Results and factors associated with adverse outcome after tricuspid valve replacement

Background and aim: This retrospective analysis of patients with severe tricuspid valve disease, who underwent tricuspid valve replacement (TVR) for either tricuspid regurgitation or stenosis, has been designed to determine the factors that predict poor hospital and long-term survival. Methods: The study population comprised 86 patients, 65 women and 21 men, who underwent TVR with or without concomitant surgical procedures between 2000 and 2010 at our institution. Patients with Ebstein's or other complex congenital anomalies were excluded from the study. Results: Average age at operation was 58.5 +/- 12.5 (range 16-78) years. Fifty (58.1%) patients had undergone previous cardiac surgery. Forty-two patients were in New York Heart Association (NYHA) class III functional capacity, and 18 were in class IV. Symptoms of right heart failure (HF) were present in 66 patients, of whom 19 had ascites. Bioprosthetic tricuspid valves were implanted in 84 patients and mechanical prostheses in two. The choice to proceed to TVR instead of repair was individualised and based on the surgeon's preference. In-hospital mortality was 18 (20.9%) patients, caused mainly by multi-organ and HF, and was significantly related to NYHA class and symptoms of right HF before surgery, with no mortality in patients with NYHA class I and II, 19% mortality with NYHA class III, and 55.6% mortality with NYHA class IV. Eighteen (20.9%) patients died during postoperative follow-up. The main risk factors associated with perioperative mortality were: the presence of severe symptoms at the time of surgery, low preoperative haematocrit, postoperative complications, postoperative ventilation time longer than 72 h, and renal failure requiring dialysis. Elevated pulmonary artery pressure, preoperative symptoms of right HF, and low haematocrit unfavourably affected the long-term results. Conclusions: Many earlier studies reported high mortality and morbidity after TVR in both early and late postoperative periods. Our main finding is that good outcomes for TVR are achievable in properly selected patients. Sixty of 86 patients in our group had preoperative NYHA functional class III and IV, which suggests that surgical timing was late in many patients. Based on our observations, we propose that surgical correction of severe tricuspid valve disease should be considered before the development of advanced HF, when patients are asymptomatic or only oligosymptomatic.