ELLIPSE1Impella vs. IABP pour choc cardiogénique sous VA-ECMO

Over the last decade, veno-arterial extracorporeal membrane oxygenation (VA-ECMO) has become the mechanical circulatory support of choice for cardiogenic shock (CS) refractory to medical treatment. VA-ECMO allows to supplement cardiac function until myocardium recovers, or in bridge to long-term left ventricular assist device (LVAD) or to heart transplant. However, recent randomized controlled trials (RCT) and meta-analysis have failed to demonstrate any benefit of VA-ECMO in terms of 30-day survival compared with optimal medical treatment in CS related to myocardial infarction. Reasons for these disappointing results are multifactorial. The associated risk of temporary mechanical circulatory support (t-MCS)-related complications might counterbalance any hemodynamic benefit. In addition to hemorrhagic and ischemic complications, the significant myocardial impact due to VA-ECMO drawbacks should not be overlooked. Indeed, while VA-ECMO restores systemic perfusion, it can also lead to an increase in left ventricle (LV) loading conditions. Although LV distension is not consistently observed, recent publications strongly suggest that the origin of LV overload seems to be multifactorial and contributes partially to cardiac remodeling through the modulation of cardioprotective cellular pathways, resulting in reduced cardiomyocyte apoptosis. Furthermore, peripheral VA-ECMO affects myocardial contractility and increases myocardial work (potential energy, stroke work, and pressure-volume area), particularly at native low blood flow. Thus, cardiac recovery may be compromised and weaning from VA-ECMO delayed. This vicious cycle affects the patients overall prognosis, as delayed and possibly failure of VA-ECMO weaning and may convert the initial medical strategy of recovery towards heart transplantation or LVAD. To optimize the chance of VA-ECMO weaning, early LV unloading may be a good therapeutic option. It aims at increasing coronary flow directly and, indirectly, by improving sub-endocardial myocardial perfusion through decreasing LV wall stress and myocardial oxygen consumption. Currently, 2 techniques are mainly used, the intra-aortic balloon pump (IABP) and the microaxial flow pump as Impella device familly (ECMELLA). These invasive techniques carry risks of bleeding and thromboembolic complications, and the benefit/risk ratio of their use for myocardial recovery is not clearly established. A few retrospective studies suggest that early left ventricular unloading during VA-ECMO could improve prognosis. In the absence of RCT, the choice of the technique (Impella or IABP) is mainly driven by center practice. A multicenter randomized clinical trial would be the best choice to address this question. However, due to the acute poor prognosis of these patients (50% of early deaths), reliable preliminary data about the expected effect size are necessary to design the most efficient clinical trial. Recent epidemiologic developments of observational studies, the emulated trials, allow a better control of immortal time bias and indication bias. This innovative multicenter study will compare the effectiveness and safety of LV unloading by Impella (ECMELLA) versus IABP in terms of survival with myocardial recovery in patients with refractory CS