Glioblastoma Surgery Imaging-Reporting and Data System: Validation and Performance of the Automated Segmentation Task

For patients with presumed glioblastoma, essential tumor characteristics are determined from preoperative MR images to optimize the treatment strategy. This procedure is time-consuming and subjective, if performed by crude eyeballing or manually. The standardized GSI-RADS aims to provide neurosurgeons with automatic tumor segmentations to extract tumor features rapidly and objectively. In this study, we improved automatic tumor segmentation and compared the agreement with manual raters, describe the technical details of the different components of GSI-RADS, and determined their speed. Two recent neural network architectures were considered for the segmentation task: nnU-Net and AGU-Net. Two preprocessing schemes were introduced to investigate the tradeoff between performance and processing speed. A summarized description of the tumor feature extraction and standardized reporting process is included. The trained architectures for automatic segmentation and the code for computing the standardized report are distributed as open-source and as open-access software. Validation studies were performed on a dataset of 1594 gadolinium-enhanced T1-weighted MRI volumes from 13 hospitals and 293 T1-weighted MRI volumes from the BraTS challenge. The glioblastoma tumor core segmentation reached a Dice score slightly below 90%, a patientwise F1-score close to 99%, and a 95th percentile Hausdorff distance slightly below 4.0 mm on average with either architecture and the heavy preprocessing scheme. A patient MRI volume can be segmented in less than one minute, and a standardized report can be generated in up to five minutes. The proposed GSI-RADS software showed robust performance on a large collection of MRI volumes from various hospitals and generated results within a reasonable runtime.

Prediction of ventricular arrhythmia in phospholamban p.Arg14del mutation carriers-reaching the frontiers of individual risk prediction

Aims

This study aims to improve risk stratification for primary prevention implantable cardioverter defibrillator (ICD) implantation by developing a new mutation-specific prediction model for malignant ventricular arrhythmia (VA) in phospholamban (PLN) p.Arg14del mutation carriers. The proposed model is compared to an existing PLN risk model.

Methods and results

Data were collected from PLN p.Arg14del mutation carriers with no history of malignant VA at baseline, identified between 2009 and 2020. Malignant VA was defined as sustained VA, appropriate ICD intervention, or (aborted) sudden cardiac death. A prediction model was developed using Cox regression. The study cohort consisted of 679 PLN p.Arg14del mutation carriers, with a minority of index patients (17%) and male sex (43%), and a median age of 42 years [interquartile range (IQR) 27–55]. During a median follow-up of 4.3 years (IQR 1.7–7.4), 72 (10.6%) carriers experienced malignant VA. Significant predictors were left ventricular ejection fraction, premature ventricular contraction count/24 h, amount of negative T waves, and presence of low-voltage electrocardiogram. The multivariable model had an excellent discriminative ability {C-statistic 0.83 [95% confidence interval (CI) 0.78–0.88]}. Applying the existing PLN risk model to the complete cohort yielded a C-statistic of 0.68 (95% CI 0.61–0.75).

Conclusion

This new mutation-specific prediction model for individual VA risk in PLN p.Arg14del mutation carriers is superior to the existing PLN risk model, suggesting that risk prediction using mutation-specific phenotypic features can improve accuracy compared to a more generic approach.

Development and external validation of prediction models to predict implantable cardioverter-defibrillator efficacy in primary prevention of sudden cardiac death

Aims

This study was performed to develop and externally validate prediction models for appropriate implantable cardioverter-defibrillator (ICD) shock and mortality to identify subgroups with insufficient benefit from ICD implantation.

Methods and results

We recruited patients scheduled for primary prevention ICD implantation and reduced left ventricular function. Bootstrapping-based Cox proportional hazards and Fine and Gray competing risk models with likely candidate predictors were developed for all-cause mortality and appropriate ICD shock, respectively. Between 2014 and 2018, we included 1441 consecutive patients in the development and 1450 patients in the validation cohort. During a median follow-up of 2.4 (IQR 2.1–2.8) years, 109 (7.6%) patients received appropriate ICD shock and 193 (13.4%) died in the development cohort. During a median follow-up of 2.7 (IQR 2.0–3.4) years, 105 (7.2%) received appropriate ICD shock and 223 (15.4%) died in the validation cohort. Selected predictors of appropriate ICD shock were gender, NSVT, ACE/ARB use, atrial fibrillation history, Aldosterone-antagonist use, Digoxin use, eGFR, (N)OAC use, and peripheral vascular disease. Selected predictors of all-cause mortality were age, diuretic use, sodium, NT-pro-BNP, and ACE/ARB use. C-statistic was 0.61 and 0.60 at respectively internal and external validation for appropriate ICD shock and 0.74 at both internal and external validation for mortality.

Conclusion

Although this cohort study was specifically designed to develop prediction models, risk stratification still remains challenging and no large group with insufficient benefit of ICD implantation was found. However, the prediction models have some clinical utility as we present several scenarios where ICD implantation might be postponed.

The circulatory small non-coding RNA landscape in community-acquired pneumonia on intensive care unit admission

Community‐acquired pneumonia (CAP) is a major cause of sepsis. Despite several clinical trials targeting components of the inflammatory response, no specific treatment other than antimicrobial therapy has been approved. This argued for a deeper understanding of sepsis immunopathology, in particular factors that can modulate the host response. Small non‐coding RNA, for example, micro (mi)RNA, have been established as important modifiers of cellular phenotypes. Notably, miRNAs are not exclusive to the intracellular milieu but have also been detected extracellular in the circulation with functional consequences. Here, we sought to determine shifts in circulatory small RNA levels of critically ill patients with CAP‐associated sepsis and to determine the influence of clinical severity and causal pathogens on small RNA levels. Blood plasma was collected from 13 critically ill patients with sepsis caused by CAP on intensive care unit admission and from 5 non‐infectious control participants. Plasma small RNA‐sequencing identified significantly altered levels of primarily mature miRNAs in CAP relative to controls. Pathways analysis of high or low abundance miRNA identified various over‐represented cellular biological pathways. Analysis of small RNA levels against common clinical severity and inflammatory parameters indices showed direct and indirect correlations. Additionally, variance of plasma small RNA levels in CAP patients may be explained, at least in part, by differences in causal pathogens. Small nuclear RNA levels were specifically altered in CAP due to Influenza infection in contrast to Streptococcus pneumoniae infection. Pathway analysis of plasma miRNA signatures unique to Influenza or Streptococcus pneumoniae infections showed enrichment for specific proteoglycan, cell cycle, and immunometabolic pathways.

Abnormal aortic hemodynamics at predilection sites for dissection in patients with Marfan Syndrome: a 4D flow study

Funding Acknowledgements

Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): AMC Foundation Horstingstuit Foundation

Introduction

Patients with Marfan syndrome (MFS) may develop aneurysmatic dilatation and dissection of the aorta with a consequence of sudden death at relatively young age. We performed an aortic 4D flow MRI analysis, providing a comprehensive quantification and visualization of abnormal aortic velocity and wall shear stress (WSS) magnitude and direction with recently developed techniques (1,2). We hypothesize that abnormal hemodynamics are found at predilection sites for aortic dissection in MFS patients.

Methods

This prospective study included 56 MFS patients and 25 healthy subjects as controls. Aortic 4D flow MRI was performed on a 3T Philips Ingenia system (Best, Netherlands). The aorta was manually segmented on time-averaged phase contrast MR angiogram images (phase contrast images multiplied by absolute velocity) by thresholding, watershed, and manual voxel in-/exclusion. The segmentations were used to mask the velocities, calculate WSS, and co-registration for quantification of abnormal hemodynamics (3).

Abnormally elevated velocity and WSS were defined as higher than the three-dimensional 95% confidence interval as determined in the control group. Abnormally directed velocity and WSS were defined as vector angle differences higher than 120°. The aorta was subdivided in six regions of interest (ROIs) for total multiple linear regression with age, aortic diameter, and blood pressure characteristics. Independent predictors were defined as characteristics that were significant in the total model. Significance was defined as p < 0.05 with Bonferonni correction. The 3D-maps with abnormal hemodynamics were co-registered and added to create 3D-maps that show the incidence of abnormal hemodynamics.

Results

Figure 1 shows examples of maps with abnormal velocity and WSS magnitude and direction respectively. Ascending elevated velocity was associated with age, aortic diameter and blood pressure characteristics, whereas elevated WSS was associated with blood pressure characteristics only. No independent predictors were found for abnormally directed hemodynamics. Figure 2 shows the incidence maps for abnormally elevated velocity and abnormally directed WSS in two patients. The maximum incidence for elevated velocity and WSS were 32% and 20%, respectively, and found in the ascending aorta. The maxima for abnormally directed velocity and WSS were 18% and 39%, respectively, and found in the inner proximal descending aorta.

Conclusion

Altered aortic geometry and wall properties in MFS patients cause detectable hemodynamic effects in 30% of our cohort at known predilection sites for aortic dissection in MFS patients: the ascending aorta and proximal descending aorta. Independent measures of altered hemodynamics could possibly indicate individual patients at risk for aortic dissection.

On the cutting edge of glioblastoma surgery: where neurosurgeons agree and disagree on surgical decisions

OBJECTIVE

The aim of glioblastoma surgery is to maximize the extent of resection while preserving functional integrity. Standards are lacking for surgical decision-making, and previous studies indicate treatment variations. These shortcomings reflect the need to evaluate larger populations from different care teams. In this study, the authors used probability maps to quantify and compare surgical decision-making throughout the brain by 12 neurosurgical teams for patients with glioblastoma.

METHODS

The study included all adult patients who underwent first-time glioblastoma surgery in 2012-2013 and were treated by 1 of the 12 participating neurosurgical teams. Voxel-wise probability maps of tumor location, biopsy, and resection were constructed for each team to identify and compare patient treatment variations. Brain regions with different biopsy and resection results between teams were identified and analyzed for patient functional outcome and survival.

RESULTS

The study cohort consisted of 1087 patients, of whom 363 underwent a biopsy and 724 a resection. Biopsy and resection decisions were generally comparable between teams, providing benchmarks for probability maps of resections and biopsies for glioblastoma. Differences in biopsy rates were identified for the right superior frontal gyrus and indicated variation in biopsy decisions. Differences in resection rates were identified for the left superior parietal lobule, indicating variations in resection decisions.

CONCLUSIONS

Probability maps of glioblastoma surgery enabled capture of clinical practice decisions and indicated that teams generally agreed on which region to biopsy or to resect. However, treatment variations reflecting clinical dilemmas were observed and pinpointed by using the probability maps, which could therefore be useful for quality-of-care discussions between surgical teams for patients with glioblastoma.

Glioblastoma Surgery Imaging-Reporting and Data System: Standardized Reporting of Tumor Volume, Location, and Resectability Based on Automated Segmentations

Treatment decisions for patients with presumed glioblastoma are based on tumor characteristics available from a preoperative MR scan. Tumor characteristics, including volume, location, and resectability, are often estimated or manually delineated. This process is time consuming and subjective. Hence, comparison across cohorts, trials, or registries are subject to assessment bias. In this study, we propose a standardized Glioblastoma Surgery Imaging Reporting and Data System (GSI-RADS) based on an automated method of tumor segmentation that provides standard reports on tumor features that are potentially relevant for glioblastoma surgery. As clinical validation, we determine the agreement in extracted tumor features between the automated method and the current standard of manual segmentations from routine clinical MR scans before treatment. In an observational consecutive cohort of 1596 adult patients with a first time surgery of a glioblastoma from 13 institutions, we segmented gadolinium-enhanced tumor parts both by a human rater and by an automated algorithm. Tumor features were extracted from segmentations of both methods and compared to assess differences, concordance, and equivalence. The laterality, contralateral infiltration, and the laterality indices were in excellent agreement. The native and normalized tumor volumes had excellent agreement, consistency, and equivalence. Multifocality, but not the number of foci, had good agreement and equivalence. The location profiles of cortical and subcortical structures were in excellent agreement. The expected residual tumor volumes and resectability indices had excellent agreement, consistency, and equivalence. Tumor probability maps were in good agreement. In conclusion, automated segmentations are in excellent agreement with manual segmentations and practically equivalent regarding tumor features that are potentially relevant for neurosurgical purposes. Standard GSI-RADS reports can be generated by open access software.

Electrocardiogram-based mortality prediction in patients with COVID-19 using machine learning

Funding Acknowledgements

Type of funding sources: Public hospital(s). Main funding source(s): The Netherlands Organisation for Health Research and Development (ZonMw) University of Amsterdam Research Priority Area Medical Integromics

OnBehalf

CAPACITY-COVID19 Registry

Background

The electrocardiogram (ECG) is an easy to assess, widely available and inexpensive tool that is frequently used during the work-up of hospitalized COVID-19 patients. So far, no study has been conducted to evaluate if ECG-based machine learning models are able to predict all-cause in-hospital mortality in COVID-19 patients.

Purpose

With this study, we aim to evaluate the value of using the ECG to predict in-hospital all-cause mortality of COVID-19 patients by analyzing the ECG at hospital admission, comparing a logistic regression based approach and a DNN based approach. Secondly, we aim to identify specific ECG features associated with mortality in patients diagnosed with COVID-19. 

Methods and results

We studied 882 patients admitted with COVID-19 across seven hospitals in the Netherlands. Raw-format 12-lead ECGs recorded after admission (<72 hours) were collected, manually assessed, and annotated using pre-defined ECG features. Using data from five out of seven centers (n = 634), two mortality prediction models were developed: (a) a logistic regression model using manually annotated ECG features, and (b) a pre-trained deep neural network (DNN) using the raw ECG waveforms. Data from two other centers (n = 248) were used for external validation. Performance of both prediction models was similar, with a mean area under the receiver operating curve of 0.69 [95%CI 0.55–0.82] for the logistic regression model and 0.71 [95%CI 0.59–0.81] for the DNN in the external validation cohort. After adjustment for age and sex, ventricular rate (OR 1.13 [95% CI 1.01–1.27] per 10 ms increase), right bundle branch block (3.26 [95% CI 1.15–9.50]), ST-depression (2.78 [95% CI 1.03–7.70]) and low QRS voltages (3.09 [95% CI 1.02-9.38]) remained as significant predictors for mortality.

Conclusion

This study shows that ECG-based prediction models at admission may be a valuable addition to the initial risk stratification in admitted COVID-19 patients. The DNN model showed similar performance to the logistic regression that needs time-consuming manual annotation. Several ECG features associated with mortality were identified.

Figure 1: 

Overview of methods, using and example case: (left) logistic regression and (right) deep learning. This specific case had a high probability of in-hospital mortality (above the threshold of 30%). Follow-up of this case showed that the patient had died during admission.

Preoperative chemoradiotherapy to improve overall survival in pancreatic cancer: Long-term results of the multicenter randomized phase III PREOPANC trial

4016

Background: Preoperative chemoradiotherapy (CRT) may improve overall survival in resectable pancreatic cancer (RPC) and borderline resectable pancreatic cancer (BRPC). Long term results are presented. Methods: In this multicenter phase III trial, patients with RPC or BRPC were randomized between preoperative CRT, (gemcitabine 1000 mg/m2 weekly for 7 of 10 weeks, and 15x2.4 Gy radiotherapy in week 4 to 6), followed by surgery and four cycles of adjuvant gemcitabine (1000 mg/m2 weekly for 3 of 4 weeks), or immediate surgery followed by 6 cycles of adjuvant gemcitabine (1000 mg/m2 weekly for 3 of 4 weeks). Primary endpoint was overall survival by intention-to-treat (ITT). Results: From April 2013 to July 2017, 246 eligible patients were accrued by 16 Dutch centers and randomized, 119 to preoperative CRT and 127 to immediate surgery. After a median follow up of 56 months (35.3-92.0 months), 210 patients have died, 93 (78%) in the preoperative CRT arm and 117 (92%) in the immediate surgery arm. Three- and five-year overall survival ITT was 27.7% and 20.5% after preoperative CRT versus 16.5% and 6.5% after immediate surgery (HR 0.73; 95% CI 0.56 to 0.96; p = 0.025). In addition, disease-free survival (HR 0.70; p = 0.009) locoregional failure-free interval (HR 0.57; p = 0.004) and distant metastases free interval (HR 0.74; p = 0.071) improved after preoperative CRT. Also in the stratified subsets RPC and BRPC, preoperative CRT improved OS: RPC (n = 133, HR 0.79; 95% CI 0.54 to 1.16; P = 0.23). BRPC (n = 113, HR 0.67; 95% CI 0.45 to 0.99; p = 0.045). We could not demonstrate a difference in treatment effect between these subsets (interaction test p = 0.56). Conclusions: Preoperative gemcitabine-based CRT for RPC or BRPC improves long term overall survival compared to immediate surgery with adjuvant gemcitabine. Clinical trial information: NTR3709.

Associations between gut microbiota, faecal short-chain fatty acids, and blood pressure across ethnic groups: the HELIUS study

Aims 

Preliminary evidence from animal and human studies shows that gut microbiota composition and levels of microbiota-derived metabolites, including short-chain fatty acids (SCFAs), are associated with blood pressure (BP). We hypothesized that faecal microbiota composition and derived metabolites may be differently associated with BP across ethnic groups.

Methods and results 

We included 4672 subjects (mean age 49.8 ± 11.7 years, 52% women) from six different ethnic groups participating in the HEalthy Life In an Urban Setting (HELIUS) study. The gut microbiota was profiled using 16S rRNA gene amplicon sequencing. Associations between microbiota composition and office BP were assessed using machine learning prediction models. In the subgroups with the largest associations, faecal SCFA levels were compared in 200 subjects with lower or higher systolic BP. Faecal microbiota composition explained 4.4% of the total systolic BP variance. Best predictors for systolic BP included Roseburia spp., Clostridium spp., Romboutsia spp., and Ruminococcaceae spp. Explained variance of the microbiota composition was highest in Dutch subjects (4.8%), but very low in South-Asian Surinamese, African Surinamese, Ghanaian, Moroccan and Turkish descent groups (explained variance <0.8%). Faecal SCFA levels, including acetate (P < 0.05) and propionate (P < 0.01), were lower in young Dutch participants with low systolic BP.

Conclusions 

Faecal microbiota composition is associated with BP, but with strongly divergent associations between ethnic groups. Intriguingly, while Dutch participants with lower BP had higher abundances of several SCFA-producing microbes, they had lower faecal SCFA levels. Intervention studies with SCFAs could provide more insight in the effects of these metabolites on BP.

Long-term clinical outcomes of losartan in patients with Marfan syndrome: follow-up of the multicentre randomized controlled COMPARE trial

AIMS

The COMPARE trial showed a small but significant beneficial effect of 3-year losartan treatment on aortic root dilatation rate in adults with Marfan syndrome (MFS). However, no significant effect was found on clinical endpoints, possibly due to a short follow-up period. The aim of the current study was therefore to investigate the long-term clinical outcomes after losartan treatment.

METHODS AND RESULTS

In the original COMPARE study (inclusion 2008-2009), adult patients with MFS (n = 233) were randomly allocated to either the angiotensin-II receptor blocker losartan® on top of regular treatment (β-blockers in 71% of the patients) or no additional medication. After the COMPARE trial period of 3 years, study subjects chose to continue their losartan medication or not. In a median follow-up period of 8 years, 75 patients continued losartan medication, whereas 78 patients, originally allocated to the control group, never used losartan after inclusion. No differences existed between baseline characteristics of the two groups except for age at inclusion [losartan 34 (interquartile range, IQR 26-43) years, control 41 (IQR 30-52) years; P = 0.031], and β-blocker use (losartan 81%, control 64%; P = 0.022). A pathological FBN1 mutation was present in 76% of patients and 58% of the patients were male. Clinical endpoints, defined as all-cause mortality, aortic dissection/rupture, elective aortic root replacement, reoperation, and vascular graft implantation beyond the aortic root, were compared between the two groups. A per-patient composite endpoint was also analysed. Five deaths, 14 aortic dissections, 23 aortic root replacements, 3 reoperations, and 3 vascular graft implantations beyond the aortic root occurred during follow-up. Except for aortic root replacement, all endpoints occurred in patients with an operated aortic root. Patients who used losartan during the entire follow-up period showed a reduced number of events compared to the control group (death: 0 vs. 5, P = 0.014; aortic dissection: 3 vs. 11, P = 0.013; elective aortic root replacement: 10 vs. 13, P = 0.264; reoperation: 1 vs. 2, P = 0.463; vascular graft implantations beyond the aortic root 0 vs. 3, P = 0.071; and composite endpoint: 14 vs. 26, P = 0.019). These results remained similar when corrected for age and β-blocker use in a multivariate analysis.

CONCLUSION

These results suggest a clinical benefit of combined losartan and β-blocker treatment in patients with MFS.

Antihypertensives and their relation to mortality by SARS-CoV-2 infection

The role of antihypertensives, especially Renin-Angiotensin-Aldosterone System inhibitors, is still debatable in COVID-19-related severity and outcome. Therefore, we search for a more global analysis of antihypertensive medication in relation to SAS-CoV-2 severity using prescription data worldwide. The association between the percentage use of different types of antihypertensive medications and mortality rates due to a SARS-CoV-2 infection during the first 3 weeks of the pandemic was analyzed using random effects linear regression models for 30 countries worldwide. Higher percentages of prescribed angiotensin receptor blockers (ARBs) (β, 95% confidence interval [CI]; -0.02 [-0.04 to -0.0012]; p = .042) and calcium channel blockers (CCBs) (β, 95% CI; -0.023 [-0.05 to -0.0028]; p = .0304) were associated with a lower first 3-week SARS-CoV-2-related death rate, whereas a higher percentage of prescribed angiotensin-converting enzyme inhibitors (ACEis) (β, 95% CI; 0.03 [0.0061-0.05]; p = .0103) was associated with a higher first 3-week death rate, even when adjusted for age and metformin use. There was no association between the amount of prescribed beta-blockers (BBs) and diuretics (Diu) and the first 3-week death rate. When analyzing the combination of drugs that is used by at least 50% of antihypertensive users, within the different countries, countries with the lowest first 3-week death rates had at least an angiotensin receptor blocker as one of the most often prescribed antihypertensive medications (ARBs/CCBs: [β, 95% CI; -0.02 [-0.03 to -0.004]; p = .009], ARBs/BBs: [β, 95% CI; -0.03 [-0.05 to -0.006]; p = .01]). Finally, countries prescribing high-potency ARBs had lower first 3-week ARBs. In conclusion, ARBs and CCB seem to have a protective effect against death from SARS-CoV-2 infection.

Dutch Outcome in Implantable Cardioverter-Defibrillator Therapy: Implantable Cardioverter-Defibrillator-Related Complications in a Contemporary Primary Prevention Cohort

Background

One third of primary prevention implantable cardioverter‐defibrillator patients receive appropriate therapy, but all remain at risk of defibrillator complications. Information on these complications in contemporary cohorts is limited. This study assessed complications and their risk factors after defibrillator implantation in a Dutch nationwide prospective registry cohort and forecasts the potential reduction in complications under distinct scenarios of updated indication criteria.

Methods and Results

Complications in a prospective multicenter registry cohort of 1442 primary implantable cardioverter‐defibrillator implant patients were classified as major or minor. The potential for reducing complications was derived from a newly developed prediction model of appropriate therapy to identify patients with a low probability of benefitting from the implantable cardioverter‐defibrillator. During a follow‐up of 2.2 years (interquartile range, 2.0–2.6 years), 228 complications occurred in 195 patients (13.6%), with 113 patients (7.8%) experiencing at least one major complication. Most common ones were lead related (n=93) and infection (n=18). Minor complications occurred in 6.8% of patients, with lead‐related (n=47) and pocket‐related (n=40) complications as the most prevailing ones. A surgical reintervention or additional hospitalization was required in 53% or 61% of complications, respectively. Complications were strongly associated with device type. Application of stricter implant indication results in a comparable proportional reduction of (major) complications.

Conclusions

One in 13 patients experiences at least one major implantable cardioverter‐defibrillator–related complication, and many patients undergo a surgical reintervention. Complications are related to defibrillator implantations, and these should be discussed with the patient. Stricter implant indication criteria and careful selection of device type implanted may have significant clinical and financial benefits.

Investigating the relationships between unfavourable habitual sleep and metabolomic traits: evidence from multi-cohort multivariable regression and Mendelian randomization analyses

BACKGROUND

Sleep traits are associated with cardiometabolic disease risk, with evidence from Mendelian randomization (MR) suggesting that insomnia symptoms and shorter sleep duration increase coronary artery disease risk. We combined adjusted multivariable regression (AMV) and MR analyses of phenotypes of unfavourable sleep on 113 metabolomic traits to investigate possible biochemical mechanisms linking sleep to cardiovascular disease.

METHODS

We used AMV (N = 17,368) combined with two-sample MR (N = 38,618) to examine effects of self-reported insomnia symptoms, total habitual sleep duration, and chronotype on 113 metabolomic traits. The AMV analyses were conducted on data from 10 cohorts of mostly Europeans, adjusted for age, sex, and body mass index. For the MR analyses, we used summary results from published European-ancestry genome-wide association studies of self-reported sleep traits and of nuclear magnetic resonance (NMR) serum metabolites. We used the inverse-variance weighted (IVW) method and complemented this with sensitivity analyses to assess MR assumptions.

RESULTS

We found consistent evidence from AMV and MR analyses for associations of usual vs. sometimes/rare/never insomnia symptoms with lower citrate (- 0.08 standard deviation (SD)[95% confidence interval (CI) - 0.12, - 0.03] in AMV and - 0.03SD [- 0.07, - 0.003] in MR), higher glycoprotein acetyls (0.08SD [95% CI 0.03, 0.12] in AMV and 0.06SD [0.03, 0.10) in MR]), lower total very large HDL particles (- 0.04SD [- 0.08, 0.00] in AMV and - 0.05SD [- 0.09, - 0.02] in MR), and lower phospholipids in very large HDL particles (- 0.04SD [- 0.08, 0.002] in AMV and - 0.05SD [- 0.08, - 0.02] in MR). Longer total sleep duration associated with higher creatinine concentrations using both methods (0.02SD per 1 h [0.01, 0.03] in AMV and 0.15SD [0.02, 0.29] in MR) and with isoleucine in MR analyses (0.22SD [0.08, 0.35]). No consistent evidence was observed for effects of chronotype on metabolomic measures.

CONCLUSIONS

Whilst our results suggested that unfavourable sleep traits may not cause widespread metabolic disruption, some notable effects were observed. The evidence for possible effects of insomnia symptoms on glycoprotein acetyls and citrate and longer total sleep duration on creatinine and isoleucine might explain some of the effects, found in MR analyses of these sleep traits on coronary heart disease, which warrant further investigation.

Skin autofluorescence of advanced glycation end products and mortality in affective disorders in the lifelines cohort study: A mediation analysis

OBJECTIVE

Life expectancy in patients suffering from affective disorders is considerably diminished. We investigated whether skin autofluorescence (SAF), indicating concentration of advanced glycation end products in the skin and oxidative stress, mediates the association between affective disorders and excess mortality.

METHODS

Included were 81,041 participants of the Lifelines cohort study. Presence of major depressive disorder, dysthymia, generalised anxiety disorder, panic disorder or social phobia was assessed with the Mini-International Neuropsychiatric Interview. SAF was assessed as mediator in Cox proportional hazards models for all-cause or natural-cause mortality.

RESULTS

Mortality was increased in cases with major depression compared to controls (36.4 vs. 22.5 per 100,000 person years). Partial mediation by SAF of the association between affective disorders and mortality was shown (9.0-10.5%, P<.001-.002), although attenuated by cardiometabolic parameters and history of physical illness. For major depressive disorder, partial mediation by 5.5-10.3% was shown (crude model: P<.001; fully adjusted model: P=.03).

LIMITATIONS

The relatively short duration of follow-up and the relatively young cohort resulted in a lack of power to detect an association between mortality and dysthymia, social phobia and two or more comorbid disorders.

CONCLUSION

Evidence of partial mediation by SAF of the association between affective disorders and all-cause and natural-cause mortality was demonstrated, although attenuated by health factors. For major depression, mediation by SAF was largest and remained significant after adjustment for sociodemographic and health factors, identifying oxidative stress as possible determinant of premature death.

European first responder systems and differences in return of spontaneous circulation and survival after out-of-hospital cardiac arrest: A study of registry cohorts

BACKGROUND

In Europe, survival-rates after out-of-hospital cardiac arrest (OHCA) vary widely between regions. Whether a system dispatching First Responders (FRs; main FR-types: firefighters, police officers, citizen-responders) is present or not may be associated with survival-rates. This study aimed to assess the association between having a dispatched FR-system and rates of return of spontaneous circulation (ROSC) and survival across Europe.

METHODS

Results of an inventory of dispatched FR-systems for OHCA in Europe were combined with aggregate ROSC and survival data from the EuReCa-TWO study and additionally collected data. Regression analysis (weighted on number of patients included per region) was performed to study the association between having a dispatched FR-system and ROSC and survival-rates to hospital discharge in the total population and in patients with shockable initial rhythm, witnessed OHCA and bystander cardiopulmonary resuscitation (CPR; Utstein comparator group). For regions without a dispatched FR-system, the theoretical survival-rate if a dispatched FR-system would have existed was estimated.

FINDINGS

We included 27 European regions. There were 15,859 OHCAs in the total group and 2,326 OHCAs in the Utstein comparator group. Aggregate ROSC and survival-rates were significantly higher in regions with an FR-system compared to regions without (ROSC: 36% [95%CI 35%-37%] vs. 24% [95%CI 23%-25%]; P<0.001; survival in total population [N=15.859]: 13% [95%CI 12%-15%] vs. 5% [95%CI 4%-6%]; P<0.001; survival in Utstein comparator group [N=2326]: 33% [95%CI 30%-36%] vs. 18% [95%CI 16%-20%]; P<0.001), and in regions with more than one FR-type compared to regions with only one FR-type. All main FR-types were associated with higher survival-rates (all P<0.050).

INTERPRETATION

European regions with dispatched FRs showed higher ROSC and survival-rates than regions without.

FUNDING

This project/work has received funding from the European Union's Horizon 2020 research and innovation programme under acronym ESCAPE-NET, registered under grant agreement No 733381 (IO, HLT and MTB) and the European Union's COST programme under acronym PARQ, registered under grant agreement No CA19137 (IO, DC, HLT, MTB). HLT and MTB were supported by a grant from the Netherlands CardioVascular Research Initiative, Dutch Heart Foundation, Dutch Federation of University Medical Centres, Netherlands Organization for Health Research and Development, Royal Netherlands Academy of Sciences - CVON2017-15 RESCUED (HLT), and CVON2018-30 Predict2 (HLT and MTB).

Between-hospital variation in rates of complications and decline of patient performance after glioblastoma surgery in the dutch Quality Registry Neuro Surgery

Introduction

For decisions on glioblastoma surgery, the risk of complications and decline in performance is decisive. In this study, we determine the rate of complications and performance decline after resections and biopsies in a national quality registry, their risk factors and the risk-standardized variation between institutions.

Methods

Data from all 3288 adults with first-time glioblastoma surgery at 13 hospitals were obtained from a prospective population-based Quality Registry Neuro Surgery in the Netherlands between 2013 and 2017. Patients were stratified by biopsies and resections. Complications were categorized as Clavien-Dindo grades II and higher. Performance decline was considered a deterioration of more than 10 Karnofsky points at 6 weeks. Risk factors were evaluated in multivariable logistic regression analysis. Patient-specific expected and observed complications and performance declines were summarized for institutions and analyzed in funnel plots.

Results

For 2271 resections, the overall complication rate was 20 % and 16 % declined in performance. For 1017 biopsies, the overall complication rate was 11 % and 30 % declined in performance. Patient-related characteristics were significant risk factors for complications and performance decline, i.e. higher age, lower baseline Karnofsky, higher ASA classification, and the surgical procedure. Hospital characteristics, i.e. case volume, university affiliation and biopsy percentage, were not. In three institutes the observed complication rate was significantly less than expected. In one institute significantly more performance declines were observed than expected, and in one institute significantly less.

Conclusions

Patient characteristics, but not case volume, were risk factors for complications and performance decline after glioblastoma surgery. After risk-standardization, hospitals varied in complications and performance declines.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11060-021-03697-8.

Identification of patients at risk of sudden cardiac death in congenital heart disease: The PRospEctiVE study on implaNTable cardIOverter defibrillator therapy and suddeN cardiac death in Adults with Congenital Heart Disease (PREVENTION-ACHD)

BACKGROUND

Sudden cardiac death (SCD) is the main preventable cause of death in patients with adult congenital heart disease (ACHD). Since robust risk stratification methods are lacking, we developed a risk score model to predict SCD in patients with ACHD: the PRospEctiVE study on implaNTable cardIOverter defibrillator therapy and suddeN cardiac death in Adults with Congenital Heart Disease (PREVENTION-ACHD) risk score model.

OBJECTIVE

The purpose of this study was to prospectively study predicted SCD risk using the PREVENTION-ACHD risk score model and actual SCD and sustained ventricular tachycardia/ventricular fibrillation (VT/VF) rates in patients with ACHD.

METHODS

The PREVENTION-ACHD risk score model assigns 1 point each to coronary artery disease, New York Heart Association class II/III heart failure, supraventricular tachycardia, systemic ejection fraction < 40%, subpulmonary ejection fraction < 40%, QRS duration ≥ 120 ms, and QT dispersion ≥ 70 ms. SCD risk was calculated for each patient. An annual predicted risk of ≥3% constituted high risk. The primary outcome was SCD or VT/VF after 2 years. The secondary outcome was SCD.

RESULTS

The study included 783 consecutive patients with ACHD (n=239 (31%) left-sided lesions; n=138 (18%) tetralogy of Fallot; n=108 (14%) closed atrial septal defect; median age 36 years; interquartile range 28-47 years; n=401 (51%) men). The PREVENTION-ACHD risk score model identified 58 high-risk patients. Eight patients (4 at high risk) experienced the primary outcome. The Kaplan-Meier estimates were 7% (95% confidence interval [CI] 0.1%-13.3%) in the high-risk group and 0.6% (95% CI 0.0%-1.1%) in the low-risk group (hazard ratio 12.5; 95% CI 3.1-50.9; P < .001). The risk score model's sensitivity was 0.5 and specificity 0.93, resulting in a C-statistic of 0.75 (95% CI 0.57-0.90). The hazard ratio for SCD was 12.4 (95% CI 1.8-88.1) (P = .01); the sensitivity and specificity were 0.5 and 0.92, and the C-statistic was 0.81 (95% CI 0.67-0.95).

CONCLUSION

The PREVENTION-ACHD risk score model provides greater accuracy in SCD or VT/VF risk stratification as compared with current guideline indications and identifies patients with ACHD who may benefit from preventive implantable cardioverter-defibrillator implantation.

Skin autofluorescence of advanced glycation end products and course of affective disorders in the lifelines cohort study, a prospective investigation

BACKGROUND

Skin autofluorescence (SAF), indicating concentration of advanced glycation end products in the skin and oxidative stress, is cross-sectionally associated with affective disorders. Prospective studies of oxidative stress markers will help to clarify the pathophysiological role of oxidative stress.

METHODS

Data of a population-based cohort study were used. Presence of major depressive disorder, dysthymia, generalised anxiety disorder, panic disorder or social phobia was assessed at baseline and at 5-year follow-up with the Mini-International Neuropsychiatric Interview. Associations between SAF at baseline and incidence and persistence/recurrence of affective disorders were assessed with logistic regression.

RESULTS

Of 43,267 participants with no disorder at baseline, 2885 (6.7%) developed an incident disorder during follow-up. In 1360 of 3648 participants (37.3%) with an affective disorder at baseline, a persisting/recurrent disorder was present at follow-up. A modest association existed between SAF and incident affective disorders (OR=1.07 [95%CI 1.03-1.12], P<.001), specifically major depressive disorder (OR=1.11 [95%CI 1.04-1.19], P=.003); this association lost statistical significance after adjustment for sociodemographic factors. Associations between SAF and persistence/recurrence were not significant.

LIMITATIONS

Many confounders might also act as intermediate: extensive adjustment for confounders caused overfitting and possibly masked effects of SAF on course of affective disorders. Relatively small sample sizes for analyses of SAF and persistence/recurrence of affective disorders resulted in a low power.

CONCLUSIONS

Increased SAF modestly raises the odds of incident affective disorders, particularly major depressive disorder, providing evidence that oxidative stress plays a role in subsequent occurrence of affective disorders. However, significance of effects faded after adjustment for socioeconomic status.

Predicting Poor Outcome Before Endovascular Treatment in Patients With Acute Ischemic Stroke

Background: Although endovascular treatment (EVT) has greatly improved outcomes in acute ischemic stroke, still one third of patients die or remain severely disabled after stroke. If we could select patients with poor clinical outcome despite EVT, we could prevent futile treatment, avoid treatment complications, and further improve stroke care. We aimed to determine the accuracy of poor functional outcome prediction, defined as 90-day modified Rankin Scale (mRS) score ≥5, despite EVT treatment.

Methods: We included 1,526 patients from the MR CLEAN Registry, a prospective, observational, multicenter registry of ischemic stroke patients treated with EVT. We developed machine learning prediction models using all variables available at baseline before treatment. We optimized the models for both maximizing the area under the curve (AUC), reducing the number of false positives.

Results: From 1,526 patients included, 480 (31%) of patients showed poor outcome. The highest AUC was 0.81 for random forest. The highest area under the precision recall curve was 0.69 for the support vector machine. The highest achieved specificity was 95% with a sensitivity of 34% for neural networks, indicating that all models contained false positives in their predictions. From 921 mRS 0–4 patients, 27–61 (3–6%) were incorrectly classified as poor outcome. From 480 poor outcome patients in the registry, 99–163 (21–34%) were correctly identified by the models.

Conclusions: All prediction models showed a high AUC. The best-performing models correctly identified 34% of the poor outcome patients at a cost of misclassifying 4% of non-poor outcome patients. Further studies are necessary to determine whether these accuracies are reproducible before implementation in clinical practice.

Association between skin autofluorescence of advanced glycation end products and affective disorders in the lifelines cohort study

BACKGROUND

Oxidative stress may be a mechanistic link between affective disorders (depressive and anxiety disorders) and somatic disease. Advanced glycation end products are produced under the influence of oxidative stress and in the skin (measured by skin autofluorescence [SAF]) serve as marker for cumulative oxidative stress. Aim of study was to determine whether SAF is associated with presence of affective disorders.

METHODS

Participants in the Lifelines cohort study who had completed the Mini-International Neuropsychiatric Interview for affective disorders and a SAF-measurement were included. Cross-sectional associations between SAF and presence of the following psychiatric disorders were investigated through logistic regression analyses adjusted for sociodemographic factors, cardiometabolic parameters, and somatic morbidities: major depressive disorder, dysthymia, generalised anxiety disorder, panic disorder or social phobia.

RESULTS

Of 81,041 included participants (41.7% male, aged 18-91 years), 6676 (8.2%) were cases with an affective disorder. SAF was associated with presence of affective disorders (OR=1.09 [95%CI 1.07-1.12], P<.001 adjusted for sociodemographic factors). Association with major depressive disorder was strongest and significant after adjustment for all confounders (OR=1.31 [95%CI 1.25-1.36], P<.001 in the crude model; OR=1.12 [95%CI 1.07-1.17], P<.001 in the fully adjusted model). For other disorders, associations lost significance after adjustment for cardiometabolic parameters and somatic morbidities.

LIMITATIONS

Persons of non-Western descent and severely (mentally or physically) ill individuals were underrepresented.

CONCLUSIONS

SAF was associated with presence of affective disorders, suggesting a link between these disorders and cumulative oxidative stress. For major depressive disorder, this association was strongest and independent of sociodemographic, cardiometabolic factors, and somatic morbidities.

Medication in adults after atrial switch for transposition of the great arteries: clinical practice and recommendations

Aims

Heart failure is the main threat to long-term health in adults with transposition of the great arteries (TGA) corrected by an atrial switch operation (AtrSO). Current guidelines refrain from recommending heart failure medication in TGA-AtrSO, as there is insufficient data to support the hypothesis that it is beneficial. Medication is therefore prescribed based on personal judgements. We aimed to evaluate medication use in TGA-AtrSO patients and examine the association of use of renin–angiotensin–aldosterone system (RAAS) inhibitors and β-blockers with long-term survival.

Methods and results

We identified 150 TGA-AtrSO patients [median age 30 years (interquartile range 25–35), 63% male] included in the CONCOR registry from five tertiary medical centres with subsequent linkage to the Dutch Dispensed Drug Register for the years 2006–2014. Use of RAAS inhibitors, β-blockers, and diuretics increased with age, from, respectively, 21% [95% confidence interval (CI) 14–40], 12% (95% CI 7–21), and 3% (95% CI 2–7) at age 25, to 49% (95% CI 38–60), 51% (95% CI 38–63), and 41% (95% CI 29–54) at age 45. Time-varying Cox marginal structural models that adjusted for confounding medication showed a lower mortality risk with use of RAAS inhibitors and β-blockers in symptomatic patients [hazard ratio (HR) = 0.13 (95% CI 0.03–0.73); P = 0.020 and HR = 0.12 (95% CI 0.02–0.17); P = 0.019, respectively]. However, in the overall cohort, no benefit of RAAS inhibitors and β-blockers was seen [HR = 0.93 (95% CI 0.24–3.63); P = 0.92 and HR = 0.98 (0.23–4.17); P = 0.98, respectively].

Conclusion

The use of heart failure medication is high in TGA-AtrSO patients, although evidence of its benefit is limited. This study showed lower risk of mortality with use of RAAS inhibitors and β-blockers in symptomatic patients only. These findings can direct future guidelines, supporting use of RAAS inhibitors and β-blockers in symptomatic, but not asymptomatic patients.

Computer versus cardiologist: Is a machine learning algorithm able to outperform an expert in diagnosing a phospholamban p.Arg14del mutation on the electrocardiogram?

BACKGROUND

Phospholamban (PLN) p.Arg14del mutation carriers are known to develop dilated and/or arrhythmogenic cardiomyopathy, and typical electrocardiographic (ECG) features have been identified for diagnosis. Machine learning is a powerful tool used in ECG analysis and has shown to outperform cardiologists.

OBJECTIVES

We aimed to develop machine learning and deep learning models to diagnose PLN p.Arg14del cardiomyopathy using ECGs and evaluate their accuracy compared to an expert cardiologist.

METHODS

We included 155 adult PLN mutation carriers and 155 age- and sex-matched control subjects. Twenty-one PLN mutation carriers (13.4%) were classified as symptomatic (symptoms of heart failure or malignant ventricular arrhythmias). The data set was split into training and testing sets using 4-fold cross-validation. Multiple models were developed to discriminate between PLN mutation carriers and control subjects. For comparison, expert cardiologists classified the same data set. The best performing models were validated using an external PLN p.Arg14del mutation carrier data set from Murcia, Spain (n = 50). We applied occlusion maps to visualize the most contributing ECG regions.

RESULTS

In terms of specificity, expert cardiologists (0.99) outperformed all models (range 0.53-0.81). In terms of accuracy and sensitivity, experts (0.28 and 0.64) were outperformed by all models (sensitivity range 0.65-0.81). T-wave morphology was most important for classification of PLN p.Arg14del carriers. External validation showed comparable results, with the best model outperforming experts.

CONCLUSION

This study shows that machine learning can outperform experienced cardiologists in the diagnosis of PLN p.Arg14del cardiomyopathy and suggests that the shape of the T wave is of added importance to this diagnosis.