The Western Thoracic Surgical Association

Back to 2018 Program


Clinical Predictors of Successful Weaning from Veno-arterial Extracorporeal Membrane Oxygenation in patients with Refractory Cardiogenic Shock
Patrick Kohtz1, Mark Lucas1, Christopher Pierce1, Mohamed Eldeiry1, Yihan Lin1, Michael Bronsert2, Andreas Brieke3, Amrut Ambardekar3, *Michael Weyant1, *T. Brett Reece1, *Jay Pal1, *David Fullerton1, *Joseph Cleveland1, Muhammad Aftab1
1Division of Cardiothoracic Surgery, University of Colorado School of Medicine, Aurora, CO;2Adult and Child Consortium for Health Outcomes Research and Delivery Science and Surgical Outcomes and Applied Research, University of Colorado, Aurora, CO;3Division of Cardiology, University of Colorado School of Medicine, Aurora, CO

Background: Venoarterial-extracorporeal membrane oxygenation (VA-ECMO) is a lifesaving procedure in patients with refractory cardiogenic shock. This therapy is employed as a bridge-to-recovery, decision, ventricular assist devices (VADs) or transplant. Data on the predictors of successful weaning from VA-ECMO is lacking. This retrospective study aims to identify these predictors to recognize which patients would benefit from this therapy.
Methods: From 2013 to 2017, 123 patients diagnosed with severe, refractory, cardiogenic shock underwent VA ECMO support at our institution. Our institutional database was developed to assess their outcomes. Each patient has >650 variables evaluated during the index hospitalization. The etiologies of cardiogenic shock included post-cardiotomy 32(26%), acute myocardial infarction 27(22%), non-ischemic cardiomyopathy 35(28.5%), ischemic cardiomyopathy 17(13.8%) and miscellaneous 12 (9.7%). The mean duration of ECMO support was 6.77.0 days. We performed a bivariate analysis of the association of successful weaning from the VA-ECMO for 48 hours with several potential variables. We used Fisher's exact for categorical and t-test for continuous variables. In addition, we conducted forward stepwise logistic regression analyses for selected grouping of potential predictors. Multiple imputations were used for missing values for continuous variables. All statistical tests were considered to be significant at a 2-sided p < .05.
Results: Among 123 patients (66% male, aged 53.215.2 years), 75 patients (61%) were successfully separated from the VA-ECMO. Five patients needed re-institution of VA-ECMO support after initial successful weaning, two of these required it during the same hospitalization. Adjunct cardiac procedure including ventricular assist device, valve surgery, percutaneous coronary interventions, coronary artery bypass grafting, heart or lung transplantation were performed on 52 of 75 (70%) patients successfully separated from the VA-ECMO (P<0.001). Comparing the patients who were successfully weaned from VA-ECMO to those who were not, a total of 20 variables (Fig. 1) were identified as significantly associated with successful weaning after bivariate analysis. Multivariate modeling with four different models identified nine of these as predictors associated with successful or unsuccessful (Fig. 2) weaning from the VA-ECMO. The predictors associated with successful VA-ECMO removal include patients undergoing adjunct cardiovascular procedures, no vascular complications, sinus rhythm within 72 hours after the ECMO initiation, decompensated non-ischemic cardiomyopathy as etiology of cardiogenic shock and longer hospitalizations. Risk factors associated with VA ECMO weaning failure include high lactic acid, pre-existing coronary artery disease, post-cardiotomy cardiogenic shock and need for left ventricle venting during ECMO support.
Conclusions: VA-ECMO provides good clinical outcomes in patients with refractory cardiogenic shock, particularly in non-ischemic cardiomyopathy. Adjunct cardiac procedures are significantly associated with higher rates of successful weaning from VA-ECMO. Early identification of the predictors of successful weaning will help recognize the patients who would benefit from the prolonged VA-ECMO support. Further studies are needed to validate these predictors.


Back to 2018 Program