Background Treatment options for lupus nephritis (LN) are limited by imperfect efficacy and multiple toxicities; new approaches are needed. Autoreactive B cells are attractive therapeutic targets given their pathogenic role in LN. However, a prospective randomised LN trial of B cell depletion with rituximab did not demonstrate improved efficacy. It is likely that the high levels of BAFF that exist following B cell depletion facilitate the maturation of autoreactive B cells. This autoreactivity may contribute to the limited clinical response. We have hypothesised that administration of an anti-BAFF reagent following B cell depletion will result in an enhanced, sustained clinical response.

Materials and methods CALIBRATE is a trial assessing the safety of belimumab following B cell depletion in patients with non-naïve LN (relapsed LN or LN that is resistant to therapy). Forty subjects will receive treatment with rituximab and cyclophosphamide; 20 will be randomised to receive belimumab for up to 1 year, 20 will receive no other therapy except for glucocorticoids. Efficacy and mechanistic outcomes at 1 and 2 years will also be evaluated. Given the ability of both rituximab and belimumab to depress serum IgG levels, levels of IgG are determined at screening in order to follow the combined effect of rituximab and belimumab on IgG and risk of infection during the CALIBRATE trial. We now present the characteristics of subjects screened for this study.

Results CALIBRATE has screened 27 patients. Demographic and clinical features at screening are shown in Table 1. Patients with a Up/c ratio >3 were 16 times (95% CI: 1.4, 767.3) more likely to have IgG levels ≤450 mg than patients with Up/c ≤3. Furthermore Up/c correlated inversely with serum IgG (r = −0.524, p = 0.0050).

Conclusions Physicians caring for lupus patients do not routinely measure IgG, even when considering initiation of rituximab or belimumab, therapies known to reduce serum IgG levels . We report that 33% of patients with non-naïve LN and >50% of patients with Up/c >3 are hypogammaglobulinemic. Transient hypogammaglobulinemia was previously demonstrated in almost 1/3 of LN subjects participating in a study of abatacept and cyclophosphamide (ACCESS study) and did not associate with serious infections. Proteinuria correlated inversely with serum IgG. Our preliminary findings demonstrate that a considerable number of patients with active, non-naïve LN are hypogammaglobulinemic, and confirm an inverse association between IgG levels with proteinuria. The CALIBRATE trial will follow levels of serum IgG and urinary protein prospectively and will monitor patients for the potential development of infectious events.

Acknowledgements This abstract is presented on behalf of the CALIBRATE study investigators. This study is being conducted by the Immune Tolerance Network. Research reported in this publication was supported by the National Institute of Allergy And Infectious Diseases of the National Institutes of Health under Award Number UM1AI109565. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Study medication and financial support was also provided by Genentech, Inc.

Trial Registration ClinicalTrials.gov Identifier: NCT02260934