Biomarker studies to predict outcomes of patients with COVID-19 related acute respiratory distress syndrome measured pre and post initiation of veno venous extracorporeal membrane oxygenation

Veno-venous extracorporeal membrane oxygenation (V-V ECMO) is a resource intensive and expensive therapy increasingly used to treat a select group of patients with severe acute respiratory distress syndrome (ARDS). High levels of Interleukin-10 (IL-10) are reported to be predictive of mortality in non-COVID ARDS patients receiving ECMO. In this study we evaluated if levels of biomarkers including IL-10 can predict mortality if measured prior to ECMO initiation in patients with COVID-19 ARDS. A total of sixteen patients with COVID-related ARDS undergoing treatment with ECMO were included in the pre-ECMO biomarker analysis, while samples from a total of 22 patients with post-ECMO samples were analyzed. Nine of sixteen patients with pre-ECMO samples (56%) survived to hospital discharge. There was no difference in baseline demographics between survivors and non-survivors. In univariate analysis, survivors had no difference in IL-10 levels prior to ECMO initiation although we identified several other biomarkers (Ang 1 and TNF-alpha with logistical regression, Mitochondrial DNA and IP-10 with Kaplan-Meier analysis) that differed significantly between survivors and non-survivors. Moreover, as opposed to a prior study, area under the curve (AUC) analysis showed that IL-10 levels were not predictive of survival. We advise caution for utilizing a single biomarker as outcome predictor in this highly complex population with often long clinical courses despite our promising results.