Abstract Background International organisations advocate the use of extracorporeal cardio-pulmonary resuscitation (ECPR) with veno-arterial extracorporeal membrane oxygenation (VA-ECMO) in selected patients with therapy-refractory cardiac arrest [1–3]. Although VA-ECMO allows for full circulatory support, it is inherent to increased left ventricular (LV) pressure due to retrograde aortic perfusion, which may hamper myocardial recovery and aggravate pulmonary oedema. In order to mitigate these negative sequelae, adjunct LV unloading with an Impella microaxial flow pump may be considered. The effects of concomitant treatment with VA-ECMO and Impella (ECMELLA) in patients with therapy-refractory cardiac arrest due to acute myocardial infarction (AMI) remains unclear. Objectives To the best of our knowledge this is the first study to investigate whether treatment with ECMELLA is associated with improved 30-day mortality rate in patients with therapy-refractory cardiac arrest caused by AMI, compared to treatment with VA-ECMO alone. Methods Patients treated with ECMELLA were propensity score (PS)-matched to patients receiving VA-ECMO based on age, electrocardiogram (ECG) rhythm, cardiac arrest location (out-of-hospital or in-hospital) and Survival After Veno-Arterial ECMO (SAVE) score. Cox proportional-hazard and Poisson regression models were used to analyse 30-day mortality rate (primary outcome), hospital and intensive care unit (ICU) length of stay (LOS) (secondary outcomes). Multiple sensitivity analyses on patient demographics and cardiac arrest parameters were performed. Results 95 adult patients from three tertiary care centers were included, out of whom 34 pairs were PS-matched. ECMELLA treatment was associated with 47% decreased 30-day mortality risk [95% Confidence Interval (CI) 0.31–0.91, P=0.021], 71% prolonged hospital [95% CI 1.50–1.95, P<0.001] and 81% prolonged ICU LOS [95% CI 1.57–2.08, P<0.001]. Kaplan-Meier analyses (Figure 1) and multiple sub-group analyses (age, sex, initial ECG rhythm, Charlson comorbidity index, body mass index, SAVE score, cardiac arrest location, lactate and pH levels) confirmed survival benefits in the ECMELLA group. Especially patients with prolonged low-flow time and high initial lactate benefited from ECMELLA therapy. Moreover, LV ejection fraction strongly improved in the ECMELLA group between ICU admission and ICU discharge from 15% to 40%, compared 15% and 20% in the VA-ECMO group. Conclusion In this multicenter propensity score-matched cohort of patients with ECPR during therapy-refractory cardiac arrest caused by AMI, treatment with ECMELLA was associated with improved survival compared to treatment with VA-ECMO alone. These findings support current guideline recommendations on early evaluation of ECPR in well selected patients with therapy-refractory cardiac arrest. A clinical trial is urgently needed to further evaluate the role of LV unloading in patients with therapy-refractory cardiac arrest. Funding Acknowledgement Type of funding sources: None.