Highlight
– A multicenter, randomized pilot trial (N = 26) showed feasibility of enrolling and maintaining separation between low- and moderate-intensity anticoagulation strategies in adults on venovenous ECMO.
– Major bleeding occurred less frequently in the low-intensity group (8.3% vs 28.6%); thromboembolic events were infrequent and did not clearly differ between groups.
– The study was designed to assess feasibility, not provide definitive efficacy or safety estimates; a larger multicenter trial is justified and appears achievable.
Background and clinical context
Bleeding is one of the most common and clinically consequential complications of venovenous extracorporeal membrane oxygenation (ECMO). Anticoagulation is employed to limit clot formation in the extracorporeal circuit and oxygenator but increases the risk of hemorrhage in critically ill patients who already have multiple factors predisposing to bleeding (coagulopathy, thrombocytopenia, invasive procedures). Balancing anticoagulation intensity to minimize both bleeding and thrombosis is a central, unresolved problem in ECMO management.
Practice varies widely among centers and clinicians: some programs favor systemic heparin with traditional laboratory targets, while others have explored lower-intensity or even anticoagulation-minimizing strategies, particularly when bleeding risk is high or in the context of oxygenator changes and newer circuit coatings. Observational registry data and single-center series have raised equipoise about whether lower-intensity anticoagulation might reduce hemorrhagic complications without a clinically important rise in thrombotic events, but randomized trial evidence has been lacking.
Study design
This multicenter, parallel-group, randomized pilot trial enrolled 26 critically ill adults receiving venovenous ECMO at three United States centers. Participants were randomized to either low-intensity or moderate-intensity systemic anticoagulation for the duration of ECMO. The trial primarily evaluated feasibility outcomes: enrollment rate across sites and adherence to assigned anticoagulation intensity (i.e., separation between arms). Clinical endpoints were assessed between enrollment and 24 hours after decannulation.
Primary clinical outcomes for the pilot were defined as:
– Primary efficacy outcome: major bleeding (as pre-defined by the trial—consistent with common critical care bleeding definitions).
– Primary safety outcome: thromboembolic events (including clinically significant circuit or patient thrombosis).
The trial report (Gannon et al., Chest 2025) provides event counts, absolute risk differences, 95% confidence intervals (CIs), and P values for the predefined outcomes.
Key findings
Enrollment and adherence
– All 26 randomized patients received the assigned anticoagulation intensity, demonstrating excellent protocol adherence and separation between groups. This confirms feasibility of an interventional randomized approach to anticoagulation intensity in venovenous ECMO across multiple centers.
Bleeding outcomes
– Major bleeding occurred in 1 of 12 patients (8.3%) in the low-intensity anticoagulation group versus 4 of 14 patients (28.6%) in the moderate-intensity group.
– Absolute risk difference was -20.2 percentage points (95% CI, -48.6 to 8.1); P = .33.
– Although the absolute point estimate favors low-intensity anticoagulation (a 20-percentage-point lower risk), the confidence interval is wide and crosses the null, reflecting the small sample size and the pilot nature of the study.
Thromboembolic events
– One patient in the low-intensity group (8.3%) experienced a thromboembolic event compared with none in the moderate-intensity group (difference, 8.3 percentage points; 95% CI, -7.3 to 24.0; P = .46).
– Event counts were small; the trial was not powered to exclude a clinically important increase in thrombosis with reduced anticoagulation intensity.
Mortality and other clinical outcomes
– No patients in the low-intensity group died before hospital discharge, whereas 2 patients (14.3%) in the moderate-intensity group died before discharge; both of these deaths followed major bleeding events.
– The relationship between anticoagulation strategy and mortality must be interpreted cautiously given the sample size and the possibility of chance imbalance in patient risk.
Interpretation of effect sizes
– The pilot provides a signal that lower-intensity anticoagulation may reduce bleeding without a clear and large excess of thrombotic events, but uncertainty remains high.
– Confidence intervals are wide; both clinically important benefit and potential harm remain compatible with trial data.
Mechanistic rationale and biological plausibility
ECMO circuits activate coagulation cascades through blood–material interactions, shear stress, and exposure of blood to nonendothelial surfaces. Systemic anticoagulation (most commonly unfractionated heparin) reduces circuit thrombus formation but also impairs hemostasis. Many bleeding events in ECMO arise from mucosal sites, surgical cannulation sites, or spontaneous hemorrhage in patients with underlying coagulopathy. Lowering anticoagulation intensity could plausibly reduce bleeding frequency or severity while modern circuit coatings, improved oxygenator design, and careful circuit surveillance may mitigate thrombosis risk. However, insufficient anticoagulation may increase the need for circuit exchanges, oxygenator failures, or clinically significant thrombotic events, which would offset bleeding reductions.
Expert commentary and comparison to current practice
This pilot trial addresses a high-priority question for ECMO clinicians. Current guidelines and practice documents (including ELSO guidance) acknowledge limited high-quality evidence to specify anticoagulation targets and emphasize individualized decision-making. The randomized pilot confirms that a pragmatic multicenter randomized trial is achievable, which is an essential step toward generating definitive evidence.
Strengths of the study
– Randomized allocation removes selection bias that afflicts observational comparisons.
– Multicenter conduct increases generalizability compared with single-center series.
– High adherence and preserved separation between strategies indicate operational feasibility and clinician acceptability.
Limitations and caveats
– Small sample size: with only 26 patients, the trial is underpowered to detect plausible differences in major bleeding or thrombosis and cannot provide definitive safety conclusions.
– Event definitions and anticoagulation targets (specific laboratory goals, dosing ranges) are essential to interpret applicability; the pilot report outlines strategies but a definitive trial should pre-specify targets and management algorithms to ensure reproducibility.
– Center expertise, circuit components (coating, oxygenator model), transfusion practices, and threshold for circuit exchange vary and could materially affect both bleeding and thrombosis rates, potentially limiting external validity.
– Short surveillance window (to 24 hours after decannulation) captures most in-hospital events tied to ECMO but may miss late sequelae related to thrombotic complications or bleeding-related morbidity beyond that timeframe.
Implications for clinicians
– The pilot data suggest that lower-intensity anticoagulation is feasible and may reduce bleeding, but clinicians should not change practice universally on the basis of this small trial alone.
– Centers facing high bleeding burden may consider protocolized attempts at lower anticoagulation intensity within quality-improvement frameworks, with rigorous monitoring for thrombotic complications and transparent reporting.
Design considerations for a definitive trial
A large, multicenter randomized trial is indicated. Key design elements to consider include:
– Clearly prespecified anticoagulation targets (e.g., anti-Xa, aPTT ranges, ACT) and protocolized algorithm for adjustments and for management of bleeding or suspected thrombosis.
– Sufficient sample size to detect clinically meaningful differences in major bleeding and thrombotic complications with acceptable power; sample-size estimation should consider expected event rates from registries and pilot data.
– Standardized definitions of major bleeding and thrombotic events and blinded adjudication of clinical events.
– Stratification by key confounders (e.g., baseline bleeding risk, indication for ECMO, body mass index, center) to reduce imbalance.
– Secondary outcomes including transfusion requirements, circuit exchange frequency, oxygenator lifespan, duration of ECMO, ICU/hospital LOS, and functional outcomes after discharge.
– Embedded economic evaluation to understand resource implications (e.g., fewer transfusions vs more circuit changes).
Conclusion and takeaways
The Strategies for Anticoagulation During Venovenous Extracorporeal Membrane Oxygenation pilot trial demonstrates that randomized comparison of low- versus moderate-intensity anticoagulation in venovenous ECMO is feasible across multiple U.S. centers, with excellent adherence and separable treatment strategies. The pilot observed fewer major bleeding events with low-intensity anticoagulation and no clear excess of thromboembolic events, but the trial was not powered for efficacy or safety conclusions. A large, pragmatic, multicenter randomized trial with predefined anticoagulation targets, robust event adjudication, and clinically meaningful endpoints is justified and appears achievable based on these feasibility data.
Until results of a definitive trial are available, clinicians should individualize anticoagulation intensity in ECMO patients, balancing bleeding risk and thrombosis risk, and document institutional protocols and outcomes to inform practice and future trial design.
Funding and trial registration
Clinical trial registry: ClinicalTrials.gov No.: NCT04997265. Funding sources are reported in the original publication (Gannon et al., Chest 2025).
References
Gannon WD, Pratt EH, Vogelsong MA, Adkisson WH, Bacchetta M, Bloom SL, Ford DJ, Guenthart BA, Landsperger JS, Qian ET, Rackley CR, Rice TW, Fielding-Singh V, Stokes JW, Stollings JL, Semler MW, Casey JD; Pragmatic Critical Care Research Group. Low-Intensity vs Moderate-Intensity Anticoagulation for Venovenous Extracorporeal Membrane Oxygenation: The Strategies for Anticoagulation During Venovenous Extracorporeal Membrane Oxygenation Pilot Trial. Chest. 2025 Sep;168(3):639-649. doi: 10.1016/j.chest.2025.02.032. Epub 2025 Mar 11. PMID: 40081660; PMCID: PMC12489345.
(Readers are also referred to contemporary ELSO guidance and institutional protocols for anticoagulation during ECMO; comprehensive guideline references and registry data are discussed in the trial manuscript.)