Targeting interleukin-17 in patients with active rheumatoid arthritis: rationale and clinical potential

Abstract

Clinical and experimental evidence suggest that interleukin-17A (IL-17A; also known as IL-17) is an attractive therapeutic target in rheumatoid arthritis (RA). Rheumatoid synovial tissue produces IL-17A, which causes cartilage and bone degradation in synovial and bone explants. Overexpression of IL-17A induces synovial inflammation and joint destruction in animal RA models. These effects are attenuated in IL-17A-deficient animals and by agents that block IL-17A. Serum IL-17A levels and, to a greater extent, synovial fluid IL-17A levels are elevated in many patients with RA. In some RA cohorts, higher IL-17A levels have been associated with a more severe clinical course. Several IL-17A blockers, including the anti-IL-17A monoclonal antibodies secukinumab and ixekizumab, and the anti-IL-17 receptor subunit A monoclonal antibody brodalumab have been evaluated in phase II clinical trials. Of these, secukinumab is the most advanced with respect to clinical evaluation in RA, with phase III trials ongoing in patients on background methotrexate who had inadequate responses to previous tumor necrosis factor blocker therapy.

Introduction

Rheumatoid arthritis (RA) is a chronic inflammatory disease associated with progressive joint destruction, functional disability, reduced health-related quality of life (HRQOL), systemic complications, premature mortality, and a high economic burden [Gullick and Scott, 2011; Koivuniemi et al. 2009; McInnes and Schett, 2011; Zhang and Anis, 2011]. Current guidelines recommend initial therapy with a conventional disease-modifying antirheumatic drug (DMARD), typically methotrexate, but responses are often inadequate due to its inability to stop progression of established disease [Smolen et al. 2010]. Biologics are recommended in patients unsuccessfully treated with multiple DMARD regimens, and in patients with an inadequate response to a single DMARD, but with poor prognostic features.

Several biologic classes are now approved for use in RA, including the tumor necrosis factor (TNF) blockers (i.e. adalimumab, etanercept, infliximab, golimumab, and certolizumab pegol), and agents targeting the interleukin (IL)-6 receptor (tocilizumab), IL-1 receptor (anakinra), cytotoxic T-lymphocyte antigen (CTLA)-4-mediated costimulation (abatacept), and B-cell depletion (rituximab). It is generally standard practice to start biologic therapy with a TNF blocker added to background DMARD therapy. However, approximately 30–40% of patients do not respond to TNF blockers, and even among patients who do respond initially, few achieve complete remission and many lose their responses over time [Aaltonen et al. 2012; Alonso-Ruiz et al. 2008; Finckh et al. 2006]. Patients with inadequate responses to initial TNF blocker therapy are often switched to a second TNF blocker or a biologic with an alternative mechanism. Although some patients respond, few achieve major durable responses [Salliot et al. 2011]. This clinical situation underscores the need for new biologics with novel mechanisms that can provide greater and longer lasting treatment responses. IL-17A has emerged as an attractive therapeutic target in RA. This paper reviews the rationale for targeting IL-17A and then describes the profile of several IL-17A blockers in initial clinical trials.

T-helper-17 cells and the interleukin-17A pathway

Historical perspective

The discovery of IL-17A and its role as the key effector of T-helper (Th)-17 cells occurred relatively recently compared with other major cytokines (e.g. interferon γ, TNFα, and IL-1β and IL-6) and T-cell helper subsets (i.e. Th1 and Th2; Figure 1). IL-17A was first identified from a clone of activated murine T cells in 1993, when it was termed CTLA-8 [Rouvier et al. 1993]. Two years later, IL-17A was shown to interact with a novel receptor that was unrelated to previously identified cytokine receptor families and is now known as IL-17RA [Yao et al. 1995a]. In 1999, rheumatoid synovial explants were shown to produce functional IL-17A, with IL-17-producing cells found in T-cell-rich areas of the synovium [Chabaud et al. 1999]. Two key experimental observations were made in 2001, suggesting that IL-17A may play an important role in mediating joint degradation in RA. First, in the collagen-induced arthritis (CIA) model in mice (a widely accepted experimental RA model), IL-17A overexpression accelerated development and enhanced severity of synovial inflammation, and radiographic analysis showed enhanced bone erosion [Lubberts et al. 2001]. Conversely, blocking endogenous IL-17A with a soluble IL-17 receptor fusion protein suppressed arthritis development and joint damage. Second, in human rheumatoid synovial and bone explants, IL-17A enhanced collagen degradation and bone resorption, and blocked collagen synthesis and bone formation [Chabaud et al. 2001]. Blocking IL-17A protected against these effects. The next major advance came in 2005 when IL-17A was shown to be produced by a new lineage of Th cells, termed Th17, which arise via a distinct pathway from the Th1 and Th2 subsets [Harrington et al. 2005; Park et al. 2005]. Importantly, Th17 cells were shown to be essential for the development of autoimmune inflammation in animal models [Langrish et al. 2005].

Figure 1.

Key events in understanding the role of interleukin (IL)-17A in rheumatoid arthritis (RA) [Rouvier et al. 1993; Yao et al. 1995a; Chabaud et al. 1999, 2001; Lubberts et al. 2001; Nakae et al. 2003; Park et al. 2005; Harrington et al. 2005; Langrish et al. 2005; Hueber et al. 2010a]. CIA, collagen-induced arthritis; CTLA, cytotoxic T-lymphocyte antigen.

T-helper-17 cell differentiation

Differentiation of Th17 cells from naïve cluster-of-differentiation (CD)-4-positive T cells is mediated by different cytokines and transcription factors than those involved in differentiation of other Th-cell lineages [Korn et al. 2009; Miossec et al. 2009]. The process was first characterized in murine cells, where a combination of transforming growth factor β and IL-6 activated retinoid-related orphan receptor (ROR)-γt, a unique transcription factor needed for expression of both IL-23R and IL-17A on developing Th17 cells. Subsequent exposure of these cells to IL-23 was necessary for full commitment to the Th17 phenotype, leading to enhanced IL-17A production, as well as secretion of other Th17 cytokines including IL-17F, IL-21, and IL-22. In human CD4-positive cells, IL-1β plus either IL-23 or IL-6 are needed to induce RORc, the human counterpart of RORγt, leading to Th17 commitment [Acosta-Rodriguez et al. 2007; Annunziato and Romagnani, 2009; Wilson et al. 2007]. Besides IL-17A, human Th17 cells produce IL-17F, IL-22, and IL-26. They also produce chemokine ligand 20 and express several chemokine receptors (CCRs), including CCR4 and CCR6. Because CCR6 is the receptor for chemokine ligand 20, Th17 cells may have the potential to promote migration of additional Th17 cells to inflammatory sites.

The CD4-positive regulatory T cells (Tregs) appear to play a reciprocal role in regulating Th17 cells. Specifically, induction of Tregs is associated with expression of transcription factor forkhead box P3 and suppression of RORγt [Ichiyama et al. 2008]. In the presence of proinflammatory cytokines, however, Tregs can be induced to differentiate into Th17 cells, which may help perpetuate a chronic inflammatory state [Bovenschen et al. 2011; Koenen et al. 2008].

Interleukin-17A pathway

Human IL-17A is a 15 kDa glycoprotein that contains 155 amino acids and has 63% homology to murine CTLA-8 [Yao et al. 1995b]. It is part of a structurally related cytokine family with six members (IL-17A through IL-17F). Of these, IL-17F has the highest homology with IL-17A at 50% [Iwakura et al. 2011]. Th17 cells secrete the IL-17A cytokine as a disulfide-linked homodimer, but heterodimers composed of IL-17A and IL-17F subunits, as well as IL-17F homodimers, also exist [Gaffen, 2009; Wright et al. 2007]. IL-17A, IL-17F, and the heterodimeric IL-17A/F cytokine all bind to a receptor complex consisting of IL-17RA and IL-17RC subunits [Gaffen, 2009; Pappu et al. 2010]. Binding stimulates downstream signaling, leading to activation of multiple transcription factors, including activator protein-1, nuclear factor-κB, and cytidine-cytidine-adenosine-adenosine-thymidine/enhancer-binding protein, and ultimately to enhanced expression of multiple chemokines, cytokines, and antimicrobial peptides. Although IL-17A and IL-17F bind to the same receptor complex, the former is approximately 10–30 times more potent than the latter, with the heterodimer having intermediate potency [Chang and Dong, 2007]. Moreover, as shown in rheumatoid synoviocytes, IL-17A regulates a much larger number of inflammation-related genes than IL-17F [Hot et al. 2011]. These observations suggest that the biological profiles of IL-17A and IL-17F are distinct, although they likely overlap in some regards.

Other IL-17 family members interact with other receptor complexes. Of note, IL-17E, also known as IL-25 and thought to play a role in stimulating Th2 responses, binds to a receptor complex that includes IL-17RA as one of its subunits [Rickel et al. 2008]. The other IL-17 family members have poorly characterized proinflammatory properties, but do not interact with receptors containing the IL-17RA subunit [Gaffen, 2009; Pappu et al. 2010].

Rationale for targeting interleukin-17A in rheumatoid arthritis

Evidence from experimental rheumatoid arthritis models

In addition to the effect of IL-17A overexpression mentioned above [Lubberts et al. 2001], several other observations in animal models point to the importance of IL-17A in driving synovial inflammation and joint destruction. The development of CIA in IL-17A-deficient mice was markedly suppressed, with reductions in synovial hyperplasia, cellular infiltration, and bone erosion compared with wild-type controls [Nakae et al. 2003]. Similarly, in streptococcal cell-wall-induced arthritis, a model of chronic relapsing arthritis that is unaffected by TNF blockers, joint inflammation and cartilage damage were significantly reduced in IL-17 receptor knockout mice compared with wild-type controls [Plater-Zyberk et al. 2009].

Anti-IL-17A treatment started after onset of CIA resulted in reduced histologic measures of joint inflammation, and radiographic evidence of cartilage and bone destruction [Lubberts et al. 2004]. Of note, the reduction in bone damage was associated with fewer osteoclasts in inflamed joints, and fewer Th17 cells in apposition to activated osteoclasts in subchondral regions, consistent with observations in rheumatoid synovial biopsy specimens [Pöllinger et al. 2011]. Agents that neutralize IL-17A have also been shown to reduce disease severity and joint destruction in other animal models, including adjuvant-induced arthritis in rats [Bush et al. 2002; Chao et al. 2011], antigen-induced arthritis in mice [Koenders et al. 2005], and glucose-6-phosphate isomerase-induced arthritis in mice [Ishiguro et al. 2011].

Clinical evidence

Concentrations of IL-17A in serum and synovial fluid are higher in patients with RA than with osteoarthritis (OA) or in healthy subjects (Table 1) [Metawi et al. 2011; Moran et al. 2009; Park et al. 2012; Suurmond et al. 2011; Ziolkowska et al. 2000]. Several observations suggest that elevated IL-17A levels may be associated with a more severe clinical course. Patients with RA and anticitrullinated protein antibodies (ACPAs) have a more erosive clinical course than patients with ACPA-negative disease [van Venrooij et al. 2011]. In a cohort of 59 patients with established RA, synovial IL-17A concentrations were significantly higher in patients with ACPA-positive RA than in those with ACPA-negative RA [Suurmond et al. 2011]. Serum and synovial fluid IL-17A levels showed significant positive correlations with Disease Activity Score in 28 joints (DAS28) in a cohort of 30 patients with active RA with knee effusions [Metawi et al. 2011]. Moreover, poorer functional status was associated with higher IL-17A levels. In another RA cohort, synovial fluid IL-17A levels were significantly correlated with C-reactive protein (CRP) levels and disease duration [Moran et al. 2009].

Table 1.

Serum and synovial fluid interleukin-17A levels in rheumatoid arthritis compared with control cohorts.

Reference	Results
Ziolkowska et al. [2000]	Serum and SF IL-17A levels > 10 pg/mL detected in 40% (6/15) and 80% (12/15) of patients with RA, respectively, versus 0% (0/8) of patients with OA; serum and SF IL-17A not significantly correlated (r = 0.299; p > 0.05)
Moran et al. [2009]	Serum IL-17A levels detected in 20% (8/40) of patients with IA versus 0% (0/8) of healthy controls; SF IL-17A levels detected in 57% (28/49) of patients with IA versus 20% (1/5) of patients with OA; SF IL-17A levels significantly higher than paired serum IL-17A levels (all p < 0.01)
Metawi et al. [2011]	Mean serum IL-17A significantly higher in 30 patients with RA versus 13 healthy controls (11.25 versus 0.6 pg/mL; p < 0.001); SF IL-17A levels averaged 8.4 pg/mL in patients with RA (range 6.4–13 pg/mL); positive correlation between serum and SF IL-17A (r = 0.501; p = 0.005)
Suurmond et al. [2011]	SF IL-17A levels significantly higher in patients with ACPA-positive versus ACPA-negative RA (p < 0.01) and patients with OA (p < 0.05); serum IL-17A levels not measured
Park et al. [2012]	Serum and SF IL-17A significantly higher in patients with RA versus patients with OA (p < 0.05 and < 0.01, respectively); mean SF IL-17A levels more than two times higher than mean serum IL-17A levels in patients with RA

ACPA, anticitrullinated protein antibody; IA, inflammatory arthritis (comprising RA and spondyloarthropathy); IL, interleukin; OA, osteoarthritis; RA, rheumatoid arthritis; SF, synovial fluid.

The levels of IL-17A and Th17 cells may differ across RA subsets. In a recent study, Th17 levels were measured in 33 patients with early RA and 20 with established RA, and 53 healthy control subjects [Arroyo-Villa et al. 2012]. The percentage of circulating Th17 cells was reduced in patients with early RA compared with healthy subjects, showing a negative correlation with ACPA titers. In turn, ACPA-positive disease was associated with the presence of radiographic bone erosions. However, levels of Th17 were not associated with rheumatoid factor, DAS28, erythrocyte sedimentation rate, or CRP. In the subset with established disease, elevated Th17 levels were found in synovial fluid, but not in the circulation. Taken together, these findings raise the hypothesis that Th17 cells may migrate to the inflamed joint in early erosive disease and then become sequestered there.

Recent studies suggest that mast cells, and not Th17 cells, are the major source of IL-17A in the rheumatoid synovium. In rheumatoid synovial biopsies, IL-17A-positive cells were found in the synovial sublining and at the periphery of lymphocytic aggregates, but only 1–8% expressed the T-cell marker CD3 and less than 1% expressed CD4 or CCR6 (i.e. markers found on Th17 cells) [Hueber et al. 2010a]. In contrast, most IL-17A-positive cells showed strong staining for mast cell tryptase. Consistent with this observation, stimulation of CD34-positive-derived mast cells resulted in RORγt-dependent IL-17A production, but other Th17 cytokines were not detected. In another study, nearly all IL-17A-positive cells in rheumatoid synovial specimens stained positive for the mast cell marker CD117, whereas only a small fraction stained positive for CD3, CD4, or the macrophage marker CD68 [Suurmond et al. 2011].

The biological effects of IL-17A in the rheumatoid joint are consistent with the synovial inflammation, cartilage destruction, and bone erosion seen in RA. Besides enhancing expression of a wide range of proinflammatory chemokines and cytokines, IL-17A stimulates production of matrix metalloproteinases by synovial tissue. IL-17A upregulated production of metalloproteinases-1, -2, -9, and/or -13 in rheumatoid synovial explants and synovial fibroblasts, and promoted matrix turnover and cartilage destruction in human cartilage cultures, particularly when evaluated in the presence of other cytokines (which mimic the microenvironment found in the rheumatoid joint) [Moran et al. 2009]. IL-17A may promote osteoclast differentiation leading to bone erosion by upregulating receptor activator of nuclear factor κB (RANK)-ligand expression on osteoblasts or RANK expression on osteoclast precursors, or indirectly by stimulating cytokine release from rheumatoid synovial fibroblasts or macrophages [Adamopoulos et al. 2010; Sato et al. 2006].

TNF-like weak inducer of apoptosis (TWEAK), a member of the TNF superfamily, has been suggested to play an important role in joint inflammation and bone erosion in RA [Dharmapatni et al. 2011]. Serum and synovial fluid TWEAK concentrations are significantly higher in patients with RA versus OA, and in patients with active versus inactive RA [Dharmapatni et al. 2011; Park et al. 2012]. Like IL-17A, soluble TWEAK induced RANK-ligand expression on osteoblasts, suggesting its potential for promoting bone resorption [Dharmapatni et al. 2011]. TWEAK has been shown to act synergistically with IL-23 or IL-21 to promote Th17 differentiation and IL-17A production, both of which are suppressed by blocking the TWEAK receptor [Park et al. 2012]. Moreover, IL-17A-positive Th17 cells expressing the TWEAK receptor were identified in synovial biopsies from patients with RA. These observations suggest that TWEAK may act, at least in part, through production of IL-17A.

Interleukin-17A and T-helper-17 cells as biomarkers

Initial findings suggest that increased circulating levels of Th17 cells and IL-17A may be predictive biomarkers in patients with an inadequate response to TNF blockers (TNF-IR). Levels of Th17 and IL-17A were measured at baseline and after 6 months of TNF-blocker therapy in 48 patients with RA [Chen et al. 2011]. Mean circulating levels of both parameters decreased significantly in parallel with reductions in DAS28 and ACPA titers in patients responding to TNF-blocker therapy. In contrast, Th17 and IL-17A levels increased significantly in patients with a TNF-IR, although TNFα levels declined. On logistic regression, high baseline IL-17A level (≥40 pg/mL) was identified as the only independent predictor of poor therapeutic response to TNF blockers. In another study, two independent cohorts (n = 24 and 19) with active RA were evaluated before and after 4 and 8–12 weeks of TNF-blocker therapy, respectively [Alzabin et al. 2012]. The percentage and total number of Th17 cells (but not Th1 cells) increased significantly in both cohorts after TNF-blocker therapy. Patients were classified according to European League Against Rheumatism (EULAR) response criteria. Nonresponse to TNF-blocker therapy was associated with increased p40 levels (subunit common to IL-12 and IL-23), as well as a trend for greater ex vivo IL-17 production from peripheral blood mononuclear cells. Moreover, higher Th17 cell levels at baseline were negatively correlated with changes in DAS28 erythrocyte sedimentation rate at 4 weeks. Large prospective studies are still needed to confirm and validate these findings before either Th17 or IL-17A levels can be used to predict which patients are unlikely to respond to TNF-blocker therapy and which may be candidates for treatment with IL-17A blockers.

Safety considerations

Despite the extensive rationale for targeting IL-17A, it is important to remember that this cytokine is the principal effector of Th17 cells and, therefore, plays an important role in host defense against extracellular bacteria and fungi at mucosal surfaces [Miossec et al. 2009; Onishi and Gaffen, 2010]. Accordingly, targeting IL-17A raises the potential risks of infection and immune dysfunction. Other biologics used in RA, including the TNF blockers, also have targets involved in host defense. Some reassurance about the safety of targeting IL-17A can be gleaned from patients with genetic disorders caused by IL-17RA or IL-17F deficiency. Such patients develop recurrent or persistent skin, nail, and mucosal infections caused by Candida albicans and, to a lesser extent, Staphylococcus aureus, but show no other infectious or autoimmune manifestations [Puel et al. 2011]. Chronic mucocutaneous candidiasis has also been reported in patients with autoimmune polyendocrine syndromes associated with production of autoantibodies against Th17 cytokines [Kisand et al. 2010; Puel et al. 2010]. Mucocutaneous candidiasis, therefore, remains a potential risk for IL-17A blockers, but may require concomitant inhibition of IL-17F.

Secukinumab

Secukinumab is a fully human immunoglobulin (Ig)-G1κ monoclonal antibody that binds with high affinity and selectivity to human IL-17A, resulting in neutralization of the cytokine’s activity [Hueber et al. 2010]. Secukinumab was initially evaluated in a randomized, placebo-controlled proof of concept (POC) trial involving 52 patients who had active RA despite receiving a stable dose of methotrexate (≤25 mg/week for ≥3 months) [Hueber et al. 2010b]. Patients were randomly allocated to receive placebo or intravenous secukinumab 10 mg/kg at weeks 0 and 3, and then were followed through week 16. The 3-week dosing interval reflected the drug’s expected serum half-life of 3–4 weeks. Patients receiving TNF blockers or other biologics within the previous 2–3 months, or intra-articular or systemic steroids, or conventional DMARDs other than methotrexate within 1 month were excluded. The primary endpoint was response rate at week 6 based on 20% improvement in American College of Rheumatology criteria (ACR20), with a treatment difference between secukinumab and placebo considered to be significant a priori if p was less than 0.20. The ACR20 rates at week 6 were 27% and 46% with placebo and secukinumab, respectively (p = 0.12). Responses to secukinumab versus placebo occurred rapidly (week 4 ACR20 rate: 50% versus 31%; p = 0.13) and were maintained at the final assessment (week 16 ACR20 rate: 54% versus 31%; p = 0.08). The DAS28 scores and CRP levels declined over time, with greater reductions seen with secukinumab versus placebo (p = 0.16 and = 0.001, respectively). When measured by area under the treatment response–time curve, secukinumab significantly improved ACR20 rates (p = 0.01), baseline-adjusted DAS28 score (p = 0.03), and baseline-adjusted CRP level (p = 0.002) compared with placebo.

The overall incidence of adverse events (AEs) was slightly higher in the secukinumab versus placebo group (81% versus 65%) [Hueber et al. 2010b]. Most individual AEs, however, occurred at similar rates in both groups, and none of the patients discontinued secukinumab due to an AE. No serious infections were reported. The rates of nonserious infections, consisting mostly of nasopharyngitis and upper respiratory tract infections, were identical (35%) in both groups.

On the basis of the positive findings in the POC trial, secukinumab was subsequently evaluated in a randomized, double-blind, placebo-controlled, dose-finding phase II trial that was conducted at 54 centers in Europe, Asia, and the United States [Genovese et al. 2012a]. In all, 237 patients with active RA on stable doses of methotrexate (7.5–25.0 mg/week) were randomly allocated to receive placebo, or subcutaneous secukinumab 25, 75, 150, or 300 mg every 4 weeks. Patients on stable corticosteroid therapy (prednisone ≤10 mg/day) were eligible, as were those unsuccessfully treated with biologics or other DMARDs after an appropriate washout period. The primary endpoint was ACR20 rate at week 16 for each secukinumab dose group versus placebo.

The ACR20 rates at week 16 were numerically, but not significantly, higher with secukinumab 75, 150, and 300 mg versus placebo (47%, 47%, and 54% versus 36%); the rate with the 25-mg dose was 34% [Genovese et al. 2012a]. Although the primary efficacy endpoint was not achieved, secukinumab produced clinically relevant reductions in several secondary endpoints compared with placebo. The DAS28-CRP scores were significantly reduced over the initial 16-week treatment period with secukinumab 75, 150, and 300 mg compared with placebo (Figure 2). These reductions were seen as early as week 2 and were maintained through week 16. The DAS28–erythrocyte sedimentation rate scores were similarly reduced from baseline during this period. Serum high-sensitivity CRP levels were also significantly lower with secukinumab at these three dose levels compared with placebo. Secukinumab 150 and 300 mg showed greater ACR20 and DAS28-CRP responses in patients with baseline high-sensitivity CRP levels at least 10 mg/liter versus less than 10 mg/liter. These trends were not evident with the other secukinumab dose levels.

Figure 2.

Change from baseline in Disease Activity Score in 28 joints–C-reactive protein (DAS28-CRP) with placebo, and secukinumab 25, 75, 150, and 300 mg administered subcutaneously every 4 weeks in a dose-ranging phase II trial. Reprinted with permission from Genovese et al. [2012a].

Patients who achieved ACR20 responses on secukinumab had statistically significant and clinically meaningful improvements in HRQOL, which were seen across all Short Form 36 version 2 (SF-36) Health Survey domains and the Functional Assessment of Chronic Illness Therapy-Fatigue scale [Gnanasakthy et al. 2012]. Higher levels of ACR response were associated with incrementally greater HRQOL improvements. When evaluated by dose level, the improvements in HRQOL on SF-36 exceeded minimum clinically important differences for the Physical and Mental Component Summary scores, and seven of eight domain scores with secukinumab 75 mg, and for all eight domains with 150 mg [Strand et al. 2012]. Improvements in HRQOL with these doses also exceeded minimum clinically important differences on the Short Form 6D instrument. Improvements in HRQOL with secukinumab 25 and 300 mg were less pronounced.

The safety profile of secukinumab was generally comparable to placebo [Genovese et al. 2012a]. AEs were reported in 47–61% of patients in the secukinumab groups compared with 58% in the placebo group. Infections, consisting mostly of nasopharyngitis and upper respiratory tract infections, were reported more often with secukinumab versus placebo, but were not dose related (18–29% versus 16%). Discontinuations due to AEs occurred in 2% of patients across treatment groups.

In the phase II trial, patients who achieved ACR20 responses at week 16 were treated monthly with the same dose until week 52, whereas nonresponders received a higher dose of secukinumab starting at week 20 [Genovese et al. 2012a]. Nonresponders on the 25- and 75-mg doses were titrated to 150 mg; those on 150 mg were titrated to 300 mg; and those on 300 mg continued on the same dose. All patients in the placebo group were switched to secukinumab 150 mg every 4 weeks. Patients who received secukinumab 150 mg throughout the study had the highest ACR50 rates at weeks 24 (50%) and 52 (55%) [Genovese et al. 2012b]. Health Assessment Questionnaire-Disability Index (HAQ-DI) scores also improved over time in responders to secukinumab 150 mg, with changes from baseline of -0.6 and -0.8 at weeks 24 and 52, respectively. Similarly, the proportions with EULAR remission increased over time; in the secukinumab 150 mg group, remission increased from 12% at week 16 to 30% at week 24 and 40% at week 52. Patients who switched from placebo to secukinumab achieved similar EULAR remission rates at weeks 24 and 52 (29% and 39%, respectively). In contrast, nonresponders did not derive much benefit from secukinumab dose escalation.

Phase III plans

The results of the dose-finding study indicated that secukinumab was worthy of further evaluation in phase III clinical trials. Doses of 75 and 150 mg delivered monthly via subcutaneous injection were selected for these trials based on their efficacy and HRQOL benefits, as well as the drug’s overall favorable safety profile. Two phase III clinical trials are currently ongoing.

The REASSURE 1 trial is comparing the safety and efficacy of secukinumab 75 and 150 mg versus placebo when added to background methotrexate therapy (7.5–25 mg/week) in patients with a TNF-IR and active RA [ClinicalTrials.gov identifier: NCT01377012]. Study treatment is being administered for up to 2 years. The primary efficacy outcome is ACR20 rate at week 24, and key secondary outcomes include changes from baseline in HAQ-DI, radiographic progression, and major clinical response rate (defined as 70% improvement in ACR response over 6 continuous months). Planned accrual is 630 patients.

The NURTURE 1 trial is also comparing the safety and efficacy of secukinumab 75 and 150 mg versus placebo in patients with a TNF-IR and active RA [ClinicalTrials.gov identifier: NCT01350804]. However, this study also includes a fourth arm with the active comparator abatacept, and study treatments are being added to background therapy with stable doses of methotrexate (7.5–25 mg/week) or another single conventional DMARD. Study treatment is being administered for up to 1 year. The ACR20 rate at week 24 is the primary efficacy outcome, and changes from baseline in HAQ-DI and major clinical response rate are key secondary outcome measures. Planned accrual is 548 patients. Patients completing this study will be eligible to enter a 4-year extension trial, which was designed to evaluate the long-term efficacy, safety, and tolerability of secukinumab 75 and 150 mg delivered monthly from prefilled syringes [ClinicalTrials.gov identifier: NCT01640938].

Ixekizumab

Ixekizumab is a humanized IgG4 anti-IL-17A monoclonal antibody [Leonardi et al. 2012]. This agent was initially evaluated in a two-part POC trial that was conducted at 17 sites in Australia, Belgium, and Romania [Genovese et al. 2010]. The first part was a single-dose escalation study, and the second part was a randomized, double-blind, placebo-controlled study. In the latter, 77 patients with active RA on stable doses of at least one DMARD received intravenous ixekizumab at doses of 0.2, 0.6, or 2 mg/kg, or placebo at weeks 0, 2, 4, 6, and 8, and then were followed for an additional 8 weeks. The primary efficacy endpoint, the change from baseline to week 10 in DAS28 scores, was -2.3, -2.2, and -2.4 at the three respective dose levels of ixekizumab compared with -1.7 with placebo (p ≤ 0.05 for lowest and highest doses versus placebo). Corresponding ACR20 rates were 74%, 70%, and 90% with ixekizumab and 56% with placebo (p ≤ 0.05 for 2 mg/kg versus placebo). AEs in the ixekizumab groups were not dose related. Leukopenia and vertigo were the most common AEs in the combined ixekizumab group, each occurring in 6.8% of patients. In addition, pharyngitis, rhinitis, and urinary tract infections each occurred at rates of 5.1%.

In the subsequent phase II trial, ixekizumab or placebo was added to background DMARD therapy in 260 patients with RA who were biologic naïve and 188 with a TNF-IR [Genovese et al. 2011]. Subcutaneous ixekizumab was administered at a dose of 3, 10, 30, 80, or 180 mg at weeks 0, 1, 2, 4, 6, 8, and 10. In the primary efficacy analysis, ixekizumab produced a statistically significant dose-related response based on logistic regression of ACR20 at week 12 in patients who were biologic naïve (p = 0.03). When analyzed by dose level, only the 30-mg dose significantly increased ACR20 at week 12 compared with placebo (70% versus 35%; p = 0.001). The ACR20 rates with the other ixekizumab doses ranged from 43% to 54%. Ixekizumab was administered only at the two higher dose levels (i.e. 80 and 180 mg) to the TNF-IR cohort, producing significantly higher ACR20 rates versus placebo [40% (p = 0.03) and 39% (p = 0.047), respectively, versus 23%]. Secondary endpoints, including changes from baseline in CRP levels and DAS28-CRP score, favored ixekizumab over placebo. AE rates through week 12 were similar across treatment arms, with infection rates slightly higher with ixekizumab versus placebo in the biologic-naïve (25% versus 19%) and TNF-IR (27% versus 23%) cohorts. Development of this agent is currently focused on psoriasis and psoriatic arthritis.

Brodalumab

Brodalumab is a fully human IgG2 anti-IL-17RA monoclonal antibody [Papp et al. 2012b]. Clinical results with this agent in RA were recently reported. In a randomized, double-blind, dose-escalation study, 40 patients with moderate to severe RA were randomized in a 3:1 ratio to receive brodalumab or placebo [Churchill et al. 2012]. Brodalumab was administered at 50, 140, or 210 mg subcutaneously every 2 weeks for six doses or at 420 or 700 mg intravenously every 4 weeks for two doses. The primary endpoint was safety. Overall, treatment-related AEs were reported in 23% of patients in the brodalumab groups (with leukocytosis the most common; 7%) and in 30% of patients in the placebo group (with headache the most common; 20%). At the doses administered, brodalumab occupied IL-17RA on circulating leukocytes and inhibited IL-17-mediated signaling. ACR response rates were an exploratory endpoint in this phase IB study. At week 13, the ACR20 rates were 37% with brodalumab and 22% with placebo.

Brodalumab was subsequently evaluated in a randomized, double-blind, placebo-controlled phase II trial [Pavelka et al. 2012]. A total of 252 patients with active RA despite methotrexate treatment who were biologic naïve were randomized to receive brodalumab 70, 140, or 210 mg subcutaneously or placebo at weeks 0, 1, 2, 4, 6, 8, and 10. The primary endpoint was ACR50 at week 12, which was achieved by 10–16% of patients in the brodalumab groups compared with 13% of those in the placebo group. Mean changes from baseline in DAS28 also did not differ significantly between brodalumab and placebo groups. AEs were similar across treatment arms. The study investigators concluded that there was no evidence of meaningful clinical efficacy, and therefore these preliminary results are not supportive of further evaluation of this agent in RA. As for ixekizumab, no trials in RA are ongoing and development is currently focused on psoriasis and psoriatic arthritis.

Although at least two of the drugs targeting IL-17 pathways described above continue to be developed for treatment of RA, all three drugs have demonstrated efficacy in psoriasis [Leonardi et al. 2012; Papp et al. 2012a, 2012b; Rich et al. 2012].

Conclusion

Clinical and experimental evidence indicates that IL-17A is a rational target for therapeutic intervention in RA. The Th17 cells are likely to be the major source of IL-17A in the circulation and, possibly, in synovial fluid, but mast cells are the major IL-17A source in rheumatoid tissue. Several IL-17A blockers have been evaluated in clinical trials; however, it is important to remember that although monoclonal antibodies may have the same molecular target (e.g. both secukinumab and ixekizumab target IL-17A) or overlapping targets (e.g. IL-17A versus IL-17RA), their clinical profiles can differ considerably due to differences in pharmacokinetics, antibody isotype, immunogenicity, and binding specificity and avidity. The results of the ongoing phase III trials should help to shed light on whether IL-17A is truly a viable therapeutic target in RA.