Abstract 4362985: Waitlist Outcomes in Patients on VA-ECMO With and Without Left Ventricular Unloading as a Bridge to Heart Transplantation

Abstract

Background: Venoarterial extracorporeal membrane oxygenation (VA-ECMO) is increasingly used to bridge patients to heart transplantation following the 2018 UNOS heart allocation policy that prioritized patients on VA-ECMO support due to their high waitlist mortality. Many VA-ECMO patients employ concomitant left ventricular (LV) unloading strategies, such as intra-aortic balloon pump (IABP) or Impella (ECPella); however, the clinical utility of these methods in improving waitlist outcomes remains unclear. Research Question: To compare clinical outcomes of waitlisted patients between those who received an LV unloading strategy and those who did not while being supported by VA-ECMO as a bridge to heart transplantation. Methods: Using the UNOS registry, we retrospectively analyzed adult first-time heart transplant candidates supported by VA-ECMO alone (Loading) versus VA-ECMO plus LV unloading (Unloading) from October 18, 2018, to March 30, 2022. Competing risks analysis and multivariable Cox proportional hazards models were used to evaluate outcomes of death/deterioration and transplantation on the waitlist while adjusting for age, gender, ethnicity, and mechanical ventilation. Results: In total, 369 patients were included: 252 in the Loading group, and 117 in Unloading. Baseline differences indicated higher illness severity in the Unloading group (more mechanical ventilation [35.9% vs. 22.2%, p=0.008] and inotrope use [64.1% vs. 48.0%, p=0.006]). Despite this, there was no statistically significant difference in waitlist death/deterioration between groups (adjusted subdistribution hazard ratio [sdHR]: 1.304; 95% CI: 0.709–2.398; p=0.39). However, the adjusted cause-specific hazard ratio (csHR) was 1.618 (95% CI: 0.860–3.043; p=0.14), suggests a nonsignificant trend toward higher instantaneous risk of death/deterioration in the ECMO unloading alone group. Subgroup analysis comparing ECMO only, ECMO+IABP, and ECPella also showed no significant differences. Conclusion: LV unloading in patients supported by VA-ECMO did not significantly alter waitlist outcomes in this cohort despite being used in sicker patients. Nonetheless, we did identify a trend towards increased risk of death/deterioration on the waitlist among the Unloading group. Larger studies are warranted to determine if an unloading device should be considered in waitlist prioritization for future organ allocation policy iterations.