ID:128 FC-D05-9 nfluence of Drainage and Return Cannulation Sites on Cannula-Associated Deep Vein Thrombosis After Veno-Venous Extracorporeal Membrane Oxygenation: A Retrospective Study

Background: Veno-venous extracorporeal membrane oxygenation (V-V ECMO) often requires prolonged support, during which cannula-associated deep vein thrombosis (CaDVT) can occur. However, the impact of cannulation configuration on CaDVT incidence has not been fully evaluated. We therefore aimeds to compare the incidence of CaDVT after V-V ECMO decannulation between femorojugular (F-J; femoral drainage–internal jugular return) and femorofemoral (F-F; femoral drainage–femoral return) configurations. Methods: This single-center retrospective observational study included adult patients successfully weaned from V-V ECMO in our ICU between January 2020 and December 2025. Post-decannulation contrast-enhanced CT was used to diagnose CaDVT as thrombosis in veins related to cannulation sites. Thrombosis extending into the superior or inferior vena cava was defined as great-vessel thrombosis. The primary outcome was CaDVT; secondary outcomes were great-vessel thrombosis and pulmonary embolism. Group comparisons were performed using the Mann–Whitney U test and Fisher’s exact test. Results: Twenty-eight patients were analyzed (F-J, n=20; F-F, n=8). Median [IQR] ECMO duration was 18.0 [13.8–26.5] days in the F-J group and 46.5 [26.0–93.5] days in the F-F group. Median [IQR] APTT during ECMO was similar (F-J, 43.9 [39.9–51.9] s; F-F, 43.3 [35.9–52.0] s). Overall CaDVT incidence was 67.9%, with no difference between configurations (F-J 70.0% vs F-F 62.5%, p>0.99). Great-vessel thrombosis occurred in 50.0% and pulmonary embolism in 10.7%, without between-group differences. Conclusions: CaDVT after V-V ECMO decannulation was frequent, but incidence did not differ between F-J and F-F configurations. Systematic post-decannulation thrombus assessment, including evaluation of the superior and inferior vena cava, is warranted regardless of configuration.