In Reply:—
In their letter to the editor, Soriano et al. state that in our recent counterpoint editorial,1we “mainly questioned [their] assertion that repeated large doses of ketamine were responsible for mediating the neurodegenerative changes noted in neonatal rat brains.” This is not what we questioned; we know that repeated large doses of ketamine can cause extensive neuroapoptosis. What we mainly questioned was whether repeated large doses are necessary, or whether neuroapoptosis can be triggered by a single low dose of ketamine. We then presented evidence that a single subcutaneous dose of ketamine (20, 30, or 40 mg/kg)—one that does not fully immobilize, anesthetize, or abolish pain responses in infant mice—does trigger a significant increase in the rate of neuroapoptosis. The scientifically appropriate response would be for Soriano et al. to administer these single subanesthetic doses of ketamine to infant mice and, using the same methods we used, to see whether they could reproduce our findings. Instead, Soriano et al. imply that it requires “repeated large doses” to trigger neuroapoptosis and argue that ketamine is safe for pediatric anesthesia because such large “doses and durations . . . are never used in pediatric anesthesia.”
To bolster their claim that a single dose is ineffective, they cite a recent report of Scallet et al. 2in which a single dose of ketamine at 20 mg/kg did not trigger apoptosis in infant rats, although repeated 20-mg/kg doses did. Because, in our single-dose experiments, 20 mg/kg was the threshold dose for triggering apoptosis, it is not surprising or very meaningful that one laboratory would report a barely significant effect and another would report a barely insignificant effect at this dose. What is surprising is that instead of directly acknowledging and discussing the implications of our finding that a single dose of ketamine at 20, 30, or 40 mg/kg does trigger apoptosis in a dose-dependent manner, Soriano et al. continue to promote their original position that ketamine is safe because, in their hands, a single dose as high as 75 mg/kg does not trigger neuroapoptosis. It is difficult to reconcile this position with their introductory statement that “there is no doubt regarding the scientific validity” of our findings.
Soriano et al. suggest that we should have measured blood ketamine concentrations in our mouse experiments. However, Soriano et al. did not measure ketamine blood concentrations in their rodent experiments, and anesthesiologists do not routinely measure, much less rely on, ketamine blood concentrations to determine depth of anesthesia. We reported that a single dose of ketamine, in the range of 20–40 mg/kg, that does not fully immobilize, anesthetize, or render an infant mouse insentient to pain does trigger neuroapoptosis in the infant mouse brain. This is a message that is not difficult to understand. Presumably, we can all agree that regardless of ketamine blood concentrations, it would be unacceptable to perform surgery on an infant mouse, or infant human, whose depth of anesthesia is such that the infant is squirming around, flailing the extremities, and responding to skin pinch by vigorous antalgic movements.
Soriano et al. continue to argue that the neuroapoptosis response to anesthetic drugs is due to hypoxia/ischemia. How is this possible in light of our demonstration1that arterial oxygen saturation remains in the 97–99% range over a 4-h period after a dose of ketamine that triggers neuroapoptosis within this same time interval? Soriano et al. postulate that the oxygen saturation fleetingly decreased to brain-damaging levels during intervals between our sampling time points but abruptly resumed normal levels at each time point (15, 30, 60, 120, 180, 240 min) just before we drew blood. We doubt that the readership of Anesthesiology will be persuaded by this argument, especially because we are talking about a subanesthetic dose of ketamine, a drug that reputedly, even at anesthetic doses, does not compromise cardiorespiratory function.
Even if severe hypoxia/ischemia did occur, it could not account for the neuroapoptosis response to ketamine because 4–6 h after ketamine administration, an increase in apoptotic profiles is evident both as a caspase-3 activation response and as ultrastructurally confirmed apoptotic morphology. However, when one intentionally induces hypoxia/ischemia and examines the developing brain 4–6 h later, there is no increase in apoptotic profiles, either by caspase-3 activation or ultrastructural criteria. It is illogical to argue that anesthesia-induced apoptosis is caused by hypoxia/ischemia if one cannot demonstrate that intentionally induced hypoxia/ischemia reproduces the anesthesia-induced apoptosis phenomenon. What one does find in the brain 4–6 h after hypoxia/ischemia, as we have demonstrated previously,3and also very recently,4is excitotoxic neurodegeneration. (See Young et al. 4for a detailed presentation of evidence directly addressing and clarifying this issue.) Soriano et al. challenge our position by citing works from other laboratories that they believe contradict our observations. We have examined the cited works, some of which are in vitro studies, and find that these works either support our position or are irrelevant to the issue in contention. We stand by our own observations, which are based on a three-decade-long direct investigation of the specific issue in contention: in vivo excitotoxic versus apoptotic neurodegeneration in the developing brain.3–10
Regarding the nutritional deprivation issue, we stated1that in a “typical” experiment, we expose infant rodents to a single dose of saline or an apoptogenic anesthetic drug and, without returning the pups to the maternal nest, kill them 4–8 h later. Because both the control and experimental pups are exposed to the same degree of maternal/nutritional deprivation during this 4- to 8-h period, nutritional deprivation cannot explain the higher rate of neuroapoptosis in the experimental pups. In an apparent effort to refute this interpretation, Soriano et al. note that in one study we killed animals not only at 4 and 8 h but also at 12, 16, 24, and 48 h, and determined, using a staining procedure that detects cumulative neuronal degeneration, that apoptosis became increasingly more prominent at 12, 16, and 24 h. We do not understand how this reference to our earlier comprehensive evaluation of the apoptotic response to large doses of MK80111refutes our current interpretations pertaining to “typical” experiments focusing on the very early response to low subanesthetic doses of ketamine.
Soriano et al. point out that our most recent findings1pertaining to threshold conditions for inducing developmental neuroapoptosis were conducted in mice and suggest that species differences between rats and mice and between rodents and humans may be of paramount importance. We have tested rats and mice and find no appreciable differences between these species, but we agree that differences between rodents and primates may be very important. Of course, species differences can go in either direction—humans may be less vulnerable or they may be more vulnerable than rodents.
Soriano et al. conclude that only human studies can provide the final answer. We do not contest the importance of human experiments, but such experiments will require many years to complete and, because of design limitations, may provide equivocal results that defy interpretation. Therefore, we recommend that the issue be addressed in nonhuman primate studies designed to test the sensitivity of the primate brain to anesthesia-induced developmental neuroapoptosis. If the primate brain proves sensitive, additional studies designed to evaluate the neurobehavioral consequences of graded degrees and controlled patterns of apoptotic neurodegeneration in the developing nonhuman primate brain would be informative. Such studies would provide the anesthesiology community with reliable information and guidance in the conduct of much-needed human studies.
* Washington University School of Medicine, St. Louis, Missouri. olneyj@wustl.edu
