Pre-procedural risk stratification for percutaneous mitral valve repair: a mortality score integrating baseline risk and procedural success
P Stoeckl, I Toskas, J Fischer Pireddu, M Bladt, P Jaeger, I Buchner, J Henes, M Gawaz, T Geisler, D RathAbstract
Background
Mitral transcatheter edge-to-edge repair (M-TEER) is an established treatment for symptomatic mitral regurgitation (MR) in patients at high surgical risk. This study aimed to develop a risk stratification tool to predict one year-mortality and to investigate how baseline risk modifies the prognostic impact of procedural success.
Purpose
To (1) develop the CHAMPIONS risk score for predicting one-year mortality after M-TEER and (2) evaluate the interaction between baseline risk and residual MR.
Methods
Score development followed a machine learning-guided approach. A random forest classifier was trained on 80% of the overall cohort (n=939). The remaining 20% served as an internal validation subset. The top 20 variables were selected for logistic regression. Regression coefficients were translated into a point-based score (0-14 points), with higher beta-coefficients corresponding to higher point allocations. The score incorporates nine routinely available parameters. Patients were categorized into low, intermediate and high-risk groups based on their total score.
Results
The CHAMPIONS score demonstrated moderate discrimination (AUC: 0.67) and excellent calibration (Hosmer-Lemeshow test: x2 = 0.41, p=0.995). One-year mortality rates increased stepwise across risk categories: low (7.6%), intermediate (16.7%) and high risk (30.9%) (log-rank p<0.001). A significant interaction was found between baseline risk and residual MR. Achieving residual MR ≤ 1 was associated with survival benefits primarily in the intermediate-risk group (mortality 10.4% vs 25.1%, p<0.001). Residual MR did not significantly change survival for patients in the low- or high-risk categories.
Conclusion
The CHAMPIONS score is a simple tool for pre-procedural risk stratification in M-TEER patients. It identifies intermediate-risk patients as the primary group were achieving residual MR ≤ 1 leads to a significant survival benefit. External validation is needed to confirm its generalizability and refine its predictive performance.For image description, please refer to the figure legend and surrounding text.For image description, please refer to the figure legend and surrounding text.