Study objectives included determining demographic, phenotypic, diagnostic and treatment features, and outcomes of all patients who underwent serum/CSF myelin oligodendrocyte glycoprotein (MOG) antibody testing.

In MOG antibody-associated disease (MOGAD), uncertainty remains regarding antibody titers and disease behaviour, in particular, the value of CSF MOG antibodies in diagnosis and prognosis. Studies have shown heterogeneity in CSF positive results including CSF restricted MOG antibodies, raising the possibility of intrathecal MOG antibody production.

Of 4494 MOG assays, 413 were on CSF samples in 402 people (268 patients with ≥ 1 serum sample). Mean time between S/C testing was 20.9 days (90% within 30 days). Five patients were S+/C+, 4/5 with acute disseminated encephalomyelitis, 1/5 with longitudinally extensive myelitis, all meeting MOGAD diagnostic criteria. Twenty-three patients were S+/C-, 13/23 with optic neuritis, 15/23 meeting MOGAD diagnostic criteria. MOG S+/C+ patients were male, relatively young (mean age 15.8 years), none with CSF oligoclonal banding, and 60% with persistent seropositivity (vs. 26% of S+/C- patients). No patient was S-/C+. 

CSF MOG positivity was rare in our cohort. Isolated optic neuritis was not observed in CSF MOG+ patients. MOGAD was the most common diagnosis, particularly in S+/C+ patients. Our findings do not support CSF MOG testing, particularly in optic neuritis. Limitations include selection bias and a fixed cell-based assay. Prospective studies with a live cell-based assay will clarify the utility of CSF MOG antibody testing.