LAUSR

Abstract Background Decreased functional magnetic resonance (fMRI) signaling in ventral striatum during reward anticipation has been a consistent finding in unmedicated patients with schizophrenia. Antipsychotic medication may normalize this, which has been associated with improvement in psychotic symptoms. Whether this initial treatment response predict long term clinical outcome have not been examined previously. We here present unique data from a cohort of antipsychotic näive patients, who was examined before and after six weeks of antipsychotic monotherapy, and again after six years of naturalistic treatment. We hypothesized that improved reward related fMRI-signaling in striatum during initial treatment will be associated with less severe symptoms and a higher level of function after six years. Methods Antipsychotic-naïve, first episode patients with schizophrenia and matched healthy controls were examined at three timepoints: Baseline, after six weeks and after six years. During the first six weeks, patients were treated with a D2/3 antagonist, after which they switched to a naturalistic clinical setting. At all three timepoints, fMRI was obtained while participants played a monetary rewarding game. Mean contrast signal during anticipation of salient events was extracted from predefined regions of interest in nucleus accumbens and caudatus. Additionally, symptom severity and level of function were assessed in patients. Remission after six years was defined by the Andreasen criteria. Long term changes in patients were analyzed with paired t-test. Repeated measures ANOVA was used to compare fMRI signal in patients and controls over time. Changes in fMRI signal during the first six weeks was compared between remitted and unremitted patients with students t-test. Correlations between symptom severity, level of function and fMRI signal up were analyzed with Spearman or Pearson as appropriate. Results Long term follow-up data on fMRI and psychopathology was obtained from 25 patients and 28 controls 6.7 (4.3–8.7) years after baseline examinations. In patients, symptomatology and level of function was improved (all p<0.001), and remission criteria was met by 12 (48%) of the patients. Repeated measures ANOVA of the extracted fMRI contrast signal at baseline and six years showed a group*time interaction in left Nucleus accumbens (F=4.8, p= 0.038) and left caudatus (F= 4.08, p=0=049). Mean contrast signal increased in patients but decreased in healthy controls. Change in contrast signal in left caudatus during the first six weeks of treatment differed between remitted and non-remitted patients (t=2.2, p=0.044). An increase in signal during initial treatment was found in patients later being in remission, which was not observed in non-remitted patients. Decreased fMRI contrast signal in left caudatus at six years follow up was associated with a higher level of negative symptoms (r=-0.44, p=0.02) and a lower level of function (r=0.45, p=0.02). Discussion In this long-term follow up study, half of the patients were in remission 4–8 years after their first schizophrenia diagnosis. These individuals were characterized by a normalization of reward anticipation signaling in left striatal regions during their initial treatment with a selective D2/3 antagonist. Schizophrenia is a diagnosis with a heterogeneous outcome, and prognostic biomarkers are still missing. The present results suggest that normalization in reward processing during initial antipsychotic treatment may predict remission six years later. Although these results need to be reproduced, they do support the notion, that patients responding to antipsychotic treatment are characterized by changes in striatal function, which may be predictive for long-term outcome.