Background: Venoarterial extracorporeal membrane oxygenation (VA-ECMO) is increasingly used in patients with cardiogenic shock despite the lack of evidence from adequately powered randomised clinical trials. Three trials reported so far were underpowered to detect a survival benefit; we therefore conducted an individual patient-based meta-analysis to assess the effect of VA-ECMO on 30-day death rate.
Methods: Randomised clinical trials comparing early routine use of VA-ECMO versus optimal medical therapy alone in patients presenting with infarct-related cardiogenic shock were identified by searching MEDLINE, Cochrane Central Register of Controlled Trials, Embase, and trial registries until June 12, 2023. Trials were included if at least all-cause death rate 30 days after in-hospital randomisation was reported and trial investigators agreed to collaborate (ie, providing individual patient data). Odds ratios (ORs) as primary outcome measure were pooled using logistic regression models. This study is registered with PROSPERO (CRD42023431258).
Findings: Four trials (n=567 patients; 284 VA-ECMO, 283 control) were identified and included. Overall, there was no significant reduction of 30-day death rate with the early use of VA-ECMO (OR 0·93; 95% CI 0·66-1·29). Complication rates were higher with VA-ECMO for major bleeding (OR 2·44; 95% CI 1·55-3·84) and peripheral ischaemic vascular complications (OR 3·53; 95% CI 1·70-7·34). Prespecified subgroup analyses were consistent and did not show any benefit for VA-ECMO (pinteraction ≥0·079).
Interpretation: VA-ECMO did not reduce 30-day death rate compared with medical therapy alone in patients with infarct-related cardiogenic shock, and an increase in major bleeding and vascular complications was observed. A careful review of the indication for VA-ECMO in this setting is warranted.
Funding: Foundation Institut für Herzinfarktforschung.
Copyright © 2023 Elsevier Ltd. All rights reserved.