The functional neuroanatomy of autism
Paolo Brambillaa Antonio Y. Hardanb Stefania Ucelli di Nemic,d Edgardo Caverzasic Jair C. Soarese,f Jorge Pereza Francesco Baralec,d
be instrumental in furthering understanding of the pathophysiology of autism and in exploring the importance of dimensional measures of the broader phenotype currently defined as autism.
a
Introduction
Biological Psychiatry Unit, IRCCS S. Giovanni di Dio, Fatebenefratelli, Brescia, Italy b Department of Psychiatry, Western Psychiatric Institute and Clinic, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA c Department of Psychiatry, IRCCS S. Matteo Hospital, University of Pavia, Italy d ‘Cascina Rossago’ Rehabilitation Center for Autism, Pavia, Italy e Department of Psychiatry, University of Texas Health Science Center, San Antonio, TX, USA f South Texas Veterans Health Care System, Audie L. Murphy Division, San Antonio, TX, USA Reprint requests to: Dr Paolo Brambilla, (current address) Dipartimento di Patologia e Medicina Sperimentale e Clinica, Sezione di Psichiatria, Università di Udine, Policlinico Universitario Via Colugna, 50 - 33100 Udine - Italy E-mail:
[email protected]
Accepted for publication: November 25, 2003
Summary Autism is a neurodevelopmental syndrome characterized by impaired social and executive functions. Functional magnetic resonance imaging (fMRI) is a non-invasive technique that allows investigation of the neural networks underlying cognitive impairments in autism. In this article, brain imaging studies investigating the functional brain anatomy of autism are reviewed. Face recognition, theory of mind and executive functions have all been explored in functional neuroimaging studies involving autistic patients. The available literature suggests an involvement of abnormal functional mechanisms in face recognition, mentalization and executive functions in adults with highfunctioning autism or Asperger’s syndrome, possibly due to brain maturation abnormalities, and resulting in dysfunctional reciprocal cortico-subcortical connections. Future functional neuroimaging research should investigate subgroups of autistic children and adolescents longitudinally and attempt to integrate genetic, cognitive and empirical approaches. Such studies will Functional Neurology 2004; 19(1): 9-17
KEY WORDS: brain imaging, fMRI, MRS, neuroimaging, PET, SPECT.
Autism is a pervasive neurodevelopmental syndrome mainly characterized by poor social communication, inadequate response to others’ emotions, and stereotypic and obsessional behaviors (1). Individuals with classic autism also have delayed language development and, in most cases, mental retardation or learning disabilities; people with Asperger’s syndrome, on the other hand, present no history of significant language delays or abnormalities and have normal or superior intellectual abilities, but still show impairments of social interaction. Structural magnetic resonance imaging (MRI) investigations have reported increased total brain, parieto-temporal lobe, and cerebellar hemisphere volumes in autism (2). Magnetic resonance spectroscopy (MRS) studies have found evidence for decreased synthesis and increased degradation of prefrontal cortical membranes (3) and reduced concentration of N-acetyl-aspartate (NAA) in the amygdala, hippocampus, cingulate, Wernicke’s area and cerebellum (4-7) in subjects with autism. Positron emission tomography (PET) and single photon emission computed tomography (SPECT) studies have shown temporal and frontal lobe hypoperfusion (8-10) and abnormal temporal cortex activation during auditory tests (11,12). On the basis of these findings, it has been suggested that structural and biochemical abnormalities in neural networks involving the fronto-temporo-parietal cortex, limbic system, and cerebellum underlie the pathophysiology of autism (2). These changes could possibly result from abnormal neuronal development during early life, which may be a result of or triggered by genetic and environmental factors. Functional MRI (fMRI) is an imaging tool that, allowing non-invasive examination of the functional brain architecture of individuals with autism during performance of experimental tasks, makes it possible to study the relationship between regional brain activation and psychopathological symptoms (12). Technically, fMRI explores oxygen consumption and distribution in biological tissues, exploiting the para-magnetic properties of the oxygenated and deoxygenated hemoglobin molecule (blood oxygen level dependent technique, BOLD). During the activation of a specific brain region there is increased blood flow in that region with relatively re-
9
P. Brambilla et al.
duced oxygen extraction fraction, and consequently reduced deoxyhemoglobin levels (13). The BOLD effect is the basis of most fMRI investigations that attempt to elucidate the neural basis of higher cognitive functions. This review summarizes findings from functional neuroimaging studies performed in autism, divided into three categories of investigation: face recognition, theo ry of mind and executive functions. We discuss the limi tations of the fMRI studies conducted to date and outline perspectives for future investigations in this field.
Searching and evaluating fMRI studies in autism Through a comprehensive Medline search of the period 1966-October 2003, and manual search of bibliographic cross-references, we identified all original fMRI research papers, published in English, that have included autistic patients. We rated each paper for completeness on the basis of a 12-point check-list divided into 3 categories: subjects (items 1-4), methods used for image acquisition and analysis (items 5-10) and results and conclusions (items 11,12) (2). The 12 criteria in the check-list were: 1) Patients were evaluated prospectively, specific diagnostic criteria were applied, and demographic data are reported; 2) Healthy comparison subjects were evaluated prospectively, psychiatric and medical illnesses were excluded, and demographic data are reported; 3) Major confounding factors (e.g., age, gender, IQ, handedness, socioeconomic status, educational level, height, total brain measures) were controlled, either by stratification or statistically; 4) Sample size per group ≥ 10; 5) Tasks administered during the fMRI session were standardized and related to the hypothesis of the study; 6) Measures of brain activation are reported; 7) MRI tesla ≥ 1.5; 8) Task procedures and measurement of region of interest (ROI) activity are described clearly enough to be reproducible; 9) Measurements are described clearly enough to be reproducible; 10) ROIs were defined a priori or an exploratory approach is justified; 11) Statistical parameters for significant and important non significant differences are given; 12) Conclusions are consistent with results and limitations are discussed. Without any intent to criticize either the investigators or the studies themselves, the completeness of these published studies was thus rated with a numeric score. Half points were assigned when criteria were partially met.
ropsychological studies have shown specific impairments in face recognition in patients with autism, such as decreased ability to identify faces, and lack of ability to process faces as whole images, these subjects instead concentrating on parts of the face (e.g., the lower face and mouth area) (15-19). Abnormal brain activation during face discrimination has been found, in functional neuroimaging studies, in adults with high-functioning autism or Asperger’s syndrome (Table I). Specifically, increased activity in the bilateral inferior temporal gyrus (ITG), right thalamus, left superior temporal gyrus (STG), and left peristriate visual cortex has been found in subjects with autism, but not in healthy controls, when they process facial features (identity) and facial expressions (emotion) (20-22). In controls, on the other hand, the ITG has been found to be the area most strongly associated with object processing (21). Furthermore, control subjects showed activation in the right fusiform gyrus (FG) (21-23), left cerebellum, left amygdalohippocampal complex, left middle temporal gyrus (MTG) (24), and left inferior and middle frontal gyrus (MTG) (22,25) during face discrimination, while the autistic patients did not. Physiologically, face recognition (perceived holistically) and object recognition (processed analytically) are subtended by separate neural systems employing different strategies or algorithms (26-28). The fusiform, lateral occipital, parahippocampal and posterior inferior temporal gyri are involved in common object perception (29,30). On the other hand, the neural system involved in recognition of faces includes the ITG (early perception of facial features), lateral FG (invariant aspects of faces, i.e., identity), superior temporal sulcus (STS) (changeable aspects of faces, e.g., eye gaze) and amygdala (emotional significance of facial expressions) (31-35). None of the brain regions subtending face processing in individuals without social disability was found to be significantly active in autistic populations (20-23). Specifically, abnormal activity within the medial temporal lobe (i.e., amygdalohippocampal junction, ITG, STG, and MTG), peristriatal visual cortex, and thalamus has been described in autism, possibly explaining some of the social deficits observed in these patients. It is thus possible that individuals with autism do not have a cortical specialization for face recognition, and that they employ, instead, the same neurocircuitry that is used for object processing. A critical test would be to explore, using fMRI, face recognition performance before and after interventions intended to motivate young subjects with autism to look at, and to take an interest in, faces (‘Look at me’ followed by a reward for doing so) (18). Eye tracking movement investigations during fMRI would also be crucial to further explore the correlation of configural/segmental strategies with activation of face processing brain systems.
Face and emotion recognition Recognition of individual facial expressions facilitates the understanding of emotional information that is necessary for adequate interpersonal and social relationships. Individuals with autism and Asperger’s syndrome make little eye contact and seem to treat people as if they were objects, tendencies that could be related to their aberrant social and affective functions (14). Neu-
10
Theory of Mind ‘Theory of Mind’ (ToM) or ‘mentalization’ is the cognitive capacity to infer others’ mental states (thoughts, desires and beliefs) in order to explain their behavior (36). This function is usually acquired in normal children in the first 3-4 years of life. ToM can be seen as a meaFunctional Neurology 2004; 19(1): 9-17
Functional brain imaging in autism
sure of social intelligence, that is, of the ability to interact in complex social groups and in close relationships, to mediate self reference and to predict how others will feel, think and behave (18,37). People with autism have deficits in their ability to think of other people in terms of mental states (36,38). Three functional brain imaging studies have investigated the role of ToM in adults with high-functioning autism or Asperger’s syndrome, showing impaired task performances and abnormal brain activation (Table II, see over). These tests investigated the ability to attribute mental states to others, using the story comprehension paradigm (39), photographs of other people’s eyes (40), and animated sequences (41). In general, high-functioning autistic or Asperger’s syndrome subjects were found to show reduced activation of the STG, STS, and basal temporal areas, and no activation of the amygdala (40) and left medial prefrontal cortex (Brodmann’s area 8/9) (39,41). In contrast, hyperactivation of an unusual prefrontal area was noted (Brodmann’s area 9/10) (39). Also, fewer and less accurate
interpretations during mentalizing tasks were revealed in subjects with autism compared to normal controls. However, it should be noted that individuals with Asperger’s syndrome have also been shown to be capable of conversing with others about mental states (42). This finding could reflect possession of a second-order ToM, which may help performance on experimental tasks, but still does not facilitate social adaptation in real life (43). Additionally, the relationship between the deficits in ToM and the degree of social dysfunction in autistic patients remains unclear. In normal individuals, mentalization has been shown to be associated with a complex brain network including the cingulate, medial prefrontal/orbitofrontal cortex, amygdala, STS and STG (44-49), in which the prefrontal regions may be the execution component and the posterior regions (temporal and parietal) the representational component (50). In particular, the cingulate is active during error monitoring and selection between competing responses, the orbitofrontal cortex mediates decision making, the dorsomedial prefrontal cortex is
Table I - Functional brain imaging studies investigating face processing in autism. Study
Autistic patients
Controls
Methods
Findings in autism
Rating
Schultz et al. (21)
14 male ADI, ADOS & ICD-10 AU (8) or AS (6), right-handed 23.8±12.4 y-o IQ=109.1±19.5
28 normal males right-handed 21.6±8.9 y-o IQ=109.6±16.8
fMRI 1.5T Tasks: object and face discrimination
– abnormal face performance – active in patients during face task: – right and left ITG – active in controls during face and – object task: right FG and ITG, – respectively
12
Critchley et al. (20)
9 male ADI & ICD-10 AU (2) or AS (7) right-handed 37±7 y-o IQ=102±15
9 normal males right-handed 25.5±2.8 y-o IQ=116±10
fMRI 1.5T Task: face emotion recognition
– normal task performance – active in patients: left STG – and left peristriate visual cortex – active in controls: right FG and left – cerebellum, amygdalohippocampal – region and MTG
9
Pierce et al. (23)
7 male ADI & ADOS DSM IV AU 4 right-handed 29.5±8.0 y-o (21-41) IQ=83.7±10.9 (73-102)
8 healthy subjects 4 right-handed 28.3 y-o (20-42) IQ=supposed normal
fMRI 1.5T Task: face and shape perception
– normal task performances – reduced bilateral FG and left – amygdala activation – normal activation of ITG and MTG
9
Hall et al. (22)
8 male DSM IV AU (6) or AS (2), right-handed 20-33 y-o IQ=105±18 (80-130)
8 healthy males right-handed IQ=109±16 (90-135)
H215O PET Task: face and voice emotion recognition
– increased number of errors during 10.0 – task – increased activation in right thalamus, – right anterior temporal lobe and left – AC during emotion recognition
Ogai et al. (25)
5 DSM IV AU right-handed 21.8±5.9 y-o IQ=112.4±10.5
9 normal controls right-handed 23.0±5.2 IQ=113.3±5.2
fMRI 1.5T Task: face emotion recognition
– normal task performances – reduced activation in left MFG during – fear recognition and in left insula, left – IFG and left putamen during disgust – test – normal activation during happy face – recognition
10.0
Abbreviations: AU=autism, AS=Asperger; ADI=Autism Diagnostic Interview-Reviewed; ADOS=Autism Diagnostic Observation Schedule; ITG=inferior temporal gyrus; FG=fusiform gyrus; STG=superior temporal gyrus; MTG=middle temporal gyrus; AC=anterior cingulate; MFG=middle frontal gyrus; IFG=inferior frontal gyrus; y-o=years old; PET=positron emission tomography. Functional Neurology 2004; 19(1): 9-17
11
P. Brambilla et al.
critical for self reference, the amygdala modulates social emotional stimuli, the STS is crucial in processing signals of intention coming from motion of other agents (e.g., human body movements) and the STG is involved in phonological and auditory processing. Thus, on the basis of the neurophysiology of ToM and functional neuroimaging studies exploring ToM in autism, it is conceivable that the brain system subserving ToM in humans is abnormally hypoactive in autistic subjects, and does not involve the amygdala and left medial prefrontal cortex.
Executive functions Generally, the term executive functions refers to a group of higher-level neuropsychological capacities that allow the maintenance of a problem-solving mindset appropriate to the attainment of a goal (18), like planning of behavior, inhibiting automatic actions and retaining information in the working memory (51). Neuropsychological studies have shown impairments in executive and language abilities in autism (52,53), which may underlie inflexibility and difficulty adapting to novel situations, while functional neuroimaging studies have been employed in adults with high-functioning autism to elucidate the brain regions that subtend abnormal executive functions. These studies are described below. Working memory In an fMRI study, Ring et al. (54) found that people with autism or Asperger’s syndrome show a trend to slightly better performance than normal controls on the Embedded Figures Task (EFT), a test – to find simple shapes hidden within complex forms – that investigates information processing relating to object feature analysis.
The group with autism showed a right-sided activation of an area near the junction of the middle occipital and middle temporal regions (Brodmann’s area 17/18/19), whereas in normal controls the right dorsolateral prefrontal cortex (DLPFC) and bilateral parietal and occipital cortex were activated (Table III). Also, the subjects with autism employed an approach characterized by fewer demands on working memory during the EFT task, but greater activity in some of the regions involved in object perception. In another fMRI report, Luna et al. (55) found that patients with autism were significantly less able than healthy individuals to shift their eyes accurately to remembered target locations in the oculomotor delayed response (ODR) task (assessing spatial working memory). On the other hand, their performance in the visually guided saccade (VGS) task (assessing basic sensorimotor and attention processes) was unimpaired (Table III). Furthermore, the autistic patients pre sented no deficits in saccade velocity, duration or laten cy in either the ODR or the VGS tasks, indicating normal sensorimotor processes and preserved cerebellar and subcortical mechanisms. Furthermore, in the ODR versus VGS comparison, individuals with autism were found to show significantly less activation bilaterally in the DLPFC and posterior cingulate cortex compared to healthy subjects. In normal individuals, parietal regions are involved in vi sual attention and in spatial processing, whereas the ventrolateral prefrontal cortex (Brodmann’s area 47), DLPFC (Brodmann’s area 46), inferior parietal lobe (Brodmann’s 40) and MFG (Brodmann’s 6) have been reported to be implicated in working memory (56-58). In particular, during visuospatial working memory processing, the DLPFC integrates sensory information and response plans over time, while the posterior cingulate monitors sensory events related to spatial orientation and memory (59,60). The findings of Ring et al. (54)
Table II - Functional brain imaging studies investigating theory of mind in autism. Study
Autistic patients
Happe et al. (40)
Controls
Methods
Findings in autism
Rating
5 male right-handed AS 6 retrospective male 24 y-o (20-27) controls, right-handed IQ=100 (87-112) 38 y-o (24-65)
H215O PET Task: story comprehension
– worse scores on ToM stories, but – not on physical stories or unconnected – sentences – no activation of medial prefrontal cortex during ToM
BaronCohen et al. (39)
6 DSM IV & ICD-10 AU or AS 4 M, right-handed 26.3±2.1 y-o IQ=108.5±10.5
12 normal controls 6 M, right-handed 25.5±2.8 y-o IQ=110±8.5
fMRI 1.5T – impaired task performances Task: photographs of – abnormal activation during mental other person’s eyes – state recognition: bilateral STG (↑), – left amygdala (↓), right insula (↓), – left IFC (↓)
8
Castelli et al. (41)
10 DSM IV AU or AS 33±7.6 y-o IQ=not reported (high-functioning)
10 normal controls 25±4.8 y-o IQ=supposed normal
H215O PET – ↓ mental state attribution during task Task: silent animation – ↓ left FG, right TmP/Am, bilateral STS sequences – and medial SFG activation
9
7.5
Abbreviations: AU=autism, AS=Asperger; ADI=Autism Diagnostic Interview-Reviewed; ADOS=Autism Diagnostic Observation Schedule; M=males, ToM=theory of mind; STG=superior temporal gyrus; IFG=inferior frontal gyrus; FG=fusiform gyrus; TmP/Am=temporal pole adjacent to amygdala; STS=superior temporal sulcus; SFG=superior frontal gyrus; y-o=years old; PET=positron emission tomography; ↑ or ↓=abnormal enhanced or reduced activation.
12
Functional Neurology 2004; 19(1): 9-17
Functional brain imaging in autism
Table III - Functional brain imaging studies investigating executive functions in autism. Study
Autistic patients
Controls
Methods
Findings in autism
Rating
Working memory Ring et al. (54)
6 AU or AS DSM IV & ICD-10 4 M, all right-handed 26.3±2.1 y-o IQ=108.5±10.5
12 normal controls 6 M, all right-handed 25.5±2.8 y-o IQ=110±8.5
fMRI 1.5T Task: EFT
– normal task performance – active in patients: right cuneus, right – middle and inferior occipital gyrus – active in controls: right DLPFC, – bilateral parietal and occipital cortex
10.5
Luna et al. (55)
11 ADI & ADOS AU 9 M, 9 right-handed 32.3±9.3 y-o IQ=100.3±8.9
6 healthy males, right-handed 30.3±11.8 y-o IQ=104.64±6.7
fMRI 3T Task: ODR and VGS
– reduced accuracy in shifting eyes to 9.5 – remembered target locations in the – ODR task; normal VGS task – reduced bilateral DLPFC and posterior cingulate cortex activation in the ODR versus VGS comparison
Motor control, language and attention Muller et al. (65)
4 DSM IV AU men 3 right-handed 25.5 y-o (18-31) IQ=77.3
5 healthy males, right-handed 26.2 y-o (23-30) IQ=supposed normal
H215O PET Task: listening, repeating, generating sentences
– listening sentences: reduced right – dentate nucleus activation – repeating sentences: increased left – DLPFC and right dentate nucleus – activation – generating sentences: reduced left DLPFC and left thalamus activation
9.5
Muller et al. (62)
7 male DSM IV & ADI AU, 4 right-handed 29.4±9.0 y-o (15-41) IQ=88.3±11.1 (73-103)
7 healthy males 4 right-handed 28.5 y-o (21-43) IQ=supposed normal
fMRI 1.5T Task: visually paced finger movement
– increased bilateral parieto-occipital – & controlateral prefrontal activation – increased spatial variability of most – pronounced activations
9.5
Allen & 8 ADI, ADOS & Courchesne DSM IV AU, 7 M (64) 5 right-handed 26.89±8.59 y-o (14-38) IQ=85±11.95
8 healthy controls, 7 M 5 right-handed 26.77±8.22 y-o (13-39) IQ=113.75±9.21
fMRI 1.5T Task: attention, motor and sensory tests
– normal attention and motor task – performance accuracy – cerebellar activation: minimal during – attention task, increased during motor – task, normal during sensory task
10
Muller et al. (63)
8 male ADI & DSM IV AU 5 right-handed 28.4±8.9 y-o (15-41) IQ=86.5±11.4
8 healthy males 5 right-handed 21-43 y-o IQ=supposed normal
fMRI 1.5T Task: visuomotor learning tests
– increased number of errors during 10.5 – task – increased activation in bilateral inferior – and posterior parietal lobes, bilateral – premotor area, right medial frontal – area and left MFG and SFG – reduced activation in superior parietal and occipital cortices – increased spatial variability of activation – foci
Belmonte & YurgelunTodd (66)
6 ADI & DSM IV AU (5) or AS (1) right-handed 32.7±9.8 y-o (24-50) IQ=not reported
6 right-handed normal controls 27.2±4.4 y-o (23-36) IQ=not reported
fMRI 1.5T Task: visual spatial attention test
– normal task performance – increased activation in bilateral VOC – and striate cortex – activation of left VOC during left – attention instead of during right – attention
10.5
Abbreviations: AU=autism, AS=Asperger; ADI=Autism Diagnostic Interview-Reviewed; ADOS=Autism Diagnostic Observation Schedule; EFT=Embedded Figures Task; ODR=oculomotor delayed response; VGS=visually guided saccade; DLPFC= dorsolateral prefrontal cortex; VOC=ventral occipital cortex; MFG=middle frontal gyrus; SFG=superior frontal gyrus; y-o=years old; PET=positron emission tomography. Functional Neurology 2004; 19(1): 9-17
13
P. Brambilla et al.
and Luna et al. (55) suggest that dysfunctional integration among neocortical regions (DLPFC, posterior cingulate, and parietal cortex) underlies the visual search and spatial components of working memory in autism. On the contrary, no abnormalities were found in the circuitry participating in basic sensorimotor control of saccadic eye movements (55). This could be consistent with Frith’s theory of ‘weak central coherence’ (61) according to which the distinctive cognitive style of autism is characterized by predominance of piecemeal analysis over configural processing. Thus, people with autism would seem to employ a more local approach to the processing of complex stimuli, involving an inchmeal process rather than a ‘normal’ use of working memory and a global approach. Motor control, language and attention One functional brain imaging study reported the presence, during visually paced finger movement, of significant hyperactivation in parieto-occipital lobes and controlateral prefrontal cortices in the autistic group, as opposed to significantly increased activation in the controlateral sensorimotor cortex (i.e., peri-rolandic and supplementary motor areas) in the controls (62). Moreover, the strongest activations were located along the controlateral central sulcus in the primary motor and somatosensory cortices in control subjects, while their location varied from patient to patient in the autistic group, being observed particularly in the prefrontal and superior parietal regions. Subsequently, the same group of researchers (63) showed that, during visuomotor learning tasks, i.e., visually driven motor sequence learning (finger press movements), subjects with autism showed greater activation of prefrontal and posterior/inferior parietal cortices and less activation of superior parietal and occipital foci than normal individuals. Furthermore, increased spatial variability of activation foci was found in autism. Exploring cerebellar functions, Allen and Courchesne (64) showed diffuse hypoactivation during motor task and hyperactivation during attention condition in autism. Specifically, abnormal patterns of activation were seen in the right superior hemisphere VIIa (superior semi-lunar lobule) and in lobule VI. On the contrary, only minimal cerebellar activation during a sensory test (i.e., only visual stimuli without the request to pay attention or respond) was observed both in the autistic and in the control subjects (Table III). In a study investigating language functions (i.e., listening, repeating, and generating sentences), high-functioning adult autistic men showed abnormally low activation in the left DLPFC, left thalamus and right dentate nucleus during the receptive and expressive conditions and abnormally high activation in the left DLPFC and right dentate nucleus during motor speech functions (65). Moreover, the normal left-sided lateralization of language activation in the DLPFC during verbal auditory task, and in the thalamus during expressive language condition, were reversed in the autistic group. Interestingly, a recent fMRI study (66) found that, during a visual spatial attention task, high-functioning adult individuals with autism activated ventral occipital and striatal cortices instead of the normal network (i.e., the
14
DLPFC, superior parietal cortex, MTG, MFG, and premotor cortices). Furthermore, in the autistic group, the normal tendency to activate the left ventral occipital cortex during right attention was reversed (Table III). In normal individuals, basic motor functions are subtended by the supplementary and pre-supplementary motor areas, primary sensorimotor cortices and paleocerebellum (67). Also, cortical activation associated with language functions is strongly lateralized to the left hemisphere and involves a network of regions in the frontal, temporal and parietal lobes (i.e., Broca's area, located in the inferior frontal gyrus; Wernicke’s area, encompassing the superior temporal, supramarginal and middle temporal gyri; and the posterior superior temporal gyrus) (68-70). The altered neurofunctional organization reported in autism in relation to motor control and language functions (62,64,65) may potentially represent the functional substrate compensating the complex stereotyped behavior and delayed language acqui sition associated with this condition. In particular, impaired development of the primary motor and somatosensory cortices (located along the central sulcus) and cerebellum may underlie motor anomalies, whereas maldevelopment of the dentato-thalamo-prefrontal pathway and reversed dominance in right hemisphere may underlie language deficits (Table III).
Limitations of and future perspectives for fMRI studies in autism Although our check-list ratings, presented in the last column of the tables, show that the quality of fMRI investigations in this field has improved over recent years, available fMRI studies nevertheless present several methodological limitations. The majority of the studies enrolled small groups of adult high-functioning patients, often mixed with individuals with Asperger’s syndrome, without comorbid epilepsy or other neurological disorders. Thus, the sample is not completely representative of the general autistic population and the results cannot be generalized to low-functioning individuals with autism or to autistic children/adolescents. Regarding fMRI studies in general, it should be noted that the experimental tasks implemented in these investigations were not always specifically related to the hypotheses the studies set out to test (71). Additionally, the exact spatial specificity of fMRI, the accuracy of the functional maps when compared with the actual sites of neuronal activity, and the nature of the fMRI signals when correlated with processes that define neuronal signaling are all aspects that remain to be determined (72,73). Future functional neuroimaging studies should attempt to overcome the sample and design limitations of earlier investigations. First, fMRI investigations should investigate longitudinally larger samples of young individuals with autism, using individuals very closely matched for age, gender, socioeconomic status and IQ. However, as fMRI is very difficult to perform in children with autism, one possible strategy could be to use appropriate neuropsychological tasks outside the scanner for children and to utilize similar tasks inside the scanner for older children, adolescents and adults. Use of this strategy may not solve the problem, but it would help in Functional Neurology 2004; 19(1): 9-17
Functional brain imaging in autism
the generalization of the results to all subgroups. Considering homogeneous subgroups of individuals affected by autism would facilitate understanding of the time frame of development of specialized brain functions underlying cognitive abilities in autism, allowing better understanding of the links between neurofunctional anato my and behavioral features in this condition. Also, areas of activation should be considered individually by manually tracing the relevant anatomy, in order to increase the spatial localization of functional brain maps. Improvements to task design and imaging sequences could include shorter spin-echo times, the averaging of a greater number of signals to boost the signal-to-noise ratio, and gradient-coil shimming tailored to individual subjects.
Concluding remarks The available functional brain imaging literature points to the existence of aberrant neural networks subtending facial recognition, mentalization and executive functions in autism. Specifically, abnormal activity within the medial temporal lobe and peristriatal visual cortex may underlie deficits in face recognition, while abnormally low activation in a brain system that includes the orbitofrontal cortex, STS and STG may subtend mentalization. With regard to executive functions, dysfunctional integration of the DLPFC, posterior cingulate cortex, and parietal cortex may underlie working memory impairments, while disturbances of the primary motor and somatosensory cortices and cerebellum may underlie motor anomalies. Finally, language deficits may be subtended by anomalies in the dentato-thalamo-prefrontal pathway and reversed dominance in the right hemisphere, while selective attention may activate an aberrant network including the ventral occipital cortex and striatum. Thus, information is not processed normally by multimodal brain systems, even though the visual system is intact. According to the ‘experience-expectant’ model of neural plasticity, cortical regions are genetically pre-programmed to become specialized for a particular function, but exposure to appropriate stimuli is necessary for that specialization to occur (74). Synaptic connections are produced in these brain areas during the period of cortical specialization and are rapidly and selectively strengthened or eliminated, depending on the stimuli received. The perceptual neural system thus adapts to input received from the environment (e.g., ex posure to faces). Also crucial to the correct development of this brain system are visuo-spatial abilities and the natural inclination for social communication and interaction. It can be hypothesized that an aberrant ‘expectant’ neural system involved in face recognition, mentalization and executive functions would not be able to process correctly environmental stimuli, and that this would ultimately lead to the activation of alternative (and aberrant) brain circuitry. Early impairment of brain maturation, gene mutations, inappropriate expression of neurotrophins, and environmental factors could all play a major role in this process (2,75). Moreover, the innate failure to respond to external stimuli would mean failure to stimulate synaptic plasticity and pruning during sensitive periods for the development of neural cog Functional Neurology 2004; 19(1): 9-17
nitive networks. In summary, future functional neuroimaging research should investigate longitudinally subgroups of autistic children and adolescents with homogeneous demographical and behavioral features. Also, brain imaging techniques should be integrated with genetic, cognitive and empirical approaches. This will be instrumental in furthering understanding of the pathophysiology of autism and in exploring dimensional measures of the broader autistic phenotype, such as face recognition or social brain.
Acknowledgements This work was supported by grants from the Biological Psychiatry Unit, Fatebenefratelli-Brescia (Ministry of Health) to Dr Perez and Dr Brambilla and the Italian Society for Neuroimaging in Psychiatry (SINIP, ‘F. Tenconi’ Award) to Dr Brambilla. Dr Hardan’s contribution was partly supported by NIMH grant M64027. Dr Soares was supported by NIMH MH 01736, the Krus Endowed Chair in Psychiatry, and the Veterans Administration.
References 11. Minshew NJ, Payton JB. New perspectives in autism, Part II: the differential diagnosis and neurobiology of autism. Curr Probl Pediatr 1988;18:613-694 12. Brambilla P, Hardan A, di Nemi SU, Perez J, Soares JC, Barale F. Brain anatomy and development in autism: review of structural MRI studies. Brain Res Bull 2003; 61:557-569 13. Minshew NJ, Goldstein G, Dombrowski SM, Panchalingam K, Pettegrew JW. A preliminary 31P MRS study of autism: evidence for undersynthesis and increased degradation of brain membranes. Biol Psychiatry 1993;33:762-773 14. Chugani DC, Sundram BS, Behen M, Lee ML, Moore GJ. Evidence of altered energy metabolism in autistic children. Prog Neuropsychopharmacol Biol Psychiatry 1999; 23:635-641 15. Otsuka H, Harada M, Mori K, Hisaoka S, Nishitani H. Brain metabolites in the hippocampus-amygdala region and cerebellum in autism: an 1H-MR spectroscopy study. Neuroradiology 1999;41:517-519 16. Hisaoka S, Harada M, Nishitani H, Mori K. Regional magnetic resonance spectroscopy of the brain in autistic individuals. Neuroradiology 2001;43:496-498 17. Friedman SD, Shaw DW, Artru AA et al. Regional brain chemical alterations in young children with autism spectrum disorder. Neurology 2003;60:100-107 18. Mountz JM, Tolbert LC, Lill DW, Katholi CR, Liu HG. Functional deficits in autistic disorder: characterization by technetium-99m-HMPAO and SPECT. J Nucl Med 1995; 36:1156-1162 19. Ohnishi T, Matsuda H, Hashimoto T et al. Abnormal regional cerebral blood flow in childhood autism. Brain 2000; 123:1838-1844 10. Zilbovicius M, Boddaert N, Belin P et al. Temporal lobe dysfunction in childhood autism: a PET study. Positron emission tomography. Am J Psychiatry 2000; 157:19881993 11. Muller RA, Behen ME, Rothermel RD, Chugani DC, Muzik O, Mangner TJ, Chugani HT. Brain mapping of language
15
P. Brambilla et al.
12. 13.
14.
15. 16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27. 28. 29.
30.
31. 32.
33.
16
and auditory perception in high-functioning autistic adults: a PET study. J Autism Dev Disord 1999;29:19-31 Boddaert N, Zilbovicius M. Functional neuroimaging and childhood autism. Pediatr Radiol 2002;32:1-7 David A, Blamire A, Breiter H. Functional magnetic resonance imaging. A new technique with implications for psychology and psychiatry. Br J Psychiatry 1994;164:2-7 Dawson G, Webb S, Schellenberg GD et al. Defining the broader phenotype of autism: genetic, brain, and behavioral perspectives. Dev Psychopathol 2002;14:581-611 Hobson RP, Ouston J, Lee A. What’s in a face? The case of autism. Br J Psychol 1988;79:441-453 Davies S, Bishop D, Manstead AS, Tantam D. Face perception in children with autism and Asperger’s syndrome. J Child Psychol Psychiatry 1994;35:1033-1057 Klin A, Sparrow SS, de Bildt A, Cicchetti DV, Cohen DJ, Volkmar FR. A normed study of face recognition in autism and related disorders. J Autism Dev Disord 1999;29:499-508 Klin A, Jones W, Schultz R, Volkmar F, Cohen D. Defining and quantifying the social phenotype in autism. Am J Psychiatry 2002;159:895-908 Bolte S, Poustka F. The recognition of facial affect in autistic and schizophrenic subjects and their first-degree relatives. Psychol Med 2003;33:907-915 Critchley HD, Daly EM, Bullmore ET et al. The functional neuroanatomy of social behaviour: changes in cerebral blood flow when people with autistic disorder process facial expressions. Brain 2000;123:2203-2212 Schultz RT, Gauthier I, Klin A et al. Abnormal ventral temporal cortical activity during face discrimination among individuals with autism and Asperger syndrome. Arch Gen Psychiatry 2000;57:331-340 Hall GB, Szechtman H, Nahmias C. Enhanced salience and emotion recognition in Autism: a PET study. Am J Psychiatry 2003;160:1439-1441 Pierce K, Muller RA, Ambrose J, Allen G, Courchesne E. Face processing occurs outside the fusiform ’face area’ in autism: evidence from functional MRI. Brain 2001;124: 2059-2073 Critchley H, Daly E, Phillips M et al. Explicit and implicit neural mechanisms for processing of social information from facial expressions: a functional magnetic resonance imaging study. Hum Brain Mapp 2000;9:93-105 Ogai M, Matsumoto H, Suzuki K et al. fMRI study of recognition of facial expressions in high-functioning autistic patients. Neuroreport 2003;14:559-563 Rhodes G, Tan S, Brake S, Taylor K. Expertise and configural coding in face recognition. Br J Psychol 1989; 80:313-331 Schwarzer G. Development of face processing: the effect of face inversion. Child Dev 2000;71:391-401 Marcus DJ, Nelson CA. Neural bases and development of face recognition in autism. CNS Spectrums 2001;6:36-59 Malach R, Reppas JB, Benson RR et al. Object-related activity revealed by functional magnetic resonance imaging in human occipital cortex. Proc Natl Acad Sci U S A 1995;92:8135-8139 Dolan RJ, Fink GR, Rolls E et al. How the brain learns to see objects and faces in an impoverished context. Nature 1997;389: 596-599 Farah MJ. Is face recognition ’special’? Evidence from neuropsychology. Behav Brain Res 1996;76:181-189 Haxby JV, Hoffman EA, Gobbini MI. The distributed human neural system for face perception. Trends Cogn Sci 2000;4:223-233 Adolphs R, Baron-Cohen S, Tranel D. Impaired recognition of social emotions following amygdala damage. J
Cogn Neurosci 2002;14:1264-1274 34. LaBar KS, Crupain MJ, Voyvodic JT, McCarthy G. Dynamic perception of facial affect and identity in the human brain. Cereb Cortex 2003;13:1023-1033 35. Rossion B, Caldara R, Seghier M, Schuller AM, Lazeyras F, Mayer E. A network of occipito-temporal face-sensitive areas besides the right middle fusiform gyrus is necessary for normal face processing. Brain 2003;126:23812395 36. Baron-Cohen S. Mindblindness: an essay on autism and theory of mind. Cambridge, MA; MIT Press 1995 37. Grady CL, Keightley ML. Studies of altered social cognition in neuropsychiatric disorders using functional neuroimaging. Can J Psychiatry 2002;47:327-336 38. Baron-Cohen S. Social and pragmatic deficits in autism: cognitive or affective? J Autism Dev Disord 1988;18:379402 39. Baron-Cohen S, Ring HA, Wheelwright S et al. Social intelligence in the normal and autistic brain: an fMRI study. Eur J Neurosci 1999;11:1891-1898 40. Happe F, Ehlers S, Fletcher P et al. ‘Theory of mind’ in the brain. Evidence from a PET scan study of Asperger syndrome. Neuroreport 1996;8:197-201 41. Castelli F, Frith C, Happe F, Frith U. Autism, Asperger syndrome and brain mechanisms for the attribution of mental states to animated shapes. Brain 2002;125:1839-1849 42. Bowler DM. “Theory of mind” in Asperger’s syndrome. J Child Psychol Psychiatry 1992;33:877-893 43. Klin A. Attributing social meaning to ambiguous visual stimuli in higher-functioning autism and Asperger syndrome: The Social Attribution Task. J Child Psychol Psychiatry 2000;41:831-846 44. Fletcher PC, Happe F, Frith U, Baker SC, Dolan RJ, Frackowiak RS, Frith CD. Other minds in the brain: a functional imaging study of “theory of mind” in story comprehension. Cognition 1995;57:109-128 45. Brunet E, Sarfati Y, Hardy-Bayle MC, Decety J. A PET investigation of the attribution of intentions with a nonverbal task. Neuroimage 2000;11:157-166 46. Castelli F, Happe F, Frith U, Frith C. Movement and mind: a functional imaging study of perception and interpretation of complex intentional movement patterns. Neuroimage 2000;12:314-325 47. Gallagher HL, Happe F, Brunswick N, Fletcher PC, Frith U, Frith CD. Reading the mind in cartoons and stories: an fMRI study of ’theory of mind’ in verbal and nonverbal tasks. Neuropsychologia 2000;38:11-21 48. Vogeley K, Bussfeld P, Newen A et al. Mind reading: neural mechanisms of theory of mind and self-perspective. Neuroimage 2001;14:170-181 49. Saxe R, Kanwisher N. People thinking about thinking people. The role of the temporo-parietal junction in "theory of mind". Neuroimage 2003;19:1835-1842 50. Abu-Akel A. A neurobiological mapping of theory of mind. Brain Res Brain Res Rev 2003;43:29-40 51. Happe F. Autism: cognitive deficit or cognitive style? Trends Cogn Sci 1999;3:216-222 52. Hughes C, Russell J, Robbins TW. Evidence for executive dysfunction in autism. Neuropsychologia 1994;32: 477-492 53. Minshew NJ, Goldstein G, Siegel DJ. Neuropsychologic functioning in autism: profile of a complex information processing disorder. J Int Neuropsychol Soc 1997;3:303-316 54. Ring HA, Baron-Cohen S, Wheelwright S et al. Cerebral correlates of preserved cognitive skills in autism: a functional MRI study of embedded figures task performance. Brain 1999; 122:1305-1315 55. Luna B, Minshew NJ, Garver KE et al. Neocortical system Functional Neurology 2004; 19(1): 9-17
Functional brain imaging in autism
56.
57.
58.
59.
60. 61. 62.
63.
64.
65.
abnormalities in autism: an fMRI study of spatial working memory. Neurology 2002;59:834-840 Cohen MS, Kosslyn SM, Breiter HC et al. Changes in cortical activity during mental rotation. A mapping study using functional MRI. Brain 1996;119:89-100 Thompson-Schill SL, Jonides J, Marshuetz C et al. Effects of frontal lobe damage on interference effects in working memory. Cogn Affect Behav Neurosci 2002; 2:109-120 Manjaly ZM, Marshall JC, Stephan KE, Gurd JM, Zilles K, Fink GR. In search of the hidden: an fMRI study with implications for the study of patients with autism and with acquired brain injury. Neuroimage 2003;19:674-683 Vogt BA, Finch DM, Olson CR. Functional heterogeneity in cingulate cortex: the anterior executive and posterior evaluative regions. Cereb Cortex 1992;2:435-443 Smith EE, Jonides J. Storage and executive processes in the frontal lobes. Science 1999;283:1657-1661 Frith U. Cognitive explanations of autism. Acta Paediatr Suppl 1996;416:63-68 Muller RA, Pierce K, Ambrose JB, Allen G, Courchesne E. Atypical patterns of cerebral motor activation in autism: a functional magnetic resonance study. Biol Psychiatry 2001;49:665-676 Muller RA, Kleinhans N, Kemmotsu N, Pierce K, Courchesne E. Abnormal variability and distribution of functional maps in autism: an FMRI study of visuomotor learning. Am J Psychiatry 2003;160:1847-1862 Allen G, Courchesne E. Differential effects of developmental cerebellar abnormality on cognitive and motor functions in the cerebellum: an FMRI study of autism. Am J Psychiatry 2003;160:262-273 Muller RA, Chugani DC, Behen ME et al. Impairment of dentato-thalamo-cortical pathway in autistic men: language activation data from positron emission tomography. Neurosci Lett 1998;245:1-4
Functional Neurology 2004; 19(1): 9-17
66. Belmonte MK, Yurgelun-Todd DA. Functional anatomy of impaired selective attention and compensatory processing in autism. Brain Res Cogn Brain Res 2003;17: 651-664 67. Rivkin MJ, Vajapeyam S, Hutton C et al. A functional magnetic resonance imaging study of paced finger tapping in children. Pediatr Neurol 2003;28:89-95 68. Calandra-Buonaura G, Basso G, Gorno-Tempini ML et al. Human brain language processing areas identified by functional magnetic resonance imaging using a lexical decision task. Funct Neurol 2002;17:183-191 69. Heim S, Opitz B, Muller K, Friederici AD. Phonological processing during language production: fMRI evidence for a shared production-comprehension network. Brain Res Cogn Brain Res 2003;16:285-296 70. Stippich C, Mohammed J, Kress B et al. Robust localization and lateralization of human language function: an optimized clinical functional magnetic resonance imaging protocol. Neurosci Lett 2003;346:109-113 71. Honey GD, Fletcher PC, Bullmore ET. Functional brain mapping of psychopathology. J Neurol Neurosurg Psychiatry 2002;72:432-439 72. Ugurbil K, Toth L, Kim DS. How accurate is magnetic resonance imaging of brain function? Trends Neurosci 2003; 26:108-114 73. Attwell D, Iadecola C. The neural basis of functional brain imaging signals. Trends Neurosci 2002;25:621-625 74. Black JE, Jones TA, Nelson TA et al. Neuronal plasticity and the developing brain. In Alessi NE, Coyle JT, Harrison SI eds Handbook of Child Adolescent Psychiatry. Basic Psychiatric and Science and Treatment. New York; John Wiley & Sons 1998:31-53 75. Nelson KB, Grether JK, Croen LA et al. Neuropeptides and neurotrophins in neonatal blood of children with autism or mental retardation. Ann Neurol 2001;49:597-606
17