Search

Background: Identifying people at risk of cardiovascular diseases (CVD) is a cornerstone of preventative cardiology. Risk prediction models currently recommended by clinical guidelines are typically based on a limited number of predictors with sub-optimal performance across all patient groups. Data-driven techniques based on machine learning (ML) might improve the performance of risk predictions by agnostically discovering novel risk predictors and learning the complex interactions between them. We tested (1) whether ML techniques based on a state-of-the-art automated ML framework (AutoPrognosis) could improve CVD risk prediction compared to traditional approaches, and (2) whether considering non-traditional variables could increase the accuracy of CVD risk predictions., Methods and Findings: Using data on 423,604 participants without CVD at baseline in UK Biobank, we developed a ML-based model for predicting CVD risk based on 473 available variables. Our ML-based model was derived using AutoPrognosis, an algorithmic tool that automatically selects and tunes ensembles of ML modeling pipelines (comprising data imputation, feature processing, classification and calibration algorithms). We compared our model with a well-established risk prediction algorithm based on conventional CVD risk factors (Framingham score), a Cox proportional hazards (PH) model based on familiar risk factors (i.e, age, gender, smoking status, systolic blood pressure, history of diabetes, reception of treatments for hypertension and body mass index), and a Cox PH model based on all of the 473 available variables. Predictive performances were assessed using area under the receiver operating characteristic curve (AUC-ROC). Overall, our AutoPrognosis model improved risk prediction (AUC-ROC: 0.774, 95% CI: 0.768-0.780) compared to Framingham score (AUC-ROC: 0.724, 95% CI: 0.720-0.728, p < 0.001), Cox PH model with conventional risk factors (AUC-ROC: 0.734, 95% CI: 0.729-0.739, p < 0.001), and Cox PH model with all UK Biobank variables (AUC-ROC: 0.758, 95% CI: 0.753-0.763, p < 0.001). Out of 4,801 CVD cases recorded within 5 years of baseline, AutoPrognosis was able to correctly predict 368 more cases compared to the Framingham score. Our AutoPrognosis model included predictors that are not usually considered in existing risk prediction models, such as the individuals' usual walking pace and their self-reported overall health rating. Furthermore, our model improved risk prediction in potentially relevant sub-populations, such as in individuals with history of diabetes. We also highlight the relative benefits accrued from including more information into a predictive model (information gain) as compared to the benefits of using more complex models (modeling gain)., Conclusions: Our AutoPrognosis model improves the accuracy of CVD risk prediction in the UK Biobank population. This approach performs well in traditionally poorly served patient subgroups. Additionally, AutoPrognosis uncovered novel predictors for CVD disease that may now be tested in prospective studies. We found that the "information gain" achieved by considering more risk factors in the predictive model was significantly higher than the "modeling gain" achieved by adopting complex predictive models., Competing Interests: The authors have declared that no competing interests exist.

Background: Cardiovascular disease is a major cause of morbidity and mortality in COPD patients. Systemic inflammation associated with COPD, is often hypothesised as a causal factor. p38 mitogen-activated protein kinases play a key role in the inflammatory pathogenesis of COPD and atherosclerosis., Objectives: This study sought to evaluate the effects of losmapimod, a p38 mitogen-activated protein kinase (MAPK) inhibitor, on vascular inflammation and endothelial function in chronic obstructive pulmonary disease (COPD) patients with systemic inflammation (defined by plasma fibrinogen >2·8g/l)., Methods: This was a randomised, double-blind, placebo-controlled, Phase II trial that recruited COPD patients with plasma fibrinogen >2.8g/l. Participants were randomly assigned by an online program to losmapimod 7·5mg or placebo tablets twice daily for 16 weeks. Pre- and post-dose 18F-Fluorodeoxyglucose positron emission tomography co-registered with computed tomography (FDG PET/CT) imaging of the aorta and carotid arteries was performed to quantify arterial inflammation, defined by the tissue-to-blood ratio (TBR) from scan images. Endothelial function was assessed by brachial artery flow-mediated dilatation (FMD)., Results: We screened 160 patients, of whom, 36 and 37 were randomised to losmapimod or placebo. The treatment effect of losmapimod compared to placebo was not significant, at -0·05 for TBR (95% CI: -0·17, 0·07), p = 0·42, and +0·40% for FMD (95% CI: -1·66, 2·47), p = 0·70. The frequency of adverse events reported was similar in both treatment groups., Conclusions: In this plasma fibrinogen-enriched study, losmapimod had no effect on arterial inflammation and endothelial function at 16 weeks of treatment, although it was well tolerated with no significant safety concerns. These findings do not support the concept that losmapimod is an effective treatment for the adverse cardiovascular manifestations of COPD.

Aims: The association of coronary arterial calcification with cardiovascular morbidity and mortality is well-recognized. Lower limb arterial calcification (LLAC) is common in PAD but its impact on subsequent health is poorly described. We aimed to determine the association between a LLAC score and subsequent cardiovascular events in patients with symptomatic peripheral arterial disease (PAD)., Methods: LLAC scoring, and the established Bollinger score, were derived from a database of unenhanced CT scans, from patients presenting with symptomatic PAD. We determined the association between these scores outcomes. The primary outcome was combined cardiac mortality and morbidity (CM/M) with a secondary outcome of all-cause mortality., Results: 220 patients (66% male; median age 69 years) were included with follow-up for a median 46 [IQR 31-64] months. Median total LLAC scores were higher in those patients suffering a primary outcome (6831 vs. 1652; p = 0.012). Diabetes mellitus (p = 0.039), ischaemic heart disease (p = 0.028), chronic kidney disease (p = 0.026) and all-cause mortality (p = 0.012) were more common in patients in the highest quartile of LLAC scores. The area under the curve of the receiver operator curve for the LLAC score was greater (0.929: 95% CI [0.884-0.974]) than for the Bollinger score (0.824: 95% CI [0.758-0.890]) for the primary outcome. A LLAC score ≥ 4400 had the best diagnostic accuracy to determine the outcome measure., Conclusion: This is the largest study to investigate links between lower limb arterial calcification and cardiovascular events in symptomatic PAD. We describe a straightforward, reproducible, CT-derived measure of calcification-the LLAC score.