Background/Purpose: ABT-122 is a dual variable domain (DVD-Ig) biologic which inhibits TNF-α and IL17A. In a 12-wk phase 2 study (NCT02141997) in MTX-IR patients (pts) with RA, ABT-122 120 mg administered either every other week (eow) or every week (ew) (but not 60 mg eow) had efficacy at least comparable to adalimumab (ADA) 40 mg eow. However, dual inhibition was not significantly differentiated from TNF-α inhibition alone. Biomarker investigations were designed to assess the on-target effects of ABT-122 and the contribution of the anti-TNF-α and anti-IL17A components to pts ACR20 response at wk-12 in this trial.

Methods: Consenting pts provided whole blood and Paxgene blood tube samples (baseline [BL], wk-4) and serum samples (BL, wk-4, wk-12). Gene expression (Affymetrix GeneChip Human Gene 2.0 ST arrays), DNA methylation (Illumina Infinium Human Methylation 450K BeadChips), and multiplex serum protein (Milliplex or Myriad-RBM) analyses were conducted in 132 ACR20, 96 ACR20, and 30 ACR20 responders, respectively. Cell populations were estimated from DNA methylation profiles in whole blood (Genome Biol. 15:R31). Wk-4 changes in gene expression and DNA methylation were correlated with wk-12 continuous DAS28 change. Ingenuity Pathway Analysis software was used for pathway analysis. Repeated measures analysis of log-transformed serum protein changes used linear mixed models.

Results: Wk-4 changes in expression of genes associated with IL17A signaling (eg, IL23, MMP9, S100A8, S100A9, LCN2) indicated inhibition of IL17A by ABT-122. Serum protein changes (S100A8/A9, LCN2) were seen at wk-4 and wk-12 of ABT-122 treatment. Expression of 3756 genes was changed significantly at wk-4 in responders. Few genes (eg, MMP9, NAMPT; associated mostly with IL17A signaling) showed statistically significant differences in the ABT-122 arms vs the ADA arm. Differentially expressed genes plotted for ABT-122 responders were highly concordant with those from ADA responders. Pathway analysis of the nearest neighbor genes to these CpG sites indicated that ABT-122 inhibited IL17A pathways. Upstream driver pathway analysis of nearest neighbor genes and DAS28 found strong correlation with TNF-α responsive pathways (second highest rank) and weak correlations with IL17A- and IL17 receptor-responsive pathways (223rd and 1691st ranks, respectively). ABT-122 and ADA drove similarly robust changes in several protein biomarkers (eg, MMP3, CXCL13, CXCL9, IL6, S100A8/A9), suggesting TNF-α–driven effects. Methylation signatures in whole blood indicated significant changes in CD4 T-cell and granulocyte populations in ABT-122 vs ADA responders, indicating ABT-122–mediated inhibition of IL17A. Methylation of 6190 CpG sites was significantly correlated with DAS28 in responders at wk-4 (false detection rate <5%).

Conclusion: ABT-122 successfully engaged TNF-α  and IL17A qualitatively, compared with adalimumab. However, correlation of wk-4 changes in gene expression and DNA methylation with wk-12 DAS28 response indicated that response to ABT-122 in pts with MTX-IR RA was primarily driven by TNF-α inhibition, with little added contribution from IL17A inhibition.