As a lifesaving treatment, VV ECMO is becoming more and more essential for children and adults with refractory respiratory failure, and such patients have a high prevalence of cardiac dysfunction [11, 12]. VV ECMO does not provide direct cardiac support, which depends on intrinsic cardiac function to maintain cardiac output. Therefore, the condition of cardiac function during VV ECMO is closely associated with the outcome. We present the first detailed laboratory investigation in the effect of VV ECMO on the heart in nonneonatal population.
In this study, the volume of fluid intake was significantly more in the ECMO group. Golej et al.  found MAP decreased at the initiation of VV ECMO and increased in a short time. In consideration of the numerous influence factors, we haven`t analyzed the immediate hemodynamic response. But we have observed similar MAP fluctuation at the initiation of VV ECMO and restored quickly after fluid bolus, which contributed a lot to the amount of the fluid taken in the ECMO group, and may associated with myocardium interstitial edema.
In addition, the ECMO group disorderd and dissolved of focal myofilament, morphological deformations of mitochondria and decreased activities of mitochondrial complexes compared with the control group, suggesting that VV ECMO therapy is associated with myocardium injury.
The mechanisms of these findings require further research. Several factors may contribute to it, such as activation of systemic inflammatory response, increase of reactice oxygen species (ROS), and release of toxic substances by ECMO.
The instigation of a systemic inflammatory state with exposure to ECMO is well accepted [17–21]. Fortenberry et al.  found that neutrophil was activated and concentration of circulating interleukin (IL)-8 increased significantly after ECMO initiation. Adrian et al.  perfused blood in vitro ECMO circuit performed for 24 hours, and found cytokines, including IL-1β, IL-1 receptor antagonist, IL-8, IL-6, tumor necrosis factor α (TNF-α) were increased. Mu et al.  found concentrations of circulating IL-6 and IL-10 increased significantly in a hemorrhage-reperfusion piglet model on modern ECMO. The inflammatory cytokines have been implicated in mediating myocardial injury [22–24]. Deng et al.  and Hennein et al.  found that preoperative left ventricular dysfunction is associated with the degree of proinflammatory cytokine release. Liakopoulos et al.  found that cardiac dysfunction occurred after cardiopulmonary bypass was associated with the increase of systemic and myocardial TNF-α, and anti-inflammatory pretreatment with methylprednisolone abolished the increase of TNF-α attenuated myocardial dysfunction.
It is well established that cardiopulmonary bypass is associated with increased production of reactice oxygen species (ROS) [28, 29], and similar results were verified during ECMO therapy. Hirthler et al.  and Underwood et al.  detected that systemic free radical was increased during ECMO. Moller et al.  found free oxygen radical scavenging enzymes such as superoxide dismutase and glutathione reductase were decreased during ECMO in healthy lambs, resulted in increased lipoperoxide level. ROS are central mediators of cardiac injury, especially for mitochondria, and finally resulting in cardiac dysfunction . Precious researches indicate that lacking of ROS scavenging enzyme aggravate myocardial injury , and antioxidant potentiate cardiac protection .
In addition, it is not clear whether the ECMO circuit adds some toxic substances such as endotoxin  and plasticizers  impact the myocardium.
Despite changes of ultrastructure and function of cardiomyocyte and mitochondria induced by VV ECMO, we found hemodynamics during VV ECMO therapy was stable, suggesting that the injury was mild, and had no effect on the cardiac performance for healthy piglets, consistently with the results of Strieper  and Roberts .
However, energy is the base to maintain normal cardiac pump function, continuously produced by mitochondrial respiration . Therefore, any alteration of mitochondrial structure and function are fundamental for cardiac function. The changes of mitochondria in this study do not impact the cardiac performance, but may impact the cardiac reserved function, which may essential for a marginally functional heart or when cardiac demand increased under disease conditions.
Furthermore, under disease conditions with severe hypoxia, hypoxia itself plays an important role in myocardial injury , and the advantage of VV ECMO in providing adequate oxygenation to the myocardium may be of protective effect. Shen et al.  found cardiac dysfunction due to hypoxia coronary perfusion during venoarterial ECMO in a hypoxemic swine model, postulating that VV ECMO may be more adequate for hypoxemic condition. However, abrupt hyperoxia for a hypoxemic heart may induce reoxygenation injury, causing further cardiac injury. Allen et al.  and Trittenwein et al.  found increased amounts of oxygen free radicals after reoxygenation on cardiopulmonary bypass and ECMO respectively.
Therefore, the effect of VV ECMO on the myocardium and cardiac performance would be more complicated for patients under hypoxemic condition, and should be further investigated. However, we present the potentially adverse effect of VV ECMO alone on the heart for the first time, calling for attention to the impact on heart during VV ECMO, and provide a potentially aspect to improve the survival of VV ECMO.