Long-term outcomes of tricuspid valve replacement after previous left-side heart surgery.

OBJECTIVES To assess long-term outcomes of tricuspid valve replacement (TVR) after previous left-side heart surgery. METHODS We reviewed reoperative TVR after left-side heart surgery performed at our institution between March 1997 and June 2012. In-hospital data were retrieved from our institutional database or medical records; follow-up was performed through telephone call, surviving patients being asked to provide a recent (≤6 months) echocardiogram. RESULTS Reoperative TVR was performed in 117 patients. Preoperative characteristics included: mean age 63.7 years, median logistic EuroSCORE (LES) 11.8, New York Heart Association (NYHA) class >2 in 79.5% of patients, right ventricle (RV) dysfunction >mild in 23.9% of patients and mean systolic pulmonary artery pressure (sPAP) 48.4 mmHg. A mechanical prosthesis was implanted in 5.1% of patients. A right thoracotomy was preferred to median sternotomy in 8.6% of cases. Isolated-TVR (I-TVR) was performed in 52.1% of patients, a beating-heart approach being used in 85.2% of I-TVR cases. Postoperative RV failure occurred in 46.1% of patients. Median length-of-stay was 11.5 days. Thirty-day mortality was 6.0% overall and 8.2% in the I-TVR group. Higher preoperative LES (P = 0.002), ascites (P = 0.004), RV dysfunction (P = 0.033) and sPAP (P = 0.046) were associated with acute mortality. No significant difference in acute outcomes was observed between beating and arrested-heart I-TVR, except for postoperative median length-of-stay (9 vs 28 days, respectively, P = 0.007). Among survivors median follow-up time was 5.1 years. Five-year and 10-year freedom from cardiac death were 79.4 and 61.0%, freedom from tricuspid reoperation were 97.3 and 87.5%, freedom from bioprosthesis degeneration were 92.8 and 74.3%, respectively. Five-year and 10-year survival in the I-TVR subgroup were respectively 74.4 and 61.6%. Higher preoperative sPAP was associated with increased follow-up mortality (P = 0.048). At the last follow-up, NYHA class I-II was found in 86.1% of surviving patients. CONCLUSIONS In selected cases, TVR is currently feasible with low acute mortality, especially if performed in the absence of ascites, significant RV dysfunction and pulmonary hypertension. Long-term mortality remains more difficult to predict, although it appeared to be also associated with higher preoperative pulmonary pressure. The global high-complexity profile of these patients is likely to impair long-term outcomes.