In Reply:—
We read with considerable interest the critical commentary of Kopp with respect to our editorial in the April issue of Anesthesiology.1That editorial presented a brief introduction to the research articles that were presented at the Anesthesiology/Foundation for Anesthesia Education and Research Symposium on Anesthetics and the Developing Brain; the intent was to summarize current research in anesthetic neurotoxicity with an emphasis on the molecular mechanisms that underlie the adverse impact of anesthetics. The central concern expressed by Kopp is the potential toxicity of oxygen. Given that oxygen administration is a routine practice in the clinical practice of anesthesia, it is highly probable that a state of hyperoxemia is induced in subjects undergoing anesthesia and surgery. Kopp suggests that it is this hyperoxemia that can injure the brain, and in particular the developing brain.
There is growing evidence that the administration of oxygen in concentrations that produce hyperoxemia is associated cellular injury. The adverse impact of high concentrations of oxygen on retinopathy of prematurity2and on bronchopulmonary dysplasia3has long been recognized. In susceptible neonates, the incidence of cerebral palsy is increased in association with hyperoxemia.4More recent evidence also indicates that resuscitation of premature neonates with a high fraction of inspired oxygen (Fio2) is associated with greater mortality and worse outcomes.5Indeed, the authors of a recent metaanalysis concluded that the available data support the use of room air for resuscitation of asphyxiated neonates in place of 100% oxygen.5Importantly, the use of room air for this purpose does not seem to be associated with worse cognitive outcomes.6Preclinical studies in adult animals also suggest that resuscitation from global ischemia with high Fio2leads to greater neurologic injury.7
In the investigations of Kalkman et al. 8and Wilder et al. ,9the concentration of oxygen that was administered is not clear. It is reasonable to assume, based on the current standard of practice, that supplemental oxygen was administered and some degree of hyperoxemia did occur. Could the association between anesthetic exposure and adverse outcomes be explained by oxygen toxicity rather than anesthetics? Although Kopp's contention is feasible, it is difficult to separate the effects of oxygen from those of the patients' primary disease, anesthetics, surgery, postsurgical inflammation, and use of analgesics. The question of whether oxygen can injure the otherwise normal developing brain is best answered in the laboratory.
Of significant interest are the observations of Felderhoff-Mueser et al. ,10who demonstrated oxygen toxicity in the developing brain. An inspired concentration of oxygen of 80% resulted in widespread neurodegeneration; toxicity was apparent with as little as 2 h of exposure. The pattern of injury was similar to that produced by anesthetics. Moreover, the period of vulnerability, as with anesthetics, was approximately postnatal day 7, with little injury seen at postnatal day 14. By contrast, injury was not observed with the administration of 40% oxygen for as long as 12 h. This begs the question of whether anesthetic toxicity observed in previously published studies might be due to oxygen.
In published studies to date, the reported inspired concentrations of oxygen were 30%,1150%,12,13and 21%.14,15The duration of exposure ranged from 4 to 6 h. In these studies, injury produced with anesthesia was significantly greater than that in control nonanesthetized animals. With the exception of the studies of Stratmann et al. ,12,13the concentration of oxygen used was well below the level that has been shown to produce injury to the developing brain. Furthermore, the duration of exposure is well below the 12-h exposure to 40% oxygen in the study of Felderhoff-Mueser et al. 10in which injury was not observed. The available data indicate, therefore, that in experimental models, the toxicity produced by anesthetic exposure is not due to oxygen administration but due to anesthetics.
There is a remote possibility that there might be a relative increase in brain tissue partial pressure of oxygen (Po2) during anesthesia, even with the administration of air. Anesthetics decrease the cerebral metabolic rate for oxygen substantially and, depending on the inspired concentration of inhaled agents, cerebral blood flow may increase. Whether this relative increase in tissue Po2is detrimental in the developing brain is not clear. However, it is not outside of the realm of possibility that relative tissue hyperoxia might reduce the antioxidant defenses of neurons16and thereby make them more vulnerable to anesthetic neurotoxicity. This question will have to be addressed experimentally. We therefore invite Dr. Kopp to join us in our efforts to more definitely characterize anesthetic (and oxygen) toxicity in the developing brain and to develop the means and practices by which this toxicity can be prevented. This would, to paraphrase Kopp, allow us to bring more balance to the discussion.
*VA San Diego Healthcare System, San Diego, California. ppatel@ucsd.edu