Online Letters to the Editor
To the Editor:
We read with great interest the article published in a recent issue of Critical Care Medicine on the prevalence of intracranial hemorrhage (ICH) in COVID-19 patients on extracorporeal membrane oxygenation (ECMO) by Seeliger et al (1).
We were struck by the high prevalence of ICH in COVID-19 patients on ECMO. This retrospective article analyzes ICH occurrence rate and clinical outcomes in patients on ECMO due to COVID-19–induced acute respiratory distress syndrome (C-ARDS) compared with other viral acute respiratory distress syndrome (ARDS). The article reported ICH in 29 of 142 COVID-19 patients (20%), including 15 major bleeds.
We have not observed a similar predilection to ICH in COVID-19 at our institution. We have cared for 69 COVID patients requiring venovenous ECMO support for C-ARDS, of whom 66.6% survived to hospital discharge. Of these patients, only two patients (2.9%) suffered ICH, one while on ECMO and the other following her ECMO run. Both patients recovered completely without neurologic sequelae. The reason for the discrepant rates of ICH in C-ARDS patients is unclear. Our patient population was younger, with a mean age of 43 (range, 16–68). Severity of ARDS at the time of cannulation appears similar between the groups. Our practices for screening for ICH mirror those of Seeliger et al (1). However, only 13 patients in our cohort had head imaging which is a lower percentage of imaging than the authors’ report, so some subclinical bleeds could be missed. Regardless, the rate of major/fatal bleeds observed by Seeliger et al (1) still far exceeds our observed rate.
Perhaps the biggest difference between cohorts is the anticoagulation strategy. Unfractionated heparin (UFH) was primarily used by Seeliger et al (1). In contrast, ~80% (n = 55) of our patients were anticoagulated with bivalirudin targeting an activated partial thromboplastin time between 50 and 80 seconds. In our patients, 14.5% (n = 10) started out with UFH and were transitioned to bivalirudin, whereas only 5.8% (n = 4) were anticoagulated with UFH exclusively.
Our institutional data does not show an increased prevalence of ICH in C-ARDS patients on venovenous ECMO. Based on our experience, we postulate that bivalirudin may represent a safer strategy for anticoagulation in C-ARDS patients on VV-ECMO. Further study with prospective clinical trials is warranted to confirm these findings.