10.2 USING THE RDOC PARADIGM TO IDENTIFY SOCIAL COGNITIVE BRAIN-BEHAVIOR RELATIONSHIPS IN PEOPLE WITH SCHIZOPHRENIA SPECTRUM DISORDERS. (9th April 2019)

• Record Type: Journal Article
• Authors: Voineskos, Aristotle
• Is Part Of: Schizophrenia bulletin. Volume 45(2019)Supplement 2
• Page Start: S102
• Page End: S103
• Publication Date: 2019-04-09
• Subjects: Schizophrenia -- Periodicals; Schizophrenia -- Research -- Periodicals; 616.898005
• DOI: 10.1093/schbul/sbz022.035
• Languages: English
• ISSNs: 0586-7614

Abstract

Abstract: Background: The case-control model in psychiatric research, most notably in schizophrenia, has not led to biomarker identification (recent retrospective mega-analyses notwithstanding) or therapeutic innovation. The Research Domain Criteria (RDoC) approach argues for using a dimensional strategy toward identification of brain-behavior subtypes, less-wedded to diagnostic categories, that may translate better toward therapeutic innovation. Here, a large, multi-center, prospectively designed study aiming to identify brain-behavior subtypes in the social processes RDoC domain will be described. Methods: The prospective design of a large, multi-center study that aims to recruit 300 people with a schizophrenia spectrum disorder and 180 healthy controls (total n=480) with a full range of social functioning from highly impaired to high performing will be described. Inter-site harmonization for clinical, social cognitive, neurocognitive, social functioning, and most important, neuroimaging approaches will be described including a traveling human phantom study. Imaging tasks aiming to identify lower-level and higher-level social cognitive circuit brain function and social cognitive tasks out of the scanner aim to identify lower-level and higher-level constructs with the overall aim of relation connectivity within and among social brain circuits with social cognitive performance out of the scanner. In addition, more complex tasks are being used to disentangle the relationship among social cognitive circuits and their relationship to social cognitive performance and social function out of the scanner. Results: We were able to successfully combine neuroimaging data across sites, bolstered by results from our traveling human phantom study, showing that T1-weighted, diffusion-weighted, and fMRI data, through a clustering algorithm successfully identified each individual's scan. Combining data across sites, we found significantly different neural strategies related to efficiency across the full range of social impairment, independent of diagnosis in lower-level social cognitive tasks in the MRI scanner. We also found that machine learning approaches incorporating both lower-level and higher-level social cognitive networks during the resting state fMRI condition successfully distinguished between poorer and better social function. Finally, we found that an fMRI task engaging both lower and higher level social cognitive networks added to the neural strategy heuristic, regarding optimal network utilization based on lower vs. higher level task performance. In this work we found significant positive associations between corresponding putative circuits and social cognitive performance (e.g. higher-level circuit functional connectivity was associated with higher level task performance), and inverse association of the lower level circuit functional connectivity with higher level task performance. Conclusions: Overall, our results support the value of prospectively designed multi-center research work using clinical neuroimaging approaches. Our results also support the value of the RDoC approach, such that brain-behavior subtypes related to brain network utilization and task demand can identify groups of people based on social cognitive performance and social function. The work also fulfills the RDoC mandate in that brain circuitry that is targetable and is associated with performance in a behavioral domain, can be carried forward into target engagement studies to test promising new interventions.