Unloading the left ventricle in VA-ECMO: a comparative analysis of outcomes

Abstract

Abstract Funding Acknowledgements None. Introduction Venoarterial Extracorporeal Membrane Oxygenation (VA-ECMO) has radically changed the way we see and treat cardiogenic shock, despite the challenges it presents. The retrograde delivery of oxygenated blood to the aorta via the arterial cannula imposes increased afterload on an already strained left ventricle (LV). Unloading techniques aim to reduce the preload and afterload caused by VA-ECMO on the LV. Purpose Our study aims to compare the clinical outcomes of patients on VA-ECMO where additional invasive techniques were employed to unload the LV, against those without unloading strategies. Methods Patients treated with VA-ECMO in our hospital since 2012 were categorized into two groups: the Venting Group (VG) that received LV unloading interventions and the Non-Venting Group (NVG) without such interventions. We collected comprehensive clinical data, including demographics, baseline characteristics, ECMO-related parameters, in-hospital complications, and patient outcomes. Statistical analyses involved chi-square tests for categorical variables and independent samples t-tests for continuous variables to assess the impact of unloading strategies on patient outcomes. Results Our hospital treated 100 patients with VA-ECMO (mean age 58.4 ± 11.9 years, 60% male). Approximately 39% of patients received LV unloading support. The choice of unloading device was guided by our center's expertise. Most commonly, Intra-Aortic Balloon Pump (IABP) was used (61.5%), followed by Impella CP (23.1%), percutaneous atrial septostomy (2.6%) and surgical LV venting (12.8%). Unloading techniques were implemented either concurrently with VA-ECMO cannulation (58.9%), before (17.9%), or after its insertion (23.2%). Most patients (68%) received inotropic drugs, predominantly dobutamine. The mean duration of venting device use was 4.2 ± 3.8 days. Airway bleeding (3.3% NVG vs. 17.9% VG, p=0.032) and thromboembolic complications (19.7% NVG vs. 41.0% VG, p=0.036) were significantly more prevalent in the VG. There were no significant differences in the others reported ECMO-related complications. No significant differences were observed in the mean duration of support with VA-ECMO (3 days NVG vs. 4 days VG, p=0.147) or in the mean duration of hospitalization (13 days NVG vs. 12 VG, p=0.619) between the two groups. The 30-day survival rates were 42% (95% CI, 0.30-0.59) in NVG and 29% (95% CI, 0.18-0.49) in VG, which did not display statistical significance, p=0.24 [Figure 1]. Conclusions In our cohort, incorporating unloading devices alongside VA-ECMO did not yield major differences in survival rates, yet led to a considerable increase in airway bleeding and thromboembolic complications. The precocity of unloading device implantation occurred either as an initial strategy or to prevent increased afterload in hearts with limited contractile reserve.Characteristics and Clinical OutcomesSurviving Curves VG vs. NVG