Recovery of left ventricular function is associated with improved outcomes in LVAD recipients

Background The significance of recovered left ventricular ejection fraction (LVEF) in LVAD recipients, outside of pump explantation, is unclear. Methods Patients undergoing first LVAD implantation at Duke University Hospital between 2006 and 2017 were evaluated for LVEF recovery up to 2 years following implant. Occurrence of gastrointestinal bleeding (GIB), hospitalization for heart failure (HF), pump thrombosis and death were assessed before and after LVEF recovery. Results Of 286 patients who met inclusion criteria, 9.8% reached a “threshold” of recovery with an LVEF ≥ 40%. 17.4% achieved “relative” recovery with an increase in LVEF ≥ 10% since LVAD implantation. For either definition, recovered patients had a lower incidence of a composite endpoint of GIB, HF hospitalization, pump thrombosis, or death compared to patients without recovery. Patients with “threshold” recovery had 4.7 events per 100 patient-years (95% CI, 0.7-33.6) compared to 48.8 events per 100 patient-years (95% CI, 39.5-60.3) without “threshold” recovery [p = .020]. Those with “relative” recovery had 14.1 events per 100 patient-years [95% CI, 5.9-33.8] versus 50.7 events per 100 patient-years (95% CI, 40.7-63.0) without “relative” recovery [p = 0.005]. However, improved outcomes in the “relative” recovery group were limited to those who also met the “threshold” definition. Importantly, among patients who achieved “threshold” recovery, the incidence of the composite endpoint declines in the postrecovery period, suggesting that LVEF recovery mechanistically results in improved outcomes. Conclusions An LVEF ≥ 40% associates with better outcomes in LVAD recipients. Methods to promote recovery could reduce morbidity and mortality related to LVAD support. The significance of recovered left ventricular ejection fraction (LVEF) in LVAD recipients, outside of pump explantation, is unclear. Patients undergoing first LVAD implantation at Duke University Hospital between 2006 and 2017 were evaluated for LVEF recovery up to 2 years following implant. Occurrence of gastrointestinal bleeding (GIB), hospitalization for heart failure (HF), pump thrombosis and death were assessed before and after LVEF recovery. Of 286 patients who met inclusion criteria, 9.8% reached a “threshold” of recovery with an LVEF ≥ 40%. 17.4% achieved “relative” recovery with an increase in LVEF ≥ 10% since LVAD implantation. For either definition, recovered patients had a lower incidence of a composite endpoint of GIB, HF hospitalization, pump thrombosis, or death compared to patients without recovery. Patients with “threshold” recovery had 4.7 events per 100 patient-years (95% CI, 0.7-33.6) compared to 48.8 events per 100 patient-years (95% CI, 39.5-60.3) without “threshold” recovery [p = .020]. Those with “relative” recovery had 14.1 events per 100 patient-years [95% CI, 5.9-33.8] versus 50.7 events per 100 patient-years (95% CI, 40.7-63.0) without “relative” recovery [p = 0.005]. However, improved outcomes in the “relative” recovery group were limited to those who also met the “threshold” definition. Importantly, among patients who achieved “threshold” recovery, the incidence of the composite endpoint declines in the postrecovery period, suggesting that LVEF recovery mechanistically results in improved outcomes. An LVEF ≥ 40% associates with better outcomes in LVAD recipients. Methods to promote recovery could reduce morbidity and mortality related to LVAD support.