Immunoglobulin A Antibodies Against Myelin Oligodendrocyte Glycoprotein in a Subgroup of Patients With Central Nervous System Demyelination

This cross-sectional study examines laboratory and imaging data for patients with demyelinating central nervous system disease to investigate the frequency of MOG-IgA and associated clinical features.

Recent evidence suggests that IgA may play a role in the pathogenesis of inflammatory disorders. 9,10However, the role of autoreactive IgA antibodies in CNS demyelination is still unclear.Here, we conducted an observational, retrospective, longitudinal multicenter study to investigate the frequency of MOG-IgA and its association with clinical features in demyelinating CNS syndromes.

Study Participants
We cross-sectionally screened serum samples from 1344 patients with suspected or confirmed multiple sclerosis (MS), 11 MOG antibody-associated disease, 8 or NMOSD 7 at sampling and 110 healthy controls from 5 centers in a discovery and confirmation setup.Patients were assessed from September 2012 to April 2022 (median follow-up time, 39 months; range, 0-227 months).Both CSF and longitudinal serum samples were measured when available.Five patients were excluded from the study (eMethods in Supplement 1).This study was approved by the institutional review boards of the participating centers.All patients provided written informed consent.

Clinical and Imaging Data
Retrieval and analysis of available clinical and other data, magnetic resonance images, and retinal optical coherence tomography are described in the eMethods and eTable 1 in Supplement 1.

Live Cell-Based MOG Assay
Serum samples (1:100) and CSF (1:5) were examined for IgA/ IgG/IgM reactivity against full-length human MOG using a live cell-based assay as previously described 3,5 (eMethods in Supplement 1).For each sample, the ratio of the geometric mean channel fluorescence intensity of the human MOGtransfected cell line divided by the geometric mean channel fluorescence intensity of the control cell line was calculated.Geometric mean channel fluorescence ratio cutoffs were set to 3 SDs and a 25% surplus above the mean values for the healthy controls of the discovery cohort (IgA ≥2.4,IgG ≥3, IgM ≥1.6).

Statistical Analysis
We used χ 2 and Fisher exact tests for categorical variables.For continuous variables, we used unpaired t tests.The significance cutoff was set at P < .05.For optical coherence tomography analyses, we performed linear mixed models at eye level with correction for age and sex (fixed effects) to account for intraparticipant, intereye dependencies.We used Prism 9 version 9.4.1 or R version 4.1.3(packages: ellipsis, pastecs, readxl, ggplot2, car, lmerTest, MuMIn, Matrix, carData and lme4).Further details are described in the eMethods in Supplement 1.

Key Points
Question What is the frequency of immunoglobulin (Ig) A antibodies against myelin oligodendrocyte glycoprotein (MOG) in patients with central nervous system (CNS) demyelination, and do these antibodies associate with a distinct clinical phenotype?
Findings In this longitudinal study, a subgroup of patients with demyelinating disorders was double-seronegative for aquaporin 4 (AQP4) IgG and MOG-IgG but seropositive for MOG-IgA.These patients presented with frequent myelitis and brainstem syndrome, infrequent optic nerve involvement, and a low percentage of cerebrospinal fluid-specific oligoclonal band positivity.

Meaning
The findings suggest that MOG-IgA may be a novel diagnostic biomarker in a distinct subgroup of AQP4-/MOG-IgG double-seronegative patients with CNS demyelination.layer and ganglion cell-inner plexiform layer thicknesses in eyes of patients with isolated MOG-IgA and optic neuritis were not different from those of MOG-IgG patients with optic neuritis (eFigure 3 in Supplement 1).Additionally, no significant differences in the frequency of disease manifestations were detected in other MOG-Ig isotype groups (MOG-IgM, MOG-IgG/A, MOG-IgG/M), except for a difference in optic neuritis frequency comparing isolated MOG-IgA with isolated MOG-IgG (35/55 [64%]) (eFigure 2 in Supplement 1).

Discussion
We identified isolated MOG-IgA in a small subset of patients presenting with myelitis, brainstem syndrome, and infrequent optic neuritis overlapping with core clinical features of NMOSD 7 and MOG antibody-associated disease. 8While the coexistence of MOG-IgM and MOG-IgA has previously been described 12 in a similar frequency as detected in our cohort, we expand on the existing literature by reporting isolated MOG-IgA seropositivity in patients seronegative for MOG-IgG/-IgM and AQP4-IgG.
Unlike IgG, which is mounted systemically, IgA is mainly produced in mucosal tissues where it serves as a first-line barrier against pathogens and commensals, raising questions about the different mechanisms of immune activation that lead to divergent MOG-Ig responses.Although a high frequency of patients who were seropositive for isolated MOG-IgA showed records of attacks preceded by infections or vaccinations, we did not observe associations with specific triggers.An alternative explanation for the occurrence of isolated MOG-IgA could be subsequent seroconversion from MOG-IgM or MOG-IgG induced by the inflammatory milieu.While our longitudinal data of unchanged MOG-Ig isotype patterns over time argue against this, little is known about diseasespecific induction of isolated IgA responses. 9Future studies are required to investigate the clinical relevance of both isolated and coexisting MOG-IgG/-IgA seropositivity.
In contrast to IgG, which is known for its proinflammatory role through complement activation, 4 ,6 the pathogenic potential of IgA is debated. 9evidence suggests that IgA may target neuronal and myelin antigens 13,14 in CNS inflammation, and a proinflammatory role via IgA immune complex formation and subsequent immune activation has been desc ribed in several diseases. 9The distinct clinical syndrome in patients seropositive for isolated MOG-IgA, characterized by frequent inflammation of the brainstem and spinal cord, areas with high blood-brain barrier permeability, 15 further suggests that IgA may have a pathogenic role in CNS inflammation.Prospective studies investigating immune activation mechanisms and transferring MOG-IgA into animals will be important steps to assess pathogenicity and clarify the etiology of MOG-IgA-associated disease.

Limitations
Our study has several limitations.First, the clinical data were mostly obtained retrospectively with some unavailable clinical variables; therefore, we cannot exclude the possibility of recollection bias.Second, serum samples were not always collected from untreated patients, possibly underestimating the detected frequency of MOG-IgA/-IgG/-IgM.Further, the small number of patients seropositive for isolated MOG-IgA may have underpowered the detection of additional clinical and other differences, compromising the generalizability of the findings.

Conclusions
In this study, MOG-specific IgA was identified in a subg ro u p o f p at i e nt s w h o we re d o u b l e -s e ro n e g at ive for AQP4-/MOG-IgG and presented with distinct clinical features.This finding suggests a potential use of MOG-IgA as a biomarker in AQP4-/MOG-IgG double-seronegative CNS demyelination.Further prospective studies are required to enhance the characterization of the syndrome and decipher underlying pathogenic mechanisms.

Figure 1 .
Figure 1.Study Design and Frequency of Isolated Myelin Oligodendrocyte Glycoprotein (MOG) Immunoglobulin (Ig) A in Central Nervous System Demyelination

Table .
Demographic and Clinical Features of Patients Who Were Seropositive for MOG-IgA and MOG-IgG aAll P values for comparisons of characteristics between groups were nonsignificant.b Patients who were seropositive for MOG-IgG regardless of coexistence of MOG-IgA and/or MOG-IgM.