Is it worth repairing rheumatic mitral valve disease in children? Long-term outcomes of an aggressive approach to rheumatic mitral valve repair compared to replacement in young patients†

OBJECTIVES: Contemporary experience in mitral valve (MV) repair for children with rheumatic heart disease (RHD) is limited, despite the potential advantages of repair over replacement. We reviewed our long-term outcomes of rheumatic MV repair and compared them with the outcomes of MV replacement in children with RHD. METHODS: This study is a review of 419 children (≤18 years) with RHD who underwent primary isolated MV surgery between 1992 and 2015, which comprised MV repair (336 patients; 80.2%) and MV replacement (83 patients; 19.8%). The replacement group included mechanical MV replacements (MMVRs) (n = 69 patients; 16.5%) and bioprosthetic MV replacements (n = 14 patients; 3.3%). The mean age with standard deviation at the time of operation was 12.5 ± 3.5 (2-18) years. Mitral regurgitation (MR) was predominant in 390 (93.1%) patients, and 341 (81.4%) patients showed ≥3+ MR. The modified Carpentier reconstructive techniques were used for MV repair. RESULTS: Overall early mortality was 1.7% (7 patients). The mean follow-up was 5.6 years (range 0-22.3 years; 94.7% complete). Survival of patients who underwent repair was 93.9% both at 10 and 20 years, which was superior than that of replacement (P < 0.001). Freedom from reoperation at 10 and 20 years after MV repair was 81.7% and 72.6%, respectively, compared to 83.2% for MV replacement (P = 0.580). Forty patients underwent reoperation after the initial surgery with no operative deaths. Mixed mitral lesion and postoperative residual MR (≥2+) were the predictors for reoperation in the repair group, whereas lower body surface area and usage of bioprosthesis were significant factors for the replacement group. Freedom from thrombotic, embolic and haemorrhagic events at 10 and 20 years for patients with repair was 98.2% compared to 90.1% in patients with replacement and 67.6% for patients with MMVR (P = 0.004). CONCLUSIONS: Twenty-three years of follow-up shows that MV repair is superior to MMVR in children with RHD. Hence, the rheumatic MV should be repaired when technically feasible to maximize the survival and reduce the valve-related morbidity with comparable durability to MMVR.