LY2439821, a humanized anti–interleukin-17 monoclonal antibody, in the treatment of patients with rheumatoid arthritis: A phase I randomized, double-blind, placebo-controlled, proof-of-concept study

Authors


Abstract

Objective

We undertook this study to evaluate safety, tolerability, pharmacokinetics, pharmacodynamics, and efficacy of LY2439821, a humanized anti–interleukin-17 (anti–IL-17) monoclonal antibody, in a first in-human trial in rheumatoid arthritis (RA) patients taking oral disease-modifying antirheumatic drugs (DMARDs).

Methods

This randomized, double-blind, placebo-controlled study consisted of 2 parts. In part A, 20 patients received 1 intravenous (IV) dose of LY2439821 (0.06, 0.2, 0.6, or 2.0 mg/kg, escalating) or placebo followed by 8 weeks of evaluation. End points included safety, tolerability, and pharmacokinetics. In part B, 77 patients received 1 IV dose of LY2439821 (0.2, 0.6, or 2.0 mg/kg) or placebo every 2 weeks for a total of 5 doses, with a total evaluation period of 16 weeks. End points included safety, tolerability, pharmacokinetics/pharmacodynamics, and efficacy (Disease Activity Score in 28 joints [DAS28] and percentages of patients meeting American College of Rheumatology 20%, 50%, or 70% improvement criteria [achieving an ACR20, ACR50, or ACR70 response]). The primary efficacy end point was the DAS28 at week 10.

Results

Baseline characteristics were similar across all groups. Changes in the DAS28 were significantly greater in the 0.2 mg/kg, 2.0 mg/kg, and all-LY2439821–combined groups (−2.3, −2.4, and −2.3, respectively) than in the placebo group (−1.7) at week 10 (P ≤ 0.05), and these differences were significant as early as week 1. Percentages of ACR20, ACR50, and ACR70 responses as well as improvements in the ACR core set of measures were greater in LY2439821-treated patients than in placebo-treated patients at multiple time points. There was no apparent dose-response relationship in treatment-emergent adverse events.

Conclusion

LY2439821 added to oral DMARDs improved signs and symptoms of RA, with no strong adverse safety signal noted. This first evaluation of LY2439821 supports neutralization of IL-17 as a potential novel goal for the treatment of RA.

The treatment of rheumatoid arthritis (RA) has traditionally included corticosteroids and conventional disease-modifying antirheumatic drugs (DMARDs). While these therapies provide some benefit, their efficacy has been limited. Newer, biologically-based therapies include molecules that inhibit cytokine activity (tumor necrosis factor [TNF] inhibitors, interleukin-1 receptor antagonist [IL-1Ra], or anti–IL-6R monoclonal antibody), block T cell–mediated costimulation (CTLA-4Ig), or modify B cell biology (anti-CD20) (1–3). These therapies have been effective in moderate-to-severe RA and have slowed disease progression, as determined radiographically, particularly when combined with methotrexate (MTX). However, less than two-thirds of the patients treated in pivotal trials meet at least the American College of Rheumatology 50% improvement criteria (achieve at least an ACR50 response) (4), and disease remission (a Disease Activity Score in 28 joints [DAS28] <2.6 [5]) is achieved by only a small minority of these patients (3, 6–9).

IL-17 (or IL-17A) is a proinflammatory cytokine produced primarily by a subset of CD4+ T cells, called Th17 cells, which represent a third subset of CD4+ “helper” lymphocytes in addition to the classically described Th1 and Th2 populations (10, 11). Th17 cells likely evolved to provide adequate immunity to specific classes of pathogens, such as extracellular bacteria and fungi, through the role of IL-17 in the generation and recruitment of neutrophils (12–14). However, aberrant Th17 responses and IL-17 production have been implicated in a variety of autoimmune diseases, including RA (15, 16).

IL-17 could contribute to the pathogenesis of RA in several ways, some in synergy with TNFα and/or IL-1β and others independently of these 2 cytokines. Under normal conditions, levels of IL-17 are extremely low or undetectable in human sera; however, 2 studies have shown elevated levels in sera and synovial fluid in a subset of patients with RA (17, 18). Additionally, Chabaud et al demonstrated that RA synovial tissue explants produce biologically active IL-17 in addition to IL-6, TNF, and IL-1β (19). Immunohistochemical staining of RA synovium has identified a subset of infiltrating T cells expressing IL-17 (19, 20).

In preclinical studies, IL-17 has been shown to contribute to joint inflammation and destruction, as well as to bone erosion, and these activities are, in part, independent of IL-1β. Additionally, blockade of IL-17 by antibody or soluble receptors reduces inflammation and bone erosion in various animal arthritis models (21). Thus, blocking the biologic activity of IL-17 is predicted to be beneficial in the treatment of RA, potentially reducing joint inflammation and preventing bone erosion.

LY2439821 is a humanized monoclonal antibody (IgG4) that binds to and neutralizes the biologic effects of IL-17. It is highly specific to IL-17, with no binding to other IL-17 family members (unpublished data). Based on the above-described data, blocking IL-17 is predicted to be beneficial in the treatment of RA and other autoimmune diseases.

This study is the first evaluation of the tolerability, pharmacokinetics, and pharmacodynamics of LY2439821 in humans, and was a combined single/multiple dose administration, randomized, double-blind, placebo-controlled, phase I study in RA patients taking background oral DMARDs. Tolerability and pharmacokinetics were evaluated after a single escalating dose in part A, and tolerability, pharmacokinetics, pharmacodynamics, and efficacy were evaluated after multiple dosing in part B.

PATIENTS AND METHODS

Patients.

Patients enrolled were ages 18–75 years and were diagnosed as having RA according to the 1987 revised criteria of the ACR (formerly, the American Rheumatism Association) (22) with a global functional class of I, II, or III according to the ACR 1991 revised criteria (23). Inclusion criteria included weight of 40–86 kg (inclusive) and the stable use of ≥1 DMARDs for a minimum of 4 weeks prior to the study (MTX 7.5–25 mg/week, hydroxychloroquine [HCQ] ≤400 mg/day, sulfasalazine [SSZ] 1,000–3,000 mg/day, leflunomide 5–20 mg/day, or azathioprine ≤150 mg/day or 2 mg/kg/day). Combinations of MTX, HCQ, and/or SSZ were allowed. Females of childbearing potential used appropriate contraception, as defined in the protocol. Additionally, patients entering part B of the study must have experienced an insufficient response to at least 1, but not more than 5, oral DMARDs and must have had ≥6 tender joints and ≥6 swollen joints (28-joint assessment) as well as either C-reactive protein >1.5 mg/dl or erythrocyte sedimentation rate ≥28 mm/hour at the time of screening for the study.

Primary exclusion criteria included the use of rituximab within 6 months, abatacept within 3 months, infliximab, adalimumab, cyclosporine, or mycophenolic acid within 2 months, and etanercept or anakinra within 28 days prior to study randomization. Evidence of any active or recent infections, or a history of malignancy or other autoimmune disease, were also exclusion criteria. Patients in part B were also excluded if the use of etanercept, adalimumab, infliximab, anakinra, or abatacept was discontinued for lack of efficacy. Patients could participate in only one part of the study.

Study design.

This was a randomized, double-blind, placebo-controlled study of LY2439821 conducted at 17 sites in Australia, Belgium, and Romania between November 2006 and February 2008. LY2439821 and placebo were administered by intravenous infusion (1 hour). Randomization during part A was performed manually by a study drug coordinator, and part B used an interactive voice response system for randomization. All study personnel remained blinded to treatment until study completion. The dose levels of LY2439821 administered in the study were determined based on the results of previous preclinical studies (unpublished data).

In part A, patients were given a single escalating intravenous (IV) dose of LY2439821 or placebo (5 patients per dose level; 1 randomized to placebo and 4 randomized to LY2439821 at 0.06, 0.2, 0.6, or 2.0 mg/kg). Escalation to the next LY2439821 dose level required acceptable safety data from all patients in the previous cohort through 24 hours postdosing and from at least 2 LY2439821-treated patients through 1 week postdosing. In order to maintain the blind, safety data through 1 week postdosing for the first 3 dosed patients were required prior to each safety review. Safety data from all patients in part A through a minimum of 3 weeks postdosing were reviewed by a formally constituted internal safety assessment committee prior to the commencement of part B. Primary end points included safety and tolerability. Pharmacokinetic parameters after a single dose were also evaluated.

Patients (total of 77) in part B were assigned by parallel randomization to treatment with placebo or with 0.2, 0.6, or 2.0 mg/kg LY2439821 administered IV every 2 weeks for a total of 5 treatments. Patients were then followed up for an additional 8 weeks after the final infusion. Primary end points included safety and the DAS28 at week 10. Secondary end points included percentages of patients achieving ACR20, ACR50, and ACR70 responses as well as improvements in the ACR core set of measures (24) (swollen joint count, tender joint count, patient's assessment of pain, patient's and physician's global assessments of disease activity, Health Assessment Questionnaire disability index [HAQ DI] score [25], and levels of acute-phase reactants) at various time points. Pharmacokinetic parameters were also assessed.

Regulatory and Ethical Review Board approvals from competent authorities in each country were obtained for the study protocol. All patients signed an informed consent document, and the study was conducted in accordance with the Declaration of Helsinki and following Good Clinical Practice guidelines.

Statistical analysis.

Baseline values were compared using an analysis of variance model. Covariates taken into consideration included treatment and center. For numeric continuous efficacy measures in part B, an analysis of covariance (ANCOVA) model was used that included treatment and center as fixed effects and the baseline measurement as a covariate. Pairwise comparisons between each dose and placebo were made using the ANCOVA model. For categorical efficacy response measures, a 2-sided Cochran-Mantel-Haenszel test was used, adjusted for center between each LY2439821 group and the placebo group. If any analytic group had fewer than 5 subjects, the 1-sided Fisher's exact test was used in place of the Cochran-Mantel-Haenszel test. No adjustments for multiplicity were made. Pharmacokinetic parameters were derived using a 2-compartment pharmacokinetic model employing WinNonlin Professional, version 5.2 (Pharsight, Mountain View, CA). Safety assessments were summarized descriptively by treatment group.

In part A, 20 patients were considered sufficient to determine safety and pharmacokinetics. In part B, completion by 17 patients in each treatment group provided 80% power to detect a treatment difference in the DAS28 of 1.2 at a significance level of 0.05, using a 1-sided 2-sample t-test. This assumed a common SD of 1.35 and did not include any adjustments for multiplicity. Such power assumptions were deemed appropriate for a first in-human proof-of-concept study. For a 2-sided analysis, with all other assumptions as above, 21 patients per group would have been required.

RESULTS

Patient baseline demographics, characteristics, and disposition.

Patients in part A were generally similar across treatment groups (Table 1). The only significant differences in treatment groups occurred in disease duration and in HCQ and glucocorticoid use. For disease duration, the 0.06 mg/kg treatment group had RA for an average of 22 years compared with an average of 5.6–8 years in all other groups. For HCQ use, the placebo and the 0.2 mg/kg treatment groups had mean daily doses of 300 mg and 200 mg, respectively (2 of 4 patients per group taking HCQ), while no patients in the other treatment groups were taking HCQ. For glucocorticoid use, the placebo and the 2.0 mg/kg groups had mean daily doses of 50 mg and 0.5 mg, respectively, compared with mean daily doses of 5–10 mg for the other groups. In the placebo group, only 1 patient took a concomitant glucocorticoid, at a dose of 25 mg twice a day. Dose for concomitant glucocorticoid use was not restricted as a criterion for entry into part A of the study. In part B, there was a statistically significant difference in the mean weekly dose of MTX; otherwise, patient demographics and baseline characteristics were similar across treatment groups (Table 2).

Table 1. Demographic and baseline characteristics of the patients for part A; single dose escalation*
 Placebo (n = 4)LY2439821P
0.06 mg/kg (n = 4)0.2 mg/kg (n = 4)0.6 mg/kg (n = 4)2.0 mg/kg (n = 4)
  • *

    Except where indicated otherwise, values are the mean ± SD. BMI = body mass index; ACR = American College of Rheumatology; MTX = methotrexate; HCQ = hydroxychloroquine; DMARD = disease-modifying antirheumatic drug.

  • P values are for overall comparison, and nonsignificance indicates randomization was done appropriately for demographic categories (age, sex, etc.).

  • By analysis of variance with treatment and center as fixed factors.

  • §

    By Fisher's exact test.

  • By analysis of covariance with treatment and center as fixed effects for continuous variables and by Fisher's exact test for categorical variables.

  • #

    Unable to estimate SD or SD not calculated.

Age, years41.3 ± 13.865.3 ± 7.851.7 ± 13.051.1 ± 14.249.3 ± 9.70.069
Weight, kg75.5 ± 11.072.3 ± 12.870.3 ± 12.374.3 ± 11.864.5 ± 5.50.606
BMI, kg/m227.8 ± 3.427.2 ± 5.225.6 ± 6.027.6 ± 4.423.9 ± 2.50.619
No. of men/women1/32/22/22/20/40.637§
ACR functional class, no. (%)     0.346
 I2 (50.0)3 (75.0)2 (50.0)2 (50.0)2 (50.0)
 II0 (0.0)1 (25.0)2 (50.0)2 (50.0)0 (0.0)
 III2 (50.0)0 (0.0)0 (0.0)0 (0.0)2 (50.0)
Disease duration, years7.7 ± 5.6622.0 ± 8.025.6 ± 3.768.0 ± 5.686.0 ± 4.600.004
Glucocorticoid use      
 Yes, no. (%)1 (25.0)4 (100.0)1 (25.0)3 (75.0)1 (25.0)
 Daily dose, mg50.0#6.3 ± 2.995.0#10.0 ± 5.000.5#<0.001
MTX use      
 Yes, no. (%)4 (100.0)4 (100.0)3 (75.0)2 (50.0)4 (100.0)
 Weekly dose, mg19.4 ± 5.1512.5 ± 2.8911.7 ± 2.8920.0 ± 0.0019.4 ± 14.770.684
HCQ use      
 Yes, no. (%)2 (50.0)0 (0.0)2 (50.0)0 (0.0)0 (0.0)
 Daily dose, mg300.0 ± 141.4200.00 ± 0.00<0.001
Leflunomide use      
 Yes, no. (%)0 (0.0)1 (25.0)0 (0.0)2 (50.0)0 (0.0)
 Daily dose, mg20.0#20.0 ± 0.00
Concomitant DMARD use,  no. (%)      
 One DMARD1 (25.0)3 (75.0)3 (75.0)4 (100.0)4 (100.0)
 Two DMARDs2 (50.0)1 (25.0)1 (25.0)0 (0.0)0 (0.0)
 Three DMARDs1 (25.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)
Table 2. Demographic and baseline characteristics of the patients for part B; multiple dose*
 Placebo (n = 18)LY2439821P
0.2 mg/kg (n = 19)0.6 mg/kg (n = 20)2.0 mg/kg (n = 20)
  • *

    Except where indicated otherwise, values are the mean ± SD. DAS28 = Disease Activity Score in 28 joints; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate; RF = rheumatoid factor; HAQ DI = Health Assessment Questionnaire disability index (see Table 1 for other definitions).

  • P values are for overall comparison, and nonsignificance indicates randomization was done appropriately for demographic categories (age, sex, etc.).

  • By analysis of covariance (ANCOVA) with treatment and center as fixed factors.

  • §

    By 2-sided Cochran-Mantel-Haenszel test adjusted for center (general association).

  • By ANCOVA with treatment and center as fixed effects for continuous variables and by Cochran-Mantel-Haenszel test for categorical variables.

  • #

    Unable to estimate SD or SD not calculated.

Age, years54.4 ± 9.759.6 ± 8.457.7 ± 11.555.6 ± 8.20.464
Weight, kg68.1 ± 12.767.3 ± 13.267.6 ± 11.565.3 ± 11.60.882
BMI, kg/m226.6 ± 4.525.4 ± 4.327.2 ± 3.924.7 ± 4.20.272
Sex, no. (%)    0.080§
 Male0 (0)3 (15.8)1 (5.0)5 (25.0)
 Female18 (100)16 (84.2)19 (95.0)15 (75.0)
ACR functional class, no. (%)    0.211
 I3 (16.7)3 (15.8)5 (25.0)1 (5.0)
 II5 (27.8)8 (42.1)5 (25.0)13 (65.0)
 III10 (55.6)8 (42.1)10 (50.0)6 (30.0)
Disease duration, years6.5 ± 5.6710.5 ± 13.0210.9 ± 10.216.1 ± 5.480.268
Swollen joint count, 0–2812.9 ± 4.1812.4 ± 5.3513.7 ± 5.3811.8 ± 4.950.314
Tender joint count, 0–2818.2 ± 4.7117.3 ± 6.3016.8 ± 6.3415.7 ± 7.550.458
DAS286.1 ± 0.746.0 ± 0.686.0 ± 0.835.8 ± 0.860.517
CRP, mg/dl2.47 ± 3.211.90 ± 2.161.80 ± 1.812.15 ± 3.010.891
ESR, mm/hour64.3 ± 25.2461.0 ± 26.0363.6 ± 23.9969.1 ± 26.260.783
RF titer, IU/ml125.2 ± 243.9569.1 ± 82.13275.6 ± 597.49130.1 ± 181.540.266
HAQ DI score1.8 ± 0.481.7 ± 0.551.8 ± 0.661.4 ± 0.740.065
Glucocorticoid use     
 Yes, no. (%)6 (33.3)4 (21.1)6 (30.0)8 (40.0)
 Daily dose, mg5.7 ± 2.664.9 ± 1.753.7 ± 0.825.1 ± 2.590.485
MTX use     
 Yes, no. (%)7 (38.9)9 (47.4)8 (40.0)7 (35.0)
 Weekly dose, mg12.9 ± 3.9310.5 ± 2.679.7 ± 3.1212.1 ± 2.670.006
HCQ use     
 Yes, no. (%)0 (0.0)0 (0.0)1 (5.0)0 (0.0)
 Daily dose, mg200.0#
Leflunomide use     
 Yes, no. (%)10 (55.6)6 (31.6)8 (40.0)10 (50.0)
 Daily dose, mg20.0 ± 0.0020.0 ± 0.0020.0 ± 0.0019.0 ± 3.160.510
Concomitant DMARD use, no. (%)     
 One DMARD1 (5.6)1 (5.3)2 (10.0)1 (5.0)
 Two DMARDs15 (83.3)16 (84.2)14 (70.0)17 (85.0)
 Three DMARDs2 (11.1)2 (10.5)4 (20.0)2 (10.0)

Part A enrolled 20 patients across 5 treatment groups at 2 study centers in Australia. No patients withdrew from part A due to adverse events (AEs); 2 patients withdrew for other causes (1 relocated and 1 withdrew based on investigator's decision). Part B enrolled 77 patients across 4 treatment groups (5 subjects from 3 study centers in Belgium and 72 subjects from 11 study centers in Romania). Study drug treatment was discontinued for 3 patients due to AEs (1 patient in the 0.6 mg/kg group experienced worsening of arthritis; 2 patients in the 2.0 mg/kg group experienced moderate leukopenia [white blood cell count >2,200/μl but <3,000/μl]). Treatment for 1 patient in the 2.0 mg/kg group was discontinued due to an abnormal pretreatment alanine aminotransferase concentration of 158 units/liter discovered after the first dose of study drug had been administered.

Safety.

AEs.

Table 3 presents treatment-emergent AEs occurring in ≥2 of the all-LY2439821–combined patients in part A or in >3% of the all-LY2439821–combined patients in part B. The term “all-LY2439821–combined patients” refers to the combined results from all of the LY2439821-treated groups within either part A or part B of the study.

Table 3. Treatment-emergent adverse events by preferred term occurring in ≥2 patients from all LY2439821-treated groups combined (part A) or in >3% of patients from all LY2439821-treated groups combined (part B)*
Preferred termPlaceboLY2439821
0.06 mg/kg0.2 mg/kg0.6 mg/kg2.0 mg/kgAll LY2439821-treated groups
  • *

    Values are the number (%) of patients experiencing an event. All adverse events listed were coded using the Medical Dictionary for Regulatory Activities, version 10.0. In part A, there were 4 patients in the placebo-treated group and 4 patients in each of the LY2439821-treated groups, for a total of 16 patients in all of the LY2439821-treated groups combined. In part B, there were 18 patients in the placebo-treated group, 19 patients in the 0.2 mg/kg LY2439821-treated group, and 20 patients in each of the 0.6 mg/kg and 2.0 mg/kg LY2439821-treated groups, for a total of 59 patients in all of the LY2439821-treated groups combined.

Part A      
 Diarrhea0 (0.0)1 (25.0)1 (25.0)1 (25.0)1 (25.0)4 (25.0)
 Abdominal pain0 (0.0)0 (0.0)2 (50.0)0 (0.0)0 (0.0)2 (12.5)
 Vomiting0 (0.0)1 (25.0)0 (0.0)0 (0.0)1 (25.0)2 (12.5)
 Fatigue0 (0.0)0 (0.0)0 (0.0)2 (50.0)0 (0.0)2 (12.5)
 Back pain0 (0.0)1 (25.0)1 (25.0)0 (0.0)0 (0.0)2 (12.5)
 Headache4 (100.0)1 (25.0)2 (50.0)2 (50.0)3 (75.0)8 (50.0)
 Paresthesia0 (0.0)1 (25.0)1 (25.0)0 (0.0)0 (0.0)2 (12.5)
Part B      
 Leukopenia0 (0.0)1 (5.3)1 (5.0)2 (10.0)4 (6.8)
 Neutropenia0 (0.0)1 (5.3)1 (5.0)0 (0.0)2 (3.4)
 Vertigo0 (0.0)1 (5.3)1 (5.0)2 (10.0)4 (6.8)
 Pharyngitis0 (0.0)1 (5.3)1 (5.0)1 (5.0)3 (5.1)
 Rhinitis1 (5.6)0 (0.0)3 (15.0)0 (0.0)3 (5.1)
 Urinary tract infection1 (5.6)1 (5.3)1 (5.0)1 (5.0)3 (5.1)
 Influenza0 (0.0)1 (5.3)1 (5.0)0 (0.0)2 (3.4)
 Headache1 (5.6)1 (5.3)1 (5.0)1 (5.0)3 (5.1)
 Hypotension0 (0.0)1 (5.3)1 (5.0)0 (0.0)2 (3.4)

In part A, headache was the most common treatment-emergent AE and occurred most frequently in the placebo group (Table 3). The most common treatment-emergent AEs occurring in patients who received LY2439821 were headache (8 patients compared with all 4 placebo-treated patients, all of mild or moderate intensity, and all resolved) and diarrhea (4 patients compared with no placebo-treated patients, all of mild or moderate intensity, and all resolved). There were no serious AEs noted in part A.

In part B, the most common treatment-emergent AEs that occurred in all-LY2439821–combined patients were leukopenia and vertigo (4 patients for both, compared with no placebo-treated patients for either), while rhinitis was the most common treatment-emergent AE occurring in any individual LY2439821 treatment group (3 patients, all in the 0.6 mg/kg group, compared with 1 placebo-treated patient) (Table 3). There was no apparent dose-response relationship in treatment-emergent AEs. One serious AE occurred in the 0.6 mg/kg treatment group. Three days after the last study drug infusion, the patient reported a skin ulcer that had been present for the previous 10 days. The patient was hospitalized due to the skin ulcer 13 days after the last dose of study drug. The ulcer improved with treatment, and the investigator did not consider the event related to study drug or protocol procedures.

Laboratory parameters.

Evaluation of laboratory parameters indicated that mean changes from baseline were generally small and similar across treatment groups, except for white blood cell, neutrophil, and platelet counts, which were slightly but significantly decreased in LY2439821-treated groups compared with the placebo group at several times during the study. Mild-to-moderate leukopenia and mild neutropenia were reported as treatment-emergent AEs in 4 and 2 patients receiving LY2439821, respectively, compared with none in the placebo-treated group (Medical Dictionary for Regulatory Activities [MedDRA], version 10.0) (Table 3). Based upon Common Terminology Criteria for Adverse Events, version 3.0, 3 patients (all treated with LY2439821) experienced white blood cell counts <3,000/μl during the study, and 5 patients (all treated with LY2439821) experienced neutrophil counts <1,500/μl. There were no cases of thrombocytopenia reported during the study.

Of the 3 patients with white blood cell counts <3,000/μl, one received 0.2 mg/kg LY2439821 and two received 2.0 mg/kg LY2439821 (the lowest recorded white blood cell count was 2,290/μl at week 6). Both patients in the 2.0 mg/kg treatment group were deemed to have moderate intensity leukopenia possibly/probably related to study treatment, and further study drug treatment was discontinued for both patients (one after 3 doses of LY2439821; one after 4 doses of LY2439821).

Of the 5 patients with neutrophil counts <1,500/μl, two received 0.2 mg/kg LY2439821, two received 0.6 mg/kg LY2439821, and one received 2.0 mg/kg LY2439821 (the lowest recorded neutrophil count was 1,040/μl at week 6). None of the patients discontinued the study due to neutropenia.

Infections.

The frequency of infections was low in this study. Pharyngitis, rhinitis, urinary tract infections, and influenza were the only infections reported in >3% of the all-LY2439821–combined patients in part B. Rhinitis and urinary tract infections also occurred in the placebo group; however, pharyngitis and influenza did not. The most frequently occurring infection in any one treatment group was rhinitis, which occurred in 3 patients taking 0.6 mg/kg LY2439821 (Table 3). Of the patients with low white blood cell or neutrophil counts, only 1 (with both mild leukopenia and mild neutropenia as defined by MedDRA coding) had an infection (of the urinary tract), which presented and resolved prior to the onset of low white blood cell and neutrophil counts, as well as a mild bacteriuria, determined by the investigator to be possibly related to the study drug, which presented after the onset of low white blood cell and neutrophil counts.

Immunogenicity and allergic and infusion reactions.

Anti-LY2439821 antibodies were detected in 2 patients in part B of the study and at both postdosing time points analyzed (week 8 and week 16). Antibodies were detected in 1 patient each in the 0.2 mg/kg (titer of 1:2) and 0.6 mg/kg (titer of 1:10) treatment groups. Neither of these patients experienced treatment-emergent AEs, and no alterations in pharmacokinetics were noted. No antibodies to LY2439821 were detected in any placebo-treated patients.

One patient in part A (0.2 mg/kg treatment group) experienced an adverse drug reaction, deemed by the investigator to be a case of moderate immune type III reaction. This patient experienced a skin rash and generalized pruritus, preceded by various symptoms (headache, abdominal pain, diarrhea, somnolence, and anorexia) and followed by a delayed increase in transaminases. All clinical symptoms resolved completely, and there were no detectable anti-LY2439821 antibodies in this patient. Another patient in the 0.2 mg/kg treatment group in part A experienced infusion site pain. There were no other reports of allergic or infusion reactions during the study.

Pharmacokinetics.

The pharmacokinetic properties of LY2439821 were similar in both parts of the study. The mean terminal elimination half-life ranged from 15 to 18 days in part A and from 14 to 16 days in part B. Because of the long half-life, accumulation of drug was observed in part B, in which dosing occurred every 2 weeks. Maximum concentration and area under the curve in both parts of the study showed dose proportionality and linear pharmacokinetics across the doses tested.

Efficacy.

DAS28 scores.

In part B, DAS28 scores decreased as early as week 1 in all LY2439821 dose groups relative to placebo, and this observation persisted to the end of the study (Figure 1A). At week 10, the changes from baseline in DAS28 values were −2.3, −2.2, −2.4, and −2.3 in the 0.2 mg/kg, 0.6 mg/kg, 2.0 mg/kg, and all-LY2439821–combined groups, respectively (P ≤ 0.05 versus placebo for all except the 0.6 mg/kg group), compared with −1.7 in the placebo group (least squares mean) (Figure 1A). At this time, 73.7%, 65.0%, 85.0%, and 74.6% of patients in the 0.2 mg/kg, 0.6 mg/kg, 2.0 mg/kg, and all-LY2439821–combined groups, respectively, experienced improvements in the DAS28 of ≥1.2 compared with 44.4% of placebo-treated patients.

Figure 1.

Change from baseline in Disease Activity Score in 28 joints (DAS28) by treatment over time, and percentage of patients meeting the American College of Rheumatology 20%, 50%, or 70% improvement criteria (achieving an ACR20, ACR50, or ACR70 response). Values are the least squares mean per treatment group. A, Change in DAS28 over time. B, ACR20 responses at week 10. The percentages of patients achieving an ACR20 response were 73.7%, 70.0%, and 90.0% for the 0.2 mg/kg, 0.6 mg/kg, and 2.0 mg/kg LY2439821 (LY)–treated groups, respectively, compared with 55.6% for the placebo-treated group (P ≤ 0.05 versus placebo for the 2.0 mg/kg group). C, ACR50 responses at week 10. The percentages of patients achieving an ACR50 response were 42.1%, 40.0%, and 35.0% for the 0.2 mg/kg, 0.6 mg/kg, and 2.0 mg/kg LY2439821-treated groups, respectively, compared with 16.7% for the placebo-treated group. D, ACR70 responses at week 10. The percentages of patients achieving an ACR70 response were 26.3%, 20.0%, and 25.0% for the 0.2 mg/kg, 0.6 mg/kg, and 2.0 mg/kg LY2439821-treated groups, respectively, compared with 5.6% for the placebo-treated group. ∗ = P ≤ 0.05 versus placebo.

ACR20, ACR50, and ACR70 responses.

Significantly greater percentages of patients achieved an ACR20 response at multiple time points and doses of LY2439821. Significantly greater percentages of patients achieved an ACR20 response at all dose levels by week 2 (26.3%, 30.0%, 45.0%, and 33.9% for 0.2 mg/kg, 0.6 mg/kg, 2.0 mg/kg, and all-LY2439821–combined groups, respectively, compared with 0.0% for the placebo group) (Figure 1B). At week 10, the percentages of patients achieving an ACR20 response were 73.7%, 70.0%, 90.0%, and 78.0% for the 0.2 mg/kg, 0.6 mg/kg, 2.0 mg/kg, and all-LY2439821–combined groups, respectively (P ≤ 0.05 versus placebo for the 2.0 mg/kg group), compared with 55.6% for the placebo group. Also at week 10, percentages of patients achieving ACR50 and ACR70 responses in the LY2439821-treated groups (for the ACR50 response, 42.1%, 40.0%, 35.0%, and 39.0% for the 0.2 mg/kg, 0.6 mg/kg, 2.0 mg/kg, and all-LY2439821–combined groups, respectively; for the ACR70 response, 26.3%, 20.0%, 25.0%, and 23.7% for the 0.2 mg/kg, 0.6 mg/kg, 2.0 mg/kg, and all-LY2439821–combined groups, respectively) were higher than those in the placebo group (16.7% and 5.6% for ACR50 and ACR70 responses, respectively) (Figures 1C and D). Throughout the study, most patients showed improvements in the ACR core set of measures, with LY2439821-treated patients showing significant improvements compared with placebo-treated patients at multiple time points (Figures 2 and 3).

Figure 2.

Responses in American College of Rheumatology core set parameters by treatment over time. A, Swollen joint counts. B, Tender joint counts. C, Health Assessment Questionnaire disability index (HAQ DI) score. D, Change from baseline in C-reactive protein. Values are the least squares mean per treatment group. * = P ≤ 0.05 versus placebo. LY = LY2439821.

Figure 3.

Responses in American College of Rheumatology core set parameters by treatment over time. A, Patient's assessment of pain. B, Patient's global assessment of disease activity. C, Physician's global assessment of disease activity. Values are the least squares mean per treatment group. ∗ = P ≤ 0.05 versus placebo. LY = LY2439821.

DISCUSSION

IL-17 is a unique proinflammatory cytokine amenable to targeting inflammatory and autoimmune diseases. This phase I, two-part, single/multiple dose administration study in patients with active RA provides preliminary insight into the safety and potential efficacy of targeting this novel pathway. With currently available disease-modifying therapies, less than two-thirds of RA patients reach at least an ACR50 response. Furthermore, disease remission (DAS28 <2.6) is achieved by only a small minority of these patients, emphasizing the need to develop novel therapies. The results of this study demonstrate that neutralization of IL-17 could provide a viable goal for future treatment of RA. IL-17 acts as an “effector” cytokine, much like TNFα. IL-17 can synergize with other proinflammatory cytokines to stimulate release of additional proinflammatory cytokines and chemokines, nitric oxide, and matrix metalloproteinases in many cell types. Since the IL-17 receptor is ubiquitously expressed, virtually all cell types have been demonstrated to have a biologic response to IL-17. Evidence suggests that IL-17 may have several roles including the promotion of neutrophil homeostasis, host defense against extracellular bacteria and some fungi, and chronic pathogenic inflammation such as that found in autoimmune disease, asthma/allergy, and lung inflammation (11). There is considerable evidence that IL-17 plays an important and nonredundant role in the pathogenesis of arthritis in animal models (21).

The two-staged design of this trial allowed the assessment of pharmacokinetics and tolerability with single dose administration in RA patients receiving concomitant DMARD therapy without requiring a particular level of active disease to be present. Upon satisfactory completion of part A, and with no clear concerns about tolerability, the study continued with a parallel design in part B, which explored multiple dose administration and allowed both safety and efficacy assessments in a population of patients with active RA receiving concomitant oral DMARD therapy.

Pharmacokinetic parameters were reproducible across both parts of the study. The half-life was similar after single and multiple dosing, ranging from 15 to 18 days and from 14 to 16 days, respectively, indicating the potential for dosing every 2 weeks or less frequently. Tolerability appeared to be favorable, with only 1 serious AE reported in part B of the study. AEs were also few, mild to moderate in severity, and did not increase with drug dose escalation. White blood cell, platelet, and neutrophil counts were slightly but significantly decreased in LY2439821 treatment groups versus the placebo group at several time points during the study. Four patients experienced mild-to-moderate leukopenia, only 2 patients experienced mild neutropenia according to MedDRA, version 10.0 coding (Table 3), and no patients experienced thrombocytopenia. Given the low incidence of patients with leukopenia or neutropenia, and the dosing with concomitant medications known to be associated with decreased blood cell counts, the causal relationship of these changes to LY2439821 is unclear. However, IL-17 induces production of granulocyte colony-stimulating factor and other chemokines and cytokines, contributing to neutrophil progenitor differentiation and maturation (14). Therefore, blocking such effects with LY2439821 could possibly result in reduced neutrophil counts.

Because the study was an initial assessment of LY2439821 in humans, the focus on safety was conservative, and the patients experiencing moderate leukopenia were withdrawn due to the potential higher risk of infection. In a relatively short experimental period, patients treated with LY2439821 demonstrated robust and significant improvements in DAS28 scores as well as significantly greater percentages of ACR20 responses. Both parameters showed improvements that occurred early in the study (as early as week 1) and that were sustained through at least 16 weeks (8 weeks after the last dose of drug). LY2439821-treated patients also showed significant improvements at multiple time points in tender joint count, swollen joint count, patient's assessment of pain, patient's global assessment of disease activity, and HAQ DI score. The physician's global assessment of disease activity decreased in LY2439821-treated patients but with fewer significant differences versus placebo than observed for the patient's global assessment of disease activity.

Since this was a phase I study, there was a primary focus on safety and tolerability after both single and multiple dose administration; thus, efficacy parameters were not addressed until part B. Further, despite the observed efficacy, the study failed to demonstrate a clear dose-response relationship between LY2439821 and efficacy parameters or with respect to tolerability. The 2.0 mg/kg dose, however, did show consistent improvements compared with the other dose levels, and findings were not meaningfully different with regard to safety. A high placebo response (55.6% of placebo-treated patients had an ACR20 response at weeks 10 and 16) was observed in this study. It is not clear whether this observation represents unique features of this study, the patient population, the investigators, or access to care in the populations concerned. Despite the placebo response, a significant difference was demonstrated between the active treatment and placebo groups in both the DAS28 and the percentages of ACR responders. The DAS28 at week 10 was chosen as the primary efficacy end point because, as a validated continuous outcome measure and a frequently used composite index in RA outside of the US, it allowed for a comparison of efficacy with the smaller sample size consistent with phase I safety studies. Furthermore, the study was not powered to draw definitive conclusions on dose effects or to assess accurately the proportion of ACR responders, but rather to provide an early opportunity to assess potential benefit in a multiple dose phase I trial.

Recently, results were presented from a proof-of-concept trial evaluating another anti–IL-17 antibody (26). Patients in that study also had active disease despite taking concomitant MTX. The data indicated that treatment with anti–IL-17 improved signs and symptoms compared with placebo, further validating this cytokine as a potentially valuable target for the treatment of RA.

The results of the current study indicate that LY2439821 may offer a therapeutic alternative for patients with active RA. The ultimate safety and efficacy profile of this approach will need to be further explored in large scale clinical trials. The positive results observed, coupled with a good tolerability profile, provide a solid proof of concept for LY2439821 in the treatment of RA. The utility of this compound in RA therefore warrants further investigation.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Genovese had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Genovese, Roberson, Sloan-Lancaster.

Acquisition of data. Van den Bosch, Roberson, Bojin, Biagini, Ryan.

Analysis and interpretation of data. Genovese, Sloan-Lancaster.

ROLE OF THE STUDY SPONSOR

As sponsor of the study, Chorus, a Division of Eli Lilly, was responsible for the creation of the study design, in close collaboration with Dr. Genovese. Chorus oversaw the data collection and analysis, as well as data interpretation. Chorus actively participated in the writing of the manuscript. Eli Lilly approved the content of the submitted manuscript prior to its submission for publication. Publication of the manuscript was contingent on Eli Lilly having an opportunity to review it prior to its submission.

Acknowledgements

We wish to thank the patients and the principal investigators for this study, and to acknowledge all of the enrollment contributions by investigators, who were as follows: Dr. Peter Ryan, Dr. Susan Proudman, Dr. Iuliana Mirela Biagini, Dr. Silvia Bojin, Dr. Monica Sanziana Capraru, Dr. Liana Maria Chicea, Dr. Dorica Cristei, Dr. Eugen Dumitru, Dr. Ruxandra Ionescu, Dr. Mihail Bogdan Jantes, Dr. Gavrila Mirea, Dr. Gabriela Udrea, Dr. Eugenia Mociran, Dr. Ilse Hoffman, Dr. Jan Lenaerts, and Dr. Filip Van Den Bosch. We wish to acknowledge Ling Liu, PhD, and Jirong Lu, PhD, for their contributions to the functional and physical characterization of LY2439821; and Susan Strobel, PhD, Stephen L. Myers, MD, John Polzer, PhD, Kristy Kikly, PhD, Pierre-Yves Berclaz, MD, Eiry Roberts, MD, and Olivier Benichou, MD, for their contributions in the preparation of the manuscript.

Ancillary