The Relationship between Early Exposure to General Anesthesia and Neurobehavioral Deficits

General anesthesia has become an integral part of surgical procedures in contemporary clinical practice. It not only greatly improves patient comfort during surgery by effectively reducing intraoperative pain but also opens up the possibility for surgeons to perform more complex and time-consuming procedures. Modern anesthetic protocols and strategies have been proven to be safe and recommended in many pathological conditions, but certain commonly used intravenous and inhalation general anesthetic agents were found to cause neuronal apoptosis and impair synaptogenesis in immature animals (including nonhuman primates), affecting brain development and leading to cognitive decline as well as a variety of long-lasting (or even permanent) behavioral deficits, which can severely impact normal life [1, 2]. Being very vulnerable at the critical stage of synaptogenesis, the brain is highly susceptible to general anesthesia [3]. The “window of vulnerability” in humans has been reported to be between the last third trimester of pregnancy and 2–3 years of age [4‒6]. The undesirable impact of general anesthesia imposed on the neurobehavioral development of these groups of populations is worth being studied.

With the development of society and the continuous improvement of medical care, a considerable number of children undergo general anesthesia each year in the USA alone [7], and given the population density of China was 4 times that of the USA, it was expected that the number of infants and children receiving surgical procedures under general anesthesia in China would be several times higher than that in the USA. Considering the large population, despite the current uncertainty about the long-term neurotoxicity of early exposure to general anesthesia, the US Food and Drug Administration (FDA) has warned that the multiple (two or more) or prolonged (more than 3 h) use of general anesthetics and sedative medications during surgical procedures in pregnant women in their last trimester of pregnancy, and children under the age of 3 years may affect the brain development of both the fetus and the child [8]. Meanwhile, guidelines from the UK and Ireland [9] and a statement from the European Institutions [10], though recommending avoiding unnecessary general anesthesia, do not suggest adjustments to clinical practice. Determining the risk of neurotoxicity from general anesthesia is therefore an urgent public health issue.

To date, clinical reports have generally shown that children exposed to general anesthesia during procedures at least bear moderate risk of neurodevelopmental impairment but do not show cognitive decline on intelligence tests compared with their peers not exposed to general anesthesia [11‒13]. However, the neurotoxic impact induced by general anesthesia can still be monitored through more comprehensive cognitive assessments during later childhood development. As more clinical investigations on general anesthesia have been conducted, we are beginning to observe that exposure to general anesthesia during infancy may exert adverse effects on children’s long-term behavioral development compared with those who have not experienced such exposure, ranging from internalizing behavioral problems (e.g., anxiety, depression, or somatization symptoms) to externalizing behavioral problems (e.g., impulsive and disruptive behaviors) which are exacerbated by impairments in executive functioning [14, 15]. Of greater concern, in addition to the neurobehavioral deficits observed, a growing body of research points to the possibility that early exposure to general anesthesia may also induce social and emotional deficits, such as attention-deficit hyperactivity disorder (ADHD) and autism spectrum disorder (ASD). Furthermore, a large body of preclinical evidence suggests that a variety of animals (from nematodes to nonhuman primates) exposed to general anesthesia during critical stages of brain development experience long-term neurobehavioral changes [2]. Thus, it is reasonable to hypothesize that exposure to general anesthesia at an early age may be associated with neurodevelopmental behavioral disorders in children.

This review retrospects a large number of clinical studies of early exposure to general anesthesia affecting neurobehavioral development in children, summaries the toxicological mechanisms of general anesthetics that have been closely associated with clinically significant neurobehavioral changes in basic research, and focuses on fine motor disability, ADHD, impulse control disorders (ICDs), ASD, and developmental coordination disorder. Five key neurobehavioral disorders are discussed, and future perspectives are presented to provide references for more rational and safer clinical use of general anesthesia in infants and young children.

The possibility that fetuses, infants, and young children exposed to general anesthesia may suffer from persistent neurotoxicity, which may induce long-term neurobehavioral disorders and developmental deficits, requires broad attention and in-depth research from both the academic and societal perspectives. We will explore the long-term effects of early exposure to general anesthesia on the neurobehavioral functions of children with five representative neurobehavioral disorders, including fine motor disability, ADHD, ICDs, ASD, and DCD.

Fine motor disability is the abnormality in the coordination of small muscle groups that hinders the body’ to perform tasks that require some degree of manual dexterity [16, 17]. Patients usually present with symptoms such as poor finger movements, impaired hand-eye coordination, and lack of accuracy in hand manipulation. The execution of fine motor skills requires the synergistic action of multiple brain regions, with even the simplest fine motor skills involving effective connections among the premotor cortex, motor cortex, cerebellum, basal ganglia, pyramidal tract, and peripheral nerves, not to mention visuospatial, sensory, and executive functions, which are more complex skills [18, 19]. Fine motor deficit usually appears as one of the symptoms of an underlying disease process and is rarely discussed as an independent disorder, but negative consequences it brings can affect the normal life of the child. Therefore, special attention should be paid to any risk factors that may induce fine motor deficit.

Given that a significant number of children undergo general anesthesia each year, whether general anesthesia has an impact on the long-term fine motor skills of children has become a widespread concern in the society. A series of clinical studies have been conducted to provide valuable references. In 2018, the Mayo Anesthesia Safety in Kids (MASK) study retrospectively analyzed nearly 1,000 cases, with changes in intelligence being the primary outcome, while behavioral changes being the secondary outcome. Evaluating intellectual impact in children predominantly relies on standardized full-scale intelligence quotient (IQ) scores derived from the Wechsler Intelligence Scale. In contrast, determining the deficits in fine motor skill necessitates a comprehensive neuropsychological assessment combined with parental reports to ensure a thorough evaluation. The study ultimately found that there were no significant differences in intelligence between children who received general anesthesia twice or more and those who experienced single or no anesthetic exposure before 3 years of age. Whereas, the multiple exposure group, as they grew into adolescence, showed a decrease in problem-solving speed as well as compromised fine motor skills comparing with none or single exposure group [12]. The conclusion drawn from the secondary indicators is preliminary and exploratory, requiring more accurate analytical methods to further validate. Therefore, the team conducted a more thorough and comprehensive statistical analysis of nearly 1,000 cases using factor analysis and cluster analysis in 2019. The results of the analysis remain consistent with the observation that early repeated exposures to general anesthesia does not lead to cognitive deficits but may cause impairments in specific behavioral patterns, such as reduced problem processing speed and fine motor skills [20]. In 2020, a cohort study fully assessed the neurodevelopmental outcomes of 13,433 children between the ages of 7 and 16 across multiple dimensions of learning, cognition, motor, social behavior, and language using dozens of different tests and scales. Similarly, the study revealed that children exposed to general anesthesia at an early age did not show significant differences in neurocognitive aspects but exhibited specific impairments in motor skills and fine motor abilities (hand dexterity) [21]. Another cohort study in the same year evaluated child development with the Japanese translation of the Ages and Stages Questionnaires, third edition (J-ASQ-3) and demonstrated that infants who underwent three or more general anesthesia exposure had neurodevelopmental delays at 1 year of age, which may result in a significantly increased risk of delays in the development of communicative, gross motor, fine motor, problem-solving, and interpersonal skills, especially gross motor skills [22]. The above clinical studies indicated that early exposure to general anesthesia may cause fine motor disability to some children.

A series of preclinical studies have also been initiated to delve into the mechanisms of general anesthesia-induced neurotoxicity on fine motor skills. Sevoflurane, as one of the most commonly used general anesthetic agents in pediatric surgery, is a preferred target for exploring the neurotoxicity of pediatric general anesthetics. Published in 2022, the study on the possibility that sevoflurane exposure may cause dendritic spine dysplasia and impaired fine motor ability in developing mice via the PI3K/AKT/mTOR pathway has attracted much attention. In this experimental investigation, neonatal C57BL/6 mice at 7 days of age received intraperitoneal injections 30 min prior to exposure to either 2% sevoflurane or 100% oxygen. Specifically, the mice were administered either the PI3K/AKT/mTOR activator IGF-1 or the mTOR phosphorylation inhibitor KU0063794. Subsequently, the mice were exposed to either 2% sevoflurane or 100% oxygen for 3 h. Within 6 h after the cessation of sevoflurane or pure oxygen exposure, the hippocampus was rapidly collected for subsequent Western blotting experiments to detect protein expression of PI3K/AKT/mTOR pathway phosphorylation. In addition to this, the team performed Golgi staining of the brain tissue of the mice at postnatal day 14 and observed the morphological features of the dendritic spine of hippocampal neurons using oil immersion microscopy. As the mice grew into adolescence (at postnatal day 48), fine motor skills were measured by the balance beam test. The results of the study illustrated that early exposure to sevoflurane reduced the proportion of fine dendritic spine and increased the proportion of mushroom-shaped dendritic spine but did not change the density of dendritic spine. In addition, early sevoflurane exposure inhibited phosphorylation of the PI3K/AKT/mTOR pathway in the hippocampus of young mice, ultimately leading to long-term fine motor disability. This impairment can be rescued by pretreatment with IGF-1, whereas pretreatment with KU0063794 exacerbated the damage [23]. In conclusion, this study investigated the role of changes in dendritic spine plasticity in sevoflurane-induced neurodevelopmental toxicity and provided new insights into the neurotoxic effects of sevoflurane. A study published in 2021 explored general anesthetics and mechanisms of postoperative long-range fine motor injury using juvenile nonhuman primates and rodents, and for the first time, highlighted the effects and mechanisms of m6A methylation on fine motor injury in sevoflurane anesthesia [24]. m6A modification is the most abundant form of RNA methylation modification, and YTHDF1 is one of the recognition proteins for m6A methylation and has been testified to be involved in regulating the formation and development of neurocognition [25‒27]. This study found that the expression of YTHDF1 was significantly downregulated in the brains of juvenile primates and rodents after sevoflurane anesthesia. Furthermore, single-cell sequencing results manifested that the decline in YTHDF1 expression was most pronounced in sp8-positive interneurons. The results of RNA-binding protein immunoprecipitation assay and m6A-seq assay showed that m6A in the synaptophysin mRNA where existed YTHDF1-binding sites was highly enriched. Considering that the main function of YTHDF1 is to recognize methylation sites on RNA, a strong correlation between synaptophysin and neurodevelopmental toxicity of general anesthetic agents was found in previous studies [28], and the overexpression of YTHDF1 rescued sevoflurane-induced fine motor dysfunction, cognitive deficits, and changes in synaptophysin in young mice was detected in this study, it was elucidated that YTHDF1 regulated the expression of its downstream target gene synaptophysin in an m6A methylation-dependent manner, and early sevoflurane exposure interfered with the interaction of YTHDF1 with the downstream genes, which led to fine motor skill deficits [24]. This finding opened up a whole new area of research to unveil the potential mechanisms of the effects of general anesthesia on the neurodevelopment of the young brain.

The integrity of the myelin sheath is a key structure for the brain’s processing speed and motor coordination function. Thus, developmental defects in myelin sheaths will likely result in fine motor deficits. As mentioned earlier, a series of clinical studies have found that early exposure to general anesthetics was associated with postoperative fine motor dysfunction long after the surgery. Consequently, the hypothesis that general anesthetics inducing developmental defects in myelin sheaths may lead to fine motor deficits was raised. In 2019, the concept of myelin developmental toxicity of general anesthetics was first introduced in a study of the neurodevelopmental toxicity of general anesthetics. The study found that young rhesus monkeys and young rats exposed to sevoflurane both had reduced expression of Ermin protein-coding gene (ERMN), a key gene for myelin development, in the prefrontal cortex. The result was validated in the classical model of rats exposed to sevoflurane. The myelin development of those rats was found to be impaired, and they showed long-term neurobehavioral changes [29]. This finding attracted widespread attention, setting off a wave of research into the effects of general anesthetics on myelin development [30‒32], which have brought the concept of myelin developmental toxicity of general anesthetics to the forefront. Meanwhile, the idea that well- or poorly-developed myelination in the brain is closely linked to fine motor skills has also been widely discussed [33, 34]. For example, in 2021, Cullen et al. [34] put forward that mice with conditional knockout of the Kif3a gene in oligodendrocyte (OL) precursor cells demonstrate fine motor dysfunction, manifested by a significant reduction in stride length and an increase in step frequency during gait analysis. Additionally, these mice exhibit increased error counts in specific balance and coordination tasks. OL precursor cells have the capacity to proliferate and differentiate into mature OLs, which migrate along axons to form myelin sheaths. Kif3a knockout mice have impaired primary cilia assembly and OL production, leading to defective myelin development and ultimately fine motor coordination [34]. Taken together, these clinical and preclinical studies are important in exploring the relationship between early general anesthesia exposure and long-term fine motor deficits, as well as possible underlying mechanisms.

ADHD, also known as mildly dysfunctional brain syndrome, is a neuropsychological developmental disorder characterized by hyperactivity, impulsivity, inattention, lack of vigilance, and an inability to adapt to rapid change [35]. It has a fairly high prevalence rate. Approximately 6%–9% of children and adolescents were affected globally, with a prevalence of 5.29% in population aged under 18 [36], among which approximately 50%–80% have ADHD symptoms till their adulthood, manifesting severe behavioral and psychiatric problems. It is therefore crucial to identify people with ADHD timely and provide necessary supports and treatments to reduce the burden on patients, their families, and the society.

The idea that there was a link between early exposure to general anesthesia, and the incidence of ADHD was first proposed 12 years ago when Sprung et al.’s [37] team found that children exposed to general anesthesia multiple times (two or more times) in the early years were at a three-fold increased risk of developing ADHD, while there was no significant association between a single exposure to general anesthesia and the development of ADHD. Similar conclusions were drawn from another paper published in 2015. Through Cox regression analysis of data collected from 1,146 children who had received general anesthesia and 3,438 participants in age- and gender-matched control group, the study evidenced that the risk of being diagnosed with ADHD in children who were exposed to general anesthesia once and for less than 3 h did not increase. However, for children who were exposed to general anesthesia multiple times and for more than 3 h, the risk of developing ADHD in the future significantly increased [38]. Nevertheless, the results of numerous studies on the association between single general anesthesia exposure and the subsequent development of ADHD have been inconsistent. Studies with large sample sizes have demonstrated an increased risk of future development of ADHD in children exposed to general anesthesia on a single occasion [39], whereas studies with small sample sizes have found no such increased risk [40, 41]. The difference may lie in the smaller risk ratio. Specifically, only when the sample size of the study was large enough can the relevance between general anesthesia and ADHD be clearly determined. Therefore, a detailed analysis of studies with large sample sizes would be provided. An article published in 2017 examined the relationship between exposure to surgery and anesthesia at a young age and the diagnosis of mental disorders in children. It conducted survival analysis on 38,493 children who were exposed to general anesthesia and surgery at least once, and 192,465 children who have not been exposed to general anesthesia before the age of 5, with two groups being matched in age and gender. Ultimately, the result indicated that children who underwent surgery and anesthesia before the age of 5 had a slightly increased risk of psychological disorders, including ADHD, but the increased risk was not dependent on the specific timing of the surgery and anesthesia [39]. Another study similarly noted that children undergoing general anesthesia before the age of 5 years were at an increased risk of developing ADHD and that multiple general anesthesia exposures further intensified this potential risk. It also found that race was a key moderator of this risk that regardless of the age of first exposure, non-white children being more likely to develop ADHD after general anesthesia than white children [42]. A recent publication filled the research gap in children aged 0–3 years. Using a population-based cohort study and propensity score matching methodology, the study aimed to investigate the potential association between general anesthesia exposure and the development of ADHD in children between 0 and 3 years of age. The result revealed that exposure to general anesthesia before the age of 3 years significantly raised the risk of developing ADHD in the future. This finding was consistent with the results of the aforementioned study, which used five-year-old children as subjects. Meanwhile, the study identified several risk factors for ADHD, including times of exposure to general anesthesia, duration of exposure, gender, and central nervous system surgery [43]. In addition to the clinical studies reviewed above, an experimental animal study reported that early ketamine exposure caused hyperactive behavior similar to ADHD, which further supported the idea that repeated exposures to general anesthesia in children was associated with the development of ADHD [44]. Therefore, in clinical practice, appropriate general anesthesia strategies, preventive measures, and follow-up measures should be adopted for children undergoing surgical procedures that include general anesthesia. Furthermore, the potential mechanisms by which early exposure to general anesthesia leads to ADHD should be elucidated and ways to minimize the possible threat posed by general anesthesia should be explored.

From a molecular genetic perspective, ADHD is one of the most heritable psychiatric disorders. Children whose parents have ADHD are subjected to a two- to eight-fold increased risk of developing ADHD, and siblings of children with ADHD also face an increased risk of ADHD [45]. Whereas the specific genetic mechanisms of ADHD may involve interactions between multiple genes and environmental factors, which are complex and remain to be disclosed. Faraone’s team has been dedicated to exploring the molecular genetics of ADHD and has identified candidate genes associated with ADHD, such as DRD4, DRD5, SLC6A3, SNAP-25, HTR1B, DBH, MAOA, SLC6A2, TPH-2, SLC6A4, CHRNA4, GRIN2A, and ADRA2A. Despite their weak effects, these candidate genes confirmed that genetic susceptibility to ADHD stemmed from the combined action of multiple small effector genes. These candidate genes contained those associated with dopamine, norepinephrine, 5-hydroxytryptamine-related receptors and transporter proteins, brain-derived neurotrophic factors, and nicotinic acetylcholine receptor α4 subunit genes, constructing a map of genes related to the regulation of neurotransmitter pathways associated with ADHD [46]. Nonetheless, the current understanding of the molecular genetics of ADHD is still limited, and studies with more robust statistical methods to detect small effector genes, large-scale collaborative studies, and genome-wide association studies can be conducted to further uncover the molecular genetics of ADHD.

From a molecular biology perspective, the pathogenesis of ADHD involves the central dopamine neuronal pathway, the noradrenergic neuronal pathway, the serotonergic neuronal pathway, the glutamatergic neuronal pathway, as well as brain-derived neurotrophic factors. Numerous genetic studies have indirectly substantiated the reliability of these systems’ association with ADHD [46]. Recent studies have shown that abnormalities in the regulation of the neural network comprising the frontal cortex of the brain and its downstream nervous system were closely linked to the development of ADHD [47]. In the neural network, the critical functions of dopaminergic neurons in the central nervous system were implied through 5 different dopamine receptors. Damage to specific brain regions, such as the ventral tegmental area of the midbrain, can result in diminished dopamine neurotransmission within the prefrontal cortex. This reduction in neurotransmission can lead to an increase in dopamine receptor sensitivity, potentially triggering ADHD. Some studies have revealed that different dopamine receptor subtypes were closely linked to the pathogenesis of ADHD. A deeper understanding of the relationship between these five different dopamine receptors and ADHD will help elucidate the different roles of these receptors and inspire more therapeutic approaches for ADHD. In 2012, Wu et al. [48] published an article on the role of dopamine receptors in ADHD, which pointed out that the activity of D2 and D3 receptor subtypes was increased in the striatum of adolescents with ADHD, and D2, D4, and D5 receptor subtypes were closely related to the onset of ADHD. Methylphenidate and amphetamine are currently used clinically to treat ADHD, and they act by stimulating D1 receptors in the prefrontal cortex and D2 receptors in the striatum. In addition to clinical studies, an animal study also found that mice with conditional knockout of circadian clock genes exhibited decreased activity of dopaminergic neurons in the ventral striatum, hypothalamus, and prefrontal cortex. These mice displayed behavioral manifestations similar to ADHD. Furthermore, it was observed that damage to dopaminergic neurons exacerbated symptoms of ADHD [49]. This finding further testified the involvement of central dopaminergic neuronal pathways in the pathogenesis of ADHD. Recent studies have illustrated that the central dopamine pathway was involved in general anesthesia. Therefore, early exposure to general anesthesia may lead to alterations in the structure and function of the central dopaminergic system, which may predispose patients to ADHD. The conclusions drawn from this study were actually in line with similar propositions put forth as early as 2008 [50], suggesting that dysregulation of the catecholamine system may lead to ADHD. ADHD was considered a consequence of abnormality in the brain’s frontal lobe regions, particularly the prefrontal cortex and associated subcortical structures and circuits. The underlying anomalies were attributed to disruptions in catecholamine neurotransmission. Patients with ADHD exhibited decreased levels of dopamine and norepinephrine, primarily due to impairments in their respective neurotransmitter transport systems. The efficacy of catecholaminergic stimulants further confirmed that neurotransmitter abnormalities in ADHD patients primarily stemmed from the catecholamine system. This system comprises the dopamine system as well as the norepinephrine and adrenaline systems, which aligns with the implications of the gene map related to neurotransmitter pathway regulation associated with ADHD mentioned earlier [50]. Besides, abnormalities in glutamatergic pyramidal neurons, a major cellular component of the prefrontal cortex and associated subcortical structures, are often cited as a cause of ADHD. Prefrontal cortex exerts its functions largely on glutamatergic neurotransmission. In 2017, an article stated that the impairment of AMPA receptors in pyramidal neurons in the prefrontal cortex of animal models of ADHD led to dysfunctional glutamatergic transmission and ultimately to behavioral abnormalities. Methamphetamine can normalize ADHD-like behavior by improving AMPA receptor function and activating pyramidal neurons in the prefrontal cortex. These findings demonstrated that dysfunction of AMPA receptor in the prefrontal cortex may be responsible for adolescent ADHD behaviors, and enhancing the prefrontal cortex activity may be one of the therapeutic strategies for ADHD [51]. The neurotoxic effects of general anesthesia cause impaired neuronal development in the prefrontal cortex, resulting in cognitive and behavioral developmental deficits. It is therefore reasonable to speculate that AMPAR may be involved in the pathogenesis of ADHD from general anesthesia exposure. Another study manifested that ketamine and other NMDAR antagonists can induce ADHD-like behaviors in mice, accompanied by alterations in dendritic spine morphology in the striatum [52]. Symptoms of ADHD in mice can be altered by the activation of pyramidal NMDAR [51]. Both AMPAR and NMDAR belong to the family of glutamatergic receptors, indicating that the glutamatergic system may lie behind the pathogenesis of general anesthesia exposure-induced ADHD. In addition to this, BDNF is also closely related to the development of ADHD. The formation and plasticity of neuronal spine may underlie the pathogenesis of ADHD and are regulated by brain-derived neurotrophic factors [53]. During the development of the frontal lobe of the brain, the reduction of BDNF may induce ADHD [54]. Moreover, the severity of ADHD is strongly associated with reduced BDNF concentrations, especially in the midbrain [55]. Anesthesia may diminish BDNF, whereas methamphetamine increases plasma BDNF concentrations in children with ADHD [56]. Therefore, general anesthesia may weaken the expression of BDNF and interfere with the normal process of synaptogenesis, thus inducing ADHD. This conclusion provides us with important clues for further investigation of the pathological mechanisms of ADHD. Taken together, these clinical and preclinical studies are significant for combing the correlation between early exposure to general anesthesia and ADHD, as well as possible underlying mechanisms.

ICDs are seldom discussed as an independent disorder and are often considered as clinical diagnostic and pathophysiological markers of some neurobehavioral disorders, such as ADHD, borderline personality disorder, obsessive-compulsive disorder, trichotillomania, pathological gambling, and chronic substance abuse [57, 58]. Therefore, the research on ICDs is often in conjunction with ADHD. ADHD, a chronic neurodevelopmental disorder, is one of the most prevalent neurobehavioral disorders in children, presenting a growing global concern. The core symptoms of ADHD include inattention, hyperactivity, and impulsivity [35], with impulsivity being a critical symptom for diagnosing the disorder. Impulsivity is characterized by difficulties in inhibiting behaviors and delayed aversion [59], such as impulsive responses before task completion, challenges in waiting for turns, or a tendency to interrupt and disturb others. Adolescents and adults with ADHD-like impulsive behaviors have been found to have a significantly increased propensity to engage in undesirable behaviors such as drug addiction, pathological gambling, and suicide, resulting in significant negative consequences for the individual, family, and society [60].

Currently, a multitude of clinical investigations have highlighted a correlation between early repeated exposures to general anesthesia and the subsequent development of ADHD, but few studies concentrated specifically on the relationship between early exposure to general anesthesia and ADHD-like impulse control deficits. Given the potential implications for public mental health and social stability, it is crucial to establish the connection between early exposure to general anesthesia and ADHD-like ICDs and further explore the underlying mechanisms.

As the most widely used general anesthetic in clinical practice for infants and young children, sevoflurane exerts its efficacy by both inhibiting neuronal activity through its action on gamma-aminobutyric acid (GABA) receptors and attenuating neuronal excitatory effects via the reduction of NMDA receptor activity [61]. Multiple studies have demonstrated that early exposure to sevoflurane may engender synaptic alterations and neuronal morphological abnormalities in various brain regions during development, particularly in the prefrontal cortex [1]. In 2012, Briner et al. [62] put forward that general anesthetics could increase dendritic spine density in the prefrontal cortex of rats during the “synaptogenesis” process. During critical developmental periods, neurons in the prefrontal cortex are particularly vulnerable to the neurotoxic effects induced by sevoflurane exposure [63], and these effects are enduring [64]. Of further significance, current research has indicated the involvement of the prefrontal cortex in the regulation of impulsive behavior [65], with patients exhibiting ADHD-like impulsive behaviors often displaying hyperactivity in the prefrontal cortex [66]. The prefrontal cortex integrates input from various brain regions, encodes and regulates higher cognitive functions, including inhibitory responses and impulse control, to achieve corresponding physiological functions [67, 68]. The normal functioning of the brain relies on a delicate dynamic balance between excitatory and inhibitory signals. Any disruption in this balance within the neural circuits of the brain during development can easily precipitate related neurological disorders. In 2020, Xie et al. [64] investigated the induction of ADHD-like impulsive behavior in adult mice following early repeated exposures to general anesthesia, along with the underlying mechanisms. In their study, postnatal day 6 (a critical period for brain development) C57 mice were exposed to 3% sevoflurane for 2 h per day over 3 consecutive days. Mice in the control group were exposed to air at an equivalent flow rate, while all other experimental conditions remained consistent with the sevoflurane group. Subsequently, the cliff avoidance reactions of 8-week mice were tested to assess their impulsivity. The results revealed that compared to mice in the air control group, those in the sevoflurane group exhibited a significant increase in the number of cliff jumps during the testing period, along with a significant reduction in the latency to the first cliff jump. Moreover, in order to eliminate impulsive behavior caused by anxiety or fear, the research team performed elevated plus maze tests (EPM) following the cliff avoidance reaction tests. No significant differences were detected between the two groups in terms of either the time spent exploring the open arms or the number of entries into the open arms. Additionally, compared to mice in the air control group, those in the sevoflurane group exhibited shorter durations of freezing behavior, indicative of fear, during the EPM tests. This suggested that early repeated exposures to sevoflurane after birth can induce ADHD-like impulsive behavior in adulthood. Furthermore, the study utilized immunofluorescence staining experiments to further investigate whether the medial prefrontal cortex region was involved in sevoflurane-induced ADHD-like impulsive behavior. The experimental results revealed a significant increase in c-Fos signal expression in the medial prefrontal cortex region of mice in the sevoflurane group compared to the air control group, with a greater number of co-localized cells showing both c-Fos signal and excitatory neuron marker (CaMKIIα+) signal. This proved that the ADHD-like impulsive behavior induced by early sevoflurane exposure was mediated by excitatory neurons in the prefrontal cortex region. Finally, the team utilized designer receptors exclusively activated by designer drugs technology to activate excitatory neurons in the prefrontal cortex region of mice in the air control group and simultaneously inhibit excitatory neurons in the prefrontal cortex region of mice in the sevoflurane group. The results demonstrated that mice with activated excitatory neurons in the prefrontal cortex exhibited ADHD-like impulsive behavior, while inhibiting excitatory neurons in the prefrontal cortex alleviated the ADHD-like impulsive behavior induced by sevoflurane exposure. This further suggests a close association between the activity of excitatory neurons in the medial prefrontal cortex region and the ADHD-like impulsive behavior induced by sevoflurane exposure. In summary, this study manifested that early repeated exposures to sevoflurane may disrupt the development of medial prefrontal cortex neurons, leading to an increased ratio of excitatory to inhibitory signals in the medial prefrontal cortex neurons area, thereby triggering ADHD-like impulsivity behaviors [64]. The research undertaken by Xie’s team represents a rare focus on early-life general anesthesia exposure and its relationship with ICDs, pioneering an examination into the functional involvement of the medial prefrontal cortex in ADHD-like impulsive behaviors and filling the current research gap. However, the medial prefrontal cortex is intricately connected to numerous neural circuits characterized by complex afferent and efferent projections, which have been inextricably associated with the pathogenesis of psychiatric disorders [69]. The precise neural circuits underlying impulsive behavior control disorders remain elusive. As a consequence, future investigations may use chemogenetics or optogenetics techniques to reveal the interactions between medial prefrontal cortex projection neurons and downstream target brain regions, elucidating potential pathogenic pathways behind ADHD-like impulsive behaviors.

In addition, it is also evident that both dopamine and serotonin systems play indispensable roles in ADHD-like impulsive behaviors. While the noradrenergic system is implicated in the onset and treatment of ADHD, its regulatory role in impulsive behavior remains poorly understood. Methylphenidate, the most commonly prescribed medication for ADHD, acts through the modulation of monoaminergic neurotransmitter imbalances, particularly dopamine or norepinephrine dysregulation. Most studies have substantiated that atomoxetine has a significant impact on impulsive behavior tests in rodents, such as the stop-signal task, delay discounting task, and 5-choice serial reaction time task. Yet, the specific neural mechanisms involved remained to be unearthed [70, 71]. This observation superficially reinforced the intricate relationship between methylphenidate’s dopaminergic modulation and impulse regulation. Beyond mere pharmacological inquiry, several clinical investigations have unveiled anomalies in dopamine transporter function and dopamine dehydrogenase activity within the prefrontal cortex of individuals with ADHD [72, 73]. Additionally, supporting evidence from animal models underscores the significant correlation between the dopamine system and impulsive behaviors. Both dopamine transporter-deficient mice (DAT KO) and rats subjected to neonatal 6-hydroxydopamine lesions demonstrate impulsive traits across behavioral paradigms [74]. Collectively, clinical and preclinical findings highlighted the irreplaceable contribution of the dopamine system to ADHD-like impulsivity. Similarly, the serotonin system emerges as a pivotal player in ADHD-like impulsivity. On the other hand, Daniel and his colleagues investigated the impact of reduced levels of serotonin (5-HT) in the central nervous system following rapid tryptophan depletion (RTD) on impulsive aggressive behavior in boys with ADHD. The study employed a double-blind crossover design, where 22 male adolescents with ADHD underwent RTD testing on 1 day and were given a tryptophan-balanced placebo on another day. Competitive reaction time games were used within 270 min after ingesting the RTD/placebo to stimulate aggressive behavior. The results revealed that RTD significantly increased aggression in individuals with low impulsivity [75]. Furthermore, investigations reflecting the therapeutic efficacy of selective serotonin reuptake inhibitors in ADHD treatment have yielded equivocal outcomes, elucidating serotonin’s nuanced role in ADHD etiology. However, selective serotonin reuptake inhibitors demonstrate therapeutic efficacy in alleviating other ICDs such as intermittent explosive disorder [76]. Notably, substantial evidence supports that 5-HT modulates DA levels, and their interaction alters the equilibrium between them, inducing different levels or types of impulsive behavior. This may also be one of the key factors in the neurochemical basis of ADHD [77]. In summary, the intricate interplay between the dopamine and serotonin systems emerges as a cornerstone in elucidating the mechanisms underpinning ADHD-like impulsive behaviors. Taken together, these related studies serve as valuable resources for exploring the correlation between early exposure to general anesthetics and ADHD-like impulsive behavior, as well as possible underlying mechanisms.

ASD is a neurodevelopmental disorder characterized by deficits in social communication, restricted interests, and repetitive behaviors [78]. According to the World Health Organization, the estimated international prevalence of ASD was 0.6% [79]. The steadily climbing prevalence over the past decades has garnered widespread attention globally.

ASD is a highly heritable disorder. Genetic factors play an undeniable role in the etiology of ASD, but the phenotypic expression of genetic susceptibility shows considerable heterogeneity among individuals with ASD [80‒82]. Moreover, in specific patient populations, genetic factors may be modulated by obstetric and perinatal factors as well as early neonatal environmental exposures [83]. Therefore, the correlation between cesarean section delivery and ASD has become a direction of academic interests. In addition, cesarean section is a surgical procedure used in childbirth to prevent risks of compromising the health of the mother or baby during the process of vaginal delivery, and the global incidence of cesarean section takes a mounting trend, with an estimated annual growth of 4% [84]. The proportion of cesarean section in China has rapidly increased to approximately 40% [85]. All these add to the imperative to clarify the relationship between cesarean section and ASD. A meta-analysis indicated that cesarean section was associated with a 23% increased risk of ASD compared to vaginal delivery [85]. Regional anesthesia is often the preferred choice for cesarean section surgeries, while general anesthesia is primarily used for emergency deliveries or in cases where there is a maternal preference, insufficient regional anesthesia, or failed attempts at regional anesthesia [86]. Whereas, in recent years, due to elevated risks to the mother, the frequency of using regional anesthesia in cesarean section surgeries has been decreasing [87]. Meanwhile, more animal experiments showed that several general anesthesia techniques can lead to abnormal apoptosis of neurons and behavioral defects. Investigating the correlation between general anesthesia and the risk of ASD has been particularly necessary. In 2015, Chien and colleagues [88] evaluated the subsequent risk of ASD in newborns under general anesthesia during cesarean section by comparing it with the risk of ASD under natural delivery or regional anesthesia. The results showed that compared to vaginal delivery, newborns delivered via cesarean section under general anesthesia bore a 52% increased risk of developing ASD. Additionally, the incidence of autism was higher in them compared to those delivered via cesarean section with RA. Therefore, the use of general anesthesia during cesarean section was associated with an increased risk of neurodevelopmental disorders in children, such as ASD. It is worth noting that this study did not demonstrate a causal relationship between newborns delivered via cesarean section under general anesthesia and the subsequent risk of developing ASD; it only showed a statistical association between them [88]. In 2021, Yang and colleagues [89] conducted a study examining the relationship between different anesthesia methods and ASD, revealing a potential association between general anesthesia and increased risk of ASD. The study employed multivariable conditional logistic regression analysis to process data from 950 ASD patients and 764 controls. After adjusting for confounding factors, the team found that newborns delivered via cesarean section with general anesthesia had a significantly higher risk of developing ASD compared to those delivered via cesarean section with regional anesthesia. This suggests that, after considering other potential influencing factors, general anesthesia may be an independent risk factor for ASD [89]. While there may be conflicting findings in the above studies, they provided valuable insights into the complex relationship between exposure to general anesthesia during pregnancy and the subsequent development of ASD in children, and the possible guidance for future clinical practices aimed at lowering the potential risk of ASD.

In addition to obstetric and perinatal factors influencing ASD, early neonatal exposure may also modulate ASD development. Among these, exposure to general anesthesia is one of the common exposures in newborns. Current research suggests that early childhood exposure to general anesthesia may lead to neurodevelopmental disorders such as ASD. Nevertheless, this potential link was rarely studied, and the results were not convincing due to small sample sizes [90, 91]. In 2023, Pikwer et al. [92] conducted a large-scale study on the relationship between early childhood exposure to general anesthesia and the risk of ASD. This study represents the first population-based research of its kind, encompassing 401,750 children aged 0–5 years who underwent general anesthesia, and a matched control group of 1,187,796 children. After the thorough consideration of potential confounding factors, the study revealed that the risk of developing ASD in children exposed to general anesthesia early in life was nearly doubled compared to those not exposed to general anesthesia. Additionally, it was observed that children exposed to general anesthesia but not undergoing surgery had a higher risk of developing autism subsequently compared to those who received both general anesthesia and surgery. This implied that general anesthesia may represent an independent risk factor distinct from the surgical procedure itself. Furthermore, the study found that children exposed to both general anesthesia and surgery still had a significantly elevated risk of developing ASD compared to those who did not receive them. This indicated a significant association between early exposure to general anesthesia and the subsequent development of ASD, but the association is statistical instead of casual and the absolute risk is small [92]. Whether there are other unknown factors influencing this outcome requires further observational and research efforts.

ASD is a complex neurodevelopmental disorder with a wide range of risk factors. Despite the cumulative research on ASD in recent decades [93], a precise consensus regarding its core pathogenic mechanisms and effective therapeutic approaches has yet to be established. Among various hypotheses in ASD research, the genetic hypothesis has been widely recognized. Nevertheless, ASD is still believed to be a multifactorial and complex disease, and that in addition to the interplay of genetic and environmental factors, various mechanisms such as synaptic developmental disorders and dysregulation of the excitatory/inhibitory (E/I) balance in the brain may also be involved in the pathogenesis of ASD. Regarding the synaptic developmental disorder hypothesis, numerous genetic studies have indirectly affirmed its reliability. These studies have found associations between frequent variant genes in ASD and aspects such as synaptic structure establishment, neurotransmitter receptor function, and synaptic plasticity regulation. For example, as early as 2011, the team led by Gilman identified a large ASD-related functional gene network that primarily comprises genes involved in synaptic formation and function, including IP3R2, NRXN1, NLGN3, and the SHANK protein family [94]. Additionally, recent preclinical studies have found alterations in astrocytes in the pathophysiological processes of ASD [95, 96]. As one of the primary cell types in the central nervous system, astrocytes play crucial roles in supporting neural network construction, participating in synaptic formation and removal, maintaining neuronal function, and stabilizing the brain’s internal environment. An increase in Ca²⁺ serves as a hallmark of astrocyte activation, a process primarily driven by the second type of 1,4,5-triphosphate receptor (IP3R2). Interestingly, the IP3R2 gene has long been found to exhibit abnormalities in patients with ASDs [94]. In 2021, Wang et al. [97] probed the relationship between calcium signaling pathways in astrocytes and ASD-like behaviors through methods such as astrocyte-specific IP3R2 conditional knockout mouse models and drug interventions. Their study found that mice with conditional knockout of IP3R2 specific to astrocytes exhibited atypical social and repetitive behaviors. To further delve into the neurobiological mechanisms, the team utilized in vivo microdialysis techniques to analyze the levels of neurotransmitter transmission in the medial prefrontal cortex of rodents involved in social behavior and observed significantly reduced ATP levels in knockout mice. Additionally, treatment with ATP mediated by P2X2 receptors can rescue social behavior in mice [97]. From this, it could be seen that impaired calcium release led to astrocyte activation failure, synaptic changes, and subsequently, the manifestation of ASD-like behaviors in mice. In 2019, Zhou et al. [98] investigated the specific mechanistic pathways by which sevoflurane, a representative general anesthetic for children, causes astrocyte damage and consequently behavioral disorders. This study indicated that sevoflurane exposure had both immediate and long-term disruptive effects on Ca²⁺ homeostasis in developing astrocytes, including reducing basal Ca²⁺ levels and altering spontaneous and neurotransmitter-induced Ca²⁺ signals in astrocytes. Moreover, changes in Ca²⁺ signaling led to downregulation of ezrin protein expression, affecting the morphogenesis of astrocytes, especially in the development of fine processes [98]. Ezrin, as a membrane-associated protein binding to actin, plays a critical role in the fine processes of astrocytes [99, 100]. The morphological changes in astrocytes are associated with synaptic overgrowth and functional impairment in the cortical regions of the brain, affecting the interaction between astrocytes and neurons. This has a negative impact on the stability and maturation of synapses, ultimately resulting in social behavioral deficits in mice. Further research revealed that specific enhancement of ezrin protein expression in astrocytes can mitigate the morphological and functional damage caused by sevoflurane exposure and completely correct behavioral defects in developing mice [98], suggesting that astrocytes may be an important target for general anesthetics, and their morphological integrity is crucial for synaptic formation and neural behavioral development. Additionally, there is evidence that early exposure to general anesthesia can lead to social impairments, which are core symptoms of ASD [101]. To sum up, early exposure to general anesthetics disrupts Ca²⁺ homeostasis in astrocytes, leading to downregulation of ezrin protein expression, which subsequently affects the morphology and function of astrocytes, alters synaptic formation, and may ultimately result in ASD-like behaviors. This discovery presents novel avenues for the neurodevelopmental and behavioral disturbances induced by anesthetics.

As mentioned earlier, GABA is the most common inhibitory neurotransmitter, while glutamate is the most common excitatory neurotransmitter. The two maintain a delicate balance in the brain, collectively sustaining the E/I balance of the central nervous system to support normal neural network function and behavioral performance. Disruption of this balance may induce serious behavioral disorders. Recent research illustrated that E/I imbalance may be a key factor underlying ASD characteristic behaviors and cognitive abnormalities [102‒104]. In related studies, abnormal expression of glutamate and GABA receptors in the postmortem brains of ASD patients has been observed [105]. Changes in the levels of glutamate and glutamine in the cortex and basal ganglia of ASD children, as well as in the basal ganglia of adult patients, using proton magnetic resonance spectroscopy have also been identified [106‒109]. Additionally, reductions in GABA levels in several brain regions of children with ASD have been reported [110, 111]. While studies supported the hypothesis of E/I imbalance in ASD, the heterogeneity of the E/I ratio revealed by them was not always consistent, and the biological mechanisms behind the imbalance remain unclear. In 2018, Horder et al. [112] investigated abnormalities in glutamate and GABA in ASD. The study compared glutamate and GABA levels in the striatum of adult human ASD patients and 6 different rodent models of ASD using MRS technology. Reduced glutamate concentration in the striatum of adult ASD patients was inspected, which was correlated with the severity of social symptoms. Meanwhile, reduced glutamate concentration in the striatum was also observed in certain rodent models [112]. The findings of this study supported the hypothesis of E/I imbalance in ASD and suggested that the reduction of glutamate in the striatum may be one of the key pathological mechanisms of ASD. The association between general anesthesia and ASD can be inferred from Satomoto et al.’s [113] report in 2009 that neonatal exposure to sevoflurane may lead to social behavioral abnormalities, such as ASD. Briefly, general anesthetics may lead to overactivation or impairment of GABAergic neurons, resulting in E/I imbalance, which in turn can cause social behavioral abnormalities such as ASD.

In a nutshell, the genetic hypothesis, the synaptic developmental disorder hypothesis, and the E/I imbalance hypothesis interact with each other and collectively contribute to the pathophysiological processes of ASD. This aligns with the widely accepted view that ASD is a complex disorder caused by multiple factors.

DCD is a neurobehavioral developmental disorder characterized by slow, awkward, or inaccurate movements, as well as learning difficulties (new motor tasks or adapting previously learned gestures to modifications or additional constraints), which severely affects a child’s ability to perform coordinated movements [114]. The prevalence of DCD is estimated to be around 4.3% [115], and such high prevalence has aroused worry and concern among families and society.

Currently, the diagnosis of DCD is based on abnormal performance in manual dexterity, ballistic motor skills deficiencies, and balance skills deficit [114]. Recent studies have shown that abnormal brain function in children with DCD involves multiple brain regions, including but not limited to the cerebellum, basal ganglia, parietal lobe, and frontal lobe [116]. As a critical stage of brain development, infancy and early childhood are highly susceptible to general anesthetics which may ignite a range of cognitive and behavioral developmental disorders [3]. Clarification of the correlation between early exposure to general anesthesia and DCD has become requisite.

More than 2 decades ago, Ivry’s [117] team had already identified cerebellar injury as an important cause of DCD that was intimately linked to uncoordinated behavior, clumsiness, and postural control. This finding provided insight into the pathogenesis of DCD. Five years later, Zwicker et al.’s [118] team further verified the pivotal function of the cerebellum in DCD by reviewing and comparatively analyzing data from relevant animal and human studies. It has been suggested that early exposure to general anesthesia may lead to neuronal degenerative necrosis in several brain regions, such as the cingulate cortex, hippocampus, thalamus, and parietal lobe, but the effects on the cerebellum have not been clearly explicated [119]. Therefore, whether early exposure to general anesthesia provokes apoptotic degeneration of cerebellar neurons and the subsequent development of DCD has become an interesting line of research. In 2017, Xiao et al. [120] conducted a study on the effects of early isoproterenol exposure on cerebellar development in mice. Immunofluorescence experiments were performed to observe and assess Purkinje cell dendritogenesis and Bergmann glial cell development in the cerebellum of isoproterenol-anesthetized mice. The experimental results showed that Purkinje cells and their dendritic length were significantly reduced in the cerebellum, and the filopodia formation and cellular phenotypic differentiation of Bergmann glial cells were also disturbed. Among them, Purkinje cells were the only output neurons in cerebellar circuits, which were the central cell populations integrating motor coordination and learning [121]. In contrast, Bergmann glial cells specifically promoted the development of dendrites of Purkinje cells [122]. Both of them were fundamental in maintaining cerebellar functions. This specified that early exposure to general anesthesia may lead to abnormal development of key neurons in the cerebellum, ultimately resulting in DCD.

In addition, some neuroimaging findings manifested that there were alterations in the white matter microstructure of the brain in children with DCDs, especially in the corticospinal tract, posterior thalamic radiation, and sensory-motor tracts in the parietal subregion of the corpus callosum [114]. In 2017, Wilson et al. [123] analyzed the diffusion-weighted magnetic resonance imaging data and found that sensory-motor-related white matter organization in children with DCD was decreased, along with changes in the structural connectivity of brain networks. These findings denoted that cerebral white matter damage may be an important contributor to DCD. The neurotoxic effects of multiple general anesthetic exposures in early life leading to developmental deficits in the white matter of the brain have long been the focus of academic research. In 2023, Kim et al. [124] conducted a longitudinal structural magnetic resonance imaging analysis of the effects of multiple early general anesthesia exposures on the development of the brain in rhesus monkeys and demonstrated that multiple general anesthesia exposures during infancy may delay the maturation of white matter and impair its microstructural integrity. Taken together, early general anesthesia exposure may trigger developmental deficits in cerebral white matter, and there may be a correlation between damage to cerebral white matter and DCDs. However, there is a lack of specialized studies revealing the exact link between early general anesthesia exposure and DCD, which illuminates a direction for future research.

In addition to cerebellar developmental disorders and damage to the white matter of the brain, damage to brain regions such as the basal ganglia and parietal lobe may also be associated with DCD [125]. A review has now explored in detail the association of parietal lobes with DCD and explicated that children with DCD have deficits in visuospatial processing, face recognition and executive functioning, all associated with the parietal lobe [116]. Nonetheless, the developmental toxicity imposed by general anesthesia exposure on the parietal lobe is still an area that deserves in-depth exploration, and further studies are required to reveal its potential mechanisms and effects.

Apart from structural and functional damage to brain regions, insufficient activation of functional networks between brain regions has been associated with DCD [114]. In 2016, in children with DCD, Biotteau et al.’s [116] team found reduced cerebellar dysfunction and parietal activation patterns. Frontal-parietal and frontal-cerebellar neural networks play a crucial role in motor planning, motor control (especially internal modeling), visuomotor mapping, and automation [126]. At the same time, these regions overlap with the mirror neuron system and are involved in developing learned behaviors [127]. Therefore, underactivation of functional networks between these brain regions may be one of the reasons for developing DCD. Meanwhile, general anesthetics disrupt frontal-parietal brain communication [128]. Accordingly, we speculated that early exposure to general anesthesia may cause dysfunctional networks among different brain regions and lead to DCD. Whereas there lacks definitive research evidence, which awaits further exploration. In summation, early exposure to general anesthesia is strongly associated with the DCD, and the neurotoxic effects of general anesthesia leading to neurobehavioral developmental disorders remain to be unveiled.

To avoid the neurological toxicity effects and long-term negative consequences resulting from early exposure to general anesthesia, screening and monitoring of vulnerable populations should be prioritized, particularly children with specific genetic backgrounds or a history of neurological disorders. Simultaneously, it is essential to develop and employ more precise assessment tools to identify patients at higher risk of neurological toxicity when being exposed to general anesthetics. Tailored anesthesia plans that take into consideration the unique characteristics of each patient should be crafted. Furthermore, long-term neurobehavioral tracking of patients receiving general anesthesia is necessary to timely identify and take interventions for developmental behavioral disorders.

For researchers, animal models with high similarity to human neurobiology such as macaques should be developed first to provide basic tools for studying the neurotoxicity and mechanisms of general anesthetics. Second, the molecular and cellular mechanisms that drive the neurotoxic effects of general anesthetics should be explored, including their impact on critical processes like neuronal survival, differentiation, synaptogenesis, and function. Finally, based on a comprehensive understanding of neurotoxicity mechanisms, new intervention measures and drugs to alleviate or prevent the adverse effects of general anesthetics should be researched and developed.

In this review, a vast array of literature on the effects of general anesthesia on long-term neurobehavioral function in children was reviewed. Through analyzing those researches, it summarized the toxicological mechanisms closely linked to clinical neurobehavioral changes that may be induced by general anesthetic agents in basic research, offering clinical insights for the appropriate and safe use of these drugs in infants and young children. Moreover, it identified existing research gaps and shed light on future research directions which are expected to provide scientific evidence for the formulation of safer anesthesia practice policies and guidelines.

The authors have no conflicts of interest to declare.

This work was supported by the National Natural Science Foundation of China (Grant No. 82371276, 82171173); the Natural Science Foundation of Shanghai (Grant No. 22ZR1437200); and Clinical Research Program of Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University School of Medicine (Grant No. JYLJ202308).

Jinnan Xu conceptualized and outlined the review article. Jinnan Xu, Fan Wang, Yi Gao, Chuanyu Qi, and Tiannan Chen performed material preparation and literature search. Jinnan Xu wrote the first draft, and Jia Yan commented and edited. All authors approved the final draft.