Background: There is a lack of specific tools for risk assessment of patients receiving transcatheter edge-to-edge repair (M-TEER) for severe mitral regurgitation (MR).

Objective: The study aims to develop a multivariable risk model to predict long-term survival after successful M-TEER.

Methods: Consecutive patients with severe MR (grade III+) undergoing successful M-TEER between 08/10 and 04/21 were included and followed prospectively. Primary endpoint was all-cause mortality after a median follow-up period of 19.6±10 months. A lasso-penalized logistic regression model was used to identify independent predictors for primary endpoint. Discrimination and calibration of the developed risk model was assessed using receiver operating characteristic (ROC) analysis and calibration plots. Model validation was performed using internal bootstrap resampling. 

Results: 589 patients (77±8.4 years, 56.9% male) were enrolled. All-cause mortality rate was 33.6% (n=189). Five independent predictors of all-cause mortality were identified: NT-proBNP levels, left ventricular stroke work index (LVSWi), the ratio of mean arterial pressure to right atrial pressure (MAP/RAP ratio), tricuspid annular plane systolic excursion (TAPSE) and chronic obstructive lung disease (COPD). The developed risk model showed good discriminative capacity (Area under the curve [AUC] 0.72) even after internal bootstrap resampling validation, whereas the discriminative ability of the Logistic EuroSCORE was modest (AUC 0.61). Calibration plots of this novel risk model showed a good agreement between predicted to observed probabilities. 
Conclusions: A tailored risk model instead of classical surgical risk scores may be useful for risk assessment of patients undergoing M-TEER. External validation of study results for broad clinical application is needed.