T144. Correlation Between Cerebellar-Prefrontal Network Connectivity and Negative Symptoms Associated With the Stages of Schizophrenia

Abstract

Background To identify different network-symptom relationship may be a model of how precision medicine approach. In the July 2019 issue of AJP, Brady et al. successfully identified a network biomarker of negative symptom severity in a sample of patients with schizophrenia using resting state functional connectivity (FC) fMRI, which pushed forward the psychiatric research from purely correlational studies. As we known, medications and different episodes of the disease affect the brain function, so the breakdown of cerebellar-prefrontal network connectivity may not directly correspond to negative symptom. At this point, we would strengthen empirical support for a causal relationship between dysfunctional connectivity and psychopathology using different stages of patients with schizophrenia. Methods We used three independent cohorts with schizophrenia, one of which is 54 medication-naïve patients with first episode schizophrenia (FES) in Shanghai Mental Health Center (SH), the other is 112 medicated patients with FES in Guangzhou Huiai Hospital (GZ1), and the third is 35 chronic patients with schizophrenia in Guangzhou Huiai Hospital (GZ2). All patients were Mandarin-speaking Han Chinese and met the criteria for schizophrenia disorder based on the structured clinical interview for DSM-IV-TR. Negative symptom severity was assessed with the Scale for Assessment of Negative Symptoms (SANS) in SH cohort and with the Positive and Negative Syndrome Scale (PANSS) in GZ1 and GZ2 cohorts. MRI imaging was respectively conducted on Siemens 3.0-T (SH) and Philips 3.0-T (GZ1+GZ2) MRI systems. MRI data were preprocessed and analyzed using the DPABI toolbox, which are the same as the study of Brady et al. Furthermore, we selected the regions of interest (ROI) from the results of Brady et al. for our ROI-wise functional connectivity analysis: right dorsolateral prefrontal cortex (dlPFC, 36, 24, 30), left dorsolateral prefrontal cortex (-33, 30, 42) and midline cerebellar cortex (-9, -96, -27). Results We modeled the effect of negative symptom severity on FC between left or right dlPFC and midline cerebellar cortex while covaried the effects of head motion, age and sex. Our results showed that FC between right dlPFC and cerebellar cortex positively correlated with negative symptom severity in medication-naïve patients (SH cohort: r = 0.343, p = 0.014) and the FC tended to be significantly positively correlated with negative symptom severity in medication patients with FES (GZ1 cohort: r = 0.179, p = 0.061). However, in chronic patients with schizophrenia, the FC between right dlPFC and cerebellar cortex negatively correlated with negative symptom severity (GZ2 cohort: r = -0.390, p = 0.021). The FC between left dlPFC and cerebellar cortex did not correlate with negative symptom severity in all three cohorts. Discussion Our data of chronic patients with schizophrenia validated Brandy’s findings: negative symptom severity was found inversely correlated with FC between right dlPFC and cerebellar cortex. However, our results showed a significantly positive correlation in medication free cohort, and the significance become weaker in the medicated cohort. Our finding proved that the prefrontal cortex – cerebellum network circuit linked directly to negative symptoms, further it is a positive correlation in FES patients, and it is a negative correlation in chronic patients which are the same as Brandy’s. We speculate that medication affects function of the brain network, and then reverses the correlationship between it and the symptoms. We will follow up our two cohorts of FES patients for further fMRI data collection and symptom assessment, and test whether the network could correspond to negative symptom severity.