Development of De-Novo Aortic Insufficiency in Patients with Durable Left Ventricular Assist Device Who Were Bridged from Impella Device

Abstract

<h3>Purpose</h3> As the incidence of INTERMACS I profile patients receiving durable left ventricular assist devices(VADs) increases, there has been a rising use of Impella (Abiomed, Daners, MA) for acute support. While impella allows for end-organ recovery and safe bridging to VAD it has also been associated with the development of de-novo aortic insufficiency (AI) post VAD in small series. We evaluate the incidence and clinical significance of de-novo AI in post VAD patients at our center. <h3>Methods</h3> All patients implanted with durable VAD at our center from 2011-2021 were retrospectively reviewed. Echocardiogram were reviewed prior to VAD implant and 3 months post VAD for the development of de-novo AI. <h3>Results</h3> Forty-eight patients were implanted with Heartmate II or Heartmate 3 (Abbott, Chicago, IL) durable VAD. Twenty-seven patients were bridged with an Impella prior to VAD, while 21 were not. By definition, the Impella-bridged patients were INTERMACS 1 at time of VAD; 24 were male and the mean age was 65.3-years with 62.9% (N=17) with ischemic cardiomyopathy. In this group, 25.9% (N=7) of patients had at least moderate de-novo AI and 48% (N=13) had at least mild de-novo AI at 3 months post VAD. Three of the Impella bridged patients received aortic valve plication during VAD surgery due to significant AI, and at 3 months post VAD, all 3 patients had trace to no AI. Of the 21 patients not bridged with Impella, 17 were INTERMACS 2 and 4 were INTERMACS 3; 18 were male and the mean age was 62.7 years-old, with 61.9%(N=13) having ischemic cardiomyopathy. In this group, 19% (N=4) had at least moderate de-novo AI and 23.8% (N=5) had at least mild de-novo AI a 3 months post LVAD echocardiogram. One patient received prophylactic aortic valve plication during LVAD implantation due to significant baseline AI. No patients required aortic valve intervention post LVAD in the impella-bridged group, while 1 patient required TAVR for AI in the non-impella group. <h3>Conclusion</h3> This retrospective analysis of our single center experience shows that while there is an increased incidence of de-novo AI in VAD patients that are bridged with Impella, the presence of AI may not be clinically significant. By proper afterload reduction and diuresis as well as consideration of aortic plication in patients with at least moderate AI at time of VAD, morbidity of AI can be significantly minimized.