Old and new echocardiographic parameters to predict successful VA ECMO weaning

Abstract

Abstract Funding Acknowledgements Type of funding sources: Other. Main funding source(s): ESC Research Grant 2018, App000021138. Introduction Venoarterial extracorporeal membrane oxygenation (VA ECMO) is a type of temporary mechanical circulatory support (MCS) for acute cardiovascular failure. The population of VA ECMO patients is characterised by high mortality rates. No protocol for weaning from VA ECMO has been validated worldwide. The decision to wean a patient from VA ECMO is challenging because it is necessary to understand whether the patient can tolerate VA ECMO removal, but it is also important to avoid any delay to avoid complications. Purpose the aim of this pilot study was to assess whether strain by speckle tracking could give additional information to haemodynamic, and conventional echocardiographic parameters in identifying patients who develop adverse outcomes within 3 (± 1) months from VA ECMO removal. Methods observational prospective pilot study delivered over 3 years in 2 ECMO centres. The VA ECMO patients who underwent VA ECMO removal (not for palliation) after a weaning trial have been recruited and followed up. Conventional echocardiographic, haemodynamic and strain parameters (recorded during the VA ECMO weaning trial) of patients developing a composite clinical outcome (ClO) within 3 (± 1) months post VA ECMO removal (death for any reason, new necessity of high dose of ino-vasopressors or MCS after VA ECMO removal, new hospitalisation for heart failure) have been compared with those of patients free from ClO. A sub-analysis on composite cardiac outcome (CaO) development and an exploratory ROC analysis have been performed. Results Over 3 years of recruitment, 92 VA ECMO patients have been screened. 21 patients met the eligibility criteria for the study and 19 patients could be analysed. 5 patients experienced ClO (3 had CaO, 2 were complicated by septic shock necessiting high dose of ino-vasopressors). As reported in Figure 1, there were no differences between ClO+ and ClO– patients apart from indexed end diastolic volume (iEDV). iEDV was significantly lower in ClO– patients compared to the others (respectively 44.7 ml m-2 IQR 17.9; 70.4 ml m-2 IQR 43.0, p < 0.01). Focusing on CaO development, the median iEDV remained significantly higher in CaO+ (112.5 ml min-1m-2 IQR 47.1; p = 0.01). Furthermore, the median circumferential strain (absolute value) and the ejection fraction (EF) were significantly lower in CaO+ (Figure 2). At the ROC analysis, the best cut point to discriminate ClO+ and ClO- patients was: for EF 26.32% (AUC 0.79), for cardiac index 2.5ml min-1 m-2 (AUC 0.71) and for right ventricle (RV) free wall longitudinal strain -12.0%. (AUC 0.75). Conclusions VA ECMO weaning procedure is complex and requires the assessment of multiple variables. According to our analysis, large iEDV and low EF predispose to the development of CaO. Furthermore, low absolute values of circumferential strain and RV free wall longitudinal strain may be indicative of development of CaO and ClO respectively. The study is limited by the small number of events.