Early B Cell and Plasma Cell Kinetics Upon Treatment Initiation Portend Flares in Systemic Lupus Erythematosus: A Post-Hoc Analysis of Three Phase III Clinical Trials of Belimumab

OBJECTIVE: To investigate changes in B cell subsets in relation to disease flares upon initiation of standard therapy (ST) plus belimumab or placebo in patients with systemic lupus erythematosus (SLE). PATIENTS AND METHODS: Using data from the BLISS-76, BLISS-SC and BLISS Northeast Asia trials, we investigated associations of relative to baseline rapid (through week 8) and early (through week 24) changes in peripheral B cell subsets, anti-dsDNA and complement levels with the occurrence of disease flares from week 24 through week 52 (Mann-Whitney U tests) or the entire study follow-up (Cox regression analysis), assessed using the SELENA-SLEDAI Flare Index. RESULTS: Patients on ST alone who flared displayed less prominent early decreases in CD19(+)CD20(-)CD138(+) long-lived plasma cells (-16.1% versus -35.1%; P=0.012). In all arms combined, patients who developed severe flares showed less prominent early decreases in CD19(+)CD20(-)CD138(+) long-lived plasma cells (-23.5% versus -39.4%; P=0.028) and CD19(+)CD27(bright)CD38(bright) SLE-associated plasma cells (-19.0% versus -27.8%; P=0.045). After adjustment for rapid changes, early increases in overall CD19(+)CD20(+) B cells (HR: 1.81; 95% CI: 1.08–3.05; P=0.024) and early increases or no return after a rapid expansion in CD19(+)CD20(+)CD27(+) memory B cells (HR: 1.58; 95% CI: 1.18–2.11; P=0.002) portended subsequent severe flares. Patients who developed flares of any severity showed no or less prominent rapid (0.0% versus -12.5%; P<0.001) or early (-1.9% versus -21.7%; P<0.001) decreases in anti-dsDNA levels, and patients who developed severe flares showed no or less prominent early decreases in anti-dsDNA levels (0.0% versus -13.3%; P=0.020). Changes in complement levels exhibited no ability to distinguish flaring from non-flaring patients. CONCLUSIONS: Increase or lack of decrease in certain circulating B cell subsets or anti-dsDNA levels upon treatment initiation for active SLE heralded subsequent severe disease flares. A rapid expansion of memory B cells may signify sustained response to therapy when followed by a subsequent drop, while no return or delayed increases in memory B cells may portend flaring. Peripheral B cell and serological marker kinetics may help identify patients in whom therapeutic modifications could protect against flare development, and may hence prove a useful complement to traditional surveillance and early treatment evaluation in SLE.