Using Electrocardiogram to Assess Diastolic Function and Prognosis in Mitral Regurgitation.

Background: The determination of left ventricular diastolic function (LVDF) in patients with significant (≥moderate) mitral regurgitation (MR) poses a complex challenge. We recently validated an artificial intelligence-enabled electrocardiogram (AI-ECG) algorithm to estimate LVDF.
Objectives: This study sought to evaluate the risk of all-cause mortality across AI-ECG LVDF-derived myocardial disease (MD) grades in MR.
Methods: This was a retrospective study including all patients in the AI-ECG LVDF study testing group who underwent comprehensive transthoracic echocardiography confirming significant MR and electrocardiogram within 14 days of each other at the Mayo Clinic between September 2001 and April 2023. AI-ECG LVDF status was determined based on the index electrocardiogram and used to categorize patients into 3 stages of MD: MD-1, normal or grade 1 LVDF; MD-2, grade 2 LVDF; and MD-3, grade 3 LVDF.
Results: Of 4,019 patients with significant MR (mean age 69.8 years; 49.0% women), 1,175 (29.2%), 1,881 (46.8%), and 963 (24.0%) were classified by AI-ECG as MD-1, MD-2, and MD-3, respectively. The median mitral effective regurgitant orifice area was 26 mm ² (Q1-Q3: 20-36 mm ² ). Over a median follow-up of 3.5 years, 1,636 (40.7%) patients died. In multivariable survival analysis adjusted for multiple risk factors, a higher diastolic function grade was independently associated with an increased death risk (MD-2, adjusted HR [aHR]: 1.99; 95% CI: 1.62-2.45; MD-3, aHR: 2.65; 95% CI: 2.11-3.34). These findings were consistent when accounting for mitral valve intervention and across various sensitivity and subgroup analyses.
Conclusions: In patients with significant MR, the grading of LVDF by AI-ECG is independently associated with all-cause mortality.
Competing Interests: Funding Support and Author Disclosures This work was funded by the Mayo Clinic. This work was partly supported by the David M.C. Ju Foundation, a nonprofit organization. The study's funders had no role in study design, data collection, data analysis, data interpretation, or writing of the report. Drs Lee, Lopez-Jimenez, Friedman, and Attia Oh have invented algorithms licensed to ANUMANA and may benefit from algorithm commercialization via Mayo Clinic. Drs Lopez-Jimenez, Asirvatham, Friedman, and Attia have served on the scientific advisory board for ANUMANA. Dr Oh has served as a consultant for Medtronic's valve projects. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
(Copyright © 2024 American College of Cardiology Foundation. Published by Elsevier Inc. All rights reserved.)