A novel prediction model for heart failure-related death in hypertrophic cardiomyopathy
Y Zhang, Y Guan, Y Zheng, F Qin, X Wang, Y Ge, M Zhang, X Ma, L SongAbstract
Background
Heart failure (HF) is one of the major causes of death in patients with hypertrophic cardiomyopathy (HCM). However, a validated prediction model for HF-related death in this population remains unavailable.
Purpose
This study aimed to develop and validate a prediction model for HF-related death in patients with HCM.
Methods
A total of 4,020 patients with HCM from five medical centers were enrolled. The development cohort contained 2,885 patients from one large center, and other four independent centers (n=1,135) served for external validation. The endpoint was HF-related death. A multivariable Cox proportional hazards model was constructed to develop the prediction model. External validation was conducted to assess model generalizability in independent cohorts. Model performance was evaluated using the time-dependent area under the receiver operating characteristic curve (AUROC).
Results
During a median follow-up of 5.3 years, HF-related death occurred in 57 patients (2.0%) in the development cohort. In the combined external validation cohort, 20 patients (1.8%) experienced HF-related death over a median follow-up of 3.5 years. The final HF-related death prediction model incorporated age at visit, New York Heart Association functional class, systolic blood pressure, N-terminal pro-B-type natriuretic peptide, left ventricular ejection fraction (LVEF), and left atrial diameter. The model achieved AUROCs of 0.88 (0.81−0.95) and 0.86 (0.81−0.92) for predicting HF-related death at 1 and 5 years in the development cohort, and 0.90 (0.80−0.99) and 0.85 (0.73−0.97) in the external validation cohort. Among patients with preserved LVEF, the model also demonstrated good performances at both time points, with AUROCs of 0.85 (0.77−0.93) and 0.86 (0.79−0.92) in the development cohort, and 0.86 (0.74−0.99) and 0.80 (0.64−0.96) in the external validation cohort, respectively.
Conclusions
In this multicenter study, we developed and externally validated a six-variable prediction model for HF-related death in patients with HCM. The model also showed robust performance in patients with preserved LVEF.graphical abstract