Phase II Randomized Trial of Rituximab Plus Cyclophosphamide Followed by Belimumab for the Treatment of Lupus Nephritis

Abstract

Objective: To assess the safety, mechanism of action, and preliminary efficacy of rituximab followed by belimumab in the treatment of refractory lupus nephritis (LN).

Methods: In a multicenter, randomized, open-label clinical trial, 43 patients with recurrent or refractory LN were treated with rituximab, cyclophosphamide (CYC), and glucocorticoids followed by weekly belimumab infusions until week 48 (RCB group), or treated with rituximab and CYC but no belimumab infusions (RC group). Patients were followed up until week 96. Percentages of total and autoreactive B cell subsets in the patients' peripheral blood were analyzed by flow cytometry.

Results: Treatment with belimumab did not increase the incidence of adverse events in patients with refractory LN. At week 48, a complete or partial renal response occurred in 11 (52%) of 21 patients receiving belimumab, compared to 9 (41%) of 22 patients in the RC group who did not receive belimumab (P = 0.452). Lack of improvement in or worsening of LN was the major reason for treatment failure. B cell depletion occurred in both groups, but the percentage of B cells remained lower in those receiving belimumab (geometric mean number of B cells at week 60, 53 cells/μl in the RCB group versus 11 cells/μl in the RC group; P = 0.0012). Percentages of total and autoreactive transitional B cells increased from baseline to week 48 in both groups. However, percentages of total and autoreactive naive B cells decreased from baseline to week 48 in the belimumab group compared to the no belimumab RC group (P = 0.0349), a finding that is consistent with the observed impaired maturation of naive B cells and enhanced censoring of autoreactive B cells.

Conclusion: The addition of belimumab to a treatment regimen with rituximab and CYC was safe in patients with refractory LN. This regimen diminished maturation of transitional to naive B cells during B cell reconstitution, and enhanced the negative selection of autoreactive B cells. Clinical efficacy was not improved with rituximab and CYC in combination with belimumab when compared to a therapeutic strategy of B cell depletion alone in patients with LN.

Trial registration: ClinicalTrials.gov NCT02260934.

Figure 1

Flow diagram according to the Consolidated Standards of Reporting Trials (CONSORT) statement, showing the distribution of patients with recurrent or refractory lupus nephritis (LN) at each stage of the study from the time of informed consent to week 96. Reasons for exclusion of patients at each stage are provided. Samples from the per protocol population were evaluated at weeks 24, 48, and 96. RC = treatment with rituximab and cyclophosphamide but no belimumab infusions; RCB = treatment with rituximab, cyclophosphamide, and glucocorticoids followed by weekly belimumab infusions until week 48; SLE = systemic lupus erythematosus.

Figure 2

Total numbers of B cells within peripheral blood mononuclear cells (PBMCs) and relative frequencies of B cell subpopulations following treatment with RC versus RCB in samples from the per protocol population of patients with lupus nephritis. A, B cell counts before treatment and at week 12 (left) and during reconstitution in the peripheral blood at weeks 24–60 (right) following RC or RCB treatment. Each data point represents CD19+ B cell counts as determined by clinical laboratory testing in the peripheral blood from individual patients (number of PBMCs ranging 8–17). * = P < 0.05; ** = P < 0.01; *** = P < 0.001 by analysis of variance (ANOVA) on log values for comparisons at week 12, and by repeated‐measures ANOVA on log values (with baseline adjustment) for comparisons at weeks 24 through 60. Tukey‐Kramer post hoc adjustment was applied for multiple comparisons. B, Mean frequencies of each B cell subpopulation within total B cells from individual patients, including a per protocol sample analyzed at weeks 0 and 24 and per protocol sample analyzed at weeks 48 and 60, in each treatment group at each time point, as determined by flow cytometric analysis of cryopreserved PBMCs (number of PBMCs ranging 2–16). B cell subpopulation data were analyzed for subpopulations with >50 cells at each of the time points evaluated. * = P < 0.01 between treatment groups. For more details, see Supplementary Tables 2 and 3, available on the Arthritis & Rheumatology website at http://onlinelibrary.wiley.com/doi/10.1002/art.41466/abstract. Double neg = CD27−IgD− double‐negative (see Figure 1 for other definitions).

Figure 3

Reconstitution of autoreactive antinuclear antibody–positive (ANA+) B cell subsets following treatment with RC versus RCB in the per protocol population at week 48. A, Mean frequencies of B cell subpopulations within total ANA+ B cells from each group before treatment and at week 48 (number of peripheral blood mononuclear cells [PBMCs] ranging 8–13). * = P < 0.05; ** = P < 0.01 between treatment groups at week 48. B, ANA+ transitional and ANA+ naive B cells as a percentage of total B cells in the peripheral blood before treatment and at week 48. Each data point represents the relative frequency of ANA+ transitional B cells (left) or ANA+ naive B cells (right) in an individual patient at each time point (number of PBMCs ranging 8–13). * = P < 0.05 between treatment groups, using Fisher’s exact test. For more details, see Supplementary Tables 4–6, available on the Arthritis & Rheumatology website at http://onlinelibrary.wiley.com/doi/10.1002/art.41466/abstract. Double neg = CD27−IgD− double‐negative (see Figure 1 for other definitions).