According to the Diagnostic and Statistical Manual of Mental Disorders (DSM-5), autism spectrum disorder (ASD) is a neurodevelopmental disorder, characterized by the presence of impaired social communication and unusually repetitive behaviors or restricted interests (American Psychiatric Association, 2013).

However, the term ASD includes a wide variability of clinical presentation (Lord et al., 2000), which seems to reflect a multifactorial etiopathogenesis, still unclear, encompassing different genetic risk factors and susceptibility to environmental factors (Dietert et al., 2011; Tordjman et al., 2014).

Over the years, disruption of brain development and the maturation time course has been recognized as mechanism of autism (Amaral et al., 2008), suggesting the existence of common neurobiological substrates, such as defective synaptic structure and aberrant brain connectivity (Ecker et al., 2016).

Therefore, several studies have assessed the utility of brain magnetic resonance imaging (MRI) in the evaluation of region-specific structural changes and quantification of specific alterations in gray (GM) or white matter (WM) [in terms of cortical thickness, gyrification index, GM surface and volume (Pagnozzi et al., 2018)].

The identification of an MRI biomarker could provide insight into the underlying mechanisms of symptoms and could additionally provide crucial support in early clinical diagnosis and dividing patients, sharing common features, into different groups to tailor specific interventions.

In literature, consensus exists on increased growth of total GM and WM volume (Courchesne et al., 2001; Sparks et al., 2002; Hardan et al., 2006; Hazlett et al., 2011; Nordahl et al., 2012; Mahajan and Mostofsky, 2015; Ismail et al., 2016; Libero et al., 2016; Lucibello et al., 2019), which involves specific brain regions, such as temporal lobes (Lai et al., 2013; Ecker et al., 2014; Riddle et al., 2017; Pappaianni et al., 2018), frontal lobes (Ecker et al., 2014; Zhou et al., 2014; Eilam-Stock et al., 2016; Patriquin et al., 2016), or both (Katuwal et al., 2015; Postema et al., 2019).

A comparable number of studies finds opposite results, reporting decreased volumes of the frontotemporal (Hardan et al., 2006; Ecker et al., 2010; Mak-Fan et al., 2012), and parietal cortex (Ecker et al., 2010), significantly smaller right (Haznedar et al., 1997), and bilateral (Ecker et al., 2010; Jiao et al., 2010), anterior cingulate gyrus, and decreased cortical volume in the orbitofrontal cortex bilaterally (Ecker et al., 2012).

Recently, many studies have focused their efforts on measurement of cortical thickness (Doyle-Thomas et al., 2013; Khundrakpam et al., 2017; Pereira et al., 2018; Prigge et al., 2018), and among them, the one with the largest ASD population (Van Rooij et al., 2018), showed promising results. Specifically, the authors found complex developmental trajectories involving different brain regions with significant differences in terms of increased cortical thickness in the frontal cortex and decreased thickness in the temporal cortex between ASD patients and controls during adolescence. Other authors focus their study on cortical gyrification, suggesting the involvement of both genetic and non-genetic factors in definition of cortical gyral and sulcal patterns (Lohmann et al., 1999; White et al., 2002; Kates et al., 2009; Kremen et al., 2010; Mata et al., 2010; Docherty et al., 2015; Kuhn et al., 2016; Bernardoni et al., 2018; Duan et al., 2020; Kruggel and Solodkin, 2020). Although analysis of gyrification has led to the identification of different cortical regions involved in ASD, there is high heterogeneity across studies.

However, no studies have investigated the correlation between clinical, genetic, and radiological findings in a well-selected pediatric population.

After image quality check, five patients were excluded. Data from 34 patients, 4 females (11.8%) and 30 males (88.2%), were eventually included in the analysis. Their age range was between 3.4 and 6 years. All subjects of the sample selected had normal MRI findings and normal FMR1 and CGH array analysis.

Seven patient (20.6%) had a normal intellectual quotient, 12 (35.3%) a mild intellectual disability, and three (8.8%) a moderate intellectual disability.

About ADOS valuation, 5/34 ASD patients (14.7%) had a low level of severity, 20/34 (58.8%) had a medium level of severity, and 5/34 (14.7%) had a high level of severity (Table 1).

Autism is a complex and heterogeneous neurodevelopmental disorder with widely varied clinical features and a multitude of possible etiological factors.

Emerging evidence supports that ASD undergoes an atypical trajectory of brain maturation (Doyle-Thomas et al., 2013; Ecker et al., 2014) or abnormal lateralization of brain networks (Conti et al., 2015). Findings of atypical brain morphology in ASD, however, are highly heterogeneous and there are still not clear neuroanatomical markers to accurately identify individuals with ASD.

Whereas most neuroradiological studies on autism have focused on finding a specific pattern of brain involvement, only a few studies aimed to establish a correlation between clinical severity and cortical thickness (Doyle-Thomas et al., 2013). Additionally, many of these studies exclusively investigate well-known cortical areas involved in specific function, whereas we have tried to assess the correlation of several cortical areas to verify some correspondence with already known brain circuits. Whereas some results have reinforced well-known concepts already widely described in literature, others are quite difficult to frame.

The correlation between cingulate and frontal regions, particularly the left medial orbito-frontal area, with a social affect score of both ADOS and ADI-R scales is not surprising as the involvement of this area in social-affective processing is already described as part of the ventral social affective processing system (Belger et al., 2011). These results are consistent with what is already known about cortical gyrification morphology in neurodevelopmental disorders (Kohli et al., 2019).

Less clear is the correlation with other areas, such as the left lateral-occipital area, even if other authors have already shown this result in general in ASD patients (Kohli et al., 2019).

Among the results obtained, it is useful to underline a partial congruence between the results of the ADOS2 and ADI-R subtests investigating social affective behaviors and cortical gyrification index. This highlights how the two diagnostic tests support each other and how it is essential to always try to perform a complete patient assessment.

In the same way, it is easy to understand the correlation found between the restricted and repetitive behavior indices of both scales with gyrification of para-hippocampal, temporal, and middle frontal area.

Many human neuroimaging studies have indeed shown increased activation in ventral striatum and ventromedial frontal cortex in response to unexpected negative feedback that implies a change of behavior strategy (Glascher et al., 2009; D’Cruz et al., 2016; Sharda et al., 2017).

Conversely, it is unclear how the cingulate cortex could be involved in communication tasks although some neuroimaging studies suggest the association with this region and indirect language and social inferential capacity, such as the comprehension of non-literal language or construction-based pragmatic information (D’Cruz et al., 2011).

In general, the assessment of cortical areas based on the ADOS and ADI scores cannot be univocal. These scores, indeed, assess other skills, such as language or different non-verbal communication ability that could be underpinned by more cortical areas. Certainly, many of these areas, such as temporal and frontal or middle temporal areas, are already described as involved in patients with ASD.

This study confirms and expands previous results while reducing the variability of cohort patients as, in other studies a systematic selection of ASD subjects was not performed, and combining imaging with more detailed clinical data. The well-selected population of our study allowed us to reduce the variability observed between different ages.

Exclusion criteria, such as dysmorphic features or epilepsy or other major neurological signs, have led to less variability related to secondary factors, and focusing on only preschool-age patients has reduced the bias linked to the noted changes in the thickness of the cerebral cortex related to age (Shibata et al., 2010).

Neuroimaging studies indeed report numerous region-specific alterations in cortical thickness in patients with ASD. However, there are many inconsistent findings, and this is probably due to atypical CT developmental trajectories in ASD, characterized by decreased cortical thinning during early adolescence and increased thinning at later stages, involving mostly frontal and parietal areas (Mensen and Khatami, 2013; Ecker et al., 2014).

Limitations of our study include a small sample size, which reduces the power of statistical analysis.

Though the study has investigated the brain structure–function correlation with the aim of catching a possible clinical-related sensitivity marker at this age, the lack of controls constitutes a limitation of the study. The need for sedation during MRI scans, indeed, limits the opportunity to collect age-matched typical subjects.

Moreover, because pediatric brains are different in size and shape from adult brains commonly used as frameworks for spatial normalization (e.g., Talairach space), specific cortical areas could show correlation of deformation with age. In this context, several studies assessed the influence of age on various spatial normalization parameters (Wilke et al., 2002; Fonov et al., 2011), and consequently, our results should be taken with caution.

In conclusion, our results confirm that recent brain MRI techniques integrated to clinical data can provide some useful information in defining the cerebral regions involved in ASD although large sample studies characterized by homogeneous analytical and multisite approaches are needed.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

The studies involving human participants were reviewed and approved by Ethics Committee of Fondazione Policlinico Gemelli Hospital – Catholic University, Rome, Italy. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin.

RB, SL, and EM contributed to conception and design of the study. SL, TV, and GB organized the database. EP, MC, DC, and RD’A contributed to investigation. AN and ML performed the statistical analysis. SL wrote the first draft of the manuscript. TV and AN wrote sections of the manuscript. PM, CC, and EM contributed to the supervision of manuscript. All authors contributed to manuscript revision, read, and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.