IMAGING GENETICS TOWARDS A NEW DIAGNOSIS OF SCHIZOPHRENIA

Background Current diagnoses of Schizophrenia (SZ) and related psychiatric disorders are catalogued by phenomenological principles and clinical descriptions. Complicated gene and environmental factor interactions likely play an important role in the onset of SZ and induce great heterogeneity. This unclear etiology and heterogeneity prevent the establishing of accurate subgroups within SZ through clinical observation alone; simultaneously, the boundaries among psychiatric disorders previously drawn by phenomenological findings are merging because there is great overlap, especially in genetic variations and brain imaging findings. By characterizing and quantifying brain regions affected in psychiatric disorders, imaging genetics is a valuable research method to fulfill a new diagnostic system. We review contributions of imaging genetics to SZ diagnosis, limitations, and possible future directions. Methods Among numerous imaging genetic studies, we reviewed recent studies with relatively clear and consistent findings. They were either done in task-phase imaging to indicate certain cognitive impairments of SZ, or the genes included were highlighted in SZ itself or an important cognitive domain related to SZ. Results Some pioneering studies contribute to subtypes and disorder boundaries. The clustering study from Arnedo et al. formed correlations between disjoined networks of symptom groups and Single Nucleotide Polymorphism (SNP) sets, and it may be promising for future SZ subtypes. Bipolar-SZ Network on Intermediate Phenotypes studies combining various clinical biomarkers and brain imaging divided SZ, bipolar disorder and schizoaffective disorder into three different “biotypes” regardless original diagnoses, and it provided meaningful insight into the disorder boundaries. For imaging genetics research, large numbers of associations between functional imaging intermediate phenotypes and genes have been found in working memory, episodic memory, emotion, attention, cognitive control, and theory of mind in SZ or healthy risk allele carriers. Such associations were difficult to identify while using structural brains measures. Discussion Imaging genetics will continue to shape the diagnosis of SZ and related disorders. However, there are still problems to solve. Most importantly, the logic behind imaging genetics has been questioned. Franke et al. reminds us the structural brain deficits believed to be important pathological alterations of SZ may reflect prenatal and later development environmental effects not specific to SZ, or the diagnostic category of SZ may not be uniformly organized. The following aspects need future effort. Firstly, more brain measures need to be introduced, and the criteria of intermediate phenotypes should be carefully checked. The specificity of brain alterations will need careful examination in clinical settings. Secondly, more genetic factors need to be collected. Polygenic risk scores, parallel independent component analysis and other clustering method may be promising. Factors other than SNPs, such as copy number variations and regulatory factors will need to be included. Finally, imaging genetics will need compatible phenomenological assessment scales which need not distinguish current diagnoses but provide a comprehensive scan of clinical features.