Predictors of Clinical Outcomes Following Transcatheter Aortic Valve Implantation

Background: Despite good predictive scores for surgical aortic valve replacement (SAVR), they were not designed for transcatheter aortic valve implantation (TAVI). We sought to determine factors predictive of worse outcomes post-TAVI. Methods: Between 2008 and 2017, all patients who underwent TAVI in two experienced centres in Melbourne were prospectively included in a registry and followed-up for one year. Outcomes were based on the Valve Academic Research Consortium (VARC)-2. The primary end-point was one-year mortality. Results: A total of 588 patients (mean age 84 ± 6 years, mean STS-PROM 5.1 ± 3.5, 295 (50%) women) were included. The primary end-point occurred in 43 (7.3%) patients (baseline mean STS-PROM 6.3 ± 3.8 for patients who died vs. 4.9 ± 3.5 for survivors, p = 0.01). On univariate analysis, baseline characteristics that predicted mortality included larger valve size (29-34mm, HR = 2.44, 95%CI 1.13-5.28, p = 0.02), atrial fibrillation (AF, HR = 2.35, 95%CI 1.26-4.36, p = 0.007), moderate-severe mitral regurgitation (MR, HR = 2.34, 95%CI 1.18-4.65, p = 0.02) and chronic obstructive pulmonary disease (COPD, HR = 2.08, 95%CI 1.03-4.20, p = 0.04). Post-procedure ICU admission (HR = 4.81, 95%CI 2.23-10.38, p < 0.001), acute kidney injury (HR = 4.32, 95%CI 1.54-12.11, p = 0.005) and moderate-severe MR (HR = 2.34, 95%CI 1.18-4.65, p = 0.02) predicted worse outcomes. Multivariate analysis confirmed moderate-severe MR (HR = 4.84, 95%CI 2.15-10.9, p>0.001), AF (HR = 2.46, 95%CI 1.19-5.09, p = 0.015), and COPD (HR = 2.37, 95%CI 1.04-5.37, p = 0.039) as predictors of one-year mortality. Conclusion: While STS-PROM is a reasonable surrogate, a novel predictive score specific for TAVI incorporating clinical and echocardiographic factors should be considered for optimal patient selection.