Long term outcomes of reoperative tricuspid valve surgery

Objective:Reoperation on the isolated tricuspid valve is known to be fraught with high perioperative risk. The determinants of this outcome remain unclear. Long-term prognosis also needs to identify risk factors for accurate patient selection. The present work aims to highlight potential risk factors for perioperative and long-term mortality. Methods: this is a retrospective bicentric study of consecutive patients who underwent re-intervention on isolated tricuspid valve after CABG or left-sided valve surgery. The approach and type of surgery were left to the preference of the team. The main parameters were compared, with particular attention to the MELD-score. Results: From March 2004 to July 2020, 115 patients (median age 67 years) underwent isolated tricuspid valve reoperation at two high-volume centers. Previous operations were CABG (6.1%), aortic valve replacement (43%), mitral valve surgery (83%) and tricuspid surgery (23%) in their possible combinations. The surgical redo approach to the tricuspid valve was median sternotomy (82%) or left minithoracotomy. Most patients had atrial fibrillation (77%). Median cardiopulmonary bypass time was 64 min (IQR 51-90 min) and 85 patients were operated on beating heart approach. Most of the patients received a biological prosthesis (75%), followed by ring annuloplasty (22%) and had mechanical valve replacement in 4 cases (3.5%). A significant proportion of patients experienced post-operative acute kidney injury (28%) and 11% required a de-novo implantation of definitive pacemaker. In-hospital mortality was 7.8%. No single preoperative or intraoperative factor emerged as associated with in-hospital mortality at multivariable analysis. Interestingly, MELD score of 15 or above shows good discriminatory ability (AUC 0.68) for in-hospital mortality although less specific. Mean follow-up time is 5 years (range 3 months - 16 years) and 88% complete. Overall survival at 1, 5, and 10 years is 92.6%, 71.1% and 43.9%, respectively. When stratified for preoperative MELD Score with 15 points as cut-off value, a clear survival advantage is present for those within the first stratum (log-rank analysis p = 0.029). Conclusions: Isolated tricuspid valve surgery in case of reintervention is feasible with low in-hospital mortality, although the incidence of postoperative acute renal failure is not negligible. Despite its low specificity, the MELD score is useful in stratifying patients, because those who underwent surgery with a score <15 have a clear long-term survival advantage.