Is it MS or Neuromyelitis Optica? A New Diagnostic Assay

Abstract & Commentary

Commentary By Brian R. Apatoff, MD, PhD, Associate Professor of Neurology, New York  Presbyterian Hospital-Cornell Campus, Assistant Editor, Neurology Alert.

Synopsis: NMO-IgG immunofluorescence staining for an additional 14 patients with multifocal neurologic findings were initially suspected to be paraneoplastic disorder.

Source: Lennon VA, et al. A Serum Autoantibody Marker of Neuromyelitis Optica: Distinction From Multiple Sclerosis. Lancet. 2004;364:2106-2112.

Lennon and colleagues at the Mayo Clinic tested serum from 102 North American patients with neuromyeltis optica (NMO) and 12 Japanese patients with optic-spinal multiple sclerosis with an indirect immunofluorescence assay, using a composite substrate of mouse tissues that identified a distinctive NMO-IgG staining pattern. Control serum samples were also tested from more conventional multiple sclerosis (MS) patients and other neurological disease. With their assay methodology, the NMO-IgG staining localized to the blood-brain barrier, outlining central nervous system (CNS) microvessels, pia, subpia, and Virchow-Robin space, and partly co-localizing with laminin. The sensitivity of the serum assay was 73% (95%; CI, 60-86), and specificity was 91% (CI, 79-100%) for NMO. For the Japanese variant of optic-spinal MS, the sensitivity of the serum assay was 58% (95%; CI, 30-86), and specificity was 100% (CI, 66-100%). Some of the patients appeared to have high antibody titers in the thousands, although a relatively low titer of 1:60 was the considered to be positive. Lennon et al further established NMO-IgG immunofluorescence staining for an additional 14 patients with multifocal neurologic findings that were initially suspected to be paraneoplastic disorder.


Neuromyelitis optica (Devic’s syndrome) is a severe idiopathic inflammatory demyelinating pathology that preferentially affects the optic nerves, often bilaterally, and the spinal cord, usually with extensive lesions spanning multiple vertebral segments. There is usually an absence of clinical brain involvement, and brain MRIs lack the typical periventricular demyelinating lesions seen with MS. The cerebrospinal fluid may demonstrate inflammatory cells in NMO, but usually lacks increased parameters of IgG synthesis or oligoclonal IgG banding. It has been reported that NMO may respond better to corticosteroids and cytotoxic immunosuppressive agents than the immunomodulatory treatments interferon-beta or glatiramer acetate used in MS.

It is surprising that Lennon et al selected only 22 patients with classical MS as controls. No information is given as to how these few MS serum samples were presumably, randomly selected. It would have been reasonable to include at least as many MS patients as NMO patients, particularly patients presenting with clinically isolated syndromes of severe optic neuritis or transverse myelitis. If this assay can be validated and reproduced in other laboratories, then it may have helpful diagnostic and therapeutic implications. For example, patients with NMO IgG staining might be more effectively treated with interventions directed at an IgG that is presumably pathogenic for this inflammatory pathology. This might include corticosteroids, cyclophosphamide, plasmapheresis, rituxamab, or IgG-absorption methods. By more precisely defining the antigen specificity of the NMO-IgG in the CNS, we can hope to have a more complete understanding of this unique neuroimmunological disorder.