The Janus Face of Coronavirus Disease 2019–Associated Coagulopathy

To the Editor:

We congratulate Doyle et al (1) on their exciting and clinically relevant report published in a recent issue of Critical Care Medicine comparing thromboembolic complications and bleeding events in coronavirus disease (COVID)–19 acute respiratory distress syndrome (ARDS) with influenza A patients on veno-venous extracorporeal membrane oxygenation (ECMO). The key finding that COVID-19–induced ARDS is associated with higher thromboembolic events despite therapeutic anticoagulation is of high relevance to the community of intensivists. That being said, the reported high rate of both incident intracranial hemorrhages (ICHs) in eight of 51 COVID-19 patients (16%) early after ECMO cannulation and the additional three cases of ICHs (6%) during ECMO support astonishes us. Besides the morbidity attributable to thromboembolic complications, the fact that a total of 22% of COVID-19 patients experienced an ICH during ECMO support is alarming and raises several important questions.

First, is there a reasonable pathophysiologic explanation for this unexpected high rate of ICHs in COVID-19? As a reminder, ICHs were observed only in 2% in the ECMO to Rescue Lung Injury in Severe ARDS (EOLIA) trial (2) and 4% in patients with H1N1 influenza associated ARDS (3). General features of bleeding vulnerability such as therapeutic anticoagulation and ECMO-associated hemotrauma should be taken into account together with COVID-19–induced unique features such as a potential endothelial susceptibility in the sense of a vasculitis.

Second, the severity of the ICHs and the clinical response is not elaborated. Did these patients receive heparin-free ECMO support or was heparin dose merely reduced? Were the findings on cranial CT minor or did they require surgical intervention? In light of the global resource-shortness during the COVID-19 pandemic and the assumed unfavorable outcome associated with ICH on ECMO, we would be very much interested in the survival of these ECMO patients.

Third, a broad range of prophylactic (0.3–0.7 U/L) and therapeutic (0.6–1.0 U/L) anti-Xa levels were targeted during ECMO support in this study. Did the authors adjust this target range according to the supposed risk of present or impending thromboembolism and bleeding? One could hypothesize that the theoretical rational regarding endotheliopathy and bleeding vulnerability might justify 1) a narrow and 2) a lower range of anti-Xa. Interestingly, recent evidence supports the notion that target levels above 0.46 IU/mL might represent an independent risk factor for severe bleeding under ECMO support (4). At the same time, hemostaseologists would probably argue that anti-Xa levels do not fully reflect the important effects heparin has on other coagulation enzymes and the expression of tissue factor and tissue factor pathway inhibitor (5).

The complex coagulopathy in light with the endothelialitis found in critically ill COVID-19 patients might literally be “Janus faced” with regard to the simultaneous risk of thrombosis together with potentially fatal bleeding events (particularly if additional factors such as an extracorporeal circulation comes into play). Both the associated mortality and the long-term morbidity for thrombosis versus bleeding highlight the utmost importance of a tailored individualized approach to choose the right degree of anticoagulation.