In Reply:--Our computer simulations showed that the relation between cerebral venous oxygen saturation (SVO2) and CMROsub 2 depends on temperature. Hypothermia increases hemoglobin's oxygen affinity. During profoundly hypothermic cardiopulmonary bypass (CPB), high SVO2can be the result of impaired oxygen transfer from hemoglobin to brain. SVO2may, under some circumstances, be an accurate monitor of CMRO2. However, our results showed that validation of near-infrared and internal jugular measurements of SVO2during normothermia does not imply that they have been validated under hypothermic conditions. We successfully used a computer simulation of a mathematical model to do a technology assessment. The model warned clinical investigators about potential misinterpretation of near-infrared and internal jugular measurements of SVO2.
This model prediction has since been verified clinically. Du Plessis et al. simultaneously measured cerebral mitochondrial and cerebral hemoglobin saturations in children undergoing profoundly hypothermic CPB. During cooling and initiation of low-flow CPB, SVOsub 2 increased. In contrast, cerebral cytochrome saturation decreased. Therefore, oxygen availability at the mitochondrial level was diminished, despite high SVO2. Together, our studies and those of Du Plessis et al. show that SVO2can be a poor monitor of the adequacy of cerebral oxygenation during profoundly hypothermic CPB.
Gunter correctly points out that choice of certain of the model parameters can affect the quantitative predictions of the model. However, we emphasize that choice of the parameters does not affect the usefulness of the model. Choices of these parameter values do not change whether temperature affects the relation between SVO2and CMR sub O2. Changes in parameters' values force the relation between SV sub O2and CMRO2to change for all temperatures. We quote from our discussion section, with the italic being in the original paper :“It is important to appreciate predicted SVO2at given levels of CMRO2should be considered as rough estimates only …[For example,] actual SVO2values required to preserve CMR sub O2may be higher than those given for profoundly hypothermic CPB. However, determination of these estimates was not the goal of the study. Instead, our aim was to determine the effect of hypothermia on the relationship between CMRO2and SVO2.” We addressed whether conventional interpretation of SVO2data, obtained either from jugular venous catheters or near-infrared spectroscopy, reliably indicates the adequacy of cerebral oxygenation during hypothermic CPB. Our assessment of these SVO2-measuring technologies clearly shows that they may not.
We agree with Gunter's observation that dissolved oxygen provides almost all cerebral oxygen requirements during profoundly hypothermic CPB. We also made this potentially important observation by examining results of the mathematical model. We then contacted Drs. Frank Kern and William Greeley at Duke University, who have measured SV sub O2, CMRO2, and cerebral blood flow in hundreds of infants and children. We are collaborating with them to analyze statistically their clinical data to test this model prediction.
However, these observations regarding dissolved oxygen probably do not account for the failure of Diaspirin cross-linked hemoglobin to increase oxygen utilization during hypothermic CPB in rabbits. The effect of dissolved oxygen becomes significant at profoundly hypothermic temperatures (i.e., less than 20 degrees Celsius). At warmer temperatures, the brain has an extraordinary ability to extract oxygen. We did our Diaspirin experiments during CPB at 27 degrees Celsius. Diaspirin probably failed to increase CMR sub O2because cerebral oxygen delivery was not limiting oxygen utilization at this temperature.
A perfluorocarbon emulsion would increase oxygen availability. However, we do not know whether the brain needs any more oxygen during CPB. We are not aware of evidence that, during CPB, the brain has inadequate cerebral oxygen delivery, even during low-flow conditions. Even then, the brain may be able to sufficiently increase extraction of oxygen from hemoglobin to maintain a normal CMRO2. We may not be able to use the mathematical model to study this question reliably. The result would probably depend on our choice of physiologic parameters. As Gunter pointed out, there is a limitation to the analyses that we can do using the mathematical model.
Franklin Dexter, M.D. Ph.D., Assistant Professor, Bradley J. Hindman, M.D., Associate Professor, Department of Anesthesia, University of Iowa, Iowa City, Iowa 52242.
(Accepted for publication July 1, 1996.)