Phenotyping of Mitral Valve Prolapse Without Severe Mitral Regurgitation Using Electrocardiographic and Echocardiographic Data (JCEP October 2025)-Article: Phenotyping of Mitral Valve Prolapse Without Severe Mitral Regurgitation Using Electrocardiographic and Echocardiographic Data

Article: Phenotyping of Mitral Valve Prolapse Without Severe Mitral Regurgitation Using Electrocardiographic and Echocardiographic Data

Back to course

Pdf Summary

This study investigates phenotyping of mitral valve prolapse (MVP) without severe mitral regurgitation (MR) using standard electrocardiographic (ECG) and echocardiographic data through an unsupervised machine learning approach. MVP, affecting 2-3% of the population, is often benign without severe MR, but a subset is at risk of ventricular arrhythmias and sudden cardiac death (SCD). Traditional risk factors capture only a minority of these at-risk patients, motivating new methods for risk stratification.<br /><br />Researchers analyzed 343 MVP patients with trace to mild MR and comprehensive ECG and echocardiographic data from the University of California-San Francisco registry (2013-2022). Applying hierarchical clustering on 32 demographic, ECG, and echocardiographic variables, three distinct phenotypic clusters emerged: Cluster 1 (83% of cases) with normal left atrial (LA), left ventricular (LV), and right ventricular (RV) function (low risk); Cluster 2 (9%) showed intermediate risk with prolonged ECG intervals, increased LV mechanical dispersion, RV dysfunction, and elevated pulmonary pressures; Cluster 3 (8%) had advanced LA and LV remodeling and dysfunction, despite mostly mild MR, representing a high-risk group.<br /><br />Arrhythmic events (including sudden cardiac arrest, ventricular tachycardia/fibrillation, frequent premature ventricular contractions) occurred in 19% of Cluster 1, and were significantly higher in Clusters 2 (38%) and 3 (43%). Mortality was also elevated in Clusters 2 and 3 compared to Cluster 1, with adjusted hazard ratios around 5 for all-cause death. Key features driving clustering included ECG intervals (QRS, QTc), LA systolic strain, LA function index, and RV systolic function, indicating the importance of electrical and structural abnormalities beyond mitral valve morphology.<br /><br />This work highlights that standard ECG and echocardiographic metrics, analyzed with machine learning, can reveal MVP subgroups with distinct arrhythmic and mortality risks, independent of MR severity. Notably, the study identifies a “primary atriopathy” (LA dysfunction) as a novel risk factor. These insights support more personalized MVP management and suggest that traditional focus on valve anatomy alone may miss broader myocardial and atrial involvement linked to arrhythmic risk. The approach requires external validation but offers a practical, noninvasive stratification framework for MVP patients without severe MR.

Keywords

mitral valve prolapse

mitral regurgitation

electrocardiogram

echocardiography

machine learning

phenotyping

arrhythmic risk

left atrial dysfunction

ventricular arrhythmias

risk stratification