Timing of Left Ventricular Unloading with Impella Device in Patients with VA-ECMO: A Propensity Score–Matched Analysis

Objectives: Venoarterial extracorporeal membrane oxygenation (VA-ECMO) or percutaneous microaxial Impella pump provide mechanical circulatory support (MCS) in patients with refractory cardiogenic shock (RCS). Recent reports suggest that MCS with both venoarterial VA-ECMO combined with concomitant Impella percutaneous ventricular assist device (pVAD) to unload the left ventricle may favorably modify patient outcomes. We sought to investigate the impact of different timing strategies (VA-ECMO first vs. pVAD first) on outcomes of patients under dual MCS.

Methods: Between January 2014 and August 2019, all consecutive patients who had concomitant treatment with VA-ECMO and Impella pVAD were evaluated. Depending on the primarily implanted device, patients were divided into two groups (Impella first vs. VA-ECMO first). Thirty-day mortality and morbidity rates were evaluated. Secondary end-points included duration of dual MCS and bridging rate to myocardial recovery or transplantation/durable MCS. Because of non-randomized group assignment, nearest neighbor 1:1 propensity score matching was performed to account for potential confounding variables.

Results: A total of 139 patients underwent dual MCS therapy. Of these, 74 cases were eligible for propensity score matching of statistically significant variables (standardized difference > 0.20) and 37 patients were assigned to each group. Patients in the Impella-first group showed significantly lower 30-day mortality (27 vs. 62%, p = 0.002), and had a certain trend toward lower risk of experiencing stroke (16 vs. 30%, p = 0.167). Myocardial recovery/successful bridge to transplantation or durable MCS was possible in 42 patients (56.8%), more frequent in Impella first group (70 vs. 43%, p = 0.019). Major, minor surgical, and vascular (e.g., bleeding, ischemia) complications were comparable between the two groups. Pneumonia rates, necessity of renal replacement therapy or liver dialysis, gastrointestinal complications, and sepsis rates were also equivalent in these two groups.

Conclusion: First-line Impella therapy escalated by ECMO demonstrated significantly lower mortality than patients in the ECMO first group. The beneficial effects of a primary versus secondary LV unloading strategy with Impella ventricular assist device therapy has to be evaluated in the setting of prospective, randomized trials. In the absence of more compelling evidence, an early left ventricular unloading strategy may be considered in selected RCS patients.