A Disease Waiting For Its Assay: The History of MOGAD

For roughly twenty years, MOG antibodies were considered noise. Studies kept finding them — in patients with MS, in patients with other demyelinating diseases, in healthy controls. The conclusion the field drew was reasonable: these antibodies probably aren’t doing much. That conclusion was wrong. The antibodies were real. The assay was broken.

The protein

Myelin oligodendrocyte glycoprotein, MOG, is expressed on the outermost surface of the myelin sheath. Its location matters: it sits on the very outside of oligodendrocytes, fully exposed to the immune system. This makes it a structurally logical target for antibody-mediated attack. Researchers noticed this in the 1980s, when MOG was identified as a potent inducer of experimental autoimmune encephalomyelitis — EAE — the classic animal model used to study multiple sclerosis. MOG-immunized animals developed demyelinating disease. The inference seemed obvious: MOG must be important in human MS, too.

That inference was wrong, or at least overstated. But it launched decades of research into MOG antibodies in human demyelinating disease — research that was almost immediately complicated by the tools available to detect them.

The assay problem

The problem was ELISA.

Enzyme-linked immunosorbent assay, ELISA, is a workhorse of antibody detection. It works by coating a solid surface with the antigen of interest — in this case, MOG protein — and then exposing it to patient serum. If antibodies are present, they bind. The trouble is that coating a surface with purified protein requires denaturing it: stripping it out of its native environment, unfolding it, and adhering it flat. What was once a three-dimensional glycoprotein sitting in a lipid bilayer is now a linearized string of amino acids on a plastic plate.

For MOG, this matters enormously. The antibodies that are actually relevant in MOGAD recognize a conformational epitope — the specific three-dimensional shape of MOG’s extracellular domain as it exists in a cell membrane. Denatured MOG doesn’t have that shape. So ELISA-based assays were detecting antibodies against linear epitopes, finding them in patients with MS, patients with other demyelinating diseases, and healthy controls. The field grew appropriately skeptical. MOG antibodies looked like noise.

The fix: cell-based assays

The correction came in 2011 and 2012, with the development of cell-based assays. The approach is straightforward in principle: instead of adhering purified protein to a plate, you transfect cells to express full-length, native MOG on their surface. Patient serum is then incubated with these cells. If MOG-specific IgG is present, it binds to the correctly folded extracellular domain. A fluorescently labeled secondary antibody tags the bound IgG, and flow cytometry — FACS — quantifies the signal. The protein stays where it belongs, embedded in a lipid bilayer, presenting the same conformation the immune system encounters in vivo.

The improvement in specificity was dramatic. False positives largely disappeared. A real signal emerged.

The door that opened first

This methodological breakthrough landed in fertile soil, because the field had already been primed to look.

In 2004, Lennon and colleagues published a landmark finding: antibodies against aquaporin-4 — AQP4 — were present in a substantial subset of patients with neuromyelitis optica, or NMO. NMO had long been considered a severe variant of MS. The AQP4 discovery proved otherwise. Here was an antibody-mediated demyelinating disease, clinically and serologically distinct from MS, hiding in the seronegative-MS wastebasket. The discovery raised an obvious question: what was driving disease in the patients who were AQP4-seronegative?

A disease takes shape

Starting around 2011, groups from Oxford, Munich, and Melbourne began identifying patients — many AQP4-seronegative — with MOG-IgG detected by cell-based assay. Their clinical features were distinctive. Bilateral or simultaneous optic neuritis, sometimes with severe disc edema. Longitudinally extensive transverse myelitis. Acute disseminated encephalomyelitis, particularly in children. Cortical encephalitis. Between attacks, patients often recovered surprisingly well — better than typical MS or AQP4-positive NMOSD. The disease appeared steroid-responsive in ways that also distinguished it from its neighbors on the demyelinating spectrum.

This was not MS. It was not AQP4-positive NMOSD. It was something new, or rather, something old that we had finally developed the tools to see.

The term MOGAD was formally adopted around 2018 and 2019 to reflect this recognition — a distinct nosological entity with its own clinical phenotype, its own demographic predilections, and its own emerging treatment ladder. In 2023, Banwell and colleagues published international consensus diagnostic criteria in The Lancet Neurology, formalizing what years of cohort data had been building toward.

Where apheresis enters

Acute attacks are typically treated with high-dose corticosteroids. For patients who don’t respond, intravenous immunoglobulin is a reasonable next step. For steroid-refractory cases, therapeutic plasma exchange enters the picture — mechanistically sensible for an antibody-mediated disease, since removing circulating IgG directly targets what appears to be the primary effector. There is a growing body of case series and retrospective data supporting PLEX in refractory MOGAD attacks, though robust prospective trial data remains limited.

The bottom line

That last sentence captures something true about MOGAD more broadly. We have a name. We have diagnostic criteria. We have a treatment ladder with reasonable mechanistic logic supporting each rung. What we are still working out — actively, with ongoing trials — is the natural history, the optimal long-term immunosuppression, and the full spectrum of what the disease can look like, particularly in its cortical presentations.

MOGAD went from animal model curiosity to false lead to distinct human disease over roughly four decades. The trajectory is a good reminder that the biology usually gets there before we do. Sometimes we’re just waiting for the right assay.