Regularized parametric survival modeling to improve risk prediction models.

We propose to combine the benefits of flexible parametric survival modeling and regularization to improve risk prediction modeling in the context of time-to-event data. Thereto, we introduce ridge, lasso, elastic net, and group lasso penalties for both log hazard and log cumulative hazard models. The log (cumulative) hazard in these models is represented by a flexible function of time that may depend on the covariates (i.e., covariate effects may be time-varying). We show that the optimization problem for the proposed models can be formulated as a convex optimization problem and provide a user-friendly R implementation for model fitting and penalty parameter selection based on cross-validation. Simulation study results show the advantage of regularization in terms of increased out-of-sample prediction accuracy and improved calibration and discrimination of predicted survival probabilities, especially when sample size was relatively small with respect to model complexity. An applied example illustrates the proposed methods. In summary, our work provides both a foundation for and an easily accessible implementation of regularized parametric survival modeling and suggests that it improves out-of-sample prediction performance.

Lipoprotein(a) levels and carotid intima-media thickness in children: A 20-year follow-up study.

Elevated lipoprotein(a) [Lp(a)] is independently associated with cardiovascular disease (CVD). In a recent long-term follow-up study involving children with familial hypercholesterolemia, Lp(a) levels contributed significantly to early atherosclerosis, as measured by carotid intima-media thickness (cIMT). To determine if this holds true for children without FH, we conducted a 20-year follow-up study, examining 88 unaffected siblings (mean age: 12.9 years) of children with FH. No significant association was found between Lp(a) and cIMT during follow-up (ß-adjusted [95% CI] = 0.0001 [-0.008 to 0.008] mm per 50 nmol/L increase Lp(a), p = 0.97). In conclusion, our findings suggest that elevated levels of Lp(a) do not play a significant role in arterial wall thickening among children without FH during the 20-year follow-up period. This leads us to consider the possibility that cIMT may not be a suitable marker for detecting potential subtle changes in the arterial wall mediated by Lp(a) in the young, general population. However, it could also be that elevated Lp(a) is only a significant risk factor for atherosclerosis in the presence of other risk factors such as FH.

A European-Japanese study on peach allergy: IgE to Pru p 7 associates with severity.

BACKGROUND: Pru p 3 and Pru p 7 have been implicated as risk factors for severe peach allergy. This study aimed to establish sensitization patterns to five peach components across Europe and in Japan, to explore their relation to pollen and foods and to predict symptom severity. METHODS: In twelve European (EuroPrevall project) and one Japanese outpatient clinic, a standardized clinical evaluation was conducted in 1231 patients who reported symptoms to peach and/or were sensitized to peach. Specific IgE against Pru p 1, 2, 3, 4 and 7 and against Cup s 7 was measured in 474 of them. Univariable and multivariable Lasso regression was applied to identify combinations of parameters predicting severity. RESULTS: Sensitization to Pru p 3 dominated in Southern Europe but was also quite common in Northern and Central Europe. Sensitization to Pru p 7 was low and variable in the European centers but very dominant in Japan. Severity could be predicted by a model combining age of onset of peach allergy, probable mugwort, Parietaria pollen and latex allergy, and sensitization to Japanese cedar pollen, Pru p 4 and Pru p 7 which resulted in an AUC of 0.73 (95% CI 0.73-0.74). Pru p 3 tended to be a risk factor in South Europe only. CONCLUSIONS: Pru p 7 was confirmed as a significant risk factor for severe peach allergy in Europe and Japan. Combining outcomes from clinical and demographic background with serology resulted in a model that could better predict severity than CRD alone.

Hospital utilisation and the costs associated with complications of ICD implantation in a contemporary primary prevention cohort.

INTRODUCTION: Implantation of an implantable cardioverter defibrillator (ICD) is standard care for primary prevention of sudden cardiac death. However, ICD-related complications are increasing as the population of ICD recipients grows. METHODS: ICD-related complications in a national DO-IT Registry cohort of 1442 primary prevention ICD patients were assessed in terms of additional use of hospital care resources and costs. RESULTS: During a median follow-up of 28.7 months (IQR 25.2-33.7) one or more complications occurred in 13.5% of patients. A complication resulted in a surgical intervention in 53% of cases and required on average 3.65 additional hospital days. The additional hospital costs were €6,876 per complication or €8,110 per patient, to which clinical re-interventions and additional hospital days contributed most. Per category of complications, infections required most hospital utilisation and were most expensive at an average of €22,892. The mean costs were €5,800 for lead-related complications, €2,291 for pocket-related complications and €5,619 for complications due to other causes. We estimate that the total yearly incidence-based costs in the Netherlands for hospital management of ICD-related complications following ICD implantation for primary prevention are €2.7 million. CONCLUSION: Complications following ICD implantation are related to a substantial additional need for hospital resources. When performing cost-effectiveness analyses of ICD implantation, including the costs associated with complications, one should be aware that real-world complication rates may deviate from trial data. Considering the economic implications, strategies to reduce the incidence of complications are encouraged.

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

BACKGROUND AND PURPOSE: The electrocardiogram (ECG) is frequently obtained in the work-up of COVID-19 patients. So far, no study has evaluated whether ECG-based machine learning models have added value to predict in-hospital mortality specifically in COVID-19 patients. METHODS: Using data from the CAPACITY-COVID registry, we studied 882 patients admitted with COVID-19 across seven hospitals in the Netherlands. Raw format 12-lead ECGs recorded within 72 h of admission were studied. With data from five hospitals (n = 634), three models were developed: (a) a logistic regression baseline model using age and sex, (b) a least absolute shrinkage and selection operator (LASSO) model using age, sex and human annotated ECG features, and (c) a pre-trained deep neural network (DNN) using age, sex and the raw ECG waveforms. Data from two hospitals (n = 248) was used for external validation. RESULTS: Performances for models a, b and c were comparable with an area under the receiver operating curve of 0.73 (95% confidence interval [CI] 0.65-0.79), 0.76 (95% CI 0.68-0.82) and 0.77 (95% CI 0.70-0.83) respectively. Predictors of mortality in the LASSO model were age, low QRS voltage, ST depression, premature atrial complexes, sex, increased ventricular rate, and right bundle branch block. CONCLUSION: This study shows that the ECG-based prediction models could be helpful for the initial risk stratification of patients diagnosed with COVID-19, and that several ECG abnormalities are associated with in-hospital all-cause mortality of COVID-19 patients. Moreover, this proof-of-principle study shows that the use of pre-trained DNNs for ECG analysis does not underperform compared with time-consuming manual annotation of ECG features.

Clinical features and prognostic factors in Covid-19: A prospective cohort study.

BACKGROUND: Mortality rates are high among hospitalized patients with COVID-19, especially in those intubated on the ICU. Insight in pathways associated with unfavourable outcome may lead to new treatment strategies. METHODS: We performed a prospective cohort study of patients with COVID-19 admitted to general ward or ICU who underwent serial blood sampling. To provide insight in the pathways involved in disease progression, associations were estimated between outcome risk and serial measurements of 64 biomarkers in potential important pathways of COVID-19 infection (inflammation, tissue damage, complement system, coagulation and fibrinolysis) using joint models combining Cox regression and linear mixed-effects models. For patients admitted to the general ward, the primary outcome was admission to the ICU or mortality (unfavourable outcome). For patients admitted to the ICU, the primary outcome was 12-week mortality. FINDINGS: A total of 219 patients were included: 136 (62%) on the ward and 119 patients (54%) on the ICU; 36 patients (26%) were included in both cohorts because they were transferred from general ward to ICU. On the general ward, 54 of 136 patients (40%) had an unfavourable outcome and 31 (23%) patients died. On the ICU, 54 out of 119 patients (45%) died. Unfavourable outcome on the general ward was associated with changes in concentrations of IL-6, IL-8, IL-10, soluble Receptor for Advanced Glycation End Products (sRAGE), vascular cell adhesion molecule 1 (VCAM-1) and Pentraxin-3. Death on the ICU was associated with changes in IL-6, IL-8, IL-10, sRAGE, VCAM-1, Pentraxin-3, urokinase-type plasminogen activator receptor, IL-1-receptor antagonist, CD14, procalcitonin, tumor necrosis factor alfa, tissue factor, complement component 5a, Growth arrest-specific 6, angiopoietin 2, and lactoferrin. Pathway analysis showed that unfavourable outcome on the ward was mainly driven by chemotaxis and interleukin production, whereas death on ICU was associated with a variety of pathways including chemotaxis, cell-cell adhesion, innate host response mechanisms, including the complement system, viral life cycle regulation, angiogenesis, wound healing and response to corticosteroids. INTERPRETATION: Clinical deterioration in patients with severe COVID-19 involves multiple pathways, including chemotaxis and interleukin production, but also endothelial dysfunction, the complement system, and immunothrombosis. Prognostic markers showed considerable overlap between general ward and ICU patients, but we identified distinct differences between groups that should be considered in the development and timing of interventional therapies in COVID-19. FUNDING: Amsterdam UMC, Amsterdam UMC Corona Fund, and Dr. C.J. Vaillant Fonds.

Enhanced PeriOperative Care and Health protection programme for the prevention of surgical site infections after elective abdominal surgery (EPO2CH): statistical analysis plan of a randomised controlled multicentre superiority trial.

BACKGROUND: Surgical site infections (SSI) are frequent complications after elective abdominal surgery. We designed the Enhanced PeriOperative Care and Health Protection programme (EPO2CH) care bundle, comprising of intraoperative high fractional inspired oxygen; intraoperative goal-directed fluid therapy; active preoperative, intraoperative and postoperative warming; glucose control and treatment of hyperglycaemia (> 10 mmol L- 1) in diabetics as well as non-diabetics; and wound irrigation before closure using an aqueous antiseptic. We hypothesise that EPO2CH added to standard care reduces the incidence of SSI compared to standard care alone for elective abdominal surgery. METHODS: This trial is designed as an open label, pragmatic randomised controlled parallel-group multicentre superiority trial. The primary endpoint is the incidence of SSI, defined by the Centers for Disease Control and prevention, within 30 days after surgery. The incidence of SSI is assessed using the Dutch national complication register and medical chart review. Secondary endpoints include the SSI incidence within 90 days, incidence of anastomotic leakage at 30 and 90 days, the incidence of incisional hernia within 1 year, mortality within 1 year and 5 years, quality of life, health and disability, and cost-effectiveness. Primarily, an intention-to-treat analysis will be performed to estimate the relative risk using a log binomial model. If not feasible, a logistic regression will be used to estimate the odds ratio. A per-protocol analysis will also be performed. Furthermore, the attributive effect of the distinct interventions will be explored. DISCUSSION: The results of the EPO2CH trial will determine if the EPO2CH bundle is effective to prevent SSI incidence for patients undergoing elective abdominal surgery. Details of the statistical analysis are described in this Statistical Analysis Plan (SAP). TRIAL REGISTRATION: Registration number: Dutch Trial Register Trial NL5572 . Registered on March 3, 2016. SAP version: V1.0, January 8, 2020. This SAP has been written based on study protocol V10.

[Medical disciplinary boards punish more often and more severely; an observational study of 10 years of medical disciplinary law]. / Medische tuchtcolleges straffen vaker en strenger.

OBJECTIVE: To examine if the medical disciplinary law in the Netherlands is becoming tougher. DESIGN: Observational study. METHOD: We read the annual reports of the Dutch Disciplinary Committees for the Healthcare Sector 2007-2017, and registered the numbers of disciplinary cases and those cases upheld, the measures imposed and the degree of consistency in the event of an appeal. Over 400 medical disciplinary cases were reviewed in more detail. Based on these cases, we determined those variables that increase or decrease the likelihood of conviction, such as the gender of the defendants, the region in which the disciplinary case was heard, and if objective norms and professional practice guidelines had been used in arriving at the verdict. RESULTS: Each year a disciplinary case is brought against an average of 0.4% of all health care providers - mainly physicians (an average of approximately 1400 disciplinary cases each year). One-third of all disciplinary cases were dealt with in court and about half of them were upheld. Over time, the number of disciplinary cases has increased (36%), as has the percentage of cases that are upheld (27%). Additionally, heavier measures were more often imposed and the degree of consistency between the initial ruling and the ruling on appeal also increased (56%). Those factors that increased the likelihood of a case being upheld were: being a physician of male gender, the disciplinary case being heard in The Hague or Eindhoven, and the non-implementation of an objective norm or professional practice guideline in arriving at the verdict. CONCLUSION: Since 2007 the Dutch medical disciplinary boards have been punishing more often and more severely. This may be because the subjective demands made on health care professionals have changed. Arriving at an opinion subjectively is not a problem when it comes to verifying compliance with standards of due care. However, if this method is used to determine the penalty, the health care system will become defensive - and this will not improve the quality of care.

External Validation of the Dutch SOURCE Survival Prediction Model in Belgian Metastatic Oesophageal and Gastric Cancer Patients.

The SOURCE prediction model predicts individualised survival conditional on various treatments for patients with metastatic oesophageal or gastric cancer. The aim of this study was to validate SOURCE in an external cohort from the Belgian Cancer Registry. Data of Belgian patients diagnosed with metastatic disease between 2004 and 2014 were extracted (n = 4097). Model calibration and discrimination (c-indices) were determined. A total of 2514 patients with oesophageal cancer and 1583 patients with gastric cancer with a median survival of 7.7 and 5.4 months, respectively, were included. The oesophageal cancer model showed poor calibration (intercept: 0.30, slope: 0.42) with an absolute mean prediction error of 14.6%. The mean difference between predicted and observed survival was -2.6%. The concordance index (c-index) of the oesophageal model was 0.64. The gastric cancer model showed good calibration (intercept: 0.02, slope: 0.91) with an absolute mean prediction error of 2.5%. The mean difference between predicted and observed survival was 2.0%. The c-index of the gastric cancer model was 0.66. The SOURCE gastric cancer model was well calibrated and had a similar performance in the Belgian cohort compared with the Dutch internal validation. However, the oesophageal cancer model had not. Our findings underscore the importance of evaluating the performance of prediction models in other populations.

Meta-analysis and field synopsis of genetic variants associated with the risk and severity of acute pancreatitis.

BACKGROUND: Genetic risk factors can provide insight into susceptibility for acute pancreatitis (AP) and disease progression towards (infected) necrotizing pancreatitis and persistent organ failure. The aim of the study was to undertake a systematic review of the genetic evidence for AP. METHODS: Online databases (MEDLINE, Embase, BIOSIS, Web of Science, Cochrane Library) were searched to 8 February 2018. Studies that reported on genetic associations with AP susceptibility, severity and/or complications were eligible for inclusion. Meta-analyses were performed of variants that were reported by at least two data sources. Venice criteria and Bayesian false-discovery probability were applied to assess credibility. RESULTS: Ninety-six studies reporting on 181 variants in 79 genes were identified. In agreement with previous meta-analyses, credible associations were established for SPINK1 (odds ratio (OR) 2·87, 95 per cent c.i. 1·89 to 4·34), IL1B (OR 1·23, 1·06 to 1·42) and IL6 (OR 1·64, 1·15 to 2·32) and disease risk. In addition, two novel credible single-nucleotide polymorphisms were identified in Asian populations: ALDH2 (OR 0·48, 0·36 to 0·64) and IL18 (OR 1·47, 1·18 to 1·82). Associations of variants in TNF, GSTP1 and CXCL8 genes with disease severity were identified, but were of low credibility. CONCLUSION: Genetic risk factors in genes related to trypsin activation and innate immunity appear to be associated with susceptibility to and severity of AP.

ANTECEDENTES: Los factores de riesgo genético pueden contribuir a determinar la susceptibilidad para desarrollar una pancreatitis aguda (acute pancreatitis, AP) y de su progresión a pancreatitis necrotizante (infectada) e insuficiencia orgánica crónica. Nuestro objetivo fue revisar de forma sistemática la evidencia genética de la pancreatitis aguda. MÉTODOS: Se revisaron las bases de datos electrónicas (MEDLINE, Embase, BIOSIS, Web of Science, Cochrane Library) hasta febrero de 2018. Se incluyeron estudios que presentaban información de las asociaciones genéticas y la susceptibilidad de AP, gravedad y/o complicaciones. Se realizó un metaanálisis de las variantes genéticas descritas en al menos dos fuentes. Se aplicaron los criterios de Venecia y la probabilidad bayesiana de falsa alarma para la valoración de la credibilidad. RESULTADOS: Se identificaron 96 estudios que analizaron 181 variantes en 79 genes. De acuerdo con un metaanálisis previo, se establecieron asociaciones creibles con el riesgo de enfermedad para SPINK1 (razón de oportunidades, odds ratio, OR 2,87, i.c. del 95% 1,89-4,34), IL1B (OR 1,23, i.c. del 95% 1,06-1,42) e IL6 (OR 1,64, i.c. del 95% 1,15-2,32). Además, en poblaciones asiáticas, se identificaron dos nuevos polimorfismos de nucleótico único (SNP) creibles en ALDH2 (OR 0,48, i.c. del 95% 0,36-0,64) e IL18 (OR 1,47, i.c. del 95% 1,18-1,82). En cuanto a la gravedad de la enfermedad se identificaron variantes en los genes TNF, GSTP1 y CXCL8, pero de baja credibilidad en función de nuestra evaluación. CONCLUSIÓN: Los factores de riesgo genéticos en genes relacionados con la activación de la tripsina y la inmunidad innata parecen ser estar asociados con la susceptibilidad y gravedad de la pancreatitis aguda.

Data-efficient deep learning of radiological image data for outcome prediction after endovascular treatment of patients with acute ischemic stroke.

Treatment selection is becoming increasingly more important in acute ischemic stroke patient care. Clinical variables and radiological image biomarkers (old age, pre-stroke mRS, NIHSS, occlusion location, ASPECTS, among others) have an important role in treatment selection and prognosis. Radiological biomarkers require expert annotation and are subject to inter-observer variability. Recently, Deep Learning has been introduced to reproduce these radiological image biomarkers. Instead of reproducing these biomarkers, in this work, we investigated Deep Learning techniques for building models to directly predict good reperfusion after endovascular treatment (EVT) and good functional outcome using CT angiography images. These models do not require image annotation and are fast to compute. We compare the Deep Learning models to Machine Learning models using traditional radiological image biomarkers. We explored Residual Neural Network (ResNet) architectures, adapted them with Structured Receptive Fields (RFNN) and auto-encoders (AE) for network weight initialization. We further included model visualization techniques to provide insight into the network's decision-making process. We applied the methods on the MR CLEAN Registry dataset with 1301 patients. The Deep Learning models outperformed the models using traditional radiological image biomarkers in three out of four cross-validation folds for functional outcome (average AUC of 0.71) and for all folds for reperfusion (average AUC of 0.65). Model visualization showed that the arteries were relevant features for functional outcome prediction. The best results were obtained for the ResNet models with RFNN. Auto-encoder initialization often improved the results. We concluded that, in our dataset, automated image analysis with Deep Learning methods outperforms radiological image biomarkers for stroke outcome prediction and has the potential to improve treatment selection.

PREVENTION-ACHD: PRospEctiVE study on implaNTable cardioverter-defibrillator therapy and suddeN cardiac death in Adults with Congenital Heart Disease; Rationale and Design.

BACKGROUND: Many adult congenital heart disease (ACHD) patients are at risk of sudden cardiac death (SCD). An implantable cardioverter-defibrillator (ICD) may prevent SCD, but the evidence for primary prevention indications is still unsatisfactory. STUDY DESIGN: PREVENTION-ACHD is a prospective study with which we aim to prospectively validate a new risk score model for primary prevention of SCD in ACHD patients, as well as the currently existing guideline recommendations. Patients are screened using a novel risk score to predict SCD as well as current ICD indications according to an international Consensus Statement. Patients are followed up for two years. The primary endpoint is the occurrence of SCD and sustained ventricular arrhythmias. The Study was registered at ClinicalTrials.gov (NCT03957824). CONCLUSION: PREVENTION-ACHD is the first prospective study on SCD in ACHD patients. In the light of a growing and aging population of patients with more severe congenital heart defects, more robust clinical evidence on primary prevention of SCD is urgently needed.

[Hydrochlorothiazide and skin cancer]. / Hydrochloorthiazide en huidkanker.

Hydrochlorothiazide and skin cancer Hydrochlorothiazide is a frequently prescribed diuretic with known photosensitizing properties. Recently, a large case-control study in a Danish population found an association between the use of hydrochlorothiazide and an increased risk of developing non-melanoma skin cancer. These data suggest that it may be wise to limit the use of hydrochlorothiazide for the treatment of hypertension. We reviewed the current literature to examine whether a causal relationship between hydrochlorothiazide and non-melanoma skin cancer exists. We consider that the evidence for a causal relationship is limited and contradicting. Moreover, we found that other antihypertensive agents such as calcium blockers and angiotensin receptor blockers are also associated with basal cell carcinoma. Based on the current literature, there seems to be insufficient evidence to advice against the use of hydrochlorothiazide.

Prevalence of and factors associated with adult-onset asthma in different ethnic groups: The HELIUS study.

Little is known about adult-onset asthma in different ethnic groups. The aim of this study was to examine ethnic differences in the prevalence of adult-onset asthma and factors associated with this phenotype. Cross-sectional data of 23,356 participants of the HELIUS study were used, including Dutch, South-Asian Surinamese, African Surinamese, Moroccan, Turkish and Ghanaian origin participants. Adult-onset asthma was defined as: self-reported asthma symptoms or start of asthma-medication at age ≥18 years combined with a smoking history <10 pack years. The prevalence of adult-onset asthma and its association with potential risk factors were assessed by logistic regression analyses. The adjusted prevalence of adult-onset asthma was higher in the Turkish, Moroccan and South-Asian Surinamese groups (4.9-6.0%) compared to the Dutch, Ghanaian and African Surinamese origin groups (2.4-2.6%). In addition to ethnicity, age, female sex, BMI, and doctors' diagnosis of nasal allergy/hay fever and chronic sinusitis/polyps were independently associated with adult-onset asthma. There are significant differences in the adjusted prevalence of adult-onset asthma among six ethnic groups.

Validation of Reference Charts for Mid-Trimester Fetal Biometry.

OBJECTIVE: The aim of our study was to assess charts proposed for international use in the Intergrowth-21st Project. METHODS: Ultrasound data were collected from 43,923 healthy singleton pregnancies examined at 18-23 weeks of gestation in the Netherlands. Fetal measurements were converted into Z-scores using previous and current Dutch reference charts and Intergrowth charts. The distributions of the Z-scores were compared with the expected standard normal distribution. RESULTS: In the Dutch population, Intergrowth curves perform well for head circumference and biparietal diameter, but not for abdominal circumference (AC, Z- score = 0.43) and femur length (FL, Z-score = 0.26). Similar findings have been reported in other European countries. Compared with the population in the Intergrowth study, Dutch women are relatively tall (170 vs. 162 cm) and sturdy (67 vs. 61 kg) with a moderately high BMI. Maternal size, in particular maternal height, is positively correlated with birthweight. CONCLUSIONS: Whilst the establishment of the Intergrowth charts is an important step towards worldwide uniformity, for now locally derived charts still perform better, especially for AC and FL. Results from our validation study indicate that distinction between normal and pathologically small babies may be improved by taking maternal size into account.

Machine learning improves prediction of delayed cerebral ischemia in patients with subarachnoid hemorrhage.

BACKGROUND AND PURPOSE: Delayed cerebral ischemia (DCI) is a severe complication in patients with aneurysmal subarachnoid hemorrhage. Several associated predictors have been previously identified. However, their predictive value is generally low. We hypothesize that Machine Learning (ML) algorithms for the prediction of DCI using a combination of clinical and image data lead to higher predictive accuracy than previously applied logistic regressions. MATERIALS AND METHODS: Clinical and baseline CT image data from 317 patients with aneurysmal subarachnoid hemorrhage were included. Three types of analysis were performed to predict DCI. First, the prognostic value of known predictors was assessed with logistic regression models. Second, ML models were created using all clinical variables. Third, image features were extracted from the CT images using an auto-encoder and combined with clinical data to create ML models. Accuracy was evaluated based on the area under the curve (AUC), sensitivity and specificity with 95% CI. RESULTS: The best AUC of the logistic regression models for known predictors was 0.63 (95% CI 0.62 to 0.63). For the ML algorithms with clinical data there was a small but statistically significant improvement in the AUC to 0.68 (95% CI 0.65 to 0.69). Notably, aneurysm width and height were included in many of the ML models. The AUC was highest for ML models that also included image features: 0.74 (95% CI 0.72 to 0.75). CONCLUSION: ML algorithms significantly improve the prediction of DCI in patients with aneurysmal subarachnoid hemorrhage, particularly when image features are also included. Our experiments suggest that aneurysm characteristics are also associated with the development of DCI.

Novel endotypes in heart failure: effects on guideline-directed medical therapy.

Aims: We sought to determine subtypes of patients with heart failure (HF) with a distinct clinical profile and treatment response, using a wide range of biomarkers from various pathophysiological domains. Methods and results: We performed unsupervised cluster analysis using 92 established cardiovascular biomarkers to identify mutually exclusive subgroups (endotypes) of 1802 patients with HF and reduced ejection fraction (HFrEF) from the BIOSTAT-CHF project. We validated our findings in an independent cohort of 813 patients. Based on their biomarker profile, six endotypes were identified. Patients with endotype 1 were youngest, less symptomatic, had the lowest N-terminal pro-B-type natriuretic peptide (NT-proBNP) levels and lowest risk for all-cause mortality or hospitalization for HF. Patients with endotype 4 had more severe symptoms and signs of HF, higher NT-proBNP levels and were at highest risk for all-cause mortality or hospitalization for HF [hazard ratio (HR) 1.4; 95% confidence interval (CI) 1.1-1.8]. Patients with endotypes 2, 3, and 5 were better uptitrated to target doses of beta-blockers (P < 0.02 for all). In contrast to other endotypes, patients with endotype 5 derived no potential survival benefit from uptitration of angiotensin-converting enzyme-inhibitor/angiotensin-II receptor blocker and beta-blockers (Pinteraction <0.001). Patients with endotype 2 (HR 1.29; 95% CI 1.10-1.42) experienced possible harm from uptitration of beta-blockers in contrast to patients with endotype 4 and 6 that experienced benefit (Pinteraction for all <0.001). Results were strikingly similar in the independent validation cohort. Conclusion: Using unsupervised cluster analysis, solely based on biomarker profiles, six distinct endotypes were identified with remarkable differences in characteristics, clinical outcome, and response to uptitration of guideline directed medical therapy.

Variation in coagulation and fibrinolysis genes evaluated for their contribution to cerebrovascular complications in adults with bacterial meningitis in the Netherlands.

OBJECTIVE: To study whether genetic variation in coagulation and fibrinolysis genes contributes to cerebrovascular complications in bacterial meningitis. METHODS: We performed a nationwide prospective genetic association study in adult community-acquired bacterial meningitis patients. The exons and flanking regions of 16 candidate genes involved in coagulation and fibrinolysis pathways were sequenced. We analyzed whether genetic variation in these genes resulted in a higher risk of cerebrovascular complications, unfavorable outcome and differences in thrombocyte count on admission. RESULTS: From 2006 to 2011, a total of 1101 bacterial meningitis patients were identified of whom 622 supplied DNA for genotyping and passed genetic quality control steps. In 139 patients (22%) the episode of bacterial meningitis was complicated by cerebral infarction, and 188 (30%) had an unfavorable outcome. We identified the functional variant rs494860 in the protein Z (PROZ) gene as our strongest association with occurrence of cerebral infarction (odds ratio (OR) 0.49 (95% confidence interval 0.33-0.73), p = 5.2 × 10-4). After Bonferroni correction for multiple testing no genetic variant was significantly associated (p-value threshold 2.7 × 10-4). CONCLUSION: Our study suggests a functional genetic variation in the PROZ gene, rs494860, may be of importance in bacterial meningitis pathogenesis and cerebral infarction risk. Replication of this finding in other cohort studies populations is needed.

Prediction models for patients with esophageal or gastric cancer: A systematic review and meta-analysis.

BACKGROUND: Clinical prediction models are increasingly used to predict outcomes such as survival in cancer patients. The aim of this study was threefold. First, to perform a systematic review to identify available clinical prediction models for patients with esophageal and/or gastric cancer. Second, to evaluate sources of bias in the included studies. Third, to investigate the predictive performance of the prediction models using meta-analysis. METHODS: MEDLINE, EMBASE, PsycINFO, CINAHL, and The Cochrane Library were searched for publications from the year 2000 onwards. Studies describing models predicting survival, adverse events and/or health-related quality of life (HRQoL) for esophageal or gastric cancer patients were included. Potential sources of bias were assessed and a meta-analysis, pooled per prediction model, was performed on the discriminative abilities (c-indices). RESULTS: A total of 61 studies were included (45 development and 16 validation studies), describing 47 prediction models. Most models predicted survival after a curative resection. Nearly 75% of the studies exhibited bias in at least 3 areas and model calibration was rarely reported. The meta-analysis showed that the averaged c-index of the models is fair (0.75) and ranges from 0.65 to 0.85. CONCLUSION: Most available prediction models only focus on survival after a curative resection, which is only relevant to a limited patient population. Few models predicted adverse events after resection, and none focused on patient's HRQoL, despite its relevance. Generally, the quality of reporting is poor and external model validation is limited. We conclude that there is a need for prediction models that better meet patients' information needs, and provide information on both the benefits and harms of the various treatment options in terms of survival, adverse events and HRQoL.

Dynamic prediction of recurrent events data by landmarking with application to a follow-up study of patients after kidney transplant.

This paper extends dynamic prediction by landmarking to recurrent event data. The motivating data comprised post-kidney transplantation records of repeated infections and repeated measurements of multiple markers. At each landmark time point ts, a Cox proportional hazards model with a frailty term was fitted using data of individuals who were at risk at landmark s. This model included the time-updated marker values at ts as time-fixed covariates. Based on a stacked data set that merged all landmark data sets, we considered supermodels that allow parameters to depend on the landmarks in a smooth fashion. We described and evaluated four ways to parameterize the supermodels for recurrent event data. With both the study data and simulated data sets, we compared supermodels that were fitted on stacked data sets that consisted of either overlapping or non-overlapping landmark periods. We observed that for recurrent event data, the supermodels may yield biased estimates when overlapping landmark periods are used for stacking. Using the best supermodel amongst the ones considered, we dynamically estimated the probability to remain infection free between ts and a prediction horizon thor, conditional on the information available at ts.