Midterm outcomes of transaortic and transapical TAVI in patients with unsuitable vascular anatomy for femoral access: A propensity score inverse probability weight study

Aims: Transaortic (TAo) and transapical (TA) implantation can be used in transcatheter aortic valve implantation (TAVI) when the transfemoral (TF) approach is precluded. We compare the safety and efficacy of these alternative techniques. Methods: From October 2007 to February 2016, TAo and TA patients’ data were collected. Propensity score inverse probability of treatment weight (IPTW) method was employed to minimize the impact of no-randomization bias. Results: From our single-center non-TF-TAVI registry, 282 patients were included: 235 (83.3%) underwent TAo and 47 (16.7%) TA. Differences in baseline characteristics were statistically significant in age, sex, risk profile according to logistic-EuroSCORE, and previous cardiac surgery. No difference in hospital morbidity and mortality, but lower stroke-rate in TAo (1.27% vs. 8.5% p <.01) was observed. This was confirmed at logistic regression after IPTW adjustment (odds ratio [OR]: 0.16, 95% CI 0.03–0.71, p =.01), together with reduced risk of the paravalvular leak (PVL) (OR: 0.14, 95% confidence interval [CI]: 0.02–0.81, p =.02). Kaplan–Meier estimates did not demonstrate differences in long-term mortality among access routes (logrank test p =.13). At the IPTW-Cox regression model, long-term mortality was related to New York Heart Association III–IV (hazard ratio [HR]: 2.92, 95% CI: 1.15–7.40, p =.026), chronic renal failure (HR: 3.25; 95% CI: 1.02–10.32 p =.046), previous transient ischemic attack/stroke (HR: 2.29, 95% CI: 1.25–4.20 p =.007). Sapien-3 device resulted to be a protective factor, reducing long-term mortality (HR: 0.18, 95% CI 0.04– 0.90 p =.03). Conclusions: TAo is safe and feasible in case of contraindication to femoral approach demonstrating comparable midterm outcomes to TA, thus representing a central access alternative, to increase the overall safety of high-risk TAVI procedures.