Introduction
Understanding of autism spectrum disorder (ASD) has changed significantly since first described by Hans Asperger and Leo Kenner (Asperger, Reference Asperger1938; Kanner, Reference Kanner1943). Currently, we define ASD as a neurodevelopmental condition with persistent deficits in social communication and social interaction across multiple domains, including deficits in social reciprocity, non-verbal communicative behaviours used for social interaction, and skills in developing, maintaining, and understanding relationships. Furthermore, the diagnosis requires the presence of restricted, repetitive patterns of behaviour, interests, or activities (American Psychiatric Association, 2013).
The prevalence of ASD also has been increasing over the years. The most recent findings of the Centres for Disease Control, issued in March 2020, estimated a prevalence rate of ASD of 1 in 54 in child population age 8, for the surveillance year 2016. This represents a near three-fold increase in the last 15 years (Maenner et al. Reference Maenner, Shaw, Baio, Washington, Patrick and DiRienzo2020). The findings from the Autism and Developmental Disabilities Network in 2016 (for the surveillance year 2012) had previously estimated that the percentage of children identified with ASD was about 1 in 68 or 1.5% (Baio et al. Reference Baio, Wiggins, Christensen, Maenner, Daniels and Warren2018).
The estimated economic burden associated with ASD is substantial, affecting multiple sectors of society. The total cost per year for children with ASD, based on US studies, was estimated to be between $11.5 billion and $60.9 billion (US dollars) (Lavelle et al. Reference Lavelle, Weinstein, Newhouse, Munir, Kuhlthau and Prosser2014). This cost represents a variety of direct and indirect costs, from medical care, to compensatory education, to lost parental productivity. The findings from an Irish study show the average annual cost per child for families amounted to €28,464.89 (Roddy & O’Neill, Reference Roddy and O’Neill2019).
Additionally, ASD frequently presents with multiple comorbidities, which may increase difficulties during study participation, as well as during diagnosis, management, and treatment processes. Comorbid psychopathologies significantly overrepresented in ASD include anxiety, depression, attention-deficit/hyperactivity disorder, eating disorder, gender dysphoria, intellectual disability, behaviours that challenge, as well as medical comorbidities – seizures, sleep difficulties, gastrointestinal disorders, mitochondrial dysfunctions, and immune system disorders (Mannion & Leader, Reference Mannion and Leader2013). Children with ASD present more often with toileting and feeding issues (Emond et al. Reference Emond, Emmett, Steer and Golding2010; Richardson, Reference Richardson2016).
In previous studies, researchers have described dissimilarities in brain structure and function in people with ASD (Brieber et al. Reference Brieber, Neufang, Bruning, Kamp-Becker, Remschmidt, Herpertz-Dahlmann, Fink and Konrad2007; Ameis & Catani, Reference Ameis and Catani2015; Hong et al. Reference Hong, de Wael, Bethlehem, Lariviere, Paquola, Valk, Milham, Di, Margulies, Smallwood and Bernhardt2019); however, the underlying biomolecular basis of the condition remains unknown. This has been causing a lack of specific treatment targets. Currently, therapeutic options for ASD are restricted to supportive use of psychotropics to reduce specific symptoms in children, such as use of antipsychotics – risperidone and aripiprazole – for challenging behaviour (Matson et al. Reference Matson, Sipes, Fodstad and Fitzgerald2011) or selective serotonin reuptake inhibitors (SSRI) for obsessive and repetitive behaviour (Hollander et al. Reference Soorya, Chaplin, Anagnostou and Je2012). Efficacy of such management remains controversial (Williams et al. Reference Williams, Brignell, Randall, Silove and Hazell2013; Reddihough et al. Reference Reddihough, Marraffa, Mouti, O’Sullivan, Lee, Orsini, Hazell, Granich, Whitehouse, Wray, Dossetor, Santosh, Silove and Kohn2019).
Hence, there is a need for better aimed, ASD-specific treatment, and this will only be possible with a better understanding of its physiology. Some of the most advanced techniques used in understanding mechanisms in autistic brains are nuclear imaging techniques – positron emission tomography (PET) and single-photon emission computed tomography (SPECT).
In both of these techniques, radioactive material (radiotracer) is administered through intravenous injection, and it binds to proteins that could be of interest (specific receptors, neurotransmitters, enzymes, etc.) or it could be metabolised and accumulated by an organ which helps to monitor specific activities, for example, radiolabelled glucose metabolised by the brain (Lammertsma, Reference Lammertsma2001). While SPECT uses gamma camera detectors to detect the gamma ray emissions from the tracer, PET detects photons from the annihilation process of positrons (the product of the decay of the radiotracers) and electrons from the organ.
Methods
In this article, the authors would like to familiarise readers with the most recent molecular imaging studies and discuss further research opportunities in ASD, as well as challenges.
Selection criteria
This literature review aimed to include all studies that were published in English language in peer review journals published between January 2009 and January 2020. Researchers conducted a literature search in PubMed, EMBASE, Scopus, and PsychINFO using the search terms: ASD, autism spectrum disorder, PET, and SPECT in various combinations.
Authors initially identified 299 studies in total. After removal of duplicates, 139 article titles and abstracts were found to be rated 4 or 5 on the scale by two reviewers. During the screening process, we used Likert-like scale (from 1 to 5) to assess quality of the articles, where 4 and 5 represented high relevance studies. Articles with lowest score (1 to 3) were excluded.
Inter-rater reliability between two researchers was also calculated for all marked studies. The result was a Cohen’s κ = 0.79 (95% CI 0.68, 0.87) which suggest moderate to strong level of agreement. Cohen suggested the κresult be interpreted as follows: values ≤ 0 as indicating no agreement and 0.01–0.20 as none to slight, 0.21–0.40 as fair, 0.41– 0.60 as moderate, 0.61–0.80 as substantial, and 0.81–1.00 as almost perfect agreement (Cohen, Reference Cohen1960).
References of the articles were also cross-checked. We focused on original studies on humans from the 10-year-period, available in English, but we also included some older relevant, innovative studies (4 PET and 5 SPECT; please see Fig. 1). After reviewing the full articles, the total number of articles included in this study was reduced to 31. For the full list of reviewed studies, please refer to Tables 1 (PET) and 2 (SPECT).
PRISMA flow diagram. IRR, inter-rater reliability.
PET studies in autism
5HT, serotonin; 5HT2, serotonin receptor type 2; AChE, acetylcholinesterase; SERT, serotonin transporter; DAT, dopamine active transporter; CBF, cerebral blood flow; GABAa, γ-aminobutyric acid receptor type a; mGluR5, metabotropic glutamate receptor 5; [18F]FPEB-[18F]-3-fluoro-5-[(pyridin-3-yl)ethynyl]benzonitrile; [11C]McN-5652-(11C)-labelled trans-1,2,3,5,6,10-beta-hexahydro-6-[4-(methylthio)phenyl]pyrrolo-[2,1-a]isoquinoline; [11C]WIN-35,428-2Beata-carbomethoxy-3-beta-(4-fluorophenyl) tropane; [18F]FDG-2-[18F]Fluoro-2-deoxy-d-glucose; [11c]AMT-alpha[11C]methyl-L-tryptophan; [18F]MPPF-4-(2´-Methoxyphenyl)-1-[2´-(N-2”-pyridinyl)-p-[18F]fluorobenzamidoethylpiperazine; [11C]MP4a-N-[11C]Methylpiperidin-4-yl acetate; [11C]DASB-(11C)23-amino-4-(2[(dimethylamino)methyl]phenylthio) benzonitrile; [11C]Ro15–4513-(11C)-labelled 8-Azido-5,6-dihydro-5-methyl-6-oxo-4 H-imidazo[1,5-a][1,4]benzodiazepine-3-carboxylic acid ethyl ester; [11C]PK11195- ([1-(2-chlorophenyl)-N-methyl-N-(1-methylpropyl]-3-isoquinoline carboxamide); [11C]MDL-4-(1-hydroxy-1-(2,3-dimethoxyphenyl)methyl)-N-2-(4-fluorophenylethyl)piperidine; ADOS, Autism Diagnostic Observation Schedule; ADI-R, Autism Diagnostic Interview-Revised; ICD-10, International Statistical Classification of Diseases and Related Health Problems, 10th revision; DSM, Diagnostic and Statistical Manual of Mental Disorders; GARS, Gilliam Autism Rating Scale; CARS, Childhood Autism Rating Scale; VABS, Vineland Adaptive Behavior Scales; ABC, Autism Behaviour Checklist; IQ, intelligence quotient.
SPECT studies in autism
5HT2a, 5-hydroxytryptamine receptor type 2a; SERT, serotonin transporter; DAT, dopamine active transporter; CBF, cerebral blood flow; GABAa, γ-aminobutyric acid receptor; 123I-5-I-R91150-Iodine-123 labelled; 4-amino-N-[1-[3-(4-fluorophenoxy)propyl]-4-methyl-4-piperidinyl]-5-io do-2-methoxybenzamide; [123 I]nor-beta-CIT-123iodinated N-(2-fluoroethyl)-2b-carbomethoxy-3b-(4-iodophenyl)-nortropane; Tc-99 m HMPAO-Technetium-99 m-hexamethylpropyleneamine oxime; Tc-99 m ECD-Technetium-99 m-ethyl cysteinate dimer; [123I]IMZ, 123I-Iomazenil; ICD-10, International Statistical Classification of Diseases and Related Health Problems, 10th revision; DSM, Diagnostic and Statistical Manual of Mental Disorders; ADI-R, Autism Diagnostic Interview-Revised; CARS, Childhood Autism Rating Scale; ABC, Autism Behaviour Checklist; IQ, intelligence quotient.
Results
As rightly noted in previous systematic reviews, compared with other conditions like Parkinson’s disease, Schizophrenia, or Alzheimer’s disease, there are few studies that investigate ASD (Zürcher et al. Reference Zürcher, Bhanot, McDougle and Hooker2015). Brain metabolism and cerebral blood flow (CBF) using 2-[18 F]Fluoro-2-deoxy-d-glucose (FDG) account for nearly three quarters of all the studies. In this literature review, authors would additionally like to present several newer investigations involving neurotransmitters and other proteins with clinical relevance.
PET in autism research
The vast majority of the studies investigated glucose metabolism or CBF. From the beginning, researchers were looking for proof of altered metabolism and dysfunctional maintenance processes.
[15O] H2O and CBF
By using water with radioactive oxygen-15 atom, researchers are able to look for parts of the brain with hypo- or hyperperfusion. In 1 of the studies, 21 children with ASD were compared with 10 neurotypical children. Findings suggested that, in the children with ASD, there was significant hypoperfusion in both temporal lobes localised in the associative auditory and adjacent multimodal cortex (Zilbovicius et al. Reference Zilbovicius, Boddaert, Belin, Poline, Remy, Mangin, Thivard, Barthelemy and Samson2000). Similar results were obtained on 12 participants in an extension study. Some other researchers in their study used multivariate machine-learning methods, where 45 participants with ASD were scanned and compared with 13 neurotypical participants; they noticed reduced perfusion in the right superior sulcus (temporal lobe), and increased perfusion in the left post-central area in the parietal lobe, which is the main sensory receptive area for the sense of touch (Duchesnay et al. Reference Duchesnay, Cachia, Boddaert, Chabane, Mangin, Martinot, Brunelle and Zilbovicius2011).
[18F]FDG and glucose metabolism
From studies where 2-deoxy-2-(18F)Fluoro-D-glucose (FDG) was used, we have some knowledge about glucose metabolism in the brain. It is a tracer mainly used in measuring cancer metabolism. However, the findings have not been consistent throughout the years. When 17 ASD participants were compared with 17 neurotypical participants, findings suggested lower glucose level metabolic rates bilaterally in the ventral caudate, putamen, and thalamus (Mehmet Haznedar et al. Reference Haznedar, Buchsbaum, Hazlett, LiCalzi, Cartwright and Hollander2006). In another study with 15 children with ASD, moderate to severe intellectual disability, and comorbid infantile spasm, hypometabolism was described in the frontal and parietal lobes (Dilber et al. Reference Dilber, Çalışkan, Sönmezoğlu, Nişli, Motavalli Mukaddes, Tatli, Aydinli, Ekici and Özmen2013). Findings from a study involving 23 children with ASD suggested low glucose uptake predominantly in the hippocampus and amygdala, followed by the mesial temporal lobe and cerebellum, while there was significant uptake in the frontal lobes (Chivate et al. Reference Chivate, Thakrar, Narang, Kumar, Verma, Patkar and Chivate2016). In an adult ASD study involving 25 participants, a decreased metabolic rate was found in the parietal lobe, frontal premotor, visual fields area, and amygdala (Mitelman et al. Reference Mitelman, Bralet, Mehmet Haznedar, Hollander, Shihabuddin, Hazlett and Buchsbaum2018). Glucose metabolism was increased in the posterior cingulate, occipital cortex, hippocampus, and basal ganglia. In another study, looking also at sensory processing abnormalities, involving 100 participants with ASD and 100 controls, hypometabolism was observed in the temporal lobe as well as diffuse cerebral changes (Kadwa et al. Reference Kadwa, Sahu, Singhi, Malhi and Mittal2019). Hypermetabolism was observed in the frontal lobe.
These at times contradictory differences arise from different study designs. There are discrepancies in diagnosing ASD, protocols for scanning, including sedation, controlling for cofounding factors like age, sex, epilepsy, medication prescribed, or intelligence quotient (IQ). The power of the studies with few participants is questionable as well. Additionally, ethical concerns arise regarding enrolling healthy participants, especially children, in studies involving radioisotopes, which can be controversial. It is worth remembering that ASD is highly heterogeneous in presentation, as is the time course of its core deficits.
In one study, [18F] FDG PET was used to investigate changes in glucose metabolism in the brain after participants were subjected to a ketogenic diet. After 12 months of this diet, there was a markedly decreased uptake diffusely in the whole cerebral cortex with relatively low reduction in the basal ganglia (Żarnowska et al. Reference Żarnowska, Chrapko, Gwizda, Nocuń, Mitosek-Szewczyk and Gasior2018).
Serotonin (5-HT) and oxytocin (Oxt)
Serotonin (5-hydroxytryptamine, 5-HT) is a neurotransmitter which plays a regulatory role in prenatal and postnatal brain development. Its disruption of signalling involved in the pathogenesis of ASD during the early development of the brain has been well documented (Yang et al. Reference Yang, Tan and Du2014).
In order to determine whether there are serotonergic abnormalities in the brains of children with ASD, researchers evaluated serotonin synthesis capacity in vivo using the tryptophan analogue – alpha[11C]Methyl-L-tryptophan (AMT) as a radiotracer (Chandana et al. Reference Chandana, Behen, Juhász, Muzik, Rothermel, Mangner, Chakraborty, Chugani and Chugani2005). Data from 117 children with ASD suggested that 5-HT synthesis capacity increased gradually between the ages of 2 years to 15 years to reach 1,5 times normal adult values. Also, locally, asymmetries of AMT uptake were observed in the frontal cortex, cerebellum, and thalamus. It has been advocated that those changes are related to the language function.
In other studies, using [18F]Setoperone, cortical serotonin receptor type-2 (5-HT2) density was measured in parents of children with ASD. Findings suggest that cortical 5-HT2 binding potentials throughout the cerebral cortex were significantly lower in parents compared to controls as well as platelet serotonin levels were significantly negatively correlated with cortical 5-HT2 binding potentials in parents (Goldberg et al. Reference Goldberg, Anderson, Zwaigenbaum, Hall, Nahmias, Thompson and Szatmari2009). The same tracer used in another study on eight adult participants with high functioning ASD showed decreased serotonin binding in the thalamus only (Beversdorf et al. Reference Beversdorf, Nordgren, Bonab, Fischman, Weise, Dougherty, Felopulos, Zhou and Bauman2012). Significant differences were not detected in other brain regions.
Using the radiotracer 4-(2´-Methoxyphenyl)-1-[2´-(N-2”-pyridinyl)-p-[18F]fluorobenzamidoethyl-piperazine ([18F]MPPF), marking the 5-HT 1 A receptor (5-HT1AR), researchers also tried to measure MPPF binding potentials (Lefevre et al. Reference Lefevre, Mottolese, Redouté, Costes, Le Bars, Geoffray, Leboyer and Sirigu2018). At baseline, patients with ASD did not differ from controls for 5-HT1AR concentration and distribution. However, while intranasal oxytocin was used, MPPF binding potentials significantly increased in several brain regions of neurotypical participants. The lack of oxytocin effect at both peripheral and central levels of serotonin was suggestive of disturbed interactions.
Oxytocin is a neurohormone synthesised in the hypothalamus; its importance for social behaviour has gained a lot of attention since the first evidence of oxytocin-induced improvements of social skills in patients with ASD (Watanabe et al. Reference Watanabe, Abe, Kuwabara, Yahata, Takano, Iwashiro, Natsubori, Aoki, Kawakubo, Kamio, Kato, Miyashita, Kasai and Yamasue2014; Lefevre & Sirigu, Reference Lefevre and Sirigu2016). Oxytocin also leads to increased activity in dopaminergic brain regions associated with reward, including the striatum and the nucleus accumbens. In another study using [11C]DASB – 23-amino-4-(2-dimethylaminomethylphenylsulfanyl)-benzonitrile that binds to serotonin transporter (SERT), after a 10-week-long exposure to oxytocin of 10 adult patients with ASD, findings showed significant elevated binding in the left inferior gyrus extending to the left middle frontal gyrus. However, that increased synthesis of the SERT had no significant correlation with any clinical variables (Hirosawa et al. Reference Hirosawa, Kikuchi, Ouchi, Takahashi, Yoshimura, Kosaka, Furutani, Hiraishi, Fukai, Yokokura, Yoshikawa, Bunai and Minabe2017). The SERT is responsible for the reuptake of 5-HT from the synaptic cleft back into the synaptic boutons. Thus, it terminates the effects of 5-HT and at the same time enables its reuse by the presynaptic neurons.
In other studies, looking at SERT binding properties, results were suggestive of the reduction in the anterior and posterior cingulate cortices which were also associated with the impairment of social cognition in participants with ASD. Correlation was also found between obsessive and repetitive behaviours, interests, and the reduction of SERT binding in the thalamus (Nakamura et al. Reference Nakamura, Sekine, Ouchi, Tsujii, Yoshikawa, Futatsubashi, Tsuchija, Sugihara, Iwata, Suzuki, Matsuzaki, Sugiyama, Takei and Mori2010).
There are also studies using tracers [11C]MDL 100907 and [11C]DASB binding to 5HT2 receptor and SERT that found no significant alterations in binding parameters in adults with ASD (Girgis et al. Reference Girgis, Slifstein, Xu, Frankle, Anagnostou, Wasserman, Pepa, Kolevzon, Abi-Dargham, Laruelle and Hollander2011).
Dopamine
Dopamine is a hormone and neurotransmitter that plays a pivotal role in multiple processes, including executive functions, motor control, arousal, motivation, reinforcement, and reward. Genetic studies report an association between dysfunction of dopamine active transporters (DATs) and ASD (Hamilton et al. Reference Hamilton, Campbell, Sharma, Erreger, Herbor, Saunders, Belovich, Sahai, Cook, Gether, McHaourab, Matthies, Sutcliffe and Galli2013).
In the previously mentioned study (Hirosawa et al. Reference Hirosawa, Kikuchi, Ouchi, Takahashi, Yoshimura, Kosaka, Furutani, Hiraishi, Fukai, Yokokura, Yoshikawa, Bunai and Minabe2017) where radioligands [11C]McN-5652 and [11C]WIN-35,428 were used with adult participants, findings suggested that dopamine transporter binding was significantly higher in the orbitofrontal cortex, where dopamine transporter binding was significantly inversely correlated with SERT binding. Those results that have higher activity in the dopaminergic system are comparable with previous PET studies, which found increased striatal dopamine D2 receptor binding in ASD children (Fernell et al. Reference Fernell, Watanabe, Adolfsson, Tani, Bergstrom, Hartvig, Lilja, von Knooring, Gillberg and Långström1997) and increased dopamine synthesis and storage in the striatum and frontal cortex of adults with ASD (Nieminen-Von Wendt et al. Reference Nieminen-Von Wendt, Metsìhonkala, Kulomìki, Aalto, Autti, Vanhala, Eskola, Bergman, Hietala and Von Wendt2004).
Gamma-aminobutyric acid (GABA) and glutamate (Glu)
GABA is the main inhibitory neurotransmitter. Its principal role is reducing neuronal excitability throughout the nervous system. Excitation/inhibition (E/I) imbalance remains a widely commented hypothesis in ASD conditions (Rubenstein & Merzenich, Reference Rubenstein and Merzenich2003). The E/I balance is assumed to be continuously controlled by the interaction of GABA-ergic and glutamatergic functions.
Despite the popularity of this thesis, there are a limited number of PET studies. Researchers at King’s College London reported significantly lower radiotracer binding to GABA in the left and right nucleus accumbens and the left and right subcallosal area on three adult participants (Mendez et al. Reference Mendez, Horder, Myers, Coghlan, Stokes, Erritzoe, Howes, Lingford-Hughes, Murphy and Nutt2013). Investigators concluded that GABA Aα5 expression in adult males with ASD is suggestive of defects in GABA system signalling. In subsequent studies on 27 adult participants, the same group, in collaboration with researchers in Stockholm, suggested that GABA receptor availability may be normal in adults with ASD, although GABA signalling may be malfunctioned (Horder et al. Reference Horder, Andersson, Mendez, Singh and Borg2018).
Recent studies focusing on the role of glutamate in ASD demonstrate that metabotropic glutamate receptor 5 (mGluR5) binding is altered in the post-central gyrus and cerebellum of subjects with ASD, suggesting abnormal glutamate signalling in these regions. Lastly, the correlations between altered radiotracer binding potential in the praecuneus and the cerebellum indicate that some ASD symptoms may be influenced by abnormal glutamate signalling (Fatemi et al. Reference Fatemi, Wong, Brašić, Kuwabara, Mathur, Folsom, Jacob, Realmuto, Pardo and Lee2018).
Acetylcholinesterase
To our knowledge so far, there has only been one study investigating the role of acetylcholinesterase (AChE) in ASD. Studies using AChE inhibitors reported lower aggression levels and reduced attention deficit in children with ASD (Nicolson et al. Reference Nicolson, Craven-Thuss and Smith2006). In PET studies on a group of 20 adult participants, it was reported that lower acetylcholine activity in the bilateral fusiform gyri was negatively correlated with social functioning. Further studies are warranted.
Protein synthesis
As far as we are aware, there is only one study looking at protein synthesis in the brain in young patients diagnosed with developmental delay and pervasive developmental disorder (PDD) by using L-[1-11C]-Leucine. The incorporation of this radioactive exogenous amino acid into brain proteins is indicative of the protein synthesis rate.
Intensified synthesis and deposition of specific proteins, such as Rho GTPase, have been reported to affect morphology of dendritic spines and, furthermore, numerous components of the GTPase network have been associated with ASD and other neurodevelopmental conditions (Pinto et al. Reference Pinto, Pagnamenta, Klei, Anney, Merico and Regan2010). For example, in Fragile X syndrome, the abnormal spine morphology is associated with dysregulated protein synthesis, cognitive impairment, and other neurological signs.
In the study on eight children with developmental delay with PDD, researchers looked at protein synthesis rates in specific regions (Grossman et al. Reference Grossman, Aldridge, Weiler and Greenough2006). They found an increased protein synthesis rate in the perisylvian language zone (temporal region) of children with developmental delay without PDD, and these findings may be associated with the pervasive developmental symptoms. These results are interesting as they confirm previous findings in similar regions as other molecular imaging investigations, but further research is required.
Neuroinflammation
Cases of ASD have frequently been reported in association with congenital rubella and other infections. These observations stimulated further studies into markers of immune function in ASD. In researches on animals, findings suggested that there are altered central and peripheral inflammatory responses in mouse models of ASD (Lucchina & Depino, Reference Lucchina and Depino2014). Various genetic studies linked ASD with genes that are associated with several immune functions, including the HLA antigen or complement C4b – a major histocompatibility complex protein (Stigler et al. Reference Stigler, Sweeten, Posey and McDougle2009).
Studies would also investigate microglia – glial cells that sense pathologic tissue alterations. Neuroinflammation, which involves microglial activation, is thought to play a key role in the development and progression of neurodegenerative diseases and other brain pathologies. Once brain tissue is exposed to any form of injury, the microglia are rapidly activated releasing several cytokines, that is, tumour necrosis factor and interleukin 1 beta or oxygen species with other proteins, that is, translocator protein 18 kDa (TSPO) that binds to radioligand (R)-[11C]PK11195 used in the study by Suzuki and colleagues (Reference Suzuki, Sugihara, Ouchi, Nakamura, Futasubashi, Takebayashi, Toshihara, Omata, Matsumoto, Tsuchiya, Iwata, Tsujii, Sugiyama and Mori2013). In their study on 20 adult participants with ASD, and IQs within the normal range, results suggested increased binding potential values in multiple brain regions. Brain areas with increased binding potentials included the cerebellum, pons, fusiform gyri, midbrain, and the anterior cingulate and orbitofrontal cortices. The most significant increase was detected in the cerebellum. Those findings supported augmented but not altered diffuse microglial activation in the brain in the subjects with ASD which may have important implications in ASD research.
SPECT in autism research
Tracers used in SPECT typically have longer half-lives than the one used in PET; therefore, they should be more accessible and less expensive. However, SPECT has 2–3 times lower sensitivity than PET and slightly lower spatial resolution.
Over recent years, there has been a lack of SPECT studies in ASD research. Similar to PET studies, CBF is mostly studied using SPECT; however, this has become less popular due to the development of less expensive other MRI techniques like arterial spin labelling (ASL), also known as arterial spin tagging, dynamic contrast enhanced, or dynamic susceptibility contrast, which also allow blood flow to be measured, but are less invasive.
Tc-99 m HMPAO, Tc-99 m ECD, and brain perfusion
While both Technetium-99 m-hexamethylpropyleneamine oxime (Tc-99 m HMPAO) and Technetium-99 m-ethyl cysteinate dimer (Tc-99 m ECD) have been used to investigate regional CBF (rCBF), their perfusion patterns differ from rCBF-PET images. This reflects differences in the pharmacokinetics of each medicinal radiocompounds in the central nervous system. For clinical diagnoses, these patterns should be taken into consideration (Koyama et al. Reference Koyama, Kawashima, Ito, Ono, Sato, Goto, Kinomura, Yoshioka, Sato and Fukuda1997).
A study by Degirmenci and colleagues (2008) investigated perfusion patterns in 10 children with ASD and their relatives. Results suggested the existence of regional brain perfusion alterations in the temporal, frontal, parietal cortices and in the caudate nucleus in children with ASD and in their first-degree family members (Degirmenci et al. Reference Degirmenci, Miral, Capa Kaya, İyilikçi, Arslan, Baykara, Evren and Durak2008). In another study with 10 children with ASD and intellectual disability, generalised hypoperfusion of the brain was reported in all 10 cases when compared to control group with the lowest perfusion in the frontal and prefrontal regions. Subcortical areas also exhibited hypoperfusion (Gupta & Ratnam, Reference Gupta and Ratnam2008).
Studies where Tc-99 m ECD was used showed a significant reduction of global cerebral perfusion in young participants with ASD in comparison with neurotypical controls and the presence of left hemispheric dysfunction, mostly in the temporo-parietal regions, which are associated with language processing and the comprehension of music and sounds (Burroni et al. Reference Burroni, Orsi, Monti, Hayek, Rocchi and Vattimo2008). In studies involving 15 children with ASD and comorbid epilepsy, findings suggested mixed hypoperfusion patterns, predominantly in the prefrontal cortex, anterior cingulate cortex, medial frontal cortex, medial parietal cortex, and anterior temporal cortex. The data were compared with databases of neurotypical subjects from the institute (Sasaki et al. Reference Sasaki, Nakagawa, Sugai, Shimizu, Hattori, Nonoda and Sato2010). On the other hand, the study conducted by Yang Wen-Han and colleagues indicated that there was a decrease in rCBF as a global event within 23 child participants. The main regions involved were the frontal and temporal lobes bilaterally, the limbic system, and basal ganglia. An asymmetry was also noted in hypoperfusion among milder forms of ASD (Yang et al. Reference Yang, Jing, Xiu, Cheng, Wang, Bao and Wang2011). Interestingly, brain perfusion as an outcome was used in a Chinese study on the effect of electroacupuncture in ASD. In 55 child participants, perfusion improved bilaterally in the prefrontal cortex and temporal lobes. Unfortunately, there was no control group in this study (Zhao et al. Reference Zhao, Jia, Hu and Sun2014).
Serotonin receptor, 5-HT2
Using 123iodinated 4-amino-N-1-[3-(4-fluorophenoxy)propyl]-4-methyl-4-piperidinyl]5-iodo-2-methoxybenzamide (123I-5-I-R91150) with high affinity and selectivity for serotonin 5-HT2A receptors, researchers from King’s College London were able to look closely at responsivity of the serotonergic system in adult participants with ASD (Murphy et al. Reference Murphy, Daly, Schmitz, Toal, Murphy, Curran, Erlandsson, Eersels, Kerwin, Ell and Travis2006). Their findings suggested that people with ASD have a significant reduction in cortical 5-HT2A receptors, which supported the previously discussed theory of reduced serotonergic system response and platelet 5-HT2A receptor binding. In addition, reduced receptor binding was correlated with abnormal social communication. These findings are clinically relevant as very frequently patients with ASD are treated with medication acting on the serotonergic system (SSRIs, Serotonin-norepinephrine reuptake inhibitors - SNRIs, etc.), but response rates differ.
Transporters for serotonin and dopamine
There is lack of comprehensive data on SERTs and DATs in ASD, especially using SPECT. To our knowledge, the last study looking at the proteins was performed in 2008. The researchers used 123iodinated N-(2-fluoroethyl)-2b-carbomethoxy-3b-(4-iodophenyl)-nortropane (123-I nor-beta-CIT) to measure transporter binding capacity in 15 participants (Makkonen et al. Reference Makkonen, Riikonen, Kokki, Airaksinen and Kuikka2008). The findings interestingly showed reduced SERT binding in the frontal cortex; however, DAT binding did not differ significantly between children with and without ASD.
Gamma-aminobutyric acid receptors
Over the past years, in researches looking at fragmented brain areas, there have been reported neuropathological changes in patients with ASD, especially reductions in and denaturation of the GABAA receptors. 123I-Iomazenil is a ligand with high affinity for central benzodiazepine receptor. That receptor together with GABAA receptor forms a complex; hence, the information can be obtained indirectly about the GABA receptor activity. Mori and colleagues (Reference Mori, Mori, Fujii, Toda, Miyazaki, Harada, Hashimoto and Kagami2012) investigated changes in 25 children with ASD in comparison to 10 nonsymptomatic epilepsy patients. The main results suggested significantly decreased accumulation of 123I-Iomazenil in the superior and medial frontal cortex. Those regions are known to be responsible for inference of thoughts, feelings, and intentions of others. Also, patients with intellectual disabilities showed a greater decrease in the radiotracer’s accumulation.
Discussion
Challenges in molecular imaging in autism research
Over the years, there have been several radiotracers produced with good specificity which can be easily used in the research; however, ASD remains a behaviourally defined diagnosis and spectrum of the condition is very wide. It is important for persons making the diagnosis that they interpret symptoms correctly. The Autism Diagnostic Observation Schedule (ADOS), together with the Autism Diagnostic Interview–-Revised (ADI-R), is considered to be the gold standard in diagnostic evaluations of ASD. However, researchers also frequently highlighted that social and cultural factors influenced the diagnostic rates (Bernier et al. Reference Bernier, Mao and Yen2010). Their use, although not complicated, requires some training to improve validity.
Differences in cultural beliefs about appropriate social behaviour can affect the accurate diagnosis of ASD (Grinker, Reference Grinker2007). For example, in Asian culture, direct eye contact with authorities is recognised as a sign of disrespect; therefore, children would be expected to avoid it. In studies conducted among African parents, it was concluded that carers might be more likely to emphasise on disruptive behaviour more during the assessment rather than social oddities in their children. In other cultures, children might be considered mature for their age, as they respond well to adults, but there are no concerns that children do not relate socially with peers.
It is also worth to consider that several factors like multilanguage families, immigration, or social depravation can affect presentation in a way that an unexperienced rater could interpret as symptoms of ASD. Despite the aforementioned difficulties, it has been recommended that all researchers should consider using ADOS and ADI-R in their studies which has not been the case to date.
Due to deficits in social interaction, anxiety around new experiences, impaired language abilities, and hypersensitivity to sensory stimuli, some careful considerations should be given to overall experience in participation in the studies. Taking part in the study would involve not only attending unfamiliar environment but also some invasive procedures (blood samples, arterial lines, iv cannulas). Scanning would require participants to maintain position for a certain amount of time, which is not always possible, especially for people with more severe forms of ASD. Research centres continuously develop protocols to improve quality of data and overall participants’ experience (Smith et al. Reference Smith, Bhanot, Norman, Mullett, Bilbo, McDougle, Zurcher and Hooker2019).
ASD, as mentioned earlier, is a highly heterogeneous condition, with multiple aetiologies, sub-types, and developmental trajectories. The symptomatology of ASD is pervasive and extensive with a variable onset that could be considered a multi-dimensional process. Cross-sectional studies provide some information; however, there is still a lack of longitudinal observations of the condition. Developing brains create new neural pathways that affect the findings. There are well-documented cases where children meeting ASD criteria initially “grew out of them” (Masi et al. Reference Masi, Demayo, Glozier and Guastella2017). Most studies include children with ASD; however, there is limited number of studies looking at adults with the condition. It is imperative that studies take into consideration increased prevalence of multiple comorbidities among ASD population.
The authors were pleased to note development of the project “European Autism Interventions - A Multicentre Study for Developing New Medications” (EU-AIMS, 2020), a collaboration between organisations representing people with ASD, their families, academia, and industry that would help to address difficulties as mentioned above, allowing professionals with different background to integrate their knowledge. When set up, EU-AIMS was the largest single grant for ASD in the world, and for the study of any neurodevelopmental condition in Europe. The project is therefore a great platform for a variety of professionals and participants to investigate successful treatment options for ASD in the near future, which is long-awaited.
In well-designed studies, controlling for comorbidities is extremely important especially in the context of the increased prevalence of these in ASD. As findings depend on the severity of ASD, and related intellectual disability, it is wise to include IQ in patients’ characteristics and try to find matching controls. As molecular imaging carries a risk of unnecessary exposure to radioisotopes, it is ethically questionable to include healthy volunteers in the studies, especially at a young age.
To date, studies focus on males with ASD, mainly. Numerous researches, together with anecdotal evidence, suggest that the ratio of males to females with ASD ranges from 2:1 to 16:1. The most-up-to-date approximate calculation is 3:1 (Loomes et al. Reference Loomes, Hull and Mandy2017). Females remain underdiagnosed and should be equally represented in research.
Future research
The disparity between the prevalence of ASD and the number of molecular imaging studies conducted to date warrants investments in methods to better address and overcome the practical and technical challenges. Nuclear medicine can provide important information in terms of changes in metabolism in certain part of the brain in response to any therapy, both pharmacological and nonpharmacological. This has not been investigated fully.
Different theories involve GABA, serotonin, oxytocin, or dopamine systems in pathophysiology of ASD. Newer radiotracers are able to bind specifically to designated receptors, giving some understanding of their role in the process as well as response to used medications. Animal models suggest role of acetylcholine in attentional and social deficits in ASD; however, medication acting on nicotinic and muscarinic receptors is not routinely prescribed. With PET and SPECT molecular or metabolic changes in the brain, associated with symptom reduction, during supportive therapies in ASD, are easily visualised.
In terms of technology development, combining MRI/PET imaging would enable concurrent measures of blood flow (with ASL MRI), structural connectivity (diffusion MRI), and functional connectivity (resting state fMRI), along with neurotransmitter receptor synthesis or density (with PET). Combined MRI/PET imaging has the advantage of requiring only a single session, which is a benefit for every participant, but is especially relevant for children or individuals with more severe forms of ASD.
Limitations
The authors have systematically searched several databases and identified studies from across different settings. All type of studies were discussed, not always with the highest quality data. Only articles published in English were included; however, it is worth mentioning that a large amount of molecular imaging studies have been published in Chinese. It is difficult to compare outcomes of the studies due to large heterogeneity and small sample sizes.
Despite several challenges within the molecular imaging research field, these modalities can provide a substantial amount of information about ASD pathophysiology. It is only a short matter of time before pharmacological approaches would be considered to modify responses of specific brain receptors to allow people with ASD deal better with difficulties specific to the condition.
Acknowledgements
The authors would like to thank Dr Ruth Ceannt for her invaluable help and comments on the manuscript.
Conflict of interest
The authors declare that there is no conflict of interest regarding the publication of this article.
Ethical standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committee on human experimentation with the Helsinki Declaration of 1975, as revised in 2008. The authors assert that ethics committee approval was not required for publication of this paper.
Financial support
The authors received no financial support for the research, authorship, and publication of this article.
