Predicting Success and Long-Term Outcomes of Percutaneous Mitral Valvuloplasty: A Multifactorial Score

Abstract

Background Percutaneous mitral valvuloplasty (PMV) success depends on appropriate patient selection. A multifactorial score derived from clinical, anatomic/echocardiographic, and hemodynamic variables would predict procedural success and clinical outcome. Methods Demographic data, echocardiographic parameters (including echocardiographic score), and procedure-related variables were recorded in 1085 consecutive PMVs. Long-term clinical follow-up (death, mitral valve replacement, redo PMV) was performed. Multivariate regression analysis of the first 800 procedures was performed to identify independent predictors of procedural success. Significant variables were formulated into a risk score and validated prospectively. Results Six independent predictors of PMV success were identified: age less than 55 years, New York Heart Association classes I and II, pre-PMV mitral area of 1 cm 2 or greater, pre-PMV mitral regurgitation grade less than 2, echocardiographic score of 8 or greater, and male sex. A score was constructed from the arithmetic sum of variables present per patient. Procedural success rates increased incrementally with increasing score (0% for 0/6, 39.7% for 1/6, 54.4% for 2/6, 77.3% for 3/6, 85.7% for 4/6, 95% for 5/6, and 100% for 6/6; P < .001). In a validation cohort (n = 285 procedures), the multifactorial score remained a significant predictor of PMV success ( P < .001). Comparison between the new score and the echocardiographic score confirmed that the new index was more sensitive and specific ( P < .001). This new score also predicts long-term outcomes ( P < .001). Conclusion Clinical, anatomic, and hemodynamic variables predict PMV success and clinical outcome and may be formulated in a scoring system that would help to identify the best candidates for PMV.