Left ventricular unloading during venoarterial extracorporeal membrane oxygenation implantation: benefits or complications?

Abstract

Abstract Funding Acknowledgements None. Introduction Peripheral venoarterial extracorporeal membrane oxygenation (VA ECMO) is a therapeutic option for refractory cardiogenic shock (CS). However, the retrograde flow increases afterload, leading to pulmonary edema, blood stasis in the left ventricle (LV) and formation of thrombi. Several strategies try to reduce LV pressure, including the percutaneous transvalvular microaxial pump (p-MCS) and intra-aortic balloon pump (IABP). However, these interventions are associated to complications, such as limb ischemia, embolism, bleeding and hemolysis. Purpose This study aims to investigate whether simultaneous LV unloading at ECMO implant yields clinical benefits or complications compared to ECMO implantat alone. Methods Consecutive patients (pts) with CS requiring peripheral VA ECMO from January 2020 to August 2023 were prospectively included, comparing the group that received LV unloading through IABP or p-MCS to the group that received ECMO only. Standard statistical methods were used. Results A total of 38 pts required VA ECMO. 18 of them received LV unloading since cannulation (8 with IABP and 10 with p-MCS). In the group without initial LV-unloading (20), 3 pts required IABP and 1 p-MCS during support and 1 was changed to VAV ECMO. Change to axillar configuration was perfomed in 1 pts in the unloaded group and in 2 in the only-ECMO group. A higher prevalence of SCAI E (72.2%) was observed in the unloaded-ECMO, compared to the only-ECMO, where SCAI D was more prevalent (60%) (p=0.024). In the LV unloaded group, acute myocardial infarction (AMI) was the leading cause of CS, while myocarditis was the predominant one in the other group (p=0.01). The simultaneous support had a higher rate of intubation (94.4% vs 50% p=0.004) and more acute pulmonary edema (44.4% vs 10% p=0.027). In the ECMO-only 2 pts required intubation during support and LV stasis was observed in 1, whereas no cases were reported in the LV-unloaded group (p=0.036). The mean PaO2/FiO2 during support were higher in the not unloaded group (284.54 vs 200.50 mmHg p=0.036), necessitating lower FiO2 levels (40.89 vs 61.47% p=0.003). In terms of complications, we included the 4 patients from the ECMO-only group who eventually required LV unloading in the unloaded-ECMO. There were no significant differences in the duration of support, in-hospital mortality, stroke, embolism, systemic and local bleeding, or hemolysis. Conclusions Pts with ECMO and LV unloading since cannulation debuted with more severe CS, more AMI as the underlying cause, higher rates of acute pulmonary edema, and intubation. Despite these challenges, they did not experience worse outcomes in terms of in-hospital mortality or ischemic and hemorrhagic events. Notably, 4 patients initially managed with ECMO eventually required subsequent LV unloading. Consequently, the simultaneous implantation of ECMO and a LV unloading device appears to be a reasonable alternative for select patients with a higher risk profile.