MOG Antibody-Associated Disease (MOGAD): ICMOG 2023 Criteria, Acute Steroids, and Long-Term Immunotherapy
A comprehensive guideline to MOG-IgG testing, distinguishing MOGAD from NMOSD and MS, acute attack treatment, and maintenance immunotherapy.
Executive Summary
MOG antibody-associated disease (MOGAD) is a distinct inflammatory demyelinating disorder of the central nervous system mediated by autoantibodies against myelin oligodendrocyte glycoprotein, formally separated from multiple sclerosis and aquaporin-4 antibody NMOSD by the International MOGAD Panel 2023 (ICMOG 2023) criteria. Clinical phenotypes span bilateral simultaneous optic neuritis, longitudinally extensive transverse myelitis, ADEM, brainstem syndromes, and the distinctive FLAMES (FLAIR-hyperintense lesions in anti-MOG-associated encephalitis with seizures). Diagnosis hinges on a positive serum cell-based assay for MOG-IgG using full-length human MOG, and titer interpretation is decisive because low positives carry a substantial false-positive rate. Acute therapy is IV methylprednisolone 1000 mg daily for 5 days with a slow oral prednisone taper over 6 to 8 weeks, with PLEX or IVIG for severe attacks. Maintenance immunotherapy is selective — many patients are monophasic — with monthly IVIG, mycophenolate, tocilizumab, or (less effectively) rituximab when relapse risk is established.
Table of Contents
- Pathophysiology and Six Clinical Phenotypes
- ICMOG 2023 Diagnostic Criteria
- MOG-IgG Cell-Based Assay: Titers, Pitfalls, and Interpretation
- MOGAD vs AQP4 NMOSD vs MS — Differentiation
- Acute Attack Treatment and Long-Term Immunotherapy
- Practice Insights — Lessons from the Neuroimmunology Clinic
- Frequently Asked Questions
S1. Pathophysiology and Six Clinical Phenotypes
Myelin oligodendrocyte glycoprotein is expressed on the outermost surface of CNS myelin sheaths and the surface of oligodendrocytes. In MOGAD, IgG1-predominant autoantibodies against conformational epitopes of full-length human MOG mediate complement activation, antibody-dependent cellular cytotoxicity, and direct disruption of myelin. The pathology is distinct from multiple sclerosis (T cell-driven plaque demyelination) and from AQP4-IgG NMOSD (astrocytopathy at perivascular foot processes with secondary demyelination). Histologically, MOGAD lesions show perivenous demyelination with relatively preserved axons, complement deposition, and variable inflammatory infiltrates that often resemble the pattern of acute disseminated encephalomyelitis.
Six Core Clinical Phenotypes
The ICMOG 2023 framework recognizes six core demyelinating attack phenotypes, any one of which (in the right serologic and supporting context) qualifies as a MOGAD-defining event. Recognizing the phenotype shapes both the urgency of testing and the differential against AQP4 NMOSD and MS.
- Optic neuritis — the most common phenotype (60 to 70 percent of presentations). MOGAD ON is often bilateral and simultaneous, with prominent optic disc edema, painful eye movements, and longitudinally extensive optic nerve enhancement with characteristic perineural sheath enhancement on orbital MRI. Visual acuity recovery is typically good after steroids, in contrast to AQP4-IgG NMOSD where outcomes are often poor.
- Transverse myelitis — the second most common phenotype. MOGAD myelitis is often longitudinally extensive (3 or more vertebral segments), frequently involves the conus medullaris (a more specific feature than longitudinal extent), and produces characteristic H-shaped gray matter lesions on axial cord T2 imaging. Bowel and bladder symptoms are common, motor recovery is generally better than in AQP4-IgG NMOSD.
- Acute disseminated encephalomyelitis (ADEM) — the dominant childhood phenotype but also seen in adults. Multifocal large T2 hyperintense lesions, encephalopathy, and depressed consciousness with frequent involvement of deep gray matter (thalami, basal ganglia) distinguish it from MS pseudo-ADEM presentations.
- Cerebral monofocal or polyfocal deficits — tumefactive demyelinating lesions, large confluent supratentorial lesions, with focal neurologic deficits referable to lesion location.
- Brainstem or cerebellar syndrome — ophthalmoplegia, ataxia, dysarthria, dysphagia. Important to distinguish from AQP4-IgG area postrema syndrome (intractable hiccups, nausea, vomiting from dorsal medulla lesion), which is much more specific for NMOSD.
- Cortical encephalitis with seizures (FLAMES) — the most recently recognized MOGAD phenotype. Patients present with new-onset focal seizures, headache, and confusion, with MRI showing unilateral cortical FLAIR hyperintensity often involving the frontal or temporal lobe, leptomeningeal enhancement, and elevated CSF pleocytosis. Often misdiagnosed initially as viral encephalitis or autoimmune encephalitis until MOG-IgG returns positive.
Approximately 40 to 50 percent of adult MOGAD patients have a monophasic course, particularly those who seroconvert to MOG-IgG negative within 12 months. The remaining 50 to 60 percent relapse, with the highest risk in the first year after the initial attack. Children with MOGAD are more often monophasic, especially those presenting with ADEM under age 11.
S2. ICMOG 2023 Diagnostic Criteria
The International MOGAD Panel published consensus diagnostic criteria in 2023 (Lancet Neurology) that formally separate MOGAD from MS and NMOSD. The criteria operate as a structured three-element framework: a core demyelinating attack, MOG-IgG positivity by cell-based assay, and exclusion of better diagnoses.
Core Diagnostic Framework
| Element | Requirement |
|---|---|
| 1. Core demyelinating attack | Any one of: optic neuritis, transverse myelitis, ADEM, cerebral monofocal/polyfocal deficits, brainstem/cerebellar syndrome, cortical encephalitis (FLAMES) |
| 2. MOG-IgG positivity | Serum cell-based assay (live or fixed) using full-length human MOG, IgG1-specific secondary antibody |
| 3. Exclusion of better diagnoses | Active rule-out of MS, AQP4-IgG NMOSD, glial fibrillary acidic protein (GFAP) antibody disease, infection, malignancy |
Titer-Stratified Diagnostic Confidence
The 2023 criteria explicitly stratify diagnostic confidence by MOG-IgG titer because low-positive titers (1:10 to 1:40 on serum cell-based assay) carry a meaningful false-positive rate (15 to 30 percent), particularly in patients with MS-typical presentations.
| MOG-IgG Serum Titer | Diagnostic Confidence | Additional Requirements |
|---|---|---|
| Clear positive (1:100 or higher) | Diagnostic if core attack + exclusion criteria met | No additional supporting features required |
| Low positive (1:10 to 1:40) | Requires supporting features | At least one supporting clinical or MRI feature (see below) |
| Negative serum, positive CSF only | Insufficient for diagnosis alone | Use only as adjunctive; requires very suggestive phenotype |
Supporting Features for Low-Positive Titer Cases
When MOG-IgG titer is low positive, at least one of the following supporting clinical or imaging features is required to establish MOGAD diagnosis:
- Bilateral simultaneous optic neuritis
- Optic nerve perineural enhancement on orbital MRI
- Longitudinally extensive optic nerve T2 hyperintensity (more than half the length)
- Longitudinally extensive transverse myelitis (3 or more vertebral segments) involving the conus medullaris
- H-sign on axial spinal cord T2 (gray matter involvement)
- FLAMES cortical encephalitis pattern
- ADEM phenotype
- Absence of MS-typical brain MRI lesions (no periventricular ovoid, no Dawson fingers, no juxtacortical, no infratentorial typical lesions)
- Absence of oligoclonal bands in CSF (less than 80 percent specific but supportive)
Exclusion of Better Diagnoses
Before accepting MOGAD, the clinician must actively rule out alternative explanations: infectious meningoencephalitis (HSV PCR, VDRL, HIV, TB), glial fibrillary acidic protein (GFAP) antibody disease in cases with autoimmune encephalitis features, CNS lymphoma in tumefactive lesions, paraneoplastic syndromes when atypical features cluster, and AQP4-IgG positive NMOSD. AQP4-IgG must always be tested concurrently because dual positive cases are extremely rare and almost always represent a single true diagnosis (usually AQP4 NMOSD if both are weakly positive).
S3. MOG-IgG Cell-Based Assay: Titers, Pitfalls, and Interpretation
MOG-IgG testing is the linchpin of MOGAD diagnosis and is also where clinicians most frequently err. The technical specifications of the assay matter enormously: not all MOG-IgG tests are equivalent, and ordering the wrong test or misinterpreting a low-positive result can lead to either missed diagnoses or inappropriate long-term immunosuppression.
Required Assay Specifications
- Cell-based assay (CBA), live or fixed — gold standard. ELISA and Western blot are not acceptable for MOGAD diagnosis because of unacceptable false-positive rates against linear epitopes that do not mediate disease.
- Full-length human MOG as the antigen — truncated or non-human MOG variants have inferior specificity.
- IgG1-specific secondary antibody — minimizes cross-reactivity with low-affinity IgM and IgA that are not pathogenic.
- Titer reporting from 1:10 through 1:1000 or above with semi-quantitative endpoint dilution.
- Serum is the primary specimen; CSF testing has limited additional yield (10 to 15 percent of serum-negative cases are CSF-positive but with reduced specificity).
Titer Interpretation Table
| Serum CBA Titer | Interpretation | Action |
|---|---|---|
| 1:1000 or higher | Strongly positive, highly specific | MOGAD diagnosis established if core attack present |
| 1:100 to 1:320 | Clear positive, high specificity | MOGAD diagnosis established if core attack + exclusion |
| 1:10 to 1:40 (low positive) | Borderline; 15-30% false positive rate | Require supporting clinical or MRI features per ICMOG 2023 |
| Negative serum | MOGAD unlikely but not excluded if very suggestive phenotype | Consider CSF testing; repeat serum at attack relapse |
When to Repeat MOG-IgG Testing
Repeat MOG-IgG testing is informative at three time points:
- At attack onset, off steroids — ideally before or within 14 days of IV methylprednisolone, because steroids can transiently lower titer.
- At 6 and 12 months after initial attack — persistent high positivity predicts relapsing course; seroconversion to negative predicts monophasic course and informs maintenance therapy decisions.
- At suspected new attack — rising titer or new positivity in a previously seronegative patient supports a true MOGAD relapse.
S4. MOGAD vs AQP4 NMOSD vs MS — Differentiation
The clinical, serologic, MRI, and CSF differentiation between MOGAD, AQP4-IgG NMOSD, and multiple sclerosis is the most important diagnostic exercise in adult neuroimmunology. The three conditions have entirely different acute treatment urgency, maintenance therapy choices, and disability trajectories, so distinguishing them at the first attack matters more than at any subsequent time point.
| Feature | MOGAD | AQP4-IgG NMOSD | Multiple Sclerosis |
|---|---|---|---|
| Antibody | MOG-IgG (CBA, IgG1) | AQP4-IgG (CBA) | None disease-specific |
| Age at onset | Bimodal: children + adults 20s-40s | Median 40s; female predominance 9:1 | Adults 20s-40s |
| Optic neuritis | Often bilateral simultaneous, disc edema, perineural enhancement, good recovery | Often unilateral, chiasmal, longitudinally extensive, poor recovery | Unilateral, retrobulbar, short segment, generally good recovery |
| Spinal cord | LETM, conus, H-sign on axial | LETM, cervical-thoracic, central cord | Short-segment, peripheral, eccentric |
| Brain MRI | Large fluffy lesions, deep gray matter, FLAMES cortical, brainstem (no area postrema) | Area postrema, dorsal medulla, periependymal, hypothalamic | Periventricular ovoid, juxtacortical, Dawson fingers, infratentorial typical |
| Area postrema syndrome | Absent | Specific (intractable hiccups/vomiting) | Absent |
| FLAMES (cortical encephalitis with seizures) | Specific to MOGAD | Absent | Absent |
| CSF oligoclonal bands | Negative (>80%) | Negative (~80%) | Positive (>90%) |
| CSF pleocytosis | Often (especially FLAMES, ADEM) | Often during attacks | Mild, lymphocytic |
| Course | 40-50% monophasic; relapses possible | Uniformly relapsing | Relapsing-remitting then progressive |
| Maintenance therapy | IVIG, MMF, tocilizumab; rituximab less effective | Rituximab, eculizumab, satralizumab, inebilizumab | Multiple DMTs (high-efficacy preferred) |
| Disability accumulation | Generally good recovery; relapse-related | Stepwise relapse-related; severe | Relapse + slow progressive accumulation |
S5. Acute Attack Treatment and Long-Term Immunotherapy
Acute Attack Therapy
MOGAD attacks are highly steroid-responsive, and the speed of treatment initiation correlates with outcome. The standard acute regimen is IV methylprednisolone 1000 mg daily for 5 days, started as soon as the diagnosis is suspected (often before MOG-IgG returns). For pediatric patients use weight-adjusted dosing at 30 mg/kg/day (maximum 1000 mg/day) for 5 days.
The slow oral prednisone taper over 6 to 8 weeks after the IV course is essential and frequently neglected. Rapid taper within 2 to 4 weeks is the single most common cause of post-attack rebound, with relapse occurring as steroids decline. A typical taper schedule is prednisone 1 mg/kg/day (maximum 60 mg) for 2 weeks, then 40 mg for 1 week, 30 mg for 1 week, 20 mg for 1 week, 10 mg for 1 week, 5 mg for 1 week, off.
Adjunctive Therapy for Severe Attacks
Severe attacks (severe vision loss with NEDA-positive optic neuritis, severe motor or sphincter compromise in transverse myelitis, depressed consciousness in ADEM, status epilepticus in FLAMES) warrant adjunctive therapy beyond IV methylprednisolone:
| Adjunctive Therapy | Regimen | When to Use |
|---|---|---|
| Plasma exchange (PLEX) | 5-7 sessions every other day, 1.5 plasma volume per session | Severe attack with inadequate steroid response at 5-7 days; or upfront in fulminant cases |
| IVIG | 2 g/kg total over 5 days (typically 0.4 g/kg/day x 5) | Alternative to PLEX, especially when vascular access difficult; first choice in pediatric patients |
| Continued high-dose IV methylpred | Extend to 7-10 days for inadequate response | When PLEX/IVIG not available; less common practice now |
Long-Term Maintenance Immunotherapy
Maintenance immunotherapy is selective in MOGAD because 40 to 50 percent of adults are monophasic. The decision rests on three considerations: number of attacks (most clinicians wait for a second attack), MOG-IgG titer trajectory (persistently high positives predict relapse), and severity of attacks (a single fulminant attack with poor recovery may justify maintenance even before second attack).
| Maintenance Agent | Dosing | Comments |
|---|---|---|
| IVIG (preferred first-line) | 1 g/kg IV monthly | Well tolerated; emerging evidence shows relapse rate reduction of 70% in adults; first choice in pediatric MOGAD |
| Mycophenolate mofetil | 1000 mg PO BID (target trough 1.5-3 mg/L) | Oral convenience; takes 3-6 months to full effect; cover with steroid bridge |
| Azathioprine | 2-3 mg/kg PO daily (check TPMT before starting) | Lower cost alternative to MMF; bone marrow and LFT monitoring required |
| Rituximab | 1000 mg IV x 2 doses 2 weeks apart, then every 6 months | LESS effective in MOGAD than NMOSD (~40-60% relapse reduction vs 80-90%); reserve for failure/intolerance of IVIG, MMF |
| Tocilizumab | 8 mg/kg IV monthly OR 162 mg SC weekly | IL-6 receptor antagonist; emerging data show efficacy in MOGAD; consider for rituximab failures |
| Avoid: MS-specific DMTs | Natalizumab, fingolimod, dimethyl fumarate, alemtuzumab | May paradoxically worsen MOGAD; never use in confirmed MOGAD |
When to De-Escalate Maintenance Therapy
After 2 to 3 years of relapse-free maintenance therapy with seroconversion to MOG-IgG negative, careful de-escalation can be considered in selected patients. The Histiocyte Society and ICMOG recommendations are cautious here because relapse-after-de-escalation is well documented. A reasonable approach is to extend IVIG dosing interval (every 6 weeks instead of monthly) for 6 to 12 months before discontinuation, with serum MOG-IgG titer monitoring every 3 months and brain/spinal cord MRI at 6 months post-discontinuation.
S6. Practice Insights — Lessons from the Neuroimmunology Clinic
1. Always order MOG-IgG and AQP4-IgG together — sequential testing burns weeks
Early in experienced practice clinicians would order AQP4-IgG first and add MOG-IgG only if NMOSD was confirmed negative. That sequential approach added 1 to 2 weeks of uncertainty during which empiric maintenance decisions were being made on incomplete information. Experienced practice now is to order both antibodies by cell-based assay on the same serum draw for any patient with optic neuritis, transverse myelitis, ADEM-like presentation, or new-onset focal cortical seizures. The cost of two assays is trivial against the cost of an inappropriate DMT decision.
2. Low-positive MOG-IgG in a typical MS phenotype is usually noise
In experienced hands, a MOG-IgG titer of 1:10 or 1:20 in a patient who otherwise has textbook MS — Dawson fingers, juxtacortical lesions, positive oligoclonal bands, retrobulbar ON with good recovery — is almost always a false positive and should not derail the MS diagnosis. clinicians have seen patients reclassified to MOGAD on low-positive titer alone, denied appropriate MS DMT, and then declared MOGAD treatment failures when in fact they had progressive MS. The ICMOG 2023 supporting features framework exists exactly to prevent this misclassification.
3. The 6 to 8 week oral steroid taper is non-negotiable
The single intervention that has reduced early MOGAD relapses in experienced practice is committing to an oral prednisone taper of at least 6 to 8 weeks after IV methylprednisolone, with a written taper schedule given to the patient at discharge. Earlier in experienced practice clinicians would discharge with a 2-week taper and see post-attack relapse in a meaningful fraction; the longer taper has changed that. The corollary lesson is that MOGAD is more steroid-dependent than steroid-responsive, and managing the taper carefully matters more than maximizing the IV dose.
4. IVIG has become the default maintenance, not rituximab
When clinicians trained, rituximab was the default maintenance for both NMOSD and MOGAD, transplanted directly from the NMOSD experience. After watching multiple MOGAD patients relapse on adequate B cell depletion, experienced practice has shifted decisively to IVIG 1 g/kg monthly as the first-line maintenance agent in MOGAD, with mycophenolate as a steroid-sparing oral alternative and tocilizumab as second-line for IVIG failures. clinicians reserve rituximab for the small subset who cannot access or tolerate IVIG and mycophenolate. The mechanistic basis is that MOG-IgG comes from long-lived plasma cells that survive CD20 depletion.
Personal protocol adjustments beyond the published guideline
- Bilateral simultaneous optic neuritis = MOG-IgG send and start steroids today. The bilateral simultaneous pattern is so MOGAD-typical that in practice experienced clinicians send MOG-IgG and AQP4-IgG concurrently and start IV methylprednisolone the same day rather than waiting for the antibody result. Visual outcome depends on time to treatment, and the antibody result will not change the initial management.
- FLAMES presentations get MOG-IgG before HSV PCR. When a previously well adult presents with new focal seizures, headache, and unilateral cortical FLAIR hyperintensity, experienced practice is to send MOG-IgG immediately alongside HSV PCR and start empiric acyclovir — but seasoned clinicians do not wait for HSV PCR to return before considering MOG-IgG as a parallel hypothesis. Several FLAMES patients in clinical experience were initially treated for viral encephalitis for days before MOG-IgG was sent.
- Eight-week oral prednisone taper as the standard, never the exception. The discharge order for every MOGAD attack now includes a written 8-week taper schedule on the discharge instruction sheet. Rapid tapers have caused more early relapses in clinical practice than treatment failures of acute therapy.
- Repeat MOG-IgG at 6 and 12 months — the result drives maintenance decisions. Persistent MOG-IgG above 1:100 at 12 months in experienced practice prompts a conversation about initiating IVIG maintenance even without a second attack, because the relapse risk in this population is high. Seroconversion to negative supports continued observation without maintenance.
Four-panel personal framework: six core phenotypes that should prompt MOG-IgG testing (panel 1), titer interpretation ladder with false-positive caution (panel 2), differentiation from NMOSD and MS by clinical and MRI feature (panel 3), and the four-step cascade experienced practitioners follow from suspicion through maintenance decision (panel 4). Approximate ranges; not institutional outcome data.
S7. Frequently Asked Questions
- Unruptured Cerebral Aneurysm Screening: Who, When, and How
- Aneurysmal Subarachnoid Hemorrhage: Chronic & Long-Term Management
- Brain AVM: Spetzler-Martin Grading, Hemorrhage Risk, and Treatment Decision-Making
- Intracerebral Hemorrhage: Acute Blood Pressure, Anticoagulation Reversal, and Surgical Indications
- Migraine Prophylaxis: Beta-Blockers, Topiramate, CGRP Monoclonals, Gepants, and Botox for Chronic Migraine
- Guillain-Barre Syndrome: Diagnosis, Immunotherapy, and Management