Anticoagulation in Extracorporeal Membrane Oxygenation for Trauma: Striking a Balance

Extracorporeal membrane oxygenation in a venovenous configuration (VV-ECMO) is increasingly recognized as an important therapeutic option for patients with severe acute respiratory distress syndrome (ARDS).1 Accumulating scientific evidence as well as experience garnered during both the 2009 influenza A (H1N1) and coronavirus 2019 (COVID-19) pandemics, have resulted in widespread adoption of VV-ECMO for patients with ARDS.2–5 In this context, ECMO centers have expanded the application of VV-ECMO to other indications, including severe respiratory failure due to traumatic injury.6–8 The article by Niles et al.9 offers an important contribution to the field of ECMO for respiratory support in traumatically injured patients. In this single-center, retrospective cohort study, 75 patients with trauma-related acute respiratory failure were managed with VV-ECMO over an 8 year period. Pulmonary contusion was the most common etiology of respiratory failure (49% of patients) with aspiration (15%) and transfusion-related acute lung injury (11%) as other leading causes. Thirty-two percent of the patients experienced cardiac arrest before initiation of VV-ECMO and approximately 50% of patients had traumatic brain injury (TBI). Overall survival to hospital discharge was 68%. Thirty-eight (51%) patients received no anticoagulation during ECMO. Of the 37 patients who received systemic anticoagulation, 27 (73%) had anticoagulation infusion held at some point during the ECMO run. In multivariable logistic regression, there was no association between exposure to anticoagulation and survival. Nearly half of the total cohort developed a hemorrhagic complication and 28% developed a thrombotic complication during ECMO, most commonly oxygenator thrombus requiring exchange. Notably, there was no significant difference in hemorrhagic or thrombotic events based on anticoagulation status. Nearly all patients were transfused packed red blood cells (pRBCs) during the first 72 hours of ECMO support regardless of exposure to anticoagulation. Patients receiving systemic anticoagulation required a significantly lower number of transfusions of pRBCs and other blood products than did patients not on anticoagulation. Survivors were less likely to require transfusions of pRBCs and other blood products than were nonsurvivors. Traumatically injured patients pose management challenges due to the risk of bleeding (related to the injury itself or from surgical sites) and thrombosis (due to the potential for a trauma-related coagulopathic state resulting from vascular injury and activation of pro-inflammatory cytokines). The use of ECMO introduces another level of complexity; systemic anticoagulation is recommended during ECMO to prevent circuit and systemic thrombosis, yet may not be advisable in trauma patients due to the nature of the injury or ongoing procedural interventions.10 In this study, Niles et al.9 demonstrate the safety and feasibility of withholding systemic anticoagulation in patients with trauma-related acute respiratory failure supported with VV-ECMO. Evidence supporting the deferral of anticoagulation during ECMO has been limited, although described in patients with TBI.11,12 Niles et al.9 report a high overall survival, which is similar to the survival reported in other cohorts of VV-ECMO for traumatic injury and for patients receiving VV-ECMO for acute respiratory failure for other indications.5,7 After adjusting for relevant variables, survival was not related to anticoagulation status. By using a low activated partial thromboplastin time (aPTT) target rather than full therapeutic anticoagulation, Niles et al.9 provide further support for this approach in VV-ECMO patients with traumatic injury and suggest it appropriately balances bleeding and thrombotic risk.13–15 Niles et al.9 report that patients who received anticoagulation required fewer transfusions than their counterparts who never received anticoagulation. Although patients receiving anticoagulation may have been less critically ill and less likely to suffer from ongoing bleeding (thus rendering them better candidates for ongoing anticoagulation at baseline), the case-by-case decision-making described by Niles et al.9 reflects the dilemma that ECMO practitioners often face at the bedside: what level of bleeding risk renders the initiation of systemic anticoagulation inappropriate? When should anticoagulation be held in the event of bleeding or coagulopathy? Niles et al.9 show that a personalized approach to decision-making based on individual risk assessment may offer a safe strategy for anticoagulation management during VV-ECMO for trauma-related acute respiratory failure. Furthermore, whereas the conclusions of the study should be extrapolated with caution in other populations before additional evaluation, the safety profile observed with holding anticoagulation in this cohort may offer guidance when considering the need to hold anticoagulation for patients receiving VV-ECMO for respiratory failure in the absence of traumatic injury. This study has several limitations including its single-center, retrospective design and the potential for unmeasured confounders that may have impacted patient outcomes. Furthermore, the authors note that their analysis only included thrombotic events occurring during the ECMO run. Prior studies have reported an incidence of cannula-associated deep vein thromboses (CaDVT) as high as 85%.16 In this population, it would be informative to measure thrombotic events following ECMO decannulation to assess if the incidence of CaDVT or systemic thromboembolism differed by anticoagulation status. Finally, the decision of whether to initiate ECMO and administer systemic anticoagulation may be affected by institutional experience in caring for both trauma and ECMO patients. Practice patterns may differ at other institutions and impact outcomes. Future studies may be informed by use of multi-institutional registry data or multicenter prospective analyses to minimize the risk of confounding and to broaden the generalizability of the results. Although prospective, randomized controlled trials of patients receiving ECMO come with many challenges, a clinical trial of anticoagulation during VV-ECMO for traumatic injury may best offer definitive guidance on the safest approach to managing these critically ill patients. In conclusion, this study suggests that anticoagulation can be safely held in patients with traumatic injury receiving VV-ECMO without increasing mortality, complications, or transfusion requirements. Niles et al.9 offer important insights and provide further evidence supporting the use of VV-ECMO in patients with severe respiratory failure caused by traumatic injury.