Article Text

THU0103 Synovial plasma cells predict relapse after rituximab therapy in rheumatoid arthritis
  1. E. Vital,
  2. Y.M. El-Sherbiny,
  3. K. Henshaw,
  4. S. Dass,
  5. S. Das,
  6. M. Buch,
  7. F. Ponchel,
  8. P. Emery
  1. Section of Musculoskeletal Disease, University of Leeds, Leeds, United Kingdom


Background Response to rituximab in patients with RA is followed by relapse after a variable interval. Previous studies have associated initial response to rituximab with lower numbers of CD79a+ B cells at baseline and CD138+ plasma cells at 16 weeks. We recently demonstrated that numbers of plasmablasts, rheumatoid factor titre, and DAS28 at 6 months predicted subsequent outcome [1]. Synovial predictors of duration of response have not previously been described.

Objectives To identify synovial predictors of response and relapse after rituximab

Methods 25 patients received 2×1000 mg rituximab in combination with 2×100 mg methylprednisolone and oral methotrexate. Synovial biopsy was performed before, and either 16 (n=16) or 26 (n=9) weeks after rituximab. Sections were stained to identify Plasma cells (CD138), B cells (CD19, CD20cy), T cells (CD3) and macrophages (CD68) and analysed by digital image analysis. Clinical response was categorised by EULAR response criteria at 6 months. Patients with EULAR moderate or good response at 6 months were categorised at 12 months as relapse (EULAR non-response) or non-relapse (EULAR moderate or good response). Non-parametric statistical tests were used.

Results At 6 months, EULAR non, moderate and good responses were achieved by 7 (28%), 9 (36%)and 9 (36%) patients respectively. Of the 18 patients with EULAR good or moderate response at 6 months, 6 (33%) had relapse at 12 months.

In synovium, following rituximab there was a significant reduction in CD19/CD20cy (p=0.001) and CD3 (p=0.004), and no significant change in CD138 (p=0.506). Follow up biopsies collected at 16 or 26 weeks were compared. No significant differences in DAS28 or synovial markers were identified between these time points.

Post rituximab histology and response at 6 months: No marker was significantly associated with EULAR response, other than a trend to higher CD68 in non-responders after treatment

Post rituximab histology and relapse at 12 months: Comparing relapse and non-relapse groups, there was no difference in CD138 scores at baseline. In the no relapse group there was a trend to reduction in CD138 (p=0.060), whereas in the relapse group there was a trend to increase in CD138 (p=0.180), resulting in significantly higher post-rituximab CD138 scores in patients with relapse (p=0.047).

CD138 after rituximab did not correlate with DAS28, but did correlate with RF titre (R=0.637). Patients with RF >108 IU/mL (the level most predictive of relapse in our previous study) had significantly higher synovial CD138 after rituximab (p=0.010), despite no difference in DAS28 at 6 months (p=0.324).

Conclusions After rituximab, synovial B cells are generally depleted efficiently. In contrast, synovial plasma cells were only partially reduced in a subset of patients, with high post-rituximab numbers predicting relapse The ability of the synovium to support and retain plasma cells is a determinant of clinical outcome of rituximab therapy.

  1. Vital EM et al. Arthritis and Rheumatism 2011: 63(10):1643

Disclosure of Interest E. Vital Grant/Research support from: Roche, Speakers Bureau: Roche, Y. El-Sherbiny: None Declared, K. Henshaw: None Declared, S. Dass Speakers Bureau: Roche, S. Das: None Declared, M. Buch Speakers Bureau: Roche, F. Ponchel: None Declared, P. Emery Grant/Research support from: Roche, Speakers Bureau: Roche

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.