[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
March 2013

Interleukin 6 Receptor Blockade in Patients With Neuromyelitis Optica Nonresponsive to Anti-CD20 Therapy

JAMA Neurol. 2013;70(3):394-397. doi:10.1001/jamaneurol.2013.1246

Objective To report first experiences with interleukin 6 receptor inhibition in therapy-resistant neuromyelitis optica (NMO).

Design Retrospective case series.

Setting Neurology department at a tertiary referral center.

Patients Patients with an aggressive course of NMO switched to tocilizumab after failure of anti-CD20 therapy.

Main Outcome Measures Annualized relapse rate and disability progression measured by the Expanded Disability Status Scale.

Results We report 3 female patients with a median age of 39 years (range, 26-40 years) and aquaporin 4–positive NMO. All patients had been treated with different immunosuppressive and immunomodulating agents, followed by 1 to 3 cycles of rituximab. Despite complete CD20-cell depletion during rituximab therapy, the median annualized relapse rate was 3.0 (range, 2.3-3.0) and the median Expanded Disability Status Scale score increased from 5.0 (range, 4.5-7.0) to 6.5 (range, 5.0-7.0). After the switch to tocilizumab (median duration of therapy, 18 months), the median annualized relapse rate decreased to 0.6 (range, 0-1.3). A total of 2 relapses occurred; however, they were mild and there were no changes in clinical disability.

Conclusions Interleukin 6 receptor–blocking therapy can be effective in therapy-resistant cases of NMO. Larger controlled studies are needed to confirm the efficacy of tocilizumab.

Neuromyelitis optica (NMO) is a rare, severely disabling autoimmune disease, preferentially affecting optic nerves and the spinal cord, structures highly expressing the target antigen aquaporin 4 (AQP4). Because of the severity of NMO attacks, prophylactic therapy must be started as soon as possible. Therapy response in NMO differs from that in multiple sclerosis; immunomodulatory medications (interferon beta and glatiramer acetate) seem to be not effective and natalizumab appears to even exacerbate disease activity.1,2 Several small studies and case report series demonstrated moderate efficacy of a number of immunosuppressive agents.3 More promising results have been shown for B cell–depleting therapy, yet approximately 20% of patients experience relapses despite the complete depletion of B cells.4-6

Recently, specific AQP4-secreting plasmablasts exhibiting a CD19intCD27high CD38highCD180 phenotype have been described in NMO.7 Secretion of AQP4 antibodies by these cells is dependent on interleukin 6 (IL-6), a cytokine also known to be increased in NMO relapses.7,8 To evaluate if the IL-6 receptor–blocking antibody tocilizumab (ACTEMRA) could be a new therapeutic option for NMO, we analyzed the clinical course of 3 patients who were switched to tocilizumab after ongoing aggressive disease despite rituximab therapy.


Patients were identified from a cohort of 18 patients with confirmed NMO treated at a tertiary referral center. We searched for patients with aggressive disease who did not respond to rituximab therapy and had been switched to tocilizumab thereafter. Annual relapse rate and progression of disability, scored by the Expanded Disability Status Scale (EDSS), before and during rituximab and tocilizumab treatment were analyzed. Routine brain and spinal cord magnetic resonance imaging was evaluated for contrast-enhancing new or enlarged lesions before and during both therapies.

Ethical statement

The presented 3 patients were treated on an individual basis according to internal recommendations. All data were analyzed retrospectively from patient files. All patients gave written informed consent for publication.

Demographic and clinical features

Three female patients were diagnosed with AQP4 antibody–positive NMO (Table). All patients had longitudinal spinal cord lesions (>3 segments) and a history of optic neuritis and transverse myelitis. No intrathecal immunoglobulin synthesis or oligoclonal bands were in the cerebrospinal fluid of any of the 3 patients. No patient had concomitant autoimmune diseases. Previous long-term therapies included interferon beta-1a and 1b (n = 2), glatiramer acetate (n = 1), azathioprine (n = 1), and mitoxantrone hydrochloride (n = 1). All patients had experienced pronounced disease activity (median, 13 relapses; range, 4-16; median annualized relapse rate, 2.6; range, 1.7-2.7; median EDSS score, 5.0; range, 4.5-7.0) prior to receiving rituximab. The median disease duration at initiation of rituximab was 8.2 years (range, 2.5-9.4 years) and the median age was 39 years (range, 26-40 years).

Table. Patients' Baseline Characteristics and Clinical Course During Rituximab and Tocilizumab Treatment
Table. Patients' Baseline Characteristics and Clinical Course During Rituximab and Tocilizumab Treatment
Table. Patients' Baseline Characteristics and Clinical Course During Rituximab and Tocilizumab Treatment
Treatment response to rituximab

Rituximab was administered at a dose of 2000 mg/cycle (1000 mg + 1000 mg, patient 1) or 1500 mg/cycle (500 mg + 1000 mg, patients 2 and 3), with a 2-week interval between infusions. Patients 1 to 3 received rituximab for 32 (3 cycles), 8 (1 cycle), and 4 (1 cycle) months, respectively. Despite complete CD20-cell depletion, patients continued to experience relapses (median, 2; range, 1-6), with at least 1 severe relapse, defined by an EDSS score progression of 1.0 or more; the median annualized relapse rate was 3.0 (range, 2.3-3.0) (Table and Figure). Interestingly, in 2 patients, relapses occurred shortly after the first application of rituximab (after 18 and 40 days). Magnetic resonance imaging confirmed acute gadolinium-enhancing spinal lesions in patients 1 and 3. The clinical disability progressed in patients 1 and 2 by 1.5 and 0.5 points, respectively, and remained unchanged, with an EDSS score of 7.0, in patient 3.

Figure. Relapses before and during treatment with rituximab and tocilizumab. Shown are all relapses from month −72 to +21 (patient 1), +18 (patient 2), and +9 (patient 3), relative to the start of tocilizumab treatment. Bars depict duration of rituximab and tocilizumab treatment. In patient 3, disease manifested 31 months prior to the start of tocilizumab treatment.

Figure. Relapses before and during treatment with rituximab and tocilizumab. Shown are all relapses from month −72 to +21 (patient 1), +18 (patient 2), and +9 (patient 3), relative to the start of tocilizumab treatment. Bars depict duration of rituximab and tocilizumab treatment. In patient 3, disease manifested 31 months prior to the start of tocilizumab treatment.

Treatment response to tocilizumab

Because of ongoing disease activity, patients were further switched to the anti–IL-6 receptor antibody tocilizumab. Tocilizumab was administered at a dose of 6 mg/kg every 6 weeks in patients 1 and 2 and every 4 weeks in patient 3. Duration of tocilizumab therapy was 21, 18, and 9 months in patients 1 to 3, respectively. Patient 2 was relapse-free and demonstrated improvement of her gait and EDSS score from 5.0 to 4.0 (Table and Figure). Two patients experienced a mild relapse without EDSS score progression, clinically presenting as isolated optic neuritis in patient 1 and moderate progression of the preexisting motor deficit in patient 3. The relapses occurred 3 and 4 months after tocilizumab treatment initiation, respectively. In both cases, deficits were reversible after receiving steroid pulse therapy. After the therapy regimen in patient 1 had been changed from every 6 to every 4 weeks, she was stable for the next 18 months. Compared with rituximab therapy, the median annualized relapse rate decreased from 3.0 (range, 2.3-3.0) to 0.6 (range, 0-1.3). Tocilizumab therapy was ongoing in all 3 patients.

Tocilizumab was well tolerated. There was no serious infection, malignancy, hypersensitivity reaction, or elevation of transaminase levels. Patient 3 experienced urinary tract infection in the fourth month and mild oral mucosis in the seventh month of therapy. Cholesterol levels had been elevated in patients 1 and 3 prior to tocilizumab therapy and demonstrated no significant increase after therapy initiation.


In this case series, we demonstrated the efficacy of the anti–IL-6 receptor antibody tocilizumab in rituximab-refractory, aggressive cases of NMO. The exact mechanism of action of rituximab in NMO remains unknown. Despite complete depletion of CD20+ B cells, it spares CD20-negative plasma cells and has only moderate and probably indirect effects on antibody production.5,6 Other rituximab effects include monocyte activation and increased synthesis of B-cell activating factor, a key molecule supporting differentiation and survival of B cells as well as immunoglobulin production.6,9 In line with this, an early transient increase of AQP4 antibodies after rituximab application has been reported.10 Notably, in 2 of our patients, relapses occurred early after rituximab initiation.

Inhibition of CD20-negative plasmablasts, which produce pathogenic AQP4 antibodies, might be an alternative treatment strategy. In the presented cases, a switch to tocilizumab led to definite clinical improvement. Recently, another patient successfully treated with tocilizumab was reported by Araki et al.11 They demonstrated a substantial reduction of the frequency of CD19intCD27highCD38highCD18 plasmablasts and titer of anti-AQP4 antibodies within 1 month after tocilizumab initiation. Similarly, tocilizumab decreased the frequency of CD27highCD38highIgD plasmablasts and the titer of anti–double-stranded DNA antibodies in systemic lupus erythematosus as well as titers of rheumatoid factor in patients with rheumatoid arthritis.12

Despite overall clinical stabilization, mild relapses occurred in 2 of our patients as well as in the Araki et al case.11 In patient 1, we reduced the application intervals from every 6 to 4 weeks and the patient was relapse-free afterward. Accordingly, an escalation study demonstrated further clinical improvement in rheumatoid arthritis after dose escalation from 4 to 8 mg/kg.13

As demonstrated in rheumatoid arthritis, tocilizumab has a good safety and tolerability profile. Infections are the most common adverse events; however, rates of serious infections remain low at least for 5 years, demonstrating safety of continuous tocilizumab therapy.13 Possible tuberculosis reactivation and opportunistic infections make careful observations essential. Importantly, C-reactive protein is directly downregulated by tocilizumab and cannot be used as a sensitive diagnostic marker. In our patients, no serious adverse events were observed. Reported urinary tract infection and mild oral mucosis have been treated on an ambulatory basis without any complications.

We propose that IL-6 receptor blockade with tocilizumab is a promising therapeutic option for aggressive, therapy-resistant cases of NMO. Larger controlled studies with longer follow-up periods are needed to confirm the efficacy, safety, and optimal responders for tocilizumab in NMO.

Back to top
Article Information

Correspondence: Ingo Kleiter, MD, Department of Neurology, St Josef-Hospital, Ruhr-University Bochum, 44791 Bochum, Germany (ingo.kleiter@rub.de).

Accepted for Publication: October 4, 2012.

Published Online: January 28, 2013. doi:10.1001/jamaneurol.2013.1246

Author Contributions:Study concept and design: Kleiter, Chan, and Gold. Acquisition of data: Ayzenberg, Kleiter, Schröder, Hellwig, and Chan. Analysis and interpretation of data: Ayzenberg, Kleiter, Chan, and Yamamura. Drafting of the manuscript: Ayzenberg and Kleiter. Critical revision of the manuscript for important intellectual content: Ayzenberg, Kleiter, Schröder, Hellwig, Chan, Yamamura, and Gold. Administrative, technical, and material support: Hellwig, Chan, and Gold. Study supervision: Kleiter and Gold.

Conflict of Interest Disclosures: Dr Kleiter reports receiving travel reimbursements and speaker and consulting honoraria from Bayer HealthCare, Biogen Idec, Merck Serono, and Novartis as well as research support from Bayer HealthCare, Novartis, and Biogen Idec. Dr Hellwig reports research support form Biogen Idec, Bayer HealthCare, Merck Serono, sanofi-aventis, and Novartis Pharma as well as speaker honoraria from Biogen Idec, Bayer HealthCare, Merck Serono, sanofi-aventis, and Novartis Pharma. Dr Chan received speaker and consulting honoraria and scientific grant support from Bayer HealthCare, Biogen Idec, Merck Serono, Novartis Pharma, and Teva Pharma. Dr Gold received speaker and consulting honoraria and scientific grant support from Bayer HealthCare, Biogen Idec, Merck Serono, Novartis Pharma, Chugai Pharmaceuticals, Teva Pharma, and Genzyme.

Funding/Support: This work was supported by grants from the German Ministry for Education and Research (“German Competence Network Multiple Sclerosis”) (Dr Gold).

This article was corrected for errors on April 29, 2013.

Papeix C, Vidal JS, de Seze J,  et al.  Immunosuppressive therapy is more effective than interferon in neuromyelitis optica.  Mult Scler. 2007;13(2):256-25917439893PubMedGoogle ScholarCrossref
Kleiter I, Hellwig K, Berthele A,  et al; Neuromyelitis Optica Study Group.  Failure of natalizumab to prevent relapses in neuromyelitis optica.  Arch Neurol. 2012;69(2):239-24522332191PubMedGoogle ScholarCrossref
Morrow MJ, Wingerchuk D. Neuromyelitis optica.  J Neuroophthalmol. 2012;32(2):154-16622617743PubMedGoogle ScholarCrossref
Jacob A, Weinshenker BG, Violich I,  et al.  Treatment of neuromyelitis optica with rituximab: retrospective analysis of 25 patients.  Arch Neurol. 2008;65(11):1443-144818779415PubMedGoogle ScholarCrossref
Kim SH, Kim W, Li XF, Jung IJ, Kim HJ. Repeated treatment with rituximab based on the assessment of peripheral circulating memory B cells in patients with relapsing neuromyelitis optica over 2 years.  Arch Neurol. 2011;68(11):1412-142021747007PubMedGoogle ScholarCrossref
Pellkofer HL, Krumbholz M, Berthele A,  et al.  Long-term follow-up of patients with neuromyelitis optica after repeated therapy with rituximab.  Neurology. 2011;76(15):1310-131521482945PubMedGoogle ScholarCrossref
Chihara N, Aranami T, Sato W,  et al.  Interleukin 6 signaling promotes anti-aquaporin 4 autoantibody production from plasmablasts in neuromyelitis optica.  Proc Natl Acad Sci U S A. 2011;108(9):3701-370621321193PubMedGoogle ScholarCrossref
Içöz S, Tüzün E, Kürtüncü M,  et al.  Enhanced IL-6 production in aquaporin-4 antibody positive neuromyelitis optica patients.  Int J Neurosci. 2010;120(1):71-7520128675PubMedGoogle ScholarCrossref
Lehmann-Horn K, Schleich E, Hertzenberg D,  et al.  Anti-CD20 B-cell depletion enhances monocyte reactivity in neuroimmunological disorders.  J Neuroinflammation. 2011;8(8):14622027448PubMedGoogle ScholarCrossref
Nakashima I, Takahashi T, Cree BA,  et al.  Transient increases in anti-aquaporin-4 antibody titers following rituximab treatment in neuromyelitis optica, in association with elevated serum BAFF levels.  J Clin Neurosci. 2011;18(7):997-99821565508PubMedGoogle ScholarCrossref
Araki M, Aranami T, Matsuoka T, Nakamura M, Miyake S, Yamamura T. Clinical improvement in a patient with neuromyelitis optica following therapy with the anti-IL-6 receptor monoclonal antibody tocilizumab [published online July 11, 2012].  Mod Rheumatol22782533PubMedGoogle Scholar
Shirota Y, Yarboro C, Fischer R, Pham TH, Lipsky P, Illei GG. Impact of anti-interleukin-6 receptor blockade on circulating T and B cell subsets in patients with systemic lupus erythematosus [published online August 2, 2012].  Ann Rheum Dis22858586PubMedGoogle Scholar
Ogata A, Hirano T, Hishitani Y, Tanaka T. Safety and efficacy of tocilizumab for the treatment of rheumatoid arthritis.  Clin Med Insights Arthritis Musculoskelet Disord. 2012;5:27-4222438671PubMedGoogle Scholar