Neuromyelitis Optica

In Devic neuromyelitis optica (NMO), a severe myelopathy is associated with acute or subacute optic neuropathy, with or without recovery. For more than a century, neurologists have debated the precise nosologic classification of NMO among inflammatory demyelinating diseases in general, and multiple sclerosis (MS) in particular. Recently, new criteria have been proposed for both NMO and MS, allowing the 2 entities to be differentiated in most cases. Because a small proportion of cases nevertheless fulfill the criteria for both NMO and MS, challenges persist for clear definitions, especially concerning the roles of immunomodulatory and immunosuppressive treatment.

In 1870, Sir T. Clifford Allbutt, MD,¹ first alluded to an association between spinal cord disease and visual loss. Writing in The Lancet, he said:

The use of the ophthalmoscope as an aid in the diagnosis of encephalic disease is gaining ground quickly, and attention is now being given also to its indications in diseases of the spine . . . I have myself examined, from this point of view, 30 well-marked cases of spinal injury, and in 8 of these I found secondary disturbances within the eye. Of these cases, 17 were severe injuries which proved fatal within a few weeks.^{1 (p76)}

Furthermore, he raised the probing question: "What are the processes which, following the changes in the spine and preceding the changes in the eye, link the two events together in a chain of causation?"^{1 (p77)}

Eugène Devic, a French physician, subsequently described a case² of combined myelopathy and optic neuropathy and inspired his student Fernand Gault³ to write a thesis on the topic. Devic first suggested a separate clinical entity termed neuromyelitis optica, which subsequently became known as Devic neuromyelitis optica.

Neuromyelitis optica may be a monophasic or a multiphasic disease characterized by a severe acute transverse myelitis and acute or subacute, unilateral or bilateral optic neuropathy, with or without recovery. The patient described by Devic in a report published in a French medical journal in 1894² was a 45-year-old woman who presented with "neurasthenia" followed by "bilateral optic neuritis and acute myelopathy."^{2 (p1033)} He gave the following description:

One day later, ophthalmologic examination showed bilateral optic neuritis shortly thereafter complicated by retinal hemorrhage. Bladder dysfunction was associated with transient then permanent motor paralysis. Neuropathological examination revealed an acute and widespread lumbar myelitis and showed a severe bilateral optic neuritis.^{2 (p1033)}

In 1949, Stansbury published an extensive review of the literature⁴ and concluded that:

(1) Neuromyelitis optica may occur at any age but is commoner in persons between 30 and 50 years of age . . . . (2) The disease appears to be commoner in persons of the female sex . . . and is not limited to white race . . . . (3) The incidence of the initial disability was distributed equally between the visual system and the spinal cord.^{4 (pp333-334)}

On the basis of this review of the literature, criteria for NMO were formulated and included patients with a monophasic illness comprising acute or subacute optic neuritis and transverse myelitis occurring within an 8-week period, with demyelinating lesions restricted to the optic nerves and spinal cord at postmortem examination in patients aged between 9 and 60 years.⁴

Recently, however, new diagnostic criteria have been proposed in light of a large series of patients with NMO seen at the Mayo Clinic in Rochester, Minn. These new criteria are less restrictive and propose the inclusion of patients with a longer interval between the myelitis and the optic involvement, thereby enlarging the spectrum of NMO.⁵

There is a consensus in the literature regarding the salient pathological features of NMO seen at postmortem examination. In 1927, Beck⁶ described a case of NMO with pathological findings and said that his case was:

clinically unlike disseminated sclerosis but that pathologically it differs in four main points: (1) Rarefaction going on to the cavitation was found in both spinal cord and the optic nerve and chiasma. (2) Cell exudations containing many polymorphonuclear leucocytes were present in the lumbar regions of the spinal cord and optic chiasma. (3) Demyelination and destruction of the spinal cord extended continuously through fourteen segments. (4) Round-celled perivascular infiltration was found throughout the central nervous system.^{6 (p700)}

Hoffman⁷ also reported, "In some cases of neuromyelitis optica there are degenerative phenomena in the form of malacia in the cord, and the relationship to necrotic myelopathy is a close one."^{7 (p388)}

In 1937, Hassin⁸ claimed that:

a clinical differential diagnosis between MS and neuroptic myelitis is sometimes difficult. Even pathologically, because of the gross appearance of the patches, these two morbid conditions do resemble each other to such an extent that on the basis of macroscopic changes alone the conclusion would be permissible that neuroptic myelitis and MS are one disease process, the former acute and the latter chronic . . . .^{8 (p1093)}

He nevertheless added,

Such a view, however, is not sustained by the microscopic characteristic of the patches and other considerations. In neuroptic myelitis the patches invade the gray matter with as great constancy and severity as they do the white matter, affecting not only the nerve fibers but also the ganglion cells of the anterior and posterior cornua.^{8 (p1094)}

He concluded that, "For these reasons, neuroptic myelitis should be considered a definite morbid entity, different from multiple sclerosis and disseminated encephalomyelitis."^{8 (p1098)}

Despite the many extensive reviews during the past century, the status of NMO as a distinct disease entity has remained uncertain. In a proportion of cases, a specific cause has been identified, and in most cases of uncertain cause there has been much debate as to whether NMO is a variant of MS or a distinct clinical syndrome. In his conclusion, Hassin stated, "Neuroptic myelitis is a definite toxic-infectious disease process, different from multiple sclerosis . . . that may be classified as a form of disseminating softening . . . ."^{8 (p1099)}

Several causes have been proposed, such as infectious diseases (measles, tuberculosis),^{9 - 10} systemic diseases,¹¹ and, more recently, endocrinopathies.¹² In 1957, Miller et al,⁹ referring to a French case report,¹³ stated, "Neuromyelitis optica, as well as occurring spontaneously and after banal infections, has also been observed as a complication of measles."^{9 (p670)} In 1976, April and Vansonnenberg¹¹ described the first case of NMO in systemic lupus erythematosus. The neuropathological examination showed "a severe widespread demyelinative and necrotic myelopathy, with grossly visible shrinkage . . . ."^{11 (p1068)}Other reports have since described cases of NMO with systemic lupus erythematosus or Sjögren syndrome,^{14 - 15} and Hutchinson et al¹⁶ recently suggested that NMO should be considered a primary autoimmune disease.

Neuromyelitis optica carries a poor prognosis. Stansbury⁴ emphasized in 1949,

Neuromyelitis optica may take one of three courses: (1) Regression to complete recovery (2) One or more remissions, with an eventual fatal termination or (3) A progressive downhill course to death . . . . It is generally accepted that the mortality rate of neuromyelitis optica is high, and most authors quote 50 per cent as the expected death rate.^{4 (pp329-330)}

On the subject of therapy, he stated,

When the etiology and pathogenesis of a disease are unknown, therapy is usually unsatisfactory . . . . Arsenic, commonly employed for multiple sclerosis, is contraindicated in cases of NMO. A cure in one instance was attributed to vitamin B therapy.^{4 (p331)}

Currently, high-dose intravenous corticosteroids, immunosuppressive treatment, plasmatic exchange, or intravenous immunoglobulin is proposed, but, as the etiology and pathogenesis of NMO remain unknown, the outcomes summarized by Stansbury in 1949 still hold true.

Although clinicians have been able to delineate clinical, biological, radiologic, and pathological differences between NMO and MS, the relationship between these 2 entities remains a contentious issue, more than a century after the first decriptions by Sir T. Clifford Allbutt and Eugène Devic. While we await the results of further research, the most suitable option is to view NMO simply as a clinical syndrome that can occur in isolation or in the context of other disorders. As the molecular biology of MS is unraveled, comparisons and contrasts can be directed to the rarer and more deadly disorder of NMO.

Corresponding author and reprints: Jérôme de Seze, MD, Department of Neurology, Hôpital R. Salengro, CHRU de Lille, 59037 Lille CEDEX, France (e-mail: j-deseze@chru-lille.fr).

Accepted for publication November 26, 2002.