Novel Prediction Model Accurately Forecasts Early LV Systolic Dysfunction Progression in HCM Patients

Hypertrophic cardiomyopathy (HCM) is a common inherited myocardial disorder characterized by left ventricular hypertrophy, often progressing to heart failure (HF). A critical and heterogeneous risk in HCM patients is early left ventricular systolic dysfunction progression (ELVSDP), which significantly impacts prognosis. Current clinical practice lacks accurate short- and medium-term prediction tools to identify patients at high risk of ELVSDP, hindering timely intervention and individualized management strategies. This gap necessitates a robust predictive model to improve risk stratification.

Researchers enrolled 314 HCM patients without ELVSDP at baseline, randomly dividing them into training and validation sets. They used LASSO-Cox regression to identify independent predictors of ELVSDP. Based on these predictors, a nomogram and an interactive dynamic prediction tool were developed. Model performance was rigorously evaluated using the concordance index (C-index), time-dependent receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). Risk stratification was further assessed via the Kaplan-Meier method.

The study identified several independent predictors for ELVSDP: age (HR = 1.17), smoking history (HR = 2.79), B-type natriuretic peptide (BNP) level (HR = 1.002), and left ventricular outflow tract obstruction (HR = 2.24). The developed model demonstrated strong predictive performance in both the training and validation sets.

The time-dependent area under the curve (AUC) for the model exceeded 0.88 at 6, 12, and 18 months, with C-indices of 0.94 and 0.93 in the training and validation sets, respectively.

Bootstrap validation confirmed the model's stability, and calibration curves showed excellent agreement between predicted and observed outcomes. DCA further indicated a significant net clinical benefit. Crucially, the incidence of ELVSDP was significantly higher in the high-risk group compared to the low-risk group (P < 0.0001), underscoring the model's ability to stratify risk effectively.