A randomized, phase II study of sequential belimumab and rituximab in primary Sjögren's syndrome

Abstract

BACKGROUNDPrimary Sjögren's syndrome (pSS) is characterized by B cell hyperactivity and elevated B-lymphocyte stimulator (BLyS). Anti-BLyS treatment (e.g., belimumab) increases peripheral memory B cells; decreases naive, activated, and plasma B cell subsets; and increases stringency on B cell selection during reconstitution. Anti-CD20 therapeutics (e.g., rituximab) bind and deplete CD20-expressing B cells in circulation but are less effective in depleting tissue-resident CD20+ B cells. Combined, these 2 mechanisms may achieve synergistic effects.METHODSThis 68-week, phase II, double-blind study (GSK study 201842) randomized 86 adult patients with active pSS to 1 of 4 arms: placebo, s.c. belimumab, i.v. rituximab, or sequential belimumab + rituximab.RESULTSOverall, 60 patients completed treatment and follow-up until week 68. The incidence of adverse events (AEs) and drug-related AEs was similar across groups. Infections/infestations were the most common AEs, and no serious infections of special interest occurred. Near-complete depletion of minor salivary gland CD20+ B cells and a greater and more sustained depletion of peripheral CD19+ B cells were observed with belimumab + rituximab versus monotherapies. With belimumab + rituximab, reconstitution of peripheral B cells occurred, but it was delayed compared with rituximab. At week 68, mean (± standard error) total EULAR Sjögren's syndrome disease activity index scores decreased from 11.0 (1.17) at baseline to 5.0 (1.27) for belimumab + rituximab and 10.4 (1.36) to 8.6 (1.57) for placebo.CONCLUSIONThe safety profile of belimumab + rituximab in pSS was consistent with the monotherapies. Belimumab + rituximab induced enhanced salivary gland B cell depletion relative to the monotherapies, potentially leading to improved clinical outcomes.TRIAL REGISTRATIONClinicalTrials.gov NCT02631538.FUNDINGFunding was provided by GSK.

Keywords: Autoimmune diseases; Clinical Trials; Drug therapy; Immunology.

Figure 1. Study design.

LLN, lower limit of normal.

Figure 2. Patient flow through the study.

*Patients may only have one primary reason;.†Completer population (patients who completed the 52-week treatment and general follow-up period of the study, including the visit at week 68). AE, adverse event.

Figure 3. Median (IQR).

(A–D) Total B cells (CD19⁺), memory B cells (CD20⁺CD27⁺), naive B cells (CD20⁺CD27^–), and plasmablasts (CD27⁺CD38⁺CD19⁺) over time by flow cytometry (completer population, n = 60). Flow cytometry data were analyzed using the Hodges-Lehmann method to provide a nonparametric 95% CI for the treatment comparisons of interest. For clear presentation of results, data in A–C are presented as cells/μL (with different y axes maximum values), and data in D are presented as cells/mL. ^†N = 7 at weeks 4, 12, and 52. ^‡N = 16 at weeks 1, 36, 44, 68. ^§N = 17 at weeks 1, 44, and 68. N = 18 at weeks 4, 12, 28, and 36. N = 16 at week 52. ^¶N = 13 at week 24. **N = 7 at weeks 4, 8, and 12. N = 6 at week 52. ^††N = 16 at weeks 1, 8, 44, and 68. N = 15 at week 36. ^‡‡N = 17 at weeks 1, 44, and 68. N = 18 at weeks 4, 8, 28, and 36. N = 16 at week 12. N = 15 at week 52. ^§§N = 13 at week 24. N = 14 at week 36. ^¶¶N = 16 at weeks 1, 8, 28, 44, and 68. N = 15 at weeks 24 and 36. IQR, interquartile range.

Figure 4. Key serological biomarkers over time.

(A–C) Free BLyS, CXCL13, IgG, and RF (patients positive at baseline) (completer population, n = 60). *N = 16 at week 36. ^†N = 18 at week 28. ^‡N = 16 at weeks 36 and 44. ^§N = 15 at week 24. ^¶N = 7 at weeks 4 and 48. **N = 16 at week 8. ^††N = 17 at week 20. N = 18 at week 36. ^‡‡N = 15 at weeks 10, 24, 40, and 48. N = 14 at week 32. ^§§N = 6 at weeks 0, 36, and 52. N = 7 at week 12. N = 5 at week 24. ^¶¶N = 7 at week 0. N = 6 at week 12. N = 5 at weeks 24, 36, and 52. ***N = 14 at week 0. N = 12 at week 12. N = 10 at weeks 24 and 36. N = 9 at week 52. ^†††N = 11 at week 0. N = 9 at weeks 12, 24, and 36. N = 10 at week 52. BLyS, B-lymphocyte stimulator; CXCL13, chemokine (C-X-C motif) ligand 13; IgG, immunoglobulin G; RF, rheumatoid factor.

Figure 5. Absolute CD20⁺ B cells in the salivary gland and representative immunofluorescence images.

(A and B) Absolute CD20⁺ B cells in the salivary gland and representative immunofluorescence (Hoechst CD20) histological images. (completer population, n = 60). Median (IQR): includes all baseline/week 24 completer data. Only data for patients with paired baseline/week 24 biopsies. Minimum values = 0.1. When CD20⁺ B cells were undetectable, values were input as 0.1 to allow logarithmic display. Changes in absolute CD20⁺ B cells in the salivary gland were analyzed using the Hodges-Lehmann method to provide a 95% CI for treatment comparisons of interest. For the histological images, original slides were imaged at 20× using a Zeiss Axio Scan Z1 slide scanner and are included in Supplemental Figure 3. IQR, interquartile range. Scale bars: 100 µm.

Figure 6. Clinical efficacy over time as measured by mean (standard error).

(A–D) ESSDAI total score, unstimulated salivary flow, stimulated salivary flow, and ESSPRI total score (completer population, n = 60). *N = 15 at week 12. ^†N = 15 at week 36. ^‡N = 15 at weeks 36 and 68. ^§N = 16 at week 52. ESSDAI, EULAR Sjögren’s syndrome disease activity index; ESSPRI, EULAR Sjögren’s Syndrome Patient Reported Index.

Figure 7. Proportion of responders with an ESSDAI reduction.

(A and B) Proportion of responders with an ESSDAI reduction of ≥3 points and ≥5 points versus baseline (completer population, n = 60). ESSDAI responder analyses utilized a generalized estimating equation model. ESSDAI, EULAR Sjögren’s syndrome disease activity index.