Very-long-term outcomes of mechanical valves in mitral position focusing on valve-related complications

OBJECTIVES: This study aimed to examine very-long-term outcomes of a mechanical valve at the mitral position. METHODS: This study included all patients who underwent mitral valve replacement (MVR) using a mechanical valve including urgent operation at the Department of Cardiovascular Surgery, Hyogo Brain and Heart Center, Himeji, from January 1987 to December 2015. RESULTS: Five hundred and eighty-three patients (277 men [47.51%]; age, 61 [54-67] years) were included in this study. The implanted valve models were as follows: SJM, 221 (37.91%); ATS, 35 (6.00%); On-X, 68 (11.66%); and Carbomedics 194, (33.28%).The median clinical follow-up duration was 13.3 (7.4-19.6) years. The survival rates at 10, 15, 20 and 25 years were 81.42%, 69.27%, 56.34% and 45.03%, respectively. Thromboembolism was observed in 38 patients, and the linearized ratio for each event was 0.626%/patient-year [95% confidence interval (CI), 0.443-0.859%]. Intracranial haemorrhage and gastrointestinal bleeding were observed in 26 and 9 patients, and the linearized ratio for each event was 0.425%/patient-year (95% CI, 0.277-0.006%) and 0.145%/patient-year (95% CI, 0.067-0.276%), respectively. Major paravalvular leak was observed in 32 patients, and the linearized ratio was 0.532%/patient-year (95% CI, 0.364%-0.751%). The cumulative incidence rate of major paravalvular leak at 10, 15, 20 and 25 years was 3.7%, 5.6%, 6.4% and 10.4%, respectively. Multivariable Cox regression analysis revealed that repeated MVR and male gender were associated with major paravalvular leak. CONCLUSIONS: Male gender and repeated MVR were risk factors for paravalvular leak after mechanical MVR. Paravalvular leak could have occurred regardless of postoperative period even at 25 years after implantation. Lifelong clinical follow-up is considered necessary.