Anticoagulation Strategies in Adults Undergoing Extracorporeal Membrane Oxygenation: A Systematic Review and...

Extracorporeal membrane oxygenation (ECMO) is a life-saving intervention for patients with severe cardiac or respiratory failure, but it is associated with a high risk of thrombotic and bleeding complications. Unfractionated heparin (UFH) remains the most frequently used anticoagulant, largely due to historical practice and longstanding clinical familiarity, despite the absence of robust evidence from randomized controlled trials. Alternative strategies-including direct thrombin inhibitors (DTIs), low-molecular-weight heparins (LMWHs), nafamostat mesylate (NM), and no anticoagulation-are increasingly being explored. However, a comprehensive and contemporary comparison of these approaches has been lacking.

We conducted a comprehensive search of PubMed/MEDLINE, EMBASE, and CENTRAL through March 2025, supplemented by manual searches of reference lists. We included randomized controlled trials and observational studies including patients 16 years or older receiving ECMO for ≥ 24 h that compared any anticoagulation strategy with another or with no anticoagulation. The primary outcome was thromboembolic events; secondary outcomes included bleeding and mortality. Risk of bias was assessed using validated tools. Meta-analyses were performed using a multivariable random-effects model, with prespecified subgroup analyses by ECMO modality (venoarterial [VA], venovenous [VV], or mixed) and subsequent sensitivity analyses.

Twenty-one observational studies involving 2 775 adult ECMO patients were included, with UFH serving as the comparator in all studies. DTIs showed reduced thromboembolism (OR 0.73; 95% CI: 0.53-0.99) and were associated with significantly lower bleeding (OR 0.51; 95% CI: 0.39-0.67) and mortality (OR 0.70; 95% CI: 0.52-0.94), confirmed in a sensitivity analysis. LMWH was associated with significantly reduced odds of thromboembolic events (OR 0.26; 95% CI: 0.13-0.55), as well as showing a favorable bleeding profile (OR 0.46; 95% CI: 0.25-0.87), yet no significant difference in mortality (OR 0.80; 95% CI: 0.30-2.14). While no anticoagulation reduced the odds for bleeding (OR 0.17; 95% CI: 0.07-0.38) and NM showed no significant differences, both had wide confidence intervals, limiting interpretation. Heterogeneity for LMWH was low for efficacy and bleeding, yet substantial for mortality; for DTI, low to moderate; for no anticoagulation, low; and for NM, substantial to high on all comparisons.

In this systematic review and meta-analysis, DTIs and LMWH appear to be effective and safe alternatives to UFH in adult ECMO, with consistent reductions in thromboembolic and bleeding events and a survival benefit observed for DTIs. These findings support reconsideration of UFH as the default anticoagulant and favor a more prominent role for alternative strategies in clinical practice.

PROSPERO CRD42022363588.