Daily Cardiology Research Analysis

BACKGROUND: Although statistical models have been commonly used to identify patients at risk of cardiovascular disease for preventive therapy, these models tend to over-recommend therapy. Moreover, in populations with pre-existing diseases, the current approach is to indiscriminately treat all, as modelling in this context is currently inadequate. This study aimed to develop and validate the Transformer-based Risk assessment survival (TRisk) model, a novel deep learning model, for predicting 10-year risk of cardiovascular disease in both the primary prevention population and individuals with diabetes. METHODS: An open cohort of 3 million adults aged 25-84 years was identified using linked electronic health records from 291 general practices, for model development, and 98 general practices, for validation, across England from 1998 to 2015. Comparison against the QRISK3 score and a deep learning derivation of it was done. Additional analyses compared discriminatory performance in other age groups, by sex, and across categories of socioeconomic status. FINDINGS: TRisk showed superior discrimination (C index in the primary prevention population 0·910; 95% CI 0·906-0·913). TRisk's performance was found to be less sensitive to population age range than the benchmark models and outperformed other models also in analyses stratified by age, sex, or socioeconomic status. All models were overall well calibrated. In decision curve analyses, TRisk showed a greater net benefit than benchmark models across the range of relevant thresholds. At the widely recommended 10% risk threshold and the higher 15% threshold, TRisk reduced both the total number of patients classified at high risk (by 20·6% and 34·6%, respectively) and the number of false negatives as compared with recommended strategies. TRisk similarly outperformed other models in patients with diabetes. Compared with the widely recommended treat-all policy approach for patients with diabetes, TRisk at a 10% risk threshold would lead to deselection of 24·3% of individuals, with a small fraction of false negatives (0·2% of the cohort). INTERPRETATION: TRisk enabled a more targeted selection of individuals at risk of cardiovascular disease in both the primary prevention population and cohorts with diabetes, compared with benchmark approaches. Incorporation of TRisk into routine care could potentially reduce the number of treatment-eligible patients by approximately one-third while preventing at least as many events as with currently adopted approaches. FUNDING: None.

IMPORTANCE: Lipoprotein(a) [Lp(a)] is independently associated with atherosclerotic cardiovascular disease (ASCVD) risk but is not included in the new American Heart Association Predicting Risk of Cardiovascular Disease Events (PREVENT) equations for CVD risk assessment. OBJECTIVE: To evaluate the performance of these equations in individuals with elevated Lp(a). DESIGN, SETTING, AND PARTICIPANTS: Cohort study involving 314 783 participants from the multicenter Multi-Ethnic Study of Atherosclerosis (MESA, 2000-2018; n = 6670) and the population-based UK Biobank (UKB, 2006-2022; n = 308 113) without known cardiovascular disease with available Lp(a) measurements. Analyses were conducted March 25, 2025. EXPOSURE: Elevated Lp(a) level of 125 nmol/L or higher. MAIN OUTCOMES AND MEASURES: Coronary heart disease (CHD), ASCVD, heart failure (HF), and total CVD. Participants were categorized as low (<5%), borderline (5% to <7.5%) intermediate (7.5% to <20%), and high (≥20%) risk of each outcome. Ten-year observed event rates were calculated, and the association between elevated Lp(a) and outcomes overall and by risk category was evaluated in age- and sex-adjusted Cox proportional hazards models. Improvement in risk prediction with the addition of elevated Lp(a) was evaluated using continuous and categorical net reclassification improvement (NRI) (using the above cut points). RESULTS: Among the 314 783 participants (mean [SD] age, 62.1 [10.2] years and 3523 females [53%] in MESA; mean [SD] age, 56.3 [8.1] years; 169 648 females [55%] in the UKB), observed 10-year ASCVD event rates generally fell within the bounds of predicted risk categories regardless of Lp(a) level, although participants with elevated Lp(a) had higher event rates than did those with nonelevated Lp(a) (hazard ratio [HR], 1.30; 95% CI, 1.22-1.38) with similar results for CHD, HF, and total CVD. For CHD, the strongest association was among low-risk individuals (P for interaction = .31). The addition of elevated Lp(a) values to PREVENT modestly improved ASCVD risk prediction (category-free NRI, 0.058; 95% CI, 0.043-0.065; categorical NRI, 0.006, 95% CI, 0.004-0.011) with the greatest improvement in borderline-risk; when Lp(a) was evaluated continuously, the greatest improvement in prediction was among individuals at low risk. For CHD, the greatest improvement in prediction was in low- and high-risk individuals. CONCLUSIONS AND RELEVANCE: In this analysis of 2 cohort studies, the novel PREVENT equations performed well for risk prediction overall, including among individuals with elevated Lp(a). However, Lp(a) values remain independently associated with higher risk, and Lp(a) may improve personalized risk assessment, particularly among specific subgroups.

IMPORTANCE: Vascular cell adhesion molecule 1 (VCAM-1) and intracellular cell adhesion molecule 1 (ICAM-1) are responsible for immune cell-cell interactions. Systemic levels of VCAM-1 are associated with incident heart failure (HF). OBJECTIVES: To determine if VCAM-1 and ICAM-1 levels are associated with progression of established HF. DESIGN, SETTING, AND PARTICIPANTS: Participants enrolled in the biomarker substudy of the Dapagliflozin and Prevention of Adverse Outcomes in Heart Failure (DAPA-HF) randomized clinical trial had VCAM-1 and ICAM-1 levels measured at baseline and 12 months. The DAPA-HF trial was conducted at 410 sites in 20 countries. Patients with HF and reduced ejection fraction (HFrEF) in New York Heart Association (NYHA) class II to IV with elevated natriuretic peptides were enrolled between February 15, 2017, and August 17, 2018, with final follow-up on June 6, 2019. Data were analyzed from January 2023 to January 2025. INTERVENTIONS: Dapagliflozin, 10 mg, once daily vs placebo. MAIN OUTCOMES AND MEASURES: The primary outcome was the composite of a worsening HF event or cardiovascular death. The associations between VCAM-1 and ICAM-1 levels at baseline and the primary outcome, its components, and all-cause death were analyzed using Cox proportional hazards regression models adjusted for known prognostic variables including estimated glomerular filtration rate (eGFR), N-terminal pro-B-type natriuretic peptide (NT-proBNP), and high-sensitivity troponin T (hs-TnT), as well as high-sensitivity C-reactive protein. RESULTS: A total of 3051 participants (mean [SD] age, 67.2 [10.5] years; 2386 male [78.2%]) were included in this study. Mean (SD) follow-up time was 17.6 (5.2) months. The median (IQR) baseline VCAM-1 level was 997 (816.7-1218.8) ng/mL. Compared with patients with lower concentrations of VCAM-1, those with higher concentrations of VCAM-1 were older (mean [SD] age T3 vs T1, 69.7 [9.7] years vs 64.1 [10.7] years; P < .001), in worse NYHA class (T3 vs T1, NYHA class III/IV 35.6% [362 of 1017] vs 26.5% [269 of 1017]; P < .001), and had higher NT-proBNP (median [IQR] T3 vs T1, 2018 [1126-3753] pg/mL vs 1118 [693-1830] pg/mL) and hs-TnT (median [IQR] T3 vs T1, 24.7 [17.1-37.5] ng/L vs 16.6 [11.6-24.9] ng/L) concentrations, and lower eGFR (mean [SD] T3 vs T1, 58.4 [17.6] mL/min/1.73 m2 vs 71.7 [18.0] mL/min/1.73 m2). Patients in tertile 3 of VCAM-1, compared with tertile 1, had the highest risk of each outcome (eg, adjusted hazard ratio [HR] for primary outcome 1.40; 95% CI, 1.11-1.77; P = .004). ICAM-1 level was not associated with an elevated risk of any outcome. The benefit of dapagliflozin vs placebo in reducing the risk of the primary outcome was consistent across VCAM-1 tertiles: HR, 0.76 (95% CI, 0.54-1.06), 0.82 (95% CI, 0.59-1.12), and 0.77 (95% CI, 0.61-0.98) for tertiles 1, 2 and 3, respectively (P for interaction = .93). There was no significant change in VCAM-1 level with dapagliflozin at 52 weeks. CONCLUSIONS AND RELEVANCE: Results of this substudy of the DAPA-HF randomized clinical trial demonstrate that higher VCAM-1 levels, possibly reflecting a distinct inflammatory/immune pathophysiological pathway in HFrEF, were associated with worse outcomes, even after adjustment for conventional prognostic variables. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT03036124.