Late Results of Percutaneous Mitral Commissurotomy up to 20 Years: Development and Validation of a Risk Score Predicting Late Functional Results From a Series of 912 Patients

Study Questions:

What clinical and procedural parameters predict a good clinical outcome following balloon percutaneous mitral commissurotomy (PMC)?


Data were available for 1,024 consecutive patients undergoing PMC at a single center who were followed for up to 20 years. A good functional result was defined as survival with no cardiovascular death or need for further mitral valve (MV) intervention, and in New York Heart Association (NYHA) class I or II. Other endpoints included overall survival, survival free from cardiovascular death, and survival with no cardiovascular death or repeat mitral intervention. Valves were characterized by the Cormier classification: 1) flexible valves and mild subvalvular disease; 2) flexible valve and extensive subvalvular disease; and 3) calcified valves on fluoroscopy. MV area (MVA) was determined by planimetry of two-dimensional echocardiographic images.


For the entire cohort, the average patient age was 49 ± 14 years and 83% were female. Atrial fibrillation was noted in 40% and the distribution of Cormier valve anatomy was 14%, 55%, and 31% for groups 1, 2, and 3. A good immediate result was defined as a final MVA ≥1.5 cm2 without mitral regurgitation >2/4. A poor immediate result was noted in 112 patients. Good immediate results were noted in 912 patients, who subsequently were divided to a derivation cohort of 609 patients for modeling predictive variables and a validation cohort of 303. Cardiovascular death occurred in 17% of those with poor immediate results compared with 8.6% of those with good immediate results, and 58% of those with poor immediate results required subsequent MV replacement compared with 27% of those with good immediate results. Sustained good functional results were noted in 43%, 39%, and 15% of Cormier group 1, 2, and 3 patients (p < 0.001). Multivariable analysis, taking into account interactions, identified seven predictive factors for a good functional result. These included age and its interaction with MVA, valve calcification, and its interaction with gender, NYHA class, and its interaction with atrial fibrillation, and final mean MV gradient (≤3, 3-6, ≥6 mm Hg). A total of 11 subsets among the 7 variables were identified, each of which was given a score of 0-5, from which a global score was calculated. Patients were grouped into a low- (0-2), intermediate- (3-5), and high-risk global score (6-13). Low-, intermediate-, and high-risk groups comprised 182 (30%), 286 (47.2%), and 138 (22.8%) of the patients, who accounted for 48, 153, and 108 events, respectively, during follow-up. The hazard ratio for poor late functional results was 2.8 for the intermediate-risk group when compared to the low-risk group, and was 7.4 for the high-risk group (both p < 0.0001). Distribution of scores and predictive ability of the score was statistically identical in the validation cohort compared to the derivation cohort.


With follow-up of up to 20 years, 30% of patients with an initially good functional result from PMC maintained a good functional result. Multiple parameters predicted late functional results, which were strongly influenced by age and the quality of the immediate PMC results.


This large study from a single center nicely demonstrates that in a sizable proportion of patients, long-term results from percutaneous balloon valvotomy for rheumatic MV stenosis can result in a good prolonged functional result. Multiple parameters were identified, which predicted a persistent improvement over time, including age, NYHA class, and immediate post-procedure gradient. A number of interactions were noted, including the fact that an initially good MVA is a strong predictor of late functional results in younger patients, but no longer significant in patients after the age of 70. Similarly, final mean gradient was a predictor of long-term good functional results, but conversely, did not vary according to age. The scoring system proposed by the authors is relatively straightforward and identified a high-risk group with a low likelihood of long-term good functional results, who perhaps would be better served by a traditional MV replacement rather than an attempt at PMC.

Clinical Topics: Invasive Cardiovascular Angiography and Intervention, Valvular Heart Disease, Interventions and Structural Heart Disease, Mitral Regurgitation

Keywords: Balloon Valvuloplasty, Mitral Valve Insufficiency, Cardiology, Mitral Valve Stenosis, Calcinosis, Angioplasty, Balloon, Coronary

< Back to Listings