Long-term Cardiac Remodeling and Outcomes after LVAD Discontinuation for Myocardial Recovery with Device Left In-Situ

LVADs through mechanical unloading can lead to reverse remodeling and may culminate in myocardial recovery permitting LVAD deactivation. Minimally invasive LVAD discontinuation with device left in-situ has been reported by our groups to be safe with respect to risk of TE or infection post-discontinuation. Whether this approach reduces long-term safety or recovery sustainability is unknown. This is a dual-center retrospective analysis from May 2010 to January 2020 of 17 patients who had elective minimally invasive LVAD discontinuation for myocardial recovery. All patients had outflow graft surgical ligation and driveline discontinuation with device left in-situ. LVAD discontinuations for infection or device thrombosis were excluded. Patients were followed with serial echocardiography or cardiac CT to monitor for cardiac remodeling. Primary composite endpoint was survival free of HF hospitalization, LVAD reimplantation or transplant after discontinuation. Median age of patients at VAD implantation was 42 years (IQR: 25 to 54) years, all had a nonischemic cardiomyopathy (100%), and 5 (29%) were female. At VAD implantation, median EF was 10.0% (5.0 to 15.0) and LVIDD 6.6cm (5.8 to 7.1). With 11 HVAD (65%) and 6 HeartMate II patients (35%), median duration of LVAD support was 28.7 months (13.5 to 36.2). During LVAD weaning study, median LVEF was 55% (50 to 60), LVIDD 4.8cm (4.3 to 5.1), PCWP 8mmHg (8 to 12), CI 2.4 L/m2 (2.0 to 2.7) and peak VO2 21.2 mL/kg/min (15.7 to 23.8). Median follow-up time after discontinuation was 28.0 months (13.9 to 49.1). One month post-discontinuation, median EF decreased to 50% (49 to 55, p=0.04) and LVIDD increased to 5.2 cm (4.6 to 5.8, p=0.10). By 4 years, median EF was 40% (19 to 50, p=0.014) and LVIDD 6.0cm (5.5 to 6.3, p=0.001). Post-discontinuation, 1 (6%) patient required transplantation and 3 (18%) died, all from infection related causes. Probability of survival free of primary composite endpoint at 1, 2 and 4 years after discontinuation was 94, 87 and 62%, respectively (Figure). In this analysis, we found LVAD discontinuation with device left in-situ for myocardial recovery is associated with long-term cardiac remodeling. Importantly, probability of survival free of primary composite endpoint was 87% at 2 and 62% at 4 years post-discontinuation. These data show LVAD discontinuation without device explant for myocardial remission appears to be safe with low risk of recurrent HF, though infection may be an important cause of mortality. Factors leading to long-term remodeling must be elucidated.