Article Text

Download PDFPDF

Less frequent rituximab retreatment maintains remission of neuromyelitis optica spectrum disorder, following long-term rituximab treatment
  1. Su-Hyun Kim,
  2. Yeseul Kim,
  3. Gayoung Kim,
  4. Na Young Park,
  5. Hyun-Min Jang,
  6. Hyun-June Shin,
  7. Jae-Won Hyun,
  8. Ho Jin Kim
  1. Department of Neurology, Research Institute and Hospital of National Cancer Center, Goyang, South Korea
  1. Correspondence to Dr Ho Jin Kim, Department of Neurology, Research Institute and Hospital of National Cancer Center, Goyang 410-769, South Korea; hojinkim{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Rituximab, a chimeric monoclonal antibody that selectively targets CD20+ B cells, has exhibited robust efficacy and an acceptable safety profile in neuromyelitis optica spectrum disorder (NMOSD).1–3 Previously, we reported that the therapeutic response of B cell depletion varied among 100 patients with NMOSD, which resulted from multiple factors including Fc gamma polymorphism, and that monitoring CD27+ memory B cell appears to improve treatment outcomes via individualised treatment.2 Our treatment strategy has also proven to effectively prevent relapse at a lower cumulative dose, compared with the fixed maintenance therapy every 6 months, in a French cohort.4 Currently, we have a group of patients with NMOSD who have undergone long-term rituximab treatment for more than 7 years. We found that retreatment interval became significantly prolonged over time when we targeted depletion of memory B cells. Here, we analysed clinical outcomes and changes in B cell reconstitution over time, following long-term repeated rituximab treatment.


Through December 2017, sixty-eight patients with NMOSD had been treated with rituximab for at least 7 years. Twenty-one patients who had been treated with mitoxantrone, prior to rituximab, were excluded; the remaining 47 patients were included. Written informed consent was obtained from all patients. After induction therapy, patients received a single infusion of rituximab, at a dose of 375 mg/m2, as maintenance therapy; this was performed whenever the frequency of re-emerging CD27+ memory B cells in peripheral blood mononuclear cells (PBMC) exceeded 0.05% as revealed by flow cytometry for the initial 2 years, and, subsequently, when the frequency exceeded 0.1%. Blood samples were obtained every 6 weeks throughout the first year, every 8 weeks throughout the second year and every 10 weeks thereafter, to evaluate lymphocyte subsets.

We analysed the frequency and number of CD19+ B cells, CD27+ memory B cells and CD27 naïve B cells before each retreatment by year, retreatment interval of rituximab by year, clinical outcome, anti-aquaporin-4 antibody (cell-based immunofluorescence assay) and serum immunoglobulin levels after rituximab therapy. Statistical analysis was performed using GraphPad Prism software (V.6.0; San Diego, CA, USA). The Wilcoxon matched pairs signed-rank test was used for analysis of paired data.


Forty-seven patients had received a median of 12 times (range, 5–19) of repeated rituximab treatment for a median of 9 years (range, 7–12). Clinical characteristics and outcome of patients treated with rituximab are outlined in online supplementary table 1. The annualised relapse rate was reduced by 97% and 29 (62%) patients became relapse-free after rituximab treatment. Disability improved or stabilised in 98% of patients. A total of 29 relapses occurred in 18 patients (online supplementary figure 1). Among 45 (96%) patients with anti-aquaporin-4 antibody, 93% maintained seropositivity after 7 years of treatment. The retreatment interval was significantly extended following repeated cycles of therapy, as the time to repopulation of memory B cells above the therapeutic target became longer: 22 weeks in the first year, 26 weeks in the second year, 32 weeks in the third year, 35 weeks in the fourth year, 36 weeks in the fifth year, 37 weeks in the sixth year and 41 weeks after 6 years (figure 1A).

Supplemental material

Figure 1

(A) Time interval between each rituximab treatment each year. Results are presented as median and range (minimum-maximum). (B) Median naïve/memory B cell ratio each year. Results are presented as median and IQR. (C) Frequency of memory B cells and naïve B cells in peripheral blood mononuclear cells (PBMC) during extended rituximab treatment in patient 25. **p<0.01, ***p<0.001.

The median frequency of memory B cells remained low, as the therapeutic target was constant, whereas the median frequency of naïve B cells exhibited increasing reconstitution before each retreatment, through repeated cycles of therapy (figure 1C). Consequently, an increasing ratio of naïve to memory B cells over time was observed (figure 1B). Due to predominant reconstitution of naïve B cells before each retreatment, the median frequency of CD19+ B cells increased to 2.35%, 3.59% and 4.61% of PBMCs at 3, 5 and 7 years after rituximab treatment, respectively (online supplementary figure 2).

Supplemental material

After 7 years of treatment, 49%, 32% and 23% of patients exhibited low IgM, IgG and IgA levels, respectively. No serious adverse events leading to discontinuation were observed during follow-up.


The aim of the study was to assess long-term clinical outcomes and changes in the reconstitution ratio of naïve and memory B cell subsets over multiple courses of rituximab in patients with NMOSD. After repeated cycles of rituximab targeting depletion of memory B cells, we found sustained clinical response with fewer retreatments than during the initial treatment phase. Given that the memory B cell pool is maintained by homeostatic proliferation,5 there was an additive effect of repeated treatments targeting memory B cells, with delays in reconstitution of memory B cells in peripheral blood. After repeated cycles of rituximab, a shift towards naïve B cells in the reconstituted B cell population is associated with sustained clinical remission, despite significant reconstitution of CD19+ B cells, which increased up to median 4.6% of PBMCs following 7 years of treatment. Taken together, these results suggest that the proportion of memory B cells, rather than the number of total B cells, is a more robust biomarker of rituximab response.

Conventional treatment strategy (fixed-interval rituximab retreatment every 6 months, or retreatment targeting depletion of CD19+ B cells to <1% of lymphocytes) can be insufficient to prevent relapse in some patients particularly during the early phase of rituximab treatment.1 3 Despite depletion of CD19+ B cells (<1% of lymphocytes), more rapid reconstitution of memory B cell subsets, or insufficient depletion of memory B cells during early phases of rituximab treatment, may lead to incomplete clinical response. However, we showed here that less frequent retreatment with rituximab, due to delayed reconstitution of memory B cells, is sufficient to maintain remission after long-term repeated treatment with rituximab. Thus, the conventional strategy may lead to unnecessarily frequent retreatment following 5 years of rituximab treatment which might increase the risk of adverse effects of long-term immunosuppression and associated costs in the majority of the patients. The present results reopen the issue of retreatment timing, especially after several years of treatment, and suggest that monitoring memory B cells might anticipate, and eventually avoid, relapse with the least frequency of rituximab retreatment.

Supplemental material



  • Contributors HJK and S-HK had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: HJK, S-HK. Acquisition of data: HJK, S-HK, GK, NYP, H-MJ, H-JS, J-WH. Analysis and interpretation of data: HJK, S-HK, GK, NYP, H-MJ, H-JS, J-WH. Drafting of the manuscript: S-HK, HJK. Critical revision of the manuscript for important intellectual content: HJK, S-HK, GK, NYP, H-MJ, H-JS, J-WH. Statistical analysis: S-HK. Study supervision: HJK.

  • Funding This work was supported by the National Research Foundation of Korea (Grant No NRF-2016R1D1A1A09916480) and by the Bio and Medical Technology Develop­ment Program (M3A9B6069339) through the Ministry of Science and ICT, Republic of Korea.

  • Competing interests S-HK has received a grant from the National Research Foundation of Korea. S-HK has lectured, consulted and received honoraria from Bayer Schering Pharma, Biogen, Celltrion, Eisai, Genzyme, HanAll BioPharma, MedImmune, Merck Serono, Novartis, Teva-Handok and UCB; received a grant from the Ministry of Science and ICT; accepted research funding from Genzyme, Kael-GemVax, Merck Serono, Teva-Handok and UCB; serves on a steering committee for MedImmune; is a coeditor for the Multiple Sclerosis Journal-Experimental, Translational and Clinical, and an associated editor for the Journal of Clinical Neurology.

  • Patient consent Obtained.

  • Ethics approval This study was approved by the Institutional Review Board of the National Cancer Center.

  • Provenance and peer review Not commissioned; externally peer reviewed.