] Table 2: Levels and quality of the evidences available about the relationship between specific... [/fig_ref] summarizes the main risks for obstetric and perinatal adverse outcomes in women who receive ART and specific procedure for high technology infertility treatments according to the level [bib_ref] Evidence-based mechanistic reasoning, Howick [/bib_ref] and the quality [bib_ref] GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables, Guyatt [/bib_ref] of available evidence.
# Results
## Overall art-related complications singleton pregnancies
Many data from systematic reviews with and without meta-analyses have demonstrated that ART singletons are at increased risk of pregnancy and perinatal complications. In 2004, an initial systematic review of controlled studies found a relative risk of 3.27 (95 % CI 2.03 to 5.28) for early-preterm birth (EPTB) (< 32 weeks) in singleton pregnancies from assisted conceptions [bib_ref] Perinatal outcome of singletons and twins after assisted conception: a systematic review..., Helmerhorst [/bib_ref]. [bib_ref] Obstetric and perinatal outcomes in singleton pregnancies resulting from IVF/ICSI: a systematic..., Pandey [/bib_ref]. Also in this last meta-analysis, only in some of the included studies were the pregnancies resulting from ovulation induction excluded from non-IVF/ICSI conceptions [bib_ref] Obstetric and perinatal outcomes in singleton pregnancies resulting from IVF/ICSI: a systematic..., Pandey [/bib_ref].
The most recently published meta-analysis about the risk of pregnancy-related complications and adverse perinatal outcomes in singleton pregnancies obtained with ART involves 50 cohort studies for a total of 161,370 ART singleton compared with 2,280,241 NC singleton pregnancies [bib_ref] Assisted reproductive technology and the risk of pregnancy-related complications and adverse pregnancy..., Qin [/bib_ref]. This meta-analysis revealed that the singleton ART pregnancies experienced a significantly increased risk of placenta previa [bib_ref] Assisted reproductive technology and the risk of pregnancy-related complications and adverse pregnancy..., Qin [/bib_ref]. Of note, the risk of EPTB (RR 2.12, 95 % CI 1.73 to 2.59) and of VLBW (RR 2.12, 95 % CI 1.84 to 2.43) was two-fold higher in ART conceptions than in NC [bib_ref] Assisted reproductive technology and the risk of pregnancy-related complications and adverse pregnancy..., Qin [/bib_ref]. The results of this study are consistent with those of the previous reviews but it presents important strengths, such as the large sample size, 64 % of the included studies were considered of high methodological quality and the association between ART and obstetric risk persisted and remained statistically significant in sensitivity analysis based on various exclusion criteria [bib_ref] Assisted reproductive technology and the risk of pregnancy-related complications and adverse pregnancy..., Qin [/bib_ref]. Nevertheless, as relevant biases, patients who achieved a pregnancy with ovulation induction and/or intrauterine insemination (IUI) were included in the NC category, resulting in an underestimation of the association between ART and adverse outcomes. When data were restricted to studies that did not include these patients in the NC group, the risk of GDM, placental abruption, PTB, EPTB, LBW, VLBW, and perinatal mortality resulted in a further increase. Unfortunately, more than half of the included studies did not specify whether they were included, thereby restricting this subgroup analysis [bib_ref] Assisted reproductive technology and the risk of pregnancy-related complications and adverse pregnancy..., Qin [/bib_ref]. Finally, recent retrospective studies confirmed in ART pregnancies a risk of PIH/PE about 20 % higher (aOR 1.17, 95 % CI 1.10 to 1.24) [bib_ref] Increased incidence of gestational hypertension and preeclampsia after assisted reproductive technology treatment, Wang [/bib_ref] and demonstrated a risk of peripartum hysterectomy about six-fold increased (OR 5.98 95 % CI 2.18 to in comparison with non-ART pregnancies [bib_ref] Risk of peripartum hysterectomy in births after assisted reproductive technology, Cromi [/bib_ref].
## Multiple pregnancies
Initial systematic reviews with meta-analyses [bib_ref] Perinatal outcome of singletons and twins after assisted conception: a systematic review..., Helmerhorst [/bib_ref] [bib_ref] Perinatal outcomes of in vitro fertilization twins: a systematic review and meta-analyses, Mcdonald [/bib_ref] demonstrated that ART twins had a higher risk of PTB compared with NC twins but the perinatal mortality, contrary to singletons, was unchanged [bib_ref] Perinatal outcomes of in vitro fertilization twins: a systematic review and meta-analyses, Mcdonald [/bib_ref] or reduced [bib_ref] Perinatal outcome of singletons and twins after assisted conception: a systematic review..., Helmerhorst [/bib_ref].
A recent meta-analysis including 39 cohort studies explored the risk of adverse pregnancy outcomes in ART conceptions compared with NC in a sample of 146,008 multiple births [bib_ref] Pregnancy-related complications and adverse pregnancy outcomes in multiple pregnancies resulting from assisted..., Qin [/bib_ref]. ART were associated with a higher risk of premature rupture of membranes [bib_ref] Pregnancy-related complications and adverse pregnancy outcomes in multiple pregnancies resulting from assisted..., Qin [/bib_ref]. When data were restricted to matched and/or high quality studies, the risk of PIH, GDM, PTB, and LBW/VLBW increased further. Moreover, the results of main outcomes had a significant heterogeneity among studies and were significantly influenced by inclusion/exclusion of the pregnancies achieved after ovulation induction with and without IUI in NCs.
In all meta-analytic data [bib_ref] Perinatal outcome of singletons and twins after assisted conception: a systematic review..., Helmerhorst [/bib_ref] [bib_ref] Perinatal outcomes of in vitro fertilization twins: a systematic review and meta-analyses, Mcdonald [/bib_ref] [bib_ref] Pregnancy-related complications and adverse pregnancy outcomes in multiple pregnancies resulting from assisted..., Qin [/bib_ref] [bib_ref] Assisted reproductive technology and risk of adverse obstetric outcomes in dichorionic twin..., Qin [/bib_ref] perinatal mortality was not increased. A recent cohort study [bib_ref] Perinatal outcomes associated with assisted reproductive technology: the Massachusetts Outcomes Study of..., Declercq [/bib_ref] , however, reported an adjusted risk of perinatal death for twins 45 and 85 % lower among ART births than fertile or subfertile births, respectively. This finding can be explained because of the decreased proportion of monochorionic twins in ART pregnancies (2-7 % vs. 22-30 %) [bib_ref] Do assisted-reproduction twin pregnancies require additional antenatal care?, Jauniaux [/bib_ref] showing that in twins studies it could be crucial to control/adjust data for zygosity. Furthermore, ART monochorionic twins are also at increased risk of adverse perinatal outcomes compared with spontaneous monochorionic twins. In fact, in monochorionicity ART appears to increase the already high risk of PTB [bib_ref] Outcome of monochorionic twins conceived by assisted reproduction, Simões [/bib_ref]. One way to overcome the zygosity as confounder could be to perform comparisons restricted to differentsex twin pairs in order to exclude all the monochorionic twins. To this regard, a recent meta-analysis [bib_ref] Assisted reproductive technology and risk of adverse obstetric outcomes in dichorionic twin..., Qin [/bib_ref] including 15 cohort studies and involving 6,420 dichorionic twins after ART and 13,650 dichorionic NC twins confirmed a risk of placenta previa [bib_ref] Dizygotic twin pregnancies after medically assisted reproduction and after natural conception: maternal..., Bensdorp [/bib_ref]. This did not translate into a difference in PE or delivery mode, and there was no evidence of a difference in PTB rates because women delivered prematurely in 51 % of cases both after IVF/ ICSI and after NC [bib_ref] Dizygotic twin pregnancies after medically assisted reproduction and after natural conception: maternal..., Bensdorp [/bib_ref].
## Number of embryos transferred
An excellent meta-analysis of individual data from RCTs demonstrated that the elective single embryo transfer (SET) is less effective (of about 50 %) than double embryo transfer (DET) in women aged more than 33 years but not in younger, and that the odds of a multiple live birth in women randomized to eSET were significantly smaller than for women who received DET [bib_ref] Clinical effectiveness of elective single versus double embryo transfer: metaanalysis of individual..., Mclernon [/bib_ref]. After eSET, the adjusted risk to deliver a LBW baby (aOR 0.36, 95 % CI 0.15 to 0.87) and to have a PTB (aOR 0.33, 95 % CI 0.20 to 0.55) was significantly reduced, whereas the adjusted odds of a term singleton birth after eSET were almost five times higher than those after DET (aOR 4.93, 95 % CI 2.98 to 8.18) [bib_ref] Clinical effectiveness of elective single versus double embryo transfer: metaanalysis of individual..., Mclernon [/bib_ref]. Data obtained from the US National Assisted Reproductive Technology Surveillance System, a retrospective population-based study analyzed the 82,508 fresh autologous ART cycles, confirmed that a higher chance of a good perinatal outcome, defined as a live birth at or after 37 weeks of gestation of a normal birth weight (2,500 g or greater) singleton, is associated with SET in patients with a favorable prognosis who were aged younger than 35 years [bib_ref] Number of embryos transferred after in vitro fertilization and good perinatal outcome, Kissin [/bib_ref]. The same conclusions were obtained after the analysis of the 2013 data, i.e. increasing use of elective SET, when clinically appropriate (typically age < 35 years), might reduce multiple births and related adverse consequences of ART [bib_ref] Assisted Reproductive Technology Surveillance-United States, Sunderam [/bib_ref]. Moreover, the average rate of SET was 21.4 % also among good prognosis patients and varied considerably among States (from 4 to 77.5 %) [bib_ref] Assisted Reproductive Technology Surveillance-United States, Sunderam [/bib_ref].
Irrespective of the risk related to the multiple pregnancies, the number of embryos transferred can be an independent and potential factor influencing the obstetric and perinatal outcomes. Available data were very heterogeneous showing better perinatal outcomes [bib_ref] Birthweight of singletons after assisted reproduction is higher after single-than after double-embryo..., Sutter [/bib_ref] [bib_ref] Perinatal outcomes after assisted reproductive technology treatment in Australia and New Zealand:..., Wang [/bib_ref] or no difference [bib_ref] Obstetric and neonatal outcome after single embryo transfer, Poikkeus [/bib_ref] [bib_ref] Obstetric outcome after in vitro fertilization with single or double embryo transfer, Sazonova [/bib_ref] between SET singletons and DET singletons. A meta-analysis demonstrated no difference in the incidence of PTB (aOR 0.83, 95 % CI 0.64 to 1.06) between singletons born after SET and singletons born after DET [bib_ref] Why do singletons conceived after assisted reproduction technology have adverse perinatal outcome?..., Pinborg [/bib_ref]. However, an effect of vanishing twin syndrome (VTS) [bib_ref] Ultrasonic assessment of the high rate of human multiple pregnancy in the..., Levi [/bib_ref] , as consequence of DET or multiple embryo transfer, cannot be excluded. This is particularly true in consideration of the higher risk of pregnancy complications observed in VTS pregnancies from IVF/ ICSI cycles in comparison with VTS pregnancies from NC [bib_ref] Prevalences and pregnancy outcome of vanishing twin pregnancies achieved by in vitro..., Márton [/bib_ref]. A large cohort study including 7,757 deliveries following IVF/ICSI procedures found more adverse outcomes among VTS survivors occurring after 8 weeks of gestation, even after adjustment for maternal age and parity [bib_ref] Consequences of vanishing twins in IVF/ICSI pregnancies, Pinborg [/bib_ref]. However, that study did not differentiate between "true" (first trimester) cases of the VTS and those occurring at the second trimester since embryonic losses from 9 weeks until midgestation were grouped together, and a case-mix of di-and monochorionic twins is probable. On the other hand, similar maternal and perinatal complications, such as gestational duration and birth weights, were observed between VTS survivors and singletons in a retrospective series including IVF/ICSI patients when only well selected patients with initial heart beat on both embryos and with clear chorionicity were included [bib_ref] Effect of the mode of assisted reproductive technology conception on obstetric outcomes..., Sala [/bib_ref].
In singletons with a 'vanishing co-twin' , the risk of PTB (aOR 1.73, 95 % CI 1.54 to 1.94) and of other adverse perinatal outcomes (aOR ranging from 1.73 to 2.88) was significantly higher than in singletons from a single gestation [bib_ref] The effect of early fetal losses on singleton assisted-conception pregnancy outcomes, Luke [/bib_ref]. The obstetric and perinatal risks in VTS seem to increase in accordance with the number of reduced fetuses [bib_ref] The effect of early fetal losses on singleton assisted-conception pregnancy outcomes, Luke [/bib_ref]. Other case-control and cohort studies on singleton deliveries after IVF/ICSI showed that VTS was associated with higher risk of vaginal bleeding and preterm premature rupture of membranes, and that VTS survivors had a birth weight significantly lower, and a rate of LBW and SGA about double when compared to singleton controls [bib_ref] Vanishing twins: a predictor of small-for-gestational age in IVF singletons, Pinborg [/bib_ref] [bib_ref] Birth weight is lower for survivors of the vanishing twin syndrome: a..., Shebl [/bib_ref] [bib_ref] Predicting adverse obstetric outcome after early pregnancy events and complications: a review, Van Oppenraaij [/bib_ref]. Of note, no difference was observed in terms of duration of gestation suggesting a direct effect of the VTS on the survived embryos/fetus [bib_ref] Birth weight is lower for survivors of the vanishing twin syndrome: a..., Shebl [/bib_ref].
A well designed retrospective study on a very large cohort of 252,994 singleton deliveries evaluated the pregnancy and perinatal outcomes, adjusted for confounders such as fertility treatment and maternal age, in VTS in comparison with those recorded in singletons and dichorionic twins [bib_ref] Vanishing twin syndrome: is it associated with adverse perinatal outcome?, Evron [/bib_ref]. VTS was found to be an independent risk factor for several adverse perinatal outcomes [bib_ref] Vanishing twin syndrome: is it associated with adverse perinatal outcome?, Evron [/bib_ref]. Specifically, the risk of GDM (aOR 1.4, 95 % CI 1.01 to 2.0), intrauterine growth restriction (IUGR) (aOR 2.7, 95 % CI 1.7 to 4.3), VLBW (aOR 6.9, 95 % CI 4.7 to 10.2), low Apgar scores (aOR 1.9, 95 % CI 1.1 to 3.3), and perinatal mortality (aOR 2.4, 95 % CI 1.2 to 4.6) was significantly increased in VTS when compared with singletons. In addition, VTS significantly influenced also the risk of malpresentation, placental abruption, and cesarean section [bib_ref] Vanishing twin syndrome: is it associated with adverse perinatal outcome?, Evron [/bib_ref]. However, that study [bib_ref] Vanishing twin syndrome: is it associated with adverse perinatal outcome?, Evron [/bib_ref] included mainly NC, and the effect of VTS in IVF/ICSI vs. NC pregnancies was not tested. To this regard, a recent retrospective matched (for maternal age, previous gravidity, parity, and prepregnancy BMI) analysis demonstrated that the incidence of the VTS in infertile IVF/ ICSI patients is lower than in NC (12.6 % vs. 18.2 %, respectively; estimated for twin pregnancies) but the perinatal outcomes resulted worse. In fact, the risk of GDM (aOR 3.0, 95 % CI 1.6 to 5.6 vs. aOR 0.46, 95 % CI 0.2 to 1.1) and of PE (aOR 1.6, 95 % CI 0.7 to 6.1 vs. aOR 1.00, 95 % CI 0.8 to 1.8) in VTS pregnancies was very high in IVF/ICSI patients but lower or unchanged in NC women [bib_ref] Prevalences and pregnancy outcome of vanishing twin pregnancies achieved by in vitro..., Márton [/bib_ref]. Also the incidence of IUGR (aOR 3.0, 95 % CI 1.8 to 5.2 vs. aOR 9.2, 95 % CI 5 to 22) and of LBW (aOR 4.0, 95 % CI 1.8 to 7.1 vs. aOR 2.1, 95 % CI 1.6 to 4.0) in VTS pregnancies was higher in IVF/ ICSI patients than in NC women [bib_ref] Prevalences and pregnancy outcome of vanishing twin pregnancies achieved by in vitro..., Márton [/bib_ref]. Very interesting data from multiple logistic regression analyses that identified in VTS pregnancies obtained after IVF/ICSI cycles a risk of IUGR 28-fold higher (aOR 28.2, 95 % CI 2.2 to 14.5) [bib_ref] Prevalences and pregnancy outcome of vanishing twin pregnancies achieved by in vitro..., Márton [/bib_ref].
## Controlled ovarian hyperstimulation (coh)
In comparison with children born after natural-cycle IVF, IVF children conceived after stimulation with clomiphene citrate (CC) and with CC plus gonadotropins had a higher risk of LBW [bib_ref] Implications of assisted reproductive technologies on term singleton birth weight: an analysis..., Nakashima [/bib_ref]. On the other hand, IVF children conceived after ovarian stimulation with gonadotropins alone did not have a higher risk of LBW compared to natural-cycle [bib_ref] Implications of assisted reproductive technologies on term singleton birth weight: an analysis..., Nakashima [/bib_ref] confirming IUI data [bib_ref] Perinatal outcomes in 6,338 singletons born after intrauterine insemination in Denmark, Malchau [/bib_ref]. In addition, no effect of the duration of stimulation, amount of drug administered, and number of oocytes retrieved was detected on SGA or LBW in a cohort study on 32,000 singletons born from gonadotropin IVF/ICSI cycles [bib_ref] Ovarian stimulation for IVF has no quantitative association with birth weight: a..., Griesinger [/bib_ref]. A retrospective cohort study on 392 women less than 40 years of age demonstrated the safety, in terms of obstetric and neonatal complications, of the use of gonadotropin-releasing hormone (GnRH) agonist to trigger ovulation in antagonist cycles. Specifically, there were no significant differences in the rate of maternal (27.6 vs. 20.8 %) or neonatal complications .0 %) between the GnRH agonist and hCG [bib_ref] Maternal and neonatal outcomes after gonadotropin-releasing hormone agonist trigger for final oocyte..., Budinetz [/bib_ref]. In addition, luteal GnRH antagonist administration in women with severe early ovarian hyperstimulation syndrome (OHSS) is associated not only with not different live birth rates but also with similar duration of gestation (36.9 ± 0.9 vs. 36.9 ± 2.4 weeks) and neonatal weight (2392 ± 427 vs. 2646 ± 655 g) when compared to patients who do not receive the treatment [bib_ref] Pregnancy and neonatal outcomes following luteal GnRH antagonist administration in patients with..., Lainas [/bib_ref]. However, in patients undergoing COH using the GnRH antagonist protocol, GnRH agonist trigger were found to be independent risk factors for ectopic pregnancy [bib_ref] Ectopic pregnancy risk factors for ART patients undergoing the GnRH antagonist protocol:..., Weiss [/bib_ref].
The ovarian response to COH has been also related to pregnancy complications. A recent observational United Kingdom (UK) study [bib_ref] Increased risk of preterm birth and low birthweight with very high number..., Sunkara [/bib_ref] higher in women with a high number (>20) of oocytes retrieved when compared with women with a normal response (10-15 oocytes) [bib_ref] Increased risk of preterm birth and low birthweight with very high number..., Sunkara [/bib_ref]. Moreover, the observation that higher rates of LBW infants in singleton gestations after IVF cycles was only present in patients who had a fresh ET when compared with those with frozen-thawed cycles [bib_ref] Ovarian stimulation and low birth weight in newborns conceived through in vitro..., Kalra [/bib_ref] has suggested that the gonadotropin stimulated multifollicular development and the production of supraphysiologic levels of sex steroid hormones immediately before embryo implantation might represent an independent contributing factor to that increased risk (and not only an effect of fresh or frozen embryos as below detailed). In fact, in IVF cycles the elevated peak serum estradiol (E 2 ) levels on the day of the human chorionic gonadotropin (hCG) trigger were closely related to the risks of SGA and PE in singleton pregnancies [bib_ref] High serum oestradiol concentrations in IVF cycles increase the risk of pregnancy..., Farhi [/bib_ref] [bib_ref] Peak serum estradiol level during controlled ovarian hyperstimulation is associated with increased..., Imudia [/bib_ref]. Thus, the supraphysiologic hormonal milieu at the time of implantation and placentation during a fresh IVF cycle may modulate trophoblast invasion and lead to abnormal placentation, whereas hormonal levels in a frozen ET cycle are much more akin to the endocrine environment of NC. This hypothesis has been tested by a small cohort study on a population at high risk for OHSS for high E 2 levels; the elective cryopreservation of all embryos with subsequent thawed ET reduced significantly the risk of PE (21.9 % vs. 0 %, respectively; OR not calculable), and to deliver SGA infants (OR 0.09, 95 % CI 0.01 to 0.65) as compared with the patients who had fresh ET [bib_ref] Elective cryopreservation of all embryos with subsequent cryothaw embryo transfer in patients..., Imudia [/bib_ref]. However, the published case-control studies on the obstetric and perinatal outcomes in patients with OHSS showed divergent results [bib_ref] Outcome of pregnancies complicated by severe ovarian hyperstimulation syndrome (OHSS): a follow-up..., Wiser [/bib_ref] [bib_ref] Obstetric outcome of women with in vitro fertilization pregnancies hospitalized for ovarian..., Courbiere [/bib_ref] [bib_ref] Is severe OHSS associated with adverse pregnancy outcomes? Evidence from a case-control..., Haas [/bib_ref]. In line with previous data, no increased risk of the adverse outcomes among women with a poor ovarian response (≤ 3 oocytes) compared with women with a normal response was observed [bib_ref] Increased risk of preterm birth and low birthweight with very high number..., Sunkara [/bib_ref] , even if several important confounders, such as smoking, body mass index (BMI), polycystic ovarian syndrome (PCOS), and gonadotropin doses were not available for an optimal data adjustment. On the other hand, a retrospective cohort study of the Society for Assisted Reproductive Technology (SART) on 14,086 patients demonstrated an inverse relationship between maximal basal follicle stimulating hormone (FSH) levels and the risk for PTB and LBW in singleton IVF gestations with the lowest risk of PTB (aRR 0.87, 95 % CI 0.76 to 1.01) and of LBW (aRR 0.89, 95 % CI 0.73 to 1.04) in patients with the lowest ovarian reserve, defined as highest serum basal and/or CC-induced FSH levels [bib_ref] Examining the relationship between ovarian reserve, as measured by basal FSH levels,..., Calhoun [/bib_ref].
Finally, the hitherto published literature regarding the relation between OHSS and pregnancy complications is limited and controversial [bib_ref] Is severe OHSS associated with adverse pregnancy outcomes? Evidence from a case-control..., Haas [/bib_ref]. This disparity in the observed outcomes probably results from the inclusion of patients with OHSS, not exclusively limited to those with significant increase in vascular permeability, as reflected by third space fluid accumulation necessitating drainage. A recent study [bib_ref] Pregnancy outcome in severe OHSS patients following ascitic/plerural fluid drainage, Haas [/bib_ref] has demonstrated that severe OHSS, complicated by third space fluid sequestration necessitating drainage, is not associated with adverse late pregnancy outcome (i.e. IUGR and PIH/PE), except for PTB. Therefore, following resolution of the OHSS, pregnancies should be regarded as any pregnancy resulting from IVF treatment, with special attention to identify and treat PTB.
## Blastocyst vs. cleavage state et
In a first systematic review with meta-analysis [bib_ref] Why do singletons conceived after assisted reproduction technology have adverse perinatal outcome?..., Pinborg [/bib_ref] , the risk of PTB was not significantly different in singletons conceived after day 5 vs. day 2 embryo culture (aOR 1.14, 95 % CI 0.80 to 1.64). A further meta-analysis [bib_ref] Obstetric and perinatal outcomes in singleton pregnancies resulting from the transfer of..., Maheshwari [/bib_ref] aimed to assess specifically the risk of pregnancy complications in singleton pregnancies resulting from ET at the blastocyst stage (days 4-5 or 5-6) compared with those resulting from ET at the cleavage stage (days 2-3 or day 3 or days 2-4) demonstrated, after data synthesis of 7 retrospective cohort studies, that pregnancies occurring as a result of ET at the blastocyst stage were associated with a higher risk of PTB (RR 1.27, 95 % CI 1.22 to 1.31) and VPTB (RR 1.22, 95 % CI 1.10 to 1.35) delivery and a lower risk of SGA [bib_ref] Obstetric and perinatal outcomes in singleton pregnancies resulting from the transfer of..., Maheshwari [/bib_ref]. However, these data were not adjusted for confounders and included very different study protocols. Another more recent meta-analytic study [bib_ref] Neonatal outcomes among singleton births after blastocyst versus cleavage stage embryo transfer:..., Dar [/bib_ref] aimed to evaluate the perinatal outcomes among singleton births following blastocyst stage (day 5 or 6) ET compared with cleavage stage (day 2-4) ET and included 6 observational studies with low to moderate risk of bias. It demonstrated, after adjusting data for main confounders, that only the risk of PTB (aOR 1.32, 95 % CI 1.19 to 1.46) was significantly higher among births after blastocyst transfer than after cleavage stage transfer [bib_ref] Neonatal outcomes among singleton births after blastocyst versus cleavage stage embryo transfer:..., Dar [/bib_ref]. No further difference in perinatal/neonatal outcome was observed [bib_ref] Neonatal outcomes among singleton births after blastocyst versus cleavage stage embryo transfer:..., Dar [/bib_ref].
At the moment, the most complete data on maternal and child health adjusted for confounders were from a large retrospective cohort study [bib_ref] Extended embryo culture and an increased risk of preterm delivery, Kalra [/bib_ref] [bib_ref] Extended embryo culture and an increased risk of preterm delivery, Kalra [/bib_ref]. These results were confirmed also in twins (aOR 1.81, 95 % CI 1.63 to 2.00; and aOR 1.75, 95 % CI 1.50 to 2.04 for PTB and VPTB, respectively) [bib_ref] Extended embryo culture and an increased risk of preterm delivery, Kalra [/bib_ref]. However, pregnancies that did not result in a live birth were not analyzed and the number of embryos transferred was not considered. To this regard, a retrospective cohort study analyzed the transfer of 693 fresh single cleavage embryos comparing to 850 fresh single blastocysts [bib_ref] Extended embryo culture is not associated with increased adverse obstetric or perinatal..., Oron [/bib_ref]. A higher live birth rate (33.5 % vs. 13.8 %) was detected after single blastocysts transfer and after adjusting data for several confounders, while no effect of the extended culture was observed on obstetric or perinatal outcomes [bib_ref] Extended embryo culture is not associated with increased adverse obstetric or perinatal..., Oron [/bib_ref]. These data have been confirmed also more recently in a retrospective matched case-control study by the same researchers [bib_ref] Obstetric and perinatal outcome from single cleavage transfer and single blastocyst transfer:..., Oron [/bib_ref] and by a large retrospective population-based study including a total of 50,788 infants (43,952 singleton and 3,418 twin deliveries) [bib_ref] Risk of preterm birth after blastocyst embryo transfer: a large population study..., Chambers [/bib_ref]. Conversely, there were lower odds of PTB among twins (aOR 0.80, 95 % CI 0.70 to 0.93) conceived after blastocyst transfer than after cleavage stage ET [bib_ref] Risk of preterm birth after blastocyst embryo transfer: a large population study..., Chambers [/bib_ref].
An important issue observed in the studies comparing the transfer of embryos at blastocyst vs. cleavage state was the discrepancy between the increased incidence in PTB and the similar incidence of LBW neonates in singletons [bib_ref] Extended embryo culture and an increased risk of preterm delivery, Kalra [/bib_ref]. Regarding this, an effect of the extended culture on the neonatal weight has been demonstrated. In fact, the incidence of SGA in neonates born after extended embryo culture compared with cleavage stage ET was significantly lower (OR 0.80, 95 % CI 0.74 to 0.87) [bib_ref] Extended embryo culture and an increased risk of preterm delivery, Kalra [/bib_ref]. The effect of extended embryo culture on fetal growth has been specifically assessed in a cohort study on 1,079 singleton neonates born after fresh ET [bib_ref] Does long in vitro culture promote large for gestational age babies?, Mäkinen [/bib_ref]. The risk of large-for-gestational age (LGA) (but not for SGA) adjusted for mother's age, BMI, parity, type of treatment (IVF or ICSI), main cause of infertility and embryo culture period was increased after extended embryo culture (aOR 2.22, 95 % CI 1.14 to 4.34), and the length of the embryo culture was a strong and significant factor determining the gestation and genderadjusted birth weight of the IVF babies [bib_ref] Does long in vitro culture promote large for gestational age babies?, Mäkinen [/bib_ref].
On the other hand, a recent retrospective registry study [bib_ref] Neonatal and maternal outcome after blastocyst transfer: a population-based registry study, Ernstad [/bib_ref] [bib_ref] Neonatal and maternal outcome after blastocyst transfer: a population-based registry study, Ernstad [/bib_ref].
The findings of obstetric and perinatal complications in pregnancies achieved after ET at the blastocyst stage are unexpected because good quality embryos from women with the best prognosis tend to be selected for extended culture. The specific mechanism underlying the association between extended embryo culture and PTB is unknown. It could be the result of the transfer at blastocyst stage inducing a defective implantation for asynchrony between the endometrium and embryo [bib_ref] Assisted reproductive technology, epigenetics, and long-term health: a developmental time bomb still..., Grace [/bib_ref] , trigger genetic and epigenetic modifications in the preimplantation embryos due to the culture medium of the extended culture [bib_ref] Assisted reproductive technology, epigenetics, and long-term health: a developmental time bomb still..., Grace [/bib_ref] [bib_ref] Environmental and epigenetic effects upon preimplantation embryo metabolism and development, Chason [/bib_ref] and/or in decidual/trophodermal cells due to abnormal embryo-endometrium interaction [bib_ref] Analysis of differential messenger RNA expression between bovine blastocysts produced in different..., Rizos [/bib_ref]. In fact, extended culture up to blastocyst stage increases the incidence of monozygotic twins (OR 3.04, 95 % CI 1.54 to 6.01) and the male-to-female ratio (OR 1.29, 95 % CI 1.10 to 1.51 [bib_ref] Impact of blastocyst transfer on offspring sex ratio and the monozygotic twinning..., Chang [/bib_ref] , and both are associated with PTB [bib_ref] Multiple pregnancy, short cervix, part-time worker, steroid use, low educational level and..., Shiozaki [/bib_ref]. Although an effect of the procedure of embryo freezing and thawing cannot be excluded (see below), all these factors, alone and/or in concert, can be responsible for the higher risk of PTB detected in children born after blastocyst transfer.
Well-designed studies [bib_ref] Birthweight distribution in ART singletons resulting from embryo culture in two different..., Lemmen [/bib_ref] [bib_ref] The protein source in embryo culture media influences birthweight: a comparative study..., Zhu [/bib_ref] [bib_ref] The type of culture medium and the duration of in vitro culture..., Vos [/bib_ref] have recently evaluated the effect of different types of in vitro culture medium and of the duration of in vitro culture on obstetric and perinatal outcomes in IVF/ICSI populations showing conflicting results, suggesting that data cannot be generalized for all commercial culture media. The different culture medium (Medicult vs. Vitrolife) and the duration of culture (day 3 vs. day 5) had no effect on singleton newborns birthweight, as well as on the incidence of PTB and VPTB, was detected on a large population of 2,098 infertile women treated with fresh IVF and ICSI cycles after controlling for confounders [bib_ref] The type of culture medium and the duration of in vitro culture..., Vos [/bib_ref]. Unfortunately, all these studies had a retrospective design and the risk of residual confounding cannot be excluded.
## Frozen-thawed vs. fresh et
The proportion of frozen ET has increased in consideration of the better reproductive outcomes compared to IVF fresh cycles [bib_ref] Fresh embryo transfer versus frozen embryo transfer in in vitro fertilizationcycles: a..., Roque [/bib_ref] [bib_ref] Fresh versus frozen embryo transfer: backing clinical decisions with scientific and clinical..., Evans [/bib_ref]. Analyzing the SART database from 2008 to 2011, odds of extra-uterine pregnancy 65 % lower in women who had a frozen compared with a fresh transfer in autologous cycles were observed [bib_ref] Ectopic pregnancy after in vitro fertilization: differences between fresh and frozenthawed cycles, Londra [/bib_ref] , that risk is lower for day 5 blastocyst vs. day 3 embryos [bib_ref] Frozen-thawed day 5 blastocyst transfer is associated with a lower risk of..., Fang [/bib_ref].
The first systematic reviews [bib_ref] Children born after cryopreservation of embryos or oocytes: a systematic review of..., Wennerholm [/bib_ref] available in the literature, including mostly pregnancy studies achieved by using slow-frozen embryos and only a few vitrification studies, concluded that the risk of adverse obstetric outcome was similar between cryopreserved embryos/ oocytes and NC children or fresh ET. A more recent systematic review with meta-analysis [bib_ref] Obstetric and perinatal outcomes in singleton pregnancies resulting from the transfer of..., Maheshwari [/bib_ref] , including 11 observational studies, suggested that singleton pregnancies arising from frozen embryos seem to offer better obstetric and perinatal outcome than those obtained after fresh oocyte cycles. The clinical benefits regarded the risk of antepartum hemorrhage (RR 0. [bib_ref] Assisted reproductive technology, epigenetics, and long-term health: a developmental time bomb still..., Grace [/bib_ref] [bib_ref] Obstetric and perinatal outcomes in singleton pregnancies resulting from the transfer of..., Maheshwari [/bib_ref]. The use of frozen ET was also subsequently confirmed to be a predictor of reduced risk of PTB when compared to fresh ET after adjusting data for main confounders (aOR 0.85, 95 % CI 0.76 to 0.94) [bib_ref] Why do singletons conceived after assisted reproduction technology have adverse perinatal outcome?..., Pinborg [/bib_ref]. However, a significant heterogeneity in the technique of freezing (slow freezing vs. vitrification), in embryo stage (cleavage stage vs. blastocyst stage vs. both), and in regimen in replacement cycles (natural cycles vs. hormone replacement cycles) was observed among studies. In addition, a bias due to difference in prognosis between populations cannot be excluded. In fact, although the risk of pregnancy complications was higher in the paired comparison of the same patient who conceived after fresh vs. subsequent frozen-thawed ET (including donor cycles) [bib_ref] Perinatal morbidity after in vitro fertilization is lower with frozen embryo transfer, Kalra [/bib_ref] , it is likely that the frozen embryo populations are more likely composed of the better prognosis patients. Similarly, also the selection bias due to the physical effects of freezing and thawing on the worse embryos cannot be excluded.
In a large retrospective cohort study [bib_ref] Perinatal outcomes of children born after frozen-thawed embryo transfer: a Nordic cohort..., Wennerholm [/bib_ref] , the perinatal outcome of 6,647 singletons conceived after frozenthawed ET were compared with 42,242 singletons born after fresh ET and 288,542 singletons born after spontaneous conception. Data, adjusted for several confounding factors, confirmed that singletons conceived after frozen-thawed ET had better perinatal outcomes than after fresh ET [bib_ref] Perinatal outcomes of children born after frozen-thawed embryo transfer: a Nordic cohort..., Wennerholm [/bib_ref]. Specifically, albeit the perinatal outcomes were worse when compared to NC, singletons born after frozen-thawed ET had a lower risk of LBW (aOR 0. [bib_ref] Obstetric and perinatal outcomes in singleton pregnancies resulting from the transfer of..., Maheshwari [/bib_ref] [bib_ref] Perinatal outcomes of children born after frozen-thawed embryo transfer: a Nordic cohort..., Wennerholm [/bib_ref]. The increased perinatal mortality rate has been showed only in few studies [bib_ref] Obstetric outcome in singletons after in vitro fertilization with cryopreserved/thawed embryos, Sazonova [/bib_ref] but not in others [bib_ref] Perinatal outcome of children born after frozen and fresh embryo transfer: the..., Pelkonen [/bib_ref] [bib_ref] Infant outcome of 957 singletons born after frozen embryo replacement: the Danish..., Pinborg [/bib_ref] [bib_ref] Neonatal outcome and birth defects in 6623 singletons born following minimal ovarian..., Kato [/bib_ref] [bib_ref] International Committee for Monitoring Assisted Reproductive Technologies world report: Assisted Reproductive Technology, Mansour [/bib_ref]. Specifically, the perinatal mortality rate resulted 25.2 per 1,000 births for fresh IVF/ICSI vs. 17.5 per 1,000 for frozen-thawed cycles [bib_ref] International Committee for Monitoring Assisted Reproductive Technologies world report: Assisted Reproductive Technology, Mansour [/bib_ref].
A Japanese large-scale registry-based study not only confirmed that SET of frozen-thawed blastocyst was associated with significantly lower odds of PTB, LBW and SGA, but, for the first time, suggested a risk three-fold higher of placenta accreta (aOR 3.16, 95 % CI 1.71 to 6.23) after frozen-thawed ET than after fresh ET [bib_ref] Impact of frozen-thawed single-blastocyst transfer on maternal and neonatal outcome: an analysis..., Ishihara [/bib_ref]. A study [bib_ref] Cryopreserved embryo transfer is an independent risk factor for placenta accreta, Kaser [/bib_ref] explored specifically this complication in patients using IVF/ICSI, with autologous or donor oocytes, undergoing fresh or cryopreserved transfer. After multivariate analysis (including many predictors such as non-Caucasian race, uterine factor infertility, prior myomectomy and placenta previa), a significant and specific association between frozen-thawed ET and placenta accreta (aOR 3.20, 95 % CI 1.14 to 9.02) was detected [bib_ref] Cryopreserved embryo transfer is an independent risk factor for placenta accreta, Kaser [/bib_ref]. On the other hand, a decreased risk of placenta previa (RR 0.71, 95 % CI 0.53 to 0.95) and placental abruption (RR 0.44, 95 % CI 0.24 to 0.83) in pregnancies after frozen thawed ET was reported [bib_ref] Obstetric and perinatal outcomes in singleton pregnancies resulting from the transfer of..., Maheshwari [/bib_ref]. Unfortunately, the demographics of patients having these complications is lacking in these studies.
The frozen-thawed embryo cycles showed an increased association with PIH/PE (aOR 1.58, 95 % CI 1.35 to1.86) [bib_ref] Impact of frozen-thawed single-blastocyst transfer on maternal and neonatal outcome: an analysis..., Ishihara [/bib_ref] , similarly to data from the Swedish IVF registry [bib_ref] Obstetric outcome in singletons after in vitro fertilization with cryopreserved/thawed embryos, Sazonova [/bib_ref]. In comparison with pregnancies following fresh cycles, those following frozen-thawed ET had a higher risk of PIH/PE with a risk difference of 1.8 % (95 % CI 1.2 to 2.8) and 5.1 % (95 % CI 3.0 to 7.1) in singleton and twin pregnancies, respectively [bib_ref] Risk of hypertensive disorders in pregnancies following assisted reproductive technology: a cohort..., Opdahl [/bib_ref]. These data were confirmed when compared with fresh cycle pregnancies by the same mother after sibling analysis (OR 2.63, 95 % CI 1.73 to 3.99) [bib_ref] Risk of hypertensive disorders in pregnancies following assisted reproductive technology: a cohort..., Opdahl [/bib_ref] and, interestingly, also in PCOS patients [bib_ref] Fresh versus Frozen Embryos for Infertility in the Polycystic Ovary Syndrome, Chen [/bib_ref]. In fact, a large multicenter RCT on a total of 1,508 infertile women with PCOS assigned to undergo either fresh-embryo transfer or embryo cryopreservation (vitrification) followed by frozen-embryo transfer demonstrated an incidence of PE significantly higher with frozen-embryo transfer [rate ratio (ReRa) [bib_ref] Assisted reproductive technologies and perinatal morbidity: interrogating the association, Barnhart [/bib_ref] [bib_ref] Perinatal outcomes of children born after frozen-thawed embryo transfer: a Nordic cohort..., Wennerholm [/bib_ref] [bib_ref] Impact of frozen-thawed single-blastocyst transfer on maternal and neonatal outcome: an analysis..., Ishihara [/bib_ref]. Of note, even if not statistically significant, the risk of LGA was higher in boys than in girls (41 % vs. 17 %) [bib_ref] Perinatal outcomes of children born after frozen-thawed embryo transfer: a Nordic cohort..., Wennerholm [/bib_ref] confirming previous data that showed an odd for LBW in frozen-thawed ET significantly higher in female than that in male neonates [bib_ref] Implications of assisted reproductive technologies on term singleton birth weight: an analysis..., Nakashima [/bib_ref]. The first paper reporting the increased risk of LGA newborns in frozenthawed ET compared with fresh ET, irrespective of maternal BMI and abnormal oral glucose tolerance test (OGTT), was published in 2010 [bib_ref] Perinatal outcome of children born after frozen and fresh embryo transfer: the..., Pelkonen [/bib_ref]. At the moment, there is not a conclusive hypothesis for explaining that finding and many potential explanations of the increased rate of LGA in frozen-thawed embryos have been proposed. Different media used for culturing of human embryos can differently affect the birth-weight of the newborns [bib_ref] Effect of in vitro culture of human embryos on birthweight of newborns, Dumoulin [/bib_ref] with an effect of the embryo freezing and thawing [bib_ref] Further evidence that culturemedia affect perinatal outcome: findings after transfer of fresh..., Nelissen [/bib_ref]. However, the effect on human embryos of the many components of the culture media is not known. A clinical study [bib_ref] The influence of the type of embryo culture medium on neonatal birthweight..., Vergouw [/bib_ref] demonstrated no difference in birth-weight between two culture media but a higher birth-weight after embryo freezing and thawing in both groups when compared with fresh cycles suggesting generic effect of cryoprotectants on enzymatic pattern involved in epigenetic programming [bib_ref] Comparative birth weights of singletons born after assisted reproduction and natural conception..., Geyter [/bib_ref]. However, a specific effect of media for extended culture cannot be excluded since babies born after day 5-6 transfer had significantly higher risk of LGA compared with babies born after day 2 transfer [bib_ref] Does long in vitro culture promote large for gestational age babies?, Mäkinen [/bib_ref] (see before).
Recently, a national register-based controlled cohort study with further meta-analysis was designed in order to clarify whether the "large baby syndrome" was caused by intrinsic maternal factors or related to the freezing/ thawing procedures [bib_ref] Large baby syndrome in singletons born after frozen embryo transfer (FET): is..., Pinborg [/bib_ref]. In order to avoid biases due to different frozen techniques and embryo culture/stage, only singletons born after slow freezing and ET of cleavage stage embryos (day 2 or 3) were included in the meta-analysis [bib_ref] Large baby syndrome in singletons born after frozen embryo transfer (FET): is..., Pinborg [/bib_ref]. The increased risk of LGA in frozenthawed embryos singletons was firstly confirmed also in a sibling cohort after adjusting data for maternal age, parity and year of birth (aOR 3.45, 95 % CI 1.33 to 8.33) [bib_ref] Large baby syndrome in singletons born after frozen embryo transfer (FET): is..., Pinborg [/bib_ref]. The pooled risk for LGA (aOR 1.54, 95 % CI 1.31 to 1.81) and macrosomia (aOR 1.64, 95 % CI 1.26 to 2.12) was confirmed in frozen-thawed embryos vs. fresh embryos [bib_ref] Large baby syndrome in singletons born after frozen embryo transfer (FET): is..., Pinborg [/bib_ref]. These risks were the same extension as when compared to NC [bib_ref] Large baby syndrome in singletons born after frozen embryo transfer (FET): is..., Pinborg [/bib_ref]. Unfortunately, these data were not adjusted for maternal BMI and GDM that can act as residual confounders [bib_ref] Large baby syndrome in singletons born after frozen embryo transfer (FET): is..., Pinborg [/bib_ref].
A recent retrospective UK cohort study [bib_ref] Obstetric and perinatal outcomes after either fresh or thawed frozen embryo transfer:..., Maheshwari [/bib_ref] analyzed a total of 112,432 cycles (95,911 fresh and 16,521 frozen cycles) using multivariate logistic regression and detected no association between type of embryo transferred (frozen vs. fresh) and PTB (aRR 0. [bib_ref] The influence of the type of embryo culture medium on neonatal birthweight..., Vergouw [/bib_ref] [bib_ref] Obstetric and perinatal outcomes after either fresh or thawed frozen embryo transfer:..., Maheshwari [/bib_ref]. Moreover, that data were anonymized and regarded outcomes of IVF cycles and not of IVF patients, and did not included pregnancy complications. Thus, a bias due to more cycles of the same patient cannot be excluded, and the influence of the obstetric complications on perinatal outcomes cannot be assessed.
Finally, two retrospective recent cohort studies suggest that frozen cycles are associated with a reduction of the incidence of ectopic pregnancies and monozygotic twins [bib_ref] Risk of ectopic pregnancy lowest with transfer of single frozen blastocyst, Li [/bib_ref]. Specifically, a registry study [bib_ref] Risk of ectopic pregnancy lowest with transfer of single frozen blastocyst, Li [/bib_ref] based on a total of 44,102 pregnancies from Australian and New Zealand Assisted Reproduction Technology Database demonstrated that the SET of frozen blastocyst is associated with lowest risk of ectopic pregnancy (aOR 0.70, 95 % CI 0.54 to 0.91; vs. fresh blastocyst transfer), and that the characteristics influencing that risk were the number of embryo transferred, and the use of fresh and/or cleavage stage embryos. In addition, the analysis of 6,103 cycles revealed that frozen ET was associated with a significant reduction in monozygotic twins incidence (aOR 0.48, 95 % CI 0.29 to 0.80).
## Vitrified vs. slow frozen vs. fresh cycles
Initially, there were some safety concerns about old vitrification procedures because the use of very high concentrations of cryoprotectants required to optimize the efficacy could induce toxic effects [bib_ref] Overall efficiency of in vitro embryo production and vitrification in cattle, Vajta [/bib_ref]. New vitrification methods requiring very small volumes and small concentrations of cryoprotectants and extreme cooling rates seem safer [bib_ref] Highly efficient vitrification method for cryopreservation of human oocytes, Kuwayama [/bib_ref] [bib_ref] Improving cryopreservation systems, Vajta [/bib_ref]. However, new generation devices may require direct contact between samples and liquid nitrogen during vitrification with the risk due to the potential exposure to contaminants.
## Vitrified vs. slow frozen vs. fresh oocytes
Specific data on the influence of oocyte vitrification on obstetric risk are few and poorly reported, and head-tohead studies are lacking. A systematic review intercepted only uncontrolled or case reports reporting very limited data on the health of the newborns from vitrified and slow frozen oocytes. Also successive studies [bib_ref] Obstetric outcomes following vitrification of in vitro and in vivo matured oocytes, Chian [/bib_ref] reported reassuring pregnancy and neonatal data, even if the population studied were very little and poorly or not controlled. Only more recently, a retrospective study [bib_ref] Comparative analysis of fetal and neonatal outcomes of pregnancies from fresh and..., Setti [/bib_ref] on 954 pregnancies achieved after ET from slow frozen (n. 197) or fresh (n. 757) oocytes reported data on pregnancy and perinatal health. No differences were found between the use of fresh and frozen oocytes in the rates of ectopic pregnancies (3.6 % vs. 2.9 %) and spontaneous abortions (17.6 % vs. 26.9 %), and in gestational age at delivery [bib_ref] Comparative analysis of fetal and neonatal outcomes of pregnancies from fresh and..., Setti [/bib_ref]. However, the mean birth weights were significantly lower with fresh oocyte than with frozen-thawed oocytes, both in singleton (2.725 ± 727 g vs. 3.231 ± 615 g) and twins (2,128 ± 555 g 2,418-± 92 g) [bib_ref] Comparative analysis of fetal and neonatal outcomes of pregnancies from fresh and..., Setti [/bib_ref]. Of interest, the analysis of the 63 patients who obtained pregnancies both in fresh and thawed cycles showed no differences in the abortion rate and in the mean birth weight [bib_ref] Comparative analysis of fetal and neonatal outcomes of pregnancies from fresh and..., Setti [/bib_ref].
Only in 2014, a retrospective cohort study [bib_ref] Obstetric and perinatal outcome of babies born from vitrified oocytes, Cobo [/bib_ref] compared the obstetric and perinatal outcomes of infertile patients who received IVF cycles after oocyte vitrification compared with those who received fresh oocytes. In particular, data from a total of 804 and 996 pregnancies, and from 1,027 and 1,224 children, respectively for vitrified and fresh oocyte group, were analyzed. No differences were found between the two techniques of cryopreservation in the rate of pregnancy complications, including obstetric, perinatal and puerperal problems [bib_ref] Obstetric and perinatal outcome of babies born from vitrified oocytes, Cobo [/bib_ref]. In the vitrified oocytes group, the incidence of invasive procedures (aOR 2.12, 95 % CI 1.41 to 3.20), such as chorionic villous sampling or amniocentesis, and of urinary tract infections (aOR 0.51, 95 % CI 0.28 to 0.91) were respectively increased and reduced [bib_ref] Obstetric and perinatal outcome of babies born from vitrified oocytes, Cobo [/bib_ref]. The proportion of female offspring was significantly higher in the vitrification group (53.8 % vs. 47.4 %) [bib_ref] Obstetric and perinatal outcome of babies born from vitrified oocytes, Cobo [/bib_ref]. Main data did not change after sub-analysis for singleton and multiple pregnancies [bib_ref] Obstetric and perinatal outcome of babies born from vitrified oocytes, Cobo [/bib_ref].
## Vitrified vs. slow frozen vs. fresh embryos
Comparable gestational age at delivery, weight at birth, and congenital anomaly rate were observed analyzing 435 blastocyst vitrification cycles using an open system, with the fresh oocyte control cycles [bib_ref] Perinatal outcome of blastocyst transfer with vitrification using cryoloop: a 4-year follow-up..., Takahashi [/bib_ref]. Other researchers have confirmed those results in further studies after the vitrification of both cleavage stage embryos [bib_ref] Neonatal outcome after vitrified day 3 embryo transfers: a preliminary study, Rama Raju [/bib_ref] [bib_ref] Clinical pregnancy and live births after transfer of embryos vitrified on day..., Desai [/bib_ref] [bib_ref] Perinatal and neonatal outcomes of 494 babies delivered from 972 vitrified embryo..., Shi [/bib_ref] [bib_ref] Obstetric and neonatal outcomes after transfer of vitrified early cleavage embryos, Liu [/bib_ref] and blastocysts [bib_ref] The freezing method of cleavage stage embryos has no impact on the..., Kaartinen [/bib_ref] [bib_ref] Obstetric outcomes after transfer of vitrified blastocysts, Wikland [/bib_ref] , or a combination of both [bib_ref] Neonatal outcome and birth defects in 6623 singletons born following minimal ovarian..., Kato [/bib_ref]. One large study including 6,623 infants conceived after vitrification, found no difference also in terms of gestational perinatal mortality, although the incidence of SGA was lower and the weight at birth was heavier after vitrification in comparison with fresh ET [bib_ref] Neonatal outcome and birth defects in 6623 singletons born following minimal ovarian..., Kato [/bib_ref]. Of note, all pregnancies and babies analyzed were conceived after SET.
One retrospective, single-center study of children born after day 3 ET from fresh, slow frozen or vitrified embryos showed that the incidence of PIH/PE, GDM, placenta previa and abruptio placenta were similar in all groups [bib_ref] Obstetric and neonatal outcomes after transfer of vitrified early cleavage embryos, Liu [/bib_ref]. Specifically, no difference between vitrified and slow freezing day 3 embryos was observed in gestational ages at delivery, PTB, and perinatal mortality, even if the birth-weight of the babies born from vitrified embryos was higher in comparison with those of the babies born from slow frozen and fresh embryos, and the rate of LBW in vitrified embryos arm was significantly lower than in fresh embryos arm [bib_ref] Obstetric and neonatal outcomes after transfer of vitrified early cleavage embryos, Liu [/bib_ref]. In twins, the birthweight for children born from vitrified day 3 embryos was higher than that from the slow freezing, even if the difference cannot be considered clinically significant [bib_ref] Obstetric and neonatal outcomes after transfer of vitrified early cleavage embryos, Liu [/bib_ref]. Unfortunately, a crucial recall bias cannot be excluded because the obstetric and perinatal data were obtained by questionnaires sent to the parents without checking medical records. Conversely, another recent retrospective study [bib_ref] The freezing method of cleavage stage embryos has no impact on the..., Kaartinen [/bib_ref] demonstrated that the freezing methods of cleavage stage embryos did not exert any effect on the neonatal weight.
No difference in almost all pregnancy and perinatal/ neonatal outcomes was observed in a Swedish retrospective study [bib_ref] Obstetric outcomes after transfer of vitrified blastocysts, Wikland [/bib_ref] comparing pregnancies achieved from vitrified blastocysts with those achieved from fresh blastocysts and slow frozen cleavage stage embryos. However, in pregnancies from vitrified blastocysts the incidence of SGA was lower (12.1 vs. 3 %) in comparison with fresh blastocyst, and the rate of major postpartum hemorrhage was higher (25 vs. 6 and 7.5 %) in comparison with fresh blastocyst and slow frozen cleavage stage embryos [bib_ref] Obstetric outcomes after transfer of vitrified blastocysts, Wikland [/bib_ref].
The largest and most recent population-based cohort study compared obstetric and perinatal outcomes of 16,845, 2,766 and 6,537 clinical pregnancies, and 13,049, 2,065 and 4,955 live deliveries of fresh, slow frozen and vitrified blastocysts [bib_ref] Clinical outcomes following cryopreservation of blastocysts by vitrification or slow freezing: a..., Li [/bib_ref]. Singletons born after transfer of vitrified but not slow frozen blastocysts had a lower risk of PTB (aRR 0.86, 95 % CI 0.77 to 0.96), LBW (aRR 0.67, 95 % CI 0.58 to 0.78) and SGA (aRR 0.60, 95 % CI 0.53 to 0.68) in comparison with singletons born after transfer of fresh blastocysts, [bib_ref] Clinical outcomes following cryopreservation of blastocysts by vitrification or slow freezing: a..., Li [/bib_ref]. The beneficial effect of vitrification on the birth weight of babies (and on the pregnancy duration) have been confirmed in a smaller retrospective study on 1,209 infertile patients who received a total of 1,157 fresh and 645 vitrified-warmed single blastocyst (day 5) transfers [bib_ref] Single-embryo transfer of vitrifiedwarmed blastocysts yields equivalent live-birth rates and improved neonatal..., Roy [/bib_ref]. A higher infant birth weight has been demonstrated not only in singletons but also in twins after retrospective analysis of 784 fresh transfers and 382 vitrified-warmed DET at blastocyst stage [bib_ref] Perinatal outcomes after fresh versus vitrified-warmed blastocyst transfer: retrospective analysis, Ozgur [/bib_ref].
An interesting retrospective study [bib_ref] Neonatal health including congenital malformation risk of 1072 children born after vitrified..., Belva [/bib_ref] analyzed the perinatal and neonatal health, adjusted for treatment variables and maternal characteristics, after 960 vitrified cycles and 1,644 fresh cycles according to the day of vitrification or fresh transfer, i.e. day 3 (cleavage stage) vs. day 5 (blastocyst stage). Singletons, but not twins, born after vitrification showed only lower SGA rate (aOR 0.55, 95 % CI 0.34 to 0.90) in comparison with peers born after fresh ET. The embryonic stage at vitrification or at fresh transfer did not influence any perinatal outcome [bib_ref] Neonatal health including congenital malformation risk of 1072 children born after vitrified..., Belva [/bib_ref]. These data seem to suggest that both vitrification and extensive culture at blastocyst stage can act and interact on fetal/neonatal weight.
## In vitro maturation (ivm) of oocytes
Any adverse effects of the IVM procedure on pregnancy and perinatal outcomes are a partially unknown field of research. From a mechanicistic point of view, IVM could affect the oocyte development, as any intervention in the growth phase could affect oocyte maturation, fertilization and subsequent embryo development [bib_ref] Oocyte maturation: gamete-somatic cells interactions, meiotic resumption, cytoskeletal dynamics and cytoplasmic reorganization, Coticchio [/bib_ref]. On the other hand, the minimal gonadotropin stimulation, which very mildly stimulates the endometrium mimics natural environmental.
A systematic reviewof head-to-head RCTs aimed to compare IVM followed by IVF/ICSI vs. conventional IVF/ICSI in terms of live birth rate and/or other efficacy and safety maternal/perinatal outcomes did not find any such study. Cha et al. [bib_ref] Obstetric outcome of patients with polycystic ovary syndrome treated by in vitro..., Cha [/bib_ref] firstly reported on obstetric and perinatal outcomes in 41 IVM pregnancies suggesting no adverse effect of the technique. However, that data were uncontrolled and obtained in PCOS patients. Subsequent controlled studies confirmed the maternal and perinatal safety of IMV in comparison to IVF and/or ICSI, even if only few obstetric and perinatal data have been reported [bib_ref] Favourable pregnancy results with insemination of in vitro matured oocytes from unstimulated..., Soderstrom-Anttila [/bib_ref] [bib_ref] Obstetric outcomes and congenital abnormalities after in vitro maturation, in vitro fertilization,..., Buckett [/bib_ref] [bib_ref] Pregnancy loss in pregnancies conceived after in vitro oocyte maturation, conventional in..., Buckett [/bib_ref]. A slightly increased rate of GDM (17 % vs. 11 % and 10 % for IVM vs. IVF and ICSI, respectively) and of PE (12 % vs. 5 % and 8 % for IVM vs. IVF and ICSI, respectively) was observed in 55 IVM patients when compared with BMI and parity matched IVF/ICSI subjects (n. 217 and n. 160, respectively) [bib_ref] Obstetric outcomes and congenital abnormalities after in vitro maturation, in vitro fertilization,..., Buckett [/bib_ref]. No difference was observed all other pregnancy and perinatal outcomes among ART groups [bib_ref] Obstetric outcomes and congenital abnormalities after in vitro maturation, in vitro fertilization,..., Buckett [/bib_ref]. In singleton IVM pregnancies, the cesarean section (39 vs. 26 %) and instrumental delivery (9.5 vs. 6.5 %) rates were higher than in NC pregnancies but no difference between IVM and NC was detected in LBW (3 vs. 9 %), VLBW (0 vs. 2 %), gestational age (39 ± 3 vs. 39 ± 6 weeks), VPTB (0 vs. 2 %) [bib_ref] Obstetric outcomes and congenital abnormalities after in vitro maturation, in vitro fertilization,..., Buckett [/bib_ref]. Of interest, contrarily to IVF/ICSI neonates, the birth weight at delivery resulted heavier in IVM neonates (3,482 g vs. 3,260 g) than in NC neonates, and the PTB incidence (5 vs. 5 %) was not different from NC pregnancies [bib_ref] Obstetric outcomes and congenital abnormalities after in vitro maturation, in vitro fertilization,..., Buckett [/bib_ref]. A possible role of COH in determining a reduced birthweight in IVF/ICSI children has been also more recently suggested in a small but well controlled retrospective cohort study [bib_ref] Comparison of the obstetric and perinatal outcomes of children conceived from in..., Fadini [/bib_ref] performed on 196 IVM cycles compared with those of ICSI treatments. Although couples with a maternal age higher than 39 years or affected by azoospermia were excluded, the sample was heterogeneous as IVM cycles had different priming regimens and maturation techniques [bib_ref] Comparison of the obstetric and perinatal outcomes of children conceived from in..., Fadini [/bib_ref]. Only a difference in birth weight was observed between IVM and ICSI babies (3091 ± 669 g vs. 3269 ± 619 g) [bib_ref] Comparison of the obstetric and perinatal outcomes of children conceived from in..., Fadini [/bib_ref]. That difference may represent an inevitable influence of drug administration [bib_ref] Comparison of the obstetric and perinatal outcomes of children conceived from in..., Fadini [/bib_ref] but also an increased incidence of imprinting disorders [bib_ref] Oocyte maturation: gamete-somatic cells interactions, meiotic resumption, cytoskeletal dynamics and cytoplasmic reorganization, Coticchio [/bib_ref] [bib_ref] Culture of oocytes and risk of imprinting defects, Anckaert [/bib_ref].
Although many studies on the safety of IVM are available in the literature, all have a retrospective design, very small sample sizes, and are not controlled and include a number of biases and (unadjusted) confounders. Obviously a high prevalence of PCOS cases in these studies could influence the results [bib_ref] Pregnancy complications in spontaneous and assisted conceptions of women with infertility and..., Palomba [/bib_ref] [bib_ref] Pregnancy complications in women with polycystic ovary syndrome: new clinical and pathophysiologic..., Palomba [/bib_ref]. In fact, a retrospective study [bib_ref] Pregnancy loss in pregnancies conceived after in vitro oocyte maturation, conventional in..., Buckett [/bib_ref] on 1,581 positive pregnancy tests demonstrated a rate of miscarriage after IVM significantly higher than after IVF/ICSI (25.3 % vs. 15.7 % and 12.6 %, respectively) but directly related to the PCOS rather than to the IVM procedure [bib_ref] Pregnancy loss in pregnancies conceived after in vitro oocyte maturation, conventional in..., Buckett [/bib_ref].
## Assisted hatching
Although hatching of the blastocyst is a crucial step for embryo implantation, and the failure to hatch can be considered one of limiting factors of the embryo development, the artificial disruption of the zona pellucida (assisted hatching) can potentially affect embryo competence and, thus, increase the risk of pregnancy complications. Although each technique used for assisted hatching can potentially exert specific effects on maternal and perinatal/neonatal health, very few data are available in literature about the effects of assisted hatching on obstetric and perinatal outcomes.
In a double-blind RCT [bib_ref] A prospective, randomized, double-blinded study of assisted hatching in women younger than..., Hagemann [/bib_ref] , no effect on assisted hatching using acidic Tyroide's solution was detected in women younger than 38 years on neonatal complications. Similarly, a retrospective cohort analysis [bib_ref] No adverse effects were identified on the perinatal outcomes after laser-assisted hatching..., Zhou [/bib_ref] including a total of 699 women found no statistically significant effect of laser-assisted hatching, performed on day 3 in frozen (slow frozen and vitrified) embryos on gestational age, birth weight and Apgar score. The safety of the laser-assisted hatching was confirmed in singleton and twin pregnancies. Moreover, these data were limited by the heterogeneous subject selection due to inclusion criteria and patient choice.
Finally, the use of assisted hatching has been associated with an increased incidence of monozygotic twinning [bib_ref] What makes them split? Identifying risk factors that lead to monozygotic twins..., Knopman [/bib_ref]. Moreover, available data are contrasting.
## Pre-implantation genetic diagnosis (pgd)
The European Society of Human Reproduction and Embryology (ESHRE) PGD Consortium reported that pregnancies and babies born after ICSI/PGD are not different from pregnancies obtained and babies born after ICSI treatment [bib_ref] The ESHRE PGD Consortium: 10 years of data collection, Harper [/bib_ref]. Similarly, gestational age, birth weight, and perinatal death were not different in a study analyzing 995 children born after ICSI/PGD and 1,507 children born after ICSI [bib_ref] Neonatal follow-up of 995 consecutively born children after embryo biopsy for PGD, Desmyttere [/bib_ref]. Interestingly, fewer multiples born after PGD presented a LBW [bib_ref] Neonatal follow-up of 995 consecutively born children after embryo biopsy for PGD, Desmyttere [/bib_ref]. These results, unfortunately, were not controlled for main confounders and did not have a NC group as reference.
A recent study [bib_ref] Neonatal outcome after preimplantation genetic diagnosis, Eldar-Geva [/bib_ref] compared the perinatal outcomes of 245 neonates born after ICSI/PGD with those of neonates born to mothers matched for age, BMI and parity during the same period after ICSI (n. 242) and after NC (n. 733). The incidence of pregnancy complications after ICSI/PGD was low and not different from control groups whereas PTB and IUGR babies were more frequent in ICSI conceptions than in NC [bib_ref] Neonatal outcome after preimplantation genetic diagnosis, Eldar-Geva [/bib_ref]. In singleton pregnancies, the weight of ICSI/PGD neonates, the rate of IUGR and LBW neonates and of PTB were not different from NC neonates but higher in comparison with ICSI neonates [bib_ref] Neonatal outcome after preimplantation genetic diagnosis, Eldar-Geva [/bib_ref].
LGA was more frequent in the PGD group than after NC [bib_ref] Neonatal outcome after preimplantation genetic diagnosis, Eldar-Geva [/bib_ref]. In twin pregnancies, no difference in LBW was detected between ICSI/ PGD and NC twins, whereas significantly more ICSI twins presented with LBW [bib_ref] Neonatal outcome after preimplantation genetic diagnosis, Eldar-Geva [/bib_ref]. That result did not change after sub-analysis for fresh and cryopreserved ET or for type of biopsy [bib_ref] Neonatal outcome after preimplantation genetic diagnosis, Eldar-Geva [/bib_ref]. Of note, the mechanical partial zona dissection was the method for zona breaching in all cases of PGD.
A very recent retrospective cohort study, using data from the U.S. National ART Surveillance System on fresh autologous cycles with blastocyst transfer, compared 97,069 non-PGD procedures with other 9,833 PGD procedures demonstrating the influence on perinatal outcome of the specific indication for PGD [bib_ref] Outcomes of in vitro fertilization with preimplantation genetic diagnosis: an analysis of..., Chang [/bib_ref]. Specifically, in women aged less than 35 years who had a live birth after ICSI/PGD the odd ratio of LBW infant was lower (aOR 0.73, 95 % CI 0.54 to 0.98) and higher (aOR 1.25, 95 % CI 1.01 to 1.54) in comparison with non-PGD cycles, when the indication was the detection of genetic disorders and aneuploidy screening, respectively [bib_ref] Outcomes of in vitro fertilization with preimplantation genetic diagnosis: an analysis of..., Chang [/bib_ref]. On the other hand, no difference in any perinatal outcome was observed in women aged more than 35 years according to PGD indication [bib_ref] Outcomes of in vitro fertilization with preimplantation genetic diagnosis: an analysis of..., Chang [/bib_ref]. Another recent multicenter-cohort study confirmed that the risk of adverse perinatal outcomes was mainly related to the underlying parental conditions rather than the PGD procedure [bib_ref] Preimplantation genetic diagnosis: a national multicenter obstetric and neonatal follow-up study, Bay [/bib_ref]. In fact, after stratification of data for PGD indication, adverse obstetric and neonatal outcomes were only present in children conceived by PGD due to parental monogenetic disorder but not in children conceived after PGD due to parental chromosomal aberrations [bib_ref] Preimplantation genetic diagnosis: a national multicenter obstetric and neonatal follow-up study, Bay [/bib_ref]. Nevertheless, there was a consistent risk of placenta previa (aOR 9.1, 95 % CI 3.4 to 24.9) after PGD and IVF/ICSI, suggesting that parental factors cannot explain all adverse outcomes [bib_ref] Preimplantation genetic diagnosis: a national multicenter obstetric and neonatal follow-up study, Bay [/bib_ref].
Notwithstanding these available scientific data, definitive conclusions on the safety of the PGD cannot be drawn since couples undergoing PGD are usually fertile and in good health, data on fetal/perinatal growth and weight can be biased by timing of the biopsy, extended embryo culture, type of ET, etc., and long-term data on offspring are totally lacking. To this regard, a recent retrospective questionnaire analysis [bib_ref] Obstetric and neonatal outcomes in blastocyst-stage biopsy with frozen embryo transfer and..., Jing [/bib_ref] showed a higher incidence of PIH (aOR 4.85, 95 % CI 1.34 to 17.56) in singleton pregnancies after blastocyst-stage biopsy and frozen ET than after cleavage-stage biopsy and fresh ET. In this regard, the Society of Obstetricians and Gynecologists of Canada [bib_ref] Pregnancy outcomes after assisted human reproduction, Society Of Obstetricians [/bib_ref] underlined that the data on an improved pregnancy outcome are inconsistent.
## Gamete and embryo donation oocyte donation (od)
Initial studies on OD populations, aiming to assess their obstetric and perinatal risk, had as crucial confounders advanced maternal age and the high rate of multiple pregnancies and VTS [bib_ref] Factors influencing the obstetric and perinatal outcome after oocyte donation, Sheffer-Miouni [/bib_ref] [bib_ref] The "vanishing embryo" phenomenon in an oocyte donation programme, Rodríguez-González [/bib_ref] [bib_ref] Outcomes of pregnancies achieved by donor egg in vitro fertilization-a comparison with..., Wiggins [/bib_ref] [bib_ref] Increased risk of pregnancy induced hypertension in young recipients of donated oocytes, Keegan [/bib_ref] [bib_ref] Placental pathology in egg donor pregnancies, Gundogan [/bib_ref] [bib_ref] Reproductive performance in oocyte donors and their recipients: comparative analysis from implantation..., Zegers-Hochschild [/bib_ref]. In addition, they were based on relatively small cohorts, which did not allow subdivision into singletons and twins, with a higher twin rates in the OD group compared with the control groups [bib_ref] Obstetric and perinatal outcome after oocyte donation: comparison with in-vitro fertilization pregnancies, Soderstrom-Anttila [/bib_ref] [bib_ref] The role of embryonic origin in preeclampsia, Klatsky [/bib_ref] [bib_ref] Association between oocyte donation and maternal and perinatal outcomes in women aged..., Ray [/bib_ref] [bib_ref] Oocyte donation in Israel: a study of 1001 initiated cycles, Yaron [/bib_ref].
An increased risk of obstetric complications, such as first trimester vaginal bleeding, PE, PIH, PTB and cesarean delivery, was reported after OD, although hypertensive disorders seem to be the most frequent one . With regard to perinatal outcomes, data are few and sparse, although most studies demonstrated no difference between OD and standard IVF [bib_ref] Factors influencing the obstetric and perinatal outcome after oocyte donation, Sheffer-Miouni [/bib_ref] [bib_ref] Obstetric and perinatal outcome after oocyte donation: comparison with in-vitro fertilization pregnancies, Soderstrom-Anttila [/bib_ref]. A large SART register study [bib_ref] Toward understanding obstetrical outcome in advanced assisted reproduction: varying sperm, oocyte, and..., Gibbons [/bib_ref] indicated that children born after OD compared with IVF have lower birth weights and gestational ages at delivery. In particular, an increased risk of LBW (aOR 1.21, 95 % CI 1.13 to 1.30) and VLBW (aOR 1.28, 95 % CI 1.10 to 1.49) in OD singletons compared with IVF pregnancies was demonstrated [bib_ref] Toward understanding obstetrical outcome in advanced assisted reproduction: varying sperm, oocyte, and..., Gibbons [/bib_ref]. Similar results were obtained previously in a cohort of 2,368 South American OD children [bib_ref] Reproductive performance in oocyte donors and their recipients: comparative analysis from implantation..., Zegers-Hochschild [/bib_ref]. Higher rates of LBW and PTB among OD singletons than in IVF/ICSI singletons was observed, although the perinatal mortality was similar [bib_ref] Reproductive performance in oocyte donors and their recipients: comparative analysis from implantation..., Zegers-Hochschild [/bib_ref].
Five meta-analyses on pregnancy complications and perinatal outcomes after OD cycles are available in the literature [bib_ref] Oocyte donation: a risk factor for pregnancyinduced hypertension: a meta-analysis and case..., Pecks [/bib_ref] [bib_ref] A meta-analysis of neonatal health outcomes from oocyte donation, Adams [/bib_ref] [bib_ref] Donor oocyte conception and pregnancy complications: a systematic review and meta-analysis, Jeve [/bib_ref] [bib_ref] Is oocyte donation a risk factor for preeclampsia? A systematic review and..., Blázquez [/bib_ref] [bib_ref] Obstetric and neonatal complications in pregnancies conceived after oocyte donation -a systematic..., Storgaard [/bib_ref]. Unfortunately, these meta-analytic data were obtained without any adjustment for confounders and including studies with very high risk of biases. The first meta-analysis [bib_ref] Oocyte donation: a risk factor for pregnancyinduced hypertension: a meta-analysis and case..., Pecks [/bib_ref] was performed in order to confirm the high risk of severe hypertensive disorders observed in OD pregnancies in their daily clinical practice. After analysis of 28 studies, the odds for PIH/ PE after oocyte donation was more than two-fold higher (OR 2.57, 95 % CI 1.91 to 3.47) in comparison with other infertility treatments, and more than six-fold higher in comparison with NC (OR 6.60, 95 % CI 4.55 to 9.57) [bib_ref] Oocyte donation: a risk factor for pregnancyinduced hypertension: a meta-analysis and case..., Pecks [/bib_ref]. The second meta-analysis [bib_ref] A meta-analysis of neonatal health outcomes from oocyte donation, Adams [/bib_ref] included 23 observational studies that compared the birth weight, mean gestational age at delivery and birth defects for conceptions after OD to those conceived from autologous oocytes [bib_ref] A meta-analysis of neonatal health outcomes from oocyte donation, Adams [/bib_ref]. Data were not controlled for confounders but analyzed only according to singleton or twin pregnancies. OD neonates were at an increased risk of PTB (RR 1.26, 95 % CI 1.23 to 1.30), LBW (RR 1.18, 95 % CI 1.14 to 1.22) and VLBW (RR 1.24, 95 % CI 1.15 to 1.35) when compared to autologous oocytes [bib_ref] A meta-analysis of neonatal health outcomes from oocyte donation, Adams [/bib_ref]. Very interesting data from sub-analysis demonstrated that in donor cycles the incidence of LBW was increased in PTB (RR 1.24, 95 % CI 1.19 to 1.29) but decreased (RR 0.86, 95 % CI 0.8 to 0.93) in term deliveries. The third meta-analysis included 11 observational studies for a total of 81,752 cycles comparing obstetric complications of pregnancies achieved after OD and autologous oocyte, and controlled for specific ART procedure (IVF or ICSI) [bib_ref] Donor oocyte conception and pregnancy complications: a systematic review and meta-analysis, Jeve [/bib_ref]. In OD pregnancies the risk of developing PIH/PE, considered the primary endpoint, was significantly higher (OR 3.92, 95 % CI 3.21 to 4.78) also after sub-analysis for singleton [bib_ref] Donor oocyte conception and pregnancy complications: a systematic review and meta-analysis, Jeve [/bib_ref]. Meta-regression for the covariate of age suggested that risk was independent of age [bib_ref] Donor oocyte conception and pregnancy complications: a systematic review and meta-analysis, Jeve [/bib_ref]. Similarly, also the fourth meta-analysis [bib_ref] Is oocyte donation a risk factor for preeclampsia? A systematic review and..., Blázquez [/bib_ref] including 11 retrospective and prospective cohort studies confirmed that OD increase the risk of PE (in comparison with homologous IVF cycles) of about 70 % and that neither multiple pregnancies nor patient age can explain that effect by meta-regression analysis.
Finally, the last systematic review with meta-analysis [bib_ref] Obstetric and neonatal complications in pregnancies conceived after oocyte donation -a systematic..., Storgaard [/bib_ref] included, after search for original studies reporting at least five OD pregnancies with a control group of pregnancies conceived by conventional IVF/ICSI or NC and case series with > 500 cases, 35 studies reporting one or more pregnancy and perinatal complications were analyzed. [bib_ref] Obstetric and neonatal complications in pregnancies conceived after oocyte donation -a systematic..., Storgaard [/bib_ref].
One of the largest and better controlled cohort study published on perinatal outcomes of children born after OD included 375 OD babies, and clinical data compared with three control cohorts of children, i.e. NC [bib_ref] Obstetric outcome after in vitro fertilization with single or double embryo transfer, Sazonova [/bib_ref] babies matched by date and year of birth) and born after either IVF (11,060 singletons, and 6,532 twins) or ICSI (5,866 singletons, and 3,101 twins) [bib_ref] Perinatal outcomes in 375 children born after oocyte donation: a Danish national..., Malchau [/bib_ref]. An increased risk of PTB (aORs 1.8, 95 % CI 1.2 to 2.3; aOR 2.5, 95 % CI 1.7 to 3.6; and aOR 3.4, 95 % CI 2.3 to 4.9, respectively) and LBW (aOR 1.4, 95 % CI 0.9 to 2.2; aOR 1.8, 95 % CI 1.2 to 2.8; and aOR 2.6, 95 % CI 1.7 to 4.0, respectively) was detected in OD pregnancies vs. control pregnancies [bib_ref] Perinatal outcomes in 375 children born after oocyte donation: a Danish national..., Malchau [/bib_ref]. Of note, the risk of PE was also increased three-fold in OD pregnancies (aOR 2.9, 95 % CI 1.8 to 4.6; aOR 2.8, 95 % CI 1.7 to 4.5; and aOR 3.1, 95 % CI 1.9 to 4.9) [bib_ref] Perinatal outcomes in 375 children born after oocyte donation: a Danish national..., Malchau [/bib_ref]. The risk remained higher also after adjusting for maternal characteristics and after subanalysis for twin pregnancies. Moreover, when the perinatal risk was adjusted for maternal PE the results improved, demonstrating no direct effect of OD on perinatal outcomes [bib_ref] Perinatal outcomes in 375 children born after oocyte donation: a Danish national..., Malchau [/bib_ref]. These data have been recently confirmed by a systematic review [bib_ref] Maternal and fetal outcomes in oocyte donation pregnancies, Savasi [/bib_ref] showing that OD is an independent risk factor for PIH/PE, especially in twin pregnancies, and that its effect on fetal birthweight or growth is minimal after adjusting for obstetric complications. Finally, a very recent national registry study confirmed that OD recipients are more likely to have PTB (aOR 1.28, 95 % CI 1.12 to 1.46) and VPTB (aOR 1.30, 95 % CI 1.03 to 1.64) when compared with autologous patients, whereas the risk of having a SGA baby (aOR 0.72, 95 % CI 0.58 to 0.89) and of perinatal death (aOR 0.29, 95 % CI 0.09 to 0.94) was lower after adjusting data for gestational age [bib_ref] Donor oocytes are associated with preterm birth when compared to fresh autologous..., Dude [/bib_ref].
## Sperm donation
Sparse available data in the literature seems to suggest an increased risk of hypertensive disorders during pregnancy, with a specific increase of the PE risk, in nulliparous and in multiparous pregnancies with changed paternity [bib_ref] The influence of donated gametes on the incidence of hypertensive disorders of..., Salha [/bib_ref] [bib_ref] Etiology and pathogenesis of preeclampsia: current concepts, Dekker [/bib_ref]. In this view, it could be hypothesized that the exposure to the paternal semen before conception has a protective effect, whereas the use of donor insemination after a previous pregnancy with paternal semen could increase the risk with an immune mechanism similar to hypertensive disorders seen in OD pregnancies. Unfortunately, at the moment data on sperm donation regard low technology interventions [bib_ref] Pregnancy complications in spontaneous and assisted conceptions of women with infertility and..., Palomba [/bib_ref] , and showed that the use of donor sperm in IUI cycles is associated with a risk of perinatal complications lower to those of the children born after partner sperm IUI and comparable to those of the NC children [bib_ref] Perinatal outcomes in 6,338 singletons born after intrauterine insemination in Denmark, Malchau [/bib_ref].
## Embryo donation
Very little is known about the relationship between embryo donation and pregnancy and perinatal complications. In fact, available data can be extrapolated from infertile populations who conceived after mixed procedures of gamete and embryo donations [bib_ref] The influence of donated gametes on the incidence of hypertensive disorders of..., Salha [/bib_ref]. At the moment, it very difficult to draw conclusions on the obstetric risk in women who had an embryo donation because the number of confounders and biases is so frequent and the detrimental effect of the double gamete donation can be only supposed. Commonly, the recipients are highly selected women with few medical comorbidities but who had had probably many previous ART failed attempts and a longer time-to-pregnancy. In addition, embryos donated are almost always frozen embryos. Finally, because embryo donation is more cost-effective than oocyte donation in case of male factor [bib_ref] A costeffectiveness comparison of embryo donation with oocyte donation, Finger [/bib_ref] , the comparison in terms of pregnancy complication can be favorable for pregnancies obtained after embryo donation.
## Surrogate pregnancy
A systematic review on gestational surrogacy has been very recently published, and the pregnancy and perinatal outcomes of gestational carriers compared, when possible, with those of standard IVF and OD cycles [bib_ref] Surrogacy: outcomes for surrogate mothers, children and the resulting families-a systematic review, Söderström-Anttila [/bib_ref]. The incidence of PIH ranged from 4.3 to 10 % and from 2.9 and 7.4 %, and the incidence of placenta previa and/or placental abruption from 1.1 to 7.9 % and from 1.1 to 3.7 %, respectively, in singleton and twin surrogate pregnancies [bib_ref] Surrogacy: outcomes for surrogate mothers, children and the resulting families-a systematic review, Söderström-Anttila [/bib_ref] , and resulted not different from those observed in IVF pregnancies and lower than that usually reported in OD pregnancies (ranged from 16 to 40 %) [bib_ref] Assisted reproduction involving gestational surrogacy: an analysis of the medical, psychosocial and..., Dar [/bib_ref]. Three cases of hysterectomies related to delivery were also reported in gestational carriers and were due to uterine atony, placenta accreta and uterine rupture [bib_ref] Surrogacy: outcomes for surrogate mothers, children and the resulting families-a systematic review, Söderström-Anttila [/bib_ref]. In surrogate singletons, the incidence of PTB and of LBW resulted, respectively, ranging from 0 to 11.5 % and from 0 to 11.1 % [bib_ref] Surrogacy: outcomes for surrogate mothers, children and the resulting families-a systematic review, Söderström-Anttila [/bib_ref]. When compared to control groups, the risk of PTB and LBW was not different from IVF singletons (incidence of PTB of 14 % and of 13.6-14.0 %, respectively) and the risk of LBW was also not different from OD singletons (incidence of 14.0 %) [bib_ref] Surrogacy: outcomes for surrogate mothers, children and the resulting families-a systematic review, Söderström-Anttila [/bib_ref]. However, a very recent US cohort study [bib_ref] National Assisted Reproductive Technology Surveillance System (NASS) Group. Trends and outcomes of..., Perkins [/bib_ref] underlined that the increased risk of PTB (aRR 1.14, 95 % CI 1.05 to 1.23) observed in gestational carriers is significantly influenced not only by multiple pregnancies but also by OD.
# Discussion
Despite the level and the quality of the current evidence it is generally suboptimal due to the presence of biases, confounders and limitations in study design [fig_ref] Table 3: Specific biases, confounders and limitations present and/or declared in the available studies [/fig_ref] , current comprehensive review confirms that subfertile women who conceived after the use of high technology infertility treatments have an overall increased risk of pregnancy and perinatal complications, and highlights that every single step and/or procedure can play an independent and crucial role. In addition, several concomitant risk factors are frequently present in the same woman and influence the clinical and biological strategy of treatment. Thus, it is virtually impossible to define the weight of each reproductive treatment's phase determining the whole patient's risk. In fact, the infertility condition represents a bias per se in every study dealing with infertility treatments [bib_ref] Pregnancy complications in spontaneous and assisted conceptions of women with infertility and..., Palomba [/bib_ref] and the presence of many confounding factors cannot be always adequately controlled through multivariate analysis because in many studies they are not clinically available, missing, or not collected.
Ideally, the knowledge of the pathogenesis of the increased risk of pregnancy and perinatal complications in women who receive high technology infertility treatments and of the specific mechanisms of action could be crucial for preventing them. Unfortunately, few data are available regarding the biological explanations of that increased risk. Many mechanisms have been hypothesized and regard the alterations of the early placentation including not only alterations in endometrial receptivity, Case-control study The study reports the experience of a tertiary referral center; therefore, this factor may have inflated the observed rates of peripartum hysterectomy.
The control populations have not been separated into two groups (normal fertile couples and infertile couples who conceive without treatment) to determine the degree to which observed associations are specifically related to the ART procedures vs. infertility per se.
-Detailed information on the infertility treatments was not available.
-Small study number.
Pandey et al., 2012 [bib_ref] Obstetric and perinatal outcomes in singleton pregnancies resulting from IVF/ICSI: a systematic..., Pandey [/bib_ref] Review of 20 matched cohort studies and 10 unmatched cohort studies Ascertainment bias with the findings of increased complications with IVF/ICSI; i.e. women may be more anxious following fertility treatment and therefore more likely to report problems.
The quality of most of the studies was high and they have adjusted for most important confounders of age and parity.
-Some authors excluded pregnancies resulting from ovulation induction whereas others were not able to identify them from all non-IVF/ICSI conceptions.
-The review cannot determine whether the increased risk is due to the inherent infertility itself or the process of ovarian stimulation and/or embryo culture. Longitudinal, retrospective cohort study
The focus was not on procedure specifics, even if each artificial procedure has profound effect on the splitting of the zygote, which represents a bias in the comparison of spontaneous and IVF-ICSI VT-pregnancies.
The relatively low incidence of VTS denotes the low power of the statistical analyses. Retrospective study based on national registry
The effect of the subfertility has not been prevented.
The dataset had information on few confounders.
Detailed information on the infertility treatments was not available.
Sunkara et al., 2015 [bib_ref] Increased risk of preterm birth and low birthweight with very high number..., Sunkara [/bib_ref] Prospective cohort study
The effect of the different gonadotropin dosages has not been excluded.
The dataset had no information on important confounders such as smoking, BMI and the medical history of women during pregnancy.
-The dataset did not allow specific identification of women with PCOS and its anonymized nature did not make it permissible to analyze one cycle per woman.
-Individual women would have contributed to more than one cycle and outcome in the data set which means that the true sample size is unknown.
Kalra et al., 2012 [bib_ref] Extended embryo culture and an increased risk of preterm delivery, Kalra [/bib_ref] Retrospective cohort study
To attempt to control selection bias, subanalyses were performed.
-Adjusted analyses were performed and included variables that were considered clinically important, because they are associated with adverse outcome.
-Data not adjusted for gonadotropin dose.
## Mäkinen et al., 2013 [65]
Retrospective crosssectional cohort study
The effect of the subfertility has not been prevented.
-The study was not adjusted for smoking and for gonadotropin doses.
-Because of insufficient perinatal data, the authors were no able to control additional factors known to affect pregnancy outcome such as PIH, PE and GDM.
Lack of control of the duration of infertility. Retrospective study based on national registry
The different protocols and criteria for the use of frozen ET and blastocyst transfer potentially could bias the data.
The Japanese registry is cycle based with complete anonymity, therefore, it is impossible to know the detailed background of the patients who underwent ART, e.g.,
-Wide variability of data compiled from almost 600 clinics that are different in size, location, and other characteristics. National registerbased controlled cohort study with meta-analysis A bias is very unlikely as data coding was based on national registers
The data were not adjusted for confounding factors, such as maternal BMI and GDM.
The size of the frozen ET/fresh sibling cohort was limited.
Cobo et al., 2014 [bib_ref] Obstetric and perinatal outcome of babies born from vitrified oocytes, Cobo [/bib_ref] Retrospective cohort study
To avoid any selection bias, the study included the entire population of women from the two analyzed cohorts as were originally present in the Clinic.
-The study analyzed all the births for which there was notification, and not the whole series of IVF/ICSI pregnancies achieved in the Institution during the study period.
-The conclusions are based on retrospective data that were partially obtained through medical questionnaires.
-Information on pregnancy losses before 24 weeks is lacking (i.e., ectopic pregnancies, early and late miscarriages, and terminations of pregnancy due to fetal abnormalities).
-The statistical power may be limited to detect a minor increase in the incidence of negative rare outcomes (i.e., major congenital malformations)
Li et al., 2014 [bib_ref] Clinical outcomes following cryopreservation of blastocysts by vitrification or slow freezing: a..., Li [/bib_ref] A population-based cohort study
The effect of the subfertility has not been prevented.
The study used each treatment cycle as the unit of analysis where one woman could be included in both fresh and thaw cycles.
-Lack of information available on clinical-specific cryopreservation protocols and processes for slow freezing-thaw and vitrification-warm of blastocysts and the potential impact on outcomes. The lack of consistent cryopreservation protocols and comparison of embryo qualities might over-estimate the successful rate of vitrification and under-estimate the successful rate of slow freezing of blastocysts.
-The data are observational and hence conclusions concerning the biological effects of vitrification and slow freezing cannot be drawn from our study Observational study Risk of an ascertainment bias. The database includes women with PCOS and the effect on birth weight may be a result of the inherent PCOS, rather than as a direct result of the treatment modality.
-Retrospective design.
-Small sample size.
Jing et al., 2016 [bib_ref] Obstetric and neonatal outcomes in blastocyst-stage biopsy with frozen embryo transfer and..., Jing [/bib_ref] Retrospective cohort study
The effect of the subfertility has not been prevented.
To reduce the significant differences in genetic disorders, number of transferred embryos, methods of genetic testing, and vanishing twin between the two groups a logistic regression was applied in the study.
-Retrospective design.
-Small sample size.
-The study focused on obstetrics and neonatal out-comes and only included patients who were > 28 weeks pregnant.
Storgaard et al., 2016 [bib_ref] Obstetric and neonatal complications in pregnancies conceived after oocyte donation -a systematic..., Storgaard [/bib_ref] Systematic review with meta-analysis of 22 cohort studies and 13 annual report of ASRM
The effect of the subfertility has not been tested.
-OD patients are very heterogeneous regarding age, inheritance and infertility history.
-Oocyte donors also constitute a heterogeneous group. This affects -To ensure reliable results only studies of high and medium quality were included in the meta-analyses (of the 21 included cohort studies comparing an OD group to a control genetic and epigenetic mechanisms of implantation, invasion and growth of the trophoblast but also genetic and/or epigenetic alterations of oocyte/embryos due to biological manipulations (extended culture, culture media, techniques of cryopreservation, etc.), and immunological intolerance in case of OD because the fetal genome is allogenic to the carrier [bib_ref] Pregnancy complications in spontaneous and assisted conceptions of women with infertility and..., Palomba [/bib_ref] [bib_ref] Cryopreserved embryo transfer is an independent risk factor for placenta accreta, Kaser [/bib_ref] [bib_ref] Pregnancy complications in women with polycystic ovary syndrome: new clinical and pathophysiologic..., Palomba [/bib_ref] [bib_ref] Placental pathology in egg donor pregnancies, Gundogan [/bib_ref] [bib_ref] Interand intra-individual variation in allele-specific DNA methylation and gene expression in children..., Turan [/bib_ref] [bib_ref] Clinical and immunologic aspects of egg donation pregnancies: a systematic review, Van Der Hoorn [/bib_ref] [bib_ref] Macroscopic and microscopic findings of the placenta in women with polycystic ovary..., Palomba [/bib_ref]. Thus, in the next future, still remains the need and an effort should be made to understand the reasons of these risks in order to minimize or prevent them. However, from the clinical point of view, the priority is not to precisely estimate the amount of the obstetric risk but to recognize the presence of one or more risk factors (infertility and subfertility causes, patient's characteristics and specific ARTs-associated risks), to correct those modifiable and to strictly follow the resulting pregnancies with appropriate prenatal cares. In fact, the delay in receiving prenatal care increased the PTB risk, while more-frequent use of prenatal care significantly improved the birth weight among pregnancies at high risk including subfertile women [bib_ref] Prenatal care effectiveness and utilization in Brazil, Wehby [/bib_ref]. Recently, a large nationwide population-based study demonstrated that an adequate and intensive prenatal care reduces the risk of adverse pregnancy outcomes in women with history of infertility [bib_ref] Adequate prenatal care reduces the risk of adverse pregnancy outcomes in women..., Alibekova [/bib_ref]. Specifically, less than six prenatal visits (compared with equal or more than six prenatal visits) and prenatal visits performed after the 12th week of gestation (compared with prenatal visits performed at or before the 12 th week of gestation) are related with a risk lower of VLBW neonates (aOR15.1, 95 % CI 8.8 to 25.8; aOR 2.1, 95 % CI 1.2 to 3.8; respectively), LBW neonates (aOR 2.1, 95 % CI 1.5 to 2.5; aOR 1.6, 95 % CI 1.3 to 1.9; respectively), and preterm birth (aOR 2.2, 95 % CI 1.9 to 2.6; aOR 1.1, 95 % CI 0.9 to 1.3; respectively) [bib_ref] Adequate prenatal care reduces the risk of adverse pregnancy outcomes in women..., Alibekova [/bib_ref]. That data, however, were limited to singleton pregnancies.
At the moment, well established strategies to identify, follow and manage infertile patients and/or patients who have receive an infertility treatment are lacking and only few papers suggest potential strategies of management consisting in a generic close pregnancy monitoring and diagnostic testing [bib_ref] Pregnancy outcomes after assisted human reproduction, Society Of Obstetricians [/bib_ref] [bib_ref] Maternal and fetal outcomes in oocyte donation pregnancies, Savasi [/bib_ref]. Physicians should assess the pregestational risk of infertile women before start any fertility enhancement treatment and discuss with the couples the increased risk for maternal and perinatal complications in a view of risk to benefit ratio and the potential alternatives, suggesting also to avoid any medical intervention in case of high-risk patients [bib_ref] Pregnancy outcomes after assisted human reproduction, Society Of Obstetricians [/bib_ref] , such as in case of women of very advanced reproductive age (> 55 years) or advanced reproductive age (> 45 years) with medical conditions.
# Conclusion
Subfertile women who conceived after the use of high technology infertility treatments are at increased risk of pregnancy complications, and every single/specific step and/or procedure can play an independent and crucial role. Thus, all infertile patients scheduled for high technology infertility treatments should be clearly informed of that increased obstetric and perinatal risk in case of pregnancy, regardless of multiple pregnancy. A careful preconceptional counselling aimed to optimize the general health status of the pre-pregnant women is needed (to stop smoking, reduce BMI in overweight/obese patients, and so on), identifying and treating modifiable [fig_ref] Table 3: Specific biases, confounders and limitations present and/or declared in the available studies [/fig_ref] Specific biases, confounders and limitations present and/or declared in the available studies (Continued) pregnancy rates, but it is not known whether it influences obstetric and neonatal outcomes. group only two were of high quality and 11 were of medium quality).
-Some outcomes were poorly defined, e.g. only three out of 11 studies included a strict definition of gestational diabetes Retrospective cohort study
The effect of the subfertility has not been prevented.
To limit the confounding variables, women who conceived with donated gametes were compared to age-and paritymatched controls from similar demographic backgrounds who conceived with their own gametes.
-Retrospective design.
-Small sample size.
Söderström-Anttila et al., 2016 [bib_ref] Surrogacy: outcomes for surrogate mothers, children and the resulting families-a systematic review, Söderström-Anttila [/bib_ref] Systematic review of observational studies (cohort studies and case-series) Cohort studies, but not case series, were assessed for methodological quality, in terms of risk of bias.
-Lack of high quality studies.
-Most studies have small sample size, lack of controls and a low response rate. -Gestational and traditional surrogacy was not always separated in the studies.
ART assisted reproductive technologies, ASRM American society of reproductive medicine, BMI body mass index, DET double embryo transfer, ET embryo transfer, GDM gestational diabetes mellitus, ICSI intracytoplasmic sperm injection, IUI intrauterine insemination, IVF in vitro fertilization, NC natural conception, OD oocyte donation, OI ovulation induction, PE preeclampsia, PIH pregnancy-induced hypertension, SET single embryo transfer, VTS vanishing twin syndrome reproductive disorders [bib_ref] Pregnancy complications in spontaneous and assisted conceptions of women with infertility and..., Palomba [/bib_ref] and, finally, an effort should be made to optimize the infertility treatments (milder stimulation, OHSS prevention, elective SET) in order to prevent or reduce the risk of pregnancy complications in these infertile women. Finally, further large cohort prospective studies are required to clarify the contribution of each single factor on pregnancy and perinatal outcomes.
[fig] Figure 1: PRISMA 2009 [13] flow diagram [/fig]
[table] Table 1: Key words used to study the relationship between specific procedure of ARTs for female or couple infertility and obstetric and perinatal outcomes [/table]
[table] Table 2: Levels and quality of the evidences available about the relationship between specific procedure for ARTs and risk of the main obstetric and perinatal adverse outcomes [/table]
[table] Table 3: Specific biases, confounders and limitations present and/or declared in the available studies [/table]
|
Mucin expression in gastric- and gastro-oesophageal signet-ring cell cancer: results from a comprehensive literature review and a large cohort study of Caucasian and Asian gastric cancer
Background The literature on the prognostic relevance of signet-ring cell (SRC) histology in gastric cancer (GC) is controversial which is most likely related to inconsistent SRC classification based on haematoxylin-eosin staining. We hypothesised that mucin stains can consistently identify SRC-GC and predict GC patient outcome. Methods We performed a comprehensive literature review on mucin stains in SRC-GC and characterised the mucin expression in 851 Caucasian GC and 410 Asian GC using Alcian Blue (AB)-Periodic Acid-Schiff (PAS), MUC2 (intestinal-type mucin), and MUC5AC (gastric-type mucin). The relationship between mucin expression and histological phenotype [poorly cohesive (PC) including proportion of SRCs, non-poorly cohesive (non-PC), or mucinous (MC)], clinicopathological variables, and patient outcome was analysed. Results Depending on mucin expression and cut-offs, the positivity rates of SRC-GC reported in the literature varied from 6 to 100%. Patients with MUC2 positive SRC-GC or SRC-GC with (gastro)intestinal phenotype had poorest outcome. In our cohort study, PC with ≥ 10% SRCs expressed more frequently MUC2, MUC5AC, and ABPAS (p < 0.001, p = 0.004 and p < 0.001, respectively). Caucasians with AB positive GC or combined ABPAS-MUC2 positive and MUC5AC negative had poorest outcome (all p = 0.002). This association was not seen in Asian patients. Conclusions This is the first study to suggest that mucin stains do not help to differentiate between SRC-GC and non-SRC-GC. However, mucin stains appear to be able to identify GC patients with different outcome. To our surprise, the relationship between outcome and mucin expression seems to differ between Caucasian and Asian GC patients which warrants further investigations.
# Introduction
Gastric cancer (GC) is a heterogenous disease with respect to epidemiology, morphology, and clinical behaviour. Despite declining incidence, GC remains one of the major causes of cancer-related death worldwide. The incidence of poorly cohesive GC, including signet-ring cell (SRC) cancer, appears to be rising in the Western World. Several studies investigating the prognostic relevance of SRC histology reported conflicting results. In some studies, SRC histology was associated with poor outcome, which was not confirmed in other studies. Furthermore, it has been suggested that the relationship between SRC histology and outcome may depend on the disease stage in GC patients. The clinical utility of the proportion of SRCs to predict response to preoperative chemo(radio)therapy in GC patients remains a matter of debate.
A recent expert panel hypothesised that these inconclusive results could be related to inconsistencies in the histological classification of SRC-GC. Whilst SRC-GC have always been typed as diffuse-type cancers in the Lauren classification, the WHO definition of SRC-GC changed several times between the 1st edition in 1977and the 4th edition in 2010. Up to the 4th edition, when SRC-GC became a subcategory of poorly cohesive GC, SRC-GC was classified as a separate specific subtype of GC. Furthermore, the definition of the extent of SRCs to qualify as SRC-GC changed over the years from 'predominant', to more than 50% SRCs in the 2nd edition WHOand back to "predominantly" or "exclusively" in the 4th and 5th editions WHO. In an attempt to achieve a more consistent classification of SRC-GC, a multidisciplinary expert panel recently proposed specific cut-off values for the percentage of SRCs to distinguish bona-fide SRC-GC (more than 90% SRCs) from poorly cohesive GC with SRC component (between 10 and 90% SRCs) and poorly cohesive GC not otherwise specified (less than 10% SRCs). However, what remains particularly challenging is the unequivocal definition of what constitutes a SRC based on routine histology as exemplified by the 5 different types of SRCs described in the 3rd edition WHO. Therefore, there remains an urgent clinical need to identify a specific biomarker for SRCs to standardise SRC-GC classification and establish the clinical importance of SRC-GC.
We hypothesised that (1) SRC containing gastric cancers have a different mucin expression compared to non-SRC gastric cancers, andthere is an association between SRC mucin expression, clinicopathological variables, and patient outcome.
The present study consists of two parts: (1) a comprehensive literature search to establish the frequency and clinical importance of mucin stains in SRC containing GC, and (2) a large cohort study in Asian and Caucasian GC where the histological phenotype was classified according to recently published consensus guidelines and the expression of several different mucin stains and its relationship to clinicopathological variables, patient outcome, and ethnicity was investigated.
# Materials and methods
## Literature review
A comprehensive literature search was conducted in the PubMed database including all publications up to October 1st, 2018 using synonyms and MESH terms for 'gastric' and 'signet-ring cell cancer' (see Online . The title and abstract of resulting articles were screened based on the following inclusion criteria: the abstract or title mentioned SRC-GC, and results from histochemical or immunohistochemical mucin stains were reported separately for SRC-GC (see. If a study provided both, data from an SRC-GC and non-SRC-GC cohorts, only the SRC-GC data were extracted and analysed. We excluded studies reporting results from less than 10 SRC-GC, referring to hereditary diffuse GC, metastases with SRCs or unknown primary tumour, cell culture-based studies, animal studies, diagnoses based on cytology, and letters to editors containing no additional information. We also excluded studies where we were unable to retrieve the full-text version of the article, articles written in languages other than English or Japanese, and articles reporting only on gastro-oesophageal junction adenocarcinomas.
From the included studies, we extracted information about the definition of SRC-GC, classification system used, frequencies of positivity for histochemical and immunohistochemical mucin stains, differences in mucin expression between SRC-GC and relationship between mucin expression in SRC-GC, patient outcome, and other clinicopathological factors.
## Gastric cancer cohort study
## Patients
We included 851 Caucasian patients from Leeds Teaching Hospitals NHS Trust (LTHT), Leeds, UK, and 410 Asian patients from the Kanagawa Cancer Center Hospital (KCCH), Yokohama, Japan, diagnosed with adenocarcinoma of the stomach or gastro-oesophageal junction. All patients underwent potentially curative gastrectomy or gastro-oesophagectomy with lymph node dissection. 76 patients had neoadjuvant chemotherapy. Clinical variables (age, sex, treatment, overall and 5-year survival, and mortality status) and histopathological variables [(y)pT, (y)pN, grade of differentiation, and tumour location] were retrieved from hospital records and pathology reports. The study protocol was approved by the relevant local ethics committees.
Tissue microarrays (TMAs) had been previously constructed from the resection specimen. Three 0.7 mm cores where sampled from each LTHT GC and two 1.2 mm cores from each KCCH GC. The cores were taken from areas with the highest tumour density in both cohorts. Four micron thick sections were cut and stained with haematoxylin-eosin (H&E) using a standard protocol and subjected to histochemical stains and immunohistochemistry as described below. Slides were scanned at × 40 magnification using an Aperio XT Scanner (Aperio Technologies, Vista, CA, USA) at the University of Leeds slide scanning facility and viewed using Aperio ImageScope (version 12.3.2.8013). Scoring of the immunohistochemical stains was performed independently by two observers (D.L. and K.K.), and discrepant scores were reviewed and discussed jointly until agreement was reached.
## Histological classification
The H&E stain was used to classify GC based on 5th edition WHO Classification of Tumours of the Digestive Systemin combination with the recently published consensus. A poorly cohesive GC was defined as a cancer composed of isolated neoplastic cells or small aggregates of neoplastic cells. To be classified as a mucinous GC, more than 50% of tumour area had to be occupied by pools of extracellular mucin. GC not fulfilling the criteria for either poorly cohesive or mucinous GC were classified as non-poorly cohesive GC.
A classical SRC was defined as a cell with ample optically clear cytoplasmic mucin on H&E stain and eccentrically placed nucleus. To compare the mucin expression characteristics of SRCs in poorly cohesive GCs with lookalikes SRCs in non-poorly cohesive and mucinous GCs, we decided to quantify the proportion of SRCs in all histological subtypes based on the H&E appearance in 4 categories: < 10% SRCs, ≥ 10-50% SRCs, ≥ 50-90% SRCs, and ≥ 90% SRCs. This classification was performed blinded to (immuno)histochemical expression results or clinicopathological information.
## Mucin histochemistry and immunohistochemistry
All GCs were stained for Alcian Blue (AB)-Periodic Acid-Schiff (PAS) combined with pan-cytokeratin (CK-ABPAS), MUC2, and MUC5AC. We chose ABPAS to differentiate between acidic and neutral mucin. ABPAS stain was combined with immunohistochemistry for pan-cytokeratin to distinguish between mucin within epithelial cells, mucin within non-epithelial cells such as macrophages and extracellular mucin. We chose MUC2 as intestinal-type mucin stain and MUC5AC as gastric-type mucin stain, since these
## Papers
Mucin stains in SRC-GC 49 papers -Not reporƟng on mucin stains stains were most frequently used in studies in the literature. After deparaffinisation in xylene and rehydration following a standard protocol, the pan-cytokeratin stain was performed first (see below) followed by the ABPAS stain. We used 1% AB solution (pH 2.5), 1% periodic acid solution, and Schiff's solution following our routine laboratory protocol. Slides were dehydrated, coverslipped, and scanned.
For all antibodies, antigen retrieval was performed in a microwavable pressure cooker using 10 mM citrate buffer, pH 6.0, and full pressure for 5 min. Hydrogen peroxide and egg white solution were used to block endogenous peroxidase activity and endogenous biotin, respectively. The sections were incubated with antibodies against MUC2 (dilution 1:100, clone CCP58; Agilent/Dako), MUC5AC (dilution 1:100, clone CLH2; Agilent/Dako), and pancytokeratin (dilution 1:200, clone AE1/AE3; Agilent/Dako) for 1 h at 37 °C. Dako Real streptavidin-biotin detection kit was used as detection system and 3,3′-Diaminobenzidine (DAB) as chromogen (Dako) according to the instructions of the manufacturer. Slides were counterstained with Mayer's haematoxylin, dehydrated, coverslipped, and scanned.
## Scoring of mucin stains
For all three stains (MUC2, MUC5AC, and ABPAS), a core with ≥ 10% stained tumour cells was classified as positive, irrespective of expression intensity. For this scoring, all tumour cells were considered irrespective of their morphology (signet-ring cell or not). If one of the cores of a case was classified as being positive, the whole case was classified as positive.
ABPAS-positive cores were initially subdivided into 5 categories: (1) AB positive; (2) PAS positive; (3) mixed positivity-same cells AB and PAS positive; (4) mixed positivity-different cells AB and PAS positive; (5) combination of categories 3 and 4. A case was classified as positive for both ABPAS, if AB and PAS positivity was seen in the same core (category 3-5) or if one core was AB positive (category 1) and another core was PAS positive (category 2). As the number of cases in certain ABPAS subcategories was very small, we subsequently combined groups (see.
Furthermore, we created additional variables by combining the results of the different mucin stains (ABPAS, MUC2, and MUC5AC) (see Online Resource 2).
## Statistical analyses
Data were analysed using SPSS Statistics for Windows version 25.0 (IBM, Armonk, NY, USA). Patient characteristics, histological tumour types, and mucin expression were compared between the LTHT and the KCCH cohorts using the Mann-Whitney U test for continuous variables and the Pearson Chi-square test for categorical variables. For the analysis of the association between histological phenotype and mucin expression, results from both cohorts (LTHT and KCCH) were combined. Associations between mucin expression and clinicopathological variables were assessed using the Kruskal-Wallis test. For survival analyses, individual mucin expression and combinations of mucin expressions (see Online Resource 2) were used. Kaplan-Meier survival analysis and log-rank test were used to compare 5-year and overall survival between patients with different histological tumour types or different mucin expression. 16 LTHT patients and 60 KCCH patients underwent neoadjuvant chemotherapy, respectively. To determine whether neoadjuvant treatment influenced the results, all analyses were repeated excluding these 76 patients. As all associations remained the same, we report here results based on the whole cohort. p values < 0.05 were considered statistically significant.
# Results
## Literature review of mucin stains in gastric signet-ring cell cancers
The literature search in PubMed resulted in only 35 studies published over a period of 40 years (between 1977 and 2017) (see. The median number of SRC-GC patients per study was 37 (range 11-317), and the percentage of SRC-GC within individual studies ranged from 7 to 100%. The majority of studies (n = 25, 71%) originated from Asia.
## Definition of gastric signet-ring cell cancers and classification system used
The definition of SRCs and the classification system used varied between studies (see. Some authors provided a very detailed description mentioning intracellular accumulation of PAS, AB, and/or mucicarmine positive mucintogether with an eccentric nucleus as defining factorsor described 'classical SRCs' next to small SRCs with deep eosinophilic cytoplasm and a round larger hyperchromatic nucleus. Other authors neither described the definition of SRC nor mentioned the classification system used. Some studies used cut-off values to define SRC-GC, either in concordance with the 2nd and 3rd editions WHO classification (50% rule)or using the 4th edition WHO classification(predominant or exclusive rule without specific percentage). One of the studies used a cut-off value of 90% to define a GC as SRC-GC.
## Gastric signet-ring cell cancer and histochemical mucin stains
We identified 12 studies reporting on histochemical mucin stains in SRC-GC published between 1977 and 2013. The median number of included SRC-GC cases was 28 (range 11-102).
Different studies used single stain or combination of stains to classify SRC-GC (see.
Three studies combined morphological features and histochemical mucin stains to classify SRC-GC into several subtypes. Kubota et al.used AB, AB-PAS, and LNAase stains to classify 64 SRC-GC as type A (immature: PAS weak positive, AB negative, LNAase positive, small cell size, and high nuclear/cytoplasmic ratio), type B (intermediate: stronger PAS positivity, AB negative or weak positive, LNAase positive, and smaller nuclear/cytoplasmic ratio) or type C (mature: PAS strong positive, AB positive, and eccentric nucleus). The C-type was further divided in C1 and C2 subtypes with the C1 subtype showing weaker AB positivity combined with LNAase positivity and the C2 subtype showing stronger AB positivitycombined histochemical stains for PCS, GOS, and sialidase GOS with immunohistochemical stains for pepsinogen I and II to classify 127 SRC-GC as gastric phenotype, intestinal phenotype, or mixed gastrointestinal phenotype. The gastric phenotype was the most prevalent subtype (in both studies approximately 74%). However, the above mentioned detailed histomorphological subclassifications of SRC-GC have not been validated in any subsequent studies and the relationship with clinical variables has only been investigated by Akamatsuwho did not find any associations.
## Frequency of histochemical mucin positivity in gastric signet-ring cell cancers
Fujiyoshi et al.categorised SRCs into 'classical' (33.3%) and 'non-classical' based on AB positivity, whereas Bakkelund et al.used PAS positivity as defining feature of SRCs, although the amount of PAS positive material varied between SRCs in the same GC and between GCs. Work by Terada et al.suggested that SRCs were positive for PAS-D, AB and mucicarmine, whereas Santini et al.showed that 51% of SRC-GCs were positive for PAS, AB, and high-iron diamine (HID), 12% positive for AB or PAS and HID, and 37% showed PAS positivity only. Takenoshita et al.compared PAS and AB positivity between 'pure' SRC-GC and tubular or poorly differentiated GC containing SRCs. All pure SRC-GC were strongly PAS positive and 71% were also strongly AB positive. Interestingly, the AB and PAS positivity rate was similar in SRC containing tubular or poorly differentiated GCs, and only the expression intensity was found to be lower. Furthermore, SRC-GC seemed to have stronger AB expression compared to PAS expression (p < 0.01).
In summary, the frequency of AB positive SRC-GC varied from 56%to 64%, and PAS positivity varied from 61%to 95%. One study described HID positivity in 63% of GCs.
## Relationship of histochemical mucin positivity with clinicopathological variables
We did not find any study investigating the relationship between AB and/or PAS positivity and clinicopathological variables or patient outcome.
A single study combined histochemical mucin stains (PCS, GOS, and sialidase GOS) and immunohistochemical stains [pepsinogen II, SH-9 (surface mucous cell stain), and TKH-2 (goblet cells and parietal cell stain)] to define the phenotype of 203 SRC-GC as gastric [> 90% surface mucous cell type or pyloric gland cell type, n = 130 (64%)], intestinal [> 90% goblet cell type or microcyst type, n = 4 (2%)], or mixed gastrointestinal [10-90% gastric and/or intestinal cell types, n = 69 (34%)]. The proportion of gastric phenotype SRC-GC decreased with increasing depth of invasion, whereas the proportion of mixed phenotype SRC-GC increased with depth of invasion. This could suggest that the progression of SRC-GC may be associated with a phenotypic shift from gastric to intestinal-type mucin expression.
## Immunohistochemical stains for mucin expression of gastric signet-ring cell cancers
Twenty-six studies using immunohistochemistry were published between 1997 and 2017, and the median (range) number of SRC-GC was 31 . The results are summarised in.
We noted a wide variation of cut-off values used to classify a cancer as being positive for a particular marker which most likely explains the wide range in reported positivity frequencies. MUC4, STn, and trefoil factor family peptide (TFF) 1 and TFF3 were only investigated in a single study and reported to be positive in 57%, 57% to 71%, 33.3% and 100%SRC-GC, respectively.
The frequency of single mucin stain positive SRC-GC ranged from 13%to 95%for MUC2, 11%to 100%for MUC5AC, 29%to 71%for MUC6 and 6%to 97%for EMA/MUC1. A single study by compared the frequency of mucin expression between pure SRC-GC and SRC-GC combined with < 50% tubular or papillary component and did not find any differences for MUC2, MUC5AC, or MUC6.
Some studies suggested a relationship between mucin expression and depth of invasion. MUC5AC expression was seen in tumour cells in the superficialpart of the gastric wall, whereas tumour cells positive for MUC6or PCS IIIwere more frequently found in deeper parts of the wall. The location of MUC2 positive tumour cells appeared to be more variable and could be superficially, central/marginal, or in a mosaic pattern.
Several studiesinvestigated combinations of mucin stains in SRC-GC (seeand described four SRC-GC phenotypes: gastric (G), intestinal (I), gastrointestinal/mixed (GI), and unclassified (UC). The G-type showed expression of one or more gastric mucin stains (MUC5AC and/or MUC6 and/or PCSIII and/or M-GGMC-1) and absence of intestinal-type mucin stain MUC2; the I-type showed expression of intestinal mucin stains and absence of gastric mucin stains; the GI-type showed a combination of stains and the unclassified type is negative for any mucin stains (for details, see Online Resource 3). However, the reported frequency of these phenotypes in SRC-GC varied: G-type ranged from 0%to 81%, I-type from 0%to 18%, and GI-type 0%to 69%. Aihara et al.found a difference in mucin phenotype when comparing SRC-GC with non-SRC-GC (SRC-GC: 41 G-type, 28 GI-type; non-SRC-GC: 20 G-type, 31 GI-type, p = 0.029).
Seki et al.divided 35 intramucosal SRC-GC in three groups based on the type of ILS (as described by Akamatsu et al.: complete type only (group A-40%), both complete and incomplete type (group B-49%), and incomplete type only (group C-11%). MUC2 expression was different (29%, 76%, and 100% of GC in groups A, B, and C, respectively. p = 0.0006).
## Relationship between immunohistochemical mucin stains and clinicopathological variables including patient outcome
Some studies reported a poorer outcome for patients with MUC1 positive SRC-GC. GI and I-type SRC-GC were associated with poorer overall outcome compared to G and UC-type (31.82% vs. 68.75%, p = 0.0146). No association was found between patient outcome and MUC2 or STn positivity.
MUC2 positive or GI-type SRC-GC was associated with larger tumour diameter, increased depth of invasion, presence of lymph-node metastases, or lymphovascular invasion. Patients with G-type SRC-GC had smaller tumours, lower rates of lymph-node metastasis, or vascular invasion compared to other phenotypes (both p < 0.01). Furthermore, Xiong et al.found that MUC5AC expression was inversely associated with depth of invasion. Bamba et al.showed that mucosal tumour size was related to abundance of I-type tumour cells, whereas no such relationship was seen for tumour cells in deeper parts of the wall.
To the best of our knowledge, the relationship between mucin positivity and response to therapy has not been investigated.
## Comparison of immunohistochemical mucin stains between gastric signet-ring cell cancer and other histological tumour types
The relationship between immunohistochemical mucin expression and histological tumour type in GC is still controversial. Some studies reported that tubular/papillary/ glandular adenocarcinomas were more frequently MUC1, MUC2and/or MUC5ACpositive compared to SRC-GC. Whereas other studies found that SRC-GC were more often MUC2, TFF3, or MUC5ACpositive compared to other types of GC, Ilhan et al.found no relationship of MUC1 expression and histological tumour type. Zhangdid not find an association between MUC2, MUC5AC, and MUC6 expression and histological tumour type, whereas other studies showed that MUC2 expression was associated with mucinous cancers.
In conclusion, the existing number of studies on (immuno)histochemical mucin expression in SRC-GC is limited, and results are controversial most likely related to sample size, the use of different mucin stains, or combinations of stains and variable cut-offs. We noted differences in frequencies of mucin positivity in SRC-GCs as well as differences in the reported association with clinicopathological variables including patient outcome. The majority of published studies were performed using material from Asian GC patients; thus, it is not clear whether results in Caucasian patients would be similar.
All the above motivated us to complement the comprehensive literature review with a large cohort study on more than 1000 patients with gastric or gastro-oesophageal cancers.
## Results from our own gastric and gastro-oesophageal cancer cohort study
This study included material from 958 (76%) patients with GC and 303 (24%) patients with junctional/lower oesophageal adenocarcinoma. For clinicopathological and demographic cohort characteristics, see. This cohort included 851 Caucasian patients (LTHT cohort: 709 gastric and 142 junctional/lower oesophageal cancer) and 410 Asian patients (KCCH cohort: 249 gastric and 161 junctional/ lower oesophageal cancer).
## Frequency of mucin positivity
Using the presence of expression in more than 10% tumour cells as cut-off for all 3 stains (MUC2, MUC5AC, AB/PAS), 670 (53%) cancers were classified as 'triple negative'. 172 (14%) cancers were classified as MUC2 positive, 383 (31%) MUC5AC positive, and 253 (20%) AB and/or PAS positive (see. 145 AB and/or PAS positive cancers showed a mixed expression with both, the same and different cells being positive for AB and/or PAS. 330 (26%) cancers were positive for only one of the three mucin stains (48 (4%) only MUC2 positive, 217 (17%) only MUC5AC positive, and 65 (5%) only ABPAS positive). 176 (14%) cancers were positive for two mucin stains whereby combined positivity of MUC5AC and ABPAS was most frequently seen (94 (8%) cancers). 42 (3%) cancers were triple positive (see also Online Resource 2).
## Histological phenotypes and associations with mucin stains
259 (21%) cancers were classified as poorly cohesive, 943 (75%) as non-poorly cohesive, and 41 (3%) as mucinous. An overview of results from the non-poorly cohesive and mucinous cancers can be found inand. In the group of poorly cohesive cancers, 192 (74%), 36 (14%), 21 (8%), and 10 (4%) cancers contained < 10% SRCs, ≥ 10-50% SRCs, ≥ 50-90% SRCs, and ≥ 90% SRCs, respectively. Signet-ring cells were seen in mucinous cancers: < 10% SRCs (n = 14, 34%), ≥ 10-50% SRCs (n = 12, 29%), ≥ 50-90% SRCs (n = 9, 22%), and ≥ 90% SRCs (n = 6, 15%). As expected, the number of non-poorly cohesive cancers with cells looking like signet-ring cells was very low, only 38 (3.8%) cancers contained more than 10% lookalike SRCs, and none of the non-poorly cohesive cancers contained more than 50% lookalike SRCs.
Because of relatively small number of cases in subgroups, we compared poorly cohesive cancer with < 10% SRCs (n = 192, 15%), poorly cohesive cancer with ≥ 10% SRCs (n = 67, 5%), non-poorly cohesive cancer with < 10% SRCs (n = 905, 72%), non-poorly cohesive cancer with ≥ 10% SRCs (n = 38, 3%), mucinous cancer with < 10% SRCs (n = 14, 1%), and mucinous cancer with ≥ 10% SRCs (n = 27, 2%) (see also.
Both, poorly cohesive and non-poorly cohesive cancers with ≥ 10% SRCs, were more frequently MUC2, MUC5AC, or ABPAS positive compared to cancers with < 10% SRCs (see. MUC2 was highly expressed in all mucinous cancers irrespective of the percentage of SRCs. Mucinous cancers with < 10% SRCs were less often ABPAS positive compared to mucinous cancers with ≥ 10% SRCs.
## Comparison between the ltht cohort and the kcch cohort
The proportion of cancers with a particular histomorphological phenotype was similar in the LTHT and KCCH cohorts (see Online Resource 4). In both cohorts, non-poorly cohesive cancers were the most common histological subtype [LTHT n = 608 (71%); KCCH n = 297 (72%)]. MUC2-and MUC5AC-positive cancers were more frequent in the LTHT cohort (MUC2 LTHT n = 129 (15%) vs. KCCH n = 44 (11%), p = 0.043; MUC5AC LTHT n = 289 (34%) vs. KCCH n = 98 (24%), p < 0.001), whereas ABPAS-positive cancers were more frequent in the KCCH cohort (KCCH n = 96 (23%) vs. LTHT n = 159 (19%); p = 0.037) (see also.
AB positivity was associated with worse 5-year survival only in the LTHT cohort (p = 0.002) (see. PAS, MUC2, or MUC5AC positivity was not related to outcome in any of the cohorts. Presence of combined ABPAS positivity, MUC2 positivity, and MUC5AC negativity (n = 35, 4%) was related to poorer 5-year survival in the LTHT cohort (p = 0.002) (seeand Online . No other associations with patient outcome were found in either cohort.
# Discussion
There is an on-going debate whether patients with signetring cell (SRC) type gastric or gastro-oesophageal cancer have a different prognosis and response to chemotherapyand, therefore, should be treated differently to patients with other gastric cancer (GC) subtypes. The inconsistent results reported in the literature could be related to the fact that the histopathological classification of SRC-GC can be challenging. This is due to the variable morphological appearances of individual SRCswhich may or may not be recognised as SRCs by some investigators, as well as changing criteria in the WHO classification over the last decades. Whilst an expert panel recently For PC and non-PC cases with ≥ 10%, RCs showed more MUC2 (p < 0.001 and p = 0.003), MUC5AC (p = 0.004 and p < 0.001), and ABPAS (p < 0.001 and p < 0.001) positivity compared to cases with < 10% SRCs. MC showed high MUC2 expression. MC with ≥ 10% SRCs were more often ABPAS positive compared to MC with < 10% SRCs (p = 0.003). PC poorly cohesive cancer; non-PC non-poorly cohesive cancer; MC mucinous cancer; SRC signet-ring cell proposed cut-offs to enable separating pure SRC-GC (≥ 90% SRCs) from GC with an SRC component, a biomarker to unequivocal identify SRC-GC would be of potential great value to clinicians and patients.
In search of a potential promising biomarker, we decided to focus on mucin-related stains as mucin appeared to be a defining characteristic of SRCs. Our group was the first to conduct a comprehensive literature review into the reported frequency and clinical importance of different mucin stains in SRC-GC. To our surprise, the number of studies describing (immuno)histochemical mucin expression in SRC-GC was very limited and results varied between studies making firm conclusions difficult. However, the existing literature seemed to suggest that mucin is not only present in SRC-GC, but can also be seen in other types of GC. Furthermore, some SRC-GC appeared to contain no mucin. Similarly to the well described morphological heterogeneity of GC, we found evidence that combinations of different kinds of mucins can be seen in the same GC. Our literature review supports the previous suggestion of an expert panel that inconsistent clinicopathological findings can at least partly be explained by differences in the histological haematoxylin-eosin-based classification of SRC-GC together with different cut-off values used for considering a stain positive. Furthermore, even if the same cut-off value was used, results remained contradictorywhich could be related to the use of different primary antibodies. When comparing reported results from Asian and Caucasian SRC-GC, we saw similar wide ranges of mucin positivity, suggesting that differences might not be simply related to ethnic origin of the cancers. Sampling of the tumours (luminal versus centre versus invasive front) could potentially explain different results in mucin expression as it has been reported by several investigators that intramucosal SRC-GC showed a 'layered structure' both morphologically and (immuno) histochemically. Mucin characteristics of SRC-GC varied depending on tumour size and disease stage, thus results might vary depending on the case selection for the study.
In order to close the gap in the current literature revealed by our comprehensive review, we decided to conduct our own gastric and gastro-oesophageal cohort study into mucin stains. Our study is the largest study to date where all cancers were re-classified in a standardised manner according to WHO classification and recent consensus. Furthermore, our study is the first study to include both, Asian and Caucasian patients, enabling us to directly compare patient characteristics, histological tumour types, mucin expression, and relationship between mucin expression and patient outcome. Previous studies had either investigated Caucasian cohorts or Asian cohorts.
Our study provides further evidence that there is no (immuno)histochemical mucin stain unique to SRC cancer, since (1) a proportion of SRC containing GC was negative for the (immuno)histochemical mucin stains and (2) a relatively large percentage of GC without SRCs were positive for one or more mucin stains, similarly to what had been reported in the literature. However, this is the first study to suggest that the mucin expression might be related to the quantity of SRCs within a given tumour as we saw more frequently mucin expression in poorly cohesive GC containing ≥ 10% SRCs.
When directly comparing Caucasian (LTHT) and Asian (KCCH) GC cohorts, MUC2 and MUC5AC positivity was more frequent in the LTHT cases, whereas ABPAS positivity was more frequent in the KCCH cases. This is the first study to report a relationship between AB positivity and outcome in GC patients. Most interestingly, AB positivity alone or in combination with other mucin expressions was only related to poor outcome in Caucasian GC providing further support for the hypothesis that different outcome in Caucasian and Asian GC patients may be related to an underlying biological difference. As AB stains acidic mucins which are considered to be present in an intestinal phenotype, our finding would support the described association between (gastro) intestinal mucin phenotype and unfavourable outcome. Due to relatively low patient numbers in mucin-defined subgroups, we were unable to explore whether the AB expression related outcome difference between cohorts was related to a different disease stage mix.
Based on the literature review and the results of our cohort study, we had to reject our first working hypothesis and conclude that SRC containing GC do not have a different mucin expression compared to non-SRC-GC. Further studies are needed to address this clinical need. On the other hand, results reported in the literature and from our own cohort study confirmed a relationship between SRC mucin expression and patient outcome in the surgery alone setting. Furthermore, our cohort study suggests that irrespective of histological phenotype, the mucin expression is different in Caucasian and Asian GC patient and is associated differently with outcome in different ethnic groups.
We recommend further studies comparing Caucasian with Asian GC to validate our findings and explore underlying molecular mechanisms for difference in mucin expression and outcome. Also, we recommend investigating whether the different mucin expression in SRC containing GC is related to variable treatment response.
## 3
|
Assessment of the In Vitro and In Vivo Antifungal Activity of NSC319726 against Candida auris
Re: Spectrum01395-21 (Assessment of the in vitro and in vivo antifungal activity of NSC319726 against Candida auris) Dear Dr. Frederic Lamoth:Thank you for submitting your manuscript to Microbiology Spectrum.As you can see below, both reviewers were generally in favor of the manuscript and would like to see only minor modifications before it can be published. Both reviewers were in favor of moving the supplemental figure to the main text, and both would like to see more methodological details on some of the experiments.When submitting the revised version of your paper, please provide (1) point-by-point responses to the issues raised by the reviewers as file type "Response to Reviewers," not in your cover letter, and (2) a PDF file that indicates the changes from the original submission (by highlighting or underlining the changes) as file type "Marked Up Manuscript -For Review Only". Please use this link to submit your revised manuscript -we strongly recommend that you submit your paper within the next 60 days or reach out to me. Detailed information on submitting your revised paper are below.Link Not AvailableThank you for the privilege of reviewing your work. Below you will find instructions from the Microbiology Spectrum editorial office and comments generated during the review.
E-mail: [email protected] Reviewer comments:
Reviewer #1 (Comments for the Author):
This manuscript describes the antifungal activity of a investigational chemotherapeutic NSC319726 against Candida auris clinical isolates both in vitro and in vivo. Overall, while the antifungal activity of this agent has been described previously, as mentioned by the authors of this manuscript, these studies expand what is known to an emerging healthcare associated and multidrug resistant pathogen of great clinical concern. The manuscript is well written and the procedures are easy to follow and understand.
-In the methods the authors describe that MIC testing was performed per CLSI methods, however the description of the endpoint used for micafungin was complete growth inhibition rather than thẽ 50% inhibition as detailed by CLSI M27 S4. IF the authors did indeed use an alternative and more stringent endpoint for micafungin than is standard, this should be clearly mentioned and justification for the decision to do so detailed.
-The authors should include a description of the media, incubation time, and incubation temperatures used for MIC testing particularly if the differ from the CLSI M27 S4 as with the previous point. These parameters all can significantly impact MIC determination, particularly with Candida auris, and as the comparative MIC values are a key aspect of the manuscript they should be as clearly defined/ described as possible.
-The supplementary figure should be moved to the main text. The finding that this investigational agent is fungistatic against Candida auris but was previously shown to be fungicidal against Candida albicans is important and of interest.
-The arrows denoting the significance of comparisons between treatments in are a bit confusing as they are placed between the endpoint of the graph and the legend. These should be moved or changed to more clearly show which of the two they are corresponding to.
-In line 35-36, the authors describe the clades of C. auris to be geographic clades. While references like this are common, this is somewhat misleading. While the clades where each associated with a geographic region where they were first found, assignment of isolates to clades is done based up on genetic relatedness and the global dissemination of C. auris has resulted in isolates from various clades being identified in many countries (as seen in . It would seem simply changing the word "geographic" to "genetic" would be more appropriate.
-lines 82 through 86 are very redundant and should be revised/ reworded to address this.
Reviewer #2 (Comments for the Author):
The manuscript is well written and presents novel preliminary data on NSC319726 activity against Candida auris. Please see my detailed comments below.
1. It seems the tested drug concentration ranges were rather unusual -judging by fluconazole and micafungin MICs {greater than or equal to} 1024. Please clarify (in the Materials and Methods section) what drug concentration ranges were applied for each drug.
2. The time-kill curves should be performed in biological triplicates, and error bars presented on the graph. Moreover, it is not clear why the graph presenting time-kill curves was made a Supplementary Material and not a Figure in the main text.
3. Please provide some explanation why stains I.3 and IV.1 were selected for in vivo studies.
4. Please provide a comment (in the discussion) on the in vivo results in light of the discovery of fungistatic activity of NSC319726.
5. Lines 36-37: grammar; should be: South Asian, South African, South American.
## Staff comments:
## Preparing revision guidelines
To submit your modified manuscript, log onto the eJP submission site at https://spectrum.msubmit.net/cgi-bin/main.plex. Go to Author Tasks and click the appropriate manuscript title to begin the revision process. The information that you entered when you first submitted the paper will be displayed. Please update the information as necessary. Here are a few examples of required updates that authors must address:
- Point-by-point responses to the issues raised by the reviewers in a file named "Response to Reviewers," NOT IN YOUR COVER LETTER. - Upload a compare copy of the manuscript (without figures) as a "Marked-Up Manuscript" file. - Each figure must be uploaded as a separate file, and any multipanel figures must be assembled into one file. For complete guidelines on revision requirements, please see the journal Submission and Review Process requirements at https://journals.asm.org/journal/Spectrum/submission-review-process. Submissions of a paper t hat does not conform t o Microbiology Spect rum guidelines will delay accept ance of your manuscript . "
Please return the manuscript within 60 days; if you cannot complete the modification within this time period, please contact me. If you do not wish to modify the manuscript and prefer to submit it to another journal, please notify me of your decision immediately so that the manuscript may be formally withdrawn from consideration by Microbiology Spectrum.
If your manuscript is accepted for publication, you will be contacted separately about payment when the proofs are issued; please follow the instructions in that e-mail. Arrangements for payment must be made before your article is published. For a complete list of Publicat ion Fees, including supplemental material costs, please visit our website.
Corresponding authors may join or renew ASM membership to obtain discounts on publication fees. Need to upgrade your membership level? Please contact Customer Service at [email protected].
Thank you for submitting your paper to Microbiology Spectrum.
## Spectrum 01395-21 : responses to reviewers comments
Line numbering refers to the "clean" revised manuscript.
## Reviewer #1 (comments for the author):
-In the methods the authors describe that MIC testing was performed per CLSI methods, however the description of the endpoint used for micafungin was complete growth inhibition rather than the ~50% inhibition as detailed by CLSI M27 S4. IF the authors did indeed use an alternative and more stringent endpoint for micafungin than is standard, this should be clearly mentioned and justification for the decision to do so detailed.
Response: We agree that the CLSI M27 S4 document defines the cut-off at 50% inhibition for micafungin. Actually, we did not observe any trailing effect regarding micafungin (clear cut inhibition) for all isolates, which means that the MIC 50% inhibition and MIC 100% inhibition are the same. We agree that strict adherence to CLSI method is warranted and we have modified the MIC definitions for micafungin (50% instead of complete) in methods (lines 126-129), which actually does not affect the results presented in text and in .
-The authors should include a description of the media, incubation time, and incubation temperatures used for MIC testing particularly if the differ from the CLSI M27 S4 as with the previous point. These parameters all can significantly impact MIC determination, particularly with Candida auris, and as the comparative MIC values are a key aspect of the manuscript they should be as clearly defined/ described as possible.
Response: We have adhered to CLSI procedure. We have now added details about media, inocula, incubation time and temperature in methods (lines 118-129).
-The supplementary figure should be moved to the main text. The finding that this investigational agent is fungistatic against Candida auris but was previously shown to be fungicidal against Candida albicans is important and of interest.
Response: We agree. The paper was initially submitted as an "observation", which according to journal guidelines allow only 2 figures/tables. This is why we have moved the time-kill curve as Supplemental material. We agree that it makes perfect sense to move it as a figure in the main manuscript (now inserted as new .
-The arrows denoting the significance of comparisons between treatments in are a bit confusing as they are placed between the endpoint of the graph and the legend. These should be moved or changed to more clearly show which of the two they are corresponding to.
Response: We have moved the arrows in the graph to clearly show the curves that have been compared. We have also added more precisions in the legend.
-In line 35-36, the authors describe the clades of C. auris to be geographic clades. While references like this are common, this is somewhat misleading. While the clades where each associated with a geographic region where they were first found, assignment of isolates to clades is done based up on genetic relatedness and the global dissemination of C. auris has resulted in isolates from various clades being identified in many countries (as seen in . It would seem simply changing the word "geographic" to "genetic" would be more appropriate.
Response: We agree and we have changed the sentence as follows: five genetic clades have been identified from distinct geographical origin: South Asian (1), etc… (lines 36-38).
-lines 82 through 86 are very redundant and should be revised/ reworded to address this. Response: We agree. We have suppressed the first sentence.
Reviewer #2 (Comments for the Author):
1. It seems the tested drug concentration ranges were rather unusual -judging by fluconazole and micafungin MICs {greater than or equal to}1024. Please clarify (in the Materials and Methods section) what drug concentration ranges were applied for each drug.
Response: We have indeed tested some higher concentrations than those recommended in the range of the CLSI procedure. For a strict adherence to CLSI recommendations (as also suggested by Reviewer 1), we have now provided only the results within the CLSI ranges (see modifications in methods, lines 123-125, results lines 53-56, and . Thank you for resubmitting your revised manuscript. All of the reviewer's previous comments were addressed.
## The
Your manuscript has been accepted, and I am forwarding it to the ASM Journals Department for publication. You will be notified when your proofs are ready to be viewed.
The ASM Journals program strives for constant improvement in our submission and publication process. Please tell us how we can improve your experience by taking this quick Author Survey.
As an open-access publication, Spectrum receives no financial support from paid subscriptions and depends on authors' prompt payment of publication fees as soon as their articles are accepted. You will be contacted separately about payment when the proofs are issued; please follow the instructions in that e-mail. Arrangements for payment must be made before your article is published. For a complete list of Publicat ion Fees, including supplemental material costs, please visit our website. |
A novel alignment-free method for detection of lateral genetic transfer based on TF-IDF
Lateral genetic transfer (LGT) plays an important role in the evolution of microbes. Existing computational methods for detecting genomic regions of putative lateral origin scale poorly to large data. Here, we propose a novel method based on TF-IDF (Term Frequency-Inverse Document Frequency) statistics to detect not only regions of lateral origin, but also their origin and direction of transfer, in sets of hierarchically structured nucleotide or protein sequences. This approach is based on the frequency distributions of k-mers in the sequences. If a set of contiguous k-mers appears sufficiently more frequently in another phyletic group than in its own, we infer that they have been transferred from the first group to the second. We performed rigorous tests of TF-IDF using simulated and empirical datasets. With the simulated data, we tested our method under different parameter settings for sequence length, substitution rate between and within groups and post-LGT, deletion rate, length of transferred region and k size, and found that we can detect LGT events with high precision and recall. Our method performs better than an established method, ALFY, which has high recall but low precision. Our method is efficient, with runtime increasing approximately linearly with sequence length.Many microbes can acquire DNA from their environment and incorporate it into their genome via processes of lateral genetic transfer (LGT; also known as horizontal gene transfer, HGT) 1 . Circumstantial evidence for LGT was first reported more than a century ago 2 , and the phenomenon gained widespread attention in the 1950s with the emergence and spread of multi-drug resistance in bacteria 3 . With the uptake of genome sequencing over the last two decades, it has become increasingly clear that LGT plays a central role in the evolution of microbial genomes 1,4-6 .LGT not only contributes to the spread of antibiotic resistance, but is also responsible for a range of metabolic innovations involving carbon and nitrogen metabolism, ion transport and other core processes 7 , which in turn can define microbial physiology and thus ecosystem function.The recognised mechanisms of LGT (transformation, transduction and conjugation) can introduce exogenous regions of very different lengths, from short fragments to large chromosomal blocks 8 . Recombination need not be constrained by gene boundaries 9 , and there is little evidence to suggest that entire genes, or structurally based regions within genes, are privileged units of transfer 10,11 . In any event, genomic regions of lateral origin can be overwritten, wholly or in part, by subsequent LGT events. Thus microbial genomes can become mosaics, with regions of different lengths reflecting the history of LGT events, transfer mechanisms and donors in each lineage. Further, over time, sequence regions of lateral origin will evolve to become indistinguishable from the non-lateral background, a process known as amelioration 12 .This complex biology presents challenges for the detection and delineation of genomic regions of lateral origin. As typically applied, approaches based on the topological comparison of inferred phylogenetic trees implicitly take genes (gene families) as the unit of analysis. Extensions that test for recombination breakpoints are computationally intensive, yet fail to identify the specific lineage(s) affected by transfer and/or subsequent overwriting. Directionality of transfer can also be difficult or impossible to determine by any phylogenetic approach. More broadly, computational methods are differentially sensitive to the extent of amelioration 13,14 . Considerable scope thus remains for the development of new methods that are sensitive, directional, scalable, informative on individual genomes or lineages, and do not require the units of analysis to be delineated a priori.Alignment-free approaches to detect LGT at genome level have been developed in recent years. ALFY (ALignment-Free local homologY) 15,16 uses Kr 17 based on shustrings (SHortest Unique subSTRINGS) to calculate pairwise evolutionary distances between genomes, which can then serve as input into a neighbor-joining algorithm 18 to compute a phylogenetic tree. Then ALFY compares the generated tree with a reference, inferring topological incongruence as instances of LGT.Another alignment-free method for LGT detection is based on the so-called purity measure 19 . This is a concept from text mining, and is used to detect unusual regions of a string without recourse to domain knowledge. If most substrings of string x, which is itself a substring of string T, appear with the same frequency as x, then the purity value of x is high, i.e. subpatterns in x occur infrequently in T outside whole occurrences of x, as would be expected if x had arisen by LGT. Both of these alignment-free methods use suffix trees 20 for scalability on large sequence datasets. However, they consider only one target sequence (although ALFY incorporates a pairwise comparison between query and multiple subject sequences) and do not take into account any natural group structure of the dataset, whether taxonomic (a hierarchy of species, genera etc.), ecological or otherwise.In this paper, we propose a novel alignment-free method for LGT detection based on concepts from TF-IDF (Term Frequency-Inverse Document Frequency). TF-IDF is a numerical statistic from document analysis that reflects the importance of a word (term) to a document within a collection or corpus, by comparing the frequency of a word in a document with its occurrence in other documents.Term frequency (TF) is used to indicate the topic of a document 21 . The TF of term t in document D is simply the raw frequency of t in D, denoted by tf (t, D). The inverse document frequency (IDF) 22 is used to distinguish a word from the prevalent vocabulary in the corpus. If t appears in D t articles, then its IDF is idf(t) = D * /D t , where D * is the number of all documents in the corpus. Thus a high IDF indicates that the term appears infrequently, and as such carries more importance for a specific article. Salton and Buckley combined the TF and IDF statistics into a single statistic that is widely used as a weighting factor in text mining and information retrieval 22-24 .Here we apply concepts from TF-IDF to develop an algorithm to detect LGT events in microbial genomes. Using simulated datasets, we test this algorithm and compare its performance with ALFY on sets of sequences of different length, from the size of a single gene (1000 nucleotides) up to 300-fold longer, and evaluate its performance over k-mer length and a biologically relevant range of values for parameters including substitution rate between groups, within groups and post-LGT. We find that with appropriate parameter values, the algorithm performs with good precision and recall; furthermore, runtime increases approximately linearly with sequence length, and in most cases TF-IDF performs much better than ALFY 15 . We also apply this method to an empirical dataset composed of seven Staphylococcus aureus genomes, and recover putative regions of lateral origin that correspond to genes involved in transport, antibiotic resistance, pathogenicity and virulence. Our results are comparable with those found with ALFY, and include two genomic regions independently confirmed by Holden et al.25.
# Results
Performance with different parameter values. As described in Methods, we varied branch length at three stages of the simulation process (variation between groups, variation within groups, and variation post-LGT) and examined the effect on precision and recall. The results are shown in Figs 1-4 for simulations under the HYK85 26 model of sequence change; the corresponding plots for F84 [bib_ref] A hidden Markov model approach to variation among sites in rate of..., Felsenstein [/bib_ref] are in the Supplementary file. Since TF-IDF does not detect LGT between sequences within a group, for the comparison we ignore such regions that are detected by ALFY; and if an atypical region is equally predicted in several sequences of potential donor groups, we treat this result as a single prediction for the calculation of precision and recall. [fig_ref] Figure 1: Performance of TF-IDF with variation between groups [/fig_ref] shows that when variation between groups is less than 0.05, the average distance accumulated between groups is less than 15%; at this degree of between-group similarity, the precision of our TF-IDF method is low (less than 50%) because the high similarity makes lateral regions harder to distinguish in the recipient group. Precision increases to a high level when variation between groups is above 0.1. Recall is high throughout (approximately 90%) and is less affected by variation; however, at the shortest sequence length examined here (1000), some simulated LGT segments are less than 50 nt in length, too short to contain enough information to make them distinct. As a consequence, recall is significantly lower for this sequence length only.
The precision of ALFY is low, around 0.35, and stable across all branch lengths, but its recall is high. There are two reasons for this. Firstly, ALFY cannot infer the direction of transfer, and may correctly predict one transfer from donor to recipient, but then (erroneously) predict it again from recipient to donor, effectively halving its precision. In the accompanying article [bib_ref] Exploring lateral genetic transfer among microbial genomes using TF-IDF, Cong [/bib_ref] we compare TF-IDF with another directional LGT inference approach [bib_ref] Directed networks reveal genomic barriers and DNA repair bypasses to lateral gene..., Popa [/bib_ref] applied to genome-scale empirical data. Secondly, ALFY predicts all most-similar regions as lateral transfers without using a threshold to determine if the similarity is significant or not. As such, it is apparent that ALFY is a useful tool for determining areas which should be further studied for transferred segments, but as a stand-alone detector of LGT it is inferior to TF-IDF. For sequences of length 1000 nt, ALFY's default sliding window size is too large, leading to reduced performance. [fig_ref] Figure 2: Performance of TF-IDF with variation within groups [/fig_ref] shows the effect of variation within groups on precision and recall. Here, the precision of TF-IDF increases as variation increases. As above, the sequences must be sufficiently dissimilar for the TF statistic to support a decision of LGT. Recall is high, and stable when the sequence length is ≥ 3000 nt. Again, at sequence length 1000, some short LGT events (< 50 nt) are ignored, resulting in decreased recall. The precision of ALFY is stable for variation above 0.005, but again low. TF-IDF shows greater stability and better performance than ALFY in almost all cases, and increasingly outperforms it as the variation increases. As in [fig_ref] Figure 1: Performance of TF-IDF with variation between groups [/fig_ref] , ALFY displays better recall than TF-IDF at sequence lengths greater than 1000 nt, but the gap is not large. When the variation within groups is low and the sequence length is short (1000 nt), ALFY again fails to detect most LGT events, leading to extremely low recall (see Supplementary file). [fig_ref] Figure 3: Performance of TF-IDF with variation post-LGT and deletion [/fig_ref] shows the performance of TF-IDF against variation post-LGT and deletion rate for sequences of length 300,000 nt. Plots for other sequence lengths are similar in nature and can be found in the Supplementary file. As variation increases, both precision and (especially) recall decrease substantially, as substitutions progressively obscure the regions of lateral origin. When the branch length post-LGT reaches 0.05 (i.e. one nucleotide in ten is expected to have changed, as this is a two-level tree), almost all k-mers (for k = 40) have been changed, whether in lateral regions or not. In this case, all k-mer based methods, including TF-IDF, will fail (and indeed, even alignment-based methods will struggle).
As the amount of deletion increases, precision remains stable and recall decreases slightly. Deletion can move an LGT segment within a sequence, or delete part (or parts) of it. Moving an LGT region does not change its k-mers, so this will not affect the performance of TF-IDF. Deletions within a lateral region affect only the immediately adjacent k-mers, with little effect on precision unless the region becomes so fragmented that k-mer counts are reduced to the point where they are ignored by TF-IDF, degrading the recall.
Precision and recall increase slightly with sequence length, but length does not appear to interact substantially with the substitution-rate parameters. Since there is no interaction between variation post-LGT and deletion [fig_ref] Figure 3: Performance of TF-IDF with variation post-LGT and deletion [/fig_ref] , we can fix one of these parameters and vary the other. [fig_ref] Figure 4: Performance of ALFY with variation post-LGT [/fig_ref] shows that for TF-IDF and ALFY, both precision and recall decrease as variation post-LGT increases. The precision of ALFY is worse than that of TF-IDF, but its recall is higher and more stable. When deletion is varied [fig_ref] Figure 5: Performance of ALFY with deletion [/fig_ref] , precision is stable except at sequence length 1000, while recall decreases slightly for TF-IDF. As before, TF-IDF is more precise than ALFY, whereas ALFY exhibits higher recall (except at sequence length 1000). k-mer size. k-mer size also affects the performance of TF-IDF. As shown in [fig_ref] Figure 6: Performance of TF-IDF with k-mer size [/fig_ref] , precision increases with k, but recall decreases. This effect is roughly consistent for every sequence length we examined. The two plots indicate that in this simulation, precision and recall are best balanced at k = 40. Indeed, in our experience (as shown and unpublished) k = 40 is a useful default setting, in the absence of conditions that argue otherwise. However, if LGT is sufficiently obscured by substitution such that nearly all k-mers are unique, TF-IDF will not be able to find sets of k-mers that appear frequently in distant groups, and no LGT will be predicted. In such cases, shorter k may give better performance. Note that larger k imposes a greater memory cost, and more computational time is spent indexing unique k-mers.
Computation time. [fig_ref] Figure 7: Log-log plot of sequence length against walltime [/fig_ref] compares computation time (walltime) for various sequence lengths L for ALFY and TF-IDF. All experiments were done on a virtual machine with a single AMD Opteron 2.3-GHz processor and 256 GB memory. As noted below, TF-IDF is expected to scale as O(nL log U), where U is the number of unique k-mers in the dataset. U is highly dependent on variation at all levels of the simulation, which also leads to variation of time consumption in each experiment; if the final sequences are sufficiently dissimilar, we expect U to LGTs between all sequences in a dataset. In a dataset with tens of sequences or more, ALFY will take much longer than TF-IDF, and this is shown in [fig_ref] Figure 7: Log-log plot of sequence length against walltime [/fig_ref]. [fig_ref] Figure 8: Log U against time divided by sequence length [/fig_ref] shows how walltime depends on U. As above, we expect time divided by L to have a linear relationship with log U, and this is clearly shown.
Analysis of an empirical dataset. We also tested our algorithm on an empirical dataset that had previously been examined by the developers of ALFY [bib_ref] Alignment-free detection of horizontal gene transfer between closely related bacterial genomes, Domazet-Lošo [/bib_ref]. We used a subset of their dataset, seven genomes of Staphylococcus aureus, because this dataset contains strong group information (six genomes from Clonal Complex 8 (CC8) and one multi-drug resistant strain from CC30, S. aureus MRSA252) and showed LGT in their analysis. We investigate potential LGT into S. aureus TW20, a member of CC8, from MRSA252.
Setting k = 40, we identify 1421 regions of TW20 as of lateral origin. Many of these are short and, in this simple example (where the donor group is of size 1, reducing the efficacy of the IDF component) potentially due to noise; but 173 are of length ≥ 2000, 52 of ≥ 4000 and 20 of ≥ 6000 nt . It is unclear how to optimise selection of the length threshold, but setting it at ≥ 2000 nt we infer as lateral 35.6% of the genome, which incorporates 67% (4/6) of the TW20 penicillin-binding genes, and ≥ 50% (i.e. > 1.5-fold over-representation) of the annotated genes encoding efflux proteins (2/4), metalloproteinases and -peptidases (3/3), permeases (31/45) and uptake proteins (2/4), types of functions known to be mobilised by LGT [bib_ref] Lateral genetic transfer: open issues, Ragan [/bib_ref] [bib_ref] Horizontal gene transfer among genomes: The complexity hypothesis, Jain [/bib_ref]. For details see . By contrast, hypothetical proteins, which might be expected to show no bias for or against lateral origin, are not enriched at any of the length thresholds mentioned above. Ribosomal proteins, which are not expected to be lateral (Jain et al. [bib_ref] Horizontal gene transfer among genomes: The complexity hypothesis, Jain [/bib_ref] , are rarely represented in our lateral regions (8/60). Phage proteins are not represented in our detected lateral regions; recalling that our approach can discover LGT only within the dataset, these results might accurately reflect the history of genetic relationships among these seven genomes. Scope remains for further analysis with other empirical data, and with different settings for k and gap size.
Both our TF-IDF method and ALFY identify most of the genomic region from 2.80-0.42 Mb (TF-IDF) or 2.8-0.5 Mb (ALFY) as lateral [fig_ref] Figure 9: Comparison of TF-IDF and ALFY with an empirical dataset [/fig_ref] ; this region includes two transposons, SCC elements and genes encoding methicillin and penicillin resistance. Robinson and Enright 31 hypothesised that the methicillin resistance, at least, had been transferred from CC30 into a CC8 background as part of a large chromosomal replacement. The region from 1.75-1.80 Mb includes the transposon Tn554 25 , which encodes resistance to erythromycin and spectinomycin. A region from 2.11-2.15 Mb incorporating a number of annotated phage genes was likewise identified. Regions identified as lateral by TF-IDF but not by ALFY include 1.06-1.17 Mb (transport protein genes) and 2.64-2.65 Mb (a transporter and a member of the TetR family of regulatory proteins, which control the expression of genes involved in multidrug resistance and pathogenicity).
# Discussion and conclusion
We have developed a fast alignment-free method to infer LGT events. Our method is based on TF-IDF, one of the most important methods used in information retrieval. TF-IDF has been widely applied in search engines, document classification and related applications including relevance decision-making. Here we apply TF-IDF to sequence analysis for the first time, treating a sequence or genome as an article and each k-mer as a word. Using simulated datasets, we show that TF-IDF can effectively find LGT events with good precision and recall, outperforming ALFY in most biologically realistic situations. We also analyse an empirical dataset and show that TF-IDF finds essentially all regions identified by ALFY as of lateral origin. TF-IDF further detects other regions that, based on annotated gene content, may also have arisen via LGT. Our method is alignment-free and scales very well in both length and number of sequences, i.e. to many entire genomes. It is worth noting that in each simulated dataset, all sequences share the same length and group size. For the empirical dataset, the group sizes and lengths of the seven S. aureus are of the same magnitude. For this reason, we did not normalise the count of k-mers in the IDF step. However, in other empirical datasets the sequence length and group size may vary greatly, and normalisation might be considered [bib_ref] Exploring lateral genetic transfer among microbial genomes using TF-IDF, Cong [/bib_ref].
Our method is purely data-driven, its performance relying strongly on sequence and group information in the dataset. In our simulations, when sequences are relatively similar within-group (variation 0.005-0.02) and relatively dissimilar between-group (variation > 0.1), group boundaries are clear, and the precision and recall of our algorithm is high. When speciation is modest (< 0.05), within-group divergence high (< 0.1) or LGT events obscured by subsequent evolution (> 0.02), TF-IDF loses precision in inferring LGT events.
In the accompanying article [bib_ref] Exploring lateral genetic transfer among microbial genomes using TF-IDF, Cong [/bib_ref] we apply this method to larger empirical datasets. TF-IDF could further be applied to environmental data, e.g. to study the flow of genetic material in communities and across the biosphere. We anticipate that significant scope remains for further algorithmic and implementational improvements. LGT, but is unaffected by deletion. Recall (B) decreases greatly with variation post-LGT and slightly with deletion. Variation between groups is 0.1, and variation within groups is 0.01. Sequence length is 300,000 nt.
Scientific RepoRts | 6:30308 | DOI: 10.1038/srep30308 Methods Notation. Here we establish some notation. We start with a dataset of n sequences, each of length L. For empirical datasets (and for some approaches to simulation) the length may vary among sequences; in those cases we use L to denote the average length. The sequences in the dataset are divided into m groups corresponding to closely related genomes (e.g. belonging to the same clonal group, species or genus). We denote each sequence as S i,j , where i = 1, 2, … , n is the number of the sequence in the dataset and j = 1, 2, … , m is the number of the group to which the sequence belongs. The number of sequences in group j is denoted by h j .
Our method proceeds by comparing substrings (words) of a fixed length k, called k-mers. We encode each sequence as a frequency vector of k-mers, counting only those k-mers that actually appear in the sequence, and denoting the number of unique k-mers appearing in the dataset by U. In general, U is much smaller than 4 k , the total number of all possible k-mers.
Although we illustrate our approach here using nucleotide sequences, the method is easily adapted for amino acids, requiring only a change of alphabet.
TF-IDF on texts. As mentioned above, TF-IDF was introduced to indicate the topic of a document, and distinguish that document from others in the same corpus for a specific query. The classical usage of TF-IDF is as a smart retrieval system and for automatic document categorisation. A variant uses prototype vectors to calculate relevance between documents with a nearest-neighbor learning method [bib_ref] Developments in automatic text retrieval, Salton [/bib_ref]. PrTFIDF 36 is an improved version of TF-IDF founded on a probabilistic model for text categorization, and there are other variants for calculating TF-IDF [bib_ref] Exploring the similarity space, Zobel [/bib_ref]. In recent years, TF-IDF has also been applied in other areas including decision-making and sentiment analysis [bib_ref] Interpreting TF-IDF term weights as making relevance decisions, Wu [/bib_ref].
TF-IDF is widely used in text mining and information retrieval because it allows the identification of terms that are characteristic of (and hence important for) one text or a set of texts. It is not sufficient for a term to be frequent in a text (TF); it must also be rare in other texts in the corpus (IDF). Importantly, IDF depends only on the occurrence of terms, not on their numerical frequencies. Drawing on analysis of documents in three independent domains, Salton and Yang 39 identified five situations relevant to the performance of TF-IDF:
1. Terms that appear frequently across a corpus contribute little to performance because they do not discriminate between relevant and non-relevant documents; 2. Terms that appear in a moderate number of texts and show somewhat skewed distributions provide good retrieval performance; 3. Terms with sharply skewed distribution occurring in very few documents are important only for those documents; 4. Rare terms are important for the few queries and documents in which they occur; and 5. Terms of low or moderate frequency, but with a flat distribution across documents, are similarly useful for the documents in which they occur.
Classically, the frequencies of terms in a corpus follow a power law (Zipf law), in which case TF-IDF performs well. However, TF-IDF can perform adequately even when this is not the case: TF-IDF requires only that terms relevant to the query are distributed intensively in a subset of documents within the corpus [bib_ref] Term-weighting approaches in automatic text retrieval, Salton [/bib_ref] [bib_ref] A theory of term importance in automatic text analysis, Salton [/bib_ref] ; this might include the query terms themselves (e.g. happy), or related terms in the corpus (pleased, delighted). TF-IDF on sequences. Molecular sequences have long been analogised with natural language 41 or treated as texts. Alternatively, both molecular sequences and texts have been subsumed within a broader class of objects [bib_ref] An overview of sequence comparison -time warps, string edits, and macromolecules, Kruskal [/bib_ref]. The analogy is not precise: in sequences, "terms" must be recognized computationally, e.g. by extracting k-mers. Fast approaches exist for extracting k-mers [bib_ref] A fast, lock-free approach for efficient parallel counting of occurrences of k-mers, Marçais [/bib_ref] [bib_ref] Blue: correcting sequencing errors using consensus and context, Greenfield [/bib_ref] , and k-mer distribution in empirical sequences has been studied at some length [bib_ref] Genomic DNA k-mer spectra: models and modalities, Chor [/bib_ref] [bib_ref] The distribution of word matches between Markovian sequences with periodic boundary conditions, Burden [/bib_ref] [bib_ref] A new method to compute K-mer frequencies and its application to annotate..., Kurtz [/bib_ref]. Like words in text, short k-mers (k between three and eight) in DNA sequences show Zipf-like scaling [bib_ref] Linguistic features of noncoding DNA-sequences, Mantegna [/bib_ref] , although this is not sufficient to confirm DNA sequences as a natural language [bib_ref] Is DNA a language?, Tsonis [/bib_ref].
Although there is dispute whether DNA is a language or not, some methods in text mining have been successfully applied to DNA analysis. For example, the first (to our knowledge) software to identify lateral transfer in biological datasets 51 was repurposed from the analysis of textual contamination in manuscripts, which in turn was built on software for phylogenetic inference from DNA sequences 52 (PHYLIP).
Sequences (genomes, genes, proteins) do, however, differ from texts in some properties. For example, k-mer frequency distributions in sequences are usually much flatter than term frequencies in texts. Experience from text mining indicates that this is not critical, but this remains to be explored and is in fact a goal of the current work. In the specific application here, genomic regions of lateral origin are expected to have k-mers that appear frequently in genomes of the donor taxon, but rarely in the host. This is analogous to conditions 2 and/or 4 above 39 .
Our algorithm works by comparing the frequencies of identified k-mers in a group of sequences (our TF) with their frequencies in other groups (our IDF). If a k-mer in one sequence is prevalent in a different group but not in its own, then it may have arisen by LGT from the group in which it is prevalent, and the direction of the transfer should be from that (donor) group to the recipient sequence. We compare these TF and IDF statistics to appropriate thresholds to optimize detection performance, i.e. to balance precision and recall.
Our algorithm consists of four steps: extracting all k-mers from genomes within one dataset calculating inverse document frequencies, constructing potential LGT segments, and calculating term frequencies. For pseudocode, see the Supplementary file. Extracting k-mers. To extract k-mers we scanned all the genomes, incrementing one nucleotide at one time.
If the genome length is L, then L-k + 1 k-mers are found. Unique k-mers were indexed in a red-black tree 53 for further searching.
Calculating IDF. To calculate the inverse document frequency, we construct an n × m relationship matrix R, denoting the frequency (number of occurrences) at which k-mers in each sequence appear in each group. Each row in R corresponds to a sequence, and each column corresponds to a group. Suppose sequence i consists of k-mers w i,1 , w i,2 , … , w i,L-k+1 . If the word w appears in group j with frequency f j (w), then the entries of the relationship matrix are The entries in R are our IDF values. The larger the R ij , the more likely that sequence i contains a region transferred laterally from group j. Note that this is in contrast to the original definition of IDF, where a higher IDF indicates that the word appears less frequently in other documents.
[formula] ∑ = . = − + R f w ( ) (1) [/formula]
To detect potential lateral-transfer events, we compare the IDF values against a threshold t. This threshold is the average value of all entries in R: IDF values that are above the average are used for further analysis.
[formula] ∑∑ = . = = t nm R 1 (2) i n j m ij [/formula]
Constructing potential LGT segments. We then mark potential lateral segments in each sequence. For each sequence i and group j with a sufficiently high IDF value, we examine each k-mer in sequence i to see if it appears in group j. Then we join all consecutive k-mers which do, forming potential lateral segments. Because mutations or other genomic events may disrupt the perfect matching, we allow gaps between blocks of k-mers of size up to a threshold G. Here we set G = 2k, a value at which the total number of detections is not greatly affected in real application [bib_ref] Exploring lateral genetic transfer among microbial genomes using TF-IDF, Cong [/bib_ref]. We then assess the significance of these potential lateral segments using term frequency. LGT only from outside the target group, so no region is in grey. Both plots were generated from analysis of the seven S. aureus genome dataset.
Scientific RepoRts | 6:30308 | DOI: 10.1038/srep30308
Calculating TF. For each potential lateral segment σ in a sequence, we calculate the frequency (number of occurrences) at which each of its component k-mers appears in sequences of its own group, say j. Our TF statistic for σ is the sum of these:
[formula] ∑ δ = . σ σ ∈ f w ( ) (3) w j [/formula]
If δ σ is higher than some threshold, then σ occurs frequently in its own group, and as such is considered not to be the consequence of a lateral event; otherwise it is considered to be of lateral origin.
To set the threshold, we calculate the average frequency of all unique k-mers in the recipient group j, denoted by τ j . Then we compare δ σ to lτ j , where l is the number of k-mers contained in the segment. If δ σ is smaller, we consider σ to have been transferred laterally from the other group into this sequence. Other approaches to setting the threshold are possible, but we do not consider them here.
Note that our method considers lateral transfers only within the dataset; like most other LGT methods, it is silent on potential transfers from sources external to the dataset. In addition, it can detect transfers only between groups, not between sequences in the same group.
Runtime analysis. The computational complexity of the algorithm is dominated by extraction of the unique k-mers in the dataset. To find these, we scan each of the n sequences of length L. As each unique k-mer is found it is added to a library, which is stored in a red-black tree [bib_ref] A dichromatic framework for, Guibas [/bib_ref]. A red-black tree is an approximately balanced tree, which guarantees that searching and insertion are efficient. On average, this step takes O(nL log U) time, where U is the number of unique k-mers stored in the tree. The frequency of each k-mer is also computed at this time. The remaining calculations are much quicker because most of the frequency (f) terms are zero. Thus for biological sequences of standard complexity, runtime increases about log-linearly with sequence length. Note that the k-mer profiles of each sequence could in principle be stored and retrieved for future use.
## Implementation.
We have implemented this algorithm in C+ + . The program can be compiled using GCC 4.8.2 and run on Unix, Unix-like and Windows platforms. We use the map template from STL (Standard Template Library) to index all distinct k-mers in a dataset. The inner implementation of map is a red-black tree [bib_ref] A dichromatic framework for, Guibas [/bib_ref].
Comparisons with ALFY. ALFY finds putative homology (shared DNA segments) between pairs of sequences by matching shustrings (shortest unique substrings). If a match is found with a region in an otherwise distant sequence, it will be judged as a potential lateral transfer. This method shows high efficiency and effectiveness for LGT detection [bib_ref] Alignment-free detection of horizontal gene transfer between closely related bacterial genomes, Domazet-Lošo [/bib_ref] [bib_ref] Alignment-free detection of local similarity among viral and bacterial genomes, Domazet-Lošo [/bib_ref] , so we use it to benchmark our method.
The inputs to both TF-IDF and ALFY are sequences. For TF-IDF the group information is compulsory, while ALFY requires a query sequence and subject sequences. Both TF-IDF and ALFY can process DNA sequences; TF-IDF can also process amino-acid sequences, but ALFY does not currently implement evolutionary models of amino-acid change. Only k-mer frequencies will be taken into consideration for calculating the value of TF-IDF.
In TF-IDF, if a k-mer has low a frequency in its own group but high frequencies in other groups, then this k-mer will be judged atypical. A set of contiguous atypical k-mers will be inferred as lateral, with the direction of the transfer from the k-mer prevalent group. In contrast, ALFY computes the average shustring length between segments of only two sequences at a time. The longer the average shustring, the closer the two segments; and if the sequences themselves are otherwise distant in the reference tree, the segment in question will be inferred as lateral, without any implication of which sequence was donor or recipient.
If the sequences are grouped such that each group is compact and boundaries between groups are clear, then TF-IDF should find lateral segments easily. ALFY does not use group information, so grouping does not affect its performance.
The computational complexity of TF-IDF is O(nL log nL), where n is total number of sequences in a dataset, and L the average length of sequences in a dataset. The computational complexity of the ALFY algorithm is O(nL). However, TF-IDF will process all sequences and infer all potential lateral regions over an entire dataset, whereas ALFY makes all pairwise comparisons between a single query sequence and the others. For fairness of comparison, all sequences in a dataset should be set as queries to find all LGTs in a dataset, in which case the complexity of ALFY increases to O(n 2 L), which in practice is much slower than TF-IDF.
## Simulation of datasets.
In order to test the performance of TF-IDF in different situations, and to compare with ALFY, we simulated datasets under the HYK85 26 and F84 27 evolutionary models. Our simulation process is as follows:
(1) We start with one random sequence, which will become the ancestor of all sequences in the dataset. We vary the length L of this sequence from 1000 to 300000 characters to simulate sequences from a single gene to a significant part of a genome (but our algorithm can be applied to sequences of any length). (2) To establish phyletic groups (i.e. to simulate speciation), the ancestral sequence is allowed to evolve along a balanced binary tree with four levels of equal branch lengths, using the evolutionary model. The branch length varies from 0.01 to 0.20 (substitutions per site) in steps of 0.05. We refer to this parameter as variation between groups. (3) To populate these groups with sequences, each descendant (leaf) in the initial tree (above) is allowed to evolve along another phylogenetic tree under the same evolution model. Again we use a balanced binary tree with four levels of equal branch length, which vary from 0.001 to 0.020 in steps of 0.005. We refer to this parameter as variation within groups. (4) We then simulate LGT events between groups. For the sake of simplicity, here we make transfers only into sequences in Group 1. We fix the number of LGT events at 20, with lengths normally distributed around mean 0.1 of the sequence length, and standard deviation half that amount. For each simulated event the recipient Scientific RepoRts | 6:30308 | DOI: 10.1038/srep30308 sequence (in Group 1) is selected at random, with (typically) several sequences receiving multiple transfers and others receiving none. Transfer events overwrite the equivalent positions in the recipient sequence, but (to simplify our simulation) cannot themselves be subsequently overwritten. Five of the 20 LGT events are simulated to come from the group (of 16 sequences) arising from the most-recent common ancestor on the binary tree (from Step 2), five from descendants of the second-most recent ancestor (32 sequences), five from the third (64 sequences) and five from the deepest bifurcation (128 sequences). Thus the probability of transfer decreases with increasing distance (on the tree) between donor and target. (5) In a final evolutionary process, we further evolve each of the 256 sequences along a balanced two-level tree, with branch lengths varying from 0 to 0.1 in steps of 0.025. We refer to this parameter as variation post-LGT. [bib_ref] Highways of gene sharing in prokaryotes, Beiko [/bib_ref] In some simulations, Step 5 also includes a stochastic process (implemented by using a shell script to call ALF [bib_ref] ALF-a simulation framework for genome evolution, Dalquen [/bib_ref] , not to be confused with ALFY) which deletes from 0 to 10% of a sequence. The proportion was varied using the deletion rate setting in ALF, while keeping deletion length distribution at its default value. We refer to this parameter as deletion. We did not simulate duplications here because bacterial genomes contain very few repetitive components. Step 1: The simulation starts with a single ancestor and generates 16 sequences, which serve as ancestors for each group (variation between groups).
Step 2: Within each group we generate 16 descendants (variation within groups), then add LGT events between these groups.
Step 3: Finally we simulate variation post-LGT, which may include deletion. From each initial ancestor the simulation generates 256 sequences. Symbols: DNA sequences which are ancestors of the sequence groups. Phylogenetic tree used to generate populations of each group. DNA sequences that constitute groups.
LGTs events are added between them.
LGT between two sequences.
Phylogenetic tree on which the evolutionary process post-LGT is simulated. This process tends to obscure the LGT events. Branch length determines the 'age' of the LGT events. Regions of the sequences may be deleted at this step. DNA sequences generated by the simulation.
## Figure 11. an example of simulated and inferred
LGTs. The x-axis displays the nucleotide position, and the yaxis the sixteen sequences generated in our first (recipient) group. The wide bars show the lateral regions actually simulated, and the narrow black bars the regions inferred as lateral by TF-IDF. Here, variation between groups is 0.1, variation within groups is 0.001, variation post-LGT is 0.01, deletion is zero, k = 40 and sequence length is 1000 nt.
Scientific RepoRts | 6:30308 | DOI: 10.1038/srep30308
After the above steps, we select one descendant of each tree to yield our final dataset (256 sequences per simulation).
In addition to varying the parameters mentioned above for both TF-IDF and ALFY, for TF-IDF only we also varied the word length k, in steps of 10 from 20 to 50. As the number of possible parameter combinations above is very large, at Step 2 we varied only the variation between groups parameter while keeping all others fixed at minimal-impact settings. Similarly at Step 3 we varied only the variation within groups parameter. For each parameter combination we simulated 50 datasets under the F84 model of sequence change, and 50 under HYK85. This process is depicted in [fig_ref] Figure 1: Performance of TF-IDF with variation between groups [/fig_ref] , and is explained in greater detail in the Supplementary file. We also analysed smaller datasets omitting Step 4, to examine whether TF-IDF inferred LGT when none was present; no segments met the IDF (k-mers frequent in donor groups) and TF (k-mers infrequent in the recipient group) criteria simultaneously, so no LGT was inferred.
Performance measures. We assessed the performance of our algorithm on simulated data using two measures. Precision is the proportion of inferred LGT events which are real (i.e. were actually simulated):
[formula] = + Precision tp tp fp /( )(4) [/formula]
where tp and fp are the total lengths of all true and false positives respectively. Recall is the proportion of true (simulated) LGTs which were inferred by the algorithm:
[formula] = + Recall tp tp fn /( )(5) [/formula]
where fn is the total length of false negatives (simulated LGTs which were not found). [fig_ref] Figure 1: Performance of TF-IDF with variation between groups [/fig_ref] illustrates the output of TF-IDF analysis of a simulated dataset, showing the 20 regions of (simulated) lateral origin of which 19 were detected (wholly or in part) by TF-IDF. Positions 797-877 of Sequence 11 represent a false positive inference of LGT, and positions 58-117 of Sequence 2 a false negative. Overall for this dataset (i.e.
LGT from Groups 2-16 into Group 1), precision was 0.82 and recall 0.95. Complete details (start and end coordinates) are presented in the Supplementary file.
[fig] Figure 1: Performance of TF-IDF with variation between groups. Precision (A) increases with variation between groups. Recall (B) is not substantially affected by variation between groups. Variation within groups is 0.01, variation post-LGT is zero, and deletion is zero. Error bars are 2× standard error.Scientific RepoRts | 6:30308 | DOI: 10.1038/srep30308 increase as the number of possible k-mers in the dataset, i.e. as nL. Thus, we expect the time to have an O(L log L) dependence on L, and this is verified inFig. 7; the slope of the fitted line is 1.07. For ALFY, the time consumption is O(n 2 L) for detecting [/fig]
[fig] Figure 2: Performance of TF-IDF with variation within groups. Precision (A) increases with variation within groups, while recall (B) is essentially unchanged. Variation between groups is 0.1, variation post-LGT is zero, and deletion is zero. Error bars are 2× standard error.Scientific RepoRts | 6:30308 | DOI: 10.1038/srep30308 [/fig]
[fig] Figure 3: Performance of TF-IDF with variation post-LGT and deletion. Precision (A) decreases with variation post- [/fig]
[fig] Figure 4: Performance of ALFY with variation post-LGT. Precision (A) and recall (B) decrease with variation post-LGT. Variation between groups is 0.1, variation within groups is 0.01, and deletion is zero. Error bars are 2× standard error.Scientific RepoRts | 6:30308 | DOI: 10.1038/srep30308 [/fig]
[fig] Figure 5: Performance of ALFY with deletion. Precision (A) is stable with deletion. Recall (B) decreases with deletion. Variation between groups is 0.1, variation within groups is 0.01, and variation post-LGT is 0.01. Error bars are 2× standard error.Scientific RepoRts | 6:30308 | DOI: 10.1038/srep30308 [/fig]
[fig] Figure 6: Performance of TF-IDF with k-mer size. Precision (A) increases with k, while recall (B) decreases with k. Error bars are 2× standard error. [/fig]
[fig] Figure 7: Log-log plot of sequence length against walltime. See text for details. Scientific RepoRts | 6:30308 | DOI: 10.1038/srep30308 [/fig]
[fig] Figure 8: Log U against time divided by sequence length. The slope of the regression line is 0.0002, and the grey area is the 95% confidence interval. [/fig]
[fig] Figure 9: Comparison of TF-IDF and ALFY with an empirical dataset. Both A and B represent the genome of Staphylococcus aureus TW20. A shows the result of ALFY analysis 15 ; regions inferred to have been transferred from MRSA252 are represented in black, while regions homologous between TW20 and USA300.TCH15156 are shown in grey. B shows the result of TF-IDF analysis. TF-IDF can infer [/fig]
|
Comparison of the effects of twice-daily exenatide and insulin on carotid intima-media thickness in type 2 diabetes mellitus patients: a 52-week randomized, open-label, controlled trial
Background: Exenatide, a glucagon like peptide 1 analog, has been suggested to reduce the cardiovascular disease risk factors, such as body weight, blood pressure and subclinical atherosclerosis in patients with type 2 diabetes mellitus (T2DM). This was the first randomized, open-label, controlled trial to compare the effects of exenatide versus insulin on subclinical atherosclerosis, as assessed by carotid-intima media thickness (CIMT), in patients with T2DM.Methods: A total of 66 patients with T2DM admitted from March 10, 2015 to June 20, 2017 in the Department of Endocrinology, Beijing Hospital were randomized to receive twice-daily exenatide or aspartate 70/30 insulin for 52 weeks. The primary endpoint was change from baseline in CIMT, and secondary endpoints included changes at week 52 from baseline in body weight, glycemic markers, lipid metabolism markers, blood pressure, C-reactive protein, fibrinogen, 8-hydroxydeoxyguanosine, irisin, and brain natriuretic peptide.Results: Exenatide more significantly reduced the CIMT from baseline compared with insulin after 52 weeks, with a mean difference of − 0.14 mm (95% interval confidence: − 0.25, − 0.02; P = 0.016). Weight and body mass index were both significantly reduced in the exenatide group over 52 weeks. Exenatide reduced total lipoprotein and low-density lipoprotein cholesterol levels more significantly than insulin at weeks 16 and 40. Correlation analyses showed that CIMT was positively correlated with low-density lipoprotein cholesterol.Conclusions: Twice-daily exenatide could prevent atherosclerosis progression in patients with T2DM over a 52-week treatment period compared with insulin therapy.
# Background
Type 2 diabetes mellitus (T2DM) is a chronic and progressive metabolic disease characterized by hyperglycemia due to the defects of insulin secretion and/or action. The incidence rates of atherosclerotic cardiovascular disease and peripheral arterial disease are elevated among patients with T2DM, and some large-scale studies have shown that exogenous insulin therapy may worsen cardiovascular outcomes in T2DM patients.
Human glucagon like peptide 1 (GLP-1) receptor agonists have been developed to augment insulin secretion and inhibit glucagon secretion to control glycemic excursions. It has been reported that fasting total GLP-1 is significantly negatively correlated with CIMT in male T2DM patients. Exenatide twice-daily, a short-acting GLP-1 analog, has been approved for the treatment of T2DM as a GLP-1 receptor agonist. The EXSCEL trial demonstrated that the incidence of major adverse cardiovascular events was not significantly different between patients treated with extended-release exenatide and placebo. In addition, exenatide delays gastric emptying, inhibits food intake, and limits weight gain. A retrospective study revealed that twice-daily exenatide treatment reduced the risks of cardiovascular diseases in patients with T2DM. Exenatide twicedaily showed significant improvement in cardiovascular risk markers including weight, high-density lipoproteincholesterol (HDL-C) level and high-sensitivity C-reactive protein (hsCRP) level. Carotid intima-media thickness (CIMT) is a surrogate marker for subclinical atherosclerosis worldwide using simple and noninvasive B-mode carotid ultrasound. Liraglutide, also a GLP-1 analog, improved the CIMT in T2DM patients in an 18-month prospective study. Exenatide once weekly significantly improved fasting glycemia, glycosylated hemoglobin (HbA1c), body mass index (BMI), lipid profile and CIMT in patients with T2DM in an 8-month prospective study. However, effects of exenatide twice-daily compared with insulin therapy on CIMT in T2DM patients have not been evaluated by a randomized trial so far. In the present study, we aimed to compare the efficacy of exenatide to that of insulin for improving atherosclerosis markers (e.g., CIMT, hsCRP, fibrinogen, and 8-hydroxydeoxyguanosine ), body weight, blood pressure, glycemic control and dyslipidemia (e.g., HbA1c, fasting plasma glucose level, and lipid profile) in a randomized, open-label, controlled trial in T2DM patients.
# Methods
## Study population
This single-center randomized, open-label, controlled trial was performed in the Department of Endocrinology, Beijing Hospital and was performed in accordance with the Declaration of Helsinki. The research ethics committee of Beijing Hospital reviewed and approved the study protocol before the enrollment of patients (No. 2013 BJYYEC-017A-03). All participants were informed of the details of the study and signed the corresponding consent forms.
Patients with T2DM admitted from March 10, 2015 to June 20, 2017 were screened for enrollment. T2DM patients were included if they met the following criteria:
(1) diagnosed with T2DM according to the 1999 WHO criteria; (2) aged between 20 and 75 years; (3) glucose control was not satisfactory with HbA1c level between 7.5 and 11%; (4) had taken at least two oral hypoglycemic drugs with higher than 1/2 of the maximum dose for at least 3 months. Patients were excluded if they had any of the following clinical conditions: type 1 diabetes; > 75% stenosis of any segment of the carotid artery by high frequency B mode ultrasound; an acute cardiovascular event within 30 days prior to randomization; currently planned cardiovascular, carotid or peripheral artery revascularization or cardiac valvular surgery; previous use of insulin or exenatide more than 1 month; an alanine aminotransferase (ALT) or aspartate aminotransferase (AST) level > 2.5 times the upper limit of normal range; serum creatinine concentration ≥ 133 µmol/L for males or ≥ 106 µmol/L for females; history of pancreatitis; currently participating in or having completed another clinical trial within 3 months; or positive for human urinary chorionic gonadotropin or could not adopt a contraceptive method during the study.
## Study design
Eligible patients were randomized 1:1 to receive exenatide or insulin aspart 70/30 using computer codes, and patients were stratified according to the severity of disease (7.5% and 9%). Patients allocated to the exenatide group were given exenatide 5 µg twice-a-day (administration 60 min before breakfast and dinner) subcutaneously, and the dose was increased to 10 µg twice-a-day after 4 weeks. Patients allocated to the insulin group were given insulin aspartate 70/30 subcutaneously. The initial dose of insulin was 0.2-0.4 IU/kg per day and was titrated according to self-monitoring blood glucose and HbA1c detected every 12 weeks. The titration of insulin could be completed in a visit or a telephone follow-up by the investigators following the protocol shown in Additional file 1:. All of the included patients were free to take hypoglycemic drugs except for sulfonylureas and nateglinide drugs. All the participants were educated on a suitable diet and exercise.
The primary outcome of the study was the change in CIMT from baseline to week 52. The secondary objectives included change from baseline to week 52 in atherosclerosis markers (e.g., hsCRP, fibrinogen and 8-OHdG), body weight, blood pressure (diastolic and systolic blood pressure), glycemic control (e.g., HbA1c, fasting plasma glucose level) and dyslipidemia (e.g., total cholesterol [TC], HDL-C, triglyceride , and low-density lipoprotein cholesterol [LDL-C]).
The CIMT was measured using a Philips iU22 Color Doppler ultrasound (Bothell, WA, USA). Two physicians from the Ultrasound Department of Beijing Hospital performed the ultrasound examination after receiving uniform training. Briefly, the patients assumed the supine position, and the IMTs of the bilateral carotid arteries were measured. Measurements were made for three segments: the carotid artery to the dilated portion of the carotid artery, the enlargement of the carotid artery, and the internal carotid artery, which was within a 1-cm range from the distal portion of the carotid artery to the dilated portion of the carotid artery. The IMT of the posterior wall was measured. The maximum IMT values for the posterior wall of the carotid arteries within the three segments were measured, and the bilateral maximum mean value was used for statistical analysis. Each participant was examined by the same sonographer using the same equipment throughout all visits.
# Serum sample analysis
TG, TC, LDL, and HDL were assessed using standard commercial oxidase method (TC intra-assay coefficient of variation [CV] 0.8%, inter-assay CV 2.22%; TG intraassay CV 0.48%, inter-assay CV 1.46%; LDL intra-assay CV 0.71%, inter-assay CV 1.59%; and HDL intra-assay CV 0.83%, inter-assay CV 1.13%;). A kit from Trinity Biotech was used for assessing HA1c (intra-assay CV < 2%, inter-assay CV < 3%), and the glucose oxidase method was used for measuring FPG (intra-assay CV 0.43%, inter-assay CV 1.81%). Hs-CRP was measured by immune turbidimetric assay (intra-assay CV 4%, interassay CV 7%). The evaluation of the fibrinogen level was performed with Kanto Kagaku's assay (intra-assay CV 2.8%, inter-assay CV 5.5%). Determinations of 8-OHdG, Irisin, and brain natriuretic peptide (BNP) was carried out by ELISA (8-OHdG intra-assay CV 5.7%, inter-assay CV 7.2%; Irisin intra-assay CV 5.4%, inter-assay CV 6.7%; and BNP intra-assay CV 5.2%, inter-assay CV 6.3%). All measurements were performed according to the manufacturer's instructions.
# Statistical analysis
All analyzes were performed with SAS V.9.10 software (SAS, Cary, NC, USA). Regarding the sample size estimation, we assumed a 0.15 mm difference in the CIMT between the exenatide and insulin group based on previous liraglutide and DPP-IV studies. Considering a standard deviation of 0.2 mm, 28 patients per treatment group were needed for the primary endpoint analysis to warrant a power of 80% with a two-sided significance level of 0.05. Assuming a drop-out rate of 15%, the sample size needed for each group would be 33, for a total sample size of 66 patients.
The full analysis set was used for statistical analysis. For the primary endpoints, the least-squares mean change from baseline to 52 weeks and associated 95% confidence intervals and P values for exenatide versus insulin were derived from a mixed model for repeated measures with age, sex, duration of T2DM and CIMT at baseline as fixed covariates. Normally distributed data are expressed as means and standard deviations, and the t-test was used for comparison. Skewed data were compared with a nonparametric test. The count data are expressed as proportions, and the frequencies were compared with a Chisquare test. A P < 0.05 indicated a statistically significant difference. All statistical analyses were carried out with IBM SPSS statistical software V.22 for Windows (IBM Corp., Armonk, NY, USA).
# Results
Overall, 80 patients were screened and 14 were excluded for not meeting the inclusion criteria. The remaining 66 patients were randomized into the exenatide group or insulin group at a 1:1 ratio. Finally, 27 patients treated with exenatide and 32 patients treated with insulin were included in the analysis. The baseline characteristics of the patients are shown in. Patients in the two groups were well balanced for most of the baseline characteristics except for gender and diastolic blood pressure. A total of 6 patients did not complete the study due to rash (n = 1), non-response to study drug (n = 3) or loss to follow-up (n = 2) in the exenatide group. In the insulin group, one patient was lost to follow-up .
Exenatide more significantly reduced CIMT than did insulin, with a mean difference of − 0.14 mm (95% interval confidence: − 0.25, − 0.02) after 52 weeks (P = 0.016; . Additionally, exenatide more significantly reduced body weight than did insulin at each time point, P < 0.01), with a mean difference of − 2.21 kg after 52 weeks. Similarly, exenatide reduced BMI significantly more than insulin at each time point, P < 0.01), with a mean difference of − 0.5 kg/m 2 after 52 weeks.
The diastolic and systolic blood pressures were both not significantly reduced in either the exenatide or insulin group after 52 weeks compared with baseline values, with no significant differences between the two groups for either diastolic or systolic blood pressure.
Although exenatide significantly reduced HbA1c from baseline (P < 0.01), this reduction was not significantly greater than that achieved with insulin. For fasting plasma glucose level change, no significant difference was observed between the two groups.
Total cholesterol and LDL-C levels were both reduced more significantly in the exenatide group than in the insulin group at weeks 16 and 40, d, P < 0.05 and P < 0.01, respectively). However, there was no significant difference between the two groups at week 52.
Changes in the hsCRP and fibrinogen levels from baseline were both not significantly different between the two groups after 52 weeks. Both exenatide and insulin significantly reduced the 8-OHdG level from baseline to week 52. Exenatide was associated with a more significant reduction in the 8-OHdG level compared with insulin at week 28 (P < 0.01,, while no significant difference was observed at week 52. The irisin level was increased both in the exenatide and insulin groups after 52 weeks (P < 0.01). The brain natriuretic peptide level was decreased in both the exenatide and insulin groups after 52 weeks. However, no significant difference was observed between the exenatide and insulin group for the irisin or brain natriuretic peptide.
We further performed a correlation analysis to assess the association between CIMT and other markers in this study and found CIMT was positively correlated with LDL-C (r = 0.441,P = 0.021) and Fibrinogen (r = 0.605, P < 0.01). Hypoglycemia occurred in one patient treated with exenatide and five patients treated with insulin. No severe hypoglycemia events were reported in the trial.
# Discussion
An observational study for multi-center (71 centers) demonstrated that 20 weeks of treatment with shortacting exenatide was well tolerated and showed a significant body weight and glucose reduction in T2DM patients whose glycemia had been inadequately controlled with oral hypoglycemic agents. Once-weekly exenatide resulted in a nominal 9% relative reduction in major adverse cardiovascular events and a 14% relative reduction in all-cause mortality compared to placebo in T2DM with and without known cardiovascular disease, whereas dipeptidyl peptidase-4 inhibitors had no effect on cardiovascular risk outcomes but increased risks of acute pancreatitis and hypoglycemia.
Another study showed that the use of dipeptidyl peptidase-4 inhibitors were associated with a reduced risk of heart failure hospitalization compared to GLP-1RAs. Moreover, postoperative exenatide did not provide any additional cardioprotective effect compared to intravenous insulin in coronary artery bypass grafting patients. The effects of GLP-1RA on the cardiovascular risk are still controversial. In this randomized controlled trial, for the first time, exenatide twice-daily more significantly reduced the CIMT in T2DM patients compared with insulin therapy over 52 weeks. To our knowledge, this is the first randomized study to show improvement in subclinical atherosclerosis, as assessed by CIMT, in patients with T2DM over a long period of 52 weeks. In this study, exenatide improved the surrogate atherosclerotic marker CIMT in T2DM patients. The results were consistent with one prospective study for exenatide once weekly, which reported an improvement in CIMT Flow diagram of the study Adjusted mean change from baseline in carotid artery thickness after 52 weeks in the exenatide or insulin group. *P < 0.05 (compared to insulin group). Data are adjusted least-square mean difference in T2DM patients after 8 months of treatment. Recently, several studies have investigated the effects of anti-diabetic agents, especially for GLP-1 receptor agonist, on CIMT in T2DM patients. An observational study showed that exenatide twice-daily improved another surrogate atherosclerotic marker, arterial dilation, in patients with T2DM.
The weight and BMI were both decreased after exenatide treatment for 52 weeks, whereas they were increased after insulin treatment for 52 weeks. The reductions in both weight and BMI were significantly different between the exenatide and insulin groups. The glucose-lowering effect of exenatide was equal to that of premixed insulin used in our study, as evidenced by the non-significant difference in the HbA1c level between the two groups. These results are consistent with previous studies showing a similar decrease in HbA1c between exenatide and insulin treatment in T2DM patients, and exenatide-treated patients lost weight while insulin-treated patients gain weight. In addition, we observed that FPG was not significantly reduced at week 52 from baseline in the exenatide group, whereas it was significantly reduced with insulin treatment. It has been reported that exenatide supports a modest reduction in FBG in T2DM patients while insulin predominantly affects the FBG. Taken together, these results showed that exenatide might not be inferior to insulin in glucose-lowering activity and is superior to insulin in reducing body weight.
Interestingly, we found significant associations between the change in CIMT and LDL-C and fibrinogen concentrations by correlation analysis. These results were not consistent with a previous study, which showed no significant correlation between the CIMT and LDL-C level. However, in liraglutide-treated T2DM patients with metabolic syndrome, CIMT was significantly correlated with the TG level. It has been reported that the fibrinogen concentration increases with the development and progression of T2DM. However, neither exenatide nor insulin could reduce the fibrinogen concentration in our study. We also found that exenatide reduced the total cholesterol and LDL-C levels more significantly than did insulin at weeks 16 and 40. The total cholesterol and LDL-C levels were both reported to be significantly related with cardiovascular disease and atherosclerosis risk. Exenatidecould reduce the serum total cholesterol level by about 5%, TG level by about 12%, and LDL-C level by about 6%, while increasing the HDL-C level by about 24% in T2DM patients for at least 3 years. There were no significant differences between these two groups in total cholesterol and LDL-C level at week 52. However, the average values for total cholesterol and LDL-C were both in normal range at baseline, which may explain lack of effect of exenatide on these cholesterol levels. Thus, the effect of exenatide on total cholesterol and LDL-C levels remains to be investigated in future studies, preferably in patients with hypercholesteremia.
The diastolic and systolic blood pressures were both not significantly reduced in the exenatide group after 52 weeks compared with baseline values, with no significant differences between the exenatide and insulin groups for either diastolic or systolic blood pressure. Notably, the systolic blood pressure was increased in the insulin group after 52 weeks of treatment. A post hoc analysis showed that the exenatide twice-daily dose did not affect blood pressure in T2DM patients, while another previous study reported that exenatide could reduce the diastolic and systolic blood pressures in T2DM patients. In our study, the lack of efficacy of exenatide on blood pressure compared with insulin may because the diastolic or systolic blood pressure of most patients at baseline was well regulated. Additionally, we found that exenatide was associated with a more remarkable reduction in 8-OHdG, a marker of oxidative stress to DNA and increased risk of atherosclerosis, compared with insulin. Taken together, our results suggest a potential role for exenatide in preventing atherosclerosis progression in T2DM patients.
Our study has some limitations. First, this singlecenter study did not have enough power to evaluate the influences of exenatide on some metabolic outcomes. Second, the effects of exenatide on clinical outcomes, particularly on cardiovascular outcomes, should be further compared with those of insulin. Finally, study participants were free to use metformin, and whether metformin and exenatide combination therapy affects the levels of metabolic and atherosclerotic markers deserves further investigation.
# Conclusions
In conclusion, exenatide inhibits atherosclerotic progression to delay the development of cardiovascular disease in patients with T2DM compared to insulin therapy, in addition to its benefits of glucose-lowering, body weight control, and dyslipidemia improvement.
## Supplementary information
Supplementary information accompanies this paper at https ://doi. org/10.1186/s1293 3-020-01014 -7.
Additional file 1:. Titration protocol for insulin in patients with type 2 diabetes mellitus.
## Abbreviations
T2DM: Type 2 diabetes mellitus; CIMT: Carotid intima-media thickness; GLP-1: Glucagon like peptide 1; 8-OHdG: 8-Hydroxydeoxyguanosine; ARB: Angiotensin receptor blocker; ACEI: Angiotensin I Converting Enzyme Inhibitor. |
Prognostic value of procalcitonin in respiratory tract infections across clinical settings
Introduction: Whether the inflammatory biomarker procalcitonin provides prognostic information across clinical settings and different acute respiratory tract infections (ARIs) is poorly understood. In the present study, we investigated the prognostic value of admission procalcitonin levels to predict adverse clinical outcome in a large ARI population.Methods: We analysed data from 14 trials and 4,211 ARI patients to study associations of admission procalcitonin levels and setting specific treatment failure and mortality alone at 30 days. We used multivariable hierarchical logistic regression and conducted sensitivity analyses stratified by clinical settings and ARI diagnoses to assess the results' consistency.Results: Overall, 864 patients (20.5%) experienced treatment failure and 252 (6.0%) died. The ability of procalcitonin to differentiate patients with from those without treatment failure was highest in the emergency department setting (treatment failure area under the curve (AUC): 0.64 (95% confidence interval (CI): 0.61, 0.67), adjusted odds ratio (OR): 1.85 (95% CI: 1.61, 2.12), P <0.001; and mortality AUC: 0.67 (95% CI: 0.63, 0.71), adjusted OR: 1.82 (95% CI: 1.45, 2.29), P <0.001). In lower respiratory tract infections, procalcitonin was a good predictor of identifying patients at risk for mortality (AUC: 0.71 (95% CI: 0.68, 0.74), adjusted OR: 2.13 (95% CI: 1.82, 2.49), P <0.001). In primary care and intensive care unit patients, no significant association of initial procalcitonin levels and outcome was found. Conclusions: Admission procalcitonin levels are associated with setting specific treatment failure and provide the most prognostic information regarding ARI in the emergency department setting.
# Introduction
The assessment of disease severity and prediction of adverse outcome in patients with acute respiratory tract infections (ARIs) is essential to improve patient management, including therapeutic and diagnostic steps and site-of-care decisions [bib_ref] Strategies for the treatment of sepsis, Warren [/bib_ref] [bib_ref] Septic shock in humans: advances in the understanding of pathogenesis, cardiovascular dysfunction,..., Parrillo [/bib_ref] [bib_ref] Duration of hypotension before initiation of effective antimicrobial therapy is the critical..., Kumar [/bib_ref]. For this purpose, different blood biomarkers have been evaluated to predict mortality in smaller patient cohorts and critically ill patients, but results are still somewhat controversial in regard to type of biomarker for specific ARI subpopulation and optimal cutoff levels for clinical routine [bib_ref] Presepsin (soluble CD14 subtype) and procalcitonin levels for mortality prediction in sepsis:..., Masson [/bib_ref] [bib_ref] Prognostic and diagnostic value of eosinopenia, C-reactive protein, procalcitonin, and circulating cell-free..., Garnacho-Montero [/bib_ref] [bib_ref] Hospital mortality prognostication in sepsis using the new biomarkers suPAR and proADM..., Suberviola [/bib_ref] [bib_ref] Blood biomarkers for personalized treatment and patient management decisions in community-acquired pneumonia, Schuetz [/bib_ref] [bib_ref] Comparison between white blood cell count, procalcitonin and C reactive protein as..., Magrini [/bib_ref] [bib_ref] Serum soluble triggering receptor expressed on myeloid cells-1 and procalcitonin can reflect..., Li [/bib_ref] [bib_ref] Evaluation of the Mortality in Emergency Department Sepsis score combined with procalcitonin..., Zhao [/bib_ref].
Procalcitonin (PCT), a precursor protein of calcitonin, is currently one of the most frequently used infectious disease biomarkers in clinical practice [bib_ref] Sepsis biomarkers: a review, Pierrakos [/bib_ref]. PCT levels mirror severity and evolution of infection and are thought to be associated with poor prognosis in patients with sepsis and ARI [bib_ref] the Finnsepsis Study Group. Predictive value of procalcitonin decrease in patients with..., Karlsson [/bib_ref]. Changes in PCT level in response to therapeutic treatment have also been reported, which suggests prognostic significance in a variety of clinical settings [bib_ref] Early changes of procalcitonin may advise about prognosis and appropriateness of antimicrobial..., Georgopoulou [/bib_ref] [bib_ref] Prognostic value of procalcitonin (PCT) and/or interleukin-6 (IL-6) plasma levels after multiple..., Haasper [/bib_ref]. However, there is a lack of clinical data comparing PCT in different clinical settings (primary care, emergency department (ED), intensive care unit (ICU)) and across ARI diagnoses to aid in better understanding its (prognostic) value in daily clinical practice.
In the present study, we investigated the prognostic value of admission PCT levels to predict treatment failure and mortality alone in a large ARI patient population across different settings, ARI diagnoses and countries by performing an individual patient data meta-analysis.
# Material and methods
## Patients and setting
This analysis includes all patients from a previous individual patient data meta-analysis [bib_ref] Procalcitonin to guide initiation and duration of antibiotic treatment in acute respiratory..., Schuetz [/bib_ref]. The initial metaanalysis was prespecified in collaboration with the Cochrane Database of Systematic Reviews [bib_ref] Procalcitonin to initiate or discontinue antibiotics in acute respiratory tract infections, Schuetz [/bib_ref]. In brief, the aim of the meta-analysis was to assess the safety and efficacy of using PCT to guide initiation and duration of antibiotic treatment in patients with ARI assigned to routine PCT measurement or standard of care without PCT measurement. This approach was used over a large range of patients with varying severities of disease in different clinical settings. Patients with a clinical diagnosis of either upper or lower ARI derived from 14 randomised or quasi-randomised trials were included. Trials focused exclusively on paediatric patients or on a purpose other than initiation and duration of antibiotic therapy were not included. Further details about identifying suitable trials were published previously [bib_ref] Procalcitonin to initiate or discontinue antibiotics in acute respiratory tract infections, Schuetz [/bib_ref]. No ethical approval was needed for this meta-analysis. Written informed consent was obtained from all participants within the initial trials, including consent to participate in further analyses.
## Search for and identification of trials
The initial search was prespecified and published previously [bib_ref] Procalcitonin to initiate or discontinue antibiotics in acute respiratory tract infections, Schuetz [/bib_ref]. In brief, suitable trials were identified by a formal search of the Cochrane Controlled Trials Registry (CCTR), MEDLINE and Embase (all from their inception to May 2011), as well as reference lists of reports describing such trials. In addition, conference proceedings (from 2006 to 2011) of the Interscience Conference on Antimicrobial Agents and Chemotherapy, the European Congress of Clinical Microbiology and Infectious Disease, the American Thoracic Society, the American Association of Respiratory Care and the American College of Chest Physicians were hand-searched. Trial registries were checked and experts contacted for further eligible trials. Two independent reviewers assessed trial eligibility on basis of titles, abstracts, full-text reports and further information obtained from investigators as needed. Further, the protocol, case report forms and unedited databases from investigators of all eligible trials were requested. Data from each trial were first checked against reported results, and queries were resolved with the principal investigator, trial data manager or statistician.
## Aims and endpoints
The aim of the present analysis was to study associations of admission PCT levels and adverse clinical outcomes. In line with the initial Cochrane meta-analysis protocol [bib_ref] Procalcitonin to initiate or discontinue antibiotics in acute respiratory tract infections, Schuetz [/bib_ref] , the predefined primary endpoints were defined as all-cause mortality and setting specific treatment failure at 30 days. For the primary care setting, treatment failure was defined as occurrence of at least one of the following events: death, hospitalisation, ARI-specific complications (for example, empyema for lower ARI, meningitis for upper ARI), recurrent or worsening infection, and patients reporting any symptoms of an ongoing respiratory infection (for example, fever, cough, dyspnoea) at 30-day follow-up. For the ED setting, treatment failure was defined as death, ICU admission, rehospitalisation after index hospital discharge, ARI-associated complications (for example, empyema or acute respiratory distress syndrome for lower ARI), or recurrent or worsening infection within 30 days of follow-up. For the ICU setting, treatment failure was defined as death within 30 days of follow-up.
In regard to PCT as a baseline predictor, we used initial PCT levels corresponding to the PCT level drawn at the primary care visit (primary care), ED admission (ED trials) and ICU admission (ICU trials). In all trials, PCT was measured using a rapid, sensitive assay with a functional assay sensitivity of 0.06 μg/L (KRYPTOR PCT; B- R- A- H- M- S, Hennigsdorf, Germany) and an assay time of less than 20 minutes. We used different a priori defined PCT cutoffs (0.1 μg/L, 0.25 μg/L, 0.5 μg/L, 2.0 μg/L) that correspond to cutoffs used in previous antibiotic stewardship trials and also in practice guidelines on the use of PCT.
# Statistical analysis
We used descriptive statistics, including mean with standard deviation, median with interquartile range (IQR) and frequencies, to describe the populations, as appropriate.
Patients were divided in two groups depending on having experienced a treatment failure or not. Because the rate of loss to follow-up in the different trials was low (<10%), we assumed that patients lost to follow-up did not undergo an event. This assumption was also verified in a time-to-event analysis where we used censoring for patients lost to follow-up.
For the primary endpoint of setting-and diagnosisspecific treatment failure or all-cause mortality alone, we calculated adjusted odds ratios (ORs) (with age and sex as additional fixed effects) and 95% confidence intervals (CIs) using multivariable hierarchical logistic regression. Discrimination of PCT levels for treatment failure or mortality was investigated by area under the receiver operating characteristic (ROC) curve and 95% CIs. For the different clinical settings (primary care, ED, ICU) and most prevalent ARI subgroups (that is, acute bronchitis, exacerbated chronic obstructive pulmonary disease (ECOPD), community acquired pneumonia (CAP)), we calculated sensitivity and specificity analyses. We used different clinically established PCT cutoffs (0.1 μg/L, 0.25 μg/L, 0.5 μg/L, 2.0 μg/L) to estimate risk prediction. If subgroups were too small, values were labelled as 'not applicable'. For graphical display, we calculated Kaplan-Meier curves for time to death and time to adverse outcome.
Tests were carried out at 5% significance levels. Analyses were performed with STATA 12.1 software (StataCorp, College Station, TX, USA).
# Results
## Population
The study sample comprised all 4,211 intention-to-treat patients (median age: 63 years, 54.2% male) with a final diagnosis of ARI. Regarding the clinical setting, a total of 1,008 (23.9%) patients were from primary care, 2,605 (61.9%) from the ED and 598 (14.2%) from the ICU. Overall, 864 patients (20.5%) had a treatment failure and 252 (6.0%) patients died. Baseline characteristics of the overall population and stratified based on the occurrence of a treatment failure are summarized in .
## Procalcitonin levels in different settings and acute respiratory tract infections
As shown in [fig_ref] Figure 1: Admission procalcitonin levels [/fig_ref] , PCT levels significantly increased with higher-acuity clinical settings (that is, primary care, ED, ICU). Similar results were found when we looked at PCT levels in different types of ARI with higher levels in more severe ARIs (that is, ECOPD, CAP, ventilator-associated pneumonia) [fig_ref] Figure 1: Admission procalcitonin levels [/fig_ref]. [fig_ref] Table 2: summarizes the results of logistic regression analysis as a measure of association... [/fig_ref]. shows the results of the adjusted regression analysis for treatment failure and mortality in the different populations studied. PCT in the different cutoff ranges significantly separated ED patients in regard to the time to treatment failure and time to mortality (P <0.0001 and P = 0.02, respectively). Similar analyses did not show significant results in the primary care and ICU settings (data not shown). In the overall lower ARI population, as well as in CAP patients, PCT also showed significant separation in regard to the time to treatment failure (P <0.0001 and P <0.01, respectively) and time to mortality (P <0.0001 and P <0.01, respectively) [fig_ref] Figure 4: Association between time to severe adverse events and admission procalcitonin levels in... [/fig_ref]. [fig_ref] Table 3: Procalcitonin cutoffs [/fig_ref] gives an overview of diagnostic measures (including sensitivity, specificity, negative predictive values (NPVs) and positive predictive values (PPVs)) of PCT at different PCT cutoff levels. In primary care patients, a cutoff of 0.25 μg/L had a high PPV of 45.8% (95% CI: 25. [bib_ref] Hospital mortality prognostication in sepsis using the new biomarkers suPAR and proADM..., Suberviola [/bib_ref]
## Association of admission procalcitonin levels and adverse outcome
# Discussion
The findings of this analysis, which included a large population with different types and severities of ARI from previous randomised controlled trials, are threefold. First, we found an increase in PCT levels across clinical settings and ARI diagnoses, demonstrating that the current practice of using different cutoff levels in different clinical settings is mandatory. Second, initial PCT levels were significantly associated with setting-specific treatment failure and mortality in the overall population. Third, the prognostic information derived from initial PCT levels in primary care and ICU patients was rather low, whereas in the ED setting and in ECOPD and bronchitis patients, admission PCT levels provided prognostic information and thereby may improve initial risk stratification in these patient populations.
An accurate and fast assessment of disease severity and predictions regarding a patient's clinical course potentially assist patients and physicians with settingappropriate expectations regarding the illness. Such predictions are of particular importance to ensuring efficient use of health care and hospital resources and are indispensable for choosing optimal therapeutic options in the initial management of ARIs [bib_ref] Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management..., Mandell [/bib_ref]. This includes decisions regarding site of care, diagnostic evaluation and assessment for appropriate early discharge [bib_ref] Biomarker-guided personalised emergency medicine for all -hope for another hype, Schuetz [/bib_ref]. The role of prognostication is also acknowledged by respiratory infection guidelines, which recommend stratifying patients with CAP on the basis of predicted risk for mortality using validated risk scores (that is, the Pneumonia Severity Index (PSI) or CURB-65 score) [bib_ref] Guidelines for the management of adults with communityacquired pneumonia: diagnosis, assessment of..., Niederman [/bib_ref] [bib_ref] Guidelines for the management of adult lower respiratory tract infections, Woodhead [/bib_ref]. Clinical risk scores are somewhat limited by practicality and risk for miscalibration due to different patient populations, and therefore they have only moderate operational characteristics [bib_ref] Predicting mortality with pneumonia severity scores: importance of model recalibration to local..., Schuetz [/bib_ref]. Also, these scores are validated only for CAP and not for other ARIs. Thus, there is interest in additional prognosticating mechanisms, throughout all clinical settings and subgroups of ARIs, using newly available biomarkers (such as PCT) that are objectively and rapidly measurable, as well as responsive to clinical recovery, and that add relevant, reliable, real-time information.
The prognostic potential of PCT in patients with respiratory infections has been evaluated in different studies, mainly in patients with CAP and sepsis. Most studies have found PCT levels to be increased in patients not surviving their disease compared with survivors, but with only moderate prognostic accuracy. In line with our results, in a large CAP cohort in the United States [bib_ref] Risk prediction with procalcitonin and clinical rules in community-acquired pneumonia, Huang [/bib_ref] , the greatest benefit of PCT was found in patients classified as high risk on the basis of PSI score. Having a PCT <0.1 μg/L virtually excluded mortality in these high-risk patients. In a Swiss study, researchers found that initial PCT levels did not improve clinical risk scores for mortality prediction [bib_ref] Prognostic value of procalcitonin in community-acquired pneumonia, Schuetz [/bib_ref]. Subsequent repeated measurements of PCT in this population demonstrated improved clinical outcomes with falling PCT levels. In addition, the study found that PCT was more helpful in predicting treatment failure other than mortality, such as ICU admission or CAP-related complications. For these outcomes, PCT significantly improved clinical risk scores. In another large CAP study, done in Germany and including mostly low-risk patients, researchers found that PCT was a fair predictor of mortality and significantly improved clinical risk scores [bib_ref] Assessment of inflammatory markers in patients with community-acquired pneumonia-influence of antimicrobial pre-treatment:..., Krüger [/bib_ref]. According to these studies, our findings are generally congruent. In low-risk patients (primary care) with respiratory infections, low PCT levels <0.1 μg/L correctly identified patients at lower risk for treatment failure with a high NPV. In higher-risk populations (CAP, ICU), increased levels of PCT had high PPV mainly for mortality prediction. In ED patients and in patients with lower ARIs, PCT was a good predictor of treatment failure and mortality.
Nonetheless, several limitations should be considered when interpreting our results. First, we carried out a secondary analysis of a meta-analysis. It was designed to focus on a different question, that is, whether using PCT to guide initiation and duration of antibiotic treatment in patients with ARIs in primary care, ED and ICUs is safe and efficient in a broader patient population. Second, we did not investigate the effect of repeated measurements that might add supplementary prognostic information to admission levels. Third, we focused only on interventional trials as specified in the meta-analysis protocol and did not include any observational data that would potentially allow for larger sample sizes in the different subgroups and increases in the patient spectrum, as in observational studies less rigorous exclusion criteria usually apply. Fourth, we focused on adverse outcomes as specified in the original trials, but we were not able to look into some adverse outcomes, such as septic shock and respiratory failure, among others. Fifth, clinical risk scores such as CURB-65 and PSI for CAP and Acute Physiology and Chronic Health Evaluation for ICU patients were not routinely available and thus were not included in this analysis. Similarly, we were not able to obtain more detailed baseline data on all patients, including comorbidities and other risk factors, which would have allowed a more rigorous adjustment in the statistical models. Although this study provides new and clinically relevant information about the prognostic value of initial admission PCT values in different clinical setting and for different upper and lower ARIs, it remains unclear whether an improved initial prognostic assessment based on PCT level would translate into better triage decisions and outcomes in patients. Sixth, randomised controlled outcome studies need to be conducted for patients with ARIs in the ED and ICU to investigate whether real-life measurement of PCT adds useful prognostic information and thereby improves the daily clinical management and outcomes of patients.
# Conclusions
This is the first large-scale study in which the prognostic value of admission PCT levels has been investigated throughout all clinical settings in the field of ARIs. PCT levels mainly provide prognostic information for risk stratification of patients in the ED setting and in patients with low to moderate ARIs. Future randomised controlled studies must address whether adding PCT to initial risk assessment can improve patient management and prognostication.
## Key messages
We found an increase in PCT levels across the clinical settings and ARI diagnoses, supporting the current practice of using specific PCT cutoff levels in different clinical settings. PCT levels are significantly associated with treatment failure and mortality in the overall population. The prognostic information of initial PCT levels in primary care and ICU patients is rather low, whereas admission PCT levels provide important prognostic information in the ED setting and in patients with low to moderate ARIs.
## Competing interests
No commercial sponsor had any involvement in the design or conduct of this study, namely, the collection, management, analysis and interpretation of the data; and the preparation, decision to submit, review, or approval of the manuscript. AK, SS, PS, MC-C and BM received support from B- R- A- H- M- S to attend meetings and fulfil speaking engagements. PS, MC-C and BM received support from bioMérieux to attend meetings and fulfil speaking engagements. BM has served as a consultant to and received research support from B- R- A- H- M- S and bioMérieux. MB and HCB report receiving grants from B- R- A- H- M- S during the conduct of the study. DS and MT received research support from B- R- A- H- M- S. TW reports receiving personal fees from Thermo Fisher and Becton Dickinson and grants from Siemens Diagnostics during the conduct of the study. Outside the presently reported study, TW reports receiving personal fees from Bayer, Novartis, Basilea Pharmaceutica, Pfizer, Cubist, AstraZeneca, GSK and MSD. C-EL reports receiving personal fees from bioMérieux, grants from Janssen and grants from Bayer outside the submitted work. MW reports receiving personal fees from Pfizer and Gilead outside the submitted work. FT reports receiving grants from Abbott, AstraZeneca, Pfizer and Schering-Plough outside the submitted work. All other authors declare that they have no competing interests.
# Authors' contributions
All authors made substantive intellectual contributions to this study. AK, BM and PS conducted statistical analyses and drafted the first manuscript. MB, MC-C, DS, LB, MW, KBB, LW, OB, TW, SS, VN, MT, HB, C-EL, JC and FT were in charge of the acquisition of patient data in the included trials, provided individual patient data from their respective trials and resolved queries about their trial data as needed. For this article, they made substantial contributions to conception and design and took an active part in acquisition, analysis and interpretation of data. NB is a librarian who did the search for our metaanalysis and helped in the search and interpretation of studies. She gave feedback on the manuscript and was part of the team that developed the ideas for the meta-analysis. All authors contributed to the interpretation of data and to the revising of the manuscript critically for important intellectual content. All authors approved the final version of the manuscript and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
[fig] Figure 1: Admission procalcitonin levels. (A) Procalcitonin (PCT) levels stratified by setting (P = 0.0001). (B) PCT levels stratified by diagnosis (P = 0.0001). ECOPD, Exacerbated chronic obstructive pulmonary disease; ED, Emergency department; ICU, Intensive care unit; CAP, Community-acquired pneumonia; PC, Primary care; VAP, Ventilator-associated pneumonia. [/fig]
[fig] Figure 2, Figure 3: Multivariate regression analysis for estimation of predictive value of procalcitonin levels on admission stratified by adverse events and mortality in different settings and diagnoses. *Treatment failure;~Mortality; Adj., Adjusted; ARI, Acute respiratory tract infection; CAP, Community-acquired pneumonia; CI, Confidence interval; ICU, Intensive care unit; COPD, Chronic obstructive pulmonary disease OR, Odds ratio; VAP, Ventilator-associated pneumonia. Association between time to severe adverse events and admission procalcitonin levels in emergency department patients. (A) Time to treatment failure is significantly shorter in emergency department (ED) patients with higher procalcitonin (PCT) levels on admission. (P <0.0001. (B) Time to death is significantly shorter in ED patients with higher PCT levels on admission (P <0.02). Treatment failure is defined as death, ICU admission, rehospitalisation, complications or recurrent or worsening infection within 28 days. [/fig]
[fig] Figure 4: Association between time to severe adverse events and admission procalcitonin levels in lower acute respiratory tract infection and community-acquired pneumonia patients. (A) Time to treatment failure is significantly shorter in lower acute respiratory tract infection (ARI) patients with higher procalcitonin (PCT) levels on admission (P <0.0001). (B) Time to death is significantly shorter in lower ARI patients with higher PCT levels on admission (P <0.0001). (C) Time to treatment failure is significantly shorter in patients with communityacquired pneumonia (CAP) with higher PCT levels on admission (P <0.01). (D) Time to death is significantly shorter in patients with CAP with higher PCT levels on admission (P <0.01). ED, Emergency department. Treatment failure is defined as death, ICU admission, rehospitalisation, complications or recurrent or worsening infection within 28 days. [/fig]
[fig] 4: Procalcitonin cutoffs (0.5 and 2.0 μg/L) for risk prediction in different clinical settings and in various acute respiratory infection subgroups a [/fig]
[table] Table 2: summarizes the results of logistic regression analysis as a measure of association and AUC as a measure of discrimination for both endpoints: treatment failure and mortality. PCT levels in ED patients were significantly as-95% CI: 1.45, 2.29), P <0.001), whereas no significant effect was seen in primary care or ICU patients. InTable 1Baseline characteristics a [/table]
[table] Table 3: Procalcitonin cutoffs (0.1 and 0.25 μg/L) for risk prediction in different clinical settings and in various acute respiratory infection subgroups a [/table]
|
Genomic Comparison of the P-ATPase Gene Family in Four Cotton Species and Their Expression Patterns in Gossypium hirsutum
Plant P-type H + -ATPase (P-ATPase) is a membrane protein existing in the plasma membrane that plays an important role in the transmembrane transport of plant cells. To understand the variety and quantity of P-ATPase proteins in different cotton species, we combined four databases from two diploid cotton species (Gossypium raimondii and G. arboreum) and two tetraploid cotton species (G. hirsutum and G. barbadense) to screen the P-ATPase gene family and resolved the evolutionary relationships between the former cotton species. We identified 53, 51, 99 and 98 P-ATPase genes from G. arboretum, G. raimondii, G. barbadense and G. hirsutum, respectively. The structural and phylogenetic analyses revealed that the gene structure was consistent between P-ATPase genes, with a close evolutionary relationship. The expression analysis of P-ATPase genes showed that many P-ATPase genes were highly expressed in various tissues and at different fiber developmental stages in G. hirsutum, suggesting that they have potential functions during growth and fiber development in cotton. ), P 2D (Na + or Ca 2+ ATPase), and P 3B (Mg 2+ ATPase)[5,6]. Forty-six P-ATPase genes have been identified in Arabidopsis [7], and 43 P-ATPase genes have been identified in rice[8]. The gene numbers of five subfamilies (P 1B , P 2A , P 2B , P 3A , and P 4 ) are high, but there are few P 5 subfamily genes, while the genes of the P 2C , P 2D , P 1A and P 3B subfamilies were absent in Arabidopsis and rice[8].Some studies have revealed that P-ATPase genes play an important role in primary metabolic and secondary metabolic processes in plants. For example, some genes of the P 3A subfamily can affect the transport of anthocyanins and proanthocyanidins (PAs) in plants, such as TT13 in Arabidopsis[9]and PH5 in petunia[10]. However, there are only a few detailed reports about P 3A -ATPase in cotton, especially involving the transport of PAs, while another study reported that pigment formation in brown cotton fiber is related to the transport of PAs[11]. Cotton is the most important fiber crop in China with wide varieties. The whole genome sequencing of two diploid cotton species (G. arboretum and G. raimondii) and two tetraploid cotton species (G. barbadense and G. hirsutum) has been completed[12][13][14][15]. In this study, the P-ATPase genes of four cotton species were identified, screened and analyzed using bioinformatics, the evolutionary relationship of the P-ATPase family in cotton has been preliminarily discussed, and the expression pattern of the P-ATPase gene in upland cotton has been analyzed. To search for possible genes related to PA accumulation, fluorescent quantitative analysis of the P 3A -ATPase subfamily of genes of brown cotton fibers was used, combined with the accumulation of PAs in brown cotton fibers. This study lays the foundation for further cloning and utilization of P-ATPase genes to cultivate new varieties of colored cotton.
# Introduction
H + -ATPase is a membrane protein that is widely present in the plant plasma membrane and various endomembranes, and plays an important role in cell metabolism. The representative H + -ATPases are currently divided into three types according to structure, namely the vacuolar-type H + -ATPases (V-ATPases), H + -pyrophosphatases (H + -PPases) and P type H + -ATPases (P-ATPase). The H + -PPases is located in vacuolar or endomembranes; the vacuole H + -ATPase is a V-type ATPase; and the plasma membrane H + -ATPase belongs to the P-type ATPase. As a P-type pump, P-ATPase can generate and maintain the electrochemical gradient of H + on both sides of the cell membrane by decomposing intracellular ATP and providing energy for the transmembrane transport of various nutrients and ions. In addition, the plant P-ATPase also participates in the regulation of various physiological processes, such as intracellular pH, cell elongation, stomatal opening and closing, and plant response to environmental stress.
To date, 10 subfamilies of P-ATPase have been identified according to their transport of various ions, heavy metals and possibly lipids: P 1B (heavy metal ion ATPases [HMA]), P 2A (endoplasmic reticulum Ca 2+ ATPase [ECA]), P 2B (autoinhibited Ca 2+ ATPase [ACA]), P 3A (autoinhibited H + ATPase (AHA)), P 4 (putative aminophospholipid ATPase [ALA]), P 5 (unknown ATPase), P 1A (K + ATPase), P 2C
# Results
## Identification of p-atpase genes in cotton
We first queried the specific domains of Arabidopsis AHA10 (PF00122, PF00702, and PF00690) on the Pfam website and used the amino acid sequences of these domains to search for candidate genes in the local databases of G. arboretum, G. raimondii, G. barbadense and G. hirsutum, which were previously established. We screened these candidate genes using the PFAM and SMART databasesand identified 301 P-ATPase genes with specific domains of P-ATPase genes in four cotton species. Fifty-one P-ATPase genes were found in G. raimondii and 53 P-ATPase genes were found in G. arboretum, similar to those in Arabidopsis and rice, containing 46 and 43 P-ATPase genes, respectively. However, 99 and 98 P-ATPase genes were found in the tetraploid cotton G. barbadense and G. hirsutum, respectively. These genes in cotton were classified and named according to the classifications and names of the P-ATPase genes in Arabidopsis.
In addition, the lengths and molecular weights of proteins encoded by the P-ATPase genes were analyzed. The characteristics of 301 P-ATPase genes in G. raimondii, G. hirsutum, G. arboreum and G. barbadense suggested that the lengths of all P-ATPase proteins differed largely, ranging from 800 to 1301 aa; the molecular weights of the P-ATPase proteins ranged from 90 to 140 kDa in G. arboretum and G. raimondii, and the lengths and molecular weights of most P-ATPase proteins in G. barbadense and G. hirsutum had a similar trend as those in G. arboretum and G. raimondii, but some proteins were different. For example, GhHMA10 encodes 420 amino acids with a molecular weight of 44.9 kDa, and GbAHA23 encodes 1646 aa with a molecular weight of 183.48 kDa. The isoelectric points of all P-ATPase proteins were predicted; the minimum was 4.82, and the maximum was 8.81 .
## Chromosome distribution of p-atpase genes
To better understand the distribution of P-ATPase genes on chromosomes, according to the genome database information of G. raimondii, G. arboretum and G. hirsutum, the chromosome localization was mapped. The results showed that the P-ATPase genes of G. hirsutum were tagged on 21 chromosomes, and the P-ATPase genes of G. raimondii and G. arboretum were tagged on 12 and 13 chromosomes. The P 1B -ATPase genes of G. hirsutum were distributed in nine chromosomes (At1, Dt1, At4, At5, Dt5, At8, Dt8, At9 and Dt9), and each chromosome had only one gene, while the P 1B -ATPase genes of G. raimondii and G. arboretum were all distributed in 6 chromosomes, with only one or two genes on each chromosome. The distributions of the P 2A -ATPase, P 2B -ATPase, and P 3A -ATPase genes in G. raimondii, G. arboretum and G. hirsutum were similar to that of the P 1B -ATPase gene, but a few chromosomes had different distributions. For instance, there were four P 2B -ATPase genes on the 6 th chromosome of G. raimondii, 4 P 2B -ATPase genes on the 3 rd chromosome of G. arboretum, and six P 3A -ATPase genes on the Dt9 chromosome. The P 4 subfamily genes in four cotton species are evenly distributed on chromosomes, with one or two genes per chromosome. and P3A-ATPase genes in G. raimondii, G. arboretum and G. hirsutum were similar to that of the P1B-ATPase gene, but a few chromosomes had different distributions. For instance, there were four P2B-ATPase genes on the 6 th chromosome of G. raimondii, 4 P2B-ATPase genes on the 3 rd chromosome of G. arboretum, and six P3A-ATPase genes on the Dt9 chromosome. The P4 subfamily genes in four cotton species are evenly distributed on chromosomes, with one or two genes per chromosome.
## Phylogenetic analysis of the p-atpase genes
To determine the subfamily classification of P-ATPase genes in cotton species and the number of genes in each subfamily, a phylogenetic tree was constructed with the sequences of 51 P-ATPase proteins in G. raimondii, 53 P-ATPase proteins in G. arboretum, 98 P-ATPase proteins in G. hirsutum, 99 P-ATPase proteins in G. barbadense and 45 P-ATPase proteins in Arabidopsis. As shown in the result, all P-ATPases genes in each cotton species could be clustered into six subfamilies: P1B, P2B, P2A, P3A, P4 and P5. The gene numbers of each subfamily in G. raimondii and G. arboretum were very similar, and there were large numbers in the P2B and P3A subfamilies, which were approximately twice those in the P1B, P2A and P4 subfamilies.. In G. hirsutum, the gene numbers were approximately twice that of G. raimondii and G. arboretum, but the distribution of genes in each
## Phylogenetic analysis of the p-atpase genes
To determine the subfamily classification of P-ATPase genes in cotton species and the number of genes in each subfamily, a phylogenetic tree was constructed with the sequences of 51 P-ATPase proteins in G. raimondii, 53 P-ATPase proteins in G. arboretum, 98 P-ATPase proteins in G. hirsutum, 99 P-ATPase proteins in G. barbadense and 45 P-ATPase proteins in Arabidopsis. As shown in the result, all P-ATPases genes in each cotton species could be clustered into six subfamilies: P 1B , P 2B , P 2A , P 3A, P 4 and P 5. The gene numbers of each subfamily in G. raimondii and G. arboretum were very similar, and there were large numbers in the P 2B and P 3A subfamilies, which were approximately twice those in the P 1B, P 2A and P 4 subfamilies.. In G. hirsutum, the gene numbers were approximately twice that of G. raimondii and G. arboretum, but the distribution of genes in each subfamily was similar to G. raimondii, with the most genes in the P 3A subfamily and the least in the P 2A subfamily. In G. barbadense, P 2B -ATPase genes were most abundant, while P 2A -ATPase genes were least abundant. Interestingly, the member of P 5 subfamily was very few in the five plants, only one or two genes . subfamily was similar to G. raimondii, with the most genes in the P3A subfamily and the least in the P2A subfamily. In G. barbadense, P2B-ATPase genes were most abundant, while P2A-ATPase genes were least abundant. Interestingly, the member of P5 subfamily was very few in the five plants, only one or two genes .
## Evolutionary relationship analysis of p-atpase genes
To gain insight into the evolutionary relationship of P-ATPase genes in G. arboretum, G. raimondii, G. barbadense and G. hirsutum, a phylogenetic tree was constructed according to the protein sequences of the P-ATPase genes in four cotton species, and the P-ATPase gene structures were analyzed with their CDS sequences using the GSDS online tool. The results showed that most P-ATPase genes contained introns and exons, and most P-ATPase genes in G. raimondii and G. barbadense also contained untranslated regions (UTRs). Almost all P3A subfamily genes contained 10-20 introns, the P1B subfamily genes contained 5-20 introns, and most P2A subfamily genes contained 5-10 introns, but some genes contained more than 20 introns. Interestingly, the structures of the P1B subfamily genes on different branches had significant differences. Most genes in class C did not contain introns or contained only one, and most genes in class A contained 5-7 introns, but all genes in class B contained more than 30 introns. In the phylogenetic tree, the gene structures of most paralogous gene pairs between G. hirsutum and G. arboretum/G. raimondii were very similar, which suggests that their genetic relationship is close, and these paralogous genes had high homology. However, the gene structures of paralogous gene pairs between G. barbadense and G. arboretum/G.
## Evolutionary relationship analysis of p-atpase genes
To gain insight into the evolutionary relationship of P-ATPase genes in G. arboretum, G. raimondii, G. barbadense and G. hirsutum, a phylogenetic tree was constructed according to the protein sequences of the P-ATPase genes in four cotton species, and the P-ATPase gene structures were analyzed with their CDS sequences using the GSDS online tool. The results showed that most P-ATPase genes contained introns and exons, and most P-ATPase genes in G. raimondii and G. barbadense also contained untranslated regions (UTRs). Almost all P 3A subfamily genes contained 10-20 introns, the P 1B subfamily genes contained 5-20 introns, and most P 2A subfamily genes contained 5-10 introns, but some genes contained more than 20 introns. Interestingly, the structures of the P 1B subfamily genes on different branches had significant differences. Most genes in class C did not contain introns or contained only one, and most genes in class A contained 5-7 introns, but all genes in class B contained more than 30 introns. In the phylogenetic tree, the gene structures of most paralogous gene pairs between G. hirsutum and G. arboretum/G. raimondii were very similar, which suggests that their genetic relationship is close, and these paralogous genes had high homology. However, the gene structures of paralogous gene pairs between G. barbadense and G. arboretum/G. raimondii were not close, which indicates that G. hirsutum has a closer genetic relationship to G. arboretum/G. raimondii than G. barbadense. different cotton species. Twenty-one pairs of orthologous genes were identified between G. hirsutum and G. arboretum, and 16 pairs of orthologous genes were identified between G. hirsutum and G. raimondii. The results showed significant collinear relationships between G. hirsutum and G. arboretum or G. raimondii, and G. hirsutum was more closely related to G. arboretum than G. raimondiiand . In addition, the remaining genes in G. hirsutum do not have duplicative relationships with those in G. arboretum and G. raimondii, indicating that independent duplications might occur in G. hirsutum. To further analyze the evolutionary relationship of P-ATPase genes in G. hirsutum, G. arboretum and G. raimondii, the collinearity of P-ATPase genes was drawn, using Circos software, between different cotton species. Twenty-one pairs of orthologous genes were identified between G. hirsutum and G. arboretum, and 16 pairs of orthologous genes were identified between G. hirsutum and G. raimondii. The results showed significant collinear relationships between G. hirsutum and G. arboretum or G. raimondii, and G. hirsutum was more closely related to G. arboretum than G. raimondiiand . In addition, the remaining genes in G. hirsutum do not have duplicative relationships with those in G. arboretum and G. raimondii, indicating that independent duplications might occur in G. hirsutum. arboreum, and G. raimondii. Red, blue and green rectangles represent chromosomes of three cotton species. Blue, orange and gray lines represent genetic duplication in three cotton species themselves, and green lines indicate genetic duplication between different cotton species.
## Expression patterns of p-atpase genes in g. hirsutum
P-ATPase plays a variety of roles in plant cells. Thus, to realize the function of P-ATPase genes in G. hirsutum, the expression of P-ATPase genes in the different tissues (leaf, petal, pistil, root, stamen, stem and torus) and fibers was analyzed at various developmental stages (5DPA,10 DPA, 20 DPA and 25 DPA). Among the 98 genes, 84 P-ATPase genes were at least expressed in one tissue or at one developmental stage of fiber because the value of log2 FPKM was more than 1. The heat map was plotted using R software. The results indicated that 84 P-ATPase genes might play a role at various growth stages of cotton, and the remaining 14 genes might be expressed in a special environment or were only pseudogenes. More than one-third of the 84 P-ATPase genes were widely expressed in the various tissues of cotton and at three developmental stages of cotton fiber, which suggested that these genes may participate in the whole growth and fiber development of cotton. In addition, some genes were highly expressed in the petal and pistil or stamen, suggesting that these genes may play a role in flower development. A few genes had higher expression at the three developmental stages of cotton fibers compared to other tissues, indicating that these genes are involved in the regulation of cotton fiber development. In the stem, leaf, and root, a few genes were highly expressed, showing that these genes may be involved in stress resistance. Interestingly, seven P-ATPase genes (GhAHA1, 2,were only expressed in the stamen, suggesting that these genes are associated with a specific role in the development of the cotton stamenand . A significantly different expression pattern in P-ATPase members in the same subfamily was observed, signifying that these genes performed functional differentiation, while a similar expression pattern of some genes from different subfamilies was observed, which showed that these genes may have a kind of synergy.
## Expression patterns of p-atpase genes in g. hirsutum
P-ATPase plays a variety of roles in plant cells. Thus, to realize the function of P-ATPase genes in G. hirsutum, the expression of P-ATPase genes in the different tissues (leaf, petal, pistil, root, stamen, stem and torus) and fibers was analyzed at various developmental stages (5 DPA,10 DPA, 20 DPA and 25 DPA). Among the 98 genes, 84 P-ATPase genes were at least expressed in one tissue or at one developmental stage of fiber because the value of log 2 FPKM was more than 1. The heat map was plotted using R software. The results indicated that 84 P-ATPase genes might play a role at various growth stages of cotton, and the remaining 14 genes might be expressed in a special environment or were only pseudogenes. More than one-third of the 84 P-ATPase genes were widely expressed in the various tissues of cotton and at three developmental stages of cotton fiber, which suggested that these genes may participate in the whole growth and fiber development of cotton. In addition, some genes were highly expressed in the petal and pistil or stamen, suggesting that these genes may play a role in flower development. A few genes had higher expression at the three developmental stages of cotton fibers compared to other tissues, indicating that these genes are involved in the regulation of cotton fiber development. In the stem, leaf, and root, a few genes were highly expressed, showing that these genes may be involved in stress resistance. Interestingly, seven P-ATPase genes (GhAHA1, 2, 4-8) were only expressed in the stamen, suggesting that these genes are associated with a specific role in the development of the cotton stamenand . A significantly different expression pattern in P-ATPase members in the same subfamily was observed, signifying that these genes performed functional differentiation, while a similar expression pattern of some genes from different subfamilies was observed, which showed that these genes may have a kind of synergy.
## Accumulation of pas and expression analysis of p3a-atpase genes in brown cotton.
Some P3A-ATPase genes were found in Arabidopsis and petunia, such as TT13 and PH5, which were related to the accumulation of PAs. Thus, we explored whether such genes are present in brown cotton. First, the contents of PAs at different development stages of brown cotton fiber were measured. As shown in, the contents of PAs increased gradually with the development of fiber, reached the highest level at 12 DPA, and thereafter decreased gradually. The accumulation of PAs showed a trend that increased initially and then decreased gradually back to baseline. To find the genes in the P3A subfamily associated with the accumulation of PAs in brown cotton fiber, primers were designed according to the sequences of P3A-ATPase genes, and qRT-PCR was performed with the fibers at 6, 12, 24, and 30 DPA in brown cotton, respectively. The results suggested that the relative expression levels of most P3A-ATPase genes were not consistent with the accumulation of PAs in brown cotton fibers. However, GhAHA5, GhAHA 7, GhAHA 9, GhAHA10 and GhAHA11 showed a
## Accumulation of pas and expression analysis of p 3a -atpase genes in brown cotton.
Some P 3A -ATPase genes were found in Arabidopsis and petunia, such as TT13 and PH5 [9,10], which were related to the accumulation of PAs. Thus, we explored whether such genes are present in brown cotton. First, the contents of PAs at different development stages of brown cotton fiber were measured. As shown in, the contents of PAs increased gradually with the development of fiber, reached the highest level at 12 DPA, and thereafter decreased gradually. The accumulation of PAs showed a trend that increased initially and then decreased gradually back to baseline. To find the genes in the P 3A subfamily associated with the accumulation of PAs in brown cotton fiber, primers were designed according to the sequences of P 3A -ATPase genes, and qRT-PCR was performed with the fibers at 6, 12, 24, and 30 DPA in brown cotton, respectively. The results suggested that the relative expression levels of most P 3A -ATPase genes were not consistent with the accumulation of PAs in brown cotton fibers. However, GhAHA5, GhAHA 7, GhAHA 9, GhAHA10 and GhAHA11 showed a relative expression tendency similar to the PAs' accumulation. Therefore, we hypothesized that one or more of these five genes may be related to the accumulation of PAs.
Molecules 2018, 23, x 8 of 14 relative expression tendency similar to the PAs' accumulation. Therefore, we hypothesized that one or more of these five genes may be related to the accumulation of PAs. relative expression tendency similar to the PAs' accumulation. Therefore, we hypothesized that one or more of these five genes may be related to the accumulation of PAs.
# Discussion
In recent years, many reports have described the structure and function P-ATPase genes, but a comprehensive study of their roles is still lacking in cotton. In this study, we identified and analyzed the P-ATPase gene family in G. arboretum, G. raimondii, G. barbadense and G. hirsutum at the genomic level. In all, 53, 51, 99 and 98 P-ATPase genes were identified in G. arboretum, G. raimondii, G. barbadense and G. hirsutum, respectively. The rich members and their high homology determined the diversity of expression and the specificity of function for the P-ATPase gene family. The homologous genes of various P-ATPases have been isolated from various plants. For example, AtECA1-4 can be expressed in all tissues of Arabidopsis. These genes play a role in the entire growth process of Arabidopsis, such as transporting calcium or manganese ions. In upland cotton, there were similar expression of P 2A genes, GhECA1, 2, 7, 8, 11, 12 were expressed in all tissues, they might play a similar function like AtECA1-4. In plants, some P-ATPase genes showed tissue-specific expression, suggesting they have specific function in a tissue. In Arabidopsis, the AHA3 is only expressed in phloem cells; the expression pattern of AHA2, AHA9 and AHA10 had well defined the interaction/role in root hair, anther and developing seeds, respectively; In rice, OsHMA3, OsHMA4 and OsHMA5 have high expression in root; In Hevea brasiliensis, HbHA2 and HbHA4 showed high transcript abundance in latex; in tomato, SlHA5, 6, 7, show tissue-specific pattern, its transcripts can only be detected in flowers; In Arabidopsis, ALA10 is only expressed in guard cells and root epidermal cells. These P-ATPase genes have been shown to play a role in specific tissues. The expression patterns of P-ATPase in upland cotton showed that some genes had obvious tissue-specific expression. For instance, GhAHA1, 2, 4-8 were expressed in stamen and GhAHA14, GhAHA15 had a specific expression in stem, suggesting these genes may play a role in stamen or stem, like P3A genes of Arabidopsis in different tissues. Some genes of other subfamilies also showed tissue-specific expression. In all, the tissue-specific expression indicates that the P-ATPase genes play multiple functions in upland cotton. found that three P-ATPase genes form different subfamilies and could be expressed in the same type of cells at the same development stage. Interestingly, the similar expression patterns in upland cotton were found, GhAHA2 and GhALA10 were only expressed in stamens, GhAHA21 and GhHMA1 showed similar expression patterns in different tissues, suggesting these genes may have similar or synergistic effects in specific tissue. In addition, the expression of some P-ATPase genes in specific tissues was also influenced by different developmental stages. For example, tobacco pma1 is mainly expressed in the guard cells of seedling cotyledons and stems, whereas the pma1 gene in fruit is only expressed in the phloem, and pma6 is mainly expressed in the vegetative phase of trichomes. In upland cotton, GhHMA12 and GhHMA13 showed a similar expression in developmental fiber, which had an expression level at 20 and 25 DPA, but little expression at 5 and 10 DPA. Upland cotton, as an important economic crop in the world, had complex gene regulation networks. Many P-ATPases have been shown to participate in regulating growth and physiological processes in different plants. The analysis of P-ATPases expression patterns in upland cotton showed that vast majority of P-ATPases had a certain function in cotton growth and development. Combining with the tissue-specific expression and functional analysis of P-ATPases in other plants, the potential functions of some P-ATPases in upland cotton could be inferred, such as regulating root and fiber development, and participating in some abiotic stresses response, etc.
Interestingly, in brown cotton fiber, the expression patterns of major P 3A -ATPase genes by qRT-PCR were found to be inconsistent with the expression pattern of P 3A -ATPase genes in upland cotton. The material used for fluorescence quantification experiments was brown cotton fiber, and white cotton was used for transcriptome analyses. We speculated that the differences in cotton varieties may lead to the changes in expression patterns of P 3A genes, and the difference in the environment may also be the cause of the changes.
In Arabidopsis and rice, the P-ATPase gene family can be divided into six subfamilies: P 1B , P 2B , P 2A , P 3A , P 4 , and P 5 .. Phylogenetic analysis showed that the P-ATPases genes of four cotton species could be also clustered into the same six subfamilies. There was only one or two P 5 members in four cotton species, which were similar with Arabidopsis and rice.believed that there is a relationship between intron structure and gene evolution. There are two different hypotheses about the origin of introns. One posits that the gene encoding the protein is initially discontinuous, namely that introns were present at the beginning, before the gene encodes the protein; and the other speculation is that the gene encoding the protein is continuous, and thus, there is no intron while the gene encodes the protein.showed that the original genes had few introns, and with the ongoing replication of genes, intron number gradually increased. The study also indicated that some introns were already present, and others were generated or increased with the insertion of transposons during gene duplication. By comparing the exon and intron distributions of P-ATPase genes in diploid and tetraploid cotton genotypes, most P-ATPase genes on the same branch were found to have a nearly identical structure, which suggested that the evolution of cotton species was highly conserved. In addition, the P-ATPase genes on different branches or the same branches in diploid and tetraploid cotton species were found to have some structural diversity, indicating that the structural diversity of the P-ATPase genes ensured that some new functional genes appeared in the evolution of the P-ATPase gene family.
In a previous study, it was suggested that proanthocyanidin is synthesized in the cytoplasm, transported into vacuoles by transporters, and is then accumulated in vacuoles. The P-ATPase genes play important roles during this process. First, H + -ATPase decomposes ATP and releases energy to generate and maintain an electrochemical gradient of H + on both sides of the cell membrane and to provide energy to transporters for the transmembrane transport of PAs. Second, H + -ATPase is used to adjust cell pH. A relatively stable pH can maintain cell activity and affect the activity of H + -ATPase. pH can affect the accumulation of PAs in the vacuole, whether it is too high or too low.showed that TT13 played an important role in the accumulation of PAs in the Arabidopsis seed coat. When TT13 is mutated, PAs will not accumulate in the vacuoles of the seed coat. Some researchers found that vacuole pH increased, and the petal color changed from red to purple color when the expression of PH5 was reduced in petunia. In the present study, it is speculated that P 3A -ATPase, the homologous gene to PH5, may also play an important role in the PAs biosynthesis of brown cotton fiber.analyzed the expression levels of five key genes (CHI, F3H, DFR, ANS and ANR) participating in PAs synthesis pathways at different developmental stages of brown cotton fiber and found that these genes have higher expression levels in the early stage of fiber development. After reaching the peak, the expression began to gradually decrease. This result is consistent with the accumulation of PAs during the development of brown cotton fibers. We speculate that if there are one or more P 3A genes involved in the proanthocyanidin synthesis pathway, they might have an expression trend similar to the accumulation of PAs. Fluorescence quantitative analysis showed that the trend of relative expression of GhAHA5, GhAHA7, GhAHA9, GhAHA10, and GhAHA11 in brown cotton fiber was similar to that for PAs accumulation, especially the changes in expression levels of GhAHA10 and GhAHA11 were basically consistent with changes in the accumulation of PAs. Therefore, it was speculated that one or more of these genes may affect the accumulation of PAs in cotton fiber. Other P 3A genes may also have certain functions in the development of fibers, such as GhAHA6 and GhAHA24, which had the highest expression at 18 DPA; they may play a role in the stage of fiber development.
# Materials and methods
## Genomic data sources
The genome data for Arabidopsis were downloaded from The Arabidopsis Information Resource (http://www.arabidopsis.org/). The genome data for four cotton species were obtained from their genomes' respective sequence websites: Gossypium raimondii from the Phytozome database (http: //www.phytozome.net/); Gossypium arboreum and Gossypium hirsutum from the Cotton Genome Project (http://cgp.genomics.org.cn); and Gossypium barbadense from the Cotton Crop Databases (http://cotton.cropdb.org).
## Sequence identification of p-atpase genes
First, we used DNATOOLS software to establish four local databases with the whole genome amino acid sequences of G. arboretum, G. raimondii, G. barbadense and G. hirsutum. Then, we used the amino acid sequence of Arabidopsis P-ATPase gene (AHA10)-specific domains as a query sequence to preliminarily identify P-ATPase genes in four local databases using DNAtools software (E-value < 0.001). These candidate genes were examined to determine whether they contained the P-ATPase domains using the PFAM (http://pfam.wustl.edu/hmmsearch.shtml)and SMART databases (http://smart.embl-heidelberg.de).
## Phylogenetic analysis of p-atpase genes
These P-ATPase sequences were aligned using the ClustalW toolof MEGA 5.0 software to remove repeat sequences, and the phylogenetic tree of Arabidopsis and four cotton species was constructed using the Neighbor-Joining method (Bootstrap = 1000) to analyze the evolutionary relationships of the P-ATPase gene family. The naming and classification of P-ATPase genes in cotton refers to Arabidopsis P-ATPase genes.
## Amino acid sequence analysis of cotton p-atpase genes
To analyze the basic properties and characteristics of cotton P-ATPase genes, the data for isoelectric point, protein molecular weight and amino acid length were obtained using the ExPAsy (http://www. expasy.org) online tools and the Cotton Functional Genomics Database (https://cottonfgd.org/). The basic structure of P-ATPase genes was analyzed using the GSDS online tool (http://gsds.cbi.pku. edu.cn).
## Chromosome localization and gene duplication of cotton p-atpase genes
The genome annotation information for G. arboretum, G. hirsutum and G. raimondii was downloaded from the Phytozome database (http://www.phytozome.net/) and the Cotton Genome Project (http://cgp.genomics.org.cn). The syntenic information was obtained using OrthoMCL (http://orthomcl.org/orthomcl/). The chromosome localizations and syntenic blocks were mapped using MapInspect (http://mapinspect.software.informer.com) and Circos software.
## Expression patterns of p-atpase in g. hirsutum
The RNA-seq data derived from the TM-1 transcriptome of the Cotton Functional Genomics Database (https://cottonfgd.org/) were used to analyze P-ATPase gene expression profiles in G. hirsutum, which included the expression analyses of P-ATPase genes in various tissues and fibers at different development stages. If the value of log 2 FPKM was more than 1, we considered the gene expressed, and if the value was less than or equal to 1, we considered it not expressed. R software was used to visualize the expression patterns with normalization method.
## Qrt-pcr of p 3a -atpase genes in g. hirsutum
The PCR primers were designed according to the sequences of P 3A -ATPase genes . The RNA of brown cotton fibers at 6 days post-anthesis (DPA), 12 DPA, 18 DPA, 24 DPA and 30 DPA was extracted with an RNA extraction kit (TransGen Biotech, Beijing, China) and used to perform qRT-PCR. The total volume was 20 µL, including 10 µL of SYBR ® Premix Ex Taq ™ II (2x), 2 µL of cDNA, and 0.8 µL of upstream and downstream primers. The reaction procedure was as follows: 50 - C for 2 min, 95 - C for 30 s, 40 cycles of 95 - C for 15 s, 60 - C for 20 s and 68 - C for 20 s, and 4 - C for storage. The UBQ7 gene was used as an internal reference gene. For each experiment, three biological replicates were executed, and the relative expression levels were calculated using the 2 − Ct method.
## Determination of pa content
According to methods from Ikegami, the soluble and insoluble PAs of brown cotton fibers at different developmental stages (6 DPA, 12 DPA, 18 DPA, 24 DPA, and 30 DPA) were extracted, and the content of PAs was determined by spectrophotometry according to the standard curve of catechins, which were used as controls. For each experiment, three biological replicates were executed.
# Conclusions
In the study, a total of 301 P-ATPase genes were identified according to the whole genome sequences of G. arboretum, G. raimondii, G. barbadense and G. hirsutum using TBlastN, PFAM, and ClustalW tools, and the isoelectric point and molecular weight of 301 genes were predicted. The phylogenetic and structural analyses of P-ATPase genes in four cotton species showed that most genes located in the same branch of the evolutionary tree had a similar structure, and the numbers of exons and introns contained in the genes were also similar. The expression pattern analysis of P-ATPase genes in G. hirsutum indicated that most genes participated in the growth and development of cotton. In addition, the qRT-PCR analyses suggested that some members of the P 3A subfamily in brown cotton might participate in the transport and accumulation of PAs. The function of P 3A -ATPase genes in transporting the pigment of brown cotton fiber was initially explored, laying the foundation for selecting a functional P-ATPase gene and breeding a new brown cotton variety.
Supplementary Materials: The following are available online. : Characteristics of 301 P-ATPase genes in G. raimondii, G. hirsutum, G. arboreum and G. barbadense. "At" and "Dt" represent Group A and Group D chromosomes in G. hirsutum, respectively, and "sca" indicates that there is no chromosomal location information for the gene. : Gene numbers of 5 P-ATPase subfamilies in Arabidopsis, G. raimondii, G. hirsutum, G. arboreum and G. barbadense. : Homologous analyses of P-ATPase genes between different cotton species. The RNA-seq data of P-ATPases expression patterns in upland cotton. : Primers used in qRT-PCR. |
Comparison Between Percutaneous Transhepatic Rigid Cholangioscopic Lithotripsy and Conventional Percutaneous Transhepatic Cholangioscopic Surgery for Hepatolithiasis Treatment
Background: Percutaneous transhepatic cholangioscopy (PTCS) is one option for treating hepatolithiasis without surgical resection. This approach can use conventional biliary drainage methods over a long period, but a shorter procedure needs to be evolved.Objective: To evaluate the short-term and the long-term therapeutic outcomes of percutaneous transhepatic cholangioscopic lithotripsy (PTCSL) in comparison with conventional PTCS.Methods: In this retrospective study, 118 patients with hepatolithiasis were enrolled who underwent treatment in our hospital between March 2007 and July 2014. About 67 of them received PTCSL and the remaining 51 patients received conventional PTCS. Preoperative data, surgical operation-related records, the postoperative therapeutic effect, and the long-term hepatolithiasis recurrence rate were collected for comparison between the 2 groups.Results:The age, sex, and surgical history were similar between the 2 groups, but there was a significant difference in the Child-Pugh score, with more grade 3 patients in the PTCS group (P = 0.002). However, the operation time, intraoperative blood infusion, and the blood loss were similar between the 2 groups. The final clearance ratio of calculus in the PTCSL group was significantly better than in the PTCS group after multivariate analysis (P = 0.021; OR = 0.201; 95% CI, 0.051-0.785). Calculus recurrence was 9% (PTCSL) and 22% (PTCS). The postoperative hospital stay was significantly shorter in the PTCSL group (P = 0.001; OR = 1.337; 95% CI, 1.132-1.58).Conclusions: PTCSL was a satisfactory therapeutic option for hepatolithiasis treatment, with less operation time and a superior long-term therapeutic effect compared with conventional PTCS.
H epatolithiasis is the presence of gallstones (calculus) in the biliary ducts of the liver, and is common in East Asia where it has been considered endemic, but rare in Western countries. [bib_ref] Gallstone disease: management of intrahepatic stones, Mori [/bib_ref] In China and other parts of East Asia, 20% to 45% of the patients who underwent surgical operations for gallstones had hepatolithiasis. [bib_ref] Hepatolithiasis: epidemiology and classification, Pausawasdi [/bib_ref] [bib_ref] Comparison of treatments for hepatolithiasis: hepatic resection versus cholangioscopic lithotomy, Otani [/bib_ref] [bib_ref] Hepatolithiasis in East Asia. Retrospective study, Nakayama [/bib_ref] [bib_ref] Relative prevalence of gallstone diseases in Taiwan. A nationwide cooperative study, Su [/bib_ref] Clinical progression of the disease may lead to liver parenchymal destruction due to recurrent cholangitis or subsequently result in biliary cirrhosis and even cholangiocarcinoma. [bib_ref] Laparoscopic hepatectomy with bile duct exploration for the treatment of hepatolithiasis: an..., Tian [/bib_ref] In contrast to gallbladder stones, hepatolithiasis is considered to be intractable, with frequent recurrence, [bib_ref] Gallstone disease: management of intrahepatic stones, Mori [/bib_ref] and patients with concomitant biliary stricture or residual stones are at a high risk of complications after treatment. [bib_ref] Risk factors for long-term outcomes after initial treatment in hepatolithiasis, Park [/bib_ref] Current treatment strategies include surgical resection of the affected lobe, and this is recommended in many cases. [bib_ref] Left-sided hepatic resection for hepatolithiasis: a longitudinal study of 110 patients, Shah [/bib_ref] [bib_ref] Hepatolithiasis: analysis of Japanese nationwide surveys over a period of 40 years, Suzuki [/bib_ref] Alternatively, in cases where resection is difficult, multiple lobes are affected, the patient has undergone previous biliary surgery, or has a high preoperative risk, nonresection methods can be used. These involve endoscopic retrograde cholangiopancreatography, peroral cholangioscopy, and percutaneous transhepatic cholangioscopy (PTCS). [bib_ref] Endoscopic or laparoscopic approach for hepatolithiasis in the era of endoscopy in..., Tan [/bib_ref] [bib_ref] Percutaneous transhepatic removal of bile duct stones: results of 261 patients, Ozcan [/bib_ref] [bib_ref] Ten-year longterm results after non-surgical management of hepatolithiasis, including cases with choledochoenterostomy, Tsuyuguchi [/bib_ref] PTCS is particularly useful when locations cannot be approached in a retrograde manner. [bib_ref] Gallstone disease: management of intrahepatic stones, Mori [/bib_ref] [bib_ref] Percutaneous transhepatic cholangioscopy: does its role still exist?, Choi [/bib_ref] The conventional PTCS approach is to expand the sinus gradually and perform percutaneous transhepatic biliary drainage. [bib_ref] Ultrasonically guided cholangiography and bile drainage, Makuuchi [/bib_ref] [bib_ref] Percutaneous transhepatic cholangioscopy in the treatment of complicated intrahepatic biliary strictures and..., Jeng [/bib_ref] This procedure has a relatively long operation time, usually between 2 and 3 weeks, with multiple expanding sheath sizes, leading to the potential for bleeding, biliary fistula, biliary tract infection, and peritonitis. Alternatively, the same approach can be used with lithotripsy to break up the calculus in percutaneous transhepatic cholangioscopic lithotripsy (PTCSL). [bib_ref] Intrahepatic stones: the percutaneous approach, Neuhaus [/bib_ref] Compared with conventional PTCS, PTCSL has the merits of minimal invasiveness, a shorter operation time, lesser intraoperative blood loss, and blood transfusion, all due to an optimized operation channel establishment and a shorter time period.
The objective of this study was to evaluate the shortterm and the long-term therapeutic outcomes of PTCSL, by comparison with that of conventional PTCS. The results should provide important information on which technique is more beneficial for the treatment of hepatic calculus.
# Materials and methods
The study was approved by the Ethics Committee of the First Affiliated Hospital of Guangzhou Medical University (Guangzhou province, China), and all the information obtained during the study was used only for scientific research.
## Patients
Patients with hepatolithiasis, who underwent treatment in the First Affiliated Hospital of Southern Medical University or in First People's Hospital of Shunde City affiliated to Guangzhou Medical University between March 2007 and July 2014, were enrolled.
Inclusion criteria were as follows: (1) patients diagnosed with complicated hepatolithiasis; (2) adult patients;
(3) patients with liver function preoperatively assessed as Child-Pugh A or B grades, or C grade without coagulation disorders; and (4) patients who underwent PTCS.
Exclusion criteria were as follows: (1) patients lost to follow-up; and (2) patients with other concurrent diseases diagnosed preoperatively, such as diabetes, hypertension, and cancer.
The diagnostic method and the standard for hepatolithiasis involved preoperative assessments including the B-ultrasound test, computed tomography (CT), magnetic resonance imaging (MRI), magnetic resonance cholangiography (MRCP), 3-dimensional digital image construction, endoscopic retrograde cholangiopancreatography, and PTCS, to determine the location, size, and number of gallstones, and the location of any stenosis.
The diagnostic criteria for complicated hepatolithiasis in this study combined multiple criteria including the Chinese guidelines of diagnosis and treatment for hepatolithiasis revised in 2011,the Japanese classification, [bib_ref] Gallstone disease: epidemiology, pathogenesis, and classification of biliary stones (common bile duct..., Tazuma [/bib_ref] and the Tsunoda classification 19 : (1) Location (L): the localization of the lithiasis or the bile duct lesion (distention or stenosis) was pinpointed accurately by the rebuilt CT images on hepatic fragments I to VIII. (2) Stenosis (S): the diameter of the secondary normal bile duct ranges between 2 and 3 cm; therefore, 2 cm was used as the standard to define stenosis of the bile duct. S0 was defined as no stenosis; S1 was relatively stenosed with a diameter ratio of the proximal bile duct stenosis and the distal bile duct distention >1/2; S2 was absolute stenosis with a diameter ratio of the proximal bile duct stenosis and distal bile duct distention <1/2. (3) Distention (D): the internal diameter of the bile duct of 10 mm was used as the criterion to distinguish between mild and severe distention; D0 was no distention; D1 was mild distention with an internal diameter between 2 and 10 mm; D2 was moderate distention with an internal diameter between 10 and 15 mm; and D3 was severe distention with an internal diameter larger than 15 mm. (4) Hepatic cirrhosis (C) and portal hypertension: portal hypertension and splenomegaly were the result of extensive hepatolithiasis, chronic hepatocholangeitis, and cholestatic cirrhosis, which were important prognostic factors of hepatolithiasis.
In the current study, if hepatolithiasis affected 2 or more hepatic lobe bile ducts and was complicated with S1, S2 or D2, D3, or C, it was considered to be complicated hepatolithiasis
The joint diagnosis was made and confirmed on the basis of images by the same 3 experts from the First Affiliated Hospital of Guangzhou Medical University: P.W. (13-y experience, from the Department of Hepatobiliary Surgery), Y.L. (33-y experience, from the Department of Hepatobiliary Surgery), and Yu Deng (13-y experience, from the Radiological Department).
Patients were divided into group PTCSL and group PTCS according to the treatment they received.
## The operative procedure
All procedures were performed at the Department of Hepatobiliary Surgery, the First Affiliated Hospital of Guangzhou Medical University.
## Ptcsl
The procedure was undertaken by P.W., B.S. (18-y experience), and B.H. (5-y experience). Instruments and equipments used during the procedure included a Nephrolithotome (12-degree Ultra-Wide-Angle Ureterorenoscope; Richard Wolf GmbH, Germany), which was used to replace a rigid choledochoscope, with a distal diameter of 2.8 mm. Other tools for lithotripsy used with nephrolithotomy included a ballistic lithoclast (Intracavity Lithotripsy Machine, type APL; Guangzhou Jielun Medical Equipment Co., Ltd., Guangdong Province, China) including electrohydraulic lithotripsy, ultrasonic lithotripsy, and a holmium laser. Other instruments were also utilized such as a C-arm x-ray machine (SIREMOBIL Compact L; Siemens, Germany), the light source (Voicecontrol compatible X8000 Xenon Light Source of 300 W; Stryker Corporation, MI), the camera system (1088i HD Camera Control Unit, PAL 220 V; Stryker Corporation), an adjustable pressure pump (type APL; Guangzhou Jielun Medical Equipment Co., Ltd, Guangdong Province, China), a zebra guidewire (type HAW; Cook Medical, Bloomington, IN), an expander of 8 to16 Fr series (type PLVW, Cook Medical), sheaths of 14, 16, and 18 Fr, and a biliary balloon dilator (ATB Advance; Cook Medical).
# The operation method
The puncture point for cholangioscopy entrance was selected carefully and determined according to preoperative CT and B-ultrasound images, usually at the lateral abdominal wall close to the caudal of the xiphoid process. A biliary drainage catheter of 8 Fr was placed through PTCS guided by B-ultrasound or a C-arm x-ray machine under epidural or intravenous anesthesia. The correct positioning for cholangioscopy was confirmed once the bile was drained out, and the puncture point was enlarged to 6 mm. The zebra guidewire was placed while the drainage catheter was removed. Biliary expanders from 8 to 16 Fr were used along the guidewire to expand the sinus gradually until it reached 16 or 18 Fr. Then a 16 or 18 Fr catheter covering the dilator was sent into the intrahepatic biliary duct together with the dilator. One end of the catheter remained in the duct and the other end was fixed to the skin after the dilator was removed, so that a fistulous channel from outside to the intrahepatic duct was established. Physiological saline was injected into the targeted biliary duct through the guidewire and the sheath of a rigid choledochoscope. Lithotripsy was performed once the calculus was found, and the shattered calculi were flushed out with physiological saline by "wash and suction" technology. The calculus could also be taken out with a basket (NTSE-045065-UDH; Cook Medical) or a clamp (Ureteral Grasping Forceps; Richard Wolf Medical Instruments Corporation, Vernon Hills, IL). Stenosis of the biliary duct was expanded with a series of soft expanders or a biliary balloon dilator, and the stricture could be cut open by an electric knife for columnar stenosis (tubular stricture), which is usually found in the scar stricture of the bile duct, before expansion (mannitol was injected into biliary tract before cutting it open). A drainage catheter of 14 Fr was indwelled beyond the stricture fragment, and the support catheter was indwelled for 8 to 11 months. All operational manipulations were performed within the sheath or biliary ducts, and the instruments did not come in contact with the wall of the sinus. The remnant calculus was evaluated by postoperative imaging examination, and a suitable withdrawal time of the indwelled drainage catheter was determined depending on the postoperative condition.
## Ptcs
The procedure was undertaken by Y.L., Tianling Fang (12-y experience), and J.X. (5-y experience).
The sinus was expanded gradually in a conventional PTCS approach, and a drainage catheter of 10 Fr was replaced by a 10 Fr or a 11 Fr PTCS catheter 1 week after percutaneous transhepatic biliary drainage. The sinus was further expanded using a catheter of larger diameter twice or thrice a week, until a 16 Fr catheter could be placed around 2 or 3 weeks later. Then, the calculus could be taken out by fibrocholangioscopy (Olympus, Japan) through a 16 Fr catheter.
## Outcomes
## Procedure assessments
Operation-related data were collected for both groups, including the operation time, the intraoperative blood loss, and intraoperative blood infusion. Other information was also recorded such as the immediate clearance ratio, the final clearance ratio, the postoperative hospitalization time, the stricture rate of the intrahepatic biliary duct, the mortality, the complication rate (including the recurrence of calculus and the cholangitis recurrence rate), and preoperative and postoperative laboratory tests (including the serum transaminase level, the serum bilirubin level, the serum albumin level, the hemoglobin level, and the prothrombin time).
## The immediate clearance ratio
An intraoperative choledochoscope instead of cholangiography was used as the routine examination to investigate the clearance of calculus.
## The final clearance ratio
Final imaging examination to confirm whether the stone was clean at the end of the therapeutic course.
## Recurrence
The biliary drainage tube was indwelt postoperatively for each patient as a routine procedure to facilitate choledochoscope examination or cholangiography. Cholangiography, ultrasound examination, or CT inspection were performed during the follow-up examination; if hepatolithiasis was found within 3 months after surgery, it was defined as remnant calculus; if hepatolithiasis recurrence or a cholangitis flare up was found 3 months after surgery, it was defined as a recurrence.
## Postoperative follow-up
Patients were followed up by telephone interview and doctor visits in the outpatient department every 3 months; all follow-up examination data were collected until July 2014. Examinations included ultrasound, CT scanning, MRI, hepatobiliary contrast imaging, and 3D reconstruction. The rate of remnant calculus, the recurrence time and the rate of calculus, the incidence of cholangitis, and the cause of death were also recorded.
# Statistical analysis
SPSS software (version 19.0; SPSS Inc., Chicago, IL) was utilized in the current study. Quantitative data are presented as mean ± SD. Numeric data are presented as the rate or the ratio. Comparisons between 2 groups were performed by the Student t test, the w 2 test, and the Fisher exact test for statistical analysis. Possible factors of postoperative outcomes were entered into a multiple logistic regression model, including age, sex, surgical history, the Child-Pugh score, DBIL, ALT, g-GGT, the immediate clearance ratio, the final clearance ratio, hospital stays, and the postoperative hospital stay. P < 0.05 was considered as statistically significant.
# Results
## Patient characteristics and preoperative data
About 67 out of 118 patients received PTCSL and were classified into the PTCSL group, including 35 men and 35 women with an average age of 55.2 ± 13.5 years and a disease course that ranged between 3 months and 10 years. The other 51 patients, including 24 men and 27 women, with an average age of 58.8 ± 16.5 years and a disease course that ranged between 4 months and 9.5 years, underwent conventional PTCS and were classified into the PTCS group.
The comparison of patient characteristics and preoperative data are shown in [fig_ref] TABLE 1: Comparison of Preoperative Data Between the 2 Groups HBsAg, surface antigen of... [/fig_ref]. There was no difference between the patients in terms of their age, sex, and surgical history. However, there was a significant difference in the Child-Pugh grade score as the majority of the patients in both groups were grade 2, but this was much higher in the PTCSL group than in the PTCS group (86.6% vs. 58.8%, respectively) and more patients were grade 3 in the PTCS group than in the PTCSL group (39.2% vs. 11.9%, respectively). There were also some significant differences in the preoperative laboratory tests and in the number of patients with 6 and 7 biliary strictures, which was significantly higher in the PTCSL group. Significantly more patients had right-lobe stones in the PTCSL group (70.1%) compared with the PTCS group (49.0%).
## Intraoperative data
Intraoperative data of the 2 groups are shown in [fig_ref] TABLE 2: Patient Outcomes *P < 0 [/fig_ref]. No significant differences were found between the groups in terms of the immediate clearance ratio, the operation time, intraoperative blood infusion, or the intraoperative blood loss.
## Outcomes
The postoperative hospital stay was similar in the 2 groups as shown in [fig_ref] TABLE 2: Patient Outcomes *P < 0 [/fig_ref]. The final clearance ratios and recurrence were also similar.
Both groups had no perioperative mortality. The complications are shown in [fig_ref] TABLE 2: Patient Outcomes *P < 0 [/fig_ref] and were similar for both groups, with 6 cases of cholangitis in the PTCSL group compared with 2 in the PTCS group and 1 case of postoperative intrahepatic biliary stricture in the PTCSL group.
Multivariate analysis revealed that the final clearance ratios and the postoperative hospital stay of the PTCSL group were significantly better than those for the PTCS group, even if DBIL was concerned, as shown in [fig_ref] TABLE 3: Multivariate Analysis GGT indicates g-glutamyl transpeptidase [/fig_ref].
# Follow-up results
All enrolled patients were followed up for a period that ranged between 2 and 85 months (mean 31.1 mo). There were 17 patients with recurrence of calculus as diagnosed by ultrasound, CT scan, or cholangiography: 6 patients from the PTCSL group and 11 patients from the PTCS group; however, there was no significant difference in terms of the recurrence rate, with 9.0% and 21.6%, respectively (P = 0.053). All recurrent calculi were located in the right lobe of the liver, and one of the recurrent patients from the PTCSL group had hepatolithiasis recurrence at the right posterior hepatobiliary duct. In the PTCSL group, there were 3 recurrences within the first postoperative year, and 1 recurrence in the second, third, and fifth years, respectively. Two recurrent calculi were located in the right posterior hepatobiliary duct, 1 located in the right anterior common hepatobiliary duct, 2 located in both the right and the left lobes of the liver, and another one located in the right caudate lobe.
The recurrence of hepatolithiasis in 2 patients was revealed by a follow-up CT scan, and the patients were reluctant to receive further treatment because they experienced no symptoms. Three patients with recurrent hepatolithiasis had concurrent symptoms and underwent left hemihepatectomy and plastic cholangiojejunostomy. Another patient received liver transplantation. There was 1 patient from the PTCSL group who had concurrent cholangiocarcinoma, and died from multiple organ dysfunction syndrome 13 months after cholangiojejunostomy. No other
# Discussion
PTCSL demonstrated its clinical value on hepatobiliary duct stenosis, calculi with hepatic segment atrophy, diffused intrahepatic biliary calculi, calculi located within biliary branches of grades III and IV, calculi concurrent with cirrhosis or portal hypertension, and for patients who were intolerant to partial hepatectomy. More importantly, the preoperative CT, MRCP, and B-ultrasound images provided the accurate entrance route for PTCSL, which enabled a thorough cholangioscopic lithotripsy, while protecting the hepatic function from damage. In the current study, PTCSL demonstrated an advantage over PTCS on lithotripsy of calculus in the right hepatic lobe, bilateral lobes, or in patients with hepatic function at Child-Pugh stage C and intolerant to conventional surgery.
A critical measure of the success of hepatolithiasis treatment is the number of patients who have recurrence as this can generally be high and lead to the risk of complications. [bib_ref] Risk factors for long-term outcomes after initial treatment in hepatolithiasis, Park [/bib_ref] Surgical resection is usually considered to be the optimal treatment for low recurrence rates of around 5% to 6%, 20,21 although these procedures often also involve other stone removal methods such as lithotripsy. The recurrence of calculus was reduced in the PTCSL group (9%) compared with the PTCS group (22%). Although the decrease in recurrence was not significant, this rate of recurrent calculus after PTCSL was also lower than that published in previous studies using PTCSL (29%, 3 20%, 22,23 18%, [bib_ref] Reappraisal of percutaneous transhepatic cholangioscopic lithotomy for primary hepatolithiasis, Chen [/bib_ref] 17%, . Those studies may have less success because methods have improved since they were published. We found that our method of individualized PTCSL could remove lesions thoroughly, get rid of calculi, and unblock stenosis. In the current study, the combination of a rigid choledochoscope and a protective sheath was utilized. The protective sheath was placed within the dilated fistulous tract, and all surgical operations were performed within the sheath and dilated biliary ducts; meanwhile, the biliary duct could be "straightened" by the sheath and formed an artificial channel to the outside. The shattered calculi were washed and sucked away by the rigid choloedochoscope from the sheath. Moreover, any stenosis or stricture of the biliary duct was corrected successfully by PTCSL at the same time, which can avoid multiple operations. Using CT, MRI, MRCP, and B-ultrasound, accurate locations of blood vessels and biliary ducts prevented accidental damage of the portal vein and hepatic vein during operation on a stenosis. The rigid percutaneous transhepatic choloedochoscope could reach the fragment of biliary duct stenosis from the dilated part, and as most of the stenoses were membrane type and short in length, they could be dilated directly by forceps. For severe stenosis, the biliary duct was dilated firstly by a balloon, and then guidewires with a series of diameters could be used if necessary. A stenosis with scar tissue could be opened by an electric or a laser knife, followed by a balloon for dilation; then, the placement of a supportive sheath helped avoid damage to blood vessels on the biliary duct wall and reduce hemobilia. Overall, our method of PTCSL reduced immediate and final remnant calculus in the PTCSL group effectively. The PTCSL group had a lower recurrence of cholangeitis compared with the PTCS group; however, it was not significant.
Although, unexpectedly, we found no significant difference in terms of the operation time between the 2 groups, there was a difference in postoperative hospitalization. The longer hospital stay for the patients in the PTCS group has the potential to cost more and indicate higher levels of pain and discomfort; although we did not measure the cost of treatment or pain experienced by the patients in this study, it would be worth considering in future studies.
Our method of percutaneous transhepatic operation through an expanded fistulous tract was performed under the guidance of CT or B-ultrasound. The puncture site at the dilated biliary duct with an optimal entrance angle was selected; the lateral abdominal wall close to the caudal of the xiphoid process was usually the primary entrance site; B2, B3a, and B3b, the dilated biliary ducts in the left lateral lobe, were usually used as puncture sites or the right midclavicular line was the entrance site, and the dilated biliary ducts B7a and B6c were used as puncture sites to avoid blood vessels, the intestine, and the chest cavity. The cholangioscopic lithotripsy was direct from the skin to targeted biliary ducts from an expanded fistulous tract. All PTCSL operations were successful in the current report. In our study, PTCSL shortened the whole lithotripsy period, the operation time, and the distance between skin entrance sites and targeted biliary ducts, although these differences were not significant in this population. Meanwhile, biliary duct dilation times and intraoperative blood losses were also slightly reduced, so that the minimal invasive treatment for cholangioscopic lithotripsy was realized by PTCSL.
There were some limitations in this study. First, this was a retrospective study of patients undergoing PTCS and PTCSL; hence, there was some inevitable selection bias between the groups that may impact the results. Second, the number of patients undergoing this treatment in this singlecenter study was quite small, and so the sample size was limited. These results should be confirmed in larger populations in multiple centers.
In summary, PTCSL demonstrated significant merits compared with PTCS, with reduced rates of immediate and final remnant calculus, a lower recurrence rate of hepatolithiasis, and a shorter postoperative hospitalization time, suggesting that PTCSL might be an ideal option for patients who are intolerant to conventional surgery or with postoperative remnant calculi.
[table] TABLE 1: Comparison of Preoperative Data Between the 2 Groups HBsAg, surface antigen of the hepatitis B virus; HCV, hepatitis C virus; HGB, hemoglobin; PT, prothrombin time; PTCS, percutaneous transhepatic cholangioscopy; PTCSL, percutaneous transhepatic cholangioscopic lithotripsy; TBIL, total bilirubin. [/table]
[table] TABLE 3: Multivariate Analysis GGT indicates g-glutamyl transpeptidase; ALT, alanine aminotransferase; CI, confidence interval; DBIL, direct bilirubin; OR, odds ratio. [/table]
[table] TABLE 2: Patient Outcomes *P < 0.05. PTCS indicates percutaneous transhepatic cholangioscopy; PTCSL, percutaneous transhepatic cholangioscopic lithotripsy. [/table]
|
Vaccinia Virus as a Master of Host Shutoff Induction: Targeting Processes of the Central Dogma and Beyond
The synthesis of host cell proteins is adversely inhibited in many virus infections, whereas viral proteins are efficiently synthesized. This phenomenon leads to the accumulation of viral proteins concurrently with a profound decline in global host protein synthesis, a phenomenon often termed "host shutoff". To induce host shutoff, a virus may target various steps of gene expression, as well as pre-and post-gene expression processes. During infection, vaccinia virus (VACV), the prototype poxvirus, targets all major processes of the central dogma of genetics, as well as pre-transcription and post-translation steps to hinder host cell protein production. In this article, we review the strategies used by VACV to induce host shutoff in the context of strategies employed by other viruses. We elaborate on how VACV induces host shutoff by targeting host cell DNA synthesis, RNA production and processing, mRNA translation, and protein degradation. We emphasize the topics on VACV's approaches toward modulating mRNA processing, stability, and translation during infection. Finally, we propose avenues for future investigations, which will facilitate our understanding of poxvirus biology, as well as fundamental cellular gene expression and regulation mechanisms.
## Introduction to host shutoff
Viruses are obligatory intracellular parasites that only replicate in host cells. Although viruses encode various numbers of genes to perform replication, they rely on cellular translation machinery for protein synthesis. In other words, competition exists between viruses and their hosts to utilize the limited amount of cellular translation machinery. This situation instigates many viruses to induce a rapid and profound decline in global host protein synthesis while continuously synthesizing the viral proteins, a phenomenon often termed "host shutoff". Viruses encode proteins that lead to host shutoff by affecting many cellular gene expression processes. Studies on virus-induced host shutoff to date have indicated that some viruses allocate proteins to inhibit DNA synthesis and transcription; more often, many viruses encode an arsenal of proteins targeting mRNA processing and translation. One of the most energy-consuming processes in the cell is mRNA translation. By usurping host mRNA translation, viruses can gain a competitive advantage over available energy and translational machinery, thereby enhancing the expression of viral mRNA to protein. Recently, it has also been evident that viruses can limit host cell proteins post-translationally by inducing rapid protein degradation. Because the targeted proteins include newly synthesized proteins, such as antiviral proteins, it also contributes to host shutoff.
DNA is the cellular genetic material encoding information that proceeds to transcripts and proteins. Some viruses have evolved means to modulate host DNA. showed that poxvirus infection induces a rapid decrease in host cell DNA synthesis. To decrease host cell DNA synthesis, VACV infection suppresses host nuclear DNA polymerase activity as early as 2 h postinfection. The studies measuring 14 C-or 3 H-labeled thymidine incorporation into the host and viral DNA during VACV infection observed an immediate decrease in labeled thymidine incorporation in host DNA. In contrast, labeled thymidine incorporation in viral DNA increased until 3 h postinfection, corresponding to the time when viral DNA replication occurs. Heat-inactivated and UV-irradiated non-infectious VACV could inhibit host DNA synthesis, indicating that protein(s) in the virion are responsible for the inhibition. However, the factors are not yet identified.
Some other viruses also decrease host DNA synthesis or even degrade the host cell DNA. Notable examples include the virion-associated protein of frog virus 3 (FV3), S1 gene of reovirus, ICP10 of herpes simplex virus type 2 (HSV-2), and a small segment at the 3' end of the vesicular stomatitis virus (VSV) genome inhibits host DNA synthesis via largely unknown mechanisms. Bacteriophage T4D-induced exo-and endodeoxynucleases degrade bacterial host DNA within 5 min. Summary of vaccinia virus (VACV)-induced host protein synthesis shutoff. VACV targets all the macromolecules in the central dogma of genetics, i.e., DNA, RNA, and proteins. VACV inhibits DNA replication (virion proteins), inhibits transcription (early protein(s)), interferes with mRNA processing, such as polyadenylation (VP55/VP39) and splicing (H1 phosphatase), and induces mRNA degradation (D9/D10). Further, VACV hinders 43S preinitiation complex formation (F17), decreases polysome bound mRNAs (surface tubular element), and inhibits translation (VACV 169). Additionally, VACV infection leads to the redistribution and post-translational modifications of translation initiation factors, which confers a translational advantage to viral mRNAs and a disadvantage to cellular mRNAs. Furthermore, VACV infection accelerates cellular protein degradation, including newly synthesized proteins. Abbreviations: HDAC 4/5, histone deacetylase 4/5; SR protein, serine (S)/arginine (R)-rich protein; eIF, eukaryotic translation initiation factor; eIF4A and eIF4E are abbreviated to 4A and 4E, respectively; tRNA, transfer RNA; tRNA_fMet, initiatormethionine tRNA; mRNA, messenger RNA; snRNA, small non-coding RNA; POLADS, VACV targets all the macromolecules in the central dogma of genetics, i.e., DNA, RNA, and proteins. VACV inhibits DNA replication (virion proteins), inhibits transcription (early protein(s)), interferes with mRNA processing, such as polyadenylation (VP55/VP39) and splicing (H1 phosphatase), and induces mRNA degradation (D9/D10). Further, VACV hinders 43S preinitiation complex formation (F17), decreases polysome bound mRNAs (surface tubular element), and inhibits translation (VACV 169). Additionally, VACV infection leads to the redistribution and post-translational modifications of translation initiation factors, which confers a translational advantage to viral mRNAs and a disadvantage to cellular mRNAs. Furthermore, VACV infection accelerates cellular protein degradation, including newly synthesized proteins. Abbreviations: HDAC 4/5, histone deacetylase 4/5; SR protein, serine (S)/arginine (R)-rich protein; eIF, eukaryotic translation initiation factor; eIF4A and eIF4E are abbreviated to 4A and 4E, respectively; tRNA, transfer RNA; tRNA_fMet, initiator-methionine tRNA; mRNA, messenger RNA; snRNA, small non-coding RNA; POLADS, polyadenylated short sequences; PABP, poly(A) binding protein; 40S, eukaryotic small ribosomal subunit.
## Prevention of cellular rna synthesis
Different classes of RNAs are transcribed by three DNA-dependent RNA polymerase enzymes in eukaryotic cells. Transcription by RNA polymerase I (RNAPI) synthesizes ribosomal RNAs (rRNAs). RNA polymerase II (RNAPII) yields messenger RNAs (mRNAs), and RNA polymerase III (RNAPIII) produces transfer RNAs (tRNAs), small non-coding RNAs (snRNAs), micro RNAs (miRNAs), 5S RNA (5S), and small nucleolar RNAs (snoRNAs). The prevention of cellular mRNA synthesis provides viral mRNAs with the advantage of accessing cellular translation machinery with fewer competitors. Many viruses target RNAPII enzymes to inhibit cellular mRNA synthesis, while other viruses use their RNA polymerases for viral mRNA transcription. In two previous studies, the rate of RNA synthesis was determined using 3 H-and 14 C-uridine uptake experiments. These experiments demonstrated that a 60% and 90% reduction in uridine uptake during VACV infection occurred at 3 and 9 h post-infection, respectively, compared to the uninfected condition. Unlike the inhibition of DNA polymerase activity, infectious VACV and the expression of early viral genes are required to inhibit RNAPII activity. Again, the mechanism of VACV-induced host transcription silencing remains unanswered, and further studies are needed to elucidate the viral and cellular mechanisms involved in this process. Interestingly, Teferi et al. showed that the viral K7 protein promotes histone methylation associated with heterochromatin formation, suggesting epigenetic modulation is involved in this process. Other epigenetic and genetic mechanisms may also lead to VACV-induced transcription silencing.Additionally, VACV infection induces a global degradation of host and viral mRNA, which is discussed in more detail in a later section.
VACV is not alone in the transcriptional silencing of host cells. Foot-and-mouth disease virus (FMDV) encodes 3C protease that cleaves histone H3, thus affecting the regulatory domain necessary for transcription. Influenza A virus (IAV) uses multiple strategies to diminish host transcription. In the early stages of infection, IAV inhibits RNAPII elongation, followed by RNAPII degradation in the later stages. IAV infection also induces stress that leads to the failure of RNAPII termination at the polyadenylation signal (PAS), which interferes with host mRNA maturation. A recent study indicated that the deregulation of RNAPII is dependent on the viral NS1 protein and can be augmented by the post-translational SUMOylation of an intrinsically disordered region of the IAV NS1 protein. Like IAV, herpes simplex virus type 1 (HSV-1) also induces the failure of RNAPII to terminate at PAS likely through the binding between the immediate early protein ICP27 and 3 processing factor CPSF to prevent the 3 cleavage of mRNA. Additionally, HSV-1 infection decreases RNAPII occupancy in two-thirds of cellular genes. In fact, HSV-1 proteins ICP4 and ICP22 are known to dysregulate RNAPII-mediated transcription initiation and elongation.
## Targeting mrna processing
Messenger RNA that codes for a protein is synthesized as pre-mRNA. Pre-mRNA undergoes multiple processing steps to form mature mRNA. VACV encodes proteins that modulate many of these steps, including capping, polyadenylation, splicing, and the export of mRNA from the nucleus to the cytoplasm.
## Capping
During transcription, when the length of cellular mRNA reaches 25-30 nucleotides, 7-methylguanosine cap (m 7 G) is added to the 5 end of the transcript, which not only protects the mRNA from 5 -3 exoribonucleases, such as Xrn1, it also helps the mRNA to be efficiently translated through the cap-dependent initiation mode. VACV mRNA capping is carried out in three reactions performed by viral enzymes. In this process, the newly synthesized transcript with a 5 triphosphate is cleaved to produce diphosphate by RNA triphosphatase. Guanosine monophosphate (GMP) is then added by RNA guanylyl-transferase and subsequently methylated at the N7 position by RNA (guanine-N7) methyltransferase. VACV encodes a heterodimeric capping enzyme complex consisting Pathogens 2020, 9, 400 5 of 23 of D1 and D12 proteins. Additionally, VACV encodes the VP39 protein (J3R), which adds a methyl group at the 2 -O position of the first transcribed nucleotide (cap 1) adjacent to the 5 cap (cap 0). This modification is present in higher eukaryotic mRNAs. However, apart from poxviruses, some other viruses, including HIV-1 (by host FTSJ3 protein), coronaviruses (by NSP16), and flaviviruses (by NS5 protein), have also been reported to contain 2 -O methylation in cap 1. The host's innate immune system recognizes hypomethylated viral RNA via melanoma differentiation-associated protein 5 (MDA5), retinoic acid-inducible gene I (RIG-1), and interferon-induced proteins with tetratricopeptide repeats 1 (IFIT-1) to distinguish between self and non-self mRNAs, thereby preventing non-self mRNA translation. Through methylation at the 2 -O position of cap 1, VACV may evade the host immune response and sustain its mRNA translation. Another strategy that VACV uses is to remove the 5 cap of mRNA using two decapping enzymes (D9 and D10). The decapped mRNAs are then degraded by 5 -3 exonucleases, such as Xrn1. The decapping process leading to mRNA degradation is discussed further in the mRNA degradation section.
Many viruses have developed strategies to remove the 5 cap to induce mRNA degradation. The influenza virus cleaves cellular mRNA 5 caps and uses the cleaved oligonucleotide as a primer for its transcripts through a mechanism known as "cap snatching". To perform cap snatching, polymerase basic protein 2 (PB-2) binds to the host mRNA cap through an extensive conformational change, followed by cleavage of the cap from host mRNA using the endonuclease in polymerase acidic protein (PA). PB-1 uses the cleaved cap as a primer for viral mRNA transcription. Hantavirus also performs cap snatching using its N protein, which binds to the cellular mRNA cap and rescues capped mRNA fragments from the processing-bodies' post-mRNA degradation. The capped mRNA fragment is further processed using viral RNA-dependent RNA polymerase, which is used as the primer for viral mRNA synthesis. Recently, the African swine fever virus (ASFV) was shown to encode a decapping protein, ASFV-DP, which removes the m 7 G cap from both the host and viral mRNAs. However, the physiological role of the decapping enzyme in ASFV infection is not elucidated yet.
## Polyadenylation
Polyadenylation is the process of adding a stretch of adenine bases at the 3 end of mRNA. It is an essential process for mRNA maturation that is tightly coupled with transcription termination. Polyadenylation protects mRNA from 3 -5 exoribonucleases and assists in the optimal translation of mRNAs. To protect viral mRNA and allow for optimal translation, VACV encodes a heterodimeric poly(A) polymerase complex of VP55/VP39 proteins early during infection that adds a poly(A) tail at the 3 end of viral mRNAs. The VP55 (E1L) protein functions as a catalytic component, whereas VP39 (J3R) protein acts as a processivity factor. During VACV infection, apart from catalyzing polyadenylation in viral mRNA VACV poly(A), viral polymerase polyadenylates cellular RNAs, such as tRNA, small nuclear RNA (snRNA), and mRNA, as well as small mRNAs that originated from the VACV genome to form non-translating polyadenylated short sequences (POLADS). The POLADs may affect host mRNA translation that will be discussed in detail in a later section. A separate interesting finding is that VACV VP55 protein polyadenylates host miRNAs, thereby leading to miRNA degradation. While the degradation of these miRNAs may abolish their effects on cellular and viral mRNAs, it is not clear how this mechanism may selectively affect host protein synthesis.
Other viruses have evolved means to affect mRNA stability by targeting the polyadenylation process. Nonstructural protein 1 (NS1a) of IAV restricts the polyadenylation of the nascent host mRNA by interacting with cleavage-polyadenylation specificity factor 30 (CPSF30). CPSFs are responsible for recognizing a polyadenylation signal in pre-mRNA and cleave the pre-mRNA, followed by the polyadenylation of upstream cleaved products. In contrast to IAV's strategy of blocking polyadenylation, Kaposi's sarcoma-associated herpesvirus (KSHV) hyper-adenylates host transcripts in the nucleus after nucleolar translocation of PABP due to the SOX-protein-induced reduction of cytoplasmic mRNA, which leads to nuclear retention and the decreased stability of transcripts.
## Splicing
A pre-mRNA contains coding exons and noncoding introns. During the maturation of pre-mRNA, the introns are removed, and exons are joined through a process called splicing. VACV genes lack introns; hence, splicing is dispensable for VACV. However, VACV can target host-splicing machinery to induce host shutoff. The serine (S)/arginine (R)-rich protein (SR protein) family is an important protein family that is required for spliceosome assembly. SR proteins are hyper-phosphorylated at a serine residue that is required for its function. VACV-encoded protein phosphatase H1 may inactivate SR protein, thereby dysregulating the host cell's RNA splicing machinery. The incubation of VACV H1 with SR protein from HeLa cell lysate de-phosphorylates SR protein, while the mechanism in VACV-infected cells has not been elucidated. Since splicing is not required for VACV, it likely selectively inhibits cellular mRNA splicing that contributes to host shutoff without interfering with viral mRNAs.
HSV-2 encodes the ICP27 protein that binds directly to pre-mRNAs and prevents their splicing. Unlike HSV-2, the human immunodeficiency virus 1 (HIV-1) Vpr protein binds to the spliceosome protein SAP145 to inhibit host pre-mRNA splicing. Recently, it was found that IAV endoribonuclease PA-X protein targets splicing machinery that leads to preferable degradation of cellular mRNAs with more splicing sites. However, the exact molecular mechanism is yet to be fully elucidated. Splicing blocking may lead to mRNA with a premature stop codon or disruption of the ORF, resulting in the production of aberrant proteins that are promptly degraded.
## Mrna export
Pre-mRNA processing occurs in the nucleus, after which, the mature mRNA must be exported to the cytoplasm such that translation can occur. Some viruses can target the export of host mRNAs by disrupting the nuclear-cytoplasmic transport machinery. Rhinovirus 2A and 3C proteases cleave vital constituents of nucleocytoplasmic mRNA export machinery. The IAV protein NS1 and VSV matrix protein downregulate and competitively bind the Nup98 protein, respectively, which is required for mRNA export. Additionally, IAV's NS1 protein interacts with and sequesters key mRNA export machinery proteins (p15, E1B-AP5, Rae1, and NXF1) to form an inhibitory complex, thereby inducing mRNA export blockage. During VACV infection, the nuclear pore complex is required for efficient VACV replication; however, whether this involves mRNA transport is not clear and is worthy of investigation.
## Induction of mrna degradation
VACV infection promptly induces host cellular RNA degradation. The mRNAs of cellular housekeeping genes, such as β-actin and α-tubulin mRNAs, are progressively degraded over time during VACV infection, and they are almost entirely degraded by 10 h post-infection. This process is highly efficient considering that the mammalian cellular mRNAs have an average half-life of approximately 6.9 h. Earlier studies from Rice et al. postulated that RNA degradation during VACV infection could occur due to several reasons: (1) the synthesis of viral RNases; (2) rapid turnover of mRNAs due to re-compartmentalization of mRNA in morphologically altered infected cells; and (3) interferon-induced 2 5 -oligoadenylate synthetase (OAS) through the activation of endonuclease RNase L, which does not discriminate between host and viral mRNA.
Although the aforementioned postulates could be true, subsequent findings have demonstrated that VACV encodes two decapping enzymes (D9R and D10R) that contain the Nudix hydrolases motif, which are likely the major driving force of mRNA degradation in VACV-infected cells. These decapping enzymes containing a Nudix motif that hydrolyzes a nucleoside diphosphate linked to any moiety X. Initial reports demonstrated that D9 and D10 negatively regulated gene expression, independent of the promoter used, suggesting they target a post-transcriptional step to decrease gene expression. However, the expression of a gene under an encephalomyocarditis (EMC) virus leader sequence that underwent 5 -cap-independent translation using the internal ribosome entry site (IRES) was not suppressed by D9 or D10 overexpression. Additional findings demonstrated that D9 and D10 lead to accelerated degradation of transcripts that are capped (m 7 GpppN) at the 5 end, which is a feature of both VACV and host cell mRNAs. In fact, by cleaving the 5 -cap, D9 and D10 render the mRNAs for degradation by the 5 -3 exonuclease Xrn1. The importance of D9 and D10 is further accentuated by the fact that a homolog of D10 is found in all chordopoxviruses, whereas the D9 homolog is found in most chordopoxviruses. We recently showed that the depletion of cellular mRNA is a significant contributor to VACV-induced host protein synthesis shutoff. Degradation of host mRNA leading to host shutoff is vital for VACV replication, as it not only usurps the host innate immune response but also reallocates the translation machinery to viral mRNAs. Interestingly, D9 and D10 likely also induce the rapid degradation of viral mRNAs. How VACV mitigates D9 and D10's effect to speed up viral mRNA turnover for viral protein synthesis is an interesting area we are actively pursuing. Another related open question is whether D9 and D10 target any specific mRNA population to induce degradation.
The induction of host mRNA degradation is a frequently used strategy by many viruses. The SARS-CoV NSP1 protein, IAV PA-X protein, HSV virion host shutoff (VHS) protein, Epstein-Barr virus (EBV) BGLF5 protein, and KSHV SOX protein have exo-or endoribonuclease activities and decrease global host protein synthesis by decimating host mRNAs. Interestingly, most of these ribonucleases could not distinguish between cellular and viral mRNAs. Other mechanisms are needed to cause a profound host shutoff while still keeping efficient viral protein synthesis.
## Usurping host mrna translation machinery
## An overview of eukaryotic mrna translation and diverse strategies employed by different viruses to inhibit host translation
All viruses rely on their infected host cells for mRNA translation as they do not encode genes for translation machinery. In many cases, viral and cellular mRNA translation represents a significant conflict of interest to compete for translation machinery. Not surprisingly, many viruses target translation processes to gain a translational advantage for viral mRNAs over cellular mRNAs. More efficient utilization of the translation machinery by viral mRNAs would ultimately put host mRNA translation at a disadvantage and contribute to host protein synthesis shutoff.
Messenger RNA translation is the most energy-consuming process in a cell. Cap-dependent translation is the dominant mode of eukaryotic mRNA translation with three major steps: initiation, elongation, and termination. In eukaryotic cells, the initiation factor 2 (eIF2) forms a ternary complex (TC) with initiator-methionine tRNA (Met-tRNAi) and GTP. Consequently, TC binds to the 40S ribosome in complex with eIF1, eIF3, and eIF5 to form the 43S pre-initiation complex. A rate-limiting translation initiation step is followed by the recruitment of a hetero-trimeric complex called eIF4F on mRNA. The eIF4F complex consists of an m 7 G cap-binding protein eIF4E, RNA helicase eIF4A, and scaffold protein eIF4G. Once formed, the eIF4F complex binds the m 7 G cap of the mRNA, and the scaffold protein eIF4G interacts with the poly(A) binding protein (PABP) bound to the 3 poly(A) tail to promote transient 5 -3 communication, which is known as the closed-loop model. Subsequently, by binding to the eIF3 complex, eIF4G helps to recruit the 43S pre-initiation complex on the mRNA to form the 48S pre-initiation complex. The 48S pre-initiation complex then scans the mRNA in the 5 →3 direction until it reaches the start codon, usually AUG. Once the start codon is recognized, eIF5B mediates the hydrolysis of eIF2-bound GTP. This change prompts the joining of the 60S ribosome to the 48S pre-initiation complex to form an elongation-competent 80S complex. This sophisticated process is followed via translation elongation, termination, and ribosome recycling. Since translation initiation is the rate-limiting step of translation, many viruses target different proteins involved in this process.
The multi-subunit complex eIF4F is a primary target for viruses to hijack host translation. All three proteins in the eIF4F complex are targeted by different viruses belonging to different families. FMDV protease 3C cleaves eIF4A to block host translation. HSV-1 VHS protein binds to eIF4A to gain Pathogens 2020, 9, 400 8 of 23 proximity to cleave and degrade host mRNAs. Enterovirus (EV) 2A, various retrovirus proteases, FMDV leader protease, and feline calicivirus 3C protease cleave eIF4G. Viral proteins, such as rhinovirus 2A and rotavirus NSP3, bind eIF4G and displace PABP to prevent the closed-loop conformation required for translation in many host mRNAs. Cap-binding protein eIF4E is a common target for many viruses to reduce host translation. Some viral proteins bind and recruit eIF4E to viral mRNA to gain a translation advantage. Notable examples include the Adenovirus shutoff protein 100K and turnip mosaic virus (TMV) VPg protein, while Cripavirus and influenza virus perform this function using unknown proteins. Enteroviruses (EV) use a distinctive approach to target eIF4E, by which virus-induced miR-141 suppresses eIF4E mRNA translation to limit the availability of this protein. Many viruses also target the phosphorylation state of translation repressor protein eIF4E binding protein 1 (4EBP1). 4EBP1 is a translation repressor that limits the availability of eIF4E to form the eIF4F complex. Hyperphosphorylated (four sites) 4EBP1 releases eIF4E for eIF4F complex formation. SV40 small t antigen carries out the PP2A-dependent dephosphorylation of 4EBP1, whereas VSV M protein and reovirus p17 dephosphorylate 4EBP1 through inactivating Akt-mTOR, which is required for 4EBP1 hyperphosphorylation. In doing so, viruses induce host shutoff since the vast majority of host mRNAs depend on eIF4F-complex-reliant cap-dependent translation.
Viruses are known to alter the function of multifactor initiation complexes other than eIF4F to induce host shutoff. Alphaviruses (Sindbis and Semliki Forest virus) induce phosphorylation of eIF2α to block global host translation. eIF3 is a multiprotein complex composed of 13 different subunits, several of which are known targets for viruses to decimate host mRNA translation. Measles N protein binds eIF3g, whereas rabies M protein binds eIF3h to impede host translation. Similarly, SARS-CoV and infectious bronchitis virus (IBV) spike protein binds eIF3f to impair host mRNA translation. Enteroviruses encode 3C proteases that cleave eIF5B, thus impairing translation by preventing eIF5B from interacting with eIF1A and the ribosome to accurately position met-tRNA on the start codon of an mRNA. Some viruses, such as FMDV, use multiple modes to induce host shutoff. In addition to cleaving eIF4A and eIF4G, FMDV infection induces the cleavage of eIF3a, eIF3b, and PABP.
PABP binds to the 3 poly(A) tail of an mRNA to enhance RNA stability, and this interaction is also vital for translation initiation complex formation in the cytoplasm. Many viruses target PABP to induce host shutoff. Calicivirus and enterovirus protease 3C or 3C-like protein cleaves PABP to inhibit its function. HIV-1 protease, Rubella capsid protein, and influenza NS1 bind PABP and suppress host translation. Rotavirus NSP3 displaces PABP from eIF4G and interacts with RoXaN to cause nuclear accumulation of PABP. The ORF57 (HSV-1 ICP27 homolog) and K10 proteins of KSHV and HSV-1 UL47 protein bind PABP to cause nuclear accumulation and abolish its function in the cytoplasm.
To efficiently produce viral proteins during the shutoff, many viruses have evolutionarily acquired cis-elements in mRNA to induce selective viral protein synthesis. The IRES is used by many RNA viruses, as well as some DNA viruses. Notable viruses that were suggested to be able to use IRES to mediate translation initiation via cap-independent mode to produce proteins include coronavirus, HCV, CSFV, HIV, and CrPV. Influenza B virus induces the combined translation of both M1 and BM2 protein via the base pairing of mRNA with 18S ribosomal RNA to promote the re-initiation of translation. Adenovirus uses ribosome shunting to enhance viral mRNA translation, where the tripartite leader in the non-coding region of viral late mRNAs exhibits high complementarity with 18S rRNA, which promotes the ribosome shunting mechanism. Viruses are notorious for recruiting translation initiation factors to their mRNAs to promote viral mRNA translation. Calicivirus VPg protein binds and recruits eIF3 and eIF4E to viral mRNAand HSV-1 ICP27 and UL47 proteins bind PABP. As such, the recruitment of initiation factors to viral mRNAs stimulates their translation. During replication, viruses produce dsRNA that is sensed by innate immune molecule PKR, leading to PKR and downstream eIF2α phosphorylation. When eIF2α is phosphorylated, global inhibition of translation initiation occurs. To avert this situation, viruses use various strategies to hinder PKR-mediated eIF2α-phosphorylation. Influenza NS1 and HCMV's two related proteins (TRS1 and IRS1) sequester dsRNA and prevent PKR phosphorylation. EBV SM and KSHV ORF57 directly sequester PKR to thwart PKR activation. HPV E6 and HSV ICP34.5 proteins regulate eIF2α phosphatase to dephosphorylate eiF2α. VACV encodes two PKR inhibitors, namely E3L and K3L. E3L binds to dsRNA to prevent PKR dimerization, whereas K3L, with its homology to eIF2α, acts as a pseudo-substrate for PKR. Many great details on how VACV, as well as other poxviruses, relaxes PKR-mediated translation inhibition have been revealed.
## Suppression of host cell translation during vacv infection
VACV uses multiple tactics to modulate cellular mRNA translation. Our transcriptome-wide analysis showed that VACV mRNAs have a higher translation efficiency than host mRNAs. During VACV infection, prompt inhibition of cellular protein synthesis occurs. Earlier findings have shown that VACV infection results in the inhibition of protein synthesis via a surface tubular element (STE) displayed on the VACV membrane. However, VACV STE did not affect either cellular RNA or DNA synthesis. When the authors exposed cells to purified STE, a decrease in the polyribosome occurred, accompanied by an increase in the free ribosome pool.
A remarkable factor present in virions implicated in inducing host shutoff is the phosphorylated 11 kDa protein, encoded by F17R. Purified 11 kDa/F17 protein from VACV virion and cell-free extract from VACV-infected cells prevents methionyl-tRNA_fMet-40S initiation complex formation, thereby inducing host shutoff. The importance of this small protein is further augmented by the finding that preventing the expression of F17 protein interrupts VACV morphogenesis. It was recently shown that F17 sequestering of Rictor and Raptor dysregulates mTOR to counter the antiviral response while retaining mTOR-mediated enhancement of viral protein synthesis. These findings indicate that F17 has multiple roles ranging from inducing host shutoff and countering the antiviral response to enhancing viral protein synthesis during infection.
VACV expresses a protein using its early gene VACWR169, which is confined in the host cell cytoplasm. In the cytoplasm, protein 169 impairs host protein synthesis, thereby facilitating the inhibition of host antiviral responses. Unlike other factors that induce host shutoff, protein 169 targets translation initiation by affecting both cap-dependent and cap-independent mechanisms, although how this protein manages to do it is yet to be elucidated. Protein 169 is not vital to VACV replication and spread in culture cells; however, it is required to regulate virulence as a VACV lacking protein 169 causes severe infections that induce stronger immune responses and are thus promptly cleared. By inducing host protein synthesis shutoff, VACV protein 169 suppresses host antiviral response and hence regulates virulence. Interestingly, full-length 169 is not encoded in all VACV strains; for example, a search of a VACWR169 homolog in the VACV Copenhagen strain reveals a premature stop-codon that would result in a truncated, possibly non-functional, protein product.
As discussed earlier, VACV-poly(A)-polymerase-induced POLADS may sequester PABP to make them inaccessible for host mRNA translation. In a cell-free protein-synthesizing system, the addition of POLADS inhibits cellular mRNA translation by up to 70%, while viral mRNA displayed minimal inhibition. The selective host protein synthesis inhibitory property of the POLADS is due to the poly(A) tail, as the increased length of the poly(A) tail in POLADS corresponds to increased inhibitory activity. Moreover, the addition of PABP reversed the host cell mRNA translation inhibition by POLADS, suggesting that PABP becomes the limiting factor that may be sequestered by these POLADS for host cell mRNA translation during VACV infection. These results also suggest that VACV mRNA translation is less dependent on PABP, which is in agreement with the finding that PABP1 is localized outside of viral factories, where viral mRNA translation is concentrated. However, the molecular mechanisms involved, as well as whether there is another poly(A)-binding protein that can substitute for PABP, are yet to be discovered.
Our genome-wide identification of VACV transcription start sites and polyadenylation sites revealed pervasive transcription initiation and termination. Most of them are not transcripts of the ≈200 annotated genes. The findings indicate a vast number of "dark" transcripts that are both capped and polyadenylated, which likely include some of the POLADS described above. These transcripts could be competitors of cap and poly(A)-tail-binding proteins necessary for cap-dependent mRNA translation. Because VACV post-replicative mRNA translation is less dependent on these factors, as shown by several groups, including us, these "dark transcripts" can place VACV mRNA at a translational advantage. This hypothesis is under active investigation in our laboratory.
Recently, antiviral granules (AVGs) and RNA granules were suggested to be present in VACV-infected cells. In fact, AVGs contain translation initiation proteins (eIF3h, eIF4E, and PABP), leading to speculation that such redistribution of translation initiation factors could be a mechanism used to limit the availability of these factors for host mRNA translation, although further investigation is needed.
## Preferential translation of vacv mrnas
A recent review by Meade et al. very nicely summarized how VACV infection modulates host translation machinery to selectively translate viral mRNAs. We will only give a brief overview of the strategies used by VACV to preferentially translate VACV mRNAs.
Several studies showed that VACV's post-replicative mRNA translation is enriched in or near virus replication sites called "viral factories," which are characterized by the intense staining of VACV viral DNA in the cytoplasm. It has been shown that translation initiation factors, such as eIF4E and eIF4G, could be recruited to the viral factories, likely enhancing cap-dependent translation initiation. Another study suggests that translation outside of the viral factories could also occur. VACV infection stimulates eIF4F complex formation. The eIF4E is repressed by hypo-phosphorylated eIF4E-binding proteins (4EBPs). In the early stages of infection, VACV induces surface integrin-β1-mediated PI3K activation, leading to hyperphosphorylation of 4EBP1 and the subsequent release of cap-binding protein eIF4E, which consequently augments the formation of the eIF4F complex enhancing VACV protein synthesis. Another poxvirus, namely myxoma virus (MYXV), activates AKT using the host range protein MT-5, although the role of AKT activation on mRNA translation during MYXV infection has not been studied yet. Additionally, VACV infection activates the MAPK/ERK pathway that stimulates the phosphorylation of eIF4E by MNK1. Phosphorylation of eIF4E at the serine 209 residue may lead to an increase in translation initiation of VACV mRNAs.
The poly(A) leader at the 5 end of transcripts is a unique feature of all VACV post-replicative mRNAs, which was discovered three decades ago. Only recently, we and others found that the 5 -poly(A) leader confers a selective translational advantage to viral post-replicative mRNAs, specifically in poxvirus-infected cells. Although the mechanism is still largely unknown, it was suggested that post-translational phosphorylation of small ribosomal protein RACK1 by VACV kinase B1 is necessary for the poly(A)-leader-mediated translational advantage. We found that the poly(A)-headed mRNAs can be efficiently translated in cells with impaired cap-dependent translation, suggesting that it is a cap-independent translation-enhancing element. Moreover, the 5 -poly(A) leader is not an IRES. An A-tract, omega prime, found in the tobacco mosaic virus (TMV) 5 untranslated region, could also enhance translation. The TMV omega prime enhances translation via promoting the recruitment of the eIF4F complex. Recently, it was found that mRNAs of yeast-virus-like elements contain a similar non-templated 5 -poly(A) leader, which also drives eIF4E-independent translation. It is of note that these VACV post-replicative mRNAs have 5 -caps. As discussed above, VACV modulates cap-dependent translation initiation factors and recruits them to viral replication sites for efficient viral mRNA translation. How the cap-independent translation element of the 5 -poly(A) leader coordinates with cap-dependent translation promotion to facilitate selective translation of viral mRNAs is an active area of investigation in our laboratory. We hypothesize that VACV can utilize the advantages of both cap-dependent and cap-independent translation modes to achieve a maximal translational advantage for viral mRNAs, which is facilitated by viral and cellular factors.
## Selective synthesis of viral and cellular proteins during vacv-induced host shutoff
Virus-induced global host shutoff is beneficial for viruses since it helps evade the host antiviral immune response. Additionally, shutoff also leads to the reapportioning of cellular machinery and critical host processes to confer a replication advantage to the virus. Although shutoff provides advantages to the virus regarding immune evasion and the reallocation of cellular resources, the selective synthesis of crucial cellular proteins and viral proteins is necessary to sustain virus replication. During VACV-induced host shutoff, viral and specific cellular proteins are selectively synthesized to drive efficient viral replication. The viral post-replicative mRNAs use an evolutionarily optimized 5 -poly(A) leader at the 5 -UTR region to mediate the translation advantage to VACV post-replicative mRNAs. However, during VACV infection, viral proteins augment the translational initiation to provide a translational advantage to the 5 -poly(A) leader bearing the VACV post-replicative mRNAs. Cellular proteins crucial for VACV replication are those of different complexes of oxidative phosphorylation (OXPHOS). The mRNAs encoding for proteins of OXPHOS have a shorter and less complex secondary structure bearing 5 -UTR, which provides a higher translation efficiency during VACV infection. These findings advanced our understanding of selectively synthesized proteins during poxvirus-induced host shutoff. We previously reviewed this topic in more detail.
## Induction of systematic host protein degradation
In addition to manipulating host gene expression at both the DNA and RNA levels, VACV also targets cellular proteins for degradation. A properly functioning ubiquitin-proteasome system has previously been shown to be necessary for viral replication. VACV infection in the presence of proteasome inhibitors significantly reduces the expression of post-replicative genes and fails to produce virus factories, but viral titers can be rescued after treatment removal.
Using a tandem-mass tag (TMT) labeling method, Soday et al. were recently able to determine that 265 host proteins are degraded during VACV infection, including histone deacetylase 5 (HDAC5) by VACV early protein C6, interferon-induced proteins with tetratricopeptide repeats (IFITs), tripartite motif (TRIM) proteins, and cell surface collagens, about 70% of which were determined to be targeted by the 26S proteasome via rescue with the proteasome inhibitor MG132. Histone deacetylases have previously been shown to play a role as a restriction factor in other viruses, including EBV, HCMV, and HPV, among others, C6 has been shown to inhibit interferon regulatory factors (IRF) 3 and 7 by interacting with proteins of the TANK binding complex (TBK1), as well as being responsible for the proteasomal degradation of HDAC4. Liu et al. also found that IFITs are targeted for degradation by a VACV ankyrin-repeat/F-box protein C9. The degradation likely helps VACV evade antiviral responses since IFITs and TRIMs are important players in a host's defense against viruses. There is still much work to be done to determine the VACV proteins responsible for the proteasomal degradation of the many other proteins identified in the above study.
# Conclusions and perspectives
During infection, VACV makes host cells conducive to viral replication through multiple measures, including inhibiting host cell death, altering cellular metabolism, and causing host shutoff, which is marked by global inhibition of host protein synthesis. To induce host shutoff, VACV employs wide-ranging strategies that target the key processes of gene expression, as well as pre-and post-gene expression processes: inhibiting host DNA synthesis and mRNA transcription, interfering with host mRNA processing and maturation, hijacking host translation machinery to preferentially synthesize viral proteins, and systematically promoting host protein degradation. A combination of these strategies leads to an expeditious, profound, and sustained host shutoff during VACV infection. Accelerated mRNA degradation and inhibition of cellular mRNA translation likely play a key role in the rapid induction of host shutoff, while host mRNA transcription and processing interference, cellular protein degradation, and cellular DNA synthesis inhibition can make the shutoff more profound and lasting during infection. Next, we briefly discuss several topics that we believe are important in this area, but of course, these are not exhaustive.
For most of the tactics VACV employs to induce host shutoff, the factors involved and their underlying molecular mechanisms are still largely unknown. What are the viral factors and how do these factors inhibit cellular DNA replication and transcription? Does VACV interfere with mRNA export? Likewise, how does VACV promote viral mRNA translation, and stimulate host protein degradation? What are the cellular mechanisms exploited by these viral factors? For those processes with identified viral factors responsible-for example, viral decapping enzymes that cause accelerated mRNA degradation, VACV 169 that causes translational repression, the poly(A) leader of viral mRNAs that confers translational advantage-the molecular mechanisms are largely elusive. How these different tactics cooperate during VACV replication to achieve the profound host shutoff is also a fruitful area for exploration. Additionally, while we know a great deal about VACV-induced host-shut shutoff, not much has been revealed in other poxviruses. It is conceivable that many of the strategies used by VACV are also used by other poxviruses. However, we may not assume all processes in VACV infection also occur in other poxvirus infections. In fact, it is of great interest to explore whether other poxviruses have evolved different genes/strategies to fulfill the mission of shutting off the host protein synthesis. Supporting this area of investigations, different poxviruses have evolutionarily acquired a broad range of different genes to address various cellular environments they face.
The viral decapping enzymes (D9 and D10)-induced mRNA degradation is probably the most well-studied host shutoff induction mechanism, with many papers published on this topic. However, with a greater understanding of this process, more intriguing questions arise that demand answers. The knockout of individual decapping enzymes or both enzymes presents various effects on VACV replication in cultured cells and infected animals. Interestingly, overexpression of these decapping enzymes also causes a decrease in VACV replication. A compelling question is how these decapping enzymes are regulated and coordinated to keep the balance for efficient viral replication. Our unpublished data indicate that the decapping enzymes are also necessary for highly efficient viral mRNA translation during infection. We are actively investigating the mechanism involved in this aspect, as well as the relationship between the induction of mRNA decay and stimulation of the mRNA translation. Another intriguing aspect is whether the decapping enzymes have any selectivity in inducing mRNA decay during infection, as previous studies showed that they could induce both cellular and viral mRNA decay. Supporting this possibility, in vitro studies have shown that D9 decapping activity is more susceptible to RNA inhibition, while D10 is more susceptible to an m 7 G cap inhibition.
The 5 -poly(A) leader bearing post replicative mRNAs are selectively translated during poxvirus-induced host shutoff. However, the exact mechanism and the host/viral factors involved in this process are still elusive. As mentioned above, we are investigating how the decapping enzymes may stimulate the translation of mRNAs with a poly(A) leader. Other outstanding questions include: Are there specific cellular factors facilitating 5 -poly(A)-mediated translation? Does the 5 -poly(A) leader have the intrinsic property of recruiting ribosomes with low (or no) requirement of eukaryotic translation initiation factors? Does the poly(A) leader facilitate viral mRNA escape from decapping enzyme-induced mRNA decay? It has been shown that the first A of the 5 -poly(A) leaders during VACV infection is N6-methyladenosine (m6A)-modified. As the m6A modification has been suggested in various aspects of mRNA degradation and mRNA translation, it is important to evaluate whether this modification facilitates a viral mRNA advantage during VACV-induced host shutoff. In fact, there is increased stability of m 6 Am RNAs due to resistance to the Dcp2 decapping enzyme in cells.
It has been over half of a century since the first few studies described the phenomena of host shutoff upon VACV infection . Since then, VACV has kept amazing us by revealing how many means it can deploy to achieve this goal. Studies on this subject have revealed several valuable molecular tools produced by VACV, for example, the virus-encoded decapping enzymes, mRNA translation modulators, and the 5 -poly(A) leaders on viral post-replicative mRNAs. However, many outstanding questions deserve further inquiries, which would not only facilitate understanding poxvirus replication but also provide invaluable tools to dissect many aspects of gene expression regulation and cell biology. Funding: This work was supported by a grant from the National Institutes of Health (R01AI143709, www.nih.gov) to Z.Y. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. |
Pathway-guided analysis identifies Myc-dependent alternative pre-mRNA splicing in aggressive prostate cancers
# Supplementary methods
## Gene ontology (go) analysis with background correction for expressed genes
The GO annotation was queried via the EnrichR API in R (1). A customized background gene list is required for the proper calculation of over-and under-representation of a GO term [bib_ref] Ontological analysis of gene expression data: current tools, limitations, and open problems, Khatri [/bib_ref]. For the alternative splicing analysis in this study, the background genes were selected by having sufficient coverage at splice junctions to meet the filtering criteria described above. With this customized background list, a corrected p-value can be computed using the hypergeometric test. The Benjamini-Hochberg procedure was used to control for the false discovery rate (FDR) at 5%. To reduce complexity, the resulting GO terms were required to contain at least 10 genes, with an exception for , where the minimum term size was increased to 100 to display the most representative terms. To visualize GO results, the REVIGO web server was employed with customized R plotting scripts for (3).
## Overlap enrichment assessment
Hypergeometric test p-value is used to measure the significance of the overlap between two groups of alternative splicing events. The triple intersection p-value is calculated by R package "SuperExactTest" based on hypergeometric test (4).
## Breast cancer and lung cancer myc-correlated alternative splicing analysis
The RNA-Seq processing framework described above was applied to quantify gene expression and alternative splicing of GTEx normal breast and lung samples, and TCGA BRCA and LUAD tumor-adjacent normal samples and tumor samples that are matched to tumor-adjacent normal samples. These datasets are de-identified. The Myc pathway-dependent splicing analysis was performed as described above.
## Lentiviral constructs
The myrAKT1 lentiviral vector has been described previously [bib_ref] The Sca-1 cell surface marker enriches for a prostateregenerating cell subpopulation that..., Xin [/bib_ref]. The inducible Myc lentiviral vector was cloned by inserting MYC into the BamHI site of the PSTV lentiviral backbone. Lentiviruses were prepared and titered as described [bib_ref] The Sca-1 cell surface marker enriches for a prostateregenerating cell subpopulation that..., Xin [/bib_ref].
## Organotypic human prostate transformation assay
This assay was conducted as previously described (6, 7) with de-identified human prostate samples. Doxycycline (1 ug/mL, Calbiochem 324385) was added to all culture media and renewed every 3 days.
## Xenograft outgrowth of transformed cells and cell line derivation
The xenograft and cell line derivation protocols have been previously described and were modified only to accommodate the doxycyline-inducible vector [bib_ref] Reprogramming normal human epithelial tissues to a common, lethal neuroendocrine cancer lineage, Park [/bib_ref]. Mice were fed sterile doxycycline chow (Bio-Serv S3888) continuously starting 3 days before xenograft implantation. Cell line initiation was performed on harvested tumors with the addition of 1 ug/mL doxycycline to all media.
## Cell line exon annotations
Exon annotations of known stop codons and the middle exon length were generated based on the same GENCODE gene annotation file used for alignment. Potential frameshift annotation is determined if the middle exon length cannot be divided by three. Potential RNA binding proteins were labeled according to the GO annotation term 'RNA binding'.
## Cell line propagation
The engineered cell lines were grown in stem cell media, composed of advanced DMEM/F12K (Gibco 12634028) base media with addition of B27 (Gibco 17504044), EGF (10 ng/mL, Peprotech 100-47), and FGF2 (10 ng/mL, Peprotech 100-18B) as well as Glutamax (Gibco 35050061). Doxycycline (1 ug/mL) was added to cultures to maintain MYC expression. Media was renewed every 3 days.
## Myc withdrawal experiments
Cells were collected by centrifugation and washed with media three times to remove doxycycline. 1 million cells were plated for each condition. Doxycycline was added back to the appropriate wells and then harvested at the appropriate time point (0-24 h).
## Histology
Portions of xenograft outgrowths were fixed in formalin overnight and transferred to 70% ethanol solution before submission for further processing by the Tissue Procurement Core Laboratory at UCLA (TPCL). Organoids were collected by dispase dissociation from Matrigel, washed three times with PBS, and then formalin-fixed for 30 min at room temperature. The fixed organoids were again collected by centrifugation and resuspended in HistoGel and submitted to TPCL. All samples were paraffin-embedded, sectioned at 4 µm, and mounted on glass slides. Hematoxylin and eosin staining was conducted according to standard protocols.
## Immunohistochemistry
Immunohistochemical studies were conducted as previously described (7). Briefly, unstained slides were subjected to deparaffinization, rehydration, and heat-activated citric acid antigen retrieval. Rehydrated slides were blocked with 1% horse serum in PBS before overnight incubation with primary antibodies also diluted in 1% horse serum/PBS. Primary and secondary antibodies and their dilutions are listed below. Antibody binding was detected using an HRPconjugated secondary antibody and a chromogenic substrate.
## Immunoblotting
Portions of tumor xenografts or 10 million cultured cells were placed in 8M urea lysis buffer with protease inhibitors (Sigma-Aldrich 4693159001) and homogenized with a Dounce apparatus. The lysate was cleared by ultracentrifugation at 30,000 x g for 30 min. Samples were denatured by boiling in SDS loading buffer under reducing conditions for 1 min and subjected to polyacrylamide gel electrophoresis. Wet transfer to nitrocellulose membrane was followed by blocking in 1% milk/0.1% Tween/PBS and overnight primary antibody incubation at 5 °C in the same buffer. HRPconjugated secondary antibodies were applied after washing and the blot visualized with a proluminescent substrate. Semi-quantitative blots of SRSF3 protein levels used PVDF membrane. Fluorescence levels were measured by Typhoon scanner and normalized to GAPDH levels. Antibody sources and dilutions are described below.
## Antibodies for flow cytometry, immunohistochemistry and immunoblotting
Antibodies used for flow cytometry were the fluorochrome conjugates CD49f-PE (12-0495-82; eBiosciences) and Trop2-APC (FAB650A; R&D Systems).
Primary antibodies used for immunohistochemistry included CK8 (1:1,000, Covance MMS-162P), AR (1:250, Santa Cruz sc-816), PSA (KLK3) (1:2000, Dako A0562), CK5 (1:1000, Covance PRB-160P), and p63 (1:250, Santa Cruz sc-8431). Secondary antibodies used were ImmPRESS anti-rabbit Ig peroxidase and anti-mouse Ig peroxidase (Vector Labs). Liquid DAB+ substrate reagent (Dako) was used to perform direct chromogenic visualization.
The following primary antibodies were used for immunoblotting (all at 1:1000 dilution, unless otherwise noted): Myc (Abcam ab32072), pan-AKT (Cell Signaling 4691), p53 (Cell Signaling Technology 2527), PARP1 (AbCam ab32138), cleaved PARP1 (AbCam, ab32064), anti-Cdk2 (AbCam ab32147), anti-Cdk2 (phospho Y15) (AbCam ab76146), p21 anti-p21 [EPR3993] (ab109199), and GAPDH (1:5,000, GeneTex GT239). HRP-conjugated goat anti-rabbit and goatanti-mouse secondary antibodies (BioRad) were used for luminescent detection. For semiquantitative Western blots, goat anti-mouse-cy5 (1:5000, Sigma-Aldrich GEPA45009) was used.
# Cell cycle analysis
One million cells were withdrawn from doxycycline as described above and harvested by centrifugation at the appropriate time-point. Cell pellets were washed three times with PBS and then singly dissociated with trypsin prior to fixation in 10% cold ethanol. After overnight fixation at 5° C, cells were pelleted and rehydrated in PBS. RNAse was added and the suspension incubated at room temperature for 4 h before staining with 20 ng/mL 7AAD and analysis by flow cytometry.
## Cell growth assay
Cells were washed with PBS, withdrawn from doxycycline, and plated at a density of 100,000 cells per well. Cells were lysed with CellTiterGlo luciferase reagent at the appropriate time and submitted for luminometry.
## Whole transcriptome sequencing analysis
Total RNA was isolated by guanidinium thiocyanate-phenol-chloroform extraction, followed by column clean-up. Isolated RNA was submitted for RNA integrity number (RIN) analysis. Only samples with RIN > 9 were carried forward. cDNA libraries were prepared from isolated RNA after poly-A selection using the TruSeq RNA Sample Prep Kit v2 (Illumina). High-throughput sequencing with 150 bp paired-end reads was performed using an Illumina HiSeq 2500. At least 100 million reads were collected for each sample.
## Cell line exon annotations
Exon annotations of known stop codons and the middle exon length were generated based on the same GENCODE gene annotation file used for alignment. Potential frameshift annotation is determined if the middle exon length cannot be divided by three. Potential RNA binding proteins were labeled according to the GO annotation term 'RNA binding'. . Comparison of count-based and ratio-based isoform-level analyses of prostate RNA-Seq datasets. (A) Unsupervised analysis of count-based isoform expression from a combined prostate cancer dataset (left panel). The same methodology applied to the ratio-based alternative splicing approach from
# Supplementary figures
[fig] Figure S2: Gene signature analysis identifies a common set of exons correlated with Myc, E2F, or mTOR pathways. (A) Violin plot depiction of gene signature scores of AR, Myc Targets V2, and mTOR sets across prostate cancer datasets. Dashed lines indicate averages across datasets profiling a disease phenotype (normal prostate, benign prostate, primary prostate cancer, mCRPC, and NEPC).(A) Fig. S3. Validation of Myc signature score and exon conservation across phylogeny and tumor type. (A) Box-and-whisker plot depiction of Myc signature scores in benign prostate tissues and primary prostate cancers stratified by Myc status. Samples with genomic amplifications of the Myc locus or single-gene overexpression are compared to samples without these alterations and adjacent benign tissues. (B) Kaplan-Meier disease-free survival plots of prostate cancers stratified by Myc signature score (first panel), Myc amplification status (second panel), or single-gene Myc expression (third panel). (C) Unsupervised two-way hierarchical clustering heatmap depiction of exon usage of 1,039 Myc-correlated exons across prostate cancer datasets in healthy tissue, and in primary, metastatic, and neuroendocrine prostate cancers. Columns depict patient samples. The Myc score annotation is colored from white (low) to black (high) based on the rank-transformed signature score of patient samples across the data sets. Rows represent exon inclusion events. Both are ordered by hierarchical clustering. (D) UCSC Genome Browser tracks depicting ultraconservation of Myc-regulated exons in SRSF3 (top panel) and HRAS (bottom panel) from humans to lamprey. (E) Box-and-whisker plot depiction of the Myc Targets V2 signature scores for breast and lung tissues. Left panel depicts normal breast (GTEx), tumor-adjacent normal breast (TCGA-BRCA), and breast adenocarcinomas (TCGA-BRCA). Right panel depicts normal lung (GTEx), tumor-adjacent normal lung (TCGA-LUAD), and lung adenocarcinomas (TCGA-LUAD). (F) Heatmap of Myc-correlated exons in the prostate meta-dataset alongside tissues from normal breast and lung as well as breast and lung adenocarcinomas. Dashed line indicates separation between two cancer types. The Myc score annotation is colored from white (low) to black (high) based on the rank-transformed signature of patient samples across the datasets. [/fig]
[fig] Figure S4: Establishment of engineered human tumor model with regulated Myc expression. (A) Representative scatterplot from florescence-activated cell sorting isolation of CD49f-high/Trop2high basal cells from total dissociated benign human prostate. (B) Florescent photomicrograph of doubly transduced prostate organoids as well as single and untransduced controls. "UT" = untreated, "C" = c-Myc transduction (GFP), "A" = myrAKT1 (RFP), "CA" = c-Myc and myrAKT1 (merge = yellow). (C) Photomicrograph of fixed organoids to show histology. Hematoxylin and eosin staining. (D) Immunohistochemical staining of transformed xenograft outgrowth compared to normal prostate tissue controls. [/fig]
[fig] Figure S5: Characterization of the response to Myc withdrawal in vitro. (A) Immunoblot of Myc expression levels in engineered cell line ICA1 in response to doxycycline titration. Data are representative of all three cell lines. (B) Growth response of ICA1 cell line in response to doxycycline titration as measured in a luciferase-based assay. (C) Stacked column chart depicting the change in cell cycle distribution over time after doxycycline withdrawal as measured by flow cytometry. [/fig]
[fig] Figure S6: Individual exon incorporation changes in response to Myc withdrawal. (A) Semiquantitative immunoblot of SRSF3 protein levels in response to Myc withdrawal for 24 h. Quantitation is the average reduction in SRSF3 levels measured in each cell line over the three independent replicates shown. [/fig]
|
Malignant Hyperthermia During Laparoscopic Adjustable Gastric Banding
Background:We report a rare case of malignant hyperthermia during laparoscopic adjustable gastric banding.Case Description: A 32-y-old female with no previous history of adverse reaction to general anesthesia underwent laparoscopic adjustable gastric banding. Intraoperative monitoring revealed a sharp increase in end-tidal carbon dioxide, autonomic instability, and metabolic and respiratory acidosis, along with other metabolic and biochemical derangements. She was diagnosed with malignant hyperthermia. Desflurane, the anesthetic agent was discontinued, and the patient was started on intravenous dantrolene.Results:The surgery was completed, and the patient was brought to the surgical intensive care unit for continued postoperative care. She developed muscle weakness and phlebitis that resolved prior to discharge.Conclusion:Prompt diagnosis and treatment of malignant hyperthermia leads to favorable clinical outcome. This clinical entity can occur in the bariatric population with the widely used desflurane. Bariatric surgeons and anesthesiologists alike must be aware of the early clinical signs of this rare, yet potentially fatal, complication.
# Introduction
Laparoscopic adjustable gastric banding (LAGB) is the least-invasive form of bariatric surgery. It is safe and effective. [bib_ref] Laparoscopic adjustable gastric banding in the treatment of obesity: a systematic literature..., Chapman [/bib_ref] [bib_ref] Laparoscopic bariatric surgery in super-obese patients (BMIϾ50) is safe and effective: a..., Parikh [/bib_ref] [bib_ref] Lap-band: outcomes and results, O'brien [/bib_ref] [bib_ref] Safety of the laparoscopic adjusted gastric band: 7-year data from a U.S...., Carelli [/bib_ref] [bib_ref] US experience with the LAP-BAND system, Ren [/bib_ref] [bib_ref] Favorable early results of gastric banding for morbid obesity: the American experience, Ren [/bib_ref] Moreover, it is increasingly being performed on an outpatient basis. Malignant hyperthermia (MH) is a rare but potentially life-threatening complication that can occur from administration of certain general anesthetics. [bib_ref] Malignant hyperthermia, Denborough [/bib_ref] [bib_ref] Malignant hyperthermia, Rosenberg [/bib_ref] [bib_ref] Malignant hyperthermia-associated diseases: state of the art uncertainty, Litman [/bib_ref] Its incidence ranges from 1:5000 to 1:50,000 to 100,000 administrations of general anesthetics. [bib_ref] Malignant hyperthermia, Rosenberg [/bib_ref] Since its introduction as an anesthetic agent, desflurane has been classified as a weak trigger of MH. [bib_ref] Delayed onset of malignant hyperthermia induced by isoflurane and desflurane compared with..., Wedel [/bib_ref] Its widespread use has changed the clinical presentation of MH. [bib_ref] Delayed onset of malignant hyperthermia induced by isoflurane and desflurane compared with..., Wedel [/bib_ref] Tachycardia and late onset presentations are some features of MH induced by desflurane. [bib_ref] Delayed onset of malignant hyperthermia induced by isoflurane and desflurane compared with..., Wedel [/bib_ref] [bib_ref] A suspected case of delayed onset malignant hyperthermia with desflurane anesthesia, Papadimos [/bib_ref] Nevertheless, its facilitation of rapid postoperative recovery, particularly in obese patients, makes it an ideal choice during bariatric surgery. [bib_ref] Effect of increased body mass index and anesthetic duration on recovery of..., Mckay [/bib_ref] [bib_ref] Faster washout and recovery for desflurane vs. sevoflurane in morbidly obese patients..., Colla [/bib_ref] [bib_ref] Desflurane versus sevoflurane for laparoscopic gastroplasty in morbidly obese patients, Vallejo [/bib_ref] Here we report a case of MH during LAGB, with desflurane as the maintenance anesthetic agent. Review of the literature reveals one publication in which a case of MH was described as a complication during LAGB. [bib_ref] Safety of the laparoscopic adjusted gastric band: 7-year data from a U.S...., Carelli [/bib_ref] The utilization of an anesthesiology team familiar with the biochemical and metabolic derangements associated with MH will permit timely diagnosis and treatment of the disorder, thus leading to better outcomes. Given the potential fatal outcome of MH, treatment with dantrolene should be initiated once the diagnosis is suspected. [bib_ref] Complications associated with the administration of dantrolene 1987 to 2006: a report..., Brandom [/bib_ref]
## Case report
A 32-y old, 96 kg female with a body mass index of 41 kg/m 2 , was evaluated at our institution for bariatric consultation. Her past medical history included asthma, hyperlipidemia, and infertility. Her past surgical history consisted of knee arthroscopy and dilation and curettage; both were performed while she was under general anesthesia, uneventfully. Preoperative evaluation revealed no personal history or family history of adverse reaction to anesthetics.
The patient underwent LAGB. After adequate preoxygenation, general anesthesia was induced with a combination of propofol (200 mg), fentanyl (100mcg), and ro-curonium (50 mg). The trachea was intubated and general anesthesia was maintained using a combination of oxygen and air (1:1), desflurane (5% to 6%), and fentanyl, as needed.
Towards the end of the laparoscopic portion of the operation (35 min from induction of general anesthetic), she experienced rapid physiologic changes consistent with MH. Her end-tidal carbon dioxide (ETCO2) increased from 32 mm Hg to 80 mm Hg within 5 min, and peaked at 115 mm Hg. The patient's heart rate increased from 70 to 80 beats per minute to sinus tachycardia at 129 beats per minute. Her blood pressure trended up from an average 110/50 mm to 162/60 mm Hg, and her temperature increased from 35.8°C to 37.5°C. Arterial blood gas was consistent with mixed acidosis; her arterial blood gas demonstrated a pH of 7.15, partial pressure of carbon dioxide (PCO2) of 49 mm Hg, and bicarbonate (HCO3-) of 17.3 mmol/L. Her urine was straw-colored with a red tinge. The volatile anesthetic was immediately discontinued, and 175 mg of intravenous dantrolene was administered; furthermore, minute volume and the fraction of inspired oxygen (FiO2) were increased. She immediately responded to the treatment. Her ETCO2 subsequently decreased to 31 mm Hg, and her pulse (129 to mid 80s), temperature (37.5°C to 36.0°C), and blood pressure returned to her baseline of 115/50 mm Hg [fig_ref] Figure 1: Intraoperative anesthetic course [/fig_ref]. The surgery was completed uneventfully, with a total operative time of 63 min.
Biochemically, the patient had elevated creatine kinase MB (CKMB) and creatine kinase of 84.2ng/mL and 22233u/L, respectively, which peaked at 91.1ng/mL and 49220u/L on the first postoperative day 1. Postoperative management with intravenous dantrolene (1 mg/kg every 6 h) was continued for an additional 36 h. The patient remained hemodynamically stable. Her acidosis resolved, and the remainder of her metabolic and biochemical derangements progressively improved to normal values. She developed muscle weakness, which improved with physical therapy. She also developed phlebitis of her forearm, where dantrolene was infused. The phlebitis was treated conservatively with arm elevation and warm compresses. The patient was discharged home on postoperative day 8 in good and stable condition.
# Discussion
We present a rare, but potentially life threatening, intraoperative complication of bariatric surgery. The patient's rapid and severe metabolic and biochemical derangements and their subsequent rapid reversal with administration of dantrolene are strongly suggestive of malignant hyperthermia. [bib_ref] Trends and outcomes of malignant hyperthermia in the United States, Rosero [/bib_ref] [bib_ref] Clinical presentation, treatment, and complications of malignant hyperthermia in North America from..., Larach [/bib_ref] Desflurane will likely remain in widespread use in the bariatric population given its quick offset and effectiveness, [bib_ref] Effect of increased body mass index and anesthetic duration on recovery of..., Mckay [/bib_ref] [bib_ref] Faster washout and recovery for desflurane vs. sevoflurane in morbidly obese patients..., Colla [/bib_ref] [bib_ref] Desflurane versus sevoflurane for laparoscopic gastroplasty in morbidly obese patients, Vallejo [/bib_ref] and its use is justified because it is a weak trigger of MH; moreover, other inhalation anesthetics, such as halothane, sevoflurane, isoflurane, and enflurane, are all potential triggers of MH. Additionally, succinylcholine, the depolarizing neuromuscular blocking agent, has been reported to be itself a trigger of MH. [bib_ref] Delayed onset of malignant hyperthermia induced by isoflurane and desflurane compared with..., Wedel [/bib_ref] [bib_ref] A suspected case of delayed onset malignant hyperthermia with desflurane anesthesia, Papadimos [/bib_ref] [bib_ref] Malignant hyperthermia: pharmacology of triggering, Hopkins [/bib_ref] Nevertheless, it is important to note that MH can be triggered up to 6 h after general anesthesia induction when desflurane is used. [bib_ref] A suspected case of delayed onset malignant hyperthermia with desflurane anesthesia, Papadimos [/bib_ref] In view of increasing same-day discharge after LAGB, particular emphasis should be placed on the potential for delayed presentation of this complication of general anesthesia and the potential severe adverse consequences that can occur if MH is not recognized and treated promptly.
The potentially fatal nature of MH mandates treatment as soon the diagnosis is suspected. The principles of treatment of MH start with immediate discontinuation of the trigger agent. The patient is then hyperventilated to decrease the ETCO2, which is then followed by administration of dantrolene. The initial dose of dantrolene is 2.5 mg/kg, and this dose is repeated as needed; the dose can be titrated up to 10 mg/kg until resolution of the patient's tachycardia and hypercarbia. [bib_ref] Clinical presentation, treatment, and complications of malignant hyperthermia in North America from..., Larach [/bib_ref] Dantrolene is then continued at 1 mg/kg every 4 to 8 h for 24 to 48 h. While dantrolene is being infused, hyperthermia must be controlled by all means available to the treating physician. Rapid cooling continues until a core body temperature of 38.5°C is reached. Continued care for the next 36 h to 48 h is then performed in the intensive care unit, where the patient is monitored and treated for arrhythmias, biochemical disturbances and metabolic disturbances. Urine output is increased to 2 mL/kg/hour with administration of fluids, mannitol, and furosemide as needed. [bib_ref] Malignant hyperthermia, Rosenberg [/bib_ref] Once the patient is stabilized with documented improvement of metabolic and biochemical derangements, attention to possible complications of dantrolene and ways to curtail their progression are warranted.
Our patient developed muscle weakness and phlebitis, the 2 most common complications of dantrolene administration. [bib_ref] Complications associated with the administration of dantrolene 1987 to 2006: a report..., Brandom [/bib_ref] Administration of dantrolene as well as aggressive fluid resuscitation was discontinued after 36 h once she demonstrated objective evidence of improvement. Although these maneuvers may have decreased the potential complications from dantrolene, continued administration of the drug through a central line could have decreased the chance of phlebitis development. [bib_ref] Complications associated with the administration of dantrolene 1987 to 2006: a report..., Brandom [/bib_ref]
# Conclusion
Malignant hyperthermia is a very rare phenomenon, but bariatric surgeons, anesthesiologists, and all medical personnel involved in perioperative care of bariatric patients should be aware of early signs of MH. Early recognition and prompt treatment reduce the morbidity and potential mortality of this serious complication. |
Prevalence and molecular characteristics of drug-resistant Mycobacterium tuberculosis in Beijing, China: 2006 versus 2012
Background: As the epidemic of MDR-TB and XDR-TB becomes increasingly severe, it is important to determine the clinical characteristics and molecular epidemiology of MDR-TB and XDR-TB. Recently, many studies have shown that clinical features and molecular characteristics of drug-resistant strains vary in different geographical areas, however, further information is needed to assess the dynamic evolution of drug-resistant TB. Comparative studies between different time periods are necessary to elucidate the development of drug-resistant TB. Results: A total of 255 and 537 strains were collected from Beijing Chest Hospital in 2006 and in 2012, respectively. Drug-resistance rates and mutations associated with resistance to first-line anti-tuberculosis (TB) drugs were compared. The overall rate of drug resistance among strains of TB in 2012 was 54.4 %, significantly higher than that in 2006 (34.9 %, P < 0.001). Rates of resistance to each first-line drug (isoniazid, rifampicin, streptomycin and ethambutol) and to second-line drug ofloxacin increased significantly from 2006 to 2012. The overall MDR rate also increased significantly from 2006 (14.9 %) to 2012 (27.0 %). The rate of MDR increased significantly between these two time periods in previously treated cases (P = 0.023) but not in new cases (P = 0.073), and the rate of XDR was similar in new cases at the two time periods, but was marginally higher in 2012 in previously treated cases (P = 0.056). Previous treatment was found to be a risk factor for drug-resistant TB, especially for MDR-TB. In addition, the proportion of drug resistant isolates in which katG, the mabA-inhA promoter, oxyR-ahpC intergenic region, rpoB, rpsL, and embB were mutated was similar in 2006 and 2012, however patterns of mutation in these loci were more diverse in 2012 compared to 2006. Conclusions: Our data suggests that the prevalence of drug resistant TB remains high in Beijing, China, and that increasing rates of resistance in M. tuberculosis to all anti-TB drugs should be considered when choosing an optimal anti-TB regimen. Moreover, acquired multi-drug resistance may play a primary role in the MDR-TB epidemic in Beijing, China. Consequently, this highlights the importance of an earlier start to effective and supervised treatment in order to reduce the burden of retreatment.
# Background
Although the incidence and mortality of tuberculosis (TB) have declined over the past decade, there were an estimated 9.0 million incident cases of TB and 1.5 million deaths in 2013 according to the World Health Organization. Two major factors, including the lethal association of HIV with active TB disease, and the worldwide dissemination of multidrug-resistant (MDR) and extremely drug-resistant (XDR) strains of Mycobacterium tuberculosis, contribute to the severe TB epidemic. Globally, an estimated 3.5 % of new cases and 20.5 % of previously treated cases have MDR-TB. In 2013 there were an estimated 480, 000 new cases of MDR-TB, and about 210, 000 associated deaths. China is one of the 27 MDR-TB high-burden countries. According to the 2007 national survey of drug-resistant TB in China, 5.7 % of new cases and 25.6 % of previously treated cases were diagnosed with MDR-TB. Based on these survey results, it is estimated that there are 120, 000 new cases of MDR-TB in China per year [bib_ref] National survey of drug-resistant tuberculosis in China, Zhao [/bib_ref]. In addition, XDR-TB cases were reported by over 100 countries in 2013, and up to 9.0 % of MDR-TB cases had XDR-TB in 2013. It is therefore urgent to control the epidemic of MDR-TB and XDR-TB.
In order to understand how drug-resistant TB develops and to find better ways to control MDR-TB and XDR-TB, it is essential to determine the clinical characteristics and molecular epidemiology of MDR-TB and XDR-TB. Recently, many studies have shown that clinical features and molecular characteristics of drugresistant strains vary in different geographical areas [bib_ref] Comparison between molecular epidemiology, geographical regions and drug resistance in Mycobacterium tuberculosis..., Ramazanzadeh [/bib_ref] [bib_ref] Molecular characteristics of rifampicin-and isoniazidresistant Mycobacterium tuberculosis isolates from the Russian Federation, Afanas'ev [/bib_ref] , such as, for example in different provinces in China [bib_ref] Molecular characterization of multidrug-and extensively drug-resistant Mycobacterium tuberculosis strains in Jiangxi, Yuan [/bib_ref] [bib_ref] Prevalence and molecular characteristics of drug-resistant Mycobacterium tuberculosis in Hunan, China, Zhao [/bib_ref] [bib_ref] Molecular characterization of multidrug-resistant Mycobacterium tuberculosis isolates from China, Zhao [/bib_ref] [bib_ref] Molecular characterization of multidrug-resistant Mycobacterium tuberculosis isolated from south-central in China, Yu [/bib_ref] [bib_ref] Molecular and phenotypic characterization of multidrug-resistant Mycobacterium tuberculosis isolates resistant to kanamycin,..., Zhang [/bib_ref]. These studies, however, have focused on the clinical and molecular characteristics of drugresistant M. tuberculosis strains from a single time period, and are thus unable to assess the dynamic evolution of drug-resistant TB. Comparative studies between different time periods are thus necessary to elucidate the development of drug-resistant TB.
To better understand changes in the clinical and molecular characteristics of M. tuberculosis isolates in Beijing, China, we analyzed all strains collected from TB inpatients admitted to Beijing Chest Hospital (tertiary TB referral hospital) over two time periods separated by six years . Clinical information, drugresistant phenotypes (including MDR and XDR phenotypes) and first-line drug resistance-associated mutations were compared between these two time periods.
# Results
## Demographic and clinical characteristics of enrolled subjects
Our purpose in this study was to evaluate changes in the clinical characteristics and molecular epidemiology of drug-resistant TB and thus assess its dynamic evolution. We thus performed a comparative study of M. tuberculosis isolates collected at the Beijing Chest Hospital over two time periods separated by 6 years. The MTB strain bank at Beijing Chest Hospital was first established in 2005, and complete datasets were available for each year from 2006 to 2012 when we initiated this project. We reasoned that a 6-year window would reveal possible changes in the rate of occurrence of drug-resistant TB, so chose to analyze all isolates collected from inpatients in the Beijing Chest Hospital from 2006 to 2012. A total of 792 isolates were selected, including 255 and 537 isolates from 2006 to 2012, respectively. The mean age of patients in 2006 and 2012 was 48.8 ± 19.5 years and 51.4 ± 19.6 years, and the male to female ratio was 2.7 and 2.9, respectively.
As shown in [fig_ref] Table 1: General demographic characteristics of patients enrolled in 2006 and 2012 [/fig_ref] , there were no significant differences in the clinical characteristics (including gender, age and treatment history) of each subgroup (any drug-resistant TB, MDR-TB and pan-susceptible TB cases) between cases in 2006 and 2012 (P > 0.05). Of note, however, the proportion of previously treated cases in 2012 (40.8 %) was significantly higher than that in 2006 (31.0 %, P = 0.008).
## Drug-resistance patterns differ in 2006 and 2012
The overall rate of resistance in the MTB strains examined to any drug was 54.4 % in 2012, significantly higher than that in 2006 (34.9 %, P < 0.001), indicating the severe and worsening situation of drug-resistance in TB in China. Furthermore, the rate of resistance to any drug in 2012 was higher than that in 2006 in both new and previously treated cases (P < 0.05).
The proportion of drug-resistant M. tuberculosis isolates in 2006 was compared with that in 2012 [fig_ref] Table 2: Comparison of drug susceptibility patterns between clinical M [/fig_ref]. The overall resistance level to each drug was significantly higher in 2012 (P < 0.05). Rates of resistance to rifampicin (RIF), streptomycin (STR), ethambutol (EMB) and ofloxacin (OFX) were significantly higher in new cases in 2012 (P < 0.05), while the proportion of isolates resistant to isoniazid (INH) were marginally higher (P = 0.054), and resistance to capreomycin (CAP) and amikacin (AMK) was not significantly changed (P > 0.05). The proportions of isolates resistant to INH, RIF, OFX and CAP in previously treated cases were significantly higher in 2012 (P < 0.05), but the proportions of isolates resistant to STR, EMB and AMK were not significantly changed (P > 0.05).
The overall rate of MDR-TB was 27.0 % in 2012, significantly higher than that in 2006 (14.9 %, P < 0.05). The proportion of MDR-TB in previously treated cases was significantly higher in 2012 compared to 2006 (P = 0.023), but there was no significant increase in new cases (P = 0.073). The overall pre-XDR rate was 13.6 % in 2012, again, significantly higher than that in 2006 (6.7 %, P = 0.004). However, the pre-XDR rate in 2012 did not increase significantly in either new cases (P = 0.165) or previously treated cases (P = 0.090) compared with 2006. The rate of XDR-TB also increased significantly from 2006 (2.4 %) to 2012 (6.9 %). While the rate of XDR was similar in new cases during the two time periods, it was marginally higher in 2012 in previously treated cases (P = 0.056).
## Factors associated with drug-resistant tb and mdr-tb
The risk factors associated with resistance to any drug and MDR-TB were analyzed based on pooled demographic data for all patients [fig_ref] Table 3: Factors associated with drug-resistant tuberculosis in all patients OR of 0 [/fig_ref]. By univariate analysis, age and treatment history were significantly associated with resistance to any drug and MDR-TB (P < 0.05). Multivariate analysis confirmed that age and treatment history were independently associated with resistance to any drug and MDR-TB (P < 0.05).
Patients older than 64 years had a significantly lower risk of developing drug resistance relative to patients in the younger age group (<25 years), with an adjusted OR of 0.52 (95 % CI: 0.31-0.86, P = 0.002). This risk decreased significantly in MDR-TB cases, with an adjusted
# Discussion
In this hospital-based study, the rate of resistance to any drug and the MDR rate of M. tuberculosis isolates was found to increase significantly from 2006 to 2012, reflecting the serious drug-resistant TB epidemic in China. Overall, the percentage of previously treated cases in 2012 was higher than that in 2006, suggesting that treatment of previously treated cases is still a big challenge in controlling the TB epidemic in China. Furthermore, the proportion of previously treated cases among the MDR-TB cases increased to 40.8 % in 2012 compared to that in 2006 (31.0 %). This implies that acquired multi-drug resistance may play an increasing role in the MDR-TB epidemic in China.
Among the new TB cases in this study, rates of resistance to each first-line drug in the 2006 isolate collection were consistent with those found in the National survey of drug-resistant TB conducted in 2007 [bib_ref] National survey of drug-resistant tuberculosis in China, Zhao [/bib_ref]. However, rates of resistance to each first-line drug were significantly increased in 2012. The rate of resistance to the second-line drug OFX also increased from 7.4 % in 2006 to 13.2 % in 2012 among new TB cases. The increasing rate of OFX resistance rate should thus be considered when choosing an optimal anti-TB regimen. As the new cases in this study had not received therapy or were in treatment for less than 1 month, the rate of drug resistance among new cases should reflect the transmission of drug-resistant TB.
Among previously treated cases examined in our study, rates of resistance of M. tuberculosis isolates to INH, RIF, OFX and CAP were also higher in the 2012 collection compared with the 2006 collection. Overall rates of drug-resistance to each first-line drug in 2006 were much higher than that in the National survey [bib_ref] National survey of drug-resistant tuberculosis in China, Zhao [/bib_ref]. This difference might be attributed to differences in sampling methods and the subjects targeted. For the national survey, isolates were selected from across 10 provinces of China using a cluster-randomized sampling method [bib_ref] National survey of drug-resistant tuberculosis in China, Zhao [/bib_ref]. In contrast, in this study, patients were recruited from one tertiary TB referral hospital, i.e., most subjects were hospitalized TB patients with relatively serious symptoms. Overall, the increasing rate of drugresistance in previously treated cases underlines the importance of standard treatment and the necessity of optimizing the treatment regimen according to the results of drug susceptibility testing.
The MDR and XDR rates in the 2006 isolates were similar to those found in the National survey [bib_ref] National survey of drug-resistant tuberculosis in China, Zhao [/bib_ref] among new cases, but were much higher for previously treated cases. This may also be due to differences in the patients targeted as discussed above. Moreover, the rates of MDR and XDR MTB isolates had increased in 2012 compared to 2006 both among new cases and previously treated cases, although the difference was not significant for new cases. This data suggests that acquired MDR might outweigh primary resistance in the MDR epidemic in China. This finding is different from that of Gao et al. [bib_ref] Transmission of Mycobacterium tuberculosis in China: A Population-Based Molecular Epidemiologic Study, Yang [/bib_ref] who found that it is recent transmission of M. tuberculosis including transmission of MDR strains, that contributes most highly to the TB epidemic in China. This difference in findings may be due to variation in regions sampled: Gao et al. collected strains from five counties within five different provinces [bib_ref] Transmission of Mycobacterium tuberculosis in China: A Population-Based Molecular Epidemiologic Study, Yang [/bib_ref] , while the study population examined here came from one hospital in Beijing. Previous studies have shown that drug-resistance rates vary in different provinces [bib_ref] Prevalence of tuberculosis drug resistance in 10 provinces of China, He [/bib_ref]. Pre-XDR rates increased significantly from 2006 to 2012. Since pre-XDR is an important step in the development of drug resistance from MDR to XDR, this finding presents an additional alarming indication of the worsening situation of XDR-TB in China. Previous treatment is a well-known risk factor for drug-resistant TB and MDR-TB, and the prevalence of MDR-TB can be up to 10 times higher after unsuccessful treatment [bib_ref] Global surveillance for antituberculosis-drug resistance, 1994-1997. World Health Organization-International Union against Tuberculosis..., Pablos-Mendez [/bib_ref]. In this sense, results obtained here are consistent with previous observations: the risk of suffering MDR-TB among previously treated cases was 32.64 times higher than that in new cases. The implementation of DOTS is thus still very important for effectively controlling drug-resistant TB and MDR-TB, especially with respect to supervising patients to complete the treatment. We also found that people older than 64 years of age had a lower risk of drug-resistant TB and MDR-TB. This is consistent with the conclusion of a systematic review of European studies which concluded that MDR-TB cases are more likely to occur in patients younger than 65 years of age [bib_ref] Risk factors for multidrug resistant tuberculosis in Europe: a systematic review, Faustini [/bib_ref]. The higher risk of getting MDR-TB in people under 65 years may be attributed to the use of RIF for anti-TB treatment from around 1965. TB cases in older patients are usually considered as relapse cases, and the infecting strains may be more ancient, and carry a lower risk of becoming resistant to RIF.
Between 2006 and 2012, there was more or less no difference in the molecular detection rate for first line drug-resistance in our study. It should be noted that mutations in embB were also found in EMB-susceptible isolates [bib_ref] Detection of embB306 mutations in ethambutol-susceptible clinical isolates of Mycobacterium tuberculosis from..., Mokrousov [/bib_ref]. Here, embB306 mutations were found in 4.5 % and 6.5 % of EMB-susceptible isolates in 2006 and 2012, respectively. This percentage is lower than previously reported in Russia (31.2 %) [bib_ref] Detection of embB306 mutations in ethambutol-susceptible clinical isolates of Mycobacterium tuberculosis from..., Mokrousov [/bib_ref] and Singapore (20.0 %) [bib_ref] Novel mutations within the embB gene in ethambutol-susceptible clinical isolates of Mycobacterium..., Lee [/bib_ref] , and may be due to different percentages of multi-drug resistance among EMB-susceptible isolates. Hazbon et al. collected 807 M. tuberculosis isolates from Colombia, Mexico, New York and Texas, and found that the association between embB306 mutations and resistance to increasing numbers of anti-TB drugs was significant in each region, suggesting the role of embB306 mutations in broad drug resistance [bib_ref] Role of embB codon 306 mutations in Mycobacterium tuberculosis revisited: a novel..., Hazbon [/bib_ref]. The frequency of the embB306 mutation in one setting may thus be influenced by the percentage of multi-drug resistant isolates. Accordingly, the embB306 mutation does not appear to be a reliable marker for predicting EMB resistance in Beijing, China.
Mutation rates at common loci in specific genes associated with drug-resistance were similar in 2006 and 2012. It has previously been shown that antibiotic resistance associated mutations can impair bacterial fitness [bib_ref] Physiological cost of rifampin resistance induced in vitro in Mycobacterium tuberculosis, Billington [/bib_ref] [bib_ref] Effect of rpoB mutations conferring rifampin resistance on fitness of Mycobacterium tuberculosis, Mariam [/bib_ref]. However, acquisition of compensatory mutations in drug resistant strains can restore their ability to survive. It is possible that the reason why the mutation rate at common loci in the drug-resistant isolates in our study was unchanged may be that mutations in these common loci are associated with other compensatory mutations that lead to lower fitness costs.
Our study, however, has some limitations. The isolates examined were collected from only one TB referral hospital in Beijing. Patients admitted to this hospital tend to be severe cases or to have received therapy in other hospitals but with poor effect. Thus the incidence of drugresistant TB may be overestimated, and may not reflect the average level of the whole country. Well-designed studies with a wide coverage of different regions in China should thus be conducted in the future.
# Conclusions
Results from this study indicate that the prevalence of drug resistant TB remains high in Beijing, China, and suggest that increasing rates of resistance in M. tuberculosis to all anti-TB drugs should be considered when choosing optimal anti-TB regimens for treatment. The rate of MDR and XDR in M. tuberculosis isolates was higher in 2012 compared to 2006, especially in isolates from previously-treated cases, suggesting that acquired multi-drug resistance may increasingly be playing a primary role in the MDR-TB epidemic in China. These findings highlight the importance of an earlier start on effective and supervised treatment in order to reduce the burden of retreatment.
# Methods
## M. tuberculosis isolates and drug susceptibility testing (dst)
A total of 255 and 537 M. tuberculosis isolates collected in 2006 and 2012, respectively, were obtained from Beijing Bio-Bank of clinical resources on Tuberculosis at Beijing Chest Hospital. All isolates were recovered from inpatients diagnosed with pulmonary TB. If several isolates had been recovered from the same patient at different time points, only the earliest isolate was included in this analysis. Clinical investigations were conducted in accordance with the principles expressed in the Declaration of Helsinki, and this study was approved by the Ethics Committee of Beijing Chest Hospital. Written informed consent was not obtained from patients as the data were analyzed anonymously.
DST was performed using the proportion method on Löwenstein-Jensen medium, according to WHO guidelines, with the following concentrations of anti-TB drugs: INH -0.2 μg/ml, RIF -50 μg/ml, STR -10 μg/ml, EMB -5.0 μg/ml, OFX -2.0 μg/ml, levofloxacin (LFX) -2.0 μg/ml, CAP -40 μg/ml, AMK -30 μg/ml. Strains were deemed to be resistant to a specific drug when the growth rate was ≥1 % that of the control. Both OFX and LFX susceptibility testing were performed, but as results showed that all LFX-resistant isolates were also resistant to OFX, LFX-resistance data were not included in the analysis.
## Data collection and definitions
Demographic and clinical information on enrolled patients, including gender, age, address and TB treatment history, were obtained from inpatients' medical records. New cases were TB patients who had never been treated with anti-TB drugs or that had been treated for less than 1 month. Previously treated cases were TB patients who had been treated with anti-TB drugs for 1 month or longer. MDR-TB was defined as resistance to at least INH and RIF. Although kanamycin resistance is included in the WHO definition of pre-XDR and XDR, this drug is rarely used to treat TB at the Beijing Chest Hospital. In this study, XDR-TB was therefore defined as resistance to INH and RIF plus OFX and at least one injectable second-line drug (CAP or AMK). Pre-XDR TB was defined as resistance to INH and RIF plus either OFX or a second-line injectable drug (CAP or AMK), but not both.
## Detection of drug resistance associated gene mutations
Loci associated with hot-spots of drug-resistance to first line anti-TB drugs (INH, RIF, STR and EMB), including katG, the mabA-inhA promoter, oxyR-ahpC intergenic region, rpoB RRDR (RIF-resistance-determining region), rpsL and embB were sequenced in this study. Genomic DNA was extracted from freshly cultured M. tuberculosis using a conventional cetyltrimethylammonium bromide (CTAB) method [bib_ref] Extraction of Mycobacterium tuberculosis DNA: a question of containment, Somerville [/bib_ref]. All the primers for amplification of target nucleotide positions and DNA sequencing are listed in [fig_ref] Table 5: Primers for PCR amplification and DNA sequencing [/fig_ref]. For each target gene, the volume of PCR mixture was 25 μL, containing 12 μL of 2× Taq Master Mix, 1 μL of forward and reverse primers (10 μΜ), 10 μL of distilled H 2 O and 1 μL of genomic DNA. For katG, rpoB RRDR, embB, and rpsL, the PCR program comprised an initial denaturation at 95°C for 3 min, followed by 35 cycles of 94°C for 45 s, 62°C for 45 s and 72°C for 35 s, and a final step of 72°C for 4 min. For the mabA-inhA promoter and oxyR-ahpC intergenic region, the PCR program comprised an initial denaturation at 95°C for 5 min, followed by 35 cycles of 94°C for 1 min, 62°C for 1 min and 72°C for 1 min, and a final step of 72°C for 4 min. All PCR products were sent for sequencing using primers which were the same as those used for PCR amplification. Sequencing data was aligned with the corresponding sequences of the M. tuberculosis H37Rv reference strain using BLASTn optimized for megablast on the National Center for Biotechnology Information website (http://blast.ncbi.nlm.nih.gov/Blast.cgi).
# Statistical analysis
Pearson chi-square tests or Fisher exact tests were used to compare drug-resistance rates between isolates collected in 2006 and in 2012. Univariate analysis of categorical variables was performed with the Pearson chi-square test or Fisher exact test as appropriate. Univariate and multivariate logistic regression analyses were used to analyze drug-resistanceassociated risk factors. Variables with a P value less than 0.05 in the univariate analysis were analysed further by multivariable logistic regression analysis. A two-sided P value of <0.05 was considered statistically significant. Statistical analyses were performed using SPSS statistics 19.0.
## Ethics approval and consent to participate
This study was approved by the Ethics Committee of Beijing Chest Hospital. Written informed consent was not obtained from patients as the data were analyzed anonymously.
## Availability of data and materials
All of the data are complete in this study, no supplementary data are attached.
[table] Table 1: General demographic characteristics of patients enrolled in 2006 and 2012 [/table]
[table] Table 2: Comparison of drug susceptibility patterns between clinical M. tuberculosis isolates in 2006 and in 2012 [/table]
[table] Table 3: Factors associated with drug-resistant tuberculosis in all patients OR of 0.12 (95 % CI: 0.05-0.27, P < 0.001). Previously treated cases were associated with a higher risk of developing drug resistance, with an adjusted OR of 7.11 (95 % CI: 5.09-9.92, P < 0.001), and the risk increased significantly in MDR-TB cases, with an adjusted OR of 32.64 (95 % CI: 19.40-54.92, P < 0.001). [/table]
[table] Table 4: Most frequently identified mutations within first-line drug-resistance associated loci among drug-resistant M. tuberculosis isolates [/table]
[table] Table 5: Primers for PCR amplification and DNA sequencing [/table]
|
Relatives Education And Coping Toolkit - REACT. Study protocol of a randomised controlled trial to assess the feasibility and effectiveness of a supported self management package for relatives of people with recent onset psychosis
Background: Mental health problems commonly begin in adolescence when the majority of people are living with family. This can be a frightening time for relatives who often have little knowledge of what is happening or how to manage it. The UK National Health Service has a commitment to support relatives in order to reduce their distress, but research studies have shown that this can lead to a better outcome for service users as well. Unfortunately, many relatives do not get the kind of support they need. We aim to evaluate the feasibility, acceptability and effectiveness of providing and supporting a Relatives' Education and Coping Toolkit (REACT) for relatives of people with recent onset psychosis.Methods:The study is a randomised control trial. Trial Registration for Current Controlled Trials ISRCTN69299093. Relatives of people receiving treatment from the Early Intervention Service for psychosis are randomly allocated to receive either Treatment As Usual (TAU) or TAU plus the REACT intervention. The main aims of the study are to: (i) determine the acceptability of a supported self-management intervention; (ii) determine preference for type of support; (iii) assess the feasibility of the design; (iv) identify the barriers and solutions to offering support for selfmanagement approaches within the NHS; (v) estimate the likely effect size of the impact of the intervention on outcome for relatives; (vi) gain detailed feedback about the barriers and solutions to using a self-management approach; (vii) describe the way in which the intervention is used. Outcomes will be assessed from baseline and at 6 month follow-up. Discussion: The intervention is compared to current treatment in a sample of participants highly representative of relatives in routine early intervention services across the UK. The intervention is protocolised, offered within routine practice by existing staff and extensive process data is being collected. Randomisation is independent; all assessments are made by blind raters. The limitations of the study are the lack of control over how the intervention is delivered, the short follow-up period, and the lack of assessment of service user outcomes. Despite these, the findings will inform future effectiveness trials and contribute to the growing evidence base for supported self-mangement interventions in mental health.
# Background
Psychosis affects approximately 1% of people and is the third most disabling condition worldwide . First episode commonly occurs in adolescence at which time it is estimated that 60-70% will be living with their family [bib_ref] Lifetime Prevalence, Demographic Risk Factors, and Diagnostic Validity of Nonaffective Psychosis as..., Kendler [/bib_ref]. The government recognises the very important role of relatives and is committed to providing them with appropriate support via NHS services. Family interventions (FIs) are effective in improving outcome for people with psychosis and their relatives. As an adjunct to pharmacotherapy, FIs reduce relapse and hospitalisation rates [bib_ref] Efficacy of psychological therapy in schizophrenia: Conclusions from meta-analyses, Pfammatter [/bib_ref] [bib_ref] Family intervention for schizophrenia, Pharoah [/bib_ref]. FIs generally focus on cognitive and behavioural techniques to modify appraisals that relatives hold about the behaviour of the person with psychosis and develop coping strategies. Research is limited in its focus on people with more chronic mental health difficulties, and lack of attention to outcomes for relatives. However, interventions that are well integrated into Early Intervention Services (EIS) show reductions in relatives' distress [bib_ref] The first episode of psychosis: the experience of relatives, Addington [/bib_ref] , and appraisals held by relatives at first episode of psychosis are significant predictors of important determinants of outcome [bib_ref] Expressed emotion at first-episode psychosis: investigating a carer appraisal model, Raune [/bib_ref].
Significant barriers still exist to the dissemination of effective interventions through NHS EIS. These include clinicians with high caseloads and lack of confidence and training in working with relatives [bib_ref] Training and disseminating family interventions for schizophrenia: developing family intervention skills with..., Fadden [/bib_ref]. As a result, relatives report significant negative impact on many areas of their life, and the risk of distress is even higher at first episode than at later stages [bib_ref] The psychological well-being of family members of individuals with schizophrenia, Martens [/bib_ref].
There is a clear need for an intervention that can be widely available to relatives, is easy to use, phase specific, recovery focussed, does not require extensive clinical resources, targets key appraisals and coping strategies and empowers relatives. Self-management interventions that have the flexibility to be used alongside other work and family commitments and augment other forms of support are ideally suited to meet the needs of relatives. Self-management refers to health technologies (written/ audio/video/computer/internet) to assist users to manage a particular health problem, with little or no professional input. They can be used as stand-alone interventions or as an adjunct to other forms of intervention. This is a rapidly growing area and a recent meta-analysis of studies evaluating such approaches for depression shows promising results [bib_ref] What makes self-help interventions effective in the management of depressive symptoms? Meta-analysis..., Gellatly [/bib_ref]. Greater effectiveness is associated with using a 'guided' model with low-level contact with a professional/paraprofessional, and a CBT (Cognitive Behaviour Therapy) rather than educational model [bib_ref] What makes self-help interventions effective in the management of depressive symptoms? Meta-analysis..., Gellatly [/bib_ref]. Self-management approaches can increase dissemination of evidence based interventions to large numbers of people, and foster empowerment. Although much self-help literature has been written for a wide range of mental health problems, there is little development in the area of psychosis and few high quality evaluations from which conclusions can be drawn regarding effectiveness. We are not aware of any studies which have systematically evaluated the use of a supported self-management approach for relatives of people with psychosis.
This paper describes the rationale and protocol for a randomised controlled trial in which relatives of people at first episode psychosis in EIS receive current treatment or current treatment plus the REACT supported selfmanagement toolkit. The main aims of the study are (i) to determine the acceptability of a supported selfmanagement intervention and outcome measures to relatives of people with recent onset psychosis; (ii) to determine preference for type of support (email/telephone); (iii) to assess the feasibility of the design as measured by rates of recruitment, retention, attendance and direct feedback from participants; (iv) to identify the barriers and solutions to offering support for self-management approaches for relatives of people with psychosis in the NHS; (v) to estimate the likely effect size of the impact of the intervention on outcome for relatives; (vi) to gain detailed feedback from relatives about the barriers and solutions to using a self-management approach; (vii) to describe the way in which the intervention is used. This is not intended as a 'definitive' randomised controlled trial and will not assess the impact on patient outcomes or cost effectiveness. Further funding will be sought to address these issues if this study supports the feasibility.
# Methods
This trial is conducted by a multidisciplinary team of researchers, clinicians, statistician and relatives based across academic institutions and NHS Trusts in the North West of England. The research team are responsible for the content of the intervention and have worked with a design company to produce the toolkit and website. The trial is supported by an independent Trial Steering Committee.
## Design
## Randomisation, treatment allocation and blinding
This is a stratified randomised controlled trial in which participants are allocated to receive either Treatment As Usual (TAU) or TAU plus the REACT intervention. Randomisation is done using permuted blocks within Trust with block sizes varying randomly, and is carried out by an independent Clinical Trials Unit at The Christie NHS Foundation Trust, Manchester. Assessments are carried out in face-to-face interviews at baseline and at 6 months follow-up. Participants are referred into the study by Care Coordinators or self-referral. Potential participants are contacted by a research assistant (RA1) who presents them with verbal and written information about the study. If they wish to take part, the participant is asked to give written consent. Following this, a baseline interview is conducted to assess eligibility and to complete all the measures. A second research assistant (RA2) then contacts the Clinical Trials Unit and provides information about which Trust the relative is in. They are given the trial allocation by telephone. RA2 contacts the relative by telephone and post to inform them which arm of the trial they have been allocated to. RA2 contacts the relevant support worker who will guide the relative through the intervention and instructs them to arrange the first appointment with the participant. RA2 also arranges the 6 month follow-up interview date. In order to ensure that outcomes are collected blind to the treatment allocation, a letter is sent prior to this interview reminding the participant about the importance of not letting RA1 know which group they were allocated to, and making sure that the REACT toolkit is not visible in the house. This is followed-up with a phone call the day before the interview to check the appointment time is convenient and to reiterate the need to maintain blindness. RA1 conducts all of the follow-up interviews and remains blind to allocation throughout the study. To ensure blindness, all communication between relatives and NHS staff is via RA2. The RAs are housed in separate offices and receive individual supervision. Any instances of unblinding will be recorded.
This study was reviewed and approved by the UK NHS Ethics Committee process (REC ref: 08/H1001/ 147).
## Qualitative evaluation
Qualitative interviews will be used to help us understand how the intervention is used and experienced. Relatives participating in the intervention arm of trial will form the strategic sampling pool and will be invited to take part in lightly structured interviews to explore key domains around their use of and experience of the intervention, focusing on the barriers to using the intervention and potential solutions to overcome these. Recruitment will continue until data saturation is reached (estimate approx 10-15 relatives).
## Participants
## Inclusion criteria
Three NHS Mental Health Trusts in the North West UK are taking part in this study. Each trust has an Early Intervention Service for Psychosis team which supports young adults who are experiencing symptoms of psychosis for a period of up to 3 years. Participants are relatives, partners or close friends of people currently receiving support from one of these teams. Additional criteria include: first contact with EIS within the last 24 months; sufficient understanding of written and spoken English to be able to use the intervention; aged over 18.
## Recruitment and consent
Only relatives who already have direct contact with each EIS, and for whom the service has current contact details are invited to take part in the research. This includes relatives currently on the "family and friends group" lists, currently attending groups, and currently receiving face-to-face input from clinical staff and support workers. No relatives are contacted using details taken from patient medical notes without patient consent. No information is collected about the service user. Therefore, service users will not be required to consent into the study. However, to ensure service users are aware of the study and could invite any of their relatives not currently involved with the service, each service will write to all service users informing them about the study and send them a copy of the relatives' information sheet with a covering letter asking them to pass this to any relatives that they would be happy for us to invite and who they feel may be interested in taking part.
Care Coordinators will be informed about the study in a series of presentations given by the research team. They will be asked to discuss the study with relatives they are working with and to refer any relatives interested in taking part who consent to their details being given to the research team. A series of presentations will also be made to relatives at any service user and care events that happen in the Trusts during the recruitment phase. These will encourage relatives to refer themselves into the study if they wish to take part. A website will also be set up that enables confidential referrals to be made either by Care Coordinators or directly from relatives.
## Sample size
The main aim of the study is to assess the feasibility of the design and intervention. The sample required for this is not based on a formal power calculation but on a pragmatic decision balancing sufficient numbers to be able to identify likely barriers to carrying out a larger scale trial, and cost. We aim to recruit 100 participants which will also allow us to estimate the likely effect size of the intervention on a range of outcome measures.
## Outcome measures
Relatives will be assessed on a range of measures at both baseline and follow-up. All measures are conducted in a face-to-face interview.
1) The General Health Questionnaire (GHQ-28) [bib_ref] Scaled Version of the General Health Questionnaire, Goldberg [/bib_ref] is a 28-item version of General Health Questionnaire derived by factor analysis from the full 60-item version. It is used to assess the psychological aspect of quality of life or in the detection of psychiatric distress. Participants indicate whether their current state differs from their usual state, thereby assessing recent changes in state as opposed to long-term traits or illnesses.
2) The Family Questionnaire [bib_ref] The Family Questionnaire (FQ): a scale for measuring symptom appraisal in relatives..., Quinn [/bib_ref] presents participants with a range of symptoms of mental illness that have the potential to pose a problem in family life. It assesses the extent to which each of these symptoms cause the relative concern and how well the relative is able to cope with their concerns.
3) The Relationship Quality Scale [bib_ref] Predictors of relapse in unipolar depressives: Expressed emotion, marital distress, and perceived..., Hooley [/bib_ref] asks service users and relatives to rate their perception of their relationship with one another on scales of supportiveness and criticism. This measure has been shown to predict outcome for people with severe depression [bib_ref] Predictors of relapse in unipolar depressives: Expressed emotion, marital distress, and perceived..., Hooley [/bib_ref] and schizophrenia [bib_ref] Patients Perceptions of Family Emotional Climate and Outcome in Schizophrenia, Lebell [/bib_ref]. 4) The Brief Illness Perception Questionnaire [bib_ref] Assessing cognitive representations of mental health problems. II The Illness Perception Questionnaire..., Lobban [/bib_ref] adapted from the original Illness Perception Questionnaire [bib_ref] The illness perception questionnaire: A new method for assessing the cognitive representation..., Weinman [/bib_ref] measures the insight of participants into their relative's illness. Components include the participant's views on the consequences of the illness, comprehension of the illness, knowledge of prognosis, the extent to which they attribute blame, knowledge of potential causes and their own concern and emotional responses to the illness. 5) The Herth Hope Index [bib_ref] Abbreviated Instrument to Measure Hope -Development and Psychometric Evaluation, Herth [/bib_ref] consists of 12 items designed to assess hope in adults in clinical settings. 6) The Experience of Caregiving Inventory [bib_ref] Caring for relatives with serious mental illness: The development of the experience..., Szmukler [/bib_ref] measures the experience of caring for a relative with a serious mental illness. The 66-items are divided into 10 sub-scales, 8 negative (difficult behaviours, negative symptoms, stigma, problems with services, effects on the family, the need to provide back-up, dependency, loss) and 2 positive (rewarding personal experiences, good aspects of the relationship with the patient). 7) The Carer Well-Being and Support Questionnairemeasures the experience of carers of people with severe mental health problems. It is designed to cover all aspects of the carer's experience including relationships, roles, financial concerns, physical/ emotional health, stigma, worries about safety, their satisfaction with support offered and ease of obtaining information. 8) The Relatives' Satisfaction Questionnaire (adapted from CSQ-8 [bib_ref] The client Satisfaction Questionnaire-8 and the Service Satisfaction Questionnaire-30, Attkinson [/bib_ref] is used to assess satisfaction with services in both arms of the trial to test the hypothesis that the intervention will lead to an increased level of satisfaction in relatives. 9) The Treatment As Usual Checklist is a short questionnaire designed for this study to assess the amount and type of support that participants have received from the Early Intervention Service while in the Treatment as Usual arm of the trial.
## The intervention development
During the first year of the study a CBT oriented, supported self-management package for relatives of people with a recent onset psychosis was developed and finalised. First, a systematic review of studies of psychological interventions for relatives of people with recent onset psychosis was conducted. This identified the key components of effective interventions and distinguished them from those of ineffective interventions. Second, relatives of people who have experienced psychosis were invited to take part in focus groups to allow in-depth analysis of relatives' views and experiences of a selfmanagement approach. The questions asked, focussed around (i) their experiences of self-management approaches; (ii) preferred format (or "health technology"); (iii) what support they would like to receive; (iv) perceived barriers. Further details on the findings of this study are reported elsewhere [bib_ref] REACT: The views of relatives of young people with psychosis on how..., Lobban [/bib_ref]. Finally, participants from the focus groups were invited to be part of an Intervention Reference Group that was involved in an iterative process of feedback and development to produce the supported self-management package. An independent design company was used to style the package professionally and an independent publisher was used to produce the required number of intervention packs.
## Features
The finalised package is modular in design, providing a range of sections or "tools" that relatives can draw on as needed. These modules have been designed to operate independently of each other and contain topics such as essential information about psychosis, ways to identify and challenge beliefs that may cause distress, ways to manage common difficulties faced by relatives, and coping strategies to aid recovery for both the relative and the person with psychosis. The package is also well referenced to guide people to existing support in related domains such as legal advice, advocacy, charities etc. Building on previous research in self-management for depression, the intervention also gives relatives the opportunity to incorporate their personal experience into the process, to situate this intervention within a context of previous experiences, build on existing selfmanagement strategies to facilitate engagement, and highlight the self as the key agent of change [bib_ref] Guided self-help in primary care mental health -Meta-synthesis of qualitative studies of..., Khan [/bib_ref].
The intervention has been designed to be used by relatives in their own homes and at their own convenience. As such, it has been produced as a hard copy format and as a website and participants are able to choose to use either or both of these formats. Both versions contain the same information and resources. Support is provided by an NHS support worker trained and supervised by the research team. The support worker will offer an initial face-to-face introductory session in which they guide the relative through the materials and how to use them. Following this, support in using the package will be offered via email or telephone depending on the relatives' preference.
# Analysis
## Quantitative data
Analysis of outcomes will be carried out using simple descriptive statistics, tabulation and simple graphical display. Tests of statistical significance will be carried out using Student's t-test (quantitative outcomes) or Pearson chi-square (binary outcomes). Further analyses to allow for pre-randomisation (baseline) measures and to investigate the effects of missing outcomes will be carried out using analyses of covariance or logistic regression, respectively.
## Qualitative data
Analysis will be carried out by a multidisciplinary team of psychologists, nurses, relatives and researchers. To ensure that the analysis is grounded in the data, rather than reflecting pre-existing ideas, analysis will occur in parallel with data collection so that aspects of the developing analysis can be tested in subsequent interviews. Categorisation and thematic analysis of the data will be developed by cycling between the analysis and transcripts and periodic 'testing' of the analysis by discussion amongst the entire team so as to meet accepted criteria for trustworthiness of the analysis. In addition, we will assess the validity of the final analysis by examining coherence and catalytic validity, which is essentially the utility of the analysis in identifying implications for clinical practice and research that can be tested.
# Discussion
The REACT toolkit has the potential to offer relatives the information and support they need during the crucial period of early psychosis, in a format which is highly accessible, free to access and can be used flexibly to meet their individual needs. It offers NHS Trusts the potential to meet government guidelines for supporting relatives, whilst overcoming many of the barriers associated with training and resources in offering face-to-face family interventions. The content of the intervention is based on already established effective interventions, but important questions need to be answered about whether these interventions can be offered in a supported self-management format. This trial will answer key questions of feasibility that need to be addressed before a large scale clinical and cost effectiveness evaluation of this approach. Specifically, the trial will provide extensive quantitative and qualitative data on the acceptability of the intervention for relatives, exactly how they use the intervention, and their preferences for how this should be delivered, including format of the toolkit and type and amount of support required to use it. Barriers and facilitators to using the toolkit and support will be identified from the perspective of the relatives, but also from the clinical staff attempting to deliver REACT. Important data for future research will include feasibility of recruitment, and retention to a large scale trial, and estimates of effect sizes on key outcome variables which can inform future power calculations.
This trial benefits from rigorous design in terms of independent randomisation, blind rated assessments, and a clearly protocolised complex intervention. In addition, the research is conducted within a "real world" setting, recruiting relatives from existing NHS services, delivering the intervention alongside current treatment, and offering support via existing clinical staff. This increases the external validity of the findings, but is also the source of some potential limitations. These include the variation that will undoubtedly exist in the way in which the intervention is supported, and variation within the current treatment offered as the control condition. Variation in both is likely to be present at an operational level between the Trusts taking part, but also for each participant, given their individual journeys through mental health services. Extensive process measures are being collected in an attempt to measure this variation with a view to understanding potential confounds in future effectiveness trials. Additional limitations include a follow-up period of 6 months only, and lack of service user outcomes. Both can be addressed in future trials but are beyond the scope and resources of a feasibility trial.
REACT is designed to provide information and strategies for relatives to build upon their existing coping strategies. It is not designed to replace face-to-face contact with Care Coordinators who provide valuable emotional support and detailed specific information about the service user that relatives are so keen to understand. Neither is this approach designed to replace intensive family therapy that should be offered to those families where difficulties arising from the psychosis have caused major breakdown in communication within the family, or severe psychological distress for the relatives. All relatives should have direct access to crisis services. Finally, many relatives derive benefit from receiving support from other relatives who have also experienced psychosis within their family. In its current form, this toolkit fails to offer this peer support that is available in many Early Intervention Services via family and friends support groups. Although REACT has been designed with user involvement to meet the needs of relatives, it is important the toolkit is offered as part of a comprehensive service for relatives.
[fig] Figure 1: Consort diagram showing progression of participants through the trial. [/fig]
[fig] Figure 2: Development of the REACT Intervention. [/fig]
|
Impact of primary colorectal Cancer location on the KRAS status and its prognostic value
Background: Colorectal cancer (CRC) originating from the right-sided or left-sided colon is distinct clinicopathological entity. The KRAS status and its prognostic value in CRC remain controversial. This study aimed to investigate the association of KRAS status with clinicopathological features and prognostic value in CRC. Methods: 178 colon cancer and 145 rectal cancer patients were enrolled. KRAS mutation test was performed on paraffin-embedded tumor samples using PCR methods. The colon cancer was divided into right-sided colon cancer (RCC) and left-sided colon cancer (LCC). Studies that reported the association of KRAS mutation with CRC clinical features and prognosis in databases were searched prior to 2018. The data of the present study was combined with the data of published studies using meta-analysis methods. Results: No significant difference between colon cancer and rectal cancer regarding the KRAS status. The KRAS mutation was much frequent in RCC than in LCC (p = 0.010). 17 studies with 11,385 colon cancer patients were selected, the pooled results of our data and previous published data showed that KRAS mutation was more frequent in RCC compared with in LCC (p < 0.01); KRAS mutation was not associated with the prognosis in RCC patient; however, KRAS mutation indicated a poor prognosis in LCC patients compared with KRAS wild type (p < 0. 01). Conclusion: KRAS status has no difference between colon cancer and rectal cancer. KRAS mutation was more frequent in RCC than in LCC, and associated with a poor prognosis in LCC patients, but not in RCC patients.
# Background
Colorectal cancer (CRC) is the third most common malignancy globally, accounting for approximately.0% of all new cancer cases [bib_ref] Colorectal cancer statistics, Siegel [/bib_ref]. CRC can be divided into colon cancer and rectal cancer base on their primary tumor location within the colon and rectum. The colon cancer can further be classified into right-sided colon cancer (RCC) and left-sided colon cancer (LCC) divided at the site of splenic flexure of colon. In recent years, a growing evidences revealed that there were significant differences between RCC and LCC with regard to the clinical findings, pathology, genetic mutations and survival time [bib_ref] Understanding the role of primary tumour localisation in colorectal cancer treatment and..., Stintzing [/bib_ref]. Thus, the location of tumor is an important factor that affects the prognosis of CRC.
Knowledge has shown that CRC tumorigenesis was characterized by the accumulation of genetic mutations, and Kirsten rat sarcoma viral oncogene homolog (KRAS) mutation was an early event in tumorigenesis [bib_ref] RAF/RAS oncogenes and mismatch-repair status, Rajagopalan [/bib_ref]. KRAS mutation occur in approximately 30 to 50% of CRC, and 90% of mutation occur in codon 12 or 13 [bib_ref] Kirsten ras mutations in patients with colorectal cancer: the 'RASCAL II' study, Andreyev [/bib_ref] [bib_ref] Comprehensive biostatistical analysis of CpG island methylator phenotype in colorectal cancer using..., Nosho [/bib_ref] [bib_ref] Cetuximab and chemotherapy as initial treatment for metastatic colorectal cancer, Van Cutsem [/bib_ref]. At present, anti-epidermal growth factor receptor (EGFR) antibody has been showed to be an effective therapy in the treatment of CRC patients. However, patients with KRAS mutation are unlikely to benefit from anti-EGFR therapy [bib_ref] Association of KRAS p.G13D mutation with outcome in patients with chemotherapy-refractory metastatic..., De Roock [/bib_ref] [bib_ref] K-ras mutations and benefit from cetuximab in advanced colorectal cancer, Karapetis [/bib_ref] , thus the KRAS status is used as an important biomarker for the selection of suitable patients.
To date, many studies reported the clinicopathological features of CRC, and some studies further analyzed the KRAS status in RCC and LCC. However, the results of KRAS status in RCC and LCC remained inconsistent [bib_ref] Molecular profiling of patients with advanced colorectal Cancer: Princess Margaret Cancer Centre..., Chiu [/bib_ref] [bib_ref] Clinicopathological and molecular differences between right-sided and leftsided colorectal cancer in Japanese..., Natsume [/bib_ref]. Here, we reported our results of clinicopathological features and KRAS status in Chinese patients with CRC, and further compared the KRAS status in RCC and LCC.
Although some studies showed that the tumor location and KRAS status could affect the effectiveness of patients treated with cetuximab [bib_ref] Impact of KRAS and TP53 mutations on survival in patients with left-and..., Bleeker [/bib_ref] [bib_ref] Differences of protein expression profiles, KRAS and BRAF mutation, and prognosis in..., Gao [/bib_ref] , the association between KRAS mutation and patients' survival remained controversial, as some reports have failed to show any prognostic value of KRAS [bib_ref] Relevance of Ki-67 antigen expression and K-ras mutation in colorectal liver metastases, Petrowsky [/bib_ref] [bib_ref] Influence of KRAS mutation status in metachronous and synchronous metastatic colorectal adenocarcinoma, Rose [/bib_ref] [bib_ref] BRAF mutation is a prognostic biomarker for colorectal liver metastasectomy, Teng [/bib_ref] [bib_ref] The impact of KRAS mutations on prognosis in surgically resected colorectal cancer..., Kim [/bib_ref]. For example, there was a study found that KRAS mutations were not associated with risk of death in the overall patients of CRC, but LCC patients harboring KRAS mutation have a greater risk of death [bib_ref] KRAS testing, tumor location, and survival in patients with stage IV colorectal..., Charlton [/bib_ref]. These results suggested that the prognostic value of KRAS status in CRC patients might also depending on the location of primary tumor; however, due to the limited of reports, this result still need to be further validated. Therefore, in this study, we combined the published data to further explore the prognostic value of KRAS status in both RCC and LCC.
# Methods
## Patients and data extraction
This study was retrospective design, patients who diagnosed with CRC and undergoing radical surgery in the Affiliated Tumor Hospital of Guangxi Medical University from January 2015 to January 2018 were included. The inclusion criteria were: CRC was confirmed by historical biopsy; Patients with inflammatory bowel disease or a known history of familial adenomatous polyposis were excluded. Patients with unknown KRAS status or receiving anti-EGFR agents in the perioperative period were also excluded. Detailed information was obtained on patients' age,
## Search strategy for articles
Because the association of KRAS status with clinicopathological features and prognostic value in CRC might depend on the primary tumor location, we retrieval articles that analyzing the KRAS status and the prognostic value in RCC and LCC prior to April 2018 by searching the following electronic databases, PubMed, Cochrane Library, Web of Science, EBSCO and Chinese National Knowledge Infrastructure (CNKI). The following search terms were employed: "colon cancer", "KRAS", "left-side" or "right-side," "prognosis". Included articles were limited to human studies but not limited by language. The first author, year of publication, study location, number
# Statistical analysis
Demographic and clinicopathological characteristics of the patients were stratified according to primary tumor location and KRAS mutation status. Continuous variables were presented as mean ± standard deviation, and compared using a Student's t-test. Summary statistics for the patients were presented as totals for categorical variables. The differences between wild-type KRAS (wt-KRAS) and mutant-type KRAS (mt-KRAS) in each group were assessed by the χ 2 test. The analyses were performed using R software version 3.4.3.
The meta-analysis of KRAS status in RCC and LCC, and the prognostic value of KRAS status in RCC and LCC was performed using Stata 11.2 software (Stata Corp, College Station, TX) with 2-tailed p-values. The pooled odds ratio (OR) with the corresponding 95% CI were used to estimate then KRAS status in RCC and LCC. The pooled HR with the corresponding 95% CI was used to assess the prognostic value of KRAS status in RCC and LCC. The p-value < 0.05 was considered statistically significant.
# Results
## Clinicopathological characteristics of crc patients
There were 178 colon cancer and 145 rectal cancer patients finally enrolled in this study. The mean age of colon cancer and rectal cancer was (57.62 ± 12.67) years and (59.57 ± 11.89) years, respectively. No significant different between colon cancer and rectal cancer in the KRAS status (p = 0.393). Most of the colon cancer and rectal cancer was moderate differentiation, but no significant difference between colon cancer and rectal cancer (p = 0.099). There was no obvious difference in nodal status, distant metastases, AJCC stage between colon cancer and rectal cancer (p > 0.05). However, the number of advanced tumor infiltration patients (T3 + 4 stage) of colon cancer was greatly larger than that of rectal cancer (p = 0.018). Detail of included CRC patients was listed in [fig_ref] Table 1: Characteristics of colon and rectal cancer patientsColon [/fig_ref].
## Clinicopathological characteristics of colon cancer in different status of kras
Of the 178 colon cancer patients, 62 occur KRAS mutation, the remained116 were KRAS wild-type. We divided these patients based on the status of KRAS, and found that no obvious differences in patients' gender, age, histological differentiation, nodal status, distant metastases, AJCC stage (p > 0.05). However, significant difference was found between RCC and LCC regarding the KRAS status, with KRAS mutation in RCC was 46.4% (32/69), and in LCC was 27.5% (30/109), p = 0.010. [fig_ref] Table 2: Characteristics of colon cancer in different KRAS status [/fig_ref] showed the detail of the characteristics in different status of KRAS.
## Characteristics of included studies
Seventeen studies [bib_ref] Molecular profiling of patients with advanced colorectal Cancer: Princess Margaret Cancer Centre..., Chiu [/bib_ref] [bib_ref] Clinicopathological and molecular differences between right-sided and leftsided colorectal cancer in Japanese..., Natsume [/bib_ref] [bib_ref] Impact of KRAS and TP53 mutations on survival in patients with left-and..., Bleeker [/bib_ref] [bib_ref] Differences of protein expression profiles, KRAS and BRAF mutation, and prognosis in..., Gao [/bib_ref] [bib_ref] KRAS testing, tumor location, and survival in patients with stage IV colorectal..., Charlton [/bib_ref] [bib_ref] Mutation spectra of RAS gene family in colorectal cancer, Chang [/bib_ref] [bib_ref] Retrospective analysis of KRAS, NRAS and BRAF gene mutations and clinicopathological features..., Hou [/bib_ref] [bib_ref] The impact of microsatellite instability status and sidedness of the primary tumor..., Kim [/bib_ref] [bib_ref] Association between specific KRAS mutations and the clinicopathological characteristics of colorectal tumors, Kodaz [/bib_ref] [bib_ref] Prognostic implication of KRAS status after hepatectomy for colorectal liver metastases varies..., Sasaki [/bib_ref] [bib_ref] Relationship between KRAS/NRAS/BRAF gene mutations and clinical pathological characteristics in colorectal cancer, Sun [/bib_ref] [bib_ref] KRAS and BRAF gene mutations and DNA mismatch repair status in Chinese..., Ye [/bib_ref] [bib_ref] Prognostic significance of alterations in KRAS isoforms KRAS-4A/4B and KRAS mutations in..., Abubaker [/bib_ref] [bib_ref] Characterization of rare transforming KRAS mutations in sporadic colorectal cancer, Tong [/bib_ref] [bib_ref] Left-sided primary tumors are associated with favorable prognosis in patients with KRAS..., Von Einem [/bib_ref] [bib_ref] KRAS and BRAF gene mutations in correlation with clinicopathologic features of colorectal..., Zhu [/bib_ref] with 11,385 colon cancer patients were included in this study based on the included criteria. Among them, sixteen studies [bib_ref] Molecular profiling of patients with advanced colorectal Cancer: Princess Margaret Cancer Centre..., Chiu [/bib_ref] [bib_ref] Clinicopathological and molecular differences between right-sided and leftsided colorectal cancer in Japanese..., Natsume [/bib_ref] [bib_ref] Impact of KRAS and TP53 mutations on survival in patients with left-and..., Bleeker [/bib_ref] [bib_ref] Differences of protein expression profiles, KRAS and BRAF mutation, and prognosis in..., Gao [/bib_ref] [bib_ref] Mutation spectra of RAS gene family in colorectal cancer, Chang [/bib_ref] [bib_ref] Retrospective analysis of KRAS, NRAS and BRAF gene mutations and clinicopathological features..., Hou [/bib_ref] [bib_ref] The impact of microsatellite instability status and sidedness of the primary tumor..., Kim [/bib_ref] [bib_ref] Association between specific KRAS mutations and the clinicopathological characteristics of colorectal tumors, Kodaz [/bib_ref] [bib_ref] Prognostic implication of KRAS status after hepatectomy for colorectal liver metastases varies..., Sasaki [/bib_ref] [bib_ref] Relationship between KRAS/NRAS/BRAF gene mutations and clinical pathological characteristics in colorectal cancer, Sun [/bib_ref] [bib_ref] KRAS and BRAF gene mutations and DNA mismatch repair status in Chinese..., Ye [/bib_ref] [bib_ref] Prognostic significance of alterations in KRAS isoforms KRAS-4A/4B and KRAS mutations in..., Abubaker [/bib_ref] [bib_ref] Characterization of rare transforming KRAS mutations in sporadic colorectal cancer, Tong [/bib_ref] [bib_ref] Left-sided primary tumors are associated with favorable prognosis in patients with KRAS..., Von Einem [/bib_ref] [bib_ref] KRAS and BRAF gene mutations in correlation with clinicopathologic features of colorectal..., Zhu [/bib_ref] with 5, 835 patients provided the data of KRAS status in colon cancer, with 3961 LCC patients and 1874 RCC patients, respectively. Four studies [bib_ref] Clinicopathological and molecular differences between right-sided and leftsided colorectal cancer in Japanese..., Natsume [/bib_ref] [bib_ref] KRAS testing, tumor location, and survival in patients with stage IV colorectal..., Charlton [/bib_ref] [bib_ref] Prognostic implication of KRAS status after hepatectomy for colorectal liver metastases varies..., Sasaki [/bib_ref] [bib_ref] Left-sided primary tumors are associated with favorable prognosis in patients with KRAS..., Von Einem [/bib_ref] with 6697 patients provided the survival data of KRAS status in colon cancer, with 3670 LCC patients and 3027 RCC patients, respectively. [fig_ref] Table 3: Characteristics of included studiesLCC left-side colon cancer, RCC right-side colon cancer, wt... [/fig_ref] presented the detail of the characteristics of included studies. A flow chart of the article selection process was shown in [fig_ref] Figure 1: Flow diagram of the study selection process [/fig_ref].
## Pooled results of kras status in rcc and lcc
To clarify the KRAS status in RCC and LCC, we combined the data from sixteen studies that provided the KRAS status in RCC and LCC with our data. The pooled results showed that KRAS mutation was much more frequent in RCC compared with LCC (OR = 1.68, 95%CI = 1.50-1.88, p < 0.01), and no significant heterogeneity across the studies (I 2 = 34.3%, p = 0.082). See . There was little publication bias across the studies (Begg's Test = 0.343; Egger's test = 0.575). See [fig_ref] Figure 3: Begg's funnel plot for publication bias test [/fig_ref].
## Pooled results of prognostic value of kras status in rcc and lcc
To estimate the difference prognostic value of KRAS status in RCC and LCC, we combined the data from four studies that provided the data of overall survival (OS) of RCC and LCC patients with different KRAS status. All the studies included metastatic CRC cases, and patients receiving chemotherapy and/or radiotherapy after surgical resection. The multivariate analysis was performed by adjusting the confounding prognostic factors in each study. The pooled results showed that RCC patients with KRAS mutation has no significant different OS compared with patients with KRAS wild type (HR = 0.77, 95%CI = 0.58-1.02, p = 0.073; I 2 = 62.1%), see ; however, LCC patients with KRAS mutation has a Forest plot of the KRAS mutation between right-sided colon cancer and left-sided colon cancer. The squares and horizontal lines correspond to the study-specific OR and 95% CI. The diamond represents the summary OR and 95% CI. The diamond locates to the right of vertical line means the KRAS mutation was much more frequent in RCC compared with in LCC (OR < 1)
shorter OS than patients with KRAS wild type (HR = 1.21, 95%CI = 1.08-1.36, p < 0.01; I 2 = 29.0%), see .
# Discussion
In this study, by analyzing the clinicopathological features of 178 colon cancers and 145 rectal cancers, we failed to find the difference between colon cancer and rectal cancer regarding the KRAS status. By dividing the CRC based on the KRAS status, we did not observe the difference between KRAS mutation and wild type regarding the clinicopathological features, but found that RCC harboring more KRAS mutation compared with LCC (46.4% vs. 37.5%). We next combined the data of KRAS status in RCC and LCC, by pooling the data of Forest plot of the KRAS mutation in the prediction of right-sided colon cancer patients. The squares and horizontal lines correspond to the study-specific HR and 95% CI. The diamond represents the summary HR and 95% CI. The diamond locates to the left but touches the vertical line means the no significant difference between patients with and without KRAS mutation regarding the OS sixteen studies and our data, we found that KRAS mutation was much more frequent in RCC than in LCC. In addition, by pooling the data of four studies, we found an obvious difference of OS in RCC and LCC regarding the KRAS status, that is, LCC patients with KRAS mutation has a shorter OS than with KRAS wild type, while RCC patients with KRAS mutation has no significant different OS compared with patients with KRAS wild type. These results indicated that both tumor location and KRAS status play important roles in the prognosis of CRC patients.
Knowledge has shown that the right and left sides of the colon have different embryologic origins. Tumor that origins from the two sites of the colon has different molecular carcinogenic characters, including KRAS, BRAF mutations and microsatellite instability (MSI) [bib_ref] Impact of KRAS and TP53 mutations on survival in patients with left-and..., Bleeker [/bib_ref] [bib_ref] Differences between right-and left-sided colon cancer in patient characteristics, cancer morphology and..., Nawa [/bib_ref] [bib_ref] Value of mismatch repair, KRAS, and BRAF mutations in predicting recurrence and..., Hutchins [/bib_ref]. KRAS has been confirmed as a proto-oncogene which induces tumorigenesis in several cancers. In CRC, KRAS mutations status and tumor location are associated with targeted therapy effectiveness. In this study, KRAS status has no obvious difference in colon cancer or rectal cancer, but showed significant difference in RCC and LCC, which was in consistent with Natsume et al. [bib_ref] Clinicopathological and molecular differences between right-sided and leftsided colorectal cancer in Japanese..., Natsume [/bib_ref] and Tong et al. [bib_ref] Characterization of rare transforming KRAS mutations in sporadic colorectal cancer, Tong [/bib_ref] reports, but in contrast to Cushman-Vokoun et al.report. Then the following meta-analysis with larger patients further verified the different KRAS status in RCC and LCC, indicating that the KRAS mutation was more frequent in RCC than LCC.
Since the effect of anti-EGFR therapy on CRC is associated with KRAS status, many studies have estimated the prognostic value of KRAS status in CRC patients [bib_ref] K-ras mutations and benefit from cetuximab in advanced colorectal cancer, Karapetis [/bib_ref] [bib_ref] Molecular profiling of patients with advanced colorectal Cancer: Princess Margaret Cancer Centre..., Chiu [/bib_ref] , and some studies showed that mutation of KRAS indicated a poor prognosis of CRC patients, but there were also some reports have failed to show the similar result [bib_ref] Relevance of Ki-67 antigen expression and K-ras mutation in colorectal liver metastases, Petrowsky [/bib_ref] [bib_ref] Influence of KRAS mutation status in metachronous and synchronous metastatic colorectal adenocarcinoma, Rose [/bib_ref] [bib_ref] BRAF mutation is a prognostic biomarker for colorectal liver metastasectomy, Teng [/bib_ref] [bib_ref] The impact of KRAS mutations on prognosis in surgically resected colorectal cancer..., Kim [/bib_ref] [bib_ref] Mutation spectra of RAS gene family in colorectal cancer, Chang [/bib_ref] , thus, the current conclusions regarding the prognostic value of KRAS status remain inconclusive. Because the distinct genetic alteration between RCC and LCC, both of the location of tumor and KRAS status are proposed to influence the prognostic value CRC. As Sasaki et al. [bib_ref] Prognostic implication of KRAS status after hepatectomy for colorectal liver metastases varies..., Sasaki [/bib_ref] pointed that, KRAS mutation in RCC was not associated with the prognosis of CRC, while KRAS mutation in LCC indicated a poor prognosis of CRC patients. However, this result was based on a relative small sample size; thus, the robustness of the conclusion was undermined. In this study, by combining the data from four studies with 6697 patients, we found that LCC patients with KRAS mutation has a poor prognosis, but RCC patients with KRAS mutation did not show the similar results, which can partly explain the inconsistent results of the prognostic value of KRAS status in CRC patients. And these results further verified that both the KRAS status and location of tumor could affect the treatment effectiveness and prognosis in CRC patients.
Although this study showed the different KRAS status in RCC and LCC, and found the prognostic value of KRAS mutation was depending on the location of tumor, there were several limitations should be considered when interpreting the results. First, due to lack of the survival data in our center, we did not combine our data with published studies, thus, the sample size of the Forest plot of the KRAS mutation in the prediction of left-sided colon cancer patients. The squares and horizontal lines correspond to the study-specific HR and 95% CI. The diamond represents the summary HR and 95% CI. The diamond locates to the right of vertical line means the LCC patients with KRAS mutation has a shorter OS than patients with KRAS wild type (HR > 1) analysis for the prognostic value of KRAS status was relative small. Second, due to the limited studies available, we did not divide the patients based on the their ethnicity, so we did not know whether various ethnicity could affect the prognostic value of KRAS status in CRC, since evidence has shown that there were many differences in CRC between Asian and Caucasian ethnicity [bib_ref] Racial differences in colorectal cancer incidence and mortality in the Women's Health..., Simon [/bib_ref] [bib_ref] Racial differences in the incidence of colorectal cancer, Virk [/bib_ref]. Third, the present study only included the data of mutation of KRAS codons 12 and 13 in exon 2, other mutations, such as NRS and BRAF mutation, were not included. Although these type of mutations were fewer compared with the KRAS mutation, the lack of data of other mutations might lead to selected bias in the analysis. Fourth, the pooling analysis included all stages of CRC patients without stratified them into different stage, that is, early stage or advanced stage of CRC, hence the data very heterogeneous and would reduce the robustness of the results. Fifth, some of the included studies were retrospective design, which may lead selected bias and undermine the robustness of the results. Therefore, future research should be conducted to address the aforementioned limitations.
# Conclusion
This study demonstrated that no significant difference of KRAS status between colon cancer and rectal cancer. KRAS mutation was much more frequent in RCC compared with LCC, and LCC patients with KRAS mutation has a poor prognosis compared with KRAS wild type, but RCC patients did not show the similar effect.
Abbreviation CRC: Colorectal cancer; EGFR: Epidermal growth factor receptor; HR: Hazard ratio; KRAS: Kirsten rat sarcoma viral oncogene homolog; LCC: Left-sided colon cancer; OR: Odds ratio; RCC: Right-sided colon cancer
[fig] Figure 1: Flow diagram of the study selection process [/fig]
[fig] Figure 3: Begg's funnel plot for publication bias test [/fig]
[fig] Figure 4: Forest plot of the KRAS mutation in the prediction of right-sided colon cancer patients. The squares and horizontal lines correspond to the study-specific HR and 95% CI. The diamond represents the summary HR and 95% CI. The diamond locates to the left but touches the vertical line means the no significant difference between patients with and without KRAS mutation regarding the OS [/fig]
[table] Table 1: Characteristics of colon and rectal cancer patientsColon (n = 178) Rectal (n = 145) t/χ 2 value P value [/table]
[table] Table 2: Characteristics of colon cancer in different KRAS status [/table]
[table] Table 3: Characteristics of included studiesLCC left-side colon cancer, RCC right-side colon cancer, wt KRAS wild type, mut KRAS mutant type, HR hazard ratio, CI confidence interval, P prospective, R retrospective of KRAS status in LCC and RCC patients, hazard ratio (HR) and the corresponding 95%CI of prognostic value of KRAS status in LCC and RCC were extracted. [/table]
|
Two cases of an internal hernia after laparoscopic sigmoid resection
## Summary
Two patients with a laparoscopic resection of the sigmoid colon in their surgical history were diagnosed with an internal hernia because of a mesenteric gap underneath the descending neocolon. While the first case demonstrated a strongly symptomatic patient with a closed-loop obstruction of the small bowel, the second case was less obvious and correct diagnosis was achieved after a few weeks by finally performing a diagnostic laparoscopy. Since internal hernias after laparoscopic sigmoid resection are a rare complication and as presented in the second case can display very unspecific symptoms, the surgeon's awareness is not immediately raised towards it. Therefore, an asymptomatic internal hernia bears a constant risk of a life-threatening outcome allowing to recommend an initial closure of the mesenteric gap. In both cases, successful treatment was achieved by surgery and intraoperatively closing the remaining mesenteric gap. Subsequently recovery and follow-up were uneventful.
# Background
Colonic diverticulitis is a gastrointestinal disease characterised by inflammation of abnormal pouches-diverticula-which can develop within the wall of the large intestine. The disease is common in western countries. [bib_ref] Epidemiology and pathophysiology of diverticular disease, Munie [/bib_ref] Due to a successful medical treatment of complicated diverticulitis, the indication for a surgical intervention is reserved for complicative situations including diverticular perforation, recurrent paracolic abscess formation or chronic symptomatic diverticulitis. If a patient qualifies for surgical treatment of diverticulitis, a laparoscopic approach is generally preferred due to lower perioperative morbidity and faster postoperative recovery. [bib_ref] State-Of-The-Art surgery for sigmoid diverticulitis, Cirocchi [/bib_ref] Nowadays, the technique of laparoscopic sigmoid resection is well established and represents the gold standard. To achieve a tension-free descendorectostomy, extensive mobilisation of the splenic flexure and transposition of the descending colon into the small pelvis are necessary. Complications after laparoscopic sigmoid resection include anastomotic leakage (2.9%-15.3%), 5 bleeding (0.5%-1%), [bib_ref] Management of complications following emergency and elective surgery for diverticulitis, Holmer [/bib_ref] ureteral injury (1%), [bib_ref] Incidence of iatrogenic ureteral injury during open and laparoscopic colorectal surgery: a..., Marcelissen [/bib_ref] and surgical site infections (5%). [bib_ref] Infective complications in laparoscopic surgery, Boni [/bib_ref] Surgeon's awareness of these complications needs to be high and, therefore, immediate steps for diagnosis and therapy are usually initiated.
However, other complications such as an internal hernia do occur less frequently and thus can be easily overseen due to the low specificity of symptoms. [bib_ref] Internal hernia following laparoscopic colorectal surgery: single center experience, Svraka [/bib_ref] We herein present two cases of patients with internal hernia following laparoscopic sigmoid resection demonstrating the large variety of clinical findings and emphasising the need to consider this rare differential diagnosis in patients with abdominal pain postoperatively.
## Case presentation case presentation 1
A female patient in her 80s presented at the emergency department with acute epigastric pain and nausea without vomiting for 4 hours. An association with eating or drinking was not observed. The patient's further medical history showed that similar problems had occurred several times in the last weeks and months. However, these had only lasted for a few minutes. The patient's surgical history solely indicated a previous laparoscopic resection of the sigmoid due to diverticulitis disease 9 years ago. The last defecation was reported on the previous day without any traces of blood and with normal consistency. Physical examination demonstrated a painful palpation of the upper abdomen with a slightly distended, but soft abdomen. No metallic or high-pitched bowel sounds could be heard. Blood results were normal despite a mild leucocytosis (11 g/L).
A CT of the abdomen was performed and showed a closed-loop obstruction of the small bowel (figure 1).
The subsequent emergency laparoscopy demonstrated an internal hernia with a jejunal loop obstructed in a residual mesenteric gap (figure 2) with signs of fatal ischaemia (figure 3). The resection of the necrotic jejunal loop with subsequent end-to-end anastomosis was performed after laparotomy. Subsequently the residual mesenteric gap was closed by a running suture with a nonresorbable thread (Ethibond 2-0).
Postoperative recovery was uneventful, and the patient was discharged 9 days after surgery.
## Case presentation 2
A female patient in her late 70s presented with recurring abdominal pain during a period of 3 weeks. According to the patient, the pain was located predominantly in the lower abdomen, and she did not observe any association with eating or drinking. There was no nausea and emesis. Her medical history indicated a laparoscopic sigmoid resection 4 years ago due to diverticulitis disease. Furthermore, appendectomy and hysterectomy were performed more than 10 years ago. Physical examination did not indicate any pathological findings; blood results were normal.
Diagnostics via colonoscopy and CT abdomen also did not reveal the origin of complaints. The patient observed irregular bowel movement and therefore constipation was assumed which was treated medically without improvement of her complaints.
Since symptoms did not improve under conservative therapy, diagnostic laparoscopy was performed revealing a large mesenteric gap (figure 4) through which small bowel loops had migrated underneath the descending neocolon to the left side of the abdomen. The intestines were slightly distended, but without signs of acute obstruction. Position of the small bowel was corrected towards the right side and closure of the mesenteric gap was performed from Treitz ligament towards the retroperitoneum on the left side downward to the iliac vessels with a running and non-resorbable suture (Ethibond 2-0).
## Outcome and follow-up
Both patients had an uneventful recovery and were discharged 9 and 8 days after surgery.
Follow-up was administered after 6-8 weeks and displayed a satisfactory outcome without any signs of abdominal discomfort.
# Discussion
In these two case reports, we describe the rare complication of an internal hernia with small bowel obstruction and chronic subacute abdominal pain, respectively, following laparoscopic resection of the sigmoid colon without primary closure of the mesenteric defect.
Laparoscopic resection of the sigmoid colon in diverticulitis is a very common and often performed surgery. The laparoscopic approach is generally preferred due to lower operative trauma which is associated with a decreased incidence of significant adhesions, less postoperative pain, early discharge from hospital and lower morbidity rate than open surgery. [bib_ref] Abdominal adhesions: a practical review of an often overlooked entity, Tabibian [/bib_ref] [bib_ref] The Sigma-trial protocol: a prospective double-blind multi-centre comparison of laparoscopic versus open..., Klarenbeek [/bib_ref] [bib_ref] The evolving role of laparoscopy in colonic diverticular disease: a systematic review, Gaertner [/bib_ref] Nevertheless, it remains a surgical intervention with potential complications such as wound infection, anastomotic leakage, bleeding, ureteral injury, intraabdominal adhesions and internal hernia, with the latter being a rare complication. [bib_ref] Prevalence of internal hernias after laparoscopic colonic surgery, Sereno Trabaldo [/bib_ref] A retrospective data analysis of almost 1300 patients following laparoscopic left-sided colonic resection demonstrates a prevalence of internal hernia with small bowel obstruction and closed loop syndrome of 0.38%. 14 The time between preceding surgery and clinical presentation was 0.3-10 months and thus significantly shorter than in our cases. This finding was confirmed by another analysis of almost 200 patients following left-sided laparoscopic colonic resection reporting a prevalence of 0.5% of acute onset internal hernia requiring subsequent emergency surgery. However, this analysis also described a prevalence of 21% of asymptomatic internal hernias as confirmed by CT within 1 year postoperatively. [bib_ref] High prevalence of asymptomatic internal hernias after laparoscopic anterior resection in a..., Däster [/bib_ref] There are only a few reported cases and reviews of internal hernia following laparoscopic resection of the sigmoid colon in diverticulitis. However, they all demonstrate that clinical presentation can range from asymptomatic to acute with severe bowel obstruction and subsequent ischaemia. [bib_ref] Internal hernia after laparoscopic colorectal surgery: an under-reported potentially severe complication. A..., Portale [/bib_ref] The non-specificity and varying intensity of symptoms of internal hernias after laparoscopic sigmoid resection allow comparison with other surgeries in which internal hernias may also occur. Our group itself has described the problem of internal hernia after laparoscopic inguinal hernia surgery. [bib_ref] Mind the gap -Small bowel obstruction due to preperitoneal herniation following laparoscopic..., Thalheimer [/bib_ref] Also the presence of internal hernia after laparoscopic Roux-Y gastric bypass for morbid obesity is a well-known and diagnostically challenging complication since symptoms can vary widely too. [bib_ref] Closure of mesenteric defects in laparoscopic gastric bypass: a multicentre, randomised, parallel,..., Stenberg [/bib_ref] As demonstrated in the second case, a conclusive diagnosis can often only be made by diagnostic laparoscopy. Since minor laparoscopic surgery nowadays poses a low operative risk, [bib_ref] Laparoscopic entry: a literature review and analysis of techniques and complications of..., Molloy [/bib_ref] an early diagnostic laparoscopy is advisable in unclear cases and has been implemented in our surgical department ever since.
Considering a possible life-threatening situation with incarcerated internal hernia, primary closure of the mesenteric gap between the neocolon descendens and retroperitoneum (figure 5) appears to be reasonable. If a pre-emptive closure of such a mesenteric gap can significantly reduce the incidence of internal hernia after a laparoscopic sigmoid resection can only be clarified by a controlled, randomised study in the future.
In the context of gastric bypass, these data exist and recommend the consistent closure of the mesenteric gaps. [bib_ref] Closure of mesenteric defects in laparoscopic gastric bypass: a multicentre, randomised, parallel,..., Stenberg [/bib_ref] In summary, we describe two cases of the rare complication of internal hernia after laparoscopic sigmoid resection and considering that a potentially complication-prone internal hernia occurs more frequently than assumed, we recommend primary closure of the mesenteric gap.
## Learning points
► Laparoscopic resection of the sigmoid colon in diverticulitis is a very common and often performed surgery. While certain complications such as anastomotic leakage are well known internal hernias are rare and therefore an awareness should be raised. ► These two cases show that symptoms of internal hernias can display a wide range and particularly cases with mild symptoms can dangerously prolong the time until correct therapy is initiated. ► An early diagnostic laparoscopy is advisable in cases with unspecific abdominal symptoms after laparoscopic sigmoid resection. ► Asymptomatic hernias present a constant risk of a possible life-threatening outcome, allowing to recommend a primary closure of the mesenteric gap.
[fig] Figure 1: CT image of the abdomen: arrows point towards the closed loop (small bowel). [/fig]
[fig] Figure 2: Intraoperative image of the mesenteric gap (A) after laparoscopic sigmoid resection. [/fig]
[fig] Figure 3: Closed-loop obstruction of the small bowel with ischaemic jejunal loop (B). [/fig]
[fig] Figure 4: Laparoscopic view at the mesenteric gap (C) underneath the neocolon descendens (D). [/fig]
[fig] Figure 5: Illustration (by L. Plath) of sigmoid resection (1), descending neocolon (2) with mesenteric gap and closure of mesenteric gap by suture (3). [/fig]
|
Efficacy of regular professional brushing by a dental nurse for 3 months in nursing home residents—A randomized, controlled clinical trial
Objectives:The oral health of nursing home residents is poor for various reasons.Many require help for oral hygiene. Regular professional brushing by a dental nurse should improve oral hygiene. This study aimed to determine the efficacy of regular tooth brushing by a dental nurse on the oral health of nursing home residents.Methods: This controlled trial randomized participants (n = 50; mean age 83 ± 8 years)to brushing by a dental nurse every 2 weeks for 3 months (n = 25; test group) or oral hygiene procedures performed/controlled by nursing home staff (n = 25; control group). Personal, general and oral health, as well as various oral hygiene parametersplaque index (PI), gingivitis index (GI), papilla bleeding index (PBI), oral hygiene index (OHI) and Volpe-Manhold Index (VMI)-were evaluated at baseline, after initial professional dental cleaning and before last brushing.Results: At baseline, oral health was impaired according to investigated indices in both groups. After professional brushing for 3 months, there were improvements in PI, GI and PBI, with significant increases compared with the control group in OHI and VMI (P = 0.017 and P < 0.001, respectively). Among the control group, the number of teeth decreased while the root caries index increased (P = 0.002 between groups).Conclusions: Regular professional brushing every 2 weeks by a dental nurse can be recommended for nursing homes residents to improve oral health parameters and to help reduce root caries incidence as a basis to preserve the number of teeth. Such oral hygiene procedures will maintain and improve the oral health of nursing home residents.K E Y W O R D Sdentistry nursing, gerodontology, nursing home, oral hygiene| INTRODUC TI ONThe number of people that depend on care in nursing homes will increase dramatically over the next few decades. In 2009, over two million inhabitants in Germany alone needed such care.1The oral health of nursing home residents has been well documented and is poorer than that of older community-dwelling people. In particular, periodontitis, root caries, dry mouth and existence of prostheses have shown higher prevalence. 2,3 Furthermore, oral care in nursing homes is inadequate, aggravated by the lack of awareness of good 328 |
## | study design and procedures
Personal health parameters were obtained and documented from the medical files of the nursing home, including care dependency level (according to German law concerning the care for the elderly, care dependency is classified into grades 1-5, from 1 = little impairment of independency to 5 = highest impairment of independency with special needs for nursing care); number of months living in the nursing home, mobility and oral hygiene practices (self-brushing with a manual toothbrush or using interdental devices (usually interdental brushes) without supervision by staff, brushing by staff with a manual toothbrush or supervision of resident's brushing with a manual toothbrush by staff); and general health parameters (including cognitive status, prescribed medications, systemic diseases). Study participants as unit of randomization were randomized to either the treatment or the control group.
A block randomization was carried out and the block size varied with a maximal number of four. Afterwards, all participants received a baseline oral examination by the cooperating dentist, followed by professional dental cleaning to reach a standardized level of oral hygiene. An additional oral examination was then performed to document cleaning success and acceptance among study participants. Subsequently, nursing home staff received in-house training regarding oral hygiene knowledge and practice guidelines for nursing home residents.
During the intervention phase, the control group received oral hygiene procedures performed or controlled by the nursing home staff (treatment as usual). In the treatment group, a dental nurse performed one brushing session every 2 weeks. Also every 2 weeks, every study participant received an oral examination; the final examination was performed after 3 months, directly before the final brushing session. A study flow chart, including dropouts, is shown in . The dental nurse was instructed and calibrated according to the educational guidelines for dental students of the Medical Dental University of Cologne.
## | in-house training
Before the start of the intervention phase and after performing professional dental cleaning for all study participants, all nursing home staff received an in-house training session regarding oral hygiene recommendations for daily practice in nursing homes. This was done to provide standardized knowledge regarding providing oral hygiene for residents, supervising resident's brushing or controlling results of self-brushing residents. Residents did not receive any additional training, since clean results were supervised by the educated staff.
First, the cooperating dentist of the nursing home held one educational session regarding the general necessity of regular oral hygiene provision to nursing home residents, and the actual oral hygiene and oral health situation in this nursing home. Second, a standardized lecture regarding oral health and oral hygiene among older people, distributed by the German Society of Gerodontology, 13 was held.
Afterwards, instructional films were shown addressing correct oral hygiene procedures, prosthetic reincorporation and replacement, and their correct cleaning procedures (https://www.bzaek.de/fuer-medien/video-audio.html, last assessed 15 November 2018). Finally, jaw models and toothbrushes were used to perform practical exercises.
## | professional dental cleaning (initial preintervention procedure)
At the cleaning appointment, all participants were visited in their rooms by the dental nurse. First, the nurse prepared the necessary F I G U R E 1 Study flow chart Were asked to parƟcipate in the study and agreed equipment (scaler, brushes, ultrasonic cleaning device, interdental brushes) in a central bathroom. Afterwards, participants walked or were transported to the central bathroom where they were seated and the oral examination was completed. Professional dental cleaning was performed with brushes, ultrasonic cleaning, scalers and interdental brushes until a macroscopic clean situation was reached controlled and documented by the cooperating dentist. If participants refused to continue the procedure, the nurse would stop immediately (nobody did).
## | control group
Among study participants in the control group, besides in-house training of nursing home staff and pretreatment in form of the professional dental cleaning session, no other parameters were changed regarding their oral hygiene practices from before study participation. Oral hygiene practice in the control group was self-brushing with a manual toothbrush or using interdental devices (usually interdental brushes) without supervision by staff, brushing by staff with a manual toothbrush or supervision of resident's brushing with a manual toothbrush by staff.
## | brushing group
One mouth mirror, the toothbrush (Oral B Pro-Expert CrossAction, Schwalbach, Germany) and the toothpaste (Elmex Caries Prevention Professional; GABA, Hamburg, Germany) were prepared prior to the brushing session. Because of its mild taste, we assumed it would be well accepted by the participants, and 1450 ppm provided sufficient fluoride supply. Additionally, interdental brushes were available in different sizes. First, removable dentures were taken out of the mouth and cleaned. In advance, participants rinsed their mouth with water to remove food debris. The dental nurse placed residents on a chair and performed brushing from behind, supporting the resident's head with her arm. Occlusal surfaces were brushed first. For cleaning of the other surfaces, the modified Bass technique was applied. 14 After again rinsing the mouth with water, interdental brushes adapted to interdental spaces were used. The length of denture brushing did not exceed 3 minutes, and dental brushing did not exceed 3 minutes according to timeframes specified from the nursing director of this nursing home for self-brushing, while the use of interdental brushes did not exceed an additional 2 minutes.
## | outcome parameters assessed
## | oral examination
The total number of teeth and prostheses prevalence were documented before professional dental cleaning, before the first dental brushing session, and before and after the last tooth brushing session. The periodontal status according to the community periodontal index of treatment needs (CPITN) [bib_ref] Development of the World Health Organization (WHO) community periodontal index of treatment..., Ainamo [/bib_ref] should also be documented at these timepoints; because of missing participants' compliance, CPITN was only documented before professional dental cleaning.
## | dementia status
The exact diagnosis of dementia was inconsistently documented in the nursing home. For some residents, information on dementia type and grade was documented, but often this information was missing.
In this nursing home, on the day of moving in, residents are assigned to rooms according to the information "dementia yes/no," based on the available diagnoses made by a neurologist. When planning this study, we decided to use this available binary assignment for all participating residents.
## | indices
Plaque index (PI), [bib_ref] Periodontal disease in pregnancy. II. Correlation between oral hygiene and periodontal condition, Silness [/bib_ref] gingivitis index (GI), [bib_ref] Periodontal disease in pregnancy. I. Prevalence and SEverity, Loe [/bib_ref] Quigley-Hein index (QHI) [bib_ref] Reduced plaque formation by the chloromethyl analogue of victamine C, Turesky [/bib_ref] and the papilla bleeding index (PBI) [bib_ref] Motivation and education, Saxer [/bib_ref] were obtained as described in detail elsewhere. The root caries index (RCI) was graded on a scale from RC1 (hard surface) to RC5 (soft surface), 20 and the dental nurse was calibrated according to earlier approaches. [bib_ref] High-fluoride toothpaste: a multicenter randomized controlled trial in adults, Srinivasan [/bib_ref] Oral hygiene was documented by the oral hygiene index (OHI) [bib_ref] Nutrition in a collegiate basic nursing curriculum, Greene [/bib_ref] and the amount of calculus by the Volpe-Manhold Index (VMI). [bib_ref] In vivo calculus assessment. I. A method and its examiner reproducibility, Volpe [/bib_ref] Indices were obtained before professional dental cleaning, before the first brushing and before the last brushing session.
## | xerostomic visual analogue scale (vas)
All participants were asked "How dry is your mouth?" independent from their cognitive status, and answers were recorded as continuous variables from 0 cm = "not dry at all" to 10 cm = "no saliva at all." [bib_ref] Van den Bogaert W. Efficacy of the BioXtra dry mouth care system..., Dirix [/bib_ref] No objective salivation rates were investigated.
## | food debris (vestibulum, upper prostheses, lower prostheses)
Food debris in the vestibulum was documented after rinsing once with water via a six-grade scale. For participants without prostheses, the scale contained from 1 = no food debris, 2 = smallest singular pieces of food debris, 3 = vestibulum covered up to one-third with food debris, 4 = vestibulum covered up to two thirds with food debris, 5 = vestibulum covered completely with food debris and 6 = vestibulum covered up to occlusal surface with food debris. If participants had prostheses, food debris was also documented via a six-grade scale from 1 = no food debris, 2 = smallest singular pieces of food debris not reaching prosthetic teeth, 3 = prosthetic teeth covered up to 25% with food debris, 4 = prosthetic teeth covered up to 50% with food debris, 5 = prosthetic teeth covered up to 75% with food debris and 6 = prostheses completely covered with food debris up to occlusal surfaces. The maximum reached index number was documented and taken for statistical analysis.
# | data analysis
Absolute and relative frequencies are given for qualitative variables, and mean ± standard deviation (SD) are given for quantitative variables. Group differences were tested using unpaired t test or
Fisher's exact test, respectively. According to the analysis pre/ post-mean and SD, mean difference with 95% confidence intervals (CI) or P-values are presented. Regarding index differences between baseline examination, examination before first brushing and examination before last brushing, Wilcoxon signed rank test and Friedmann's test with alpha adjustment were performed. All reported P-values are two-sided and considered statistically significant if lower than 5%; at Friedmann's test after alpha adjustment, P-values were considered statistically significant if lower than 1.7%. All calculations were done with SPSS Statistics 24 (IBM Corp., Armonk, NY, USA). Data were entered twice and reconciled in case of inconsistencies.
## | re sults
## | clinical characteristics
Fifty nursing home residents or their legal advisors provided written informed consent and participated in the study .
Of the participants, 68% were female, the mean age was 83 (SD: 8) years, and the mean time spent living in the nursing home was 8 (SD: 8) months, with a care dependency level of 3 (SD: 1). All residents had medical and dental health insurance. Dementia was diagnosed by their neurologist in 70% of participants, 84%
were mobile, and they used 8 (SD: 5) prescribed daily medications, while 5 (SD: 2) comorbidities were documented from the medical files. Overall, 81% of study participants reported that they would brush their teeth by themselves, 73% with and 100% without diagnosed dementia. However, participants with dementia received help or supervision by the nursing home staff.
Regarding the dental clinical examination, participants had 17 (SD: 9) remaining teeth, the RCI was 1.5 (SD: 1.6), and 78% suffered from periodontitis. The mean xerostomic VAS value was 1 (SD: 2). 14% of participants had total removable prostheses of the upper or lower jaw, 48% had partial removable dentures, and 84% had fixed prosthodontics . In demented residents, the mean PI was 2.7 (SD: 0.4), without any difference compared to non-demented residents 2.4 (SD: 1). There were no differences in clinical characteristics between intervention and control group.
## | differences between residents with and without dementia
Residents with and without dementia showed differences in base- illustrates the changes observed in oral health over 3 months.
## | oral health indices
At baseline, mean indices were high in all patients. Two weeks after professional dental cleaning in all participants, but before the first interventional brushing, the indices were similar or had been reduced.
After 3 months of professional brushing, there were significant reductions compared with baseline in PI (P = 0.027), GI (P = 0.008), OHI (P < 0.001) and VMI (P < 0.001). In the control group, significant reductions compared with baseline were recorded for GI (P = 0.013), OHI (P < 0.001) and VMI (P < 0.001). Significant between-group differences were noted for OHI before last brushing (P = 0.017) and VMI before last brushing (P < 0.001), in favour of the intervention group .
## | vestibulum index
Regarding the investigated food debris values, no differences could be shown between treatment and control groups, nor between different time points during the study (Supplementary file 1). TA B L E 2 Investigated oral health indices at baseline, before first brushing and before the last brushing session Brushing group, mean (SD)
## | number of teeth and root caries index
## Controls, mean (sd) p-value
## F i g u r e 2 mean oral health index (ohi) and volpe-manhold
Index (VMI) values at baseline, before first brushing and before the last brushing session. *P < 0.017 after alpha adjustment; BFB, before first brushing; BL, Baseline; BLB, before last brushing
## | d iscuss i on
Our hypothesis was that regular professional brushing would be efficient in maintaining or improving individual oral health status.
Accordingly, we have shown the beneficial effects of regular professional brushing every 2 weeks by a dental nurse for 3 months on oral health parameters such as plaque index, gingivitis index, oral hygiene index and Volpe-Manhold Index. We even provide some evidence of effects on long-term parameters such as the number of teeth and the incidence of root caries. Our findings agree with those from other studies, which show that regular application of oral hygiene methods, possibly combined with fluoride application, can be successful among nursing home residents to improve oral health. [bib_ref] Improving oral hygiene in the long-term care of the elderly-a RCT, Zenthofer [/bib_ref] [bib_ref] A study on the effects of short-, medium-and long-term professional oral care..., Ueda [/bib_ref] In a study by Ekstrand et al [bib_ref] Development and evaluation of two root caries controlling programmes for home-based frail..., Ekstrand [/bib_ref] oral hygiene, independent of whether remotivation was performed by a dentist or staff educated in dental hygiene and even when no remotivation was performed. In our study, there was also improvement in oral hygiene parameters in the control group, and there is a question whether the success in oral hygiene parameters might be due to the Hawthorne effect 28 or at least be influenced in both groups: that is knowing they are involved in a study might lead to better brushing/supervising by staff but also by residents during the 2 weeks between brushing sessions. Also, the implementation of in-house training-if seen as another intervention-may partly explain these improvements among the control group. However, the staff's interest in this training was very low; only one training session with a study duration of 3 months took place, and there were no changes regarding other parameters such as daily workflow and available time slots for oral hygiene or responsibilities. Thus, we doubt that these effects might have had an impact on the results.
Our findings suggest a small difference regarding root caries development, shown by the RCI, and number of teeth over the short study duration of 3 months, a finding that seems surprising and might be regarded as coincidental due to the small number of teeth lost in the whole population. On the other hand, tooth loss while living in a nursing home is a well-known, quickly-occurring event and our result might cautiously be considered evidence of long-term effects of oral hygiene on parameters such as the number of teeth and root caries incidence. Regularly performed oral hygiene provision not only helps to maintain good oral hygiene but ensures regular supervision of oral health. This additional effect might also lead to reduced tooth loss by early detection of teeth with treatment needs. Certainly, one might ask if it is necessary to have additional staff to brush teeth.
Other studies have reported that in-house-solutions might include greater time provision and better training for care staff. [bib_ref] Barriers and facilitators in providing oral care to nursing home residents, from..., Hoben [/bib_ref] We therefore conclude that with high amounts of oral biofilm in this special patient population, the QHI might be an unnecessary parameter for future studies.
In our study, the xerostomic burden was very low, especially compared to other older populations. [bib_ref] Van den Bogaert W. The influence of xerostomia after radiotherapy on quality..., Dirix [/bib_ref] and 27%-30% in medicated people (a statistically significant higher proportion than in non-medicated populations), 33,34 these results do not seem reliable in light of the high number of residents diagnosed with dementia in our study. In addition, even though statistically significant differences between groups were reached, results were not of clinical relevance. However, when planning the study, it was unknown how many participants would be diagnosed with dementia, and therefore, this VAS was included. If we assume that more participants suffered from xerostomia or hyposalivation, this might also have influenced the differences in root caries that could be reported in our 3-month study period.
There is and will be much discussion regarding the topic of delegation and task reallocation, both issues that have become increasingly relevant in medicine and dentistry. [bib_ref] Systematic review of whether nurse practitioners working in primary care can provide..., Horrocks [/bib_ref] with an oral health assessment that can control oral health and any problems that may occur. Obviously, quality assurance by a dentist in Germany is mandatory since the responsibility of the dentist cannot be delegated.
There are limitations to our study. Firstly, the 3-month study duration is insufficient to provide results regarding long-term efficiency. Furthermore, only one dental nurse provided all services to the nursing home residents. Regarding the external validity of this data, there might be person-dependent differences in quality regarding the treatment success achieved. On the other hand, with the same dental nurse performing all examinations, inter-examiner differences were eliminated. Although the dental nurse in the whole-study procedure was not blinded, the dentist performing all clinical oral examinations was supposed to be blinded.
However, due to possible conversation with the residents before and after the examinations, blinding could not always be guaranteed, which is a considerable risk of bias. Also, there might be differences between the test and control groups that were not investigated in this study, such as equipment, practices, outcome quality and daily cognitive condition of residents, which may be a potential risk of bias that should be taken into account when interpreting results. Practical experience in our study showed that residents or their caregivers or legal advisors were very positive about the study and all but one person that we asked to participate gave written informed consent. At almost every session, the first question raised was regarding costs. If residents had to pay for the procedures, there might be less enthusiasm to participate in these cleaning sessions; therefore, regarding the external validity we described an optimal scenario that might not be transferrable to a real-life situation. There is a need for further prospective longitudinal research to better evaluate oral hygiene strategies, including different approaches regarding their feasibility and effectiveness to maintain the best possible oral health in residential care setting.
## | con clus ions
Regular additional professional brushing every 2 weeks by a dental nurse may be recommended for nursing homes residents to maintain and improve oral hygiene parameters as possible basis to preserve the number of teeth and reduce root caries incidence. Such oral hygiene procedures may help to maintain and improve oral health of nursing home residents.
## | clini c al rele van ce
## | scientific rationale for study
The oral health of nursing home residents is poor for various reasons. Many require help for oral hygiene.
## | principal findings
Regular additional professional brushing every 2 weeks by a dental nurse can be recommended for nursing homes residents to improve short-term oral health parameters and to help reduce root caries incidence as a basis to preserve the number of teeth.
## | practical implications
Such oral hygiene procedures will maintain and improve the oral health of nursing home residents.
## Ack n owled g em ents
We thank Dr. S. Kottmann as the attending dentist in the nurs- |
Self-Assembly Hydrosoluble Coronenes: A Rich Source of Supramolecular Turn-On Fluorogenic Sensing Materials in Aqueous Media
Water-soluble coronenes, that form nanoparticles by self-association, work as new fluorescent materials by complexation with cucurbit[7]uril, as well as selective turn-on fluorogenic sensors for nitroaromatic explosives with remarkable selectivity, by using only water as solvent.
S elf-assembling of organic dyes is a rich source of nanomaterials for practical applications. [bib_ref] Fluorescent Sensors Based on Aggregation-Induced Emission: Recent Advances and Perspectives, Gao [/bib_ref] Aggregation of perylenediimides has been deeply studied and these studies constitute a paradigm in organic nanoaggregation. The closely related coronenediimide derivatives have been studied from a structural point of view or in the preparation of aromatic acceptors for solar cells and high-end electronics. [bib_ref] Nonfullerene Acceptor Molecules for Bulk Heterojunction Organic Solar Cells, Zhang [/bib_ref] [bib_ref] Fluorine Tuning of Morphology, Energy Loss, and Carrier Dynamics in Perylenediimide Polymer..., Zhu [/bib_ref] Simpler coronenediimides have been used for the preparation of discotic liquid crystals 12 or water-soluble dendrimers. [bib_ref] A Facile Route to Water-Soluble Coronenes and Benzo[ghi]perylenes, Schmidt [/bib_ref] However, the excellent characteristics for π−π stacking or selfassembling make coronenediimides the best choice for molecular recognition studies. We are interested in the preparation of sensing devices for the detection of explosives or toxins in water or vapor. 14−16 To prepare useful new fluorescent nanomaterials, we have designed an easy route to coronenediimide derivatives having an extended aromatic core surrounded by a hydrophilic periphery, suitable for applications as sensing materials. In this paper, we want to introduce their synthesis and self-assembling characteristics in comparison to related hemicoronene-or perylenediimides. Their unique applications as discrete nanoparticles to supramolecular turn-on fluorescent recognition and sensing of nitroaromatic explosives in water will be presented.
Our synthesis started by the Suzuki reaction of the dibromoperylenediimide 1 with two equivalents of a N-Boc protected piperazinyl-pyrimidine boronic ester 2 in conditions used for related Suzuki reactions [fig_ref] Figure 1: Synthesis of hemicoronene-and coronenediimides [/fig_ref]. [bib_ref] Surface functionalized silica nanoparticles for the off−on fluorogenic detection of an improvised..., García-Calvo [/bib_ref] [bib_ref] Synthesis of a Tetracorannulene-perylenediimide That Acts as a Selective Receptor for C60..., García-Calvo [/bib_ref] The bissubstituted, four N-Boc protected derivative 3 (85% yield) was obtained after workup of the reaction. Irradiation of 3 in dichoromethane (DCM) under visible light (halogen lamp, 50W, 4 cm distance) and air for 7 h gave the N-Boc protected coronene 4 (89% yield). Traces of an unexpected intermediate of cyclization were also detected. The monocyclized intermediate product 5 (76%) was subsequently obtained as the main product under a shorter irradiation time . N-Boc deprotection of all compounds by treatment with trifluoroacetic acid (TFA) in DCM for 20 min quantitatively gave the unprotected compounds 6, 7, and 8, bearing four secondary amine groups on the periphery [fig_ref] Figure 1: Synthesis of hemicoronene-and coronenediimides [/fig_ref]. Extended hemicoronediimides from monosubstituted perylenediimides have shown interest as semiconducting materials. After study of the physicochemical characteristics of the obtained compounds, we realized that the N-Boc protected compounds 3, 4, and 5 were bright fluorescent compounds, soluble in common organic solvents, that showed high quantum yields (0.3−0.6 in DCM or CHCl 3 ) and lifetimes similar to the starting material 1 (3−10 ns) [fig_ref] Figure 2: AFM images of samples deposited on mica sheets from water solutions 0 [/fig_ref] , and S29). However, the unprotected compounds 6, 7, and 8 were almost non fluorescent compounds, soluble in water but almost insoluble in most organic solvents [fig_ref] Figure 3: Fluorescent titration curves and titration profiles of 10 μM solutions of 7... [/fig_ref] , and S95). As an exception, 6 initially gave a nonfluorescent solution in water, but the water solution became brightly fluorescent after 24 h. The kinetic study showed the appearance and continuous growth of a fluorescent band at 502 nm and a shoulder at 520 nm after dissolving 6 in water [fig_ref] Figure 3: Fluorescent titration curves and titration profiles of 10 μM solutions of 7... [/fig_ref]. Then 6, 7 and 8 showed the presence of stable spherical nanoparticles in AFM by deposition of 2 μL of aqueous solutions of samples on mica sheets and evaporation of water (tapping mode, force constant 2.8 N m −1 , room temperature, scan rate 1−2 lines per second) [fig_ref] Figure 2: AFM images of samples deposited on mica sheets from water solutions 0 [/fig_ref]. Because of the large ability of these compounds to selfassociate in water, NMR spectra were not sufficiently informative, especially for 7; therefore, we prepared a derivative of 7 with long hydrophilic tails that prevented selfassociation, 9 [fig_ref] Figure 2: AFM images of samples deposited on mica sheets from water solutions 0 [/fig_ref] , which permitted a full characterization by spectroscopy, confirming the structures. The watersoluble nanoparticles from 6, 7, and 8, were also characterized by DLS [fig_ref] Figure 4: Fluorescent titration curves and ratiometric titration profiles of 10 μM solutions in... [/fig_ref] , S59−S60, and S98−S101).
Low molecular weight organic nanoparticles in water constitute an excellent material for studies of aggregation/ disaggregation linked to fluorescence variations on the way to new supramolecular sensing devices. Therefore, we studied in deep disaggregation mechanisms based on molecular recognition. The piperidine/piperazine groups on the periphery of the compounds are expected to be good guests for host−guest chemistry of cucurbiturils in water. [bib_ref] Cucurbit[8]uril (CB[8])-Based Supramolecular Switches, Pazos [/bib_ref] [bib_ref] Cucurbit[n]uril-Based Microcapsules Self-Assembled within Microfluidic Droplets: A Versatile Approach for Supramolecular Architectures..., Liu [/bib_ref] Cucurbiturils have been used for the preparation of supramolecular luminescent sensors [bib_ref] Supramolecular Luminescent Sensors, Mako [/bib_ref] or for enhancing fluorescence of perylenediimides in water. [bib_ref] Strongly Fluorescent, Switchable Perylene Bis-(diimide) Host−Guest Complexes with Cucurbit[8]uril In Water, Biedermann [/bib_ref] Therefore, we performed several tests by adding aqueous solutions of cucurbit[n]urils, CB [bib_ref] A Water-Soluble, NIR-Absorbing Quaterrylenediimide Chromophore for Photoacoustic Imaging and Efficient Photothermal Cancer..., Liu [/bib_ref] , CB [bib_ref] Coronene-Containing N-Heteroarenes: 13 Rings in a Row, Endres [/bib_ref] , CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] , and CB [bib_ref] A Helicene Nanoribbon with Greatly Amplified Chirality, Schuster [/bib_ref] , in 1:1, 1:5, 1:10, and 1:20 dye/cucurbituril molar proportions to 10 μM aqueous solutions of 6, 7, and 8. Except The fluorescent titration curves and titration profiles of 10 μM solutions of 7 and 8 in water, with CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] , showed asymptotic increases of fluorescence when a large excess of CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] was added [fig_ref] Figure 3: Fluorescent titration curves and titration profiles of 10 μM solutions of 7... [/fig_ref]. Additional UV−vis absorption titrations are shown in Figures S62 and S109. In this way, 7 and 8 solutions in water became brightly fluorescent in the presence of a large excess of CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] , expanding the narrow range of fluorescent hemicoronene/coronenediimides in water solutions. CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] is expected to host small molecular guests; [bib_ref] Noninnocent Role of Na+ Ions in the Binding of the N-Phenyl-2-naphthylammonium Cation..., Thomas [/bib_ref] therefore, in this series, the complexation with CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] should happen by the piperidine/piperazine groups,giving rise to disaggregation to form individual complexes in solution. The disaggregation effect worked in the opposite way in the case of 6, where the presence of excess CB [bib_ref] A Water-Soluble, NIR-Absorbing Quaterrylenediimide Chromophore for Photoacoustic Imaging and Efficient Photothermal Cancer..., Liu [/bib_ref] , CB [bib_ref] Coronene-Containing N-Heteroarenes: 13 Rings in a Row, Endres [/bib_ref] , or CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] in 10 μM water solutions of 6 hindered the development of fluorescence [fig_ref] for 6 ,: compounds 7 and 8 showed a neat increase of fluorescence in the... [/fig_ref]. The type of complex formed between CB[7] and 7 or 8 was studied by Job's plot experiments and isothermal titration calorimetry (ITC) measurements, but the results were inconclusive. Instead, accurate mass spectrometry measurements afforded high resolution m/z peaks of the lowest terms in the series of expected complexes, 7@CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] (m/z 2040.7227) and 8@ CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] (m/z 2043.7303), a low resolution peak of the complex 7@2CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] (m/z 3208.1), and self-associated compounds such as [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] 2 (m/z 1757.6), [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] 3 (m/z 2634.9), and [7] 4 (m/z 3515.9) (Figures S51−S53 and S92−S94). With these results, we looked for new applications of the aggregation/ disaggregation mechanism that could afford light on the mechanism as well as new sound applications in sensing. In the [fig_ref] Figure 4: Fluorescent titration curves and ratiometric titration profiles of 10 μM solutions in... [/fig_ref]. In addition, the lifetime decay changed from an initial value of 4.95 to 4.62 ns after addition of TNB or TNT to solutions of 7 in water, but there were no significant changes in the absorbance values under the same conditions. The evolution of changes seems to be correlated to selective disaggregation processes, which produced ratiometric changes in fluorescence, with increases of the final red fluorescence from almost 0 to 2/3%. From the titration plots [bib_ref] Fluorescence Enhancement from Nitro-Compound-Sensitive Bacteria within Spherical Hydrogel Scaffolds, Kim [/bib_ref] [bib_ref] A single molecular fluorescent probe for selective and sensitive detection of nitroaromatic..., Zhang [/bib_ref] but in this case with the unprecedented feature of using only water in the titration. We also calculated the LODs within the values measured (different than 0) by adjusting the initial values to a mean square linear regression and using the R program (SI page S114), which can be considered as a good approach to the limits of quantification of the system. By this way, measured values were LOQ = 0.046 μM for TNB and LOQ = 0.21 μM for TNT. With a solubility of 100 mg/L for TNT (0.44 mM) (SI, page S115) in pure water or seawater at 20°C, the LOQs found are suitable for the detection of traces of TNT in environmental aquatic samples by a turn-on fluorescence mechanism. The system can also be applied to TNB, which is roughly as three-times more soluble in water/seawater than TNT (SI, page S115). Indeed, 7 was not sensitive to the presence of common cations, anions, or usual interferents found in water [fig_ref] Figure 2: AFM images of samples deposited on mica sheets from water solutions 0 [/fig_ref] ; therefore, its presence did not interfere the TNB/TNT detection. Attempts of determining binding constants from the fluorescence titration experiments gave the best fitting from the titration of 7 and TNT for a 1:2 model (supramolecular.org), being K 1 = (2.3 ± 1.3) × 10 5 M −1 and K 2 = (3.4 ± 1.0) × 10 3 M −1 the apparent formation constants of the 7/TNT (1:1) and 7/(TNT) 2 (1:2) complexes, respectively. Because of the large uncertainties, they have only a qualitative value, indicating that multiple association accounted for the disaggregation of 7 in the presence of TNT. Aggregation of 7 was studied by ITC measurements . The binding isotherm was fitted by a dimer dissociation model (Nano-Analyze Software, TA Instruments). We obtained thermodynamic parameters, K agg = (1.39 ± 0.4) × 10 4 M −1 , ΔH = −66 ± 2 kJ mol −1 , and ΔS = −142 ± 9 J mol −1 K −1 , that agreed with the aggregation process.
To understand the interactions, we performed theoretical DFT calculations (SI, p S76) of the complexes between two self-aggregated 7, and then 7 with two CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] or one TNB.
The results showed that dimer [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] 2 had stabilization energy of −52.24 kcal/mol with respect to the two isolated molecules. The most stable calculated structure for the complex 7@2CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] was 7@2(apical)CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] , followed by 7@2(equatorial)CB [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] In conclusion, we have prepared a new series of perylene-, hemicoronene-, and coronene diimides that were soluble in water, giving nanoparticles by self-association, which in turn worked as new fluorescent materials by self-aggregation or complexation with cucurbit [bib_ref] Perylene Diimide-Embedded Double [8]Helicenes, Liu [/bib_ref] uril, as well as selective turn-on fluorogenic sensors for explosive trinitroaromatic compounds, with remarkable selectivity, by using only water as solvent.
[fig] for 6 ,: compounds 7 and 8 showed a neat increase of fluorescence in the presence of CB[7] (Figures S45, S61, and S102). Then we performed fluorescent titrations of 10 μM solutions of 7 and 8 in water by adding increasing amounts of concentrated solutions of CB[7] (Figure 3). [/fig]
[fig] Figure 1: Synthesis of hemicoronene-and coronenediimides. [/fig]
[fig] Figure 2: AFM images of samples deposited on mica sheets from water solutions 0.1 μg/mL: (a) 6 (b) 7 (c) 8. (d) Derivatization of coronene 7 to 9.initial tests, 7 or 8 were not sensitive to common cations, anions, acids or oxidants in water, but 7 showed a dramatic appearance of red fluorescence in the presence of 1,3,5trinitrobenzene (TNB), a common explosive. Consequently, the study was extended to trinitrotoluene (TNT), a commonly used explosive that is usually detected by quenching of fluorescence from suitable fluorophores 27−32 with very few exceptions.33−35 For this reason, TNT lacks a turn-on fluorogenic method for its detection in water with practical use. Taking into account the large number of sunken warfare materials still existing in the oceans from the Word Wars I and II,36 the design of new fluorogenic sensing materials for the detection of TNB/TNT in water is worthy of study. Titration of a 10 μM solution of 7 in water with TNB, 0 to 6 mg/mL TNB, showed a decrease in a 690 nm far red band and an increase of a 632 nm red band in fluorescence with the addition of increasing amounts of TNB. Analogously, titration of a 10 μM solution of 7 in water with TNT, 0 to 1.77 mg/mL TNT, also showed a decrease in the 700 nm band and an increase of a 595 nm band in fluorescence with the addition of increasing amounts of TNT (Figure 4). The fluorescent titration curves and ratiometric titration profiles of 10 μM solutions in water of 7 with increasing amounts of TNB and TNT are shown in [/fig]
[fig] Figure 3: Fluorescent titration curves and titration profiles of 10 μM solutions of 7 and 8 in water with increasing amounts of CB[7], 0 to 100 equiv CB[7] for titration of 7 and 0 to 400 equiv CB[7] for titration of 8. Inset: solution samples of 7 and 8 under UV light, 365 nm, before and after titrations with CB[7]. [/fig]
[fig] Figure 4: Fluorescent titration curves and ratiometric titration profiles of 10 μM solutions in water of 7 with increasing amounts of TNB and TNT, with expansions of the first parts of the titration plots. Insets: water solution samples of 7 under UV light, 365 nm, before and after titration with TNB/TNT. [/fig]
[fig] Figure 5: (a) Calculated structure of dimer[7] 2 seen from a parallel plane to the structure and (b) from a perpendicular plane. (c) Calculated structure of complex 7@2(apical)CB[7], relative energy (0 kcal/mol), and (d) 7@2(equatorial)CB[7], relative energy (25.16 kcal/mol). (e) Calculated structure of complex 7·TNB seen from a parallel plane to the aromatic structure or (f) [/fig]
[table] ■ ASSOCIATED CONTENT: * sı Supporting Information The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.orglett.1c03175. Compound structures (XYZ) Synthetic procedures, complete characterization of all compounds, additional experiments, theoretical calculations (PDF) Tomás Torroba − Department of Chemistry, Faculty of Science, University of Burgos, 09001 Burgos, Spain; Institut de Química Teorica i Computacional (IQTCUB), Universitat de Barcelona, 08028 Barcelona, Spain; orcid.org/0000-0002-5018-4173; Email: ttorroba@ ubu.es [/table]
|
Hyperspectral image compressed processing: Evolutionary multi-objective optimization sparse decomposition
In the compressed processing of hyperspectral images, orthogonal matching pursuit algorithm (OMP) can be used to obtain sparse decomposition results. Aimed at the time-complex and difficulty in applying real-time processing, an evolutionary multi-objective optimization sparse decomposition algorithm for hyperspectral images is proposed. Instead of using OMP for the matching process to search optimal atoms, the proposed algorithm explores the idea of reference point non-dominated sorting genetic algorithm (NSGA) to optimize the matching process of OMP. Take two objective function to establish the multiobjective sparse decomposition optimization model, including the largest inner product of matching atoms and image residuals, and the smallest correlation between atoms. Utilize NSGA-III with advantage of high accuracy to solve the optimization model, and the implementation process of NSGA-III-OMP is presented. The experimental results and analysis carried on four hyperspectral datasets demonstrate that, compared with the state-of-the-art algorithms, the proposed NSGA-III-OMP algorithm has effectively improved the sparse decomposition performance and efficiency, and can effectively solve the sparse decomposition optimization problem of hyperspectral images.OPEN ACCESSCitation: WANG L, WANG W (2022) Hyperspectral image compressed processing: Evolutionary multiobjective optimization sparse decomposition. PLoS ONE 17(4): e0267754. https://doi.org/10.much information as possible from as little data as possible, combine compression and sampling at the sampling end, and reconstruct the original signal with high precision at the reconstruction end. Random sampling is performed at a lower sampling rate than the Nyquist sampling frequency, and reconstruct the original signal with optimization algorithms. The sampling of signals can break through the limitations of the traditional Nyquist law and greatly reduce the amount of data, but its application premise is that the signal meets the sparseness or compressibility conditions.The first task of applying compressed sensing theory to process hyperspectral images is achieving sparse decomposition. Tensor decomposition are used to take advantage of the low rank and sparsity of images to obtain sparse representation of hyperspectral images, to realize the hyperspectral image unmixing [9], fusion [10] and restoration[11]. Sparse decomposition can also be used for image denoising[12].Hyperspectral image is a complex signal with multiple structural components, and it is difficult to effectively sparsely represent it using orthogonal transform. Increasing the number of atoms to form a redundant dictionary is conducive to sparse representation of complex signals. Among the many sparse decomposition algorithms, Orthogonal Matching Pursuit (OMP)[13,14]is a commonly used sparse decomposition algorithm for compressed reconstruction.The redundant dictionary is used to sparsely represent the hyperspectral image signal, which can improve the accuracy of sparse decomposition, but it also brings about problems. The high complexity of the algorithm poses new challenges to the computational efficiency of the sparse decomposition algorithm. In fact, sparse decomposition is a process of objective function optimization, so we consider using intelligent optimization algorithms to complete the solution.We construct a multi-objective sparse decomposition optimization model that takes the largest inner product of matching atoms and image residuals, and the smallest correlation between atoms as the optimization goal. Aiming at this multi-objective sparse decomposition optimization model, the multi-objective evolutionary algorithm[15][16][17]is good choice to solve it. Considering the accuracy and efficiency, the non-dominated sorting genetic algorithm (NSGA-III)[18,19]is explored to complete the solution of optimization model. An improved sparse decomposition algorithm with the help of OMP and NSGA-III is proposed, denoted as NSGA-III-OMP. The innovation of this paper is, abandoning the traditional optimal atom matching method, choosing the optimal atom selection using the evolution process of NSGA-III, so that the atom describing the complex hyperspectral image can be found more accurately.The remainder of the manuscript is organized into six sections. In Section 2, the mathematical model, including the Gabor dictionary basis construction and the multi-objective sparse decomposition optimization model are presented. In Section 3, the NSGA-III method is introduced firstly, with its selection operator, crossover operator, mutation operator, and its steps are illustrated. The NSGA-III-OMP is described and the implementation process of algorithm are shown. In Section 4, experimental datasets and evaluation metrices are presented. The parameter selection is described in Section 5, with results and comparisons shown in Section 6. Section 7 is about conclusion.Multi-objective sparse decomposition optimization modelConstruction of redundant dictionaryThe sparse representation ability of single orthogonal basis cannot satisfy image signals with complex features, therefore, redundant dictionary composed of non-orthogonal atoms is used for sparse decomposition and representation.PLOS ONEHyperspectral image compressed processing and evolutionary multi-objective optimization PLOS ONE | https://doi.
# Introduction
Hyperspectral images (HSIs) contain rich spatial geometric information and spectral feature information, which are suitable for target detection and recognition, image classification and other fields [bib_ref] Low-Rank and Sparse Matrix Decomposition With Orthogonal Subspace Projection-Based Background Suppression for..., Yang [/bib_ref] [bib_ref] Graph Convolutional Networks for Hyperspectral Image Classification, Hong [/bib_ref] [bib_ref] Hyperspectral Images Denoising via Nonconvex Regularized Low-Rank and Sparse Matrix Decomposition, Xie [/bib_ref]. The expansion of the application field proposes complete spatial resolution and spectral resolution, needs to reduce the dimensionality [bib_ref] Unsupervised Dimensionality Reduction for Hyperspectral Imagery via Local Geometric Structure Feature Learning, Shi [/bib_ref] [bib_ref] Dimensionality Reduction of Hyperspectral Imagery Based on Spatial-Spectral Manifold Learning, Huang [/bib_ref] [bib_ref] A Review of Virtual Dimensionality for Hyperspectral Imagery, Chang [/bib_ref] of hyperspectral data. In the process of dimensionality reduction, how to ensure the integrity of information is a big problem. Furthormore, sharp increase in the amount of information, further brings challenges to data storage and transmission. The traditional Shannon sampling theorem is no longer suitable for large data volume hyperspectral image processing.
Compressed Sensing (CS) theory [bib_ref] Robust Uncertainty Principles: Exact Signal Reconstruction from Highly Incomplete Frequency Information, Candès [/bib_ref] was proposed in 2006, with the characteristics of directly collecting data information. The essence of compressed sensing theory is to extract as Using the Gabor function to generate a redundant dictionary for sparse representation of the signal can ensure the global information of the signal and the intensity of the signal change in any local time. The generation function of the Gabor dictionary is [bib_ref] Robust Face Recognition with Kernelized Locality-Sensitive Group Sparsity Representation, Tan [/bib_ref] ,
[formula] g g n ð Þ ¼ 1 ffi ffi s p win n À u s � � cos un þ o ð Þð1Þ [/formula]
Where, win n ð Þ ¼ e À pn 2 is the Gaussian window function, n = 0,1,. . ., N, N is the signal length, γ = (s,u,υ,ω) is time-frequency parameter. Let D = {g γ } γ2Γ denote the redundancy dictionary, in which g γ is the atom defined by the parameter group γ, Γ is the set of parameter groups. The time-frequency parameters are discretized according to the following methods: g ¼ a a 1 ; a 2 a a 1 Du; a 3 a À a 1 Du; a 4 Do ð Þ, whereas, a = 2, Δu = 1/2, Δυ = π, Δω = π/6, 0 < α 1 � log 2 N, 0 � a 2 � 2 À a 1 þ1 N, 0 � a 3 < 2 a 1 þ1 , 0 < α 4 � 12. The Gabor dictionary is highly redundant, and the number of atoms is N atom = 52(Nlog 2 N + N − 1).
## Sparse decomposition of omp
The basic idea of OMP for sparse decomposition is to find the optimal atoms from the redundant dictionary that can best match the image signal. The calculation process is as follows.
First, select the atom g γ0 that best matches the image signal f to be decomposed from the redundant dictionary, and meet the following conditions,
[formula] j f; g g0 � � j ¼ max g2G j f; g g � � jð2Þ [/formula]
Where, h- i represents the inner product. The image can be decomposed into two parts: the component on the best atom g γ0 and the residual, namely,
[formula] f ¼ f; g g0 � � g g0 þ R 1 fð3Þ [/formula]
Where, R 1 f is the residual error after matching the original image with the best atom. Continuously carry out the above decomposition process on the residual,
[formula] R k f ¼ R k f; g gk � � g gk þ R kþ1 fð4Þ [/formula]
Where, R k f; g g k D E represents the component of the image residual R k f on the corresponding atom g g k .
After K times decomposition, the image is decomposed into,
[formula] f ¼ X KÀ 1 k¼0 R k f; g g k D E g g k þ R K fð5Þ [/formula]
Under the condition that the signal satisfies the limited length, the exponential decay kR k fk is becomes zero with the increase of k. A small number of atoms can represent the main components of the image, namely,
[formula] f � X KÀ 1 k¼0 R k f; g g k D E g g kð6Þ [/formula]
Where, K << N.
Assume the optimal atom set is Θ K , the original signal could be represented by the K atoms, denoted as,
[formula] f ¼ Θ K θð7Þ [/formula]
Where, θ is the sparse coefficient vector. Utilize least square method to solve the sparse coefficient vector,
[formula] θ ¼ Θ þ K fð8Þ [/formula]
The reconstructed signal could be expressed as,
[formula] f ¼ Θ K Θ þ K fð9Þ [/formula]
The decomposition steps of OMP can be summarized in Algorithm 1.
[formula] ¼ max g2G j R k f; g g � � j. [/formula]
4: Update the optimal atoms set with g γk to obtain
[formula] Θ k = Θ k−1 S g γk . 5: Update the residual with R k f ¼ f À Θ k Θ þ k f, where Θ þ k [/formula]
represents the pseudo-inverse of the matrix Θ k . 6: Compute k = k + 1. 7: end 8: Compute the reconstructed signalf according to Eq [bib_ref] Sparse and Low-Rank Constrained Tensor Factorization for Hyperspectral Image Unmixing, Zheng [/bib_ref].
## Sparse decomposition optimization model
Analyzing the sparse decomposition process of the OMP algorithm, we find that the selection of the optimal atom is completed by calculating the inner product of all the atoms in the redundant dictionary and the decomposition residuals. In order to improve the accuracy and efficiency of sparse decompositionprocess, this paper takes, 1) the maximum inner product of the matching atom and the image residual, and 2) the minimum correlation between the atoms, as the two optimization goal, to construct a multi-objective sparse decomposition optimization model.
According to Algorithm 1, after k − 1 sub-decomposition, the residual is R k−1 f, and the optimal atom set is Θ k−1 , then in the kth decomposition process, the first objective function that needs to be optimized is the maximum inner product of matching atoms and residuals. The first objective function is computed as,
[formula] f 1 g g À � ¼ max g2G j R kÀ 1 f; g g � � jð10Þ [/formula]
In order to ensure the orthogonality between atoms and make the decomposition process more accurate, the second objective function that needs to be optimized is the minimum correlation between atoms. The second objective function is computed as,
[formula] f 2 g g À � ¼ min g2G j Θ kÀ 1 ; g g � � jð11Þ [/formula]
In order to ensure the unity of the multi-objective solution process, taking the minimum of each objective function as the optimization objective. The multi-objective sparse decomposition optimization model is expressed as,
[formula] f ¼ min g2G 1 j R kÀ 1 f; g g j � � ; min g2G j Θ kÀ 1 ; g g � � j " #ð12Þ [/formula]
Proposed NSGA-III-OMP algorithm
The basic idea of NSGA-III-OMP algorithm is: in the iterative process of OMP sparse decomposition, the NSGA-III evolution process is used to replace the original atom matching process. The main advantage of the algorithm NSGA-III is that, by defining a hyperplane and multiple reference points, it selects the individuals which can best adapt the environment, in order to maintain the diversity of the population and ensure that decision makers can find the optimal solution. Firstly, we describe the NSGA-III algorithm in detail, including the initilization, selection, the crossover and mutation operation. Following that, the flow chart and implementation of NSGA-III-OMP is at the last.
## Nsga-iii algorithm
In view of the characteristics of the evolutionary algorithm to solve the multi-objective sparse decomposition model, there is no need to generate a Gabor dictionary in advance, and a realvalued encoding method is used to construct chromosomes, and each chromosome represents an atom. According to the generating formula of Gabor atom, the length of chromosome is D = 4, and its value is the vector represented by (s,u,υ,ω). Suppose the population size is M, the maximum evolutionary generation is G max , and the number of objective functions is P. Genetic operator is the key to ensure the optimization of chromosome update, usually including selection, crossover and mutation [bib_ref] Rerounting Path Network Planning under Dangerous Weather, Wang [/bib_ref] [bib_ref] Optimization of Wind-Thermal Economic-Emission Dispatch Problem using NSGA-III, Amorim [/bib_ref].
Initialization operation. The initial parent population X par 0 is generated in a random manner. The mth element in X par 0 is expressed in Eq (13),
[formula] X par 0 n; m ð Þ ¼ 1 ffi ffi s p e À p nÀ u s ð Þ 2 cos un þ o ð Þð13Þ [/formula]
Where, (s,u,υ,ω) is the random number satisfying the range of specific parameter, m = 1,2,. . ., M. Selection operation. The selection mechanism of the NSGA-III algorithm [bib_ref] An Improved NSGA-III Algorithm using Genetic K-Means Clustering Algorithm, Liu [/bib_ref] is different from the traditional genetic algorithm. The specific steps include: non-dominant sorting, define the reference point, normalize objective function, association operation and filter operation.
Before the first selection operation, crossover and mutation operations are performed on the initial parent population X par 0 to generate offspring population X spr 0 . Merge the parent population X par 0 and the offspring population X spr 0 to form a composite population
[formula] X com 0 ¼ X par 0 S X spr 0 . 1) Non-dominant sorting. [/formula]
Compute the fitness and take the top M members from the composite population X com 0 . The individuals of the composite population X com 0 are divided into L layers according to the non-dominated rules [bib_ref] An Evolutionary Many-Objective Optimization Algorithm using Reference-Point-Based Nondominated Sorting Approach, Part I:..., Jain [/bib_ref]. The first L − 1 layers of individuals is denoted as X sel and the number is |X sel |. If |X sel | = M, the next parent population is started with X par Gþ1 ¼ X sel , else we need to select other M − |X sel | members from the Lth layer.
2) Define the reference point. We employ an efficient technique that spaces points on a normalization hyperplane with an intercept of one on each axis and is equally inclined to all objective axes. The number of objective functions is P, the length of chromosome is D, then the number of reference points is J ¼ C D PþDÀ 1 , j = 1,2,. . .,J, and the reference points can be set on the normalized hyperplane of dimension P − 1. We believe that the optimal solutions are widely scattered across the complete normalized hyper-plane.
3) Normalize objective function. Calculate the ideal point in population X sel by recognizing the minimum value b min p for each objective function p = 1,2,. . .,P, and build the ideal point
[formula] b min 1 ; b min 2 ; . . .; b min P � � , then the translated objective is computed by Eq (15), b min p ¼ min f p z ð Þ; z 2 X sel ð14Þ f 0 p z ð Þ ¼ f p z ð Þ À b min pð15Þ [/formula]
The intercept a P of the p-th objective axis and the linear hyper-plane can then be computed according to the reference [bib_ref] An Evolutionary Many-Objective Optimization Algorithm using Reference-Point-Based Nondominated Sorting Approach, Part I:..., Jain [/bib_ref] , and the normalized objective functions can be represented as,
[formula] f n p z ð Þ ¼ f 0 p z ð Þ a p À b min p ¼ f p z ð Þ À b min p a p À b min p ; for p ¼ 1; 2; . . .; Pð16Þ [/formula]
Assume the structured reference points is H s , we utilize Eq (16) to map each reference point onto the normalized hyperplane, and save the reference points in H n .
4) Association operation. Connect the reference point and the origin of hyperplane to form a reference line, calculate the perpendicular distance from all individuals in X sel to the reference lines. Select the nearest reference line, and associate the individual with the corresponding reference point. Through association operation, we associate individuals with reference points. It may appear that a reference point is associated with one or more population individuals, or some reference points are not associated with any population individual. At this point, we need to perform filter operation. 5) Filter operation. Determine the number ρ j of first l − 1 layer individuals associated with each reference point j, identify the reference points set J min = {j:argmin j ρ j } having minimum ρ j . Randomly select a reference point from j 1 , if r j 1 ¼ 0 and j 1 is associated with multiple individuals in lth layer, then select the individual closest to the reference line j 1 O to enter the next generation. If r j 1 ¼ 0 and j 1 is not associated with multiple individuals in lth layer, then randomly select an individual to enter the next generation. Repeatedly select reference points from J min , until the individuals are supplemented [bib_ref] An Evolutionary Many-Objective Optimization Algorithm using Reference-Point-Based Nondominated Sorting Approach, Part I:..., Jain [/bib_ref] and the parent population X par Gþ1 operation is completed.
Crossover and mutation operation. Since we have already performed an elitist selection operator on the parent population, the population diversity could be maintained. Therefore, we apply usual crossover and mutation operators to create the offspring population X spr Gþ1 . According to the crossover probability, it is judged whether the individual has crossover and the gene position where the crossover occurs. The analog binary crossover operator is used to generate the progeny population. According to the mutation probability, determine whether the individual has mutation and determine the gene location where the mutation occurred, and adopt the basic mutation method to determine the individual value after mutation.
NSGA-III process. After continous iteration in NSGA-III evolution process, we obtain the optimal offspring solution set X spr G max . Compute the contribution degree of the individual objective function value in the total objective function value, the individual X optimal (optimal = 1,2,. . .,M) with the smallest value of f optimal is selected as the optimal atom. The smallest value of f optimal is computed as,
[formula] f optimal ¼ X P p¼1 f p X optimal � � À f p X m ð Þ � � min � � f p X m ð Þ � � minð17Þ [/formula]
Where, X m m ¼ 1; 2; . . .; M ð Þ 2 X spr G max . Combined with the description of the above operators, the implementation steps of NSGA-III is summarized in Algorithm 2.
Algorithm 2: NSGA-III Inputs: the evolution generation G max , the initial parent population X par 0 , the objective function f p , p = 1,2,. . .,P, the structured reference points H s . Output: the optimal solution X optimal . 1: Perform crossover and mutation operations to generate the initial offspring population X spr 0 , set iteration G = 0.
[formula] 2: While G � G max do 3: X sel = �, l = 1 4: X com G ¼ X par G [ X spr G 5: Select one individual X l from X com G according non-dominant sorting. 6: repeat 7: X sel = X sel S X l 8: Compute l = l + 1 9: until L = l − 1 10: if |X sel | = M then 11: X par Gþ1 ¼ X sel , break; 12: else 13: X par Gþ1 ¼ X sel 14: [/formula]
Points to be choosen from Lth layer: M − |X sel | 15:
Normalize the objective functions using Eq [bib_ref] Multi-objective Porfolio Optimization of System-of-Systems based on Robust Capabilities, Li [/bib_ref] and save the reference points H n .
## 16:
Associate each member of X sel with a reference point.
## 17:
Excute the filter operator to obtain the parent population X par Gþ1 . 18: endif 19: Perform crossover and mutation operations to generate the offspring population X spr Gþ1 . 20: Compute G = G + 1 21: endWhile 22: Find the optimal solution X optimal using Eq (17) from X spr G max .
## Implementation of nsga-iii-omp
We use the NSGA-III algorithm to replace the atomic matching process in OMP, and propose the NSGA-III-OMP algorithm. The implementation of NSGA-III-OMP is summarized in Algorithm 3. According to Algorithm 2: NSGA-III, with the parameters M and G max , the optimal individual X optimal is obtained as the optimal atom g k . 3: Update the optimal atoms set with g γk to obtain
[formula] Θ k = Θ k−1 S g γk . 4: Update residual with R k f ¼ f À Θ k Θ þ k f, where Θ þ k [/formula]
represents the pseudoinverse of the matrix Θ k . 5: Compute k = k + 1. 6: end 7: Compute reconstructed signalf according to Eq (9).
## Experimental datasets and evaluation metrices
## Hyperspectral datasets
Four hyperspectral datasetswith diverse signatures are included in the investigations, allowing for a complete quantitative and quantitative comparative evaluation of the proposed scheme. The datasets are namely Cuprite1, Cuprite2, Indian Pines collected by AVIRIS and Pavia University collected by ROSIS. The water absorption and noisy bands in the original data set have been removed, and the image has been spatially cropped to the block size B = 16 for computation convenience. The basic conditions of the 4 sets of data are shown in S1 The original image of the 50th band of hyperspectral data is shown in [fig_ref] Fig 1: Original 50th images of four datasets [/fig_ref].
## Plos one
## Evaluation metrices
The performance of the algorithm is evaluated using the average peak signal-to-noise ratio (PSNR) [bib_ref] Compressed Sensing Reconstruction of Hyperspectral Images Based on Spectral Unmixing, Wang [/bib_ref] , average structural similarity (SSIM) [bib_ref] Image Quality Assessment: From Error Visibility to Structural Similarity, Wang [/bib_ref] , average vector-based SNR [bib_ref] Reconstruction of Hyperspectral Imagery from Random Projections Using Multihypothesis Prediction, Chen [/bib_ref] and average vector-based spectral-angle distortion (SAD) [bib_ref] Reconstruction of Hyperspectral Imagery from Random Projections Using Multihypothesis Prediction, Chen [/bib_ref] , running time and reconstructed image. The running software and hardware environment of the experiment is: AMD quadcore CPU, 3.80 GHz, 16 G memory, Matlab2012b.
In the next discussions, we use F = [F 1 , F 2 , . . .F l ,. . .F L ] 2 R N×L to represent the hyperspectral images, N is the pixels of each band image and L is the number of bands, F l is the vector representation of one band image, F[n] is the vector representation of one pixel.
The band-based PSNR measured in dB is defined as,
[formula] PSNR F l ;F l � � ¼ 20 log 10 max F lð [/formula]
Þ ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffiffi
[formula] MSE F l ;F l � � rð18Þ [/formula]
Where, F l andF l are the original and reconstructed band images, max(F l ) is the peak value of
[formula] F l , MSE F l ;F l � � [/formula]
is the mean squared error [bib_ref] Compressed Sensing Reconstruction of Hyperspectral Images Based on Spectral Unmixing, Wang [/bib_ref] and is computed as,
[formula] MSE F l ;F l � � ¼ 1 N F l ÀF l � � � � � � 2 2ð19Þ [/formula]
The band-based SSIM between F l andF l is computed as,
[formula] SSIM F l ;F l � � ¼ 2m 1 m 2 þ C 1 ð Þ 2s 12 þ C 2 ð Þ m 2 1 þ m 2 2 þ C 1 À � s 2 1 þ s 2 2 þ C 2 À �ð20Þ [/formula]
Where, μ 1 and μ 2 are the mean values of F l andF l , σ 1 and σ 2 are the standard deviation values of F l andF l , σ 12 represents the correlation coefficient between F l andF l , C 1 and C 2 are constants related to the dynamic range of the pixel values. The details for these parameters can refer to [bib_ref] Image Quality Assessment: From Error Visibility to Structural Similarity, Wang [/bib_ref]. The average PSNR and average SSIM are calculated by averaging the band-based PSNR and band-based SSIM over all bands of the hyperspectral dataset.
Vector-based SNR measured in dB is defined as,
[formula] SNR F n ½ �;F n ½ � � � ¼ 10 log 10 var F n ½ � ð Þ MSE F n ½ �;F n ½ � � �ð21ÞMSE F n ½ �;F n ½ � � � ¼ 1 L F n ½ � ÀF n ½ � � � � � � � 2 2ð22Þ [/formula]
The vector-based SAD measured in degree is presented as,
[formula] SAD F n ½ �;F n ½ � � � ¼ ff F n ½ � ÀF n ½ � � �ð23Þ [/formula]
The average SNR and average SAD are calculated by averaging the vector-based SNR and vector-based SAD over all spectral vectors in the hyperspectral dataset.
## Parameter selection
The proposed algorithm NSGA-III-OMP needs to set the number of populations, the maximum evolutionary algebra and the number of decompositions (ie, the optimal number of atoms), and the effects of these parameters on the performance of the algorithm are studied separately.
## Evolution parameter
Firstly, the algorithm NSGA-III-OMP is used to sparsely decompose the 40th band image of the 4 groups of hyperspectral data, and the influence of the maximum evolutionary generation, population size and decomposition times on the performance of the algorithm is analyzed. The variation range of population size M is 5~50, the interval is 5, the variation range of the maximum evolutionary generation G max is 5~50, the interval is 5, the variation range of the decomposition number K is 20~100, and the interval is 20. [fig_ref] Fig 2: Influence of evolution parameters on NSGA-III-OMP [/fig_ref] the variation of the average PSNR of the reconstructed image of Cuprite2 with the parameters. When the number of decompositions is 100, the average reconstructed PSNR varies with the maximum evolutionary generation and population size is shown in [fig_ref] Fig 2: Influence of evolution parameters on NSGA-III-OMP [/fig_ref]. Under the same maximum evolutionary generation, the PSNR has a small oscillation with the increase of the number of populations. While under the small population number, with the increase of the evolutionary generation, the PSNR has a slow increase, but the overall change is not large.
When the population size is 10, the influence of the maximum evolutionary generation and decomposition times on the reconstructed PSNR is shown in [fig_ref] Fig 2: Influence of evolution parameters on NSGA-III-OMP [/fig_ref]. When the maximum evolutionary generation is 5, the influence of population size and decomposition times on reconstruction performance is shown in [fig_ref] Fig 2: Influence of evolution parameters on NSGA-III-OMP [/fig_ref]. [fig_ref] Fig 2: Influence of evolution parameters on NSGA-III-OMP [/fig_ref] and 2(c), it can be clearly found that the influence of the number of decompositions on the reconstructed PSNR is much greater than the influence of the maximum evolutionary generation and the number of populations. [fig_ref] Fig 2: Influence of evolution parameters on NSGA-III-OMP [/fig_ref] shows the influence of the maximum evolutionary generation and population size on the reconstruction time when the decomposition number is 100. It is found that the maximum evolutionary generation has a greater impact on the reconstruction efficiency than the population size. The experimental results of the other three groups of hyperspectral images are similar to Cuprite2, considering the reconstruction accuracy and computational complexity, the number of selected population size in NSGA-III-OMP algorithm is M = 10, and the maximum evolutionary generation is G max = 5.
## Decomposition number
The NSGA-III-OMP algorithm and the OMP algorithm are used to sparsely decompose the 40th band image of the hyperspectral data sets, and the algorithm is set to terminate when the number of decompositions reaches 150. [fig_ref] Algorithm 3: NSGA-III-OMP Inputs [/fig_ref] the changes of the reconstructed PSNR obtained by the two algorithms. The thick line is the accuracy of the reconstructed image obtained by the 50 best atoms using OMP. The experimental results show that the NSGA-III-OMP algorithm cannot achieve the reconstruction accuracy of the OMP algorithm with only 50 atoms. This is because: in each decomposition process of the OMP algorithm, it can find the best matching atoms from the redundant dictionary to guarante the performance. Due to the randomness, the atom searched by NSGA algorithm may not be the atom that best matches the residual. Therefore, more atoms need to be found to fully characterize the original image characteristics and achieve the reconstruction accuracy of the OMP algorithm.
If the PSNR of the reconstructed image obtained by the OMP algorithm using 50 atoms is used as the standard, Cuprite1 and Cuprite2 need about 100 atoms to achieve this standard, while Indian Pines and Pavia University need about 150 atoms to achieve the accuracy of the OMP algorithm. Accordingly, the maximum decomposition number of the OMP algorithm is set to K = 50. For the 4 sets of hyperspectral data, the decomposition number for the algorithm NSGA-III-OMP is set to [fig_ref] Fig 1: Original 50th images of four datasets [/fig_ref] ].
# Results and analysis
Three decomposition techniques are compared in order to better understand the possibilities of the proposed NSGA-III-OMP. The first is the OMP algorithm, described in Algorithm 1. This approach searches the best atoms by traversing all the atoms in the redundant dictionary. The second is the particle swarm optimization based on OMP, denoted as PSO-OMP algorithm, described in literature [bib_ref] Sparse Decomposition of Images based on Particle Swarm Optimization, Wang [/bib_ref]. The core idea of this algorithm is within the framework of OMP, achieving the best atoms by evolution of particle swarms. The third one is our proposed algorithm, presented in Algorithm 3. This algorithm is within the framework of OMP, exploring the genetic evolution to search the best atoms. Compared with PSO-OMP, except for the differences in evolutionary principles, our NSGA-III-OMP algorithm also optimizes a multiobjective model shown in subsection 2.3, while PSO-OMP solves a single objective model. In addition, the selection operation adopts a non-dominated sorting to maintain the diversity of population and ensure the accuracy of the optimal solution.
The decomposition number of OMP algorithm is K = 50. With reference [bib_ref] Sparse Decomposition of Images based on Particle Swarm Optimization, Wang [/bib_ref] , the population size of PSO-OMP algorithm is 10, the maximum evolutionary generation is 5, and the decomposition number is K = [100,100,150,150] for four datasets, respectively. The parameters of NSGA-III-OMP algorithm are: the population size is M = 10, the maximum evolution generation is G max = 5, and the decomposition number is K = [100,100,150,150] for four datasets.
## Reconstruction performance comparison
The band-based PSNR of four datasets is shown in [fig_ref] Fig 4: Reconstructed band-based PSNR vs band number using OMP, PSO-OMP and NSGA-III-OMP [/fig_ref] Seen from the figure, the results of PSO-OMP are almost the same as OMP for Cuprite1 and Cuprite2, shown in [fig_ref] Fig 4: Reconstructed band-based PSNR vs band number using OMP, PSO-OMP and NSGA-III-OMP [/fig_ref]. We can clearly find that the proposed NSGA-III-OMP has obvious advantage over PSO-OMP and OMP. In regard to Indian Pines in [fig_ref] Fig 4: Reconstructed band-based PSNR vs band number using OMP, PSO-OMP and NSGA-III-OMP [/fig_ref] , we discover that for first half bands (maybe 1-100 bands), the NSGA-III-OMP is better than PSO-OMP, and OMP is worst result. Whereas, in the last bands (maybe 101-200 bands), the three algorithms obtain the very consistent accuracy. For the dataset Pavia University, we can clearly see the advantages of evolution algorithms, especially our NSGA-III-OMP has absolute advantage over the other two methods. The average PSNR listed in [fig_ref] Table 1: Average PSNR obtained by using OMP, PSO-OMP and NSGA-III-OMP [/fig_ref] can get the same conclusion as in [fig_ref] Fig 4: Reconstructed band-based PSNR vs band number using OMP, PSO-OMP and NSGA-III-OMP [/fig_ref] The average PSNR are computed by the band-based PSNR of four datasets with three methods (see S1 Dataset). These results shows that our algorithm can better find the optimal atoms by solving a multi-objective model using non-dominated sorting selection operation.
The band-based SSIM of four datasets is shown in [fig_ref] Fig 5: Reconstructed band-based SSIM vs band number using OMP, PSO-OMP and NSGA-III-OMP [/fig_ref] For Cuprite1 and Cuprite2 images, the SSIM of PSO-OMP is higher than OMP, which is different from the PSNR comparison in [fig_ref] Fig 4: Reconstructed band-based PSNR vs band number using OMP, PSO-OMP and NSGA-III-OMP [/fig_ref]. Especially, the results of our proposed algorithm NSGA-III-OMP are superior to the other two methods. The average SSIM are presented in [fig_ref] Table 2: Average SSIM obtained by using OMP, PSO-OMP and NSGA-III-OMP [/fig_ref]. The average SSIM are computed by the band-based SSIM of four datasets with three methods (see S2 Dataset). This shows that our algorithm can recover the spatial structural features of the original image, which is a compelling proof of the proposed algorithm's usefulness.
The average SNR and average SAD are shown in [fig_ref] Fig 6: Reconstructed average SNR using OMP, PSO-OMP and NSGA-III-OMP for four datasets [/fig_ref] The average SAD are computed by the vector-based SAD of four datasets with three methods (see S4 Dataset). The comparisons between the original spectral vector and the reconstructed spectral vectors are shown in [fig_ref] Fig 8: Comparison between reconstructed spectral vectors and original sepctral vector [/fig_ref] In this figure, without loss of generality, we choose one spectral vector randomly from the datasets as a representive for comparison. The reconstructed spectral vectors are very consistent with the original spectral vectors, which is an addition proof of the algorithm's avalibility.
The following comparison with several other state-of-the-art algorithms is presented to further showcase the algorithm's performance. The algorithm 3D-DWT [31] uses three dimension wavelet transform for the sparse decomposition of hyperspectral images. The JOMP [31] computes the common support by sparse coding the vector consisting of patches of all bands, leading to high complexity. The FJOMP [32] finds the common support by sparse coding the
## Plos one
training band, improving the computational efficiency. Note that, the Cuptite1 image used in this compariosn is cut from the upper corner to be subimage of size 512×512 pixels for keeping in line with the literatures.
## Plos one
capabilities of orthogonal basis is not enough. For the first 12~20 bands, the PSNR of JOMP and FJOMP are higher than NSGA-III-OMP, while the two algorithms dropped significantly with the band number increasing. In contrast to our algorithm, the performance remains stable and obtains the highest average PSNR. Although the comparison is not detailed enough, the results can still reflect the effectiveness of our proposed algorithm.
In conclusion, numerous comparisons between the proposed NSGA-III-OMP and some state-of-the-art algorithms have demonstrated the reliability of the algorithm. In comparison with the OMP and PSO-OMP algorithms, the results and analysis fully indicate that it is feasible to use NSGA-III to solve the constructed multi-objective sparse decomposition optimization model. The optimal atoms searched by NSGA-III not only can represent the spatial features of the band images, but also describe the spectral features of spectral vectors. These results are able to prove the decomposition accuracy of NSGA-III-OMP.
## Reconstruction efficiency comparison
The runtime of three decomposition algorithms is shown in [fig_ref] Fig 1: Original 50th images of four datasets [/fig_ref] From the histogram, the evolution algorithms, PSO-OMP and NSGA-III-OMP, can greatly improve the efficiency of optimal atoms matching. In terms of efficiency, NSGA-III-OMP has a slight advantage over PSO-OMP. Compared with OMP, the proposed NSGA-III-OMP can increase the calculation speed by an order of magnitude.
## Reconstruction image comparison
As compared to other metrices, a viewer's preference for pleasant visual quality is more relevant. After the data set Cuprite1 is sparsely decomposed, the comparison between the reconstructed image and the original image is shown in [fig_ref] Fig 1: Original 50th images of four datasets [/fig_ref] The figure shows the original image and the 40th band of the reconstructed image. The reconstructed PSNR of OMP algorithm, The comparison between the reconstructed image and the original image of Pavia University is shown in [fig_ref] Fig 1: Original 50th images of four datasets [/fig_ref] The figure shows the original image and the 40th band of the reconstructed image. The reconstructed PSNR of OMP algorithm, PSO-OMP algorithm and NSGA-III-OMP algorithm can reach 39.66dB, 42.96dB and 46.38dB respectively.
Seen from these reconstruction images, the reconstructed image can well describe the detailed features of the original image, which fully demonstrates that the NSGA-III-OMP algorithm can find the optimal atoms by using the evolution process of the chromosome, complete the high-precision sparse decomposition of the image, and fully demonstrate the effectiveness of the algorithm.
# Conclusions
Analyzing the characteristics of hyperspectral images, a sparse decomposition strategy based on NSGA-III-OMP is proposed. The algorithm utilizes the reference point non-dominated sorting genetic method to solve the constructed multi-objective sparse decomposition optimization model. The algorithm uses the chromosomes of the genetic algorithm to simulate the atom matching process in OMP, and explores the selection, crossover and mutation operators to search the optimal atom. The influence of the population size, the maximum evolutionary generation and the decomposition times on the performance of the algorithm is studied, and reasonable parameters are set in the following experiments. Performance comparisons between proposed NSGA-III-OMP algorithm and other algorithms demonstrate that, our proposed algorithm can effectively improve the reconstruction accuracy and computational efficiency of sparse decomposition.
## Supporting information
[fig] Algorithm 3: NSGA-III-OMP Inputs: the original signal f, the atom number K, the parameters of NSGA-III (including the population size M, the maximum evolution generation G max ). Output: the reconstructed signalf. 1: Initialize residual R 0 f = f, the optimal atoms set Θ k = [], number of decompositions k = 1. 2: While k � K do 3: [/fig]
[fig] Fig 1: Original 50th images of four datasets. (a) Cuprite1. (b) Cuprite2. (c) Indian Pines. (d) Pavia University. https://doi.org/10.1371/journal.pone.0267754.g001 [/fig]
[fig] Fig 2: Influence of evolution parameters on NSGA-III-OMP. (a) Influence of maximum evolution generation and population size on reconstructed PSNR, when the decomposition number is 100. (b) Influence of maximum evolution generation and decomposition number on reconstructed PSNR, when the population size is 10. (c) Influence of population size and decomposition number on reconstructed PSNR, when the maximum evolution generation is 5. (d) Influence of maximum evolution generation and population size on reconstructed time, when the decomposition number is 100. https://doi.org/10.1371/journal.pone.0267754.g002 [/fig]
[fig] Fig 3: Reconstructed PSNR vs decomposition number of algorithm OMP and algorithm NSGA-III-OMP. (a) Reconstructed PSNR vs decomposition number for Cuprite1. (b) Reconstructed PSNR vs decomposition number for Cuprite2. (c) Re-constructed PSNR vs decomposition number for Indian Pines. (d) Reconstructed PSNR vs decomposition number for Pavia University. Note: Thick line is the accuracy of the reconstructed image obtained by the 50 best atoms using OMP. [/fig]
[fig] Fig 7: Seen from Fig 6, our NSGA-III-OMP has the highest SNR among the three algorithms, which illustrates our algorithm can recover the spectral vestors accurately. The results in Fig 7 further explained the effectiveness of proposed NSGA-III-OMP, because the spectal angle distortion between the original spectral vector and the reconstructed spectral vector is the smallest. The average SNR are computed by the vector-based SNR of four datasets with three methods (see S3 Dataset). [/fig]
[fig] Fig 4: Reconstructed band-based PSNR vs band number using OMP, PSO-OMP and NSGA-III-OMP. (a) Reconstructed PSNR for Cuprite 1. (b) Reconstructed PSNR for Cuprite 2. (c) Reconstructed PSNR for Indian Pines. (d) Reconstructed PSNR for Pavia University. https://doi.org/10.1371/journal.pone.0267754.g004 [/fig]
[fig] Fig 5: Reconstructed band-based SSIM vs band number using OMP, PSO-OMP and NSGA-III-OMP. (a) Reconstructed SSIM for Cuprite 1. (b) Reconstructed SSIM for Cuprite 2. (c) Reconstructed SSIM for Indian Pines. (d) Reconstructed SSIM for Pa-via University. https://doi.org/10.1371/journal.pone.0267754.g005 [/fig]
[fig] Fig 6: Reconstructed average SNR using OMP, PSO-OMP and NSGA-III-OMP for four datasets. https://doi.org/10.1371/journal.pone.0267754.g006 [/fig]
[fig] Fig 8: Comparison between reconstructed spectral vectors and original sepctral vector. (a) Reconstructed spectral vectors for Cuprite 1. (b) Reconstructed spectral vectors for Cuprite 2. (c) Reconstructed spectral vectors for Indian Pines. (d) Recon-structed spectral vectors for Pavia University. https://doi.org/10.1371/journal.pone.0267754.g008 [/fig]
[fig] Fig 9: Comparison between our proposed NSGA-III-OMP and other three methods. https://doi.org/10.1371/journal.pone.0267754.g009 Fig 10. Reconstructed average SAD using OMP, PSO-OMP and NSGA-III-OMP for four datasets. https://doi.org/10.1371/journal.pone.0267754.g010 PSO-OMP algorithm and NSGA-III-OMP algorithm can reach 40.11dB, 40.15dB and 42.60dB respectively. [/fig]
[fig] S1: Table. Basic condition of four hyperspectral datasets used in experiments. (DOCX) Dataset. Band-based PSNR. (XLSX) S2 Dataset. Band-based SSIM. (XLSX) S3 Dataset. Vector-based SNR. (XLSX) S4 Dataset. Vector-based SAD. (XLSX) Fig 12. Comparison between original image and reconstructed images of Pavia University. (a) Original image. (b) Reconstructed image by OMP. (c) Reconstructed image by PSO-OMP. (d) Reconstructed image by NSGA-III-OMP. [/fig]
[table] PLOS ONE: | https://doi.org/10.1371/journal.pone. [/table]
[table] Table 1: Average PSNR obtained by using OMP, PSO-OMP and NSGA-III-OMP. [/table]
[table] Table 2: Average SSIM obtained by using OMP, PSO-OMP and NSGA-III-OMP. [/table]
|
C1q/TNF‐related peptide 8 (CTRP8) promotes temozolomide resistance in human glioblastoma
The C1q/TNF-related peptide 8 (CTRP8) has recently emerged as a novel ligand of the G protein-coupled receptor RXFP1 in the fatal brain tumor glioblastoma (GBM). We previously demonstrated that the CTRP8-RXFP1 ligand-receptor system promotes motility and matrix invasion of patient GBM and U87 MG cells by specific phosphorylation of PI3 kinase and protein kinase C. Here, we demonstrate a novel role for CTRP8 in protecting human GBM cells against the DNA alkylating damage of temozolomide (TMZ), the standard chemotherapy drug used to treat GBM. This DNA protective role of CTRP8 required a functional RXFP1-STAT3 signaling cascade in GBM cells. We identified N-methylpurine DNA glycosylase (MPG), a monofunctional glycosylase that initiates base excision repair pathway by generating an apurinic/apyrimidinic (AP) site, as a new CTRP8-RXFP1-STAT3 target in GBM. Upon TMZ exposure, treatment with CTRP8 reduced the formation of AP sites and double-strand DNA breaks in GBM cells. This CTRP8 effect was independent of cellular MGMT levels and was associated with decreased caspase 3/7 activity and increased survival of human GBM. CTRP8-induced RXFP1 activation caused an increase in cellular protein levels of the anti-apoptotic Bcl members and STAT3 targets Bcl-2 and Bcl-XL in human GBM. Collectively, our results demonstrate a novel multipronged and clinically relevant mechanism by which the CTRP8-RXFP1 ligand-receptor system exerts a DNA protective function against TMZ chemotherapeutic stress in GBM. This CTRP8-RXFP1-STAT3 axis is a novel determinant of TMZ responsiveness/chemoresistance and an emerging new drug target for improved treatment of human GBM.Abbreviations AP site, apurinic/apyrimidinic site; APE1, AP endonuclease 1; ARP, aldehyde reactive probe; ATM, ataxia telangiectasia mutated kinase; Bcl-2, B-cell lymphoma 2; Bcl-XL, B-cell lymphoma-extra large; BER, base excision repair; cAMP, cyclic 3 0 -5 0 adenosine monophosphate; CI, cell index; CTRP8, C1q/tumor necrosis factor-related peptide 8; dI, deoxyinosine; DNA pol b, DNA polymerase b; DNA, deoxyribonucleic acid; GBM, glioblastoma; KD, knockdown; MGMT, O 6 -methylguanine-DNA methyltransferase; MPG, N-methylpurine DNA glycosylase; MTIC, 3methyl-(triazen-1-yl) imidazole-4-carboxamide; OTM, olive tail moments; PCR, polymerase chain reaction; PI3K, phosphatidyl inositol 3 kinase; PKC, protein kinase C; RLN2, relaxin-2; RTCA, real-time cell analysis; RXFP1, relaxin family peptide receptor 1; siRNA, small interfering RNA; STAT3, signal transducer and activator of transcription 3; TMZ, temozolomide; XRCC1, X-ray repair cross-complementing group 1; cH2AX, phosphorylated histone 2A (Ser 139).
The C1q/TNF-related peptide 8 (CTRP8) has recently emerged as a novel ligand of the G protein-coupled receptor RXFP1 in the fatal brain tumor glioblastoma . We previously demonstrated that the CTRP8-RXFP1 ligand-receptor system promotes motility and matrix invasion of patient GBM and U87 MG cells by specific phosphorylation of PI3 kinase and protein kinase C. Here, we demonstrate a novel role for CTRP8 in protecting human GBM cells against the DNA alkylating damage of temozolomide (TMZ), the standard chemotherapy drug used to treat GBM. This DNA protective role of CTRP8 required a functional RXFP1-STAT3 signaling cascade in GBM cells. We identified N-methylpurine DNA glycosylase (MPG), a monofunctional glycosylase that initiates base excision repair pathway by generating an apurinic/apyrimidinic (AP) site, as a new CTRP8-RXFP1-STAT3 target in GBM. Upon TMZ exposure, treatment with CTRP8 reduced the formation of AP sites and double-strand DNA breaks in GBM cells. This CTRP8 effect was independent of cellular MGMT levels and was associated with decreased caspase 3/7 activity and increased survival of human GBM. CTRP8-induced RXFP1 activation caused an increase in cellular protein levels of the anti-apoptotic Bcl members and STAT3 targets Bcl-2 and Bcl-XL in human GBM. Collectively, our results demonstrate a novel multipronged and clinically relevant mechanism by which the CTRP8-RXFP1 ligand-receptor system exerts a DNA protective function against TMZ chemotherapeutic stress in GBM. This CTRP8-RXFP1-STAT3 axis is a novel determinant of TMZ responsiveness/chemoresistance and an emerging new drug target for improved treatment of human GBM.
Abbreviations AP site, apurinic/apyrimidinic site; APE1, AP endonuclease 1; ARP, aldehyde reactive probe; ATM, ataxia telangiectasia mutated kinase; Bcl-2, B-cell lymphoma 2; Bcl-XL, B-cell lymphoma-extra large; BER, base excision repair; cAMP, cyclic 3 0 -5 0 adenosine monophosphate; CI, cell index; CTRP8, C1q/tumor necrosis factor-related peptide 8; dI, deoxyinosine; DNA pol b, DNA polymerase b; DNA, deoxyribonucleic acid; GBM, glioblastoma; KD, knockdown; MGMT, O 6 -methylguanine-DNA methyltransferase; MPG, N-methylpurine DNA glycosylase; MTIC, 3methyl-(triazen-1-yl) imidazole-4-carboxamide; OTM, olive tail moments; PCR, polymerase chain reaction; PI3K, phosphatidyl inositol 3 kinase; PKC, protein kinase C; RLN2, relaxin-2; RTCA, real-time cell analysis; RXFP1, relaxin family peptide receptor 1; siRNA, small interfering RNA; STAT3, signal transducer and activator of transcription 3; TMZ, temozolomide; XRCC1, X-ray repair cross-complementing group 1; cH2AX, phosphorylated histone 2A (Ser 139).
# Introduction
The relaxin family peptide receptor 1 (RXFP1) is a G protein-coupled receptor and relaxin-2 (RLN2) is a major ligand in human tissues of normal and neoplastic origin [bib_ref] International Union of Basic and Clinical Pharmacology. XCV. Recent advances in the..., Halls [/bib_ref] [bib_ref] Relaxin-like ligandreceptor systems are autocrine/paracrine effectors in tumor cells and modulate cancer..., Klonisch [/bib_ref]. Relaxin was shown to promote vasodilation, cardioprotection, antifibrotic wound healing, and angiogenesis [bib_ref] Relaxin inhibits early steps in vascular inflammation, Brecht [/bib_ref] [bib_ref] Effects of relaxin on arterial dilation, remodeling, and mechanical properties, Kp [/bib_ref] [bib_ref] Cardiovascular effects of relaxin: from basic science to clinical therapy, Du [/bib_ref]. In various tumors, the RLN2-RXFP1 has emerged as an important ligand-receptor system involved in controlling growth, migration/tissue invasion, angiogenesis, and metastasis [bib_ref] Relaxin-like ligandreceptor systems are autocrine/paracrine effectors in tumor cells and modulate cancer..., Klonisch [/bib_ref]. Contrary to other tumors, malignant brain tumors such as Grade III anaplastic astrocytoma and Grade IV glioblastoma (GBM) express RXFP1 but fail to express RLN2 [bib_ref] C1q-tumour necrosis factor-related protein 8 (CTRP8) is a novel interaction partner of..., Glogowska [/bib_ref]. Instead, we recently identified secreted adiponectin paralog C1q/tumor necrosis factor-related peptide 8 (CTRP8) as a novel RXFP1 agonist in human GBM [bib_ref] C1q-tumour necrosis factor-related protein 8 (CTRP8) is a novel interaction partner of..., Glogowska [/bib_ref] [bib_ref] CTRP8 and CTRP9B are novel proteins that hetero-oligomerize with C1q/TNF family members, Peterson [/bib_ref]. Of all 16 currently known CTRP members, CTRP8 is the least studied, in part, due to the fact that CTRP8 is a pseudogene in mice [bib_ref] CTRP8 and CTRP9B are novel proteins that hetero-oligomerize with C1q/TNF family members, Peterson [/bib_ref]. CTRPs are emerging as important regulators in metabolism, immune responses, and cancer [bib_ref] C1q and tumor necrosis factor superfamily: modularity and versatility, Kishore [/bib_ref] [bib_ref] CTRP family: linking immunity to metabolism, Schaffler [/bib_ref] [bib_ref] Metabolic function of the CTRP family of hormones, Seldin [/bib_ref] [bib_ref] RXFP1 is targeted by complement C1q tumor necrosis factor-related factor 8 in..., Thanasupawat [/bib_ref]. All CTRPs are composed of four distinct structural domains and can form homo-or heterotrimers and multimeric complexes. CTRP8 shares close phylogenetic and sequence conservation with CTRP1 and CTRP6, and their C-terminal globular domains share high conformational similarity with complement component C1q and tumor necrosis factor (TNF) [bib_ref] C1q and tumor necrosis factor superfamily: modularity and versatility, Kishore [/bib_ref] [bib_ref] The crystal structure of a complement-1q family protein suggests an evolutionary link..., Shapiro [/bib_ref]. Located at the N terminus of the C1q/TNF globular domain of CTRP8 is the putative RXFP1 interacting site 'AYAAFSV' [bib_ref] Discovery and validation of novel peptide agonists for G-protein-coupled receptors, Shemesh [/bib_ref]. In human GBM cells, the CTRP8-mediated autocrine/paracrine RXFP1 activation resulted in elevated intracellular cAMP levels, PI3 kinase pathway activation, and the phosphorylation of PKC isoforms [bib_ref] C1q-tumour necrosis factor-related protein 8 (CTRP8) is a novel interaction partner of..., Glogowska [/bib_ref]. Like RLN2 in other neoplastic models, the interaction of CTRP8 with RXFP1 promoted GBM matrix invasion and coincided with increased production and secretion of lysosomal protease cathepsin-B, a known prognostic marker of GBM [bib_ref] C1q-tumour necrosis factor-related protein 8 (CTRP8) is a novel interaction partner of..., Glogowska [/bib_ref].
Glioblastoma is the most frequent and most aggressive form of primary brain tumor of the astrocytic lineage with a patient survival time of only 15-17 months. Treatment consists of extensive surgical resection followed by radiation and chemotherapy [bib_ref] Long-term survival with glioblastoma multiforme, Krex [/bib_ref]. The drug of choice in the treatment of GBM is temozolomide (TMZ), an imidazole derivative and second-generation alkylating prodrug which undergoes spontaneous hydrolysis to the active metabolite 3-methyl-(triazen-1-yl) imidazole-4-carboxamide (MTIC). Treatment with TMZ results in DNA base methylation. The methylation at the N 7 position of guanine (N 7 -MeG; 80-85%) and the N 3 position of adenine (N 3 -MeA; 8-18%) constitute the majority of TMZ-induced DNA methylations repaired by the base excision repair (BER) pathway. BER is the predominant DNA repair system in mammalian cells and repairs small cytotoxic DNA base lesions resulting from oxidized, alkylated, or deaminated nucleotides [bib_ref] Overview of base excision repair biochemistry, Kim [/bib_ref] [bib_ref] Base excision repair, Krokan [/bib_ref]. The remaining 5-10% of TMZ-induced DNA-methylated lesions occur as O 6 -MeG which is the substrate for the enzyme O 6 -methylguanine-DNA methyltransferase (MGMT) [bib_ref] Mechanisms of chemoresistance to alkylating agents in malignant glioma, Sarkaria [/bib_ref]. The TMZinduced purine base alkylations N 3 -MeA and N 7 -MeG are the substrates for the monofunctional glycosylase N-methylpurine DNA glycosylase (MPG, also known as alkylpurine-DNA-N-glycosylase [APNG] or 3-alkyladenine DNA glycosylase . MPG initiates the first step of BER by removing the methylated base to generate an apurinic/pyrimidinic (AP) abasic site. The glycosylic backbone of the AP site is then cleaved by an AP lyase, like AP endonuclease 1 (APE1). This generates a cytotoxic 5 0 -deoxyribosyl phosphate (dRP) residue which is commonly removed by the dRP lyase activity of DNA polymerase b [bib_ref] The lyase activity of the DNA repair protein beta-polymerase protects from DNA-damage-induced..., Sobol [/bib_ref]. DNA polymerase b adds the complementary base and the X-ray repair cross-complementing group 1 (XRCC1)/DNA ligase III complex performs the phosphodiester bond formation to complete BER [bib_ref] Base excision repair, Krokan [/bib_ref]. Inhibition of BER in MPG overexpressing human glioma sensitizes these cells to TMZ in vitro and in vivo, but this cytotoxic effect is diminished at higher cellular levels of the rate-limiting BER enzyme DNA polymerase b [bib_ref] Overview of base excision repair biochemistry, Kim [/bib_ref] [bib_ref] N-methylpurine DNA glycosylase and DNA polymerase beta modulate BER inhibitor potentiation of..., Tang [/bib_ref].
In the present study, we have identified a novel role of the CTRP8-RXFP1 ligand-receptor system in promoting the repair of TMZ-induced alkylating DNA base damage in GBM. CTRP8 activated a newly discovered RXFP1-STAT3 signaling pathway which caused enhanced resistance to DNA alkylating stress and increased survival in GBM upon TMZ treatment. This CTRP8-RXFP1-STAT3 signaling cascade may serve as a new mediator of TMZ chemoresistance in human GBM.
# Materials and methods
## Isolation of patient gbm cells and cell culture
Human GBM tissues were obtained from GBM patients treated at the local Health Science Centre. The study was approved by the University and Pathology ethics boards (ethics approval # H2010:116). Human GBM cells isolated from two GBM patients (GBM-1/ 2) and the human U87MG glioblastoma cell line [bib_ref] Origin of the U87MG glioma cell line: good news and bad news, Allen [/bib_ref] were cultured in DME/F12 containing 10% FBS at 37°C in a humidified 5% CO 2 atmosphere. The medium was changed to DME/F12 with 1% FBS 24 h prior to the treatments. U87MG cells had been authenticated prior to this study.
## Chemicals and reagents
Temozolomide (TMZ) was purchased from Sigma (Oakville, ON, Canada) and used at 1.5 mM which caused cell damage 24 h of incubation, respectively. STAT3 inhibitor VI, S3I-201, was from EMD Millipore (Billerica, MA, USA). Cells were preincubated with inhibitors at 25 lM each for 60 min prior to additional treatment.
## Recombinant protein production
Recombinant human full-size C-terminally Flag-tagged CTRP8 in pET28a vector was produced in Escherichia coli. Recombinant CTRP8 was purified by His-Gravity kit (GE Healthcare, Mississauga, ON, Canada) according to manufacturer's protocol and dialyzed against Tris buffer (50 mM Tris/HCl, 150 mM NaCl, pH 7.4) before determining the concentration by NanoVue spectrophotometer (GE Healthcare). The purity of the recombinant CTRP8 was assessed with 15% SDS/PAGE following Coomassie staining and immunoblot for anti-Flag detection.
## Rna silencing and pcr
For knockdown (KD) of RXFP1 in patient GBM cells, 5 9 10 4 cells in six-well plates were transfected with two different RXFP1 siRNA at a concentration of 100 nM [RXFP1-1: (5 0 ?3 0 ) sense CCGUUUACCUGAUAAA CCUtt, antisence AGGUUUAUCAGGUAAACGGgt; siRXFP1-2: (5 0 ?3 0 ) sense GGAAGUAAUAAGAUU GAAAtt, antisense UUUCAAUCUUAUUACUU CCta (Ambion, Ottawa, ON, Canada) using siLentFect lipid reagent (Bio-Rad, Mississauga, ON, Canada)]. Total RNA was collected for the detection of RXFP1 expression levels using RT-PCR and quantitative realtime PCR (qPCR) with the following primers: RXFP1 forward AAAAGAGATGATCCTTGCCAAACG, reverse CCACCCAGATGAATGATGGAGC; MPG forward GGTCCTAGTCCGGGGACTTCC, reverse CTTGTCTGGGCAGGCCCTTGC; and GAPDH forward CATCACCATCTTCCAGGAGCG, reverse TGACCTTGCCCACAGCC TTG. The qPCR was performed with a QuantStudio Ò 3 system (Applied Biosystems, Ottawa, ON, Canada). The comparative C T (DDC T ) method was used for data analysis using QUANTSTUDIO Ò DESIGN & ANALYSIS software (Applied Biosystem, Ottawa, ON, Canada). Samples were normalized to the expression of GAPDH.
## Xcelligence â real-time cell analysis (rtca)
We performed xCELLigence real-time cell cytotoxicity assays (ACEA Biosciences, Inc., San Diego, CA, USA). Patient GBM cells and U87MG cells were cultured on E-plates and treated with CTRP8, TMZ, and siRXFP1 as indicated. Changes in cellular impedance are represented as cell index (CI) and were recorded every 15 min for 24 h upon treatment using RTCA software (ACEA Biosciences, Inc., San Diego, CA, USA).
## Caspase 3/7 activity assay
Caspase 3/7 activity assay was performed using Caspase-Glo 3/7 reagent (Promega, Madison, WI, USA) according to the manufacturer's instructions. Caspase-Glo 3/7 reagent was added to the samples in 96-well plates and incubated for 4 h at room temperature (RT). Plates were spun at 350 rpm prior to detecting luminescence signal with a luminometer (Wallac, Per-kinElmer, Boston, MA, USA).
2.7. Single-cell gel electrophoresis assay (alkaline comet assay) GBM cells (5 9 10 4 cells) plated in 6-well plates were treated, and DNA damage was assessed using a Comet assay kit (Trevigen, Gaithersburg, MD, USA). Cells were embedded in low-melting-point agarose on glass slides. Once the agarose was solidified, slides were maintained in prechilled lysis solution at 4°C for 45 min before being incubated in an alkaline solution for 20 min at RT followed by single-cell gel electrophoresis with fresh electrophoresis buffer for 15 min at 25 V 0.8 amps. Slides were dehydrated with 70% and 100% ethanol for 20 min and stained with SYBR green. Comet images were acquired using a Z2 microscope (Zeiss, Jena, Germany). Comet olive tail moments (OTM; product of the tail length and the fraction of total DNA in the tail), a measure of DNA damage, were quantified for 50 cells per treatment using the COMET ASSAY IV software (Perceptive, Bury St Edmunds, UK).
## Immunofluorescence
Immunofluorescence detection of cH2AX was described previously [bib_ref] High Mobility Group AT Hook 2 protein (HMGA2) mediates temozolomide resistance in..., Thanasupawat [/bib_ref]. Briefly, patient GBM cells on cover slips were treated with siRNA for RXFP1 silencing 24 h prior to treatment with CTRP8 and/or TMZ. Cells were fixed with 3.7% formaldehyde for 20 min at RT. Nonspecific antibody binding sites were blocked for 2 h at RT with 1% BSA in 0.01% Triton X-100 plus 5% rabbit normal serum (blocking buffer; Sigma). GBM cells were immunostained overnight with 1 : 5000 cH2AX (EMD; Millipore) in blocking buffer at 4°C prior to incubation for 1 h at RT with AlexaFlour-594-conjugated rabbit anti-mouse (Invitrogen, Thermo Scientific). For nuclear staining, cells were counterstained with 0.1 lgÁmL À1 DAPI and mounted with Fluoromount aqueous mounting medium (both Sigma). Cells were imaged with a Z2 microscope and ZEN IMAGING software (Zeiss). Intensity quantification of immunofluorescence signal for cH2AX foci was analyzed using IMAGE J software (National Institutes of Health, Bethesda, MD, USA). A total of 100 nuclei per each treatment for all cell lines were analyzed. The results are represented as a graph with relative fluorescence intensity.
# Western blot analysis
Proteins were separated on 10% and 12% SDS/PAGE gels and transferred to nitrocellulose membranes. For immunodetection, nonspecific protein binding sites were blocked by incubation with 5% nonfat milk in TBS/T for 1 h at RT. Primary antibodies [1 : 1000 of pSTAT3 Tyr705 , pSTAT3 Ser727 , total STAT3, cH2AX, XRCC1, MGMT, Bcl-2, Bcl-XL (all Cell Signaling Technologies, Boston, MA, USA), 1 : 2000 of APE1, 1 : 3000 of MPG, 1 : 500 of DNA Polb (all Abcam, Toronto, ON, Canada), and 1 : 10 000 for b-actin (Sigma)] were incubated at 4°C overnight. Membranes were washed 39 in TBS/T for 5 min each at RT before adding HRP-conjugated secondary antibodies for 1 h. Specific binding was visualized with ECL solution (Thermo Scientific). All western blots were performed using Bio-Rad Laboratories Inc (Bio-Rad, Mississauga, ON, Canada) system, including ChemiDoc MP gel documentation and IMAGE LAB software for quantitate analysis of proteins signals. Percentage of relative intensity was display as graphs representing three independent experiments for each of the cell line use in the study.
## N-methylpurine dna glycosylase (mpg) molecular beacon activity assay
The MPG activity assay was performed on U87MG cells as described previous [bib_ref] Quantitative, real-time analysis of base excision repair activity in cell lysates utilizing..., Svilar [/bib_ref]. Beacon oligodeoxyribonucleotides (MPG probe: 5 0 -6-FAM/ GCACT/X/TTGAATT GACACGCCATGTCGAT-CAATTCAATAGTGC/3Dab/-3 0 , control probe: 5 0 -6-FAM/GCACTATTGAATTGACACGCCATGTC-GATCAATTCAATAGTGC/3Dab/-3 0 ; X is deoxyinosine) were purchased from Integrated DNA Technologies (Coralville, IA, USA). Stem loop formation of the beacons was confirmed by heating the oligonucleotides to 95°C for 3 min followed by slow cooling overnight at RT. Upon hairpin loop formation, no fluorescence signal was emitted and the beacon remained stable at 37°C. When the hairpin loop beacon was reheated to 95°C, fluorophore and quencher separated as the oligonucleotides unfolded, resulting in maximum fluorescence signal. Nuclear protein lysates were extracted with NE-PER nuclear/cytoplasmic extraction reagents (Thermo Scientific). Ten micrograms of nuclear protein lysates was incubated with 40 nM beacon probe, and fluorescence was detected at 37°C every 20 s for 120 min using a QuantStudio Ò 3 system.
## Detection of ap sites in genomic dna
U87MG cells were pretreated with 100 ngÁmL À1 of CTRP8 in 1% FBS for 24 h prior to treatment with 1.5 mM TMZ for 15 min. Genomic DNA was extracted using Genomic DNA Mini Kit (Thermo Scientific) according to the manufacturer's instructions. The ARP labeling and quantification of AP sites were performed by AP sites assay kit (Dojindo Molecular Technologies, Burlington, ON, Canada). Ten microliters of genomic DNA (100 lgÁmL À1 ) in TE buffer was incubated with 10 lL of 5 mM ARP solution at 37°C for 1 h. ARPlabeled DNA in DNA binding solution was added to a 96-well plate and incubated at 37°C overnight in the dark. Wells were washed 5 times with washing buffer, HRP-Streptavidin solution was added and incubated at 37°C for 1 h. Wells were washed again 5 times and incubated with substrate solution at 37°C for 1 h prior to absorbance was measured at 630 nm with Synergy H1 microplate reader (BioTek, Winooski, VT, USA). Standard ARP DNA kit solutions determined 1-40 AP sites per 100 000 bp, and data are presented as number of AP sites per 100 000 nucleotides. All treatments were performed in triplicate.
# Statistical analysis
All experiments were carried out at least in triplicate. Results are showed as mean AE standard deviation (SD). Data were analyzed with GRAPHPAD PRISM 6 statistical software using two-way ANOVA with post hoc Tukey's HSD (honestly significant difference). P values less than 0.05 was considered significant. The level of significance was defined as *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.
# Results
## Ctrp8 activates a novel rxfp1-stat3 signaling pathway in gbm cells
Aberrant STAT3 signaling is a hallmark of gliomagenesis and has important therapeutic implications in GBM [bib_ref] STAT3 tyrosine phosphorylation influences survival in glioblastoma, Birner [/bib_ref]. Treatment of patient GBM-1/2 [fig_ref] Figure 1: CTRP8 promotes STAT3 signaling in GBM [/fig_ref] , C, E; [fig_ref] Figure 1: CTRP8 promotes STAT3 signaling in GBM [/fig_ref] and U87MG [fig_ref] Figure 1: CTRP8 promotes STAT3 signaling in GBM [/fig_ref] , D, F) with CTRP8 resulted in STAT3 activation with pSTAT3 Y705 phosphorylation as early as 5 min after treatment. A subtle phosphorylation was observed for STAT3 S727 upon stimulation with CTRP8 in patient GBM-1 [fig_ref] Figure 1: CTRP8 promotes STAT3 signaling in GBM [/fig_ref] , B) and U87MG [fig_ref] Figure 1: CTRP8 promotes STAT3 signaling in GBM [/fig_ref]. STAT3 inhibitor S3I-201 effectively blocked STAT3 phosphorylation in CTRP8-treated patient GBM-1 [fig_ref] Figure 1: CTRP8 promotes STAT3 signaling in GBM [/fig_ref] and U87MG [fig_ref] Figure 1: CTRP8 promotes STAT3 signaling in GBM [/fig_ref] cells but had no effect on total STAT3 levels. CTRP8-mediated STAT3 activation was critically dependent on the presence of RXFP1 in human GBM cells. Specific siRNA-mediated RXFP1 KD in patient GBM-1 and U87MG with two different siRXFP1-1/2 constructs abolished the ability of CTRP8 to cause STAT3 phosphorylation in patient GBM ( . This CTRP8 effect was more pronounced for the pSTAT3 Y705 than pSTAT3 S727 residue. Similarly, siRXFP1 knockdown (KD; siRXFP1-1) diminished phosphorylation of both pSTAT3 Y705/S727 residues and abolished the ability of CTRP8 to induce STAT3 phosphorylation in patient GBM-1 (C) and U87MG cells (D). b-Actin served as loading control in all blots. Representative examples of qPCR results demonstrate the significant downregulation of RXFP1 transcripts upon siRXFP1-1 treatment in patient GBM-1 (E) and U87MG (F) cells. Quantitative analysis from three independent experiments (two-way ANOVA; data are shown as mean AE SD; ****P < 0.0001) are shown.
the successful siRXFP1 KD with siRXFP1/2 in patient GBM-1 [fig_ref] Figure 1: CTRP8 promotes STAT3 signaling in GBM [/fig_ref] and U87MG cells [fig_ref] Figure 1: CTRP8 promotes STAT3 signaling in GBM [/fig_ref] and demonstrated that CTRP8 did not alter endogenous RXFP1 mRNA levels [fig_ref] Figure 1: CTRP8 promotes STAT3 signaling in GBM [/fig_ref]. Similar results were obtained in patient GBM-2 cells treated with siRXFP1-2, indicating that the effects detected with siRXFP1 treatment were likely not the result of siRNA-mediated off-target effects . Collectively, these results identified CTRP8 as a novel inducer of an RXFP1-STAT3 signaling cascade in human GBM.
## Ctrp8 protects gbm cells against dna damage by the alkylating drug temozolomide
The STAT3 signaling pathway is associated with TMZ chemoresistance in GBM, but the underlying mechanisms are unclear [bib_ref] STAT3 is essential for the maintenance of neurosphere-initiating tumor cells in patients..., Villalva [/bib_ref]. Here, we show that RXFP1 agonist CTRP8 [bib_ref] C1q-tumour necrosis factor-related protein 8 (CTRP8) is a novel interaction partner of..., Glogowska [/bib_ref] mitigated the ability of first-line GBM drug TMZ to induce DNA damage. Patient GBM-1/2 cells , B;, B) and U87MG, E) exposed to TMZ demonstrated strong immunofluorescence for nuclear cH2AX, an established marker for doublestrand (ds) DNA breaks. However, GBM-1/2 cells cotreated with TMZ and CTRP8 showed markedly reduced nuclear cH2AX fluorescence, while CTRP8 alone did not elicit dsDNA breaks in patient GBM, B) or U87MG. The CTRP8 protective effect against dsDNA damage resulting from unrepaired TMZ-induced DNA lesions was RXFP1-dependent and abolished by siRXFP1 KD in patient GBM-1/2, B) and U87MG. Corresponding IgG control experiments failed to show specific immunofluorescence as shown for patient GBM-2, F) and U87MG. Quantitative western blot analysis revealed that treatment with CTRP8 of patient GBM-1/2 , E, G; [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] and U87MG resulted in a marked reduction in phosphorylated cH2AX protein upon TMZ treatment as compared to TMZ treatment alone. The presence of RXFP1 was critical for CTRP8 to elicit its DNA protective effect in the presence of TMZ and was lost upon siRXFP1-1 KD in GBM-1/2 [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] , C) and U87MG . Similar results were obtained in patient GBM-2 cells upon treatment with a siRXFP1-2 . In addition, STAT3 inhibitor S3I-201 blocked the ability of CTRP8 to attenuate cH2AX protein levels upon TMZ treatment in patient GBM-1/2 [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] and U87MG .
We assessed the extent of CTRP8-mediated protection against DNA damage induced by TMZ at the single cell level [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] , E). Comet assay permits the quantification of dsDNA fragmentation in the nucleus at the level of a single GBM cell [bib_ref] Analysis by alkaline comet assay of cancer patients with severe reactions to..., Alapetite [/bib_ref]. Exposure to TMZ increased the olive tail moment (OTM) in patient GBM-1/2 [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] , E) and U87MG . Quantitative analysis of the comets showed that the TMZinduced OTM was markedly reduced in the presence of CTRP8 in patient GBM-1/2 [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] and U87MG . This DNA protective function of CTRP8 in the presence of TMZ was abrogated by the specific STAT3 inhibitor S3I-201 in patient GBM-1/2 [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] and U87MG , while CTRP8 or S3I-201 alone had no effect [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref]. These data demonstrated a novel protective role of the CTRP8-RXFP1-STAT3 signaling pathway against TMZ chemotherapeutic stress in human GBM.
## Tmz resistance induced by ctrp8 involves increased mpg and ber activity
The removal of TMZ-induced methylated DNA bases by the base excision repair (BER) pathway generates apurinic/apyrimidinic (AP) single-stranded (ss) DNA sites which are fragile and can progress into dsDNA breaks [bib_ref] Base excision repair, Krokan [/bib_ref]. We reasoned that a DNA protective role of CTRP8 may involve a reduction in the number of detectable AP sites in genomic DNA. Here, we show that TMZ significantly increased the number of AP sites per 10 5 nucleotides from an average of 15 (control) to 24 AP sites. CTRP8 markedly reduced the number of AP sites in TMZ-treated patient GBM-1 cells by >40% to levels of untreated GBM cells. To determine a molecular mechanism that can account for this remarkable DNA protective effect of CTRP8, we assessed the amounts of key cellular BER proteins in the patient GBM-1/2 modelsand U87MG. Exposure to CTRP8 caused an exclusive increase in protein production of the monospecific DNA glycosylase N-methylpurine DNA glycosylase (MPG), a key BER-initiating enzyme which removes altered DNA bases and generates AP sites [bib_ref] Overview of base excision repair biochemistry, Kim [/bib_ref] [bib_ref] Base excision repair, Krokan [/bib_ref] , whereas other BER members, including APE1, XRCC1, or DNA polymerase B (DNA pol b), remained unchanged as determined by quantitative western blot analysis in patient GBM-1/2, D, F; [fig_ref] Figure 5: Schematic model of the CTRP8-RXFP1-STAT3 signaling axis in human GBM [/fig_ref] and U87MG. This increase in MPG protein coincided with the ability of CTRP8 to enhance MPG gene expression in patient GBM-1/2and U87MG. The presence of a functional RXFP1 in GBM was required for increased MPG protein production as determined by siRXFP1 treatment in patient GBM-1/2and U87MG. Similar results were obtained with a different siRNA (siRXFP1-2) to suppress RXFP1 expression in patient GBM-2 . Quantitative western blot data showed that the increase in MPG protein was also sensitive to STAT3 activation. STAT3 inhibitor S3I-201 blocked this increase as shown in patient GBM-1and U87MG. Next, we employed a real-time molecular beacon assay to specifically measure MPG activity and show that the observed increase in MPG protein content observed upon CTRP8 treatment translated into increased MPG enzymatic activity. The MPG molecular beacon assay uses a specific MPG deoxyribonucleotide probe with a fluorophore (6-FAM) attached at the 5 0 -end and a quencher (3Dab) at the 3 0 -end [bib_ref] Quantitative, real-time analysis of base excision repair activity in cell lysates utilizing..., Svilar [/bib_ref]. The MPG molecular probe contains a deoxyinosine (dI) base that is specifically recognized and exclusively cleaved by the MPG enzyme to release a short DNA fragment and the attached 5 0 -fluorophore from its quencher to generate a fluorescence signal which can then be quantified by qPCR. An identical molecular beacon lacking the MPG recognition site was used as a control. When nuclear lysates of U87MG were incubated with the MPG beacon, we observed a steady increase in fluorescence signal over time with the MPG probe in cell lysates of untreated U87MG, reflecting endogenous MPG activity. A further significant increase in MPG activity was observed upon exposure to CTRP8 which was reduced to endogenous MPG activity levels by STAT3 inhibitor SI3-201. The absence of fluorescence signals in cell lysates incubated with the control probe lacking the MPG cleavage site showed that the MPG beacon assay specifically detected MPG activity. These results demonstrated that CTRP8 can increase both MPG protein content and MPG activity in human GBM cells. The DNA protective role of CTRP8 was not attributable to changes in MGMT protein. Both MGMT and STAT3 have been shown to mediate TMZ resistance in glioblastoma [bib_ref] MGMT gene silencing and benefit from temozolomide in glioblastoma, Hegi [/bib_ref] [bib_ref] STAT3 inhibition overcomes temozolomide resistance in glioblastoma by downregulating MGMT expression, Kohsaka [/bib_ref]. In the patient GBM cells studied, CTRP8 failed to alter MGMT protein levels and U87MG is devoid of MGMT [bib_ref] Dovitinib enhances temozolomide efficacy in glioblastoma cells, Thanasupawat [/bib_ref]. Collectively, we identified the key BER monofunctional glycosylase MPG as a novel target and mediator of DNA protection of the CTRP8-RXFP1-STAT3 signaling cascade in human GBM cells.
## Ctrp8 promotes glioblastoma survival
We reasoned that the DNA protective function of the CTRP8-RXFP1-STAT3 signaling cascade would be most effective if CTRP8 also promoted cell survival mechanisms in glioblastoma. Real-time cell analysis (RTCA) cytotoxicity assays revealed that TMZ treatment caused significant cell death, indicated by a marked decrease in cell index (CI) compared to untreated controls in patient GBM-1/2 [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] ; Figs S2E and S5E) and U87MG cells [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref]. TMZinduced cell death was abrogated by CTRP8 in GBM cells, and this cytoprotective role of CTRP8 was RXFP1 dependent and blocked by RXFP1 KD with siRXFP1-1 or siRXFP1-2 in patient GBM-1/2 . CTRP8 attenuates TMZ-induced DNA damage. cH2AX, a marker of double-strand (ds) DNA breaks, was detected by immunofluorescence in patient GBM (A). Treatment with TMZ (1.5 mM) resulted in a significant increase in cH2AX foci (red) in DAPI-stained nuclei (blue) compared to medium controls (A). Pretreatment for 24 h with CTRP8 (100 ngÁmL À1 ) caused a marked reduction in cH2AX foci upon exposure to TMZ compared to patient GBM-1 cells treated with TMZ alone (A). This CTRP8-mediated DNA protective effect was abolished upon siRXFP1-1 KD (A). The results of the quantification of fluorescence intensity of cH2AX foci for 100 nuclei per treatment group are shown (B). Upon TMZ treatment, western blot analysis (C-F) revealed a marked reduction in cH2AX protein in the presence of CTRP8 in patient GBM-1 (C, E) and U87MG (D, F). This DNA protective effect of CTRP8 was blocked upon siRXFP1-1 treatment in patient GBM-1 (C) and U87MG (D). Similar results were obtained with STAT3 inhibitor S3I-201 in patient GBM-1 (E) and U87MG (F). b-Actin served as loading control. Representative western blot images are shown. Densitometry of western blot signal for cH2AX in patient GBM-1 (G) and U87MG (H) was normalized to b-actin. We employed Comet assay to quantify DNA fragmentation as determined by olive tail moment (OTM) in individual human GBM-1 cells (I, J). Pretreatment for 24 h with CTRP8 (100 ngÁmL À1 ) prior to TMZ exposure (1.5 mM) resulted in decreased OTM, as shown in representative agarose gel images of comets from patient GBM-1 cells (I, J). The OTM was determined as an index of both comet tail length and the amount of DNA in tail as quantified by SYBR green fluorescence dye (L). Quantitative analysis of OTM from 50 cells for each treatment group revealed that CTRP8 caused a marked reduction in dsDNA breaks. This protective CTRP8 function was lost upon treatment with S3I-201 in patient GBM-1 (K) and U87MG (M). CTRP8 and S3I-201 alone failed to cause dsDNA breaks and resulted in negligible OTM (K, M). Quantitative analysis from three independent experiments (two-way ANOVA; data are shown as mean AE SD; ****P < 0.0001) are shown. . CTRP8 reduces the number of AP sites and enhances BER. We quantified AP sites in genomic DNA to determine the cause of TMZ-induced DNA damage. The endogenous level of AP sites was about 12-15 sites/10 5 bp in our patient GBM-1 model (A). TMZ alone triggered a marked upregulation of AP sites in patient GBM-1 cells (A). A significant reduction in the number of AP sites comparable to levels detected in untreated control cells was observed upon pretreatment of patient GBM-1 cells with CTRP8 followed by TMZ exposure (A). CTRP8 treatment alone did not alter the number of AP sites (A). Western blot analysis of important BER proteins showed that CTRP8 (100 ngÁmL À1 ) induced the specific and exclusive upregulation of N-methylpurine DNA glycosylase (MPG) protein, a key factor in initiating BER, in patient GBM-1 (B) and U87MG (C). Importantly, siRXFP1 KD and STAT3 inhibition abolished this CTRP8 mediated increase in MPG protein (B-G) as shown for patient GBM-1 (B, D, F) and U87MG (C, E, G). The increase in MPG protein coincided with a significant upregulation of MPG transcripts upon CTRP8 treatment as revealed by QPCR analysis in patient GBM-1 (H) and U87MG (I). We used an MPG-specific molecular beacon activity assay on U87MG nuclear lysates which utilizes a specific MPG deoxyribonucleotide probe with a fluorophore (6-FAM) attached at the 5 0 -end and a quencher (3Dab) at the 3 0 -end . This MPG oligo probe contains a deoxyinosine (dI) base as MPG recognition site. The exclusive MPG cleavage at this site releases a 6-bp DNA fragment with attached 5 0fluorophore which dissociates from its quencher to generate a fluorescence signal which is proportional to MPG activity and can be quantified by qPCR (J). An identical control oligo lacking the dI base and, thus, is not cleaved by MPG was used as control. Quantification of fluorescence intensities reflecting MPG activities measured at 60, 120, and 180 min is shown (K). Endogenous MPG exclusively cleaved the MPG probe, and CTRP8 treatment caused a further significant increase in MPG activity which was absent in the presence of S3I-201 (K). Quantitative analysis from three independent experiments (two-way ANOVA; data are shown as mean AE SD; *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001) are shown. [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] ; Figs S2E and S5E) and U87MG [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref]. Treatment with CTRP8 or siRXFP1 alone did not show cytotoxicity in patient GBM-1/2 [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] ; Figs S2E and S5E) and U87MG [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref]. Coinciding with the cytotoxicity data, TMZ induced a significant increase in caspase 3/7 activity in patient GBM-1/2 [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] and S5F) and U87MG [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref]. CTRP8 significantly curtailed the TMZ-induced caspase 3/7 activation in patient GBM-1/2 [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] ; Figs S2F and S5F) and U87MG [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref]. This anti-apoptotic function of CTRP8 was critically dependent on the presence of a functional RXFP1-STAT3 signaling cascade. Specific KD of RXFP1 using two different specific siRNA or treatment with STAT3 inhibitor S3I-201 abolished this CTRP8 protective effect [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref]. Quantitative western blot analysis demonstrated an upregulation of the anti-apoptotic STAT3 targets Bcl-XL and Bcl-2 proteins in patient GBM-1/2 [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] , G-J; [fig_ref] Figure 5: Schematic model of the CTRP8-RXFP1-STAT3 signaling axis in human GBM [/fig_ref] and U87MG [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref]. The upregulation of these Bcl members was abolished by RXFP1 KD [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] , F, G, I, K, M; [fig_ref] Figure 5: Schematic model of the CTRP8-RXFP1-STAT3 signaling axis in human GBM [/fig_ref] , H, J) and STAT3 inhibition [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref] , F, H, J, L, N; [fig_ref] Figure 5: Schematic model of the CTRP8-RXFP1-STAT3 signaling axis in human GBM [/fig_ref] , I, K). In summary, the ability of the novel CTRP8-RXFP1-STAT3 signaling axis to guard against TMZ-induced DNA damage and promote survival pathways provide evidence for a novel role of the CTRP8-RXFP1 system in TMZ chemoresistance in human GBM [fig_ref] Figure 5: Schematic model of the CTRP8-RXFP1-STAT3 signaling axis in human GBM [/fig_ref].
# Discussion
The current gap in our understanding of the cellular mechanisms employed by human GBM to effectively thwart drug-mediated DNA alkylating damage contributes to a lack in therapeutic improvement and the dismal prognosis of GBM patients. Here, we demonstrate a novel mechanism which links our recently discovered autocrine/paracrine CTRP8 activation of the G protein-coupled receptor RXFP1 with the oncogenic STAT3 signaling pathway predictive of poor clinical outcome in human GBM patients [bib_ref] STAT3 tyrosine phosphorylation influences survival in glioblastoma, Birner [/bib_ref] [bib_ref] C1q-tumour necrosis factor-related protein 8 (CTRP8) is a novel interaction partner of..., Glogowska [/bib_ref]. Our data provide first evidence that an activated CTRP8-RXFP1-STAT3 axis promotes BER and increases resistance to the first-line chemotherapeutic drug TMZ in human GBM cells.
Key clinical features of GBM pathology include extensive cerebral dissemination and resistance to treatment with chemotherapeutic drugs such as TMZ. CTRP8-RXFP1 is emerging as a new ligand-receptor system which promotes GBM migration [bib_ref] C1q-tumour necrosis factor-related protein 8 (CTRP8) is a novel interaction partner of..., Glogowska [/bib_ref] and, as shown here, protects against the cytotoxic effects of the DNA alkylating drug TMZ. Likely initiated by an interaction of RXFP1 with the small G protein Ga i3 to activate the Ga i3 -Gbc-PI3K signaling pathway [bib_ref] Relaxin stimulates protein kinase C zeta translocation: requirement for cyclic adenosine 3..., Nguyen [/bib_ref] , our discovery of a novel CTRP8-RXFP1-STAT3 signaling cascade in human GBM links this CTRP8-RXFP1 system to oncogenic STAT3 functional outcomes, including GBM cell survival, angiogenesis, and cell migration/invasion [bib_ref] Protein kinase Cvarepsilon mediates Stat3Ser727 phosphorylation, Stat3-regulated gene expression, and cell invasion..., Aziz [/bib_ref] [bib_ref] Protein kinase C zeta regulates human pancreatic cancer cell transformed growth and..., Butler [/bib_ref] [bib_ref] Role of STAT3 in genesis and progression of human malignant gliomas, Ouedraogo [/bib_ref]. CTRP8-activated RXFP1 may utilize PI3K to mediate STAT3 activation as PI3K and its target BMX TEC kinase were recently shown to mediate the phosphorylation of STAT3 [bib_ref] C1q-tumour necrosis factor-related protein 8 (CTRP8) is a novel interaction partner of..., Glogowska [/bib_ref] [bib_ref] Essential role of Stat3 in PI3K-induced oncogenic transformation, Hart [/bib_ref]. We also previously identified lysosomal cathepsins as targets of H2 relaxin, the cognate ligand of RXFP1, and CTRP8 in human thyroid cancer (cathepsin-D and cathepsin-L) and GBM (cathepsin-B), respectively [bib_ref] C1q-tumour necrosis factor-related protein 8 (CTRP8) is a novel interaction partner of..., Glogowska [/bib_ref] [bib_ref] Relaxin enhances the oncogenic potential of human thyroid carcinoma cells, Hombach-Klonisch [/bib_ref]. High cathepsin-B serum levels are associated with poor prognosis in GBM patients [bib_ref] Cathepsin L in glioma progression: comparison with cathepsin B, Strojnik [/bib_ref]. . This CTRP8 protective function was dependent on the presence of functional RXFP1, as siRXFP1 KD abolished this effect, and treatment with CTRP8 and siRXFP1 alone had no effect in GBM-1 (A, B). Caspase 3/7 activity assays (C, D) confirmed that CTRP8 protected patient GBM-1 (C) and U87MG (D) against the cytotoxic effects of TMZ. This protective function CTRP8 was lost upon siRXFP1 KD (C) and STAT3 inhibition (D) as demonstrated in representative results for siRXFP1-1 KD in patient GBM-1 (C) and S3I-201 treatment in U87MG (D). Western blot analysis revealed that CTRP8 treatment caused the RXFP1-and STAT3-dependent upregulation of anti-apoptotic Bcl members and STAT3 targets Bcl-XL and Bcl-2 (E, F). Densitometry results of western blots for Bcl-XL are shown for patient GBM-1 (G, H) and U87MG (K, L) and for Bcl-2 with patient GBM-1 (I, J) and U87MG (M, N). Quantitative analysis from three independent experiments (two-way ANOVA; data are shown as mean AE SD; *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001) are shown.
Stat3 upregulates the expression of lysosomal proteases cathepsin-B and cathepsin-L under physiological conditions [bib_ref] Stat3 controls lysosomalmediated cell death in vivo, Kreuzaler [/bib_ref] and, thus, may facilitate cathepsin-B enhanced tissue invasion and lysosomal-mediated cell death regulation in brain tumors [bib_ref] Comparison of potential biological markers cathepsin B, cathepsin L, stefin A and..., Levicar [/bib_ref]. Although TMZ is the drug of choice in the treatment of GBM patients, frequent treatment failures result in resistance to this drug and fatal GBM recurrences [bib_ref] Malignant astrocytic glioma: genetics, biology, and paths to treatment, Furnari [/bib_ref] [bib_ref] Mechanisms of chemoresistance to alkylating agents in malignant glioma, Sarkaria [/bib_ref]. Major DNA adducts generated by TMZ are N 7methylguanine (N 7 -MeG; 60-80%), N 3 -methyladenine (N 3 -MeA; 10-20%), and O 6 -methylguanine (O 6 -MeG; 5-10%) [bib_ref] Repair of 3-methyladenine and abasic sites by base excision repair mediates glioblastoma..., Bobola [/bib_ref]. Excision of a modified base generates an apurinic/apyrimidinic (AP) DNA site, and consecutive AP endonuclease 1 (APE1) activity creates a single-stranded DNA site which has the propensity to develop into a double-strand break if not processed expediently by BER [bib_ref] Deoxyribonucleic acid damage and repair: capitalizing on our understanding of the mechanisms..., Helena [/bib_ref]. BER is the predominant DNA repair pathway for the repair of single cytotoxic DNA base lesions which includes oxidized, deaminated, and N 7 -MeG/ N 3 -MeA alkylated nucleotides [bib_ref] Overview of base excision repair biochemistry, Kim [/bib_ref] [bib_ref] Base excision repair, Krokan [/bib_ref] , whereas TMZ-induced cytotoxic, radio-sensitizing, and base-mispairing O 6 -MeG sites are removed by the O 6 -MeG DNA methyltransferase (MGMT). MGMT is the sole enzyme dedicated to the demethylation of O 6 -meG to guanine by transferring this methyl group to an internal cysteine residue which inactivates the MGMT enzyme [bib_ref] O 6 -Methylguanine DNA methyltransferase protein expression in tumor cells predicts outcome..., Spiegl-Kreinecker [/bib_ref]. Although MGMT promoter hypermethylation in GBM tumors is clinically associated with a better initial TMZ treatment response [bib_ref] MGMT gene silencing and benefit from temozolomide in glioblastoma, Hegi [/bib_ref] , even MGMT-negative GBM cells do not sufficiently respond to TMZ. This demonstrates the need to identify additional molecular mechanisms contributing to TMZ resistance. Our data showed that the DNA protective role of CTRP8 was as effective in MGMT-negative U87MG and MGMT-positive patient GBM cells [bib_ref] High Mobility Group AT Hook 2 protein (HMGA2) mediates temozolomide resistance in..., Thanasupawat [/bib_ref] , excluding altered cellular MGMT enzyme levels as a target of activated RXFP1 and cause for the CTRP8mediated protection against TMZ-induced DNA damage in our GBM models. We therefore focused on N 7 -MeG and N 3 -MeA adducts which constitute over 90% of TMZ-induced base alterations. These are recognized by the BER glycosylase MPG which performs the initial cleavage of the glycosylic bond between the damaged base and deoxyribose to generate an AP site [bib_ref] Repair of 3-methyladenine and abasic sites by base excision repair mediates glioblastoma..., Bobola [/bib_ref] [bib_ref] Overview of base excision repair biochemistry, Kim [/bib_ref]. Treatment of human GBM cells with RXFP1 agonist CTRP8 increased MPG protein levels and MPG activity as determined by MPG molecular beacon assay, while the cellular levels of other BER proteins, including APE1, DNA polymerase b (polb), and XRCC1, remained unaffected by CTRP8-mediated RXFP1 activation. The increased MPG activity resulted in enhanced BER capacity, reflected by the reduced number of AP sites and DNA double-strand breaks with resulting decrease in apoptosis. This indicated sufficient activity of BER factors downstream of MPG to ensure enhanced BER capacity in GBM with activated RXFP1 [bib_ref] Human methyl purine DNA glycosylase and DNA polymerase beta expression collectively predict..., Trivedi [/bib_ref]. TMZ resistance in GBM is associated with DNA damage-induced activation of the serine/threonine ataxia telangiectasia mutated kinase (ATM). Phosphorylation of MPG by ATM coincides with increased MPG activity and has been linked to alkylating drug resistance in pediatric GBM [bib_ref] ATM regulates 3-methylpurine-DNA glycosylase and promotes therapeutic resistance to alkylating agents, Agnihotri [/bib_ref]. However, it should be noted that this kinase-mediated mechanism of MPG activation utilizes preexisting cellular MPG as phosphorylation substrate, likely as a fast response to DNA damage. By contrast, the activation of the CTRP8-RXFP1-STAT3 signaling cascade initiated MPG gene activation and increased MPG protein production, consequently resulting in higher We propose a model in which the interaction of CTRP8 with membrane-anchored RXFP1 triggers a pSTAT3 signaling cascade in human GBM. STAT3 activation enhances MPG-dependent BER, thereby reducing DNA damage and promoting GBM survival. The latter includes the upregulation of anti-apoptotic STAT3 targets Bcl-2 and Bcl-XL. Collectively, this establishes the CTRP8-RXFP1-STAT3 cascade as a novel oncogenic signaling pathway that promotes TMZ chemoresistance in human high-grade glioma.
and possibly more sustained MPG glycosylase activity in stressed GBM. MPG has been described as an unfavorable independent prognostic factor for glioma patients and MPG gene and protein expression increase from low-to high-grade gliomas [bib_ref] Aberrant expression of Nmethylpurine-DNA glycosylase influences patient survival in malignant gliomas, Liu [/bib_ref]. Glioma patients undergoing TMZ treatment with low MPG levels, possibly due to MPG promoter methylation, have a better outcome compared to those with high MPG expression [bib_ref] Alkylpurine-DNA-Nglycosylase confers resistance to temozolomide in xenograft models of glioblastoma multiforme and..., Agnihotri [/bib_ref]. This points to the importance of a proper balance of BER factors and links increased MPG activity and AP site formation to enhanced TMZ resistance in GBM [bib_ref] N-methylpurine DNA glycosylase and DNA polymerase beta modulate BER inhibitor potentiation of..., Tang [/bib_ref]. Recently, the drug salinomycin was shown to downregulate the expression of DNA repair factors MPG, MGMT, and Rad51 recombinase and induce endoplasmic reticulum (ER) stress. Combined salinomycin/TMZ treatment of GBM cells resulted in enhanced TMZ sensitivity, DNA damage, apoptosis, and increased survival of mice with orthotopic GBM xenografts [bib_ref] Endoplasmic reticulum stress-inducing drugs sensitize glioma cells to temozolomide through downregulation of..., Xipell [/bib_ref]. The ability to enhance TMZ resistance and mount an anti-apoptotic Bcl2-like response identifies CTRP8-RXFP1 as a new and powerful defense against TMZ stress in glioma. All the results presented here for CTRP8 were replicated with H2 relaxin in our human GBM models and yielded similar results (T. . While this confirmed a key role for GBM-expressed CTRP8 as oncogenic driver in human glioblastoma, it is conceivable that H2 relaxin has a similar effect on therapeutic resistance in RXFP1+ tumors outside of the brain. The Rembrandt database identifies RXFP1 as constitutively expressed gene in all human GBM, suggesting subtype independent roles of RXFP1 in GBM.
# Conclusions
Ligand binding to the G protein-coupled receptor RXFP1 and downstream STAT3 pathway activation protects GBM against the DNA alkylating drug temozolomide. This TMZ resistance is mediated by STAT3 upregulation of MPG glycosylase with enhanced BER and increases Bcl-2/Bcl-XL anti-apoptotic pathway activation in human GBM cells. Thus, the RXFP1 ligand-receptor system should be considered an attractive new drug target to curb TMZ resistance in GBM.
## Supporting information
Additional Supporting Information may be found online in the supporting information section at the end of the article. [fig_ref] Figure 1: CTRP8 promotes STAT3 signaling in GBM [/fig_ref]. CTRP8 promotes STAT3 signaling in another patient GBM-2 cell model. . Different siRNA confirms essential role of RXFP1 in CTRP8 function.. CTRP8 attenuates TMZ induced DNA damage in different human GBM models. [fig_ref] Figure 4: CTRP8 promotes GBM cell survival [/fig_ref]. CTRP8 diminishes dsDNA breaks in the patient GBM-2 model. [fig_ref] Figure 5: Schematic model of the CTRP8-RXFP1-STAT3 signaling axis in human GBM [/fig_ref]. CTRP8 enhances MPG and promotes survival in the second patient GBM model. . O 6 -methylguanine DNA methyltransferase (MGMT) is not a target of CTRP8.
[fig] Figure 1: CTRP8 promotes STAT3 signaling in GBM. Exposure of human GBM-1 with human recombinant CTRP8 (100 ngÁmL À1 ) resulted in the phosphorylation of STAT3 at Tyr705 and Ser727 in patient GBM cells (A, C) and U87MG (B, D), whereas total STAT3 protein levels remained unchanged (A-D). Pretreatment with the specific STAT3 inhibitor S3I-201 abolished the ability of CTRP8 to cause STAT3 phosphorylation in patient GBM-1 cells (A) and U87MG (B) [/fig]
[fig] Figure 3: CTRP8 reduces the number of AP sites and enhances BER. We quantified AP sites in genomic DNA to determine the cause of TMZ-induced DNA damage. The endogenous level of AP sites was about 12-15 sites/10 5 bp in our patient GBM-1 model (A). TMZ alone triggered a marked upregulation of AP sites in patient GBM-1 cells (A). A significant reduction in the number of AP sites comparable to levels detected in untreated control cells was observed upon pretreatment of patient GBM-1 cells with CTRP8 followed by TMZ exposure (A). CTRP8 treatment alone did not alter the number of AP sites (A). Western blot analysis of important BER proteins showed that CTRP8 (100 ngÁmL À1 ) induced the specific and exclusive upregulation of N-methylpurine DNA glycosylase (MPG) protein, a key factor in initiating BER, in patient GBM-1 (B) and U87MG (C). Importantly, siRXFP1 KD and STAT3 inhibition abolished this CTRP8 mediated increase in MPG protein (B-G) as shown for patient GBM-1 (B, D, F) and U87MG (C, E, G). The increase in MPG protein coincided with a significant upregulation of MPG transcripts upon CTRP8 treatment as revealed by QPCR analysis in patient GBM-1 (H) and U87MG (I). We used an MPG-specific molecular beacon activity assay on U87MG nuclear lysates which utilizes a specific MPG deoxyribonucleotide probe with a fluorophore (6-FAM) attached at the 5 0 -end and a quencher (3Dab) at the 3 0 -end (Svilar et al., 2012) (J). This MPG oligo probe contains a deoxyinosine (dI) base as MPG recognition site. The exclusive MPG cleavage at this site releases a 6-bp DNA fragment with attached 5 0fluorophore which dissociates from its quencher to generate a fluorescence signal which is proportional to MPG activity and can be quantified by qPCR (J). An identical control oligo lacking the dI base and, thus, is not cleaved by MPG was used as control. Quantification of fluorescence intensities reflecting MPG activities measured at 60, 120, and 180 min is shown (K). Endogenous MPG exclusively cleaved the MPG probe, and CTRP8 treatment caused a further significant increase in MPG activity which was absent in the presence of S3I-201 (K). Quantitative analysis from three independent experiments (two-way ANOVA; data are shown as mean AE SD; *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001) are shown. [/fig]
[fig] Figure 4: CTRP8 promotes GBM cell survival. Real-time xCELLigence assays were used to quantify the effect of CTRP8 on TMZ-induced cytotoxicity. Measurements were taken every 15 min, and cell indices collected at 24 h of incubation are shown (A, B). We observed a strong TMZ-mediated cytotoxicity in patient GBM-1 (A) and U87MG (B). Treatment with CTRP8 largely abolished the cytotoxic effect of TMZ in patient GBM-1 (A) and U87MG (B) [/fig]
[fig] Figure 5: Schematic model of the CTRP8-RXFP1-STAT3 signaling axis in human GBM. [/fig]
|
Psoriasis Carries an Increased Risk of Venous Thromboembolism: A Danish Nationwide Cohort Study
Background: Psoriasis is an immunoinflammatory disease associated with cardiovascular risk factors, atherothrombotic events, and hypercoagulability. Venous thromboembolism (VTE) is potentially lethal and shares risk factors with psoriasis, but the risk of VTE associated with psoriasis is unknown. The present study investigated the potential association between psoriasis and VTE.Methods and Findings: Information from nationwide prospectively recorded registers of hospitalization, drug dispensing from pharmacies, socio-economic data, and causes of death was linked on an individual level. In an unselected nationwide cohort, we used multivariate Poisson regression models controlling for age, gender, comorbidity, concomitant medication, socio-economic data, and calendar year, to assess the risk of VTE associated with psoriasis. A total of 35,138 patients with mild and 3,526 patients with severe psoriasis were identified and compared with 4,126,075 controls. Patients with psoriasis had higher incidence rates per 1000 person-years of VTE than controls (1.29, 1.92, and 3.20 for controls, mild psoriasis, and severe psoriasis, respectively). The rate ratio (RR) of VTE was elevated in all patients with psoriasis with RR 1.35 (95% confidence interval [CI] 1.21-1.49) and RR 2.06 (CI 1.63-2.61) for mild and severe psoriasis, respectively. Exclusion of patients with malignancies, and censoring of patients undergoing surgery did not alter the results.Conclusion: This nationwide cohort study indicates that patients with psoriasis are at increased risk of VTE. The risk was highest in young patients with severe psoriasis. Physicians should be aware that patients with psoriasis may be at increased risk of both venous and arterial thromboembolic events. Citation: Ahlehoff O, Gislason GH, Lindhardsen J, Charlot MG, Jørgensen CH, et al. (2011) Psoriasis Carries an Increased Risk of Venous Thromboembolism: A Danish Nationwide Cohort Study. PLoS ONE 6(3): e18125.
# Introduction
Psoriasis is a prevalent chronic immunoinflammatory disease affecting approximately 2% of the population [bib_ref] Current and future management of psoriasis, Menter [/bib_ref]. It is associated with cardiovascular risk factors, atherothrombotic events, and markers of hypercoagulability, including platelet activation and hyperhomocysteinemia [bib_ref] AJC editor's consensus: psoriasis and coronary artery disease, Friedewald [/bib_ref] [bib_ref] Unfavorable cardiovascular risk profiles in untreated and treated psoriasis patients, Wakkee [/bib_ref] [bib_ref] Cardiovascular disease and risk factors among psoriasis patients in two US healthcare..., Kimball [/bib_ref] [bib_ref] Psoriasis and occlusive vascular disease, Mcdonald [/bib_ref] [bib_ref] Risk of myocardial infarction in patients with psoriasis, Gelfand [/bib_ref] [bib_ref] The risk of stroke in patients with psoriasis, Gelfand [/bib_ref] [bib_ref] Association of psoriasis with coronary artery, cerebrovascular, and peripheral vascular diseases and..., Prodanovich [/bib_ref] [bib_ref] Increased risk for cardiovascular mortality in psoriasis inpatients but not in outpatients, Mallbris [/bib_ref] [bib_ref] Patients with severe psoriasis are at increased risk of cardiovascular mortality: cohort..., Mehta [/bib_ref] [bib_ref] Psoriasis is associated with clinically significant cardiovascular risk: a Danish nationwide cohort..., Ahlehoff [/bib_ref] [bib_ref] Psoriasis and atherothrombotic diseases: disease-specific and non-disease-specific risk factors, Gisondi [/bib_ref] [bib_ref] Inflammation and hypercoagulable state in adult psoriatic men, Karabudak [/bib_ref] [bib_ref] Platelet activation in patients with psoriasis: increased plasma levels of platelet-derived microparticles..., Tamagawa-Mineoka [/bib_ref]. Venous thromboembolism (VTE), i.e., deep venous thrombosis and pulmonary embolism, is prevalent and potentially lethal, and is associated with various conditions including cancer, prolonged immobilization, and major surgery or trauma (secondary VTE) [bib_ref] Risk factors for venous thromboembolism, Goldhaber [/bib_ref]. Interestingly, VTE is also associated with certain cardiovascular risk factors (e.g., obesity, hypertension, and smoking), and arterial cardiovascular events [bib_ref] Risk factors for venous thromboembolism, Goldhaber [/bib_ref] [bib_ref] Risk factors for venous thromboembolism: results from the Copenhagen City Heart Study, Holst [/bib_ref] [bib_ref] Venous thromboembolism and subsequent hospitalisation due to acute arterial cardiovascular events: a..., Sorensen [/bib_ref]. Moreover, the risk of VTE is associated with elevated levels of C-reactive protein [bib_ref] C-reactive protein and risk of venous thromboembolism in the general population, Zacho [/bib_ref] , and increased risk of VTE was recently demonstrated in patients with inflammatory bowel disease [bib_ref] Venous thromboembolism during active disease and remission in inflammatory bowel disease: a..., Grainge [/bib_ref] [bib_ref] Risk of Venous Thromboembolism With Inflammatory Bowel Disease, Saleh [/bib_ref]. The potential impact of psoriasis on the risk of VTE, however, has not been examined in detail previously [bib_ref] Psoriasis and occlusive vascular disease, Mcdonald [/bib_ref]. We therefore examined the risk of VTE in patients with psoriasis in a cohort study based on nationwide prospectively recorded registers with the underlying hypothesis that patients with psoriasis carry an increased risk of VTE.
# Methods
## Ethics
This study was approved by The Danish Data Protection Agency , and data at the individual case level were made available to us by the national registers in anonymized form. Register studies do not require ethical approval in Denmark. The authors had full access to all data and take full responsibility for its integrity.
## Study population and data sources
The study population comprised the entire Danish population aged $18 years on January 1, 1997. The population was followed until December 31, 2006, or death. Patients with prevalent psoriasis, a history of previous VTE, and patients receiving vitamin K antagonist treatment at baseline were not included. The study was conducted and reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations [bib_ref] The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines..., Von Elm [/bib_ref]. In Denmark, all citizens have a unique personal civil registration number that enables individual level-linkage of information across nationwide prospectively recorded registers. All medications dispensed from pharmacies were obtained from the National Prescription Registry (the Danish Registry of Medicinal Product Statistics), where all dispensed prescriptions have been recorded since 1995 ensuring complete registration. Patients with psoriasis were identified by their claims of prescriptions for vitamin D derivatives. Patients were included when claiming their second prescription for these agents to ensure persistent medical treatment for psoriasis as previously accepted [bib_ref] Psoriasis is associated with clinically significant cardiovascular risk: a Danish nationwide cohort..., Ahlehoff [/bib_ref]. Subjects with prevalent psoriasis were defined as patients fulfilling the psoriasis criteria prior to study start. Morbidity was obtained from the Danish National Patient Register in which all hospital contacts, diagnoses, and invasive procedures have been recorded since 1978 according to the International Classification of Diseases (ICD), i.e. ICD-8 until 1994 and ICD-10 thereafter. Patients with severe psoriasis were identified by hospitalizations (including out-patient visits) for psoriasis (ICD-10 L40) or psoriatic arthritis (M070-M073) and included at the time of their third diagnosis. The severe psoriasis classification has previously been validated and pose an acceptable measure of severe disease [bib_ref] Psoriasis is associated with clinically significant cardiovascular risk: a Danish nationwide cohort..., Ahlehoff [/bib_ref]. Comorbidity at study entry was described by Charlson's index, as defined by 19 prespecified diagnoses at study entry and up to 1 year previously [bib_ref] Charlson scores based on ICD-10 administrative data were valid in assessing comorbidity..., Nuttall [/bib_ref]. All deaths were identified from the Central Population Register, in which deaths are recorded within 2 weeks. Causes of death, recorded according to ICD codes, were obtained from the National Causes of Death Register. Socioeconomic status was defined by the individual average yearly gross income in a 5-year period prior to inclusion and patients were divided into quintiles according to their income.
## Medical treatment
Prescriptions claimed for topical vitamin-D derivatives (ATC D05AX), i.e., topical treatment used exclusively for psoriasis [bib_ref] Current and future management of psoriasis, Menter [/bib_ref] , and vitamin K antagonists (B01AA) were used for in-and exclusion of subjects as described above. Pharmacologically managed cardiovascular diseases and cardiovascular risk factors, including hypertension, dyslipidemia, and diabetes mellitus were identified by prescriptions for platelet inhibitors (B01AC), betablockers (C07), angiotensin-converting enzyme inhibitors (ACEI)/angiotensin 2 receptor antagonists (C09), loop diuretics (C03C), spironolactone (C03D), statins (C10A), and glucose-lowering drugs (A10) claimed up to 6 months prior to study initiation [fig_ref] Table 1: Baseline characteristics of the study population [/fig_ref].
## Study endpoints
The following primary endpoint was assessed: first time inhospital discharge diagnosis of VTE (ICD-10 I26 and I80.1-I80.9). VTE diagnoses made in emergency departments were not included. The diagnosis of VTE (deep venous thrombosis and pulmonary embolism) in hospitalized patients has previously been validated in the Danish National Patient Register with a positive predictive value of 75% when excluding diagnoses from emergency departments [bib_ref] Venous thromboembolism discharge diagnoses in the Danish National Patient Registry should be..., Severinsen [/bib_ref]. Hospitalizations with the specific diagnosis of pulmonary embolism (ICD-10 I26) were evaluated as a secondary endpoint.
# Statistical analysis
Baseline characteristic are presented as percentages and means with standard deviations. Unadjusted event rates are summarized as events per 1000 person-years. The rate ratios (RRs) and 95% confidence interval (CI) of VTE were estimated by time-dependent Poisson regression models adjusted for age, calendar year, concomitant medication, comorbidity, socioeconomic data, and gender. Psoriasis status was included as a time-dependent variable, i.e., patients were only considered at risk from the time they complied with the inclusion criteria. Age and calendar year were also included as time-dependent variables. Comorbidity, socioeconomics, and concomitant medication were included as fixed variables obtained at baseline.
Sensitivity analyses with exclusion of patients at increased risk of secondary VTE at baseline, including those with cancer (ICD-10 C00-C97), the risk of which may be increased in psoriatic patients [bib_ref] Psoriasis and risk of incident cancer: an inception cohort study with a..., Brauchli [/bib_ref] , and rheumatological disease including psoriatic arthritis (ICD-10 M05-07.3, M32-34, and M35.3) were conducted. The potential impact of immobilization and hemostatic activation due to surgery (any surgical procedure code) was assessed in a sensitivity analysis with censoring of subjects at the time of the surgical procedure. Surveillance bias was addressed in analyses with inclusion of patients with psoriasis at the time of their first vitamin D prescription claim or first psoriasis diagnosis, and by exclusion of all subjects with a history of hospitalizations up to 1 year prior to study start. A two-sided p-value,0.05 was considered statistically significant. Interaction between severe psoriasis and age was found, and therefore both overall and age-stratified estimates are presented. All statistical analyses were performed with the use of SAS statistical software version 9.2 and STATA software version 11.
# Results
The cohort study had a maximum follow-up of 10 years. A total of 35,138 patients with mild psoriasis, 3526 patients with severe psoriasis, and 4,126,075 controls were identified. Details on study population selection are presented in [fig_ref] Figure 1: Flowchart of study population selection [/fig_ref]. Patients with severe psoriasis were more often men and more often treated with cardiovascular medication and glucose-lowering drugs [fig_ref] Table 1: Baseline characteristics of the study population [/fig_ref]. We have previously demonstrated an age-and severity-dependent increase in risk of cardiovascular and all-cause mortality with psoriasis in this nationwide cohort [bib_ref] Psoriasis is associated with clinically significant cardiovascular risk: a Danish nationwide cohort..., Ahlehoff [/bib_ref]. Patients with psoriasis had higher overall VTE rates than controls, i.e., 1.29, 1.92, and 3.20 per 1000 person-years for controls, mild psoriasis, and severe psoriasis, respectively. This pattern of increased VTE rates was sustained in the age-stratified event rates [fig_ref] Table 2: Age-stratified incidence rates per 1000 person-years [/fig_ref] , and in analyses of the secondary endpoint of pulmonary embolism, with corresponding overall pulmonary embolism rates 0.44, 0.61, and 1.02 per 1000 person-years.
## Overall and age-stratified time-dependent multivariable adjusted poisson regression
Psoriasis was associated with a severity-and age-dependent increased risk of VTE. The overall RR was 1.35 (CI 1.21-1.49) and 2.06 (CI 1.63-2.61) for mild and severe psoriasis, respectively. The corresponding overall risks for pulmonary embolism were comparable with RRs 1.14 (CI 0.95-1.37) and 1.88 (CI 1.22-2.89). Age-stratified estimates for VTE are presented in [fig_ref] Table 3: Adjusted age-stratified rate ratios and 95% confidence intervals of venous thromboembolism [/fig_ref].
## Sensitivity analyses: influences of secondary vte and surveillance bias
Exclusion of all subjects with a history of cancer or rheumatological disease did not attenuate the association between psoriasis and VTE, with RRs 1.34 (CI 1.21-1.49) and 1.99 (CI 1.56-2.53) for mild and severe psoriasis, respectively. Furthermore, censoring of participants undergoing a surgical procedure had no significant influence on the VTE risk estimates, i.e., RR 1.20 (CI 0.96-1.51) and 2.55 (CI 1.53-4.24) for mild and severe psoriasis, respectively.
With use of the less restrictive psoriasis diagnosis criterion, where patients with psoriasis were identified by their first vitamin D prescription claim or first psoriasis diagnosis, a total of 55,422 patients with mild psoriasis and 11,532 patients with severe psoriasis were included, and 22,018 patients with prevalent psoriasis were excluded. This psoriasis diagnosis definition corresponded to a 10-year psoriasis prevalence of 2.2%, and the VTE risk estimates obtained hereby were comparable to the results of the primary analyses, including the observed doseresponse relationship with psoriasis severity, with overall RRs 1.44 (CI 1.34-1.56) and 1.88 (CI 1.62-2.19) for mild and severe psoriasis, respectively. Exclusion of all subjects with a history of out-patient and/or in-patient hospital contacts up to 1 year prior to study start also did not alter the results (mild psoriasis RR 1.
# Discussion
This nationwide study of VTE risk in patients with psoriasis suggested an association between psoriasis and increased risk of VTE. The present study is, to our knowledge, the first to examine the psoriasis-related risk of VTE in a large unselected cohort. We demonstrated a clear dose-response relationship between VTE risk and psoriasis severity, and the results were further corroborated by sensitivity analyses addressing the influence of secondary VTE and surveillance bias, respectively. These findings indicate that VTE should be added alongside the list of adverse atherothrombotic events, e.g., acute myocardial infarction, which has been linked to psoriasis [bib_ref] AJC editor's consensus: psoriasis and coronary artery disease, Friedewald [/bib_ref] [bib_ref] Unfavorable cardiovascular risk profiles in untreated and treated psoriasis patients, Wakkee [/bib_ref] [bib_ref] Cardiovascular disease and risk factors among psoriasis patients in two US healthcare..., Kimball [/bib_ref] [bib_ref] Psoriasis and occlusive vascular disease, Mcdonald [/bib_ref] [bib_ref] Risk of myocardial infarction in patients with psoriasis, Gelfand [/bib_ref] [bib_ref] The risk of stroke in patients with psoriasis, Gelfand [/bib_ref] [bib_ref] Association of psoriasis with coronary artery, cerebrovascular, and peripheral vascular diseases and..., Prodanovich [/bib_ref] [bib_ref] Increased risk for cardiovascular mortality in psoriasis inpatients but not in outpatients, Mallbris [/bib_ref] [bib_ref] Patients with severe psoriasis are at increased risk of cardiovascular mortality: cohort..., Mehta [/bib_ref] [bib_ref] Psoriasis is associated with clinically significant cardiovascular risk: a Danish nationwide cohort..., Ahlehoff [/bib_ref].
Psoriasis is an immunoinflammatory disease characterized by Thelper (Th)1-and Th17-driven inflammation with a striking overlap of inflammatory markers and mediators with atherosclerosis [bib_ref] Inflammation, atherosclerosis, and coronary artery disease, Hansson [/bib_ref]. Indeed, these mechanisms may form the background, in part, for the increasing evidence linking psoriasis to risk of atherothrombotic cardiovascular events, e.g., acute myocardial infarction and cardiovascular mortality [bib_ref] AJC editor's consensus: psoriasis and coronary artery disease, Friedewald [/bib_ref] [bib_ref] Unfavorable cardiovascular risk profiles in untreated and treated psoriasis patients, Wakkee [/bib_ref] [bib_ref] Cardiovascular disease and risk factors among psoriasis patients in two US healthcare..., Kimball [/bib_ref] [bib_ref] Psoriasis and occlusive vascular disease, Mcdonald [/bib_ref] [bib_ref] Risk of myocardial infarction in patients with psoriasis, Gelfand [/bib_ref] [bib_ref] The risk of stroke in patients with psoriasis, Gelfand [/bib_ref] [bib_ref] Association of psoriasis with coronary artery, cerebrovascular, and peripheral vascular diseases and..., Prodanovich [/bib_ref] [bib_ref] Increased risk for cardiovascular mortality in psoriasis inpatients but not in outpatients, Mallbris [/bib_ref] [bib_ref] Patients with severe psoriasis are at increased risk of cardiovascular mortality: cohort..., Mehta [/bib_ref] [bib_ref] Psoriasis is associated with clinically significant cardiovascular risk: a Danish nationwide cohort..., Ahlehoff [/bib_ref]. While it is wellestablished that inflammation plays a central role in the pathogenesis of atherosclerosis and atherothrombosis [bib_ref] Inflammation, atherosclerosis, and coronary artery disease, Hansson [/bib_ref] , the potential association between inflammation as determined by circulating levels of inflammatory markers (e.g., C-reactive protein) and risk of VTE is, however, unclear at present [bib_ref] C-reactive protein and risk of venous thromboembolism in the general population, Zacho [/bib_ref] [bib_ref] Review: C-reactive protein and venous thromboembolism: causal or casual association?, Lippi [/bib_ref] , More than three decades ago, a small case-control study suggested that the risk of VTE and other adverse cardiovascular events was increased in psoriatic patients, but the risk of VTE was not specifically addressed in that particular study of highly selected patients [bib_ref] Psoriasis and occlusive vascular disease, Mcdonald [/bib_ref] , On the other hand, VTE has been associated with the increased risk of atherothrombotic events and was recently also linked to inflammatory bowel disease [bib_ref] Risk factors for venous thromboembolism, Goldhaber [/bib_ref] [bib_ref] Risk factors for venous thromboembolism: results from the Copenhagen City Heart Study, Holst [/bib_ref] [bib_ref] Venous thromboembolism and subsequent hospitalisation due to acute arterial cardiovascular events: a..., Sorensen [/bib_ref] [bib_ref] C-reactive protein and risk of venous thromboembolism in the general population, Zacho [/bib_ref] [bib_ref] Venous thromboembolism during active disease and remission in inflammatory bowel disease: a..., Grainge [/bib_ref] , In addition to increased markers of systemic inflammation, a psoriasis severitydependent increase in platelet activation, e.g., platelet hyperaggregability and increased levels of platelet-derived microparticles and soluble P-selectin, has been demonstrated [bib_ref] Inflammation and hypercoagulable state in adult psoriatic men, Karabudak [/bib_ref] [bib_ref] Platelet activation in patients with psoriasis: increased plasma levels of platelet-derived microparticles..., Tamagawa-Mineoka [/bib_ref]. Along this line, platelet activity regeneration time following aspirin ingestion may be shorter in psoriasis patients compared to controls [bib_ref] Platelet activation in psoriasis, Berrettini [/bib_ref] , and hyperhomocysteinemia, which may be associated with atherothrombosis and VTE, has been reported in these patients, suggestive of a prothrombotic predisposition [bib_ref] Psoriasis and atherothrombotic diseases: disease-specific and non-disease-specific risk factors, Gisondi [/bib_ref] [bib_ref] Inflammation and hypercoagulable state in adult psoriatic men, Karabudak [/bib_ref]. Moreover, methotrexate is commonly used in severe psoriasis and though this agent may be associated with lower risk of atherothrombosis in patients with psoriasis and psoriatic arthritis, it is also associated with hyperhomocysteinemia [bib_ref] AJC editor's consensus: psoriasis and coronary artery disease, Friedewald [/bib_ref] [bib_ref] Methotrexate reduces incidence of vascular diseases in veterans with psoriasis or rheumatoid..., Prodanovich [/bib_ref]. Finally, it is also interesting, that statins, which have pleiotropic effects including antiinflammatory properties, may exert beneficial actions on psoriatic skin manifestations and may lower the risk of VTE in healthy individuals and in patients with atherosclerosis [bib_ref] Efficacy of simvastatin in plaque psoriasis: A pilot study, Shirinsky [/bib_ref] [bib_ref] Statins, fibrates, and venous thromboembolism: a meta-analysis, Squizzato [/bib_ref]. In view of the considerations presented above, several mechanisms may contribute to the increased risk of VTE in patients with psoriasis, e.g., coincident risk factors, inflammation and hypercoagulability, and more studies are clearly warranted to examine individual merits of these contributions and, for example, the potential contribution of VTE to the increased cardiovascular and overall mortality observed in patients with severe psoriasis [bib_ref] Increased risk for cardiovascular mortality in psoriasis inpatients but not in outpatients, Mallbris [/bib_ref] [bib_ref] Patients with severe psoriasis are at increased risk of cardiovascular mortality: cohort..., Mehta [/bib_ref] [bib_ref] Psoriasis is associated with clinically significant cardiovascular risk: a Danish nationwide cohort..., Ahlehoff [/bib_ref] [bib_ref] The risk of mortality in patients with psoriasis: results from a population-based..., Gelfand [/bib_ref]. In addition to the plausible pathophysiological mechanisms, the likelihood of a causal relationship between psoriasis and augmented risk of VTE as found in our study was strengthened by our study cohort design which controlled for important measured confounders, the finding of a consistent dose-response relationship between psoriasis severity and risk of VTE, and the sensitivity analyses that indicated that the elevated VTE risk was not driven by secondary VTE. Some further strengths and limitations of the study should be discussed. The main limitation of our study is inherent to its observational nature that precludes conclusions on causality. The large number of participants, the nationwide coverage of prospectively recorded registries, and the complete follow-up strengthens the study. Specifically, the use of nationwide registries of hospitalizations and dispensed prescriptions from all pharmacies in Denmark where healthcare is readily accessible and essentially free of charge enabled us to reduce potential surveillance bias and avoid selection bias related to e.g., subject gender, age, socioeconomic status, healthcare insurance, and labor market association. This was further supported by sensitivity analyses addressing this key issue. The use of prospectively identified patients with psoriasis ensured a rational basis for allocation of exposure time and, due to the very large sample size, did not imply any significant loss of information. We were, however, unable to identify patients with psoriasis treated with topical corticosteroids alone (due to the liberal indications for these agents), and the results may not apply to these patients. Since data of in-hospital treatment are, at present, not available and as systemic antiinflammatory psoriasis treatment, i.e., methotrexate and biological agents, is generally provided directly to the patients by the respective hospital departments and are therefore not captured by the registers, we were unable to use systemic anti-inflammatory treatment as a measure of disease severity and could not address the potential impact of various systemic treatment strategies on the risk of VTE. Indeed, future studies to address the impact of various psoriasis treatment strategies on cardiovascular risk are likely to provide important insights [bib_ref] More than skin deep: atherosclerosis as a systemic manifestation of psoriasis, Alexandroff [/bib_ref]. Also, the use of hospitalizations and prescription claims to describe comorbidity, concomitant medication, and outcomes implies a possibility of underestimation of these variables since we were unable to capture data from those who did not come to medical attention. The possibility of residual confounding is always present in observational studies, and we were unable to account for some important risk factors of VTE, e.g., obesity and smoking. Importantly, we controlled for socioeconomic status that correlate with the presence of obesity and smoking. Finally, the Danish population is predominantly of Caucasian descent and extrapolation to other ethnicities should only be done with caution.
# Conclusion
This first nationwide cohort study indicates that patients with psoriasis are at increased risk of VTE. The risk was highest in young patients with severe disease. Further prospective studies are needed to confirm this association, but physicians should be aware that patients with psoriasis may be at increased risk of both venous and arterial thromboembolic events.
[fig] Figure 1: Flowchart of study population selection. VTE: venous thromboembolism. doi:10.1371/journal.pone.0018125.g001 [/fig]
[table] 33 [: CI 1.19-1.48] and severe psoriasis 1.93 [CI 1.50-2.48]). [/table]
[table] Table 1: Baseline characteristics of the study population. [/table]
[table] Table 2: Age-stratified incidence rates per 1000 person-years. IR: incidence rate; CI: 95% confidence interval; VTE: venous thromboembolism. doi:10.1371/journal.pone.0018125.t002 [/table]
[table] Table 3: Adjusted age-stratified rate ratios and 95% confidence intervals of venous thromboembolism (VTE). RR: rate ratio; CI: 95% confidence interval. doi:10.1371/journal.pone.0018125.t003 [/table]
|
Tuberculous Dactylitis in Adult – A Case Report and Review of Literature
Introduction: Tubercular dactylitis is an uncommon form of musculoskeletal tuberculosis (TB), especially after the age of 5 years without any risk factors.Case Report: A case of 49-year-old male presented with pain, swelling, and discharging sinus of proximal phalanx of third digit of the right hand with no constitutional symptoms for 2 years, with multiple failed treatment in form of antibiotic therapy. There were no risk factors such as immunodeficiency or any co-morbidities. The plain radiograph was suggestive of increased bone density with mild periosteal reaction; magnetic resonance imaging was suggestive of tubercular osteomyelitis. A biopsy was performed, the gene expert of the sample revealed TB with no drug resistance. The patient was managed with anti-tubercular drugs with complete resolution of clinical and radiological symptoms at 1-year followup. Conclusion: TB should be considered a differential in patients with chronic soft-tissue or skeletal lesions even in the absence of the usual risk factors because with treatment it still carries a good prognosis.AbstractCase ReportCase PresentationOsteoarticular tuberculosis (TB) accounts for only 1-3% of all the TB infections [1]. Hand and feet are involved in 14 and 10 % of the case of osteoarticular TB[1,2]. Tuberculous infection of the metatarsal, metacarpal, and phalanges of hand and feet is known as tuberculous dactylitis. Tubercular dactylitis is rare and it constitutes 2-4% of skeletal TB[3,4]. The hand is more commonly involved than a foot in tuberculous dactylitis[5,6,7]. Tuberculous dactylitis is uncommon after the age of 5 years. It poses a serious diagnostic challenge due to its unusual presentations, inadequate diagnostic skills, limited understanding of the management, and many overlapping features with other diseases. Here, we report a rare case of TB dactylitis in adults without any risk factor diagnosed and treated successfully with anti-tubercular therapy.
# Introduction
A 49-year-old male presented with pain, swelling, and discharging sinus over the proximal phalanx of the third digit of the right hand for 2 years. The patient was apparently asymptomatic 2 years back when he developed pain and swelling of the third digit which was insidious in onset and gradually progressive finally culminating with a discharging sinus 1 year back. There was no history of trauma, cough, weight loss, decreased appetite, and any other risk factors. The patient was previously treated at other centers with incision and drainage followed by antibiotics. On examination, there was a firm, globular swelling at the base of the third digit with a discharging sinus over the proximal phalanx of the third digit of the right hand. No local inflammatory signs were found.
The finger range of motion was painful and restricted. Plain radiographs showed increased bone density, mild periosteal reaction with an osteolytic lesion of proximal phalanx of the third fingerTuberculous dactylitis usually affects children <5 years of age; it needs to be considered as a differential diagnosis in adults with unusual soft-tissue or skeletal lesions even in the absence of u s u a l r i s k f a c to r s s u c h a s t u b e r c u l a r c o n t a g i o n , immunodeficiencies, and primary foci of TB elsewhere in the body. A high degree of suspicion accompanied by proper clinical examination and radiology helps in diagnosing this rare entity. Once the diagnosis is confirmed by biopsy, Antitubercular chemotherapy is the treatment of choice.
# Conclusion
Anti-tubercular therapy was started which included four drugs (isoniazid, rifampicin, pyrazinamide, and ethambutol) for 2 months, followed by three drugs (isoniazid, rifampicin, and ethambutol) for 7 months. The patient finger was immobilized for 1 month. The patient responded to the anti-tubercular therapy with complete resolution of the pain, swelling, and discharging sinus was satisfactorily healed. Radiographs of the right hand after the completion of the treatment course at 1 year revealed a decrease in the size of the osteolytic lesion with the resolution of associated sclerosis. At 1 year follow-up, the patient had a full range of motion of fingers with no other symptoms.
Tuberculous infection of the metatarsal, metacarpal, and phalanges of hand and feet is known as tuberculous dactylitis. Tuberculous dactylitis is a rare <2-4% of skeletal TB. It is most common by hematogenous spread from the lungs. Fingers are more commonly affected than the toes, proximal phalanx of the index, and middle fingers are the most common sites involved. Tuberculous dactylitis is uncommon beyond the age of 5 years after the formation of the epiphyseal center. It usually becomes symptomatic after 1-3 years following the initial infection. In children, the disease usually involves multiple bones. Children tubular bone has an abundant blood supply through a large nutrient artery. Infection is usually in the center of the marrow cavity and short tubular bone is gradually becomes tuberculous granuloma. When the nutrient artery of involved bone is occluded it leads to formation of the sequestrum, there is endosteal destruction and simultaneous new bone formation. It is also sometimes called "Spina ventosa" because radiologically it appears to be a cystic expansion of the phalanges. Typically, the disease presents as a chronic, painful, and progressive swelling of the fingers or toes, affecting most commonly the proximal phalanx and metacar pal s of the hand. Multifocal involvement , sequestration, discharging sinus, and mutilation of the fingers occur more often in children or patients who are suffering from concomitant immunodeficiency. Abscess and sinus formation is common leading to secondary bacterial infections. A radiologic picture usually shows an osteolytic lesion, mild to no periosteal reaction, areas of bone sclerosis, and destruction. MRI is the modality of choice to see early marrow and softtissue involvement. In spina ventosa, the bone may take shape of honey combing, diffuse infiltration, cystic lesion, or rarely bone atrophy. Tuberculous dactylitis commonly occurs in the background of tubercular contag ion or a state of immunosuppression and for it to present in the absence of both is a rare entity, as it was in our case. The mainstay of diagnosis of tubercular dactylitis remains the histopathological examination of the biopsy specimens, demonstration of mycobacterium TB on Gene-Xpert studies, localizing acid-fast bacilli on Z-N stain. The culture of mycobacterium from bone tissue remains the Gold standard. Laboratory investigations such as white blood counts, differential counts, ESR, and CRP do not aid in the diagnosis of tubercular dactylitis; however, they tend to help in ruling out the other possible diagnosis. The treatment of TB dactylitis consists of long-term anti-tubercular therapy consisting of four drugs isoniazid, rifampicin, pyrazinamide, and ethambutol for 2 months, followed by isoniazid and rifampicin for 6-12 months. Excellent clinical-radiological response is seen irrespective of the stage of the disease in which it is started. Ankylosis of the joint remains one of the dreaded complications of the disease; however, in most cases; complete remission is seen following the completion of antitubercular chemotherapy.
Khanna et al. reported a case of TB of distal phalanx of great toe presenting as painless paronychia with pseudopterygium and nail dystrophy. We have done a review of literature in the past 10 years on tuberculous dactylitis in adults and tabulated.
were sent for histopathology which revealed chronic granulomatous inflammation. Gene-Xpert was done which detected TB and no resistance to first-line drugs was found.
# Discussion
|
Anti-Tumor Activities of Bioactive Phytochemicals in Sophora flavescens for Breast Cancer
Patients with breast cancer and breast cancer survivors are frequent users of botanicals and their bioactive phytochemicals. In China, active ingredients in Sophora flavescens like matrine (MT), oxymatrine (OMT), other Sophora flavescens alkaloids andCompound Kushen Injection (CKI) are extensively used for multiple malignant tumors. In vivo and in vitro studies have confirmed that these activities or injection have significant effects on relieving symptoms, alleviating side effects after chemotherapy and improving the quality of life of breast cancer patients, where there is evidence for efficacy. A large number of experimental studies have also revealed that they can inhibit the proliferation, invasion and migration of breast cancer cells according to different mechanisms. This provides promising valuable supportive therapies for prevention, treatment and postoperative recovery of breast cancer. Rigorous clinical research and experimental studies reflect integrative care as it is used in hospital is needed to responsibly move this field forward. This review summarizes an up to date knowledge of the available bioactive phytochemicals, their discovery, current clinical and experimental status.
# Introduction
Application of botanicals to treat solid tumors all over the world has a long history. Many anti-cancer agents have been identified from botanicals, although most of their anti-tumor mechanisms are under discussion. Botanicals are widely used in combination with radiotherapy or chemotherapy to improve the curative efficacy of cancer and reduce complications and side effect. Traditional Chinese Medicine (TCM), including herbal remedies, has been practised for more than five thousand years and many of the herbal medicines have been intensively investigated. In China, TCM formulas are extensively used for modulating immune function and improving the quality of life of cancer patients who undergoing chemotherapy or radiotherapy.Because of the low toxicity and wide range of biological activities of botanicals, many natural products have been applied as alternative treatments for many cancers including breast cancer.Breast cancer is the most common noncutaneous malignancy among women worldwide, and the cure and prevention of cancer effects of botanicals are more and more thought highly of now. Although the mortality rate of breast cancer is decreasing and survival rate is increasing year by year, it is also estimated that more than 1000 thousand women are newly diagnosed with breast cancer each year worldwide and that over 400,000 cases will die from breast cancer. China National Cancer Center announced that the incidence of breast cancer is about 34,000 in 2015; Breast cancer accounted for 17.10% of all cancers and was the fifth leading cause of death in female malignant tumors, accounting for 8.16% of all cancer deaths. It has been a serious burden for societies of China and the whole world for a long time. The main treatment options for breast cancer cases include surgery, chemotherapy, hormone and immunological therapy, radiotherapy. Multidrug chemotherapy can reduce the recurrence rate while increase multidrug resistance. Even treated at an early stage, approximately 30% of breast cancer patients suffer from metastasis and relapse. Resistance to chemotherapy and subsequent recurrence and metastasis are two main intractable problems in most patients with breast cancer, which made patients bear additional economic and spiritual stress. The high toxicity of the treatments of breast cancer to normal tissues is practically inevitable. Seeking for effective and safe treatment strategies for breast cancer is an urgent task for medical researchers.In the past several decades, many botanicals are used for breast cancer treatment. So far, it has been proved that there are many kinds of herbal extracts that have effects on the treatment of breast cancer, involving various mechanisms, such as Saikosaponin A, Emodin, Sanguisorba, Oridonin, Curcumol, Triptolide, artesunate, cycloartane triterpenoid derivative.Breast cancer prevention is equally important for high-risk women with older age and postmenopause status. Glycyrrhiza Species (Licorice) and its three medicinally used species of licorice (G. glabra, G. uralensis, and G. inflata), Silybum marianum (Milk Thistle) and its extract silibin B or silymarin, Angelica sinensis (Dong Quai) and its dong quai's phytoconstituents ferulic acid or Z-ligustilide, Epimedium, ginger, black cohosh and flaxseed have been reported for breast cancer prevention. These botanicals and their bioactive phytochemicals play the role of breast cancer prevention through four major mechanisms including hormonal, chemical, inflammatory, and epigenetic.Vinca alkaloids and taxol are two anticancer drugs extracted from naturally occurring compounds; however, toxic side effects place restrictions on their use. Therefore, developing novel natural with low toxicity and high therapeutic selectivity is clinically important. Novel molecular markers and genetic mutations are now comprehensively identified in breast cancer genome sequencing projects. At present, most studies on the underlying mechanisms of botanicals on breast cancer are limited to the molecular level, and the role of traditional medicine in gene level is less studied.
## Sophora flavescens
Kushen (Radix Sophorae Flavescentis/Sophorae flavescens), a traditional Chinese herbal medicine, is the dried root of Sophora flavescens Ai. In 200 A.D., Kushen was described in the traditional Chinese book Shen Nong Ben Cao Jing for the first time and it was used to treat inflammation, solid tumors, and other compounds. The main chemical components in Kushen include alkaloids, flavonoids, alkylxanthones, quinones, triterpene glycosides, fatty acids, and essential oils. Among them, Kushen alkaloids (KS-As) account for 3% and Kushen flavonoids (KS-Fs) account for 1.5%, which are two characteristic components of Kushen. MT and OMT consist of approximately 20% of the total alkaloids in Kushen. Sophora flavescens water decoction recorded in Chinese Pharmacopoeia can treat diseases such as leukorrhea, jaundice, enteritis, scabies, dysentery, carbuncles as well as pyogenic infections of the skin. KS-As or KS-Fs preparations from different manufacturers are more or less different, but their anti-tumor function was consistent with that in published reports.
Looking for the active substances is a critical approach to the development of new drugs to treat cancers. As unique tetracyclo-quinolizidine alkaloids, MT and OMT are found only in Sophora species thus far. Various patents of Kushen extracts have been applied in China, USA and other countries.In 2005, Kushen extracts were approved by the Chinese FDA for malignant tumor treatment. Kushen alkaloids are effective anticancer drugs and widely used in cancer treatment in China now. The mechanism of breast cancer treatment and prevention includes inhibition of cell proliferation, induction of cell differentiation or apoptosis, ant-inflammatory, antioxidant, and antiangiogenesis. This literature will introduce the anticancer effects of CKI and other phytochemicals extracted from Sophora flavescens .
## Compound kushen injection
Compound Kushen Injection (CKI), commonly known as the Yanshu Injection, is exacted from Kushen (Radix Sophorae Flavescentis). The chemical fingerprint of CKI contains at least 8 different components, with primary 4 principles including MT, OMT, sophoridine and oxysophocarpine, which exhibit a variety of pharmacological activities. In China, CKI has been extensively used integrated with both TCMs and modern medicines when treating tumors, including cardiovascular and rheumatic diseases asExperimental and clinical research has shown that CKI and its compounds exhibit anti-cancer actions, such as inhibiting cancer cell proliferation, metastasis and invasion, accelerating cell apoptosis, inducing cell differentiation and cycle arrest, restraining angiogenesis, reversing multidrug resistance, and reducing chemotherapy-induced toxicity. In China, the study of CKI is mostly focused on clinical practice. The involvement of CKI and its mechanisms of breast cancer treatment are reviewed below. Uncontrolled migration of cells away from the primary tumor leads to cancer progression, which is the main cause of cancer-related deaths. Reconstitution and fractionation of CKI highlight the importance of combinations of multiple compounds for antitumor activity. Live-cell imaging confirmed CKI strongly reduced the migration of MDA-MB-231 cells, with down-regulation of actin cytoskeletal and focal adhesion genes.The cell cycle was identified as the potential primary target pathway of CKI in MCF-7 cells. CKI may induce apoptosis in MCF-7 cells via a p53 independent mechanism; furthermore, lncRNA H19, which over-expressed in breast cancer, was significantly down-regulated in cells treated with CKI.In Hep G2 and MDA-MB-231 cells, protein levels in cell cycle pathway, DNA repair pathway and DNA double-strand breaks (DSBs) could be altered by CKI. It declined protein levels for cell cycle and DNA repair while increasing the level of DSBs. At the same time, energy metabolism was reduced because of reduced glucose consumption and cellular energy charge. These results support a model of action of CKI that multiple compounds rather than individual components affect multiple targets and the synergistic.Breast cancer stem cells (CSCs) are the sources of oncogenesis, cancer metastasis and relapse. SP cells exhibit properties similar to CSCs because of the capabilities to proliferate more tumorigenic cells than other populations. Compared with cisplatin and saline (as a control) group, CKI suppressed the size of MCF-7 SP population (approximately 90%), and down-regulated the main genes of Wnt signaling pathway to suppress tumor growth, which suggested that CKI may serve as a novel drug targeting CSCs.Anti-angiogenesis may be one of the important mechanisms of CKI in inhibiting tumor growth. Different doses of CKI injected into the mice with tumor exhibited different angiogenesis effects: With the increase of compound Kushen injection dose, the tumor mass was decreased significantly, and the tumor inhibition rate was obviously increased.Combination of CKI and doxorubicin could overcome the multidrug resistance in breast carcinoma MCF-7 cells. They showed an obvious synergistic effect, with an increase in the accumulation of doxorubicin and reduction of the expression of P-glycoprotein.In two animal experiments, CMI could obviously decrease the levels of PDCD4, P53, E2, survivin and Ki67 in the model rats of breast cancer.The detection of the expression level of Survivin was expected to be the predictive index of paclitaxel chemotherapy in breast cancer. For patients with high expression of Survivin, combining with CKI may improve the efficiency of chemotherapy, and even prolong PFS.In addition, the serum level of VEGF-A decreased in breast cancer patients treated with CKI during radiotherapy, and the decrease was greater than that treated with radiotherapy alone.CKI is widely used in China. It is often used in combination with operative radiotherapy, chemotherapy and neoadjuvant chemotherapy. The effect is usually better than that of CKI alone, which may be related to improving the immune function of patients.Patients with CKI treatment had higher levels of CD3 + , CD4 + , CD4 + /C8 + and NK cells than those with only postoperative radiotherapy.Yasheng et al assessed relevant randomized controlled trials from 2000 to 2017 and the results showed that patients given CKI combined with chemotherapy (therapy group) had a higher performance status improvement rate than that given only chemotherapy (control group). In the analysis adverse drug reactions (ADRs), CKI combined with chemotherapy was indicated to significantly reduce the rate of liver dysfunction, kidney dysfunction, the reaction of fatigue, diarrhea, bone marrow suppression, hair loss, nausea and vomiting, oral mucositis, abnormal elevation of tumor marker CEA and CA153 and platelet decrease. 37-40 Lymphocyte transformation ratio, immunoglobulin level, 41 life quality, KPS and QOL score had also been improved.What's more, CKI had a good analgesic effect on bone and muscle pain caused by the paclitaxel chemotherapy for breast cancer patients, with their life quality improvement.But these clinical studies face many problems to be solved urgently. Ongoing challenges include the inability to bind participants to the study, because most measures are subjective, which are susceptible to suggestive biases. Patients who received it may perceive a benefit to the intervention.
## Matrine
As one of the major quinolizidine alkaloids of Sophora flavescens, MT (C 15 H 24 N 2 O) is extracted from the root of Kushen. MT has pharmacological effects such as antidiuretic, anti-asthma, anti-bacterial, anti-inflammatory and immune suppressive activities. What's more, it has therapeutic effects on cardiac arrhythmia, hepatitis and various solid tumors. For years, MT has been reported to have promising pharmaceutical efficacies for various cancer cells, such as acute myeloid leukaemia, nasopharyngeal Carcinoma, hepatocellular carcinoma, cervical cancer, colorectal cancer, bladder cancer, prostate cancer, pancreatic cancer.However, the antitumor function about the molecular mechanism of MT still remains unclear. Like other extracts of botanicals, it has some moderate side effects, such as toxicity to the central nervous, coupled with system low water solubility, bioactivity and bioavailability. Opening of the D-ring and breaking of the amide bond can destroy the anti-proliferative activities, which mean amide bond may be required for the antitumor activities of MT. Annexin A2 has been identified as a directbinding target of MT in cancer cells using the photo-affinity labeling approach.Autophagy might cooperate, aggravate or antagonize apoptosis.MT has been shown to promote autophagic cell death and induce autophagic in cancer cells in recent years. Autophagy was a protective cellular response to MT treatment; 57 therefore, inhibition of matrine-induced autophagy by treating the cancer cells with chloroquine enhanced the cytotoxicity. The molecular mechanism of autophagy induced by the effect of MT in human breast cancer Bcap-37 cells may be via down-regulating PI3K/ Akt/mTOR signaling pathway.Earlier studies suggest that the mechanism of inhibiting growth and inducing apoptosis of MCF-7/ADR cell growth is related to the blocking of S period cells.Bcl-2, a novel anti-cancer target, is commonly associated with breast cancer. When treated with MT, three types of breast cancer cells (MCF-7 cells, BT-474 cells, MDA-MB-231) were all induced to death. The obvious effects on inhibiting the growth of MCF-7 cells and promoting the apoptosis might be achieved by up-regulating the expression of Fas protein, inhibiting telomerase activity, and downregulating the expression of VEGF protein to inhibit the tumor vascular formation. It displayed synergistic effects with trichostatin A, celecoxib and rosiglitazone against VEGF excretions and the cell proliferation via EGF/ VEGF-VEGFR1-Akt-NF-κB signaling pathway. Other studies revealed that the suppression of the cancer cell proliferation and invasion by MT is associated with inhibition of Akt signaling: downstream targets such as Bcl-2/Bax, NF-κB, p-65, MMP-9 and MMP-2, and upstream targets such as EGF, VEGF and VEGFR1 in the breast cancer cells.Compared to the controls, tumors treated with MT had a greater apoptosis index and a less microvessel density by downregulating the expression of Bcl-2/Bax, p53, VEGF and VEGFR-2, increasing the activation of caspase-3 and caspase-9.In another experimental study, MT inhibited the phosphorylation of AKT level to regulate the downstream apoptosis factors of PI3K/AKT signal pathway; the expressions of MRP1, P-GP, Bcl-2, p-AKT were down-regulated and PTEN level was up-regulated.Conversely, in one study, Bax is overexpressed while Bcl-2 expression is decreased: MT inhibited the cell growth and induced apoptosis through up-regulation of Bax and down-regulation of Bcl-2 expression.Animal research revealed that MT inhibited the expression of VEGF and downregulated the Wnt/βcatenin signaling pathway in vivo. The 4T1-tumor-bearing mice treated with MT (100 mg/kg) exhibited a significant reduction in tumor volume and weight compared with the control group, and the expression levels of cytochrome c, cleaved caspase-9 and caspase-3 increased significantly.The treatment with MT resulted in the degradation of the inhibitor of κB (IκB) kinase β (IKKβ), which is one of the activation of IkB kinases (IKKs) and results in the phosphorylation of IkB.The ER unfolded protein response (UPR) is a cellular signaling pathway in the ER, which is a survival mechanism utilized by tumors to buffer proteotoxic stress; overload ER stress is considered to be a key regulator of tumor cell death pathways in response to anticancer drugs. Endoplasmic reticulum stress is mainly involved in the cell apoptosis induced by MT. It could inhibit mitochondrial activity by promoting the release of cytochrome C and enhancing caspase-3 function to induce ER stress-mediated apoptosis. At the same time, MT inhibits the expression of hexokinase II and glycometabolism to down-regulate energy metabolism in MCF-7 cells.The apoptosis induction effect may be caused by reducing mitochondrial transmembrane potential and leaving the MMP channel open.What's more, it may regulate the proliferation and apoptosis of 4T1 cells by inhibiting the expression of ANXA3 mRNA and downregulating the expression of ANXA3 protein.Recent evidence has indicated that miRNAs can act as mediators of the therapeutic efficacy of botanicals. MiR-21 is overexpressed in the vast majority of cancers such as colorectal cancer, bladder cancer, lung cancer, prostate cancer and breast cancer. Inactivation of PTEN promoted the accumulation of PIP 3 , resulting in phosphorylation of Akt, and this process is related to the high expression level of miR-21. An experimental study showed that MT upregulated PTEN and down-regulated PTEN p21 /WAF1/CIP1 and p27 /KIP1 , which resulted in the inhibition of cell cycle progression and proliferation. It may inhibit breast cancer cell growth by interfering miR-21/PTEN/Akt signaling pathway.Upregulation of ZIC1 gene expression or MT alone can significantly inhibit the proliferation, adhesion and migration of MDA-MB-231 cells, and induce the apoptosis of human breast cancer cells, while the antitumor effect of ZIC1 gene combined with MT is more obvious.Breast cancer patients receiving chemotherapy like anthracyclines and taxanes bear severe cumulative toxicities and tolerability disturbance. MT and chemotherapeutic drugs had synergistic effects and it reversed multidrug resistance in breast cancer, which involved regulating the downstream apoptosis factors (Bcl-2, p65, Bax, PARP, GSK-3β, Caspase-3), inhibiting the activation of PI3K/ AKT pathway, and reducing P-glycoprotein expression in breast carcinoma MCF-7 cells.MT injection has a protective effect on liver function of breast cancer patients undergoing chemotherapy after operation.Other types of pharmaceutical preparations of Kushen such as Sophora flavescens alkaloid gels also can restrain cancer cell proliferation, inhibit metastasis and induce cellular apoptosis.Further clinical studies with different concentrations and types of MT are warranted to obtain a better picture of antitumor activities of it.
## Oxymatrine
Oxymatrine (molecular formula: C 15 H 24 N 2 O 2 ) is another major alkaloid component extracted from the roots of Kushen, which belongs to quinolizidine alkaloid with relative molecular mass of 264.4. OMT has been shown a wide range of pharmacological effects, such as antitumor, antihypertension, anti-inflammation, heart strengthening, antipyresis and analgesia, anti-allergy, anti-arrhythmia, asthma relieving, sedation and hypnosis, inhibition of myocardial ischemia and infarction, antivirus and sterilization. Many in vivo and in vitro experiments suggest that OMT exhibits antineoplastic properties in several cancers via different signaling pathways, including suppression of proliferation and promotion of apoptosis.Previous study had confirmed that the antineoplastic effect of OMT was related to the induction of apoptosis of cancer cells by blocking the process of cell cycle, initiating cell self-regulation procedures.A recent study revealed that it involved the inhibition of epithelial-mesenchymal transformation of cancer cells.Similarly, one study demonstrated that it induced apoptosis in breast cancer cells by down-regulating apoptosis-related protein, including poly (ADP-ribose) polymerase (PARP) and cleaved Caspase-3, cleaved Caspase-9.Like other antiapoptotic proteins (Bcl-XL, Mcl-1, Bcl-W), BCL-2 regulates cell differentiation and apoptosis through the intrinsic (or mitochondrial) pathway. Highest concentration of OMT (100 μg/mL) regulated Bax mRNA abundance by 169% at 72 hrs, and regulated Bax mRNA abundance by 169%, resulting in promoting apoptosis in MCF-7 cells."sidepopulation (SP)" cells, which are used for identification and analysis of the CSCs, show the characteristics of the cancer stem-like cells and represent a population with a capacity of limited maturation and self-renewal. Ying Zhang et al explored the effect of OMT on the SP of MCF-7 cells and this study demonstrated that OMT treatment increased the expression of phosphorylated β-catenin, which is related to the inactivation of Wnt signaling pathway, with the decrease of total β-catenin and deactivation of Wnt signaling pathway. It decreased the viability of MCF-7 cells in a time-and dose-dependent manner. What's more, compared with cisplatin, OMT showed a lower inhibitory effect on non-SP cells and a higher inhibitory effect on SP cells.On the other hand, OM could clearly inhibit the proliferation and increases apoptosis of human MCF-7 cells, and the mechanism of this effect probably related to the activity of Wnt/β-catenin signaling pathway,with the expression of downstream cytokines c-myc and cyclinD1 decreasing.
OM had direct anti-tumor effects and partly reversed MDR. The mechanism of reversing MDR was associated with the behavior of OMT down-regulating P-gP expression on the cell membrane.The nonflavonoid compounds consist of alkylxanthones, triterpene glycosides, quinones, essential oils and fatty acids. Recent studies illustrated the antineoplastic effects of novel Sophora flavescens alkaloids, including matrine derivative WM622, WM-127, matrine derivatives containing benzo-α-pyrone structure (6aS, 10S, 11aR, 11bR, 11cS)-10-methylaminododecahydro-3a,7a-diaza-benzo (de) (MASM), which beared therapeutic potentials for lung and hepatocellular carcinoma.Kur and KS-Fs were able to enhance the effects of Taxol and adriamycin on tumor proliferation in vitro. Either combination of low dose of Taxol (5 mg/kg) or high dose of taxol (5 mg/kg), KS-Fs (200 mg/kg/day) were shown to enhance the effect of Taxol on tumor growth in vivo.Other matrine derivatives from sophoridine have also been developed. Many novel N-substituted sophoridinic acid derivatives were synthesized, and their cytotoxicity and anticancer activities were evaluated. An introduction of an aliphatic acyl on the nitrogen might enhance the antitumor activity. Among the synthesized derivatives, COCH 2 Br bearing bromoacetyl side-chain inhibited the activity of DNA topo I to suspend the cell cycle at S-phase, which lead the breast cancer tumor cells to apoptosis.A study revealed this (2S)-7,2ʹ,4ʹ-trihydroxy-5methoxy-8-dimethylallyl flavanone can inhibit proliferation, migration, adhesion, and tube formation of endothelial cells.Because of its anti-angiogenesis activity and antiproliferative properties this new flavonoid can be a good anticancer candidate. New sophoridinol derivatives and N-substituted sophoridinic acid/ester such as sophoridinol 7i (R1:9-methylanthracene, R2:CH2OH) arrest the cell cycle at the G0/G1 phase and showed an antiproliferative activity with an IC50 of 3.1 μM.It is well documented that several quinolizidine alkaloids have potential antitumor activity. A series of 12-chlorobenzyl sophoridinic derivatives with novel chemical scaffolds such as compounds 5a(C7H4OCl-) and 7b (C13H8F2p2-) exhibited reasonable antiproliferative and anticancer activities; they displayed an equipotency in both adriamycin (AMD)-susceptible and resistant MCF-7 breast carcinoma cells.In addition, three novel compounds (YF3-5, YF3-7 and YF3-9) all demonstrated anti-proliferation activity, of which YF3-5 showed the strongest anti-proliferation properties against the four human cancer cell lines (BT20 breast, MCF-7 breast, A549 lung and U2OS osteosarcoma cells).Cell cycle arrest and induction of mitochondria-mediated apoptosis are common mechanisms of inhibiting cell proliferation. Recently, sophoraflavanone G (SG), a compound isolated from Sophora flavescens, induced reactive oxygen species, DNA fragmentation, nuclear condensation, showing anti-tumor and anti-inflammatory properties in MDA-MB-231. SG increased expression of cleaved caspase-3, caspase-9, caspase-8, and decreased Bcl-2 and Bcl-xL expression, which inhibited the migration and invasion of breast cancer cells.
## Conclusions and future directions
Traditional medicine, such as Ayurvedic medicine, Chinese medicine, uses unidentified chemical entities with unknown mechanisms of action. TCM has been applied in many countries and regions and has achieved certain results. From the point of view of the basic principles of evidence-based medicine and the provision of clinical scientific evidence, there are still some problems to be solved in the practice of TCM application, such as the lack of sufficient experimental data on the effectiveness and safety of TCM, not an exhaustive study on side effects of TCM. The clinical trials of TCM are short of large sample randomized controlled data and exact evaluation index system and evaluation method in this field.
This review summarizes multiple anti-cancer effects of active ingredients in Sophora flavescens. Effective extracts of medicinal plants, such as MT, OMT and their derivatives, have been extensively studied in breast cancer. There is a growing body of evidence supporting the use of Kushen and its active ingredients as effective supportive care strategies during breast cancer treatment. CKI and other phytochemicals extracted from Sophora flavescens may become potential complementary and alternative therapy for breast cancer in the future. In China, patients with breast cancer commonly use them as supportive care during cancer treatment and to manage treatment-related side effects.
However, evidence supporting the application of MT, OMT and other Sophora flavescens alkaloids in the oncology setting is limited. The specific anti-cancer mechanism of them needs further systematic studies. Some of the reported clinical studies about them were not well-controlled studies and not blinded. Although they are perceived as harmless, more studies are needed on safety rather than efficacy. What's more, little information is available on toxicities and cost effectiveness. Drugbotanical interactions, pharmacokinetic and distribution profile of active constituents have to be determined to help reach safe and efficacious concentrations in clinical settings. Evaluation and exploration of the clinical impact that is more truly generalizable must be made to avoid potentially toxic side effects or loss of activity of active ingredients. Additionally, seeking international cooperation in the research of them is also an important way to promote these compounds more widely used.
## Human and animal rights and informed consent
This article does not contain any studies with human or animal subjects performed by any of the authors.
## Abbreviations
MT, matrine; OMT, oxymatrine; CKI, Compound Kushen Injection; TCM, Traditional Chinese Medicine; KS-As, Kushen alkaloids; KS-Fs, Kushen flavonoids; DSBs, double-strand breaks; CSCs, Breast cancer stem cells; IκB, inhibitor of κB; IKKs, IkB kinases; UPR, unfolded protein response; SG, sophoraflavanone G. |
The CDK inhibitor CR8 acts as a molecular glue degrader that depletes cyclin K
[bib_ref] Crabtree GR Chemically induced proximity in biology and medicine, Stanton [/bib_ref] [bib_ref] Ebert BL The novel mechanism of lenalidomide activity, Fink [/bib_ref] [bib_ref] The myeloma drug lenalidomide promotes the cereblon-dependent destruction of ikaros proteins, Lu [/bib_ref] [bib_ref] Lenalidomide causes selective degradation of IKZF1 and IKZF3 in multiple myeloma cells, Krönke [/bib_ref] [bib_ref] A novel cereblon modulator recruits GSPT1 to the CRL4 CRBN ubiquitin ligase, Matyskiela [/bib_ref] [bib_ref] Thomä NH Structural basis of lenalidomide-induced CK1α degradation by the CRL4 CRBN..., Petzold [/bib_ref] [bib_ref] Structure of the DDB1-CRBN E3 ubiquitin ligase in complex with thalidomide, Fischer [/bib_ref] [bib_ref] Catalytic in vivo protein knockdown by small-molecule PROTACs, Bondeson [/bib_ref]
## Cr8 induces proteasomal cyck degradation
To identify small molecules that mediate protein degradation through an E3 ubiquitin ligase, we correlated drug sensitivity data for 4,518 clinical and pre-clinical drugs tested against 578 cancer cell lines [bib_ref] High-throughput identification of genotype-specific cancer vulnerabilities in mixtures of barcoded tumor cell..., Yu [/bib_ref] [bib_ref] Non-oncology drugs are a source of previously unappreciated anti-cancer activity, Corsello [/bib_ref] with the mRNA expression levels for 499 E3 ligase components [bib_ref] Next-generation characterization of the Cancer Cell Line Encyclopedia, Ghandi [/bib_ref] (Extended Data. DCAF15 gene expression correlated with indisulam and tasisulam toxicity, consistent with its known function as a degrader of the essential protein RBM39 by the CRL4 DCAF15 E3 ubiquitin ligase, thus demonstrating the potential of the approach (Extended. We sought to validate the high-scoring ligase-drug correlations by examining whether CRISPR-mediated inactivation of the identified E3 ligase component would rescue the respective drug-induced toxicity (Extended Data. These experiments confirmed that sgRNAs targeting DCAF15 confer resistance to indisulam and tasisulam. In addition, we observed a correlation between cytotoxicity of the CDK-inhibitor R-CR8 [bib_ref] CR8, a potent and selective, roscovitine-derived inhibitor of cyclin-dependent kinases, Bettayeb [/bib_ref] and mRNA expression levels of the CUL4 adaptor DDB1and Extended Data. Consistently, sgRNAs targeting DDB1 conferred resistance to R-CR8.
As DDB1-dependent cytotoxicity of R-CR8 implicated ubiquitin ligase-mediated degradation of one or more essential proteins, we performed quantitative proteome-wide mass spectrometry to evaluate protein abundance following compound treatment. Of the >8,000 quantified proteins, cycK was the only protein that consistently showed decreased abundance following R-CR8 addition, Extended Data. As expected, R-CR8 did not alter cycK mRNA levels (Extended Dataand compound-induced cycK degradation could be rescued by inhibition of the E1 ubiquitin-activating enzyme (MLN7243), cullin neddylation (MLN4924) and the proteasome (MG132). Together, these results suggest that R-CR8 triggers rapid proteasomal degradation of cycKthrough the activity of a DDB1-containing cullin-RING ubiquitin ligase.
To dissect the molecular machinery required for R-CR8 toxicity, we performed genomewide and E3 ubiquitin ligase-focused CRISPR-Cas9 resistance screens, Extended Data. SgRNAs targeting DDB1, CUL4B, RBX1, the cullin-RING activator NEDD8, and the NEDD8-activating enzyme (NAE1/UBA3) were significantly enriched in the R-CR8-resistant cell population. As all of these proteins are required for CRL activity, our results provide genetic evidence for the involvement of a functional CUL4-RBX1-DDB1 ubiquitin ligase complex in mediating R-CR8 cytotoxicity.
Thus far, all known cullin-RING ligases engage their substrates through specific substrate receptors, and DDB1 serves as an adaptor protein able to bind over 20 such receptors (also known as DDB1-CUL4-associated-factors, DCAFs) [bib_ref] Molecular architecture and assembly of the DDB1-CUL4A ubiquitin ligase machinery, Angers [/bib_ref] to recruit them to the CUL4-RBX1 ligase core. As no DCAF was identified in our viability screens, we constructed a fluorescent reporter of cycK stability (Extended Data, in which R-CR8-mediated degradation of endogenous cycK could be recapitulated with a cycK eGFP fusion protein, e, Extended Data. Using the stability reporter, in which the extent of degradation can be determined by measuring cycK eGFP levels normalised to mCherry expression, we found that Sand R-CR8 facilitated cycK eGFP degradation to the same extent (Extended Data; henceforth, CR8 refers to R-CR8). We then performed a genomewide CRISPR-Cas9 screen for genes involved in cycK reporter stability and validated the involvement of DDB1 in CR8-mediated cycK degradation, but not in compound-independent cycK degradation (Extended Data. In addition, we identified cyclin-dependent kinase 12 (CDK12), which is a known target of CR8 [bib_ref] The structure and substrate specificity of human Cdk12/Cyclin K, Bösken [/bib_ref] and whose activity depends on the interaction with cycK 20 , as a crucial component for CR8induced cycK eGFP destabilisation, Extended Data.
As neither the cycK eGFP stability reporter screen nor the CR8 resistance screen identified a substrate receptor, we performed additional CRISPR screens targeting 29 genes encoding known DCAFs or DCAF-like candidate proteins in four different cell lines. While sgRNAs targeting the previously identified components of the CUL4-RBX1-DDB1 complex consistently caused resistance to CR8, a DCAF substrate receptor could not be identified (Extended Data .
## Cr8 directs cdk12 to crl4 core component
Since none of our genetic screens highlighted a DCAF required for cycK degradation, we tested whether CR8-engaged CDK12-cycK directly binds one of the CUL4-RBX1-DDB1 ligase components in the absence of a substrate receptor. We therefore performed in vitro coimmunoprecipitation experiments using recombinantly purified proteins. The kinase domain of CDK12 (CDK12 713−1052 ) bound to cycK 1−267 did not markedly enrich DDB1 over the bead binding control in the absence of CR8, whereas equimolar amounts of the compound led to stoichiometric complex formation. DDB1 β-propeller domains A (BPA) and C (BPC) [bib_ref] Molecular architecture and assembly of the DDB1-CUL4A ubiquitin ligase machinery, Angers [/bib_ref] , which are otherwise involved in DCAF binding, were sufficient for drug-induced CDK12-cycK recruitment. DDB1 β-propeller B (BPB), which binds CUL4 and is not involved in DCAF binding, was dispensable for the interaction. In vitro ubiquitination assays confirmed that the CUL4A-RBX1-DDB1 ligase core alone is sufficient to drive robust cycK ubiquitination. Quantification of the interaction showed that CR8 stimulated binding between CDK12-cycK and DDB1 in the range of 100-500 nM depending on the experimental setup. While weak CDK12-cycK-DDB1 interaction was still detectable in the absence of the compound in vitro, CR8 strengthened complex formation 500-to 1000-fold as estimated by isothermal titration calorimetry (ITC) (Extended Data. Thus, our data indicate that CR8-engaged CDK12-cycK is recruited to the CUL4-RBX1-DDB1 ligase core through DDB1, and the compound tightens the complex sufficiently to drive CR8-induced cycK degradation in the absence of a canonical DCAF substrate receptor.
We then crystallised CDK12 713−1052 -cycK 1−267 bound to CR8 and DDB1 ΔBPB and determined the 3.5 Å resolution structure of this complex, Extended Data . In the structure, CDK12 forms extensive protein-protein interactions (~2100 Å 2 ) with DDB1. CR8 binds the active site of CDK12 and bridges the CDK12-DDB1 interface, while cycK binds CDK12 on the opposite site and does not contact DDB1. The N-and C-lobes of CDK12 are proximal to DDB1 residues located in a loop of the BPA domain (amino acid (aa) 111-114), BPC-helix 2 (aa 986-990), and a loop in the C-terminal domain (aa 1078-1081) that are otherwise involved in DCAF binding (Extended Data . In addition, the C-terminal extension of CDK12 binds the cleft between the DDB1 domains BPA and BPC, a hallmark binding site of DDB1-DCAF interactions (Extended Data . The density for this region could only be tentatively assigned, likely due to the presence of multiple conformations, but the CDK12 C-terminal tail clearly engages with DDB1 and assumes a conformation different from those seen in isolated CDK12-cycK structures (Extended Data [bib_ref] The structure and substrate specificity of human Cdk12/Cyclin K, Bösken [/bib_ref] [bib_ref] Morin GB & Bullock AN Structures of the CDK12/ CycK complex with..., Dixon-Clarke [/bib_ref]. Structure-guided mutational analyses combined with time-resolved fluorescence resonance energy transfer (TR-FRET) assays were used to assess the contribution of these interactions to CR8-dependent CDK12-DDB1 complex formation, Extended Data . Taken together, our data demonstrates that CDK12 assumes the role of a glue-induced substrate receptor and places cycK in a position typically occupied by CRL4 substrates. This renders CDK12-cycK binding to DDB1 mutually exclusive with that of DCAFs and provides a structural framework for why a canonical substrate receptor is dispensable for cycK ubiquitination.
## Cdk12-ddb1 interface imparts selectivity
CR8 is a pleiotropic CDK inhibitor reported to bind CDK1/2/3/5/7/9/12 [bib_ref] CR8, a potent and selective, roscovitine-derived inhibitor of cyclin-dependent kinases, Bettayeb [/bib_ref] [bib_ref] The structure and substrate specificity of human Cdk12/Cyclin K, Bösken [/bib_ref] , yet in cells we observed selective cycK destabilization in the presence of the drug. As cycK is reported to associate with CDK9, CDK12, and CDK13 [bib_ref] The structure and substrate specificity of human Cdk12/Cyclin K, Bösken [/bib_ref] , we tested whether the other cycK-dependent kinases are also recruited to DDB1. The closely related CDK13 (90.8 % sequence identity), but not the more divergent CDK9 (45.5 % sequence identity) (Extended Data , was recruited to DDB1 in the presence of CR8, albeit with a lower binding affinity (Extended Data . Analogously, less productive in vitro cycK ubiquitination was observed for CDK13 compared to CDK12 (Extended Data . The key difference between CDK9 and CDK12/13 primary sequence lies in the C-terminal extension (Extended Data , b), which in our structure nestles up against DDB1 BPA and BPC propellers, Extended Data . Mutations in, or truncation of, the CDK12 C-terminal extension (Extended Data abolished basal binding between CDK12 and DDB1, whereas complex formation could still be facilitated by CR8 to a varying extent (Extended Data , i). Hence, our data show that the pan-selective CDK inhibitor CR8 induces specific protein-protein interactions between CDK12/13 and DDB1 and suggest that the Cterminal extension, while contributing to binding, is not essential for drug-dependent kinase recruitment.
## Cr8 phenylpyridine confers glue activity
CR8 occupies the ATP binding pocket of CDK12 and forms discrete contacts with residues in the BPC domain of DDB1 (~150 Å 2 ) through its hydrophobic phenylpyridine ring system . Mutation of DDB1 residues Ile909, Arg928, and Arg947 each diminished druginduced recruitment of the kinase (Extended Data , highlighting the contribution of the phenylpyridine moiety to complex formation. To evaluate the structure-activity relationship underlying the gain-of-function activity of CR8, we probed other CDK inhibitors for their ability to drive complex formation between DDB1 and CDK12. DRF053 [bib_ref] Roscovitine-Derived, Dual-Specificity Inhibitors of Cyclin-Dependent Kinases and Casein Kinases 1, Oumata [/bib_ref] , a CR8-related inhibitor that carries a differently linked phenylpyridine ring system , induced binding with two-fold lower affinity than CR8 (Extended Data . Roscovitine 23 , the parent compound of CR8 that lacks the 2-pyridyl substituent but retains the phenyl ring proximal to Arg928 , also facilitated complex formation, albeit with a three-fold lower apparent affinity (Extended Data . The affinity rankorder observed in our TR-FRET assay correlated with the degree of cycK ubiquitination in vitro, in which DRF053 and roscovitine showed less processive ubiquitination . As neither DRF053 nor roscovitine induced degradation of the cycK eGFP reporter in cells , our results demonstrate that the presence and correct orientation of the 2-pyridyl on the surface of CDK12 confer the gain-of-function activity of CR8 leading to cycK degradation.
To probe whether any ligand could in principle drive the interaction of CDK12 with DDB1, we tested the endogenous CDK nucleotide cofactor ATP for its ability to promote complex formation. ATP neither facilitated nor abrogated the interaction over basal binding observed in the presence of DMSO (Extended Data, suggesting that although the nucleotidebound conformation of CDK12 seems incompatible with approaching DDB1 (Extended Data, its C-terminal extension is free to adopt multiple conformations [bib_ref] Morin GB & Bullock AN Structures of the CDK12/ CycK complex with..., Dixon-Clarke [/bib_ref]. THZ531 24 , a bulky covalent CDK12/13 inhibitor predicted to clash with DDB1 (Extended Data, locks the CDK12 C-terminal extension in a conformation incompatible with DDB1 recruitment (Extended Data. Consistently, THZ531 further decreased the TR-FRET signal and diminished cycK ubiquitination in vitro below DMSO control levels , Extended Data, f) [bib_ref] Covalent targeting of remote cysteine residues to develop CDK12 and CDK13 inhibitors, Zhang [/bib_ref]. Flavopiridol 25 , a natural product-derived inhibitor structurally distinct from CR8 , also stimulated the binding of CDK12-cycK to DDB1 (Extended Data . While flavopiridol gave rise to moderate cycK ubiquitination in vitro , it failed to degrade cycK in cells . Our results thus show that the DDB1-compound interactions display significant plasticity and that structurally diverse surface-exposed moieties in CR8, DRF053, roscovitine and flavopiridol can facilitate CDK12-cycK recruitment. Small differences in their ability to stabilise the DDB1-CDK12 complex translate, in an almost binary fashion, into cellular degradation of cycK or lack thereof. This behaviour is reminiscent of CRL4 CRBN and thalidomide analogues [bib_ref] Thomä NH Structural basis of lenalidomide-induced CK1α degradation by the CRL4 CRBN..., Petzold [/bib_ref] [bib_ref] Defining the human C2H2 zinc finger degrome targeted by thalidomide analogs through..., Sievers [/bib_ref] , where an apparent affinity threshold must be overcome to drive drug-induced target degradation.
## Cyck degradation adds to cr8 toxicity
Finally, to delineate the contribution of CRL4-mediated cycK degradation to CR8 cytotoxicity over non-degradative CDK inhibition, we compared compound toxicity in wildtype HEK293T Cas9 cells to cells that were pre-treated with MLN4924 (NEDD8-activating enzyme inhibitor), genetically-depleted for DDB1, or subject to DCAF overexpression. Global inhibition of CRL activity by MLN4924 had only minor effects on cell viability (Extended Data , but resulted in decreased sensitivity to CR8 (Extended Data , showing that CRL neddylation significantly contributes to CR8 toxicity. Overexpression of the substrate receptor CRBN also affected sensitivity to CR8 and decreased cycK degradation (Extended Data , presumably by reducing the free pool of DDB1. As expected, CR8-induced endogenous cycK degradation was dependent on DDB1 and, consistently, we found that cytotoxicity of CR8, but not that of the other CDK inhibitors, was ten-fold lower in cells depleted for DDB1 . Together, the data demonstrate that the CRL4-dependent gain-of-function glue degrader activity of CR8 strongly contributes to its cellular potency and provides an additional layer of ortholog-specific CDK inactivation through cycK degradation.
Kinase inhibitors have long been suspected to have a degradation component to their mode of action [bib_ref] chemical biology view of bioactive small molecules and a binder-based approach to..., Schreiber [/bib_ref] , and our work provides the first characterization and structural dissection of how a kinase inhibitor scaffold acquires degrader properties. Molecular glue degraders have thus far only been shown to engage substrate-recruiting E3 ligase modules. CDK12 is not a constitutive E3 ligase component, but rather serves as a drug-induced substrate receptor, linking DDB1 to the ubiquitination target. CR8 thus bypasses the requirement for a canonical DCAF and instead hijacks the essential adaptor protein DDB1. Although cycK is the primary ubiquitination target, CDK12 may become subject to autoubiquitination upon prolonged compound exposure similar to canonical DCAFs [bib_ref] Identification of a primary target of thalidomide teratogenicity, Ito [/bib_ref] [bib_ref] The molecular basis of CRL4DDB2/CSAubiquitin ligase architecture, targeting, and activation, Fischer [/bib_ref].
While previously reported molecular glue degraders engage a ligandable pocket on the ligase to recruit target proteins, CR8 instead binds the active site of CDK12 and does not rely on an independent ligand binding site on DDB1 (Extended Data. This suggests that the repertoire of targets and ubiquitin ligases accessible to targeted protein degradation can be expanded by target-binding small molecules that induce de novo contacts with a ligase or strengthen existing weak protein-protein interactions. Kinase inhibitors in particular often show poor selectivity and small molecule-induced kinase inactivation that leverages complementary protein-protein interfaces offers a path towards improved drug selectivity, which might, for example, facilitate selective inactivation of CDK12, an emerging therapeutic target [bib_ref] CDK12 Inhibition Reverses De Novo and Acquired PARP Inhibitor Resistance in BRCA..., Johnson [/bib_ref].
The gain-of-function glue degrader activity of CR8 is attributed to a 2-pyridyl moiety exposed on the kinase surface. Surface-exposed single residue mutations have been shown to promote the formation of higher-order protein complexes, as the haemoglobin Glu to Val mutation, for example, induces polymerization in sickle cell anaemia [bib_ref] Edelstein SJ Three-dimensional reconstruction of the 14-filament fibers of hemoglobin S, Dykes [/bib_ref]. Accordingly, single residue mutations designed to increase surface hydrophobicity give rise to ordered protein assemblies [bib_ref] Proteins evolve on the edge of supramolecular self-assembly, Garcia-Seisdedos [/bib_ref]. Bound compounds, such as enzyme inhibitors, can in principle mimic such amino acid changes with dramatic effects on the protein interaction landscape, suggesting that compound-induced protein-protein interactions may be more common than previously recognised. Taken together, our results suggest that modifications of surface-exposed regions in target-bound small molecules offer a rational strategy to develop molecular glue degraders for a given protein target.
# Methods
## Mammalian cell culture
The human HEK293T cell lines were provided by the Genetic Perturbation Platform, Broad Institute and K562 Cas9 , THP1 Cas9 , P31FUJ Cas9 cell lines were provided by Zuzana Tothova
## Antibodies
The following antibodies were used in this study: anti-cycK (Bethyl Laboratories, A301-939A for full length cycK), anti-cycK (abcam, ab251652, for cycK 1−267 ), anti-beta-actin (Cell Signaling, #3700), anti-CRBN (Sigma prestige, HPA045910), anti-mouse 800CW (LI-COR Biosciences, 926-32211), anti-rabbit 680LT (LI-COR Biosciences, 925-68021), antirabbit IgG antibodies (abcam, ab6721).
## Reporter vectors
The following reporter were used in this study:
## Oligos
List of all oligonucleotides used in this study can be found in Supplementary Oligo .
## Bioinformatic screen
We computed Pearson correlations of viability of PRISM repurposing compounds in 8 doses and 578 cell lines 4 with gene expression (GE) and copy number variation (CN) of all detectable protein-coding genes of matched cell lines from The Cancer Cell Line Encyclopedia (CCLE) [bib_ref] Next-generation characterization of the Cancer Cell Line Encyclopedia, Ghandi [/bib_ref]. Z score was computed for each pair of compounds, dose viability, and genomic feature (GE or CN) across all cell lines. For each compound-genomic feature pair, the most extreme correlations are ranked from negative to positive. To focus on novel compound-gene relationships, we restricted genes to a curated list of 499 E3 ligase components and compounds that are not one of "EGFR inhibitor", "RAF inhibitor" or "MDM inhibitor" based on PRISM repurposing annotation [bib_ref] Next-generation characterization of the Cancer Cell Line Encyclopedia, Ghandi [/bib_ref]. Hit compounds were selected if either the Z score was less than −6 or ranked in the top 15 with Z score less than −4. The resulting list of 158 E3 gene-compound pairs was further curated and shortened manually to 96 E3 gene-compound pairs, which included 95 unique E3 ligases and 85 unique compounds. . Suspension cells were directly subjected to analysis. 10 μL of cell suspension was subjected to the flow analysis with FACSCanto equipped with High Throughput Sampler (BD Bioscience). The percentage of sgRNA transduced cells in the drug treatment wells was normalised to the DMSO control. Wells with fluorescent drug and samples with less than 120 viable cells events or less than 6% fluorescent cells were removed from analysis. All E3-drug pairs were ranked based on the number of experimental conditions (cell line and drug dose) with more than 50% of sgRNA transduced cells in drug treatment wells in comparison to corresponding DMSO control.
## Validation of ddb1 resistance phenotype
For validation experiments, virus was produced in a 6-well plate format, as described above with the following adjustments: 2.5 × 10 5 HEK293T cells/well in 2 mL DMEM medium, 3 μL/well of TransIT-LT1, 15 μL/well of OPTI-MEM, 500 ng/well of the desired plasmid, 500 ng/well psPAX2, and 50 ng/well pVSV-G in 32.5 μL/well OPTI-MEM. After collecting the virus, 10 × 10 3 HEK293T Cas9 cells in 100 μL DMEM medium were transduced with 10 μL of virus supernatant. The transduced HEK293T Cas9 cells were then mixed with untransduced control cells at a 1:9 ratio. Nine days after sgRNA transduction, cells were treated for 3 days with DMSO or 1 μM CR8 and analysed by flow cytometry to determine the percentage of BFP + cells. sgRNAs targeting DDB1 provide partial depletion of DDB1 (50% DDB1 alleles modified, reducing DDB1 levels by roughly 50%), which suggests selection towards heterozygous or hypomorphic clones.
Whole proteome quantification using tandem mass tag mass spectrometry 10 × 10 6 Molt-4 cells were treated with DMSO (triplicate) or 1 μM CR8 (single replicate) for 1 or 5 hours and later were harvested by centrifugation. Samples were processed, measured and analysed as described before [bib_ref] Thalidomide promotes degradation of SALL4, a transcription factor implicated in Duane radial..., Donovan [/bib_ref]. Data are available in the PRIDE repository (PXD016187 and PXD016188).
## Quantitative pcr
HEK293T Cas9 cells were treated with DMSO or 1 μM CR8 for 2 hours, collected by centrifugation, washed with PBS, and snap-frozen at −80°C. mRNA was isolated using a QIAGEN RNA kit (Qiagen, 74106). For cDNA synthesis, total RNA was reverse transcribed using a High-Capacity cDNA Reverse Transcription Kit (Thermo Fischer) before qPCR analysis with TaqMan Fast Advanced Master Mix (ThermoFisher Scientific, 4444557) for CCNK (TaqMan, Hs00171095_m1, Life Technologies) and GAPDH (TaqMan, Hs02758991_g1). Reactions were run and analysed on a CFX96 Real Time system (Bio-Rad).
## Immunoblots for whole protein lysate
Cells were washed with phosphate buffered saline (PBS) and lysed (150 mM NaCl, 50 mM Tris (pH 7.5), 1% NP-40, 1% glycerol, 1x Halt Cocktail protease and phosphatase inhibitors) for 20 minutes on ice. The insoluble fraction was removed by centrifugation, protein concentration was quantified using a BCA protein assay kit (Pierce), and equal amount of lysate was run on SDS-PAGE 4-12% Bis-Tris Protein Gels (NuPAGE, Thermo Fisher) and subsequently transferred to nitrocellulose membrane with Trans-Blot Turbo System (BIO-RAD). Membranes were blocked in Odyssey Blocking Buffer/PBS (LI-COR Biosciences) and incubated with primary antibodies overnight at 4°C. The membranes were then washed in TBS-T, incubated for 1 hour with secondary IRDye-conjugated antibodies (LI-COR Biosciences), and washed three times in TBS-T for 5 minutes prior to Near-Infrared Western blot detection on an Odyssey Imaging System (LI-COR Biosciences).
## Cyck stability reporter analysis
HEK293T Cas9 expressing the cycK eGFP degradation reporter were transduced with experimental sgRNAs. Nine days after infection the cells were dosed for 2 hours with DMSO or 1 μM CR8 and fluorescent signal quantified by flow cytometry (CytoFLEX, Beckman or LSR Fortessa flow cytometer BD Biosciences). Using FlowJo (flow cytometry analysis software, BD), the geometric mean of eGFP and mCherry fluorescent signal for round and mCherry positive cells was calculated. The ratio of eGFP to mCherry was normalised to the average of three DMSO-treated controls. [bib_ref] Hochberg Y Controlling the False Discovery Rate: A Practical and Powerful Approach..., Benjamini [/bib_ref].
## Bison crispr -cr8 resistance screen
The BISON CRISPR library targets 713 E1, E2, and E3 ubiquitin ligases, deubiquitinases, and control genes and contains 2,852 guide RNAs. It was cloned into the pXPR003 as previously described [bib_ref] Optimized sgRNA design to maximize activity and minimize off-target effects of CRISPR-Cas9, Doench [/bib_ref] by the Broad Institute Genome Perturbation Platform (GPP). The virus for the library was produced in a T-175 flask format, as described above with the following adjustments: 1.8 × 10 7 HEK293T cells in 25 mL complete DMEM medium, 244 μL of TransIT-LT1, 5 mL of OPTI-MEM, 32 μg of library, 40 μg psPAX2, and 4 μg pVSV-G in 1 mL OPTI-MEM. 10% (v/v) of BISON CRISPR library was added to 6 × 10 6 HEK293T Cas9 cells in triplicates and transduced. Samples (n=3) were processed as describe above for the genome wide resistance screen.
## Genome wide crispr -cyck stability reporter screen
A single clone of cycK eGFP HEK293T Cas9 was transduced with the genome wide Brunello library as described above with the following modification: 4.5 × 10 8 cycK eGFP HEK293T Cas9 cells in 225 mL of medium. Nine days later cells were treated with CR8 (n=3) or DMSO (n=3) for at least 2 hours and the cycK stable population was separated using fluorescence activated cell sorting. Four populations were collected (top 5%, top 5-15%, lowest 5-15% and lowest 5%) based on the cycK eGFP to mCherry mean fluorescent intensity (MFI) ratio on MA900 Cell Sorter (Sony). Sorted cells were harvested by centrifugation and subjected to direct lysis as described above. The screen was analysed as described below by comparing stable populations (top 5% eGFP/mCherry expression) to unstable populations (lowest 15% eGFP/mCherry expression). Hits enriched in cycK stable population with FDR < 0.05 are labelled on the plot.
## Pooled crispr screen data analysis
The data analysis pipeline comprised the following steps: (1) Each sample was normalised to the total number of reads.
(2) For each guide, the ratio of reads in the stable vs. unstable sorted gate was calculated, and sgRNAs were ranked. (3) The ranks for each guide were summed for all replicates. (3) The gene rank was determined as the median rank of the four guides targeting it. (4) P-values were calculated by simulating a distribution with guide RNAs that had randomly assigned ranks over 100 iterations. R scripts can be found in the Supplementary Information.
## Dcaf arrayed screen
An arrayed DCAF library (targeting DCAFs substrate receptors, DCAF-like and control genes) was constructed as described above with the appropriate oligos (Supplementary Oligo . K562 Cas9 , P31FUJ Cas9 , THP1 Cas9 and MM1S Cas9 were individually transduced and treated with DMSO or 1 μM CR8 (K562 Cas9 , P31FUJ Cas9 , THP1 Cas9 ) or 0.1 μM CR8 (MM1S Cas9 ). The analysis was performed as described above for validation of DDB1 resistance phenotype.
## Protein purification
Human wild-type and mutant versions of DDB1 (Uniprot entry Q16531), CDK12 (Q9NYV4, K965R) and CCNK (O75909) were subcloned into pAC-derived vectors [bib_ref] A set of baculovirus transfer vectors for screening of affinity tags and..., Abdulrahman [/bib_ref]
## Co-immunoprecipitation assay
The purified His 6 -CDK12/StrepII-cycK complex was mixed with equimolar concentrations of full-length His [bib_ref] CR8, a potent and selective, roscovitine-derived inhibitor of cyclin-dependent kinases, Bettayeb [/bib_ref] Plates were incubated at 19°C and crystals appeared 5-13 days after setup. Crystals were flash cooled in liquid nitrogen in reservoir solution supplemented with 25% (v/v) glycerol as a cryoprotectant prior to data collection. Diffraction data were collected at the Swiss Light Source (beamline PXI) with an Eiger 16M detector (Dectris) at a wavelength of 1 Å and a crystal cooled to 100 K. Data were processed with DIALS, scaled with AIMLESS supported by other programs of the CCP4 suite [bib_ref] Overview of the CCP4 suite and current developments, Winn [/bib_ref] , and converted to structure factor amplitudes with STARANISO 42 , applying a locally weighted CC 1/2 = 0.3 resolution cutoff.
## Structure determination and model building
The DDB1 ΔBPB -R-CR8-CDK12 713−1052 -cycK 1−267 complex formed crystals belonging space group P3 1 21, with three complexes in the crystallographic asymmetric unit (ASU). Their structure was determined using molecular replacement (MR) in PHASER 43 with a search model derived from PDB entry 6H0F for DDB1 ΔBPB , and PDB entry 4NST for CDK12-cycK. The initial model was improved by iterative cycles of building with COOT 44 , and refinement using phenix.refine 45 or autoBUSTER, with ligand restraints generated using eLBOW through phenix.ready_set 47 . The final model was produced by refinement with autoBUSTER. Analysis with MOLPROBITY 48 . indicates that 93.9% of the residues in final model are in favourable regions of the Ramachandran plot, with 0.6% outliers. Data processing and refinement statistics are in Extended Data . Interface analysis was performed using PISA [bib_ref] Henrick K Inference of Macromolecular Assemblies from Crystalline State, Krissinel [/bib_ref].
## Biotinylation of ddb1
Purified full-length StrepII-Avi-DDB1 was biotinylated in vitro at a concentration of 8 μM
## Time-resolved fluorescence resonance energy transfer (tr-fret)
Increasing concentrations of Alexa488-SpyCatcher-labelled 26 His 6 -Spy-cycK/His 6 -CDK12 ( Alexa488 cycK-CDK12) were added to a mixture of biotinylated DDB1 ( biotin DDB1) at 50 nM, terbium-coupled streptavidin at 4 nM (Invitrogen) and kinase inhibitors at 10 μM (final concentrations) in 384-well microplates (Greiner, 784076) in a buffer containing 50 mM Tris (pH 7.5), 150 mM NaCl, 0.1% pluronic acid and 0.5% DMSO (see also figure legends). CR8 titrations were carried out by adding increasing concentrations CR8 (0-25 μM) into premixed 500 μM Alexa488 cycK-CDK12, 50 nM biotin DDB1, and 4 nM terbium-coupled streptavidin. Before TR-FRET measurements, reactions were incubated for 15 minutes at room temperature. After excitation of terbium (Tb) fluorescence at 337 nm, emissions at 490 nm (Tb) and 520 nm (Alexa488) were measured with a 70 μs delay to reduce background fluorescence and the reactions were followed by recording 60 data points of each well over 1 hours using a PHERAstar FS microplate reader (BMG Labtech). The TR-FRET signal of each data point was extracted by calculating the 520:490 nm ratio. Data were analysed with Prism 7 (GraphPad) assuming equimolar binding of biotin DDB1 to Alexa488 cycK-CDK12 using the equations described previously 8 .
Counter-titrations with unlabelled proteins were carried out by mixing 500 μM Alexa488 cycK-CDK12 with 50 nM biotin DDB1 in the presence of 4 nM terbium-coupled streptavidin and 1 μM compound for DDB1 titrations or 12.5 μM compound for CDK12 titrations. After incubation for 15 minutes at room temperature, increasing amounts of unlabelled cycK-CDK12 or DDB1 (0-10 μM) were added to the preassembled Alexa488 cycK-CDK12/ biotin DDB1 complexes in a 1:1 volume ratio and incubated for 15 minutes at room temperature. TR-FRET data were acquired as described above. The 520/490 nm ratios were plotted to calculate the half maximal inhibitory concentrations (IC 50 ) assuming a single binding site using Prism 7 (GraphPad). IC 50 values were converted to the respective K i values as described previously [bib_ref] a web-based tool for converting IC50 to Ki values for inhibitors of..., Cer [/bib_ref]. Three technical replicates were carried out per experiment.
## Ddb1-cul4-rbx1 reconstitution and in vitro cul4 neddylation
In vitro CRL4 reconstitution and CUL4 neddylation were performed as described [bib_ref] Thomä NH Structural basis of lenalidomide-induced CK1α degradation by the CRL4 CRBN..., Petzold [/bib_ref]. His 6 -CUL4A/His 6 -RBX1 at 3.5 μM was incubated with His 6 -DDB1 at 3 μM in a reaction mixture containing 3.8 μM NEDD8, 50 nM NAE1/UBA3 (E1), 30 nM UBC12 (E2), 1 mM ATP, 50 mM Tris (pH 7.5), 100 mM NaCl, 2.5 mM MgCl 2 , 0.5 mM DTT and 5% (v/v) glycerol for 1.5 hours at room temperature. Neddylated and gel filtration-purified DDB1-CUL4-RBX1 ( N8 DDB1-CUL4-RBX1) was concentrated to 7.6 μM, flash frozen and stored at −80°C.
## In-vitro ubiquitination assays
In rpm. An initial injection of the ligand (4 μL) was made and discarded during data analysis. For probing DDB1-CDK12-cycK complex formation, DDB1 ΔBPB (100 μM, in the syringe) was titrated into the cell containing CDK12-cycK (10 μM) or CDK12-cycK (10 μM) pre-incubated with CR8 (30 μM). The heat change accompanying the titration was recorded as differential power by the instrument and determined by integration of the peak obtained. Titrations of ligand to buffer only and buffer into protein were performed to allow baseline corrections. The heat change was fitted using nonlinear least-squares minimization to obtain the dissociation constants, K d , the enthalpy of binding, ΔH, and the stoichiometry, n. Between one and three replicates were performed per titration.
## Bioluminescence resonance energy transfer (bret) analyses
Bioluminescence resonance energy transfer (BRET) experiments were using a NanoBRET PPI starter kit (Promega N1821) according to the manufacturer's instructions and as previously described [bib_ref] Patterns of substrate affinity, competition, and degradation kinetics underlie biological activity of..., Sperling [/bib_ref].
## Drug sensitivity assays
## Cyck stability reporter analysis with crbn overexpression
HEK293T Cas9 cells expressing the cycK eGFP degradation reporter were transiently transfected with pLX307-Luc or pLX307-CRBN (for flow experiment) as described above and 48 hours post infection treated with CR8 for 2 hours and analysed by flow cytometry.
## Extended data
Extended Dataa, Drug sensitivity of K562 Cas9 , P31FUJ Cas9 , THP1 Cas9 and MM1S Cas9 cells exposed to CR8 for 3 days (n=3). b, mRNA expression levels for genes in DCAF library. "|" represents mean (n=4). c, Flow analysis of K562 Cas9 , P31FUJ Cas9 , THP1 Cas9 and MM1S Cas9 cells expressing sgRNAs and a BFP marker (blue fluorescent protein) after a 3-day treatment with 1 μM CR8. "|" represents mean (n>2). a, Sequence alignment of CDK12 and CDK13. In this and later panels asterisks denote contacts with CR8 and circles indicate contacts with DDB1 (coloured according to DDB1 domains, see. Arrows mark differences at the DDB1-CR8-CDK interface. b, Sequence alignment of CDK12 and CDK9. c, Multiple sequence alignment of different human CDKs. d, Titration of CDK12-Alexa488 cycK (0-3.75 μM) to 50 nM terbium DDB1 and 5 μM CR8 or DMSO (n=3). No DDB1 only contains streptavidin-terbium and shows concentration-dependent fluorophore effects. e, Titration of CDK13-Alexa488 cycK (0-3.75 μM) to 50 nM terbium DDB1 and 5 μM CR8 or DMSO (n=3). f, Titration of CDK9-
## Extended
# Supplementary material
Refer to Web version on PubMed Central for supplementary material. a, Pearson correlation between CR8 toxicity and mRNA expression of DDB1. Dots represent cancer cell lines. Smaller area under the curve (AUC) corresponds to higher drug toxicity. TPM, transcripts per million (n=471). b, Flow analysis of HEK293T Cas9 cells expressing sgRNAs and a BFP marker (blue fluorescent protein) after a 3-day treatment with 1 μM CR8 (bars represent mean, n=3). c, Whole proteome quantification of Molt-4 cells treated with 1 μM CR8 (n=1) or DMSO (n=3) for 5 hours (two-sided moderated t-test, n=3). d, Immunoblots of CycK degradation in HEK293T Cas9 cells pre-treated with 0.5 μM MLN7243, 1 μM MLN4924, or 10 μM MG132 for 4 hours followed by exposure to 1 μM CR8 for 2 hours (n=3). e, Immunoblots of CycK degradation time course in HEK293T Cas9 cells treated with 1 μM CR8 (n=3). f, Genome-wide CRISPR/Cas9 viability screen for CR8 resistance in HEK293T Cas9 cells. Guide counts were collapsed to gene-level (n=4 guides/ gene; two-sided empirical rank-sum test-statistics). Black dots denote DCAF substrate receptors. g, Genome-wide CRISPR/Cas9 reporter screen for cycK eGFP stability upon 1μM CR8 treatment in HEK293T Cas9 cells. Guide counts were collapsed to gene-level (n=4 guides/gene; two-sided empirical rank-sum test-statistics). Black dots denote DCAF substrate receptors.
[fig] 34: were previously published. HEK293T cells were cultured in DMEM Gibco) ALX-430-161-M005) from Enzo Life Sciences, roscovitine HY-30237), THZ531 HY-103618), LDC00067 HY-15878) from MedChem Express and DRF053 D6946) from Sigma. [/fig]
[fig] Data, Figure 4 |: Characterization of DDB1-CDK12-cycK complex formation. a, Schematic of the TR-FRET setup. Positions of the FRET donor (terbium-streptavidin (Tb)) and acceptor ( Alexa488 SpyCatcher (A)) are indicated in the structural model. b, Titration of CDK12-Alexa488 cycK (0-3.75 μM) to 50 nM terbium DDB1 and 5 μM CR8 or DMSO (n=3 [/fig]
[fig] Extended Data, Figure 6 |: CDK12 C-terminal extension adopts different conformations. a, Conformation of the C-terminal extension in the structure of the CDK12-CR8-DDB1 ΔBPB complex. b, Structure of CDK12 bound to AMP-PNP (PDB entry 4CXA) superimposed onto CDK12 in the CDK12-CR8-DDB1 ΔBPB complex. c, Titration of CDK12-Alexa488 cycK (0-3.75 μM) to 50 nM terbium DDB1 in the presence of 5 μM THZ531, ATP or DMSO (n=3).d, Structure of CDK12 bound to THZ531 (PDB entry 5ACB) superimposed onto CDK12 in the CDK12-CR8-DDB1 ΔBPB complex. e, THZ531 binding pose in the active site of CDK12 as in d. f, Chemical structure of THZ531. Extended Data Figure 7 |. Differences between CDK12 and other CDKs highlight residues involved in CR8-induced DDB1 recruitment. [/fig]
[fig] Figure 1 |: CR8-induced degradation of cycK depends on DDB1 and CDK12. [/fig]
[fig] Figure 2 |: CR8-bound CDK12 binds DDB1 in a DCAF-like manner. a, Co-immunoprecipitation (IP) experiments with recombinant proteins (n=3). b, In vitro ubiquitination of cycK by the RBX1 N8 CUL4-DDB1 ubiquitin ligase core (n=2). c, TR-FRET signal for CDK12-Alexa488 cycK titrated to Terbium DDB1 in DMSO or 10 μM CR8 (n=3). No DDB1 only contains streptavidin-terbium and shows concentration-dependent fluorophore effects. d, Cartoon representation of the DDB1 ΔBPB -R-CR8-CDK12-cycK crystal structure. e, TR-FRET counter titration of unlabelled wild-type or mutant CDK12-cycK (0-10 μM) into pre-assembled Terbium DDB1-CR8-CDK12-Alexa488 cycK complex (n=3). f, Structural models of CRL4 CRBN bound to lenalidomide and CK1α and RBX1-CUL4-DDB1 (CRL4) bound to the R-CR8-CDK12-cycK complex (bottom). The E2 active site cysteine (red spheres) binds ubiquitin through a thioester bond. [/fig]
[table] Extended Data Table 1 |: Data collection and refinement statistics. ΔBPB -CR8-CDK12 713-1052 -cycK Footnotes: * Data collected from a single crystal † Values in parentheses are for the highest-resolution shell ‡ From STARANISO 42 assuming a local weighted CC 1/2 = 0.3 resolution cut-off [/table]
|
The Subcutaneous Implantable Cardioverter-defibrillator: New Features and Implant Techniques and Future Developments
The use of subcutaneous implantable cardioverter-defibrillators (S-ICDs) has increased over time. Device-based algorithms have been developed to reduce inappropriate shocks. New implant techniques have been developed including the two-incision technique and the placement of the generator submuscularly. More patients are being implanted without general anesthesia. This review summarizes the newest S-ICD features, surgical implantation methods, and device-related safety and efficacy findings.
# Introduction
Traditional implantable cardioverter-defibrillators (ICDs) consist of a pulse generator placed in the pectoral area and endocardial leads with high-voltage shocking coils. This necessitates the implantation of permanent intravascular hardware. Some of the periprocedural complications related to vascular access and intracardiac lead placement include pneumothorax, cardiac perforation, and lead dislodgement, while long-term complications such as device infection and lead malfunction due to exposure to perpetual cardiac motion as well stress from upper extremity bony and soft tissue compression can also occur.These complications are more common in younger, active patients. Of particular concern, device-related endocarditis carries significant mortality (up to 40%), with complete system extraction generally required for cure. Leads that are present for more than one year usually require laser-or mechanically-assisted extraction. The subcutaneous ICD (S-ICD) was designed to provide effective defibrillation while avoiding the complications related to the presence of hardware in the intravascular space. This device represents a novel development in the field of device therapy, and operator experience with the use of this technology has accrued significantly since its inception and approval. This review summarizes the newest S-ICD features, surgical implantation methods, and device-related safety and efficacy findings.
## Subcutaneous implantable cardioverterdefibrillator development
The first S-ICD, the SQRX from Cameron Health (San Clemente, CA, USA), was approved by the United States Food and Drug Administration (US FDA) in 2012. It was 15.7 mm thick, weighed 145 g, and was 69 mL in volume. Following the acquisition of Cameron Health by Boston Scientific (Natick, MA, USA), the second-generation EMBLEM™ device became available in 2015. This device was 12.7 mm thick, weighed 130 g, was 59.5 mL in volume, and included remote monitoring capabilities. In addition, software changes included the SMR8 algorithm, which, in a computer model, decreased T-wave oversensing (TWOS) by 40% with no change in the sensitivity of ventricular arrhythmia events.Then, in 2016, the current third-generation EMBLEM™ MRI S-ICD (Boston Scientific, Natick, MA, USA) was released. It is the same size as the second-generation device with a 7.3-year projected longevity and includes magnetic-resonance-imaging-conditional scanning and software changes, as S-ICD: New Features and Implant Techniques described below. Separately, a new version of the S-ICD lead has also been released in which the suture sleeve is integrated directly on the lead, allowing the implanter to save time by alleviating the need to suture this down. This is in contrast to the old version of the device, in which the suture sleeve had to be manually sewn onto the lead.
A new feature of the current EMBLEM™ S-ICD device (Boston Scientific, Natick, MA, USA) is the SMART Pass arrhythmia discrimination algorithm, which is designed to reduce inappropriate shocks due to TWOS.It also includes an additional high-pass filter to reduce the amplitude of lower-frequency signals such as T-waves while maintaining sensing of higher-frequency R-waves. Limited data suggest that this algorithm reduces inappropriate shocks by 82% in comparison with the first-generation device and 71% in comparison with the second-generation device, respectively, with no change in time to detection or ventricular arrhythmia detection accuracy.The AF Monitor, another algorithm, stores episodes of suspected atrial fibrillation using ventricular scatter analysis, a measure of heart rate variability. To diagnose atrial fibrillation, the criterion of more than 80% of beats in a 192-beat window must be met, which may lead to the undercounting of shorter-duration episodes. In addition, a relatively regular ventricular response while in atrial fibrillation may be missed by the device. At least six minutes of atrial fibrillation must be present for the event to be stored, but the six minutes do not need to be continuous. One episode per day is stored for retrieval, with a total of up to seven episodes lasting 44 seconds in duration being kept. The device will report the percentage of the time spent in atrial fibrillation and the number of days where atrial fibrillation was detected within the last 90 days. In abstract form, this algorithm showed 100% specificity and 94% sensitivity for atrial fibrillation. 4
## Subcutaneous implantable cardioverterdefibrillator screening
Screening continues to be important to reduce the risk of inappropriate shock. For this, electrodes are placed on the skin to approximate S-ICD sensing. The left leg electrode is placed in the fifth intercostal space on the midaxillary line, the left arm electrode is placed 1 cm lateral to the left side of the xiphoid midline, and the right arm electrode is placed 14 cm above the left arm electrode. If right-sided lead placement is to be completed, screening should be performed with the left and right arm electrodes located to the right of the sternum. The grounding electrode should be placed on the right lower abdomen.
Traditionally, screening is completed manually and can still be performed in this manner. The sweep speed is set at 25 mm/s with a gain of 5 mm/mV to 20 mm/mV, with a preference for the largest gain as long as there is no clipping of the QRS complex. The electrocardiogram (ECG) should be recorded for 10 seconds to 20 seconds with the patient both supine and standing, although sitting, left or right decubitus, or a prone position may be considered in circumstances in which there is concern of significant QRS or T-wave variability with body positioning. A patient screening tool made of a transparent plastic with colored profiles is used to assess the QRS complex and T-wave. The QRS is placed within the first part of the profile, with the specific profile chosen based on the size of the QRS complex. A screening pass is defined when all QRS complexes and T-waves fall within a profile, excepting occasional failures for premature ventricular complexes or paced beats (if a pacemaker is present). The same ECG lead should pass in the supine and standing positions. In general, the ZOOM™ Programmer (Boston Scientific, Natick, MA, USA) represents a good tool for screening. A 12-lead ECG machine can also be used, but, if borderline screening failure occurs due to small changes in filtering between machines, repeat screening with the ZOOM™ Programmer (Boston Scientific, Natick, MA, USA) may produce a positive result.There is now an automated screening tool (AST) built into the ZOOM™ Programmer (Boston Scientific, Natick, MA, USA). It uses the 3-Hz to 40-Hz band-pass filter in the programmer to match the sensing algorithm of the S-ICD. The AST needs at least six beats over 15 seconds to check the mean and standard deviation of the R-wave amplitude. If an R-wave is greater than one standard deviation in size from the mean, it is discarded and the mean is remeasured. This new average has to be greater than 0.47 mV or it is considered to be a screening failure. The second part of the AST is a measurement of the QRS:Twave ratio; a screening pass is considered to demonstrate a result of more than 3.5:1.
The AST was designed to increase the speed of screening and reduce interobserver variability as a cause of screening failure. In addition, it more closely mimics the sensing filters of the S-ICD versus printing with manual evaluation.
## Subcutaneous implantable cardioverterdefibrillator implantation
Initially, most operators used the three-incision technique for placement of the S-ICD lead. However, over time, the two-incision technique has become the preferred technique, with more than half of the implants in the US using this.With this method, after the xiphoid incision is made, an 11-French or 12-French peel-away sheath is placed over the tunneling tool . This tool is then advanced parasternally approximately 14 cm superiorly toward the manubrium. Next, the tunneling tool is removed, leaving the splittable sheath behind in the subcutaneous tissue, and the lead is inserted through the sheath, which can then be removed. The benefit of this technique is that it reduces procedure duration with equivalent safety and successful placement rates .Submuscular generator placement is an option to prevent erosion or for a more cosmetic appearance. One method is to position the generator between the serratus anterior and latissimus dorsi muscles.This leads the device to be located more posteriorly than the standard subcutaneous generator location but cushions it to limit erosion.
A more complicated approach involves placing the device underneath the serratus anterior muscle.With this technique, the muscle slips of the serratus anterior are separated and a submuscular pocket is created beneath it. After the generator is placed in the pocket, the muscle slips are sutured together to keep the device in place.The generator can also be placed underneath the relatively thick fascia of the serratus anterior without going underneath the muscle. Care must be taken to prevent damage to the long thoracic nerve with these approaches.
There do not appear to be differences in outcomes between generator placement locations, but the use of the submuscular position is more painful for patients and the available data are limited with respect to long-term follow-up of the submuscular and subfascial generator locations.Overall, the depth of generator placement can be decided based on physician experience and patient factors such as body habitus to enhance patient acceptance of the device.
## Anesthesia and analgesia management
Most of the initial implant S-ICD procedures were performed under general anesthesia. Creation of the pocket and tunneling of the lead can be painful for patients, and most patients undergo defibrillation threshold testing (DFT) at implant. General anesthesia requires an anesthesiologist but allows the operator to focus solely on the procedure. However, general anesthesia and postanesthesia nursing care may not be widely available, and general anesthesia may lead to more hypotension, longer in-room time, and increased cost.In the S-ICD System Postapproval Study conducted by Boston Scientific (Natick, MA, USA), 64% of patients had general anesthesia.Several other options for anesthesia have been described in recent years.
Monitored anesthesia care (MAC) may be used in some cases, frequently with propofol. 14 Mild to moderate sedation may be employed during some parts of the procedure, but, for tunneling and DFT, deeper sedation is often required. Local anesthesia should also be used to reduce sedation needs. Care should be taken to avoid oversedation and respiratory decompensation. The operative team should be familiar with airway maneuvers and an anesthesia team should be available if complications occur.
Nonanesthesiologist-administered sedation and analgesia (NASA) does not involve an anesthesiologist but involves sedation delivered by or under the guidance of the implanting physician. It may involve various agents including benzodiazepines, opioids, propofol, methohexital, etomidate, and/or nalbuphine.However, similar to using MAC, concerns about respiratory compromise or hypotension also exist, and NASA requires comparable cardiovascular and respiratory monitoring and a nurse dedicated solely to this task. Although this option is appealing for many hospitals, as it does not require an anesthesiologist and is less costly, there can be safety concerns. Staff must be appropriately trained to handle any potential complications, and the patient must be deemed eligible as a candidate to undergo NASA.
Recent reports have described regional nerve blocks used for breast surgery as an alternative method for S-ICD implantation anesthesia.These techniques target the thoracic intercostal nerves to provide a 13-hour to 14-hour analgesic effect. A recent report of 19 patients undergoing S-ICD implantation with two different nerve block techniques suggested it is feasible, safe, and can be performed in about 10 minutes.The serratus anterior plane block provides analgesia to the area in which the generator pocket is made by targeting the lateral cutaneous branches. Under ultrasound guidance, the T4-T5 level at the midaxillary line is imaged to find the plane between the latissimus dorsi and serratus anterior muscles.After confirming that this location is correct, 20 mL of 0.25% bupivacaine is injected.
The transversus thoracic muscle plane block provides analgesia to the sternum where the xiphoid incision and defibrillator coil portion of the lead are located by targeting the anterior branches. The plane between the transversus thoracis and internal intercostal muscles can be found near the left parasternal border at the T3-T4 level, and bupivacaine is injected here. 18 MAC, NASA, or general anesthesia is still needed, but it is possible that regional nerve blocks can be used to limit postoperative pain.
## Efficacy and safety of the subcutaneous implantable cardioverter-defibrillator
The S-ICD underwent more than 10 years of preclinical and early clinical studies. Initial work demonstrated that the implantation position with the lowest defibrillation threshold was that of a left lateral pulse generator and an 8-cm shocking coil tunneled lateral to the left sternal border, which has subsequently become the currently used configuration.The device received Conformité Européenne approval in 2009. The trial that led to US FDA approval of the device in 2012 was an Investigational Device Exemption (IDE) study.This was a nonrandomized cohort study of 321 patients with ICD indications with an 11-month median follow-up period. The primary efficacy endpoint, acute conversion rate of induced ventricular fibrillation, was 94.7% with a single shock. Freedom from serious device-related complications at 180 days was 99% and the all-cause complication-free rate was 92.4%. The US FDA mandated a postapproval registry, and initial results of this observational study were published in 2017.The perioperative outcomes of 1,637 relatively sick patients from the US showed a similarly high, 30-day freedom from device-or procedure-related complications rate of 96.2%. Successful DFT, when attempted, occurred in 98.7% of patients.
A multinational registry of S-ICD recipients, mostly from European countries, called the Evaluation of Factors Impacting Clinical Outcome and Cost-effectiveness of the S-ICD (EFFORTLESS) study, has to date reported its early and midterm results.The latest report from this large registry was published in 2017.It included 985 patients, with a median follow-up of 3.1 years. The implantation indication in this cohort was primary prevention in 65%. The overall rate of inappropriate shocks was 11.7% and appeared to be lowered significantly by dual therapy zone programming. Most inappropriate shocks were due to TWOS. It is worth noting that only a minority (7.8%) of patients followed up with in this cohort had been implanted with the second-generation S-ICD, which has algorithms designed to reduce the rate of inappropriate shock by 30% to 40%.DFT is currently recommended for all patients undergoing S-ICD placement.In the S-ICD System Postapproval Study, there was a 98.7% conversion success rate.However, 1.6% of patients were noninducible and 12.3% did not undergo testing. In the EFFORTLESS cohort, 99.7% had a successful shock during testing, but some did require repositioning of the lead or generator to be successful, whereas 11.1% did not undergo defibrillating testing. 22 Recent data from the ICD Registry from September 28, 2012 to April 1, 2016 demonstrated that only 70.7% of patients undergoing S-ICD implant in the US have undergone DFT testing.The greatest determinant of whether testing was performed was the implanting facility. Almost 7% of patients had an inappropriate safety margin, defined as the lowest energy successfully converting out of ventricular fibrillation of > 65 J. Importantly, DFT testing was not associated with in-hospital complications. Two smaller studies have suggested the safety of S-ICD implantation without DFT testing with reasonable patient outcomes and good shock efficacy.Overall, these findings show that, while recommended, not every patient can be a suitable candidate for DFT testing. For instance, the patient may be in atrial fibrillation and has not received appropriate anticoagulation, and there is concern that the shock will convert the rhythm to normal sinus. Some patients may also be thought to have too severe a degree of cardiac disease to undergo DFT. Larger studies are needed, however, before recommending that DFT testing is not needed for the S-ICD population.
One major concern with S-ICD therapy is regarding the lack of antitachycardia pacing (ATP) capability to treat monomorphic ventricular tachycardia (MMVT). Only 2.2% of patients in the EFFORTLESS cohort experienced more than one shock for MMVT during the 3.1 years of follow-up, a rate similar to the rate of 1.5% per year reported in the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT).The lack of ATP capability may not need to be a major deciding factor regarding what type of ICD should be offered. However, the lowest programmable treatment zone in the S-ICD is 170 bpm. Therefore, a history of slow ventricular tachycardia, especially if not controlled with antiarrhythmic therapy or ablation, would be a contraindication for S-ICD implant at present.
## The subcutaneous implantable cardioverter-defibrillator in hypertrophic cardiomyopathy
Patients with hypertrophic cardiomyopathy (HCM) are at varying risk for sudden cardiac death, and the ICD has been the main prophylactic measure for sudden cardiac death.The S-ICD seems to be a particularly attractive option in this patient population, given the frequently young age at implantation, the anticipated three to four decades of ICD therapy and lead failures, and the general lack of pacing needs. As mentioned previously, TWOS is the major cause of inappropriate shock delivery seen with the S-ICD system, with increased QRS duration representing a risk factor for this untoward outcome. Patients with HCM can often have striking
## S-icd: new features and implant techniques
QRS and T-wave abnormalities. In a study of 27 patients with HCM who were screened for S-ICD candidacy, 85% passed the screening test and 59% proceeded with device implantation.In early studies, patients were programmed to a single shock zone, and it was recognized that dual-zone programming significantly reduced the rate of inappropriate shocks. In comparison, results from a large meta-analysis that included 2,190 patients with HCM showed that the rate of inappropriate shock using transvenous devices was on the order of 5%.Given the accumulating experience with S-ICD implantation as well as improved device programming and sensing algorithms, the rate of inappropriate shocks is expected to decrease. Judicious screening of HCM patients will also help to minimize the risk of inappropriate therapies. It has been reported that 7% to 16% of HCM patients fail screening in all three sensing vectors.This rate may be up to 38% in children with HCM.Although exercise ECG screening may not be necessary in the majority of S-ICD implants,patients with HCM may benefit from this additional screening step given the dynamic changes that may occur in the R-and T-wave morphologies in this patient population. These studies illustrate that a significant number of patients with HCM are potentially eligible for S-ICD implantation and showed device efficacy in terminating induced ventricular fibrillation.
## Future directions
With the introduction of the S-ICD came the beginning of the elimination of the "weakest link" in the ICD system, the transvenous lead.The S-ICD continues to have shortcomings, however, one of which is that it does not offer bradycardia pacing or ATP. Futures versions of the device could potentially be paired with a leadless pacemaker, with communication occurring between the devices. This has been demonstrated in an animal model wherein the S-ICD sensed the ventricular arrhythmia and commanded the leadless pacemaker to deliver ATP.In 99% of episodes, there was successful communication between the S-ICD and leadless pacemaker. It is possible that future versions of this combination system will include sensing from the leadless pacemaker to enhance discriminatory ability and reduce inappropriate shocks from TWOS.
Some randomized controlled trials evaluating the efficacy of the S-ICD will shed light on important clinical issues. The Multicenter Automatic Defibrillator Implantation Trial with S-ICD (MADIT S-ICD) study is currently ongoing (NCT02787785).Patients with diabetes mellitus and ischemic cardiomyopathy have a higher rate of sudden cardiac death than do those without diabetes mellitus, even when the ejection fraction is more than 35%. 42 MADIT S-ICD was expected to randomize 1,800 patients with diabetes mellitus, prior myocardial infarction, ejection fraction of 36% to 50%, and age ≥ 65 years to an S-ICD or usual medical therapy arm with a primary endpoint of all-cause mortality. In addition, the Prospective, Randomized Comparison of Subcutaneous and Transvenous ICD Therapy (PRAETORIAN) study is an ongoing clinical trial (NCT01296022) randomizing patients with a class I or IIa indication for an ICD without pacing indications to either a transvenous or subcutaneous device.The primary endpoint is ICD-related adverse events.
# Conclusion
The S-ICD is no longer an emerging therapy.At this point, enough data support its use in many different populations. Surgical and anesthetic/analgesic techniques are evolving to improve speed of implant, enhance safety, and reduce postoperative pain. Future iterations will continue to enhance arrhythmia discrimination, possibly with the pairing of a leadless pacemaker. |
Network analysis of the relationship between negative life events and depressive symptoms in the left-behind children
Background: There are 68.77 million left-behind children in China, who are at a great risk of depression associated with negative life events. Our study aims to investigate the most central symptoms of depression in left-behind children and the relationship between depressive symptoms and negative life events using network analysis. Method: A cross-sectional data set (N = 7255) was used, which included children and adolescents aged 7 to 17. Network analysis was used to evaluate: 1) the most central symptoms among the items included in Child Depression Inventory (CDI) of the left-behind children; 2) bridge symptoms between depressive disorder and Adolescent Self-Rating Life Events Check List (ASLEC) of the left-behind children; 3) differences in networks of depressive disorders between left-behind and non-left-behind children, and 4) differences in the network of depression and negative life events between left-behind and non-left-behind children. The stability and centrality indices of the network were also evaluated in the study. Results: The most central symptoms in the CDI among the left-behind children included self-hatred, crying, fatigue, and sadness. The items with the highest bridge strength centrality in the CDI-ASLEC network included academic stress, discrimination, and school performance decrement. Higher bridge strength values indicate a greater risk of contagion to other communities. The connections in the CDI-ASLEC network are denser in the left-behind children than in non-left-behind children. Limitations: The study which was conducted based on cross-sectional data shows that network analysis can only make undirected estimation, but not causal inferences. Conclusions: We identified the core symptoms of depression and the bridge symptoms between negative life events and depression in the left-behind children. These findings suggest that more attention should be paid to self-hatred, sadness, and fatigue in the treatment of depression in left-behind children. Intervention for academic stress and discrimination of the left-behind children may help to reduce the contagion of negative life events to depression symptoms.
# Introduction
China's reform and opening-up and the subsequent social and economic development have led to the prosperity of many cities. A large amount of the population in the undeveloped areas in central and western China began to move to the developed area for employment opportunities. According to the Report on China's Migrant Population Development, there were more than 200 million floating populations in China in 2020, and over 70% of them gathered in the developed region in eastern China. The data show that the average monthly income of the floating population is 4598 yuan. After paying for food, clothing, housing, and transportation, their disposable income is too low to afford their children's education and housing in the developed region. Therefore, a large number of children are left behind by their parents in their hometown, who are known as leftbehind children. The left-behind children are defined as children who have lived in undeveloped areas for more than 6 months and have one parent or both work in developed areas. In 2015, there were 68.77 million leftbehind children in China, including 40.51 million in rural areas. These children are usually taken care of by their relatives, especially grandparents, who pay more attention to their physiological needs rather than their psychological needs.
Although studies have shown that working in the developed area can increase family income and raise family status in the local area, it is rarely beneficial for leftbehind children. Researches show that left-behind children have more mental health problems than nonleft-behind children, including a higher risk of depression, and anxiety, a stronger sense of loneliness, and higher suicide risk, highlighting severe psychological problems of these children and tough challenges posed by the problems.
Depression is the most common mental health problemin childhood, and depression in children and adolescents is usually the first episode of depression. Early prevention and intervention can help reduce depression symptoms and the risk of recurrence. A previous systematic review and meta-analysis showed that left-behind children have higher depression risk and depression scores than non-left-behind children (RR 1.52 [95% CI 1.27-1.82]; SMD 0·16 [0·10-0·21]). However, at present, most studies only compare the differences in the score of depression scale between leftbehind and non-left-behind children, while few studies have made a detailed comparison of depressive symptoms between left-behind and non-left-behind children. Therefore, it remains unknown whether the symptoms of depression differ between them.
Depression of children is associated with many factors, such as environmental stress, life trauma, bullying, etc. Studies show that children who are exposed to negative life events are at a higher risk of depression. It is also reported that left-behind children are more likely to be exposed to negative life events. For example, in rural areas of Mexico, the children whose father works in other places have 39% higher incidence of disease than the non-left-behind children, and their incidence of diarrhea is increased by 51%. In addition, the risk of physical and mental disorders in the left-behind children is higher than that in the non-leftbehind children, such as car accidents, falls, sexual abuse, neglectand bullying, and less communication with parents. It is worth noting whether the negative life events experienced by left-behind children aggravate their depressive symptoms. Studies have shown that reducing these negative life events can alleviate depressive symptoms. For example, more frequent and longer parent-child communication can significantly reduce the incidence of depression in left-behind children. Therefore, it is significant to investigate the effect of negative life events on the depression of the left-behind children.
We found in our previous studies that the overall prevalence of depression in the left-behind children was 24.8%. Meanwhile, we also found that high income, frequent parent-child communication, telephone communication or talking about learning experiences, school life and emotional experiences are protective factors for depression. In addition, prevalence of depressive symptoms is higher in the left-behind children than in non-left-behind children, and negative life events are risk factors for depression in left-behind children. Despite the work we have done above, it is still unclear which negative life event plays a major role in certain type of depressive symptom.
Few studies have investigated the relationship between negative life events and depression symptoms of leftbehind children, and to our knowledge, no studies have been conducted to explore their relationship using network analysis. Therefore, the understanding remains limited regarding their relationship, as well as the major types of the negative life events experienced by the leftbehind children and their core depression symptoms. This study aims to explore the relationship between negative life events and characteristics of depression symptoms of left-behind children using network analysis. The network usually consists of symptoms (nodes) of mental disorders and connections (edges) between them. According to the network theory, diseases are usually caused by the interaction between symptoms, and the symptoms themselves are not mental diseases, but components of mental diseases. Alleviating the symptoms of disease plays an important role in disease treatment. Targeting the core symptom derived from network analysis helps us treat the underlying disease.
We investigated the network structure of depressive symptoms of a large sample (N = 2517) of left-behind children and explored the relationship between depression and negative life events using network analysis. We aimed to: a) construct depression networks to identify the main symptoms of depressive disorders of leftbehind children; b) compare the differences in depression networks between left-behind children and nonleft-behind children; and c) use bridge centrality to identify the disease pathway linking negative life events and depressive symptoms of left-behind children.
In this study, two networks were constructed: the network of depressive symptoms and the network of both depressive symptoms and negative life events. These networks were used to identify the most central symptoms of depression disorders of left-behind children, to identify the negative life events with a stronger correlation with depressive symptoms of left-behind children using bridge symptoms, and to compare the differences in the two networks between left-behind children and non-leftbehind children.
# Methods
# Ethics statement
The current study was reviewed and approved by the Medical Ethics Committee of the Department of Medical Psychology, Army Medical University (No. CWS20J007). The study was conducted in accordance with the Declaration of Helsinki guidelines. The approval from officials of sampled schools was obtained in a written form. Parents or legal guardians were contacted to provide consent on behalf of the children to participate in the study. After reading the informed consent, participants can complete the following survey if they want to further participate in this study. The participants were also assured that the survey was anonymous and personal information would not be disclosed.
## Participants
This cross-sectional study using a dataset from a previous studyconducted a three-phase survey in districts and counties of Chongqing. In this study, we selected samples from 19 districts and counties: Chengkou, Wuxi, Wushan, Fengjie, Yunyang, Liangping, Fengdu, Dianjiang, Changshou, Shizhu, Pengshui, Qianjiang, Wulong, Youyang, Xiushan, Tongnan, Dazu, Hechuan and Jiangjin, with a total of 9383 participants. All participants completed the Child Depression Inventory (CDI) and the Adolescent Self-Rating Life Events Checklist (ASLEC), as well as the demographic information questionnaire. Left-behind children were selected according to three criteria: 1) Both parents have been working outside the home for more than 6 months; 2) They live with family members other than their parents; and 3) They are aged between 7 and 17 years old. Participants were excluded if they missed responses to more than 50% items or had a single parent going out. Only 7255 children were included at last.shows the flow of participants for this study.
## Assessments
## Child depression inventory (cdi)
The depressive symptoms of left-behind children were determined based on the Chinese version of CDI, which is the most commonly used questionnaire for assessing depression among children and adolescents aged 7-17. CDI is a self-report questionnaire of 27 items on five dimensions, which is designed to assess the loss of pleasant sensation (e.g. symptoms like unhappiness), inefficiency (e.g. less motivation), interpersonal problems (e.g. unable to maintain relationships), negative emotions (e.g. sadness), and negative self-assessment (e.g. feeling of being unable to do anything). Each item has three statements of different degree and requires the child to choose the one that best describes himself or herself in the past 2 weeks. Each item is scored on a scale of 0 (low), 1 (medium) and 2 (high), with a total score ranging from 0 to 54. Half of the items are scored in reverse order, with higher scores indicating more severe depressive symptoms. The generally recommended cut-off point for depression is 19 or 20 points. However, the cut-off point in this study was 12 points as this study was not aimed to diagnose depression, and a lower cut-off point could include more likely depressed individuals. In the current study, the α coefficient of the questionnaire was 0.83, showing good consistency.
## Adolescent self-rating life events check list (aslec)
ASLEC was developed by Liu Xianchen et al.to assess the frequency and stress intensity of stressful life events among adolescents, especially middle school students. The scale consists of 27 items on six dimensions, i.e. interpersonal relationship, academic stress, punishment, loss, health adjustment and others. ASLEC is a self-rating scale, requiring participants to determine whether the event described in each item has occurred within the past 3 months; if so, they need to tick on a scale of 1 (no influence) to 5 (significant influence), and if not, to tick the box of "not happening" scored as "no influence". The total score of the scale ranges from 27 to 135, with higher scores indicating greater total stress. The scale shows good reliability and validity. In the current study, the α coefficient of the scale was 0.927.
# Data analysis
Missing data All data were analyzed in R 4.0.0. Using the mice software package in R language, we adopted the insertion method for the data with less than 50% missing individual-level data. In this study, there were missing data for 3754 options, accounting for 0.95% of the total options (391,770).
Glasso network The EBIC glasso function of the qgraph software package in the R language was used to estimate the two networks, both of which used datasets (N = 7255). GGM was regularized by graphical LASSO (Least Absolute Shrinkage and Selection Operator) algorithm. This can shrink all edges and make small edges become zero-weight edges to obtain a more stable and interpretable network. The GGM adjustment parameter was set to the recommended value of 0.5 to well judge and measure the sensitivity and specificity of discovering true edges. In the visual network, the red edge represented the negative partial correlation between nodes, and the blue edge represented the positive partial correlation between nodes, with thicker edges indicating the stronger correlation between nodes. We estimated the network containing depressive symptoms (i.e. CDI network), and the depression network containing negative life events (i.e. CDI-ASLEC network).
Stability and accuracy analysis The stability and accuracy of the network were calculated with the bootnet function of R software. First, the accuracy of the edge weights was evaluated by calculating 95% confidence intervals (CI) using bootstrapping. The narrower the CI was, the more accurate the estimation of the edge weights and the centrality index would be. The stability of the centrality index was then estimated by calculating the correlation stability coefficient using case-dropping bootstrapping. The correlation stability coefficient (CScoefficient) is the maximum percent of cases that can be excluded if the correlation between the centrality indices of the original sample and the subset of the sample is 0.70 or higher (95% probability). The CS-coefficient (how much data can be discarded) should not be less than 0.25, preferably higher than 0.50.
Centrality and difference analysis The qgrath package of R software was used to calculate the centrality index, and the bootnet package was used to test the node centrality difference. Studies have shown that strength centrality is more stable than compactness and intermediation. Therefore, strength centrality was used as an index in this study. Strength centrality represents the sum of edge weights of each node (e.g. correlation coefficients), reflecting the possibility that the activation of a certain symptom may lead to the activation of others.
Bridge symptoms Bridge symptoms are thought to be the overlapping symptoms of two psychiatric disorders. In this study, we used bridge centrality statistics to determine the overlapping symptoms of negative life events with depressive symptoms. Bridge strength centrality is the best index to identify bridge nodes, so we calculated bridge strength centrality and bridge expected influence centrality. According to the report by Jones et al., eliminating bridge symptoms can prevent the spread of one disease to another.
## Comparison between left-behind children and nonleft-behind children
In order to compare the differences in depression network between left-behind children and non-left-behind children, we used the R "network contrast test (NCT)" package. NCT uses permutation test to compare the invariance in global strength (i.e. the sum of all edge weights) and structure between two networks. In order to compare the differences between the two samples, we generated a depression network and a depression-negative life event network to explore whether their depressive symptoms share the same correlation, whether their strongest symptoms are consistent, and whether they experience the same negative life events.
# Results
The left-behind children accounted for 56.86% (N = 4125) and non-left-behind children accounted for 43.14% (N = 3130) of the total sample (N = 7255). The cut-off value for depressed individuals was 12 points in the depression scale. The final analysis included 2517depressed left-behind children and1716 depressed nonleft-behind children. More information on age, gender, and depression level can of the sample can be found in.
## Network stability
The stability of edge weights of the two networks met the requirements, greater than the recommended 0.5. A higher CI of edge weights indicated a higher accuracy of network estimation. The edge weight of network 1 and network 2 was 0.59 and 0.75, respectively. In addition, since the strength centrality index was previously reported to have higher repeatability and stability, we focused more on the strength centrality index. The CS-coefficient of strength centrality index in network 1 and network 2 was 0.594 and 0.75, respectively, both higher than the recommended critical value of 0.25.
## Centrality
## Network 1 (cdi network)
The structure and centrality of depression network was shown inand, respectively. Network 1 only contained 27 central depressive symptoms, and the symptoms with the highest strength centrality included self-hatred (S = 2.28), crying (S = 1.68), fatigue (S = 1.25), and sadness (S = 1.23). The results of the strength centrality difference test showed that the estimated strength centrality of self-hatred was significantly higher than that of other symptoms (P < 0.05).
## Network 2 (cdi-aslec network)
The structure and centrality of the depressionadolescent life event network was shown inand, respectively. The network included depressive symptoms and negative life events, and the symptoms with the highest strength centrality were academic stress (S = 1.46), public humiliation (S = 1.35), and self-hatred (S = 1.24). The results of the strength centrality difference test showed that the estimated strength centrality of self-hatred was significantly higher than that of other symptoms (P < 0.05).
## Bridge symptoms
Network 2 There were some bridge symptoms between depression and negative life events. The symptoms with the highest bridge strength centrality were: academic stress (BS = 0.15), discrimination from others (BS = 0.14) and school performance decrement (BS = 0.12) due to depression. The symptoms with the highest bridge expected influence centrality included academic stress (EI = 0.14) caused by negative life events and school performance decrement (EI = 0.12) due to depression.
# Discussion
We used network analysis to study the characteristics of depression network and depression-negative life events network with a large sample (N = 2517) of left-behind children. We found that differences existed in the strength of symptoms in the CDI network of left-behind children, and self-hatred has the highest centrality in the network, consistent with the previous research which shows that self-hatred also has the highest centrality in the CDI of the non-left-behind children. This means that left-behind and non-left-behind children have similar central symptoms in the CDI depression network. NCT analysis in our study also shows that there is no significant difference in the CDI depression network between left-behind and non-left-behind children. These findings may be explained by the fact that left-behind and non-left-behind children share the same developmental stage, such as puberty. On the one hand, the left-behind children begin to have self-awareness, and the derived self-related internal informationplays an important role in depression. On the other hand, with the development of self-identity in adolescence, the children tend to pay more attention to selfachievements, family atmosphere, etc.. Indeed, previous studies have shown that lower self-worth is associated with depression. The negative self-information has become a potential risk factor of adolescent depression, while positive self-identity information can prevent negative effects. Last but not least, depression may also be associated with physical development during puberty, such as the development of secondary sexual characteristics and height, and with excessive attention paid to appearance.
In addition, we also found some other high-intensity symptoms, such as crying, fatigue and sadness. Among them, crying and fatigue are not the most central symptoms according to the previous report. Our finding about fatigue and sadness can be explained by the fact that left-behind children experience more bullying or abuse, and they can only reduce the risk of depression through self-sympathy. Fatigue symptoms stem from more manual labor they need to shoulder when living with grandparents. Compared with non-left-behind children, left-behind ones reported increased working hours, and the fatigue increased the risk of depression. Therefore, we should improve the social support for left-behind children, including protecting them against abuse and bullying and reducing their extra working hours, to reduce their risk of depression. It is interesting to find that these central symptoms in the depression network of the left-behind children differed from those in CDI-ASLEC network which included the negative life events. In CDI-ASLEC network, academic stress and public humiliation in ASLEC and school performance decrement in CDI have the highest centrality reflecting great academic stress and impaired self-esteem of the left-behind children. It should be noted that these networks are undirected rather than directed (causal networks). The most central symptoms were not obtained based on all symptoms, but on the symptoms we input.
We also identified bridge symptoms in CDI-ASLEC network. Bridge symptoms are considered to be an illness pathway for one disorder to spread to another. Therefore, when one disorder appears, intervention of potential bridge symptoms can effectively prevent the spread of disorders and the development of complications. In this network, academic stress, discrimination and school performance decrement have the strongest bridge strength centrality in the left-behind children, consistent with the previous studies. Parents' inquiry about study accounts for a high proportion of their communication with their children, which increases the children's academic stress, and in turn increases the risk of depression. In addition, the selfesteem of left-behind children will be impaired by discrimination by others, and low self-esteem is more likely to lead to depression. Taking corresponding measures to intervene bridge symptoms, especially those with high intensity centrality, can effectively reduce the risk of depression. For example, some studies have reported that regulating self-esteem bridge symptoms can reduce the impact of negative life stress on depression. However, future studies are warranted to confirm if academic pressure and difficulties can serve as targets for intervention. Finally, we used the network comparison test to compare the differences between left-behind children and non-left-behind children in CDI network and CDI-ASLEC network. It is found that, firstly, in CDI network, there is no difference in network structure and global strength between left behind and non-left behind children; secondly, in CDI-ASLEC network, there is no difference in network structure between the left-behind and non-left-behind children. However, there is difference in global strength, indicating that the characteristics of left-behind children network have stronger connection than those of the non-left-behind children. These results show that compared with non-left-behind children, negative life events have a greater impact on left-behind children. One reason is that left-behind children are exposed to more negative life events; another reason is that left-behind children show more depressive symptoms when they encounter negative life events than non-left-behind children. Therefore, stress caused by the negative life events is a common risk factor for depression in the left-behind children.
In short, this study uses network analysis to obtain more subtle results. Our previous research found that negative life events are a risk factor for depression, and proper and adequate communication between the left-behind children and their parents can effectively reduce the impact of negative life events on depression. This study further explored the relationship between negative life events and depression from the perspective of symptoms. The results regarding bridge symptoms indicate that academic stress and school performance decrement are the main pathways linking negative life events and depression. This result further supports our previous findings that the parent-child communications about learning experiences and school life are protective factors for depression. When communicating with the left-behind children, the parents should avoid increasing their academic stress, and meanwhile, they should show more care about their study and life. By doing so, they can reduce the depression of left-behind children caused by negative life events more effectively.
# Limitations
There are some limitations in this study. First, we used sample data from the results of a cross-sectional survey, thus we cannot infer the causality dynamically. Second, the data is from previous research, with a poor timeliness. Therefore, it is necessary to repeat the investigation within a certain period of time to explore the evolvement of depressive symptoms of the left-behind children. In addition, due to the limitations of the study method itself, we did not input all the symptoms into the network model while other surveys, using different survey tools, found network structures differ from those of CDI.
# Conclusions
This study explores the characteristics of CDI network and CDI-ASLEC network of the left-behind children aged 7-17 using network analysis. The results show that negative life events of left-behind children are closely associated with depression. Besides self-hatred, attention should also be paid to sadness and fatigue in the treatment of depression of the left-behind children. The intervention of academic stress and discrimination by others should also be considered in the treatment of depression, which may help alleviate the effect of negative life events on depression. The identification of core symptoms of depression facilitates the diagnosis and treatment of depression, offers suggestions for the reform of public policies for the left-behind children, and provides guidance on the education of their children for parents and guardians. |
PCR-based rapid genotyping of Stenotrophomonas maltophilia isolates
Background: All bacterial genomes contain repetitive sequences which are members of specific DNA families. Such repeats may occur as single units, or found clustered in multiple copies in a head-to-tail configuration at specific loci. The number of clustered units per locus is a straindefining parameter. Assessing the length variability of clusters of repeats is a versatile typing methodology known as multilocus variable number of tandem repeat analysis (MLVA).Results:Stenotrophomonas maltophilia is an environmental bacterium increasingly involved in nosocomial infections and resistant to most antibiotics. The availability of the whole DNA sequence of the S. maltophilia strain K279a allowed us to set up fast and accurate PCR-based diagnostic protocols based on the measurement of length variations of loci carrying a variable number of short palindromic repeats marking the S. maltophilia genome. On the basis of the amplimers size, it was possible to deduce the number of repeats present at 12 different loci in a collection of S. maltophilia isolates, and therefore label each of them with a digit. PCR-negative regions were labelled 0. Coamplification of two pairs of loci provided a 4-digit code sufficient for immediate subtyping. By increasing the number of loci analyzed, it should be possible to assign a more specific digit profile to isolates. In general, MLVA data match genotyping data obtained by PFGE (pulsed-field gel electrophoresis). However, some isolates exhibiting the same PCR profiles at all loci display distinct PFGE patterns.Conclusion:The utilization of the present protocol allows to type several S. maltophilia isolates in hours. The results are immediately interpretable without the need for sophisticated softwares. The data can be easily reproducible, and compared among different laboratories.
# Background
After years of debate regarding its appropriate taxonomic position, the nonfermentative, gram-negative bacillus previously known as Pseudomonas maltophilia or Xanthomonas maltophilia, has been definitively classified as Stenotrophomonas maltophilia. This species is found in a wide variety of environments, and has been isolated from different sources, including water, sewage, soil and plant rhizosphere environments [bib_ref] Microbiological and clinical aspects of infection associated with Stenotrophomonas maltophilia, Denton [/bib_ref]. S. maltophilia is increasingly prevalent in hospitals, and is often isolated in hospitalized patients, as well as in cystic fibrosis (CF), burn, and immunosuppressed patients. The presence of S. maltophilia in CF patients is not associated with a worse clinical outcome. However, the organism contributes to chronic airway inflammation [bib_ref] Immunostimulatory properties of the emerging pathogen Stenotrophomonas maltophilia, Waters [/bib_ref]. Moreover, in mixed infection formed in the CF lungs, S. maltophilia has been shown to influence the architecture of Pseudomonas aeruginosa biofilms by producing a diffusable signal factor [bib_ref] Interspecies signalling via the Stenotrophomonas maltophilia diffusible signal factor influences biofilm formation..., Ryan [/bib_ref].
S. maltophilia isolates exhibit high genetic diversity. Genotypic profiles have been determined by a variety of methods, including AFLP (amplified fragment length polymorphism) fingerprinting [bib_ref] Genomic diversity of the genus Stenotrophomonas, Hauben [/bib_ref] , RFLP (restriction fragment length polymorphism) analysis of the gyrase B gene [bib_ref] Vandamme P: Identification of genomic groups in the genus Stenotrophomonas using gyrB..., Coenye [/bib_ref] or the intergenic region between smeD and smeT genes [bib_ref] Analysis of sequence variation among sme DEF multi drug efflux pump genes..., Gould [/bib_ref] , ERIC-PCR [bib_ref] Stenotrophomonas maltophilia : antimicrobial resistance and molecular typing of an emerging pathogen..., Gylmez [/bib_ref] , and PFGE (pulsed-field gel electrophoresis) analysis of Xba I genomic digests [bib_ref] Stenotrophomonas maltophilia : antimicrobial resistance and molecular typing of an emerging pathogen..., Gylmez [/bib_ref] [bib_ref] Molecular typing of Stenotrophomonas (Xanthomonas) maltophilia by DNA macrorestriction analysis and random..., Yao [/bib_ref] [bib_ref] Molecular epidemiology of Stenotrophomonas maltophilia in a university hospital, Crispino [/bib_ref] [bib_ref] High genetic diversity among Stenotrophomonas maltophilia strains despite their originating at a..., Valdezate [/bib_ref].
Genome-wide analyses showed that in many bacterial genomes short DNA segments are amplified in tandem at specific chromosomal loci http://minisatellites.u-psud.fr. Changes in the number of repeats among isolates can be monitored by PCR, and MLVA (Multi locus variable number of tandem repeat analysis) surveys are widely used for subtyping purposes [bib_ref] Utility of fast mycobacterial interspersed repetitive unit-variable number tandem repeat genotyping in..., Allix [/bib_ref] [bib_ref] Multiple-locus variable-number tandem-repeats analysis of Salmonella enterica subsp. Enterica serovar Typhimurium using..., Lindstedt [/bib_ref] [bib_ref] Multiple-locus variable-number tandem repeat analysis of Dutch Bordetella pertussis strains reveals rapid..., Schouls [/bib_ref] [bib_ref] Evaluation and selection of tandem repeat loci for Streptococcus pneumoniae MLVA strain..., Koeck [/bib_ref] [bib_ref] Use of variablenumber tandem repeats to examine genetic diversity of Neisseria meningitidis, Yazdankhah [/bib_ref] [bib_ref] Multiple-locus variable-number tandem repeat analysis for typing of Staphylococcus epidermidis, Johansson [/bib_ref] [bib_ref] Multiple-locus variable-number tandem repeat analysis of Salmonella Enteritidis isolates from human and..., Cho [/bib_ref].
The sequence of the genome of the S. maltophilia strain K279a has been completed [bib_ref] The complete genome, comparative and functional analysis of Stenotrophomonas maltophilia reveals an..., Crossman [/bib_ref]. Genome inspection allowed us to set up a simple, fast and accurate PCR-based diagnostic protocol which relies on the measurement of length heterogeneity of specific intergenic regions of the S. maltophilia genome.
The present protocol allows typing several S. maltophilia isolates in hours. The assignment of a digit code to each isolate could be used to easily compare data among different laboratories.
# Results
## Specific sequence repeats punctuate the genome of s. maltophilia
The whole DNA sequence of the S. maltophilia strain K279a has been determined [bib_ref] The complete genome, comparative and functional analysis of Stenotrophomonas maltophilia reveals an..., Crossman [/bib_ref]. The genome is 4,851,126 bp in length, and has an average G+C content of 66.3%. We found that the K279a chromosome hosts an abundant family of small, palindromic repeats fitting the consensus GTAGTGCCGGCCGCTGGCCGGCA (complementary residues are underlined) that we called SMAG (for Stenotrophomonas MAltophilia GTAG) because they carry the tetranucleotide GTAG at one terminus, similarly to small repetitive extragenic palindromic sequences (REPs) identified in the genomes of Escherichia coli and other microrganisms [bib_ref] Repetitive extragenic palindromic sequences in the Pseudomonas syringae pv. tomato DC3000 genome:..., Tobes [/bib_ref]. SMAGs make up approximately 0.5% of the K279a genome, and are spread throughout the chromosome either as single units, or in pairs, separated by 5-80 bp long spacers. The size of the SMAG family allows to hypothesize that some of these repeats may function as regulatory signals either at the DNA or the RNA level, as shown for REPs [bib_ref] Repetitive extragenic palindromic sequences in the Pseudomonas syringae pv. tomato DC3000 genome:..., Tobes [/bib_ref].
## Smags and the pcr-based genotyping of s. maltophilia isolates
In the K279a chromosome, monomeric and dimeric SMAGs are reiterated in tandem at multiple chromosomal loci, along with tracts of variable length of flanking DNA. We exploited the occurrence of SMAG arrays to set up PCR-based typing protocols, and focused our attention on 12 such loci, labelled I to XII in accord to their location on the K279a chromosome [fig_ref] Table 2: Loci analyzed by PCR in S [/fig_ref]. No rule in the pattern of amplification of SMAG sequences at the different loci could be discerned . Thus, region I features 50 bp long repeats, resulting from the duplication of a monomeric 24 bp long SMAG along with 26 bp of flanking DNA. In contrast, region XII features repeats which are 103 bp in length, and results from the duplication of a dimeric 72 bp SMAG and 31 bp of flanking DNA. In pilot experiments, regions IX and X were amplified by PCR from the DNA of the control strain K279a and five different S. maltophilia isolates. As shown in , the size of the SMAG-positive regions varies, and this correlates with changes in the number of repeating units as confirmed by sequence analysis. Isolates could thus be marked by a digit corresponding to the number of SMAG repetitions present at a given locus . Prompted by these results, we monitored the twelve SMAG-positive loci by PCR. Analyses were carried out on DNAs derived from 38 S. maltophilia strains, including the K279a strain, isolated from different sources [fig_ref] Table 1: Source and origin of the S [/fig_ref]. On the basis of the amplimers size, it had been possible to deduce the number of repeats present at the loci in the various isolates, and therefore label each of them with a digit [fig_ref] Table 3: PCR analysis of SMAG+ loci in S [/fig_ref]. In some instances, we could not detect an amplification product for one or more loci in different isolates. PCR-negative regions were labelled 0. To confirm our findings, alternative primers were used for some of these regions, but no reliable amplification product could be detected. The lack of amplification may reflect either an extensive polymorphism or deletions occurred in the regions analyzed.
A few PCR products, derived from the amplification of regions VII, XI and XII, were slightly different in length, and could not be assigned to a size class. In these instances, amplimers were assigned to the nearest size class, and marked with the letters a, b and c to denote size differences among them [fig_ref] Table 3: PCR analysis of SMAG+ loci in S [/fig_ref]. To clarify this issue, we determined the sequence of the PCR products derived by amplification of region VI in strains 915, 1029 and LMG959 (classified in [fig_ref] Table 3: PCR analysis of SMAG+ loci in S [/fig_ref] as 1a, 1b and 1c, respectively). In the control strain K279a, region VI contains four SMAG dimers. In the 915 strain, the amplified DNA (1a amplimer) was similar to the K279a interval, but only one SMAG dimer was present. In the LMG959 strain, the region amplified (1c amplimer) was 45 bp shorter, because the SMAG dimer was replaced by a SMAG monomer. In strain 1029, the size change of region VI (1b amplimer) was due to replacement of the SMAG dimer and 10 bp flanking sequence by a 66 bp palindromic element, that is a member of a distinct, less abundant family of GTAG+ repeats in S. maltophilia. Thus, size variations of loci analyzed may correlate with recombinational events which replace SMAGs with members of the same family, or related DNA families.
In order to assess the stability of the SMAG-positive regions, the strains 528, 916 and 1039 were sub-cultured for 5 days, and the DNA extracted from single colonies of each strain was analyzed by PCR (data not shown). No changes in the pattern of amplification at loci III, IV, IX and XII were observed. [fig_ref] Table 3: PCR analysis of SMAG+ loci in S [/fig_ref] , the information derived from the survey of loci II, V, I and VII was sufficient to obtain a 4digit code, that assigned the 38 DNAs analyzed to 23 different PTs (PCR Types). Some strains, such as 528 and 571, belong to the same PT type, and exhibited the same PCR profile at all the other loci. The same holds true for the three PT type 14 strains 916, 1019 and 1053, and for 3 out of 7 of the PT-19 strains. In contrast, strains 714 an 262 belong to different PTs (17 and 19, respectively), but, aside from differences in region VII, were identical at all loci.
## As shown in
The finding that a comparatively relatively robust typing can be achieved by analysing only the four loci II, V, I and VII, highlighted in [fig_ref] Table 3: PCR analysis of SMAG+ loci in S [/fig_ref] , is relevant, mostly in view of the fact that they can be co-amplified in pairs. In both instances, the amplimers corresponding to either locus can be easily distinguished because of their size range, allowing an immediate typing .
## Pfge-typing of s. maltophilia isolates
PFGE is the gold standard for strain genotyping also for S. maltophilia. PCR data shown in [fig_ref] Table 3: PCR analysis of SMAG+ loci in S [/fig_ref] partially complemented genotyping data obtained by PFGE. OBGTC9 and OBGTC10 strains exhibited the same PCR profile at all the SMAG-positive regions analyzed, and their PFGE profiles are undistinguishable . Strains 916, 1019 and 1053, which belong to PFGE A-type, and strains 528 and 571, which belong to PFGE B-type (data not shown; for the PFGE relatedness of these strains, see [bib_ref] Molecular epidemiology of Stenotrophomonas maltophilia in a university hospital, Crispino [/bib_ref] similarly exhibited the same PCR profile at all the loci [fig_ref] Table 3: PCR analysis of SMAG+ loci in S [/fig_ref]. Other strains exhibited similar PCR profiles, but different PFGE patterns. XbaI digests of OBGTC13, OBGTC23 and OBGTC30 DNAs produced undistinguishable PFGE patterns. Their PCR profiles were similar on the whole, but OBGTC23 featured a PCR type different from OBGTC13 and OBGTC30 (PT-10 vs PT-13), and the three isolates differed from each other at multiple additional loci. Moreover, K279a DNA was identical to OBGTC9 and OBGTC10 DNAs at all SMAG loci, but clearly differed from both DNAs when analyzed by PFGE .
# Discussion
The MLVA technique involves amplification and size analysis of polymorphic DNA regions containing variable numbers of tandemly repeated sequences, and is an established method to classify isolates of microbial species for which complete genome information is available. The determination of the complete genome sequence of the S. maltophilia K279a strain allowed us to set up a MLVA assay also for this organism.
All the approaches so far utilized for the genotyping of S. maltophilia clinical isolates have led to the conclusion that S. maltophilia strains are highly heterogeneous [bib_ref] Genomic diversity of the genus Stenotrophomonas, Hauben [/bib_ref] [bib_ref] Vandamme P: Identification of genomic groups in the genus Stenotrophomonas using gyrB..., Coenye [/bib_ref] [bib_ref] Analysis of sequence variation among sme DEF multi drug efflux pump genes..., Gould [/bib_ref] [bib_ref] Stenotrophomonas maltophilia : antimicrobial resistance and molecular typing of an emerging pathogen..., Gylmez [/bib_ref] [bib_ref] Molecular typing of Stenotrophomonas (Xanthomonas) maltophilia by DNA macrorestriction analysis and random..., Yao [/bib_ref] [bib_ref] Molecular epidemiology of Stenotrophomonas maltophilia in a university hospital, Crispino [/bib_ref] [bib_ref] High genetic diversity among Stenotrophomonas maltophilia strains despite their originating at a..., Valdezate [/bib_ref].
According to AFLP fingerprinting [bib_ref] Genomic diversity of the genus Stenotrophomonas, Hauben [/bib_ref] and gyrB RFLP [bib_ref] Vandamme P: Identification of genomic groups in the genus Stenotrophomonas using gyrB..., Coenye [/bib_ref] analysis, S. maltophilia can be assigned to 8-10 genomic groups. The majority of CF isolates grouped in two clusters [bib_ref] Vandamme P: Identification of genomic groups in the genus Stenotrophomonas using gyrB..., Coenye [/bib_ref] , suggesting that isolates of specific groups have an increased potential for the colonization of the respiratory tract of CF patients. Considering the interest paid to S. maltophilia as an emerging opportunistic pathogen associated with nosocomial infections, we wanted to develop a fast, accurate and unexpensive method of genotyping which could be adopted for strain classification and comparisons. The chromosomal regions carrying clusters of SMAGs, an abundant sequence repeat spread in the S. maltophilia genome, vary in size among isolates, allowing MLVA-based typing surveys. The repeat units found at the various loci analyzed are sufficiently large to discriminate length variation among isolates by low molecular weight agarose gel electrophoresis. In most instances, PCR data The coordinates of the loci on the genome of the K279a strain, the forward (f) and reverse (r) primers in the 5'-3' orientation, their 5' end position, and the annealing temperatures (Ta) are shown. enabled to set a direct correlation between the length of the amplimers and the number of SMAG repeats present at the analyzed loci. The latter could be directly marked by the number of repeats, setting the basis for a simple, numerical classification of the strains analysed. Broad sorting of a large number of isolates may suggest to restrict MLVA analyses to a few SMAG-positive loci. Allelic variants of the four loci II, V, I and VII [fig_ref] Table 3: PCR analysis of SMAG+ loci in S [/fig_ref] are detectable by two PCR co-amplifications reaction, providing 4-digit typing profiles which could turn out to be effective for simple typing purposes. By increasing the number of loci analyzed, it could be possible to obtain a more discriminating digit profile, as in MLST analyses [bib_ref] Multi locus sequence typing of bacteria, Maiden [/bib_ref].
According to our typing scheme, regions not responding to the PCR approach have been assigned the digit 0. While uninformative on the integrity of the region under scrutiny, the lack of amplification of a certain locus is fully exploitable in a multi-typing system. A "caveat" may be represented by cases in which it is not possible to immediately correlate the size of the amplicon with the number of repeats as observed for amplimers slightly differing in length occasionally found at loci VII, XI and XII. However, this could represent a problem in the analyses of large populations of isolates, calling for a highly discriminating profiling. Minor size differences among amplicons allow discrimination for typing purposes as the major ones, and [fig_ref] Table 1: Source and origin of the S [/fig_ref] can be indicated by marking amplimers also with letters. However, accurate measuring of small size differences among amplimers which have been analyzed in different electrophoretic runs may be cumbersome. Thus, it would be advisable to assign to amplicons which may slightly differ in size the same digit, which should correspond to the most likely number of repeats present. This may ensure to rapidly proceed in classifying the different isolates, eventually further distinguishing them by means of additional analyses.
MLVA assays provide results that parallel PFGE data, although some differences have been noticed, since MLVA and PFGE measure different types of chromosomal modifications, and, for example, recombination events within a genome could be detected by PFGE, but overlooked by MLVA analyses [bib_ref] Multiple-locus variable-number tandem repeat analysis for typing of Staphylococcus epidermidis, Johansson [/bib_ref] [bib_ref] Comparison of multiple-locus variable-number tandemrepeat analysis with pulsed-field gel electrophoresis, spa typing,..., Malachowa [/bib_ref] [bib_ref] Multiple-locus variable-number tandem-repeat assay analysis of methicillin-resistant Staphylococcus aureus strains, Tenover [/bib_ref]. The same holds true in our study, as strains such as K279a, OBGTC9 and OBGTC10, while identical according to MLVA data, differ when analyzed by PFGE. The method we have devised is simpler, less time-consuming and economically more advanta-geous than PFGE. As suggested by Tenover et al. [bib_ref] Multiple-locus variable-number tandem-repeat assay analysis of methicillin-resistant Staphylococcus aureus strains, Tenover [/bib_ref] , MLVA approaches could be particularly helpful to identify strains responsible for outbreaks in hospital settings, and to determine the relatedness of isolates collected over short periods of time. In contrast, PFGE could be priviliged for long time period analyses of bacterial populations.
# Conclusion
The utilization of the present protocol will be useful for fast and efficient typing purposes. Several S. maltophilia isolates could be typed in hours, and the results interpreted de visu without the need for sophisticated software. Data would be easily reproducible, and immediately comparable among different laboratories.
# Methods
## S. maltophilia strains
S. maltophilia strains analyzed in this study are listed in [fig_ref] Table 1: Source and origin of the S [/fig_ref]. Clinical isolates were identified as S. maltophilia by using the VITEK II system (bioMerieux, Morey-l'Etoile, France). The identification was confirmed by PCR amplification and sequence analysis of the 16S rDNA. Strains were routinely grown in brain heart infusion at 37°C, except for the environmental strains LMG959, LMG10871, LMG10879 and OBGN1 which were grown at 30°C. In order to analyze the stability of the genomic regions under scrutiny, a few strains were sub-cultured in brain heart infusion at 37°C 5 times for 18-24 hrs.
## Pcr amplification
The DNA of single colonies derived from the final subcultures was analyzed by PCR amplification of DNA regions of interest. Genomic DNA was extracted as described by De Gregorio et al. [bib_ref] Di Nocera PP: Structural organization and functional properties of miniature DNA insertion..., Gregorio [/bib_ref]. PCR reactions were carried out by incubating 20 ng of DNA with 160 ng of each primer in the presence of dXTPs (200 nanomoles), 1.5mM magnesium chloride and the Taq DNA polymerase Recombinant (Invitrogen). Because of the high GC content of the S. maltophilia genome (> 66%), all PCR reactions were carried out in GC-rich buffer (Roche). The oligomers used as primers, and the annealing temperatures, are listed in [fig_ref] Table 2: Loci analyzed by PCR in S [/fig_ref]. Samples were incubated at 95°C for 5', and subsequently for 1' at 95°C, 1' at the annealing temperature and 1' at 72°C, for a total of 30 cycles. At the end of the cycle, samples were kept at 72°C for 7' before harvesting.
PCR products were electrophoresed on 1.5-2% agarose gels in 0.5×TBE buffer (45 mM Tris pH 8, 45 mM Borate, 0.5 mM EDTA) at 120 V (constant voltage). The 100 bp ladder (Fermentas) was used as molecular weight marker.
# Pfge analysis
Preparation of agarose plugs containing chromosomal DNA for PFGE analysis was performed using the PulseNet Co-amplification of SMAG-positive loci Co-amplification of SMAG-positive loci. Amplimers deriving from the dual amplification of II and V, and I and VII loci from the DNA of the indicated strains were analyzed by elctrophoresis as in .
standardized procedure http://www.cdc.gov/pulsenet The DNA plugs were digested with 60 U of XbaI (Roche Diagnostics) at 37°C fo 16 h. Genomic DNA fragments were separated by PFGE at 14°C on agarose 1.2% w/v gels in a clamped homogeneous field electrophoresis apparatus (CHEF-DRII system; Bio-Rad, Hemel Hempstead, UK), with pulse times ramped from 1 to 20 s over 21 h at 6.0 V/ cm in 0.5× TBE. DNA fragments obtained from XbaI digestion of plugs containing chromosomal DNA of Salmonella braenderup strain H9B12 were used as molecular weight markers [bib_ref] Establishment of a universal size standard strain for use with the PulseNet..., Hunter [/bib_ref].
Electrophoretic patterns were analyzed by UPGMA (Unweighted Pair Group Method with Arithmetic mean) using the Gel Compar II version 4.5 software (Applied Maths).
[table] Table 1: Source and origin of the S. maltophilia strains analyzed in this study [/table]
[table] Table 2: Loci analyzed by PCR in S. maltophilia strains [/table]
[table] Table 3: PCR analysis of SMAG+ loci in S. maltophilia strains. [/table]
|
SAMD9 and SAMD9L in inherited predisposition to ataxia, pancytopenia, and myeloid malignancies
Germline mutations in the SAMD9 and SAMD9L genes, located in tandem on chromosome 7, are associated with a clinical spectrum of disorders including the MIRAGE syndrome, ataxia-pancytopenia syndrome and myelodysplasia and leukemia syndrome with monosomy 7 syndrome. Germline gain-of-function mutations increase SAMD9 or SAMD9L's normal antiproliferative effect. This causes pancytopenia and generally restricted growth and/or specific organ hypoplasia in nonhematopoietic tissues. In blood cells, additional somatic aberrations that reverse the germline mutation's effect, and give rise to the clonal expansion of cells with reduced or no antiproliferative effect of SAMD9 or SAMD9L include complete or partial chromosome 7 loss or loss-of-function mutations in SAMD9 or SAMD9L. Furthermore, the complete or partial loss of chromosome 7q may cause myelodysplastic syndrome in these patients. SAMD9 mutations appear to associate with a more severe disease phenotype, including intrauterine growth restriction, developmental delay and hypoplasia of adrenal glands, testes, ovaries or thymus, and most reported patients died in infancy or early childhood due to infections, anemia and/or hemorrhages. SAMD9L mutations have been reported in a few families with balance problems and nystagmus due to cerebellar atrophy, and may lead to similar hematological disease as seen in SAMD9 mutation carriers, from early childhood to adult years. We review the clinical features of these syndromes, discuss the underlying biology, and interpret the genetic findings in some of the affected family members. We provide expert-based recommendations regarding diagnosis, followup, and treatment of mutation carriers.
# Introduction
Aberrations of chromosome 7 in myeloid malignancies was first described in the 1960's, but the exact gene(s) in this region driving myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) development remain elusive [bib_ref] /7q-syndrome in myeloidlineage hematopoietic malignancies: attempts to understand this complex disease entity, Honda [/bib_ref]. Besides total chromosomal loss (monosomy 7) or loss of the long arm of chromosome 7 (7q-), different commonly deleted regions (CDR) have been identified, including 7q22 (CDR1), 7q34 (CDR2), and 7q35-36 (CDR3) [bib_ref] Identification of a common microdeletion cluster in 7q21.3 subband among patients with..., Asou [/bib_ref] [bib_ref] Loss of heterozygosity in 7q myeloid disorders: clinical associations and genomic pathogenesis, Jerez [/bib_ref] [bib_ref] Recurrent genetic defects on chromosome 7q in myeloid neoplasms, Hosono [/bib_ref]. Interestingly, aberrations of chromosome 7 are found in 30-40% of pediatric MDS patients [bib_ref] Childhood myelodysplastic syndrome in Denmark: incidence and predisposing conditions, Hasle [/bib_ref] [bib_ref] Myelodysplastic syndrome, juvenile myelomonocytic leukemia, and acute myeloid leukemia associated with complete..., Hasle [/bib_ref] , but only in 10% of adult MDS, for unknown reasons [bib_ref] Cytogenetic features in myelodysplastic syndromes, Haase [/bib_ref]. Despite decades of research striving to identify the gene(s) on 7q contributing to myeloid malignancies, very few mutations in the remaining haploinsufficient alleles have been found: in essence, only recurrent mutations in CUX1 [bib_ref] Recurrent genetic defects on chromosome 7q in myeloid neoplasms, Hosono [/bib_ref] [bib_ref] Mutation analysis of CUTL1 in childhood myeloid neoplasias with monosomy 7, Hindersin [/bib_ref] and EZH2 [bib_ref] Somatic mutations of the histone methyltransferase gene EZH2 in myelodysplastic syndromes, Nikoloski [/bib_ref] have been identified. This indicates that haploinsufficiency for genes located in the CDR1-3 is the principal mechanism driving MDS development in these cases.
For the myelodysplasia and leukemia syndrome with monosomy 7 (MLSM7; OMIM #252270), the presence of a sibling pair both diagnosed with a myeloid neoplasm displaying monosomy 7 is a prerequisite for diagnosis [bib_ref] Myelodysplasia and leukemia syndrome with monosomy 7: a genetic perspective, Gilchrist [/bib_ref]. No candidate gene(s) has been ascertained for MLSM7 and the disease does not follow a clear monogenic pattern [bib_ref] Two autopsy cases of atypical myeloproliferative disorder with group C monosomy occurring..., Kamiyama [/bib_ref]. It has even been suggested that the predisposing locus is not located on chromosome 7 due to retainment of different parental chromosomes 7 between siblings with leukemic bone marrows [bib_ref] Familial bone marrow monosomy 7. Evidence that the predisposing locus is not..., Shannon [/bib_ref].
The ataxia-pancytopenia (ATXPC) syndrome (OMIM #159550) was first described in 1978 by Dr. Frederik Li in a family with neurologic symptoms and pancytopenia of unknown origin, with some family members developing a myeloid neoplasm with monosomy 7 [bib_ref] A family with acute leukemia, hypoplastic anemia and cerebellar ataxia: association with..., Li [/bib_ref] , Exclusion and linkage analyses combined with whole exome sequencing (WES) of an ATXPC kindred revealed mutations in the gene SAMD9L as the cause of this syndrome, and mutations in this gene were also found in the original family described by Dr. Li [bib_ref] Ataxia-pancytopenia syndrome is caused by missense mutations in SAMD9L, Chen [/bib_ref].
We were part of a collaborative team describing two families-from Sweden and Finland, respectively-with SAMD9L mutations and neurological symptoms and fluctuating cytopenias, infections, and MDS [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref] [bib_ref] Ataxia-pancytopenia syndrome with SAMD9L mutations, Gorcenco [/bib_ref]. In light of these recent findings, it is unlikely that ATXPC and MLSM7 really represent two distinct entities, but merely reflect a variation in the penetrance and severity of symptoms in one and the same disease spectrum that includes germ line mutations in e.g., GATA2, SAMD9L, and SAMD9, leading to monosomy 7. In this review, we summarize the disease phenotype in SAMD9L and SAMD9 mutation carriers, and we attempt to provide recommendations for the identification, initial screening, follow-up and treatment of patients with, based on the limited data currently available about these disorders.
## Samd9l
Function of SAMD9L SAMD9L, a protein widely expressed across human tissues, contains a SAM domain that binds RNA as well as oligomerizes both with SAM-containing and non-SAMcontaining proteins. The exact role of SAMD9L is currently unknown, but the gene has a general antiproliferative function and has been demonstrated to function as tumor suppressor in breast, hepatocellular and in squamous cell carcinoma, being repressed by the p53 pathway [bib_ref] SAMD9L-related ataxia-pancytopenia syndrome, Phowthongkum [/bib_ref]. In hematopoetic tissue SAMD9L regulates cell proliferation by being a crucial component of a protein complex that facilitates the degradation of cytokine receptors through the homotypic fusion of endosomes [bib_ref] Haploinsufficiency of SAMD9L, an endosome fusion facilitator, causes myeloid malignancies in mice..., Nagamachi [/bib_ref].
## Heterozygous germline mutations in samd9l in atxpc syndrome
To date, four families with the ATXPC syndrome and SAMD9L mutations have been described [bib_ref] Ataxia-pancytopenia syndrome is caused by missense mutations in SAMD9L, Chen [/bib_ref] [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref] [bib_ref] Ataxia-pancytopenia syndrome with SAMD9L mutations, Gorcenco [/bib_ref] [bib_ref] SAMD9L-related ataxia-pancytopenia syndrome, Phowthongkum [/bib_ref]. In total, [bib_ref] Human sterile alpha motif domain 9, a novel gene identified as down-regulated..., Li [/bib_ref] in SAMD9L result in a gain-of-function (GOF) [bib_ref] Ataxia-pancytopenia syndrome is caused by missense mutations in SAMD9L, Chen [/bib_ref] , not a loss-of-function (LOF) and haploinsufficiency. In the latter theory, LOF/haploinsufficiency for SAMD9L in ATXPC would be the first hit towards leukemic transformation, followed by monosomy 7 or 7q-that includes other genes involved in the pathogenesis of MDS. This idea was however proven wrong by functional assays ascertaining that the germline mutations actually are GOF aberrations increasing the antiproliferative effect of SAMD9L [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref].
SAMD9L mutations are lost due to aberrations of chromosome 7 during progression towards MDS
The SAMD9L gene is located on chromosome 7q21.2 in the MDS CDR1 [bib_ref] Haploinsufficiency of SAMD9L, an endosome fusion facilitator, causes myeloid malignancies in mice..., Nagamachi [/bib_ref]. In the three patients identified to date with inherited GOF SAMD9L mutations who developed MDS, the mutant allele was lost due to aberrations of chromosome 7 [bib_ref] Ataxia-pancytopenia syndrome is caused by missense mutations in SAMD9L, Chen [/bib_ref] [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref]. Importantly, a clear correlation between metaphases carrying the aberration of chromosome 7 and the change in the ratio between wild type (wt) and mutant SAMD9L ratio could be observed with disease progression [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref]. One patient developed a der(1;7) which resulted in loss of genetic material on chromosome 7, including the mutant SAMD9L allele [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref]. The other two patients that developed MDS lost the mutant SAMD9L copy by monosomy 7 and deletion of 7q, respectively [bib_ref] Ataxia-pancytopenia syndrome is caused by missense mutations in SAMD9L, Chen [/bib_ref] [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref]. Again, we could clearly show a correlation between the metaphases carrying the aberration of chromosome 7 and the change in the variant allele frequency (VAF) towards the wt allele [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref].
## Reversion of the germline samd9l mutation
Sequencing revealed a mutant SAMD9L VAF lower than the expected 50% in the blood in some of the unaffected family members. Furthermore, SNP array B-allele frequencies demonstrated clear skewing towards the wt allele, without net genomic loss in the region on chromosome 7q21.2. This implies that the low mutant VAF was due to a copy neutral loss of heterozygosity (CN-LOH) of chromosome 7q [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref]. Chen et al. also observed decreased frequency of the missense mutant SAMD9L in two of their study subjects and suspected hematopoietic mosaicism. Long-term in vitro culturing of transformed lymphocytic lines from heterozygous carriers demonstrated that CN-LOH of 7q, resulting in net loss of mutant allele, occurred after 3 weeks to 6 months in culture. This is most likely due to the selective growth advantage attributed to hematopoietic precursor cells that have lost their mutated copy and duplicated the wt counterpart of SAMD9L by uniparental disomy (UPD) [bib_ref] Ataxia-pancytopenia syndrome is caused by missense mutations in SAMD9L, Chen [/bib_ref]. Moreover, in some individuals the GOF mutation was retained but instead counteracted by frameshift/stop-gain mutations in SAMD9L in cis [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref]. Normally frameshift/stop-gain mutations would result in destruction of the mRNA from the mutant allele by nonsense-mediated decay but this requires that the mutation occurs in a defined distance to intron/exon boundaries [bib_ref] NMD: a multifaceted response to premature translational termination, Kervestin [/bib_ref]. In the case of SAMD9L, functionally being a one exon gene, nonsense-mediated decay is thus likely not responsible for any LOF. As an additional mechanism of reversion, we identified in cis mutations in the same codon that carried the GOF mutation . These resulted in amino acid changes probably inducing less GOF, a hypothesis supported by proliferative effects seen in cells positive for in cis mutations when overexpressing the different SAMD9L variants in 293T cells [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref]. Strikingly, some patients carried multiple clones with different genetic mechanisms of reversion and in one case a patient even carried a reversion by CN-LOH and additionally an MDS clone with aberrations of chromosome 7 [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref]. These findings show the tremendous selection pressure of the cells to neutralize the GOF SAMD9L mutation and the advantage inferred to the cells that have undergone genetic reversion . Family members that did not develop MDS all showed genetic reversions in some form or another, which is relevant in regard to the diagnostic criteria that at least two siblings have to be affected by MDS with aberration of chromosome 7 to make the diagnosis MLSM7. It appears likely that the common occurrence of genetic reversion has led to underdiagnosis of MLSM7. It can also be speculated, as in the study by Shannon et al. that investigated three families with MLSM7 and found retainment of mixed parental alleles in leukemic bone marrow between siblings, actually reflected mosaicism of clones harboring mutational reversal events and 7q-, respectively [bib_ref] Familial bone marrow monosomy 7. Evidence that the predisposing locus is not..., Shannon [/bib_ref].
The origin of SAMD9L mutations in hematopoietic hierarchy and development SAMD9L is most highly expressed in NK cells and monocytes and subsequently symptomatic carriers usually present with low numbers of NK cells and monocytes [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref]. To investigate the effect of the genetic reversion in mature cells; B-, T-, NK cells, monocytes, and granulocytes were isolated from two individuals affected by either CN-LOH or reversion mutations. Detection of allelic ratios of the wt and antiproliferative effect may underly the general intrauterine growth reduction or organspecific hypoplasias or atrophies observed in these syndromes. In SAMD9L-associated disease the non-hematological disease manifestations may be mild and patients may not report symptoms, but some degree of nervous system involvement was noted almost all ATXPC patients mutant SAMD9L showed that NK cells showed a higher percentage of gene reversion, indirectly arguing for the essential gene function of SAMD9L in NK-cell differentiation [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref].
To investigate when the reversion of the inherited GOF SAMD9L mutations occurred, mutant VAF in peripheral blood (PB) at birth (using Guthrie cards) was investigated. Equal fractions of wt and mutant SAMD9L in PB directly after birth indicated that the low mutant SAMD9L VAF detected later in life was not caused by mosaicism at birth [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref] , but instead by a clonal genetic event that occurred most likely between 1 and 4 years of age. This also coincided with a time when many individuals were experiencing a transient pancytopenia, probably undergoing genetic reversion leading to a fitness advantage of the reverted cells, resulting in their expansion and recovery of the PB counts.
Cell sorting of hematopoietic stem cells (HSC), multipotent progenitors (MPP), common myeloid progenitors, common lymphoid progenitors and granulocyte-monocyte progenitors followed by analysis of the mutant VAF using digital droplet PCR was performed to investigate at which cellular level the reversion has occurred. Surprisingly, this analysis suggested that the molecular reversion occurred at the MPP rather than at the HSC level [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref] , although more patients would have to be studied to draw more definitive and generalized conclusions. We hypothesize that the reversion of the inherited SAMD9L mutation might result in increased proliferation of MPPs (at least when compared to mutant cells), leading to clonal expansion. Alternatively, a rare HSC clone would represent the initial revertant, with the correction leading to a selective recruitment of these HSCs into differentiation but with no evident advantage in terms of self-renewal. In both cases, MPPs must associate with a high maintenance capacity. While most hematopoietic homeostasis has previously been attributed to HSCs, mainly based on transplantation experiments, such scenarios appear in line with more recent lineage-tracing studies and selective HSC-depletion strategies in mice, demonstrating substantial long-term contribution of MPPs not only in steady-state but also the context of stress hematopoiesis [bib_ref] Fundamental properties of unperturbed haematopoiesis from stem cells in vivo, Busch [/bib_ref] [bib_ref] Clonal dynamics of native haematopoiesis, Sun [/bib_ref] [bib_ref] The bulk of the hematopoietic stem cell population is dispensable for murine..., Schoedel [/bib_ref] [bib_ref] Mitotic history reveals distinct stem cell populations and their contributions to hematopoiesis, Säwén [/bib_ref].
## Samd9l knock-out mice develop mds
It has been suggested that monosomy 7, 7q-, or 7q21 microdeletions increase the signaling through the cytokine [bib_ref] Haploinsufficiency of SAMD9L, an endosome fusion facilitator, causes myeloid malignancies in mice..., Nagamachi [/bib_ref]. This observation probably explains the advantage of Samd9l −/− and Samd9l +/− cells in the setting of a competitive transplantation, even though the advantage is rather small and is probably enhanced by the competition in the posttransplantation setting and/or by infections post transplantation that trigger interferon associated Samd9l expression. The observation that cells with reduced or absent levels of SAMD9L have a growth advantage compared to wt cells fits well with the observation that GOF SAMD9L mutations are leading to growth inhibition [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref].
Neurological phenotype of the ATXPC syndrome Some degree of nervous system involvement has been documented in almost all carriers of pathogenic SAMD9L mutations reported to date . Balance problems and nystagmus are most common, followed by mild pyramidal signs [bib_ref] Ataxia-pancytopenia syndrome with SAMD9L mutations, Gorcenco [/bib_ref]. A few patients developed mild cerebellar-type dysarthria. Other cerebellar motor signs such as impairment of rapid alternating movements, dysmetria, cerebellar tremor were typically less pronounced than the balance problems, or entirely absent, why the usage of the term "ataxia" is actually debatable [bib_ref] Ataxia-pancytopenia syndrome with SAMD9L mutations, Gorcenco [/bib_ref]. Moreover, we have identified paracentral retinal dysfunction in two mutation carriers and additional reported patients had clinical signs and symptoms highly suggestive of this. We therefore suggest that part of the dysmetria reported in the other families could be motor problems caused by limited visual control. Behavioral symptoms have been described in carriers of specific SADM9L mutations, especially during childhood and adolescence.
Magnetic resonance imaging of individuals with ATXPC revealed cerebellar atrophy in all patients and white matter hyperintensities in some. Cerebellar atrophy started during early childhood and was slowly progressive. The cerebellar atrophy is surprisingly pronounced compared to the generally mild balance impairment. Neuropathological examination of four individuals with ATXPC has been reported and brains showed pronounced loss of cerebellar Purkinje cells [bib_ref] A family with acute leukemia, hypoplastic anemia and cerebellar ataxia: association with..., Li [/bib_ref] [bib_ref] Ataxia-pancytopenia: syndrome of cerebellar ataxia, hypoplastic anemia, monosomy 7, and acute myelogenous..., Li [/bib_ref] [bib_ref] Ataxia-pancytopenia syndrome is caused by missense mutations in SAMD9L, Chen [/bib_ref]. Granule cell loss was found in three patients but was less pronounced. There was depletion of white matter in the cerebellum in two individuals [bib_ref] A family with acute leukemia, hypoplastic anemia and cerebellar ataxia: association with..., Li [/bib_ref] [bib_ref] Ataxia-pancytopenia: syndrome of cerebellar ataxia, hypoplastic anemia, monosomy 7, and acute myelogenous..., Li [/bib_ref] and in the hippocampus in one patient [bib_ref] Ataxia-pancytopenia syndrome is caused by missense mutations in SAMD9L, Chen [/bib_ref]. While the cerebellar atrophy can be readily explained by loss of Purkinje cells, cerebellar white matter and granule cells, these patients have been autopsied at ages when there commonly are no or no marked white matter hyperintensities in MRI. Thus, the exact neuropathological correlate of the white matter hyperintensities remains unknown, but we hypothesize that SAMD9L mutations may slow down the physiological myelination that physiologically occurs in childhood and adolescence.
## Samd9
Normal function of SAMD9 SAMD9, a highly conserved gene contigous to SAMD9L, shares a common gene structure to its paralogue and both genes encode proteins with 58% amino acid identity. Most mammals have both Samd9 and Samd9l. However, mice have only Samd9l, and cows encode only Samd9 suggesting that the genes partially complement each other's functions [bib_ref] Human sterile alpha motif domain 9, a novel gene identified as down-regulated..., Li [/bib_ref]. SAMD9, like SAMD9L, is involved in control of cell proliferation and functions as a tumor suppressor in some cancers. Deleterious mutations in the SAMD9 gene are known to cause normophosphatemic familial tumoral carcinosis (NFTC), a rare autosomal recessive disease. NFTC is characterized by abnormal inflammation of the skin and gingiva that is induced by excessive signaling via interferon pathways-in line with this it has been experimentally demonstrated that the function SAMD9 is tightly regulated by interferon-γ [bib_ref] Functional characterization of SAMD9, a protein deficient in normophosphatemic familial tumoral calcinosis, Hershkovitz [/bib_ref].
## Samd9 mutations cause the mirage syndrome
Recent work has identified mutations in SAMD9 in children with myelodysplasia, infection, restriction of growth, adrenal hypoplasia, genital phenotypes, and enteropathy, named MIRAGE syndrome (OMIM #617053) (Figs. 1b and 2; [fig_ref] Table 1: Summary of clinical phenotype of reported patients with ATXPC /MLSM7 and MIRAGE [/fig_ref] [bib_ref] SAMD9 mutations cause a novel multisystem disorder, MIRAGE syndrome, and are associated..., Narumi [/bib_ref]. In the initial report the majority of the affected children died in early childhood, but two individuals with an expanded lifespan developed MDS with mononsomy 7, resulting in a loss of the mutant SAMD9 allele. A recent study also reported that in seven MIRAGE children in whom parental DNA was available, the mutations were de novo and located in highly conserved residues, often affecting arginines (6 of 8) with half of all mutations clustering in a hotspot of codons 982 and 983 of SAMD9 [bib_ref] Somatic mutations and progressive monosomy modify SAMD9-related phenotypes in humans, Buonocore [/bib_ref]. The conclusion is therefore that SAMD9 mutations underlying the MIRAGE syndrome are GOF and the negative effect on cellular growth puts a selection pressure on the cells to lose their mutated copy. In addition, this argues that the functions of these gene homologs are very similar. However, besides the similar hematologic phenotype, it remains elusive why the mutations in SAMD9L result in cerebellar atrophy while mutations in SAMD9 are associated with complex multi-organ defects. We hypothesize that possible neurological manifestations in SAMD9-associated disease may have remained unnoticed because most patients died at a very young age when the nervous system is still immature and mild balance problems or visual problems would not have been evident.
Buonocore et al. describes eight MIRAGE patients with intrauterine growth restriction and severe testicular dysfunction leading to female external genitalia in six of the eight children, while the remaining two children had atypical external genitalia (Figs. 1b and 2) [bib_ref] Somatic mutations and progressive monosomy modify SAMD9-related phenotypes in humans, Buonocore [/bib_ref]. The majority of 19 de novo MIRAGE patients identified to date have had signs of bone marrow failure (various cytopenias) and infections in their medical history. Four children in total have had a longer life span and developed MDS with aberrations of chromosome 7. Strikingly, in the three non-MDS patients in which cytogenetics were performed, hematopoietic cells showed some form of aberration of chromosome 7, either monosomy 7 or 7q-. These three patients had an initial VAF around 50% but showed a lower VAF in a later sample indicating expansion of clones that that had lost their GOF SAMD9 mutation due to monosomy 7 or 7q-. The reason that these patients did not develop MDS could be time related, since all died at 5-21 months of age. Similar to the observations made for GOF SAMD9L mutations in ATXPC, GOF SAMD9 mutations were also reverted by other genetic mechanisms such as frame-shift or stop gain mutations in the blood of four individuals with MIRAGE syndrome , including patients with nonmanifest hematopoietic phenotype [bib_ref] Somatic mutations and progressive monosomy modify SAMD9-related phenotypes in humans, Buonocore [/bib_ref].
## Inherited samd9 mutation causing a milder phenotype
Schwartz et al. have recently identified a family with an inherited SAMD9 mutation, in which three children developed MDS with monosomy 7 at an early age [bib_ref] Germline SAMD9 mutation in siblings with monosomy 7 and myelodysplastic syndrome, Schwartz [/bib_ref]. While the mother was completely unaffected, both by MDS and by genetic reversion, all three children displayed clonal somatic reversion in the blood by different mechanisms. One affected child had a CN-LOH of chromosome 7 and simultaneous monosomy 7, while the other two children had monosomy 7 as well as in cis mutations in SAMD9, arguing for neutralization of the germline GOF variant by both mechanisms.
It can only be speculated that the differences between de novo and familial SAMD9 mutations is the severity of the GOF. Weaker GOF variants are probably tolerated while more severe ones can only occur de novo due to embryonic lethality or is not fixated in the population due to infertility by the accompanying genital malformation. However, it is unclear in the reported family, why the mother did not develop any phenotype and did not show any genetic reversion. Unfortunately, RNA was not available for testing of uni-or biallelic expression of SAMD9 (J. Klco, Personal communication, May 2017). Importantly, this latest publication also provides additional evidence that different clones which have undergone genetic reversion by different mechanism, exist in the same patient and compete with each other resulting in the end in an outgrowth of the MDS clone with monosomy 7 [bib_ref] Germline SAMD9 mutation in siblings with monosomy 7 and myelodysplastic syndrome, Schwartz [/bib_ref].
## Diagnosis and follow-up
Screening for SAMD9 and SAMD9L mutations All patients with pediatric MDS with monosomy 7, del 7q and der(1;7) and probably all adult MDS patients with these chromosomal aberrations should be tested with gene panels not only covering GATA2 associated disorders, but also including SAMD9 and SAMD9L mutations on genetic material not obtained from blood cells. In our experience skin fibroblasts are the best source of germline DNA, whereas buccal swabs on some occasions have showed non representative skewing towards wt copies, likely due contamination of blood cells that have undergone reversion by CN-LOH [bib_ref] Buccal swabs but not mouthwash samples can be used to obtain pretransplant..., Thiede [/bib_ref]. Accompanying monocytopenia and low or absent NK cells shall even raise more suspicion about GATA2 deficiency syndrome, ATXPC and MIRAGE syndromes, but this should not be a prerequisite for genetic testing because many patients with ATXPC and MIRAGE have "reverted" the GOF mutation resulting in a recovery of cell numbers in the PB.
With diagnosis of MIRAGE or ATXPC syndrome in one family member, a thorough family history should be taken and members of the family should, after consent, be offered genetic counseling and testing regarding the mutation identified in the family. Counseling should include the information that they may be mutation carriers, possibly at risk for MDS or cytopenias with potentially severe consequences, even if they have not experienced any symptoms previously. In ATXPC families, relatively subtle neurological signs may indicate carrier status. We consider testing family members because they may either be affected, at risk for developing MDS to our present knowledge, or since they might represent potential HSC donors for a family member with ATXPC syndrome. It should also be noted that some of the patients in the described families were suffering from clinical symptoms associated with low cell numbers of macrophages/monocytes as well as NK cells. So the recommendation for chest X-ray to exclude macrophage deficiency related alveolar proteinosis and screening vaccination for NK-cell deficiency related HPV infection is at the moment based on our clinical experience with a very limited patient cohort. However, we still think that the benefits of these screenings outweighs their risks and would recommend them until better data is available. In addition, for every carrier of pathogenic SAMD9 or SAMD9L mutations without MDS we recommend frequent (every 6-12 months) follow-up physician visits and laboratory tests, as suggested in [fig_ref] Table 2: Recommendations for clinical care and surveillance of carriers of pathogenic mutations in... [/fig_ref] [bib_ref] How I diagnose and manage individuals at risk for inherited myeloid malignancies, Churpek [/bib_ref].
## Treatment of mds in samd9l and samd9 mutation carriers
Most difficult, regarding both syndromes, is to give recommendations on whom, when and how to perform hematopoietic stem cell transplantation (HSCT), because of a lack of evidence and extensive clinical experience. The 4year ATXPC patient in the Swedish family developed severe neurological problems after allogeneic HSCT from a matched unrelated donor, although we still remain uncertain if there was a causal relationship [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref]. Conditioning was performed according to the EWOG-MDS guidelines on HSCT in Childhood MDS and JMML, essentially using anti-thymocyte globulin, fludarabine and thiotepa for conditioning. Due to the fact that we were not aware of the definite diagnosis prior to the transplantation, a thorough neurological examination and a brain MRI were not performed during the pre-transplantation evaluation. However, MRI of the brain after the allogeneic HSCT showed pronounced cerebellar atrophy that correlated with onset of ataxia and nystagmus. This hypoplasia could be induced by a CMV infection the patient suffered during infancy, with prolonged problems due to his primary deficiency. Moreover, no worsening of neurological symptoms in the transplanted boy in the Finnish family was observed [bib_ref] Gain-of-function SAMD9L mutations cause a syndrome of cytopenia, immunodeficiency, MDS, and neurological..., Tesi [/bib_ref]. However, it could be speculated that the conditioning triggered early onset of ATXPC related cerebellar atrophy, possibly by chemotherapy inducing an interferon response and that this in combination with the cellular damage imposed by these treatments will prevent cellular repair. Whether the same would apply to cellular damage induced by irradiation is of course unknown, but we would currently advise physicians to avoid higher doses of irradiation, especially to the brain, in patients with ATXPC syndrome.
With the currently available information it is impossible to draw any evidence-based conclusions, but we are worried that the only curative treatment for patients with ATXPC that develop MDS might aggravate the neurological symptoms and will severely affect their quality of life. Unfortunately, one can argue for both an early and a late time point for allogeneic HSCT. When transplanting before transformation to MDS, non-myeloablative regimens similar to the ones used for Fanconi's anemia or GATA2 deficiency syndrome can probably be used [bib_ref] Successful allogeneic hematopoietic stem cell transplantation for GATA2 deficiency, Cuellar-Rodriguez [/bib_ref] [bib_ref] Fanconi anaemia: new strategies, Dufour [/bib_ref] , even though evidence for this assumption are missing at the moment. This idea is based on the observation that like in GATA2 deficiency syndrome, the competitiveness of HSC is greatly reduced in ATXPC and MIRAGE syndrome and monocytes and NK cells are also reduced in number and possibly function. These effects reduce the risk for rejection and confer an advantage of the incoming HSCs in competition for niches in the bone marrow. Such reduced intensity conditioning will hopefully lessen any risk of developing cerebellar atrophy associated with allogeneic HSCT. However, once transformation to MDS with monosomy 7 has occurred, most transplant physicians would recommend a myeloablative conditioning, which at least in our patient possibly could have triggered cerebellar atrophy. We are reluctant to advice against this, but a careful risk-benefit analysis should of course be made, and of note should be that non-myeloablative conditioning regimens for HSCT after transformation are now in use for Fanconi anemia [bib_ref] Sequential treatment for allogeneic hematopoietic stem cell transplantation in Fanconi anemia with..., Talbot [/bib_ref].
## Treatment of neurological symptoms in atxpc
The neurological symptoms of ATXPC are generally mild and very slowly progressive, causing far less disability than other types of genetic cerebellar syndromes [bib_ref] Self-rated health status in spinocerebellar ataxia--results from a European multicenter study, Schmitz-Hubsch [/bib_ref] [bib_ref] Atypical Friedreich ataxia in patients with FXN p.R165P point mutation or comorbid..., Ygland [/bib_ref] [bib_ref] Autosomal dominant cerebellar ataxia with slow ocular saccades, Wictorin [/bib_ref]. For the neurologist, considering the diagnosis and enquiring about any of the non-neurological disease features in the patient or family members is paramount. Once the diagnosis is verified, the patient can be reassured that the neurological disease usually will be slowly progressive and often remain mild to moderate until high age. Rehabilitation programs may be beneficial for those who experience balance problems. Awareness and specific measures for the less well-known form of vision impairment due to retinal dysfunction may reduce overall impairment. Access to dedicated neurology, ophthalmology and rehabilitation services should be provided. In the presently known families, there was complete or near-complete penetrance of neurological signs in SAMD9L mutation carriers. Thus, including neurological examination for, at the very least, balance and eye-movement abnormalities, in a hematology workup for cytopenia or MDS may faster lead to a suspicion of SAMD9L-related disease.
# Conclusions
GOF mutations in SAMD9L, located in the chromosome 7q, are responsible for the ATXPC syndrome. In all of the patients affected by myelodysplasia, aberrations of chromosome 7 resulted in a loss of the mutated SAMD9L allele during MDS transformation. Analysis of mutation carriers who went through a hematopoietic crisis at a young age, and who showed no symptoms at time of analysis, revealed CN-LOH of chromosome 7q not present at birth. Additionally, a LOF mutation in the same codon as the germline SAMD9L mutation occurred in one of these patients. De novo as well as familial germ-line mutations in the SAMD9 gene have been discovered to cause the MIRAGE syndrome. The majority of MIRAGE patients died in infancy, but some developed a progressive loss of the mutated SAMD9 allele through monosomy 7 or 7q-, and secondary LOF mutations. The latter eliminated the growth-restricting effects on the mutant SAMD9 protein in hematopoietic cells. However, a number of patients developed MDS with monosomy 7 or 7q-similar to what was observed in ATXPC. This illustrates that the elevated risk for MDS and AML in both syndromes are likely driven by monosomy 7 or 7q-, as a reversion mechanism gone awry, resulting in loss of the growth restricting imposed by the GOF mutation, but by secondary loss of genes in CDR1-3 on chromosome 7q contribute to leukemogenesis.
[table] Table 1: Summary of clinical phenotype of reported patients with ATXPC /MLSM7 and MIRAGE [/table]
[table] Table 2: Recommendations for clinical care and surveillance of carriers of pathogenic mutations in SAMD9 or SAMD9L• Clinical examination including basic neurological examination (test for balance and eye movement abnormalities)• Thorough family history • CBC with differential • Lymphocyte subpopulation analysis (to detect deficiencies in NK-cell and other populations) • Sequencing of SAMD9 or SAMD9L on genomic DNA obtained from a non-hematopoietic tissue • Sequencing of SAMD9 or SAMD9L on DNA from hematopoietic tissue to measure mutant VAF of the GOF mutation and to detect genetic reversion by somatic LOF mutations • Bone marrow biopsy with aspirate for karyotype and FISH for chromosome 7q as well as molecular analysis using targeted sequencing for genes mutated in myeloid malignancies to evaluate the risk for MDS transformation• SNP array on peripheral blood or bone marrow to estimate the extent of the CN-LOH of chromosome 7 or 7q • HLA typing • CXR (to look for alveolar proteinosis)• MRI of brain with T2/FLAIR imaging (for baseline assessment of cerebellar atrophy and white matter abnormalities; SAMD9L only)• Clinical examination (hematology) every 6-12 months • CBC with differential every 6-12 months • CXR every third year (maybe every fifth year if the initial CXR was negative to minimize the exposure to radiation)• Gynecological examination once a year (but this is depending on the initial results, HPV vaccination status and on the identification of HPV strains)• Offer neurological and ophthalmological rehabilitation support as appropriate• Bone marrow aspirations for karyotype every other year or if any significant persistent change in CBC is detected • Targeted sequencing for genes mutated in myeloid malignancies to evaluate risk for MDS transformation every ATXPC ataxia-pancytopenia syndrome,CBC complete blood count, CN-LOH copy number neutral loss of heterozygosity,CXR chest x-ray, FISH fluorescent in situ hybridization, GOF gain-of-function, HLA human leukocyte antigen, [/table]
|
Isthmin—A Multifaceted Protein Family
Isthmin (ISM) is a secreted protein family with two members, namely ISM1 and ISM2, both containing a TSR1 domain followed by an AMOP domain. Its broad expression pattern suggests diverse functions in developmental and physiological processes. Over the past few years, multiple studies have focused on the functional analysis of the ISM protein family in several events, including angiogenesis, metabolism, organ homeostasis, immunity, craniofacial development, and cancer. Even though ISM was identified two decades ago, we are still short of understanding the roles of the ISM protein family in embryonic development and other pathological processes. To address the role of ISM, functional studies have begun but unresolved issues remain. To elucidate the regulatory mechanism of ISM, it is crucial to determine its interactions with other ligands and receptors that lead to the activation of downstream signalling pathways. This review provides a perspective on the gene organization and evolution of the ISM family, their links with developmental and physiological functions, and key questions for the future.
## Gene organization
Isthmin (ISM) is a secreted protein that was first detected through an unbiased screening for secreted proteins in Xenopus embryos and initially named Xenopus Isthmin (xIsm). The ISM protein family has two members, namely ISM1 (~60 kDa) and ISM2 (~63.9 kDa). Both of these proteins contain a hydrophobic signal peptide at the N-terminus along with a centrally positioned thrombospondin type 1 repeat (TSR1) domain. Sequence analysis has revealed that along with the TSR1 domain, ISM also contains another C-terminus domain called the Adhesin-associated domain in MUC4 and other protein (AMOP) domains [bib_ref] Isthmin is a novel secreted protein expressed as part of the Fgf-8..., Pera [/bib_ref].
## Evolutionary history
The Ism1 gene has a conserved sequence in various organisms, including humans, chickens, zebrafish, dogs, and cows [bib_ref] Distinct spatiotemporal expression of ISM1 during mouse and chick development, Osório [/bib_ref]. Ism1 is located at human chromosome 20 (20p12.1) and the amino acid (aa) sequence length of 464 is comprised of six exons covering 77.7 kb. In mice, Ism1 is located at chromosome 2 (2;2F3) and the aa sequence length is 454. In chickens, the location of this gene is at chromosome 3 and the aa sequence is 443, while in zebrafish it is at chromosomal location 13 and the aa sequence is 443 [bib_ref] Distinct spatiotemporal expression of ISM1 during mouse and chick development, Osório [/bib_ref]. Phylogenetic analysis demonstrates that both ISM1 and ISM2 proteins are highly conserved in various organisms, thus suggesting the role of the ISM protein family in cellular events. (See.
The reliability of each node was estimated by bootstrapping with 1000 replications. The numbers shown at each mode indicate the bootstrap values (%). The alignment was generated by using MUSCLE before constructing the phylogenetic tree and the pairwise deletion method was used for the treatment of missing data. The accession numbers used for Ism1 are indicated as follows: NP_543016. [bib_ref] Isthmin is a novel secreted protein expressed as part of the Fgf-8..., Pera [/bib_ref] The reliability of each node was estimated by bootstrapping with 1000 replications. The numbers shown at each mode indicate the bootstrap values (%). The alignment was generated by using MUSCLE before constructing the phylogenetic tree and the pairwise deletion method was used for the treatment of missing data. The accession numbers used for Ism1 are indicated as follows: NP_543016. [bib_ref] Isthmin is a novel secreted protein expressed as part of the Fgf-8..., Pera [/bib_ref]
## Tsr1 domain
The TSR domain is present within the extracellular proteins or in the membranebound proteins, and they are associated with multiple functions, including immunity, neuronal development, cell proliferation, cell to cell interaction, cell to matrix interaction, cellular adhesion, migration, and apoptosis [bib_ref] The thrombospondin type 1 repeat superfamily, Tucker [/bib_ref] [bib_ref] The structure of human thrombospondin, an adhesive glycoprotein with multiple calcium-binding sites..., Lawler [/bib_ref] [bib_ref] The layered fold of the TSR domain of P. falciparum TRAP contains..., Tossavainen [/bib_ref] [bib_ref] The cell biology of thrombospondin-1, Chen [/bib_ref]. TSR1 contains 60 residues that are comprised of a reverse three-stranded β-sheet and its core is formed by a conserved amino acid residue that includes tryptophan (W), arginine (R), and cysteine (C). A domain-based analysis of the TSR1 domain in ISM1 and ISM2 along with 25 other proteins can be seen in [fig_ref] Figure 2: Schematic representation of human TSR1 domain sequence alignment of 27 proteins drawn... [/fig_ref]. The TSR1 domain that forms part of various protein families including spondin, UNC, semaphorins, ADAMTS, brain angiogenesis inhibitor 1 (BAI1), and the human thrombospondin-1 (hTSP-1) protein has a wide range of functions. A previous study found that the TSR domain mediates the binding of F-spondin to the extracellular matrix (ECM) [bib_ref] ISM1 regulates NODAL signaling and asymmetric organ morphogenesis during development, Osório [/bib_ref] , while it is also involved in the binding of SCO-spondin to another extracellular matrix (ECM) named Reissner's fiber [bib_ref] Regulation of N-glycosylation and secretion of Isthmin-1 by its C-mannosylation, Yoshimoto [/bib_ref] , eventually facilitating the removal of monoamines from cerebrospinal fluid. Furthermore, TSRs are involved in motor neuron and distal tip cell (DTC) migration in UNC-5 as well as axon guidance in the UNC-6 protein, while the presence of multiple TSR domains is liable for diverse functions in various tissues [bib_ref] N -glycosylation is required for secretion and enzymatic activity of human hyaluronidase1, Goto [/bib_ref]. In the ADAMTS protein family, the TSR domain containing multiple proteins The TSR1 domain that forms part of various protein families including spondin, UNC, semaphorins, ADAMTS, brain angiogenesis inhibitor 1 (BAI1), and the human thrombospondin-1 (hTSP-1) protein has a wide range of functions. A previous study found that the TSR domain mediates the binding of F-spondin to the extracellular matrix (ECM) [bib_ref] Plasmin-mediated Release of the Guidance Molecule F-spondin from the Extracellular Matrix, Tzarfaty-Majar [/bib_ref] , while it is also involved in the binding of SCO-spondin to another extracellular matrix (ECM) named Reissner's fiber [bib_ref] The thrombospondin type 1 repeat superfamily, Tucker [/bib_ref] , eventually facilitating the removal of monoamines from cerebrospinal fluid [bib_ref] Reissner fiber binds and transports away monoamines present in the cerebrospinal fluid, Caprile [/bib_ref]. Furthermore, TSRs are involved in motor neuron and distal tip cell (DTC) migration in UNC-5 as well as axon guidance in the UNC-6 protein, while the presence of multiple TSR domains is liable for diverse functions in various tissues [bib_ref] UNC-5 Function Requires Phosphorylation of Cytoplasmic Tyrosine 482, but Its UNC-40-Independent Functions..., Killeen [/bib_ref]. In the ADAMTS protein family, the TSR domain containing multiple proteins is involved in proteolysis [bib_ref] Characterization of ADAMTS-9 and ADAMTS-20 as a Distinct ADAMTS Subfamily Related to..., Somerville [/bib_ref] , whereas another study has suggested that the TSR domain in the papilin protein mediates inhibition of proteolytic activity in the ADAMTS protein.
Moreover, anti-angiogenic properties of the TSR domain were observed in the BAI1 protein [bib_ref] Brain Angiogenesis Inhibitor 1 Is Differentially Expressed in Normal Brain and Glioblastoma..., Kaur [/bib_ref]. hTSP-1 is a glycosylated matricellular protein involved in the regulation of ECM function, whereas the TSR protein interacts with CD36 and negatively regulates angiogenesis, further inducing endothelial cell apoptosis [bib_ref] Contribution of Human Thrombospondin-1 to the Pathogenesis of Gram-Positive Bacteria, Binsker [/bib_ref] [bib_ref] Molecular Basis of Antiangiogenic Thrombospondin-1 Type 1 Repeat Domain Interactions With CD36, Klenotic [/bib_ref] [bib_ref] Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin-1, Jiménez [/bib_ref]. TSP1 performs significant roles in tissue repair processes through the TSR domain, which is a recognized regulator in the activation of latent TGF-β and thus in wound healing and fibrosis [bib_ref] Thrombospondin1 in tissue repair and fibrosis: TGF-β-dependent and independent mechanisms, Sweetwyne [/bib_ref].
Two motifs of the TSR domain in TSPI, KRFK and WxxW, are requisite for this latent TGF-β activation. An earlier study demonstrated that the WSXW motif facilitates binding of the TSR domain to the intact heparin [bib_ref] Specificities of Heparin-binding Sites from the Amino-Terminus and Type 1 Repeats of..., Yu [/bib_ref].
The WSLW motif is present in the TSR1 domain of ISM1 and WSPW in ISM2. It is worth investigating whether these two motifs have the same functions, including TGF-β activation. A motif-based analysis of ISM1 and ISM2 can be seen in [fig_ref] Figure 3: Schematic representation of thrombospondin type 1 repeat [/fig_ref]. [bib_ref] Simizu, S. C-mannosylation of R-spondin3 regulates its secretion and activity of Wnt/β-catenin..., Fujiwara [/bib_ref].
Moreover, anti-angiogenic properties of the TSR domain were observed in the BAI1 protein [bib_ref] N-Glycosylation Regulates ADAM8 Processing and Activation, Srinivasan [/bib_ref]. hTSP-1 is a glycosylated matricellular protein involved in the regulation of ECM function, whereas the TSR protein interacts with CD36 and negatively regulates angiogenesis, further inducing endothelial cell apoptosis [bib_ref] Contribution of Human Thrombospondin-1 to the Pathogenesis of Gram-Positive Bacteria, Binsker [/bib_ref] [bib_ref] Molecular Basis of Antiangiogenic Thrombospondin-1 Type 1 Repeat Domain Interactions With CD36, Klenotic [/bib_ref] [bib_ref] Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin-1, Jiménez [/bib_ref]. TSP1 performs significant roles in tissue repair processes through the TSR domain, which is a recognized regulator in the activation of latent TGF-β and thus in wound healing and fibrosis [bib_ref] N-Glycans of ADAMTS13 modulate its secretion and von Willebrand factor cleaving activity, Zhou [/bib_ref].
Two motifs of the TSR domain in TSPI, KRFK and WxxW, are requisite for this latent TGF-β activation. An earlier study demonstrated that the WSXW motif facilitates binding of the TSR domain to the intact heparin [bib_ref] Transcriptomic landscape of the primitive streak, Alev [/bib_ref].
The WSLW motif is present in the TSR1 domain of ISM1 and WSPW in ISM2. It is worth investigating whether these two motifs have the same functions, including TGF-β activation. A motif-based analysis of ISM1 and ISM2 can be seen in [fig_ref] Figure 3: Schematic representation of thrombospondin type 1 repeat [/fig_ref]. Additionally, the CSVTCG motif of the TSR domain facilitates interaction with CD36 that triggers the anti-angiogenic activity via apoptosis of endothelial cells [bib_ref] Thrombospondin-1-induced apoptosis of brain microvascular endothelial cells can be mediated by TNF-R1, Rege [/bib_ref]. ISM1 also has a CSVTCG motif in the TSR1 domain, and it is also worthy of investigation whether it is also involved in triggering anti-angiogenic function. In ISM1, TSR1 has a DGE motif at the position of 218 to 220, and this is an α2β1 ligand sequence, which is a major collagen receptor on platelets as well as a key co-stimulatory pathway of effector T cells that has been implicated in the pathogenesis of arthritis. The α2β1 integrin is expressed by human Th17 cells and it promotes the survival of effector T cells in humans through the inhibition of Fas-induced apoptosis [bib_ref] Alpha2beta1 Integrin (VLA-2) Protects Activated Human Effector T Cells From Methotrexate-Induced Apoptosis, Abderrazak [/bib_ref]. Integrin α2β1 facilitated the interaction between embryonic retinal cells and collagen in an avian model. Furthermore, the EPQ motif is present in the TSR1 domain of ISM2, and this motif is known to determine the carbohydrate binding specificity [bib_ref] A novel C-type lectin from crab Eriocheir sinensis functions as pattern recognition..., Wang [/bib_ref].
## Amop domain
AMOP is only observed in extracellular-domain-containing proteins engaged in cell adhesion, which suggests extracellular localization of the protein. The AMOP domain is~100 residues long and encompasses eight invariant residues of cysteine, which was initially found to be present in a few splice variants of MUC4 and in four other proteins, Mesh, SUSD2, ISM1, and ISM2 [bib_ref] The septate junction protein Mesh is required for epithelial morphogenesis, ion transport,..., Jonusaite [/bib_ref] (see [fig_ref] Figure 4: Schematic representation of human AMOP domain sequence alignment in 4 proteins drawn... [/fig_ref].
Additionally, the CSVTCG motif of the TSR domain facilitates interaction with CD36 that triggers the anti-angiogenic activity via apoptosis of endothelial cells [bib_ref] Thrombospondin-1-induced apoptosis of brain microvascular endothelial cells can be mediated by TNF-R1, Rege [/bib_ref]. ISM1 also has a CSVTCG motif in the TSR1 domain, and it is also worthy of investigation whether it is also involved in triggering anti-angiogenic function. In ISM1, TSR1 has a DGE motif at the position of 218 to 220, and this is an α2β1 ligand sequence, which is a major collagen receptor on platelets as well as a key co-stimulatory pathway of effector T cells that has been implicated in the pathogenesis of arthritis. The α2β1 integrin is expressed by human Th17 cells and it promotes the survival of effector T cells in humans through the inhibition of Fas-induced apoptosis [bib_ref] Isthmin1, a secreted signaling protein, acts downstream of diverse embryonic patterning centers..., Kesavan [/bib_ref]. Integrin α2β1 facilitated the interaction between embryonic retinal cells and collagen in an avian model [bib_ref] The Sp1-Related Transcription Factors sp5 and sp5-like Act Downstream of Wnt/β-Catenin Signaling..., Weidinger [/bib_ref]. Furthermore, the EPQ motif is present in the TSR1 domain of ISM2, and this motif is known to determine the carbohydrate binding specificity [bib_ref] Nodal signaling activates differentiation genes during zebrafish gastrulation, Bennett [/bib_ref].
## Amop domain
AMOP is only observed in extracellular-domain-containing proteins engaged in cell adhesion, which suggests extracellular localization of the protein. The AMOP domain is ~100 residues long and encompasses eight invariant residues of cysteine, which was initially found to be present in a few splice variants of MUC4 and in four other proteins, Mesh, SUSD2, ISM1, and ISM2 [bib_ref] Vascularization of engineered tissues: Approaches to promote angio-genesis in biomaterials, Moon [/bib_ref] (see [fig_ref] Figure 4: Schematic representation of human AMOP domain sequence alignment in 4 proteins drawn... [/fig_ref]. AMOP-containing proteins have conserved cysteine (C) residues, while AMOP in both ISM1 and ISM2 also possesses a KGD motif that binds to the integrin αIIbβ3 present in the multiple antagonists of platelet aggregation and is engaged in the integrin-mediated cellular adhesion and tumour metastasis [bib_ref] Angiogenesis in the female reproductive organs: Pathological implications, Reynolds [/bib_ref] (see [fig_ref] Figure 5: Schematic representation of AMOP domain of ISM1 [/fig_ref]. AMOP-containing proteins have conserved cysteine (C) residues, while AMOP in both ISM1 and ISM2 also possesses a KGD motif that binds to the integrin α IIb β 3 present in the multiple antagonists of platelet aggregation and is engaged in the integrin-mediated cellular adhesion and tumour metastasis [bib_ref] The Effect of the Single Substitution of Arginine within the RGD Tripeptide..., Lu [/bib_ref] (see [fig_ref] Figure 5: Schematic representation of AMOP domain of ISM1 [/fig_ref].
RGD and KGD are two integrin-bound motifs that were recently identified in the S proteins of SARS-CoV-2 and the ACE2 protein [bib_ref] A potential inhibitory role for integrin in the receptor targeting of SARS-CoV-2, Luan [/bib_ref] [bib_ref] Molecular interaction and inhibition of SARS-CoV-2 binding to the ACE2 receptor, Yang [/bib_ref]. The AMOP domain encompasses an RKD motif that is involved in integrin-dependent cell adhesions [bib_ref] Blockade of αvβ3 and αvβ5 integrins by RGD mimetics induces anoikis and..., Maubant [/bib_ref]. However, the RKD motif present in the AMOP domain of ISM1 has been shown to be selective in binding with the extracellular surface of αvβ5 integrins that are engaged in vascular permeability as well as cell migration [bib_ref] Identification of Isthmin 1 as a Novel Clefting and Craniofacial Patterning Gene..., Lansdon [/bib_ref]. Additionally, the AMOP domain in ISM2 also has an WSRL motif that is known to be involved in autophagy induction. RGD and KGD are two integrin-bound motifs that were recently identified in the S proteins of SARS-CoV-2 and the ACE2 protein [bib_ref] A potential inhibitory role for integrin in the receptor targeting of SARS-CoV-2, Luan [/bib_ref] [bib_ref] Molecular interaction and inhibition of SARS-CoV-2 binding to the ACE2 receptor, Yang [/bib_ref]. The AMOP domain encompasses an RKD motif that is involved in integrin-dependent cell adhesions [bib_ref] Blockade of αvβ3 and αvβ5 integrins by RGD mimetics induces anoikis and..., Maubant [/bib_ref]. However, the RKD motif present in the AMOP domain of ISM1 has been shown to be selective in binding with the extracellular surface of αvβ5 integrins that are engaged in vascular permeability as well as cell migration. Additionally, the AMOP domain in ISM2 also has an WSRL motif that is known to be involved in autophagy induction [bib_ref] Roles of main pro-and anti-angiogenic factors in tumor angiogenesis, Huang [/bib_ref].
## Post-translational modifications
Post-translational modifications (PTMs) are considered as backbones of proteins because of their regulatory behaviour over the localization and function of their respective proteins [bib_ref] Isthmin is a novel secreted angiogenesis inhibitor that inhibits tumour growth in..., Xiang [/bib_ref]. N-glycosylation is a well-recognized type of post-translational modification in which N-glycans are directly enclosed in Asparagine amino acid residue in proteins, and it is essential for the functional features of proteins, including stability, folding, and secretion [bib_ref] Function and interactions of integrins, Van Der Flier [/bib_ref]. ISM1 is demonstrated as an N-glycosylated secreted protein based on biochemical analysis, and two putative N-glycosylation sites at the positions N39 and N28 were identified, while site-directed mutagenesis by mimicking Asparagine (N) to Glutamine (Q) caused reduction in the ISM1 protein size, thereby indicating that N-
## Post-translational modifications
Post-translational modifications (PTMs) are considered as backbones of proteins because of their regulatory behaviour over the localization and function of their respective proteins. N-glycosylation is a well-recognized type of post-translational modification in which N-glycans are directly enclosed in Asparagine amino acid residue in proteins, and it is essential for the functional features of proteins, including stability, folding, and secretion [bib_ref] Mutually exclusive locales for N-linked glycans and disorder in human glycoproteins, Goutham [/bib_ref]. ISM1 is demonstrated as an N-glycosylated secreted protein based on biochemical analysis, and two putative N-glycosylation sites at the positions N39 and N28 were identified, while site-directed mutagenesis by mimicking Asparagine (N) to Glutamine (Q) caused reduction in the ISM1 protein size, thereby indicating that N-glycosylation is required for the secretion of ISM1 which was further validated by tunicamycin treatment [bib_ref] ISM1 regulates NODAL signaling and asymmetric organ morphogenesis during development, Osório [/bib_ref]. Following that, another study identified two C-mannosylation sites at the TSRI domain, Trp 223 and Trp 226 , wherein site-directed mutagenesis and tunicamycin treatment demonstrated that C-mannosylation is required for N-glycosylation of ISM1 [bib_ref] Regulation of N-glycosylation and secretion of Isthmin-1 by its C-mannosylation, Yoshimoto [/bib_ref]. Moreover, it has been reported that C-mannosylation facilitates protein folding and stabilizing, particularly in the TSR domain. In addition, N-glycosylation in Hyaluronidase 1 (HYAL-1) is required for its secretion and thus regulates enzymatic activity, while upregulation of HYAL-1 is reported to be linked to tumour cell proliferation and angiogenesis in multiple cancers [bib_ref] N -glycosylation is required for secretion and enzymatic activity of human hyaluronidase1, Goto [/bib_ref] [bib_ref] Simizu, S. C-mannosylation of R-spondin2 activates Wnt/β-catenin signaling and migration activity in..., Mizuta [/bib_ref] [bib_ref] Simizu, S. C-mannosylation of R-spondin3 regulates its secretion and activity of Wnt/β-catenin..., Fujiwara [/bib_ref]. Moreover, N-glycosylation of ADAM8 and ADAM13 proteins is crucial for its functional features, including processing, stability, and activity [bib_ref] N-Glycosylation Regulates ADAM8 Processing and Activation, Srinivasan [/bib_ref] [bib_ref] N-Glycans of ADAMTS13 modulate its secretion and von Willebrand factor cleaving activity, Zhou [/bib_ref].
## Expression and function
ISM1 is expressed in a large number of human tissues, such as lung, liver, breast, brain, stomach, muscle, skin, bone marrow, and colon. Spatiotemporal analysis indicated that in mouse embryos, Ism1 transcripts were observed in the anterior mesendoderm (AM), paraxial and lateral plate mesoderm (LPM), mid-brain hindbrain boundary (MHB), and trunk neural tube, while in adult stages, the highest expression level of Ism1 was found in the lung and the brain and a moderate expression level was observed in the heart, kidney, ovary, testis, and bone marrow [bib_ref] Distinct spatiotemporal expression of ISM1 during mouse and chick development, Osório [/bib_ref]. On the other hand, a different expression pattern was observed in chicken embryos, such as that vigorous expression of Ism1 transcripts was observed in the anterior head mesoderm, optic vesicles, LPM, trunk neural tube, and dermomyotome, whereas a strong expression level was observed in the ear, eye, and spinal cord primordia in later stages [bib_ref] Distinct spatiotemporal expression of ISM1 during mouse and chick development, Osório [/bib_ref]. Moreover, another previous study identified ism1 amongst the top 20 genes which are expressed in the site of anterior primitive streak (APS) in the chicken embryo [bib_ref] Transcriptomic landscape of the primitive streak, Alev [/bib_ref]. Interestingly, asymmetric expression of Ism1 led to the study on the embryonic role of ISM1 on chicken embryos, whereas defective NODAL signalling in the left LPM as well as impaired asymmetric heart morphogenesis were triggered by ISM1, and thus, it was identified as an antagonist of NODAL signalling which plays role in asymmetric organ morphogenesis [bib_ref] ISM1 regulates NODAL signaling and asymmetric organ morphogenesis during development, Osório [/bib_ref]. This study also revealed that through the C-terminus AMOP domain, ISM1 negatively regulates NODAL signalling followed by suppressing activation of effector molecule SMAD2 [bib_ref] ISM1 regulates NODAL signaling and asymmetric organ morphogenesis during development, Osório [/bib_ref].
Additionally, ISM1 is expressed since the initial stages of embryonic development in both Xenopus and zebrafish. Ism1 has strong maternal expression in Xenopus eggs and other expression domains in the embryo, including the blastopore lip, paraxial mesoderm, neural folds, cranial neural crest, and ear placode. While the expression pattern is similar to fgf8, ISM1 was demonstrated as part of the FGF8 synexpression group, including the Spry and Sef genes [bib_ref] Isthmin is a novel secreted protein expressed as part of the Fgf-8..., Pera [/bib_ref].
In addition, ISM1 expression was also found in the brain, heart, eye, and spleen tissues of zebrafish. In zebrafish embryonic development, Ism1 was identified at very early developmental stages, including the tail bud stage, shield stage, and early somitogenesis, while the expression of Ism2 could not be identified up to 24 h post fertilization (hpf). This is despite the fact that Ism2 expression was most vigorous in two bilateral streams of the mesenchymal cells in the head region and moderate expression was found in trunk [bib_ref] Isthmin1, a secreted signaling protein, acts downstream of diverse embryonic patterning centers..., Kesavan [/bib_ref]. Furthermore, Ism1 knockdown in zebrafish embryos gave rise to the aberrant formation of intersegmental vessel (ISV) in the trunk. Previously, upregulation of Ism1 in the zebrafish was identified in relation to the WNT/β-catenin signalling pathway [bib_ref] The Sp1-Related Transcription Factors sp5 and sp5-like Act Downstream of Wnt/β-Catenin Signaling..., Weidinger [/bib_ref] , and in another study, Ism1 was also identified as a NODAL-controlled gene [bib_ref] Nodal signaling activates differentiation genes during zebrafish gastrulation, Bennett [/bib_ref]. Additionally, Ism1 expression in blastoderm required nodal signalling and its subsequent expression, even though the authors observed a compensated function in embryonic development [bib_ref] Isthmin1, a secreted signaling protein, acts downstream of diverse embryonic patterning centers..., Kesavan [/bib_ref] which is opposite to that observed in chicken embryos. Spatiotemporal expression of Ism1 and being a target gene of multiple signalling pathways as mentioned above indicates there might be other functions of ISM in multiple biological processes. Over the past few years, multiple studies have focused on the functional analysis of ISM1 in several events, including angiogenesis, metabolism, organ homeostasis, immunity, craniofacial development, and cancer.
## Angiogenesis
Angiogenesis is a tightly controlled process of forming new capillary blood vessels from pre-existing ones which generates oxygen and nutrients for cells [bib_ref] Vascularization of engineered tissues: Approaches to promote angio-genesis in biomaterials, Moon [/bib_ref]. Following birth, angiogenesis continues to facilitate organ growth; nevertheless, in adulthood, blood vessels mainly persist quiescently, while angiogenesis only occurs in multiple physiological circumstances, such as wound healing, damaged tissue repair, embryonic development, and organogenesis [bib_ref] Angiogenesis in the female reproductive organs: Pathological implications, Reynolds [/bib_ref]. A plethora of growth factors, cytokines, and secreted proteins participate in the angiogenesis process [bib_ref] Roles of main pro-and anti-angiogenic factors in tumor angiogenesis, Huang [/bib_ref]. In addition, angiogenesis could be inhibited by endogenous inhibitors, which are proteins formed and secreted in the body that may inhibit blood vessel formation.
A previous study has demonstrated ISM1 as a novel inhibitor of angiogenesis [bib_ref] Isthmin is a novel secreted angiogenesis inhibitor that inhibits tumour growth in..., Xiang [/bib_ref] , whereas recombinant mouse ISM1 (rISM1) was expressed in E. coli and used to treat human umbilical vein endothelial cells (ECs) in vitro. Here, it was found that through the AMOP domain, ISM1 can inhibit capillary network formation in the initial stages of angiogenesis via its interaction with αVβ5 integrin without affecting EC migration. Additionally, rISM1 is reported to inhibit vascular endothelial growth factor (VEGF)-mediated angiogenesis, while overexpression of ISM1 suppressed tumour angiogenesis in B16 melanoma cells and tumour growth in mice [bib_ref] Isthmin is a novel secreted angiogenesis inhibitor that inhibits tumour growth in..., Xiang [/bib_ref].
Furthermore, integrins are cell-surface-signalling receptors involved in the activation of multiple intra-cellular signalling pathways as well as mediation of adhesion to the ECM and adjacent cells [bib_ref] Function and interactions of integrins, Van Der Flier [/bib_ref]. Additionally, integrins also serve as regulators of multiple processes, including cell growth, proliferation, migration, differentiation, apoptosis, and tissue repair [bib_ref] The role of integrins in inflammation and angiogenesis, Mezu-Ndubuisi [/bib_ref]. Integrins are also reported to be instantly inhibited by means of antibodies, peptides, and peptidomimetics [bib_ref] A reevaluation of integrins as regulators of angiogenesis, Hynes [/bib_ref].
Immobilized ISM1 in the ECM promotes EC survival, while conversely, soluble ISM1 induces EC apoptosis via αvβ5 integrin as an antagonist and indicates that the antiangiogenic functions of ISM1 are diminished when it is present in immobilized form. Contrary to ISM1, other signalling molecules including vitronectin and fibronectin serving as agonists do not exhibit dual function based on their soluble or insoluble state.
Glioma is a common central nervous system (CNS) tumour that accounts for 30% of all intracranial tumour incidences, and angiogenesis is vital to the formation of malignant gliomas. Furthermore, ECM molecules promote cell signalling via activation of particular cell adhesion receptors, access modulation to soluble factors, and modification of mechanical properties of the tissue . Additionally, an earlier study [bib_ref] Isthmin inhibits glioma growth through antiangiogenesis in vivo, Yuan [/bib_ref] investigated the role of ISM1 in glioma angiogenesis and found that it can hinder HUVEC proliferation by the inhibition of VEGF. Furthermore, it was also demonstrated that ISM1 activated the HUVEC apoptosis through the caspase-3 pathway, and thus, inhibits tumour angiogenesis in vivo [bib_ref] Isthmin inhibits glioma growth through antiangiogenesis in vivo, Yuan [/bib_ref].
At present, the role of ISM2 in angiogenesis still remains elusive, while it is not surprising that it also contains the TSR1 domain, which has been observed in multiple anti-angiogenic proteins. For instance, the anti-angiogenic function of ADAMTS1 requires the TSR1 domain [bib_ref] Full-length ADAMTS-1 and the ADAMTS-1 fragments display pro-and antimetastatic activity, respectively, Liu [/bib_ref] , while it has anti-angiogenic activity in primary gastric cancer [bib_ref] Expression of ADAMTS1 and Its Correlation with Angiogenesis in Primary Gastric Cancer..., Chen [/bib_ref]. Furthermore, ADAMTS2 is reported to inhibit the formation of the capillary network in EC 3D culture models [bib_ref] ADAMTS-2 functions as anti-angiogenic and anti-tumoral molecule independently of its catalytic activity, Dubail [/bib_ref] , whereas ADAMTS4 is also reported to suppress angiogenesis and melanoma growth [bib_ref] Ge, R. ADAMTS4 and its proteolytic fragments differentially affect melanoma growth and..., Rao [/bib_ref]. Additionally, ADAMTS5 was also reported to inhibit angiogenesis through the TSR1 domain [bib_ref] The first but not the second thrombospondin type 1 repeat of ADAMTS5..., Sharghi-Namini [/bib_ref]. Moreover, SCO-spondin (SCOP) is also reported to function as an angiogenesis inhibitor in glioblastoma [bib_ref] SCO-spondin oligopeptide inhibits angiogenesis in glioblastoma, Bibes [/bib_ref] , while the UNC5B protein is involved in angiogenesis inhibition [bib_ref] Activation of the UNC5B receptor by Netrin-1 inhibits sprouting angiogenesis, Larrivée [/bib_ref] [bib_ref] Netrin-4 inhibits angiogenesis via binding to neogenin and recruitment of Unc5B, Lejmi [/bib_ref]. In summary, TSR1-domain-containing proteins seem to show anti-angiogenic functions, while according to our TSR1 domain sequence alignment analysis, ISM2 has a WSPW motif that is found to be conserved in ADAMTS12, SBSPON (somatomedin B and TSP-1-domain-containing protein), RSPO1 (R-spondin-1), and SEMA5A. It is noteworthy that this heparin-binding motif WSPW in the TSR domain was reported to mediate the inhibition of angiogenesis of ECs [bib_ref] Binding and displacement of vascular endothelial growth factor (VEGF) by thrombospondin: Effect..., Gupta [/bib_ref] and anti-proliferative activity of TSPs [bib_ref] Modulation of endothelial cell proliferation, adhesion, and motility by recombinant heparin-binding domain..., Vogel [/bib_ref]. The participation of ISM1 and the above-mentioned TSR1 domain containing secreted and ECM proteins in anti-angiogenesis suggests that this domain might be considered for the development of novel anti-angiogenesis therapeutic strategies.
## Pathological processes
In the studies listed above, ISM1 has been demonstrated as an endogenous angiogenesis inhibitor, and it might be interesting to elucidate whether this protein inhibits cancer to generate its own blood vessels. However, over the past few years, increasing evidence has revealed that the atypical expression of ISM1 may affect cancer.
Long non-coding RNAs (lncRNAs) perform a pivotal role in the progression and metastasis of a variety of carcinomas, while lncRNA H19 (H19) is highly prevalent in gastric cancer tissues [bib_ref] Long non-coding RNA H19 regulates cell growth and metastasis via the miR-22-3p/Snail1..., Gan [/bib_ref]. Another study demonstrated that ISM1 is a binding protein of lncRNA H19, which mediates ISM1 upregulation and boosts carcinogenesis and metastasis of gastric cancer.
Furthermore, in hepatocellular carcinoma (HCC), cell proliferation and migration are reported to be promoted by ISM1 and regulated by the interaction with circular RNA (circRNA) and micro RNA (miRNA), namely hsa_circ_0091570/miR-1307 hsa_circ_0091570, which competitively binds to miR-1307 by serving as the ceRNA (competing endogenous RNA) that then regulates ISM1 expression [bib_ref] hsa_circ_0091570 acts as a ceRNA to suppress hepatocellular cancer progression by sponging..., Wang [/bib_ref]. This study summarized that downregulation of hsa_circ_0091570 occurs in HCC and could act as ceRNA through sponging miR-1307 for regulating ISM1 expression and thereby taking part in HCC progression [bib_ref] hsa_circ_0091570 acts as a ceRNA to suppress hepatocellular cancer progression by sponging..., Wang [/bib_ref].
Additionally, colon adenocarcinoma (COAD) is a malignant tumour of the digestive tract that is associated with an extremely high incidence rate and is ranked third amongst all tumours globally, and miRNA is reported to play a major role in this tumour cell proliferation and apoptosis [bib_ref] MicroRNA miR-4779 suppresses tumor growth by inducing apoptosis and cell cycle arrest..., Koo [/bib_ref] [bib_ref] miRNAs and apoptosis: RNAs to die for, Jovanovic [/bib_ref]. The Wnt/β-catenin signalling pathway serves as a regulatory pathway in tumorigenesis because of its involvement in multiple cellular processes [bib_ref] The Wnt signaling pathway in tumorigenesis, pharmacological targets, and drug development for..., Wang [/bib_ref] , while Wnt3a is substantially enhanced in colon cancer cells in promoting tumour angiogenesis and metastasis [bib_ref] Wnt3a expression is associated with epithelial-mesenchymal transition and promotes colon cancer progression, Qi [/bib_ref]. A recent study [bib_ref] miR-1307-3p overexpression inhibits cell proliferation and promotes cell apoptosis by targeting ISM1..., Zheng [/bib_ref] has shown that miR-1307-3p inhibits the activation of the Wnt3a/β-catenin signalling pathway by targeting and thus downregulation of ISM1, which then inhibits proliferation and promotes the apoptosis of COAD cells. Conversely, ISM1 overexpression promoted activation of the Wnt3a/βcatenin signalling pathway along with proliferation and decelerated cell apoptosis in COAD cells.
Furthermore, a recent study has also reported the elevated expression of ISM1 in the CRC tissue of patients, and on top of that, a positive correlation of ISM1 with cancerassociated signalling pathways including EMT, hypoxia, KRAS, Notch, and Hedgehog were observed [bib_ref] Effect of ISM1 on the Immune Microenvironment and Epithelial-Mesenchymal Transition in Colorectal..., Wu [/bib_ref].
The inner surface of the micro vessels is lined by endothelial cells to generate a semi-permeable barrier that allows exchange of fluids and proteins in blood and tissue, while tight regulation of this endothelial permeability is crucial in the maintenance of organ homeostasis, and thus, dysfunction including endothelial hyperpermeability leads to vascular inflammation related to multiple diseases, such as respiratory distress, acute lung injury, sepsis, diabetes, and cancer [bib_ref] Molecular mechanisms of endothelial hyperpermeability: Implications in inflammation, Kumar [/bib_ref] [bib_ref] IL-4 Causes Hyperpermeability of Vascular Endothelial Cells through Wnt5A Signaling, Skaria [/bib_ref]. Identification of functional signalling molecules involved in vascular permeability (VP) for circulatory homeostasis might have therapeutic potential. ISM1 is demonstrated as a novel VP inducer which acts via cell-surface GRP78facilitated activation of Src followed by Src-mediated tyrosine phosphorylation of adherensjunction (AJ) proteins and the consequent dissociation of these AJ proteins, therefore resulting in barrier disruption [bib_ref] Isthmin is a novel vascular permeability inducer that functions through cell-surface GRP78-mediated..., Venugopal [/bib_ref]. Furthermore, this study [bib_ref] Isthmin is a novel vascular permeability inducer that functions through cell-surface GRP78-mediated..., Venugopal [/bib_ref] also reported that a lipopolysaccharide (LPS)-induced acute lung injury (ALI) mouse model had elevated upregulation associated to ALI and thus suggested the inhibition of ISM1 overexpression as a therapeutic strategy in ALI.
On the top of that, hypoxia is one of the most common reasons for vascular hyperpermeability and a crucial risk factor of several pathological characteristics in lung disorders [bib_ref] Hypoxia Induces Permeability and Giant Cell Responses of Andes Virus-Infected Pulmonary Endothelial..., Gavrilovskaya [/bib_ref]. Hyperpermeability of pulmonary microvascular endothelial cell (PMVEC) monolayers induced by hypoxia is vital for vascular leakage and leads to pulmonary diseases including ALI and high-altitude pulmonary oedema (HAPE). AECII are cuboidal cells that comprise about 15% of the overall lung cells, and they account for epithelium reparation and facilitate lung homeostasis through the secretion of several lysozymes and proteins. The elevated ISM1 level in AECII was accountable for hypoxia triggered PMVEC monolayer hyperpermeability in an AECII/PMVEC co-culture system that indicated substantial function of alveolar epithelial cells and a modulatory role of ISM1 in hyperpermeability featured lung diseases [bib_ref] Novel HIF-1-target gene isthmin1 contributes to hypoxia-induced hyperpermeability of pulmonary microvascular endothelial..., Li [/bib_ref]. Hypoxia-inducible factor-1α (HIF1α) is a master transcriptional regulator of hypoxia, while the same study demonstrated that elevated HIF1α expression transcriptionally activated Ism1 gene expression and thus identified Ism1 as a novel HIF1α target gene [bib_ref] Novel HIF-1-target gene isthmin1 contributes to hypoxia-induced hyperpermeability of pulmonary microvascular endothelial..., Li [/bib_ref].
It is already clear that inflammation control and maintenance of homeostasis are of utmost importance for lung health, whereas prolonged lung inflammation leads to chronic obstructive pulmonary disease (COPD), and the severity of disease in COPD patients is directly linked to the accumulation of alveolar macrophages (AMs). It is important to highlight that, by using genetic (Ism1) and pathological (COPD) mouse models, one study reported that the secreted protein ISM1 is lung-resident, having a high expression that safeguards lung homeostasis through controlling AM numbers and an efficient phenotype through cell-surface GRP78 (csGRP78)-facilitated apoptosis [bib_ref] ISM1 protects lung homeostasis via cell-surface GRP78-mediated alveolar macrophage apoptosis, Lam [/bib_ref]. Furthermore, the same study revealed that intratracheal delivery of recombinant ISM1 (rISM1) exhibited effective suppression of lung inflammation through the depletion of the pro-inflammatory cs-GRP78 high AMs by targeted apoptosis, as well as prevented emphysema development and thus retained pulmonary function in cigarette-smoke-induced COPD mice. Furthermore, Ism1 knockout in mice exhibited an increase in cs-GRP78 high AMs along with upregulation of MMP9, MMP12, and NF-κB p65, in addition to a moderate increase in TGF-β1 and VEGF-A, prolonged lung inflammation, and progressive emphysema [bib_ref] ISM1 protects lung homeostasis via cell-surface GRP78-mediated alveolar macrophage apoptosis, Lam [/bib_ref].
On the other hand, GRP78 is a stress-response protein that belongs to the heat-shock protein family, which has the ability to modulate protein folding and is upregulated in cells under stress. This protein is known to be highly expressed in various human cancers, including melanoma and breast, prostate, lung, and ovarian cancer, and thus it is associated with chemoresistance, increased malignancy, and inadequate patient outcomes [bib_ref] The unfolded protein response regulator GRP78/BiP is required for endoplasmic reticulum integrity..., Li [/bib_ref]. Of note is that ISM1 selectively promotes cellular apoptosis, harbouring elevated cell-surface GRP78 in activated ECs as well as in metastatic and aggressive cancer cells, while systemic delivery of the GRP78-specific cyclic peptide BC71 effectively suppressed the growth of subcutaneous tumours in mice.
Moreover, orofacial clefts have a complex aetiology and are one of the most prominent birth defects, affecting 1-2 children per 1000 births. Complex diseases have been reported to be linked with copy number variants (CNVs) and ISM1 heterozygous deletions were shown to be enriched in cleft lip and palate cases compared to the controls, therefore indicating that the loss of even one copy predisposes to this disorder (Lansdon, Darbro, Petrin, Hulstrand, Standley, Brouillette, Long, Mansilla, Cornell, Murray, et al., 2018). SM1 is expressed in the same synexpression group as other clefting genes, including spry1, spry2, and fgf8, and knockdown of ISM1 causes craniofacial dysmorphologies in frogs, including cleft-like phenotypes, hence confirming the role of ISM1 in craniofacial development Additionally, trophoblastic cells in placental tissues collected from patients with gestational hypertension and preeclampsia have a strong expression of ISM1, while reduced serum concentrations of ISM2 were observed in preeclampsia patients and, quite the opposite, ISM2 showed prominent expression in choriocarcinoma, thus suggesting a possible contrast in function. Furthermore, strong expression of ISM2 is observed in choriocarcinoma, while moderate expression in lung and prostate adenocarcinoma is observed and mild expression is indicated in colon adenocarcinoma and cohesive gastric carcinoma.
In summary, ISM1 upregulation and overexpression are reported in multiple cancers, including gastric cancer, hepatocellular carcinoma (HCC), colon adenocarcinoma, and colorectal cancer, while ISM2 overexpression was reported in choriocarcinoma. However, multiple signalling molecules and pathways are also reported as being involved in the regulation and synexpression. Even though ISM1 is reported as an endogenous angiogenesis inhibitor, it is quite the opposite that overexpression of ISM1 and ISM2 is found in the above-mentioned cancers, while it could be interesting to explore whether tight regulation over the tissue-specific expression of these proteins might play a role in cancer therapy.
## Metabolism
Endocrine tissues have enhanced expression of ISM1, including the thyroid, pituitary, and adrenal glands, whereas the thyroid gland is reported to have the highest expression of this protein. Furthermore, as discussed earlier, increased expression of ISM1 is found in adipose tissue, pancreas, spleen, liver, and kidney compared to other organs. Secreted by adipose tissue, adipokines participate in systemic metabolism and tissue homeostasis in a paracrine or endocrine manner, while imbalance in adipokines causes metabolic dysfunction. A recent study [bib_ref] Isthmin-1 is an adipokine that promotes glucose uptake and improves glucose tolerance..., Jiang [/bib_ref] reported ISM1 as an adipokine secreted by mouse adipocytes that promotes glucose uptake by a signalling cascade which involves an unidentified receptor kinase such as insulin. Furthermore, recombinant ISM1 led to a vigorous rise in GLUT4-dependent glucose uptake in both murine and human adipocytes as well as primary human muscle cells, whereas Ism1 knockout in adipocytes exhibited a reduction in glucose uptake and insulin-dependent phosphorylation of protein kinase AKT at the serine residue 473 [bib_ref] Isthmin-1 is an adipokine that promotes glucose uptake and improves glucose tolerance..., Jiang [/bib_ref].
Furthermore, complete knockout of Ism1 in mice exhibited a drop in glucose tolerance as well as a decline in glucose uptake in the brown adipose tissue and muscle, while on the other hand, overexpression of Ism1 in diet-induced mice led to the reduction in adipose tissue mass along with improved glucose tolerance, insulin sensitivity, and boosted suppression of hepatic glucose production in an insulin clamp [bib_ref] Isthmin-1 is an adipokine that promotes glucose uptake and improves glucose tolerance..., Jiang [/bib_ref].
On the other hand, even though mouse adipocytes and hepatocytes exhibited ISM1dependent signalling cascade associated improvised glucose uptake, astonishingly ISM1 suppressed insulin-dependent de novo lipogenesis (DNL). In addition, it was suggested that ISM1 facilitated hepatic DNL inhibition, which prevented lipid accumulation in an NAFLD mouse model, and thus the authors argued that ISM1 signalling indirectly modulates the hepatic lipid accumulation by inhibiting the fatty acid releases from the adipose tissue [bib_ref] Isthmin-1 is an adipokine that promotes glucose uptake and improves glucose tolerance..., Jiang [/bib_ref].
Furthermore, a study conducted on Spanish pubertal boys and girls revealed higher ISM1 levels in obese pubertal boys as compared to normal weight and overweight boys, while there were no significant changes observed in pubertal girls, thus indicating a plausible function of ISM1 in male restricted obesity.
However, a cohort study conducted on Canadian young adults with childhood obesity history revealed a positive correlation of obesity with risks of T2DM and coronary artery disorders [bib_ref] Childhood obesity leads to adult type 2 diabetes and coronary artery diseases, Fang [/bib_ref]. In contrast and as previously mentioned, placental tissues from patients with gestational hypertension and preeclampsia are associated with a strong expression of ISM1. However, a cohort study conducted on women with gestational diabetes mellitus (GDM) demonstrated that women who developed preeclampsia did not show more insulin resistance as compared to non-preeclamptic women [bib_ref] Insulin Resistance and Preeclampsia in Gestational Diabetes Mellitus, Montoro [/bib_ref]. However, ISM1 function in glucose homeostasis is enigmatic, no receptor has been identified so far for this adipokine in functional studies, and there is also an absence of studies reporting on ISM2 and its role in metabolism and any other related processes.
## Immune system
Haematopoiesis is the formation of blood cells in the marrow that involves the replication as well as specialization of hematopoietic stem cells (HSCs) towards the downstream progenitor cells [bib_ref] Concise Review: Epigenetic Regulation of Hematopoiesis: Biological Insights and Therapeutic Applications, Antoniani [/bib_ref]. Multiple classes of secreted proteins, among them the colony stimulating factors (CSFs) or haematopoiesis, are known to function as important regulators during hematopoietic differentiation. Furthermore, these secreted molecules are of clinical significance because of their application in stimulating haematopoiesis in patients with neutropenia and several other haematological diseases [bib_ref] FAM20: An evolutionarily conserved family of secreted proteins expressed in hematopoietic cells, Nalbant [/bib_ref]. Consequently, identifying the function of novel secreted proteins that show specific spatiotemporal expression patterns within the hematopoietic cells is of considerable interest. ISM1 is reported as required for the normal formation of HSPCs towards their downstream progeny during haematopoiesis in zebrafish because Ism1 knockout resulted in the drop in frequency of mature blood cell populations, including neutrophil, macrophage, and erythrocyte, while elevated expression of Ism1 was reported in stromal cell lines [bib_ref] Isthmin 1 (ism1) is required for normal hematopoiesis in developing zebrafish, Berrun [/bib_ref]. However, the study was conducted on a zebrafish morpholino-based model, and it remains unclear how ISM1 interacts with HSPCs and the underlying mechanism requiring ISM1 during haematopoiesis in zebrafish [bib_ref] Isthmin 1 (ism1) is required for normal hematopoiesis in developing zebrafish, Berrun [/bib_ref].
Moreover, ISM1 was found to be produced in human and mice barrier tissues, such as mucosa, skin, and some lung lymphocytes, which might be linked to the NK, NKT, and Th17 cell lineages. In addition, the ISM1 expression increases significantly in CD4+ T cells when the cells are polarized to the Th17 lineage in vitro, indicating that ISM1 is a mediator of lymphocyte effector functions and might be involved in both innate and adaptive immune responses [bib_ref] Isthmin 1 Is a Secreted Protein Expressed in Skin, Mucosal Tissues, and..., Valle-Rios [/bib_ref].
On top of that, a recent study found a role of ISM1 in immunity against viral infections, as its expression was induced by Grass carp reovirus (GCRV) both in vitro and in vivo, while they further demonstrated that rISM1 reduced the cytopathic effects in GRCV-infected cells and promoted the Ifn gene and IFN-inducible antiviral protein Mxa gene [bib_ref] Zebrafish Ism1 is a novel antiviral factor that positively regulates antiviral immune..., Li [/bib_ref]. Additionally, the same study proposed that ISM1 may first induce the activation of the Tbk1-Irf3-Ifn antiviral signalling pathway, which then leads to enhanced expression of Ifn and Mxa, both of which in turn suppress viral replication [bib_ref] Zebrafish Ism1 is a novel antiviral factor that positively regulates antiviral immune..., Li [/bib_ref].
Additionally, as discussed earlier, a study demonstrating ISM1 upregulation in CRC patients also revealed the associations between suppressive immune cells (M2 macrophages, T-regs, and T cell exhaustion), and ISM1 overexpression detected in PD1-resistant patients further indicated that ISM1 upregulation can possibly play a crucial role in the formation of an inhibitory immune microenvironment [bib_ref] Effect of ISM1 on the Immune Microenvironment and Epithelial-Mesenchymal Transition in Colorectal..., Wu [/bib_ref].
## Mechanism and receptors
Secreted proteins can play a direct role in the regulation of biological processes, including cell proliferation, development, adhesion, migration, and apoptosis [bib_ref] Numerous growth factors, cytokines, and chemokines are secreted by human CD34+ cells,..., Majka [/bib_ref]. As from the above-mentioned studies and discussions, it is clear that ISM1 is involved in multiple biological processes.
However, since ISM1 is a circulating ligand, it is crucial to elucidate its downstream tissue-specific functions mechanistically. In an earlier study [bib_ref] Isthmin is a novel secreted angiogenesis inhibitor that inhibits tumour growth in..., Xiang [/bib_ref] , the authors reported that ISM1 can bind to αvβ5, although they did not demonstrate the signalling properties of ISM1, while further suggesting that ISM1 can function either as an antagonist or agonist which modulates EC apoptosis or survival depending on its physical state.
A previous study conducted by our lab revealed that ISM1 regulates the NODAL signalling by binding to activin receptor type IB during chicken foetal development and is involved in organ asymmetry. Furthermore, it is noteworthy that ISM1 has minimal effect on ACTIVIN-A, BMP4, or TGF-β1 signalling but significantly inhibits NODAL-induced phosphorylation of SMAD2 [bib_ref] ISM1 regulates NODAL signaling and asymmetric organ morphogenesis during development, Osório [/bib_ref].
In addition, a previous study reported that ISM1 activates the Src pathway through the novel ISM-GRP78 apoptosis pathway, triggering apoptosis not only in endothelial cells but also in in cancer cells with an elevated expression of cell-surface GRP78 [bib_ref] Isthmin targets cell-surface GRP78 and triggers apoptosis via induction of mitochondrial dysfunction, Chen [/bib_ref].
Moreover, a critical role of ISM1-csGRP78 signalling is reported in lung homeostasis by controlling alveolar macrophage (AM) number and function, whereas the association between ISM1 and COPD pathogenesis in mice was observed as rISM1 delivery suppressed inflammation, attenuated emphysema, and preserved lung function by specifically targeting csGRP78 on stress-activated csGRP78 high AMs in CS-induced COPD mice [bib_ref] ISM1 protects lung homeostasis via cell-surface GRP78-mediated alveolar macrophage apoptosis, Lam [/bib_ref]. Furthermore, ISM1 is also reported to promote vascular permeability through cell-surface GRP78-mediated Src activation [bib_ref] Isthmin is a novel vascular permeability inducer that functions through cell-surface GRP78-mediated..., Venugopal [/bib_ref]. Additionally, ISM1 is reported to promote the Tbk1-Irf3-Ifn antiviral signalling pathway, which then leads to enhanced expression of Ifn and Mxa, both of which in turn suppress viral replication [bib_ref] Zebrafish Ism1 is a novel antiviral factor that positively regulates antiviral immune..., Li [/bib_ref].
Of note, it was [bib_ref] Isthmin-1 is an adipokine that promotes glucose uptake and improves glucose tolerance..., Jiang [/bib_ref] reported that ISM1, as an adipokine secreted by mouse adipocytes, promotes glucose uptake by induction of the PI3K-AKT pathway, which requires mTORC2.
In zebrafish, ISM1 upregulation of ISM1 was observed to be linked to the overexpression of Wnt signalling during embryonic development [bib_ref] The Sp1-Related Transcription Factors sp5 and sp5-like Act Downstream of Wnt/β-Catenin Signaling..., Weidinger [/bib_ref]. It is noteworthy that Ism1 synexpression with FGF8 was observed in Xenopus embryos [bib_ref] Isthmin is a novel secreted protein expressed as part of the Fgf-8..., Pera [/bib_ref] , while Ism1 was identified as a NODAL controlled gene [bib_ref] Nodal signaling activates differentiation genes during zebrafish gastrulation, Bennett [/bib_ref].
Furthermore, Wnt3a is substantially enhanced in colon cancer cells in promoting tumour angiogenesis and metastasis [bib_ref] Wnt3a expression is associated with epithelial-mesenchymal transition and promotes colon cancer progression, Qi [/bib_ref]. A previous study [bib_ref] miR-1307-3p overexpression inhibits cell proliferation and promotes cell apoptosis by targeting ISM1..., Zheng [/bib_ref] has shown that miR-1307-3p inhibits the activation of the Wnt3a/β-catenin signalling pathway via targeting ISM1 downregulation and thus inhibiting proliferation and promoting the apoptosis of COAD cells. Conversely, ISM1 overexpression promoted the activation of the Wnt3a/β-catenin signalling pathway along with the proliferation and decelerated cell apoptosis in COAD cells. Furthermore, another study has also reported the elevated expression of ISM1 in the CRC tissue of patients, and on the top of that, the positive correlation of ISM1 with cancerassociated signalling pathways including EMT, hypoxia, KRAS, Notch, and Hedgehog was observed [bib_ref] Effect of ISM1 on the Immune Microenvironment and Epithelial-Mesenchymal Transition in Colorectal..., Wu [/bib_ref].
Transforming growth factor-β (TGF-β) is a conserved family of secreted polypeptide factors which regulate several aspects of physiological embryogenesis and adult tissue homeostasis. This family includes many important proteins, such as TGF-β, nodal, lefty, activin, growth and differentiation factor, and bone morphogenetic proteins [bib_ref] TGF-beta signaling: A complex role in tumorigenesis (Review), Liu [/bib_ref]. Taking into account previous findings about the relationship of ISM1 with the nodal [bib_ref] ISM1 regulates NODAL signaling and asymmetric organ morphogenesis during development, Osório [/bib_ref] and TGF-β [bib_ref] ISM1 protects lung homeostasis via cell-surface GRP78-mediated alveolar macrophage apoptosis, Lam [/bib_ref] pathways and the importance of these pathways in several crucial processes, it is important to further explore and elucidate ISM1's role in the TGF-β pathway, particularly focusing on the interaction between ISM1 and TGF-β receptors.
## Future perspectives
Even though ISM was identified two decades ago, there remains a lack of understanding with regards to the roles of ISM1 and ISM2 in embryonic development and other physiological processes. To address the role of ISM, functional studies have begun but unresolved issues for researchers remain. Initially, its role in development [bib_ref] Identification of Isthmin 1 as a Novel Clefting and Craniofacial Patterning Gene..., Lansdon [/bib_ref] [bib_ref] ISM1 regulates NODAL signaling and asymmetric organ morphogenesis during development, Osório [/bib_ref] [bib_ref] Isthmin1, a secreted signaling protein, acts downstream of diverse embryonic patterning centers..., Kesavan [/bib_ref] and physiology [bib_ref] ISM1 protects lung homeostasis via cell-surface GRP78-mediated alveolar macrophage apoptosis, Lam [/bib_ref] [bib_ref] Isthmin-1 is an adipokine that promotes glucose uptake and improves glucose tolerance..., Jiang [/bib_ref] [bib_ref] Zebrafish Ism1 is a novel antiviral factor that positively regulates antiviral immune..., Li [/bib_ref] has been addressed but still needs further detailed studies.
The truth is, although multiple studies have identified the spatiotemporal expression of ISM during embryonic development of chickens, mice, zebrafish, and Xenopus, the exact roles and signalling cascades working behind the expression are yet to be discovered, except that one study revealed that ISM1 negatively regulates NODAL signalling in asymmetric organ morphogenesis of chicken embryos [bib_ref] ISM1 regulates NODAL signaling and asymmetric organ morphogenesis during development, Osório [/bib_ref]. As the functional studies of ISM have been studied in embryonic development, it is possible that several human birth defects and genetic diseases will be attributed to the mutations of the Ism gene family.
To elucidate the regulatory mechanism of ISM, it is crucial to determine its interactions with other ligands and receptors that lead to the activation of downstream signalling pathways. Even though ISM1 is reported to bind with GRP78 [bib_ref] Isthmin is a novel vascular permeability inducer that functions through cell-surface GRP78-mediated..., Venugopal [/bib_ref] [bib_ref] ISM1 protects lung homeostasis via cell-surface GRP78-mediated alveolar macrophage apoptosis, Lam [/bib_ref] [bib_ref] Isthmin targets cell-surface GRP78 and triggers apoptosis via induction of mitochondrial dysfunction, Chen [/bib_ref] and αVβ5, its diffusion into the ECM to activate specific receptors remains elusive.
Another major frontier will be to elucidate its role in angiogenesis and cancer. ISM1 was demonstrated as an endogenous inhibitor of angiogenesis by two studies [bib_ref] Isthmin is a novel secreted angiogenesis inhibitor that inhibits tumour growth in..., Xiang [/bib_ref] [bib_ref] Isthmin inhibits glioma growth through antiangiogenesis in vivo, Yuan [/bib_ref] , while one study demonstrated that the solubility of the protein itself determines its angiogenic properties. However, both in cells and in intact tissues, receptor-based studies are needed to establish it as an endogenous inhibitor of angiogenesis. Endogenous angiogenesis inhibitors affect the pathological angiogenesis process without affecting quiescent capillary vessels for tissue homeostasis, and thus, based on that, the development of novel anti-angiogenic proteins is needed along with determining the ways to regulate the expression by identification of associated signalling molecules. This can be carried out by generating knockout mice models, particularly tissue specific knock-out studies can be more significant.
Even though ISM1 is reported as anti-angiogenic, multiple studies have reported the prominent expression of ISM1 in several cancers and the elevated expression of ISM2 in choriocarcinoma. At this point, the function of these proteins remains enigmatic, and thus there is a need to explore the bioactivity of ISM as well as the regulatory mechanism involved in the secretion of the protein at both the cellular and tissue level. Additionally, future studies are required to find whether ISM activation itself is an aetiological agent in initiating the primary tumour or a progression factor in cancer pathogenesis. These studies may further lead to the development of pharmacogenetic agents to treat various diseases and cancers.
Recently, Lam et al. [bib_ref] ISM1 protects lung homeostasis via cell-surface GRP78-mediated alveolar macrophage apoptosis, Lam [/bib_ref] found that ISM1 possesses therapeutic potential in vivo, as its administration alleviated pathological conditions in a COPD mouse model. Moreover, dietary intake of rISM1 in N. guentheri was shown to prolong the mean lifespan by 7.5% [bib_ref] Identification of Isthmin1 in the small annual fish, Nothobranchius guentheri, as a..., Li [/bib_ref]. Future investigations should elucidate the appropriate method of delivery, its long-term consequences, and its optimal dosage level for therapeutic applications.
[fig] Figure 1: (Homo sapiens), NP_001012376.1 (Danio rerio), NP_001082228.1 (Xenopus laevis), F1NGZ0 (Gallus gallus), F7AJE9 (Macaca mulatta), A0A8C0Q7L4 (Canis lupus familiaris), XP_016792946.1 (Pan troglodytes), NP_001157407.1 (Bos taurus), A2ATD1 (Mus musculus), and A0A8I5ZXL1 (Rattus norvegicus). The phylogenetic tree of ISM1 (a) and ISM2 (b) proteins were drawn by Molecular Evolutionary Genetic Analysis (MEGA 11.0) using the neighbour-joining method. [/fig]
[fig] Figure 2: Schematic representation of human TSR1 domain sequence alignment of 27 proteins drawn by Molecular Evolutionary Genetic Analysis (MEGA 11.0) using MUSCLE alignment. The accession numbers used for these 27 proteins are mentioned as follows: NP_543016.1 (ISM1), NP_872315.2 (ISM2), NP_922932.2 (ADAMTS6), NP_001269281.1 (ADAMTS10), NP_001311440.1 (ADAMTS12), NP_055087.2 (ADAMTS7), NP_955387.1 (ADAMTS18), NP_620687.2 (ADAMTS16), NP_008969.2 (ADAMTS5), NP_008919.3 (ADAMTS1), NP_001305710.1 (ADAMTS9), NP_079279.3 (ADAMTS20), NP_775733.3 (PAPILIN), NP_620594.1 (ADAMTS13), NP_694957.3 (SBSPON), NP_056019.1 (THSD7A), NP_001033722.1 (RSPO1), NP_001269792.1 (RSPO2), NP_001025042.2 (RSPO4), NP_001138724.1 (PROPERDIN), NP_003957.2 (SEMA5A), NP_001693.2 (ADGRB1), NP_588610.2 (UNC5A), NP_003719.3 (UNC5C), NP_061146.1 (THSD1), NP_116173.2 (RSPO3), and NP_001231818.1 (UNC5B). [/fig]
[fig] Figure 3: Schematic representation of thrombospondin type 1 repeat (TSR1) domain of ISM1 (a) and ISM2 (b). The TSR1 domain is depicted by green-coloured shape. The amino acid sequence of TSR1 domain motifs is shown as coloured with their existing and plausible functions. [/fig]
[fig] Figure 4: Schematic representation of human AMOP domain sequence alignment in 4 proteins drawn by Molecular Evolutionary Genetic Analysis (MEGA 11.0) using MUSCLE alignment. The accession numbers used for these 4 proteins are mentioned as follows: NP_543016.1 (ISM1), NP_872315.2 (ISM2), NP_004523.3 (MUC4), and NP_062547.1 (SUSD2). [/fig]
[fig] Figure 5: Schematic representation of AMOP domain of ISM1 (a) and ISM2 (b). The AMOP domain is depicted by blue-coloured shape. The amino acid sequence of AMOP domain motifs is shown as coloured with their existing and plausible function. [/fig]
[fig] Author: Contributions: H.M.S. and Z.H. equally contributed to the conceptualization, methodology, bioinformatics analysis, writing original draft preparation and editing; Z.Z. edited and directed writing of the manuscript. All authors have read and agreed to the published version of the manuscript. Funding: Innovation and Technology Commission, RGC Theme Based Research. [/fig]
|
The Effect of Upgrades to Childcare Outdoor Spaces on Preschoolers’ Physical Activity: Findings from a Natural Experiment
Physical inactivity is a significant risk factor for childhood obesity. Preventing obesity in the early years reduces the risk of developing chronic health conditions later. Early childhood education and care (ECEC) services are important settings to establish good preschooler physical activity behaviors. This natural experiment investigated the influence of ECEC outdoor physical environment upgrade on preschoolers' physical activity (aged 2-5 years). Centers implemented upgrades without researcher input. Physical activity was measured by 7-day accelerometry for intervention (n = 159; 6 centers) and control (n = 138; 5 centers) groups. ECEC outdoor space was assessed using a modified Environment and Policy Assessment and Observation (EPAO) Instrument. Key outcomes were measured at baseline and 6-12 months follow-up. Fixed sandboxes, balls, portable slides, portable floor play equipment (e.g., tumbling mats), and natural grassed areas were positively associated with activity levels; fixed tunnels and twirling equipment were negatively associated with activity levels (all p < 0.05). Post-upgrade portable play equipment (balls, twirling equipment, slides, floor play equipment) increased intervention preschoolers' moderate-vigorous physical activity (MVPA) levels compared to control (p < 0.05). Intervention preschoolers were more active than control at follow-up (58.09 vs. 42.13 min/day increase in total physical activity; 30.46 vs. 19.16 min/day increase in MVPA (all p < 0.001)). Since few preschoolers meet daily activity recommendations while at ECEC, the findings may help ECEC providers to optimize outdoor physical environments and encourage more active play among preschoolers.
# Introduction
The number of children attending some form of early childhood education and care (ECEC) in most developed countries has increased dramatically in recent decades [bib_ref] Correlates of children's objectively measured physical activity and sedentary behavior in early..., Tonge [/bib_ref]. According to the OECD (Organization for Economic Co-operation and Development), the number of Western European children enrolled in ECEC increased from 20% to 90% over a [bib_ref] Canadian 24-Hour Movement Guidelines for Children and Youth: An Integration of Physical..., Tremblay [/bib_ref] [bib_ref] The Brain + Body Equation: Canadian kids need active bodies to build..., Participaction [/bib_ref] year period between 1994 and 2014. In Australia, 54% of 2-3-year-olds and 85% of 4-5-year-olds attend ECEC; in the United States, 40% of 3-year-olds are enrolled in ECEC; in Canada, 54% of children aged 4 years and under are enrolled; and in France, 100% of 3-year-olds are enrolled.
In 2015, excess body weight resulted in about 4 million deaths and 120 million disability-adjusted life-years worldwide. Promoting physical activity behaviors in early childhood is crucial to preventing obesity and is an international priority. Childhood overweight and obesity are associated with an increased risk of developing several preventable health conditions, such as cardiovascular disease, diabetes, depression, arthritis, and premature mortality. In addition, overweight and obese children are more likely to remain obese as they enter adulthood. Regular physical activity mitigates such risks and is also associated with various positive physical health outcomes (such as improved bone health and cardiovascular fitness), socio-emotional development (such as enhanced social skills and emotional intelligence), mental health (such as reduced depression and anxiety problems), and better sleep in children [bib_ref] Physical activity for preschool children-How much and how?, Timmons [/bib_ref] [bib_ref] Resurrecting free play in young children looking beyond fitness and fatness to..., Burdette [/bib_ref]. Early childhood is a critical time for establishing healthy patterns of wellbeing and physical activity behaviors [bib_ref] Correlates of children's objectively measured physical activity and sedentary behavior in early..., Tonge [/bib_ref] [bib_ref] Does early physical activity predict body fat change throughout childhood?, Moore [/bib_ref] [bib_ref] Tracking of physical activity in young children, Pate [/bib_ref] ; therefore, there is a significant opportunity to influence behaviors of young children while attending ECEC [bib_ref] Physical Activity at Child Care Settings: Review and Research Recommendations, Ward [/bib_ref].
Recently, the World Health Organization (WHO, Geneva, Switzerland) and countries such as Australia, Canada, and New Zealand released 24-h movement guidelines for the early years [bib_ref] Canadian 24-Hour Movement Guidelines for Children and Youth: An Integration of Physical..., Tremblay [/bib_ref]. The Guidelines recommend children aged 2-5 years should spend at least 180 min in a variety of physical activity each day, of which 60 min is energetic play. Estimates of the proportion of young children meeting physical activity guidelines vary considerably. In Australia, 34% of preschoolers achieve recommended physical activity levels each day. In Canada, 62% of three to four-year-old preschoolers meet physical activity guidelines [bib_ref] The Brain + Body Equation: Canadian kids need active bodies to build..., Participaction [/bib_ref]. These differences in the proportion of preschoolers meeting physical activity guidelines could be due to differing methodologies (subjective vs. objective measures), in particular, different cut-points for light-intensity physical activity. Despite this, most studies show that preschoolers accumulate most of their physical activity in low-intensity activity [bib_ref] A review of preschool children's physical activity and sedentary time using objective..., Hnatiuk [/bib_ref] [bib_ref] Objectively measured physical activity in four-year-old British children: A cross-sectional analysis of..., Hesketh [/bib_ref] [bib_ref] Physical activity and sedentary behavior during the early years in Canada: A..., Colley [/bib_ref] and a significant proportion do not meet the guidelines.
Outside of the home, ECEC is an important setting to promote young children's physical activity. Correlates of preschoolers' physical activity while at ECEC include child age (older children are often more active), sex (boys are often more active), better fundamental motor coordination, educator behaviors (i.e., prompts and feedback), and the provision of active opportunities for physical activity and certain features of the outdoor environment (e.g., size, use of and presence of portable play equipment like bikes and balls), and these correlates have all shown positive associations [bib_ref] Correlates of children's objectively measured physical activity and sedentary behavior in early..., Tonge [/bib_ref]. Modifying the outdoor physical environment has significant potential to influence preschoolers' physical activity while attending ECEC [bib_ref] Correlates of children's objectively measured physical activity and sedentary behavior in early..., Tonge [/bib_ref]. However, there are few intervention studies involving changes to the ECEC outdoor physical environment as such studies are difficult and expensive to implement [bib_ref] A natural experiment to examine the impact of park renewal on park-use..., Veitch [/bib_ref]. Findings from these intervention studies have been mixed. A recent Canadian study found that adding novel portable equipment, in addition to staff training and modifying outdoor playtime, increased objectively measured physical activity levels in children attending the ECECs that had been modified [bib_ref] Impact of the Supporting Physical Activity in the Childcare Environment (SPACE) Intervention..., Tucker [/bib_ref]. A similar finding was reported by a US study, which found that adding portable play equipment to the ECEC playground increased children's physical activity [bib_ref] Increasing preschoolers' physical activity intensities: An activity-friendly preschool playground intervention, Hannon [/bib_ref]. Another US study found adding pathways and improving the overall quality of the physical environment increased physical activity levels in children [bib_ref] Childcare outdoor renovation as a built environment health promotion strategy: Evaluating the..., Cosco [/bib_ref]. Conversely, a Belgian study found adding portable play equipment and playground markings were not positively associated with physical activity [bib_ref] De Bourdeaudhuijm, I. Promoting physical activity at the pre-school playground: The effects..., Cardon [/bib_ref]. The mixed findings from previous studies could be attributed to methodological differences including differing study designs, small sample sizes, subjective measures of children's physical activity, and short follow-up periods. More rigorous intervention studies are required to address the evidence gap and better understand how changes to the ECEC outdoor physical environment impacts upon children's activity levels.
Due to the high expense and difficulty in implementation, no such studies have been conducted in Australia. Natural experiments using quasi-experimental research designs have been identified as a research priority for demonstrating the causal relationships between built environments and physical activity [bib_ref] Work group IV: future directions for measures of the food and physical..., Story [/bib_ref]. The Western Australian Play Spaces and Environments for Children's Physical Activity (PLAYCE) study provided a unique opportunity to conduct a natural experiment in ECEC. This study had two aims: [bib_ref] Correlates of children's objectively measured physical activity and sedentary behavior in early..., Tonge [/bib_ref] to measure the change in the ECEC outdoor physical environment and (2) examine the impact of changes to the ECEC outdoor physical environment on children's physical activity levels while attending care.
# Materials and methods
## Study design and participants
This was a sub-study of the Play Spaces and Environments for Children's Physical Activity (PLAYCE) study. The PLAYCE cross-sectional study investigated the relative and cumulative influence of the ECEC, home, and neighborhood environment on preschoolers' physical activity. Full details of the PLAYCE study methods have been published [bib_ref] Influence of the day care, home and neighbourhood environment on young children's..., Christian [/bib_ref]. Briefly, between 2015 and 2017, 1596 preschoolers aged 2-5 years and their parents were recruited from 104 ECEC services across metropolitan Perth, Western Australia. The sampling and recruitment of centers were stratified by socioeconomic areas (low, medium and high) and size of center [bib_ref] Influence of the day care, home and neighbourhood environment on young children's..., Christian [/bib_ref]. Tiered consent was utilized, requiring consent from center directors first before parental consent (response rate 24%). Preschoolers were excluded if they were attending full-time school and had any intellectual, emotional, physical, or behavioral disabilities, which hindered participation in physical activity. An ethics amendment to the PLAYCE study for this natural experiment study was granted by The University of Western Australia Human Research Ethics Committee (#RA/4/1/7417).
This current study was an opportunistic natural experiment conducted between 2016 and 2017. The study used a pre-post-test design with intervention and matched control centers to evaluate the effect of changes to the ECEC outdoor physical environment on children's physical activity behavior while at ECEC. Centers were invited to join the intervention group if they informed the study team that they were undertaking a major upgrade of their outdoor physical environment within six months of their first data collection, which was conducted as part of the main study. This was considered the baseline assessment for the current study. After intervention centers were recruited, control centers were then selected based on two criteria: (1) had their baseline assessment completed between July and December 2016, and (2) matched one-to-one to intervention centers based upon SES tertile (low, middle, or high). Matching of centers on center SES was done to ensure that intervention and control centers had access to similar funding and thereby should have similar outdoor physical environment features at baseline. Once centers provided consent to a follow-up assessment of their center, all parents within the center were invited to participate, i.e., parents who had previously participated in the PLAYCE study and parents who had not previously participated in the PLAYCE study. The current study was an opportunistic pilot natural experiment nested within the larger PLAYCE cross-sectional study; it was not the original design of the PLAYCE study to evaluate the impact of ECEC interventions on children's physical activity. Therefore, not all children who completed the baseline assessments as part of the PLAYCE study were available for follow-up assessment as part of the current study, the main reasons for this included children leaving the center to commence kindergarten or full-time school, or parent unwillingness to re-participate. Therefore, 'new children' were recruited to the study at follow-up to ensure a sufficient sample size.
Six intervention centers (4 middle SES, 2 low SES; 4 large size, 1 middle size, and 1 small size) with 159 intervention children and 5 control centers (3 middle SES, 2 low SES; 3 large size, 1 middle size, and 1 small size) with 138 children participated [fig_ref] Figure 1: Overview of study design and recruitment process [/fig_ref]. As control centers were recruited after the baseline assessment was completed, only five centers met the selection criteria to be included as control centers: two intervention centers from the middle SES tertile were matched to one control center of the same SES tertile.
## Instruments
## Physical activity
Physical activity was objectively measured using ActiGraph GTX3+ (ActiGraph, Pensacola, US) accelerometers, which have demonstrated validity for measuring physical activity in young children
## Instruments
## Physical activity
Physical activity was objectively measured using ActiGraph GTX3+ (ActiGraph, Pensacola, US) accelerometers, which have demonstrated validity for measuring physical activity in young children [bib_ref] Measurement of physical activity in preschool children, Pate [/bib_ref]. A 15 s sampling interval (epoch) to accommodate the typical nature of children's physical activity was used. Physical activity intensity was classified based on the following cut points developed by Pate and colleagues: sedentary (<200 counts/15 s), light intensity (200-419 counts/15 s), and moderate to vigorous intensity physical activity (MVPA) (420 counts/15 s) [bib_ref] Validation and calibration of an accelerometer in preschool children, Pate [/bib_ref] ; the cut points have been validated for use in preschoolers by Trost and colleagues. Non-wear time was defined as strings of consecutive zero counts lasting 20 min or longer. ECEC monitoring days were considered valid based on at least 1 day at ECEC with 75% wear time [bib_ref] Influence of the day care, home and neighbourhood environment on young children's..., Christian [/bib_ref]. Wear time was calculated by taking non-wear time from monitoring time. Accelerometers were worn on the right hip for 7 days during waking hours. Parents reported the days and times their child attended ECEC in the accelerometer diary. Only accelerometer data for the time children were at ECEC were analyzed for which there was a minimum of at least one day at day ECEC with 75% wear time [bib_ref] Influence of the day care, home and neighbourhood environment on young children's..., Christian [/bib_ref].
A diary was used to record if the accelerometer was removed, the amount of time the accelerometer was worn and the days, and times children attended ECEC. Accelerometer data were analyzed using a custom-built excel and SAS macro developed for analyzing the intensity and amount of physical activity children accumulate while attending ECEC and overall. Data for the average minutes of total physical activity (TPA) (calculated as sum of light physical activity and MVPA) and MVPA per ECEC day were used in analyses.
## Ecec outdoor physical environment
The main PLAYCE study measured the ECEC outdoor physical environment using a slightly modified version of the Environment and Policy Assessment and Observation (EPAO) Instrument'Physical Environment' subscale to ensure relevance for the Australian context [bib_ref] Influence of the day care, home and neighbourhood environment on young children's..., Christian [/bib_ref]. The EPAO instrument was developed to assess the quality of ECEC centers physical activity and eating behaviors in the US [bib_ref] An instrument to assess the obesogenic environment of child care centers, Ward [/bib_ref] and has been validated [bib_ref] The childcare environment and children's physical activity, Bower [/bib_ref] [bib_ref] Nutrition and physical activity in child care: Results from an environmental intervention, Ward [/bib_ref]. The revised tool has been previously described in the PLAYCE study methods paper [bib_ref] Influence of the day care, home and neighbourhood environment on young children's..., Christian [/bib_ref]. The audit tool assessed the presence of physical activity opportunities in the physical environment [bib_ref] Interaction between physical environment, social environment, and child characteristics in determining physical..., Gubbels [/bib_ref]. The revised audit items have been shown to have excellent intra-class correlation coefficients (ICCs) for intra-rater reliability, and good to excellent ICCs for inter-rater reliability. For intra-rater reliability of the outdoor physical environmental audit, the 'Fixed play equipment' subscale ICC was 0.94, the 'Portable play equipment' subscale was 0.94, the 'Natural physical features' subscale was 0.84, and the 'Outdoor play spaces' subscale was 0.80. For inter-rater reliability of the outdoor physical environmental audit, the 'Fixed play equipment' subscale ICC was 0.7, the 'Portable play equipment' subscale was 0.74, the 'Natural physical features' subscale was 0.79, and the 'Outdoor play spaces' subscale was 0.79.
The current study included data from five outdoor physical environment subscales in the audit tool: 'Fixed play equipment' and 'Portable play equipment' from the EPAO, 'Total size of playing area' [bib_ref] Play Equipment, Physical Activity Opportunities, and Children's Activity Levels at Childcare, Gubbels [/bib_ref] , 'Outdoor play spaces' [bib_ref] Physical activity in children attending preschools, Olesen [/bib_ref] [bib_ref] Behaviour mapping: a method for linking preschool physical activity and outdoor design, Cosco [/bib_ref] , and 'Natural elements' [bib_ref] Physical activity in children attending preschools, Olesen [/bib_ref] [bib_ref] Behaviour mapping: a method for linking preschool physical activity and outdoor design, Cosco [/bib_ref]. The physical environment subscales were created using the scoring tool and guidelines provided by the original authors. Each subscale was created at the child level. While some intervention centers had more than one outdoor area that a child could access, upgrades were implemented in all outdoor areas, therefore, all children were exposed to the upgrade intervention.
The 'Fixed play equipment' subscale score was based on the availability of eight types of equipment: structured tracks (e.g., playground markings), climbing structures (e.g., jungle gyms), see-saws, slides, tunnels, balancing surfaces (e.g., balance beams), sandboxes, and swinging equipment (e.g., swings, ropes). Merry-go-rounds measured in the EPAO were not included in analyses as none of the ECEC outdoor environment had this feature. Items were coded 1 if present and 0 if not; total score for 'Fixed play equipment' was calculated as the sum, divided by eight (number of items) and multiplied by 10 (to obtain a score from 0 to 10, higher score indicated more activity opportunities). The 'Portable play equipment' subscale score was calculated in a similar manner, and included nine items: balls, climbing structures (e.g., ladders), floor play equipment (e.g., tumbling mats), jumping equipment (e.g., jump ropes, hula hoops), push/pull toys (e.g., wagons), riding toys (e.g., tricycles, cars), slides, sand/water toys (e.g., buckets, scoops), and twirling equipment (e.g., ribbons, batons). 'Total playing area' was rated on a scale from 0 (no playing area) to 10 (very large area) by comparing all ECEC outdoor playing areas and dividing them into 10 tertiles. The 'Natural elements' subscale score was calculated in a similar manner and included eight items: mature trees, other plants, vegetable/herb/fruit garden, rocks/stones/pebbles, natural grassed area, artificial grassed area, potted plants, and flower beds. The 'Outdoor play spaces' subscale score was also calculated in a similar fashion, and included five items: open areas, water play areas, sloping ground, a variety of ground surfaces (e.g., mulch, artificial covering), and playground constructed at different levels.
Research personnel conducting the environmental audits were trained in using the EPAO. A senior research assistant was present on site to answer any queries and ensure consistency. Intra-rater and inter-rater reliability intraclass correlations (ICCs) were good to excellent (ICC = 0.70-0.94).
## Confounders
Child age, sex, and parental education variables were collected via established items in the PLAYCE parent survey [bib_ref] Influence of the day care, home and neighbourhood environment on young children's..., Christian [/bib_ref].
# Statistical analysis
All analyses were conducted using SPSS V.24, p values < 0.05 were considered statistically significant. Analyses were conducted at the child level to adjust for child-related confounders (age, sex, parent education, and accelerometer wear time). All data were analyzed using the intention-to-treat principle.
Descriptive statistics were used to examine various background characteristics. The significance of differences in preschooler activity levels between groups and within groups was examined using t-tests. The distribution of the outdoor physical environment measures was explored and the significance of differences between features at baseline and follow-up were examined using t-tests. Effect sizes were interpreted using the classification defined by Cohen [bib_ref] A power primer, Cohen [/bib_ref].
Multiple multivariate linear regression analyses with three levels (i.e., child level; room level; center level) were conducted to examine the associations between the ECEC outdoor physical environment features and child characteristics as independent variables, and physical activity levels (TPA and MVPA) as outcome variables, and random intercepts at the child, room, and center level. Analyses were conducted with time (baseline and follow-up), group (intervention or control), and the interaction between group and time as covariates. Models were run separately for each ECEC outdoor physical environment subscale: 'Fixed play equipment' and 'Portable play equipment', 'Total size of playing area, 'Outdoor play spaces' and 'Natural elements'. Insignificant independent variables were stepwise deleted from the model in order of their significance, starting with the least significant variable. This procedure was repeated until all remaining independent variables were significant.
# Results
## Participant baseline characteristics
At baseline, the average age of participants was 2 years 10 months (±SD = 0.82) and 51.1% were girls (results not shown). There were no significant differences between intervention and control group sociodemographic characteristics at baseline except that the intervention group had fewer parents with a bachelor's degree or higher qualification compared with the control group (44% vs. 59%, p = 0.04). [fig_ref] Table 1: Child activity levels at baseline and follow-up [/fig_ref] All physical activity differences within groups were significant (p < 0.05) with a large effect size in favor of follow-up (d between 0.86 and 1.70) [fig_ref] Table 1: Child activity levels at baseline and follow-up [/fig_ref]. Children in both groups accumulated more activity at follow-up, but it was higher in the intervention group (TPA and MVPA). No significant effects were found for between-group differences for change in activity levels between follow-up and baseline. [fig_ref] Table 2: Early childhood education and care [/fig_ref] provides an overview of changes in ECEC outdoor physical environment features between intervention and control groups, and within groups between baseline and follow-up. At baseline, the intervention group had higher scores for all subscales compared with the control group. Environmental scores for all subscales, except the control group's 'Total Outdoor Playing Area' subscale score, decreased at follow-up for both intervention and control groups. All within-group differences, except 'Total Outdoor Playing Area', between baseline and follow-up measurements for intervention and control groups were significant (p < 0.05). All effect sizes were in favor of baseline: the intervention group had a larger effect size than the control group (d = −0.95 to −0.79 vs. −0.77 to −0.43). Between-group effect sizes were small (d between −0.12 to −0.32) and in favor of the control group. Significant mean changes were found for 'Portable Play Equipment' and 'Outdoor Play Spaces' subscale scores (p < 0.05): for both subscales, the control group's score decreased less than the intervention group's. [fig_ref] Table 3: Multivariate regression analyses of the association between the ECEC outdoor physical environment,... [/fig_ref] shows the associations of the ECEC outdoor physical environment and child background factors with outdoor physical activity levels. Fixed sandbox (Model 1), balls (Model 2), and presence of natural grassed area (Model 3) were positively associated with TPA and MVPA. Portable slides (Model 2) and portable floor play equipment (e.g., tumbling mats) (Model 2) were positively associated with MVPA only. Fixed tunnels (Model 1) and twirling equipment (Model 2) were negatively associated with TPA and MVPA. None of the 'Outdoor Play Spaces' features were significantly associated with children's physical activity levels. Post-upgrade portable equipment comprising of slides, balls, twirling equipment, and floor play equipment was positively associated with more MVPA in intervention children than control children.
## Child physical activity level
## Ecec outdoor physical environment features
## Ecec outdoor environment features and physical activity
There are also significant age and gender effects consistently across all models. Older children accumulated more physical activity outdoors than younger children, and boys were significantly more active than girls. Note. [bib_ref] Correlates of children's objectively measured physical activity and sedentary behavior in early..., Tonge [/bib_ref] Outdoor physical environment subscale score (range from 0 to 10), higher score indicates more activity opportunities; 2 Total outdoor playing area rated on scale of 1-10, 0 (no playing area) and 10 (very large area); * p < 0.05; *** p < 0.001. Ref = control group.
# Discussion
This natural experiment examined the influence of upgrades to the ECEC outdoor space on children's activity levels. Results showed that post-upgrade portable play equipment (balls, slides, twirling equipment, portable floor play equipment like tumbling mats) had a beneficial effect on MVPA levels of the intervention group compared to the control group. This is in line with previous studies that showed that the presence of portable equipment encourages active play in children [bib_ref] Impact of the Supporting Physical Activity in the Childcare Environment (SPACE) Intervention..., Tucker [/bib_ref] [bib_ref] The childcare environment and children's physical activity, Bower [/bib_ref] [bib_ref] Environmental factors associated with physical activity in childcare centres, Henderson [/bib_ref]. Portable play equipment, also known as 'loose parts', refers to open-ended play equipment that children can use in a variety of ways [bib_ref] How not to Cheat Children: the Theory of Loose Parts, Nicholson [/bib_ref]. Children prefer playing equipment with moveable features as they are action-oriented (compared to static equipment) and can be used for various different functions [bib_ref] Equipment Choices of Primary-Aged Children on Conventional and Creative Playgrounds, Frost [/bib_ref]. Therefore, portable equipment encourages variation in activities over time, combatting the possible novelty effect of other new equipment such as fixed play equipment. In addition, floor play equipment such as tumbling mats can be used in a variety of ways to encourage active play like rolling or somersaulting or even provide soft landing surfaces for children to jump onto. The study also found that boys and older children tended to be more active [bib_ref] Correlates of children's objectively measured physical activity and sedentary behavior in early..., Tonge [/bib_ref] [bib_ref] Physical activity among children attending preschools, Ridgers [/bib_ref] [bib_ref] Activating Childcare Environments for All Children: The Importance of Children's Individual Needs, Gubbels [/bib_ref].
In line with previous studies, we found several outdoor environmental features to be positive predictors of children's activity levels. Children were more active (TPA and MVPA) when balls and a natural grassed area were present. Balls provide opportunities for children to engage in play while developing essential fundamental movement skills such as throwing and kicking [bib_ref] The influence of centre-based childcare on preschoolers' physical activity levels: A cross-sectional..., Vanderloo [/bib_ref] [bib_ref] Effectiveness of physical activity interventions for preschoolers: A meta-analysis, Gordon [/bib_ref] [bib_ref] Effects of Child Care Policy and Environment on Physical Activity, Trost [/bib_ref]. Previous research has shown that the presence of natural elements encourages more activity and less sedentary behavior in children [bib_ref] Activating Childcare Environments for All Children: The Importance of Children's Individual Needs, Gubbels [/bib_ref] [bib_ref] Impact of preschool environment upon children's physical activity and sun exposure, Boldemann [/bib_ref] , and could provide more activity opportunities than non-natural environments. However, there are studies that showed negative [bib_ref] Physical activity in children attending preschools, Olesen [/bib_ref] or no association [bib_ref] Are preschool children active enough? Objectively measured physical activity levels, Cardon [/bib_ref] between the presence of vegetation and children's activity. It has been suggested that difference in types of natural elements being assessed could explain the conflicting results from studies [bib_ref] Activating Childcare Environments for All Children: The Importance of Children's Individual Needs, Gubbels [/bib_ref] : functional natural elements such as grass and hills may encourage more activity [bib_ref] Activating Childcare Environments for All Children: The Importance of Children's Individual Needs, Gubbels [/bib_ref] ; conversely aesthetic natural elements such as plants and trees may obstruct open areas available for play [bib_ref] Play Equipment, Physical Activity Opportunities, and Children's Activity Levels at Childcare, Gubbels [/bib_ref] [bib_ref] Impact of preschool environment upon children's physical activity and sun exposure, Boldemann [/bib_ref] [bib_ref] Schoolyard characteristics, physical activity and sedentary behaviour: Combining GPS and accelerometry, Van Kann [/bib_ref]. Grassed areas can be used in a variety of ways such as running, playing with balls or games such as 'follow the leader'. Future research into the functionality of various natural features in ECEC would be helpful to optimize future outdoor space design.
The presence of fixed tunnels and twirling equipment were negative environmental predictors of children's activity levels. Tunnels allow for crawling/sliding movement that is not as active as running but is still important for developing fundamental movement skills. Our results suggest that children were engaged in mostly light activity when interacting with tunnels, where they could be playing hide-and-seek or imaginative play. Similarly, playing with twirling equipment could be more creative rather than physical activity promoting.
We found that portable slides and fixed sandboxes were positive predictors of children's activity levels. This was in contrast to Gubbels et al.'s cross-sectional study that found that sandboxes and portable slides were associated with less activity in children attending childcare [bib_ref] Play Equipment, Physical Activity Opportunities, and Children's Activity Levels at Childcare, Gubbels [/bib_ref]. The authors suggested that sandboxes could have a confounding effect on children's activity levels as their results showed a positive bivariate association between the sandbox and children's activity levels [bib_ref] Play Equipment, Physical Activity Opportunities, and Children's Activity Levels at Childcare, Gubbels [/bib_ref]. ECEC centers with sandboxes tend to have larger budgets and larger spaces for play facilities, therefore, they are likely to also have many other outdoor features that are associated with more activity in children. Secondary chi-square tests showed that ECEC centers that had a sandbox also had many other environmental features including those which were found to be positively associated with activity levels (p < 0.001; results not shown). In practice, fixed sandboxes in Australian ECEC centers may also include large play equipment or see-saws/swings, so children could be moving actively between features within the sand play area.
In addition, portable slides can be placed in new and interesting locations within the play area to encourage children to interact with the equipment more, for example, within a large fixed play structure or between balancing beams. Portable slides tend to be shorter than fixed slides, so children may do more 'laps' running back up the stairs to slide down again. These 'heavier type' portable equipment have to be physically manipulated by educators, possibly also encouraging more involvement from educators when children are playing outside. Previous studies have shown positive associations between educator involvement and children's activity levels [bib_ref] Impact of the Supporting Physical Activity in the Childcare Environment (SPACE) Intervention..., Tucker [/bib_ref] [bib_ref] Use of the Environment and Policy Evaluation and Observation as a Self-Report..., Ward [/bib_ref]. Encouraging preschoolers to be more active requires an interaction between the availability of various outdoor space features and educator support. An outdoor environmental space feature on its own may not be positively associated with activity, however, when placed in proximity with other features it may offer more opportunities for play.
Intervention children accumulated significantly more activity (TPA and MVPA) at follow-up compared to baseline. This may be due to the availability of new portable play equipment such as balls, slides, floor play equipment, and twirling equipment at follow-up, which showed to have a beneficial influence on the intervention group's activity levels. However, only a very small proportion of intervention children achieved the recommended three hours of physical activity while in ECEC at follow-up (4.44% of the intervention group; 6.25% of the control group). Considering that a large proportion of children attend ECEC services, it is concerning that only a small proportion of young children are sufficiently physically active. More focus should be given to support children in accumulating sufficient amounts of physical activity while attending ECEC. The results of this study suggest a number of physical environmental features of ECEC that could be targeted to promote more activity in children attending care. Even though intervention children's TPA levels increased significantly at follow-up, the average TPA accumulated was still below the recommended three hours of physical activity while in ECEC [bib_ref] Canadian 24-Hour Movement Guidelines for Children and Youth: An Integration of Physical..., Tremblay [/bib_ref].
Unexpectedly, control children's activity levels also increased significantly at follow-up (TPA and MVPA). This could be in part due to different children participating at baseline and follow-up: 24 of control children participated at both time points, 35 of them participated only at follow-up. This may have meant that sociodemographic and physical activity differences in children participating at baseline compared with follow-up could have influenced changes in group physical activity levels regardless of any intervention effect from the environmental upgrade. Moreover, the smaller number of children participating at follow-up compared to baseline assessment (approximately 50% less children participated at follow-up for both groups) could have influenced the findings resulting in a less accurate representation of change in children's physical activity post-intervention.
At baseline, the intervention group's outdoor environment provided more activity opportunities than the control group's. However, the intervention group's subscale scores were still relatively low: none of the subscales scored more than 5 out of the total 10 points available. The low levels of activity opportunities in the intervention group's outdoor environments suggest there was a significant opportunity for improvement via an environmental upgrade.
Environmental subscale scores of the intervention group, except 'Total Outdoor Playing Area', significantly changed at follow-up, however, changes were in the unexpected direction. We hypothesized intervention centers would have greater availability of activity-promoting features at follow-up, however, all subscale scores decreased at follow-up. A potential reason could be related to this being a natural experiment, i.e., centers implemented changes to the outdoor space independent of the research team. Intervention centers may have made changes to features of the outdoor space that were not associated with children's physical activity. For example, comparison of before and after photographs taken by study personnel showed that one center replaced a climbing structure with a swing set that has been shown to be associated with lower physical activity levels in children [bib_ref] Play Equipment, Physical Activity Opportunities, and Children's Activity Levels at Childcare, Gubbels [/bib_ref]. The EPAO tool assessed the availability of activity opportunities and thus measured presence of different types of items and not number of individual items. Intervention centers could have purchased many new balls and bicycles, however, this change would not have been captured in the tool. This could explain the significantly large increase in TPA and MVPA in intervention children, even though environmental subscale scores decreased at follow-up. That is, while there may have been less different types (or variety) of activity opportunities in the ECEC outdoor space at follow-up, a sheer increase in the number of existing types of activity promoting features such as more balls and bicycles may have encouraged children to be more active. Furthermore, many of the changes to the outdoor space appeared to be part of center routine maintenance rather than a significant change to the area: 'before and after' photographs show instances where older equipment was replaced with newer equipment, but no major changes were made to the types of features available for children to play with. Such changes may have been viewed by centers as an 'upgrade' as structural changes (such as turf and structures were replaced) were made to the outdoor space, albeit there was no real increase in the number of different types of activity promoting features.
Surprisingly, environmental subscale scores for the control group, except 'Outdoor play space', also changed significantly at follow-up: scores at follow-up were lower than at baseline. We hypothesized that there would be no change in the outdoor physical environment of control centers between baseline and follow-up. A potential reason could be control centers may have made changes to their outdoor space between baseline and follow-up (average of 12 months) but not reported it to study personnel. In practice, ECEC centers tend to undergo an annual maintenance review that includes their outdoor space. Centers may have removed older equipment, which could, in part, explain reduced scores at follow-up.
Unexpectedly, control centers had a smaller reduction in subscale scores compared to intervention centers: the control group had a favorable but small change in subscale scores for 'Portable play equipment' and 'Outdoor play space' at follow-up compared to the intervention group. As discussed earlier, intervention centers may have removed more different types of older portable play equipment as part of their upgrade but not yet replaced them with new equipment at the time of the follow-up assessment. An apparent lack of difference in 'Total outdoor playing area' is consistent with the idea that given outdoor play spaces are fixed areas, it is unlikely that they would change in size significantly after an upgrade. If anything, they may possibly decrease if intervention centers added more fixed play equipment to their outdoor play space. Furthermore, although the outdoor space was measured accurately using a laser distance measure, the calculation of the actual area was difficult due to the unusual shapes of many outdoor spaces. This may have led to small measurement errors. Future studies may consider using methods such as geographic information-based systems to measure the actual size of the outdoor space. Finally, the lack of observed effect of other outdoor environment features (e.g., sloping ground, water play areas) may in part be explained by them only being present in very few ECEC centers.
Even though this study focused on how the ECEC outdoor physical environment influences children's physical activity, future studies should investigate the impact of changes to the ECEC outdoor physical environment on different types of play as well as other health and developmental outcomes [bib_ref] Effects of play equipment and loose parts on preschool children's outdoor play..., Maxwell [/bib_ref]. It may be possible to expand the existing EPAO tool to encompass a child's development outcomes, by also examining the influences of outdoor environmental features on children's socio-emotional development. In line with the holistic development of young children [61], ECEC services should attempt to find a balance between providing outdoor physical environment features that develop gross motor skills as well as those which encourage creative and imaginative play [bib_ref] Effects of play equipment and loose parts on preschool children's outdoor play..., Maxwell [/bib_ref]. Furthermore, in accordance with the social-ecological framework [bib_ref] Translating social ecological theory into guidelines for community health promotion, Stokols [/bib_ref] , the physical activity behavior of children attending ECEC is influenced by not only the physical environment but also pedagogical intentions and effects. Future intervention studies are required to examine the interaction of physical, policy, and practice level interventions on the physical activity behavior of children while attending ECEC.
Finally, cultural differences between Australia (where the study was conducted) and the US (in which the EPAO instrument was developed)mean that some features were present in the Australian ECEC environment but not captured by the EPAO instrument, and vice versa. For example, merry-go-rounds are not common in ECEC outdoor spaces in Australia but are included in the EPAO instrument. Furthermore, natural features such as nature strips/corners or chicken coops (which were installed in intervention centers as part of the environmental upgrade) were not captured in the environmental audit. Future studies could examine the influence of various emerging natural elements on children's activity levels and developmental outcomes.
# Strengths and limitations
A strength of this study was the natural experiment design that included intervention and control groups. Intervention studies examining changes to the built environment are difficult and expensive to implement; therefore, natural experiment is a preferred study design. Furthermore, objective measures of physical activity (accelerometry) and direct observation of physical environment features rather than less reliable methods such as self-report were used. Activity opportunities in ECEC settings were assessed by analyzing a wide range of individual features based on the existing validated observation instrument EPAO rather than summarizing facilities into a single measure.
A limitation was that study personnel could not control the type of features added during the intervention and these varied widely across centers. As a result, many of the changes made to the outdoor space were not activity promoting. Findings may also have been impacted as different children participated at baseline and follow-up. It was not the original intention of the PLAYCE study to evaluate the impact of center upgrade interventions on children's physical activity, thus follow-up children for both intervention and control were recruited after baseline assessments were completed. Baseline physical activity levels of both groups may have impacted the result; however, intervention children accumulated more TPA and MVPA at follow-up than control children despite it being a nonsignificant increase. The study sample did not include ECEC centers from high SES areas, as intervention centers were from low and middle SES areas only. Previous studies have shown that physical activity in young children is not related to SES status [bib_ref] A review of correlates of physical activity of children and adolescents, Sallis [/bib_ref] ; therefore, the lack of representation of children from high SES areas is unlikely to have had a significant impact on the results.
In addition, the study focused only on the influence of the outdoor physical environment on children's activity levels. However, in line with the social-ecological framework, physical activity behaviors of young children are also influenced by educators' physical activity-related practices and ECEC physical activity-related policies. To ensure changes implemented by ECEC services are physical activity promoting, future studies could recruit one or two large ECEC providers (i.e., one provider operating many ECEC centers), instead of many independent ECEC providers. This may help research teams and ECEC providers to work together to design outdoor spaces with features that have been shown through research and practice to positively influence children's physical activity and health. The amount of outdoor playtime provided by ECEC centers could have moderated the results. None of the ECEC centers studied had a policy on outdoor time, however, it is possible that baseline levels of educator-led outdoor play-time may have impacted results. Future research should examine the impact of environmental changes on educators' physical activity related practices (e.g., amount of outdoor play-time, role modeling, and self-efficacy).
One of the aims of this pilot study was to test study recruitment and data collection methods. The follow-up assessment occurred on average later for the control compared with the intervention group as control centers could only be recruited once the intervention centers were confirmed (i.e., when the upgrade was to be completed). The mean time between baseline and follow-up for intervention centers was 10.7 (SD 6.0) months and for control centers was 12.4 (SD 1.1) months. Reasons for the variation in the length of time between baseline data collection and upgrade completion included budgetary constraints, delays in obtaining quotes, plan and budget approvals and installation of the upgrade, and changes to the timeline of the upgrade installation. Lastly, the seasonal variation could have impacted baseline and follow-up measurements, however, almost all centers had their baseline and follow-up assessments in the same season; and data collection was in a similar season for matched intervention and control centers. Furthermore, Perth's (Western Australia) Mediterranean climate means there are less extreme weather conditions impacting children's physical activity levels at ECEC. Future studies should take into account the seasonal variation in their analyses.
# Conclusions
This study is one of the first to date to examine the influence of changes to the ECEC outdoor physical environment on preschoolers' activity levels, using an intervention design with matched controls to assess the causal relationships between the ECEC outdoor physical environment and physical activity while attending ECEC. Outdoor play spaces provide an important opportunity for children's physical activity, play, learning, and development. The addition of new portable equipment comprising balls, slides, twirling equipment, and floor play equipment resulted in intervention children being more active at follow-up. Features such as fixed sandboxes and real grass were also found to be beneficial for activity levels. Conversely, children were observed to be less active when fixed tunnels and twirling equipment were available for play.
Even though many preschool-aged children attend and spend a large part of their waking day at ECEC, very few of them achieve the recommended activity level while at care. Findings from the current study may provide some direction for ECEC providers for optimizing their outdoor physical environment to encourage more active play among preschoolers. Holistic development of the child, including physical and socio-emotional aspects, requires a balance between ECEC physical environment features that promote gross motor development of the child, and other features that nurture creative and imaginative play. Future intervention studies are required to examine the interaction of physical, policy, and practice level interventions on the physical activity behavior of children while attending ECEC.
[fig] Figure 1: Overview of study design and recruitment process. [/fig]
[fig] Author: Contributions: H.C. led the conceptualization and funding acquisition with contributions from M.R., S.G.T., L.L., and A.T.; H.C. led the methodology with contributions from M.R., A.T., L.L., P.B., and M.N.; M.N. led the investigation and analyses. M.N. led the writing of the original draft. All authors contributed to the writing, reviewing, and editing the manuscript at each stage. All authors have read and agreed to the published version of the manuscript. Funding: The PLAYCE study is funded by the Western Australian Health Promotion Foundation (Healthway; No. 32018). H.C. is supported by an Australian National Heart Foundation Future Leader Fellowship (100794). [/fig]
[table] Table 1: Child activity levels at baseline and follow-up (total physical activity (TPA) and moderate-vigorous physical activity (MVPA)). [/table]
[table] Table 2: Early childhood education and care (ECEC) outdoor physical environment features by intervention or control centers at baseline and follow-up. [/table]
[table] Table 3: Multivariate regression analyses of the association between the ECEC outdoor physical environment, children's characteristics, and children's activity levels. : natural elements variables. Variables excluded because they were nonsignificant: fixed climbing structures; fixed see-saws; fixed slides; fixed play structure; fixed swings; fixed balancing beams; portable climbing structures; portable jumping equipment; portable floor play equipment (TPA only); portable push/pull toys; portable sand play toys; portable riding toys; portable slides (TPA only); fake grass; other plants; potted plants; vegetation; rocks; flower beds; trees; open areas; water play area; sloping grounds; variety in ground surface; playground built on different levels; size of outdoor space; interaction between time and intervention (except for portable play equipment as independent variable). [/table]
|
Training of quantum circuits on a hybrid quantum computer
Generative modeling is a flavor of machine learning with applications ranging from computer vision to chemical design. It is expected to be one of the techniques most suited to take advantage of the additional resources provided by near-term quantum computers. Here, we implement a data-driven quantum circuit training algorithm on the canonical Bars-and-Stripes dataset using a quantum-classical hybrid machine. The training proceeds by running parameterized circuits on a trapped ion quantum computer and feeding the results to a classical optimizer. We apply two separate strategies, Particle Swarm and Bayesian optimization to this task. We show that the convergence of the quantum circuit to the target distribution depends critically on both the quantum hardware and classical optimization strategy. Our study represents the first successful training of a high-dimensional universal quantum circuit and highlights the promise and challenges associated with hybrid learning schemes.
# Introduction
Hybrid quantum algorithms (1) use both classical and quantum resources to solve potentially difficult problems. This approach is particularly promising for current quantum computers of limited size and power. Several variants of hybrid quantum algorithms have recently been demonstrated, such as the Variational Quantum Eigensolver for quantum chemistry and related applications [bib_ref] A variational eigenvalue solver on a photonic quantum processor, Peruzzo [/bib_ref] [bib_ref] Quantum chemistry calculations on a trapped-ion quantum simulator, Hempel [/bib_ref] [bib_ref] Scalable quantum simulation of molecular energies, O'malley [/bib_ref] [bib_ref] Self-verifying variational quantum simulation of lattice models, Kokail [/bib_ref] and the Quantum Approximate Optimization Algorithm for graph or other optimization problems. Hybrid quantum algorithms can also be used for generative models, which aim to learn representations of data to make subsequent tasks easier. Applications of generative modeling include computer vision [bib_ref] Unpaired image-to-image translation using cycle-consistent adversarial networks, Zhu [/bib_ref] , speech synthesis, the inference of missing text [bib_ref] Generating sentences from a continuous space, Bowman [/bib_ref] , denoising of images [bib_ref] Generalized denoising autoencoders as generative models, Bengio [/bib_ref] , and chemical design [bib_ref] Automatic chemical design using a data-driven continuous representation of molecules, Gómez-Bombarelli [/bib_ref]. Here, we apply a hybrid quantum learning scheme on a trapped ion quantum computer [bib_ref] Demonstration of a small programmable quantum computer with atomic qubits, Debnath [/bib_ref] to accomplish a generative modeling task.
Data-driven quantum circuit learning (DDQCL) is a hybrid framework for generative modeling of classical data where the model consists of a parameterized quantum circuit [bib_ref] A generative modeling approach for benchmarking and training shallow quantum circuits, Benedetti [/bib_ref]. The model is trained by sampling the output of a quantum computer and updating the circuit parameters using a classical optimizer. After convergence, the optimal circuit produces a quantum state that captures the correlations in the training datasets. Hence, the trained circuit serves as a generative model for the training data. Theoretical results suggest that such generative models have more expressive power than widely used classical neural networks [bib_ref] A quantum machine learning algorithm based on generative models, Gao [/bib_ref]. This is because instantaneous quantum polynomial circuits-special cases of the parameterized quantum circuits used for generative modeling-cannot be efficiently simulated by classical means.
The Bars-and-Stripes (BAS) dataset is a canonical body of synthetic data for generative modeling. It can be easily visualized in terms of images containing horizontal bars or vertical stripes, where each pixel represents a qubit. Here, we use the uniformly distributed two by two BAS shown in in a proof-of-principle generative modeling task on a trapped-ion quantum computer. This is the first successful demonstration of generative quantum circuits trained on multiqubit quantum hardware. We note that there has been a single-qubit experiment in this context [bib_ref] Quantum generative adversarial learning in a superconducting quantum circuit, Hu [/bib_ref]. We compare the performance of different classical optimization algorithms and conclude that Bayesian optimization (BO) shows substantial advantages over Particle Swarm Optimization (PSO) for this task.
The experiment is performed on four qubits within a seven-qubit fully programmable trapped ion quantum computer (21) (see Materials and Methods). With individual addressing and readout of all qubits, the system can perform sequences of gates from a universal gate set, composed of Ising gates and arbitrary rotations [bib_ref] Demonstration of a small programmable quantum computer with atomic qubits, Debnath [/bib_ref]. To run the large number of variational circuit instances necessary for the data-driven learning, we calibrate single-and two-qubit gates and execute lists of circuits in an automated fashion.
The training pipeline is illustrated in . The quantum circuits are structured as layers of parameterized gates. We use two types of layers, involving single-qubit rotations and two-qubit entangling gates. A single-qubit layer sandwiches an X-rotation between two Z-rotations on each qubit i, orR ðiÞ z ða i ÞR ðiÞ x ðb i ÞR ðiÞ z ðg i Þ, involving 12 rotation parameters for the four qubits (see . An entangling layer applies Ising or XX gates between all pairs of qubits according to any imposed connectivity graph. This is expressed as a sequence of XX i, j (c i, j ) operations as shown in , with up to six entangling parameters (15) for four qubits. Because of the universality of this gate set, a sufficiently long sequence of layers of these two types can produce arbitrary unitaries.
At the start of DDQCL, all the rotation and entangling parameters are initialized with random values. Next, the circuit is repeatedly executed on the trapped ion quantum computer to reconstruct the state distribution. A classical computer then compares the measured distribution with the target distribution and quantifies the difference using a cost function (see Materials and Methods for details). A classical optimization algorithm then varies the parameters. We iterate the entire process until convergence.
We impose two distinct connectivity graphs in a four-qubit circuit: all to all and star, as shown in . With star connectivity, entanglement between certain qubit pairs cannot occur within a single gate layer, which means that more layers are necessary for certain target distributions. Comparing the training process between circuits of different connectivity provides insight into the performance of DDQCL algorithms on platforms with more limited interaction graphs.
For each connectivity graph, we add layers until the goal of reproducing the BAS data with the trained model is achieved. The match between training data and model is limited by noise, experimental throughput rate (how fast the system can process circuits), and sampling errors. The cost function used in optimization scores the result, but a successful training process must be able to generate data that can be qualitatively recognized as a BAS pattern to ensure that the system provides usable results in the spirit of generative modeling in machine learning.
We now describe the classical optimization strategies for the training algorithm. Although gradient-based approaches were recently proposed for DDQCL (23), we use gradient-free optimization schemes that appear less sensitive to noise and experimental throughput. We explore two such schemes: PSO [bib_ref] Particle swarm optimization, Kennedy [/bib_ref] and BO. PSO is a stochastic optimization scheme commonly used in machine learning that works by creating many "particles" randomly distributed across parameter space that explore the landscape collaboratively. We limit the number of particles to twice the number of parameters. BO is a global optimization paradigm that can handle the expensive sampling of manyparameter functions. It works by maintaining a surrogate model of the underlying cost function and, at each iteration, updates the model to guide the search for the global minimum. Essentially, the problem of optimizing the real cost is replaced with that of optimizing the surrogate model, which is designed to be a much easier optimization problem. We use OPTaaS, a BO software package developed by Mind Foundry and adapted for this work. [fig_ref] Figure 3: Quantum circuit training results with PSO, with simulations [/fig_ref]. We first simulate the training procedure using a classical simulator in place of the quantum processor (orange plots in [fig_ref] Figure 3: Quantum circuit training results with PSO, with simulations [/fig_ref]. Since the PSO method is sensitive to the initial "seed" values of the particles, we simulate the convergence for many different random seeds (see [fig_ref] Figure 3: Quantum circuit training results with PSO, with simulations [/fig_ref]. We choose a seed that converges quickly and reliably under simulated sampling error to start the training procedure on the trapped ion quantum computer illustrated in . We iterate the training until it converges (blue plots in [fig_ref] Figure 3: Quantum circuit training results with PSO, with simulations [/fig_ref]. In practice, which seeds are successful is unknown, and different seeds need to be tried experimentally until a good model is . DDQCL is a hybrid quantum algorithm scheme that can be used for generative modeling, illustrated here by the example of two by two BAS data. From top left, clockwise: A parametrized circuit is initialized at random. Then, at each iteration, the circuit is executed on a trapped ion quantum computer. The probability distribution of measurement is compared on a classical computer against the BAS target dataset. Next, the quantified difference is used to optimize the parametrized circuit. This learning process is iterated until convergence. . Connectivity graphs and corresponding training circuits. Top: Fully connected training circuit layer, with layers of rotations (square boxes) and entanglement gates (rounded boxes) between any pair of the four qubits. Bottom: Star-connected training circuit layer, with restricted entangling gates. In either case, each rotation (denoted by X or Z) and each entanglement gate (denoted by XX) include a distinct control parameter, for a total of 18 parameters for the fully connected circuit layer and 15 parameters for the star-connected circuit layer. We remove the first Z rotation (dashed square boxes) acting on the initial state |0>, resulting in 14 and 11 parameters. The connectivity figures on the left define the mapping between the four qubits and the pixels of the BAS images (see .
# Results
## Results from pso optimization are shown in
obtained. This incurs an additional cost in the form of multiple independent DDQCL training rounds.
For all-to-all connectivity, we find that a circuit with one rotation gate layer and one entangling gate layer is able to produce the desired BAS distribution [fig_ref] Figure 3: Quantum circuit training results with PSO, with simulations [/fig_ref]. This is not the case for the star-connected circuit, with the closest state having two additional components in the superposition (states 6 and 9 in [fig_ref] Figure 3: Quantum circuit training results with PSO, with simulations [/fig_ref]. With two additional layers, the star-connected circuit is able to model the BAS distribution (orange plots of [fig_ref] Figure 3: Quantum circuit training results with PSO, with simulations [/fig_ref]. In the experiment, however (blue plots in [fig_ref] Figure 3: Quantum circuit training results with PSO, with simulations [/fig_ref] , the PSO is unable to converge to an acceptable solution even using the best prescreened seed value and sufficient sample statistics. We conclude that PSO fails because the throughput rate is too low for effectively training the circuit in the face of gate imperfections.
For these reasons, we instead use a BO scheme for the circuit training procedure. We find that all circuits experimentally converge in agreement with the simulations, as shown in [fig_ref] Figure 4: Quantum circuit training results with BO, with simulations [/fig_ref]. Moreover, even the star-connected circuit with four layers now produces a recognizable BAS distribution [fig_ref] Figure 4: Quantum circuit training results with BO, with simulations [/fig_ref]. In contrast to PSO, BO markedly reduces the number of samples needed for training and does not require any preselection of random seeds or other prior knowledge of the costfunction landscape.
BO updates the surrogate model using the experimental result of every iteration. Therefore, the classical part of each BO iteration con-sumes more time than with PSO, where the time cost on the classical optimizer is negligible. However, the BO procedure converges faster to the desired BAS distribution. More generally, these examples highlight the need to balance quantum and classical resources to produce acceptable performance and run time in a hybrid quantum algorithm.
As a measure of the performance of the various training procedures, we compute the Kullback-Leibler (KL) divergence (D KL ) [bib_ref] On information and sufficiency, Kullback [/bib_ref] and the qBAS score [an alternative performance measure suggested in [bib_ref] A generative modeling approach for benchmarking and training shallow quantum circuits, Benedetti [/bib_ref] ] of the experimental results at the end of each DDQCL training run, as shown in . We also compute the entanglement entropy (S) averaged over all two plus two qubit partitions assuming a pure state [bib_ref] How entangled can two couples get?, Higuchi [/bib_ref] , estimated via simulation of the quantum state from the trained circuits. The entanglement entropy quantifies the level of entanglement of a state, and thus indicates how difficult it is to produce such state. This metric shows that the successfully trained circuits generate states that are consistent with a high level of entanglement. As a reference, the entanglement entropy of a GHZ state over any partition is S = 1.
# Discussion
This demonstration of generative modeling using reconfigurable quantum circuits of up to 26 parameters is one of the most powerful hybrid quantum applications to date. With ongoing engineering improvements . Column (A) corresponds to a circuit with one layer of single qubit rotations (square boxes) and one layer of entanglement gates (rounded boxes) of all-to-all connectivity. The circuit converges well to produce the BAS distribution. Columns (B) and (C) correspond to a circuit with two and four layers and star connectivity, respectively. In (B), the simulation shows imperfect convergence with two extra state components (6 and 9), due to the limited connectivity, and the experimental results follow the simulation. In (C), the simulation shows convergence to the BAS distribution, but the experiment fails to converge despite performing 1400 quantum circuits. The optimization is sensitive to the choice of initialization seeds. To illustrate the convergence behavior, the shaded regions span the 5th to 95th percentile range of random seeds [500 for (A) and (B), 1000 for (C)], and the orange curve shows the median. The two-layer circuits have 14 and 11 parameters for (A) all-to-all and (B) star connectivity, respectively while the (C) star-connectivity circuit with four layers has 26 parameters. The number of PSO particles used is twice the number of parameters, and each training sample is repeated 5000 times. Including circuit compilation, controller-upload time, and classical PSO optimization, each circuit instance takes about 1 min to be processed, in addition to periodic interruptions for the recalibration of gates.
(28), we expect the system to grow in both qubit number and gate quality. This approach can be scaled up to handle larger datasets with increased qubit number by adapting the cost function for sparser sampling [bib_ref] A generative modeling approach for benchmarking and training shallow quantum circuits, Benedetti [/bib_ref]. Moreover, this procedure can be adapted for other types of hybrid quantum algorithms.
Classical optimization techniques for hybrid quantum algorithms on intermediate-scale quantum computers do not always succeed. Recent work suggests that typical cost functions for medium to large scale variational quantum circuits landscape resemble "barren plateaus" (30), making optimization hard. As quantum computers scale up for larger problems, the cost of classical optimization such as BO must be weighed against the quantum algorithmic advantage.
# Materials and methods
## Trapped ion quantum computer
The trapped ion quantum computer used for this study consists of a chain of seven single 171 Yb + ions confined in a Paul trap and laser-cooled close to their motional ground state. Each ion provides one physical qubit in the form of a pair of states in the hyperfine-split 2 S 1/2 ground . Column (A) corresponds to a circuit with two layers of gates and all-to-all connectivity. Columns (B) and (C) correspond to a circuit with two and four layers and star connectivity, respectively. Convergence is much faster than with PSO [fig_ref] Figure 3: Quantum circuit training results with PSO, with simulations [/fig_ref]. Unlike the PSO results, the four-layer star-connected circuit in (C) is trained successfully, and no prior knowledge enters BO process. As before, the two-layer circuits have 14 and 11 parameters for (A) all-to-all and (B) star connectivity, respectively while the (C) star-connectivity circuit with four layers has 26 parameters. We used a batch of five circuits per iteration, and each training sample was repeated 5000 times. Including circuit compilation, controller-upload time, and BO classical optimization, each circuit instance takes 2 to 5 min, depending on the amount of accumulated data. . D KL (see Materials and Methods), qBAS score, and entanglement entropy (S) for the state obtained at the end of each of the DDQCL training on hardware for various circuits and classical optimizers used.
## S c i e n c e a d v a n c e s | r e s e a r c h a r t i c l e
level with an energy difference of 12.642821 GHz, which is insensitive to magnetic fields to first order. The qubits are collectively initialized into |0> through optical pumping, and state readout is accomplished by state-dependent fluorescence detection [bib_ref] Manipulation and detection of a trapped yb hyperfine qubit, Olmschenk [/bib_ref]. Qubit operations are realized via pairs of Raman beams, derived from a single 355-nm mode-locked laser [bib_ref] Demonstration of a small programmable quantum computer with atomic qubits, Debnath [/bib_ref]. These optical controllers consist of an array of individual addressing beams and a counter-propagating global beam that illuminates the entire chain. Single qubit gates are realized by driving resonant Rabi rotations of defined phase, amplitude, and duration. Single-qubit rotations about the z axis are performed by classically advancing/regarding the phase of the optical beatnote applied to the particular qubit. Two-qubit gates are achieved by illuminating two selected ions with beat-note frequencies near motional sidebands and creating an effective Ising spin-spin interaction via transient entanglement between the two qubits and the motion in the trap [bib_ref] Multiparticle entanglement of hot trapped ions, Mølmer [/bib_ref] [bib_ref] Deterministic bell states and measurement of the motional state of two trapped..., Solano [/bib_ref] [bib_ref] Ion trap quantum computing with warm ions, Milburn [/bib_ref]. Since our particular scheme involves multiple modes of motion, we use an amplitude modulation scheme to disentangle the qubit state from the motional state at the end of the interaction [bib_ref] Optimal quantum control of multimode couplings between trapped ion qubits for scalable..., Choi [/bib_ref]. Typical single-qubit gate fidelities are 99.5(2)%. Typical two-qubit gate fidelities are 98 to 99%, with fidelity mainly limited by residual entanglement of the qubit states to the motional state of the ions, coherent cross-talk, and driving intensity noise from classical imperfections in our optical controllers.
In our experiment, the effect of the gate errors is seen as an offset in the cost function after convergence. An improvement in gate fidelity will reduce this offset. However, the convergence behavior of an ideal system (as shown in the simulations in [fig_ref] Figure 3: Quantum circuit training results with PSO, with simulations [/fig_ref] is not significantly faster than the actual experimental system. This is because it is limited by the classical optimization routine.
The trapped ion quantum architecture is scalable to a much larger number of qubits, as atomic clock qubits are perfectly replicable and do not suffer idle errors (T1 and T2 times are essentially infinite). All of the errors in scaling arise from the classical controllers, such as applied noise on the trap electrodes and laser beam intensity fluctuations. Fundamental errors (such as spontaneous scattering from the control laser beams) are not expected to play a role until our gates approach 99.99% fidelity. However, as the qubit number grows beyond about 20 to 30, we expected to sacrifice full connectivity, as gates will only be performed with high fidelity between any qubit and its 15 to 20 nearest neighbors.
Another limitation is the sampling rate on the quantum computer. This is limited by technical issues on the current experiment and can be improved, e.g., by increasing the upload speed of the experimental control system.
## Classical optimizers: pso and bo
We explored two different classical optimizers in this study: PSO and BO.
PSO is a gradient-free optimization method inspired by the social behavior of some animals. Each particle represents a candidate solution and moves within the solution space according to its current performance and the performance of the swarm. Three hyperparameters control the dynamics of the swarm: a cognition coefficient c 1 , a social coefficient c 2 , and an inertia coefficient w (24).
Concretely, each particle consists of a position vector q i and a velocity vector v i . At iteration t of the algorithm, the velocity of particle i for the coordinate d is updated as
In our problem, each particle corresponds to a point in parameter space of the quantum circuit. For example, in the fully connected circuit with two layers, each particle consists of an instance of the 14 parameters. Recall, however, that parameters are angles and therefore periodic; we customized the PSO updates above to use this information. In Eq. 1 , p ðtÞ i;d and q ðtÞ i;d can be thought of as two points on a circle. Instead of using the standard displacement p ðtÞ i;d À q ðtÞ i;d , we used the angular displacement, that is, the signed length of the minor arc on the unit circle. We used the same definition of displacement for the swarm's best position g ðtÞ i;d . Last, in Eq. 2, we made sure to express angles always using their principal values.
In our experiments, we set the number of particles to twice the number of parameters of the circuit. Position and velocity vectors of each particle were initialized from the uniform distribution. For the coefficients, we used c 1 = c 2 = 1 and w = 0.5.
BO is a powerful global optimization paradigm. It is best suited to finding optima of multimodal objective functions that are expensive to evaluate. There are two main features that characterize the BO process: the surrogate model and an acquisition function.
The surrogate model is nonparametric model of the objective function. At each iteration, the surrogate model is updated using the sampled points in parameter space. The package used in this study is OPTaaS by Mind Foundry. It implements the surrogate model as regression using Gaussian process. A kernel (or correlation function) characterizes the Gaussian process, we used a Matern 5/2 as it provides the most flexibility.
The acquisition function is computed from the surrogate model. It is used to select points for evaluation during the optimization. It trades off exploration against exploitation. The acquisition function of a point has a high value if the cost function is expected to give a notable improvement over historically sampled points or if the uncertainty of the point is high, according to the surrogate model. A simple and well-known acquisition function, Expected Improvement (37), is used here.
In our case, OPTaaS also leverages the cyclic symmetry of the angles by embedding the parameter space into a metric space with the appropriate topology, effectively allowing the Gaussian process surrogate model to be placed over a hypertorus rather than a hypercube. This greatly alleviates the so-called curse of dimensionalityand allows for much more efficient use of samples of the objective function.
It is the key in BO to adequately optimize the acquisition function during each iteration. OPTaaS puts considerable computational resources toward this nonconvex optimization problem.
There are two major reasons why the BO out performs PSO in our specific case. First, PSO spends significant amount of computation resource exploring trajectories far from optimal, while BO mitigates it by the use of acquisition function. Second, the maintenance of the surrogate model enables us to make much better use of the information from the historical exploration of the parameter space.
## Cost functions
We used a cost function to quantify the difference between the target BAS distribution and the experimental measurements of the circuit. Here, p and q are two distributions. D KL (p, q) is an information theoretic measure of how two probability distribution differ. If base 2 for the logarithm is used, then it quantifies the expected number of extra bits required to store samples from p when an optimal code designed for q is used instead. It can be shown that D KL (p, q) is nonnegative and is zero if and only if p = q. However, it is asymmetric in the arguments and does not satisfy the triangle inequality. Therefore, D KL (p, q) is not a metric.
The D KL is a very general measure, but it is not always well defined, e.g., if an element of the domain is supported by p and not by q, then the measure will diverge. This problem may occur quite often if D KL (p, q) is estimated from samples and if the dimensionality of the domain is large. For PSO, we used the clipped negative log-likelihood cost function (16),
[formula] C nll ¼ À∑ i pðiÞlog fmax ½D; qðiÞgð4Þ [/formula]
Here, we set p as the target distribution. Thus, Eq. 4 is equivalent to Eq. 3 up to a constant offset, so the optimization of these two functions is equivalent. D is a small number (0.0001 here) used to avoid a numerical singularity when q(i) is measured to be zero. For BO, we used the clipped symmetrized D KL as the cost functioñ D KL ðp; qÞ ¼ D KL ½max ðD; pÞ; max ðD; qÞ þ D KL ½max ðD; qÞ; max ðD; pÞ ð5Þ
This is found to be the most reliable variant of D KL for BO.
[fig] Figure 3: Quantum circuit training results with PSO, with simulations (orange) and trapped ion quantum computer results (blue) [/fig]
[fig] Figure 4: Quantum circuit training results with BO, with simulations (orange) and trapped ion quantum computer results (blue) [/fig]
|
Eva‐1 homolog A promotes papillary thyroid cancer progression and epithelial‐mesenchymal transition via the Hippo signalling pathway
## | material s and me thods
## | patients and samples
This study included 43 patients with PTC who underwent surgi-
## | cell lines and cell culture
## | rna extraction and qrt-pcr
## | rna interference
The siRNA and the non-targeting siRNA (negative control
[NC]siRNA) used for gene knockdown were obtained from Genepharma (Genepharma). The siRNA sequences of EVA1A
and negative control siRNA were as follows: EVA1A Sense-1: 5′-CCUAGCGGCCUAUUCCUUUTT-3′; EVA1A Antisense-1:
[formula] 5′-AAAGGAAUAGGCCGCUAGGTT-3′; EVA1A Sense-2: [/formula]
5′-GAGCCUGAAUCGCUACUAUTT-3′; and EVA1A Antisense-2:
5′-AUAGUAGCGAUUCAGGCUCTT-3′. The cell lines were seeded in six-well plates for 24 hours, and transfection was performed. The knockdown efficacy was determined using qRT-PCR.
## | cell proliferation assay
Cell proliferation was determined using the Cell Counting Kit-8 (CCK-
## | colony formation assay
The two transfected or control cells (1.5 × 10 3 cells for KTC-1 and TPC-1) were seeded into six-well plates and incubated for 7-14 days.
Each well was fixed with 4% paraformaldehyde (Sigma-Aldrich Co.)
for 30 minutes, stained with 0.01% crystal violet and photographed.
All experiments were performed in triplicate.
## | cell migration and invasion assays
The migration assay was performed using the Transwell chambers (Corning Incorporated). For the wound healing assay, 1.5 × 10 5 cells were cultured into a 24-well plate until 90% confluency was reached.
Scratch wounds were created using 0.
## | apoptosis assay
## | gsea
The
## | protein extraction and western blot analysis
The
## | re sults
## | eva1a expression is overexpressed in ptc
In our previous study, RNA sequencing analysis is performed on the primary PTC and the adjacent normal tissues obtained from 19 patients with PTC.After analysing the data, we have found that EVA1A was significantly up-regulated . This result coincided with the data shown in the TCGA cohort. Afterwards,
we detected the relative expression of EVA1A in 43 patients via qRT-PCR to verify the sequencing results. The expression level of EVA1A in tumour tissues was significantly higher compared with that in adjacent normal tissues. The ROC curve analysis was used to evaluated the diagnostic capability of EVA1A. As shown in,D, EVA1A expression had excellent diagnostic value to distinguish non-cancerous tissue from cancer tissue in the TCGA data set (AUC = 81.2%) and our local cohort (AUC = 94%).
## | eva1a is associated with the clinicopathological features of ptc
TCGA data were analysed to determine the relationship between the expression level of EVA1A and the clinicopathological features and further explore the role of EVA1A in PTC. We divided the patients with PTC into two groups, namely high (n = 251) and low (n = 251)
EVA1A expression in accordance with the median value of the EVA1A expression level in TCGA cohort. Results showed that a high EVA1A expression was related to histological type (P < .001), age (P = .025), tumour size (P = .06) and lymph node metastasis (LNM; P < .001; . Nevertheless, the associations of EVA1A expression with gender, distant metastasis and disease stage (AJCC7) were not significant. In our validated cohort, high EVA1A expression was correlated with LNM (P = .009) and disease stage (AJCC7; P = .011; .
These results supported that EVA1A gene was an oncogene in PTC.
## | eva1a overexpression aggravates the risk of lnm in ptc
Further investigation was carried out to find the relationship of EVA1A expression with LNM. We conducted a univariate logistic regression analysis in TCGA data. Results showed that the significant variables for LNM were EVA1A expression (odds
## | eva1a knockdown suppresses cell proliferation and colony formation
We evaluated the expression level of EVA1A in several PTC and normal thyroid cell lines (HTORI-3) in the mRNA and protein levels.
Results showed that EVA1A expression was higher in TPC-1, KTC-1, BCPAP and FTC-133 than that in HTORI-3
## | eva1a knockdown inhibits migration and invasion
The migration and invasion assays were performed on TPC-1 and KTC-1 to further confirm the function of EVA1A in PTC. As expected, the PTC cells with underexpressed EVA1A had significantly inhibited ability to migrate and invade compared with those in the Si-NC control group.
## | eva1a underexpression effectively induces ptc cell line apoptosis in vitro
## | eva1a knockdown induces the activation of the intrinsic apoptosis pathway and suppresses emt progression via the hippo signalling pathway
Accumulating number of studies have shown that EMT and apoptosis plays a critical role in PTC metastasis and proliferation.GSEA indicated that the expression of EVA1A was related to EMT and apoptosis . Related proteins were detected to explore the mechanism of the EVA1A on EMT and apoptosis progression. Recent studies have provided enormous evidences suggesting that the Hippo signalling pathway can induce EMT.Results revealed that compared with the control group, the Si-EVA1A group showed decreased N-cadherin, vimentin and Bcl-xL expression and increased Bax expression . Furthermore, the YAP and TAZ levels in the Si-EVA1A group were lower than those in the control group . These results suggested that EVA1A may inhibit apoptosis and promote the metastasis and the proliferation of PTC by suppressing the intrinsic apoptosis pathway and activating the EMT via the Hippo signalling pathway.
## | d iscuss i on
With the development of molecular biology, our understanding of PTC molecular mechanisms is gradually developing. Next-generation sequencing has paved the way for the discovery that PTC is driven by several characteristic genetic alterations, such as point mutations in proto-oncogenes and chromosomal rearrangements.However, despite the enormous progress in PTC genetic research, the current non-surgical treatment for thyroid cancer is still limited. Therefore, further research on the pathogenetic mechanism of PTC is essential for the diagnosis and therapy of PTC.
According to Cancer Statistics 2019, thyroid cancer incidence is rising faster than other cancers in men and women and accounts for 4% of the total estimated female cancer burden in 2019.In our previous study, whole-genome sequencing has been performed on 19
pairs of PTC tumours and adjacent normal tissues.We have found that the expression level of EVA1A is up-regulated in patients with PTC in Wenzhou. This finding supports the hypothesis that EVA1A may have a role as an oncogene in PTC. The function of EVA1A in apoptosis is complex and suggested as essential in the process.
However, the molecular mechanisms of EVA1A in thyroid carcinoma are still unclear.
In the present study, the statistical results of the clinicopatho- In summary, we first found that EVA1A in PTC was a potent oncogene of enhanced tumour aggressiveness. This evidence suggested that EVA1A may be an effective molecular marker of PTC for diagnosis and therapy.
## Ack n owled g em ents
The authors would like to thank all the patients who participated in this study.
## Co n fli c t o f i nte r e s t
The authors declare that they have no conflict of interest.
## Auth o r co ntr i b uti
## E th i c a l a pprova l
The ethical approval for this study was obtained from the Ethics
## Committee of the first affiliated hospital of wenzhou medical
University.
## I n fo r m e d co n s e nt
Written informed consent was obtained from each participant (approval no. 2012-57).
## Data ava i l a b i l i t y s tat e m e n t
The data sets used during this study and additional images are included in this article. Raw data are available upon reasonable request.
## O rci d
Bang-Yi Lin https://orcid.org/0000-0001-8833-9380
Jia-Liang Wen https://orcid.org/0000-0003-1493-1757
Chen Zheng https://orcid.org/0000-0002-4540-2711
Cheng-Ze Chen https://orcid.org/0000-0003-2006-6915
Jin-Miao Qu https://orcid.org/0000-0003-4290-7551 |
Comparison of pro- and anti-inflammatory responses in paired human primary airway epithelial cells and alveolar macrophages
Background: Airway epithelial cells and alveolar macrophages (AMs) are the first line of defense in the lung during infection. Toll-like receptor (TLR) agonists have been extensively used to define the regulation of inflammation in these cells. However, previous studies were performed in non-paired airway epithelial cells and AMs. The major goal of our study was to compare the pro-and anti-inflammatory responses of paired human primary airway epithelial cells and AMs to TLR3 and TLR4 agonists. Methods: Tracheobronchial epithelial cells (TBEC) and AMs from four smokers and four non-smokers without lung disease were cultured with or without Poly(I:C) (PIC) (a TLR3 agonist) or LPS (a TLR4 agonist) for 4, 24 and 48 h. The immune responses of paired cells were compared. Results: TBEC and AMs showed stronger pro-inflammatory cytokine (e.g., IL-8) responses to PIC and LPS, respectively. TLR3 and TLR4 mRNA levels were similar in non-stimulated TBEC and AMs. However, PIC stimulation in AMs led to sustained up-regulation of the immune negative regulators Tollip and A20, which may render AMs less sensitive to PIC stimulation than TBEC. Unlike AMs, TBEC did not increase NF-κB activation after LPS stimulation. Interestingly, smoking status was correlated with less TLR3 and IRAK-M expression in non-stimulated TBEC, but not in AMs. PIC-stimulated TBEC and LPS-stimulated AMs from smokers vs. non-smokers produced more IL-8. Finally, we show that expression of A20 and IRAK-M is strongly correlated in the two paired cell types.Conclusions: By using paired airway epithelial cells and AMs, this study reveals how these two critical types of lung cells respond to viral and bacterial pathogen associated molecular patterns, and provides rationale for modulating immune negative regulators to prevent excessive lung inflammation during respiratory infection.
# Background
Airway epithelial cells along with alveolar macrophages serve as the first line of host innate immune defense against airborne pathogens and other airborne environmental hazards [bib_ref] Innate immunity and pathogen-host interaction, Basset [/bib_ref]. These lung cells are able to recognize pathogen associated molecular patterns (PAMPs) using receptors that include the Toll-like receptor (TLR) family. TLR-mediated recognition of PAMPs leads to the activation of downstream signaling cascades, the subsequent activation of pro-inflammatory transcription factors including NF-κB, and ultimately the production of pro-inflammatory cytokines including CXCL8 (IL-8), CXCL10 (IP-10) and TNF-α. While this inflammatory response is important for combating the pathogens, inflammation must be appropriately regulated to prevent excessive inflammation and tissue damage. One regulatory mechanism that ensures that inflammation is self-limiting is the PAMP-mediated induction of negative regulators. Numerous negative regulators of TLR-mediated signaling have been identified including Toll-interacting protein (Tollip), TNF alphainduced protein 3 or TNFAIP3 (A20) and interleukin-1 receptor-associated kinase 3 (IRAK-M). These negative regulators down-regulate the transcription and translation of TLR-induced genes during infection and inflammation [bib_ref] Pathogen recognition and inflammatory signaling in innate immune defenses, Mogensen [/bib_ref]. Hosts are protected from hyper-inflammation and autoimmunity by the inhibitory effect of these negative regulators [bib_ref] Pathogen recognition and inflammatory signaling in innate immune defenses, Mogensen [/bib_ref].
It has been proposed that airway epithelial cells and alveolar macrophages may respond similarly or differently to various microbes and microbial PAMPs [bib_ref] Innate immunity and pathogen-host interaction, Basset [/bib_ref] [bib_ref] Contribution of lung macrophages to the inflammatory responses induced by exposure to..., Hiraiwa [/bib_ref] [bib_ref] Innate immune responses of the airway epithelium, Ryu [/bib_ref]. However, few studies have been conducted that directly compare the response of these two cell types. Notably, negative regulators of TLR signaling pathways have not been previously investigated in airway epithelial cells and alveolar macrophages from the same human subject (paired cells) to clarify their effect on inflammatory responses.
In the current study, we used paired airway epithelial cells and alveolar macrophages from the same healthy donors to test the hypothesis that functional differences exist between airway epithelial cells and alveolar macrophages with respect to pro-inflammatory cytokine release after TLR stimulation. In particular, we tested if the response to TLR3 and TLR4 agonists differs in these two cell types and if this difference may be explained by altered expression of negative regulators of TLR signaling. We also explored the effect of smoking status in these paired cells on the immune response after TLR stimulation. A full understanding of how inflammation is regulated by these negative regulators in different host cell types will facilitate the design of new therapeutics to balance the beneficial and detrimental effects of inflammation in various lung diseases.
# Methods
# Materials
Bronchial epithelial cell growth medium (BEGM) with antibiotics was purchased from Lonza (Walkersville, MD). The BEGM was prepared following manufacturer's guideline, which contained all the supplements (BPE, hEGF, epinephrine, transferrin, insulin, retinoic acid, triiodothyronine, GA) except hydrocortisone to avoid any inhibitory effect of corticosteroids on cell pro-inflammatory responses. RNA lysis buffer RLT was from Qiagen (Hilden, Germany). RIPA western lysis buffer was purchased from Thermo-Fisher Scientific (Waltham, MA). DMEM (high glucose) for making D10 (DMEM + 10% FBS + 1% Pen/Strep + 1% Amphotericin B + 1% L-Glutamine+ 0.5% Gentamicin) was from GE Life Sciences (Logan, UT). The nuclear extraction kit and TransAM NF-κB p65 assay kit were from Active Motif (Carlsbad, CA). IL-8, IP-10 and TNF-α ELISA kits were obtained from R&D systems (Minnieaplois, MN).
## Human donor information
To isolate human primary airway epithelial cells and alveolar macrophages, we obtained human lungs from de-identified organ donors whose lungs were not suitable for transplantation and were donated for medical research through the National Disease Research Interchange (Philadelphia, PA), the International Institute for the Advancement of Medicine (Edison, NJ), or Donor Alliance of Colorado. The Institutional Review Board (IRB) at National Jewish Health deemed this research as non-human subjects research. Donors were chosen based on lung function with a Pa O2 /Fi O2 ratio of > 225, no history of clinical lung diseases, a chest radiograph indicating no infection, and a time on the ventilator of < 5 days. The sex, age, race, and smoking history of donors were variable and were not selection criteria.
## Isolation and culture of human tracheobronchial airway epithelial cells
To isolate tracheobronchial epithelial cells (TBEC), tracheal and main bronchial tissue was digested with 0.1% protease in DMEM overnight at 4°C, and processed as previously described [bib_ref] IL-13 dampens human airway epithelial innate immunity through induction of IL-1 receptor-associated..., Wu [/bib_ref] [bib_ref] Cell surface marker profiling of human tracheal basal cells reveals distinct subpopulations,..., Van De Laar [/bib_ref] [bib_ref] Characterization of side population cells from human airway epithelium, Hackett [/bib_ref]. TBEC at passage 1 from the frozen stock were cultured and expanded in collagen-coated 60-mm tissue culture dishes in BEGM medium at 37°C, 5% CO 2 . When the cells were 80% confluent, they were trypsinized and seeded onto 12-well plates for submerged culture. In our culture model (primary submerged culture), TBEC grown in a monolayer did not differentiate into the mucociliary phenotype, but showed the feature of basal cells expressing KRT5 [fig_ref] Figure 1: Characterization of primary human tracheobronchial epithelial cell [/fig_ref].
## Submerged culture and stimulation of tbec
TBEC at 1 × 10 5 cells/well in BEGM media were seeded into 12-well plates. After 48 h, the medium was changed to refresh BEGM, and LPS (10 ng/ml) or Poly(I:C) (PIC) (1 μg/ml) was added. LPS was used to mimic bacterial infection as use of live bacteria could cause cell death, thus compromising data interpretation. We chose to use Poly I:C (PIC) as a dsRNA mimic of RNA viruses to broadly test the pro-inflammatory response in paired airway epithelial cells and alveolar macrophages. Given the varying susceptibility of the two types of cells to common respiratory viruses such as influenza A viruses and rhinoviruses, it would have been difficult to compare their responses to the live viruses. It has been reported that human AMs are less susceptible to infection by live influenza A viruses or rhinoviruses compared to epithelial cells [bib_ref] Human alveolar macrophages may not be susceptible to direct infection by a..., Ettensohn [/bib_ref]. PAMP doses were chosen following LPS and PIC dose response experiments; the lowest concentrations yielding a pro-inflammatory response in both cell types were chosen. The doses we chose were comparable to previous studies using LPS and PIC stimulation in cell culture experiments [bib_ref] Response of human pulmonary epithelial cells to lipopolysaccharide involves toll-like receptor 4..., Guillott [/bib_ref] [bib_ref] Differences in LPS-induced activation of bronchial epithelial cells (BEAS-2B) and type II-like..., Schulz [/bib_ref]. The cells were harvested after 4, 24 and 48 h in RNA lysis buffer (RLT) or in RIPA buffer with protease and phosphatase inhibitors. The supernatants were collected and stored at − 80°C for ELISA.
Isolation, culture, and stimulation of human alveolar macrophages Bronchoalveolar lavage (BAL) was performed on the right middle lobe or lingula of the donor lungs by completely filling the lobe three times with balanced salt solution and EDTA, and then three times with the salt solution alone [bib_ref] Flow cytometric analysis of mononuclear phagocytes in nondiseased human lung and lungdraining..., Desch [/bib_ref] [bib_ref] Innate immune response of human alveolar macrophages during influenza a infection, Wang [/bib_ref]. After each instillation, lavage fluid was drained from the lung, collected, pooled, and centrifuged to obtain BAL cells including alveolar macrophages. The BAL cells were frozen in 90% FBS and 10% DMSO. Based on the protocols established by us [bib_ref] Innate immune response of human alveolar macrophages during influenza a infection, Wang [/bib_ref] [bib_ref] Differentiated human alveolar type II cells secrete antiviral IL-29 (IFN-lambda 1) in..., Wang [/bib_ref] and others [bib_ref] Human alveolar macrophages may not be susceptible to direct infection by a..., Ettensohn [/bib_ref] [bib_ref] Bronchoalveolar lavage cell pattern from healthy human lung, Heron [/bib_ref] , macrophages were enriched from the BALF after lysis of the RBCs, and adhered onto the plastic surface in 12-well culture plates, and then washed to remove non-adherent cells. By using the CD68 immunofluorescent staining as shown in [fig_ref] Figure 1: Characterization of primary human tracheobronchial epithelial cell [/fig_ref] , d and in our previous publication [bib_ref] Differentiated human alveolar type II cells secrete antiviral IL-29 (IFN-lambda 1) in..., Wang [/bib_ref] , nearly 99% of cells were positive for CD68.
Isolated alveolar macrophages were thawed and seeded into 12-well plates at densities of either 1 × 10 5 cells/well or 5 × 10 5 cells/well in D10 media at 37°C in 5% CO 2 . After 48 h, the medium was changed to remove the non-adherent cells and treatments with LPS (10 ng/ml) or PIC (1 μg/ml) in D10 media were started on adhered AM. After 4, 24 and 48 h, the supernatants were collected and stored at − 80°C for ELISA. The cells were harvested at these same time points in RNA lysis buffer (RLT) or in RIPA buffer with protease and phosphatase inhibitors.
## Characterization of the tbec cells and bal macrophages with immunofluorescence
Primary paired TBEC and alveolar macrophages were cultured on coverslips. After 48 h, the TBEC were stained with an anti-cytokeratin antibody (KRT5) (Abcam, 1:500) and, AM were stained with an anti-CD68 antibody (Bioscience, 1:200) following a published protocol [bib_ref] Cell surface marker profiling of human tracheal basal cells reveals distinct subpopulations,..., Van De Laar [/bib_ref] [fig_ref] Figure 1: Characterization of primary human tracheobronchial epithelial cell [/fig_ref].
## Nf-κb activity assay
Following the treatments of the paired cells as described above, nuclear proteins at each of the time points (4, 24, 48 h) were extracted using the Active Motif kit as per the manufacturer's instructions. The extracted nuclear proteins (10 μg/condition) were tested for NF-κB p65 transcription factor activity using the TransAM NF-κB a b c d Relative mRNA expression and fold change levels were calculated using the delta Ct method for target genes and GAPDH. Target gene expression was normalized to GAPDH [bib_ref] Analysis of relative gene expression data using realtime quantitative PCR and the..., Livak [/bib_ref].
# Western blot analysis
Equal amounts of protein from samples with different treatments were separated by electrophoresis on 10% SDS-polyacrylamide gels. The proteins were then transferred onto nitrocellulose membranes and probed with either a goat polyclonal anti-Tollip antibody (sc27315), a mouse monoclonal anti-A20 antibody (sc69980) or a mouse anti-β-actin antibody (sc47778) (Santa Cruz Biotechnology, Inc.). Blots were then incubated with appropriate HRP-linked secondary antibodies and developed with the ECL Western blotting substrate. The blots were scanned using a Fotodyne imaging system, and densitometry was performed using the NIH Image-J software.
# Statistical analysis
Pairwise comparisons were performed using the Student's t-test. For multiple comparisons, analysis of variance (ANOVA) or the Kruskal-Wallis test was used for normally distributed data or nonparametric data, respectively. False discovery rate (FDR) was used for correction (Benjamini and Hochberg). Single variable linear regression was used to test for an association between gene expression in AMs and TBECs. All statistical analyses and graphs were performed using GraphPad Prism software (GraphPad Software, La Jolla, CA, USA). P-values less than 0.05 were considered significant.
# Results
## Tracheobronchial epithelial cells (tbec) and alveolar macrophages display different responses to poly(i:c) and lps
The pro-inflammatory cytokines IL-8 (CXCL8) and IP-10 (CXCL10) were measured in the supernatants of cultured epithelial cells and macrophages from eight healthy subjects (four smokers and four non-smokers) as described in [fig_ref] Table 1: Characteristics of Research Subjects [/fig_ref]. TBEC responded to PIC stimulation by producing IL-8 starting at 4 h, peaking at 24 h (P < 0.001), and then maintaining cytokine production at 48 h. [fig_ref] Figure 2: Human tracheobronchial epithelial cells [/fig_ref]. Similarly, TBEC increased IP-10 production after PIC stimulation [fig_ref] Figure 2: Human tracheobronchial epithelial cells [/fig_ref]. In contrast to PIC, LPS stimulation in TBEC did not significantly increase IL-8 production at any of the time points examined [fig_ref] Figure 2: Human tracheobronchial epithelial cells [/fig_ref].
Alveolar macrophages responded to both PIC and LPS by producing IL-8, but the induction of IL-8 was stronger after LPS at 24 and 48 h (P = 0.01) [fig_ref] Figure 2: Human tracheobronchial epithelial cells [/fig_ref] and e). Alveolar macrophages also responded to PIC by producing IP-10 at 24 and 48 h, but the increase did not reach statistical significance [fig_ref] Figure 2: Human tracheobronchial epithelial cells [/fig_ref].
Although the two cell types were initially seeded at the same cell density (1 × 10 5 /well), alveolar macrophages, unlike the TBEC, do not proliferate. As a complementary approach to monitoring cytokine protein levels produced by the two cell types, IL-8 and IP-10 mRNA levels were monitored at the peak time (24 h) of cytokine induction. Consistent with the protein data, in TBEC, IL-8 mRNA increased (> 20-fold) after PIC stimulation (P = 0.01), but did not increase after LPS stimulation . Macrophages showed a significant increase in IL-8 mRNA expression (> 15-fold) after LPS stimulation at 24 h (P = 0.01). Macrophages also increased IL-8 mRNA after PIC, but the induction level was about 50% of that in TBEC . Similar to IL-8, IP-10 mRNA levels after PIC stimulation at 24 h were higher in stimulated TBEC (about 3000-fold) than macrophages that showed about 1000-fold induction . Interestingly, IP-10 protein was not detectable after LPS stimulation in both alveolar macrophages and airway epithelial cells, and LPS did not significantelly induce IP-10 mRNA in both cell types . Together, our data suggest that TBEC are a better responder to PIC, and alveolar macrophages have a stronger response to LPS.
Nuclear factor-kappaB (NF-κB) activation differs in paired TBEC and alveolar macrophages stimulated with Poly(I:C) or LPS
To determine the possible mechanism of varying pro-inflammatory cytokine responses of the paired airway epithelial cells and alveolar macrophages, NF-κB activity was measured with or without LPS or PIC stimulation. We chose to monitor NF-κB activation because NF-κB is activated by LPS and PIC and plays a key role in induction of pro-inflammatory cytokines. We monitored NF-κB activation by measuring NF-κB levels in nuclear extracts using an ELISA. Consistent with the IL-8 data, levels of NF-κB in the nuclei were increased in TBEC after stimulation with PIC (P < 0.05) at 24 and 48 h, but not LPS [fig_ref] Figure 4: Differential activation of NF-κB in human tracheobronchial epithelial cells [/fig_ref]. On the other hand, LPS, but not PIC, significantly increased NF-κB activation in alveolar macrophages at 24 h (P = 0.01) [fig_ref] Figure 4: Differential activation of NF-κB in human tracheobronchial epithelial cells [/fig_ref].
Levels of TLR4 and TLR3 expression do not explain the differences in IL-8 production and NF-κB activation in TBEC and alveolar macrophages TLR4 and TLR3 are the respective receptors for LPS and PIC. Binding of TLRs to their ligands activates TLR signaling pathways, leading to activation of NF-κB and other transcriptional factors, and thus pro-inflammatory cytokine production [bib_ref] TLR signalling and activation of IRFs: revisiting old friends from the NF-κB..., Moynagh [/bib_ref]. In order to determine why TBEC and alveolar macrophages respond more strongly to PIC and LPS, respectively, we compared the baseline (untreated) expression of TLR3 and TLR4 in the two cell types. No significant differences of TLR3 or TLR4 mRNA expression were found between the two cell types (Additional file 1: [fig_ref] Figure 1: Characterization of primary human tracheobronchial epithelial cell [/fig_ref]. This suggests that the differing responsiveness of TBEC and alveolar macrophages to PIC and LPS may be regulated by something other than TLR expression. One possibility that we explored below is that altered expression of negative regulators of TLR signaling may account for this difference.
## Differences in pamp-mediated induction of negative regulators in tbec
To potentially explain the different pro-inflammatory responses to PIC and LPS in TBEC and macrophages, we examined the expression of Tollip, A20 and IRAK-M, which are known to down-regulate TLR3 and TLR4 signaling pathways. In TBEC, there was no significant change in Tollip mRNA expression after LPS or PIC stimulation at all the time points measured [fig_ref] Figure 5: PAMP stimulation induces expression of negative regulators of TLR signaling in human... [/fig_ref]. In contrast, the other two negative regulators did exhibit altered expression following PAMP challenge. A20 mRNA levels were not significantly changed after LPS stimulation in TBEC. In contrast, PIC up-regulated A20 mRNA expression (p < 0.005) after 4, 24 and 48 h [fig_ref] Figure 5: PAMP stimulation induces expression of negative regulators of TLR signaling in human... [/fig_ref]. A20 mRNA expression in TBEC was highest at 4 h after PIC stimulation and remained significantly increased, albeit at a lower level, 24 and 48 h after PIC challenge [fig_ref] Figure 5: PAMP stimulation induces expression of negative regulators of TLR signaling in human... [/fig_ref].
LPS treatment of TBEC did lead to a slight but not statistically significant increase in IRAK-M mRNA expression at 4 and 24 h after challenge but not at 48 h [fig_ref] Figure 5: PAMP stimulation induces expression of negative regulators of TLR signaling in human... [/fig_ref]. In contrast, PIC treatment did not alter IRAK-M mRNA levels at the earliest time point but did significantly increase IRAK-M mRNA at the later 24 and 48 h time points [fig_ref] Figure 5: PAMP stimulation induces expression of negative regulators of TLR signaling in human... [/fig_ref]. Thus, A20 expression was induced early after PIC challenge while IRAK-M expression was induced later after PIC challenge. Tollip expression was not altered at any time point examined. We also examined Tollip and A20 protein levels by western blot in the subset of subjects (n = 3) with relatively abundant cells that allowed us to perform additional cell culture studies due to limited availability of alveolar macrophages in other subjects. In support of the mRNA data, Tollip protein levels were not changed significantly after LPS or PIC stimulation [fig_ref] Figure 5: PAMP stimulation induces expression of negative regulators of TLR signaling in human... [/fig_ref]. Consistent with the mRNA data, A20 protein was induced significantly by PIC stimulation in TBEC with a significant increase at 4 and 24 h but not 48 h after challenge [fig_ref] Figure 5: PAMP stimulation induces expression of negative regulators of TLR signaling in human... [/fig_ref]. Western blots for two other subjects are shown in Additional file 1: [fig_ref] Figure 2: Human tracheobronchial epithelial cells [/fig_ref] A-C.
## Pamp-mediated induction of negative regulators in alveolar macrophages overlaps but is distinct from that in tbec
In the paired alveolar macrophages, Tollip mRNA was increased after 24 and 48 h of PIC treatment; in contrast, Tollip mRNA levels were not altered significantly by LPS treatment [fig_ref] Figure 6: PAMP stimulation induces expression of negative regulators of TLR signaling in alveolar... [/fig_ref]. Thus, Tollip expression is induced by PIC in alveolar macrophages but not in TBEC. A20 regulation in alveolar macrophages was similar to that in TBEC: PIC but not LPS induced A20 mRNA expression in macrophages at all the time points (4, 24 and 48 h) [fig_ref] Figure 6: PAMP stimulation induces expression of negative regulators of TLR signaling in alveolar... [/fig_ref]. IRAK-M mRNA marginally increased at 24 h post LPS and PIC stimulation but returned to baseline levels by 48 h [fig_ref] Figure 6: PAMP stimulation induces expression of negative regulators of TLR signaling in alveolar... [/fig_ref].
To confirm the mRNA expression data, we also monitored Tollip and A20 protein levels in the macrophages after LPS and PIC stimulation. We found that both Tollip and A20 protein levels were increased after PIC challenge, but this only reached statistical significance for A20 after 4 h of challenge [fig_ref] Figure 6: PAMP stimulation induces expression of negative regulators of TLR signaling in alveolar... [/fig_ref]. Western blots for two other subjects are shown in Additional file 1: [fig_ref] Figure 2: Human tracheobronchial epithelial cells [/fig_ref] A-C. The more robust effects at the mRNA level rather than at the protein level may be due to the larger variability observed in the protein data.
Together, the above data showed that after PIC stimulation, Tollip, A20 and IRAK-M were all up-regulated in alveolar macrophage, while only A20 and IRAK-M were increased in TBEC. Moreover, there were temporal differences in this induction. LPS, on the other hand, did not increase production of any of these three negative regulators in either cell type.
## Correlation in expression of negative regulators between paired airway epithelial cells and alveolar macrophages
We and others have observed substantial variability in the response of AMs and TBECs from different donors to PAMP stimulation [bib_ref] Contribution of lung macrophages to the inflammatory responses induced by exposure to..., Hiraiwa [/bib_ref] [bib_ref] Innate immune responses of the airway epithelium, Ryu [/bib_ref]. However, to our knowledge, prior studies have not addressed if AMs and TBECs from individual donors acted similarly regarding their responses to PAMPs. We therefore analyzed our cytokine, TLR, and negative regulator expression data to see if expression correlated in the two cell types from individual donors. We found that IL-8, IP-10, TLR3 and TLR4 expression in these paired analyses did not correlate (Additional file 1: [fig_ref] Table 1: Characteristics of Research Subjects [/fig_ref] with p values). In contrast, the negative regulators A20 and IRAK-M showed striking positive correlations after 24 h of PIC stimulation (P < 0.01) .
## The impact of smoking status on tlr expression and pro-inflammatory responses to pic and lps
To determine if smoking status alters the regulation of inflammation in TBEC or alveolar macrophages, we compared cytokine production following PIC or LPS stimulation, the levels of TLR3 and TLR4, and the levels of negative regulators in cells from donors with or without smoking history. PIC treatment induced more IL-8 in TBEC from smokers than non-smokers at the protein and mRNA levels (Additional file 1: . Likewise, LPS treatment induced more IL-8 in alveolar macrophages from smokers than non-smokers at the protein and mRNA levels. (Additional file 1: Thus, smoking enhanced IL-8 production in both cell types. While smoking enhanced IL-8 production in both cell types, this effect was not observed for other pro-inflammatory cytokines. Smoking weakened LPS and PIC-induced TNF-α production in alveolar macrophages , and smoking did not significantly alter IP-10 production in either PIC-stimulated TBEC or alveolar macrophages (Additional file 1: . Although TNF-α was increased in supernatants of alveolar macrophages stimulated with TLR agonists, it was not detectable in airway epithelial cell supernatants [fig_ref] Figure 2: Human tracheobronchial epithelial cells [/fig_ref] under any conditions. Thus, we cannot compare the production of TNF-α between alveolar macrophages and airway epithelial cells stimulated with TLR agonists.
To determine how smoking altered the inflammatory response, we monitored expression of TLR3 and TLR4 and negative TLR regulators. Unstimulated TBEC from non-smokers had greater mRNA expression of TLR3 and TLR4 than TBEC from smokers (P < 0.05) [fig_ref] Figure 9: Smoking history alters TLR expression in human tracheobronchial epithelial cells [/fig_ref]. Moreover, after LPS stimulation, TLR3 and TLR4 mRNA levels significantly increased at 4 h in TBEC from non-smokers compared with smokers (P < 0.05) [fig_ref] Figure 9: Smoking history alters TLR expression in human tracheobronchial epithelial cells [/fig_ref]. In contrast, PIC did not alter TLR3 or TLR4 mRNA levels significantly in either smokers or non-smokers [fig_ref] Figure 9: Smoking history alters TLR expression in human tracheobronchial epithelial cells [/fig_ref]. In alveolar macrophages, TLR3 and TLR4 mRNA levels were not altered significantly by smoking status in the absence or presence of PAMP stimulation [fig_ref] Figure 9: Smoking history alters TLR expression in human tracheobronchial epithelial cells [/fig_ref] , although there was a weak trend towards more TLR3 and TLR4 at 4 h after LPS stimulation [fig_ref] Figure 9: Smoking history alters TLR expression in human tracheobronchial epithelial cells [/fig_ref]. Thus, changes in TLR expression could not account for all the effect of smoking on inflammatory cytokine production in these two cell types.
We also measured the levels of the negative regulators in non-stimulated TBEC and macrophages from smokers and non-smokers. No differences in Tollip and A20 expression were found between cells from the smokers and non-smokers (data not shown). However, IRAK-M mRNA expression in TBEC was lower in the smokers than the non-smokers at 24 h (P = 0.01) (Additional file 1: [fig_ref] Figure 4: Differential activation of NF-κB in human tracheobronchial epithelial cells [/fig_ref].
# Discussion
The present study leverages the use of paired airway epithelial cells and alveolar macrophages from the same donors in order to clearly demonstrate how these two types of critical innate immune cells respond to two major TLR agonists (PIC and LPS) that are relevant to bacterial and viral lung infections.
Paired TBEC and macrophages showed differential immune responses to PIC and LPS stimulation. While a b TBEC have a greater pro-inflammatory response (IL-8, IP-10) to the TLR3 agonist PIC, alveolar macrophages are a stronger responder to LPS, a TLR4 agonist. These differential responses at the inflammatory cytokine level were likely driven by differences in NF-κB activation in the two cell types. In the TBEC, NF-κB activity was higher after PIC stimulation; in contrast, macrophages showed higher NF-κB activity after LPS stimulation. In order to understand the mechanisms behind this differential immune response in the paired cells after LPS and PIC stimulation, we monitored their TLR3 and TLR4 expression. There were no significant differences in the mRNA expression of TLR3 or TLR4 between the paired cells that were not stimulated with either PIC or LPS. Thus baseline expression levels of TLR3 and TLR4 may not be responsible for the differential cell type-specific responses to TLR agonists. Because of the limited availability of primary human alveolar macrophages, we analyzed TLR mRNA but not protein expression. Previous studies indicate that mRNA expression of TLRs such as TLR3 in epithelial cells is consistent with protein expression determined using flow cytometry [bib_ref] Functional effects of toll-like receptor (TLR)3, 7, 9, RIG-I and MDA-5 stimulation..., Tengroth [/bib_ref] , suggesting that TLR mRNA is a reasonable surrogate for TLR protein levels. Nevertheless, this is one limitation of our study. It has been found that TLR4 in AEC is normally localized in the endosomal compartment, but is translocated to the cell surface to recognize pathogens or environmental LPS exposure [bib_ref] Epithelial expression of TLR4 is modulated in COPD and by steroids, salmeterol..., Macredmond [/bib_ref] [bib_ref] LipopolysaccharideActivates Distinct Signaling Pathways in Intestinal Epithelial Cell Lines Expressing Toll-Like Receptors, Cario [/bib_ref]. In contrast, macrophages express TLR4 primarily at the cell surface membrane [bib_ref] LipopolysaccharideActivates Distinct Signaling Pathways in Intestinal Epithelial Cell Lines Expressing Toll-Like Receptors, Cario [/bib_ref]. While TLR3 in macrophages is usually observed intracellularly, it is found on the cell surface as well as in the cytoplasm of AEC [bib_ref] Functional effects of toll-like receptor (TLR)3, 7, 9, RIG-I and MDA-5 stimulation..., Tengroth [/bib_ref] [bib_ref] TLR-dependent human mucosal epithelial cell responses to microbial pathogens, Mcclure [/bib_ref]. The different localization of TLR3 and TLR4 in various types of lung cells may provide an additional explanation for their different responses to TLR agonists.
Despite the fact that the two cell types showed no differences in TLR4 and TLR3 expression following LPS or PIC stimulation, their NF-κB activity and subsequent inflammatory cytokine production differed. These data suggest the involvement of other mechanisms or regulators in the differential activation of epithelial cells and macrophages exposed to TLR agonists. One possibility that we explored was expression of negative regulators of TLR signaling.
The literature suggests that TLR-induced expression of immune negative regulators represents a critical negative feedback mechanism to prevent excessive TLR signaling. These negative regulators function at multiple levels in the TLR signaling pathway ranging from inhibition of receptor complex protein formation to NF-κB a c b d activation [bib_ref] A20 is a negative regulator of IFN regulatory factor 3 signaling, Saitoh [/bib_ref] [bib_ref] A20-mediated negative regulation of canonical NF-κB signaling pathway, Pujari [/bib_ref] [bib_ref] Negative regulation of toll-like receptor-mediated signaling by Tollip, Zhang [/bib_ref] [bib_ref] Differential response to bacteria, and TOLLIP expression, in the human respiratory tract, Moncayo-Nieto [/bib_ref]. Previous studies have elucidated the role of immune negative regulators such as A20, Tollip and IRAK-M in the regulation of LPS-mediated pro-inflammatory responses of unpaired airway epithelial cells and lung macrophages. For example, A20 was found to attenuate airway epithelial cell responses to TLR2 and TLR4 agonists [bib_ref] A20 inhibits toll-like receptor 2-and 4-mediated interleukin-8 synthesis in airway epithelial cells, Gon [/bib_ref] [bib_ref] Induction of the inflammatory regulator A20 by gibberellic acid in airway epithelial..., Reihill [/bib_ref] as well as endotoxin tolerance in macrophages [bib_ref] Immune responsive gene 1 (IRG1) promotes endotoxin tolerance by increasing A20 expression..., Li [/bib_ref]. The role of A20 in antiviral responses has not been well studied. To the best of our knowledge, our study is the first one to investigate the time course of immune negative regulator expression in the response of paired lung cells (AM, AEC) to the viral mimic PIC as well as LPS. This allowed us to determine the potential mechanisms of differential responses of airway epithelial cells and alveolar macrophages to various TLR agonists. We found that the expression of these negative regulators differed with respect to time and cell type. Alveolar macrophages up-regulated multiple immune negative regulators including Tollip, A20 and IRAK-M after PIC stimulation. While A20 mRNA and protein were induced rapidly after PIC stimulation, Tollip and IRAK-M were induced at later times. Importantly, A20 induction was maintained throughout the entire 48 h post PIC stimulation. In contrast, TBEC showed significant induction of A20 only at the earliest time point after PIC stimulation, but at later times (24 and 48 h), A20 expression significantly declined. Moreover, unlike macrophages, PIC-stimulated TBEC did not significantly change the expression of Tollip. We speculate that the lack of PIC-induced Tollip induction coupled with the transient A20 induction in TBEC may contribute to their more robust pro-inflammatory responses than the alveolar macrophages. The early induction and late reduction of A20 in PIC-stimulated TBEC are consistent with a previous publication performed by in the human cytomegalovirus (HCMV) infection model [bib_ref] Biphasic regulation of A20 gene expression during human cytomegalovirus infection, Gu [/bib_ref]. Together, our data indicate the possibility that these immune negative regulators could affect the differential response of various types of lung immune cells to PAMP stimulation. Future functional studies using the RNA interference may further our understanding of these immune negative regulators in modulating TLR agonist-mediated pro-inflammatory responses in paired alveolar macrophages and airway epithelial cells.
The design of our current study was also aimed to address the effect of smoking status on pro-inflammatory response in paired TBEC and alveolar macrophages as smoking has been linked to changes in the immune response in airway cells [bib_ref] Cigarette smoke inhibits airway epithelial cell innate immune responses to bacteria, Kulkarni [/bib_ref] [bib_ref] Cigarette smoke attenuation of poly I:C-induced innate antiviral responses in human PBMC..., Mian [/bib_ref]. Macrophages from smoking subjects produced less TNF-α but more IL-8 after LPS stimulation. However, in TBEC, IL-8 induction was significantly higher in smoker's cells than non-smoker's cells after PIC but not LPS stimulation. One possible explanation for these differences in the cells previously exposed to cigarette smoke was the difference in TLR3 and TLR4 expression. As observed previously [bib_ref] Toll-like receptor 2 expression is decreased on alveolar macrophages in cigarette smokers..., Droemann [/bib_ref] [bib_ref] Effect of cigarette smoke extract on lipopolysaccharide-activated mitogenactivated protein kinase signal transduction..., Li [/bib_ref] [bib_ref] Smoking decreases the response of human lung macrophages to double-stranded RNA by..., Todt [/bib_ref] [bib_ref] Effects of cigarette smoke on toll-like receptor (TLR) activation of chronic obstructive..., Metcalfe [/bib_ref] , smoking down-regulated TLR3 and TLR4 expression in TBEC with and without stimulation. The down-regulation of TLRs by smoking was consistent with the decreased TNF-α and IP-10 production in the smokers, but not the increased IL-8 production following stimulation. Many studies have shown that cigarette smoking can inhibit inflammatory cytokine production (TNF-α, IL-6) and host defense responses (IFN-β, IP-10) in response to TLR stimulation [bib_ref] Cigarette smoke attenuation of poly I:C-induced innate antiviral responses in human PBMC..., Mian [/bib_ref] [bib_ref] Impact of tobacco-smoke on key signaling pathways in the innate immune response..., Birrell [/bib_ref] [bib_ref] Human primary airway epithelial cells isolated from active smokers have epigenetically impaired..., Wu [/bib_ref]. However, there are discrepancies in the literature regarding the effect of smoking on IL-8 production. While some studies have demonstrated a reduction in IL-8 production or no effect from smoking [bib_ref] Tobacco smoking inhibits expression of proinflammatory cytokines and activation of IL-1R-associated kinase,..., Chen [/bib_ref] , other studies have reported increased production of IL-8 and other neutrophil chemokines from smoker's airway cells after TLR agonist stimulation [bib_ref] Smoking decreases the response of human lung macrophages to double-stranded RNA by..., Todt [/bib_ref] [bib_ref] Effects of cigarette smoke on toll-like receptor (TLR) activation of chronic obstructive..., Metcalfe [/bib_ref] [bib_ref] Cigarette smoke induces IL-8, but inhibits eotaxin and RANTES release from airway..., Oltmanns [/bib_ref]. The signal transduction mechanisms controlling TNF-α and IL-8 production appear to be different in the alveolar macrophage. In macrophages, smoking has been shown to activate p38 MAPK that controls transcription, stabilization of mRNA and secretion of IL-8 but not TNF-α [bib_ref] Effects of cigarette smoke on toll-like receptor (TLR) activation of chronic obstructive..., Metcalfe [/bib_ref].
The induction of the negative regulators observed in our study may also contribute to the differences in cytokine production in cells from smokers compared to non-smokers. Our data showed that smoking decreased expression of IRAK-M without stimulation in TBEC, which is consistent with the increased IL-8 production that we observed in smokers.
A significant strength of our study was our bank of paired TBECs and AMs, which allowed us to determine if the response to TLR agonists in different cell types correlated within individual donors. We did not observe any correlation in inflammatory cytokine production between the two paired cell types. Thus, the robustness of an individual's immune response in one key lung innate immune cell type did not correlate with the robustness of that response in the other cell type. Despite this observation, we did identify a very strong correlation in PIC-induced expression of negative regulators in the two cell types. Consistent with the mRNA data, in preliminary studies on a small number of samples, we also observed a trend of positive correlations of A20 and Tollip protein levels between the two cell types (data not shown). This indicates that an individual's ability to restrain inflammation in the lung may extend to multiple cell types.
One limitation of our study is that cultured cells may not maintain the phenotype of the in vivo cells, including the loss of the full differentiation status. Unfortunately, this is an inherent issue for every cell culture study. Nevertheless, one advantage of the cell culture model is to allow us to quantitatively analyze the impact of multiple TLR agonists. In the future study, we plan to validate our results by using other culture methods such as air-liquid interface to maintain the differentiation status of airway epithelial cells. Also, it will be interesting in the future to study how airway epithelial cells and alveolar macrophages cooperate in the effective lung defense against any environmental hazards using the co-culture model.
# Conclusions
In summary, we find that TBEC and alveolar macrophages exhibit different responses to different TLR agonists. While these cells did not exhibit different levels of TLR mRNAs, they did exhibit different expression of negative regulators of TLR signaling, which could impact the extent of the immune response in these two cell types. Addtionally, our data suggest a mild impact of smoking status on the pro-(e.g., IL-8 and IP-10) and anti-inflammatory (e.g., A20 and Tollip) responses at the baseline and/or after TLR agonist stimulation. We propose the involvement of potential pathways in regulating the different responses of airway epithelial cells and alveolar macrophages to TLR agonists in [fig_ref] Figure 1: Characterization of primary human tracheobronchial epithelial cell [/fig_ref]. Understanding the role of immune negative regulators in primary human lung cells may provide new and promising therapeutic strategies to control pulmonary inflammation and infection.
## Additional file
Additional file 1: [fig_ref] Figure 1: Characterization of primary human tracheobronchial epithelial cell [/fig_ref]. Comparison of baseline (untreated) expression of TLR3 and TLR4 in human tracheobronchial epithelial cells (TBEC) and alveolar macrophages (AM). No significant differences of TLR3 or TLR4 mRNA expression were found between the two cell types. [fig_ref] Figure 2: Human tracheobronchial epithelial cells [/fig_ref]. Detection of A20, Tollip in human tracheobronchial epithelial cells (TBEC) and alveolar macrophages before and after PAMP stimulation at different time points. Western blot showing A20, Tollip, and β-actin proteins from TBEC and alveolar macrophages before (−) and after treatments with LPS, Poly(I:C) (PIC) for 4, 24 and 48 h (N = 3). These blots were also used for desitometry in Figs. 5 and 6. . Comparison of IL-8 and IP-10 expression in smokers and non-smokers human tracheobronchial epithelial cells (TBEC) and macrophages. mRNA expression of IL-8 and IP-10 in TBEC and alveolar macrophages in the absence (−) or presence of LPS or Poly(I:C) (PIC) at 24 h was compared between smokers (S, n = 4) and non-smokers (NS, n = 4). These data are a re-analysis of the data displayed in [fig_ref] Figure 2: Human tracheobronchial epithelial cells [/fig_ref]. [fig_ref] Figure 4: Differential activation of NF-κB in human tracheobronchial epithelial cells [/fig_ref]. The effect of smoking status on IRAK-M expression. IRAK-M mRNA expression was examined after 24 h of culture in non-stimulated human tracheobronchial epithelial cells from smokers (S, n = 4) and non-smokers (NS, n = 4). NS have higher IRAK-M expression than S. [fig_ref] Table 1: Characteristics of Research Subjects [/fig_ref]. Correlation analysis between paired airway epithelial cells and alveolar macrophages from the same donors with P-values.
[fig] Figure 1: Characterization of primary human tracheobronchial epithelial cell (TBEC) and alveolar macrophages (AM) under submerged cultures. a TBEC showing features of basal cells with positive staining for cytokeratin-5 (KRT5 in red, DAPI in blue, 400X.); b Negative control for KRT5 stained with DAPI in blue (400X); c Primary alveolar macrophages (AM) stained positive for CD68 (red) and DAPI (blue, 400X); d Negative control for CD68, stained with DAPI in blue (400X) p65 kit following the manufacturer's instructions. The data were expressed as optical density (OD) value. ELISA for human IL-8 (CXCL8), IP-10 (CXCL10) and TNF-α IL-8, IP-10 and TNF-α protein levels were measured in cell supernatants using specific DuoSet ELISA kits (R&D Systems, Minneapolis, MN) as per the manufacturer's instructions. RNA extraction and RT-PCR for human TLRs, Tollip, A20, IRAK-M, IL-8 and IP-10 RNA was extracted from cells stored in RLT using an RNeasy Plus kit (Qiagen). RNA was reverse transcribed using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, California). Real-time PCR was performed on the CFX96 (Bio-Rad) using TaqMan gene expression assays from Applied Biosystems (Life Technologies, Foster City, CA, USA). An identical threshold cycle (Ct) was applied for each gene of interest (TLR3, TLR4, Tollip, A20, IRAK-M, IL-8 and IP-10). [/fig]
[fig] Figure 2: Human tracheobronchial epithelial cells (TBEC) and alveolar macrophages display different responses to different pathogen associated molecular patterns (PAMPs). IL-8 and IP-10 protein production in supernatants of cultured human TBEC (a, b and c) and cultured alveolar macrophages (d, e and f) in the absence (−) and presence of Poly(I:C) (PIC) or LPS treatment for 4, 24 and 48 h. N = 8 donor subjects a b Fig. 3 PAMP stimulation increases inflammatory cytokine mRNA levels in primary human lung cells. mRNA expression of IL-8 (a) and IP-10 (b) in human tracheobronchial epithelial cells (TBEC) and alveolar macrophages in the absence (−) and presence of LPS and Poly(I:C) (PIC) at 24 h. N = 8 donor subjects. The horizontal lines indicate the medians [/fig]
[fig] Figure 4: Differential activation of NF-κB in human tracheobronchial epithelial cells (TBEC) and alveolar macrophages in responses to PAMPs. NF-κB p65 levels were measured in nuclear extracts from paired TBEC (a) and alveolar macrophages (b) in the absence (−) and presence of treatments with LPS and Poly(I:C) (PIC) at 4, 24 and 48 h. N = 3 donor subjects [/fig]
[fig] Figure 5: PAMP stimulation induces expression of negative regulators of TLR signaling in human tracheobronchial epithelial cells (TBEC). mRNA expression of negative regulators Tollip (a), A20 (b) and IRAK-M (c) in TBEC in the absence (−) and presence of LPS or Poly(I:C) (PIC) at 4, 24 and 48 h. N = 8 donor subjects. Median values are shown as horizontal lines. Densitometric analysis of Tollip (d) and A20 (e) western blots on lysates of TBEC in the absence (−) or presence of treatments with LPS, PIC at 4, 24 and 48 h. β-actin was included as a protein loading control to normalize Tollip or A20 expression. N = 3 donor subjects (additional blots in Additional file 1:Figure S2) [/fig]
[fig] Figure 6: PAMP stimulation induces expression of negative regulators of TLR signaling in alveolar macrophages. mRNA expression of the negative regulators Tollip (a), A20 (b) and IRAK-M (c) in macrophages in the absence (−) and presence of LPS, Poly(I:C) (PIC) at 4, 24 and 48 h. N = 8 donor subjects. Median values are shown as horizontal lines. Densitometric analysis of Tollip (d) and A20 (e) western blots on lysates of macrophages in the absence (−) or presence of treatments with LPS, PIC at 4, 24 and 48 h. β-actin was included as a protein loading control to normalize Tollip or A20 expression. N = 3 donor subjects (additional blots in Additional file 1: [/fig]
[fig] Figure 8, Figure 7: Smoking history alters IL-8 expressionat the protein level in human tracheobronchial epithelial cells (TBEC) and macrophages, and decreases TNF-α production in macrophages treated with LPS. a IL-8 production was measured in TBEC and alveolar macrophages in the absence (−) or presence of LPS or poly(I:C) (PIC) at 24 h, and compared between smokers (S, n = 4) and non-smokers (NS, n = 4). These data are a re-analysis of the data displayed inFig. 3a. There is significant induction of IL-8 in smokers' TBEC after PIC stimulation, and in smokers' macrophages after LPS stimulation. b TNF-α production in supernatants of cultured alveolar macrophages. The cells from smokers (S, n = 4) and non-smokers (NS, n = 4) were treated in the absence (−) or presence of LPS or PIC for 4, 24 and 48 h. NS trend to have higher levels of TNF-α at all the time points (P = 0.1) Expression of the negative regulators A20 and IRAK-M are strongly correlated between paired human tracheobronchial epithelial cells (TBEC) and alveolar macrophage samples from the same individual donors. A20 and IRAK-M mRNA expression showed significant correlation between the two cell types after 24 h of PIC stimulation. N = 8 donor subjects [/fig]
[fig] Figure 9: Smoking history alters TLR expression in human tracheobronchial epithelial cells (TBEC) but not macrophages. TLR3 and TLR4 mRNA expression in TBEC (a and b) and alveolar macrophages (c and d) that were not treated, or treated with LPS or Poly(I:C) (PIC) for 4 and 24 h. N = 8 donor subjects including four smokers (S) and four non-smokers (NS) [/fig]
[table] Table 1: Characteristics of Research Subjects [/table]
|
Construction and Accuracy Assessment of Patient-Specific Biocompatible Drill Template for Cervical Anterior Transpedicular Screw (ATPS) Insertion: An In Vitro Study
Background: With the properties of three-column fixation and anterior-approach-only procedure, anterior transpedicular screw (ATPS) is ideal for severe multilevel traumatic cervical instabilities. However, the accurate insertion of ATPS remains challenging. Here we constructed a patient-specific biocompatible drill template and evaluated its accuracy in assisting ATPS insertion.Methods: After ethical approval, 24 formalin-preserved cervical vertebrae (C2-C7) were CT scanned. 3D reconstruction models of cervical vertebra were obtained with 2-mm-diameter virtual pin tracts at the central pedicles. The 3D models were used for rapid prototyping (RP) printing. A 2-mm-diameter Kirschner wire was then inserted into the pin tract of the RP model before polymethylmethacrylate was used to construct the patient-specific biocompatible drill template. After removal of the anterior soft tissue, a 2-mm-diameter Kirschner wire was inserted into the cervical pedicle with the assistance of drill template. Cadaveric cervical spines with pin tracts were subsequently scanned using the same CT scanner. A 3D reconstruction was performed of the scanned spines to get 3D models of the vertebrae containing the actual pin tracts. The deviations were calculated between 3D models with virtual and actual pin tracts at the middle point of the cervical pedicle. 3D models of 3.5 mm-diameter screws were used in simulated insertion to grade the screw positions.Findings: The patient-specific biocompatible drill template was constructed to assist ATPS insertion successfully. There were no significant differences between medial/lateral deviations (P = 0.797) or between superior/inferior deviations (P = 0.741). The absolute deviation values were 0.8260.75 mm and 1.1060.96 mm in axial and sagittal planes, respectively. In the simulated insertion, the screws in non-critical position were 44/48 (91.7%).Conclusions:The patient-specific drill template is biocompatible, easy-to-apply and accurate in assisting ATPS insertion. Its clinical applications should be further researched.
# Introduction
Surgical interventions are common for degenerative cervical spine diseases [bib_ref] The incidence of adjacent segment degeneration after cervical disc arthroplasty (CDA): a..., Yang [/bib_ref] [bib_ref] Instrumentation of the osteoporotic spine: biomechanical and clinical considerations, Ponnusamy [/bib_ref] [bib_ref] Anterior Cervical Discectomy with Arthroplasty versus Arthrodesis for Single-Level Cervical Spondylosis: A..., Fallah [/bib_ref]. Due to the three-column fixation property, transpedicular screw fixation via the posterior approach has been shown to have superior stabilization capabilities in several biomechanical and clinical surveys [bib_ref] Effect of constrained posterior screw and rod systems for primary stability: biomechanical..., Schmidt [/bib_ref] [bib_ref] A biomechanical comparison of three surgical approaches in bilateral subaxial cervical facet..., Kim [/bib_ref] [bib_ref] Cervical pedicle screws versus lateral mass screws. Anatomic feasibility and biomechanical comparison, Jones [/bib_ref]. However, due to the posterior musculature stripping, the posterior approach can cause significant myofascial pain and lead to significant postoperative axial symptoms and neck pain [bib_ref] -and 5-level anterior fusions of the cervical spine: review of literature and..., Koller [/bib_ref] [bib_ref] Investigation of axial symptoms after cervical laminoplasty, using questionnaire survey, Ohnari [/bib_ref] [bib_ref] Operative treatment of cervical spondylotic myelopathy, Rao [/bib_ref]. On the contrary, the anterior approach is less traumatic with no damage to the paravertebral muscles and allows for anterior instrumentation as far as T1 [bib_ref] Distractive flexion injuries of the subaxial cervical spine treated with anterior plate..., Henriques [/bib_ref] [bib_ref] Preoperative magnetic resonance imaging screening for a surgical decision regarding the approach..., Fraser [/bib_ref]. However, since the screws in the anterior approach are anchored in the cancellous vertebral body, the biomechanical stability of anterior plate fixation is limited, leading to significant failure rates [bib_ref] Comparison of multisegment anterior cervical fixation using bone strut graft versus a..., Brazenor [/bib_ref]. Thus, for a successful $2-level corpectomy or operation for severe traumatic three-column instabilities, a combined anterior and posterior approach was found to be desirable [bib_ref] One stage anteriorposterior approach for traumatic atlantoaxial instability combined with subaxial cervical..., Wang [/bib_ref] [bib_ref] Combined anterior and posterior surgery for treatment of cervical fracture-dislocation in patients..., Lv [/bib_ref] [bib_ref] Posterior cervical fixation with nitinol shape memory loop in the anterior-posterior combined..., Yu [/bib_ref] but it would require a secondary posterior approach procedure that might lead to a significant increase in morbidity.
Anterior transpedicular screw (ATPS) in clinical application was first reported by Aramoni et al. [bib_ref] Anterior pedicle screw fixation for multilevel cervical corpectomy and spinal fusion, Aramomi [/bib_ref]. After corpectomy at one to three levels in 9 patients, Aramoni et al. placed ATPS under visualization of the pedicles to affix fibular grafts to cervical pedicles [bib_ref] Anterior pedicle screw fixation for multilevel cervical corpectomy and spinal fusion, Aramomi [/bib_ref]. Koller et al. demonstrated the anatomical feasibility of ATPS and found the pull-out strength of ATPS to be 2.5-fold that of vertebral body screws [bib_ref] Cervical anterior transpedicular screw fixation (ATPS)-Part II. Accuracy of manual insertion and..., Koller [/bib_ref] [bib_ref] Cervical anterior transpedicular screw fixation. Part I: Study on morphological feasibility, indications,..., Koller [/bib_ref]. The ATPS technique merges the biomechanical merits of posterior transpedicular fixation with the surgical benefits of anterior-approach-only procedures [bib_ref] Cervical anterior transpedicular screw fixation. Part I: Study on morphological feasibility, indications,..., Koller [/bib_ref] , because it can increase initial construct stability in an anterior surgery which is believed to be best beneficial for some severe multilevel cervical instabilities [bib_ref] Long-term biomechanical stability and clinical improvement after extended multilevel corpectomy and circumferential..., Acosta [/bib_ref] [bib_ref] Comparison of multisegment anterior cervical fixation using bone strut graft versus a..., Brazenor [/bib_ref] [bib_ref] -and 5-level anterior fusions of the cervical spine: review of literature and..., Koller [/bib_ref].
Accurate insertion is a key to successful application of ATPS in clinic. In fluoroscopy-guided manual insertion of ATPS, a percentage of 78.3% was reported for correctly placed screws and non-critical pedicle breaches in the axial plane [bib_ref] Cervical anterior transpedicular screw fixation (ATPS)-Part II. Accuracy of manual insertion and..., Koller [/bib_ref]. Yukawa et al. reported successful insertion of anterior pedicle screws in 6 patients with the aid of fluoroscopic images of the pedicle axis [bib_ref] Anterior cervical pedicle screw and plate fixation using fluoroscope-assisted pedicle axis view..., Yukawa [/bib_ref]. However, the sample number was small and the surgery required much time and experienced physicians. Koller et al., using the electronic conductivity device (ECD), found a high accuracy rate of ATPS insertion with no critical screw positioned in the axial or sagittal plane [bib_ref] In vitro study of accuracy of cervical pedicle screw insertion using an..., Koller [/bib_ref]. However, the physical property of ECD prevents it from penetrating the dense cortical bone. Patientspecific drill templates were developed to assist screw insertion using 3D reconstruction, computer aided design (CAD) and rapid prototyping (RP) techniques and good accuracy of screw insertion was obtained [bib_ref] Efficacy and accuracy of a novel rapid prototyping drill template for cervical..., Lu [/bib_ref] [bib_ref] A novel patient-specific navigational template for cervical pedicle screw placement, Lu [/bib_ref] [bib_ref] Accuracy and efficacy of thoracic pedicle screws in scoliosis with patient-specific drill..., Lu [/bib_ref] [bib_ref] A novel computer-assisted drill guide template for lumbar pedicle screw placement: a..., Lu [/bib_ref] [bib_ref] A novel computer-assisted drill guide template for placement of C2 laminar screws, Lu [/bib_ref] [bib_ref] Rapid prototyping drill guide template for lumbar pedicle screw placement, Lu [/bib_ref]. However, these techniques have been used just in the posterior pedicle approach but not in the anterior pedicle approach. Also, the materials used in them, such as photosensitive resin, possess significant cytotoxicity [bib_ref] The biocompatibility of resin-modified glassionomer cements for dentistry, Nicholson [/bib_ref] [bib_ref] In vitro cytotoxicity and in vivo biocompatibility of contemporary resin-modified glassionomer cements, Souza [/bib_ref]. In the process of surgical drilling, debris of the drill template can get in contact with the wounded area, a potential danger if the debris cannot be totally rinsed.
Therefore, accurate and biocompatible insertion of ATPS remains a challenge. To address this challenge, we first adopted a new strategy to construct a biocompatible drill template for ATPS insertion using 3D reconstruction, RP production and reverse mold manufacture techniques. Secondly, we evaluated the accuracy of the drill template in assisting ATPS insertion.
# Materials and methods
# Ethics statement
Ethical approval was obtained from the Human Research Ethics Committee, Southern Medical University, Guangzhou, China. The subjects gave informed consent. And all consent was written in nature regarding body donation for research.
## Specimens
Twenty four formalin-preserved cervical vertebrae (range C2-C7) from four human cadavers (3 males and 1 female, from 52 to 68 years of age, mean 61.5 years) were obtained. The entire specimens were imaged using a Brilliance CT 64-channel scanner (Philips, Eindhoven, The Netherlands). In-plane pixel size was 0.5 mm and slice thickness was 0.705 mm. All the 48 pedicles in the 24 cervical vertebrae were used after the CT scan images showed no significant bone defects.
## Three-dimensional reconstruction of cervical models with virtual pin tracts
The datasets of cervical specimens were processed and edited with Mimics software v14.11 (Materialise Corp., Leuven, Belgium). 3D reconstructions were obtained from the 2D CT images. An interactive image processing strategy, such as ''Threshold'' and ''Region growth'', was used to segment the contours of each vertebra to obtain the 3D-reconstructed models.
The 2 cylinders, 2 mm in diameter and pre-designed in Unigraphics NX 6.0 (Siemens PLM Software, Plano, TX), were imported into Mimics software where it could be freely translated and rotated. It was made sure that the cylinders were located at the central cervical pedicle by visual observation. With the tool ''subtraction'' under ''Boolean operation'' in the Geomagic studioH software, 3D models of cervical vertebrae with bilateral pin tracts were obtained [fig_ref] Figure 1: 3D model of cervical vertebra with virtual pin tracts [/fig_ref] and saved in group 1 in stereolithography file format (.stl) supported by many software packages and widely used for rapid prototyping and computeraided manufacturing.
## Production of biocompatible drill templates
The ''.stl'' files of cervical vertebrae were processed using the software Zprint 7.10 (Z Corporation, Burlington, MA) and printed on the Z Corporation 3D printer Spectrum Z TM 510 (Z Corporation, Burlington, MA). The 3D models were virtually cut into thin layers of 0.0875 mm intervals with the Zprint 7.10 software and transferred into the Spectrum Z510 for rapid prototyping [fig_ref] Figure 2: Production of biocompatible navigation template with RP model [/fig_ref]. A Kirschner wire was then inserted into the pin tract of the RP model. Polymethylmethacrylate (PMMA), which is usually used as bone cement and has good biocompatibility [bib_ref] Biocompatibility and bone formation with porous modified PMMA in normal and irradiated..., Lye [/bib_ref] [bib_ref] Fatigue and biocompatibility properties of a poly(methyl methacrylate) bone cement with multi-walled..., Ormsby [/bib_ref] [bib_ref] Antibiotic release from an experimental biodegradable bone cement, Gerhart [/bib_ref] , was used to construct the drill templates [fig_ref] Figure 2: Production of biocompatible navigation template with RP model [/fig_ref]. Also, to allow for convenience and easy handling, a grip was created at the top of the drill template. In addition, since the surgery field was narrow, the base of the drill template was not allowed to exceed the juncture of the vertebral body and the transverse process. After the PMMA was solidified, the Kirschner wire was pulled out to finalize the drill template.
## Cadaveric kirschner wire insertion
Anterior soft tissue was removed from the vertebrae. The drill template was put in place by hand and compressed slightly to the anterior surface of cervical vertebrae. A 2-mm-diameter Kirschner wire was then drilled into the cervical pedicle with the assistance of the drill template.
## Secondary 3d reconstruction of cervical models with pin tracts
After all pin tracts were drilled, the cadaveric cervical specimens were scanned with the same CT scanner using the same parameters. Since the Kirschner wires produced image artifacts, they were pulled out before the image acquisition. 3D models of each vertebra with pin tracts were obtained with same segmentation and reconstruction strategy, and saved in group 2 in ''.stl'' file format.
## Assessment of accuracy of screw insertion
Accuracy of ATPS insertion with the assistance of the drill templates was evaluated by a reverse engineering process using the software Geomagic studioH, version 11 (Geomagic, Inc., Morrisville, NC). The 3D models of groups 1 and 2 were imported into the Geomagic software, and the deviations at the middle point of the pedicle in the axial and sagittal planes were calculated. The axial plane's deviations towards the lateral side were recorded as positive values and the deviations towards the medial side as negative values. The sagittal plane's deviations towards the superior and inferior sides were recorded as positive and negative values, respectively.
Aligned with the pin tract of the 3D model of group 2, a predesigned 3D screw model (3.5 mm in diameter) was imported into Mimics to simulate the screw insertion [fig_ref] Figure 4: Accuracy evaluation with screw simulation insertion in the axial [/fig_ref]. A grading was used to distinguish non-critical and critical screw positions [bib_ref] Cervical anterior transpedicular screw fixation (ATPS)-Part II. Accuracy of manual insertion and..., Koller [/bib_ref] [bib_ref] In vitro study of accuracy of cervical pedicle screw insertion using an..., Koller [/bib_ref] [bib_ref] The vertebral artery and the cervical pedicle: morphometric analysis of a critical..., Tomasino [/bib_ref]. Briefly, the grading consists of the following [fig_ref] Figure 5: Illustration of grades of ATPS positions in cervical pedicle [/fig_ref] :
Grade 1: Screw positioned at the center of the pedicle. Grade 2: Less than one-third of the screw cross-section (#1.2 mm with a 3.5-mm diameter screw) penetrating the cortex.
Grade 3: Between one-third and one-half of the screw crosssection penetrating the cortex (or deviation ,2 mm).
Grade 4: More than one-half of the screw cross-section penetrating the cortex (or deviation $2 mm).
Grade 5: Deviation equal or greater than the screw diameter. Non-critical pedicle breaches corresponded to grades 1 and 2. Critical pedicle breaches, with the potential of posing a risk to the vertebral artery (VA), nerve root or dural sac, corresponded to grades 3-5 [bib_ref] In vitro study of accuracy of cervical pedicle screw insertion using an..., Koller [/bib_ref].
# Statistical analysis
Independent-sample T test was used to analyze the screw direction differences between the deviations towards lateral and medial in the axial plane and towards superior and inferior in the sagittal plane. A P value ,0.05 was considered as statistically significant. And, to show the real deviations, the absolute deviation values were calculated to get their means and standard deviations in the axial and sagittal planes, respectively.
With the scoring system [bib_ref] Cervical anterior transpedicular screw fixation (ATPS)-Part II. Accuracy of manual insertion and..., Koller [/bib_ref] [bib_ref] In vitro study of accuracy of cervical pedicle screw insertion using an..., Koller [/bib_ref] , assessment of pedicle screw position could be performed both in the axial and sagittal plane. Each point was assigned to each of the five grades of screw position. The accuracy score in the axial plane (range 1-5 points) and the accuracy score in the sagittal plane (range 1-5 points) was summed and described as the accuracy sum score with its
# Results
With the 3D reconstruction, rapid prototyping production and mold manufacture techniques, the patient-specific biocompatible drill templates were constructed successfully. During the operation, the drill template fit the position easily to allow no significant free motion between the drill template and the anterior cervical surface. The Kirschner wires were inserted into the cervical pedicle easily with the assistance of the patient-specific biocompatible drill template.
Calculation showed no significant difference between the deviations towards lateral and medial in the axial plane (t = 20.258, P = 0.797). The absolute deviation value in axial plane was 0.8260.75 mm [fig_ref] Figure 6: Absolute deviation values in the axial and sagittal planes [/fig_ref]. There was no significant difference either between the deviations towards superior and inferior in the sagittal plane (t = 0.332, P = 0.741). The absolute deviation value in the sagittal plane was 1.1060.96 mm [fig_ref] Figure 6: Absolute deviation values in the axial and sagittal planes [/fig_ref].
In the simulated insertion of 3.5 mm-diameter screw, one screw position was grade 3 (2.1%), one was grade 2 (2.1%) and the others were grade 1 (95.8%) in the axial plane. The mean accuracy score in the axial plane was 1.02. Three pedicle screw positions were grade 3 (6.25%), 4 were grade 2 (8.3%) and the others were grade 1 (85.4%) in the sagittal plane. The mean accuracy score in the sagittal plane was 1.22. The accuracy sum score showed a mean of 2.27 points with a possible total of 10 points. Summing up the screw positions in the axial and sagittal planes, the screws in a noncritical position were 44/48 (91.7%) and those in a critical position were 4/48 (8.3%).
# Discussion
Pedicular screw insertion has generally been considered to be very risky because it can injure VA, the spinal cord or nerve roots seriously. Because of few landmarks on the anterior surface of cervical vertebra and also a relative long distance between the anterior surface and the pedicle, insertion of ATPS is more difficult and dangerous than posterior transpedicular fixation. Several methods have been explored for precise anterior cervical transpedicular screw placement, including the fluoroscopy-guided manual insertion [bib_ref] Cervical anterior transpedicular screw fixation (ATPS)-Part II. Accuracy of manual insertion and..., Koller [/bib_ref] [bib_ref] Cervical anterior transpedicular screw fixation. Part I: Study on morphological feasibility, indications,..., Koller [/bib_ref] , fluoroscopic images of the pedicle axis [bib_ref] The vertebral artery and the cervical pedicle: morphometric analysis of a critical..., Tomasino [/bib_ref] and ECD [bib_ref] In vitro study of accuracy of cervical pedicle screw insertion using an..., Koller [/bib_ref]. It is also possible to use CT-based and fluoroscopy-based computer assisted surgeries (CAS) to assist insertion of ATPS because they have been used to assist screw insertion in posterior transpedicular fixation and yielded high accuracy [bib_ref] Accuracy of pedicle screw placement: a systematic review of prospective in vivo..., Gelalis [/bib_ref] [bib_ref] Clinical accuracy of three-dimensional fluoroscopy-based computer-assisted cervical pedicle screw placement: a retrospective..., Ishikawa [/bib_ref]. The rate of pedicle perforation in posterior cervical or cervical-thoracic fixation using pedicle screws was 8.6% in the conventional group and 3.0% in the CAS group. Gelalis et al. [bib_ref] Accuracy of pedicle screw placement: a systematic review of prospective in vivo..., Gelalis [/bib_ref] found the percentages of screws fully contained in the pedicle ranged from 89 to 100% using CT-based CAS and 81 to 92% using fluoroscopy-based CAS. However, several caveats should be considered: (1) the learning curve to master these complex techniques is relatively long; (2) errors may occur when adjacent segments of the spine shift intraoperatively or if the registration frame and optical array shift; (3) tracking of optical array devices can be obscured by the surgeon or surgical tools; (4) cost of equipment is high; and (5) surgical time is long [bib_ref] Efficacy and accuracy of a novel rapid prototyping drill template for cervical..., Lu [/bib_ref].
The patient-specific drill templates eliminated the need for expensive equipment and a time-consuming procedure in an operating room [bib_ref] A novel patient-specific navigational template for cervical pedicle screw placement, Lu [/bib_ref]. They were initially used in hip and knee surgery [bib_ref] Computer assisted orthopaedic surgery with image based individual templates, Radermacher [/bib_ref] and so far have been developed to assist screw insertion in cervical surgery, yielding good accuracy of screw insertion [bib_ref] Efficacy and accuracy of a novel rapid prototyping drill template for cervical..., Lu [/bib_ref] [bib_ref] A novel patient-specific navigational template for cervical pedicle screw placement, Lu [/bib_ref] [bib_ref] Personalised image-based templates for intra-operative guidance, Berry [/bib_ref] [bib_ref] Evaluation of a transpedicular drill guide for pedicle screw placement in the..., Mac-Thiong [/bib_ref] [bib_ref] Rapid prototype patient-specific drill template for cervical pedicle screw placement, Owen [/bib_ref]. In the present study we further improved the non-biocompatible drill templates into biocompatible ones. We first produced a non-biocompatible RP model of cervical vertebra with pin tracts. After inserting the Kirschner wire, we used PMMA, a biocompatible material [bib_ref] Biocompatibility and bone formation with porous modified PMMA in normal and irradiated..., Lye [/bib_ref] [bib_ref] Fatigue and biocompatibility properties of a poly(methyl methacrylate) bone cement with multi-walled..., Ormsby [/bib_ref] [bib_ref] Antibiotic release from an experimental biodegradable bone cement, Gerhart [/bib_ref] , to produce the reverse mold of the RP model and construct a biocompatible drill template. Moreover, we improved the complex design of previous patient-specific drill templates which is beyond the ability of a surgeon because it requires not only medical knowledge but also knowledge of reverse engineering and CAD design [bib_ref] Efficacy and accuracy of a novel rapid prototyping drill template for cervical..., Lu [/bib_ref]. First the drill template we constructed is easy to produce. It does not need surgeons to design the reverse surface modeling with complicated CAD technology. Since . Actual drill with navigation template in cadaveric cervical specimens. Anterior soft tissue was removed from the vertebrae. The drill template was put in place by hand and compressed slightly to the anterior surface of cervical vertebrae. A 2 mm-diameter Kirschner wire was then drilled into the cervical pedicle with the assistance of the drill template. doi:10.1371/journal.pone.0053580.g003 every step is programmed in certain software, an operator can just produce it step by step. Secondly, our drill template is easy to use. During a surgery, our drill template can find its position easily and fit the anterior cervical surface so well that the entry point and direction can be accurately determined. Next the Kirschner wire can be drilled into the cervical pedicle. Consequently it is possible for a surgeon to design customized surgical plan preoperatively. The technique we used can thus eliminate the need for complex equipment and markedly reduce the long duration of the surgery.
To ensure the accuracy of our drill template, we calculated the deviations at the central point of the cervical pedicles to get quantitative data. In this way we could evaluate the accuracy more directly. In the present study, we also calculated the absolute value of the deviations to show the real deviations because the deviations at different directions had positive and negative values, which resulted in a significantly smaller mean. Our study showed that the absolute value deviations in the axial plane (0.8260.75 mm) and in the sagittal plane (1.1060.96 mm) might be within an acceptable range in clinical application.
Gelalis et al. [bib_ref] Accuracy of pedicle screw placement: a systematic review of prospective in vivo..., Gelalis [/bib_ref] found that the screws positioned with freehand technique tended to perforate the cortex medially whereas the screws placed with CT navigation guidance seemed to perforate laterally more often. However, in our study, we found no significant differences between the medial and lateral deviations in the axial plane or between the superior and inferior deviations. Our finding means that there is no specific direction guidance resulting from our biocompatible drill templates.
The simulated insertion of 3.5 mm-diameter screw can show the real implant-pedicle anchorage. Our grade results are very close to the ECD results [bib_ref] In vitro study of accuracy of cervical pedicle screw insertion using an..., Koller [/bib_ref]. Unfortunately, we had 4 screw positions in grade 3. This means the screws penetrated the pedicle cortex from 1/3 to K of the screw cross-section. The deviations are relatively large although it is reported that surgeons judged the pedicle screw position of grade 3 as 'indeterminate' or 'borderline' because it would not cause injury to the VA or nerve root but rather would push the either one away [bib_ref] The vertebral artery and the cervical pedicle: morphometric analysis of a critical..., Tomasino [/bib_ref]. We also found the virtual pin tracts had some deviations from the midline of the pedicle due to the error from our visual observation. This is one of the limitations of our research. In further research, optimal feature of the pin tract should be extracted from the irregular morphology of the cervical pedicle to make sure the pin tract at the midline of the cervical pedicle [bib_ref] Efficacy and accuracy of a novel rapid prototyping drill template for cervical..., Lu [/bib_ref] [bib_ref] A novel patient-specific navigational template for cervical pedicle screw placement, Lu [/bib_ref]. Another limitation of ours is that the shape of drill templates was not optimized for clinical application. [fig_ref] Figure 4: Accuracy evaluation with screw simulation insertion in the axial [/fig_ref] are the border of cervical pedicle in axial plane. The structures inside and outside the blue lines are vertebral canal and vertebral artery, respectively. And the red lines in [fig_ref] Figure 4: Accuracy evaluation with screw simulation insertion in the axial [/fig_ref] are the border of cervical pedicle in sagittal plane. The structures upper and lower the red lines are foramen intervertebrale. A pre-designed 3D screw model (3.5 mm in diameter), which aligned with the pin tract of the 3D model of group2, was imported into Mimics to simulate the screw insertion. The screw positions were graded according to the distance between the screw thread and the border of pedicle cortex. doi:10.1371/journal.pone.0053580.g004 This is a primary study of ours on the biocompatible drill template. Future shape optimizing will follow the easy-to-apply and artistic principles. And pin tracts will be extended as long as possible to further improve the accuracy in assisting ATPS insertion.
# Conclusion
In this in vitro study, the patient-specific drill template we constructed is compatible, easy-to-apply and accurate. Further research should be done to test its clinical applications.
[fig] Figure 1: 3D model of cervical vertebra with virtual pin tracts. 3D model of each vertebra was reconstructed in Mimics software. The 2 cylinders with 2 mm-diameters were then imported and their locations were ensured at the central cervical pedicle by visual observation. With the tool ''subtraction'' under ''Boolean operation'' in the Geomagic studioH software, the 3D model of the cervical vertebrae with bilateral pin tracts was obtained and saved in group 1 in (.stl) file format. doi:10.1371/journal.pone.0053580.g001 maximum being 10 points and the minimum being 2, delineating most accurate screw placement. [/fig]
[fig] Figure 2: Production of biocompatible navigation template with RP model. The 3D model of cervical vertebrae with virtual pin tracts was rapid-prototyped with Z Corporation 3D printer Spectrum Z TM 510. A Kirschner wire was then inserted into the pin tract of the RP model and polymethylmethacrylate (PMMA) was used to construct the drill template. doi:10.1371/journal.pone.0053580.g002 [/fig]
[fig] Figure 3: Actual drill with navigation template in cadaveric cervical specimens. Anterior soft tissue was removed from the vertebrae. The drill template was put in place by hand and compressed slightly to the anterior surface of cervical vertebrae. A 2 mm-diameter Kirschner wire was then drilled into the cervical pedicle with the assistance of the drill template. doi:10.1371/journal.pone.0053580.g003 [/fig]
[fig] Figure 4: Accuracy evaluation with screw simulation insertion in the axial (A) and sagittal (B) planes. The blue lines in [/fig]
[fig] Figure 5: Illustration of grades of ATPS positions in cervical pedicle. Grade 1: Screw positioned at the center of the pedicle. Grade 2: Less than one-third of the screw cross-section (#1.2 mm with a 3.5-mm diameter screw) penetrating the cortex. Grade 3: Between one-third and one-half of the screw cross-section penetrating the cortex (or deviation ,2 mm). Grade 4: More than one-half of the screw cross-section penetrating the cortex (or deviation $2 mm). Grade 5: Deviation equal or greater than the screw diameter. doi:10.1371/journal.pone.0053580.g005 [/fig]
[fig] Figure 6: Absolute deviation values in the axial and sagittal planes. doi:10.1371/journal.pone.0053580.g006 [/fig]
|
Characteristics and outcomes of patients admitted for acute heart failure in a single‐centre study
Aims Acute heart failure represents a medical condition with very high mortality. Accurate risk stratification can help physicians to improve the health care about these patients. The aim of our study was to characterize real-life patients admitted for acute heart failure in a specific region with one tertiary medical centre and to describe risk factors of short-term and long-term mortality.Methods and resultsWe performed a retrospective analysis of patients admitted from January 2017 to December 2017 to Department of cardiology of the tertiary medical centre University Hospital in Hradec Kralove. We identified 385 patients admitted for acute heart failure to the standard care and intensive care unit. The median of age was 74 years (IQR 67.5-80) and 34% of patients were female. Hospital admission was due to de novo heart failure in 222 (57.7%) patients. The most common comorbidities were arterial hypertension (77.7%), dyslipidaemia (67.3%) and coronary artery disease (63.1%). Coronary artery disease (52.7% of cases) and valve disease (28.1% of cases) were the most common aetiologies of heart failure. The all-cause in-hospital mortality was 12.7%, 30-day mortality was 14.6% and 1 year mortality was 34%. Among risk factors of in-hospital mortality, the most significant factors were haemodialysis during the hospitalization [odds ratio (OR) 15.82, 95% confidence interval (CI) 2.96-84.57, P = 0.0008], chronic heart failure (OR 4.27, 95% CI 1.66-11.03, P = 0.001) and STEMI as a precipitating factor of heart failure (OR 4.19, 95% CI 1.23-14.25, P = 0.023). Haemodialysis during the hospitalization (OR 4.28, 95% CI 1.17-15.61, P = 0.025) and the comorbidity depression and anxiety (OR 3.49, 95% CI 1.45-8.39, P = 0.005) were the most significant risk factors of long-term mortality. Conclusions Our study confirms very high mortality rates among patients with acute heart failure underlying poor prognosis of these patients. Comorbidities (peripheral artery disease, atrial fibrillation, chronic heart failure and depression and anxiety), precipitating factors of heart failure (myocardial infarction with ST segment elevation), complications occurring during the hospitalization (acute kidney injury, pulmonary ventilation for respiratory failure and haemodialysis) and the age of patients should be included in the risk stratification of in-hospital, 30 day and 1 year mortality.
# Introduction
Acute heart failure (AHF) represents one of the most common causes of hospital admissions. Despite successful progress in the therapy of chronic heart failure during last years, the morbidity and mortality of patients admitted with AHF remain still high.Moreover, heart failure care's overall cost continues to rise as the incidence and prevalence of heart failure have increased during last years, 2 with hospitalization for heart failure representing a significant part of the resource burden. [bib_ref] Hospitalization costs for patients with acute congestive heart failure in Japan, Kanaoka [/bib_ref] Hospitalization for AHF (de novo or decompensation of chronic heart failure) itself means a negative prognostic factor in the natural history of the disease. Thus, the management of these events warrants careful attention to properly evaluate patients, identify underlying aetiologies and precip-itating factors and consider treatment options with respect to the patient's prognosis and quality of life. The knowledge of risk factors of adverse events may potentially contribute to correct risk stratification, help to guide the treatment and finally improve a clinical outcome. To date, many predictive models with different variables have been proposed to estimate the risk of adverse events in patients with AHF. Their application in standard clinical practice is largely underused, primarily because of the elusive applicability in evaluating an individual patient's risk. [bib_ref] Predicting mortality in patients with acute heart failure: role of risk scores, Passantino [/bib_ref] To provide recent data about AHF patients' characteristics and prognosis, we present results of a retrospective study from the University Hospital in Hradec Kralove. We collected data about patients admitted to the Department of Cardiology during 12 consecutive months. Our hospital represents the only hospital in the city of Hradec Kralove (92 929 inhabitants in 2017) and it also serves as the only tertiary medical centre for Hradec Kralove Region (550 804 inhabitants in 2017) with the opportunity of an urgent cardiac surgery procedure, coronary catheterization or cardiac stimulation procedure. In our study, we focused on patients' profile on admission and differences between mortality groups, and we clarified their contribution to patient's prognosis.
## Study design and methods
The aim of this study was to provide a characterization of hospitalized patients with AHF in a single-centre study and determine risk factors for short-term (in-hospital and 30-day mortality) and long-term mortality (1-year mortality). We performed a retrospective analysis of patients hospitalized for AHF at the Department of Cardiology at the University Hospital in Hradec Kralove during 12 consecutive months from January 2017 to December 2017. Using data collected from Hospital information system and health records of patients we reviewed 3413 hospitalizations and subsequently identified in total 385 patients hospitalized for an AHF during 12 months in the standard care and intensive care unit. It accounted for 422 hospitalizations, but patients were included only once during the year.
We collected data regarding demographic characteristics, comorbidities, medications and basic clinical and imaging methods. The mortality of patients was assessed either directly by the information about the patient's death in his medical record or by the information about health insurance termination during 12 months after the index hospitalization.
Qualified cardiologists performed the diagnosis of clinical conditions and treatment of patients according to the personal clinical judgement and cardiology guidelines concerning heart failure from 2016. 5
# Statistical analysis
Categorical data are presented as numbers of patients and percentages, continuous data are presented as median and interquartile range (IQR) because of non-normal distribution. Categorical data were compared by the Pearson's χ 2 test or Fisher's exact test if at least one cell had an expected value less than 5. Continuous data were compared by non-parametrical Mann-Whitney U test or Kolmogorov-Smirnov test. Multivariate logistic regression was performed to assess the effect of several variables on the all-cause in-hospital, 30 day and 1 year mortality, results are presented as odds ratios (ORs) with corresponding 95% confidence intervals (CIs). Variables with between-group significant differences and with relevant clinical interest were included in multivariate analysis. Variables with a significant association with death are characterized as independent predictors of mortality. A P value less than 0.05 is considered as a statistically significant difference.
Statistical analysis was performed using statistical software NCSS 2019 version 19.0.6.
# Results
## General characterization
During 12 months from January 2017 to December 2017 totally 385 patients with AHF were admitted to our department, from which 131 (34%) were females and the median age at the time of hospitalization was 74 years with 77 years (IQR 72-83) in women and 72 years (IQR 65-77) in men (P < 0.0001). In total 247 (64.2%) patients were older than 70 years. The vast majority of patients (369, 95.8%) were from Hradec Kralove Region and 156 (40.5%) patients were from the city of Hradec Kralove. With respect to the size of the population, with an AHF, we thus hospitalized 0.07% of the population of Hradec Kralove Region and 0.17% of inhabitants of the city of Hradec Kralove. The median length of hospitalization was 10 days and 183 (47.5%) patients spent at least 1 day in the intensive care unit. Most of the patients (241, 71.7%) were discharged home and 75 (22.3%) patients required further hospitalization at another department or hospital. During 12 months, 67 (19.9%) patients surviving the index hospitalization required rehospitalization for an AHF at our hospital. Detailed characteristics are summarized in [fig_ref] Table 1: General characteristics of patients at admission [/fig_ref].
Of 385 patients, 49 (12.7%) died during the hospitalization with the median day of death 5. Following 30 days after admission, the death occurred in 56 (14.6%) patients and 131 (34%) patients died in 12 months following the hospitalization [fig_ref] Table 2: Mortality rates [/fig_ref]. 71.4% of patients (35 of 49) died in the first 10 days of hospitalization. The cardiogenic shock (developed during 24 h after admission) was present in 28 (7.3%) patients with the in-hospital mortality accounting for 60.7% (17 patients).
The AHF was classified as de novo in 222 (57.7%) patients (without a previous diagnosis of heart failure), the rest of the patients had a recognized chronic heart failure before admission and presented as a decompensation of chronic heart failure (163, 42.3%).
The most common aetiologies of heart failure and precipitating factors are listed in . More than one causative and precipitating factor can participate in the development and sudden decompensation of heart failure in many patients. In our cohort of patients, totally 26 aetiologies and 30 triggers were recognized. Coronary artery disease was assessed as a causative factor in more than half of patients. STEMI as an aggravating factor was considered in 42 (10.9%) patients, but an acute coronary syndrome (composed of STEMI, non-STEMI, myocardial infarction in a subacute phase, global ischaemia and myocardial infarction with LBBB) was present in 79 (20.5%) patients.
Pharmacological, diagnostic and other therapeutic interventions during the patient's hospitalization are reported in [fig_ref] Table 4: Treatment and diagnostic procedures during the hospitalization [/fig_ref]. Among complications associated with the hospitalization, at least a mild hepatic injury (defined as the elevation of both ALT and AST or GMT and ALP above the URL) occurred in 164 (42.6%) patients. Acute respiratory insufficiency (defined as resting saturation of O 2 below 90%) at admission or occurring during the hospitalization because of heart failure occurred in 156 (40.5%) patients. Electrolyte disturbances (at least two of hyponatraemia <135 mmol/L, hypokalaemia <3. [bib_ref] ESC guidelines for the diagnosis and treatment of acute and chronic heart..., Ponikowski [/bib_ref] mmol/L, hyperkalaemia >5.2 mmol/L or hypochloraemia <97 mmol/L) occurred in 102 (26.5%) pa-
## Comparison of in-hospital, 30 day and 1 year mortality groups
Differences between mortality groups in specific characteristics are shown in [fig_ref] Table 5: Comparison between mortality groups [/fig_ref]. There are many significant differences in demographic and clinical characteristics between survivors and non-survivors. Those, who died in the hospital, compared with those, who survived, had a significantly higher rate of atherosclerotic diseases such as coronary artery disease and peripheral artery disease. This difference persisted after 1 year (but not in the 30 day mortality group for peripheral artery disease). The median of the length of hospitalization was 11 days and 5 days for survivors and non-survivors, respectively, for both in-hospital and 30 day mortality with a statistically significant difference. Patients who presented with a decompensation of chronic heart failure had significantly higher mortality during the hospital stay, after 30 days and even after 1 year. The deceased also had significantly lower systolic and diastolic blood pressure compared with survivors in all three groups. Although there is not any difference in BMI or BSA in patients who survived or died in the hospital or after 30 days, the difference is statistically significant in 1-year mortality group.
## Predictors of mortality
Results of multivariate logistic regression with clinically relevant variables are performed in [fig_ref] Table 6: Risk factors of short-term and long-term mortality [/fig_ref]. This analysis revealed that a longer hospital stay and higher systolic blood pressure were related to a better in-hospital outcome. In contrast, patients with an acute kidney injury or patients requiring haemodialysis had a worse prognosis. These factors also persisted in a 30 day mortality model. For 1 year mortality, only a higher BMI and higher systolic blood pressure at admission were associated with a better outcome.
# Discussions
In this single-centre retrospective study, we analysed real-life AHF patients admitted to the tertiary medical centre and we yield new and recent information about their demographic characteristics, comorbidities and outcomes. The value and main contribution of our study are in the enrolment of a consecutive and well-defined population of patients in a specific period. We further considered these characteristics for reporting prognostic factors of short-term (in-hospital and 30 day) and long-term (1 year) mortality. Considering the retrospective design, we provide results without large treatment modifications during the relatively short period of data collection.
Most of the data that we know today about the epidemiology of AHF are derived from large-scale registries mainly from the first decade of this century, 1,7-17 selected data are presented in [fig_ref] Table 7: Acute heart failure registries [/fig_ref]. More recent registries from Europe or Asia regions [bib_ref] Demographics, management, and prospective, multicenter Kyoto Congestive Heart Failure (KCHF) registry, Yaku [/bib_ref] [bib_ref] Clinical characteristics and outcome of acute heart failure in Korea: results from..., Lee [/bib_ref] follow the trends in outcomes and demographic characteristics observed in those pivotal studies, but differences in local clinical practice or different clinical thresholds for hospital admission may vary among registries and they may also change over a period of time. Comparison of these results with our study is partially limited because of different sizes of the population, multicentre setting and prospective design. However, many similar or distinct findings from our study can be reported.
The median age of 74 years in our study was slightly higher than that observed in most registries. Markedly higher age was reported only in the prospective registry in Japan. [bib_ref] Demographics, management, and prospective, multicenter Kyoto Congestive Heart Failure (KCHF) registry, Yaku [/bib_ref] We also observed significantly older patients among the deceased compared to survivors in 30 day and 1 year mortality groups. The age (as a continuous variable) was subsequently a mild but significant predictor of 30 day and 1 year mortality.
With the ageing of the population, there is an increase in the prevalence of chronic conditions such as coronary artery disease, diabetes mellitus, arterial hypertension or chronic kid- Similarly, the prevalence of coronary artery disease was higher than in most of these registries. Higher age and comorbidities are generally considered negative prognostic factors in patients with AHF or other critically ill cardiac patients. [bib_ref] Age and shock severity predict mortality in cardiac intensive care unit patients..., Padkins [/bib_ref] Thus, they are important variables in many mortality prediction risk scores, including short-term and long-term outcomes. [bib_ref] Prognostic markers of acute decompensated heart failure: the emerging roles of cardiac..., Cohen-Solal [/bib_ref] [bib_ref] Risk stratification in acute heart failure, Lee [/bib_ref] In the multivariate analysis [fig_ref] Table 6: Risk factors of short-term and long-term mortality [/fig_ref] , many comorbidities were significantly associated with all-cause mortality. Notably, the presence of chronic heart failure is reported with an odds ratio higher than 4 in the analysis of in-hospital and 30-day mortality. This observation is in contrast with previous reports where patients with de novo heart failure may have higher in-hospital mortality. [bib_ref] Heart Failure Association, European Society of Cardiology. EuroHeart Failure Survey II (EHFS..., Nieminen [/bib_ref] Another cardiovascular comorbidity, peripheral artery disease (PAD), which occurred in one third of patients who died during the hospitalization, was significantly associated with in-hospital mortality. The presence of PAD is often described as a marker of generalized atherosclerosis 24 together with coronary artery disease and Characteristics and outcomes of patients admitted for acute heart failure in a single-centre study cerebrovascular disease. In a state of acute hemodynamic decompensation, which occurs in AHF, limited blood flow to target organs may enhance an adverse course of the disease. On the other hand, the optimal modern pharmacotherapy of PAD with the use of antithrombotics, statins or invasive treatment may result in the non-significant impact on 1-year mortality. Although presented in relatively low numbers of patients, depression or anxiety as a comorbidity in patients with AHF represents another independent predictor of 30 day or 1 year mortality. The prevalence of depression and anxiety disorders in patients with heart failure is significantly higher than in the general population, [bib_ref] Depression in heart failure, Rutledge [/bib_ref] accounting for approximately one fifth in heart failure patients. According to the previous reports, depression not only represents a predictor of mortality but is also associated with an increased risk of hospitalization and emergency department visits. [bib_ref] Depression, healthcare utilization, and death in heart failure: a community study, Moraska [/bib_ref] Once heart failure develops, a correct treatment of these comorbidities together with proper adherence to guideline-directed medical therapy, especially for patients with heart failure with reduced ejection fraction, represents a crucial factor affecting the quality of life, a number of rehospitalizations and long-term survival in patients admitted for AHF. [bib_ref] ESC guidelines for the diagnosis and treatment of acute and chronic heart..., Ponikowski [/bib_ref] Above all, haemodialysis during the hospitalization represents a significant predictor of mortality in all three subgroups with the highest odds ratio among all risk factors. In our study, we assessed the factor of haemodialysis, which included patients who required chronic haemodialysis as well as a continuous or intermittent renal replacement therapy as a consequence of acute kidney injury during the hospitalization. The acute kidney injury was also observed as a negative prognostic factor for a short-term prognosis (in-hospital and 30 day mortality). New renal function decline in patients admitted for AHF develops in 18-40% of patients. [bib_ref] Acute kidney injury, heart failure, and health outcomes, Gudsoorkar [/bib_ref] The impact of elevated creatinine at admission or worsening of renal function during hospitalization is a well-defined prognostic factor associated with a poor outcome. [bib_ref] Characteristics, outcomes, and predictors of 1-year mortality in patients hospitalized for acute..., Siirila-Waris [/bib_ref] [bib_ref] Renal impairment, worsening renal function, and outcome in patients with heart failure:..., Damman [/bib_ref] According to the recent meta-analysis, in patients who achieved decongestion before hospital discharge, this decongestion was associated with mitigating the harmful effects of kidney injury in patients with AHF. [bib_ref] Systematic review of the association between worsening renal function and mortality in..., Yamada [/bib_ref] The current evidence also does not support the routine use of ultrafiltration in patients with AHF 31 because of no improvement in mortality compared with standard diuretic treatment. It may be explained by a multiorgan involvement and injury in chronic kidney disease or a chronic volume overload limiting a proper decongestion.
In the multivariate analysis, we also identified several variables associated with reduced mortality in both short and long terms: systolic blood pressure, body mass index and length of hospitalization. Higher systolic blood pressure (SBP) at admission was associated with a better outcome, as confirmed in previous studies. [bib_ref] ESC Heart Failure Long-Term Registry Investigators. Clinical phenotypes and outcome of patients..., Chioncel [/bib_ref] [bib_ref] ADHERE Scientific Advisory Committee, Study Group, and Investigators. Risk stratification for in-hospital..., Fonarow [/bib_ref] Its significance was higher when considering the in-hospital mortality. Patients with heart failure and elevated SBP at admission might have a greater myocardial reserve and lower mortality risk. [bib_ref] The relationship between systolic blood pressure on admission and mortality in older..., Vidán [/bib_ref] A longer follow-up would be beneficial because of the well-known unfavourable outcomes of patients with uncontrolled arterial hypertension with further progression of heart failure. [bib_ref] Blood pressure and heart failure, Oh [/bib_ref] On the other hand, lower SBP prevents the administration of heart failure treatment at discharge, leading to a worse long-term outcome. [bib_ref] In-hospital and long-term mortality for acute heart failure: analysis at the time..., Lombardi [/bib_ref] Furthermore, low SBP may be associated with signs of low cardiac output or peripheral hypoperfusion. Management of these patients often requires the use of inotropes or vasopressors. Their use is commonly associated with a worse prognosis 37 because of the administration in patients in a more severe condition. The percentage of use of inotropes or vasopressors in different heart failure registries can vary according to the availability of these agents, the number of patients with cardiogenic shock or the rate of complications other than cardiogenic shock (e.g. haemorrhagic shock, sepsis or hypovolemia). Compared with our study, the use of inotropic agent (dobutamine) was higher in ALARM-HF [bib_ref] Clinical presentation, management and outcomes in the acute heart failure global survey..., Follath [/bib_ref] or ATTEND 13 (11.3%) but lower in IN-HF 15 (7.7%). The use of vasopressor agent (noradrenaline) was markedly lower in ALARM-HF 14 (4.2%) or ATTEND 13 (4.7%). The use of both was comparable with the Characteristics and outcomes of patients admitted for acute heart failure in a single-centre study AHEAD 16 registry (19% and 10% for noradrenaline and dobutamine). Obesity is a well-known risk factor for cardiovascular morbidity, including coronary artery disease, stroke or heart failure. [bib_ref] Obesity as an independent risk factor for cardiovascular disease: a 26-year follow-up..., Hubert [/bib_ref] On the other hand, among patients with heart failure and obesity, a significant reduction in all-cause and cardiovascular mortality was reported. [bib_ref] Obesity paradox in heart failure: a heavy matter, Nagarajan [/bib_ref] [bib_ref] Body mass index and mortality in heart failure: a meta-analysis, Oreopoulos [/bib_ref] This obesity paradox was also shown in patients with AHF regarding inhospital 41 or long-term mortality. [bib_ref] Association between body mass index and prognosis of patients hospitalized with heart..., Seko [/bib_ref] [bib_ref] GREAT (Global Research on Acute Conditions Team) Network. Body mass index and..., Shah [/bib_ref] Our study confirmed that a higher BMI is associated with a better outcome in 1 year following the hospitalization.
Regarding the all-cause mortality rates (in-hospital, 30 day and 1 year) [fig_ref] Table 2: Mortality rates [/fig_ref] observed in our study, all of them were markedly high. Although the short-term and long-term mortality of patients with AHF remains high, with approximately one-quarter of AHF patients dying in the following year, a recent meta-analysis of 285 AHF studies between 1980 and 2017 reported a decline in 30 day all-cause death that persisted at 1 year. [bib_ref] Temporal trends in mortality and readmission after acute heart failure: a systematic..., Kimmoun [/bib_ref] The authors reported that the 30 day and 1 year all-cause deaths were 7% and 24%. According to the published outcomes from registries [fig_ref] Table 7: Acute heart failure registries [/fig_ref] , the in-hospital mortality varies from 4 to 7.1%, with a higher rate (up to 11-12.7%) in registries with a higher proportion of cardiogenic shock. [bib_ref] Clinical presentation, management and outcomes in the acute heart failure global survey..., Follath [/bib_ref] [bib_ref] Baseline characteristics and hospital mortality in the acute heart failure database (AHEAD)..., Spinar [/bib_ref] The trend in lower short-and long-term mortality in patients with AHF was also observed in the United Kingdom National heart failure audit.Despite this, the in-hospital all-cause mortality was similar to the results observed only in the AHEAD registry. [bib_ref] Baseline characteristics and hospital mortality in the acute heart failure database (AHEAD)..., Spinar [/bib_ref] The in-hospital mortality typically reflects the quality of in-patient care with respect to the severity and stage of the disease, together with proper management of complications during the hospital stay. As mentioned above, the character of our hospital (university hospital and tertiary centre) involves the care about patients with a more severe or complicated course of the disease with a potentially higher risk of adverse outcomes or requiring a higher level of medical cardiology care.
# Limitations
There were several limitations in our study. First, this registry was designed in a retrospective setting. Therefore, the management and follow-up were not standardized and decisions about the treatment, which could further influence the patient's prognosis, were made individually based on the patient's clinical state. Second, we assessed the all-cause mortality as the final endpoint without considering other detailed causes (such as heart failure, sudden death or non-cardiac causes). Third, we included patients admitted only to our Department of cardiology and we did not analyse patients admitted to other internal wards during the year. With respect to this, our cohort of patients might not represent the general population in our region.
# Conclusion
Acute heart failure represents a severe medical condition with significant public consequences. Patients, who present with AHF, are at a very high risk of adverse in-hospital and out-of-hospital outcomes. In our single-centre retrospective study with 385 patients admitted to Department of cardiology of the tertiary medical centre, we observed a very high rate of comorbidities, simultaneous complications and causative and precipitating factors with further influence on patient's prognosis. With regard to the short-time prognosis (in-hospital or 30 day mortality), beside traditional laboratory or clinical markers of unfavourable outcome, our results emphasize further consideration of comorbidities (peripheral artery disease, atrial fibrillation and chronic heart failure), precipitating factors of heart failure (myocardial infarction with ST segment elevation) and complications occurring during the hospitalization (acute kidney injury, pulmonary ventilation for respiratory failure and haemodialysis) in the risk stratification of hospitalized patients with AHF. In the stratification of long-term mortality, higher age, haemodialysis during the hospitalization and a frequently omitted comorbidity depression and anxiety represent a negative prognostic factor. Although there is a global trend in the improvement of care about patients during the hospitalization and followup, a proper risk evaluation of patients at the time of admission is necessary. It represents a crucial point in a physician's approach. Our study provides recent real-life data about patients' characteristics and outcomes and may contribute to the planning of further clinical trials.
[fig] 3: The -adherence, n (%) 1 (.4) Hyperhydration, n (%) 12 (.1) Uncontrolled arterial hypertension, n (%) 9 (2.) COPD, chronic obstructive pulmonary disease; non-STEMI, myocardial infarction without elevations of ST segment; STEMI, myocardial infarction with elevations of ST segment [/fig]
[table] Table 1: General characteristics of patients at admission [/table]
[table] Table 2: Mortality rates [/table]
[table] Table 4: Treatment and diagnostic procedures during the hospitalization [/table]
[table] Table 5: Comparison between mortality groups [/table]
[table] Table 6: Risk factors of short-term and long-term mortality [/table]
[table] Table 7: Acute heart failure registries [/table]
|
Development of functionalised polyelectrolyte capsules using filamentous Escherichia coli cells
Background: Escherichia coli is one of the best studied microorganisms and finds multiple applications especially as tool in the heterologous production of interesting proteins of other organisms. The heterologous expression of special surface (S-) layer proteins caused the formation of extremely long E. coli cells which leave transparent tubes when they divide into single E. coli cells. Such natural structures are of high value as bio-templates for the development of bio-inorganic composites for many applications. In this study we used genetically modified filamentous Escherichia coli cells as template for the design of polyelectrolyte tubes that can be used as carrier for functional molecules or particles. Diversity of structures of biogenic materials has the potential to be used to construct inorganic or polymeric superior hybrid materials that reflect the form of the bio-template. Such bio-inspired materials are of great interest in diverse scientific fields like Biology, Chemistry and Material Science and can find application for the construction of functional materials or the bio-inspired synthesis of inorganic nanoparticles.Results: Genetically modified filamentous E. coli cells were fixed in 2% glutaraldehyde and coated with alternating six layers of the polyanion polyelectrolyte poly(sodium-4styrenesulfonate) (PSS) and polycation polyelectrolyte poly (allylamine-hydrochloride) (PAH). Afterwards we dissolved the E. coli cells with 1.2% sodium hypochlorite, thus obtaining hollow polyelectrolyte tubes of 0.7 μm in diameter and 5-50 μm in length. For functionalisation the polyelectrolyte tubes were coated with S-layer protein polymers followed by metallisation with Pd(0) particles. These assemblies were analysed with light microscopy, scanning electron microscopy, energy dispersive X-ray spectroscopy and transmission electron microscopy. Conclusion: The thus constructed new material offers possibilities for diverse applications like novel catalysts or metal nanowires for electrical devices. The novelty of this work is the use of filamentous E. coli templates and the use of S-layer proteins in a new material construct.
# Background
Escherichia coli are bacteria which naturally colonise the colon of mammalians. The typical cells of E. coli are rod-shaped with dimensions of 1.1-1.5 μm × 2.0-6.0 μm. In molecular biology E. coli is generally used as a tool to produce proteins of interest of other organisms in a simple and high efficient way. In a previous study we described the formation of filamentous E. coli that are surrounded by tube-like structures consisting of outer membrane and surface (S-) layer proteins [bib_ref] Heterologous expression of the surface-layer-like protein SllB induces the formation of long..., Lederer [/bib_ref].
Although one of the best studied microorganisms only few reports describe the filament formation of Escherichia coli [bib_ref] Contraction of Filaments of Escherichia coli after Disruption of Cell Membrane by..., Koch [/bib_ref] [bib_ref] Alterations of the outer membrane composition in Escherichia coli lacking the histone-like..., Painbeni [/bib_ref] [bib_ref] Role of the rfaG and rfaP Genes in Determining the Lipopolysaccharide Core..., Parker [/bib_ref] [bib_ref] Form und Größe des Kernäquivalentes von Escherichia coli in Abhängigkeit von den..., Preusser [/bib_ref]. The formation of filamentous E. coli cells is in most cases a result of the inhibition of proteins that are naturally involved in bacterial cell division processes [bib_ref] Bacterial cell division and the Z ring, Lutkenhaus [/bib_ref] [bib_ref] Analysis of ftsZ mutations that confer resistance to the cell division inhibitor..., Bi [/bib_ref]. The previously described morphological changes were induced by the heterologous high level expression of the S-layer proteins of the uranium mining waste pile isolate Lysinibacillus sphaericus JG-A12. It was suggested that the expression of the S-layer protein SllB inhibits cell division and induces the secretion of these S-layer proteins to the surface of the E. coli cells. The stability of the filaments is a result of the S-layer proteins in the cell wall. The filaments that have a uniform thickness of 0.8-1 μm and can reach a length of several 100 μm have been discussed as promising bio-template e.g. for the production of catalytic active composites or metal microwires [bib_ref] Heterologous expression of the surface-layer-like protein SllB induces the formation of long..., Lederer [/bib_ref] [bib_ref] E. coli filament formation induced by heterologous S-layer expression, Lederer [/bib_ref].
In previous studies cells of different organisms such as erythrocytes, bacteria and spores have been used as biotemplate for the production of polyelectrolyte capsules [bib_ref] Permeability and conductivity of red blood cell templated polyelectrolyte capsules coated with..., Georgieva [/bib_ref] [bib_ref] Layer-by-layer nanoencapsulation of microbes: Controlled cell surface modification and investigation of substrate..., Franz [/bib_ref] [bib_ref] Encapsulation of bacterial spores in nanoorganized polyelectrolyte shells, Balkundi [/bib_ref]. Several studies describe the polyelectrolyte encapsulation and surface modification of living microbial and human cells. Protecting effects of these modifications against phagocytosis, increasing pH values or ultra violet radiation were analysed [bib_ref] Cyborg cells: functionalisation of living cells with polymers and nanomaterials, Fakhrullin [/bib_ref] [bib_ref] Face-lifting" and "make-up" for microorganisms: layer-by-layer polyelectrolyte nanocoating, Fakhrullin [/bib_ref]. The stepwise polyelectrolyte adsorption to different materials such as cells or polymer particles is a useful way to create polymer multilayer films with defined chemical and physical properties. Decher and co-workers proposed this technique originally for the combination of linear polycations and polyanions [bib_ref] Buildup of ultrathin multilayer films by a self-assembly process: III. Consecutively alternating..., Decher [/bib_ref] [bib_ref] Fuzzy Nanoassemblies: Toward Layered Polymeric Multicomposites, Decher [/bib_ref]. The combination of multilayer systems with proteins was described later [bib_ref] Assembly of Multicomponent Protein Films by Means of Electrostatic Layer-by-Layer Adsorption, Lvov [/bib_ref]. The starting material for this method is a solid substrate with a negatively charged planar surface. The formation of the first polyelectrolyte layer is started by addition and adsorption of cationic polyelectrolytes to the substrates. The adsorption is carried out at relatively high polyelectrolyte concentrations. A number of ionic groups remain exposed to the interface towards the solution that affects the effectively reserved surface charge. Substrate rinsing in pure water is followed by incubation of the substrate in an anionic polyelectrolyte solution. Multilayer assemblies are obtained by repeating these steps. Additionally, organic molecules and biocomponents such as proteins, particles, bio-polymers and surfactants can be incorporated in these films, thus realising a multi-functionalisation of these layers [bib_ref] Sequential Actions of Glucose Oxidase and Peroxidase in Molecular Films Assembled by..., Onda [/bib_ref].
In the present study we designed bio-based polyelectrolyte capsules by using filamentous E. coli as bio-template for the assembly of polyelectrolytes. The capsules were bio-functionalised by coating with bacterial S-layer proteins. S-layers are composed of two-dimensional, regularly arranged proteins or glycoproteins [bib_ref] Crystalline surface layers on bacteria, Sleytr [/bib_ref] [bib_ref] Three-dimensional structure of the tetragonal surface layer of Sporosarcina ureae, Engelhardt [/bib_ref] [bib_ref] Structural Research on Surface Layers: A Focus on Stability, Surface Layer Homology..., Engelhardt [/bib_ref] , which are the outer component of the cell wall of many bacteria and are a universal attribute of all archaea [bib_ref] Crystalline Surface Layers in Procaryotes, Sleytr [/bib_ref] [bib_ref] Bacterial S-layers, Sleytr [/bib_ref]. These proteins hold the ability to self-assemble into 2D arrays [bib_ref] Molecular biology of S-layers, Bahl [/bib_ref] [bib_ref] S-Layer Proteins, Sára [/bib_ref] [bib_ref] S-layers of Bacillus species, Sidhu [/bib_ref] and were found generally as mixture of monomer and polymer protein. Special characteristic of S-layers is the formation of regular arranged pores of identical size. S-layer proteins fulfil several functions like working as molecular sieve [bib_ref] Molecular Sieving through S Layers of Bacillus stearothermophilus Strains, Sára [/bib_ref] and binding of toxic heavy metal ions [bib_ref] Loss of Virulence During Culture of Aeromonas salmonicida at High Temperature, Ishiguro [/bib_ref] [bib_ref] Characterization of the surface protein layers of the mosquito-pathogenic strains of Bacillus..., Lewis [/bib_ref] [bib_ref] A general strategy for identification of S-layer genes in the Bacillus cereus..., Mesnage [/bib_ref] [bib_ref] Complexation of Uranium by Cells and S-Layer Sheets of Bacillus sphaericus JG-A12, Merroun [/bib_ref]. The applications of S-layers is multifaceted and include the usage as ultrafiltration membranes [bib_ref] Dynamics in Oxygen-Induced Changes in S-Layer Protein Synthesis from Bacillus stearothermophilus PV72..., Sára [/bib_ref] , drug microcontainers [bib_ref] S-layer stabilized lipid membranes, Schuster [/bib_ref] , filter materials [bib_ref] Biosorption of Uranium and Copper by Biocers, Raff [/bib_ref] or patterning structures in nanotechnology [bib_ref] Selenska-Pobell S: Secondary Structure and Pd(II) Coordination in S-Layer Proteins from Bacillus..., Fahmy [/bib_ref].
In the present study, bio-functionalised polyelectrolyte tubes were used as template for the bio-inspired synthesis of nanoparticular palladium as an example for hybrid material preparation. The outstanding effectiveness of the palladium nanoparticles as catalyst has received particular attention to this metal [bib_ref] Novel supported Pd hydrogenation bionanocatalyst for hybrid homogeneous/ heterogeneous catalysis, Creamer [/bib_ref]. Wahl and others previously described the formation of biogenic palladium nanoparticles in pores of S-layer proteins at the surface of Lysinibacillus sphaericus JG-A12 [bib_ref] Electron-Beam Induced Formation of Highly Ordered Palladium and Platinum Nanoparticle Arrays on..., Wahl [/bib_ref]. Using the hydrogenation of itaconic acid Creamer and others demonstrated the superior catalytic activity of these materials [bib_ref] Novel supported Pd hydrogenation bionanocatalyst for hybrid homogeneous/ heterogeneous catalysis, Creamer [/bib_ref]. Nanoparticles are very attractive for the development of new materials since their properties usually differ significantly from those of the bulk material. In particular, their physical behaviour can be drastically changed and the catalytic activity can be significantly enhanced due to the altered volume/surface ratio. The development of cluster-assembled materials with discrete, size-selected nanoparticles is of great interest to enable the fine-tuning of the properties of the nanoparticles. Especially the design of bio-nanohybrid materials by the combination of bio-molecules with nanoparticles is an emerging topic at the overlaps of biology, material sciences, and nanotechnology. Previous studies describe the design of such bio-nanohybrid materials like the assembling of colloidal gold nanoparticles to the surface of growing fungi [bib_ref] Nutrition-Driven Assembly of Colloidal Nanoparticles: Growing Fungi Assemble Gold Nanoparticles as Microwires, Sugunan [/bib_ref] , the coating of polyelectrolyte encapsulated E. coli with gold and silver nanoparticles [bib_ref] Layer-by-layer coating of bacteria with noble metal nanoparticles for surface-enhanced Raman scattering, Kahraman [/bib_ref] or the functionalisation of bacterial cells using magnetic nanoparticles [bib_ref] Functionalization of whole-cell bacterial reporters with magnetic nanoparticles, Zhang [/bib_ref].
In our work we demonstrate the potential of the use of the S-layer induced filamentous cell structures for the construction of functional conductive metallic wires that potentially can be used for electronic devices or as new catalysts. The possibility to combine such inorganic structures with biological functions opens up new perspectives for multifunctional hybrid materials.
# Results
## Preparation of polyelectrolyte capsules
Coating of filamentous E. coli cells with the polyelectrolytes PSS and PAH followed by treatment with deproteinising NaOCl solution resulted in the formation of filamentous tubes . Approximately one coated capsules per image section still contained bacteria .
These tubes were in average 0.6-0.7 μm in diameter and 5-50 μm in length. Polyelectrolyte capsules showed marginal agglomeration and were stable for several days in deionised water at 4°C. The tubes were transparent and the presence of polyelectrolyte layers is indicated by higher contrast and more acute borders. For the development of the hollow polyelectrolyte capsules different parameters were tested. Especially the fixation of the E. coli filaments with glutaraldehyde in combination with the use of polycationic solution as first polyelectrolyte induced an irreversible agglomeration of the cells. In contrast, suspensions with well separated capsules were obtained when using a polyanionic solution as starting polyelectrolyte. For capsule preparation we tried different combinations of polyanions and polycations. The alternating coating of the cells with the polyanion PSS (poly(sodium-4styrenesulfonate) and the polycation PEI (Poly(ethylenimine)) caused severe cell agglomerations. Such agglomerations were avoided when using a combination of PSS as polyanion and PAH (poly(allylamine hydrochloride)) as polycation. In addition, agglomerations were prevented by thoroughly washing of the samples with 100 mM NaCl after each coating step and the pre-solution of the cells in 100 mM NaCl before each coating step.
## Coating of polyelectrolyte capsules with s-layer proteins
The polyelectrolyte capsules were successfully coated with S-layer using a protein polymer solution as shown in [fig_ref] Figure 2: Fluorescence microscopic images of S-layer coated polyelectrolyte capsules [/fig_ref]. Approximately 80 μg per millilitre S-layer polymer protein adsorb to the polyelectrolyte capsules and potentially the S-layer sheets form a monolayer at the surface. However, the degree of surface coating with S-layer polymer proteins is not known. To visualise the protein layers on the polyelectrolyte tube surface the proteins were coupled with the fluorescence dye HiLyte Fluor ™ 488 and unbound fluorescence dye was removed prior to coating. Fluorescence microscopic images [fig_ref] Figure 2: Fluorescence microscopic images of S-layer coated polyelectrolyte capsules [/fig_ref] present partial uniformly coated hollow capsule surfaces. S-layer polymer proteins labelled with fluorescence dye induce the reversible agglomeration of the coated capsules [fig_ref] Figure 2: Fluorescence microscopic images of S-layer coated polyelectrolyte capsules [/fig_ref] , B). These analyses prove the formerly described high affinity of S-layer proteins to the polyelectrolyte tube surfaces [bib_ref] Recrystallization of Bacterial S-Layers on Flat Polyelectrolyte Surfaces and Hollow Polyelectrolyte Capsules, Toca-Herrera [/bib_ref]. In comparison, fluorescence dyes bound only sparse to polyelectrolyte capsules without protein coating as demonstrated by fluorescence microscopy [fig_ref] Figure 2: Fluorescence microscopic images of S-layer coated polyelectrolyte capsules [/fig_ref] , D). Light microscopic analyses of S-layer polymer coated hollow polyelectrolyte capsules show nearly same proportions but few differences to uncoated Light microscope images of filamentous E. coli cells and polyelectrolyte capsules in phase contrast mode. Image A presents filamentous E. coli cells in the exponential growth phase before polyelectrolyte coating. The polyelectrolyte coated E. coli filaments before NaOCl treatment are shown in image B. Image C presents polyelectrolyte tubes after the treatment with 1.2% NaOCl. The S-layer polymer protein coated polyelectrolyte capsules are shown in image D. Image E shows S-layer polymer protein coated polyelectrolyte tubes with synthesised palladium particles and image F presents polyelectrolyte capsules with synthesised palladium particles without S-layer proteins.
polyelectrolyte tubes . These S-layer coated polyelectrolyte capsules seem to exhibit more compact tube walls than those without proteins , .
## Synthesis of pd(0) particles
After incubation of S-layer coated polyelectrolyte capsules in Pd(II)-solution the tubes turned from colourless to brownish colour, indicating the binding of Pd(II)-complexes. After addition of the reducing agent, the brownish colour changed to black, indicating the formation of Pd(0). S-layer coated polyelectrolyte tubes with synthesised palladium particles are visible in the , [fig_ref] Figure 3: SEM images of filamentous polyelectrolyte capsules [/fig_ref] and , however particles were identified distinct in the [fig_ref] Figure 3: SEM images of filamentous polyelectrolyte capsules [/fig_ref] and . These tubes show uniform dark surfaces, pointing to the presence of Pd (0) and are in average 1-1.3 μm in diameter and 5-50 μm in length . In order to get more information on particle formation, surface appearance but also the interior of the capsules the materials were investigated by SEM, EDX and TEM. These analyses [fig_ref] Figure 3: SEM images of filamentous polyelectrolyte capsules [/fig_ref] , [fig_ref] Table 1: Palladium particle size analysis [/fig_ref]. These analyses indicated a higher number of small palladium particles at the surface of polyelectrolyte capsules without S-layer proteins. However, those polyelectrolyte capsules with additional S-layer protein coating exhibit distinct higher numbers of large Pd particles.
The compositions of polyelectrolyte capsules without additional S-layer proteins with palladium particles [fig_ref] Figure 3: SEM images of filamentous polyelectrolyte capsules [/fig_ref] and of polyelectrolyte capsules with additional S-layer protein with palladium particles [fig_ref] Figure 3: SEM images of filamentous polyelectrolyte capsules [/fig_ref] were analysed by energy dispersive X-ray spectroscopy (EDX). Among other things like carbon, nitrogen and oxygen the analyses verified the presence of palladium. As reference, capsules without protein coating were used as template for the synthesis of Pd(0) particles. These materials are presented in , [fig_ref] Figure 3: SEM images of filamentous polyelectrolyte capsules [/fig_ref] and . These tubes are 0.8-1.1 μm in diameter and 5-50 μm in length. Similar to the protein coated samples larger particles are visible at the surface of the uncoated capsules. However, in contrast to the protein samples, these particles are formed to a significantly less amount and showed a lower density.
# Discussion
Previous studies discovered the formation of unusual long Escherichia coli cell filaments induced by S-layer protein expression [bib_ref] Heterologous expression of the surface-layer-like protein SllB induces the formation of long..., Lederer [/bib_ref] [bib_ref] E. coli filament formation induced by heterologous S-layer expression, Lederer [/bib_ref]. Such biological structures provide a promising matrix for technical applications such as the development of microcontainers or hollow metallic microwires. Especially gram-negative cells like E. coli are attractive for such applications. They possess a comparatively fragile cell wall that can be easily destroyed. E. coli can be easily cultivated giving a high yield of biomass and can be used for multifaceted applications. In the present study we used the cells for the synthesis of polyelectrolyte hollow capsules and investigated the possibility to use them as substrate for the functionalisation with proteins and metal nanoparticles.
The development of polyelectrolyte capsules was investigated by several groups using different kinds of templates postulating that those capsules are ideal candidates for applications in the areas of drug delivery, sensing and catalysis. Sukhorukov and co-workers coated polystyrene and melamine formaldehyde latex particles with polyelectrolyte multilayers and dissolved the core, while Yu and others described the production of polymeric capsules with pre-loaded proteins based on mesoporous silica capsules which were finally removed [bib_ref] Nanoassembly of biocompatible microcapsules for urease encapsulation and their use as biomimetic..., Yu [/bib_ref]. The encapsulation of spores was described by Balkundi and co-workers aiming the development of environmental compatible materials for agriculture [bib_ref] Encapsulation of bacterial spores in nanoorganized polyelectrolyte shells, Balkundi [/bib_ref]. Franz and others investigated the encapsulation of microbes with different polyelectrolyte combinations and the following substrate uptake properties of enclosed bacteria [bib_ref] Layer-by-layer nanoencapsulation of microbes: Controlled cell surface modification and investigation of substrate..., Franz [/bib_ref]. These studies used the benefit of layer-bylayer technique which enables the variation of thickness, composition, and function of these assemblies by tuning the layer number, the species deposited, and the assembly conditions.
The present study describes the development of polyelectrolyte hollow tubes based on S-layer expressing E. coli cells which were fixed in glutaraldehyde and combined with the polyelectrolytes PSS (sodium poly(styrene sulfonate)) and PAH (poly(allylamine hydrochloride)) and a final NaOCl treatment. Other papers that used cells as template described the combination of negatively charged surfaces which were afterwards coated with the polycation followed by washing steps and a polyanion [bib_ref] Buildup of ultrathin multilayer films by a self-assembly process: III. Consecutively alternating..., Decher [/bib_ref]. In contrast, the assembly of polyelectrolyte layers on E. coli filaments necessitated the starting with a polyanion to a probably negatively charged cell surface [bib_ref] Polyelectrolyte multilayer capsules templated on biological cells: core oxidation influences layer chemistry, Moya [/bib_ref]. The combination of the glutaraldehyde fixed cells with polycationic solution induced an irreversible agglomeration of the cells. In comparison they stayed in suspension well separated when they were initially incubated with a polyanionic solution. Responsible for cell agglomerations which were observed after polycation incubation are potentially single positive groups at the mainly negative charged bacterial cell surface. Potentially, in the presence of polycations very high attractive forces operate between these cells which lead to agglomerations. However, negative polymers will saturate the few positive groups at the bacterial cell surface resulting in a very consistent charge distribution. So, the negative polymer works potentially as solubiliser.
Moya et al. described that treatment of polyelectrolyte encapsulated cells with NaOCl solution changed the chemical composition of the capsules dramatically. They observed the oxidation of the amino groups of polyallylamine to nitriles, nitroso-, nitro-, azo-and carbonyl groups and the disappearance of positive charges. Coevally the polymer chains were cross-linked with covalent bonds. Finally, the amount of PSS is strongly reduced to 10% of the original value. Moya et al. justified the stability of these capsules with the combination of cross-linking and hydrophobic interaction [bib_ref] Polyelectrolyte multilayer capsules templated on biological cells: core oxidation influences layer chemistry, Moya [/bib_ref]. In our work, the use of the polyelectrolytes PSS and PAH in combination with sodium hypochlorite resulted evidently in the formation of uniformly coated stabile filamentous hollow capsules. However, round about 1% of the coated cells remain intact during NaOCl treatment. This observation leads to the assumption that these cells were not treated efficiently with NaOCl, perhaps because of their localisation in the lid of the reaction tube during incubation.
The surface coating of these tubes with surface layer polymer proteins aimed the synthesis of two dimensional crystal lattice which hold regular ordered nanopores with uniform bonding characteristics. Toca-Herrera and co-workers described the recrystallisation of S-layer proteins on polyelectrolyte surfaces and demonstrated by AFM that the combination of a final PAH layer with TEM images of filamentous polyelectrolyte capsules. Polyelectrolyte capsules without additional S-layer polymer protein with palladium particles are shown in A and with additional S-layer polymer protein with palladium particles are shown in B. The darker tube shows a filament that contains bacterial cells that were not removed during the procedure. TEM image C shows the surface of S-layer polymer protein coated polyelectrolyte tubes with crystalline palladium particles. In D a TEM micrograph of single Pd-particles and an insert of a Fourier transformation analysis of such a particle are shown. The latter indicates by the measured distance of the lattice planes the crystalline nature of these particles. surface layer proteins hinder the recrystallisation of the proteins [bib_ref] Recrystallization of Bacterial S-Layers on Flat Polyelectrolyte Surfaces and Hollow Polyelectrolyte Capsules, Toca-Herrera [/bib_ref]. However, our light microscopic studies indicate that the binding of S-layer polymer proteins to polyelectrolyte capsules is enhanced with PAH as final polyelectrolyte capsule coating. It can be assumed that the constitution of PAH was influenced by sodium hypochloride treatment. Probably the uniform negative charges of the polyelectrolyte surface support the binding of S-layer polymer proteins via electrostatic attractive forces. The complete S-layer coating of the polyelectrolyte capsule surface is not assumed. S-layers were used to bio-functionalise the new-designed polyelectrolyte tubes.
In previous works self-assembling of bio-molecules to capsules or filaments has been reported several times and methods to functionalise these structures have been established. Mbindyo and co-workers reported the DNA-directed assembly of gold nanowires 0.2 μm in diameter and up to 6 μm in length [bib_ref] DNA-Directed Assembly of Gold Nanowires on Complementary Surfaces, Mbindyo [/bib_ref] , while the recognition capabilities of DNA, which induced the targeted attachment of functional wires were described by Braun and others [bib_ref] DNA-templated assembly and electrode attachment of a conducting silver wire, Braun [/bib_ref]. Vauthey and co-workers described the molecular self-assembly of surfactant-like peptides to form nanotubes and nanovesicles. The ability of protein coated peptide tubules to recognise and bind the protein complementary molecules in solution was investigated by Douberly and co-workers [bib_ref] Fabrication of Protein Tubules: Immobilization of Proteins on Peptide Tubules, Douberly [/bib_ref] , while Yang and others analysed microtubules as templates for fabricating metallic nanowires [bib_ref] Electroless metal plating of microtubules: Effect of microtubuleassociated proteins, Yang [/bib_ref]. Sugunan and others describe the formation of microwires of gold nanoparticle coated hyphea of Aspergillus strains while growing of initial spores in colloidal gold solution [bib_ref] Nutrition-Driven Assembly of Colloidal Nanoparticles: Growing Fungi Assemble Gold Nanoparticles as Microwires, Sugunan [/bib_ref]. The assembly of nanoparticles on filamentous fungi generates microwires with extraordinary length. However, the diameter of the distinct shorter E. coli filament based polyelectrolyte capsules is smaller. The removal of the inner organic material of the E. coli filaments is much easier than the one of gold nanoparticle encapsulated filamentous fungi. The final synthesis of palladium nanoparticles in the pores of S-layer polymer proteins seems to produce distinct smaller nanoparticles than the glutamate stabilized gold nanoparticles. Kahraman and others studied the polyelectrolyte encapsulation of E. coli and Staphylococcus cohnii with additional gold and silver nanoparticles [bib_ref] Layer-by-layer coating of bacteria with noble metal nanoparticles for surface-enhanced Raman scattering, Kahraman [/bib_ref] while Zhang and co-workers analysed the functionalisation of bacterial cell walls with magnetic nanoparticles [bib_ref] Functionalization of whole-cell bacterial reporters with magnetic nanoparticles, Zhang [/bib_ref]. Fakhrullin and co-workers gave in their review a detailed overview over the studies which focus the functionalisation of living cells with polymers and nanoparticles [bib_ref] Cyborg cells: functionalisation of living cells with polymers and nanomaterials, Fakhrullin [/bib_ref].
The application of surface layer proteins as template for the synthesis of nanoparticles is a well established method [bib_ref] Selenska-Pobell S: Secondary Structure and Pd(II) Coordination in S-Layer Proteins from Bacillus..., Fahmy [/bib_ref] [bib_ref] Electron-Beam Induced Formation of Highly Ordered Palladium and Platinum Nanoparticle Arrays on..., Wahl [/bib_ref] [bib_ref] Manufacturing and characterization of Pd nanoparticles formed on immobilized bacterial cells, Pollmann [/bib_ref]. S-layers are an interesting starting material for the synthesis of bio-inorganic composite materials that are promising for various applications, e.g. catalysts [bib_ref] Novel supported Pd hydrogenation bionanocatalyst for hybrid homogeneous/ heterogeneous catalysis, Creamer [/bib_ref]. The proteins that are decorated with catalytic active nanoparticles can be fixed on carrier materials. The S-layer properties (amino acid composition, array symmetry and pore size) determine the nanoparticle properties like size and distribution. In previous work EXAFS and ATR-FT-IR analyses proved that carboxyl groups of the proteins are involved in the binding of the Pd(II) complexes [bib_ref] Selenska-Pobell S: Secondary Structure and Pd(II) Coordination in S-Layer Proteins from Bacillus..., Fahmy [/bib_ref] [bib_ref] Manufacturing and characterization of Pd nanoparticles formed on immobilized bacterial cells, Pollmann [/bib_ref]. In the present study we used S-layer coated polyelectrolyte filaments as carrier material for synthesis of Pd(0) particles. The immobilised S-layer proteins are able to bind Pd(II) complexes, thus enabling the synthesis of palladium particles by the addition of a reducing agent.
The newly designed bio-functionalised polyelectrolyte tubes that are described in this paper are unique due to its starting material. Specific regulations of template organism, temperature and amount of activator induce the formation of Escherichia coli filaments with defined diameter and cell wall stability. The template bacteria provide up to several 100 μm long structures with defined 0.8-1 μm in diameter which were encapsulated by layer-by-layer method with polyelectrolytes. After removing the bacterial core these polyelectrolyte hollow capsules can be bio-functionalised with S-layer polymer proteins which support the synthesis of metal nanoparticles in the protein pores. In conclusion, these filamentous polyelectrolyte tubes may provide an interesting matrix for the development of microcontainers and metal microwires with possibly novel physical and chemical properties. In combination with S-layer coupled palladium nanoparticles these materials could find application as novel catalysts or in the preparation of conductive metal microwires in electrical devices. Such developments are part of future work.
# Conclusion
In the present study we describe for the first time the use of filamentous E. coli as template for the assembly of polyelectrolytes. A method was developed that enables the synthesis of polyelectrolyte wires with a uniform diameter. These hollow fibres can be functionalised with proteins as well as with Pd(0) particles. These features make the filaments promising for future developments such as novel catalysts or metal nanowires for electrical devices.
# Methods
## Strains and culture conditions
Escherichia coli BL21(DE3), that express the S-layer protein SllB of the uranium mining waste pile isolate Lysinibacillus sphaericus JG-A12, were routinely grown in LB medium supplemented with kanamycin (35 μg ml -1 ) and 100 μM IPTG at room temperature for approximately 24 hours.
## Preparation of polyelectrolyte capsules
Escherichia coli cells were harvested in the stationary phase at OD 600 =2 and a pellet of at least 100 mg biomass was washed twice with 1 ml of 100 mM NaCl solution pH 7. The cells were fixed in the following step in 1 ml of 2% glutaraldehyde (Serva, Heidelberg, Germany) at room temperature for one hour as described elsewhere [bib_ref] Polyelectrolyte multilayer capsules templated on biological cells: core oxidation influences layer chemistry, Moya [/bib_ref] [bib_ref] The action of lower aldehydes on the human erythrocyte, Heard [/bib_ref]. Afterwards the fixed cells were washed twice in 1 ml of 100 mM NaCl solution pH 7. The polyelectrolytes (PE) sodium poly(styrene sulfonate) (PSS) (Sigma, Aldrich, St. Louis, MO) of M w 70,000 Da and poly(allylamine hydrochloride) (PAH) (Sigma) of M w~5 6,000 Da were dissolved to a concentration of 1 mg ml -1 in 100 mM NaCl (Roth, Karlsruhe, Germany) solution pH 7. The final pH value of PSS solution was pH 6 and the pH value of PAH solution was pH 5. Six layers of freshly prepared PSS and PAH solutions were adsorbed onto the cells in the presence of 100 mM NaCl beginning with the polyanion. Each coating step lasted 10 minutes and was followed by four washing steps with 100 mM NaCl. After each step the cell pellet was concentrated by centrifugation at 12,000 g at room temperature for 3-5 min. To avoid cell agglomeration the cell pellet was resuspended in 150 μl of 100 mM NaCl before addition of polyelectrolyte solution. In the following deproteinisation step with 1.2% NaOCl (Sigma) [bib_ref] Permeability and conductivity of red blood cell templated polyelectrolyte capsules coated with..., Georgieva [/bib_ref] the cells were destroyed, while the hollow polyelectrolyte capsules remained. Capsules were washed four times in 100 mM NaCl to remove residual NaOCl.
## Linking of fluorescence dye to s-layer proteins
The fluorescence dye HiLyte Fluor™ 488 amine (Mobi-Tec, Göttingen, Germany) was chosen for labelling of the S-layer proteins. For coupling reactions the S-layer proteins were dissolved in 50 mM MES-buffer (pH 5.6) and linked with the help of 200 μM cross-linker EDC (1-Ethyl-3-(3-dimethylaminopropyl)carbodiimid) (Sigma) to HiLyte Fluor™ 488 amine. The reaction took two hours. Afterwards uncoupled fluorescence dyes were removed by centrifugation and fluorescence labelled Slayer protein polymers were washed with buffer.
## Coating of polyelectrolyte capsules with surface layer proteins
The natural S-layer proteins of Lysinibacillus sphaericus JG-A12 were purified as described elsewhere [bib_ref] Biosorption of Uranium and Copper by Biocers, Raff [/bib_ref]. Briefly, the S-layer protein expressing cells were grown in NB medium at 30°C, harvested by centrifugation and washed. Bacterial flagella were removed by treating the cells with the rotating-blade blender IKA T8 (IKA Labortechnik, Staufen, Germany) and following centrifugation steps. The cells were disintegrated by using the high-shear fluid processor at a pressure of 960 bar (M-110S Microfluidizer processor, Microfluidics, Newton, MA, USA). The cell fragments were washed, treated with Triton X-100 and washed again. Peptidoglycan was lysed by treatment with lysozyme. The S-layer containing fraction was washed several times, mixed with guanidine hydrochloride and non-protein compounds were removed by centrifugation. In order to remove guanidine hydrochloride the S-layer containing supernatant was dialysed several times against 1.5 mM Tris and 10 mM CaCl 2 , pH 8. Recrystallised S-layer proteins were collected by centrifugation and stored at 4°C for later applications. The isolated S-layer proteins are of high purity and were found in balance as monomer and polymer proteins (U. Weinert, pers. communication).
The polyelectrolyte tubes were washed and resuspended in 1 ml of 10 mM CaCl 2 solution. Subsequently, 200 μg ml -1 of native or fluorescence labelled S-layer polymers were added to the polyelectrolyte capsule solution and bound to the surface of the polyelectrolyte tubes. The solution was stirred at room temperature for 20-24 hours. Afterwards the solution was concentrated by centrifugation at 12,000 g at room temperature for 3-5 min and washed twice with distilled water. The supernatants were removed.
## Synthesis of pd (0) particles
Pd(0) particles were synthesised as described elsewhere [bib_ref] Selenska-Pobell S: Secondary Structure and Pd(II) Coordination in S-Layer Proteins from Bacillus..., Fahmy [/bib_ref]. Briefly, the S-layer polyelectrolyte tubes were concentrated by centrifugation. The Pd-solution was prepared 24 hours before usage. For this 2 mM Na 2 PdCl 4 (Sigma) were dissolved in water and incubated overnight in the dark. The coating was started by addition of 10 ml Na 2 PdCl 4 solution to the polyelectrolyte capsules. After 4 hours of incubation at room temperature under shaking in the darkness the tubes were washed twice in deionised water. Afterwards the bound Pd(II) was reduced by the addition of 30 μl of 100 mM dimethylamine-borane (Merck, Darmstadt, Germany) [bib_ref] Selenska-Pobell S: Secondary Structure and Pd(II) Coordination in S-Layer Proteins from Bacillus..., Fahmy [/bib_ref]. The directly observed sample colour change indicated the successful reaction. The sample was centrifuged and the pellet was washed twice and finally stored in deionised water.
Characterisation of polyelectrolyte capsules and Pd (0) particles Surface texture, height and uniformity of the polyelectrolyte coated capsules and the Pd(0) particle structures were analysed by light microscopy, scanning electron microscopy (SEM), transmission electron microscopy (TEM) and energy dispersive X-ray spectroscopy (EDX). Light microscopic images of cells and polyelectrolyte capsules were taken with the Olympus BX61 microscope (Olympus, Hamburg, Deutschland) in phase contrast mode. Fluorescence microscope images were taken with the filters U-MSWG (480-570 nm) and U-MNIBA (470-525 nm). Scanning electron microscopy (SEM) images of polyelectrolyte capsules and Pd(0) particles were obtained using the crossbeam workstation NVision 40 (Carl Zeiss SMT, Germany) at 5 keV. The morphology and chemical composition of the polyelectrolyte capsules and the Pd(0) particles was evaluated using a Titan 80-300 transmission electron microscope (FEI, Eindhoven, The Netherlands) at 300 keV. Energy dispersive X-ray spectroscopy (EDX) analyses were obtained after activation scanning electron microscopy (SEM) with the EDX system Quantax 400 (Bruker AXS, Karlsruhe, Germany) with the Si-drift detector XFlash 123 eV.
Samples for scanning electron microscopy investigations were applied to RCA purified Si wafers, each [bib_ref] Cleaning solutions based on hydrogen peroxide for use in silicon semiconductor technology, Kern [/bib_ref]. Samples were dried for about 24 hours at room temperature and analysed later with the scanning electron microscope. For transmission electron microscopy the samples were dried for about 24 hours at room temperature on carbon-coated copper grids.
[fig] Figure 2: Fluorescence microscopic images of S-layer coated polyelectrolyte capsules. Images of filamentous polyelectrolyte capsules with HiLyte Fluor™ 488 amine linked S-layer polymer proteins in phase contrast mode (A), and excited by light in the 480-570 nm wavelength range using the filter U-MSWG (B). Polyelectrolyte capsules without S-layer proteins treated with HiLyte Fluor™ 488 amine in phase contrast mode are shown in image (C) and excited by light in the 480-570 nm wavelength range using the filter U-MSWG are shown in image (D). [/fig]
[fig] Figure 3: SEM images of filamentous polyelectrolyte capsules. Polyelectrolyte capsules without additional S-layer polymer protein with palladium particles are shown in (A) and with additional S-layer polymer protein with palladium particles are shown in (C). The images B and D present EDX analyses of parts of polyelectrolyte capsules with palladium particles without additional S-layer proteins (B) and with S-layer proteins (D). [/fig]
[table] Table 1: Palladium particle size analysis [/table]
|
Determining the prevalence of palliative needs and exploring screening accuracy of depression and anxiety items of the integrated palliative care outcome scale – a multi-centre study
Background: patients with palliative needs often experience high symptom burden which causes suffering to themselves and their families. Depression and psychological distress should not be considered a "normal event" in advanced disease patients and should be screened, diagnosed, acted on and followed-up. Psychological distress has been associated with greater physical symptom severity, suffering, and mortality in cancer patients. A holistic, but short measure should be used for physical and non-physical needs assessment. The Integrated Palliative care Outcome Scale is one such measure. This work aims to determine palliative needs of patients and explore screening accuracy of two items pertaining to psychological needs.
## (continued from previous page)
Conclusions: main palliative needs were psychological, family related and spiritual. This suggests that clinical teams may better manage physical issues and there is room for improvement regarding non-physical needs. Using the Integrated Palliative care Outcome Scale systematically could aid clinical teams screening patients for distressing needs and track their progress in assisting patients and families with those issues.
Keywords: Palliative care, Screening accuracy, Depression, Anxiety, Patient centred outcome measures, Outcome measurement, End of life care, Suffering Background Palliative care is a holistic approach of care, which can be integrated early in the disease trajectory, alongside active, curative treatment, [bib_ref] Improving patient outcomes through palliative care integration in other specialised health services:..., Oliver [/bib_ref] and aims to alleviate physical and non-physical symptoms of patients and their families [bib_ref] Between hope and acceptance: the medicalisation of dying, Clark [/bib_ref] [bib_ref] specialist palliative care is more than drugs: a retrospective study of ILD..., Bajwah [/bib_ref] [bib_ref] Quality of life in palliative care cancer patients: a literature review, Jocham [/bib_ref] [bib_ref] The concept of quality of life of dying persons in the context..., Stewart [/bib_ref] [bib_ref] Patient satisfaction and patient-centered care: necessary but not equal, Kupfer [/bib_ref]. Though physical symptoms may be more easily identified by healthcare professionals, patients and carers, non-physical symptoms can equally disrupt the patient and family's quality of life and cause suffering [bib_ref] Characteristics and outcomes of psychology referrals in palliative care department, Bruera [/bib_ref]. Regarding effectiveness of psychological interventions to improve quality of life in people with long term conditions, results from a rapid systematic review show that there was a significant improvement in at least one quality of life outcome post-intervention and maintained at follow-up [bib_ref] Effectiveness of psychological interventions to improve quality of life in people with..., Anderson [/bib_ref]. In another study, Ann-Yi S and colleagues report that 24% of palliative care in-patients and 19% of palliative care outpatients in a major cancer centre benefited from psychology services [bib_ref] Characteristics and outcomes of psychology referrals in palliative care department, Bruera [/bib_ref]. Also, psychological distress has been associated with greater physical symptom severity, suffering, and mortality in cancer patients [bib_ref] Psychological distress and cancer mortality, Hamer [/bib_ref]. Thus, these symptoms and needs should be properly assessed with validated outcome measures, and, the intervention(s) to solve or manage them should be selected accordingly [bib_ref] EAPC white paper on outcome measurement in palliative care: improving practice, attaining..., Bausewein [/bib_ref]. Depression and psychological distress are two examples of such needs, which should not be considered a "normal event" in an advanced disease trajectory [bib_ref] Psychological distress and quality of life of palliative cancer patients and their..., Götze [/bib_ref]. Rather, they should be screened, diagnosed, acted on and followedup by appropriate support services and specialised healthcare professionals, whether by pharmacologic treatment, psychological treatment, psychiatric treatment or a combination of those [bib_ref] Psychological distress and quality of life of palliative cancer patients and their..., Götze [/bib_ref]. Patient centred outcome measures should be the first choice to measure subjective symptoms and needs, given that the patient is the best person to assess how the symptoms bother them. If the patient is cognitively impaired, a proxy version of the measure of choice can be used [bib_ref] on behalf of EUROIMPACT. Implementing patient-reported outcome measures in palliative care clinical..., Antunes [/bib_ref] [bib_ref] Capture, transfer, and feedback of patient centred outcomes data in palliative care..., Etkind [/bib_ref] [bib_ref] Integrated palliative care outcome scale: Protocol Validation for the Portuguese population, Antunes [/bib_ref]. These measures are short, but multidimensional and some items may be used to screen for certain palliative needs, common in this population [bib_ref] Screening for depression in advanced disease: psychometric properties, sensitivity and specificity of..., Antunes [/bib_ref].
The aim of this study is to determine the main palliative care needs of patients being treated in portuguese health care services and to explore screening properties of two items pertaining to psychological needs, using the Integrated Palliative care Outcome Scale (IPOS). We hypothesized that 1) anxiety and depression items will score highest among the non physical symptoms and 2) the area under the curve (AUC) to be > 0.7 for Items 3 (anxiety) and 5 (depression) in relation to the Portuguese Hospital Anxiety and Depression Scale (HADS) [bib_ref] Validation study of a Portuguese version of the hospital anxiety and depression..., Pais-Ribeiro [/bib_ref] Anxiety subscale score and the HADS-Depression subscale total scores, respectively.
# Methods
## Patients and settings
This was a multi-centred observational study. Data were collected in nine portuguese centres spread out from north to south and rural to urban locations to maximise generalizability using convenience sampling. There were seven hospital based palliative care services, one oncology service and one primary care facility (health centre). All patients attending the participant services were screened for eligibility by the participating healthcare professionals. Inclusion criteria were as follows: ≥18 years, mentally fit to give consent judged as such by the participating healthcare professional, diagnosed with an incurable, potentially life-threatening illness, read, write and understand Portuguese. Exclusion criteria were: patient in distress (unable to maintain a conversation during a period of time) with uncontrolled physical or emotional symptoms, and/or cognitively impaired, judged as such by the participating healthcare professional. A standard operating procedures manual was developed and distributed to all centres in the person of the facilitator/champion leading the study locally. The detailed protocol has been published elsewhere [bib_ref] Integrated palliative care outcome scale: Protocol Validation for the Portuguese population, Antunes [/bib_ref].
## Measures
The Portuguese version of the IPOS reported by the patient, previously developed, was used. The original measure has been culturally adapted and validated to European Portuguese. The protocol of the latter study has been published elsewhere [bib_ref] Integrated palliative care outcome scale: Protocol Validation for the Portuguese population, Antunes [/bib_ref] which contains the full questionnaire in the appendix. Next, we present a summary of main procedures and results. Two native Portuguese speaking translators, one clinical and one non-clinical independently created two Portuguese versions. A consensus Portuguese version was developed by two native Portuguese speaking independent reviewers blind to the original IPOS. This consensus version was sent to two other independent native Portuguese speaking translators, also blind to the original English IPOS, who back translated it into English. A second Portuguese consensus version was developed by the same reviewers. Three clinical revisions were performed by one specialist palliative care doctor, one specialist palliative care nurse and one non-clinical researcherall native Portuguese. A final Portuguese version was created. There were grammatical and content differences in the first translation stage, in the items/questions text as well as in the response categories. These were resolved by discussion by both reviewers. There were also differences in the backward translation, namely verb tenses and the use of synonyms rather than the direct translation of words. These were resolved by discussion by the same reviewers. The clinical revisions flagged differences in verb tenses in three items. Those were discussed, and changes were made to create the final version. A Portuguese version of the IPOS was developed. Next, psychometric characteristics were assessed, namely, Internal consistency (excluding open questions, which are free text data) Cronbach's alpha varied between 0.68 and 0.72, reliability between patients and healthcare professionals scores was assessed by intraclass correlation which was higher for mobility (ICC = 0.726 and lowest for practical problems (ICC = 0.088). Regarding construct validity, Items with similar constructs showed convergent validity and items with different constructs showed divergent validity. Spearman's Rho varied between .390 and .631 with p ≤ .000. The measure also displayed good sensitivity to change, as Wilcoxon ranked test showed significant statistically differences between T1 and T2 in three symptoms. IPOS It is a brief, 19-item, multidimensional scale that captures core concerns in palliative care. The first item is an open question on the three main problems or worries the respondent might have had in the past week (results of this item are not reported in the present study given that data are free text); items 2 to 9 are set on a 5 point Likert scale based on descriptors (zeronot at all, 1slightly, 2moderately, 3severely, 4overwhelmingly), item two is a list of 10 of the most common physical symptoms in a palliative population, with the possibility of adding up to three more symptoms which are not present in the list; item 3 pertains to anxiety, item 4 asks about family/friends worry, item 5 is on depression; item 6 is about being at peace; item 7 relates to sharing feelings with significant people; item 8 is about information needs and item 9 concerns practical problems related to their illness. In the patient version (as opposed to the healthcare professional version), the questionnaire has an extra item asking if the respondent filled the questionnaire alone or with help. At the very end, there is a footnote to trigger the patient to talk to their healthcare professional, if they feel they are worried about any of the issues raised by the items in the questionnaire. This feature allows for real time clinical utility of the measure.
The Portuguese Hospital Anxiety and Depression Scale (HADS) is a 14 item screening measure for anxiety and depression states. The two subscales are comprised of 7 items each, scored separately, with descriptive answers based in a 4 point Likert scale. The authors propose a cut-off threshold of 11 for depression and anxiety. The authors conclude that the Portuguese HADS is a reliable and valid measure for assessing anxiety and depression in different medical settings and disease populations [bib_ref] Validation study of a Portuguese version of the hospital anxiety and depression..., Pais-Ribeiro [/bib_ref].
# Analysis
After checking data quality and performing Little's Missing Completely At Random (MCAR) test to check if data were missing at random, descriptive statistics were used to examine the distribution of demographic and clinical variables of interest. To determine accuracy of the items under study, we previously tested all possible cut-offs (results not presented) and decided to use the most appropriate, a cut-off score of 2/3. Then we compared all 5 psychological, emotional and spiritual needs items (IPOS items 3 to 7) against the Portuguese HADS. Receiver operating characteristic curves (ROC) were used to determine the 2 items displaying the highest AUC and assess discriminant properties, namely, sensitivity and specificity of the cut-off 2/3 for both items against cut-off 10/ 11 of the HADS Anxiety subscale and the HADS Depression subscale respectively. Positive (PPV) and negative predictive values (NPV), false negative (FNR) and false positive rates (FPR) and positive and negative likelihood ratios weighted by prevalence were also computed for both items. For sensitivity, specificity, PPV, and NPV we considered 70% or above values to be acceptable and 80% or above to be high. For FPR and FNR of 30% or less, we considered them to be low. 95% confidence intervals were used. No sample size calculation was performed given that the only study found in the literature used the Palliative care Outcome Scale questionnaire (not the IPOS) and the HADS to screen for depression and anxiety was a secondary analysis of several independent datasets.
## Ethics procedures
Ethical approval was granted from all relevant Research Ethics Committees and was in accordance with the 1964 Helsinki declaration and its later amendments or comparable ethical Standard. All participants gave informed signed consent. SPSS, version 22 (SPSS/IBM Corp., Armonk, NY) software was used.
# Results
A total of 1703 individuals were screened between July 1st, 2015 and February 2016, the majority of which in a health centre, a primary health care facility. There were 18 (1.1%) patients eligible for the study who declined participation and 140 (8.2%) were excluded. A total of 135 (7.9%) patients were included (See [fig_ref] Table 1: Patients screened and included in the study by participating centre [/fig_ref].
Main reason for exclusion was being healthy (75.2%). This is expected given that the primary care centre screened most individuals, as they have the highest rates of daily patients and most of them are healthy. Mean age is 66.8 years (SD 12.7), 58 (43%) are female, 74 (54.8%) have up to 4 years of formal education, 74 (54.8%) are from the Northern region. Most patients (N = 109, 80.7%) had a cancer diagnosis and came from the 7 hospital palliative care services (See [fig_ref] Table 2: Demographic and clinical information of participants Demographic and clinical information N [/fig_ref].
The main reasons for ineligibility and exclusion from the study are presented in [fig_ref] Table 3: Reasons for ineligibility and exclusion [/fig_ref]. Most patients were approached to participate in the study whilst in external consultation, 98 (72.6%). About 31.1% were able to fill the questionnaires without help.
Data were missing at random (Little's MCAR test showed Chi-Square = 2452.946, DF = 2398, Sig. = .213). Missing data varied between 1 and 5% (rates < 1% are trivial, 1-5% are manageable, 5-15% require sophisticated statistical methods to handle, and > 15% may severely impact any form of interpretation.) [bib_ref] The treatment of missing values and its effect on classifier accuracy, Acuñna [/bib_ref]. As expected in palliative populations, most questionnaire items presented a non-parametric distribution, so the imputation of the median was used to handle missing data.
## Prevalence of needs
In terms of prevalence of needs, IPOS items scoring the highest (=4) were: family or friends anxious or worried (36.3%); feeling anxious or worried about illness (13.3%); feeling depressed (9.6%); feeling at peace (9.6%); share feelings (8.9%) and pain (7.4%). IPOS items scoring the lowest (=0) were: vomiting (77%); shortness of breath (67.4%); nausea (65%); information needs (60.7); practical problems (45.2%) and constipation (43%). (See [fig_ref] Figure 1: IPOS scores for prevalence of main palliative needsFig [/fig_ref].
## Screening for anxiety and depression
Item 3 (anxiety) and item 5 (depression) presented the highest AUCs (see . The prevalence of depression was 24.4% (C.I. 17.6-32.7%). The AUC curve was 0.72 [fig_ref] Figure 3: Area Under the Curve for Portuguese IPOS item 5, depression [/fig_ref]. Sensitivity was 51.5% and specificity was 78.4%. Positive predictive value was 43.6% and negative predictive value was 83.3%. As for the anxiety item, the prevalence was 23.7% (C.I. [bib_ref] Capture, transfer, and feedback of patient centred outcomes data in palliative care..., Etkind [/bib_ref].9%-.31.9%) and the AUC was 0.70 (C.I.:0.60-0.80), p < 0.001 (see [fig_ref] Figure 4: Area Under the Curve for Portuguese IPOS item 3, anxiety Antunes et... [/fig_ref]. Sensitivity was 65.6% and specificity was 68.0%. Positive predictive value was 38.8% and negative predictive value was 86.4% (see .
# Discussion
The main palliative needs of patients cared for by palliative care teams were psychological, family related and spiritual. IPOS systematically identified main needs in this population. Clinical teams seemed to solve or manage physical issues well. This is an extremely positive find. There is evidence that once physical needs are well managed, and the patient is more comfortable, non- physical palliative needs arise [bib_ref] What matters most for end-of-life care? Perspectives from community-based palliative care providers..., Mistry [/bib_ref] [bib_ref] Unmet care needs of advanced cancer patients and their informal caregivers: a..., Wang [/bib_ref]. These are also possible to and should be captured systematically in clinical practice, using a patient centred outcome measure and acted upon. In our study, the most stressful non physical issue was family or friends anxious or worried (36.3%). Given that Portuguese culture is based on strong family ties and that decisions are usually made within the family core, this was not surprising. Feeling anxious and depressed were the second and third most stressful issues, even though one of the exclusion criteria to be approached for invitation to participate in the study was being clearly in distress (physical or emotional) judged by direct observation of the healthcare professional. This reinforces the evidence that anxiety is often present in patients with advanced disease due to uncertainties in diagnostic, treatment and prognostic, [bib_ref] Psychological distress and quality of life of palliative cancer patients and their..., Götze [/bib_ref] and, depression is also common among patients with advanced disease [bib_ref] The development of evidencebased European guidelines on the management of depression in..., Rayner [/bib_ref]. In our sample the prevalence of these issues resonates with Ann-Yi S study in terms of the percentage of patients cared for in the psychology service [bib_ref] Characteristics and outcomes of psychology referrals in palliative care department, Bruera [/bib_ref] , although that study was conducted only with cancer patients. On the other hand, clusters of physical and non physical symptoms occur and are common in both cancer and non cancer patients [bib_ref] Are there differences in the prevalence of palliative care-related problems in people..., Moens [/bib_ref] [bib_ref] Symptoms and problem clusters in cancer and non-cancer patients in specialized palliative..., Stiel [/bib_ref]. Also, the prevalence of depression in the present study was somewhat higher than the one presented in the Antunes study on screening for depression using one item of the Palliative care Outcome Scale, namely 17.5% (C.I. 14.1-21.6%), but lower than the 30% estimated by Hotopf and colleagues for prevalence of all depressive disorders in advanced disease [bib_ref] Depression in advanced disease: a systematic review. Part 1. Prevalence and case..., Hotopf [/bib_ref]. Both AUC for anxiety and depression were acceptable (70% or above), although both C.I.s lower levels were slightly below 70%. For cut-off 2/3, both items did not perform well regarding sensitivity, which means these might not be good to identify true positive cases. However, specificity and NPV were good. Both items seem to be good excluding true negative cases, which is one component of screening [bib_ref] Are one or two simple questions sufficient to detect depression in cancer..., Mitchell [/bib_ref]. These results also reinforce external validity of IPOS.
The main limitation in our study is that the optimal gold standard to screen for anxiety and depression -psychiatric interview for depression as determined by the Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition) was not available to use due to low resources available. The present study used the HADS, a screening measure well accepted in practice and that has been used extensively both in practice and research for several years, nevertheless this is not a diagnostic tool.
Using IPOS systematically could aid clinical teams to track their progress in assisting patients and families with physical and non-physical symptoms. Like other screening tools the Portuguese Integrated Palliative care Outcome Scale seems to be good for excluding true negative cases of depression (item 5) and anxiety (item 3) and can be used to screen patients with advanced disease [bib_ref] Are one or two simple questions sufficient to detect depression in cancer..., Mitchell [/bib_ref] [bib_ref] Sensitivity and specificity of a twoquestion screening tool for depression in a..., Payne [/bib_ref].
# Conclusions
Patient centred outcome measures are powerful communication tools, allowing all those involved in patient care to use a common language, serving not only patients and families, but aiding healthcare professionals, health institutions and policy makers to make evidence supported decisions and improve patient centred care. Building evidence of screening properties of these measures allows not only for patient clinical care, but also to conduct more robust research studies. This study determined screening accuracy properties of the Portuguese Integrated Palliative care Outcome Scale for two psychological related items and shows that this measure can be used to screen patients with advanced disease. Authors' contributions BA performed study procedures and collection of data, analysis of data, and was responsible for drafting and editing of the manuscript; PPR revised the manuscript; IJH revised the manuscript; PLF revised of the manuscript. All authors read and approved the final manuscript.
# Funding
This study was funded by Calouste Gulbenkian Foundation -Programa Gulbenkian Inovar em Saúde.
## Availability of data and materials
The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.
Ethics approval and consent to participate Ethical approval was granted from all Research Ethics Committees' included centres and was in accordance with the 1964 Helsinki declaration and its later amendments or comparable ethical Standard: Comissão de Ética para a Saúde do Centro Hospitalar de S. João -Entidade Pública Empresarial (no reference number); Comissão de Ética para a Saúde Administração Regional
[fig] Figure 1: IPOS scores for prevalence of main palliative needsFig. 2 Area Under the Curve for IPOS items 3 to 7 [/fig]
[fig] Figure 3: Area Under the Curve for Portuguese IPOS item 5, depression [/fig]
[fig] Figure 4: Area Under the Curve for Portuguese IPOS item 3, anxiety Antunes et al. BMC Palliative Care (2020) 19:69Table 4 -Estimated Values for items 3 and 5, cut-off 2/3, against the HADS subscales, using cut-off 107% (C.I. 16.9-31.9%) 4% (C.I. 17.6-32.7%) [/fig]
[table] Table 1: Patients screened and included in the study by participating centre [/table]
[table] Table 2: Demographic and clinical information of participants Demographic and clinical information N(%) [/table]
[table] Table 3: Reasons for ineligibility and exclusion [/table]
|
Parental occupational exposure to endocrine disrupting chemicals and male genital malformations: A study in the danish national birth cohort study
Background: Sex hormones closely regulate development of the male genital organs during fetal life. The hypothesis that xenobiotics may disrupt endogenous hormonal signalling has received considerable scientific attention, but human evidence is scarce. Objectives: We analyse occurrence of hypospadias and cryptorchidism according to maternal and paternal occupational exposure to possible endocrine disrupting chemicals.Methods:We conducted a follow-up study of 45,341 male singleton deliveries in the Danish National Birth Cohort during 1997-2009. Information on work during pregnancy was obtained by telephone interviews around gestational week 16. Parents' job titles were classified according to DISCO-88. A job exposure matrix for endocrine disrupting chemicals (EDCs) was implemented to assess occupational exposures. The Medical Birth and National Hospital Register provided data on congenital anomalies diagnosed at birth or during follow-up, which ended in 2009. Crude and adjusted hazard ratios (HR) were obtained from Cox regression models.Results: Among all pregnancies, 6.3% were classified as possibly or probably exposed to EDCs. The most prevalent occupations conferring possible exposure were cleaners, laboratory technicians, hairdressers and agricultural workers (58% of all potentially exposed). The final cumulative incidence of cryptorchidism in boys was 2.2% (1002 cases), and of hypospadias 0.6% (262 cases). The occurrence of hypospadias increased when mothers were probably [HRa = 1.8 (95% CI 1.0-2.6)] or possibly exposed to one or more EDCs [HRa = 2.6 (95% CI 1.8-3.4). Possible paternal exposure to heavy metals increased the risk of hypospadias [HRa 2.2 (95% CI: 1.0-3.4)] and cryptorchidism [HRa 1.9 (95% CI: 1.1-2.7)]. None of the exposure groups reached statistical significance.Conclusion:The study provides some but limited evidence that occupational exposure to possible endocrine disrupting chemicals during pregnancy increases the risk of hypospadias.
# Background
Cryptorchidism (incomplete testicular descent) is a common congenital disorder, but may also be acquired [bib_ref] Acquired cryptorchidism is frequent in infancy and childhood, Wohlfahrt-Veje [/bib_ref]. The prevalence at three months of age was 1.9% in a Danish sample [bib_ref] Difference in prevalence of congenital cryptorchidism in infants between two Nordic countries, Boisen [/bib_ref] , 1.0 in a sample from USAand between 1.6% [bib_ref] Prevalence and natural history of cryptorchidism, Berkowitz [/bib_ref] and 2.4% [bib_ref] The descriptive epidemiology of congenital and acquired cryptorchidism in a UK infant..., Acerini [/bib_ref] in large samples from UK. Hypospadias (abnormal location of the urethral orifice) is observed in 2-4 per 1,000 male births in Europe [bib_ref] A high hypospadias rate in The Netherlands, Pierik [/bib_ref]. There are indications that the prevalence of cryptorchidism and hypospadias increased from the 60s to the 80s in Europe and in the USA, although data providing this information have important methodological limitations [bib_ref] Is the incidence of hypospadias increasing? Analysis of Finnish hospital discharge data..., Aho [/bib_ref] [bib_ref] Trends in the incidence of cryptorchisdism and hypospadias, and methodological limitations of..., Toppari [/bib_ref].
The fetal development of the male reproductive organs is controlled by sex hormones, and in particular androgens play a crucial role during the first trimester of pregnancy [bib_ref] Environmental endocrine disrupters and disorders of sexual differentiation, Toppari [/bib_ref]. Fetal exposure to chemicals with anti-androgen or estrogen-like activity may interfere with normal hormonal signalling, which may increase the risk of cryptorchidism, hypospadias and other male reproductive disorders [bib_ref] Testicular dysgenesis syndrome: an increasingly common developmental disorder with environmental aspects, Skakkebaek [/bib_ref] [bib_ref] The 'oestrogen hypothesis'-where do we stand now?, Sharpe [/bib_ref].
Many widespread chemicals including dioxins and furans, polychlorinated biphenyls, organochlorine pesticides, phthalate esters, brominated flame-retardants and some heavy metals have been identified as possible endocrine disrupters [bib_ref] A job-exposure matrix for potential endocrinedisrupting chemicals developed for a study into..., Van Tongeren [/bib_ref]. Therefore, it remains an important issue to corroborate or refute the hypothesis role of these chemicals in male reproductive disorders [bib_ref] Figà-Talamanca I: Maternal exposures to endocrine disrupting chemicals and hypospadias in offspring, Giordano [/bib_ref] [bib_ref] Parental occupational exposure to potential endocrine disrupting chemicals and risk of hypospadias..., Nassar [/bib_ref] [bib_ref] Exposure to endocrine disrupting chemicals and human health: a review of epidemiological..., Kurahashi [/bib_ref] [bib_ref] Maternal serum level of the DDT metabolite DDE in relation to fetal..., Longnecker [/bib_ref] [bib_ref] Maternal pregnancy levels of polychlorinated biphenyls and risk of hypospadias and cryptorchidism..., Mcglynn [/bib_ref] [bib_ref] Maternal exposure to a brominated flame retardant and genitourinary conditions in male..., Small [/bib_ref] [bib_ref] Flame retardants in placenta and breast milk and cryptorchidism in newborn boys, Main [/bib_ref] [bib_ref] Risk factors for congenital cryptorchidism in a prospective birth cohort study, Damgaard [/bib_ref] [bib_ref] Human breast milk contamination with phthalates and alterations of endogenous reproductive hormones..., Main [/bib_ref]. In this study, we examine occupational exposure to potentially endocrine disrupting compounds, utilizing data from a large Danish National Birth Cohort. The objective was to estimate the risk of hypospadias and cryptorchidism according to maternal and paternal occupational exposures to chemicals interfering with hormonal homeostasis with specific focus on fetal exposure during pregnancy.
# Methods
## Study population
We used data from the Danish National Birth Cohort (DNBC), which is a nationwide study among pregnant women and their offspring [bib_ref] The Danish National Birth Cohort-its background, structure and aim, Olsen [/bib_ref]. Between March 1997 and November 2002 pregnant women across Denmark were informed about the study during their first antenatal visit to the general practitioner. About 60% of invited women accepted the invitation by signing an informed consent form. The only exclusion criteria were not having access to a telephone, not speaking Danish well enough to complete the interview, and not intending to carry the pregnancy to term.
A total of 101,052 pregnant women were enrolled in the study, and 92,892 participated in the first interview at approximately 16 weeks of gestation (interquartile range, [bib_ref] Environmental endocrine disrupters and disorders of sexual differentiation, Toppari [/bib_ref] [bib_ref] Testicular dysgenesis syndrome: an increasingly common developmental disorder with environmental aspects, Skakkebaek [/bib_ref] [bib_ref] The 'oestrogen hypothesis'-where do we stand now?, Sharpe [/bib_ref] [bib_ref] A job-exposure matrix for potential endocrinedisrupting chemicals developed for a study into..., Van Tongeren [/bib_ref] [bib_ref] Figà-Talamanca I: Maternal exposures to endocrine disrupting chemicals and hypospadias in offspring, Giordano [/bib_ref] [bib_ref] Parental occupational exposure to potential endocrine disrupting chemicals and risk of hypospadias..., Nassar [/bib_ref] [bib_ref] Exposure to endocrine disrupting chemicals and human health: a review of epidemiological..., Kurahashi [/bib_ref] [bib_ref] Maternal serum level of the DDT metabolite DDE in relation to fetal..., Longnecker [/bib_ref] [bib_ref] Maternal pregnancy levels of polychlorinated biphenyls and risk of hypospadias and cryptorchidism..., Mcglynn [/bib_ref] [bib_ref] Maternal exposure to a brominated flame retardant and genitourinary conditions in male..., Small [/bib_ref] [bib_ref] Flame retardants in placenta and breast milk and cryptorchidism in newborn boys, Main [/bib_ref] [bib_ref] Risk factors for congenital cryptorchidism in a prospective birth cohort study, Damgaard [/bib_ref] [bib_ref] Human breast milk contamination with phthalates and alterations of endogenous reproductive hormones..., Main [/bib_ref] [bib_ref] The Danish National Birth Cohort-its background, structure and aim, Olsen [/bib_ref]. Women were contacted by trained female telephone interviewers. Interviews were classified as missing if women were not reached at the scheduled time or after three additional attempts to make contact. Interviews were cancelled if the contacted woman was no longer pregnant. All Regional Science Ethics Committees in Denmark have approved the DNBC and before we initiated this study we obtained approval from the Danish Data Protection Agency.
## Exposure assessment
In the first study interview, mothers answered questions about their work three months before pregnancy and during pregnancy, and also about the father's work. Mothers provided information on their current or most recent jobs and about the father's job. This information was coded according to the Danish version of International Standard Classification of Occupation (DISCO-88), which contains 348 job titles.
Maternal and paternal occupational codes were classified into categories of potential exposure to possible endocrine disrupting chemicals (EDCs), using a job exposure matrix that was developed by a Dutch-British group [bib_ref] A job-exposure matrix for potential endocrinedisrupting chemicals developed for a study into..., Van Tongeren [/bib_ref]. Three occupational exposure experts classified independently all the DISCO-88 job titles into three exposure categories: "unlikely", "possible" or "probable" exposure to one or more of seven groups of EDCs: pesticides, organochlorine compounds, phthalate esters, alkyl phenols, bis-phenols, heavy metals (cadmium, lead, mercury), and other compounds (hormone disrupting chemicals). The classification was scored according to the following criteria: (0) Exposure among the workers with this job title is very "unlikely" (1) There is a "possibility" that some of the workers with this job title are exposed (but the probability is fairly low) (2) There is some "probability" that at least a proportion of the workers with this job title are somewhat exposed. The job exposure ratings of the three experts were compared. If there was disagreement exceeding one, discussion to reach consensus was performed but differences of one category were allowed. The final code assigned each occupation was the median value of the three expert scores. Two-step analyses were made of firstly: combined EDC exposure, and secondly: to each of the chemical groups.
## Ascertainment of outcome
Pregnancy outcome was identified and assessed using Danish national registers, by using the unique identifier code given to each individual at birth. The Medical Birth Register and the Civil Registration System were used to obtain data on live births, stillbirths and emigration of the mother before pregnancy ended. Other pregnancy outcomes were identified in the National Hospital Discharge Register during the period 1997 through 2009. Thus, the youngest child was 6 years old and the oldest was 13 years old at the end of follow-up. Less than one percent of the study pregnancies could not be linked to registry data, in which case information from the pregnancy interview was used instead.
The National Hospital Discharge Register included information about congenital anomalies based on the 10 th Revision of the International Classification of Diseases (ICD10 codes DQ00-DQ99). The registry covered 100% of Denmark's hospitals during the study period including all inpatient and outpatient clinic diagnoses and surgeries performed. We identified 262 cases of hypospadias (ICD10 codes DQ54-DQ549, with the exeption of DQ544) and 1002 cases of cryptorchidism (ICD10 codes Q53, Q531, Q531A, Q532, Q532A, and Q539). Twenty-two boys had both anomalies. Information on surgical correction of cryptorchidism, orchiopexy (codes KKFH00, KKFH01, and KKFH10 in the Nordic Classification of Surgical Procedures) was included in the analyses. Orchiopexy indicates that the disease (cryptorchidism) is persisting and requires surgery.
# Statistical analysis
A Cox regression analysis with the boys' age as the time variable was performed to compare the occurrence of hypospadias and cryptorchidism in the "possible" and "probable" exposure categories using the baseline "unlikely" exposure category as reference. The boys entered the risk set at birth and were followed until their age at first diagnosis, surgery, death, emigration from Denmark, or end of follow up (October 21, 2009), whichever came first. Separate analyses were made for cryptorchidism with and without orchiopexy. We present the estimated HRs with 95% confidence intervals (95% CI). The following fixed set of covariates was defined a priori and included in the models regardless of their effect on the risk estimates: Age of the mother and the father in first trimester (≤20, 20.1-30, 30.1-35, 35.1-40 and over 40.1 years); mothers' pre-pregnancy body mass index (kg/m 2 ) according to three body mass index categories (< 20, 20-30 and > 30); Previous spontaneous abortion (yes/no); parity (primiparous (1), 2-3, more than 4); birth weight of boys (<2500 g, 2500-4000 g and >4000 g); gestational age (< 24 weeks, 24.1-32 weeks, 32.1 -37 weeks and ≥37.1 weeks) oral contraceptives used in the past (yes/no); treatment of infertility (yes/no); time to pregnancy (unexpected pregnancy, immediately, 1-2 months, 3-5 months, 6-12 months, ≥ 13 months); mother's alcohol consumption during pregnancy (yes/no); binge drinking defined as intake of at least five drinks at one occasion at least one time (yes/no); maternal smoking during pregnancy (no smoking, ≤10 cigarettes/day, 11-19 cigarettes/day, and ≥ 20 cigarettes/day); paternal smoking (no, yes-not every day, yes-every day), we considered: one cigarette = one cigarette equivalent, one cherrot = two cigarettes equivalent, one cigar = two cigarettes equivalent, one pipe = one cigarette and a half equivalent; gynecological disease (yes/no); maternal vegetarian diet (yes/no). The mothers' job titles were used to define their socio-occupational status. Women with a higher level of education (4 years beyond high school), or who worked in management positions, were classified as "high socio-occupational status". The "medium" category included skilled workers and women with medium-ranged training/level of education, while the "low" category included unskilled workers. All the analyses were performed using the SPSS software (version 14.0; SPSS Inc. Chicago III).
# Results
Of the 92,892 women participating in the first telephone interview, we excluded all subsequent pregnancies (n = 2,425), women whose pregnancies ended with induced abortion (n = 93), hydatidiform mole (n = 42), ectopic pregnancy (n = 24), multiple birth with no live born infants (n = 3), women who died during pregnancy (n = 1), and unknown outcomes of the pregnancy (n = 9). Women who did not work during pregnancy (n = 1,431) were also excluded. From the remaining 88,864 live-born singletons, 45,341 boys were available for analysiswith information about EDC exposure and follow-up time. Characteristics of the study population are presented in [fig_ref] Table 1: Characteristics of 45,341 pregnancies according to maternal occupational exposure to ECDs [/fig_ref] , and show that low socio-occupational status (SES) is associated with the risk of being exposed to ECDs. Other characteristics seemed to occur at similar proportions in the various EDC exposure categories.
The cumulative incidence of cryptorchidism at the end of follow-up was 2.2% (1002 cases) and of hypospadias it was 0.6% (262 cases). Almost 0.6% of all women were classified as being possibly exposed to at least one of the seven classes of EDC's, while 5.7% were classified as being probably exposed. The most prevalent occupations conferring likely exposure to EDC's among women were cleaners, laboratory technicians, hairdressers and agricultural workers (58.0% of all the exposed women). Among fathers, 14.9% were classified as being possibly exposed to one or more classes of EDCs, while 11.9% fell into the probably exposed category.
The age distributions for first hospital referral for cryptorchidism, orchiopexi and hypospadias are presented in [fig_ref] Figure 1: Age distribution of first hospital referral for cryptorchisdism, orchiopexy and hypospadiasin the... [/fig_ref]. The figures show the age when cases were diagnosed in the health care system and not the age when the anomalies became clinically manifest.
The risk of hypospadias but not of cryptorchidism was significantly elevated following possible and probable maternal exposure to one or more classes of EDCs [fig_ref] Table 2: Hazard ratios [/fig_ref]. The most prevalent occupations conferring possible exposure were cleaners, laboratory technicians, hairdressers and agricultural workers (58% of all potentially exposed). The increased risk of hypospadias was mostly related to potential exposure to pesticides, phthalate esters, alkyl phenols, bis-phenols and other compounds, although the confidence intervals for all the subgroup analyses were wide. Paternal exposure were not consistently related to increased risk of hypospadias, but an association with probable exposure to heavy metals [HRa 2.2 (95% CI: 1.0-3.4)] was observed. Cryptorchidism was associated with fathers' probable exposure to heavy metals [HRa 1.9 (95% CI: 1.1-2.7)] but not with fathers' possible exposure or with maternal exposure.
There was no difference in risk of cryptorchidism whether an orchiopexy was performed or not [fig_ref] Table 3: Hazard ratio [/fig_ref].
# Discussion
We observed a modestly increased risk for hypospadias in relation to maternal occupational EDC exposure and paternal exposure to heavy metals while the risk of cryptorchidism was not increased. These results are not entirely consistent with the hypothesis that the two male disorders share EDC exposure as etiological factor. If occupational exposure to combined EDCs is increasing the risk of hypospadias and cryptorchidism, the risk should be elevated consistently according to several of the exposure subcategories, since all have expected EDC exposure in common but we don't see it in our results. Few and scattered associations do not indicate a strong association with endocrine disruption. It must of course be acknowledged that the various exposures may impact many biological pathways and that exposure prevalence and exposure levels within and across occupations may vary substantially. The cumulative incidence of cryptorchidism of 2.2% in the studied sample is close to the rate of 1.9% observed in a prospective study of a subset of the DNBC [bib_ref] Difference in prevalence of congenital cryptorchidism in infants between two Nordic countries, Boisen [/bib_ref] , but less than the 3,2%, observed in a large representative sample of the Danish population [bib_ref] Cryptorchidism concordance in monozygotic and dizygotic twin brothers, full brothers, and half-brothers, Jensen [/bib_ref]. This difference may be explained by shorter follow-up time in our study.
Congenital cases of transient cryptorchidism may not be clinically recognized and reported to the register, as spontaneous descent is frequent during the first 3 months of life [bib_ref] Difference in prevalence of congenital cryptorchidism in infants between two Nordic countries, Boisen [/bib_ref]. Cases are reported later if the condition persists, and the register-based cumulative incidence of cryptorchidism is increasing until 15 years of age [bib_ref] Cryptorchidism concordance in monozygotic and dizygotic twin brothers, full brothers, and half-brothers, Jensen [/bib_ref]. Thus, the register-based cryptorchidism endpoint underascertains mild and transient cases of cryptorchidism compared to cohorts with clinical examinations, but the register does include both congenital and acquired cases [bib_ref] Acquired cryptorchidism is frequent in infancy and childhood, Wohlfahrt-Veje [/bib_ref]. With a main focus on persisting cryptorchidism, a lower sensitivity to transient cases will bias results if the conditions have different aetiology, or if the ascertainment rate varies by exposure level. We have no reason to suspect that the diagnosing or routine reporting of this condition is related to maternal occupation. There is evidence that different hormonal signals are involved in the abdominal and inguinal testicular descent [bib_ref] Hormones in male sexual development, Nef [/bib_ref] , but this will hardly affect risk estimates, as testes retained in the abdomen only comprise a few percent of all cases of cryptorchidism [bib_ref] Male reproductive disorders in humans and prenatal indicators of estrogen exposure A..., Storgaard [/bib_ref]. The Danish National Birth Cohort (1997-2009). CIP = Cumulative Incidence Proportion per 100 boys. CI = Confidence Interval. HRa Hazard Ratio adjusted by age of the mother and the father; mothers' pre-pregnancy body mass index (kg/m 2 ); earlier spontaneous abortion; parity; birth weight of boys; gestational age; oral contraceptives used; treatment of infertility; time taken to conceive; mother's alcohol consumption during pregnancy; binge drinking; maternal smoking during pregnancy; paternal smoking; gynecological disease; maternal vegetarian diet.
The observed cumulative incidence rate of hypospadias of 0.6% does not include mild glandular and coronal cases and is therefore substantially lower than the prevalence of 4.7% that was reported in a study of newborn boys from the Copenhagen area [bib_ref] Hypospadias in a cohort of 1072 Danish newborn boys: Prevalence and relationship..., Boisen [/bib_ref]. Although the sensitivity of hospital register-based reporting is low, the specificity is probably high. As for cryptorchidism, the reported analyses of relative effect estimates are therefore unlikely to be biased, unless diagnosing and reporting is related to maternal occupation, which is quite unlikely. The registry ascertainment of the hypospadias diagnosis mainly addresses the more severe cases.
This study assesses maternal and paternal occupational exposure to potential EDCs by a job exposure matrix (JEM). The same occupational exposure criteria were applied to mothers and fathers. Thus, the JEM approach ignores possible gender-specific exposure profiles within occupations. Moreover, little is known about actual exposure levels and possible interactions among multiple endocrine disrupters [bib_ref] A job-exposure matrix for potential endocrinedisrupting chemicals developed for a study into..., Van Tongeren [/bib_ref]. In addition, the JEM neither distinguishes substances with different mechanisms or potency for endocrine disruption nor incorporates any possible changes in exposure over time. In short, a JEM based approach will yield underestimated associations if the exposure causes the diseases we study.
Occupational studies addressing the endocrine disruption hypothesis are few in spite of the fact that the occupational setting often confers much higher exposure levels than environmental sources, which probably also applies to EDCs. A limitation of our study is the lack of information on actual exposure levels at the workplace. Moreover, we do not know if jobs that according to the hygienic experts may to involve exposure to phthalates, for instance, are conferring higher exposures than the ubiquitous exposure of the general population to these compounds. Although the lack of (consistent) associations in this study indicates that EDCs defined without referring to specific mechanisms (estrogenic, anti-androgenic etc) have little importance in the occupational setting, such findings do not rule out that environmental exposures could be important. We did observe discordant effects of parental exposure to combined EDCs for hypospadias -maternal exposure was associated with elevated occurrence and paternal exposure was not. This would be expected for a causal maternal exposure, whereas confounding or indirect effects might present elevated risks for both maternal and paternal exposure.
The risk of hypospadias according to occupational exposure during pregnancy has been examined in five published studies that have applied the EDC job exposure matrix developed by van Tongeren et al in 2002 [bib_ref] A job-exposure matrix for potential endocrinedisrupting chemicals developed for a study into..., Van Tongeren [/bib_ref]. While no increased risk related to maternal exposure was found in a Dutch nationwide register-based study [bib_ref] Risk of hypospadias in relation to maternal occupational exposure to potential endocrine..., Vrijheid [/bib_ref] and a Dutch case-referent study [bib_ref] Maternal and paternal risk factors for cryptorchidism and hypospadias: a case-control study..., Pierik [/bib_ref] , three subsequent case-referent studies all reported increased risk related to exposure to one or more classes of endocrine disrupters [bib_ref] Figà-Talamanca I: The possible role of endocrine disrupting chemicals in the aetiology..., Carbone [/bib_ref] , to heavy metals [bib_ref] Parental occupational exposure to potential endocrine disrupting chemicals and risk of hypospadias..., Nassar [/bib_ref] or to phthalates [bib_ref] Endocrine disruptors in the workplace, hair spray, folate supplementation, and risk of..., Ormond [/bib_ref]. A Spanish nested case-control study presented an increased risk of hypospadias or cryptorchidism (OR = 2.8; 95% CI: 1.1-7.2) in relation to the *HRa.-Hazard Ratio Adjusted by age of the mother and the father; mothers' pre-pregnancy body mass index (kg/m 2 ); earlier spontaneous abortion; parity; birth weight of boys; gestational age; oral contraceptives used; treatment of infertility; time taken to conceive; mother's alcohol consumption during pregnancy; binge drinking; maternal smoking during pregnancy; paternal smoking; gynecological disease; maternal vegetarian diet. measured amount of xenobiotic estrogenic activity in blood [bib_ref] Human exposure to endocrinedisrupting chemicals and prenatal risk factors for cryptorchidism and..., Fernandez [/bib_ref]. The limited size of this study did not allow for separate analyses of the two urogenital malformations. Phthalates are of particular interest since several of the above studies reported elevated risk related to this class of anti-androgenic chemicals [bib_ref] Figà-Talamanca I: Maternal exposures to endocrine disrupting chemicals and hypospadias in offspring, Giordano [/bib_ref] [bib_ref] Parental occupational exposure to potential endocrine disrupting chemicals and risk of hypospadias..., Nassar [/bib_ref] [bib_ref] Maternal and paternal risk factors for cryptorchidism and hypospadias: a case-control study..., Pierik [/bib_ref] [bib_ref] Figà-Talamanca I: The possible role of endocrine disrupting chemicals in the aetiology..., Carbone [/bib_ref]. Our data were also indicative of an adverse effect of phthalates, but estimates were not statistically significant. Some phthalates and several phthalate metabolites inhibit androgen synthesis in the fetal Leydig cell at environmental exposure levels [bib_ref] Dose-dependent alterations in androgen-regulated male reproductive development in rats exposed to Di(n-butyl)..., Mylchreest [/bib_ref] , and the compounds have been related to decreased anogenital distance [bib_ref] Human breast milk contamination with phthalates and alterations of endogenous reproductive hormones..., Main [/bib_ref] [bib_ref] Study for Future Families Research Team: Decrease in anogenital distance among male..., Swan [/bib_ref] , which is associated with hypospadias [bib_ref] Fetal testosterone insufficiency and abnormal proliferation of Leydig cells and gonocytes in..., Mylchreest [/bib_ref] [bib_ref] Possible impact of phthalates on infant reproductive health, Lottrup [/bib_ref]. Hairdressers belong to the largest single occupational group with a probable exposure to phthalates and possibly other endocrine disrupting chemicals [bib_ref] A job-exposure matrix for potential endocrinedisrupting chemicals developed for a study into..., Van Tongeren [/bib_ref].
A study of concordance rates of cryptorchidism in twin brothers, full brothers and half brothers clearly point towards important etiologic factors in the intrauterine environment provided by the mother. Recently, persistent pesticides and brominated flameretardants in human breast milk have been linked to cryptorchidism in a large prospective study [bib_ref] Flame retardants in placenta and breast milk and cryptorchidism in newborn boys, Main [/bib_ref] [bib_ref] Nordic Cryptorchidism Study Group: Persistent pesticides in human breast milk and cryptorchidism, Damgaard [/bib_ref]. A high cryptorchidism frecuency was indicated among sons of women working in greenhouses during pregnancy compared to an external reference group [bib_ref] Nordic Cryptorchidism Study Group: Persistent pesticides in human breast milk and cryptorchidism, Damgaard [/bib_ref]. Furthermore, occupational exposure to pesticides was also associated with other adverse effects such as decreased penile length, testicular volume and serum concentrations of testosterone and inhibin B. Contrary to the above results, three large case-referent studies have failed to demonstrate consistent associations between blood concentrations of a number of biopersistent xenobiotics and occurrence of cryptorchidism [bib_ref] Maternal pregnancy levels of polychlorinated biphenyls and risk of hypospadias and cryptorchidism..., Mcglynn [/bib_ref] [bib_ref] Maternal exposure to a brominated flame retardant and genitourinary conditions in male..., Small [/bib_ref] [bib_ref] Maternal serum level of 1,1-dichloro-2,2-bis(p-chlorophenyl) ethylene and risk of cryptorchidism, hypospadias, and..., Longnecker [/bib_ref] , but these biomarkers have all been measured at time points outside pregnancy. We observed no excess risk of cryptorchidism by maternal occupational EDC exposure, but instead some inconsistent association on never observe effects of paternal exposures. Stratifying cryptorchidism cases by orchiopexy yielded comparable results in the two groups.
In conclusion, this study provides some but limited evidence that occupational exposure to endocrine disrupting chemicals in general increases the risk of hypospadias.
# Conclusions
The study provides some evidence that occupational exposure to possible endocrine disrupting chemicals during pregnancy increases the risk of hypospadias.
Abbreviations EDC: endocrine disrupting chemical; JEM: job exposure matrix; DNBC: Danish National Birth Cohort.
[fig] Figure 1: Age distribution of first hospital referral for cryptorchisdism, orchiopexy and hypospadiasin the DNBC (1997-2009). [/fig]
[table] Table 1: Characteristics of 45,341 pregnancies according to maternal occupational exposure to ECDs [/table]
[table] Table 2: Hazard ratios (HR) for hypospadias and cryptorchidism in boys according to prenatal exposure to endocrine disrupting chemicals conferred by parental occupational exposure (n = 45,341) [/table]
[table] Table 3: Hazard ratio (HR) for a cryptorchidism diagnosis with or without orchipexy among 45,341 Danish boys, according to parental occupational exposure to ECDs before and during pregnancy CI = Confidence Interval. CIP = Cumulative Incidence Proportion per 100 boys. [/table]
|
Peri‐operative management of the surgical patient with diabetes 2015
Diabetes affects 10-15% of the surgical population and patients with diabetes undergoing surgery have greater complication rates, mortality rates and length of hospital stay. Modern management of the surgical patient with diabetes focuses on: thorough pre-operative assessment and optimisation of their diabetes (as defined by a HbA1c < 69 mmol.mol À1 ); deciding if the patient can be managed by simple manipulation of pre-existing treatment during a short starvation period (maximum of one missed meal) rather than use of a variable-rate intravenous insulin infusion; and safe use of the latter when it is the only option, for example in emergency patients, patients expected not to return to a normal diet immediately postoperatively, and patients with poorly controlled diabetes. In addition, it is imperative that communication amongst healthcare professionals and between them and the patient is accurate and well informed at all times. Most patients with diabetes have many years of experience of managing their own care. The purpose of this guideline is to provide detailed guidance on the peri-operative management of the surgical patient with diabetes that is specific to anaesthetists and to ensure that all current national guidance is concordant.
# Introduction
The demographics describing the dramatic increase in the number of patients with diabetes are well known. Patients with diabetes require surgical procedures more frequently and have longer hospital stays than those without the condition [bib_ref] Prevalence and clinical outcome of hyperglycemia in the perioperative period in noncardiac..., Frisch [/bib_ref]. The presence of diabetes or hyperglycaemia in surgical patients has been shown to lead to increased morbidity and mortality, with perioperative mortality rates up to 50% greater than the non-diabetic population [bib_ref] Prevalence and clinical outcome of hyperglycemia in the perioperative period in noncardiac..., Frisch [/bib_ref]. The reasons for these adverse outcomes are multifactorial, but include: failure to identify patients with diabetes or hyperglycaemia [bib_ref] Deaths associated with diabetic ketoacidosis and hyperosmolar coma, 1973-1988, Hamblin [/bib_ref] ; multiple co-morbidities including microvascular and macrovascular complications [bib_ref] Diabetes, other risk factors, and 12-yr cardiovascular mortality for men screened in..., Stamler [/bib_ref] ; complex polypharmacy and insulin prescribing errors; increased peri-operative and postoperative infections [bib_ref] Prevalence and clinical outcome of hyperglycemia in the perioperative period in noncardiac..., Frisch [/bib_ref] [bib_ref] Mortality and utilisation of critical care resources amongst highrisk surgical patients in..., Jhanji [/bib_ref] [bib_ref] Identification and characterisation of the high-risk surgical population in the United Kingdom, Pearce [/bib_ref] ; associated hypoglycaemia and hyperglycaemia [bib_ref] Prevalence and clinical outcome of hyperglycemia in the perioperative period in noncardiac..., Frisch [/bib_ref] ; a lack of, or inadequate, institutional guidelines for management of inpatient diabetes or hyperglycaemia [bib_ref] Prevalence and clinical outcome of hyperglycemia in the perioperative period in noncardiac..., Frisch [/bib_ref] [bib_ref] A national survey of in-patient diabetes services in the United Kingdom, Sampson [/bib_ref] ; and inadequate knowledge of diabetes and hyperglycaemia management amongst staff delivering care [bib_ref] Lack of confidence among trainee doctors in the management of diabetes: the..., George [/bib_ref].
Anaesthetists and other peri-operative care providers should be knowledgeable and skilled in the care of patients with diabetes. Management of diabetes is a vital element in the management of surgical patients with diabetes. It is not good enough for the diabetic care to be a secondary, or sometimes forgotten, element of the peri-operative care package.
## Previous guidelines
In April 2011 NHS Diabetes (now part of NHS Improving Quality) published a document: NHS Diabetes Guideline for the Peri-operative Management of the Adult Patient with Diabetes, in association with the Joint British Diabetes Societies (JBDS) [bib_ref] NHS Diabetes guideline for the perioperative management of the adult patient with..., Dhatariya [/bib_ref] (an almost identical version, Management of Adults with Diabetes Undergoing Surgery and Elective Procedures: Improving Standards, is available at www.diabetologists-abcd.org.uk/JBDS/JBDS.htm). This comprehensive guideline provided both background information and advice to clinicians caring for patients with diabetes. Some of the recommendations in that document were due for review in the light of new evidence and, in addition, it was felt that anaesthetists and other practitioners caring for patients with diabetes in the peri-operative period needed shorter, practical advice. The Association of Anaesthetists of Great Britain and Ireland (AAGBI) offered to co-author this shortened guideline, in collaboration with colleagues involved with the 2011 document. The previous 2011 NHS Diabetes guidelines will also be updated in 2015.
## The risks of poor diabetic control
Studies have shown that high pre-operative and perioperative glucose and glycated haemoglobin (HbA1c) levels are associated with poor surgical outcomes. These findings have been seen in both elective and emergency surgery including spinal [bib_ref] Prevalence of previously unknown elevation of glycosylated hemoglobin in spine surgery patients..., Walid [/bib_ref] , vascular [bib_ref] Haemoglobin A1c (HbA1C) in non-diabetic and diabetic vascular patients. Is HbA1C an..., O'sullivan [/bib_ref] , colorectal [bib_ref] Haemoglobin A1c as a predictor of postoperative hyperglycaemia and complications after major..., Gustafsson [/bib_ref] , cardiac [bib_ref] Elevated preoperative hemoglobin A1c level is associated with reduced long-term survival after..., Halkos [/bib_ref] [bib_ref] Elevated glycosylated haemoglobin (HbA1c) is a risk marker in coronary artery bypass..., Alserius [/bib_ref] , trauma [bib_ref] The role of admission blood glucose in outcome prediction of surviving patients..., Kreutziger [/bib_ref] , breast [bib_ref] Hyperglycemia as a risk factor for surgical site infections in patients undergoing..., Vilar-Compte [/bib_ref] , orthopaedic [bib_ref] Factors associated with nonunion, delayed union, and malunion in foot and ankle..., Shibuya [/bib_ref] , neurosurgical, and hepatobiliary surgery [bib_ref] Risk factors for wound infection after cholecystectomy, Chuang [/bib_ref] [bib_ref] Poor postoperative blood glucose control increases surgical site infections after surgery for..., Ambiru [/bib_ref]. One study showed that the adverse outcomes include a greater than 50% increase in mortality, a 2.4-fold increase in the incidence of postoperative respiratory infections, a doubling of surgical site infections, a threefold increase in postoperative urinary tract infections, a doubling in the incidence of myocardial infarction, and an almost twofold increase in acute kidney injury [bib_ref] Prevalence and clinical outcome of hyperglycemia in the perioperative period in noncardiac..., Frisch [/bib_ref]. Paradoxically, there are some data to show that the outcomes of patients with diabetes may not be different from, or may indeed be better than, those without diabetes if the diagnosis is known before surgery [bib_ref] Importance of perioperative glycemic control in general surgery: a report from the..., Kwon [/bib_ref]. The reasons for this are unknown, but may be due to increased vigilance surrounding glucose control for those with a diagnosis of diabetes.
## Referral from primary care and planning surgery
The aim is to ensure that diabetes is as well controlled as possible before elective surgery and to avoid delays to surgery due to poor control. The Working Party supports the consensus advice published in the 2011 NHS Diabetes guideline that the HbA1c should be < 69 mmol.mol À1 (8.5%) for elective cases [bib_ref] NHS Diabetes guideline for the perioperative management of the adult patient with..., Dhatariya [/bib_ref] , and that elective surgery should be delayed if it is ≥ 69 mmol.mol À1 , while control is improved. Changes to diabetes management can be made concurrently with referral to ensure the patient's diabetes is as well controlled as possible at the time of surgery. Elective surgery in patients with diabetes should be planned with the aim of minimising disruption to their self-management.
- Recommendation: Glycaemic control should be checked at the time of referral for surgery. Information about duration, type of diabetes, current treatment and complications should be made available to the secondary care team.
## Surgical outpatient clinic
The adequacy of diabetes control should be assessed again at the time of listing for surgery, ideally with a recorded HbA1c < 69 mmol.mol À1 in the previous three months. If it is ≥ 69 mmol.mol À1 , elective surgery should be delayed while control is improved. In a small number of cases it may not be possible to improve diabetic control pre-operatively, particularly if the reason for surgery, such as chronic infection, is contributing to poor control, or if surgery is semi-urgent. In these circumstances, it may be acceptable to proceed with surgery after explanation to the patient of the increased risks. Patients should be managed as a day case if the procedure is suitable and the patient fulfils the criteria for day-case surgery management. Well-controlled diabetes should not be a contra-indication to day-case surgery.
Patients with poorly controlled diabetes at the time of surgery will need close monitoring and may need to start a variable-rate intravenous insulin infusion (VRIII).
- Recommendation: Patients with diabetes should be identified early in the pre-operative pathway.
## Pre-operative assessment
Appropriate and early pre-operative assessment should be arranged. A pre-operative assessment nurse may undertake the assessment with support from either an anaesthetist or a diabetes specialist nurse. It should occur sufficiently in advance of the planned surgery to ensure optimisation of glycaemic control before the date of proposed surgery. The aim is to ensure that all relevant investigations are available and checked in advance of the planned surgery, that the patient understands how to manage his/her diabetes in the peri-operative period, and that the period of pre-operative fasting is minimised.
- Recommendation: Tests should be ordered to assess co-morbidities in line with National Institute for Health and Care Excellence (NICE) guidance on pre-operative testing. This should include urea and electrolytes and ECG for all patients with diabetes; however, a random blood glucose measurement is not indicated.
## Planning admission (including day surgery)
The aim is to minimise the fasting period, ensure normoglycaemia (capillary blood glucose (CBG) 6-10 mmol.l À1 ) and minimise as far as possible disruption to the patient's usual routine. Ideally, the patient should be booked first on the operating list to minimise the period of fasting. If the fasting period is expected to be limited to one missed meal, the patient can be managed by modification of his/her usual diabetes medication (see below). Patients should be provided with written instructions from the pre-operative assessment team about management of their diabetes medication on the day of surgery, the management of hypo-or hyperglycaemia in the peri-operative period, and the likely effects of surgery on their diabetes control. Patients should be advised to carry a form of glucose that they can take in case of symptoms of hypoglycaemia that will not cause surgery to be cancelled, for example a clear, sugar-containing drink (glucose tablets may be used instead, but some anaesthetists may feel they should not be taken within 6 h of the start of anaesthesia). Patients should be warned that their blood glucose control may be erratic for a few days after the procedure.
- Recommendation: When possible, admission should be planned for the day of surgery, with both the patient and the ward staff aware of the planned peri-operative diabetes care, including a plan to manage hypo-and hyperglycaemia. Surgery should be scheduled at the start of the theatre list to minimise disruption to the patient's glycaemic control.
## Management of existing therapy
With appropriate guidance, patients with diabetes should be allowed to retain control and possession of, and continue to self-administer, their medication. Many patients will have several years' experience and be expert in self-medication.
The aim is to avoid hypo-or hyperglycaemia during the period of fasting and the time during and after the procedure, until the patient is eating and drinking normally. In people who are likely to miss one meal only, this can often be achieved by manipulating the patient's normal medication using the guidance provided in Glycaemic control in patients with diabetes is a balance between their carbohydrate intake and utilisation (for example, exercise). It also depends on what medication they take and how those medications work. Some agents (e.g. sulphonylureas, meglitinides, Guideline for peri-operative adjustment of insulin (short starvation periodno more than one missed meal). insulin and to some extent, thiazolidinediones) act by lowering glucose concentrations, and doses need to be modified or the agents stopped during periods of starvation. Others work by preventing glucose levels from rising (e.g. metformin, glucagon-like peptide-1 analogues, dipeptidyl peptidase-4 inhibitors); these drugs may be continued without the risk of hypoglycaemia. The time of day and the expected duration of the operation need to be considered, as will whether a VRIII will be needed. Patients with continuous subcutaneous insulin infusions only missing one meal should be advised to maintain their CBG at 6-10 mmol.l À1 . If longer periods of starvation are predicted, a VRIII should be used and specialist advice sought. have been designed to take all of these factors into consideration. They are a pragmatic approach to the pre-operative management of all the available classes of agent used to manage diabetes.
## Use of a variable-rate intravenous insulin infusion
Variable-rate intravenous insulin infusions are preferred in: patients who will miss more than one meal; Guideline for peri-operative adjustment of oral hypoglycaemic agents (short starvation periodno more than one missed meal).
## Intra-operative care and monitoring
The aim of intra-operative care is to maintain good glycaemic control and normal electrolyte concentrations, while optimising cardiovascular function and renal perfusion. If possible, multimodal analgesia should be used along with appropriate anti-emetic prophylaxis, to enable an early return to a normal diet and the patient's usual diabetes regimen.
- Recommendation: An intra-operative CBG range of 6-10 mmol.l À1 should be aimed for (an upper limit of 12 mmol.l À1 may be tolerated at times, e.g. if the patient has poorly controlled diabetes and is being managed by a modification of his/her normal medication without a VRIII). It should be understood by all staff that a CBG within the range of 6-10 mmol.l À1 is acceptable and that there is no requirment for a CBG of 6 mmol.l À1 to be the target. The CBG should be checked before induction of anaesthesia and monitored regularly during the procedure (at least hourly, or more frequently if the results are outside the target range). The CBG, insulin infusion rate and substrate infusion should be recorded on the anaesthetic record. Some charts use colour-coded areas to highlight abnormal results requiring further intervention or a change of treatment (see Appendix 2).
## Management of intra-operative hyperglycaemia and hypoglycaemia
If the CBG exceeds 12 mmol.l À1 and insulin has been omitted, capillary blood ketone levels should be measured if possible (point-of-care devices are available). If the capillary blood ketones are > 3 mmol.l À1 or there is significant ketonuria (> 2+ on urine sticks) the patient should be treated as having diabetic ketoacido-sis (DKA). Diabetic ketoacidosis is a triad of ketonaemia > 3.0 mmol.l À1 , blood glucose > 11.0 mmol.l À1 , and bicarbonate < 15.0 mmol.l À1 or venous pH < 7.3. Diabetic ketoacidosis is a medical emergency and specialist help should be obtained from the diabetes team. If DKA is not present, the high blood glucose should be corrected using subcutaneous insulin (see below) or by altering the rate of the VRIII (if in use). If two subcutaneous insulin doses do not work, a VRIII should be started.
Treatment of hyperglycaemia in a patient with type-1 diabetes Subcutaneous rapid-acting insulin (such as Novorapid â , Humalog â or Apidra â ) should be given (up to a maximum of 6 IU), using a specific insulin syringe, assuming that 1 IU will drop the CBG by 3 mmol.l À1 . Death or severe harm as a result of maladministration of insulin, including failure to use the specific insulin syringe, is a 'Never Event'. If the patient is awake, it is important to ensure that the patient is content with proposed dose (patients may react differently to subcutaneous rapid-acting insulin). The CBG should be checked hourly and a second dose considered only after 2 h.
## Treatment of hyperglycaemia in a patient with type-2 diabetes
Subcutaneous rapid-acting insulin 0.1 IU.kg À1 should be given (up to a maximum of 6 IU), using a specific insulin syringe. The CBG should be checked hourly and a second dose considered only after 2 h. A VRIII should be considered if the patient remains hyperglycaemic.
## Treatment of intra-operative hypoglycaemia
For a CBG 4.0-6.0 mmol.l À1 , 50 ml glucose 20% (10 g) should be given intravenously; for hypoglycaemia < 4.0 mmol.l À1 a dose of 100 ml (20 g) should be given.
## Fluid management
There is a limited evidence base for the recommendation of optimal fluid management of the adult diabetic patient undergoing surgery. It is now recognised that Hartmann's solution is safe to administer to patients with diabetes and does not contribute to clinically significant hyperglycaemia [bib_ref] Peri-operative i.v. fluids in diabetic patients -don't forget the salt, Simpson [/bib_ref].
## Fluid management for patients requiring a vriii
The aim is to provide glucose as a substrate to prevent proteolysis, lipolysis and ketogenesis, as well as to optimise intravascular volume status and maintain plasma electrolytes within the normal range. It is important to avoid iatrogenic hyponatraemia from the administration of hypotonic solutions. Glucose 5% solution should be avoided. Use of glucose 4% in 0.18% saline can be associated with hyponatraemia.
The substrate solution to be used should be based on the patient's current electrolyte concentrations. While there is no clear evidence that one type of balanced crystalloid fluid is better than another, halfstrength 'normal' saline combined with glucose is, theoretically, a reasonable compromise to achieve these aims. Thus, the initial fluid should be glucose 5% in saline 0.45% pre-mixed with either potassium chloride 0.15% (20 mmol.l À1 ) or potassium chloride 0.3% (40 mmol.l À1 ), depending on the presence of hypokalaemia (< 3.5 mmol.l À1 ).
The Working Party recognises that these fluids may not be available in all institutions. It is our view that they should be made available in all areas where patients with diabetes will be managed. (Hospitals caring for children will usually have these solutions already available for general paediatric use).
Fluid should be administered at the rate that is appropriate for the patient's usual maintenance requirementsusually 25-50 ml.kg À1 .day À1 (approximately 83 ml.h À1 for a 70-kg patient).
Very occasionally, the patient may develop hyponatraemia without signs of fluid overload. In these circumstances, it is acceptable to prescribe one of the following solutions as the substrate solution: glucose 5% in saline 0.9% with pre-mixed potassium chloride 0.15% (20 mmol.l À1 ); or glucose 5% in saline 0.9% with premixed potassium chloride 0.3% (40 mmol.l À1 ). (Again, hospitals caring for children will usually have these solutions available).
Additional Hartmann's solution or another balanced isotonic crystalloid solution should be used to optimise intravascular volume status.
## Fluid management for patients not requiring a vriii
The aim is to avoid glucose-containing solutions unless the blood glucose is low. It is important to avoid hyperchloraemic metabolic acidosis; Hartmann's solution should be administered to optimise the intravascular volume status. If the patient requires prolonged postoperative fluids (> 24 h), a VRIII should be considered and glucose 5% in saline 0.45% with pre-mixed potassium chloride given as above.
## Returning to 'normal' (pre-operative) medication and diet
The postoperative blood glucose management plan, and any alterations to existing medications, should be clearly communicated to ward staff. Patients with diabetes should be involved in planning their postoperative care. If subcutaneous insulin is required in insulin-na€ ıve patients, or the type of insulin or the time it is to be given is to change, the specialist diabetes team should be contacted for advice.
## Transferring from a vriii back to oral treatment or subcutaneous insulin
If the patient has type-1 diabetes and a VRIII has been used, it must be continued for 30-60 min after the patient has had their subcutaneous insulin (see below). Premature discontinuation is associated with iatrogenic DKA.
## Restarting oral hypoglycaemic medication
Oral hypoglycaemic agents should be recommenced at pre-operative doses once the patient is ready to eat and drink; withholding or reduction in sulphonylureas may be required if the food intake is likely to be reduced. Metformin should only be restarted if the estimated glomerular filtration rate exceeds 50 ml.min À1 .1.73 m À2 [bib_ref] Use of metformin in the setting of mild-to-moderate renal insufficiency, Lipska [/bib_ref].
## Restarting subcutaneous insulin for patients already established on insulin
Conversion to subcutaneous insulin should commence once the patient is able to eat and drink without nausea or vomiting. The pre-surgical regimen should be restarted, but may require adjustment because the insulin requirement may change as a result of postoperative stress, infection or altered food intake. The diabetes spe-cialist team should be consulted if the blood glucose levels are outside the acceptable range (6-12 mmol.l À1 ) or if a change in diabetes management is required.
The transition from intravenous to subcutaneous insulin should take place when the next meal-related subcutaneous insulin dose is due, for example with breakfast or lunch.
## For the patient on basal and bolus insulin
There should be an overlap between the end of the VRIII and the first injection of subcutaneous insulin, which should be given with a meal and the intravenous insulin and fluids discontinued 30-60 min later.
If the patient was previously on a long-acting insulin analogue such as Lantus â , Levemir â or Tresbia â , this should have been continued and thus the only action should be to restart his/her usual rapid-acting insulin at the next meal as outlined above. If the basal insulin was stopped, the insulin infusion should be continued until a background insulin has been given.
For the patient on a twice-daily, fixed-mix regimen The insulin should be re-introduced before breakfast or before the evening meal, and not at any other time. The VRIII should be maintained for 30-60 min after the subcutaneous insulin has been given.
## For the patient on a continuous subcutaneous insulin infusion
The subcutaneous insulin infusion should be recommenced at the patient's normal basal rate; the VRIII should be continued until the next meal bolus has been given. The subcutaneous insulin infusions should not be re-started at bedtime.
## Resumption of normal diet
The key to successful management of the surgical patient with diabetes is resumption of his/her usual diet. This allows resumption of normal diabetes medication. Hospital discharge is only feasible once the patient has resumed eating and drinking.
## Other anaesthetic considerations
## Use of dexamethasone
Dexamethasone can lead to hyperglycaemia and should be used with caution in patients with diabetes [bib_ref] Mechanisms of glucocorticoid-induced insulin resistance, Geer [/bib_ref]. If it is decided that the possible benefits of earlier resumption of normal diet outweigh the disadvantages of hyperglycaemia, it should be ensured that the CBG is measured hourly for at least 4 h after administration.
## Use of regional anaesthesia
Local anaesthetic techniques, when used as the sole anaesthetic, reduce the risk of postoperative nausea and vomiting. Furthermore, when used as part of the multimodal technique in general anaesthesia, they have opioid-sparing actions, with resultant reduction of opioid-related side-effects. However, evidence suggests that patients with diabetes are more prone to epidural abscesses and haemodynamic instability after central neuraxial blockade (in patients with an autonomic neuropathy), and, possibly, increased risk of neuropathy after peripheral nerve blocks [bib_ref] Neurologic complications after neuraxial anesthesia or analgesia in patients with pre-existing peripheral..., Hebl [/bib_ref] [bib_ref] Spinal epidural abscess: a meta-analysis of 915 patients, Reihsaus [/bib_ref].
## Enhanced recovery after surgery
Enhanced recovery of patients undergoing surgery utilises several strategies to promote earlier resumption of normal diet, earlier mobilisation, and reduced length of stay [bib_ref] The enhanced recovery after surgery (ERAS) pathway for patients undergoing major elective..., Varadhan [/bib_ref]. This has particular relevance for patients with diabetes as it promotes eating and drinking (reducing the risk of iatrogenic harm from a VRIII), and by promoting earlier discharge.
Use of oral carbohydrate loading drinks Carbohydrate loading may compromise blood glucose control and should not be used in patients with insulin-treated diabetes who are likely to have a short period of fasting. There is some evidence that oral carbohydrate loading may be safe in patients with type-2 diabetes [bib_ref] Pre-operative carbohydrate loading may be used in type 2 diabetes patients, Gustafsson [/bib_ref] , and if a VRIII is to be used in any case there is no reason to withhold oral carbohydrate loading drinks in any patient with diabetes.
## Patients with diabetes requiring emergency surgery
There are three ways of managing diabetes in the patient requiring emergency surgery:
1 Modification of normal medication. This is only possible if the patient is physiologically well and is being operated on a scheduled list, for example a regular trauma list for minor hand surgery, or evacuation of retained products of conception on a regular gynaecology list. 2 Use of a VRIII. This should be the default technique to manage a patient undergoing emergency surgery because of the unpredictability of the starvation period caused by the erratic nature of emergency lists. 3 Use of a fixed-rate intravenous insulin infusion (see below). This should only be used if the patient requires immediate surgery and has concurrent DKA.
The aim is for the patient to be taken to the operating theatre with a CBG of 6-10 mmol.l À1 (6-12 mmol.l À1 may be acceptable), without overt DKA, and having been adequately resuscitated.
- Recommendation: In patients undergoing emergency surgery, the CBG should be checked regularly (hourly as a minimum whilst acutely unwell), and a VRIII established using dextrose 5% in saline 0.45% with pre-mixed potassium chloride as the substrate.
The patient should be checked for ketonaemia (> 3.0 mmol.l À1 ) or significant ketonuria (> 2+ on urine sticks) if the CBG exceeds 12 mmol.l À1 . If the patient has DKA and requires emergency surgery, the involvement of senior multidisciplinary input from intensivists, anaesthetists, surgeons and diabetologists should be considered to agree optimal peri-operative management. Operating on a patient with DKA carries significant mortality and should be avoided if at all possible. The discussion should include: the requirement for surgery, because DKA can be associated with abdominal pain (negative laparotomies have been reported); the area in which the patient should be resuscitated before theatre; whether saline 0.9% with premixed potassium chloride or Hartmann's solution should be the main resuscitation fluid; whether a VRIII or a fixed-rate infusion should be used to treat DKA in theatre; and the area in which the patient will be recovered after surgery. Recent guidelines and evidence suggest that a fixed-rate intravenous insulin infusion is superior to a VRIII in treating DKA in a ward environment. However, use of a fixed-rate infusion is asso-ciated with hypoglycaemia and requires administration of glucose 20% if the CBG is under 14 mmol.l À1 .
Early involvement of the diabetes inpatient specialist team should be sought.
If possible, long-acting insulins (Levemir, Lantus, Tresiba) should be continued in all patients at 80% of the usual dose. This is to avoid rebound hyperglycaemia when intravenous insulin is stopped.
## Safety
Errors in insulin prescribing are common and insulin has been identified as one of the top five high-risk medications in the inpatient environment. The wide range of preparations and devices available for insulin administration (currently > 60) increases the potential for error. One third of all inpatient medical errors leading to death within 48 h of the error involve insulin administration. Common errors include mistakes when abbreviating the word 'units', failure to use a specific insulin syringe, and errors when preparing insulin infusions. Use of pre-filled syringes of insulin for infusion may reduce the risk and should be considered by Trusts.
## Safe use of vriiis
Variable-rate intravenous insulin infusions are overused in the peri-operative setting. Their use is associated with hypoglycaemia, hyperglycaemia and ketosis on cessation, and hyponatraemia. There seems to be a significant risk of hypoglycaemia in patients with a CBG of 4-6 mmol.l À1 and on a VRIII. Patients often return to surgical wards from theatre with an intravenous insulin infusion in place but no directions for its withdrawal. Hospitals should have written guidelines for both the safe use of the VRIII and conversion from the VRIII to the usual diabetes treatment.
To ensure a steady supply of substrate it is recommended that glucose 5% in saline 0.45% with potassium chloride 0.15% or 0.3% should be administered concurrently with the VRIII, at a rate to meet the patient's maintenance fluid requirements. It is imperative that there is hourly monitoring of the CBG (to keep the CBG at 6-10 mmol.l À1 ), and that in patients with type-1 diabetes, the VRIII is not discontinued until alternative subcutaneous insulin has been administered.
## Metformin
A number of guidelines available for the use of metformin recommend withdrawing treatment peri-operatively. However, evidence for this approach is lacking and there is some evidence that peri-operative continuation of metformin is safe. Metformin is renally excreted and renal failure may lead to high plasma levels which are associated with an increased risk of lactic acidosis [bib_ref] Recent metformin ingestion does not increase in-hospital morbidity or mortality after cardiac..., Duncan [/bib_ref]. Metformin should be withheld if there is pre-existing renal impairment or significant risk of the patient developing acute kidney injury. Anaesthetists should be vigilant about the dangers of the use of nephrotoxic agents, including contract media.
## Non-steroidal anti-inflammatory drugs (nsaids)
There are several additional considerations to the use of NSAIDs in patients with diabetes, including the redistribution of renal blood flow in the presence of hypovolaemia and the risk of oedema, especially if given concurrently with metformin and glitazones.
## Quality control and audit
A number of bodies have published quality standards pertinent to the peri-operative care of patients with diabetes. These include NICE, the Royal College of Anaesthetists (RCoA) and NHS Diabetes (the latter no longer exists and the Joint British Diabetes Societies Inpatient Care Group has taken over some of its roles).
## National institute for health and care excellence
Quality Standard 6, Diabetes in Adults, includes an auditable standard relating to inpatient care: "People with diabetes admitted to hospital are cared for by appropriately trained staff, provided with access to a specialist diabetes team, and given the choice of selfmonitoring and managing their own insulin".
## Royal college of anaesthetists
The RCoA's Guidelines for the Provision of Anaesthetic Servicescontains much advice relating to the perioperative care of patients, including those with dia-betes. In addition, the RCoA publication Raising the Standarda Compendium of Audit Recipes contains details on audit and quality improvement topics pertinent to patients with diabetes. These focus on peri-operative care and are suitable for use by anaesthetists wishing to assess the quality of local provision. For example, topics include availability of guidelines for, and management of, chronic medication during the peri-operative period, and peri-operative fasting.
## Nhs diabetes
The NHS Diabetes guideline contains a number of suggested audit standards covering medication, safety, training and institutional issues [bib_ref] NHS Diabetes guideline for the perioperative management of the adult patient with..., Dhatariya [/bib_ref].
## National diabetes inpatient audit
Whilst the audit standards suggested above may be used within an organisation, national audit projects provide a broader picture and allow benchmarking. The National Diabetes Inpatient Audit (NaDIA) is an annual snapshot audit of surgical and medical inpatients with diabetes. The majority of Trusts in England and Wales are now taking part. The audit is commissioned by the Healthcare Quality Improvement Partnership as part of the National Clinical Audit and Patient Outcomes Programme. The results are available online in the spring of each year and enable organisations to assess their performance in caring for inpatients with diabetes.
Included in the audit are questions relating to whether diabetes management minimises the risk of avoidable complications, harm resulting from the inpatient stay, patient experience and the change in patient feedback on the quality of care since NaDIA began. Of particular relevance to anaesthetists, NaDIA includes comparative data on the prevalence of medication errors, intravenous insulin infusions, glycaemic control and patient harm. In addition, there are considerable data from patients' feedback questionnaires concerning their experience and involvement in their own care.
- Recommendation: Anaesthetists interested in perioperative diabetes management should access the National Diabetes Inpatient Audit and look up the results for their own organisation. |
Neurocognitive Signatures of Naturalistic Reading of Scientific Texts: A Fixation-Related fMRI Study
How do students gain scientific knowledge while reading expository text? This study examines the underlying neurocognitive basis of textual knowledge structure and individual readers' cognitive differences and reading habits, including the influence of text and reader characteristics, on outcomes of scientific text comprehension. By combining fixation-related fMRI and multiband data acquisition, the study is among the first to consider self-paced naturalistic reading inside the MRI scanner. Our results revealed the underlying neurocognitive patterns associated with information integration of different time scales during text reading, and significant individual differences due to the interaction between text characteristics (e.g., optimality of the textual knowledge structure) and reader characteristics (e.g., electronic device use habits). Individual differences impacted the amount of neural resources deployed for multitasking and information integration for constructing the underlying scientific mental models based on the text being read. Our findings have significant implications for understanding science reading in a population that is increasingly dependent on electronic devices.Reading expository texts remains a primary means for students to acquire scientific knowledge. Learning from such texts crucially depends on the reader's ability to construct a mental representation that can maximally capture the knowledge structure (KS) inherent in the text. The text's KS reflects the author's conceptual knowledge associations, and the text KS interacts with the reader's cognitive abilities that together impact the learning outcome of the reader's representation of the scientific knowledge after reading 1,2 . The current study is designed to examine this interaction, specifically how the KS of the text (referred to as textual KS henceforth) interacts with the individual reader's abilities in executive function and his or her reading habits (including electronic device usage). To understand this complex interaction properly, we studied expository science text reading at both the behavioural and the neurocognitive level, combining methods of network analyses of the reading material with statistical analyses of the data collected from self-paced naturalistic reading.Until now, neurocognitive studies of reading comprehension have focused on narrative texts, and the major theories in the field have also been based on analyses of narrative texts 3 . When reading expository texts, the reader's task is to identify the different possible relationships among often quite abstract concepts. These relationships can be correlational, temporal sequential, causal, or hierarchical, and can exist between pairs or clusters of concepts. Comprehension of these relationships in the text (in addition to understanding the meaning of words and facts about the world) is thus key to the reader's success in expository text comprehension 4 .An influential model of reading comprehension, the Construction-Integration model 5,6 , suggests that text comprehension is organized in cycles, roughly corresponding to short sentences or phrases 7,8 . The construction process takes place early in the cycle, in which the reader forms concepts and propositions from the linguistic input. Later in the cycle, the integration process establishes an elaborated propositional representation that is internally coherent and reasonably consistent with the discourse context and with the reader's world knowledge. However, this early construction vs. late integration dual-stage processing view has been recently challenged by 1 the view of parallel integration of information at different levels and scales. For example, Kuperberg and Jaeger 9 proposed that during reading, predictive candidates are activated before the incoming new information is processed (i.e., top-down processing). The predictive pre-activation encompasses multiple levels of representations including syntactic, semantic, phonological, orthographic and perceptual. From the perspective of memory, Hasson et al. 10 argued that all cortical circuits are involved in information accumulation in a hierarchical organization. The primary perceptual-motor systems have short process memory, while the higher order areas such as temporoparietal junction, angular gyrus, and medial prefrontal cortex have long process memory. The primary process areas are modulated by the fronto-parietal network of attentional control, while the higher-order areas by the medial temporal lobe (hippocampal) circuit of binding and consolidation. Therefore, the construction and integration processes might not be temporally dissociable, and instead, it is the brain regions that integrate information at different time scales of memory processes that should and can be empirically differentiated, such as distinct neural networks involved in the integration of local and global contexts 11 .For expository texts, the information integration process encompasses analogous transfer 12,13 or knowledge revision 3 , where updated situation models 14 or mental models 15 are generated. The extent to which the reader generates an appropriate situation model, an integrative mental representation of the text knowledge, depends on the one hand on how the knowledge is conveyed to them (e.g., text properties) and on the other, the reader's cognitive abilities, including abilities to retain information in memory, sustain attention during reading, and formulate abstract concept relations (i.e., reader characteristics). These knowledge-specific and reader-specific characteristics can be examined, as in this study, under the umbrella of textual KS, and executive function and reasoning abilities, respectively.Knowledge Structure as Network MapsTextual KS refers to how concepts/units of information are organized in an expository text 16 . Kintsch and van Dijk first proposed this idea using graphs to represent the network of coherent propositions 8 in texts and Ferstl and Kintsch were among the first to apply network measures to estimate a reader's situation model 17 . Network maps are one common explicit visual representation of KS, which consist of pairs of concepts (represented as nodes) joined by link lines (represented as edges) indicating relationships between pairs of concepts. This type of KS representation is now well established in the literature (see Kinchin et al. 18 for discussion). In this study, we extracted the textual KS as network maps according to Clariana 19 . This process involves several steps as described in the Methods section.Among numerous network metrics that could be derived, centrality has been proposed as one of the most basic and pragmatic ways to describe network maps 20 . The centrality of a node in a network indicates the relative importance of that node in relation to all other nodes, and this measure has been used as a way to quantify the structure or shape of concept maps 21 . For example, Kinchin et al. 18 categorized concept maps in terms of the network topologies of spoke, chain, and net(Fig. 1), and a major discriminating criterion was graph centrality. The spoke type concept map has a large graph centrality value, meaning that one central concept is connected to a large proportion of all other concepts. It represents a KS of simple associations, with no hierarchy and little integration of concepts. On the opposite end of this spectrum is the chain type map, which has a small graph centrality value, representing a sequential KS of isolated conceptual understanding with few associations among the concepts. Such concept maps are susceptible to "meltdown" from a single broken link, and are unlikely to then reorganize. In between these two extreme types is the net type map that has a medium graph centrality value, representing a KS of higher integrity with several levels of hierarchy and with complex interactions between levels. Reorganization of KS by incorporating would-be knowledge is well supported in maps of the net type, and missing links can more easily be compensated with redundant paths. For example, four behavioural studies 22-25 that have considered this relationship have reported an 'inverted U-shape function' (as suggested by Rikers et al. 26 ) for network graph centrality (abscissa, x-axis) and post-test measures (ordinate, y-axis), with the function's maximum agreeing with the network graph centrality of the experts' network. In sum, the shape of the network (spoke, net, chain) is related to the degree of conceptual integrity in the KS, and can be represented by different centrality scores.Applying this logic in this study, we consider network maps with medial graph centrality values to represent near optimal textual KS, whereas maps with (extremely) high or low graph centrality values represent sub-optimal textual KS. Specifically, we use the maximal betweenness centrality (MBC) 27 value, the highest betweenness centrality values of all nodes in a network to describe the characteristic of each textual network map. Also, we use the quadratic terms of the mean-centred/normalized MBC values as a measure of textual KS optimality: higher quadratic centrality values (further away from zero) indicate sub-optimal KS, while lower quadratic centrality values (closer to zero) indicate more optimal KS; the adoption of the quadratic terms is based on several established observations in previous studies[22][23][24][25]. Note that the definition of sub-optimal text here does not automatically mean a 'bad' or 'incoherent' text. Rather, the text KS typology depends on the nature of the domain knowledge; for example, optimal KS here refers to texts that have a KS structure hierarchically organized as central versus peripheral concepts.
## Executive function, reasoning, and text comprehension
Text comprehension results from how executive functions and analogical reasoning are employed by the reader to process the textual information. Executive functions consist of a set of dissociable processes that coordinate cognition and facilitate goal-oriented behaviour. Follmer's 30 meta-analysis showed positive correlations between reading comprehension and the following components of executive function: working memory, shifting, inhibition, and sustained attention and monitoring. In particular, working memory is needed to maintain and update textually relevant information on a constant basis, thereby facilitating the reader's development of a mental representation of the text. In the current study, we assess these four important components of executive functions www.nature.com/scientificreports www.nature.com/scientificreports/ through widely used standardized tests, the 'attention network test'for measuring shifting and inhibition and the 'letter-number sequencing test'for measuring working memory.
Another cognitive ability, analogical reasoning, also significantly affects reading comprehension, although it is less well examined as compared with executive function. In analogical transfer, the existing KS serves as the source or reference, and the newly formed textual KS is the target in the analogical process. For example, in chemistry classes, the solar system is often used as the source/referential analogy when explaining atomic structure (target concept). Analogical reasoning is also involved in reading when readers revise or update existing KS based on the new textual KS through reading comprehension. They compare and detect any inconsistency between the two, and if successful, further convert and incorporate the text information into prior knowledge for future use. In this study, we assess analogical reasoning by using a standardized test, the Raven's Progressive Matrices.
Although no neuroimaging work has examined text comprehension based on the reader's analogical reasoning ability, there is a sizable literature on the neural correlates of analogical reasoning. An aggregated meta-analysis of 7 studiesshowed neural correlates of semantic analogy in left IFG, MFG, frontopolar cortex (FPC), dorsolateral prefrontal cortex (DLPFC), and bilateral caudate heads. In particular, the left FPC is also involved in analogical reasoning of matrix problem tasks (e.g., based on Raven's task) and visuospatial domains. This finding is consistent with the proposal that FPC is critical for integrating the outcomes of separate cognitive operations to facilitate long-term goal oriented behaviour.
## Electronic device and reading
Individual differences also exist in areas other than executive function and analogical reasoning, and in a recent study, Follmer et al.investigated how different reading background variables relate to the individual's reading comprehension of STEM (Science, Technology, Engineering, Mathematics) texts. Using a large sample of Mechanical Turk participants, they showed that STEM text comprehension was negatively correlated with reported frequency of reading on electronic devices (e.g., smartphones, tablets, computers) as well as with reported frequency of non-reading behaviour on electronic devices (e.g., watching television). At the same time, STEM text comprehension was positively correlated with self-reported level of reading attitudes and preferences (e.g., enjoyment of challenging books, learning difficult things via reading). These disturbing findings provided initial evidence of how the emerging electronic reading habits may fundamentally alter readers' comprehension of expository scientific texts.
Previous studies have investigated the effect of paper vs. screen-based reading comprehension (see Sidl et al.for a review), with findings indicating that reading on a screen, as compared with reading on paper, may lead to. (A) Spoke -a radial structure in which all the related aspects of the topic are linked directly to the core concept, but are not directly linked to each other. (B) Chain -a linear sequence of understanding in which each concept is only linked to those immediately above and below. (C) Net -a highly integrated and hierarchical network worse performance (or more reading time for achieving the same level of performance). Such discrepancies have been attributed to aspects of the technology such as visual fatigue and less convenient navigation, and also to the impact of electronic devices on metacognitive processes (e.g., overconfidence and reduced self-regulation and monitoring). However, recent studies have taken personal preferences of platforms into consideration: when not under time pressure, some readers who prefer the electronic platform actually show an effect of screen superiority. In the current study, our focus with regard to the relationship between electronic device and reading will be on individual differences in the habits (and daily duration) of using electronic device, and the effect of these differences on reading comprehension.
## The current study
This study systematically investigates the relationships among executive functions, analogical reasoning, and electronic and non-electronic reading behaviour, and their impact on reading comprehension at both the behavioural and the neurocognitive levels. As previously mentioned, most neurocognitive studies of text comprehension have focused on narrative texts. For example, the Extended Language Network hypothesissuggests that the classic language network, the semantic control and integration network, and the executive function network are simultaneously engaged during narrative text comprehension. Swett et al.was among the first to investigate the neural correlates of expository text comprehension. Consistent with the idea of multiple networks, Swett et al. reported patterns of co-activation in the brain's key regions of cognitive control, visual processing, and language/ semantic integration. Specifically, expository text comprehension also engages the core semantic-processing network for integrating word-and sentence-level semantic information, and additional multi-modal regions that create and update the situation/mental models for the text being read. The authors further reported different patterns for central versus peripheral text concepts, which implies that good readers notice and use the implicit textual KS of the expository text by focusing on the central and peripheral concepts differently (i.e., recruiting different regions of the brain).
In fMRI studies of reading, it is important to know the exact onset time of words and phrases to convolve the hemodynamic response function (HRF) with specific task-related variance and isolate it from unexplained variance. To this end, we employed a paradigm called "fixation-related fMRI" 45 (see Methods for more details). Previous neuroimaging studies of texts dealt with the stimulus timing issue by controlling the presentation rate of the stimuli, typically with individual words, phrases, or sentences shown in a rapid-serial-visual-presentation or RSVP paradigm. But reading every word for half a second in succession of one another is not a natural reading experience. To overcome this problem, we have taken advantage of an emerging paradigm that explores simultaneous eye-tracking and fMRI data acquisition (fixation-related fMRI). With this paradigm, participants are allowed to self-pace materials during reading in the scanner in a more naturalistic manner than reading via RSVP. To match the fast speed of eye-movements and the cognitive processes during reading, we further used the multiband echo-planar imaging (EPI) acquisition techniqueto reduce the fMRI repetition time (TR) to 400 ms, in contrast to the typical TR of 2000 ms used in task-based fMRI studies. Multiband EPI provides greater within-participant statistical power with a higher sampling rate, a higher temporal Nyquist frequency to detect fast oscillatory neurally generated BOLD signals, and better removal of spurious non-BOLD high frequency signal content. By integrating eye-movement and high sampling-rate fMRI data in a naturalistic paradigm, our study is poised to provide neurocognitive insights into naturalistic scientific text comprehension.
To analyse the data collected from fixation-related fMRI, we incorporated a parametric modulator of the index of word position in sentences (starting from 1) in our fMRI GLM analysis. This approach aims to capture the variance in the HRF that changes along the time course of sentential processing across the text. It corresponds to the hypothesis of the Construction-Integration model 5,6 that cycles of text comprehension roughly corresponds to short sentences or phrases. Note that such a regressor, even though it is temporally based, would also capture variances associated with other concomitant cognitive processes which evolve along the time course of sentence reading (e.g., predictive pre-activation at syntactic, phonological, orthographic and perceptual levels 9 ). Neural patterns negatively correlated with this regressor would be more involved in the early stage of sentential processing, which could be associated with the construction phase of the cycle or the integration of local information within the sentence. Neural patterns positively correlated with this regressor would be involved in the late stage of sentential processing, which could be associated with the integration phase, as well as the integration of the sentential information with more global context of the current textual representation or world knowledge. The beta images of this regressor (variance along the time course of reading a sentence) could be further used to investigate the effects of stimuli (e.g., textual KS) and individual differences (e.g., executive function). In a naturalistic reading paradigm such as used in the current study, these concomitant cognitive processes are not dissociable, and they are indeed vital in language comprehension.
Given the approaches reviewed thus far, we make the following hypotheses. First, regarding the effects of textual KS, we hypothesize that when processing expository scientific texts with sub-optimal KS, cognitive demands of executive function should be higher due to the construction of a situation/mental model from the text; as a result, the associated neural correlates will be reflected as stronger activation in the executive control network, including the prefrontal cortex and the cingulate cortex. Second, regarding the effects of reader characteristics and individual differences, we hypothesize that executive function, analogical reasoning, and positive reading attitude will be positively correlated with reading comprehension performances. Neurocognitively, such correlations should be reflected as co-activation in areas including the left IFG, MFG, FPC, dorsolateral prefrontal cortex (DLPFC), and bilateral caudate heads, areas that are critical for executive function, analogical reasoning, and linguistic-semantic integration when processing scientific text.
# Results
Behavioural performances and individual differences. Participants read five expository texts in the scanner. Every participant made at least six correct answers to the 10 multiple-choice assessment questions at the end of each text during in-scanner reading. The accuracy for the questions for each text was as follows (mean% ± SD, n = 51): Mathematics, 94.71 ± 7.84, GPS, 90.98 ± 9.22, Mars, 91.76 ± 9.10, Electric Circuit, 95.10 ± 7.03, and Supertanker, 88.40 ± 11.14. ANOVA showed significant differences among participants' performance accuracy on the texts (F (4,250) = 5.32, p = 0.0004). Specifically, post-hoc Tukey's HSD test showed that performance accuracy differed significantly between Electric Circuit and Supertanker (lower and upper confidence limit = 2.17, 11.94, p = 0.0009) and between Mathematics and Supertanker (LCL = 1.78, UCL = 11.55, p = 0.002).
Participants' mean performance accuracy varied depending on the individual difference scores: it was positively correlated with GSRT scores (n = 49, ρ = 0.65, p < 0.0001), with Raven's score of analogical reasoning (n = 49, ρ = 0.28, p = 0.027), and with reading preference index (n = 49, ρ = 0.27, p = 0.029). Further, the GSRT was also positively correlated with the working memory LNS task (n = 46, ρ = 0.36, p = 0.007) and with Raven's scores (n = 49, ρ = 0.33, p = 0.011). GSRT scores also showed a positive trend though not significant correlation with the reading preference index (n = 49, ρ = 0.23, p = 0.059).
fMRI Results: Main effects and individual differences of integrative processing. Neural correlates of reading (Content Word fixation) were reflected in the strong activity in bilateral visual cortex and medial supplementary motor area (SMA), along with left precentral gyrus, superior and middle temporal gyrus (STG and MTG), anterior temporal lobe (aTL), inferior frontal gyrus (IFG) pars triangularis, and hippocampus,.
Neural correlates of Integrative processing were reflected in two different patterns: the first one, negatively correlated with the word position regressor, was associated with strong activities in bilateral occipital pole, posterior cingulate cortex (PCC), pregenual anterior cingulate cortex (pgACC), as well as left fusiform and precentral gyrus, red). One cluster in the left insula and IFG pars triangularis showed negative correlation between the E-device reading index and Integrative processing (MNI: ;,.
fMRI Results: Main effects and individual differences of KS optimality. After the linearly correlated variance of MBC (maximum betweenness centrality) was partialled out, the quadratic term of MBC represented the optimality of textual KS (with values closer to 0 being more optimal; see Introduction). Neural correlates of the processing of texts with optimal KS revealed strong activity in the left DLPFC and left middle STG, while the processing of sub-optimal KS led to greater activity in the left frontopolar cortex (FPC) and bilateral dorsal ACC ,. Furthermore, left FPC and bilateral SMA were correlated with the processing of sub-optimal KS texts among participants with higher GSRT scores, suggesting an interaction between textual KS properties and reader characteristics (e.g., of high-vs-low reading competence). Finally, this text-reader interaction was also reflected in the regression results of E-device reading index: during processing of sub-optimal KS texts, neural responses in the left temporoparietal junction (TPJ,increased with E-device reading index, while responses in the right claustrumdecreased. These interactions have significant implications for student science concept learning, as discussed below.
# Discussion
The current study investigated the neurocognitive processes underlying the interaction between properties of expository texts and characteristics of the reader, specifically between the textual KS (network structure of the texts to be read) and the individual readers' executive function, reasoning, and reading habits. Our study also showed that readers' electronic device usage is negatively correlated with the involvement of key brain regions for integrative information processing. To our knowledge, this study is the first systematic behavioural and neurocognitive investigation of expository texts of scientific concepts with a naturalistic reading paradigm that combines both fMRI and eye-tracking. First, at the behavioural level, we found that student performance in reading comprehension is correlated with individual differences in executive functions, analogical reasoning, and positive reading attitude. The GSRT general reading ability scores are correlated with analogical reasoning and positive reading attitude, for both in-scanner performance and immediate post-test assessment questions. GSRT scores were also correlated with individual differences in working memory. These patterns are in line with previous studies that have identified the relationships between reading comprehension and executive functionsand between comprehension and reading behaviour.
The relationships among reading comprehension and executive function, reasoning, and reading attitudes are not one-to-one, but are multidirectional and complex. For example, better executive function might lead to superior reading comprehension, and conversely, better reading experience could improve readers' reasoning, attention, and working memory. Readers with a positive reading attitude engage in more reading activity, which leads to more rewarding experiences and in turn more positive reading attitude. Different reader characteristics could also be related to each other: for example, reasoning has been proposed to require working memory capacity in the mental model theory, engaging working memory's underlying executive processes. Finally, reading comprehension performance may be correlated with the student's success in other domain disciplines: reasoning abilities have been found to be predictive of academic achievements in Mathematics, Biology, Physics, History, and English 52,53 . Our findings that scientific text reading comprehension is correlated with individual differences in working memory and analogical reasoning are consistent with these general findings but also more specifically www.nature.com/scientificreports www.nature.com/scientificreports/ demonstrate that individual difference variables impact scientific reading. Although these correlations could have underlying causal relations, the current study was not designed to test causal relationships, which need to be investigated in future studies. Second, at the neurocognitive level, we found dynamic neural correlates of integrative information processing, suggesting a local predictive focus on surface form analysis (visual cortex and fusiform gyrus) and a global predictive focus on semantic, syntactic analysis and integration (frontoparietal network and SMA). Such change in focus across the time course of processing is in line with the different time-scale analysis of text comprehension. At the beginning of reading a new sentence, integration of local information within the sentence takes place, which demanded primary perceptual-motor areas with short process memory. The more the reader proceeds along the sentence, the more the integration of sentential information with global context of the current textual representation or world knowledge takes place, which demanded higher order areas with long process memory, as predicted by models of memory and text comprehension. Thus, temporal ordering and integrative processing may be related at multiple levels and time scales, although the predictive pre-activation hypothesis 9 emphasizes that integrative processing is due to parallel integration rather than staged processing across time.
With regard to the impact of text properties, texts that have optimal textual KS recruit regions associated with linguistic, semantic (IFG and temporal lobe), and integrative processing (DLPFC). Texts with sub-optimal textual KS recruit regions that are critical for dual-tasking, monitoring, and attention (FPC and dACC), suggesting that these texts elicit more effortful processing during mental model construction. Furthermore, reading www.nature.com/scientificreports www.nature.com/scientificreports/ competence (as measured by GSRT scores) is reflected clearly in the processing of texts with sub-optimal KS: high-competence readers activate regions in integrative information processing in the SMA and FPC, as well as regions for linguistic processing in IFG, insula and STG, suggesting the engagement of multiple brain networks for conceptual integration.
Due to the nature of the hemodynamic response function, we used the content word fixation regressor to capture the variance of neural responses throughout text reading. Neural correlates of this regressor included the typical fronto-temporal circuit engaged in language, syntactic and semantic processing (IFG, STG, MTG, aTL), but also the SMA and hippocampus. SMA, including the supplementary eye field (SEF) and the pre-SMA which has traditionally been implicated in motor planning and motor learning. However, in the context of semantic retrieval, Danelli et al.found the SMA, premotor, and left IFG to be involved in both grapheme-to-phoneme and lexical-semantic routes of lexical access. Further, pre-SMA has been proposed to be part of a network including thalamus and caudate nucleus that govern aspects of semantic retrieval of object memories, supported by EEG data. The left SMA is also associated with syntactic processing as shown in a recent meta-analysis. The SMA and pre-SMA activity could be part of the on going predictive pre-activation process across multiple levels during reading comprehension. In addition, Duff and Brown-Schmidt 58 proposed that the hippocampal declarative memory system is a critical contributor to language use and processing because of its capacity for relational binding, representational integration, flexibility, and maintenance. In Hasson et al. 's memory processing hierarchy 10 , the medial temporal hippocampal region would also interact with regions with long process memory, and facilitate binding and consolidation of incoming information with global context and world knowledge. Given these findings in the literature, it is not surprising that SMA and the hippocampus both play crucial roles in expository text comprehension as shown in our current study, since the predictive and integrative processes take place irrespective of the text genre (i.e., narrative or expository).
Augmented by the high-sampling rate (400 ms TR, a Nyquist frequency of 800 ms) of multiband EPI acquisition in our current design (see Methods), the parametric modulator of word position in sentences successfully captured the dynamic change of neurocognitive integrative processes along different time scales during reading comprehension (mean reading time for each sentence = 3.33 ± 0.86 s). Our results indicated that the temporal evolution of integrative processes shifted from relatively shallow, form-oriented and local processing (e.g., involving the occipital cortex and fusiform gyrus) to more global processing that involves semantic retrieval, information integration, and situational/mental model updating that engage the DLPFC, IFG, IPL, and SMA. Previous work based on narrative text reading has implicated the frontoparietal network in situation model building, an integrative mental representation of the text, with a rough division of labour in situation model construction www.nature.com/scientificreports www.nature.com/scientificreports/ (the posterior parietal and anterior temporal regions) and situation model maintenance (frontal regions). Our finding of the dynamic changes at the sentential level, although from scientific rather than narrative text reading, is consistent with the theoretical framework that the situation/mental model is constantly updated as reading comprehension unfolds in time. Such dynamic changes are seen in cognitive domains other than language or reading: for example, Fangmeier et al.showed a similar pattern of shift in neural correlates during different stages of reasoning in which the initial processing of the premise involves occipital and temporal regions, whereas the validation of a given conclusion based on the premise engages the frontoparietal network (DLPFC, IPL, and precuneus).
By modelling the knowledge structure of a text as network maps (e.g., textual KS), we were able to capture the differences in the neural correlates of expository science text reading as a function of text structure. Specifically, the graph-theoretical measure MBC (referred to as graph centrality) of a textual KS network allowed us to represent texts with optimal (network-like maps) vs. sub-optimal (spoke-or chain-like) KS, and such KS differences directly impact the neurocognitive substrates of reading. Previous behavioural studieshave suggested an inverted U-shape function between network graph centrality of knowledge structure and reading comprehension performances. By using the U-shaped quadratic term of knowledge structure as regressor, we found that the processing of optimal KS texts recruits classical language processing brain regions (left M/STG), along with regions that involve situation/mental model construction and information integration (left DLPFC), whereas processing of sub-optimal KS texts engaged activities in the left FPC and bilateral dorsal ACC.
In the context of multitasking research, FPC and ACC have been proposed to serve complementary but dissociable roles in allocating resources for cognitive control of the primary and subgoals/tasks. While ACC has been frequently implicated in language processing (especially conflict monitoring in bilingual speech production), the role of FPC (Brodmann Area 10) has been traditionally linked to a variety of higher-order cognitive functions based on human and primate research. Specifically, FPC has been associated with the ability to hold a primary goal while performing concurrent subgoals, playing an important role in multitasking and multiple resource allocation 61,65-67 , including reasoning and integration of multiple disparate mental relations. Given this role of FPC in integrative processing, it is no surprise that we see it involved in the processing of sub-optimal KS texts that have (1) spoke-like KS, where a core concept is associated with multiple isolated concepts, and (2) chain-like KS, where concepts are serially associated one by one. In these cases, multitasking is required of the reader so as to retain the core concept while processing and integrating multiple isolated sub-concepts across www.nature.com/scientificreports www.nature.com/scientificreports/ the text. Note that the quadratic effect of graph centrality (as measured with MBC) in FPC and ACC in our data cannot be accounted for by its relation with other psycholinguistic variables such as word length or word frequency, although the latter have also been shown to have curvilinear/quadratic effect on both behavioural 69 and neuroimaging correlates of reading. It is important to note that MBC measures of the texts are largely collinear with the text-wise mean values of key psycholinguistic variables such as word frequency, AoA, and word length (see Section Materials in Method). However, in our subject-level regression model, we included both linear and quadratic terms of MBC, and the linear term was included as a covariate of non-interest. Therefore, the confounding linear effects of the psycholinguistic variables were partialled out before the group-level multiple regression.
The impact of electronic device usage is evident in our results. Across all texts, we found a negative correlation between frequency in electronic device usage and BOLD activity in left insula and IFG pars triangularis. The anterior insula is part of the salience network, which responds to the degree of information saliency (and subsequent attention) in a variety of domains including cognitive and emotional processing 72-74 . Sridharan et al.used Grainger Causality to estimate effective connectivity, proposing that the fronto-insular cortex plays a critical and causal role in switching between the central-executive network and the default-mode network. In addition, our data indicate that individuals with higher electronic device usage, on the one hand, have decreased engagement in insula and IFG, and on the other, recruit more left TPJ and less right claustrum when processing texts with sub-optimal KS. The claustrum has the highest connectivity in the brain by regional volume, especially with the frontal lobe and cingulate regions, and it has been proposed to be the "gate keeper" of neural information for conscious awareness 77 . Considering the potential negative effects of excessive daily usage of electronic device (especially texting on smartphones), the neural patterns in our data regarding insula and claustrum, along with the behavioural data of Follmer et al., could point to the readers' reduced or inefficient coordination of cognitive resources and switching between the central executive and default mode networks. At the same time, the result of over-engagement of the TPJ, part of the executive network, might suggest that these same readers required more effortful processing, especially for texts with sub-optimal KS of the spoke or chain types.
Finally, we found that individuals with higher GSRT scores engage the left FPC and bilateral SMA more strongly when reading texts with sub-optimal KS. As discussed above, FPC and SMA may be significant for expository text comprehension given their important roles in multi-tasking, cognitive resource allocation, and visuospatial processing. Our neurocognitive patterns suggest that better reading ability is associated with the engagement of neural substrates responsible for highly integrative cognitive processes as well as for reasoning. By www.nature.com/scientificreports www.nature.com/scientificreports/ contrast, readers who report excessive daily electronic device usage may not activate these critical brain regions for integrative cognitive processing. As discussed in the Introduction, behavioural work on the immediate effect of media (paper vs. screen) has, by and large, shown that excessive use of screen-based devices is associated with lower quality of metacognitive processes. Our findings provided the first neurocognitive evidence that habitual electronic device usage might adversely affect high-level cognitive processing required for scientific text comprehension. Future investigation is needed to identify the causal relationships among reading habits, preferences of media types, metacognitive performances, and expository text comprehension.
# Methods
Participants. Sixty-two right-handed native English speakers were recruited. Seven participants did not finish the first session due to eye-tracker or MR scanner technical issues. One participant was excluded due to very low accuracy (50%) for an in-scanner comprehension test and poor behavioural testing results outside the scanner. One participant was found to be left-handed after the behavioural session, leaving 51 participants aged between 18 and 40 years in our analysis. Eye-tracking data were missing for one participant during one run containing one text, leading to its exclusion for the analysis for KS. Forty-nine out of the 51 participants completed the behavioural testing session, of which only 46 correctly performed the Letter Number Sequencing task. Therefore, behavioural data analysis included 49 participants (23 males, mean age ± SD = 22.69 ± 4.57). fMRI data for neural correlates of Reading and Integrative Processing included 51 participants (24 males, mean age ± SD = 22.67 ± 4.52). Forty-six participants (21 males, mean age ± SD = 22.84 ± 4.63) were included in the fMRI multiple regression models for neural correlates of individual differences in Integrative Processing. Fortyfive participants (21 males, mean age ± SD = 22.47 ± 3.88) were included in fMRI regression models for neural correlates of individual differences in sentential processing of texts with different KS optimality.
All participants had normal or corrected to normal vision, and had no history of mental or neurological disorder. The study was approved by the Pennsylvania State University Institutional Review Board (IRB) and was performed in accordance with the ethical standards described in the IRB. Written informed consent was obtained from all participants before they took part in the study.
# Materials.
Prior to the experiment, five expository texts of STEM contents were modified from previous research stimuli (see Follmerwww.nature.com/scientificreports www.nature.com/scientificreports/ r = 0.91, p = 0.034; frequency, r = 0.89, p = 0.0453; AoA, r = 0.88, p = 0.049; length, r = 0.87, p = 0.053, concreteness, r = 0.88, p = 0.0487). Stimuli were presented using E-Prime 2.0 82 , sentence by sentence onto a screen which was then projected onto a reflective mirror mounted above the participants' eyes in the MRI scanner (see section Eye-tracking Data Acquisition and Processing for details).
## Ks quantified as maximal betweenness centrality (mbc, graph centrality).
Fifteen key terms were selected as nodes from each of the five texts, along with their synonyms and metonyms. The key terms were aggregated from a key-term generating task of a previous Amazon MTurk study of 403 participantsand a key-term generating task of the authors of the current study (with a general overlap of 88%). The edges between the nodes are defined as proximity associations between nodes, operationalized as follows: a forward pass is made through the text without regard to sentence boundaries, and for every key term that is found, it is linked to the immediate previous key term by entering a "1" (binary coding) in a 15 by 15 term proximity array, indicating www.nature.com/scientificreports www.nature.com/scientificreports/ that there is a link (edge) between the two terms. Textual network maps were thus generated with Analysis of Lexical Aggregates Reader (ALA-Reader). Maximal Betweenness Centrality per map/text as a measure of graph centrality (and measure of KS) was calculated using the NodeXL software (Microsoft Inc., 2018). For a node k in a network, its partial betweenness with respect to the other two nodes i and j is defined as the probability that node k falls on a randomly selected path linking nodes i and j. The betweenness centrality value of node k is the sum of the partial betweenness values in respect to all pairs of nodes in the network except for k. Each node in a network has a betweenness centrality value. Note that the betweenness centrality measure depends on the number of nodes in the graph, and the absolute value of MBC per se does not indicate the optimality of KS. In the current study, the lengths of all the texts were made comparable (roughly 300 words), and we used 15 nodes (key concepts) to construct concept maps for all five texts so that the graph centrality values and the range of optimal KS values are also comparable across the texts. To operationalise the optimality of textual KS, the centrality values were normalised and quadratic terms were calculated. Higher quadratic centrality values (further away from zero, which is the average in the normalised distribution) indicate sub-optimal KS, while lower quadratic centrality values (closer to zero) indicate more optimal KS. Procedure. After providing consent, participants underwent a structural MRI scan, followed by a practice session for self-paced reading in the scanner. They were instructed to click a button to advance from one sentence to the next. Each sentence was presented for up to 8 seconds after which the next sentence automatically appeared on the screen. At the end of each text they answered 10 comprehension questions. Once the practice session ended, the participants completed five self-paced reading sessions, during which time simultaneous fMRI and eye-tracking data were collected. On a second visit, which was usually one week after the in-scanner reading session, participants completed a battery of behavioural tests.
Behavioural data collection and processing. In the behavioural session, the Gray Silent Reading Test, Raven's Progressive Matrices, Letter Number Sequencing and Attention Network tests were presented to participants via E-Prime 2.0, and the Reading Background Questionnaire was completed on an internet browser. Detailed information of each test is as below.
Gray Silent Reading Test (GSRT). The GSRT test measures reading comprehension competence. Up to 13 narrative texts were provided the in GSRT, and each text was presented alongside five assessment questions. Adult participants started with Text No. 8 (a text of middle-level difficulty) and were tested downward (e.g., Text No. 7) until the basal was reached (i.e., when all five questions were answered correctly), and upward (e.g., Text No. 9) until the ceiling was reached (i.e., 3 out of 5 answers were wrong). Because all participants were in the same age group (18 and beyond), conversion of scores to quotient according to age groups was not necessary, and the raw scores were used.
Raven's progressive matrices. The Raven's test measures analogical reasoning. In each of the sixty-five tests, a matrix of relations, from which part is omitted, is presented. Subjects have to choose, from a group of six or eight alternatives, the one which completes the matrix. The problems are arranged in five sets, each of which has a distinctive theme: (A) continuous patterns, (B) analogies between pairs of figures, (C) progressive alterations of patterns, (D) permutations of figures and (E) resolution of figures into constituent parts. The first problem in a set is intended to be self-evident, and it is succeeded by twelve problems of increasing difficulty. The testing time was limited to 10 minutes, and the number of corrected trials was used as the score.
Letter number sequencing (LNS). The LNS task measures working memory. The task was adapted from the Wechsler Adult Intelligence Scale (WAIS-III). Participants heard a series of alternating letters and numbers and were asked to recall the numbers first in ascending order and then the letters in alphabetical order. The task began with a set size of two (one letter plus one number) and increased by one for every three trials until a set size of eight was reached. The participants' outputs were corrected for using capital letters (if lower-case letters were the targets) and accidental usage of arrow keys. To properly reflect the difficulty of different items, size-weighted scores were calculated as the summation of correct items' set size. For example, if the participant was correct in three items with the size of two, one item with the size of three, and two items with the size of four, the score will be calculated as 3
[formula] × 2 + 1 × 3 + 2 × 4. [/formula]
Attention network test (ANT). The ANT tests measure the alerting and orienting skills of attention and the inhibitory control ability of executive function. It consisted of a flanker test in which a central arrow was presented with congruent or incongruent flanking arrows, and the participants were asked to give indicate the direction of the central arrow as fast and as accurately as possible. The row of arrows could appear above or below the fixation cross. In some trials before the arrows appeared, one or two asterisks would appear. They could either alert the participants that the arrows will appear soon but without orienting the location of the arrows, or alert them that the arrows will appear soon and direct attention to the correct location (orienting). Three scores were derived according to Fan et al., reflecting the RT differences caused by alerting, orienting, and conflicting manipulations; for example, the higher the conflict effect on RT, the lower the participant's inhibitory control is.
Reading Background Questionnaire (RBQ). Participants were administered 20 questions constructed based on previous research 84,to assess readers' general reading habits and background, using a Google Form. The items asked about participants' reading habits on electronic media (e.g., computers, smartphones), their electronic non-reading behaviour (e.g., time spent texting friends, watching television), and their reading habits (amount www.nature.com/scientificreports www.nature.com/scientificreports/ of time spent on reading), preferences (e.g., enjoyment of types of books, enjoyment of books about different cultures), attitudes towards reading, and reading ability. Items were administered using either a 4-point or a 5-point Likert scale.
Correlational analyses showed significant correlations between E-device reading and non-reading time, and pair-wise correlations among reading time, reading preference and reading attitude/ability (see also s analyses 38 of how these variables impact reading). Exploratory factor analysis yielded two factors: Factor 1 explains 34.45% of variance, including reading preference (loading = 0.90), reading attitude/ability (loading = 0.66) and reading time (loading = 0.64); Factor 2 explains 21.67% of variance, including E-device reading (loading = 0.97) and non-reading time (loading = 0.31). Given these two factors, we simplified the RBQ variables into two scores: E-device reading index (summation of E-device reading and non-reading time) and reading preference index (summation of reading preference, reading attitude/ability and reading time).
## Behavioural data analyses.
To test what cognitive measures contribute to participants' reading comprehension behaviourally, we performed non-parametric correlation tests checking correlations between GSRT or question-answering accuracy with the Raven's scores, LNS scores, the ANT Alerting, Orienting, and Conflict scores, the RBQ E-device reading and reading preference indices. Because the mean accuracy of the performance assessment scores and the GSRT scores violated the assumption of normality (Shapiro-Wilk W test, both ps < 0.01), one-tailed non-parametric Spearman's correlations were used.
Eye-tracking data acquisition and processing. The basic idea of fixation-related fMRI paradigm, as first explored by Marsman et al., is to use self-paced eye-movements to convolve the hemodynamic responses and model the psychological regressors to analyse fMRI data of visual processing. Later studieshave further demonstrated the validity of simultaneous eye-tracking and fMRI paradigms in naturalistic word and text reading. Eye movements were recorded with an Eye-Link 1000 Plus long-range mount MRI eye tracker (SR-Research) with a sampling rate of 1 kHz. The camera was placed at the rear end of the scanner bore, and captured eye movements via a reflective mirror above the head coil. The distance between the camera and the participant's eyes via the reflective mirror was 120 cm. Recording was monocular (from the right eye), and the participant's head was stabilized in the head coil. A 13-point calibration routine preceded the experiment. Before each reading session, a validation procedure is performed, and re-calibration is done when the validation error is larger than 1 degree.
Data adjustment was later performed to address drifting issues caused by the calibration accuracy decline over time. For fixations falling outside (above or below) the range of predefined target regions, manual adjustment was performed using the Data Viewer software. Instead of using auto-adjustment which brings all fixations onto one horizontal line, we performed trial-by-trial correction adjusting all of the fixations in a single try only along the y axis (vertical adjustment) so as to maintain readers' original eye fixation patterns. MRI data acquisition. Data were acquired using a 3 T Siemens Magnetom Prisma Fit scanner with a 64-channel phased array coil. We acquired a MPRAGE scan with T 1 weighted contrast [176 ascending sagittal slices with A/P phase encoding direction; voxel size = 1 mm isotropic; FOV = 256 mm; TR = 1540 ms; TE = 2.34 ms; acquisition time = 216 s; flip angle = 9°; GRAPPA in-plane acceleration factor = 2; brain coverage is complete for cerebrum, cerebellum and brain stem]. After the T 1 , we acquired five functional runs of T 2 * weighted echo planar sequence images [30 interleaved axial slices with A/P phase encoding direction; voxel size = 3 × 3 × 4 mm; FOV = 240 mm; TR = 400 ms; TE = 30 ms; acquisition time varied on the speed of self-paced reading, maximal 306 s; multiband acceleration factor for parallel slice acquisition = 6; flip angle = 35°; brain coverage misses the top of the parietal lobe and the lower end of the cerebellum]. Additionally, we collected a pair of spin echo sequence images with A/P and P/A phase encoding direction [30 axial interleaved slices; voxel size = 3 × 3 × 4 mm; FOV = 240 mm; TR = 3000 ms; TE = 51.2 ms; flip angle = 90°] to calculate distortion correction for the multiband sequences fMRI data preprocessing and analyses. Data preprocessing and analysis were performed in SPM12 v6906 (http://www.fil.ion.ucl.ac.uk/spm). Functional imaging preprocessing consisted of correction of field inhomogeneity artefacts with the HySCO toolbox (Hyperelastic Susceptibility Artifact Correction)using the pair of spin echo sequence images and realignment for motion correction. The structural image was coregistered to the mean functional image, and segmented into grey matter, white matter, cerebrospinal fluid, bone, soft tissue, and air/background to estimate the forward deformation parameters to MNI space. Images were normalized with the 4 th degree B-Spline Interpolation algorithm and further smoothed with a Gaussian kernel of 8 mm full-width-at-half-maximum (FWHM).
In the GLM analysis, the design matrix contained one psychological regressor of interest, the "Content Word" condition, specifying the onsets and gaze durations of first pass fixations and regressions for content words (informed by eye-tracking data). The index of word position in sentences (starting from 1) was incorporated as a parametric modulator of the "Content Word" condition. We also included two psychological regressors of non-interest: "Non-Content Word" and "Instructions": the "Non-Content Word" condition modelled fixations on non-content (function) words and ocular regressions, and the "Instructions" condition modelled two seconds of instructions presented at the beginning of each run. Because of the self-paced reading, all psychological regressors at the first level were subject-specific. Finally, we included six motion parameters and three physiological regressors (white matter, ventricular, and non-ventricular CSF space signal). We then applied a high pass filter with a cut off period of 128 s, and the temporal autocorrelation was accounted for with the FAST option in SPM12. Then, we calculated fixed effects across all runs for each subject. At the group level, two random-effect one sample t-tests (N = 51) were performed for the effects of reading in general (Content Word fixation), and www.nature.com/scientificreports www.nature.com/scientificreports/ Integrative Processing (parametric effect of word positions). We applied peak-level family-wise error (FWE) correction of p < 0.05, minimal cluster size = 5 voxels, for the main effects of both one sample t-tests.
At the group-level, the beta maps of the Integrative Processing obtained at the subject-level were entered into one multiple regression model as the dependent variable (N = 46). The following eight independent variables were included to checked the effect of individual differences: (1) GSRT, (2) Raven's, (3) span-weighted LNS, the (4) Alerting, (5) Orienting, and (6) Conflict effects of the ANT, (7) the RBQ E-device reading index and (8) the RBQ reading preference index. At the whole brain level, we applied cluster-level FWE-correction p < 0.05, using a cluster-defining threshold of p = 0.001.
To further investigate Integrative Processing due to the effects of textual KS (measured as MBC, see Materials in the Methods), the beta maps of Integrative Processing of each text were entered into a subject-level regression model including the linear and quadratic terms of MBC as the independent variable. At the group level, the beta maps of quadratic MBC correlates of the Integrative processes were entered into an one-sample t-test (N = 50) for the main effect and a multiple regression model (N = 45) with the same eight independent variables for individual differences as mentioned before. We applied cluster-level FWE-correction p < 0.05, using a cluster-defining threshold of p = 0.001, for the main effects of the one-sample t-test and for each individual difference in the multiple regression model of MBC.
## Data availability
All behavioural, eye-tracking and neuroimaging data (with personal information de-identified) have been made available on OpenNeuro (https://openneuro.org/datasets/ds001980/versions/1.0.1) and on the PI's lab website (http://blclab.org/). |
Assessment of potential risk levels associated with U.S. Environmental Protection Agency reference values.
The U.S. Environmental Protection Agency (U.S. EPA) generally uses reference doses (RfDs) or reference concentrations (RfCs) to assess risks from exposure to toxic substances for noncancer health end points. RfDs and RfCs are supposed to represent lifetime inhalation or ingestion exposure with minimal appreciable risk, but they do not include information about the estimated risk from exposures equal to the RfD/RfC. We used results from benchmark dose modeling approaches recently adopted for use in developing RfDs/RfCs to estimate the risk levels associated with exposures at the RfD/RfC. We searched the U.S. EPA Integrated Risk Information System (IRIS) database and identified 11 chemicals with oral RfDs and 12 chemicals with inhalation RfCs that used benchmark dose modeling. For assessments with sufficient model information, we found that 16 of 21 (76%) of the dose-response models were linear or supralinear. We estimated the risk from exposures at the established RfDs and RfCs for these chemicals using a linear dose-response curve to characterize risk below the observed data. Risk estimates ranged from 1 in 10,000 to 5 in 1,000 for exposures at the RfDs, and from 1 in 10,000 to 3 in 1,000 for exposures at the RfCs. Risk estimates for exposures at the RfD/RfC values derived from sublinear dose-response curves ranged from 3 in 1,000,000,000 to 8 in 10,000. Twenty-four percent of reference values corresponded to estimated risk levels greater than 1 in 1,000; 10 of 14 assessments had points of departure greater than the no-observed-adverse-effect levels. For policy development regarding management of cancer risks, the U.S. EPA often uses 1 in 1,000,000 as a de minimis risk level. Although noncancer outcomes may in some instances be reversible and considered less severe than cancer, our findings call into question the assumption that established RfD and RfC values represent negligibly small risk levels. Key words: benchmark dose, noncancer, risk assessment. Environ Health Perspect 111:1318-1325 (2003). doi:10.1289/ehp.6185 available via http://dx.doi.org/ [Online 12 May 2003]
Methods for evaluating risks from exposure to toxic substances for noncancer health end points (such as birth defects, respiratory effects, and hepatotoxicity) are based on the theory that there is a threshold below which there is negligible risk of adverse health effects from environmental exposures. At the U.S. Environmental Protection Agency (U.S. EPA), this negligible risk is quantified through the use of reference doses (RfDs) and reference concentrations (RfCs). The RfD or RfC is defined as an estimate of daily or continuous exposure to the human population (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime (U.S. EPA 1999a). The value of the RfD or RfC is derived by determining a point of departure divided by uncertainty factors (UFs), which are used to account for uncertainties in the available studies, such as limitations in the database, variability within humans, and differences in species response (i.e., animal-to-human extrapolation).
The point of departure in environmental health risk assessment is meant to represent the lowest dose within the range of experimental data. In past practices, the point of departure was exclusively based on a no-observedadverse-effect level (NOAEL) or a lowestobserved-adverse-affect level (LOAEL), derived from animal or epidemiologic studies. The NOAEL is the highest dose for which there are no observed statistically or biologically significant increases in the frequency or severity of adverse effects between the exposed population and its control. Similarly, the LOAEL is the lowest dose at which there are statistically or biologically significant increases in the frequency or severity of adverse effects between the exposed population and its appropriate control group. The NOAEL/LOAEL structure, however, does not provide sufficient information to quantify the equivalent risk levels from exposure at the RfD/RfC because there is no estimated risk at the NOAEL or LOAEL itself. Several authors have criticized the use of the NOAEL because of its sensitivity to sample size, the high sampling variability from experiment to experiment, and the inability to use all of the available doseresponse data [bib_ref] Benchmark dose workshop: criteria for use of a benchmark dose to estimate..., Barnes [/bib_ref] [bib_ref] A new method for determining allowable daily intakes, Crump [/bib_ref] [bib_ref] Procedures for calculating benchmark doses for health risk assessment, Gaylor [/bib_ref] [bib_ref] Statistical properties of the NOAEL, Leisenring [/bib_ref]. [bib_ref] Statistical properties of the NOAEL, Leisenring [/bib_ref] have shown that average risk levels associated with the NOAEL may be substantial. The true risk of exposure at the NOAEL can vary from zero to > 20%, depending on the end point, spacing of doses, and numbers of animals used [bib_ref] Statistical properties of the NOAEL, Leisenring [/bib_ref]. In many cases, an adverse effect may not be detected in a critical effect study because of insufficient statistical power.
Although the NOAEL/LOAEL structure does not provide sufficient information to quantify risk levels from exposure [bib_ref] Percentiles of the product of uncertainty factors for establishing probabilistic reference doses, Gaylor [/bib_ref] , the resulting RfDs and RfCs are assumed to be equivalent to negligible or de minimis risks. As a point of comparison, the U.S. EPA has defined 1 in 1,000,000 excess cancer risk as a de minimis risk level for cancer [bib_ref] Application of hazard identification information for pollutants modeled in EPA's Cumulative Exposure..., Caldwell [/bib_ref] [bib_ref] Extending the threshold of regulation concept: de minimis limits for carcinogens and..., Fiori [/bib_ref] , although regulatory actions are sometimes limited to instances where risk exceeds 1 in 100,000.
Over the past several years, the U.S. EPA has been in the process of developing the benchmark dose (BMD), which is derived from modeling the exposure-response data, as an alternative to the NOAEL/LOAEL as the point of departure for noncancer risk assessments. The BMD is the dose that corresponds to a specified level of increased response [the benchmark response (BMR)] compared with background. This dose is calculated by fitting a mathematical model to the dose-response data, which can be continuous or quantal. The BMD method has several advantages over the NOAEL/LOAEL method, including making better use of dose-response information and reflecting sample size more appropriately [bib_ref] Benchmark dose workshop: criteria for use of a benchmark dose to estimate..., Barnes [/bib_ref] [bib_ref] A new method for determining allowable daily intakes, Crump [/bib_ref]. In singlechemical hazard assessments, the BMD allows for consideration of the dose-response over the entire exposure range, and furthermore, when a dose derived from benchmark modeling is used as the point of departure, actual risk levels can be calculated as an alternative to the hazard index (which is typically based on comparisons of human exposures with an RfD or RfC).
In 2000, the U.S. EPA published a draft guidance document on the application of the BMD approach in determining the point of departure for all types of health effects data (U.S. EPA 2000a). Although a BMR of 10% has most often been used by the U.S. EPA in its assessments, it is anticipated that 5% or 1% would be a more appropriate response level for some end points and designs. Furthermore, in these draft guidelines, a lower statistical confidence limit on the BMD (BMDL) is specifically proposed as a replacement for the NOAEL/LOAEL in setting the point of departure, which is used to determine acceptable levels of human exposure to environmental toxicants. A lower confidence limit is placed on the BMD to obtain a dose (BMDL) that assures with high confidence (e.g., 95%) that the BMR is not exceeded (U.S. EPA 2000a). In addition to ensuring an added measure of protection, this process rewards better experimental design and procedures that provide more precise estimates of the BMD.
Most new and revised RfDs and RfCs in the U.S. EPA Integrated Risk Information System (IRIS) assessments are based on BMD modeling. Certain health end points, however, are not amenable to modeling, and the NOAEL/LOAEL approach will continue to be used in some cases (U.S. EPA 2000a). For this article, we have reviewed and synthesized currently available risk assessment information on the chemicals for which U.S. EPA reference values are based on BMD modeling.
We estimate the equivalent risk levels expected from hypothetical human exposures at established RfD and RfC values using the BMD dose-response information to investigate whether these levels represent negligible risks, and to underscore some of the potential strengths of using benchmark modeling in environmental health risk assessment.
# Methods
We searched the U.S. EPA IRIS database (U.S. EPA 2000b) to identify the chemicals for which current regulatory reference values relied on BMD modeling. We identified 11 chemicals with RfDs based on oral BMD values, and 12 chemicals with RfCs based on benchmark concentration (BMC) values. The BMD and Environmental Health Perspectives - VOLUME 111 | NUMBER 10 | August 2003 Abbreviations: ADJ, adjusted for duration of exposure; BMD 10 , BMD that equals a BMR of 10%; BMDL 10 , lower confidence limit on the BMD 10 ; BMD 05 , BMD that equals a BMR of 5%; BMDL 05 , lower confidence limit on BMD 05 ; C max , peak blood concentration; EGBE, ethylene glycol monobutyl ether; HED, human dose of an agent that is believed to induce the same magnitude of toxic effect as the experimental animal species concentration or dose (this adjustment may incorporate toxicokinetic information on the particular agent, if available, or use a default procedure, such as assuming that daily oral doses experienced for a lifetime are proportional to body weight raised to the 0.75 power). a The reported number of significant figures is not standardized in IRIS. b Point of departure. c RfD derived from the LOAEL (1,000-fold UF) = 0.01 mg/kg/day; BMD-based RfD (100-fold UF) = 0.01 mg/kg/day. d Oral BMDL was derived from the BMCL (8.2 mg/m 3 ) by route-to-route extrapolation with the assumptions that inhalation absorption was 50% and oral absorption was 100% in the dose range near the BMC; BMDL (adjusted for continuous exposure) = (8.2 mg/m 3 × 20 m 3 /day × 0.5)/70 kg = 1.2 mg/kg/day; former RfD derived from the LOAEL (1,000-fold UF) = 0.001. e HED was calculated as follows: using C max for 2-butoxyacetic acid as the dose metric, the BMDL 05 was determined to be 64 µM; physiologically based pharmacokinetic modeling was used to "back-calculate" a human equivalent dose of 5.1 mg/kg/day, assuming that rats and humans receive their entire dose of EGBE from drinking water over a 12-hr period each day. f RfDs were based on the NOAEL and BMD 10 values adjusted for duration (e.g., BMD × 5/7 days); the RfD in IRIS is derived from the NOAEL (3,000-fold UF) = 0.02 mg/kg/day; prospective RfD derived from BMD (3,000-fold UF) = 0.03 mg/kg/day. g BMD was based on a benchmark response of a 1-standard-deviation change from the control mean. Abbreviations: EGBE, ethylene glycol monobutyl ether; ELE, effect level extrapolation factor. a Interspecies extrapolation, intraspecies differences (human variability), subchronic-to-chronic extrapolation, database deficiencies, and ELE. b UFs assigned a value of 10 1/2 were rounded to 3. BMC values and other risk assessment information for these chemicals are presented in [fig_ref] Table 1: Risk assessment information for chemicals with oral RfDs derived from benchmark modeling [/fig_ref]. We have included risk assessment information for naphthalene, a chemical with an IRIS assessment containing an established RfD based on a NOAEL (RfD = 0.02 mg/kg/day) as well as a prospective RfD based on BMD modeling (RfD = 0.03 mg/kg/day) (U.S. EPA 1998i).
We determined whether the principal study identified for each chemical's IRIS assessment was derived from quantal (dichotomous) or continuous critical effect data [fig_ref] Table 2: End points and UFs for chemicals with oral RfDs derived from benchmark... [/fig_ref].
For compounds with BMD or BMC values based on quantal data, the BMR is expressed in terms of a percent increase in risk of adverse outcome compared with background. For compounds with BMD or BMC values based on continuous data, the BMR may be expressed as a percent change in mean response compared with control (e.g., immunosuppression with tributyltin oxide) or in terms of a 1 standard deviation change from the control mean response (e.g., decreased lymphocyte count with benzene). The BMR is a response level used to define a BMD, which is used as the point of departure, from which an RfD or RfC can be developed. The BMR is typically set at the lower end of the range of responses (e.g., 10% or 5% change) that can be detected experimentally. This can help to avoid uncertainties associated with low-dose extrapolation using models that may not reflect biologic realities [bib_ref] A new method for determining allowable daily intakes, Crump [/bib_ref].
Using the benchmark modeling information described above from the IRIS assessments of 19 chemicals, we estimated the equivalent risk levels expected from hypothetical human exposures at the established RfDs and RfCs. We also evaluated whether each of the models used for BMD calculations was linear, sublinear, or supralinear in the observed dose range 1320 VOLUME 111 | NUMBER 10 | August 2003 - Environmental Health Perspectives Abbreviations: ADJ, dose that has been adjusted for duration of exposure; BMC 10 , BMC that equals a BMR of 10%; BMCL 10 , lower confidence limit on the BMC 10 ; BMC 05 , BMC that equals a BMR of 5%; BMCL 05 , lower confidence limit on BMC 05 ; HEC, human equivalent concentration [the human concentration (for inhalation exposure) of an agent that is believed to induce the same magnitude of toxic effect as the experimental animal species concentration or dose; this adjustment may incorporate toxicokinetic information on the particular agent, if available, or use a default procedure, such as assuming that daily oral doses experienced for a lifetime are proportional to body weight raised to the 0.75 power]; MVh, human ambient default minute volume; MVho, human occupational default minute volume; RDDR, regional deposited dose ratio (the ratio of the regional deposited dose calculated for a give exposure in the animal species of interest to the regional deposited dose of the same exposure in a human; this ratio is used to adjust the exposure effect level for interspecies dosimetric differences to derive a human equivalent concentration for particles . To estimate the risk level of the derived RfDs and RfCs, which have a linear or supralinear dose-response curve at the BMD/BMC, we assumed that the doseresponse curves for these compounds are linear at doses below the BMD or BMC . For two assessments, sufficient information was not available to determine the shape of the dose-response curve (carbon disulfide and 1,1,1,2-tetrafluoroethane), and we assumed linearity [this assumption did not overly bias our results because linearity or supralinearity was the shape of the dose-response curve in approximately three-quarters (16 of 21) of cases in which the shape of the dose-response curve was discernable]. This assumption of linearity at the relevant part of the dose-response curve is necessary to extrapolate equivalent risk levels from U.S. EPA reference values derived from BMD modeling, and it is consistent with methods proposed in the U.S. EPA draft cancer risk assessment guidelines (U.S. EPA 1999d). Such risk-level extrapolation is not possible using the NOAEL/LOAEL approach. Methods for risk-level estimation varied between reference values based on quantal end points and those based on continuous end points. For BMD/BMC values derived from quantal critical effect data, we estimated risk from exposure at concentrations equal to established RfD and RfC values by extrapolating linearly from the point represented by the BMR at the BMDL/BMCL to the established RfD and RfC values . We divided the risk at the BMR by the composite UF for those BMD models that were linear or supralinear. For example, to estimate risk from exposure to chloroform's RfC, we divided the estimated risk level at the point of departure (1 in 10 for BMR = 10%) by the composite UF of 100, to arrive at a risk of 1 in 1,000. For BMD/BMC values derived from continuous critical effect data (normally distributed), a change in response of 1 standard deviation from control is considered roughly equivalent to a 10% increase in risk of adverse response from exposure (e.g., benzene's BMR = change of 1 standard deviation in lymphocyte count compared with control mean) (U. . Therefore, when data quality and distribution allowed, we treated the dose that resulted in a 1-standard-deviation change from control as equivalent to BMD 10 /BMC 10 (BMD that equals a BMR of 10%/BMC that equals a BMR of 10%) values derived from quantal data. For assessments based on sublinear dose-response curves, we estimated risk of exposure at the RfD/RfC dose levels by extrapolating the BMD model response function to the RfD/RfC (i.e., using the BMD model, we estimated risk by putting the exposure equal to the RfD/RfC in the model).
For all chemicals in our assessment group with adequate data, we calculated the ratio of the central estimate (BMD or BMC) to the lower statistical confidence limit on the benchmark dose (BMDL) or concentration (BMCL) [fig_ref] Figure 3: Estimated risk levels from exposure at the RfC based on the BMC... [/fig_ref]. This ratio (e.g., BMD/BMDL) provides a metric to compare the relative impact on estimated risk levels resulting from the selection of the BMD/BMC versus the BMDL/BMCL as the point of departure. Abbreviations: EGBE, ethylene glycol monobutyl ether; NA, not available. a Models were defined for an observed range of experimental data. b BMCs were obtained using both Weibull and linear models; the models gave similar goodness of fit to the data and BMC estimates. c Linear dose-response curve because assessment is based on a polynomial model with β i = 0 for i > 1. d Model restricted to not allow supralinear forms. e Essentially linear dose-response curve because squared coefficient term of second-degree polynomial model is insignificantly small (β 2 ≅ 0). f Shape of dose-response curve determined supralinear based on visual inspection; because slope parameter is small, curve approaches linear at low doses. g Shape of dose-response curve determined supralinear based on visual inspection; BMD model response function was not available. h Shape of dose-response curve information was not available; assumed linear or supralinear.
# Results
We found that 13 out of 23 (57%) of the BMD and BMC values were derived from quantal versus continuous data. A 10% additional risk or 10% change from control mean response was selected as the BMR for 17 of the 23 assessments, and a 5% BMR was selected for 3 of the remaining 6 assessments [fig_ref] Table 6: Estimated risk levels from exposure at the RfD [/fig_ref] [National Toxicology Program (NTP) 1993; U.S. . The BMR values for benzene's BMC and BMD and for phenol's BMD were based on a 1-standard-deviation change in acute lymphocyte count and maternal body weight, respectively, compared with the control mean . Of the 21 BMD and BMC values for which sufficient dose-response information was available, we found that 16 (76%) were derived from dose-response data fitted to linear or supralinear models in the observed dose range [fig_ref] Table 5: [National Research Council [/fig_ref]. Seven assessments were based on linear models (two linear models; three polynomial models with β i ≅ 0 for i > 1; two power models with k ≅ 1), nine were based on supralinear models, and five were based on sublinear models. Sufficient information was not available to determine the shape of the fitted model for two assessments, carbon disulfide and 1,1,1,2-tetrafluoroethane (i.e., the response function and model parameters were not provided) .
We calculated 17 RfD and RfC equivalent risk levels (for 14 compounds) assuming linear dose-response curves. These risk estimates ranged from 1 in 10,000 to 5 in 1,000 for the oral route of exposure for compounds with RfDs based on BMD values, and from 1 in 10,000 to 3 in 1,000 for inhalation for compounds with RfCs based on BMC values . [fig_ref] Figure 2: Estimated risk levels from exposure at the RfD based on the BMD... [/fig_ref] EGBE, ethylene glycol monobutyl ether. a Because IRIS assessment is based on sublinear dose-response curve, linearity was not assumed for low dose extrapolation, and risk estimate was derived from the BMD model response function. b BMR is expressed in terms of the standard deviation (in the absence of a clear definition for an adverse effect for this continuous end point, a default BMR of 1-standard-deviation change from the control mean was selected). This default definition of a BMR for continuous end points corresponds to an excess risk of approximately 10% for the proportion of individuals below the second percentile (or above the 98th percentile) of the control distribution for normally distributed effects. Benzene's oral BMDL was derived from the BMCL (8.2 mg/m 3 ) by route-to-route extrapolation with the assumptions that inhalation absorption was 50% and oral absorption was 100% in the dose range near the BMC. c BMR represents a 5% increased risk of neuropsychologic impairment compared to background. d Insufficient data available to estimate risk level (i.e., the response function, underlying distribution of the end point, or mean response and standard deviation of the treatment group and controls were not provided). Naphthalene's RfD in IRIS derived from the NOAEL (3,000-fold uncertainty factor) = 0.02 mg/kg/day; Tthe prospective RfD derived from the BMD (3,000-fold uncertainty factor) = 0.03 mg/kg/day. EGBE, ethylene glycol monobutyl ether. a Because IRIS assessment is based on sublinear dose-response curve, linearity was not assumed for low dose extrapolation, and risk estimate was derived from the BMC model response function. b BMC was based on a BMR of 1-standard-deviation change from the control mean (in the absence of a clear definition for an adverse effect for this continuous end point, a default BMR of 1-standard-deviation change from the control mean was selected). This default definition of a BMR for continuous end points corresponds to an excess risk of approximately 10% for the proportion of individuals below the second percentile (or above the 98th percentile) of the control distribution for normally distributed effects. c A 10% relative change was selected as an appropriate BMR for the nerve conduction velocity measurements because this level is about equal to a difference of 1 standard deviation from the control, and because a change of about 10% would likely raise concern in a clinical setting. d Insufficient data available to estimate risk level (i.e., the response function, underlying distribution of the end point, or mean response and standard deviation of the treatment group and controls were not provided).
logarithmic scale for these chemicals with linear or supralinear dose-response curves at the BMD/BMC. For four RfD and RfC equivalent risk levels (for 3 compounds), we used the sublinear dose-response model to calculate risk at the RfD or RfC. These risk estimates ranged from 3 in 1,000,000,000 to 1 in 100,000 from the oral route of exposure for compounds with RfDs based on BMD values, and from 3 in 1,000,000 to 8 in 10,000 from inhalation for compounds with RfCs based on BMC values [fig_ref] Table 6: Estimated risk levels from exposure at the RfD [/fig_ref]. Five of 21 reference values (24%) reviewed for this assessment corresponded to estimated risk levels greater than 1 in 1,000. Insufficient information was available to estimate risk from exposure at two reference values that were based on continuous response data, chromium VI particulates and naphthalene (i.e., the response function, underlying distribution of the end point, or mean response and standard deviation of the treatment group and controls were not provided). [fig_ref] Figure 4: Distribution of estimated risk levels from human exposures at established RfD and... [/fig_ref] presents the distribution of estimated risk levels from human exposures at established RfD and RfC values for compounds with a linear or supralinear dose-response curve at the BMD. Risk estimates for four assessments derived from sublinear dose-response curves are presented in [fig_ref] Table 6: Estimated risk levels from exposure at the RfD [/fig_ref].
Among the chemicals for which the RfD was based on a BMD, the BMD/BMDL ratio ranged from 1.3 to 3.0. Among the chemicals for which the RfC was based on a BMC, the BMC/BMCL ratio ranged from 1.1 to 2.3 [fig_ref] Table 1: Risk assessment information for chemicals with oral RfDs derived from benchmark modeling [/fig_ref]. Thus, using the central estimate of the BMD or BMC (maximum likelihood estimate) instead of the lower statistical confidence limit (BMDL or BMCL) as the point of departure would result in a 1-to 3-fold difference in the estimated risk levels [fig_ref] Figure 2: Estimated risk levels from exposure at the RfD based on the BMD... [/fig_ref].
The effect level extrapolation factor (ELE) is an UF analogous to the LOAEL-to-NOAEL extrapolation factor. ELEs were applied in the assessments of three compounds, 1,3-butadiene (RfC), benzene (RfC and RfD), and ethylene glycol monobutyl ether (RfC) (Tables 2 and 4) . Thus, no ELE factor was assigned for 16 of 17 assessments that were based on a BMR of 10%.
When we compared the points of departure (i.e., BMDL/BMCL values) with the NOAELs, we found that the points of departure were higher than the NOAELs in 10 of the 14 studies with identified NOAELs [fig_ref] Figure 5: Ratio of the points of departure [/fig_ref]. The BMDL values were up to 4.6 times higher than the empirical NOAELs (range, 0.5-4.6), and the BMCL values were up to 3.9 times higher than the empirical NOAELs (range, 0.2-3.9) [fig_ref] Table 1: Risk assessment information for chemicals with oral RfDs derived from benchmark modeling [/fig_ref]
## Pod < noael pod > noael
# Discussion
To determine whether U.S. EPA reference values represent negligibly small risk levels, we reviewed and synthesized currently available risk assessment information on chemicals for which established RfD and RfC values are based on BMD modeling. For RfDs and RfCs derived from linear or supralinear doseresponse curves, our risk estimates ranged from 1 in 10,000 to 5 in 1,000 for the oral route of exposure, and from 1 in 10,000 to 3 in 1,000 for inhalation. Risk estimates for RfDs and RfCs derived from sublinear dose-response curves ranged from 3 in 1,000,000,000 to 8 in 10,000. Twenty-four percent of reference values reviewed for this assessment corresponded to estimated risk levels greater than 1 in 1,000. The estimated risk of exposure to 1,1-dichloroethylene at its RfC, for example, corresponded to a 3-in-1,000 risk of adverse effect [liver toxicity (fatty change)]. The BMD methodology is the first step in the development of continuous risk functions that can be used to estimate risks at different exposures rather than using an RfD/RfC approach, which has limited use in the decision-making process. For example, BMD and BMC values that are based on the same level of adverse response (e.g., BMR = 10%) can be used to rank the potential hazard of exposure to multiple toxicants. Further application of BMD models, such as has been done here, can be used for estimating adverse noncancer health outcomes from different exposures for other risk-ranking exercises, regulatory policy development, and cost/benefit analyses.
The U.S. EPA used a variety of fitted models to calculate the BMD/BMC values found in IRIS (e.g., K power, linear, quantallinear, exponential polynomial, and Weibull). To compare RfD and RfC equivalent risk levels, we assumed that the dose-response curves for the chemicals in our assessment group are linear at doses below the point of departure. This assumption could have resulted in both underestimates and overestimates of risk. In the case of a supralinear dose-response curve at low doses, for example, this assumption may have resulted in an underestimate of risk. Among the chemicals we reviewed, 9 of the 21 assessments with sufficient information to determine the shape of the dose-response curve were based on supralinear functions. In the case of a sublinear dose-response, this assumption may have resulted in a marked overestimate of risk. For the 5 assessments based on sublinear dose-response curves, therefore, we did not assume linearity for low dose extrapolation and risk estimation.
We believe that the assumption of linearity in the relevant part of the dose-response curve is justified and useful to compare risk levels among this group of compounds. Seventy-six percent of BMD and BMC values considered in this assessment were derived from dose-response data fitted to linear or supralinear models. Furthermore, the range of extrapolation for the RfD/RfC calculations was not large among this group, with most based on points of departure extrapolated to 2 orders of magnitude or less (7 were extrapolated to 1 order of magnitude, 13 were extrapolated to 2 orders of magnitude, and 3 were extrapolated to 3 orders of magnitude). The average and median composite UFs among the compounds in our assessment are 340 and 100, respectively. This implies that even if the dose-response curve for a particular compound is not strictly linear at much lower doses, we could expect the potential impact on the risk estimate to be relatively small. Current U.S. EPA methodology for reference value derivation assumes that the established RfDs/RfCs represent negligibly small risk levels. For assessments that have linear dose-response curves, the extrapolation from the point of departure is typically 2 orders of magnitude or less. Therefore, for the RfD/RfC values to represent risk levels that are below regulatory concern, the dose-response curve would have to drop off sharply after the point of departure. This assumption seems unlikely, especially given our finding that a large number of the assessments we reviewed (9 of 21) were based on supralinear dose-response functions. Although this supralinearity may carry significant implications for risk assessment, more research is needed to determine whether these dose-response relationships remain supralinear at very low doses. On the other hand, assessments based on dose-response curves that are not monotonic may have sublinear or stepwise relationships below the observed data.
Using the BMDL or BMCL as the point of departure in the risk assessment of noncarcinogenic compounds rather than the BMD or BMC central estimate is generally characterized as a conservative assumption (in the health-protective sense). We found that using the central estimate of the BMD (maximum likelihood estimate) instead of the lower bound estimate as the point of departure results in a 1-to 3-fold difference in the risk estimates. According to the U.S. EPA draft BMD guidelines (U.S. EPA 2000a), a lower confidence limit is placed on the BMD to obtain a dose (BMDL) that assures with high confidence (e.g., 95%) that the BMR is not exceeded. This process of using the BMDL rewards better experimental design and procedures that provide more precise estimates of the BMD, resulting in tighter confidence intervals and thus BMDLs that are closer to the central estimate. Our results suggest that the current practice of using the statistical lower bound estimate versus the maximum likelihood estimate as the point of departure is reasonable and does not substantially bias the risk estimate.
For carcinogens, the U.S. EPA typically develops a linear estimate of the slope of the dose-response curve, under the assumption that the curve is linear at low doses. This allows for quantification of risk at any given level of exposure. The U.S. EPA has defined 1 in 1,000,000 excess cancer risk as a de minimis risk level for cancer [bib_ref] Application of hazard identification information for pollutants modeled in EPA's Cumulative Exposure..., Caldwell [/bib_ref] [bib_ref] Extending the threshold of regulation concept: de minimis limits for carcinogens and..., Fiori [/bib_ref] , although regulatory actions are sometimes limited to instances where risk exceeds 1 in 100,000. Among compounds in IRIS with RfDs and RfCs based on BMD modeling, however, we found risk estimates as great as 5 in 1,000. Although noncancer outcomes may in some instances be reversible and considered less severe, this finding calls into question the assumption that noncancer RfD and RfC values represent "acceptable levels" of exposure. In addition, some of the noncancer health end points considered in this assessment are severe and irreversible events, for example, ovarian atrophy (1,3-butadiene) and developmental neuropsychologic impairment (methylmercury), highlighting the need for a renewed discussion within the public health community about what should be considered an "acceptable level" of risk from exposure to toxicants with noncancer health end points.
Most of the BMDLs and BMCLs used as points of departure in IRIS are based on 10% BMRs, many with values higher than the empirically derived NOAELs. This research should help inform discussions about whether this level of BMR is adequately protective of the public health, and whether human exposures at concentrations equal to the resulting reference values do in fact represent negligibly small risk levels.
[fig] Figure 2: Estimated risk levels from exposure at the RfD based on the BMD and the BMDL derived from linear or supralinear dose-response curves. EGBE, ethylene glycol monobutyl ether. [/fig]
[fig] Figure 3: Estimated risk levels from exposure at the RfC based on the BMC and the BMCL derived from linear or supralinear dose-response curves. EGBE, ethylene glycol monobutyl ether. a BMC central estimate not provided by the U.S. EPA. [/fig]
[fig] Figure 4: Distribution of estimated risk levels from human exposures at established RfD and RfC values for reference doses derived from linear or supralinear dose-response curves. a Risk estimates for assessments based on sublinear dose-response curves are not included. [/fig]
[fig] Figure 5: Ratio of the points of departure (PODs) to empirically derived NOAELs. PODs based on bench- [/fig]
[table] Table 1: Risk assessment information for chemicals with oral RfDs derived from benchmark modeling. a [/table]
[table] Table 2: End points and UFs for chemicals with oral RfDs derived from benchmark modeling. [/table]
[table] Table 5: [National Research Council (NRC) 2000; U.S. [/table]
[table] Table 3: Risk assessment information for chemicals with RfCs derived from benchmark modeling. a [/table]
[table] Table 4: End points and UFs for chemicals with RfCs derived from benchmark modeling.Abbreviations: EGBE, ethylene glycol monobutyl ether; ELE, effect level extrapolation factor. a Interspecies extrapolation, intraspecies differences (human variability), subchronic-to-chronic extrapolation, database deficiencies, and ELE. [/table]
[table] Table 6: Estimated risk levels from exposure at the RfD. [/table]
[table] Table 7: Estimated risk levels from exposure at the RfC. [/table]
|
Anapole nanolasers for mode-locking and ultrafast pulse generation
## Reviewer #3 (remarks to the author):
This is an interesting paper reporting on nanoscale lasers based on tightly-confined anapoles. The authors show on the basis of semiclassical simulations that is it possible to engineer a nanolaser running on anapole modes. In a design using conventional InGaAs nanodisks they suggest on-chip sources to efficiently couple light into waveguide channels with high intensity and a route to generate ultrafast sub-picosecond pulses. The paper is well written and merits being considered for publication. However, before publication, the authors are required to sufficiently address the following points:
(1) The laser dynamics seems to be dominated by an anapole mode. What is the radiating (farfield) mode? Which pattern will be coupled out? Is there a quadupolar component?
(2) The periods of pulses in and are 190fs (seemingly from simulations) and 1.65ps (experimentally measured). Why? This analysis is very important is very important for an understanding of the process and the authors should provide an explanation to fully justify their experimental observation.
(3) shows a 3 -level diagram, but the authors state that they used a 4-level model in the method. Which is correct? Moreover, further details should be provided about the method beyond (a) simple reference(s).
(4) The authors should discuss and cite a number of recent papers reporting on (the spatiotemporal dynamics of) nano-lasing and amplified spontaneous emission on and with non-radiating ('dark') near-field modes in systems such as in -Nano-fishnet metamaterials:
## Response to reviewers' comments
and a summary of the changes made in the manuscript Referee #1:
Paper: Juan S. Gongora, Andrey E. Miroshnichenko, Yuri S. Kivshar and A. Fratalocchi, "Anapole nanolasers for mode-locking and ultrafast pulse generation", Nature . This paper reports a novel type of laser based on a scattering-free anapole (superposition of electric and toroidal dipoles) in a gain medium. The authors claimed that the anapole nanolaser can enable efficient coupling to waveguides and a new mechanism of modelocking for ultrafast laser pulse generation. In general, the anapole nanolaser concept is very interesting and can potentially lead to many exciting applications. I think the idea is sufficient to warrant publication if the authors convincingly demonstrate its unique capability compare to other nanolaser systems. However, I am not yet fully convinced that the theory and simulation results have shown that. The following are my specific comments:
Based on my understanding of the paper, anapole nanolasers are unique because stimulated emission occurs at a scattering-free state which does not provide any optical feedback. Can the authors clarify how it is conceptually different from other systems such as a laser that relies on the dark mode, or bound state in the continuum (BIC)?
## Our reply:
We thank Referee #1 for the positive assessment, as well as for many comments that helped us to significantly improve the outreach of this work. There are many differences between an anapole laser and lasers relying on dark modes or BIC states, as we summarize in the following.
Nanolasers based on dark modes are related to resonance effects, as discussed in the following articles for nonlinear metamaterials and SPASERs: .
In an ideal dark laser, dark states are uncoupled to radiation and non-radiating transitions are necessary to transfer energy from the optical amplifier to the dark resonance, as discussed in the original concept of SPASER introduced in [A2] and in the metamaterial laser theoretically analyzed in [A1]. Non-radiating transitions typically require the presence of plasmonic resonances in (most of the cases noble) metals. These resonances, such as localized dark plasmon polaritons, allow for non-radiative energy coupling from an optical amplifier to a dark plasmon.
Contrary to a dark-mode, the anapole is a non-resonant state of the system, and it is typically associated to a strong minimum in the scattering from the material. The anapole is generated by the destructive interference of two radiating components --an electric dipole and a toroidal dipole--that exist at the same frequency and that cancel each other in the far field. Unlike a darkresonance, the anapole state allows for radiative mechanisms of direct excitation, as discussed in Ref.
[16] of the revised paper:
[16] Wei, L., Xi, Z., Bhattacharya, N. & Urbach, H. P. Excitation of the radiationless anapole mode. Optica 3, 799 (2016).
In the anapole laser, the energy down-converted by stimulated emission at the anapole frequency directly excites the electric and toroidal dipoles that interfere destructively, generating a radiation-less state that does not coexist with additional radiating components and that is observed in a purely dielectric structure. This is a unique working principle of this laser that is not observed with other types of systems relying on resonance effects and/or non-radiative transitions. To clarify this statement quantitatively, we analyzed in more detail the profile of the energy emitted from the anapole laser ( To quantify the degree of similarity between our laser emission and the ideal anapole state, we computed the cross-correlation coefficients of their spatial distributions. More specifically, we compared the ideal anapole of [fig_ref] Figure a2: shows the multipole decomposition for an ideal anapole in InGaAs in the... [/fig_ref] and the electromagnetic distribution shown in of the main text, which was calculated considering a real InGasAs nanodisk laser. The cross-correlation coefficient is 96.1%, which shows that the anapole laser allows to create a practically ideal radiation-less anapole state, without the need of any external radiating component, but just destructive interference of specific multipoles that are suitably excited by stimulated emission. The high cross-correlation value observed in our numerical experiments originates from the fact that, as shown in [fig_ref] [ A1 ]: Wuestner, Sebastian and Hamm, Joachim M [/fig_ref] , the presence of losses diminishes (of almost the same factor) the energy of all multipolar components and increases the absorption scattering cross-section of the nanodisk. This implies that the energy difference that is not coupled to ED and TD dipolar components is absorbed inside the nanodisk, without altering the nature of the state being amplified that is 96.1% close to an ideal anapole.
We revised the main text on lines 52-56, 171-175, and added Refs. .
A fundamental difference between our system and nanolasers emitting on BIC states lies in the fact that an anapole laser is compact and localized. On the contrary, standard BIC implementations require extended structures, where the formation of a BIC state is either achieved by symmetries in the photonic band structure or by engineering supercavities.
Another important difference consists in the fact that in a BIC laser, localized BIC states usually exist and compete with a large number of propagating modes. In the case of the anapole-based nanolaser, conversely, the non-radiating state dominates the system response and radiating components are negligible, as discussed previously.
We revised the manuscript on lines 56-58, and added Refs.
[A3]-[A4] as Refs. .
## Referee #1:
From the laser theory perspective, does the anapole nanolaser still require population inversion? Which parameters will dictate the threshold carrier density? And what would be the corresponding cavity quality factor? How does the emission get amplified without any optical feedback?
## Our reply:
Our first principle simulations show that the lasing characteristics of the anapole laser (emission and linewidth vs pumping rate), follow the same behavior of a standard laser. The system requires population inversion, in order for the optical amplifier to provide the necessary gain to the structure, as depicted in revised . The carrier threshold density to lasing, as observed in of the main text, depends on the pumping rate ρ 0 , which is proportional to the population inversion density, or equivalently, to the input pumping rate. By increasing the latter, we increase ρ 0 and control the laser threshold condition.
As discussed above, the anapole state does not originate from a resonance of the system. This specific condition makes the definition of the quality factor Q for an anapole an ill-posed problem. The quality factor Q, in fact, is defined for resonant states that radiate electromagnetic energy in time proportionally to !"#! ! ! ! !! , being ! the frequency of the electromagnetic resonance. This implies that in the scattering cross section of the system at the resonance ! , an observer would measure a Lorentzian peak with full width half maximum equal to ! / . An anapole state, conversely, does not possess a Lorentzian shape and is observed in points of the spectrum where the scattering cross section vanishes. This implies that it is not possible to associate a Q factor to this particular state.
The amplification of anapole states does not require optical feedback. This depends on the fact that the anapole is a localized state and, at such, it can be directly amplified by stimulated emission. The steady state amplification of the anapole results from a balance between fieldenhancement originated by the gain material and the losses of the system, including absorption and near-field energy leakage. This is the same scenario that is observed in standard nanolasers such as, e.g., core-shell spasers: We revised the main text on lines [A5] as Ref. in the revised text.
## Referee #1:
In the optical router , can the authors provide more details on how the pump beam can selectively excite the anapole with field polarization perpendicular to the pump polarization? While I understand that the pump is chosen at a different frequency corresponding to the maximum scattering, the symmetry of the mode patterns at this frequency and how it couples to the specific spatial field of the anapole should be clarified.
## Our reply:
We apologize with the Referee if the manuscript was not clear on this point. The anapole symmetry selection results only from the polarization of the pump beam. The anapole state, in fact, is originated from the interference of an electric dipole and toroidal dipole: the latter has cylindrical symmetry, while the former has a strongly asymmetric shape elongated in the spatial direction perpendicular to the incident polarization. The result of this interference, as illustrated in of the main text and in , shows a three-lobe asymmetric shape perpendicular to the electric field polarization, with energy localized at opposite spatial sides on the nanodisc. As a consequence to this result, the total symmetry direction of the excitation of the electric dipole is orthogonal to the input polarization, as shown in of the main text. When we nonlinearly amplify the anapole in our laser, the input pump polarization selectively excites the corresponding symmetry direction in the electric dipole, leading to different energy distributions in the final anapole state. In order to clear this point in the text, we revised of the main text by adding a clear indication of the polarization selection mechanism in Panel a:
## Revised fig. 4
We then revised the main text on lines 186-194 and explained in more detail the polarization selection rule of the anapole state amplified in the laser.
## Referee #1
The authors showed a four orders of magnitude improved efficiency in selective coupling to the waveguides compare to a InGaAs spherical nanoparticle. However, the coupling efficiency generally depends on the mode matching, so I am not sure if one single comparison (for a specific geometry) is convincing enough to prove the superiority of the anapole. The authors should also comment on its advantage compare to other similar systems such as a metallodielectric nanolaser-waveguide design or a plasmon laser circuit.
## Our reply:
We agree with the Referee that another comparison can help to demonstrate the superior performance of this system. The nanosphere in , for the same radius of the cylinder of the anapole, possesses a strong scattering peak at the same frequency of the anapole, and therefore permits a comparison in similar optical conditions. Following the Referee suggestion, we revised the manuscript and included a direct comparison with another integrated structure, composed by a cylindrical nanolaser emitting at a frequency with strongly dipolar energy emission. The latter well couples with the fundamental waveguide mode of the nanowire, and therefore constitutes an interesting benchmark for the anapole laser. In our analysis, we considered an InGaAs cylinder with the same geometrical parameters of the anapole laser (radius d=220 nm and height h=100 nm), but emitting at the wavelength 1125 nm, located on a resonant peak as in a conventional nanolaser structure. The nonlinear emission of the semiconductor is set to this value by considering an Indium fractional concentration x=0.32.
Changing the In concentration of the semiconductor induces a slight variation of the complex refractive index n(λ), which we have explicitly considered in our simulations. Due to the material variation, the scattering spectrum of the InGaAs cylinder is slightly modified, as can be seen in in : Scattering cross-section for a In 0.32 Ga 0.68 As cylinder. The In concentration x=0.32 corresponds to an emission at the wavelength 1125nm. As a reference, we report the scattering efficiency for the proposed anapole laser (blue dotted line), based on In 0.15 Ga 0.75 As.
In this configuration, the nanolaser emits on a strongly scattering dipolar mode, which is dominated by an electric dipole spatial distribution. This is further illustrated in [fig_ref] [ A5 ]: Noginov, M [/fig_ref] , where we report the volumetric plot of the electromagnetic energy at the emission wavelength. As can be seen from the figure, most of the electromagnetic energy is radiated into space, while a small percentage of energy is coupled into the waveguides via butt-coupling. In addition to that, the system couples energy in all the waveguides, as can be easily observed by : Surface plot of (x, y) section of the electromagnetic energy in logarithmic color scale.
The energy coupled in the nanowire channel in this configuration is 15 times lower (more than one order of magnitude) than the one observed with the anapole laser, thus proving the superior performance of this system. This result depends on the fact that the anapole laser takes advantage from the mechanism of side waveguide coupling, which is superior to butt coupling with propagating modes. The latter process relies on perfect mode matching, which is very challenging to achieve with integrated nanostructures.
Compared to plasmonics systems, the main advantage of the anapole lies in direct integrability and fully CMOS compatibility for silicon photonics circuits. The anapole laser does not employ noble metals and at such the cost is also reduced for this structure. In addition to these points, another important advantage lies in the fabrication: the anapole has an all-dielectric structure that can be easily controlled by e-beam lithography with high nanoscale precision.
We revised the text on lines 207-213 added Figs. a4-a6 as .
## Referee #1:
For the mode-locking part, the authors stated that the anapole chain performs like weakly coupled nonlinear oscillators, and the weak coupling is due to the spatial overlap of the anapole emitted fields. While I agree that the system has the advantage of not needing external design elements, I wonder how it is different from other integrated phase-locked lasers that rely on evanescent coupling of multiple laser sources?
## Our reply:
In phase-locked laser arrays, each laser typically emits at the same frequency and the phase locking condition affects the spatial distribution of the emitted radiating field. This a powerful technique that is normally used to control the spatial distribution the laser, as discussed, e.g., in in Ref.
[51] of the revised text, which we report here for completeness: In the anapole laser array, conversely, each nanolaser emits at a different frequency and the phase locking condition is spontaneously developed in the time dynamics of the field, opening the possibility to control at the ultrafast scale the time evolution of the electromagnetic energy emitted from the laser. This difference makes an anapole laser array more similar to an oscillatory neural network, which has been suggested as the basic unit for neurocomputers, i.e., a future generation of computers that are based brain functions, as introduced in:
[A6] Hoppensteadt, F. C. & Izhikevich, E. M. Oscillatory Neurocomputers with Dynamic Connectivity. Phys. Rev. .
To provide more details on this scenario, and clearly illustrate the differences with a more conventional laser array, we revised the section related to mode-locking of anapoles (lines 224-228, 235-311), employing a simplified model that explains the main dynamics of the Anapole system, which is obtained by generalizing the theory of multi-modal laser systems, as discussed in e.g., Section II of:
[A7] Javaloyes, J., Mandel, P. & Pieroux, D. Dynamical properties of lasers coupled face to face. Physical .
to the multi-frequency emission regime. We added also one new section in the supplementary material titled "Nonlinear dynamics of the anapole chain", which explains the derivation of the model and a new supplementary figure 10, which contains simulation results on the nonlinear phase-locking dynamics of the anapoles. We also added Refs. [A6,A7] as Refs. .
## Referee #1:
For practical application, electrically pumped nanolasers are preferred. Can the authors comment on any possible route towards electrical injection without perturbing the anapole? As the Referee correctly points out, electrical pumping presents several advantages, especially with reference to integrated applications. Our choice of InGaAs is primarily driven by the largescale use of this material in electrically pumped diode lasers, such as: To realize an electrically pumped anapole laser, the main point is to design a semiconductor heterostructure that ensures three-dimensional confinement of the anapole mode. As an alternative to a typical vertical cavity setup, a possible solution is to consider a core-shell geometry, such as the one proposed in Ref. . As discussed in:
[A9] Liu, W., Zhang, J., Lei, B., Hu, H. & Miroshnichenko, A. E. Invisible nanowires with interfering electric and toroidal dipoles. Optics Letters 40, 2293 (2015), anapole states can be excited in core-shell structures by acting on the multilayer thicknesses. By employing this strategy, an electrically pumped anapole laser can be composed of InGaAs/GaAs or InGaAs/InP heterostructures. This structure, after adding top and bottom electrodes for electric carriers injection, can provide an initial setup for an electrically pumped anapole laser, where the anapole state is generated in the intrinsic layer InGaAs.
Another important point is the optical footprint of the electrical circuitry. To this extent, dielectric conductors such as, e.g., indium tin oxide (ITO), are preferred to metal contacts. The use of dielectric contacts helps in minimizing optical losses. An intriguing possibility would be the introduction of graphene-based contact layers, such as the ones proposed in:
## Referee #1:
Similarly, will the anapole mode be severely affected by having a realistic lattice-matched substrate beneath it?
## Our reply:
This is an interesting technological question. If the lattice mismatch affects few lattice sites only, we do not expect any perturbation in the optical properties of the system and in the resulting dynamics of the anapole laser. However, to reduce lattice-mismatch non-idealities, the InGaAs layer can be grown on a lattice-matched substrate, and then transferred on a low-refractive index substrate that is compatible with Si technology. At the proposed x=0.15 concentration of Indium, the InGaAs is lattice matched to both InP and GaAs, which could be used as ultra-thin epitaxial growth substrates. By introducing a sacrificial layer (such as, e.g., AlAs), the InGaAs/GaAs or InGaAs/InP substrate is easily transferred to a low-refractive substrate. Such technique is widely used in the fabrication of III-V semiconductor nanolasers on silicon platforms, as discussed in [A8] and in the article:
## Referee #2:
The authors present a numerical study of "anapole lasers" and suggest a number of applications, including couplers, splitters, and mode-locking. Since anapoles correspond to non-radiating configurations, such systems are uncoupled to free-space allowing amplification of the nearfield. The paper deals with a timely topic, which is currently attracting substantial attention in the nanophotonics research community. The manuscript introduces intriguing ideas, however, the set of results presented is not sufficient to assess the importance of the work. In particular:
The main issue is how good an approximation to an ideal anapole is the configuration of Figs 3c-e. The authors should provide a multipole decomposition of this configuration including electric quadrupole, magnetic dipole and quadrupole in addition to the electric and toroidal dipole contributions.
In , the authors present a multipole decomposition for the ideal lossless configuration. They should include higher order multipoles (see comment 1) and also present results for the case of realistic losses in the system.
In , it is shown that the scattering cross-section does not depend strongly on the presence of losses. The authors should provide a justification for this. Also, why does the scattering crosssection retain high values even at the anapole resonance?
## Our reply:
We thank the Referee for the positive appreciation of our work, as well as for providing timely and extremely valuable comments.
Following the Referee suggestion, we performed a complete multipolar expansion of the field in the presence of losses. The [fig_ref] Figure b1: Multipole analysis of an ideal anapole state [/fig_ref] shows the multipolar decomposition of the field in terms of fundamental and high order dipoles in the ideal case, while [fig_ref] Figure b2: Effect of losses in the Anapole laser [/fig_ref] -e provides the multipolar decomposition in the case of losses, with [fig_ref] Figure b2: Effect of losses in the Anapole laser [/fig_ref] showing the corresponding absorption cross section of the system. The role of losses and dissipation in the formation of anapole states is a subject of research interest, as discussed in the recent paper:
[B1] Tribelsky, M. I. & Miroshnichenko, A. E. Giant in-particle field concentration and Fano resonances at light scattering by high-refractive-index particles. Physical Review A 93, (2016).
In our system, as we see from [fig_ref] Figure b2: Effect of losses in the Anapole laser [/fig_ref] , the net effect of losses in the linear regime is to slightly reduce (of almost the same factor) the energy coupled into all the multipolar components of the electromagnetic field. As shown in [fig_ref] Figure b2: Effect of losses in the Anapole laser [/fig_ref] , the energy difference that is not coupled to electromagnetic components contributes to increase the absorption scattering cross section of the system at the anapole frequency. In our particular case, this variation is little as the losses of the system are small. b3 shows FDTD simulations of the system linear response when a plane wave impinges on the nanodisk supporting an anapole state in the (a) lossless and (b) real InGaAs medium with loss k=0.15: Effect of losses and scattering suppression of an anapole state. a,b Electromagnetic energy distribution for an In0:15Ga0:85As disc of diameter d = 440 nm at the anapole wavelength = 498nm. The difference between the (a) lossless and (b) realistic lossy configuration is minimal. The imaginary refractive index is (a) k = 0 and (b) k = 0:15.
As seen from FDTD results, the linear response of the system is almost the same in both cases.
More interesting is the effect of losses in the anapole laser, where the anapole state is nonlinearly amplified via stimulated emission. To evaluate this effect quantitatively, we calculated by FDTD simulations to what extent the nonlinearly amplified anapole state of is close to an ideal anapole, as requested by the Referee. [fig_ref] Figure b4: Scattered Field from an Ideal anapole [/fig_ref] shows the scattered electromagnetic energy in the case of an ideal anapole. This calculation is performed by extracting the anapole state from the scattered fields of the InGaAs nanocylinder in the ideal lossless case with k=0. We then computed the cross correlation between the ideal anapole of [fig_ref] Figure b4: Scattered Field from an Ideal anapole [/fig_ref] and the electromagnetic distribution shown in of the amplified anapole in the realistic, lossy InGaAs laser with k=0.15. The cross correlation coefficient is 96.1%, which shows that the anapole laser practically amplifies an almost ideal anapole state. This results from the fact that the electromagnetic losses, as shown in [fig_ref] Figure b2: Effect of losses in the Anapole laser [/fig_ref] , increase the amount of energy trapped inside the nanodisk, and this mechanism provides an additional contribution to local field enhancement in the nonlinear processes of light-matter interaction, which contributes to amplify a radiationless anapole state in the InGaAs material.
The scattering cross section values associated to the anapole state and presented in
## Referee #2:
In , it is not clear under which conditions the scattering cross-section is calculated. Is this assuming plane wave illumination?
## Our reply:
The Referee is correct: our linear scattering cross-section calculations were performed by considering a monochromatic plane-wave illumination. We revised the main text on lines 317-319.
## Referee #2:
The authors suggest that changing the geometry of the anapole chain from the waveguide allows to control the temporal dynamics of the system. The authors should discuss the main mechanism for this behavior. For example, is it due to a change in the interactions within the chain, or simply an interference effect of the emitted fields (or both)?
## Our reply:
The Referee is correct; the temporal dynamics of the system is controlled by changing the interactions among the chain. We followed the Referee suggestion and revised the section related to the ultrafast pulse generation by discussing in more details the mechanism of this behavior. We illustrated this dynamics by resorting to a simplified model of light-matter interaction, which is obtained by generalizing the theory of multi-modal laser systems, as discussed in e.g., Section II of:
[B4] Javaloyes, J., Mandel, P. & Pieroux, D. Dynamical properties of lasers coupled face to face. Physical .
to the multi-frequency emission regime. We added one new section in the supplementary material titled "Nonlinear dynamics of the anapole chain", which explains the derivation of the model and a new supplementary figure 10, which contains simulation results that illustrate in more details the mechanisms of phase-locking of the anapoles for different coupling conditions. The dynamics of the Anapole chain is very similar to an oscillatory neural network, as introduced in:
[B5] Hoppensteadt, F. C. & Izhikevich, E. M. Oscillatory Neurocomputers with Dynamic Connectivity. Phys. , as basic building block for future neurocomputing architecures, i.e., computers that mimic brain functions.
We revised the main text on lines .
## Referee #2:
In page 10, the authors mention a statistical campaign of simulations. It would be useful to provide some more information on this.
## Our reply:
We followed the Referee suggestion and revised the text on lines 284-288 furnishing more details.
## Referee #2:
7. Prior art is not properly referenced. For example, active sources of electromagnetic fields uncoupled or weakly coupled to free-space have been suggested before, see for example . Sub-diffraction-limit laser sources have been studied in many works, such as Nature . Moreover, an active source based on toroidal modes was introduced in Sci. . The dynamic anapole was introduced in J. and first observed in . For a history of the field see Nature and references therein. The authors should discuss advantages and disadvantages of the suggested scheme compared to other approaches in the literature (see comment 7).
## Our reply:
We thank very much the Referee for the suggested literature, which we cited as Refs. We also found the following relevant additional papers:
[B6] Savinov, V., Fedotov, V. A. & Zheludev, N. I. Toroidal dipolar excitation and macroscopic electromagnetic properties of metamaterials. [B7] Kaelberer, T., Fedotov, V. A., Papasimakis, N., Tsai, D. P. & Zheludev, N. I. Toroidal Dipolar Response in a Metamaterial. Science 330, 1510-1512 (2010), which we cited as Refs. in the revised paper.
Concerning the advantages of the anapole laser, a first point comes from the lack of radiating components, which allows the anapole laser to shape the beam controllably, opening to applications that are not possible with standard systems based on radiating modes. An example is provided in revised , discussing a spontaneously polarized laser with routing functionalities.
Another advantage of this near-field emitting platform lies in the possibility to engineer anapoleanapole interactions and create nanolasers with advanced functionalities, such as in the example presented in this paper for the generation of ultrafast pulse generation on chip, which is not observed in classical nanolasers and represents an attractive feature of this design, opening also interesting connections with optical neurocomputing.
Another advantage of the anapole laser relies in a compact and integrated all-dielectric structure, which is fully compatible with CMOS technology and Silicon Photonics. This leads to important benefits in the fabrication of the samples, which can be easily controlled in 2D with high precision through electron beam lithography and directly integrated on-chip.
We revised the main text on lines 52-58, 72-74, 182-183, 310-314.
## Referee # 2:
For the cases of Figs. 4-6 the authors should normalize the output power to the input (pump) power.
## Our reply:
We re-plotted the Figures by normalizing the intensity with respect to the intensity of the electromagnetic field inside the nanodisc, which measures the electromagnetic energy stored in the anapole state within the gain material. This choice provides the adimensional quantity that is typically employed in the study of nonlinear light-matter interactions in optical amplifiers following first principle Maxwell-Bloch equations, see, e.g.:
Moloney, J. V. & Newell, A. C. Nonlinear Optics. .
In the applications described by Figs 4-6, this normalization allows to directly measure the main figure of merit of the interaction, represented by the energy transfer between the anapole and the guided modes in the nanowire channels.
## Referee #3:
This is an interesting paper reporting on nanoscale lasers based on tightly-confined anapoles. The authors show on the basis of semiclassical simulations that is it possible to engineer a nanolaser running on anapole modes. In a design using conventional InGaAs nanodiscs they suggest on-chip sources to efficiently couple light into waveguide channels with high intensity and a route to generate ultrafast sub-picosecond pulses. The paper is well written and merits being considered for publication. However, before publication, the authors are required to sufficiently address the following points:The laser dynamics seems to be dominated by an anapole mode. What is the radiating (far-field) mode? Which pattern will be coupled out? Is there a quadupolar component?
## Our reply:
We Thank the Referee for the positive appreciation of our manuscript and for the valuable comments and remarks.
The anapole state is radiationless and it is generated by the superposition of toroidal (TD) and electric (ED) dipoles, which cancel each-other in the far-field. Electromagnetic components that do not couple to ED or TD radiate to the far field. To quantify the amount of radiative emission, we performed a full multipolar decomposition of the linear anapole for different values of the losses (k=0: ideal lossless material, k=0.15, real absorption of InGaAs) as shown in [fig_ref] Figure c1: Effect of losses in the Anapole laser [/fig_ref] (ED: electric dipole, TD: toroidal dipole, MD: magnetic dipole, EQ: electric quadrupole, MQ: magnetic quadrupole and C abs : absorption cross section) below.
As can be seen from the figure, the contribution of quadrupolar modes is negligible, and it is further decreased by the introduction of material losses (panels d,e).
The spatial profile of the anapole state, as shown in of the revised manuscript and in , shows a typical three-lobes asymmetric profile perpendicular to the electric field polarization, with energy localized at opposite spatial sides on the nanodisk. To quantify the degree of similarity between the laser emission and the ideal anapole state, we computed the cross-correlation coefficients of their spatial distributions. More specifically, we compared the electromagnetic distribution shown in of the main text to that of an ideal anapole, calculated from [fig_ref] Figure c1: Effect of losses in the Anapole laser [/fig_ref] in the case of no losses (k=0). The cross-correlation coefficient is 96.1%, which shows that the anapole laser allows to create an almost ideal radiationless anapole state, without the need of any external radiating component, but just destructive interference of specific multipoles that are suitably excited by stimulated emission. This result originates from the fact that, as shown in [fig_ref] Figure c1: Effect of losses in the Anapole laser [/fig_ref] , the presence of losses decreases (of almost the same factor) the energy of all multipolar components and increases the absorption scattering cross-section of the nanodisk. This implies that the energy difference that is not coupled to ED and TD components, is absorbed inside the nanodisk. This process does not change the nature of the state being nonlinearly amplified, which is 96.1% close to an ideal anapole.
We revised the main text on lines 171-175 and added [fig_ref] Figure c1: Effect of losses in the Anapole laser [/fig_ref] as . We also added a new section in the supplementary titled "Effects of losses in the anapole laser", where we discuss in more detail the formation of the radiationless state in the nonlinear regime of light amplification.
## Referee #3:
The periods of pulses in and are 190fs (seemingly from simulations) and 1.65ps (experimentally measured). Why? This analysis is very important is very important for an understanding of the process and the authors should provide an explanation to fully justify their experimental observation.
## Our reply:
We apologize if the manuscript was not clear enough on this point. The results refer to different configurations of anapoles, where the temporal dynamics is controlled by changing the interactions among the anapoles in the chain. We followed the Referee suggestion and revised the section related to the ultrafast pulse generation by discussing in more details the mechanism of this behavior. We illustrated this dynamics by resorting to a simplified model of light-matter interaction, which is obtained by generalizing the theory of multi-modal laser systems, as discussed in e.g., Section II of:
[C1] Javaloyes, J., Mandel, P. & Pieroux, D. Dynamical properties of lasers coupled face to face. Physical .
to the multi-frequency emission regime. We added one new section in the supplementary material titled "Nonlinear dynamics of the anapole chain", which explains the derivation of the model and a new supplementary figure 10,
[fig] [ A1 ]: Wuestner, Sebastian and Hamm, Joachim M. and Pusch, Andreas and Renn, Fabian and Tsakmakidis, Kosmas L. and Hess, Ortwin, "Control and dynamic competition of bright and dark lasing states in active nanoplasmonic metamaterials", Phys. Rev. B 20, 201406 (2016), [A2] Bergman, D. J. & Stockman, M. I. Surface Plasmon Amplification by Stimulated Emission of Radiation: Quantum Generation of Coherent Surface Plasmons in Nanosystems. Phys. [/fig]
[fig] Figure a2: shows the multipole decomposition for an ideal anapole in InGaAs in the case of no losses k=0.Fig. a2 Multipole analysis of an ideal anapole state. The multipole components are computed in Cartesian coordinates and they correspond to: Electric Dipole (blue line), Toroidal Dipole (orange line), Magnetic Dipole (green line), Electric Quadrupole (cyan line), and Magnetic Quadrupole (violet line). At the anapole frequency (dashed vertical line), toroidal and electric dipole cancel each other producing the radiation-less state. [/fig]
[fig] [ A5 ]: Noginov, M. A. et al. Demonstration of a spaser-based nanolaser. Nature 460, 1110-1112 (2009). or semiconductors nanolasers, Ref. [18] of the revised manuscript: [18] Chen, R. et al. Nanolasers grown on silicon. Nat Photon 5, 170-175 (2011). [/fig]
[fig] Figure a5: Electromagnetic energy of a cylinder emitting on a dipolar scattering state. [/fig]
[fig] [ 51 ]: Kao, T.-Y., Reno, J. L. & Hu, Q. Phase-locked laser arrays through global antenna mutual coupling. Nature Photonics 10, 541-546 (2016). [/fig]
[fig] [ A8 ]: Zhou, Z., Yin, B. & Michel, J. On-chip light sources for silicon photonics. Light Sci Appl 4, e358 (2015). [/fig]
[fig] [ A10 ]: Kim, Y.-H. et al. Graphene-contact electrically driven microdisk lasers. Nature Communications 3, 1123 (2012),which is a promising design for a room-temperature electrically pumped anapole laser.We revised the manuscript on lines 116-129 and added Refs. [A8-A10] as Refs.[35,26,37]. [/fig]
[fig] [ A11 ]: Kim, J. et al. Ultra-thin flexible GaAs photovoltaics in vertical forms printed on metal surfaces without interlayer adhesives. Appl. Phys. Lett. 108, 253101 (2016). We revised the main text on lines 106-115 and added Ref. [A11] as Ref. [36]. [/fig]
[fig] Figure b1: Multipole analysis of an ideal anapole state. The multipole components are computed in Cartesian coordinates and they correspond to: Electric Dipole (blue line), Toroidal Dipole (orange line), Magnetic Dipole (green line), Electric Quadrupole (cyan line), Magnetic Quadrupole (violet line). At the anapole frequency (dashed vertical line), toroidal and electric dipole cancel each other producing the radiation-less state. [/fig]
[fig] Figure b2: Effect of losses in the Anapole laser (a)-(e) Multipole decomposition for InGaAs cylinder of height h=100 nm and radius 220 nm vs. the incident wavelength and material losses k. Multipole components are: (a) Electric Dipole, (b) Toroidal Dipole, (c) Magnetic Dipole, (d) Electric Quadrupole (e) Magnetic Quadrupole. (f) Absorption cross-section. [/fig]
[fig] Figure b4: Scattered Field from an Ideal anapole (k=0) for an InGaAs nanodisk of diameter d=440nm and height 100nm. [/fig]
[fig] Figure c1: Effect of losses in the Anapole laser (a)-(e) Multipole decomposition for an InGaAs cylinder of height h=100 nm and radius 220 nm vs. the incident wavelength and of the material losses k. The multipole components are: (a) Electric Dipole, (b) Toroidal Dipole, (c) Magnetic Dipole, (d) Electric Quadrupole (e) Magnetic Quadrupole. (f) Absorption cross-section. [/fig]
|
Neurenteric Cyst: Case Report and Operative Video
Neurenteric cysts are rare, congenital lesions of the spinal axis composed of endodermal tissue arising from poor segmentation of the notochord. A 36-year-old patient presented with arm paresthesias and incontinence with imaging revealing a lesion in the C6-C7 region most consistent with neurenteric cyst. After partial resection of the lesion, the patient regained all neurological function. Here, we provide a brief overview of this rare neuropathologic entity and demonstrate surgical resection of neurenteric cyst through operative video.
# Introduction
Neurenteric cysts are uncommon congenital spinal lesions which account for only 0.7%-1.3% of spinal axis tumors. These cysts result from inappropriate segmentation of the notochord during embryogenesis causing endodermal tissue to remain in the spinal canal. Neurenteric cysts present more commonly in males, cause focal sensorimotor symptoms at their spinal level, and typically appear as non-contrast-enhancing lesions that are T1 isointense and T2 hyperintense. Gross total resection is the most common management. We present a case of neurenteric cyst and highlight surgical management of this lesion through operative video.
## Case presentation
A 36-year-old male presented with worsening left arm and leg paresthesias and issues with bladder control. Symptoms were noted shortly after a hospitalization for bacterial meningitis two months prior. Initially, the patient felt tingling and numbness in the left posterior arm, medial forearm, and left hand involving the third, fourth, and fifth digits. These symptoms eventually progressed to the left leg and were accompanied by episodes of urinary incontinence and ataxia.
On physical exam, the patient had weakness of the left-hand intrinsic muscles (3/5), wrist extensors/flexors (4/5), and triceps (4/5); his other extremities had full strength. Sensation was diminished in the left C8 dermatome. Upper and lower extremity reflexes were normal, gait was normal, and Hoffman's sign was negative. MRI cervical spinerevealed a large anterior intradural, extramedullary cystic spinal lesion at C6/C7 with compression of the spinal cord posteriorly. The lesion is T2 hyperintense and T1 hypointense and measures 1.6 x 2.7 x 3.5 cm in size. Also notable was fusion of the C6 and C7 vertebrae. The lesion was most consistent with neurenteric cyst, though expanded differential could also include schwannoma, epidermoid cyst, and arachnoid cyst. Treatment of the cyst involved partial laminectomies at C5 and T1 and a C6-C7 laminectomy. This was followed by fenestration and excision of the cyst (Video 1) and a C6-C7 laminoplasty.
## Video 1: operative video of neurenteric cyst resection.
Orientation: right = cranial, left = caudal View video here: https://vimeo.com/420816405
Careful opening of the arachnoid was performed with sharp dissection using microscissors. Gentle dissection down the right aspect of the spinal cord, in between the C6 and C7 nerve roots, allowed for identification of the cyst. The cyst was then opened sharply and a significant amount of fluid was evacuated. At the dural entry of the right C7 nerve root, a small amount of mucinous material was resected and sent for frozen section, which confirmed neurenteric cyst. Given the ventral location of the cyst under the spinal cord, complete resection of the entire cyst was not possible.
Postoperatively, the patient showed immediate improvement in left arm paresthesias and slight residual numbness in digits 3-5. The patient was discharged on postoperative day 3 and was ambulatory and voiding. Full strength was reached at two-week follow-up. On pathological analysis, the lesion demonstrated mucin-producing goblet cells surrounded by a central cystic cavityand was formally diagnosed as a congenital, non-segmenting neurenteric cyst.
# Discussion
Neurenteric cysts are rare, congenital lesions that are usually ventral to the spinal cord in the intradural, extramedullary compartment of the cervical and thoracic regions of the spinal canal. Composed of heterotopic endodermal tissue that results from incomplete separation of the notochord and endoderm during the third week of embryogenesis, neurenteric cysts eventually present during the second or third decade of life, as did our patient, and are more common in males. Consistent with our patient's presentation, most common symptoms are pain, radiculopathy, or myelopathy, but symptoms may wax and wane. In 50% of cases, other bony abnormalities of the spine such as spinal dysraphism, scoliosis, spina bifida, split cord malformation, or Klippel-Feil syndrome can be observed.
Radiographically, neurenteric cysts are typically hypo-or isointense, non-enhancing lesions on T1 MRI and hyperintense on T2 MRI. Histologically, neurenteric cysts are composed of a central cystic cavity surrounded by mucin-producing simple columnar or cuboidal goblet cells that can be ciliated or non-ciliated. Surgical management is the recommended course of treatment, with a posterior approach the most widely reported technique used. Gross total resection via a posterior approach is usually not possible, as in our case, with recurrences reported in the literature likely due to the ventral location of these lesions. In general, post-surgical recurrence has been reported in up to 37% of cases. In a case series of 23 patients, partial excision had higher recurrence rates but shared similar clinical outcomes due to the benign nature and favorable prognosis of this entity. In a series of 121 spinal neurenteric cysts, gross total resection was possible in 44.6%, recurrence occurred in 22.7%, and progression free survival rate at 10 years was 66.2%. Ultimately, surgical intervention allows for resolution of neurological symptoms, and though gross total resection may be desired, subtotal resection allows for similar outcomes.
# Conclusions
A 36-year-old patient presenting with arm paresthesias and incontinence was found to have a C6-C7 neurenteric cyst. We demonstrate partial surgical resection of this rare congenital tumor located in the ventral spinal cord in an accompanying surgical video.
# Additional information disclosures
Human subjects: Consent was obtained by all participants in this study. Baylor College of Medicine IRB issued approval H-43183. The patient consented to the use of anonymized/nonidentifiable information being used in an academic or research setting.
## Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work. |
Prevalence of Shigella, Salmonella and Cmpylobacter Species and Their Susceptibility Patters Among Under Five Children With Diarrhea in Hawassa Town, South Ethiopia
BACKGROUND: Diarrhea is the leading cause of morbidity and mortality in under-five children in developing countries including Ethiopia. Therefore, up-to-date data on etiologic agent and susceptibility pattern are important for the management of bacterial diarrhea in under-five children, which was the main objective of this study.METHOD: A cross-sectional study was conducted at Hawassa Adare Hospital and Millennium Health Center from June 6 to October 28, 2011. A total of 158 under-five children with diarrhea were selected using convenient sampling technique. Demographic and clinical data were collected using questionnaire. Fecal samples were collected and processed for bacterial isolation, and antimicrobial susceptibility testing following standard bacteriological techniques.RESULT: A total of 158 fecal samples were collected from 81(51.3%) males and 77(48.7%) females of underfive children with diarrhea. Of the 158 fecal samples, 35(22.2%) bacterial pathogens were isolated. The isolated bacteria were Campylobacter species, 20 (12.7%), Shigella species, 11 (7.0%), and Salmonella species, 4 (2.5%). The majority of the isolates were sensitive to Chloramphenicol, Ciprofloxacin, Nalidixic acid and Cotrimoxazol and high rate of drug resistance was observed against Erythromycin and Amoxicillin.CONCLUSIONS: The finding of this study indicates that Campylobacter species were the predominant etiologies and the presence of bacterial isolates resistant to the commonly prescribed drugs for treating diarrhea in children. Therefore, periodic monitoring of etiologic agent with their drug resistant pattern is essential in the management of diarrhea in children.
# Introduction
Diarrhea is one of the main causes of morbidity and mortality of under-five children in developing countries. Here, the average number of episodes of diarrhea per child per year within this age group is 3.2. Bacterial diarrhea in under-five children is commonly caused by Salmonella, Shigella Campylobacter species and diarrheogenic E. coli. Of the pathogens causing diarrhea, Shigella continues to play a major role in etiology of inflammatory diarrhea and dysentery. Thus, it presents a serious challenge to public health authorities worldwide [bib_ref] Etiology of Diarrhea in Children Younger than 5 years of age Admitted..., Mandomando [/bib_ref]. The prevalence of Salmonella infection varies depending on the water supply, waste disposal, food preparation practices and climate. The commonest illness among children caused by Salmonella is gastroenteritis. Campylobacter is one of the most frequently isolated bacteria from stools of infants with diarrhea in developing countries as result of contaminated food and water.
There were different prevalence rates of causative agents in different regions. A study conducted in Gondar on children with diarrhea isolated Campylobacter, Salmonella and Shigella species with prevalence of 10.5%, 5.2% and 5.2%respectively [bib_ref] Campylobacter enteritis among children in Dembia district, northwest Ethiopia, Mitikie [/bib_ref].
Another related study conducted in Jimma also reported an isolation rate of 11.6%, 4.9% and 5.8% for Campylobacter, Salmonella and Shigella species respectively [bib_ref] Antimicrobial sensitivity Pattern of Campylobacter species among children in Jimma University Specialized..., Beyene [/bib_ref]. Shigella species was also isolated from 34.6% of the patients who attended health facilities in Hawassa Town [bib_ref] Antimicrobial susceptibility pattern of Shigella isolates in Hawassa, Roma [/bib_ref]. Development of antimicrobial resistance by enteric pathogens like Shigella, Campylobacter and Salmonella species against easily accessible and commonly prescribed drugs has become a major concern throughout the world, particularly in developing countries like Ethiopia [bib_ref] Antimicrobial susceptibility of Shigella flexneri and S. dysenteriae isolated from stool specimens..., Temu [/bib_ref].
According to [bib_ref] Shigella and Salmonella serogroups and their antibiotic susceptibility patterns in Ethiopia, Asrat [/bib_ref] , the most strains of Shigella species in Ethiopia was resistant to Erythromycin (100.0%), Tetracycline (97.3%), Cephalothin (86.7%), Ampicillin (78.7%), Chloramphenicol (74.7%) and Sulfonamide (54.7%) [bib_ref] Shigella and Salmonella serogroups and their antibiotic susceptibility patterns in Ethiopia, Asrat [/bib_ref]. Moreover, Yismaw in Gondar also revealed that there was high resistance of Shigella species against Ampicillin (79.9%), Tetracycline (86 %) and Cotrimoxazole (73.4%) [bib_ref] A Five-Year Antimicrobial Resistance Pattern Observed In Shigella Species Isolated From Stool..., Yismaw [/bib_ref]. In addition, a similar study done in Harar on drug susceptibility of Shigella species showed high resistance against Amoxicillin (100), Ampicillin (100%) and Tetracycline (70.6%) [bib_ref] Antibiotic susceptibility patterns of Salmonella and Shigella isolates in Harar, Eastern Ethiopia, Reda [/bib_ref] , whereas resistance for Salmonella species against Ampicillin (81.2%), Cephalothin (86.4%), Chloramphenicol (83.7%), Erythromycin (100.0%), Gentamicin (75.6%), Sulfonamide (81.1%), Tetracycline (94.5%) and Trimethoprimsulfamethoxazole (75.7%) was reported by [bib_ref] Shigella and Salmonella serogroups and their antibiotic susceptibility patterns in Ethiopia, Asrat [/bib_ref] done in Ethiopia [bib_ref] Shigella and Salmonella serogroups and their antibiotic susceptibility patterns in Ethiopia, Asrat [/bib_ref]. Likewise, the resistance to Amoxicillin (100%), Ampicillin (100%), Tetracycline (71.4%) and Chloramphenicol (62.3%) of Salmonellas species was reported by Reda et al in Harar [bib_ref] Antibiotic susceptibility patterns of Salmonella and Shigella isolates in Harar, Eastern Ethiopia, Reda [/bib_ref]. Moreover, a study on antimicrobial susceptibility of Shigella species in Hawassa, indicated the presence of high resistance against Gentamicin (96%), Nalidixic acid (90%), Ampicillin (93%), Erythromycin (90%) and Tetracycline (90%) [bib_ref] Antimicrobial susceptibility pattern of Shigella isolates in Hawassa, Roma [/bib_ref]. Multiple drug resistant Shigella isolates which showed resistance to six antibiotics (Ampicillin, Erythromycin, Cephalothin, Chloramphenicol, Tetracycline and Trimethoprim-sulfamethoxazole) have also been observed [bib_ref] Antimicrobial susceptibility pattern of Shigella isolates in Hawassa, Roma [/bib_ref]. Another study conducted in Jimma on drug susceptibility profile of Campylobacter species reported that the isolates showed 50% and 60% resistance rate to Ampicillin and Trimethoprim-sulfamethoxazole respectively [bib_ref] Antimicrobial sensitivity Pattern of Campylobacter species among children in Jimma University Specialized..., Beyene [/bib_ref].
According to Ethiopia's Demographic and Health Survey of 2005, children who lived in the Southern Nation Nationalities and People (SNNP) Region were more susceptible to episodes of diarrhea (25.1 %) than those who lived in other regions. To our knowledge, there are no published studies on etiologies of diarrhea in under-five children in SNNPR.
Therefore, the purpose of this study was to identify bacterial etiologies of diarrhea and determine their antimicrobial susceptibility in under-five children against commonly prescribed drugs in Hawassa, South Ethiopia.
# Materials and methods
A prospective cross-sectional study was conducted in Adare Hospital and Millennium Health Center in Hawassa, Ethiopia, from June 6 to October 28, 2011 to isolate common bacterial etiologies of diarrhea in under-five children, and assess their susceptibility patterns.
A total of 158 under-five children affected by diarrhea were included in the study. But, children who had taken antibiotic within seven days before data collection, those who were above 5 years old, and children whose parents/guardians were not voluntary were excluded from the study. Medical history was taken from each child and informed consent was obtained from the parents or guardians before sample collection was attempted. All the relevant demographic, clinical and laboratory data were recorded and transferred to the questionnaire prepared for this study.
Specimen Collection and Pathogen Identification: Freshly passed stool and rectal swab were collected, placed immediately in Cary Blair transport medium (Oxoid Ltd, Basingstoke, UK) and transported to the laboratory within six hours of collection. For identification of Shigella and Salmonella species, specimens were placed in Selenite F enrichment broth (Oxoid) and incubated at 37°C for 24 hours, then subcultured onto deoxycholate agar (DCA) and xylose lysine deoxycholate agar (XLD) (Oxoid) agar and then incubated at 37°C for 18-24 hours.
The growth of Salmonella and Shigella species was detected by their characteristic appearance on XLD agar (Shigella: red colonies, Salmonella red with a black centre) and DCA (Shigella: pale colonies, Salmonella black centre pale colonies). The suspected colonies were further tested through a series of biochemical tests to identify Shigella and Salmonella species. Salmonella species were further confirmed by agglutination test with polyvalent anti-sera and Shigella isolates were sero-grouped by the slide agglutination test using commercially available antisera (Denka Seikn Co. Ltd, Japan).
For isolation and identification of Campylobacter species, the samples were inoculated on Charcoal Cefoperazone Deoxycholate (CCD) agar and incubated at 42°C for 48 hours. Campylobacter species were identified by growth on CCD medium with small gray colonies, at microaerophilic condition, growth at 42 0 C, positive oxidase and catalase test and Gram-staining reaction and morphology.
Susceptibility testing: Antimicrobial drug susceptibility testing was carried out by using disk diffusion method, according to guidelines of Clinical Laboratory Standards Institute (CLSI). The antibiotic discs used and their concentration were: Ampicillin (AMP, 10-μg), Chloramphenicol (C, Ethical clearance was secured from the Ethical Clearance Committee of the College of Public Health and Medical Sciences, Jimma University. Permission was also obtained from Hawassa Regional Health Bureau, Hawassa Adare Hospital, and Hawassa Millennium Health Center.
Data were entered and analyzed using SPSS version 16.0 computer software. Comparisons were made using Chi-square test. P-value of <0.05 was considered indicative of a statistically significant difference.
# Results
A total of 158 under-five children with diarrhea participated in the study. Seventy-one (44.9%) of the children were from Millennium Health Center and 87 (55.1%) of them were from Adare Hospital. Out of the 158 study participants, 81 (51.3%) were males and 77 (48.7%) were females resulting in an overall male to female ratio of 1.1:1. The age of the participants ranged from 1-58 months with a mean of 19.59 months (+SD14.89): thirty (19%) of them were younger than 6 months, 42 (26.6%) were between 12 and 23 months, and 30(19%) were between 24 and 35 months old .
## Fig 1: distribution of participants by age and sex
The bacterial pathogens were identified from 35 (22.2%) study participants of which Campylobacter species was the leading isolate that accounted for 20 (12.7%) [fig_ref] Table 1: Frequency of bacterial pathogens isolated from fecal sample of children with diarrhea... [/fig_ref]. resistant to four, five and six drugs respectively [fig_ref] Table 4: Antibiogram of bacterial pathogens isolated from under-five children with diarrhea at Adare... [/fig_ref]. The overall rate of resistance of Salmonella species was high for Erythromycin (100%) and Ceftriaxone (75%). But, lower resistance rate was observed against Nalidixic acid (25%). Among Salmonella serogroups, Serogroup B showed high resistance against Erythromycin (100%), and Ceftriaxone (100%). Similarly Serogroup A showed 100% resistance against both Erythromycin, and Nalidixic acid [fig_ref] Table 3: Drug resistance pattern of bacterial species from under-five children at Adare Hospital... [/fig_ref].
Of the twenty Campylobacter species tested, high rate of resistance was observed against Amoxicillin (80%) and Erythromycin (55%). But, relatively low rate of resistance was seen against Ampicillin (30%), Nalidixic acid (20%), and Cotrimoxazole (20%), and there was no resistance detected against Chloramphenicol, and Ceftriaxone [fig_ref] Table 3: Drug resistance pattern of bacterial species from under-five children at Adare Hospital... [/fig_ref]. Multiple drug resistance was detected in 19 (95%) of the Campylobacter isolates of which 9 (45%), 4(20%) and 1(5%) showed resistance to three, four and five drugs respectively [fig_ref] Table 4: Antibiogram of bacterial pathogens isolated from under-five children with diarrhea at Adare... [/fig_ref].
# Discussion
Diarrhea is among the leading causes of mortality and morbidity in under-five children and continues to be a health problem worldwide, especially in developing countries [bib_ref] Etiology of Diarrhea in Children less than five years of age in..., Vargas [/bib_ref] [bib_ref] Bacterial Pathogens Associated with Bloody Diarrhea in Uruguayan Children, Mota [/bib_ref] where enteric bacteria are among the major causes of diarrhea among under-five children. The overall prevalence of enteric bacteria isolated in this study is comparable with previous studies conducted in Jimma (22.3%) [bib_ref] Antimicrobial sensitivity Pattern of Campylobacter species among children in Jimma University Specialized..., Beyene [/bib_ref] and Dembia District in Gondar (20.9%) [bib_ref] Campylobacter enteritis among children in Dembia district, northwest Ethiopia, Mitikie [/bib_ref]. However, the prevalence of bacterial pathogen is lower compared with previous studies conducted in other developing countries such as Tanzania (42.7%) [bib_ref] Etiology of Diarrhea in Children less than five years of age in..., Vargas [/bib_ref] and Mozambique (27.2%).The possible reason for such difference could be the inclusion of diarrheogenic E. coli in these studies.
According to results of related studies, among enteropathogenic bacteria, Salmonella, Shigella, Campylobacter species and diarrheogenic E. coli were the most frequently isolated bacteria in underfive children with diarrhea. In line with this, Campylobacter, Shigella and Salmonella species were isolated at the rates of 12.7%, 7% and 2.5% respectively. Comparable bacterial rates of isolates were reported from studies conducted in Jimma (Campylobacter species 11.6%, Shigella species 4.9% and Salmonella species 5.8%) (9), Gondar (Campylobacter species 10.5%, Shigella species 5.2% and Salmonella species 5.2%) [bib_ref] Campylobacter enteritis among children in Dembia district, northwest Ethiopia, Mitikie [/bib_ref] and Tikur Anbessa, Ethio-Swedish Children's Hospital (Campylobacter species 13.7%, Shigella species 11.7% and Salmonella species 3.8%) [bib_ref] Studies on enteric campylobacteriosis in Tikur Anbessa and Ethio-Swedish children's hospital, Asrat [/bib_ref].
The rate of Campylobacter infections globally has been increasing with the number of cases often exceeding those of Salmonellosis and Shigellosis [bib_ref] Human Campylobacteriosis in Developing Countries, Coker [/bib_ref]. High rates of Campylobacter species was reported from children in Jimma (11.6%) (9), Addis Ababa (13.7%) (21), North West Ethiopia (13.8%) [bib_ref] Aseffa Campylobacter enteritis among children in north-west Ethiopia: a 1-year prospective study, Gedlu [/bib_ref] and Gondar (10.5%) [bib_ref] Campylobacter enteritis among children in Dembia district, northwest Ethiopia, Mitikie [/bib_ref]. Generally, Campylobacter isolation rates in developing countries range from 5 to 20% [bib_ref] Human Campylobacteriosis in Developing Countries, Coker [/bib_ref] , and the isolation rate of Campylobacter species (12.7%) in this study lies within the indicated range and is comparable with locally conducted studies such as the ones in Jimma (9), Addis Ababa [bib_ref] Studies on enteric campylobacteriosis in Tikur Anbessa and Ethio-Swedish children's hospital, Asrat [/bib_ref] , and North West Ethiopia [bib_ref] Aseffa Campylobacter enteritis among children in north-west Ethiopia: a 1-year prospective study, Gedlu [/bib_ref]. In this study, the majority (80%) of the Campylobacter species were isolated from children of less than 24 month of age, which agrees with the report of the World Gastroenterology Organization 2011.
The prevalence rate of Shigella species in this study was 7% which is closer to the studies conducted in Jimma by Beyene et al [bib_ref] Antimicrobial sensitivity Pattern of Campylobacter species among children in Jimma University Specialized..., Beyene [/bib_ref] , in Gondar by , in North west Ethiopia by Andualem et al [bib_ref] The prevalence and Antimicrobial responses of Shigella Isolates in HIV-1 infected and..., Andualem [/bib_ref] and in Harar, by Reda et al [bib_ref] Antibiotic susceptibility patterns of Salmonella and Shigella isolates in Harar, Eastern Ethiopia, Reda [/bib_ref] , in which the prevalence rates were 5%, 5.2%, 8.7% and 6.7% respectively. But, it is lower than the rate reported in studies carried out at Tikur Anbessa, Ethio-Swedish Children's Hospital by , in Jimma by , in Hawassa by , in Indonesia by , and in Iran by where the prevalence rates were 11.7%, 20.1%, 34.5%, 9.3% and 9.8% respectively. The lower isolation rate could be due to difference in study participants and study time.
In this study, Serogroup B (S. flexneri) was the only species (100%) isolated from the study subjects. Its dominance was also reported in the studies by in Hawassa, with the isolation rate of 99% and by in Addis Ababa where Serogroup B (S. flexneri) was a dominant isolate (54.0%). This feature of S. flexneri was also reported in studies conducted in Indonesia, Iran [bib_ref] Epidemiology Of Shigella Species Isolated From Diarrheal Children And Drawing Their Antibiotic..., Mashouf [/bib_ref] and South India [bib_ref] Changing Patterns Of Antimicrobial Susceptibility Of Shigella Serotypes Isolated From Children With..., Mamatha [/bib_ref]. However, contrary to the current study, lower isolation of S. flexneri species (27%) was revealed by a study conducted in Botswana [bib_ref] Shigella and Salmonella strains isolated from children under 5 years in Gaborone,..., Urio [/bib_ref].
The overall prevalence of Salmonella species in this study was 2.5%. It is lower than the findings of other similar studies conducted by Mitikie et al in Gondar (5.2%), Beyene et al in Jimma (5.2%), and Asrat et al in Addis Ababa (3.8%) [bib_ref] Antimicrobial sensitivity Pattern of Campylobacter species among children in Jimma University Specialized..., Beyene [/bib_ref] [bib_ref] Studies on enteric campylobacteriosis in Tikur Anbessa and Ethio-Swedish children's hospital, Asrat [/bib_ref]. But it is comparable with studies carried out in other developing countries such as Mozambique 2.5 % (1), Tanzania 1.4% [bib_ref] Etiology of Diarrhea in Children less than five years of age in..., Vargas [/bib_ref] , Botswana 3% (29) and Palestine 2% [bib_ref] Detection and Identification of Bacterial Enteropathogens by Polymerase chain reaction and Conventional..., Elamreen [/bib_ref]. Among the four Salmonella isolates, there were 3(1.9%) Serogroup B and 1(0.6%) Serogroup A isolates, which is comparable with the finding of a study reported from Botswana where Serogroup B and Serogroup A were isolated with frequency of 2% and 1% respectively from under five children with diarrhea [bib_ref] Shigella and Salmonella strains isolated from children under 5 years in Gaborone,..., Urio [/bib_ref].
Antimicrobial resistance by enteric pathogens is of major concern because of indiscriminate use of drugs [bib_ref] Antimicrobial susceptibility of Shigella flexneri and S. dysenteriae isolated from stool specimens..., Temu [/bib_ref]. In this study, Shigella isolates revealed reasonably high rate of resistance to a number of commonly used antibiotics in Ethiopia such as Amoxicillin (100%), Erythromycin (90.9%), Ampicillin (63.6%), Ceftriaxone (54.5%) and Tetracycline (54.5%).
The development of high resistance of Shigella species against commonly used antibiotics was witnessed by other investigators in different times. In Hawassa, a high rate of resistance of Shigella species to Ampicillin (93%), Erythromycin (90%), Tetracycline (90%) and Cotrimoxazole (56%) was reported by . In Gondar, high antibiotic resistance was documented against Ampicillin (79.9%), Tetracycline (86 %), and Cotrimoxazole (73.4%) by . Asrat reported isolation of Shigella species with a high resistance to Erythromycin (100%), Tetracycline (97.3%), and Ampicillin (78.7%) in Addis Ababa [bib_ref] Shigella and Salmonella serogroups and their antibiotic susceptibility patterns in Ethiopia, Asrat [/bib_ref].
High resistance against Amoxicillin (100%), Ampicillin (100%) and Tetracycline (70.6%) was also reported by Reda et al in Harar [bib_ref] Antibiotic susceptibility patterns of Salmonella and Shigella isolates in Harar, Eastern Ethiopia, Reda [/bib_ref].
In the current study, all Shigella isolates showed susceptibility to Ciprofloxacin, Nalidixic acid, and Cotrimoxazole. According to the Standard Treatment Guideline for General Health Facilities Treatment of Common Diseases in Ethiopia, Ciprofloxacin is the choice of drug as first line for bacillary dysentery caused by Shigella species and other pathogens. Cotrimoxazole and Ceftriaxone are among the other alternatives. Unlike the findings of previous studies, no resistance was observed against Cotrimoxazole, which could be associated with frequent use of this drug in previously reported study areas. Resistance to Cotrimoxazole, Tetracycline, Nalidixic acid and Ampicillin was also reported from South India [bib_ref] Changing Patterns Of Antimicrobial Susceptibility Of Shigella Serotypes Isolated From Children With..., Mamatha [/bib_ref]. In this study, the rate of resistance of Shigella species against Gentamicin was 27.3%, which is relatively high compared with resistance rates reported in Hawassa (10), Jimma (25), [bib_ref] Shigella and Salmonella serogroups and their antibiotic susceptibility patterns in Ethiopia, Asrat [/bib_ref] and Harar [bib_ref] Antibiotic susceptibility patterns of Salmonella and Shigella isolates in Harar, Eastern Ethiopia, Reda [/bib_ref] , all from Ethiopia, where the reported resistance rates were 2%, 1.3%, 0%, and 0% respectively. The possible reason could be the wide utilization of Gentamicin in the study area and frequent exposure of Shigella species to this drug.
Among the four Salmonella isolates, the overall rate of resistance was high for Erythromycin (100%), Ceftriaxone (75%) and Nalidixic acid (25%) which is comparable with the result of Asrat's study, where 100% and 37.8% of the isolates were resistant to Erythromycin and Nalidixic acid respectively [bib_ref] Shigella and Salmonella serogroups and their antibiotic susceptibility patterns in Ethiopia, Asrat [/bib_ref]. Contrary to our findings, high resistance rates of Salmonella species to Ampicillin (81.2%), Tetracycline (75.5%), Gentamicin (75.6%), Chloramphenicol (83.7%) and Amoxicillin (100%) were reported in other parts of Ethiopia [bib_ref] Shigella and Salmonella serogroups and their antibiotic susceptibility patterns in Ethiopia, Asrat [/bib_ref] [bib_ref] Antibiotic susceptibility patterns of Salmonella and Shigella isolates in Harar, Eastern Ethiopia, Reda [/bib_ref]. This variability could be because of differences in prescription of drug pattern.
Most of the time, in immunocompetent individuals, Campylobacter enteritis is a selflimiting disease and need not to be treated with antimicrobial agents. However, in some patients, Campylobacter may cause severe complications and increase the risk of death, and, therefore, requires treatment [bib_ref] Human Campylobacteriosis in Developing Countries, Coker [/bib_ref]. In this study, eleven antibiotics were tested against 20 isolates of Campylobacter species. Compared with findings of similar studies conducted in other regions of Ethiopia, this study showed increased resistance against Amoxicillin (80%) and Erythromycin (55%),which are used for treatment of diarrhea due to Campylobacter species [bib_ref] Antimicrobial sensitivity Pattern of Campylobacter species among children in Jimma University Specialized..., Beyene [/bib_ref] [bib_ref] Aseffa Campylobacter enteritis among children in north-west Ethiopia: a 1-year prospective study, Gedlu [/bib_ref]. However, the majority of the isolates were sensitive to Tetracycline and Ciprofloxacillin which are considered to be an alternative treatment.
To sum up, in the present study, Campylobacter, Shigella and Salmonella species were isolated at different rates and played a dominant role in causing diarrhea in children of the study area. The overall antibiotic resistance levels against some commonly prescribed drugs were higher than those reported from other regions in Ethiopia, possibly due to the higher levels of exposure and usage of those antimicrobials in the study area.
Therefore, up-to-date information on etiologies and periodic monitoring of the antimicrobial susceptibility pattern is very important for the management of diarrhea in under-five children.
[table] Table 1: Frequency of bacterial pathogens isolated from fecal sample of children with diarrhea at Adare Hospital and Millennium Health Center from June to October 2011, Hawassa, South EthiopiaIn this study, Shigella species was the second dominant bacterial etiology in under-five children with diarrhea and sero-grouping data indicating that all Shigella isolates were found to be S. flexneri. S. flexneri was frequently isolated from children with dysentery, above 24 months of age and with three days' duration of diarrhea. Isolation rate of S. flexneri showed statistically significant association with dysenteric stool [P=<0.001] and children aged 36-47 months [P=0.025] (Table 2). [/table]
[table] Table 2: Demographic and clinical findings in association with culture result for Salmonella, Shigella and Campylobacter species among under-five children at Adare Hospital and Millennium Health Center, June to October 2011, Hawassa, South Ethiopia (*Blood and Mucoid, ** p = value statistical significance association ) [/table]
[table] Table 3: Drug resistance pattern of bacterial species from under-five children at Adare Hospital and Millennium Health Center, June to October 2011, Hawassa, South Ethiopia [/table]
[table] Table 4: Antibiogram of bacterial pathogens isolated from under-five children with diarrhea at Adare Hospital and Millennium Health Center June to October 2011, Hawassa, South Ethiopia = Resistance for one drug, R 2 = Resistance for two drugs, R 3 = Resistance for three drugs, R 4 = Resistance for four drugs, R 5 = Resistance for five drugs, R 6 = Resistance for six drugs, -=No resistance [/table]
|
Multicenter study of re‐irradiation using carbon‐ions for head and neck malignancies after photon radiotherapy
Purpose:The goal of this multicenter retrospective study of patients with head and neck malignancies was to evaluate the efficacy and safety of carbon-ion (Cion) radiotherapy (RT) after photon RT.Methods:We enrolled 56 patients with head and neck malignancies who underwent re-irradiation (re-RT) using C-ions between November 2003 and March 2019, treated previously with photon RT. The tumors at re-RT were located in the sinonasal cavities (n = 20, 35.7%), skull base (n = 12, 21.4%), and orbit (n = 7, 12.5%). The tumors at the initial RT were located in the sinonasal cavities (n = 13, 23.2%), skull base (n = 9, 16.1%), and orbit (n = 9, 16.1%). The median period between the initial RT and re-RT was 41 (4-568) months. The most common histology of re-RT was squamous cell carcinoma (n = 11, 19.6%). The most commonly used protocol was 57.6 Gy (relative biological effectiveness) in 16 fractions (n = 23, 41.1%). Surgery preceded re-RT in three patients (5.4%). One patient with malignant melanoma received concurrent chemotherapy.Results:The 2-year local control, progression-free survival, and overall survival rates were 66.5%, 36.9%, and 67.9%, respectively. The median follow-up time was 28 months. Two patients (3.6%) developed grade ≥ 3 acute toxicities, and 14 (25.0%) developed grade ≥ 3 late toxicities. A single patient had confirmed grade 5 dermatitis with infection.Conclusion: Re-RT using C-ions for head and neck malignancies after photon RT is an effective treatment with tolerable toxicity.K E Y W O R D Sclinical cancer research, head and neck, multicenter study, salvage treatmentSUPPORTING INFORMATIONAdditional supporting information may be found in the online version of the article at the publisher's website.
# | introduction
Treatment strategies for head and neck malignancies must consider functional and cosmetic preservation, as well as tumor control. Photon radiotherapy (RT), widely used for functional and cosmetic preservation, can be added to the standard for treatment protocol for unresectable tumors [bib_ref] An intergroup phase III comparison of standard radiation therapy and two schedules..., Adelstein [/bib_ref] and postoperative adjuvant therapies. [bib_ref] Postoperative concurrent radiotherapy and chemotherapy for high-risk squamous-cell carcinoma of the head..., Cooper [/bib_ref] However, the management of malignancies in patients with prior photon RT is challenging. Several advanced treatments (including stereotactic RT and intensity-modulated RT) have been developed and re-irradiation (re-RT) for head and neck malignancies has improved. [bib_ref] A multi-institutional comparison of SBRT and IMRT for definitive reirradiation of recurrent..., Vargo [/bib_ref] [bib_ref] Reirradiation using robotic image-guided stereotactic radiotherapy of recurrent head and neck cancer, Yamazaki [/bib_ref] High energy X-rays, used in conventional RT, have the ability to penetrate the human body, while carbon ions (C-ions) can reach desirable depths. [bib_ref] Biophysical characteristics of HIMAC clinical irradiation system for heavy-ion radiation therapy, Kanai [/bib_ref] Additionally, C-ions present a Bragg peak, the peak region that occurs where their range ends and have little dose distribution beyond their designated depth. C-ions have the ability to achieve dose localization for deep-seated tumors, being accelerated by the designed energy to develop a high dose at their target depth.
C-ions are classified as high-linear energy transfer (LET) radiation and exhibit higher relative biological effectiveness (RBE) than X-rays. [bib_ref] Biophysical characteristics of HIMAC clinical irradiation system for heavy-ion radiation therapy, Kanai [/bib_ref] Because plateaus of C-ions in the superficial layer of a body show low LET and RBE, C-ion RT (CIRT) can achieve both high dose radiation at the target region and high biological effects. [bib_ref] Biophysical characteristics of HIMAC clinical irradiation system for heavy-ion radiation therapy, Kanai [/bib_ref] CIRT is, therefore, considered useful for photon-resistant tumors. [bib_ref] Multicenter study of carbonion radiation therapy for mucosal melanoma of the head..., Koto [/bib_ref] Based on these characteristics, re-RT using C-ions for head and neck malignancies can be considered to have efficacy against photon-resistant tumors and superiority over photon RT in terms of safety. In a multicentric in silico trial of re-RT for head and neck cancers, it was reported that C-ions showed better dose localization than protons and X-rays. [bib_ref] Benefit of particle therapy in re-irradiation of head and neck patients. Results..., Eekers [/bib_ref] Single institutional outcomes of re-RT using C-ions for head and neck malignancies have been reported in Germany, China, and Japan. Held et al., in a clinical study of patients previously treated with a course of irradiation including CIRT, reported that the median overall survival (OS) was 26.1 months, and 14.5% of patients experienced grade ≥3 late toxicity. [bib_ref] Carbon ion reirradiation for recurrent head and neck cancer: a single-institutional experience, Held [/bib_ref] Gao et al. reported results in patients previously treated with definitive photon-based RT and showed that the 1-year OS rate was 95.9%. [bib_ref] Salvage carbon-ion radiation therapy for locoregionally recurrent head and neck malignancies, Gao [/bib_ref] Hayashi et al. reported clinical results in patients previously treated with CIRT, and showed that the 2-year OS rate was 59.6%; 37.5% of patients experienced grade ≥3 late toxicities. [bib_ref] Feasibility of reirradiation using carbon ions for recurrent head and neck malignancies..., Hayashi [/bib_ref] In Japan, several CIRT centers, including the Hyogo Ion Beam Medical Center, QST Hospital, SAGA-HIMAT Foundation, and Gunma University Heavy Ion Medical Center, treat head and neck tumors. We conducted a multicentric study to assess retrospectively clinical data of CIRT for head and neck malignancies after photon RT (J-CROS study: 1903 HN).
# | methods
## | eligibility
Patients provided informed consent for the use of personal data. This study was approved by each of the relevant institutional review boards and was conducted in compliance with the Declaration of Helsinki. The identification number of this trial is UMIN000038950. Fifty-six patients were enrolled, each had undergone re-RT to the head and neck using C-ions between November 2003 and March 2019, treated previously with photon RT.
## | carbon-ion radiotherapy
The gross tumor volume (GTV) was based on computed tomography, positron emission tomography, and magnetic resonance imaging findings. The clinical target volume (CTV) was set as the GTV plus 0-5-mm margins. The planning target volume (PTV) had 2-5-mm margins around the CTV.
The doses of C-ions were shown as a photon-equivalent dose in Gy (RBE), which was defined as the physical dose multiplied by the RBE of the C-ion. [bib_ref] Preclinical biological assessment of proton and carbon ion beams at Hyogo ion..., Kagawa [/bib_ref]
## | clinical outcome and toxicity
We determined local control (LC) if the PTV showed no tumor regrowth. If neither local recurrence nor metastasis in the regional lymph nodes was observed, locoregional control was determined. Toxicities were evaluated based on the Common Terminology Criteria for Adverse Events, version 4.0, and data on grade ≥3 acute toxicity and grade ≥2 late toxicities were collected. All patients underwent restaging based on the eighth edition of the tumor-node-metastasis staging system (International Union Against Cancer, 2017).
## | statistical analyses
The LC, locoregional control, progression-free survival (PFS), and OS rates were computed from the first day of CIRT and analyzed using the Kaplan Meier method. All statistical tests were two-sided. Univariate analyses, using the log-rank test, were conducted to investigate prognostic factors for LC, PFS, and OS. All factors with statistically significant associations in the log-rank test were analyzed by the multivariate analysis using the Cox-proportional hazards model. Statistical significance was set at p < 0.05. The number of patients with grade ≥3 late toxicities was counted to determine the cumulative incidence using the Kaplan-Meier method. Statistical analyses were completed using R 4.0.3 (R Core Team). shows summaries of the patients' characteristics. The tumors treated with re-RT using C-ions were located in the sinonasal cavities (n = 20, 35.7%), skull base (n = 12, 21.4%), and orbit (n = 7, 12.5%). The tumors treated with the initial RT were located in the sinonasal cavities (n = 13, 23.2%), skull base (n = 9, 16.1%), and orbit (n = 9, 16.1%). The median period between initial RT and re-RT was 41 (4-568) months. Twenty-four patients (42.9%) were treated for <3 years from the initial RT. PTV overlap was confirmed in 48 patients (85.7%). The most common histology of the re-RT was squamous cell carcinoma (n = 11, 19.6%). Thirteen (23.2%) patients presented with a second primary tumor, while 43 (76.8%) presented with recurrence of the original tumors. The most commonly used radiation protocol was 57.6 Gy (RBE) in 16 fractions (n = 23, 41.1%). Surgery preceded re-RT in three patients (5.4%). Six patients received neoadjuvant chemotherapy, and these regimens included TS-1 + docetaxel (DTX), cisplatin (CDDP) + fluorouracil (5-FU), dacarbazine + nimustine + vincristine (DAV), nedaplatin (NDP) + 5-FU, cetuximab + CDDP + 5-FU, and unknown. One patient received concurrent chemotherapy with DAV.
# | results
## | cohort
## | clinical outcomes
The median follow-up time was 28 (3-147) months. The 2year LC, PFS, and OS rates were 66.5%, 36.9%, and 67.9%, respectively . The 2-year locoregional control rate was 52.3%. The 2-year cumulative incidence of distant metastasis was 31.0%. In patients with recurrent tumors, the 2-year LC, PFS, and OS rates were 62.0%, 35.2%, and 70.9%, respectively. In patients with second primary tumors, the 2-year LC, PFS, and OS rates were 84.6%, 42.3%, and 55.9%, respectively. No significant difference was observed between recurrent tumors and second primary tumors . shows the result of the analyses for each factor. The univariate analyses revealed that surgery prior to re-RT and the interval between initial RT and re-RT (< 36 months) were prognostic factors of a low LC. Additionally, the site of irradiation at initial RT (sinonasal cavities) and the interval between the initial RT and re-RT (<36 months), were prognostic factors of a low PFS and OS. In the multivariate analysis based on the Cox proportional hazards model, all aforementioned factors were identified as significant predictors of LC, PFS, and OS.
## | acute and late toxicities
Two patients (3.6%) developed grade ≥3 acute toxicities: one grade 3 acute dermatitis, and the other grade 3 acute pharyngeal mucositis. All 56 patients completed re-RT using C-ions. Regarding late toxicities, the 2-year cumulative incidence of grade ≥3 late toxicities was 25.2% using the Kaplan Meier method. . Fourteen patients (25.0%) developed grade ≥3 toxicities. Grade 5 dermatitis
## T a b l e 1 (continued)
with infection was confirmed in one patient. Grade 4 vision loss, hemorrhage, and mucositis developed in four, one, and one patient(s), respectively. Vision loss in three patients developed only on the affected side. All late toxicities with their details are summarized in Tables S1-S3.
# | discussion
Re-RT is challenging because of the complexity involved in the tolerance of various normal tissues. The decision between achieving therapeutic efficacy and minimizing toxicities must be carefully considered. Proton RT and CIRT are expected to improve dose conformity, [bib_ref] International recommendations on reirradiation by intensity modulated radiation therapy for locally recurrent..., Ng [/bib_ref] and further studies are required to evaluate the benefit of re-RT. This may be the first multi-institutional study to examine the clinical outcomes and toxicities of re-RT using C-ions for head and neck malignancies. It has been reported that re-RT for head and neck malignancies is tolerable and feasible. 9,10,14- [bib_ref] Reirradiation for recurrent head and neck cancers using charged particle or photon..., Yamazaki [/bib_ref] We evaluated the efficacy and toxicities in our cohort, and our findings demonstrated that re-RT using C-ions may represent an effective and tolerable treatment .
The most limiting factor of re-RT is grade 5 toxicities. McDonald et al. reported that the probability of carotid rupture was 2.6%, and 76% of them were grade 5 20 in patients receiving re-RT using photons. Grade 5 carotid ruptures were observed in both acute and late toxicities. Gao et al. reported that grade 5 mucosal necrosis led to hemorrhage in four patients receiving re-RT using C-ions as late toxicities. [bib_ref] Salvage carbon-ion radiation therapy for locoregionally recurrent head and neck malignancies, Gao [/bib_ref] In this study, one patient developed grade 5 dermatitis with infection as a late toxicity. As previously reported, severe skin necrosis led to death due to sepsis. [bib_ref] High-dose reirradiation with intensity-modulated radiotherapy for recurrent head-and-neck cancer: disease control, survival..., Duprez [/bib_ref] Grade 4 vision loss, hemorrhage, and mucositis developed in four, one and one patient(s), respectively. Vision loss can develop from the use of any type of beam. [bib_ref] Outcomes of visual acuity in carbon ion radiotherapy: analysis of dose-volume histograms..., Hasegawa [/bib_ref] [bib_ref] Analysis of vision loss caused by radiation-induced optic neuropathy after particle therapy..., Demizu [/bib_ref] [bib_ref] Radiationinduced optic neuropathy after stereotactic and image guided intensity-modulated radiation therapy (IMRT), Brecht [/bib_ref] Even with CIRT, it is sometimes difficult to avoid vision loss if an optic nerve is adjacent to a tumor. [bib_ref] Outcomes of visual acuity in carbon ion radiotherapy: analysis of dose-volume histograms..., Hasegawa [/bib_ref] [bib_ref] Analysis of vision loss caused by radiation-induced optic neuropathy after particle therapy..., Demizu [/bib_ref] All patients who had the potential to develop vision loss were informed about the risks of vision loss induced by re-RT.
Acute toxicity is also the most serious side-effect in RT. Held et al. reported that treatment was canceled due to acute laryngeal edema in patients receiving re-RT using C-ions. [bib_ref] Carbon ion reirradiation for recurrent head and neck cancer: a single-institutional experience, Held [/bib_ref] It has been reported that the completion rates of re-RT for recurrent head and neck cancer cases using photons, protons, and C-ions were 95%, [bib_ref] Refining patient selection for reirradiation of head and neck squamous carcinoma in..., Ward [/bib_ref] 94.6% 17 to 98.3%, [bib_ref] Reirradiation of head and neck cancers with proton therapy: outcomes and analyses, Phan [/bib_ref] and 96.9%, [bib_ref] Rare entities in head-andneck cancer: salvage re-irradiation with carbon ions, Held [/bib_ref] respectively. In this study, all 56 patients completed re-RT using C-ions.
Comparing the toxicities and effectiveness found in this study with those of previous studies is complicated because of this study's longer follow-up time (median follow-up, 28 months). As indicated in , there was one study with a comparable follow-up period reported by Held et al. (median follow-up, 28.5 months). [bib_ref] Carbon ion reirradiation for recurrent head and neck cancer: a single-institutional experience, Held [/bib_ref] Our study showed that the 2-year LC and OS rates on re-RT using C-ions were 66.5% and 67.9%, respectively. These clinical outcomes appear to be better than those reported by Held et al. (1.5-year LC, 44.7% and OS, 59.2%) but lead to a higher grade ≥3 late toxicity (25.0% vs. 14.5%). There is no clear solution to prioritize minimizing toxicities or achieving efficacy, and the treatment plan must be decided with consideration for an individual patient's condition.
Photon RT plus concurrent cisplatin treatment is the standard for head and neck squamous cell carcinoma. [bib_ref] Postoperative concurrent radiotherapy and chemotherapy for high-risk squamous-cell carcinoma of the head..., Cooper [/bib_ref] Concurrent chemotherapy was administered with re-RT using photons and protons. [bib_ref] Final report of RTOG 9610, a multi-institutional trial of reirradiation and chemotherapy..., Spencer [/bib_ref] [bib_ref] Refining patient selection for reirradiation of head and neck squamous carcinoma in..., Ward [/bib_ref] [bib_ref] Proton beam reirradiation for recurrent head and neck cancer: multi-institutional report on..., Romesser [/bib_ref] [bib_ref] Reirradiation of head and neck cancers with proton therapy: outcomes and analyses, Phan [/bib_ref] However, there are only a few reports on concomitant chemotherapy with CIRT. [bib_ref] Phase I/II trial evaluating concurrent carbon-ion radiotherapy plus chemotherapy for salvage treatment..., Kong [/bib_ref] [bib_ref] Phase I dose-escalation trial of s-1 combined with carbon-ion radiotherapy for sinonasal..., Takahashi [/bib_ref] [bib_ref] Carbon-ion radiotherapy combined with chemotherapy for head and neck mucosal melanoma: prospective..., Takayasu [/bib_ref] [bib_ref] Phase II study of induction chemotherapy with TPF followed by radioimmunotherapy with..., Jensen [/bib_ref] In this study, only one patient with malignant melanoma received concurrent treatment with dacarbazine, nimustine, and vincristine (DAV therapy), as previously reported. [bib_ref] Carbon-ion radiotherapy combined with chemotherapy for head and neck mucosal melanoma: prospective..., Takayasu [/bib_ref] Our clinical outcomes were not clearly inferior to outcomes of re-RT using protons in combination with concurrent chemotherapy. [bib_ref] Proton beam reirradiation for recurrent head and neck cancer: multi-institutional report on..., Romesser [/bib_ref] [bib_ref] Reirradiation of head and neck cancers with proton therapy: outcomes and analyses, Phan [/bib_ref] Gao et al. published the 1-year results of re-RT using C-ions, including a cohort with concurrent chemotherapy, and long-term data are expected. [bib_ref] Salvage carbon-ion radiation therapy for locoregionally recurrent head and neck malignancies, Gao [/bib_ref] Proton beams have similar physical properties to those of CIRT and may be useful for re-irradiation. [bib_ref] Proton beam reirradiation for recurrent head and neck cancer: multi-institutional report on..., Romesser [/bib_ref] [bib_ref] Reirradiation of head and neck cancers with proton therapy: outcomes and analyses, Phan [/bib_ref] However, the rapid distal fall-off and sharper lateral penumbra of CIRT can achieve more conformal irradiation than that in proton therapy. [bib_ref] Biophysical characteristics of HIMAC clinical irradiation system for heavy-ion radiation therapy, Kanai [/bib_ref] [bib_ref] Benefit of particle therapy in re-irradiation of head and neck patients. Results..., Eekers [/bib_ref] Therefore, CIRT is considered more F I G U R E 1 Kaplan-Meier curves of local control, progressionfree survival, and overall survival following re-irradiation using carbon ions for head and neck malignancies. LC, local control; PFS, progression-free survival; OS, overall survival advantageous for malignancies located close to organs at risk. There was no significant difference in LC, PFS, and OS between patients with recurrent tumors and those with second primary tumors. Our results indicate that a longer RT interval significantly improved OS, PFS, and LC, a finding also informed by Ward et al. [bib_ref] Refining patient selection for reirradiation of head and neck squamous carcinoma in..., Ward [/bib_ref] and Held et al. [bib_ref] Carbon ion reirradiation for recurrent head and neck cancer: a single-institutional experience, Held [/bib_ref] [bib_ref] Rare entities in head-andneck cancer: salvage re-irradiation with carbon ions, Held [/bib_ref] Patients who underwent surgery prior to re-RT had a significantly lower LC rate, but there were only a few such patients. The site of the initial RT (sinonasal cavities) was identified as a significant prognostic factor of worse PFS and OS. Because this factor did not significantly affect LC, it is possible that this significant difference was not correlated with CIRT. Further studies are required to understand these two factors.
Our study has two limitations: our data were retrospectively analyzed and various radiation dosages were adopted, which may have influenced the clinical outcomes.
# | conclusions
Our findings suggest that re-RT using C-ions for head and neck malignancies after photon RT is an effective treatment with tolerable toxicities. Further investigations, preferably in prospective trials, are required for greater reliability.
# Acknowledgments
|
Transcription Factor SCL Is Required for c-kit Expression and c-Kit Function in Hemopoietic Cells
In normal hemopoietic cells that are dependent on specific growth factors for cell survival, the expression of the basic helix-loop-helix transcription factor SCL/Tal1 correlates with that of c-Kit, the receptor for Steel factor (SF) or stem cell factor. To address the possibility that SCL may function upstream of c-kit, we sought to modulate endogenous SCL function in the CD34 ϩ hemopoietic cell line TF-1, which requires SF, granulocyte/macrophage colony-stimulating factor, or interleukin 3 for survival. Ectopic expression of an antisense SCL cDNA (as-SCL) or a dominant negative SCL (dn-SCL) in these cells impaired SCL DNA binding activity, and prevented the suppression of apoptosis by SF only, indicating that SCL is required for c-Kit-dependent cell survival. Consistent with the lack of response to SF, the level of c-kit mRNA and c-Kit protein was significantly and specifically reduced in as-SCL-or dn-SCLexpressing cells. c-kit mRNA, c-kit promoter activity, and the response to SF were rescued by SCL overexpression in the antisense or dn-SCL transfectants. Furthermore, ectopic c-kit expression in as-SCL transfectants is sufficient to restore cell survival in response to SF. Finally, enforced SCL in the pro-B cell line Ba/F3, which is both SCL and c-kit negative is sufficient to induce c-Kit and SF responsiveness. Together, these results indicate that c-kit, a gene that is essential for the survival of primitive hemopoietic cells, is a downstream target of the transcription factor SCL.
CL, also known as TAL1 or Tcl-5, is associated with 25% of chromosomal rearrangements in childhood T cell acute lymphoid leukemia [bib_ref] Disruption of the SCL gene by a t(1;3) translocation in a patient..., Aplan [/bib_ref] [bib_ref] Chromosomal translocation in a human leukemic stem-cell line disrupts the T-cell antigen..., Begley [/bib_ref] [bib_ref] Two distinct mechanisms for the SCL gene activation in the t(1; 14)..., Bernard [/bib_ref] [bib_ref] The tal gene undergoes chromosome translocation in T cell leukemia and potentially..., Chen [/bib_ref] [bib_ref] Involvement of the TCL5 gene on human chromosome 1 in T-cell leukemia..., Finger [/bib_ref]. SCL codes for a hemopoietic-specific transcription factor of the basic helix-loophelix family, and can enhance or suppress transcription [bib_ref] Positive and negative transcriptional control by the TAL1 helixloop-helix protein, Hsu [/bib_ref]. SCL expression is detected in primitive pluripotent hemopoietic precursors [bib_ref] Opposing effects of the basic helix-loop-helix transcription factor SCL on erythroid and..., Hoang [/bib_ref] , in a subset of CD34 ϩ cells, in mast cells and megakaryocytes [bib_ref] Expression of TAL-1 proteins in human tissues, Pulford [/bib_ref] [bib_ref] Expression of tal-1 and GATA-binding proteins during human hematopoiesis, Mouthon [/bib_ref] and in maturing erythroid cells [bib_ref] The SCL gene product: a positive regulator of erythroid differentiation, Aplan [/bib_ref] [bib_ref] The gene SCL is expressed during early hematopoiesis and encodes a differentiation-related..., Begley [/bib_ref] [bib_ref] SCL is coexpressed with GATA-1 in hemopoietic cells but is also expressed..., Green [/bib_ref] [bib_ref] Erythroid expression of the 'helix-loop-helix' gene, Green [/bib_ref] [bib_ref] Steel factor affects SCL expression during normal erythroid differentiation, Miller [/bib_ref]. Consistent with a role for SCL in establishing the hemopoietic lineage, a targeted disruption of the SCL gene specifically abrogates blood cell formation, resulting in embryonic lethality at day E8.5 [bib_ref] Absence of yolk sac hematopoiesis from mice with a targeted disruption of..., Robb [/bib_ref] [bib_ref] Absence of blood formation in mice lacking the T-cell leukaemia oncoprotein tal-1/SCL, Shivdasani [/bib_ref]. Moreover, scl Ϫ / Ϫ embryonic stem cells fail to contribute to hemopoietic stem cells and blood cells in chimeric mice, further suggesting a critical role of SCL for the development of all hemopoietic lineages [bib_ref] The T cell leukemia oncoprotein SCL/tal-1 is essential for development of all..., Porcher [/bib_ref] [bib_ref] The SCL gene product is required for the generation of all hematopoietic..., Robb [/bib_ref]. Together, these results suggest that SCL is required for mesodermal cell commitment into hemopoietic stem cells and/or is essential for the survival of hemopoietic stem cells.
SCL binds DNA only upon interaction with the ubiquitous basic helix-loop-helix transcription factors E2A or HEB [bib_ref] Preferred sequences for DNA recognition by the TAL1 helix-loop-helix proteins, Hsu [/bib_ref]. The SCL-E2A heterodimer preferentially binds the TAL1 consensus AA CAGATG GT defined by in vitro binding site selection (CASTing). SCL represses the E2A-dependent activity directed from the TAL1 consensus in front of a minimal promoter, but relieves the inhibition conferred by Id1 on E2A [bib_ref] Positive and negative transcriptional control by the TAL1 helixloop-helix protein, Hsu [/bib_ref]. However, the TAL1 consensus has not been found in any known hemopoietic pro-S moter. SCL also associates with Lmo-2 or RBTN-2, a RING finger protein [bib_ref] The LIM protein RBTN2 and the basic helix-loop-helix protein TAL1 are present..., Valge-Archer [/bib_ref] [bib_ref] Specific in vivo association between the bHLH and LIM proteins implicated in..., Wadman [/bib_ref] [bib_ref] The LIMonly protein Lmo2 is a bridging molecule assembling an erythroid, DNA-binding..., Wadman [/bib_ref] [bib_ref] The oncogenic cysteinerich LIM domain protein rbtn2 is essential for erythroid development, Warren [/bib_ref] in a multiprotein complex that includes E2A, GATA-1, an erythroid zinc finger transcription factor [bib_ref] Cloning of cDNA for the major DNA-binding protein of the erythroid lineage..., Tsai [/bib_ref] , and a novel LIM-binding protein, Ldb1 [bib_ref] The LIMonly protein Lmo2 is a bridging molecule assembling an erythroid, DNA-binding..., Wadman [/bib_ref] , and selectively binds an Ebox-GATA motif also defined by cycle amplification and selection of targets (CASTing). Although similar motives are found on two erythroid promoters, those of the glycophorin A and B genes and the porphobilinogen deaminase gene, their contribution to promoter activity in response to SCL has not been established [bib_ref] The LIMonly protein Lmo2 is a bridging molecule assembling an erythroid, DNA-binding..., Wadman [/bib_ref]. Finally, in T-ALL, SCL upregulates the expression of a cell surface marker, TALLA-1 [bib_ref] Transcriptional activity of TAL1 in T cell acute lymphoblastic leukemia (T-ALL) requires..., Ono [/bib_ref]. However, it is not clear whether TALLA-1 is a direct target of SCL. Thus, despite observations that indicate a crucial role for SCL in normal hemopoiesis and T cell leukemogenesis, no natural binding site or SCL target gene has been as yet reported.
The survival of hemopoietic cells both in culture and in vivo is critically dependent on the presence of hemopoietic growth factors that act to suppress apoptosis [bib_ref] Product of the steel locus suppresses apoptosis in hemopoietic cells. Comparison with..., Caceres-Cortes [/bib_ref] [bib_ref] Role of protein kinase C and the Na ϩ /H ϩ antiporter..., Rajotte [/bib_ref] [bib_ref] Haemopoietic colony stimulating factors promote cell survival by suppressing apoptosis, Williams [/bib_ref]. The importance of such growth factors for blood cell development was first revealed by the genetics of hereditary anemias in mice. Perhaps the best studied examples are white spotting and Steel mice that are severely anemic due to mutations in the genes encoding the tyrosine kinase receptor c-kit [bib_ref] The proto-oncogene c-kit encoding a transmembrane tyrosine kinase receptor maps to the..., Chabot [/bib_ref] or its ligand Steel factor (SF), 1 respectively [bib_ref] Regulation of p53-mediated apoptosis and cell cycle arrest by Steel factor, Abrahamson [/bib_ref] [bib_ref] Mast cell growth factor maps near the steel locus on mouse chromosome..., Copeland [/bib_ref]. Severe mutations result in lethality in homozygotes (for review see references [bib_ref] From white spots to stem cells: the role of the Kit receptor..., Fleischman [/bib_ref] [bib_ref] Hereditary anemias of the mouse: a review for geneticists, Russel [/bib_ref] , indicating a crucial role for these two genes in hemopoiesis. We and others have previously shown that SF suppresses apoptosis in early hemopoietic cells that express the surface antigen CD34 [bib_ref] Regulation of p53-mediated apoptosis and cell cycle arrest by Steel factor, Abrahamson [/bib_ref] [bib_ref] Stem cell factor enhances the survival but not the self-renewal of murine..., Li [/bib_ref] , and in the CD34 ϩ cell line TF-1 [bib_ref] Product of the steel locus suppresses apoptosis in hemopoietic cells. Comparison with..., Caceres-Cortes [/bib_ref]. SF has also been shown to synergize with IL-7 in delaying apoptosis in primitive thymocytes [bib_ref] Phenotypic and functional characterization of c-kit expression during intrathymic T cell development, Godfrey [/bib_ref] [bib_ref] Thymic hyperplasia in transgenic mice caused by immortal epithelial cells expressing c-kit..., Moll [/bib_ref] [bib_ref] Developmentally regulated cell surface expression and function of c-kit receptor during lymphocyte..., Palacios [/bib_ref].
The expression pattern of SCL in primary hemopoietic cells parallels that of c-kit [bib_ref] Analysis of gene expression in a complex differentiation hierarchy by global amplification..., Brady [/bib_ref] , suggesting that SCL functions in concert with c-Kit. Furthermore, a two-to fourfold augmentation of SCL protein was observed in maturing erythroid progenitors stimulated with SF [bib_ref] Steel factor affects SCL expression during normal erythroid differentiation, Miller [/bib_ref]. However, it is not known whether upregulation of SCL is a cause or consequence of c-Kit activation, cell survival, or cell proliferation. Because of the correlation between c-Kit and SCL expression, we directly investigated the relationship of SCL function relative to c-Kit through the attenuation of SCL protein levels in TF-1 cells using an antisense SCL (as-SCL) expression vector, or through expression of a dominant negative SCL (dn-SCL). We reasoned that SCL function is likely to be determined by the nature of its binding partners, which have yet to be defined in CD34 ϩ cells. We therefore chose to investigate this through the disruption of endogenous SCL function in the CD34 ϩ hemopoietic cell line TF-1, which should express appropriate SCL binding partners. Moreover, TF-1 has retained one of the most im-portant characteristics of primary hemopoietic cells [bib_ref] Establishment and characterization of a unique human cell line that proliferates dependently..., Kitamura [/bib_ref] , i.e., their requirement in hemopoietic growth factors for cell survival [bib_ref] Product of the steel locus suppresses apoptosis in hemopoietic cells. Comparison with..., Caceres-Cortes [/bib_ref] [bib_ref] Interleukin-1 beta suppresses apoptosis in CD34 positive bone marrow cells through activation..., Rodriguez [/bib_ref]. Our approach to defining SCL function and SCL target(s) was to disrupt SCL in TF-1 cells and to screen stable transfectants with a functional assay, i.e., cell survival in response to different growth factors.
# Materials and methods
Cell Lines, Growth Factors, and Antibodies . The TF-1 cell line [bib_ref] Establishment and characterization of a unique human cell line that proliferates dependently..., Kitamura [/bib_ref] was a gift from Dr. T. Kitamura (DNAX, Palo Alto, CA). The cells were maintained in IMDM (GIBCO BRL, Gaithersburg, MD) supplemented with 10% FCS (GIBCO BRL) and 5 ng/ml human GM-CSF. The cells were passaged every second day at 1.5 ϫ 10 5 /ml. Jurkat T cells were also maintained in IMDM supplemented with FCS (10%), and were passaged three times weekly at a concentration of 1.5 ϫ 10 5 /ml.
Purified recombinant human GM-CSF was a gift from Dr. Steve Clark (Genetics Institute, Cambridge, MA) and purified recombinant SF was from Dr. K. Langley (Amgen, Thousand Oaks, CA). The monoclonal antibody 2TL136 specific for human SCL was provided by Dr. Danièle Mathieu-Mahul (IN-SERM, Marseille, France; references 10, 42). The monoclonal antibody ACK2 is specific for murine c-Kit (GIBCO BRL). The monoclonal mouse anti-human c-Kit was from Cedarlane Labs. (Hornby, Ontario, Canada). The monoclonal mouse anti-rabbit eIF-4E cross-reacts with the human protein (Transduction Labs., Lexington, KY). The monoclonal anti-CD45 (clone 4B2; ATCC number HB 196, American Type Culture Collection, Rockville, Maryland) was used at 1:10 dilution of hybridoma supernatant. Monoclonal anti-hGM-CSF receptor  chain was purchased from Upstate Biotechnology, Inc. (Lake Placid, NY) and used at a 1:100 dilution. Annexin V was purchased from Biodesign (Kennebunk, ME), and used at a concentration of 400 mg/ml.
Retrovirus Production and Infection. The human SCL cDNA, the dn-SCL devoid of DNA binding domain [bib_ref] The SCL gene product: a positive regulator of erythroid differentiation, Aplan [/bib_ref] and the as-SCL were all cloned in the EcoRI site of the murine stem cell virus (MSCV)-neo vector [bib_ref] The Hox-11 homeobox-containing gene of human leukemia immortalizes murine hematopoietic precursors, Hawley [/bib_ref]. High titer amphotropic viruses (10 6 PFU/ml) were produced by transient transfection into BING cells [bib_ref] Production of high-titer helper-free retroviruses by transient transfection, Pear [/bib_ref]. The cells were irradiated and used for coculture with TF-1 cells for 48 h. Murine c-kit cloned in the LXSN retroviral vector was a gift from Dr. A. Bernstein (Samuel Lunenfeld Research Institute, Toronto, Ontario, Canada). Ecotropic viruses were produced by transfection into the BOSC23 packaging cell line [bib_ref] Production of high-titer helper-free retroviruses by transient transfection, Pear [/bib_ref].
For retroviral infection, 10 6 exponentially growing cells (TF-1 or A31) were presensitized with polybrene at 2 g/ml for 24 h and cocultured with virus-producing clones for another 24 h. Nonadherent TF-1 cells were separated from the infected fibroblasts. A polyclonal population was analyzed 7 d after selection in G418 at 1 mg/ml. For long-term experiments, cloning was performed through limiting dilution [bib_ref] The SCL gene product: a positive regulator of erythroid differentiation, Aplan [/bib_ref].
Plasmids and Transfection . The human SCL cDNA was cloned in the antisense orientation under the control of the metallothionein promoter by digesting the plasmid pMSCL [bib_ref] The SCL gene product: a positive regulator of erythroid differentiation, Aplan [/bib_ref] with EcoRI and religation. The EcoRI sites come from the plasmid polylinker.
Stable TF-1 transfectants harboring the as-SCL were obtained through Lipofectin-mediated DNA transfer (GIBCL BRL) as previously described [bib_ref] The SCL gene product: a positive regulator of erythroid differentiation, Aplan [/bib_ref]. Cells were cloned immediately after gene transfer by limiting dilution, and the selective pressure was applied the following day, at a concentration of 1 mg/ml G418 (GIBCO BRL). GM-CSF was present throughout the gene transfer procedure. After selection, G418 concentrations were lowered to 400 g/ml in order to expand the cells and prepare a large stock of frozen cells immediately after characterization. Each cell line was kept for not more than 2 mo in culture.
Nuclear Staining for SCL. 3 ϫ 10 5 TF-1 and as-SCL transfectants were fixed in 500 l of Bouin's fixative for 15 min at room temperature. The cells were then pelleted at 200 g for 5 min and resuspended in 200 l of 0.2% Triton X-100, followed immediately with 800 l of PBS. The cells were pelleted again, washed with 1 ml of PBS, and resuspended in 100 l of 1:10 dilution of anti-SCL monoclonal antibody (clone BTL73, provided by Dr. Danièle Mathieu-Mahul). After 30 min of incubation on ice, cells were washed twice with 1 ml of PBS followed by a 10-min wash in 1 ml of PBS, resuspended in 100 l of FITC-coupled goat anti-mouse antibody at the recommended dilution (Caltech, San Francisco, CA), and incubated for 30 min at 4 Њ C. The cells were washed as above before flow cytometry analysis.
Surface Marker Staining. For surface marker staining, 3-5 ϫ 10 5 cells were washed once with PBS supplemented with 2% FBS and 0.05% sodium azide (staining buffer), labeled with the primary antibody for 30 min on ice in a total volume of 100 ml, washed three times with 1 ml of staining buffer, and labeled with a biotinylated goat anti-mouse antibody (Cedarlane, Labs.) in staining buffer supplemented with 1% normal goat serum (Sigma-Aldrich, Oakville, Ontario, Canada) for 30 min on ice. After washing, the cells were incubated with PE-conjugated streptavidin (Sigma) for an additional 30 min on ice, washed, and analyzed on the FACScan ® flow cytometer (Becton Dickinson, San Jose, CA).
Northern Blotting Analysis. Cells were homogenized in the presence of guanidium and RNA isolated by acid-phenol extraction as previously described [bib_ref] Single-step method of RNA isolation by acid guanidinium thiocyanate-phenolchloroform extraction, Chomczynski [/bib_ref]. Northern blotting and hybridization were performed as previously described [bib_ref] Opposing effects of the basic helix-loop-helix transcription factor SCL on erythroid and..., Hoang [/bib_ref]. Probes used were the 1.2-kb Hind-XbaI human SCL cDNA fragment (nucleotides 109-1293; reference 12), the 1.4-kb SalI-XbaI fragment of human GATA-1 cDNA (26), a 1.5-kb DraI-KpnI human c-kit cDNA fragment from human c-kit-BS plasmid (gift from Dr. J. Simms, Immunex, Seattle, WA), and the 1.2-kb PstI-XbaI fragment of the rat glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA [bib_ref] Isolation and characterization of rat and human glyceraldehyde-3-phosphate dehydrogenase cDNAs: genomic complexity..., Tso [/bib_ref]. Blots were exposed to a Phosphor-Imager screen for quantitation.
Western Blotting Analysis. Cell lysates were prepared as described previously [bib_ref] Loss of TAL-1 protein activity induces premature apoptosis of Jurkat leukemic T..., Leroy-Viard [/bib_ref]. Protein concentrations were determined with the Bio-Rad Protein Assay reagent (Bio-Rad, Hercules, CA). Equal amounts of proteins (20 g) were loaded on a 12% SDS-polyacrylamide gel and transferred to nitrocellulose using the mini "Transblot" (Bio-Rad) for 1 h at 100 V.
Membranes were blocked in 5% nonfat dry milk and 1% BSA, incubated with 2TL13G hybridoma tissue culture medium (monoclonal anti-SCL) diluted 1:300 for 2.5 h at room temperature, washed, incubated with a goat anti-mouse alkaline phosphataselinked antibody (Bio-Rad), diluted 1:1,000 for 1.5 h, and washed extensively. Western blots were then developed by incubating membranes in the dark with 10 ml of BCIP (165 g/ml) and NBT (330 g/ml) (GIBCO BRL) for 2 min, and the reaction was terminated by several washes with water as described previously [bib_ref] Loss of TAL-1 protein activity induces premature apoptosis of Jurkat leukemic T..., Leroy-Viard [/bib_ref]. In parallel, Western blotting was performed with a monoclonal antibody against the eukaryotic translation initiation factor (eIF-4E) (Transduction Labs.) as a control for protein loading.
## Electrophoretic mobility shift assays
Nuclear extracts were prepared as previously described [bib_ref] The atrial natriuretic factor promoter is a downstream target for Nkx-2.5 in..., Durocher [/bib_ref]. Protein concentrations were assayed with the Bio-Rad Protein Assay reagent. Binding reactions for electrophoretic mobility shift assays were performed as previously described [bib_ref] Loss of TAL-1 protein activity induces premature apoptosis of Jurkat leukemic T..., Leroy-Viard [/bib_ref]. In brief, the binding reactions were allowed to proceed at room temperature for 15 min in the presence of 0.5 g of poly(dI-dC) as nonspecific competitor DNA in 20 mM Hepes (pH 7.5), 50 mM KCl, 1 mM dithiothreitol, 1 mM EDTA, 5% glycerol, 10 g BSA, 100,000 cpm double-stranded synthetic oligonucleotide, and 25 g nuclear extract in a total volume of 10 l. 50-fold molar excess of unlabeled oligonucleotides were used for self-competition and added before nuclear extracts. For antibody supershift assays, 200 ng of affinity-purified anti-E2A (Santa Cruz Biotechnology, Santa Cruz, CA), 1 l of the monoclonal anti-SCL antibody (BTL73; reference 48), or 1 l of an unrelated antibody were added to the binding reaction. Protein complexes were resolved by electrophoresis on 4% polyacrylamide gel buffered in 0.25 ϫ Tris-borate-EDTA, 195 mM glycine (pH 8.5) at 4 Њ C. The sequences of the probes (coding strand) are: ACCTGAA CAGATG GTCGGCT TAL1 consensus [bib_ref] Preferred sequences for DNA recognition by the TAL1 helix-loop-helix proteins, Hsu [/bib_ref] ; CTAGGGAG CACCTG CCAGGTG GCTGGCCC murine c-kit probe [bib_ref] Involvement of transcriptional factor encoded by the mi locus in the expression..., Tsujimura [/bib_ref] ; and an E box probe derived from the rat POMC E box also referred to as DE2c [bib_ref] Cell-specific helix-loophelix factor required for pituitary expression of the pro-opiomelanocortin gene, Therrien [/bib_ref].
c-kit Promoter Constructs, Transfection Protocols, and Luciferase Assay. The human c-kit promoter was cloned by PCR from genomic DNA to generate a fragment that extends from 1146 bp upstream of the transcription initiation site to 43 bp downstream. In brief, genomic DNA was amplified in two steps, the first PCR providing a fragment that extends from position Ϫ 634 to ϩ 60 (PCR1), and the second PCR a fragment that covers Ϫ 1146 to Ϫ 379 (PCR2). PCR1 was cut with SalI and BamHI and PCR2 with HindIII (site added) and SalI, and the fragments were cloned into the HindIII and BglII sites of the promoterless luciferase expression vector pXPII to generate kit 1146. The chimeric construct was verified through sequencing. TF-1 and A31 cell lines were transfected by electroporation. Cells were passaged 24 h before gene transfer at a concentration of 3 ϫ 10 5 cells/ml. Exponentially growing cells were then concentrated at 2.5 ϫ 10 7 cells/ml and electroporated at 900 F and 350 mV using a Bio-Rad electroporator with 12 g of reporter DNA and 1.5 g of CMV  -galactosidase used as an internal control for the experiment, with or without different molar ratios of MSCV-SCL as shown; the total amount of transfected DNA was then filled to 25 g with pGEM4 as a carrier. Cell lysates were prepared 15 h after transfection, normalized for  -galactosidase content, and assayed for luciferase activity. Rous Sarcoma virus-driven luciferase (RSV-luc) was used as an external control for all the transfections and pXP2 as a negative control as described previously [bib_ref] The atrial natriuretic factor promoter is a downstream target for Nkx-2.5 in..., Durocher [/bib_ref].
Apoptosis Assays. Analysis of DNA fragmentation by agarose gel electrophoresis was performed as previously described [bib_ref] Product of the steel locus suppresses apoptosis in hemopoietic cells. Comparison with..., Caceres-Cortes [/bib_ref]. In brief, equal cell numbers were lysed in Sarkosyl buffer and treated with Proteinase K and RNase A before loading on a 1.2% agarose gel (20 cm), followed by a 16-h electrophoresis at 40 V. Cell viability was independently confirmed through trypan blue exclusion and counting [bib_ref] Product of the steel locus suppresses apoptosis in hemopoietic cells. Comparison with..., Caceres-Cortes [/bib_ref] [bib_ref] Role of protein kinase C and the Na ϩ /H ϩ antiporter..., Rajotte [/bib_ref] , or the double fluorochrome staining assay, as described below. Apoptosis was also assessed by flow cytometry analysis of cells labeled with Annexin V-FITC (1 g/ml), shown previously to be an early marker of apoptosis [bib_ref] Annexin V for flow cytometric detection of phosphatidylserine expression on B cells..., Koopman [/bib_ref] [bib_ref] Early redistribution of plasma membrane phosphatidylserine is a general feature of apoptosis..., Martin [/bib_ref]. Immediately before acquisition, 1 ml of propidium iodide solution (1 mg/ml) was added in order to detect dead cells.
Double Fluorochrome Staining Assay. A stock solution of dye mix containing 100 g/ml acridine orange and 100 g/ml ethidium bromide was prepared in PBS. Cell were resuspended at 5 ϫ 10 5 to 5 ϫ 10 6 in IMDM. 1 l of dye mix was added to 25 l of cell suspension, and 10 l of the mixture was placed on a microscope slide for examination by fluorescent microscopy (Leica, Wetzlar, Germany). Cells with bright green chromatin were considered viable cells while those with bright orange chromatin or collapsed chromatin were considered dead cells. A minimum of 200 cells were counted.
# Results
Decreased SCL Protein Levels and c-Kit Levels in TF-1 Cells Expressing an as-SCL. To address SCL function and define SCL target genes in primitive hemopoietic cells, we chose to disrupt SCL function in the CD34 ϩ /c-kit ϩ cell line TF-1. To this end, we took two complementary approaches. The first one consists of expressing an as-SCL that interferes with SCL protein levels, and the second one involves a dn-SCL devoid of its DNA binding domain [bib_ref] The SCL gene product: a positive regulator of erythroid differentiation, Aplan [/bib_ref] , which prevents SCL function without affecting protein levels. We reasoned that loss of SCL function may be detrimental to the cells and therefore decided to analyze the cells as early as 1 wk after gene transfer, at a time when the selective pressure (G418) is already optimal. To minimize cell loss during the selective pressure and, consequently, the generation time required for in vitro expansion, we chose retroviral infection as a method for high efficiency gene delivery for which we typically obtain 30-50% transfer efficiency.
We first determined the efficiency of as-SCL in disrupting SCL protein levels, which were assessed through immunostaining of permeabilized cells with the monoclonal anti-TAL1 antibody BTL73 [fig_ref] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL [/fig_ref]. As expected, TF-1 cells express high levels of endogenous SCL. In contrast, 1 wk after transduction of as-SCL the endogenous SCL protein level was significantly decreased, with little if any overlap between the two populations. The cells were further analyzed for expression of several cell surface markers. Data shown in [fig_ref] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL [/fig_ref] indicate that the fluorescence intensity for c-Kit (CD117) was at least three-to fourfold lower in dn-SCL and as-SCL transfectants when compared with parental TF-1 cells (data not shown) or control cells expressing the vector alone ( MSCV ). In contrast, the level of CD45 was not affected by as-SCL or dn-SCL expression and that of CD116 (GM-CSF receptor ␣ chain) was either similar or slightly higher. Together, the results indicate that decreased SCL protein levels ( as-SCL ) or decreased SCL function ( dn-SCL ) specifically lower c-Kit levels without affecting other surface markers.
Since the GM-CSF receptor is not affected by SCL levels, we proceeded to derive stable TF-1 cell lines expressing dn-SCL (TF1-dn) following long term selection in G418 in the presence of GM-CSF as a survival factor, in order to directly address SCL function at the molecular and cellular levels. We also established clones that express as-SCL driven by a weaker promoter, the mouse metallothionein promoter, instead of the strong promoter contained within the retroviral LTR. Despite this, most of the 50 clones screened displayed weak levels of transgene expression, consistent with the reported difficulty in obtaining stable as-SCL clones in another hemopoietic cell line, K562 (9, 53; and Kirsch, I.R., unpublished data). The two highest antisense expressor clones, A30 and A31, were selected.
We next confirmed that SCL DNA binding activity was indeed reduced in the antisense clone A31 and the dn-SCL-expressing transfectants (TF1-dn) through electrophoretic mobility shift assays using the TAL1 consensus probe (reference 21; . TF-1 nuclear extracts produced four specific complexes (C1-C4), as shown by selfcompetition [fig_ref] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL [/fig_ref]. In contrast, a probe with divergent sequences (mDE2C) did not compete for binding (data not shown), indicating that these complexes were specific. All four complexes were also found using nuclear extracts from Jurkat cells [fig_ref] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL [/fig_ref] and TF-1 transfectants harboring the vector alone (TF1neo, [fig_ref] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL [/fig_ref]. The identity of each complex (C1-C4) was verified by antibody supershifting. Preincubation of TF1-neo extracts with either anti-SCL or anti-E2A antibody supershifted the two most slowly migrating complexes, C1 and C2 [fig_ref] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL [/fig_ref] , indicating that these contain SCL/E2A heterodimers as previously described [bib_ref] Preferred sequences for DNA recognition by the TAL1 helix-loop-helix proteins, Hsu [/bib_ref] [bib_ref] The LIMonly protein Lmo2 is a bridging molecule assembling an erythroid, DNA-binding..., Wadman [/bib_ref]. These complexes were absent in HL-60 cells, which were included as a negative control (lane 16).
In the antisense-expressing clone A31, the SCL-containing complex C1 decreased significantly, whereas the C2 complex was only marginally affected [fig_ref] Figure 5: Survival response to SF rescued by ectopic murine c-kit expression in as-SCL... [/fig_ref] , suggesting that under conditions where SCL is limiting, the smaller complex, C2, is preferentially formed over C1. Rescuing clone A31 wild-type SCL (A31-SCL) restored C1 and C2 binding to the TAL1 probe [fig_ref] Figure 4: as-SCL or dn-SCL prevents cell survival in response to SF [/fig_ref]. In all cell lines, the two fastest migrating complexes, C3 and C4, were not supershifted with anti-SCL or anti-E2A, suggesting that these did not contain the corresponding proteins. These complexes were not affected in TF1-neo but were decreased in A31, possibly as a consequence of clonal variation, since they were also decreased in A31-SCL, a subclone of A31.
In the transfectant expressing a dn-SCL, which is a polyclonal population, both SCL-containing complexes, C1 and C2, were significantly decreased relative to TF1-neo whereas the non-SCL-containing complexes C3 and C4 were not affected [fig_ref] Figure 5: Survival response to SF rescued by ectopic murine c-kit expression in as-SCL... [/fig_ref]. Together, our results indicate that ectopic expression of as-SCL or dn-SCL significantly interferes with the formation of the two SCL-containing complexes, C1 and C2, on the TAL1 probe.
SCL Protein Levels Specify c-kit Expression. As observed with transient transfectants [fig_ref] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL [/fig_ref] , flow cytometry analysis . Decreased SCL DNA binding activity in TF-1 cells expressing as-SCL (A31) or a dn-SCL (TF1-dn). TF-1 cells and the different transfectants were maintained in GM-CSF containing medium. Exponentially growing cells were treated with ZnCl 2 for 24 h before preparation of nuclear extracts. A double-stranded 32 P-labeled TAL1 consensus probe was incubated with 25 g nuclear extracts from TF-1, TF1-neo, A31, A31-SCL, TF1-dn, Jurkat, and HL-60 as described in Materials and Methods. Unlabeled double-stranded oligonucleotide probes used as competitors were added at 50-fold molar excess. SCL-containing complexes (black arrows) formed on the TAL1 consensus probe were supershifted (gray arrows) by the monoclonal antibody anti-SCL/TAL1 or anti-E2A, which were added to the binding reaction. These complexes were displaced by a 50-fold molar excess of self (lanes 4, 13, and 17) and of the c-kit probe but not by a control probe with divergent sequences, mDE2C (data not shown). of the A31 and TF-1 cells or TF1-dn and TF1neo (data not shown) using a monoclonal antibody to human c-Kit indicated a quantitative decrease in c-Kit levels in A31 and TF1-dn as compared with control cells. There was a direct correlation with c-kit mRNA levels as determined through Northern blotting . Thus, parental TF-1 cells or TF-1 cells harboring the vector alone (TF1pac and TF1-neo) exhibited comparable levels of c-kit mRNA . In contrast, c-kit expression was significantly lower in A31 and T-dn . To exclude the possibility that c-kit mRNA expression was affected by the site(s) of integration in these antisense clones, we attempted to rescue c-kit expression by wild-type SCL through retrovirus-mediated gene transfer (MSCV; reference 43), using puromycin resistance as a second selective marker. Expression of the puromycin resistance gene alone did not affect c-kit expression in TF-1 cells [fig_ref] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL [/fig_ref] pac) nor in A31 cells (data not shown). In contrast, elevating SCL expression in clone A31 restored c-kit expression [fig_ref] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL [/fig_ref] and further increased c-kit mRNA levels in parental TF-1 cells (data not shown). Together, the results indicate that the targeted attenuation of SCL protein levels and DNA binding activity directly leads to decreased c-kit expression.
Sequence analysis of the c-kit promoter revealed two adjacent E boxes at Ϫ374 and Ϫ381 that are conserved between human and mouse [bib_ref] Involvement of transcriptional factor encoded by the mi locus in the expression..., Tsujimura [/bib_ref] , and fit the core binding motif of MyoD/E2A. Electrophoretic mobility shift assays indicate that SCL-E2A complexes bind these two E boxes (our unpublished results), suggesting that these E boxes may contribute to c-kit promoter activity in response to SCL. We directly tested SCL transactivation properties on a chimeric construct in which the luciferase reporter gene was placed under the control of the human c-kit promoter (kit1146). Since the endogenous levels of SCL are elevated in TF-1 cells, we used TF1-dn for transient transfection. The reporter construct kit1146 was codelivered with varying amounts of the SCL expression vector MSCV-SCL. Results shown in D indicate that SCL induced a dose-dependent increase in c-kit promoter activity, which was twofold at an optimum ratio of activator to reporter. Conversely, cotransfection of dn-SCL and kit1146 in parental TF-1 cells caused a twofold decrease in luciferase activity (data not shown). Together, these results are consistent with the hypothesis that SCL positively regulates c-kit transcription through binding to two adjacent E boxes in the c-kit promoter.
Disrupted SCL Function Specifically Prevents SF-dependent but not GM-CSF or IL-3-dependent Cell Survival. We have previously shown that both SF and GM-CSF suppress apoptosis in TF-1 cells. Parental cells and the different transfectants were, therefore, compared for their survival in response to optimal concentrations of SF, GM-CSF, or IL-3. Apoptotic death was revealed by the presence of a DNA ladder after agarose gel electrophoresis (data not shown). In the absence of growth factor, there was an intense DNA ladder, suggesting that the cells underwent apoptosis. At a concentration of 100 pM of SF, chosen to be near maximal suppression of apoptosis, parental TF-1 cells and the control TF1-neo line behaved similarly, with no detectable DNA degradation. Under identical conditions, apoptosis was evident in both antisense clones, A30 and A31 (data not shown), in which SCL protein levels (data not shown) and DNA binding activity were decreased. Cell viability was restored in clone A31-SCL, consistent with increased SCL DNA binding activity . Reduced c-kit expression in clones expressing an as-SCL: rescue with wild-type SCL. Cells were maintained in culture with 200 pM GM-CSF in the presence of ZnCl 2 for 24 h. (A) Indirect immunofluorescence detection of c-Kit by flow cytometry with the monoclonal antibody against human c-Kit indicates decreased surface expression in A31 and in TF1-dn as compared with parental TF-1 cells or TF1-neo (data shown for A31 and TF-1). Cells labeled with the second antibody alone (goat anti-mouse FITC) served as controls, which were similar for A31 and TF-1 (data shown for TF-1). (B and C) RNA extraction and Northern blotting were performed as described in Materials and Methods. The blots were sequentially hybridized with full-length cDNA for human c-kit, GATA-1, Lyl-1 (data not shown), and GAPDH. The blots were exposed for 24 h to a PhosphorImager screen. (B and C) After normalization with GAPDH and taking TF-1 as 1, c-kit mRNA levels were: 1.05 (TF1-pac), 0.39 (A31), 1.82 (A31-SCL), 1 (TF1-neo), and 0.5 (TF1-dn). (B) Viable cells were determined through double fluorochrome staining as described in Materials and Methods. (D) The c-kit promoter construct kit1146 driving the luciferase reporter gene are codelivered with SCL into exponentially growing TF1-dn cells by electroporation as described in Materials and Methods at the indicated activator to reporter ratio. Data are normalized to that of RSV-GH, cotransfected as an internal control, and corrected for the effect of SCL on RSV-luciferase, used as an external standard. Luciferase activity in the absence of activator was taken as 1. Data are the mean of two independent experiments that were performed in triplicates. lane 4) and SCL protein levels (data not shown) in this clone, whereas control cells expressing the vector alone (A31-pac) behaved like the parental A31 cell line.
Cell viability was therefore quantitated using the double fluorochrome staining technique . Full doseresponse curves for SF and GM-CSF were performed in the two antisense clones, A30 and A31, as well as in control cells. Data are shown for growth factor concentrations that provide 80% survival in TF-1 cells . As observed with agarose electrophoresis, TF-1 and TF1-neo lines remained viable with SF, whereas A30 and A31 underwent apoptosis. As above, ectopic SCL expression in clone A31 (A31-SCL) restored cell viability in SF-containing cultures, indicating that apoptotic death was due to as-SCL. In contrast, cell survival was the same in control and antisense clones in GM-CSF or IL-3 stimulated cultures (data not shown).
Apoptotic cells were also detected through Annexin V staining of membrane phosphatidyl serine, which is exteriorized during apoptotic death, both in stable clones (data not shown) and in transient transfectants 1 wk after G418 selection. Results shown in [fig_ref] Figure 4: as-SCL or dn-SCL prevents cell survival in response to SF [/fig_ref] indicate that parental cells and the different transfectants survive well in GM-CSF-containing cultures, consistent with their staining pattern for GM-CSF receptor (GM-R) observed in [fig_ref] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL [/fig_ref]. In contrast, when the cells were maintained with SF, 80% of the population expressing dn-SCL or as-SCL was apoptotic, whereas both controls (TF-1 and MSCV transfectants) survived readily. Impaired survival in response to SF stimulation in these transfectants is directly correlated with decreased surface c-Kit expression as shown in [fig_ref] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL [/fig_ref]. Together, the results suggest that SCL specifically regulates c-Kit expression and cell survival in response to SF.
Cell Survival in Response to SF Is Rescued by Ectopic SCL or Ectopic c-kit Expression. As discussed above, the A31 subclone in which endogenous c-kit levels were restored after ectopic SCL expression, A31-SCL, survived well in cultures stimulated with SF. We, therefore, addressed the possibility whether c-kit expression by itself was sufficient to rescue SF-dependent cell viability. Murine c-kit was delivered to A31 cells through retrovirus-mediated gene transfer, using the LXSN retroviral vector. Despite a high homology between murine and human c-kit, ligand binding is species specific such that cells expressing human c-Kit require five-to eightfold higher concentrations of murine SF as compared with human SF [fig_ref] Figure 5: Survival response to SF rescued by ectopic murine c-kit expression in as-SCL... [/fig_ref]. After a 2-d selection in low concentrations of murine SF, c-kit-infected A31 and TF-1 cells (A31-kit and TF1-kit, respectively) were expanded with human GM-CSF. Flow cytometry analysis indicated that Ͼ75% of TF1-kit or A31-kit cells express murine Kit [fig_ref] Figure 5: Survival response to SF rescued by ectopic murine c-kit expression in as-SCL... [/fig_ref]. In GM-CSF-containing cultures, all clones behaved similarly (data not shown). As expected, TF1-kit responded better than parental TF-1 cells to murine SF [fig_ref] Figure 5: Survival response to SF rescued by ectopic murine c-kit expression in as-SCL... [/fig_ref]. As with human SF, the response of A31 cells to murine SF was significantly lower than that of TF-1 cells, even at saturating concentrations of murine SF. In contrast, the response of A31-kit to murine SF was similar to that of TF1-kit, suggesting that c-kit overexpression was sufficient to overcome the effect of reduced SCL expression. These observations are compatible with the view that the major survival function of SCL in TF-1 cells is mediated through c-kit expression.
SCL Induces c-Kit Expression in a Negative Cell Line. Ba/F3 is a pro-B cell line that expresses the IL-3 receptor but not c-Kit [bib_ref] The c-kit receptor transduces the stem cell factor-triggered growth signal in murine..., Ohashi [/bib_ref] [bib_ref] Cytoplasmic domains of the interleukin-2 receptor beta and gamma chains mediate the..., Nelson [/bib_ref] or SCL (data not shown). Therefore, we transduced the human SCL cDNA into Ba/F3 cells through retroviral infection in order to address the possibility that SCL can induce c-kit expression in a negative cell line. Control cells were transduced with the MSCV vector alone. 2 wk after selection in G418, polyclonal populations were analyzed for surface expression of murine c-Kit by flow cytometry [fig_ref] Figure 6: SCL induces c-Kit expression in pro-B cell line Ba/F3 [/fig_ref]. As expected, Ba/F3-MSCV cells were c-Kit negative. In contrast, Ba/F3-SCL cells were clearly c-Kit positive.
We next determined whether the cells have acquired responsiveness to murine SF [fig_ref] Figure 6: SCL induces c-Kit expression in pro-B cell line Ba/F3 [/fig_ref]. In the absence of growth factor, Ba/F3-SCL transfectants continuously die over a 3-d period, with the same kinetics as parental Ba/F3 or MSCV transfectants maintained with or without SF. In contrast, when Ba/F3-SCL transfectants were maintained with SF, cell numbers remained at input levels for 2 d and declined slightly on day 3, consistent with the presence of surface c-Kit on these cells. Together, our observations indicate that enforced SCL expression induces c-Kit and confers SF responsiveness in Ba/F3 cells.
# Discussion
Role of SCL as a Survival Gene: Upregulation of c-kit Expression. This study provides evidence for a mechanism whereby SCL secures the survival of primitive hemopoietic cells through upregulation of the tyrosine kinase receptor c-Kit, which is itself essential for the generation of hemopoietic cells both in vivo and in vitro [bib_ref] Absence of yolk sac hematopoiesis from mice with a targeted disruption of..., Robb [/bib_ref] [bib_ref] Absence of blood formation in mice lacking the T-cell leukaemia oncoprotein tal-1/SCL, Shivdasani [/bib_ref] [bib_ref] The T cell leukemia oncoprotein SCL/tal-1 is essential for development of all..., Porcher [/bib_ref] [bib_ref] The SCL gene product is required for the generation of all hematopoietic..., Robb [/bib_ref]. Several lines of evidence indicate a direct link between SCL and cell survival in response to SF. First, the apoptotic phenotype is induced by both as-SCL and dn-SCL. Second, apop-tosis in the antisense clone A31 is rescued by ectopic SCL expression in the sense orientation. Third, apoptosis is specific to the c-Kit pathway triggered by SF, and is not observed with GM-CSF or IL-3, which indicates exquisite biological specificity. Indeed, SF is essential for the normal development of hemopoietic cells, whereas IL-3 and GM-CSF are produced by activated T cells as part of an inflammatory/stress response, suggesting that SCL function is more important in steady-state hemopoiesis. Finally, enforced SCL expression in the pro-B cell line Ba/F3, which is c-Kit Ϫ , clearly induces c-Kit expression in these cells. In all, our study provides a direct link between SCL and the survival response of hemopoietic cells to SF, mostly through upregulation of c-kit transcription.
SCL and Transcription Regulation. The conservation of two everted E boxes at positions Ϫ381 and Ϫ374 between the human and mouse promoter sequences (49) suggests a functional role in transcription regulation, which was confirmed through binding studies (data not shown) and promoter deletion analysis. Our gel shift assays indicate that SCL-containing complexes that are formed on the c-kit probe also include E2A, consistent with the view that SCL does not bind DNA on its own and requires interaction with E12, E47, or HEB, another ubiquitous basic helix-loop-helix [bib_ref] Preferred sequences for DNA recognition by the TAL1 helix-loop-helix proteins, Hsu [/bib_ref] [bib_ref] The LIMonly protein Lmo2 is a bridging molecule assembling an erythroid, DNA-binding..., Wadman [/bib_ref]. Furthermore, SCL is also found in association with the LIM-only protein Rbtn2/Lmo2 [bib_ref] The LIM protein RBTN2 and the basic helix-loop-helix protein TAL1 are present..., Valge-Archer [/bib_ref] [bib_ref] Specific in vivo association between the bHLH and LIM proteins implicated in..., Wadman [/bib_ref] in a multiprotein complex with GATA-1 and a LIM-binding protein, Ldb1 [bib_ref] The LIMonly protein Lmo2 is a bridging molecule assembling an erythroid, DNA-binding..., Wadman [/bib_ref]. An E box-GATA motif was identified through in vitro binding site selection with nuclear extracts instead of isolated proteins [bib_ref] The LIMonly protein Lmo2 is a bridging molecule assembling an erythroid, DNA-binding..., Wadman [/bib_ref] , which corresponds strikingly to the c-kit everted E box sequences at Ϫ381. Sequence analysis of the human c-kit promoter also reveals the presence of several potential GATA sites at Ϫ900. Although these are found upstream instead of downstream of the E box as in the CASTing experiment, it is possible that a looping mechanism favored by protein-protein association allows for optimal spatial arrangements such as those defined by CASTing.
Transactivation assays indicate that SCL represses the E2A-dependent activity of a promoter construct containing the TAL1 motif. It is only in the presence of Id that SCL was shown to relieve the inhibition conferred by Id1 on E2A (7). In addition, despite its high affinity binding to the Ebox-GATA motif, the SCL-E2A heterodimer has no activity on a reporter construct with a minimal promoter linked to two such motives, either in the absence or presence of Lmo2. This multiprotein complex, however, provides a twofold enhancement of luciferase activity driven by GATA-1 on the same construct [bib_ref] The LIMonly protein Lmo2 is a bridging molecule assembling an erythroid, DNA-binding..., Wadman [/bib_ref]. Thus, SCL appears to be a weak transactivator, consistent with our results indicating a twofold contribution to the activity of the proximal c-kit promoter in TF-1 cells. This twofold contribution was confirmed by three distinct approaches, Northern blotting for steady state c-kit mRNA levels, flow cytometry analysis of surface c-Kit protein, and analysis of the c-kit promoter driving the luciferase reporter gene. A recent report using semiquantitative reverse transcriptase PCR did not detect significant differences in c-kit mRNA levels in embryoid bodies grown from scl ϪրϪ embryonic stem and wild-type embryonic stem cells [bib_ref] Hematopoietic-specific genes are not induced during in vitro differentiation of scl-null embryonic..., Elefanty [/bib_ref]. However, it should be noted that flow cytometry analysis and luciferase activity can quantitatively reveal a two-to fourfold difference, whereas reverse transcriptase PCR remains semiquantitative and requires much larger differences for detection. Implication for Leukemogenesis. It was previously shown that loss of SCL function is associated with premature apoptosis upon nutrient deprivation in the lymphoid cell line Jurkat [bib_ref] Loss of TAL-1 protein activity induces premature apoptosis of Jurkat leukemic T..., Leroy-Viard [/bib_ref] and decreased colony formation in the erythroid cell line K562 (53), suggesting a role for SCL in mediating cell survival. As observed with the antiapoptotic gene bcl-2, which infrequently induces lymphomas in transgenic mice after a long latency period of 18 mo (57), elevating SCL expression in thymocytes is not sufficient to cause tumors in transgenic mice [bib_ref] Begley. 1995. scl, a gene frequently activated in human T cell leukaemia,..., Robb [/bib_ref]. Directing SCL expression in the thymus can nonetheless shorten the time of appearance of T cell tumors in LMO-2 (59), LMO-1 (58), or casein kinase II transgenic mice [bib_ref] Tal-1 induces T cell acute lymphoblastic leukemia accelerated by casein kinase IIa, Kelliher [/bib_ref]. Although these observations suggest a role for SCL in leukemogenesis, its target genes remain unknown.
The cells that repopulate the thymus and a subpopulation of triple negative thymocytes are c-Kit ϩ [bib_ref] Phenotypic and functional characterization of c-kit expression during intrathymic T cell development, Godfrey [/bib_ref]. Interestingly, SF is expressed by fetal thymic stromal cells and epithelial cells [bib_ref] Thymic hyperplasia in transgenic mice caused by immortal epithelial cells expressing c-kit..., Moll [/bib_ref] [bib_ref] Developmentally regulated cell surface expression and function of c-kit receptor during lymphocyte..., Palacios [/bib_ref] [bib_ref] Expression of cytokines and their receptors by human thymocytes and thymic stromal..., Wolf [/bib_ref]. More importantly, T cell differentiation in thymic lobe reconstitution with fetal liver precursors is inhibited by anti-Kit [bib_ref] Phenotypic and functional characterization of c-kit expression during intrathymic T cell development, Godfrey [/bib_ref] and is impaired in white spotting mice, indicating the functional importance of c-Kit [bib_ref] Intrathymically expressed c-kit ligand (stem cell factor) is a major factor driving..., Rodewald [/bib_ref]. In parallel, the presence of SCL mRNA in a subset of thymocytes [bib_ref] Expression of tal-1 and GATA-binding proteins during human hematopoiesis, Mouthon [/bib_ref] and the absolute requirement in SCL for the generation of T cells in vivo [bib_ref] The T cell leukemia oncoprotein SCL/tal-1 is essential for development of all..., Porcher [/bib_ref] [bib_ref] The SCL gene product is required for the generation of all hematopoietic..., Robb [/bib_ref] are indicative of the importance of SCL for thymocyte development. Our observations suggest that SCL may also specify c-Kit expression in primitive thymocytes to sustain cell survival in response to SF. Consistent with this, elevating SCL levels in thymocytes in double CD2-SCL and CD2-Lmo2 transgenics results in a twofold expansion of the double negative CD4 Ϫ /CD8 Ϫ thymocyte population relative to CD2-Lmo2 transgenics [bib_ref] Tal-1 induces T cell acute lymphoblastic leukemia accelerated by casein kinase IIa, Kelliher [/bib_ref] , which, in light of our results, may be due to an increase in c-Kit ϩ subpopulations. We propose that constitutive SCL expression in the T cell lineage caused by chromosomal rearrangements results in constitutive c-kit expression in primitive thymocytes and prolongs their survival, which may represent an initiating event in T cell ALL. Secondary events are probably involved in the full development of T cell leukemia [bib_ref] Begley. 1995. scl, a gene frequently activated in human T cell leukaemia,..., Robb [/bib_ref] [bib_ref] Reconstitution of mice with bone marrow cells expressing the SCL gene is..., Elwood [/bib_ref] , a mechanism commonly observed in tumor biology [bib_ref] Molecular themes in oncogenesis, Bishop [/bib_ref].
Tissue-specific Transcription Factors and Cell Survival. The regulation of cell survival is central to normal developmental processes and stress response. It is widely accepted that survival cues are crucial for the differentiation process [bib_ref] Differentiation and cell death: lessons from the immune system, Linette [/bib_ref] , and emerging evidence suggests a novel role for tissue-specific transcription factors, otherwise important for driving a defined pattern of gene expression, in regulating cell survival. Thus, the zinc finger transcription factor GATA-1 is required for terminal maturation in the erythroid lineage [bib_ref] Development of hematopoietic cells lacking transcription factor GATA-1, Pevny [/bib_ref]. GATA-1 gene ablation in embryonic stem cells blocks the differentiation at the proerythroblast stage [bib_ref] Development of hematopoietic cells lacking transcription factor GATA-1, Pevny [/bib_ref] [bib_ref] Transcription factor GATA-1 permits survival and maturation of erythroid precursors by preventing..., Weiss [/bib_ref] and causes apoptotic death [bib_ref] Transcription factor GATA-1 permits survival and maturation of erythroid precursors by preventing..., Weiss [/bib_ref]. Similarly, GATA-4 is required for cardiomyocyte differentiation from embryonal carcinoma cells [bib_ref] Enhanced cardiogenesis in embryonic stem cells overexpressing the GATA-4 transcriptional factor, Grépin [/bib_ref] and cell survival in the cardiomyocyte pathway but not the neuronal lineage. Hence, apoptosis may be linked to insufficient survival signals, but also to inefficient or abortive cell differentiation. Similarly, SCL is required early in hemopoiesis, possibly in specifying ventral mesoderm to the hemopoietic fate [bib_ref] The T cell leukemia oncoprotein SCL/tal-1 is essential for development of all..., Porcher [/bib_ref] , and is shown here to be crucial for SF-mediated cell survival. Therefore, it is proposed that lineage-specific transcription factors orchestrate both lineage-specific gene expression and the survival of the specified cell types.
[fig] Figure 1: Flow cytometry analysis of TF-1 transfectants expressing as-SCL or dn-SCL. (top left) Decreased SCL protein in TF-1 cells expressing as-SCL. TF-1 cells were infected with a retrovirus harboring the human SCL cDNA cloned in the antisense orientation and maintained as a polyclonal population in GM-CSF containing medium in the presence of G418 (1 mg/ml). 1 wk after infection, as-SCL transfectants were fixed and permeabilized with Triton X-100 for nuclear staining with the monoclonal anti-TAL1 at a final dilution of 1:10, followed by labeling with a goat anti-mouse antibody coupled to FITC (Sigma). Cells were washed extensively for 10 min after each labeling step and analyzed by flow cytometry. Parental TF-1 cells were stained in parallel as positive controls. Both cell types were also stained with the second antibody alone (2 nd ) as negative controls. In contrast to the shift in SCL nuclear staining between as-SCL transfectants and parental TF-1 cells, those of MSCV (vector alone) transfectants and TF-1 cells were comparable (data not shown).(top right) Decreased c-Kit expression in TF-1 cells expressing as-SCL or dn-SCL. The same cells as in the top left panel were analyzed for surface labeling with a biotinylated monoclonal antibody against human c-Kit and streptavidin-coupled PE (S-PE). Cells labeled with S-PE alone serve as negative controls. (middle and bottom) Surface expression of CD45 and CD116 (GM-CSF receptor [GM-R] ␣ chain) is unaltered in TF-1 transfectants.Cells were labeled with a mouse anti-human CD45 and anti-human CD116 (GM-R), followed by FITC-labeled goat anti-mouse IgG. Labeling with the second antibody alone (2 nd ) is shown. [/fig]
[fig] Figure 4: as-SCL or dn-SCL prevents cell survival in response to SF. Apoptotic cell death is analyzed quantitatively by staining with Annexin V 6 d after retroviral infection. The indicated transfectants were incubated for 44 h with a concentration of SF chosen to provide 70-80% survival in parental TF-1 cells (25% Annexin ϩ cells), or kept in GM-CSF containing medium. Cells were stained with propidium iodide and Annexin V-FITC as described in Materials and Methods. The proportions of Annexin ϩ cells are shown in each histogram. Results shown are typical of two independent infections. [/fig]
[fig] Figure 5: Survival response to SF rescued by ectopic murine c-kit expression in as-SCL transfectants. TF-1 cells and the as-SCL-expressing cells (A31 cells) were infected with MSCV harboring the murine c-kit gene, or with the pac resistance gene (control) as described in Materials and Methods. A31-Kit and TF1-Kit (A) as well as A31 and TF-1 (B) were labeled with ACK2, a monoclonal antibody specific for murine Kit, and a goat anti-rat Ig coupled to FITC (solid lines). Cells labeled with the second antibody alone served as negative controls (broken lines). For apoptosis assay, cells were maintained with the indicated concentrations of murine SF for 2 d. Viable cells were determined by trypan blue exclusion. (C) The half efficient concentration of murine SF for the different clones were: TF-1, 17.8 Ϯ 2.1; A31, 13 Ϯ 4; TF1-kit, 3.5 Ϯ 0.5; A31-kit, 4.3 Ϯ 0.6. [/fig]
[fig] Figure 6: SCL induces c-Kit expression in pro-B cell line Ba/F3. (top) After retroviral-mediated gene transfer, polyclonal populations with enforced SCL expression (Ba/F3-SCL) or control cells harboring the vector alone (Ba/F3-MSCV) were labeled with the monoclonal rat anti-murine c-Kit as in Fig. 5.(bottom) Viable cell counts from cultures stimulated with optimal concentrations of murine SF were determined through trypan blue exclusion. Cell counts for growth factor-deprived Ba/F3 and Ba/F3-MSCV (data not shown) were comparable to those observed with murine SF. [/fig]
|
Nature versus urban hiking for Veterans with post-traumatic stress disorder: a pilot randomised trial conducted in the Pacific Northwest USA
# Introduction
Post-traumatic stress disorder (PTSD) is a common, chronic mental health condition that affects up to 30% of military Veterans and is frequently comorbid with anxiety, depression and substance misuse. 1-3 PTSD increases the risk of suicide as well as obesity, physical inactivity and cardiovascular and metabolic disorders. [bib_ref] Posttraumatic stress disorder in veterans and military personnel: epidemiology, screening, and case..., Gates [/bib_ref] [bib_ref] The prevalence of posttraumatic stress disorder in operation enduring Freedom/ Operation Iraqi..., Fulton [/bib_ref] [bib_ref] The prevalence of posttraumatic stress disorder in the Vietnam generation: a multimethod,..., Schlenger [/bib_ref] [bib_ref] Physical activity, and eating behaviors, Hall [/bib_ref] [bib_ref] Health benefits of physical activity: a systematic review of current systematic reviews, Warburton [/bib_ref] [bib_ref] Health of US veterans of 1991 Gulf war: a follow-up survey in..., Kang [/bib_ref] [bib_ref] Trauma is a public health issue, Magruder [/bib_ref] [bib_ref] Review of somatic symptoms in post-traumatic stress disorder, Gupta [/bib_ref] [bib_ref] Comorbidity of posttraumatic stress disorder, anxiety and depression: a 20-year longitudinal study..., Ginzburg [/bib_ref] [bib_ref] Burden of medical illness in women with depression and posttraumatic stress disorder, Frayne [/bib_ref] [bib_ref] Posttraumatic stress disorder, cardiovascular, and metabolic disease: a review of the evidence, Dedert [/bib_ref] Clinical practice guidelines recommend treatment with several evidencebased psychotherapies and medications,but many Veterans who need PTSD treatment do not receive it. [bib_ref] Combat duty in Iraq and Afghanistan, mental health problems, and barriers to..., Hoge [/bib_ref] Barriers to obtaining treatment include concerns about medication side effects, desire for self-management approaches, stigma about receiving mental healthcare and a lack of confidence in mental health treatment in general. [bib_ref] Combat duty in Iraq and Afghanistan, mental health problems, and barriers to..., Hoge [/bib_ref] [bib_ref] Stigma as a barrier to seeking health care among military personnel with..., Sharp [/bib_ref] [bib_ref] Failure and delay in initial treatment contact after first onset of mental..., Wang [/bib_ref] [bib_ref] Treatment-seeking barriers for veterans of the Iraq and Afghanistan conflicts who screen..., Stecker [/bib_ref] These and other factors adversely impact engagement, contributing to low initiation of and high drop-out rates from treatment. [bib_ref] A systematic review of dropout from psychotherapy for posttraumatic stress disorder among..., Goetter [/bib_ref] Identifying a wider range of approaches that are acceptable and effective is key to reducing the burden PTSD places on individuals and their communities.
Strengths and limitations of this study ► By using group-based urban hiking as a comparison group to control for the effects of physical activity and social cohesion (present in both interventions), this study was designed to isolate benefits specifically due to the environment (which differed between the interventions). ► We used population-based recruitment methods to enroll a representative sample of Veterans with post-traumatic stress disorder. ► Because of its small size and focus on feasibility, the study was not large enough to determine the effectiveness of nature hiking on outcomes.
## Open access
There is growing interest in nature contact as a potential therapy for Veterans with PTSD and robust evidence that nature contact improves physical and psychological health among healthy individuals and those with mental health disorders. [bib_ref] Nature contact and human health: a research agenda, Frumkin [/bib_ref] Nature contact has been shown to increase subjective well-being; decrease stress, anxiety, depression and negative affect; and promote adaptive shifts in emotion regulation. [bib_ref] Nature contact and human health: a research agenda, Frumkin [/bib_ref] [bib_ref] The affective benefits of nature exposure, Bratman [/bib_ref] Benefits of nature contact are generally posited to occur based on two theories: attention restoration theory (ART) and stress recovery theory (SRT). [bib_ref] The restorative benefits of nature: toward an integrative framework, Kaplan [/bib_ref] [bib_ref] Stress recovery during exposure to natural and urban environments, Ulrich [/bib_ref] ART theorises that nature contact improves cognitive function through a replenishment of 'directed attention', a capacity that is overly taxed in urban environments due to the need to block out distracting stimuli (eg, noise) to focus on a specific task or cognitive process. This depleted attention capacity can be restored in natural environments through the engagement of 'soft fascination', with implications for both cognitive and emotional well-being. SRT is based on psychoevolutionary principles, and posits that many types of nature exposure enhance psychological wellbeing through a precognitive, positive affective response and activation of the parasympathetic nervous system in ways that reduce stress and sympathetic nervous system arousal. [bib_ref] The affective benefits of nature exposure, Bratman [/bib_ref] Like nature contact, physical activity (PA) is considered to be a promising approach to improve outcomes for individuals with PTSD. PA reduces anxiety and depression and improves stress regulation, sleep and cognitive functioning in the general population, and in people with PTSD, though only 8 studies have involved randomised controlled trial (RCT) designs, and 5 of the RCTs were pilot studies or included fewer than 30 people. [bib_ref] Ptsd symptom reduction with mindfulness-based stretching and deep breathing exercise: randomized controlled..., Kim [/bib_ref] [bib_ref] A pilot study of a randomized controlled trial of yoga as an..., Mitchell [/bib_ref] [bib_ref] Exercise augmentation of exposure therapy for PTSD: rationale and pilot efficacy data, Powers [/bib_ref] Furthermore, we are aware of only one RCT focused on Veterans. [bib_ref] Warrior wellness: a randomized controlled pilot trial of the effects of exercise..., Hall [/bib_ref] Group-based PA interventions may be particularly well-suited for military Veterans, due to (1) proportionally higher rates of PTSD among Veterans, [bib_ref] Post-Traumatic stress disorder by gender and veteran status, Lehavot [/bib_ref] (2) consistency of PA interventions with values cultivated during military service and (3) benefits of social interaction with veteran peers. [bib_ref] Veterans' perspectives on benefits and drawbacks of peer support for posttraumatic stress..., Hundt [/bib_ref] To our knowledge, no PA interventions in those with PTSD investigated the PA environment as a component of treatment. This is an important omission, because the environment in which PA takes place may play an important role in its benefits. [bib_ref] The mental and physical health outcomes of green exercise, Pretty [/bib_ref] Green exercise, defined as activity that takes place in natural environments, is a burgeoning area of research. [bib_ref] Does participating in physical activity in outdoor natural environments have a greater..., Thompson Coon [/bib_ref] [bib_ref] The effects of green exercise on physical and mental wellbeing: a systematic..., Lahart [/bib_ref] [bib_ref] Nature-assisted therapy: systematic review of controlled and observational studies, Annerstedt [/bib_ref] [bib_ref] What is the best dose of nature and green exercise for improving..., Barton [/bib_ref] [bib_ref] Nature-Based physical Recreation leads to psychological well-being: evidence from five studies, Wolsko [/bib_ref] [bib_ref] Walking for well-being: are group walks in certain types of natural environments..., Marselle [/bib_ref] A number of studies have documented benefits from green exercise in Veteran populations and among individuals with PTSD. [bib_ref] Nature-assisted therapy: systematic review of controlled and observational studies, Annerstedt [/bib_ref] [bib_ref] What is the best dose of nature and green exercise for improving..., Barton [/bib_ref] [bib_ref] Nature-Based physical Recreation leads to psychological well-being: evidence from five studies, Wolsko [/bib_ref] [bib_ref] Walking for well-being: are group walks in certain types of natural environments..., Marselle [/bib_ref] [bib_ref] How individuals with self-reported anxiety and depression experienced a combination of individual..., Kyriakopoulos [/bib_ref] [bib_ref] Addressing posttraumatic stress among Iraq and Afghanistan veterans and significant others: an..., Bennett [/bib_ref] [bib_ref] Outcomes of a therapeutic Fly-Fishing program for veterans with combat-related disabilities: a..., Bennett [/bib_ref] [bib_ref] The promise of river running as a therapeutic medium for veterans coping..., Dustin [/bib_ref] [bib_ref] Enhancing the well-being of Veterans using extended group-based nature recreation experiences, Duvall [/bib_ref] [bib_ref] Nature adventure rehabilitation for combat-related posttraumatic chronic stress disorder: a randomized control..., Gelkopf [/bib_ref] [bib_ref] Effects of outward bound experience as an adjunct to inpatient PTSD treatment..., Hyer [/bib_ref] The specific interventions studied (from care farming to river rafting), dose/duration and inclusion of additional, explicit therapeutic components vary substantially among studies. A 2019 systematic review that examined evidence for the proposed additive effects of exercise in the presence of nature observed some benefits (eg, lower perceived exertion and enjoyment). Nevertheless, the authors concluded that there was a high risk of bias across trials and an overall low quality of evidence. [bib_ref] The effects of green exercise on physical and mental wellbeing: a systematic..., Lahart [/bib_ref] Thus, uncertainty about the duration and impacts of green exercise remains due to methodological issues and because most interventional studies tested only a single bout of exercise. Furthermore, in the studies including Veterans, important limitations include low retention for follow-up, absence of control groups and insufficient statistical power. [bib_ref] The promise of river running as a therapeutic medium for veterans coping..., Dustin [/bib_ref] [bib_ref] Enhancing the well-being of Veterans using extended group-based nature recreation experiences, Duvall [/bib_ref] [bib_ref] Nature adventure rehabilitation for combat-related posttraumatic chronic stress disorder: a randomized control..., Gelkopf [/bib_ref] [bib_ref] Effects of outward bound experience as an adjunct to inpatient PTSD treatment..., Hyer [/bib_ref] [bib_ref] Awe in nature heals: evidence from military veterans, at-risk youth, and college..., Anderson [/bib_ref] [bib_ref] Whatever happened to the soldiers? Nature-assisted therapies for veterans diagnosed with post-traumatic..., Poulsen [/bib_ref] [bib_ref] Effectiveness of Care Farming on Veterans' Life Satisfaction, Optimism, and Perceived Loneliness, Greenleaf [/bib_ref] In addition to nature contact and PA, a third important constituent of many green exercise interventions is a group component. Some recent research suggests that increased social cohesion and connectedness may mediate benefits of green exercise, 59 but findings are inconsistent. [bib_ref] Psychological benefits of walking: moderation by company and outdoor environment, Johansson [/bib_ref] Social support forged through group activity could be particularly relevant for Veterans, as camaraderie and solidarity are critical components of military culture, and ones that are frequently lost in the return to civilian life.Social support is associated with reduced PTSD symptoms and improved treatment response [bib_ref] An examination of social support and PTSD treatment response during prolonged exposure, Price [/bib_ref] and may directly impact stress response by increasing personal resources, [bib_ref] The impact of social support, unit cohesion, and trait resilience on PTSD..., Zang [/bib_ref] and/ or may indirectly impact PTSD symptom severity and response to treatment through buffering the potentially harmful impacts of stressful events. [bib_ref] Stress, social support, and the buffering hypothesis, Cohen [/bib_ref] Adequately powered studies involving ongoing green exercise that are designed to distinguish between benefits due to PA and those due to the physical (eg, nature) and social (eg, group cohesion) environment are needed. Thus, our goal was to design and conduct a pilot study to test the feasibility and acceptability of a green exercise intervention for PTSD symptoms in military Veterans, regardless of PTSD aetiology. The intervention (nature hiking) and the active control (urban hiking) were group based and involved similar amounts of PA, to ensure control of the potential benefits of the group-based social support and of PA. [fig_ref] Figure 1: Conceptual model. [/fig_ref] depicts our conceptual model. This paper describes the results of the initial pilot study designed to emulate important elements of the future envisioned full-scale randomised trial.
# Methods
## Identification and recruitment of participants
We used active and passive methods to identify and recruit Veterans to participate. While receiving care at a Department of Veterans Affairs (VA) healthcare facility was not an inclusion criterion, we used VA electronic medical records as a key source to identify potentially eligible Veterans. We identified VA enrollees (identified using electronic medical records) with at least one encounter with a diagnosis of PTSD in the prior 2 years; a zip code in one of three Seattle-Tacoma area counties (King, Snohomish and Pierce); no hospitalisations in the prior 3 months; and no diagnoses of schizophrenia, bipolar disorder or other psychotic disorder. We randomly selected 1001 individuals who met these criteria from a total of approximately 7000 and mailed them a letter informing them about the study and inviting them to participate. We followed the mailing with up to three phone calls until the recruitment period ended. We also placed study recruitment flyers in clinics in the VA Puget
## Open access
Sound and mailed flyers to four local organisations and clinics serving Veterans. Individuals who expressed an interest were mailed an invitation letter.
We initially screened all Veterans who expressed an interest in participating for eligibility over the phone; inclusion criteria assessed included a history of PTSD, ability/willingness to comply with study procedures (eg, complete questionnaires, wear and sync an activity monitor, drive to hikes and walk at least 2 hours at an easy/ moderate pace over uneven terrain). Exclusion criteria assessed included a diagnosis of schizophrenia, bipolar disorder or other psychotic disorder; hospital admission in the prior 3 months and inability to perform unsupervised PA based on the PA Readiness Questionnaire. [bib_ref] Revision of the physical activity readiness questionnaire (PAR-Q), Thomas [/bib_ref] We invited those who passed all criteria except for the PA Readiness Questionnaire to obtain approval to participate from their primary care provider. Though some of this information was available in VA medical records, because we also included Veterans who did not have VA medical records, we employed methods that allowed us to evaluate eligibility without medical record access. Those who passed initial screening were mailed consent forms and given a link to complete a more extensive screening questionnaire online. Via the online screening questionnaire, PTSD symptoms, drug use, alcohol misuse and suicidality were assessed. PTSD was determined based on a PTSD-checklist-5 66 score >33. We excluded those with drug abuse in past year (Drug Abuse Screening Test-10 67 score >=3); alcohol disorder/dependence (current/ past year; Alcohol Use Disorders Identification Test-10 68 score >16) and moderate/severe suicidality (past month; Mini International Neuropsychiatric Interview (MINI) Suicidality module 69 score >5). Those who were eligible and returned signed consents were considered enrolled in the study.
## Study design
We conducted a two-arm randomised controlled pilot trial. The two interventions were group nature and group urban hiking. The random 1:1 allocation sequence was generated using simple randomisation in random blocks of 2, 4 and 6. Randomisation assignments were placed in opaque sequentially numbered envelopes. Once an individual was determined to be eligible, the study coordinator selected the next envelope to determine the individual's group assignment. We did not blind participants, the study coordinator or the study statistician to group assignment. [fig_ref] Figure 2: Depiction of study design and assessments [/fig_ref] presents an overview of the study, including timing of assessments.
## Description of hike locations and amenities
The criteria used to select the hike locations (which applied to both nature and urban hikes) included duration, elevation change, availability of facilities (eg, toilets and water), distance from participants' homes and access to parking. Nature hikes were held in State Parks, National Wildlife Refuges and Natural Resources Conservation areas in the US Pacific Northwest. The nature hikes were in forest habitat, including old growth forest, saltwater shoreline, waterfalls and alpine lakes. Elevations ranged from sea level to 2200 feet above sea level. Urban hikes were held in primarily built environments, avoiding urban parks or primarily residential neighbourhoods with substantial greenery or water features. Urban hikes comprised areas that included sports stadiums, urban art and retail establishments and were mainly on sidewalks rather than separated bike/pedestrian paths/rail-trails. It was not feasible to match nature and urban hikes on elevation change; instead, we aimed to have similar hike durations to match total exertion. Generally, nature hikes involved somewhat shorter distances but included more elevation gain/loss.
## Hiking intervention
A total of six hikes over 12 weeks (one every other week) were offered between August and October 2019. We chose to offer six hikes (vs more or fewer) because this number was thought to be feasible and would be sufficient to assess feasibility and acceptability. The standard structure for hikes was: (1) 'ice breakers' (short, guided conversations), (2) overview of the planned hike, including distance, unique features and planned stops,
## Open access
(3) hike and (4) posthike debrief and administration of questionnaires. There were no additional group/therapeutic activities.
Hike durations increased gradually to account for anticipated increases in participants' physical fitness. Initial hikes were 60-90 min (2-3 miles), and later hikes were 2-3 hours (5-6 miles). To ensure safety and inclusion, one hike leader was in sight and hearing of the first participant and a second leader accompanied the last participant. The group stopped at least every 30 min to keep everyone together and offer opportunities to rest, regroup and inquire about and address any issues or concerns arising since the last check-in.
The same hike leaders, who were non-clinicians, led both nature and urban hikes to control for hike leader effects. On every hike, at least one of the leaders was a woman. Leaders were experienced outdoor educators who were employed by a Seattle-based outdoor organisation that provides outdoor recreation activities for people with disabilities. While the leaders were not Veterans, the organisation received grants from the VA as part of the Adaptive Sports Programme 70 and had previously led programmes for Veterans. Leaders were trained to handle physical and mental health emergencies by the PIs (AJL and GNB) and a co-investigator who is a licensed clinical psychologist (KL). AJL and GNB supervised the hike leaders during the study.
To reduce barriers to attendance, a US$35 incentive was provided to defray parking costs. We provided a rain jacket and technical shirt as well as well as an activity monitor (Garmin vivosmart V.4) at the participant's first hike.
## Outcomes
The primary outcomes of interest were feasibility and acceptability. Feasibility was assessed based on recruitment statistics (the proportion of individuals who were contacted, eligible and enrolled, as well as reasons for ineligibility), retention (questionnaire completion), hike attendance and safety (eg, adverse events). We aimed to recruit 36-45 participants (12-15 people allocated to each of the three groups-nature hiking, urban hiking and a no-hiking control group) and complete enrolment by July 2019 (approximately 3 months after recruitment began) due to concerns about weather for hikes later in the fall. Because of lower-than-anticipated recruitment numbers, in late June, we decided to eliminate the no-hiking control group. At this time, only one person was randomised to the no-hiking control group and informed of their group assignment; that person was re-randomised after this decision was made. The target for retention and attendance was 70%, a commonly cited standard for trials. To assess acceptability, in the 6-week and 12 week questionnaires, we included questions created for the study about the difficulty of the hikes' distance, pace and the terrain (rated on a 5-point scale from extremely difficult to effortless), and satisfaction with the locations of hikes (rated on a 5-point scale from extremely unsatisfied to very satisfied). Lastly, prehike and trailhead information and communication were assessed on a 5-point scale (eg, from very poor to excellent). We also included open-ended questions for participants to report what they thought went well and what could have been better. Additionally, after the final hike, the lead author Open access (AJL) conducted semistructured telephone interviews, with a goal to interview 10-15 participants. To include a range of perspectives, we purposively sampled participants from both arms, aimed to include men and women, and participants who varied in terms of hike attendance. Questions inquired about participants' impressions of the hikes, including difficulty, location, length of time, distance from home, hike leaders and reasons hikes were missed (if applicable); study communications; enrolment process; assessments and other thoughts/impressions. Determination of efficacy was not a goal of this pilot RCT. The primary outcome of the future planned study is PTSD symptoms, assessed by the PTSD-Checklist for Diagnostic and Statistical Manual 5 (PCL-5), a 20-item instrument that assesses PTSD symptoms in the past month (range 0-80, with higher scores indicating greater symptom severity). Other outcome measures of interest for the future planned study, which are detailed in online supplemental table 1, include quality of life, 73 positive and negative affect, sleep, 76 rumination 77 and cognitive reappraisal. [bib_ref] The emotion regulation questionnaire: psychometric properties in general community samples, Preece [/bib_ref] Baseline and follow-up assessments We conducted assessments online using commercial software (QuestionPro) at baseline (before hikes began), and then weekly for 12 weeks, starting with the week of the first hike and ending the week after the sixth hike, and finally at week 24; questionnaires completed immediately after the hikes were completed on paper. See [fig_ref] Figure 2: Depiction of study design and assessments [/fig_ref] for an overview and online supplemental table 1 for measures at each time point. Questionnaires at weeks 6, 12 and 24 took approximately 30 min to complete. Questionnaires administered at weeks 1-5 and 7-11 included fewer measures and/or shortened versions and took 5-10 min to complete. Participants received gift cards worth US$10 for completing questionnaires in weeks 1-5 and 7-11, US$20 for the 6-week questionnaire, and US$50 for the 12-week and 24-week questionnaires. In addition to questionnaires, to obtain objective information about PA (a potential mechanism of benefit, which we would want to measure precisely in a future study), we asked participants to wear a wrist worn-activity monitor (Garmin vivosmart V.4) every day, for at least 10 hours per day, for the first 12 weeks of the study. No incentives were provided for wearing or synching the monitor.
## Data analysis quantitative analysis
The primary purpose of the extensive data collection was to evaluate feasibility of data collection rather than to estimate effect sizes because estimating effect sizes from small pilot studies is inherently imprecise. [bib_ref] The role and interpretation of pilot studies in clinical research, Leon [/bib_ref] Thus, instead of conducting hypothesis tests for effectiveness outcomes for which we were underpowered, we present descriptive statistics (eg, medians and IQRs) for the primary outcome (PCL-5) only. For acceptability measures related to communication, we categorised responses as positive if respondents chose one of the two most favourable response options (eg, satisfied/very satisfied; good/ excellent) and not positive if they chose one of the other response options (extremely unsatisfied/unsatisfied/ neither satisfied or unsatisfied; inadequate/very poor/ adequate). We then calculated the proportion of urban and nature participants with favourable responses for each question. In addition to proportions, we also calculated the mean scores for hike locations, distance, pace, prehike information, prehike communication and trailhead communication by group.
# Qualitative analysis
All interviews were recorded, and the interviewer took notes during interviews. For both the comments shared via open-ended questions on the questionnaire and comments shared orally during the interviews, we conducted inductive content analysis, which involves open coding of data, organising codes and data into categories and comparing data across participants to identify patterns and themes in the data. [bib_ref] The qualitative content analysis process, Elo [/bib_ref] Patient and public involvement Patients were involved in the design and conduct of this study. The study question and design were informed by a Veteran with PTSD who served as a co-investigator. The design and messaging for this pilot study were also informed by a focus group of patients/Veterans who participated in a prior unpublished feasibility study.
# Results
## Feasibility
Recruitment statistics Recruitment took place between April and August 2019 (16 weeks total). Of the 1001 patients mailed an invitation letter, we were unable to assess interest or eligibility in 586 (because they did not respond to the mailings and/ or answer the phone when called; see [fig_ref] Figure 3: CONSORT diagram [/fig_ref] for Consolidated Standards of Reporting Trials diagram). Of the 436 with whom we made contact (including 21 who contacted the study), 159 were not interested, 102 had health conditions that limited their walking/hiking, 37 had time conflicts (eg, work or church on Sundays or travel that would prevent participation), and 41 had other reasons that they were unable to participate (eg, moving out of the area, did not have PTSD). Of the 97 (81 from letters+16 from passive recruitment) interested who passed initial screening, 48 completed the online screening questionnaire. Twenty individuals were not eligible, 2 decided that they no longer wanted to participate and 26 were eligible and randomised. Of the 20 who were not eligible, 13 were ineligible because of a moderate/high risk of suicide or skipping the question on suicidality, and 6 did not meet the threshold for PTSD. Compared with those contacted and not randomised, a greater proportion of those randomised were women (27% randomised vs 15% of those contacted), white (73% vs 63%) and Hispanic (8% vs 6%). Additionally, those who were randomised Open access were younger (mean age=47, range 25-65) than those not randomised (mean age=52, range: 21-95). [fig_ref] Table 1: Baseline characteristics of Veterans in the urban and nature hiking groups [/fig_ref] presents characteristics of Veterans who were randomised and includes self-reported race/ethnicity, which differed from race/ethnicity in the electronic medical record (reported above). Specifically, 42% of those randomised self-reported being white, whereas the electronic medical record data indicated that 73% were white. Thirty per cent had a college degree or more. Less than half worked full time and 46% had 100% VA serviceconnected disability, indicating severe impairment in ability to work. Nearly two-thirds of participants had served in combat and 68% had depressive symptoms based on the Patient Health Questionnaire (PHQ)-8. Based on self-report, nearly 70% met or exceeded PA guidelines of at least 75 min per week of vigorous-intensity activity or 150 min per week of moderate-intensity activity or an equivalent combination of the two. At baseline (prior to study initiation), 81% of participants reported hiking at least one time and 27% completed seven or more hikes in the prior year.
## Retention (questionnaire completion)
Mean completion of all questionnaires was 91% in the nature hiking group and 68% in the urban hiking group.
Completion rates were similar for the shorter weekly questionnaires and the longer questionnaires.
## Hike attendance
Over the course of the intervention, participants in the nature and urban groups attended an average of 56% and 58%, respectively, of scheduled hikes. In the nature group, one person attended no hikes, four (31%) attended 1-2 hikes, one attended 3 hikes and seven (54%) attended 4-6 hikes. In the urban group, one person attended no hikes, four (31%) attended 1-2 hikes, no one attended only 3 hikes and eight (62%) attended 4-6 hikes. Attendance was lower among women in the nature group (n=5, mean: 43%) than among women in the urban group (n=2, 67%), whereas among men, hike attendance was similar in the two groups (65% vs 56%). Common reasons for missing hikes included work, childcare and prior plans.
Safety/adverse events One participant in the urban hiking group reported increased anxiety/PTSD symptoms in connection with hiking in the urban environment and withdrew from the study. Open access above, they liked getting to know other Veterans and having a 'mission'. Veterans wanted to find more ways of connecting with one another socially during hikes as well as outside of hikes. Hike logistics (eg, distance from home) were noted as potential barriers to attendance.
## Open access
## Efficacy measures
Median PCL-5 scores decreased from baseline to week 12 and remained at the 12-week level at week 24 for those in the nature hiking group (baseline=41, 12-weeks=32, 24 weeks=31). Among those in the urban hiking group, PCL-5 scores decreased from baseline to 12 weeks but increased nearly back to baseline levels at 24 weeks (base-line=48, 12-weeks=43, 24 weeks=47) (online supplemental [fig_ref] Figure 1: Conceptual model. [/fig_ref]. We did not test the statistical significance of the changes because this pilot study was not designed to answer this question. [bib_ref] Caution regarding the use of pilot studies to guide power calculations for..., Kraemer [/bib_ref]
# Discussion
This study was an important step in establishing feasibility and acceptability and identifying changes to consider in the development of a rigorous, fully powered study to evaluate the impact of nature hiking on PTSD symptoms. The results of this pilot study generally supported feasibility and acceptability. Participants reported high acceptability, enjoyment and value, based on quantitative and qualitive data. In both arms, more than half of participants completed most hikes. Qualitative feedback about improving the social component supports the hypothesis that social connection is an important aspect of hikes, indicating a need to further develop the social component and continue to study group interventions like this one. Additionally, the decrease in median scores on the PCL-5 among those in the nature group after the 12-week hiking intervention, and 12 weeks later (week 24) is promising. This preliminary finding should be investigated more thoroughly in future, larger-scale versions of our study. The indication that improvements may persist after the conclusion of the intervention is especially compelling given the current unknowns regarding the duration of effects of nature interventions. Nevertheless, several issues need to be considered related to feasibility and acceptability for the next iteration of this research.
## Feasibility of recruitment
We fell short of our goal of recruiting at least 36 people over 4 months. Failing to meet recruitment goals in the planned timeframe is a common problem in RCTs. [bib_ref] Recruitment and retention of participants in randomised controlled trials: a review of..., Walters [/bib_ref] Barriers to recruitment included unexpected delays, insufficient resources and an inefficient recruitment process. Regarding delays, we had to wait weeks for institutional review board approval for each proposed modification to recruitment materials/protocol. Regarding Key themes and findings from qualitative data
## Themes findings
A positive experience ► Both nature ('All expectations were exceeded') and urban study participants ('LOVE THE GROUP') provided positive feedback.
Perceived benefits ► Participants reported on how the hikes helped them to be more active, lose weight, reduce stress and feel more connected to others.
Hike logistics ► Participants suggested that prior to hikes, we ensure parking access, availability of toilets and locate the hikes closer to participants' homes.
► Others suggested that we consider organising carpools and/or covering gas/mileage costs.
Difficulty of hikes ► Most found the difficulty just right. ► Some felt that the hikes were too short/easy.
Location of hikes ► Nature group: One participant wished that there was more of a 'reward' ('like a waterfall', 'when you have a view, it seems more profound'), because some were just 'walks through the woods'.
► Urban group: One person noted that some neighbourhoods were 'sketchy' and they were 'constantly walking around garbage' for one hike. Others noted that they really enjoyed exploring different neighbourhoods, areas around sports stadium and learning about the history of areas.
Group composition ► A few participants suggested that we enrol more women or organise women-only groups and/or groups for survivors of sexual assault.
Incentives for completing questionnaires ► Participants suggested that we offer the option to receive a single gift card that accumulated value instead of separate ones each time a questionnaire was completed.
## Assessments
► Several participants had trouble with the online software (eg, getting 'kicked out' of the survey mid-way through).
► Some participants reported that they would have liked text prompts instead of email, since they did not regularly check their email.
► Some participants found some questions to be difficult to answer (eg, the Perceived Cohesion Scale) or they were confused by differences in the time frame for different instruments (eg, on the weekly questionnaires, some questions asked participants how they felt 'right now' while others asked about the prior week).
## Activity monitors
► Several participants noted having problems programming and syncing the activity monitor.
## Fostering interaction/connections between participants in a group
► Participants suggested facilitating more structured ways to get to know other members of the group, including a social gathering prior to the initial hike, reintroductions before each hike, gathering for lunch or other meal after hikes and organising a social media group.
Open access resources, we only had 20 hours per week of paid staff time for recruitment. The addition of two volunteers in the final 2 months helped to accelerate enrolment, but more resources earlier in recruitment would have been necessary to meet our goal. One contributor to inefficiency in recruitment was the broad, population-based approach we employed for active recruitment. To identify patients for the introductory mailing, the only inclusion criteria were having a single encounter in VA's electronic medical record with a PTSD diagnosis and living in one of three Puget Sound counties. The only exclusion criteria were a diagnosis of schizophrenia, bipolar disorder or other psychotic disorder. Likely in part because of this broad approach, which did not include upper age limits, approximately one-quarter of contacted individuals reported a health condition that impaired their walking. Burdensome study procedures may have also impacted recruitment. About half of interested participants failed to complete the online screening questionnaire and others informed us that they had trouble completing the online questionnaire. Imposing an upper age limit (eg, 65 years) and restructuring the recruitment process to make it faster and easier for potential participants may be necessary.
Accessibility of the intervention and restrictive eligibility criteria may have also impacted recruitment. In addition to being able to walk over uneven ground for at least 2 hours, participants also had to be available during the times selected, have low suicide risk and be free from physical conditions such as high blood pressure (or obtain their primary care provider's permission) among several other criteria. Changing inclusion criteria (eg, eliminating restrictions related to suicidality) might improve recruitment and generalisability, but would require tradeoffs related to safety and retention that must be considered carefully.
Lastly, about 38% (159/415) of those for whom we were able to assess eligibility and interest declined participation. While some of these people may have declined because of the additional burdens of a research study, this statistic indicates that hiking may only appeal to a segment of the population, just as psychotherapy and pharmacotherapy only appeal to subsets of the population. [bib_ref] First-Line psychotherapies for Military-Related PTSD, Steenkamp [/bib_ref] Because of differences in treatment preferences, offering options is important and nature hiking merits consideration so that we can rigorously assess its efficacy.
## Retention
Retention varied by group and was poorer for the wristworn activity monitor than for the questionnaires. The activity monitor had a substantial amount of missing data, which is a common problem for activity monitors, [bib_ref] Physical activity in the United States measured by Accelerometer, Troiano [/bib_ref] and may have been related to the number of technical steps required for setting up the monitor and syncing it, as many participants needed additional help to troubleshoot problems. Providing more support to set up the monitor and incentives to wear and sync it may help to obtain more complete data. While overall questionnaire completion was high, it was higher in the nature hiking group (91%) than in the urban hiking group (68%). Though the small sample and our inability to conduct interviews with those who did not complete follow-up measures makes inference difficult, the retention differences could be a marker of commitment to the study. Future studies should pay careful attention to marketing the study to ensure that both interventions are perceived as helpful. Enhancing the social aspects of the interventions may help achieve that goal. The difference in incentives provided for questionnaire completion vs the other aspects of the study may also have played a role in retention for different study aspects. However, many participants shared that they participated to help fellow Veterans, indicating altruistic/intrinsic motivators for participation, reinforcing the importance of understanding drivers of participation, and reducing barriers and enhancing facilitators.
## Acceptability of the hiking interventions
Hike attendance (56%) was lower than our target (70%) and women had lower attendance in the nature hiking group than men. While we were unable to ascertain reasons for missing hikes for each person, some reasons reported (eg, other plans, work) were hard to avoid, while others (driving distance to hikes) could be addressed in the future by restricting the geographical area of recruitment and hikes and/or organising small groups at different times to accommodate individuals' schedules. Our study, unfortunately, does not shed light on the optimal hike 'dose'. We suspect that 8-12 hikes (similar to a standard psychotherapy course) may be optimal for achieving clinically meaningful results. Additional research will be necessary to examine this important question. There were also an indication of lower acceptability/ratings for information sharing in the urban hiking versus the nature hiking groups. For nature hikes, we listed a trail. For urban hikes, we shared information about the urban area, but did not provide an exact hiking route, which may have made it more difficult for participants to research urban hikes. Providing a map of the route might help participants feel prepared. Regarding differences in attendance by gender, a history of military sexual trauma, which is common among women Veterans, 85 may have impacted some women participants' comfort and perception of safety of hiking in nature with a majority male group. Ensuring a greater proportion of women in each group or organising women-only groups (as was suggested by some participants) could address this concern. These changes, would, however, result in additional costs and trade-offs that would need to be carefully considered.
# Conclusions
This pilot study provided useful information related to feasibility and acceptability, including common factors that resulted in exclusion; resources and procedures needed for recruitment; factors to consider for selection Open access of nature and urban hikes; and barriers and facilitators to achieving high completion in follow-up assessments and the hikes. These insights can be harnessed to increase participation and rigour in future, scaled-up iterations of the study and ensure that environments are safe (ie, non-triggering). Future studies with larger sample sizes are needed to isolate the ways nature contact, PA and social support conferred by the group impact outcomes to develop and provide well-tailored interventions.
Contributors AJL, GNB, CCE, JCF, KL, JB, and HF conceptualised the study and contributed to the intervention development and design. AJL and GNB oversaw the conduct of the trial all the authors contributed to the ongoing management of the trial. AJL, GNB, MM, AP, and CK oversaw data collection and data analysis. The manuscript was drafted by AJL. All the authors contributed to the interpretation of the findings and revised and reviewed the paper.
Funding This work was funded by Recreational Equipment, Inc. (REI; Award/Grant number is not applicable) and supported by equipment and outfitting contributions from Outdoor Research.
Disclaimer The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs, the US Government, Recreational Equipment (REI), or Outdoor Research.
## Competing interests none declared.
Patient consent for publication Not applicable.
Ethics approval This study was approved by the institutional review boards at the VA Puget Sound Healthcare System (MIRB 01738) and the University of Washington (6951).
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available on reasonable request.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
[fig] Figure 1: Conceptual model. [/fig]
[fig] Figure 2: Depiction of study design and assessments. PTSD, post-traumatic stress disorder. [/fig]
[fig] Figure 3: CONSORT diagram. CONSORT, Consolidated Standards of Reporting Trials. [/fig]
[table] Table 1: Baseline characteristics of Veterans in the urban and nature hiking groups [/table]
|
Neurological and Psychological Characteristics of Young Nitrous Oxide Abusers and Its Underlying Causes During the COVID-19 Lockdown
Background: The COVID-19 pandemic has a serious impact on the mental health of the public due to its economic and social impact. And psychological effects have led to drug and alcohol abuse. After the city lifted the lockdown, we consecutively encountered several young nitrous oxide abusers admitted to hospital for neurological treatment.Purpose: To inform physician decisions and social intervention, this observational study aimed at investigating the neurological and psychological characteristics of nitrous oxide abusers and its underlying causes during the COVID-19 lockdown.Methods: The nitrous oxide abusers who sought neurological treatment at our hospital between May 2020 and June 2020 were enrolled. Clinical data including socio-demographic, physical examination, laboratory examination, electromyography and neuroimaging were collected. Their motivations for inhaling nitrous oxide, knowledge about the nitrous oxide abuse and the accompanying of family were investigated face to face. Psychological status was assessed by the Symptom Checklist 90 (SCL-90) psychological evaluation.Results: Six nitrous oxide abusers were enrolled and the age was 22 ± 4.3. Clinical presentations included varying degrees of limb numbness and an ataxic gait. Laboratory examination revealed that all the patients did not have pernicious anemia, 4 patients had decreased vitamin B12 while 3 patients exhibited elevated homocysteine levels. MR of the spinal cord revealed that 4 patients had abnormal signals in the cervical spinal cord of high symmetry with splayed or inverted V sign after T2WI. Electromyogram (EMG) test showed 5 patients had peripheral nerve damage. The SCL-90 psychological evaluation results indicated that all patients had severe anxiety, depression and psychosis and they had severer psychological problems than ordinary citizens. Their motives for inhaling nitrous oxide are to relieve boredom, curiosity and buddy pressure. Their family spent <1 day per week to stay with them during city lockdown. Conclusion: The enrolled patients caused by abuse of nitrous oxide presented with symptoms of subacute combined with spinal degeneration. They had more serious psychological problems related to the COVID-19 pandemic. These cases make us value the psychological problems of young people under the outbreak and take multi-layered measures from families, schools (companies), hospitals, and governments to address it.
# Methods:
The nitrous oxide abusers who sought neurological treatment at our hospital between May 2020 and June 2020 were enrolled. Clinical data including socio-demographic, physical examination, laboratory examination, electromyography and neuroimaging were collected. Their motivations for inhaling nitrous oxide, knowledge about the nitrous oxide abuse and the accompanying of family were investigated face to face. Psychological status was assessed by the Symptom Checklist 90 (SCL-90) psychological evaluation.
Results: Six nitrous oxide abusers were enrolled and the age was 22 ± 4.3. Clinical presentations included varying degrees of limb numbness and an ataxic gait. Laboratory examination revealed that all the patients did not have pernicious anemia, 4 patients had decreased vitamin B12 while 3 patients exhibited elevated homocysteine levels. MR of the spinal cord revealed that 4 patients had abnormal signals in the cervical spinal cord of high symmetry with splayed or inverted V sign after T2WI. Electromyogram (EMG) test showed 5 patients had peripheral nerve damage. The SCL-90 psychological evaluation results indicated that all patients had severe anxiety, depression and psychosis and they had severer psychological problems than ordinary citizens. Their motives for inhaling nitrous oxide are to relieve boredom, curiosity and buddy pressure. Their family spent <1 day per week to stay with them during city lockdown.
# Introduction
The COVID-19 pandemic and the ensuing lockdown have had a serious impact on the physical and mental health of the public [bib_ref] The psychological status of people affected by the COVID-19 outbreak in China, Zhu [/bib_ref]. And the psychological effects have led to drug and alcohol abuse [bib_ref] Substance use and abuse, COVID-19-related distress, and disregard for social distancing: a..., Taylor [/bib_ref]. After the city lifted the lockdown, we consecutively encountered some young nitrous oxide abusers who were admitted to hospital for neurological treatment, which was a significant increase compared with the same period. It is important to characterize their neurological and psychological outcomes and explore the underlying causes in order to improve the clinical management during the COVID-19 pandemic period.
# Materials and methods
## Study design and participants
Patients attending Taizhou Hospital of Zhejiang Province for care due to nitrous oxide abuse were consecutively enrolled. Clinical data including sociodemographic characteristics, physical examination, laboratory examination, electromyography, neuroimaging, and psychological assessment were obtained. The duration and frequency of nitrous oxide use, the sources of laughter were inquired. Their motivations for inhaling nitrous oxide were investigated face to face. The family environment, siblings, interpersonal relationships, personality traits, financial conditions, and academic performance were investigated. The survey about the time their parents or family member spent to stay with them and knowledge about the nitrous oxide abuse was carried out. The history of physical illness and family history were recorded. After the city was unsealed, the first two nitrous oxide abuser came to the hospital for neurological treatment at the same time. We thought it might be a phenomenon and therefore started this observational study. The study was done between May 2020 and June 2020.
All data were anonymized to comply with the provisions of personal data protection. The participants have provided their consent to publish the observational study, and the consent procedure was approved by the Ethics Committee of Taizhou Hospital of Zhejiang Province. All procedures were performed according to the guidelines of the institutional ethics committee
## Magnetic resonance imaging (mri)
An MRI scan of the cervical spine and brain was done to all patients. T1WI sequences included MRI sequences with and without gadolinium. Sagittal and axial images were obtained using T2-weighted MRI sequences. Data on the affected spinal cord segments (number of segments of the spine) and their positions on the sagittal image (cervical and thoracic vertebrae) were recorded.
## Electromyogram (emg)
Neurologic manifestations such as muscle weakness, sensory loss, and cognitive decline were recorded. Nerve conduction studies were performed on the median nerve, ulnar nerve, peroneal nerve, tibial nerve, and sural nerve depending on the clinical manifestations of patients. Compound muscle action potential (CMAP) amplitude, distal latency, sensory nerve action potential (SNAP) amplitude, and conduction velocity were detected using a full range functional EMG evoked potentiometer (Keypoint 9033A07, Denmark).
## Psychological assessment
Using Symptom Checklist 90 (SCL-90), the mental state of the patients and ordinary citizens was assessed by a professional psychiatrist. The severity of symptoms (normal, mild, moderate, partial severe, severe, degree from light to heavy) is determined by the number of standard deviations of the dimension score from the norm group mean.
## Statistical analyses
Data were analyzed by the Statistical Package for Social Sciences (IBM SPSS 16.0). Descriptive statistics and one-sample t-test were performed for data comparison between nitrous oxide abusers and ordinary citizens. Statistical significance was set at P ≤ 0.05.
# Results
## Sociodemographic and laboratory characteristics of the patients
From May 2020 to June 2020, six patients with nerve damage caused by nitrous oxide inhalation were consecutively admitted to our hospital. As shown in [fig_ref] TABLE 1 |: Sociodemographic and laboratory characteristics of the case series [/fig_ref] , the mean age of the 6 patients was 22 ± 4.3, four were college students while two were high school graduates. The average duration of nitrous oxide Half of the patients were only children and half had one sibling. One had a poor interpersonal relationship with his family, and one was doting by his parents. All the patients did not live with their family and their parents or family members spent <2 h a day or 1 day per week to stay with or care for them during city lockdown. Their personality traits were either introverted, or withdrawn, or perverted. Five patients were in good economic condition and one was moderate. Five patients had moderate academic performance and one was lower. In addition, they didn't know that nitrous oxide abuse could lead abnormality of neurological function.
The patients stated that the reasons for nitrous oxide abuse were the lack of employment or study during the pandemic, a history of nitrous oxide abuse and relapse during the pandemic, boredom, curiosity and peer pressure.
## Neurological characteristics of the patients and therapeutic process
In the physical examine, all patients presented with limb numbness and varying degrees of walking instability. Varying degrees of sensory impairment and sensory ataxia were exhibited among the patients. There was no case of a positive pathological sign or obvious damage to the pyramidal tract. EMG examination showed peripheral nerve damage in patients except case 6. The abusers had multiple motor and sensory axonal damage and myelin sheath change, or motor nerve damage, or partial nerve damage. One case (case 4) had decreased muscle strength. Mild memory loss was documented in 2 patients (data not shown).
The imaging results showed that four patients had high symmetric signals with splayed or inverted V signs in the cervical spinal cord after MR T2WI. Case 4 exhibited a slightly high signal while case 1 did not exhibit any imaging abnormalities (see [fig_ref] FIGURE 1 |: Results of MR T2WI showing high symmetry signal in the cervical spinal... [/fig_ref]. All the patients presented no symptoms caused by autoimmune encephalitis, intracranial infection, cerebrovascular disease, brain trauma, tumor, or other toxic/metabolic causes, etc. All were diagnosed with subacute combined degeneration of the spinal cord (SCD). The six patients had improved neurological outcomes after vitamin B12 and adenosine cobalamin therapies for 5-8 days and discharged. They were prescribed vitamin B12 on discharge and told to return do a follow-up check on time.
All the patients had no history of physical illness or family history.
## Psychological characteristics of the nitrous oxide abusers
The SCL90 psychological evaluation results showed that the total score was 303.7 ± 43.1, each case was more than 250 points, and all cases had various psychological problems. The major severe psychological manifestations were anxiety, depression, hostility and psychosis (see [fig_ref] TABLE 2 |: The results of psychological assessment by Symptom Checklist 90 evaluation [/fig_ref].
To analyze the differences in the psychological status of nitrous oxide abusers and normal people, the SCL-90 score (1). It was noted that the SCL-90 scores of nitrous oxide abusers in anxiety, hostility, depression, interpersonal relationships, paranoia, psychosis and somatization were significantly higher than those of health controls, P < 0.01 [fig_ref] TABLE 3 |: Psychological status of the case series according to SCL-90, compared to ordinary... [/fig_ref] , indicating that these young nitrous oxide abusers presented severer psychological problems than ordinary citizens of the same age. In view of this situation, all the patients were asked to go to the psychological department for check-up after discharge from hospital.
# Discussion
Due to the COVID-19 pandemic and the lockdown that followed, public psychological problems cannot be ignored. In addition to the heightened mental stresses among patients and healthcare workers during the COVID-19 pandemic, the mental health of healthy people was also affected. There was a drastic increase in public fear, a decline in social and economic activities that triggered psychosocial sequelae. Quarantined individuals exhibited depression, fear, guilt and anger (4). Psychosocial [bib_ref] A longitudinal study on the mental health of general population during the..., Wang [/bib_ref]. The psychological effects lead to drug and alcohol abuse [bib_ref] The role of stress in drug addiction. An integrative review, Ruisoto [/bib_ref]. And it is notable that these problems are more likely to happen among children and adolescents [bib_ref] Adolescent inhalant abuse: environments of use, Mcgarvey [/bib_ref]. After the city lifted the lockdown, we consecutively encountered 6 nitrous oxide abusers who were hospitalized for neurological therapy within 1 month and they were all youth. It is notable that there were only 6 patients of nitrous oxide abuse were treated between October 2017 and December 2019 in our hospital and it cued the effect of the COVID-19 pandemic and the lockdown on public health especially the young. For physician decisions and social intervention, it was urgent to investigate the neurological characteristics and psychological state of them and the underlying causes of nitrous oxide abuse during the COVID-19 lockdown.
For more than 170 years, nitrous oxide has been used as an anesthetic in clinical practice. Its inhalation causes feelings of euphoria, involuntary laughter, distorted voices and mild hallucinations and it gradually becomes a popular way to relieve the pressures among the youth [bib_ref] The sharp rise of neurological disorders associated with recreational nitrous oxide use..., Zheng [/bib_ref] [bib_ref] Up: The rise of nitrous oxide abuse. An international survey of contemporary..., Kaar [/bib_ref]. A global drug survey (GDS2014) conducted in 17 countries involving 74,864 patients confirmed that the prevalence of nitrous oxide use as a recreational drug in the UK and US was 38 and 29.4%, respectively [bib_ref] Up: The rise of nitrous oxide abuse. An international survey of contemporary..., Kaar [/bib_ref]. Incidences of nitrous oxide abuse in China are gradually increasing, with the majority of the abusers being the youth [bib_ref] The sharp rise of neurological disorders associated with recreational nitrous oxide use..., Zheng [/bib_ref].
The adverse effects of exposure to nitrous oxide include slowed reaction rate, dizziness, nausea and vomiting. Inhalation of large quantities of nitrous oxide at a high pressure may lead to suffocation. Long-term adverse effects include nerve damage due to vitamin B12 deficiency, cobalamin reactive psychosis, and homocysteine accumulation [bib_ref] Nitrous oxide-induced B1 2 deficiency myelopathy: Perspectives on the clinical biochemistry of..., Hathout [/bib_ref]. Vitamin B12 is an important cofactor of cellular methionine synthase. Extremely low levels of vitamin B12 leads to methionine consumption and homocysteine accumulation. Methionine consumption leads to a decrease in downstream S-adenosine, which is required for myelin production and maintenance. Deficiency in vitamin B12 leads to demyelination and gliosis of the central nervous system (especially the dorsal spinal cord), as well as demyelination of peripheral nerves. Homocysteine accumulation increases the risk of stroke and peripheral neuropathy [bib_ref] Vitamin B12 and homocysteine levels and 6-year change in peripheral nerve function..., Leishear [/bib_ref].
Pernicious anemia and neurological damage caused by nitrous oxide are very common. Clinical manifestations of these conditions include paresthesia in limbs, gait instability or difficult walking, weakness, falls or balance disorders, Lhermitte's Sign and ataxia [bib_ref] Global burden related to nitrous oxide exposure in medical and recreational settings:..., Oussalah [/bib_ref]. Occasionally there is cognitive impairment and optic atrophy [bib_ref] Vitamin B12 and homocysteine levels and 6-year change in peripheral nerve function..., Leishear [/bib_ref] [bib_ref] Whippits, nitrous oxide and the dangers of legal highs, Thompson [/bib_ref]. In this study, all the 6 nitrous oxide abusers presented with limb numbness and varying degrees of walking instability. Two patients presented with mild memory loss, 4 presented with increased T2 signal in cervical spinal cord, 3 presented with extensive peripheral nerve damage, while 1 exhibited mild anemia. In terms of treatment, the neurological symptoms could be ameliorated by in time vitamin B12 supplementary [bib_ref] Nitrous oxide-induced B1 2 deficiency myelopathy: Perspectives on the clinical biochemistry of..., Hathout [/bib_ref] [bib_ref] Inhaled nitrous oxide 'Whip-Its!' causing subacute combined degeneration of spinal cord, Al-Sadawi [/bib_ref]. All the patients had improved neurological outcomes after vitamin B12 therapies and discharged.
Considering the impact of the COVID-19 epidemic on people's mental health [bib_ref] COVID-19 and its consequences on mental health (Review), Tsamakis [/bib_ref] , the psychological assessment was carried out. The results indicated varying degrees of anxiety, depression, hostility and psychosis and one case presented obvious suicidal tendency. The SCL-90 score of the 6 patients was significantly higher compared to that of healthy individuals. During COVID-19 pandemic and city lockdown, stressors such as university closures, social distancing, pessimism on the economic prospects are susceptible to lead development of mental health symptoms [bib_ref] COVID-19 and its consequences on mental health (Review), Tsamakis [/bib_ref]. Compared with the past, the increase in the number of hospitalizations caused by nitrous oxide abuse, and the increase of nitrous oxide dose or relapse reflected to a certain extent the impact of the COVID-19 pandemic on people's psychological status.
To explore the potential causes of nitrous oxide abuse during the COVID-19 lockdown, motivations for inhaling nitrous oxide, sociodemographic characteristics, family environment, siblings, interpersonal relationships, personality traits, financial conditions, and academic performance were investigated. The results showed that risk factors for nitrous oxide abuse included the lack of employment or study during the pandemic, a history of nitrous oxide abuse and relapse during the pandemic, boredom, curiosity and peer pressure, parental or family inconcern or doting, and possible good economic situation.
Although there were still many debates about the lockdown policy [bib_ref] A literature review and meta-analysis of the effects of lockdowns on COVID-19..., Herby [/bib_ref] , it did inhibit the spread of the SARS-CoV-2 and reduce the absolute number of deaths (20). We should focus more on solving the problems caused by the city lockdown such as the psychological problems and take effective measures for the above potential causes. It should enhance the combination of meaning-based coping and spirituality processes to mitigate the adverse effects of coronavirus stress on wellbeing [bib_ref] Meaning-based coping and spirituality during the COVID-19 pandemic: mediating effects on subjective..., Arslan [/bib_ref]. Multi-layered interventional measures from families, schools, hospitals, and governments should be implemented as early as possible. It is worth emphasizing that the patients' family rarely communicated with them during the lockdown. Loneliness is strongly associated with mental health problems [bib_ref] Advancing our understanding of loneliness and mental health problems in young people, Pitman [/bib_ref]. Therefore, the company of the family appears to be extremely important [bib_ref] Parenting in a time of COVID-19, Cluver [/bib_ref]. Indoor games, read and physical sports with the family are recommended. Despite of social distancing and school closures, on-line courses and virtual workshops where clinicianled mental health and psychosocial services such as stress control, drug abuse education are conducted should be encouraged. For those with obvious suicidal tendency, severe depression or other serious psychological problems, drug therapy intervention by psychiatrist needs to be involved [bib_ref] The Suicidal Risks and Treatments, Seek Medications from Multi-disciplinary, Lu [/bib_ref]. Lastly, government's measures should be taken to control the nitrous oxide flooding from the source such as recreation places [bib_ref] The sharp rise of neurological disorders associated with recreational nitrous oxide use..., Zheng [/bib_ref].
Disadvantage of this study: nitrous oxide abuse not only leads to peripheral neuropathy, SCD and other physiological diseases [bib_ref] Another extensive nitrous oxide (N2O) myelopathy: the wood for the trees?, Demas [/bib_ref] [bib_ref] No laughing matter: subacute degeneration of the spinal cord due to nitrous..., Keddie [/bib_ref] [bib_ref] Pulmonary embolism and deep vein thrombosis caused by nitrous oxide abuse: a..., Sun [/bib_ref] but also causes a series of abnormal mental symptoms, including personality changes, mood disorientations (such as anxiety, depression, mania), impulsive and aggressive behaviors, hallucinations, delusions and other psychotic symptoms [bib_ref] Neurologic, psychiatric, and other medical manifestations of nitrous oxide abuse: a systematic..., Garakani [/bib_ref]. We failed to obtain the psychological assessment data of the patients before the pandemic and before they started abusing nitrous oxide. The causal relationships between the pandemic and psychological changes, and between nitrous oxide abuse and psychological changes could not be explained. The second disadvantage is that the psychological status of the general population in the city was not obtained at the time of psychological assessment of the patients.
# Conclusion
The nitrous oxide abusers during the COVID-19 pandemic and lockdown presented SCD neurological symptoms and more serious psychological problems than healthy controls. In addition to the neurological therapy, more attention should be payed to the mental health of them. These young cases make us value the psychological problems of young people under the outbreak and it is imperative to take multi-layered, three-dimensional measures from families, schools (companies), hospitals, and governments to address it.
# Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
# Ethics statement
The studies involving human participants were reviewed and approved by Ethics Committee of Taizhou Hospital of Zhejiang Province. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.
[fig] FIGURE 1 |: Results of MR T2WI showing high symmetry signal in the cervical spinal posterior cord, presenting splayed or inverted V sign. On the left of each case is sagittal image and on the right is axial image. Red arrows indicate a clearly high signal. Yellow arrows indicate slightly higher or suspicious signals. [/fig]
[table] TABLE 1 |: Sociodemographic and laboratory characteristics of the case series. [/table]
[table] TABLE 2 |: The results of psychological assessment by Symptom Checklist 90 evaluation. [/table]
[table] TABLE 3 |: Psychological status of the case series according to SCL-90, compared to ordinary citizens during COVID-19. [/table]
|
Human papillomavirus and posttransplantation cutaneous squamous cell carcinoma: A multicenter, prospective cohort study
# | introduction
Cutaneous squamous cell carcinoma (cSCC) is the most common malignancy in solid organ transplant recipients (OTRs), followed by basal cell carcinoma (BCC). [bib_ref] Skin cancers after organ transplantation, Euvrard [/bib_ref] [bib_ref] Incidence of skin cancer after renal transplantation in The Netherlands, Hartevelt [/bib_ref] The incidence of cSCC and BCC increases with duration of immunosuppressive therapy. [bib_ref] Skin cancers after organ transplantation, Euvrard [/bib_ref] Important risk factors are male sex, increasing age, fair skin type, sun exposure, and smoking. [bib_ref] Keratotic skin lesions and other risk factors are associated with skin cancer..., Bavinck [/bib_ref] cSCCs are usually preceded by multiple viral warts from human papillomavirus (HPV) infection and by actinic keratoses, which also often contain HPV. [bib_ref] Keratotic skin lesions and other risk factors are associated with skin cancer..., Bavinck [/bib_ref] The virome component of the human microbiome is known to play an important role in disease, [bib_ref] The molecular revolution in cutaneous biology: investigating the skin microbiome, Kong [/bib_ref] [bib_ref] Biogeography and individuality shape function in the human skin metagenome, Oh [/bib_ref] and HPV is particularly prevalent in the normal skin virome. [bib_ref] Human skin microbiota: high diversity of DNA viruses identified on the human..., Foulongne [/bib_ref] [bib_ref] Human papillomavirus community in healthy persons, defined by metagenomics analysis of human..., Ma [/bib_ref] HPVs are double-stranded DNA viruses that are classified into 5 genera (α, β, γ, μ, and ν). [bib_ref] Human papillomavirus molecular biology and disease association, Doorbar [/bib_ref] [bib_ref] Beta genus papillomaviruses and skin cancer, Howley [/bib_ref] αHPVs (eg, mucosal HPV types are responsible for the development of cervical carcinoma and other mucosal SCCs, including anogenital and oropharyngeal carcinomas. [bib_ref] Five-year cervical (pre)cancer risk of women screened by HPV and cytology testing, Uijterwaal [/bib_ref] The role of skin βPVs in cSCC carcinogenesis is controversial, [bib_ref] Beta genus papillomaviruses and skin cancer, Howley [/bib_ref] [bib_ref] Cutaneous HPV and skin cancer, Accardi [/bib_ref] [bib_ref] Association between beta-genus human papillomavirus and cutaneous squamous cell carcinoma in immunocompetent..., Chahoud [/bib_ref] but evidence for etiologic involvement is accumulating. [bib_ref] Milk of human kindness?-HAMLET, human papillomavirus, and warts, Bavinck [/bib_ref] [bib_ref] Papillomavirus-associated squamous skin cancers following transplant immunosuppression: one Notch closer to control, Connolly [/bib_ref] [bib_ref] Human Betapapillomavirus infection and keratinocyte carcinomas, Quint [/bib_ref] [bib_ref] Spontaneous tumour development in human papillomavirus type 8 E6 transgenic mice and..., Marcuzzi [/bib_ref] [bib_ref] E6 and E7 from beta HPV38 cooperate with ultraviolet light in the..., Viarisio [/bib_ref] βPV cause latent, persistent skin infections [bib_ref] Betapapillomaviruses frequently persist in the skin of healthy individuals, De Koning [/bib_ref] and were first discovered in cSCCs from patients with epidermodysplasia verruciformis, a rare genetic skin disease characterized by increased susceptibility to βPV with a high risk of cSCC on sun-exposed skin. [bib_ref] Beta genus papillomaviruses and skin cancer, Howley [/bib_ref] βPVs are ubiquitous, with more than 90% of people carrying these viruses as part of their normal skin virome. [bib_ref] Human papillomavirus community in healthy persons, defined by metagenomics analysis of human..., Ma [/bib_ref] [bib_ref] Prevalence and associated factors of betapapillomavirus infections in individuals without cutaneous squamous..., De Koning [/bib_ref] [bib_ref] Case-control study of genusbeta human papillomaviruses in plucked eyebrow hairs and cutaneous..., Iannacone [/bib_ref] Hair follicles are the likely reservoir. [bib_ref] Betapapillomaviruses frequently persist in the skin of healthy individuals, De Koning [/bib_ref] βPV-infected cells may have impaired DNA repair and decreased sensitivity to apoptosis induced by ultraviolet radiation (UVR), facilitating escape from normal cellular defense mechanisms. [bib_ref] Beta genus papillomaviruses and skin cancer, Howley [/bib_ref] [bib_ref] Milk of human kindness?-HAMLET, human papillomavirus, and warts, Bavinck [/bib_ref] A substantial body of epidemiologic data show an association between HPV infection and cSCC, [bib_ref] Association between beta-genus human papillomavirus and cutaneous squamous cell carcinoma in immunocompetent..., Chahoud [/bib_ref] [bib_ref] Case-control study of genusbeta human papillomaviruses in plucked eyebrow hairs and cutaneous..., Iannacone [/bib_ref] [bib_ref] Multicenter study of the association between betapapillomavirus infection and cutaneous squamous cell..., Bavinck [/bib_ref] [bib_ref] Genus beta human papillomaviruses and incidence of basal cell and squamous cell..., Karagas [/bib_ref] [bib_ref] Prospective study of human papillomavirus seropositivity and risk of nonmelanoma skin cancer, Andersson [/bib_ref] [bib_ref] Cutaneous alpha, beta and gamma human papillomaviruses in relation to squamous cell..., Farzan [/bib_ref] in particular, in OTRs. [bib_ref] A case-control study of betapapillomavirus infection and cutaneous squamous cell carcinoma in..., Proby [/bib_ref] [bib_ref] The presence of betapapillomavirus antibodies around transplantation predicts the development of keratinocyte..., Genders [/bib_ref] In most studies, HPV infection has been determined by detection of serologic responses to a large series of βPV and sometimes γPV types [bib_ref] Association between beta-genus human papillomavirus and cutaneous squamous cell carcinoma in immunocompetent..., Chahoud [/bib_ref] [bib_ref] Multicenter study of the association between betapapillomavirus infection and cutaneous squamous cell..., Bavinck [/bib_ref] [bib_ref] Genus beta human papillomaviruses and incidence of basal cell and squamous cell..., Karagas [/bib_ref] [bib_ref] Prospective study of human papillomavirus seropositivity and risk of nonmelanoma skin cancer, Andersson [/bib_ref] [bib_ref] A case-control study of betapapillomavirus infection and cutaneous squamous cell carcinoma in..., Proby [/bib_ref] [bib_ref] The presence of betapapillomavirus antibodies around transplantation predicts the development of keratinocyte..., Genders [/bib_ref] [bib_ref] Seropositivity for human papillomavirus and incidence of subsequent squamous cell and basal..., Paradisi [/bib_ref] [bib_ref] The Association between cutaneous squamous cell carcinoma and betapapillomavirus seropositivity: a cohort..., Plasmeijer [/bib_ref] [bib_ref] Case-control study of cutaneous human papillomaviruses in squamous cell carcinoma of the..., Iannacone [/bib_ref] [bib_ref] Sunlight exposure and cutaneous human papillomavirus seroreactivity in basal cell and squamous..., Iannacone [/bib_ref] [bib_ref] Cutaneous human papillomaviruses and squamous cell carcinoma of the skin: nested case-control..., Faust [/bib_ref] and less frequently by detecting genomic DNA from 1 or multiple HPV types in eyebrow hair or skin scrapings. [bib_ref] Case-control study of genusbeta human papillomaviruses in plucked eyebrow hairs and cutaneous..., Iannacone [/bib_ref] [bib_ref] Multicenter study of the association between betapapillomavirus infection and cutaneous squamous cell..., Bavinck [/bib_ref] [bib_ref] A case-control study of betapapillomavirus infection and cutaneous squamous cell carcinoma in..., Proby [/bib_ref] [bib_ref] Human papillomavirus load in eyebrow hair follicles and risk of cutaneous squamous..., Neale [/bib_ref] [bib_ref] Deep sequencing extends the diversity of human papillomaviruses in human skin, Bzhalava [/bib_ref] [bib_ref] Human papillomavirus type 197 is commonly present in skin tumors, Arroyo Muhr [/bib_ref] Case-control studies in both OTR and immunocompetent populations have shown that the presence of βPV DNA or antibodies was associated with a 1.5-to 3-fold increased risk of cSCC. [bib_ref] Association between beta-genus human papillomavirus and cutaneous squamous cell carcinoma in immunocompetent..., Chahoud [/bib_ref] [bib_ref] Case-control study of genusbeta human papillomaviruses in plucked eyebrow hairs and cutaneous..., Iannacone [/bib_ref] [bib_ref] Multicenter study of the association between betapapillomavirus infection and cutaneous squamous cell..., Bavinck [/bib_ref] [bib_ref] Genus beta human papillomaviruses and incidence of basal cell and squamous cell..., Karagas [/bib_ref] [bib_ref] Prospective study of human papillomavirus seropositivity and risk of nonmelanoma skin cancer, Andersson [/bib_ref] [bib_ref] Cutaneous alpha, beta and gamma human papillomaviruses in relation to squamous cell..., Farzan [/bib_ref] [bib_ref] A case-control study of betapapillomavirus infection and cutaneous squamous cell carcinoma in..., Proby [/bib_ref] [bib_ref] The presence of betapapillomavirus antibodies around transplantation predicts the development of keratinocyte..., Genders [/bib_ref] [bib_ref] Case-control study of cutaneous human papillomaviruses in squamous cell carcinoma of the..., Iannacone [/bib_ref] [bib_ref] Cutaneous human papillomaviruses and squamous cell carcinoma of the skin: nested case-control..., Faust [/bib_ref] [bib_ref] Human papillomavirus load in eyebrow hair follicles and risk of cutaneous squamous..., Neale [/bib_ref] It was not possible, however, to determine whether the βPV infection increases risk of cSCC development or whether cSCC formation promotes active proliferation of βPV. To our best knowledge, there is only 1 published cohort study in OTRs that investigated the influence of βPV infection on the development of cSCC. [bib_ref] The presence of betapapillomavirus antibodies around transplantation predicts the development of keratinocyte..., Genders [/bib_ref] In a single-center study in 445 patients, OTRs who were βPV seropositive around the time of transplantation had an almost 3-fold increased risk of developing cSCC during the 22-year follow-up period. [bib_ref] The presence of betapapillomavirus antibodies around transplantation predicts the development of keratinocyte..., Genders [/bib_ref] There are no prospective studies in which the associations between cSCC and both the number and DNA load of βPV types have been examined.
In this study, we present data from 2 prospective multicenter cohort studies confirming that the presence of human skin βPV infection is associated with cSCC development in OTRs.
# | methods
## | participants and study design
As part of a European Union-funded fifth framework program (5FP) collaborative research grant (QLK2-CT-2002-0117), the EPI-HPV-UV-CA study group collected data for a prospective [bib_ref] Longitudinal study of seroprevalence and serostability of 34 human papillomavirus types in..., Antonsson [/bib_ref] and a case-control study between 2003 and 2006. [bib_ref] Keratotic skin lesions and other risk factors are associated with skin cancer..., Bavinck [/bib_ref] [bib_ref] A case-control study of betapapillomavirus infection and cutaneous squamous cell carcinoma in..., Proby [/bib_ref] A description of the inclusion and exclusion criteria is provided in the supplementary material (Figures S1A-S1C). Cohort 1 was assembled from 441 OTRs followed until death or cessation of follow-up in 2016. βPV infection was assessed in eyebrow hair by using polymerase chain reaction-based methods. βPV IgG seroresponses were determined with multiplex serology. A competing risk model with delayed entry was used to estimate cumulative incidence of histologically proven cSCC and the effect of βPV by using a multivariable Cox regression model. Results are reported as adjusted hazard ratios (HRs). OTRs with 5 or more different βPV types in eyebrow hair had 1.7 times the risk of cSCC vs OTRs with 0 to 4 different types (HR 1.7, 95% confidence interval 1.1-2.6). A similar risk was seen with high βPV loads (HR 1.8, 95% confidence interval 1. . No significant associations were seen between serum antibodies and cSCC or between βPV and basal cell carcinoma. The diversity and load of βPV types in eyebrow hair are associated with cSCC risk in OTRs, providing evidence that βPV is associated with cSCC carcinogenesis and may present a target for future preventive strategies. [bib_ref] Evaluation of a novel highly sensitive, broad-spectrum PCR-reverse hybridization assay for detection..., De Koning [/bib_ref] and for viral load determination using quantitative polymerase chain reaction. [bib_ref] Human papillomavirus load in eyebrow hair follicles and risk of cutaneous squamous..., Neale [/bib_ref] Because more than 95% of OTRs had 1 or more βPV types in their eyebrow hair, [bib_ref] A case-control study of betapapillomavirus infection and cutaneous squamous cell carcinoma in..., Proby [/bib_ref] we dichotomized the number of βPV as 0 to 4 and as 5 or more, as previously published. [bib_ref] A case-control study of betapapillomavirus infection and cutaneous squamous cell carcinoma in..., Proby [/bib_ref] High viral load was defined as 1 βPV copy per 1 to 19 cells and low load as 1 copy per 20 or more cells. Serum was collected and tested for 16 βPV, 6 γPV, 4 αPV, 2 μPV, and 1 νPV IgG antibodies with a multiplex serology technique based on glutathione s-transferase capture ELISA in combination with fluorescent bead technology. 37,38 All 3 techniques used have been previously described in more detail. [bib_ref] A case-control study of betapapillomavirus infection and cutaneous squamous cell carcinoma in..., Proby [/bib_ref] [bib_ref] Human papillomavirus load in eyebrow hair follicles and risk of cutaneous squamous..., Neale [/bib_ref] [bib_ref] Longitudinal study of seroprevalence and serostability of 34 human papillomavirus types in..., Antonsson [/bib_ref] [bib_ref] Evaluation of a novel highly sensitive, broad-spectrum PCR-reverse hybridization assay for detection..., De Koning [/bib_ref] [bib_ref] Seroprevalence of 34 human papillomavirus types in the German general population, Michael [/bib_ref] [bib_ref] Multiplex human papillomavirus serology based on in situ-purified glutathione s-transferase fusion proteins, Waterboer [/bib_ref] In cohort 1, eyebrow hair and serum were collected at 1, 3, 6, 9, A competing risk model with death as competing event and delayed entry (also referred to as left truncation) was used to estimate the cumulative incidence of first cSCC or BCC since transplantation. [bib_ref] Revision surgery is overestimated in hip replacement, Keurentjes [/bib_ref] [bib_ref] Tutorial in biostatistics: competing risks and multi-state models, Putter [/bib_ref] Competing risk models take into account situations where more than 1 cause of failure is possible. In our study, the 2 competing events were cSCC or BCC and death, as patients might have died beforethe occurrence of cSCC or BCC. Delayed entry was used to adjust for the fact that in cohort 2 the HPV DNA testing and serology were not performed at the time of transplantation and that we had taken the 12-month time point in cohort 1.
# | statistical analysis
We had previously defined a new variable based on concordance of βPV DNA and serology. We classified OTR according to whether they were both antibody and DNA positive for the same βPV type as follows: antibody negative, regardless of DNA status; antibody positive but with no types for which DNA was also found; and antibody positive with at least 1 type concordant for DNA. [bib_ref] A case-control study of betapapillomavirus infection and cutaneous squamous cell carcinoma in..., Proby [/bib_ref] The final analyses were adjusted for sex, age, skin type, and study center (study center being an important composite variable representing differences in patient populations, immunosuppression regimens, and sun exposure between the 3 clinical sites). We also investigated the other clinical variables, and none changed the estimates of interest. These were, therefore, not included in the analysis. Analyses of cohort 2 or of the combined cohorts showed that adjustment for type of immunosuppression, type of organ transplanted, average sun exposure, number of sunburns, smoking, or alcohol consumption did not substantially change the hazard ratios (HRs) (data not shown). We did not adjust for keratotic skin lesions and common viral warts, as these potentially lie on the causal pathway between HPV infection and cSCC or BCC and their inclusion in the model may lead to overadjustment, biasing the result toward the null. [bib_ref] Overadjustment bias and unnecessary adjustment in epidemiologic studies, Schisterman [/bib_ref] The analyses concerning the competing risk model were performed in R environment (http://www.R-project.org) with the mstate library. [bib_ref] The mstate package for estimation and prediction in non-and semi-parametric multi-state and..., De Wreede [/bib_ref] All other analyses were performed with SPSS Statistics for Windows version 23.0 (IBM Corp., Armonk, NY).
# | results
The median and maximum follow-up times since transplantation were 9.5 and 12.4 years in cohort 1 and 18.7 and 29.9 years in cohort 2, respectively. The median and maximum follow-up times since DNA and serum sampling did not differ between the cohorts and were 9. The clinical profile of transplanted patients differed between cohorts 1 and 2 , but there were no significant differences in virologic results and association with cSCC. OTRs in cohort 1 were more often female, had a darker skin type, were less often smokers, and had lower alcohol consumption. Almost all received mycophenolate mofetil in combination with tacrolimus, rather than azathioprine, which was the main immunosuppressive drug in cohort 2. The numbers of keratotic lesions and common viral warts were higher in cohort 2, consistent with their longer duration of immunosuppression . After adjustment, sun exposure, painful sunburns, and smoking were not associated with cSCC in the combined cohort [fig_ref] Table 2: shows the cause-specific HRs for prognostic factors as-we had insufficient statistical power... [/fig_ref].
The number of βPV types in eyebrow hair, βPV DNA load, and serologic responses to HPV did not significantly differ between the 2 cohorts . shows a higher overall cumulative incidence of cSCC in OTRs infected with multiple (≥5) HPV types mea- High βPV load was also associated with a significantly higher cumulative incidence of cSCC compared with a low load or absent βPV types and [fig_ref] Table 3: reports the association between presence of βPV DNA in eyebrow hair and... [/fig_ref]. In the combined cohort, the adjusted HR was 1.8 (95% CI 1.2-2.8). The βPV load risk factor in cohort 1 was probably driving the HRs in the combined analyses.
HPV seropositivity at the time of sampling was not associated with cSCC risk and [fig_ref] Table 3: reports the association between presence of βPV DNA in eyebrow hair and... [/fig_ref]. The association was stronger but still not significant for concordant serologic responses and [fig_ref] Table 3: reports the association between presence of βPV DNA in eyebrow hair and... [/fig_ref]. We observed a weak association between γPV seropositivity and cSCC [fig_ref] Table 3: reports the association between presence of βPV DNA in eyebrow hair and... [/fig_ref] but no association with αPV, μPV, and νPV seropositivity [fig_ref] Table 3: reports the association between presence of βPV DNA in eyebrow hair and... [/fig_ref]. We examined possible heterogeneity of βPV effects between cohorts 1 and 2, but this was not the case in the univariate and adjusted analyses for any of the variables in [fig_ref] Table 3: reports the association between presence of βPV DNA in eyebrow hair and... [/fig_ref]. We were unable to assess the influence of type of organ transplanted because our study populations were mostly kidney transplant recipients. There was no evidence of violation of the proportional hazard assumption. .
# | discussion
In this observational hospital-based study of 2 OTR cohorts with a follow-up period after HPV sampling of longer than 10 years, we found that the diversity and load of infecting βPV types were associated with an approximately doubled risk of cSCC. Our results were robust after adjustment for multiple potential confounders. This level of increased risk is equivalent to that associated with other well-established risk factors such as skin phototype 3 and points to an important role for the skin virome and specifically HPV in the pathogenesis of cSCC.
HPV constitute an important part of the normal human skin virome. [bib_ref] Human skin microbiota: high diversity of DNA viruses identified on the human..., Foulongne [/bib_ref]
## (1.3-4.1) 2.5 (1.3-5.0)
Missing values
All hazard ratios (HRs) are calculated using delayed entry and adjusting for competitive risk of death. Significant HRs are indicated in bold. Aza, azathioprine; CyA, cyclosporin A; MMF, mycophenolate mofetil; SCC, squamous cell carcinoma; Tac, tacrolimus.
a The adjusted HRs are calculated with the factor of interest and sex, age at physical examination, skin type, and study center included in the model. that HPV plays a role in cSCC carcinogenesis, principally through synergy with UVR. [bib_ref] Milk of human kindness?-HAMLET, human papillomavirus, and warts, Bavinck [/bib_ref] [bib_ref] Papillomavirus-associated squamous skin cancers following transplant immunosuppression: one Notch closer to control, Connolly [/bib_ref] [bib_ref] Human Betapapillomavirus infection and keratinocyte carcinomas, Quint [/bib_ref] In normal cells, UVR upregulates cellular defense processes leading to p53 activation, cell cycle arrest, apoptosis, or DNA repair. Several βPV types (eg, HPV5, HPV8, HPV38, and HPV49) deregulate these crucial cellular regulatory pathways by targeting multiple transcription factors and/or transcriptional regulators, resulting in infected cells that are highly susceptible to chromosomal instability and malignant transformation by UVR. [bib_ref] Beta genus papillomaviruses and skin cancer, Howley [/bib_ref] [bib_ref] Human Betapapillomavirus infection and keratinocyte carcinomas, Quint [/bib_ref] [bib_ref] Viral oncogenesis and its role in nonmelanoma skin cancer, St [/bib_ref] Synergy between βPV and UVR has also been documented in animal models. In transgenic mice expressing the HPV8 early region or E6 gene alone, a single dose of UVR rapidly promoted papillomas and cSCC formation, [bib_ref] Spontaneous tumour development in human papillomavirus type 8 E6 transgenic mice and..., Marcuzzi [/bib_ref] and in HPV38
E6/E7 transgenic mice, UVR also induced the development of actinic keratoses and cSCC. [bib_ref] E6 and E7 from beta HPV38 cooperate with ultraviolet light in the..., Viarisio [/bib_ref] Certain βPV types also appear to have enhanced replication in the context of host immunosuppression, likely accompanied by increased oncogene expression and oncogenic activity. [bib_ref] Beta-papillomavirus DNA loads in hair follicles of immunocompetent people and organ transplant..., Weissenborn [/bib_ref] Previous epidemiologic studies investigating the association between βPV and cutaneous squamous carcinogenesis have generated inconsistent findings. [bib_ref] Beta genus papillomaviruses and skin cancer, Howley [/bib_ref] [bib_ref] Cutaneous HPV and skin cancer, Accardi [/bib_ref] [bib_ref] Association between beta-genus human papillomavirus and cutaneous squamous cell carcinoma in immunocompetent..., Chahoud [/bib_ref] One of the challenges is the diversity of recognized HPV types within the normal skin virome. [bib_ref] Deep sequencing extends the diversity of human papillomaviruses in human skin, Bzhalava [/bib_ref] [bib_ref] Human papillomavirus type 197 is commonly present in skin tumors, Arroyo Muhr [/bib_ref] More than 200 HPV types have been identified, and the number of putative new types is increasing. [bib_ref] Deep sequencing extends the diversity of human papillomaviruses in human skin, Bzhalava [/bib_ref] [bib_ref] Human papillomavirus type 197 is commonly present in skin tumors, Arroyo Muhr [/bib_ref] [bib_ref] Proteomic analysis of the gamma human papillomavirus type 197 E6 and E7..., Grace [/bib_ref] We tested for only a subset of HPV types. Our polymerase chain reaction-based βPV DNA detection technique was state-of-theart when we initiated this study and remains a reliable and reproducible method. It is possible, however, that the true association between βPV infection and cSCC is even stronger, as we may have failed to detect additional relevant HPV infections. We also observed that seropositivity to γPV was associated with subsequent cSCC, supporting recent findings that γPV type HPV197 is commonly present in skin tumors [bib_ref] Human papillomavirus type 197 is commonly present in skin tumors, Arroyo Muhr [/bib_ref] and that its E6 and E7 proteins interact with a set of cellular proteins similar to those encoded by genital HPVs linked to human carcinogenesis. [bib_ref] Proteomic analysis of the gamma human papillomavirus type 197 E6 and E7..., Grace [/bib_ref] Another limitation of this study was the lack of follow-up data in Berlin and Verona. While this decreased the power of the study, we believe that it is unlikely to have introduced bias because all patients in these 2 centers were excluded, independent of their medical history.
Although cohort 1 was ideally designed for analysis of the involvement of HPV in cSCC development, we were hampered by the low number of events during the first 10 years after transplantation. Therefore, cohorts 1 and 2 were combined, which in the crude associations between the presence, number, and load of βPV DNA with the later development of cSCC appeared similar. There were also differences between the 2 cohorts. In recent years, older patients were more frequently transplanted, the immunosuppressive regimen changed, and a larger proportion of patients with a darker skin phototype were transplanted, especially in Leiden and London.
Other potential limitations of this study are missing data in the clinical characteristics included in the multivariable analyses. Of All hazard ratios (HRs) are calculated using delayed entry and adjusting for competitive risk of death. Significant HRs are indicated in bold.
a The detailed numbers of squamous cell carcinoma per cohort and per exposure are provided in the supplementary material [fig_ref] Table 3: reports the association between presence of βPV DNA in eyebrow hair and... [/fig_ref]. b Adjusted HRs are not available for cohort 1 because of low number of events. c The HRs are adjusted for sex, age at physical examination, skin type, and study center. particular concern is the 28% missing data for alcohol consumption, but we found no evidence that adjusting for alcohol in those with complete information made any difference to the estimates of association between HPV and cSCC. We have no good explanation why we did not find a statistically significant association with βPV serology. We have shown a 40% increased risk of cSCC overall in seropositive OTRs but had to exclude a substantial number of patients because of missing data, decreasing the power of our study, which may be a possible explanation of our failure to show a statistically significant association between βPV serology and cSCC.
This cohort study provides evidence that infection with βPV is associated with the development of cSCC. βPV vaccines are now in development, and our data provide a rationale for routine pretrans-
[fig] ported earlier, 3: the number of keratotic skin lesions and common viral warts was strongly associated with the development of cSCC, with an adjusted HR of 4.7 (95% confidence interval [CI] 2.0-10.9) for OTRs with 50 or more lesions compared with those OTRs with fewer lesions. [/fig]
[table] Table 2: shows the cause-specific HRs for prognostic factors as-we had insufficient statistical power to assess differences in risk of cSCC according to type of organ transplant received. As we have re- [/table]
[table] Table 3: reports the association between presence of βPV DNA in eyebrow hair and subsequent development of cSCC. The combined [/table]
|
Immunomodulatory Factors in Primary Endometrial Cell Cultures Isolated from Cancer and Noncancerous Human Tissue–Focus on RAGE and IDO1
Citation: Tkaczuk-Włach, J.; Kędzierski, W.; Jonik, I.; Sadok, I.; Filip, A.; Kankofer, M.; Polkowski, W.; Ziółkowski, P.; Gamian, A.; Staniszewska, M. Immunomodulatory
# Introduction
Endometrial microenvironment constitutes very dynamic composition of several cell types, i.e., fibroblasts, epithelial, endothelial, and immune cells like macrophages or NK cells [bib_ref] Implications of uterine NK cells and regulatory T cells in the endometrium..., Kofod [/bib_ref]. The architecture of the endometrium depends on the phase of menstrual cycle, status of pregnancy, or pathology [bib_ref] Decidualization of the human endometrium, Okada [/bib_ref] regulated by multiple factors, including hormones, immune suppression molecules, or residing microbiota [bib_ref] Progesterone Receptor Regulation of Uterine Adaptation for Pregnancy, Wu [/bib_ref] [bib_ref] Localisation of Indoleamine 2,3-dioxygenase and Kynurenine Hydroxylase in the Human Placenta and..., Ligam [/bib_ref]. Our knowledge on the significance of different components in endometrial pathology, including benign and malignant abnormalities, is still far from complete and awaits further studies using adequate models to mimic appropriate endometrial phenotype. Learning about the expression and relationship of important factors modulating immune status in human endometrium brings novel findings and help in the development of effective therapies.
The microenvironment of the endometrium is rich in mucins expressed on the surface of epithelial cells. These transmembrane glycoproteins, characteristic for endometrium, protect from pathogens, providing a barrier against environmental challenges. The expression patterns of different types of mucins have been extensively studied in human endometrium in order to verify their role in pathologies like infertility, endometriosis, and cancer [bib_ref] MUC4 gene polymorphism and expression in women with implantation failure, Koscinski [/bib_ref] [bib_ref] MUC1 as a Discriminator between Endometrium from Fertile and Infertile Patients with..., Margarit [/bib_ref] [bib_ref] Expression of the transmembrane mucins, MUC1, MUC4 and MUC16, in normal endometrium..., Dharmaraj [/bib_ref]. MUC1 was associated with the increased proliferation [bib_ref] MUC1 stimulates EGFR expression and function in endometrial cancer, Engel [/bib_ref] and metastatic potential [bib_ref] Biosynthesis of MUC1 mucin in human endometrial adenocarcinoma is modulated by estradiol..., Paszkiewicz-Gadek [/bib_ref] of cancer cells. Knowledge about changes in mucin expression patterns between normal and pathological states is still incomplete and needs further investigation [bib_ref] Expression of the transmembrane mucins, MUC1, MUC4 and MUC16, in normal endometrium..., Dharmaraj [/bib_ref].
The role of HER2 in different cancers has been well established [bib_ref] HER2 Expression Beyond Breast Cancer: Therapeutic Implications for Gynecologic Malignancies, English [/bib_ref]. It is a heterodimeric receptor activated upon binding with several ligands resulting in MAPK, PI3K, and PKC pathways activation [bib_ref] Human Epidermal Growth Factor Receptor 2 (HER2) in Cancers: Overexpression and Therapeutic..., Iqbal [/bib_ref]. HER2 overexpression is linked to poor prognosis in endometrial serous carcinoma patients [bib_ref] Amplification of c-erbB2 oncogene, Santin [/bib_ref]. Reports indicate that HER2 is an important cellular factor worth investigating as a therapeutic target for endometrial cancer [bib_ref] Targeting HER2 in ovarian and uterine cancers: Challenges and future directions, Teplinsky [/bib_ref]. Advances in targeted cancer therapy and successful treatments using anti-HER2 antibodies, i.e., trastuzumab, bring more interest in this potent cellular target also in the context of endometrium [bib_ref] HER2 aberrations in cancer: Implications for therapy, Yan [/bib_ref].
The receptor for advanced glycation end-products (RAGE) and indoleamine 2,3dioxygenase 1 (IDO1) have been shown to contribute to regulation of immune status of endometrium are currently considered among new molecules of clinical relevance [bib_ref] Antibody drug conjugates against the receptor for advanced glycation end products (RAGE),..., Healey [/bib_ref] [bib_ref] Indoleamine 2,3-dioxygenase in endometrial cancer: A targetable mechanism of immune resistance in..., Mills [/bib_ref].
Soluble factors secreted by immune cells, with a prominent role of INF-γ, enhance IDO1 expression [bib_ref] Kynurenine Pathway of Tryptophan Metabolism: Regulatory and Functional Aspects, Badawy [/bib_ref]. Activation of the kynurenine pathway (KP), responsible for the generation of several toxic metabolites called kynurenines, modulates local immune tolerance [bib_ref] The role of decidual immune cells on human pregnancy, Liu [/bib_ref] [bib_ref] Crosstalk between decidual NK and CD14+ myelomonocytic cells results in induction of..., Vacca [/bib_ref]. IDO1 activity is of great interest for scientists studying biology and pathology of endometrium [bib_ref] Indoleamine 2,3-dioxygenase suppresses the cytotoxicity of 1 NK cells in response to..., Liu [/bib_ref]. Recently, KP has been also recognized as a novel factor in endometrial cancer tissue. IDO expression and KP activation might be prognostic factors correlating with negative clinical outcome [bib_ref] Indoleamine 2,3-dioxygenase is a novel prognostic indicator for endometrial cancer, Ino [/bib_ref].
Contribution of surface receptor RAGE signaling is an additional component contributing to endometrial microenvironment [bib_ref] The immunobiology of the receptor of advanced glycation end-products: Trends and challenges, González [/bib_ref] [bib_ref] Immunohistochemical Expression of RAGE and Its Ligand (S100A9) in Cervical Lesions, Zhu [/bib_ref] [bib_ref] Effects of receptor for advanced glycation endproducts on microvessel formation in endometrial..., Zheng [/bib_ref]. This is a transmembrane receptor expressed on several cell types, including epithelial, endothelial, and immune cells activating migration, proliferation, and inflammatory response [bib_ref] Immunohistochemical Expression of RAGE and Its Ligand (S100A9) in Cervical Lesions, Zhu [/bib_ref] [bib_ref] Receptor for Advanced Glycation End Products (RAGE) Prevents Endothelial Cell Membrane Resealing..., Xiong [/bib_ref] [bib_ref] Induction of endothelial RAGE ex-pression in pterygium, Al-Swailem [/bib_ref] [bib_ref] Advanced Glycation End Products Subspecies-Selectively Induce Adhesion Molecule Expression and Cytokine Production..., Takahashi [/bib_ref] [bib_ref] Tumor Necrosis Factor-α Blocks Differentiation and Enhances Suppressive Activity of Immature Myeloid..., Sade-Feldman [/bib_ref]. Several ligands activate RAGE, i.e., advanced glycation end-products (AGEs) or high-mobility group box 1 (HMGB1) and contribute to cancer progression [bib_ref] RAGEs and s-RAGE; friend or foe for cancer, Ahmad [/bib_ref] [bib_ref] RAGE Mediates S100A7-Induced Breast Cancer Growth and Metastasis by Modulating the Tumor..., Nasser [/bib_ref] [bib_ref] HMGB1 and RAGE in Inflammation and Cancer, Sims [/bib_ref]. Interestingly, RAGE is also present on endometrial cells and is upregulated in some endometrial cancer cells in vivo as well as in culture [bib_ref] Effects of receptor for advanced glycation endproducts on microvessel formation in endometrial..., Zheng [/bib_ref]. In that paper, knockdown of the receptor present on the endometrial cancer cells xenografted into mice resulted in decreased proliferation and vessel formation in vivo. Metabolic abnormalities in the case of insulin resistance (IR) and obesity have been shown to enhance negative effects of RAGE-mediated inflammation and oxidative stress [bib_ref] Diamanti-Kandarakis, E. Implications and Future Perspectives of AGEs in PCOS Pathophysiology, Merhi [/bib_ref]. Finally, RAGE is considered as a therapeutic target for the development of novel targeted drugs [bib_ref] Antibody drug conjugates against the receptor for advanced glycation end products (RAGE),..., Healey [/bib_ref].
Identification and verification of the applicability of therapeutic targets in endometrial cancer require using adequate research models that in the case of human endometrium are scarce. Some examples of the xenotransplant models of human endometrial cells into mice have been described, but they present multiple challenges. The simpler models of the in vitro cultured cells derived from endometrium are adopted for this purpose [bib_ref] Culture characters, genetic background, estrogen/progesterone receptor expression, and tumorigenic activities of frequently..., Qu [/bib_ref]. Most of them are monocultures of one cell type that do not reflect the holistically complicated microenvironment of endometrium. Growing a monoculture of one cell type is also not trivial due to slow proliferation, thus modifications of the approach are continuously being explored. The most widely used endometrial cell lines are Ishikawa, HEC-1A, KLE (adenocarcinoma cells) and T-HESC (immortalized stromal cells) available from European Collection of Authenticated Cell Cultures (ECACC) and
American Tissue Culture Collection (ATCC). Good alternatives for human are animal cells, like bovine monocultures of epithelial endometrial cells [bib_ref] Rifaximin anti-inflammatory activity on bovine endometrium primary cell cultures: A preliminary study, Flammini [/bib_ref]. For the obvious reason, this is not an ideal model for studying human endometrial biology. Recently explored there are 3D human cell culture models called organoid culture [bib_ref] Self-renewing endometrial epithelial organoids of the human uterus, Fitzgerald [/bib_ref] [bib_ref] The Use of Endometrial Cancer Patient-Derived Organoid Culture for Drug Sensitivity Testing..., Girda [/bib_ref] that might be also a promising set up for studying endometrial abnormalities or testing drugs. Some models contain several cell types to generate an interactive model supporting communication between different cell types. However, this model has some limitations due to the great heterogeneity of individual organoids that present different growth progress and require sophisticated microscopic techniques to be evaluated. In addition, using a medium without serum might favor expanding only some cell fraction and does not necessarily reflect in vivo tumor composition.
The aim of our study was to identify the possibility of studying the relevance of immune modulating factors within endometrial microenvironment in the patient-derived primary endometrial cells cultured in vitro. We focus especially on the selected aspects that contribute to endometrial cancer biology, such as IDO1 activity (expressed as L-Kyn level) and RAGE implicated with inflammatory and metabolic status. Expression of the clinically relevant cancer markers such as HER2 and MUC1 was also assessed in the context of the presented model of endometrial cancer. Our data, especially on the enhanced activity of kynurenine pathway, bring new insights into molecular events in endometrial cancer.
# Materials and methods
## Patient characteristic
A total of 22 tissue samples [fig_ref] Table 1: Characteristics of the study group [/fig_ref] were obtained from women patients age 30-86 years, recommended to the surgical intervention due to a variety of endometrial abnormalities (benign and malignant), who were admitted to the Chair and Department of Gynecology and Gynecological Endocrinology, Medical University of Lublin, Poland. The used methodologies conformed to the standards set by the Declaration of Helsinki and the protocol was approved by the Bioethical Committee of the Medical University of Lublin. Written informed consent was obtained from each patient prior to the tissue collection. The study included patients referred for diagnostic curettage of the uterine cavity due to abnormal uterine bleeding with suspected endometrial cancer. Previous medical history indicated that some patients suffered from diabetes or hyperglycemia [fig_ref] Table 1: Characteristics of the study group [/fig_ref].
After curettage of the uterine cavity, the tissue material subjected to immunohistochemistry was fixed in formalin, apart from a small piece further processed for in vitro cell culturing. Based on the histopathological results, patients were divided into a cancer group (10 patients) with endometrial adenocarcinoma type I and noncancerous group (12 patients) with non-neoplastic changes in the endometrium, i.e., straight hypertrophy of the endometrium, chronic endometrial inflammation, different degree of dysplasia, and endometrial polyp. One sample within the noncancerous group was from the benign change of endometrium (polyps) but there was metaplasia in the cervix identified at the same time (E9-17). There was also a case with mixed type endometrial changes, i.e., cancer cells and dysplasia (E4-11). In all patients with diagnosis of endometrial glandular cancer, attention was drawn to additional tests, such as fasting hyperglycemia, obesity-related history, menstrual irregularities (polycystic ovarian syndrome in the procreative age), and insulin resistance. In this respect, all studied patients showed classic risk factors for the development of endometrial cancer. Patients within the noncancerous group (without negative histopathology result in endometrial cancer) showed no history of the menstrual disorder, insulin resistance, and diabetes, with the exception of one woman with postmenopausal diabetes [fig_ref] Table 1: Characteristics of the study group [/fig_ref] , biopsy E1-6).
## Immunohistochemistry
In addition to the diagnostic histopathology performed at the clinic, a portion of some specimens collected for the study was subjected to the immunohistochemistry (IHC) staining to confirm the status of the selected markers, i.e., CK8/18, HER2, RAGE, and IDO1 in the obtained specimen. The piece of the collected tissue was fixed in 4% formaldehyde in PBS (Phosphate-Buffered Saline) at the time of surgical removal and was embedded in paraffin (FFPE) for cutting into 4 µm sections. The slices mounted on poly-L-lysine-coated glass slides were first deparaffinized by heating at 60 - C and then immersed in xylene for 9 min, followed by antigen retrieval in DAKO Retrieval Solution pH 9.0 by heating in microwave Daewoo at 350 Watt and leaving at room temperature. The sections were next immunostained utilizing the ABC DAKO kit (Glostrup, Denmark). The endogenous peroxidase was first blocked with the Peroxidase-Blocking Solution (DAKO, Glostrup, Denmark) for 10 min at room temperature and washed twice with distilled water. After 15 min of blocking with Protein Block Serum-Free (DAKO, Glostrup, Denmark) at room temperature and washing with distilled water, the sections were incubated with one of the following antibodies: Cytokeratin 8/18 (CK8/18, clone EP1, DAKO, Glostrup, Denmark, ready to use), HER2 (DAKO, Glostrup, Denmark, Hercept Test), IDO1 (Merck, Darmstadt, Germany, diluted 1:80), RAGE (Merck, Darmstadt, Germany diluted 1:80) over night at 4 - C. The sections washed with PBS were next subjected to the 30 min incubation with DAKO, Glostrup, Denmark, EnVision + System-HRP at room temperature and after another PBS washing (2 × 5 min) the reaction was developed with 3,3 -diaminobenzidine tetrahydrochloride (DAB, DAKO, Glostrup, Denmark,) for 5 min. Finally, the slides were counterstained with hematoxylin-eosin (HE) and mounted under coverslips. The slides were observed under the Olympus BX51 microscope and pictures were recorded with the camera. The positive or negative reactivity was determined for each marker. IHC results of the collected endometrial tissue samples are presented in [fig_ref] Table 2: Immunochemistry [/fig_ref]. na-samples not analyzed; + indicates tissue with positive IHC reactivity; − indicates tissue with negative IHC reactivity; L-Kyn was quantified in the cell-culturing medium by HPLC-DAD; <LOQ-under limit of quantification of the applied HPLC-DAD method.
## Isolation and culture of cells from human endometrium
The tissue subjected for cell culture immediately after removal was placed into a sterile container filled with Dulbecco's Modified Eagle Medium/Nutrient Mixture F12 (DMEM/F12), supplemented with penicillin/streptomycin (pen/strep, Sigma, Saint Louis, MO, USA) and amphotericin (Sigma, Saint Louis, MO, USA), and was transported on ice for in vitro cell culturing. The tissue (pea size) was next placed on a plastic dish and cut into 1 mm pieces. The minced sample was collected with 5-10 mL of PBS containing pen/strep and amphotericin and transferred into a 15 mL conical tube for centrifugation at 200× g for 5 min. The resulting pellet was reconstituted in 2 mL of digestion solution containing 1 mg/mL of Collagenase I (Sigma, Saint Louis, MO, USA) in DMEM/F12 and incubated for 2-2.5 h at 37 - C to dissociate cells. The reaction was stopped by the addition of the complete culturing medium and cells were spun down at 200× g for 5 min. The cell pellet reconstituted in 0.5 mL of culturing medium was plated on the 6-well plate. When needed, sample of the endometrial tissue was treated with erythrocyte lysis buffer ELB (8.3 g ammonium chloride, 20 mg EDTA, 1 g sodium bicarbonate in 1 L) for 5-10 min at 37 - C. Finally, the cell pellet after centrifugation at 200× g for 5 min was reconstituted in 0.5 mL/well of culturing medium consisting of DMEM/F12 (PAN Biotech, Germany), 5% fetal bovine serum (FBS), 1 mg/mL bovine serum albumin (BSA), 10 ng/mL cholera toxin, 0.5 µg/mL hydrocortisone, 5 µg/mL insulin, 5 ng/mL epidermal growth factor (EGF), 5 µg/mL apo-transferrin (all from Sigma, Saint Louis, MO, USA), without antibiotics. The prepared culturing medium was used within 4 weeks. The plated tissue was incubated at 37 - C with constant 5% CO 2 and routinely observed under the light microscope. The outgrown cells were passaged when reached 70% confluency and tested at passage 1-8.
The human ovarian cancer cell line SK-OV-3 from American Type Culture Collection (ATCC) was cultured at 37 - C with 5% CO 2 in DMEM containing 4.5 g/L of D-glucose, supplemented with 10% (v/v) fetal bovine serum (FBS), 2 mmol/L L-glutamine and 1 U penicillin/streptomycin.
## Determination of population doubling time (pdt)
The cells were plated on 6-well plates at density 50,000 cells/well on individual plates that were subjected to cell counting on the particular day 1-10. The cells were collected from 3 wells by trypsinization for each time point and were counted manually. The Population Doubling Time (PDT) was calculated from the time frame of the cell linear growth using Equation (1):
[formula] PDT = T × ln2 ln Xe Xb (1) [/formula]
where T-hours between beginning and end point; Xe-number of cells at end time; Xb-number of cells at the beginning time.
## Karyotyping
The cells were subjected to hypotonic shock by 20 min incubation with 0.075 mol/L KCl (Sigma, Saint Louis, MO, USA) at 37 - C. Then, cells were fixed with a cold mixture of methanol/acetic acid (3:1, v/v), sprayed on the microscope slides and left to dry on air. The RHG (R-bands by heating using Giemsa reagent) banding method, according to Sehested, has been employed for chromosome staining. Briefly, the slides were incubated with 1 mol/L PBS for 3-10 min at 88 - C and washed with ultrapure water. The Giemsa (AQUA-MED, Łódź, Poland) solution (5%) was applied to obtain R band staining. The GTG pattern (G-bands by trypsin using Giemsa was obtained by incubation with 0.25% trypsin solution in PBS (BIOMED, Lublin, Poland) for 2-3 min. After washing with PBS, the slides were incubated with 5% Giemsa. The chromosome pattern was analyzed under Nikon ECLIPSE-Ni light microscope. Data collected from 15-25 slides for each cell line were analyzed using Applied Spectral Imaging software.
## Immunofluorescent staining
The cells were cultured on chamber slides and after washing with PBS were fixed for 15 min at room temperature with 4% paraformaldehyde diluted in PBS. After 3 washing cycles with PBS, cells were permeabilized with 0.5% Triton X-100 for 15 min at room temperature and the chamber slides were incubated with a blocking solution (3% non-fat dry milk (NFDM), 1% BSA, 0.1% Triton X-100 in PBS) for 30 min at room temperature. Next, primary antibody against RAGE (Merck, Darmstadt, Germany, diluted 1: were applied in the blocking solution, and chamber slides were incubated overnight at 4 - C. After washing (3 times) with PBS, the chamber slides were incubated for 1 h, at room temperature, with secondary Ab anti-rabbit IgG-AF555 or anti-mouse IgG-AF488 (ThermoFisher, Rockfold, IL, USA) for rabbit or mouse primary Ab, respectively. The washed 3 times with PBS chamber slides were next mounted with the mounting medium containing DAPI for nuclear staining. The slides were observed under Zeiss Axio Observer, LSM700 confocal microscope with 20× magnification or under the Nikon Eclipse Ti-2U Fluorescent microscope. The CK8-and vimentin-positive cells were counted from 3 fields as the % of total DAPI-stained nuclei.
## Western blotting
The cells were lysed with RIPA buffer containing protease inhibitor cocktail (Sigma, Saint Louis, MO, USA) and 100 nmol/L PMSF (Sigma, Saint Louis, MO, USA) followed by sonication. The lysate was loaded on 12% polyacrylamide gel and resolved in the presence of SDS. After protein transfer onto PVDF, the membrane was blocked with 5% NFDM/PBS containing 0.5% Tween-20 (PBS-T) for 1 h, at room temperature. The membrane was further left overnight at 4 - C with a mixture of rabbit anti-HER2 Ab (Abcam, Cambridge, UK 1: 500) and rabbit anti-β-actin Ab (Elabscience, Huston, TX, USA, 1:1000) diluted in 5% NFDM/PBS-T. Following 3 washing cycles with PBS-T, the membrane was incubated with a mixture of secondary Ab, i.e., anti-mouse IgG-AP and anti-rabbit IgG-AP, to detect mouse and rabbit Ab, respectively. The reaction was developed by incubation in a standard Alkaline Phosphatase substrate solution (Pierce Biotechnology, Rockford, IL, USA). The calibration curve prepared in DMEM/12 was used to determine L-Kyn concentration. Matrix-matched calibration solutions (prepared analogously as experimental culture medium samples) contained fixed amounts of 3NT and variable concentrations of L-Kyn. Standards of crystalline L-Kyn and 3NT were obtained from Sigma-Aldrich (Saint Louis, MO, USA). Stock solutions of L-Kyn (1 mg/mL) and 3NT (1 mg/mL) were prepared daily by dissolving appropriate amounts of the substances in dimethyl sulfoxide (DMSO, Merck, Germany) or PBS (pH 7.7), respectively. A working solution of L-Kyn was prepared in DMSO. The blank response (L-Kyn in DMEM/F12) was subtracted from each sample to calculate L-Kyn concentration in the medium collected from cells. The limit of detection (LOD) and quantification (LOQ) estimated for L-Kyn in DMEM/F12 matrix were 0.074 and 0.24 nmol/mL, respectively.
## Quantification of l-kyn generated by endometrial cells cultured in vitro
# Statistical analysis
The Mann-Whitney U test was used to evaluate difference in concentration of L-Kyn present in culture medium of the in vitro grown cells isolated from cancer and the noncancerous endometrium. For calculations, the results above LOQ were considered.
Data were processed using the PQStat software. Differences were considered significant at p < 0.05.
# Results
## Development of the experimental in vitro model
We have successfully developed and secured for further studies several cultures of human endometrial cells obtained from patients with cancer (10 cultures) and noncancerous endometrium (12 cultures). The success rate of cell culture was 90% and 69% in the cancer and the noncancerous group, respectively.
The limited amount of tissue mass (often large sample is unavailable) was an important determinant of the employed method of cell isolation. We started the cultures from a pea-sized sample of endometrial tissue. Such a sample without separation naturally contains different cell types present in the endometrial environment. Such an approach allows for retaining a composition of factors and cellular interactions which is natural for the organ and unique for the individual patient. To avoid fibroblast overgrowth, the medium components have been optimized to favor the growth of the epithelial cells. The complete culture medium contained DMEM/F12 with lower (5%) concentration of FBS was supplemented with BSA, cholera toxin (growth stimulator of epithelioid cells), insulin, apo-transferrin, hydrocortisone, and EGF. Additionally, in the first step of cell isolation, the tissue samples were minced and collagenase-treated for better cell disintegration and outgrowth. The endometrial samples that contained a substantial amount of blood were treated with ELB to deplete residual erythrocytes. Noteworthy, in order to assure better tissue adherence to the plate surface, the amount of medium at the beginning of the culture was kept at minimum. This minimized lifting-up of the pieces of tissue while transferring the plate and daily examination under the microscope. The first cell outgrowth was usually noticed about 2-5 days post plating. It was observed that the cells obtained from cancer tissue were easier to expand and resulted in higher number of passages as well as number of frozen vials, in contrast to the noncancerous cells [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref]. Five out of twenty-two biopsies initially attempted to culture (23%) were not able to further expand in vitro, and eventually ceased. The cultures were carried through several passages, some even reached 15 passages, although the kinetics of growth varied [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref] and calculated Population Doubling Time (PDT) was from 28 to 67 h, when tested at passage around p3-p7.
The morphology of the growing in vitro cells was evaluated under the light microscope [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref]. Appearance of the obtained cell cultures varied from typical epithelial morphology, cells with irregular shape, or mixture-containing cells with elongated morphology characteristic for fibroblasts. We have noticed that lowering serum concentration to 5% resulted in a less intense proliferation of fibroblast-like cells allowing for expanding other cell types present in endometrial tissue. Some of the isolates at the lower passages had a tendency to form the regular shape 3D structures similar to organoids reported elsewhere [bib_ref] Self-renewing endometrial epithelial organoids of the human uterus, Fitzgerald [/bib_ref] that contained highly packed cellular mass [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref]. These cellular aggregates were further dissociated during cell passaging and eventually the culture became a monolayer [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref] , lower panel E1-6). The morphology of the growing in vitro cells was evaluated under the light microscope [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref]. Appearance of the obtained cell cultures varied from typical epithelial morphology, cells with irregular shape, or mixture-containing cells with elongated morphology characteristic for fibroblasts. We have noticed that lowering serum concentration to 5% resulted in a less intense proliferation of fibroblast-like cells allowing for expanding other cell types present in endometrial tissue. Some of the isolates at the lower passages had a tendency to form the regular shape 3D structures similar to organoids reported elsewhere [bib_ref] Self-renewing endometrial epithelial organoids of the human uterus, Fitzgerald [/bib_ref] that contained highly packed cellular mass [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref] , arrow). These cellular aggregates were further dissociated during cell passaging and eventually the culture became a monolayer [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref] , lower panel E1-6). The karyotype of selected cultures isolated form noncancerous (E1-6, E22-38) and cancer (E10-18, E12-23) tissue have also been analyzed by G-and R-band staining [bib_ref] A simple method for R banding of human chromosomes, showing a pH-dependent..., Sehested [/bib_ref]. In noncancer, as well as cancer isolates, we have reported normal number of chromosomes without aberrations as expected for women's phenotype [fig_ref] Figure 3: Karyotype characterization of cancer cells isolated from patients and cultured in vitro... [/fig_ref]. The karyotype of selected cultures isolated form noncancerous (E1-6, E22-38) and cancer (E10-18, E12-23) tissue have also been analyzed by G-and R-band staining [bib_ref] A simple method for R banding of human chromosomes, showing a pH-dependent..., Sehested [/bib_ref]. In noncancer, as well as cancer isolates, we have reported normal number of chromosomes without aberrations as expected for women's phenotype [fig_ref] Figure 3: Karyotype characterization of cancer cells isolated from patients and cultured in vitro... [/fig_ref]. To characterize the obtained cultures, we performed immunofluorescent staining for Cytokeratin 8 (CK8) that is abundantly present in epithelial cells and overexpressed on several types of cancers and in vitro cultured epithelial cells [bib_ref] Developing in vitro models of human ductal carcinoma in situ from primary..., Brown [/bib_ref]. The cells cultured on chamber slides were fixed and stained with anti-CK8 antibody. Our results indicated that CK8 was expressed in most of the cells present in primary cultures isolated from cancer and noncancerous endometrial tissue [fig_ref] Figure 4: Expression of CK8 and MUC1 in the cultured in vitro endometrial cells [/fig_ref] and [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref]. It was with the agreement for the IHC staining performed on the corresponding tissue specimens preserved for this purpose [fig_ref] Table 2: Immunochemistry [/fig_ref]. All the tested tissue samples showed positive reactivity with anti-CK8/18 antibody. To characterize the obtained cultures, we performed immunofluorescent staining for Cytokeratin 8 (CK8) that is abundantly present in epithelial cells and overexpressed on several types of cancers and in vitro cultured epithelial cells [bib_ref] Developing in vitro models of human ductal carcinoma in situ from primary..., Brown [/bib_ref]. The cells cultured on chamber slides were fixed and stained with anti-CK8 antibody. Our results indicated that CK8 was expressed in most of the cells present in primary cultures isolated from cancer and noncancerous endometrial tissue [fig_ref] Figure 4: Expression of CK8 and MUC1 in the cultured in vitro endometrial cells [/fig_ref] and [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref]. It was with the agreement for the IHC staining performed on the corresponding tissue specimens preserved for this purpose [fig_ref] Table 2: Immunochemistry [/fig_ref]. All the tested tissue samples showed positive reactivity with anti-CK8/18 antibody. Staining with anti-vimentin antibody [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref] that indicates stroma cells showed 5-47.3% of positive cells depending on the culture. Some cells presented intense staining [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref] , while there were cells with weak or no vimentin expression [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref] , arrowheads and stars, respectively). Staining with anti-vimentin antibody [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref] that indicates stroma cells showed 5-47.3% of positive cells depending on the culture. Some cells presented intense staining [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref] , while there were cells with weak or no vimentin expression [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref] , arrowheads and stars, respectively).
Finally, to assess contribution of immune cells, we used the anti-CD45 antibody to identify leukocytes [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref]. However, there were only few CD45+ cells present with a negligible number in cancer (E10-18, E12-23) and noncancerous cultures (E17-33).
We also observed a positive staining with anti-MUC1 antibody of the in vitro cultured endometrial cells, from the cancer and noncancerous group [fig_ref] Figure 4: Expression of CK8 and MUC1 in the cultured in vitro endometrial cells [/fig_ref]. However, the expression of this sialoprotein was culture-dependent, with variable intensity and homogeneity of staining.
## Her2 and rage status of the endometrial cells cultured in vitro
The expression of HER2 in the endometrial primary cell cultures was assessed by WB as the band of~185 kDa. We have identified a relatively low amount of HER2 in endometrial cell cultures [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref] , as compared to the commercial cell line (SK-OV-3) of human ovarian cancer known for abundant expression of this receptor [fig_ref] Figure 5: Expression of RAGE and HER2 on endometrial cells cultured in vitro [/fig_ref] , lane 1). In the primary cultures, there was variable expression of HER2 and mostly a rather weak band of HER2 was seen, even in the culture derived from HER2-positive biopsy [fig_ref] Figure 5: Expression of RAGE and HER2 on endometrial cells cultured in vitro [/fig_ref] , lane 2, E12-23). Moderate expression was observed in one culture (E10-18, lane 3) that was derived from HER2 negative biopsy. However, this specimen tested in IHC with the anti-HER2 antibody was negative [fig_ref] Table 2: Immunochemistry [/fig_ref]. Importantly, there was no HER2 expression in cells derived from the control benign tissue [fig_ref] Figure 5: Expression of RAGE and HER2 on endometrial cells cultured in vitro [/fig_ref] , lane [bib_ref] MUC4 gene polymorphism and expression in women with implantation failure, Koscinski [/bib_ref] [bib_ref] Expression of the transmembrane mucins, MUC1, MUC4 and MUC16, in normal endometrium..., Dharmaraj [/bib_ref] [bib_ref] MUC1 stimulates EGFR expression and function in endometrial cancer, Engel [/bib_ref]. The specimen E22-38 [fig_ref] Figure 5: Expression of RAGE and HER2 on endometrial cells cultured in vitro [/fig_ref] , lane 9) was also negative in IHC staining with anti-HER2 antibody [fig_ref] Table 1: Characteristics of the study group [/fig_ref]. Next, we aimed to determine the RAGE status of the obtained endometrial cell cultures. As demonstrated in [fig_ref] Figure 5: Expression of RAGE and HER2 on endometrial cells cultured in vitro [/fig_ref] , there was RAGE present in most tested lysates of the primary cultures (lane 3-9) and in the SK-OV-3 cells (lane 1). It was in agreement with the IHC staining for the corresponding tissues, i.e., E13-24, E24-41, E9-17 [fig_ref] Table 2: Immunochemistry [/fig_ref] , sample 5, 6, 15, respectively). There were some cell cultures (E10-18, E22-38) that showed RAGE expression in WB [fig_ref] Figure 3: Karyotype characterization of cancer cells isolated from patients and cultured in vitro... [/fig_ref] despite being derived from RAGE-negative tissue [fig_ref] Table 1: Characteristics of the study group [/fig_ref]. Interestingly, the lowest RAGE expression in WB was observed in the lysate of the cell culture derived from the HER2-positive biopsy [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref]. RAGE expression in the primary cell cultures derived from the cancer and the non- Next, we aimed to determine the RAGE status of the obtained endometrial cell cultures. As demonstrated in [fig_ref] Figure 5: Expression of RAGE and HER2 on endometrial cells cultured in vitro [/fig_ref] , there was RAGE present in most tested lysates of the primary cultures (lane [bib_ref] Progesterone Receptor Regulation of Uterine Adaptation for Pregnancy, Wu [/bib_ref] [bib_ref] Localisation of Indoleamine 2,3-dioxygenase and Kynurenine Hydroxylase in the Human Placenta and..., Ligam [/bib_ref] [bib_ref] MUC4 gene polymorphism and expression in women with implantation failure, Koscinski [/bib_ref] [bib_ref] MUC1 as a Discriminator between Endometrium from Fertile and Infertile Patients with..., Margarit [/bib_ref] [bib_ref] Expression of the transmembrane mucins, MUC1, MUC4 and MUC16, in normal endometrium..., Dharmaraj [/bib_ref] [bib_ref] MUC1 stimulates EGFR expression and function in endometrial cancer, Engel [/bib_ref] and in the SK-OV-3 cells (lane 1). It was in agreement with the IHC staining for the corresponding tissues, i.e., E13-24, E24-41, E9-17 [fig_ref] Table 2: Immunochemistry [/fig_ref] , sample 5, 6, 15, respectively). There were some cell cultures (E10-18, E22-38) that showed RAGE expression in WB [fig_ref] Figure 3: Karyotype characterization of cancer cells isolated from patients and cultured in vitro... [/fig_ref] despite being derived from RAGE-negative tissue [fig_ref] Table 1: Characteristics of the study group [/fig_ref]. Interestingly, the lowest RAGE expression in WB was observed in the lysate of the cell culture derived from the HER2-positive biopsy [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref].
RAGE expression in the primary cell cultures derived from the cancer and the noncancerous tissue was also observed by immunofluorescence after staining of the fixed cells with the same antibody that recognizes the extracellular RAGE protein domain [fig_ref] Figure 5: Expression of RAGE and HER2 on endometrial cells cultured in vitro [/fig_ref].
## Ido1 expression and l-kynurenine generation by the primary endometrial cell cultures
The endometrial cells derived from cancer [fig_ref] Figure 6: IDO expression and activity in endometrial cells cultured in vitro [/fig_ref] , upper panel) and noncancerous [fig_ref] Figure 6: IDO expression and activity in endometrial cells cultured in vitro [/fig_ref] , lower panel) tissue were fixed on chamber slides and stained with anti-IDO1 antibody to observe protein expression. Most of the cells present in the tested cultures were found positive for IDO1 in IF, despite being derived from cancer or noncancerous tissue [fig_ref] Figure 6: IDO expression and activity in endometrial cells cultured in vitro [/fig_ref]. Interestingly, the positive IF result for cultured cells did not corelate with IHC staining of the original tissue. We observed no IDO1 staining of the original tissue available for IHC [fig_ref] Table 2: Immunochemistry [/fig_ref] , sample 4, 6, 21) although the corresponding derived cells displayed IDO1 expression [fig_ref] Figure 6: IDO expression and activity in endometrial cells cultured in vitro [/fig_ref] , E10-18, E13-24, E22-38). However, the enzymatic activity of the IDO1 enzyme measured indirectly as the level of L-Kyn production showed a difference between the cancer and noncancerous cells. Using the HPLC-DAD approach, we analyzed L-Kyn level in conditioned medium from 6 cancer and 9 noncancerous endometrial cultures [fig_ref] Table 2: Immunochemistry [/fig_ref]. Two cell samples, E21-37 and E14-25, contained L-Kyn below level of quantification (LOQ) of the applied HPLC-DAD method and were not included in statistical analysis. Comparison using the Mann-Whitney U test showed that cultures derived from cancer tissue secrete significantly greater amounts of L-Kyn (p < 0.05) than cells derived from the noncancerous endometrial tissue [fig_ref] Figure 6: IDO expression and activity in endometrial cells cultured in vitro [/fig_ref]. Higher L-Kyn secretion suggests more active IDO1 enzyme in the cancer-derived cells. It was also interesting that cells derived from the IDO1-negative biopsy secreted L-Kyn into the medium (E21-37, E11-21, E22-38), suggesting that the IHC staining of the tissue might not completely reflects the phenotype of cells that are present in tumor. This observation highlights the importance of analytical approach in determination of the IDO1 activity in cancer tissue.
# Discussion
Immunotherapy presents a promise for more effective anti-cancer treatment however, regulation of immune mechanism in cancer microenvironment is still not
# Discussion
Immunotherapy presents a promise for more effective anti-cancer treatment [bib_ref] Targeting indoleamine-2,3-dioxygenase in cancer: Scientific rationale and clinical evidence, Ricciuti [/bib_ref] , however, regulation of immune mechanism in cancer microenvironment is still not fully known. IDO1 is involved in immune regulation and in endometrial cancer it is often associated with the check point protein PD-L1 [bib_ref] Indoleamine 2,3-dioxygenase in endometrial cancer: A targetable mechanism of immune resistance in..., Mills [/bib_ref] as determined by IHC staining. In our study, IDO1 staining was observed only in some cancer tissues and in some noncancer ones as well, which is in agreement with the previous reports [bib_ref] Indoleamine 2,3-dioxygenase in endometrial cancer: A targetable mechanism of immune resistance in..., Mills [/bib_ref] [bib_ref] Mapping the immunosuppressive environment in uterine tumors: Implications for immunotherapy, Vanderstraeten [/bib_ref]. Interestingly, in the in vitro cultures, derived from the endometrial tissue, the enzyme was active regardless of the IHC-negative IDO1 status of the original tissue [fig_ref] Table 2: Immunochemistry [/fig_ref] , as quantified by measuring of L-Kyn secreted to culturing medium. We recorded L-Kyn concentration to be higher in medium from the cancer group compared to the noncancer derived cultures [fig_ref] Figure 6: IDO expression and activity in endometrial cells cultured in vitro [/fig_ref] translating into the higher enzyme activity. We believe our paper is the first demonstration that in vitro cultured primary endometrial cells recapitulate IDO1 overexpression seen in tissue stained by IHC [fig_ref] Table 2: Immunochemistry [/fig_ref] , sample 3). Kynurenine secretion by cells derived from IHC-negative tissue for IDO1 suggests underestimation of IDO1 status in the original tissue. The primary cultures have the advantage over the established cancer cell lines and consist of a mixture of epithelial, stromal cells, and other residual cell types. A possible scenario is enhanced proliferation of the IDO1 expressing stromal cells [bib_ref] (IDO1) enhances survival and invasiveness of endometrial stromal cells via the activation..., Mei [/bib_ref] in vitro. Kynurenine might also be generated by other tryptophan degrading enzymes like IDO2 and TDO [bib_ref] Kynurenine Pathway of Tryptophan Metabolism: Regulatory and Functional Aspects, Badawy [/bib_ref]. The origin of L-Kyn generation by the endometrial cell cultures remains to be verified. However, analytical approaches of IDO1 determination, i.e., the used in this study HPLC-DAD might better reflect the status of IDO1 enzyme in the tumor and should be considered in final tissue assessment.
The presented endometrial cell cultures as demonstrated in [fig_ref] Figure 5: Expression of RAGE and HER2 on endometrial cells cultured in vitro [/fig_ref] showed expression of RAGE. The signaling initiated by RAGE ligands like advanced glycation end-products (AGEs), that accumulate in hyperglycemia and metabolic abnormalities, mediate oxidative stress and induce inflammation [bib_ref] The immunobiology of the receptor of advanced glycation end-products: Trends and challenges, González [/bib_ref] associated with endometriosis [bib_ref] Gene Expression Analysis of Endometrium Reveals Progesterone Resistance and Candidate Susceptibility Genes..., Burney [/bib_ref] [bib_ref] Implication of the RAGE-EN-RAGE axis in endometriosis, Sharma [/bib_ref] and endometrial cancer [bib_ref] Effects of receptor for advanced glycation endproducts on microvessel formation in endometrial..., Zheng [/bib_ref]. RAGE-initiated signaling is also considered as an immune modulating factor in cancer microenvironment, thus we tested RAGE expression in the cultured in vitro human endometrial cells. There was a difference in original tissue RAGE expression compared to the corresponding cell cultures. Most cultures expressed RAGE (determined by WB) although some tissues [fig_ref] Table 2: Immunochemistry [/fig_ref] , sample 4 21) remained negative for RAGE in IHC. Interestingly, the lowest amount of RAGE was seen for the cell culture derived from the HER2-positive tissue [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref]. This RAGE-HER2 inverse correlation was also observed in commercial SK-BR-3 breast cancer cell line known to overexpress HER2, while in the other breast cancer cell line MDA-MB-231 with low HER2 expression there was high RAGE (data not shown). The same low RAGE expression was seen in the tissue of the HER2-negative breast cancer subtype [bib_ref] Analysis of AGE modified proteins and RAGE expression in HER2/neu negative invasive..., Korwar [/bib_ref].
The presented in [fig_ref] Figure 5: Expression of RAGE and HER2 on endometrial cells cultured in vitro [/fig_ref] SK-OV-3 cell line showed expression of both receptors HER2 and RAGE. This implies that RAGE expression might be promoted in the in vitro cultured endometrial cells, unless there is low HER2 expression. This might be different for other cancers, such as ovarian cancer, where RAGE and HER2 coexist at high level (SK-OV-3 cells).
It would be also worth investigating in the future whether the negative IHC staining for RAGE observed in the studied inhere tumor tissue biopsies is associated with an increased level of AGEs that upon binding to RAGE can hamper tissue staining with anti-RAGE antibody. On the other hand, since hyperglycemia induces RAGE expression through ROS generation [bib_ref] Hyperglycemia-Induced Reactive Oxygen Species Increase Expression of the Receptor for Advanced Glycation..., Yao [/bib_ref] , it might be possible that in cancer patients with co-existing hyperglycemia (like in the present studies) there is an upregulation of RAGE in cancer cells that is further sustained during in vitro cell culture.
In vitro culture of the established cell lines, usually derived from cancer cells or immortalized primary cells [bib_ref] A tissue-engineered human endometrial stroma that responds to cues for secretory differentiation,..., Schutte [/bib_ref] , is an approach providing a constant source of material with minimal variability. However, it has several disadvantages over the tissue directly obtained from patient, i.e., it is usually monoculture [bib_ref] Culture characters, genetic background, estrogen/progesterone receptor expression, and tumorigenic activities of frequently..., Qu [/bib_ref] and hard to completely mimic in in vivo conditions. While the established cell lines are relatively easy and predictable to work with, they might be used to study only some aspects of the tumor environment, i.e., they lack several cellular components and factors supplied by other cell types present in the original cancer tissue.
Here we present the data on the in vitro cultured primary cells that can be used as models for studying the immune components of the endometrial environment. Using the presented approach, we were able to propagate not only cancer but also noncancerous endometrial cells, although the cancer cells had a higher success rate and greater expansion properties compared to the cells isolated from noncancerous tissue [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref]. This was supported by the differential staining with anti-CK8 and vimentin markers indicating epithelial and stromal cells, respectively. In our cultures, the majority of cells showed epithelial phenotype, although some relevant number of stromal cells was also present [fig_ref] Figure 1: Potential of cell expansion in the in vitro culture conditions [/fig_ref] , . Since we did not perform CK8/vimentin co-staining, it is difficult to say that the vimentin+ cells are CK8+; however, it might be assumed based on calculated percentage in both types of staining . It can be explained by cytoskeletal rearrangement in the rather dynamically behaving partial EMT cells undergoing epithelial-to-mesenchymal transition (EMT) [bib_ref] Molecular mechanisms of epithelial-mesenchymal transition, Lamouille [/bib_ref]. . Identified % of CK8+ and vimentin+ cells in cancer (E10-18, E12-23, E13-24) and noncancerous (E1-6, E17-33, E22-38) cells stained on chamber slides with anti-CK8 or anti-vimentin antibody and co-stained with DAPI. Importantly, we did not observe chromosomal changes in our primary cultured cells, indicating that the model system does not introduce major gene aberrations [fig_ref] Figure 3: Karyotype characterization of cancer cells isolated from patients and cultured in vitro... [/fig_ref] , as it is observed in the immortalized cell cultures [bib_ref] Culture characters, genetic background, estrogen/progesterone receptor expression, and tumorigenic activities of frequently..., Qu [/bib_ref] [bib_ref] Successful Immortalization of Endometrial Glandular Cells with Normal Structural and Functional Characteristics, Kyo [/bib_ref]. The genetic changes occurring during cell immortalization might negatively influence the experimental results and are often raising a question of the specificity of the observed effects [bib_ref] Sequential cytogenetic and molecular cytogenetic characterization of an SV40T-immortalized nasopharyngeal cell line..., Zhang [/bib_ref].
## Cell
The previously described markers important in endometrial receptivity [bib_ref] Decreased MUC1 in endometrium is an independent receptivity marker in recurrent implantation..., Wu [/bib_ref] and overexpressed in cancer metastasis [bib_ref] Biosynthesis of MUC1 mucin in human endometrial adenocarcinoma is modulated by estradiol..., Paszkiewicz-Gadek [/bib_ref] , such as MUC1, we confirmed to be abundant also in the endometrial cells cultured in vitro in our hands [fig_ref] Figure 4: Expression of CK8 and MUC1 in the cultured in vitro endometrial cells [/fig_ref]. The model cells also show expression of HER2 protein in most of the cancer cultures but not in cells derived from benign tissue. Surprisingly, the cells derived from the cancer tissue diagnosed as HER2+ also showed a low level of HER2 staining [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref] , suggesting that the number of cells overexpressing HER2 was not dominant in this isolate.
Furthermore, the generated number of cells was sufficient to characterize the obtained culture and perform multiple assays, including protein expression, analysis of factors involved in immune regulation (kynurenine), inflammatory pathways, and metabolic abnormalities (RAGE) [fig_ref] Figure 4: Expression of CK8 and MUC1 in the cultured in vitro endometrial cells [/fig_ref].
The cultured cells displayed heterogeneous and changing morphology allowing for the 3D structure formation that mimics the native endometrial microenvironment [fig_ref] Figure 2: Morphology of the in vitro cultured endometrial cells [/fig_ref]. This is important for studying endometrial cancer differentiation processes since the changes in cellular architecture mark the potency to metastasis with the involvement of multiple factors, i.e., TGF-β in epithelial-to-mesenchymal transition (EMT). The opposite effect is implicated also in the regeneration of endometrium [bib_ref] Mesenchymal-to-Epithelial Transition Contributes to Endometrial Regeneration Following Natural and Artificial Decidualization, Patterson [/bib_ref] or embryo implantation [bib_ref] MFGE8 regulates TGF-β-induced epithelial mesenchymal transition in endometrial epithelial cells in vitro, Yu [/bib_ref]. Thus, the presented in vitro model with no need for additional extracellular matrix components unlike other models [bib_ref] Successful Immortalization of Endometrial Glandular Cells with Normal Structural and Functional Characteristics, Kyo [/bib_ref] will be useful to study mechanisms governing immune response in the endometrial microenvironment.
Finally, in cells isolated from cancerous tissue, we have shown increased generation of L-Kyn [fig_ref] Figure 6: IDO expression and activity in endometrial cells cultured in vitro [/fig_ref] that is the first stable metabolite of the immune regulatory pathway mediated by IDO1. The enzyme expression has also been confirmed in the cultured cells [fig_ref] Figure 6: IDO expression and activity in endometrial cells cultured in vitro [/fig_ref] , likewise expression of HER2, another cancer marker associated with 30% of endometrial cancers [bib_ref] Endometrial cancer: Molecular markers and management of advancedstage disease, Arenda [/bib_ref]. This is an important observation indicating that the cultured cells retain malignant phenotype with high IDO1 [bib_ref] Indoleamine 2,3-dioxygenase is a novel prognostic indicator for endometrial cancer, Ino [/bib_ref] [bib_ref] Targeting indoleamine-2,3-dioxygenase in cancer: Scientific rationale and clinical evidence, Ricciuti [/bib_ref] and HER2 expression observed in vivo [bib_ref] HER2 expression patterns in paired primary and metastatic endometrial cancer lesions, Halle [/bib_ref]. Together, it makes this model suitable for identification and testing novel immunotherapeutics for HER2-and IDO1-associated pathology shown to be important in endometrial cancer [bib_ref] Mapping the immunosuppressive environment in uterine tumors: Implications for immunotherapy, Vanderstraeten [/bib_ref] or endometriosis [bib_ref] Indoleamine 2,3-dioxygenase suppresses the cytotoxicity of 1 NK cells in response to..., Liu [/bib_ref].
# Conclusions
The described model of primary human endometrial cells presents an opportunity for investigating biology of endometrium, especially the immune component (IDO1 and RAGE), in normal and cancer cells.
[fig] Figure 1: Potential of cell expansion in the in vitro culture conditions. (A) Comparison of the maximal number of passages and obtained number of frozen vials derived from cancer and the noncancerous endometrial tissue. (B) Growth kinetics of the selected endometrial cell cultures (cancer cells: E10-18, E12-23, E13-24 and control cells: E1-6, E22-38, E17-33) was assessed by plating 50,000 cells per well on 6-well plates and recording at the indicated time post-plating the cell number in the individual wells after trypsinization. [/fig]
[fig] Figure 2: Morphology of the in vitro cultured endometrial cells. The cells derived from cancer (E10-18, E12-23, E13-24) and noncancerous (E1-6, E22-38, E17-33) tissue were cultured at p7 (A) and p0 (B); the arrow indicates 3D cellular aggregate; scale bar 50 µm. [/fig]
[fig] Figure 3: Karyotype characterization of cancer cells isolated from patients and cultured in vitro cells. The fixed on slide noncancerous endometrial cells (A) E1-6, (B) E22-38, and cancer endothelial cells (C) E10-18 and (D) E12-23 were stained for R-banding (RHG) and G-banding (GTG). The analysis was performed on 15-20 fields and showed a female pattern without abnormalities. The figure presents representative pictures for each cell culture. Scale bars 10 μm [/fig]
[fig] Figure 4: Expression of CK8 and MUC1 in the cultured in vitro endometrial cells. The cells derived from noncancerous (E1-6, E22-38, E17-33) and cancer (E10-18, E12-23, E13-24) tissue were cultured on chamber slides, fixed and immunostained with (A) anti-CK8 (green) or (B) anti-MUC1 Ab (green); nuclei were co-stained with DAPI (red). [/fig]
[fig] Figure 5: Expression of RAGE and HER2 on endometrial cells cultured in vitro. (A) WB was performed on cancer cell lysates (1) SK-OV-3, (2) E12-23, (3) E10-18, (4) E13-24, (5) E24-41, (7) E9-17, and noncancerous cells (6) E1-6, (8) E17-33, (9) E22-38 grown to 70-100% confluency, harvested in RIPA buffer, and subjected to WB analysis with anti-RAGE Ab and anti-β-actin Ab as a loading control. (B) The immunofluorescent staining of the cancer (E10-18, E12-23, E13-24) and noncancerous (E1-6, E17-33, E22-38) endometrial cells was performed with anti-RAGE Ab (green) and nuclei were co-stained with DAPI (red). [/fig]
[fig] Figure 6: IDO expression and activity in endometrial cells cultured in vitro. (A) Immunofluorescent staining of cells obtained from cancer (E10-18, E12-23, E13-24) and noncancerous (E1-6, E17-33, E22-38) endometrial tissue probed with anti-IDO1 Ab (green) and co-stained with DAPI (red). (B) L-Kyn concentration in the medium from 5 cancer and 8 noncancerous (control) endometrial cell cultures was determined by HPLC-DAD. The Mann-Whitney U test was employed to compare median level of L-Kyn within cancer and noncancerous cells. Asterisk (*) indicates significant difference with p = 0.045. [/fig]
[table] Table 1: Characteristics of the study group. [/table]
[table] Table 2: Immunochemistry (IHC) of the tissue and L-Kyn level generated by in vitro cultured cells determined by High-Performance Liquid Chromatography with Diode Array Detector (HPLC-DAD). [/table]
|
Tailing and degradation of Argonaute-bound small RNAs protect the genome from uncontrolled RNAi
[fig] g: Argonaute-bound small RNAs from wild type cells were analyzed by high-throughput sequencing and classified as indicated below the pie charts. Pie charts illustrate percentages for the individual small RNA classes relative to the total number of reads [/fig]
[fig] e: Growth assay showing strong reduction in viability of cid14Δ cells that over-express rdp1 gene. Cells were growing for three days before imaging. [/fig]
[fig] f: ChIPseq experiment showing that H3K9me2 is increased at rDNA in cid14Δ cells [/fig]
[table] b: Quantification of Argonaute-bound small RNAs that map to TF2 or LTR elements in indicated strains. RNAi does not target transposable elements in fission yeast wild type cells. Error bars indicate s.e.m. of two independent small RNA sequencing experiments. Quantification of small RNAs that have non-templated nucleotides at the 3' end in different classes of Argonaute-bound small RNAs. Small RNAs that originate from centromeric region, mRNA and ncRNA and antisense rRNA are abundantly modified at the 3' end. Small RNAs originating from tRNAs and sense rRNA are modified less frequently. (e) Quantification of small RNAs that have non-templated nucleotides at the 3' end in different classes of size selected fraction. Only small RNAs that originate from centromeric region are abundantly modified at the 3' end. Small RNAs originating from mRNA, ncRNA, tRNAs and rRNA are modified less frequently suggesting that these small RNAs are not loaded on Argonaute. [/table]
|
Deucravacitinib for the Treatment of Psoriatic Disease
Psoriasis is an immune-mediated disease, with the interleukin (IL)-23/IL-17 axis currently considered its main pathogenic pathway. Tyrosine kinase 2 (TYK2) is responsible for mediating immune signalling of IL-12, IL-23 and type I interferons, without interfering with other critical systemic functions as other JAK proteins do. This article aims to review the current knowledge on deucravacitinib, a new oral drug that selectively inhibits TYK2, granting it a low risk of off-target effects. After good efficacy and safety results in a phase II, placebo-controlled trial, two phase III, 52-week trials evaluated deucravacitinib 6 mg against placebo and apremilast-an active comparator. POETYK PSO-1 and PSO-2 involved 1688 patients with moderate-to-severe psoriasis. After 16 weeks, in both studies, over 50% of patients treated with deucravacitinib reached PASI75, which was significantly superior to placebo and apremilast. In POETYK PSO-1, these results improved until week 24 and were maintained through week 52, with over 65% of patients achieving PASI75 at this point. A reduction in signs and symptoms was also reported by patients, with greater impact on itch. Deucravacitinib was well tolerated and safe. There were no reports of serious infections, thromboembolic events, or laboratory abnormalities, which are a concern among other JAK inhibitors. Persistent efficacy and consistent safety profiles were reported for up to 2 years. Despite advances in the treatment of psoriasis, namely among biologic agents, an oral, effective and safe new drug can bring several advantages to prescribers and patients. Further investigation is required to understand where to place deucravacitinib among current psoriasis treatment options.
# Introduction
Psoriasis is a chronic, inflammatory, immune-mediated disease with a high negative impact on patients' quality of life [bib_ref] Pathophysiology, clinical presentation, and treatment of psoriasis: a review, Armstrong [/bib_ref]. Psoriasis impairs patients' psychosocial wellbeing and work productivity, as a result of persistent lesions and social stigma [bib_ref] Absolute and relative psoriasis area and severity index (PASI) treatment goals and..., Gerdes [/bib_ref].
Psoriasis arises from a complex immunological disarray that is still not fully understood. Interleukin (IL)-23/IL-17 signalling pathway is currently considered the main pathogenic pathway [bib_ref] Discovery of the IL-23/IL-17 signaling pathway and the treatment of psoriasis, Hawkes [/bib_ref] [bib_ref] The role of IL-23 and the IL-23/TH17 immune axis in the pathogenesis..., Girolomoni [/bib_ref] [bib_ref] Psoriasis pathogenesis and the development of novel targeted immune therapies, Hawkes [/bib_ref]. Environmental stimuli or loss of selftolerance via autoantibodies trigger plasmacytoid dendritic cells (DCs) to produce tumour necrosis factor-α (TNF-α) and
## Key points
Tyrosine kinase 2 (TYK2) is a member of the Janus kinases family that is responsible for mediating the immune response associated with psoriasis through interleukin (IL)-12, IL-23 and type I interferons (IFN-α and IFN-β).
Deucravacitinib, an oral TYK2 selective inhibitor, has shown a good efficacy and safety profile up to 52 weeks in moderate to severe psoriasis.
This new drug may address the existing unmet need for oral options for psoriasis, but long-term evaluation and comparison trials with biologic agents might determine what place it will take among current therapeutic solutions.
All JAKs play an essential role in many immune responses. TYK2 mediates signalling of inflammatory cytokines of both adaptive (IL-12, IL-23) and innate (type I IFNs) immune responses [bib_ref] Molecular and clinical effects of selective tyrosine kinase 2 inhibition with deucravacitinib..., Catlett [/bib_ref]. Unlike TYK2, JAK1, 2 and 3 are also responsible for mediating a series of signals that support broader systemic responses, namely haematopoiesis, myelopoiesis, lipid metabolism and bone regulation. In fact, according to studies with mice models, complete deficiency of JAK1 and JAK2 seems to be incompatible with life, and there are no reports of human beings born with this genetic condition [bib_ref] Disruption of the Jak1 gene demonstrates obligatory and nonredundant roles of the..., Rodig [/bib_ref] [bib_ref] Biallelic JAK1 mutations in immunodeficient patient with mycobacterial infection, Eletto [/bib_ref] [bib_ref] Conditional deletion of Jak2 reveals an essential role in hematopoiesis throughout mouse..., Park [/bib_ref]. JAK3 inactivating mutations result in immunodeficiency syndromes in both mice and humans, and they present with life-threatening infections in the first months of life [bib_ref] Defective lymphoid development in mice lacking Jak3, Nosaka [/bib_ref] [bib_ref] Mutation of Jak3 in a patient with SCID: essential role of Jak3..., Russell [/bib_ref] [bib_ref] The International Journal of Biochemistry JAK3: a two-faced player in hematological disorders, Cornejo [/bib_ref].
JAK1 can pair with all other JAKs, being associated with type I IFN-α/β (JAK1-TYK2), IFN-γ (JAK1-JAK2) and IL-6 (JAK1/JAK2-TYK2) receptors. JAK3 only pairs with JAK1 and they become activated by IL-2, IL-4, IL-7, IL-9, IL-15 and IL-21 receptors, being crucial in lymphocyte function [bib_ref] JAK-STAT signaling as a target for inflammatory and autoimmune diseases: current and..., Banerjee [/bib_ref] [bib_ref] Janus kinases in immune cell signaling, Ghoreschi [/bib_ref].
The JAK2 pair (JAK2-JAK2) plays an important role in erythropoiesis and myelopoiesis signalling as it is connected to the receptor of erythropoietin, thrombopoietin and granulocyte macrophage-colony stimulating factor. The JAK2-TYK2 pair is triggered by IL-12 and IL-23 [fig_ref] Figure 1: Schematic representation of Janus Kinase Signal Transducer and Activator of Transcription [/fig_ref] [bib_ref] JAK-STAT signaling as a target for inflammatory and autoimmune diseases: current and..., Banerjee [/bib_ref] [bib_ref] Janus kinases in immune cell signaling, Ghoreschi [/bib_ref].
In line with this knowledge, JAK 1, 2 and 3 inhibitors such as tofacitinib (JAK 1/3), baricitinib (JAK 1/2) and ruxolitinib (JAK 1/2) raised safety concerns due to reports of important adverse effects like cytopenia, dyslipidaemia, gastric perforation, thromboembolic events and infections, namely herpes zoster [bib_ref] Tuberculosis and other opportunistic infections in tofacitinib-treated patients with rheumatoid arthritis, Winthrop [/bib_ref] [bib_ref] JAK inhibitors in rheumatoid arthritis: an evidence-based review on the emerging clinical..., Harrington [/bib_ref] [bib_ref] Incidence of venous and arterial thromboembolic events reported in the tofacitinib rheumatoid..., Mease [/bib_ref] [bib_ref] Herpes zoster in psoriasis patients treated with tofacitinib, Winthrop [/bib_ref] [bib_ref] an oral Janus kinase inhibitor, for the treatment of chronic plaque psoriasis:..., Papp [/bib_ref] [bib_ref] Thromboembolic events in Janus kinase inhibitors: a pharmacovigilance study from 2012 to..., Dong [/bib_ref]. Their clinical research in psoriasis has, therefore, been abandoned, mostly due to their unfavourable efficacy/safety ratio [bib_ref] JAK inhibitors for treatment of psoriasis: focus on selective TYK2 inhibitors, Nogueira [/bib_ref].
TYK2 participates almost exclusively in immune cytokine signalling pathways. Mouse models with TYK2 deficiency have a compromised immune cytokine signalling which results in a mild immunodeficiency phenotype. According to published case reports, human patients with TYK2 homozygous mutations present a higher susceptibility to mycobacterial and viral infections. Concomitantly, hyper-IgE syndrome occurred in some of the cases, depending on which one of five null mutations was present. These findings were attributed to impairment of the IL-23 pathway [bib_ref] Resolving TYK2 locus genotype-to-phenotype differences in autoimmunity, Dendrou [/bib_ref] [bib_ref] Human TYK2 deficiency: mycobacterial and viral infections without hyper-IgE syndrome, Kreins [/bib_ref] [bib_ref] A patient with tyrosine kinase 2 deficiency without hyper-IgE syndrome, Kilic [/bib_ref].
Considering TYK2 is involved in the signalling process of IFN-α, IL-12 and IL-23, it was accepted as a viable therapeutic target for psoriasis [bib_ref] Review treatment of dermatologic conditions: the evolution of JAK inhibitors, Jo [/bib_ref].
A study with mouse models suggested that TYK2 inactivation provides protection from multiple autoimmune diseases, including psoriasis. An association between TYK2 gene and psoriasis susceptibility was also reported in a genome-wide association study [bib_ref] A genome-wide association study identifies new psoriasis susceptibility loci and an interaction..., Strange [/bib_ref]. This evidence was the starting point for researching TYK2 as a potential therapeutic target for psoriasis [bib_ref] Resolving TYK2 locus genotype-to-phenotype differences in autoimmunity, Dendrou [/bib_ref].
interferon-α (IFN-α). These inflammatory cytokines activate myeloid DCs to release IL-23 and IL-12. IL-23 promotes the proliferation of T helper (Th) [bib_ref] Molecular and clinical effects of selective tyrosine kinase 2 inhibition with deucravacitinib..., Catlett [/bib_ref] [fig_ref] Table 2: [52]. [/fig_ref] [bib_ref] Discovery of the IL-23/IL-17 signaling pathway and the treatment of psoriasis, Hawkes [/bib_ref] [bib_ref] The role of IL-23 and the IL-23/TH17 immune axis in the pathogenesis..., Girolomoni [/bib_ref] [bib_ref] Genetics of psoriasis: evidence for epistatic interaction between skin barrier abnormalities and..., Bergboer [/bib_ref].
Plaque psoriasis is the most common form of the disease, accounting for 80% of the cases. Most patients have a mildto-moderate presentation of the disease, which can usually be managed with topical treatments. However, when topical therapy fails to control the disease and in extensive and severe cases, systemic treatment is required [bib_ref] Pathophysiology, clinical presentation, and treatment of psoriasis: a review, Armstrong [/bib_ref].
Current systemic therapies include oral conventional options, namely cyclosporine, methotrexate, fumaric acid esters and acitretin, newer targeted oral small molecules such as apremilast and parenteral biologic agents. The efficacy of current oral treatments in psoriasis is generally lower than that of biologics. In addition, conventional systemic therapies are associated with high risk of drug interactions, cumulative organ toxicities (hepatic, pulmonary, hematologic and renal) and potential severe adverse effects, and require extensive analytic monitoring, while apremilast is often not well tolerated. Thus, there is still an unmet need for accessible, efficacious and safe therapies for patients with moderate-to-severe psoriasis [bib_ref] Pathophysiology, clinical presentation, and treatment of psoriasis: a review, Armstrong [/bib_ref] [bib_ref] Efficacy and safety of acitretin in three fixed doses of 25, 35..., Dogra [/bib_ref] [bib_ref] Apremilast, an oral phosphodiesterase 4 (PDE4) inhibitor, in patients with moderate to..., Papp [/bib_ref] [bib_ref] An intensified dosing schedule of subcutaneous methotrexate in patients with moderate to..., Warren [/bib_ref] [bib_ref] National Psoriasis Foundation * Consensus Conference, Rosmarin [/bib_ref].
New small molecules are the subject of recent research. They have the potential to be administered orally or topically, which can be more convenient for some patients [bib_ref] Psoriasis pathogenesis and the development of novel targeted immune therapies, Hawkes [/bib_ref] [bib_ref] Small molecules in the treatment of psoriasis, Torres [/bib_ref] [bib_ref] New topical therapies for psoriasis, Le [/bib_ref].
This article aims to review the current knowledge on deucravacitinib, a new oral small molecule that selectively inhibits TYK2, for the treatment of psoriasis.
## Tykinhibition
Janus kinases (JAKs) are cytoplasmatic proteins involved in cytokine receptor signalling: JAK1, JAK2, JAK3 and TYK2. They act as signal transducers, translating extracellular stimuli into intracellular cascade, which results in altered gene expression. They are constituted by multiple domains. The highly conserved active domain is similar in all JAKs, while the regulatory domain has a specific configuration in each one of them [bib_ref] JAK-STAT signaling as a target for inflammatory and autoimmune diseases: current and..., Banerjee [/bib_ref].
When a circulating extracellular cytokine binds to its cell receptor, a combination of two members of the JAK family is recruited and phosphorylated. Subsequently, intracellular signal transducer and activator of transcription (STAT) proteins become phosphorylated and activated. This process enables STAT proteins to move into the cell nucleus and modulate gene expression. Each set of JAKs is associated with specific cytokine receptors [bib_ref] JAK-STAT signaling as a target for inflammatory and autoimmune diseases: current and..., Banerjee [/bib_ref] [bib_ref] Janus kinases in immune cell signaling, Ghoreschi [/bib_ref].
## Deucravacitinib
Deucravacitinib is a new oral small molecule that selectively inhibits TYK2 by uniquely binding to the TYK2 regulatory pseudokinase (JH2) domain (allosteric inhibition). By not binding to the conserved active domain (competitive inhibitors), like other JAK inhibitors, deucravacitinib is considerably more selective to TYK2 than to the other JAKs. Other JAK inhibitors like tofacitinib, upadacitinib and baricitinib variably inhibit JAK 1, 2 and 3, but not TYK2. The use of deucravacitinib in psoriatic patients decreases IL-23 pathway biomarkers in lesional skin towards non-lesional levels and normalizes IFN and IL-12 pathway genes, without affecting JAK 1, 2 and 3 biomarkers. Through this mechanism of action, deucravacitinib blocks IL-12, IL-23 and type I interferon downstream signalling, without interfering with JAK2-dependent hematopoietic functions [bib_ref] Molecular and clinical effects of selective tyrosine kinase 2 inhibition with deucravacitinib..., Catlett [/bib_ref] [bib_ref] Selectivity profile of the tyrosine kinase 2 inhibitor deucravacitinib compared with janus..., Chimalakonda [/bib_ref]. Preclinical studies initially showed its efficacy in mouse models with lupus erythematosus and inflammatory bowel disease [bib_ref] BMS-986165 is a highly potent and selective allosteric inhibitor of Tyk2, blocks..., Gillooly [/bib_ref] [bib_ref] Tyrosine kinase 2-mediated signal transduction in T lymphocytes is blocked by pharmacological..., Tokarski [/bib_ref] [bib_ref] Highly selective inhibition of tyrosine kinase 2 (TYK2) for the treatment of..., Wrobleski [/bib_ref].
## Phase i trial
Deucravacitinib's safety and good tolerability was shown in a phase I trial involving 108 healthy participants (Clini-calTrials.gov identifier: NCT02534636). For approximately 3 months, there were no reports of serious adverse effects and non-serious adverse events were evenly described in the active and the placebo groups (75% and 76%, respectively). Headache, nausea, rash and upper tract respiratory infections were the most frequently reported side effects.
Another phase I study was later developed to determine whether deucravacitinib had a clinically relevant effect on electrocardiographic parameters. A single oral dose of deucravacitinib 12 mg (therapeutic dose) or 36 mg (supratherapeutic dose) did not produce clinically relevant changes on Circulating cytokines couple to their receptors in the cell membrane. That connection triggers a conformational change of the receptor, which recruits a combination of autophosphorylated JAKs. Different receptors associate with different JAKs, each of them associated with specific biologic functions. JAKs create con-ditions to phosphorylate STAT proteins, causing their dimerization, which enables them to move into the cell nucleus and modulate gene expression. EPO erythropoietin, G-CSF granulocyte colony-stimulating factor, GH growth hormone, GM-CSF granulocyte-macrophage colony-stimulating factor, IL interleukin, IFN interferon, TPO thrombopoietin, TYK tyrosine kinase the corrected QT interval or other measured parameters in healthy adults [bib_ref] Lack of electrocardiographic effects of deucravacitinib in healthy subjects, Chimalakonda [/bib_ref].
## Phase ii trials
A phase II multicentre, double-blind, placebo-controlled trial was conducted in 267 adult patients with moderateto-severe psoriasis (NCT02931838) [bib_ref] Phase 2 trial of selective tyrosine kinase 2 inhibition in psoriasis, Papp [/bib_ref]. All 267 patients presented a Psoriasis Area Severity Index (PASI) score of 12 or higher, a Physician's Global Assessment (PGA) score > 2 and a body surface area (BSA) of 10% or worse. They were randomized into six arms, one corresponding to placebo and the other five evaluating different doses of oral deucravacitinib (3 mg every other day, 3 mg each day, 3 mg twice a day, 6 mg twice a day and 12 mg each day). At the end of 12 weeks of treatment, patients treated with deucravacitinib in a dose of 3 mg daily or higher achieved significantly greater rates of PASI75 (achieving 75% improvement in PASI from baseline) than with placebo. PASI75 was achieved in 39% of patients on 3 mg of deucravacitinib daily (p < 0.001 vs placebo). Among the groups treated with doses of 3 mg twice daily or higher, over 60% of patients reached a PASI75 response (p < 0.001 vs placebo) [bib_ref] Phase 2 trial of selective tyrosine kinase 2 inhibition in psoriasis, Papp [/bib_ref].
Additionally, the Dermatology Life Quality Index (DLQI) questionnaire was used to assess symptom-related discomfort and impact on daily functioning. A greater percentage of patients in the groups receiving 3 mg of deucravacitinib twice daily, 6 mg twice daily, or 12 mg daily presented a DLQI score of 0 or 1 (normal or near-normal quality of life) in comparison to the placebo group (42%, 60% and 64% in the respective deucravacitinib groups vs 4% in the placebo one-95% confidence intervals of 20-54, 38-71 and 41-74, respectively) [bib_ref] Phase 2 trial of selective tyrosine kinase 2 inhibition in psoriasis, Papp [/bib_ref].
A post-hoc analysis of this phase II trial evaluated the impact of deucravacitinib on patient-reported quality of life. The percentage of patients achieving a DLQI 0/1 in the three highest dosage groups combined increased through weeks 4-12. Improvement in quality of life followed a similar pattern of response to treatment to that of the clinical outcomes [bib_ref] Deucravacitinib in moderate to severe psoriasis: clinical and quality-of-life outcomes in a..., Thaçi [/bib_ref].
Regarding safety data, the percentage of patients with reported adverse events ranged from 55% in the group treated with 3 mg each day to 80% in the group on 6 mg twice a day, whereas the placebo group presented the lowest rate of adverse effects (51%). The most common were nasopharyngitis, headache and diarrhoea. Acne was reported only in the active treatment groups (n = 8), with four cases (9%) in the highest dose arm. Only four patients reported serious adverse effects and they belonged either to the placebo group or to groups treated with the lowest doses of deucravacitinib. No cases of herpes zoster infection, tuberculosis, opportunistic infections or cardiovascular events were reported during the 12-week period of the trial [bib_ref] Phase 2 trial of selective tyrosine kinase 2 inhibition in psoriasis, Papp [/bib_ref].
## Phase iii trials
Two large, phase III, double-blinded, 52-week trials on plaque psoriasis were recently completed: NCT03624127 (POETYK PSO-1) and NCT03611751 (POETYK PSO-2) [bib_ref] Deucravacitinib versus placebo and apremilast in moderate to severe plaque psoriasis: efficacy..., Armstrong [/bib_ref].
In POETYK PSO-1, 666 patients with moderate to severe plaque psoriasis (BSA ≥ 10%, PASI ≥ 12, PGA ≥ 3) were randomly assigned to deucravacitinib 6 mg once daily, placebo or an active comparator in a 2:1:1 proportion. The first group maintained the same dose of 6 mg once a day for 52 weeks, while the placebo group switched to deucravacitinib 6 mg once a day after 16 weeks. In the active comparator group, patients were treated with apremilast 30 mg twice a day, titrated from 10 mg over the first 5 days. At the end of 24 weeks, PASI response was evaluated and patients reaching PASI50 were maintained on apremilast. All patients from this arm who did not reach PASI50 at this point were switched to deucravacitinib treatment [bib_ref] Deucravacitinib versus placebo and apremilast in moderate to severe plaque psoriasis: efficacy..., Armstrong [/bib_ref].
POETYK PSO-2 involved 1022 patients with moderate to severe plaque psoriasis who were also randomly assigned to one of three different arms. Patients assigned to receive placebo switched to deucravacitinib 6 mg once a day after 16 weeks. The other two groups started with deucravacitinib 6 mg daily or with apremilast 30 mg twice a day, titrated up from 10 mg over the first 5 days. POETYK PSO-2 included a randomized withdrawal phase in which patients originally randomized to deucravacitinib who had achieved PASI75 response at week 24 were re-randomized 1:1 to placebo or deucravacitinib, whereas those who did not achieve PASI75 response at week 24 continued being treated with deucravacitinib. Regarding patients on apremilast for the first 24 weeks, those who achieved PASI75 were maintained under the same treatment, while those who did not were started on deucravacitinib.
According to POETYK PSO-1 results at the end of 16 weeks, response rates were significantly higher with deucravacitinib versus placebo and apremilast regarding PASI75 (58.4% vs 12.7% vs 35.1%, respectively; p < 0.0001) and PGA 0/1 (53.6% vs 7.2% vs 32.1%, respectively; p < 0.0001) .
At week 16, there were more deucravacitinib patientsversus placebo and apremilast patients-achieving an absolute PASI ≤ 1 (PSO-1: 24.4%, 1.8%, 10.1%, respectively) and ≤ 5 (PSO-1: 59.3%, 14.5%, 35.7%, respectively).
Deucravacitinib-treated patients reported greater quality of life improvement in comparison with other groups, with a significantly greater DLQI 0/1 response rate at week 16 (41.0%) versus patients who received placebo (10.6%; p < 0.0001) or apremilast (28.6%; p = 0.0088) [bib_ref] Deucravacitinib versus placebo and apremilast in moderate to severe plaque psoriasis: efficacy..., Armstrong [/bib_ref].
These responses were maintained through week 52 in the deucravacitinib group in POETYK PSO-1, with 65.1% of patients achieving PASI75 response and 52.7% a PGA of 0/1. Patients who switched from placebo to deucravacitinib at week 16 had PASI75 and PGA 0/1 responses at week 52 comparable to those observed in patients who received continuous deucravacitinib treatment from day 1 [bib_ref] Deucravacitinib versus placebo and apremilast in moderate to severe plaque psoriasis: efficacy..., Armstrong [/bib_ref]. POETYK PSO-2 obtained similar results: 53.6% of patients in the deucravacitinib arm achieved PASI75 at week 16 in comparison with 9.4% (p < 0.0001) of patients receiving placebo and 40.2% (p = 0.0003) of those receiving apremilast . Among deucravacitinib-treated patients who achieved PASI75 at week 24 and were re-randomized to continue treatment, responses were maintained at week 52 in the most patients (PASI75 of 80.4%). Information regarding patients who switched to placebo is still unavailable.
Efficacy outcomes on scalp psoriasis were separately analysed in both POETYK studies. Significantly more patients receiving deucravacitinib achieved scalp-specific PGA 0/1 and Psoriasis Scalp Severity Index (PSSI 90), in comparison with placebo and apremilast. At the end of 16 weeks, 70.8% of deucravacitinib-treated patients in POEYK PSO-1 and 60.3% in POETYK PSO-2 reached scalp-specific-PGA 0/1 (p < 0.0001).
Daily evolution of symptoms (burning, itch, pain, skin tightness, stinging) and signs (bleeding, cracking, dryness, redness, scaling, shedding or flaking) was also evaluated during these trials by using a patient-reported numerical scale ranging from 0 (absent) to 10 (worst imaginable). At week 16, in both POETYK PSO-1 and POETYK PSO-2, patients treated with deucravacitinib experienced significantly greater improvements in mean change from baseline in symptom scores (− 32.0 vs − 6.3 for placebo and − 23.7 for apremilast in POETYK PSO-1 and − 31.3 vs − 3.2 for placebo and − 23.0 for apremilast in POETYK PSO-2; p < 0.0001 for each) and sign scores (− 34.3 vs − 7.7 for placebo and − 25.4 for apremilast in POETYK PSO-1 and − 35.0 vs − 6.3 for placebo and − 26.5 for apremilast in POETYK PSO-2; p < 0.0001 for each). The greatest improvements in psoriasis symptoms were observed for itch and skin tightness.
Preliminary safety results suggested a good tolerability and safety profile in both trials. The most reported adverse effects were nasopharyngitis, upper respiratory tract infections, headache and diarrhoea. According to peerreviewed data from POETYK PSO-1, herpes zoster infections occurred at a low rate in the deucravacitinib arm (n = 5) and all cases were mild to moderate. Acne was reported in 15 out of 531 patients on deucravacitinib against 0 among placebo and apremilast groups. Serious adverse events were more frequent with placebo (5.5%) than apremilast (2.4%) or deucravacitinib (2.1%) at week 16. Discontinuation due to adverse effects over weeks 0 to 52 were lower with deucravacitinib versus placebo and apremilast. Changes from baseline levels of standard hematologic parameters (lymphocytes, neutrophils, platelets and haemoglobin) and chemistry parameters including lipid panel and creatine phosphokinase (CPK) were globally not clinically relevant from weeks 0 to 16 in both POETYK PSO-1 and 2 [bib_ref] Deucravacitinib versus placebo and apremilast in moderate to severe plaque psoriasis: efficacy..., Armstrong [/bib_ref].
At the end of the 52-week POETYK PSO-1 and 2, 1221 patients were switched to an open-label deucravacitinib extension trial for up to 240 weeks. Recently, data concerning long-term extension (LTE) results was presented [bib_ref] Poster-Deucravacitinib long-term efficacy and safety in plaque psoriasis: 2-year results from the..., Warren [/bib_ref].
From week 0 to 60 of POETYK PSO-LTE (NCT04036435), patients who were already on deucravacitinib at week 0 kept improving (PASI75 from 70.8 to 79.4% and PASI90 from 43.6% to 50.7%). Patients who switched Efficacy results at 16 weeks of deucravacitinib 6 mg once daily in comparison with placebo and with the active comparator apremilast (30 mg twice a day, titrated from 10 mg over the first 5 days), in two large, phase III, double-blinded trials with patients with moderate to severe psoriasis [bib_ref] Deucravacitinib in moderate to severe psoriasis: clinical and quality-of-life outcomes in a..., Thaçi [/bib_ref] [bib_ref] Deucravacitinib versus placebo and apremilast in moderate to severe plaque psoriasis: efficacy..., Armstrong [/bib_ref] POETYK PSO-1 (n = 666) and POETYK PSO-2 (n = 1020) PASI75 at least a 75% improvement from baseline in Psoriasis Area and Severity Index scores, PGA 0/1 Physician's Global Assessment score of clear or almost clear; ss-PGA 0/1 scalp-specific Physician's Global Assessment score of clear or almost clear in those with ss-PGA of at least 3 (moderate) at baseline; DLQI Dermatology Life Quality Index 0/1 scores reflect no effect at all on patient's life in patients with a baseline DLQI score of ≥ 2 from apremilast to deucravacitinib at week 0 also clinically improved (PASI75 from 73.8 to 87.1% and PASI90 from 40.0 to 62.9%) [bib_ref] Poster-Deucravacitinib long-term efficacy and safety in plaque psoriasis: 2-year results from the..., Warren [/bib_ref]. Aside from COVID-19, the safety profile was consistent across POETYK PSO-1, PSO-2 and LTE. An increase in serious infections was observed, which is attributable to COVID-19 infections due to the ongoing pandemic [bib_ref] Poster-Deucravacitinib long-term efficacy and safety in plaque psoriasis: 2-year results from the..., Warren [/bib_ref].
A smaller, phase III, week, open-label, single-arm, clinical trial with 80 Japanese patients with moderate-tosevere psoriasis (NCT03924427) was also recently completed, with no published results to date.
## Psoriatic arthritis
There is also emerging data suggesting good efficacy and safety results of deucravacitinib in patients with active psoriatic arthritis. A total of 203 patients with psoriatic arthritis were randomized 1:1:1 to placebo, deucravacitinib 6 mg once a day and deucravacitinib 12 mg once a day in a double-blind phase II trial (NCT03881059). American College of Rheumatology-20 (ACR-20) response at week 16 was considered the primary endpoint, which refers to an improvement from baseline in the number of tender and swollen joints and patient's pain, among other parameters. ACR-20 response was significantly higher with deucravacitinib 6 mg once a day (52.9%; p = 0.0134) and 12 mg once a day (62.7%; p = 0.0004) against the placebo (31.8%). Higher numbers of patients treated with deucravacitinib versus placebo also achieved enthesitis and dactylitis resolution. Adverse events were more frequently reported at both deucravacitinib doses (65.7%) compared with placebo (42.4%). The most common ones in deucravacitinib-treated patients were nasopharyngitis, upper respiratory tract infections, sinusitis, bronchitis, rash, diarrhoea and headache. Acne was reported in 2 of 70 (2.9%) patients in the 6-mg, once-a-day deucravacitinib group, 1 of 67 (1.5%) in the 12-mg, once-aday group and dermatitis acneiform was reported in 2 of 70 (2.9%) and 2 of 67 (3.0%), respectively. No placebo patients developed acne or dermatitis acneiform [bib_ref] Efficacy and safety of selective TYK2 inhibitor, deucravacitinib, in a phase II..., Mease [/bib_ref].
Phase III studies are currently in the recruiting phase (NCT04908202 and NCT04908189) (
# Discussion
Systemic therapies for psoriatic disease have evolved rapidly in recent years, and several highly effective biologic drugs have been developed. However, there is still an unmet need for effective and safe oral therapies, as conventional systemic treatments are associated with important side effects, long-term toxicity and drug interactions. Apremilast, a new targeted small molecule, is only moderately effective [bib_ref] Efficacy and safety of acitretin in three fixed doses of 25, 35..., Dogra [/bib_ref] [bib_ref] Apremilast, an oral phosphodiesterase 4 (PDE4) inhibitor, in patients with moderate to..., Papp [/bib_ref] [bib_ref] An intensified dosing schedule of subcutaneous methotrexate in patients with moderate to..., Warren [/bib_ref] [bib_ref] National Psoriasis Foundation * Consensus Conference, Rosmarin [/bib_ref].
JAK inhibitors have been of particular interest due to their ability to target multiple cytokines simultaneously. JAK inhibitors, as a class, although proven to be effective in the treatment of psoriasis, are also associated with some off-target effects that translate into changes in laboratory parameters (haemoglobin, platelets, neutrophils, lipid panel, creatine phosphokinase), higher risk of infections, malignancy and thromboembolic events. Available evidence suggests that these unwanted effects are a consequence of JAK 1, 2 and 3 pathways inhibition. Therefore, due to safety concerns with JAK 1, 2 and 3 inhibitors, interest has shifted to more selective inhibition of TYK2. At clinically relevant doses and exposures, deucravacitinib demonstrates highly selective inhibition of TYK2 [bib_ref] JAK inhibitors for treatment of psoriasis: focus on selective TYK2 inhibitors, Nogueira [/bib_ref] [bib_ref] Review treatment of dermatologic conditions: the evolution of JAK inhibitors, Jo [/bib_ref] [bib_ref] Selectivity profile of the tyrosine kinase 2 inhibitor deucravacitinib compared with janus..., Chimalakonda [/bib_ref].
Deucravacitinib, a novel oral TYK2 selective inhibitor, has shown a good efficacy and safety profile in moderate to severe psoriasis trials. A phase II study showed that a daily dose of at least 3 mg was necessary to accomplish efficacy purposes. Considering higher rates of adverse effects in the 6-mg twice-a-day group, 6 mg once a day was the dose chosen for phase III studies [bib_ref] Phase 2 trial of selective tyrosine kinase 2 inhibition in psoriasis, Papp [/bib_ref]. Phase III trials evaluated deucravacitinib 6 mg versus placebo and apremilast as active comparator. Deucravacitinib showed superior and faster improvement of psoriasis lesions than placebo and apremilast. The clinical response improved through week 24 and was maintained in patients who were under treatment thorough week 52. The impact on scalp psoriasis was separately analysed with even stronger efficacy outcomes.
A subjective reduction in signs and symptoms was also noted. The greatest symptom improvement was consistently observed for the itch domain; this may be particularly meaningful to patients given the prevalence and burden of this symptom. The positive repercussion of this therapy on patients' quality of life was also superior compared with apremilast and placebo.
Deucravacitinib was globally well tolerated and safe. In contrast, tolerability is often a cause of discontinuation for both conventional agents and apremilast. Nasopharyngitis and upper respiratory tract infections were the most frequently reported adverse events. Acne was first described in the active treatment groups of the phase II study, with greater incidence in the arms with the highest doses. It has been suggested to be a consequence of commensal bacteria proliferation allowed by cytokine inhibition [bib_ref] Phase 2 trial of selective tyrosine kinase 2 inhibition in psoriasis, Papp [/bib_ref]. Results of phase III studies using deucravacitinib 6 mg a day report this adverse effect as uncommon. A higher selectivity for TYK2 versus JAK 1, 2 and 3 reduces the chance of offtarget effects, and the reported safety profile was consistent with its mechanism of action. There was no description of herpes zoster, opportunistic infections, thromboembolic events, or hematologic and lipid laboratory abnormalities, in contrast to trials of JAK 1, 2 and 3 inhibitors in psoriasis and other diseases [bib_ref] Tuberculosis and other opportunistic infections in tofacitinib-treated patients with rheumatoid arthritis, Winthrop [/bib_ref] [bib_ref] JAK inhibitors in rheumatoid arthritis: an evidence-based review on the emerging clinical..., Harrington [/bib_ref] [bib_ref] Incidence of venous and arterial thromboembolic events reported in the tofacitinib rheumatoid..., Mease [/bib_ref] [bib_ref] Herpes zoster in psoriasis patients treated with tofacitinib, Winthrop [/bib_ref] [bib_ref] an oral Janus kinase inhibitor, for the treatment of chronic plaque psoriasis:..., Papp [/bib_ref]. Additionally, no clinically significant laboratory parameter abnormalities were observed with deucravacitinib treatment, suggesting that routine laboratory monitoring during deucravacitinib treatment may not be warranted.
Deucravacitinib demonstrated persistent efficacy and consistent safety profiles in patients with psoriasis for up to 2 years. Information from the POETY PSO-2 trial regarding patients who responded to deucravacitinib and were rerandomized to placebo is still unavailable. These results will be essential to evaluate time until relapse after drug withdrawal. Currently available oral drugs have not been able to induce a long maintenance of clinical response after drug withdrawal, as we have seen with IL-23 inhibitor biologic agents. Being an IL-23 inhibitor as well, if deucravacitinib demonstrates this phenomena it would greatly impact clinical practice.
Deucravacitinib clinical efficacy in psoriatic arthritis is under research, with promising results in a phase II study. It is also being studied for lupus, ulcerative colitis and Crohn's disease. Moreover, this compound's positive phase III data led to an increased interest in molecules targeting the TYK2 JH2 domain, and several companies are announcing phase I and II studies with other highly selective TYK2 inhibitors.
At this point, biologic therapies offer effective, safe and convenient options for psoriatic patients. However, oral options bring some advantages, such as no risk of immunogenicity or injection-site reactions, easier transportation and storage and may be more comfortable for patients who prefer an oral treatment they can take at home or for those with trypanophobia. Although they do not need expert personnel drug administration or patient training, treatment response might be easily compromised as a result of incorrect adherence [bib_ref] TYK2 inhibition: changing the treatment landscape for psoriasis?, Abduelmula [/bib_ref]. Further pharmacoeconomic studies are required.
The role of deucravacitinib in the treatment landscape for psoriasis is still to be determined. It will probably be used as a first-line agent, but also after biologic exposure, since at least one third of the POETYK PSO-1 and POETYK PSO-2 study population had prior biologic exposure. In addition to clinical data, price and regulatory factors are likely to be decisive in defining the place of deucravacitinib in the treatment of psoriasis.
# Conclusion
Deucravacitinib has the potential to become an efficacious, safe and well-tolerated treatment for patients with moderate to severe psoriasis. Evaluation of longer-term response, maintenance of clinical response after drug withdrawal and comparison trials with biologic agents are important missing steps to understand this drug's positioning in psoriasis treatment.
Celgene, Fresenius-Kabi, Janssen, JS BIOCAD, LEO Pharma, Lilly, Mylan, Novartis, Pfizer, Regeneron, Roche, Sandoz, Samsung-Bioepis, Sanofi and UCB. Tiago Torres has received consultancy and/or speaker's honoraria from and/or participated in clinical trials sponsored by AbbVie, Amgen, Almirall, Amgen, Arena Pharmaceuticals, Biocad, Biogen, Boehringer Ingelheim, Bristol Myers Squibb, Celgene, Fresenius-Kabi, Janssen, LEO Pharma, Eli Lilly, MSD, Mylan, Novartis, Pfizer, Samsung-Bioepis, Sanofi-Genzyme, Sandoz and UCB.
Ethics approval Not applicable.
[fig] Figure 1: Schematic representation of Janus Kinase Signal Transducer and Activator of Transcription (JAK/STAT) pathway and respective activating cytokines. [/fig]
[table] Table 2: [52]. [/table]
|
Fine Needle Aspiration Cytology for Neck Masses in Childhood. An Illustrative Approach
The primary indication of fine-needle aspiration cytology of the head and neck region is a thyroid nodule or a mass located in the cervical area or the head. Although a thyroid nodule may raise the suspicion of malignancy, less than one in 20 cases results in a carcinoma. In addition, the list of differential diagnoses is quite different according to the age of the patient. A number of benign lesions, such as branchial cysts, sialadenosis, and sialoadenitis are often seen in childhood and youth. The malignant lesions that are on the top of the list of a pediatric mass of the head and neck (H&N) region include rhabdomyosarcoma, neuroblastoma, and papillary carcinoma of the thyroid gland. This critical review of the diagnostic features of a pediatric mass of the H&N region is accompanied by panels of several cytology features that may be of help to the cytopathologist and clinician.
# Introduction
In the era of transcriptomics and advancements identified at increasing pace in this 2nd decade of the 3rd millennium, an apparently simple approach with a needle, a slide, and some staining tools may be awkward. It seems that the fine needle aspiration biopsy was performed first in 1857, although several techniques and new protocols have transformed this seemingly simple procedure into a practice that is impressively diffuse worldwide. Aside from scarring in some patients, there are complications due to unnecessary surgery which is well known to many pediatricians. In adults, fine needle aspiration cytology (FNAC) has been used with great success as the primary screening test for thyroid nodules. In non-thyroid nodules, FNAC has also increased rates of sensitivity and specificity in numerous studies. However, such a structured approach has remained unused in the evaluation of pediatric nodules and instead, urgent analysis using scalpel and surgery is performed. This critical review of the literature highlights the information we can receive using FNAC in pediatrics and shows some illustrative cytologic features that may be extremely useful for both training or experienced cytologists and pediatricians. The present study was designed to critically review the role of FNAC and its utility in pediatric head and neck (H&N) lesions. It is not a comprehensive atlas and has no aim to fill all gaps of the H&N lesions in pediatrics, which are covered in excellent texts and books. Nevertheless, our paper emphasizes the procedure and specifies the spectrum of the H&N lesions in the pediatric age group, highlighting some studies gathered from the literature comparing cytology and histology of the H&N lesions.
# Methodology
A neck mass that is resistant to antibiotics for more than four weeks should be considered the primary indication to proceed with non-thyroid FNAC. Other signs should also include cases of asymmetric benign lymphadenopathy as well as cervical lymph nodal enlargement showing unusual location, fast expansion, evidence of weight loss or loss of appetite, and night sweats. The identification of skin changes and/or a fixed/immobile mass may favor a malignant lesion and is also an indication to proceed with FNAC. An FNAC with a critical result may often be managed in our institution within a few hours planning for surgery to be performed in the same day or the day after. Nonthyroidal FNAC should be conducted while keeping in mind that the patient may be a potential surgical candidate. Other indications for an FNAC may involve an atypical presentation with systemic symptoms; the urgency made clear by the medical team or family and the willingness to diagnose the lesion before the typical 2-7 days of surgical processing. In consideration of the high percentage of papillary thyroid carcinomas in childhood and youth, FNAC should be offered as the first procedure for all pediatric thyroid nodules following an ultrasound scan showing a lesion with a diameter of 1 cm or more. In the case of pediatric thyroid nodules with lesions measuring less than 1 cm in diameter, an FNAC with a critical result processing (same day or day after or within 24 h) should be offered in the event that concerning ultrasonographic results are present. Ultrasound anomalies will include hypoechogenicity, irregularity of the margins, or increased vascularity of the lesion. FNAC is usually performed in the Otolaryngology/ear-nose-throat (ENT) clinical department or inpatient ward by specialized and well-trained personnel, including a cytopathologist, an interventional radiologist, or a (pediatric) surgeon. Topical (4% lidocaine cream) anesthesia is used with some general low sedation, and all biopsies should be performed using image guidance. In our opinion, the practice to use palpation should be discouraged or left to very palpable nodules to reduce the annual false negativity rate. The procedure involves the use of a 25-to 27-gauge needle, and a range of 3 to 5 passes are performed for each targeted site, representing a single FNA procedure. If a patient has multiple targeted nodules or masses, various aspirates are usually performed at the discretion of the medical personnel. The aspiration of the material is then used for smear preparation, which includes air-dried slides stained with Diff-Quik and alcohol-fixed slides stained with the Papanicolaou stain. It is paramount to perform additional testing at the same time to avoid unnecessary delay in the diagnosis. In fact, residual material is typically used for ThinPrep processing (Hologic Inc., Marlborough, MA, USA), microbiology, and molecular biology studies. To confirm whether the material is adequate in a lymph node, a small drop of the aspirate is placed onto the preparation slides for immediate assessment (also technically labelled 'rapid on-site evaluation'), while the remainder is rinsed in a cell preservative. Typically, 10 million cells are considered adequate for cytopathology assessment, which requires 2-3 passes indeed.
The cell block preparation is routinely performed in our institution and not only in case of ambiguous interpretation of the aspirates. It is also important to advise the lab technician to stain some tissue sections of the cell block with hematoxylin-eosin and keep some unstained tissue sections for histochemical (special) stains, immunohistochemical stains, and further ancillary testing as required. The cytology aspirates are interpreted initially by a cytotechnologist and subsequently by a pediatric pathologist with an intra-institutional cytopathology consultation adjured by a cytopathologist. FNAC cases receive diagnoses with an adequacy interpretation (unsatisfactory, satisfactory), a primary interpretation (nondiagnostic, negative for malignant cells, atypical cells present, suspicious for cancerous cells, or positive for malignant cells), and a free text explanatory diagnosis. The category 'less than optimal' is not often used in our institutions to avoid non-categorical decisions.
It is also advisable to perform a yearly outcome rating using spreadsheet-based correlation to surgical histopathology and clinical follow-up. In the case of clinical monitoring, laboratory values and clinical presentation are part of a 6-month-1-year follow-up, while surgical histopathology, which is based on either an incisional surgical biopsy or an excisional surgical biopsy, is considered the reference standard in the diagnosis of tumors of the H&N region. In the setting of a quality assurance program, true positives are cytopathologic results that warrant surgical treatment and are confirmed as such using surgical histopathologic biopsy. True negatives are cytopathologic results that did not designate a need for surgery and are established as such using the surgical histopathologic biopsy. Clinical true negatives are negative conditions on cytology that resolve or did not progress without surgical intervention. In case of false positives cytology, cytopathology results are positive, but there would have been no need for surgical treatment, while false negatives are unremarkable cytopathologic results without a need for surgery, although a surgical intervention would have been the right option. To avoid bias in any cytology-pathology correlation quality assurance program, it is prudent that pathologists interpreting surgical histopathology slides are distinct from cytopathologists interpreting the FNAC. Some infective or non-malignant conditions, such as atypical mycobacterial infection, cervicofacial abscess, or lymphatic malformations, may require surgical diagnosis or treatment but may also appropriately be managed by medical therapy. If the appropriate medical treatment is initiated and results efficacious and efficient, the result is labeled as an actual clinical positive. The 2015 Standards for Reporting of Diagnostic Accuracy (STARD) guideline for reporting diagnostic accuracy studies should be used.
In the setting of procedure complications, care should be taken when considering the side effects of FNAC, e.g., infection and damage to nearby structures. By reviewing the literature, we found a limited core of evidence for such side effects if a safe procedure is applied. In particular, an honest and straightforward report provided by Chen et al., who reported a 19-year-old girl with a painful goiter, which became more painful after fine needle aspiration (FNA). The patient's second FNAC revealed only many neutrophils and an antibiotic treatment improved the pain, but the goiter persisted. The third FNA revealed cytology of papillary carcinoma cells and the total thyroidectomy showed ischemic necrosis with neutrophilic aggregation around the needle track other than a papillary thyroid carcinoma. This report reveals the possibility of missing malignant cells in a background of an infection following an improper FNA. Secondary infection and ischemic necrosis may indeed occur after FNA, and aseptic procedures are necessary to prevent bacteria from seeding into the thyroid gland. Disinfection of the FNAC site and experienced personnel are mandatory for laboratories that harbor periodic accreditation by the College of the American Pathologists (CAP) operating in healthcare institutions in United States of America and Canada. Thus, side effects may occur following an FNA procedure. Nevertheless, we consider that FNAC in children and adolescents is accurate, safe, and well-accepted in a wide range of lesions of the head and neck regions. A CAP accreditation or a similar accreditation outside of North America should be requested by all pediatricians and parents of the affected child.
## Diagnostics
In most of the cases, FNAC of the H&N region is performed to investigate clinically suspicious lymphadenopathy, and the primary differential diagnoses include congenital anomaly, reactive/infectious lymphadenopathy, lymphoma, and metastatic disease, but also thyroid gland disease and salivary gland disease.
## Congenital lesion
The age is also particularly important to keep in mind, and 75% of branchial cysts occur in 20-40 year-old patients. Benign squamous-lined cysts are usually characterized by abundant inflammatory cells (neutrophils), with few unremarkable squamous epithelial cells with bland nuclear features, and cholesterol crystals, while a large number of squamous cells with or without occasional nuclear hyperchromasia, nuclear membrane irregularity, and high nucleus to cytoplasm ratio portend for a malignant diagnosis.
## Reactive/infectious lymphadenopathy and malignant lymphoma
In about 2/3 of cases, de novo lymphadenopathies of the H&N region are benign, and even a history of malignancy present in a child with a de novo lymphadenopathy will have a rate of benignancy probably approaching half of the cases. Moreover, additional surgical procedures have been avoided in up to 61% of cases in one review. The cytopathological diagnosis of reactive is mostly accurate in children and youth with less in 1 out of 20 (~5%) subsequent (false-negative) malignant diagnoses. Cytologic features of reactive lymphadenopathy include polymorphic lymphoid cells showing a clear-cut sequence of maturation, prominent reactive cells arising from germinal centers, the presence of tingible body macrophages, the lack of a monomorphic lymphoid population and the simultaneous presence of a variegated (non-homogeneous) cell population smear, and the lack of a subpopulation (even minimal) of large, irregular lymphoid cells potentially indicative of Hodgkin's lymphoma, anaplastic T-cell lymphoma, large B-cell lymphoma). Necrotizing granulomata are often characteristic of underlying tuberculous infection (particularly systemic), although they may also be found in fungal infection and other acid-fast bacilli (AFB) (e.g., mycobacteria causing scrofula). In a child, the finding of granulomata may suggest non-tuberculous AFB, especially in a child, and cat-scratch disease should be considered in this age group. Granulomas without evidence of necrosis are suggestive of sarcoidosis, toxoplasmosis, or foreign-body giant cell reaction. In all cases, a thorough microbiological investigation is paramount (e.g., Bartonella henselae for cat-scratch disease). The correlation between clinical and laboratory findings using electronic medical charts or liaising with the clinician is essential. In fact, serological tests may be useful in a subset of cases for differentiating systemic lupus erythematosus lymphadenitis from Kikuchi-Fujimoto's disease, which is histiocytic necrotizing lymphadenitis. Finally, it is important not to forget that a florid granulomatous pattern can mislead the cytopathologist disguising both Hodgkin's lymphoma and Non-Hodgkin's lymphoma.
## Rhabdomyosarcoma and neuroblastoma
In a pediatric setting, the most fearful malignant diagnoses in the H&N region, apart from malignant lymphoma, are rhabdomyosarcoma and neuroblastoma. Rhabdomyosarcoma is considered to harbor three main subtypes, including embryonal, which is the most frequent type, the alveolar, and the pleomorphic, which most commonly seen in adults. Rhabdomyosarcoma occurs very often in the head and neck region and the extremities, with the trunk reserved for the third location. Cytomorphology features of the embryonal rhabdomyosarcoma include a variability of patterns including round to spindle-shaped cells, some pleomorphism (different from the marked pleomorphism of the pleomorphic type of the rhabdomyosarcoma of the adult), a variable number of rhabdomyoblasts may be the distinctive clue for the cytopathologist, and, occasionally, inclusion-like cytoplasmic condensation, which is also known as myogenic differentiation. Rhabdomyoblastic differentiation is characterized by elongated, strap-or tadpole-shaped cytoplasm, and nuclear eccentricity. The nuclear diagnostic clues also rely on dense hyperchromatic chromatin and nuclear membrane irregularity. The alveolar rhabdomyosarcoma is distinctively identified by larger, uniformly round to variably polygonal cells with an increased number of mitotic figures in the background with scattered rhabdomyoblasts and giant cells with wreath-like nuclei. Metastatic neuroblastoma to the neck is quite rare in adulthood, but it is a principal differential diagnosis in infants and children, especially in infants. Although most neuroblastomas arise in the adrenal gland or the paravertebral sympathetic chain bilaterally, this pediatric tumor can occur anywhere along the sympathetic chain. Cytomorphology of neuroblastoma may be entirely different according to the degree of ganglionic differentiation of the neuroblastoma. These features include a fibrillary matrix, dyshesive small undifferentiated cells, sparse Homer-Wright rosettes, a nuclear pattern labeled as "salt and pepper" for the variability of the granularity of the nuclear chromatin, and, occasionally, ganglion-like cells. In rare occasions, necrotic debris and dysmorphic calcifications are seen. It is paramount to distinguish reactive/reparative changes from the diagnoses mentioned above. A spectrum of appearance needs to be kept in mind in case of benign fibrous proliferation. The retention of a low nucleus to cytoplasm ratio is crucial in case of prior radiation where cells may be large, showing irregularity of the nuclei, multinucleation, all features that may be quite worrisome. In fact, the resolution may decrease the inflammatory background, making the diagnosis particularly distressful for an inexperienced pathologist. Moreover, the myofibroblasts that occur in benign fibrous proliferations may become slender with less conspicuous nucleoli during the healing process. In all cases, the identification of pleomorphism, nuclear hyperchromasia, high nucleus to cytoplasm ratio, and atypical mitoses suggest malignancy that needs to be surgically acted upon as soon as possible.
Diagnostics 2018, 8, x 5 of 11 cytoplasm ratio, and atypical mitoses suggest malignancy that needs to be surgically acted upon as soon as possible.
## Thyroid gland disease
The introduction of synoptic reports has been crucial in improving healthcare worldwide, allowing a common platform for adequately assessing patients with neoplastic disease. Thyroid cases receive an adequacy interpretation, a first interpretation using "The Bethesda System for Reporting Thyroid Cytopathology" (TBSRTC), and a free text explanatory diagnosis with a comment according to the pathologist or cytopathologist. It is also encouraged that the cytopathologist or the pathologist liaise with the clinical team in case an FNAC may indicate suspicious malignant cells or some distinct cancerous cells. TBSRTC categories include insufficient for diagnosis, benign, atypical cells of undetermined significance, suspicious for a follicular neoplasm, suspicious for a Hürthle cell neoplasm, suspicious for malignancy (papillary carcinoma, medullary carcinoma, lymphoma, metastatic tumor, other), and malignant. Despite iodine supplementation, multinodular goiter (MNG) occurs in North America. MNG is a common disorder of the thyroid gland with multiple nodules. The macrofollicular pattern of a follicular neoplasm is indistinguishable from MNG Cytologically, a benign follicular nodule shows mainly macrofollicles, low/moderate cellularity, cohesive follicular cells with uniformity and even spacing, coarse chromatin, as well as colloid, which may be copious. The differential diagnosis of benign follicular nodules includes suspicious follicular nodule, suspicious Hürthle cell nodule, and papillary carcinoma. In some benign nodules, focal cytologic atypia may be present and becomes a diagnostic challenge. Of note, Hürthle cell metaplasia is often seen in MNG and the cells may show at places marked nuclear atypia. Hashimoto or chronic lymphocytic thyroiditis is characterized by a
## Thyroid gland disease
The introduction of synoptic reports has been crucial in improving healthcare worldwide, allowing a common platform for adequately assessing patients with neoplastic disease. Thyroid cases receive an adequacy interpretation, a first interpretation using "The Bethesda System for Reporting Thyroid Cytopathology" (TBSRTC), and a free text explanatory diagnosis with a comment according to the pathologist or cytopathologist. It is also encouraged that the cytopathologist or the pathologist liaise with the clinical team in case an FNAC may indicate suspicious malignant cells or some distinct cancerous cells. TBSRTC categories include insufficient for diagnosis, benign, atypical cells of undetermined significance, suspicious for a follicular neoplasm, suspicious for a Hürthle cell neoplasm, suspicious for malignancy (papillary carcinoma, medullary carcinoma, lymphoma, metastatic tumor, other), and malignant. Despite iodine supplementation, multinodular goiter (MNG) occurs in North America. MNG is a common disorder of the thyroid gland with multiple nodules. The macrofollicular pattern of a follicular neoplasm is indistinguishable from MNG Cytologically, a benign follicular nodule shows mainly macrofollicles, low/moderate cellularity, cohesive follicular cells with uniformity and even spacing, coarse chromatin, as well as colloid, which may be copious. The differential diagnosis of benign follicular nodules includes suspicious follicular nodule, suspicious Hürthle cell nodule, and papillary carcinoma. In some benign nodules, focal cytologic atypia may be present and becomes a diagnostic challenge. Of note, Hürthle cell metaplasia is often seen in MNG and the cells may show at places marked nuclear atypia. Hashimoto or chronic lymphocytic thyroiditis is characterized by a mixed population of lymphocytes, tingible-body macrophages, dendritic-lymphocytic clusters, Hürthle cells and scant colloid. Hyperplastic Hürthle cells nodules may occur in Hashimoto thyroiditis mimicking a Hürthle cell tumor. The differential diagnosis of Hashimoto thyroiditis includes a reactive lymph node, MNG with predominant Hürthle cell change, primary malignant lymphoma of the thyroid gland, Hürthle cell neoplasm, and papillary thyroid carcinoma. A subacute thyroiditis and Riedel disease are not a common event in pediatric age or youth. The category of suspicious for a follicular neoplasm has some cytologic features including marked cellularity, scant colloid, predominant microfollicles or epithelial trabeculae, and enlarged, crowded follicular cells (differently from the uniformity and even spacing of the benign lesions). In a pediatric setting, a suspicious for a follicular neoplasm poses a differential diagnosis of a benign follicular nodule, papillary thyroid carcinoma, and parathyroid adenoma or carcinoma. Other malignant masses have other distinguishing features in childhood and youth. A suspicious for Hürthle cell neoplasm diagnosis has cytologic features showing an almost complete cell population of Hürthle cells, usually dyshesive cells with prominent nucleolus, and pseudopsammoma bodies. Cytological features of papillary thyroid carcinoma include sheets, microfollicles, and papillae, although the nuclear changes need to be searched carefully. The nuclear changes include grooves, pseudoinclusions, "powdery" chromatin, and thickened and irregular nuclear membrane. There is also nuclear crowding or molding. Cells show a variable amount of cytoplasm and psammoma bodies may be present. The differential diagnosis of papillary carcinoma includes a benign follicular nodule, a follicular neoplasm, Hashimoto thyroiditis, radiotherapy effect, and a hyalinizing trabecular tumor. Although the anaplastic carcinoma does not play any role in childhood or youth, an important entity to keep in mind is the medullary carcinoma. Cytologically, this endocrine parafollicular cells derived tumor shows numerous isolated cells, loose aggregates of cells with epithelioid, plasmacytoid, or spindle-shaped morphology, nuclear changes (rounding or elongation, granularity of the chromatin, inconspicuous nucleolus, pseudoinclusions, and multiple nuclei), cytoplasmic changes (red granules) on Diff-Quik stained preparations, and amyloid. Although rare, two primary malignant lymphomas of the thyroid gland need to be kept in mind and include the diffuse large B cell lymphoma and the extra-nodal marginal zone B-cell lymphoma.
## Salivary gland disease
Salivary gland lesions are rare in children, although an increase of salivary gland disease in adolescence and youth has been reported recently. Previous research has shown that salivary gland tumors are rare in the young population, but the incidence of all primary salivary gland carcinomas increased with increasing patient age, particularly in patients younger than 30 years. Lesions of the major cephalic salivary glands, with the exception of mumps and cytomegaly, are unusual in children and adolescents, but these lesions may give rise to some different tentative diagnoses. Since malignant salivary gland tumors are relatively more frequently recognized in younger patients, a safe approach is advisable, and intradepartmental consultation is crucial. Cytologic diagnoses of malignant tumors are confirmed histologically in 93% of cases, while benign tumor diagnoses are confirmed on histology in 95% of cases. Inflammatory lesions are confirmed on histology in 73%, while benign salivary gland tissue is ascertained as such histologically in 18% of cases. Non-rare malignant salivary gland tumors in childhood and youth are a mucoepidermoid carcinoma, adenoid cystic carcinoma, acinic cell carcinoma, malignant lymphomas, and metastatic tumors from a tumor of the H&N region or paraganglia (e.g., neuroblastoma). Benign neoplasms include pleomorphic adenoma and Warthin's tumor. In childhood, 80-90% of all malignant salivary gland tumors are constituted by mucoepidermoid carcinomas, adenoid-cystic carcinomas, and acinic cell carcinomas. (e,f) Mucoepidermoid carcinoma (Papanicolaou stain, ×400 and Papanicolaou stain, ×400, e and f, respectively). The mucin in the background and the nuclear detail are supportive of a diagnosis of mucoepidermoid carcinoma. (g,h) Adenoid cystic carcinoma (Diff-Quik, ×100 and ×100, g and h, respectively). For the details of the single salivary gland tumors, please refer to the text.
In adults, the corresponding figure is only 45%. Despite the low incidence of 15-25% of malignant neoplasms for adults identified in 1969, in a study carried out on 2632 patients following teams and investigations reported a significantly higher relative proportion in young patients. Acinic cell carcinoma is characterized by a hypercellular aspirate with a clean background with tumor cells seen in disorganized clusters with loss of round groupings and lack of an associated ductal epithelium and on higher magnification cells show some uniformity resembling normal serous acinar cells with the cytoplasm being foamy or bubbly and harboring fine dark granules. The examination of the background highlights many naked nuclei. Apart from clear cell tumors, mainly low-grade mucoepidermoid carcinoma and epithelial-myoepithelial carcinoma), oncocytic tumors, sialadenosis, and normal salivary gland tissue need to be kept in the differential diagnosis list. Smears of mucoepidermoid carcinoma are usually low in cellularity with a particularly striking dirty background of mucin and debris. In the smears, the presence of scattered cell clusters of intermediate cells with overlapping epithelial groups, some mucin-coated cells (goblet-cells-like), and few squamous epithelial cells are particularly evident features for the diagnosis of mucoepidermoid carcinoma(e,f) Mucoepidermoid carcinoma (Papanicolaou stain, ×400 and Papanicolaou stain, ×400, e and f, respectively). The mucin in the background and the nuclear detail are supportive of a diagnosis of mucoepidermoid carcinoma. (g,h) Adenoid cystic carcinoma (Diff-Quik, ×100 and ×100, g and h, respectively). For the details of the single salivary gland tumors, please refer to the text.
In adults, the corresponding figure is only 45%. Despite the low incidence of 15-25% of malignant neoplasms for adults identified in 1969, in a study carried out on 2632 patients following teams and investigations reported a significantly higher relative proportion in young patients. Acinic cell carcinoma is characterized by a hypercellular aspirate with a clean background with tumor cells seen in disorganized clusters with loss of round groupings and lack of an associated ductal epithelium and on higher magnification cells show some uniformity resembling normal serous acinar cells with the cytoplasm being foamy or bubbly and harboring fine dark granules. The examination of the background highlights many naked nuclei. Apart from clear cell tumors, mainly low-grade mucoepidermoid carcinoma and epithelial-myoepithelial carcinoma), oncocytic tumors, sialadenosis, and normal salivary gland tissue need to be kept in the differential diagnosis list. Smears of mucoepidermoid carcinoma are usually low in cellularity with a particularly striking dirty background of mucin and debris. In the smears, the presence of scattered cell clusters of intermediate cells with overlapping epithelial groups, some mucin-coated cells (goblet-cells-like), and few squamous epithelial cells are particularly evident features for the diagnosis of mucoepidermoid carcinoma. However, both components may be quite bland or missing in one or all slides. According to the grade of differentiation, nuclear features may vary from bland to hyperchromasia with a high nucleus to cytoplasm ratio, although this tumor is usually low-grade in childhood and adolescence. Some potential pitfalls in diagnosing the mucoepidermoid carcinoma are some cystic nature of these tumors, which may harvest only hypocellular (even acellular in some cases) mucoid material. Moreover, extracellular mucin is typically copious, mimicking the fibrillary stroma seen in pleomorphic adenomas, although the mucin of pleomorphic adenomas stains less intensely and has no fibrillary pattern. Finally, squamous metaplasia is not only characteristic of mucoepidermoid carcinoma but may be found in other tumors, such as pleomorphic adenoma and Warthin's tumor. The adenoid cystic carcinoma features are large globules of extracellular matrix with or without surrounding basaloid cells. In FNAs with a predominance of basaloid tumor cells, both benign and malignant salivary gland tumors of epithelial-myoepithelial differentiation should be considered in the differential diagnosis. If TBSRTC has received a warming acceptance in many countries, the Milan System for Reporting Salivary Gland Cytology (MSRSGC) seems that will have similar success in the coming years. The MSRSGC is an evidence-based tiered classification system that comprises six diagnostic categories associated with an average risk of malignancy (ROM) and clinical management strategies. It is expected that the MSRSGC will improve communication between the cytopathologist and the treating pediatrician, facilitating cytologic-histologic correlation, and lead to overall improved patient care.
## Study comparison
FNAC is now being considered as a valuable diagnostic aid in many children's hospitals across continents, not only in North America. In most studies, FNAC reveals good aspects, showing the early availability of results, its simplicity, minimal trauma or injury, and the low rate of complications. Ancillary techniques, including flow cytometry, cytogenetics, immunohistochemistry on the cell block, and electron microscopy can be easily applied for the characterization of tumors. Furthermore, FNA does not usually need heavy sedation, and general anesthesia is not necessary. FNAC of the H&N region is a well-accepted technique that has high specificity in most studies. In, a comparison of four studies is provided.
# Limitations
Most of the studies have several limitations that are inherent in their retrospective nature. First, the lack of standardization of the enrolled patients is reflective of both the diversity of pediatric nodules and the variable diagnostic and treatment approaches of different clinical and surgical teams. In fact, there is often a significant variation in the use, duration, and timing of antibiotic treatment, and the accessibility of the clinical and surgical services was also variable and change over time was present in most of the retrospective studies. The lack of universal follow-up is also an essential drawback of numerous FNAC studies. The importance of the monitoring may be enshrined for all patients with negative FNAC results that need to be instructed to follow up if the mass persists, increases in size, or any concerns remain.
# Conclusions
The potential avoidance of surgery with associated scarring, early and late septic and non-septic surgical complications, general anesthetic risk, recovery time, and significant expense bills have all been heralded as benefits of FNAC, especially given the high prevalence of non-neoplastic pediatric tumor and masses of the H&N region. FNAC shows an impressive diagnostic accuracy in numerous studies. FNAC is supported by an active lab safety profile and carries a very limited number of drawbacks considering that in the case of a dubious cytology result, a surgery can be planned in a very short time identifying FNAC as a useful triage procedure. In conclusion, FNAC is a safe, well-tolerated, and accurate tool for diagnosing pediatric masses of thyroid and non-thyroid origin in childhood and youth. FNAC plays a vital role in giving comfort to obviate the need for unnecessary immediate surgery in benign nodules (congenital/reactive/inflammatory) and plan a primary surgical intervention in malignant lesions. In consideration of significant health care changes worldwide, we would like to strongly emphasize that when such a triage of pediatric masses of the H&N region is handled, there is a potential decrease in both surgical morbidity and health care costs. |
Efficacy and safety of asenapine in Asian patients with an acute exacerbation of schizophrenia: a multicentre, randomized, double-blind, 6-week, placebo-controlled study
Rationale Asenapine is a second generation anti-psychotic approved in the USA in 2009 for the treatment of schizophrenia, but its efficacy has not been proven in Asian patients.Objectives The objectives of this study are to evaluate the efficacy and tolerability of asenapine in Asian patients experiencing an acute exacerbation of schizophrenia. Methods In this prospective, double-blind study, patients in Japan, Korea, and Taiwan were randomized (1:1:1) to asenapine 5 mg twice daily (bid), 10 mg bid or placebo for 6 weeks after a 3-to 7-day washout/screening period. The primary endpoint was the mean change in the positive and negative syndrome scale (PANSS) total score from baseline to day 42/treatment end. Results Of the 532 participants randomized, 530 received treatment. The primary endpoint was significantly greater with asenapine 5 and 10 mg bid than with placebo (−12.24 and −14.17 vs. −0.95; p < 0.0001). The results of secondary endpoints including PANSS negative subscale scores and PANSS responders at the end of treatment supported the results of the primary endpoint. There were no significant differences in the incidence of treatment-emergent adverse events reported with asenapine 5 and 10 mg bid and placebo (84.6, 80.7, and 81.6 %). There was a mean (± standard deviation) change in weight of −1.76 ± 2.45 kg for placebo, +0.42 ± 2.65 kg for asenapine 5 mg bid, and +0.81 ± 2.89 kg for asenapine 10 mg bid group. Conclusions Asenapine was effective and generally well tolerated when used for the treatment of acute exacerbations of schizophrenia in Asian patients.
# Introduction
Schizophrenia is a complex psychiatric disorder associated with variable degrees of functional impairment and social disability [bib_ref] Current therapy issues and unmet clinical needs in the treatment of schizophrenia:..., Lublin [/bib_ref] [bib_ref] Schizophrenia, Bjust the facts^4. Clinical features and conceptualization, Tandon [/bib_ref]. It is a chronic relapsing disorder, with recurrent exacerbations of positive symptoms against a background of persistent negative symptoms, cognitive dysfunction, and depressive symptoms [bib_ref] Current therapy issues and unmet clinical needs in the treatment of schizophrenia:..., Lublin [/bib_ref] [bib_ref] Schizophrenia, Bjust the facts^4. Clinical features and conceptualization, Tandon [/bib_ref].
Anti-psychotic medication is the mainstay of treatment for schizophrenia. Historically, pharmacological treatments have focused on dopamine D 2 receptor blockade to control the positive symptoms of schizophrenia, while negative and cognitive symptoms often persist . Second generation anti-psychotics, which also inhibit a range of other receptors, have been developed, but they vary in their effectiveness across mood domains [bib_ref] Current therapy issues and unmet clinical needs in the treatment of schizophrenia:..., Lublin [/bib_ref] [bib_ref] Unmet treatment needs in schizophrenia patients: is asenapine a potential therapeutic option?, Pompili [/bib_ref]. In addition, second generation anti-psychotics have differing levels of risk for weight gain and hyperprolactinemia [bib_ref] Body weight and metabolic adverse effects of asenapine, iloperidone, lurasidone and paliperidone..., Hert [/bib_ref] [bib_ref] Schizophrenia, Bjust the facts^5. Treatment and prevention. Past, present, and future, Tandon [/bib_ref] , which can negatively affect treatment adherence [bib_ref] Effectiveness of antipsychotic drugs in patients with chronic schizophrenia, Lieberman [/bib_ref].
Asenapine is a second generation anti-psychotic with a unique receptor-binding profile that has been approved in the USA for the treatment of schizophrenia [bib_ref] Asenapine review, part I: chemistry, receptor affinity profile, pharmacokinetics and metabolism, Citrome [/bib_ref] [bib_ref] Asenapine: a novel psychopharmacologic agent with a unique human receptor signature, Shahid [/bib_ref]. It has antagonistic activity at dopaminergic, serotonergic, α-adrenergic, and histaminic receptors but no appreciable affinity for muscarinic receptors [bib_ref] Asenapine: a novel psychopharmacologic agent with a unique human receptor signature, Shahid [/bib_ref]. The efficacy and safety of asenapine in schizophrenia has been established in a global clinical trial program [bib_ref] Asenapine review, part II: clinical efficacy, safety and tolerability, Citrome [/bib_ref]. In two 6-week fixed-dose trials, asenapine 5 mg twice daily (bid) was more effective than placebo in improving positive and negative syndrome scale (PANSS) total scores [bib_ref] The positive and negative syndrome scale (PANSS) for schizophrenia, Kay [/bib_ref] in patients with acute exacerbations of schizophrenia, with only modest effects on weight and metabolic variables [bib_ref] Efficacy and safety of asenapine in a placebo-and haloperidol-controlled trial in patients..., Kane [/bib_ref] [bib_ref] Efficacy and tolerability of asenapine in acute schizophrenia: a placebo-and risperidonecontrolled trial, Potkin [/bib_ref].
In a dedicated study, the pharmacokinetic and safety profiles of asenapine were similar in healthy Japanese and Caucasian subjects (Merck, meaning no dosage adjustment is required on the basis of race. However, generally Asian subjects account for only a small percentage of the total populations enrolled in the phase II and III clinical studies of asenapine, and studies conducted in specific ethnic populations are still needed. Therefore, the study described herein was conducted to confirm the efficacy and safety of asenapine in Asian patients experiencing an acute exacerbation of schizophrenia.
# Methods
## Patients
The study was conducted from May 2010 to April 2014 at 112 centers: Japan, 81 sites; Taiwan, 15 sites; and Korea, 16 sites. Male and female patients aged 20-64 years with a Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) (American Psychiatric Association 2000) diagnosis of schizophrenia of paranoid (295.30), disorganized (295.10), catatonic (295.20), or undifferentiated (295.90) subtypes were eligible for randomization. The current acute exacerbation of schizophrenia had to be of ≤2 months duration, with symptoms that represented a dramatic and substantial change from the state prior to the exacerbation, and there was a requirement for a change in medication or dosage to treat new or worsened positive symptoms.
Other key inclusion criteria were a PANSS total score ≥60, with scores of ≥4 in two or more of five items on the PANSS positive subscale (delusions, conceptual disorganization, hallucinatory behavior, grandiosity, suspiciousness/persecution) at the initial screening assessment and at baseline, and a score of ≥4 on the clinical global impressions-severity of illness (CGI-S) scale [bib_ref] Addressing side effects from antipsychotic treatment in schizophrenia, Guy [/bib_ref] at baseline. Patients who had received previous anti-psychotic medication for a prior episode of acute exacerbation of schizophrenia were required to have had a positive response. The use of all prohibited concomitant medications (anti-psychotics, anti-depressants, mood stabilizers, anti-epileptics, monoamine oxidase inhibitors, St. John's Wort, anti-emetics that are dopamine antagonists, and traditional herbal medication) had to be discontinued, with the last dose taken no later than the evening prior to the baseline visit, for inclusion in the trial (for depot neuroleptic, discontinuation must have occurred more than 3 months prior to randomization).
Patients with a diagnosis of schizoaffective disorder (295.70), schizophrenia of residual subtype (295.60), schizophreniform disorder (295.40), or a psychiatric disorder other than schizophrenia were excluded from the study. Women who were pregnant were ineligible for inclusion in the trial. Patients were also excluded if they had taken any experimental medication within 12 weeks before baseline, were defined as having treatment-refractory schizophrenia, had received treatment with ≥3 anti-psychotic drugs within the previous month, or had an uncontrolled, unstable clinically significant medical condition (e.g., renal, endocrine, hepatic, respiratory, cardiovascular, hematologic, immunologic or cerebrovascular disease, or malignancy) or abnormal laboratory, vital sign, physical examination, or electrocardiogram (ECG) findings at screening. Other exclusion criteria included the following: current (past 6 months) substance abuse; a body mass index (BMI) <16.0 or >35.0 at baseline; a diagnosis of Parkinson's disease; a history of or current treatment for narrow angle glaucoma; a history of any seizure disorder beyond childhood; a history of allergy or sensitivity to drugs such as psychotropics and anti-psychotics; a ≥20 % decrease in PANSS total score from screening to baseline; imminent risk of self-harm or harm to others; current involuntary inpatient confinement; known diagnosis of borderline personality disorder, mental retardation, or organic brain disorder; and previous treatment with asenapine.
Patients classed as treatment-refractory (has been treated with at least two different atypical anti-psychotic agents at dosages equivalent to or greater than 600 mg/day of chlorpromazine [12 mg/day of haloperidol] for more than 4 weeks, each without clinical response, or has received clozapine for 12 weeks immediately preceding the screening) and those who received treatment with three or more anti-psychotic drugs, or dose equivalents higher than 18 mg/day of haloperidol (equivalent 900 mg/day of chlorpromazine) within 1 month prior to randomization were also excluded.
A number of measures were implemented to minimize the risks to patients randomized to placebo: hospitalization for a minimum 3 weeks, with the possibility of extending the hospitalization to 6 weeks if the patient was not clinically stable enough to be discharged; regular assessment of symptom severity; requirement that a caregiver oversees the patient's compliance with trial medication during outpatient dosing; allowance of concomitant benzodiazepines to treat agitation and anxiety; and exclusion of high-risk subjects including those that are actively suicidal, homicidal, under involuntary commitment, or have a recent history of aggressive behavior.
## Randomization and treatment
The study consisted of a 3-to 7-day washout/screening period, a 6-week double-blind treatment phase, and a follow-up period. Placebo was administered single blind for the duration of the screening period, and patients were tapered off their prestudy anti-psychotic medication, anti-depressant medication, and any anti-parkinsonian drugs to treat extrapyramidal symptoms (EPS).
After the baseline assessments were completed, patients considered eligible by an investigator were randomized (1:1:1) to receive sublingual asenapine 5 mg bid, 10 mg bid, or placebo. The allocation of patients was performed using permuted-block randomization (block size of 6 and allocation ratio of 1:1:1) using SAS. Participants and investigators remained blinded to treatment assignment, in accordance with the doubleblind study design. The asenapine and placebo tablets were identical in appearance and characteristics (i.e., all tablets were fast dissolving) and had identical packaging. During the treatment period, participants received their allocated dosage of asenapine or placebo twice a day every day for 6 weeks. Tablets were administered sublingually without water.
Participants were hospitalized during screening and for the first 3 weeks of the treatment period, with the possibility of extending hospitalization to 6 weeks in the absence of a responsible caregiver to provide support and ensure compliance with study medication. Compliance during inpatient treatment was recorded by hospital staff and monitored by returned tablet counts during outpatient treatment.
Concomitant lorazepam and short-acting benzodiazepines were permitted to treat agitation and anxiety, and concomitant EPS medication was permitted if EPS worsened or appeared.
An interim analysis was built into the protocol to assess whether the trial should be discontinued (in case of lack of efficacy). In May 2012, an independent data-monitoring committee was convened to review the efficacy data for half of the planned number of participants collected up to the day of database lock . Upon review of the data, the committee concluded that continuation of the trial was appropriate.
## Efficacy measurement
The primary efficacy outcome was change in the PANSS total score from baseline to day 42 (end of treatment). Secondary efficacy outcomes included changes from baseline to end of treatment in PANSS subscale scores (positive symptoms, negative symptoms and general psychopathology), PANSS Marder factor scores (positive symptoms, negative symptoms, disorganized thought, hostility/excitement, anxiety/depression), and CGI-S scores. Additional secondary variables were the percentage of PANSS responders (those with ≥30 % decrease in the total PANSS total score from baseline to end of treatment) and the percentage of CGI-global improvement (CGI-I) responders (those who are assessed as very much improved, much improved or minimally improved by the investigator) at day 42. Post-baseline efficacy assessments were obtained on day 4 and 7 and weekly thereafter (day 14, 21, 28, 35, and 42). Efficacy assessments were performed by an investigator or a trained rater assigned by the investigator. Using the same rater for assessment of a subject throughout their participation in the study was strongly encouraged.
## Safety assessments
Key safety parameters were weight gain, body mass index (BMI), EPS, glycosylated hemoglobin (HbA1c), fasting glucose, insulin, and prolactin. For prolactin, only clinically relevant abnormal values were reported. Other safety variables were the frequency of the onset of treatment-emergent adverse events (TEAEs) and clinically relevant changes in physical findings, vital signs, electrocardiogram parameters, laboratory values, and use of anti-parkinsonian drugs.
Laboratory, ECG, and BMI assessments were made at baseline and on days 14, 28, and 42. Assessments of vital signs were conducted at baseline and on days 14, 21, 28, and 42. Physical examinations were conducted at screening and at end of treatment (day 42). The incidence, nature, and severity of AEs were recorded throughout the screening and treatment periods until 7 days after the last dose of study drug. EPS were reported as AEs and rated according to the druginduced extrapyramidal symptoms scale (DIEPSS) (Inada 2009) at baseline and on days 21 and 42.
## Sample size calculations
Based on the results of previous phase II and III clinical trials of asenapine in patients with schizophrenia [bib_ref] Evaluation of the clinical efficacy of asenapine in schizophrenia, Minassian [/bib_ref] [bib_ref] Efficacy and tolerability of asenapine in acute schizophrenia: a placebo-and risperidonecontrolled trial, Potkin [/bib_ref] , the range of the difference in the change in PANSS total score from baseline to end of treatment at 42 days between the asenapine 5 mg bid and placebo groups was between −3.4 and −9.7. Referring to the study by [bib_ref] Efficacy and safety of asenapine in a placebo-and haloperidol-controlled trial in patients..., Kane [/bib_ref] , we assumed that the change from baseline in PANSS total score would be −6, and the standard deviation (SD) of the total of change in PANSS score would be 20 (effect size = 0.3). The sample size required to show superiority of asenapine over placebo was estimated be 176 per group, with a two sided significance level of 5 and 80 % power. Therefore, approximately 530 participants were expected to be randomized.
# Statistical analysis
Several sets were defined for the analysis of outcomes. The subjects as treated (i.e., safety analysis set; All Subjects Treated, AST) consisted of all participants enrolled in the study who received at least part of one dose of the study drug, while the full analysis set (FAS) consisted of all participants in the AST who had a baseline and at least one post-baseline PANSS measurement.
The primary efficacy outcome was analyzed in the FAS using analysis of covariance (ANCOVA) with dropout or missing data imputed by using the last observation carried forward (LOCF). No patient had ≥6 missing PANSS items; for 5 or fewer missing PANSS items, the PANSS score was calculated by multiplying the total for the non-missing items by the total number of items and then dividing by the number of non-missing items.
The model included change from baseline in PANSS total score as a response variable, baseline PANSS total score as a covariate, and treatment groups and regions (Japan, Taiwan, and Korea) as explanatory variables. Pairwise comparisons for the primary efficacy outcome were performed in the following order: (1) asenapine 5 mg bid vs. placebo and (2) asenapine 10 mg bid vs. placebo. The result of the second comparison was considered only when the result of asenapine 5 mg bid vs. placebo was statistically significant. Further analysis using a mixed model for repeated measures (MMRM) was performed using change from baseline in PANSS total score at each visit from 7 days as the response variables and treatment groups, regions, visits, and treatment by visit interaction as the fixed effect variables and baseline PANSS total score as the covariates. Changes from baseline to end of treatment in selected secondary efficacy outcomes (PANSS subscale, Marder factor, and CGI-S scores) were also assessed using ANCOVA with LOCF. The MMRM method was also performed for PANSS subscale and Marder factor scores. PANSS and CGI-I responder rates in the placebo versus each asenapine group were compared using the Cochran-Mantel-Haenszel (CMH) test with an adjustment for region. Regarding the analysis of secondary endpoints, these were exploratory in nature and no adjustment for multiplicity was made. Safety variables in the AST were summarized using descriptive statistics. TEAEs and categorical variables for demographics were compared the placebo and each asenapine group using Fisher's exact test. Continuous variables for demographics were compared the placebo and each asenapine group using ANOVA. A p value less than 0.05 was considered statistically significant (two-tailed).
# Results
Of a total of 573 patients screened, 532 participants were randomized to treatment. The AST population included 530 participants (asenapine 5 mg bid: n = 175; asenapine 10 mg bid: n = 181; placebo: n = 174) and the FAS included 525 participants (asenapine 5 mg bid: n = 173; asenapine 10 mg bid: n = 178; placebo: n = 174). The flow of participants throughout the study is presented in [fig_ref] Figure 1: Patient disposition [/fig_ref]. A total of 303 participants completed treatment, with a greater percentage of those in the asenapine 5 mg bid (65 %) or 10 mg bid (60 %) groups completing the trial compared with those receiving placebo (46 %). The incidence of withdrawal due to lack of efficacy was lower for participants receiving asenapine 5 mg bid (7.4 % of treated participants) or 10 mg bid (5.0 %) than for those receiving placebo (15.5 %).
The key baseline characteristics of the AST population are presented in [fig_ref] Table 1: Baseline characteristics and demographics [/fig_ref]. No meaningful differences in baseline demographics or clinical characteristics including PANSS total scores were noted between the treatment groups. Of the concomitantly used medications, all the groups (71.3 % in the placebo, 62.9 % in the asenapine 5 mg bid, and 70.2 % in the asenapine 10 mg bid. .06) in the asenapine 5 mg bid, asenapine 10 mg bid, and placebo groups, respectively. The improvements from baseline in PANSS total score were significantly greater in participants receiving asenapine 5 mg bid or asenapine 10 mg bid, compared with placebo from days 14 and 7, respectively. Overall, the efficacy profile of the asenapine 5 and 10 mg groups were similar. Analysis of the change in PANSS total score from baseline over time using MMRM showed that improvements from baseline in PANSS total score were significantly larger in the asenapine 5 and 10 mg bid groups compared with placebo from day 14 and 7, respectively (p < 0.05) and were sustained through to day 42 . [fig_ref] Table 3: Responder rates for change in positive and negative syndrome scale [/fig_ref] describes the proportion of responders using different thresholds to define response (i.e., ≥20, ≥30, ≥40, and ≥50 % decrease in PANSS total score).
## Secondary efficacy outcomes
Positive and negative syndrome scale subscale scores and responders
Changes in the PANSS subscale scores and PANSS Marder factor scores supported the results of the primary efficacy outcome analysis , whereby significantly more participants were classified as PANSS responders (≥30 % decrease in score) at the end of treatment in the asenapine 5 mg bid (p = 0.0001) and 10 mg bid (p < 0.0001) groups compared with the placebo group. This was evident for PANSS positive, negative, and general scores and all PANSS Marder subscores.
## Clinical global impressions-severity of illness scores and responders
CGI-S scores also supported the results of the primary efficacy outcome analysis. The percentage of CGI-I responders at end of treatment was significantly higher in both asenapine treatment groups than in the placebo group (both p < 0.0001).
## Safety and tolerability
There were no differences in the overall incidence of TEAEs across the treatment groups; TEAEs were reported in 81.6 % of participants receiving placebo, 84.6 % of participants receiving asenapine 5 mg bid, and 80.7 % of participants receiving asenapine 10 mg bid. Serious adverse events (SAEs) were reported in 5.7 and 2.8 % of participants in the asenapine 5 mg bid and 10 mg bid groups, and in 7.5 % of those receiving placebo. No SAE occurred in more than one subject, apart from aggravated schizophrenia, which occurred in 4.6 and 2.2 % of participants receiving asenapine 5 mg bid and 10 mg bid, respectively, and 4.6 % of the placebo group. As per withdrawals for lack of efficacy, the incidence of withdrawal due to adverse events (AEs) was lower for those receiving asenapine 5 mg bid (17.1 % of treated participants) or 10 mg bid (17.7 %), compared with those receiving placebo (24.7 %).
## Disease exacerbation and extrapyramidal symptoms
TEAEs occurring in ≥5 % of participants in any treatment group are shown in [fig_ref] Table 4: Incidence of treatmentemergent adverse events occurring in ≥5 % of patients in... [/fig_ref]. The most commonly reported TEAEs for asenapine were aggravated schizophrenia, akathisia, oral hypoesthesia, and somnolence. As shown in [fig_ref] Table 4: Incidence of treatmentemergent adverse events occurring in ≥5 % of patients in... [/fig_ref] , aggravated schizophrenia occurred in at least 10 % of subjects in each treatment group but was less frequent in the asenapine groups than in the placebo group. In contrast, rates of EPS and akathisia were higher in participants receiving asenapine 5 mg (5.1 and 11.4 %) and 10 mg (7.7 and 10.5 %) bid than in those receiving placebo (1.7 and 5.2 %). The incidence rates of dizziness and increase in blood creatine phosphokinase were higher in the asenapine 10 group, but there were no differences in the other TEAEs between the treatment groups [fig_ref] Table 4: Incidence of treatmentemergent adverse events occurring in ≥5 % of patients in... [/fig_ref].
## Weight
There was a mean (± standard deviation [SD]) change in weight of −1.76 ± 2.45 kg for placebo, +0.42 ± 2.65 kg for asenapine 5 mg bid, and +0.81 ± 2.89 kg for asenapine 10 mg bid. Clinically significant weight gain (≥7 % of baseline body weight) was reported in 4.7 to 7.3 % of participants in the asenapine groups and none in the placebo group. BMI decreased by 0.66 ± 0.91 kg/m 2 with placebo and increased by 0.16 ± 1.03 and 0.32 ± 1.09 kg/m 2 with asenapine 5 mg bid and 10 mg bid.
## Laboratory values and vital signs
There were no clinically significant differences in laboratory values and vital signs between the treatment groups. Mean ± SD changes from baseline in insulin were 2.37 ± 14.08, 2.24 ± 13.78, and 0.22 ± 20.34 μIU/mL with asenapine 5 mg bid and 10 mg bid and placebo, respectively; mean ± SD changes in fasting glucose were −0.03 ± 0.85, 0.11 ± 0.77, and 0.18 ± 0.97 mmol/L; and mean ± SD changes in HbA1c were −0.01 % ± 0.41 %, −0.01 ± 0.49 % and −0.04 ± 0.45 %. Mean ± SD changes from baseline in prolactin were −17.92 ± 45.31, −13.27 ± 43.93 and −27.79 ± 46.25 μg/L with asenapine 5 mg bid and 10 mg bid and placebo, respectively. The percentage of participants who had normal prolactin levels at baseline but levels above the reference range at the end of treatment (day 42) was higher in those receiving asenapine 5 mg bid (27.9 %) and asenapine 10 mg bid (31.7 %) than in those receiving placebo (10.3 %). No adverse event related to QT prolongation (SMQ) was observed during the monitoring of the current trial; however, the data are not shown.
## Stratified analyses
Given that the study was performed over a number of different sites and countries, the potential for a country-by-treatment interaction was examined, but no significant influences were found (i.e., results were p > 0.05). Furthermore, stratified analyses were performed for sex, disease duration, and schizophrenia subtypes; the results of each indicated that there were no demographic or disease subtype factors that influenced the overall results.
# Discussion
In this randomized, double-blind clinical trial, asenapine 5 and 10 mg bid was significantly more effective than placebo in treating acute schizophrenia in Asian patients, as demonstrated by statistically significant improvements from baseline in PANSS total scores. In addition, asenapine 5 and 10 mg bid was effective in controlling both positive and negative symptoms, providing statistically significant improvements from baseline in PANSS positive and negative scores compared with placebo (by 11-13 points), as well as significant improvements in general scores and PANSS Marder 5 factor scores. Moreover, the asenapine treatment groups showed significant improvements versus placebo in other secondary efficacy variables (though it should be noted that these secondary variables were only part of an exploratory analysis); overall the efficacy profiles of the asenapine 5 and 10 mg groups appeared to be similar. Asenapine was safe and well tolerated, with minimal impact on body weight and no notable effects on other metabolic parameters. The incidence of discontinuation because of AEs or lack of efficacy was greater in the placebo group than in the asenapine treatment groups. With the exception of elevated blood creatine phosphokinase and increased dizziness, there were no differences between the 5 and 10 mg groups for TEAEs.
Results of the current study, conducted in Asian countries, complement the robustly positive results of two similarly designed 6-week studies conducted in North America and Europe in patients with an acute exacerbation of schizophrenia [bib_ref] Efficacy and safety of asenapine in a placebo-and haloperidol-controlled trial in patients..., Kane [/bib_ref] [bib_ref] Efficacy and tolerability of asenapine in acute schizophrenia: a placebo-and risperidonecontrolled trial, Potkin [/bib_ref]. While the study by found significantly greater reductions in mean PANSS total scores, no significant difference from placebo was observed in negative scores, or hostility/excitement subscores [bib_ref] Efficacy and safety of asenapine in a placebo-and haloperidol-controlled trial in patients..., Kane [/bib_ref]. However, in our study, asenapine significantly reduced the scores on the negative factor and hostility/ excitement factor, suggesting that asenapine has efficacy across a broad range of symptoms in schizophrenia. Two other short-term trials have been conducted with asenapine in the 5-10 mg bid dose range: one was negative (active control, but not asenapine, separated from placebo) and one failed (neither asenapine nor active control separated from placebo) [bib_ref] Asenapine review, part II: clinical efficacy, safety and tolerability, Citrome [/bib_ref] [bib_ref] Meta-analyses of the efficacy of asenapine for acute schizophrenia: comparisons with placebo..., Szegedi [/bib_ref]. A meta-analysis of pooled data from all four trials found asenapine to be superior to placebo with regard to mean change in PANSS total score [bib_ref] Meta-analyses of the efficacy of asenapine for acute schizophrenia: comparisons with placebo..., Szegedi [/bib_ref]. Both asenapine doses were safe and well tolerated, with minimal effects on weight and no notable effects on metabolic parameters, in all studies.
In one of the previous asenapine studies, a high placebo response rate was reported; the authors suggested that this may have contributed to the finding that asenapine 10 mg bid did not have an advantage over placebo in that trial [bib_ref] Efficacy and safety of asenapine in a placebo-and haloperidol-controlled trial in patients..., Kane [/bib_ref]. Low effect sizes versus placebo are also illustrated in the previously mentioned failed and negative studies, reflecting the finding that placebo response tended to be larger than historically expected [bib_ref] Meta-analyses of the efficacy of asenapine for acute schizophrenia: comparisons with placebo..., Szegedi [/bib_ref]. High placebo response rates underscore the need to demonstrate that active treatment can separate from placebo [bib_ref] Efficacy and safety of asenapine in a placebo-and haloperidol-controlled trial in patients..., Kane [/bib_ref]. In order to exclude initial placebo responders in the present study, oral placebo tablets were administered during the screening period. Restriction of the duration of the current acute exacerbation of schizophrenia to ≤2 months may have contributed to the low placebo response rate and efficacy separation of the asenapine treatment groups from placebo seen in this trial. In the present trial, asenapine 5 mg bid and 10 mg bid were both found to be significantly effective in reducing PANSS scores compared with placebo, which is clearly in line with the previous pooled analysis of placebo-controlled trials of asenapine in acute schizophrenia showing that the two doses had similar efficacy [bib_ref] Modeling and simulation of the time course of asenapine exposure response and..., Friberg [/bib_ref].
The results of the present study demonstrate that asenapine is well tolerated, with rates of SAEs and withdrawals due to AEs lower in both active treatment groups compared with placebo. The AE profile of asenapine seen in this study was as expected; reviews of the literature show that asenapine treatment is associated with EPS and akathisia, increases in prolactin versus placebo, oral hypoesthesia, and weight gain (a known class effect) [bib_ref] Asenapine review, part II: clinical efficacy, safety and tolerability, Citrome [/bib_ref] , and the results reported herein confirm that asenapine recipients experience these AEs more frequently than those receiving placebo. However, other active controlled studies suggest that the rates of EPS and akathisia vary depending on the drug, and that asenapine induces these events at a lower rate than some anti-psychotics [bib_ref] Asenapine review, part II: clinical efficacy, safety and tolerability, Citrome [/bib_ref]. Similarly, the effect of asenapine on weight has been reported as modest compared with other antipsychotics such as olanzapine [bib_ref] Asenapine review, part II: clinical efficacy, safety and tolerability, Citrome [/bib_ref]. In the present study, only modest weight gain was observed in the asenapine group. Asenapine's unique functional activities at diverse neurotransmitter receptors, including various serotonin receptors, may contribute to the relatively benign effects on body weight [bib_ref] The preclinical profile of asenapine: clinical relevance for the treatment of schizophrenia..., Tarazi [/bib_ref].
Limitations of this study include the absence of an active control arm and ethical questions regarding the use of placebo in patients with acute schizophrenia. Another limitation of this study was the reliance on pill counts to monitor adherence rather than more reliable measures, such as measurement of blood drug levels.
In conclusion, results of this double-blind, placebocontrolled 6-week study showed that asenapine 5 mg bid and 10 mg bid are effective and well tolerated in the treatment of Asian patients with an acute exacerbation of schizophrenia.
[fig] Figure 1: Patient disposition: numbers of patients who were screened, randomized to treatment, received treatment, and completed treatment, with reasons for discontinuation shown [/fig]
[fig] Figure 3, Figure 3: Secondary efficacy outcomes: changes from baseline in a PANSS positive symptom factor, b PANSS negative symptom factor, c PANSS general psychopathology score, d PANSS Marder positive symptom factor, e PANSS Marder negative symptom factor, f PANSS Marder disorganized thought factor, g PANSS Marder Hostility/Excitement Factor, and h PANSS Marder anxiety/depression factor. BL baseline, End end of treatment, LSM least squares mean, SD standard deviation. *p < 0.05; **p < 0.01 vs. placebo [/fig]
[table] Table 1: Baseline characteristics and demographics (all patients as treated population) [/table]
[table] Table 3: Responder rates for change in positive and negative syndrome scale (PANSS) total scores, across different threshold values. Responder categories are separated by set limits for the magnitude of decrease in the total PANSS score from baseline to day 42/end of treatment bid twice daily, CI confidence interval [/table]
[table] Table 4: Incidence of treatmentemergent adverse events occurring in ≥5 % of patients in at least one treatment group (all patients as treated population)Comparison between the 5 and 10 mg groups a Fisher's exact test [/table]
|
Fluid Therapy and Acute Respiratory Distress Syndrome
# Introduction
Acute respiratory distress syndrome (ARDS) is a common critical illness encountered in intensive care units (ICUs). [bib_ref] National incidence rates for acute respiratory distress syndrome (ARDS) and ARDS cause-specific..., Eworuke [/bib_ref] ARDS is a heterogenous syndrome characterized by an inflammatory response of the lungs in response to an acute pathophysiologic insult. [bib_ref] Acute respiratory distress syndrome, Matthay [/bib_ref] The optimal fluid management for acute respiratory distress syndrome (ARDS) is unknown. There are risks and benefits to liberal and conservative fluid management strategies. Studies have shown that liberal fluid management may be more harmful in ARDS patients by increasing pulmonary edema and prolonging mechanical ventilation days and intensive care unit and hospital stay. Conservative fluid management has a risk of increasing nonpulmonary end organ damage. Studies suggest preventing fluid overload may lead to improved outcomes, although no prospective randomized controlled trial has shown mortality benefit to date. Different phenotypes of ARDS may respond differently to fluid management. Recent research suggests that hypoinflammatory and hyperinflammatory phenotypes may differ in their fluid responsiveness and may be helpful in determining optimum volume status. The heterogeneity of treatment effect raises concerns for bedside application of appropriate management. Future studies further refining ARDS phenotypes and their associated differential responses to fluid administration may help guide optimal fluid management strategies in ARDS.
The acute inflammatory response damages the microvascular endothelium and alveolar epithelium of the alveolar-capillary barrier, leading to increased vascular permeability and subsequent edema. ARDS was first described in 1967 as hypoxemia in the setting of bilateral pulmonary opacities on chest radiograph not attributable to cardiac failure. [bib_ref] Acute respiratory distress in adults, Ashbaugh [/bib_ref] In 1994, the American-European Consensus Conference (AECC) formally defined ARDS and acute lung injury (ALI). [bib_ref] The American-European Consensus Conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical..., Bernard [/bib_ref] In 2012, the Berlin Definition redefined ARDS using 3 categories (mild, moderate, and severe) to classify patients based on the degree of their hypoxemia. [bib_ref] Acute respiratory distress syndrome: the Berlin definition, Ards Definition Task Force [/bib_ref] The prevalence of ARDS is 5 to 35 cases per 100,000 individuals annually in the United States, and the incidence continues to rise. [bib_ref] National incidence rates for acute respiratory distress syndrome (ARDS) and ARDS cause-specific..., Eworuke [/bib_ref] The mortality rate ranges from 30% to 50%, although there is a wide variability depending on multiple factors, including patient risk factors, ARDS severity, and the etiology of ARDS. [bib_ref] Current incidence and outcome of the acute respiratory distress syndrome, Villar [/bib_ref] [bib_ref] Past and present ARDS mortality rates: a systematic review, Má Ca [/bib_ref] Following an acute insult such as sepsis, pneumonia, aspiration of gastric contents, or severe trauma, a dysregulated inflammatory response leads to increased lung endothelial and epithelial permeability. [bib_ref] Pathogenesis of acute respiratory distress syndrome, Huppert [/bib_ref] The pathogenesis occurs in 3 sequential phases with overlapping features: acute exudative/inflammatory phase, proliferative phase, and fibrotic phase. [bib_ref] The acute respiratory distress syndrome: pathophysiology, current clinical practice, and emerging therapies, Derwall [/bib_ref] In the acute phase, there is endothelial and epithelial injury to the alveoli and capillaries, alveolar macrophages secrete cytokines such as interleukin-1, 6, 8 and 10 (IL-1, 6, 8, and 10), and tumor necrosis factor a (TNF-a). These immunomodulatory proteins activate neutrophils to release proinflammatory molecules and stimulate the production of the extracellular matrix by fibroblasts. [bib_ref] The acute respiratory distress syndrome, Ware [/bib_ref] Alveolar-capillary permeability increases, which leads to the accumulation of protein-rich edematous fluid in the alveoli and interstitium. [bib_ref] The acute respiratory distress syndrome, Ware [/bib_ref] These acute-phase injuries decrease pulmonary compliance and increase ventilation/perfusion (V/Q) mismatch. [bib_ref] The acute respiratory distress syndrome: pathophysiology, current clinical practice, and emerging therapies, Derwall [/bib_ref] The protein-rich alveolar fluid also disrupts pulmonary oncotic forces, making the alveoli more vulnerable to increased hydrostatic pressure and the development of noncardiogenic pulmonary edema. During the proliferative phase, type II pneumocytes repopulate alveoli; alveolar edema is resolved by the active sodium and chloride transport and water channels, and protein is cleared from the small airways to restore alveolar architecture and function. [bib_ref] The acute respiratory distress syndrome, Ware [/bib_ref] A small subset of ARDS patients will progress to the fibrotic phase, which is characterized by the gradual remodeling and resolution of intra-alveolar and interstitial granulation tissue. This phase occurs inconsistently and delays functional recovery, and the presence of fibrosis is associated with increased mortality. [bib_ref] The acute respiratory distress syndrome, Ware [/bib_ref] [bib_ref] Alveolar macrophages contribute to alveolar barrier dysfunction in ventilator-induced lung injury, Frank [/bib_ref] [bib_ref] Pulmonary fibrosis correlates with outcome in adult respiratory distress syndrome. A study..., Martin [/bib_ref] This article explores the history of fluid therapy in ARDS, with a focus on liberal versus conservative fluid management strategies. It outlines the challenges and clinical application of the pertinent ARDS literature. Finally, it explores novel study designs such as latent class analysis (LCA) and machine learning to identify ARDS phenotypes.
## The history of fluid management for acute respiratory distress syndrome
The optimal fluid management strategy for ARDS is unknown. The American Thoracic Society/European Society of Intensive Care Medicine/Society of Critical Care Medicine (ATS/ESICM/SCCM) Clinical Practice Guidelines make no specific recommendation for fluid management in ARDS patients, and clinical practice varies widely. Experts have debated whether a liberal or conservative fluid management strategy improves clinical outcomes for ARDS patients for over 4 decades. [bib_ref] ARDS: the therapeutic dilemma, Hyers [/bib_ref] [bib_ref] Fluid balance during pulmonary edema. Is fluid gain a marker or a..., Schuller [/bib_ref] [bib_ref] Fluid management strategy in acute lung injury, Hudson [/bib_ref] [bib_ref] The case for and against fluid restriction and occlusion pressure reduction in..., Schuster [/bib_ref] Liberal fluid management, historically the conventional practice, does not restrict fluid administration during the resuscitative phase or actively seek to remove fluid during the deresuscitative phase. The theoretic argument for a liberal fluid management strategy is that it can increase stroke volume and thereby improve end organ perfusion and oxygen delivery. [bib_ref] The case for and against fluid restriction and occlusion pressure reduction in..., Schuster [/bib_ref] This practice pattern prevailed prior to recognition in the 1990s that fluids may worsen refractory hypoxemia in ARDS. [bib_ref] The case for and against fluid restriction and occlusion pressure reduction in..., Schuster [/bib_ref] Physician-guided early liberal resuscitative practices during that era make it difficult to quantify how much intravenous fluid was routinely given. Data from prior ARDS research provide a window into historic practice patterns. A 1987 randomized controlled trial (RCT) of ARDS patients that allowed for provider practice variation in resuscitative/deresuscitative practices observed 14-day fluid balances ranging from 5 L to 20 L. [bib_ref] Fluid balance and the adult respiratory distress syndrome, Simmons [/bib_ref] The 2000 Ventilation with Lower Tidal Volumes as Compared with Traditional Tidal Volumes for Acute Lung Injury and the Acute Respiratory Distress Syndrome (ARMA) and the Higher versus Lower Positive End-Expiratory Pressures in Patients with the 2004 Acute Respiratory Distress Syndrome (ALVEOLI) trials observed more moderate fluid balances of 4 L and 6 L at day 4, respectively. [bib_ref] Ventilation with lower tidal volumes as compared with traditional tidal volumes for..., Brower [/bib_ref] [bib_ref] Higher versus lower positive endexpiratory pressures in patients with the acute respiratory..., Brower [/bib_ref] As early as the late 1980s, observational data demonstrated an association between a liberal fluid strategy and worse clinical outcomes for ARDS. In 1 study, a lower cumulative fluid balance and a negative trend in body weight during hospitalization were associated with improved survival. [bib_ref] Fluid balance and the adult respiratory distress syndrome, Simmons [/bib_ref] A subsequent observational study found that a 25% reduction in pulmonary capillary wedge pressure (PCWP) among ARDS patients during their ICU course was associated with reduced mortality. [bib_ref] Improved survival in ARDS patients associated with a reduction in pulmonary capillary..., Humphrey [/bib_ref] However, the observational nature of these studies limits the ability to make any statements of causation. A higher PCWP or more positive fluid balance might be a marker of illness severity and confound the early data. A 1992 RCT of 101 critically ill patients with pulmonary artery catheterization to extravascular lung water (EVLW) group and pulmonary capillary wedge pressure (WP) group and looked at the impact of fluid restriction and diuresis on resolution of EVLW and ventilator and ICU days. [bib_ref] Improved outcome based on fluid management in critically ill patients requiring pulmonary..., Mitchell [/bib_ref] Although significantly confounded by including patients with congestive heart failure, fluid restriction and diuresis were associated with lower positive fluid balance and fewer ventilator and ICU days. These studies, although weakened by their observational design and likely confounded by severity, suggest that higher positive fluid balance is associated with worse clinical outcomes in ARDS.
## Conservative fluid management: a paradigm shift
Compared with a liberal fluid management strategy, a conservative strategy restricts fluid administration during the resuscitative phase and employs treatments to reduce the total body fluid balance during the deresuscitative phase. This strategy seeks to reduce the pulmonary ventilation/perfusion mismatch by limiting pulmonary edema but may risk and end-organ damage from decreased cardiac perfusion. [bib_ref] The case for and against fluid restriction and occlusion pressure reduction in..., Schuster [/bib_ref] There are few data examining the association of a liberal or conservative intravenous fluid resuscitation strategy and the development of ARDS. A small cohort study of 296 septic patients, in which 25% developed ARDS within 72 hours, showed no association between the amount of resuscitative intravenous fluid administered in the first 24 hours and the development of ARDS. [bib_ref] Volume of fluids administered during resuscitation for severe sepsis and septic shock..., Chang [/bib_ref] These findings are limited by the small difference in volume of resuscitative fluid between study groups (5.5 vs 4.7 L) and the study's limited sample size. An observational study of 879 patients undergoing elective lung resections found that positive fluid balance was an independent risk factor for developing ARDS. [bib_ref] Risk factors for acute lung injury after thoracic surgery for lung cancer, Licker [/bib_ref] A study of 1366 mechanically ventilated ICU patients, of whom 152 developed ARDS following intubation, found that a positive fluid balance was an independent risk factor for progression to ARDS. [bib_ref] Risk factors for ARDS in patients receiving mechanical ventilation for > 48..., Jia [/bib_ref] Additionally, a case-control study of 414 patients with hospital-acquired ARDS matched with intubated non-ARDS controls found that a greater cumulative fluid balance (7.3 vs 3.6 L) was a modifiable hospital exposure that increased the risk of developing ARDS. [bib_ref] The role of potentially preventable hospital exposures in the development of acute..., Ahmed [/bib_ref] The Crystalloid Liberal or Vasopressors Early Resuscitation in Sepsis (CLOVERS) trial sponsored by the National Heart, Lung and Blood Institute is currently enrolling Fluid Therapy and ARDS septic shock patients and randomizing them to a conservative intravenous fluid resuscitation strategy that uses vasopressors to achieve target blood pressure goals versus a liberal intravenous fluid resuscitation strategy for the first 24 hours of care. [bib_ref] Liberal versus restrictive intravenous fluid therapy for early septic shock: rationale for..., Self [/bib_ref] The primary end point is 28-day mortality. The CLOVERS trial seeks to enroll 2320 participants and will track the development of ARDS over the first 7 days. The primary and anticipated secondary analysis of the CLOVERS participants who develop ARDS and their relationship to intravenous fluids will likely provide the strongest causal data available on the resuscitative phase.
Several RCTs have compared a conservative or deresuscitative fluid to a liberal fluid strategy for septic and/or ARDS patients. These trials, with 1 exception, the Network Fluid and Catheters Treatment Trial (FACTT), were generally smaller proof-of-concept trials. They employed variations of a conservative/deresuscitation strategies that both used and did not use pulmonary catheters to guide fluid removal and found mixed results. [bib_ref] Extravascular lung water and pulmonary arterial wedge pressure for fluid management in..., Hu [/bib_ref] [bib_ref] Effect of different liquid management strategies on the prognosis of acute respiratory..., Wang [/bib_ref] [bib_ref] Effectiveness of treatment based on PiCCO parameters in critically ill patients with..., Zhang [/bib_ref] In a systematic review and meta-analysis of these trails, a conservative/deresuscitative fluid strategy did not demonstrate a mortality benefit. [bib_ref] Conservative fluid management or deresuscitation for patients with sepsis or acute respiratory..., Silversides [/bib_ref] However, a conservative/deresuscitative fluid strategy was associated with increased ventilatorfree days and a shorter ICU length of stay. The combined treatment effect of a conservative deresuscitation on these outcomes was heavily influenced by the inclusion of FACTT trial, which accounted for approximately 50% of the included participants. Other small proof-of-concept RCTs comparing the use of albumin and furosemide versus to placebo or furosemide only suggest that the use of albumin and furosemide may also increase ventilator-free days. [bib_ref] Albumin and furosemide therapy in hypoproteinemic patients with acute lung injury, Martin [/bib_ref] [bib_ref] A randomized, controlled trial of furosemide with or without albumin in hypoproteinemic..., Martin [/bib_ref] The defining trial that tested the effect of conservative fluid strategy in ARDS was the 2006 ARDS Network Fluid and Catheters Treatment Trial. [bib_ref] Comparison of two fluid-management strategies in acute lung injury, Wiedemann [/bib_ref] FACTT enrolled 1000 participants with ARDS over 40 hours after admission to the ICU and excluded patients with ongoing shock. The trial randomized participants to a conservative versus liberal fluid strategy that used a strict protocol of active diuresis, fluid bolus, vasopressor, and/or inotrope based on varying ranges of central venous pressure (CVP) and pulmonary artery occlusion pressures (PAOP). Diuresis was held for 12 hours when patients demonstrated evidence of shock and received vasopressors and/or fluid bolus. At 7 days, the trial produced a large difference in the cumulative fluid balance between the conservative and liberal deresuscitation groups (À136 AE 491 mL vs 6992 AE 502 mL; P<.001). The daily cumulative fluid balance in the liberal group was similar with prior contemporary ARDS trials (4 L and 6 L by day 4 in ARMA and ALVEOLI, respectively) and consistent with usual care at the time. There was no difference in the primary outcome of 60-day mortality in these groups (25% in conservative strategy vs 28% in liberal strategy, P5.30).The conservative strategy group, however, had significantly more ventilator-free days (14.6 AE 0.5 vs 12.1 AE 0.5, P<.001) and ICU-free days compared with the liberal strategy group. Despite the aggressive conservative/deresuscitation strategy, which targeted CVP less than 4 mm Hg and a PAOP less than 8 mm Hg, there was no increase in organ failure between the conservative and liberal arms of the study. Moreover, there were no significant differences in the percentage of patients receiving renal replacement therapy (10% in conservative vs 14% in liberal, P5.06) or the average number of days of renal support. These findings suggest that active deresuscitation may mitigate the lung injury associated with excess intravenous fluids without compromising organ perfusion.
In the FACTT protocol, deresuscitation was held when enrolled patients developed shock for any reason. When the results were further analyzed to compare the impact of conservative and liberal fluid strategies in baseline shock versus nonshock patients, 60-day hospital mortality was lower in the conservative arm than liberal arm in nonshock patients (19% vs 24%), but higher in the conservative arm than liberal arm in shock patients (39% vs 37%). However, a test for interaction of baseline shock and the treatment effect of fluid therapy was not significant for these outcomes.
Since this landmark study, fluid management in ARDS has undergone a clear paradigm shift from liberal to conservative/deresuscitative strategy among clinicians managing critically ill patients with ARDS. However, many concerns about the use of a conservative/deresuscitative strategy remain. It is important to note that this paradigm shift has occurred largely because of this single trial. Although FACTT suggests a conservative strategy may liberate patients from the ventilator earlier without evidence of harm, the clinical implications of conservative/deresuscitative strategy for ARDS patients with shock are not known. In addition, the secondary outcome findings of FACTT have not been prospectively validated in an RCT that evaluates a conservative/resuscitation strategy with ventilator-free days as the primary outcome. Subsequent prospective trials have failed to show mortality benefit of conservative/deresuscitative strategy, although most of these trials are limited by their small sample size. [bib_ref] Extravascular lung water and pulmonary arterial wedge pressure for fluid management in..., Hu [/bib_ref] [bib_ref] Effect of different liquid management strategies on the prognosis of acute respiratory..., Wang [/bib_ref] [bib_ref] Effectiveness of treatment based on PiCCO parameters in critically ill patients with..., Zhang [/bib_ref] Importantly, and perhaps because of the lack of a strong evidence base, most of the guidelines for ARDS management (ATS/ESICM/SCCM) do not recommend specific fluid management strategies, and the British Thoracic Society (BTS) and Japanese Society of Respiratory Care Medicine and the Japanese Society of Intensive Care Medicine (JSRCM/JSICM) make weak recommendations for conservative fluid management. [bib_ref] Official executive summary of an American Thoracic Society/American College of Chest Physicians..., Schmidt [/bib_ref] [bib_ref] Guidelines on the management of acute respiratory distress syndrome, Griffiths [/bib_ref] [bib_ref] The clinical practice guideline for the management of ARDS in Japan, Hashimoto [/bib_ref] The discordance in these guidelines underscores the need for further investigation, including the necessity of identifying subpopulations of ARDS patients with differing responses to fluid administration.
## Heterogeneity of treatment effect and acute respiratory distress syndrome phenotypes
One of the largest challenges facing ARDS research, as well as many other topics in critical care, is the complex heterogeneity of the diseases of interest. Even a rigorously conducted RCT can produce outcomes that do not accurately answer more nuanced clinical questions because of the heterogeneity of participant enrollment. It is crucial to understand that the primary outcome of any study is an average effect estimate across the enrolled study population. Beneficial or harmful effects to specific subgroups from the intervention may be masked within the same RCT. [bib_ref] Implications of heterogeneity of treatment effect for reporting and analysis of randomized..., Iwashyna [/bib_ref] Identification of patient phenotypes may improve understanding of disease syndromes and enable the development of a precision-based approach to clinical trial design. The heterogenous mixture of patients that comprises ARDS may be grouped in many ways. Examples of ARDS phenotypes include severity of hypoxia, precipitating risk factors (eg, sepsis, trauma, pancreatitis, or transfusion), direct versus indirect lung injury, timing of onset (less than or more than 48 hours from admission), radiographic appearance, genotypes, biomarkers, and hyperinflammatory versus non-or hypoinflammatory. [bib_ref] Acute respiratory distress syndrome phenotypes, Reilly [/bib_ref] Identifying such phenotypes and assessing the treatment effects in specific phenotypes have the potential to lead to meaningful and clinically applicable results. This is evidenced by the Prone Positing in Severe Acute Respiratory Distress Syndrome (PROSEVA) trial, where the investigators demonstrated that prone positioning reduced mortality in severely hypoxic patients with PaO 2 /FiO 2 ratio less than 150 mm Hg. [bib_ref] Prone positioning in severe acute respiratory distress syndrome, Gué Rin [/bib_ref] Latent class analysis is a form of mixture modeling that uses available data to identify unmeasured or latent subgroups in a heterogeneous population. [bib_ref] Latent class analysis in health research, Kongsted [/bib_ref] LCA attempts to identify the optimal number of subgroups that best fit a population. Two distinct phenotypes of ARDS have been identified using the latent class analysis, which are Fluid Therapy and ARDS hypoinflammatory and hyperinflammatory. [bib_ref] Subphenotypes in acute respiratory distress syndrome: latent class analysis of data from..., Calfee [/bib_ref] [bib_ref] Acute respiratory distress syndrome subphenotypes respond differently to randomized fluid management strategy, Famous [/bib_ref] [bib_ref] Identification and validation of distinct biological phenotypes in patients with acute respiratory..., Bos [/bib_ref] [bib_ref] Latent class analysis of ARDS subphenotypes: a secondary analysis of the statins..., Sinha [/bib_ref] [bib_ref] Acute respiratory distress syndrome subphenotypes and differential response to simvastatin: secondary analysis..., Calfee [/bib_ref] The hypoinflammatory and hyperinflammatory phenotypes were derived in a secondary analysis from 2 large ARDS RCTs (ARMA and ALVEOLI) and demonstrated different treatment effects in mortality, ventilator-free days, and organ failure-free days when exposed to different ventilation strategies. [bib_ref] Subphenotypes in acute respiratory distress syndrome: latent class analysis of data from..., Calfee [/bib_ref] Famous and colleagues used FACTT and LCA to assess the mortality outcomes of a conservative versus liberal fluid strategy among the subphenotypes. [bib_ref] Acute respiratory distress syndrome subphenotypes respond differently to randomized fluid management strategy, Famous [/bib_ref] Their revised secondary analysis of the FACTT cohort found that the hyperinflammatory group had higher 60-and 90-day mortality and fewer ventilator-free days when compared with the hypoinflammatory group. There was no significant difference in 60-day mortality rates between conservative and liberal fluid strategies in each group. Their data show that most ARDS patients are classified as the hypoinflammatory subphenotype (73%) compared with the hyperinflammatory subphenotype (23%); however, their 301 factor model that includes novel biomarkers makes clinical identification of these phenotypes currently not feasible. Subsequent work is exploring a more parsimonious 3-variable model consisting of IL-8, bicarbonate and protein C to facilitate the clinical integration of subphenotype identification. [bib_ref] Development and validation of parsimonious algorithms to classify acute respiratory distress syndrome..., Sinha [/bib_ref] With the predominance of ARDS patients belonging to the hypoinflammatory subphenotype, and until classification becomes clinically feasible, many clinicians will continue to manage ARDS patients with diuresis and try to achieve even or negative fluid balance.
## Precision based medicine as future treatment of acute respiratory distress syndrome
Clinical trials often require significant investments of time and money to complete. It is important to invest the limited time and resources available in trials that maximize the probability of detecting clinically meaningful treatment effects. One may accomplish this by identifying relevant subphenotypes and then targeting treatment toward specific patient populations most likely to benefit. [bib_ref] Clinical trials in acute respiratory distress syndrome: challenges and opportunities, Matthay [/bib_ref] [bib_ref] Why have clinical trials in sepsis failed?, Marshall [/bib_ref] Robust research design in the fields of genomics, proteomics, and metabolomics is dedicated to identifying biomarkers and deriving biological phenotypes. [bib_ref] Personalized diagnostics and biosensors: a review of the biology and technology needed..., Ahmed [/bib_ref] [bib_ref] Endothelial biomarkers in human sepsis: pathogenesis and prognosis for ARDS, Hendrickson [/bib_ref] Additionally, machine learning has the potential to identify and study the various phenotypes of ARDS through unsupervised learning methods that may uncover associations in data that are not intuitive to researchers. [bib_ref] Machine learning classifier models can identify acute respiratory distress syndrome phenotypes using..., Sinha [/bib_ref] The ongoing evolution of ARDS phenotypes and the utilization of machine learning and adaptive trial platforms hold great promise for the future of enhanced clinical trial design. This in turn will allow for the evaluation of targeted therapies in ARDS and further understanding of how one should best use intravenous fluids to treat patients with ARDS.
## Clinics care points
Fluid management is an important component in management of critically ill patients; however the optimal fluid management for ARDS remains unknown.
Although no prospective, RCTs have shown mortality benefit, it is suggested that conservative fluid management improves outcomes related to ICU stay and mechanical ventilation days.
Further studies addressing differential responses to fluid management in ARDS phenotypes will help guide fluid management for optimal outcomes.
# Disclosure
The authors have nothing to disclose. |
Hydrogen-deuterium exchange mass spectrometry reveals three unique binding responses of mAbs directed to the catalytic domain of hCAIX
Human carbonic anhydrase (hCAIX), an extracellular enzyme that catalyzes the reversible hydration of CO 2 , is often overexpressed in solid tumors. This enzyme is instrumental in maintaining the survival of cancer cells in a hypoxic and acidic tumor microenvironment. Absent in most normal tissues, hCAIX is a promising therapeutic target for detection and treatment of solid tumors. Screening of a library of anti-hCAIX monoclonal antibodies (mAbs) previously identified three therapeutic candidates (mAb c2C7, m4A2 and m9B6) with distinct biophysical and functional characteristics. Selective binding to the catalytic domain was confirmed by yeast surface display and isothermal calorimetry, and deeper insight into the dynamic binding profiles of these mAbs upon binding were highlighted by bottom-up hydrogendeuterium exchange mass spectrometry (HDX-MS). Here, a conformational and allosterically silent epitope was identified for the antibody-drug conjugate candidate c2C7. Unique binding profiles are described for both inhibitory antibodies, m4A2 and m9B6. M4A2 reduces the ability of the enzyme to hydrate CO 2 by steric gating at the entrance of the catalytic cavity. Conversely, m9B6 disrupts the secondary structure that is necessary for substrate binding and hydration. The synergy of these two inhibitory mechanisms is demonstrated in in vitro activity assays and HDX-MS. Finally, the ability of m4A2 to modulate extracellular pH and intracellular metabolism is reported. By highlighting three unique modes by which hCAIX can be targeted, this study demonstrates both the utility of HDX-MS as an important tool in the characterization of anti-cancer biotherapeutics, and the underlying value of CAIX as a therapeutic target.ARTICLE HISTORY
# Introduction
Carbonic anhydrases are integral to the regulation of intracellular pH and fluidics in both physiological and pathological disease states.In humans, this family of 15 ubiquitous metallo-enzymes, which exist either as cytosolic, membranebound or secreted isoforms, share diverse organ and tissue distribution profiles and catalytic efficiencies.All enzymes catalyze the hydration of carbon dioxide (CO 2 ) to protons (H + ) and bicarbonate (HCO 3 − )following a twostep mechanism.In particular, human carbonic anhydrase IX (hCAIX), one of the four extracellular and membrane-bound zinc-containing carbonic anhydrase isozymes, has long been recognized as a tumor-associated protein. hCAIX expression is, in most cases, triggered by hypoxia (primarily through the H1F-1 transcription factor),and the protein can be found in many types of solid tumors where it is correlated with poor prognosis and therapeutic outcome.hCAIX is key to the survival of tumor cells in the acidic hypoxic tumor microenvironment, playing a role in cell proliferation, cell adhesion, and other tumor promoting processes.With the exception of gastric, gall bladder, and intestinal cells, hCAIX is mostly absent in tissues under normoxic conditions.Taken together, hCAIX can be considered a promising therapeutic target for the detection and treatment of solid tumors. 10,Full-length hCAIX exists as a disulfide stabilized membrane-bound dimer with an extracellular domain (ECD) that harbors a proteoglycan (PG) and catalytic (CA) domain,both of which are integral to hCAIX's enzymatic function. The PG domain, a unique structural feature among carbonic anhydrases only found in hCAIX, is an intrinsically disordered motif that has a role in cell adhesion,and that has been suggested to act as a "proton antenna" at the entrance of the active site of the catalytic domain. The active site itself is a 12 Å wide and 13 Å deep conical cavity bordered by a twisted βsheet, with a nucleophilic, tetra-coordinated zinc ion located at the bottom.This catalytic pocket is characterized by several distinct hydrophobic (L91, V121, V131, L135, L141, V143, L198, P202) and hydrophilic (N62, H64, S65, Q67, T69, Q92) amino acids, which have been suggested to facilitate efficient substrate turnover.Several small molecules that inhibit hCAIX's enzyme activity by targeting the catalytic cleft have been extensively studied.These include agents that 1) directly bind the catalytically active zinc ion and its coordinated water/hydroxide ion, 2) sterically hinder entrance of the CO 2 substrate into the catalytic cleft, or 3) less commonly engage sites distal from the hCAIX active site.In contrast to small molecule inhibitors that can lack specificity and may be more likely to cause off-target toxicity, antibody-based therapies generally show improved safety profiles and increased selectivity. Biotherapeutics have emerged as key players in the treatment of cancer.There has been growing interest into developing antibody-based therapeutics targeting hCAIX. Two of the most studied anti-hCAIX monoclonal antibodies (mAbs) are the chimeric cG250which, as an adjuvant treatment, has demonstrated clinical potential against renal cell carcinoma,and the M75 mouse mAb, which binds the PG domain and is widely used for immunohistochemistry and imaging purposes.In addition, A3 and CC7, two notable humanized hCAIX's ECD binding antibodies derived from phage display, have been developed as non-immunogenic imaging tools,MSC8, which has demonstrated efficient in vitro enzyme inhibitory activity by modulating the intrinsic enzyme activity of hCA-IX,and VII/20 which has been shown to induce significant hCAIX internalization,a key characteristic for the development of an antibody-drug conjugate (ADC).In-depth biophysical characterization of antibody-antigen interactions, both in terms of the binding epitope and downstream effects, is an integral step in the development of biotherapeutics. Methods such as phage display,site-directed mutagenesis,and synthetic peptide arrayshave proven invaluable for assigning linear epitopes. In addition, conformational epitopes and allosteric effects can be mapped with high-resolution techniques such as X-ray crystallography, nuclear magnetic resonance, or by using more recently developed techniques such as cryo-electron microscopy.However, sample size and protein complexity often limit the routine application of these methods.
Mass spectrometry has recently emerged as a powerful tool for the characterization of biotherapeutics and their interactions with antigens.In this regard, hydrogendeuterium exchange mass spectrometry (HDX-MS) has proven its merit. This technology provides detailed insights into protein interactionsand is well-suited for the assignment of both linear and conformational epitopes.HDX-MS examines conformational changes induced in the antigen upon antibody binding by monitoring the exchange rate of backbone amide-associated hydrogens with solvent deuterons.The exchange rate is largely determined by the degree of its involvement in secondary protein structure and/or the antigen-antibody binding interface.Therefore a comparison of amide hydrogen exchange rates between different protein states (i.e., antibody-bound vs. free), provides a powerful insight into the relative conformational stability of an antigenantibody complex.In this study, we describe the bottom-up HDX-MS characterization of a subset of therapeutic anti-hCAIX antibody candidates, designated c2C7, m4A2 and m9B6, which target the catalytic domain of hCAIX.HDX-MS kinetics complemented by an in-depth biophysical profiling of the hCAIX-antibody interactions by surface plasmon resonance (SPR), isothermal calorimetry (ITC) and yeast surface display (YSD), and an in vitro functional assessment of enzymatic inhibition. Our efforts identified three unique modes by which the catalytic activity of hCAIX can be targeted with an antibody-based therapeutic. MAbs that target an epitope distal from the catalytic cleft can be used to develop an ADC (e.g., mAb c2C7), whereas functional inhibition of hCAIX can either be achieved by sterically hindering its substrate (i.e., CO 2 ) from entering the catalytic cleft (e.g., m4A2), or by destabilizing key structural features that decreases the rate of substrate turnover (e.g., m9B6).Hydrophilic residues in the catalytic cavity are shown in magenta, while hydrophobic are shown in cyan. Three arginine residues (R58, R60, R130), unique to CAIX within the CA family, that border the active site are highlighted in Orange. View is looking down into the catalytic cavity from above the membrane.
# Results
## Preliminary characterization of antibodies targeting the caix catalytic domain
A set of antibodies (c2C7, m4A2 and m9B6) that exclusively bind the hCAIX catalytic (CA) domain were selected from a larger panel of hCAIX binding antibodies for further characterization.An initial attempt was made to qualitatively identify the antibody epitopes on hCAIX ECD using YSD by expressing the eight hCAIX fragments as shown in. Fragments 1-3 consisted of only the proteoglycan (PG) domain, fragments 4 and 5 covered the CA domain with different N-terminal extensions, fragments 6 and 7 covered both the CA similar to fragments 4 and 5 with C-terminal extension just to the transmembrane (TM) domain, and fragment 8 included all domains except for the signal peptide (SP). Antibodies m9B6, m4A2 and c2C7 all demonstrated binding to only those fragments containing the CA domain as highlighted in, while no binding was observed to fragments containing only the PG domain. Next the structural nature of epitopes, linear versus conformational, was assayed by repeating the YSD assay using heat-denatured hCAIX antigen fragments. Interestingly, only m9B6 was able to bind efficiently to the denatured hCAIX fragments, while m4A2 and c2C7 completely lost their binding to these fragments following denaturation. These results provide evidence that m9B6 binds to a contiguous, or linear, epitope, whereas m4A2 and c2C7 recognize conformational epitopes.
In addition, antibody binding was also evaluated by isothermal calorimetry using both the monomeric and dimeric forms of hCAIX. Binding constants ranging from 0.76 to 35 nMand 0.85 to 96.5 nM, respectively. These calculated binding constants are in close agreement with those previously determined by SPR, for all but m9B6 (K D 0.74 nM), whose K D is weaker by an order of magnitude.Nonetheless, both SPR and ITC indicate that the overall binding kinetics of m9B6 are poor compared to the other antibodies.
## Hdx-ms kinetic screening with δpg-hcaix cys38ser
The binding profile of each of the antibodies was further assessed by HDX-MS. This method is a powerful tool for monitoring changes in the average conformational state of proteins, and is routinely applied in the characterization of antibody-antigen interactions.To facilitate the HDX-MS interrogation we expressed a monomeric form of hCAIX that is devoid of its PG domain by removing the inter-dimer disulfide bond at Cys 38, ΔPG-hCAIX Cys38Ser ).This construct maintained catalytic activity and structural integrity and allowed for a simplified HDX-MS experimental design to be implemented.Overall, peptic digestion of ΔPG-hCAIX Cys38Ser resulted in 75 identified peptides with 78% sequence coverage and a redundancy of 4 (Table 2,. Gaps in sequence coverage were attributed to N-linked glycosylation (N210) and insufficient reduction of cysteines. Raw deuteration data can be found in Tables S1 and S2, while a complete set of kinetics plots are shown in. After 60 min of labeling, normalized deuteration of peptides ranged from minimal deuterium incorporation (~1%), to almost complete labeling ~90%. Technical replicates allowed the calculation of average standard deviation values which range from 0.03-0.07 for all deuteration measurements for each protein state. Preliminary data has shown that the mAbs can be separated into two distinct categories: ADC candidates (c2C7 and cG250) and functional enzyme inhibitors (m4A2 and m9B6).HDX-MS findings for each category of hCAIX-binding mAbs are described in the following sections. All structural references are based on assignments made by Alterio et al.Similar binding profiles to those described below were extracted from preliminary differential HDX-MS using the full-length hCAIX dimer protein, including the PG domain (data not shown) suggesting that neither dimerization nor inclusion of the PGdomain significantly alters the binding mode of c2C7, cG250, m4A2, or m9B6.
## C2c7 and cg250 bind an allosterically silent discontinuous epitope
C2C7 and cG250 both bind the CA domain, internalize but do not inhibit CAIXs enzyme activity.These two mAbs can thus be considered good ADC candidates. We found some striking similarities in the HDX binding profiles of these two antibodies. First, both cG250 and c2C7 elicited strong stabilizations (decrease in deuteration of the bound state relative to the control) for which the magnitude increased over time in the region spanning amino acid residues L215-L237, blue). A smaller stabilization was also observed across residues N238-E251; c2C7 stabilized the complex at shorter timepoints whereas cG250 demonstrated a slightly stronger stabilization at the longest timepoint, magenta). These stabilized regions encompass the αG-helix (residues S219-T229), as well as a long unstructured region corresponding to residues L229-L251.Second, both c2C7 and cG250 stabilized residues L157-S178, dark blue), with cG250 inducing a stronger stabilization in I163-S178, light blue) at later timepoints. This stabilization covers the C-terminal half of the αE-helix (N150-A164), as well as the βA-sheet feature (E169-V172). Third, stabilization by both c2C7 and cG250 ranging from −4% to −6%, respectively, was also observed in residues H90-E120, cyan). This region consists of a buried β-sheet (βD, residues R82-W93) and an unstructured loop that resides immediately adjacent to the αG-helix. cG250, but not c2C7, slightly stabilizes residues R60-Q67 at the longest time point, purple), covering two β-sheets that sit between the catalytic cleft and αE-helix/βA-sheet. This inference is based on overlapping peptide sequences: significant changes in deuteration were observed in peptides spanning residues L47-Q67 but not those covering residues L47-L59. No structural relaxation, or increased solvent exposure was observed for either c2C7 or cG250. It is evident within the timepoints sampled here that the strongest stabilizations are localized to a motif consisting of three structural features (αE/αG/βA) with additional perturbations in the proximal residues. Given the surface-exposure of this motif and its distal location from the catalytic cavity and dimerization interfaces, we hypothesize that the described αE/ αG/βA motif represents a discontinuous epitope, where the mAb engagement results in minimal allosteric effects, and no effect on the intrinsic enzymatic activity of hCAIX.
## M4a2 and m9b6 inhibit hcaix through different binding mechanisms
m4A2 and m9B6 both bind the CA domain and inhibit hCAIX activity.Three-dimensional projections of significant shifts for both inhibitory antibodies (m4A2 and m9B6) are shown in . First, changes in the region spanning residues F123-L139 bordering the entrance to the catalytic cavity and including the αD-helix (R125-L130), βF-sheet (L136-E145), and an arginine residue (R130) that may be involved in steric control of substrate accessibility 12 were observed in response to both antibodies , magenta). A strong stabilization was observed for all peptides in this region at nearly all labeling timepoints for m4A2. This region has previously been identified in a linear peptide mapping assay as one of two plausible epitopes for m4A2 (1D-5 W or 127D-132 R).In contrast, . Differential HDX-MS of potential inhibitors m4A2 and m9B6 bound to ΔPG-CAIX Cys38Ser (PDB 3IAI). Regions demonstrating significant stabilization (3x SD, p = .01) across overlapping peptides at any single time point are color coded for clarity. Regions with no significant change are shown in dark gray, while white corresponds to missing coverage. Key residues are shown as stick models (R58, R60, R130, H90, H92, H115), and zinc is shown as a gray sphere. Kinetics plots corresponding to select peptides within structurally significant region are shown with arrows indicating associated structural features. Unbound hCAIX control is shown in black, m4A2 in blue, and m9B6 in Orange, and significant changes in deuteration are marked by asterisks. Data collected in triplicate. a slight destabilization (4-8%) was observed in this region for m9B6. We also observed a structural destabilization in residues L47-L59 only in response to m4A2 binding , light blue). This region consists primarily of unstructured loops, but also, importantly, the N-terminus of the βD-sheet, and structurally relevant arginine residues, R58 and R60. Similar to R130, these arginines have been proposed to engage in activesite gating activity. 12 m9B6 binding destabilized the regions that were stabilized by cG250 and c2C7. In particular, a strong destabilization at later timepoints was evident in residues L215-L237 , purple) and in residues Y84-E102 , cyan).
In summary, binding of the m9B6 antibody resulted in destabilization of multiple ΔPG-CAIX Cys38Ser motifs, while the m4A2 antibody stabilized the entrance of the catalytic pocket, likely through direct binding to the αD-helix. To deduce the functional relevance of these conformational "hotspots", we performed further biophysical and functional characterization.
## M4a2 and m9b6 exhibit synergistic inhibition of hcaix
Given the contrasting binding profiles and unique inhibitory mechanisms highlighted by HDX-MS for m4A2 and m9B6, we investigated their ability to simultaneously bind and inhibit hCAIX. ADC candidates were excluded from the following assays, due to their demonstrated lack of inhibitory activity.Surface plasmon resonance (SPR) experiments, in which both antibodies were either immobilized or flowed, confirmed the non-competing nature of the binding of m9B6 and m4A2 to hCAIX . We next investigated the functional inhibitory effect of m4A2 and m9B6 alone and in combination in an in cell catalytic assay . This assay assesses hCAIX catalytic activity directly in live 67NR murine breast cancer cells engineered to constitutively express hCAIX by measuring pH changes in the cellular environment resulting from conversion of carbon dioxide (CO 2 ) to bicarbonate (HCO 3 − ) and protons (H + ) by hCAIX.The presence of either m4A2 or m9B6 inhibited extracellular acidification relative to the IgG2b control. Inhibition was even stronger when the cells were exposed to a combination of m4A2 and m9B6, indicating synergistic action of the two antibodies . This synergistic behavior was also observed in in vitro enzyme activity assays that used a purified hCAIX dimer and 4-methylumbelliferyl acetate (4-MU-Ac) as a surrogate substrate .
The underlying mechanism of synergistic inhibition was further probed by HDX-MS. As the m4A2/m9B6 antibody combo was not included in the initial screen, an additional peptide map was generated, covering 82% across 77 peptides, and a separate non-binding reference state was collected under similar conditions. Full kinetics plots for the HDX dataset using the mAb combination can be found in . Compared to the unbound state, the m9B6/m4A2 combo demonstrated binding profiles that were . (a-b) Competitive SPR binding demonstrating that mAbs m4A2 and m9B6 bind to independent epitopes. (c) Analysis of CAIX catalytic activity by 67NR murine breast cancer cells expressing hCAIX incubated with anti-CAIX mAb m4A2 and/or m9B6, followed by addition of buffer and CO 2 -saturated water and monitoring the decrease in pH using the "in-cell" carbonic anhydrase activity assay. Cells incubated in the presence of IgG2b were used as a control. Data show the mean ± SEM of technical replicates (n = 3/group) and are representative of 2 independent experiments. consistent with the major shifts observed in each of the single antibody configurations. Significant stabilization in residues F123-L136 was observed, despite both 4A2 and 9B6 exerting opposing conformational influences in this region , magenta). Since we did not correct for intrinsic exchange rates in both HDX-MS assays performed in this report, a quantitative comparison of the stabilization of m4A2 alone or in conjunction with m9B6 is not possible. Further, destabilization observed in residues L215-L237, which can be attributed to m9B6 binding, and residues L47-L59, attributed to m4A2, was preserved. Conformational shifts resulting in destabilizations in response to either m9B6 (residues H90-E102) and m4A2 (residues L68-E75) were not classified as significant in the antibody combination. This is likely attributable to a relative attenuation of the overall mass shifts. Our HDX-MS assay profiles and binding assays provide evidence for simultaneous occupancy and synergistic inhibition of hCAIX by both m4A2 and m9B6, a finding that suggests potential for complementary therapeutic applications.
## M4a2 emerges as leading candidate for therapy
A slightly stronger inhibitory effect was observed with m4A2 and with more favorable binding kineticsandcompared to m9B6, suggesting that the m4A2 binding mechanism may be favorable for the development of this mAb as a CAIX inhibitor. CAIX activity contributes to acidification of the tumor microenvironment, and previous studies have shown that pharmacologic or genetic inhibition of CAIX reduces extracellular acidification and impairs glycolysis by cancer cells. 57,58 Therefore, we investigated the effects of m4A2 on extracellular acidification and metabolism. We cultured MDA-MB-231 human breast cancer cells engineered to constitutively express CAIX (CAIXpositive) in the presence of m4A2 and analyzed the extracellular acidification rate (ECAR) and glycolytic metabolism using a Glycolytic Rate Assay. This assay calculates the contribution of mitochondrial-derived CO 2 to the overall extracellular acidification rate, allowing for the determination of the proton efflux rate (PER) attributed to glycolysis (glycoPER). We observed reduced basal levels of ECAR in CAIX-positive cells cultured in the presence of m4A2, relative to cells exposed to isotype-matched control IgG. Consistent with the reduced basal levels of ECAR, analysis of the PER profiles calculated using ECAR and oxygen consumption rate (OCR) measurements demonstrated lower values for both basal total PER and glycoPER by m4A2-treated cells. Indeed, we found that basal PER and basal glycoPER were significantly reduced in cells cultured with m4A2, while the mitoOCR to glycoPER ratio was significantly increased, suggesting that m4A2-mediated inhibition of CAIX activity functions to reduce glycolytic metabolism by CAIX-expressing cancer cells. Similar results were observed using breast cancer cells in which CAIX expression was silenced using CRISPR-mediated gene-editing strategies,confirming the effect of 4A2-mediated inhibition of CAIX on extracellular acidification and glycolytic metabolism.
In addition to changes in metabolism, previous studies have shown that depletion of CAIX expression leads to increased death of cancer cells in hypoxia.To extend our observations on the biological consequences of inhibiting CAIX function with m4A2 using a model of endogenous CAIX expression, we cultured U87 MG glioblastoma cells in hypoxia, an environment that drives strong upregulation of CAIX,in the presence of m4A2 and assessed cell death using a terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL). Quantification of the TUNEL positive cells showed an approximate 1.8-fold increase in the apoptotic cell population in response to m4A2 compared to cells exposed to an isotype control IgG. Together, the glycolytic rate assay results in combination with the TUNEL assay results indicate that inhibition of CAIX activity leads to decreased glycolysis and an increase in cell death.
# Discussion
Biophysical characterization of antibody-antigen interactions is integral to the design and development of antibody-based therapeutics. Once a library of antibodies with desired binding kinetics and functional activity is selected, the fine-details of their binding modes at the antigen level must be elucidated. Such information is invaluable in classifying mAbs by their binding epitopes and for selecting antibodies with desired binding characteristics. Multiple modes of hCAIX engagement have been observed for anti-hCAIX biotherapeutics, and different modes of engagement lead to different therapeutic outcomes. 10,17 This diversity of therapeutic outcomes makes hCAIX an attractive antibody-based therapeutics anti-cancer target. We have described the selection of a panel of novel antibodies targeting both the catalytic and proteoglycan domains of hCAIX.The study presented here focuses on the detailed biophysical characterization of three antibodies selected from this panel that specifically target hCAIX's catalytic domain. YSD provided data that the epitopes were in the hCAIX and were likely (with the exception of the m9B6 epitope) conformational, but did not provide sufficient granularity to pinpoint the epitope or how the antibodies engage hCAIX differently. A previous report suggested that the binding epitopes of antibodies targeting the catalytic domain are predominantly conformational, while those targeting the PG domain are linear.Linear peptide scanning has proven effective for mapping epitopes in the PG-domain,but size limitations and the conformational nature of the epitopes on CAIX's catalytic domain render this technology ineffective in this context.
HDX-MS can efficiently screen antibody-antigen interactions under native conditions with peptide-level resolution.We used HDX-MS to interrogate the binding profiles of our three catalytic domain-specific antibodies, in addition to the cG250 mAb. The carbonic anhydrase family has been previously interrogated by HDX-MS, where the catalytic role of zinc in CAII was investigated.However, to our knowledge, this is the first report of applying the HDX-MS technology to resolve the binding epitopes of antibodies targeting hCAIX. It is evident that the subtle details in protein dynamics revealed by HDX-MS would have been impossible to obtain with a crystal structure-based epitope determination.
Both c2C7 and cG250 have demonstrated activity as antibody conjugates. For example, c2C7 demonstrated significant reduction in tumor volume in vivo when conjugated to the microtubule inhibitor mertansine. 52 cG250-conjugates have shown promise in clinical trials for renal cell carcinoma, either as an imaging tool ( 124 I-Girentuximab)or as a radioimmunotherapy ( 177 Lu-Girentuximab).A shared binding mode was observed for c2C7 and cG250. The discontinuous nature of this epitope, formed by the αE/αG/βA motif as determined by HDX-MS, is in agreement with the yeast-display findings. This epitope is far from the catalytic cavity, dimerization interface, PG-and transmembrane-domain,and binding to it likely leaves access to the catalytic cavity unimpeded under physiological conditions. This accounts for the lack of inhibitory action, and reinforces the value of these mAbs as ADC candidates. Mapping of the precise residues engaged by c2C7 and cG250 was not possible with our current HDX-MS dataset. Further experiments, such as molecular docking and alanine scanning mutagenesis may be warranted to resolve the epitope.
A conformational epitope on the catalytic domain for cG250 has previously been identified.A specific epitope identified by linear peptide scanning consisting of A77-L87 and L119-V126 has recently been proposed for cG250. 6667 These sequences are near or overlap with the dimerization interface, and may not be optimal for the binding of a relatively large antibody. Due to the sensitivity of HDX-MS to conformational epitopes and limitations of linear peptide scanning, we propose our findings as the most probable binding mode of cG250. Finally, negatively charged substitutions in hCAIX in residues L157-G167, relative to all α-carbonic anhydrase isoforms, as well as low sequence conservation in the structural features within the discontinuous epitope, may impart specific binding to the hCAIX isoform.Indeed, c2C7 was found to be specific for hCAIX ectodomain.Antibody-mediated functional inhibition of hCAIX has previously been reported; for example, the MSC8 mAb was suggested to directly inhibit hCAIX activity, rather than induce its internalization.Here we report on two non-competing hCAIX enzyme inhibiting mAbs m4A2 and m9B6, whose two binding profiles suggest divergent mechanisms of action. Importantly, both of these antibodies have demonstrated specificity for hCAIX across species and among secreted and membrane-bound isoforms.Interestingly, m4A2 epitope shares the lowest homology with other members of the CA family, and has previously been identified as a potential target for rational drug design.It is intriguing that conformational perturbations are observed in or near all three of the positively charged arginine residues at the catalytic cavity entrance. Along with the highly negatively charged PG-domain, these residues have been postulated to play an active role in gating the catalytic cavity.We therefore hypothesize that m4A2-mediated enzymatic inhibition of hCAIX is driven by steric gating, thereby obstructing the entry of substrate into its catalytic cleft by direct binding to those residues that play a critical role in the function of the catalytic cavity.m9B6 caused an ensemble of destabilizations of hCAIX and, although this mAb is almost equally potent as m4A2 in inhibiting hCAIX , it stands out from all the mAbs tested as having the weakest binding kinetics, S1). This likely explains why, contrary to m4A2, only relatively minor destabilizations (~ −5%) were observed in a previously proposed linear epitope (S122 -128E).Despite the lack of a clear canonical binding event, i.e., stabilization of the binding interface, we can make some inferences about the nature of the m9B6 binding mode. First, the YSD experiments show that m9B6 binds a linear epitopethat does not overlap with that of m4A2 . Second, m9B6/hCAIX interaction may exhibit a so-called "type-2" binding behavior, described by where binding energy is reinvested to drive structural destabilization.From this point of view one could envision that m9B6 only engages hCAIX after an unfavorable unfolding event,where the conformational distribution of the m9B6/hCAIX complex is subsequently biased toward a state with reduced solvent turnover efficiency. This is further supported by the relatively weak binding behavior, and structural loosening of two histidines (H90 and H92; , cyan) within the previously described catalytic triad.Importantly, this destabilized state remains accessible to m4A2 binding. The absence of evidence of a direct antibody interaction does not preclude a stable 9B6/hCAIX complex.We cannot rule out that m9B6 binds a sequence for which HDX coverage is missing. Further, HDX-MS is limited to changes in the average conformational state of a protein, and therefore this technology is blind to binding events driven purely by side-chain or electrostatic interactions.The observation of synergistic behavior and non-competing binding between m4A2 and m9B6 suggests that both of the outlined inhibitory mechanisms are engaged by nonoverlapping epitopes. These findings demonstrate that these antibodies can be used alone or as combinatorial therapy , or can form the basis for the rational design of biparatopic antibodies. Further studies are required to fully understand what fundamental mechanisms are employed by our antibodies, particularly m9B6, in the inhibition of hCAIX activity. This knowledge, in addition to their possible pHdependent behavior, is critical for the design of antibodybased biologics that target the acidic tumor microenvironment.
In conclusion, the HDX-MS evaluation of a set of functionally different hCAIX targeting antibodies highlights two groups of mAbs with three modes of binding, and demonstrates power of HDX-MS to deliver conformational insights in combination with complementary biophysical and functional assays. The work presented here thus lays the groundwork for future mAb engineering efforts targeting both hCAIX and other membrane-bound enzymes.
# Materials and methods
## Recombinant antibody production and purification
Anti-human CAIX mouse mAbs were sequenced and full-size heavy and light human IgG chains were ordered from a commercial vendor (Genscript); V H and V L CDR1-3 regions of 2C7 was grafted on a human IgG1 framework (designated chimeric, 'c') whereas those of 4A2 and 9B6 were grafted on a murine IgG2a framework (designated as mouse, 'm').MAbs were produced by co-transfection of CHO-3E7 cells, which were then grown to a density of 2 ×10 6 cell/mL in 150 mL of F17 medium in 1 L shaker flasks. Cells were transfected with 2 μg total DNA (containing 500 ng each of the heavy and light chain constructs) using PEI MAX™ (Polysciences, Inc.), maintained at 37°C for 24 h, then fed with Tryptone N1 at 1% w/v, after which they were transferred to 32°C for 6 days. Cell culture supernatant was harvested and analyzed by SDS-PAGE for expression.
MAbs were purified from cell-culture supernatants by protein-A affinity chromatography in batch mode (cG250 and c2C7) or by using packed columns (m9B6 and m4A2). Supernatants were incubated and agitated overnight at 4°C with MabSelect SuRe resin (Cytiva Life Sciences) or loaded onto MabSelect SuRe packed columns, previously equilibrated with HyClone TM Dulbecco's phosphate-buffered saline (DPBS; Cytiva Life Sciences) or 10 mM sodium phosphate, 140 mM NaCl buffer pH 7.0. Resin was washed with DPBS or 10 mM sodium phosphate, 140 mM NaCl buffer pH 7.0 and mAbs were eluted with 100 mM citrate buffer pH 3.0 (cG250, c2C7, m4A2) or 100 mM Glycine buffer pH 2.6 (m9B6). Fractions containing mAbs were pooled and neutralized with 1 M HEPES. Protein-A purified material was buffer-exchanged against DPBS by Tangential Flow Filtration using 30 kDa MWCO cassettes (Millipore Sigma: m9B6; Pall Corporation: m4A2), or using 30 kDa MWCO Vivaspin® centrifugal concentrators (Sartorius -cG250 and c2C7). cG250 and c2C7 mAbs were further purified by preparative size exclusion chromatography (SEC) on Superdex-200 columns (Cytiva Life Sciences), using DPBS as mobile phase. Purified mAbs were concentrated to 5-10 mg/mL by ultrafiltration using Vivaspin® centrifugal concentrators (MWCO 30 kDa; Sartorius), and sterilized by filtration through 0.2 μm filters. During the process, the protein concentration was monitored on a NanoDrop™ 2000 spectrophotometer (ThermoFisher Scientific) using absorbance at 280 nm and the calculated specific extinction coefficient of each mAbs. MAbs were kept in aliquots at −80°C for long-term storage, and reevaluated by UPLC-SEC upon each thaw prior to use.
## Hcaix ectodomain expression and purification
The hCAIX extracellular domain cDNA (HGNC:1383; a.a. 1-414), was cloned in the PTT5 expression vector, produced as a C-terminal penta-His-tagged protein in CHO-3E7 cells,purified by immobilized metal affinity chromatography. SDS/ PAGE (reducing and non-reducing conditions) indicated the presence of monomeric and dimeric subpopulations which were further purified by preparative SEC to separate dimers from monomers. SEC purification was performed on Superdex-200 Increase column (Cytiva Life Sciences), using DPBS as mobile phase. Preparative SEC fractions were analyzed by SDS-PAGE to select which fractions to pool. Fractions containing monomer or dimer only were pooled together and concentrated by ultrafiltration using 10 kDa MWCO Vivaspin® centrifugal concentrator (Sartorius). Concentrated material was sterilized by filtration through 0.2 μm filters. During the process, the protein concentration was monitored on a NanoDrop™ 2000 spectrophotometer (ThermoFisher Scientific) using absorbance at 280 nm and the calculated specific extinction coefficient of the hCAIX monomer or dimer. Both preparations were kept in aliquots at −80°C for long-term storage, and reevaluated by UPLC-SEC upon each thaw.
## Isothermal calorimetry
ITC experiments were performed on a MicroCal Auto-ITC 200 (Malvern). In total 19 injections were performed using the following instrument settings: cell temperature 25°C, reference power 10 µCal/second, initial delay 60 seconds, stirring speed 750 rpm, feedback mode/gain high, and injection volume 2 µL for 4 seconds spaced at 120 second intervals with a filter period of 5 seconds. The syringe volume was 120 µL and the cell volume was 400 µL. Data were analyzed using MicroCla PEAQ-ITC Analysis Software v1.1.0.1262 (Malvern).
## Domain-level epitope mapping by yeast surface display
To map the epitopes of the anti-hCAIX mAbs, fragments of hCAIX extracellular domain (ECD) were PCR amplified from the hCAIX DNA sequence corresponding to various combinations of hCAIX protein domains and sections between domains. The DNA fragments were then cloned into the YSD vector by in vivo recombination (IVR) in yeast. The hCAIX ECD and its fragments were expressed and covalently displayed on the surface of yeast cells using YSD.The YSD vector (pPNL6) was from The Pacific Northwest National Laboratory, USA. The hCAIX fragments were expressed as fusion proteins (Aga2-HA -(CAIX)-MYC) on the yeast cell surface. The binding of the mAbs to yeast cells was performed using a whole yeast cell ELISA. Briefly, the various clones of the induced yeast cells were loaded into a 96-well filter plate (1x10 7 cells per well), blocked, incubated with the anti-hCAIX antibodies, the binding of these mAbs was revealed by a horseradish peroxidase (HRP)conjugated secondary antibody followed by incubating with the HRP substrate tetramethyl benzidine (TMB) according to the manufacturer's conditions and read at OD 450 . The relative amount of displayed fusion protein was measured by probing with an anti-MYC antibody (ab19233, Abcam), followed by an HRP-conjugated secondary antibody (ab97135, Abcam) to be used to normalize the binding signal for anti-hCAIX mAbs. For determination the nature of epitopes, linear versus conformational, yeast cells with displayed hCAIX fragments were heated at 80°C for 30 min, then chilled on ice for 20 min prior to labeling with antibodies.The binding levels for anti-hCAIX mAbs were normalized to the amount of Myc tag (ratio anti-hCAIX/anti-Myc).
## Hdx-ms
ΔPG-hCAIX Cys38Ser was equilibrated with the mAbs at a 1:1 molar ratio in phosphate-buffered saline (PBS, pH 7.4). hCAIX-mAb complexes (4 μL) were diluted with 4 μL of labeling buffer (90% D2O, 0.1x PBS, pD 7.0). The labeling reaction (0.25, 1, 3, 10, 60 min) was quenched with a 5x dilution of reducing buffer (100 mM TCEP, 100 mM Gly-HCl, pH 2.2). Samples were manually injected into an ice-cooled loop, digested online with a Poroszyme Immobilized pepsin cartridge (Thermo) and trapped on a C18 Pepmap100 (1 x 5 mm, 5 μm, 100 Å) at 100 μL/min for 3 min in mobile phase A (0.23% formic acid). Peptic peptides were eluted from a Halo-Peptide ES analytical column (50 x 0.3 mm, 2.7 μm, 160 Å, Sciex) with a 10-35% acetonitrile gradient (0.23% formic acid) over 7.5 min at 10 μL/ min. Data acquisition was performed with a Waters Q-Tof Ultima API MS. Peptides were identified by data-dependent acquisition of an undeuterated sample which subsequently searched against a Mascot database. Data was collected in triplicate and analyzed with MS Studio.Significant changes in deuteration were defined by ± 3 SD (p = .02).
## Competitive spr binding assay
All SPR experiments were performed using the ProteOn XPR36 system (Bio-Rad Laboratories Inc.). PBS with 0.05% Tween 20 (Teknova) and with 3.4 mM ethylenediaminetetraacetic acid (EDTA) disodium salt (Teknova), pH 7.4 was used as running buffer and all experiments were performed at 25°C. Reagents used for the antibody immobilizations were obtained from Bio-Rad Laboratories Inc.
The CAIX antibody clones 4A2 and 9B6 were immobilized on a ProteOn GLC sensor chip using standard amine coupling chemistry (All reagents were obtained from Bio-Rad Laboratories Inc). Using a flow rate of 100 µL/min for 140 s, all ligand channels were activated simultaneously by injecting a fresh mixture of 20 mM 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) and 5 mM N-hydroxysulfosuccinimide (Sulfo-NHS) followed by 25 µg/ mL of the antibodies diluted in 10 mM sodium acetate buffer (pH 4.5; 25 µL/min for 240 s). The sensor chip surface was then deactivated with 1 M ethanolamine HCl (30 µL/min for 300 s). Approximately 5000 RU each of the 4A2 and 9B6 antibodies were immobilized. One ligand channel was used for a mock immobilization for subsequent reference subtraction. Next, using the Co-Inject Ligand function of the ProteOn Manager TM software v3.1.0.6, the CAIX monomer (100 nM) and a buffer blank was simultaneously injected (50 µL/min, 210 s) over the two immobilized CAIX antibodies. Immediately after, the 4A2, 9B6 and cG250 (100 nM) and a buffer blank were injected followed by an 800 s dissociation phase (50 µL/ min, 180 s). Sensorgrams were aligned, and the resulting sensorgrams were analyzed using ProteOn Manager Software v3.1.0.6.
## Cell culture
The MDA-MB-231 human breast cancer cell line and U87 MG human glioblastoma cell line were originally obtained from the American Type Culture Collection (ATCC) and cultured as described previously.The 67NR murine breast cancer cell line was originally provided by Dr. Fred Miller (Karamanos Cancer Institute, Detroit, MI) and was cultured as described previously., The generation of CAIX-negative MDA-MB-231 cells using CRISPR-mediated gene editing and the generation of CAIX-positive MDA-MB-231 cells through the re-introduction of a wild-type CAIX construct have been described previously.To generate 67NR cells constitutively expressing human CAIX (67NR-hCAIX), a human wildtype CAIX plasmid (gift from Dr. Jacques Pouyssegur, University of Nice, Nice, France) was transfected into 67NR cells using Lipofectmine 2000 as per the manufacturer's instructions. Transfected cells were selected using Zeocin and then subcloned for selection of cells with high levels of CAIX expression. For culture in normoxia, cells were incubated in a humidified incubator at 37°C with 5% CO 2 . For studies in hypoxia, cells were grown at 37°C in an atmosphere of 1% O 2 , 5% CO 2 , 94% N 2 in a humidified incubator inside a sealed workstation as described previously.Cell lines were routinely tested for mycoplasma contamination using the LookOut Mycoplasma PCR detection kit (Sigma; Cat. No. MP0035). The 67NR and MDA-MB-231 cell lines have been authenticated using short tandem repeat DNA profiling (DNA fingerprinting) by a commercial testing facility (Genetica). In addition, the cell lines were routinely tested for viability, morphology and hypoxia-induced endogenous CAIX expression or over-expression of CAIX.
## "in cell" caix activity assay
In cell CAIX catalytic activity assays were carried out as described previously with the following modifications.Assays were performed using 67NR murine breast cancer cells exogenously expressing human CAIX (67NR-hCAIX). For each replicate sample (n = 3/Ab), 5.0 × 10 5 cells were harvested with 0.5 mM EDTA in PBS and re-suspended in 200 µl CO 2 -free isotonic buffer (20 mM Hepes, 130 mM NaCl, 5 mM KCl, pH 8.0) in a 2 ml flat-bottom glass vial equipped with an 8 mm magnetic stir bar. 10 µl of a 50 µg/ml stock solution of the anti-CAIX Abs or isotype control Abs were added to the cell suspension and the sample was incubated with gentle stirring at room temperature for 45 min. The remainder of the assay was carried out on ice. 600 µl ice cold buffer was added to the cell-Ab mixture and a narrow pH electrode (Accumet, Cat # 13-620-850) was immersed in the sample and equilibrated for 3-5 min. 200 µl of ice-cold CO 2saturated water was added to initiate the assay and pH readings were immediately recorded at 2 to 5 s intervals. The final Ab concentration in the assay was 0.50 µg/ml. Data were analyzed using Graphpad Prism v8 (Graphpad Software). To determine the rate of spontaneous CO 2 hydration, measurements were performed on cell-free samples. The increase in hydration rate above the spontaneous rate is a measure of CA activity.
## Glycolytic rate assay
Metabolic flux was assessed using Seahorse XF technology as described previouslywith the following modifications. MDA-MB-231 human breast cancer cells engineered to constitutively express hCAIX (CAIX positive) were seeded at 2.0 × 10 4 cells/ well in a 96-well XF microplate (Agilent Technologies) and allowed to attach overnight. The following day, the cells were washed with complete assay medium (XF DMEM medium, pH 7.4, 10 mM glucose, 1 mM pyruvate, 2 mM glutamine; Agilent Technologies) and incubated in 200 µl/well complete assay medium containing 10 µg/ml m4A2 anti-hCAIX or IgG2b control antibodies for 60 min at 37°C without CO 2 . Media was then removed, 160 µl/well fresh media containing 10 µg/ ml m4A2 anti-hCAIX or IgG2b control antibodies was added, cell layer integrity was verified by microscopic observation and cells were subjected to a Seahorse XF Glycolytic Rate Assay (Agilent Technologies) according to instructions provided by the manufacturer. Assays were carried out using the Seahorse XFe96 Extracellular Flux Analyzer (Agilent Technologies). Data were acquired using Wave v2.6 (Agilent Technologies). After completion of the assay, cell nuclei were stained with HCS NuclearMask Red Stain (ThermoFisher Scientific,), cell layers were imaged using the Incucyte Zoom live cell imaging system (Essen Biosciences, Ann Arbor, MI) and cell nuclei were enumerated as a measure of cell number for normalization of metabolic data. Normalized data were exported to the Glycolytic Assay Report Generator for conversion of experimentally derived ECAR and OCR data into Proton Efflux Rate (PER) and calculation of specific glycolytic rate assay parameters. These parameters include basal PER (last PER measurement prior to injection of Rot/AA), basal glycolysis (last glycoPER measurement prior to injection of Rot/AA) and mitoOCR/glycoPER [(Last OCR measurement before first injection) -(Minimum OCR after Rot/AA injection)]/(Basal Glycolysis). Data for calculated parameters were exported to GraphPad Prism 8 (Graphpad Software) for statistical analysis. Statistical significance between two groups was assessed by unpaired Student's t-test. Data are presented as mean ± s.e.m. and P values less than 0.05 were considered statistically significant.
## Tunel assay
U87 MG glioblastoma cells cultured for 72 hours in hypoxia to upregulate CAIX expression. Cells were then seeded onto coverslips and incubated with 100 μg/ml m4A2 antibody in media containing 0.1% serum in hypoxia for an additional 72 hours. Mouse IgG2b (100 μg/ml) was used as an isotype control. Cells were then fixed and a TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labeling) assay (Roche Applied Sciences) was performed according to the instructions provided by the manufacturer to assess cell death. For quantification, 3-4 random fields from triplicate samples were imaged at 20x using a Zeiss Colibri inverted fluorescence microscope. TUNEL-positive cells and total cells were enumerated and the percentage of TUNEL positive cells was calculated using ImageJ (v1.48, National Institutes of Health, USA). Statistical significance between two groups was assessed by unpaired Student's t-test. Data are presented as mean ± s.e.m. and P values less than 0.05 were considered statistically significant.
## Abbreviations
|
Anti-Sa antibodies and antibodies against cyclic citrullinated peptide are not equivalent as predictors of severe outcomes in patients with recent-onset polyarthritis
The prognostic value of two antibodies targeting citrullinated antigens, anti-Sa and anti-cyclic citrullinated peptide (CCP), present at inclusion, was evaluated prospectively in a cohort of 165 consecutive patients with recent-onset or early polyarthritis (EPA) followed for up to 30 months. Patients were treated according to current Good Clinical Practice standards. Predefined outcomes were severe arthritis and persistent arthritis. At inclusion, a median of 3 months after disease onset, 133 (81%) patients fulfilled at least four American College of Rheumatology criteria for rheumatoid arthritis and 30 (18%) had erosive changes on radiographs of hands and feet. Diseasemodifying anti-rheumatic drugs were used in close to 80% of the patients at 30 months. Joint damage increased linearly over time, whereas disease activity declined markedly and remained low at each follow-up. Autoantibodies were identified in 76 (46%) patients: rheumatoid factor (RF) in 68 (41%), anti-CCP in 53 (33%), and anti-Sa in 46 (28%). All three antibodies were correlated, but anti-Sa antibodies best predicted severity at 18 and 30 months. RF and anti-CCP performed less well. For both outcomes, anti-Sa alone performed better than any combination of antibodies. The presence of any autoantibody identified about 50 to 60% of the patients with poor outcomes. In multivariate analysis, anti-Sa (odds ratio (OR) 8.83), the presence of erosions at inclusion (OR 3.47) and increasing age (OR 1.06/ year) were significantly associated with severity, whereas RF and anti-CCP were not significant predictors. Persistent arthritis was present in up to 84% of patients; autoantibodies were specific but poorly sensitive predictors of this outcome. We conclude that assays for antibodies against citrullinated antigens differ in their ability to predict poorer outcomes in patients with EPA. In our EPA cohort treated in accordance with current standards, detection of anti-Sa but not of RF or anti-CCP antibodies, in combination with clinical and radiological variables present at the first encounter, allowed the identification of a subgroup of EPA patients suffering more rapid and more severe joint damage over 30 months. ACR = American College of Rheumatology; BSA = bovine serum albumin; CCP = cyclic citrullinated peptide; CHUS = Centre hospitalier universitaire de Sherbrooke; CRP = C-reactive protein; DAS28-3 = Disease Activity Score using 28 joints and 3 variables (tender and swollen joint counts and CRP); DMARD = disease-modifying anti-rheumatic drug; ELISA = enzyme-linked immunosorbent assay; EPA = early (or recent-onset) polyarthritis; LR = likelihood ratio; MBP = myelin basic protein; M-HAQ = modified Health Assessment Questionnaire; OR = odds ratio; PADI = peptidylarginine deiminase; PBS = phosphate-buffered saline; RA = rheumatoid arthritis; RF = rheumatoid factor; SvH = Sharp-van der Heijde; WB = western blotting.Arthritis Research & Therapy Vol 7 No 3 Boire et al.
# Introduction
Rheumatoid arthritis (RA) is a chronic inflammatory arthritis that frequently starts at the peak of productive life and is a major cause of invalidity, morbidity, and premature mortality [bib_ref] The social and economic consequences of rheumatic diseases, Callahan [/bib_ref]. RA is characterized by an early inflammatory stage that is frequently responsive to disease-modifying anti-rheumatic R593 drugs (DMARDs). At a temporally undefined later stage, the RA process evolves towards pannus formation responsible for the joint destruction. Once established, pannus may progress on its own, independently of the apparent response to DMARDs. No available treatment can reverse significant joint damage. These observations gave rise to the notion of a therapeutic 'window of opportunity' during which the rheumatoid process would be more likely to be stopped or retarded [bib_ref] Rheumatoid arthritis. Pathophysiology and implications for therapy, Harris [/bib_ref]. This notion is supported by the fact that aggressive treatment of early RA decreases both mortality and long-term invalidity and can increase the rate of long-term remission [bib_ref] COBRA combination therapy in patients with early rheumatoid arthritis: long-term structural benefits..., Landewe [/bib_ref] [bib_ref] Retardation of joint damage in patients with early rheumatoid arthritis by initial..., Korpela [/bib_ref] [bib_ref] Impact of initial aggressive drug treatment with a combination of disease-modifying antirheumatic..., Puolakka [/bib_ref] [bib_ref] Radiological outcome after four years of early versus delayed treatment strategy in..., Van Aken [/bib_ref]. The identification, early into disease, of those patients likely to evolve rapidly to pannus formation and destructive/disabling arthritis would allow the most cost-effective use of expensive treatments and, reciprocally, would minimize unnecessary exposure of spontaneously remitting patients to the risks of aggressive treatments.
Sufficiently specific and sensitive prognostic markers that could be used with confidence in the individual patient with early or recent-onset polyarthritis (EPA) and recent-onset RA are still lacking [bib_ref] The longterm outcomes of rheumatoid arthritis: Work disability: a prospective 18 year..., Wolfe [/bib_ref] [bib_ref] Quantifying the exact role of HLA-DRB1 alleles in susceptibility to inflammatory polyarthritis:..., Thomson [/bib_ref] [bib_ref] Long-term outcome in rheumatoid arthritis: a simple algorithm of baseline parameters can..., Drossaers-Bakker [/bib_ref]. For example, even within patients fulfilling the 1987 revised classification criteria for RA of the American College of Rheumatology (ACR; formerly the American Rheumatism Association) [bib_ref] others: The American Rheumatism Association 1987 revised criteria for the classification of..., Arnett [/bib_ref] , chronic arthritis presents wide variations in response to treatments, degree of inflammation, and potential for joint destruction and functional impairment [bib_ref] How many types of patients meet classification criteria for rheumatoid arthritis?, Pincus [/bib_ref]. Classification criteria have even more limited value in predicting outcomes of patients with recent-onset polyarthritis [bib_ref] Estimating the incidence of rheumatoid arthritis: trying to hit a moving target, Wiles [/bib_ref]. A second challenge is the frequent occurrence of spontaneous remission in early polyarthritis present for up to 3 to 6 months. This relatively benign evolution is well documented, both in population-based studies [bib_ref] A four year follow up of suspected rheumatoid arthritis: the Tecumseh, Michigan,..., Mikkelsen [/bib_ref] [bib_ref] The prevalence of rheumatoid arthritis: follow up evaluation of the effect of..., O'sullivan [/bib_ref] and in cohort studies of patients with polyarthritis of recent onset [bib_ref] The prognosis of rheumatoid arthritis and undifferentiated polyarthritis syndrome in the clinic:..., Wolfe [/bib_ref] [bib_ref] Early undifferentiated connective tissue disease. IV. Musculoskeletal manifestations in a large cohort..., Alarcon [/bib_ref]. As a consequence, clinicians frequently adopt a watch-and-see attitude with patients during the first months of disease, delaying treatment with irreversible detrimental consequences [bib_ref] Retardation of joint damage in patients with early rheumatoid arthritis by initial..., Korpela [/bib_ref].
Recently, antibodies targeting determinants resulting from the deimination of peptidylarginine to peptidylcitrulline residues have been described in the serum of patients with RA [bib_ref] Citrulline is an essential constituent of antigenic determinants recognized by rheumatoid arthritisspecific..., Schellekens [/bib_ref]. These include antibodies targeting cyclic citrullinated peptide (CCP) [bib_ref] Autoantibody systems in rheumatoid arthritis: specificity, sensitivity and diagnostic value, Van Boekel [/bib_ref] and antibodies targeting in vivo citrullinated proteins, such as anti-keratin antibodies, antiperinuclear factor, anti-citrullinated (pro)filaggrin and anti-Sa/citrullinated vimentin [bib_ref] The Sa system: a novel antigen-antibody system specific for rheumatoid arthritis, Despres [/bib_ref]. The specificity of these antibodies for established RA, either rheumatoid factor (RF)-positive or RF-negative, and their presence in patient sera in the preclinical and early clinical phases of disease have been documented [bib_ref] Antibodies against cyclic citrullinated peptide and IgA rheumatoid factor predict the development..., Rantapaa-Dahlqvist [/bib_ref] [bib_ref] A combination of autoantibodies to cyclic citrullinated peptide (CCP) and HLA-DRB1 locus..., Berglin [/bib_ref]. The serum of most patients with diseases other than RA, either RFpositive or RF-negative, does not contain antibodies against citrullinated peptides or proteins [bib_ref] Autoantibody systems in rheumatoid arthritis: specificity, sensitivity and diagnostic value, Van Boekel [/bib_ref]. Assays to detect these anti-citrullinated peptides/proteins might therefore be useful to predict poor outcomes in patients with early polyarthritis [bib_ref] How to diagnose rheumatoid arthritis early: a prediction model for persistent (erosive)..., Visser [/bib_ref] [bib_ref] Anti-citrullinated protein/peptide antibody assays in early rheumatoid arthritis for predicting five year..., Meyer [/bib_ref] [bib_ref] Autoantibodies to cyclic citrullinated peptides predict progression to rheumatoid arthritis in patients..., Van Gaalen [/bib_ref] , although their added value relative to RF was found to be modest or even controversial [bib_ref] The prognostic value of the antiperinuclear factor, anti-citrullinated peptide antibodies and rheumatoid..., Van Jaarsveld [/bib_ref] [bib_ref] Rheumatoid arthritis associated autoantibodies in patients with synovitis of recent onset, Goldbach-Mansky [/bib_ref] [bib_ref] The prognostic value of anti-cyclic citrullinated peptide antibody in patients with recent-onset..., Kroot [/bib_ref] [bib_ref] Prognostic factors for remission in early rheumatoid arthritis: a multiparameter prospective study, Gossec [/bib_ref].
The purpose of our study was to assess the prognostic role of anti-Sa and of anti-CCP antibodies, relative to RF, in patients with polyarthritis evaluated and treated early after the onset of disease, irrespective of the fulfillment or not of diagnostic criteria for a specific disease entity. We now present an analysis of our cohort at 30 months of follow-up.
# Methods
## Patients
Consecutive adult patients with synovitis affecting at least three joints for at least 1 month and less than 12 months and evaluated at the Centre hospitalier universitaire de Sherbrooke (CHUS) were asked to participate to the study. The CHUS is the single regional rheumatology referral center for about 400,000 people, and is the site of practice of six of the seven rheumatologists in the area. Patients with bacterial or crystalinduced arthritis, patients with a defined connective tissue disease (looked for both clinically and by autoantibody testing, patients with systemic vasculitis according to ACR criteria [bib_ref] The American College of Rheumatology 1990 criteria for the classification of vasculitis..., The American College Of Rheumatology [/bib_ref] , and those who declined or were unable to consent were excluded. Failure to fulfill ACR criteria for RA [bib_ref] others: The American Rheumatism Association 1987 revised criteria for the classification of..., Arnett [/bib_ref] , the presence of skin or nail lesions typical of psoriasis or psoriasislike lesions, inflammatory bowel disease, clinical features suggestive of a spondylarthropathy [bib_ref] Calin A, the European Spondylarthropathy Study Group: The European Spondylarthropathy study Group..., Dougados [/bib_ref] , signs or symptoms suggestive of a connective tissue disease without fulfillment of specific ACR criteria, and the presence of positive HLA-B27 were not reasons for exclusion. Most of the included patients were regularly followed by rheumatologists and treated with the current approach of early and intensive treatment with DMARDs [bib_ref] Therapeutic strategies for rheumatoid arthritis, O'dell [/bib_ref]. Patients, rheumatologists, coordinating nurses and treating physicians remained blinded to the patients' HLA-DR, anti-CCP and anti-Sa status. Serum and DNA materials were coded to ensure confidentiality and blinding of investigators. All patients gave their informed consent and the ethics review board of the CHUS approved the study.
## Disease variables
A rheumatologist completed a structured physical examination, including 68 tender and 66 swollen-joint count assessments and identification of extra-articular manifestations. A trained nurse coordinator performed a structured interview at the inclusion visit and at each of the follow-up visits scheduled at 18, 30, 42 and 60 months after the onset of inflammatory arthritis. The time of onset of arthritis was assumed to be the month during which the patient indicated that joint symptoms/ signs had appeared or, in patients with previous musculoskeletal complaints (such as osteoarthritis), at the time that the signs or symptoms of inflammatory arthritis appeared, additive to the previous signs and symptoms. Scheduling the follow-up evaluations relative to the onset of symptoms rather than to the establishment of diagnosis sought to ensure that the cohort was more homogeneous in disease duration on follow-up.
Variables assessed included demographics, tender and swollen joint counts, duration of morning stiffness, use of DMARDs and corticosteroids at and between each visit, modified Health Assessment Questionnaire (M-HAQ) [bib_ref] Assessment of patient satisfaction in activities of daily living using a modified..., Pincus [/bib_ref] , erythrocyte sedimentation rate (according to the Wintrobe method), serum Creactive protein (CRP; lower limit of detection 3.0 mg/L, upper normal limit 8.0 mg/L), serum RF (latex agglutination, RapiTex RF; Dade Behring Inc, Newark, DE, USA; threshold for positive results set at 40 IU/ml on the basis of clinical experience and confirmed by its optimal prognostic value), genomic typing of the HLA-DR (see below), measure of anti-Sa antibodies (see below) and anti-CCP antibodies (QuantaLite™; Inova Diagnostics, San Diego, CA, USA; in accordance with the manufacturer's recommendations). Standardized radiographs of the hands and feet were obtained at inclusion and at each scheduled assessment. Joint space narrowing and erosions were scored by the Sharp-van der Heijde (SvH) method, with a maximum score of 448 [bib_ref] How to read radiographs according to the Sharp/van der Heijde method, Van Der Heijde [/bib_ref]. All radiographs were read in pairs in a known time sequence by a trained reviewer who was blinded to the corresponding clinical information. The smallest detectable difference was shown to be 5 units; this was assumed to be identical to the minimal clinically important difference [bib_ref] Detecting radiological changes in rheumatoid arthritis that are considered important by clinical..., Bruynesteyn [/bib_ref] [bib_ref] Progression of rheumatoid arthritis on plain radiographs judged differently by expert radiologists..., Bruynesteyn [/bib_ref]. Functional status was determined at each visit by using validated French-Canadian or English versions of the M-HAQ with a range of 0 to 3 (good functional status to maximal disablement). Disease Activity Score using 28 joints and 3 variables (tender and swollen joint counts and CRP; DAS28-3) was calculated with the appropriate formulas http:/ /www.das-score.nl [bib_ref] Modified disease activity scores that include twenty-eight-joint counts. Development and validation in..., Prevoo [/bib_ref]. A DAS28-3 score below 2.6 defined clinical remission [bib_ref] Value of Disease Activity Score 28 (DAS28) and DAS28-3 compared to American..., Balsa [/bib_ref] [bib_ref] Remission in rheumatoid arthritis: agreement of the disease activity score (DAS28) with..., Fransen [/bib_ref] , whereas a score above 3.2 indicated more than mild disease.
## Predefined outcomes
Persistent arthritis was defined as the presence of at least one joint with synovitis and/or the current use of DMARDs or at least 10 mg of prednisone equivalent per day [bib_ref] How to diagnose rheumatoid arthritis early: a prediction model for persistent (erosive)..., Visser [/bib_ref]. Fulfillment of RA criteria required at least four ACR criteria for RA. Severity required an M-HAQ score of at least 1.0 and/or belonging to the upper third of the SvH radiological score.
Genomic PCR typing at the HLA-DR and HLA-DQ Genomic DNA was extracted from leukocytes present in 2.5 ml of EDTA-treated whole blood using Wizard DNA extraction kit (Promega Corporation, Madison, WI, USA), then stored at 4°C. Genomic typing was performed by using PCR with sequence-specific primers specific for HLA class II molecules [bib_ref] Identification of the HLA-DRB1*04, -DRB1*07, and -DRB1*09 alleles by PCR amplification with..., Zetterquist [/bib_ref]. Primer sets for low-resolution typing of HLA-DR and HLA-DQ antigens, and high-resolution typing of DRB1*01, DRB1*04, DRB1*14 and DQB1 loci were from Pel-Freez Clinical Systems (Brown Deer, WI, USA). After amplification, PCR products (10 µl) were resolved in a 2% agarose gel containing ethidium bromide and detected by ultraviolet transillumination. Amplification patterns were interpreted in accordance with the manufacturer's instructions. HLA-DRB1 alleles encoding the amino acid sequences QRRAA (DRB1*0101, DRB1*0102, DRB1*0105, DRB1*0404, DRB1*0405 and DRB1*0408), QKRAA (DRB1*0401 and DRB1*0409) and RRRAA (DRB1*1001) were considered to encode the shared epitope.
## Anti-sa/citrullinated vimentin
The Sa antigen was first detected in human spleen/placenta extracts by western blotting (WB) [bib_ref] The Sa system: a novel antigen-antibody system specific for rheumatoid arthritis, Despres [/bib_ref] and subsequently identified as citrullinated vimentin. Most citrullinated argininerich proteins are adequate but not absolutely equivalent reagents for the detection of antibodies against citrullinated epitopes, as discussed in. Because access to human tissues and consistent extraction of the Sa antigen from tissues proved problematic, and commercial human vimentin is very expensive, we developed standardized and less expensive assays. We screened several cell lines and identified the ECV 304 endothelial cell line as rich in both vimentin and peptidylarginine deiminases (PADIs). Auto-citrullinated extracts from ECV 304 cells (see below) paralleled human placenta extracts for anti-Sa detection in WB. We then used this ECV 304based WB assay to monitor the performance of an in-house enzyme-linked immunosorbent assay (ELISA), designed with in vitro citrullinated bovine myelin basic protein (MBP). MBP and vimentin have similarly high proportions of arginine. Most positive results in MBP-ELISA correlated with positive anti-Sa results on the original WB method using human placenta.
Anti-Sa WB assay ECV 304 cells were cultured at 37°C under 5% CO 2 in IMDM (Sigma Chemical Co., St Louis, MO, USA) enriched with 10% fetal bovine serum containing 3.024 g/L sodium bicarbonate (pH 7.2). At confluence, cells were trypsinized, washed, incubated for 5 min on ice in 1 ml of buffer containing 40 mM Tris-HCl pH 7.5, 1 mM EDTA pH 8.0, 0.150 NaCl, and mechanically detached. The resultant cell pellet was submitted to three freeze-thaw cycles, and the supernatant was clarified by centrifugation. The concentration of protein was measured by the Bradford assay (Bio-Rad), and the extracts were stored at -80°C. In preparation for WB, the extracts (2.5 µg per tested serum) were auto-citrullinated for 3 hours at 37°C in 50 mM citrullination buffer (Tris-HCl pH 7.5, 10 mM CaCl 2 , 10 mM dithiothreitol). The reaction was stopped with EDTA (100 mM final concentration). An equal volume of 2× WB loading buffer was added to the citrullinated extract for long-term storage in aliquots at -20°C. WB was performed as described [bib_ref] The Sa system: a novel antigen-antibody system specific for rheumatoid arthritis, Despres [/bib_ref] , with ECV 304 native and citrullinated extracts being run in parallel. Sera were tested in duplicate at a dilution of 1:100 and results are expressed as either positive or negative.
## Anti-sa elisa
Bovine MBP (Sigma) was citrullinated in vitro by incubation for 3 hours at 37°C in citrullination buffer containing 0.2 U rabbit PADI 2 (Sigma) per µg of MBP. Microtiter plates (96 wells; Nunc Maxisorp, WWR International Ltd., Mississauga, ON, Canada) were coated overnight at 4°C with native or citrullinated bovine MBP at 1 µg per well. The plates were blocked for 1 hour at room temperature (20 to 25°C) with PBS containing 1% (w/v) BSA, then washed three times with PBS containing 0.05% Tween 20 (PBST). The plates were incubated for 1 hour at room temperature with 100 µl per well of human serum diluted 1:300 in PBS containing 1% (w/v) BSA (PBS-BSA), and then washed three times with PBST. Bound human IgG was detected with a goat anti-human IgG, alkaline phosphatase-labeled conjugate (Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA) diluted 1:1000 in PBS-BSA. The reaction was detected with 1 mg/ml p-nitrophenyl phosphate (Sigma) in substrate buffer (10% diethanolamine, 0.5 mM MgCl 2 , pH 9.8) for 20 min at room temperature. The absorbance, A, was read at 405 nm. Each sample was run in duplicate and in parallel on native MBP (N), on citrullinated MBP (C), and on BSA (B). Results were given by the equation A 405 = A C -A N . A positive test was defined as A 405 > 0.2. A known positive control serum was used on each plate for reagent quality control. The BSA control was used to detect spurious intrinsic individual serum stickiness for which some RA sera are notorious. When a serum gave high A 405 when used on BSA alone, a positive anti-Sa ELISA result had to be confirmed on WB to be reported as positive. All the anti-Sa assays described here were developed and performed on coded sera at the McGill University Autoimmune Research Laboratory.
# Statistical analysis
The principal independent variables were anti-Sa (MBP-ELISA), anti-CCP (positive if more than 20 U/ml), and RF (positive if at least 40 IU/ml) antibodies. The degree of association between anti-Sa and anti-CCP assays was measured by using the phi coefficient. Sensitivity, specificity and positive likelihood ratio (LR) were calculated for each of the three outcome measures at 18 and 30 months into disease and for each of the 13 following putative prognostic markers present at the inclusion visit: anti-Sa, anti-CCP, RF, sex, age (in years), duration of disease (in months), duration of morning stiffness (positive when at least 1 hour), symmetry of joint swelling, number of joints with synovitis, M-HAQ score (positive when at least 1.0), RA-associated HLA-DRB1 alleles ('shared epitope'), CRP levels (positive when higher than 8.0 mg/L), and an erosion component of the SvH score of at least 5. The predictive ability of anti-Sa, anti-CCP, and RF in the early identification of patients with each outcome was also studied by using the area under a receiver operating characteristics curve. Confidence intervals were calculated for these estimates. After these estimations, a logistic regression approach was used to estimate the independent contribution of RF, anti-Sa, and anti-CCP present at inclusion to predict the development of outcomes at 30 months. Each autoantibody was evaluated separately ACR, American College of Rheumatology; CCP, cyclic citrullinated peptide; CRP, C-reactive protein; DAS28-3, Disease Activity Score using 28 joints and 3 variables (tender and swollen joint counts and C-reactive protein); M-HAQ, modified Health Assessment Questionnaire; RA, rheumatoid arthritis; RF, rheumatoid factor; SvH, Sharp-van der Heijde. and in combination, after inclusion in the multivariate statistical model of prognostic factors listed above. A final statistical model was constructed with the stepwise logistic regression approach. By using the variables included in the final model, with and without the inclusion of anti-Sa and anti-CCP antibodies, the odds ratio (OR) was calculated to estimate the contribution of each marker adjusted for the other markers in the prediction of the severity outcome.
# Results
## Patient characteristics at study entry
Up to May 2004, 260 patients agreed to participate and were included. Eleven additional patients declined to participate or had very incomplete inclusion data and were excluded. At this time, 165 patients had reached the 30-month evaluation mark. Of these patients,5 had died and 11 failed to come to their scheduled appointment. We therefore report on the 149 patients (retention rate 90.3%) with a complete 30-month assessment.
The baseline characteristics at entry of the 165 patients are summarized in [fig_ref] Table 1: Baseline characteristics of the study population [/fig_ref]. The 16 patients who died or who were lost to follow-up did not differ significantly on any of the inclusion variables from the 149 who were followed up (data not shown). Men (42.4% of our patients) were slightly older than women (61.7 versus 58.0 years). These patients had a disease of short duration (mean 4.4 months, median 3 months). Nevertheless, at inclusion, 80.6% of the patients already fulfilled at least four ACR criteria for RA, 47% were functionally disabled (M-HAQ at least 1.0), and 18% had significant erosive changes (SvH erosion score of 5 or more).
The prevalence of autoantibodies was low. RF was found in 68 (41.2%) patients, anti-CCP in 53 (33.1%), and anti-Sa in 46 (28.0%). As reported previously [bib_ref] Autoantibody systems in rheumatoid arthritis: specificity, sensitivity and diagnostic value, Van Boekel [/bib_ref] [bib_ref] Rheumatoid arthritis associated autoantibodies in patients with synovitis of recent onset, Goldbach-Mansky [/bib_ref] , the presence of anti-CCP, RF, and anti-Sa was moderately to highly correlated (Kendall's tau-b > 0.6 for all comparisons). A total of 76 sera contained either one of RF, anti-CCP, or anti-Sa antibodies, and 35 sera contained all three. RF and anti-CCP co-existed in 46 patients, RF and anti-Sa in 41, and anti-Sa and anti-CCP in [bib_ref] Progression of rheumatoid arthritis on plain radiographs judged differently by expert radiologists..., Bruynesteyn [/bib_ref]. This means that, despite a good correlation between assays, 13 (19.1%) of the 68 sera with RF did not contain antibodies against citrullinated antigens, CCP or Sa. Interestingly, the degree of association between anti-Sa and anti-CCP (titer higher than 21 U/ml; phi coefficient 0.6881) was very similar to the degree of association between anti-Sa and high titer anti-CCP only (titer higher than 100 U/ml; phi coefficient 0.6339). This absence of variation of the degree of association with increasing titers suggests that anti-Sa and anti-CCP assays differ qualitatively. The shared epitope at the HLA
## Evolution of the clinical status from inclusion to the 18month and 30-month evaluations
Upon follow-up, the clinical status of the patients improved markedly [fig_ref] Table 2: Clinical assessments and outcomes at inclusion and during follow-up [/fig_ref]. The DAS28-3, the M-HAQ, and the swollen joint counts all significantly decreased. Despite this apparent control of clinical disease activity relative to inclusion, the radiological SvH score increased steadily at each follow-up, both in the total score and in its erosion component [fig_ref] Figure 1 Total: Sharp-van der Heijde [/fig_ref] According to our definition (see above and similar to that in [bib_ref] How to diagnose rheumatoid arthritis early: a prediction model for persistent (erosive)..., Visser [/bib_ref] , persistent arthritis was said to be present when swelling was detectable in at least one joint and/or when at least one DMARD or at least a moderate dose of corticosteroids was used at the time of follow-up. Thus, 88% and 84% of the patients were considered as having persistent arthritis at 18 and 30 months, indicating that long-term spontaneous remission was unusual in our EPA cohort, as reported previously in [bib_ref] Inflammatory polyarthritis in the community is not a benign disease: predicting functional..., Harrison [/bib_ref]. Close to half (namely 53 of 137 and 58 of 125, at 18 and 30 months, respectively) of the patients labeled 'persistent arthritis' had no swollen joints at each follow-up. The good clinical control of disease in our patients was also evident when using the DAS28-3 score: 76 (48.7%) had a score of less than 2.6 ('remission' [bib_ref] Value of Disease Activity Score 28 (DAS28) and DAS28-3 compared to American..., Balsa [/bib_ref] [bib_ref] Remission in rheumatoid arthritis: agreement of the disease activity score (DAS28) with..., Fransen [/bib_ref] at 18 months, and 82 (55.0%) at 30 months. Thus, more than half of our initial cohort was in remission as defined by DAS28-3 at 30 months, arguably the highest proportion of remission in a cohort of patients with EPA [bib_ref] Retardation of joint damage in patients with early rheumatoid arthritis by initial..., Korpela [/bib_ref].
Severe arthritis was present in 56 (38%) patients at 30 months. Our inclusive definition of severity (upper third of SvH score and/or M-HAQ ≥ 1.0) did not make any a priori assumption about the rate of damage progression in our patients, as this should be highly dependent on the characteristics of the patients included and (possibly) the treatments used. The threshold for the upper third of SvH score was calculated as a total score of more than 10 at 18 months and more than 15 at 30 months, that is, about 2% and 3.5%, respectively, of the maximal SvH score. Not surprisingly, functional impairment was the criterion most determinant for selection in the group with severe disease at inclusion, whereas radiological damage became the major reason for selection during follow-up.
Only 28 (18.79%) patients still fulfilled at least four ACR criteria for RA at 30 months. At inclusion or on follow-up, 20 patients fulfilled criteria for inflammatory rheumatic diseases other than RA: 13 had arthritis associated with skin psoriasis (present at inclusion in 12), 3 had benign sarcoid arthritis, 3 had spondylarthropathy, and 1 had scleroderma. Autoantibodies were present at low frequency in these patients: RF in four, anti-Sa in two, and anti-CCP in one. At each of the follow-up visits, more than two-thirds of the patients (namely 101 of 156 and 101 of 149 at 18 and 30 months, respectively) did not fulfill criteria for a specific diagnosis and might thus be classified as undifferentiated arthritis.
## Autoantibodies as predictors of poor outcomes
We first determined the sensitivity, specificity and positive LR of the presence of autoantibodies at onset as prognostic markers for predefined outcomes [fig_ref] Table 3: Autoantibodies as prognostic markers of poor outcomes [/fig_ref]. Anti-Sa stood out as the single moderately good marker (positive LR 2.16) to predict the development of severe arthritis. In this regard, anti-CCP (positive LR 1.38) and RF (positive LR 1.50) gave similarly poor results. It is remarkable that, despite the lower prevalence of anti-Sa in our cohort, its sensitivity for the severity outcome was very similar to that of RF and anti-CCP. Any combination of RF and/or anti-CCP with anti-Sa was no better than anti-Sa alone in predicting severe outcomes. In contrast, antibodies present at inclusion were specific but poorly sensitive predictors of persistent arthritis at 30 months. Owing to our inclusive a priori definition (see above and similar to that in [bib_ref] How to diagnose rheumatoid arthritis early: a prediction model for persistent (erosive)..., Visser [/bib_ref] , persistent arthritis was observed in 84% of the patients at 30 months. Thus, despite their strong positive LR, antibodies were not clinically helpful in predicting persistence in our cohort at the 30-month evaluation mark. We therefore looked at alternative definitions of persistent arthritis, such as nonremission defined by DAS28 (namely presenting a DAS28-3 score of at least 2.6). Non-remission defined by DAS28-3 was present in 45% of our aggressively DMARD-treated patients at the 30-month evaluation. None of the three antibodies was correlated with non-remission defined by DAS-28 (data not shown), suggesting that the presence of antibodies does not correlate with a poorer response to (conventional) DMARD treatments. Persistence defined as non-remission defined by DAS28-3 was therefore not used further. Finally, no autoantibody correlated with the fulfillment of RA criteria, mostly because of poor specificity. The best serological markers for the fulfillment of criteria for RA at 30 months were RF and anti-Sa, which gave low but very similar positive LR values (1.53 and 1.54, respectively). This result probably reflects the dampening impact of treatment on the clinical activity of arthritis.
In all three evaluated outcomes, RF had a higher sensitivity than anti-Sa, whereas anti-Sa had a higher specificity than RF (except for the fulfillment of RA criteria). Anti-CCP also had a slightly better specificity than RF for predefined outcomes. However, despite their high specificity for established RA, anti-CCP antibodies performed poorly. The same differences between anti-CCP and anti-Sa antibodies, although present, were less marked at the 18-month evaluation (data not shown). As noted above, anti-Sa alone performed better than any combination of antibodies in predicting severity. The conclusions were not significantly different when the analysis was restricted to patients fulfilling RA criteria at inclusion (data not shown). As noted in [fig_ref] Table 2: Clinical assessments and outcomes at inclusion and during follow-up [/fig_ref] , specific avoidance of DMARDs in anti-Sa patients could be discarded as a possible explanation for their more rapid radiographic deterioration, because these patients were as intensively treated as patients with RF.
## Multivariate prognostic models of severity
As illustrated in the multiple regression model for severity [fig_ref] Table 4: Odds ratio estimates of severe arthritis at 30 months from independent variables... [/fig_ref] , model A), anti-Sa (OR 8.832) clearly outperformed RF and anti-CCP. Indeed, the presence of anti-Sa conferred an even higher OR than erosion SvH abnormalities (OR 3.472), and was the most powerful independent predictive variable for the severity outcome. The OR of anti-Sa increased after adjustment for all other included variables. Although anti-Sa was the best individual marker for severity, a combination of anti-Sa, an SvH erosion score of at least 5 and increasing age best explained severe disease development. Worth noting is the absence of a HLA-DR 'shared-epitope', disability defined by M-HAQ, and the fulfillment of ACR criteria for RA from the short list of variables present at inclusion that were predictive of the predefined severe outcome.
However, anti-Sa antibodies are not widely available yet, and their presence in the model might have wiped out significant associations of RF and/or anti-CCP with severe arthritis. After the removal of anti-Sa from the analysis, neither RF nor anti-CCP was significantly predictive of severe arthritis [fig_ref] Table 4: Odds ratio estimates of severe arthritis at 30 months from independent variables... [/fig_ref] , model B). At that time, we suspected some heterogeneity among patients with anti-CCP. Receiver operating characteristics curves indicated that a cut-off of 61 U/ml of anti-CCP, Combinations of antibodies (rheumatoid factor (RF) and/or anti-Sa and/or anti-cyclic citrullinated peptide (anti-CCP)) are not shown, because none brought any improvement over the detection of anti-Sa alone. LR, likelihood ratio; RA, rheumatoid arthritis.
rather than 21 U/ml, would yield an improved positive LR of 2.01 for the severity outcome (data not shown). Similarly, after deletion of anti-Sa results and restriction to titers of anti-CCP higher than 60 U/ml, the predictive value of anti-CCP improved but remained below that of RF; neither reached a statistically significant level [fig_ref] Table 4: Odds ratio estimates of severe arthritis at 30 months from independent variables... [/fig_ref] , model C).
Finally, we tested the interaction of erosion SvH score and anti-Sa, the two major initial predictors of severe arthritis on the development of joint damage [fig_ref] Figure 1 Total: Sharp-van der Heijde [/fig_ref]. Each of the two predictors influenced the 30-month SvH score, with significant erosive damage at inclusion being the most determinant for the rate of progression of radiological damage. In patients in whom erosion SvH scores were abnormal at inclusion, the rate of yearly increase of the total SvH score was about 10 units. This rate was about 4 units/year in those whose initial erosion SvH scores were less than 5 units. These data support the importance of early damage as evidence for aggressive disease.
In the multiple regression model for persistence at 30 months (data not shown), disease duration for at least 4 months at inclusion was the single significant prognostic marker. No anti-body was statistically associated with the outcome of persistence at that time.
# Discussion
The main conclusion from our EPA study is that assays targeting different citrullinated antigens have distinct prognostic values for poor outcomes, in addition to their differences in sensitivity and specificity [bib_ref] Autoantibody systems in rheumatoid arthritis: specificity, sensitivity and diagnostic value, Van Boekel [/bib_ref]. When present, anti-Sa antibodies are useful markers of poor prognosis in EPA patients, even when rapidly treated intensively with conventional DMARDs. In our EPA population, anti-Sa outperformed anti-CCP antibodies for the prediction of each of the predetermined clinical outcomes, at least up to 30 months into disease. Despite their more specific association with RA [bib_ref] Autoantibody systems in rheumatoid arthritis: specificity, sensitivity and diagnostic value, Van Boekel [/bib_ref] [bib_ref] The diagnostic properties of rheumatoid arthritis antibodies recognizing a cyclic citrullinated peptide, Schellekens [/bib_ref] , anti-CCP antibodies did not carry any significant advantage over RF. It might thus be premature to consider replacing RF by anti-CCP antibodies in the clinical evaluation of EPA patients [bib_ref] Prospective two year follow up study comparing novel and conventional imaging procedures..., Backhaus [/bib_ref]. Differences between RF and anti-Sa antibodies were also significant. As a rule, RF proved to be the most sensitive assay for all outcomes, whereas anti-Sa was the most specific.
At first sight, the lower prevalence of anti-Sa (28% versus 41% for RF) seemed a significant disadvantage. However, the , and without anti-Sa but with a higher threshold of 61 U/ml for anti-cyclic citrullinated peptide (anti-CCP) (model C). Confidence limits are 95% Wald confidence limits. Disease duration was set as positive when present for at least 4 months. Morning stiffness was set as positive when it lasted for at least 60 min. The swollen joint count represents the contribution of each additional joint above the inclusion requirement of three. HLA shared epitope indicates the presence of at least one of the HLA DR alleles associated with RA, as listed in the Methods section. Erosion Sharp-van der Heijde (SvH) score was set as positive when the value was 5 or more. CRP, C-reactive protein; M-HAQ, modified Health Assessment Questionnaire; RF, rheumatoid factor. *P < 0.02; **P < 0.01; ***P < 0.005. sensitivity of anti-Sa for severe arthritis increased over time in our cohort, as 29%, 39%, and 45% of all severe patients had anti-Sa at inclusion, 18 months, and 30 months, respectively. This observation held true even though anti-Sa positive patients received as aggressive DMARD treatments as the other patients with persistent disease. The lack of statistically significant association of antibodies with persistence was probably due in part to the dual definition of persistence that we used, namely the presence of synovitis and/or being treated. In the absence of a more specific diagnostic test for persistence, we consider this definition to be the best currently available in treated recent-onset arthritis. The major limitation with our definition of persistence was the unexpectedly high (80%) frequency of ongoing use of DMARDs at 30 months. We expect that most patients in protracted remission will slowly taper their DMARDs over the planned follow-up of up to 5 years. The confounding influence of prolonged treatment on our definition of persistence should thus progressively attenuate over these additional 30 months, and the prognostic value of antibodies for persistence might then become assessable.
The second important conclusion is that autoantibodies were present early in only 50 to 60% of the patients who developed any of the predefined poor outcomes at 30 months. Autoantibodies present early into disease therefore characterize a large, but limited, subset of EPA patients with poor outcomes. Detection of autoantibodies in EPA patients must thus be used in combination with additional variables present at inclusion (for example elevated erosion SvH scores and increasing age) to best predict, and possibly prevent, the development of severe disease at 30 months (and beyond).
The third conclusion is that our results should be generally applicable to EPA patients evaluated by rheumatologists. Several reasons support that opinion. First, selection biases were minimized by the strict exclusion of patients with diseases known to have a good prognosis (such as crystal-induced arthritis and monoarthritis) and by a retention rate of 90% at 30 months, without individual missing data. Second, the patients were thoroughly evaluated with widely available tools (except for anti-Sa), early in their disease (median 3 months), and at consistent intervals relative to disease onset. Third, the predefined severity outcome included both functional disability and radiographic damage. Functional impairment and invalidity are the best indicators of severity, but they occur too late into disease to be useful prognostic indicators. Radiographic changes, such as those quantified with the SvH score, are therefore used as surrogates of severity in early disease. Because disability and radiographic damage are poorly correlated in early arthritis [bib_ref] Inflammatory polyarthritis in the community is not a benign disease: predicting functional..., Harrison [/bib_ref] [bib_ref] Disability in patients with early inflammatory polyarthritis cannot be 'tracked' from year..., Wiles [/bib_ref] , both contribute in relatively independent ways to the full picture of severity during the first years of chronic arthritis.
Our observations regarding the prognostic usefulness of anti-CCP in EPA patients are somewhat at variance with other reports. The reasons for differences in prognostic value between anti-CCP and anti-Sa assays therefore merit further discussion. A first possible explanation relates to deficiencies of the specific commercial anti-CCP assay used in this paper. That explanation is unlikely. We observed no significant differences in sensitivities or specificities between commercial anti-CCP assays (G. Boire, unpublished observations). Such differences would be unlikely, because all commercially available anti-CCP assays use the same antigen strips [bib_ref] Autoantibody systems in rheumatoid arthritis: specificity, sensitivity and diagnostic value, Van Boekel [/bib_ref] and differ only in technical aspects of the ELISA itself. Thus, the prognostic performance of the particular anti-CCP assay we used is likely to be generally applicable to commercial anti-CCP assays.
A second and more likely possibility for the poor performance of anti-CCP might reside in the design of the assay. The objectives of current anti-CCP assays are to attain the maximal sensitivity for patients with established RA or RA-like disease, while maintaining a reasonable specificity. This approach favors a low threshold to report positive results. Because EPA and RA are very heterogeneous diseases, a more appropriate design would be to identify subsets of patients with worse outcomes. In our cohort, most anti-CCP positive sera had high titers (mean 135.6 U/ml; median 116 U/ml) but, significantly, 11 patients had low or moderate titers (60 U/ml and below). The presence of low or moderate titers of anti-CCP at inclusion did not seem to be associated with severe arthritis at 30 months. Indeed, using a cut-off of 61 U/ml afforded the best OR estimate for the development of severe arthritis, although the predictive value of this level of anti-CCP was still not statistically significant [fig_ref] Table 4: Odds ratio estimates of severe arthritis at 30 months from independent variables... [/fig_ref] , model C). It was previously suggested that many non-RA sera with low anti-CCP titers bind similarly to citrullinated and non-modified peptides. These sera would present as false positive anti-CCP results. As a consequence, we suggest that the threshold for positive results of commercial anti-CCP assays should be adjusted upward to increase their prognostic value in EPA patients.
A third explanation for the prognostic discrepancy between anti-CCP and anti-Sa implies hypothetical qualitative differences between the anti-Sa and the anti-CCP assays, as suggested by the moderate degree of association between anti-Sa and anti-CCP results remaining constant across different cut-offs for positive anti-CCP (phi coefficient 0.6881 for titers higher than 21 U/ml, and 0.6339 for titers higher than 100 U/ ml). These qualitative differences between assays would translate into genuine consequent associations with outcomes. Antibodies present in sera from RA patients recognize citrullinated peptidic residues in the context of adjoining amino acid residues and of peptide conformation [bib_ref] Citrulline is an essential constituent of antigenic determinants recognized by rheumatoid arthritisspecific..., Schellekens [/bib_ref] [bib_ref] Autoantibody systems in rheumatoid arthritis: specificity, sensitivity and diagnostic value, Van Boekel [/bib_ref]. The short synthetic peptides used in anti-CCP ELISA were circularized and covalently bound to plastic strips to circumvent some of the problems inherent to the use of peptides in solid-phase assays [bib_ref] The diagnostic properties of rheumatoid arthritis antibodies recognizing a cyclic citrullinated peptide, Schellekens [/bib_ref]. This design might not be appropriate in very early polyarthritis, at a time when the immune response to citrullinated antigens is still being matured in vivo by modified arginine-rich proteins. At present it remains to be confirmed whether citrullinated proteins in general might perform better than CCP in ELISA at identifying severe subsets of EPA patients.
A fourth contributing factor to explain the poor performance of anti-CCP in our cohort is the influence of aggressive treatment on EPA prognosis. An earlier time of introduction, a higher intensity of use, and a more prolonged maintenance of DMARDs, with the objective of controlling disease, distinguish our current cohort from previous EPA cohorts established during the early 1990s. This more aggressive approach is illustrated by the high prevalence of DMARDs still used at 30 months, as well as by the marked decrease in disease activity over time, as assessed by the DAS28-3, the swollen joint counts, and the M-HAQ. Possibly because of this relatively intensive use of DMARDs, more than half of the patients were in clinical remission at 30 months. According to this hypothesis, the presence of anti-CCP antibodies would indeed identify patients with a poor natural course, as suggested in previous EPA or early RA cohorts [bib_ref] How to diagnose rheumatoid arthritis early: a prediction model for persistent (erosive)..., Visser [/bib_ref] [bib_ref] Anti-citrullinated protein/peptide antibody assays in early rheumatoid arthritis for predicting five year..., Meyer [/bib_ref] [bib_ref] Autoantibodies to cyclic citrullinated peptides predict progression to rheumatoid arthritis in patients..., Van Gaalen [/bib_ref] [bib_ref] Association between HLA class II genes and autoantibodies to cyclic citrullinated peptides..., Van Gaalen [/bib_ref]. In contrast, anti-CCP would fail to identify patients who do poorly when exposed relatively early to intensive treatment with current DMARDs. Such an effect of aggressive treatment was previously reported to explain the loss of HLA-DR association with severity in aggressively DMARD-treated patients [bib_ref] Early and aggressive treatment of rheumatoid arthritis patients affects the association of..., Lard [/bib_ref]. If this is true for anti-CCP, however, patients with anti-Sa would not exhibit the same generally good response to DMARDs or, alternatively, the association of anti-Sa with severe outcomes would be so large that the use of DMARDs would not completely erase it. Because intensive and effective DMARD treatments capable of slowing the progression of radiological and functional deterioration are increasingly being used in clinical practice, progressively longer periods of observation in larger samples of patients should therefore be required to delineate the true predictive value of potential prognostic markers.
Finally, our data stress the importance of intrinsic aggressiveness of arthritis in causing progressive joint damage. At inclusion, 18% of the patients already demonstrated clinically significant erosive changes. That represents a very early progression to the pannus phase, and an already missed opportunity to intervene during the optimal 'therapeutic window' in these patients. Severe outcomes at 30 months happened mostly in those patients who already had significant erosive joint damage at first evaluation. Early erosive changes are therefore an excellent surrogate marker for an aggressive arthritis. Whether the use of more sensitive imaging techniques such as ultrasound or magnetic resonance imaging [bib_ref] Ultrasonography in rheumatoid arthritis: a very promising method still needing more validation, Ostergaard [/bib_ref] would increase the sensitivity of detection without affecting its specificity for poor outcomes remains to be defined [bib_ref] Ultrasonography in rheumatoid arthritis: a very promising method still needing more validation, Ostergaard [/bib_ref] [bib_ref] Magnetic resonance imaging in the evaluation of bone damage in rheumatoid arthritis:..., Goldbach-Mansky [/bib_ref] [bib_ref] Ultrasonography of the metatarsophalangeal joints in rheumatoid arthritis: comparison with magnetic resonance..., Szkudlarek [/bib_ref]. In contrast, a serological marker such as anti-Sa antibod-ies, more highly associated with severe outcomes than RF or anti-CCP, and present at inclusion in a patient without detectable erosions, would probably represent a pre-pannus surrogate marker for aggressive arthritis. Such a marker would be extremely useful in clinical trials and in real-world practice as well, to select patients for appropriate treatment. In that situation, additional independent markers (for example inflammatory markers, possibly genetic markers) will still be needed if we are to attain a more complete prediction of the outcomes in individual EPA patients.
# Conclusion
This study reports the first direct comparison of RF, anti-CCP and anti-Sa antibodies as prognostic markers in a cohort of patients with recent-onset inflammatory polyarthritis treated according to current standards. In this cohort, anti-Sa antibodies were present at inclusion in 45% of the patients who subsequently developed severe outcomes at 30 months. Anti-Sa antibodies were the single serological marker independently predictive of poor outcomes and, together with the presence of joint erosions at inclusion and with increasing age, characterized a subgroup of patients with more rapid and more severe joint damage. In contrast, anti-CCP antibodies were not independently associated with severe outcomes in our patients. Using higher thresholds for positive results slightly improved the performance of the anti-CCP assay, but the predictive value of anti-CCP antibodies remained inferior to that of RF. Although anti-Sa and anti-CCP assays both use citrullinated antigens, our data suggest that the two assays differ in their ability to predict poor outcomes in patients treated aggressively early after the onset of inflammatory polyarthritis.
[fig] Figure 1 Total: Sharp-van der Heijde (SvH) score at inclusion and at 18 and 30 months Total Sharp-van der Heijde (SvH) score at inclusion and at 18 and 30 months. Mean values and 95% confidence limits are illustrated. The four curves represent groups of patients at inclusion who had an erosion SvH score of at least 5 and positive anti-Sa (15, 14, and 14 patients at each visit) (squares), an erosion SvH score of at least 5 and negative anti-Sa (15, 13, and 13 patients at each visit) (circles), an erosion SvH score of 4 or less and positive anti-Sa (32, 30, and 30 patients at each visit) (triangles), and an erosion SvH score of 4 or less and negative anti-Sa (103, 97, and 92 patients at each visit) (diamonds). [/fig]
[table] Table 1: Baseline characteristics of the study population (n = 165) [/table]
[table] Table 2: Clinical assessments and outcomes at inclusion and during follow-up [/table]
[table] Table 3: Autoantibodies as prognostic markers of poor outcomes [/table]
[table] Table 4: Odds ratio estimates of severe arthritis at 30 months from independent variables present at inclusion [/table]
|
The role of color in the perception of three-dimensional shape
## Lightness saturation r of rgb
## G of rgb
## B of rgb
[fig] Figure S2: Effects of specular reflections and biased sub-surface scattering distributions on realism and clarity of 3D surfaces. Related to Figure 2. A: [/fig]
[fig] Figure S3: The effect of illumination direction, albedo, color, and grayscale conversion on surface appearance. Related to Figure 2 and 3. A: The 7 levels of illumination direction used in Experiment 1. B: The same surfaces as in [/fig]
|
Limits on determining the skill of North Atlantic Ocean decadal predictions
This work is innovative and worth publication. It highlights the limitations of skill in the North Atlantic. This work is very rigorous and highlights the limitations of the skill compared with two ocean reanalysis. In those reanalysis, the density metrics in the Labrador sea is either temperature or salinity driven and thus the skill of the climate model or decadal predictions model is largely or not dependent on those.I will have few comments on the article.Line 43 andTable 1: You mention the two main types of predictions, which have the bias, removed in anomaly or FF method. Do the types of data or reanalysis or model used for initialization be important to?Line 73: Do you have some idea why the bcc-csm1-1 does not show this warming?Line 98-100: Do you have some explanation for this? Is it model dependent? Or from the initialization method?Line 113-117: In this work, two reanalysis are investigated which are both based on the NEMO ocean model. Maybe it could have been interesting to use also one using another ocean model such as those describe in Karspeck et al. (2015). You mention 5 predictions systems that use the NEMO model do you have similar conclusion with the prediction system using other ocean model?Line 125: How many years of good reanalysis do we need or a necessary for a skillful multi-annual predictions system? Line 126: Maybe not in this study but maybe from the literature, do you know if some reanalysis are better to represent processes in this region?Line 143-146: Maybe it could be nice to have a spread of this behavior with a large range of ocean reanalysis.In the conclusion, do you think you will get better agreement and skill if you look at a more integrated quantity such as the MLD or volume of dense water formed in the Labrador sea?Reviewer #2 (Remarks to the Author):Limits on determining the skill of the North Atlantic Ocean decadal predictonsThe authors investigate the mechanism of predictability in the Labrador Sea in 15 state-of-the-art climate prediction systems and two widely used reanalysis (ORAS4 and DP3). In particular they try to isolate the variable and the time scale responsible for the density change in the region that is known to influence the Atlantic Meridional Overturning Circulation. The period of study covers the period 1960-2010. Both reanalysis and hindcasts (exception made of one) show similar evolution Additional Points: Line 46 -the northwest corner is a name used for a specific region where the North Atlantic Current retroflects, and isn't thus the same as the Labrador Sea. I'd thus suggest using a different choice of words for describing the location of the Labrador Sea in the sub-polar gyre.References -many issues with lack of capitalizationThere is an error in the way citation display in the test ("author?" is appearing). See for example in the label of Table 1; L 325, 326 328, 334 etc … Line 80 units are missing.Could the axis range added for the regression ofFigure 2L 194, When referring to "control simulations"; are the authors referring to the pre-industrial simulation ? If so this is rather obvious because of global warning. Does the hindcasts revert to their historical simulation?Sentence starting L 286 is confusing, could it be re-phrase ?Line 279: This statement is not entirely correct because a prediction system consists of a model and its initialisation strategy. Here the hindcasts revert because the preferred variability of the model. The reversal is also not immediate; it takes < 5 years. I agree that : "Predictions systems revert to their innate preferred variability mode within less than 5 years."Reviewer #3 (Remarks to the Author):The authors have done a good job revising their manuscript. I only have a couple of minor comments at this stage.Line 50 -LSW contributes to the upper part of lower limb of the AMOC. The authors should be clear with their sentence here.Line 83 -I'd prefer less well observed to poorly observed.Line 154-158 -There seems to be an issue with this sentence.
for the heat content but there are large disagreement in the variability of the salt content. Observations are crucially lacking there and it is unclear which is correct. In one of the reanalysis the density changes are mainly driven by temperature variability while it is controlled by salinity variability in the other one. Such disagreement is also found in the hindcasts. Once the impact of initialisation dissipates, the controlling variable in the hindcast may revert (in the 2-5 year lead time) to its preferred mechanism of variability (estimated by looking to the pre-industrial simulation of the model). The variable controlling the density change have implication on the pattern of the meridional overturning circulation. The manuscript is interesting and cast new understanding on the potential limitation of current climate projection and prediction systems. It is very well written; entertaining and the figure are of high quality and comprehensive. However, some parts of the manuscript are unclear and would need further investigation/clarification. I would recommend that the authors address the comments below before the manuscript be accepted for publication.
L18 I disagree with the statement that the key metrics are independent of the details of the prediction system. Here the only 2 aspects of a prediction system area investigated: which model is used and whether full field or anomaly initialisation is used. Some other detail can be equally or more important (see comment below).
Sometime the authors use the accronym NASPG (L29) some other times NA SPG (L33).
L 52: There is no such thing as perfect observation. Observations will always have some degree of uncertainty.
L41: There may be earlier reference to that statement ? [fig_ref] Table 1: Considering the information provided in [/fig_ref] as an exhaustive description of the details of a prediction system is wrong. While the choice of the model or FF/Anom are important one, there are many other choice that are equally important. The parametrisation or the running of a model can have very large influence. Also on the initialisation side : Are a system assimilating data or are they initialised from a reanalysis performed with a different system. If data assimilation method is used, what method is used and what observations are assimilated ? [fig_ref] Fig 1: A lot of the analysis in the manuscript is based anomaly correlation... [/fig_ref] A lot of the analysis in the manuscript is based anomaly correlation of rho500. Correlation can be sometime misleading. Please show the time series of rho500 (e.g. in [fig_ref] Fig 1: A lot of the analysis in the manuscript is based anomaly correlation... [/fig_ref] for the hindcast and reference reanalysis.
On Fig2: - Why did the author use the control simulations with pre-industrial forcing and not the historical forcing simulation? Which of T or S drive the density change can evolve with global warming.
- Is the period of calculation for the regression comparable to that of the hindcast ? - Following on the first comment of one may expect that the driving variable for density change depends on the mean temperature of the area. Have the authors attempted to relate the driving variable that to the temperature bias of each system in the region ? Temperature bias are extremely stable ? There is also a relatively strong temperature difference between ORAs4 and DP3.
- Can the authors show the histogram of the regression value calculated over a running 50-year windows from the long PI simulation for each model in the supplementary material. I would be curious to know if some models transit from a temperature driven density to salinity driven density (or vise-versa) with time.
- Some more details: o The authors should estimate confidence interval for ORAS4 and DP3 regression estimate (L143). o The correlation of DP3 with ORAS4 and vice versa should be shown. It is very hard to see the difference between DP3 and ORAS4, I would recommend using a different symbol and if possible a color that is more different.
L90 Some of the details of DP3 assimilation are given here. If so it should be explained why the system differ from ORAS4. "DP3 assim assimilation" is redundant. Cross covariance is unclear (covariance are always cross). Does the author means "Use of covariance to estimate the multivariate updates".
L98 here the ensemble mean is used. With a predominance of temperature observations, one merely expect the ensemble spread of salinity to be larger and the ensemble mean flatter. Does the ensemble averaging influence the variability of salinity ? if so, maybe the median would be better suited ? L125 This is likely because there are sufficient data to controlled the system during the ARGO period (see for example Zhang, S., A. Rosati, and T. Delworth. "The adequacy of observing systems in monitoring the Atlantic meridional overturning circulation and North Atlantic climate." Journal of Climate 23.19 (2010): 5311-5324.) L140 Remove the extra coma L 169 do not use contraction for is not.
L204 Can you explain the behavior of FGOALS-g2 where the mean drift seems to increase over time ?
Section 5 is relatively confusing. Up to now we look at the bias-corrected hindcast and now we look at the impact of the density change w.r.t to year-to-year variability ? The ranges of density change between models are very different (on and for most models both temperature and salinity seems equally important. The authors should try to improve the clarity of this section.
L204 Are not MIROC5 and MIROC 4h belonging to that category as well ? They do not show such behavior in L213 MPI-ESM did seem to switch.
L242 add a coma after the Labrador Sea ?
In , it seems that MIROC 4h belongs to group1 L299 is the relaxation time scale the same for the atmosphere and sea ice ? L 302 Does the ensemble evolves from the previous assimilation cycle or are they initialized on the first iteration from the previous analysis mean + perturbations? It seems that the method is what is standardly call "bred vector data assimilation method" if so mention it and add a reference. Please indicate the localization radius (vertically and horizontally). L313 It is not surprising that ORAS4 underestimate salinity variability prior to the ARGO period because they uses a 1-year relaxation to climatology.
L 343 Please clarify the meaning of the sentence starting by "Specifically …"
In Supplementary , it would be most valuable if the number of observations available with time were added. If there are times with good observation covered (for example a hydrographic mission) in the Labrador Sea; the value could be reported on [fig_ref] Fig 1: A lot of the analysis in the manuscript is based anomaly correlation... [/fig_ref] Reviewer #3 (Remarks to the Author):
This paper aims to look at determining the skill of decadal predictability in the northern North Atlantic Ocean. This is an important region because of its links to the large scale Atlantic overturning circulation. These linkages also mean there is potential for predictability with up to a decade lead time. The authors examine this question using a large suite of CMIP5 models, as well as two ocean reanalysis products. The authors show that the predictive skill is not a functional of the predictive system but instead depends on the reanalysis, and more specifically how the interannual density variations are set in the underlying climate model and chosen reanalysis. This is an interesting paper, important for those wanting to use climate models to make decadal predictions. It brings up important issues related to bias and sensitivity that then will need to be further explored by the community. The paper is well written and the figure quality is reasonable. However, I do have a couple of major concerns that I think would need to be addressed by the authors before this manuscript is suitable for publications. These concerns, as well as more minor points, are detailed below.
The authors analyze the Labrador Sea, and define their study region in the caption to figure 1. Why do the authors pick a region that combines both the interior and the boundary currents, given those regions have very different properties and dynamics? Why not use an isobaths/interior based definition like many studies to allow for a focus on the interior region? Otherwise contrasting signals in the two regions might be cancelling each other out. Furthermore why the choice of the top 500 m for vertical depth averaging. Most observational studies generally consider ~200 m as the boundary between the fresher colder upper layer and the saltier and warmer layer below containing Irminger Water. The surface layer also strongly responds to the air-sea fluxes. See Straneo (2006) for a good example of such a discussion. Thus I worry that signals will be masked by the use of a 500 m layer that averages through watermasses with very different properties.
Additionally, in multiple places, the authors make statements like "poorly observed boundary current regions", "variability which is by definition unobserved", etc. Although more data is always more desirable, I think these statements are far too extreme. There are multiple annual sections that have been taken across both boundary currents, at least in summer, for longer than a decade. The main AR7W section has an even longer record. Long term mooring records exist at 53N. Argo floats have now been around for 15+ years. Additionally, past major experiments (such as the Labrador Sea Convection Experiments, the weather ships, etc.) mean that there are periods with extensive data that then allows some level of analysis of long term variability in the region. See recent papers by Yashayaev et al for example, which show mapped estimates of interior T and S variability back to the 1940s. Given that, I find the lack of an attempt to at least evaluate which reanalysis might be closer to the real world as a major limitation. I'm not arguing there is enough data that there might not be large potential errors. But to just wave it away as poorly unobserved without trying to examine the issue in detail is not correct either.
## Authors' comments:
We thank all three reviewers for their time in reviewing our work. These comments have proved useful in improving our manuscript. A point by point response to your comments is provided below -original comments in black and responses in red.
Note that all reviewer line numbers refer to the original manuscript. All response line numbers (red) refer to the Tracked Changes version of the updated manuscript.
## Reviewers' comments:
Reviewer #1 (Remarks to the Author):
This work is innovative and worth publication. It highlights the limitations of skill in the North Atlantic. This work is very rigorous and highlights the limitations of the skill compared with two ocean reanalysis. In those reanalysis, the density metrics in the Labrador sea is either temperature or salinity driven and thus the skill of the climate model or decadal predictions model is largely or not dependent on those. Thank you for reviewing our work! I will have few comments on the article.
Line 43 and [fig_ref] Table 1: Considering the information provided in [/fig_ref] : You mention the two main types of predictions, which have the bias, removed in anomaly or FF method. Do the types of data or reanalysis or model used for initialization be important to? We find no systematic relationships but it is hard to make broad groupings based on other factors -the ones we highlight in [fig_ref] Table 1: Considering the information provided in [/fig_ref] are already the broadest ways to characterise decadal prediction systems. Indeed, it is this difficulty that inspires us to approach the problem from a process-based standpoint (i.e. the driver of density variability). We have added a reference to the table where the reader can find further details of the systems investigated here.
Line 73: Do you have some idea why the bcc-csm1-1 does not show this warming? Further analysis suggests that the Labrador Sea in bcc-csm1-1 becomes increasingly disconnected from the wider subpolar gyre throughout the twentieth century. Instead, it appears to co-vary more strongly with the East Greenland Current, which is cooler and fresher than the wider SPG and would explain the cooling/freshening that is seen in bcc-csm1-1.
Line 98-100: Do you have some explanation for this? Is it model dependent? Or from the initialization method? This is due to the underlying model, as noted in the discussion of . It does not appear to be due to the initialisation methodology, at least in the broadest FF-versus-Anom sense.
Line 113-117: In this work, two reanalysis are investigated which are both based on the NEMO ocean model. Maybe it could have been interesting to use also one using another ocean model such as those described in . You mention 5 predictions systems that use the NEMO model do you have similar conclusion with the prediction system using other ocean model? We attempted to get hold of further full 3D monthly temperature and salinity data for the period 1960-2013+ for other reanalyses but were not able to procure it. However, as our analysis focuses on the drivers of density variability, and we have two reanalyses with opposing drivers, we feel this is sufficient to investigate the hindcasts in this context. Note that the focus of our study is the prediction systems (hindcasts) and not the reanalyses themselves. On the second point, we again find no systematic relationship (note added to main text). L134-135
Line 125: How many years of good reanalysis do we need or a necessary for a skillful multi-annual predictions system? Many decades according to other studies, which we have now cited. L143.
Line 126: Maybe not in this study but maybe from the literature, do you know if some reanalysis are better to represent processes in this region? We have investigated this somewhat in , which suggests that DP3-assim is a closer fit to the available observations than ORAS4. However, a more thorough analysis would need to be done to account for the uncertainties involved. L104-109.
Line 143-146: Maybe it could be nice to have a spread of this behavior with a large range of ocean reanalysis. As above, this would be nice but is outside the scope of this paper. We hope our results (for the prediction systems) will inspire others to analyse the same metrics in the ocean reanalyses.
In the conclusion, do you think you will get better agreement and skill if you look at a more integrated quantity such as the MLD or volume of dense water formed in the Labrador sea? Previous work has highlighted the very large diversity in MLDs in CMIP5 models and that MLDs can drift very rapidly in a prediction system (Menary and Hermanson, 2016) so it's not clear that such a metric would lead to improved agreement between the models. In addition, many modelling centres did not upload mixed layer diagnostics to the CMIP archive, which would make such an analysis very difficult.
## Reviewer #2 (remarks to the author):
Limits on determining the skill of the North Atlantic Ocean decadal predictions
The authors investigate the mechanism of predictability in the Labrador Sea in 15 state-of-the-art climate prediction systems and two widely used reanalysis (ORAS4 and DP3). In particular they try to isolate the variable and the time scale responsible for the density change in the region that is known to influence the Atlantic Meridional Overturning Circulation. The period of study covers the period 1960-2010. Both reanalysis and hindcasts (exception made of one) show similar evolution for the heat content but there are large disagreement in the variability of the salt content. Observations are crucially lacking there and it is unclear which is correct. In one of the reanalysis the density changes are mainly driven by temperature variability while it is controlled by salinity variability in the other one. Such disagreement is also found in the hindcasts. Once the impact of initialisation dissipates, the controlling variable in the hindcast may revert (in the 2-5 year lead time) to its preferred mechanism of variability (estimated by looking to the pre-industrial simulation of the model). The variable controlling the density change have implication on the pattern of the meridional overturning circulation. The manuscript is interesting and cast new understanding on the potential limitation of current climate projection and prediction systems. It is very well written; entertaining and the figure are of high quality and comprehensive. However, some parts of the manuscript are unclear and would need further investigation/clarification. I would recommend that the authors address the comments below before the manuscript be accepted for publication. Thank you for the nice comments! L18 I disagree with the statement that the key metrics are independent of the details of the prediction system. Here the only 2 aspects of a prediction system are investigated: which model is used and whether full field or anomaly initialisation is used. Some other detail can be equally or more important (see comment below). We have rephrased this to highlight we are looking at some of the details of the prediction system. Nonetheless, the broad initialisation method and underlying climate model are undoubtedly some of the first order details. L19-20
Sometime the authors use the acronym NASPG (L29) some other times NA SPG (L33). This has been fixed.
L 52: There is no such thing as perfect observation. Observations will always have some degree of uncertainty. This has been rephrased. L54 L41: There may be earlier reference to that statement ? If you mean Original Line 41, then the current reference predates these. If you mean Original Line 43 then the reference we use is appropriate for tropical storms, which we specifically mention.
## L71
, please clarify what all-valid year means. Do you mean " the average of all hindcast available at that year?
Yes we do. This has been clarified. L76-77 [fig_ref] Table 1: Considering the information provided in [/fig_ref] : Considering the information provided in [fig_ref] Table 1: Considering the information provided in [/fig_ref] as an exhaustive description of the details of a prediction system is wrong. While the choice of the model or FF/Anom are important one, there are many other choice that are equally important. The parametrisation or the running of a model can have very large influence. Also on the initialisation side : Are a system assimilating data or are they initialised from a reanalysis performed with a different system. If data assimilation method is used, what method is used and what observations are assimilated? Similar to the comment for L18, we are merely highlighting some of the first order measures of the systems and do not intend for the table to be exhaustive. We have added text to refer the reader to where they can find more detailed information (see . It is also important to note that we are not specifically testing FF v Anom (for example) but are testing emergent properties of the decadal prediction systems. We show the details in [fig_ref] Table 1: Considering the information provided in [/fig_ref] to orient the reader and to highlight that we are aware of the first order descriptors of the systems. - Why did the author use the control simulations with pre-industrial forcing and not the historical forcing simulation? Which of T or S drive the density change can evolve with global warming. We use control simulations in order to characterise the underlying preferred nature of the models, in the absence of any forcings. Nonetheless, other appropriate measures would be a twentieth century control, or the transient simulations, as suggested. Nonetheless, for the period 1960-2013, we find no systematic changes in the historical simulations (not shown).
- Is the period of calculation for the regression comparable to that of the hindcast ? It uses the full period, but we have now added a supplementary figure investigating the stationarity of this relationship (see also below).
- Following on the first comment of one may expect that the driving variable for density change depends on the mean temperature of the area. Have the authors attempted to relate the driving variable that to the temperature bias of each system in the region ? Temperature bias are extremely stable ? There is also a relatively strong temperature difference between ORAs4 and DP3. We tested this prior even to testing the relationship with variability. We find no systematic relationship with the mean state. For anomaly-assimilation methods, the models are already at/around their mean state. For full-field assimilation, this is likely because, although the models drift after initialisation, they do not reach their preferred (i.e. control) mean state within 5 years, and in fact are generally still not close to it. This is related to the fact that the control mean states are actually quite well separated (see [fig_ref] Fig 1: A lot of the analysis in the manuscript is based anomaly correlation... [/fig_ref] of "Exploring the impact of CMIP5 model biases…" http://onlinelibrary.wiley.com/doi/10.1002/2015GL064360/full) - Can the authors show the histogram of the regression value calculated over a running 50-year windows from the long PI simulation for each model in the supplementary material. I would be curious to know if some models transit from a temperature driven density to salinity driven density (or vise-versa) with time. We have added this as requested .
- Some more details: o The authors should estimate confidence interval for ORAS4 and DP3 regression estimate (L143). We have added this as requested o The correlation of DP3 with ORAS4 and vice versa should be shown. It is very hard to see the difference between DP3 and ORAS4, I would recommend using a different symbol and if possible a color that is more different. The (lack of) correlation can now be inferred from . We have changed the symbols so as to make this figure clearer.
L90 Some of the details of DP3 assimilation are given here. If so it should be explained why the system differ from ORAS4. "DP3 assim assimilation" is redundant. Cross covariance is unclear (covariance are always cross). Does the author means "Use of covariance to estimate the multivariate updates". We have fixed "DP3-assim assimilation". We mean that temperature is allowed to co-vary with salinity, rather than just with temperature. This has been clarified. L97-98 L98 here the ensemble mean is used. With a predominance of temperature observations, one merely expect the ensemble spread of salinity to be larger and the ensemble mean flatter. Does the ensemble averaging influence the variability of salinity ? if so, maybe the median would be better suited ? We recomputed [fig_ref] Fig 1: A lot of the analysis in the manuscript is based anomaly correlation... [/fig_ref] and 2 using just the first ensemble member and found little change. Note that the hindcasts are free-running and so although the temperature initialisation may be better (i.e. closer to reality) it does not necessarily follow that the ensemble member evolution will be smaller in temperature than in salinity. This has been changed.
L204 Can you explain the behavior of FGOALS-g2 where the mean drift seems to increase over time ? This implies that the salinity (or rather the salinity component of the linearised density EOS) is showing an exponential drift with time. Given the importance of convection in this region, it is possible that the model is systematically initialised in a state not conducive to convection and that this spins up over time, with a feedback on to salinity. Unfortunately, there is no MLD data provided in order to test this hypothesis. However, we also note that the FGOALS-g2 has a strong AMOC when free-running ("Oceanic climatology in the coupled model FGOALS-g2: Improvements and biases" Lin et al. 2013) driven by strong mixing, which is consistent with the hypothesis above.
Section 5 is relatively confusing. Up to now we look at the bias-corrected hindcast and now we look at the impact of the density change w.r.t to year-to-year variability ? The ranges of density change between models are very different (on and for most models both temperature and salinity seems equally important. The authors should try to improve the clarity of this section.
- For , this is largely an artefact of showing mean state and variability changes on the same scales. Nonetheless, comparison of with either or 4 is consistent. Note also that the "density-driver" can show small but important changes in this context: temperature and salinity anomalies generally co-vary in this region such that they are warm/saline or cool/fresh. As such, a small change in the characteristic density changes associated with T/S can have a very large impact if one now becomes bigger than the other. - For the general clarity of the section, we have reworded this section.
L204 Are not MIROC5 and MIROC 4h belonging to that category as well ? They do not show such behavior in The MIROC models both show a shift to increasingly temperature-driven density variability . This is consistent with Fig 2 panels c and/or d, in which the MIROC models show increasingly temperature-driven (and decreasingly salinity-driven) density variability, even if not a complete sign change.
L213 MPI-ESM did seem to switch. Similar to the above, in MPI-ESM-MR shows increasingly salinity -driven density variability, although there isn't a sign change here (i.e. a crossing of the solid red and blue lines). In , an increase in salinity-driven density variability is again seen, although here the sign does change. Nonetheless, in both figures the tendency is the same and consistent. In addition, it is the power of multimodel analyses such as ours that the overall signal-to-noise is improved compared to single models in which there are undoubtedly other important processes also at play.
L242 add a comma after the Labrador Sea ? This has been done.
In , it seems that MIROC 4h belongs to group1 It does but there was no AMOC data on the CMIP5 archive for this model and so it doesn't contribute to the streamfunction plots. Only models that uploaded streamfunction data are used in -f, which fortunately results in an equal number of models in each group (5 in each).
L299 is the relaxation time scale the same for the atmosphere and sea ice ? No it isn't -this has been clarified. L323-324 L 302 Does the ensemble evolves from the previous assimilation cycle or are they initialized on the first iteration from the previous analysis mean + perturbations? It seems that the method is what is standardly call "bred vector data assimilation method" if so mention it and add a reference. Please indicate the localization radius (vertically and horizontally). It is not a bred vector method. We have attempted to clarify our description further.
L313 It is not surprising that ORAS4 underestimate salinity variability prior to the ARGO period because they uses a 1-year relaxation to climatology. We have now added a comment on this in the main text. L101-102 L 343 Please clarify the meaning of the sentence starting by "Specifically …" Lead time dependent bias correction for a hindcast must be done against a baseline reanalysis. Here, we mean that we have done this against ORAS4, but that one could use DP3-assim and get very similar results. This has been clarified. This paper aims to look at determining the skill of decadal predictability in the northern North Atlantic Ocean. This is an important region because of its links to the large scale Atlantic overturning circulation. These linkages also mean there is potential for predictability with up to a decade lead time. The authors examine this question using a large suite of CMIP5 models, as well as two ocean reanalysis products. The authors show that the predictive skill is not a functional of the predictive system but instead depends on the reanalysis, and more specifically how the inter-annual density variations are set in the underlying climate model and chosen reanalysis. This is an interesting paper, important for those wanting to use climate models to make decadal predictions. It brings up important issues related to bias and sensitivity that then will need to be further explored by the community. The paper is well written and the figure quality is reasonable. However, I do have a couple of major concerns that I think would need to be addressed by the authors before this manuscript is suitable for publications. These concerns, as well as more minor points, are detailed below. We are pleased that you found the paper interesting and appreciate your time in reviewing it.
The authors analyze the Labrador Sea, and define their study region in the caption to figure 1. Why do the authors pick a region that combines both the interior and the boundary currents, given those regions have very different properties and dynamics? Why not use an isobaths/interior based definition like many studies to allow for a focus on the interior region? Otherwise contrasting signals in the two regions might be cancelling each other out. Furthermore why the choice of the top 500 m for vertical depth averaging. Most observational studies generally consider 200 m as the boundary between the fresher colder upper layer and the saltier and warmer layer below containing Irminger Water. The surface layer also strongly responds to the air-sea fluxes. See Straneo (2006) for a good example of such a discussion. Thus I worry that signals will be masked by the use of a 500 m layer that averages through watermasses with very different properties. We used this region because it has been used on several previous occasions in the literature, in particular in the current line of work (Menary et al., "Exploring the impact of CMIP5…"; Menary, Hermanson, Dunstone, "The impact of Labrador Sea…"). In addition, in the latter paper we tested the effect of using just the Central Labrador Sea, as defined by and found it made little difference to the relationships between density and either the salinity or temperature components. Nonetheless, to further test the effect of both the use of an interior region and a more restricted depth range we have recomputed the density are just its temperature and salinity components on comparable figures for both the original region top 500m (top) and Central Labrador Sea top 200m (bottom) in both reanalyses (Supplementary . The figure legends highlight the correlation between the density (black lines) and its components (salinity component in red, temperature component in blue). When switching from the original (top) to interior (bottom) definitions it can be seen that DP3-assim remains strongly salinity driven (high correlation with ρ S ). Likewise, ORAS4 remains temperature-driven (higher correlation with ρ T than with ρ S ) but the correlation is slightly weakened by the different box choice. As such, the general result (namely that DP3-assim is primarily salinity driven and ORAS4 is primarily temperature driven) remains. Additionally, in multiple places, the authors make statements like "poorly observed boundary current regions", "variability which is by definition unobserved", etc. Although more data is always more desirable, I think these statements are far too extreme. There are multiple annual sections that have been taken across both boundary currents, at least in summer, for longer than a decade. The main AR7W section has an even longer record. Long term mooring records exist at 53N. Argo floats have now been around for 15+ years. Additionally, past major experiments (such as the Labrador Sea Convection Experiments, the weather ships, etc.) mean that there are periods with extensive data that then allows some level of analysis of long term variability in the region. See recent papers by Yashayaev et al for example, which show mapped estimates of interior T and S variability back to the 1940s. Given that, I find the lack of an attempt to at least evaluate which reanalysis might be closer to the real world as a major limitation. I'm not arguing there is enough data that there might not be large potential errors. But to just wave it away as poorly unobserved without trying to examine the issue in detail is not correct either.
1. We have toned down the statements on the paucity of observations. Our aim was not to denigrate the observational network but merely to note that the Labrador Sea is a harder region to make measurements in than either other parts of the ocean or land/atmosphere. In addition, as noted above, in our previous work we made a comparison between our present definition of the Labrador Sea and the interior definition of Yashayaev & Loder and found little difference in the reanalyses. L54. L84. L92. L104. L138 2. In we have subsampled the raw (quality controlled) observational profiles to compare against the two reanalyses. We have also added information on the observational density to this figure. In addition, we have added further comment on which (if either) of the reanalyses performs better against the available observations. We find that DP3-assim is a better fit against the observations and have now mentioned this specifically. A more in-depth comparison between the reanalyses and available observations in this region would be worthwhile but would be a whole paper in itself. L104-109 3. We have added an additional supplementary figure (Supplementary that explores the effect in the reanalyses of subsampling to a more interior and nearer-surface region. This is discussed in response to the point above. L164-167
Additional Points: Line 46 -the northwest corner is a name used for a specific region where the North Atlantic Current retroflects, and isn't thus the same as the Labrador Sea. I'd thus suggest using a different choice of words for describing the location of the Labrador Sea in the sub-polar gyre. This has been changed.
References -many issues with lack of capitalization These have been fixed
## Limits on determining the skill of the north atlantic ocean decadal predictons
It is the second review iteration of the manuscript. While the authors have addressed some of my concerns, I still disagree with some of the statements. Furthermore figures in the supplementary material may have highlighted a major concern in the dynamical feasibility of one of the reanalysis product (DP3 assim). In my view the manuscript is still not ready for publication, but has the potential to make a nice contribution to Nature Communications.
I still disagree with the broad statement starting L 13: "Skill in key metrics …, are largely independent of details of the prediction system such as model or initialisation." In my view the authors show the contrary. First, two reanalyses products using a similar model version but different initialisation strategy (MOSORA and NEMOVAR) suggest that density variability is controlled by a different state variable. The difference seems to be routed in the details of the covariance matrix of the data assimilation methods (so in the initialisation). Second, I do not consider all models that are based on the NEMO community model to be the same model. Using a different sea ice compartment or using different parameters changes completely the behaviour of the model. This is nicely exemplified by your of the supplementary material, where all NEMO based models depict very different behaviour. In this sense key metric are thus not independent of details in the models.
The Supplementary shows the density variability of the two reanalyses. I find it extremely worrying that the variability in DP3-assim is mostly explained by the density-increasing trend.
There is no sign of the notorious 1995 SPG shift. Density changes there are supported by observations (e.g. Van Aken et al. 2011 using tracer and temperature and salinity observations), altimetry data, proposed mechanism of variability (Yeager and Robson 2017 for a review on that topic). This merely suggests that the choices in MOSORA for the covariance between T and S are not suitable for the region studied. It was for instance advanced in Counillon et al. 2016 that in the Labrador Sea, the covariance between T and S changes with the phase of the SPG. I acknowledge that finding which of the reanalysis is most realistic is not the main scope of the paper, but in this case it seems that DP3-ASSIM is suspicious, which is important for the following understanding of the paper.
It is said that the reanalysis agrees well in the recent well observed period (L 129) but they are not for density. I strongly insist that the density evolution of the two reanalyses is shown in the main manuscript, since most of the following figure are based on that time series. I agree that salinity observations are lacking making it is hard to identify which is more realistic. What about altimetry data? Which of the reanalyses and prediction systems (temperature driven or salinity driven ones) shows best skill in explaining the variability of the sea surface height in the NA subpolar gyre region.
## Reviewers' comments:
Reviewer #2 (Remarks to the Author):
Limits on determining the skill of the North Atlantic Ocean decadal predictons It is the second review iteration of the manuscript. While the authors have addressed some of my concerns, I still disagree with some of the statements. Furthermore figures in the supplementary material may have highlighted a major concern in the dynamical feasibility of one of the reanalysis product (DP3 assim). In my view the manuscript is still not ready for publication, but has the potential to make a nice contribution to Nature Communications.
I still disagree with the broad statement starting L 13: "Skill in key metrics …, are largely independent of details of the prediction system such as model or initialisation." In my view the authors show the contrary. First, two reanalyses products using a similar model version but different initialisation strategy (MOSORA and NEMOVAR) suggest that density variability is controlled by a different state variable. The difference seems to be routed in the details of the covariance matrix of the data assimilation methods (so in the initialisation). Second, I do not consider all models that are based on the NEMO community model to be the same model. Using a different sea ice compartment or using different parameters changes completely the behaviour of the model. This is nicely exemplified by your figure 4 of the supplementary material, where all NEMO based models depict very different behaviour. In this sense key metric are thus not independent of details in the models. "I do not consider all models that are based on the NEMO community model to be the same model" -we agree. Indeed, this is exactly our point -prediction systems based on the same ocean submodel do not always trend towards ORAS4. Their behaviour is "independent" of the ocean submodel but depends on an emergent property that is their T/S variability.
We find ourselves agreeing entirely with your statements in the above paragraph despite having reached opposite conclusions.
Part of the confusion may be that when we say "skill" at any point in the manuscript we are referring to the skill in the hindcasts as measured against these reanalyses, NOT the skill of the reanalyses against independent observations. The latter would be an entirely different paper.
Most importantly, in our analysis, it does not matter what the initialisation of the hindcasts is. All that is important are two things: 1) That a multi-year free-running hindcast is conducted, and 2) that an unforced, long control simulation using the same/similar model has also been conducted.
Essentially, we then compare these two sets of simulations and find that the skill of the hindcast system (assessed against reanalysis) depends not on details of the hindcast system but on whether its control simulation looks similar to the reanalysis to which one compares.
If we have understood your argument correctly, it is (very briefly):
1. Skill depends on the reanalysis chosen (and whether the control is like this) 2. The reanalysis can also be called the initialisation 3. Thus skill depends on the initialisation However, it is the link between 2 and 3 that does not hold. The reanalyses only act as the initialisation for a few of the models, and as previously noted, even in those (e.g. DP3-assim and DePreSys3 hindcasts) they can result in poor skill when both are compared.
The Supplementary shows the density variability of the two reanalyses. I find it extremely worrying that the variability in DP3-assim is mostly explained by the density-increasing trend. There is no sign of the notorious 1995 SPG shift. Density changes there are supported by observations (e.g. Van Aken et al. 2011 using tracer and temperature and salinity observations), altimetry data, proposed mechanism of variability (Yeager and Robson 2017 for a review on that topic). This merely suggests that the choices in MOSORA for the covariance between T and S are not suitable for the region studied. It was for instance advanced in Counillon et al. 2016 that in the Labrador Sea, the covariance between T and S changes with the phase of the SPG. I acknowledge that finding which of the reanalysis is most realistic is not the main scope of the paper, but in this case it seems that DP3-ASSIM is suspicious, which is important for the following understanding of the paper. Differences in trends in this location depend greatly on the specifics of the location chosen, such as the depth range and horizontal extent (cf. papers cited below). As such, we are not surprised that the density evolution in our Figures does not precisely match those in different regions. In response to the previous reviews, we added supplementary figures exploring this sensitivity . Nonetheless, what is most important is that all of our work with 1) reanalyses, 2) hindcasts, and 3) control simulations are self consistent in using the same location.
On the specific question of whether DP3-assim should be considered suspect in this region:
We have extended Supp to include now salinity, temperature, and density. Our analysis suggests that when using independent observations in precisely the same region, rather than being suspect, DP3-assim appears to reproduce these subsampled observations much better than ORAS4. That is -density changes that are supported by observations look a lot like DP3-assim, when using the same definitions of depth range and horizontal extent.
It is said that the reanalysis agrees well in the recent well observed period (L 129) but they are not for density. I strongly insist that the density evolution of the two reanalyses is shown in the main manuscript, since most of the following figure are based on that time series.
We have now included the density evolution in [fig_ref] Fig 1: A lot of the analysis in the manuscript is based anomaly correlation... [/fig_ref]. Note that even if one or other of the reanalyses show an trend in density, this is removed during the lead-time dependent bias correction (when assessing the skill of the hindcasts). I agree that salinity observations are lacking making it is hard to identify which is more realistic. What about altimetry data? Which of the reanalyses and prediction systems (temperature driven or salinity driven ones) shows best skill in explaining the variability of the sea surface height in the NA subpolar gyre region. We feel the present manuscript is already long enough and as previously stated and acknowledged -assessing which reanalysis produce is more realistic is not the goal of this paper. We note that DP3-assim does not assimilate SSH data (but the skill has been assessed in Roberts et al. "On the Drivers and Predictability of Seasonal-to-Interannual Variations in Regional Sea Level" (2015)). ORAS4 does assimilate SSH. As such a comparison of the two reanalysis products would be somewhat unfair. We have previously assessed the skill of the DePreSys3 hindcasts against the DP3-assim in dynamic sea level in the same region and found that this followed the skill in density, i.e. it reverted to the underlying model. The mechanisms were the same as described in the present paper. "The impact of Labrador Sea temperature and salinity variability on density and the subpolar AMOC in a decadal prediction system" (2016).
There is an error in the way citation display in the test ("author?" is appearing). See for example in the label of Could the axis range added for the regression of The figure is already quite busy -but we have added a note on the range in the caption. The ranges are fixed and the same in all the inlaid panels.
L 194, When referring to "control simulations"; are the authors referring to the pre-industrial simulation? If so this is rather obvious because of global warming. Does the hindcasts revert to their historical simulation? Yes, we are referring to the PI Control simulations. The difference between the PI Control simulations and the historical simulations is much smaller than the difference between either the PI Control (or historical) and the hindcasts/reanalyses. This may be because the historical simulations are initialised from the PI Controls but the ocean state has not changed much and has thus retained the underlying drifts/biases. As such, the results are insensitive to the choice of baseline (control or historical). We choose PI Control as the baseline because there are no transient forcings to consider and because we are using it to understand the underlying nature of the models.
Sentence starting L 286 is confusing, could it be re-phrase ? This has been rephrased. L291-293
Line 279: This statement is not entirely correct because a prediction system consists of a model and its initialisation strategy. Here the hindcasts revert because the preferred variability of the model. The reversal is also not immediate; it takes < 5 years. I agree that : "Predictions systems revert to their innate preferred variability mode within less than 5 years." We agree that a prediction system includes both the initialisation strategy and the submodels. Despite the differing initialisation strategies, we find no link between them and the skill. As such, the skill is independent of the initialisation strategy (for the metrics we have investigated and for the broad definitions of strategy (i.e. FF or Anom) we have considered). Nonetheless, we have slightly rephrased this sentence. L279-286. See also our general response, above.
Reviewer #3 (Remarks to the Author):
The authors have done a good job revising their manuscript. I only have a couple of minor comments at this stage.
Line 50 -LSW contributes to the upper part of lower limb of the AMOC. The authors should be clear with their sentence here. This has been clarified. L43
Line 83 -I'd prefer less well observed to poorly observed. This has been changed. L83
Line 154-158 -There seems to be an issue with this sentence. This has been fixed.
I am sorry to say that I am not satisfied with the answer provided. Again, I really like the paper and I think it contains original material that is valuable for the community, but I disagree with some statements (for example L251, L280), which are nicely summarised by the sentence in the abstract. L 13: "Skill in key metrics …, are largely independent of details of the prediction system such as model or initialisation."
First, this statement is too vague. What is the meaning of model, initialisation, skill (matching the reanalysis used initially) and at which time scale this statement is correct. It is important because such statement can very easily be misinterpreted.
Second and more problematic, I do not feel that sufficient evidence has been shown to hold this statement.
- The skill in matching either of the two reanalysis depends largely on the choice of the model used for running the hindcasts. Beyond 5 year lead time, hindcasts have reverted to the mechanism of variability dictated by the model selected; (mechanism depicted in PI simulation).
- The two reanalysis converges to different mechanism of variability. Their variability is controlled by a mix of the model behaviour, observations (sampling and variability) as well as assumption in the data assimilation method. Here, the two reanalyses differ in their choice of data assimilation leading to divergence in the mechanisms driving the density variability in the SPG region.
- Regarding the role of initialisation (as mean by the authors; i.e. FF or Anomaly), there does not seems to be sufficient evidence to demonstrate that skill is largely independent of the choice of initialisation. There are no clean comparisons (see [fig_ref] Table 1: Considering the information provided in [/fig_ref] to hold such statement and all FF and anomaly hindcasts are done with a different model, which has been shown to have a large influence. However, I agree that it seems to have comparatively less influence than the choice of the model beyond few years lead time.
It would be good to mentioned the name and version of the model used for ORAS4 (L319).
[fig] Fig 1: A lot of the analysis in the manuscript is based anomaly correlation of rho500. Correlation can be sometime misleading. Please show the time series of rho500 (e.g. inFig 1)for the hindcast and reference reanalysis. This has been added to Supplementary Figure 1.On Fig2: [/fig]
[fig] L125: This is likely because there are sufficient data to controlled the system during the ARGO period (see for example Zhang, S., A. Rosati, and T. Delworth. "The adequacy of observing systems in monitoring the Atlantic meridional overturning circulation and North Atlantic climate." Journal of Climate 23.19 (2010): 5311-5324.) Thanks. We have added this. L101-102 L140 Remove the extra coma Done.L 169 do not use contraction for is not. [/fig]
[fig] L372 In: Supplementary Figure 2, it would be most valuable if the number of observations available with time were added. If there are times with good observation covered (for example a hydrographic mission) in the Labrador Sea; the value could be reported on Fig 1. We have added this information to supplementary Figure 2 as requested. We thought Figure 1 was busy enough as it is, but instead we have added mention of the times of best observational coverage in the main text. L94 Reviewer #3 (Remarks to the Author): [/fig]
[fig] •: van Aken, Hendrik M., M. Femke De Jong, and Igor Yashayaev. "Decadal and multi-decadal variability of Labrador Sea Water in the north-western North Atlantic Ocean derived from tracer distributions: Heat budget, ventilation, and advection." Deep Sea Research Part I: Oceanographic Research Papers 58.5 (2011): 505-523. Yeager, S. G., and J. I. Robson. "Recent Progress in Understanding and Predicting Atlantic Decadal Climate Variability." Current Climate Change Reports 3.2 (2017): 112-127. Counillon, François, et al. "Flow-dependent assimilation of sea surface temperature in isopycnal coordinates with the Norwegian Climate Prediction Model." Tellus A: Dynamic Meteorology and Oceanography 68.1 (2016): 32437. [/fig]
[table] Table 1: Considering the information provided in [/table]
|
Transmission of Nonconjugative Virulence or Resistance Plasmids Mediated by a Self-Transferable IncN3 Plasmid from Carbapenem-Resistant Klebsiella pneumoniae
Klebsiella pneumoniae poses a critical challenge to clinical and public health. Along with conjugative plasmids, nonconjugative resistance or virulence plasmids associated with carbapenem-resistant K. pneumoniae (CRKP), hypervirulent K. pneumoniae (hvKP), and even carbapenem-resistant and hypervirulent K. pneumoniae (CR-hvKP) strains have been spreading globally. In this study, a clinical CRKP strain KP2648 was isolated, and the transferability of its plasmids was assessed using conjugation experiments. The transconjugants were characterized by polymerase chain reaction (PCR) detection, XbaI and S1-pulsed-field gel electrophoresis (PFGE), and/or whole-genome sequencing. Genetically modified IncN3 plasmids were employed to elucidate the self-transferability and the mobilization mechanisms. KP2648 has three natural plasmids: a nonconjugative IncFIB/IncHI3B virulence plasmid, a nonconjugative IncFII/IncR carbapenem-resistant plasmid, and a selftransferable IncN3 plasmid with a high conjugation frequency (7.54 6 1.06) Â 10 21 . The IncN3 plasmid could mobilize the coexisting nonconjugative virulence/resistance plasmids either directly or by employing intermediate E. coli with two forms: a hybrid plasmid fused with IncN3 or a cotransfer with the helper plasmid, IncN3. Various mobile genetic elements, including ISKpn74, ISKpn14, IS26, ISShes11, ISAba11, and Tn3, are involved in the genetic transposition of diverse hybrid plasmids and the cotransfer process during the intra/interspecies transmission. IMPORTANCE Nowadays, the underlying mobilization mechanism and evolutionary processes of nonconjugative virulence or resistance plasmids in Klebsiella pneumoniae remain poorly understood. Our study revealed the high conjugation ability of IncN3 plasmid isolated from carbapenem-resistant K. pneumoniae and confirmed its capability to mobilize the nonconjugative virulence or resistance plasmids. The self-transferable IncN3 plasmid could facilitate the transmission of pathogenicity and genetic evolution of carbapenem-resistant and hypervirulent K. pneumoniae, including hv-CRKP (virulence plasmid obtained by carbapenem-resistant K. pneumoniae) and CR-hvKP (resistance plasmid obtained by hypervirulent K. pneumoniae), warranting further monitoring.
throughout his hospitalization. Staphylococcus aureus, Enterococcus faecalis, carbapenem-susceptible K. pneumoniae, and E. coli were successively isolated from the skin wounds of the patient. CRKP strain KP2648 colonized the intestinal tract without causing infection manifestations. The patient showed typical symptoms of skin infections and was empirically treated with meropenem, vancomycin, fluconazole, and ganciclovir. His condition improved, and he was eventually discharged. KP2648 produced a negative string test and showed high resistance to a wide range of b-lactam antibiotics and carbapenem while remaining susceptible to colistin and tigecycline [fig_ref] TABLE 1: Phenotypic and genotypic characteristics of K [/fig_ref].
Polymerase chain reaction (PCR) and whole-genome sequencing (WGS) showed that KP2648 belongs to ST11 and serotype K64. It has three natural plasmids, pKP2648-Vir (216 Kb), pKP2648-KPC (102 Kb), and pKP2648-34 . The IncFIB/IncHI3B plasmid pKP2648-Vir encoded the known virulence factors, such as the aerobactin siderophore, the regulator of mucoid phenotype A (rmpA), and the putative metabolite transporter (peg344). These virulence factor genes are usually co-localized on a large virulence plasmid and serve as reliable identifiers for hypervirulence [bib_ref] Detection of multiple hypervirulent Klebsiella pneumoniae strains in a New York City..., Parrott [/bib_ref]. It also harbored the following ISs: ISkpn74, IS1 family, IS5075, IS630 family, IS903B, and others. The genomic sequence of pKP2648-Vir showed 91% coverage and 99.95% identity with that of the classical nonconjugative virulence plasmid pLVPK [bib_ref] Sequencing and analysis of the large virulence plasmid pLVPK of Klebsiella pneumoniae..., Chen [/bib_ref] , and there was no iro deletion or rmpA2 mutation [fig_ref] FIG 1: Sequence map of pKP2648-34 compared with other similar plasmids from GenBank [/fig_ref]. The IncFII/IncR plasmid pKP2648-KPC contained several AMR genes, such as bla KPC-2 , bla CTX-M-65 , and bla SHV-1 . Tn3 and ISs (ISkpn14, IS903B, IS5075, ISkpn27, IS26, and ISKpn28) were dispersed throughout the plasmid . Sequence alignment showed that pKP2648-KPC is a pKPC-LK30/ pHN7A8 plasmid that evolved from several recombination events of the nonconjugative bla KPC-2 -carrying plasmid pKPC-LK30 and the conjugative bla fosA3 -carrying plasmid pHN7A8. Notably, a part of the tra gene cluster of pHN7A8 (including TrbC, TraU, TraW, TraC, TraV, TraB, and TraK) was lost during the recombination [fig_ref] FIG 3: Transmission of the nonconjugative virulence or resistance plasmids to CRKP or hvKP,... [/fig_ref]. Thus, the pKPC-LK30 and pHN7A8 hybrid plasmid pKP2648-KPC was presumed to be nonconjugative, which was confirmed by a conjugation assay. Such nonconjugative pKPC-LK30/pHN7A8-like plasmids were also reported in other K. pneumoniae CG258 strains [bib_ref] Complete sequence of a novel IncR-F33:A-:B-plasmid, pKP1034, harboring fosA3, blaKPC-2, blaCTX-M-65, blaSHV-12,..., Xiang [/bib_ref] [bib_ref] Comparative analysis of bla (KPC-2)-and rmtB-carrying IncFII-family pKPC-LK30/pHN7A8 hybrid plasmids from Klebsiella..., Shi [/bib_ref]. The IncN3 plasmid pKP2648-34 did not contain any resistance/virulence genes; however, it did harbor a complete conjugative transfer-related module, including an oriT region, a relaxase of the MOBF family, a type IV coupling protein (T4CP) of the TraD family, and a trw gene cluster that coded for the IVB type T4SS. A BLASTN search showed that this IncN3 plasmid exhibited high sequence similarity (greater than 98% identity and coverage) to plasmid FDAARGOS_80 from Proteus mirabilis (accession: NZ_CP026060), Plasmid_C_Kpneumoniae_MS6671 from K. pneumoniae (LN824136), pR17.1451_p34k from Salmonella (CP063297), pR16.0676_34k from Salmonella (NZ_CP029801), and pN-Cit from Citrobacter freundii (JQ996149) [fig_ref] FIG 1: Sequence map of pKP2648-34 compared with other similar plasmids from GenBank [/fig_ref].
Transferability of the nonconjugative virulence or resistance plasmids from KP2648 to E. coli. Conjugation experiments were performed to evaluate the transferability of the nonconjugative virulence or resistance plasmids from KP2648. pKP2648- Vir was transferred from KP2648 to E. coli C600 . Two different patterns of transconjugants, EC2648-Vir1 and EC2648-Vir2, were observed with a conjugation efficiency of 10 25 during the selection of the virulence plasmid. The two virulence transconjugants were susceptible to carbapenem, and PCR results were rmpA-positive and bla KPC-2 -negative [fig_ref] TABLE 1: Phenotypic and genotypic characteristics of K [/fig_ref]. S1 nuclease pulsed-field gel electrophoresis (S1-PFGE) showed that EC2648-Vir1 contained a hybrid virulence plasmid (represented as pVir-fusion), which was speculated to have been derived from the fusion of pKP2648-Vir and pKP2648-34. EC2648-Vir2 contained two separate plasmids with sizes similar to those of pKP2648-Vir and KP2648-34, respectively. The virulence plasmids from both EC2648-Vir1 and EC2648-Vir2 were stable after serial batch culturing for 30 passages. In addition, both of these transconjugants could subsequently transfer their virulence plasmids to J53 in the same structures with higher frequencies of 10 22 and 10 23 , respectively. We further observed that the The resistance plasmid pKP2648-KPC, which does not carry the intact tra gene cluster, was transferred to C600 . Three different types of resistance transconjugants (EC2648-R1, EC2648-R1-1, and EC2648-R2) were selected with a conjugation efficiency of 10 26 [fig_ref] TABLE 1: Phenotypic and genotypic characteristics of K [/fig_ref]. EC2648-R1 and EC2648-R1-1 harbored similar hybrid resistance plasmids, which were speculated to be the fusion of pKP2648-KPC and pKP2648-34 (pKPC-fusion and pKPC-fusion-1). EC2648-R2 contained two separate plasmids with sizes similar to those of pKP2648-KPC and KP2648-34, respectively. Also, the resistance plasmids from EC2648-R1
FIG 2 Transfer of the nonconjugative virulence or resistance plasmids of KP2648 to E. coli. (a) XbaI and S1-PFGE of K. pneumoniae KP2648, E. coli C600, E. coli J53, and their corresponding transconjugants. EC2648-Vir1 and EC2648-Vir2 were detected by conjugation to select the virulence plasmid, whereas EC2648-R1 and EC2648-R2 were obtained by selecting resistance transconjugants. These transconjugant plasmids could be transferred to J53, represented as J-Vir1, J-Vir-2, J-R1, and J-R2. Differently colored triangles denote different plasmids; subsequent WGS confirmed that plasmids represented by the same triangles had similar backbones with or without additional ISs. (b) Schematic representation of the conjugation assays of KP2648 to E. coli. Rounded rectangles of the same color represent the same strains. Blue, green, and red circles denote pKP2648-Vir or its derivatives, pKP2648-KPC or its derivatives, and pKP2648-34, respectively. and EC2648-R2 were stable. Interestingly, pKPC-fusion from EC2648-R1 could be subsequently transferred to J53, whereas pKPC-fusion-1 could not be transferred at all. Similarly, the transfer of the resistance plasmid in EC2648-R2 was accompanied by pKP2648-34.
However, we did not observe the simultaneous dissemination of pKP2648-Vir and pKP2648-KPC in a single event using the selection with the meropenem, potassium tellurite (K 2 TeO 3 ), and rifampicin screening agar.
Transmission of the nonconjugative virulence or resistance plasmids to CRKP or hvKP, respectively. To test whether the virulence or resistance plasmids of the E. coli transconjugants obtained above could be transferred to K. pneumoniae strains, one CRKP strain (HS11286-pKPHS2DoriT) and two hvKP strains (RJF293 and RJF999) were employed as individual recipients [fig_ref] FIG 3: Transmission of the nonconjugative virulence or resistance plasmids to CRKP or hvKP,... [/fig_ref] and [fig_ref] TABLE 1: Phenotypic and genotypic characteristics of K [/fig_ref]. EC2648-Vir1 and EC2648-Vir2 To investigate whether these plasmids could be directly transferred from KP2648 to other CRKP or hvKP, we observed that the pKP2648-Vir was successfully transferred from KP2648 to HS11286-pKPHS2DoriT in a fusion form with a conjugation frequency of (7.24 6 0.82) Â 10 27 [fig_ref] TABLE 1: Phenotypic and genotypic characteristics of K [/fig_ref]. However, the transfer of pKP2648-KPC was not observed in the direct conjugation between KP2648 and RJF293H. To avoid the inhibitory effects of hyperviscosity, the rmpA-deficient RJF293ZH was also used as the recipient strain, but pKP2648-KPC was still not observed.
Genetic basis of hybrid plasmid generation. To explore the possible mobilization mechanism of the nonconjugative pKP2648-Vir and pKP2648-KPC plasmids, five virulenceand resistance-related transconjugants (EC2648-Vir1, EC2648-Vir2, EC2648-R1, EC2648-R1-1, and EC2648-R2) were sequenced. Genomic sequence alignments revealed that the hybrid plasmids represented various fusion products of the nonconjugative virulence/resistance plasmids with the conjugative IncN3 plasmid pKP2648-34. Notably, ISKpn74 elements located in pKP2648-Vir presumably attacked a 9 bp target site duplication (CAGCAAGAG) of pKP2648-34 through replicative transposition, resulting in the formation of a hybrid pVirfusion in EC2648-Vir1 [fig_ref] FIG 4: Genetic structures of the conjugative hybrid virulence plasmid pVir-fusion and the hybrid... [/fig_ref]. Similarly, ISKpn14 located in pKP2648-KPC might attack a 9 bp target site duplication (GATTGCGTA) of pKP2648-34, leading to the formation of pKPC-fusion [fig_ref] FIG 4: Genetic structures of the conjugative hybrid virulence plasmid pVir-fusion and the hybrid... [/fig_ref]. However, we did not observe the transfer of pKPC-fusion-1 in EC2648-R1-1. It might be due to the interruption of the trwJ gene that codes for the T4SS component by an IS26 insertion [fig_ref] FIG 4: Genetic structures of the conjugative hybrid virulence plasmid pVir-fusion and the hybrid... [/fig_ref].
In addition, we noticed that the Tn3 transposon was only present in pKP2648-KPC among the plasmids of KP2648, though it was added to pKP2648-VirR and to pKP2648-34R in the transconjugants EC2648-Vir2 and EC2648-R2. This observation suggests that the replicative transposition of Tn3 might contribute to the dissemination of these virulence and resistance plasmids [fig_ref] FIG 5: Mobilization of the nonconjugative pLVPK-like virulence plasmid and the nonconjugative resistance plasmid... [/fig_ref].
Mobilization of the nonconjugative virulence/resistance plasmids by the IncN3 plasmid. From the above-described experiments, we speculate that the IncN3 plasmid has a strong conjugation ability and plays a critical role in mobilizing nonconjugative virulence or resistance plasmids. To identify the transfer ability of the IncN3 plasmid, the hygromycin resistance gene was inserted into the IncN3 plasmid, named pKP2648-34H, and the modified KP2648 strain was named KP2648H, accordingly. Modified pKP2648-34H and KP2648H strains were constructed for further conjugation screening [fig_ref] FIG 5: Mobilization of the nonconjugative pLVPK-like virulence plasmid and the nonconjugative resistance plasmid... [/fig_ref]. Finally, pKP2648-34H could be transferred to C600 independently with a high conjugation frequency of (7.54 6 1.06) Â 10 21 . Furthermore, the self-transferable pKP2648-34H could also be transferred to E. coli J53, RJF293, KP3038, and HS11286-pKPHS2DoriT with a high conjugation efficiency of 10 21 [fig_ref] TABLE 1: Phenotypic and genotypic characteristics of K [/fig_ref].
In the conjugation between C600 and RJF293-34, various modes of pRJF293 transconjugants were observed : (i) pRJF293 fused with pKP2648-34H in the transconjugant TC293-34-1; (ii) pRJF293 cotransfer with pKP2648-34H in TC293-34-2; and (iii) TC293-34-3, a shortened form of pRJF293 transferred with pKP2648-34H. Similarly, pKP2648-34H also contributes to the mobilization of the nonconjugative KPC plasmids in HS11286-pKPHS2DoriT-34 and KP3038-34, likely via a mechanism of plasmid fusion . PCR and sequencing analysis showed that ISShes11 and ISAba11 participate in the fusion formation of TC293-34-1 and TC11286-pKPHS2DoriT-34. We then knocked out the traO gene of pKP2648-34H, which encodes the VirB9 protein that is responsible for the core complex of T4SS. The traO-deficient pKP2648-34HDtraO failed to transfer during the conjugation between KP2648HDtraO and C600. Moreover, neither the virulence plasmids nor the resistance plasmids were transferred in the conjugation. Altogether, these results showed that the self-transferable IncN3 plasmid could mobilize the nonconjugative virulence or KPC resistance plasmids either via cotransfer or via hybrid formation mediated by various MGEs. In addition, we designed the PCR primers specific to the conserved region of the IncN3 plasmid to investigate if it is commonly present in clinical K. pneumoniae. A total of 509 clinical K. pneumoniae isolates were screened, and 1.57% (8/509) of the isolates contained the IncN3 plasmid, including three CRKP strains, five CSKP strains, and one of CSKP strains is hvKP. Although the IncN3 plasmids were detected sporadically, it was not surprising that most strains (6/8) were observed among the critically ill patients in ICUs.
# Discussion
CRKP has been registered in the critical priority tier by the World Health Organization and has become a major menace of a public health concern (18). Moreover, CR-hvKP has recently become even more prevalent than previously assumed. Several strains are emerging with different carbapenemase-producing genes, including bla KPC , bla NDM , and bla OXA , posing a great challenge to infection control [bib_ref] Carbapenem resistance conferred by OXA-48 in K2-ST86 hypervirulent Klebsiella pneumoniae, Beyrouthy [/bib_ref]. The gut microbiome is the major reservoir and hot spot for exchanging virulence and carbapenemase-encoding genes during in-host evolution. Gastrointestinal colonization of CRKP or CR-hvKP, particularly in ICU wards, may lead to subsequent infections in at-risk patients, potentially playing a role in clone or plasmid outbreaks [bib_ref] Gastrointestinal carriage is a major reservoir of Klebsiella pneumoniae infection in intensive..., Gorrie [/bib_ref] [bib_ref] Pervasive transmission of a carbapenem resistance plasmid in the gut microbiota of..., Leon-Sampedro [/bib_ref]. Therefore, although collected from a stool sample, KP2648 should not be neglected, as it was isolated from an immunodeficient patient in an ICU. Thus, active culture surveillance and innovative strategies are urgently needed to curtail the further widespread transmission of CR-hvKP in nosocomial settings in China.
In addition to chromosomal mutations, the acquisition of pathogenic genetic element resistance and virulence phenotypes is mainly driven by plasmids (4). According to their mobility ability, plasmids can be assigned to three groups: conjugative, mobilizable, and nonmobilizable (4). Most notoriously, conjugative plasmids are especially relevant and indispensable for pathogenic gene transmission in different bacterial species [bib_ref] Conjugative plasmids interact with insertion sequences to shape the horizontal transfer of..., Che [/bib_ref]. However, it should be noted that nonconjugative resistance or virulence plasmids associated with CRKP or even CR-hvKP exhibit global dissemination. However, the underlying mobilization mechanisms and evolutionary processes of these nonconjugative plasmids remain poorly understood.
The recognition of conjugative components or conjugative helper plasmids is essential in predicting the likelihood of the dissemination of nonconjugative plasmids [bib_ref] Carbapenem resistance-encoding and virulence-encoding conjugative plasmids in Klebsiella pneumoniae, Yang [/bib_ref]. Previous studies on conjugative helper plasmids mainly focused on resistance plasmids. For example, the Incl1 conjugative helper plasmid has been shown to facilitate the transmission of the nonconjugative ciprofloxacin resistance plasmid in Salmonella [bib_ref] Transmission of ciprofloxacin resistance in Salmonella mediated by a novel type of..., Chen [/bib_ref]. To the best of our knowledge, our study is the first to demonstrate that the IncN3 plasmid plays a unique functional role as a conjugative helper plasmid and that it could transfer either the nonconjugative pLVPK-like virulence plasmid or the nonconjugative pKPC-LK30/pHN7A8-like resistance plasmid during the evolution of CR-hvKP. IncN plasmids have a broad host range. The IncN3 plasmid carries a complete conjugative transfer module. It is confirmed as self-transmissible with a high conjugation efficiency of (7.54 6 1.06) Â 10 21 and is stably inherited in the transconjugants [bib_ref] Plasmid evolution in carbapenemase-producing Enterobacteriaceae: a review, Kopotsa [/bib_ref]. A BLASTN search showed that the IncN3 plasmid is also present in other Enterobacteriaceae species, including Proteus mirabilis, Salmonella spp., Enterobacter hormaechei, Shigella dysenteriae, and Citrobacter freundii. The IncN3 plasmid was first reported and designated pN-Cit in C. freundii. In this report, pN-Cit was predicted to promote the mobilization of the nonconjugative pT-OXA-181 plasmid [bib_ref] Complete sequence of the IncT-type plasmid pT-OXA-181 carrying the blaOXA-181 carbapenemase gene..., Villa [/bib_ref]. It was also reported to probably facilitate the mobilization of the nonconjugative resistance plasmid pR16.0676_90k in Salmonella [bib_ref] New multidrug-resistant Salmonella enterica serovar Anatum clone, Chiou [/bib_ref]. In addition, a mosaic bla KPC-2 -harboring plasmid might have emerged via the incorporation of Tn4401 by an IncN1-type replicon and an IncN3-type tra system [bib_ref] Complete sequence of blaKPC-2-harboring plasmid with a mosaic of IncN1-and IncN3-type plasmids..., Kim [/bib_ref]. Collectively, these findings indicate that the IncN3 plasmid might have a powerful ability to transfer various nonconjugative plasmids or mobile elements across different species in Enterobacteriaceae. Furthermore, our study demonstrated that IncN3 plasmids might have greatly accelerated the formation of hv-CRKP and CR-hvKP by mobilizing nonconjugative plasmids either directly or indirectly. In addition, the IncN3 plasmid and the novel transconjugant plasmids might transfer freely in the gut microbiota of patients, thereby drastically broadening the transmission efficiency of pathogenicity genes and expanding the host spectrum. Thus, exploiting the conjugative delivery function of the IncN3 plasmid and interrupting plasmid-mediated transmission are warranted in further microbiome modification research [bib_ref] Exploiting CRISPR-Cas nucleases to produce sequence-specific antimicrobials, Bikard [/bib_ref].
Plasmids often contain multiple MGEs that enable them to undergo frequent genetic transposition, resulting in plasmid fusion or rearrangement, which leads to a better adaptation to the bacterial host (4). ISs are prevalent and dynamically distributed at a higher abundance on plasmids. ISs could capture and attack the hot spot in the conjugative helper plasmid to trigger homologous recombination and intermolecular transposition [bib_ref] Conjugative plasmids interact with insertion sequences to shape the horizontal transfer of..., Che [/bib_ref]. It is reported that ISs participate in 63.2% of the transferred AMR genes between plasmids and chromosomes [bib_ref] Conjugative plasmids interact with insertion sequences to shape the horizontal transfer of..., Che [/bib_ref]. For instance, IS26 and IS1216E (IS6 family) could mediate plasmid fusion and transmit resistance-encoding genes [bib_ref] Transmission of ciprofloxacin resistance in Salmonella mediated by a novel type of..., Chen [/bib_ref] [bib_ref] Plasmid fusion and recombination events that occurred during conjugation of poxtAcarrying plasmids..., Shan [/bib_ref]. Besides, replicative transpositions were reported in other presumed MGEs, such as Tn4401 (Tn3 family), IS903B (IS5 family), and ISKpn14 (IS1 family) [bib_ref] Emergence of a multidrug-resistant hypervirulent Klebsiella pneumoniae sequence type 23 strain with..., Shen [/bib_ref]. In our study, a diverse range of MGEs, including ISKpn74, ISKpn14, IS26, ISShes11, ISAba11, and Tn3, allowed for the identification of the hot spot of the IncN3 plasmid. These MGEs foster opportunities for plasmid fusion and transposition events across different genetic backgrounds. The continuous emergence and diversity of the IncN3 plasmid-mediated mobilization mechanism in Enterobacteriaceae should be further studied.
In summary, we have provided a report on the emergence of a self-transferable IncN3 plasmid with a broad host range that serves as a conjugative helper plasmid in K. pneumonia. It could directly mobilize the nontransmissible pLVPK-like or pKPC-LK30/ pHN7A8-like plasmids or employ intermediate E. coli in the evolutionary processes of CR-hvKP. Two transfer forms were conserved: a hybrid plasmid fusion with the IncN3 plasmid through a replicative transposition mechanism and a simultaneous but separate transfer with the IncN3 plasmid. We further confirmed the detailed mobilization mechanisms of the IncN3 plasmid in different types of K. pneumoniae and outlined the genetic evolution processes of hv-CRKP and CR-hvKP. The high conjugation ability to a new host and the stable inheritance of IncN3 would significantly speed up the transmission of pathogenicity in Enterobacteriaceae, warranting continuous surveillance.
# Materials and methods
Ethics approval. The K. pneumoniae strain KP2648 and other clinical isolates were collected from patients admitted to the Ruijin Hospital during routine clinical treatments. All personally identifiable information was removed before analysis. The study protocol was approved by the Ethics Committee of Ruijin Hospital, School of Medicine, Shanghai Jiao Tong University (RJ2019NO1-3). The requirement for informed consent was waived.
Bacterial strains and plasmids. The bacterial strains and plasmids used are listed in . KP2648 was collected from a stool swab and confirmed as K. pneumoniae by matrix-assisted laser desorption ionization-time of flight mass spectrometry (bioMérieux, Marcy-l' Etoile, France). For the different conjugation assays, rifampicin-resistant E. coli C600, sodium azide-resistant E. coli J53, K. pneumoniae strains, including RJF293, RJF999, HS11286, its derivative HS11286-pKPHS2DoriT (bla KPC-2 encoding plasmid pKPHS2 cannot be transferred due to the deletion of oriT), and clinical isolate KP3038, were employed as the recipient strains [bib_ref] Mobilization of the nonconjugative virulence plasmid from hypervirulent Klebsiella pneumoniae, Xu [/bib_ref]. hvKP strain RJF293 (NCBI accession number: PRJNA307277) (34) had a capsular serotype K2 and ST374, whereas RJF999 (PRJNA307276) was of serotype K1 and ST23. RJF293H is a derivative of RJF293 with a chromosomal insertion of the hygromycin resistance gene (hph), while RJF293ZH is a derivative of RJF293H with rmpA gene deletion. The ST11 CRKP strain HS11286 (PRJNA78789) was isolated from the human sputum [bib_ref] Mapping the resistance-associated mobilome of a carbapenem-resistant Klebsiella pneumoniae strain reveals insights..., Bi [/bib_ref]. The clinical CRKP strain KP3038 was isolated from an anal swab specimen that was positive for ST11 and bla KPC-2 ; however, the transferability of the bla KPC-2 gene of KP3038 cannot be detected by a conventional conjugation assay. In addition, 509 clinical K. pneumoniae isolates were collected as part of the routine clinical work with hospitalized patients from July to December 2020.
Multilocus sequence typing (MLST), antimicrobial susceptibility testing (AST), and string testing. Seven housekeeping genes for K. pneumoniae were amplified, sequenced, and analyzed according to the MLST database [bib_ref] MLST-based inference of genetic diversity and population structure of clinical Klebsiella pneumoniae, Guo [/bib_ref]. ASTs of the strains were initially performed using the VITEK2 compact system (bioMérieux, Marcy-l' Etoile, France) and were then determined by the broth microdilution method. Results were interpreted according to the Clinical and Laboratory Standards Institute guidelines (M100-ED30), except for tigecycline. In addition, the hypermucoviscosity phenotype was determined by a string test, conducted as described in our previous work [bib_ref] Whole-genome-sequencing characterization of bloodstream infection-causing hypervirulent Klebsiella pneumoniae of capsular serotype K2..., Wang [/bib_ref].
PCR. The oligonucleotide primers used are listed in . PCR was conducted to detect key carbapenem-resistance genes, virulence genes, and specific and important backbone genes in plasmids pKP2648-Vir, pKP2648-KPC, and pKP2648-34. PCR results were confirmed by direct Sanger sequencing.
Plasmid conjugation assay and plasmid stability. The transferability of pKP2648-Vir, pKP2648-KPC, pKP2648-34H, and its derivative pKP2648-34HDtraO was assessed by conjugation experiments [bib_ref] A conjugative plasmid that augments virulence in Klebsiella pneumoniae, Yang [/bib_ref]. We further tested the conjugative mobilization function of the self-transferable IncN3 plasmid. Donor and recipient isolates and antibiotics used for each pair of conjugation assays are listed in . Transconjugants were validated by AST, PCR, XbaI, S1 nuclease PFGE, and/or WGS [bib_ref] Mobilization of the nonconjugative virulence plasmid from hypervirulent Klebsiella pneumoniae, Xu [/bib_ref]. The conjugation frequency was calculated as the ratio of the number of transconjugants to recipient cells. Data were presented as mean 6 standard deviation (SD) based on three independent experiments. Furthermore, serial culturing of the purified transconjugants EC2648-Vir1, EC2648-Vir2, EC2648-R1, and EC2648-R2 was carried out for 30 generations in lysogeny broth (LB) by transferring 100 mL of bacterial culture to 10 mL of fresh LB every 12 h. Plasmid stability was assessed by randomly selecting single bacterial colonies for antibiotic resistance verification and PCR detection of key resistance/virulence genes or specific backbone genes of the IncN3 plasmid.
XbaI and S1 nuclease PFGE. XbaI and S1-PFGE were performed to confirm the genetic relatedness and the acquisition of different plasmids by the recipient strains. Chromosomal and plasmid DNA of the strains were prepared in agarose blocks and digested with XbaI and S1 nuclease, respectively. DNA fragments were separated by PFGE on the CHEF Mapper XA system (Bio-Rad, Hercules, CA, USA) [bib_ref] Mobilization of the nonconjugative virulence plasmid from hypervirulent Klebsiella pneumoniae, Xu [/bib_ref].
WGS, assembly, and annotation. The genomic DNA of KP2648 and five transconjugants (EC2648-Vir1, EC2648-Vir2, EC2648-R1, EC2648-R1-1, and EC2648-R2) were extracted and sequenced using a combination of the 150 bp paired-end Illumina NovaSeq 6000 platform and the PacBio RSII single-molecule long-read sequencing platform. Reads were reassembled using HGAP and the Canu 2.0 software package [bib_ref] Canu: scalable and accurate long-read assembly via adaptive k-mer weighting and repeat..., Koren [/bib_ref]. The genomic data were annotated using Prokka 1.1.3 [bib_ref] Prokka: rapid prokaryotic genome annotation, Seemann [/bib_ref]. Plasmid incompatible types were analyzed by PlasmidFinder 2.1 [bib_ref] In silico detection and typing of plasmids using PlasmidFinder and plasmid multilocus..., Carattoli [/bib_ref]. MGEs, virulence, and antibiotic resistance determinants were predicted by VRProfile [bib_ref] VRprofile: gene-clusterdetection-based profiling of virulence and antibiotic resistance traits encoded within genome..., Li [/bib_ref]. oriTfinder was used to determine the conjugative transfer-related modules of the plasmids, including the relaxase gene, the type IV coupling protein (T4CP) gene, and the tra gene cluster associated with the type IV secretion system (T4SS) [bib_ref] oriTfinder: a webbased tool for the identification of origin of transfers in..., Li [/bib_ref]. Plasmid sequences were aligned using the BLAST Ring Image Generator and Easyfig [bib_ref] BLAST Ring Image Generator (BRIG): simple prokaryote genome comparisons, Alikhan [/bib_ref] [bib_ref] Easyfig: a genome comparison visualizer, Sullivan [/bib_ref].
Genetically modified IncN3 plasmid. To verify the conjugation ability of the IncN3 plasmid, the hygromycin resistance gene hph was inserted into pKP2648-34. Thermo-sensitive helper plasmid pKOBEG carries the apramycin resistance gene and overexpresses the exo, bet, and gam genes of l phage [bib_ref] Mobilization of the nonconjugative virulence plasmid from hypervirulent Klebsiella pneumoniae, Xu [/bib_ref] [bib_ref] A rapid method for efficient gene replacement in the filamentous fungus Aspergillus..., Chaveroche [/bib_ref]. pKOBEG was electrically transferred into KP2648 competent cells, and positive cells were identified by apramycin. Positive cells were cultured in LB containing apramycin, and arabinose was added to induce the expression of the Exo/BET/GAM proteins for homologous recombination. Homologous recombinant fragments (upstream homologous arm-hph-downstream homologous arm) were prepared by gene splicing overlap extension-PCR and were electrically transferred into competent cells. The KP2648H strain with hph inserted was selected using a hygromycin-containing medium. The traO gene encodes the VirB9 protein, which is responsible for the core complex of the T4SS of the IncN3 plasmid [bib_ref] Structure of the outer membrane complex of a type IV secretion system, Chandran [/bib_ref]. The traO gene was knocked out from the KP2648H isolate using similar methods. The PCR primers used for the construction of the modified IncN3 plasmids are listed in .
Data availability. The genome sequences of K. pneumoniae KP2648, E. coli EC2648-Vir1, EC2648-Vir2, EC2648-R1, EC2648-R1-1, and EC2648-R2 have been deposited in the GenBank database under accession numbers CP072555-CP072558, CP083610-CP083611, CP083990-CP083992, CP084205-CP084206, CP084207-CP084208, and CP084209-CP084211, respectively.
# Supplemental material
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 1.4 MB.
[fig] FIG 1: Sequence map of pKP2648-34 compared with other similar plasmids from GenBank. FDAARGOS_80 from P. mirabilis (accession number: NZ_CP026060), Plasmid_C_Kpneumoniae_MS6671 from K. pneumoniae (accession number: LN824136), pR17.1451_p34k from Salmonella (accession number: CP063297), pR16.0676_34k from Salmonella (accession number: NZ_CP029801), and pN-Cit from C. freundii (accession number: JQ996149). of pKP2648-Vir in EC2648-Vir2 was accompanied by pKP2648-34 in conjugation experiments. [/fig]
[fig] FIG 3: Transmission of the nonconjugative virulence or resistance plasmids to CRKP or hvKP, respectively. (a) XbaI and S1-PFGE of EC2648-R1, EC2648-R2, EC2648-Vir1, EC26480Vir2, HS11286-pKPHS2DoriT, RJF293, RJF999, and their corresponding transconjugants. The meaning of different colored triangles is consistent with that inFig. 1. (b)Schematic of the process for the formation of hv-CRKP and CR-hvKP. The open yellow triangle represents the deletion of the oriT region. successfully transferred to HS11286-pKPHS2DoriT, and EC2648-R1 and EC2648-R2 could also be transferred to RJF293 and RJF999.Fig. 3boutlines the formation process of hv-CRKP and CR-hvKP by an intermediate E. coli. [/fig]
[fig] FIG 4: Genetic structures of the conjugative hybrid virulence plasmid pVir-fusion and the hybrid resistance plasmid pKPC-fusion. Alignment of pVir-fusion (a) and pKPC-fusion (b) with parental plasmids from KP2648. Linear comparison figures were created in Easyfig (http://easyfig.sourceforge.net/). Blue, green, and yellow arrows indicate tra genes, virulence genes, and ISs, respectively. (c) Proposed mechanisms of plasmid fusion in pVir-fusion: ISkpn74 on pKP2648-Vir attacked a 9 bp target site duplication (CAGCAAGAG) of pKP2648-34, leading to the formation of pVir-fusion through a replicative transposition mechanism. (d) Proposed mechanisms of plasmid fusion in pKPC-fusion: ISKpn14 located in plasmid pKP2648-KPC may attack a 9 bp target site duplication (GATTGCGTA) of pKP2648-34, leading to the formation of pKPC-fusion. [/fig]
[fig] FIG 5: Mobilization of the nonconjugative pLVPK-like virulence plasmid and the nonconjugative resistance plasmid with the help of pKP2648-34. (a) Mutant pKP2648-34H could transfer to E. coli C600 and then E. coli J53 independently; traO-deficient pKP2648-34HDtraO failed to transfer and detect in the conjugation between KP2648HDtraO and E. coli C600. (b) pKP2648-34H could transfer to RJF293, enabling the transmission of a nonconjugative virulence plasmid. (c) pKP2648-34H could transfer to KP3038 and HS11286-pKPHS2DoriT and helped the transmission of a nonconjugative resistance plasmid. The green square denotes the hygromycin resistance gene on the IncN3 plasmid. [/fig]
[table] TABLE 1: Phenotypic and genotypic characteristics of K. pneumoniae strain KP2648, E. coli C600, J53, and their corresponding transconjugants a [/table]
[bib_ref] Carbapenem resistance-encoding and virulence-encoding conjugative plasmids in Klebsiella pneumoniae, Yang [/bib_ref] |
The Expanding Cell Diversity of the Brain Vasculature
The cerebrovasculature is essential to brain health and is tasked with ensuring adequate delivery of oxygen and metabolic precursors to ensure normal neurologic function. This is coordinated through a dynamic, multi-directional cellular interplay between vascular, neuronal, and glial cells. Molecular exchanges across the blood-brain barrier or the close matching of regional blood flow with brain activation are not uniformly assigned to arteries, capillaries, and veins. Evidence has supported functional segmentation of the brain vasculature. This is achieved in part through morphologic or transcriptional heterogeneity of brain vascular cells-including endothelium, pericytes, and vascular smooth muscle. Advances with single cell genomic technologies have shown increasing cell complexity of the brain vasculature identifying previously unknown cell types and further subclassifying transcriptional diversity in cardinal vascular cell types. Cell-type specific molecular transitions or zonations have been identified. In this review, we summarize emerging evidence for the expanding vascular cell diversity in the brain and how this may provide a cellular basis for functional segmentation along the arterial-venous axis.
# Introduction
Continued expansion and sophistication of the mammalian brain has resulted in an astonishing level of cell diversity. It is believed that this cell diversity helps provide the sub-specialization of cellular function required to face the everchanging contexts of life-such as brain development, aging, and responses to injury or environmental insults. Despite this sophistication, the brain remains vitally dependent on the cerebrovasculature for metabolic exchange with circulating blood-including the delivery of oxygen, glucose, and other metabolites and clearance of toxic metabolic by-products. Disruption of the cerebral circulation for even minutes may have profound neurologic implications. To meet dynamic metabolic needs, the cerebrovasculature has developed a highly evolved blood-brain barrier (BBB) to permit regulated molecular transport while preventing influx of circulating blood cells or potentially toxic pathogens or plasma proteins. Coordinated communication between neurons, glial, and vascular cells facilitates local regulation of cerebral blood flow (CBF) to ensure blood supply is tightly matched with metabolic demand-a process known as "neurovascular coupling".
Over the past two decades, increased recognition that interconnection between multiple cell types is essential for brain vascular function and has resulted in advancement of the concept of the "neurovascular unit". Much of this work, however, has focused on cardinal cell types without acknowledgment of further sub-specialization or vascular cellular diversity. Development of single cell RNA sequencing (scRNAseq) technologies have now allowed unbiased characterization of transcriptional heterogeneity in brain vascular cells. While heterogeneity in glial and neurons have been reviewed elsewhere, we summarize the emerging evidence of added cell diversity within brain vascular cells highlighting how cellular sub-specialization contributes to regionalized vascular function along the arterialvenous axis.
## Cardinal cell types of the cerebrovasculature
The cerebrovasculature consists of multiple cell types-including endothelial cells (ECs), pericytes (PCs), and vascular smooth muscle cells (vSMCs); . Newly defined perivascular macrophages (PVMs) and perivascular fibroblastlike cells (PVFBs) also reside along the vascular wall. Astrocyte foot processes wrap around the vessel wall and create the Virchow Robin space-a perivascular space important for brain interstitial fluid and cerebrospinal fluid exchange known as the "glymphatic system". Parenchymal glial or neuronal cell bodies are closely apposed to the brain vasculature and rarely exceed > 15 µm from each vessel. Here, we summarize established characteristics of vascular cells and briefly highlight their brain-specific features and functions.
## Endothelial cells
Blood vessels are made up of a single layer of ECs that form a tubular structure. The blood facing surface, or luminal surface, is covered by the glycocalyx-a gel-like covering comprised of glycoproteins, proteoglycans, and glycosaminoglycans. Most brain ECs have a continuous cell membrane which lacks pores or fenestrations. Adjacent ECs are securely connected to one another by adherens junctional protein complexes and tight junction protein complexes occludes the intercellular cleft. Brain ECs have a high density of tight junction proteins which precludes significant paracellular flow and is a key feature regulating size-restricted transcellular passage of molecules. Unlike other organs, ECs also display lower rates of pinocytosis or non-specific bulk-flow vesicular transport. The brain-facing or abluminal EC membrane is embedded in a protein-rich basement membrane (BM) serving as a cellular scaffolding essential for interactions with neighboring cells.
Low rates of pinocytosis and membrane permeability make brain ECs the anatomic site of the BBB. The BBB permits gases, e.g., carbon dioxide and oxygen, and lipophilic and small molecules (< 400 Da) to freely enter the brain, but limits circulating cells and other molecules from crossing the endothelium without regulated transport systems. ECs contain a number of different transmembrane transporters to regulate influx of circulating nutrients and proteins-including carrier-mediated, receptormediated, and active transporter proteins. Active efflux transporters or other transport proteins also help remove potentially toxic waste products from brain into the blood. ECs sense shear stress from blood flow or receive signaling cues from neurons or glia to secrete vasoactive substances to help modulate vasomotor responses in adjacent mural cells and titrate local blood flow.
## Pericytes
In the microvasculature, e.g., capillaries, venules, and select arterioles, PCs are embedded in the shared BM and extend foot processes which cover much of the vessel wall. Unlike vSMCs which surround blood vessels circumferentially, PCs generally extend thin longitudinal processes with further delicate elaborations to cover more of the vascular surface. Interestingly, PCs cover the vasculature in a mutually exclusive pattern and do not overlap; upon experimental ablation of one pericyte, the neighboring pericytes will then grow their surface area to take over the newly absent territory. Pericytes contribute to endothelial BBB properties through upregulation of tight junctional protein complexes and/or downregulation of transcytotic pathways-such as those mediated by plasmalemma vesicle associated protein (Plvap) or caveolin-1. They also secrete BM proteins to stabilize vessels. To further regulate brain molecular exchanges, PCs are enriched in multiple carrier-and receptor-mediated transporters, active transporter proteins, or ion channels.
PCs modulate capillary diameter in response to neuronal or astrocytic cues, but the functional significance of this is unclear. Some have shown that PCs regulate local CBF regulation and neurovascular coupling. Others have found this is principally the role of vSMCs, such as those in arterioles, or pre-capillary sphincters. This discrepancy likely stems from a lack of clarity in regards to the exact definition FIGURE 1 | Brain vascular cell diversity along the arterial-venous axis. (A) Large arteries (red) and cortical veins (blue) course within the subarachnoid space (gray mesh) inside the dura mater (gray). Cerebral arteries course along the brain's pial surface and within sulci giving rise to pial and penetrating arterioles which pierce the brain parenchyma. (B) Arterioles progressively branch to form capillaries. Capillaries then converge to form venules which converge further to form veins. Each vessel segment along the arterial-venous axis are comprised of different cell types-including endothelium (shown in red, gray, and blue to denote transcriptional zonations), vascular smooth muscle cells (dark gray, vSMCs), pericytes (green), perivascular fibroblasts (yellow), and perivascular macrophages (light blue). Pericytes further adopt different morphologic configurations along the arteriovenous axis, such as ensheathing pericytes (pre-capillary arterioles) and mesh-like and stellate pericytes in more distal arterioles, capillaries, and venules. vSMCs form concentric rings in arteries but are discrete and stellate shaped in veins. Perivascular fibroblasts and macrophages are found in arteries, arterioles, venules, and veins, but not capillaries. (C-E) Cross section of cellular components of artery (C), capillary (D), and veins (E). Endothelium depicted in artery, capillary, and vein are shown in red, light gray, and blue, respectively, to denote transcriptional differences or molecular zonations defined and validated by independent groups with single cell RNA-sequencing. Dark gray, vascular smooth muscle cells; Navy blue, internal elastic lamina; Green, pericytes; Orange, perivascular fibroblasts; Light blue, perivascular macrophages; Yellow, basement membrane; Purple, astrocyte end feet. of vSMCs vs. PCs. PCs supply trophic support to neurons and also modulate immune responses. For example, PCs modulate expression of endothelial cell adhesion molecules, such as intercellular adhesion molecule-1 (Icam1), and may also play a direct role in the phagocytosis of extracellular proteins. This has been suggested to increase after brain injury, such as ischemic stroke, and pericytes may assume a microglial-like phenotype. Others have suggested pericytes may serve as multi-potent stem cells in vitro. However, lineage tracing has shown that pericytes do not significantly contribute to other cell lineages in vivo and do not differentiate into microglia following acute brain injuries.
## Vascular smooth muscle cells
vSMCs from concentric rings in larger arteries and become less layered and more sparse as vessels progressively branch to form pial and penetrating arterioles. In veins, vSMCs remain as discrete cells. Due to their location and composition, vSMCs contribute much of structural stability to the vessel wall and mediate synthesis and turnover of extracellular matrix proteins, such as collagen and elastin.
vSMCs serve as contractile cells and express a number of contractile proteins or associated regulatory proteins, such as smooth muscle alpha actin (ACTA2), and receptors to numerous vasoactive molecules, including adenosine, prostaglandins, or catecholamines, and to myogenic or flow-related stimuli. Through these contractile properties, vSMCs contribute to regulation of cerebral blood flow and autoregulation. Arterial pulsations and/or oscillations in relaxation or contraction of vSMCs contribute to perivascular fluid flow via the glymphatic system. Studies have supported that arteriolar vSMCs are the predominate site of regional CBF regulation, but the relative contributions of vSMCs and PCs to neurovascular coupling continues to be debated.
## Perivascular fibroblast-like cells
A recent constellation of unbiased scRNAseq studies have identified PVFBs on all vessels except capillaries. PVFBs are located within the Virchow-Robin space and loosely adhere to vessels through small, fine cellular processes. PVFBs express platelet-derived growth factor receptorα (Pdgfrα) and a number of extracellular matrix proteinsincluding fibrillar and non-fibrillar collagens, collagen-modifying enzymes, and small leucin-rich proteoglycans, such as lumican (lum) and decorin (Dcn). The function of these new-defined cells has not been completely characterized. Emerging evidence early in zebrafish development in other vascular beds, such as intersegmental vessels, has suggested that PVFBs stabilize the vasculature prior to appearance of PCs, and lineage tracing has suggested a subpopulation of PVFBs may serve as PC progenitor cells. In the mouse brain, however, collagen type 1 alpha 1 chain (Col1a1)-positive PVFBs are not detected until after the birth. Therefore, it is presently unclear as to the role if any PVFBs play in the developing or adult mammalian brain vasculature under physiologic conditions. In response to traumatic injury, PVFBs have been suggested to contribute to scar formation. Others have ascribed scar formation to a subpopulation of PCs which may highlight difficulties in distinguishing perivascular cell types.
## Perivascular macrophages
PVMs are myeloid cells distinct from microglia that reside in the Virchow-Robin space surrounding arterioles and venules. PVMs serve a vital role in the phagocytosis of blood-borne pathogens and antigen presentation to induce protective immune responses. They also limit EC production of pro-inflammatory mediators suggesting a dual regulatory role in immune response. PVMs promote EC BBB properties and tight junction protein expression. Others have shown that PVMS modulate EC expression of nutrient receptors-such as solute carrier family 2 member 1 (Slc2a)-playing a homeostatic role in cerebral glucose metabolism. In brain regions devoid of a BBB, such as the circumventricular organs, PVM uptake restricts influx of blood-borne proteins and macromolecules providing a size selective barrier. Phagocytic properties are not limited to blood-borne molecules and may help clear other extracellular proteins, such as amyloid-β.
## Astrocytes
Brain vascular cells are ensheathed by astrocyte end feet which form the outer limit of the perivascular Virchow-Robin space. Adjacent astrocyte end feet are connected via both tight and gap junctions which create a semi-permeable barrier-known as the glia limitans. This secondary barrier further restricts radial diffusion of large solutes and circulating inflammatory cells. Expression of transmembrane pores within astrocyte end feet, e.g., aquaporin-4, or ion pumps, e.g., Kir 4.1 potassium channels, regulate molecular exchanges between cerebrospinal fluid (CSF) within the perivascular space and brain interstitial fluid underlying the "glymphatic system". Interactions with adjacent PCs help maintain astrocyte end foot polarity and transport systems. Astrocytes also contribute directly to EC BBB properties and secrete BM proteins, such as laminin.
Astrocytes express metabotropic glutamate receptors (mGluR) and purinergic receptors (P2YR) which detect neuronal by-products of neuronal activation, such as glutamate or adenosine triphosphate, which help trigger section of vasoactive molecules-including arachidonic acid, associated derivatives, and prostaglandin E2. Low frequency stimuli or resting state fluctuations in neuronal activity have been shown to not provoke calcium signaling in astrocytes . Thus, it has been hypothesized that slow hemodynamic fluctuations in blood flow are driven largely by neurons and astrocytes play a modulating, but not triggering role in neurovascular coupling .
## Variations in cytoarchitecture and cell morphology
Large muscular arteries, such as the internal carotid arteries and vertebral arteries which coalesce to form the basilar artery, enter the subarachnoid space and form a complex anastomotic loop known as the "circle of Willis." Large cerebral arteries course along the brain's pial surface tapering and branching giving rise to pial and penetrating arterioles. Arterioles branch to give rise to an immense capillary tree. Capillaries ultimately converge to form venules which further converge to form veins. Large cortical veins within the subarachnoid space ultimately connect to venous sinuses contained within the dura mater to facilitate egress of blood from the brain. In the ensuing subsection, we describe the diversity in cell morphology of each cell type along these vascular transitions or regional variations when described in the central nervous system (summarized in .
## Endothelial cells
For most brain regions, the EC membrane remains continuous without interruption. Expression of the tight junction proteins remains constant throughout the arterial-venous axis in the brain, but show some ultrastructural differences-such as a greater overlap in arteries. Higher permeability and reduced tight junction protein expression is reported for select CNS regions-such as the spinal cord. In brain circumventricular organs-such as the subfornical organ, organum vasculosum of the lamina terminalis and area postrema-or choroid-plexus, ECs contain fenestrations or intracellular pores which permit high permeability.
Preservation of BBB integrity also occurs through suppression of caveolae and transcytosis. Classically low rates of vesicular transport were thought to be uniform. However, higher rates of vesicular transport have been observed in arteries/arterioles whereas receptor-mediated transcytosis is primarily detected in post-capillary venules. A recent work has shown that suppression is not uniform and that abundant caveolae are selectively abundant in arteriolar ECswhich facilitates the signaling between neurons, ECs, and vSMCs underlying neurovascular coupling. Whether this phenomenon varies regionally has not been reported.
## Mural cells
Considerable variation is observed in mural cells-a term referring to both vSMCs and PCs in part to account for varying cell identity along the arterial-venous axis; . In arteries with diameters > 100 µm, vSMCs are spindle shaped and form concentric rings ∼ 4-10 cells thick. As arteries progressively branch and taper forming arterioles, vSMC cell layers become thinner and ultimately become discontinuous cells with rod-like or thin lateral processes in more terminal branches; . In venules and veins, vSMCs remain discontinuous and are stellate in configuration; .
Whether a continuum of mural cell phenotypes exists and the point of transition between vSMC to pericyte remains controversial. Some have suggested that the transition from vSMCs to pericytes is discrete. Others have described a transitional cell with shared attributes between both cell types. Transitional cells have also been suggested in humans. However, classification of these or closely related cells differs between groups with a number of other terms-including ensheathing PCs, pre-capillary PCs, vSMC-PCs hybrids, or precapillary vSMCs. Despite these differences, most agree that the contractile cells in pre-capillary arterioles are responsible for regional CBF regulation and neurovascular coupling.
Brain PCs also display considerable morphologic heterogeneity-including ensheathing or transitional, mesh, thin-strand, or helical and stellate configurations; . Ensheathing PCsanalogous to the transitional cells above-predominately localize to pre-capillary arterioles. Like other PCs, they have an ovoid cell body, but with cell processes that are larger and broader which enwrap and cover much of the endothelial wall (> 90%). They also express contractile proteins-such as smooth muscle α-actin (αSMA). Contractile protein expression terminates as arterioles branch further to form smaller capillaries. Throughout capillaries and post-capillary venules, PC cell processes assume a mesh-like or thin-strand appearance which covers ∼70 or ∼50% of the endothelial cell well, respectively. The functional significance of these configurations is presently unknown but may reflect a facilitatory role in transport.
PC abundance and coverage of the EC wall is relatively constant in the adult brain-including the cortex, hippocampus, striatum, and cerebellum. However, PCs are less abundant in the spinal cord and most reduced in the gray matter of the anterior horn. Whether these alterations are the result of the relative abundance or presence of different pericyte subtypes is presently unknown. Much of this work has also been done in rodents and whether species-specific differences as with other cell types exists has also yet to be determined.
## Perivascular fibroblast-like cells
PVFBs share some cell marker expression with PCs-including N-aminopeptidase (also known as CD13, encoded by Anpep) and platelet-derived growth factor receptor-β (Pdgfrβ). Unlike PCs, they are found along all vessels expect capillaries . Morphologically they are more globular appearance with shorter and finer processes which enwrap the vessel wall. They are also more loosely associated with the vessels-described to be an "awkward hug". Whether different morphologic subtypes exist along the arterial-venous axis or in different brain regions remains to be characterized.
## Perivascular macrophages
PVMs are confined to arterioles and venules and are one of several specialized macrophages known as "border associated macrophages"-including leptomeningeal, dural, and choroid plexus macrophages; . PVMs adopt a relatively simple ameboid morphology. They are non-motile, but able to extend cellular processes along the Virchow-Robin Space. Other border associated macrophages are stellate or dendriform appearance. Whether PVMs has regional or arterial-venous variation in morphology has not been reported.
## Astrocytes
Morphologic variations in astrocytes have been described for > 100 years and were first introduced by Golgi and Cajal. Classically astrocytes were divided into two morphologic groups. Protoplasmic astrocytes of the gray matter are highly ramified contacting neuronal cell bodies, dendrites, and the vasculature, whereas fibrous astrocytes characterized as smaller with fewer branches are organized largely around white matter tracts. Between species, protoplasmic astrocytes are larger and extend 10-fold as many cellular processes in humans than rodents. Even greater morphological diversity is observed in different cortical layers. For example, layer 1 astrocytes known as subpial or marginal zone astrocytes display unique molecular expression patterns (Garcia-. Another subgroup of astrocytes unique to primatesknown as interlaminar astrocytes-are located in layer 1 and extend long fibers which extend throughout the cortex and terminate in layers 3/4. Extent of ramification or synaptic ensheathment of protoplasmic astrocytes varies between cortical layers. In layers 5-6, varicose projection astrocytes also extend long fibers with regularly spaced varicosities.
With respect to the brain vasculature, astrocyte cell bodies rarely exceed ∼6-10 µm from blood vessels and foot processes form a continuous sheathe around all vessels below the pia-including arterioles, capillaries, and venules. Within the somatosensory cortex, the perivascular astrocytic sheath is thickest for arterioles and thinnest for capillaries. However, capillaries had the greatest density of astrocyte foot processes and separation between astrocytes and capillaries decreased with increasing cortical depth. In other CNS regions, e.g., the retina, astrocytes along veins are large and occur with greater density than arteries. Whether this relationship is maintained in the brain or differs between CNS regions remains to be characterized. During development, the vasculature has been shown to interact with different subtypes of astrocytes or astrocyte-like radial-glial cells depending on location, such as truncated and outer radial glia. How different subtypes of astrocytes interact with different aspects of the arterial-venous axis in the adult brain has yet to be completely characterized.
## Transcriptomic diversification and subspecialization
Cell diversity in brain vascular cells transcends what can be seen with conventional histologic evaluations. Development of highthroughput single cell transcriptomic technologies identified considerable transcriptional variability of subpopulations within cardinal cell types. Distinct vascular segments have unique functional properties. For example, terminal or pre-capillary arterioles are believed to be the site of regional CBF regulation.
Capillaries are the site of bidirectional molecular transport. Venules and veins are the site of leukocyte migration and vascular immune responses. Dural based sinuses facilitate continued immune surveillance through meningeal lymphatics. In this subsection, we summarize newly defined transcriptional variation and functional distinctions within subpopulations of brain vascular cells (summarized in.
## Endothelial cells
EC heterogeneity is observed across organs. Brain ECs display a number of distinct attributes consistent with their specialized function. For example, brain ECs specifically express a number of transmembrane transporters-such as those involved in the transport of glucose (Slc2a1), amino acids (Slc3a2, Slc7a5), and fatty acids (Mfsd2a). Brain ECs also demonstrate a unique metabolic gene signature. Others have demonstrated enrichment of translated transcripts specific to neurotransmission, such as synapse organization and neurotransmitter transport. Enrichment for Wnt signaling and brain EC-specific transcription factors associated with maturation of the BBB are also observed, such as forkhead box F2 (Foxf2), forkhead box Q1 (Foxq1) or zic family member 3 (Zic3). Brain specific epigenetic regulatory networks, including chromatin accessibility and DNA methylome landscapes, have also been described in brain ECs suggesting added regulatory mechanisms.
Within the brain, EC gene expression changes along the arterial-venous axis. Capillary and venule ECs preferentially express solute transporters and inflammatory mediators, respectively. scRNAseq studies have demonstrated a transcriptional continuum or zonation characterized by gradual phenotypic changes along the arterialvenous axis. Closer analyses showed that transcription factors were overrepresented in arterial ECs, while transporters were enriched in ECs from capillary and veins. Three additional EC clusters were also identified but were outside the arteriovenous zone and the significance of these were unclear . scRNAseq technologies have evolved to permit characterization of a greater number of cells, such as droplet based scRNAseq. Studies employing these techniques in the adult murine brain vasculature have similarly showed continuous gene expression changes consistent with smooth EC molecular transitions along the arterial-venous axis. A recent study has identified nine distinct EC clusters. Seven correspond to arterial-venous graduationsuch as large arteries, arterial shear stress, capillary artery, capillary, capillary venous, and large vein. Two additional clusters were also noted-including choroid plexus ECs and interferon-activated ECs. Others have identified seven distinct subclusters of brain ECs. In addition to observing an arterial-venous continuum, the authors Venous Slc38a5, Nr2f2, Lcn2, Ccl19, Mafb, Mmm 1, VwfTip found evidence for further functional subspecialization within subclusters. For example, a subcluster of arterial ECs displayed selective expression of genes implicated in growth factor dependent remodeling, such as matrix gla protein (Mgp), fibulin 5 (Fbln5), elastin (Eln), insulin-like growth factor binding protein 4 (Igfbp4), and clusterin (Clu). Other processes, such as host immunity or interferon signaling, showed similar expression across vessel types. In the early postnatal brain vasculature, others have demonstrated six EC clusters or subtypes-including arterial, capillary-arterial, capillary-venous, venous, mitotic, and tip cells. Angiogenic cues specify some endothelial cells to become tip cells-which are motile and able to navigate through tissues. These are followed by less motile stalk ECs which maintain connections with the preexisting vasculature. scRNAseq has shown tip cells are enriched for transcripts encoding membrane or secreted proteins-including ion channels, cell adhesion proteins, receptor tyrosine kinases, extracellular matrix proteins, and other signaling molecules. Lineage tracing demonstrated that tip cells are more closely related to arterial than venous ECs. During angiogenesis, proliferating cells have recently been shown to arise from veins. Consistent with this, shared expression patterns are observed between mitotic, capillary venous, and venous ECs.
Regional variability has also been reported in the brain EC transcriptome. For example, brain EC expression of transporter genes was upregulated in arterial, capillary, and venous ECs, but not in choroid plexus. Consistent with their fenestrated morphology, there were also selective upregulation in high permeability genes-such as Plvap. Other studies have shown similar relative abundances of EC subtypes across 9 distinct brain regions-including the frontal cortex, posterior cortex, hippocampus, striatum, thalamus, globus pallidus externus, and nucleus basalis, subthalamic nucleus, substantia nigra, and ventral tegmental area, and cerebellum. Other targeted scRNA approaches geared toward the neuronalstem cell enriched subventricular zone show EC expression of certain stem cell markers-such as prominin 1 (Prom1) and nestin (Nes).
## Mural cells
RNAseq experiments have confirmed a number of transcriptomic differences between vSMCs and PCs. For example, arteriolar vSMCs show enrichment in gene products involved in pathways mediating cell contractility, vascular remodeling, and responses to hypoxia or oxidative stress, whereas mid-capillary PCs are enriched for immune regulatory processes and responses to viruses or toxic substances. Early scRNAseq studies distinguished vSMCs from PCs based on expression of the contractile protein smooth muscle α-actin (Acta2). In contrast to ECs, a more recent scRNAseq study has suggested a punctuated continuum for mural cells with an abrupt transition between arterial/arteriolar vSMCs and PCs. However, PCs appear to form a transcriptional continuum with venous vSMCs . This study also demonstrated differences in expression between arterial and venous vSMCs, such as levels of calponin-1 (Cnn1), Acta2, and smooth muscle protein 22-α (Tagln) . With different scRNAseq methodologies, others have shown even greater mural size diversity and identified seven distinct subpopulations. These approaches have identified specific markers for brain PCs in rodents-such as ATP binding cassette subfamily C member 9 (Abcc9), gammaglutamyltransferase 1 (Ggt1), potassium inwardly rectifying channel subfamily J member 8 (Kcnj8), vitronectin (vtn) and interferon-induced transmembrane protein 1 (Ifitm1), and have found differences in brain PCs in comparison to other organs-such as enrichment in SLC, ABC, and ATP transporters.
Many scRNAseq studies have not found transcriptionally distinct brain PC subtypes as supported by morphologic studies. A more recent scRNAseq mouse brain atlas has shown three distinct PC populations which vary in relative abundance across brain regions. Techniques-such as fluorescent activated cell sorting-have supported molecularly distinct PC subtypes based on transmembrane protein expression. Others have shown that another subpopulation of brain PCs express the gap junction protein connexin 30. A separate group has classified PCs into two subtypestype A and type B subtypes-based on expression of Pdgfrα, Pdgfrβ, Anpep, desmin (Des) and αSMA. Type A PCs have been shown to form scars after spinal cord and/or brain injury. However, some have suggested that this classification are not distinct PC populations, but rather more newly defined PVFBs. These studies support distinct subpopulations of brain pericytes. However, future studies are need to more completely characterize regulatory mechanisms contributing to pericyte diversity. Perivascular Fibroblast-Like Cells scRNAseq first identified 2 PVFBs subpopulations defined by expression of collagens [e.g., collagen type I alpha 1 chain (Col1a1) and collagen type I alpha 2 chain (Col1a2)], lumican (Lum), decorin (Dcn), and Pdgfrα . With newer droplet-based scRNAseq, up to 7 distinct subtypes have been identified. Two subclusters were enriched in membrane transporters and pumps, while others showed higher levels in collagen genes or different extracellular matrix proteins, angiogenesis, and contraction. This suggests possible functional subspecialization which has been shown to vary regionally. PVFBs enriched for membrane transport functions and collagen expression were observed with higher relative abundance in the hippocampus/cortex and basal ganglia/thalamus, respectively. Others have shown PVFBs are transcriptionally similar to populations of vascular leptomeningeal cells (pial and arachnoid cells) that reside in the meninges. Expression of the cytokine interleukin 33 (Il33) and the prostaglandin D2 synthetase (Ptgds) and markers of pial, arachnoid, and dural meningeal fibroblasts, may also help distinguish between PVFBs subpopulations located in the brain vs. fibroblasts in the meninges. Both Col1a1-positive and Rgs5-positive populations of mural cells representing PVFBs and PCs, respectively, can sense inflammatory stimuli and signal to the brain through C-C motif chemokine ligand 2 (Ccl2). Apart from this study, however, the functional significance of this heterogeneity has yet to be defined and future studies are needed.
## Perivascular macrophages
PVMs are transcriptionally closely related to microglial, but may be distinguished by expression of mannose receptor C-type 1 (Mrc, encodes CD206) and lymphatic vessel endothelial hyaluronan receptor 1 (Lyve1) or Cd36. Some have shown that expression of LYL1 basic helix-loop-helix family member (Lyl1) and Spi-C transcription factor (Spic) were specific to PVMs in the brain. Others have distinguished CNS-associated macrophages from peripheral immune cells through an absence of α4-integrin (Itga4) and Cd44.
Profiling of all CNS-associated macrophage populationsperivascular, meningeal, and choroid plexus, demonstrates three transcriptionally distinct clusters which share a core signature consisting of Mrc1, platelet factor 4 (Pf4), membranespanning four domains subfamily A member 7 (Ms4a7), stabilin 1 (Stab1), and carbonyl reductase 2 (Cbr2) . Other distinct PVM subclasses have also been defined in neuroinflammation-such as expression of antigenpresenting MHC class II molecules . Other brain inflammatory cells-such as microglia-are transcriptional similar across brain regions in adults but display added heterogeneity in different developmental periods . Whether regional or context-dependent heterogeneity exists specifically within PVMs has yet to be reported.
## Astrocytes
Astrocyte heterogeneity has been identified both morphologically and transcriptionally. For example, fibrous astrocytes of the white matter more highly express glial fibrillary acidic protein (Gfap) than protoplasmic astrocytes of cortical gray matter. Early scRNA seq experiments transcriptional defined two separate populations of cortical astrocytes distinguished by expression of glial fibrillary acidic protein (Gfap) and milk fat globule-EGF factor 8 protein (Mfge8). A more recent scRNAseq study in the cortex and hippocampus has identified 5 transcriptionally distinct astrocyte subtypes in adult rodents. Each with a unique spatial distribution, and with some exceptions roughly coincide with morphologically defined subtypes. Others have similarly supported the presence of five molecularly distinct astrocytes with different techniques-such as fluorescence-activated cell sorting.
With inclusion of additional brain regions, others have found seven distinct subtypes of astrocytes with scRNAseq-including Bergmann glia of the cerebellum, olfactory-specific astrocytes, two telencephalon specific astrocytes, two non-telencephalon astrocytes, and a myocilin (Myoc) expressing astrocyte of the dorsal midbrain. Expression of Mfge8 and angiotensinogen (Agt) sharply distinguished astrocytes from the telencephalon and diencephalon. Others have defined further regionally distinct astrocyte subtypes in other CNS regions-such as the spinal cord and retina. Additional astrocytic transcriptional heterogeneity has also been identified within cytoarchitecturally defined brain subregions. For example, laminar spatial gene expression is observed within astrocytes from distinct cortical layers which also varies between functionally distinct regions of cortex. With emerging sequencing technologies it is therefore likely additional astrocyte subtypes will be defined. How different transcriptional diversity of astrocytes influences vascular interactions has yet to be systematically studied.
## Discussion and future directions
Advances in scRNAseq technology have identified heterogeneity within cells comprising the brain vasculature beyond what can be seen with a microscope. The functional significance of transcriptomic variation has been implied based on enriched signaling or ontologic pathways. Future studies with subtype specific enhancer constructs and/or viruses are needed to directly understand the functional delineation of cell subtypes. Platforms to survey the function of vascular cells are also lacking. While organoids have rapidly increased in popularity to study neuronal brain cell types, these "mini-brains" have only been vascularized with ECs from other organs to date and further lack blood flow. Regulatory mechanism responsible for the genesis or maintenance of transcriptomic diversity in the context of brain vascular cells is also presently poorly defined. Only a single report has described epigenetic regulatory mechanisms in brain ECs. Studies have supported that signaling from neighboring glia and neurons contribute to brain specific transcriptomic networks. Alterations in blood flow-induced wall shear stress may lead to further transcriptional augmentation and in part help delineate cells along the arterial-venous axis. However, the relative contributions of flow-mediated biomechanical properties, circulating blood cells, and/or brain parenchyma to these observations is presently unknown.
Much of the variation of brain vasculature has been defined in rodents. Whether further diversity is seen in other species-such as humans-is also presently unknown. A number of studies have described alterations of vascular cells in neurovascular disorders-such as neurodegenerative diseases (Alzheimer's disease and amyotrophic lateral sclerosis), intracerebral hemorrhage, and vascular malformations. How vascular cell heterogeneity is altered in diseases affecting the cerebrovasculature has yet to be defined, and continued efforts are also needed to create in vitro human models which retain vascular cell diversity are needed to facilitate disease modeling.
# Author contributions
JR, CK, and EW designed the review outline, performed the literature search, and wrote the manuscript. DA, EC, KN, DC, AA, and TN provided the critical reviews, revised the manuscript, and provided relevant edits. All authors contributed to the article and approved the submitted version.
# Funding
The work of EW was supported by a Brain Vascular Malformation Consortium (BVMC) Pilot Feasibility Project Grant and Brain Aneurysm Foundation grant. The BVMC (U54NS065705) was a part of the NCATS Rare Diseases Clinical Research Network (RDCRN) and was supported by the RDCRN Data Management and Coordinating Center (DMCC) (U2CTR002818). RDCRN was an initiative of the Office of Rare Diseases Research (ORDR), NCATS, funded through a collaboration between NCATS and NINDS. |
TNF-α gene polymorphisms and expression
Tumor necrosis factor alpha (TNF-α) is a proinflammatory cytokine with an important role in the pathogenesis of several diseases. Its encoding gene is located in the short arm of chromosome 6 in the major histocompatibility complex class III region. Most of the TNF-α gene polymorphisms are located in its promoter region and they are thought to affect the susceptibility and/or severity of different human diseases. This review summarizes the data related to the association between TNF-α gene and its receptor polymorphisms, and the development of autoimmune diseases. Among these polymorphisms the −308G/A TNF-α promotor polymorphism has been associated several times with the the development of autoimmune diseases, however some discrepant results have been recorded. The other TNF-α gene polymorphisms had little or no association with autoimmune diseases. Current results about the molecules controlling TNF-α expression are also presented. The discrepancy between different records could be related partly to either the differences in the ethnic origin or number of the studied individuals, or the abundance and activation of other molecules that interact with the TNF-α promotor region or other elements. which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
# Background
Tumor necrosis factor (TNF), first termed by [bib_ref] Action of bacterial polysaccharide on tumors. II. Damage of sarcoma 37 by..., O'malley [/bib_ref] , was initially reported to induce programmed cell death or apoptosis. Currently, this molecule is thought to be involved in the regulation of many important cellular processes such as proliferation, differentiation, growth, and the immune response [bib_ref] A systems biology approach to suppress TNF-induced proinflammatory gene expressions, Hayashi [/bib_ref]. TNF-α is produced by various types of cells including macrophages, monocytes, neutrophils, T cells, and NK-cells. The gene encoding TNF-α is located in the class III region of the major histocompatibility complex on chromosome 6 between the HLA-B and HLA-DR genes [bib_ref] Association between TNF α gene polymorphisms and the risk of duodenal ulcer:..., Zhang [/bib_ref]. TNF binds to two types of outer membrane bound receptors on the target cells, TNFR1 and TNFR2, and triggers the cell survival and proinflammatory NF-κB and MAP kinases activations [bib_ref] The TNF and TNF receptor superfamilies: integrating mammalian biology, Locksley [/bib_ref]. The molecule activates phagocytes to engulf and clear infectious agents and cellular debris [bib_ref] Tumor necrosis factor alpha −308 gene locus promoter polymorphism: an analysis of..., Elahi [/bib_ref]. It also increases the expression of adhesion molecules on the vascular endothelium to allow immune cells, in particular neutrophils and macrophages, to translocate to the sites of tissue damage and infection [bib_ref] Tumour necrosis factor-alpha (TNFalpha): the good, the bad and potentially very effective, Barbara [/bib_ref].
The roles that TNF-α play seem to be contradictory and this was related to the genetic polymorphisms in the genes regulating its production and effect [bib_ref] Tumor necrosis factor alpha −308 gene locus promoter polymorphism: an analysis of..., Elahi [/bib_ref] , and the polymorphisms in TNF locus itself. It has been reported that the genetic alterations in the TNF-α locus are involved in high TNF-α production [bib_ref] A highly polymorphic CA repeat marker at the human tumor necrosis factor..., Tsukamoto [/bib_ref].
Several TNF-α polymorphisms have been identified inside the TNF-α promoter at the positions, relative to the transcription start site, −1031 [bib_ref] Tumor necrosis factor alpha −308 gene locus promoter polymorphism: an analysis of..., Elahi [/bib_ref].
[formula] (T/C), −863 (C/A), −857 (C/A), −851 (C/T), −419 (G/C), −376 (G/A), −308 (G/A), −238 (G/A), −162 (G/A), and −49 (G/A) [/formula]
In this article, we review the association between the genetic polymorphisms in TNF-α and the development of autoimmune diseases, and the relation between these polymorphisms and TNF-α expression.
## Open access
## Polymorphisms in tnf-α receptors
TNF-α interacts with the TNF receptors TNF-RI and TNF-RII [bib_ref] Association of polymorphisms of the tumour necrosis factor receptors I and II..., Bayley [/bib_ref]. TNF-α receptors (TNF-Rs) are active both in membrane-bound and soluble forms, and the soluble receptors act as physiological attenuators of TNF activity [bib_ref] The potential biological and clinical significance of the soluble tumor necrosis factor..., Aderka [/bib_ref]. With respect to their chromosomal location, TNF-RI gene is located at 12p13 and the TNF-RII gene is located at 1p36.2 [bib_ref] Association of polymorphisms of the tumour necrosis factor receptors I and II..., Bayley [/bib_ref].
With respect to the relation between TNF-Rs and autoimmune diseases, no association was reported between TNF-RI +36 and rheumatoid arthritis (RA) in Dutch and UK Caucasian population as approved by [bib_ref] Association of polymorphisms of the tumour necrosis factor receptors I and II..., Bayley [/bib_ref] and [bib_ref] Association between rheumatoid arthritis and polymorphism of tumor necrosis factor receptor II,..., Barton [/bib_ref] respectively. No association was found between the TNF-RII +1690 polymorphism and susceptibility to or severity of RA in the Dutch population [bib_ref] Association of polymorphisms of the tumour necrosis factor receptors I and II..., Bayley [/bib_ref].
However, TNF-RII 196M/R SNP was found to be associated with susceptibility to familial RA [bib_ref] Association between rheumatoid arthritis and polymorphism of tumor necrosis factor receptor II,..., Barton [/bib_ref] , but not associated with sporadic RA severity in Caucasian patients (Van der . [bib_ref] No association of polymorphisms in the tumor necrosis factor receptor I and..., Glossop [/bib_ref] and [bib_ref] Tumor necrosis factor receptor II gene polymorphism and severity of rheumatoid arthritis, Constantin [/bib_ref] have reported conflicting results about considering TNF-RII 196 M/R SNP as a genetic factor in RA severity. The 676TT genotype of TNF-α RII was related to a better response to anti-TNF-α drugs when compared to 676TG [bib_ref] Can tumor necrosis factor receptor II gene 676T > G polymorphism predict..., Ongaro [/bib_ref].
## Association between tnf-α genetic polymorphisms and autoimmune diseases
## Systemic lupus erthymatosus (sle)
In SLE patients, an increased level of TNF-α was reported and strongly correlated with the parameters of disease activity [bib_ref] Tumour necrosis factor alpha and its soluble receptors parallel clinical disease and..., Studnicka-Benke [/bib_ref]. A significant genetic association between TNF-α promoter −308A/G polymorphism and SLE susceptibility in Asian populations, and in European-derived populations was detected in [bib_ref] Metaanalysis of TNF-α promoter-308A/G polymorphism and SLE susceptibility in Asian populations, Zou [/bib_ref] and [bib_ref] Meta-analysis of TNF-alpha promoter −308 A/G polymorphism and SLE susceptibility, Lee [/bib_ref] metaanalysis respectively.
The −308A allele of TNF-α was associated with the production of TNF-α and susceptibility to SLE [bib_ref] TNF-308A and HLA-DR3 alleles contribute independently to susceptibility to systemic lupus erythematosus, Rood [/bib_ref] [bib_ref] A promoter polymorphism of tumor necrosis factor α associated with systemic lupus..., Sullivan [/bib_ref]. This allele contributed to susceptibility to SLE in South African patients [bib_ref] Study on the association between tumor necrosis factor alpha gene polymorphism and..., Wang [/bib_ref]. [bib_ref] TNF-308 G/A polymorphism and risk of systemic lupus erythematosus in the Polish..., Piotrowski [/bib_ref] indicated the presence of significant contribution of TNF-α −308 A allele to arthritis and renal SLE manifestation, and that the TNF-α −308 G/A polymorphism may be a HLA-DRB1*03:01 haplotype-dependent genetic risk factor for SLE in a cohort of Polish population.
The +489A allele of TNF-α was also thought to have a genetic contribution to the susceptibility to SLE in the Chinese population [bib_ref] Association of TNF-α gene polymorphisms with systemic lupus erythematosus in Taiwanese patients, Lin [/bib_ref].
## Rheumatoid arthritis (ra)
TNF-α is thought to play a central role in inflammation and it has been directly implicated in the pathogenesis of RA [bib_ref] Cytokine production in the rheumatoid joint: implications for treatment, Feldmann [/bib_ref]. High concentrations of TNF-α were detected in serum and synovial fluid of RA patients and TNF-α blood concentration correlated with RA disease activity [bib_ref] Polymorphism in the tumor necrosis factor-α gene promoter is associated with severity..., Nemec [/bib_ref].
Studies on the association between −238G/A and −308G/A TNF polymorphisms and Juvenile Idiopathic Arthritis (JIA) showed conflicting results. For example, the −238 G/A polymorphism did not have an effect on the patients' outcome in either Turkish or Czech patients while the −308 G/A polymorphism was significantly associated with a poor outcome in the Turkish group but not in the Czech patients [bib_ref] Tumour necrosis factor α G → A − 238 and G →..., Ozen [/bib_ref] or with systemic JIA in Japanese population [bib_ref] Identification of a genetic risk factor for systemic juvenile rheumatoid arthritis in..., Date [/bib_ref]. However, TNF-α −308A allele was significantly associated with JIA (JimÊnez-Morales et al. 2009), with systemic JIA [bib_ref] TNF-α promoter gene polymorphisms in Spanish children with persistent oligoarticular and systemic-onset..., Modesto [/bib_ref]. [bib_ref] Identification of a genetic risk factor for systemic juvenile rheumatoid arthritis in..., Date [/bib_ref] demonstrated that the −863A, −1013C, and −857T alleles were significantly higher in systemic JIA Japanese patients.
TNF-α was also considered as a major candidate gene in psoriatic arthritis (PsA) [bib_ref] Tumor necrosis factor-α in psoriasis and psoriatic arthritis: a clinical, genetic, and..., Kane [/bib_ref] as TNF-α was present in high levels in serum, synovial fluid, and synovial membrane in patients with PsA [bib_ref] Macrophage-derived cytokine and nuclear factor κB p65 expression in synovial membrane and..., Danning [/bib_ref] [bib_ref] Highly increased levels of tumor necrosis factor-alpha and other proinflammatory cytokines in..., Partsch [/bib_ref]. However, studies on the association between TNF-α polymorphisms and PsA showed conflicting results. A strong association between −308G/A and −238G/A promoter polymorphisms and susceptibility to PsA was demonstrated in some studies [bib_ref] Association of TNF −238 and −308 promoter polymorphisms with psoriasis vulgaris and..., Mössner [/bib_ref] [bib_ref] TNFα polymorphisms and risk of psoriatic arthritis, Rahman [/bib_ref]. In contrast, [bib_ref] TNF-α gene polymorphisms: association with disease susceptibility and response to anti-TNF-α treatment..., Murdaca [/bib_ref] did not find significant association between −238 and −308G/A SNPs and PsA in Italian patients but they found a significant association between +489A allele and both PsA susceptibility and severity. [bib_ref] TNFα polymorphisms and risk of psoriatic arthritis, Rahman [/bib_ref] could not detect any significant association between −1031T/C, −863C/A and −857C/T SNPs and PsA.
Also, studies on the relation between TNF-α polymorphisms and RA showed conflicting results. For example, susceptibility to RA was associated with the −308A allele in some studies [bib_ref] Tumor necrosis factor-α is a common genetic risk factor for asthma, juvenile..., Jimênez-Morales [/bib_ref] [bib_ref] Tumor necrosis factor-alpha promoter −308 A/G polymorphism and rheumatoid arthritis susceptibility: a..., Lee [/bib_ref] , with the G allele in others [bib_ref] Association of tumour necrosis factor-alpha −308 G/A promoter polymorphism with susceptibility and..., Mosaad [/bib_ref] , but neither with A allele nor G allele in others [bib_ref] Tumor necrosis factor-alpha and interleukin-10 gene promoter polymorphisms in Turkish rheumatoid arthritis..., Ates [/bib_ref] [bib_ref] Association of tumor necrosis factor alpha and IL-10 promoter polymorphisms with rheumatoid..., Gambhir [/bib_ref] [bib_ref] Polymorphism in the tumor necrosis factor-α gene promoter is associated with severity..., Nemec [/bib_ref] [bib_ref] Tumour necrosis factor a −308 promoter polymorphism in patients with rheumatoid arthritis, Rezaieyazdi [/bib_ref].
Heterozygous genotype GA of −308A/G SNP was found to be associated with more severe course of RA disease [bib_ref] Susceptibility for and clinical manifestations of rheumatoid arthritis are associated with polymorphisms..., Cvetkovic [/bib_ref] , with increased number of erosions, and with the progression of radiographic damage in patients with early seropositive RA . RA severity was associated with the presence of −308A allele [bib_ref] Association of tumour necrosis factor-alpha −308 G/A promoter polymorphism with susceptibility and..., Mosaad [/bib_ref] [bib_ref] Tumor necrosis factor-alpha −308 promoter polymorphism contributes independently to HLA alleles in..., Rodríguez-Carreón [/bib_ref] and with −308G allele [bib_ref] Polymorphism in the tumor necrosis factor-α gene promoter is associated with severity..., Nemec [/bib_ref]. The −308G allele showed a trend toward worse radiological outcome by 5 years in patients with inflammatory arthritis as indicated by the presence/absence of erosions [bib_ref] Polymorphisms in the tumour necrosis factor gene are not associated with severity..., Barton [/bib_ref]. In Han Chinese population, susceptibility of RA increased in patients with TNF-α −308G allele, especially in the females, and the patients containing both HLA-DRB1*04 and TNF-α-308 GG genotype showed a significant increase in risk for RA regardless of their sex [bib_ref] Association of TNF-α gene polymorphisms with the risk of rheumatoid arthritis in..., Li [/bib_ref]. On the other hand, in a cohort of Argentinean patients with RA, the −308A allele was neither associated with suscetibility to RA nor with the course and outcome of the disease [bib_ref] The −308 G/A polymorphism in the tumor necrosis factor-α gene is not..., Aranda [/bib_ref]. In addition, there was no significant association between −308 G/A polymorphism and RA risk in a cohort of Bulgarian population [bib_ref] Association of single nucleotide polymorphism at position −308 of the tumor necrosis..., Manolova [/bib_ref].
−238G/A polymorphism was associated with more severe course of RA [bib_ref] Tumor necrosis factor-alpha gene polymorphism in severe and mild-moderate rheumatoid arthritis, Fabris [/bib_ref] , however, [bib_ref] Polymorphisms in the tumour necrosis factor gene are not associated with severity..., Barton [/bib_ref] showed that both −238G/A and −376G/A were not associated with RA severity. No association was reported between −238G/A and −376G/A polymorphisms and susceptibility to RA in Egyptian population [bib_ref] TNF-α genetic polymorphisms and its expression in Egyptian rheumatoid arthritis patients, Mousa [/bib_ref]. [bib_ref] Polymorphisms in the tumour necrosis factor gene are not associated with severity..., Barton [/bib_ref] failed to find any difference in either allele or genotype frequences of −1031, −863, −857, +489, +851 and +1304 SNPs between patients with inflammatory arthritis developing erosions and those remaining non erosive at 5 years. [bib_ref] Contribution for new genetic markers of rheumatoid arthritis activity and severity: sequencing..., Fonseca [/bib_ref] demonstrated an association between −238, −308, −857, and −863 SNPs and systemic manifestations, functional status, radiological damage, work disability, and joint surgeries.
## Ankylosing spondylitis (as)
TNF-α level in blood and its expression by peripheral T cells correlated well with AS activity [bib_ref] Comparison of serum IL-1β, sIL-2R, IL-6, and TNF-α levels with disease activity..., Bal [/bib_ref] [bib_ref] Low T cell production of TNFα and IFNγ in ankylosing spondylitis: its..., Rudwaleit [/bib_ref]. Studies on the relation between TNF-α SNPs and AS have shown controversial results. For example, TNF-α polymorphisms had no independent effect on AS susceptibility [bib_ref] Polymorphisms of tumor necrosis factor-α promoter region for susceptibility to HLA-B27-positive ankylosing..., Chung [/bib_ref] [bib_ref] The association between TNF-α promoter polymorphisms and ankylosing spondylitis: a meta-analysis, Li [/bib_ref] but their modulating effect on TNF-α expression were well relevant to the phenotypic diversity in AS [bib_ref] Lack of association of TNF-α promoter polymorphisms with ankylosing spondylitis: a meta-analysis, Lee [/bib_ref] [bib_ref] Relation of HLA-B27, tumor necrosis factor-α promoter gene polymorphisms, and T cell..., Poddubnyy [/bib_ref]. In contrast, Vargas-Alarcón et al. and [bib_ref] Association of tumour necrosis factor α promoter polymorphisms with ankylosing spondylitis in..., Shiau [/bib_ref] showed an association of −308G/A polymorphism with susceptibility to AS. Moreover, the A allele was thought to have a protective role against AS [bib_ref] Polymorphisms of tumor necrosis factor-α promoter region for susceptibility to HLA-B27-positive ankylosing..., Chung [/bib_ref] [bib_ref] Tumor Necrosis factor-α promoter −308/238 polymorphism association with less severe disease in..., Nossent [/bib_ref] , and was associated with a lower risk of developing AS, and with the age at disease onset, disease severity and response to anti-TNF treatment [bib_ref] Association of single nucleotide polymorphism at position −308 of the tumor necrosis..., Manolova [/bib_ref].
In conclusion, the data from the studies on TNF-α genetic polymorphisms seem to vary from one study to another. This variation could be related to the differences in the ethnic orgin or the number of the indivisulas under study. Although these polymorphisms do affect the expression level of TNF-α, the activation and abundance of other molecules that interact directly or indirectly with the promotor sequence must affect the expression of TNF-α.
## Control of tnf-α gene expression
Several studies tried to address the relation between TNF-α polymorphisms and its expression, and the mechanisms controlling its expressions in many cell types and diseases. High TNF-α expression level was associated with the −238G allele in multiple sclerosis patients [bib_ref] TNF-α promoter polymorphisms, production and susceptibility to multiple sclerosis in different groups..., Huizinga [/bib_ref] and with −863A and −1031C alleles in healthy Japanese and non-Japaneses individuals [bib_ref] Polymorphism of the 5′-flanking region of the human tumor necrosis factor (TNF)-α..., Higuchi [/bib_ref]. While the −238A allele was reported to down regulate TNF-α expression [bib_ref] Different transcriptional activity and in vitro TNF-α production in psoriasis patients carrying..., Kaluza [/bib_ref]. In appearently healthy individuals, [bib_ref] Inflammatory bowel disease is associated with a TNF polymorphism that affects an..., Van Heel [/bib_ref] demonstrated that the −857T (but not the −857C) allele inhibits TNF-α transcription through its strong binding with the transcription factor OCT1, which blocks the interaction of nuclear factorkappa-B (NF-κB) to the nearby region −873 to −863 . [bib_ref] TNF-α genetic polymorphisms and its expression in Egyptian rheumatoid arthritis patients, Mousa [/bib_ref] detected a significant increase of TNF-α expression in RA patients compared to healthy individuals, but this increase in expression was not linked to a certain allele of the −238G/A and −376G/A SNPs. Besides, several polymorphisms in some TNF-linked genes were also thought to regulate the expression of TNF directly or indirectly [bib_ref] Impact of the-308 TNF promoter polymorphism on the transcriptional regulation of the..., Abraham [/bib_ref].
According to [bib_ref] Targeting the PIAS1 SUMO ligase pathway to control inflammation, Liu [/bib_ref] , TNF-α can be autoregulated by activating PIAS1 [a member of the protein inhibitor of activated STAT1 (signal transducer and activator of transcription 1) family] SUMO E3 ligase. It is though that once activated, PIAS1 is then recruited to the TNF-α gene promoter to repress transcription [bib_ref] Targeting the PIAS1 SUMO ligase pathway to control inflammation, Liu [/bib_ref].
The neurohormonal molecule, angiotensin II (Ang II), was thought to play a role in up regulation of TNF-α gene expression as Ang II activates two transcriptional factors that are important in mediating TNF-α gene expression, nuclear factor (NF)k B and activator protein-1 (AP-1) [bib_ref] Upregulation of vascular endothelial growth factor by angiotensin II in rat heart..., Chua [/bib_ref] [bib_ref] Angiotensin-converting enzyme inhibition prevents arterial nuclear factor-κB activation, monocyte chemoattractant protein-1 expression,..., Hernandez-Presa [/bib_ref] [bib_ref] Angiotensin II participates in mononuclear cell recruitment in experimental immune complex nephritis..., Ruiz-Ortega [/bib_ref]. Angiotensin II also provokes TNF biosynthesis in various nonmyocyte cell types [bib_ref] Angiotensin II induces TNF production by the thick ascending limb: functional implications, Ferreri [/bib_ref] [bib_ref] Angiotensin II and gene expression in the kidney, Klahr [/bib_ref].
Oxidized low density lipoprotein (LDL) and LDL were demonstrated to decrease TNF mRNA expression in NK cells [bib_ref] Oxidized low-density lipoproteins affect natural killer cell activity by impairing cytoskeleton function..., Malorni [/bib_ref]. Mitogen-activated Protein Kinase (MAPK) has been found to play a central role in the induced TNF-α expression in monocytes, macrophages, mast cells and T cells [bib_ref] Different mitogen-activated protein kinase signaling pathways cooperate to regulate tumor necrosis factor..., Hoffmeyer [/bib_ref] [bib_ref] Mast cell tumor necrosis factor α production is regulated by MEK kinases, Ishizuka [/bib_ref] [bib_ref] N-terminal kinase/stressactivated protein kinase (JNK/SAPK) is required for lipopolysaccharide stimulation of tumor..., Swantek [/bib_ref] [bib_ref] Mitogen-activated Protein (MAP) kinase regulates production of tumor necrosis factor-α and release..., Zhang [/bib_ref] [bib_ref] Regulation of TNF expression by multiple mitogen-activated protein kinase pathways, Zhu [/bib_ref].
The 3′ UTR of TNF-α contains a sequence element that is though to affect posttranslational control of TNF through mRNA stability and translation efficiency [bib_ref] Tumor necrosis factor alpha −308 gene locus promoter polymorphism: an analysis of..., Elahi [/bib_ref]. TNF mRNA has two protein-binding regions [bib_ref] Characterization of the RNA binding proteins forming complexes with a novel putative..., Hel [/bib_ref] [bib_ref] Two distinct regions in the 3′ untranslated region of tumor necrosis factor..., Hel [/bib_ref] located in the AU-rich element (ARE) within 3′ UTR [bib_ref] Fragile X-related protein FXR1P regulates proinflammatory cytokine tumor necrosis factor expression at..., Garnon [/bib_ref]. Both AREs were reported to interact with several proteins including TIAR and AUF1 (DeMaria and Brewer 1996; [bib_ref] Identification of TIAR as a protein binding to the translational regulatory AU-rich..., Gueydan [/bib_ref] which control TNF mRNA post transcription. Pituitary adenylate cyclase-activating polypeptide is another protein that was found to inhibit TNF-α expression [bib_ref] Inhibition of constitutive TNF production is associated with PACAP-mediated differentiation in PC12..., Manecka [/bib_ref].
cAMP has been reported to play an important role in regulating TNF-α expression, for example the elevation of cellular cAMP suppresses TNF-α production [bib_ref] Chronic ethanol-mediated decrease in cAMP primes macrophages to enhanced LPS-inducible NF-κB activity..., Gobejishvili [/bib_ref] [bib_ref] Adenosine-cyclic AMP pathways and cytokine expression, Zidek [/bib_ref].
Interferon regulatory factor-5 (IRF5) has been also reported to play a key role in the induction of TNF-α [bib_ref] Multiple regulatory domains of IRF-5 control activation, cellular localization, and induction of..., Barnes [/bib_ref]. [bib_ref] IRF5 is required for late-phase TNF secretion by human dendritic cells, Krausgruber [/bib_ref] has reported that TNF secretion in human monocyte-derived dendritic cells is mediated by cooperative action of IRF5 and RelA at the 5′ upstream and 3′ downstream regions of the TNF gene.
In conclusion, the expression of TNF-α seems to be controlled by the activation of other cellular molecules including signal transducer molecules, nuclear factors and second messenger molecules. More studies to address the interaction between these molecules and TNF-α are required to finally map a TNF-α pathway. Profiling the expression of the key molecules inside such pathway can open an avenue to control TNF-α over or downregulation.
# Conclusion
TNF-α plays an important role in the pathogenesis of autoimmune diseases. Several studies show that TNF-α gene promoter polymorphisms affect the susceptibility and/or severity of autoimmune diseases. Polymorphisms in the TNF receptors does not seem to be associated with the development of autoimmune diseases. The −308 G/A TNF-α promotor polymorphism seems to be highly associated with the development of these diseases, however some discrepant results have been recorded. Other TNF-α gene polymorphisms had little or no association Schematic representation showing some of the molecules thought to be involved in the interaction with TNF-α promoter region. The transcriptional induction of TNF-α is thought to be controlled by some transcription factors, including the transcription factor OCT1, the nuclear factor κB (NF-κB), the signal transducer and transcription activator (PIAS1) and activator protein-1 (AP-1). The transcription factor OCT1 can strongly bind with the allele -857T (but not the -857C) and thus blocks the interaction of NF-κB to the nearby region -873 to -863 leading to inhibition of TNF-α transcription. PIAS1 possesses SUMO E3 ligase activity and can repress NF-κB by blocking the DNA-binding activity of NF-κB to TNF-α promoter. Ang II can activate the 2 transcription factors NF-kB and AP-1that are important in mediating TNF-α gene expression. Alleles associated with upregulation of TNF-α gene are designated with arrows with heads directed up while those alleles associated with downregulation are designated with arrows with heads directed down. Molecules involved in the posttranscriptional and posttranslational control of TNF-α are mentioned in the text and not shown here with autoimmune diseases. This discrepancy might be explained by the differences in the ethnic orgin or number of the studied individuals. TNF-α gene expression is controlled by the presence of some polymorphisms in its promoter region and by several types of signalling molecules or nuclear factors that interact with the TNF-α promotor region or other elements.
## Abbreviations
Ang II: angiotensin II; ARE: AU-rich elements; AS: ankylosing spondylitis; AUF1: AU-rich element RNA-binding protein 1; HLA-B: human leukocyte antigen-B; HLA-DR: human leukocyte antigen-DR; JIA: juvenile idiopathic arthritis; LDL: low density lipoprotein; MAP kinases: mitogen-activated protein kinases; NF-κB: nuclear factor kappa-light-chain-enhancer of activated B cells; NK-cells: natural killer cells; OCT1: organic cation transporter 1; PIAS 1: protein inhibitor of activated STAT-1; STAT-1: signal transducer and activator of transcription-1; PSA: psoriatic arthritis; RA: rheumatoid arthritis; SLE: systemic lupus erthymatosus; SNPs: single nucleotide polymorphisms; SUMO E3 ligase: small ubiquitin-like modifier E3 ligase; TNFR-I: tumor necrosis factor receptor-1; TNF-α: tumor necrosis factor alpha; TIAR: T-cell-restricted intracellular antigen 1-related protein.
Authors' contributions RE searched the literature and wrote the initial draft of the manuscript. AG structured, revised and finalized the manuscript. NM took part in revising the manuscript. All authors read and approved the final manuscript |
Molecular Characterization of the First Alternavirus Identified in Fusarium oxysporum
Citation: Wen, C.; Wan, X.; Zhang, Y.; Du, H.; Wei, C.; Zhong, R.; Zhang, H.; Shi, Y.; Xie, J.; Fu, Y.; et al. Molecular Characterization of the First Alternavirus Identified in Fusarium oxysporum. Viruses 2021, 13, 2026.Abstract: A novel mycovirus named Fusarium oxysporum alternavirus 1(FoAV1) was identified as infecting Fusarium oxysporum strain BH19, which was isolated from a fusarium wilt diseased stem of Lilium brownii. The genome of FoAV1 contains four double-stranded RNA (dsRNA) segments (dsRNA1, dsRNA 2, dsRNA 3 and dsRNA 4, with lengths of 3.3, 2.6, 2.3 and 1.8 kbp, respectively). Additionally, dsRNA1 encodes RNA-dependent RNA polymerase (RdRp), and dsRNA2-dsRNA3and dsRNA4-encoded hypothetical proteins (ORF2, ORF3 and ORF4), respectively. A homology BLAST search, along with multiple alignments based on RdRp, ORF2 and ORF3 sequences, identified FoAV1 as a novel member of the proposed family "Alternaviridae". Evolutionary relation analyses indicated that FoAV1 may be related to alternaviruses, thus dividing the family "Alternaviridae" members into four clades. In addition, we determined that dsRNA4 was dispensable for replication and may be a satellite-like RNA of FoAV1-and could perhaps play a role in the evolution of alternaviruses. Our results provided evidence for potential genera establishment within the proposed family "Alternaviridae". Additionally, FoAV1 exhibited biological control of Fusarium wilt. Our results also laid the foundations for the further study of mycoviruses within the family "Alternaviridae", and provide a potential agent for the biocontrol of diseases caused by F. oxysporum.
# Introduction
Mycoviruses are a group of viruses that infect filamentous fungi, yeasts and oomycetes. Until recently, the National Center of Biotechnology Information (NCBI) database contained more than 300 mycoviral sequences, divided into at least 19 families by the International Committee for Taxonomy of Viruses (ICTV). Mycoviruses contain double-stranded RNA (dsRNA), positive single-stranded RNA (+ssRNA), negative single-stranded RNA (-ssRNA) and single-stranded DNA (ssDNA) as their genome. Double-stranded RNA (dsRNA) mycoviruses are classified into seven families: Totiviridae, Endornaviridae, Partitiviridae, Chrysoviridae, Megabirnaviridae, Quadriviridae and Reoviridae, and one genus, Botybirnavirus.
Many mycoviruses affect fungal virulence and some can cause alterations in host phenotypes, reducing or enhancing virulence. This property of mycoviruses, which reduces the ability of their fungal hosts to cause disease, is termed hypovirulence. Hypovirulent mycoviruses have the potential to elicit biological control of plant pathogenic fungal diseases. Typical examples of hypovirulent mycovirus application include Cryphonectria hypovirus 1 (CHV1), which is used to control chestnut blight caused by Cryphonectria parasitica in Europe, and the Sclerotinia sclerotiorum hypovirulence-associated DNA virus 1 (SsHADV-1), which is used to control stalk break caused by Sclerotiorum sclerotiorum. In recent years, many hypovirulence-associated mycoviruses have been found in other plant pathogenic fungi, such as Botryosphaeria cinerea debilitation-related virus (BcDRV), Botryosphaeria dothidea Chrysovirus 1 (BdCV1), Rhizoctonia solani partitivirus 2 (RSPV2), Rhizoctonia solani endornavirus 1 (RsEV1) and Colletotrichum liriopes partitivirus 1 (ClPV1).
Further studies have revealed many novel viruses that differ obviously from alreadyknown viruses. For example,reported a novel mycovirus, Aspergillus mycovirus 341 (AsV341), from Aspergillus niger. The closest known relative of the AsV341 was a totivirus, Sphaeropsis sapinea RNA virus 2 (SsRV2), from Sphaeropsis sapinea. SsRV2 contains only one genomic RNA segment (5202 bp) and is a typical totivirus, based on the sequence characteristics identified and phylogenetic analysis with other known totiviruses. Moreover, AsV341 contains four segments (between 1.5 and 3.6 kbp), and the amino acid identity and similarity levels between AsV341 and SsRV2 are only 15.2% and 27.8%, respectively. Thus, AsV341 is obviously different from the known totivirus, even though its closest relative is SsRV2.reported a novel mycovirus, Alternaria alternata virus 1 (AaV1), from Alternaria alternata, which is similar to AsV341, following an analysis of its sequence characteristics and multiple alignment. However, phylogenetic analysis showed that AaV1 and AsV341 did not cluster with any mycovirus family. In 2013, Kozlakidis et al.reported a novel mycovirus (Aspergillus foetidus virus (AfV)), which was similar to both AaV1 and AsV341, and proposed a new genus, Alternavirus, and a new family, "Alternaviridae", to accommodate the three viruses. Subsequently, several mycoviruses were sequenced and identified as potential members of the proposed "Alternaviridae" family, including Fusarium poae alternavirus 1 (FpAV 1), Fusariumm graminerarum alternairus 1 (FgAV 1), Aspergillus heteromorphus alternavirus 1 (AheAV1) and Fusarium incarnatum alternavirus 1 (FiAV 1). Until recently, mycovirus classification into the proposed family "Alternaviridae" was based on phylogenetic analysis of RNA-dependent RNA polymerase (RdRp) sequences. Some genomes of viruses belonging to the proposed family "Alternaviridae" contained three dsRNA segments, while others comprised four dsRNA segments.
Fusarium oxysporum is an important plant pathogenic fungus that causes vascular wilt in a wide variety of agricultural crop species. Biocontrol agents, such as plant growthpromoting rhizobacteria (PGPR) strains and nonpathogenic F. oxysporum strains, have proven to be effective tools for controlling plant diseases caused by F. oxysporum. As more hypovirulence-associated mycoviruses have been investigated in attempts to control crop diseases, new insights and explorations regarding biological control using hypovirulenceassociated mycoviruses to control plant diseases caused by F. oxysporumhave become available. To our knowledge, six mycoviruses have been reported in F. oxysporum. These include Fusarium oxysporum chrysovirus 1 (FoCV1), Fusarium oxysporum f.sp. dianthi mycovirus 1 (FodV1), classified as family Chrysoviridae; Fusarium oxysporum f. sp. dianthi mitovirus 1 (FodMV1), classified as family Narnaviridae; Fusarium oxysporum f. sp. dianthi hypovirus 2 (FodHV2), classified as family Hypoviridae; Hadaka Virus 1 (HadV1), classified as family Polymycoviridae; and Fusarium oxysporum ourmia-like virus 1 (FoOuLV1), classified as family Botourmiaviridae. Only FodV1 and FoOuLV1 are known to cause hypovirulence.
A collection of F. oxysporum isolates (BH19) were obtained from fusarium wilt-diseased stems of Lilium brownii in the Liaoning province of China. Our preliminary study showed that a similar banding pattern of dsRNA may exist in this isolate. Sequencing and analysis of this dsRNA showed that it corresponded with a novel alternavirus-the first to be described that infects a F. oxysporum strain-which was provisionally named Fusarium oxysporium alternavirus 1 (FoAV1). Accordingly, we conducted genome characterization to elucidate the molecular features of FoAV1, investigate its impact on virulence and understand its biological properties in F. oxysporum.
# Materials and methods
## Strains and culture conditions
Seven isolates of F. oxysporiumwere used in the study. Strains BH19 and BH3 were isolated from a fusarium wilt-diseased stem of Lilium brownii in Liaoning province of China in 2019. Strains BH19-4V and BH19-3V were single-ascospore isolates of strain BH19. MC-1 was a F. oxysporum f. sp. momordicae strain incorporating a hygromycin-resistance gene (hygromycin B phosphotransferase), which has a normal colony morphology and high virulence in its hosts. Strains MC-1-4V and MC-1-3V were derivative strains isolated from MC-1 following confrontation training with BH19-4V and BH19-3V. All strains were stored at −80 - C in glycerol and cultured on potato dextrose agar (PDA) medium at 28 - C. For dsRNA extraction, mycelium was cultured on a PDA plate covered with cellophane membranes at 28 - C for 4-5 days.
## Dsrna extraction and purification
The extraction of dsRNA was performed as described previously. Strains were grown for 4-5 days on cellophane membranes on PDA medium. Fresh mycelia (1-2 g) were harvested to isolate dsRNA by selective absorption using cellulose powder CF-11 (Sigma-Aldrich, St. Louis, MO, USA), with nucleic acid co-precipitator added to improve the yield of dsRNA. The dsRNA was treated with RNase-free DNase I and S1 nuclease (Takara, Dalian, China). The dsRNAs were electrophoresed on a 1% agarose gel, stained with ethidium bromide, and visualized with a gel documentation and image analysis system (InGenius LhR, Syngene, UK).
## Cdna synthesis, cloning, and sequencing
The purified dsRNA samples were used as a template for cDNA synthesis. cDNA synthesis and cloning were conducted according to the methods described previously. A cDNA library was constructed using TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix ( ® TransGen Biotech, Beijing, China) according to the manufacturer's instructions. To obtain initial sequence clones, a random primer (RACE3RT) was used for RT-PCR amplification. Partial gap sequences between the initial sequences were filled by RT-PCR with sequence-specific primers designed from the obtained sequences. For the 5 and 3 terminal sequences cloning, an anchor primer PC3-T7 loop was ligated to the dsRNA using T4 RNA ligase and used for the RT reaction. Then, a PC2 primer (designed based on the corresponding sequence of the PC3-T7 loop) and sequence-specific primers (designed based on the proximal regions sequences) were used for the amplification of terminal sequences. All PCR products were cloned into a pMD18-T vector (Takara, Dalian, China), which was then transformed into E. coli DH5α cells (Takara, Dalian, China) for sequencing. To achieve high-quality consensus sequences, each nucleotide of full-length cDNA was sequenced at least three times. All the primers used for cDNA cloning and sequencing are listed in Supplementary Table S1.
## Sequence and phylogenetic analysis
Open reading frames (ORFs) and conserved domains were predicted using an ORF finder and CD-search on the NCBI website (http://www.ncbi.nlm.nih.gov) and motifs scan website (http://www.genome.jp/tools/motif/). Multiple sequence alignments were performed using sequences of known alternavirus and the CLUSTALX (2.0) program. Phylogenetic trees were constructed using MEGA7 software and generated by the maximumlikelihood (ML) method with 1000 bootstrap replicates.
## Horizontal transmission of hypovirulence traits
To assess the potential horizontal transmission of hypovirulence traits of strain BH19, dual culturing technology was used as described previously. Isolate BH19 and virus-free isolate MC-1, which is hygromycin resistant, were grown separately for 4-5 days on PDA medium, and then co-cultured for a further 4-5 days on PDA fresh plates using mycelium blocks of each isolate kept in close proximity (10-15 mm) to one another. Mixed mycelium blocks were then transferred to fresh PDA plates containing hygromycin, and mycelial growth resulting from co-cultivation were termed the derivative strains. Viral transmission in these strains was evaluated based on dsRNA extraction or RT-PCR detection.
## Rna extraction and rt-pcr detection
To extract total RNA, strains were grown for 4-5 days on cellophane membrane overlying PDA, and fresh mycelia were harvested and ground to a powder in liquid nitrogen. Total RNA was prepared using an RNA reagent (Newbio Industry, Wuhan, China), according to the manufacturer's instructions.
For RT-PCR detection of virus-transmitted strains, first-strand cDNAs were synthesized using TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix ( ® TransGen Biotech, Beijing, China) according to the manufacturer's instructions. Then, PCR amplification was performed using specific oligonucleotide primers (list in Supplementary Table S1), which were designed based on the sequences of FoAV1 dsRNA. All the amplicons were identified following electrophoresis in 1.5% agarose gels.
## Virulence assay
The virulence of virus-transmitted strains was assessed by pot experiments on bitter gourd plants, as described previously. For these experiments, bitter gourd seedlings were grown in boxes containing sterile soil, and 10 mL spores (10 7 mL −1 ) from the various F. oxysporum strains were inoculated into melon roots when the plants had grown to the second or third leaf stage. Control plants were inoculated with water. All inoculated plants were monitored for the development of typical wilt symptoms, and disease symptoms were photographed. All experiments were repeated twice.
# Results
## Dsrna segments in f. oxysporum strain bh19
The F. oxysporum strains BH19 and BH3 were both isolated from a fusarium wiltdiseased stem of L. brownii in Liaoning Province, China. Compared to BH3, the colony morphology of BH19 was clearly defined with a small aerial mycelium. Strain BH19 contained four dsRNA segments (named dsRNA1, dsRNA2, dsRNA3 and dsRNA4) while no dsRNAs were found in strain BH3. The four dsRNA bands had estimated sizes of 3.5 kbp (dsRNA1), 2.5 kbp (dsRNA2), 2.3 kbp (dsRNA3) and 1.8 kbp (dsRNA4), respectively. The segments were confirmed to be dsRNA in nature based on resistance to digestion with DNase I and S1 nuclease.
Viruses 2021, 13, x FOR PEER REVIEW 4 of 14 free isolate MC-1, which is hygromycin resistant, were grown separately for 4-5 days on PDA medium, and then co-cultured for a further 4-5 days on PDA fresh plates using mycelium blocks of each isolate kept in close proximity (10-15 mm) to one another. Mixed mycelium blocks were then transferred to fresh PDA plates containing hygromycin, and mycelial growth resulting from co-cultivation were termed the derivative strains. Viral transmission in these strains was evaluated based on dsRNA extraction or RT-PCR detection.
## Rna extraction and rt-pcr detection
To extract total RNA, strains were grown for 4-5 days on cellophane membrane overlying PDA, and fresh mycelia were harvested and ground to a powder in liquid nitrogen. Total RNA was prepared using an RNA reagent (Newbio Industry, Wuhan, China), according to the manufacturer's instructions.
For RT-PCR detection of virus-transmitted strains, first-strand cDNAs were synthesized using TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix ( ® TransGen Biotech, Beijing, China) according to the manufacturer's instructions. Then, PCR amplification was performed using specific oligonucleotide primers (list in Supplementary, which were designed based on the sequences of FoAV1 dsRNA. All the amplicons were identified following electrophoresis in 1.5% agarose gels.
## Virulence assay
The virulence of virus-transmitted strains was assessed by pot experiments on bitter gourd plants, as described previously. For these experiments, bitter gourd seedlings were grown in boxes containing sterile soil, and 10 mL spores (10 7 mL −1 ) from the various F. oxysporum strains were inoculated into melon roots when the plants had grown to the second or third leaf stage. Control plants were inoculated with water. All inoculated plants were monitored for the development of typical wilt symptoms, and disease symptoms were photographed. All experiments were repeated twice.
# Results
## Dsrna segments in f. oxysporum strain bh19
The F. oxysporum strains BH19 and BH3 were both isolated from a fusarium wiltdiseased stem of L. brownii in Liaoning Province, China. Compared to BH3, the colony morphology of BH19 was clearly defined with a small aerial mycelium. Strain BH19 contained four dsRNA segments (named dsRNA1, dsRNA2, dsRNA3 and dsRNA4) while no dsRNAs were found in strain BH3. The four dsRNA bands had estimated sizes of 3.5 kbp (dsRNA1), 2.5 kbp (dsRNA2), 2.3 kbp (dsRNA3) and 1.8 kbp (dsRNA4), respectively. The segments were confirmed to be dsRNA in nature based on resistance to digestion with DNase I and S1 nuclease.
## Molecular characterization of foav1 in strain bh19
Following cDNA cloning and sequencing, the complete sequences of the four dsRNA segments found in strain BH19 were obtained, comprising a virus nominated as Fusarium oxysporum alternavirus 1 (FoAV1). The complete sequence of FoAV1 dsRNA1 (GenBank Accession No. MT659125) was 3447 nucleotides (nt) long with a GC content of 56.6%, and it contained one ORF (ORF1) that initiated at position nt 629 and terminated at position nt 3385. Based on the universal genetic code, this sequence potentially encoded 917 amino acid (aa) residues with a calculated molecular mass (Mr) of 103.4 kDa.
## Molecular characterization of foav1 in strain bh19
Following cDNA cloning and sequencing, the complete sequences of the four dsRNA segments found in strain BH19 were obtained, comprising a virus nominated as Fusarium oxysporum alternavirus 1 (FoAV1). The complete sequence of FoAV1 dsRNA1 (GenBank Accession No. MT659125) was 3447 nucleotides (nt) long with a GC content of 56.6%, and it contained one ORF (ORF1) that initiated at position nt 629 and terminated at position nt 3385. Based on the universal genetic code, this sequence potentially encoded 917 amino acid (aa) residues with a calculated molecular mass (Mr) of 103.4 kDa. Using the homology search on BLASTP, FoAV1 ORF1 encoded a protein which was closely related to RdRps of alternaviruses, including Aspergillus foetidus dsRNA mycovirus (67.72%), Fusarium poae alternavirus 1 (67.61%) and Alternaria alternata virus 1 (66.63%); FoAV1 ORF2-and ORF3-encoded proteins were also only closely related to their equivalents. Alternavirus ORF2 and ORF3 proteins showed similar identities to those predicted for ORF1. FoAV1 ORF4 encoded a protein that had no significant similarities with any other protein in the NCBI database. These results indicated that FoAV1, which was isolated from F. oxysporum strain BH19, contained four dsR-NAs segments and was thus a new member of the proposed "Alternaviridae" family. Using the homology search on BLASTP, FoAV1 ORF1 encoded a protein which was closely related to RdRps of alternaviruses, including Aspergillus foetidus dsRNA mycovirus (67.72%), Fusarium poae alternavirus 1 (67.61%) and Alternaria alternata virus 1 (66.63%); FoAV1 ORF2-and ORF3-encoded proteins were also only closely related to their equivalents. Alternavirus ORF2 and ORF3 proteins showed similar identities to those predicted for ORF1. FoAV1 ORF4 encoded a protein that had no significant similarities with any other protein in the NCBI database. These results indicated that FoAV1, which was isolated from F. oxysporum strain BH19, contained four dsRNAs segments and was thus a new member of the proposed "Alternaviridae" family. Further similarities between FoAV1 and other members of the proposed family "Alternaviridae" were investigated using conserved domain database searches together with the multiple protein alignment of related proteins. The obtained results suggested that FoAV1 ORF1, which putatively encode an RdRp, and equivalent proteins encoded by other members of the proposed family "Alternaviridae", contained several conserved amino acid motifs. Similar conservation of several amino acid motifs was obtained when the proteins predicted from the sequences of FoAV1 ORF2 and ORF3 and related alternaviruses were compared (Supplementary. There were no significant similarities between the aa sequence of the protein predicted from FoAV1 ORF4 and the protein sequences predicted from other alternaviruses that contained four dsRNAs (Supplementary.
## Phylogenetic analysis of foav1
To examine the phylogenetic relationship between FoAV1 with other mycoviruses, a phylogenetic tree was constructed using a maximum likelihood method based on the RdRp (ORF1) sequences of FoAV1, proposed "Alternaviridae" family members and other related viruses from the families Ourmiaviridae, Partitiviridae, Megabirnaviridae, Totiviridae and Chrysoviridae. The results indicated that FoAV1 clustered with members of the proposed "Alternaviridae" family but that it was not a strain or isolate of a previously described alternavirus. The proposed "Alternaviridae" family is itself clustered into four clades (clades I, II, III and IV). Phylogenetic trees based on all the alternavirus ORF2 and ORF3 sequences, including those from FoAV1, showed similar clustering into four clades, as illustrated in,C, respectively.
## Phylogenetic analysis of foav1
To examine the phylogenetic relationship between FoAV1 with other mycoviruses, a phylogenetic tree was constructed using a maximum likelihood method based on the RdRp (ORF1) sequences of FoAV1, proposed "Alternaviridae" family members and other related viruses from the families Ourmiaviridae, Partitiviridae, Megabirnaviridae, Totiviridae and Chrysoviridae. The results indicated that FoAV1 clustered with members of the proposed "Alternaviridae" family but that it was not a strain or isolate of a previously described alternavirus. The proposed "Alternaviridae" family is itself clustered into four clades (clades Ⅰ, Ⅱ, Ⅲ and Ⅳ). Phylogenetic trees based on all the alternavirus ORF2 and ORF3 sequences, including those from FoAV1, showed similar clustering into four clades, as illustrated in,C, respectively.
## Untranslated regions sequences analysis of foav1
The coding sequences of all four FoAV1 dsRNA segments were flanked by two untranslated regions (UTRs) at the 5 -end and 3 -termini, excluding poly (A) tails. The FoAV1 5 -and 3 -UTR sequences of all four genomic dsRNAs were strictly conserved but shared little conservation with equivalent sequences of other alternaviruses which, within the clades, shared limited conservation with each other.
## Elimination of foav1 dsrna from f. oxysporum strain bh19 during subculture
During passage of the F. oxysporum strain BH19, it was discovered that some isolates (nominated BH19-3V), though not all (nominated BH19-4V), spontaneously lost the FoAV1 dsRNA4 genomic segment. The absence of FoAV1 dsRNA4 in BH19-3V compared to BH19-4V was confirmed following dsRNA extraction and agarose gel electrophoresis, as well as by RT-PCR amplification of a fragment of FoAV1 dsRNA4 in BH19-4V. The colony morphology and growth rate of F. oxysporum strains BH19-3V and BH19-4V were not significantly different. Maxim likelihood phylogenetic tree based on the FoAV1 ORF3 sequence and other "Alternaviridae" fam members. All phylogenetic analyses generated using MEGA7 with 1000 bootstrap replicates. Ⅰ to represent clades of the proposed "Alternaviridae" family. Different colors represent clustering in ferent families or clades.Ⅰ, Ⅱ, Ⅲ
and Ⅳ represent different clades of "Alternaviridae" family, resp tively.
## Untranslated regions sequences analysis of foav1
The coding sequences of all four FoAV1 dsRNA segments were flanked by two translated regions (UTRs) at the 5′-end and 3′-termini, excluding poly (A) tails. T FoAV1 5′-and 3′-UTR sequences of all four genomic dsRNAs were strictly conserved shared little conservation with equivalent sequences of other alternaviruses which, wit the clades, shared limited conservation with each other. Virulence testing of F. oxysporum on L. brownii plants proved difficult, so FoAV1 was transmitted from the BH19-3V and BH19-4V strains to F. oxysporum f.sp. momordicae MC-1 to determine any effects on virulence. Using dual culture, two derivative strains, MC-1-3V and MC-1-4V, were generated, and virus infection with, respectively, three and four genomic segments, was verified. Additionally, RT-PCR amplification was conducted. The colony morphology of strains MC-1-3V and MC-1-4V did not significantly differ from the wild-type strain MC-1.
## Elimination of foav1 dsrna from f. oxysporum strain bh19 during subculture.
During passage of the F. oxysporum strain BH19, it was discovered that some isolates (nominated BH19-3V), though not all (nominated BH19-4V), spontaneously lost the FoAV1 dsRNA4 genomic segment. The absence of FoAV1 dsRNA4 in BH19-3V compared to BH19-4V was confirmed following dsRNA extraction and agarose gel electrophoresis, as well as by RT-PCR amplification of a fragment of FoAV1 dsRNA4 in BH19-4V. The colony morphology and growth rate of F. oxysporum strains BH19-3V and BH19-4V were not significantly different. Fusarium wilt, caused by F. oxysporum, results in systemic infection and is characterized by a typical symptom: complete plant wilt. In order to determine whether the presence of FoAV1 reduced the virulence in F. oxysporum f. sp. momordicae strains, the strains MC-1, MC-1-3V and MC-1-4V were inoculated into bitter gourd seedlings at the stage of two-or three-leaf growth. Virulence was assessed before, during, and after the presentation of typical wilt symptoms. The results showed that bitter gourd plants inoculated with the virus-free MC-1 strain exhibited obvious dryness and wilting, while the plants inoculated with the MC-1-3V or MC-1-4V strains remained generally healthy with few or no symptoms.
## Effects of foav1 on the virulence of f. oxysporum f. sp. momordicae
Virulence testing of F. oxysporum on L. brownii plants proved difficult, so Fo transmitted from the BH19-3V and BH19-4V strains to F. oxysporum f.sp. momord 1 to determine any effects on virulence. Using dual culture, two derivative strain 3V and MC-1-4V, were generated, and virus infection with, respectively, three genomic segments, was verified. Additionally, RT-PCR amplification ducted. The colony morphology of strains MC-1-3V and MC-1-4V did nificantly differ from the wild-type strain MC-1.
## Effects of foav1 on the virulence of f. oxysporum f. sp. momordicae
Virulence testing of F. oxysporum on L. brownii plants proved difficult, so F transmitted from the BH19-3V and BH19-4V strains to F. oxysporum f.sp. momo 1 to determine any effects on virulence. Using dual culture, two derivative stra 3V and MC-1-4V, were generated, and virus infection with, respectively, thre genomic segments, was verified. Additionally, RT-PCR amplificatio ducted. The colony morphology of strains MC-1-3V and MC-1-4V d nificantly differ from the wild-type strain MC-1. ence of FoAV1 reduced the virulence in F. oxysporum f. sp. momordicae strains, the strains MC-1, MC-1-3V and MC-1-4V were inoculated into bitter gourd seedlings at the stage of two-or three-leaf growth. Virulence was assessed before, during, and after the presentation of typical wilt symptoms. The results showed that bitter gourd plants inoculated with the virus-free MC-1 strain exhibited obvious dryness and wilting, while the plants inoculated with the MC-1-3V or MC-1-4V strains remained generally healthy with few or no symptoms.
# Discussion
In this study, we identified and characterized a novel mycovirus, FoAV1, from F. oxysporium strain BH19, which is a causal agent of fusarium wilt disease in Lilium brownii. The genome of FoAV1 contains four dsRNA segments (dsRNA1, dsRNA2, dsRNA3 and dsRNA4) and a single open reading frame (ORF) in each dsRNA segment, namely: ORF1, ORF2, ORF3 and ORF4. BLAST searches revealed that ORF1, ORF2 and ORF3 were all similar in sequence to the ORFs predicted for several alternaviruses. We proposed that FoAV1 be considered a novel member of the proposed family "Alternaviridae".
With the advent of low cost, next-generation sequencing, there has been a dramatic increase in the generation of complete mycovirome sequences. Numerous mycoviruses have been described from different families, infecting many phytopathogenic fungi and entomopathogenic fungi. However, among the hundreds of mycoviruses described so far, only six mycoviruses have been identified in F. oxysporum, an important phytopathogenic species, with no reports of any viruses belonging to the proposed family "Alternaviridae". FoAV1 is the first alternavirus described to infect F. oxysporum.
The members of the proposed family "Alternaviridae" were all isolated from fungi, and clustered mainly according to homology BLAST, multiple alignment and phylogenetic analysis. Until now, only eight mycoviruses have been proposed as alternaviruses of the family "Alternaviridae"-and these were clustered into one branch, with no classification at the genus level. The RdRp sequence of FoAV1 was closely
# Discussion
In this study, we identified and characterized a novel mycovirus, FoAV1, from F. oxysporium strain BH19, which is a causal agent of fusarium wilt disease in Lilium brownii. The genome of FoAV1 contains four dsRNA segments (dsRNA1, dsRNA2, dsRNA3 and dsRNA4) and a single open reading frame (ORF) in each dsRNA segment, namely: ORF1, ORF2, ORF3 and ORF4. BLAST searches revealed that ORF1, ORF2 and ORF3 were all similar in sequence to the ORFs predicted for several alternaviruses. We proposed that FoAV1 be considered a novel member of the proposed family "Alternaviridae".
With the advent of low cost, next-generation sequencing, there has been a dramatic increase in the generation of complete mycovirome sequences. Numerous mycoviruses have been described from different families, infecting many phytopathogenic fungi and entomopathogenic fungi. However, among the hundreds of mycoviruses described so far, only six mycoviruses have been identified in F. oxysporum, an important phytopathogenic species, with no reports of any viruses belonging to the proposed family "Alternaviridae". FoAV1 is the first alternavirus described to infect F. oxysporum.
The members of the proposed family "Alternaviridae" were all isolated from fungi, and clustered mainly according to homology BLAST, multiple alignment and phylogenetic analysis. Until now, only eight mycoviruses have been proposed as alternaviruses of the family "Alternaviridae"-and these were clustered into one branch, with no classification at the genus level. The RdRp sequence of FoAV1 was closely related to the known proposed alternavirus, and FoAV1 contained similar conservation of several amino acid motifs, clustered into one branch by phylogenetic analysis with other alternaviruses. Additionally, with the appearance of FoAV1, the proposed family "Alternaviridae" was itself clustered into four clades through phylogenetic analysis based on all alternavirus RdRps, ORF2 and ORF3 sequences and the 5 and 3 -UTR alignment. Clade I contained Alternaria alternata virus 1 (AaV1) and Stemphylium lycopersici mycovirus (SlV); clade II contained Aspergillus foetidus dsRNA mycovirus (AfV), Aspergillus hetero-morphus alternavirus 1 (AheAV1) and Aspergillus mycovirus 341 mycovirus (AsV341); clade III contained only FoAV1; clade IV contained Fusarium incarnatum alter-navirus 1 (FiAV 1), Fusarium poae alternavirus 1 (FpAV 1) and Fusarium graminearum alternavirus 1 (FgAV 1). Therefore, we suggested that FoAV1 should be considered a new member of the proposed family "Alternaviridae" and should also be used in further studies.
In past studies, an interesting phenomenon was discovered; namely, that several segments of viruses are dispensable for replication, and these dispensable segments are considered satellite-like RNAs. Satellite RNAs may have no effect at all, while some satellite RNAs are able to play important roles in biological functions. Interestingly, according to the above classification of proposed alternavirus members, all members (AaV1 and SlV) in clade I have four segments, two members (AfV and AsV341) in clade II have four segmentsand all members (FiAV 1, FpAV 1 and FgAV 1) in clade IV have three segments. As a member of clade III, FoAV1 has four segments, but dsRNA4 can eliminate these during subculture, with no significant difference in the effect (colony morphology, growth rate and virulence) on the host regarding the presence or absence of dsRNA4. Therefore, the dsRNA4 may be a satellite-like RNA with no known biological function, but it may also be related to the evolution of alternaviruses and used as evidence for the proposed family "Alternaviridae".
Fusarium wilt caused by F. oxysporum is an important disease in a wide variety of agricultural crop species. Biological control is an effective method for combatting this disease. In recent years, a large number of hypovirulent mycoviruses, such as CHV1and SsHADV1, have been identified and explored as potential biocontrol agents against fungal diseases. However, until now, only two mycoviruses, FodV1 and FoOuLV1, isolated from F. oxysporum, exerted hypovirulent effects. FoAV1 exhibited a good biological control effect for Fusarium wilt, suggesting that it may be a good resource for controlling Fusarium wilt.
In conclusion, we reported herein that FoAV1 is a new alternavirus in F. oxysporum, and that FoAV1 infection can cause hypovirulence in a host. Our results not only lay a foundation for the further study of mycoviruses within the proposed family "Alternaviridae", but also provide a potential agent for the biocontrol of diseases caused by F. oxysporum.
# Supplementary materials:
The following are available online at https://www.mdpi.com/article/10 .3390/v13102026/s1,: A list of primers used in this study,: Multiple alignment of the ORF2 (A), ORF3 (B) and ORF4 (C) amino acid motifs encoded by FoAV1 and other "Alternaviridae" family members, respectively.
## Conflicts of interest:
The authors declare no conflict of interest. |
OX-HDL: A Starring Role in Cardiorenal Syndrome and the Effects of Heme Oxygenase-1 Intervention
In this review, we will evaluate how high-density lipoprotein (HDL) and the reverse cholesterol transport (RCT) pathway are critical for proper cardiovascular-renal physiology. We will begin by reviewing the basic concepts of HDL cholesterol synthesis and pathway regulation, followed by cardiorenal syndrome (CRS) pathophysiology. After explaining how the HDL and RCT pathways become dysfunctional through oxidative processes, we will elaborate on the potential role of HDL dysfunction in CRS. We will then present findings on how HDL function and the inducible antioxidant gene heme oxygenase-1 (HO-1) are interconnected and how induction of HO-1 is protective against HDL dysfunction and important for the proper functioning of the cardiovascular-renal system. This will substantiate the proposal of HO-1 as a novel therapeutic target to prevent HDL dysfunction and, consequently, cardiovascular disease, renal dysfunction, and the onset of CRS.
# Introduction
Admission of heart failure (HF) patients in American hospitals exceeds 1 million people annually, with heart failure or coronary artery disease more prevalent in the more advanced stages of chronic kidney disease (CKD). Patients with heart failure are more predisposed to the development of acute kidney injury (AKI) . This sets the stage for the relevance of the cardiovascular-renal system as an interrelationship between kidney function and cardiovascular health. Epidemiologic studies have investigated how renal dysfunction is a prominent risk factor for cardiovascular disease (CVD)and how heart failure predisposes to kidney damage and/or the exacerbation of chronic kidney dysfunction.However, in vitro and in vivo studies have yet to establish a definitive pathophysiological explanation for this phenomenon, otherwise known as cardiorenal syndrome (CRS).
High-density lipoprotein-cholesterol (HDL-C) levels have become promising markers for the risk of CVD and even CKD. Recent studies have cited that HDL function is more important than levels and that remodeling and dysfunction likely contribute to increased risk of CVD, CKD, and CRS. HDL utilizes its protective effect through multiple mechanisms, including lowering tissue cholesterol tissue cholesterol levels through reverse cholesterol transport, attenuation of low-density lipoprotein (LDL) oxidation, and decreasing inflammatory responses via association with paraoxonase 1 (PON1)Dysfunctional HDL can result from free radical attack or oxidation of "good" HDL, leading to Ox-HDL ("bad" HDL). Lipids and lipoproteins are the major culprits of free radical damage, which results in lipid peroxidation. Free-radical-mediated lipid peroxidation alters the biophysical properties of cell membranes, which may impair normal cellular function. Furthermore, the generation of lipid peroxidation products, i.e., F2-isoprostane from arachidonic acid, may propagate the free radical damage via covalent modification of biomolecules. It is pertinent to find an endogenous antioxidant that can prevent the remodeling of "good" HDL to proinflammatory and atherogenic "bad" HDL. For CRS, it is crucial to find an antioxidant that can directly target the cardiovascular-renal system to protect from oxidative damage and dysfunction to the cardiovascular or renal systems.
## The heme oxygenase system
Heme oxygenase-1 (HO-1) is an inducible enzyme within the body that is responsible for the catabolism of heme to equimolar parts carbon monoxide (CO) and biliverdin/bilirubin and the release of free iron. HO-1 is known as a stress response protein, and we associate its induction with protection against reactive oxygen species (ROS) and, subsequently, oxidative stress. We find heme oxygenase in the kidney in two isoforms: HO-1 (inducible form) and HO-2. Studies have shown HO-1 to be necessary and important for renal vascular and tubular function. HO-1 has been shown to be important in vascular protection and function through the induction of adiponectin. HO-1 is a potential and relevant therapeutic target for protection against and amelioration of CRS. HO-1 has been shown to decrease levels of angiotensin-II-mediated isoprostane production in endothelial cells. Heme oxygenase has been proven to be cardioprotective, with associated induction of adiponectin expression, reduction in Ox-HDL with suppression of isoprostane production, and isoprostane binding to HDL.Additionally, HO-1 may protect against heme-mediated damage in CRS, as hemoglobin has been shown to bind to "good" HDL and alter its conformation and functionality, causing HDL to become proinflammatory and atherogenic..
Adiponectin is a cardioprotective protein hormone and has been shown to have increased plasma levels and improve vascular function after induction of HO-1. Adiponectin is well established as necessary for proper cardiovascular health, with clinical investigations proving adiponectin deficiency (hypoadiponectinemia) as an independent risk factor for CVD. Researchers have also found hypoadiponectinemia to be strongly associated with renal dysfunction and CKD, and circulating adiponectin levels may be a predictor for CKD. These studies show that adiponectin is crucial for vascular and renal function, with decreased levels leading to cardiovascular-renal dysfunction. The heme oxygenase system may act as a potential therapeutic target for protection against and amelioration of cardiorenal syndrome through many pathways. Heme oxygenase is a potent endogenous antioxidant, and pharmacological induction of HO-1 improved type-1 cardiorenal syndrome in postischemic SCID mice. HO-1 has been shown to decrease levels of angiotensin-II-mediated isoprostane-induced oxidative stress production in endothelial cells.
## Structure of hdl and reverse cholesterol transport pathway
HDL is a small, dense lipoprotein particle with a high ratio of proteins to lipids, synthesized by liver hepatocytes. HDL is a powerful anti-inflammatory agent that inhibits atherogenesis. The HDL proteome is very complex, with over 550 proteins reported in HDL. HDL proteomics is a relatively novel approach to understanding the complex makeup and function of HDL in the setting of oxidative stress. Sixteen HDL species with distinct proteomic signatures have been identified. The HDL proteome is complex and separate from HDL cholesterol. Mass spectrometry has been used to understand complete HDL analysis and to identify biomarkers in order to better understand HDL function. In fact, the HDL lipoprotein proteome has a high correlation with risk factors for cardiovascular disease and atherosclerotic burden and calcification on CT angiograms of coronary arteries. The diversity of the HDL proteome is associated with clinical outcomes in patients with heart failure. Alterations in the HDL proteome have been shown to result in dysfunctional HDL particles in type I diabetics.
HDL is primarily composed of several apolipoproteins, including Apo A-1, Apo A-II, and other proteins, including the enzyme paraoxonase. The external layer of HDL comprises free cholesterol, apolipoproteins (e.g., apoA-I, apoA-II, apoC, apoE), and phospholipids and is amphipathic. The inner core of the HDL particle is highly concentrated with cholesterol esters, contains a minute amount of triglycerides, and is, therefore, hydrophobic.
HDL-C is considered the "good cholesterol" because of the physiologic function it performs in "reverse cholesterol transport". This is the process where the HDL particles move through the circulation and extract free cholesterol from less-dense particles and transport the free cholesterol to the liver to be processed and expelled, reducing the overall level of total cholesterol. HDL is the smallest of the lipoproteins but contains the highest apolipoprotein/ lipid ratio. The cholesterol delivered to the liver is excreted in bile and eventually converted into bile acids. Delivery of HDL cholesterol to organs like adrenals, ovaries, and testes is critical for steroid hormone synthesis. The efflux of cholesterol from HDL involves the following regulatory proteins.
ApoA-I has been shown to bind cholesterol in vitro and in vivo, showing its role in the uptake of extrahepatic cholesterol by HDL while traveling through the circulation. Infusing patients with pro-apo-A-I was shown to increase RTC. ApoA-I is synthesized in and interacts with the protein ATP-binding cassette transporter A 1 (ABCA1) in hepatocytesand is then secreted into circulation as a lipid-poor particle. The lipid-poor apoA-1 particle removes cholesterol from the surfaces of macrophages in the arterial wall through interaction with ABCA1, forming nascent prebeta HDL or HDL 2 particles. Phospholipid transfer protein (PLTP) is responsible for the transfer of phospholipids from triglyceride-rich lipoproteins to form nascent HDL particles and has been shown to interact with apoA-1. Apo A-1 oxidation has been shown to increase HDL oxidation and dysfunction.
The plasma enzymelecithin-cholesterol acyl transferase (LCAT) is responsible for the conversion of free cholesterol into cholesterol esters, which is a more hydrophobic form of cholesterol, making it easier to be sequestered into the core of the lipoprotein particle. This eventually causes the newly synthesized HDL to assume a spherical shape; the nascent HDL swells into a round, "mature" HDL particle. The mature HDL particle increases in size with the addition of more cholesterol and phospholipids from cells and other lipoproteins while circulating through the bloodstream.
A protein involved in the oxidation of HDL is myeloperoxidase (MPO), found in neutrophils and monocytes; it is released during acute inflammation. MPO generates Ox-HDLs, which cannot bind scavenger receptor class B type 1 (SR-B1) on the membrane of liver cells for RTC and have a proinflammatory function expressed through the upregulation of the protein (VCAM-1) on endothelial cells and the activation of NF-kB and Ox-HDL, which lose their ability to activate eNOS and to inhibit caspase-3 and, therefore, lose their antiapoptotic activity. HDL in patients with documented heart disease have elevated levels of nitrotyrosine and chlorotyrosine, both products of myeloperoxidase oxidative processes that facilitate the generation of dysfunctional HDL.
Alpha-HDL, containing free cholesterol and a cholesterol ester core, returns to the liver for selective uptake of cholesterol via contact with SR-B1. Cholesterol ester transfer protein (CETP) regulates the exchange of cholesterol esters from HDL to apo-B-containing lipoproteins for triglyceridesApo-containing lipoproteins then transfer cholesterol to the liver through interaction with hepatocyte LDL receptors.
The above processes are controlled by additional regulatory input from other proteins, including hormones and receptors. For example, the peroxisomal proliferator-activated receptor-alpha (PPARα) has been shown to upregulate transcription of the ApoA-I gene; transcription of the gene for ABCA1 is regulated by liver X receptors (LXRs) and retinoid X receptors (RXRs). Each of the HDL proteins, RCT pathway proteins, and regulatory factors highlighted above represent a potential point of intervention to raise the HDL-C number and/or promote reverse cholesterol transport.
## Sexual dimorphism and hdl
There are major differences in HDL levels between men and premenopausal women. Obesity affects cardiometabolic function in both men and women but affects premenopausal women to a much lesser degree, even when matched to age and weight controls. Sexual dimorphism is important to understand since there are major sex differences in fat distribution in visceral organs, skeletal muscle, and epicardial fatAdipose tissue distribution and adipose tissue health are responsible for differences in insulin sensitivity and consequent systemic inflammation. HDL levels are much higher in premenopausal women, but our group has shown a much higher Ox-HDL/HDL ratio in obese women, reducing the anti-inflammatory index of HDL and increasing their risk of endothelial cell dysfunction.
## Cardiorenal syndrome: a definition
Cardiorenal syndrome is an umbrella term that encompasses the interaction between the heart and kidneys, such that injury to one organ causes dysfunction in the other. There is well-established crosstalk in the pathophysiology of the heart and kidneys. Patients on dialysis with end-stage renal disease (ESRD) have ten times the risk of death by a cardiovascular event than the general population. Similarly, patients admitted with heart failure acquire renal dysfunction, and combined heart and kidney failure is associated with poor clinical outcomes. The primary dysfunctional organ can be the heart or the kidney. Cardiorenal syndrome is now defined more strictly than epidemiological outcomes in heart failure and dialysis patients. There are currently five types of cardiorenal syndrome, each with unique pathophysiology and progression of illness . . Types of Cardiorenal Syndrome.
## Type 1 acute cardiorenal syndrome
Acute decompensated heart failure Acute kidney injury
## Type 2 chronic cardiorenal syndrome
Chronic heart failure Chronic kidney disease
## Type 3 acute renocardiac syndrome
Acute kidney injury Acute heart failure
## Type 4 chronic renocardiac syndrome
Chronic kidney disease Chronic heart failure
## Type 5 secondary cardiorenal syndrome
Codevelopment of heart failure and chronic kidney disease due to acute or chronic systemic disorder Type 1 cardiorenal syndrome is acute in nature, categorized as acute heart failure that leads to AKI, clinically presenting with inadequate renal perfusion due to an increase in venous pressure or a low cardiac output state, leading to kidney congestion. Type 1 CRS is a common occurrence; AKI occurs in about 25% of hospitalized patients with HF, and declining renal function has been identified as an independent predictor of mortality. AKI activates the renin-angiotensin-aldosterone system (RAAS), salt and water imbalance, and vasoconstriction, all of which contribute to continued heart damage. HF leads to decreased renal perfusion, along with monocyte and endothelial activation, causing cytokine secretion and further depressing renal function. This bidirectional pathophysiology is also exacerbated with common therapeutic agents such as ACE inhibitors and diuretics, leading to toxicity and vasoconstriction. Production of epinephrine and angiotensin, along with a decreased sensitivity to vasodilators such as nitric oxide, causes excessive vasoconstriction, thereby worsening cardiorenal function. Fibrosis from inflammation is a common feature in HF and CKDand may well be the unifying pathophysiology of the entire CRS continuum. Evaluation of creatinine levels can be misleading, as the SOLVD trial showed that early initiation of RAAS inhibitors may reflect early changes in renal hemodynamics and may not reflect kidney injury. Signs of CRS include an increase in the serum creatinine by 0.3 mg/dl in a 48-h period, an increase in the serum creatinine to 1.5 times baseline, or a urine volume less than 0.5ml/kg/h over a six-hour period.
Diagnostics 2020, 10, x FOR PEER REVIEW 7 of 23 in a 48-h period, an increase in the serum creatinine to 1.5 times baseline, or a urine volume less than 0.5ml/kg/h over a six-hour period.
## Figure 2.
Pathophysiology of acute cardiorenal syndrome (type 1): mechanism of how acute heart failure leads to acute kidney injury (AKI) due to inadequate renal perfusion, endothelial activation, and cytokine production, which activates the RAAS, salt and water imbalance, and vasoconstriction, further exacerbating AKI. CVP= central venous pressure; IAP= intraabdominal pressure; ACEi =angiotensin converter enzyme inhibitor; RAAS= renin-angiotensin-aldosterone system; SNS= sympathetic nervous system; GFR= glomerular filtration rate.
Type 2 CRS is similar to type 1; however, where type 1 is acute, type 2 CRS is chronic. Type 2 CRS is characterized by chronic cardiac dysfunction (e.g., chronic congestive heart failure), which causes progressive CKD. The prevalence of renal dysfunction in chronic HF is significantly large; nearly 50% of patients with chronic HF appear to have decreased GFR. Even a minimal decrease in GFR is a strong independent predictor of mortality. Neurohormonal activation is also present in type 2 CRS, namely, increased production of vasoconstrictive mediators (angiotensin) and altered release of vasodilatory mediators (nitric oxide). Progression of CKD can be attributed to multiple factors including, but not limited to, low cardiac output, inflammation, endothelial dysfunction, accelerated atherosclerosis, chronic hypoperfusion, and increased renal vascular resistance. The progression of CKD further instigates cardiac dysfunction through RAAS activation, hypertension, and anemia. Pathophysiology of acute cardiorenal syndrome (type 1): mechanism of how acute heart failure leads to acute kidney injury (AKI) due to inadequate renal perfusion, endothelial activation, and cytokine production, which activates the RAAS, salt and water imbalance, and vasoconstriction, further exacerbating AKI. CVP = central venous pressure; IAP = intraabdominal pressure; ACEi = angiotensin converter enzyme inhibitor; RAAS = renin-angiotensin-aldosterone system; SNS = sympathetic nervous system; GFR = glomerular filtration rate. Type 2 CRS is similar to type 1; however, where type 1 is acute, type 2 CRS is chronic. Type 2 CRS is characterized by chronic cardiac dysfunction (e.g., chronic congestive heart failure), which causes progressive CKD. The prevalence of renal dysfunction in chronic HF is significantly large; nearly 50% of patients with chronic HF appear to have decreased GFR. Even a minimal decrease in GFR is a strong independent predictor of mortality. Neurohormonal activation is also present in type 2 CRS, namely, increased production of vasoconstrictive mediators (angiotensin) and altered release of vasodilatory mediators (nitric oxide). Progression of CKD can be attributed to multiple factors including, but not limited to, low cardiac output, inflammation, endothelial dysfunction, accelerated atherosclerosis, chronic hypoperfusion, and increased renal vascular resistance. The progression of CKD further instigates cardiac dysfunction through RAAS activation, hypertension, and anemia.
Type 3 CRS is also called renocardiac syndrome and is categorized with abrupt kidney injury being the primary illness (e.g., ischemia, hypoperfusion, glomerulonephritis), leading to acute cardiac dysfunction (e.g., HF, arrhythmia, ischemia;. AKI is a powerful predictor of hospital mortality; however, type 3 CRS is less common than type 1 CRS. AKI negatively impacts cardiac function through a variety of mechanisms. Renal ischemia has been shown to induce inflammation and apoptosis in cardiac cells. AKI contributes to acute heart dysfunction via familiar mechanisms from the abovementioned CRS types: RAAS activation, hypertension, decreased GFR, endothelial activation, and cytokine secretion. Cytokines such as tumor necrosis factor (TNF), IL-1, and IL-6 play a diagnostic role and are also a pathogenic cause of myocardial cell damage and apoptosis during ischemic AKI. Furthermore, myeloperoxidase, a biomarker of oxidative stress and inflammation in acute coronary syndrome, may cause apoptosis and play a potential role in the pathogenesis of CRS. Type 3 CRS is also called renocardiac syndrome and is categorized with abrupt kidney injury being the primary illness (e.g., ischemia, hypoperfusion, glomerulonephritis), leading to acute cardiac dysfunction (e.g., HF, arrhythmia, ischemia;. AKI is a powerful predictor of hospital mortality; however, type 3 CRS is less common than type 1 CRS. AKI negatively impacts cardiac function through a variety of mechanisms. Renal ischemia has been shown to induce inflammation and apoptosis in cardiac cells. AKI contributes to acute heart dysfunction via familiar mechanisms from the abovementioned CRS types: RAAS activation, hypertension, decreased GFR, endothelial activation, and cytokine secretion. Cytokines such as tumor necrosis factor (TNF), IL-1, and IL-6 play a diagnostic role and are also a pathogenic cause of myocardial cell damage and apoptosis during ischemic AKI. Furthermore, myeloperoxidase, a biomarker of oxidative stress and inflammation in acute coronary syndrome, may cause apoptosis and play a potential role in the pathogenesis of CRS. : mechanism involves an acute insult to kidney function that results in the acute sympathetic nervous system activating a cascade of inflammatory responses, causing acute heart failure. NO= nitrous oxide; RAAS= renin-angiotensinaldosterone system; ROS= reactive oxygen species; SNS=sympathetic nervous system.
Type 4 CRS is also known as chronic renocardiac syndrome, which involves primary CKD that contributes to worsening heart function and an increased risk for cardiovascular events. Current estimates for CKD in the general population exceed 10% of the US adult population. CKD is an independent risk factor for cardiovascular-event-related mortality in individuals at all stages of CKD, especially for ESRD, which has an increased risk of cardiac death compared to patients without CKD. CKD is considered a more significant predictor of cardiovascular disease than diabetes mellitus. Mechanisms for cardiac remodeling and decreased function from CKD are multifarious in nature. In Stages 1 and 2, risk factors resulting in CKD (e.g., obesity, hypertension, dyslipidemia, and chronic inflammation) contribute to decreased cardiac function. In Stages 3 and 4, anemia, uremia Type 4 CRS is also known as chronic renocardiac syndrome, which involves primary CKD that contributes to worsening heart function and an increased risk for cardiovascular events. Current estimates for CKD in the general population exceed 10% of the US adult population. CKD is an independent risk factor for cardiovascular-event-related mortality in individuals at all stages of CKD, especially for ESRD, which has an increased risk of cardiac death compared to patients without CKD. CKD is considered a more significant predictor of cardiovascular disease than diabetes mellitus. Mechanisms for cardiac remodeling and decreased function from CKD are multifarious in nature. In Stages 1 and 2, risk factors resulting in CKD (e.g., obesity, hypertension, dyslipidemia, and chronic inflammation) contribute to decreased cardiac function. In Stages 3 and 4, anemia, uremia toxins, nutritional status, and BMI, along with chronic inflammation, lead to increased ischemic risk, coronary calcification, and neurohormonal abnormalities, as previously discussed. In late-stage CKD leading into dialysis, we see chronic inflammation (again), renal toxicity, endothelial dysfunction, oxidative stress, and accelerated atherosclerosis as contributing factors to the vicious cycle of cardiac dysfunction with renal failure.
Diagnostics 2020, 10, x FOR PEER REVIEW 10 of 23. Pathophysiology of renocardiac syndrome (type 4): mechanism involving chronic kidney injury causing worsening heart function. The various causes of CKD, including diabetes and hypertension, accelerate cardiac remodeling, leading to poor cardiac function. This pathway involves RAAS activation and sympathetic nervous system activation.
Type 5 cardiorenal syndrome results in simultaneous renal and cardiac failure due to acute or chronic systemic disorders. Diabetes, sepsis, amyloidosis, and sarcoidosis are examples of such diseases affecting combined renal and cardiac function . Sepsis accounts for the most common condition that can acutely affect both organs, with the mechanisms being unclear but possibly involving TNF. Patients presenting with sepsis have multiorgan dysfunction over 50% of the time, especially involving cardiac and renal function, explaining the high mortality involved.We observed this during the COVID-19 pandemic with systemic vascular inflammation. . Sepsis accounts for the most common condition that can acutely affect both organs, with the mechanisms being unclear but possibly involving TNF. Patients presenting with sepsis have multiorgan dysfunction over 50% of the time, especially involving cardiac and renal function, explaining the high mortality involved.We observed this during the COVID-19 pandemic with systemic vascular inflammation.
Diagnostics 2020, 10, x FOR PEER REVIEW 11 of 23 . Summary of the pathophysiology of secondary cardiorenal syndrome (type 5): mechanism is characterized by combined cardiac and renal failure due to acute or chronic processes that cause hemodynamic instabilities, hypercoagulability, neurohormonal imbalances, and toxicity and hypoxia that cause poor renal perfusion, myocardial ischemia, and hypoxia. SVR= systemic vascular resistance; ROS= reactive oxygen species; DIC= disseminated intravascular coagulation; SNS= sympathetic nervous system; RAAS= renin-angiotensin-aldosterone system.
## Cardiorenal syndrome and ho-1
Physiological interaction between organs is necessary for the optimal equilibrium and functioning of the organism. Derangements in these interactions can initiate multiorgan dysfunction. In particular, heart and kidney functions are closely interrelated through a variety of dynamic and bidirectional mechanisms; a pathological alteration in one organ can unfavorably affect function in another distant organ.
CRS involves complex interactions at the molecular level that induce vessel inflammation, atherosclerosis, cardiac fibrosis, and hypertrophy; in addition, structural and biochemical abnormalities can adversely affect cardiovascular or renal function.
Bright initially deliberated the causal association between chronic kidney disease (CKD) and cardiovascular risk in 1836. Patients with CKD are among the highest risk groups for adverse cardiovascular events and cardiovascular-related mortality and, therefore, require particular clinical attention. A recent study provided insight into the pathogenesis of CRS type 1, emphasizing the pivotal role of oxidative stress in CRS type 1. The study revealed that levels of oxidative stress markers (myeloperoxidase, nitric oxide, copper/zinc superoxide dismutase, and endogenous peroxidase activity) were significantly higher in CRS type 1 than in acute heart failure without CRS type 1 and in healthy controls. In particular, CRS type 1 patients presented a significant increase in circulating ROS and RNS and an increased expression of the inflammatory cytokine IL-6.
[132] Monu et al. showed that ang-II-mediated recruitment of T-lymphocytes and increased oxidative stress is decreased by the upregulation of HO-1 in a model of postischemic heart failure. The results showed that HO-1 induction decreased renal vasoconstriction and fibrosis and improved renal function in both immunocompetent and T-lymphocyte-suppressed mice. Interestingly, treatment with SnMP, a known HO activity inhibitor, reversed the beneficial effects of HO-1 induction, suggesting that increased levels of HO activity play a central role in preventing MIinduced cardiac and renal damage in this CRS animal model. HO-1 induction reduces postischemic pathological cardiac remodeling and, in mice with advanced heart failure and CRS, improves cardiac . Summary of the pathophysiology of secondary cardiorenal syndrome (type 5): mechanism is characterized by combined cardiac and renal failure due to acute or chronic processes that cause hemodynamic instabilities, hypercoagulability, neurohormonal imbalances, and toxicity and hypoxia that cause poor renal perfusion, myocardial ischemia, and hypoxia. SVR = systemic vascular resistance; ROS = reactive oxygen species; DIC = disseminated intravascular coagulation; SNS = sympathetic nervous system; RAAS = renin-angiotensin-aldosterone system.
## Cardiorenal syndrome and ho-1
Physiological interaction between organs is necessary for the optimal equilibrium and functioning of the organism. Derangements in these interactions can initiate multiorgan dysfunction. In particular, heart and kidney functions are closely interrelated through a variety of dynamic and bidirectional mechanisms; a pathological alteration in one organ can unfavorably affect function in another distant organ.
CRS involves complex interactions at the molecular level that induce vessel inflammation, atherosclerosis, cardiac fibrosis, and hypertrophy; in addition, structural and biochemical abnormalities can adversely affect cardiovascular or renal function.
Bright initially deliberated the causal association between chronic kidney disease (CKD) and cardiovascular risk in 1836. Patients with CKD are among the highest risk groups for adverse cardiovascular events and cardiovascular-related mortality and, therefore, require particular clinical attention. A recent study provided insight into the pathogenesis of CRS type 1, emphasizing the pivotal role of oxidative stress in CRS type 1. The study revealed that levels of oxidative stress markers (myeloperoxidase, nitric oxide, copper/zinc superoxide dismutase, and endogenous peroxidase activity) were significantly higher in CRS type 1 than in acute heart failure without CRS type 1 and in healthy controls. In particular, CRS type 1 patients presented a significant increase in circulating ROS and RNS and an increased expression of the inflammatory cytokine IL-6. showed that ang-II-mediated recruitment of T-lymphocytes and increased oxidative stress is decreased by the upregulation of HO-1 in a model of postischemic heart failure. The results showed that HO-1 induction decreased renal vasoconstriction and fibrosis and improved renal function in both immunocompetent and T-lymphocyte-suppressed mice. Interestingly, treatment with SnMP, a known HO activity inhibitor, reversed the beneficial effects of HO-1 induction, suggesting that increased levels of HO activity play a central role in preventing MI-induced cardiac and renal damage in this CRS animal model. HO-1 induction reduces postischemic pathological cardiac remodeling and, in mice with advanced heart failure and CRS, improves cardiac function and renal vasoconstriction. This renal vasoconstriction was demonstrated in a murine model of type 1 CRS, secondary to postischemic changes of LAD ligation.
Accumulating evidence suggests that hyperuricemia is one of the important factors that may significantly contribute to the development and progression of CRS. Elevated levels of uric acid have been associated with inflammation, oxidative stress, insulin resistance, dysglycemia, endothelial dysfunction, vascular, renal and cardiac stiffness, cardiac diastolic dysfunction, renal hyperfiltration, and proteinuria, all of which are components of CRS. The significance of a westernized diet, which is high in fructose, and hyperuricemia in the development of CRS is underscored by the relationship between increased consumption of sugar-sweetened beverages, hyperuricemia, and all components of this syndrome. showed that induction of HO-1 reduced expression of xanthine oxidase and NADPH oxidase, enzymatic systems that are important for ROS production, in adipocytes treated with fructose, a fuel source that increases uric acid levels. Additionally, showed that mice following a fructose diet presented an increase in isoprostanes that was associated with a decrease in HO-1 expression and an increase in heme levels. Isoprostanes and heme are regarded as valid markers of oxidative stress. Upregulation of HO-1 presents cardioand reno-protective functions mediated by its antioxidative, anti-inflammatory, and antiapoptotic properties. In animal models of myocardial ischemia (MI), both overexpression and pharmacological induction of HO-1 reduce infarct size and ventricular remodeling after ischemia-reperfusion damage by improving cardiac metabolism. Increased HO-1 expression has a protective effect against ischemia-reperfusion injury in the kidneyand can correct blood pressure elevation following ang-II exposure.
Hyperglycemia-induced mitochondrial oxidative stress, a cause of metabolic CRS, is a contributory factor to increased risk of cardiovascular disease, which can induce cellular injury and cell dysfunction. The molecular mechanism of mitochondrial dysfunction in CRS is driven via abnormalities involving the transcriptional coactivator peroxisome proliferator-activated receptor gamma coactivator 1α (PGC-1α), which controls the biogenesis of mitochondria and mitochondrial function in a variety of tissue and cell types. PGC-1α is regulated by the endothelial NO synthase that plays an important role in mitochondrial biogenesis. Studies have shown that decreased expression of PGC-1α-associated impairment of mitochondrial biogenesis may be responsible for various metabolic abnormalities in CRS. Thus, the impairment of the complex steps in the regulation of mitochondrial biogenesis may contribute to CRS.
Activation of PGC-1α reduced mitochondrial ROS, prevented adipogenesis in adipocytes, and protected diabetic hearts from hyperglycemia-mediated oxidative stress. It has been shown that the protective effect of EETs in diabetic mice involves increased expression of PGC-1α and SIRT1. In cardiomyocytes, pharmacological inhibition of SIRT1 was followed by decreased expression of PGC-1α and increased ROS production. At the same time, Singh et al. showed that PGC-1α regulates HO-1 expression, confirming the beneficial effect of HO-1 in cardiovascular diseases and lipid metabolism. Heme oxygenase-1 upregulation of PGC-1α signaling in epicardial fat attenuated cardiovascular risk in both humans and mice. Adipocyte-specific HO-1 gene therapy is very effective in reducing oxidative stress, improving both insulin resistance and vascular function in obese mice. Finally, upregulation of HO-1 by icosapent ethyl, pomegranate seed oil, and black seed oil (thymoquine) have all reduced oxidative stress, improved insulin resistance, and improved mitochondrial function. receptor gamma coactivator 1α (PGC-1α), which controls the biogenesis of mitochondria and mitochondrial function in a variety of tissue and cell types. PGC-1α is regulated by the endothelial NO synthase that plays an important role in mitochondrial biogenesis. Studies have shown that decreased expression of PGC-1α-associated impairment of mitochondrial biogenesis may be responsible for various metabolic abnormalities in CRS. Thus, the impairment of the complex steps in the regulation of mitochondrial biogenesis may contribute to CRS. More attention is being paid to the role of Ox-HDL in the chronic inflammatory state of obesity. Ox-HDL has been shown to independently upregulate the downstream activity of ang II and has been identified as part of a biomarker profile for early endothelial dysfunction in obese women without identified cardiovascular disease. We have shown oxidized HDL to be part of a biomarker profile for cardiovascular risk in obese women. PGC-1α has been shown to improve organ function by upregulating mitochondrial enzymes, improving mitochondrial function in metabolic syndrome and nonalcoholic fatty liver disease (NAFLD), and reducing Ox-HDL. Most importantly, this improvement in mitochondrial integrity and function will aid in the reprogramming of white fat to beige-like fat
## Lifestyle interventions, weight loss medications, and nutraceuticals
Weight loss with diet and exercise has been successful in short-term supervised programs, with weight loss in excess of 5% body weight. None have approached the magic number of 10-12% of sustained body weight loss in BMIs above 35. The PREDIMED-PLUS study was a 12-month intervention that increased weight loss and improved cardiovascular risk factors. There is confusing evidence as to the effect of aerobic exercise on HDL levels and cholesterol efflux. Exercise has been shown to improve psychological health, particularly in the elderly. Many nutraceuticals (like curcumin, CoQ 10, folic acid, berberine, alpha lipoic acid, astaxanthin, and policosanol) have tried to improve lipid and plasma glucose levels, with positive effects on lipid profiles. Weight-loss drugs have been effective in supervised programs, especially those with close telemedicine follow-up. There is clearly a role for diet, exercise, weight loss medications, and nutraceuticals as preventive measures in obesity, metabolic syndrome, and type 2 diabetes. They are measures meant to prevent cardiorenal syndrome. We need more aggressive measures once we are confronted with cardiorenal syndrome.
## Summary
Mitochondrial function is important for improving the function of the electron transport chain in both the heart and kidneys and converting white adipose tissue to beige, the "browning" of white adipose tissue. This results in an improvement of adiponectin levels and a severe reduction in the release of inflammatory adipocytokines. HO-1 upregulation is the key to improving mitochondrial function in both organs and reducing oxidative stress and Ox-HDL, which are important components of success in treating and preventing cardiorenal syndrome.
## Conflicts of interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
## Abbreviations
|
A BAC end view of the Musa acuminata genome
Background: Musa species contain the fourth most important crop in developing countries. Here, we report the analysis of 6,252 BAC end-sequences, in order to view the sequence composition of the Musa acuminata genome in a cost effective and efficient manner.Results: BAC end sequencing generated 6,252 reads representing 4,420,944 bp, including 2,979 clone pairs with an average read length after cleaning and filtering of 707 bp. All sequences have been submitted to GenBank, with the accession numbers DX451975 -DX458350. The BAC endsequences, were searched against several databases and significant homology was found to mitochondria and chloroplast (2.6%), transposons and repetitive sequences (36%) and proteins (11%). Functional interpretation of the protein matches was carried out by Gene Ontology assignments from matches to Arabidopsis and was shown to cover a broad range of categories. From protein matching regions of Musa BAC end-sequences, it was determined that the GC content of coding regions was 47%. Where protein matches encompassed a start codon, GC content as a function of position (5' to 3') across 129 bp sliding windows generates a "rice-like" gradient. A total of 352 potential SSR markers were discovered. The most abundant simple sequence repeats in four size categories were AT-rich. After filtering mitochondria and chloroplast matches, thousands of BAC end-sequences had a significant BLASTN match to the Oryza sativa and Arabidopsis genome sequence. Of these, a small number of BAC end-sequence pairs were shown to map to neighboring regions of the Oryza sativa genome representing regions of potential microsynteny.Conclusion:Database searches with the BAC end-sequences and ab initio analysis identified those reads likely to contain transposons, repeat sequences, proteins and simple sequence repeats. Approximately 600 BAC end-sequences contained protein sequences that were not found in the existing available Musa expressed sequence tags, repeat or transposon databases. In addition, gene statistics, GC content and profile could also be estimated based on the region matching the top protein hit. A small number of BAC end pair sequences can be mapped to neighboring regions of the Oryza sativa representing regions of potential microsynteny. These results suggest that a largescale BAC end sequencing strategy has the potential to anchor a small proportion of the genome of Musa acuminata to the genomes of Oryza sativa and possibly Arabidopsis.
# Background
Until novel technologies that will enable extremely lowcost genomic DNA sequencing are developed, funding bodies are very selective when choosing new plant genomes to sequence. Current technologies are only able to produce the sequence of a mammalian-sized genome of the desired data quality for $10 to $50 million or more. The initial goal of many genome projects is often to gain a glimpse of the genome of interest at a low cost and in an effective manner. In plants there is often some advantage in leveraging the finished genomes of Arabidopsis thaliana and Oryza sativa through comparative genomics. A. thaliana was chosen as model for the dicotyledons due to its small genome size (125 Mb) [bib_ref] International Rice Genome Sequencing Project: The map based sequence of the rice..., Meinke [/bib_ref] and rice [2] (O. sativa) was the first cereal and monocot to be sequenced [bib_ref] BGI-RIS, An integrated information resource and comparative analysis workbench for rice genomics, Zhao [/bib_ref].
Musa species (bananas and plantains) comprise very important crops in sub-Saharan Africa, South and Central America and much of Asia. The Musa species Musa acuminata (AA genome) and Musa balbisiana, (BB genome), both with 2n = 22 chromosomes represent the two main progenitors of cultivated banana varieties. The haploid genome of Musa species was estimated as varying between 560 to 800 Mb in size [bib_ref] Flow cytometric analysis of nuclear DNA content in Musa, Lysak [/bib_ref] [bib_ref] Nuclear DNA content and base composition in 28 taxa of Musa, Kamate [/bib_ref] [bib_ref] Dolezel : Nuclear genome size and genomic distribution of ribosomal DNA in..., Bartos [/bib_ref] , over four times larger than that of the model plant A. thaliana (125 Mb)and over 30% larger than that of O. sativa (390 Mb) .
Comparative genomics in the monocots have focused on the extent of synteny between closely-related species of monocots belonging to the family of Poaceae [bib_ref] Sequence analysis of the long arm of rice chromosome 11 for ricewheat..., Singh [/bib_ref]. Extensive micro and macro synteny has been shown between O. sativa, barley, maize and wheat [bib_ref] Rapid genome evolution revealed by comparative sequence analysis of orthologous regions from..., Gu [/bib_ref] [bib_ref] New in silico insight into the synteny between rice (Oryza sativa L.)..., Salse [/bib_ref] and the degree of conservation often varies between different chromosomal locations. Synteny between distantly related plants is more bioinformatically challenging to elucidate and probably occurs less frequently.
In order to understand the sequence content and sequence complexity of the Musa genome, it is necessary to sequence a large number of randomly selected clones that are representative of the entire genome. An alternative approach is to end-sequence a large number of Bacterial Artificial Chromosomes (BACs) randomly selected from a BAC library [bib_ref] Analysis of papaya BAC end sequences reveals first insights into the organization..., Lai [/bib_ref]. This latter approach does not provide a truly random sampling of the genome since regions in which the restriction site for the particular enzyme used for library construction is under-represented will also be under-represented. Nevertheless, BAC end sequencing does provide a quasi-random sampling of the genome and carries with it the advantage that BAC clones that appear to contain targets of interest provide excellent material for other analyses such as fluorescent in situ hybridization (FISH) to metaphase or pachytene chromosomes or in depth sequencing for gene discovery. A large collection of BAC end-sequences (BES) is also an essential component of a genome sequencing project. Here, we examined whether Musa BES can lead to insights into the Musa genome composition using bioinformatic comparisons to protein, repeat, expressed sequence tags (ESTs) and other databases. From the BES, we investigate the Musa gene density, GC content, protein and SSR content and putative comparative-tile BACs that represents potential regions of microsynteny between the O. sativa and Musa species.
# Results and discussion
Sequence searches, simple sequence repeats, GC profiling and protein discovery will be discussed first, followed by an analysis of genome mapping to O.sativa and A. thaliana to identify comparative tile BACs from the Musa library that will be likely collinear (i.e. showed microsynteny).
## Bac end sequencing
End sequencing of BACs from a HindIII BAC library constructed from leaves of the wild diploid 'Calcutta 4' clone [bib_ref] Construction and characterization of a bacterial artificial chromosome library of banana (Musa..., Vilarinhos [/bib_ref] , generated 6,252 high quality reads with an average length of 707 nucleotides, giving a total length of ~ 4.4 Mb that included 2,979 paired end reads [fig_ref] Table 1: Sequence statistics of the Musa BES [/fig_ref]. All sequences have been submitted to GenBank, with the accession numbers DX451975 -DX458350.
## Database sequence searches
Comparison of the BES with the TIGR non-identical amino acid database revealed that 11% of the sequences contained "genic" regions by virtue of good matches, excluding transposons/repeats (36%). Using a stringent threshold of 1e-5, 80% identity and 80% coverage resulted in 2.6% BES matches to chloroplast/mitochondria [fig_ref] Table 2: Sequence similarity search resultsNumber of Musa BES containing hits to The TIGR... [/fig_ref]. Of the protein matches, the top BLAST match in over 50% of cases was to O.sativa and in 30% to A. thaliana proteins, consistent with the closer relatedness between Musa and O. sativa when compared to Musa and A. thaliana. This is also consistent with matches to the TIGR Plant Gene Indices where the highest level of homology was shown to O. sativa followed by barley, wheat and other monocots . Of the BES analysed, 36% were found to contain sequences homologous to transposable elements or repeats. The majority of transposable elements belonged to the Ty1 copia type (742) followed by the Ty3 gypsy (211) types of retrotransposons [fig_ref] Table 2: Sequence similarity search resultsNumber of Musa BES containing hits to The TIGR... [/fig_ref] consistent with previous data that class I retrotransposons contributing to most of the nucleotide [bib_ref] The paleontology of intergene retrotransposons of maize, Sanmiguel [/bib_ref] and from studies using papaya BAC end sequences. We also found 111 matches to miniature inverted repeat transposable elements (MITEs), the most abundant being adh-11-like (46), followed by adh type D-like [bib_ref] Main D: Frequency, type, distribution and annotation of simple sequence repeats in..., Jung [/bib_ref] and adh type G-like [bib_ref] Construction and characterization of a bacterial artificial chromosome library of banana (Musa..., Vilarinhos [/bib_ref]. Gene density predictions calculated from the number BES with protein matches (686) at E = 1e-15 estimates the presence of a gene every 6.4 kb [fig_ref] Table 3: Summary of transposon content [/fig_ref] which is consistent with previous gene density studies from one Musa BAC studied [bib_ref] Gene content and density in banana (Musa acuminata) as revealed by genomic..., Aert [/bib_ref]. In contrast, a second BAC from the same study gave a gene density of a gene in every 10 kb, however upon closer examination one half of the BAC consisted of transposon related genes while the other half was non-transposon related. The discrepancy between the data suggests that the gene organization resembling Gramineae where genes are clustered in generich regions separated by gene-poor DNA containing abundant transposons. In comparison with other plant genomes, gene density appears to be similar to reports for the automatic annotation for O. sativa of 6.2 kb per gene [bib_ref] The Institute for Genomic Research Osa1 rice genome annotation database, Yuan [/bib_ref] and different from A. thaliana with 4.5 kb per gene [bib_ref] Dolezel : Nuclear genome size and genomic distribution of ribosomal DNA in..., Bartos [/bib_ref].
## Functional annotation
Gene Ontology (GO) is a controlled vocabulary of functional terms that allows consistent annotation of gene products. In order to assign putative functional roles to the Musa acuminata sequences, we used the GO assignments of the A. thaliana proteome. Among the 686 BES that did not contain a match to the repeat or transposon databases but contained a match the TIGR comprehensive protein database, 664 had matches to A. thaliana proteins and were given GO assignments based on the top matches. The genes are shown to cover a broad range of GO categories .
## Gc profile
GC profiling was performed on the matching region between the BES and the top protein hit. Any BES not containing a match from the start codon was excluded. In par- sativa and maize showed a higher GC value at the 5' end within the first 150 bp from the predicted start site, which gradually decreased towards the 3' end. This result is consistent from previous reports where it has been shown that Gene Ontology assignments for Musa BES Gene Ontology assignments for Musa BES.
Mean GC content as a function of position (5' to 3') across 129 bp sliding windows Mean GC content as a function of position (5' to 3') across 129 bp sliding windows.
## Sliding window position, bp from atg
grasses have high mean GC content and asymmetrical distributions, while the eudicots have lower GC content and more symmetrical distributions [bib_ref] A unique set of 11,008 onion expressed sequence tags reveals expressed sequence..., Kuhl [/bib_ref] [bib_ref] Compositional gradients in Gramineae genes, Wong [/bib_ref]
## Gc content
The GC content for organisms varies between the genomic, intron and exon regions and can be as low as 22% (Plasmodium falciparum) to more than 70% (Zea mays). GC content was determined on the matching region between the BES and the top protein hit. The mean GC content of all BES was 39% and coding sequence GC content was 47% consistent with previous studies which was shown to have an overall GC content to be 38% and within exons to be 49% based on 2 BACs [bib_ref] Gene content and density in banana (Musa acuminata) as revealed by genomic..., Aert [/bib_ref]. This and the previous section have shown that BES with protein matches can allow GC content and GC profiling to be calculated with some degree of accuracy. Further confirmation using a larger dataset was carried out using ESTs,-2,280 Musa ESTs [bib_ref] Analysis of expressed sequence tags from Musa acuminata ssp burmannicoides, var. Calcutta..., Santos [/bib_ref] was downloaded from GenBank, clustered and assembled to give 1,123 unique sequences of which 179 were contigs. The unique sequences generated 1,056 potential open reading frames containing an average GC content of 51%. These results are consistent with previous studies on GC content within monocots and dicots [bib_ref] A unique set of 11,008 onion expressed sequence tags reveals expressed sequence..., Kuhl [/bib_ref].
## Simple sequence repeats
Simple sequence repeats (or microsatellites) are a class of molecular markers that are often polymorphic and are widely used for generating genetic maps [bib_ref] Exploiting EST databases for the development and characterization of genederived SSR-markers in..., Thiel [/bib_ref]. A total of 352 potential SSR markers were discovered within the BAC end-sequences [fig_ref] Table 4: Distribution of SSRs [/fig_ref]. The most abundant SSRs in all four size categories were AT-rich. This is in agreement with previous reports of microsatellite abundance in other species: poly(AT)/(TA) and AT-rich trinucleotide repeats were the most abundant repeats of their class in A. thaliana and in yeast [bib_ref] Differential distribution of simple sequence repeats in eukaryotic genome sequences, Katti [/bib_ref]. Similar to observations for Rosaceae ESTs [bib_ref] Main D: Frequency, type, distribution and annotation of simple sequence repeats in..., Jung [/bib_ref] , dinucleotide repeats represent the most abundant of the four microsatellite classes. None of the SSRs present in this study has been reported previously and no matches were found with previous identified Musa SSRs [bib_ref] Isolation and characterization of microsatellite loci from a commercial cultivar of Musa..., Creste [/bib_ref] [bib_ref] Diploid ancestors of triploid export banana cultivars: molecular identification of 2n restitution..., Raboin [/bib_ref].
## Musa bac end tiling on the o. sativa and a. thaliana genome
For a relatively uncharacterized species where there may be synteny with some chromosomal regions of well sequenced model species, high throughput BAC end sequencing offers the potential to 'tile' the genome of the uncharacterized species onto to that of the sequenced species. BES mapping to O. sativa and A. thaliana were carried out in order to further characterize our BAC library and to test whether a BAC end sequencing approach might be effective for Musa in the manner described above. When Musa BACs that fulfil the criteria of having top blast hits to the same chromosome and having no homology to mitochondria and chloroplast were deemed candidate putative comparative-tile-BACs, and potentially represent regions of highly conserved gene content and organization. The predicted size of the Musa BACs (and thus the distance between the end-sequences) was compared to the span by which the paired matches are separated in the O. sativa and A. thaliana genomes respectively. Separations in the Musa BES matches that exceeded our arbitrary cut off of 500 Kb, may represent expansions of the syntenic regions and due to rearrangements during the evolution of the two genomes.
# Conclusion
In this study, 2 major ideas were examined. Firstly, that Musa BES can lead to insights into the Musa genome with specific reference to gene density, GC content, protein and SSR discovery; and secondly, that the sequences can be used to identify regions of potential microsynteny between Musa and other species. The BAC end-sequences were shown to contain homology to proteins, expressed sequence tags, transposons, repeat sequences and to be useful for simple sequence repeat identification and estimation of gene statistics and GC content. Proteins encoded in these BES were shown to cover a broad range of GO categories. Although there is only limited microsynteny between Musa and O. sativa, the results suggest that a large-scale BAC end sequencing strategy has the potential to anchor at least a small portion of the genome of Musa onto that of the sequence of the O. sativa. Largescale BAC end sequencing would show whether there are more regions of microsynteny between the reference genome and the genome of interest and if there was support for whole genome sequencing due to unique gene features and genome characteristics. BAC end data would be one useful indicator along with existing EST or genomic sequences for funding bodies to use when selecting new plant genomes to sequence and assess the potential of leveraging the finished genomes of A. thaliana and O. sativa through comparative genomics. We expect that a similar analysis using other plant or animal species would provide insights into the genome in a very cost effective and efficient manner through database searches and synteny to model species.
# Methods
## Bac end sequencing
The BES were generated from a Musa bacterial artificial chromosome (BAC) library constructed from leaves of the wild diploid 'Calcutta 4' clone (Musa acuminata subsp. Burmannicoides 2n = 2 × = 22) with an average insert size of 100 kb [bib_ref] Construction and characterization of a bacterial artificial chromosome library of banana (Musa..., Vilarinhos [/bib_ref].
DNA template was prepared in 384-well format by a standard alkaline lysis method. End sequencing was performed using Applied Biosystems (ABI) Big Dye terminator chemistry and analyzed on ABI 3730 xl machines. Base calling was performed using TraceTuner and sequences were trimmed for vector and low quality sequences using Lucy [bib_ref] DNA sequence quality trimming and vector removal, Chou [/bib_ref].
## Bac end database searches
Sequences were compared to all entries in the TIGR Plant Gene Indices [bib_ref] The TIGR gene indices: reconstruction and representation of expressed gene sequences, Quackenbush [/bib_ref] using blat and to the TIGR non-identical amino acids database that contains non-identical protein data from a number of databases including GenBank, RefSeq and Uniprot using blastx (cut-off value 1e-5). The BAC end-sequences were also compared with repetitive sequences in the TIGR Repeat Database [bib_ref] The TIGR Plant Repeat Databases: a collective resource for the identification of..., Ouyang [/bib_ref] and an inhouse transposon database using blastx with a cut-off value of 1e-5. The BAC end-sequences were compared with the TIGR rice genome sequence assembly and the A. thaliana genome sequence from TAIR using blastn with a cut-off value of 1e-10. To identify comparative tile BACs from the Musa library that were likely collinear (i.e. showed microsynteny) with the reference genomes, the searches against the Musa genomic sequence were parsed for the top pair of BES for which both ends had the highest significant match to a stretch of O. sativa or A. thaliana sequence and where the two regions on the Musa genome were between 100 kb and 500 Kb apart. The BAC end data sets for O. sativa, A. thaliana, maize and M. truncatula used for GC profiling was originally downloaded from Gen-Bank and then the vector trimmed and cleaned sequences were downloaded from estinformatics.org .
[table] Table 1: Sequence statistics of the Musa BES [/table]
[table] Table 3: Summary of transposon content [/table]
[table] Table 2: Sequence similarity search resultsNumber of Musa BES containing hits to The TIGR Plant Gene Indices using blatFigure 1Number of Musa BES containing hits to The TIGR Plant Gene Indices using blat. [/table]
[table] Table 4: Distribution of SSRs [/table]
[table] Table 5: Musa BAC end tiling on the O. sativa genome [/table]
|
Effect of Vitamin D on Blood Pressure and Hypertension in the General Population: An Update Meta-Analysis of Cohort Studies and Randomized Controlled Trials
What is already known on this topic?The effects of vitamin D on hypertension risk and blood pressure have been explored widely in cohort studies and randomized controlled trials (RCTs), but whether the association is causal still is unknown.What is added by this report?We performed an update meta-analysis of both cohort studies and RCTs in a generally heathy population and found that the dose-response relationship between circulating 25-hydroxyvitamin D level and hypertension risk was approximately L-shaped. However, pooled results of RCTs showed that there was still no significant reduction in systolic and diastolic blood pressure.What are the implications for public health practice?Vitamin D supplementation is ineffective to prevent hypertension.AbstractBackgroundThe effect of vitamin D supplementation on blood pressure has been explored in previous meta-analyses, but whether the association is causal in the general population is still unknown. We evaluated the association comprehensively and quantitatively.MethodsWe searched PubMed and Embase for relevant cohort studies and randomized controlled trials (RCTs). We used a 2-step generalized least-squares method to assess the dose-response association of circulating 25-hydroxyvitamin D (25[OH]D) and hypertension and a fixed-effects model to pool the weighted mean differences (WMDs) and corresponding 95% confidence intervals (95% CIs) of blood pressure across RCTs.ResultsWe identified 11 cohort studies and 27 RCTs, with 43,320 and 3,810 participants, respectively. The dose-response relationship between circulating 25(OH)D levels and hypertension risk was approximately L-shaped (P nonlinearity = .04), suggesting that the risk of hypertension increased substantially below 75 nmol/L as 25(OH)D decreased, but it remained significant over the range of 75-130 nmol/L. However, pooled results of RCTs showed that there was no significant reduction in systolic blood pressure (WMD, −0.00 mm Hg; 95% CI, −0.71 to 0.71) or diastolic blood pressure (WMD, 0.19 mm Hg; 95% CI, −0.29 to 0.67) after vitamin D intervention.ConclusionsThe results of this meta-analysis indicate that supplementation with vitamin D does not lower blood pressure in the general population. RCTs with long-term interventions and a sufficient number of participants who have low levels of vitamin D are needed to validate these findings. DL, et al. Effect of vitamin D supplementation on blood pressure in blacks. Hypertension 2013;61(4):779-85. 34. Skaaby T, Husemoen LL, Pisinger C, Jørgensen T, Thuesen BH, Fenger M, et al. Vitamin D status and changes in cardiovascular risk factors: a prospective study of a general population. Cardiology 2012;123(1):62-70. 35. van Ballegooijen AJ, Kestenbaum B, Sachs MC, de Boer IH, Siscovick DS, Hoofnagle AN, et al. Association of 25hydroxyvitamin D and parathyroid hormone with incident hypertension: MESA (Multi-Ethnic Study of Atherosclerosis). J Am Coll Cardiol 2014;63(12):1214-22. 36. Margolis KL, Martin LW, Ray RM, Kerby TJ, Allison MA, Curb JD, et al.; Women's Health Initiative Investigators. A prospective study of serum 25-hydroxyvitamin D levels, blood pressure, and incident hypertension in postmenopausal women. Am J Epidemiol 2012;175(1):22-32. 37. Ke L, Graubard BI, Albanes D, Fraser DR, Weinstein SJ, Virtamo J, et al. Hypertension, pulse, and other cardiovascular risk factors and vitamin D status in Finnish men. Am J Hypertens 2013;26(8):951-6.
# Introduction
Emerging evidence suggests that vitamin D deficiency is a widespread global problem. According to the Institute of Medicine (IOM), vitamin D deficiency is defined as circulating 25-hy-droxyvitamin D (25[OH]D) level <50 nmol/L based on the optimal concentration for skeletal health. Interest has increased concerning the potential health consequences of vitamin D deficiency, such as increased risk of cardiovascular diseases, cancers, and Alzheimer's disease . Although observational data have demonstrated that poor vitamin D status is associated with increased risk of hypertension (6-9), randomized controlled trials (RCTs) have provided little support for the beneficial effect of vitamin D supplementation on blood pressure (10-13). Considering the potential residual confounding, inferring causality or reversibility of this relationship and reaching consensus from these findings is difficult.
Several meta-analyses of observational studies and RCTs have been published, but results are conflicting . Golzarand et al evaluated 30 RCTs with 4,744 participants and concluded that vitamin D has a beneficial effect in subgroups of daily doses >800 IU/d, a duration less than 6 months, or older subjects. Kunutsor et al suggested that supplementation with vitamin D significantly reduced diastolic blood pressure (DBP) by 1.31 mm Hg in participants with preexisting cardiometabolic conditions. However, another meta-analysis performed by incorporating individual data supported that vitamin D supplementation is ineffective in lowering blood pressure.
Taken together, it may be hypothesized that the increased blood pressure or risk of hypertension is partly explained by individuals' baseline vitamin D status, the sample size, the intervention dose, and the follow-up duration. Meanwhile, considering that pre-existing conditions such as diabetes, cardiovascular disease, and kidney disease may influence the physiologic mechanism of vitamin D on blood pressure, considerable variability may exist between individual patients and the general population. Therefore, restricting the participants to the general population may help to explore the true association hidden by the confounders. Analyzing the population as a whole rather than restricting analyses to certain population subgroups may help us to explore the true association hidden by confounders. In addition, results from at least 10 more studies including 1,716 participants have been published on this topic since the latest meta-analysis in 2015 (10-12,18-24).
We aimed to provide a comprehensive and quantitative meta-analysis from the published cohort studies and RCTs on the effect of vitamin D involving hypertension risk and blood pressure levels in the general population.
# Methods
We used the PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analyses) checklist to perform the meta-analysis and report the results (25).
## Data source and searches
We searched PubMed and Embase databases up to June 12, 2019, for cohort studies reporting an association between blood 25(OH)D levels and risk of incident hypertension and for RCTs examining the effect of vitamin D supplementation (alone or in combination with other nutrients) on blood pressure. The search terms "vitamin D" and "blood pressure" were used in combination to retrieve relevant records. The records were restricted to human studies, and additional studies were retrieved through manually searching the references of identified articles and relevant systematic reviews.
## Study selection
Two investigators (D.Z. and C.C.) reviewed the titles and abstracts independently to identify articles for potentially relevant sources. Full-text versions were requested to evaluate eligibility. To be included, the study had to meet the following criteria: 1) followed an RCT or a cohort study design; 2) investigated the association between vitamin D and risk of hypertension or effect of blood pressure levels; 3) included a general population (≥18 y) rather than patients with specific diseases (eg, diabetes, hypertension, stroke, heart failure); and 4) provided estimates of the risks of hypertension in at least 3 categories of blood 25(OH)D levels or reported continuous risk estimates for the dose-response analysis, or reported blood pressure for meta-analysis of RCTs. We excluded articles if they 1) measured other metabolites of vitamin D (eg, 1,25-dihydroxyvitamin D); 2) focused on pregnant women or groups with specific diseases; or 3) did not report blood pressure at baseline/end or the changes after invention from baseline for trials. Inconsistencies were resolved through group discussion or adjudicated by a third reviewer.
## Data extraction
Using predefined protocols, D.Z. extracted data from each study and C.C. checked the accuracy. For cohort studies, the following information was abstracted: first author, publication year, country, follow-up period, sample size, age, number of cases/participants, categories of 25(OH)D levels, reported risk estimates, 95% confidence intervals (CIs), and covariates adjusted for in the analyses. When several adjusted models were explored, we extracted the risk ratios from the model with largest number of covariables. If the lowest 25(OH)D level was not the reference, we recalculated the risk estimates by the method of Hamling et al (26). When the mean or median 25(OH)D level per category was not reported, we assigned the value as the midpoint of the lower and upper bound in each category (27). If the category was open-ended, we assumed the width of interval to be the same as in the adjacent category (27). If studies reported 25(OH)D levels in ng/mL, we converted the values to nmol/L by multiplying by 2.5.
For RCTs, we recorded the following data: study design (sample size of each group, blinding methods, intervention/placebo type and amount, duration of intervention, type of vitamin D, and intervention frequency); characteristics of participants (age, sex, baseline circulating 25[OH]D levels); and baseline/end blood pressure in both intervention and placebo groups and/or blood pressure changes from baseline. If studies used different doses of vitamin D, we extracted only the highest dose in the analysis. If studies measured blood pressures repetitively at different intervals during the intervention, we included only the blood pressure values at the longest follow-up point. Attempts were made to contact corresponding authors for unavailable information.
## Risk for bias assessment
We used the 9-star Newcastle-Ottawa Scale to evaluate the quality of individual cohort studies; the scale is based on 8 aspects covering selection, comparability, and outcome domains (28). Meanwhile, we assessed the risk of bias for each trial using 7 fields from The Cochrane Collaboration's tool: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other bias (29). Summary assessments for trials were assigned as "high," "low," or "unclear," according to the risk bias in each outcome. Disagreements were resolved through group discussion. Publication bias was assessed with Egger's test (30).
## Data synthesis and analysis
To provide dose-response evidence from all cohort studies, we used the 2-step generalized least-squares method (31). Study-specific slope coefficients were examined by restricted cubic splines with three knots at 25%, 50%, and 75% of the distribution of circulating 25(OH)D levels. For the dose-response analyses of 25(OH)D, the reference category was re-scaled to 75 nmol/L, which is the cutoff value between insufficient and sufficient vitamin D status. P values for nonlinearity were calculated by using the Wald χ 2 test, assuming the coefficient of the second spline was zero. We used the DerSimonian and Laird random effects model to estimate the study-specific dose-response risk, and we calculated the pooled risk of hypertension for every 25 nmol/L increment in 25(OH)D levels using a random effects model (32).
We assessed the effect of vitamin D supplementation by the mean blood pressure changes (including systolic blood pressure and DBP) in the intervention group minus the changes in blood pressure in the placebo group. The standard deviations (SDs) were obtained as reported or calculated from 95% CIs, P values for t statistics, or individual standard errors (SE) from intervention and placebo groups. If the studies did not report blood pressure changes from baseline, we calculated the mean values by using blood pressure after intervention minus blood pressure at baseline, and the SD of changes was obtained according the following formula, described in the Cochrane Handbook for Systematic Reviews of Interventions (29):
We estimated correlation by calculations from 2 studies that provided complete data for SD baseline , SD final , SD change in both intervention and placebo groups (33,34). Between-study heterogeneity was assessed with the I 2 and Q statistics. We used fixed-effects models and forest plots to pool the weighted mean differences (WMDs) and corresponding 95% CIs of blood pressure across studies.
Predefined subgroup analyses were performed to explore potential effect modification and sources of heterogeneity. We also conducted sensitivity analyses by removing one study at a time to ensure that the pooled result was not simply dependent on one large or individual case. All statistics were analyzed using Stata, version 12.1 (StataCorp, LLC). Significance was set at P < .05.
# Results
## Descriptive study characteristics
The systematic search in PubMed and Embase retrieved 8,956 publications, and 3 more were identified by manual searching. After duplicate checking and initial review of the titles and abstracts, 156 potentially relevant articles were obtained in full text for further evaluation. Finally, 119 articles were excluded and 37 publications (including 11 cohort studies in 10 publications [6-9,35-40] and 27 trials [10-13,18-24,33,34,41-54]) were eligible for inclusion.
Eleven cohort studies with 8,397 incident cases of hypertension and 43,320 participants were identified from 10 publications. With the exception of 1 study conducted in Asia, most were conducted in Europe (n = 4) and the United States (n = 6). The follow-up durations ranged from 1.3 to 15.3 years (median 5.0 years). Analyses of the quality of studies yielded an average NOS score of 7.5, nine of which were of high quality (score ≥7).
Twenty-seven studies were RCTs with 3,810 participants. Among them, 2 studies included only men, 10 included only women, and 15 included both. Five of the included trials were conducted in Asia, 12 were performed in Europe, 4 were conducted in Oceania, and the remaining 6 were performed in the United States. www.cdc.gov/pcd/issues/2020/19_0307.htm - Centers for Disease Control and Prevention als with vitamin D insufficiency, vitamin D deficiency, or both. Nine trials did not provide the final 25(OH)D concentration in intervention arms, whereas the remaining studies showed a substantial increase in circulating levels of 25(OH)D compared with the baseline assessment. All trials had low risk of bias for random allocation and selective reporting. There was insufficient information about allocation concealment in 5 trials and high risk of bias in 1 trial. One open-label trial had high risk of bias for blinding of participants and personnel and unclear bias risk for blinding of outcome assessment (43).
# Meta-analyses results
## Circulating 25(oh)d levels and hypertension risk
Quantitative results from meta-analyses of cohort studies showed that the risk of incident hypertension decreased by 7% (relative risk [RR] = 0.93; 95% CI, 0.89-0.98) per 25 nmol/L increment in 25(OH)D levels, with significant heterogeneity (I 2 = 61.6%, Pheterogeneity = .004). Ten studies reporting RR for 25(OH)D exposures in at least 3 levels were eligible for the linear trend estimation. Results from the analysis of restricted cubic splines indicated an approximate L-shaped correlation between circulating 25(OH)D levels and hypertension risk (P nonlinearity = .04,. The risk of hypertension increased substantially below 75 nmol/L as 25(OH)D decreased but remained significant over the range of 75-130 nmol/L. Subgroup analyses indicated sex (male, female, or mixed), followup duration (≤5 y or >5 y), region (America, Europe, or Asia), number of cases (<1,000 or ≥1,000), and study quality (high, me-dium, or low) as the potential sources of the heterogeneity. However, the association of 25(OH)D levels per 25 nmol/L increment showed no significance in subgroups of men (RR = 0.93; 95% CI, 0.85-1.00), women (RR = 0.88; 95% CI, 0.76-1.01), European region (RR = 0.97; 95% CI, 0.94-1.01), small number of cases (RR = 0.95; 95% CI, 0.89-1.02), and medium or low quality of study (RR = 0.91; 95% CI, 0.80-1.03). Furthermore, the pooled estimates could not be altered substantially by removing one study at a time, and we found no evidence of publication bias by Egger's test (P = .38). There was also no significant reduction in DBP after intervention, and the WMD (95% CI) was 0.19 mm Hg (−0.29 to 0.67), without evidence of significant heterogeneity (I 2 = 3.3%, P heterogeneity = .42).shows the subgroup analyses of summary WMDs in SBP and DBP. We found that the heterogeneity decreased in studies of men, studies with overweight or obese individuals, studies with a large sample size (≥200), and studies with an intervention duration of 6 months or longer. The effects of vitamin D supplementation on SBP and DBP was still insignificant in all subgroups. In sensitivity analyses, the summary results remained similar by removing one study at a time. According to Egger's test, we found no evidence of publication bias in studies of SBP (P = .60) and DBP (P = .07).
## Vitamin d supplementation and blood pressure levels
## Preventing chronic disease
# Discussion
This meta-analysis of cohort studies suggested an inverse association between 25(OH)D levels and incident hypertension, with hypertension risk reduced by 7% per 25 nmol/L increment in 25(OH)D levels. Meanwhile, summary data of RCTs indicated no evidence of blood pressure reduction by supplementation with vit-PREVENTING CHRONIC DISEASE The findings from numerous observational studies have shown that sufficient vitamin D status is a protective factor for hypertension. Analysis of Mendelian randomization also provided the causal evidence for the effect of increased circulating 25(OH)D levels on reduced blood pressure levels and risk of hypertension. However, our subgroup analyses of the cohort studies produced inconsistent results, which indicated that the quantitative data failed to provide convincing evidence of the protective effect of vitamin D on hypertension. Meanwhile, most of the interventional studies did not provide consistent evidence of blood pressure benefit from supplementing with vitamin D . Given these findings, we speculate that the beneficial effect observed in cohort studies may be partly explained by the tendency that sufficient vitamin D levels are closely related to healthy lifestyle or study participants being young. It may be also in part because of the hypothesis that low 25(OH)D levels could be the result of sub-health status rather than a precursor of diseases. Furthermore, differences exist among the various methods used (ie, liquid chromatography-mass spectrometry; high-performance liquid chromatography; and enzymoimmunoassay, radioimmunoassay, and chemiluminescence immunoassays) and in the laboratories that measured 25(OH)D levels, which would also influence the accuracy of the study results.
Similar with our results, previous meta-analyses also showed no overall lowering effect of vitamin D supplementation on blood pressure. However, they suggested that vitamin D may show a beneficial effect on blood pressure in specific subgroups, such as older people, people whose dosage of vitamin D was high (>800 IU/d), short-term interventions (<6 months), or individuals with pre-existing cardiometabolic disease. A possible reason for this discrepancy is that the recruited populations of included studies had high heterogeneity. Therefore, we restricted this meta-analysis to analyses of apparently healthy individuals. We excluded trials that have targeted patients with hypertension, diabetes, cardiovascular disease, or other diseases, because the known or unknown interaction between vitamin D and antihypertensive or cardiovascular medications may mask or attenuate the small effects of blood pressure reduction.
Complicated factors such as baseline vitamin D status, intervention design, or adiposity may modify or blunt the beneficial effect on blood pressure of improving vitamin D levels. An increasing body of evidence supports the presence of thresholds in vitamin D status. Similarly, the approximately L-shaped relationship between 25(OH)D levels and hypertension risk in our meta-analysis showed that hypertension risk increased substantially below 75 nmol/L but remained marginally significant above 75 nmol/L, which suggests that subjects with vitamin D insufficiency or deficiency show higher response to supplementation. In addition, evidence showed a therapeutic effect of cholecalciferol only in vitamin D-depleted participants by decreasing their 24-hour blood pressure by 3-4 mm Hg. Therefore, we speculated that the protective effect would only appear in subjects with low vitamin D levels. Indeed, we classified the studies according to their baseline vitamin D status, but the results indicated that vitamin D supplementation had no apparent effect on blood pressure, regardless of its baseline status. This finding is in accord with a recent metaanalysis that used individual patient data. However, considering that the number of people with low vitamin D levels may be insufficient in our study, further trials are needed to verify this finding.
Individuals who are taking vitamin D supplements should do so for at least 6 months to reach the maximum attained 25(OH)D level. It is reasonable to assume that the effect of vitamin D is time-dependent. However, our findings from subgroup analyses of RCTs suggested that response of blood pressure to vitamin D is independent of interventional duration (<6 months and ≥6 months). Similar findings have been reported . Considering these findings, we still cannot rule out that the duration of vitamin D intervention is insufficient to detect any slight but significant reduction in blood pressures, especially in the apparently healthy subjects whose normal values are less likely to be further improved. It is worth noting that until June 2019 only one RCT lasting up to 2 years was included in our study; therefore, a protective effect of longer intervention could not be studied adequately. Future RCTs with longer follow-up duration are needed to provide in-depth insight into the long-term benefits of vitamin D supplementation.
The optimal dose for vitamin D supplementation would influence the effect on blood pressure. A 4-arm trial conducted in African Americans reported dose-dependent reductions in SBP after 3 months of cholecalciferol supplementation with 1,000 IU, 2,000 IU, and 4,000 IU per day (0.66 mm Hg, 3.4 mm Hg, and 4.0 mm Hg, respectively) (34). In addition, a meta-analysis synthesizing the results of 30 RCTs suggested that vitamin D supplementation at a dose of >800 IU/d reduced blood pressures significantly. Contrary to these results, we did not find the dose-response relationship for vitamin D on blood pressure. We should consider the possibility that the supplementary doses in most included trials may be larger or smaller to observe a beneficial effect. Further studies are needed to explore the potential quantitative model. This meta-analysis of RCTs included 3,810 people from the general population, which provides a substantial statistical power to detect the potential effects and thereby enhances the generalizability of our findings. However, our study also contains several poten- tial limitations. First, because most studies did not record the changes of diet, sun exposure or latitudes, genetic factors, and educational status, we are not able to answer the questions of whether these factors would modify the effect of the intervention. Second, there are several trials that did not reach enough power (they were below 80%) to detect any weak difference between interventional and placebo groups because of the small sample size and high rate of noncompliance . In addition, although we stratified the duration of follow-up (the maximum is 2.0 years) and found no significant difference between subgroups, it remains unclear whether there are any long-term (>2 years) effects of vitamin D to improve blood pressure levels. However, we may conclude that vitamin D supplementation will not affect blood pressure short-term.
## Preventing chronic disease
The results of this meta-analysis indicate that supplementation with vitamin D does not lower blood pressure in the general population. On the basis of this finding, we do not recommend using vitamin D supplementation to prevent hypertension. However, future RCTs with long-term interventions and sufficient sample sizes of people with low vitamin D levels are needed to replicate this finding.
## Preventing chronic disease
|
Reversible platypnea-orthodeoxia syndrome induced by rapidly progressive interstitial pneumonia in a patient with polymyositis
We report a case of platypnea-orthodeoxia that developed in a 64-year-old Japanese woman during an episode of rapidly progressive interstitial pneumonia with polymyositis. Pulmonary infiltrates were predominant in the bilateral lower lobes. The patient was treated successfully with early administration of immunosuppressive therapies and polymyxin B-immobilized fiber column-direct hemoperfusion, and her platypnea-orthodeoxia improved with resolution of the underlying parenchymal lung disease. Reports of platypnea-orthodeoxia syndrome with interstitial pneumonia are extremely rare. The recognition that platypnea-orthodeoxia syndrome may occur in multiple disease states, including interstitial pneumonia, is crucial to the understanding of this perplexing disorder.
# Introduction
Platypnea and orthodeoxia are terms used to describe the rare phenomenon of dyspnea and hypoxemia accentuated by an upright posture and relieved by recumbency [bib_ref] Platypnea related to orthodeoxia caused by true vascular lung shunts, Robin [/bib_ref]. We report these features in a patient with rapidly progressive interstitial pneumonia with polymyositis.
## Case report
The patient was a 64-year-old woman who complained of fever and dyspnea and was diagnosed as having pneumonia at another hospital. The patient deteriorated with administration of antibiotics over 10 days. Afterwards, she experienced rapidly progressive extension of pulmonary infiltration and hypoxemia, and she was transferred to our hospital. Chest X-ray and computed tomography scan revealed consolidations predominantly in the lower lobes. Physical examination revealed femoral muscle atrophy and weakness, and auscultation of the chest identified audible fine crackles. Elevated levels of C-reactive protein of 13.9 mg/dL, KL-6 of 1069 U/ mL, and surfactant protein-D of 355 ng/mL were also found. Moreover, anti-threonyl-tRNA synthetase (anti-PL-7) antibody in aminoacyl-tRNA synthetase autoantibody was positive, and electromyography showed abnormalities in the electrical activity of muscles as suspected with inflammation of the muscles. We then diagnosed her as having rapidly progressive interstitial pneumonia with polymyositis. She received noninvasive positive pressure ventilation for severe hypoxemia on admission. Her respiratory distress was refractory to steroid pulse therapy and cyclosporine, and intubation and mechanical ventilation were necessary on the 8th hospital day. Additional therapy, which included intravenous cyclophosphamide pulse (IVCY) therapy and polymyxin B-immobilized fiber column-direct hemoperfusion (PMX-DHP), was administered, and her condition improved slightly. Mechanical ventilation was withdrawn on the 32nd hospital day .
On the 6th hospital day, we noticed that she experienced dyspnea and hypoxemia when in a sitting position. We suspected platypnea-orthodeoxia syndrome (POS) because there was no evidence of liver disease, hepatopulmonary syndrome, or pulmonary arteriovenous malformations. On the 54th hospital day, transesophageal echocardiography was performed in the prone and sitting positions. Microbubble opacification of the left atrium following intravenous administration of normal saline was not detected after either two or six cycles, thus excluding the presence of intracardiac or intrapulmonary shunts [fig_ref] Figure 2: Image from the transesophageal echocardiographic study conducted on the 54th hospital day [/fig_ref]. Pulmonary perfusion imaging with 99m Tc-macroaggregated albumin was also performed, and the result did not suggest the presence of a right-to-left shunt. Radiological improve-ment was observed on the 73rd hospital day , and she experienced no hypoxemia in the sitting position [fig_ref] Table 1: Serial arterial blood gas results [/fig_ref]. Therefore, we diagnosed POS, which resulted from the pulmonary parenchymal ventilation/perfusion mismatch caused by the interstitial pneumonia.
# Discussion
We describe a case of polymyositis complicated by rapidly progressive interstitial pneumonia. The patient was treated with early administration of immunosuppressive therapies including steroid pulse therapy and cyclosporine, which resulted in deterioration on radiological imaging and respiratory failure that required mechanical ventilation. Therefore, although we did not investigate the lung tissue pathologically in our patient, we thought that her . Clinical course as depicted by radiological imaging. Chest X-ray and computed tomographic images showed bilateral consolidations predominantly in the lower lobes at initial presentation (day 1). Deterioration was observed radiologically on hospital day 8, and intubation and mechanical ventilation were performed. Afterwards, the lesions had improved slightly by day 19, and mechanical ventilation was withdrawal on hospital day 32. Radiological improvement with no hypoxemia in the sitting position was observed on hospital day 73. respiratory failure showed the presence of a diffuse alveolar damage pattern. Recent reports have suggested that PMX-DHP might improve oxygenation in patients with acute lung injury/acute respiratory distress syndrome associated with fulminant interstitial pneumonia or rapidly progressive interstitial pneumonia with polymyositis/ dermatomyositis [bib_ref] Favorable outcome with hemoperfusion of polymyxin B-immobilized fiber column for rapidly progressive..., Ichiyasu [/bib_ref]. Moreover, early intervention with aggressive combination therapy including IVCY and cyclosporine might be life saving as a treatment for rapidly progressive interstitial pneumonia polymyositis/ dermatomyositis [bib_ref] Successful combination therapy with corticosteroids, biweekly intravenous pulse cyclophosphamide and cyclosporin A..., Suzuki [/bib_ref]. Treatment with these immunosuppressive therapies and PMX-DHP was successful in our patient. POS is a rare clinical phenomenon that is associated with normal oxygen saturation in the supine position (platypnea) and arterial oxygen desaturation in an upright position (orthodeoxia). The exact underlying mechanisms leading to this unusual condition are not fully understood but most probably derive from one of the following three mechanisms: intracardiac shunting, pulmonary vascular shunting, or ventilation/perfusion mismatch or a combination of these mechanisms [bib_ref] Platypnea-orthodeoxia syndrome in review: defining a new disease, Rodrigues [/bib_ref].
In our patient, anatomic shunts were sought using both microbubble injection during transesophageal echocardiography and pulmonary perfusion imaging with 99m Tcmacroaggregated albumin, but none could be identified. In addition, the POS in our patient improved spontaneously over a 2-month period along with radiological improvement of the consolidation located predominantly in the bilateral lower lobes. Therefore, we diagnosed POS because of pulmonary parenchymal ventilation/perfusion mismatch caused by the interstitial pneumonia. Parenchymal lung diseases, such as emphysema, can also be present occasionally with POS [bib_ref] Platypnea-orthodeoxia syndrome in review: defining a new disease, Rodrigues [/bib_ref]. When the patient stands upright, right ventricular preload diminishes, resulting in decreased output to the pulmonary arteries such that alveolar pressures exceed arterial and venous pressures. Moreover, gravity promotes preferential blood flow to the basal areas of the lung, and apical areas tend to act as dead space, which contributes to an increase in ventilation/perfusion mismatch and increased dyspnea [bib_ref] Platypnea-orthodeoxia syndrome in review: defining a new disease, Rodrigues [/bib_ref]. However, reports of cases of POS in patients with interstitial pneumonia are extremely rare [bib_ref] Orthodeoxia: a new finding in interstitial fibrosis, Tenholder [/bib_ref] [bib_ref] Platypnoea-orthodeoxia in cryptogenic fibrosing alveolitis, Bourke [/bib_ref] [bib_ref] Severe hypoxemia with orthodeoxia due to right to left shunt in idiopathic..., Naccache [/bib_ref]. Tenholder et al. suggested that in severe fibrosis, the air spaces may be so far separated from the pulmonary vasculature as to seriously impair diffusion of oxygen [bib_ref] Orthodeoxia: a new finding in interstitial fibrosis, Tenholder [/bib_ref]. The gravitational effect of the upright posture presumably aggravates this problem by increasing blood flow through this physiological shunt. Likely, however, other factors are involved because other patients with apparently similar basal infiltration do not manifest this phenomenon. Moreover, it has been reported that this phenomenon remains unclear for a failure of the hypoxic pulmonary vasoconstriction in interstitial pneumonia [bib_ref] Orthodeoxia: a new finding in interstitial fibrosis, Tenholder [/bib_ref]. The POS in two of these reported cases was not reversible [bib_ref] Orthodeoxia: a new finding in interstitial fibrosis, Tenholder [/bib_ref] [bib_ref] Platypnoea-orthodeoxia in cryptogenic fibrosing alveolitis, Bourke [/bib_ref]. To the best our knowledge, only one case of reversible POS in interstitial pneumonia has been previously reported [bib_ref] Severe hypoxemia with orthodeoxia due to right to left shunt in idiopathic..., Naccache [/bib_ref]. Therefore, further analysis of ventilation/ perfusion mismatch relationships in interstitial pneumonia may help to define the mechanisms for POS.
In conclusion, we describe a case of reversible POS induced by rapidly progressive interstitial pneumonia with polymyositis. POS is a rare and usually underestimated syndrome. Intracardiac shunts and anatomic pulmonary vascular shunts are the most common etiologic associations. However, severe ventilation/perfusion mismatch should also be considered as a probable explanation. The recognition that POS may occur in multiple disease states, including interstitial pneumonia, is crucial to the understanding of this perplexing disorder.
[fig] Figure 2: Image from the transesophageal echocardiographic study conducted on the 54th hospital day. Transesophageal echocardiography was performed in the prone and sitting positions. Microbubble opacification after intravenously administered normal saline was not detected in the left atrium after either two or six cycles. LA, left atrium; RA, right atrium. [/fig]
[table] Table 1: Serial arterial blood gas results.FiO2, fraction of inspired oxygen; PaCO2, partial pressure of carbon dioxide; PaO2, partial pressure of oxygen. [/table]
|
Primary Clear Cell Carcinoma with no Diethylstilbestrol Exposure; Case Series
One of the rare neoplasms of cervix uteri and vagina is clear cell carcinoma; mostly in patients with a positive past medical history of intrauterine diethylstilbestrol (DES) exposure which reveals the importance of other unknown risk factors of gynecologic neoplasms. 2 Asian women referred to the gynecology-oncology department. The first one came with a complaint of prolonged vaginal discharge and the second one with irregular bleeding. Neither did report a history of DES-exposure. On vaginal examination, it was found that both cases had mass which biopsied for histologic study. They were diagnosed with primary Clear Cell Carcinoma of Vagina in the first case and Cervix uteri in the other one. Both patients underwent cytoreductive surgery soon after diagnosis.
## -
The most common reason for clear cell carcinoma in gynecologic organs is exposure to Diethylstilbestrol in the past. Malignancies with clear cell feature manage high-grade neoplasms due to their worse prognosis.
# Introduction
Adenocarcinoma of gynecologic structures is a rare malignant tumor (less than 15% of all cervical malignancies and rarer at vagina). [bib_ref] Clear cell adenocarcinoma of the cervix in second generation young women who..., Tantitamit [/bib_ref] Primary Clear Cell Carcinoma (CCC) is too rare; it accounts for approximately 2-9% of adenocarcinoma.This cancer usually has been reported in women with past medical history of diethylstilbestrol exposure (DES), but there must be some other unknown etiology for that. On histopathologic examination, this tumor shows cells with clear cytoplasm and hobnail cells, and it must be distinguished from other differential diagnoses. Immunohistochemically staining is needed in doubtful cases.Authors had reported some rare cases which could be helpful in suggesting any association between CCC and race, infection, and sexual activity. On the other hand, CCC could also act as the best treatment modality because the best prognosis is not standardized yet. This study aimed to report two rare cases of CCC without the most well-known risk factors and compare the etiology, treatment approach, and prognosis of them with prior literature.
## Case presentation
## Case 1
A 49-year-old nulligravida woman referred to Mashhad University of Medical Sciences, Iran, Hospital, gynecologiconcology department in 2017, Aug. She had a complaint of prolonged purulent watery vaginal discharge and vaginal
## Case series
bleeding after an intercourse that had happened a year before. Since 2012, she had become menopause. She had no significant prior medical history, except the history of anti-depressant therapy (sertraline). She did not report any history of hormonal contraception usage either. She was a passive smoker by her partner. Her history of the first sexual intercourse experience was not reliable. She reported her engagement time to have happened in 2015 and did not respond to sexual field questions anymore. She had a significant low socioeconomic condition. There was no genitalia structural abnormality on her physical examination. Speculum examination showed a 2×2 cm, polypoid shape, mass, arising from the post fornix at the left side of upper vagina not affecting the uterine cervix and parametriom. Per rectal examination, no additional finding was detected. Malignancy diagnosis was the first suggestion for detecting her pathology, but its source as the primary vaginal or metastatic one was the essential question. Pap smear from Exo/Endocervical canal and the lesion sampling were performed. Pelvic Magnetic Resonance Imaging (MRI) for better estimation of the tumor local extension and respectability and abdominal computed tomography (CT) and chest X-ray (CXR) for exact assessment of distant metastasis were recommended. MRI of the pelvis showed a vaginal mass 1-2 cm in size expanded inside the vagina without parametrial extension; no distant metastasis showed on abdominal CT scan and CXR (figure 1).
Histological study of the sampled tissue revealed large cells with clear cytoplasm, enlarged nucleoli, and tubular structures lined by hobnail cells. Cervical cytology was normal. The diagnosis was early stage primary CCC of vagina due to neither cervical involvement nor prenatal DES exposure history or sign of genital anomaly.
Radical hysterectomy with at least excision of about 4 cm of the vaginal wall, with bilateral oophorectomy (due to her menopausal status), complete pelvic lymph node dissection, and Para-aortic lymph node sampling were done for her. The permanent pathology confirmed residual clear cell carcinoma foci in the vagina with no more local extension [fig_ref] Figure 2: CCC of vagina showed invasive carcinoma with tubulopapillary architecture, hobnail nuclei cell [/fig_ref].
An occasional finding in the pathologic study of uterus was adenomyosis pattern in myometrium. Neoplastic involvement of Pelvic & Para-aortic lymph node was shown in a permanent report in regard of her pre-op imaging study.
After the non-complicated surgery and uneventful post-operative hospitalization, the patient referred to the Radiotherapy department and concurrent chemo-radiotherapy treatment planned for her. The patient is still under observation after 10 months of treatment and remains free from disease.
## Case 2
A 48-year-old multiparous woman was consulted in Mashhad University of Medical Sciences, Iran, gynecology oncology department with the complaint of abnormal vaginal bleeding for 6 months during 2017, July. The patient did not have important past medical history. Her last sexual activity was 17 years ago. She reported her husband's smoking behavior (passive smoking). She had no history of hormone contraceptive usage. Gynecologic examination showed no obvious visual abnormality in genitalia, vagina, and cervix. The lone finding was a bulky cervixor uterine isthmus without parametrial involvement. Her last cervical pap smear in one month before the consultation was normal.
Pelvic ultrasonography showed fluid in the uterus with thick endometrial lines, so endometrial sampling was performed for her. Histopathological examination revealed CCC of high grade in endometrial and endocervical curettage samples which was confirmed by Immunohistochemistry staining study; negative ER, PR. Ck20 and positive P16 and CEA. No more data were obtained on pelvic MRI and abdominal CT scan study. The incidental abnormal finding was an elevated CA125 in her laboratory test. She was classified as FIGO stage IB1 of the cervix, and laparotomy was planned for her. Intra-operatively, a barrelshaped cervix versus a small uterus was seen. Thus, radical hysterectomy (type 3) with parametrial resection and complete pelvic LND and Para-aortic LN sampling were performed by Gyneco-oncologist.
Histopathological examination in Microscopic view showed different sizes of tumoral cells with clear cytoplasm, large round nucleus with pleomorphism, hyperchromasia, frequent Mitotic activity, and prominent nucleoli (hobnailing appearance) in tubulocystic growth pattern); which confirmed 'high-grade CCCC' (figure 3).
The gyneco-pathologist also reported the main bulk of tumor in the uterine cervix (the largest diameter was less than 4 cm) versus limited tumoral involvement of the isthmuses of the uterus. The other surgical sample was free of tumor. Adjuvant chemo radiotherapy was the next step in her management, and after 6 months of follow-up, she didnot show any sign of recurrences. Written, informed consents were obtained from the patients for writing and publication of this case series.
# Discussion
This study presents two new cases of rare primary clear cell carcinoma of the vagina and cervix uteri. [bib_ref] Clear cell adenocarcinoma of the cervix in second generation young women who..., Tantitamit [/bib_ref] Previous articles reported the age of 26 th as the commonest age for this malignancy, which confirms the hypothesis of early age (intrauterine) DES-exposure as the commonest etiology for this type of cancer. But, recently, the presented age have risen as it was the case in this study cases, which suggests more studies about CCC risk factors 1-3 are needed. In the present article, two cases of gynecologic The most common presentation of CCC is the same as other neoplasms of this structure: atypical vaginal discharge and/or bleeding as reported in the present case series. [bib_ref] Clear Cell Carcinoma of Cervix in a 60-year-old Woman: A Case Report, Ramesh [/bib_ref] (The simple initial step for the diagnosis after the physical examination is cervical pap-smear and HPV testing, but normal cervical cytology could not rule out malignancies of adenocarcinoma type. The HPV infection role in CCCs development is not confirmed to date and there is controversy over this issue in the literature (especially HPV type 18 & 31). The other possible etiology in some case reports are advanced age, epidemiological risk and socioeconomic status, multiple sexual partners, nuliparity, smoking and contraceptive usage in addition to adenomyosis and endometriosis for endometrial CCC. [bib_ref] Clear Cell Carcinoma of Cervix in a 60-year-old Woman: A Case Report, Ramesh [/bib_ref] [bib_ref] Clear cell carcinoma of the cervix: a multiinstitutional review in the post-DES..., Thomas [/bib_ref] [bib_ref] Clear cell adenosquamous carcinoma of the cervix: a case report with discussion..., Garg [/bib_ref] Both cases in the present study were passive smokers and had a low socioeconomic condition. Hormonal contraceptive had used only by one of them; its role must be studied more in future to confirm.
Tumor marker (CA125) was elevated in second case, but the usefulness of this tumor for this cancer diagnosis or management is based on few case reports and is questionable. [bib_ref] Clear cell adenocarcinoma of the cevix without in utero diethylstilbestrol exposure: report..., Mousavi [/bib_ref] Clear cytoplasmic cells with tubulocystic pattern are the CCCs microscopic character. The immunohistochemistry staining by CK7, CAM5.2, 34 beta E12, CEA, C-A125, Leu-M1 and Vimentin (positivity), P53 and Bcl-2 (overexpression) and CK20, β-HCG and Alpha 1-fetoprotein (negative staining) are suggestive for CCC and could distinguish this malignancy from other differential diagnoses such as Aria-Stella reaction, micro glandular hyperplasia, and mesonephric hyperplasia in adults, and Yolk sac tumor, sarcoma botryoides, and embryonal rhabdomyosarcoma in children, and metastatic renal cell carcinoma in all ages. Although staining for ER/PR and HER2 Neu shows variable positivity.Because of CCC rarity, the individual experiencebased on case reports is the main source of management; no approved guideline persists.
The prognostic factor in all malignancies are stage, tumor size, growth pattern, nuclear atypia, mitotic activity, and lymph node involvement, so the worse prognosis in CCCs in regard of other malignancies' subtypes is controversy. [bib_ref] Clear cell carcinoma of the cervix: a multiinstitutional review in the post-DES..., Thomas [/bib_ref] The prognosis in the early stage has reported to be near 80-85% by either radical hysterectomy and lymph node dissection, radiotherapy, 3 or radical trachelectomy (by the aim of maintaining fertility and with/or without chemotherapy).
Fertility-sparing surgery in early stage of CCC in young patient is recommended by the National Comprehensive Cancer Network (of Fort Washington) in 2013, too. [bib_ref] Clear Cell Carcinoma of Cervix in a 60-year-old Woman: A Case Report, Ramesh [/bib_ref] [bib_ref] Neoadjuvant chemotherapy followed by radical vaginal trachelectomy and adjuvant chemotherapy for clear..., Singh [/bib_ref] The main treatment in early-stage adenocarcinoma is radical hysterectomy, but it is important to put emphasis on adjuvant chemoradiotherapy in high-risk patients to reduce the risk of distant metastasis.Disease-free and overall survival after positive versus negative lymph node detection is 31% versus 92% (P<0.001) and 80% versus 100% (P=0.02), respectively. The most expected recurrence time is 1 year after the definite treatment. The most common site for CCC early recurrences are lung, liver, and the skeletal system, but close follow-up for a long time is necessary. Recurrences in pelvis and vagina were mostly seen in about 4 years after the remission. [bib_ref] Clear cell carcinoma of the cervix: a multiinstitutional review in the post-DES..., Thomas [/bib_ref] As the prognosis of CCC is controversial in different Studies, 1, 2 long life follow-up the is planned for the presented cases in this study, and now, after 10 months of monitoring, they have reported no recurrence.
# Conclusion
Due to the sporadic cases of no DES exposure CCC, the probable etiology and the most effective mode of treatment remains unknown and depends on the patient's factors such as tumor size, surgeon, and oncologist team working experiences.
## Conflict of interest:
None declared.
[fig] Figure 1: Sagittal view of Pelvic MRI showed vaginal fornix expansion probably due to mass in the lateral of the cervix which had expanded the vagina without cervical and parametrial extension. [/fig]
[fig] Figure 2: CCC of vagina showed invasive carcinoma with tubulopapillary architecture, hobnail nuclei cell (The red arrow), and Clear cells (The blue arrow). (H&Ex100& ×400) [/fig]
[fig] Figure 3: A) Clear cell carcinoma of cervix showed infiltrative tubulopapillary carcinoma with eosinophilic cytoplasm with hobnail nuclei. (H&Ex100 & ×400) B) Immunohistochemical staining showed diffuse Napsin A (top) and CK7 (bottom) staining of tumor cell. (×100) organ clear cell carcinoma without the history of DES-exposure with no recurrences after radical surgery and adjuvant therapy are reported. [/fig]
|
Identification of an inflammatory response signature associated with prognostic stratification and drug sensitivity in lung adenocarcinoma
Increasing evidence has confirmed the close connection between inflammatory response and tumorigenesis. However, the relationship between inflammatory response genes (IRGs) and the prognosis of lung adenocarcinoma (LUAD) as well as the response to drug therapy remains poorly investigated. Here, we comprehensively analyzed IRGs RNA expression profiling and clinical features of over 2000 LUAD patients from 12 public datasets. The Cox regression method and LASSO analysis were combined to develop a novel IRG signature for risk stratification and drug efficacy prediction in LUAD patients. Enriched pathways, tumor microenvironment (TME), genomic and somatic mutation landscape in different subgroups were evaluated and compared with each other. This established IRG signature including 11 IRGs (ADM, GPC3, IL7R, NMI, NMURI, PSEN1, PTPRE, PVR, SEMA4D, SERPINE1, SPHK1), could well categorize patients into significantly different prognostic subgroups, and have better predictive in independently assessing survival as compared to a single clinical factor. High IRG scores (IRGS) patients might benefit more from immunotherapy and chemotherapy. Comprehensive analysis uncovered significant differences in enriched pathways, TME, genomic and somatic mutation landscape between the two subgroups. Additionally, integrating the IRGS and TNM stage, a reliable prognostic nomogram was developed to optimize survival prediction, and validated in an independent external dataset for clinical application. Take together, the proposed IRG signature in this study is a promising biomarker for risk stratification and drug efficacy prediction in LUAD patients. This study may be meaningful for explaining the responses of clinical therapeutic drugs and providing new strategies for administrating sufferer of LUAD.
# Materials and methods
Data acquisition and processing. The IRGs RNA expression profiles and clinical features of LUAD patients from 12 public cohorts were retrospectively and comprehensively analyzed. And these 12 datasets information were visible in [fig_ref] Table 1: Basic clinicopathologic features of LUAD patients in the two cohorts [/fig_ref]. 200 IRGs were collected from the GSEA website (http:// www. gsea-msigdb. org/ gsea/ index. jsp) . A thorough query of the LUAD dataset was performed in the Gene Expression Omnibus (GEO) database in order to obtain as many and eligible samples as possible, and a total of 11 independent LUAD study datasets (GSE10072 [bib_ref] Gene expression signature of cigarette smoking and its role in lung adenocarcinoma..., Landi [/bib_ref] , GSE14814 17 , GSE29013 [bib_ref] Robust gene expression signature from formalin-fixed paraffin-embedded samples predicts prognosis of non-smallcell..., Xie [/bib_ref] , GSE30219 [bib_ref] Ectopic activation of germline and placental genes identifies aggressive metastasis-prone lung cancers, Rousseaux [/bib_ref] , GSE31210 20 , GSE37745 21 , GSE40791 22 , GSE42127 7,23 , GSE50081 2 , GSE68465 24 and GSE72094 [bib_ref] Differential association of STK11 and TP53 with KRAS mutation-associated gene expression, proliferation..., Schabath [/bib_ref] were finally included in our study, with GSE10072, GSE30219, GSE31210, GSE40791 and GSE68465 containing corresponding lung normal tissues, and GSE14814, GSE29013, GSE30219, GSE31210, GSE37745, GSE42127, GSE50081, GSE68465 and GSE72094 containing survival data. According to the corresponding annotation files, we converted the probes to gene symbols. For genes with multiple probe set signals, their values were averaged to generate a single expression value. To reduce the batch effects from non-biological technical biases, we finally integrated them into a meta-GEO cohort using the 'ComBat' algorithm of 'sva' R package [bib_ref] Exploration, normalization, and summaries of high density oligonucleotide array probe level data, Irizarry [/bib_ref] [bib_ref] Adjusting batch effects in microarray expression data using empirical Bayes methods, Johnson [/bib_ref]. As an independent validation dataset, TCGA-LUAD RNAseq data (FPKM format) were obtained from the Genomic Data Commons (GDC, https:// portal. gdc. cancer. gov/). TCGA somatic mutation data were also collected from the GDC. Additionally, we also downloaded the copy number variation (CNV) profiles from UCSC Xena (https:// xenab rowser. net). The CNV landscape of the key genes in human chromosomes was visualized using 'Rcircos' R package.
Establishment and validation of an IRG signature. After removing samples without clear survival information, eventually, 1615 patients were included in the meta-GEO cohort, and 500 patients were included in the TCGA cohort [fig_ref] Table 1: Basic clinicopathologic features of LUAD patients in the two cohorts [/fig_ref]. Univariate Cox analysis were respectively performed for the IRGs in the two LUAD cohorts . The common genes from the statistically significant genes (p value < 0.05) in the two cohorts, were considered prognostic-related IRGs (PRIRGs), and were then fitted into LASSO regression analysis (using 'glmnet' R package) to reduce the dimensionality of the data. Next, using 'survival' R package, multivariate Cox analysis was performed to calculate the coefficient of each gene. Subsequently, IRG scores (IRGS) was generated according to the following formula: IRGS = β1 × expressionG1 + β2 × expressionG2 + ⋯ + βn × expressionGn, where βn represented the coefficient of genen and expressionGn was the expression level of genen. With the median IRGS as cut-off, patients were classified into high-and low-IRGS groups. Probability of survival was estimated by the Kaplan-Meier method, with differences between two groups tested using the log-rank test. The IRG signature constructed in this study was compared to these signatures constructed in other publications [bib_ref] Immune signature-based risk stratification and prediction of immune checkpoint inhibitor's efficacy for..., Yi [/bib_ref] [bib_ref] Identification of a costimulatory molecule-based signature for predicting prognosis risk and immunotherapy..., Zhang [/bib_ref] [bib_ref] Identification and validation of hypoxia-derived gene signatures to predict clinical outcomes and..., Shi [/bib_ref]. The receiver operating characteristic (ROC) curve, a combination of sensitivity (true positive rate) and specificity (true negative rate), was used to assess prediction performance of the IRGS, and the corresponding area under the curve (AUC) values were also calculated. Additionally, subgroup analysis was also performed, and the Kaplan-Meier survival curves were plotted to verify the predictive performance in the different subgroups. www.nature.com/scientificreports/ www.nature.com/scientificreports/ Cell culture, and qPCR analysis. Two LUAD cell lines (A549 and H1975), and normal epithelial cell line of the human lung (beas-2B) were purchased from the Chinese Academy of Sciences in Shanghai. All cells were cultured in RPMI 1640 medium (Invitrogen, Shanghai, China) which supplemented with 10% FBS (Gibco, USA), 100 U/ml penicillin and 100 mg/ml streptomycin in a humidifed atmosphere with 5% CO2 at 37 ℃. Applying the Trizol method, the total RNA was extracted and subsequently used to synthesize cDNA and subjected to PCR reactions (all experimental procedures were performed strictly according to the instructions of the kit). GAPDH was used as the reference gene and relative gene expression was calculated by the 2−ΔΔCT method. Primer sequences of NMI, GPC3 and GAPDH were listed as following: NMI forward: Gene set variation analysis and gene set enrichment analysis. As a non-parametric and unsupervised method, gene set variation analysis (GSVA) bypassed traditional methods of explicitly modeling phenotypes in enrichment scoring algorithms 31 , could be used to estimate changes in the activity of pathways and biological processes in expression data set samples. In this study, GSVA method was also performed to investigate the difference on biological processes between different IRGS subgroups. Additionally, gene set enrichment analysis (GSEA) was also conducted to explore whether biological functions and process were significantly different between different IRGS subgroups using 'clusterProfiler' R package [bib_ref] ConsensusClusterPlus: A class discovery tool with confidence assessments and item tracking, Wilkerson [/bib_ref]
## Estimation of tumor microenvironment (tme).
To better characterize the LUAD immune landscape,
we used a single-sample GSEA (ssGSEA) algorithm to quantify the infiltrating levels of 23 immune cells in the tumor microenvironment (TME) [bib_ref] Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1, Barbie [/bib_ref] [bib_ref] Pan-cancer immunogenomic analyses reveal genotype-immunophenotype relationships and predictors of response to checkpoint..., Charoentong [/bib_ref]. Since the Estimation of Stromal and Immune Cells in Malignant Tumors using Expression Data (ESTIMATE) algorithm 35 could take advantage of the unique properties of the transcriptional profiles to infer the tumor cellularity as well as the tumor purity, we calculated immune and stromal scores for LUAD tumor samples using the 'estimate' R package. Tumor tissues with abundant infiltration of immune cells often has higher immune scores, with the opposite level of tumor purity. Moreover, we also applied another algorithm named 'cell type identification by estimating relative subsets of RNA transcripts' (CIBERSORT) [bib_ref] Robust enumeration of cell subsets from tissue expression profiles, Newman [/bib_ref] (http:// ciber sort. stanf ord. edu/) to quantify the abundance of immune cell infiltration in the LUAD tumor samples. Then the samples with P-value less than 0.05 were used for further analysis.
Estimation of drug sensitivity. IC50 (semi-inhibitory concentration) was an important indicator of assessing drug efficacy or response to sample treatment. Based on the sample transcriptome, we evaluated the IC50 for each sample using 'pRRophetic' R package 37 to compare drug sensitivity across different IRGS subgroups. Higher IC50 represented lower drug sensitivity [bib_ref] PRRophetic: An R package for prediction of clinical chemotherapeutic response from tumor..., Geeleher [/bib_ref]. Moreover, we downloaded the files named 'RNA: RNA-seq' and 'Compound activity: DTP NCI-60' from the CellMiner (https:// disco ver. nci. nih. gov/ cellm iner/ home. do), further analyzed the correlation of FDA approved drugs Z scores with the IRG expression values.
Additional bioinformatic and statistical analyses. The 'limma' R package was applied to performed the analysis of expression difference between LUAD tumor tissues and lung normal tissues, and the results were visualized by heatmaps. According to the data characteristics, Spearman or Pearson method were used for correlation analysis. Wilcoxon test and Kruskal-Wallis test were used to compare the statistical differences between two groups and multiple groups, respectively. The mutation landscape in different IRGS subgroups were presented using the 'maftools' R package. The Cox proportional hazards model was adopted to calculate the hazard ratios (HR) for each gene and clinical feature, through which the independent prognostic factors were ascertained simultaneously. We employed the 'forestplot' R package to visualize the results. Additionally, to assess the prognosis classification performance of the IRGS and other clinical factors, the 'survivalROC' R package was used to draw ROC curves, and the corresponding AUC were also estimated. Using the clinical characteristics and IRGS as input, multivariate Cox regression analysis with variable selection was implemented to identify the powerful combination of these predictors. Then, we built a quantitative nomogram with the 'rms' R package to predict the individual 1-, 3-and 5-year survival probabilities. To evaluate the prediction performance of the nomogram, the calibration curves, with the Hosmer-Lemeshow test, were used to judge the consistency between the model prediction values and the actual results. All statistical analyses were performed in R 3.6.2 software.
# Guidelines statement.
This study obtained open data from the GEO database (https:// www. ncbi. nlm. nih.
gov/ geo/) and TCGA (https:// cance rgeno me. nih. gov/). All experimental protocols were performed in accordance with the relevant guidelines and regulations and adhered to the Declaration of Helsinki.
# Results
The IRGS presents good evaluation performance in LUAD patients. We obtained a total of 200 IRGs from the GSEA website, and their RNA expression profiles were applied for univariate Cox analysis. There were 81 prognosis-related IRGs (PRIRGs) in the meta-GEO cohort and 53 in the TCGA cohort [fig_ref] Figure 2: Controduction and verification of an IRG signature for LUAD patients [/fig_ref] , . The common PRIRGs in the two cohorts (n = 32) were conducted into LASSO analysis [fig_ref] Figure 2: Controduction and verification of an IRG signature for LUAD patients [/fig_ref] www.nature.com/scientificreports/ sion). Finally, a total of 11 genes (ADM, GPC3, IL7R, NMI, NMURI, PSEN1, PTPRE, PVR, SEMA4D, SER-PINE1, SPHK1) were included in a predictive signature according to their risk coefficients [fig_ref] Figure 2: Controduction and verification of an IRG signature for LUAD patients [/fig_ref] , . We then evaluated the IRGS for each patient using the score calculation formula mentioned in the material section.
With the median IRGS as cut-off, patients were classified into high-and low-IRGS groups. The Kaplan Meier survival curves of the two cohorts clearly showed that patients in low-IRGS group presented a longer overall survival (OS) than the high-IRGS patients [fig_ref] Figure 1: Schematic diagram of the study design [/fig_ref]. The distribution of patients with different IRGS and survival status were shown in [fig_ref] Figure 2: Controduction and verification of an IRG signature for LUAD patients [/fig_ref] ,g (the meta-GEO cohort, n = 1615) and [fig_ref] Figure 1: Schematic diagram of the study design [/fig_ref] ,D (the TCGA cohort, n = 500). As in [fig_ref] Figure 2: Controduction and verification of an IRG signature for LUAD patients [/fig_ref] -k, the AUC values of IRGS constructed in this study [fig_ref] Figure 2: Controduction and verification of an IRG signature for LUAD patients [/fig_ref] were significantly better than these signatures in other studies [bib_ref] Immune signature-based risk stratification and prediction of immune checkpoint inhibitor's efficacy for..., Yi [/bib_ref] [bib_ref] Identification of a costimulatory molecule-based signature for predicting prognosis risk and immunotherapy..., Zhang [/bib_ref] [bib_ref] Identification and validation of hypoxia-derived gene signatures to predict clinical outcomes and..., Shi [/bib_ref] [fig_ref] Figure 2: Controduction and verification of an IRG signature for LUAD patients [/fig_ref] in TCGA. The AUC values of ROC curves in the meta-GEO cohort at 1 year, 3 years and 5 years were 0.694, 0.678 and 0.658 respectively [fig_ref] Figure 1: Schematic diagram of the study design [/fig_ref]. Additionally, to verify the predictive performance of the IRGS in the different LUAD subgroups, subgroup analyses and the Kaplan-Meier survival curves were performed. As shown in [fig_ref] Figure 2: Controduction and verification of an IRG signature for LUAD patients [/fig_ref] , high-IRGS patients presented shorter OS than low-IRGS patients in most subgroups (≤ 65 years, > 65 years, female, male, stage I/II). In a large cohort of 1615 patients (the meta-GEO cohort), high-IRGS patients were also observed worse prognosis in stage III/IV subgroups [fig_ref] Figure 2: Controduction and verification of an IRG signature for LUAD patients [/fig_ref]. These above analyses indicated that the IRGS had good evaluation performance for risk stratification and prognosis prediction of LUAD patients. www.nature.com/scientificreports/ Multi-omics features of IRG signature genes reveal their great importance in LUAD. Based on the above analysis, we constructed an IRG signature containing 11 genes (ADM, GPC3, IL7R, NMI, NMURI, PSEN1, PTPRE, PVR, SEMA4D, SERPINE1, SPHK1). These genes were called as IRG signature genes. Most transcription factors (TFs) are associated with the cell cycle and play vital roles in the induction of proto-oncogene and tumor suppressor genes. We thus further explored the association of these 11 genes with 318 known TFs from the Cistrome program (http:// cistr ome. org/) . By co-expression analysis, we finally identified 173 TFs associated with the 11 genes . A functional analysis based on KEGG database 39 indicated that these 173 TFs were involved in six major pathways [fig_ref] Figure 3: The landscape of genetic alterations of IRG signature genes in LUAD [/fig_ref]. All the functional categories in these pathways were important, especially cancers, signal transduction, infectious diseases and immune system [fig_ref] Figure 3: The landscape of genetic alterations of IRG signature genes in LUAD [/fig_ref]. Further analysis revealed that CNV of 11 IRG signature genes were prevalent. GPC3, IL7R, NMI, PTPRE, PVR, SEMA4D, SERPINE1 and SPHK1 showed widespread CNV amplification. In contrast, ADM, PSEN1 and NMURI had prevalent CNV deletions [fig_ref] Figure 3: The landscape of genetic alterations of IRG signature genes in LUAD [/fig_ref]. And [fig_ref] Figure 3: The landscape of genetic alterations of IRG signature genes in LUAD [/fig_ref] visually showed the CNV alteration positions of these 11 genes on the chromosome. Innovatively, we further depicted the frequency of somatic mutations in these 11 genes in LUAD. In 561 samples, there was only 56 (9.98%) experienced genetic alterations, primarily including missense mutations, nonsense mutations and multi hit. Of these 11 genes, IL7R showed the highest mutation frequency, followed by SERPINE1 [fig_ref] Figure 3: The landscape of genetic alterations of IRG signature genes in LUAD [/fig_ref]. We further explored the correlation between IRG signature gene expression and CNV as well as single nucleotide polymorphism (SNP) to clarify their potential relationship, and found that the CNV of some genes, such as PTPRE, PVR, IL7R, SEMA4D, NMI, PSEN1, significantly affected gene expression [fig_ref] Figure 3: The landscape of genetic alterations of IRG signature genes in LUAD [/fig_ref]. Mutations in the IL7R gene weaken IL7R gene expression, while mutations in the other 10 genes did not seem to affect the expression levels of the corresponding genes [fig_ref] Figure 3: The landscape of genetic alterations of IRG signature genes in LUAD [/fig_ref]. Subsequently, differentially expression analysis was performed, and it was found that many genes displayed different expression features in tumor tissues in different datasets [fig_ref] Figure 4: Enriched biological pathways and tumor immune microenvironment characteristics in different IRGS subgroups [/fig_ref]. For example, IL7R was down-regulated in LUAD tumor tissues The somatic mutation data were visualized using "maftool" R package. (f,g) Correlation analysis revealed the relationship between gene and gene, gene and IRGS in meta-GEO cohort (f) and TCGA cohort (g) (*p < 0.05; **p < 0.01; ***p < 0.001). www.nature.com/scientificreports/ from GSE10072, TCGA and GSE40791, but up-regulated in GSE68465. Of these 11 IRG signature genes, only two genes (GPC3 and NMI) showed consistent changes in RNA expression. GPC3 was significantly down-regulated in tumor tissue in all data sets [fig_ref] Figure 4: Enriched biological pathways and tumor immune microenvironment characteristics in different IRGS subgroups [/fig_ref] , while NMI was significantly up-regulated [fig_ref] Figure 4: Enriched biological pathways and tumor immune microenvironment characteristics in different IRGS subgroups [/fig_ref]. We further validated the expression level of NMI and GPC3 in LUAD from the mRNA and protein levels. As shown in [fig_ref] Figure 5: Comparison of the genetic characteristics between high-and low-IRGS groups and prediction of... [/fig_ref] -C, NMI was observably up-regulated in LUAD tumor tissue and cells (A549 and H1975). As expected, GPC3 was remarkably down-regulated in LUAD tumor tissue and cells (A549 and H1975) [fig_ref] Figure 5: Comparison of the genetic characteristics between high-and low-IRGS groups and prediction of... [/fig_ref]. Correlation analysis revealed the relationship between genes and genes as well as genes and IRGS. In meta-GEO cohort, it was observed that IRGS was positively strongly related to the expression of ADM, NMI, PSEN1, PTPRE, PVR, SERPINE1 and SPHK1, and negatively correlated with GPC3, IL7R, NMURI, PTPRE and SEMA4D [fig_ref] Figure 3: The landscape of genetic alterations of IRG signature genes in LUAD [/fig_ref]. The same findings were also observed in TCGA cohort [fig_ref] Figure 3: The landscape of genetic alterations of IRG signature genes in LUAD [/fig_ref]. Also, we noted that the expression of IL7R and SEMA4D, and the expression of SERPINE1 and SPHK1 were significantly positively correlated in the two cohorts. Consistent with the results from univariate Cox analysis , Kaplan-Meier survival analysis also revealed that the expression of ADM, GPC3, IL7R, NMI, NMURI, PSEN1, PTPRE, PVR, SEMA4D and SPHK1 significantly affected patient survival [fig_ref] Figure 3: The landscape of genetic alterations of IRG signature genes in LUAD [/fig_ref]. Multivariate Cox analysis showed that ADM, IL7R, NMI, NMUR1, PTPRE, PVR, SEMA4D and SPHK1 were independent prognostic factors of LUAD, where ADM, NMI, PTPRE, PVR, SPHK1 were prognostic risk factors, while IL7R, NMUR1 and SEMA4D were favorable prognostic factors [fig_ref] Figure 2: Controduction and verification of an IRG signature for LUAD patients [/fig_ref].
The analyses mentioned above told about the multi-omics traits of the 11 IRG signature genes in LUAD. The effect on prognosis, transcriptomic (abnormal expression of GPC3 and NMI) and genomic (mutations of IL7R, CNV of PTPRE, PVR, IL7R, SEMA4D, NMI and PSEN1) alterations suggested that these IRG signature genes were of great importance in LUAD.
## Comprehensive analyses reveal different biological processes and tme between two irgs subgroups.
To infer the biological behaviors between different IRGS subgroups, GSVA and GSEA analyses were performed. The GSVA revealed eight co-enriched pathways obtained from both cohorts (meta-GEO and TCGA cohorts) [fig_ref] Figure 4: Enriched biological pathways and tumor immune microenvironment characteristics in different IRGS subgroups [/fig_ref] , where regulation of DNA directed DNA polymerase activity, Ctf18 RFC-like complex, chromosome passenger complex and condensed chromosome outer kinetochore were enriched in the high-IRGS group, while antigen processing and presentation endogenous lipid antigen via MHC class IB, lipid antigen binding, alveolar lamellar body, and positive regulation of autophagosome maturation were enriched in the low-IRGS group. The GSEA also revealed that there were two co-enriched pathways among the top five pathways obtained from both cohorts. These two pathways including DNA replication and mismatch repair were enriched in the high-IRGS group [fig_ref] Figure 4: Enriched biological pathways and tumor immune microenvironment characteristics in different IRGS subgroups [/fig_ref]. Additionally, ESTIMATE algorithm was used to quantify the immune score and tumor purity for each sample. The wilcoxon rank sum test revealed significant differences in immune score and tumor purity between low-and high-IRGS groups. These results from ESTIMATE algorithm told about that the high-IRGS patients presented higher tumor purity, and the opposite was true for immune scores [fig_ref] Figure 4: Enriched biological pathways and tumor immune microenvironment characteristics in different IRGS subgroups [/fig_ref]. Immune-infiltrating cells played important roles in TME, and we subsequently quantified immune-infiltrating cells in each IRGS subgroup using ssGSEA and CIBERSORT algorithms to further explain the potential association between risk stratification and TME in LUAD. ssGSEA revealed the low IRGS tumors were significantly infiltrated by activated B cell, immature B cell, activated CD8 T cell, immature dendritic cell (DC), activated DC, plasmacytoid DC, Type 1T helper cell, T follicular helper cell, Type 17T helper cell, NK cell, NK T cell, eosinophil, macrophage, MDSC and mast cell. While activated CD4 T cell, CD56dim NK cell significantly infiltrated in the high IRGS tumors [fig_ref] Figure 4: Enriched biological pathways and tumor immune microenvironment characteristics in different IRGS subgroups [/fig_ref]. CIBERSORT also showed that significant differences in immune cells infiltrating level were observed between high and low IRGS tumors [fig_ref] Figure 6: Drug sensitivities comparison between different IRGS subgroups [/fig_ref]. The results from ssGSEA and CIBERSORT might help to explain why the low-IRGS group presented higher immune scores and lower tumor purity. We also compared the abundance of immune infiltrating cells between LUAD normal and tumor tissues [fig_ref] Figure 6: Drug sensitivities comparison between different IRGS subgroups [/fig_ref] , and further explored the correlation between the abundance of different immune infiltrating cells. It was found that LUAD tumor tissue was significantly infiltrated by B cells naive, macrophages M1, plasma cells, DCs resting, T cells CD4 memory activated, T cells follicular helper and Tregs. While T cells CD4 memory resting, NK cells resting, mast cells resting, monocytes, neutrophils, macrophages M0/2, eosinophils significantly infiltrated in LUAD normal tissue. As shown in [fig_ref] Figure 6: Drug sensitivities comparison between different IRGS subgroups [/fig_ref] , mast cells resting showed a significantly positive correlation with NK cells activated (Cor = 0.35), and plasma cells showed a significantly negative correlation with macrophages M2 (Cor = − 0.45). Given the differential infiltrating abundance of immune cells in different tissues (high, low IRGS tumor and normal tissues), we investigated the relationship between the immune cell infiltrating level and patient prognosis. The results showed that except T cells gamma delta, other cells did not affect the survival of patients [fig_ref] Figure 7: Construction and validation of a nomogram to optimize survival prediction [/fig_ref]. However, between the different tissues, we did not observe the difference in the T cells gamma delta infiltrating level [fig_ref] Figure 6: Drug sensitivities comparison between different IRGS subgroups [/fig_ref]. Further investigation in terms of the relationship between TME infiltrating cells and IRG signature genes as well as IRGS, we found that IRG signature genes and IRGS were positively or negatively strongly related to at least 7 cell types based on the ssGSEA method, in which IL7R was significantly positively related to a large number of immune cells, while IRGS were significantly negatively correlated with most immune cells [fig_ref] Figure 4: Enriched biological pathways and tumor immune microenvironment characteristics in different IRGS subgroups [/fig_ref]. This result varied somewhat from the results based on the CIBERSORT algorithm [fig_ref] Figure 6: Drug sensitivities comparison between different IRGS subgroups [/fig_ref] , probably due to the internality of the different algorithms and the immune cell abundance calculated by CIBERSORT method with more emphasis on the cell status and subtypes. This also further illustrated the complexity of TME. Taken together, the high-and low-IRGS groups based on IRG signature had significantly different enrichment pathways and TME. We speculated that the abnormal expression of IRG signature genes, the remodeling of TME and the change of tumor biological pathways affected each other, resulting in different prognosis of patients. www.nature.com/scientificreports/ The potential of IRGS in predicting therapeutic responses. To reveal the somatic mutational landscapes for insight into the mutational process of LUAD, we made a genome-wide somatic mutation profile comparison between different risk subgroups based on TCGA data. [fig_ref] Figure 5: Comparison of the genetic characteristics between high-and low-IRGS groups and prediction of... [/fig_ref] ,b displayed the mutation www.nature.com/scientificreports/ landscapes of the top 30 most common mutant genes in the high-and low-IRGS groups of LUAD, respectively. TP53, TTN, and MUC16 for LUAD, were the top three most frequently mutated genes identified in this study. The high-IRGS group exhibited a significantly higher proportion of specific gene mutations. This enabled us to more comprehensively describe the impact of IRGS risk stratification on genomic alterations. Additionally, the Kaplan-Meier survival analysis was performed to further understand the cross-talk among somatic mutations, IRGS and patient survival. This results revealed that tumor mutation burden (TMB) had little to do with patient prognosis [fig_ref] Figure 5: Comparison of the genetic characteristics between high-and low-IRGS groups and prediction of... [/fig_ref] , but patients with high TMB combined with high IRGS presented the worst prognosis. In contrast, patients with low TMB combined with low IRGS predicted best [fig_ref] Figure 5: Comparison of the genetic characteristics between high-and low-IRGS groups and prediction of... [/fig_ref]. Increasing evidence [bib_ref] First-line nivolumab in stage IV or recurrent non-small-cell lung cancer, Carbone [/bib_ref] [bib_ref] Tumor mutational load predicts survival after immunotherapy across multiple cancer types, Samstein [/bib_ref] has demonstrated the association between TMB and immunotherapy response. Hence, we compared the distribution of TMB in the high-and low-IRGS groups and found higher TMB in the high-IRGS group [fig_ref] Figure 5: Comparison of the genetic characteristics between high-and low-IRGS groups and prediction of... [/fig_ref] , which was consist with the results from [fig_ref] Figure 5: Comparison of the genetic characteristics between high-and low-IRGS groups and prediction of... [/fig_ref] ,b that high-IRGS group exhibited significantly higher mutation frequencies. The above analyses initially indicated that high-IRGS patients may present a higher response rate to immunotherapy. To further confirm this conclusion, we applied a website named Tumor Immune Dysfunction and Exclusion (TIDE) (http:// tide. dfci. harva rd. edu) to compute TIDE scores for different IRGS subgroups. TIDE scores could be calculated for each sample by TIDE website based on the transcriptome data [bib_ref] Signatures of T Cell dysfunction and exclusion predict cancer immunotherapy response, Jiang [/bib_ref]. A lower TIDE score showed a higher response rate against both PD-1 and anti-CTLA-4 drugs. Our analysis revealed that the TIDE score was remarkably increased in the low-IRGS group [fig_ref] Figure 5: Comparison of the genetic characteristics between high-and low-IRGS groups and prediction of... [/fig_ref]. To sum up, patients at high IRGS may have a better immune response when receiving immune checkpoints inhibitors. Chemotherapy has been widely used in the treatment of many malignancies, including LUAD. Not all patients clinically benefited from these treatments, so it was of great importance to identify subgroups of patients who may be more sensitive to certain drugs. Based on the sample transcriptome, we here evaluated and compared drug sensitivity across different IRGS subgroups. As shown in [fig_ref] Figure 6: Drug sensitivities comparison between different IRGS subgroups [/fig_ref] ,b, the IC50s of several commonly-used drugs (cisplatin, paclitaxel, docetaxel, doxorubicin, gemcitabine) were lower in the high-IRGS group, indicating that high-IRGS patients might show higher sensitivity to these drugs.
Additionally, CellMiner analysis also revealed the correlation of the 11 IRG signature genes and FDA approved drugs. -P showed the correlation between the first 16 drugs (p values small to large) and IRG signature genes. Some of these genes (such as PSEN1, PTPRE, SPHK1, PVR, SERPINE1, NMUR1) showed the correlation with the predicted drugs . The correlation network map was visualized in . The above results preliminarily revealed the relationship between risk stratification as well as IRG signature gene expression and the sensitivity of commonly-used drugs, which could provide valuable clues to the development of individualized therapeutic strategies for LUAD patients.
Establishment of a prognostic nomogram to optimize survival prediction in LUAD patients. Independent prognostic analyses were respectively performed based on data from the meta-GEO cohort of 1615 patients and the TCGA cohort of 490 patients containing detailed clinical information. As shown in [fig_ref] Figure 7: Construction and validation of a nomogram to optimize survival prediction [/fig_ref] -d, in both cohorts, clinical stage and IRGS were considered as independent prognostic factors for LUAD patients. To assess the estimated performance of IRGS in predicting the prognosis of patients, we further calculated the AUC values, and found that 3-year AUC values of the IRGS in two cohorts were 0.675 and 0.692, respectively, both higher than other clinicopathologic factors (such as age, sex and stage) [fig_ref] Figure 7: Construction and validation of a nomogram to optimize survival prediction [/fig_ref]. Based on the two independent prognostic factors, we established a prognostic nomogram for LUAD patient in the meta-GEO cohort [fig_ref] Figure 7: Construction and validation of a nomogram to optimize survival prediction [/fig_ref]. The ROC analyses indicted the nomogram presented a powerful capacity for survival prediction with high AUC values in both the meta-GEO and TCGA cohorts [fig_ref] Figure 7: Construction and validation of a nomogram to optimize survival prediction [/fig_ref]. Moreover, the calibration curves in both cohorts also presented favourable consistency with the ideal performance [fig_ref] Figure 7: Construction and validation of a nomogram to optimize survival prediction [/fig_ref] , indicating a high accuracy of the nomogram.
# Discussion
In view of the complex oncogenic mechanisms and tumor heterogeneity in LUAD, the development of personalized management strategies and accurate prediction of patient prognosis remain extremely challenging. IRGs play crucial roles in tumorigesis and progression [bib_ref] Prediction of survival of diffuse large B-cell lymphoma patients via the expression..., Zhao [/bib_ref] [bib_ref] Assembly of inflammation-related genes for pathway-focused genetic analysis, Loza [/bib_ref] , but the relationship between IRGs and LUAD prognosis as well as the response to drug therapy remains poorly investigated. Therefore, based on IRG expression profiles, this study attempts to develop a model for prognosis risk stratification and drug therapy response prediction in LUAD patients, hoping to provide reference for clinical decision-making and future studies. In the present study, based on as many as possible LUAD patients (n = 1615) collected from multiple GEO chips, we developed a prognostic IRG signature that could effectively identify high-and low-risk patients. Subsequently, prognostic power was validated in an independent TCGA cohort of 500 patients. Notably, in the meta-GEO and TCGA cohorts, we observed that the IRGS, as with TNM stage, was an independent prognostic factor for LUAD patients, and its prognostic capacity was also superior to some clinicopathologic parameters (e.g., TNM stage, age, gender). In the meta-GEO cohort, the 3-year and 5-year AUC values of the model constructed in this study were not as good as those of Yi et al. [bib_ref] Immune signature-based risk stratification and prediction of immune checkpoint inhibitor's efficacy for..., Yi [/bib_ref] , but this result was still acceptable and interpretable. Compared with the models constructed in previous studies [bib_ref] Immune signature-based risk stratification and prediction of immune checkpoint inhibitor's efficacy for..., Yi [/bib_ref] [bib_ref] Identification of a costimulatory molecule-based signature for predicting prognosis risk and immunotherapy..., Zhang [/bib_ref] [bib_ref] Identification and validation of hypoxia-derived gene signatures to predict clinical outcomes and..., Shi [/bib_ref] , the IRG signature constructed in this study still exhibited ideal predictive performance, which was a parallel comparison based on TCGA data sources. Furthermore, in some LUAD patient subgroups, the IRG signature retained its predictive power to effectively distinguish between high-and low-IRG patients.
A growing number of studies have proved that tumorigenesis and progression were closely related to TME [bib_ref] Prediction of venous metastases, recurrence, and prognosis in hepatocellular carcinoma based on..., Budhu [/bib_ref] [bib_ref] Chemokine-driven lymphocyte infiltration: An early intratumoural event determining long-term survival in resectable..., Chew [/bib_ref] [bib_ref] The adaptive immunologic microenvironment in colorectal cancer: A novel perspective, Galon [/bib_ref] [bib_ref] Tumor cell-microenvironment interaction models coupled with clinical validation reveal CCL2 and SNCG..., Hu [/bib_ref] , where various cytokines, chemokines and cells interacting with tumor cells, especially immune cells, were increasingly considered to play key roles in tumor resistance in vivo. In this study, multiple methods (CIB-ERSORT, ssGSEA and ESTIMATE) were used to depict the TME landscapes of different IRGS subgroups. We observed that low-IRGS group had more types of immune cell infiltrating, with anti-tumor cells (such as CD8 + T cells, macrophage, NK cells, dendritic cells and so on) as well as immunosuppressive cells (e.g., MDSC). Tumor www.nature.com/scientificreports/ www.nature.com/scientificreports/ tissues infiltrated by abundant immune cells had higher immune score and lower tumor purity. And the results from ESTIMATE algorithm (the low-IRGS group had higher immune score and lower tumor purity) also further confirmed the above-stated conclusion. Previous studies [bib_ref] Low tumor purity is associated with poor prognosis, heavy mutation burden, and..., Mao [/bib_ref] [bib_ref] Impact of tumor purity on immune gene expression and clustering analyses across..., Rhee [/bib_ref] [bib_ref] A prognostic nomogram combining immune-related gene signature and clinical factors predicts survival..., Song [/bib_ref] [bib_ref] Establishment and evaluation of a 6-gene survival risk assessment model related to..., Wang [/bib_ref] [bib_ref] Tumor purity as an underlying key factor in glioma, Zhang [/bib_ref] have revealed the relationship between tumor purity as well as immune score and patient survival. In TME, the percentage of tumor cells was called tumor purity. It has been reported that the poor prognosis of glioma 51 and colorectal cancer 47 was closely associated with low tumor purity. In contrast to the above results, Wang et al. [bib_ref] Establishment and evaluation of a 6-gene survival risk assessment model related to..., Wang [/bib_ref] and our previous study [bib_ref] A prognostic nomogram combining immune-related gene signature and clinical factors predicts survival..., Song [/bib_ref] observed that low tumor purity tended to suggest a better prognosis in LUAD patients. Our finding in this study was also in line with that of the above-stated studies. Tumor purity of different cancer patients presented sharply different indicative significance, seemingly highlighting the potential differences in the pattern of tumorigenesis and progression of different tumors. It is clear that tumor tissue contains not only tumor cells, but also non-tumor cells, such as stromal cells and immune cells. These non-tumor cells dilute the purity of the tumor and play an important role in the biological process of the tumor. As described by Zhang et al. [bib_ref] Tumor purity as an underlying key factor in glioma, Zhang [/bib_ref] , the purity of gliomas was associated with distinct patterns of genomic alterations. This intrinsic driving force contributed to differences in phenotype and survival among patients with different tumor purities of the same cancer type. It is well known that different cancers have significant heterogeneity, including genomic heterogeneity. And it is the existence of this heterogeneity that should be responsible for the suggestive significance of tumor purity in different cancers. In addition, an extremely important factor that should not be ignored is the local immune status of the tumor. Lower purity means stronger local immune status 51 , which is closely related to the composition, proportion and activation status of different immune cells. This further illustrates the remarkable complexity of the tumor microenvironment, and further research is needed to unravel its mysteries. Additionally, enriched biological processes and genomic alterations were depicted in different IRGS subgroups. Some critical cellular pathways indicating vital processes, such as regulation of DNA directed DNA polymerase activity, DNA replication and mismatch repair were enriched in the high-IRGS group. This might help explain why high-IRGS group presented more aggressive molecular changes than the low-IRGS group. Correlated with the differences of genomic alterations in different IRGS subgroups, it was not astonished to find that high-IRGS group had higher TMB, but further analysis revealed no dramatic correlation between patient prognosis and high/low TMB. However, the cross-talk between TMB, IRGS and patient prognosis revealed that patients with high TMB combined with high IRGS embraced the worst prognosis, and patients with low TMB combined with low IRGS had the best prognosis. This forwardly confirmed the predictive efficacy of IRGS in risk stratification, and also highlighted the complexity of the potential link between tumor genomic alterations and prognosis.
The best strategy for personalized immunotherapy should be to seek effective biomarkers to predict sensitivity to drug therapy. Reliable biomarkers have not been met in clinical practice. Accumulated evidence [bib_ref] First-line nivolumab in stage IV or recurrent non-small-cell lung cancer, Carbone [/bib_ref] [bib_ref] Tumor mutational load predicts survival after immunotherapy across multiple cancer types, Samstein [/bib_ref] has demonstrated patients with high TMB status presented durable clinical responses to immunotherapy. In our study, the high-IRGS group presented higher TMB, and TIDE score was remarkably increased in the low-IRGS group. Additionally, chemotherapy has been widely used in the treatment of many malignancies, including LUAD. Here, we predicted the sensitivity of different IRGS subgroups to several commonly-used drugs. We found the IC50s of several commonly-used drugs were lower in the high-IRGS group. Thus, the above results indicated that high-IRGS patients might benefit more from immunotherapy and commonly prescribed agents (cisplatin, paclitaxel, docetaxel, doxorubicin, gemcitabine), and fully demonstrated the values of IRGS in predicting drug www.nature.com/scientificreports/ therapeutic responses. To sum up, these results might provide additional clues for individualized treatment for LUAD patients. Additionally, our study also explored the correlation between IRG signature gene expression and drug Z-scores, which would provide directional suggestions and preliminary basis for drug development of gene-targeted therapy in human cancer. In recent years, nomogram has been widly used in prognostic assessment of cancer patients [bib_ref] Nomograms in oncology: More than meets the eye, Balachandran [/bib_ref] [bib_ref] A nomogram for predicting the risk of invasive pulmonary adenocarcinoma for patients..., Jin [/bib_ref]. To further optimize the survival prediction of LUAD patients, we established a prognostic nomogram including TNM stage and IRGS. Importantly, the nomogram demonstrated reliable accuracy and robustness in predicting survival for www.nature.com/scientificreports/ LUAD. It could therefore help clinicians accurately determine the prognosis of patients and develop individualized treatment regimens. There remained some limitations in this study, and the nature of retrospective research was an inevitable question. While as many data sets as possible were included for rigorous model building and validation, as well as 'Combat' approach to reduce batch effects, we were still unable to fully resolve the sampling deviation caused by cross-platform integration. In addition, although this study used a large number of retrospective data sets for risk stratification and efficacy prediction, there was still a lack of appropriate LUAD data sets based on immunotherapy regimen to verify the predictive robustness of the IRGS, so as to further strengthen our conclusion.
Overall, the present study identified and validated a novel IRG signature for risk stratification and efficacy prediction in over 2000 LUAD samples, described the multi-dimensional characterization of 11 IRG signature genes in LUAD, and emphasized the essential roles of IRG signature genes in shaping the complexity of TME in LUAD. More broadly, assessing the IRGS of LUAD would help to enhance our perception of immune cell infiltrating characteristics and provide important insights into the efficacy of drug therapy (immunotherapy, and chemotherapy). To make a long story short, our study might inform important treatment strategies, finally promoting the individualized management of LUAD patients.
## Data availability
This study obtained open data from the GEO database (https:// www. ncbi. nlm. nih. gov/ geo/) and TCGA (https:// cance rgeno me. nih. gov/). The data sets generated and analyzed during the present study are available from the corresponding author on reasonable request. www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.
[fig] Figure 1: Schematic diagram of the study design. [/fig]
[fig] Figure 2: Controduction and verification of an IRG signature for LUAD patients. (a) Common genes of the prognostic-related genes in the meta-GEO and TCGA cohorts. (b,c) The Lasso regression analysis of the common genes to avoid the model overfitting. (d) Multivariate Cox analysis (stepwise regression) of the filtered genes in training dataset (meta-GEO cohort). (e) Kaplan-Meier survival curves revealed the OS differences between high-and low-IRGS groups in the meta-GEO cohort. (f) The risk score and (g) OS status distribution of the IRG signature in the meta-GEO cohort. (h) The ROC curve analysis of the IRG signature for predicting OS in the TCGA cohort. (i) The ROC curve analysis of the siganture constructed in the study of Shi et al. for predicting OS in the TCGA cohort. (j) The ROC curve analysis of the siganture constructed in the study of Yi et al. for predicting OS in the TCGA cohort. (k) The ROC curve analysis of the siganture constructed in the study of Zhang et al. for predicting OS in the TCGA cohort. Scientific Reports | (2022) 12:10110 | https://doi.org/10.1038/s41598-022-14323-6 [/fig]
[fig] Figure 3: The landscape of genetic alterations of IRG signature genes in LUAD. (a,b) The KEGG enrichment analysis revealed the biological processes enriched for the IRG signature genes-related TFs. The KEGG database access link is as follows: https:// www. kegg. jp/ kegg/ kegg1. html. (c) The CNV mutational frequency of the 11 IRG signature genes was prevalent. The deletion frequency, green dot; The amplification frequency, red dot. (d) The CNV alteration positions of these 11 genes on the chromosome. (e) 56 of the 561 LUAD patients experienced genetic alterations of the 11 genes, with a frequency of 9.98%, primarily including missense mutations, nonsense mutations and multi hit. [/fig]
[fig] Figure 4: Enriched biological pathways and tumor immune microenvironment characteristics in different IRGS subgroups. (a) The GSVA revealed eight co-enriched pathways obtained from both cohorts (meta-GEO and TCGA cohorts). The heatmap was applied to visualize the enriched biological pathways, and red indicated the activated pathways and blue indicated the inhibited pathways. (b) GSEA enrichment analysis based on KEGG in meta-GEO cohort. (c) GSEA enrichment analysis based on KEGG in TCGA cohort. (d,e) Comparison of immune score and tumor purity in the high-and low-IRGS groups. (f) Comparison of tumor immune infiltrating cells based on ssGSEA algorithm in the high-and low-IRGS groups. Red indicated the high infiltrating levels of immune cells and blue indicated the low infiltrating levels. ESTIMATE score, stroma score, immune score, tumor purity and subtype are shown in annotations above. (g) Correlation between the abundance of tumor immune cells infiltrating using ssGSEA algorithm and IRG signature gene expression as well as IRGS (*p < 0.05; **p < 0.01; ***p < 0.001). Scientific Reports | (2022) 12:10110 | https://doi.org/10.1038/s41598-022-14323-6 [/fig]
[fig] Figure 5: Comparison of the genetic characteristics between high-and low-IRGS groups and prediction of immune efficacy. (a,b) Somatic mutation characteristics in the high (a) and low (b) IRGS groups based on the TCGA cohort. (c) The Kaplan-Meier survival curves revealed the survival differences in patients with high and low TMB. (d) The cross-talk among TMB, IRGS and patient survival. (e) The TMB status of the high-and low-IRGS groups were analyzed and plotted. (f) The TIDE scores of the high-and low-IRGS groups were analyzed and plotted. Scientific Reports | (2022) 12:10110 | https://doi.org/10.1038/s41598-022-14323-6 [/fig]
[fig] Figure 6: Drug sensitivities comparison between different IRGS subgroups. The estimated IC50s of cisplatin, paclitaxel, docetaxel, doxorubicin, and gemcitabine in the high-and low-IRGS groups in the meta-GEO cohort (a) and TCGA cohort (b). [/fig]
[fig] Figure 7: Construction and validation of a nomogram to optimize survival prediction. (a-d) IRGS was an independent prognostic predictor by univariate and multivariate Cox analyses (a) univariate Cox analysis in meta-GEO cohort; (b) multivariate Cox analysis in meta-GEO cohort; (c) univariate Cox analysis in TCGA cohort; (d) multivariate Cox analysis in TCGA cohort.). (e,f) ROC curves for age, gender, stage and IRGS in meta-GEO cohort (e) and TCGA cohort (f). (g) Nomogram based on IRGS and TNM stage. (h,i) ROC curves for the nomogram in meta-GEO cohort (h) and TCGA cohort (i). (j,k) Calibration curves of the nomogram for predicting 1, 3 and 5-year OS probability in meta-GEO cohort (j) and TCGA cohort (k). Scientific Reports | (2022) 12:10110 | https://doi.org/10.1038/s41598-022-14323-6 [/fig]
[table] Table 1: Basic clinicopathologic features of LUAD patients in the two cohorts. [/table]
|
A 21st century view of infection control in everyday settings: Moving from the Germ Theory of Disease to the Microbial Theory of Health
# Introduction
Due to significant advances in microbiome science over the past 2 decades, we are at the brink of a paradigm shift regarding the role of microbes in disease and health, from the Germ Theory of Disease to the Microbial Theory of Health. [bib_ref] The urgent need for microbiology literacy in society, Timmis [/bib_ref] This shift will necessitate a change in the approaches that we take to design targeted infection control. In particular, we will need to leverage our knowledge of the microbiome when attempting to reduce the risk posed by infectious agents through use of targeted hygiene, and by fostering/balancing exposure to naturally diverse microbial communities. Interestingly, health-care providers have started shifting their own emphasis in this direction by promoting critical care microbiome research/applications that treat dysbiosis in intensive care patients, an ambitious but encouraging goal. [bib_ref] Dysbiosis in the intensive care unit: microbiome science coming to the bedside, Kitsios [/bib_ref] Recently, a more pragmatic use of antibiotics for treatment/ prevention of infectious diseases has also been suggested by Rook and coworkers. [bib_ref] Evolution, humanmicrobe interactions, and life history plasticity, Rook [/bib_ref] The revised viewpoints mentioned here are likely to have a profound effect on hygiene policies. [bib_ref] The hygiene hypothesis: identifying microbial friends and protecting against microbial enemies, Bloomfield [/bib_ref] On the one hand, communitybased infections-including respiratory, gastrointestinal, and skin infections-continue to exert a heavy toll on human health and prosperity. [bib_ref] Community-based infections and the potential role of common touch surfaces as vectors..., Scott [/bib_ref] We use the term "community" in a broad sense to include both home environments as well as school, workplace, and even recreational settings. This problem is exacerbated by the aging of the population as a whole, and the associated increase in percentage (now »20%) of immunocompromised individuals living in the community. The latter are often cared for at home.Contrary to optimistic predictions made during the mid-20th century, [bib_ref] From Pasteur to genomics: progress and challenges in infectious diseases, Rappuoli [/bib_ref] infectious diseases clearly have not been eradicated. Rather, new infectious agents continue to emerge and/or reemerge globally. These have included emerging antibiotic-resistant pathogens (eg, methicillin-resistant Staphylococcus aureus and carbapenem-resistant Enterobacteriaceae). It is hard to overstate the risk associated with the emergence of such multidrug-resistant pathogens. The government of the United Kingdom has referred to this issue as a "postantibiotic apocalypse," which threatens to kill 10 million people globally by 2050.There also currently is much concern about the rapid rise in allergies (especially asthma and food allergies) and other chronic inflammatory diseases in the population. The Hygiene Hypothesis proposed by Strachan 10 postulated that a lower incidence of early childhood infection (predominantly in first-world countries) might explain the rapid rise of allergic diseases during the 20th century. As discussed in a 2016 review by Bloomfield et al,our current understanding of host-microbiome interactions and immune dysfunction suggests that increases in chronic inflammatory diseases are, instead, the combined result of lifestyle, nutrition, medical, and public health (hygiene and sanitation) changes. These changes are thought to have deprived humans of exposure to potentially beneficial microbial agents (described as Old Friends [OF]), particularly in early life.These OF microbes are not pathogenic, as argued by Strachan, 10 but rather include nonharmful diverse microbial species that inhabit the human gastrointestinal tract and our natural environment. While the identity of the most important OF microbes may not be clear, our attitudes towards such nonpathogenic microbes must change. Thus, the concept of age-appropriate and health-appropriate hygiene practices for home and everyday life has emerged.Such practices include age-appropriate vaccination and exposure to nonharmful microbes that beneficially prime the developing immune system. [bib_ref] Ontogeny of early life immunity, Dowling [/bib_ref] This strategy has the goal of balancing targeted hygiene with maintaining the natural diversity of the human and indoor microbiomes.
## From the hygiene hypothesis to the microbial theory of health
The Hygiene Hypothesis of Strachan 10 has resulted in the inherently dangerous concept of our being too clean, an idea that has persisted in the media and in the minds of the public. More recently, it has been argued that the Hygiene Hypothesis is flawed, but despite this, broad acceptance of the theory had been encouraged by a phenomenon known as citation bias. Namely, it was found that publications supportive of the Hygiene Hypothesis were cited more often than nonsupportive publications. [bib_ref] Selective citation in the literature on the hygiene hypothesis: a citation analysis..., Duyx [/bib_ref] We believe it is time to restore public understanding of hygiene, and specifically targeted hygiene, as a tool for preventing transmission of pathogens (breaking the chain of infection) and, consequently, transmission of infectious diseases.This is consistent with efforts of the infection control communities and public health agencies worldwide toward emphasizing the necessity of basic hygiene practices, at both the individual and community levels, for infection control. Most importantly, hygiene is now being seen as a key component of strategies intended to tackle the global problem of antibiotic resistance.It is hoped that, by reducing the level of infectious agent exposure, fewer people will need to seek antibiotic treatment−thereby limiting the selective pressure for generating antibiotic-resistant genes and the associated antibiotic-resistant strains of pathogens. [bib_ref] Impact of a probiotic-based hospital sanitation on antimicrobial resistance and HAI-associated antimicrobial..., Caselli [/bib_ref] [bib_ref] Fighting AMR in the healthcare environment: microbiome-based sanitation approaches and monitoring tools, D'accolti [/bib_ref] [bib_ref] Global survey of antibiotic resistance genes in air, Li [/bib_ref]
## Pathogen transmission and the cycle of infection
Before considering the concept of targeted hygiene, we need to identify the highest risk factors for pathogen transmission. Globally, the home captures a large cross-section of the human population in terms of age, health, nutritional status, and susceptibility to infectious agents. The home is therefore representative of many other community settings in terms of the necessity for hygiene practices. In fact, a constant dynamic exists between the home and other community settings (eg, day care, work, school, travel, leisure, healthcare, etc.) in terms of the dissemination of infectious agents from infected individuals, contaminated food, and domestic animals to surfaces and, via the intermediacy of human hands, through the entire cycle of reinfection.Day care settings represent an especially high risk, as young children are immunologically immature and exhibit behaviors that actually encourage transmission of infectious agents (eg, poor personal hygiene and mouthing of objects).
In determining the role that environmental surfaces play in the transmission of infectious agents, it is important first to develop a working definition of surface contamination. The chain of events leading to the occurrence and spread of infectious agents must then be considered. The prudent use of effective targeted hygienic approaches as a possible means of reducing the burden of infectious agents is described, keeping in mind that decontamination of all environmental surfaces on a continuous basis may not be necessary in all instances.
To define the risk associated with surface contamination, we must consider the types of surfaces that are most likely to become contaminated during activities of daily living [fig_ref] Fig 1: Examples of surfaces in the home ordered by risk for pathogen transmission [/fig_ref].
Pathogens travel via well-defined routes from an infected source to a new host. [bib_ref] The global war against intestinal parasites − should we use a holistic..., Alum [/bib_ref] [bib_ref] Norwalk virus shedding after experimental human infection, Atmar [/bib_ref] [bib_ref] Logistic growth of a surface contamination network and its role in disease..., Lei [/bib_ref] Numerous sampling studies have recorded the presence of both pathogenic bacteria, fungi, parasites, and viruses, as well as nonpathogens, on environmental surfaces in home and community settings. [bib_ref] Community-based infections and the potential role of common touch surfaces as vectors..., Scott [/bib_ref] Both laboratory and field studies have evaluated the rates of transfer of pathogens via hands and contaminated surfaces, as reviewed by Bloomfield et al.These studies demonstrate that the surfaces with the highest risk of transmitting pathogens, and which are therefore the critical control points in the transmission of infection, are the hands and common-touch surfaces, food-contact surfaces, and the cleaning utensils used on these surfaces, as shown in [fig_ref] Fig 1: Examples of surfaces in the home ordered by risk for pathogen transmission [/fig_ref] These high-risk environmental surfaces may serve as reservoirs for infectious agents deposited following shedding from humans or domestic animals as aerosols, or infectious agents arising from contaminated raw meats, fruits, and vegetables, as well as pathogen-contaminated air and water. Common-touch surfaces [fig_ref] Fig 2: High-risk common-touch surfaces for transmission of pathogens in the home or outside... [/fig_ref] such as door knobs, toilet flush handles, faucet handles, remote control devices, shared desks and other furniture (ie, in schools and offices), digital devices, and light switches represent an especially high risk of dissemination of these infectious agents throughout indoor spaces. Infectious agents have been found to survive on indoor environmental surfaces for extended periods of time, in many instances remaining for days to months at populations high enough to initiate infection of a new host. The duration of persistence is determined by the characteristics of the pathogen and of the surface itself, and other environmental factors, such as temperature and humidity and the organic matrix associated with the pathogens. The health implications of such persisting pathogen loads are dependent, to some extent, on the human infectious doses for the microbes, which can range from 1 to 10 infectious units for some hemorrhagic fever viruses [bib_ref] Clinical recognition and management of patients exposed to biological warfare agents, Franz [/bib_ref] (eg, Ebola virus) to thousands of infectious units for Staphylococcus aureus.The cycle of infection and reinfection involves dissemination, primarily through the intermediacy of the hand, from common-touch surfaces to new surfaces and/or other hosts. This cycle includes: (1) pathogen release from an infected source; (2) contamination of common-touch surfaces by released pathogens; (3) persistence of the pathogen on the contaminated surface; (4) transfer of pathogens to secondary common touch surfaces; (5) pathogen transfer to a new host; and (6) infection of the new host with consequent spread of the associated disease. This cycle may be interrupted through the timely use of effective targeted hygiene practices and the judicial use of effective microbicides. The global outbreak of SARS-CoV-2 and the associated disease COVID-19, emerging late in 2019, is a good example of the need for microbicidal agents to interrupt the cycle of infection. COVID-19 is primarily a lower respiratory syndrome, with some enteric and multisystem impacts, therefore SARS-CoV-2 is disseminated through the very same steps discussed above. Pathogens such as SARS-CoV-2 must be targeted for decontamination using the various microbicidal agents having adequate efficacy (see section below on microbicidal efficacy).
The points to be considered in developing targeted surface decontamination practices include: (1) the probability of significant pathogen contamination at the targeted high-risk (common-touch) surface under consideration; (2) the types of pathogens that are most likely to survive on surfaces and the time periods over which these might remain infectious at levels in excess of the minimum human infectious dose; (3) the likelihood of pathogen transfer from the contaminated surface to human hands and to other surfaces and hosts; and (4) the susceptibility of the new hosts to acquiring infection.
Targeting those surfaces at high risk for pathogen transmission/ acquisition by hosts, and applying appropriate decontamination practices, form the basis for an evidence-based hygiene policy known as targeted hygiene. The historical approach to pathogen reduction on environmental surfaces has involved attempts to indiscriminately reduce microbes naturally present in those areas without regard for whether these represent low-or high-risk surfaces. For instance, advertisements in the popular media have implied that, for maintaining health, all germs must be eradicated from our environments. Such an impractical approach has not accounted for the possible beneficial impact of nonpathogenic (OF) microorganisms comprising a naturally diverse vs dysbiotic indoor microbiome. We still have much to learn about such beneficial OF microbes. Targeted hygiene, on the other hand, is intended to manage the natural microbial diversity 24 of environmental surfaces as well as the human microbiome,incorporating the novel concept of bygiene or bi-directional hygiene. [bib_ref] Bygiene: the new paradigm of bidirectional hygiene, Al-Ghalith [/bib_ref] How is this to be accomplished?
## Bidirectional hygiene (bygiene)
The use of hygiene practices, such as surface decontamination or hand antisepsis, to reduce pathogen burden on a surface must be balanced against the need to maintain the natural microbial diversity. This is the essence of bidirectional hygiene or bygiene, a concept introduced by Al-Ghalith and Knights. [bib_ref] Bygiene: the new paradigm of bidirectional hygiene, Al-Ghalith [/bib_ref] This concept is similar to that expressed in the phrase "targeted hygiene is smart hygiene."The bygiene approach [fig_ref] Fig 3: Bidirectional hygiene [/fig_ref] is designed to reduce the risk of infection and therefore the need for antibiotics, while maintaining exposure to beneficial (OF) microbes. This is in distinct contrast to the Hygiene Hypothesis, which maintains that exposure to microbes (including some pathogens) is necessary, and certain health issues have arisen due to our overemphasis on sanitation. Another way of properly considering hygiene has been formulated by Vandegrift et al. [bib_ref] Cleanliness in context: reconciling hygiene with a modern microbial perspective, Vandegrift [/bib_ref] as "those actions and practices that reduce the spread or transmission of pathogenic microorganisms, and thus reduce the incidence of disease."
There are at least 2 approaches to pathogen reduction that are consistent with the principles of bygiene. The first of these is embodied in the term targeted hygiene. Targeted hygiene takes into account the importance of high-risk surfaces and situations [fig_ref] Fig 2: High-risk common-touch surfaces for transmission of pathogens in the home or outside... [/fig_ref] in the transmission of infectious diseases, and emphasizes the use of either hygienic cleaning of hands with soap or detergent and rinsing and the use of broad-spectrum microbicides on such high-risk surfaces selectively.
It follows that high-risk surfaces/situations should be addressed through cleaning and microbicide application, while such interventions do not need to be used for low-risk surfaces.
Another alternative to microbicidal formulations for targeted infection control is the use of smart surfaces 27,28 for high-risk activities such as food preparation. Application have been described for mitigation of the formation of biofilms [bib_ref] Targeting microbial biofilms: current and prospective therapeutic strategies, Koo [/bib_ref] and for the use of copper surfaces for high risk surfaces (eg, door knobs). It has been reported, however, that microorganisms may acquire a variety of defense mechanisms to such microbicidal metals. [bib_ref] Microbial resistance to metals in the environment, Bruins [/bib_ref] Knowledge of the impacts of disruption of microbiomes on health also can inform the use of microbicides. For instance, recent studies of the oral microbiome have suggested that altering the natural diversity of microorganisms (dysbiosis) may lead to periodontal disease. [bib_ref] The oral microbiome and the immunobiology of periodontal disease and caries, Costalonga [/bib_ref] This argues for treatments aimed at tailoring the oral microbiome, rather than simply attempting to eliminate oral microbial populations altogether. Similarly, perturbations in the microbiome within the gastrointestinal tract have been proposed to lead to inflammatory diseases. [bib_ref] Exposure of the host-associated microbiome to nutrient-rich conditions may lead to dysbiosis..., Lachnit [/bib_ref] The human skin microbiome of healthy adults is relatively stable and resilient. [bib_ref] Temporal stability of the human skin microbiome, Oh [/bib_ref] However, there appears to be a much more highly diverse skin microbiome associated with isolated populations such as the Yanomami Amerindians of Venezuela, who have been devoid of contact with individuals from developed countries and to the hygiene practices of those countries. [bib_ref] The microbiome of uncontacted Amerindians, Clemente [/bib_ref] The comparative study of such isolated communities suggests that westernization and urbanization has led to a substantial reduction in exposure to the diversity of beneficial environmental microorganisms our ancestors likely were exposed to for millennia. Such exposure is thought to be essential for appropriate development of our immune systems and prevention of acquiring allergies. [bib_ref] Bygiene: the new paradigm of bidirectional hygiene, Al-Ghalith [/bib_ref] Susceptibility to disruption of the skin microbiome of individuals living in megacities has been reported by Kim et al. [bib_ref] Fragile skin microbiomes in megacities are assembled by a predominantly niche-based process, Kim [/bib_ref] Air pollutants present in certain megacities may also impact the outdoor microbiome, though the potential impacts have yet to be elucidated. Impacts of pollutants on the pharyngeal microbiome are being investigated, and indicate alterations in taxonomic composition. [bib_ref] High-level PM2.5/PM10 exposure is associated with alterations in the human pharyngeal microbiota..., Qin [/bib_ref] The skin microbiomes were shown to be impacted by factors such as environment and socioeconomic status. Outdoor environment physical activity may also contribute to the acquisition of the human microbiome via skin and airways.On the other hand, the composition of the indoor microbiome seems to be more transient than the human microbiome. [bib_ref] Longitudinal analysis of microbial interaction between humans and the indoor environment, Lax [/bib_ref] The modern home should not be considered a natural setting, and the composition of the indoor microbiome in such dwellings is not stable, being heavily influenced by the daily activities of the human, domestic animal, and plant inhabitants, as well as by external factors such as pollution and other chemical contributors to our exposome. [bib_ref] Factors shaping the human exposome in the built environment: opportunities for engineering..., Dai [/bib_ref] The types of shelters or habitats in which our ancestors lived in the distant past would have provided continuous exposure to external elements such as dirt, animals, plants, and natural water sources. Exposure to this natural outdoor microbiome has now been reduced in our modern dwellings. The consequent impact to health is thought to include an increased incidence of asthma and other manifestations of allergies. In support of this, children growing up on traditional Amish farms in contact with animals have been shown to be less prone to asthma and other types of allergic reactions. [bib_ref] Innate immunity and asthma risk in Amish and Hutterite farm children, Stein [/bib_ref] The reduced incidence of allergic reactivity in such children has been attributed to the inhalation of air containing bacterial endotoxin (lipopolysaccharide), which is thought to reduce the overall reactivity of the immune system. [bib_ref] Innate immunity and asthma risk in Amish and Hutterite farm children, Stein [/bib_ref] This is consistent with reports demonstrating reduction in development of house dust mite−related asthma induced by an allergic stimulus in mice chronically exposed to bacterial endotoxin. [bib_ref] Farm dust and endotoxin protect against allergy through A20 induction in lung..., Schuijs [/bib_ref] A beneficial indoor microbiome product of the future might make the indoors more like the outdoors, with respect to composition of the microbiome. As of this time, however, it is not clear exactly what constitutes a healthy indoor microbiome.Management of the microbiome, whether that of environment, the oral cavity, or the skin, may represent an important intervention for optimizing human health and appropriate development of the immune system. For instance, in the future there may become available products containing probiotics, beneficial environmental bacteria, or components of these bacteria capable of interacting with the developing immune system in a manner that reduces the types of allergic reactions now being seen in children. This type of product might also provide a constant exposure to such bacteria in order to keep the immune system functioning appropriately.
Probiotics have also been used to manage the indoor microbiome. The first example of such a product, the probiotic-based cleaning hygiene system 42 is now commercially available. Probiotic-based cleaning hygiene system is based on the use of nonpathogenic probiotic bacterial spores from the Bacillus genus, and has been evaluated in a hospital setting. It is claimed to allow nonpathogenic bacteria to outcompete pathogens, including antimicrobial-resistant microbes. [bib_ref] Fighting AMR in the healthcare environment: microbiome-based sanitation approaches and monitoring tools, D'accolti [/bib_ref] [bib_ref] Impact of a probiotic-based cleaning intervention on the microbiota ecosystem of the..., Caselli [/bib_ref] The mechanism of action is thought to include the inhibition of quorum-sensing molecules of certain pathogenic bacteria. [bib_ref] Impact of a probiotic-based hospital sanitation on antimicrobial resistance and HAI-associated antimicrobial..., Caselli [/bib_ref] [bib_ref] Fighting AMR in the healthcare environment: microbiome-based sanitation approaches and monitoring tools, D'accolti [/bib_ref] Additional proposed mechanisms include competitive antagonism of pathogen growth, [bib_ref] Impact of a probiotic-based cleaning intervention on the microbiota ecosystem of the..., Caselli [/bib_ref] and production of antibacterial compounds such as bacteriocins. [bib_ref] Fighting AMR in the healthcare environment: microbiome-based sanitation approaches and monitoring tools, D'accolti [/bib_ref] This cleaning system is not intended for use in surgical suites or other areas that need to be aseptic. While environmental microorganism-based cleaning systems are becoming available in the market, with few exceptions the efficacy of such systems have not been demonstrated with experimental rigor, appropriate field studies, or long term analysis of sustainability of any changes in the indoor microbiome and subsequent health outcomes in populations. An example of such a study might include the identification of the microbiome of a prototypic high-risk environment (bathroom, kitchen sink, etc.) prior to and after the application of a probiotic product. Future research should also be conducted to identify the best mix of probiotics; the appropriate or optimal delivery systems; the safety profiles/considerations; potential for causing or preventing infections; the potential impacts of age (particularly the benefit for young children, infants, newborns, as well as older children/adults); and the potential risks for the immunocompromised, aging, and malnourished populations.
# Conclusions
Shifting the paradigm from a Germ Theory of Disease toward a Microbial Theory of Health, wellness, and disease prevention should not be allowed to undermine the critical role that targeted personal and surface hygiene practices play in interrupting the dissemination of infectious agents. As discussed above, targeted hygiene focuses on high-risk activities and surfaces, and discourages use of broad-spectrum microbicides for lower-risk activities and surfaces. It is the responsibility of subject matter experts (infection control communities, public health agencies, environmental hygienists, and manufacturers of microbicides) to remain current on the advances being made in microbiome science and the impact of microbicides on these microbiomes. In particular, the impact of microbicide use on natural microbial flora, including OF as well as pathogenic species, should be assessed. It is critical to restore the public understanding of the basic principles of good hygiene practices and the importance of the concept of targeted hygiene as a means of minimizing the dissemination of infectious agents. There is an educational component of this as well: in devising best practices for informing the public in both developed and developing countries, it will be wise to draw on field studies 43 from low-and middle-income countries. These have demonstrated the critical importance of hygiene education in addition to the use of appropriate and targeted microbicidal "tools" to achieve the desired goal of limiting infectious agents transmission. In explaining targeted hygiene to the public of developed countries with reference to common-touch surfaces such as those outlined in [fig_ref] Fig 4: Fig 4 [/fig_ref] there are a number of key questions to consider:
1. What are the key targets for microbicidal product use? 2. When is the right time and place to act? 3. Who is most at risk in home and community settings? 4. Which targeted hygiene interventions can be employed?
Education of the public within developing countries may need to include more basic microbiological information:
5. The role of pathogens in infectious disease causation and spread. 6. Where are pathogens found and what are the routes of infection? 7. How can infrastructural deficiencies be accommodated while practicing the hygiene principles articulated within this paper? Infection control and prevention education should highlight the common-touch surfaces for practicing targeted hygiene. Practicing hygiene is particularly important during and after activities such as handling food or eating, handling raw foods (such as meat, poultry, fish, eggs, fruits, and vegetables), using the toilet, diapering, contacting blood or body fluids, touching contaminated porous or nonporous surfaces, dressing a wound or administering medications, touching animals, or performing outdoor activities such as gardening (Source: Scott et al 1 ).
Focusing on selected behaviors, such as hand washing with soap and clean water, is particularly impactful. Encouraging hand washing at appropriate times before and after high-risk activities [fig_ref] Fig 1: Examples of surfaces in the home ordered by risk for pathogen transmission [/fig_ref] can help to significantly reduce the risk of exposure to infectious agents.
It is hoped that the Microbial Theory of Health will usher in a new era in which pathogen reduction can be accomplished without indiscriminate elimination of potentially beneficial/naturally diverse microbes from the human and environmental microbiomes. Implementation of the principles discussed here will depend upon the leveraging of emerging microbiome sciences, development of alternative microbicides, and education of the public in the concept of targeted hygiene.
[fig] Fig 2: High-risk common-touch surfaces for transmission of pathogens in the home or outside of the home (school, workplace, recreational settings; from Scott et al 1 ; adapted from Alum et al 20 ). [/fig]
[fig] Fig 1: Examples of surfaces in the home ordered by risk for pathogen transmission (from Scott et al. 1 ; modified from Bloomfield et al 11 ) The red and yellow dots represent pathogenic and nonpathogenic microorganisms, respectively. [/fig]
[fig] Fig 3: Bidirectional hygiene (bygiene) approach. [/fig]
[fig] Fig 4: Fig 4. Infection control and prevention education should highlight the common-touch surfaces for practicing targeted hygiene. Practicing hygiene is particularly important during and after activities such as handling food or eating, handling raw foods (such as meat, poultry, fish, eggs, fruits, and vegetables), using the toilet, diapering, contacting blood or body fluids, touching contaminated porous or nonporous surfaces, dressing a wound or administering medications, touching animals, or performing outdoor activities such as gardening (Source: Scott et al 1 ). [/fig]
|
Safety of Glycoprotein IIb-IIIa Inhibitors Used in Stroke-Related Treatment: A Systematic Review and Meta-Analysis
Background: Endovascular therapy and intravenous thrombolysis with recombinant tissue plasminogen activator are the 2 most recommended treatments for acute ischemic stroke (AIS). Glycoprotein (GP) IIb-IIIa inhibitors are short-acting selective reversible antiplatelet agents that emerged as promising therapeutic agents for AIS about 10 years ago. Given the unclear safety profile and application coverage of GP inhibitors, we conducted this meta-analysis to explore the same. Methods: We used GP IIb-IIIa inhibitors, intracranial hemorrhage, and mortality as the key words on Medline, Web of Science, and the Embase databases. Randomized controlled trials, prospective literatures, and retrospective studies in English published between 1990 and 2020 were screened. The outcomes were relative risk (RR) of death and 90-day intracerebral hemorrhage (ICH). We pooled the results in 2 categories and conducted a subgroup analysis stratified by different drugs. The choice of the effects model depended on the value of I 2 . Results: In all, 3700 patients from 20 studies were included. No GP IIb-IIIa inhibitors were found to have a remarkable influence on the ICH rate. The RR values of symptomatic ICH for abciximab and eptifibatide were 4.26 (1.89, 9.59) and 0.17 (0.04, 0.69), respectively. Both tirofiban and abciximab could decrease the mortality rate within 90 days. Age > 70 years, National Institutes of Health Stroke Scale > 15, and overall dose > 10 mg are risk factors for ICH events with tirofiban usage. Thrombectomy combined with tirofiban was safe for arterial reocclusion prevention. Conclusions: In stroke-related treatment, administration of GP IIb-IIIa inhibitors could be safe, but care should be taken regarding drug species and doses. Abciximab can increase the risk of symptomatic intracranial hemorrhage. Tirofiban and eptifibatide can be considered safe in low doses. Suitable patients should be selected using strict criteria.
# Introduction
Stroke may occur in one of 6 people in their lifetime around the world, and around 5.8 million people die annually from stroke (http://world-stroke.org). About 70% of all strokes are ischemic in nature. Acute ischemic stroke (AIS) is defined as the sudden loss of blood flow to a brain area with resulting loss of neurologic function.Thus, early vessel recanalization has been the major target in AIS treatment.Over the past 2 decades, intravenous thrombolysis has been the main method to achieve vessel recanalization for ischemia-reperfusion.Therapy with intravenous recombinant tissue plasminogen activator (rt-PA) is the first-line option when contraindications are absent.Endovascular treatment proved to be effective in improving functional outcomes for selected patients with AISand is recommended by current clinical practice guideline.However, the application of intravenous rt-PA is limited by a narrow therapeutic time window and relatively poor revascularization rates,and endovascular therapy can cause injury to endothelial cells leading to activation of local platelet aggregation and subsequent early reocclusion.Demands for more effective and safe thrombolytic agents are required.
Glycoprotein (GP) IIb-IIIa inhibitors are short-acting, reversible, selective platelet antagonists that act by antagonizing the GP IIb-IIIa receptors on the platelet surface and block the final common pathway to platelet aggregation by preventing fibrinogen molecules between adjacent platelets from binding. Thus, GP IIb-IIIa inhibitors could favor endogenous thrombolysis by suppressing thrombus growth and preventing thrombus reformation through competitive inhibition with fibrinogen. Since being approved by the Food and Drug Administration (FDA) for treatment of acute coronary syndrome, GP IIb-IIIa inhibitors have been widely used in percutaneous coronary intervention including angioplasty, thrombectomy, and stent implantation.Additional data showed that GP antagonists have neuroprotective properties and could improve cerebral microcirculatory blood flow.Hence, researchers have attempted to explore the applicability of GP IIb-IIIa inhibitors in AIS treatment on the basis of existing research in ischemic heart disease.
Tirofiban, eptifibatide, and abciximab are 3 GP IIa-IIIb receptor antagonists approved by the US FDA.Yet, their safety in treating stroke and preventing reocclusion after intervention remains debatable. A trial on abciximab therapy for stroke-Abciximab Emergent Stroke Treatment Trial (AbESTT-II)-which enrolled patients with stroke within 6 hours from onset, was stopped ahead of schedule owing to a 6-fold increase in the symptomatic intracranial hemorrhage (SICH) rate.Kellert et al reported that tirofiban raised the risk for fatal intracranial hemorrhage (ICH) in patients with stroke after mechanical thrombectomy in a prospective nonrandomized study,while another recent study found that halfdose abciximab is safe and effective as intravascular therapy for acute stroke.Ciccone et al 14 last updated the Cochrane systematic review on GP antagonists in 2014 and concluded that GP antagonists not only did not reduce mortality but also significantly increased the risk of SICH. However, their study was based on only 4 randomized controlled trials (RCTs; 3 on abciximab and 1 on tirofiban). By contrast, our analysis addresses what was lacking in the Cochrane review, by discussing the efficacy of GP antagonists combined with mechanical thrombectomy in AIS treatment. Considering the fact that safety profile of GP antagonists is unclear in both clinical trials and retrospective studies, a more comprehensive systematic review is essential
The present meta-analysis aimed to present the respective ICH risk for 3 GP IIa-IIIb receptor antagonists. The initial analysis was conducted only on RCTs, followed by another one on all study types. For tirofiban, we reported its relationship with age, National Institute of Health stroke scale (NIHSS) at admission, and total dose through regression analysis. Overall, we intended to provide an overview on tirofiban, eptifibatide, and abciximab in AIS and discuss factors that might amplify the incidence of adverse events.
# Methods
## Search strategy and selection criteria
Published literature from January 1, 1990, through January 30, 2020, were independently retrieved by 2 authors on Medline, Web of Science, and the Embase databases. We screened literature relevant to GP IIb-IIIa inhibitors (Tirofiban, Eptifibatide, Abciximab, etc), stroke, and ICH. Two reviewers independently screened relevant articles through titles and abstracts after duplicate records were eliminated on endnote. A review of the full text of eligible articles was conducted. A third reviewer was consulted if disagreement occurred. A trial or study was included if: (1) it focused on patients with AIS; (2) patients in the experiment group were given abciximab, eptifibatide, or tirofiban; (3) patients in the control group were given placebo (saline) or aspirin, or it was an open-control trial; (4) it was an RCT or a prospective nonrandomized study or a retrospective control study;additional therapies such as thrombolytic therapy (eg, rt-PA), antiplatelet agents, nonsteroidal anti-inflammation drugs, and mechanical thrombectomy were applied in all study arms; (6) ICH, SICH, or death were reported within 90 days; and (7) drug administration was described with adequate details such as dose and infusion speed.
We use the Cochrane risk of bias assessment tool to assess the methodological quality of the selected RCTs and used the Newcastle-Ottawa score to assess the methodological quality of cohort studies. This systematic review reports Preferred Reporting Items for Systematic reviews and Meta-Analyses statement.
## Data extraction and risk outcomes
The following aspects were extracted from the included studies: (1) Basic demographic characteristics including sex, mean and median age, NIHSS at admission, number of cases in each group, and onset to treatment time; (2) Disease history: hypertension, hyperlipidemia, diabetes, atrial fibrillation; (3) Medication regimen: type of GP receptor inhibitor, total dose, infusion speed, combined thrombolysis, or interventional therapy schedule; and (4) Risk outcomes: caseload of any ICH, SICH, and death within 90 days in each group. Risk ratio (RR) of any ICH and SICH was calculated and analyzed separately. If cases of ICH or SICH in the experimental group was zero, the study was not considered for analysis of ICH or SICH.
## Subgroup analysis and metaregression analysis strategy
We evaluated the GP receptor inhibitor risk in 2 sets. One merely incorporated RCTs, and another included all type of studies including RCTs, cohort, and case-control studies. Regardless of heterogeneity, we first conducted a subgroup analysis with different GP receptor inhibitor types. More comprehensive analysis was performed if heterogeneity in the subgroup was large (I 2 75%). The total dose of tirofiban was divided into ">10 mg" and "10 mg"; patient age was divided as "<65 years," "65 to 70 years," and ">70 years"; NIHSS at admission was divided as "15" and ">15." Further, based on whether thrombectomy was performed, there are groups of "thrombectomy" and "no thrombectomy". Next, we conducted a metaregression analysis to further explore the effect of continuous variables such as age and NIHSS on outcomes.
# Statistical analysis
If the lower bound of the 95% CI was >1, we considered the GP receptor inhibitors to increase the risk for ICH and mortality. If the lower bound of the 95% CI was <1, we considered the GP receptor inhibitors as be a protective factor for ICH. Cochran Q test and I 2 test were used in our analysis to quantify the heterogeneity among trials. Cochran Q test was used to verify subgroup differences. If P .05, the between-subgroup differences were considered to be statistically significant. The random effect model was used for data analysis when I 2 > 75%; otherwise, the fixed effect model was used. We carried out the Z test to evaluate statistical significance in metaregression analysis and checked publication bias through funnel plots and Egger regression asymmetry test. All statistical analysis was performed using the R package (version 3.6.1).
# Results
## Characteristics of analyzed studies
Ultimately, 7 RCTs,prospective studies,and 2 retrospective studies 31,32 with 3700 patients in total were included. In all, 5 studies used abciximab, 2 used eptifibatide, and 13 used tirofiban. In total, 5 studies discussed abciximab, 3 discussed eptifibatide, 13 discussed tirofiban, and 1 discussed eptifibatide and abciximab.
Patients in 9 studies received interventional therapy incorporating carotid stenting, angioplasty, or thrombectomy. The whole process of literature selection is presented in a flowchart in. Results of assessing the methodological quality RCTs (by Cochrane risk of bias assessment tool) and cohort studies (by Newcastle-Ottawa score) are presented in, respectively.
## Risk ratio for any ich
## Subgroup analysis and metaregression analysis for tirofiban
As mentioned above, a more comprehensive subgroup analysis was conducted to identify potential risk factors for tirofiban use. Eventually, we found that while thrombectomy is not one of the factors, overall tirofiban dose >10 mg, age > 70 years, and NIHSS score > 15 all play important roles in the incidence of ICH events. Thrombectomy combined with tirofiban showed no risk for both ICH and SICH. When the overall tirofiban dose is >10 mg, the RR value of ICH was 1.46 (1.07-1.99), while that of SICH is 1.92 (1.02-3.62). When the initial NIHSS score was >15, the RR for ICH was 1.77 (1.14-2.73), implying an evident hazard, and the RR for SICH was 2.04 (0.88-4.75). In the metaregression model, we found a positive correlation between NIHSS and RR for SICH (P ¼ .001,, as well as between age and RR for ICH (P ¼ .0024,. Thus, we believe that tirofiban is safe in low doses for selected patients. However, older patients or those with high NIHSS score should be treated with caution.
## Publication bias
No significant publication bias was noted in the analysis by funnel plotsand Egger regression asymmetry test (P ¼ .253 for ICH, P ¼ .862 for SICH).
# Discussion
To assess the safety of combined treatment with GP IIb-IIIa inhibitors in stroke-related treatment, the ICH and mortality at 90 days were considered as the evaluation criteria. As SICH is a relatively fatal bleeding condition in ICH, short-term SICH incidence was also accepted as a separate evaluation standard. We chose 3 agents that are commonly used in clinical practice, namely tirofiban, eptifibatide, and abciximab; searched the relevant literature; pooled and compared results in category; and conducted further subgroup and regression analyses. Unlike us, Ciccone et al excluded 2 eptifibatide RCTs because they speculated that a lower dose of rt-PA in the experimental group than in the control group might result in bias.No RCT yet has proven that low dose rt-PA is safer than the standard dose. Two systematic reviews (Liu et aland Liu et alalso reported no significant difference in SICH incidence between low-dose rt-PA and standard-dose rt-PA. A recent prospective study showed that bleeding risk in the 0.6 mg/kg rt-PA group and 0.9 mg/kg rt-PA group was not significantly different.Based on their results, we overlooked the dose difference of rt-PA between the CLEAR and CLEAR-ER trials and included both these studies in our analysis. The Safety of Tirofiban in acute Ischemic Stroke (SaTIS) trial was also excluded in the Cochrane systematic review, because the maximum time from onset to inclusion in SaTIS was 22 hours.We speculated that with the same distribution in the experimental and control groups, the onset-to-inclusion time would not bias the final outcomes. Thus, we also included the SaTIS trial.
The American Heart Association published a new international guideline in 2019,in which the level of evidence of tirofiban and eptifibatide in AIS treatment is B-R and the class of recommendation is IIb, indicating that benefits for patients are generally greater than the inherent risks. According to our results, for younger patients with lower NIHSS scores, the ICH risk presented a tendency to decrease. Therefore, based on multiple studies, we thought our conclusion can be used as a reference for the development of guidelines and future clinical practice. Furthermore, the administration of abciximab was proposed to be potentially harmful in the guideline according to the results of the AbESTT trial. We will discuss later in the text when we discuss AbESTT trial.
Mechanical thrombectomy has been used for AIS treatment since 2005.Accordingly, concerns about postoperative arterial reocclusion were raised. Glycoprotein antagonists, with strong antiplatelet ability, were found superior to prevent arterial reocclusion. Owing to the lack of strong evidence provided by RCT, the Cochrane SR did not investigate its application in this area. Different from them, we included nonrandomized studies in this regard and carried out a preliminary discussion. The 2019 guideline suggested that the efficacy and safety of GP IIa-IIIb inhibitors combined with endovascular therapy are unclear. However, this conclusion was based on 3 retrospective single-arm studies (no control group) with an evidence level of C-LD (low evidence grade).Of 3 studies, 2 were about abciximab and 1 was about tirofiban and abciximab. Therefore, whether the level or the quantity of evidence was inadequate. Instead, we concluded that tirofiban is safe to use when combined with thrombectomy, especially at low doses.
Subgroup results showed that abciximab can remarkably increase a 4-fold risk for SICH but may reduce the 90-day mortality. Adam et al conducted the first clinical trial for abciximab with no occurrence of SICH. This likely ascribes to the low drug dose and small sample size (15 in each experimental group).In a large international trial (AbESTT II) conducted in 2008, Adams et al reported that the experiment group with abciximab presented a high proportion of hemorrhage, because of which the trial had to be discontinued. Although the superiority of abciximab in ischemic stroke was not demonstrated, the authors, based on their research in phase II, proposed a view that such hemorrhagic ratio is acceptable provided the ultimate outcome can be improved.In AbESTT, the experimental group had 8 cases of SICH out of 195 patients; however, the ratio increased to 22 of 221 in the AbESTT-II trial.In addition, 4 studies focused on carotid stent/angioplasty.In these studies, all control groups reported no ICH, while all abciximab groups reported ICH and 3 reported SICH. Our conclusions were consistent with their theory. However, the efficacy of abciximab needs more studies and analyses to be confirmed. In this article, we exclusively focused on the safety of drugs. Interestingly, the results of the AbESTT trial indicated that the therapeutic window of abciximab can extend to 24 hours,while AbESTT II in 2008 showed a reasonable and safe window within 6 hous from onset. 12 Therefore, the time window of administration may also act as a factor for safety. It is also worth noting that AbESTT, AbESTT-II, and the aforementioned 4 case-control studies administrated full dose in cardiac trials, so the application of low-dose abciximab still deserves further research.
Although eptifibatide exhibited no statistically significant impact on any ICH and death, the incidence of SICH was reduced after administration of eptifibatide. The CLEAR trial, the first randomized blind trial on safety of low-dose rt-PA combined with GP IIb-IIIa, showed encouraging results of this combination regimen, perhaps owing to the extremely small dose in CLEAR. However, a limitation of the CLEAR-trial is that the eptifibatide group had a much lower baseline of NIHSS (median: 12.0) than the control group.In addition, the CLEAR-ER trial highlighted concerns regarding systemic bleeding, given that cases of fatal, moderate, and mild bleeding were 1, 4, and 12, respectively, in the eptifibatide group but 0 in the standard therapy group.Again, the time window of eptifibatide in stroke merits further discussion.
Tirofiban is a highly selective, fast-acting, nonpeptide GP IIa-IIIb inhibitor with short halftime and debatable safety in ischemic stroke. Our subgroup results indicated that tirofiban does not influence ICH or SICH and can lower the 90-day mortality. However, a significant interstudy heterogeneity was detected while merging data. Consequently, further analysis on tirofiban was conducted. Eventually, we found that total dose >10 mg, age >70 years, and NIHSS >15 are risks for ICH events, and thrombectomy was not one of them. Based on these findings, we believe that tirofiban is safe to use but in low doses and is not suggested in either patients with high NIHSS score or older individuals. Kellert et al reported that the occurrence of fatal hemorrhage significantly increased for patients who received tirofiban after mechanical thrombectomy. To this point, we believe it could be attributed to the relatively high drug dose and severe baseline NIHSS score. In their study, the authors noted that the risk was mainly demonstrated in patients with anterior circulation stroke.Considering those patients who had lower NIHSS score and shorter time window at admission than those with posterior circulation stroke, which could minimize the events of fatal ICH, the results contradicted our assumption. This correlation between tirofiban-induced hemorrhage and lesion location suggested by Kellert et al still needs validation. Jeong and Jin in 2013 reported that they administrated tirofiban through the arterial route and in a lower dose than recommended to avoid surgical complications such as ICH. The results eventually proved that small dose administration is safe and rewarding.A larger clinical trial in 2011 by Siebler et al also reported the safety of tirofiban in ischemic stroke. They believed tirofiban would not carry an extra risk of ICH compared to standard care. Furthermore, patients treated with tirofiban were associated with lower mortality in the 5-month follow-up and higher survival rate at 3 months posttreatment.Pharmacologically speaking, with only a 2-hour halftime period, tirofiban has a lower risk of inducing continuous bleeding. Considering the benefits for cardiovascular disease, it is very much likely that tirofiban may be more advantageous to a certain group of people. Other than these, we also found that the risk of SICH was positively associated with the NIHSS score. A similar relationship exists between age and ICH. Thus, age and NIHSS score at admission influence postoperative ICH. This is consistent with the conclusions of Kellert et al,in that age and NIHSS score of patients with stroke are both independent predictors of poor prognosis.
A common side effect of all types of GP IIb-IIIa inhibitors is thrombocytopenia. Currently, there has not been clear evidence to show that this could result in poor outcomes. In the AbESTT II trial, the incidence of thrombocytopenia was relatively low and no adverse consequences were reported.Moreover, as reported in Qureshi et al, the platelet function of patients could distinctly recover within 4 hours after eptifibatide was discontinued.Torgano et al concluded in a study of 150 patients with AIS that tirofiban administered within 6 hours of symptom onset did not further escalate the risk for these patients.Another simultaneous trial, the SaTIS, investigated the safety of low-dose tirofiban within 22 hours of symptom onset. Authors of the SaTIS trial concluded that compared to standard care, tirofiban does not increase ICH risk.A recent study suggested that because it is difficult to avoid the use of abciximab during stenting in AIS, more balance is required between the antiplatelet effects and safety of a given drug.Qureshi et al concluded that eptifibatide could be safely used intravenously after low-dose intra-arterial thrombolysis.The CLEAR trial, which assessed the safety of combination therapy within 3 hours from symptom onset, proposed that a reduced dose of rt-PA with eptifibatide was reasonably safe to allow further exploration of the dose range. 18
# Conclusion
Our results provided convincing evidence that abciximab enhances the risk of SICH when used for AIS treatment and prevention of artery reocclusion. Further, eptifibatide and tirofiban are promising therapeutic agents at a reduced dose. However, patient age and NIHSS score as independent risk factors should be taken into consideration by clinicians. Further investigations on the safety and efficacy of low-dose abciximab should be performed as the evidence available thus far is inadequate.
## Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
# Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
## Orcid id
Genmao Cao https://orcid.org/0000-0003-2199-6903 |
Interferon therapy in patients with SARS, MERS, and COVID-19: A systematic review and meta-analysis of clinical studies
Concern regarding coronavirus (CoV) outbreaks has stayed relevant to global health in the last decades. Emerging COVID-19 infection, caused by the novel SARS-CoV2, is now a pandemic, bringing a substantial burden to human health. Interferon (IFN), combined with other antivirals and various treatments, has been used to treat and prevent MERS-CoV, SARS-CoV, and SARS-CoV2 infections. We aimed to assess the clinical efficacy of IFNbased treatments and combinational therapy with antivirals, corticosteroids, traditional medicine, and other treatments. Major healthcare databases and grey literature were investigated. A three-stage screening was utilized, and included studies were checked against the protocol eligibility criteria. Risk of bias assessment and data extraction were performed, followed by narrative data synthesis. Fifty-five distinct studies of SARS-CoV2, MERS-CoV, and SARS-CoV were spotted. Our narrative synthesis showed a possible benefit in the use of IFN. A good quality cohort showed lower CRP levels in Arbidol (ARB) + IFN group vs. IFN only group. Another study reported a significantly shorter chest X-ray (CXR) resolution in IFN-Alfacon-1 + corticosteroid group compared with the corticosteroid only group in SARS-CoV patients. In a COVID-19 trial, total adverse drug events (ADEs) were much lower in the Favipiravir (FPV) + IFN-α group compared with the LPV/RTV arm (P = 0.001). Also, nausea in patients receiving FPV + IFN-α regimen was significantly lower (P = 0.03). Quantitative analysis of mortality did not show a conclusive effect for IFN/RBV treatment in six moderately heterogeneous MERS-CoV studies (log OR = − 0.05, 95% CI: (− 0.71,0.62), I 2 = 44.71%). A meta-analysis of three COVID-19 studies did not show a conclusive nor meaningful relation between receiving IFN and COVID-19 severity (log OR = − 0.44, 95% CI: (− 1.13,0.25), I 2 = 31.42%). A lack of high-quality cohorts and controlled trials was observed. Evidence suggests the potential efficacy of several combination IFN therapies such as lower ADEs, quicker resolution of CXR, or a decrease in inflammatory cytokines; Still, these options must possibly be further explored before being recommended in public guidelines. For all major CoVs, our results may indicate a lack of a definitive effect of IFN treatment on mortality. We recommend such therapeutics be administered with extreme caution until further investigation uncovers high-quality evidence in favor of IFN or combination therapy with IFN.
## A r t i c l e i n f o
## Keywords:
COVID-19 IFN Interferon MERS-CoV Middle-east respiratory syndrome SARS-CoV SARS-CoV2 Severe acute respiratory syndrome A B S T R A C T Concern regarding coronavirus (CoV) outbreaks has stayed relevant to global health in the last decades. Emerging COVID-19 infection, caused by the novel SARS-CoV2, is now a pandemic, bringing a substantial burden to human health. Interferon (IFN), combined with other antivirals and various treatments, has been used to treat and prevent MERS-CoV, SARS-CoV, and SARS-CoV2 infections. We aimed to assess the clinical efficacy of IFNbased treatments and combinational therapy with antivirals, corticosteroids, traditional medicine, and other treatments. Major healthcare databases and grey literature were investigated. A three-stage screening was utilized, and included studies were checked against the protocol eligibility criteria. Risk of bias assessment and data extraction were performed, followed by narrative data synthesis. Fifty-five distinct studies of SARS-CoV2, MERS-CoV, and SARS-CoV were spotted. Our narrative synthesis showed a possible benefit in the use of IFN. A good quality cohort showed lower CRP levels in Arbidol (ARB) + IFN group vs. IFN only group. Another study reported a significantly shorter chest X-ray (CXR) resolution in IFN-Alfacon-1 + corticosteroid group compared with the corticosteroid only group in SARS-CoV patients. In a COVID-19 trial, total adverse drug events (ADEs) were much lower in the Favipiravir (FPV) + IFN-α group compared with the LPV/RTV arm (P = 0.001). Also, nausea in patients receiving FPV + IFN-α regimen was significantly lower (P = 0.03). Quantitative analysis of mortality did not show a conclusive effect for IFN/RBV treatment in six moderately heterogeneous MERS-CoV studies (log OR = − 0.05, 95% CI: (− 0.71,0.62), I 2 = 44.71%). A meta-analysis of three COVID-19 studies did not show a conclusive nor meaningful relation between receiving IFN and COVID-19 severity (log OR = − 0.44, 95% CI: (− 1.13,0.25), I 2 = 31.42%). A lack of high-quality cohorts and controlled trials was observed. Evidence suggests the potential efficacy of several combination IFN therapies such as lower ADEs, quicker resolution of CXR, or a decrease in inflammatory cytokines; Still, these options must possibly be further explored before being recommended in public guidelines. For all major CoVs, our results may indicate a lack of a definitive effect of IFN treatment on mortality. We recommend such therapeutics be administered with extreme caution until further investigation uncovers high-quality evidence in favor of IFN or combination therapy with IFN.
# Introduction
Coronaviruses (CoVs) are single-stranded, positive-sense, RNA containing, and enveloped viruses responsible for several major global outbreaks [bib_ref] Coronavirus infections: epidemiological, clinical and immunological features and hypotheses, Raoult [/bib_ref]. Global epidemics of atypical pneumonia were first caused by [bib_ref] THE SARS EPIDEMIC AND ITS AFTERMATH IN CHINA: A POLITICAL PERSPECTIVE, Heidarpour [/bib_ref] , and continued to affect the globe with MERS-CoV reappearing in South Korea in 2015 [bib_ref] MERS outbreak in Korea: hospital-to-hospital transmission, Ki [/bib_ref]. Recently, coronavirus disease 2019 (COVID-19), a disease caused by a novel variant of SARS-CoV known as SARS-CoV2, emerged in Wuhan, China. While showing a lower mortality rate (2.3%) compared to MERS-CoV (9.5%) and SARS-CoV1 (34.4%), the COVID-19 pandemic has raised significant concern. The concern is partly due to the high spreading potential of SARS-CoV2, which influences and causes mortality in a significantly larger population. The novel virus has an undetermined clinical presentation , as the recent evidence has suggested non-respiratory and asymptomatic presentations . Hence, the diversity in the presentations and hurdles in detecting the virussuggest the high importance of an effective onset-to-treat period regarding the treatment of COVID-19 patients .
Numerous novel efforts have been carried out in the fields of drug discovery, vaccine development , and repurposing of previously suggested candidates for SARS-and MERS-CoV infections. Indeed, researchers have evaluated pharmacologic options, comprising combination interferon (IFN) therapy, traditional medicine, corticosteroid therapy, and antivirals such as ribavirin (RBV), lopinavir (LPV), ritonavir (RTV), oseltamivir, and Remdesivir (REM). However, to date, such efforts have not brought forth adequate success. Nevertheless, . Role of IFNs and other innate immunity elements in multi-system CoV disease Among innate elements, inflammasomes, ILs, and IFNs are of high importance [bib_ref] Contribution of inflammasome complex in inflammatory-related eye disorders and its implications for..., Rasoulinejad [/bib_ref]. Innate immunity helps to prevent the spread of viruses and affects ACE2 − a major entry path for SARS-CoV2. Therefore, impairments in these elements may contribute to severe clinical disease. COVID-19 similar to its ancestors, may utilize ACE2, which can be mediated by IFN secretion. IFN may inhibit the replication chain. Retrograde synaptic pathway through which SARS-CoV2 infects the central nervous system (CNS) also involves ACE2. Here, dissemination of COVID-19 and its replication have been illustrated . Also, the activation of innate pathways that upregulate IFNs, inflammasome elements, and cytokines has been illustrated (blue, I-VII). Created with BioRender.com. several protocols of past curatives are being used for COVID-19 patients due to a lack of effective treatments or alternatives when extreme adverse drug events (ADE) are indicated. The innate immune system comprises inflammasomes [bib_ref] Contribution of inflammasome complex in inflammatory-related eye disorders and its implications for..., Rasoulinejad [/bib_ref] , cytokines, and IFNs which help to clear viral disease and provide multi-system immunological protection [bib_ref] The role of interferons in inflammation and inflammasome activation, Kopitar-Jerala [/bib_ref] [bib_ref] Pathophysiological roles of chronic low-grade inflammation mediators in polycystic ovary syndrome, Rostamtabar [/bib_ref]. It has been shown that SARS-CoV2 is sensitive to type I IFN therapy in human cell lines [bib_ref] Antiviral activities of type I interferons to SARS-CoV-2 infection, Mantlo [/bib_ref]. A strong association between low type I IFN production and COVID-19 severity has been reported [bib_ref] Autoantibodies against type I IFNs in patients with life-threatening COVID-19, Bastard [/bib_ref] [bib_ref] Host-viral infection maps reveal signatures of severe COVID-19 patients, Bost [/bib_ref] [bib_ref] Inborn errors of type I IFN immunity in patients with life-threatening COVID-19, Zhang [/bib_ref]. Nuclear factor-kappa light chain enhancer B (NF-kB) activation in the dendritic cells is crucial for large scale type I IFN production. In a study of COVID-19 by or 2 mutations required hospitalization [bib_ref] Coronavirus Disease 2019 in patients with inborn errors of immunity: an international..., Meyts [/bib_ref] , highlighting the functional role of IFNs. Administration with subcutaneous IFN β-1a has been shown to reduce morbidity in COVID-19 infected patients [bib_ref] A randomized clinical trial of the efficacy and safety of interferon β-1a..., Davoudi-Monfared [/bib_ref]. Lung infection in COVID-19 may evolve into systemic involvement. Also, IFNs specially IFN-α2b are capable of preventing lung abnormalities in such patients [bib_ref] Interferon-α2b treatment for COVID-19 is associated with improvements in lung abnormalities, Zhou [/bib_ref]. All of these statements emphasize the role of IFN therapy in severe acute CoVs disease. In addition to lungs, other organs like kidneys [bib_ref] Kidney involvement in COVID-19 and its treatments, Han [/bib_ref] , liver , and the brain [bib_ref] Evidence of the COVID-19 virus targeting the CNS: tissue distribution, host-virus interaction,..., Baig [/bib_ref] are also involved. A major entry pathway for SARS-CoV2 is angiotensin-converting enzyme 2 (ACE2), which is present in multiple systems throughout the body. Research has shown IFNs can significantly alter ACE2 profile. ACE2 is regarded as an interferon-stimulated gene (ISG) . Thus, interferon-induced alteration in ACE2 production may be crucial for liability to COVID-19 or its corresponding adverse outcomes [bib_ref] Interferons and viruses induce a novel truncated ACE2 isoform and not the..., Onabajo [/bib_ref]. Taken together, noteworthy for future research is that IFNs could play a crucial role in multi-organ involvement prevention of patients with COVID-19. The probable role of IFNs in the multi-organ involvement situation has been enlaced in . Intriguingly, despite contradicting in vitro and in vivo studies and the absence of sufficient high-quality randomized controlled trials (RCTs) for the use of IFNs to treat SARS-CoV2, and that several studies indicate that it is not suggested for COVID-19 therapy, antivirals such as RBV have been commonly used in combination with IFN during epidemics [bib_ref] Clinical outcomes of current medical approaches for Middle East respiratory syndrome: a..., Morra [/bib_ref] [bib_ref] Broad-spectrum coronavirus antiviral drug discovery, Totura [/bib_ref]. Also, combination therapies in RCTs have been undertaken for the novel CoV (e.g., . Surprisingly, current Chinese guidelines include IFNs as an alternative for combination therapy . Such efforts have led to rapidly increasing clinical data on IFN administration for COVID-19 cases. Notably, CoV outbreaks share remarkable similarities, and hence, investigating the experience with the previous spreading of SARS-and MERS-CoVs may assist in discovering an effective treatment or help determine if a candidate should be removed from treatment protocols . To our knowledge, there have not been any updated systematic reviews of the literature shedding light on the effectiveness of IFN therapy with the past outbreaks in mind. In the present systematic review and quantitative analysis of the evidence, we describe the characteristics of hospitalized cases with MERS-CoV, SARS-CoV1, and SARS-CoV2 patients and assess important treatment outcomes and ADEs of various combinational and non-combinational IFN treatments.
# Materials and methods
The present systematic review has been conducted compatible with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statements. We designed the protocol to determine our scope, inclusion and exclusion criteria, and outcomes of evaluated studies. The protocol for the present study is provided in further detail in Supplementary Material.
The present study aimed to assess the outcomes of IFN treatments or IFN combination therapies in hospitalized patients infected with MERS-CoV, SARS-CoV, and SARS-CoV2. Comparator therapies comprised placebo, sham therapy, and no intervention. Moreover, researches involving no comparator group were included. Outcome measures were selected according to our protocol. We assessed the efficacy of IFN therapies with or without combination with other pharmacotherapy options. As efficacy comprises numerous parameters, we took account many clinical outcomes, including mortality, discharge, CXR, hospital durations, inflammatory state, ADEs, and disease severity. Due to limited data and the emerging situation of the COVID-19 pandemic, both published and unpublished works were included. No restrictions were considered for the date of publication and language. Our classification for treatment regimens was in line with World Health Organization (WHO) Guidelines. For SARS-CoV, these groups included RBV, LPV/RTV (Kaletra), corticosteroids, IFN, convalescent plasma, and intravenous immunoglobulin (IVIG), which have been previously utilized in similar studies [bib_ref] SARS: systematic review of treatment effects, Stockman [/bib_ref]. MERS-CoV treatments included IFNs, RBV, LPV/RTV, polyclonal anti-MERS-CoV human immunoglobulins, humanized murine anti-S monoclonal antibodies, nucleoside viral RNA polymerase inhibitors (e.g., REM), peptide inhibitors (e.g., HR2P-M2), and mycophenolate mofetil (MMF). Moreover, possible SARS-CoV2 interventions according to WHO and Centers for Disease Control and Prevention (CDC) Guidelines comprised hydroxychloroquine, chloroquine, REM, oseltamivir, tocilizumab, LPV/RTV, IFN-β, convalescent plasma, IVIG, and corticosteroids. Treatments were selected if used in combination with IFN. We included human studies designed as randomized and non-randomized clinical trials, observational clinical studies (e.g., retrospective and prospective cohorts), case reports, and case series.
## Search strategy and study selection
In May 2020, five reviewers (K.S., S.Y., E.H., M.B., M.G.) performed a systematic search. PubMed, Scopus, Cochrane's library, Web of Science (WoS), Global Index Medicus (WHO library), Google Scholar, and Scopus were searched for articles. An additional search was done for unpublished work (e.g., from BioRxiv, MedRxiv), and Reference lists were also screened (grey literature). Unpublished articles were checked, and updated with the published version of each, if available. For all articles, corrections and retractions were also checked. For Google Scholar, the following search strings were developed with the help of a skilled librarian: ("interferon" OR "IFN") AND ("Middle East respiratory syndrome" OR "Middle Eastern Respiratory Syndrome" OR "MERS-CoV" OR "Severe Acute Respiratory Syndrome" OR "SARS-CoV" OR "COVID- case-control analytic research, preferably from more than one center or study group; Level II-3: Evidence acquired from multiple time series, both with or without the intervention. Dramatic outcomes in uncontrolled trials may also be taken as such kind of evidence;
Level III: Opinions of validated authorities, in accordance with clinical experience, descriptive research, or reports of expert groups.
## Data synthesis
The protocol details methods used for narrative and quantitative syntheses (Supplementary Material) (College Station).
## Risk of bias across studies
The tool developed by the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) Working Group (www.gr adeworkinggroup.org) was selected for evaluation of bias across studies eligible for meta-analysis. GRADE enables consistent evaluation of the certainty of evidence. It also allows recommendations based on high-quality observational studies. GRADE initially ranks the evidencebased on study design. Studies are then promoted or downgraded according to criteria, including the risk of bias, indirectness, and imprecision (GradePro, 2020).
# Results
## Study selection
Our search strategy produced 2693 results from all six databases. Moreover, in addition to 42 initially included articles, our updated electronic search results identified 20 relevant results. An additional search yielded seven results. For five studies, full-text could not be obtained [fig_ref] 2: Hung IF-N, Lung K-C, Tso EY-K, Liu R, Chung TW-H, Chu M-Y,... [/fig_ref] [bib_ref] Clinical investigation of outbreak of nosocomial severe acute respiratory syndrome, Gao [/bib_ref] [bib_ref] A field trial for evaluating the safety of recombinant human interferon alpha-2b..., Qing [/bib_ref] [bib_ref] Clinical features of 96 patients with severe acute respiratory syndrome from a..., Wu [/bib_ref] [bib_ref] A hospital outbreak of severe acute respiratory syndrome in Guangzhou, Wu [/bib_ref]. Due to a lack of multilingual collaborators, we used online translators for foreign studies. All foreign articles that were sufficiently translatable via online translators were included.
## Study characteristics
Fifty-five distinct publications were included in line with our eligibility criteria. Classified by aetiology, there were 29 eligible clinical studies for SARS-CoV2 [bib_ref] COVID-19 in solid organ transplant recipients: a single-center case series from Spain, Fernández-Ruiz [/bib_ref] [bib_ref] Kidney impairment is associated with in-hospital death of COVID-19 patients, Cheng [/bib_ref] [bib_ref] Clinical features of 85 fatal cases of COVID-19 from Wuhan: a retrospective..., Du [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a..., Pan [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] [bib_ref] The correlation between viral clearance and biochemical outcomes of 94 COVID-19 infected..., Yuan [/bib_ref] [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref] , 26 studies for MERS-CoV (24 distinct reports) (Al [bib_ref] Characteristics and outcomes of Middle East respiratory syndrome coronavirus patients admitted to..., Al-Hameed [/bib_ref] [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] Ribavirin and interferon therapy in patients infected with the Middle East respiratory..., Al-Tawfiq [/bib_ref] [bib_ref] Clinical predictors of mortality of Middle East Respiratory Syndrome Coronavirus (MERS-CoV) infection:..., Alfaraj [/bib_ref] [bib_ref] Effect of Ribavirin and Interferon on the Outcome of Critically Ill Patients..., Arabi [/bib_ref] [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref] [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Absence of neutralizing activity in serum 1 year after successful treatment with..., Choi [/bib_ref] [bib_ref] Mortality rate of ICU patients with the Middle East respiratory syndrome-coronavirus infection..., Garout [/bib_ref] [bib_ref] Clinical outcomes among hospital patients with Middle East respiratory syndrome coronavirus (MERS-CoV)..., Habib [/bib_ref] [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Combination therapy with lopinavir/ritonavir, ribavirin and interferon-alpha for Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] IFN-alpha2a or IFN-beta1a in combination with ribavirin to treat Middle East respiratory..., Shalhoub [/bib_ref] [bib_ref] Critically ill healthcare workers with the middle east respiratory syndrome (MERS): a..., Shalhoub [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref] , and seven studies for SARS-CoV1 from which two articles could be retrieved in full-text [bib_ref] Description and clinical treatment of an early outbreak of severe acute respiratory..., Zhao [/bib_ref]. Three studies reported on a similar population of patients. There, they were merged [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref] [bib_ref] Critically ill healthcare workers with the middle east respiratory syndrome (MERS): a..., Shalhoub [/bib_ref]. More specifically, the report by Shalhoub et al. was based on a cohort of 32 cases derived from 330 cases previously described by in 2017 in a conference paper . The multi-center cohort by [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref] is an extended version that includes 349 cases, most of whom were Randomization process not stated. previously described in earlier publications. As a result, the three studies were merged according to the 2019 report by [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref]. A sum of 3122 cases, 1665 (53.3%) males and (46.7%) 1457 females, with either suspected or confirmed COVID-19, was explored in 29 distinct reports. The mean age for COVID-19 patients was 47.12 (n = 1046) for 12 studies [bib_ref] COVID-19 in solid organ transplant recipients: a single-center case series from Spain, Fernández-Ruiz [/bib_ref] [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref]. After calculating the point estimate of the mean for the rest of the studies, which did not report study setting [bib_ref] Dealing with missing standard deviation and mean values in meta-analysis of continuous..., Weir [/bib_ref] , the mean age for all COVID-19 cases reached 51.26 (n = 3122) [bib_ref] Kidney impairment is associated with in-hospital death of COVID-19 patients, Cheng [/bib_ref] [bib_ref] Clinical features of 85 fatal cases of COVID-19 from Wuhan: a retrospective..., Du [/bib_ref] [bib_ref] COVID-19 in solid organ transplant recipients: a single-center case series from Spain, Fernández-Ruiz [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a..., Pan [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] [bib_ref] The correlation between viral clearance and biochemical outcomes of 94 COVID-19 infected..., Yuan [/bib_ref] [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref]. In 24 distinct MERS publications, 1196 patients, including 587 males, 269 females, and 340 whose gender was not reported, were investigated [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Combination therapy with lopinavir/ritonavir, ribavirin and interferon-alpha for Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] IFN-alpha2a or IFN-beta1a in combination with ribavirin to treat Middle East respiratory..., Shalhoub [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref]. Two studies did not report the age of 18 and 8 cases , respectively. A SARS study reported a mean age of 28.6 (n = 190) for one study . The overall mean of both studies was 30.39 (n = 212) after calculating the mean for the other study [bib_ref] Description and clinical treatment of an early outbreak of severe acute respiratory..., Zhao [/bib_ref]. All studies used nucleic acid real-time polymerase chain reaction (RT-PCR) test to detect the presence of CoVs in respiratory (e.g., nasopharyngeal, throat, upper respiratory swab) or urinary specimen (Al [bib_ref] Characteristics and outcomes of Middle East respiratory syndrome coronavirus patients admitted to..., Al-Hameed [/bib_ref] [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] Ribavirin and interferon therapy in patients infected with the Middle East respiratory..., Al-Tawfiq [/bib_ref] [bib_ref] Clinical predictors of mortality of Middle East Respiratory Syndrome Coronavirus (MERS-CoV) infection:..., Alfaraj [/bib_ref] [bib_ref] Effect of Ribavirin and Interferon on the Outcome of Critically Ill Patients..., Arabi [/bib_ref] [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref] [bib_ref] Experimental Treatment with Favipiravir for COVID-19: an, Cai [/bib_ref] [bib_ref] Kidney impairment is associated with in-hospital death of COVID-19 patients, Cheng [/bib_ref] [bib_ref] Absence of neutralizing activity in serum 1 year after successful treatment with..., Choi [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Clinical features of 85 fatal cases of COVID-19 from Wuhan: a retrospective..., Du [/bib_ref] [bib_ref] Mortality rate of ICU patients with the Middle East respiratory syndrome-coronavirus infection..., Garout [/bib_ref] [bib_ref] Clinical outcomes among hospital patients with Middle East respiratory syndrome coronavirus (MERS-CoV)..., Habib [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] Combination therapy with lopinavir/ritonavir, ribavirin and interferon-alpha for Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a..., Pan [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Critically ill healthcare workers with the middle east respiratory syndrome (MERS): a..., Shalhoub [/bib_ref] [bib_ref] IFN-alpha2a or IFN-beta1a in combination with ribavirin to treat Middle East respiratory..., Shalhoub [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] [bib_ref] The correlation between viral clearance and biochemical outcomes of 94 COVID-19 infected..., Yuan [/bib_ref] [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref] , except for SARS-CoV1-infected patients who were included according to clinical inclusion criteria and IgG testing [bib_ref] Description and clinical treatment of an early outbreak of severe acute respiratory..., Zhao [/bib_ref]. A both positive and clear contact history with suspected or confirmed [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Interferon alfacon-1 plus corticosteroids in severe acute respiratory syndrome: a preliminary study, Loutfy [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a..., Pan [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Critically ill healthcare workers with the middle east respiratory syndrome (MERS): a..., Shalhoub [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] [bib_ref] The correlation between viral clearance and biochemical outcomes of 94 COVID-19 infected..., Yuan [/bib_ref] [bib_ref] Description and clinical treatment of an early outbreak of severe acute respiratory..., Zhao [/bib_ref] , antibiotics such as imipenem, meropenem, cilastatin, quinolones, cephalosporins, and macrolides (Al [bib_ref] Characteristics and outcomes of Middle East respiratory syndrome coronavirus patients admitted to..., Al-Hameed [/bib_ref] [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] Ribavirin and interferon therapy in patients infected with the Middle East respiratory..., Al-Tawfiq [/bib_ref] [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Clinical features of 85 fatal cases of COVID-19 from Wuhan: a retrospective..., Du [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Interferon alfacon-1 plus corticosteroids in severe acute respiratory syndrome: a preliminary study, Loutfy [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a..., Pan [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] [bib_ref] Description and clinical treatment of an early outbreak of severe acute respiratory..., Zhao [/bib_ref] [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref] , albumin (Jian-ya, 2020; [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] , and traditional Chinese medicine (TCM) [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref]. [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Mortality rate of ICU patients with the Middle East respiratory syndrome-coronavirus infection..., Garout [/bib_ref] [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] Critically ill healthcare workers with the middle east respiratory syndrome (MERS): a..., Shalhoub [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref] , electrolyte correction , oral or IV rehydration , prone positioning [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Critically ill healthcare workers with the middle east respiratory syndrome (MERS): a..., Shalhoub [/bib_ref] , blood transfusion (Al-Qaseer, 2015; [bib_ref] Effect of Ribavirin and Interferon on the Outcome of Critically Ill Patients..., Arabi [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Critically ill healthcare workers with the middle east respiratory syndrome (MERS): a..., Shalhoub [/bib_ref] , NO therapy [bib_ref] Critically ill healthcare workers with the middle east respiratory syndrome (MERS): a..., Shalhoub [/bib_ref] , tracheostomy [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Critically ill healthcare workers with the middle east respiratory syndrome (MERS): a..., Shalhoub [/bib_ref] , intubation [bib_ref] Interferon alfacon-1 plus corticosteroids in severe acute respiratory syndrome: a preliminary study, Loutfy [/bib_ref] , and oxygen therapy (e.g., via nasal cannula) [bib_ref] Experimental Treatment with Favipiravir for COVID-19: an, Cai [/bib_ref] [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] COVID-19 in solid organ transplant recipients: a single-center case series from Spain, Fernández-Ruiz [/bib_ref] [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref]. Several studies reported patients' initial symptoms on admission, including fever, cough, shortness of breath, sputum or phlegm production, runny nose, nose obstruction, sore throat, myalgia, headache, dizziness, asthenia, and GI symptoms (e.g., diarrhea, nausea, and vomiting) ( ( (2) dosage, and administration route of both IFN and non-IFN treatments have been summarized [fig_ref] Figure 3: Quantitative analysis for COVID-19 severity and IFN administration vs [/fig_ref].
[formula] 2) SARS- CoV-2 1(a) 1(a) 0(e) 1(a) 1(b) 1(b) 1(a) 0(d) Poor quality Lo et al. (3) SARS- CoV-2 1(b) 1(a) 1(a) 1(a) 0(c) 1(b) 1(a) 0( [/formula]
## Assessment of risk of bias
29 COVID-19 studies [bib_ref] Kidney impairment is associated with in-hospital death of COVID-19 patients, Cheng [/bib_ref] [bib_ref] Clinical features of 85 fatal cases of COVID-19 from Wuhan: a retrospective..., Du [/bib_ref] [bib_ref] COVID-19 in solid organ transplant recipients: a single-center case series from Spain, Fernández-Ruiz [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a..., Pan [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] [bib_ref] The correlation between viral clearance and biochemical outcomes of 94 COVID-19 infected..., Yuan [/bib_ref] [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref] were included, four of which were clinical trials [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref]. Among trials, two were randomized [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref]. An RCT was of a high risk of bias; due to that, the assessors were aware of the intervention, and no efforts to resolve the possibility of bias were discussed . The other was of low risk of bias . Also, there were two non-randomized trials [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] , which had a moderate risk of bias , and one was not assessed due to no statement on the randomization process (in the protocol or the publication) . A poor quality cross-sectional study was also included . 11 cohorts of COVID-19 cases were critically appraised [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] The correlation between viral clearance and biochemical outcomes of 94 COVID-19 infected..., Yuan [/bib_ref] [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref]. Of those, three had good quality [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref] , two had a fair quality [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] , and others had a poor quality [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] The correlation between viral clearance and biochemical outcomes of 94 COVID-19 infected..., Yuan [/bib_ref]. 12 case-series [bib_ref] Clinical features of 85 fatal cases of COVID-19 from Wuhan: a retrospective..., Du [/bib_ref] [bib_ref] COVID-19 in solid organ transplant recipients: a single-center case series from Spain, Fernández-Ruiz [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] were also assessed. Results showed seven to be of good quality, while four had a fair quality [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] , and one was of poor quality (Jian-ya, 2020). Importantly, a case-series was pre-printed , but was later published with a comparator group. The published version showed poor quality due to a lack of comparability according the NOS tool . The only included case report was of a high risk of bias [fig_ref] Table 4: Critical appraisal for cross-sectional studies using the NIH tool [/fig_ref].
From 26 studies of MERS (Al [bib_ref] Characteristics and outcomes of Middle East respiratory syndrome coronavirus patients admitted to..., Al-Hameed [/bib_ref] [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] Ribavirin and interferon therapy in patients infected with the Middle East respiratory..., Al-Tawfiq [/bib_ref] [bib_ref] Clinical predictors of mortality of Middle East Respiratory Syndrome Coronavirus (MERS-CoV) infection:..., Alfaraj [/bib_ref] [bib_ref] Effect of Ribavirin and Interferon on the Outcome of Critically Ill Patients..., Arabi [/bib_ref] [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref] [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Absence of neutralizing activity in serum 1 year after successful treatment with..., Choi [/bib_ref] [bib_ref] Mortality rate of ICU patients with the Middle East respiratory syndrome-coronavirus infection..., Garout [/bib_ref] [bib_ref] Clinical outcomes among hospital patients with Middle East respiratory syndrome coronavirus (MERS-CoV)..., Habib [/bib_ref] [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Combination therapy with lopinavir/ritonavir, ribavirin and interferon-alpha for Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] IFN-alpha2a or IFN-beta1a in combination with ribavirin to treat Middle East respiratory..., Shalhoub [/bib_ref] [bib_ref] Critically ill healthcare workers with the middle east respiratory syndrome (MERS): a..., Shalhoub [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref] , two were good quality case-series [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref]. There were 11 cohorts (Al [bib_ref] Characteristics and outcomes of Middle East respiratory syndrome coronavirus patients admitted to..., Al-Hameed [/bib_ref] [bib_ref] Clinical predictors of mortality of Middle East Respiratory Syndrome Coronavirus (MERS-CoV) infection:..., Alfaraj [/bib_ref] [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Mortality rate of ICU patients with the Middle East respiratory syndrome-coronavirus infection..., Garout [/bib_ref] [bib_ref] Clinical outcomes among hospital patients with Middle East respiratory syndrome coronavirus (MERS-CoV)..., Habib [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] IFN-alpha2a or IFN-beta1a in combination with ribavirin to treat Middle East respiratory..., Shalhoub [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref] , two of which were good quality [bib_ref] IFN-alpha2a or IFN-beta1a in combination with ribavirin to treat Middle East respiratory..., Shalhoub [/bib_ref] , while the other nine were of poor quality, mostly due to low comparability scores and not adjusting for confounding factors (Al [bib_ref] Characteristics and outcomes of Middle East respiratory syndrome coronavirus patients admitted to..., Al-Hameed [/bib_ref] [bib_ref] Clinical predictors of mortality of Middle East Respiratory Syndrome Coronavirus (MERS-CoV) infection:..., Alfaraj [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Mortality rate of ICU patients with the Middle East respiratory syndrome-coronavirus infection..., Garout [/bib_ref] [bib_ref] Clinical outcomes among hospital patients with Middle East respiratory syndrome coronavirus (MERS-CoV)..., Habib [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref]. From 11 case reports in MERS [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] Ribavirin and interferon therapy in patients infected with the Middle East respiratory..., Al-Tawfiq [/bib_ref] [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] Absence of neutralizing activity in serum 1 year after successful treatment with..., Choi [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Combination therapy with lopinavir/ritonavir, ribavirin and interferon-alpha for Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] , ten were of a high risk of bias, mainly because of lacking any description of suitable selection processes (Al-Qaseer, 2015; Al-Tawfiq and Hinedi, 2018; Al-Tawfiq et al., 2014; [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] Absence of neutralizing activity in serum 1 year after successful treatment with..., Choi [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Combination therapy with lopinavir/ritonavir, ribavirin and interferon-alpha for Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref]. Furthermore, only one study had a low risk of bias .
Two studies were included in SARS [bib_ref] Description and clinical treatment of an early outbreak of severe acute respiratory..., Zhao [/bib_ref]. One was a randomized clinical trial with a high risk of bias, and the other was a cohort study of good quality .
Finally, we provided the results for the quality of evidence by the LOE tool [fig_ref] Figure 3: Quantitative analysis for COVID-19 severity and IFN administration vs [/fig_ref]. For COVID-19 studies, only one study declared a conflict of interest , and others did not have any competing interests to declare. For MERS studies, two studies declared the existence of a conflict of interest [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref]. Finally, SARS studies declared no conflict of interest. A few studies did not mention whether competing interests were present [fig_ref] Table 2: Characteristics of included MERS-CoV studies [/fig_ref].
## Mortality
## Covid-19
Out of 29 clinical COVID-19 studies, three did not specify mortality [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref]. A total of 414 cases expired in 26 studies [bib_ref] Kidney impairment is associated with in-hospital death of COVID-19 patients, Cheng [/bib_ref] [bib_ref] Clinical features of 85 fatal cases of COVID-19 from Wuhan: a retrospective..., Du [/bib_ref] [bib_ref] COVID-19 in solid organ transplant recipients: a single-center case series from Spain, Fernández-Ruiz [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a..., Pan [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] [bib_ref] The correlation between viral clearance and biochemical outcomes of 94 COVID-19 infected..., Yuan [/bib_ref]. Interestingly, 14 studies reported no mortality [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] [bib_ref] The correlation between viral clearance and biochemical outcomes of 94 COVID-19 infected..., Yuan [/bib_ref] , and all cases in three studies died [bib_ref] Clinical features of 85 fatal cases of COVID-19 from Wuhan: a retrospective..., Du [/bib_ref]. This was, in part, due to that some studies strictly sampled cases with a fatal outcome or survivors. Interestingly a study of 101 non-survivors was published with 134 cases, comprising a new comparator group of 33 survivors .
A recent open-label RCT showed no mortality in both LPV/RTV + RBV + IFN-β (n = 86) and LPV/RTV groups (n = 41) (P = 1.00) . A double-blind, placebo-controlled, multicenter RCT included 158 and 78 cases as intention-to-treat population in REM + IFN and Placebo + IFN groups, respectively. Results showed a 28-day Mortality of 22 (14%) in REM group (for REM + IFN: 29 (18%)) and 10 (13%) in the placebo group (for placebo + IFN: 15 (19%)) (risk difference = 1⋅1%, 95% CI: (− 8⋅1,10⋅3)) .
## Mers
A total of 494 patients expired in all 24 studies (Al [bib_ref] Characteristics and outcomes of Middle East respiratory syndrome coronavirus patients admitted to..., Al-Hameed [/bib_ref] [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] Ribavirin and interferon therapy in patients infected with the Middle East respiratory..., Al-Tawfiq [/bib_ref] [bib_ref] Clinical predictors of mortality of Middle East Respiratory Syndrome Coronavirus (MERS-CoV) infection:..., Alfaraj [/bib_ref] [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref] [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Absence of neutralizing activity in serum 1 year after successful treatment with..., Choi [/bib_ref] [bib_ref] Mortality rate of ICU patients with the Middle East respiratory syndrome-coronavirus infection..., Garout [/bib_ref] [bib_ref] Clinical outcomes among hospital patients with Middle East respiratory syndrome coronavirus (MERS-CoV)..., Habib [/bib_ref] [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Combination therapy with lopinavir/ritonavir, ribavirin and interferon-alpha for Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] IFN-alpha2a or IFN-beta1a in combination with ribavirin to treat Middle East respiratory..., Shalhoub [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref]. A multi-center study, after adjusting for diabetes with chronic complications, liver disease, renal disease, any malignancy, SOFA score on day 1, source of infection, and year, indicated an increased day 90 mortality in the group receiving rIFN vs. the no rIFN group (logistic regression adjusted OR = 2.53, 95% CI: (1.32,4.85) (P = 0.005)) . The combination therapy of IFN/RBV was not associated with death in a recent cohort . Another study reported a CFR of 31.5% in patients who received IFN treatments, and a CFR of 40% in patients who did not receive IFN (P = 0.698) .
## Sars
In two studies, 12 (5.67%) patients died [bib_ref] Description and clinical treatment of an early outbreak of severe acute respiratory..., Zhao [/bib_ref]. A randomized trial of 190 patients treated SARS cases with the following regimens: Group A (n = 40): RBV and Cefoperazone/Sulbactam, and oxygen therapy; Group B: fluoroquinolone, rIFN-α and restricted steroid use (n = 30); Group C (n = 60): quinolone, azithromycin, rIFN-α for some patients, and steroids when symptoms worsened; and Group D (n = 60): levofloxacin, azithromycin, 45 patients were given rIFN-α, high-dose methylprednisolone was given when infiltrates affected more than one pulmonary segment or when consolidation was expanded, and broad-spectrum antibiotics if a bacterial infection was confirmed after culture. In four groups, 2 (5%), 2 (6.67%), 7 (11.67%), and 0 (0%) patients died, respectively . In the other SARS study, 1 (7.7%) patient in the corticosteroid group (n = 13) died, while all patients in the corticosteroid + IFN-Alfacon-1 group (n = 9) survived .
## Discharge
## Covid-19
953 hospital discharges were reported in 24 studies [bib_ref] COVID-19 in solid organ transplant recipients: a single-center case series from Spain, Fernández-Ruiz [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a..., Pan [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] [bib_ref] The correlation between viral clearance and biochemical outcomes of 94 COVID-19 infected..., Yuan [/bib_ref] [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref]. A RCT using a six-category ordinal scale previously defined by the authors showed that at the first level of the scale, day 28 discharge (alive) was 92 (61%) in the REM + IFN group compared to 45 (58%) in the placebo + IFN group (OR = 1⋅15, 95% CI: (0⋅67,1⋅96)) .
## Mers
A total of 33 cases were discharged in 18 studies (Al [bib_ref] Characteristics and outcomes of Middle East respiratory syndrome coronavirus patients admitted to..., Al-Hameed [/bib_ref] [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] Ribavirin and interferon therapy in patients infected with the Middle East respiratory..., Al-Tawfiq [/bib_ref] [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Absence of neutralizing activity in serum 1 year after successful treatment with..., Choi [/bib_ref] [bib_ref] Mortality rate of ICU patients with the Middle East respiratory syndrome-coronavirus infection..., Garout [/bib_ref] [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Combination therapy with lopinavir/ritonavir, ribavirin and interferon-alpha for Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref]. However, six studies did not report a clear discharge outcome [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref] [bib_ref] Clinical outcomes among hospital patients with Middle East respiratory syndrome coronavirus (MERS-CoV)..., Habib [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] IFN-alpha2a or IFN-beta1a in combination with ribavirin to treat Middle East respiratory..., Shalhoub [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref]. A recent observation, in which all cases were either discharged or deceased by the end of the study period, showed a discharge rate of 20% in the RBV + IFN-α (n = 35) group vs. 35.2% in the no RBV + IFN-α group (n = 17) .
## Chest imaging and x-ray presentations
## Covid-19
Consolidation of pneumonia was indicated in six studies [bib_ref] COVID-19 in solid organ transplant recipients: a single-center case series from Spain, Fernández-Ruiz [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] , and local or diffuse infiltrates were reported in two studies [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref]. Some publications reported ground glassy shadows [bib_ref] Clinical features of 85 fatal cases of COVID-19 from Wuhan: a retrospective..., Du [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref]. Blurred edges were also reported by one study . Speckles and patchy shadows were observed in nine studies [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref]. Thickening or disorder of textures was observed in three distinct reports [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref]. Other reported categories included unilateral or bilateral CXR involvement, pleural effusion, pneumothorax, white lung appearance, lung streak shadow, single lobe lesions, multiple solid nodules, visible band shadows, and bronchial shadow with air [bib_ref] Kidney impairment is associated with in-hospital death of COVID-19 patients, Cheng [/bib_ref] [bib_ref] Clinical features of 85 fatal cases of COVID-19 from Wuhan: a retrospective..., Du [/bib_ref] [bib_ref] COVID-19 in solid organ transplant recipients: a single-center case series from Spain, Fernández-Ruiz [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a..., Pan [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref].
A recent non-randomized open-label trial investigated the efficacy of combination therapy of IFN with FPV, and included a total of 80 patients, who received IFN-α1b in two arms of the study (FPV + IFN group (n = 35), LPV/RTV + IFN (n = 45). The results showed that CT scan scores (median, range) were 12 (4.0-14.0) for FPV + IFN group, and 10 (4.5-13.5) for the LPV/RTV + IFN group (P = 0.78). Chest CT changes showed improvement in 32 cases (91.43%) vs. 28 (62.22%) cases, deterioration in 1 case (3.23%) vs. 9 (20.00%) cases, and was constant in 2 cases (6.45%) vs. 8 (17.78%) cases in FPV + IFN group and LPV/RTV + IFN group after 2 weeks, respectively (P = 0.004) .
## Mers
Of 24 distinct reports, lung consolidation was present in 12 studies (Al [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Absence of neutralizing activity in serum 1 year after successful treatment with..., Choi [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref] , while 11 studies showed infiltrates (Al Ghamdi et al., 2016; ⨁⨁⨁x MODERATE , , ) *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). shadows [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref]. Patchy shadows were observed in two studies (Al-Tawfiq , , , , , ) *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
## Ci: confidence interval;
CoV: coronavirus; IFN: interferon; MERS: middle east respiratory syndrome; OR: Odds ratio; RBV: ribavirin.
## Grade working group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
## Explanations.
References. a Many did not control for age, sex, and disease severity. Both studies did not report a remarkable difference in total mortality. Considerable variations were present among study designs.
(2 RCTs) ⨁⨁⨁x MODERATE a Comparative assessments are lacking.
(2 non-randomized trials)
-Comparative assessments are lacking.
(25 observational studies)
⨁⨁xx
## Low discharge
Wang et al. showed higher discharge rate in IFN + REM group compared to IFN + Placebo.
(2 RCTs) ⨁⨁⨁⨁ HIGH Comparative assessments are lacking.
(2 non-randomized trials)
-Comparative assessments are lacking.
(25 observational studies)
-Chest X-ray Comparative assessments are lacking.
(2 RCTs) -Cai et al. showed FPV + IFN-α was significantly linked to improvement in CXR compared with LPV/RTV + IFN-α treatment (p = 0.004).
(2 non-randomized trials)
⨁⨁⨁⨁ HIGH Comparative assessments are lacking.
(25 observational studies)
-
## Severity
Meta-analysis conducted.
(2 RCTs) -b Jiang et al. showed cases treated with IFN-β + LPV/RTV, 39 (80%) were non-severe and 3 (38%) were severe (p = 0.045). Also, among cases treated with IFN-β + LPV/RTV + ARB, 10 (19%) were non-severe and 5 (63%) were severe (p = 0.019) (2 non-randomized trials)
⨁⨁⨁⨁ HIGH Comparative assessments are lacking.
(25 observational studies)
-*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref]. Other reported categorizations included atelectasis, hilar vascular shadow, bronchovascular marking, acute pulmonary embolism, multiple solid nodules, single or multiple lobe lesion, and pleural effusion [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Combination therapy with lopinavir/ritonavir, ribavirin and interferon-alpha for Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref].
## Ades
## Sars
In a randomized trial with four treatment groups (described in the SARS mortality section), the number of cases with unabsorbed pulmonary infiltrates was 12, 11, 13, and 4 for groups A, B, C, and D, respectively. Moreover, the difference between groups was significant (P = 0.003). This study also reported infiltrates localized in one pulmonary segment, signs in one pulmonary field, the involvement of both lungs, diffuse damage, as well as reported cases with only interstitial changes . In a recent preliminary study, patients were treated with IFN-Alfacon-1 + corticosteroid or corticosteroid alone. In this study, all cases in both groups showed abnormal chest imaging (P > 0.99). Eighteen patients did not show a full resolution of CXR abnormalities. Interestingly, the IFN-Alfacon-1 treatment group showed a reduced duration to 50% resolution of lung imaging abnormalities. The median for this duration was 4 in the IFN-Alfacon-1 + corticosteroid group vs. 9 in the corticosteroid only group (P = 0.001) .
## Disease severity
## Covid-19
Seven studies that did not report the number of severe and nonsevere cases [bib_ref] COVID-19 in solid organ transplant recipients: a single-center case series from Spain, Fernández-Ruiz [/bib_ref] [bib_ref] Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads..., Liu [/bib_ref] [bib_ref] Clinical characteristics of 80 hospitalized frontline medical workers infected with COVID-19 in..., Wang [/bib_ref] [bib_ref] Clinical findings in a group of patients infected with the 2019 novel..., Xiao-Wei [/bib_ref] [bib_ref] Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19..., Yu [/bib_ref] were excluded. 22 distinct reports, including 766 severe and 2007 non-severe cases, were studied [bib_ref] Kidney impairment is associated with in-hospital death of COVID-19 patients, Cheng [/bib_ref] [bib_ref] Clinical features of 85 fatal cases of COVID-19 from Wuhan: a retrospective..., Du [/bib_ref] [bib_ref] Clinical characteristics and drug therapies in patients with the common-type coronavirus disease, Huang [/bib_ref] [bib_ref] Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of..., Hung [/bib_ref] [bib_ref] Clinical characteristics of 60 discharged cases of 2019 novel coronavirusinfected pneumonia in..., Jiang [/bib_ref] [bib_ref] Epidemiological, clinical and virological characteristics of 74 cases of coronavirus-infected disease 2019..., Jin [/bib_ref] [bib_ref] Patients of COVID-19 may benefit from sustained lopinavir-combined regimen and the increase..., Liu [/bib_ref] [bib_ref] Evaluation of SARS-CoV-2 RNA shedding in clinical specimens and clinical characteristics of..., Lo [/bib_ref] [bib_ref] Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a..., Pan [/bib_ref] [bib_ref] Clinical and epidemiological features of 36 children with coronavirus disease 2019 (COVID-19)..., Qiu [/bib_ref] [bib_ref] Clinical characteristics of 28 patients with novel coronavirus pneumonia, Rui [/bib_ref] [bib_ref] Clinical features of severe pediatric patients with coronavirus disease 2019 in Wuhan:..., Sun [/bib_ref] [bib_ref] Clinical features and treatment of COVID-19 patients in northeast Chongqing, Wan [/bib_ref] [bib_ref] Quadruple therapy for asymptomatic COVID-19 infection patients, Wang [/bib_ref] [bib_ref] Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial, Wang [/bib_ref] [bib_ref] The correlation between viral clearance and biochemical outcomes of 94 COVID-19 infected..., Yuan [/bib_ref] [bib_ref] Interferon-α2b treatment for COVID-19, Zhou [/bib_ref].
A retrospective cohort reported mean IFN treatment durations (days) in various levels of COVID-19 severity were 10.88, 95% CI: (8.00,13.75) in the mild group (n = 8), in the moderate group (n = 75), and 15.55, 95% CI: in the severe group (n = 11), which were significantly different (one-way ANOVA, P = 0.01) . The number of non-severe (n = 52) and severe (n = 8) patients receiving various combination IFN regimens were reported in a trial. Among cases treated with IFN-β + LPV/RTV, 39 (80%) were non-severe and 3 (38%) were severe (P = 0.045). Also, among cases treated with IFN-β + LPV/RTV + ARB, 10 (19%) were non-severe and 5 In 7 studies, ADEs were recorded while using regimen containing IFNs, including multi-organ damage, adverse change in blood profile, thrombocytopenia, kidney disease, fever, and pancreatitis (Al-Tawfiq et al., .
(26 observational studies)
⨁xxx VERY LOW a,c *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
ADEs: adverse drug events, CI: Confidence interval, CRP: C-reactive protein, IFN: interferon, MERS: Middle-Eastern respiratory syndrome, RBV: ribavirin.
## Grade working group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
Explanations. a Different IFN-based regimen were used, and were compared to varied treatment options. Results should be taken as speculative, rather than as for net efficacy of IFN. b All important inflammatory elements, such as major inflammatory cytokines are required for proper assessment of inflammatory state. c Combination therapies of IFN as well as lack of a comparator group makes it difficult to determine whether such adverse events are a direct result of IFNs administration.
(63%) were severe (P = 0.019) .
## Mers
454 severe cases were reported in 10 studies [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref] [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Absence of neutralizing activity in serum 1 year after successful treatment with..., Choi [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref]. The rest of studies had an unclear number of severe cases (Al [bib_ref] Ribavirin and interferon therapy in patients infected with the Middle East respiratory..., Al-Tawfiq [/bib_ref] [bib_ref] Clinical predictors of mortality of Middle East Respiratory Syndrome Coronavirus (MERS-CoV) infection:..., Alfaraj [/bib_ref] [bib_ref] Mortality rate of ICU patients with the Middle East respiratory syndrome-coronavirus infection..., Garout [/bib_ref] [bib_ref] Clinical outcomes among hospital patients with Middle East respiratory syndrome coronavirus (MERS-CoV)..., Habib [/bib_ref] [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Combination therapy with lopinavir/ritonavir, ribavirin and interferon-alpha for Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] IFN-alpha2a or IFN-beta1a in combination with ribavirin to treat Middle East respiratory..., Shalhoub [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref]. In total, 9 non-severe cases were also reported in 10 studies [bib_ref] Ribavirin and Interferon Therapy for Critically Ill Patients with Middle East Respiratory..., Arabi [/bib_ref] [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] [bib_ref] Clinical presentation and outcomes of Middle East respiratory syndrome in the Republic..., Choi [/bib_ref] [bib_ref] Absence of neutralizing activity in serum 1 year after successful treatment with..., Choi [/bib_ref] [bib_ref] Case report Ribavirin and interferon-α2b as primary and preventive treatment for Middle..., Khalid [/bib_ref] [bib_ref] Neurological complications during treatment of Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] The clinical and virological features of the first imported case causing MERS-CoV..., Lee [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] , and the rest of studies had an unclear number of non-severe cases (Al [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] Ribavirin and interferon therapy in patients infected with the Middle East respiratory..., Al-Tawfiq [/bib_ref] [bib_ref] Clinical predictors of mortality of Middle East Respiratory Syndrome Coronavirus (MERS-CoV) infection:..., Alfaraj [/bib_ref] [bib_ref] Mortality rate of ICU patients with the Middle East respiratory syndrome-coronavirus infection..., Garout [/bib_ref] [bib_ref] Clinical outcomes among hospital patients with Middle East respiratory syndrome coronavirus (MERS-CoV)..., Habib [/bib_ref] [bib_ref] Middle eastern respiratory syndrome corona virus (MERS CoV): case reports from a..., Khalid [/bib_ref] [bib_ref] Acute management and long-term survival among Subjects with severe Middle East respiratory..., Khalid [/bib_ref] [bib_ref] Combination therapy with lopinavir/ritonavir, ribavirin and interferon-alpha for Middle East respiratory syndrome, Kim [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus superspreading event involving 81 persons, Oh [/bib_ref] [bib_ref] IFN-alpha2a or IFN-beta1a in combination with ribavirin to treat Middle East respiratory..., Shalhoub [/bib_ref] [bib_ref] Critically ill healthcare workers with the middle east respiratory syndrome (MERS): a..., Shalhoub [/bib_ref] [bib_ref] Middle East respiratory syndrome coronavirus in Al-Madinah City, Saudi Arabia: Demographic, clinical..., Sherbini [/bib_ref].
Univariable analysis of the influence of severity of disease on medications administered showed a significant negative risk association of -4.62, 95% CI: (− 8.40,− 0.84) (P = 0.018) for IFN-α, and a negative but non-significant risk association of -1.24, 95% CI: (− 6.71,4.24) (P = 0.652) for IFN-β. Moreover, a multivariable analysis, which included a biomarker of disease severity, showed a strong association between disease severity and decreased survival, and no association between treatment with IFN-β and mortality (OR = 0.68, 95% CI: (0.04,10.28)) (P = 0.778) (Al . However, another study that did not discuss the severity of included patients showed a reduction in mortality was significantly associated with IFN-α (OR = 0.16, 95% CI: One randomized trial study was included, that gives a good idea of effect of specific regimen used in the study, However, assessment of efficacy of IFN therapies, and in particular the role of IFN is not feasible in such settings. Precisely, in a trial In four groups, 2 (5%), 2 (6.67%), 7 (11.67%), and 0 (0%) patients died, respectively . (See explanation for regimens in each group.) a (1 RCT) ⨁⨁⨁⨁ HIGH b
In the other SARS study, 1 (7.7%) patient in the corticosteroid group (n = 13) died, while all patients in the corticosteroid + IFN-Alfacon-1 group (n = 9) survived .
(1 observational study)
⨁⨁xx LOW Discharge
Comparative assessments are lacking.
(1 RCT) -Comparative assessments are lacking.
(1 observational study)
-Chest x-ray Regimen (A, B, C, and D) are could be linked to unabsorbed infiltrates (p = 0.003) . However, this may be of little help for a broad range of IFN regimen due to that regimen given in the study are combinations of numerous agents.
(1 RCT) ⨁⨁⨁⨁ HIGH Addition of IFN-Alfacon-1 to corticosteroid regimen may enhance its therapeutic efficacy. In a cohort, IFN-Alfacon-1 treatment group showed a reduced duration to 50% resolution of lung imaging abnormalities (p = 0.001) .
(1 observational study)
⨁⨁xx LOW
## Inflammatory profile
Comparative assessments are lacking.
(1 RCT) -Comparative assessments are lacking.
(1 observational study)
-
## Severity
Comparative assessments are lacking.
(1 RCT) -Comparative assessments are lacking.
(1 observational study)
-
## Hospital durations
No significant difference was detected in time to discharge .
(1 RCT) ⨁⨁⨁⨁ HIGH Comparative assessments are lacking.
(1 observational study)
-
## Ades
Comparative assessments are lacking.
(1 RCT) -A cohort revealed several adverse effects while treatment. However, it may not be feasible to attribute these, merely to IFN administration. Fever, neutropenia with an absolute neutrophil count (ANC) of less than 1000/μL on the last day of treatment, a minor transient decrease in ANC, and elevation of serum transaminase levels were reported during IFN therapy in both IFN-Alfacon-1 and corticosteroid alone groups .
(1 observational study) ⨁⨁xx LOW * The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). ANC: absolute neutrophil count, CI: Confidence interval, IFN: interferon, RCT: randomized controlled trial, SARS: severe acute respiratory syndrome.
## Grade working group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect Explanations. a The randomized trial of 190 patients by Zhao et al. treated SARS cases with the following regimens: Group A (n = 40): RBV and Cefoperazone/Sulbactam, and oxygen therapy; Group B: fluoroquinolone, rIFN-α and restricted steroid use (n = 30); Group C (n = 60): quinolone, azithromycin, rIFN-α for some patients, and steroids when symptoms worsened; and Group D (n = 60): levofloxacin, azithromycin, 45 patients were given rIFN-α, high-dose methylprednisolone was given when infiltrates affected more than one pulmonary segment or when consolidation was expanded, and broad-spectrum antibiotics if a bacterial infection was confirmed after culture. b Different IFN-based regimen were used, and were compared to varied treatment options. Results should be taken as speculative, rather than as for net efficacy of IFN.
(0.02,1.38)) (P = 0.09). The lower mortality did not reach statistical significance for IFN-β (OR = 0.28, 95% CI (0.03,2.33)) (P = 0.24) .
## Inflammatory cytokines
## Covid-19
A cohort study showed that during the time interval of day 0-20 (upon onset of symptoms), on average, cases receiving the ARB only regimen had higher CRP levels than cases treated with IFN alone or both IFN and ARB, by 25.7 mg/l. Also, over the time interval between day 12 and day 42 (upon onset of symptoms), on average, cases receiving the ARB only regimen showed higher IL-6 levels than the cases who received IFN alone or both IFN and ARB, by 33.5 pg/ml. These effects were not influenced by co-morbidities for IL-6 (P = 0.456), or CRP (P = 0.420) levels . In a recent phase II trial, IL-6 levels (log10 pg/ml, median, (Q1,Q3)) were 1⋅4, (1⋅0-1⋅4) in LPV/RTV + RBV + IFN-β group (n = 86), and 1⋅4, (1⋅0-1⋅6) in LPV/RTV group (n = 41). These results did not show any significant differences between the trial arms (P = 0.43). Also, in this trial, TNF-α levels were measured for both trial arms (P = 1.00) . In another clinical study, CRP (mg/dl, median, (Q1,Q3) level was 18.6 (5.0-20.0) for all patients (n = 80). CRP levels were 15.0, (3.0-19.2) and 21.4, (5.0-23.2) in the FPV + IFN-α (n = 35) and LPV/RTV + IFN-α arm (n = 45) of the study, respectively (P = 0.33). IL-6 (ng/l, median, (Q1,Q3)) was 13.4, (4.4-16.2) in all patients. IL-6 levels were 14.0, (3.5-11.0) and 12.9, (5.3-16.8) in FPV + IFN-α and LPV/RTV + IFN-α arm, respectively (P = 0.77) .
## Mers
There were higher CRP levels (mg/l, median, (Q1,Q3)) in cases treated with IFN-α (n = 13) (86.5, (25,226)) compared to cases treated with ; However, this difference did not reach statistical significance (P = 0.61) .
## Hospitalization duration
## Covid-19
In a recent cohort, duration from the symptom onset to hospital admission (days, median, (Q1,Q3)) was 8.0, (5.5, 15.5), 6.5, (3.0, 10.0), and 10.0, (4.5, 19.5) for IFN, IFN + ARB, and ARB groups, respectively. This difference, however, was not statistically significant (P = 0.087) . A placebo-controlled RCT of IFN therapy in combination with REM showed a similar duration of hospitalization (days, median, (Q1,Q3)) in the two arms of the trial 25⋅0, (16⋅0,38⋅0) in intention-to-treat populations of REM group (for REM + IFN: 29 (18%)) vs. 24⋅0 (18⋅0,36⋅0) in placebo group (for placebo + IFN: 15 (19%)) (risk difference = 0⋅0, 95% CI: (− 4⋅0,4⋅0)) . A lately surfaced cohort indicated that the duration of hospitalization was significantly correlated with PCR negative conversion durations in the IFN-α + LPV/RTV + RBV group (P = 0.0215), as well as the IFN-α + LPV/RTV group (P = 0.012) . A recent study divided the cohort of study into patients who experienced symptoms for more or less than ten days; Furthermore, 15 (45.5%) cases with symptoms lasting longer than ten days and 19 (65.5%) cases with symptoms lasting shorter than or equal to 10 days received IFN alone or in combination with ARB, RBV, or LPV/RTV . In a recent phase II RCT, the duration of hospital stay (days, median (Q1,Q3)) was significantly lower in the LPV/RTV + RBV + IFN-β 9⋅0, (7⋅0-13⋅0), compared with 14⋅5, (9⋅3-16⋅0) in the LPV/RTV (control) group (P = 0⋅016) .
## Mers
The length of hospital stay (days, median, (Q1,Q3)) in a recent multicenter study was reported 17 (10, 28) in RBV/IFN group compared to 20 (10, 36) in the no RBV/IFN group (P = 0.48) [fig_ref] 2: Hung IF-N, Lung K-C, Tso EY-K, Liu R, Chung TW-H, Chu M-Y,... [/fig_ref].
## Sars
In a randomized trial of 4 treatment groups (regimens were described in SARS mortality section), time to discharge (days, (S.D.)) was 24⋅8, (5⋅5) in group A, 24⋅8, (6⋅4) in group B, 22⋅4, (5⋅9) in group C, and 20⋅7, (4⋅6) in group D. Also, the difference between groups was not reported significant .
## Unfavorable drug events in cov infections
## Covid-19
The total number of ADEs were significantly lower in the FPV + IFNα (4) compared to the LPV/RTV arm of the trial (25) (P = 0.001); Also, nausea in patients in the FPV + IFN-α group was lower significantly in the FPV + IFN-α (0) compared to the LPV/RTV arm of the trial (6) (P = 0.03) . In another study, self-limited nausea and diarrhea were similar between the two groups. Furthermore, in this study, ADEs were reported by 41 (48%) of cases in the LPV/RTV + RBV + IFN group and 20 (49%) of cases in the LPV/RTV group . A placebo-controlled multi-center trial for the efficacy of REM + IFN compared with placebo + IFN found that ADEs were reported in 102 (66%) of REM receivers compared to 50 (64%) of placebo receivers. Importantly, REM was stopped early because of ADEs in 18 (12%) cases compared to 4 (5%) patients who discontinued placebo early . Finally, ADEs have been provided [fig_ref] Table 2: Characteristics of included MERS-CoV studies [/fig_ref].
## Mers
MERS patients showed several ADEs during treatment with IFNs and other drugs. ADEs also included pancreatitis, which was reported in one patient , kidney injuries (e.g., AKI) in 5 cases , and renal failure in 3 cases . Hepatic jaundice occurred in one patient during therapy . Changes in laboratory data were mentioned in some studies. Alterations in the laboratory profile of patients included increased amylase and lipase in seven patients who were treated with IFN [bib_ref] Clinical implications of five cases of Middle East respiratory syndrome coronavirus infection..., Rhee [/bib_ref] , increased bilirubin , decrease in Hb in 45 cases [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] , Scr and AST-ALT elevation in one patient with IFN therapy , thrombocytopenia in 2 patients [bib_ref] A case report of a Middle East respiratory syndrome survivor with kidney..., Cha [/bib_ref] , anemia in 46 cases [bib_ref] The most effective therapeutic regimen for patients with severe Middle East Respiratory..., Al-Qaseer [/bib_ref] [bib_ref] Ribavirin and interferon alfa-2a for severe Middle East respiratory syndrome coronavirus infection:..., Omrani [/bib_ref] , and hemodynamic instability in 3 patients during treatment with IFN and other drugs were reported as ADEs . Fever was mentioned in one IFN receiver . Also, multi-organ damage as a severe side effect was reported in 5 cases treated with IFN .
## Sars
In SARS patients, fever, neutropenia with an absolute neutrophil count (ANC) of less than 1000/μl on the last day of treatment, a minor transient decrease in ANC, and elevation of serum transaminase levels were reported during IFN therapy in both IFN-Alfacon-1 and corticosteroid alone groups .
# Quantitative analysis
Three observational COVID-19 investigations were eligible for metaanalysis of IFN treatment and severity. The studies showed moderate certainty [fig_ref] Table 9: Summary of findings for COVID-10 severity and IFN administration [/fig_ref]. Fixed-effects (Mantel-Haenszel) (I 2 <35%) approach was employed. Results showed the relation between receiving IFN and severity to be inconclusive. Effect size in all three studies crossed the line of no effect, indicating inconclusiveness of the current data (log OR = − 0.44, 95% CI: (− 1.13,0.25), I 2 = 31.42%) [fig_ref] Figure 3: Quantitative analysis for COVID-19 severity and IFN administration vs [/fig_ref].
Six MERS-CoV cohorts were included in the quantitative synthesis. Evaluation of the risk of bias across studies via GRADE showed the evidence was low certainty, indicating that further research could change our estimation. Random-effects (DerSimonian-Laird) (I 2 >35%) approach was selected. Mortality was not clearly affected by the administration of RBV/IFN treatment vs. no RBV/IFN (log OR = − 0.05, 95% CI: (− 0.71,0.62), I 2 = 44.71%) [fig_ref] Figure 4: Quantitative analysis for MERS-CoV mortality and IFN/RBV administration vs [/fig_ref]. Publication bias was not analyzed due to the insufficient number of studies (n < 10). Finally, summary of findings table for GRADE assessments for narrative synthesis outcomes were conducted according to a new study [bib_ref] Rating the certainty in evidence in the absence of a single estimate..., Murad [/bib_ref] , and results were provided inand SARS, respectively.
# Discussion
SARS-CoV, MERS-CoV, and SARS-CoV2 are human CoVs (hCoVs) that have been the cause of three outbreaks during the last two decades . All of them have shown the potential to manifest as multisystem difficult to treat infections [bib_ref] Cutaneous lesions and COVID-19: cystic painful lesion in a case with positive..., Goudarzi [/bib_ref] [bib_ref] THE SARS EPIDEMIC AND ITS AFTERMATH IN CHINA: A POLITICAL PERSPECTIVE, Heidarpour [/bib_ref] [bib_ref] Occurrence of Acute Coronary Syndrome, Pulmonary Thromboembolism, and Cerebrovascular Event in COVID-19, Jenab [/bib_ref] [bib_ref] SARS-CoV-2-A Tough Opponent for the Immune System. Archives of medical research. National..., Nasab [/bib_ref] [bib_ref] Mesenchymal stem cell-derived extracellular vesicle-based therapies protect against coupled degeneration of the..., Rahmani [/bib_ref] [bib_ref] COVID-19 and multiple sclerosis: predisposition and precautions in treatment. SN Compr, Sadeghmousavi [/bib_ref] [bib_ref] Anosmia: a missing link in the neuroimmunology of coronavirus disease 2019 (COVID-19), Yazdanpanah [/bib_ref]. In particular, multisystem involvement in COVID-19 has been associated with anti-viral immunity paralysis on the one hand, and on the other hand release of pro-inflammatory cytokines, e.g., interleukin-6, and hyperinflammatory shock [bib_ref] Hyperinflammatory shock related to COVID-19 in a patient presenting with multisystem inflammatory..., Bahrami [/bib_ref] [bib_ref] Lymphopenia in COVID-19: therapeutic opportunities, Fathi [/bib_ref] [bib_ref] Interleukin-6 and severe COVID-19: a systematic review and meta-analysis, Mojtabavi [/bib_ref] [bib_ref] SARS-CoV-2-A Tough Opponent for the Immune System. Archives of medical research. National..., Nasab [/bib_ref] [bib_ref] Immune-epidemiological parameters of the novel coronavirus -a perspective, Saghazadeh [/bib_ref] [bib_ref] Kawasaki disease and multisystem inflammatory syndrome in children with COVID-19. SN Compr, Sarzaeim [/bib_ref] [bib_ref] The immune system and COVID-19: friend or foe?, Yazdanpanah [/bib_ref]. Such an unpleasant event is orchestrated by genetic and environmental factors that make individuals susceptible to develop hyper inflammatory responses [bib_ref] Genetic predisposition models to COVID-19 infection, Darbeheshti [/bib_ref] [bib_ref] Case report: death due to COVID-19 in three brothers, Yousefzadegan [/bib_ref]. Supporting this, inborn errors of immunity have not been shown to increase the risk of developing severe COVID-19 and dying from that. However, there are sporadic reports of death in patients with combined immunodeficiency [bib_ref] Death due to COVID-19 in an infant with combined immunodeficiencies, Ahanchian [/bib_ref] [bib_ref] Primary immunodeficiency diseases in COVID-19 pandemic: a predisposing or protective factor?, Babaha [/bib_ref]. For this, anti-inflammatory and immunomodulatory treatments along with monoclonal antibodies appear as potential candidates [bib_ref] Regenerative medicine in COVID-19 treatment: real opportunities and range of promises, Basiri [/bib_ref] [bib_ref] Lymphopenia in COVID-19: therapeutic opportunities, Fathi [/bib_ref] [bib_ref] Monoclonal antibody as a potential anti-COVID-19, Jahanshahlu [/bib_ref] [bib_ref] The immune system as a target for therapy of SARS-CoV-2: a systematic..., Mansourabadi [/bib_ref] [bib_ref] Immunotherapy for SARS-CoV-2: potential opportunities, Pashaei [/bib_ref] [bib_ref] Efficacy of plasmapheresis and immunoglobulin replacement therapy (IVIG) on patients with COVID-19...., Pourahmad [/bib_ref] [bib_ref] Towards treatment planning of COVID-19: rationale and hypothesis for the use of..., Saghazadeh [/bib_ref] [bib_ref] Targeted therapy strategies against SARS-CoV-2 cell entry mechanisms: a systematic review of..., Seyedpour [/bib_ref] [bib_ref] The possible double-edged sword effects of vitamin D on COVID-19: a hypothesis, Shojaeefar [/bib_ref].
The ongoing COVID-19 pandemic has led the scientific community to consider repurposing previously approved treatments such as convalescent plasma, antivirals like IFN, and LPV/RTV, and the clinical reapplication of the experience learned from previous global epidemics caused by hCoVs [bib_ref] Rapid repurposing of drugs for COVID-19, Guy [/bib_ref]. The present research has systematically investigated the efficacy of combinational or mere IFN therapy. We have reviewed the clinical literature regarding the clinical efficacy of IFN for three deadly human CoVs by analyzing the mortality, discharge, CXR presentations, onset-to-treatment duration, ADEs, and other clinically essential outcomes.
Understandably, mortality is of high clinical interest. Mortality in COVID-19 and SARS cases was not significantly affected by IFN therapy, as studies reported no mortality in all study subgroups of similar mortality. Moreover, a poor-quality cohort of SARS patients showed lower mortality and faster CXR improvement in patients receiving IFN-Alfacon-1 compared to IFN-Alfacon-1 + corticosteroids group. However, the results of this uncontrolled study should be taken with its small sample size and lack of randomization in mind. The higher discharge was indicated in a high-quality trial for combinational IFN therapy with REM vs. IFN alone in COVID-19. However, lower discharge rates for taking RBV/IFN were indicated in a poor-quality cohort. In addition to a lack of high-quality evidence backing up the use of RBV/IFN for COVID-19 patients, the calculated effect size for six MERS-CoV studies shows that IFN/RBV treatment did not prove beneficial compared with no RBV/IFN in terms of mortality. Cytokine storm in COVID-19 cases has been known for inducing a destructive immune response and is possibly responsible for unfavored clinical outcomes in COVID-19 [bib_ref] COVID-19: Pathogenesis, Cytokine Storm and Therapeutic Potential of Interferons, Nile [/bib_ref]. For COVID-19, inflammatory cytokines (e.g., IL-6, TNF-α, and CRP) were lower in patients who received IFN with or without ARB. The quality of this study was good. Moreover, a randomized trial of a high risk of bias indicated no difference between inflammatory cytokine levels. Furthermore, our results should be interpreted in light of competing interests of included literature.
Also, a comparison between the anti-inflammatory potential of two IFN types was not significant in MERS patients. A wide range of chest radiography presentations was found in both COVID-19 and MERS patients. Interestingly, a moderately biased trial by Cai et al. showed a significant improvement in CXRs in FPV + IFN-α group vs. LPV/RTV +IFN-α group (P = 0.004), while also showing significantly fewer ADEs (P < 0.001). The median for onset-to-treat times was mostly under two weeks for both MERS and COVID-19. Interestingly, combination ARB + IFN treatment was clinically effective in a cohort via reducing inflammatory cytokines despite the relatively long onset-to-treatment interval (days, median (Q1,Q3)) 17.0 (10.0, 22.0). ADEs were mostly reported for IFN in combination with other antivirals. Therefore, despite that some studies showed certain combinations are less likely to result in ADEs, they were inconclusive for the use of IFN.
Studies strictly including patients according to a strictly fatal or nonfatal outcome do not help compare drug effects. In general, we indicate that although IFN has been commonly given in combination with antiviral therapies (e.g., with RBV in MERS cases), most studies have not reported a definitive benefit for the inclusion of IFN in administered regimens. Therefore, we suggest that more placebo-controlled RCTs with larger populations are required to clarify further the efficacy of IFN for a reduction in improving clinical status, and more importantly, mortality in COVID-19 patients. Also, we recommend that in order to reduce bias and increase usability in practice, comparative observational studies should control for confounding factors, especially severity.
The major restriction in our synthesis was the high risk of bias in many observational studies. Also, most articles were observational studies, many of which were case reports or case series, and did not include a control group. Also, many cohort studies did not control any confounders, resulting in a high risk of bias. Most studies used a combination of pharmacological and non-pharmacological treatments, and utilized highly varied administration protocols along with IFN therapy or did not report outcomes classified by receiving IFN, complicating the distinguish of IFN-related harms or benefits from other interventions. MERS studies mostly reported RBV/IFN treatment groups, which does not assess the net effect of IFN therapy. Despite calling authors, six of SARS studies could not be retrieved due to the unavailability of the fulltext. A limitation in SARS studies was using solely clinical criteria for inclusion rather than confirmed laboratory results. This may lead to the dampening of any actual treatment effects of antiviral therapeutics. We highlighted our assessments in light of lessons that can be learned from past CoVs, as they share principal similarities with the novel SARS-CoV2 [bib_ref] Commentary: SARS, MERS and COVID-19-new threats; old lessons, Gilbert [/bib_ref] [bib_ref] The SARS, MERS and novel coronavirus (COVID-19) epidemics, the newest and biggest..., Peeri [/bib_ref]. However, though IFN treatments may hold significant potential for the management of hospitalized COVID-19 patients, it is challenging to approximate their worldwide acceptance in regards to evidence of this treatment, irrespective of the efficacy of such therapeutics in past outbreaks.
# Conclusion
In conclusion, the present systematic review reveals that the efficacy of IFN alone has not been investigated sufficiently for three deadly human CoVs. Still, we found that combination therapy of IFN with antivirals such as FPV, ARB, REM, or corticosteroids can have potential benefits (e.g., faster CXR improvement, lower level of inflammatory cytokines). These potentials need to be tested in larger RCTs. Also, the data regarding mortality, a crucially determining clinical outcome, seem insufficient for assessing treatment efficacy. Further investigation considering potential benefits and harms (e.g., ADEs) discussed in the present research can shed light on the path, leading to more successful conclusive trials in the strict time researchers possess during rapidly evolving outbreaks.
It is notable that many of the available therapeutic options are not specific to the COVID-19 condition and the death tolls are rising . This lack of specificity has brought about numerous efforts towards understanding the origin of the virus [bib_ref] Angiotensinconverting enzyme 2 (ACE2) receptor and SARS-CoV-2: potential therapeutic targeting, Sharifkashani [/bib_ref] and discovering more targeted approaches for treatment of the disease [bib_ref] Death due to COVID-19 in an infant with combined immunodeficiencies, Ahanchian [/bib_ref] [bib_ref] Lymphopenia in COVID-19: therapeutic opportunities, Fathi [/bib_ref] [bib_ref] The immune system as a target for therapy of SARS-CoV-2: a systematic..., Mansourabadi [/bib_ref] [bib_ref] COVID-19 and picotechnology: potential opportunities, Rabiee [/bib_ref] [bib_ref] COVID-19 and medical biotechnology, Rezaei [/bib_ref] [bib_ref] Towards treatment planning of COVID-19: rationale and hypothesis for the use of..., Saghazadeh [/bib_ref] [bib_ref] Targeted therapy strategies against SARS-CoV-2 cell entry mechanisms: a systematic review of..., Seyedpour [/bib_ref] [bib_ref] Angiotensinconverting enzyme 2 (ACE2) receptor and SARS-CoV-2: potential therapeutic targeting, Sharifkashani [/bib_ref] , and the management of comorbid diseases [bib_ref] Challenges for management of immune thrombocytopenia during COVID-19 pandemic, Sahu [/bib_ref]. Definitely, such efforts need great scientific collaborations to occur and get presented [bib_ref] Borderless collaboration is needed for COVID-19-A disease that knows no borders, Mohamed [/bib_ref] [bib_ref] All together to fight COVID-19, Momtazmanesh [/bib_ref] [bib_ref] The urgent need for integrated science to fight COVID-19 pandemic and beyond, Moradian [/bib_ref]. During this pandemic in which people are at risk of infection and re-infection [bib_ref] COVID-19 affects healthy pediatricians more than pediatric patients, Rezaei [/bib_ref] and social distancing is the most important method of prevention, utilization of hybrid methods for holding scientific events might help the data to be shared and the information to be exchanged.
## Ethics approval and consent to participate
Not applicable.
## Consent for publication
All authors have read and approved the final version of the manuscript.
## Availability of data and materials
Not applicable.
# Funding
There is no funding for the present study.
# Authors' contributions
K.S. conceptualized the study, performed data curation and analysis, and prepared the initial draft. S.Y. conceptualized the study, performed data curation, and prepared the initial draft. M.B. conceptualized the study and performed data curation. E.H. conceptualized the study and performed data curation. M.G. conceptualized the study and performed data curation. A.S. conceptualized the study, designed the project, and prepared the final draft. N.R. conceptualized the study, supervised the project, and critically appraised the manuscript.
## Declaration of competing interest
The authors declare that they have no competing interests.
[fig] 2: Hung IF-N, Lung K-C, Tso EY-K, Liu R, Chung TW-H, Chu M-Y, et al. Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of patients admitted to hospital with COVID-19: an open-label, randomised, phase 2 trial. The Lancet. 2020. 3. Wang Y, Zhang D, Du G, Du R, Zhao J, Jin Y, et al. Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial. The Lancet. 2020. 4. Zhao Z, Zhang F, Xu M, Huang K, Zhong W, Cai W, et al. Description and clinical treatment of an early outbreak of severe acute respiratory syndrome (SARS) in Guangzhou, PR China. Journal of medical microbiology. 2003;52(8):715-20.2020a;Huang et al., 2020b;Hung et al., 2020;Jiang et al., 2020;Jin et al., 2020;Khalid et al., 2016;Kim et al., 2017;Kim et al., 2016;Lee et al., 2017;Liao et al., 2020;Liu et al., 2020a;Liu et al., 2020b;Lo et al., 2020;Oh et al., 2015;Pan et al., 2020;Qiu et al., 2020;Rhee et al., 2016;Rui et al., 2020;Shalhoub et al., 2018;Shalhoub et al., 2015;Sherbini et al., 2017;Sun et al., 2020;Wang et al., 2020c;Xiao-Wei et al., 2020;Yu et al., 2020;Yuan et al., 2020;Zhao et al., 2003;Zhou et al., 2020b). COVID-19, MERS, and SARS studies mentioned 20Liao et al., 2020;Liu et al., 2020a;Lo et al., 2020;Qiu et al., 2020;Rui et al., 2020;Wang et al., [/fig]
[fig] Figure 3: Quantitative analysis for COVID-19 severity and IFN administration vs. placebo. [/fig]
[fig] Figure 4: Quantitative analysis for MERS-CoV mortality and IFN/RBV administration vs. no IFN/RBV. [/fig]
[fig] -: byYuan et al. showed IFN treatment durations varied significantly among severe, moderate, and mild cases (p = 0.al. did not find any significant difference in inflammatory profile.(2 RCTs) ⨁⨁⨁x MODERATE e In a study with FPV + IFNα and LPV/RTV + IFNα arms, Cai et al. did not show a significant different in IL6 and CRP. al. showed a higher efficacy for IFN and its combination therapies vs. therapy without IFN in reducing inflammatory elements.Hung et al. showed LPV/RTV + RBV + IFNβ group had a reduced hospitalization time compared to LPV/ RTV. Wang et al. included two arms both of which used IFN, and detected no difference in hospitalization duration. (2 RCTs) ⨁⨁⨁⨁ HIGH No significant effects were detected in 3 studies. Xu et al. 15 (45.5%) cases with symptoms lasting longer than ten days and 19 (65.5%) cases with symptoms lasting shorter than or equal to 10 days received IFN alone or in combination with ARB, RBV, or LPV/RTV. In the study by Zhou et al. IFN, IFN + ARB, and ARB arms were not significantly different in regards to symptom onset to hospital admission (p = 0.87). Yuan et al. showed that the duration of hospitalization was significantly correlated with PCR negative conversion durations in the IFNα + LPV/RTV + RBV group (p = 0.0215), as well as the IFNα + LPV/RTV group (p = 0.al. showed FPV + IFNα was significantly linked to less total ADEs compared with LPV/RTV + IFNα treatment (p = 0.001). [/fig]
[table] Table 3: Characteristics of included SARS-CoV studies (n = 2). Abbreviations: CoV: coronavirus, IFN: interferon, igG: immunoglobulin G, IM: intramuscular, IVIG: intravenous immune globulin, M: mean, O: other formats, R: range, RBV: ribavirin, SARS: severe acute respiratory syndrome, U: unit. a Age of participants is reported as reported in each study. Estimated mean values may be found in (supplementary material). [/table]
[table] Table 4: Critical appraisal for cross-sectional studies using the NIH tool. Abbreviations: CD: Cannot be determined, NR: Not Reported, severe acute respiratory syndrome coronavirus. 1. Pan L, Mu M, Yang P, Sun Y, Wang R, Yan J, et al. Clinical characteristics of COVID-19 patients with digestive symptoms in Hubei, China: a descriptive, cross-sectional, multicenter study. The American journal of gastroenterology. 2020;115. [/table]
[table] Table 6: Critical appraisal for case-series using NIH tool. Abbreviations: CD: Cannot be determined, MERS-CoV: Middle-Eastern respiratory syndrome coronavirus, NA: Not applicable, NR: Not Reported, SARS-CoV: severe acute respiratory syndrome coronavirus. Yu N, Li W, Kang Q, Xiong Z, Wang S, Lin X, et al. Clinical features and obstetric and neonatal outcomes of pregnant patients with COVID-19 in Wuhan, China: a retrospective, single-centre, descriptive study. The Lancet Infectious Diseases. 2020. 6. Rui Z, Yunguang L, Yanrong L, Ning L, Qiulian L, Youling L, et al. Clinical characteristics of 28 patients with novel coronavirus pneumonia. Chinese Journal of Infectious Diseases. 2020(12):006-. 7. Jian-ya G. Clinical characteristics of 51 patients discharged from hospital with COVID-19 in Chongqing, China. medRxiv. 2020. 8. Liu Y, Yang Y, Zhang C, Huang F, Wang F, Yuan J, et al. Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads and lung injury. [/table]
[table] Table 7: Critical appraisal for included Cohorts via the NOS tool. [/table]
[table] Table 8: Critical appraisal for case-reports via the tool recently suggested by Murad et al. (1). Murad MH, Sultan S, Haffar S, Bazerbachi F. Methodological quality and synthesis of case series and case reports. BMJ evidence-based medicine. 2018;23(2):60-3. 2. Wang L, Xu X, Ruan J, Lin S, Jiang J, Ye H. Quadruple therapy for asymptomatic COVID-19 infection patients. Expert Review of Anti-infective Therapy. 2020:1-8. 3. Choi JY, Oh JO, Ahn JY, Choi H, Kim JH, Seong H, et al. Absence of neutralizing activity in serum 1 year after successful treatment with antivirals and recovery from mers in south korea. Clinical and Experimental Vaccine Research. 2019;8(1):86-8. 4. Al-Tawfiq JA, Hinedi K. The calm before the storm: clinical observations of Middle East respiratory syndrome (MERS) patients. Journal of Chemotherapy. Kim J-E, Heo J-H, Kim H-O, Song S-H, Park S-S, Park T-H, et al. Neurological Complications during Treatment of Middle East Respiratory Syndrome. [/table]
[table] Table 9: Summary of findings for COVID-10 severity and IFN administration. [/table]
[table] Table 10: Summary of Findings table for MERS-CoV mortality and IFN/RBV administration. [/table]
[table] Table 13: GRADE assessment for narrative synthesis outcomes in SARS studies.Combination or non-combination IFN treatment compared to No IFN/A different treatment protocol including IFN for SARS [/table]
|
SARS-CoV-2 and Acute Cerebrovascular Events: An Overview
Citation: Ghasemi, M.; Umeton, R.P.; Keyhanian, K.; Mohit, B.; Rahimian, N.; Eshaghhosseiny, N.; Davoudi, V. SARS-CoV-2 and Acute Cerebrovascular Events: An Overview.
# Introduction
The coronavirus disease 2019 (COVID-19) pandemic due to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the first one of the recorded pandemics that has caused a global burden on society and healthcare professionals. As of the date of writing this article, on 7 June 2021, over 173.41 million cases have been reported across 188 countries and territories, resulting in more than 3.73 million deaths, and over 2.13 billion people have been vaccinated [bib_ref] An interactive web-based dashboard to track COVID-19 in real time, Dong [/bib_ref]. The clinical manifestations of this disease are broad, ranging from asymptomatic cases to those with the severe symptomatic disease, with a case fatality rate of 2.3%. The mortality is higher in elderly individuals, patients with medical comorbidities, and those with immunocompromised conditions [bib_ref] Clinical course and risk factors for mortality of adult inpatients with COVID-19..., Zhou [/bib_ref]. While the primary mode of attack of the SARS-CoV-2 is through the respiratory pathways, early in the pandemic, reports from Wuhan, China, revealed that patients with COVID-19 might also develop neurologic symptoms (e.g., headache, dizziness and myalgia) [bib_ref] The neuroinvasive potential of SARS-CoV2 may play a role in the respiratory..., Li [/bib_ref]. Ever since, it has been found worldwide that neurological complications affecting both the central and peripheral nervous system (CNS and PNS, respectively) may occur in a considerable number of patients with COVID-19 [bib_ref] Clinical course and risk factors for mortality of adult inpatients with COVID-19..., Zhou [/bib_ref] [bib_ref] The neuroinvasive potential of SARS-CoV2 may play a role in the respiratory..., Li [/bib_ref] [bib_ref] Potential Neurologic Manifestations of COVID-19, Nordvig [/bib_ref] [bib_ref] SARS-CoV-2 and nervous system: From pathogenesis to clinical manifestation, Keyhanian [/bib_ref]. The direct invasion of the nervous system by SARS-CoV-2 through the olfactory nerve, retrograde axonal transport, the gut-brain axis, or hematogenous spread has been suggested [bib_ref] SARS-CoV-2 and nervous system: From pathogenesis to clinical manifestation, Keyhanian [/bib_ref] [bib_ref] COVID-19 and management of neuroimmunological disorders, Hartung [/bib_ref] [bib_ref] The Role of the Gastrointestinal System in Neuroinvasion by SARS-CoV-2, Xu [/bib_ref]. Critically ill COVID-19 patients admitted to the intensive care unit (ICU) may have additional risk factors for nervous system involvement, which include deep sedation and prolonged mechanical ventilation related to severe prolonged hypoxemia, immobility, and critical illness myopathy or neuropathy related to prolonged hospitalization, social isolation and delirium [bib_ref] Insights into neurological dysfunction of critically ill COVID-19 patients, Deana [/bib_ref]. A correlation between SARS-CoV-2-related acute lung injury and brain hypoxia has been recently described, which may play an important role in the neurological dysfunction following SARS-CoV-2 infection [bib_ref] Acute Lung Injury Is an Independent Risk Factor for Brain Hypoxia after..., Oddo [/bib_ref].
Recent investigations have also indicated that some patients with COVID-19 may present with acute cerebrovascular events such as stroke [bib_ref] Characteristics of ischaemic stroke associated with COVID-19, Beyrouti [/bib_ref] [bib_ref] Status of SARS-CoV-2 in cerebrospinal fluid of patients with COVID-19 and stroke, Saiegh [/bib_ref] and cerebral venous thrombosis [bib_ref] Cerebral venous thrombosis associated with COVID-19 infection: Causality or coincidence?, Poillon [/bib_ref]. Although the mechanisms underlying such complications remain to be fully elucidated, the hypercoagulability state, hyper-inflammation, cytokine storm and cerebral endothelial dysfunction may play crucial roles [bib_ref] Dynamic changes of markers of endothelial function, hemostasis, fibrinolysis and inflammation in..., Bielosludtseva [/bib_ref] [bib_ref] The cytokine release syndrome (CRS) of severe COVID-19 and Interleukin-6 receptor (IL-6R)..., Zhang [/bib_ref] [bib_ref] Coronavirus Disease 2019 and Stroke: Clinical Manifestations and Pathophysiological Insights, Divani [/bib_ref]. In this review, we discuss the possible mechanisms underlying the acute cerebrovascular events related to SARS-CoV-2 infection, and also review the current epidemiological studies and case reports of neurovascular complications in patients with COVID-19 and relevant therapeutic approaches that have been considered worldwide.
## Hypercoagulability related to sars-cov-2
One of the important findings related to the SARS-CoV-2 infection is a widespread observation of the hypercoagulable state indicated by elevated D-dimer levels, the prolongation of the prothrombin time (PT), the activated partial thromboplastin time (aPTT), and abnormal platelet counts [bib_ref] Hypercoagulation and Antithrombotic Treatment in Coronavirus 2019: A New Challenge, Violi [/bib_ref]. Both thrombocytopenia and elevated D-dimers can be justified by the disproportionate activation of the coagulation cascade and the use of its substrates; however, the pathophysiology of SARS-CoV-2-related coagulopathy is still debatable.
While pneumonia itself can cause inflammation and a hypercoagulable state, cytokine release syndrome (CRS)-and macrophage activation-like syndrome (MAL)-like phenomena are also likely to play important roles [bib_ref] Dynamic changes of markers of endothelial function, hemostasis, fibrinolysis and inflammation in..., Bielosludtseva [/bib_ref] [bib_ref] The cytokine release syndrome (CRS) of severe COVID-19 and Interleukin-6 receptor (IL-6R)..., Zhang [/bib_ref] [bib_ref] Hypercoagulation and Antithrombotic Treatment in Coronavirus 2019: A New Challenge, Violi [/bib_ref]. When endothelial cells are damaged, generally, sub-endothelial cells-which are chromogenic-are exposed. The subendothelial cells contain Von Willebrand factor (VWF) and other thrombophilic proteins. Activated endothelial cells, along with VWF, will cause platelet aggregation and platelet plug formation as the primary homeostasis response [fig_ref] Figure 1: Figure 1 [/fig_ref]. Secondarily, the coagulation cascade is activated, and it involves both extrinsic and intrinsic pathways, followed by a common pathway [bib_ref] Overview of the coagulation system, Palta [/bib_ref]. When the coagulopathy results from hyper inflammation and not an endothelial cell injury, the activated cascade would be an extrinsic pathway by the activation of a tissue factor or CD142 [bib_ref] Advance in the Management of Sepsis-Induced Coagulopathy and Disseminated Intravascular Coagulation, Iba [/bib_ref]. Tissue factor is expressed on mononuclear cells in response to interleukin (IL)-6 and other inflammatory cytokines, and will activate the extrinsic pathway [bib_ref] The immunopathology of sepsis and potential therapeutic targets, Van Der Poll [/bib_ref]. Furthermore, inflammatory cytokines impose an inhibitory effect on anticoagulation regulators like tissue factor pathway inhibitors and ADPase [bib_ref] Monocytes, neutrophils, and platelets cooperate to initiate and propagate venous thrombosis in..., Brühl [/bib_ref] [bib_ref] Aberrant coagulation causes a hyper-inflammatory response in severe influenza pneumonia, Yang [/bib_ref]. Viral infection pro-coagulopathy seems to be both dependent on endothelial cells and innate immunity by the hyper-activation of toll-like receptors (TLRs) along the surface of monocytes, macrophages, dendritic cells and fibroblasts. Another possible contributor to this hypercoagulable phenomenon is the formation of acute reactive oxygen species and oxidized phospholipids due to acute lung injury, which seems to initiate the Toll-like receptor 4 (TLR4)−TRIF (TIR-domain-containing adapter-inducing interferon-β)−TRAF6 (TNF receptor-associated factor 6)−NF-κB (nuclear factor κ-light-chain-enhancer of activated B cells) pathway [bib_ref] Pathological inflammation in patients with COVID-19: A key role for monocytes and..., Merad [/bib_ref] [bib_ref] Identification of Oxidative Stress and Toll-like Receptor 4 Signaling as a Key..., Imai [/bib_ref]. The downstream pathway for both the hyperactivation of TLRs in response to viral infection and acute lung injury involves NF-κB activation. As a result of the NF-κB activation, there will be more IL-6 and TNF-α, which are also bi-product of CRS and MAL reactions [bib_ref] The cytokine release syndrome (CRS) of severe COVID-19 and Interleukin-6 receptor (IL-6R)..., Zhang [/bib_ref] [bib_ref] Procoagulant activity during viral infections, Subramaniam [/bib_ref]. Coagulopathy was previously observed in infection with other Coronaviridae viruses, including SARS and Middle East respiratory syndrome (MERS) [bib_ref] Pathological inflammation in patients with COVID-19: A key role for monocytes and..., Merad [/bib_ref] [bib_ref] Coagulation disorders in coronavirus infected patients: COVID-19, SARS-CoV-1, MERS-CoV and lessons from..., Giannis [/bib_ref]. It has also been suggested that COVID-19 may induce antiphospholipid antibodies, but usually these antibodies are transient and found not to be pathogenic [bib_ref] Antiphospholipid antibodies in patients with COVID-19: A relevant observation?, Devreese [/bib_ref]. Schematic representation of (A) the normal coagulation cascade triggered by an injury to the endothelial cells and (B) the hypercoagulability state related to SARS-CoV-2 infection. Generally, with an injury to the endothelial cell, the first response is platelet aggregation, followed by the activation of both the extrinsic and intrinsic pathways that lead to the generation of factor Xa and common pathway initiation. The product of the secondary coagulation cascade is a fibrin clot which will enhance platelet aggregates. ADPase and tissue factor (TF) pathway inhibitors (TFPI), as well as plasminogen, act as regulatory anticoagulants. The extrinsic and common pathways are initiated due to the increased TF and von Willebrand factor (VWF) in response to inflammatory cytokines such as interleukin (IL)-6 and the expression of TF by mononuclear cells in response to IL-6, which is elevated in SARS-CoV-2 infection. TFPI and ADPase are inhibited by the cytokines as well.
Although there are some prospective studies currently looking at the incidence of thrombotic events, early studies have already confirmed the increased frequency of intravascular thrombosis leading to pulmonary embolism, myocardial infarction, ischemic strokes and even cerebral venous sinus thrombosis. A thrombotic event has sometimes been reported as the first presentation of COVID-19 infection [bib_ref] Cerebral Venous Sinus Thrombosis as a Presentation of COVID-19, Hughes [/bib_ref] [bib_ref] Confirmation of the high cumulative incidence of thrombotic complications in critically ill..., Klok [/bib_ref]. However, whether prophylactic anticoagulation for severe SARS-CoV-2 infection is beneficial still has to be considered according to common concurrent thrombocytopenia [bib_ref] Dynamic changes of markers of endothelial function, hemostasis, fibrinolysis and inflammation in..., Bielosludtseva [/bib_ref] [bib_ref] Hypercoagulation and Antithrombotic Treatment in Coronavirus 2019: A New Challenge, Violi [/bib_ref].
## Sars-cov-2 and angiotensin-converting enzyme 2
SARS-CoV-2 has spike (S) glycoproteins on its outer envelope, which have a strong affinity toward the human angiotensin-converting enzyme 2 (ACE-2) as the host cell receptor [bib_ref] SARS-CoV-2 Receptor ACE2 Is an Interferon-Stimulated Gene in Human Airway Epithelial Cells..., Ziegler [/bib_ref] [bib_ref] Tissue distribution of ACE2 protein, the functional receptor for SARS coronavirus. A..., Hamming [/bib_ref]. The binding of SARS-CoV-2 to ACE-2 is a crucial element for viral infectivity and multi-organ damage. ACE-2 is expressed in various human tissues such as CNS (glial cells and neurons), skeletal muscle, the gastrointestinal tract and endothelial cells [bib_ref] Tissue distribution of ACE2 protein, the functional receptor for SARS coronavirus. A..., Hamming [/bib_ref]. In the cerebral vasculature, endothelial ACE-2, as part of the renin-angiotensin system (RAS), plays an important role in the modulation of cerebral blood flow. The key components of RAS are angiotensinogen, renin, angiotensin I (Ang I), angiotensin II (Ang II), ACE, ACE-2, Ang type-1 receptor (AT1R), Ang type-2 receptor (AT2R) and Mas receptor [fig_ref] Figure 2: Figure 2 [/fig_ref]. Classically, the Ang II that is produced from Ang I by ACE activity mediates vasoconstriction, neuroinflammation and oxidative stress through the activation of AT1R and AT2R. Alternatively, Ang II can be converted to Ang-(1-7) by ACE-2 activity, which in turn activates the Mas receptor, mediating vasodilation, anti-inflammatory and antioxidant responses [bib_ref] Therapeutic potential of the renin angiotensin system in ischaemic stroke, Arroja [/bib_ref]. Schematic effects of SARS-CoV-2 on the renin-angiotensin system (RAS) in the cerebral vasculature. After the binding of the SARS-CoV-2 spike (s) glycoprotein to angiotensin converting enzyme 2 (ACE-2), the virus enters the cell. The virus also downregulates ACE-2 and competes with angiotensin II (Ang II) for binding to ACE-2, which ultimately decreases the activity of the ACE-2-Ang-(1-7)-Mas receptor (alternative) axis. This also leads to the greater activation of the ACE-Ang II-AT1R (classical) axis. The outcome of such events is an aberrant renin-angiotensin system (RAS), causing vasoconstriction, inflammation, oxidative stress, and thrombogenesis, causing ischemic stroke in relation to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. SARS-COV-2. TMPRSS2, Transmembrane protease, serine 2; AT1R: angiotensin 1 receptor; AT2R: angiotensin 2 receptor.
The overactivation of ACE/Ang II/AT1R/AT2R or the dysfunction of the ACE-2/Ang (1-7)-Mas receptor axis may contribute to the pathogenesis of acute ischemic stroke via increased vasoconstriction, oxidative stress and vascular inflammation (i.e., vasculitis) [bib_ref] Therapeutic potential of the renin angiotensin system in ischaemic stroke, Arroja [/bib_ref]. In SARS-CoV-2 infection, binding to ACE-2 may downregulate ACE-2 [bib_ref] Angiotensin-converting enzyme 2 (ACE2) as a SARS-CoV-2 receptor: Molecular mechanisms and potential..., Zhang [/bib_ref] , leading to excess ACE-mediated Ang II production and lower ACE-2-mediated Ang-(1-7) production [bib_ref] Coronavirus Disease 2019 and Stroke: Clinical Manifestations and Pathophysiological Insights, Divani [/bib_ref]. This SARS-CoV-2-induced imbalance between the classical and alternative RAS axes ultimately promotes ischemia via increased cerebral vasoconstriction, hyperinflammation and oxidative stress [bib_ref] Coronavirus Disease 2019 and Stroke: Clinical Manifestations and Pathophysiological Insights, Divani [/bib_ref]. Ang II promotes thrombosis by increasing the release and secretion of plasminogen activator inhibitor type 1 (PAI-1), and enhances tissue factor (TF) expression [bib_ref] Antithrombotic Effect of Captopril and Losartan Is Mediated by Angiotensin, Kucharewicz [/bib_ref]. In contrast, the activation of the angiotensin-converting enzyme (ACE)2/angiotensin-(1-7)/Mas receptor would cause antithrombotic activity. SARS-CoV-2 decreases the activation of ACE2; the result is an imbalance between the classical and alternative RAS axes, ultimately promoting ischemia via increased cerebral vasoconstriction, hyper-inflammation, oxidative stress and thrombogenesis [bib_ref] Coronavirus Disease 2019 and Stroke: Clinical Manifestations and Pathophysiological Insights, Divani [/bib_ref]. These data raise the possibility that recombinant human ACE-2 might be beneficial in preventing ischemic stroke in COVID-19 patients with known stroke risk factors.
## Acute cerebrovascular events in covid-19
Prior articles indicate that viral respiratory infections are independent risk factors for both ischemic and hemorrhagic strokes [bib_ref] Nervous system involvement after infection with COVID-19 and other coronaviruses, Wu [/bib_ref]. Acute cerebrovascular events have been reported as one of the neurological complications that can occur in COVID-19 patients [bib_ref] Seizures associated with coronavirus infections, Asadi-Pooya [/bib_ref] [bib_ref] Clinical characteristics and outcomes of inpatients with neurologic disease and COVID-19 in..., Benussi [/bib_ref] ; there is, overall, a propensity towards the occlusion of (i) large vessels (e.g., internal carotid, middle cerebral (M1 and M2 segments), or basilar arteries), (ii) multi-territory vessels, or (iii) uncommon vessels (e.g., pericallosal artery [bib_ref] Stroke in patients with SARS-CoV-2 infection: Case series, Morassi [/bib_ref] [bib_ref] Stroke in patients with COVID-19: Clinical and neuroimaging characteristics, Vogrig [/bib_ref]. On the other hand, intracerebral hemorrhage, cerebral venous thrombosis and small-vessel brain disease develop less frequently in COVID-19 patients [bib_ref] Stroke in patients with COVID-19: Clinical and neuroimaging characteristics, Vogrig [/bib_ref]. Several cases with atypical neurovascular presentations have also been reported, including bilateral carotid artery dissection [bib_ref] Bilateral carotid artery dissection in a SARS-CoV-2 infected patient: Causality or coincidence?, Morassi [/bib_ref] , posterior reversible encephalopathy syndrome (PRES) [bib_ref] Multifocal laminar cortical brain lesions: A consistent MRI finding in neuro-COVID-19 patients, Anzalone [/bib_ref] , and vasculitis [bib_ref] Vessel Wall Enhancement and Focal Cerebral Arteriopathy in a Pediatric Patient with..., Gulko [/bib_ref] [bib_ref] Focal Cerebral Arteriopathy in a Pediatric Patient with COVID-19, Mirzaee [/bib_ref]. The pathophysiology, still unclear, is possibly related to the direct damage of the vessel mediated by the virus once it invades the CNS [bib_ref] Neurological manifestations in COVID-19 caused by SARS-CoV-2, Baig [/bib_ref] , or it could be related to the development of underlying coagulopathy and thromboembolisms, as we discussed earlier [bib_ref] Confirmation of the high cumulative incidence of thrombotic complications in critically ill..., Klok [/bib_ref] [bib_ref] Coagulopathy and Antiphospholipid Antibodies in Patients with COVID-19, Zhang [/bib_ref] [bib_ref] Venous and arterial thromboembolic complications in COVID-19 patients admitted to an academic..., Lodigiani [/bib_ref]. Another suggested mechanism is cardioembolic stroke from the direct damage of the myocardial cells, as evidenced by the cardiac dysfunction and arrhythmias in these patients [bib_ref] Cardiovascular Implications of Fatal Outcomes of Patients with Coronavirus Disease 2019 (COVID-19), Guo [/bib_ref]. Accordingly, myocardial cells express ACE-2 receptors abundantly, and are then vulnerable to SARS-CoV-2 infection [bib_ref] Tissue distribution of ACE2 protein, the functional receptor for SARS coronavirus. A..., Hamming [/bib_ref].
In a retrospective study that included 214 patients with confirmed COVID-19 infection, about 6% presented with acute cerebrovascular events, mainly ischemic strokes. Stroke symptoms tend to appear later during the hospitalization, a median of 10 days after the onset of symptoms, and this was also confirmed by a larger retrospective study [bib_ref] Incidence and consequences of systemic arterial thrombotic events in COVID-19 patients, Cantador [/bib_ref]. These patients seem to have a more severe infection with higher levels of inflammatory markers and higher D-dimer levels, older age, more comorbidities (hypertension in particular) and fewer typical symptoms associated with COVID-19 [bib_ref] Neurologic Manifestations of Hospitalized Patients with Coronavirus Disease, Mao [/bib_ref]. Indeed, for many COVID-19 patients presenting with acute strokes or other neurological manifestations, the diagnosis of infection is made after the hospital admission. The current recommendations from the American Heart Association (AHA) and American Stroke Association (ASA) include the use of personal protective equipment (PPE) for all of the stroke teams at the time of stroke code activation, as many stroke patients are unable to provide the history and information for appropriate COVID-19 screening. It is indeed suggested to treat every code stroke patient as being potentially affected by the infection in order to avoid any delay in trying to understand the infection status, following the same treating guidelines available for non-COVID-19 patients [bib_ref] Acute Stroke Care in the Coronavirus Disease, Dafer [/bib_ref]. A dedicated track for the triage and management of suspected or proven COVID-19 patients with stroke-like symptoms was also suggested and implemented in Italy with a mobile CT scan unit [bib_ref] Acute stroke management pathway during Coronavirus-19 pandemic, Baracchini [/bib_ref]. The patients eligible for neurointerventional procedures should be treated accordingly, with the minimum number of staff in the angio suite and restricted access for essential staff, only ensuring the quality control of the negative pressure environment and following appropriate precaution protocols [bib_ref] European Society of Minimally Invasive Neurological Therapy (ESMINT) recommendations for optimal interventional..., Aggour [/bib_ref] [bib_ref] Neurointervention for emergent large vessel occlusion during the COVID-19 pandemic, Fiorella [/bib_ref]. After the appropriate treatment, stroke patients should be admitted to a dedicated ward or ICU units where possible, and stroke teams should guide staff familiar with managing acute ischemic or hemorrhaging stroke patients [bib_ref] Acute Stroke Care in the Coronavirus Disease, Dafer [/bib_ref].
## Ischemic stroke
Acute ischemic stroke appears to be the most common form of stroke seen in patients with COVID-19. The initial retrospective case reports from Wuhan in China reported six cases (2.34%) of stroke among the 214 patients analyzed, five of which were ischemic in nature [bib_ref] Neurologic Manifestations of Hospitalized Patients with Coronavirus Disease, Mao [/bib_ref]. Another study from Italy reported nine ischemic strokes (2.5%) among a cohort of 388 patients [bib_ref] Venous and arterial thromboembolic complications in COVID-19 patients admitted to an academic..., Lodigiani [/bib_ref]. Different incidence rates were reported in two large studies. The first is a recent case series of 1419 patients with the diagnosis of COVID-19 admitted in a hospital in Madrid, Spain; it reported a total of 14 patients with systemic arterial thrombotic events (1% incident), of which eight presented with a cerebrovascular event (six with acute ischemic stroke and two with transient ischemic attack) [bib_ref] Incidence and consequences of systemic arterial thrombotic events in COVID-19 patients, Cantador [/bib_ref]. A similar incidence (0.9%) was reported in the second large retrospective study of 3556 COVID-19 positive patients, of which 32 were diagnosed with ischemic stroke, 65.6% were defined as the cryptogenic subtype, and 34.4% were defined as an embolic stroke of undetermined source [bib_ref] SARS-CoV-2 and Stroke in a New York Healthcare System, Yaghi [/bib_ref]. There is a possibility that the total numbers were underestimated because patients with small acute strokes may present without apparent focal neurological symptoms, and may go undiagnosed. Indeed, a case series from France documented three encephalopathic patients, with no signs suggestive of ischemic stroke, of whom the diagnosis was made after undergoing MRI to better address the cause of their encephalopathy [bib_ref] Neurologic Features in Severe SARS-CoV-2 Infection, Helms [/bib_ref]. Furthermore, the difference in the incidence rates could be explained by the different patient populations and larger cohorts. Another case series in Houston (TX, USA) reported a total of 12 patients with COVID-19 who developed stroke, among which 10 cases had an ischemic stroke (including one patient with hemorrhagic transformation), and two had intracerebral hemorrhage [bib_ref] Cerebrovascular Disease in Patients with COVID-19: A Review of the Literature and..., Reddy [/bib_ref]. The inflammatory markers (e.g., D-dimer and IL-6) were elevated in a majority of these cases [bib_ref] Cerebrovascular Disease in Patients with COVID-19: A Review of the Literature and..., Reddy [/bib_ref]. The etiology was an embolic stroke of undetermined source (ESUS, 6 cases), cardioembolic (2), carotid dissection (1), hypertension-related hemorrhage (1), the rupture of mycotic aneurysm related to infectious endocarditis (1), and unknown (one case due to limited workup) [bib_ref] Cerebrovascular Disease in Patients with COVID-19: A Review of the Literature and..., Reddy [/bib_ref].
Stroke patients with COVID-19 infection usually present with a higher National Institutes of Health Stroke Scale (NIHSS) score at admission [bib_ref] SARS-CoV-2 and Stroke in a New York Healthcare System, Yaghi [/bib_ref] [bib_ref] SARS-CoV-2 and Stroke Characteristics, Shahjouei [/bib_ref] , a more severe disease course, immunocompromisation, and with different comorbidities and cardiovascular risk factors [bib_ref] Incidence and consequences of systemic arterial thrombotic events in COVID-19 patients, Cantador [/bib_ref] [bib_ref] Neurologic Manifestations of Hospitalized Patients with Coronavirus Disease, Mao [/bib_ref]. The age range is reported to be usually over 50 years old. However, more recently, a case series from New York City showed five COVID-19 patients younger than 50 affected by a large vessel ischemic stroke presented in the emergency department within a two-week period higher than usual [fig_ref] Table 1: Case reports of acute cerebrovascular events related to COVID-19 [/fig_ref]. Two of the five patients were previously healthy; one had hypertension and hyperlipidemia, another had undiagnosed diabetes, and the last reported patient had a history of prior mild stroke and diabetes [bib_ref] Large-Vessel Stroke as a Presenting Feature of COVID-19 in the Young, Oxley [/bib_ref]. The data from a larger patient cohort from New York City reported stroke in COVID-19 positive patients, mainly in men (71.9%) and white people (70%), with an average age of 62.5 versus 70 in the COVID-19−negative stroke patients. Moreover, patients with COVID-19 and ischemic stroke appeared to have higher mortality than the controls [bib_ref] SARS-CoV-2 and Stroke in a New York Healthcare System, Yaghi [/bib_ref]. A large multicenter study reported stroke characteristics in 432 COVID-19 patients admitted to 71 centers from 17 countries. They observed a considerably higher rate of large vessel occlusions, a much lower rate of small vessel occlusion and lacunar infarction, and a considerable number of young strokes when compared with the population studies before the pandemic [bib_ref] SARS-CoV-2 and Stroke Characteristics, Shahjouei [/bib_ref]. More data and studies on the incidence of stroke in young COVID-19 patients are needed.
A large international multicenter study on 17,799 COVID-19 hospitalized patients reported 156 stroke episodes, 123 (79%) of whom presented with acute ischemic stroke, 27 (17%) of whom had intracranial hemorrhage, and 6 (4%) of whom presented with cerebral venous sinus thrombosis. The mean age for ischemic stroke among hospitalized COVID-19 patients was 68.6 years. Another multicenter prospective cohort study which included 150 patients with COVID-19 related ARDS showed 64 thrombotic complications, two of which were acute ischemic stroke despite anticoagulation [bib_ref] High risk of thrombosis in patients with severe SARS-CoV-2 infection: A multicenter..., Helms [/bib_ref].
Given the hypercoagulable state related to the infection, as a possible cause of ischemic stroke, prophylactic anticoagulation with low molecular weight heparin (LMWH) may be recommended for patients with severe COVID-19, according to the International Society of Thrombosis and Hemostasis (ISTH) [bib_ref] ISTH interim guidance on recognition and management of coagulopathy in COVID-19, Thachil [/bib_ref]. The American Society of Hematology (ASH) guideline panel recently suggested: using prophylactic-intensity over intermediateintensity or therapeutic-intensity anticoagulation for patients with coronavirus disease 2019 (COVID-19)-related critical illness who do not have suspected or confirmed venous thromboembolism (VTE) (conditional recommendation based on very low certainty in the evidence about effects).Higher mortality rates were observed in association with elevated PT and D-dimer levels, and decreased platelet counts and fibrinogen at days 10 and 14 from the onset of symptoms [bib_ref] ISTH interim guidance on recognition and management of coagulopathy in COVID-19, Thachil [/bib_ref]. The monitoring of these parameters can help determine the prognosis and the selection of patients that require admission and aggressive treatments. Interestingly, in a retrospective study that included 449 patients with severe COVID-19 infection and elevated D-dimers, the use of LMWH was associated with lower mortality [bib_ref] Anticoagulant treatment is associated with decreased mortality in severe coronavirus disease 2019..., Tang [/bib_ref]. However, the data on the efficacy of LMWH in preventing venous and arterial thromboembolic complications are conflicting [bib_ref] Venous and arterial thromboembolic complications in COVID-19 patients admitted to an academic..., Lodigiani [/bib_ref] [bib_ref] High risk of thrombosis in patients with severe SARS-CoV-2 infection: A multicenter..., Helms [/bib_ref]. In a case report of six patients who presented with acute ischemic stroke and confirmed COVID-19 infection with associated elevated D-dimer levels (≥1000 µg/L), two patients had ischemic stroke despite therapeutic anticoagulation [bib_ref] Characteristics of ischaemic stroke associated with COVID-19, Beyrouti [/bib_ref]. Data from the same study showed that the primary mechanism underlying the ischemic stroke was large-vessel occlusion, and the stroke usually occurred later in the course of the disease, between days 8 and 24 from the onset of symptoms. Further investigations are warranted to establish the actual need for therapeutic anticoagulation in patients with COVID-19 to reduce the risk of ischemic stroke.
PROTECT COVID (a randomized clinical trial of anticoagulation strategies in COVID-19) is ongoing, comparing the effectiveness of therapeutic versus prophylactic anticoagulation in patients with COVID-19 infection and mild-to-moderate elevations in D-dimer levels greater than 500 ng/mL (clinical trial identifier: NCT04359277) [bib_ref] SARS-CoV-2 and Stroke in a New York Healthcare System, Yaghi [/bib_ref]. Other randomized trials are also ongoing to investigate the anticoagulation benefits in patients with COVID-19 (NCT04362085, NCT04345848, NCT04406389, NCT04528888).
Despite the lack of defined data on the prognosis of strokes related to COVID-19 infection, the overall outcome appears to be poor, as the majority of stroke patients are older and present with severe infection and more comorbidities [bib_ref] Incidence and consequences of systemic arterial thrombotic events in COVID-19 patients, Cantador [/bib_ref] [bib_ref] Acute Stroke Care in the Coronavirus Disease, Dafer [/bib_ref] [bib_ref] Could COVID-19 represent a negative prognostic factor in patients with stroke?, Siniscalchi [/bib_ref] [bib_ref] Potential neurological symptoms of COVID-19. Ther, Wang [/bib_ref]. Nonetheless, mechanical thrombectomy for emergent large vessel occlusion could be justified, as it can improve the outcome and should be offered to all the potential candidates notwithstanding the infectious status [bib_ref] Neurointervention for emergent large vessel occlusion during the COVID-19 pandemic, Fiorella [/bib_ref].
## Hemorrhagic stroke
A small number of stroke patients with COVID-19 infection present with cerebral hemorrhage [fig_ref] Table 1: Case reports of acute cerebrovascular events related to COVID-19 [/fig_ref]. The initial retrospective case series of 214 patients from Wuhan in China [bib_ref] Neurologic Manifestations of Hospitalized Patients with Coronavirus Disease, Mao [/bib_ref] reported only one case of hemorrhagic stroke. Similarly, another retrospective case series, again from Wuhan, reported one hemorrhagic stroke within 13 patients who presented with acute cerebrovascular events. An additional five case reports of hemorrhagic stroke have been published [bib_ref] Neurologic Manifestations of Hospitalized Patients with Coronavirus Disease, Mao [/bib_ref] [bib_ref] Acute Stroke Care in the Coronavirus Disease, Dafer [/bib_ref]. A ruptured aneurysm in the pericallosal region [bib_ref] Letter to editor: Severe brain haemorrhage and concomitant COVID-19 Infection: A neurovascular..., Muhammad [/bib_ref] or posterior-inferior cerebellar artery (PICA) [bib_ref] Status of SARS-CoV-2 in cerebrospinal fluid of patients with COVID-19 and stroke, Saiegh [/bib_ref] was found in two cases. A hypothesis about the underlying pathophysiologic mechanism of cerebral hemorrhage is the reduced expression and function of ACE-2 in SARS-CoV-2 infected cells. ACE-2 is expressed in vascular endothelial cells, and its signaling is involved in the regulation of cerebral blood flow and the reduction of the body's blood pressure. In the case of COVID-19 infection, the signaling is altered with subsequent hypertension and predisposition to the development of hemorrhagic stroke from arterial wall rupture [bib_ref] Stroke in patients with SARS-CoV-2 infection: Case series, Morassi [/bib_ref]. Another possible mechanism is the underlying coagulopathy induced by the infection with thrombocytopenia [bib_ref] Potential neurological symptoms of COVID-19. Ther, Wang [/bib_ref]. Future observations may better clarify the incidence and clinical and laboratory characteristics of COVID-19 patients presenting with hemorrhagic strokes.
## Cerebral venous thrombosis
Cerebral venous thrombosis has been reported in several studies. In a multinational retrospective study, all of the cases of cerebral venous sinus thrombosis (CVST) with COVID-19 infection were collected from the start of the pandemic to the end of June 2020. They reported on 13 post-COVID-19 CVST patients and compared their characteristics with the CVST data obtained before the COVID-19 pandemic from the same centers. They concluded that compared to non-COVID-19-infected CVST patients, patients with the infection tended to be older, and had fewer CVST risk factors and worse outcomes [bib_ref] Cerebral venous sinus thrombosis associated with SARS-CoV-2; a multinational case series, Mowla [/bib_ref].
Several smaller studies reported CVST in seven adults (age range between 32 and 62 years, 62.5% female; [fig_ref] Table 1: Case reports of acute cerebrovascular events related to COVID-19 [/fig_ref] and one pediatric (a 13-year-old male) patient with COVID-19 [bib_ref] Cerebral venous thrombosis associated with COVID-19 infection: Causality or coincidence?, Poillon [/bib_ref] [bib_ref] Cerebral Venous Sinus Thrombosis as a Presentation of COVID-19, Hughes [/bib_ref] [bib_ref] Cerebral Venous Thrombosis Associated with COVID-19, Cavalcanti [/bib_ref] [bib_ref] Venous cerebral thrombosis in COVID-19 patient, Garaci [/bib_ref] [bib_ref] Cerebral Venous Sinus Thrombosis in a Pediatric Patient With COVID-19, Bastidas [/bib_ref]. Headaches were a presenting symptom in six (85.7%) cases, variably accompanied by different focal neurological deficits, confusion and impaired consciousness [bib_ref] Cerebral venous thrombosis associated with COVID-19 infection: Causality or coincidence?, Poillon [/bib_ref] [bib_ref] Cerebral Venous Sinus Thrombosis as a Presentation of COVID-19, Hughes [/bib_ref] [bib_ref] Cerebral Venous Thrombosis Associated with COVID-19, Cavalcanti [/bib_ref] [bib_ref] Venous cerebral thrombosis in COVID-19 patient, Garaci [/bib_ref]. Although in three patients the treatment and outcome were not reported [bib_ref] Cerebral venous thrombosis associated with COVID-19 infection: Causality or coincidence?, Poillon [/bib_ref] [bib_ref] Venous cerebral thrombosis in COVID-19 patient, Garaci [/bib_ref] , the condition was fatal in three out of five cases (60%) within a few days of onset despite anticoagulation and supportive therapy. Notably, in some cases, neurological symptoms occurred about two weeks after the onset of the general symptoms (i.e., fever, cough or dyspnea) of COVID-19 [bib_ref] Cerebral venous thrombosis associated with COVID-19 infection: Causality or coincidence?, Poillon [/bib_ref] [bib_ref] Venous cerebral thrombosis in COVID-19 patient, Garaci [/bib_ref]. Therefore, the possibility of this potentially life-threatening condition should not be overlooked even when patients present several days to weeks after the onset of COVID-19. A more recent multicenter 3-month cohort study of 13,500 consecutive patients with COVID-19 in New York City found an imaging-proved cerebral venous thrombosis incidence of 8.8 per 10,000 cases, which is higher than expected (i.e., 5 per million annually) [bib_ref] Cerebral Venous Thrombosis in COVID-19: A New York Metropolitan Cohort Study, Al-Mufti [/bib_ref]. In this study, despite the standard management [bib_ref] Mechanical Thrombectomy in Cerebral Venous Thrombosis, Siddiqui [/bib_ref] consisting of anticoagulation, endovascular thrombectomy and surgical hematoma evacuation, the mortality rate was 25% [bib_ref] Cerebral Venous Thrombosis in COVID-19: A New York Metropolitan Cohort Study, Al-Mufti [/bib_ref].
Overall, various degrees of elevated acute phase reactants (e.g., CRP and ferritin), hypercoagulability factors (e.g., D-dimer and aPTT) and abnormal platelet counts were found in these cases [bib_ref] Cerebral venous thrombosis associated with COVID-19 infection: Causality or coincidence?, Poillon [/bib_ref] [bib_ref] Cerebral Venous Sinus Thrombosis as a Presentation of COVID-19, Hughes [/bib_ref] [bib_ref] Cerebral Venous Thrombosis Associated with COVID-19, Cavalcanti [/bib_ref] [bib_ref] Venous cerebral thrombosis in COVID-19 patient, Garaci [/bib_ref] , suggesting a possible association with the hypercoagulability state in the setting of SARS-CoV-2 infection. It is unclear whether the monitoring of these markers has any value for the prediction of the onset or severity of cerebral venous thrombosis in these cases. This needs further detailed information on COVID-19 patients with such complications.
## Therapeutic approaches
Administering tissue plasminogen activator (tPA) in patients with COVID-19 and stroke is one of the therapeutic options. The role of other anticoagulants, such as low molecular weight heparin (LMWH) or full-dose heparin, is uncertain. There is some data to show that LMWH may be useful in sepsis-induced coagulopathy [bib_ref] Anticoagulant treatment is associated with decreased mortality in severe coronavirus disease 2019..., Tang [/bib_ref]. Although aspirin therapy in COVID-19 patients with ischemic stroke (especially in those who cannot take anticoagulants due to the risk of hemorrhagic transformation [bib_ref] Severe Acute Respiratory Syndrome Coronavirus 2 Infection and Ischemic Stroke, Valderrama [/bib_ref] or other medical limitations) can be considered as a secondary preventive approach, this medication is not indicated in patients with disseminated intravascular coagulation, a high risk of bleeding, or thrombocytopenia [bib_ref] Acute Pulmonary Embolism in Patients with COVID-19 at CT Angiography and Relationship..., Léonard-Lorant [/bib_ref] [bib_ref] Should aspirin be used for prophylaxis of COVID-19-induced coagulopathy?, Mohamed-Hussein [/bib_ref].
One reasonable treatment for COVID-19 patients is human recombinant soluble ACE-2 (hrsACE-2). There are two mechanisms of action for it: (1) preventing the SARS S protein from binding to lung and endothelial endogenous ACE-2, thereby reducing the infection of host cells; and (2) inhibiting the ACE-2 depletion by the SARS-CoV-2 virus. Considering ACE-2 is exhibited by brain endothelium and neurons, it is probable that the depletion of ACE-2 by the virus damages the endothelial function and leads to acute stroke. Along with the other known treatments, medications that affect the RAS system, such as angiotensin (1-7), may be appropriate therapies for COVID-19. Angiotensin's role is currently under evaluation in clinical trials (NCT04332666). In addition, AT1 receptor blockers (ARBs), such as losartan, could be preclusive in stroke [bib_ref] COVID-19-Related Stroke, Hess [/bib_ref]. On the other hand, another study has shown that early intravenous thrombolysis and immediate mechanical thrombectomy had poor outcomes in patients with acute ischemic stroke due to large vessel occlusion with COVID-19 [bib_ref] Treatment of Acute Ischemic Stroke due to Large Vessel Occlusion with COVID-19, Escalard [/bib_ref]. Overall, more well-designed, randomized, controlled trials are needed to provide an evidence-based approach for the prevention or treatment of acute cerebrovascular events in patients with COVID-19 [bib_ref] The COVID-19 pandemic: Is our medicine still evidence-based? Ir, Deana [/bib_ref].
# Conclusions
A growing body of evidence indicates that acute cerebrovascular events, including both ischemic and hemorrhagic strokes and cerebral venous thrombosis, may occur in patients with COVID-19. The underlying mechanisms of such events are still not completely understood. Still, they may include a hypercoagulability state, inflammation and cytokine storm, endothelial dysfunction, and an aberrant RAS axis due to the binding of SARA-CoV-2 to endothelial ACE-2. These abnormalities ultimately cause vasoconstriction, oxidative stress, inflammation and thrombogenesis. As these complications, especially cerebral venous thrombosis, are potentially life-threatening, physicians need to be vigilant when encountering patients with COVID-19 who have neurological symptoms such as headache, confusion, altered mental status, seizure and focal neurological deficits. Due to the small number of published cases or the mainly retrospective design of the previous clinical studies, (i) the functional outcome with the available therapies (e.g., LMWH) for thrombotic events and (ii) inflammatory or coagulable markers that can be efficiently used for the monitoring or prediction of such events in COVID-19 are still elusive, requiring a large cohort of patients with such complications.
[fig] Figure 1: Figure 1. Schematic representation of (A) the normal coagulation cascade triggered by an injury to the endothelial cells and (B) the hypercoagulability state related to SARS-CoV-2 infection. Generally, with an injury to the endothelial cell, the first response is platelet aggregation, followed by the activation of both the extrinsic and intrinsic pathways that lead to the generation of factor Xa and common pathway initiation. The product of the secondary coagulation cascade is a fibrin clot which will enhance platelet aggregates. ADPase and tissue factor (TF) pathway inhibitors (TFPI), as well as plasminogen, act as regulatory anticoagulants. The extrinsic and common pathways are initiated due to the increased TF and von Willebrand factor (VWF) in response to inflammatory cytokines such as interleukin (IL)-6 and the expression of TF by mononuclear cells in response to IL-6, which is elevated in SARS-CoV-2 infection. TFPI and ADPase are inhibited by the cytokines as well. [/fig]
[fig] Figure 2: Figure 2. Schematic effects of SARS-CoV-2 on the renin-angiotensin system (RAS) in the cerebral vasculature. After the binding of the SARS-CoV-2 spike (s) glycoprotein to angiotensin converting enzyme 2 (ACE-2), the virus enters the cell. The virus also downregulates ACE-2 and competes with angiotensin II (Ang II) for binding to ACE-2, which ultimately decreases the activity of the ACE-2-Ang-(1-7)-Mas receptor (alternative) axis. This also leads to the greater activation of the ACE-Ang II-AT1R (classical) axis. The outcome of such events is an aberrant renin-angiotensin system (RAS), causing vasoconstriction, inflammation, oxidative stress, and thrombogenesis, causing ischemic stroke in relation to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection. SARS-COV-2. TMPRSS2, Transmembrane protease, serine 2; AT1R: angiotensin 1 receptor; AT2R: angiotensin 2 receptor. [/fig]
[fig] ↓: Hgb (119 g/L); ↑ LDH (654 U/L), ALT (137 U/L), PT (12.5 s), fibrinogen (950 mg/dL), D-dimer (>80,000 µg/L), ferritin (4927 µg/L) & CRP (305.4 mg/L); lupus anticoagulant (+); normal CBC diff , aPTT & INR. [/fig]
[table] Table 1: Case reports of acute cerebrovascular events related to COVID-19. [/table]
|
Mycobacterium tuberculosis infection, immune activation, and risk of HIV acquisition
BackgroundAlthough immune activation is associated with HIV acquisition, the nature of inflammatory profiles that increase HIV risk, which may include responses to M. tuberculosis (Mtb) infection, are not well characterized.MethodsWe conducted a nested case-control study using cryopreserved samples from persons who did and did not acquire HIV during the multinational Step clinical trial of the MRKAd5 HIV-1 vaccine. PBMCs from the last HIV-negative sample from incident HIV cases and controls were stimulated with Mtb-specific antigens (ESAT-6/CFP-10) and analyzed by flow cytometry with intracellular cytokine staining and scored with COMPASS. We measured inflammatory profiles with five Correlates of TB Risk (CoR) transcriptomic signatures. Our primary analysis examined the association of latent Mtb infection (LTBI; IFNγ+CD4+ T cell frequency) or RISK6 CoR signature with HIV acquisition. Conditional logistic regression analyses, adjusted for known predictors of HIV acquisition, were employed to assess whether TB-associated immune markers were associated with HIV acquisition.ResultsAmong 465 participants, LTBI prevalence (21.5% controls vs 19.1% cases, p = 0.51) and the RISK6 signature were not higher in those who acquired HIV. In exploratory analyses, Mtb antigen-specific polyfunctional CD4+ T cell COMPASS scores (aOR 0.96, 95% CI 0.77, 1.20) were not higher in those who acquired HIV. Two CoR signatures, Sweeney3 (aOR PLOS ONE PLOS ONE | https://doi.
a1111111111 a1111111111 a1111111111 a1111111111 a1111111111
1.38 (1.07, 1.78) per SD change) and RESPONSE5 (0.78 (0.61, 0.98)), were associated with HIV acquisition. The transcriptomic pattern used to differentiate active vs latent TB (Sweeney3) was most strongly associated with acquiring HIV.
# Conclusions
LTBI, Mtb polyfunctional antigen-specific CD4+ T cell activation, and RISK6 were not identified as risks for HIV acquisition. In exploratory transcriptomic analyses, two CoR signatures were associated with HIV risk after adjustment for known behavioral and clinical risk factors. We identified host gene expression signatures associated with HIV acquisition, but the observed effects are likely not mediated through Mtb infection.
# Background
Several lines of evidence suggest that systemic inflammation increases risk of HIV acquisition. For example, in the HIV Vaccine Trials Network Step Study (V520-023/HVTN 502), an efficacy trial of an Adenovirus 5 (Ad5)-vectored HIV-1 vaccine, some vaccinees experienced a transient increased risk of HIV acquisition for several months after vaccination [bib_ref] Efficacy assessment of a cell-mediated immunity HIV-1 vaccine (the Step Study): a..., Buchbinder [/bib_ref]. In Step, ELISpot mock responses (interferon-gamma [(IFN-] secretion in the absence of antigen), but not HIV-antigen stimulated responses, directly correlated with a 61% increase in HIV acquisition risk [bib_ref] Immune-correlates analysis of an HIV-1 vaccine efficacy trial reveals an association of..., Huang [/bib_ref]. Furthermore, CAPRISA-004 and Partners in Prevention studies each identified different cytokine profiles associated with HIV acquisition . Other studies suggest associations between herpes simplex virus-2 and chronic filarial infections with HIV acquisition [bib_ref] Effect of Wuchereria bancrofti infection on HIV incidence in southwest Tanzania: a..., Kroidl [/bib_ref] [bib_ref] Herpes simplex virus 2 infection increases HIV acquisition in men and women:..., Freeman [/bib_ref]. Although there is some understanding of the local genital inflammation and microbiome changes associated with HIV susceptibility [bib_ref] Interactions of HIV, other sexually transmitted diseases, and genital tract inflammation facilitating..., Mayer [/bib_ref] [bib_ref] The genital tract and rectal microbiomes: their role in HIV susceptibility and..., Karim [/bib_ref] , the biologic underpinnings of peripheral blood immune profiles potentially associated with HIV risk are not well understood. It may be that other infections or exposures induce immune phenotypes that increase susceptibility of CD4 T cells and macrophages to HIV infection.
Mycobacterium tuberculosis (Mtb) latently infects approximately 24% of the world's population, with an estimated 10 million new cases of clinical disease (TB) in 2019 alone [bib_ref] The global burden of latent tuberculosis infection: a re-estimation using mathematical modelling, Houben [/bib_ref] , Latent Mtb infection (LTBI), is defined by a positive response to Mtb antigens (Purified Protein Derivative (PPD) skin test or an IFN-release assay (IGRA)) in the absence of clinical symptoms. Recent data indicates that many asymptomatic individuals have "incipient TB" with evidence of systemic inflammation [bib_ref] Sequential inflammatory processes define human progression from M. tuberculosis infection to tuberculosis..., Scriba [/bib_ref] , which is associated with an increased risk of progression to clinical TB disease [bib_ref] Considerations for biomarker-targeted intervention strategies for TB prevention, Fiore-Gartland [/bib_ref] [bib_ref] A blood RNA signature for tuberculosis disease risk: a prospective cohort study, Zak [/bib_ref]. While the effects of HIV on risk of LTBI progression have been intensively investigated, the effect of Mtb infection on risk of acquiring HIV has never been formally assessed, other than in ex-vivo experiments [bib_ref] Tuberculosis generates a microenvironment enhancing the productive infection of local lymphocytes by..., Garrait [/bib_ref] [bib_ref] Mycobacterium tuberculosis complex enhances susceptibility of CD4 T cells to HIV through..., Thayil [/bib_ref] [bib_ref] TLR2-Modulating Lipoproteins of the Mycobacterium tuberculosis Complex Enhance the HIV Infectivity of..., Skerry [/bib_ref]. In a recent study, infant macaques were vaccinated with TB vaccines and then challenged with simian immunodeficiency virus (SIV) orally. Animals vaccinated with either an Mtb-SIV auxotroph vaccine or BCG vaccine demonstrated significantly increased SIV risk compared to placebo recipients. Both vaccines resulted in immune activation, which correlated with SIV acquisition .
Because the global burden of Mtb infection is high and the lifetime risk of progression to active disease is only~10% in HIV-negative persons, no countries with co-endemic TB and HIV routinely treat LTBI in HIV-negative persons . If treating LTBI were to reduce the risk of acquiring HIV, then targeting LTBI treatment to the subset of individuals most at-risk for HIV could prevent development of TB disease as well as decrease risk of HIV. We therefore designed a study of LTBI-and Mtb-associated immune activation using prospectively collected samples from an HIV vaccine trial to address this question. We included several Correlates of Risk (CoR) transcriptomic signatures that have been validated to predict TB-disease states, including incipient active TB and TB treatment success, based on differential human gene activation.
# Methods
## Study population
We conducted a case control ancillary study nested within the completed Step study (V520-023/HVTN502). The Step study enrolled 3,000 high-risk HIV-negative males and females aged 18-45 who received the MRKAd5 HIV-1 gag/pol/nef vaccine at 34 study sites in the Americas, Caribbean, and Australia from 2004-2007. Participants received vaccine/placebo injection at 0, 4, and 26 weeks. HIV testing was done at 6-monthly visits for up to 4 years of in the Step study and the follow-on HVTN 504; stored blood from the prior visit was tested for HIV RNA to more precisely time HIV acquisition [bib_ref] Efficacy assessment of a cell-mediated immunity HIV-1 vaccine (the Step Study): a..., Buchbinder [/bib_ref]. A total of 172 incident infections were diagnosed in men (3.3/100 person-years) and 15 infections in women (0.45/100 PY). We identified as cases all participants within Step and HVTN 504 with incident HIV who had available peripheral blood mononuclear cells (PBMC) at the last study visit prior to HIV acquisition. Cases who acquired HIV before week 8 (first available PBMC) were excluded. Controls were selected by HVTN staff as persons who did not acquire HIV during study follow-up, matched to cases 2:1 on study site and treatment arm, and with a pooled gender distribution similar to cases. We acquired PBMC from the study visit corresponding to that of their matched case; if no sample from a matched visit was available a new control was selected. An unblinded team member not involved in the laboratory work constructed a batching scheme that allowed for a balanced selection of cases and controls throughout the laboratory work, while allowing our laboratory team to remain blinded to participant characteristics. The following baseline variables from the Step study were assessed: age, gender, self-reported race/ethnic identity, and study site; study treatment (vaccine/placebo); baseline Ad5 antibody titer, self-reported behavioral HIV risks: drug use, number and type of sexual partners, condomless sex, substance use; circumcision status and HSV-2 serostatus for males.
Step participants were generally not eligible for TB prophylaxis and no prophylactic or therapeutic TB drugs were documented as given to any participant during study follow up. Because this ancillary study used only deidentified data and samples from the Step trial, the work described in this manuscript was determined not to be engaged in human subjects research by the University of Washington Institutional Review Board. All participants who participated in the Step study and whose samples and data were used here were verified to have given specific informed consent for future use of stored samples at time of participation in the vaccine trial.
## Cell activation and flow cytometry
Cryopreserved PBMCs were thawed, washed and rested in RPMI 1640 media containing 10% heat-inactivated fetal bovine serum (FBS) overnight prior to antigen stimulation with a concentration of 2 x 10 6 cells /mL. PBMCs were counted with Guava easyCyte (Millipore) using Guava ViaCount reagent (Luminex) and GuavaSoft v.2.6 software. Samples with less than 66% viability were discarded. Cells were stimulated with a pool of early secretory Mtb antigen target-6 (ESAT-6) and culture filtrate protein (CFP-10) (BEI Resources). PMA (25ng/mL)/ionomycin (1ug/mL) was used as a positive control and DMSO (0.5%) was used as a negative control. In addition, costimulatory antibody anti-CD28/49d, cytokine secretion inhibitor Brefeldin A and Monensin were added to each stimulation cocktail. Cells were lysed and permeabilized with FACS Lyse and FACS Perm-II buffer and stained on a BD LSRFortessa with an antibody panel developed for analyzing CD4 and CD8 T-cell responses (ICS) including IFN-, TNF, IL-2, and IL-17a (Table A in S1 Appendix) . Flow cytometry data were analyzed in FlowJo™ v10.7. Each sample was compensated and gated manually.
## Rna isolation and rt-pcr
Approximately 2 x 10 6 PBMCs from each sample were re-suspended in RNAlater (Invitrogen) preservative and stored at -20˚C for processing . Total RNA was isolated using RNeasy spin columns (Qiagen) and cDNA templates for qRT-PCR were generated using Applied Biosystems high-capacity cDNA reverse transcriptase kits. qRT-PCR was performed with 44 primer-probe sets in S1 Appendix) from Taqman using the Fluidigm 96.96 dynamic array platform. cDNA was preamplified using a pool of specific TaqMan primer-probe sets for 16 cycles with a 15 second denaturation step at 95˚C and 4 minutes at 60˚C. The pre-amplified reaction was added to the Fluidigm 96.96 dynamic array platform with TaqMan gene expression assays and amplified for an additional 40 cycles. A positive control sample was run on every chip to monitor consistency. Primer-probe sets were included to encompass the components of 5 previously described CoR signatures from peripheral blood associated with TB disease outcomes. Signature scores were calculated from raw Ct values in R v3.6.1 as previously described [25-30] in S1 Appendix; additional primer probes not pertaining to selected CoRs were included for future analysis).
# Statistical analysis
The outcome of interest in all analyses was HIV acquisition. To evaluate the association of Mtb antigen-specific T-cell activation and HIV acquisition, we used three measures:
1. Individuals with LTBI were defined using a flow cytometry-based method that is strongly correlated with the tuberculin skin test and IGRA. Samples were classified as positive if the frequency of IFN-+ CD4+ cells stimulated with ESAT-6/CFP-10 pooled peptides doubled compared to negative control . LTBI was prespecified as the primary Mtb exposure variable.
2. We used Combinatorial polyfunctionality analysis of antigen-specific T-cell subsets (COM-PASS) to determine overall Mtb-antigen-specific T cell activation. COMPASS is an analytic tool that uses a Bayesian hierarchical framework to model all observed cell subsets. COM-PASS outputs functional scores (FS) and polyfunctional scores (PFS) that define the posterior probabilities of antigen-specific response across cell subsets, described by a single numerical score that ranges from 0 to 1 [36]. We have previously used COMPASS to investigate polyfunctionality in CD4+ T cells in an epidemiologic study of South African adolescents screened for LTBI, as well as to predict HIV-specific responses in a prior HIV vaccine trial . FS and PFS derived from COMPASS output were included as untransformed variables of secondary interest.
3. To determine whether transcriptomic evidence of Mtb-associated immune activation also predicts risk of HIV acquisition, we used Correlates of Risk (CoR) scores previously validated to detect different TB-associated states. The primary signature of interest was RISK6, a 6-gene transcriptomic CoR signature that predicts progression to TB disease (incipient TB) . Additional CoR scores assessed in exploratory analysis include Suliman4 and Maertzdorf4, both of which also predict incipient TB; Sweeny3, which accurately differentiates between current LTBI vs active TB; and RESPONSE5, which predicts cure at time of TB treatment initiation among those with active TB in S1 Appendix) . CoR scores that were not normally distributed were log 10 transformed (RISK6, and Suliman4). To account for score values of zero (n = 3; all RISK6), zeros were replaced by dividing the next lowest value by two.
For all measures, the association between exposures and HIV acquisition was estimated using univariate and multivariate conditional logistic regression accounting for the variables used to match controls to cases (country of residence and treatment arm within the Step study). Variables assessed for inclusion in multivariate regression included known predictors of HIV risk in the Step study (i.e. gender, validated behavioral risk score (scale 0-7) [39], baseline Ad5 titer, and self-reported race/ethnicity). Each variable of interest was assessed for association with HIV acquisition using conditional logistic regression, and associations that were significant at p<0.1 were included in the multivariate model. For sub-analyses restricted to males, we included HSV-2 serostatus and circumcision, as these were documented in the parent trial for males only. Analyses of the association of PFS, FS, and CoR scores with HIV acquisition were repeated after stratification by LTBI status. Stratified analyses were performed as standard logistic regression with matching variables included, as logistic regression conditioned on LTBI would have excluded all case/control sets with discordant LTBI status. PFS, FS, and CoR scores were reported as the Odds Ratio (OR) for 1 standard deviation change to allow for comparability across outcomes. As a sensitivity analysis, we repeated the primary analyses restricted to case/control sets with samples within 6 and 12 months prior to HIV diagnosis. All analyses were conducted in R v3.6.1 and Stata v15.1 (College Station, TX, USA, 2017).
# Results
We selected 155 cases who acquired HIV and 310 controls (n = 465), with baseline characteristics presented in [fig_ref] Table 1: Participant characteristics [/fig_ref]. Among cases, the median interval between sample collection date and HIV diagnosis was 287 days (IQR 161, 483). The majority of participants (88%) were from Peru or the US. By design, there was no difference in distribution of vaccine treatment arm or country in cases vs. controls, although there was a trend toward more males among cases. As expected, the behavioral risk score previously found to be associated with HIV acquisition was higher in cases than controls (3.3 vs 2.8, SD 1.2, p<0.0001), and HSV-2 seropositivity was more common in male cases than controls (40.1% vs 30.7%, p = 0.049).
To assess one of our primary hypotheses, whether Mtb infection was associated with HIV acquisition, we examined 454 participants, including 152 cases and 302 controls (n = 11 missing samples or gating failures) for LTBI status via flow cytometry. Overall, 94 (20.7%) study participants were considered LTBI positive via detection of ESAT-6-CFP-10-specific CD4+ T cells: 65 (21.5%) controls compared to 29 (19.1%) cases. LTBI positivity was not associated with HIV acquisition in unadjusted or adjusted conditional logistic regression analyses (aOR 0.85 (0.51, 1.43) per 1 SD change, p = 0.73) . When stratified by gender, the inference did not change. Among men, when adjusting for all covariates in the primary models as well as HSV-2, LTBI status was not associated with HIV acquisition (aOR: 0.99; 95% CI 0.58, 1.69). Likewise, there was no association between LTBI status and HSV-2 serostatus (31.8% seropositive in LTBI negative vs 30.0% in LTBI positive, p = 0.77).
We also performed two exploratory analyses with ESAT-6/CFP-10-specific functional and polyfunctional cytokine expression, as measured through the COMPASS FS and PFS, respectively, which were both significantly higher in LTBI-positive versus LTBI-negative samples (p<0.0001; [fig_ref] Fig 1: Combinatorial polyfunctionality analysis of antigen-specific T-cell subsets [/fig_ref]. Among all 152 cases the median FS was 0.026 (IQR 0.009, 0.077) and of 302 controls the median FS was 0.029 (0.009, 0.077). In controls, the median PFS was 0.015 vs 0.015 among cases and controls. [fig_ref] Fig 1: Combinatorial polyfunctionality analysis of antigen-specific T-cell subsets [/fig_ref] shows box plots and a heatmap of COMPASS posterior probabilities by case/control status. The distributions of mono-and poly-functional Mtb-specific subsets across persons who did or did not acquire HIV were similar. When assessing the association using conditional logistic regression, neither FS (OR 0.94 per 1 SD change, 95% CI 0.76, 1.16) nor PFS (OR 0.96, 95% CI 0.78, 1.18) were associated with HIV
## Plos one
acquisition. we found no evidence of an association of FS or PFS with HIV acquisition when evaluating only LTBI positive participants, (Table C in S1 Appendix). RISK6 score was calculated for 439 participants (n = 141 cases, n = 298 controls) in whom RNA extraction was successful. The mean of untransformed RISK6 scores among cases was 0.119 (SD 0.147) compared to 0.107 (SD 0.141) among controls . In both unadjusted and adjusted analyses, RISK6 was not associated with HIV acquisition (aOR 1.15 per 1 SD change, 95% CI 0.92, 1.44) [fig_ref] Table 3: Associations between each Correlates of Risk [/fig_ref] acquired HIV and controls for RISK6, as well as the other four CoR signatures (Suliman4, Sweeney3, Maertzdorf4, and RESPONSE5) examined in exploratory analyses. In multivariate conditional logistic regression, two of the four CoR measures demonstrated associations with HIV acquisition. A one SD increase of the Sweeney3 score was associated with a 38% increased odds of HIV acquisition (aOR 1.38, 95% CI 1.07, 1.78). In contrast, the RESPONSE5 CoR score was associated with a moderate decrease in odds of HIV acquisition (aOR 0.78, 95% CI
## Fig 2. distribution of 5 tb correlates of risk (cor) scores among participants who subsequently acquired hiv vs controls. box plots demonstrate median and interquartile range of values, with dots completing the range. risk6
and Suliman4 were log 10 transformed as this allowed for the data to be normally distributed and for all scores to be evaluable on a similar scale.
https://doi.org/10.1371/journal.pone.0267729.g002 0.61, 0.98). There were no important trends in CoR scores between those with lower baseline Ad5 titer, gender, or behavioral risk score. HSV-2 serostatus (in males only) was associated with score of one signature and vaccine vs placebo treatment was associated with a different signature. The mean of three CoR scores differed significantly by country of residence. Bivariate associations between predictors of HIV acquisition and each CoR are shown in in S1 Appendix. In adjusted analyses, Ad5 titer was not influential in evaluating associations between any CoR score and HIV acquisition. Due to the long interval between PBMC collection and HIV diagnosis in some cases, we performed sensitivity analyses including the 282 participants (n = 94 cases, n = 188 controls) with samples within 12 months, and then the 159 participants (n = 53 cases, n = 106 controls) with 6 months or less between sampling and HIV diagnosis. The aOR for RISK6 increased to 1.28 (0.94, 1.75) but neither this or any of the other primary exposures (LTBI, FS, or PFS), became statistically significant for participants sampled within a year or 6 months of diagnosis. As in the full dataset, RESPONSE5 remained associated with decreased risk of HIV and Swee-ney3 with the association strengthened by >10% [fig_ref] Fig 1: Combinatorial polyfunctionality analysis of antigen-specific T-cell subsets [/fig_ref]. Due to loss of power, the point estimates for several exploratory analyses changed by >10% for the participants sampled within 6 months, but no findings were significant (Table E in S1 Appendix).
In adjusted analyses stratified by LTBI status, a one SD increase in logRISK6 score was associated with a trend toward decreased odds of HIV acquisition among LTBI-positive participants (aOR 0.58, 95% CI 0.34, 0.99) and a trend toward increase in odds of HIV acquisition among LTBI negative participants (aOR 1.24, 95% CI 0.97, 1.59) (Table F in S1 Appendix). Stratified by LTBI status, some associations between CoR scores and HIV remained in the LTBI negative participants, but not in the smaller subset of LTBI positive participants. In univariate analyses however, none of the CoR scores were associated with LTBI status (Table G in S1 Appendix).
# Discussion
In this study, we accessed rare, cryopreserved specimens from the completed Step HIV vaccine trial to evaluate the influence of sub-clinical Mtb infection on HIV acquisition risk, employing a case-control design. Despite the fact that many of the biomarkers that have previously been associated with increased risk of acquiring HIV are also those elevated in subclinical/incipient Mtb infection or TB disease, we found little evidence that prior infection with Mycobacterium tuberculosis increases risk of acquiring HIV. To our knowledge, there are no cohorts in which LTBI status has been determined using clinical methods (TST or IGRA) among comparable persons with subsequent HIV endpoints. Our use of comprehensive flow cytometry and COMPASS analysis allowed us to retrospectively determine LTBI status and to characterize Mtb antigen-specific polyfunctional T-cell responses within a homogenous trial population with subsequent evaluation of HIV status. Neither LTBI status, nor polyfunctional cellular activation scoring was associated with risk of HIV acquisition, either bivariately, or after controlling for the previously described HIV risk factors. Transcriptional TB CoR signatures do not measure Mtb-specific responses, but rather the differential expression of genes associated with incipient or active TB disease; we hypothesized that these same signatures, many of which include interferon stimulated genes (ISGs), could also reveal HIV susceptibility states. However, the primary CoR signature of interest, the RISK6 signature, comprising several ISGs, was not associated with HIV acquisition in the overall study. Although aORs differed by LTBI status, as RISK6 was associated with lower HIV risk in those with LTBI and showed a trend towards increased HIV risk in those without LTBI, we could not rule out that this difference was due to chance. It is possible that pathways other than the specific ISG signaling characterized by RISK6 are associated with HIV risk and are also differentially regulated by co-infections. In addition, evaluation of other CoR profiles suggested associations with HIV acquisition, suggesting that there may be an unmeasured source of immune activation predictive of HIV acquisition. The different gene sets in each signature describe different activation patterns, and possibly various underlying exposures. Because LTBI status was not independently associated with any of the 5 CoRs in this cohort, especially RISK6, Suliman4, or Maetz-dorf4, it is likely that there was little incipient TB in the Step study. It is therefore possible that these CoRs were detecting ISGs provoked by stimuli other than Mtb infection. The Mtb-specific cellular responses that define LTBI are well-known to persist despite LTBI treatment or spontaneous clearance of Mtb infection , and therefore persons identified as LTBI in this study may not have had differential ISG or other gene expression measured concurrently in that sample or at the time of HIV exposure.
The results of the exploratory analyses provide some evidence that a transcriptomic biomarker could evaluate HIV risk, regardless of Mtb infection. The Sweeney3 signature was most strongly associated with HIV risk in unadjusted and adjusted analyses, and in the LTBInegative, but not the LTBI-positive group. This three-gene set predicts the presence of active TB disease versus LTBI with better accuracy (86% sensitivity, 86% specificity) than most combinations of clinical prediction tools, and with similar sensitivity to GeneXpert-MTB RIF sputum tests [bib_ref] Xpert MTB/RIF and Xpert MTB/RIF Ultra for pulmonary tuberculosis and rifampicin resistance..., Horne [/bib_ref] [bib_ref] Blood-based host biomarker diagnostics in active case finding for pulmonary tuberculosis: A..., Moreira [/bib_ref]. Because the three genes, GBP5, DUSP3, and KLF2, are associated with macrophage regulation and other immune pathways [bib_ref] Inducible GBP5 Mediates the Antiviral Response via Interferon-Related Pathways during Influenza A..., Feng [/bib_ref] [bib_ref] DUSP3 Genetic Deletion Confers M2-like Macrophage-Dependent Tolerance to Septic Shock, Singh [/bib_ref] [bib_ref] An update on the regulatory mechanisms of NLRP3 inflammasome activation, Paik [/bib_ref] , this gene set may detect a highrisk systemic immune status, provoked by genital or gastrointestinal infection or altered microbiome composition, which could also be differentially associated with LTBI status. Alternatively, the transcription of these genes could relate to other stable host factors that impact both Mtb and HIV susceptibility.
In contrast, the RESPONSE5 profile evaluates differential gene expression that predicts cure after TB treatment among those with active TB, wherein a high score represents higher innate immune activation , and in this study, higher scores were associated with lower odds of HIV acquisition. These findings suggest that Mtb infection and polyfunctional Mtb antigen-specific T-cell activation are not associated directly with HIV risk, but that other unmeasured exposures or intrinsic factors may ultimately activate these same host transcriptional pathways and explain the observed association between some transcriptional signatures and HIV acquisition. For example, some risk scores were modestly associated with receipt of the Ad5 HIV DNA vaccine in the parent trial, with HSV-2 serostatus, or with country of residence.
Our study has several limitations, the foremost of which was lack of availability of stored PBMC close to the time of HIV acquisition in some cases, which we tried to address through sensitivity analyses restricted to cases with proximal samples. Some samples were collected more than one year prior to documented HIV diagnosis. Due to wide intervals between study visits (up to 6 months between visits), many persons likely acquired HIV much closer to the sampling timepoint than indicated by their diagnosis visit, but the timing of HIV acquisition is not clearly pinpointed. That said, CoR signatures are able to predict progression to incident TB up to 24 months prior to clinical disease and show accurate prediction within 1 year [bib_ref] A blood RNA signature for tuberculosis disease risk: a prospective cohort study, Zak [/bib_ref]. As the parent Step study occurred within some countries with moderate or high TB prevalence, there is a possible risk that interval Mtb infection or sub-clinical reactivation occurred between sampling and HIV outcome. Additionally, LTBI status was not determined in the
Step study; thus, we were unable to use clinically accepted measures of LTBI status, such as IGRA or TST. Having access to a single blood sample limited our ability to assess interval LTBI status conversion or reversion prior to HIV acquisition, and there is a possibility of other unevaluated infections or other confounders. Nonetheless, the Step study represents a rare opportunity to address this issue as there are few, if any, cohorts in which Mtb infection status, future HIV outcome, and longitudinal cryopreserved specimens are available. Instead, we used a rigorous Mtb-antigen-specific flow cytometry assay that provided richer descriptive data, not only about LTBI status, but about the degree and quality of Mtb responses. Other strengths of this study include rigorous characterization of study participants as part of the parent Step study, and novel application of TB CoR signatures.
Based on this translational study we would not expect that treating LTBI in persons at risk for HIV would decrease risk of HIV acquisition substantially, especially in the context of increasingly effective and available HIV pre-exposure prophylaxis (PrEP) options. Treating LTBI has other intrinsic benefits, including prevention of TB disease and Mtb transmission, especially in those who are living with HIV or otherwise at high risk of developing active TB disease.
To our knowledge there is little published data on differential gene regulation being associated with HIV susceptibility or protection. While we found variable associations with some purpose-designed TB CoR signatures, global transcriptomic analysis could assist in discovery of similar, simpler HIV CoR signatures. Although such HIV CoR signatures would be unlikely to replace patient-reported indications for PrEP and other prevention measures, such signatures could be beneficial in risk-stratifying participants within HIV prevention and vaccine studies. Transcriptomic signatures could also be employed to detect high-risk post-vaccine immune responses in early-phase trial volunteers. Designing purpose-built signatures could help identify vaccine candidates that might induce risky immune states for HIV acquisition, such as occurred in the Step study, before moving the vaccine candidate to larger trials in HIV-exposed populations.
# Conclusions
In this study, we found no evidence that LTBI or polyfunctional Mtb-antigen-specific T cell responses were associated with risk of HIV acquisition in a population at high risk for HIV. However, the exploratory analyses provided a proof of concept that transcriptomic signatures could provide additional information about immunologic profiles of HIV susceptibility beyond known behavioral and viral co-infection risk factors identified in the Step study and other prior studies.
Supporting information S1 Appendix. Supplemental Tables A and B provide materials used in flow cytometry and transcriptional signature analysis, respectively. Tables C, E, and F provide sensitivity analyses stratified or restricted by LTBI status (C, F) and time restricted analyses (E).
[fig] Fig 1: Combinatorial polyfunctionality analysis of antigen-specific T-cell subsets (COMPASS) describesMtb-antigenspecific T cell activation. Box and whisker plots depict median and interquartile ranges for FS and PFS scores as well as the range. A. COMPASS Polyfunctional Scores (PFS) demonstrate expected higher levels of Mtb-specific cellular immune activation in persons with LTBI than those without (p<0.0001). B. COMPASS Mtb-specific Functional Scores (FS) were no different between persons who acquired HIV and controls. PFS was similarly indistinguishable between cases and controls. C. Heatmap of COMPASS posterior probabilities depicts combinations of intracellular cytokine staining composing the combinatorial polyfunctional subsets of Mtb-specific CD4+ T cells. Columns show the subsets with detectable antigenspecific responses color-coded by the cytokines they express and ordered by the degree of functionality from one function on the left to 3 of 4 functions on the right; the combinations of cytokines are found in colored boxes at the bottom. Horizontal rows depict one participant, with lime shaded rows representing control samples stacked above violet rows representing persons who acquired HIV. The intensity of purple shading of each box shows the probability that a given participant sample that expresses the subset has a Mtb-antigen specific response ranging from white (zero) to purple (one). This heatmap depicts relatively similar distributions of mono-and polyfunctional Mtb-specific subsets across persons who did or did not acquire HIV.https://doi.org/10.1371/journal.pone.0267729.g001 [/fig]
[fig] 30: Maertzdorf J, Weiner J, Mollenkopf H-J, Network T, Bauer T, Prasse A, et al. Common patterns and disease-related signatures in tuberculosis and sarcoidosis. Proceedings of the National Academy of Sciences. 2012; 109(20):7853-8. https://doi.org/10.1073/pnas.1121072109 PMID: 22547807 31. Cosmi L, Maggi L, Santarlasci V, Liotta F, Frosali F, Angeli R, et al. Detection by flow cytometry of ESAT-6-and PPD-specific circulating CD4+ T lymphocytes as a diagnostic tool for tuberculosis. International archives of allergy and immunology. 2007; 143(1):1-9. https://doi.org/10.1159/000098220 PMID: 17191004 32. Tesfa L, Koch FW, Pankow W, Volk HD, Kern F. Confirmation of Mycobacterium tuberculosis infection by flow cytometry after ex vivo incubation of peripheral blood T cells with an ESAT-6-derived peptide pool. Cytometry Part B, Clinical cytometry. 2004; 60(1):47-53. https://doi.org/10.1002/cyto.b.20007 PMID: 15221869 33. Papageorgiou CV, Anastasopoulos A, Ploussi M, Leventopoulos M, Karabela S, Fotiadis K, et al. Flow cytometry analysis of CD4+ IFN-γ+ T-cells for the diagnosis of mycobacterium tuberculosis infection. Cytometry Part B: Clinical Cytometry. 2016; 90(3):303-11. 34. Lee J, Lee SY, Won DI, Cha SI, Park JY, Kim CH. Comparison of whole-blood interferon-γ assay and flow cytometry for the detection of tuberculosis infection. Journal of Infection. 2013; 66(4):338-45. https://doi.org/10.1016/j.jinf.2012.08.020 PMID: 23010554 35. Sauzullo I, Scrivo R, Mengoni F, Ermocida A, Coppola M, Valesini G, et al. Multi-functional flow cytometry analysis of CD 4+ T cells as an immune biomarker for latent tuberculosis status in patients treated with tumour necrosis factor (TNF) antagonists. Clinical & Experimental Immunology. 2014; 176(3):410-7. 36. Lin L, Finak G, Ushey K, Seshadri C, Hawn TR, Frahm N, et al. COMPASS identifies T-cell subsets correlated with clinical outcomes. Nat Biotech. 2015; 33(6):610-6. https://doi.org/10.1038/nbt.3187 PMID: 26006008 37. Biasin M, Piacentini L, Lo Caputo S, Naddeo V, Pierotti P, Borelli M, et al. TLR activation pathways in HIV-1-exposed seronegative individuals. Journal of immunology (Baltimore, Md: 1950). 2010; 184 (5):2710-7. https://doi.org/10.4049/jimmunol.0902463 PMID: 20124101 38. Shah JA, Berrington WR, Vary JC, Jr., Wells RD, Peterson GJ, Kunwar CB, et al. Genetic Variation in Toll-Interacting Protein Is Associated With Leprosy Susceptibility and Cutaneous Expression of Interleukin 1 Receptor Antagonist. The Journal of infectious diseases. 2016; 213(7):1189-97. https://doi. org/10.1093/infdis/jiv570 PMID: 26610735 39. Huang Y, Follmann D, Nason M, Zhang L, Huang Y, Mehrotra DV, et al. Effect of rAd5-Vector HIV-1 Preventive Vaccines on HIV-1 Acquisition: A Participant-Level Meta-Analysis of Randomized Trials. PloS one. 2015; 10(9):e0136626. https://doi.org/10.1371/journal.pone.0136626 PMID: 26332672 40. Kim OH, Jo KW, Park S, Jo YH, Kim MN, Sung H, et al. Comparison of the change in QuantiFERON-TB Gold Plus and QuantiFERON-TB Gold In-Tube results after preventive therapy for latent tuberculosis infection. PloS one. 2020; 15(6):e0234700. https://doi.org/10.1371/journal.pone.0234700 PMID: 32544206 41. Xin H, Cao X, Zhang H, Liu J, Pan S, Li X, et al. Dynamic changes of interferon gamma release assay results with latent tuberculosis infection treatment. Clin Microbiol Infect. 2020; 26(11):1555.e1-.e7. https://doi.org/10.1016/j.cmi.2020.02.009 PMID: 32062048 [/fig]
[table] Table 1: Participant characteristics.� Race and ethnicity were asked in a singular question about self-identity in the parent study without option of multiselect; options including race or ethnicity therefore sum to 100%. �� Combined risk score for male and female includes theoretical and observed range 0-7 points.https://doi.org/10.1371/journal.pone.0267729.t001Table 2. LTBI status, Mtb-specific CD4 T-cell activation, and HIV acquisition. [/table]
[table] Table 3: Associations between each Correlates of Risk (COR) signature and HIV acquisition, univariate and adjusted analyses.Models were fully fit with all variables that were associated with case control status: gender, risk score, and age. [/table]
|
FABP5 deletion in nociceptors augments endocannabinoid signaling and suppresses TRPV1 sensitization and inflammatory pain
The endocannabinoid anandamide (AEA) produces antinociceptive effects by activating cannabinoid receptor 1 (CB1). However, AEA also serves as an agonist at transient receptor potential vanilloid receptor 1 (TRPV1) in nociceptive sensory neurons, which may exacerbate pain. This potential functional duality is highlighted by the failure of an inhibitor of the AEA catabolic enzyme fatty acid amide hydrolase (FAAH) to afford pain relief in a clinical trial. Consequently, it remains to be determined whether elevating AEA levels in nociceptors leads to antinociceptive or pro-nociceptive effects. Fatty acid binding protein 5 (FABP5) is an intracellular carrier that mediates AEA transport to FAAH for inactivation. Leveraging the abundant expression of FABP5 in TRPV1 + nociceptors, we employed a conditional knockout strategy to demonstrate that FABP5 deletion in nociceptors augments AEA levels, resulting in the emergence of antinociceptive effects mediated by CB1. Mechanistically, FABP5 deletion suppresses inflammation-and nerve growth factor-mediated TRPV1 sensitization via CB1, an effect mediated by calcineurin. Unexpectedly, inhibition of FAAH failed to blunt TRPV1 sensitization, uncovering functionally distinct outputs resulting from FABP5 and FAAH inhibition. Collectively, our results demonstrate that FABP5 serves a key role in governing endocannabinoid signaling in nociceptors to disrupt TRPV1 sensitization and pain, and position FABP5 as a therapeutic target for the development of analgesics. OPEN www.nature.com/scientificreports/ prostaglandin E2 increase its efficacy and potency at TRPV1, which may enhance hyperalgesia 21-24 . The structurally related FAAH substrate oleoylethanolamide (OEA) likewise serves as a potent TRPV1 agonist and can elicit pro-nociceptive effects in some settings 25 . Moreover, FAAH regulates the metabolism of N-acyl taurines that activate TRPV1 26 . Taken together, although FAAH inhibition produces antinociceptive effects in diverse preclinical models of pain, its inhibition simultaneously augments the levels of distinct classes of TRPV1 agonists. Consistent with this notion, FAAH knockout (KO) mice exhibit a pro-nociceptive phenotype characterized by exacerbated TRPV1 activation 27 .Activation of TRPV1 in nociceptors stimulates the release of neurotransmitters and neuropeptides into the dorsal horn of the spinal cord, including the pro-nociceptive neuropeptide calcitonin gene-related peptide (CGRP) whose release is enhanced in the setting of chronic pain and is tempered by analgesics 28-30 . AEA exhibits biphasic activity wherein it inhibits TRPV1-mediated CGRP release from dorsal root ganglia (DRG) at low concentrations via CB1 while enhancing CGRP release and nociceptor output at higher concentrations18,24,[31][32][33][34][35][36][37][38]. Moreover, stimulation of AEA synthesis in DRG neurons promotes TRPV1 activation 8 , suggesting that interventions that increase AEA levels during inflammation could in principle exacerbate pain. Consequently, a major question that remains to be answered is whether augmenting AEA levels in TRPV1 + nociceptors leads to CB1mediated antinociceptive effects or pro-nociceptive effects elicited by TRPV1 activation.Fatty acid binding proteins (FABPs) are a family of intracellular carriers for endocannabinoids39,40. We previously identified FABP5 as the major FABP subtype that mediates AEA transport to FAAH for inactivation 39 . In addition to AEA, FABP5 also governs the transport of the related N-acylethanolamines OEA and palmitoylethanolamide (PEA), which serve as agonists at peroxisome proliferator-activated receptor alpha (PPARα) 41,42 . Consequently, pharmacological or genetic inhibition of FABP5 elevates AEA, PEA, and OEA levels and produces antinociceptive effects via CB1 and PPARα 43-45 . FABP5 is the predominant FABP subtype in sensory neurons and is extensively distributed in TRPV1 + nociceptors45,46. Similarly, CB1, PPARα, and FAAH are widely expressed in sensory neurons while robust AEA synthesis and ensuing TRPV1 activation are observed in TRPV1 + nociceptors in vitro4,8,46,47. The overlapping distribution between FABP5 and AEA metabolizing enzymes in TRPV1 + nociceptors provides a unique opportunity to leverage FABP5 deletion to augment AEA levels in nociceptors and in turn resolve outstanding questions underlying the functional role of AEA as a modulator of TRPV1 activation and inflammatory pain.
The endocannabinoid system has emerged as an attractive target for the development of analgesics. The endocannabinoids anandamide (AEA) and 2-arachidonoylglycerol (2-AG) are endogenous ligands for cannabinoid receptors 1 and 2 (CB1 and CB2, respectively), whose activation produces antinociceptive effects in diverse models of pain [bib_ref] International Union of Basic and Clinical Pharmacology. LXXIX. Cannabinoid receptors and their..., Pertwee [/bib_ref] [bib_ref] Targeting fatty acid amide hydrolase (FAAH) to treat pain and inflammation, Schlosburg [/bib_ref] [bib_ref] Peripheral gating of pain signals by endogenous lipid mediators, Piomelli [/bib_ref]. In the periphery, the CB1 receptor is expressed in nociceptive sensory neurons wherein it mediates the antinociceptive effects of phytocannabinoids [bib_ref] Cannabinoids mediate analgesia largely via peripheral type 1 cannabinoid receptors in nociceptors, Agarwal [/bib_ref] [bib_ref] Pre-and postsynaptic distribution of cannabinoid and mu opioid receptors in rat spinal..., Hohmann [/bib_ref] [bib_ref] Localization of central cannabinoid CB1 receptor messenger RNA in neuronal subpopulations of..., Hohmann [/bib_ref]. Sensory neurons also express enzymes that catalyze AEA biosynthesis and inactivation [bib_ref] Localization of the endocannabinoid-degrading enzyme fatty acid amide hydrolase in rat dorsal..., Lever [/bib_ref] [bib_ref] Anandamide produced by Ca(2+)-insensitive enzymes induces excitation in primary sensory neurons, Varga [/bib_ref] , indicative of a functional endocannabinoid system. Nociceptors express a host of ion channels that modulate pain including transient receptor potential vanilloid receptor 1 (TRPV1), a non-selective cation channel that is a central mediator of inflammatory pain [bib_ref] Impaired nociception and pain sensation in mice lacking the capsaicin receptor, Caterina [/bib_ref] [bib_ref] Vanilloid receptor-1 is essential for inflammatory thermal hyperalgesia, Davis [/bib_ref]. During inflammation, TRPV1 undergoes sensitization that sustains hyperalgesia [bib_ref] Inhibition of TNF reduces mechanical orofacial hyperalgesia induced by Complete Freund's Adjuvant..., Lis [/bib_ref] [bib_ref] The role of TRPV1 in different subtypes of dorsal root ganglion neurons..., Yu [/bib_ref] [bib_ref] Shp-1 dephosphorylates TRPV1 in dorsal root ganglion neurons and alleviates CFA-induced inflammatory..., Xiao [/bib_ref] [bib_ref] Contribution of interleukin-1 beta to the inflammationinduced increase in nerve growth factor..., Safieh-Garabedian [/bib_ref] while disruption of TRPV1 sensitization promotes analgesia [bib_ref] Mapping the binding site of TRPV1 on AKAP79: implications for inflammatory hyperalgesia, Btesh [/bib_ref]. Therefore, proteins that modulate TRPV1 sensitization constitute promising targets for the development of analgesics [bib_ref] Mapping the binding site of TRPV1 on AKAP79: implications for inflammatory hyperalgesia, Btesh [/bib_ref] [bib_ref] Targeting dorsal root ganglia and primary sensory neurons for the treatment of..., Berta [/bib_ref] [bib_ref] Transient receptor potential ion channels in primary sensory neurons as targets for..., Sousa-Valente [/bib_ref].
It is well-established that enhancement of peripheral AEA signaling via inhibition of its catabolic enzyme fatty acid amide hydrolase (FAAH) produces antinociceptive effects mediated by CB1 and/or CB2 receptors [bib_ref] Cannabinoids reduce hyperalgesia and inflammation via interaction with peripheral CB1 receptors, Richardson [/bib_ref] [bib_ref] Anandamide suppresses pain initiation through a peripheral endocannabinoid mechanism, Clapper [/bib_ref]. However, a clinical trial of a FAAH inhibitor failed to demonstrate clinically meaningful analgesia [bib_ref] An efficient randomised, placebo-controlled clinical trial with the irreversible fatty acid amide..., Huggins [/bib_ref]. A plausible mechanism to explain this discrepancy stems from previous observations that FAAH substrates including AEA engage additional molecular targets that may counteract the antinociceptive effects resulting from CB1 activation. In addition to serving as a CB1 ligand, AEA is also a TRPV1 agonist and inflammatory mediators such as
# Results
Inflammatory hyperalgesia in FABP5 cKO mice. FABP5 Flox mice were designed with loxP sites flanking exons 2 and 3 of the FABP5 gene [fig_ref] Figure 1: Characterization of FABP5 cKO mice [/fig_ref]. The mice were born at expected Mendelian frequencies and were otherwise normal. To ablate FABP5 in TRPV1 + nociceptors, we crossed FABP5 Flox mice with TRPV1-Cre mice to generate FABP5 Flox/TRPV1-Cre mice (hereafter referred to as FABP5 cKO). Reduced FABP5 expression in DRGs of FABP5 cKO mice was confirmed via qPCR and histologically [fig_ref] Figure 1: Characterization of FABP5 cKO mice [/fig_ref]. Note that residual FABP5 immunostaining in FABP5 cKO DRGs was observed in cells surrounding sensory neurons [fig_ref] Figure 1: Characterization of FABP5 cKO mice [/fig_ref] , presumably reflecting FABP5 expression in satellite glial cells.
We previously reported that pharmacological or genetic inhibition of FABP5 produces antinociceptive effects in models of inflammatory pain [bib_ref] Targeting fatty acid binding protein (FABP) anandamide transporters-A novel strategy for development..., Berger [/bib_ref] [bib_ref] Inhibition of fatty acid binding proteins elevates brain anandamide levels and produces..., Kaczocha [/bib_ref] [bib_ref] Fatty-acid-binding protein inhibition produces analgesic effects through peripheral and central mechanisms, Peng [/bib_ref] [bib_ref] Fatty acid binding protein deletion suppresses inflammatory pain through endocannabinoid/N-acylethanolamine-dependent mechanisms, Kaczocha [/bib_ref] [bib_ref] Fatty acid-binding protein 5 controls microsomal prostaglandin E synthase 1 (mPGES-1) induction..., Bogdan [/bib_ref]. Consequently, we assessed inflammatory hyperalgesia in mice receiving an intraplantar injection of carrageenan [bib_ref] Fatty acid-binding protein 5 controls microsomal prostaglandin E synthase 1 (mPGES-1) induction..., Bogdan [/bib_ref]. As expected, WT and FABP5 Flox mice developed thermal hyperalgesia while paw withdrawal latencies were significantly higher in global FABP5 KO and FABP5 cKO mice [fig_ref] Figure 2: Deletion of FABP5 in nociceptors suppresses inflammatory hyperalgesia [/fig_ref] [bib_ref] SAR studies on truxillic acid mono esters as a new class of..., Yan [/bib_ref] , providing evidence that FABP5 deletion in sensory neurons produces antinociceptive effects. Comparable effects were observed in female mice [fig_ref] Figure 2: Deletion of FABP5 in nociceptors suppresses inflammatory hyperalgesia [/fig_ref] , indicating that the function of FABP5 in pain modulation extends to both sexes. We additionally employed the behavioral spectrometer as a non-reflexive readout of inflammatory pain. The behavioral spectrometer quantifies diverse behavioral patterns in mice, a subset of which are pain-related and can be rescued by analgesics [bib_ref] Validation and implementation of a novel high-throughput behavioral phenotyping instrument for mice, Brodkin [/bib_ref]. Of the behavioral patterns assessed, we found that carrageenan administration to male and female WT mice resulted in an increase in "still" behavior while the remaining parameters were unchanged [fig_ref] Figure 2: Deletion of FABP5 in nociceptors suppresses inflammatory hyperalgesia [/fig_ref] , suggesting that mice remained more stationary in the setting of acute inflammatory pain. Administration of naproxen restored "still" patterns to pre-carrageenan baselines, confirming that these effects were pain-specific [fig_ref] Figure 2: Deletion of FABP5 in nociceptors suppresses inflammatory hyperalgesia [/fig_ref]. Notably, while carrageenan administration increased "still" behavior in male and female FABP5 Flox mice, it failed to do so in FABP5 KO and FABP5 cKO mice [fig_ref] Figure 2: Deletion of FABP5 in nociceptors suppresses inflammatory hyperalgesia [/fig_ref].
Contribution of CB1 and PPARα to pain hypersensitivity in FABP5 cKO mice. To characterize the mechanisms underlying pain modulation by FABP5, we first quantified the levels of AEA, 2-AG, PEA, and OEA in DRGs of FABP5 Flox and FABP5 cKO mice. Compared to FABP5 Flox mice, AEA, PEA, and OEA levels were elevated in DRGs obtained from FABP5 cKO mice while 2-AG levels remained unchanged, and no sexspecific differences were noted [fig_ref] Figure 3: CB1 and PPARα mediate the antinociceptive effects in FABP5 cKO mice [/fig_ref]. Carrageenan administration did not alter AEA, PEA, or OEA levels in FABP5 cKO mice [fig_ref] Figure 3: CB1 and PPARα mediate the antinociceptive effects in FABP5 cKO mice [/fig_ref]. The expression of FAAH and the N-acylethanolamine biosynthetic enzyme NAPE-PLD were comparable between the genotypes and sexes [fig_ref] Figure 3: CB1 and PPARα mediate the antinociceptive effects in FABP5 cKO mice [/fig_ref]. To validate these results, we employed the FAAH substrate AMC-AEA [fig_ref] Figure 3: CB1 and PPARα mediate the antinociceptive effects in FABP5 cKO mice [/fig_ref] to assess FAAH activity in DRGs of both genotypes, which revealed comparable activity levels [fig_ref] Figure 3: CB1 and PPARα mediate the antinociceptive effects in FABP5 cKO mice [/fig_ref]. Similarly, we observed comparable hydrolysis of the phospholipase A1 substrate PED-A1 in DRGs of these mice [fig_ref] Figure 3: CB1 and PPARα mediate the antinociceptive effects in FABP5 cKO mice [/fig_ref]. Since PED-A1 is not specific for NAPE-PLD, we treated DRGs with the selective NAPE-PLD inhibitor LEI-401 to identify the enzymatic activity corresponding to NAPE-PLD [fig_ref] Figure 3: CB1 and PPARα mediate the antinociceptive effects in FABP5 cKO mice [/fig_ref]. Our results revealed comparable PED-A1 hydrolysis in DRGs of both genotypes treated with vehicle or LEI-401 [fig_ref] Figure 3: CB1 and PPARα mediate the antinociceptive effects in FABP5 cKO mice [/fig_ref]. Taken together, these results indicate that the elevations in N-acylethanolamine levels observed in FABP5 cKO mice stem from attenuated transport and metabolism rather than alterations in biosynthetic or catabolic enzymes. Since no sex-specific differences were observed in any of the parameters assessed, we employed mixed cohorts of male and female mice for all subsequent experiments.
## Fabp5 inhibition reduces inflammation-and trpv1-evoked cgrp release. inflammatory
hyperalgesia is characterized by exacerbated release of CGRP from sensory neuron terminals into the dorsal horn of the spinal cord [bib_ref] Gait analysis and pain response of two rodent models of osteoarthritis. Pharmacol, Ferland [/bib_ref] [bib_ref] Calcitonin gene-related peptide and pain: a systematic review, Schou [/bib_ref] [bib_ref] Enhanced release of immunoreactive CGRP and substance P from spinal dorsal horn..., Garry [/bib_ref]. CGRP levels are elevated in rodents and humans experiencing ongoing pain and are reduced by analgesics [bib_ref] Gait analysis and pain response of two rodent models of osteoarthritis. Pharmacol, Ferland [/bib_ref] [bib_ref] Calcitonin gene-related peptide and pain: a systematic review, Schou [/bib_ref] , thereby serving as a neurochemical proxy for nociceptor sensitization/output. We previously reported that the carrageenan-induced increase in spinal CGRP levels is blunted following global FABP5 deletion [bib_ref] Fatty acid-binding protein 5 controls microsomal prostaglandin E synthase 1 (mPGES-1) induction..., Bogdan [/bib_ref]. Consistent with these results, CGRP release was suppressed in both FABP5 KO and cKO mice following carrageenan challenge [fig_ref] Figure 4: FABP5 deletion attenuates carrageenan-induced and TRPV1-evoked spinal CGRP release [/fig_ref] , confirming that FABP5 deletion in nociceptors recapitulates the effects observed in global FABP5 KO mice. TRPV1 is extensively co-expressed with CGRP and FABP5 in sensory neurons [bib_ref] Fatty-acid-binding protein inhibition produces analgesic effects through peripheral and central mechanisms, Peng [/bib_ref] [bib_ref] Quantitative differences in neuronal subpopulations between mouse and human dorsal root ganglia..., Shiers [/bib_ref] [bib_ref] CGRPalpha-expressing sensory neurons respond to stimuli that evoke sensations of pain and..., Mccoy [/bib_ref] and carrageenan administration sensitizes TRPV1, leading to exacerbated CGRP release [bib_ref] Fatty acid-binding protein 5 controls microsomal prostaglandin E synthase 1 (mPGES-1) induction..., Bogdan [/bib_ref]. Consistent with this notion, while capsaicinevoked CGRP release was enhanced in WT mice, its release was blunted in spinal cords of FABP5 KO and FABP5 cKO mice [fig_ref] Figure 4: FABP5 deletion attenuates carrageenan-induced and TRPV1-evoked spinal CGRP release [/fig_ref] , indicating that FABP5 deletion in sensory neurons leads to attenuated TRPV1 sensitization. Administration of AM251 or GW6471 to FABP5 KO or cKO mice restored CGRP release to levels comparable to WT mice [fig_ref] Figure 4: FABP5 deletion attenuates carrageenan-induced and TRPV1-evoked spinal CGRP release [/fig_ref] , confirming that CB1 and PPARα activation underlies these effects.
## Fabp5 deletion blunts trpv1 sensitization via cb1 and pparα. carrageenan-evoked hyperalgesia
is dependent upon nerve growth factor (NGF) release, which plays an obligate role in TRPV1 sensitization [bib_ref] Inhibition of carrageenan-induced edema by indomethacin or sodium salicylate does not prevent..., Amann [/bib_ref] [bib_ref] Neutralization of endogenous NGF prevents the sensitization of nociceptors supplying inflamed skin, Koltzenburg [/bib_ref] [bib_ref] Carrageenan-induced thermal hyperalgesia in the mouse: role of nerve growth factor and..., Sammons [/bib_ref] [bib_ref] Bradykinin and nerve growth factor release the capsaicin receptor from PtdIns(4,5)P2-mediated inhibition, Chuang [/bib_ref] [bib_ref] NGF rapidly increases membrane expression of TRPV1 heat-gated ion channels, Zhang [/bib_ref]. NGF activates the TrkA receptor that is co-expressed with TRPV1 in numerous sensory neurons 60 . We confirmed that DRG neurons co-express FABP5 and TrkA [fig_ref] Figure 5: Co-expression of FABP5 with [/fig_ref] and subsequently employed calcium imaging to determine the impact of FABP5 deletion upon TRPV1 sensitization by NGF. TRPV1 undergoes rapid tachyphylaxis following repeated capsaicin application 61 that can be largely restored by mediators that sensitize TRPV1 www.nature.com/scientificreports/ including NGF [fig_ref] Figure 6: FABP5 deletion blunts NGF-mediated TRPV1 sensitization via CB1 and PPARα [/fig_ref]. Our results revealed that while NGF sensitized TRPV1 in WT DRGs, it failed to do so in DRGs obtained from FABP5 KO mice [fig_ref] Figure 6: FABP5 deletion blunts NGF-mediated TRPV1 sensitization via CB1 and PPARα [/fig_ref]. Comparable TRPV1 and TrkA expression was observed between the genotypes [fig_ref] Figure 6: FABP5 deletion blunts NGF-mediated TRPV1 sensitization via CB1 and PPARα [/fig_ref] , confirming that these results do not stem from altered TRPV1 or TrkA expression. Application of the CB1 antagonist AM251 restored the ability of NGF to sensitize TRPV1 in FABP5 KO DRGs [fig_ref] Figure 6: FABP5 deletion blunts NGF-mediated TRPV1 sensitization via CB1 and PPARα [/fig_ref] and is consistent with the extensive co-expression between FABP5 and CB1 [fig_ref] Figure 5: Co-expression of FABP5 with [/fig_ref]. The PPARα antagonist GW6471 was likewise effective in restoring the ability of NGF to sensitize TRPV1 [fig_ref] Figure 6: FABP5 deletion blunts NGF-mediated TRPV1 sensitization via CB1 and PPARα [/fig_ref]. CB1 and PPARα expression was comparable between the genotypes [fig_ref] Figure 6: FABP5 deletion blunts NGF-mediated TRPV1 sensitization via CB1 and PPARα [/fig_ref]. Mechanistically, exogenously applied CB1 agonists suppress TRPV1 sensitization via calcineurin [bib_ref] The cannabinoid WIN 55,212-2 inhibits transient receptor potential vanilloid 1 (TRPV1) and..., Patwardhan [/bib_ref]. Consistent with these observations, the calcineurin inhibitor cyclosporin A largely restored TRPV1 sensitization in FABP5 KO DRGs [fig_ref] Figure 6: FABP5 deletion blunts NGF-mediated TRPV1 sensitization via CB1 and PPARα [/fig_ref] , indicating that a common mechanism is shared between exogenous and endogenous CB1 agonists. Collectively, these data provide evidence that FABP5 deletion promotes CB1 and PPARα activation to blunt NGF-mediated TRPV1 sensitization.
To determine whether these effects are observed in vivo, we assessed capsaicin-evoked nocifensive (paw licking/biting) behaviors in WT and FABP5 KO mice pre-treated with vehicle or NGF. Intraplantar capsaicin induced comparable nocifensive behaviors in WT and FABP5 KO mice at baseline [fig_ref] Figure 7: Capsaicin-induced nocifensive behavior in WT and FABP5 cKO mice [/fig_ref]. In contrast, while administration of NGF potentiated capsaicin-evoked nocifensive behaviors in WT mice, it failed to do so in mice lacking FABP5 [fig_ref] Figure 7: Capsaicin-induced nocifensive behavior in WT and FABP5 cKO mice [/fig_ref]. Administration of vehicle or NGF alone did not elicit a robust behavioral response. Collectively, our results provide direct evidence that FABP5 is essential for NGF-mediated TRPV1 sensitization in vivo.
## Faah inhibition produces distinct effects upon trpv1.
Based on the role of FABP5 as an intracellular carrier that delivers AEA to FAAH 39 , inhibition of FAAH is expected to phenocopy the effects of FABP5 upon TRPV1 sensitization. However, FAAH inhibition was previously reported to enhance TRPV1 activation in DRG neurons while intraplantar capsaicin administration exacerbated nocifensive behaviors in FAAH KO mice [bib_ref] Localization of the endocannabinoid-degrading enzyme fatty acid amide hydrolase in rat dorsal..., Lever [/bib_ref] [bib_ref] A pro-nociceptive phenotype unmasked in mice lacking fatty-acid amide hydrolase, Carey [/bib_ref]. These results suggest that blockade of FAAH and FABP5 may produce non-overlapping and possibly opposing effects upon TRPV1 sensitization. We explored this possibility and found that application of the selective FAAH inhibitor PF3845 failed to blunt NGF-mediated TRPV1 sensitization despite elevating AEA levels in DRGs [fig_ref] Figure 6: FABP5 deletion blunts NGF-mediated TRPV1 sensitization via CB1 and PPARα [/fig_ref]. Moreover, administration of the peripherally-restricted FAAH inhibitor URB937 (1 mg/kg, i.p.) [bib_ref] Anandamide suppresses pain initiation through a peripheral endocannabinoid mechanism, Clapper [/bib_ref] did not suppress capsaicin-evoked CGRP release [fig_ref] Figure 6: FABP5 deletion blunts NGF-mediated TRPV1 sensitization via CB1 and PPARα [/fig_ref]. Taken together, these results provide further evidence that while inhibition of both FABP5 and FAAH results in elevated AEA levels, distinct effects upon TRPV1 sensitization are observed.
# Discussion
Augmentation of endocannabinoid signaling represents an attractive strategy to treat pain. Initial translational efforts to harness the therapeutic potential of the endocannabinoid system were dampened following the failure of a FAAH inhibitor in a clinical trial of osteoarthritis pain [bib_ref] An efficient randomised, placebo-controlled clinical trial with the irreversible fatty acid amide..., Huggins [/bib_ref]. A mechanism(s) underlying this seeming lack of efficacy has yet to emerge. While systemic elevations in AEA levels produce antinociceptive effects in most www.nature.com/scientificreports/ preclinical models of pain, inflammatory mediators can convert AEA into a potent TRPV1 agonist in nociceptive sensory neurons while TRPV1-mediated nocifensive responses are observed in mice lacking FAAH [bib_ref] Inflammatory mediators convert anandamide into a potent activator of the vanilloid type..., Singh Tahim [/bib_ref] [bib_ref] A pro-nociceptive phenotype unmasked in mice lacking fatty-acid amide hydrolase, Carey [/bib_ref]. Therefore, it is conceivable that the dual function of AEA as a CB1 and TRPV1 agonist may contribute at least in part to the observed lack of efficacy of the FAAH inhibitor. However, the molecular targets engaged by AEA in nociceptors in vivo have not been established and the goal of the current work was to determine whether FABP5 deletion and ensuing AEA elevation leads to CB1-mediated antinociceptive effects or TRPV1-mediated pro-nociceptive effects. Previous efforts to characterize the role of the peripheral endocannabinoid system in pain modulation have yielded important insights at the behavioral level. Administration of peripherally-restricted FAAH inhibitors resulted in CB1-and/or PPARα-mediated antinociceptive effects [bib_ref] Anandamide suppresses pain initiation through a peripheral endocannabinoid mechanism, Clapper [/bib_ref] , providing strong evidence that enhancement of the peripheral AEA and PEA tone is sufficient to modulate pain. Moreover, deletion of CB1 receptors from nociceptors abolished the antinociceptive effects of exogenous CB1 agonists 4 , further positioning CB1 receptors in sensory neurons as key determinants of cannabinoid analgesia. Our present results demonstrate that FABP5 mediates AEA transport in DRGs as its deletion elevates AEA levels, consistent with the high degree of co-expression between FABP5 and FAAH 45 . Our combined histological and behavioral data reveal that FABP5 is expressed in CB1 + sensory neurons while its ablation augments AEA levels to trigger antinociceptive effects www.nature.com/scientificreports/ mediated by CB1 receptors. The selective increase in AEA but not 2-AG levels likely suggests that monoacylglycerol lipase may maintain access to 2-AG on the plasma membrane while bypassing the need for cytosolic transport by FABP5. Alternatively, the possibility that other proteins contribute to cytosolic 2-AG trafficking cannot be excluded at this time. The molecular mechanisms underlying pain modulation by endocannabinoids have been partially elucidated with TRPV1 emerging as a major target of interest given its prominent role in pain. It was previously demonstrated that application of exogenous CB1 agonists disrupts TRPV1 sensitization in DRG neurons in vitro through a calcineurin-dependent mechanism 60,62,63 . Employing capsaicin-evoked CGRP release and NGF-mediated TRPV1 sensitization as readouts, our data revealed that FABP5 ablation elevates AEA levels and blunts TRPV1 sensitization via a CB1-and calcineurin-mediated mechanism. TRPV1 sensitization was also restored following antagonism of PPARα, consistent with a report demonstrating that PPARα physically interacts with TRPV1 and its activation desensitizes the channel [bib_ref] Functional and biochemical interaction between PPARalpha receptors and TRPV1 channels: Potential role..., Ambrosino [/bib_ref]. Notably, these data support our previous findings that inhibition of peripherally-expressed FABP5 produces antinociceptive effects via CB1 and PPARα [bib_ref] Fatty-acid-binding protein inhibition produces analgesic effects through peripheral and central mechanisms, Peng [/bib_ref]. Unexpectedly, the inability of FAAH inhibition to block TRPV1 sensitization coupled with previous reports that FAAH inhibition promotes TRPV1 activation and pro-nociceptive effects 24,27,34-38 highlights divergent roles of FABP5 and FAAH in controlling TRPV1 function. www.nature.com/scientificreports/ There are several potential mechanisms that could underlie these contrasting effects. It is possible that FABP5 transports AEA and/or other FAAH substrates to TRPV1 for activation. Consequently, FABP5 deletion may compromise ligand delivery to TRPV1 while this function is preserved following FAAH inhibition leading to TRPV1 engagement. Additionally, it is possible that FABP5 and FAAH regulate the transport and metabolism of non-overlapping pools of TRPV1 ligands. For example, N-acyl taurines are TRPV1 agonists that are highly elevated following FAAH inhibition [bib_ref] A FAAH-regulated class of N-acyl taurines that activates TRP ion channels, Saghatelian [/bib_ref] and could in principle counteract AEA-and CB1-mediated antinociceptive effects. If N-acyl taurines are not FABP5 ligands and their levels unaltered following FABP5 inhibition, this could permit AEA-mediated antinociceptive effects to predominate. Regardless of the mechanism(s) involved, it is tempting to speculate that the inability of FAAH inhibition to temper TRPV1 sensitization may contribute at least in part to the lack of efficacy of a FAAH inhibitor in a clinical trial for pain [bib_ref] An efficient randomised, placebo-controlled clinical trial with the irreversible fatty acid amide..., Huggins [/bib_ref]. Collectively, our results establish a key role for FABP5 in shaping endocannabinoid and N-acylethanolamine signaling in nociceptors with FABP5 emerging as a novel modulator of TRPV1 sensitization. The contrasting effects between FABP5 and FAAH coupled with previously established anti-inflammatory effects observed following FABP5 inhibition position FABP5 as a promising target for the development of analgesics 49 .
# Methods
Compounds.
## Generation of fabp5 flox and fabp5 cko mice.
Mice carrying LoxP-flanked exons 2 and 3 were generated via standard gene-targeting techniques by Cyagen. A 17.5 Kb targeting vector containing a diphtheria toxin (DTA) cassette, neomycin cassette flanked by Frt sites, and exons 2-3 flanked by LoxP sites was electroporated into C57Bl/6 ES cells, which were subsequently selected for blastocyst microinjections. Founders were confirmed as germline-transmitted via crossbreeding with Flp-deleter mice. The F1 progeny were subsequently backcrossed to C57Bl/6 mice and then crossed with Trpv1-Cre mice (Jackson Labs #017769) to ablate FABP5 in TRPV1 + nociceptors. Control C57Bl/6J mice were from Jackson Labs.
Hargreaves. Thermal hyperalgesia was assessed as previously reported [bib_ref] Targeting fatty acid binding protein (FABP) anandamide transporters-A novel strategy for development..., Berger [/bib_ref]. Briefly, inflammation was induced by a unilateral, intraplantar injection of 1% carrageenan (20 µL in saline). The average latency to withdraw the paw was measured when a focused beam of radiant heat was applied to the mid plantar surface using a Hargreaves plantar apparatus (Ugo Basile). Mice were acclimated to the experimental room and Hargreaves for 2 days prior to baseline testing and at least 1 h prior to testing.
Behavioral spectrometer. Mice were acclimated to the Behavioral Sequencer box (Behavioral Instruments Inc, New Jersey) for 2 days prior to testing [bib_ref] Validation and implementation of a novel high-throughput behavioral phenotyping instrument for mice, Brodkin [/bib_ref]. Mice were allowed to move around freely within the 40 cm by 40 cm arena for 30 min during testing. Using a combination of cameras, infrared sensors and accelerometers embedded in the floor, the frequency and duration of predetermined behaviors were analyzed with Viewer 3 Immunofluorescence. Mice were anesthetized with isoflurane and transcardially perfused with saline followed by 4% paraformaldehyde (PFA) in 0.1 M phosphate buffered saline (PBS), pH 7.4. DRGs were dissected out from the spinal column and immersed in the same fixative overnight at 4ºC. The next day the tissue was immersed in 30% sucrose in 0.1 M PB for cryoprotection. The DRGs were embedded in a gelatin-albumin mixture (3% gelatin, 30% egg albumin in dH 2 O) and frozen using a liquid nitrogen-chilled isopentane bath. Cryostat sections 16 µm in thick were thaw-mounted on gelatin/chromium-coated glass slides, air dried, and stored in a frost-free freezer at − 20 °C until further processing. For immunofluorescence staining slides were thawed at room temperature for 10 min. Sections were fixed onto the slides with 4% PFA for 3 min, rinsed with PBS three times for 10 min each, and incubated for 30 min in PBS containing 5% normal donkey serum (NDS). Slides were incubated 24-48 h at 4 °C with a mixture of primary antibodies diluted in PBS with 0.3% Triton X-100 and 5% NDS. Following primary antibody incubation, the sections were again washed three times for 10 min each in PBS followed by 30 min in 5% NDS in PBS. The slides were incubated for 1 h at 37 °C with a mixture of secondary antibodies conjugated to either Alexa 488 or 594 diluted in PBS with 0.3% Triton X-100 and 5% NDS. Slides were subsequently washed three times with PBS for 10 min each and immediately mounted with Vectashield Plus® antifade mounting medium with DAPI (Vector Labs Burlingame, CA, Cat# H-2000) and stored at 4 °C in the dark. Fluorescent immunoreactivity in cells was observed with a Zeiss Axioplan 2 epifluorescent microscope. Images were obtained using Zeiss AxioCam HRm monochrome digital camera, and AxioVision Rel. 4.6 microscope software. Images were only adjusted for brightness and contrast.
qPCR. RNA was extracted using the RNeasy mini kit (Qiagen) followed by cDNA synthesis using the Super-Script III First Strand synthesis system (ThermoFisher). qPCR was performed with PowerUp SYBR green (Ther-moFisher) on a StepOnePlus instrument (Applied Biosystems). Quantification was performed using the 2 −ΔΔCt method with actin serving as the housekeeping gene. The following primers were used: CGRP. CGRP release was performed as previously described [bib_ref] Fatty acid-binding protein 5 controls microsomal prostaglandin E synthase 1 (mPGES-1) induction..., Bogdan [/bib_ref] Dissociated DRG neurons were incubated with 5 µM fura-2AM in EC1A buffer for 30 min at 37 °C. DRG neurons loaded with fura-2AM were pulsed alternately with 340 and 380 nm wavelength illumination (Lambda XL, Sutter Instruments, Novato, CA). Images were acquired using the ORCA-Flash4.0 digital camera at a capture rate of 0.33 Hz. The fluorescence ratio for individual neurons was determined as the intensity of emission during 340 nm excitation (I340) divided by 380 nm emission (I380) and used as an indicator of change in cytoplasmic calcium. The I340/I380 ratio was calculated on a pixel-by-pixel basis using the MetaFluor software (Molecular Devices, Sunnyvale, CA, USA). NGF (2.5S murine, Gibco, Waltham, Massachusetts) at 100 ng/ml and capsaicin (Sigma-Aldrich, St Louis, MO) at 100 nM in EC1A were applied using an array of quartz fiber flow pipes (500 mm internal diameter) positioned about 1 mm away from DRGs. Drugs were added to Fura-2AM, EC1A, NGF and capsaicin solutions so that cells were continually bathed in the following concentrations: 10 µM GW6471, 1 µM AM251, 1 µM PF3845 (Cayman, Ann Arbor, MI), and 200 nM cyclosporin A (TCI, Portland, OR).
Statistical analysis. Data are represented as mean ± SD. Thermal hyperalgesia was analyzed by One-Way ANOVA followed by Tukey post-hoc while behavioral spectrometer assays were analyzed via paired t-test. Changes in lipid levels were assessed via Two-Way ANOVA followed by Bonferroni post-hoc or unpaired t-test as appropriate. Carrageenan-evoked CGRP release assays were analyzed by unpaired t-test while capsaicinevoked release was analyzed via One-Way ANOVA followed by Dunnett's post-hoc or unpaired t-test as appropriate. Calcium imaging was examined via One-Way ANOVA followed by Dunnett's post-hoc test. qPCR data were analyzed by comparing ΔCt values using an unpaired t-test.
## Ethics statement. all experiments were approved by the stony brook university animal care and use
Committee (#277150). The experiments met the United States Public Health Service Policy guidelines on Humane Care and Use of Laboratory Animals. Mice were housed in an AAALAC certified facility with ad libitum access to food and water, and lighting was maintained on a 12-h light/dark cycle. Data reporting in the manuscript follows the recommendations in the ARRIVE guideline. All experiments were performed on 8-to 12-week-old mice.
## Data availability
Data will be available from the corresponding author upon reasonable request.
[fig] Figure 1: Characterization of FABP5 cKO mice. (A) Schematic showing the generation of FABP5 Flox mice and placement of LoxP sites. (B) PCR analysis of WT, heterozygous, and homozygous FABP5 Flox mice. (C) qPCR analysis of relative FABP5 expression in DRGs of FABP5 Flox and FABP5 cKO mice. **p < 0.01 calculated using unpaired t test (n = 3). (D) FABP5 distribution in DRGs of FABP5 Flox and FABP5 cKO mice. Scientific Reports | (2022) 12:9241 | https://doi.org/10.1038/s41598-022-13284-0 [/fig]
[fig] Figure 2: Deletion of FABP5 in nociceptors suppresses inflammatory hyperalgesia. (A) Thermal withdrawal latencies in male and female WT, FABP5 KO, FABP5 Flox , and FABP5 cKO mice at baseline and 4 h after intraplantar carrageenan injection. ***p < 0.001 calculated using one-way ANOVA followed by Tukey post-hoc test (n = 6). (B) Behavioral spectrometer analysis of still behavior at baseline and after carrageenan injection in WT male and female mice administered naproxen (30 mg/kg, i.p.) or vehicle. **p < 0.01 calculated using paired t test (n = 6). (C) Still behavior in male and female WT, FABP5 KO, FABP5 Flox , and FABP5 cKO mice at baseline and 4 h after carrageenan injection. **p < 0.01; ***p < 0.001 calculated using paired t test (n = 6). [/fig]
[fig] Figure 3: CB1 and PPARα mediate the antinociceptive effects in FABP5 cKO mice. (A) Levels of AEA, PEA, OEA, and 2-AG in DRGs of male and female FABP5 Flox and FABP5 cKO mice at baseline and 4 h after intraplantar carrageenan injection. **p < 0.01; ***p < 0.001 vs WT and FABP5 Flox calculated using two-way ANOVA followed by Bonferroni post-hoc test (n = 5). (B) Percent expression of CB1, PPARα, FAAH, NAPE-PLD, and ABHD4 in DRGs of male and female FABP5 cKO mice relative to FABP5 Flox controls. (C) FAAH and NAPE-PLD activity in DRGs. Left: DRGs were incubated with the selective FAAH substrate AMC-AEA (100 µM). Right: DRGs were incubated with the phospholipase A1 substrate PED-A1 (10 µM) in the presence or absence of the selective NAPE-PLD inhibitor LEI-401 (30 µM) (n = 3). (D) Thermal withdrawal latencies in FABP5 cKO mice receiving vehicle, AM251 (3 mg/kg, i.p.), or GW6471 (4 mg/kg, i.p.). ***p < 0.001 calculated using one-way ANOVA followed by Dunnett's post-hoc test (n = 6). [/fig]
[fig] Figure 4: FABP5 deletion attenuates carrageenan-induced and TRPV1-evoked spinal CGRP release. (A) WT, FABP5 KO, and FABP5 cKO mice received intraplantar injections of vehicle or carrageenan and lumbar spinal cords were obtained 4 h later. The sections were incubated for 30 min and CGRP release was subsequently quantified. *p < 0.05 calculated using unpaired t test (n = 5). (B) WT, FABP5 KO, and FABP5 cKO mice were administered vehicle, 3 mg/kg AM251, or 4 mg/kg GW6471 prior to intraplantar injections of carrageenan. Lumbar spinal cord sections were collected 4 h later and incubated with vehicle followed by 1 µM capsaicin in the presence or absence of 3 µM AM251 or 10 µM GW6471 for 30 min. CGRP release was quantified prior to and after the addition of capsaicin. **p < 0.01; ***p < 0.001 calculated using one-way ANOVA followed by Dunnett's post-hoc test (n = 5).Scientific Reports | (2022) 12:9241 | https://doi.org/10.1038/s41598-022-13284-0 [/fig]
[fig] Figure 5: Co-expression of FABP5 with (A) TrkA and (B) CB1 in DRG neurons. Representative images of FABP5, CB1, and TrkA distribution in L4-L5 DRGs. DAPI is shown in blue. Scientific Reports | (2022) 12:9241 | https://doi.org/10.1038/s41598-022-13284-0 [/fig]
[fig] Figure 6: FABP5 deletion blunts NGF-mediated TRPV1 sensitization via CB1 and PPARα. (A) Design of calcium imaging experiment and representative calcium traces in WT and FABP5 KO DRGs incubated with NGF and capsaicin. DRG neurons received two 30 s pulses of 100 nM capsaicin (Cap, gray bars) interspersed with a 5 min incubation of vehicle (black trace) or 100 ng/ml NGF (blue trace). KCl was employed to confirm neuronal identity. (B) Capsaicin response ratio (2nd/1st peak) in WT (white bars) and FABP5 KO (gray bars) DRGs incubated with vehicle or NGF in the presence or absence of 3 µM AM251, 10 µM GW6471, or 200 nM cyclosporin A. The effect of FAAH inhibition upon capsaicin responses was evaluated by incubating WT DRGs with 1 µM PF3845 for 2 h prior to capsaicin application. ***p < 0.001 calculated using one-way ANOVA followed by Dunnett's post-hoc test. (C) qPCR analysis of relative expression levels of TRPV1, TrkA, CB1, and PPARα in male and female FABP5 KO DRGs relative to WT controls (n = 3). (D) Levels of AEA in WT DRGs incubated for 2 h with vehicle or 1 µM PF3845. *p < 0.05 calculated using unpaired t test (n = 3). (E) Capsaicin-evoked CGRP release in WT mice administered vehicle or URB937 (1 mg/kg, i.p.). Mice received URB937 1 h prior to intraplantar carrageenan administration and spinal CGRP release elicited by 1 µM capsaicin was examined 4 h later (n = 5). (F) AEA levels in DRGs and lumbar spinal cords (LSC) of mice injected with vehicle or URB937 (1 mg/kg, i.p.). *p < 0.05 calculated using unpaired t test (n = 3). Scientific Reports | (2022) 12:9241 | https://doi.org/10.1038/s41598-022-13284-0 [/fig]
[fig] Figure 7: Capsaicin-induced nocifensive behavior in WT and FABP5 cKO mice. Mice received an intraplantar injection of vehicle or 200 ng NGF followed by vehicle or 1 µg capsaicin. The duration of nocifensive behavior defined as paw licking or biting was subsequently scored for 5 min. *p < 0.05 calculated using unpaired t test (Inc, St Augustin, Germany). Mice were placed in the box 4 h after carrageenan or saline injection.Capsaicin-induced nocifensive behavior. Mice were acclimated for two days to a transparent Plexiglass enclosure (8″ × 8″) and allowed to move freely for 20 min. On the day of testing, the mice received an intraplantar injection of 200 ng NGF (Gibco, Waltham, Massachusetts) or vehicle (saline, 10 µl) followed by 1 µg capsaicin (Sigma-Aldrich, St Louis, MO) or vehicle (5% ethanol in saline, 10 µl) 30 min later. The duration of nocifensive behaviors (paw licking and/or biting) was assessed for 5 min immediately following the injections. [/fig]
|
A Risk Assessment Model for Type 2 Diabetes in Chinese
Aims: To develop a risk assessment model for persons at risk from type 2 diabetes in Chinese.Materials and Methods: The model was generated from the cross-sectional data of 16246 persons aged from 20 years old and over. C4.5 algorithm and multivariate logistic regression were used for variable selection. Relative risk value combined with expert decision constructed a comprehensive risk assessment for evaluating the individual risk category. The validity of the model was tested by cross validation and a survey performed six years later with some participants.Results: Nine variables were selected as risk variables. A mathematical model was established to calculate the average probability of diabetes in each cluster's group divided by sex and age. A series of criteria combined with relative RR value (2.2) and level of risk variables stratified individuals into four risk groups (non, low, medium and high risk). The overall accuracy reached 90.99% evaluated by cross-validation inside the model population. The incidence of diabetes for each risk group increased from 1.5 (non-risk group) to 28.2(high-risk group) per one thousand persons per year with six years followup.Discussion:The model could determine the individual risk for type 2 diabetes by four risk degrees. This model could be used as a technique tool not only to support screening persons at different risk, but also to evaluate the result of the intervention.
# Introduction
Type 2 diabetes is a worldwide public health problem resulting from both lifestyle and genetic factors. Although the pathogenesis of diabetes is unclear, fortunately, type 2 diabetes can be prevented by available lifestyle intervention [bib_ref] Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects..., Tuomilehto [/bib_ref].
High risk intervention is one of the main strategies in noncommunicable diseases prevention, which is also suitable for type 2 diabetes. One of problems is how to find and determine those at high risk within a population. Based on epidemiological research work, the risk factors associated with the onset of diabetes, especially type 2 diabetes, were clear. A set of assessment methodologies, from the simple checking list to several risk score models, was developed in recent years, such as the Finnish Risk Score, Danish Diabetes Risk Score [bib_ref] A Danish Diabetes Risk Score for Targeted Screening -The Inter99 Study, Glümer [/bib_ref] , ADA [bib_ref] Performance of recommended screening tests for undiagnosed diabetes and dysglycemia, Rolka [/bib_ref] , Cambridge Risk Score [bib_ref] Diabetes risk score: Towards earlier detection of type 2 diabetes in general..., Griffin [/bib_ref] , NHANESIII [bib_ref] Predicting impaired glucose tolerance using common clinical information: Data from the Third..., Nelson [/bib_ref] , DRC [bib_ref] Diabetes Risk Calculator: A simple tool for detecting undiagnosed diabetes and pre-diabetes, Heikes [/bib_ref] , Thailand Risk Score [bib_ref] A risk score for predicting incident diabetes in the Thai population, Aekplakorn [/bib_ref] , Spanish Diabetes Risk Score [bib_ref] Validation of the FINDRISC (Finnish Diabetes Risk Score) for prediction of the..., Soriguer [/bib_ref] and so on [bib_ref] Survey of diabetes risk assessment tools: Concept, structure and performance, Thoopputra [/bib_ref] [bib_ref] Identifying risk and preventing progression to type 2 diabetes in vulnerable and..., Taylor [/bib_ref]. All these tools are helpful in the assessment of high risk and of persons with diabetes all over the world but only a few are developed in China [bib_ref] A simple Chinese risk score for undiagnosed diabetes, Gao [/bib_ref] [bib_ref] A simple tool detected diabetes and prediabetes in rural Chinese, Xin [/bib_ref].Considering the large number of individuals with prediabetes or at high-risk of diabetes in China [bib_ref] Risk score based on self-reported or available clinical data to detect undiagnosed..., Brown [/bib_ref] , a tool, suitable for Chinese and with a high efficiency for dealing with the huge clinical data simultaneously in a precise way, is urgently needed.
The aim of this work was to develop a risk assessment model for type 2 diabetes in China, which was an available and affordable computerized model with common variables. The model could be used to assess different levels of risk for type 2 diabetes both at the individual's level and at the population's level, and therefore, to support the prevention of diabetes.
# Materials and methods
## Subjects
Form Mar to Nov. 2001, a cross-sectional baseline survey was drawn from 27 research institutes in Beijing, China. Most of the research institutes are located in the urban districts of Xicheng, Haidian and Shijingshan. A total of 16246 subjects aged 20 and older were enrolled when they participated in their annual health examination through a questionnaire on health behavior and clinical measurements. Almost all persons are researchers or graduated students (younger than 30 years old). These persons were highly educated and in a sedentary working pattern. Health behavior included age, sex, prior history of diabetes and parental or sibling history of diabetes (PSH), clinical examination included the measurements of height, weight, waist circumference, systolic blood pressure (SBP), diastolic blood pressure (DBP), cholesterol (CHOL), triglyceride (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), and fasting plasma glucose (GLU). Body mass index (BMI) was calculated dividing the weight (kg) by the height squared (m2). The investigations were executed by trained medical staff from these institutes. All the blood test were conducted by one clinical biochemistry department in an evaluated hospital. The baseline survey recruited 16246 persons. Among them were 15237 persons who were free of a history of diabetes and fasting glucose levels ,7.0 mmol/L, 1009 persons who have a history of diabetes or fasting glucose levels .7.0. Six years later, in 2007, during the same period as in 2001, the followup survey was carried out in four institutes of the above 27 institutes when about 2288 persons without diabetes in 2001 completed the follow-up investigation.
# Ethics statement
The research was approved by the 12th Five-year National Science and Technology Supporting Project and National Natural Science Foundation of China, and all participants gave written consent to participation in the study. The individual in this manuscript has given written informed consent to publish these case details.
## Model design and methods
Data mining and statistical methods combined with expert decisions were used to accomplish variable selection and analyze the impact of clinical variables on the risk of type 2 diabetes. The schematic diagram is shown in [fig_ref] Figure 1: The schematic diagram of the Risk assessment Model and Comprehensive Risk Assessment... [/fig_ref].
## Variable selection
The variable selection was firstly filtered based on the preprocessing of the variables. The variables, which had large vacant data, were deleted. Then, the remaining variables were accomplished using C4.5 decision tree [bib_ref] Feature selection with decision tree criterion, Grabczewski [/bib_ref] and multivariate logistic regression. The same variables chosen both by C4.5 and logistic regression were regarded as the final risk variables.
The total number of questionnaire and clinical variables was 96. After variables filtering, 57 remained. The variables list was shown in in File S1. Then, all the 16246 persons were divided into five sub-cohorts: (1) all persons; (2) all the males; (3) all the females; (4) all the persons under 50 year-old; (5) all the persons above 50 year-old. Therefore there were five cohorts as the training sets finally. C4.5 algorithm was applied to build the decision tree for each sub-cohort, the diabetic label was conducted as an identical class label. The C4.5 algorithm [bib_ref] Decision tree and PCA-based fault diagnosis of rotating machinery, Sun [/bib_ref] was executed for each candidate training set by Weka [bib_ref] The Weka data mining software: An update, Hall [/bib_ref] (Weka v.3-6-10, 2013, University of WAIKATO, Hamilton, New Zealand). The decision trees were obtained by 10-fold cross validation without pruning. The variables, whose frequencies of occurrence were higher than five of the first eight levels of the decision trees, would be the most relevant variables.
Meanwhile, we put the same 57 variables into multivariate logistic regression. The variables which were statistically significant (P,0.05) would be selected as the most relevant variables. The multivariate logistic regression was executed by Weka. Finally, age, diastolic blood pressure (DBP), high-density lipoprotein (HDL), waist, sex, cholesterol (CHOL), parental or sibling history, body mass index (BMI), and triglyceride (TG) were both selected by C4.5 and logistic regression simultaneously. The C4.5 and LR results of variable selection are shown in [fig_ref] Table 3: Incidence of diabetes of 6-years follow-up in 2288 subjects [/fig_ref] in File S1.
Furthermore, from the results of the C4.5 algorithm, the value around 5.85 mmol/L of glycemia appeared six times in the first eight layers and four times in the first two layers. So the value 5.85 mmol/L was chosen as the cut-off point of glycemia for persons at high risk. Concretely, if the level of an individual's glycemia was higher than 5.85 mmol/L, the individuals were considered as a person at high risk.
## Fast clustering
It was clear that all of the selected variables were not specific to diabetes. However, any single variable could not be used to identify the risk of diabetes. To extract the features of persons at risk, the fast clustering was used to divide the population into different clusters by all selected variables, which could lead to a greater similarity between persons within the same cluster and a greater diversity between persons in different clusters.
First, the 16246 data were standardized. Then, the standardized data were put into the fast clustering analysis (fastclus) [bib_ref] Using cluster analysis to examine dietary patterns: Nutrient intakes, gender, and weight..., Wirfalt [/bib_ref]. Fastclus was performed to estimate the different characteristics of the data based on nine variables which had been selected as risk variables.
Two steps were applied to determine the most appropriate number of clusters: (1) Several runs of the fastclus procedure were executed by setting different cluster numbers from 2 to 7; (2) Square of R (R 2 ) and cubic clustering criterion (CCC) [bib_ref] Clustering analysis on the dietary patterns of residents in Jiangsu Province, Tian [/bib_ref] were calculated in each run. Generally, a smaller number of clusters enhances the ease of interpretation, and a higher increment of R 2 and a higher CCC indicated a better separation of clusters. So the smaller cluster number was chosen by CCC $10, and the most significant increments of R 2 . The results of the preselected number of clusters are shown in [fig_ref] Table 1: Characteristics of model and follow-up subjects [/fig_ref] in File S1.
According to the results of R 2 and CCC, the optimal cluster number was 3. The fastclus procedure was conducted by R [bib_ref] R: A data analysis and statistical programming environment-An emerging tool for the..., Grunsky [/bib_ref] (R v3.0.2, R Foundation for Statistical Computing, Wirtschaftsuniversitä t Wien), and three cohorts were finally obtained.
## Relative risk calculation
After the fastclus procedure, the model population was divided into three groups with relative significant similarities in each group. The characteristics of the persons in the three clusters are shown in in File S1. The relative risk for individuals in each group was calculated by the following two steps.
First, the probability of diabetes for each person in each cluster was calculated. Second, relative risk was a ratio calculated by individual probability divided by average probability of the onset of diabetes, which was computed according to age and gender in his or her corresponding cluster.
3.1 Logistic Regression. Logistic regression was needed to calculate the probability of the onset of diabetes both for individuals and for each cluster primarily.
For each cluster obtained by fast clustering, 16246 data from the different clusters were put into the multivariate logistic model. Then the goodness of fit for the logistic regression models was evaluated by the Hosmer and Lemeshow test [bib_ref] AUSDRISK: An Australian type 2 diabetes risk assessment tool based on demographic,..., Chen [/bib_ref]. Generally, a larger P value represented a better match. The Beta coefficient of each independent variable could be derived from the multivariate logistic regression for each cluster. The logistic regressions were built by R. The result of the Hosmer and Lemeshow test was shown in [fig_ref] Table 2: Distribution of risk degree and metabolic features of model population [/fig_ref] in File S1.
## Average probability of the onset of diabetes for each
Cluster. The 15237 subjects without diabetes were divided into 24 groups in each cluster by age (5 years a group between 20 to . 75 years old) and gender. The means of variables for each group were put into the logistic regression of each matching cluster to calculate the average probability of the onset of diabetes in each group.
3.3 Individual Relative Risk Calculation. The individual probability (P k ) of the onset of diabetes was calculated one by one for 15237 subjects. The average probabilities (P 0 ) of diabetes of the 24 groups in each cluster were also calculated by the mean value of variables of each group. The individual relative risk (RR) was calculated dividing individual probability P k by average probability P 0 of the age-gender matching group. The formula of RR is shown as below:
[formula] RR~P k P 0 [/formula]
The ratio of probability between any subject and the age-gender matching group (RR) indicated the individual relative degree of the risk of diabetes.
A cut-off point of RR, to distinguish the individual with a risk of diabetes, was chosen by receiver-operating characteristic (ROC) curve [bib_ref] Risk assessment tools for identifying individuals at risk of developingtype 2 diabetes, Buijsse [/bib_ref]. Commonly, the optimal cut-off point was identified as the coordinate closest to the y intercept (0, 1) of the ROC curve. The ROC curve was executed by R, and the curve of all the subjects is shown in [fig_ref] Figure 1: The schematic diagram of the Risk assessment Model and Comprehensive Risk Assessment... [/fig_ref] in File S1.
The area under the curve (AUC) of the ROC curve was 0.808. According to the criteria of choosing the optimal RR, the cut-off point of the RR value is 2.2, the sensitivity was 0.794, and specificity was 0.679.
## Criteria of comprehensive risk assessment
A criterion of comprehensive risk assessment (CCRA) was the final process in the construction of the risk assessment model. The CCRA executed in three steps [fig_ref] Figure 1: The schematic diagram of the Risk assessment Model and Comprehensive Risk Assessment... [/fig_ref]. and started from (1) absolute criteria: whether the glycemia level of an individual was higher than the value 5.85 mmol/L or not; (2) relative risk assessment: whether the RR of an individual was higher than 2.2 or not; (3) relative criteria of each cluster: For the individuals whose RR was higher than 2.2, a relative criteria of each cluster was used to judge degree of the risk by selected risk variables combined with characteristics of the cluster and clinical knowledge. For example, in first cluster, if an individual satisfied all three of the following criteria (degree = 3): (1) BMI .26.8; (2) CHOL . 5.18 mmol/L; (3) TG .1.7 mmol/L, this individual was assessed as being at high risk. If an individual satisfied any two of the three criteria (degree = 2), they were assessed as medium-risk. If they satisfied any one of the three criteria (degree = 1), the individuals were assessed as low-risk.
The pseudocode of the first cluster's main steps for risk assessment is shown below, other clusters were processed in a similar way to the first cluster but have different criteria. For the individuals whose RR was lower than 2.2, they would be identified as the non-risk individuals. The risk assessment model finally divided the model population into four risk categories: high-risk, medium-risk, low-risk and non-risk of diabetes.
-
## Model validation
The model validation was conducted in two steps: jackknife validation and follow-up validation.
In the jackknife validation, all 16246 subjects were classified into three categories manually: non-risk, different-risk (low-risk, medium-risk and high-risk) and subjects with diabetes. Then, the jackknife method was applied to proceed with the cross-validation. The jackknife methodrandomly split 16245 data into one set to fit the model, and left the one last data to test the assessment accuracy. For a leave-one-out estimate, all 16246 data were used to get an overall estimate of the prediction accuracy based on the risk assessment model. Through these 16246 evaluations, the overall accuracy of these categories were calculated by the following formula. The jackknife was executed by R.
## Oa~n umber of correct assessment number of truesamples
In the follow-up validation, a longitudinal data was used to validate the risk assessment model. The incidence of diabetes was calculated from a total of 2288 subjects without diabetes from the original model population in a follow-up six years later, in 2007. 62 persons developed diabetes during six-year follow-up. The incidence of diabetes from each original risk category was used as a comparison metric for the follow-up validation. The diabetes was diagnosed using WHO criteria (2006).
# Results
## Development of model
1. Characteristics of Baseline Data. The total number of this database was 16246 (1009 subjects with diabetes). Characteristics of 15237 subjects without diabetes on the baseline survey are shown in [fig_ref] Table 1: Characteristics of model and follow-up subjects [/fig_ref].
2. Results of Variable Selection. Nine variables were selected as risk variables for the risk assessment model both by a C4.5 decision tree and by logistic regression. They were: AGE, SEX, BMI, WAIST, CHOL, TG, HDL, DBP and PSH.
The occurrence frequencies (number) of each variable in all decision trees are shown in [fig_ref] Table 3: Incidence of diabetes of 6-years follow-up in 2288 subjects [/fig_ref] in File S1. The variables whose frequencies of occurrence were higher than four in the first eight levels were chosen as the most relevant variables. The results of the multivariate logistic regression are shown in in File S1, the variables whose P value was lower than 0.05 were chosen as the strong relevant variables.
Although the P value of HDL was higher than 0.05 in the multivariate logistic regression, HDL is a well-known protective factor for diabetes and this was strongly suggested by the decision trees. Meanwhile, the empirical experience showed that the CHOL had a high relevance to diabetes in women aged around 50. So HDL and a new variable obtained by multiplying the CHOL and sex were both selected as risk variables.
The value of glycemia around 5.85 was used as the absolute criteria of pre-diabetes in the risk assessment model.
3. Results of the Fast Clustering. The diabetes-related metabolic characteristics were well represented in the three groups by fast clustering using nine selected variables in File S1). The first and the biggest group represented the generally metabolic status of the adults. The second group indicated an alteration by aging and the third group was mainly composed of the persons with relatively high value of risk variables. The feature extracted by fast clustering offered a base for further risk assessment for diabetes with these common metabolic variables.
4. Results of Relative Risk Calculation. The average probability by age and gender in the three clusters is shown in [fig_ref] Figure 2: Average probability of diabetes in three clusters with each age and gender... [/fig_ref].
As shown in [fig_ref] Figure 2: Average probability of diabetes in three clusters with each age and gender... [/fig_ref] , the average probability of diabetes shows almost the same trend rising in parallel with age in the first and second cluster either for both men and women. An intersection of average probability between male and female presented at the age 55 in the third cluster. The average probability of diabetes was higher in women than in men. Furthermore, the average probability of diabetes in groups aged over 55 in the third cluster was higher than in the other clusters, especially in females.
5. Comprehensive Risk Assessment. The procedure of the comprehensive risk assessment was implemented as in block A of [fig_ref] Figure 1: The schematic diagram of the Risk assessment Model and Comprehensive Risk Assessment... [/fig_ref].
The model population (excluding persons with diabetes) was divided into four risk categories by this assessment model, the distribution of the degree of risk and the metabolic characteristics of each category are shown in [fig_ref] Table 2: Distribution of risk degree and metabolic features of model population [/fig_ref].
In this model population, around 20% of persons have a high and middle-high risk of diabetes. Meanwhile, the mean value of each variable increased significantly from the non-risk category to the high risk category except the HDL.
## Model validation
1. Cross-validation of Baseline Population. The accuracy of all risk determination by this risk assessment model was 90.99%, as shown in in File S1.
2. Follow-up validation. After six-year follow-up, the average incidence of diabetes in the model population was around 4.5 persons/1000 persons/year. The incidence increased significantly from the non-risk group (1.5 persons/1000 persons/year) to the high-risk group (28.2 persons/1000 persons/year). The incidence increased almost 19 times and 6 times in high-risk group than that in the non-risk and total group, respectively [fig_ref] Table 3: Incidence of diabetes of 6-years follow-up in 2288 subjects [/fig_ref]. The characteristics of 2288 persons both in 2001 and 2007 are shown in [fig_ref] Table 1: Characteristics of model and follow-up subjects [/fig_ref].
# Discussion
With nine common variables and one cut-off point value of glycemia as risk variables, a computerized assessment model was developed for the risk of type 2 diabetes in Chinese. The model was constructed by very common parameters including general personal information (age, gender, BMI, waist and blood pressure) metabolic variables (CHOL, TG, and HDL) and genetic factors (PSH), so this model is both available and affordable. The model could efficiently and accurately stratify the population into four different groups. The persons at high and medium risk will be targeted, and be thought of as persons at risk in the list of intervention.
The accuracy of the model was fully validated by the incidence of diabetes in the six-year follow-up data. The incidence of diabetes was dramatically increased from the non-risk persons (1.5 persons/1000 persons/year) to the high-risk persons (28.2 persons/1000 persons/year) during six-year follow-up of the model population. Furthermore, the percentage of high-risk persons could be clearly reduced by reassessment after six-month intervention (data was not shown).
Compared with different risk assessment tools that have been developed in the last ten years, the principle of thinking was quite similar. Including assessment variables, methods were proposed to account for risk, as well as certain criteria for determination status of risk. The main differences between our model and others are focused on two points: precision and efficiency. Taking the advantage of previous work [bib_ref] Quantitative influence of risk factors on blood glucose level, Chen [/bib_ref] [bib_ref] Analysis on health information extracted from an urban professional population in Beijing, Zhang [/bib_ref] [bib_ref] Some pathological knowledge discovered in large database of type 2 diabetes, Luo [/bib_ref] , the core of this assessment was built on the comprehensive analysis from selected risk variables to calculate the probability of diabetes by age and gender, a way of assessment which is based on computer technology. This model could be used for risk determination without any limitation of numbers and give the results in several seconds either for screening or for reassessment. All these features will be greatly facilitate the practices of intervention for diabetes.
One of the problems in developing the assessment model was how to determine the status of diabetes risk not only derived from impaired fasting glucose but also from impaired glucose tolerance. However, the risk variables related to both impaired regulations of glucose metabolism were reported as being quite similar. But the sensitivity of these risk factors in the assessment of the two impaired regulations of glucose metabolism was quite different. The accuracy of assessment for impaired fasting glucose, such as in our model, reaches 90%. The accuracy, evaluated by the same model, of assessment for impaired glucose tolerance reduced to 68% (data not shown). This presents the big challenge of improving the assessment model to recognize the risk of impaired glucose tolerance.
In conclusion, a risk assessment model was developed for type 2 diabetes in Chinese. The model could stratify the population into four risk groups. The model could be used as a technique tool to support the screening of persons at risk of type 2 diabetes and to evaluate the integrating effects of intervention.
## Supporting information
File S1 Supplemental Material. File S1 contains seven tables and one figure. They are: (1) Table S1 the optimal number of fastclus; [bib_ref] A Danish Diabetes Risk Score for Targeted Screening -The Inter99 Study, Glümer [/bib_ref] [fig_ref] Figure 1: The schematic diagram of the Risk assessment Model and Comprehensive Risk Assessment... [/fig_ref] receiver operating characteristic curve of RR. (DOC)
[fig] Figure 1: The schematic diagram of the Risk assessment Model and Comprehensive Risk Assessment Procedure. doi:10.1371/journal.pone.0104046.g001 [/fig]
[fig] Figure 2: Average probability of diabetes in three clusters with each age and gender matching group. doi:10.1371/journal.pone.0104046.g002 [/fig]
[table] Table 1: Characteristics of model and follow-up subjects (Mean 6 SD). [/table]
[table] Table 2: Distribution of risk degree and metabolic features of model population. doi:10.1371/journal.pone.0104046.t002 [/table]
[table] Table 3: Incidence of diabetes of 6-years follow-up in 2288 subjects. doi:10.1371/journal.pone.0104046.t003 [/table]
[table] Table S2: Hosmer and Lemeshow test for three logistic regressions; (3) Table S3 frequency of selected variables occurrence in all decision trees; (4) Table S4 risk factors and beta coefficient derived from multivariate logistic regression; (5) S5 the characteristics of different clusters (mean6SD); (6) Table S6 the results of jackknife cross-validation in model population; (7) Table S7 the list of 96 variables in risk variable selection; (8) [/table]
|
The generation of centripetal force when walking in a circle: insight from the distribution of ground reaction forces recorded by plantar insoles
Background: Turning involves complex reorientation of the body and is accompanied by asymmetric motion of the lower limbs. We investigated the distribution of the forces under the two feet, and its relation to the trajectory features and body medio-lateral displacement during curved walking. Methods: Twenty-six healthy young participants walked under three different randomized conditions: in a straight line (LIN), in a circular clockwise path and in a circular counter-clockwise path. Both feet were instrumented with Pedar-X insoles. An accelerometer was fixed to the trunk to measure the medio-lateral inclination of the body. We analyzed walking speed, stance duration as a percent of gait cycle (%GC), the vertical component of the ground reaction force (vGRF) of both feet during the entire stance, and trunk inclination. Results: Gait speed was faster during LIN than curved walking, but not affected by the direction of the curved trajectory. Trunk inclination was negligible during LIN, while the trunk was inclined toward the center of the path during curved trajectories. Stance duration of LIN foot and foot inside the curved trajectory (Foot-In) was longer than for foot outside the trajectory (Foot-Out). vGRF at heel strike was larger in LIN than in curved walking. At mid-stance, vGRF for both Foot-In and Foot-Out was higher than for LIN foot. At toe off, vGRF for both Foot-In and Foot-Out was lower than for LIN foot; in addition, Foot-In had lower vGRF than Foot-Out. During curved walking, a greater loading of the lateral heel occurred for Foot-Out than Foot-In and LIN foot. On the contrary, a smaller lateral loading of the heel was found for Foot-In than LIN foot. At the metatarsal heads, an opposite behaviour was seen, since lateral loading decreased for Foot-Out and increased for Foot-In. Conclusions: The lower gait speed during curved walking is shaped by the control of trunk inclination and the production of asymmetric loading of heel and metatarsal heads, hence by the different contribution of the feet in producing the body inclination towards the centre of the trajectory.
# Background
Human walking has been extensively studied from both the kinematic and kinetic perspectives [bib_ref] Biomechanics and muscle coordination of human walking. Part I: introduction to concepts,..., Zajac [/bib_ref] [bib_ref] Dynamic principles of gait and their clinical implications, Kuo [/bib_ref] [bib_ref] Normal walking speed: a descriptive meta-analysis, Bohannon [/bib_ref]. Most published studies deal with walking along straight paths. Few papers focus on walking along non-linear, e.g. curved, trajectories. Under the curved-walking condition, the central nervous system organizes the movement considering not only the propulsion required but also the equilibrium constraints connected to body rotation. Turning involves complex reorientation of the head, trunk, pelvis and feet [bib_ref] Human walking along a curved path. I. Body trajectory, segment orientation and..., Courtine [/bib_ref] [bib_ref] Tuning of a basic coordination pattern constructs straight-ahead and curved walking in..., Courtine [/bib_ref] [bib_ref] Insight into the generation of the centripetal force, Schmid [/bib_ref] [bib_ref] Coordinated modulation of locomotor muscle synergies constructs straight-ahead and curvilinear walking in..., Courtine [/bib_ref] [bib_ref] Video task analysis of turning during activities of daily living, Glaister [/bib_ref] , and is accompanied by adjustments of body orientation (such as trunk inclination to the inner part of the trajectory) to counteract the centrifugal acceleration acting on the walking body, and by asymmetric motion of the lower limbs, whereby the leg on the inside of the trajectory travels a shorter pathway than that on the outside [bib_ref] Human walking along a curved path. I. Body trajectory, segment orientation and..., Courtine [/bib_ref] [bib_ref] Human walking along a curved path. II. Gait features and EMG patterns, Courtine [/bib_ref]. Not unexpectedly, recent studies requiring subjects to travel both linear and circular pathways have detected abnormalities in patients with neurological disorders. Both Parkinson's disease [bib_ref] Walking along circular trajectories in Parkinson's disease, Guglielmetti [/bib_ref] and stroke patients [bib_ref] Curved walking in hemiparetic patients, Godi [/bib_ref] show walking difficulties, more evident during circular rather than linear trajectories. Therefore, given the fundamental and clinical relevance of curved walking, the present investigation was carried out to investigate foot action during curved walking, in the hypothesis that the distribution of the vertical forces under the feet can help in producing the coordinated motion of the turning body.
Knowledge about the spatio-temporal pattern of distribution of the ground reaction forces (GRF) when walking in a circle may be useful for assessing the control of medio-lateral equilibrium and the way subjects and patients accomplish the task. It could help in detecting changes due to diseases of the central [bib_ref] Curved walking in hemiparetic patients, Godi [/bib_ref] [bib_ref] Stabilometry is a predictor of gait performance in chronic hemiparetic stroke patients, Nardone [/bib_ref] [bib_ref] Linear and angular control of circular walking in healthy older adults and..., Goodworth [/bib_ref] or peripheral nervous system [bib_ref] Afferent control of walking: are there distinct deficits associated to loss of..., Nardone [/bib_ref] , and in estimating the evolution of the walking disorder or the potential advantage of gait rehabilitation [bib_ref] Changes of plantar pressure distributions following open and closed kinetic chain exercise..., Lee [/bib_ref] [bib_ref] Ground reaction force patterns in stroke patients with various degrees of motor..., Chen [/bib_ref]. Plantar pressure analysis from different points of the foot sole [bib_ref] Analysis of human locomotion by recording sole-floor reaction forces from anatomically discrete..., Warabi [/bib_ref] has a high degree of reliability [bib_ref] Repeatability of the Pedar-X in-shoe pressure measuring system, Ramanathan [/bib_ref] [bib_ref] Test-retest reliability of an insole plantar pressure system to assess gait along..., Godi [/bib_ref] , and can be useful and appropriate for such assessment.
We hypothesised that: 1) the distribution of the forces beneath the sole differs between linear and curved trajectories, owing to the different kinematics of the two feet during curved walking; 2) the force distribution during curved walking differs between the inner and outer foot, owing to the need to exert different GRFs in order to produce the centripetal acceleration; and 3) the placement of the point of application of the forces under the foot sole should help create or modify the torques acting in the frontal plane, so as to match the body's internal inclination during curved walking.
We analyzed the GRF collected by insole devices, since it gives an overall view of the time-course of the vertical forces and its peaks [bib_ref] Insight into the generation of the centripetal force, Schmid [/bib_ref] [bib_ref] The influence of mechanical compression on the induction of osteoarthritis-related biomarkers in..., Piscoya [/bib_ref] [bib_ref] Ground reaction forces and plantar pressure distribution during occasional loaded gait, Castro [/bib_ref] [bib_ref] Plantar pressures and ground reaction forces during walking of individuals with unilateral..., Castro [/bib_ref]. The combination of spatial distribution, time-course, and peak values of GRF in selected parts of the feet provides detailed information about the features of the forces acting on the body during linear and curvilinear paths, and gives insight into the generation of the centripetal force when walking in circle. Preliminary results have been recently presented [bib_ref] Insight into the generation of the centripetal force, Schmid [/bib_ref] [bib_ref] Ground reaction forces during linear and curved walking trajectories, Godi [/bib_ref].
# Methods
## Subjects
Twenty-six healthy participants (16 women, 10 men), aged (mean ± standard error, SE) 25.1 ± 0.5 years, range 21-35, mean body weight 63.0 ± 0.8 kg, mean height 1.7 ± 0.2 m, were recruited. No subject had a history of neurological diseases. All were free from ankle or foot pathology or other impairments that could contribute to postural instability or movement dysfunction. Exclusion criteria were major trauma in the last six months or lower limb surgery at any time previously. The study was approved by the local Research Ethics Committee and all subjects gave their informed consent.
## Procedure
Subjects walked under three different conditions, in random order: linear walking (LIN) and curved walking clockwise (CW) and counter-clockwise (CCW), at selfselected speed. The circle path (1.2 m radius) was drawn with a tape stuck on the floor. Subjects executed two 20 m trials for each trajectory, making a total of 6 trials. Before data acquisition, each subject performed two short practice trials for each condition to familiarize themselves with the instrumentation and task. They were instructed to walk looking forward, head erect, and not to focus on the tape. Walking time was monitored using photocells. For the LIN trajectory, photocells were placed at the beginning and at the end of a 20 m pathway; for the curved trajectories, at the beginning of the first lap and at the end of the third lap, corresponding to an overall path of 20 m. For each subject, we collected 50 steps for each condition. The first and last two steps of each trial were excluded from data acquisition, because changes in spatial-temporal variables (albeit minimal [bib_ref] When and how does steady state gait movement induced from upright posture..., Breniere [/bib_ref] can occur at initiation and termination of gait [bib_ref] Measuring step kinematic variability on an instrumented treadmill: how many steps are..., Owings [/bib_ref] [bib_ref] Reliability of measurements of pressures applied on the foot during walking by..., Randolph [/bib_ref]. The entire session lasted about one hour.
## Data collection and treatment
Walking speed was determined from the time taken to travel a 20 m pathway for both linear and curved trajectories. Stance duration was the time interval between the heel strike and toe off of the same leg. To allow for more ready comparison between straight line walking and turning, the percentage of the total gait-cycle duration (%GC) was computed as the time-interval between two successive heel strikes of one leg. In all subjects, both feet were instrumented with insoles. Subjects wore no socks. Insoles (Pedar-X system, Novel, Germany) and shoes (Superga 2750 model, Italy) corresponding to the individual's foot size were chosen. Insoles were placed inside the shoes and connected to the Pedar box. At the beginning of the session, the insoles were calibrated using the proprietary calibration device according to the manufacturer's manual. Data were sampled at 50 Hz. The Pedar-X system used in this study has been previously shown to have good reliability for both linear and curved trajectories [bib_ref] Test-retest reliability of an insole plantar pressure system to assess gait along..., Godi [/bib_ref].
The system produces, for each sample, a force value measured as the sum of the forces registered by the active sensors from each insole: this force value was used as the vertical component of the ground reaction forces (vGRF). The time course of the vGRF was stored for each trial and subsequently analysed using proprietary software.
The insole was also divided into eight anatomical foot regions: medial and lateral heel, medial and lateral arch, I metatarsal head, II-V metatarsal heads, hallux, and lateral toes. This was done in order to describe the different contribution of each region to the total vGRF averaged across the 50 steps [bib_ref] Repeatability of the Pedar-X in-shoe pressure measuring system, Ramanathan [/bib_ref]. vGRF was normalized to body weight (%BW = vGRF /BW*100) to reduce inter-individual variability [bib_ref] Normalization of ground reaction forces, Mullineaux [/bib_ref] [bib_ref] Normalization of ground reaction forces, joint moments, and free moments in human..., Wannop [/bib_ref].
We also measured trunk inclination during linear and curved trajectories by means of a tri-axial accelerometer (MicroStrain G-LINK Wireless Accelerometer System, ± 2 g range). This was placed in a pocket fastened by an elastic belt to the sternum, in the midline and midway along its length. Data were transferred wirelessly to a PC for off-line analysis. Acceleration data were sampled at 32 Hz; inclination was estimated using a low-pass FIR filter (finite impulse response filter, cut-off frequency = 0.3 Hz, 50th order/51-tap) in order to obtain the gravitational component of the medio-lateral acceleration, followed by trigonometric transformation [bib_ref] Measuring daily behavior using ambulatory accelerometry: the activity monitor, Bussmann [/bib_ref].
# Statistical analysis
We used parametric statistics, since the variables to be compared (vGRF, stance duration, walking speed, trunk inclination) followed a normal distribution (p > 0.05, Shapiro-Wilk's Test) and variances were homogeneous (Levene's Test, p > 0.05 for all variables). One-way analysis of variance (ANOVA) assessed differences in walking speed under the three walking conditions (LIN, CW, CCW). Two-way ANOVA was performed for vGRF and stance duration of the left and right foot during the three walking conditions. During curved trajectories, meaningful functional comparisons between feet were made according to the position of the foot with respect to the trajectory, i.e. foot on the inside ('Foot-In') or on the outside ('Foot-Out'). Values of vGRF from each foot anatomical region were compared in the three conditions (LIN, Foot-In, Foot-Out) with one-way ANOVA. When ANOVA gave a significant result (p < 0.05), posthoc analysis was performed with Tukey's test. CW and CCW have been examined as different conditions. Since the major gait determinants such as speed and cadence were not different between curved trajectories (see [fig_ref] Table 1: Gait variables during different walking trajectories [/fig_ref] , the plantar vGRF of the feet were collapsed into a Foot-In and a Foot-Out prior to statistical analysis. Data are presented as means ± SE in the figures and the text. The software STATISTICA (StatSoft, version 12.0) was used.
# Results
## Gait variables during linear and curved trajectories
As shown in [fig_ref] Table 1: Gait variables during different walking trajectories [/fig_ref] , walking speed and cadence were significantly affected by the type of trajectory, both being larger during LIN than curved. On the other hand, during curved-walking, speed and cadence were not affected by the direction of the trajectory.
During LIN, stance duration was similar in the left and right foot [fig_ref] Table 1: Gait variables during different walking trajectories [/fig_ref]. During curved walking, stance duration of the Foot-Out was shorter than that of LIN foot and Foot-In, while there was no difference between LIN foot and Foot-In. The changes observed in the above gait variable according to the type of trajectory confirmed previous findings in young [bib_ref] Test-retest reliability of an insole plantar pressure system to assess gait along..., Godi [/bib_ref] and elderly subjects as well as patients [bib_ref] Walking along circular trajectories in Parkinson's disease, Guglielmetti [/bib_ref] [bib_ref] Curved walking in hemiparetic patients, Godi [/bib_ref].
Ground reaction force during the stance phase [fig_ref] Figure 1: Vertical component of the ground reaction force [/fig_ref] shows the vGRF-time profiles during the stance phase in the left foot of a representative subject during LIN and during CW and CCW walking, when the foot was Foot-Out or Foot-In. The profiles are characterized by two peaks separated by a trough: the first peak broadly corresponds to the heel strike and weight acceptance phase (henceforth expressed as heel strike, HS), the trough to the foot mid-stance (MS), and the second peak to terminal stance and toe off (TO). LIN and curved trajectories differently affected the step time profile of vGRF: while this was superimposable for both feet during LIN (not shown), compared to LIN the vGRF profiles of both Foot-In and Foot-Out showed a decrease of the first (HS) and last peak (TO) and an increase of the trough at MS. Further, the trough was relatively larger in Foot-In than Foot-Out [fig_ref] Figure 1: Vertical component of the ground reaction force [/fig_ref]. At HS, vGRF was 129.6 %BW ± 2.7, 113.6% ± 2.5 and 119.6% ± 2.8 for LIN foot, Foot-In and Foot-Out, respectively (F(2,76) = 9.36; p < 0.0005) [fig_ref] Figure 2: Average [/fig_ref]. Post-hoc analysis showed that vGRF of both Foot-In and Foot-Out were lower than that of LIN foot (respectively, p < 0.0005 and p < 0.05). In turn, vGRF was slightly smaller (not significantly so) for Foot-In than Foot-Out (p = 0.25). At MS, vGRF was 62.8% BW ± 2.3, 86.4 ± 2.3 and 80.0 ± 3.1 in LIN foot, Foot-In and Foot-Out, respectively (F(2,76) = 32.0; p < 0.0005) [fig_ref] Figure 2: Average [/fig_ref]. Both Foot-In and Foot-Out had greater vGRF values than LIN foot (post-hoc, p < 0.0001). vGRF of Foot-In was marginally higher than that of Foot-Out (p = 0.07). At TO, vGRF was 136.5%BW ± 2.7, 121.3 ± 2.7 and 127.9 ± 2.8 in LIN foot, Foot-In and Foot-Out, respectively (F(2,76) = 7.70; p < 0.005) [fig_ref] Figure 2: Average [/fig_ref]. Post-hoc analysis showed that Foot-In had smaller vGRF values than LIN foot (p < 0.005). vGRF was marginally smaller in Foot-Out than LIN foot (p = 0.08). [fig_ref] Table 2: Summary of the changes in vertical component of the ground reaction forces [/fig_ref]. Each region is coloured according to the values of the mean vGRF. The colour code is reported in the calibration bar at the bottom of the figure.
## Ground reaction force distribution in space
For LIN, the values of vGRF in each of the eight anatomical regions were distributed equally in both feet [fig_ref] Figure 3: shows the vGRF [/fig_ref]. Accordingly, the values shown in [fig_ref] Table 2: Summary of the changes in vertical component of the ground reaction forces [/fig_ref] for each anatomical region are the averages for both feet. On the whole, the lateral part of the foot was loaded to a larger extent than the medial part, except for the hallux. This basic pattern was maintained in the three trajectories. However, modest (maximum 10%BW) but significant changes in vGRF in the anatomical regions of the foot sole were present between LIN and curved walking (see [fig_ref] Table 2: Summary of the changes in vertical component of the ground reaction forces [/fig_ref] and [fig_ref] Figure 3: shows the vGRF [/fig_ref]. For curved trajectories, Foot-In and Foot-Out were differently loaded with respect to LIN foot. Firstly, at heel strike, the lateral part of the Foot-In heel was unloaded with respect to that of LIN foot and Foot-Out. On the contrary, the medial part of the heel of Foot-Out was unloaded with respect to that of LIN foot and Foot-In. Regarding the medial and lateral arch, changes between the trajectories were negligible. Changes were instead again evident in the forefoot. At the level of the metatarsal heads, the lateral part of Foot-Out was unloaded with respect to that of LIN foot and Foot-In, while the medial part of Foot-In was unloaded with respect to Foot-Out. Strikingly different behaviour between the lateral and medial part of the foot was observed at the hallux and lateral toes. The hallux showed a clear-cut unloading in Foot-In with respect to LIN foot and Foot-Out, while the changes in the outer toes were smaller.
In order to shortly describe the changes in vGRF distribution in the four medial and lateral anatomical regions of the foot during LIN and curved trajectories, we calculated an asymmetry index (AI) for each of the regions of LIN foot, Foot-In and Foot-Out, as:
Medial part of Foot-Lateral part of Foot Medial part of Foot þ Lateral part of Foot  100 [fig_ref] Figure 4: Asymmetry index [/fig_ref] shows that AI was negative for both linear and curved trajectories in the cases of the heel, arch and metatarsal heads, in keeping with the larger vGRF on the lateral than medial part of these foot regions. AI of the forefoot at the toes was positive, again regardless of the trajectory, indicating a larger vGRF on the hallux than lateral toes. Turning trajectory significantly modulated AI of the heel and the metatarsal heads, not so AI of the arch and toes. During LIN trajectory, AI of both heel and metatarsal heads showed an intermediate value between those for CW and CCW. One-way ANOVA showed a significant effect of feet (LIN foot, Foot-In, Foot-Out) on the heel (F(2,96) = 26.31; p < 0.0001), metatarsal heads (F(2,96) = 3.57; p < 0.05), but not on the arches (F(2,96) = 0.75; p = 0.47) and toes (F(2,96) = 0.37; p = 0.69). The post-hoc test showed that, during curved walking, AI of the Foot-Out heel became more negative than that of LIN foot (p < 0.0005) or Foot-In (p < 0.0005). This was in keeping with an increasing greater loading of the lateral part of the heel for Foot-Out with respect to Foot-In and LIN foot. On the contrary, AI became less negative for Foot-In with respect to LIN foot (p < 0.01), sign of a decreasing lateral loading of the heel. At the metatarsal heads, AI acted in the opposite way: for Foot-Out, it became slightly less negative (p = 0.62) with respect to LIN foot while, for Foot-In, it became more negative than LIN foot (p = 0.21) and Foot-Out (p < 0.05), in keeping with an increasing larger loading of the lateral part of the metatarsal heads of Foot-In with respect to Foot-Out and LIN foot.
## Trunk inclination during linear and curved walking
Trunk inclination in the frontal plane was dependent on the trajectory . During LIN walking, inclination ranged between −1 and +0.9 deg. For curved paths, the trunk was inclined toward the interior of the trajectory: for CCW, it was inclined to the left (negative values), ranging across subjects between −7.6 and −2.1 deg, while for CW it was inclined to the right (positive values), ranging between +0.5 and +7.6 deg. On average, trunk inclination was −0.3 deg ± 0.1, −4.6 deg ± 0.4, and +4.4 deg ± 0.5 during LIN, CCW and CW, respectively. ANOVA showed a significant effect of the trajectories on inclination (F(2,30) = 162.3; p < 0.0001). Inclination values differed from each other for the three trajectories (post-hoc, p < 0.0001). shows that, across subjects, even moderate changes in walking speed had an effect on trunk inclination, i.e. the faster the gait speed (normalized to height), the more inclined the trunk, both for CW (y = 12.63× -5.28, p < 0.01, r 2 = 0.56) and for CCW (y = − 7.03× + 1.2, p < 0.01, r 2 = 0.36). Clearly, no relationship was found between speed and trunk inclination during LIN trajectories (y = −0.18 -0.12×, p = 0.90, r 2 = 0.001). As expected, the asymmetry index (AI) of the vGRF values of each foot exhibited a clear-cut dependence on trunk inclination during the CW and CCW trajectories. This dependence was opposite for the back (heel) and for the front part (metatarsal heads) of both feet. In , the mean trunk inclination (right, positive values) is plotted against the mean AI (abscissa), for the LIN, CW and CCW trajectories. Briefly, during CW, the heel of the right foot (Foot-In, panel D) was loaded relatively more in its medial part, while during CCW, the heel of the same foot (now Foot-Out) was loaded more in its lateral part. A mirror pattern was observed for the front part of the right foot (panel B). In this case, during CW the metatarsal heads of the right foot (Foot-In) were loaded more in their lateral part, while during CCW the metatarsal heads of the right foot (now Foot-Out) were loaded more in their medial part (right bottom panel). Panels A and C show similar but opposite patterns for the left foot.
# Discussion
Several studies have shown that there is a moderate reduction in walking speed along a curved path with respect to walking in a straight line [bib_ref] Human walking along a curved path. I. Body trajectory, segment orientation and..., Courtine [/bib_ref] [bib_ref] Walking along circular trajectories in Parkinson's disease, Guglielmetti [/bib_ref] [bib_ref] Test-retest reliability of an insole plantar pressure system to assess gait along..., Godi [/bib_ref] [bib_ref] Interaction of the body, head, and eyes during walking and turning, Imai [/bib_ref] [bib_ref] Relationship between velocity and curvature of a human locomotor trajectory, Vieilledent [/bib_ref] [bib_ref] Realistic human walking paths, Brogan [/bib_ref] [bib_ref] How do humans turn? Head and body movements for the steering of..., Hicheur [/bib_ref] [bib_ref] Velocity and curvature in human locomotion along complex curved paths: a comparison..., Hicheur [/bib_ref] [bib_ref] The formation of trajectories during goal-oriented locomotion in humans I. A stereotyped..., Hicheur [/bib_ref] [bib_ref] Velocity/curvature relations along a single turn in human locomotion, Olivier [/bib_ref] [bib_ref] Comparison of reliability, validity, and responsiveness of the mini-BESTest and Berg Balance..., Godi [/bib_ref] , dependent on the angle of turn [bib_ref] Walking along curved paths of different angles: the relationship between head and..., Sreenivasa [/bib_ref]. This reduction may be due to biomechanical constraints, to the necessary tuning of the locomotor command directed to the limbs [bib_ref] Tuning of a basic coordination pattern constructs straight-ahead and curved walking in..., Courtine [/bib_ref] , and/or to the need to stabilize the head or the lower limb joints [bib_ref] Interaction of the body, head, and eyes during walking and turning, Imai [/bib_ref] [bib_ref] Local dynamic stability in turning and straight-line gait, Segal [/bib_ref]. Curved walking also involves different neural processes compared to straight-path walking. Cognitive flexibility and set-shifting processes uniquely contribute to how individuals navigate curved paths [bib_ref] Contributions of cognitive function to straight-and curved-path walking in older adults, Lowry [/bib_ref] [bib_ref] A comparison of straight-and curved-path walking tests among mobility-limited older adults, Odonkor [/bib_ref].
The present findings confirm the slight decrease in velocity during curved compared to linear walking in young healthy subjects [bib_ref] Test-retest reliability of an insole plantar pressure system to assess gait along..., Godi [/bib_ref] ; at the same time, they increase our insight into the mechanisms underlying curved walking, showing an interaction in the pattern of the vertical component of the ground reaction force (vGRF) within and between the two feet during curved walking. While the insole output of the two feet is symmetric during linear walking, the pattern diverges during curved walking. Further, the modifications of the vGRF from the linear pattern are differently distributed in the various parts of the feet, depending on the position of the foot with respect to the orientation of the curved path and on the time-course of the stance phase.
In this study, both CW and CCW trajectories have been performed and have been separately examined as different conditions, instead of collapsing CW and CCW walking into one 'turning' condition. This was made in order to exclude that a 'preferred' leg would have produced different results when matched to a 'preferred' direction, with potential differences in walking velocity for CW and CCW turning, since leg preference might contribute to balance asymmetries and be associated with different foot placements [bib_ref] Measures of frontal plane stability during treadmill and overground walking, Rosenblatt [/bib_ref] [bib_ref] Leg preference associated with protective stepping responses in older adults, Young [/bib_ref].
## Profiles of the ground reaction force during the stance phase
The vGRF time profiles during the stance phase of gait are characterized by two peaks [bib_ref] Assessment of balance control in humans, Winter [/bib_ref] : the first peak corresponds to the heel strike and weight-acceptance phase, the second peak to the last part of the stance until toe off, when the vGRF becomes zero. With respect to the straight path, the force profiles of both feet during curved walking show a decrease of the first and last peaks and an increase of force during the trough corresponding to mid-stance. Further, when compared between feet during curved walking, the peaks are slightly but significantly lower, and the trough slightly but significantly higher, in Foot-In than Foot-Out (see [fig_ref] Figure 1: Vertical component of the ground reaction force [/fig_ref]. Therefore, the outer foot would play a predominant role in steering the body, as shown by the higher value of the forces during both the first and second peak. During mid- stance, however, the higher force at the inner foot is likely due to the body pivoting on it during turning. Such a pattern of vertical forces during stance had been observed previously, as well as its dependence on curvature and velocity [bib_ref] Insight into the generation of the centripetal force, Schmid [/bib_ref] [bib_ref] Contributions of the inside and outside leg to maintenance of curvilinear motion..., Smith [/bib_ref]. The smaller amplitude of the first and last peaks (both feet) in curved with respect to linear walking may be in part related to the fact that the insole records only the vertical component of the force. This depends on the distance between the instantaneous positions of the center of mass (CoM) of the body and of the vGRF of the stance foot. In this study, we did not compute the position of the CoM of the body. However, its distance from the vGRF must have increased during curved with respect to straight walking, as clearly shown by the inclinometer measurements and by the strong relationship between trunk inclination angle and velocity of curved walking (see [bib_ref] Human walking along a curved path. I. Body trajectory, segment orientation and..., Courtine [/bib_ref]. Therefore, the diminished values of vGRF must reflect the new vertical force equal to F = F * cos inclination-angle. The inter-peak trough during curved trajectories was however increased with respect to linear walking. This must depend on the overall different distributions of the vGRF during the stance phase in curved with respect to linear trajectories. The increase of the trough would be even larger were it not for the effect of the inclination. This increase must be related to the production of the curved trajectory, whereby the body pivots in the horizontal plane on the foot arch during the yaw rotation, however briefly, for the production of a small angular deviation for each step [bib_ref] Human walking along a curved path. I. Body trajectory, segment orientation and..., Courtine [/bib_ref] , rather than rolling onto the forefoot along the parasagittal plane. Of note, the stance duration is normally slightly longer for the inner than the outer foot (see [fig_ref] Table 1: Gait variables during different walking trajectories [/fig_ref] [bib_ref] Human walking along a curved path. II. Gait features and EMG patterns, Courtine [/bib_ref]. The increased load of the mid-foot during mid-stance is common to both the inner and outer foot, so that each foot (albeit more so the inner foot) can contribute a compliant support for the body weight [bib_ref] Foot anatomy specialization for postural sensation and control, Wright [/bib_ref] step after step, as one walks a curved path.
## Trunk inclination
The generation of the centripetal force at foot level can be effective only if trunk control is adequate. This must requires a delicate coordination pattern of muscle activities along the body. That this is so has been shown some time ago by Courtine et al. [bib_ref] Coordinated modulation of locomotor muscle synergies constructs straight-ahead and curvilinear walking in..., Courtine [/bib_ref] and Orendurff et al. [bib_ref] The kinematics and kinetics of turning: limb asymmetries associated with walking a..., Orendurff [/bib_ref] , who found that amplitude and timing characteristics of limb and trunk muscle activities were significantly correlated to the spatial and temporal gait adaptations associated with curvilinear locomotion. Ultimately, the accurate and appropriate position of the inner and outer foot should be the result of the coordinated spatial and temporal modulation of muscle activities of the entire kinematic chain. This is tuned as a function of both spatial and temporal features of gait [bib_ref] Coordinated modulation of locomotor muscle synergies constructs straight-ahead and curvilinear walking in..., Courtine [/bib_ref] and assures the critical control of upper body stability [bib_ref] Contribution of sensorimotor integration to spinal stabilization in humans, Goodworth [/bib_ref]. Some authors already used these vertical force variables as a measure of balance during walking [bib_ref] A comparison of straight-and curved-path walking tests among mobility-limited older adults, Odonkor [/bib_ref] [bib_ref] Posterior-anterior body weight shift during stance period studied by measuring sole-floor reaction..., Kobayashi [/bib_ref]. In the present study, a relationship between trunk inclination and speed normalized to height was clearly present. During curved as opposed to linear walking, the body mid-point comes closer to the inner and more distant from the outer foot during the respective stance phases. The tighter the curve, the more the body midpoint shifts towards the inner foot during its stance phase until the body mid-point bypasses the inner foot toward the centre of curvature [bib_ref] Human walking along a curved path. II. Gait features and EMG patterns, Courtine [/bib_ref] [bib_ref] Insight into the generation of the centripetal force, Schmid [/bib_ref].
In addition to the variability across subjects of body kinematics and asymmetry indexes, other spatial features of gait may play a role in the relatively large scatter of the relationship between trunk inclination and gait speed shown in . These features have not been directly measured in this study. However, they might include the yaw angle of the feet and the step width [bib_ref] Human walking along a curved path. I. Body trajectory, segment orientation and..., Courtine [/bib_ref] , which may not be exactly the same in all subjects.
Spatial distribution of the ground reaction force underneath the foot sole
The subdivision of the foot print into eight regions (four for each longitudinal half ) allowed to further explore the mechanisms subserving the controlled production of centripetal force during curved walking. For simplicity, we discuss here in detail only the most salient features of vGRF during the load-acceptance phase and the late stance phase (see [fig_ref] Figure 7: Description through the asymmetry index [/fig_ref]. Undoubtedly, the measurement of vGRF per se under the feet cannot provide the generation of centripetal force, which is a horizontal force that is not measured by the insoles. However, the instantaneous position of the point of application of this vertical force during turning, and its displacement with respect to what happens during linear walking, can give insight in the production of the gap between the center of mass and this application point. This distance created between center of mass and center of pressure in the medio-lateral plane produces a disequilibrium torque driven by gravity that is responsible for accelerating the body toward the center of the trajectory, very much as a similar torque during linear walking produces a disequilibrium in the sagittal plane that accelerates the body forward [bib_ref] The role of anticipatory postural adjustments and gravity in gait initiation, Lepers [/bib_ref] [bib_ref] The functional role of the triceps surae muscle during human locomotion, Honeine [/bib_ref].
At heel strike, compared to linear walking, CCW is associated with displacement of the point of application of the vGRF toward the lateral part of the heel (as indicated by the increased AI) of the right (outer) foot. This outer displacement produces an increased medio-lateral distance between the center of mass of the body (displaced toward the interior of the trajectory) and the application point, therefore a larger torque acting in the frontal plane and pushing the center of mass toward the inner part of the trajectory, favoring centripetal acceleration [bib_ref] Stance width influences frontal plane balance responses to centripetal accelerations, Goodworth [/bib_ref]. Then, once the body rolls over the foot and the vGRF moves to the metatarsal heads, the force exhibits a moderate displacement toward the medial part of the forefoot. This relatively reduces the torque acting on the body directed to the inner part of the trajectory. Overall, it seems that a push-pull mechanism operates on the body mass: when touching the ground, the outer foot helps the body 'fall' toward the interior of the trajectory, while in the late stance phase its front part brakes any further inclination of the body to the interior of the trajectory by having its medial (inner) part put pressure on the ground.
Under the same CCW condition, compared to linear walking, the vGRF of the left (inner) foot at foot contact is shifted toward the medial part of the heel. This increases the net torque acting on the body mass directed to the interior of the trajectory. The body fall to the interior is therefore being favored by the inner foot during the weight acceptance phase. On the other hand, the small displacement of the point of application of the vGRF at the lateral metatarsal heads during the evolution of the stance phase of the inner foot (AI is increased) brakes any further fall toward the interior of the trajectory, thereby favoring the movement of center of mass toward the exterior of the trajectory.
The described features are closely mirrored in CW walking . Overall, it seems that a double pushpull mechanism operates on the body mass. At heel contact and weight acceptance, both feet create a torque pushing the body to the interior of the trajectory. On the contrary, both forefeet have an opposing action, or an action braking the inward fall, by increasing the torque directed to the outer part of the trajectory. During curved walking, the late stance phase has a braking action, promptly counteracted by the heel contact of the opposite foot when the double stance phase supervenes. The double-stance phase reinstates the appropriate inward fall due to the shift of the vGRF point of application to the lateral part of the outer foot heel when it touches the ground.
Thus the feet behave functionally in the same way, regardless of their being the inner or outer foot. This is possible because of the asymmetry of the path of the vGRF below the feet: from medial (at heel strike) to lateral (at toe off) for the inner foot, and from lateral (at heel strike) to medial (at toe off) for the outer foot. Obviously, the net effect of this push-pull pattern cannot be but an inward thrust, as witnessed by the trunk inclination toward the interior of the trajectory, which counteracts the centrifugal force during steady-state turning.
# Limitations
Precise calculation of the torques presupposes a complete quantitative description of the generation of the centripetal force. We did not record the yaw placement of the feet (inner and outer) with respect to the direction walked. Foot distance from the trajectory can have major impact on the vGRF distribution and the direction of the ensuing torques [bib_ref] Effect of stance width on multidirectional postural responses, Henry [/bib_ref]. For instance, patients affected by different ailments may exhibit different foot placements from normal subjects [bib_ref] Curved walking in hemiparetic patients, Godi [/bib_ref] [bib_ref] Afferent control of walking: are there distinct deficits associated to loss of..., Nardone [/bib_ref] [bib_ref] Mediolateral foot placement ability during ambulation in individuals with chronic post-stroke hemiplegia, Zissimopoulos [/bib_ref] and this can affect their capacity to smoothly walk a circular path. Even in our normal subjects here, inconsistent foot placement may have been the source of the observed non-negligible variability in the recorded variables [bib_ref] Stance width changes how sensory feedback is used for multi-segmental balance control, Goodworth [/bib_ref]. Another limitation is that we did not record the body segment kinematics or compute the body's center of mass. Measuring the medio-lateral trunk inclination can provide easy, comprehensive and meaningful information about gait pattern during curved paths. However, from the mere measure of body inclination, one can have only indirect information on the biomechanics of curved walking. Data on foot placement and center of mass position together would have allowed a precise estimation of the torques accompanying circular walking.
Further, by necessity, plantar insoles, as compared to force platforms, give information on the vertical component of the force produced by the walking body, and cannot reveal the size of the shear forces [bib_ref] Plantar pressure assessment, Orlin [/bib_ref]. Devices able to yield continuous measures of plantar shear-forces during walking are not readily available. Shear forces inevitably have a non-negligible value [bib_ref] Temporal characteristics of plantar shear distribution: relevance to diabetic patients, Yavuz [/bib_ref] , the more so during curved walking because of the trunk inclination accompanying this locomotor task, and do certainly play a role in the production of mechanical effects in the frontal plane [bib_ref] The kinematics and kinetics of turning: limb asymmetries associated with walking a..., Orendurff [/bib_ref] [bib_ref] Linear and angular control of circular walking in healthy older adults and..., Goodworth [/bib_ref].
Therefore, the present results can give insight into the way foot placement exploits gravity in order to produce the body inclination during curved walking, by creating the properly oriented medio-lateral torques. The effect of the shear forces in favouring or braking the inward fall cannot be addressed here. However, we would consider that shear forces are more the consequence than the cause of the body inclination, and may not be a continuously controlled variable in the production of curved walking on solid, non-slippery ground.
# Conclusions
The present findings provide comprehensive and meaningful information about the pattern of GRF in curved walking in healthy subjects and give hints about the role of trunk inclination in accomplishing this task. In spite of the limitations mentioned above, the combined use of plantar insoles and accelerometer systems is an easy and relatively low-cost method. The advantage of the insoles is that there is no constraint on foot placement and that many consecutive strides can be recorded, while knowledge about the medio-lateral trunk inclination when walking along curved pathways may be useful in addressing problems in the control of equilibrium during turning.
The hypotheses put forward in this study should be tested in patients with problems in turning while walking. We predict that in these patients both distribution of plantar pressures as measured by insoles and of trunk inclination as measured by the accelerometer would be abnormal. We would also predict that the largest deviations from normal behaviour should be more evident during curved than linear trajectories, in which balance control is less critical [bib_ref] Contributions of cognitive function to straight-and curved-path walking in older adults, Lowry [/bib_ref].
[fig] Figure 3: shows the vGRF (%BW) values in each of the eight anatomical regions of the left and right foot during LIN walking (A) and of Foot-In and Foot-Out during curved walking (B), averaged across subjects [/fig]
[fig] Figure 1: Vertical component of the ground reaction force (vGRF) during linear and curved trajectories in the outer and inner foot. A. Profiles of the vertical component of the ground reaction force (vGRF) obtained during stance in the left foot of a representative subject during linear (LIN) walking and in the same foot when it was on the outside (Foot-Out) and inside (Foot-In) of the curved trajectory. B. Average of all subjects (+ standard error, SE) of the vGRF normalized to body weight (% BW) measured during the linear trajectory (LIN) as the average of right and left foot values and during the curved trajectory in the inner (Foot-In) and outer foot (Foot-Out). Values are obtained from the peak values at heel strike and toe off, and from the trough value at mid-stance. Average of 50 steps. [/fig]
[fig] Figure 2: Average (+ standard error, SE) of the vertical component of the ground reaction force (vGRF) normalized to body weight (% BW) measured during the linear trajectory (LIN) as the average of right and left foot values and during curved trajectory in the inner (Foot-In) and outer foot (Foot-Out). Values are obtained from the peak values of vGRF at heel strike (A) and toe off (C), and from the trough value at mid-stance (B). A. At Heel Strike, vGRF of both Foot-In and Foot-Out was lower than that of LIN foot. B. At Mid-Stance, both Foot-In and Foot-Out had greater vGRF values than LIN foot. C. At Toe Off, Foot-In had smaller vGRF values than LIN foot. *, p < 0.05; **, p < 0.005; ***, p < 0.0005. [/fig]
[fig] Figure 4: Asymmetry index (AI) of the estimated ground reaction force distribution at the heel, metatarsal heads, arches and toes of the foot during linear (LIN) and curved trajectories (Foot-In, Foot-Out [/fig]
[fig] Figure 6, Figure 5: Dependence of the asymmetry index (AI) of the vGRF distribution at the metatarsal heads of the left and right foot (respectively, A and B) and at the heel of the left and right foot (respectively, C and D) on trunk inclination during linear (LIN), counter-clockwise (CCW) and clockwise (CW) trajectories. In the abscissa, larger negative values (average ± standard error, SE) of AI represent an increase in vGRF on the lateral part of the relevant foot region. In the ordinate, positive or negative values (average ± SE) represent respectively inclination towards right or left side. See text for explanation. Trunk inclination during linear and curved trajectories and its relation to walking speed. A. Trunk inclination during the linear (LIN) and curved trajectories (counter-clockwise, CCW; clockwise, CW). The curved trajectories present a similar but opposite value. B. Relationship between walking speed normalized to height and trunk inclination across subjects. A linear relationship is present between inclination and speed but only during curved trajectories. *, p < 0.05; **, p < 0.005; ***, p < 0.0005. [/fig]
[fig] Figure 7: Description through the asymmetry index (AI), of the shifts of vGRF subserving the controlled production of the centripetal force during curved walking. Black points indicate the AI during counter-clockwise trajectories; the white points indicate AI during linear trajectories. Note in the Foot-Out the displacement from medial to lateral position of vGRF at the heel and from lateral to medial position of vGRF at the metatarsal heads with respect to linear walking. The reverse occurs in the Foot-In. [/fig]
[table] Table 1: Gait variables during different walking trajectories [/table]
[table] Table 2: Summary of the changes in vertical component of the ground reaction forces (vGRF) in the eight anatomical regions of the feet during linear and curved trajectories [/table]
|
Cytokine production by human leukocytes with different expressions of natural antiviral immunity and the effect of antibodies against interferons and TNF-±
Introduction: Two activities of innate antiviral immunity were studied: the resistance of human peripheral blood mononuclear cells (PMBCs) ex vivo to viral infection and the production of cytokines. Materials and Methods: Samples of blood were taken from healthy blood donors and from persons with frequent infections of the upper respiratory system. PMBCs were isolated by gradient centrifugation. Vesicular stomatitis virus (VSV) was used as the indicatory virus to infect PMBCs. The cytokines: IFN, TNF, and IL-6 were titrated by biological methods and IL-10 by ELISA.Results: Blood donors were divided for two groups: those with VSV-resistant and those with VSV-sensitive PMBCs and secretion of cytokines by them was compared. The resistant PMBCs produced more cytokines than the sensitive ones. A statistically significant difference, was found only in the case of the IFNs. To examine the contribution of IFNs and TNF in maintaining resistance, leukocytes from both groups were treated with specific anti-cytokine antibodies. The authors' previous study showed that the elimination of spontaneous IFN-α, IFN-β, IFN-γ, and TNF-α from resistant leukocytes resulted in increased VSV replication This indicates the important role of cytokines. In VSV-sensitive PMBCs, anti-IFN-α showed the opposite effect (decreased virus replication). In the absence of spontaneous IFN-α, disturbances in cytokine production were observed. Conclusions: Complete resistance of PMBC to VSV infection is accompanied by higher cytokine release, The paradoxical effect of anti-IFN-α on virus replication in leukocytes sensitive to viral infection may be attributed to changes in the cytokine profile balance, i.e. high TNF production by VSV-infected leukocytes and a complete reduction of IL-6 production.
# Introduction
Immune responses in vertebrates, invertebrates, and plants are initiated by the recognition of a large number of pathogens by cells of the innate immune system [bib_ref] Innate immunity: cells, receptors and signaling pathway, B³ach-Olszewska [/bib_ref] [bib_ref] Innate immunity, Miedzhitow [/bib_ref]. In vertebrates, systemic reactions develop when the innate immune system is effectively activated. Phagocytosis with intracellular killing and the production and activity of cytokines such as interferons (IFNs), members of the tumor necrosis factor (TNF) superfamily, interleukins (IL-12, IL-18), chemokines, and adhesion molecules are very important. Another activity of the innate immune system is MHC-independent killing of infected cells by natural killer cells and by complement activated along the alternative or lectin pathways [bib_ref] NK cell compartments and their activation by dendritic cells, Ferlazzo [/bib_ref] [bib_ref] The lectincomplement pathway -its role in innate immunity and evolution, Fujita [/bib_ref] [bib_ref] Mannan-binding lectin activates C3 and the alternative complement pathway without involvement of..., Selander [/bib_ref] [bib_ref] Roles of the alternative complement pathway and C1q during innate immunity to..., Yuste [/bib_ref]. The resistance of cells, especially leukocytes ex vivo, to viral infections is also considered to be one of the reactions of innate immunity [bib_ref] Innate immunity: cells, receptors and signaling pathway, B³ach-Olszewska [/bib_ref] which protect the host against invasion by microorganisms. As observed in many laboratories, high levels of cytokines with antiviral activity (IFNs and TNF) are frequently considered to be indicators of innate antiviral immunity. Our own results, obtained several years ago, showed different degrees of resistance to viral infection in organ cultures of human placenta, amniotic membrane, and umbilical cord vein [bib_ref] Antiviral nonspecific immunity. Possible role of endogenous tumor necrosis factor and interferons, Paradowska [/bib_ref] [bib_ref] Production of cytokines with antiviral activity by endothelial cells, Zaczyñska [/bib_ref]. Resistance was directed against viruses belonging to different taxonomic groups: vesicular stomatitis virus (VSV, Rhabdoviridae), encephalomyocarditis virus (EMCV, Picornaviridae), and herpesvirus type 1 (Herpesviridae). Furthermore, resistance to different viral infections was found in cultures of murine and human leukocytes [bib_ref] Individual differentiation of innate antiviral immunity in humans; the role of endogenous..., Orzechowska [/bib_ref] [bib_ref] Effect of cyclosporine A on the non-specific innate immunity of mice, Zaczyñska [/bib_ref]. Therefore resistance is non-specific. It can be reduced by specific antibodies against IFN-α, IFN-β, IFN-γ, and TNF-α, which indicates a role of endogenous cytokines. Resistance is highest immediately after organ culture and leukocyte isolation and is reduced after 1-5 days of in vitro incubation before infection with virus. The non-specific character of antiviral immunity, its dependence on endogenous IFNs and TNF, and the fact that it is present in cells ex vivo suggest that it could be considered innate immunity.
Using infection of peripheral blood leukocytes with VSV, we developed a new method for detecting innate immunity [bib_ref] Individual differentiation of innate antiviral immunity in humans; the role of endogenous..., Orzechowska [/bib_ref]. VSV was selected as an indicatory virus because it does not cause natural infections in the European human population, so specific antibodies against VSV are absent in human sera, which excludes the participation of acquired immunity in this experimental model of innate immunity. The virus also infects a wide spectrum of human and animal cells, including different types of immune cells. In our study, the whole peripheral blood mononuclear cell (PBMC) population was used, which mimics an in vivo natural system. Separated cell fractions are then needed for further studies. The results of preliminary experiments indicate that the innate immunity expressed by PMBCs is lower than that of the separate cell populations. These results suggest the presence of some regulatory mechanisms in PMBCs.
We used VSV replication in PBMCs to identify three different levels of natural immunity. A VSV replication titer higher than 4 log TCID 50 was considered a deficiency of innate immunity, a titer of 2-3 log indicated partial resistance, and a titer of 0-1 log indicated complete innate antiviral immunity. Using this method, we have studied the innate immunity of more than 300 people. Innate immunity was different among individuals. A lower level was often observed in persons with frequent infections of the upper respiratory track [bib_ref] Individual differentiation of innate antiviral immunity in humans; the role of endogenous..., Orzechowska [/bib_ref] and in those with frequent activation of herpes labialis. Deficiency of innate immunity is associated with the absence of remission and short survival time of patients with acute leukemia [bib_ref] Deficiency of innate immunity of leukocytes is associated with the failure of..., B³ach-Olszewska [/bib_ref]. In contrast, all patients with complete immunity achieved long-lasting complete remission after chemotherapy and were still alive after finishing the study. An age dependence for immunity was also shown by Rybka et al. [bib_ref] Age related natural antiviral non-specific immunity of human leukocytes, Rybka [/bib_ref]. Leukocytes from the peripheral blood of healthy donors (30-40 years old) were mostly resistant to viral infection. In contrast, leukocytes isolated from cord blood vein and from elderly individuals (>60) were vulnerable to viral infection. When innate immunity is over-stimulated, the reactions may lead to significant wastage in the body. Over-stimulation accompanies and participates in the development of autoimmunity or neurodegenerative diseases [bib_ref] Interferons as pathogenic effectors in autoimmunity, Baccala [/bib_ref] [bib_ref] Activation of innate immunity in the CNS triggers neurodegeneration through a Toll-like..., Lehnardt [/bib_ref] [bib_ref] Effects of cytokines on acetylcholine receptor expression: implications for myasthenia gravis, Poea-Guyon [/bib_ref] [bib_ref] Toll-like receptors, endogenous ligands, and systemic autoimmune disease, Rifkin [/bib_ref] [bib_ref] CD40 signaling regulates innate and adaptive activation of microglia in response to..., Townsend [/bib_ref]. Therefore, innate reactions must be under rigorous control.
We decided to compare two evaluation methods of innate immunity: a direct method based on PMBC resistance to VSV infection and an indirect method based on cytokine secretion. The production of IFNs and of TNF-α, IL-6, and IL-10 was studied. The role of every cytokine in innate immunity is different. The IFNs and TNF-α are the most important cytokines in antiviral immunity, IL-6, according to Jones [bib_ref] Directing transition from innate to acquired immunity: defining a role for IL-6, Jones [/bib_ref] , directs the transition from innate to acquired immunity, and IL-10 regulates immunity [bib_ref] The primary mechanism of the IL-10--regulated antiinflammatory response is to selectively inhibit..., Murray [/bib_ref].
# Materials and methods
## Blood donors samples
Peripheral venous blood samples (10 ml) were taken from a group of healthy volunteers (12 persons) and from currently healthy persons with recurrent infections of the upper respiratory tract (9 persons). The latter group was sent to a laboratory of virology to find the reason for their low immunity. The persons were 18-72 years old.
The study was performed with permission of the local ethics committee.
## Cell cultures
PMBCs were isolated from heparinized peripheral blood by gradient centrifugation in Dextran-Uropolin (Gradisol G) with a density of 1.115 g/ml (Aqua Medica, Poznañ, Poland). Five ml of blood were layered on 3 ml of Gradisol and centrifuged for 25 min at 400 g. Leukocytes from the border of the two phases were collected, washed two times with RPMI supplemented with 2% calf serum (CS), and suspended in the medium at a concentration of 2×10 6 cells/ml. The cell suspension was cultured at 37°C (in 5% CO 2 in air).
## L 929 cells (atcc ccl 1)
The murine fibroblast-like cell line was maintained in Eagle medium with 10% CS, antibiotics (100 U/ml penicillin and 100 µg/ml streptomycin), and 2 mM L-glutamine.
## A 549 cells (atcc ccl 185)
The human epithelial-like cell line was maintained in Dulbecco medium supplemented with 10% CS, antibiotics, and L-glutamine.
## Viruses
VSV, Indiana strain, Rhabdoviridae, Vesiculovirus, was multiplied and titrated in L 929 cells. EMCV, Columbia MM strain, Picornaviridae, Cardiovirus, was multiplied and titrated in L 929 cells and used for IFN titration.
## Estimation of resistance/innate immunity of pbls
Resistance/innate immunity was estimated by infection of leukocytes (2×10 6 cells/ml) with VSV. They were infected with a dose of 100 TCID 50 (tissue culture infectious dose, based on the cytopathic effect caused by the virus in about 50% of infected L 929 cells) of VSV/ml of the leukocytes suspensions. After 40 min of adsorption, the virus was washed out three times with 5 ml EMEM and the cells were suspended in 1 ml RPMI with 2% CS. A sample of infected cells was kept at a temperature of 4°C and served as a control of the starting level of the virus. The other cells were incubated at 37°C and samples of medium above the infected cells were collected each day and titrated in L 929 . The titer of virus is expressed in TCID 50 . Based on the replication of VSV in PBMCs, we identified three different levels of natural antiviral immunity: a VSV replication titer higher than 4 log TCID 50 was considered as deficiency of resistance/innate immunity, a titer of 2-3 log indicated partial resistance, and a titer of 0-1 log indicated complete resistance/innate antiviral immunity.
## Cytokine assay
IFN, TNF, and IL-6 were titrated using biological methods and IL-10 with ELISA.
IFN was titrated using the biological method of inhibition of the cytopathic effect caused by EMCV virus on the human cell line A 549 . In each experiment, calibration against an internal laboratory standard of human IFN-α/β was performed and the obtained results were corrected according to it. The internal standard was calibrated against the international standard of human leukocyte IFN 69/19. TNF was assayed according to Flick and Gifford [bib_ref] Comparison of in vitro cell cytotoxic assays for tumor necrosis factor, Flick [/bib_ref]. The cytotoxic activity of TNF-α in the samples was measured in L 929 cells in the presence of actinomycin D (Serva, Germany) at a concentration of 2 µg/ml EMEM. Recombinant human TNF-α (Genetech, USA) was used as the standard in all assays. In our assay, 1 unit was equal to 2.5-10 pg/ml of TNF-α.
Production of IL-6 was determined using the IL-6--dependent hybridoma cell line 7TD1 38 as previously described by Van Snick et al. [bib_ref] Purification and NH 2 -terminal amino acid sequence of a T--cell-derived lymphokine..., Van Snick [/bib_ref]. Briefly, the cells were cultured in 96-well plates (Costar, MA, USA) at a concentration of 2×10 3 cells/well in the presence of serial dilutions of the supernatants from uninfected and VSV--infected leukocyte cultures. After 72 h the surviving cells were evaluated colorimetrically by staining the cells with 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT, Sigma Chemicals Co., USA) and lysis of the cells with a sodium dodecyl sulfate (SDS)/ solution in a 45% aqueous solution (v/v) DMF (N, N-dimethylformamide, Sigma Chemicals Co., USA). One unit of IL-6 activity was calculated as the inverse dilution of a supernatant sample in which a half-maximal proliferation of 7TD1 cells was registered. In our assay, one unit correspond to 0.1 ng of recombinant (E. coli) human IL-6 (R&D Systems, USA). The sensitivity of the bioassay was 5 U/ml. IL-10 was quantified using ELISA according to the producer's instructions (BD Bioscientes Pharmingen, USA).
## Antibodies against human cytokines
Anti-IFN-α (polyclonal, sheep), anti-IFN-γ (polyclonal, goat), and anti-TNF-α (polyclonal, rabbit) antibodies were from Sigma-Aldrich (St. Louis, MO, USA) and anti-IFN-β (polyclonal, sheep) from the National Institutes of Health (Bethesda, MD, USA).
# Statistical analysis
The data were presented as means, standard deviations (SD), medians, quartiles (25-75%), and ranges (min.-max.). The significance of differences between groups was determined using the Student's t-test and the Mann-Whitney U-test. Differences between pairs of measurements in the same patients were tested using Wilcoxon's matched-pairs signed-ranks test. Spearman's rank correlation test was used for determining the relationship between the levels of IL-6 and the concentration of anti-IFN. The analysis was done using the computer program Statistica Version 5.1, Stat-Soft (USA), taking p<0.05 as statistically significant.
# Results
## Human leukocytes present different levels of antiviral immunity (cytokine production and resistance to viral infection)
In our study we used leukocytes isolated from the peripheral blood of 21 currently healthy persons with different innate immunity. The cultures of leukocytes were infected with VSV (100 TCID 50 /ml) immediately after preparation (T=0) and the kinetics of virus replication was studied during three days of infection. The results obtained are presented in [fig_ref] Figure 1: Kinetics of VSV replication in human PMBCs [/fig_ref]. The kinetics of VSV replication was different, so two groups of individuals were distinguished. The first consisted of 7 persons with resistant to viral infection leukocytes. They had very low or no viral replication (0-1 log TCID 50 ). The second group consisted of 14 persons (9 from the group with frequent infections and 6 healthy blood donors) with partial immunity or no resistance. VSV replicated to a titer of 2-5 log.
In the next experiments, spontaneous and VSV-induced cytokine production in both groups (resistant and sensitive PMBCs) were measured. We found that VSV infection only stimulated IFN production, did not significantly influence TNF, and reduced IL-6 and IL-10 levels.
To determine the correlation between the level of resistance/innate immunity measured by VSV replica-tion and the production of cytokines, we monitored the levels of IL-6, IL-10, TNF, and IFNs in the supernatants of two different leukocyte cultures, uninfected and after infection with VSV. Leukocytes with stronger resistance/innate immunity produced more cytokines; how-ever, statistical significance was found only in the case of VSV-induced IFN. [fig_ref] Figure 2: Comparison of spontaneous production of cytokines by VSV-resistant [/fig_ref] presents a comparison of spontaneous cytokine production by leukocytes with high and low levels of resistance/innate antiviral immunity.
## Elimination of ifn-α caused an unexpected reduction in vsv replication in sensitive leukocytes and a disturbance in cytokine production
In the next experiment the influence of specific anti--IFNs and anti-TNF-α antibodies on VSV replication and the production of cytokines by leukocytes with different antiviral immunity was examined. Before their application, the antibodies were carefully titrated for effective neutralization. The amounts were sufficient to neutralize cytokines produced by leukocytes. The antibodies were added to the leukocytes after VSV adsorption. The antibodies stimulated viral replication in the virus-resistant cells (results not shown). This effect indicates that endogenous IFNs and TNF-α participate in the maintenance of resistance/innate immunity. The influence of the antibodies on viral replication in leukocytes sensitive to VSV infection is presented in . Instead of the expected stimulation of virus replication after endogenous IFN-α elimination, a reduction in virus titer was observed. We attempted to explain these unexpected results by studying changes in cytokine profiles in the absence of endogenous IFNs and TNF-α. It was found that in the absence of endogenous IFN-α the profile of cytokine production was changed. The presence of anti-IFN-α antibodies abolished IL-6 production. The lack of IL-6 production was observed in both resistant and sensitive leukocytes, infected as well as uninfected. The results of IL-6 production by uninfected untreated leukocytes and those treated with anti--IFNs and anti-TNF-α antibodies are presented in . The other important observation was that in the presence of anti-IFN-α antibodies, an increased level of TNF-α, remarkably so in infected leukocytes from the VSV-sensitive group, was observed . To confirm the possible role of TNF in antiviral activity, VSV--infected leukocytes sensitive to the virus were treated with 100 U of TNF-α (the median TNF level in leukocytes treated with antibody against IFN-α). The results of four independent experiments, presented in , show a reduction in VSV titer by 2-4 log. The antibodies against IFN-α also stimulated IL-10 production depending on the level of innate immunity level [fig_ref] Figure 7: Effect of the elimination of endogenous IFNs and TNF-α on IL-10 production... [/fig_ref]. The addition of antibodies against TNF-α did not markedly influence IFN production (data not shown).
The results indicated that the elimination of endogenous IFN-α caused the largest changes in cytokine profiles. The changes involving TNF-α production were suspected to be responsible for the unexpected inhibition of virus replication by the antibodies. To prove the hypothesis, we tried to reduce the antiviral effect of the anti-IFN-α antibodies with addition anti-TNF-α. The results of the experiments were, however, unclear, suggesting that not only TNF, but other unknown mediators might participate in the antiviral effect.
# Discussion
In view of cytokines' antiviral activity, the question arises whether PMBCs (resistant or sensitive to VSV) expressing high, non-specific, antiviral resistance/innate immunity are able to produce more cytokines than those expressing low innate immunity. In this context we studied the production of IFN, TNF, IL-6, and IL-10. The results of this study show differences in cytokine production by cells with different degree of resistance/innate immunity. VSV-resistant PMBCs with good innate immunity released more spontaneous cytokines than PMBCs with immune deficiency. Statistical significance, however, was found only in the case of VSV--induced IFN. This means that even though the cells were not able to replicate VSV to a high level, they were able to produce more IFN after contact with VSV than the sensitive cells with high titers of VSV.
The results of a previous study on the influence of specific anti-cytokine antibodies on viral replication in human leukocytes revealed the importance of endogenous IFNs and TNF-α in maintaining antiviral innate immunity [bib_ref] Individual differentiation of innate antiviral immunity in humans; the role of endogenous..., Orzechowska [/bib_ref]. Specific antibodies against IFN-α, IFN--β, IFN-γ, and TNF-α caused reductions in innate immunity, mimicking the phenomenon observed after stimulation of viral replication in PMBCs previously resistant to VSV. Quite different results were obtained when antibodies, especially antibodies against IFN-α, were added to cells with very low innate immunity. Instead of VSV stimulation, reduction of VSV replication was observed. We tried to find the reason for these unexpected results. As first, a possible disturbance in cytokine production and replacement of antiviral function of IFN by other cytokines was considered. In fact, neutralization of IFN-α with specific antibodies caused a disturbance in leukocyte cytokine profiles. In the absence of endogenous IFN-α production, IL-6 was completely abolished. The other consequence of IFN-α elimination was stimulation of TNF-α. This higher TNF-α production by infected leukocytes sensitive to viral infection was assumed to be probably the reason for this paradoxical inhibition of VSV replication. Within the group deficient in innate immunity, TNF-α may show an inhibitory effect on viral replication. Antiviral activity of TNF-α has been reported by dif-ferent authors [bib_ref] TNF-alpha inhibits HIV-1 replication in peripheral blood monocytes and alveolar macrophages by..., Lane [/bib_ref] [bib_ref] Antiviral effect of recombinant tumour necrosis factor in vitro, Mestan [/bib_ref]. For example, Mestan et al. [bib_ref] Antiviral effect of recombinant tumour necrosis factor in vitro, Mestan [/bib_ref] showed that recombinant TNF-α added to the Hep-2 cell line protects the cells against VSV infection. This effect was specific for TNF because the antiviral activity was abolished by anti-TNF, but not by anti-IFN-α antibodies This observation is in agreement with our results obtained in VSV-sensitive PBLs. In our previous study, however, TNF-α's effect on VSV replication was dependent on the innate immunity of the PMBCs [bib_ref] Individual differentiation of innate antiviral immunity in humans; the role of endogenous..., Orzechowska [/bib_ref]. In sensitive cells, an inhibitory effect on viral replication was observed, while in resistant PMBCs a stimulation of VSV replication. In embryonal tissues sensitive to VSV, inhibition of viral replication was also observed [bib_ref] Effect of exogenous tumor necrosis factor, interleukin 6, and interferons on vesicular..., Paradowska [/bib_ref]. Inhibition of viral replication by TNF in sensitive leukocytes was shown in this study. In cultures resistant to infection, the opposite effect and stimulation of replication occurred [bib_ref] Antiviral nonspecific immunity. Possible role of endogenous tumor necrosis factor and interferons, Paradowska [/bib_ref]. Similarly, effects depending on the level of innate immunity were noticed following culture treatments with immunomodulating substances, i.e. lipopolysaccharide and peptides from ovine colostrums [bib_ref] Effect of proline rich polypeptide from ovine colostrum on virus replication in..., Domaraczenko [/bib_ref] [bib_ref] Study on risk factors for transplacental viral infections; effect of bacterial factors..., Jatczak [/bib_ref]. The virus stimulatory effect, however, was eliminated by antibodies against TNF-α. Our previous and the present study indicate a double role of endogenous TNF-α: it is responsible for both inhibition and stimulation of viral replication by immunomodulators. The lack of IL-6 production in the absence of IFN-α indicates the necessity for this type of IFN for IL-6 production by leukocytes. Over the years, IL-6 has been assigned both pro-and anti-inflammatory properties. According to Jones [bib_ref] Directing transition from innate to acquired immunity: defining a role for IL-6, Jones [/bib_ref] , an important role of IL-6 is directing the transition from innate to acquired immunity; IFN-α, through its effect on IL-6 production, could thus indirectly influence this transition.
[fig] Figure 1: Kinetics of VSV replication in human PMBCs (n=21). Leukocytes were infected with VSV immediately after their isolation, washed several times with culture medium, and incubated at 37 o C for 3 days. The kinetics of virus replication was studied in samples collected each day after infection. The samples were titrated in L 929 cells. [/fig]
[fig] Figure 2: Comparison of spontaneous production of cytokines by VSV-resistant (n=7) and VSV-sensitive (n=14) PMBCs. The samples were collected after 24 h of cell incubation from uninfected and VSV-infected PMBCs. In the figure, however, spontaneously released cytokines are presented. Only the IFN level is presented in both the uninfected and VSV-infected PMBCs. To inactivate the virus, the samples of VSV-infected cells were UV irradiated (on ice) before titration. IFN, TNF, and IL-6 were titrated by biologic methods and IL-10 by ELISA. Statistics were performed using the Student's t-test. *Statistically significant. [/fig]
[fig] Figure 3, Figure 4, Figure 5, Figure 6: Effect of anti-IFNs and anti-TNF-α antibodies on VSV replication in PMBCs sensitive to VSV infection. The antibodies (a=anti) were applied to VSV-infected PMBCs after virus adsorption. The kinetics of viral replication in antibody-treated cells was compared with untreated cells. Samples of media were collected over three days. VSV was titrated in each sample in L 929 cells. Results from the first day after VSV infection are presented. Statistics were performed using the Mann-Whitney test. *Statistically significant. Effect of antibodies against IFNs and TNF on IL-6 production by PBLs. Anti-cytokines antibodies (a=anti) were added immediately after virus adsorption. The level of IL-6 was determined in sensitive and resistant PMBCs. Statistics were performed with the non-parametric Wilcoxon test. *,**Statistically significant. Effect of anti-IFNs antibodies on TNF production in sensitive leukocytes infected with VSV. For description, seeFig 4.*,**Statistically significant. Effect of TNF-α on VSV replication in sensitive leukocytes (n=4). 100 U of TNF-α was added to the leukocytes just after VSV adsorption and the kinetics of virus replication was studied. [/fig]
[fig] Figure 7: Effect of the elimination of endogenous IFNs and TNF-α on IL-10 production in sensitive leukocytes infected with VSV. For description, see Fig 4. *, **Statistically significant. [/fig]
|
Endoscopic injection sclerotherapy for pediatric bleeding esophageal varices complicated by gastric vein, main portal vein, splenic mesenteric junction, and splenic vein occlusion: a case report
Background: Endoscopic injection sclerotherapy (EIS) is a life-saving procedure for pediatric patients with bleeding gastric varices (GV) associated with advanced liver cirrhosis and severe portal hypertension. Because of the lack of an endoscopic banding ligation device for pediatric patients, EIS is usually performed for bleeding esophageal varices (EV) in infants with congenital biliary atresia. Case presentation: We present a case of a 15-month-old female infant with type I biliary atresia with jaundice (total serum bilirubin, 22.2 mg/dL), hypoalbuminemia (serum albumin level, 2.58 g/dL), coagulopathy (prothrombin time > 20 s compared with that of a normal control), ascites, splenomegaly, portal hypertension (portal vein velocity, 3.9-5.6 cm/sec of hepatopetal flow), and repeated bleeding of the varices after receiving three doses of intravascularly administered Histoacryl 1 ampoule mixed with Lipiodol UF 8 mL in the EV. Prominent GV and EV were occluded by EIS. The sclerosing agent was also present in the main portal vein, splenic mesenteric junction, and splenic vein, causing an engorged inferior mesenteric vein. The patient underwent total hepatectomy and living donor liver transplantation (LDLT) by left lateral segment graft (segments 2, 3, and 4 of the middle hepatic vein trunk) and left portal vein graft to the recipient inferior mesenteric vein anastomosis. Portal vein stent placement via segment 4 of the portal vein stump was performed from the inferior mesenteric vein to the umbilical portion of the left portal vein. The patient is still alive and doing well after the LDLT. Conclusions: EIS is a life-saving procedure in cases involving bleeding EV complicated by gastric, main portal vein, splenic mesenteric junction, and splenic vein occlusions; hence, it should be kept in mind as a treatment for EV complications in pediatric patients.
# Background
Endoscopic hemostasis for bleeding esophageal varices (EV) is widely used along with endoscopic banding ligation. In contrast, endoscopic injection sclerotherapy (EIS) is the treatment of choice for bleeding gastric varices (GV) [bib_ref] Endoscopic injection sclerotherapy with 1.5% Sotradecol for bleeding cardiac varices, Chiu [/bib_ref] [bib_ref] Endoscopic cyanoacrylate injection with or without lauromacrogol for gastric varices: a randomized..., Zeng [/bib_ref]. Because the devices for banding ligation are only suitable for adult patients, EIS is also a life-saving endoscopic hemostatic treatment for pediatric patients with bleeding EV and/or GV. We present a case of a pediatric patient with a rare complication treated with EIS.
## Case presentation
A 15-month-old female Filipino infant with congenital type I biliary atresia and without any other anomalies or malformations, who had not undergone Kasai's surgical procedure for biliary atresia, was referred by a liver center in the Philippines. She weighed 8.1 kg and had a height of 67.3 cm. She had jaundice (total serum bilirubin, 22.2 mg/dL), hypoalbuminemia (serum albumin level, 2.58 g/dL), coagulopathy (prothrombin time > 20 s compared to that of a normal control), ascites, splenomegaly, portal hypertension (portal vein velocity, 3.9-5.6 cm/sec with hepatopetal flow measured by Doppler ultrasound), and repeated bleeding of the varices after three doses of intravascularly administered Histoacryl 1 ampoule mixed with Lipiodol UF 8 mL (Auckland, New Zealand) in the EV [fig_ref] Figure 1: Endoscopic picture of severe esophageal varices [/fig_ref]. A Doppler ultrasound was used to investigate the portal hemodynamics before EIS. The diameter of the portal vein was 6.1 mm with reversal hepatofugal flow in portal vein velocity. After the first EIS, the portal vein diameter was 4.4 mm without thrombosis. After the third EIS, the end point of EIS was further investigated, and computed tomography Abdominal computed tomography showing that the sclerosing agent was occluded in the esophageal varices, gastric varices, main portal vein, splenic mesenteric junction, and splenic veins (black arrows) and engorged the inferior mesenteric vein angiogram revealed that the intrahepatic portion of the portal vein was not clearly demonstrated. Prominent GV and EV were occluded by EIS [fig_ref] Figure 2: Plant X-ray film showing that the sclerosing agent was occluded in the... [/fig_ref]. The sclerosing agent was not only present in the EV and GV but also retrogradely occluded the main portal vein, splenic mesenteric junction, and splenic vein, causing an engorged inferior mesenteric vein [fig_ref] Figure 3: Abdominal computed tomography showing that the sclerosing agent was occluded in the... [/fig_ref]. The patient underwent total hepatectomy and living donor liver transplantation (LDLT) via a left lateral segment graft (segments 2, 3, and 4 of the middle hepatic vein trunk) and left portal vein graft for the recipient inferior mesenteric vein anastomosis. Portal vein stent placement via segment 4 of the portal vein stump was performed from the inferior mesenteric vein to the umbilical portion of the left portal vein [fig_ref] Figure 4: Portal vein stent placement via segment 4 of the portal vein stump... [/fig_ref]. The patient is still alive and doing well after the LDLT.
# Discussion and conclusions
Thrombosis in the portal, splenic, and mesenteric veins may develop after sclerotherapy in cirrhotic patients [bib_ref] Superior mesenteric venous thrombosis treated by direct aspiration thrombectomy, Nakayama [/bib_ref] [bib_ref] Recurrent gastric varices bleeding after living-related liver transplantation successfully eradicated by splenic..., Lu [/bib_ref]. This case is interesting because it involves an infant. The insertion tube outer diameter of adult and pediatric gastroscope is 9.0-11.4 and 5.9-6.0 mm, respectively. The cap of the ligation device can only hold tightly on a gastroscope with outer diameter > 10-11.4 mm; however, such a diameter is hard to insert to a 15-month infant patient with an esophageal orifice diameter < 10 mm. Therefore, ligation of varices was only performed in older pediatric or adult patients with an esophageal orifice diameter > 10 mm. EIS was frequently used to stop bleeding in infants or those with esophageal orifice diameter < 10 mm. In fact, EIS rarely induces gastric, main portal vein, splenic mesenteric junction, and splenic vein occlusion in pediatric patients. The dose of the intravascularly administered Histoacryl mixed with Lipiodol differs between adult and pediatric patients because the size of the vascular structures is smaller in the latter. According to the current presentation, half a dose of the sclerosing agent, i.e., Histoacryl 1/2 ampoule mixed with Lipiodol 4 mL, is suggested for the treatment of bleeding varices in pediatric cases. LDLT should be the treatment of choice not only for pediatric patients with biliary atresia but also for correcting complications arising from EIS with regards to splenic and superior mesenteric vein occlusion [bib_ref] P4 stump approach for intraoperative portal vein stenting in pediatric living donor..., Chen [/bib_ref]. In our case, the sclerosing agent occluded the splenic and portal veins and retrogradely down to the splenic mesenteric junction. We placed a portal vein stent to avoid thrombosis of the mesenteric vein in this patient who underwent LDLT. Because the portal vein flow was low (9.8 cm/sec) and the inferior mesenteric vein was small during portal vein anastomosis, we also worried about the insufficient blood inflow from the splenic and superior mesenteric veins to the portal vein and compromised liver graft. A portal vein stent was placed to increase the inferior mesenteric vein size and augment the portal vein flow to ensure sufficient blood flow to the portal system of the new liver graft [fig_ref] Figure 5: Schematic picture of the portal systems and the direction of blood flow... [/fig_ref]. The portal vein stent placement via segment 4 stump approach for intraoperative portal vein stenting in pediatric LDLT is an innovative technique for this kind of complication. Hence, we would like to share our experience and contribute to the medical field. In conclusion, endoscopic hemostasis with EIS is a life-saving procedure not only for bleeding EV but also for the treatment-associated major complications with gastric vein, main portal vein, splenic mesenteric junction, and splenic vein occlusions in pediatric patients with biliary atresia.
[fig] Figure 1: Endoscopic picture of severe esophageal varices [/fig]
[fig] Figure 2: Plant X-ray film showing that the sclerosing agent was occluded in the esophageal and gastric varices and extended into the splenic vein and part of the superior mesenteric veins [/fig]
[fig] Figure 3: Abdominal computed tomography showing that the sclerosing agent was occluded in the esophageal varices, gastric varices, main portal vein, splenic mesenteric junction, and splenic veins (black arrows) and engorged the inferior mesenteric vein [/fig]
[fig] Figure 4: Portal vein stent placement via segment 4 of the portal vein stump was performed from the inferior mesenteric vein to the umbilical portion of the left portal vein [/fig]
[fig] Figure 5: Schematic picture of the portal systems and the direction of blood flow for the portal vein stent with portal vein and inferior mesenteric vein anastomosis. The X means the insufficient blood inflow from the esophageal and gastric varices, splenic and superior mesenteric veins to the portal vein compromised the liver graft. The big arrow means a portal vein stent was placed to increase the inferior mesenteric vein size and augment the portal vein flow to ensure sufficient blood flow to the portal system of the new liver graft [/fig]
|
A rare presentation of appendicitis contained within an incisional hernia post loop ileostomy reversal – A case report
Introduction and importance: Appendicitis within an incisional hernia is rare, with current literature describing a small number of cases, occurring through a variety of surgical incisions. We describe a case of appendicitis contained within an incisional hernia following reversal of a loop ileostomy, on a background of previous sigmoid cancer resection. This is the second such case we were able to identify on literature review.Case presentation:A 45 year old man presented with one day of migratory abdominal pain, predominantly focused at a tender, irreducible lump in his right lower quadrant, underlying the scar from previous reversal of loop ileostomy. CT on admission revealed an incisional hernia, containing an inflamed appendiceal tip. He underwent an uncomplicated laparoscopic appendicectomy and primary suture closure of the hernia defect, and was discharged the following day. Acute appendicitis was confirmed on histopathology. Discussion: Placement of a defunctioning ileostomy is common in the management of colonic cancers, and incisional hernias are a common complication. It is however rare for an appendix to be contained within a hernia sac, and even rarer for appendicitis to develop in this setting. As a result, the presentation of this condition may mimic that of an incarcerated or strangulated incisional hernia, with pre-operative diagnosis typically relying on diagnostic imaging. Conclusion: Incisional hernia appendicitis is rare and presents a diagnostic challenge. Early recognition of this dual pathology is necessary to allow for prompt surgical management of both the appendicitis and hernia, as well as guiding the approach for hernia repair.
# Introduction
Acute appendicitis and abdominal wall hernias are both common surgical pathologies, but the presence of an inflamed appendix within a hernia sac is a rare occurrence [bib_ref] Amyand's hernia: a review, Ivashchuk [/bib_ref]. Clinically, it may be difficult to distinguish from a strangulated incisional hernia, however accurate and timely diagnosis is necessary to appropriately guide treatment, consisting of appendicectomy as well as repair of the hernia defect. This case reports describes a patient who presented with a hernia at the site of a previously reversed loop ileostomy, containing an inflamed appendixthis is the second case of appendicitis within an ileostomy-site hernia we were able to identify on review of current literature. He subsequently underwent laparoscopic appendicectomy and primary closure of his hernia defect. We discuss considerations in both diagnosis and management of this rare condition.
# Methods
This case report was prepared in accordance with the SCARE Statement consensus-based surgical case report guidelines [bib_ref] The SCARE 2020 guideline: updating consensus Surgical CAse REport (SCARE) guidelines, Agha [/bib_ref].
## Case
We present the case of a 45-year-old Caucasian male, who presented to the emergency department of our hospital in Melbourne, Australia. He reported a 24 h history of migratory abdominal pain, which started as a sudden onset generalised cramping, gradually localising to a firm, painful lump underlying a surgical scar in the right iliac fossa (RIF). At presentation, the pain was described as constant, with no aggravating or alleviating factors. He reported nausea, without vomiting. He denied any bowel or urinary symptoms, as well as subjective fevers or rigors.
The patient's past surgical history was significant for T3N0M0 sigmoid adenocarcinoma, for which he underwent laparoscopic high anterior resection with loop ileostomy four years prior. His ileostomy was reversed 12 months later. Subsequent surveillance CT imaging and endoscopy was unremarkable, with his most recent colonoscopy 12 months prior. He denied any other medical or surgical history, did not take any regular medications, and was a non-smoker. No significant family history was noted. Vital signs were normal, and he was afebrile at presentation. The patient was obese, with a BMI of 32.5 kg/m 2 . Examination of his abdomen revealed focal tenderness in the right lower quadrant, directly underlying his ileostomy reversal site scar. A firm, 3 cm diameter mass could be felt at this point, which was tender, irreducible and did not have a cough impulse.
A provisional diagnosis of acute appendicitis was made based on clinical presentation, with a differential diagnosis of incarcerated incisional hernia considered. The patient was kept nil by mouth, given intravenous crystalloid fluids and further investigations were arranged.
No abnormality was seen on initial blood tests, including the white cell count (WCC) and C-reactive protein (CRP). A non-contrast CT scan of his abdomen and pelvis revealed an incisional hernia containing fat and a mildly dilated appendiceal tip (7 mm diameter) with surrounding fat stranding [fig_ref] Figure 1: CT images of appendix entering hernia defect [/fig_ref].
Informed consent for surgery was obtained, and the patient was taken to the operating theatre. Surgery was performed by the on-call general surgeon of the day. Intraoperatively, an incisional hernia defect at the previous loop ileostomy site was seen [fig_ref] Figure 2: Laparoscopic view of appendix entering hernia defect, and previous nylon suture used... [/fig_ref]. The hernia sac was reduced, revealing an inflamed appendix tip without perforation. The appendix was mobilised from adjacent adhesions, and routine laparoscopic appendicectomy was performed. Continuous suture repair of the hernia defect was performed using Medtronic 0 PDS V-Loc barbed sutures . Post-operative recovery was uneventful, and the patient was discharged home the following day with instructions to avoid heavy lifting. No post-operative complications were noted on routine follow up, and the patient reported a satisfactory recovery. Histopathology subsequently confirmed acute appendicitis.
# Discussion
Placement of a defunctioning stoma is common practice in bowel cancer resection with primary anastomosis. This figure is highest in patients undergoing rectal cancer resection, 89 % of whom received a stoma in a review of practice in Australia and New Zealand from 2007 to 2019 [bib_ref] Current practice in Australia and New Zealand for defunctioning ileostomy after rectal..., Grupa [/bib_ref]. This practice reduces the risk of anastomotic leakage and resultant pelvic sepsis in low pelvic anastomoses. However, a recognised drawback of this approach is the risk of incisional hernia development at the stoma site following reversal, estimated to affect 6.1 % of patients [bib_ref] Systematic review and meta analysis of incisional hernia post reversal of ileostomy, De Haes [/bib_ref].
The presence of the appendix within a hernia was first described in 1735 by Claudius Amyand, and the term Amyand's hernia has been used to describe an incarcerated inguinal hernia containing appendix (whether normal or inflamed) [bib_ref] Amyand's hernia: a review, Ivashchuk [/bib_ref]. The incidence of this is rare, estimated at 0.19-1.7 % of inguinal hernia cases, with an even smaller subset containing an inflamed appendix, affecting only 0.07-0.13 % of Amyand's hernias [bib_ref] Amyand's hernia: a review, Ivashchuk [/bib_ref].
Reports of incisional hernias containing an inflamed appendix are rare, but have been described involving a variety of incisions, including Pfannenstiel [bib_ref] Incisional hernia appendicitis: a unique case report, Paudyal [/bib_ref] [bib_ref] Appendicitis found in an incisional hernia, Galiñanes [/bib_ref] , Kocher [bib_ref] Appendiceal hernia: an extremely rare condition, Molina [/bib_ref] , and laparoscopic port-site incisions [bib_ref] Incisional hernia appendicitis: a report of two unique cases and literature review, Sugrue [/bib_ref]. Our literature review identified only one other case of appendicitis within an incisional hernia at an ileostomy reversal site [bib_ref] Complicated appendicitis within an incisional hernia, Becker [/bib_ref].
Clinical presentation in such cases may differ from the classical presentation of appendicitis. Patients often present with a tender, irreducible mass at a previous incision site as their primary complaint, which may instead raise clinical suspicion of a strangulated incisional hernia. Given the rare nature of this pathology, it is likely not to be a leading differential diagnosis on initial presentation, and would likely be identified based on imaging or intraoperative findings. In the case of our patient, CT imaging provided a clear diagnosis, demonstrating an inflamed appendix contained within the incisional hernia sac.
Identification of an inflamed appendix either intraoperatively or on pre-operative imaging suggests the need for prompt appendicectomy followed by hernia repair. Due consideration must be given to the surgeon's chosen approach for repairing such a hernia. While the use of synthetic mesh is associated with a lower recurrence rate compared to suture repair, placement of a synthetic within a contaminated or infected wound would carry an increased risk of surgical site infections. Given the rarity of this condition, no literature exists to specifically evaluate the safety of mesh repair of a hernia containing an inflamed appendix. Earlier guidelines cautioned over the use of mesh in contaminated wounds [bib_ref] HIG (SA) guidelines for the management of ventral hernias, Bougard [/bib_ref]. But more recent systematic reviews found no overall increase in surgical site infection with the use of mesh in clean-contaminated and contaminated fields in the setting of ventral hernia repair [bib_ref] Surgical site infection in mesh repair for ventral hernia in contaminated field:..., Maatouk [/bib_ref] [bib_ref] Emergency repair of complicated abdominal wall hernias: WSES guidelines, Simone [/bib_ref] ; and that biologic mesh may be considered in these circumstances [bib_ref] Surgical site infection in mesh repair for ventral hernia in contaminated field:..., Maatouk [/bib_ref] [bib_ref] Emergency repair of complicated abdominal wall hernias: WSES guidelines, Simone [/bib_ref]. In this case, the surgeon thought it prudent to opt for primary suture repair.
# Conclusion
Incisional hernia appendicitis is a rare, albeit recognised clinical entity. Early recognition may be difficult based on clinical assessment alone, as it may be easily mistaken for an incarcerated or strangulated incisional hernia. Imaging plays a key role in making an accurate preoperative diagnosis. Surgical management consists of appendicectomy, as well as incisional hernia repair.
# Ethical approval
We have consulted our institution's research ethics committee, and were advised that this project is exempt from requiring ethics approval, provided written consent has been obtained from the patient and the clinical details provided have been de-identified.
# Source of funding
This research did not receive any funding.
# Author contribution
Marek Bak -Conceptualisation, writing -original draft. Kumail Jaffry -Literature review, writingreview and editing. Pee Yau Tan -Supervision, writingreview and editing.
## Registration of research studies
Not applicable -Not a First in Man study.
## Guarantor
Marek Bak -Primary author. Pee Yau Tan -Research Supervisor.
## Consent
Written consent was obtained from the patient for publication of this case report and the included clinical images. A copy of this consent form may be provided for review by the Editor-in-Chief on request.
## Declaration of competing interest
All authors declare they have no conflict of interest.
[fig] Figure 1: CT images of appendix entering hernia defect. [/fig]
[fig] Figure 2: Laparoscopic view of appendix entering hernia defect, and previous nylon suture used in the closure of the ileostomy defect. [/fig]
|
Trapped Stent in the Left Coronary Sinus in a Myocardial Infarction Patient
Stent entrapment is a very rare complication of percutaneous coronary intervention. The interventional approach could be a treatment strategy. However, if it does not work, surgical treatment should be considered. Here, we report a case of surgical treatment of stent entrapment in the left coronary sinus of a 53-year-old male patient.A 53-year-old male was admitted to the hospital with chest discomfort that had persisted for five days. His medical history was unremarkable, except for a tuberculosis infection 10 years earlier. Elevated cardiac enzyme levels were observed, including 0.1 ng/mL of troponin T and 11.1 ng/mL of creatine kinase-myocardial band. Electrocardiography showed that the ST segment was elevated at the anterior lead. A chest X-ray showed pulmonary congestion. Transthoracic echocardiography revealed a reduction of the ejection fraction to 35%, with severe hypokinesis of the anteroseptal area. The patient was diagnosed with myocardial infarction accompanied by pulmonary edema. Selective coronary angiography showed 90% stenosis in the proximal to mid-portion of the left anterior descending (LAD) artery. The lesion was long and irregular. We decided to insert two stents into the lesions. The first stent was inserted uneventfully into the proximal LAD artery, after which an attempt was made to insert the second stent into the mid-portion of the LAD artery because of the irregularity of the lesion. However, when the stent was extracted from the balloon, it was trapped and deformed. One part of it floated in the left coronary sinus(Fig. 1). We tried to push and withdraw the stent by snaring it, but failed. We planned the surgical removal of the trapped stent and a coronary artery bypass graft. With cardiopulmonary bypass, we were able to remove the stent through the aortotomy site(Fig. 2). We also performed left internal thoracic artery bypass to the distal portion of the LAD artery. The next day, the patient was extubated and he recovered well. However, on the second postoperative day, he had a sudden cough that produced fresh blood. He was intubated again and was subjected to chest computed tomography (CT). The chest CT showed active bleeding from the right middle lobe(Fig. 3). A lesion was also found in the bronchoscopy examination. We resectioned the right middle lobe. The lesion in the right middle lobe may have existed before the coronary angiography. In the follow-up echocardiography, the wall motion of the left ventricle was good. The ejection fraction increased from 35% to 65%.The patient had postoperative pneumonia and acute renal failure.Korean J Thorac Cardiovasc Surg 2015;48:368-370 □ Case Report □ http://dx.
## Trapped stent in the left coronary sinus in a myocardial infarction patient
Sun Han, M.D., Pil Won Seo, M.D.
Stent entrapment is a very rare complication of percutaneous coronary intervention. The interventional approach could be a treatment strategy. However, if it does not work, surgical treatment should be considered. Here, we report a case of surgical treatment of stent entrapment in the left coronary sinus of a 53-year-old male patient. He was treated with active intensive care including antibiotics, hemodialysis, and pulmonary rehabilitation. His condition waxed and waned. He was improving and underwent conservative treatment including hemodialysis in the general ward.
However, unfortunately, he passed away at 5 months after the operation due to aspiration pneumonia.
# Discussion
Entrapment of a coronary stent is a rare and dangerous complication of percutaneous coronary intervention. It may be life-threatening and may sometimes require emergency surgi-cal treatment. The increased use of endovascular intervention has resulted in an increasing number of complications pertaining to unretrievable devices due to abrupt entrapment. Several bailout devices have been developed such as GooseNeck snares, basket retrieval devices, and angioplasty balloons. In a previous study, 15% to 20% of the cases of failed percutaneous retrieval of catheter remnants were referred for surgical removal [bib_ref] Serial optical coherence tomography images of trapped balloon catheter after bailout stenting, Kawamoto [/bib_ref].
An angulated coronary anatomy, coronary calcification, a long stent, and sequential stents may cause stent entrapment [bib_ref] A rare and avoidable complication of percutaneous coronary intervention: stent trapped in..., Çicek [/bib_ref]. In the case of a long lesion, the distal stent is generally inserted earlier, because if the proximal stent is inserted earlier, it may entrap the distal stent. In addition, poor stent tractability, flexibility, and conformability may cause stent entrapment in the coronary artery. In one case, the order of insertion of sequential stents in a long coronary lesion was inverted, but the patient had no other predisposing factors.
These days, to improve flexibility, stents often have fewer fixed links than the past stents. However, these designs might reduce the strength of the stent and, therefore, pose an increased risk of longitudinal deformation, which is a cause of stent entrapment [bib_ref] Trapped rotablation wire causing longitudinal stent deformation, Lempereur [/bib_ref].
An entrapped stent could cause a blood clot. This may be critical because it may cause sudden myocardial infarction [bib_ref] Removal of a ruptured, detached, and entrapped angioplasty balloon after coronary stenting, Madronero [/bib_ref]. Transcatheter removal of an entrapped stent is usually better, faster, and safer than surgical removal. The use of a snaring wire or another balloon is a treatment alternative for stent entrapment. However, if this does not work, surgical treatment should be strongly considered, particularly in the left coronary opening, as in our patient. However, surgical intervention may have catastrophic results. With respect to a coronary artery bypass graft, it is preferable to anastomose the graft to the coronary segment distal to the entrapment site. In conclusion, here, we have reported a case of stent entrapment, which is a rare complication of percutaneous coro-nary intervention.
[fig] Figure 1: revealed a reduction of the ejection fraction to 35%, with severe hypokinesis of the anteroseptal area. The patient was diagnosed with myocardial infarction accompanied by pulmonary edema. Selective coronary angiography showed 90% stenosis in the proximal to mid-portion of the left anterior descending (LAD) artery. The lesion was long and irregular. We decided to insert two stents into the lesions. The first stent was inserted uneventfully into the proximal LAD artery, after which an attempt was made to insert the second stent into the mid-portion of the LAD artery because of the irregularity of the lesion. However, when the stent was extracted from the balloon, it was trapped and deformed. One part of it floated in the left coronary sinus(Fig. 1). We tried to push and withdraw the stent by snaring it, but failed. We planned the surgical removal of the trapped stent and a coronary artery bypass graft. With cardiopulmonary bypass, we were able to remove the stent through the aortotomy site(Fig. 2). We also performed left internal thoracic artery bypass to the distal portion of the LAD artery. The next day, the patient was extubated and he recovered well. However, on the second postoperative day, he had a sudden cough that produced fresh blood. He was intubated again and was subjected to chest computed tomography (CT). The chest CT showed active bleeding from the right middle lobe(Fig. 3). A lesion was also found in the bronchoscopy examination. We resectioned the right middle lobe. The lesion in the right middle lobe may have existed before the coronary angiography. In the follow-up echocardiography, the wall motion of the left ventricle was good. The ejection fraction increased from 35% to 65%.The patient had postoperative pneumonia and acute renal failure. First stent deployed successfully in the proximal portion of the left anterior descending artery (dotted arrow). The second stent was entrapped and floated in the left coronary sinus (solid arrow). [/fig]
[fig] Figure 2: (A) Entrapped and deformed stent in the left coronary sinus (circle). (B) Deformed stent after its bailout. [/fig]
[fig] Figure 3: Chest computed tomography showing active bleeding from the right middle lobe (arrow). [/fig]
|
Use of Machine Learning in the Analysis of Indoor ELF MF Exposure in Children
Characterization of children exposure to extremely low frequency (ELF) magnetic fields is an important issue because of the possible correlation of leukemia onset with ELF exposure. Cluster analysis-a Machine Learning approach-was applied on personal exposure measurements from 977 children in France to characterize real-life ELF exposure scenarios. Electric networks near the child's home or school were considered as environmental factors characterizing the exposure scenarios. The following clusters were identified: children with the highest exposure living 120-200 m from 225 kV/400 kV overhead lines; children with mid-to-high exposure living 70-100 m from 63 kV/150 kV overhead lines; children with mid-to-low exposure living 40 m from 400 V/20 kV substations and underground networks; children with the lowest exposure and the lowest number of electric networks in the vicinity. 63-225 kV underground networks within 20 m and 400 V/20 kV overhead lines within 40 m played a marginal role in differentiating exposure clusters. Cluster analysis is a viable approach to discovering variables best characterizing the exposure scenarios and thus it might be potentially useful to better tailor epidemiological studies. The present study did not assess the impact of indoor sources of exposure, which should be addressed in a further study.We used cluster analysis, a non-parametric and unsupervised Machine Learning (ML) approach. In contrast to parametric approaches, such as correlation analysis or multivariate analysis, cluster analysis is applied in exploratory data mining to understand possible relationships within the data without assuming any linear or non-linear parametric model to explain the observed data. The application of ML approaches to problems related to electromagnetic field (EMF) exposure is at the very beginning; to the best of our knowledge, sporadic examples of applications can be found in the prediction of radiofrequency (RF) radiation effect on plants[19]or the prediction of wireless local-area network (LAN) EMF in the indoor environment[20]. This is the first time that ML has been used for the characterization of ELF exposure scenarios in children. Cluster analysis was applied to personal exposure measurements recorded in the EXPERS study[5,6]from children living in France to identify clusters of children with similar exposure levels and similar exposure scenarios (i.e., similar electric network configuration in the proximity of the home or school). The aim of the paper was neither to demonstrate the influence of ELF exposure on children Leukemia, nor to present a novel measurement study of personal ELF exposure. The aim was to further analyze data coming from an already done measurement campaign-the EXPERS study-with a novel approach in order to discover similarities in the exposure scenarios. In particular, the focus was to understand (i) what were the main characteristics of the clusters of ELF exposure with respect to the type of electric networks in the vicinity of the measurement location (i.e., a child's home and school), (ii) what were the main differences between the clusters, and (iii) which environmental factors better discriminated the clusters. The paper is organized as follows: Section 2 "Materials and Methods" reports the details of the source of the analyzed data (Section 2.1 Data Source) and the description of the method used to perform cluster analysis and its validation (Section 2.2 Data Analysis); Section 3 "Results" reports some statistics on the analyzed dataset (Section 3.1 Descriptive Results), the results of the procedure used to determine the optimal number of clusters and the evaluation of the cluster solution (Section 3.2 Choice of the Optimal Number of Clusters and Evaluation of Clustering Performance), and the detailed description of the characteristics of the clusters we found (Section 3.3 Results of Clustering Analysis); Section 4 "Discussion" reports the discussion of the results, current study limitations and future directions; the last section, Section 2 "Conclusions", summarizes the main findings of the study.
# Introduction
Interest in the analysis of children exposure to extremely low frequency magnetic fields (ELF-MF, 40-800 Hz) was raised with the first epidemiological study where exposure to ELF was found to be a possible health risk factor for childhood leukemia [bib_ref] Electrical wiring configurations and childhood cancer, Wertheimer [/bib_ref]. Although leukemia is the most common type of cancer in children younger than 15 years, its etiology is still unknown. Previous studies evidenced that daily average exposures >0.4 µT increased the risk of childhood leukemia's onset, without any causal relationship [bib_ref] A pooled analysis of magnetic fields, wire codes, and childhood leukemia, Greenland [/bib_ref] [bib_ref] A pooled analysis of magnetic fields and childhood leukemia, Ahlbom [/bib_ref] [bib_ref] Pooled analysis of recent studies on magnetic fields and childhood leukemia, Kheifets [/bib_ref]. To further investigate their possible relationship with children leukemia, several studies were conducted to measure personal exposure to ELF in children in Europe [bib_ref] Methodology of a study on the French population exposure to 50 Hz..., Bedja [/bib_ref] [bib_ref] Exposure of children to extremely low frequency magnetic fields in France: Results..., Magne [/bib_ref] [bib_ref] Extremely low-frequency magnetic fields and risk of childhood leukemia: A risk assessment..., Schüz [/bib_ref] [bib_ref] Relative contribution of residential and occupational magnetic field exposure over twenty-four hours..., Forssén [/bib_ref] [bib_ref] Exposure of children to residential magnetic fields in Norway: Is proximity to..., Vistnes [/bib_ref] , United States [bib_ref] Childhood exposure to magnetic fields: Residential area measurements compared to personal dosimetry, Friedman [/bib_ref] [bib_ref] Alternative magnetic field exposure metrics: Relationship to TWA, appliance use, and demographic..., Foliart [/bib_ref] [bib_ref] Correlation of year-to-year magnetic field exposure metrics among children in a leukemia..., Foliart [/bib_ref] [bib_ref] Assessing human exposure to power-frequency electric and magnetic fields, Kaune [/bib_ref] [bib_ref] Exposures of children in Canada to 60-Hz magnetic and electric fields, Deadman [/bib_ref] [bib_ref] Power-frequency electric and magnetic fields and risk of childhood leukemia in Canada, Mcbride [/bib_ref] , and Asia [bib_ref] Survey of residential extremely-low-frequency magnetic field exposure among children in Taiwan, Li [/bib_ref] [bib_ref] Analysis of individual-and school-level clustering of power frequency magnetic fields, Lin [/bib_ref] [bib_ref] Development of a New Personal Magnetic Field Exposure Estimation Method for Use..., Yang [/bib_ref].
The aim of the present study was to gain deeper insights on the variables (such as, sources of ELF in the environment) that characterize real-life exposure scenarios and better characterize ELF exposure in children. In particular, we focused on the characterization of electric networks in the proximity of a child's home or school as environmental factors that potentially could influence indoor exposure.
# Materials and methods
## Data source
A database of personal exposure measurements to ELF MF in children was analyzed. The data came from the EXPERS study [bib_ref] Methodology of a study on the French population exposure to 50 Hz..., Bedja [/bib_ref] [bib_ref] Exposure of children to extremely low frequency magnetic fields in France: Results..., Magne [/bib_ref] and consisted of exposure measurements recorded from children aged 0-14 years located in France. The database contained the recordings from 977 children located in rural or urban places (948 children) and in Paris (29 children) (for more details about the areas under test, see [bib_ref] Exposure of children to extremely low frequency magnetic fields in France: Results..., Magne [/bib_ref]. Measurements were performed during cold seasons (February-April 2007, October 2007-April 2008, and October 2008-January 2009). An EMDEX II (Enertech, Campbell, CA, USA (https: //www.enertech.net/html/EMDEXII.html)) personal dosimeter was used to measure broadband (40-800 Hz) and harmonic (100-800 Hz) MF amplitudes with a sensitivity of 0.01-300 µT. Measurements were recorded with a sampling rate of 3 s over 24 h for one day. For each child, the database contained the number of power lines and substations close to the measurement site, i.e., the child's home or school, as identified by the French grid operators (RTE for high voltage and Enedis for mid and low voltage) from the Lambert II coordinates of home and school addresses. . Also, the database contained a timetable reporting time, duration, and location (e.g., at home, at school, indoors, outdoors, etc.) of child activities during the 24 h of exposure measurement.
# Data analysis
Using the information available in the daily activity tables, the geometric mean (GM) of the 50 Hz component of B field was computed for exposure measurements that were recorded indoors when the child was at home and, where available, at school. As in previous reference studies on ELF MF exposure [bib_ref] Potential health impacts of residential exposures to extremely low frequency magnetic fields..., Grellier [/bib_ref] , we decided to consider GM as the metric to represent B field values to facilitate the comparison of our results with similar studies. We also decided to analyze only exposures recorded during the day and not during the night to exclude any confounding effect due to the presence of the alarm clock in night recordings, as evidenced in [bib_ref] Exposure of children to extremely low frequency magnetic fields in France: Results..., Magne [/bib_ref]. As written above, in this study, we decided to analyze as variables the B field measured indoor and the type of electric networks in the vicinity of the measurement site; at the moment, we did not analyze indoor sources of exposure. A total of 1793 measurements were available for the analysis, including 977 recordings made at home and 816 at school.
Cluster analysis was performed on a matrix of 1793 × 11 real data, each representing a child exposure (row) and its characterization (column) in terms of B field and number of power lines and substations in close proximity to home or school for each of the 10 types of the electric networks described in the section above.
Data were analyzed with K-means clustering (Matlab (ver. R2018a), MatWorks Inc., Natick MA, USA). The goal of K-means clustering is to partition the data into K clusters so that the pairwise dissimilarities between data of the same cluster are smaller than those in the other clusters. In K-means clustering, dissimilarity d(x i , x i ) between two data vectors x i and x i (i, i = 1, · · · , N) of P components is measured with the squared Euclidean distance:
[formula] d(x i , x i ) = P ∑ j=1 x ij − x i j 2 ,(1) [/formula]
where x ij and x i j are the jth components (j = 1, · · · , P) of vectors x i and x i . In K-means clustering, the user has to specify the number of partitions K < N into which the data points have to be clustered; then, the K-means algorithm assigns each data point to one and only one of the K clusters by minimizing the so-called "within cluster" point scatter W(C):
[formula] W(C) = 1 2 K ∑ k=1 ∑ i, i ∈Sk d(x i , x i ),(2) [/formula]
where Sk is the set of observations in the kth cluster (k = 1, · · · , K). Minimization of W(C) is obtained by an iterative procedure that at each step searches among the possible assignments of the N data points to the K clusters that one that improves (i.e., minimizes) W(C) from its value obtained at the previous step. The iteration is stopped when the new assignment solution is unable to improve W(C) further. In K-means clustering, it is possible to demonstratethat minimization of (2) is equivalent to minimizing the average distance of the data points of the same cluster from the cluster centroid x k = (x 1k , · · · , x Pk ), which is the mean vector of the data assigned to the same cluster. As such, each cluster centroid can be regarded as the "representative" of the characteristics of the pool of data assigned to the cluster. Centroid initialization was obtained from the algorithm published by [bib_ref] The Advantages of Careful Seeding, David [/bib_ref] (as implemented in Matlab), which uses heuristics to find centroid seeds and improves the running times of the K-means clustering algorithm. To be sure that the partitioning solution was stable, we repeated the clustering using new initial cluster centroid positions by choosing from 30 possibly different sets of seeds, according to the algorithm by [bib_ref] The Advantages of Careful Seeding, David [/bib_ref]. The average silhouette score was used to determine K (i.e., the optimal number of clusters) and the appropriateness of the cluster solution (i.e., it was used for the validation of the cluster solution). The silhouette score S i for a data point x i is a measure of the similarity between x i and the other data of the same cluster, when compared to data in other clusters. It is computed as:
[formula] S i = b i − a i max(a i , b i ) ,(3) [/formula]
where a i is the average distance of x i to the other points in the same cluster, b i is the minimum average distance of x i to all the other points in different clusters, minimized over the clusters, and −1 ≤ S i ≤ 1. A high silhouette value close to 1 indicates that x i is far away from neighboring clusters and well-matched to its own cluster, that is, x i has been assigned to the proper cluster. A value close to −1 indicates that x i has been assigned to the wrong cluster. The average of S i across all data points (i = 1, · · · , N) is a measure of the overall appropriateness of the clustering solution over the entire dataset. The average silhouette score can be used to determine the optimal number of clusters K: by computing the average silhouette score as a function of K, it is possible to determine the best value of K as the value for which the average silhouette reaches a maximum. The value of the average silhouette score is also used as a metric to assess the robustness of the clustering solution, with values greater than 0.5 indicating a good quality in the partitioning of the data.
# Results
# Descriptive results
The average value of B field (GM) across all 1793 recordings was 0.02 µT (1st quartile Q1: 0.006 µT; 3rd quartile Q3: 0.042 µT), and the maximum was 1.09 µT; one should note that B field was >0.4 µT only in 5 (five) recordings out of 1793. [fig_ref] Table 1: Distribution of electric networks in the analyzed dataset [/fig_ref] shows the distribution of the different types of electric networks in the analyzed dataset of indoor measurements. It is possible to see that more than two thirds of the measurements (66.8%) in the dataset came from children living or going to schools near underground networks of low voltage, while nearly half of them were from children living or going to school near underground networks of mid voltage (45.7%) or overhead lines of low voltage (43.8%), and fewer than 15% (13.7%) of the measurements were from children living or going to school near a substation. A small number of measurements (below 4%) were from children living or going to school near underground networks of high or extra-high voltage or overhead lines of mid to ultra-high voltage. A total of 201 out of 1793 (11.2%) measurements came from sites that apparently were far from any of the electric networks considered in the present study; in other words, none of the electric networks mentioned above was found near the measurement site of these last data. This group of data may consist of measurements from places served by local distributors of electricity rather than by RTE and Enedis, representing 5% of the distribution of electricity in France. In addition, this group included measurements made in rural areas where the precision of geolocalization might have been less than in urban areas, and thus might have influenced the localization of the electric grid near the measurement site, especially if the precision was larger than 40 m. [fig_ref] Table 1: Distribution of electric networks in the analyzed dataset [/fig_ref] also shows the maximum and mean number of power lines and substations found in the proximity of the measurement site. The maximum number of underground cables of low voltage close to the measurement site was 59, followed by 27 underground cables of mid voltage and 16 overhead lines of low voltage; the number of underground cables of high or extra-high voltage or overhead lines of mid to ultra-high voltage or substations close to the measurement site was at maximum five. The analysis of the dataset also revealed that a significant percentage of measurements came from children who live or go to schools near two different types of electric networks at the same time. Data in show that children who lived near underground networks of mid voltage also had underground networks of low voltage in the vicinity (38.8% of the total number of measurements); as expected by grid construction, those near a substation were found to have underground networks of low (12.7% of the measurements) or mid voltage (13.2% of the measurements) and overhead lines of low voltage (4.9% of the measurements) in the vicinity; those near an overhead line of low voltage also had underground networks of low (24%) or mid voltage (19.6%) in the vicinity; children living or going to schools near overhead lines of mid to ultra-high voltage were found to have underground networks of low and mid voltage (up to 1.2% of total number of measurements) and overhead lines of low voltage (up to 2.3%) in the vicinity. None of the children near overhead lines of mid to ultra-high voltage lived or went to schools that are also in the vicinity of underground networks of high and extra-high voltage. This is another finding which is due to construction on the 63 kV-400 kV grid; as a matter of fact, it is very rare, except for sites near high voltage substations, to live near underground and overhead lines of 63 kV-400 kV at the same time. Overhead lines of this type are mainly found in the country side, whereas underground lines are found in big cities. . Distribution of electric networks in the analyzed dataset (cont.d). The main diagonal (shaded background) shows the number (and %) of measurements from children whose home or school is in proximity of only a single type of network, whereas data below the main diagonal show the number (and %) of measurements in close proximity to two different types of networks at the same time.
[formula] 4 (0.2) 5 (0.3) 0 extra-high 7 (0.4) 6 (0.3) 0 0 OVHD low 431 (24.0) 352 (19.6) 4 (0.2) 1 (0.1) 228 (12.7) mid 22 (1.2) 11 (0.6) 0 0 41 (2.3) 4 (0.2) high 4 (0.2) 3 (0.2) 0 0 6 (0.3) 2 (0.1) 1 (0.1) extra-high 5 (0.3) 2 (0.1) 0 0 5 (0.3) 0 1 (0.1) 0 ultra-high 2 (0.1) 1 (0.1) 0 0 1 (0.1) 0 0 2 (0.1) 0 Substation 228 (12.7) 237 (13.2) 1 (0.1) 1 (0.1) 87 (4.9) 11 (0.6) 2 (0.1) 2 (0.1) 0 0 1 [/formula]
Number of all measurements = 1793. [fig_ref] Figure 1: Pictorial representation of the distribution of electric networks in the analyzed dataset... [/fig_ref] illustrates, with an interconnected graph, the distributions of electric network types reported in . Links in the graph show which electric networks are found in the vicinity of the others. As commented above (see , substations are typically found in the vicinity of underground networks of low and mid voltage and overhead lines of low voltage. A link (straight line) between electric networks "A" and "B" means that there are some children in the dataset whose home or school is near both "A" and "B" networks at the same time. Loop links represent children that are near to only one single type of electric networks. Link thickness is proportional to the number of children for which the link is valid. Numbers in or next to the nodes are the type of electric networks: 1 = UND_low; 2 = UND_mid; 3 = UND_high; 4 = UND_extra; 5 = OVHD_low; 6 = OVDH_mid; 7 = OVHD_high; 8 = OVHD_extra-high; 9 = OVHD_ultra-high; 10 = substations. [fig_ref] Figure 2: Average silhouette score as a function of the number of partitions [/fig_ref] shows the average silhouette score (see Equation (3)) as a function of the number of clusters used for partitioning the matrix of 1793 × 11 real data. It is possible to see that the score increases as the number of partitions increases from two to six. The increase is steep up to four clusters; at five and six clusters, there is still an increase, but it is minimal if compared to the increase observed from two to four clusters. This means that, using five or six partitions, the quality of the clustering solution would not significantly be better than that obtained with four partitions. Node size is proportional to the number of children in the dataset living or going to a school near that particular electric network. A link (straight line) between electric networks "A" and "B" means that there are some children in the dataset whose home or school is near both "A" and "B" networks at the same time. Loop links represent children that are near to only one single type of electric networks. Link thickness is proportional to the number of children for which the link is valid. Numbers in or next to the nodes are the type of electric networks: 1 = UND_low; 2 = UND_mid; 3 = UND_high; 4 = UND_extra; 5 = OVHD_low; 6 = OVDH_mid; 7 = OVHD_high; 8 = OVHD_extra-high; 9 = OVHD_ultra-high; 10 = substations. [fig_ref] Figure 2: Average silhouette score as a function of the number of partitions [/fig_ref] shows the average silhouette score (see Equation (3)) as a function of the number of clusters used for partitioning the matrix of 1793 × 11 real data. It is possible to see that the score increases as the number of partitions increases from two to six. The increase is steep up to four clusters; at five and six clusters, there is still an increase, but it is minimal if compared to the increase observed from two to four clusters. This means that, using five or six partitions, the quality of the clustering solution would not significantly be better than that obtained with four partitions. [fig_ref] Table 3: Cluster size [/fig_ref] shows the number of measurements that the K-means algorithm assigned to each cluster (i.e., the cluster size), by varying the number of partitions. As expected, the size of the clusters generally decreased as the number of partitions increased (a deviation from this behavior was seen at cluster #5 in the six-partition solution where the cluster size slightly increased from 264 to 268 if compared to the five-partition solution). In particular, the minimum size of the clusters was seven for the five-partition solution and three for the six-partition solution. [fig_ref] Table 3: Cluster size [/fig_ref] shows the number of measurements that the K-means algorithm assigned to each cluster (i.e., the cluster size), by varying the number of partitions. As expected, the size of the clusters generally decreased as the number of partitions increased (a deviation from this behavior was seen at cluster #5 in the six-partition solution where the cluster size slightly increased from 264 to 268 if compared to the five-partition solution). In particular, the minimum size of the clusters was seven for the five-partition solution and three for the six-partition solution. [fig_ref] Table 3: Cluster size [/fig_ref]. Cluster size (i.e., number of measurements) for different partitioning solutions, from 2-to 6cluster solutions. [fig_ref] Table 1: Distribution of electric networks in the analyzed dataset [/fig_ref] It is a good and consolidated practice to choose the number of partitions so as to maximize the silhouette score and, at the same time, have a reasonable cluster size. For the dataset analyzed in the study, a good compromise between silhouette and cluster size was obtained by partitioning the data with four clusters. Last but not least, it is important to note that the solution with four partitions resulted in a silhouette score of nearly 0.65 which is indicative of good quality of the partitioning.
## Choice of the optimal number of clusters and evaluation of clustering performance
## Choice of the optimal number of clusters and evaluation of clustering performance
## Cluster # number of partitions
## Results of clustering analysis
The results obtained with the four-partition solution can be seen in which displays the coordinates (in absolute values) of the centroids of the four clusters. Each centroid is a multidimensional vector with 11 components, one for each of the 11 analyzed measurement variables. The number of measurements assigned to each cluster can be found in [fig_ref] Table 3: Cluster size [/fig_ref] , in correspondence to the solution with four partitions. It is a good and consolidated practice to choose the number of partitions so as to maximize the silhouette score and, at the same time, have a reasonable cluster size. For the dataset analyzed in the study, a good compromise between silhouette and cluster size was obtained by partitioning the data with four clusters. Last but not least, it is important to note that the solution with four partitions resulted in a silhouette score of nearly 0.65 which is indicative of good quality of the partitioning.
## Results of clustering analysis
The results obtained with the four-partition solution can be seen in which displays the coordinates (in absolute values) of the centroids of the four clusters. Each centroid is a multi-dimensional vector with 11 components, one for each of the 11 analyzed measurement variables. The number of measurements assigned to each cluster can be found in [fig_ref] Table 3: Cluster size [/fig_ref] , in correspondence to the solution with four partitions. . Coordinates of the centroids of the four clusters for each of the 11 analyzed measurement variables (from "B" to "Substation"). By definition, in K-means clustering, the centroid is obtained as the mean of the data assigned to the cluster. As explained in the Methods Section, each cluster centroid can be regarded as the "representative" of the characteristics of the pool of data assigned to the cluster. The analysis of the coordinates of the centroids is thus useful to understand (i) what are the main characteristics of the clusters, (ii) what are the main differences between the clusters, and (iii) which are the variables that better discriminate the clusters. With regard to points (i) and (ii), from the inspection of , it is possible to see that cluster #1 pooled together children with the highest exposures (average GM value of B in the cluster: 0.146 µT; Q1: 0.066 µT; Q3: 0.220 µT) and the highest number of extra-high and ultra-high voltage overhead lines close to their home or school (mean OVHD_extra-high: 1 line; mean OVHD_ultra-high: 0.8 lines). Cluster #2 contained children with the highest number of high voltage overhead lines (mean OVHD_high: 1 line) in the vicinity and a mid-to-high exposure (average GM value of B in the cluster: 0.053 µT; Q1: 0.014 µT; Q3: 0.051 µT) (please note that here and in other cases reported below, it may happen that the average would be lower than Q3; this is not an error because quartiles are calculated with respect to the median value of the sample distribution and not the mean). Cluster #3 was for children with a mid-to-low GM B value (average: 0.025 µT; Q1: 0.003 µT; Q3: 0.034 µT) and with the highest number of substations (mean value: 0.9 substations) and underground networks of low and mid voltage (mean UND_low: 11.4 cables; mean UND_mid: 4.6 cables) near home or school; finally, cluster #4 contained children with the lowest exposure (average GM value of B in the cluster: 0.019 µT; Q1: 0.002 µT; Q3: 0.016 µT) and the lowest number of underground cables, overhead lines, and substations near their home or school. Cluster #4 contained all 201 measurements coming from sites that apparently were far from any electric network. For these measurements, the average GM value of B was 0.014 µT (Q1: 0.001 µT; Q3: 0.013 µT) which was well comparable to that obtained from the other measurements assigned to the same cluster #4 (average GM value of B: 0.019 µT; Q1: 0.002 µT; Q3: 0.016 µT). In addition, cluster #4 contained all five samples with B > 0.4 µT (GM). As observed in the other samples assigned to the same cluster, these latter samples corresponded to children with the lowest number of underground cables, substations, and overhead lines of mid to ultra-high voltage in the vicinity. This explains why cluster analysis assigned these latter data to cluster #4; from an overall point of view, the characteristics of these samples were very similar to the other samples of cluster #4. A deeper inspection revealed that these five samples were characterized by a slightly higher number of low voltage overhead lines (Q1: 1 line; Q3: 4 lines) than the other samples in the same cluster (Q1: 0 lines; Q2: 2 lines); this might explain the higher exposure values that characterized these latter measurements.
## Cluster # b (µt) und (n) ovhd (n) substation (n) low mid high extra-high low mid high extra-high ultra-high
With regard to point (iii), to identify which variables better discriminated each exposure cluster, the centroids displayed in were normalized by scaling each coordinate separately to the maximum of each column, i.e., to the maximum along each variable. [fig_ref] Figure 3: Int [/fig_ref] displays the normalized values of the centroid coordinates for each of the 11 variables considered in the study.
From [fig_ref] Figure 3: Int [/fig_ref] , it can be seen that the coordinates of the centroids along the variables UND_high, UND_extra-high, OVHD_low, and OVHD_mid did not vary across the clusters; this means that, on average, data across the different clusters did not show significant differences along these variables. We can thus conclude that these latter variables did not contribute to differentiating the clusters. From a practical point of view, this means that the number of underground networks of high or extra-high voltage and the number of overhead lines of low or mid voltage near home or school were not relevant, alone, to differentiating the characteristics of the four clusters (i.e., the four patterns of exposition). From [fig_ref] Figure 3: Int [/fig_ref] , it can be seen that the coordinates of the centroids along the variables UND_high, UND_extra-high, OVHD_low, and OVHD_mid did not vary across the clusters; this means that, on average, data across the different clusters did not show significant differences along these variables. We can thus conclude that these latter variables did not contribute to differentiating the clusters. From a practical point of view, this means that the number of underground networks of high or extrahigh voltage and the number of overhead lines of low or mid voltage near home or school were not relevant, alone, to differentiating the characteristics of the four clusters (i.e., the four patterns of exposition).
On the other hand (see [fig_ref] Figure 3: Int [/fig_ref] , all the remaining variables seem to contribute to differentiating the clusters; as a matter of fact, the coordinate of the centroids along these latter variables varied across the clusters, thus meaning that the data assigned to the different clusters have different and unique characteristics. In particular, it is clear that variables OVHD_extra-high and OVHD_ultrahigh contributed to differentiating and characterizing cluster #1 (the one with the highest value of B). From [fig_ref] Figure 3: Int [/fig_ref] , it is possible to comment that children living or going to schools near overhead lines of extra-high and ultra-high voltage are most likely characterized by a high value of B. Variable OVHD_high contributed to differentiating cluster #2 (the one with a B field of mid-to-high value); this suggested that children living or going to schools near overhead lines of high voltage are most likely characterized by a mid-to-high B value. Finally, variables substation, UND_low, and UND_mid contributed to differentiating and characterizing cluster #3 (the one with a B field of mid-to-low On the other hand (see [fig_ref] Figure 3: Int [/fig_ref] , all the remaining variables seem to contribute to differentiating the clusters; as a matter of fact, the coordinate of the centroids along these latter variables varied across the clusters, thus meaning that the data assigned to the different clusters have different and unique characteristics. In particular, it is clear that variables OVHD_extra-high and OVHD_ultra-high contributed to differentiating and characterizing cluster #1 (the one with the highest value of B). From [fig_ref] Figure 3: Int [/fig_ref] , it is possible to comment that children living or going to schools near overhead lines of extra-high and ultra-high voltage are most likely characterized by a high value of B. Variable OVHD_high contributed to differentiating cluster #2 (the one with a B field of mid-to-high value); this suggested that children living or going to schools near overhead lines of high voltage are most likely characterized by a mid-to-high B value. Finally, variables substation, UND_low, and UND_mid contributed to differentiating and characterizing cluster #3 (the one with a B field of mid-to-low value); this suggested that children living or going to schools near substations and underground cables of low and mid voltage are most likely characterized by a B field of mid-to-low value.
To go deeper in understanding the main features of the four clusters, [fig_ref] Figure 4: Within cluster analysis [/fig_ref] displays the distribution of indoor measurements within each cluster as a function of the electric network type found near the measurement site. value); this suggested that children living or going to schools near substations and underground cables of low and mid voltage are most likely characterized by a B field of mid-to-low value.
To go deeper in understanding the main features of the four clusters, [fig_ref] Figure 4: Within cluster analysis [/fig_ref] displays the distribution of indoor measurements within each cluster as a function of the electric network type found near the measurement site. The first panel of [fig_ref] Figure 4: Within cluster analysis [/fig_ref] shows that more than 80% of the measurements assigned to cluster #1 were from children near overhead lines of extra-high voltage and that nearly 40% of the measurements were from children near overhead lines of ultra-high voltage. More than 40% of the measurements in cluster #1 were from children near underground low voltage cables and a small percentage (below 10%) were from children close to underground networks of mid voltage or to substations. None of the measurements in cluster #1 were from children near underground networks of high and extra-high voltage or to overhead lines of mid and high voltage.
With regard to cluster #2, 100% of the measurements were from children near at least one overhead line of high voltage; 60% were from children close to overhead lines of low voltage and 40% The first panel of [fig_ref] Figure 4: Within cluster analysis [/fig_ref] shows that more than 80% of the measurements assigned to cluster #1 were from children near overhead lines of extra-high voltage and that nearly 40% of the measurements were from children near overhead lines of ultra-high voltage. More than 40% of the measurements in cluster #1 were from children near underground low voltage cables and a small percentage (below 10%) were from children close to underground networks of mid voltage or to substations. None of the measurements in cluster #1 were from children near underground networks of high and extra-high voltage or to overhead lines of mid and high voltage.
With regard to cluster #2, 100% of the measurements were from children near at least one overhead line of high voltage; 60% were from children close to overhead lines of low voltage and 40% from children close to underground networks of low voltage. Only a smaller percentage (<30%) of measurements in this cluster were from children who live close to underground cables of mid voltage or overhead lines of mid and extra-high voltage or substations. None of the measurements in cluster #2 were from children near underground networks of high and extra-high voltage or overhead lines of ultra-high voltage. Cluster #3 appears to be mainly characterized by measurements from children living near underground networks of low (97% of the measurements) and mid voltage (100%) and substations (87%). It is important to note that, as displayed in and [fig_ref] Figure 1: Pictorial representation of the distribution of electric networks in the analyzed dataset... [/fig_ref] , children living near substations are also close to underground cables of low and mid voltage. A very small proportion (1%) of the measurements were from children who live near underground networks of extra-high voltage and overhead lines of mid voltage; none of the measurements in cluster #3 were from children near underground networks of high voltage and overhead lines of high, extra-high, and ultra-high voltage. Finally, cluster #4 was characterized by measurements from children living only near underground cables of low and mid voltage and overhead lines of low voltage.
# Discussion
K-means cluster analysis revealed significant and recurrent patterns in personal exposure to ELF MF. The first pattern (cluster #1) corresponded to children living or going to school near overhead lines of extra (225 kV) and ultra-high voltage (400 kV); these children were all characterized by the highest values of MF observed in the dataset (the average GM value of B within the cluster was 0.146 µT). The second pattern (cluster #2) was found for children who live or go to school near overhead lines of high voltage (63/90/150 kV); these children were characterized by a mid-to-high value of MF (the GM value of B averaged within the cluster was 0.0523 µT). The third pattern corresponded to children with the highest number of underground cables of low (400 V) and mid voltage (20 kV) and substations in the vicinity of their home or school; they were characterized by a mid-to-low MF (the average GM value of B within the cluster was 0.025 µT). Finally, the last pattern (Cluster #4) corresponded to children far from power lines and substations; these children were characterized by the lowest values of MF (the average GM value of B within the cluster was 0.019 µT). All these results are in line with [bib_ref] Exposure of children to extremely low frequency magnetic fields in France: Results..., Magne [/bib_ref] (see, in particular, in [bib_ref] Exposure of children to extremely low frequency magnetic fields in France: Results..., Magne [/bib_ref] which reports the variables correlated with 24h GM), where the 24 h GM MF exposure was found to be correlated with overhead lines of high to ultra-high voltage, underground cables of low to mid voltage, and substations.
Interestingly, the analysis revealed that underground electric networks of high (63 to 150 kV) and extra-high voltage (225 kV) and overhead lines of low (400 V) and mid voltage (20 kV) were not relevant, alone, to differentiating the characteristics of the four clusters. With regard to underground networks of high and extra-high voltage, the probability of being close to these types of networks is typically low. As expected, because our sample size of 977 children was representative of the French population [bib_ref] Exposure of children to extremely low frequency magnetic fields in France: Results..., Magne [/bib_ref] , the resulting dataset contained a small number (12 out of 1793) of measurements made near these types of networks. This would mean that, from a methodological point of view, measurements near underground networks of high and extra-high voltage might have a lower impact on cluster analysis than those made near other types of networks that are more frequently seen in real practice. In parallel, it is also important to note that the contribution of underground networks to total indoor exposure can be small. Although the MF at the centerline of underground cables can be higher than that generated by an equivalent overhead line, the spatial attenuation of the MF is much steeper in these networks than in overhead lines. As a result, the MF to the sides of the cable is usually significantly lower than that of a power line [bib_ref] Attenuation of Low Frequency Magnetic Fields Produced by HV Underground Power Cables, Ippolito [/bib_ref] ; for example, the field at 1 m above ground level at a distance of 20 m from the centerline of a 400 kV cable is of the order of 0.10 ÷ 0.20 µT, whereas the MF of an overhead line of equivalent power is about 2 µT. If measurements are done at more than 1 m above ground level (as in the current study), it is expected that the field will be even lower and of less contribution to total exposure. A significant percentage of the measurements in the dataset were from children (43.8%) living near overhead lines of low voltage which were found to be of low efficacy in differentiating the clusters. A possible explanation for this aspect could be that these networks marginally contributed to the measured exposure. This is also in line with [bib_ref] Exposure of children to extremely low frequency magnetic fields in France: Results..., Magne [/bib_ref] , where no correlation was found between the 24 h GM MF exposure and low voltage lines. Finally, with regard to mid voltage overhead lines, they were found to have a marginal contribution in differentiating the clusters; again, one possible explanation for this result could be the small number of children in the dataset that live near mid voltage overhead lines (58 out of 1793). In addition, as commented in [bib_ref] Exposure of children to extremely low frequency magnetic fields in France: Results..., Magne [/bib_ref] , mid voltage lines were mainly found in rural areas characterized by lower exposure values if compared to the exposure measured in highly urbanized areas, such as in Paris. This last evidence would mean that, if present, mid voltage lines marginally contribute to the exposure.
As described above, the pattern observed in cluster #3 concerned children who live or go to school near substations and underground networks of low and mid voltage. As commented above, the MF of underground networks could be very low when measured at a distance of 20 m and even lower at greater distances. Our dataset contained information regarding the presence of low and mid voltage cables within 40 m from a child's home or school; thus it could be that the contribution of low and mid voltage cables to our exposure data might be very low. On the other hand, according to the network distribution observed in our dataset, children who live near substations usually have underground networks of low and mid voltage in close proximity. Thus, we can conclude that the true differentiating variable that characterized mainly cluster #3 was the presence of substations near home or school, whereas the presence of underground networks of low and mid voltage was only a consequence of the configuration of power grids in the analyzed dataset.
The scarce contribution of underground networks to MF at distances greater than 20 m and the results from cluster analysis seem to suggest that, in future personal exposure campaigns, it would be interesting to record the presence of underground networks at distances below 20 m (i.e., at a distance where their MF is not marginal). This would provide additional information needed to better determine the role of these networks in differentiating the exposure patterns.
With regard to the uncertainty due to measurements of the B field near the sensitivity threshold of the personal dosimeter, it could be that the GM calculated from the values of B measured with the device was below the sensitivity threshold (equal to 0.001 µT). However, we can assume that the impact of this aspect should be marginal because clusters were identified by the K-means algorithm by considering not only the value of B but also all the other 10 variables related to the number of electric networks in the vicinity of the measurement position.
It is important to note that the geolocalization of electric networks might have some slight uncertainties. As a matter of fact, the geolocalization of electric networks was done from the physical address of a child's home or school; that is from the mailbox, which is a point closer to the street. This means that the exact and real position of the child with respect to the electric network nearby cannot be known without uncertainties, especially in those locations where the child can move around large areas, such as in schools. It seems reasonable to assume that the impact of this uncertainty should be marginal. At the moment, it is not possible to give an exact and quantitative estimation of this aspect using the data analyzed in the present study. It could be interesting in further studies to go deeper in analyzing the impact of this uncertainty, for example by considering both the approximate locations (derived from the physical address of the child's home or school) and, if available, the exact GPS position of the child.
Last but not least, we are planning to extend this cluster analysis by also considering as additional variables indoor sources of the exposure, such as the type of heating in the child's home or school, to discover how they might impact on the exposure scenario.
# Conclusions
Cluster analysis was applied on personal exposure measurements recorded 24 h for one day from children living in France. The analysis revealed four clusters of indoor ELF MF exposure characterized by significantly different patterns of underground networks, power lines, and substations near a child's home or school. Among the electric networks considered in the study, the analysis also revealed that the presence of overhead lines of high to ultra-high voltage and substations near home or school greatly contributed to differentiating the four clusters of indoor MF exposure. The present analysis seems to indicate that the presence of underground networks and low to mid voltage overhead lines within 20 ÷ 40 m from home or school did not contribute to cluster differentiation (probably because of the spatial attenuation of the B field which is not negligible at such distances). Although the present analysis revealed that children living within 20 ÷ 40 m from these latter networks were typically characterized by the lowest exposures, it might be interesting, in future measurement campaigns, to collect more samples at distances lower than 20 ÷ 40 m, i.e., at distances where the B field generated by these networks would only be partially attenuated. This would give us the chance to assess if the lowest exposures evidenced in the present study were due to the type of network, irrespective of the distance, or if they were due to a combined effect of the type of network and the distance from the measurement location.
[fig] Figure 1: Pictorial representation of the distribution of electric networks in the analyzed dataset and their interconnections. Nodes (circles) represent the electric networks. Node size is proportional to the number of children in the dataset living or going to a school near that particular electric network. [/fig]
[fig] Figure 2: Average silhouette score as a function of the number of partitions. [/fig]
[fig] Figure 3: Int. J. Environ. Res. Public Health 2019, 16, Normalized values of the centroids of the four clusters. Each panel shows the centroid coordinate scaled to the maximum of each of the 11 analyzed variables (as displayed in the panel legends). [/fig]
[fig] Figure 4: Within cluster analysis. Black bars: Percentage of measurements within each cluster from children living (or going to schools) near underground cables, overhead lines, and substations. White bars: The complement to 100%. [/fig]
[table] Table 1: Distribution of electric networks in the analyzed dataset. The table shows the number of indoor measurements close to each network type. The last two columns on the right show, for each network type, the maximum and mean number of power lines and substations near the measurement site (i.e., home or school). [/table]
[table] Table 3: Cluster size (i.e., number of measurements) for different partitioning solutions, from 2-to 6-cluster solutions. [/table]
|
A randomized controlled trial of smartphone-based mindfulness training for smoking cessation: a study protocol
Background: Tobacco use is responsible for the death of about 1 in 10 individuals worldwide. Mindfulness training has shown preliminary efficacy as a behavioral treatment for smoking cessation. Recent advances in mobile health suggest advantages to smartphone-based smoking cessation treatment including smartphone-based mindfulness training. This study evaluates the efficacy of a smartphone app-based mindfulness training program for improving smoking cessation rates at 6-months follow-up. Methods/Design: A two-group parallel-randomized clinical trial with allocation concealment will be conducted. Group assignment will be concealed from study researchers through to follow-up. The study will be conducted by smartphone and online. Daily smokers who are interested in quitting smoking and own a smartphone (n = 140) will be recruited through study advertisements posted online. After completion of a baseline survey, participants will be allocated randomly to the control or intervention group. Participants in both groups will receive a 22-day smartphone-based treatment program for smoking. Participants in the intervention group will receive mobile mindfulness training plus experience sampling. Participants in the control group will receive experience sampling-only. The primary outcome measure will be one-week point prevalence abstinence from smoking (at 6-months follow-up) assessed using carbon monoxide breath monitoring, which will be validated through smartphone-based video chat. Discussion: This is the first intervention study to evaluate smartphone-based delivery of mindfulness training for smoking cessation. Such an intervention may provide treatment in-hand, in real-world contexts, to help individuals quit smoking.
# Background
Tobacco use is the leading cause of preventable death worldwide. Although over 70% of smokers want to quit, fewer than 5% achieve this annually. Behavioral treatments teach individuals strategies to avoid smoking triggers, or to foster positive affective states, reduce negative mood, reduce stress by exercise or relaxation, divert attention from cravings, substitute other activities for smoking, or to develop social support mechanisms [bib_ref] Comparative evaluation of American Cancer Society and American Lung Association smoking cessation..., Lando [/bib_ref]. These approaches have only modest success, with 6-12 month abstinence rates of 20-30% [bib_ref] An analysis of the effectiveness of interventions intended to help people stop..., Law [/bib_ref] [bib_ref] Smoking cessation treatment: any progress?, Shiffman [/bib_ref]. Their limited success may be due to the fact that smoking triggers are always present, making avoidance difficult; diversion of attention requires cognitive resources which are often depleted after strong affective states [bib_ref] Self-regulation and depletion of limited resources: does self-control resemble a muscle?, Muraven [/bib_ref] ; and effective substitutions are not always readily available.
Recent work suggests that mindfulness training may be an effective behavioral treatment for smoking by targeting stress, craving and learned responses. Mindfulness training has been adapted as a central component of some contemporary psychotherapy and has shown some efficacy in the treatment of psychiatric disorders relating to or involving pain, anxiety, and depression [bib_ref] Mindfulness meditation practices as adjunctive treatments for psychiatric disorders, Marchand [/bib_ref] [bib_ref] Meditation programs for psychological stress and well-being: a systematic review and meta-analysis, Goyal [/bib_ref]. Mindfulness has been operationalized into two components: (i) maintaining attention on immediate experience and (ii) maintaining an attitude of acceptance toward that experience [bib_ref] Mindfulness: A proposed operational definition, Bishop [/bib_ref]. Mindfulness training typically involves the training of attention regulation, body awareness, and emotion regulation, among other components [bib_ref] How Does Mindfulness Meditation Work? Proposing Mechanisms of Action From a Conceptual..., Hölzel [/bib_ref]. In this way, mindfulness training may aid in smoking cessation by teaching individuals to pay attention to and work mindfully with affective states and craving as they arise, rather than to react by smoking.
Preliminary evidence supports testing mindfulness training in the treatment of substance-use disorders and other addictions [bib_ref] Meditation programs for psychological stress and well-being: a systematic review and meta-analysis, Goyal [/bib_ref] [bib_ref] Craving to Quit: psychological models and neurobiological mechanisms of mindfulness training as..., Brewer [/bib_ref]. A recent review found that mindfulnessbased interventions were associated with reduced consumption of substances of misuse including alcohol, cocaine, amphetamines, marijuana, cigarettes, and opiates, as compared to several active and inactive control groups [bib_ref] Are mindfulness-based interventions effective for substance use disorders? A systematic review of..., Chiesa [/bib_ref]. Moreover, several studies included in the review [bib_ref] Are mindfulness-based interventions effective for substance use disorders? A systematic review of..., Chiesa [/bib_ref] report that mindfulness training was associated with reduced craving and increased mindfulness, suggesting a potential mechanism contributing to outcomes. Although several limitations of prior studies were identified such as small sample sizes and lack of methodological details, findings overall suggest that mindfulness training may be an effective treatment for addictions. A recent randomized controlled trial (RCT) reported preliminary evidence that mindfulness training was associated with increased smoking cessation rates as compared to another leading, widely disseminated and validated smoking cessation treatment [bib_ref] Mindfulness training for smoking cessation: Results from a randomized controlled trial, Brewer [/bib_ref]. In that trial, home mindfulness practice was also found to reduce the association between craving and smoking [bib_ref] Mindfulness training for smoking cessation: Moderation of the relationship between craving and..., Elwafi [/bib_ref].
Despite these promising findings, in-person mindfulness training is challenged by the need for experienced therapists, significant demands on time, limited access to and high costs of in-person treatment delivery, and other limitations. One way to overcome these challenges is to offer mobile mindfulness training delivered by smartphone. The global use of smartphones is growing, for example nearly half of U.S. adults (age 18+, 45%) and two-thirds of young adults (age 18-29, 66%) now use smartphones [bib_ref] Two-thirds of young adults and those with higher income are smartphone owners, Rainie [/bib_ref] (young adults also have the highest smoking rates in the U.S. (32-35%). Smartphone app-based treatments for addictions are also gaining popularity due to relative low cost, ease of use, and availability [bib_ref] Mobile applications for mindfulness training in the treatment of substance-use disorders, Garrison [/bib_ref] [bib_ref] A content analysis of popular smartphone apps for smoking cessation, Abroms [/bib_ref]. This suggests the demand for and utility of a mindfulness training smartphone app for smoking cessation. In addition to improving treatment dissemination, this approach can be standardized compared to in-person training; may reduce stigma associated with seeking treatment for substance use; improves tracking of adherence such as to home meditation practice, as compared to self-report; offers direct access to tools for self-management and social support; and gets treatment into the hand of the individual in context [bib_ref] Mobile applications for mindfulness training in the treatment of substance-use disorders, Garrison [/bib_ref]. Beginning work has evaluated the efficacy of smartphone-based treatments for smoking. A recent pilot trial compared SmartQuit, an app teaching acceptance and commitment therapy, which uses mindfulness strategies, to the National Cancer Institute's QuitGuide app for smoking cessation, and found higher engagement and promising quit rates for SmartQuit [bib_ref] Randomized, controlled pilot trial of a smartphone app for smoking cessation using..., Bricker [/bib_ref]. Therefore in this study we translate a mindfulness training program that has shown preliminary efficacy for smoking cessation into a smartphone app.
The smartphone app is also embedded with experience sampling [bib_ref] Ecological momentary assessment, Shiffman [/bib_ref] to query individuals about their behavior and experience in real time to measure psychological mechanisms of change in vivo. By sampling behavior and experience in real time, it is possible to minimize recall bias, maximize ecological validity, and document change over time [bib_ref] Ecological momentary assessment, Shiffman [/bib_ref]. Tracking cigarette use in real time also avoids known biases in retrospective reports [bib_ref] Are self-reports of smoking rate biased? Evidence from the Second National Health..., Klesges [/bib_ref]. Experience sampling has been used successfully across disciplines including mindfulness [bib_ref] A Wandering Mind Is an Unhappy Mind, Killingsworth [/bib_ref] and smoking cessation [bib_ref] Using SMS text messaging to assess moderators of smoking reduction: Validating a..., Berkman [/bib_ref] , and will be utilized here to evaluate smoking, craving, mood, and mindfulness. This trial is innovative by: (i) mobile mindfulness training merges a new treatment for smoking -mindfulness training -with a novel delivery method -smartphone app; (ii) mobile mindfulness training can potentially improve efficacy by bringing mindfulness into real-world contexts; (iii) experience sampling will capture real time, real-world smoking, craving, mood, and mindfulness, by smartphone app, whereas prior studies were limited to text messaging or web-based tools [bib_ref] A Wandering Mind Is an Unhappy Mind, Killingsworth [/bib_ref] [bib_ref] Using SMS text messaging to assess moderators of smoking reduction: Validating a..., Berkman [/bib_ref] [bib_ref] Text2Quit: results from a pilot test of a personalized, interactive mobile health..., Abroms [/bib_ref] [bib_ref] A Randomized Trial of Text2Quit: A Text Messaging Program for Smoking Cessation, Abroms [/bib_ref] ; and (iv) a novel comparison group, experience sampling-only, will be used to disentangle the efficacy of mindfulness training from that of effective self-monitoring.
The primary aim is to evaluate the efficacy of mobile mindfulness training plus experience sampling to improve smoking cessation rates at 6-months follow-up as compared to experience sampling-only. It is hypothesized that mobile mindfulness training will contribute to a significantly higher smoking cessation rate at 6months follow-up as compared to control. The secondary aim is to test the hypothesis that mobile mindfulness training moderates the decoupling of craving and smoking. It is hypothesized that mobile mindfulness training will moderate the change in the prediction of smoking by craving across study time points. It is also hypothesized that meditation practice during treatment will moderate this change.
# Methods/design
A pragmatic RCT funded by the American Heart Association will be conducted. Participants will be allocated to either the control (experience sampling-only) or intervention (mobile mindfulness training plus experience sampling) group by basic randomization, using Yale Qualtrics Survey Tool (http://www.qualtrics.com/). Ethical approval for the procedures of this trial was obtained from the Yale University Institutional Review Board Human Investigation Committee (HIC #1301011337).
## Setting
The study will be conducted online and by smartphone. Treatment delivery will be smartphone-based. Participants (n = 140) will be recruited using online advertising (e.g., http://www.google.com/ads/) and directed to the study website. This recruitment method has been used to register 89 participants and randomize 16 participants per week [bib_ref] com: evaluation of a Web-based cessation program for smokeless tobacco users, Severson [/bib_ref]. This approach will have broad reach.
## Participants
Participants will be eligible if they are: age 18-65 years; smoke at least 5 cigarettes per day; less than 3 months abstinence in the past year; own an iPhone (https://www.apple.com/ iphone/) or Android (https://www.android.com/intl/en_us/); and are motivated to quit, as indicated by a score of at least 9 of 10 on the Contemplation Ladder [bib_ref] The Contemplation Ladder: validation of a measure of readiness to consider smoking..., Biener [/bib_ref] , and at least 4 of 5 on one item of the Action subscale of the Readiness to Change Questionnaire [bib_ref] Development of a short 'readiness to change' questionnaire for use in brief,..., Rollnick [/bib_ref]. "I am trying to smoke less than I used to" from 1=Strongly disagree to 5=Strongly agree" ". If smartphone ownership is found to be a barrier to recruitment, smartphones and/or voice/data plans will be provided.
## Recruitment and randomization
All study procedures will be automated using Yale Qualtrics Survey Tool. The study website links to a short screening survey. The survey notifies an individual if they are eligible or not eligible. If eligible, they are asked for their contact information and directed to an online consent form, which includes study contact information to ask questions. Those who provide consent are sent an email with a link to the baseline survey and a copy of the consent form. This email also includes a recommendation for nicotine replacement therapy or other pharmacotherapies [bib_ref] A content analysis of popular smartphone apps for smoking cessation, Abroms [/bib_ref]. At the end of the baseline survey, participants are randomized to treatment group and emailed a link to download their app and begin the study.
## Retention strategies
Treatment retention will be defined as completing 75% of mobile mindfulness training modules. Retention rates will be ensured by: (i) emphasizing the importance of follow-up at study initiation and through follow-up; (ii) requesting three referrals at study initiation who can update a participants' contact information; (iii) paying $10 for the end of treatment survey, $20 each for surveys at 3-and 6-months follow-up, and $0.50 per experience sampling "check-in" [bib_ref] Using SMS text messaging to assess moderators of smoking reduction: Validating a..., Berkman [/bib_ref] ; and (iv) paying for participation (up to $116 total) at 6-months follow-up. All randomized participants will be followed up and their true smoking status used in the analysis, and follow-up will be collected blind to group allocation [bib_ref] Outcome criteria in smoking cessation trials: proposal for a common standard, West [/bib_ref].
## Sample size
The sample size was determined based on the prior study of in-person mindfulness training for smoking cessation [bib_ref] Mindfulness training for smoking cessation: Results from a randomized controlled trial, Brewer [/bib_ref]. Cell size was determined by a χ 2 power analysis using one-week point prevalence abstinence at end of treatment and at 4-months follow-up from that study. Point prevalence abstinence differed between groups by 20% or more at each time point. Thus it is estimated that group sizes of 59 per group will achieve 80% power to detect a 20% group difference at p < .05 using a twosided Z-test with pooled variance. 140 subjects will be recruited to allow for 12-15% attrition. Further, a sample size of 33 was sufficient to detect a significant prediction of smoking by craving at baseline and significant effects of meditation on smoking at end of treatment [bib_ref] Mindfulness training for smoking cessation: Moderation of the relationship between craving and..., Elwafi [/bib_ref] , therefore these group sizes will be sufficient to detect similar changes.
## Intervention and comparator intervention group
The intervention group will receive a smartphone app, Craving to Quit (www.cravingtoquit.com) which delivers mindfulness training and experience sampling daily for 22 days. An introductory email will explain that the app will help participants self-monitor their smoking habits, recognize when and how often they smoke, identify triggers for smoking, and learn methods to become more mindful of cravings and 'ride them out'. Upon downloading the app, participants will input the number of cigarettes smoked per day and set a quit date of 3 weeks from today. They will view a tutorial explaining the app features and begin the first day.
Mobile mindfulness training is comprised of 22 modules, 5-15 minutes each, designed to teach mindfulness for smoking cessation using audio, video, animations, and in vivo exercises . Participants have access to only one new module per day; subsequent days are locked to prevent skipping ahead. Four bonus modules become available across the study to bolster training. Overall the concept of mindfulness -paying attention to and accepting momentary experience -will be taught. Participants will learn three standard meditation practices: body scan, loving kindness, and breath awareness. Body scan is practiced by bringing awareness to different parts of the body, and is considered to foster awareness of body sensations that constitute cravings and affective states. Loving kindness is practiced by directed well-wishing by repeating phrases such as "may X be happy," and is considered to foster acceptance of oneself and others. Breath awareness is practiced by paying attention to the breath wherever one feels it most strongly in the body, and is considered to help retrain the mind away from habitually engaging in self-related thinking toward a more present-centered awareness.
Participants will also be taught an informal mindfulness practice to work mindfully with cravings: Recognize, Accept, Investigate, and Note ("RAIN") what cravings feel like as they arise and pass away. RAIN is accessible from the home page of the app so that participants may use it any time they have a craving to smoke by choosing the "Want-O-Meter." They are directed to identify their smoking trigger, rate their craving, and choose between using RAIN to ride out their craving, or completing an audio-guided exercise to 'smoke mindfully' by paying attention to the moment-to-moment experience and bodily sensations of smoking.
Social support has been shown to boost quit rates [bib_ref] Initial Evaluation of a Real-World Internet Smoking Cessation System, Cobb [/bib_ref] , therefore the app is also paired with a closed online community, built on the Ning platform (www.ning.com). Features include: (1) video tutorials of how to use the app; (2) quit buddy sign ups, for individuals to go through the process with a peer; (3) forums for posting questions and insights; (4) a "Tip of the week;" and (5) an "Ask the Doc" feature to get answers from an "expert" (JB). Participants can access the community beyond their quit date in order to encourage others to quit and to bolster their own abstinence.
Additional app features include: (i) "Activity Feed" tracks interaction with the app and can be shared on the community; (ii) "My Morning Stats" tracks daily smoking to encourage a gradual taper, and daily and monthly financial savings; (iii) "Night Reflection" to reflect on daily smoking, with links to journal on the community and practice an audio-guided body scan or loving kindness meditation; (iv) "Reminders Settings" to adjust times/days to receive reminders including "Morning Introduces the Craving to Quit program and mindfulness, introduces habit formation using an animation, and provides a guided mindful smoking exercise.
2 Setting personalized goals, e.g., list three reasons why you want to quit smoking, and provides a mindful smoking exercise.
3
Teaches body scan meditation, provides a guided body scan exercise, and provides a mindful smoking exercise.
4
Teaches how to work with cues, affective states, and craving using RAIN, and provides a RAIN exercise.
## 5
Introduces the concept of craving using an animation with the metaphor of craving as a tantrum toddler, i.e., let the toddler cry it out, and provides a RAIN exercise. [bib_ref] Smoking cessation treatment: any progress?, Shiffman [/bib_ref] Teaches how to recognize triggers, e.g., list some of your triggers, and provides a RAIN exercise.
## 7
Expands on the concept of craving using an animation with the metaphor of craving as a fire, i.e., let the fire burn out, and provides a RAIN exercise.
8
Teaches how to use noting practice, i.e., the "N" of RAIN, in everyday life, and provides a noting practice exercise.
## 9
Teaches strategies for staying on track, and provides a noting exercise.
## 10
(Un)resistance Training, and provides a noting exercise.
## 11
Builds on noting practice by teaching curiosity, a core element of mindfulness, and provides a curiosity exercise.
## 12
Expands on the concept of craving and curiosity using an animation with the metaphor of a hot coal, asks "What do you get from smoking mindfully today?"
## 13
Teaches loving kindness meditation, provides a loving kindness exercise, and provides "Wild Geese" by Mary Oliver.
14 Teaches evaluating the costs & benefits of smoking, provides a loving kindness exercise.
## 15
Misperceptions about Quitting, tell a smoking buddy you are quitting today [bib_ref] Two-thirds of young adults and those with higher income are smartphone owners, Rainie [/bib_ref] Builds on noting and curiosity by teaching noting while walking meditation, provides a walking noting practice.
## 17
Teaches open awareness of thoughts, to work mindfully with thoughts that trigger smoking, using animations such as "Thoughts like a Radio."
## 18
Builds on walking while noting with animations such as "Tripping on Thoughts," "Autobiography in 5 short chapters" by Portia Nelson, provides a noting exercise.
## 19
Asks subjects to reflect on their own experience with treatment to gather evidence for how treatment helps with smoking cessation, noting practice with a particular eye out for doubt.
## 20
Provides tips on staying motivated and maintaining mindfulness practice, subjects write down a mantra to use and set mantra reminder.
## 21
Quit day ceremony, subjects tell a friend or family member that today is their quit day.
## 22
Incorporates mindfulness practices as a new, healthy habit, and instructs the user on which modules to return to if they relapse.
Bonus "Big Mind Meditation" audio by Joseph Goldstein; Tree Analogy for reinforcing noting video; Attitude is Everything video; "Mountain Meditation" audio by Joseph Goldstein; Sitting Meditation audio Stats Motivator," "Goal Reminder," "Check-ins," and "Night Reflection;" (v) "My Quit Pact" which can be personalized across the intervention and includes quit date, money saved, reasons for quitting, triggers to watch for, and a personalized mantra; and (vi) "Tracker," to track cigarettes smoked. Experience sampling is embedded in the app as "check-ins." Based on previous experience sampling methods for smoking cessation [bib_ref] In the trenches of real-world self-control: neural correlates of breaking the link..., Berkman [/bib_ref] , participants will be prompted six times per day to check-in, each day of the 22-day program. They will set the daily start and end times for check-ins, their day will be divided into six intervals, and they will be prompted to check-in at a random time in each interval [bib_ref] A Wandering Mind Is an Unhappy Mind, Killingsworth [/bib_ref]. Each check-in will ask the following questions adapted from the Day Reconstruction Method [bib_ref] A survey method for characterizing daily life experience: the day reconstruction method, Kahneman [/bib_ref] and from prior studies using experience sampling [bib_ref] A Wandering Mind Is an Unhappy Mind, Killingsworth [/bib_ref] [bib_ref] Using SMS text messaging to assess moderators of smoking reduction: Validating a..., Berkman [/bib_ref] : (1) (3) When you started this check-in, how distracted or focused were you? VAS from very distracted to very focused. (4) When you started this check-in, how was your experience feeling? VAS from very unpleasant to very pleasant. (5) When you started this check-in, how were you relating to your experience? VAS from really wanted it to stop to really wanted it to continue. (6) When you started this check-in, how much were you craving a cigarette? VAS from not at all to very much. (7) When you started this check-in, how much were you feeling these emotions? VAS from not at all to extremely, for each: relaxed, anxious, joyful, tired, content, sad, excited, irritable. (8) The tracker says you have smoked (#) cigarettes today. Adjust your tracker below if needed. Today I have smoked: (#). Participants will be sent a reminder text message if their response rate drops below 3 checkins in 24 hours. Any day with less than 3 completed check-ins will be excluded from the analysis. Based on prior studies [bib_ref] A Wandering Mind Is an Unhappy Mind, Killingsworth [/bib_ref] a response rate of 80% is expected.
Upon completion of the 22-day program, participants will receive an email of congratulations, reminding them that they will be paid for the check-ins they completed across the program 6-months from when they started the study, and that they will receive the next survey 1month from when they started the study (the dates will be provided as a reminder). Links to the online followup surveys will be emailed to participants at 1-, 3-, and 6-months. At 6-months follow-up, participants who report one-week abstinence from smoking will take part in carbon monoxide (CO) monitoring. CO levels will be measured using piCO+ Smokerlyzer breath CO-monitors (www.bedfont.com). CO-monitors will be shipped to participants with usage instructions, and they will set up a recorded, smartphone-based video call with a researcher, during which they will use the CO-monitor and show the output to the researcher [bib_ref] Internet-based contingency management to promote smoking cessation: a randomized controlled study, Dallery [/bib_ref]. They will be provided a paid shipping return label and will be asked to return the CO-monitor in its original packaging. Once a participant has completed the 6-months followup survey and returned the CO-monitor (if applicable), they will be emailed an Amazon (www.amazon.com) gift card with their study payment.
## Control group
The control group will receive experience samplingonly, to control for: (i) expectancy effects of treatment for smoking; (ii) non-specific effects of using a smartphone for smoking cessation; and (iii) potential effects of experience sampling for smoking cessation. Experience sampling may bring awareness to craving and smoking, and is similar to other interventions that teach individuals to identify triggers and track the number of cigarettes they smoke. Experience sampling-only will be delivered by smartphone app for 22 days. An introductory email will explain that the app will help individuals self-monitor their smoking habits, recognize when and how often the smoke, and identify triggers for smoking. Upon downloading the app, participants will input the number of cigarettes smoked per day and set a quit date of 3 weeks from today. They will view a tutorial explaining the app features and begin the first day. Across treatment, experience sampling procedures will be matched to the intervention group. From day 22, all study procedures will be matched to the intervention group. In addition, along with study payment at 6-months, this group will receive a free download code for the treatment app.
## Measurements
## Primary outcome measure
The primary outcome is one-week point-prevalence abstinence from tobacco smoking at 6-months, verified by carbon monoxide monitoring [bib_ref] Cut-off levels for breath carbon monoxide as a marker for cigarette smoking, Javors [/bib_ref] [bib_ref] Using breath carbon monoxide to validate self-reported tobacco smoking in remote Australian..., Maclaren [/bib_ref] with the interviewers blinded to treatment allocation. Abstinence rates will be compared between groups using χ 2 tests. This trial will use an intention-to-treat approach [bib_ref] Outcome criteria in smoking cessation trials: proposal for a common standard, West [/bib_ref].
## Secondary outcomes
The secondary outcomes are smoking, craving, mood and mindfulness as measured by experience sampling.
Smoking will be measured using the Tracker. Craving will be measured using the check-in question "When you started this check-in, how much were you craving a cigarette?" Prediction of smoking by craving will be tested using standard regression analyses with group and craving as independent variables and smoking as the dependent variable [bib_ref] Mindfulness training for smoking cessation: Moderation of the relationship between craving and..., Elwafi [/bib_ref]. Formal (meditation) and informal (RAIN) practice will be included in the model as independent predictor variables [bib_ref] Mindfulness training for smoking cessation: Moderation of the relationship between craving and..., Elwafi [/bib_ref] to test whether meditation practice during treatment moderates the change in the prediction of smoking by craving across study time points.
## Exploratory outcomes
Theory-based potential mechanisms will be tested using mediation analyses of variables related to smoking, craving, mood and mindfulness, and other psychological variables, evaluated using 15-minute online surveys at baseline, 1-, 3-, and 6-months from treatment initiation, including: (i) smoking demographics, including self report of smoking abstinence over the follow-up period allowing up to five cigarettes total [bib_ref] Outcome criteria in smoking cessation trials: proposal for a common standard, West [/bib_ref] ; (ii) 4-item Alcohol Use Disorders Identification Test (AUDIT-C; [bib_ref] The AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for..., Bush [/bib_ref] ; (iii) 15-item Minnesota Nicotine Withdrawal Scalerevised (MNWS-R; [bib_ref] Effects of abstinence from tobacco: valid symptoms and time course, Hughes [/bib_ref] to evaluate nicotine withdrawal symptoms (i.e., craving, irritability, anxiety, depression, difficulty concentrating, appetite, insomnia, restlessness, impatience, constipation, dizziness, coughing, dreaming/ nightmares, nausea, sore throat); (iv) 4-item Perceived Stress Scale (PSS; [bib_ref] A global measure of perceived stress, Cohen [/bib_ref] to measure the degree to which situations in one's life are appraised as stressful over the last week; (v) 12-item Smoking Self Efficacy Questionnaire (SEQ; [bib_ref] Development and validation of a scale measuring self-efficacy of current and former..., Etter [/bib_ref] to evaluate the self-efficacy to resist the urge to smoke; (vi) 24-item Five Facet Mindfulness Questionnaire short form (FFMQ-SF; [bib_ref] Assessment of mindfulness by self-report: the Kentucky inventory of mindfulness skills, Baer [/bib_ref] [bib_ref] Psychometric properties of the five facet mindfulness questionnaire in depressed adults and..., Bohlmeijer [/bib_ref] to measure mindfulness related to observing, describing, acting with awareness, non-judging of inner experience, and non-reactivity to inner experience; (vii) 12-item Self Compassion Scale short form (SCS-SF; [bib_ref] The Development and Validation of a Scale to Measure Self-Compassion, Neff [/bib_ref] [bib_ref] Construction and factorial validation of a short form of the Self-Compassion Scale, Raes [/bib_ref] to evaluate how one typically acts toward oneself in difficult times; (viii) 22-item Craving Experience Questionnaire (CEQ; [bib_ref] The Craving Experience Questionnaire: a brief, theory-based measure of consummatory desire and..., May [/bib_ref] to measure desire and craving to smoke; (ix) exercise, eating, and relaxation methods questions, including prior meditation experiences; and (x) 6-item Gratitude Questionnaire (GQ-6;500 [bib_ref] The grateful disposition: a conceptual and empirical topography, Mccullough [/bib_ref] to evaluate one's proneness to experience gratitude in everyday life. Participants will be paid $10 for the 1-month survey, and $20 for each 3-and 6-months survey, the total to be paid at 6-months.
# Discussion
This trial is the first to test smartphone app-based mindfulness training to for smoking cessation. Although many meditation-related smartphone apps are available, very few have been tested for use in clinical groups such as smokers [bib_ref] Mobile applications for mindfulness training in the treatment of substance-use disorders, Garrison [/bib_ref]. In addition, there are currently few smoking cessation interventions that target the association between craving and smoking [bib_ref] Craving to Quit: psychological models and neurobiological mechanisms of mindfulness training as..., Brewer [/bib_ref]. In-person mindfulness training has been found to improve smoking cessation rates and decrease the association between craving and smoking [bib_ref] Mindfulness training for smoking cessation: Results from a randomized controlled trial, Brewer [/bib_ref]. Therefore, this trial will test whether mobile mindfulness training is associated with similar outcomes.
The choice of a control group for this trial was critical for isolating and testing variables of interest, by comparing mobile mindfulness training to experience sampling-only. Alternative comparators could be in-person mindfulness training, an online smoking cessation program, waitlist control, or another smartphone-based smoking cessation intervention. However, there is limited evidence for the efficacy of existing smartphone apps for smoking cessation, and other apps offer additional features not available in the tested intervention that introduce confounds, or vice versa. Experience sampling may bring awareness to smoking, craving, and mood, similar to mindfulness training, therefore experience sampling-only was chosen as the control, to disentangle the effects of mindfulness training from those of effective self-monitoring. Exploratory analyses will test whether experience sampling-only also leads to mindfulness, and whether increases in mindfulness are associated with smoking cessation.
If the intervention is found to be effective, this RCT will provide a platform for large-scale clinical trials to compare mobile mindfulness training to active behavioral treatments for smoking cessation, and to disseminate this treatment to the wider community.
Competing interests Judson A. Brewer and Prasanta Pal own stock in Claritas Mindsciences, the company that developed the apps used in this study. All other authors declare that they have no competing interests.
Authors' contributions KG conceived of the trial, participated in the design of the trial, wrote the paper, and is carrying out the trial. PP is the Chief Developer/Programmer of the software and participated in the design of the intervention. RR participated in the design of the trial and of the intervention. JD participated in the design of the trial. SO participated in the design of the trial. JB conceived of the trial and of the intervention, and participated in the design of the trial and of the intervention. All authors edited and approved the final manuscript. |
A Pilot Study: The Beneficial Effects of Combined Statin-exercise Therapy on Cognitive Function in Patients with Coronary Artery Disease and Mild Cognitive Decline
Objective Hypercholesterolemia, a risk factor in cognitive impairment, can be treated with statins. However, cognitive decline associated with "statins" (HMG-CoA reductase inhibitors) is a clinical concern. This pilot study investigated the effects of combining statins and regular exercise on cognitive function in coronary artery disease (CAD) patients with prior mild cognitive decline. Methods We recruited 43 consecutive CAD patients with mild cognitive decline. These patients were treated with a statin and weekly in-hospital aerobic exercise for 5 months. We measured serum lipids, exercise capacity, and cognitive function using the mini mental state examination (MMSE). Results Low-density lipoprotein cholesterol levels were significantly decreased, and maximum exercise capacity (workload) was significantly increased in patients with CAD and mild cognitive decline after treatment compared with before. Combined statin-exercise therapy significantly increased the median (range) MMSE score from 24 (22-25) to 25 (23-27) across the cohort (p<0.01). Changes in body mass index (BMI) were significantly and negatively correlated with changes in the MMSE. After treatment, MMSE scores in the subgroup of patients that showed a decrease in BMI were significantly improved, but not in the BMI-increased subgroup. Furthermore, the patients already on a statin at the beginning of the trial displayed a more significant improvement in MMSE score than statin-naïve patients, implying that exercise might be the beneficial aspect of this intervention as regards cognition. In a multivariate logistic regression analysis adjusted for age >65 years, sex, and presence of diabetes mellitus, a decrease in BMI during statin-exercise therapy was significantly correlated with an increase in the MMSE score (odds ratio: 4.57, 95% confidence interval: 1.05-20.0; p<0.05). Conclusion Statin-exercise therapy may help improve cognitive dysfunction in patients with CAD and preexisting mild cognitive decline.
# Introduction
The number of dementia cases in developed countries is dramatically increasing, and rising health care costs related to dementia are a large social problem [bib_ref] Monetary costs of dementia in the United States, Hurd [/bib_ref]. Cardiovascular diseases, as well as their risk factors, are strongly associated with cognitive decline [bib_ref] Predictors of maintaining cognitive function in older adults: the Health ABC study, Yaffe [/bib_ref]. Development of effective therapeutic strategies for cognitive dysfunction in coronary artery disease (CAD) patients is therefore an urgent issue.
Statins are effective in reducing cardiovascular events and are recommended for secondary prevention in all patients who have CAD and tolerate statin therapy [bib_ref] ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular..., Stone [/bib_ref]. However, the US Food and Drug Administration and several reviews have raised concern regarding potential adverse cognitive effects associated with statins [bib_ref] The effect of HMG-CoA reductase inhibitors on cognition in patients with Alzheimer's..., Padala [/bib_ref] [bib_ref] Statins and cognitive function: a systematic review, Richardson [/bib_ref] [bib_ref] An assessment by the Statin Cognitive Safety Task Force: 2014 update, Rojas-Fernandez [/bib_ref]. Exercise therapy is recommended as a protective factor for cognition in older adults, as well as CAD patients [bib_ref] Cardiopulmonary fitness is associated with cognitive performance in patients with coronary artery..., Swardfager [/bib_ref].
We previously reported that exercise combined with statins is clinically useful and an effective treatment for CAD patients [bib_ref] Combination treatment of rosuvastatin or atorvastatin, with regular exercise improves arterial wall..., Toyama [/bib_ref] [bib_ref] Statins combined with exercise are associated with the increased renal function mediated..., Toyama [/bib_ref]. However, the effects of statin-exercise therapy on cognitive function in CAD patients have not been fully clarified.
In the present study, we examined whether or not statinexercise therapy is beneficial for managing cognitive decline in patients with CAD and pre-existing mild cognitive impairment.
# Materials and methods
## Study population
We performed a prospective observational study to investigate the clinical significance of regular exercise and statin treatment in CAD patients with mild cognitive decline. We recruited 43 consecutive CAD patients with mild cognitive decline (mini mental state examination [MMSE] score <25). All patients were referred to our hospital for low-density lipoprotein (LDL) cholesterol-lowering therapy with statins in order to achieve guideline goals and underwent cardiac rehabilitation therapy at the Division of Cardiology, Hitoyoshi Medical Center between April 2011 and August 2014. Patients who could not perform cardiopulmonary exercises or undergo MMSE tests during the study period were excluded.
## Study design
The patients performed weekly 30-minute in-hospital aerobic exercises on ergometer-equipped bicycles. The aerobic exercises were based on anaerobic thresholds that were evaluated with cardiopulmonary exercise tests. Patients performed the exercises for 5 months. An experienced registered exercise rehabilitation instructor instructed all of the patients regarding the performance of 30-minute walking or cycle ergometer exercises at home to a Borg scale rating of 12-13. All of the patients were treated with one of a variety of statins to achieve guideline goal levels for LDL (<70 mg/ dL). Thirty statin-naive subjects were started on a de novo statin. Ten patients with pre-existing statin treatment remained on their current dose and statin, two patients were switched to similar-intensity therapy, and one patient had their dose increased. The protocol of the Hitoyoshi Cardiovascular Rehabilitation Study (HARVEST), which is evaluating the efficacy of exercise therapy on patients with cardiovascular disease and is registered with the UMIN protocol registration system (identification number UMIN 000015544; https://upload.umin.ac.jp/cgi-open-bin/ctr/ctr.cgi? function=brows&action=brows&recptno=R000018044&type =summary&language=E), was approved by the Institutional Review Board of the Hitoyoshi Medical Center in Kumamoto, Japan. A signed consent form was obtained from each patient. The methods in this study were performed in accordance with approved guidelines.
## Home exercise instructions
All of the patients were instructed to perform exercises at home under the supervision of a registered exercise rehabilitation instructor, as previously described [bib_ref] Rosuvastatin combined with regular exercise preserves coenzyme Q10 levels associated with a..., Toyama [/bib_ref]. The patients maintained personal healthcare logs in which they recorded their daily blood pressure, body weight, and exercises that were performed at home. Patients who did not complete the home exercises received repeat instruction and were asked again to perform the training.
## Blood biochemistry
Venous blood samples were collected before the exercise test.
## Cognitive assessment
The MMSE test, a brief 30-point questionnaire, was conducted by a single trained technician at baseline and at the end of treatment. Subjects who scored below 25 were considered to have mild cognitive impairment.
## Cardiopulmonary exercise test
An individually optimized 5-, 10-, 15-, and 20-Watt rampup exercise protocol on an electromagnetically braked cycle ergometer (Well Bike BE-250; Cat Eye Co., Osaka, Japan) with blood pressure and heart rate monitoring was performed in a temperature-controlled room (23-25°C), as previously described [bib_ref] Rosuvastatin combined with regular exercise preserves coenzyme Q10 levels associated with a..., Toyama [/bib_ref]. Respiratory gas was sampled from a mouthpiece, and minute ventilation, carbon dioxide in expired air, oxygen in inspired air, ventilatory equivalent for carbon dioxide, and ventilatory equivalent for oxygen were measured for each breath using a breath-by-breath gas analyzer (AE310S; Minato Medical Science, Osaka, Japan). Peak oxygen uptake, anaerobic threshold, and maximum workload were evaluated at baseline and at the end of treatment.
## Statistical analyses
We used the Shapiro-Wilk test to determine whether the samples were normally distributed. Depending on the distribution of data, the paired Student's t-test or Wilcoxon test was used to analyze the effects of 5 months' treatment. Pearson or Spearman correlation tests were used to analyze correlations between changes in the MMSE and changes in other parameters. To determine the independent association between changes in the MMSE and changes in body mass index (BMI), we adjusted for the effects of other covariates on the cognitive function with multivariate linear regression analysis with forced inclusion of the variables of age, sex, and the presence of hypertension, diabetes mellitus, and cerebral stroke. Univariate and multivariate logistic regression analyses were used to assess the baseline parameters, risk factors, and medications that were correlated with an elevation in MMSE score (ΔMMSE = [MMSE score after 5 months] -[baseline MMSE score]; cutoff value of ΔMMSE = 1.0, median). There were a total of four missing data-points for HbA1c in non-diabetic patients (two missing baselines, and two separate patients missing 5-month data). The data are expressed as the mean±standard deviation, or median value. A value of p<0.05 was considered statistically significant. All statistical analyses were performed using the IBM SPSS Statistics software program, Version 21.0 (IBM Corp., Armonk, NY, USA).
# Results
## Patients' characteristics
The patients' baseline characteristics are shown in Table 1. Forty-three CAD patients (aged 71±8 years, 81% male) were included.
## Effect of statin-exercise therapy on the cognitive function and other parameters
Five months of combined statin-exercise therapy significantly improved the cognitive decline on average, as measured by the MMSE test . [fig_ref] Table 2: Effects of Statin-exercise Treatment on Various Parameters [/fig_ref] shows the effects of statin-exercise therapy on various parameters. The cardiopulmonary function, as measured by the maximum workload capacity, significantly improved after treatment. The average total cholesterol and LDL cholesterol levels were significantly decreased after treatment compared with before treatment.
## Effect of pre-treatment statin and statin types on cognitive outcome
Thirteen patients were on pre-existing statin therapy at the beginning of the statin-exercise therapy period. Overall, this subgroup experienced little-to-no changes in their statin regimen at the initiation of the trial. These patients displayed a significant increase in the MMSE score (p<0.009), which was not observed in the statin-naïve patients
## Baseline diabetes mellitus and improvement of the cognitive function
Fifteen diabetic CAD patients were included in this study. Six diabetic patients were taking oral glycemic agents, while four separate patients were taking insulin [fig_ref] Table 1: Patients' Characteristics [/fig_ref]. While the anti-diabetic agents might have had effects on BMI and cognition, a univariate logistic regression analysis showed that neither oral hypoglycemic agents nor insulin correlated with an increase in the MMSE score [fig_ref] Table 3: Logistic Regression Analysis for Elevation of Mini Mental State Examination Scores among... [/fig_ref] or a decrease in the BMI (odds ratio: 1.85, p=0.489; and 0.53, p=0.601, respectively). Further, although the baseline fasting blood glucose level, HbA1c level, and oral hypoglycemic agents were not associated with cognitive response [fig_ref] Table 3: Logistic Regression Analysis for Elevation of Mini Mental State Examination Scores among... [/fig_ref] , a univariate logistic regression analysis showed that the presence of diabetes mellitus (DM) was significantly negatively correlated with an increase in the MMSE score (odds ratio: 0.24, p<0.05). The MMSE in non-diabetic CAD patients significantly increased after treatment compared with before treatment, with no marked changes in the diabetic patients .
## Effects of statin-exercise therapy-associated changes in bmi on cognitive decline
Although the average BMI did not change markedly across the study group [fig_ref] Table 2: Effects of Statin-exercise Treatment on Various Parameters [/fig_ref] , changes in the BMI were [fig_ref] Table 4: Correlations between Changes in the Mini Mental State Examination Score and Each... [/fig_ref]. After adjusting for age, sex, and the presence of hypertension, DM, and cerebral stroke, a multivariate linear regression analysis showed that changes in the BMI were still significantly associated with changes in the MMSE (Table 5). Furthermore, there was a significant inter-subgroup difference in changes in the BMI between the MMSEincreased (ΔMMSE >1.0) and the MMSE-decreased groups (ΔMMSE ! 1.0) [fig_ref] Table 6: Effects of Statin-exercise Treatment on Various Parameters in the Mini Mental State... [/fig_ref].
The MMSE in the BMI-decreased group (ΔBMI = [BMI after 5 months] -[baseline BMI]; ΔBMI <0.0) was significantly elevated after treatment compared with before treat- , while the abdominal circumference in the BMI-increased group did not change markedly, that of the BMI-decreased group decreased significantly . However, univariate logistic regression and linear regression analyses showed that the abdominal circumference was not associated with the MMSE score improvement [fig_ref] Table 3: Logistic Regression Analysis for Elevation of Mini Mental State Examination Scores among... [/fig_ref]. Additionally, while the average baseline renal function and left ventricular ejection fraction in the BMI-increased patients were significantly lower than in the BMI-decreased patients, a univariate logistic regression analysis showed no association between these variables and the MMSE score [fig_ref] Table 3: Logistic Regression Analysis for Elevation of Mini Mental State Examination Scores among... [/fig_ref]. Finally, an analysis showed that both a decrease in the BMI, and the absence of baseline DM were significantly correlated with an elevation in the MMSE score in response to statin-exercise therapy (BMI, odds ratio: 3.74, p<0.05; DM, 0.24, p<0.05, Table 3). In multivariate logistic regression analysis adjusted for age >65 years of age, sex, and the presence of DM, a decrease in the BMI with statin-exercise therapy remained significantly correlated with an elevation in the MMSE score (odds ratio: 4.57, p<0.05, [fig_ref] Table 3: Logistic Regression Analysis for Elevation of Mini Mental State Examination Scores among... [/fig_ref].
# Discussion
The major finding of the present pilot study is that a combination of statins and regular exercise therapy appeared to significantly ameliorate cognitive dysfunction in patients already displaying mild cognitive dysfunction. This beneficial effect correlated with treatment-associated decrease in the BMI in CAD patients with mild cognitive decline as well as in those with pre-existing statin therapy.
Statins are an established therapy for reducing the risk of cardiovascular events. We previously reported the beneficial effects of statin-exercise treatment on CAD [bib_ref] Combination treatment of rosuvastatin or atorvastatin, with regular exercise improves arterial wall..., Toyama [/bib_ref] [bib_ref] Statins combined with exercise are associated with the increased renal function mediated..., Toyama [/bib_ref]. Recently, perhaps because dietary control in statin users has been less emphasized as a therapeutic approach, high calorie and fat intake has been observed in many statin users, accompanied by higher BMI (vs. non-statin users) [bib_ref] Different time trends of caloric and fat intake between statin users and..., Sugiyama [/bib_ref]. High-calorie diets (and to a lesser extent fat intake) have a well-established role in the development of obesity and diabetes, which inturn are risk factors associated with the pathogenesis of cognitive decline [bib_ref] Obesity and vascular risk factors at midlife and the risk of dementia..., Kivipelto [/bib_ref]. The potential adverse effects of statins on the cognitive function are a concerning issue. Notably, some of the beneficial effects of exercise on the MMSE in our cohort were absent in the cohort that started taking a statin de novo.
Exercise is known to exert beneficial effects on cognition [bib_ref] Effects of aerobic exercise on mild cognitive impairment: a controlled trial, Baker [/bib_ref] [bib_ref] Aerobic exercise effects on cognitive and neural plasticity in older adults, Erickson [/bib_ref] ; indeed, a very recent meta-analysis showed that aerobic exercise had a significant protective effect on cognitive dysfunction in older adults with mild cognitive impairment [bib_ref] Aerobic exercise ameliorates cognitive function in older adults with mild cognitive impairment:..., Zheng [/bib_ref]. However, weight loss per se is often seen in the course of dementia, particularly in Alzheimer's dis- ease [bib_ref] Weight loss and Alzheimer's disease: temporal and aetiologic connections, Sergi [/bib_ref]. In fact, low BMI as well as weight-loss were the predominant risk factors for mild cognitive impairment-todementia conversion in aged subjects with cognitive decline [bib_ref] Body mass index and mild cognitive impairment-to-dementia progression in 24 months: a..., Sobow [/bib_ref]. This likely relates to a diminished appetite and not increased physical activity. To prevent further progression of dementia in CAD patients with cognitive decline, we suspect that body weight management through exercise during statin therapy may be crucial in order for patients to experience maximal benefit.
The findings of this study could lead to novel and practical therapeutic approaches to prevent development and/or limit progression of cognitive decline in CAD patients. However, to confirm the importance of combined statinexercise therapy, further large prospective clinical studies are required; for example, comparing the cognitive outcomes among independent groups receiving exercise or statin, a statin-exercise combination group, and a non-exercise (control) group. Standardization of statin dosing and type are also crucial, as the present pilot study suggests that the effects of the combination of certain statin therapy types with exercise in CAD patients should be evaluated. The MMSE score in lipophilic statins users increased significantly over the study period, while hydrophilic statin users saw no marked change. Hydrophilic statins are less permeable across the blood-brain barrier than lipophilic statins in model systems [bib_ref] Functional characterization of mouse organic anion transporting peptide 1a4 in the uptake..., Ose [/bib_ref] , raising the question of what impact the hydrophilicity of a given statin might have on brain disorders and/or healthy brain. Interestingly, Rojas-Fernandez et al. concluded in a systemic review that hydrophilic statins (pravastatin and rosuvastatin) might be less likely to contribute to cognitive impairment, due to their limited penetration across the blood-brain barrier [bib_ref] Is statin-associated cognitive impairment clinically relevant? A narrative review and clinical recommendations, Rojas-Fernandez [/bib_ref]. These elements should be assessed in combination with exercise in future investigations of the cognitive function.
Our study has several limitations. First, our study design prevented effective confirmation of which element of the in- . Effects of Statin-exercise Treatment on Various Parameters in the Body Mass Index-decreased and -increased Groups.
[fig] Figure: The effect of statin-exercise therapy on cognition in CAD patients with mild cognitive decline. (A) Box-and-whisker plot showing the effect of combined exercise and statin treatment for 5 months on the cognitive function using the MMSE test. (B) Box-and-whisker plot showing the effect of statin-exercise therapy on the MMSE score between the statin-naïve group and pre-existing statin groups. (C) Box-and-whisker plot showing the effect of combined exercise and statin therapy for 5 months on the MMSE scores between patients without DM (non-DM) and those with DM. (D) A regression analysis showing the correlation between changes in the BMI and changes in the MMSE score. (E) Box-and-whisker plot showing the effect of statin-exercise therapy on the MMSE score between the BMI-increased and BMI-decreased groups. (A, B, C, and E) In these box-and-whisker plots, the lines within the boxes represent the median values, the upper and lower lines of the boxes represent the 25th and 75th percentiles, respectively, and the upper and lower bars outside the boxes represent the 90th and 10th percentiles, respectively. MMSE: mini mental state examination, BMI: body mass index, 5 mo: 5 months, n: number of patients in each group Changes in BMI (kg/m 2 ) Changes in MMSE (score) correlated with changes in the MMSE (Figure D, [/fig]
[table] Table 1: Patients' Characteristics. [/table]
[table] Table 2: Effects of Statin-exercise Treatment on Various Parameters. [/table]
[table] Table 3: Logistic Regression Analysis for Elevation of Mini Mental State Examination Scores among Baseline Parameters. [/table]
[table] Table 4: Correlations between Changes in the Mini Mental State Examination Score and Each Parameter. [/table]
[table] Table 5: Multiple Regression Analysis of Changes in the Mini Mental State Examination. [/table]
[table] Table 6: Effects of Statin-exercise Treatment on Various Parameters in the Mini Mental State Examination Score-decreased and -increased Groups. 58.8 (46.4 to 64.1) 58.2 (48.3 to 66.7) 59.1 (54.1 to 68.9) 60.6 (50.0 to 71.5)Data are mean ± SD, or median (25% to 75%). Mini Mental State Examination Score-decreased ( E 1.0) and -increased ( E >1.0) Groups *p < 0.05, †p < 0.01 compared with the respective baseline values; ‡ p < 0.05 compared with the Mini Mental State Examination Score-decreased Group. [/table]
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.