Autoantibodies, routine and novel markers of neuroinflammation in people with atypical psychiatric disease

Background: Increasingly, the immune system is recognized as participating in the pathophysiology of psychiatric disease. There is renewed interest in biomarkers identifying immune activation. Methods: We measured serum and cerebral spinal fluid (CSF) autoantibodies with other routine and novel markers of neuroinflammation, including CSF cytokines in patients with atypical psychiatric (AP) disease who were referred by their psychiatrist. Their results were compared with cohorts of non-inflammatory (NI) controls, viral infectious controls and patients with autoimmune encephalitis. Results: Thirty-five AP patients were enrolled (29 females; 6 males), alongside 18 NI controls. Six AP patients had first episode psychosis, 7 had depression, 3 had schizophrenia. Others had a mix of both psychotic and mood disorder features, making their disease difficult to classify. Ten patients had a history of autoimmune disease and 11 a family history of autoimmunity.Low-mid titre (1:80-1:640) anti-nuclear antibodies (ANAs) without associated positive extractable nuclear antigen antibodies were noted in 20 patients. The predominant pattern was speckled (17/20, 85%); other patterns were mitotic spindle apparatus (2) or homogenous and speckled (2). One patient had a high titre 1:2560 speckled ANA associated anti-SSa/Ro60 antibodies and high thyroperoxidase antibodies (900IU/mL; normal < 35IU/mL). Two patients had raised anti-thyroglobulin antibodies and 7 patients had raised thyroperoxidase antibodies.No onconeuronal or limbic encephalitis antibodies were identified on serum or CSF. One patient had CSF pleocytosis (>5 monocytes), 5 had raised CSF protein (>0.45g/L), 3 had CSF restricted oligoclonal bands and 13 had raised CSF neopterin (>20nm/L).CSF cytokines: granulocyte-macrophage colony-stimulating factor, and interferon gamma were associated with AP patients. A subset of patients (10) had at least one CSF cytokine beyond 4 standard deviations of the NI cohort.Our data supports more extensive investigation of patients presenting with psychiatric disease across a broad diagnostic spectrum. Further study is needed to validate these results.