BASELINE LABORATORY CHARACTERISTICS FROM THE COMBINED PLACEBO GROUPS IN THE BLISS TRIALS ARE PREDICTIVE OF SEVERE FLARE AT 24 WEEKS

Background: Identification and validation of potential serologic biomarkers of clinically meaningful SLE flare are major unmet medical needs in clinical practice and trials. Objectives: To examine biomarker predictors of SLE flares at baseline in the combined placebo groups from the belimumab BLISS-52 (NCT00424476) and BLISS-76 (NCT00410384) phase 3 trials. Methods: The BLISS trials enrolled patients with active SLE (SELENA SLEDAI ≥6) who were autoantibody positive (antinuclear antibody [ANA] or anti-double-stranded DNA [anti-dsDNA]), on stable standard therapy for 30 d, and without severe active lupus nephritis or CNS SLE. Patients were randomized to placebo, or belimumab 1 or 10 mg/kg, plus standard therapy. Changes in prednisone were allowed early in the trials, but had to be within 25% or 5 mg of baseline dose. Flare (moderate-severe) was defined as 1 new BILAG A or 2 new B scores, or by modified SLE Flare Index (SFI). Laboratory biomarkers evaluated included ANA, anti-dsDNA, anti-Smith, complement (C) 3, C4, C-reactive protein (CRP), proteinuria >0.5 g/24 h, immunoglobulin, B- and T-cell subsets, and B-lymphocyte stimulator (BLyS) quartile levels. The proportion of placebo patients with abnormal tests who developed a new flare was compared with the proportion without a flare. Meaningful difference was defined as: ≥10% absolute difference (% flare - % no flare) or nominal p values for pairwise comparison (likelihood ratio test); and ≥50% increase ([% flare - % no flare]/% no flare). Results: By wk 24 in the placebo group (n=562), 25.3% of patients (n=142) had a flare by BILAG and 12.8% (n=72) by SFI. Predictors of wk-24 flare (% with vs without flare) are shown in the table. Table 1. Univariate Analysis of Predictors of 24-Wk Flare % BILAG A/2B Flare% SFI Severe Flare WithWithoutWithWithout (n=142)(n=420)(n=72)(n=490) Low C359.239.066.740.8 Low C464.850.263.952.4 CRP47.134.451.435.6 Proteinuria >0.5g/24 h33.117.143.117.9 Positive anti-dsDNA (>200 IU/mL)47.928.359.729.4 Positive anti-Smith39.028.446.528.9 BLyS ≥2 ng/mL36.421.354.220.7 p<0.001; p<0.01; p<0.05. Conclusions: SLE patients on standard therapy with low C3/4, positive CRP, proteinuria >0.5 g/24 h, positive anti-dsDNA or anti-Smith, or BLyS ≥2 ng/mL were at increased risk for SLE flare over 24 wk. Flare definitions (BILAG and SFI) had high concordance regarding predictors; univariate risk factors were the same for both definitions at 24 wk. Study limitations: laboratory tests were defined as categorical variables (present/absent) and time-varying nature of the characteristics over the 24 wk was not examined. The data suggest serologic tests predict clinically meaningful flare over 24 wk, and may be useful in the identification of patients at greater risk for flares in clinical practice and trials. Disclosure of Interest: M. Petri Consultant for: GlaxoSmithKline/Human Genome Sciences, R. van Vollenhoven Grant/Research support from: GlaxoSmithKline/Human Genome Sciences, R. Levy Speakers Bureau: GlaxoSmithKline/Human Genome Sciences, S. Navarra Speakers Bureau: GlaxoSmithKline/Human Genome Sciences, R. Cervera Grant/Research support from: GlaxoSmithKline/Human Genome Sciences, Consultant for: GlaxoSmithKline/Human Genome Sciences, Z. Zhong Shareholder of: Human Genome Sciences, Employee of: Human Genome Sciences, W. Freimuth Shareholder of: Human Genome Sciences, Employee of: Human Genome Sciences, J. Buyon Grant/Research support from: Human Genome Sciences, Consultant for: GlaxoSmithKline/Human Genome SciencesCitation: Annals of the Rheumatic Diseases, volume 71, supplement 3, year 2012, page 538Session: SLE, Sjögren's and APS – treatment (Poster Presentations )