T89. LONGITUDINAL EVALUATION OF GLUTAMATE AND N-ACETYL-ASPARTATE IN ANTIPSYCHOTIC-NAïVE FIRST EPISODE PSYCHOSIS SUBJECTS AND RELATIONSHIPS TO DURATION OF UNTREATED PSYCHOSIS AND RESPONSE TO ANTIPSYCHOTIC MEDICATION

Abstract

Background: Meta-analyses have consistently identified a relationship between longer duration of untreated psychosis (DUP) and worse longterm outcomes in patients with schizophrenia. However, the neurobiology of this phenomenon and its implications for response to antipsychotic medications remain poorly understood. We previously identified two indices of glutamatergic dysfunction: excess of glutamate + glutamine (Glx) (reference) and an abnormal correlation between N-acetyl-aspartate (NAA) and Glx 1. The aim of this longitudinal study was to measure Glx and NAA in antipsychotic-naive first episode psychosis subjects (FEP) before and after antipsychotic treatment and evaluate their relationships to DUP and treatment response. Methods: 51 FEP (age: 23.55 ± 5.92, 34M/17F) were enrolled in this proton magnetic resonance spectroscopy (1H-MRS) study. Two voxels were assessed: the anterior cingulate cortex (ACC) and the hippocampus (HIP) (PRESS, TR/TE= 2000/80 ms), at baseline and after 6 and 16 weeks of treatment with risperidone. Metabolite concentrations were quantified as absolute value with partial volume corrections 2 using TARQUIN 3. DUP was defined as the duration between discernable psychotic symptoms to the time of initial treatment contact. Treatment response was assessed as percent change of the score on positive subscale of the Brief Psychiatric Rating Scale (BPRS) between baseline and week 16. Differences in metabolite levels between sessions were investigated using MANOVA, and the NAA-Glx correlation was examined using partial correlations (controlled for age and gender). We also used partial correlations to assess relationships between baseline metabolites and DUP (log-transformed). Results: Effect of DUP on NAA and Glx levels: There were trend-level negative correlations between Glx and DUP in both the ACC (r=-0.26, p=0.07) and hippocampus (r=-0.27, p=0.07). No significant correlations were found for NAA (ACC: r=0.07, p=0.63; HIP: r=-0.18, p=0.23). Effect of medication on NAA, Glx and NAA-Glx correlation: At baseline, a correlation between NAA and Glx was found in both ACC (r=0.55, p<0.01) and HIP (r=0.41, p<0.01). No effect of treatment was found for Glx (F(2,63)=1.13, p=0.32), NAA (F(2,64)=1.66, p=0.19), or NAA-Glx correlation (F(2,60)=0.36, p=0.69) in the ACC. No overall effect of treatment for Glx (F(2,57)=1.87, p=0.16), NAA (F(2,58)=0.86, p=0.42), or the NAAGlx correlation was found (F(2,53)=1.30, p=0.27) in the hippocampus. Treatment response and NAA, Glx: There were no significant correlations between changes in Glx and changes in the positive subscale of the BPRS both for ACC and HIP (r=-0.07, p=0.70; and r=0.07, p=0.70 respectively). The same results were found for NAA (ACC: r=-0.23, p=0.20; HIP: r=0.03, p=0.86). Discussion: In both the ACC and hippocampus, we identified trend-level negative correlations between Glx and the DUP. As enrollment in this study continues, we will be able to clarify the direction of this relationship in the future. There were no significant changes in Glx over time. This is in contrast to De La Fuente-Sandoval who reported decreases in Glx with treatment, although in the associate striatum 4. We did not replicate our prior finding of a lack of correlation between Glx and NAA in the hippocampus 1, however this finding was made in chronic patients, thus it could index a downstream effect of illness chronicity.