Screening for autoantibodies in inflammatory neurological syndrome using fluorescence pattern in a tissue-based assay: Cerebrospinal fluid findings from 793 patients

Background: To evaluate indirect immunofluorescence patterns of auto-antibodies and the targeting antigens to Immunoglobulin Gs (IgGs) in the cerebrospinal fluid (CSF) by a tissue-based assay(TBA).

Methods: CSF samples were collected from 793 patients. Auto-antibody levels were measured via an immunofluorescence assay.

Results: 110 (13.9%) CSF samples with a specific response were confirmed. Of these, 37 showed a neuronal pattern, 57 an astrocyte pattern, 7 a neuronal and astrocyte pattern, and 9 samples showed an oligodendrocyte pattern. In the neuronal antibody group, 16 patients had NMDAR-IgGs, 3 had LGi1-IgGs, 2 had AMPA2-IgGs, 2 had GAD65-IgGs, 1 patient had GABA-IgGs, and 1 patient had overlapping NMDAR-IgGs and AQP4-IgGs. Of the unidentified neuronal antibodies, two were cellular surface antibodies, three were cellular surface and cytoplasm antibodies, three were cytoplasm antibodies, and four were nuclear and cytoplasm antibodies. Among the 57 patients with the astrocyte pattern, 28 patients were positive for AQP4-IgGs, 21 were positive for GFAP-IgGs, 5 patients had overlapping AQP4 and GFAP-IgGs, and 3 patients had an unidentified antigen. Seven patients showed neuronal and astrocyte patterns simultaneously; four of them had unknown neuronal antibodies. In the patients with an oligodendrocyte pattern, one was positive for MOG-IgGs and four for MBP-IgGs.

Conclusions: The TBA is helpful for diagnosing autoimmune neurological syndrome, especially in patients with unknown antibodies and antigens. Presence of unidentified antibodies against neuronal or glial cells could be an interesting finding, but should be investigated in future studies which incorporate parallel serum samples at an appropriate IgG dilution.