In these two patients, the initial mechanical support consisted of a combination of V-A ECMO with IABP for refractory cardiogenic shock. Although V-A ECMO is helpful for restoring a disturbance in systemic circulation, the prognosis of patients with LV failure is reportedly not as good as we expect, despite advanced mechanical support with V-A ECMO [8]. With it being likely that the endodiastolic LV pressure in these patients was remarkably elevated enough to induce pulmonary congestion, which indicated that an additional therapeutic option was required for LV decompression, we suggest that the main reason for this outcome is that the V-A ECMO increases the afterload to the damaged LV, leading to a further rise in LV wall stress and a subsequent reduction of subendocardial coronary flow [9].
We inserted the LV-Impella equipped with a microaxial flow pump directly across the aortic valve into the LV chamber. This served to unload the LV by venting out a certain quantity of blood from the LV cavity and delivering it into the ascending aorta, which made a retrograde flow of V-A ECMO reduced. In addition, as shown in the clinical course chart, we could reduce or eliminate the use of inotropic and vasoactive agents, which might increase wall stress and oxygen consumption in the injured myocardium under mechanical support with the combination of V-A ECMO and LV-Impella. A previous retrospective study using a multicenter cohort compared the outcome of treatment between V-A ECMO alone and a combination of V-A ECMO and LV-Impella in 157 patients suffering from severe cardiogenic shock, including five patients with fulminant myocarditis, of whom 123 (78%) received V-A ECMO alone and 34 (22%) received a combination of V-A ECMO and LV-Impella, respectively [10]. Using propensity score matching, the study showed that the combined treatment group had a lower rate of hospital mortality (47% vs. 80%, P < 0.001) and a higher rate of successful transition to either recovery or further therapy (68% vs. 28%, P < 0.001) [10].
A more extensive retrospective study in a multicenter registry enrolled 34 patients with fulminant myocarditis employing LV-Impella [11]. Despite that study being retrospective and enrolling a small number of patients so it might have had limited information regarding the mechanical support for fulminant myocarditis, the study suggested the safety and effectiveness of LV-Impella based on the survival rate of 62% (21 patients, including a successful recovery of 15 patients) in the setting of cardiogenic shock [11]. In addition, although a multicenter cohort study in 686 cardiogenic shock patients treated with V-A ECMO showed that an introduction of LV-Impella < 2 h after V-A ECMO was associated with lower mortality risk. However, the optimal timing of unloading with LV-Impella was still unclear [12]. Therefore, several questions remain to be answered, including the optimal timing of initiation of LV-Impella and the ideal duration of support necessary for recovery or to obtain a bridge to further therapy.
Myocarditis is an inflammatory disease, and in affected myocardial tissue, active inflammation such as cellular infiltrates develops, resulting in damage to cardiomyocytes. It is possible that early unloading to the affected LV during the active inflammatory phase may have a positive therapeutic effect against the progression of inflammatory insult, which is different from cases of ischemic LV dysfunction [9]. In general, the use of the LV-Impella on V-A ECMO as soon as cardiogenic shock emerges, if possible, may result in a better prognosis. Nishihara et al. reported a case of thrombosis in the ascending aorta during V-A ECMO management and suggested combined use of LV-Impella from as soon as possible especially in case of very low cardiac function [13].
However, in the cases reported here, we were unable to introduce the LV-Impella soon after the sudden development of cardiogenic shock because it took several hours to transfer the patients to our university hospital. Although there was a time delay of several hours until the initiation of LV-Impella support, these two patients achieved a full recovery of LV function within approximately 6 days after the start of LV unloading. Thus, a delay of several hours until the introduction of the LV-Impella might be acceptable, with the expectation that an almost full recovery of LV function can be achieved in a short duration, such as 6 days. Therefore, while the decision to initiate the LV-Impella should be made within at least 24 h from the establishment of V-A ECMO, once the patient is stabilized with cardiogenic shock by mechanical support with V-A ECMO, physicians have time to decide whether the LV-Impella is necessary. Subsequently, further accumulation of cases with fulminant myocarditis is required to clarify our suggestion for treatment with LV-Impella.
In these two patients, mechanical support using the LV-Impella was implemented in addition to V-A ECMO within 24 h after the development of severe cardiogenic shock. They had a full recovery of LV function within approximately 6 days after LV unloading.