Study Emphasizes Utility of MRI to Help Discriminate Long Myelitis of Neuromyelitis Optica from Sarcoidosis


Clinical and radiologic features distinguishing neuromyelitis optica (NMO) and multiple sclerosis (MS) are already well-accepted. However, there are other neurologic disorders that can be difficult to distinguish from NMO, which sometimes results in misdiagnosis—particularly spinal cord sarcoidosis (SCS), an under-recognized cause of long myelitis that commonly mimics NMO. In an unprecedented study, Mayo Clinic researchers have identified clinical, laboratory, systemic, and radiologic features that, taken together, help discriminate SCS from NMO.

Long myelitis (also known as longitudinally extensive transverse myelitis) involves inflammation of the spinal cord in which the magnetic resonance imaging (MRI) signal abnormality spans three or more vertebral segments in length. Symptoms typically include weakness and numbness in the arms and legs, and bladder problems.

Eoin Flanagan, M.B., B.Ch.

“NMO is a well-recognized cause of long myelitis among the neurology community, but other causes of long myelitis are less appreciated,” says Eoin Flanagan, M.B., B.Ch., Mayo neurologist and first author of the study paper (published in Annals of Neurology). “In our myelitis clinic, we have encountered many patients referred for suspected long myelitis from NMO, who actually had spinal cord sarcoidosis. This was the impetus for our study—to compare long myelitis in these two disorders and help clinicians distinguish between them.”

The Mayo team studied 71 adult patients in all—evaluated between 1996 and 2015, with SCS or NMO, whose first myelitis episode was accompanied by a spinal cord lesion spanning, or greater than, three vertebral segments. Of those 71 patients, 47% subsequently confirmed to have SCS were initially misdiagnosed as NMO, confirming the Mayo team’s suspicion that SCS is a common, under-recognized cause of long myelitis.

The study evaluated clinical characteristics of each case, and spine MRI was reviewed by two neuroradiologists.

“We found that the MRI was quite helpful in distinguishing the two diseases, particularly the gadolinium enhancement pattern,” says Dr. Flanagan. “When neurologists see a long myelitis, we want them to consider sarcoidosis, particularly when gadolinium enhancement extends in a linear pattern along the posterior surface of the spinal cord—something we term ‘dorsal subpial enhancement.’"

“Most patients with sarcoidosis have lung disease, and we recommend a CT chest scan in long myelitis to assess for enlarged hilar lymph nodes, a characteristic of sarcoidosis. A biopsy of these lymph nodes can confirm sarcoidosis, although in a minority of patients without lung disease, spinal cord biopsy was necessary for SCS diagnosis.”

Unlike sarcoidosis, neuromyelitis optica does not typically affect the lungs or other areas of the body. It is predominantly a disease of the spinal cord and optic nerve. Instead of a biopsy, it can often be diagnosed via a blood test called aquaporin-4-IgG. The antibody marker for this disease was first discovered at Mayo in 2004, which was a huge breakthrough in the field of neurology, and it explains why so many NMO patients come to Mayo for their evaluation.

However, 20% of patients have NMO disease without the antibodies. In such cases, the diagnosis “needs to become one of exclusion” and, again, sarcoidosis should be considered in this situation, according to Dr. Flanagan.

Misdiagnosis not only delays treatment, it has the potential to do harm if unsuitable medications are used. For example, there is a drug treatment for SCS, called infliximab, that could potentially worsen myelitis of NMO.

Conversely, plasma exchange is a common treatment for NMO; however, it is not a typical treatment for sarcoidosis, which usually continues to progress if a patient isn’t started on an appropriate (long-term oral steroids) treatment.

The study also outlines a number of demographic, clinical, and spinal fluid differences that, in combination, are highly useful to discriminate between NMO and SCS. Thus, the paper is meant to alleviate what Dr. Flanagan calls the “diagnostic conundrum” of long myelitis. To that end, the MRI pattern remains of particular importance.

“This study emphasizes the utility of MRI in distinguishing the different causes of myelitis,” he says. “In the future, we’re going to continue to study the MRI patterns accompanying myelitis as, in our experience, MRI is a very powerful tool that helps guide clinicians toward the correct diagnosis.”


Christoph Bahn

Christoph Bahn covers emerging research and discovery for Mayo Medical Laboratories. His writing has also appeared in The New York Times, Los Angeles Times, and Smithsonian Air & Space. He divides his time between Southern California and Northwest Ohio.


Please contact me:270-282-3554. Drs in California diagnosed me: 15 yr. Hx of TM(2002) & NMO(diagnosed in 2014. Went to see Dr. Michael Levy(johns hopkins in 2015) he doeant believe I have NMO(but i am tested poaitive for proteins and Aquaphorin4), dr. Harold Moses(Vanderbilt) says I DO have NMO. Not sure what else to do. Dr. Levy discontinued Imuran-Dr. Moses wants me to increase it. Any help would be appreciated.

Hi Lisa, thank you for your comment/question. Unfortunately, we cannot diagnose conditions, provide second opinions, or make specific treatment recommendations through this correspondence. If you would like to seek help from Mayo Clinic, please call one of our appointment offices (Minnesota: 507-538-3270, Arizona: 800-446-2279, or Florida: 904-953-0853).

Is this associated to cluster headache crises or suddenly blurred vision?

Is this associated to cluster headache crisis or suddenly blurred vision? I’ve been diagnosed with Sjogren on 2004, and Sarcoidosis on 2009. Additionally, I’m a Cluster Headache sufferer, what makes me almost disable during crisis… Can this be associated with any of this myelitis? Please, return to me.. Thanks a lot.

Hi Marcia, we are not aware of an association of these disorders with cluster headaches. Vision loss can occur with optic nerve involvement with either disorder, although is more common with neuromyelitis optica than sarcoidosis. The vision usually worsens over days but can sometimes be sudden. – Dr. Flanagan

Comments are closed.