PROPOSAL FOR NOVEL PATIENT CLASSIFICATION WITH ENLARGED LATERAL VENTRICLES AND COGNITIVE DECLINE BASED ON DATA-DRIVEN ANALYSIS OF NEUROIMAGING IN PSYCHIATRIC DISORDERS AND CLINICAL CHARACTERIZATION

Abstract Background The creation of a novel diagnostic system using objective biomarkers is expected to take place, due to difficulty of the current diagnostic system based on symptoms in psychiatry. Aims & Objectives We clustered the 5602 subjects based on brain subcortical volumes and showed a four-biotype classification, namely extremely (Brain Biotype [BB] 1: n=110) and moderately smaller limbic regions (BB2: n=566), larger basal ganglia (BB3: n=679), and normal volumes (BB4: n=4247), being associated with cognitive/social functioning (Mol Psychiatry, 2023). In the present study, we performed a detailed examination of the biological and clinical characteristics of these biotypes. Method A total of 5602 patients in the previous paper, including healthy subjects (n=3077), patients with schizophrenia (SZ: n=1499), bipolar disorder (BD: n=234), major depressive disorder (MDD: n=598), autism spectrum disorder (ASD: n=189) and other psychiatric disorders (O: n=5), were included. The biological and clinical characteristics of four biotypes were investigated using subcortical volumes from T1-weighted MRI images, cognitive functions, EEG, and genomic information. Results We examined the differences in subcortical volume characteristics in each biotype and found that the lateral ventricle volume was remarkably larger (Z-score = 5.1 ± 1.8) in BB1 (n=110). 99.1% of BBI subjects had the lateral ventricle volume greater than 3 standard deviation (3 SD) of normal subjects, and this brain feature discriminated BB1 with a sensitivity of 99.1% and a specificity of 98.1%. Therefore, BB1 was redefined as a brain structural feature with a lateral ventricular volume of 3 SD or greater. Of the 186 redefined subjects, 33 patients with detailed cognitive function data were stratified into two groups: 9 patients with moderate or higher cognitive decline (SZ8/ASD1) and 24 patients with mild or lower cognitive decline (SZ9/ HC9/ASD4/ MDD1/O1), and compared. EEG abnormalities (71% in the former group, 0% in the latter), a diagnosis of schizophrenia (89% in the former group, 38% in the latter group), and low intelligence (IQ=67 in the former group, IQ=103 in the latter group) were obviously observed in the former group than in the latter. Furthermore, copy number variation analysis of 7 out of 9 cases in the former group showed a high rate of pathological CNV (22q11.21 deletion, 7q11.23 duplication, and DPP6 deletion) in 3 cases, which is only seen in approximately one out of several thousand normal subjects. Discussion & Conclusion We have identified a novel candidate for classification of psychiatric disorders with high rates of EEG abnormalities, psychotic symptoms, and pathological rare genomic variants in a stratified group of patients with schizophrenia and other psychiatric disorders with lateral ventricular volumes of 3 SD or greater and moderate to severe cognitive decline. We proposed a possible novel psychiatric disorder by going back to each patient based on the clinical significance from a large-scale data-driven analysis screening, and proposed a methodology to find the concept and psychiatric disorders by objective indices for the first time in the world.This new biotype classification suggests that objective measurements based on neuroimaging data will be incorporated into psychiatric diagnosis in the future.