Redo mitral valve replacement in children

Background. Despite excellent results with current techniques of mitral valve repair, some children still require mitral valve replacement (MVR). Of necessity, a certain percentage of these children need valve rereplacement (redo MVR). Methods. Of 82 MVRs at our institution since 1988, 22 were redo MVRs in 21 children aged 1.4-21 years (mean 9.8 +/- 5.6 years). Interval from initial MVR was 1 month to 18 years (mean 5.8 +/- 4.9 years). Reoperation indication was primarily stenosis in 13 (usually pannus formation), regurgitation in 4, valve thrombosis or embolism in 4, and endocarditis in 1. The initial valve was mechanical in 18 and xenograft in 4. Results. During redo MVR, 8 of 22 (36%) patients had additional procedures including left ventricular outflow tract obstruction relief or aortic valve replacement in 4 and tricuspid valve repair in 2. The new mitral valve was mechanical in 20 and xenograft in 2. Median valve size increased from 21 rum to 25 mm although 7 valves were not upsized on rereplacement (average valve size increase was 3.3 +/- 1 mm). Only 2 patients (9%) needed a new pacemaker. There were no hospital deaths. Six children developed end-stage heart failure and underwent successful cardiac transplantation 3 to 32 (mean 12.1 +/- 10.0) months, postoperatively. On follow-up of 1 month to 7.7 years (mean 2.5 +/- 2.3 years) there has been only one late death of an unrecognized coronary artery anomaly. Conclusions. Redo MVR in children can be performed safely with low morbidity and mortality. A larger prosthesis can often be placed in these children. A surprisingly high percentage of patients eventually require transplantation, perhaps reflecting unnecessary delay in referral for redo MVR. (c) 2005 by The Society of Thoracic Surgeons.