Precuneus Hyper-Perfusion Relates to Symptom Severity and Hypoconnectivity in Individuals with Autism Spectrum Disorder

Background: Social deficitits in Autism Spectrum Disorder (ASD) have been associated with abberrant patterns of blood oxygenation level dependent (BOLD) resting state functional connectivity MRI (rs-fcMRI) (Fair, 2007; Assaf, 2010).  More recent work has used Arterial Spin Labeled (ASL) perfusion MRI to assess functional connectivity by quantifying cerebral blood flow (CBF), which is inaccessible to conventional BOLD. Comparisons of CBF in resting state networks in ASD and TD individuals may elucidate mechanisms of network dysfunction in ASD. To our knowledge, no study to date has investigated perfusion-based resting state networks in ASD. 

Objectives: This study aims to explore differences in CBF-based functional connectivity in ASD and TD individuals, and how these differences may be related to severity of social symptoms assessed with the Social Responsiveness Scale (SRS; Constantino, 2003). 

Methods:  Eleven high-functioning children and adolescents with ASD (age [years] mean±SD: 12.93±0.95; 4 unmedicated, 3 female) and eleven typically developing (TD) participants (13.73±3.42, 1 female) were recruited. MRI data was collected on a 3-T Siemens Trio Scanner. The imaging protocol included a T1-weighted structural MRI and resting-state background-suppressed 3D GRASE pseudo continuous ASL (Fernandez-Seara, 2008). Standard preprocessing and CBF quantification were performed. Data were restricted to grey-matter voxels and zero-meaned before subjected to a Group ICA using GIFT (Calhoun, 2001). ICA model order was determined using the AIC/MDL criterion. The default mode network (DMN) was identified using GIFT component labeller and visual inspection. DMN_CBFwas extracted at z-threshold>2 from each participant, controlling for globalCBF and total intracranial volume. Multiple regression analyses were performed on the TD and ASD groups together and separately. SRS scores were correlated with average baseline perfusion controlling for age, sex and global CBF. Moreover, these regional CBF values were correlated to functional connectivity maps.

Results:  ASD and TD participants showed highly similar DMNs; however, differences were observed in the posteromedial cortex (dorsal precuneus), where individuals with ASD showed reduced connectivity compared to TD individuals. Furthermore, in the ASD group, SRS scores were positively correlated with CBF rs-fcMRI in the dorsal precuneus such that greater social impairment was related to increased CBF (p< 0.01; t(6)=3.14). CBF values from the precuneus ROI were also positively correlated with SRS total score in ASD (r=.856, p<.007). Moreover, voxel-wise correlation of the precuneus ROI CBF values with individual ASD functional connectivity maps revealed a widespread reduction of DMN connectivity with increased precuneus CBF.

Conclusions: Our results support previous findings of precuneus hypoconnectivity in ASD, and a relationship to social behavior (Lynch, 2013). The relationship between social deficits and CBF in the precuneus suggests dysfunctional hyperactivity in this area in ASD. Decreased connectivity has been observed in ASD between the precuneus and the dorsal lateral and medial prefrontal cortex, regions relevant for social processing and executive/memory functions, where individuals with ASD typically show deficits (Kennedy, 2008). Taken together, our results suggest that hyperperfusion in this area might lead to altered rs-fcMRI and highlight the need to investigate underlying neuro-physiological differences in ASD.