Neuromyelitis optica spectrum disorder (NMOSD)

Doctors diagnose NMOSD from symptoms, an exam, MRI scans and blood tests. There must be at least one typical attack. Examples are inflammation of the optic nerve (optic neuritis), inflammation of the spinal cord (myelitis), or a bout of hiccups, nausea or vomiting that will not stop (area postrema syndrome). The key blood test looks for aquaporin-4 (AQP4) antibodies. An international expert panel recommends a cell-based test. In the studies it reviewed, this test found the antibody in about 3 in 4 people with NMOSD and rarely gave a false positive.

MRI of the brain and spinal cord looks for typical damage. A long spinal cord lesion that stretches over three or more vertebrae (longitudinally extensive transverse myelitis) is the most specific MRI sign of NMOSD and is very uncommon in adults with multiple sclerosis (MS). A spinal tap (lumbar puncture) can also help, because some spinal fluid findings are useful in telling NMOSD apart from MS.

When the AQP4 test is negative, the diagnosis needs stricter rules and more MRI evidence. The expert panel says retesting should be considered if a person with a negative test relapses, and before starting certain immune treatments. Doctors may also test for MOG antibodies, which point to a separate condition called MOGAD. Getting the diagnosis right matters, because some MS medicines appear to make NMOSD worse.

There is no cure, but medicines can make attacks less likely. During an attack, doctors use strong steroid medicines and sometimes plasma exchange, which filters harmful antibodies out of the blood.

Long-term medicines help prevent attacks. In the United States, four are approved for adults with AQP4 antibodies: eculizumab and ravulizumab, which block part of the immune system called complement; inebilizumab, which removes B cells, the white blood cells that make antibodies; and satralizumab, which blocks the receptor for an immune signal called IL-6. In clinical trials, antibody medicines like these greatly lowered the risk of attacks. Medicines that calm the immune system more broadly, such as rituximab, azathioprine and mycophenolate, are also used.

An autologous stem cell transplant uses strong chemotherapy to reset the immune system, then gives back the person’s own stored blood-forming cells. In 2025, European experts in MS and transplantation said the evidence is still not enough to support it for NMOSD outside clinical trials. They said this is mostly because highly effective medicines are now available, and that a transplant could be considered as a rescue option when nothing else has worked.

Results have been mixed. A European transplant registry study looked at 16 people with hard-to-treat NMOSD treated from 2001 to 2011. The transplant controlled the disease for a time, but 13 of the 16 later relapsed or worsened and needed more treatment. A small US study treated 13 people. One, who also had active lupus, died of lupus complications. Of the other 12, most were free of attacks without immune medicines five years later. None of those whose AQP4 antibodies disappeared had another attack, but the two whose antibodies stayed relapsed within two years. Transplants from a donor have been described in only a few people.

NMOSD varies from person to person. Most people have the relapsing form, with attacks months or years apart. There is usually partial recovery between attacks. About 5 to 10% have a single episode. Each attack can damage new areas of the spinal cord and brain, so the number and severity of attacks shape how a person does over time. A severe attack in the neck part of the spinal cord can affect breathing. In an Argentine registry, this was the most common cause of death.

NMOSD can leave lasting harm. In a population study in Olmsted County, Minnesota, and Martinique, 60% of people living with NMOSD at the end of 2011 were blind in at least one eye, needed a walking aid, or both. In a Danish national study, people whose disease started later in life had a higher risk of dying.

Newer long-term medicines have made attacks much less common in clinical trials. In the PREVENT trial of eculizumab, only 3 in 100 people on the medicine had an attack, compared with 43 in 100 on a placebo. The trial did not show a clear difference in disability between the two groups. Because each attack can add new damage, preventing attacks is the main aim of long-term treatment.