Prognostic Value and Thresholds for Functional Mitral Regurgitation in Non-Ischemic Cardiomyopathy by Magnetic Resonance Imaging
Author: Tom Kai Ming Wang, MBChB, MD(res),
Additional Authors**
See the oral presentation (begins at minute 17:50)
Background: Functional mitral regurgitation (FMR) is clinically important because of its poor prognosis and the rapidly evolving evidence for mitral valve interventions for its management. Echocardiography evaluation of FMR remains challenging, while the clinical utility and thresholds for cardiac magnetic resonance imaging (CMR) quantification of FMR are not well-established.
Objective: We evaluated the associations and thresholds of CMR-derived FMR-volume and fraction with adverse outcomes in patients with non-ischemic cardiomyopathy (NICM).
Methods: Consecutive NICM patients undergoing CMR during 2001/4/1-2019/3/31 were retrospectively studied. Multivariable Cox regression was performed for FMR quantification parameters and three composite endpoints: primary (time to death, heart transplant and/or left ventricular assist device), secondary A (primary endpoint and ventricular arrhythmia hospitalization), and secondary B (primary endpoint and heart failure hospitalization). Clinically significant and severe FMR cut-points for predicting the primary endpoint were determined.
Results: Mean MR-volume was 11±12% and MR-fraction was 14±13% in 840 NICM patients. There were 141 (16.8%) primary endpoints, over a mean follow-up of 3.6±3.0 years. In multivariable analysis, both MR-volume and MR-fraction were independently associated with the three composite endpoints after adjusting for clinical and other CMR parameters, for example with hazards ratios of 1.13 (95%CI 1.06-1.19), 1.12 (95%CI 1.06-1.18) respectively per 5% increase of MR-fraction. Increasing age and New York Heart Association Class were also associated with all three composite endpoints, while male sex, late gadolinium enhancement (LGE) and diabetes were associated with the primary endpoint, secondary endpoint A and both of these endpoints respectively. The clinically significant FMR cut-points were MR-volume and fraction of 10mL and 15% respectively, and severe FMR thresholds were 35mL and 40% respectively. There was a significant interaction between MR-fraction and left ventricular end-diastolic volume indexed (LVEDVi) for the primary endpoint (P=0.021), with accentuated hazards when LVEDVi≥100ml/m2, but not with LGE.
Conclusion: FMR volume and fraction by CMR were independently associated with adverse outcomes in NICM patients, and FMR severity appears to be more prognostically important than chamber quantification. The CMR-thresholds identified for FMR severity are lower than contemporary valvular heart disease echocardiography guidelines, and may help inform the modality and timing of FMR management.
**Additional Authors:
Tom Kai Ming Wang, MBChB, MD(res), Cleveland Clinic, Cleveland, Ohio
Duygu Kocyigit, MD, Cleveland Clinic, Cleveland, Ohio
Harry Choi, MD, Cleveland Clinic, Cleveland, Ohio
Zoran Popovic, MD, Cleveland Clinic, Cleveland, Ohio
Wilson Tang, MD, Cleveland Clinic, Cleveland, Ohio
Brian Griffin, MD, Cleveland Clinic, Cleveland, Ohio
Scott Flamm, MD, Cleveland Clinic, Cleveland, Ohio
Deborah Kwon, MD, Cleveland Clinic, Cleveland, Ohio
