Clinical manifestations

Presentation and course

Neurologic disorders associated with MOG antibodies: clinical presentation and disease course. The first promising results of MOG antibodies in demyelinating inflammatory CNS diseases were achieved by using self-assembling radiolabeled MOG tetramers, which identified a subgroup of mainly pediatric patients with ADEM and MOG antibodies (34).

Following studies confirmed the result, and MOG antibodies were predominantly described in children and associated with an ADEM-like presentation and a monophasic disease (04; 12; 37). However, it is now well-known that MOG antibodies were also present in a subset of adult patients with CNS demyelinating syndromes, including ADEM, neuromyelitis optica spectrum disorder, monophasic and recurrent optic neuritis, and transverse myelitis as well as demyelinating syndromes overlapping with anti-NMDA receptor encephalitis; it is also well-known that MOG antibodies can be associated with a recurrent disease course (13). Further, an encephalitis clinical phenotype with cortical lesions and epileptic seizures has been repeatedly reported (08; 45).

Within the last large observational cohort studies on MOG antibody-positive patients shed light on the clinical and demographic features of these patients.

These data indicated that an age-related phenotype distribution exists with a predominant initial presentation of ADEM-like disease in children, especially before the age of 11 years, either with or without concurrent involvement of the optic nerve followed by transverse myelitis. In summary, about 50% of MOG antibody-positive children presented with ADEM, another 25% to 35% with optic neuritis, and the remainder with myelitis and other syndromes, eg, brainstem syndromes. In contrast, seropositive adults presented with optic neuritis in 50% of cases: 25% with myelitis and the others with ADEM, brainstem, cerebral, or multifocal symptoms (19; 02).

MRI abnormalities typically relate to the clinical affected anatomical area. Typical findings in children reflect ADEM-like changes. Enhancement of brain lesions is common with MOGAD cerebral attacks and often has a nonspecific patchy appearance and rarely persists beyond 3 months (16). Optical nerve involvement frequently affects the anterior part, and often (31% to 58%) bilateral. Spinal cord lesions can be present as longitudinal extensive transverse myelitis or as short lesions, with predominance of the grey matter “H-sign” and are reported to particularly involve the conus. However, importantly, initial MRI can be normal despite severe clinical affection (09; 32).

CSF examination showed in more than 50% of adult seropositive patients elevated white cell count and in about 50% elevated albumin CSF/serum ratio. CSF-restricted oligoclonal IgG bands, which are a hallmark of multiple sclerosis, were only present in 10% of patients and in these cases, oligoclonal bands were often transient. Mononuclear cells prevailed within the CSF; however, neutrophils are not uncommon. Similar CSF changes are found in children (26; 24).

Diagnostic criteria of MOGAD. In 2023, the International MOGAD Panel proposed the following criteria for the diagnosis of MOGAD (02):

If the cell-based assay from serum show low positive MOG-IgG titers, is negative but the CSF is positive, or the serum titer is unknown, the demyelinating event must fulfill supportive clinical or MRI criteria.

Prognosis and complications

Early studies indicated a monophasic disease course in patients with MOG antibodies. However, subsequent studies revealed that MOG antibodies are also present in recurrent non-multiple sclerosis disorders. Although severe disease courses are described (27; 42), overall, MOG antibodies seemed to be associated with fewer relapses and a better functional outcome (14). Risk of relapse has been reported as ranging from around 27% to 36% over short follow-up periods, with one study reporting an 8-year risk of 36% (28; 08; 06).

Of note, most acquired disability results from the first attack (approximately 60% to 70%), indicating the importance of acute attack management.

As MOG antibodies are present in monophasic and multiphasic diseases, possible association of a temporal dynamic of antibody levels from disease onset to follow-up with prognosis has been suggested. A decline of MOG antibody titer was more common in patients with a monophasic disease course, whereas a stable or increasing titer during follow-up indicated a recurrent disease (08), such as multiphasic disseminated encephalomyelitis or ADEM, followed by optic neuritis, neuromyelitis optica spectrum disorder, and, rarely, multiple sclerosis with anti-MOG seropositivity (20). A potential use of antibodies for treatment response was encouraged by a study showing that conversion to seronegativity during immune-directed therapy was a predictor for disease activity-free follow-up (23). Two studies found a median time to convert to a seronegative status of about 1 year (01; 46). Persistence of MOG antibodies is associated with the risk of having a relapse by a factor of 2 of 10 (02).

However, as severe disease courses with insufficient treatment response to disease-modifying therapies in association with fluctuation and reappearance of antibodies has been described, the use of MOG antibodies as predictive or treatment response marker warrants further prospective studies with clear international standardized assays and cut-off values (42; 38).

The prognostic value of CSF MOG antibodies remains unclear but a retrospective observational, multicenter study suggests that CSF MOG antibodies are common in MOGAD and are associated with a more severe clinical presentation (05). In addition, the clinical outcome is worse in those patients in whom MOG antibodies are only detected in the CSF.

Biological basis

Etiology and pathogenesis

The pathogenic role of autoimmune responses to MOG has been well established in numerous studies (36). In both human diseases and experimental animal models, MOG-mediated encephalomyelitis leads to heterogenous clinical phenotypes. Studies investigating the pathogenic role of human MOG antibodies demonstrated that they can activate complement and cellular-dependent cytotoxicity in vitro but had only subtle effects after intracerebral or peripheral injection in rodents (03). However, studies have demonstrated that human MOG antibodies are pathogenic in rodents if they cross react with rodent MOG and if titers and affinity are sufficiently high. Further, there are two possible mechanisms of tissue injury: antibody mediated and augmentation of MOG-T-cell-induced inflammation.

Pathology. Histopathological findings of anti-MOG antibody-associated diseases have only been available from single case reports and revealed some information on the poorly understood underlying pathomechanism. Cases have been published with variable clinical presentations including neuromyelitis optica spectrum disorder, ADEM, relapsing-remitting multiple sclerosis, and atypical demyelinating syndrome and different MRI features, including fulminant, tumefactive lesions (36; 47).

However, two seminal studies described neuropathological changes in large cohort of MOG antibody patients (21; 43). Demyelination occurred by merging of multiple perivenous lesions, with partial axonal preservation. Complement and immunoglobulin deposition is observed at areas of active demyelination. MOG harboring macrophages are located within the perivascular spaces and in demyelinating lesions. AQP4 expression is preserved, and cortical lesions are prevalent with predominance of intracortical and subpial lesions. Importantly, these studies revealed that MOGAD-associated pathology occurs on a background of mainly CD4+ T-cells, in contrast to multiple sclerosis, supporting a humoral-mediated disease mechanism in MOG antibody-associated diseases. These pathoimmunological features strongly suggest that MOGAD is an independent autoimmune demyelinating disease entity, distinct from aquaporin-4+ neuromyelitis optica spectrum disorders and multiple sclerosis (21; 43).

In conclusion, anti-MOG antibodies indicate a distinct pathomechanism in this subgroup of inflammatory demyelinating diseases.

Epidemiology

Less is known about the exact prevalence of MOG-antibody diseases in the entirety of inflammatory demyelinating CNS diseases; numerous studies addressing these question different test methods, and patient selection considerably varies. The overall described incidence across studies is reported to be 0.2 to 1.4 in 100,000, with slight tendency prevailing affecting females and more prominent in childhood (19).

The frequency of MOG antibodies in pediatric patients with acquired demyelinating diseases were analyzed in three population-based studies from Austria and Germany, United Kingdom, and Netherlands; the studies reported seroprevalences of 31%, 32% and 22%, respectively (20; 10; 15). In contrast, MOG antibodies were less frequent (4% and 2%) in patients with optic neuritis in two population-based studies from Denmark and the United States (41; 07). Finally, two studies reported similar frequencies (12% and 11%) of MOG antibodies in neuromyelitis optica spectrum disorder (18; 39).

Prevention

No preventive interventions have been described.

Differential diagnosis

Associated or underlying disorders

The routine diagnostic use of MOG antibodies in clinical practice in pediatric patients was first recommended in 2017 (17). The authors classified patients into four main phenotypes: multiple sclerosis, AQP-4 antibody neuromyelitis optica spectrum disorder, MOG antibody-associated diseases, or antibody-negative recurrent demyelinating syndrome. With a baseline MRI, the most common disease, multiple sclerosis, could be reliably diagnosed, and if multiple sclerosis is diagnosed according to the revised McDonald criteria (44), no further diagnostic steps are necessary. MOG antibodies should be analyzed in typical AQP-4-negative neuromyelitis optica spectrum disorder and in patients with ADEM. Furthermore, poorly marginated lesions located in the cerebellar peduncle or with a leukodystrophy-like MRI pattern are suggestive for MOG antibodies. A further recommendation was published by an expert consensus for patients with a demyelinating CNS disease of suspected autoimmune etiology and an either monophasic or relapsing disease course (25). As screening of MOG antibodies in large, unselected patient populations decreased the predictive test power, testing is only recommended in a subpopulation of inflammatory demyelinating CNS diseases.

Diagnostic workup

For antibody testing only, a cell-based assay using full-length MOG expressed in human cells is recommended, ELISA or immunoblot are obsolete. Antibody analysis should be performed with immunofluorescence or fluorescence-activated cell sorting. As total anti-IgG (heavy and light chain) secondary antibodies identify all antibody subclasses and result in a lower specificity, IgG Fc or IgG1 secondary antibodies are recommended (25). MOG antibodies are more prevalent in serum but may also occur only in CSF, therefore, testing of both serum and CSF is recommended to avoid false negative results. Due to a higher prevalence in sera, a peripheral antibody production is suggested in the majority of cases. In positive cases, MOG antibody testing should be repeated after 6 months and 1 year.

Management

Until now, there have been no prospective studies regarding treatment of MOG antibody-associated diseases. Usually, the first attack is treated with intravenous methylprednisolone, (in case of nonresponder) plasma exchange and IVIg. Restitutio ad integrum after acute attack is reported to be about 50% to over 90% (27; 08; 01).

Because MOG antibodies are often associated with a monophasic disease course and there is a lack of prospective randomized treatment trials, it is unclear whether long-term treatment is required. Assuming a mainly B-cell mediated and humoral immune pathogenesis of the disease, therapy targeting antibodies and B-cells are of special interest. Patients with MOG antibodies are less often treated with a disease-modifying drug compared with patients with AQP-4 antibodies (22). However, it has been shown that after the first demyelinating event, maintenance of immunosuppression or immunomodulatory treatment, such as azathioprine or rituximab, was associated with a decrease of a relapse risk and a reduced annual relapse rate (28; 08; 14). Furthermore, a prospective observational cohort study including 79 MOG antibody-associated spectrum disease patients who received either mycophenolate mofetil or no therapy revealed a reduced risk of relapse of 86% (31).

Overall, prospective studies for the evaluation of use of long-term disease-modifying treatments in patients with MOG antibodies and the value of serial testing for prognosis and treatment monitoring are required.

Outcomes

Limited data are available about the long-term outcome of MOGAD. MOGAD patients seem to have a better disability outcome than those with AQP4 antibodies (40). In one study of 61 patients with a median follow-up of 14 to 15 years, favorable outcomes were achieved in the majority of patients, yet 37.5% had an EDSS score of 3 or greater, an EDSS of 6 or greater occurred in 12.5%, and poor visual outcome (visual acuity of 0.1 or less at least in one eye) occurred in 16% (11).

However, given the published diagnostic MOG-IgG criteria and the limited data available for long-term follow-up, prospective studies about long-term follow-up are needed to determine the true prognosis of the disease (02).