Comment

Rethinking anticoagulation intensity during extracorporeal membrane oxygenation

Mauro Panigada, Giacomo Grasselli

Lancet Available online 7 July 2026

https://doi.org/10.1016/S0140-6736(26)00872-X

Extracorporeal membrane oxygenation (ECMO) is an established therapy for refractory respiratory or circulatory failure,¹ yet bleeding and thrombotic complications continue to drive much of the residual morbidity and mortality.² Clinicians have long targeted an activated partial thromboplastin time (aPTT) of 2·0–2·5 times baseline during ECMO, a range borrowed from non-ECMO populations and written into guidelines without prospective validation in adequately powered trials.³ Observational evidence has progressively challenged this paradigm. Large multicentre prospective cohorts have consistently linked higher aPTT values to haemorrhagic complications, whereas lower intensities have not translated into excess thrombosis.⁴ A recent meta-analysis pooling seven studies and 619 patients on ECMO extended this observation, showing that low-intensity unfractionated heparin (UFH) was associated with lower mortality and intracranial haemorrhage, without an increase in systemic thrombosis.⁵ A definitive answer, however, has been missing, and few topics in critical care have generated as much debate.

In The Lancet, Olivier van Minnen and colleagues⁶ report a multicentre, three-arm, open-label, non-inferiority trial conducted in seven intensive care units in the Netherlands between November, 2020, and September, 2024.¹ Adults receiving veno-venous or veno-arterial ECMO without a compelling indication for full-dose anticoagulation were randomly assigned (1:1:1) to standard-dose UFH targeting an aPTT of 2·0–2·5 times baseline, low-dose UFH targeting 1·5–2·0 times baseline, or therapeutic subcutaneous low-molecular-weight heparin (LMWH). The primary outcome was a composite of severe bleeding, severe thrombotic events during ECMO, and all-cause mortality at 6 months, with a prespecified non-inferiority margin of 7·5 percentage points. Of 330 enrolled patients (110 in each group), 320 were analysed at 6 months (225 (70%) men and 95 (30%) women; median age across groups 55–58 years). Veno-arterial cannulation predominated (192 [60%] participants), and severe acute respiratory distress syndrome and cardiogenic shock were the main indications. Baseline characteristics were broadly balanced; ethnicity was not reported.

The composite outcome occurred in 87 (81%) of 107 patients with standard-dose UFH, 78 (72%) of 108 with low-dose UFH (absolute risk difference −9·1 percentage points, 95% CI −20·3 to 2·1), and 79 (75%) of 105 with LMWH (−6·1 percentage points, −17·2 to 5·0), meeting non-inferiority for both interventions. Severe bleeding occurred in 70 (65%) patients in the standard-dose UFH group, 63 (58%) patients in the low-dose UFH group, and 62 (59%) patients in the LMWH group, with no excess of thrombotic events (12 [11%], 11 [10%], and nine [9%]). 54 (50%) patients in the standard-dose UFH group, 45 (42%) in the low-dose UFH group, and 46 (44%) in the LMWH group had died at 6 months. LMWH had the lowest rate of oxygenator exchanges. The economic evaluation favoured both interventions across plausible willingness-to-pay thresholds. Of note, in the veno-venous ECMO subgroup the relative risk for 6-month mortality with low-dose UFH versus standard UFH was 0·49 (95% CI 0·26–0·95).

This is the first adequately powered randomised trial of anticoagulation targets during ECMO and the first to formally evaluate therapeutic LMWH in this setting, in both patients on veno-venous ECMO and those on veno-arterial ECMO, with complete 6-month follow-up. The pragmatic design adds to external validity, although some caveats deserve mention. The standard-dose comparator poorly mirrors current practice in many centres: the international PROTECMO study reported a median aPTT of 52 s across 41 centres,⁴ a value that falls within the low-dose rather than the standard range of the present trial. aPTT itself was originally designed to detect haemophilia and is affected by acute-phase factor VIII, lupus anticoagulant, and inter-reagent variability, all of which are common during ECMO.7, 8 Anti-factor Xa activity, now preferred by international guidance,⁸ was added only midway through the study and not used for dose adjustment. The open-label design and the wide CIs preclude any claim of superiority.

The trial provides prospective randomised evidence that historical high-intensity targets are no longer justified in most patients on ECMO and validates what many centres are already doing. Guidelines should be updated accordingly. How to reliably monitor reduced-intensity regimens remains unresolved. Anti-factor Xa is a step forward, but the optimal monitoring strategy at low intensities is still undefined.⁷ Viscoelastic testing, when specifically adapted and validated for the ECMO setting, might add complementary information on heparin effect, fibrinolysis, and platelet contribution that conventional assays cannot provide.⁹ More importantly, a single target is unlikely to fit all patients on ECMO. Bleeding and thrombotic risk vary widely across this population by comorbidity, indication, surgical status, and inflammatory milieu:¹⁰ two patients with severe acute respiratory distress syndrome, one from diffuse alveolar haemorrhage in antineutrophil cytoplasmic antibody-associated vasculitis and one with a mechanical valve, do not have the same anticoagulation needs. A risk-stratified approach, initially anchored on clinical criteria such as previous thromboembolism, active bleeding, or major surgery, and subsequently refined by biomarkers and inflammation profiles, seems to be the natural next step. The LMWH findings also legitimise subcutaneous anticoagulation as a viable option in selected patients, an alternative to UFH that had never been formally endorsed during ECMO.

This trial settles a long-standing controversy and should prompt an update of anticoagulation targets during ECMO. After years of converging observational signals, an adequately powered randomised trial finally provides the level of evidence the question deserved. The practical message is straightforward: less intensity is not less safety. What clinicians still have to learn is how to monitor reduced anticoagulation intensity in ECMO reliably, and how to individualise it within a population that is anything but uniform.