Hyperoxemia and Veno-Arterial Extracorporeal Membrane Oxygenation: Do Not Forget the Gut
In our opinion, caution is required in front of the apparent safety of hyperoxemia during veno-arterial ECMO for cardiogenic shock.
Information regarding the sampling site would be useful because of the risk of differential oxygenation. It is likely that patients were monitored with a right radial arterial catheter. However, in their article (1), the authors do not indicate the location of the arterial blood sample.
Patients treated for cardiogenic shock differ from eCPR patients by a residual stroke volume. Because of that, it is likely that right radial arterial PaO2 may reflect oxygenation of the upper part of the body, determined by lung function, and native cardiac outflow. However, this do not indicate the oxygenation of the lower part of the body, which is determined by the oxygenator performance, and the ECMO outflow, when cannulas are inserted at the femoral site (2). In most of the ECMO settings, current guidelines recommend the use of 100% oxygen sweep gas. So, in the setting of cardiogenic shock and cardiac arrest, the PaO2 measured peripherally might underestimate the oxygen burden of the lower part of the body. In fact, very hyperoxemic blood might irrigate the hepatosplanchnic territory.
In this population, hyperoxemic injury may worsen ischemic small bowel damage and favors the occurrence of multiple organ failure. We previously showed that shock, by a mechanism of nonocclusive mesenteric ischemia, is associated with an increase of “Intestinal Fatty-Acid Binding Protein” (iFABP) plasma level. In addition, we showed that this very specific marker of enterocyte damage was independently associated with mortality, reinforcing the old hypothesis that the gut plays a key role in the occurrence of multiple organ failure (3). Although there is a lack of clinical data concerning the impact of hyperoxemia on small bowel injury, some preclinical studies have shown that hyperoxemia might be deleterious for the gut. In a recent study, newborn rats exposed to normobaric hyperoxia were found to have more bacterial translocation and higher iFABP level than rats of the normoxic group (4). In the oxygen-ICU randomized trial, there were a higher number of shock and liver failure in the upper arm oxygen compared with the lower arm (5). Those data support the hypothesis of the hyperoxemic injury of the small bowel favoring multiple organ failure. Hypoperfusion and hyperoxemia-induced oxidative stress might have synergetic effects leading to gut damage.
To conclude, among cardiogenic shock and cardiac arrest patients on veno-arterial ECMO, an association between hyperoxemia and mortality might have been found if PaO2 had been measured at another arterial site, reflecting the oxygenation level of the hepatosplanchnic territory.