Targeting the Interleukin 6 Receptor to Treat Neuromyelitis Optica.

Abstract

The number of different antibody-associated disorders of the central nervous system and the clinical phenotypes encountered are increasing steadily. Because these patients have potentially reversible diseases, it is incumbent on the neurologist to treat them promptly and effectively. However, although some diseases respond fairly well and quickly to conventional immunotherapies, some can be very challenging. Neuromyelitis optica (NMO) is a severe and disabling condition caused by antibodies to the water channel aquaporin 4 (AQP4), which is expressed on the astrocyte end feet abutting the pia and vascular endothelium. Antibodies to AQP4 lead to complement-dependent inflammation and antigen internalization, and treatment aims to reduce antibody levels. However, as AQP4 antibodies seldom disappear, the patient remains susceptible to further inflammatory lesions of the optic nerve or spinal cord, with each relapse resulting in increased disability and a risk of permanent blindness or paralysis. Thus, relapse prevention must be a major goal of therapy. In addition, neuropathic pain, mostly related to spinal cord involvement, is a prominent and often treatment-resistant feature in NMO1 that has now become well recognized. Conventional treatments include immunosuppression with corticosteroids, azathioprine, and mycophenolate mofetil.2 CD20-directed therapeutic monoclonal antibodies such as rituximab are often used, but they seldom act quickly as they only target the B cells and not the plasma cells that synthesize the antibodies. Encouragingly, a population of CD19intCD27highCD38highCD180− plasmablasts was identified in the peripheral blood lymphocytes of patients with NMO, and these cells appeared to make some of the AQP4 antibodies.3 This process was dependent on interleukin 6 (IL6), and IL-6 levels were elevated in patients with relapsing NMO. Indeed, previous short reports have suggested that tociluzimab, a monoclonal antibody that targets the IL-6 receptor, may be a promising strategy to reduce relapses in NMO.4,5 In this issue of JAMA Neurology, Ringelstein et al6 describe an open-label, retrospective observational study of 8 female patients with NMO (n = 6) or NMO spectrum diseases (n = 2). At the time of tociluzimab administration, the patients had a long duration of disease (mean [SD], 7.9 [7.7] years) and had a high number of relapses (mean [SD], 11.5 [6.5]), with an annualized relapse rate of 4 in the preceding year. They had received various immunotherapies and all had been given rituximab so that their B cells were still depleted (mean [SD] relative amount of CD19+ B cells, 1.4% [1.3%]). Impressively, the annualized relapse rate decreased from 4 to 0.4 during the mean (SD) of 30.9 (15.9) months following tociluzimab treatment. Few adverse effects were observed in this study or in the substantially larger randomized trials of tociluzimab, which confirmed its efficacy in other autoimmune diseases.7 Overall, the results appear to be convincing, particularly in patients 1, 3, 5, and 7. Half of the total of 8 relapses following tociluzimab administration were seen in the first 2.5 months of treatment, perhaps due to a delay in the tociluzimab taking effect, although many serum parameters associated with tociluzimab infusion had been shown to change before this period.8 Alternatively, use of disease-modifying multiple sclerosis drugs may have altered the cytokine balance and increased the chance of relapse before tociluzimab administration in patients 2 and 6. In patients 4 and 8, the reductions in relapse rates following tociluzimab treatment were less clear. Interestingly, the Expanded Disability Status Scale score improved after tociluzimab treatment, but again it is difficult to assess the significance of this as the pretreament Expanded Disability Status Scale score may have been measured during a relapse; it is usual for there to be partial improvement after each relapse, although disability does not revert to baseline. Finally, AQP4 antibody titers, measured by a quantitative cell-based assay, decreased during tociluzimab treatment, consistent with a reduction in antibody synthesis. There are caveats, of course. An observational study has no control group and the authors were not blinded. Such inevitable shortfalls have been shown to result in major biases in previous studies.9 Also, there were varied durations and types of previous treatments and tociluzimab. Some of these drugs, like rituximab,mayhaveasloweronsetofaction inNMO asmany cycles of CD20+ B-cell depletionmay be required before an effect is seen on serum IgGor specific antibody levels. Also, the promising effects observed may have been dependentontheprecedingB-celldepletionbyrituximab.Thisseems less likely as similar efficacy of tociluzimab has been seen in Japanese patients without B-cell depletion.4 Anunexpectedandverypromising result of this studywas the effect on pain. Neuropathic pain is a particularly disabling feature of NMO and does not respond well to traditional antinociceptive medications, often persisting despite successful treatment of relapses. The mechanisms of pain in NMOareunclear, but IL-6hasbeen implicated in this andother forms of inflammatorymyelitis. Indeed, intrathecal adminisRelated article page 756 Opinion