Prospective Multicenter Analysis of Hemoadsorption in VA-ECMO Patients: Analysis from the international COSMOS Registry

Introduction The International COSMOS (CytOSorb TreatMent Of Critically Ill PatientS) captures real-world of haemoadsorption with the use is CytoSorb (CS) in critical care. The present sub-analysis evaluates CS use in VA-ECMO patients, focusing on timing, clinical effects, and safety. Methods Data were collected prospectively at predefined timepoints: screening, 24 hours before, during, 24 hours post-CS-treatment, ICU discharge, and at 90-day follow-up. Results are reported as mean±SD or median [IQR]. Results A total of 45 VA-ECMO patients (mean age 59.5±13.13; 24.4% female) with median APACHE II score 26.0 [19.0, 32.0] were included. On average, 4.2±3.63 CS adsorbers were used; 44.4% received ≥4 devices. CS treatment duration averaged 72 hours. CS was integrated within ECMO (11.1%), renal replacement therapy (75.6%), or as hemoperfusion (13.3%). Initiation was before (n=3), simultaneously (n=15), or after ECMO start (n=27; 17 ≤5 days, 10 >5 days). Median norepinephrine dose decreased (from 0.138 [0.063, 0.315] to 0.063 [0.0, 0.23] µg/kg/min; p=0.067) and MAP/NE index increased significantly (p=0.008). Early CS use (≤5 days after ECMO start) was associated with shorter overall ECMO duration (7 vs. 18 days; p <0.001) and ICU stay in survivors (20 vs. 57 days; p=0.004). ICU mortality was 38.6%, and 90-day mortality was 46.3%. Median platelet count changes were not significant. No device-related adverse events or deficiencies were reported. Conclusions Findings from the ongoing COSMOS Registry suggest that CS was associated with improved haemodynamic stability in VA-ECMO patients without significant safety concerns. Early CS initiation correlated with shorter ECMO duration and ICU stay in survivors.