Mark Wefers Bettink

Monitoring of mitochondrial oxygen tension in the operating theatre 8 159 capillary-venous oxygen saturation did not decline. A similar phenomenon of decreasing mitoPO 2 levels while capillary-venous oxygen saturation remained unchanged has been described earlier by Ubbink et al. 5 They observed a drop in mitoPO 2 after a decline in microvascular blood flow by a bolus of clonidine, but no changes in capillary-venous oxygen saturation were measured. An alternative explanation divergence between mitoPO2 and microcirculation could be the different wavelengths used by the three different measurement techniques. StO 2 is measured with the spectra absorbance of visible light, mitoPO 2 is measured with green light (515 nm) and velocity and flow are measured with the hemoglobin laser Doppler frequency shift red light. All these different wavelengths give different tissue penetrations 24 and thus measure in different tissue compartments and could therefore show different results. Anyhow, based on these findings it is clear that measuring oxygen availability directly at cellular level provides complementary data and new insight. In the two cases we described the oxygen transport to the cells is at risk. In the first case major blood loss was followed by normovolemic hemodilution and a drop in haemoglobin levels occurred. This normovolemic hemodilution was accompanied by a slow decrease of the mitoPO 2 to values ​below 10 mmHg. Striking is that especially mitoPO 2 decreased while other usual parameters, such as blood pressure, heart rate and pulse variation index did not change. Hemoglobin levels and SpHb were reduced but not to values indicating a direct need for blood transfusion, and even serum lactate levels remained low in the intraoperative phase. This advocates the decision not to use blood transfusion despite the large amount of blood loss. However, in the postoperative phase lactate levels rose and the patient did need a blood transfusion. We have previously observed more or less the same behavior of mitoPO 2 in a single case during major abdominal surgery 25 . These cases suggest that mitoPO 2 monitoring can be of great value for improving hemodynamic management. In the second case presented in this study, a perioperative phase with controlled hypovolemic circumstances has been presented. The hypovolemic circumstance resulted in low central venous pressure (CVP) values and low mitoPO 2 values between 30 and 20 mmHg, whereas most other parameters remained stable. Only lactate levels had risen, indicating tissue hypoxia. These examples show that mitoPO 2 can clearly respond to hemodynamically unstable situations. Furthermore, it shows that during hemodynamically unstable conditions mitoPO 2 values that are really below the normal standard deviation ​of less than 10 mmHg have been measured. In the second case, in which hypovolemic conditions have been pursued for a longer period of time, also systemic signals of hypoxia were intraoperatively observed. In the first case,