Long-term clinical outcomes of tricuspid valve replacement using bovine versus porcine valves: a nationwide population-based study

Abstract

Abstract OBJECTIVES The early and long-term clinical outcomes of bovine versus porcine tricuspid valve replacement (TVR) were compared based on the nationwide database from the National Health Insurance Service. METHODS Of 1464 patients who underwent TVR from 2002 to 2018 in Korea, 541 were enrolled after excluding mechanical TVR, re-TVR, complex congenital heart disease, Ebstein anomaly and age <19 years at operation. Bovine (group B) and porcine valves (group P) were used in 342 and 199 patients, respectively. The median follow-up duration was 4.1 years [interquartile range 1.2–9.0]. Inverse probability of treatment weighting analysis was performed for adjustment between the groups. Early and long-term clinical outcomes, including all-cause mortality, ischaemic stroke, haemorrhagic stroke, endocarditis and reoperation, were compared. RESULTS In inverse probability of treatment weighting analysis, operative mortality and early clinical outcomes were comparable between the groups. The cumulative incidence of all-cause mortality demonstrated no significant differences between the groups [36.8% vs 38.0% at 5 years in group B versus group P; adjusted hazard ratio = 0.93; P = 0.617]. The cumulative incidence of cardiac death, ischaemic stroke, haemorrhagic stroke and endocarditis also demonstrated no significant differences between the groups (28.1% vs 25.9%, 7.1% vs 1.2%, 3.2% vs 4.2% and 9.7% vs 6.0% at 5 years in group B versus group P, respectively). However, the cumulative incidence of reoperation was higher in group B than in group P (20.2% vs 3.4% at 5 years in group B vs group P; adjusted hazard ratio = 4.76; P = 0.006). CONCLUSIONS Early clinical outcomes and long-term outcomes, including all-cause mortality, cardiac death, ischaemic stroke, haemorrhagic stroke and endocarditis, were comparable between bovine and porcine TVRs. However, porcine valves demonstrated a lower cumulative incidence of reoperation than bovine valves.