Acute attacks of NMO can be treated with high-dose corticosteroid therapy (intravenous methylprednisolone 1000 mg daily for five consecutive days) [68, 69]. A recent study showed that only 17% of NMO relapses completely resolve following high-dose corticosteroid therapy; the majority of attacks demonstrate partial response (65.4%) , thereby suggesting that we once again emphasize the emerging adage that “time is tissue,” and as such, the concomitant employment of corticosteroids and other treatment modalities, as close to the inception of the “ictus” as possible, likely has the best chance to accelerate recovery.
With respect to combining corticosteroids with other treatment strategies germane to both limiting the magnitude of tissue injury and the corresponding compromise in the patient’s neurologic repertoire of capabilities. Another principal objective for the application of combination regimens is the prospect that particular measures may be effective in also accelerating the process of attenuating mechanisms that foment further inflammation, the vasogenic edematous burden within CNS tissue compartments that are inherently at greater risk of permanent damage and disorganization of their complex tissue architecture, at least in part, by virtue of their conspicuously limited compliance characteristics (i.e., small, perhaps even negligible changes in augmented tissue edema can result in escalation in compartment pressure, thereby resulting in altered flow characteristics with respect to fluid clearance and water homeostasis, both intracellularly as well as extracellularly).
Perhaps the neuroradiologic features of greatest conspicuity are those that also carry important prognostic ramifications, most specifically as they relate to the burden of residual physical disability, as a derivative of NMO-associated syndromes. Specifically, the longitudinally extensive spinal cord distribution of neuropathology associated with NMO reflects, in part, the movement of tissue water, both along and across tissue barriers.
As vasogenic edematous processes continue unabated, there is a dangerous and rapidly converging phenomenon that brings the expansion of spinal cord tissue water on a “collision course” with autoregulatory mechanisms that maintain adequate blood perfusion both across the transverse and caudal-rostral extent of the spinal axis.
Without rapid intervention to control the expansion of the vasogenic edema burden, the dynamic range of such compensatory responses is exceeded, with the potentially catastrophic consequence of embarrassed blood flow dynamics, including the failure of the circumferentially organized vascular arborization of the well- recognized vasocorona (which represents a final common vascular pathway for collateralizing ischemic changes across the transverse and longitudinal spinal axis).
Plasmapheresis (typically five full volume exchanges) has been shown to be an effective treatment for NMO relapses (as a first-line treatment or in treating steroid- refractory patients) [34, 70-76]. Since NMO is a humoral complement-mediated astrocytopathy, plasmapheresis is hypothesized to ameliorate attacks by removing pathogenic antibodies, activated complement, and cytokines from the circulation. Escalating therapy with plasmapheresis has been shown to significantly improve remission in corticosteroid-refractory patients . In fact, early plasmapheresis has been shown to produce a better clinical outcome [34, 75]; in NMO patients with TM, early plasmapheresis is critical since TM is associated with a poor outcome . In a small cohort of ten patients, intravenous immunoglobulin was shown to be beneficial in treating those who failed to stabilize with corticosteroids with or without plasmapheresis .
Once the acute attack of NMO has been stabilized, immunotherapy aimed at preventing further attacks should be instituted as soon as possible, since NMO relapses are potentially devastating. The importance of distinguishing MS from NMO is underscored by the fact that disease-modifying agents employed in MS, like interferon-beta, natalizumab, and fingolimod, are not only ineffective in NMO, but have been shown to aggravate NMO disease activity [78-81].
In MS, over time, the vast majority of patients will eventually transition from a predominantly relapsing-remitting course of both clinical and radiographic exacerbations, into the more insidious, often even imperceptible, recalcitrant, and, until recently, treatment-resistant phase of disease progression, the highly stereotyped signature of irreversible compromise (or even complete abolishment) of critical functional neurologic capabilities. In contradistinction, NMO has not been associated with a similar “progressive” course, but rather, the exacerbations themselves represent the principal corpus of activities, and the incomplete recovery from them, that drive the accrual of disability . The rapid identification of the NMO clinical syndromes, followed by the expeditious employment of intensive and often combination therapy, is in keeping with accelerated cessation of the attack and its associated mechanisms. Without equivocation, immunosuppressive therapy to reduce humoral immune activity constitutes the mainstay for NMO disease-modifying treatment.
Azathioprine (AZA) was the first agent shown to be effective in preventing attacks and is typically used at doses of 2-3 mg/kg/day, often in combination with oral prednisone (1 mg/kg/day) [23, 82]. Potential adverse effects of AZA include transaminitis, leukopenia, gastrointestinal upset, recurrent infections, myelosup- pression, and increased risk of lymphoma [82-85]. Although successful in reducing relapses, the majority of patients are unable to tolerate this regime; furthermore, many relapse when prednisone is tapered below 5-15 mg/day [69, 82]. In such cases, it would be better to consider alternative therapy to avoid the complications of chronic corticosteroid use. It is also important to remember that AZA should be avoided in patients with low thiopurine methyltransferase activity since this population is at risk of myelotoxicity .
In a retrospective study of 24 NMO patients, treatment with mycophenolate mofetil (MMF) has been shown to reduce relapses and stabilize the disease course . A median dose of 2000 mg/day was used . In our experience, MMF is a useful oral immunosuppressive agent to control disease activity in NMO and is far better tolerated than AZA. Potential adverse effects of MMF include gastrointestinal upset, photosensitivity, recurrent infections, and myelosuppression . We recommend checking blood counts, renal function, and liver function every 3 months in patients on MMF to monitor for these potential adverse reactions.
Rituximab, a chimeric anti-CD20 monoclonal antibody that depletes B-cell and plasmablast levels, has been shown to be very effective in treating NMO [83, 8894], underscoring the role of B-cell dysregulation in NMO etiopathogenesis. Rituximab is well tolerated; typical adverse reactions are infusion related (fever, chills, rash, angioedema, bronchospasm, and hypotension) and, infrequently, cardiac arrhythmias . One important safety consideration before commencing rituximab therapy is to test for hepatitis B and C since rituximab has been associated with reactivation of these diseases [95-97].
While the precise rituximab-dosing interval in NMO is not clear, our center monitors monthly CD19 cell counts, and when cell counts begin to recover (i.e., when the percent of CD19 cells approaches 1%), we initiate re-treatment. While rituximab is an antibody against CD20 localized upon pre-B-cells, we employ the monthly surveillance strategy of specifically ascertaining when the CD19+ fraction is returning and approaching 1-2% . The rationale for emphasizing the utilization of the CD19+ fraction of B-cells is related to the observation that this cell surface antigen is expressed earlier and persists later than CD20. As such, our surveillance strategy has allowed us to “bracket” our treatment in order to avoid “being late” in the re-treatment with the disease-modifying agent for NMO and to thereby avoid additional exacerbations.
It is important to underscore that beyond the 1-2% circulating composition of the CD19 fraction, the reconstitution curve (for CD19+ cells) becomes sigmoidal and thereby reflects the accelerated return of the B-cell fraction, along with the propensity to develop new exacerbations. Our group has investigated the role of rituximab dose magnitude and the duration of CD19 suppression. In essence, a 100 mg dose of rituximab administered intravenously is associated with a mean reduction (i.e., below 1%) of about 3 months, whereas a 1000 mg dose may suppress the CD19 fraction for about 9-12 months . The next generation of anti- CD20 monoclonal antibodies (ocrelizumab and ofatumumab) is currently being studied in MS  and would also be potentially useful for treating NMO.
Mitoxantrone, which is approved for the treatment of MS, is also effective in NMO, but carries significant safety risks (including cardiotoxicity and hematogenous malignancies) [90, 99-101]. Cyclophosphamide is somewhat effective but is poorly tolerated, carcinogenic, and gonadotoxic [83, 102, 103]. Other therapies that have been described in small cohort of NMO patients include oral methotrexate , low-dose periodic oral corticosteroids , cyclosporine in combination with oral corticosteroids , preventive plasmapheresis [107, 108], and glatiramer acetate with  or without  intermittent corticosteroid pulses.
Novel therapeutic strategies that are being explored include eculizumab (a monoclonal antibody that inhibits complement protein C5) , aquaporumab (a nonpathogenic monoclonal antibody that competitively inhibits NMO-IgG) , and tocilizumab (anti-IL-6 receptor monoclonal antibody which prevented NMO relapses and controlled the neurogenic pain of the disease) .
NMO is a humoral autoimmune disease characterized by antibody- and complement-mediated astrocytic destruction. Classic manifestations of NMO/ NMOSD include severe AON, LETM, and the area postrema syndrome. Serologic testing for NMO-IgG (using the most sensitive and specific methods,
i.e., cell-based assays) should always be considered in patients with such symptoms, especially in those without brain MRI lesions that are typical for MS. The vast majority of NMO patients pursue a relapsing course, and as such, immunotherapy to prevent future attacks should be instituted in every NMO patient as soon as possible to avert potentially devastating (or even lethal) relapses.
Disclosures The authors have no relevant financial disclosures.