Reconstructive surgery in congenital mitral valve insufficiency (Carpentier's techniques): Long-term results

Background: Previous publications have stressed the benefits of mitral valve repair over mitral valve replacement in children. However, few communications have reported the long-term results and none with a follow-up of more than 10 years. This article reports our results in a series of 145 patients operated on for congenital mitral valve insufficiency by means of the same technique (Carpentier's technique) in a single center. Methods: Between 1970 and 1995, 145 patients younger than 12 years old underwent surgery for congenital mitral valve insufficiency. Mean age was 5.7 +/- 3.1 years, ranging from 0.17 to 12 years. Mitral valve insufficiency associated with atrioventricular defect, atrioventricular discordance, straddling mitral valve, acquired diseases, Marfan syndrome, and degenerative disease was excluded from this study. According to Carpentier classification, 31 patients had type I mitral valve disease (normal leaflet motion), 79 patients type II (leaflet prolapse), and 35 type III (restricted leaflet motion), with 15 having normal papillary muscles and 20 abnormal papillary muscles. Associated lesions were present in 51 patients (35%). A conservative operation was possible in 138 patients (95%). Among them, 70 patients required a prosthetic annuloplasty and 21 patients valve extension with a pericardial patch. Valve replacement was necessary in seven patients (5%). Results: In-hospital mortality was 5% (95% CI: 2.5% to 9.9%) (seven patients). No early death was observed in the group of patients who underwent valvular replacement. In-hospital mortality was as follows: type I, 9.6%; type II, 2.5%; and type III, 13%. No statistically significant difference was noted among patients with the different types of disease. Mean follow-up was 9.3 +/- 6.9 years (1 to 26 years), and cumulative follow-up was 1142 patient-years. Ten late deaths occurred. Actuarial survival at 10 years was 88% in patients who underwent valve repair and 51% in patients who underwent valve replacement. Late reoperation was required in 15% (n = 21) of patients who had undergone valve repair and 28% (n = 2) in patients with valve replacement. Causes of reoperation were recurrent left ventricular failure (n = 1), residual or recurrent mitral valve insufficiency (n = 17), mitral valve stenosis (n = 3), and calcification of the bioprosthesis (n = 2). No failure resulting from leaflet extension was observed. In the repair group, actuarial freedom from reoperation was 68% (95% CL: 80.5% to 51.5%) at 15 years, and the linearized rate of exposure to reoperation was 1.9% per patient-year. No thromboembolic event was observed in any group. Conclusion: Congenital mitral valve insufficiency can be repaired in infancy with a low mortality. Conservative surgery with Carpentier's techniques is feasible in the majority of cases of congenital mitral valve insufficiency. This technique offers stable long-term results with a low rate of reoperation. Leaflet extension, associated with prosthetic ring annuloplasty could prevent reoperations in selected cases.