Share this article on:

Concerns Remain After FDA Meeting on Neurotoxic Effects of Pediatric Anesthesia

Hurley, Dan

doi: 10.1097/01.NT.0000278753.04525.11
Back to Top | Article Outline


  • ✓ An FDA meeting to assess whether anesthesia in the first few months of life might cause permanent cognitive deficits in humans did little to quell the ongoing debate on this issue between anesthesiologists and neuroscientists.

Four years after publication of a rat study sparked a heated debate between anesthesiologists and neuroscientists over whether anesthesia in the first few months of life might cause permanent cognitive deficits in humans, a meeting called by the FDA to assess the growing evidence in animal models — including the first primate study — did little to quell the dispute.

Prominent anesthesiologists said after the Mar. 29 meeting in Rockville, MD, that pediatric neurologists need not concern themselves with the issue, pointing out that the evidence in humans is entirely lacking and that they have no choice but to administer anesthesia when medically necessary.



But neuroscientists involved in the research, as well as a pediatric neurologist familiar with it, said the growing number of rodent studies, combined with the primate study, provide compelling evidence of a serious risk in humans that should be of concern to neurologists.

Coming down smack in the middle, an FDA medical officer in the Division of Anesthesia, Analgesia, and Rheumatology Products said during the meeting that the clinical relevance remains uncertain.

“A safety signal has been identified in animals for many drugs used to provide sedation and anesthesia,” said Arthur F. Simone, MD, PhD, during his presentation. “This database is growing. The relevance of the animal findings to pediatric patients is unknown.” That view echoed the conclusions he and two other FDA scientists reached in an overview of the evidence published in the March issue of A nesthesia and Analgesia (104:509–20).

After hearing the presentations, the FDA Anesthetic and Life Support Drugs Advisory Committee voted unanimously to conclude that there is insufficient evidence to recommend some agents over others.

But in an interview afterward with Neurology Today, the acting chair of the committee seemed to contradict that conclusion, even while emphasizing that the evidence is preliminary.

“The data suggest that ketamine is safe at the concentrations used clinically, and that the inhaled anesthetics are a risk if used for a long time,” said Steven L. Shafer, MD, professor of anesthesia at Stanford University Medical Center and editor-in-chief of Anesthesia and Analgesia. “I'm more concerned about the inhaled anesthetics.”

Despite those concerns, he added, “We do not have evidence of human injury, and there is no syndrome associated with children who receive anesthesia, as there is for alcohol. If my child were having anesthesia, I would pay no attention to these findings at all. Anesthesia in children involves multiple risks that are well understood and serious. I would hate to see those risks not given full weight because of concerns about a risk that is strictly hypothetical.”

Asked whether neurologists should be concerned about the data, he said, “This is not something that should be on their radar screen, except as a research topic.”

Back to Top | Article Outline


Researchers who first discovered apoptosis in the brains of neonatal rats given anesthesia disagreed.

“Pediatric neurologists should be and are concerned,” said John W. Olney, MD, the John P. Feighner Professor of Psychiatry at Washington University in St. Louis, whose research with rats first uncovered apoptotic neurodegeneration in the developing brain after administration of an NMDA (N-methyl-D-aspartate) antagonist (Science 1999;283:70–74). Later papers by him and others established that GABA (gamma-amino butyric acid) agonists, the other type of general anesthetic, caused the same kind of damage in rodents.



Dr. Olney said neurologists consulted about the effects of surgery in the first few months of life should advise delaying elective surgery when possible. “If it doesn't have to be done at the age of two weeks, but can be done at the age of five years, it should be postponed.”

Dr. Shafer disagreed. “Because anesthetics aren't safe intrinsically, no procedure in a child is performed electively,” he said. “It has always been the case that if a surgery could be delayed, it is delayed.” He noted that after the article by Dr. Simone and two other FDA scientists was published in Anesthesia and Analgesia, “We received multiple inquiries from concerned parents asking what this means for their children who were going to have surgery. The answer we gave to them is that their physicians would not have recommended surgery if they didn't believe it was necessary.”

But an anesthesiologist who co-authored several of Dr. Olney's papers, and who is also a member of the FDA's advisory committee on anesthetics, said that some surgery performed in young children can be safely delayed.

“We do nose jobs and hernia repairs and tonsillectomies,” said Vesna Jevtovic-Todorovic, MD, PhD, the Harold Carron Professor of Anesthesiology and Neuroscience at the University of Virginia. “The first two months of life are critical for the developing brain. So when people call me about a surgery they're considering, I say to them, Can you wait?”

A study she and Dr. Olney published in 2003, linking the morphologic damage caused by the anesthetic agents in rats to learning and memory deficits (J Neurosci 23:876–82), set off a firestorm of criticism from other anesthesiologists. Newborn rats exposed to a standard cocktail of anesthetic agents used in pediatric surgery — midazolam, nitrous oxide, and isoflurane — performed significantly worse on maze tests than did non-exposed rats, both at age one month and in mature animals, and hippocampal slices showed less long-term potentiation.

“When that study came out, I was absolutely crucified,” Dr. Jevtovic-Todorovic said. “Now everybody believes that in animals it happens. The question remaining is how do we make the translational jump, to see what is going on in humans.”

While not a jump, a major step forward was described at the FDA meeting by an agency scientist who led a study of rhesus monkeys exposed to ketamine in utero, at five days after birth or 35 days after birth. The earlier developmental stages, at a gestational age of 122 days or five days after birth, proved to be more sensitive to ketamine-induced neurotoxicity than at 35 days, reported William Slikker, Jr., PhD, director of the Division of Neurotoxicity at the FDA's National Center for Toxicological Research. A three-hour infusion of ketamine produced no neuronal cell death, whereas a 24-hour infusion did.



The monkey study provoked nearly contradictory responses from researchers who attended the meeting. Dr. Shafer, as noted above, concluded that ketamine at the doses used in humans should be safe. And according to Dr. Soriano, “We rarely use ketamine for as long as 24 hours. Even three hours of ketamine is rarely done.”

But both Dr. Olney and Dr. Jevtovic-Todorovic described the monkey study as an important confirmation of their results in rodents, increasing the likelihood that they will also hold true for humans.

“The only way to sort this out is to do experiments with primates at a gestational age equivalent to newborn humans and to see whether they also demonstrate this effect,” Dr. Rothman said.

Steven M. Rothman, MD, the Ernest and Jane G. Stein Professor of Developmental Neurology at Washington University, agreed that neurologists should be concerned about the emerging findings, adding that pediatric neurologists should ask about neonatal exposure to anesthesia when taking a history of children with developmental delays.

“Neurologists commonly see kids who were ill as newborns and are delayed in development at two or three years of age,” he said. “These kids can be quite complicated to figure out, because they've had multiple stresses. Now there is the additional confound, that prolonged anesthesia and sedation, especially in the case of very premature infants, could be an additional risk factor.”

Back to Top | Article Outline


Even assuming an effect in humans can be found, how big that effect will be remains uncertain, Dr. Olney conceded. “My guess is that for some of the more prolonged anesthetic exposures, say in complex cardiac procedures that can take eight hours or more, there may be a considerable risk of substantial neurodegeneration,” he said. “Under more mild circumstances, my guess would be that some neuronal dropout happens but that it would be so mild that it taxes our testing methods to really pick it up.”

Despite the uncertainties, researchers are moving quickly to find either novel anesthetics that do not cause apoptosis, or neuroprotective agents that prevent damage from existing anesthetics. Dr. Jevtovic-Todorovic published a paper last year showing that melatonin protects against apoptosis (Neurobiol Dis 2006; 21(3):522–530), and a paper published this spring reported that the gas xenon has a similarly protective effect (Anesthesiol 106(4):746–53). Dr. Olney, however, said that he will be reporting later this year that xenon causes apoptosis through another means.

Even so, Dr. Olney expressed optimism that new agents will soon be found that can provide broad-spectrum protection without any neurotoxic effects. “We are pursuing some very promising leads,” he said, “and will be submitting our findings for publication or presentation at scientific meetings in the near future.”

In the monkey study presented at the FDA meeting, Dr. Slikker reported that co-administration of an NR1 antisense oligonucleotide blocked the ketamine-induced neuronal cell death.

In the meanwhile, Dr. Jevtovic-Todorovic said, children exposed to anesthesia in the first few months of life might benefit from special care. “When you look at the animal studies that we and others have published,” she said, “the cognitive deficits we find would result in slower learning. You cannot make a direct extrapolation, but if you could, you'd say that a child in first grade might take, say, twice as long to learn a task.”

Back to Top | Article Outline


At the meeting, the advisory committee also heard presentations arguing that the evidence for cognitive deficits following anesthesia is far stronger for elders than it is for infants. Zhongcong Xie, MD, PhD, of the Genetics and Aging Research Unit at the Mass General Institute for Neurodegenerative Disease, reviewed his studies showing that isoflurane not only induces apoptosis but also increases the amyloid load in the brain. Gregory Crosby, MD, an anesthesiologist and laboratory investigator at Harvard Medical School, pointed out, “Elders are the only age group known to suffer persistent cognitive dysfunction postoperatively.”

Dr. Olney said that the human data in elders is worrisome and merits further study. But, he said, the neuronal damage to infants, if confirmed in human studies, would result in lifetime deficits, compared to the briefer effects of any deficits caused late in life.

All of the presentations at the FDA meeting can be read online at

© 2007 American Academy of Neurology