Advances in the Understanding and Diagnosis of Neuromyelitis Optica and Transverse Myelitis

Brian G. Weinshenker MD
Department of Neurology
Mayo Clinic College of Medicine
Rochester MN

I am pleased that The Transverse Myelitis Association is publishing the following three articles, which are important to patients with neuromyelitis optica (Devic’s disease) and also to patients with transverse myelitis.  Our discovery of NMO-IgG, a biomarker for patients with neuromyelitis optica, and our subsequent discovery that approximately 40% of patients with “idiopathic” long spinal cord lesion-type transverse myelitis are also positive and at risk for recurrence has been a major step in the understanding of transverse myelitis.  This discovery has allowed us to identify patients who are at risk for recurrence of transverse myelitis.  Perhaps most importantly, discovery of the molecular target of this antibody, a water-channel protein called aquaporin-4, may ultimately lead to the development of specific treatments for these conditions.

The first article was published in the Lancet in 2004.  It describes the discovery and characteristics of NMO-IgG.  This antibody, present in the serum of patients with neuromyelitis optica, reacts with brain blood vessels and the surface lining of the brain called the pia.  NMO-IgG is detected by a technique called immunofluorescence.  This involves a search for the pattern of staining of mouse tissue by a human antibody in the serum of a patient after the antibody is detected by a second anti-human antibody with a fluorescent tag.  The location of the antibody staining from patients with NMO in mouse tissue suggested that it may target a protein that is at the interface between blood and the brain substance called the blood brain barrier.  The localization both along the lining of the brain and on the outside of blood vessels further implicated a protein target in the “foot processes” of astrocytes.  Astrocytes are supporting structural cells of the central nervous system.  They have long extensions some of which abut blood vessels and the lining of the brain (glia limitans).  This pattern of staining later turned out to be a major clue as to the nature of the target protein for this antibody.  In this article, our team studied a group of patients with neuromyelitis optica.  NMO-IgG was present in approximately 70 percent of those patients.  It was present in a high proportion of patients with either recurrent or first-event transverse myelitis.  It was also present in a smaller percentage of patients who had recurrent attacks of optic neuritis but had no evidence of myelitis.  This suggested that the spectrum of neuromyelitis optica may be broader than previously recognized and may include patients who have only optic neuritis or only long spinal cord lesion-type transverse myelitis. 

The second paper published in the Journal of Experimental Medicine in 2005, details definitive experiments performed by Dr. Vanda Lennon and colleagues at Mayo Clinic that identified aquaporin-4 as the target of NMO-IgG.  They demonstrated that a specific antibody directed to aquaporin-4 gave exactly the same pattern on mouse tissues as NMO-IgG.  Furthermore, Lennon and colleagues demonstrated that there was no NMO-IgG staining of brain tissues of a “knock out” mouse that was raised from an embryo from which the aquaporin gene had been deleted.  Finally, Lennon et al were able to show that the antibody was able to directly bind to and precipitate aquaporin-4 protein while none of the other proteins which bind aquaporin-4 to the cell membrane were precipitated by NMO-IgG.  “Precipitation” means that when the antibody specifically recognizes the antigen to which it directs, it forms an aggregate of antibodies that “falls out of solution” and can be captured and measured.  Only aquaporin-4 and none of the other proteins in the complex of proteins that holds aquaporin-4 in the cell membrane were precipitated by NMO-IgG. 

This work provided definitive evidence that aquaporin-4 was the target of NMO-IgG.  It suggested that aquaporin-4 was the target of the immune attack in neuromyelitis optica.   It also raised the intriguing possibility that by binding to these water channels, NMO-IgG might lead to problems with water transport in the brain which might explain some of the unusual symptoms experienced by some patients with NMO.  The aquaporin-4 paper is rather complex and will be difficult for most lay people to understand in detail.  Hopefully, my summary comments will assist you in focusing on the results of this study and the significance of these findings. 

In the third article, published in the Annals of Neurology in 2006, the Mayo Clinic neuromyelitis optica team studied patients with a first event of long spinal cord lesion-type transverse myelitis.  These are patients who on MRI have a lesion extending over three or more vertebrae in length.  They found that approximately 40 percent of patients tested positive for NMO-IgG.  Furthermore, in follow up, over 50 percent of patients with a single event of transverse myelitis who were positive to NMO-IgG developed relapses of transverse myelitis or developed optic neuritis within one year.  None of the patients who tested negative for NMO-IgG experienced relapses.  This suggested that many patients who present with the first event of transverse myelitis may have a limited form of neuromyelitis optica and are at risk for relapse.  On this basis, our group suggested that patients with a first event of transverse myelitis who test positive for NMO-IgG should be treated with immunosuppressive drugs to lessen the chance of a second or subsequent attack of transverse myelitis even before clinical criteria for neuromyelitis optica are satisfied.

Our group is continuing to work to prove the pathogenic significance of this antibody.  Specifically, we are trying to determine whether this antibody is merely a marker of the damage seen in transverse myelitis and neuromyelitis optica, or whether it is actually the perpetrator of the damage.  We are optimistic that it may be the actual perpetrator given the close proximity of where aquaporin-4 is located and the major pathological findings in spinal cord tissue samples from patients with neuromyelitis optica.  Both aquaporin-4 and antibodies and inflammatory proteins called complement are lined up on the outside of blood vessels.  This suggests that the antibody may actually cause the damage that occurs in these conditions. 

I would like to thank Sandy Siegel and The Transverse Myelitis Association for allowing us to publish these papers for your review.  We hope that our presentation of this information on the NMO-IgG antibody adds to your understanding of these complex neuroimmunologic disorders.