Résumé
Objectives: Unfractionated heparin (UFH) remains the anticoagulation of choice at most centres for patients receiving extracorporeal membrane oxygenation (ECMO). One disadvantage of UFH relies on its individual dosing requirement to achieve target values. In this context heparin resistance has been described, defined as doses exceeding 35,000 IU UFH/d. However, the incidence of heparin resistance and its association with thromboembolic complications despite anticoagulation within target ranges remains unknown. Method(s): This retrospective study included adults receiving venovenous (VV) and venoarterial (VA) ECMO, or extracorporeal CO2-removal (ECCO2R) between 2010 and May 2022. The primary outcome was the incidence of heparin resistance (>35,000 IU of UFH/d). Secondary outcomes were heparin failure (thromboembolic complications despite anticoagulation within target ranges) and survival. A multivariable poisson regression model was fitted to analyse the effect of heparin resistance, COVID-19 and ECMO type on the incidence rate of thromboembolic events. Result(s): Of 197 included patients, 33 (16.8%) had heparin resistance. Patients with COVID-19 (n=51) had a higher rate of heparin resistance compared to nonCOVID-19 patients (37% vs. 9.6%, P<0.001). Thromboembolic complications occurred at a rate of 5.89/100 ECMO days. There was a significant effect of COVID-19 (incidence rate ratio (IRR) 2.12, 95% confidence interval (CI) 1.4 to 3.3, P<0.001) and ECMO type (VA ECMO: IRR 2.35;95% CI 1.43 to 3.87, P<0.001;ECCO2R: IRR 2.63, 95% CI 1.37 to 4.9, P=0.003;reference VV ECMO) on incidence rate of thromboembolic events, but not of heparin resistance (IRR 1.11, 95% CI 0.7 to 1.76, P=0.7). ECMO duration was longer (25d (IQR 11-33) vs. 8d (IQR 4-18), P<0.001) in patients with heparin resistance, but hospital survival did not differ (23 (70%) vs. 91 (57%), P=0.2). Conclusion(s): The study revealed a high incidence of heparin failure in ECMO patients, especially in those with COVID-19. Heparin resistance had no effect on the incidence rate of thromboembolic events, whereas our data suggest an increased risk in patients with COVID19, VA ECMO and ECCO2R.