General anesthetics and sedatives are used in millions of children every year to facilitate surgical procedures, imaging studies, and sedation in operating rooms, radiology suites, emergency departments, and ICUs. Mounting evidence from animal studies suggests that prolonged exposure to these compounds may induce widespread neuronal cell death and neurological sequelae, seriously questioning the safety of pediatricanesthesia. 
Within the span of a single generation of pediatricanesthesiologists, the approach to neonatal anesthetic care has changed from a belief that the provision of analgesia, hypnosis, and amnesia to the neonate was unnecessary.
The pendulum of anesthetic drug provision swung from its nadir to a position where multiple anesthetic agents were administered to neonates as a matter of routine.Over the past several years, however, many have begun to question whether the pendulum has swung too far and the need for anesthesia in the neonates should be reassessed.
As such, pediatricanesthesiologists are being increasingly confronted with questions regarding the safety of the drugs that, in effect, define their profession. Answering such questions first involves understanding the evidence in support of anesthetic neurotoxicity, its potential shortcomings, and current research addressing this issue.

The evidence

Data supporting anesthetic neurotoxicity have been present in the literature for over 10 years. These data originate from mechanistic studies of fetal alcohol syndrome, an accepted, well-defined, and permanent neurotoxidrome. In brief, it was found that exposure of developing rodents to ethanol, a known N-methyl-d-aspartate (NMDA) receptor antagonist and γ-aminobutyric acid (GABA) receptor agonist, during a critical period of development resulted in widespread neuroapoptosis of the central nervous system.
Given that most anesthetic agents are believed to exert their effects via these receptors, the data were subsequently reproduced using established anesthetic agents and neurodegeneration was described using a wide variety of molecular and histologic techniques. These abnormalities, however, were without defined developmental significance, until a landmark study reported long-term cognitive deficits in developing rodents exposed to a combination of midazolam, nitrous oxide, and isoflurane. These data have subsequently been reproduced in a wide variety of species using many of the anestheticagents in common clinical use.
Although criticism regarding dose, duration, degree of clinical monitoring, and generalizability of rodent data to humans remains, many of these concerns have been addressed by large animal models utilizing nonhuman primates. At present, seven published studies have reported neuroapoptosis in response to anesthetic exposure in developing monkeys, all using monitoring standards similar to that of children undergoing operative procedures.
Degeneration of oligodendrocytes has also been reported in developing monkeys, suggesting that the spectrum of anesthetic neurotoxicity may be wider than previously assumed. At the same time, there have been many attempts to prevent the histiologic and developmental abnormalities induced by anesthetics, many of which have been successful in the laboratory. These protective techniques include, but are not limited to, hypothermia, melatonin, dexmedetomidine, lithium, erythropoietin, xenon , bumetanide, and environmental enrichment.
However, it has not yet been demonstrated in monkeys that prevention of histiologicabnormalities induced by anesthetic exposure prevents the aforementioned behavioral consequences. Indeed, while isoflurane has been reported to both induce and not induce histiologic injury in monkeys, the developmental consequence of isoflurane exposure has not yet been reported in this species. As such, it has not yet been established that the absence of histiologic changes can serve as an adequate marker for the complete absence of neurocognitive injury, which may complicate the development of neuroprotective strategies.
In contrast to the hundreds of preclinical studies in animals is the relative dearth of reports studying the neurodevelopmental effect of anesthetics in humans. While the prospective randomized controlled trial is the gold standard for determination of cause and effect, cost, length of time from exposure to the measured outcome, and ethical considerations have made such studies prohibitive with regard to anesthetic neurotoxicity. These studies exhibit significant variability due to varying population selection, comparators, definition of anesthetic exposure, timing of anesthetic exposure, outcome measurements, and findings. Some of the retrospective studies encompassed a period during which halothane, a drug no longer used in the United States, was the primary anesthetic agent used, and standard monitoring devices (such as the pulse oximeter) were not available.

With regard to outcome measurement, comprehensive neurocognitive tests are the gold standard for both determining the presence or absence of neurologic deficits as well as quantifying their magnitude. The remaining studies utilized data from individual or group-administered tests of achievement (GTA), teacher/parent rating scales, and diagnostic codes. Unlike individual tests of achievement, GTAs are intended to serve as sensitive tests to screen large numbers of subjects but lack the specificity necessary for diagnostic precision.  Lastly, parent/teacher reports are overtly subjective, and information on their sensitivity and specificity is completely lacking with respect to the outcomes of interest.

Information for surgeons & parents

Results from animal studies must be extrapolated to humans with great caution. One need not look further than penicillin, perhaps one of the greatest drug discoveries in the history of medicine, to learn of the disparate effects drugs have on animals as opposed to humans. Penicillin is profoundly toxic, to the point of inducing fatal enterotoxemic and hemorrhagic syndromes, in both guinea pigs and rabbits, two mammals commonly used in laboratory research at the time this drug was discovered. Had the initial testing of penicillin been conducted in these species, this incredibly important drug likely might never have been used in humans.  
As a more recent example, fluoroquinolone antibiotics have been shown to cause irreversible degeneration of articular cartilage in a wide variety of juvenile animals including dogs, mice, and guinea pigs, an effect that has not yet been demonstrated in humans despite worldwide utilization of these drugs. Similar caution must also be extended to anesthetic neurotoxicity, given that its purported mechanism primarily involves neuronal cell death. While much attention has been given to the fact that many neurons die by apoptosis as a normal part of growth and development (as much as 50–70%), comparatively little has been given to those neurons that survive.
Unlike the cellular lining of the gastrointestinal system that is effectively replaced by mitosis every few days, neurons do not replicate and are thus among the longest living cells within the body. The lifespan of human neurons can be over 100years; neurons of a mouse in captivity live for 2years, a small fraction of their human counterparts. Such an observation suggests a considerable difference in the magnitude of neuronal survival mechanisms between these two species, a magnitude that also may govern how the neurons respond to stress and injury. This may help explain why the described neurotoxic effects of anesthetics in rodents are particularly robust, while evidence in humans is comparatively weak.
What can be said with regard to the human studies? At best, these studies have not ruled out the possibility that anesthetic neurotoxicity exists in humans. This, of course, is a far cry from stating that it has been ‘ruled in’. Parents (and some clinicians) often misunderstand the difference between association and causation. An illustrative example is the association between gray hair and death. While gray hair is certainly associated with death, it does not cause death; interpretation of such an association is, as in all retrospective analysis, subject to confounders in which the studied factor may only serve as a marker for the causative factor, in the above example that of age. There are no data that would indicate that a change of anesthetic practice should be undertaken and may have unintended consequences secondary to delaying necessary surgery or changing anesthetic practice to one that is unknowingly more risky.
As there is not yet certainty within the medical community that there is a real risk posed by anesthetics, there does not seem to be a compelling medical argument to actively bring up the purported risk with patients and their parents as part of the consent/assent process. Such a discussion would be of unclear benefit, particularly to those undergoing nonelective surgery, and may induce unnecessary and unwarranted concern on the part of the parent. The effect of preoperative discussions of anesthetic risk on ameliorating parental anxiety has been reported in a number of studies, and unlike comparable studies in adults, detailed information regarding anesthesia has not been associated consistently with decreased parental anxiety.  Nevertheless, some practitioners are beginning to question whether there is a legal responsibility to discuss the possibility of anesthetic neurotoxicity.
Parental desire for anesthetic information has been studied and virtually all studies report that the vast majority of parents want to know about the risks of anesthesia, including severe and rare risks such as death. When parents are further queried regarding what the specific content of perioperative conversations should be, it has been reported that parents most often want detailed information on readily apparent, short-term concerns—notably pain, nausea, anesthetic induction, and emergence. With regard to children, their desire for information has also been studied and similar themes emerge; the majority of children want comprehensive perioperative information, but mostly with regard to well-defined, largely short-term problems. As such, for both children and their parents, the extant literature on the discussion of anesthetic risk largely studied risks that are accepted and short-term, rather than those that are hypothetical and long-term. It is thus difficult to extend these data to the issue of anesthetic neurotoxicity.
So what do we tell the parents of our patients when asked whether their children are at risk of adverse neurocognitive outcomes as a result of anesthesia?

In INDIA alone, more than 1million children, 4years of age and under, undergo surgical procedures requiring anesthesia annually. While most children appear to recover well, findings from these animal studies call for further research to ensure the safety of every child undergoing anesthesia. Until this determination can be made, children requiring surgery essential to their health should proceed as directed by their physician. Young children usually do not undergo surgery unless the procedure is vital to their wellbeing. Therefore, postponing a necessary procedure may itself lead to significant health problems, and may not be an option for the majority of children.

References
1. Davidson AJ. Neurotoxicity and the need for anesthesia in the newborn: does the emperor have no clothes? Anesthesiology
2012; 116: 507–509.
2. Anand KJ, Hickey PR. Pain and its effects in the human neonate and fetus. N Engl J
Med 1987; 317: 1321–1329.
3. Anand KJ, Hickey PR. Halothane-morphinecompared with high-dose sufentanil for anesthesia
and postoperative analgesia in neonatalcardiac surgery. N Engl J Med 1992; 326:1–9.
4. Anand KJ, Sippell WG, Aynsley-Green A. Randomised trial of fentanyl anaesthesia in preterm babies undergoing surgery: effects on the stress response. Lancet 1987; 1: 62–66.
5. Gleich S, NemergutM, Flick R. Anesthesia related neurotoxicity in young children: an
update. CurrOpinAnaesthesiol2013; 26: 340–347.
6. Brambrink AM, Orfanakis A, Kirsch JR. Anesthetic neurotoxicity. AnesthesiolClin2012; 30: 207–228.
7. Stratmann G. Review article: Neurotoxicity of anesthetic drugs in the developing brain.
AnesthAnalg2011; 113: 1170–1179.
8. Pruett D, Waterman EH, Caughey AB. Fetal alcohol exposure: consequences, diagnosis,
and treatment. ObstetGynecolSurv2013; 68:62–69.
9. Ikonomidou C, Bittigau P, Ishimaru MJ et al. Ethanol-induced apoptotic neurodegeneration and fetal alcohol syndrome. Science 2000; 287: 1056–1060.
10. Loepke AW, Soriano SG. An assessment of the effects of general anesthetics on developings
brain structure and neurocognitive function. AnesthAnalg2008; 106:
1681–1707.
11.Michael E. Nemergut, Devon AgangaandRandall P. FlickAnesthetic neurotoxicity: what to tell the parents?PediatricAnesthesiaVolume 24, Issue 1, pages 120–126, January 2014