Comparison of Periprocedural and Intermediate-Term Outcomes of TAVI in Patients with Ejection Fraction ≤ 20% vs. Patients with 20% < EF ≤ 40%

Abstract

Treatment of congestive heart failure (CHF) with left ventricular (LV) systolic dysfunction and severe aortic stenosis (AS) is challenging, yet transcatheter aortic valve replacement (TAVR) has emerged as a suitable treatment option in such patients. We compared the periprocedural outcomes of TAVR in patients with an ejection fraction (EF) of ≤20% (VLEF group) to patients with an EF > 20% to ≤40% (LEF group). We included patients with severe AS and reduced LV ejection fraction (LVEF ≤ 40%) who underwent TAVR at four centers within Northwell Health between January 2016 and December 2020. Over 2000 consecutive patients were analyzed, of which 355 patients met the inclusion criteria. The primary composite endpoint was in-hospital mortality, moderate or greater paravalvular (PVL), stroke, conversion to open surgery, aortic valve re-intervention, and/or need for PPM. Secondary endpoints were length of stay, NYHA classification at 1 month and 1 year, mortality at 1 month and 1 year, mean valve gradient at 1 month, KCCQ score at 1 month, and ≥ moderate PVL at 1 month. There was no difference in the primary composite endpoint between the two groups (23.6% for VLEF vs. 25.3% for LEF, p = 0.29). During TAVR placement, 40% of patients in the VLEF group required ≥1 vasopressors for hypotension lasting ≥30 min vs. only 21% of patients in the LEF group (p < 0.01). Intra-aortic balloon pump (IABP) use during procedure was greater in the VLEF group (9% vs. 1%, p < 0.01)—all placed post TAVR. Emergency ECMO use was higher in the VLEF group as well (5% vs. 0%). Total length of stay was significantly different between the two groups as well (6 days vs. 3 days, p < 0.01). Both groups had a change in LVEF of ~10%. One-year outcomes were similar between the groups. All-cause mortality at 1 year was not significantly different at 1 year (13% for VLEF vs. 11% for LEF), and KCC scores were also similar (77.54 vs. 74.97). Mean aortic valve gradients were also similar (12 mmHg vs. 11 mmHg, p = 0.48). Our study suggests that patients with EF ≤ 20% can safely have TAVR with similar periprocedural outcomes compared to patients with EF > 20% to ≤40% despite higher rates of vasopressor and mechanical support.