We sought to assess the impact of the aortic root geometry on developing de novo aortic insufficiency (AI) in patients undergoing left ventricular assist device (LVAD). In total, 114 patients underwent LVAD implantation between February 2016 and January 2020 were included in this study (HeartMate3 N = 68, HeartWare N = 46). Significant aortic insufficiency was defined as mild-to-moderate or greater in echocardiography. The cohort was divided into two groups; those who developed significant AI (Group AI: n = 13) and did not (Group non-AI: n = 101). The primary outcomes of interest included late survival and predictors for significant AI. The patients in Group AI were older than Group non-AI (62.6 ± 11.9 vs 51.3 ± 14.0 years, p < 0.01). The diameter of proximal ascending aorta in Group AI was larger than Group non-AI (31.0 ± 5.0 vs 27.4 ± 4.3 mm, p < 0.01). Aortic valve remained closed in 53.8% in Group AI and 36.6% in Group non-AI (p = 0.24). The late survival was not significantly different between the groups (67.1% vs 76.0% at 3 years, log rank = 0.97). The Cox hazard model showed that larger proximal ascending aortic diameter/BSA (HR 1.55, CI 1.19-2.04, p < 0.01) and not-opening aortic valve (HR 4.73, CI 1.43-16.9, p = 0.01) were independent risk factors for significant AI. The cutoff value of proximal ascending aortic diameter/BSA was 15.5 (area under curve: 0.770, sensitivity: 0.69, specificity: 0.79). Dilated proximal ascending aorta at the time of LVAD surgery and not-opening aortic valve during follow-up were associated with the incidence of de novo significant AI.
Keywords: Aortic insufficiency; Heart failure; Left ventricular assist device; Proximal ascending aorta size.
© 2021. Springer Japan KK, part of Springer Nature.