Abstract
BACKGROUND: In some patients with left ventricular assist devices (LVADs), unloading of the left ventricle (LV) and medical therapy may lead to improvement in LV systolic function, allowing for LVAD weaning. There are no guideline-directed parameters to help identify candidates for weaning and long-term outcomes remain imperfectly documented. In this study we aimed to assess the clinical and echocardiographic characteristics of weaned patients and evaluate their event-free survival after weaning. METHODS: This investigation was a single-center retrospective study of patients who underwent a second- or third-generation LVAD implantation between 2009 and 2021. RESULTS: Ninety-eight patients were included. Fourteen patients (14%) with LV recovery underwent LVAD weaning after a median support time of 309 days. Heart failure etiologies in weaned patients included toxic (recreational drugs) (n = 8, 57%), toxic (medication) (n = 2, 14%), ischemic (n = 2, 14%), or idiopathic dilated (n = 2, 14%) cardiomyopathy. In unweaned patients, heart failure was mostly attributed to ischemic (n = 35, 42%) and idiopathic dilated (n = 27, 32%) cardiomyopathy. Three months after implantation, patients who were eventually weaned had a higher LV ejection fraction (LVEF) (35% vs 19%, P = 0.001) and lower left ventricular end-diastolic diameter (LVEDD) (52 vs 60 mm, P = 0.03) than unweaned patients. At last follow-up after weaning, mean LVEF was 44 ± 6% and no death nor heart transplant had occurred. CONCLUSIONS: LVADs can induce LV reverse remodeling leading to myocardial recovery in a significant proportion of patients, especially those with toxic and nonischemic cardiomyopathies. Early reverse remodeling with decreasing LVEDD and improving LVEF at 3 months after implantation may suggest potential candidacy for LVAD weaning. Weaned patients maintain satisfactory LVEF recovery after weaning and have good long-term event-free survival.