Why Inhibition of IL-23 Lacked Efficacy in Ankylosing Spondylitis

The term spondyloarthritis pertains to both axial and peripheral arthritis including ankylosing spondylitis (AS) and psoriatic arthritis (PsA), which is strongly linked to psoriasis and also the arthritis associated with inflammatory bowel disease. The argument supporting the role for IL-23 across the spectrum of SpA comes from 4 sources. First, genome wide associated studies (GWAS) have shown that all the aforementioned disorders exhibit IL-23R pathway SNPs, whereas HLA-B27 is not linked to all of these diseases-hence the IL-23 pathway represents the common genetic denominator. Secondly, experimental animal models have demonstrated a pivotal role for the IL-23/IL-17 axis in SpA related arthropathy that initially manifests as enthesitis, but also synovitis and axial inflammation and also associated aortic root and cutaneous inflammation. Thirdly, the emergent immunology of the human enthesis also supports the presence of IL-23 producing myeloid cells, not just at the enthesis but in other SpA associated sites including skin and gut. Finally, drugs that target the IL-23 pathway show excellent efficacy for skin disease, efficacy for IBD and also in peripheral arthropathy associated with SpA. The apparent failure of IL-23 blockade in the AS which is effectively a spinal polyenthesitis but evidence for efficacy of IL-23 inhibition for peripheral enthesitis in PsA and preliminary suggestions for benefit in axial PsA, raises many questions. Key amongst these is whether spinal inflammation may exhibit entheseal IL-17A production independent of IL-23 but peripheral enthesitis is largely dependent on IL-23 driven IL-17 production. Furthermore, IL-23 blocking strategies in animal models may prevent experimental SpA evolution but not prevent established disease, perhaps pointing towards a role for IL-23 in innate immune disease initiation whereas persistent disease is dependent on memory T-cell responses that drive IL-17A production independently of IL-23, but this needs further study. Furthermore, IL-12/23 posology in inflammatory bowel disease is substantially higher than that used in AS trials which merits consideration. Therefore, the IL-23 pathway is centrally involved in the SpA concept but the nuances and intricacies in axial inflammation that suggest non-response to IL-23 antagonism await formal definition. The absence of comparative immunology between the different skeletal sites renders explanations purely hypothetical at this juncture.