GAS, PANDA, and MASK: Comment

In their recent Editorial, Drs. Vutskits and Culley discuss the General Anesthesia or Awake-regional Anesthesia in Infancy (GAS), Pediatric Anesthesia Neurodevelopment Assessment (PANDA), and Mayo Anesthesia Safety in Kids (MASK) studies.¹  These studies are similar in that they all prospectively tested children exposed to a single anesthetic at a young age using comprehensive batteries of neurodevelopmental assessments. While this Editorial interprets these studies as showing “No Evidence of Clinical Anesthetic Neurotoxicity!” with an emphatic exclamation point, we respectfully disagree for two reasons.

First, the Editorial suggests that the GAS study found no difference in “primary and secondary outcomes” and that the PANDA and MASK studies provide strong evidence of a lack of “detectable alterations in neurodevelopmental outcome.”^2–4  This is largely true for the primary outcomes and most secondary outcomes. However, this was not true for parent reports of behavior/emotions and executive function (i.e., guiding, directing, and managing cognitive abilities, as well as behavioral and emotional functioning). All three studies reported significantly more problems in at least one of these reports.⁵  Although parent reports have definite limitations, they are an essential component of clinical neuropsychologic evaluations because children may exhibit different behavior in the laboratory environment versus other settings such as home or school. Interpreting these scores can be confusing because in contrast to other neuropsychologic tests, higher scores typically signify more problems. Four scores based on parent reports were reported in all studies: the Behavior Rating Inventory of Executive Function Global Executive Composite and the Child Behavior Checklist Internalizing Problems, Externalizing Problems, and Total Problems scores. Both the GAS and the MASK studies found significantly more problems with Executive Function as measured by the Behavior Rating Inventory of Executive Function. The PANDA study found that a higher proportion of children exposed to a single anesthetic for inguinal herniorraphy had clinically abnormal scores on the Child Behavior Checklist Internalizing Problems (e.g., depression or anxiety) scale compared with their unexposed siblings. The MASK study found that a higher proportion of children exposed to a single anesthetic had clinically abnormal scores on the Child Behavior Checklist Externalizing Problems (e.g., aggression or rule breaking behavior) scale compared with unexposed children. Furthermore, the observed effects (point estimates) were similar among all three studies, particularly for the Child Behavior Checklist outcomes. The clinical significance of these small differences in the absence of changes in other outcomes can be argued, but it is unjustified to conclude that there is “no evidence [!]” of an association between single exposures to anesthesia and any neurodevelopmental outcome.

Second, one criterion proposed for determining the biologic effect of exposure to a potential toxin is biologic gradient (i.e., a dose–response relationship). The MASK study also examined children who received multiple exposures to anesthesia before age 3. Such exposures were associated with an approximate doubling of the risk of developing attention-deficit hyperactivity disorder or learning disabilities (reproducing an earlier study). Multiple exposures were also associated with modest deficits in fine motor skills and rapid automatic naming, a proxy for processing speed later in life in some children, without deficits in other neuropsychologic domains such as general intelligence and memory. Thus, the MASK study provides compelling evidence for a dose–response relationship between exposures to anesthesia and specific neurodevelopmental outcomes. Exposure to multiple anesthetics before age 3 is not uncommon. In the population-based cohort of children used to construct the MASK study cohort, 15% of children required general anesthesia before age 3. Of these, approximately 25% received multiple anesthetics before age 3. If extrapolated to the U.S. population, this would represent approximately 160,000 children each year. The associations noted in the MASK and other studies do not prove that anesthetics are causative agents—the potential for confounding by indication and other factors has been exhaustively discussed and is very real. Yet the finding of a specific pattern of differences may argue against this, as it is not clear what common underlying condition would produce such a specific pattern. If neurotoxicity occurs with multiple exposures, either it also occurs with single exposures (even if largely subclinical), or the first exposure primes the brain for later injury.

It may be useful to recognize that there are perhaps two separate questions here: one clinical and one scientific. The clinical question is “Should concerns about potential neurotoxic effects of anesthetic drugs prompt immediate changes in clinical practice?” In the case of single exposures, we and most others would agree that the answer to the latter is “no” based on the current evidence, with the caveat that for one in four children, a single exposure presages multiple exposures. Such multiple exposures could (if anesthesia proves causative for the observed associations) have more significant effects in some children. The scientific question is “Do anesthetic drugs produce long-term changes in the brains of children?” It is important to answer this question regardless of whether these changes produce clinically-relevant effects in most children. A confident exclamatory assertion of “no evidence [!]” may make us feel better as clinicians, but may not serve the pursuit of further knowledge in this important area.

The authors declare no competing interests.