Article Text

A1.34 ACPA fine specificity is associated with increased plasmablast numbers and worse clinical response to rituximab in rheumatoid arthritis
  1. E Vital1,
  2. L Israelsson2,
  3. L Nogueira3,
  4. V Malmström2,
  5. Yasser El-Sherbiny1,
  6. AC Rawston1,
  7. G Serre3,
  8. L Klareskog2,
  9. F Ponchel1,
  10. P Emery1
  1. 1NIHR Leeds Institute of Rheumatic and Musculoskeletal Medicine & University of Leeds, Leeds, UK
  2. 2Department of Medicine, Karolinska Institute, Stockholm Sweden
  3. 3Laboratory of Epidermis Differentiation and Rheumatoid Autoimmunity, CNRS-Inserm, University Toulouse III, Toulouse, France


Background Quality and duration of response to rituximab in RF/CCP-positive RA is variable. Higher pre-and post-treatment plasmablasts are associated with non/short-response. The objective of this study was to assess the relationship of ACPA fine specificity with clinical response and B cell subsets in RA patients treated with rituximab.

Methods 82 patients with active RA received 2x1000mg rituximab. Clinical response was assessed using DAS28 and EULAR criteria at 0, 6 and 12 months. Presence of antibodies were analysed by ELISA as previously described. One hundred and fifty healthy controls were used to determine the 98th percentile for positivity.

At 0 and 6 month we measured IgM-, IgG- and IgA-RF using commercial ELISAs (EuroImmun); anti-CCP2 using Phadia Immunocap; and antibodies against citrullinated Vimentin60-75(Vim60-75), alpha-enolase (CEP-1), collagen II(CII-C1, CII-U1), a collagen II peptide (CIICitC1), and peptides of the Fibrinogen alpha and beta chain (Fib-alpha, Fib-beta) using in-house assays in University of Toulouse and Karolinska Institute.

B cell naïve (CD19+ CD27-CD38-), memory (CD19+ CD27+ CD38-) and plasmablast (CD19+/-CD27++ CD38++) subsets were enumerated by flow cytometry as previously described.

Results Mean (SD) baseline DAS28 was 5.63(1.08). Presence of each RF isotype was associated with higher DAS28, but individual ACPA epitopes were not associated with baseline disease activity. EULAR Non/Moderate/Good responses at 6 months were 26/38/37% respectively. Clinical response was not related to IgM/G/A-RF status at baseline or change in titre. CCP2 was not associated with any clinical outcome.

EULAR Non-responders had higher baseline titres of CIICitC1(p = 0.038), CEP-1(p = 0.088) and Fib beta-60-74(p = 0.093). There was a trend to lower Mod/Good response rate for patients positive for each specific ACPA epitope - this was significant only for Vim60-75 (61 vs. 84%, p = 0.031). However, the total number of positive epitopes at baseline was significantly higher in patients with EULAR Non response (median (IQR):4(2-5) vs. 2(1-4), p = 0.008).

After a second cycle of rituximab, there was a further reduction in

Vim60-75 (p = 0.043) and CEP-1 (p = 0.068), but not CCP (p = 0.382).

Number of ACPA epitopes positive was associated with higher numbers of plasmablasts in early B cell repopulation at 6 months (Figure 1).

Conclusions Here, for the first time, we associate clinical response and B cell repopulation pattern after rituximab with the presence of ACPA with particular fine specificity. These results suggest that specificity for ACPA epitopes rather than CCP2 is linked to the mechanism of action of B cell depletion in RA, and that plasmablasts in early repopulation are disease-associated.

Abstract A1.34 Figure 1

Plasmablast numbers 6 months after rituximab according to number of ACPA epitopes.

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.