RESUMEN
BACKGROUND: Granulocyte-macrophage colony-stimulating factor (GM-CSF) is a key mediator of signs and symptoms in preclinical models of osteoarthritis. We explored the efficacy, safety, and pharmacokinetics of an anti-GM-CSF antibody, otilimab, in patients with hand osteoarthritis. METHODS: This double-blind, randomised, placebo-controlled phase 2a study was done in 16 centres in the Netherlands, Germany, Poland, the UK, and the USA. Patients aged 18 years or older with inflammatory hand osteoarthritis, who had received at least one course of unsuccessful non-steroidal anti-inflammatory drugs, with two or more swollen and tender interphalangeal joints (on the same hand), signs of inflammation or synovitis identified with MRI in the affected hand, and a self-reported 24 h average hand pain intensity over the past 7 days of 5 or more on a 0-10 numerical rating scale were eligible for inclusion. Patients were randomly assigned (1:1) via interactive response technology (blocked randomisation; block size of four) to receive either subcutaneous otilimab 180 mg or placebo, administered weekly from week 0 to week 4, then every other week until week 10. Patients, investigators, and trial staff were masked to treatment; at least one administrator at each site was unmasked to prepare and administer treatment. The primary endpoint was change from baseline in 24 h average hand pain numeric rating scale averaged over 7 days before the visit at week 6. Secondary endpoints were: change from baseline in 24 h average and worst hand pain intensity at each visit; proportion of patients showing 30% and 50% reductions in 24 h average and worst hand pain intensity at each visit; change from baseline in Australian and Canadian Hand Osteoarthritis Index (AUSCAN) 3·1 numeric rating scale questionnaire components at each visit; change in number of swollen and tender hand joints at each visit; change from baseline in Patient and Physician Global Assessment of disease activity; serum concentration of otilimab; and safety parameters. Efficacy endpoints were assessed in the intention-to-treat population. The safety population included all patients who received at least one dose of study treatment. The study is registered with ClinicalTrials.gov, NCT02683785. FINDINGS: Between March 17, 2016, and Nov 29, 2017, 44 patients were randomly assigned (22 in the placebo group and 22 in the otilimab group). At week 6, difference in change from baseline in 24 h average hand pain numeric rating scale between the otilimab and placebo groups was -0·36 (95% CI -1·31 to 0·58; p=0·44); at week 12, the difference was -0·89 (-2·06 to 0·28; p=0·13). Patients receiving otilimab showed greater improvement in AUSCAN components versus placebo at each visit. The change from baseline in the 24 h worst hand pain numeric rating scale in the otilimab group at week 6 showed a difference over placebo of -0·33 (95% CI -1·28 to 0·63; p=0·49); at week 12, this difference was -1·01 (95% CI -2·22 to 0·20; p=0·098). The proportion of patients achieving 30% or higher or 50% or higher reduction from baseline in the 24 h average and worst hand pain numeric rating scale scores was consistently greater (although non-significant) with otilimab versus placebo. Similarly, patients receiving otilimab showed greater improvement in AUSCAN pain, functional impairment, and stiffness scores versus placebo at each visit. No differences were observed between otilimab and placebo in the change from baseline in the number of swollen and tender joints across the study. The Patient Global Assessment was consistently lower than placebo at all visits; the Physician Global Assessment showed reductions across the study period, but the changes were similar in both treatment groups. Median otilimab serum concentrations increased during weekly dosing from 1730 ng/mL at week 1 to a maximum of 3670 ng/mL at week 4, but declined after transitioning to dosing every other week (weeks 6-10). In total, 84 adverse events were reported by 24 patients: 54 adverse events in 13 (59%) patients in the otilimab group and 30 adverse events in 11 (50%) patients in the placebo group. The most common adverse events were cough (reported in 4 [9%] patients; 2 in each group), and nasopharyngitis (in 3 [7%] patients; 1 in the placebo group and 2 in the otilimab group). Three serious adverse events occurred in this study (all in the otilimab group) and were deemed not related to the study medication. There were no deaths during the study. INTERPRETATION: There was no significant difference between otilimab and placebo in the primary endpoint, although we noted a non-significant trend towards a reduction in pain and functional impairment with otilimab, which supports a potential role for GM-CSF in hand osteoarthritis-associated pain. There were no unexpected safety findings in this study, with no treatment-related serious adverse events reported. FUNDING: GlaxoSmithKline.
RESUMEN
BACKGROUND: Otilimab is a human monoclonal antibody that inhibits granulocyte-macrophage colony-stimulating factor (GM-CSF), a driver in many immune-mediated inflammatory conditions. We evaluated the effect of otilimab on the GM-CSF-chemokine (C-C motif) ligand 17 (CCL17) axis and synovitis in patients with rheumatoid arthritis. METHODS: This phase 2a, randomised, double-blind, multicentre, placebo-controlled, parallel-group study was done at nine sites across the USA, Poland, and Germany. Patients aged 18 years or older with rheumatoid arthritis per American College of Rheumatology-European League Against Rheumatism 2010 criteria and receiving stable methotrexate were randomly assigned (3:1) by an interactive response technology system to either subcutaneous otilimab 180 mg or placebo once weekly for 5 weeks, then every other week until week 10 (within a 12-week treatment period), followed by a 10-week safety follow-up. Randomisation was stratified by early rheumatoid arthritis (≤2 years since diagnosis) and established rheumatoid arthritis (>2 years since diagnosis). Patients and study personnel (except for an unblinded coordinator or nurse who prepared and administered the study drug) were blinded to treatment assignment; the syringe was shielded during administration. Patients were enrolled by study investigators and allocated to a treatment by central randomisation on the basis of a schedule generated by the sponsor. The primary endpoint was change over time (assessed at baseline and weeks 1, 2, 4, 6, 8, 12, and 22 of follow-up) in 112 biomarkers, including target engagement biomarkers and those that may be indicative of rheumatoid arthritis disease activity and response to otilimab. Secondary endpoints were change from baseline in synovitis, osteitis and erosion assessed by rheumatoid arthritis MRI scoring system (RAMRIS) and rheumatoid arthritis MRI quantitative score (RAMRIQ), and safety evaluation. The primary, secondary, and safety endpoints were assessed in the intention-to-treat population. Biomarker and MRI endpoints were analysed for differences between treatment groups using a repeated measures model. This study is registered with ClinicalTrials.gov, NCT02799472. FINDINGS: Between Aug 9, 2016, and Oct 30, 2017, 39 patients were randomly assigned and included in the analysis (otilimab n=28; placebo n=11). In the otilimab group, mean serum concentrations of GM-CSF-otilimab complex peaked at week 4 (138·4 ng/L, 95% CI 90·0-212·9) but decreased from week 6-12. CCL17 concentrations decreased from baseline to week 1, remained stable to week 8, and returned to baseline at week 12; least-squares mean ratio to baseline was 0·65 (95% CI 0·49-0·86; coefficient of variation 13·60) at week 2, 0·68 (0·53-0·88; 12·51) at week 4, 0·78 (0·60-1·00; 12·48) at week 6, and 0·68 (0·54-0·85; 11·21) at week 8. No meaningful change in CCL17 concentrations was observed with placebo. In the otilimab group, the least-squares mean ratio to baseline in MMP-degraded type I collagen was 0·86-0·91 over weeks 1-8, returning to baseline at week 12; concentrations remained above baseline at all timepoints in the placebo group. There were no observable differences between otilimab and placebo for all other biomarkers. At week 12, least-squares mean change in RAMRIS synovitis score from baseline was -1·3 (standard error [SE] 0·6) in the otilimab group and 0·8 (1·2) with placebo; RAMRIQ synovitis score showed a least-squares mean change from baseline of -1417·0 µl (671·5) in the otilimab group and -912·3 µl (1405·8) with placebo. Compared with placebo, otilimab did not show significant reductions from baseline to week 12 in RAMRIS synovitis, osteitis and bone erosion, or in RAMRIQ synovitis and erosion damage. Adverse events were reported in 11 (39%) of 28 otilimab-treated and four (36%) of 11 placebo-treated patients, most commonly cough in the otilimab group (2 [7%] of 28; not reported in placebo group), and pain in extremity (four [36%] of 11) and rheumatoid arthritis (two [18%] of 11) in the placebo group (not reported in otilimab group). There were no serious adverse events or deaths. INTERPRETATION: Serum concentrations of GM-CSF-otilimab complex indicated that target engagement was achieved with initial weekly dosing, but not sustained with every other week dosing. CCL17 might be a pharmacodynamic biomarker for otilimab activity in future studies. Otilimab was well tolerated and, despite suboptimal exposure, showed some evidence for improved synovitis over 12 weeks in patients with active rheumatoid arthritis. FUNDING: GlaxoSmithKline.