Neuroimaging biomarkers for the prediction of treatment response in schizophrenia

Abstract

Schizophrenia is a severe debilitating mental illness that affects approximately 1% of people worldwide. Compared to healthy controls (HC), patients with schizophrenia (SZ) have memory deficits, specifically prominent disruptions in their episodic memory. These memory disruptions are thought to be caused by altered structure and function of areas within the memory network, specifically the prefrontal and temporal lobe including the hippocampus. Recently, identifying biomarkers of treatment response by using neuroimaging has become a prominent area of interest in schizophrenia. Biomarkers in schizophrenia could help by allowing for more efficient treatment using current method and also the development of novel treatments. The primary objective of this dissertation research was to use non-invasive magnetic resonance imaging techniques, such as magnetic resonance spectroscopy (MRS), functional magnetic resonance imaging (fMRI), and structural magnetic resonance imaging (sMRI), to examine structural and functional features of schizophrenia as it relates to the memory network and treatment response. In the first study, we conducted a multimodal fMRI blood oxygen level-dependent (BOLD) and MRS study in healthy controls and chronic patients with schizophrenia stable on medication and found that patients showed decreased BOLD signal during encoding and retrieval within prominent regions of the memory network during an episodic memory task. Also, we found a significant correlation between the metabolites N-acetylaspartate/creatine (NAA/Cr) and glutamate+ glutamine (Glx/Cr) in our healthy control group but not in schizophrenia. In the second study, we used an automated algorithm, FreeSurfer, to measure the volume of the basal ganglia in unmedicated patients with schizophrenia. We found that treatment response was significantly correlated with all regions of the basal ganglia measured (caudate, putamen, and pallidum) and that the caudate volume explains a significant amount of the variance in treatment response even when controlling for baseline symptom severity and duration of illness. In the third study, we measured effective connectivity within regions of the memory network in patients with schizophrenia off-medication and again after 1-week of treatment with an APD. We found that while off medication SZ showed decreased connectivity within the memory network and that after 1-week of medication that the connectivity within the memory network increased.