Background:
Functional mitral regurgitation (MR) grading by echocardiography is often challenging, and cardiac magnetic resonance (CMR) may provide more accurate and reproducible assessment, however optimal CMR thresholds have not been established. We aimed to determine clinically significant thresholds of MR-volume and MR-fraction by CMR in patients with non-ischemic cardiomyopathy (NICM).
Methods:
All patients having CMR with confirmed NICM between 2001/4/1-2019/3/31 were retrospectively studied. Time to death, heart transplant and/or left ventricular assist device is the primary endpoint, and receiver-operative (ROC) characteristics analyses were used to identify significant cut-points for MR-volume and MR-fraction.
Results:
Mean MR-volume was 11±12% and MR-fraction 14±13% in 840 NICM patients. C-statistics for the primary endpoint (141 events) were MR-volume 0.629 (95%CI: 0.577-0.691) and MR-fraction 0.594 (95%CI 0.542-0.647). Thresholds of MR-volume≥10mL and MR-fraction≥15% had the highest products of sensitivities and specificities (57.4%, 56.8%; and 54.6%, 63.4% respectively) reflecting elevated risk. MR-volume≥35mL and MR-fraction≥40% were associated the best combination of positive and negative predictive values (40.1%, 84.2%; and 38.4%, 84.6% respectively) representing severe risk. The figure illustrates Kaplan-Meier curves of the primary endpoint, categorized into three subgroups based on aforementioned MR-volume and MR-fraction thresholds.
Conclusion:
Adverse prognosis are observed when MR-volume≥10mL or MR-fraction≥15%, with increasing risk of adverse outcomes as functional MR severity increased in NICM. The proposed MR thresholds from this study for elevated and severe risk are lower than current valvular regurgitation guidelines, suggesting the need for separate classification schema for functional MR.