PO47 Integrating Genetic Variants into Hypertrophic Cardiomyopathy Risk Stratification: A Quantitative and Comparative Analysis

Background Hypertrophic cardiomyopathy (HCM) is often associated with pathogenic or likely pathogenic (P/LP) genetic variants and variants of unknown significance (VUS). The prognostic value of gene-positive status, especially in the context of ESC Sudden Cardiac Death (SCD) risk scoring, remains underexplored. Objective To evaluate (1) the prevalence of P/LP variants and VUS in a cohort of HCM patients, and (2) the prognostic role of these variants individually or integrated into existing risk stratification models. Methods Single-center prospective cohort of 92 HCM patients who underwent genetic testing between January 2018 and November 2024. The prevalence of P/LP variants and VUS was determined. Our primary endpoint was a composite of cardiovascular events that included admission for heart failure, ventricular fibrillation/tachycardia, cardiac syncope, cardiovascular cardiac death, myocardial infarction, ischemic stroke and/or complete heart block. A combined prognostic model was built by adding P/LP status to the ESC SCD risk score; logistic regression was performed to determine the optimal weight for P/LP status. Discrimination performance was assessed using C-statistics for the ESC SCD model alone and the combined model. Integrated discrimination improvement (IDI) and calibration analysis were conducted to assess the additional value of genetic data. The net clinical benefit of the baseline and combined models was compared across thresholds using decision curve analysis (DCA). Results A total of 71 HCM patients (65±12 years, 49.3% female) were suitable for analysis, 77.5% septal HCM, 15.5% apical and 7% mixed disease. A mutation was observed in 49.3% of the patients, 66% of which were LP/P variants, 34% were VUS. The most frequent mutations involved the gene MYBPC3. Primary endpoint was met in 57.7% of patients. Mean ESC SCD risk scores were 3.40% for pathogenic, 2.33% for VUS, and 2.52% for gene-negative individuals (p > 0.05 for all comparisons). The baseline ESC SCD model achieved a C-statistic of 0.569; including a weighted P/LP variable (weight = 0.036) resulted in a C-statistic of 0.567. IDI for the weighted model was 0.001, with a modest improvement for events (+1.2%) and a decline for non-events (-1.1%). Calibration analysis showed minimal differences between the weighted model and ESC SCD risk score alone. Decision curve analysis highlighted marginal improvement in net benefit at lower risk thresholds when genetic data were included. Conclusions Genetic testing reveals a prevalence of P/LP variants in 1 every 3 HCM patients, but provides limited independent prognostic value. The ESC SCD risk score remains the dominant predictor of cardiovascular events, with minor incremental benefits from incorporating genetic pathogenicity. Larger studies are needed to validate these findings and explore advanced integration of genetic data into clinical practice.