Predictors of Long-term Cardiovascular Versus Non-cardiovascular Mortality and Repeat Intervention in Patients Having Transcatheter Aortic Valve Implantation

Abstract

There is a paucity of literature characterizing the risk of long-term mortality and reintervention after transcatheter aortic valve implantation (TAVI). Addressing this gap has become increasingly relevant with the inclusion of intermediate and low surgical risk patients and the need for data to inform their long-term management. We sought to investigate the long-term trends and predictors of cardiovascular versus noncardiovascular mortality as well as reintervention in post-TAVI patients. Our cohort consisted of 5,406 patients who underwent TAVI in Ontario, Canada from 2011 to 2018. We used Kaplan-Meier analysis to estimate 7-year all-cause mortality and a Cox proportional hazard model to identify demographic, co-morbid, and procedural predictors. Similarly, cumulative incidence functions were used to estimate cardiovascular versus noncardiovascular mortality at 5 years, with predictors identified through Fine-Gray models. The Kaplan-Meier estimate for 7-year all-cause mortality in our cohort was 67%; this was driven by a number of co-morbidities including congestive heart failure and liver disease. We found that cardiovascular death was more likely for approximately the first 2 years post-TAVI whereas noncardiovascular death was more likely from this point to the end of the study. We identified a number of factors that uniquely modified the risk of either cardiovascular or noncardiovascular mortality. Only 1.6% of patients who underwent repeat intervention. The distinct factors associated with cardiovascular versus noncardiovascular death suggest different approaches to short-term and long-term surveillance of patients post-TAVI by both the heart team and primary care providers.