Multiple lines of evidence indicate that immunological and inflammatory alterations contribute at least in a subgroup to the pathophysiology of schizophrenia. In this retrospective chart review, we investigated whether clinical factors contribute to altered cerebrospinal fluid (CSF) findings in schizophrenia-spectrum disorders. Clinical data from electronic medical records of patients with psychotic disorders (ICD-10: F20-F29) who received routine CSF diagnostics at the Department of Psychiatry and Psychotherapy, LMU Munich, Germany, were included. Chi² tests for dichotomous outcomes and independent t tests for continuous outcomes were used to compare differences between groups. A total of 331 patients were included in the analyses (43.2% female and 56.8% male). The mean age was 37.67 years (±15.58). The mean duration of illness was 71.96 months (±102.59). In all, 40% (128/320) were first-episode psychosis (FEP) patients and 60% (192/320) were multi-episode psychosis (MEP) patients. Elevated CSF protein levels were found in 19.8% and elevated CSF/serum albumin ratios (Q Alb ) in 29.4% of the cases. Pleocytosis was found in 6.1% of patients. MEP patients showed significantly higher mean Q Alb compared with FEP patients ( t (304.57) = −2.75, p = 0.006), which did not remain significant after correcting for age. Q Alb elevation occurred more frequently in men ( X 2 (1) = 14.76, p = <0.001). For treatment resistance, family history, and cMRI alterations, no significant differences in CSF-related outcomes were detected. Our work extends other retrospective cohorts confirming a relevant degree of CSF alterations in schizophrenia-spectrum disorders and shows the difficulty to relate these alterations to clinical and disease course trajectories. More research is needed to develop treatment response predictors from CSF analyses.

CSF abnormalities and a neuroinflammatory pathophysiology have been discussed for affective and non-affective psychosis for more than 30-years. Recent studies pointed towards a specific phenotype of autoimmune-antibody mediated psychosis, but evidence is still sparse. Especially CSF data investigating autoimmune antibodies in large-scale CSF cohorts of affective and non-affective psychoses are lacking. We analyzed a retrospective naturalistic cohort of 592 patients with A) schizophrenia-spectrum disorders (N = 330) or B) depressive disorders (N = 262) who underwent a lumbar puncture as part of the clinical routine in the Department of Psychiatry and Psychotherapy at the Ludwig-Maximilians University Munich between July 2012 and May 2017. We used a predefined systematic algorithm for the database search in the clinical documentation system and data was extracted by TO and AG. The study was approved by the local ethics committee. We identified 592 patients with standard CSF parameters. Schizophrenia spectrum patients did not differ from depressive patients with regard to the white blood cell count (cells/µl) (p = 0.774) or albumin quotient (p = 0.663). The general prevalence of oligoclonal bands did not differ between groups (schizophrenia: 37.0%, depression: 37.8%; p = 0.838). However, schizophrenia patients showed higher frequencies for intrathecal oligoclonal bands (32% of all oligoclonal bands) compared to depressive patients (19.1% of all oligoclonal bands. (p = 0.034). 124 schizophrenia-spectrum patients (54 first-episode patients) received CSF analyses for neural antibodies. None of the patients showed positive CSF results in any of the tested autoimmune-encephalitis panel (NMDA(N=119), AMPA-1(N=114), AMPA-2(N=114), CASPR(N=111), LGI-1(N=110) and GABA-B(N=112)-Antibodies) in CSF. The results for the intracellular onconeuronal and synaptic antibodies were also negative (Amphyphysin(N=93), Yo(N=58), Hu(N=94), Ri(N=94), CV2(N=93), Ma1(N=93) and Ma2(N=93)-Antibodies). Three of these patients with negative CSF titers did have low-titer neuronal antibodies in serum: CASPR-2-AB: 1:10, CASPR-2-AB: 1:50, Yo-AB: low band-intensity. 36 depression patients were also tested for autoimmune antibodies and again no positive reports could be identified in CSF. This is the first analyses of autoimmune antibodies in first-episode and recurrent schizophrenia and depressive mood disorder showing no positive CSF titers. However, schizophrenia patients have a higher prevalence of intrathecal oligoclonal bands compared to affective patients pointing towards more immunological disturbances in this population. The here presented analyses are exploratory and need to undergo confirmatory analyses and quality control.