Potential Impact of Direct Versus Indirect Central Venoarterial Extracorporeal Membrane Oxygenation (VA-ECMO) Cannulation in Refractory Postcardiotomy Cardiogenic Shock.

Abstract

Background Central venoarterial extracorporeal membrane oxygenation (VA-ECMO) is a commonly employed strategy to support patients in refractory postcardiotomy cardiogenic shock (RPCS). This support can be provided using either indirect central ECMO (icECMO) with a closed thorax or direct central ECMO (dcECMO) with an open thorax. Methods This single-center retrospective analysis included 60 patients undergoing central VA-ECMO for RPCS from January 2019 to December 2020. The primary endpoint of this study is to compare 30-day survival outcomes between the icECMO and dcECMO approaches in RPCS patients. Secondary endpoints include the evaluation of adverse events and the identification of predictors that influence 30-day mortality. Results The study included 60 patients, 25 received icECMO and 35 treated with dcECMO due to RPCS. The icECMO group demonstrated significantly better 30-day survival rates (icECMO; 10 [40%] vs. dcECMO; 5 [14.3%], log-rank test; p=0.042). Despite comparable ECMO flow rate and ECMO RPM (rotations per minute) in the first day between the study groups ([icECMO; 4.5 l/min vs. dcECMO; 4.6 l/min, p=0.124], [icECMO; 3510 rpm vs. dcECMO; 3800 rpm, p=0.115], respectively), lactate levels were significantly higher in the dcECMO group on the 1st and 3rd post-extracorporeal life support (ECLS) days (p=0.006 and p=0.008, respectively). Gastrointestinal ischemia was more common in the dcECMO group (p=0.036). Successful ECMO weaning was more frequent in the icECMO group (56% vs. 22.9%, p=0.014). Multivariable logistic regression identified arterial lactate on the first day with a cutoff 4 mmol/las an independent risk factor for 30-day mortality with Exp(B) of 8.9 (p=0.007). Conclusions Our findings suggest a potential survival advantage with the icECMO technique in patients undergoing central ECMO cannulation after RPCS. However, larger prospective studies are essential to confirm this observation.