Perioperative Brain Health: Anesthesia Neurotoxicity

Animal Data

• Timeline for preclinical (animal) evidence for anesthesia-induced neurotoxicity:
  • The earliest evidence for anesthesia-induced neurotoxicity was published in 1999 when Ikonomidou et al. demonstrated that exposure to NMDA receptor antagonists induced widespread neuronal apoptosis in fetal and early neonatal rats.¹
  • Jevtovic-Todorvic et al. demonstrated the potential clinical relevance of this finding. They showed that exposure to other anesthetic agents, including isoflurane, nitrous oxide, and midazolam, was associated with neuroapoptosis and persistent learning and memory impairments in rats.¹
  • Subsequent preclinical studies indicated that exposure to all commonly used anesthetic agents can cause some degree of neuronal injury and neurobehavioral impairment across various animal models.¹
  • The transition to clinical (human) studies began in 2009 (see below).
• Summary of animal data:¹
  • The majority of animal studies have demonstrated persistent neuronal injury and neurobehavioral impairment after ante- or postnatal exposure to anesthetic agents. Early studies focused on deficits in learning and memory, while more recent studies have demonstrated behavioral changes, including delayed development, social withdrawal, and anxious, depressive, or hyperactive behaviors.
  • Nearly all anesthetic drugs, including volatile agents, nitrous oxide, ketamine, propofol, benzodiazepines, barbiturates, etomidate, and dexmedetomidine, have been studied and associated with adverse outcomes.
  • The effect of anesthetic agents is typically dose-dependent, with more neuronal injury and neurobehavioral impairment observed after longer or repeated exposures and higher doses of anesthetic agents.
  • These effects have been seen across multiple animal models, including rats, mice, sheep, rabbits, guinea pigs, piglets, and nonhuman primates.
  • Animal studies have also been used to investigate potential strategies to mitigate the neurotoxic effects of anesthesia, such as pharmacological interventions, environmental enrichment, and anesthesia protocols that minimize exposure or utilize alternative agents with potentially lower neurotoxicity.¹
• Limitations of animal studies:
  • Preclinical data have been criticized for several reasons, including difficulty translating relevant findings from animal to human models, differences in the timing of neurodevelopment in different species, magnitude and duration of anesthetic exposures, and a lack of physiologic monitoring and surgical stimulation during experiments.¹
  • SmartTots, a partnership between the US Food and Drug Administration (FDA) and the International Anesthesia Research Society, has proposed standards for ongoing animal research in this field, including a selection of appropriate animal models, anesthetic exposures, and neurobehavioral outcomes and more rigorous study design and data reporting requirements.²
• Key mechanisms by which anesthetics may cause neurotoxicity include apoptosis (programmed cell death), oxidative stress, and neuroinflammation.¹ These processes contribute to neuronal damage and cognitive deficits.

Human Data

FDA Response to Evidence³

• In reaction to the preclinical evidence, the FDA released a warning in 2016 about the possible risks of repeated or prolonged exposure (longer than three hours) to general anesthesia in children younger than three years of age and pregnant women.
• However, the FDA acknowledged that anesthetic and sedative drugs are often medically necessary and that untreated pain may also adversely affect neurodevelopment. Therefore, they advised that medical providers should discuss the risks and benefits of anesthetic exposure lasting longer than three hours to pregnant women and children younger than three years and that procedures should potentially be delayed when medically appropriate.
• The FDA has also required labeling changes for specific anesthesia and sedative drugs to include information about the potential risks of neurotoxicity in young children and pregnant women. These changes are intended to provide healthcare providers with important safety information when making treatment decisions.

Human Study Design

• Human studies on anesthesia-induced neurotoxicity in the developing brain have sought to determine whether the adverse effects observed in animal models also apply to children undergoing anesthesia.
• Human studies have primarily been observational studies and retrospective analyses of pre-existing data sets, as conducting controlled, prospective experiments in infants and children is challenging for several reasons, including⁴
  • inability to randomize children to anesthetic and surgical exposures;
  • difficulty in accounting for underlying comorbidities; and
  • the need to assess outcomes many years after exposure,
• Human studies may be limited by⁴
  • heterogeneity in anesthetic exposures, surgical procedures, and outcome measures;
  • use of assessment tools that are not the most sensitive measures for evaluation of neurodevelopment or assessment prior to the development of adverse outcomes;
  • lack of pre-exposure and perioperative clinical data, which may allow the inclusion of children with major clinical anomalies or intraoperative complications that can also affect neurodevelopment; and
  • limitations of observational and retrospective studies, including unmeasured confounding variables (i.e., socioeconomic status or other environmental variables), incomplete data, and recall bias.

Landmark Studies

• The three most cited studies of anesthesia-induced neurotoxicity in humans are:
  • GAS: General anesthesia versus awake regional anesthesia in infancy
  • PANDA: Pediatric Anesthesia Neurodevelopmental Assessment
  • MASK: Mayo Anesthesia Safety in Kids
• These studies prospectively assessed neurodevelopmental outcomes after exposure to anesthesia in infancy or early childhood. The primary outcome for all three studies was intelligence, but neuropsychological testing of other domains, including executive function and behavior, was also included.^5-7
• Outcomes
  • These studies showed no difference in intelligence after exposure to anesthesia (primary outcome). However, each study did show some effect on executive and behavioral outcomes.^5-7
  • When analyzed together, these studies show that a single general anesthetic exposure was associated with statistically significant increases in parent reports of behavioral problems without effects on general intelligence. These scores reflect worse behavioral scores, increased emotional distress, and impaired executive function associated with anesthetic exposure.⁸

Summary of Human Data

• Many studies have examined anesthesia-induced neurotoxicity in humans. The results have been variable, mainly due to issues already discussed, including the difficulty of creating well-designed trials and the heterogeneity in anesthetic exposure and outcome measures used in these studies.
• A systematic review conducted in 2021 examined 56 published articles to answer four predetermined study questions with the following conclusions:⁸

Future Directions

• Although animal studies have consistently shown adverse neurodevelopmental effects related to anesthesia exposure, human studies have been inconsistent. Further research into the potential neurotoxic effects of anesthetic agents, potential strategies to mitigate these effects, and vulnerable patient populations is warranted.
• Potential avenues for study include:
  • Identification of vulnerable patient populations: This may include children who have multiple, lengthy, or prenatal anesthetic exposures; children with congenital anomalies (particularly cardiac anomalies) or significant comorbidities; children of lower socioeconomic status; and children with genetic or environmental factors that predispose them to injury.⁴
  • Identification of intermediate outcomes or biomarkers: No consistent biomarkers of anesthetic-induced neurotoxicity have been identified. However, identifying a biomarker could lead to a better understanding of the mechanisms of neurotoxicity, improved translation from animal to human research, development of targeted protective strategies, and decreased latency from exposure to measurement of outcomes.⁴
  • Investigation of neuroprotective strategies that may mitigate neurotoxic effects of anesthetic agents, including:
    • Alternative anesthetic agents: Data from preclinical studies have shown that dexmedetomidine and opioids may have fewer neurotoxic effects than other sedative anesthetic agents. One ongoing study is the TREX (toxicity of remifentanil and dexmedetomidine) trial, which is being conducted by SmartTots. This study will compare a dexmedetomidine-remifentanil-low dose sevoflurane (0.3-0.6% end-tidal) anesthetic compared to standard sevoflurane (2.5-3% end-tidal) anesthetic in children <3 years old undergoing >2 hours of anesthesia.⁹
    • Pharmacologic agents, such as lithium, erythropoietin, and progesterone, may ameliorate neurodegenerative effects.⁹
    • Nonpharmacologic approaches to ameliorate neurotoxic effects.
• Preconditioning: Pre-exposure to subanesthetic doses may induce adaptive responses that protect against higher doses of anesthetics.¹⁰
• Environmental enrichment: Postanesthesia environmental enrichment and/or socialization can enhance neuroplasticity and functional recovery.^4,10
• Physiologic neuroprotection: This includes many of the strategies we already employ for neuroprotection under anesthesia. They include mild hypothermia, avoidance of hypoxia and hypercarbia, and maintenance of normal blood pressure and glucose levels.⁹