Feature Tracking-Derived Global Longitudinal Strain Enhances Risk Stratification for Sudden Cardiac Death in Hypertrophic Cardiomyopathy.

Abstract

BACKGROUND Left ventricular (LV)-global longitudinal strain (GLS) assessed by cardiac magnetic resonance (CMR) feature tracking is an emerging marker for predicting adverse outcomes in hypertrophic cardiomyopathy (HCM), but its incremental prognostic value and mechanistic role in sudden cardiac death (SCD) risk stratification remain unclear. OBJECTIVES The study sought to evaluate whether LV-GLS adds prognostic value beyond current ESC (European Society of Cardiology) and ACC (American College of Cardiology)/AHA (American Heart Association) SCD risk models, and mediates the relationship between myocardial abnormalities and SCD risk in HCM. METHODS The authors retrospectively analyzed 2,009 patients with HCM (mean age: 50 ± 14 years, 70% men) who underwent CMR between 2010 and 2017. LV-GLS was quantified using cine CMR feature tracking. The primary endpoint included SCD and aborted SCD. Prognostic performance was assessed using time-dependent receiver-operating characteristic analysis and competing risk regression. Mediation analysis was used to investigate how LV-GLS mediated associations between myocardial hypertrophy, fibrosis, and SCD. RESULTS Over a median follow-up of 88.2 months, 85 (4.2%) patients experienced SCD events. These patients had significantly lower absolute LV-GLS values (9.0% ± 3.6% vs 11.1% ± 3.6%; P < 0.001). In competing-risk regression, LV-GLS independently predicted SCD after adjustment for ESC (subdistribution HR [sHR]: 1.12 per 1% decrease [95% CI: 1.06-1.22]; P < 0.001) and ACC/AHA risk factors (sHR: 1.09 [95% CI: 1.02-1.18]; P = 0.016). Adding LV-GLS improved the 5-year predictive accuracy of both ESC and ACC/AHA models (AUC from 0.72 to 0.77 and from 0.71 to 0.76, respectively). Absolute LV-GLS with a cutoff of 9.23% further stratified risk in patient subgroups with either class II or class III implantable cardioverter-defibrillator indications (all log-rank P < 0.001). Mediation analysis showed LV-GLS partially mediated the effect of maximum wall thickness and extent of fibrosis on SCD (proportion-mediated: 17.5% and 23.1%, respectively; both P < 0.001). CONCLUSIONS In patients with HCM, CMR-derived LV-GLS is an incremental predictor of SCD beyond current guideline-based risk models and partially mediates the association between myocardial abnormalities and SCD.