To the best of our knowledge, the present study was the first to directly address the effect of VA-ECMO on LV function in CS with AS or MR. There were two major findings. First, in CS with AS, the effect of increased VA-ECMO blood flow on LV performance parameters was either neutral (LVEDV, LVESV, LVSV, LVEF) or even beneficial (LVEDP decrease). Second, in CS with MR, increased EBF and MAP did not affect LV volume variables (i.e., LVEDV, LVESV, LVSV, LVEF), but increased LVEDP. Therefore, AS prevented LV overload and partial protection was observed also with MR, at the cost of higher LVEDP.
It has been repeatedly demonstrated—both in animal and human studies—that higher VA-ECMO blood flow may increase LV afterload, with a negative effect on LV performance, especially if LV function is already severely depressed [7, 8, 10, 12–14, 21, 22]. Therefore, in CS caused by severe LV dysfunction, VA-ECMO circulatory support may induce LV overload and pulmonary oedema, which represents a critical situation requiring urgent attention and therapeutic intervention including LV unloading and decompression [15, 16]. There are numerous factors that can influence or modulate the development of LV overload, including severity of LV dysfunction, VA-ECMO blood flow rate, use of vasopressors, inotropes, mechanical ventilation setting, presence of right heart failure, and the competence of the heart valves. Valvular heart disease is often present in patients with CS, either with causal relationship or as a concomitant and modulating disease. Severe AS can be a cause of CS, especially if decompensated with LV systolic dysfunction, or contribute to shock in case of acute myocardial injury of other etiology (e.g., myocardial infarction or myocarditis). Additionally, acute MR can clinically manifest as CS, and severe MR is often present as concomitant disease as a result of LV dilatation in advanced heart failure [15–17].
Increased VA-ECMO blood flow in CS with severe LV dysfunction and competent valves predominantly increases LVESV and reduces LVSV and LVEF; this phenomenon is largely explained by increased LV afterload at higher EBF rates [7, 14, 21]. However, if severe AS is present, LV afterload is determined almost entirely by obstruction of the aortic valve and increased VA-ECMO blood flow does not translate into an elevaton in LV afterload. This could also be a reason why we did not observe any changes in LV volumes with increasing EBF during the simulation of AS. Moreover, decreased preload caused by increased blood drainage from the right atrium at higher EBF can be responsible for reduced LVEDP in our study. Therefore, our data indicate that, in CS with severe AS, VA-ECMO restores systemic circulation and, at the same time, unloades the left ventricle, prevents LV overload, and reduces the risk for pulmonary oedema. In clinical practice, VA-ECMO is frequently used as circulatory support in CS with AS [23–27], and our data support this approach from the haemodynamic perspective.
The situation is different in CS with severe MR. An incompetent mitral valve enables translation of the effect of increased LV afterload at higher EBF to the left atrium. We did not record regurgitation volumes; however, we speculate that increased LV afterload at higher EBF resulted in increased regurgitation volume, which would fully explain our observation of unchanged LVESV and LVSV together with increased LVEDP. Importantly, higher LVEDP is associated with an increased risk for subsequent pulmonary oedema. Thus, preservation of LV function (LV volumes and LVEF) in CS with MR does not necessarily lead to stable haemodynamic conditions but more likely increased congestion. This effect could be even greater in predominantly LV failure where preserved RV would contribute to increased LV preload, hence increased LVEPD. Additionally, atrial septostomy (transseptal puncture), that we had to performe when introducing MR model, may have aleviated pulmonary congestion to some extent by venting left atrium [28]. Our results, therefore, support the recommendation to not use VA-ECMO alone in MR but rather in combination with other devices such as the Impella system [17].
Our study had several limitations. We used a model of CS caused by global hypoxia that affects not only the left but also the right ventricle, causing severe biventricular systolic dysfunction. Therefore, our model differs from other large animal models of acute heart failure, which are primarily based on the development of myocardial infarction by coronary artery occlusion; however, it reflects frequent clinical scenarios. Furthermore, LV ejections in our model supply with the hypoxemic blood not only coronary arteries but also carotid arteries at least during the initial phases of the hypoxic period. Hypoxic brain damage could therefore be anticipated. We cannot exclude that cerebral hypoxia may have influenced some of the mechanisms of central regulation of blood circulation. Moreover, we focused on the acute effects of VA-ECMO on haemodynamic and LV performance variables. We speculate that, especially in severe MR, long-term use of VA-ECMO may result in an increase in LV volumes and LV dilatation. Finally, our experimental study was conducted in young and otherwise healthy animals. Therefore, caution is advised in translating our results to clinical scenarios involving patients with advanced heart failure, LV remodelling, and comorbidities.