Background/Purpose: The efficacy of sirukumab, an anti-IL-6 cytokine antibody, was evaluated in multiple phase 3 studies in patients with rheumatoid arthritis (RA)  (SIRROUND -M, -D, -T, and -H) including the following populations: methotrexate (MTX)  inadequate responders (IR), conventional DMARD-IR, and TNF inhibitor (TNFi)-IR.  Treatment with sirukumab was hypothesized to normalize expression levels of serum analytes dysregulated at baseline in RA, with a dose response association.

Methods: Serum samples from the 4 studies were analyzed at baseline and Week 4 post-treatment using SomaLogic SOMAscan^TM (376 analytes) and validated antibody-based (Meso Scale Discovery, Luminex, and ELISA; 33 analytes) platforms.  For the SIRROUND -M, -D, -T, and -H studies, respectively, results are reported for patients  given sirukumab 100mg q2w (n=61, 205, 136, 106), sirukumab 50mg q4w (n=61, 203, 128, 115), or placebo (n=0, 122, 56, 0) and demographically-matched healthy controls samples (0, 50, 35, 145). Differences between groups were evaluated by comparing within-subject log₂ ratio of Week 4 over baseline values between treatment groups (adjusted for MTX dose; SIRROUND -D, -T) or versus 0-change (SIRROUND -M, -H), with FDR <0.05 and |fold/reference geometric means|>1.5 considered significant.

Results: At baseline, 24 analytes were significantly elevated and 6 were suppressed in RA patients compared to healthy controls (Table 1). Among the elevated analytes, 8 were suppressed by treatment with sirukumab 50mg q4w and 100 mg q2w across all 4 studies, consisting of acute phase proteins, S100A12, MMP-1, and surfactant protein D. Sixteen of the analytes elevated at baseline were not decreased by sirukumab, notably MMP-3 and lymphocyte chemoattractants CXCL10 and CXCL13, the latter 3 known to be decreased by TNF inhibitors.  Among analytes suppressed at baseline, none were increased by sirukumab.  Among analytes not significantly dysregulated in RA at baseline, 14 were significantly decreased by sirukumab, including complement components, acute phase proteins and neutrophil-associated proteins.  Changes in analyte levels were not significantly different between 100mg q2w and 50mg q4w sirukumab dose groups, with the exception of CRP, where decreases were significantly greater with the higher dose only in TNFi-IR patients .

Conclusion: Sirukumab consistently decreased 8 inflammatory proteins which were elevated at baseline across multiple RA populations, with no consistent dose response observed.  In addition to expected reductions in acute phase proteins with sirukumab, these included MMP-1 and SP-D. Several markers were not affected, including some lymphocyte chemoattractants that are known to be decreased by TNF inhibitors. The clinical implications of RA-associated analytes not normalized by sirukumab, as well as those decreased below normal levels, remain to be understood.