New OPTN/UNOS data demonstrates higher than previously reported waitlist mortality for lung transplant candidates supported with ECMO

Abstract

Use of extracorporeal membrane oxygenation (ECMO) is not currently incorporated into US allocation models due to the historical lack of complete data in the national US registry which changed in 2016 to include ECMO at the time of waitlist removal and more granular timing and configuration data. We studied adult lung transplant candidates from 01 May 2016 to 01 June 2020 with data abstracted from multiple sources in the US Scientific Registry of Transplant Recipients. Waitlist analyses included cumulative incidence functions and Cox proportional hazards (PH) models considering ECMO as a time-dependent variable. Post-transplant analyses included Kaplan Meier, Cox PH models, and observed to expected survival ratios. A total of 867 candidates were on ECMO prior to transplant; 247 were identified using new sources of data. Candidates on ECMO had a 23.9 increased adjusted likelihood of waitlist removal for being too sick or death, but only a 4.08 increased adjusted likelihood of transplant. Candidates bridged with ECMO who underwent lung transplant (N=587) experienced an increased overall hazard of post-transplant mortality with veno-arterial and veno-venous configurations conferring HR = 1.67 (95% CI, 1.16, 2.40), HR = 1.45 (95% CI, 1.15, 1.82), respectively. We identified an additional 28.5% of candidates bridged with ECMO prior to transplant using new data. This study of the newly identified full cohort of ECMO candidates demonstrates higher utilization of ECMO as well as an underestimation of waitlist mortality risk factors that should inform strategies to provide timely access to transplant for this population.