Development and Validation of a Risk Score Predicting Late Functional Results From a Series of 912 Patients

Background—Long-term follow-up after percutaneous mitral commissurotomy enables predictive factors of late results to be identified. Methods and Results—Late results of percutaneous mitral commissurotomy were assessed in 1024 consecutive patients. Good immediate results, defined as valve area 1.5 cm without mitral regurgitation 2/4, were obtained in 912 patients (89%). These 912 patients were randomly split into 2 cohorts comprising 609 and 303 patients that were used to develop and validate, respectively, a scoring system predicting late functional results. The 20-year rate of good functional results (survival without cardiovascular death, mitral surgery, or repeat percutaneous mitral commissurotomy and in New York Heart Association class I or II) was 30.2 2.0%. A multivariable Cox model identified 7 predictive factors of poor late functional results: higher final mean gradient (P 0.0001), interaction between age and final mitral valve area (P 0.0001) showing that the impact of valve area decreases with age, interaction between sex and valve calcification (P 0.0001) showing that the impact of valve anatomy is stronger in men, and interaction between rhythm and New York Heart Association class showing an impact of New York Heart Association class only in patients in atrial fibrillation (P 0.0001). A 13-point score enabled 3 risk groups to be defined, corresponding to predicted good functional results of 55.1%, 29.1%, and 10.5% at 20 years in the validation cohort. Conclusions—Twenty years after percutaneous mitral commissurotomy in a population of patients with varied characteristics, 30% still had good functional results. Prediction of late functional results is multifactorial and strongly determined by age and the quality of immediate results. A simple validated scoring system is useful for estimating individual patient outcome. (Circulation. 2012;125:2119-2127.)