Early clinical and hemodynamic predictors of successful weaning from V-A ECMO: a single-centered retrospective study

Abstract Background/Introduction - Veno-arterial extracorporeal membrane oxygenation (VA-ECMO) is increasingly used in the management of cardiogenic shock (CS) with or without refractory cardiac arrest (CA). Successful weaning from VA-ECMO is defined as device removal without further requirement for re-cannulation over the following 30 days. Few data exist on clinical presentation, early hemodynamic parameters, and their association with successful VA-ECMO removal or mortality. Purpose - We aimed to evaluate the role of clinical presentation and early hemodynamic features as predictors of successful VA-ECMO weaning. Methods - Monocentric retrospective study on adults admitted in general intensive care unit from 2012 to 2020, receiving VA-ECMO due to refractory CS and/or CA. Clinical, hemodynamic and treatment data at time of cannulation and after 24 hours were collected. Primary endpoint was successful weaning from VA-ECMO. Results - 100 patients enrolled: 33.3% of them were female, median age was 54.0 [47.3-61.0] years-old. 77% were cannulated during extracorporeal cardio-pulmonary resuscitation and 23% due to CS. The frequency of ischemic etiology was 47.0% overall. 32% of patients had intra-aortic balloon pump (IABP) inserted at the same time of VA-ECMO cannulation and 95% patients were mechanically ventilated. Successful weaning from VA-ECMO occurred in 33.3% of cases: 16 (20.8%) CA and 17 (73.9%) CS patients. Median weaning time was 76.5[61.3-216.8] hours for CA and 144.0[87.0-218.0] hours for CS patients. 25 (25%) patients died within the first 24 hours. 30-days mortality was 33% in those who were weaned; the main cause for death after weaning was multiorgan failure due to septic shock. The differences between weaned and not weaned patients, both at time of cannulation and after 24 hours, are shown and summarized in figure 1. At univariate regression analysis, a significant association with primary endpoint was observed for lactate clearance, CA, and Sequential Organ Failure Assessment (SOFA) score at presentation, systolic blood pressure (SBP), mean arterial pressure (MAP) and pulse pressure (PP), both at baseline and after 24 hours, serum lactate, pH and Vasoactive Inotropic Score (VIS) both at cannulation and after 24 hours; furthermore, a significant association for norepinephrine and epinephrine doses at 24 hours was found. At multivariate analysis only lactate clearance was retained in the model (OR .047, 95% CI .036-.980; p=0.047) (Figure 2). Conclusion(s) - In CS patients requiring VA-ECMO, lactate clearance in the first 24 hours represents the strongest early predictor of successful weaning. However, a multiparametric approach considering clinical features, hemodynamic and echocardiographic parameters is advocated to define the best management accordingly and better stratify patients’ trajectory.