On December 14, 2016, the US Food and Drug Administration (FDA) released a Drug Safety Communication about the use of general anesthetics and sedation drugs in young children and pregnant women. Consistent with federal regulations, this warning was released by the FDA independent of the SmartTots program. The warning states that “repeated or lengthy use of general anesthetic and sedation drugs during surgeries or procedures in children younger than 3 years or in pregnant women during their third trimester may affect the development of children’s brains” (http://www.fda.gov/Drugs/DrugSafety/ucm532356.htm). This Drug Safety Communication has resulted in mandatory changes to the labels of 11 commonly used general anesthetics and sedative agents, including sevoflurane, propofol, ketamine, barbiturates, and benzodiazepines.1
The FDA warning stimulated considerable controversy among clinicians and the public as to whether it appropriately informs patients and care providers about the potential neurotoxic effects of anesthetics or causes needless concern.1 , 2 The debate continues simply because the best available data remain inconclusive. These ongoing concerns call for the scientific community and SmartTots to continue to pursue whether general anesthetic drugs are toxic to the developing human brain.
It is important to consider the reasons that the controversy about anesthetic neurotoxicity continues, despite >7 years of investigation. The FDA warning was based primarily on results from studies of laboratory animals, including nonhuman primates. The data have provided ample evidence that general anesthetic drugs cause pathological changes, including cell death, reduced neurogenesis, and disruption of synaptic structures and neuronal circuitry.3 These changes are associated with neurobehavioral deficits that persist even as the animals mature. The evidence from human studies that informed the FDA warning was far less compelling. In fact, resolving the question of whether anesthetic drugs are neurotoxic to patients has proven difficult. No single randomized controlled clinical trial will provide the answer to this central question because of various confounding factors. These factors include underlying neurodevelopmental or neurological disorders in children who require multiple or prolonged exposure to anesthetics, as well as effects of disease or inflammation caused by disease or surgery, independent of drug effects. To further complicate the issue, animal studies have shown that neurotoxicity depends on the dose, the number of anesthetic exposures, the age of the subjects, and the stage of neurodevelopmental maturation at the time of exposure.3 Comorbidities such as inflammation, caused by either surgery or the underlying illness, may increase the brain’s sensitivity to anesthetics and hence the resultant neurotoxicity.4 Thus, the overall research approach needs to be iterative and nuanced. Although some progress has been made, additional studies are required, including those in the preclinical setting.
Two clinical studies partially sponsored by SmartTots offer some reassurance that a brief exposure to general anesthetic drugs does not cause overt, persistent neurocognitive deficits. The first large-scale randomized controlled study to compare cognitive outcomes in infants undergoing general anesthesia to those receiving a regional anesthetic is the General Anaesthesia Compared to Spinal Anaesthesia trial.5 The primary outcome will be scores on an intelligence quotient (IQ) test, measured at 5 years of age. Data for the primary outcome are not yet available; however, the secondary outcome (measured at 2 years of age) showed no performance deficits. More definitive answers will be available when the data for children 5 years of age are reported in 2018.
The Pediatric Anesthesia and Neurodevelopment Assessment study used an observational ambidirectional cohort design to compare neurocognitive performance and behavioral outcomes in children who underwent hernia repair before 3 years of age to the performance of their unexposed siblings, with the outcomes being assessed at 8–15 years of age.6 No overt differences in outcomes were detected between exposed and unexposed siblings.
Results from the General Anaesthesia Compared to Spinal Anaesthesia and Pediatric Anesthesia and Neurodevelopment Assessment studies suggest that a short exposure to a general anesthetic drug does not cause persistent cognitive impairment. These results are consistent with data from preclinical studies showing that pathological changes occur only when the duration of exposure is >2 hours or when subjects undergo repeated drug exposures.3
Some additional reassurance has been provided by several retrospective epidemiological studies that examined cognitive outcomes after anesthesia in large cohorts of children. One of these studies linked information gleaned from a health care database to performance on a primary school entrance examination called the Early Developmental Instrument.7 The performance of children who underwent general anesthesia at an early age was compared with the performance of age-matched controls. The primary outcome was defined as any major domain of the Early Developmental Instrument in the lowest (10th) percentile. The results showed that those <2 years of age at the time of surgery had no increase in the risk of impaired performance, whereas children who were older at the time of surgery showed only a small increase in vulnerability relative to controls (25.6% vs 25.0%). The clinical significance of this seemingly small difference between groups is uncertain.
A second epidemiological study examined children who had 1 exposure to anesthesia and surgery before 4 years of age, as well as children who underwent multiple surgeries.8 The primary outcome was school grades at 16 years of age. IQ test scores at military conscription (at 18 years of age) were also compared (males only). The results showed that 1 exposure to anesthetics before 4 years of age was associated with 0.41% lower school grades and 0.97% lower IQ test scores. Similar differences were observed in test scores after multiple exposures. Neither of these 2 retrospective studies specified which anesthetic drugs were used or the duration of surgery.
The findings described above have led investigators to initiate studies of longer or repeated exposures to anesthetic drugs. One such trial that compares long-term outcomes after 2 different anesthetics is currently underway (A Study to Compare the Long-term Outcomes After Two Different Anaesthetics [TREX] trial; ClinicalTrials.gov Identifier: NCT03089905). This study will assess neurodevelopmental outcomes after a low dose of sevoflurane (administered in combination with remifentanil and dexmedetomidine) versus a standard sevoflurane-based anesthetic (a relatively high dose of sevoflurane). Children undergoing surgery lasting ≥2.5 hours will be assessed. The primary outcome is global cognitive function, measured with the full-scale IQ score of the Wechsler Preschool and Primary School Intelligence Scale. This international multicenter study is now recruiting patients in Australia, North America, and Europe.
The concerns about possible neurotoxic effects of anesthetic drugs on the developing brain originally sprang from the results of preclinical studies. As such, robust, reliable, and reproducible data from preclinical studies are essential to inform both the need for and optimum design of clinical trials. To investigate the reproducibility of data from preclinical studies, SmartTots recently sponsored 2 teams of investigators who performed studies that (1) were undertaken in independent laboratories under similar experimental conditions; (2) used the widely studied model of 7-day-old rat pups; (3) assessed neuroapoptosis, a commonly used primary end point of neurotoxicity; (4) used methodologies familiar to the investigators; and (5) were designed to inform future clinical trials.9 , 10 Specifically, the studies sought to determine whether sevoflurane-induced neuroapoptosis was reduced by cotreatment with dexmedetomidine. The rat pups were treated with sevoflurane (2.5% for 6 hours) alone, dexmedetomidine alone, or sevoflurane in combination with various doses of dexmedetomidine. These studies were stimulated by reports showing that dexmedetomidine was neuroprotective in several models of brain injury.11–13 Studies from both laboratories showed that generally, sevoflurane but much less so dexmedetomidine alone caused overt widespread apoptosis. Strikingly, however, one of the research teams found that dexmedetomidine reduced apoptosis,9 whereas the other reported opposite findings.10 Notably, in treatment groups that received both drugs, the dose of sevoflurane was not reduced to account for the anesthetic-sparing (or anesthetic-potentiating) properties of dexmedetomidine. Moreover, because of practical concerns, no effort was made to maintain adequate levels of oxygenation or ventilation or blood pressure management in the face of drug-induced neurodepression. Thus, not surprisingly, physiological parameters, including arterial blood gases and blood pressure, were markedly disrupted, and mortality rates were high in groups cotreated with sevoflurane and dexmedetomidine. One of the possible causes for the discrepancy in results for the primary outcome (neuroapoptosis) between the 2 studies could be differences in the level of neurodepression caused by the combination of drugs.
These pioneering studies represent one of the first attempts to evaluate the reproducibility of data from preclinical studies. In addition, they suggest that marked disturbances of physiological factors could contribute to changes in the brains of immature animals after general anesthesia. The results suggest that greater rigor is required when interpreting or extrapolating results from preclinical studies. An editorial that accompanied these 2 SmartTots-sponsored studies suggested that future studies of anesthetic neurotoxicity using animal models should meet certain experimental guidelines.14 For example, preclinical studies should be preregistered and their protocols reported, as occurs now for clinical studies. Such reporting helps to ensure that data are not overly selected for publication and that negative results are reported. The editorial also recommended that standards for preclinical studies, such as the Animal Research: Reporting of In Vivo Experiments (ARRIVE) or Stroke Therapy Academic Industry Roundtable (STAIR) guidelines, should be widely adopted.15 , 16 Finally, cardiorespiratory factors that promote neuroapoptosis, such as hypoxia and hypotension, must be taken into consideration when attributing evidence of neurotoxicity to the direct effect of anesthetic drugs on the developing brain. This latter point is important because infants and children undergoing anesthesia are generally supported to avoid the adverse cardiorespiratory effects that occurred in unsupported rat pups.
Another study, sponsored in part by SmartTots, showed that neuroapoptosis was increased in the brains of immature nonhuman primates treated with sevoflurane for 3 hours.17 Unlike the rat pups, these subjects received support to avoid physiological dysregulation. In these nonhuman primates with adequately supported oxygenation, ventilation, and blood pressure, sevoflurane increased neuroapoptosis, suggesting a direct adverse effect of this anesthetic drug on the brain. Thus, disruption of physiological factors alone does not account for the adverse anesthetic drug effects.
Ongoing discussions during SmartTots-sponsored meetings of preclinical and clinical investigators have highlighted the need to identify key research priorities and coordinate activities. The investigators concluded that high priority should be given to preclinical studies examining dose–response relations of neurotoxicity, head-to-head comparisons of drugs or combinations of drugs that mitigate anesthetic neurotoxicity, and the search for “translatable” biomarkers. Such studies should be undertaken in experimental animal models that will allow physiological support to mitigate the adverse effects of anesthetics on cardiorespiratory function. Also, the role of inflammation needs to be investigated, given that inflammatory factors increase the sensitivity of neurons to anesthetic drugs and may exacerbate anesthetic neurotoxicity.4
Finally, SmartTots recently achieved an important milestone that aims to increase the pace of discovery. Through generous support provided by the FDA and the International Anesthesia Research Society, the organization was recently able to appoint a Medical Director. The Medical Director will be charged with identifying research priorities, ensuring ongoing dialogue between investigators, and contributing new knowledge to the field. We are pleased to announce that Dr Lena Sun, a pediatric anesthesiologist and clinician–investigator from the Columbia University Department of Anesthesiology, has been appointed as the inaugural Medical Director of SmartTots.
In summary, general anesthesia is an essential component of pediatric care, and millions of children undergo anesthesia and surgery each year. The most recent evidence suggests that a single, brief exposure to a general anesthetic does not cause overt neurocognitive deficits in either neonatal laboratory animals or human infants. However, it is still premature to conclude whether anesthetic drugs administered early in life cause harm. The ongoing TREX trial is 1 study of great interest. In animal models, future studies that examine neurotoxicity need to measure physiological parameters such as blood pressure and oxygenation to delineate the direct neurotoxic properties of anesthetic drugs. The appointment of Dr Lena Sun as the Medical Director for the SmartTots program is a major step forward that will facilitate discoveries.
Resolving whether anesthetics are neurotoxic to the human infant nervous system is paramount as the data will either allay unfounded fears that could discourage or delay beneficial surgery or cause substantial changes in pediatric anesthetic practice and drive the quest for strategies to reduce anesthetic-induced harm.
The authors wish to acknowledge Dr Merle Paule, Director of the Division of Neurotoxicology of the National Center for Toxicological Research (NCTR) of the US Food and Drug Administration (FDA). Dr Paule retired on December 31, 2017, after providing many years of scientific contributions and exemplary leadership to the field of anesthetic neurotoxicity.
Name: Beverley A. Orser, MD, PhD.
Contribution: This author wrote the manuscript.
Name: Santhanam Suresh, MD.
Contribution: This author wrote the manuscript.
Name: Alex S. Evers, MD.
Contribution: This author wrote the manuscript.
This manuscript was handled by: Jean-Francois Pittet, MD.
1. Andropoulos DB, Greene MF. Anesthesia and developing brains: implications of the FDA warning. N Engl J Med. 2017;376:905–907.
2. Davidson A, Vutskits L. The new FDA drug safety communication on the use of general anesthetics in young children: what should we make of it? Paediatr Anaesth. 2017;27:336–337.
3. Vutskits L, Xie Z. Lasting impact of general anaesthesia on the brain: mechanisms and relevance. Nat Rev Neurosci. 2016;17:705–717.
4. Avramescu S, Wang DS, Lecker I, et al. Inflammation increases neuronal sensitivity to general anesthetics. Anesthesiology. 2016;124:417–427.
5. Davidson AJ, Disma N, de Graaff JC, et al. GAS Consortium. Neurodevelopmental outcome at 2 years of age after general anaesthesia and awake-regional anaesthesia in infancy (GAS): an international multicentre, randomised controlled trial. Lancet. 2016;387:239–250.
6. Sun LS, Li G, Miller TL, et al. Association between a single general anesthesia exposure before age 36 months and neurocognitive outcomes in later childhood. JAMA. 2016;315:2312–2320.
7. O’Leary JD, Janus M, Duku E, et al. A population-based study evaluating the association between surgery in early life and child development at primary school entry. Anesthesiology. 2016;125:272–279.
8. Glatz P, Sandin RH, Pedersen NL, Bonamy AK, Eriksson LI, Granath F. Association of anesthesia and surgery during childhood with long-term academic performance. JAMA Pediatr. 2017;171:e163470.
9. Perez-Zoghbi JF, Zhu W, Grafe MR, Brambrink AM. Dexmedetomidine-mediated neuroprotection against sevoflurane-induced neurotoxicity extends to several brain regions in neonatal rats. Br J Anaesth. 2017;119:506–516.
10. Lee JR, Lin EP, Hofacer RD, et al. Alternative technique or mitigating strategy for sevoflurane-induced neurodegeneration: a randomized controlled dose-escalation study of dexmedetomidine in neonatal rats. Br J Anaesth. 2017;119:492–505.
11. Sanders RD, Xu J, Shu Y, et al. Dexmedetomidine attenuates isoflurane-induced neurocognitive impairment in neonatal rats. Anesthesiology. 2009;110:1077–1085.
12. Kuhmonen J, Pokorný J, Miettinen R, et al. Neuroprotective effects of dexmedetomidine in the gerbil hippocampus after transient global ischemia. Anesthesiology. 1997;87:371–377.
13. Degos V, Charpentier TL, Chhor V, et al. Neuroprotective effects of dexmedetomidine against glutamate agonist-induced neuronal cell death are related to increased astrocyte brain-derived neurotrophic factor expression. Anesthesiology. 2013;118:1123–1132.
14. Vutskits L, Sall JW. Reproducibility of science and developmental anaesthesia neurotoxicity: a tale of two cities. Br J Anaesth. 2017;119:451–452.
15. Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG. Improving bioscience research reporting: the ARRIVE guidelines for reporting animal research. PLoS Biol. 2010;8:e1000412.
16. Fisher M, Feuerstein G, Howells DW, et al. Update of the stroke therapy academic industry roundtable preclinical recommendations. Stroke. 2009;40:2244–2250.
17. Noguchi KK, Johnson SA, Dissen GA, et al. Isoflurane exposure for three hours triggers apoptotic cell death in neonatal macaque brain. Br J Anaesth. 2017;119:524–531.