Applying the Supporting Features for MOGAD Diagnosis to Patients With Multiple Sclerosis

In the 2023 diagnostic criteria, supporting clinical/MRI features are required to diagnose myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) in patients at high risk of false MOG-IgG positivity (low/unavailable titer, cerebrospinal fluid-restricted), to help differentiation from multiple sclerosis (MS). However, overlap is possible and determining the frequency of the MOGAD supporting features at MS onset may help prevent misdiagnosis. We assessed the frequency of the MOGAD supporting features and potential risk of misdiagnosis at first attack of MS/clinically isolated syndrome (CIS).

In this observational study, we retrospectively identified consecutive patients hospitalized between 2021 and 2024 at 2 Italian centers, with: (1) relapsing-remitting MS/CIS, and (2) first attack MRI available. The frequency of the MOGAD supporting features was assessed at MS/CIS presentation, and potential risk of misdiagnosis determined based on the probability of concurrent false MOG-IgG positivity. Other clinical-MRI features typical of MOGAD but not included in the 2023 diagnostic criteria were also evaluated.

A total of 244 patients with MS/CIS were included (median age, 34 [range, 13-78] years; 66% were female). Overall, 65/244 (27%) patients with MS/CIS showed at least 1 MOGAD supporting feature during the presenting attack. "Central cord lesion" (27/82 [33%]) and "deep gray nuclei involvement" (24/86 [28%]) were more represented than other supporting features (0%-7%); p < 0.001. MOG-IgG was tested in 207/244 (85%) patients, showing false positivity in 3 (1.4%) patients. The combined probability of detecting a false MOG-IgG positivity and meeting 1 of the MOGAD supporting features was 0.4%. Among other typical MOGAD features not included in the diagnostic criteria, the following had a 0% frequency at MS/CIS presentation: (1) encephalopathy, (2) marked CSF pleocytosis (>50 cells/mm³), (3) wheelchair need during myelitis, and (4) MRI T2-lesion resolution postattack.

Although the MOGAD supporting features can be met in one-fourth of patients at MS/CIS presentation, the risk of misdiagnosis in similar unselected cohorts remains low. Future refinements of the MOGAD supporting features should take into account their relative frequency in MS/CIS and possible integration with additional clinical-MRI features that are more specific for MOGAD.