MOGAD myelitis is diagnosed when a patient has neurological deficits and tempo of symptomatic development compatible with myelitis, a clear positive serum MOG-IgG test, and supportive of diagnosis MRI features (2). The latest international expert consensus advocates towards the use of live cell-based assays to increase diagnostic specificity. If not available, fixed cell-based assays can be used, with a clear positive result being a titer >1:100 (2). Titers lower than 1:100 have a lower predictive value (number of true-positive results/total positive results) for MOGAD and may lead to misdiagnosis. For example, a titer of MOG of 1:20 to 1:40 carry a positive predictive value of 51%, meaning that nearly 50% of patients with this titer may have a different etiology of their clinical presentation (13). Serological testing should ideally occur before administration of corticosteroids, intravenous immunoglobulins, or plasma exchange, as these interventions can increase the risk of false negative result. In cases of a high clinical suspicion but a negative test, if done after initiation of immune therapies, testing should be repeated about 3 months or later. Screening for the presence of serum MOG-IgG routinely in patients with clear features of multiple sclerosis is not recommended, as false positives can occur, decreasing the positive predictive value of the test and leading to misdiagnosis (2, 10).

In addition, titers of MOG antibody can fluctuate with intermittent seroconversion to negativity and can become undetectable over time (14); this, and the fact that the MRI abnormalities can disappear (up to 72% of the brain lesions; 79% of spinal cord lesions) (13, 15, 16), make a retrospective diagnosis of MOGAD impossible in some scenarios, which warrants a continuity of high degree of suspicion if new neurological symptoms arise (17).

Cerebrospinal fluid (CSF) analysis typically demonstrates lymphocytic pleocytosis, with >50 WBC/mm³ seen in about 30% of patients (9). Oligoclonal bands are frequently absent, with positivity rates being less than 10%–15% (4, 18). MOG-IgG antibody concentrations in CSF are typically low, however in some cases MOG-IgG can be positive in CSF and not in serum. Intrathecal antibody production occurs more frequently in MOGAD than in AQP4-positive NMOSD (19). Caution is advised in interpretation of the MOG-IgG positivity when clinical features are atypical of MOGAD, as false positives can occur. Testing both CSF and serum is not recommended for routine evaluations (2). However, if the serum MOG antibody is negative in a patient presenting with clinical, radiological and CSF findings typical of MOGAD, CSF testing for MOG antibody may be considered, as well as re-testing serum several months later.

Current diagnostic criteria for MOGAD require exclusion of an alternative diagnosis. Differential diagnostic considerations include multiple sclerosis, neuromyelitis optica spectrum disorder, CRMP-5 antibody positive paraneoplastic myelopathy, GFAP antibody positive encephalomyelitis, neurosarcoidosis, SLE associated TM, ischemic myelopathy, metabolic myelopathy, infections and neoplastic causes. Specific findings pointing to other inflammatory diagnoses are summarized in Table 1.

Acute myelitis due to MOGAD is typically longitudinally extensive involving 3 or more vertebral segments in length (Figure 1). Short lesions can occur in about 7% of patients, more commonly in those patients that are approximately 40 years of age (3). Multifocal spinal cord lesions can occur (2, 18). Spinal cord necrosis is typically not seen (20). Some T2 signal abnormalities may be initially subtle and contrast the severity of clinical presentation. In fact, in some cases linear T2 signal abnormalities may be difficult to distinguish from a prominent central canal. In these instances, carefully evaluating axial images may be helpful, as well as repeating the MRI at a later time to evaluate for the evolution of the lesion. MRI of the brain may also show MOGAD-associated lesions and should be considered in TM evaluation. Some patients with MOGAD TM may also have clinically silent spinal cord lesions (20). Although current data are limited by the retrospective nature, one study found silent spinal cord MRI lesions occurred in none of the patients during remission and in up to 7.3% of those during acute attack (21). Another study confirmed that finding new or enlarging spinal cord lesions inter-attack is rare and occurs in <1% of patients (22). Initial MRI of the spinal cord may be normal in up to 10% of patients, and repeat MRI within days would be warranted (13–15).

A negative MRI in a patient presenting acutely with clinical features of myelitis should not deter from testing for MOG antibody. The spinal cord lesions are typically located in the central cord, can involve both the gray and white matter, affect more than 50% of the axial section or can be restricted to the grey matter in about 50% of patients, which on axial imaging can be seen as the “H sign” (2, 23). T2 signal abnormality confined to gray matter (sagittal line and axial H sign) as well as lack of enhancement favor diagnosis of MOGAD over NMOSD or MS (8).

A straight T2 hyperintense line with surrounding hyperintense signal in the anterior and posterior gray matter horns on sagittal view can also be seen in MOGAD myelitis and is known as pseudo dilation of the central canal (23). The conus medullaris can be affected in about 26% of patients with MOGAD (24), which is a valuable radiographic clue, as the conus medullaris is less frequently affected in other demyelinating conditions, such as multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD), with rates of involvement as low as 1.3 and 6%, respectively, (8, 24–26). Gadolinium enhancement can occur in about half of patients with MOGAD myelitis and can be heterogeneous giving a cloud-like appearance (2, 9, 23). Interestingly, dorsal nerve root enhancement, cauda equina and pial enhancement can also occur (2).

Clinically silent new lesions on MRI are rare and occur in only 3% of patients with MOGAD (21, 27). Upon follow up and resolution of acute myelitis, complete resolution of spinal cord lesions can be seen in up to 79% of MOGAD cases (4). Spinal cord atrophy can be seen in severe cases, but this is rare (2).

In the acute setting, patients presenting with TM receive intravenous corticosteroids. For those with incomplete recovery or severe clinical picture overall, plasma exchange is typically used next (9, 28). Intravenous immunoglobulins (IVIG) can also be used following the plasma exchange completion, given that intravenous immunoglobulins can also serve as an effective preventive therapy for those patients who elect long-term preventive therapy option (29). Steroids are typically tapered slowly over at least a month or longer, with some data pointing to a lower risk of relapse in those with steroid use >1 month (16).

Neurological outcomes of patients with TM due to MOGAD are typically more favorable than of those with anti-aquaporin4 antibody positive NMOSD. A recent study evaluating long-term outcomes of patients with TM due to MOGAD (n = 32) vs. NMOSD (n = 57) found that MOGAD TM patients on average had a lower EDSS score than patients with AQP4-Ab TM (1.8 [1.0–8.0] vs. 3.0 [1.0–8.0]), reflecting better outcomes. Due to higher predilection of MOG positive TM to conus localization, persistent bladder dysfunction was more common in patients with MOGAD (59% with MOGAD and 48% with AQP-4 positive NMOSD), with up to 23% requiring long-term catheterization in both groups. In addition, neuropathic pain was less common in patients with MOGAD TM vs. NMOSD TM (29% vs. 13%) (32).