Arterial spin labeling provides a reliable neurobiological marker of autism spectrum disorder

• Main Authors: Benjamin E. Yerys, John D. Herrington, Gregory K. Bartley, Hua-Shan Liu, John A. Detre, Robert T. Schultz
• Format: Article
• Language: English
• Published: BMC 2018-12-01
• Series: Journal of Neurodevelopmental Disorders
• Subjects: Social cognition; Social perception; MRI; Autism; Faces; Blood flow
• Online Access: http://link.springer.com/article/10.1186/s11689-018-9250-0

Abstract Background Research on neurobiological markers of autism spectrum disorder (ASD) has been elusive. However, radionuclide studies of cerebral blood flow (CBF) have shown decreased blood flow (hypoperfusion) in the temporal lobes of individuals with ASD across ages and intelligence. This observation fits with current neuroscientific models that implicate temporal regions in social perception and social cognition. Arterial spin labeled perfusion MRI allows noninvasive quantification of regional CBF as part of a multimodal MRI protocol. This method is almost entirely absent from ASD research to date. Our a priori hypothesis was that children with ASD would present with hypoperfusion in the temporal lobes—most notably the fusiform gyrus (given its prominent role in ASD social perception deficits). We also sought to examine the reproducibility of CBF measures, and their relationship to individual differences in facial recognition and ASD symptoms. Methods A total of 58 males (33 with ASD) between the ages of 12 and 17 years participated in the study. All children completed two arterial spin labeling and structural (T1) scans using a 3 T Siemens Verio scanner approximately 8 weeks apart, as well as behavioral testing at time 1 that included diagnostic measures and the Benton Facial Recognition Test. CBF was the key dependent variable, as was facial recognition performance, and ASD symptoms. The two scans were used for reliability analyses. Results The ASD group showed hypoperfusion in the bilateral fusiform gyrus and in right inferior temporal gyrus. Intra-class correlations showed moderate to good reliability across time within both groups, and no diagnostic group × time interactions. CBF in the left fusiform gyrus was significantly positively correlated with facial recognition. No significant correlations were observed with core ASD symptoms. Conclusions Arterial spin labeling revealed hypoperfusion in children with ASD in regions critical to social perception and cognition. The left fusiform gyrus plays an important role in facial recognition, and greater CBF in this region was correlated with more normative facial recognition performance in children with ASD. This study takes an important first step in establishing CBF of the temporal lobes as a reliable marker of ASD.