This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by McGowan, F. X. Articles by Davis, P. J. Search for Related Content PubMed PubMed Citation Articles by McGowan, F. X., Jr Articles by Davis, P. J.

---

EDITORIAL

Anesthetic-Related Neurotoxicity in the Developing Infant: Of Mice, Rats, Monkeys and, Possibly, Humans

Francis X. McGowan, Jr, MD^*, and Peter J. Davis, MD

From the *Division of Cardiac Anesthesia, Children’s Hospital Boston, Harvard Medical School, Boston, Massachusetts; and Children’s Hospital of Pittsburgh, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania.

Address correspondence to Francis X. McGowan, Jr., MD, Chief, Division of Cardiac Anesthesia, Children’s Hospital of Boston, 300 Longwood Avenue, Boston, MA 02115. Address e-mail to francis.mcgowan{at}childrens.harvard.edu.

This issue of Anesthesia & Analgesia contains seven varied articles that all relate in some way to the possibility of neuronal toxicity arising from the use of inhaled and certain IV anesthetics in the immature and developing animal.1–7 Although there is a fair amount of overlap in the content, the diligent reader should nevertheless be rewarded by a significantly improved understanding of the main aspects of this topic and the controversy that surrounds it.

The origins of this debate go back at least as far as 1999 when Dr. John Olney and colleagues reported in Science that N-methyl-d-aspartate (NMDA) antagonists, including MK801 and ketamine, produced evidence of neurotoxicity.8 Subsequently, his group and others demonstrated that many anesthetics, including nitrous oxide, isoflurane, and ketamine, appear to induce apoptotic changes in neurons, particularly in the brains of infant rodents (detailed in Refs. 1–4). In fact, this debate goes back at least to 1985, when Uemura et al. demonstrated diffuse effects of halothane on synaptic development and subsequent learning behavior in rats.9

Perhaps, these results are to be expected. Most neurotransmission in the developing brain is primarily due to NMDA- and/or -aminobutyric acid-dependent mechanisms, as is the regulation of neuronal developmental and synaptogenesis. There are at least two possible general mechanisms of pro-apoptotic and/or neurotic injury. The first is that receptor blockade by anesthetic drugs decreases "trophic" stimulation at a critical point in time, leading to induction of cell death programs (analogous to the effects of growth factor deprivation in a variety of cell types). The second is that the transient receptor blockade caused by anesthetic drugs results in subsequent receptor upregulation (protein expression and/or receptor activity). When the blocking anesthetic is removed, the affected neuron is subsequently subjected to significantly increased NMDA activity, resulting in increased excitatory amino acid neurotoxicity. The mechanisms remain largely unsolved, although the available evidence weakly suggests the second.

Regardless of the exact mechanisms, the available data raise the specter that standard clinical anesthesia practice could promote brain injury in the human fetus, infant, and young child. These reports taken together led the United States Food and Drug Administration (FDA) to convene an Advisory Committee Meeting in April of 2007 (meeting transcript available at http://www.fda.gov/ohrms/docets/ac/07/transcripts/2007/4285t1.pdf). The purpose of the FDA meeting was to evaluate the potential neurodegenerative effects of anesthetics in infant and juvenile animals. The FDA panel focused on the data presented in a nearly contemporaneous comprehensive review.10

We personally believe that the evidence demonstrating the potential for different types of anesthetics to stimulate proapoptotic and neurotoxic events in infant rats and other species appears to be reasonably good. However, as the reader will also see, the extrapolation of the animal data to humans, and human infants in particular, is very tenuous. In the Pro/Con presentations from Jevtovic-Todorovic, Olney, Loepke, and Soriano, as well as the subsequent summaries by Loepke and from the FDA Group of Wang and Slikker, the reader can review the data that were presented to the FDA in great detail, as well as some new information. Reading these as companion pieces, one is able to follow the development of the evidence as well as its substantial and continued limitations. The article by Sanders et al.5 provides preliminary data that the phenomenon of anesthetic-induced neurodegeneration appears to occur, not surprisingly, in the neonatal rat spinal cord.

It is clear that the available data merit concern and further investigation. That said, many of the limitations pointed out in the articles by Loepke et al. persist. These limitations pertain to the experimental model, the anesthetic dose and/or concentration, the duration of the exposure (both absolute and in comparison to the human), and the developmental age and stage of the animal at the time of exposure. In some cases, the experimental methodologies used to assess effect and outcome have also been less than optimal.

Normal brain development requires significant expansion and growth of specific neurons and synapses as well as the targeted deletion of others. These processes are under the control of genetic, environmental, cellular, humoral, neurotransmitter, and trophic factors. Neuronal migration and synaptogenesis takes place along glial guide paths, with later-developing neurons that are migrating to cortical layers traveling past earlier developing ones that remain located in deeper brain layers closer to the germinal matrix. Specific factors that regulate this process of neuronal migration and synaptogenesis are not known, but increasing evidence again supports multiple layers of control, including genetic, neurotransmitter, excitatory amino acid, and neural cellular adhesion molecules. Depending upon the brain region, as many as half or more of all of the synaptic connections that were formed during the first one to two years of life will be absent within a few years thereafter. Which synapses are maintained and which are deleted are again influenced by a number of factors, including genetics, excitatory amino acid neurotransmitters, neuronal activity produced by external stimuli, and various trophic factors.

It appears that developing neurons can be induced to activate pathways resulting in cell death when their synaptic development, dendritic branching, and remodeling activities are interfered with. The period of rapid synaptogenesis appears to be the most vulnerable time. The intensity and duration of neurodevelopmental activity vary widely among different species. For example, in the newborn rat this peak period appears to be relatively confined to about the first postnatal week or two, peaking sharply on postnatal day 7 ("P7" in rat pups). It seems much more difficult, although not impossible, to induce neuroapoptosis with anesthetics in the rat at later developmental stages. This postconceptual age in the infant rat aligns with late second to early third trimester neurodevelopmental stages in the human. Nonetheless, in rodents, this period of rapid and critical synaptogenesis and related events is accomplished rapidly, probably within a few days to a few weeks. In subhuman primates, and probably in humans, the process appears to occur more slowly, over several months or perhaps even years.

Along with expected differences in pharmacokinetics and pharmacodynamics among species, interspecies variations make experimental differences in dose, duration of exposure, and time of exposure, both critical and problematic in terms of extrapolation to humans. Even in the subhuman primate, arguably the closest situation to humans, the "window" to produce an anesthetic-induced neuroapoptotic effect appears to be relatively narrow compared with the period of synaptogenesis in humans.

Does the slower development in humans confer increased risk, because of the longer period of risk, or increased resistance, because a child would only be exposed to anesthetics for a brief fraction of the "at risk" period? It is hard to know. The broad window of vulnerability in the human may result in fewer vulnerable neurons or synapses at any given point in time, decreasing the potential for injury, and making such injury (if any) impossible to detect. This may explain why, despite the large number of annual pediatric anesthetics (estimated to be at least 4 million each year in the United States), a "phenotype" of brain injury in children exposed to anesthetics at a young age has not been described. On the other hand, a thorough and valid search for such a phenotype has not yet been conducted. There are certainly examples in anesthetic practice (e.g., toxicity due to intrathecal lidocaine) where safety has been assumed based on a long history apparently free from adverse events, only to subsequently find toxicity when it was rigorously investigated. It is also worth noting that infants and children do not undergo "elective" surgery. There are well documented adverse consequences of procedural delay (e.g., the former premature infant with inguinal hernias). Furthermore, procedural outcome data in a number of difference arenas (e.g., cardiac, orthopedic, craniofacial) suggest enhanced results (at least surgical, at times overall outcome) can be obtained from early repair compared with delayed repair.

The few studies that have attempted to address developmental problems arising from anesthesia and surgery have, by necessity, included infants with congenital or acquired abnormalities likely to have major impact upon neuroimaging results and neurobehavioral assessment (see Table 1 in the article by Loepke and Soriano3). For example, as many as 30%–40% of infants with significant congenital heart disease have structural or biochemical brain abnormalities before any intervention or surgery.11 A neonatal procedure as relatively straightforward as a balloon atrial septostomy in such patients may be associated with an 80% or more incidence of postprocedure brain abnormalities on magnetic resonance imaging.12 Genetic syndromes such as velo-facial-cardiac syndrome (microdeletion of chromosome 22) are associated with congenital abnormalities that require infant surgery and developmental delay.13 Unpublished data from one fetal care center suggest that brain abnormalities (observed by fetal ultrasound or magnetic resonance imaging) may be found in as many as 25% of infants with a variety of nonbrain anomalies (F.X. McGowan, personal communication) The cosegregation of nonanesthetic causes of neurocognitive deficit with structural anomalies requiring surgical correction during infancy may render impossible clinical studies to assess the additional risk of anesthetic exposure. And if such risk exists, how do we weight that risk against the substantial evidence of harm to the developing brain from painful procedures in the absence of adequate pain relief.1,3,14 Lastly, some animal studies suggest that simple interventions, such as treatment with melatonin, can attenuate the risk of neurologic damage from anesthetic exposure.15 Should we give infants melatonin before surgery? Is it rational to give a drug with unknown toxicity to infants to protect them from hypothetical anesthetic toxicity that has never been shown to occur in infants?

Based on these issues, what actions are clearly indicated at the present time?

1. We need additional animal studies to better define molecular mechanisms, risks, and potential treatments for this problem. Too much time and effort have been spent asking essentially the same question in different animal models. Defining specific mechanisms should be emphasized, in part, because they are likely to identify specific treatments.
   
2. Future studies must also be relevant to human development and practice. Some therapeutic possibilities are alluded to in the review in this issue by Degos et al.6 Unlike many forms of brain injury, this injury features a single inciting event, a circumscribed time frame, and the ability to intervene before, during, and after injury. The therapy need only be periprocedural (as opposed to long-term or lifelong), further enhancing therapeutic options. Drugs that have been given to humans without apparent toxicity that may have efficacy against underlying brain apoptotic mechanisms (e.g., melatonin, minocycline, lithium)15–17 could be particularly interesting if anesthetics are in fact shown to cause a similar injury in humans.
   
3. Increased research on mechanisms of anesthetic action are needed to discover drugs with greater specificity or mechanisms of action that do not involve -aminobutyric acid and NMDA receptors.
   
4. Human studies of adequate statistical power are needed to identify any evidence of injury from exposure to anesthetics, including in utero exposure. It will be extremely difficult to find sufficient numbers of human infants having the sufficient range of age and duration exposure in the absence of important exclusionary comorbidities. Pregnant women undergoing anesthesia for procedures unrelated to the pregnancy may be a particularly good population for separating anesthetic exposure from the multifactorial risks faced by the infant with structural anomalies requiring surgical intervention. Although no studies can reduce to zero the possibility of anesthetic-induced neurocognitive injury, studies should be adequately powered to demonstrate that the risk is minor compared with the well recognized risks of anesthesia and surgery in children.
   
5. A translational research effort using currently available methods in genomics and proteomics should be initiated. The goal would be to identify sensitive and specific biomarkers for neurocognitive injury in human neonates and infants that could be detected in blood and/or urine. Identification and validation of such biomarkers would facilitate subsequent human genetic association, pharmacologic, and therapeutic studies.
   
6. Until the risk of neurocognitive injury is understood, pediatric surgical specialties, in conjunction with anesthesiologists and pediatricians, should identify surgical procedures that can be delayed until older ages without incurring additional risk. It is therefore critical to gather outcome data to define the risks and benefits of early versus delayed repair.
   
7. It is not known whether local anesthetics are associated with a risk of neuroapoptosis. However, even lacking these data, it is reasonable to consider increased use of regional anesthetic techniques, where possible.
   
8. Single centers are unlikely to have sufficient numbers of "healthy" infants undergoing surgery and diagnostic procedures to determine subtle neurocognitive changes. The lack of data on pediatric anesthetics and their outcomes strongly emphasizes the need for a national pediatric anesthesia registry to serve as a foundation for these efforts. Of note, several United States children’s hospitals, in conjunction with the Society for Pediatric Anesthesia, have founded and financed an effort to establish a national registry of pediatric anesthetic procedures, with goals of expanding participation and data collection to a wide range of locations performing pediatric anesthesia, as well as performing root cause analyses of adverse events. Such a database could obviously become quite valuable to the study of the long-term consequences of receiving anesthesia as an infant or young child.
   
9. Finally, anesthetic-induced developmental neurotoxicity must be framed within the appropriate context. At the conclusion of the FDA Panel, one of the summary questions was "Are there sufficient data to determine applicability of the findings of anesthetics in nonclinical models to humans?" The answer of the voting FDA Committee Members was a unanimous "No." The second question was "To what extent are the doses and duration of exposure to anesthetics used in nonclinical studies relevant to the clinical use of these drugs?" The Panel determined that the extent was still unclear. It was therefore determined that, based on the current knowledge and lack of appropriate alternatives, there was no scientific basis to recommend changes in clinical practice. In our view, that statement remains true a year later.
   

Thus, understanding developmental neurotoxicity as it applies to the administration of anesthesia to pregnant women, infants, and children requires acknowledging the importance of the research conducted to date as well as its limitations. It requires a plan to move forward. We must acknowledge that virtually all pediatric surgical procedures are necessary, without alternatives, and require anesthesia. And, perhaps most critical to frame this discussion in the broader clinical context, we must acknowledge that our primary goal in anesthesia risk management in children is avoiding hypoxia and cardiovascular collapse. These known and understood causes of brain injury and death in pediatric patients during anesthesia are the real enemy. We must not let our enthusiasm for understanding the possible neurocognitive risks of anesthetics in children obscure our awareness of this enemy or prevent us from alleviating pain.

We have much to do.

	Footnotes

 
Accepted for publication March 10, 2008.

Reprints will not be available from the author.

	REFERENCES

Top REFERENCES

 

1. Loepke AW, McGowan FX Jr, Soriano SG. CON: The toxic effects of anesthetics in the developing brain: the clinical perspective. Anesth Analg 2008;106:1664–9[Free Full Text]
2. Jevtovic-Todorovic V, Olney JW. PRO: Anesthesia-induced developmental neuroapoptosis: status of the evidence. Anesth Analg 2008;106:1659–63[Free Full Text]
3. Loepke AW, Soriano SG. An assessment of the effects of general anesthetics on developing brain structure and neurocognitive function. Anesth Analg 2008;106:1681–707[Abstract/Free Full Text]
4. Wang C, Slikker W, Jr. Strategies and experimental models for evaluating anesthetics: effects on the developing nervous system. Anesth Analg 2008;106:1643–58[Abstract/Free Full Text]
5. Sanders RD, Xu J, Ma D, Maze M. General anesthetics induce apoptotic neurodegeneration in the neonatal rat spinal cord. Anesth Analg 2008;106:1708–11[Abstract/Free Full Text]
6. Degos V, Loron G, Mantz J, Gressens P. Neuroprotective strategies for the neonatal brain. Anesth Analg 2008;106:1670–80[Abstract/Free Full Text]
7. Cattano D, Young C, Straiko MM, Olney JW. Subanesthetic doses of propofol induce neuroapoptosis in the infant mouse brain. Anesth Analg 2008;106:1712–4[Abstract/Free Full Text]
8. Ikonomidou C, Bosch F, Miksa M, Bittigau P, Vockler J, Dikranian K, TenkovaTI, Stefovska V, Turski L, Olney JW. Blockade of NMDA receptors and apoptotic neurodegeneration in the developing brain. Science 1999;283:70–4[Abstract/Free Full Text]
9. Uemura E, Levin ED, Bowman RE. Effects of halothane on synaptogenesis and learning behavior in rats. Exp Neurol 1985;89:520–9[Web of Science][Medline]
10. Mellon RD, Simone AF, Rappaport BA. Use of anesthetic agents in neonates and young children. Anesth Analg 2007;104:509–20[Abstract/Free Full Text]
11. Miller SP, McQuillen PS, Hamrick S, Xu D, Glidden DV, Charlton N, Karl T, Azakie A, Ferriero DM, Barkovich AJ, Vigneron DB. Abnormal brain development in newborns with congenital heart disease. N Engl J Med 2007;357:1928–38[Abstract/Free Full Text]
12. McQuillen PS, Hamrick SEG, Perez MJ, Barkovich J, Glidden DV, Karl TR, Teitel D, Miller SP. Balloon atrial septostomy is associated with preoperative stroke in neonates with transposition of the great arteries. Circulation 2006;113:280–5[Abstract/Free Full Text]
13. Shprintzen RJ, Higgins AM, Antshel K, Fremont W, Roizen N, Kates W. Vel-cardio-facial syndrome. Curr Opin Pediatr 2005;17:725–30[Web of Science][Medline]
14. Anand KJ. Effects of perinatal pain and stress. Prog Brain Res 2000;122:117–29[Web of Science][Medline]
15. Yon JH, Carter LB, Reiter J, Jevtovic-Todorovic V. Melatonin reduces the severity of anesthesia-induced apoptotic neurodegeneration in the developing rat brain. Neurobiol Dis 2006;21:522–30[Web of Science][Medline]
16. Elewa HF, Hilali H, Hess DC, Machado LS, Fagan SC. Minocycline for short-term neuroprotection. Pharmacotherapy 2006;26:515–21[Web of Science][Medline]
17. Chuang DM. The antiapoptotic actions of mood stabilizers: molecular mechanisms and therapeutic potentials. Ann NY Acad Sci 2005;1053:195–204[Web of Science][Medline]

This article has been cited by other articles:

				A. W. Loepke, G. K. Istaphanous, J. J. McAuliffe III, L. Miles, E. A. Hughes, J. C. McCann, K. E. Harlow, C. D. Kurth, M. T. Williams, C. V. Vorhees, et al. The Effects of Neonatal Isoflurane Exposure in Mice on Brain Cell Viability, Adult Behavior, Learning, and Memory Anesth. Analg., January 1, 2009; 108(1): 90 - 104. [Abstract] [Full Text] [PDF]

This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by McGowan, F. X. Articles by Davis, P. J. Search for Related Content PubMed PubMed Citation Articles by McGowan, F. X., Jr Articles by Davis, P. J.