Predictors of successful weaning from VA-ECMO: the role of CS etiology and related treatment

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Introduction Veno-arterial extracorporeal membrane oxygenation (VA-ECMO) is a short-term mechanical circulatory support for refractory cardiogenic shock (CS). Successful weaning from VA ECMO is defined as device removal without further requirement for re-cannulation over the following 30 days. Few data exist regarding timing and reliable parameters for weaning from VA-ECMO: most study focused on echocardiographic and invasively monitored parameters during weaning trial(1,2). No data exist concerning pharmacological support and standard haemodynamic parameters in the first 24 hours since VA-ECMO implantation. Purpose We sought to evaluate hemodynamic predictors of VA-ECMO weaning. Methods Single-center retrospective observational study of patients with refractory CS who underwent VA-ECMO insertion from 2013 to 2020 in Intensive Care Unit (ICU). Primary endpoint was successful weaning from VA ECMO. Demographic, clinical, hemodynamic data within the first 24 hours and outcome were collected. We also recorded maximum dose of each vasoactive medication, Vasoactive Inotropic Score (VIS)(3) after VA-ECMO implantation and the total amount of time exposition to vasoactive drugs. Results 100 patients (26% female; 52,6±12,6 y.o.) who underwent VA-ECMO cannulation were enrolled (Table 1). 31 were successfully weaned (31%) at a median time of 140,0 [74,3-229,0] hours; 30-days mortality was 47%. 79 (79%) patients were admitted for cardiac arrest (CA); 48 (48%) had acute myocardial infarction (AMI)-CS. CA patients were less likely to be weaned: 20,3% of CA patients had successful VA-ECMO removal while 79,4% was not weaned (p <0,001). In the weaned group, mean blood pressure was significantly higher (75,3±18,2 vs 61,4±18,3 mmHg; p <0,01).Patients who successfully removed VA-ECMO had lower serum lactate (8,8±5,2 vs 11,3±5,0 mmol/L; p <0,05), lower fluid balance (+2,3 [0,5-5,5] vs +4,1 [1,7-8,3] L/die; p <0,05) and lower Sequential Organ Failure Assessment (SOFA) score (11,2±2,6 vs 13,2±3,4; p <0,01). Lower doses of norepinephrine was administered in patient successfully weaned (0,27±0,17 vs 0,44±0,25 mcg/kg/min; p <0,01). Despite not significant, lower VIS and time to exposition to vasoactive drugs amongst weaned patients were observed (p=0,08 and p=0,09, respectively). Stepwise logistic regression confirmed the worst outcome of CA patients (OR 0,64; p <0,001), SOFA score (OR 0,81; p <0,05) and time to exposition to vasoactive drugs (OR 0,81; p <0,05) on successful weaning from VA-ECMO. Conclusions Mortality of patients undertaking VA-ECMO still remains high. Amongst the data in the model, the strongest negative association with successful weaning from VA-ECMO was found for CA, according to the latest classification entailing CA as CS severity modifier(4). Other results show higher requirement of norepinephrine and fluid administration for vasoplegia mostly related to hypoxic organ injury and related infection (i.e. bowel translocation).