REVIEW Open Access Conceptual And Methodological .

5m ago
396.57 KB
10 Pages

Mazzone and Curatolo Behavioral and Brain Functions 2010, tent/6/1/17REVIEWOpen AccessConceptual and methodological challenges forneuroimaging studies of autistic spectrumdisordersLuigi Mazzone1*, Paolo Curatolo2AbstractAutistic Spectrum Disorders (ASDs) are a set of complex developmental disabilities defined by impairment in socialinteraction and communication, as well as by restricted interests or repetitive behaviors. Neuroimaging studieshave substantially advanced our understanding of the neural mechanisms that underlie the core symptoms ofASDs. Nevertheless, a number of challenges still remain in the application of neuroimaging techniques to the studyof ASDs. We review three major conceptual and methodological challenges that complicate the interpretation offindings from neuroimaging studies in ASDs, and that future imaging studies should address through improveddesigns. These include: (1) identification and implementation of tasks that more specifically target the neural processes of interest, while avoiding the confusion that the symptoms of ASD may impose on both the performanceof the task and the detection of brain activations; (2) the inconsistency that disease heterogeneity in persons withASD can generate on research findings, particularly heterogeneity of symptoms, symptom severity, differences inIQ, total brain volume, and psychiatric comorbidity; and (3) the problems with interpretation of findings from crosssectional studies of persons with ASD across differing age groups. Failure to address these challenges will continueto hinder our ability to distinguish findings that outline the causes of ASDs from brain processes that representdownstream or compensatory responses to the presence of the disease. Here we propose strategies to addressthese issues: 1) the use of simple and elementary tasks, that are easier to understand for autistic subjects; 2) thescanning of a more homogenous group of persons with ASDs, preferably at younger age; 3) the performance oflongitudinal studies, that may provide more straight forward and reliable results. We believe that this would allowfor a better understanding of both the central pathogenic processes and the compensatory responses in the brainof persons suffering from ASDs.BackgroundAutistic Spectrum Disorders (ASDs) are a class of conditions that embodies Autistic Disorder, Asperger’s syndrome, and Pervasive Developmental Disorder NotOtherwise Specified (PDD-NOS). Each of these conditions is defined by the presence of complex developmental disabilities that include qualitative impairmentsin social interactions (i.e. impaired use of non-verbalbehaviors, failure to develop peer relationships, andpoor social reciprocity) and in communication (i.e.,delay in development of spoken language, inability tosustain a spoken conversation, stereotyped use of* Correspondence: [email protected] of Child Neurology and Psychiatry, Department of Pediatrics,University of Catania, Catania, Italylanguage, paucity of symbolic or imitative play), andrestricted or stereotyped interests and behaviors. Persons with ASDs must manifest symptoms by the age ofthree years. Intelligence Quotients (IQ) vary widely, butthe overall prevalence of intellectual disability in thispopulation is around 50-75% [1].Recent years have witnessed innovative approaches tothe study of ASDs, driven by the emergence of newtechnologies and methodologies for studying both normal and pathological development in children and adolescents. Among these, some of the most importanttechniques include anatomical magnetic resonance imaging (MRI) and functional MRI (fMRI), which can revealanatomical and functional abnormalities in brain development. Both of these MRI modalities have played a 2010 Mazzone and Curatolo; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (, which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.

Mazzone and Curatolo Behavioral and Brain Functions 2010, tent/6/1/17major role in advancing our knowledge of the neuralbases of ASDs [2].Anatomical studies have documented in persons withASDs increases in brain volumes, particularly in theposterior regions, and especially in the right hemisphere [3-7]. These observations are consistent withwell-replicated findings of enlarged head circumferencein very young autistic children as compared to healthychildren, that seems to derive from accelerated headgrowth at 6-14 months of age, despite normal or smaller head circumference at the time of birth [8-12].Additional anatomical imaging studies have reportedlarger volumes of both white and gray matter in thefrontal cortex [13], larger volumes of the caudatenucleus even after covarying for overall brain size[14,15], and abnormal volumes (usually larger) of theamygdala and hippocampus [16,17]. Finally, both smaller [18] and larger volumes [19,20] of the cerebellumhave been reported.Besides neuroanatomical studies, a large number offunctional neuroimaging studies, designed to investigatethe neural mechanisms underlying the core symptomsof ASDs, have recently been published.Despite the great contribution that these imagingtechnologies have provided to the knowledge of morphologycal features in persons with ASDs, the uniquephenotypic trademarks of these disorders create peculiardifficulties in designing imaging studies that providedata with univocal interpretation and new insight intothe pathogenesis of these conditions. Some of theseinclude: the difficulty that persons with ASDs have inperforming the tasks used in fMRI studies in a way thatis comparable to normal subjects; the heterogeneity inpersons with ASD who are enrolled in research studies;and the problematic interpretation of findings fromcross-sectional studies of persons with ASDs whobelong to differing age groups.Gaining a better understanding of the nature of thealterations in brain functions that occur in ASD couldbe of vital importance to try to identify the link betweenanatomical and functional abnormalities in the brain ofpersons with ASDs and their behavioral phenotypes.Our intent is to review in a critical way the knowledgeaccomplished so far in functional imaging studies ofASDs as well as to identify the gaps that still need to beaddressed, trying to overcome the problems and challenges that may have limited our knowledge on thepathogenesis of this type of disorders.fMRI task designSeveral fMRI studies, many published in the last decade,used specifically designed tasks to investigate abnormalities in discrete neural systems in persons with ASDs. Inparticular, some studies have investigated the role playedPage 2 of 10by specific perceptual, cognitive and attentional processes underlying the executive functions [21,22]. Otherstudies used specific tasks for language-comprehensionto investigate functional connectivity and semantic processes [23,24]. Finally, a great number of studies usedtasks involving face perception or emotional processes(table 1) to identify brain disturbances that mightaccount for the profound impairments, typical of persons with ASDs, to interact socially and to recognize theemotions of others. Findings from these studies, however, have been inconsistent and often contradictory.The inconsistencies across studies likely derive from thedifficulty that persons with ASDs have with processingstrategies, and from the degree of arousal, effort, frustration, or confusion that they manifest while performingthe task. This could represent one of the reasons whypersons with ASDs turn out in showing such a variablebrain activity during functional imaging tasks. Moreover,the activations associated with these differences couldbe linked to epiphenomenal effects that are associatedwith differing performance levels across groups on agiven task. Finally, it would probably be useful to selectan appropriate control task suitable for the primary taskof interest.An example: face perception tasksMany behavioral studies have consistently shown thatpersons with ASDs are selectively impaired in their ability to recognize faces [25-33]. In addition to this, recentfMRI studies have shown that this selective impairmentis associated with abnormal patterns of brain activation.Most of these studies consistently reported an atypicalpattern of activation in the fusiform gyrus, which isextensively activated during face processing in healthyindividuals but seems to be much less activated duringthe same tasks in individuals with ASDs [34,35]. One ofthe first neuroimaging studies of face processing in individuals with ASDs, reported decreased activation in thefusiform gyrus during face discrimination as comparedto control subjects [36]. This study also reported thatpersons with ASDs activate object-processing regions,such as the right inferior temporal gyrus, when viewingfaces. However, a recent clinical study showed that children with ASDs spend more time looking at an adult’smouth instead of gazing into the eye [37] and manyother clinical studies have used eye tracker to detectwhat participants were looking at during the face discrimination task. Nevertheless, before 2005, no fMRI studyhad reported on the relationship between the directionand duration of gaze fixation and the patterns of brainactivation during the processing of human faces in individuals with ASDs. The first study to address this point[38], showed that the activation of the fusiform gyrusduring face perception in ASD patients correlated positively and strongly with the duration of the participants

Mazzone and Curatolo Behavioral and Brain Functions 2010, tent/6/1/17Page 3 of 10Table 1 MRI studies published between 2005 and 2009 exploring facial processing in samples of at least 10 patientswith ASD, as determined by a PubMed searchSTUDYAUTISMGROUPN SexAge (M SD)CONTROLGROUPN SexAge (M SD)TASK DESIGN[Cognitive process]RESULTS IN ASD GROUP13 M(31.2 9.1)13 M(25.6 5.1)Faces expressing different intensities of fear[Emotional cognitive process]Abnormal functional connectivity of medial TL14 M(15.9 4.71)16 M(14.5 4.60)12 M(17.1 2.78)16 M(14.5 4.56)Facial emotion discrimination task[Emotional cognitive process]Facial recognition task[Cognitive perception]Activation in the FG and AMY strongly andpositively correlated with the time spent fixating theeyes in both studiesDapretto Met al., 2006[65]9M1F(12.05 2.5)9M1F(12.38 2.2)Face emotional recognition task[Emotional cognitive process]No activation in the IFGBölte S. etal., 2006[66]5M(29.4 5.9)5 Ma(25.8 8.0)Face recognition pre and post after FEFA[Emotional cognitive process]No significant activation changes in the FG pre- andpost-trainingBird G. et al.,2006[67]14 M 2 F(33.3 11.5)14 M 2 F(35.3 12.1)Task in which pairs of face and house stimuliwere present on every trial[Social and cognitive perception]Failure of attention to modulate connectivitybetween extra striate areas and V1Wang AT etal., 2007[68]18 M(12.5 2.9)18 M(11.8 1.9)Irony comprehension[Social and cognitive perception]Reduced activity in the medial PFC and right STGAshwin C. etal., 2007[69]13 M(31.2 9.1)13 M(25.6 5.1)Perception of fearful faces[Emotional cognitive process]Increase in the ACC and al.,2007[70]8M2F(34 11)4M3F(35 12)Passively viewing non emotional faces[Cognitive perception]Significant activation of FG and IOG; Hypoactivationin right AMY, IFC, STS, and face-relatedsomatosensory and premotor cortexDichter GSet al., 2007[71]16 M 1 F(22.9 5.2)14 M 1 F(24.6 6.5)Reaction time to arrow or gaze stimulus withsimilar flanker stimuli oriented (congruent orincongruent directions)[Cognitive perception and control]Hypoactivation in MFG, right IFG, bilateralintraparietal sulcus, and the ACC during incongruentgaze stimuliKoshino al., 2008[72]11 M(24.5 10.2)10 M 1 F(28.7 10.9)n-back working memory task involving facerecognition[Working memory]Hypoactivation in the left IPFC and in the right PTCKleinhansNM et al.,2008[73]19(23.5 7.8)21(25.1 7.6)Facial emotion discrimination (familiar,unfamiliar and new friend)[Emotional cognitive process]No between-group differences in fusiform activationto faces or houses;Significant FG-AMY and FG-STS functionalconnectivityPinkham AEet al., 2008[74]12 M(24.1 5.7)12 Mb (27.1 3.9)12 Mc (26.4 5.2)12 Md (28.0 3.9)Complex social judgments of faces[Social and cognitive perception]Reduced activation in the right AMY, FG, VLPFCHumphreysK. et al.,2008[75]13 M(27 10)15 M(29 10)1) Conventional face and object mapping2) Motion pictures experiment[Emotional cognitive process and perception]Decreased of activation, not only in fusiform facearea but also in STC and occipital areaUddin LQ etal., 2008[76]12 M(13.19 2.6)12 M(12.23 2.10)Responsiveness to images of the subjects’own face and of others’ faces[Emotional perception]Activation of right PM/PF system while viewingimages containing mostly their own faceBookheimerSY et al.,2008[77]12 M(11.3 4.0)12 M(11.9 2.4)Subjects had to match faces presented in theupright versus and inverted position[Cognitive perception]No differences in the FG;Decrease in activation in left PFC;No activation in AMY for upright taskWelchew DEet al., 2005[64]Dalton KMet al., 2005[38]Study IStudy II

Mazzone and Curatolo Behavioral and Brain Functions 2010, tent/6/1/17Page 4 of 10Table 1: MRI studies published between 2005 and 2009 exploring facial processing in samples of at least 10 patientswith ASD, as determined by a PubMed search (Continued)Pierce K. etal., 2008[78]9M2F(9.9)9M2F(9.8)Pictures of a familiar adult or child, strangeradult or child, objects[Emotional cognitive proce