The interleukin (IL)‐23/IL‐17 axis in ankylosing spondylitis: new advances and potentials for treatment

Summary

Ankylosing spondylitis (AS), the most common form of spondyloarthropathy, is a chronic, progressive multi‐system inflammatory disorder characteristically affecting the sacroiliac joints and axial skeleton. Although the exact mechanisms underlying the pathogenesis of AS remain to be elucidated, the presence of human leucocyte antigen (HLA)‐B27 is known to markedly increase its risk of development. Current treatments include non‐steroidal anti‐inflammatory drugs (NSAIDs) and tumour necrosis factor (TNF) blockers. In recent years, the interleukin (IL)‐23/IL‐17 pathway has been shown to have significance in the pathogenesis of AS and treatment modalities targeting this pathway have been shown to be beneficial in various other inflammatory conditions. This review provides an overview of the IL‐23/IL‐17 pathway in the pathogenesis of AS and summarizes new potential treatments for AS and related inflammatory diseases.

Ankylosing spondylitis (AS) and other immune mediated inflammatory diseases

Ankylosing spondylitis (AS) is a common chronic, progressive inflammatory rheumatic disease characterized by sacroileitis and axial inflammation. AS can be thought of as an immune‐mediated inflammatory disorder (IMID). IMIDs encompass a wide range of diseases, including rheumatoid arthritis (RA), psoriasis, the inflammatory bowel diseases (IBD) Crohn's disease and ulcerative colitis, type 1 diabetes mellitus (T1DM) and multiple sclerosis (MS). All IMID share features of chronic inflammation; some (T1DM, MS, RA) have clear autoimmune features following failure of tolerance or immune regulation, whereas others, such as IBD and AS, are more ‘autoinflammatory’ in nature 2.

During the past three decades, advances in molecular and immunological research have led to an increase in our understanding of these diseases and have repeatedly highlighted a key role for cytokine dysregulation in their pathophysiology. Although these molecules are known to be imperative facilitators of normal immune function, imbalances in cytokine levels can lead to both acute and chronic inflammatory disorders; however, the latter appear to predominate. An explanation as to why these disorders tend to be persistent in nature remains to be elucidated, but genetic predisposition, cytokine cascades, enduring initiating triggers and failure of inflammation resolution probably all contribute to amplification of immune responses.

Genetic factors have been shown to be vital determinants of susceptibility and, interestingly, significant numbers of these patients have several (co‐existent) IMID. AS has the greatest heritability of the IMID, estimated from twin studies at greater than 90% 3. Furthermore, recent meta‐analysis of genome‐wide association studies (GWAS) data has shown that AS shares multiple genetic associations with IBD and psoriasis 7. Possession of the class 1 human leucocyte antigen (HLA)‐B27 confers approximately 30% of the total genetic risk to AS, with polymorphisms of endoplasmic reticulum aminopeptidase 1 (ERAP1) (an enzyme that trims peptides for antigen presentation), interleukin (IL)‐23 receptor (IL‐23R) and other T helper type 17 (Th17)/23 pathway genes being the next most common 4. Multiple genetic polymorphisms within Th17/23 pathway genes have been found in common between AS, IBD and psoriasis, suggesting a common pathogenic role. More recently, research has shown that additional environmental factors (such as microbiomes, infections and drug and toxin‐exposure) also play a significant role in AS pathogenesis 5.

In parallel with these genetic advances, immunological studies have implicated Th17/23 responses in AS (and IBD and psoriasis). More recently, therapeutic strategies targeting the Th17 response have proved beneficial in AS 6. This review will summarize the biology of the Th17/23 axis and its role in AS and related IMID, before describing current and future therapeutic strategies targeting these responses in AS.

Th17 responses and the IL‐17/23 axis

Th17 cells are a subset of T helper cells, developmentally distinct from Th1 and Th2 cells, which produce IL‐17 together with other proinflammatory cytokines, such as IL‐6, IL‐22, IL‐26, interferon (IFN)‐γ and tumour necrosis factor (TNF)‐α 7. Of these, IL‐17 subsets A and F and IL‐26 are considered to be the most specific to the Th17 response 8.

IL‐17 has been shown to enhance T cell priming and to stimulate various cell types, including fibroblasts, endothelial cells, macrophages and epithelial cells, to produce proinflammatory mediators (such as IL‐1, IL‐6, TNF‐α and chemokines). IL‐17 has been shown to function principally during the effector phase of an inflammatory response 9. Interestingly, the development of murine Th17 cells from native T cells has been shown to be inhibited by IFN‐γ and IL‐27 (Th1 response cytokines), and by IL‐4 and IL‐25 (also known as IL‐17E; associated with Th2 responses) 8, confirming its distinction from the other T cell subsets. IL‐22 and granulocyte–macrophage colony‐stimulating factor (GM‐CSF) can be produced by Th17 cells or by related but probable Th populations 10.

Research into the triggers and drivers of Th17 responses has identified a major role for cytokine IL‐23, which was first described by Oppmann et al., and was found to be expressed in activated murine and human monocytes, macrophages, dendritic cells, T and B cells and endothelial cells 11. Aggarwal et al. found that activation of T cells in the presence of IL‐23 led to expansion of Th17 cells, thereby leading to an increase in IL‐17 levels (Fig. 1) 12. The work by Sherlock and Cua has shown that over‐expression of IL‐23 is sufficient to induce an inflammatory disease with enthesitis in mice, which is similar to human AS 13. Interestingly, IL‐23 has also been shown to be capable of inducing the secretion of IL‐17 by non‐T cells in an inflammatory environment and that, although IL‐23 appears to be crucial in localized tissue inflammation, it is not required for systemic inflammatory responses 14.

Figure 1.

Interleukin (IL)‐17/IL‐23 pathway.

IL‐23 is a member of the IL‐12 cytokine family and, in addition to the IL‐12 related p40 subunit, it also has a unique p19 chain 12. It is produced mainly by activated myeloid cells, endothelial and epithelial cells, and mediates signalling via its heterodimeric IL‐23 receptor complex, consisting of a unique IL‐23 receptor (IL‐23R) subunit paired with a common IL‐12Rβ1 subunit shared with the IL‐12 receptor 15. Both the IL‐12Rβ1 and IL‐23R chain lack intrinsic signalling activity and are associated with intracellular proteins to induce downstream signalling 8.

IL‐12 has been shown to have importance in the development of Th1 cells; IL‐23, however, despite having a similar molecular structure to the rest of the family, has distinct effects 6. Cua et al. undertook a study of mice deficient of the IL‐23p19 subunit, and found that upon immunization they developed normal Th1 responses but were deficient in their production of IL‐17‐producing cells, indicating that IL‐23, rather than IL‐12, induced Th17 cell development; this was later confirmed when exogenous administration of IL‐23 led to amplification in Th17 cell quantity 16. Veldhoen et al. suggested later that IL‐23 probably acts upon previously differentiated Th17 cells to induce amplification and stabilization, rather than leading to the differentiation of Th17 itself 17.

Functional analysis of IL‐17 has revealed its unique role in stimulating epithelial defence against Gram‐negative bacteria and fungi; for example, Ye et al. found that mice deficient of the IL‐17 receptor were significantly more susceptible to infections caused by Klebsiella and Candida 18. Furthermore, particular organisms (for example, Klebsiella, Mycobacterium tuberculosis and fungi) have been shown to stimulate Th17 responses 6, 19.

Thus, Th17 responses are characterized by production of IL‐17 (as well as other proinflammatory cytokines) and reliance to a significant extent on IL‐23 for expansion and maintenance of their phenotype, giving rise to the concept of the IL‐17/23 axis.

Th17 responses in immune‐mediated inflammatory diseases

A growing body of evidence supports the role of Th17 in multiple human inflammatory diseases. Thus, while early studies suggested that the development of inflammatory diseases are related to Th1 responses, as patients with these conditions demonstrate elevated levels of IL‐12 and IFN‐γ, emerging evidence from studies on numerous inflammatory conditions (such as multiple sclerosis 20, 21, 22, inflammatory bowel disease 22, psoriasis 23, 24, 25 and inflammatory arthritis 26, 27, 28, 29, 30, 31) suggests the significance of alternative cytokine pathways – in particular the Th17 response. Some of the confusion arose because the key Th1 response cytokine IL‐12 and the key Th17‐promoting cytokine IL‐23 share a common p40 subunit 12.

Pathogenesis of ankylosing spondylitis

Ankylosing spondylitis (AS) is a form of spondyloarthritis causing chronic, progressive inflammatory disease characterized by sacroileitis and axial inflammation, with potential to cause bone erosion, new bone formation and ankylosis of the spine 32. The burden of the disease is dependent upon the degree of acute inflammation causing pain and stiffness and on new bone formation causing a reduction in spinal mobility 32, and extra‐articular features involving the eyes, heart, lungs and gut may be seen.

Although the precise mechanisms behind the pathogenesis of AS remain to be elucidated, several theories have been suggested. For many years it has been known that HLA‐B27 has a strong association with AS, with evidence that 94% of patients are HLA‐B27‐positive compared to only 9·5% of the general population 33. A number of clinically related conditions also share a high prevalence of HLA‐B27 positivity and, together, are considered spondyloarthropathies. These include juvenile enthesitis‐related arthritis (approximately 70% B27⁺), reactive arthritis (30–70%), colitis‐related spondyloarthritis (33–75%) and psoriatic spondyloarthritis (40–50%) 34.

The specific role of HLA‐B27 in the inflammatory process is largely unknown, but several mechanisms have been postulated: (i) arthritogenic peptide, (ii) HLA‐B27 protein misfolding resulting in intracellular stress and (iii) innate immune recognition of aberrant HLA‐B27 34.

Further studies, however, have shown that HLA‐B27 positivity alone is not sufficient to cause inflammation. Although there is a strong hereditary component to disease development, Brown et al. demonstrated that the concordance rate for HLA‐B27 positivity in dizygotic twins was considerably lower than for monozygotic twins (24 and 63%, respectively), suggesting the presence of other relevant genes 35.

Evans et al. demonstrated that polymorphisms of ERAP1 (a gene involved in peptide trimming prior to HLA class 1 presentation) affect the risk of ankylosing spondylitis development in individuals with HLA‐B27 positivity, therefore suggesting that the aberrant processing of antigenic peptides is significant in disease pathogenesis 4.

Additionally, genes IL‐1A and IL‐23A have been shown to be relevant 36 (Table 1).

Table 1.

Examples of genes implicated in ankylosing spondylitis 36.

Gene	Locus	Function (example)
HLA‐B	6p21·3	Immune recognition
ERAP1	5q15	Aminopeptidase
IL‐23R	1p31	Cytokine receptor for Th17 responses
IL‐1A	2q14	Proinflammatory response
STAT‐3		Th17 signalling

HLA = human leucocyte antigen; ERAP = endoplasmic reticulum aminopeptidase; IL – interleukin; STAT = signal transducer and activator of transcription; Th17 = T helper type 17.

Th17 responses in ankylosing spondylitis

Cortes et al. reported a GWAS in patients with AS and identified an over‐representation of genes involved in the Th17 pathway associated with AS, and specifically identified a major role for IL‐23 37; furthermore, caspase activation and recruitment domain 9 (a gene involved in the signalling pathway following fungal induction of IL‐23 in dendritic cells) and signal transducer and activation of transcription 3 (STAT‐3; which encodes the major signalling molecule activated by IL‐23R) have also been found in association with AS (and Crohn's disease) 32.

In keeping with this, studies have revealed a significantly higher incidence of IL‐23‐positive cells in the subchondral bone marrow in patients with AS compared to those from controls, adding to earlier work demonstrating raised serum levels of IL‐17 and IL‐23, and raised concentrations of IL‐23 in synovial fluid 32. This group further found that IL‐23‐positive cells were present at considerably higher numbers than IL‐12‐positive cells, signifying that pathways involving IL‐23 are probably more significant.

Various polymorphisms in the IL‐23R gene have been shown to be associated with AS 38, 39. Because of linkage disequilibrium it is difficult to determine fully exactly which are likely to be pathogenic; it appears that there is both a ‘primary’ AS‐associated single nucleotide polymorphism (SNP), which alters the amino acid sequence of the intracellular signalling portion of IL‐23R, as well as secondary associations with a probable regulatory function 40.

Chan et al. hypothesized that expression of aberrant HLA‐B27 in spondyloarthritis may act via cells containing killer cell immunoglobulin‐like receptor 3DL2 (KIR3DL2; a natural killer cell receptor for HLA‐B27 homodimer) and, interestingly, demonstrated increased expression of this receptor on natural killer cells and CD4⁺ T cells in patients with spondyloarthritis and enthesitis‐related arthritis 41. The binding of HLA‐B27 homodimers to KIR3DL2‐positive cells can stimulate IL‐17 production, and CD4 T cells positive for KIR3DL2 in the synovial fluid of patients with spondyloarthritis demonstrate increased IL‐17 secretion 34. Notably, an increased number of IL‐17‐associated cells has also been demonstrated in the facet joints of patients with ankylosing spondylitis, including both T cells and non‐lymphocytic cells 42.

DeLay et al. demonstrated in HLA‐B27 transgenic rat models of spondyloarthropathy that the IL‐23/IL‐17 axis is activated strongly in the colon and that misfolding of HLA‐B27 proteins and that unfolded protein response activation in macrophages can lead to increased IL‐23 induction, supporting a potential link between HLA‐B27 misfolding and immune dysregulation 43.

Interestingly, Payeli et al. demonstrated that a monoclonal antibody HD6, which bound specifically to both blood and synovial fluid mononuclear cells expressing abnormal HLA‐B27, inhibited IL‐17 production in cells from AS patients 44.

In comparison to previous studies on patients with rheumatoid arthritis and psoriatic arthritis, Appel et al. demonstrated that the majority of IL‐17‐producing cells in patients with AS were myeloperoxidase‐ and CD15‐positive neutrophils rather than CD3⁺ T cells and mast cells, suggesting that IL‐17‐producing cells other than Th17 cells are relevant in local inflammation in this population 42.

Sherlock et al. showed in murine models of spondyloarthritis that IL‐23 mediates disease, and the introduction of exogenous IL‐23 to otherwise normal mice was sufficient to induce features of spondyloarthritis. They further show that IL‐23 acts on T cells positive for various receptors [IL‐23R, RAR‐related orphan receptor γt (ROR‐γt) and stem cell antigen 1 (Sca1)], which allow entheses to respond to IL‐23 in the absence of further cellular recruitment, and induce inflammatory mediators such as IL‐6, IL‐17, IL‐22 and chemokines (C‐X‐C motif ligand 1); this study supports findings by other groups that the source of IL‐23 may be from myeloperoxidase (MPO)‐positive cells, which opposes earlier studies suggesting that macrophages may be the key source 32.

Management of AS

Non‐steroidal anti‐inflammatory drugs (NSAIDs) and physiotherapy are the cornerstones of treatment for AS 45. Disease‐modifying anti‐rheumatic drugs (DMARDS), used widely in rheumatoid arthritis treatment, are ineffective for the axial inflammation in AS. In certain patients, anti‐TNF agents, including the monoclonal antibodies (mAbs) infliximab, adalumimab and golumimab and the recombinant receptor etanercept, have also been shown to be beneficial 46. However, up to 40% of patients either fail to tolerate or respond to this treatment, and no approved alternative treatment options are currently available for this group of patients, suggesting that alternative therapies are needed 45.

Current strategies targeting Th17 responses in AS

Several therapies targeting aspects of the Th17 response have been trialled in numerous IMID and so far have shown promising results (Table 2).

Table 2.

Potential therapies targeting interleukin (IL)‐17/IL‐23 pathways.

Drug action	Examples			Trials
Anti‐IL‐17A	Secukinumab	Cosentyx	Phases II/III	Psoriasis, ankylosing spondylitis, rheumatoid arthritis
	Ixekizumab		Phase III	Psoriasis, psoriatic arthritis, rheumatoid arthritis
	Perakizumab		Phase I	Rheumatoid arthritis
Anti‐IL‐17R	Brodalumab		Phase III	Psoriasis, psoriatic arthritis, Crohn's disease, rheumatoid arthritis
Anti‐IL‐23 p19	Tildrakizumab		Phase II	Psoriasis
	Guselkumab		Phase II	Psoriasis
Anti‐IL‐12/23	Ustekinumab	Stelara	Phase III	Psoriasis,* psoriatic arthritis, ankylosing spondylitis, multiple sclerosis
	Briakinumab		Phase IIb	Rheumatoid arthritis, inflammatory bowel disease, multiple sclerosis

*Currently in routine clinical use in the United Kingdom following National Institute for Health and Care Excellence (NICE) guidance.

Following earlier promising Phases II and III studies of secukinumab (a monoclonal antibody against IL‐17A) in the treatment of psoriasis and psoriatic arthritis, Baeten et al. underwent a randomized, double‐blind placebo trial with secukinumab in patients with AS and found an improvement in ASAS 20 scores (i.e. patients reaching 20% improvement in ASAS criteria) in 59% of patients compared to 24% on placebo at 6 weeks 45. Furthermore, Bayesian analysis specified a 99·8% probability of secukinumab provoking improved response rates compared to placebo. In this study, quality of life was also measured as an outcome, and an improvement was noted in 52% of the treatment group compared to 33% of the placebo group; this was associated with an improvement in both C‐reactive protein (CRP), erythrocyte sedimentation rate (ESR) and S100A9 (an inflammatory biomarker protein) levels at 6 weeks. Furthermore, an improvement in ASAS score was correlated significantly with a change in S100A8/9 levels. Radiological imaging was performed on these patients at 6 and 28 weeks and, interestingly, magnetic resonance imaging (MRI) scores of axial inflammatory lesions improved in the secukinumab group, suggesting an improvement in clinical symptoms as well as joint damage.

These encouraging results have also been confirmed in larger studies presented at the American College of Rheumatology (ACR) meeting in 2014. Baeten et al. performed a 52‐week Phase III randomized controlled trial of secukinumab on 371 patients, and found a significant improvement in both signs and symptoms of AS which persisted throughout the duration of the study 47. Furthermore, another Phase III study conducted by Deodhar et al. demonstrated that secukinumab produces both rapid and sustained improvements in patient‐reported outcomes, such as fatigue, productivity and quality of life 48.

Ustekinumab (a monoclonal antibody against the p40 subunit of IL‐12 and IL‐23) has also shown to be beneficial in the treatment of AS. Poddhubnyy et al. performed a proof‐of‐concept study with ustekinumab on 20 patients who were followed‐up over a 24‐week period 49 and found a ≥ 50% improvement in Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) score in 55% of patients. Furthermore, 50 and 20% of patients, respectively, achieved the AS Disease Activity Score (ASDAS) clinically important improvement and major improvement criteria; impressively, at week 24, 25% of patients had an ASDAS score of <1·3, implying inactive disease. Improvement in patient‐reported outcome parameters and MRI score improvement was also noted. Phase II trials of this treatment in AS are currently under way.

Risks of therapies targeting the IL‐17/23 axis

All trials thus far report a favourable risk–benefit safety profile, with main adverse events including Candida infections and neutropenia 45; leucopenia has also been noted. In the study by Baeten et al., one patient developed a Staphylococcal aureus subcutaneous abscess which needed surgical intervention 45.

Future of treatment in ankylosing spondylitis

There are numerous potential benefits in targeting specific molecular pathways in AS:

• Beneficial in patients who fail to respond to conventional treatment options

• Improvement in axial disease, compared to other treatment options

• Limitation of disease progression and potentially induction of clinical remission, rather than symptomatic improvement only

• Targets specific pathways therefore limited side effect profile

• Well tolerated by patients in current clinical trials

Thus far, trials of therapies targeting the Th17 pathway in AS have been directed only towards IL‐17A and IL‐23, inhibitors against IL‐17R and IL‐23R. In current development are small molecules targeting key transcription factors of the IL‐17 axis, such as RORγ, all worthy of detailed preclinical and potentially clinical trials 1. Small molecule inhibitors targeting key signalling molecules such as Janus kinase 2 (JAK2), STAT‐3 and tyrosine kinase 2 (Tyk2) will also be promising candidates for therapeutic trials in AS and related IMIDs.

Disclosure

P.B. has received research support from Merck, given unrestricted educational talks supported byPfizer and is an investigator on a Novartis‐funded trial.