Phenotypic Diversity and Outcomes of Transcatheter Mitral Repair for Secondary Mitral Regurgitation: Insights From the COAPT and MITRA-FR Trials.

BACKGROUND

Mitral valve transcatheter edge-to-edge repair (M-TEER) is effective for treating patients with secondary mitral regurgitation (SMR), but therapeutic response varies because of anatomy and patient heterogeneity.

OBJECTIVES

The aim of this study was to identify distinct SMR phenotypes and analyze their response to M-TEER in the COAPT (Cardiovascular Outcomes Assessment of the MitraClip Percutaneous Therapy for Heart Failure Patients With Functional Mitral Regurgitation) and MITRA-FR (Percutaneous Repair or Medical Treatment for Secondary Mitral Regurgitation) trials.

METHODS

Unsupervised hierarchical clustering was performed using 22 baseline parameters (clinical, echocardiographic, and laboratory features). Clusters were identified and described with clinical risk (Society of Thoracic Surgeons [STS] repair score), comorbidity burden (Charlson Comorbidity Index [CCI]), and cardiac damage (CD; a 4-stage classification). The primary outcome was heart failure hospitalization or cardiovascular death at 2 years.

RESULTS

Clustering identified 3 distinct phenotypes in COAPT (n = 614): cluster 1 (21%), low clinical risk with high left ventricular CD (STS score 2.0%, CCI: 2.1 points, CD stage 2.0); cluster 2 (43%), intermediate clinical risk with biventricular CD (STS score 6.1%, CCI: 3.0 points, CD stage 2.5); and cluster 3 (36%), high clinical risk with low CD (STS score 7.6%, CCI: 3.3 points, CD stage 2.1) and the least severe SMR. Cluster 1 had the highest 2-year survival probability among the 3 groups (P < 0.001). Following M-TEER, 2-year freedom from heart failure hospitalization and cardiovascular death was higher in clusters 1 (relative risk reduction 46%; P = 0.026) and 2 (relative risk reduction 49%; P < 0.001) compared with cluster 3 (relative risk reduction 34%; P = 0.033). A survival benefit was seen in clusters 1 (P = 0.047) and 2 (P = 0.018) but not in cluster 3 (P = 0.22). Validation in MITRA-FR showed similar findings, with cluster 3 being more prevalent than in COAPT (48% vs 36%).

CONCLUSIONS

Clustering analysis identified 3 distinct SMR phenotypes that were validated across 2 landmark M-TEER trials. Outcomes following M-TEER are better predicted by multidimensional patient phenotypes integrating CD, surgical risk, and comorbidity burden than by any single domain alone and may help explain differences in treatment effect observed across trials.