Special Report: Doctors Debate Study Finding IQ Lowered in Kids Post-Anesthesia
Lower IQ, reduced language comprehension, and decreased grey matter density is experienced by children having general anesthesia for surgery before age four, says new research in the journal Pediatrics [1].
“Our study is the first to directly compare cognitive performance measured by IQ tests and relate these functional findings to long-term brain structural changes in children following surgery with general anesthesia early in life,” lead author and Cincinnati Children’s Hospital Medical Center anesthesiologist Andreas Loepke, M.D., Ph.D., F.A.A.P. , told Bioscience Technology.
Some anesthesiologists say the findings, while of interest and use to the field, need to be cautiously interpreted.
“The possibility that general anesthetics might be neurotoxic in the developing brain has created huge concern and anxiety, and left anesthesia providers and parents with doubts as to whether it is safe and well-tolerated to use anesthesia in young children,” Tom Hansen, M.D., Ph.D., Head of Pediatric Anesthesiology at Odense University Hospital, told Bioscience Technology. Hansen was not involved in the study.
Read More: Electric Brain Stimulation Has Mixed Effect on IQ [2]
“Fortunately, in contrast to the plethora of positive animal studies, it has been difficult to demonstrate a human corollary to this phenomenon. If anesthesia-related neurotoxicity exists in humans, it must be vague, or present only in a subset of susceptible infants, otherwise it would have been easy to demonstrate, and it would most likely have been suspected many years ago. Some retrospective or follow-up cohort studies have suggested a link between exposure to anesthesia and surgery in early childhood and subsequently impaired neurobehavioral outcome. [But] similar studies have been unable to show any such association,” Hansen said.
Massachusetts Eye and Ear Infirmary (MEEI) Pediatric Anesthesiologist, and Harvard Medical School professor, Corey Collins, D.O., F.A.A.P., told Bioscience Technology: “This research offers new evidence that exposure to surgical anesthetics might contribute to a measurable decrement in language and performance IQ. While I find their data generally well-presented, my specific concerns are with non-significant data interpreted as note-worthy.” Collins was also uninvolved with the study.
106 children
Loepke [3] and his team have published [4] many past studies unearthing [5] widespread cell death, ablation of neurons, and neurocognitive impairment [6] in young rodents post-anesthesia.
For the new paper, the team compared scores of 53 healthy children in a language development study, all aged five to 18 years old, with no history of surgery, to scores of 53 children the same age who had surgery before age four.
The average test scores for all 106 children were within norms. But children who had had anesthesia scored significantly lower in aural comprehension, and performance IQ, than children who had not.
Lower IQ and language scores were associated with decreased gray matter density in both the brain’s occipital cortex, and the cerebellum.
The children were matched for gender, socioeconomic status, age, and handedness, all of which can impact cognition and brain structure. The authors also reported taking into consideration types of surgeries, and length of exposure. Anesthetics involved included common sevoflurane, isoflurane, halothane, and nitrous oxide.
No children had a history of neurologic or psychological illness, or head trauma. Cognitive tests included the Oral and Written Language Scales and the Wechsler Intelligence Scale. Structural comparisons were done via MRI.
Some controlled factors
“Even though our study included only slightly over 100 children, we were able to control for several factors that strongly affect neurocognitive outcome, by matching our two study groups for age, gender, and socioeconomic status,” Loepke told Bioscience Technology. Loepke is also a professor of clinical anesthesia and pediatrics at the University of Cincinnati College of Medicine. “By reducing bias due to confounding of these factors, we feel this design added substantially to the validity of our findings. But similar to all other currently available studies, our retrospective study design does not let us determine the differential effects of anesthesia, surgery, or pain and inflammation.”
The team did not expect the study’s results to be so dramatic.
“It is difficult to relate the widespread neuronal cell death and learning impairment found in animal studies following prolonged anesthetic exposure to every day pediatric anesthesia practice,” Loepke said. “We were surprised to find a change in cognitive performance, even following the substantially shorter exposure time in our children, compared with the much longer duration in animal studies. This suggests that cognitive tests may be very sensitive to assess long-term brain development following surgery early in life.”
Also unexpected, Loepke told Bioscience Technology, were the brain structural abnormalities, “which were not observed in brain regions that had previously been widely identified to be vulnerable to anesthesia-induced neuronal cell death in animals. All these questions emphasize that much more research is needed to better understand this potential pediatric health dilemma, and to improve safety.”
Ramifications
Due to difficulties examining brain structure and measuring long-term cognitive performance in such children, “we should take a two-pronged approach,” Loepke said. “On one hand, we need a much better understanding of the underlying mechanisms of the effects of anesthesia and surgery on the developing brain, which is best studied in animal models. These studies should lead to the discovery of targeted mitigating strategies or alternative anesthetic techniques safe in young children who often need to undergo surgical procedures to improve quality of life, treat serious health concerns, or save babies’ lives.”
On the other hand, Loepke said, “we need to perform additional studies to better define the human applicability of the concerning results in animal studies. We do not have enough information to recommend a change in our—by all other measures—very safe anesthesia practice. Rash changes, or delaying crucial procedures, may do more harm than good.”
Effects of surgery vs. anesthesia unknown
Hansen has reservations. A problem such studies “are struggling with,” he told Bioscience Technology: “It is virtually impossible to separate the pharmacodynamic effects of general anesthetics from a multitude of other factors that might also cause neurological damage like stress of surgery, reasons for surgery, or impairment due to inadequate anesthesia management.”
Hansen noted children never get anesthesia “for the sake of the anesthesia, but for surgical and/or pathological reasons.” So in children, “no definite causal link between certain anesthetic drugs or techniques, and poor neurological outcome, has been established. Prospective studies are underway, but will not be available for years.” Even then, he said, those studies “are likely to be inconclusive.”
Hansen noted the new paper describes a retrospective study “comprising an interesting cohort recruited from an MRI database that includes five- to 18-year-old volunteers born at term, and lacking any history of neurological or psychological comorbidities. The exposure group comprised 53 children with a history of surgery requiring anesthesia before the age of four. The control group comprised 53 age-, gender-, handedness- and socioeconomic- status matched children to non-exposed children of the same cohort. All children had details from hand-written anesthetic records reviewed, and underwent a battery of individual neurocognitive tests.”
The authors found “in exposed children, language comprehension and performance IQ were decreased compared with unexposed controls, and these functional abnormalities were associated with decreased gray matter volume, primarily in posterior brain regions. Wisely, the authors refrain from drawing clinical conclusions from these data, acknowledging the many limitations inherent in their study.”
ENT kids may skew results
Importantly, he said, “the authors included many types of surgeries and diagnoses. This is a major flaw, in that specific surgeries and diagnoses can independently impact subsequent neurocognitive outcome. It is a problem here, where the vast majority of children underwent ENT (ear, nose, and throat) procedures, because ENT children are well-known to suffer subsequent language and cognition problems. This issue will soon be highlighted in an ongoing large scale Swedish cohort study.”
Many other perioperative factors can impact neurocognitive outcome in young children exposed to surgery and anesthesia. “These factors do not receive the same attention in the media and medical world,” Hansen said. “The authors focused on some of these in their study, e.g. hypotension, bradycardia, hypoxia inflammation, pain, and genetics. Other factors are equally or more important in this respect, e.g. hyponatriemia, hypocapnia, hyperoxia, hypo- and- hyper-glycemia.”
More importantly, he said, there is no proven causal link between specific anesthetic agents or techniques, but there is “a definite [7] causal relationship between poor perioperative anesthetic care, and bad neurocognitive outcome and mortality, in neonates, infants, and small children.”
Overall, Hansen said, the Cincinnati study “does not add much new information regarding neurocognitive outcomes in young children exposed to surgery and anesthesia. But it emphasizes that such studies must be performed prospectively, in large and well-defined groups of children, with homogenous epidemiological data including type of surgery, anesthesia, and electronically recorded vital signs. Without ensuring proper conduct of anesthesia, as defined [7] by the 10N-Pediatric Anesthesia Quality Checklist, no neurocognitive or pathological MRI findings should automatically be linked to general anesthetics or surgery.”
Loepke’s responses
Loepke disagreed on several points. “I don’t understand the notion of ‘fortunately, it has been difficult to demonstrate a human corollary to this phenomenon,’” he told Bioscience Technology. “Our goal is to improve patient safety, not exculpate anesthetics from causing any harm. We frankly discuss the limitations of our study in the manuscript.” The paper noted that retrospective studies like his could not distinguish between anesthesia and surgery effects, he said.
Also, Loepke said: “Studies examining academic performance and grades have generally been negative following a single exposure, whereas studies involving individually administered tests of language—and in our study IQ—-have demonstrated an effect. This could very well suggest these domains are more sensitive to detect a signal. This was also mentioned in our discussion.”
He disagrees that it can be known any prospective studies will be inconclusive, as they “have not been completed. They will most definitely answer the question asked, but will obviously not be definitive.”
Also as discussed in his paper, Loepke said: It is “quite controversial” whether children with chronic ear infections will suffer language and cognition problems. Studies cited in the paper suggest otherwise, he said. And “ENT children” in the study were treated young, at about two years old, with myringotomies and ear tubes to prevent long-standing hearing problems that could lead to language and cognition problems. Loepke suggested kids not treated with surgery for chronic ear infections “should have more problems with language and cognition, since their hearing impairment was never treated. Since many young children experience ear infections, these children could have been in the control group, decreasing performance, and potentially making it harder to demonstrate exposure effect.”
The large-scale Swedish study is also retrospective, “like ours, so cannot distinguish between the effects of surgery and anesthesia.” It also relies on academic performance, “which has been demonstrated, by Caleb Ing at Columbia University, to be less sensitive to find an effect than individually administered tests. Interestingly, the Swedish study demonstrates a very small, yet measurable, detriment of anesthesia and surgery early in life, and subsequent lower academic performance. Like our study, this is association, not causation.”
Loepke noted his study included “very strong confounders” of cognition. No other human studies included all those factors, he contended. And “we did not claim our study was all-inclusive of all the factors, inside and outside of the hospital that could affect neurocognitive outcome. No study will ever capture all these factors. Accordingly, we never made the claim of having demonstrated a causal relationship between anesthesia exposure and the observed outcome—just the association. To state there is a ‘definitive causal relationship between poor perioperative anesthetic care and bad neurocognitive outcome’ is not correct, as there is no grading system for good or poor perioperative care.”
Loepke added that his team “respectfully, but vehemently, disagree with the statement our study ‘does not add much new information regarding neurocognitive outcomes in young children exposed to surgery and anesthesia.’ Our study is the first to directly compare IQ tests between previously exposed and unexposed children. It confirms a very recent study demonstrating that individually administered tests are more sensitive than academic performance to find a phenotype, which could explain negative results of several studies.”
Loepke stressed again that his study is the first to match for the “strong confounders” of gender, age, and socioeconomic status, and the first to assess brain structure after anesthesia in early childhood.
And while his study relied on hand-written anesthesia records, he said, “we were able to report vital signs and drug doses, which the great majority of epidemiological studies was completely lacking.”
Finally, Loepke said, his group “strongly” disagrees their study does not “significantly contribute to a very important health concern.” Anesthesiologists and pediatricians need to work together to improve understanding “of the effects of anesthesia, surgery, and the perioperative medical management on brain development. These discussions may be painful, but we have to face them. More research is needed.”
Small size of study
The main objections of Collins, who has served as MEEI pediatric anesthesiology chief, are with study size, the handwritten notes, and previous anesthesia status. The authors state that “‘previously exposed children scored lower than unexposed matched control group in all tests and subcategories,’” Collins told Bioscience Technology. “I find this statement to be unwarranted with such a small retrospective study with significant limitations. The reliance on hand-written anesthesia records introduces many concerns about accuracy, completeness, legibility, or veracity. Also, the inability to confirm the anesthesia-exposure status of all controls seems to limit the accuracy. How can we know the unexposed group did not receive an anesthetic?”
Collins said that, at the sample size provided, two to three confounding variables would eliminate the statistical significance of the data. “I note that unexposed cohort in figure 1 appears to have a performance IQ mean between 110-120, well outside normal distribution for an otherwise random sample. The MRI data seems to imply a correlation between the cognitive effect of anesthesia exposure, and a physical measurement of regional brain injury or neuronal loss. While I applaud the authors for including this data, I do not agree the methodology would support this conclusion. While brain development and MRI technique are well outside my expertise, I wonder if many other factors impact the regional brain development of children: family history, body size, obesity/nutritional status, and infectious disease history are a few questions that come to mind.”
Finally, Collins is concerned about the heterogeneous series of anesthetic exposures and ages. “Surgical times were from five to 170 minutes, and ages were from one day to 3.8 years. They did not review the individual anesthetic agents for distribution in the exposed cohorts. These issues are of importance, since such variables greatly influence the nature of exposure to anesthesia, likely complications encountered and whether such events were documented. For example, laryngospasm is much more prevalent in infants leading to hypoxemia, though many anesthesiologists consider it to be expected for pediatric cases, and will not document such events. Halothane was the primary inhaled agent for over 50 years, yet no research on neurocognitive effects have looked at it to my knowledge.”
ENT kids
When it comes to kids and ENT diseases, Collins noted “the need for anesthesia in healthy children under age one year is most often related to otitis, and the surgical rationale is under challenge from the Swedish study mentioned, and others.”
He added he thinks that toxicity issue “will be most immediately relevant when considering myringotomy in the USA, where there is good evidence to challenge the benefit of surgery. Parents are prudent to consider this before agreeing to surgery, and surgeons will need to be prepared to defend the surgical benefit. We are already having this conversation at MEEI, and discussing alternative anesthetic options for children. Parents have, as recently as last week, elected to postpone surgery because of this question.”
Soon enough, he said, the field may face “similar challenges to tonsillectomy, since Marcus and colleagues published [8] a good trial demonstrating that over 50 percent of children with clinically proven OSA (obstructive sleep apnea) resolve to normal sleep studies without surgery.”
Other “firsts”
Noting his paper addressed limits of hand-written records, Loepke added that, “since all studies have to rely on assessments in older patients, anesthetic exposures occurred in the era before electronic records.” Regarding the paper’s statement that “previously exposed children scored lower than unexposed matched control group in all tests and subcategories,” he said “one could phrase it that, on average, previously exposed children did not score higher on any of the subtests performed, whereas they scored worse on listening comprehension and performance IQ.”
Loepke said the paper also addressed “our inability to confirm non-exposure in all control subjects.” He agreed this could have “biased our study against finding a difference.” But as a difference was found despite this, his findings were “strengthened” by it, he said.
The MRI analysis, Loepke emphasized, is “the first, albeit preliminary, long-term brain structural assessment following surgery with anesthesia. We convincingly demonstrated no measurable gray matter loss was observable in brain regions with permanent neuronal deletion in animals, such as retrosplenial cortex and thalamus. While any study can be criticized as not addressing all confounders of brain development, we addressed several major confounders like age, gender, and socioeconomic status, which previous studies did not routinely address.”
Regarding the above-average performance, “our children were volunteered by their parents, who may have been better educated than the general population,” he said. But as “we adjusted for family income, this confounder had no effect on results.”
Loepke said his cohort’s heterogeneity “clearly reflects the range of children undergoing pediatric anesthesia, since it is very similar to studies like those performed by the Mayo Clinic group. However, we documented specific drugs as well as equipment doses, finding that 84 percent of patients were exposed to Halothane, but 19 percent received the more contemporary anesthetic Sevoflurane.”
Many animal studies link halothane to deleterious brain development, he said. And children in the Mayo Clinic study exhibiting “learning impairment following multiple anesthetic exposures” were “predominantly exposed to halothane.” His team wrote a review [9] of the studies.
“Profound” effect in vitro
The bottom line is that it is clear anesthetics have a “profound” effect on mammalian neuronal growth and development in vitro, said Collins.
Furthermore, Collins said, “While human development is likely far more complex than that of non-human primates and mammals, there is sufficient data to warrant careful study. I find this research similar to many others in that it suggests a possible effect. I believe the data should be regarded as a pilot, one that offers rationale for specific, carefully designed research to establish a reasonable scientific relationship between exposure to anesthesia and any practical impact on human development.”
Loepke concluded: “We clearly agree our study is not definitive. Additional research, in animals and in prospective studies in humans, is clearly warranted.”
• CONFERENCE AGENDA ANNOUNCED:
The highly-anticipated educational tracks for the 2015 R&D; 100 Awards & Technology Conference feature 28 sessions, plus keynote speakers Dean Kamen and Oak Ridge National Laboratory Director Thom Mason. Learn more. [10]