B cell phenotypes in patients with rheumatoid arthritis relapsing after rituximab: expression of B cell-activating factor-binding receptors on B cell subsets

Abstract

Serum levels of B cell-activating factor (BAFF) rise following rituximab (RTX) therapy in patients with rheumatoid arthritis (RA). Initiation of naive B cell return to the periphery and autoreactive B cell expansion leading to relapse after RTX may therefore be linked to interactions between BAFF and BAFF-binding receptors (BBR). Relationships between serum BAFF and BBR expression [(BAFFR, calcium signal modulating cyclophilic ligand interactor (TACI) and B cell maturation antigen (BCMA)] were determined on B cell subsets, defined using immunoglobulin (Ig)D/CD38. Twenty pre-RTX and 18 RA patients relapsing after B cell depletion were included. Results were analysed with respect to timing of relapse up to 7 months after peripheral B cell return (≥ 5 B cells/μl) and to serum BAFF levels. After B cell return, B cell populations from relapsing patients had significantly lower BAFFR⁺ expression compared to HC and pre-RTX patients. The percentage of BAFFR⁺ B cells increased with time after B cell return and was correlated inversely with serum BAFF levels. BAFFR expression remained reduced. The percentage of TACI⁺ memory B cells were lower in RA patients after RTX compared with healthy controls (HC). BCMA expression (% and expression) did not differ between patients and HC. Relapse following B cell return appeared largely independent of the percentage of BAFFR⁺ or percentage of BCMA⁺ B cells or serum BAFF levels. The lower percentage of TACI⁺ memory B cells may reduce inhibitory signalling for B cell differentiation. In patients relapsing at longer periods after B cell return, recovery of the B cell pool was more complete, suggesting that selection or expansion of autoreactive B cells may be needed to precipitate relapse.

Keywords: B lymphocytes; rheumatoid arthritis; rituximab.

Figure 1

Example of a flow cytometry plot showing B cell phenotypes defined on the basis of immunoglobulin (Ig)D/CD38 expression for CD19‐gated cells in a rheumatoid arthritis (RA) patient pre‐rituximab (RTX) (a), at first documented B cell return (b), 3 months post‐B cell return (c) and (d) undergoing clinical relapse 6 months post‐B cell return. B cell subpopulations are indicated as follows: transitional naive, mature naive, resting memory (IgD‐RM), IgD⁺ resting memory (IgD ⁺ RM), post‐germinal centre (post‐GC) and plasmablasts. In panels (e) and (f) results of the cross‐sectional study of B cell phenotypes of 20 RA patients studied pre‐RTX and 18 relapsing post‐RTX, either within 0–3 months first documented B cell return or ≥ 4 months later, are shown. In panel (e) The relationship between percentage of CD19⁺ B cells within each subpopulation as defined by IgD/CD38 and time (months) to relapse after the first documented sign of B cell return (> 5 CD19⁺ B cells/μl) after RTX. Linear regression was used to calculate Spearman's rank coefficients, which are shown for those with r ² > 0·4 and giving P ≤ 0·05. (f) absolute numbers CD19⁺ B cells from patients post‐RTX are shown. Statistical analysis compared values in post‐RTX patients relapsing close to (0–3 months) or ≥ 4 months after B cell return. Horizontal lines on graphs represent median values and results were analysed using the Mann–Whitney U‐test; *P < 0·05; **P < 0·01; ***P < 0·001. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 2

(a) Flow cytometry plot of B cell activating factor receptor (BAFFR) expression on CD19⁺ B cells within the lymphocyte gate in an RA patient studied pre‐rituximab (RTX) and (b) at B cell return (which occurred 6 months after RTX), (c) at 3 months after B cell return. The associated histograms for (a–c) show relative expression of BAFFR within the gated CD19⁺ population. The patient relapsed 3 months after the sample shown in (c). (d) (1–6): percentage of BAFFR⁺ expression on CD19⁺ B cell subpopulations (defined using immunoglobulin (Ig)D/CD38) in healthy controls (HC) and in pre‐ and post‐RTX rheumatoid arthritis (RA) patients. Results for post‐RTX patients were divided on the basis of proximity to first documented B cell return, as described in Fig. 1. Lines on graphs represent median values and results were analysed using the Mann–Whitney U‐test; *P < 0·05; **P < 0·01; ***P < 0·001. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 3

Flow cytometry plot of transmembrane activator and calcium signal modulating cyclophilic ligand interactor (TACI) expression on CD19⁺ B cells within the lymphocyte gate in the rheumatoid arthritis (RA) patient studied pre‐rituximab (RTX) (a) and (b) at B cell return (which occurred 6 months after RTX), and (c) at 3 months after B cell return. The associated histograms for (a–c) show relative expression of TACI within the gated CD19⁺ population. The patient relapsed 3 months after the sample shown in (c). (d) (1–6): percentage of TACI⁺ expression on CD19⁺ B cell subpopulations (defined using immunoglobulin (Ig)D/CD38) as in healthy controls (HC), pre‐RTX patients and patients relapsing post‐B cell return. Results for HC and pre‐RTX patients were compared and results between the two patient groups relapsing at different intervals after B cell return were also compared. Lines on graphs represent median values and results were analysed using the Mann–Whitney U‐test; *P < 0·05; **P < 0·01; ***P < 0·001. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 4

(a) Flow cytometry plot of B cell maturation antigen (BCMA) expression on CD19⁺ B cells within the lymphocyte gate in a rheumatoid arthritis (RA) patient studied pre‐rituximab (RTX) (b) at B cell return (which occurred 6 months after RTX) and (c) at 3 months after B cell return. The associated histograms for (a–c) show relative expression of BCMA within the gated CD19⁺ population. The patient relapsed 3 months after the sample shown in c). In (d), (1–6) percentage of BCMA⁺ expression on CD19⁺ B cell subpopulations [defined using immunoglobulin (Ig)D/CD38) in healthy controls (HC), pre‐RTX patients and patients relapsing after RTX. Results for HC and pre‐RTX patients were compared and results between the two patient groups relapsing at different intervals after B cell return were also compared. Lines on graphs represent median values and results were analysed using the Mann–Whitney U‐test; *P < 0·05; **P < 0·01; ***P < 0·001. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 5

(a) Serum B cell activating factor (BAFF) levels in 5 HC, pre‐RTX patients (n = 20) and patients relapsing after RTX (n = 18). Dotted line indicates upper limit for normal control values as defined by the manufacturers. (b) Relationship between serum BAFF levels and time after B cell return in rheumatoid arthritis (RA) patients relapsing after RTX. (c) The relationship between percentage of BAFFR⁺ B cells and serum BAFF levels are shown for each B cell subpopulation, defined using immunoglobulin (Ig)D/CD38. Linear regression analysis (Spearman's correlation) and exact P‐values are shown for significant correlations (P < 0·05) between serum BAFF levels and percentage of BAFF receptor (BAFFR)⁺ B cells.

Figure 6

In (a) the relative expression [mean fluorescence intensity (MFI)] of B cell activating factor receptor (BAFFR) and in (b) expression of transmembrane activator and calcium signal modulating cyclophilic ligand interactor (TACI) expression on B cells was measured within each B cell subpopulation, as defined using immunoglobulin (Ig)D/CD38, and plotted against time (months) after peripheral B cell return in rheumatoid arthritis (RA) patients relapsing after rituximab (RTX). Linear regression analysis (Spearman's correlation) and exact P‐values are shown for significant correlations (P < 0·05) Shaded areas indicate the ranges for MFI of the two BAFF‐binding receptors (BBR)⁺ given by the respective B cell subpopulations in five healthy control (HC) samples taken within the same time‐period.