Aims: Few studies have reported data on the optimal timing of left ventricular (LV) unloading during venoarterial extracorporeal membrane oxygenation (VA-ECMO) for cardiac arrest or shock. This study evaluated the feasibility of an early LV unloading strategy compared with a conventional strategy in VA-ECMO.
Methods and results: Between December 2018 and August 2022, 60 patients at two institutions were randomized in a 1:1 ratio to receive early (n = 30) or conventional (n = 30) LV unloading strategies. The early LV unloading strategy was defined as LV unloading performed at the time of VA-ECMO insertion. LV unloading was performed using a percutaneous transseptal left atrial cannulation via the femoral vein incorporated into the ECMO venous circuit. The early and conventional LV unloading groups included 29 (96.7%) and 23 (76.7%) patients, respectively (median time from VA-ECMO insertion to LV unloading: 48.4 h, interquartile range 47.8-96.5 h). The groups showed no significant differences in the rate of VA-ECMO weaning as the primary endpoint (70.0% vs. 76.7%; relative risk 0.91; 95% confidence interval 0.67-1.24; p = 0.386) and survival to discharge (53.3% vs. 50.0%, p = 0.796). However, the pulmonary congestion score index at 48 h after LV unloading was significantly improved only in the early LV unloading group (2.0 ± 0.7 vs. 1.7 ± 0.6 at baseline vs. at 48 h; p = 0.008).
Conclusions: Compared with the conventional approach, early LV unloading did not improve the VA-ECMO weaning rate, despite the rapid improvement in pulmonary congestion. Therefore, the results of this study do not support the application of this strategy after VA-ECMO insertion.
Keywords: Cardiogenic shock; Catheters; Extracorporeal membrane oxygenation; Heart transplantation; Pulmonary oedema.
© 2023 European Society of Cardiology.