Post Hoc Analyses of Biomarkers Predictive of a Renal Response to Intravenous (IV) Belimumab (BEL) Plus Standard Therapy (ST) in Patients with Lupus Nephritis (LN)

Le résumé fourni par la source

Background: In BLISS-LN (GSK BEL114054), a Phase 3 trial of adults with LN treated with BEL 10 mg/kg/month IV or placebo (PBO), + ST, BEL resulted in significantly more kidney responses (Primary Efficacy Renal Response [PERR] and Complete Renal Response [CRR]), and greater improvements in anti-dsDNA and anti-C1q autoantibodies and complement C3 and C4 levels than PBO over 104 weeks (wks). This post hoc analysis of BLISS-LN assessed the characteristics of patients with LN most likely to benefit from the addition of BEL to ST. Methods: Post hoc logistic regression models tested the ability of potential baseline (BL) biomarker levels and biomarker level changes in the first 12 months of treatment to predict PERR and/or CRR at Wk 104. Results: Overall, 446 patients were evaluated. High BL anti-C1q predicted a PERR with BEL and a CRR with PBO (Table). High BL levels of immunoglobulin (Ig)A and a decrease in IgA at Wk 8 with BEL were predictive of PERR, and high BL levels of naïve B cells were predictive of PERR and CRR to BEL. Low BL plasmablasts, and a decrease in IgA or IgM at Wk 24 predicted a PERR in both groups and a decrease in IgA at Wk 24 predicted CRR to BEL. A reduction in anti-C1q at Wk 52 and of uPCR at Wk 12 also predicted PERR and CRR in both groups. Changes in B-cell subsets were not predictive of a response in the BEL group. Conclusions: High levels of IgA and naïve B cells at BEL initiation, as well as early reductions in IgA, predicted attainment of PERR and/or CRR for BEL but not PBO. Early reductions in uPCR predicted responses regardless of whether patients received BEL or PBO. Funding: Commercial Support - GSK (GSK BEL114054; NCT01639339) Table. - Significant BL biomarker and percentage change from BL to stated wks biomarker prediction of kidney response at Wk 104 (treatment policy*; post hoc analysis) Predictor PERR at Wk 104 CRR at Wk 104 PBO (n=223) BEL (n=223) PBO (n=223) BEL (n=223) IgA (g/l) at BL Parameter estimate† - 0.3341 - - Odds ratio‡ ns 1.40, p=0.0173 ns ns IgM (g/l) at BL Parameter estimate† 0.6674 - - - Odds ratio‡ 1.95, p=0.0120 ns ns ns Anti-C1q (U/ml) at BL Parameter estimate† - 0.002391 0.001615 - Odds ratio‡ ns 1.0024, p=0.0148 1.0016, p=0.0192 ns Naive B cells (count/μl) at BL Parameter estimate† 0.001730 0.001688 Odds ratio‡ ns 1.0017, p=0.0433 ns 1.0017, p=0.0446 Plasmablasts (count/ml) at BL Parameter estimate† -0.000166 -0.000111 -0.000193 - Odds ratio‡ 0.9998, p=0.0124 0.9999, p=0.0398 0.9998, p=0.0268 ns Change in IgA (g/l) at 8 wks Parameter estimate† - -0.0199 - - Odds ratio‡ ns 0.98, p=0.0482 ns ns Change in IgA (g/l) at 24 wks Parameter estimate† -0.0155 -0.0332 - -0.0171 Odds ratio‡ 0.98, p=0.0471 0.97, p=0.0003 ns 0.98, p=0.0456 Change in IgM (g/l) at 24 wks Parameter estimate† -0.0132 -0.0152 - - Odds ratio‡ 0.99, p=0.0174 0.98, p=0.0171 ns ns Change in anti-dsDNA (IU/ml) at 24 wks Parameter estimate† -0.004507 - - - Odds ratio‡ 0.9955, p=0.0204 ns ns ns Change in anti-C1q (U/ml) at 24 wks Parameter estimate† -0.018157 - -0.011071 - Odds ratio‡ 0.9820, p=0.0002 ns 0.9890, p=0.0190 ns Change in anti-C1q (U/ml) at 52 wks Parameter estimate† -0.012990 -0.016034 -0.011426 -0.013112 Odds ratio‡ 0.9871, p=0.0020 0.9841, p=0.0055 0.9886, p=0.0190 0.9870, p=0.0219 Change in memory B cells (count/μl) at 24 wks Parameter estimate† 0.004601 - - - Odds ratio‡ 1.0046, p=0.0466 ns ns ns Change in uPCR (g/g) at 12 wks Parameter estimate† -0.5611 -0.3763 -0.8469 -0.3696 Odds ratio‡ 0.57, p<0.0001 0.69, p=0.0018 0.43, p<0.0001 0.69, p=0.0094 *Treatment policy population included all available data, ignoring whether the patient had discontinued study treatment or started a prohibited medication;†estimates are from a post hoc logistic regression model for the parameter with covariates induction regimen (CYC vs MMF), race (Black African ancestry vs other), baseline uPCR, and baseline eGFR. Positive parameter estimates indicate that higher BL values or an increase from BL to stated wks lead to better response. Conversely, negative parameter estimates indicate that lower BL values or a decrease from BL to stated wks lead to better response. Parameter estimates with non-significant odds ratios are not displayed;‡odds ratios are for a one unit change in the lab parameter, and due to this small scale, the odds ratios appear close to 1. Odds ratios that were not statistically significant predictors of response are not displaved.PERR at Wk 104 (Wk 100. confirmed at Wk 104) = uPCR ≤0.7, eGFR no more than 20% below pre-flare value or ≥60 ml/min/1.73 m2, no rescue therapy.CRR at Wk 104 (Wk 100, confirmed at Wk 104) = uPCR <0.5, eGFR no more than 10% below pre-flare value or ≥90 ml/min/1.73 m2, no rescue therapy.BEL, belimumab; BL, baseline; CYC, cyclophosphamide; MMF, mycophenolate mofetil; CRR, Complete Renal Response; eGFR, estimated glomerular filtration rate; Ig, Immunoglobulin; PBO, placebo; PERR, Primary Efficacy Renal Response, ns, not significant; uPCR, urine protein/creatinine ratio.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.