Low-Dose Heparin in Critically Ill Patients Undergoing Extracorporeal Membrane Oxygenation - the Help-ECMO Pilot Randomised Controlled Trial

Extra corporeal membrane oxygenation (ECMO) is a rescue therapy for reversible cardiac and/or respiratory failure. Despite improvement in management of patients undergoing ECMO, mortality remains high. Due to thrombosis risk, which includes arterial and venous thrombosis as well as in the extracorporeal circuit and components, variable intensity systemic anticoagulation with unfractionated heparin is routinely used. However, bleeding is one of the most frequent complications, can be severe and is independently associated with worse outcomes. Optimal anticoagulation to prevent thrombosis whilst minimising bleeding in adults on ECMO remains unknown. Before conducting a large randomised controlled trial to determine whether lower anticoagulation intensity is safe and effective compared with therapeutic anticoagulation, we aimed to determine the feasibility of randomising ECMO patients to these anticoagulation protocols.

Methods:

The HELP-ECMO pilot study (ACTRN12613001324707) is a randomised, controlled, unblinded trial at two Australian intensive care units (ICUs). Inclusion criteria were ICU patients who required ECMO (venous-venous [VV] or venous-arterial [VA]). Patients who did not meet any exclusion criteria were randomised to receive either therapeutic anticoagulation with heparin (target activated partial thromboplastin time [aPTT] between 50 and 70 seconds) or low dose heparin (12000 units/24 hours aiming for aPTT <45 seconds). Paired aPTT and anti-Xa assays were taken according to protocol, and at a minimum once a day. The primary endpoints for feasibility were difference in mean heparin dose, aPTT and anti-Xa levels between both study groups. Secondary endpoints included incidence of thromboembolic events, ECMO circuit thrombosis and bleeding events. All primary outcomes were log-transformed prior to analysis and reported as geometric means (95%CI) with overall differences determined using Repeat measures ANOVA.

Results:

Between May 2014 and March 2016, 31 patients who underwent ECMO (9 (29%) VA and 22 (71%) VV) were enrolled; 16 were randomised to low dose and 15 to therapeutic dose heparin. The groups were similar in age (mean 41 years [SD 16.8] vs 43 [SD 17.6] p=0.75), gender (68% vs 80% male, p=0.47), type of ECMO (31% vs 27% VA, p=0.78) and Acute Physiology and Chronic Health Evaluation III illness severity score (mean 65.4 [SD 23.5] vs 61.8 [SD 30.1], p=0.72) and sepsis-related organ failure assessment score (mean 10 [SD 3.6] vs 10 [SD 3.3], p=1.0). The mean duration of ECMO support was 9.33 days (SD 5.97) in the low dose and 9.79 days (SD 4.77) in the therapeutic dose group (p=0.82).

For the primary outcomes, there was a significant difference in the daily mean aPTT (48.1 [95% CI 43.5-53.3] vs 56.2 [95% CI 50.7-62.3], p=0.03), daily mean anti-Xa (0.11 [95% CI 0.07-0.18] vs 0.30 [IQR 0.19-0.46], p=0.003) and daily mean heparin dose (11784 units [95% CI 8693-15972] vs 22050 [IQR 16262-29899], p=0.004) in the low dose compared to therapeutic group.

There was no difference in thrombotic complications with regard to DVT (2 [12.5%] vs 3 [20%], p=0.57), PE (1 [6.3%] vs 0 [0%], p=0.33), stroke (no events), intracardiac thrombus (1 [6.3%] vs 2 [13.3%], p=0.51), acute pump (1 [6.3%] vs 1[6.7%], p=0.96) and distal perfusion cannula thrombosis (2 [12.5%] vs 0 [0%], p=0.16) in low dose compared with therapeutic group, respectively.

With regard to bleeding, there was no difference in intracranial haemorrhage (ICH) (0 [0%] vs 1 [6.7%], p=0.29), retroperitoneal bleeding (1 [6.3%] vs 1 [6.7%], p=0.96), gastrointestinal bleeding (0 [0%] vs 2 [13.3%] , p=0.13), or haemoptysis (1 [6.5%] vs 1 [6.7%], p=0.96) in low dose compared with therapeutic group, respectively.

Conclusion: In this pilot trial, administration of a low dose heparin protocol was feasible, and resulted in a significant difference in mean heparin dose administered and daily aPTT and anti-Xa levels between groups. Low dose heparin was not associated with an increase in thrombotic events nor a decrease in bleeding events; however the study was not powered for these outcomes. Our findings support the feasibility of a larger phase III study to evaluate the safety and efficacy of low-dose anticoagulation compared with therapeutic heparin with regard to thrombotic and bleeding events in patients receiving ECMO.