Validation of MitraScore in diverse mitral regurgitation subtypes: Insights from the OCEAN-mitral registry.

Background: MitraScore is a newly developed tool for predicting mortality risk in patients undergoing mitral transcatheter edge-to-edge repair (M-TEER). Although its performance has been demonstrated in a derivation study, external validation beyond functional mitral regurgitation (MR) remains limited.

Objective: To externally validate the performance of MitraScore across diverse MR subtypes.

Methods: Using the dataset from the OCEAN-Mitral registry, an ongoing multicenter study recruiting consecutive patients undergoing M-TEER, discrimination and calibration of MitraScore were assessed with Harrell's C-statistic and Gronnesby-Borgan goodness-of-fit tests, respectively. The primary outcome was all-cause mortality. Risk stratification was evaluated by classifying patients into low-risk (MitraScore: 0-2 points), moderate-risk (3-4 points), and high-risk (5-8 points) groups.

Results: In the entire population (n = 3745; age, 78.8 ± 9.6; female 45.1 %; median follow-up, 1.2 years), MitraScore showed moderate discrimination (C-statistic: 0.63) and good calibration (Chi-squared value: 2.86; P = 0.41) to predict mortality. Compared to MitraScore low-risk patients, moderate- (hazard ratio [HR], 1.92; 95 % confidence interval [CI], 1.59-2.33) and high-risk patients (HR, 4.16; 95 % CI, 3.25-5.33) showed significantly higher mortality risk. These performances were consistent in subgroup analyses of different etiologies of MR, anatomical features, and cases with hemodynamic instability. The addition of novel independent predictor, "Clinical Frailty Scale ≥ 4 points", improved the performance of conventional MitraScore, outperforming STS score, and EuroSCOREII.

Conclusions: This was the largest external validation study for MitraScore, demonstrating moderate discrimination, good calibration, and effective risk stratification across diverse MR types. Frailty assessment enhances predictive value, although additional validation is required. (250/250).