Article Text

A8.13 Changes in the relationship between BAFF, BAFF-R and B cell phenotype following rituximab in patients with systemic lupus erythematosus
  1. Venkat Reddy1,
  2. Gabriel Martin Garcia2,
  3. David Isenberg1,
  4. Maria Leandro1,
  5. Geraldine Cambridge1
  1. 1Centre for Rheumatology, University College London, Rayne Building, London, U. K. WC1E 6JF
  2. 2Hospital General Gregorio Marañón, Madrid, Spain


Background and Aims In vitro, BAFF (B-lymphocyte activating factor) differentially promotes the survival of B-cells and plasmablasts through receptors for the key B-cell cytokine BAFF. Following rituximab (RTX) therapy in patients with systemic lupus erythematosus (SLE), serum BAFF levels increase. The aim was to determine the effect of RTX treatment on relationships amongst BAFF, B-cell sub-populations and receptors for BAFF (BAFF-R).

Materials and Methods Twenty-three RTX-naïve (RTX-N) and 12 RTX-treated (RTX-T) SLE patients and six healthy controls (HCs) were included. RTX-N and RTX-T patients had comparable organ involvement, background therapy (including corticosteroids) and baseline%CD19+ B-cells. The relationships amongst serum BAFF levels (ELISA),%B-cell sub-populations (FACS analysis using monoclonal antibodies to CD19, IgD, CD27, CD38) and%BAFF-R + cells within each B-cell sub-population were compared between RTX-N, RTX-T and HCs. Data were analysed using Mann Whitney U and Spearman Rank correlation tests.

Results BAFF levels were significantly higher in RTX-T patients compared with RTX-N patients (p < 0.005) despite similar %CD19+ B-cells, and both groups had higher levels than HC (p < 0.05). RTX-N patients had lower % naïve B-cells (IgD + CD27-) and higher %IgD-CD27- B-cells compared with HCs (p = 0.01, 0.002, respectively). Following RTX treatment %CD38++/+++ cells (transitional IgD+ CD38++ (p < 0.05) and plasmablasts (IgD-CD38+++)) were significantly higher compared with HCs (p < 0.05), but with no significant difference between RTX-N and RTX-T patients in the frequency of plasmablasts. Moreover, there was a significant reduction in the frequency of BAFF-R + plasmablasts after RTX (p < 0.05). Positive correlations were found between serum BAFF and%transitional cells (r2 = 0.69, p < 0.0001) and plasmablasts (r2 = 0.61, p < 0.0001) in RTX-N patients, but not in RTX-T patients. In RTX-N patients, BAFF levels and %BAFF-R + B-cells were negatively correlated in total CD19+ B-cells (r2 = 0.49, p = 0.0003), in naive B-cells (r2 = 0.51, p = 0.0002); and in post-switched B-cells (IgD-CD27+) (r2 = 0.5, p = 0.0002). No such correlations were found after RTX. In RTX-N patients, there was a positive correlation between BAFF levels with total IgM and anti-dsDNA (r2 = 0.53, 0.48, respectively; p < 0.005 for both) (median 908 IU) but not in RTX-T patients (median anti-dsDNA 558 IU).

Conclusions These results suggest that (IgD + and IgD-) CD38++ B-cells were differently sensitive to BAFF-mediated effects in RTX-N patients, compared with RTX-T patients. RTX may therefore disrupt the aberrant interaction between BAFF and B-cells in RTX-N SLE patients. Serum anti-dsDNA antibodies and total IgM also correlated with BAFF levels in RTX-N patients but not in RTX-T. RTX may therefore act by removing autoantibodies/immune complexes, which may be responsible for altering the functional consequences of BAFF with both apparently disparate CD38++ B cell sub-populations.

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.