Background: Fenebrutinib (GDC-0853, FEN) is an oral, non-covalent, and selective inhibitor of Bruton’s tyrosine kinase (BTK) in clinical development for autoimmune diseases. Objectives: This was a randomized, placebo-controlled, multi-center study to evaluate the efficacy, safety, and pharmacodynamic effects of FEN in patients with moderate-to-severe systemic lupus erythematosus (SLE) activity. Methods: Patients who met SLICC or revised ACR SLE criteria, had ≥1 serologic marker of SLE, SLEDAI ≥8, and were on ≥1 standard of care (SOC) therapy were included; patients with renal or CNS involvement, or exposure to B cell depleting or calcineurin inhibitor therapy were excluded. Patients were randomized to placebo (PBO), FEN 150 mg QD, or FEN 200 mg BID, for 48 weeks. A corticosteroid taper was recommended, with burst and taper permitted from Week 0 (W0) to W12 and W24 to W36. The primary endpoint was SRI-4 at W48. Post hoc subgroup analyses were conducted based on patient baseline disease characteristics. Results: This study enrolled 260 patients, with the majority recruited in Latin America, USA, and Western Europe. At W48, the SRI-4 response rates for FEN 150 mg QD and FEN 200 mg BID were 51% (95% CI: -8.5, 21.2; p value 0.37) and 52% (95% CI: -7.3, 22.4; p value 0.34), respectively, compared to 44% for PBO (Table 1). Post-hoc analysis showed larger responses in subgroups of patients with higher baseline disease activity (Table 1). Safety results were similar between FEN and PBO arms, although more serious adverse events were observed in the FEN 200 mg BID arm. Study discontinuations were balanced across the 3 arms (24-26%). FEN treatment significantly reduced levels of CD19+ B cells, anti-dsDNA autoantibodies, IgG, and a BTK-dependent RNA signature highly expressed in plasmablasts by W48 compared to PBO; C4 levels modestly improved with FEN vs. PBO (Table 2). Table 1. SRI-4 Response (%) at W48 in Primary Analysis and in Post-hoc Patient Subgroups PBO FEN 150 mg QD FEN 200 mg BID SRI-4 Response (%) at W48 44 n =84 51 n =87 52 n =88 SRI-4 Response (%) in Baseline Subgroups At least 1 BILAG A 48 n =42 54 n =39 59 n =46 At least 1 BILAG A and SLEDAI increased DNA binding 37 n =19 53 n =17 65 n =26 SLEDAI arthritis with at least 4 swollen joints 39 n =57 50 n =54 57 n =54 SLEDAI arthritis with at least 4 tender joints 39 n =71 53 n =70 59 n =69 CLASI >=10 21 n =14 36 n =11 31 n =16 Table 2. Key Biomarker Results PBO FEN 150 mg QD FEN 200 mg BID Median (%) Change from Baseline at W48 Plasmablast signature -19.7% n =52 -54.3% * n =53 -51.7% * n =57 CD19 + B cells (cells/µl) -0.50 n =38 -57.0 * n =49 -57.5 * n =48 Anti-dsDNA # (IU/ml) +6.9 n =31 -38.3 * n =36 -75.7 * n =33 Total IgG (g/L) -0.20 n =65 -1.25 * n =64 -1.56 * n =64 C3 (g/L) -0.02 n =65 +0.01 n =67 -0.01 n =66 C4 (g/L) 0.00 n =65 +0.02 * n =67 +0.01 * n =66 #Patients who were positive at baseline (>30 IU/mL) *Denotes significant vs. PBO; Kruskal-Wallis false-discovery rate controlled two sided ( p -value ≤0.05) Conclusion: The primary endpoint of SRI-4 for FEN was not met despite evidence of strong BTK target and pathway inhibition. FEN had an acceptable safety profile. Several disease activity subgroups were suggestive of a greater treatment effect on SRI-4 compared to PBO Disclosure of Interests: David Isenberg Consultant of: Study Investigator and Consultant to Genentech, Richard Furie Grant/research support from: AstraZeneca, Biogen, Consultant of: AstraZeneca, Biogen, Nicholas S. Jones Shareholder of: Genentech/Roche, Employee of: Genentech/Roche, Pascal Guibord Shareholder of: Roche, Employee of: Roche, Joshua Galanter Shareholder of: Genentech/Roche, Employee of: Genentech/Roche, Chin Lee Shareholder of: Genentech/Roche and Eli Lilly, Employee of: Genentech/Roche, Anna McGregor Employee of: Genentech/Roche, Balazs Toth Shareholder of: Genentech/Roche, Employee of: Genentech/Roche, Julie Rae Shareholder of: Genentech/Roche, Employee of: Genentech/Roche, Olivia Hwang Shareholder of: Genentech/Roche, Employee of: Genentech/Roche, Armend Lokku Shareholder of: Roche, Employee of: Roche, Pedro Miranda Consultant of: Study Investigator for Genentech, Viviane de Souza Consultant of: Study investigator for Genentech, Juan Jaller-Raad Consultant of: Study investigator for Genentech, Anna Maura Fernandes Consultant of: Study investigator for Genentech, Rodrigo Garcia Salinas Consultant of: Study investigator for Genentech, Leslie Chinn Shareholder of: Genentech/Roche, Employee of: Genentech/Roche, Michael J. Townsend Shareholder of: Genentech/Roche, Employee of: Genentech/Roche, Alyssa Morimoto Shareholder of: Genentech/Roche, Employee of: Genentech/Roche, Katie Tuckwell Shareholder of: Genentech/Roche, Employee of: Genentech/Roche

Background In phase 1 trials (NCT01534884 and NCT01873443), pharmacokinetic equivalence of CT-P10, biosimilar of rituximab, to innovator rituximab (RTX) was demonstrated. In the phase 3 study, equivalence of PK and efficacy up to week 24 were achieved between CT-P10 and RTX (US and EU sourced)1,2. Objectives To investigate the long-term efficacy, pharmacodynamics, immunogenicity and safety of CT-P10 up to week 48. Methods Patients with rheumatoid arthritis were randomly assigned to CT-P10, US-RTX or EU-RTX, in combination with MTX. The patients received 2 treatment courses at Week 0 and 24, each consisting of 2 infusions of 1000mg study drug at 2-week interval. Results A total of 372 patients were randomised, and 330 patients completed the 2nd course treatment. DAS28 scores through Week 48 were comparable between CT-P10 and US/EU-RTX (Figure), as well as the proportion of ACR responses at Week 48 between the CT-P10 and combined rituximab groups; 81.3% and 79.8% for ACR 20, 55.4% and 53.9% for ACR50, and 31.7% and 33.7% for ACR 70, respectively. B-cell depletion was comparable from after the 1st infusion and up to Week 48. Number (%) of patients with positive anti-drug antibodies in the CT-P10, US-RTX, and EU-RTX was 7 (4.9), 13 (9.4), and 5 (8.6), respectively at Week 48. The safety profile was also similar across groups (Table). Conclusions This phase 3 randomised controlled trial demonstrated the comparability of CT-P10 with two rituximab in terms of efficacy, pharmacodynamics, immunogenicity and safety for 1 year. References Suh CH, et al. 2016 ACR Abstract No. 1634. Yoo DH, et al. 2016 ACR Abstract No. 1635. Disclosure of Interest C.-H. Suh Consultant for: Celltrion, Inc., E. Chalouhi El Khouri Grant/research support from: Celltrion, Inc., P. Miranda Grant/research support from: Celltrion, Inc., F. F Cons Molina Grant/research support from: Celltrion, Inc., P. Shesternya Grant/research support from: Celltrion, Inc., F. Medina-Rodriguez Grant/research support from: Celltrion, Inc., P. Wiland Grant/research support from: Celltrion, Inc., S. Jeka Grant/research support from: Celltrion, Inc., J. Chavez-Corrales Grant/research support from: Celltrion, Inc., T. Linde Grant/research support from: Celltrion, Inc., P. Hrycaj Grant/research support from: Celltrion, Inc., I. Hospodarskyy Grant/research support from: Celltrion, Inc., M. Abello-Banfi Grant/research support from: Celltrion, Inc., J. Jaworski Grant/research support from: Celltrion, Inc., M. Piotrowski Grant/research support from: Celltrion, Inc., W. Park Consultant for: Celltrion, Inc., S. C. Shim Consultant for: Celltrion, Inc., S. J. Lee Employee of: Celltrion, Inc., S. Y. Lee Employee of: Celltrion, Inc., D. H. Yoo Consultant for: Celltrion, Inc.

Background RA is associated with pulmonary comorbidity and lung function decline over time, but longitudinal assessment of pulmonary abnormalities in the context of RA treatment needs further characterisation. Mavrilimumab, an investigational human monoclonal antibody, inhibits GM-CSF by binding to the GM-CSF receptor α subunit. Objectives To investigate the pulmonary safety of mavrilimumab because of the theoretical risk of inhibiting alveolar macrophage function and causing pulmonary alveolar proteinosis (PAP). Methods Pulmonary monitoring included standardised serial pulmonary function testing (spirometry and diffusing capacity of lung carbon monoxide [DLCO]), chest X-rays, assessments of dyspnoea and pulmonary adverse events (AEs) in two randomised, double-blind studies (NCT01706926; NCT01715896) where patients (pts) with moderate to severe RA received mavrilimumab 30, 100 or 150 mg every other week (eow), or placebo and mavrilimumab 100 mg eow or golimumab 50 mg every 4 weeks, respectively. Eligible pts transferred to the open-label extension study (NCT01712399) and received mavrilimumab 100 mg eow. All studies excluded pts with clinically significant uncontrolled pulmonary disease. An Independent Pulmonary Evaluation Committee (IPEC), blinded to treatment, adjudicated pulmonary AEs and lung function abnormalities. Results Mavrilimumab was received by 442 pts with cumulative safety data exposure of approximately 900 pt-yrs and a median (range) exposure time of 2.5 (0.1–3.3) yrs. Baseline (BL) characteristics are shown (Table). Mean dyspnoea (Table), forced expiratory volume in 1 second (FEV1), forced vital capacity (FVC) and DLCO were mostly maintained within 5% of BL values for pts treated with mavrilimumab during the clinical programme. Clinically relevant decreases in predicted FEV1 and FVC (>20% from BL and <80% predicted) were demonstrated by ≤6.2% of pts at any visit (Table); decreases were mostly transient with no apparent trends. Overall, 83 pts (9.24/100 pt-yrs) reported ≥1 pulmonary AE; bronchitis was reported most frequently (34 pts [3.78/100 pt-yrs]); one AE was considered serious and treatment-related (acute bronchitis). The reported pulmonary AE rate was generally stable over time. No suspected or confirmed PAP cases were found by IPEC and no pulmonary-related deaths were reported. Conclusions We believe this is the most comprehensive longitudinal study of pulmonary function in a clinical RA programme. The BL pulmonary function profile indicates that this is not a normal population from a pulmonary health perspective. Mavrilimumab was not associated with substantial decline in pulmonary function or PAP in pts treated up to 3.3 years; its acceptable safety profile advocates initiation of Phase III studies with mavrilimumab. Further studies are now required to fully characterise pulmonary function over time in RA. Acknowledgements Funded by MedImmune. Medical writing support: R Plant, QXV Comms, an Ashfield company, funded by MedImmune. ^Senior author. Disclosure of Interest G. Burmester Consultant for: MedImmune, M. Michaels Employee of: MedImmune, D. Close Employee of: MedImmune, A. Godwood Shareholder of: AstraZeneca, Employee of: MedImmune, K. Middleton Employee of: MedImmune (contracted employment at time of study), P. Miranda Grant/research support from: Amgen, Medimmune, Janssen, Pfizer, Celltrion, Abbott, Sanofi, Actelion, Merck & Co, Boehringer, BMS, Consultant for: Pfizer [Etanecept: Fee less than USD5000], J. Vencovský Consultant for: Pfizer, Elli Lilly, MSD, Novartis, Speakers bureau: Biogen, Pfizer, MSD, Abbvie, Novartis, Boehringer, UCB, BMS, J. Kremer Shareholder of: Corrona, Grant/research support from: Abbvie, Amgen, Genentech, Lilly, Pfizer, Consultant for: Abbvie, Amgen, BMS, Genentech, Lilly, Pfizer, Employee of: Corrona, I. McInnes Grant/research support from: MedImmune [The University of Glasgow is a charity registered in Scotland, charity number SC004401 Grant/research support: Research award to University of Glasgow], Consultant for: MedImmune, M. Albulescu Shareholder of: MedImmune, Employee of: MedImmune, M. Weinblatt Grant/research support from: BMS, UCB, Crescendo Bioscience. Amgen, dxterity, Consultant for: MedImmune, Astra Zeneca, Amgen, Abbvie, BMS, Crescendo bioscience, Lilly, Pfizer, UCB, Roche, novartis

Background Mavrilimumab, a fully human monoclonal antibody targeting the granulocyte-macrophage colony-stimulating factor receptor-α, has demonstrated efficacy and safety in disease-modifying antirheumatic drug (DMARD)-inadequate responder (IR) patients (pts) with rheumatoid arthritis (RA).1 Objectives Few head-to-head studies in tumor necrosis factor antagonist (aTNF)-IR assess alternative aTNFs vs. other agents; this Phase IIb exploratory study (NCT01715896) evaluates the efficacy and safety of mavrilimumab and golimumab in aTNF-IR and DMARD-IR pts. Methods This 24-week study enrolled pts with active RA (28-joint Disease Activity Score [DAS28]–C-reactive protein [CRP]/erythrocyte sedimentation rate ≥3.2); ≥4 swollen joints; inadequate response to ≥1 DMARDs and/or 1–2 aTNFs receiving concomitant methotrexate (MTX; 7.5–25.0 mg/week). Pts received subcutaneous mavrilimumab 100 mg every other week (eow), based on data from the Phase IIa study,2 or golimumab 50 mg alternating with placebo eow. Key endpoints were ACR20/50/70 responses, DAS28–CRP <3.2 and <2.6, Health Assessment Questionnaire Disability Index improvement >0.22 at Week 24 and safety/tolerability. Treatment estimates for mavrilimumab and golimumab are presented with standard errors. Results Pts were randomized to mavrilimumab or golimumab (1:1) (Table). Data for pts receiving mavrilimumab 100 mg eow and golimumab 50 mg at Week 24 are shown for the aTNF-IR, DMARD-IR and overall strata (Table). The most common treatment-emergent adverse events (TEAEs) for mavrilimumab 100 mg and golimumab 50 mg were nasopharyngitis (5.7%, 1.5%), headache (4.3%, 2.9%), upper respiratory tract infection (4.3%, 2.9%), viral upper respiratory tract infection (4.3%, 2.9%), and hepatic enzyme increase (4.3%, 2.9%), respectively. Related serious TEAEs were pneumocystis pneumonia (n=1) and lung disorder (n=1) [both in golimumab-treated pts]. No deaths were reported and no significant pulmonary safety signals were identified. Conclusions In this exploratory study, mavrilimumab 100 mg eow and golimumab 50 mg demonstrated efficacy and an acceptable safety profile in DMARD-IR and aTNF-IR pts. The study was not powered to demonstrate statistical significance between mavrilimumab and golimumab. As mavrilimumab 100 mg eow has previously been shown to be suboptimal compared with 150 mg eow in DMARD-IR pts (EARTH EXPLORER 1),1 additional studies are needed to establish the benefit of a higher dose in pts with moderate to severe RA and inadequate response to aTNF agents. References Burmester G, et al. Arthritis Rheum. 2014;66:S1231 Burmester G, et al. Ann Rheum Dis. 2013;72:1445–1452 Acknowledgement Funded by MedImmune. Editorial assistance: K Alexander, QXV Comms, an Ashfield business, UK Joint senior authors: D. Close and G. Burmester Disclosure of Interest M. Weinblatt Grant/research support from: BMS, UCB, Crescendo Bioscience, Consultant for: MedImmune, Astra Zeneca, Amgen, Abbvie, BMS, Crescendo Bioscience, Lilly, Pfizer, UCB, Roche, I. McInnes Grant/research support from: Research award to University of Glasgow, Consultant for: MedImmune, Astra Zeneca, J. Kremer Shareholder of: Corrona, Grant/research support from: Abbvie, Amgen, Genentech, Lilly, Pfizer, Consultant for: Abbvie, Amgen, BMS, Genentech, Lilly, Pfizer, Employee of: Corrona, P. Miranda Grant/research support from: Clinical trials: Amgen, Medimmune, Janssen, Pfizer, Celltrion, Abbott, Sanofi, Actelion, Merck & Co, Boehringer, BMS, Consultant for: Pfizer for Etanecept (fee less than USD5000), J. Vencovský Consultant for: Pfizer, Servier, Samsumg, Eli lilley, BMS, Novartis, A. Godwood Shareholder of: AstraZeneca, Employee of: MedImmune, M. Albulescu Employee of: MedImmune, D. Close, Employee of: MedImmune, G. Burmester, Consultant for: MedImmune

Background Of patients with RA, ∼40% of do not achieve a minimal acceptable improvement (ACR20) despite modern biologic therapy.1,2,3 Granulocyte-macrophage colony-stimulating factor (GM–CSF) is implicated in RA pathogenesis via myeloid and granulocyte cell lineage activation. In a 12-week Phase IIa study, mavrilimumab, a first-in-class inhibitor of the GM–CSF receptor-α demonstrated a sustained effect via this novel therapeutic pathway in RA.4 Objectives To evaluate the efficacy and safety of mavrilimumab in patients with moderate to severe, adult-onset RA in a 24-week, Phase IIb study. Methods Patients (18–80 yrs; inadequate response to ≥1 DMARDs; DAS28–CRP ≥3.2; ≥4 SJC) receiving MTX were randomized to receive 1 of 3 SC mavrilimumab dosages (150, 100, 30 mg every other week [eow]) or placebo (PBO) plus MTX (7.5–25.0 mg/week). Co-primary endpoints were change in DAS28–CRP (Day 1 to Week 12) and ACR20 response rate (Week 24). Safety and tolerability were measured through assessment of AEs and pulmonary parameters. Results were analyzed using the modified ITT population. Results 326 patients from Europe, South America, and South Africa (mean [SD] age, 51.8 [11.1] yrs; female, 86.5%; mean [SD] DAS28–CRP, 5.8 [0.9]; RF+/anti-CCP+, 81.9%) received mavrilimumab 150, 100 or 30 mg eow or PBO (N=79, 85, 81 and 81, respectively). At Week 12, a statistically significant difference in DAS28–CRP change from baseline (p<0.001) was observed for all dosages of mavrilimumab vs. PBO. At Week 24, a significantly greater percentage of all mavrilimumab-treated patients also met the ACR20 co-primary endpoint vs. PBO (table). A dosage response was observed across several secondary endpoints, with separation from PBO evident as early as Week 1 and first dose. The most common treatment-emergent AEs were headache (7.6%, 4.7%, 6.2%, 2.5%), nasopharyngitis (7.6%, 3.5%, 4.9%, 7.4%) and bronchitis (5.1%, 1.2%, 3.7%, 7.4%) for mavrilimumab 150, 100, 30 mg eow or PBO, respectively. There was no increase in pulmonary AEs for mavrilimumab vs. PBO (6.3%, 3.5%, 6.2% vs. 9.9%). No serious infections were observed in the 100 and 150 mg eow groups. Two cases of pneumonia were observed (one each in mavrilimumab 30 mg eow and PBO groups). There were no deaths or anaphylaxis, and no apparent dosage relationship for AEs. >90% of patients entered a long-term, open-label extension study. Conclusions This Phase IIb study demonstrated the potential benefit of inhibiting macrophage activity via the GM–CSF receptor-α pathway on RA disease activity. The study met both co-primary endpoints with a clear dosage response. Mavrilimumb was well-tolerated over the 24-week study period. References Weinblatt ME, et al. N Engl J Med 1999;340:253–9. Lipsky PE, et al. N Engl J Med 2000;343:1594–602. Weinblatt ME, et al. Arthritis Rheum 2003;48:35–45, Burmester GR, et al. Ann Rheum Dis 2013;72:1445–52. Acknowledgements Funded: MedImmune. Editorial assistance: N Panagiotaki, QXV Communications, UK †Joint senior authors. Disclosure of Interest G. Burmester Grant/research support from: AbbVie, Pfizer, UCB, Roche, Consultant for: AbbVie, BMS, Novartis, MedImmune, MSD, Pfizer, UCB, Roche, Speakers bureau: AbbVie, BMS, Novartis, MSD, Pfizer, UCB, Roche, I. McInnes Grant/research support from: MedImmune (Research award to University of Glasgow), Consultant for: MedImmune, AstraZeneca, J. Kremer Shareholder of: Corrona, Grant/research support from: AbbVie, Amgen, Genentech, Lilly, Pfizer, Consultant for: AbbVie, Amgen, Genentech, Lilly, Pfizer, BMS, Employee of: Corrona, P. Miranda Grant/research support from: MedImmune (to support protocol), M. Korkosz: None declared, J. Vencovsky: None declared, A. Rubbert-Roth Grant/research support from: Pfizer, Chugai, Consultant for: MSD, USB, Abbott, Pfizer, Roche, BMS, Chugai, Speakers bureau: Roche, Pfizer, UCB, E. Mysler Grant/research support from: Medimmune, Roche, BMS, Pfizer, M. Sleeman Shareholder of: AstraZeneca, Employee of: MedImmune, A. Godwood Shareholder of: AstraZeneca, Employee of: MedImmune, M. Albulescu Employee of: MedImmune, D. Close Shareholder of: AstraZeneca, Employee of: MedImmune, M. Weinblatt Grant/research support from: BMS, UCB, Crescendo Bioscience, Consultant for: MedImmune, AstraZeneca, Amgen, Abbvie, BMS, Crescendo Bioscience, Lilly, Pfizer, UCB, Roche

BackgroundMacrophages are pivotal to the pathogenesis of rheumatoid arthritis (RA), and their inflammatory products drive many of the signs and symptoms of disease. Mavrilimumab inhibits macrophage activation and survival via blockade of granulocyte-macrophage colony-stimulating factor receptor-α (GM–CSFR-α) and has demonstrated sustained clinical benefit in patients with RA.1ObjectivesTo examine time to onset of clinical response to mavrilimumab in the EARTH EXPLORER 1 study.MethodsIn a 24-week, Phase IIb study (NCT01706926), patients with adult-onset RA (18–80 years; Disease Activity Score 28 C-reactive protein [DAS28–CRP] >3.2; ≥4 swollen joints; inadequate response to ≥1 disease-modifying antirheumatic drugs [DMARDs]) receiving concomitant methotrexate were enrolled. Patients received mavrilimumab (150, 100, or 30 mg every other week [eow]) or placebo, administered subcutaneously with methotrexate (7.5–25.0 mg/week). Efficacy assessments included DAS28–CRP, American College of Rheumatology (ACR)20/50/70 responses, CRP, erythrocyte sedimentation rate (ESR), swollen joint count (SJC), tender joint count (TJC), and pain assessments.Results326 patients with mean (SD) DAS28–CRP 5.8 (0.9) were randomized to mavrilimumab (150, 100, or 30 mg eow) or placebo (N=79, 85, 81, and 81, respectively). At Week 1 (first assessment), all mavrilimumab dosages demonstrated significant reductions from baseline in DAS28–CRP (p<0.001 vs. placebo), with treatment benefit increasing through Week 12. Significant improvements for mavrilimumab 150 mg eow vs. placebo were demonstrated at Week 1 for CRP, ESR, SJC, TJC, and pain (p=0.003 to p<0.001; Table). Effects of mavrilimumab 150 mg eow were near maximum compared with placebo for CRP, ESR, and pain at Week 1, and were sustained until Week 24. SJC and TJC for mavrilimumab 150 mg eow improved from baseline to Week 1, and continued to improve through Week 12. This improvement was sustained through Week 24.ConclusionsBy targeting activated macrophages via inhibition of GM–CSFR-α, mavrilimumab substantially reduced patients9 RA disease activity early in their treatment, from first dose (Week 1) to final assessment for many patients, as evaluated by several clinical outcomes.ReferencesBurmester GR, et al. Ann Rheum Dis. 2013;72:1445–52.AcknowledgementsFunded: MedImmune Ltd. Editorial assistance: N Panagiotaki, QXV Communications, UKDisclosure of InterestI. McInnes Grant/research support from: MedImmune (Research award to University of Glasgow), Consultant for: MedImmune, AstraZeneca, G. Burmester Grant/research support from: AbbVie, Pfizer, UCB, Roche, Consultant for: AbbVie, BMS, Novartis, MedImmune, MSD, Pfizer, UCB, Roche, Speakers bureau: AbbVie, BMS, Novartis, MSD, Pfizer, UCB, Roche, J. Kremer Shareholder of: Corrona, Grant/research support from: AbbVie, Amgen, Genentech, Lilly, Pfizer, Consultant for: AbbVie, Amgen, Genentech, Lilly, Pfizer, BMS, Employee of: Corrona, P. Miranda Grant/research support from: MedImmune (to support protocol), M. Korkosz: None declared, J. Vencovsky: None declared, A. Rubbert-Roth Grant/research support from: Pfizer, Chugai, Consultant for: MSD, UCB, Abbott, Pfizer, Roche, BMS, Chugai, Speakers bureau: Roche, Pfizer, UCB, E. Mysler Grant/research support from: MedImmune, Roche, BMS, Pfizer, D. Close Shareholder of: AstraZeneca, Employee of: MedImmune, M. Sleeman Shareholder of: AstraZeneca, Employee of: MedImmune, A. Godwood Shareholder of: AstraZeneca, Employee of: MedImmune, M. Albulescu Employee of: MedImmune, M. Weinblatt Grant/research support from: BMS, UCB, Crescendo Bioscience, Consultant for: MedImmune, AstraZeneca, Amgen, Abbvie, BMS, Crescendo bioscience, Lilly, Pfizer, UCB, Roche

Background Long-term treatment of anti-TNF agents in patients with immune mediated disease can increase the risk of active tuberculosis infection (ATBI) by reactivation of latent tuberculosis infection (LTBI) or de novo infection. CT-P13 is a biosimilar of innovator infliximab (INX), approved by the EMA in 2013 based on the studies PLANETAS and PLANETRA. Objectives To identify the risk of positive conversion in the QuantiFERON®-TB Gold in-tube (QTF) test over 2 years in ankylosing spondylitis (AS) and rheumatoid arthritis (RA) patients receiving CT-P13 or INX in the extension studies of PLANETAS and PLANETRA, in 17 countries with various TB incidences. Methods Of 476 enrolled subjects in extension studies of PLANETAS or PLANETRA, all patients had QTF test at baseline and at least one follow-up QTF result after study drug exposure for a 110-week period. Patients with positive QTF result at baseline received prophylaxis before study drug exposure. Patients received either CT-P13 or INX (5mg/kg in AS; 3mg/kg in RA) by usual schedule up to week 102. Countries were divided into four risk groups according to TB incidence, as listed in the 2013 WHO TB report: very low (0–19/100000 population), low (20–49), intermediate (50–124) and high (≥125). In prevalent countries, QTF was performed at weeks 14, 30, 54, 62 and 110. In countries with low risk, QTF was performed at weeks 62 and 110 (or the end of the study visit). To identify the positive conversion in QTF test, patients with negative result at baseline were included in this analysis. Results Among 458 patients with negative QTF at baseline, median dose is 15 (range 9 to 15) and all patients had at least one QTF result after 9th dose exposure. Positive conversion of QTF was observed in 16.6% (76/458) of patients (AS 18.5% [31/168]; RA 15.5% [45/290]). The results showed a tendency of higher positive conversion rate in the region where the TB incidence is high (very low 9.3%, low 11.4%, intermediate 21.7%, high 52.0%). Relative risk (RR) of positive conversion in intermediate and high vs. very low and low TB incidence countries was 2.37 (95% CI 1.57 to 3.59). Non-white ethnicity had higher RR of positive conversion in very low and low TB incidence countries (RR 3.83, 95% CI 2.04 to 7.21), and intermediate and high TB incidence countries (RR 3.11, 95% CI 1.89 to 5.10) as well. In very low, low, intermediate and high incidence countries, positive conversion was observed in 9.3%, 8.5%, 19.6% and 48.0% of patients at week 62, and 0%, 3.0%, 2.1% and 4.0% of patients after week 62 over 2 years, respectively. Amongst patients who had positive conversion, 76.3% (58/76) of patients received prophylactic TB medication. Conclusions To reduce TB incidence in patients receiving anti-TNF agents, appropriate screening and serial QTF tests at least during the first two years of treatment are necessary to reduce and minimize the risk of TB in patients residing in intermediate and high TB endemic regions. Disclosure of Interest W. Park Grant/research support: CELLTRION, Inc., Consultant for: CELLTRION, Inc., Speakers bureau: CELLTRION, Inc., D. Yoo Grant/research support: CELLTRION, Inc., Consultant for: CELLTRION, Inc., Speakers bureau: CELLTRION, Inc., P. Hrycaj Grant/research support: CELLTRION, Inc., N. Prodanovic Grant/research support: CELLTRION, Inc., P. Miranda Grant/research support: CELLTRION, Inc., E. Ramiterre Grant/research support: CELLTRION, Inc., A. Baranauskaite Grant/research support: CELLTRION, Inc., P. Wiland Grant/research support: CELLTRION, Inc., Y.-A. Lee Grant/research support: CELLTRION, Inc., S. Lee Employee of: CELLTRION, Inc. DOI 10.1136/annrheumdis-2014-eular.3492

Background CT-P13 is a biosimilar of innovator infliximab (INX), approved by the European Medicines Agency in 2013. Clinical data up to 1 year from PLANETRA have been reported at EULAR 20131 Objectives To compare the radiographic progression between CT-P13 and INX treatment and to assess its association with anti-drug antibody (ADA) and clinical disease activity in active rheumatoid arthritis (RA) patients, who participated in the PLANETRA study. Methods Radiographs obtained at baseline and week 54 were evaluated with the “paired review” method and the evaluation was performed by two independent readers without knowing the time point of the radiographs. The individual component scores of the joint damage progression (JDP) were calculated according to the van der Heijde modification of the Sharp scoring system. The analysis was performed to demonstrate comparability in JDP between two treatment groups using Student9s t-test and elucidate an association of JDP with ADA and clinical parameters such as ACR20, IgM rheumatoid factor (RF), and anti-CCP. Results Among the 606 patients, 336 patients had radiographs both at baseline and week 54. The mean change from baseline JDP was similar between CT-P13 and INX treatment groups with respect to total Sharp score (TSS), joint space narrowing (JSN) score and erosion score at week 54 (CT-P13, 1.0/0.4/0.7; INX, 0.6/0.7/0.0, respectively). At week 54, a higher progression of TSS could be observed in the ADA positive patients (CT-P13, 1.1; INX, 1.2) compared with the ADA negative patients (CT-P13, 0.9; INX, 0.0), but the scores were comparable between two groups. The ACR20 responders tended to show a similar progression of TSS between two treatment groups (CT-P13, 0.6; INX, 0.5) at week 54. The mean change of TSS in ACR20 non-responders from baseline to week 54 were also comparable between the two treatment groups (CT-P13, 2.0; INX, 0.9). ACR20 responders showed less progression of TSS compared with non-responders in both treatment groups but these results were statistically not significant with respect to response status and treatment groups. At week 54, the progression of TSS was similar between treatment groups for each of the negative and positive baseline IgM RF subgroups (CT-P13, -0.5/1.4; INX, -0.2/0.7, respectively). The patients who belong to the negative and positive baseline anti-CCP subgroups showed also a comparable radiographic progression between treatment groups (CT-P13, 0.5/1.0; INX, 0.2/0.7, respectively). In both CT-P13 and INX, there was less progression of TSS in the patients who had negative baseline IgM RF or anti-CCP but the changes were statistically not significant. Conclusions Patients treated with CT-P13 showed a comparable radiographic progression as compared to those treated with INX at week 54. The ADA and clinical parameters such as ACR20, IgM RF, and anti-CCP showed tendency of association with radiographic progression. References Yoo DH, et al. Ann Rheum Dis 2013;72(S3):73 Disclosure of Interest : D. Yoo Grant/research support: CELLTRION, Inc., Consultant for: CELLTRION, Inc., Speakers bureau: CELLTRION, Inc., W. Park Grant/research support: CELLTRION, Inc., Consultant for: CELLTRION, Inc., Speakers bureau: CELLTRION, Inc., P. Miranda Grant/research support: CELLTRION, Inc., M. Piotrowski Grant/research support: CELLTRION, Inc., E. Ramiterre Grant/research support: CELLTRION, Inc., S. Shevchuk Grant/research support: CELLTRION, Inc., A. Baranauskaite Grant/research support: CELLTRION, Inc., S. Lee Employee of: CELLTRION, Inc., U. Müller-Ladner Speakers bureau: CELLTRION, Inc. DOI 10.1136/annrheumdis-2014-eular.3056

Background Blocking TNF alpha (TNFα) with monoclonal antibodies (mAbs) has been successful in the treatment of rheumatoid arthritis. However primary and secondary resistances are frequent and impose treatment changes. Active immunization with a TNF-Kinoid that safely induces self polyclonal anti-TNFα antibodies (Abs) could be an alternative to anti-TNFα mAbs Objectives We evaluated the immunogenicity and safety of TNF-K in patients with rheumatoid arthritis and secondary resistance to TNF blockers. Methods TNFα-Kinoid (TNF-K, Neovacs SA, Paris, France) is an immunotherapeutic composed of recombinant human TNFα conjugated to KLH, inactivated and adjuvanted with ISA-51 emulsion. 40 patients with active rheumatoid arthritis (DAS28≥3.2) with history of positive clinical response to TNF-blockers followed by secondary failure (35% IFX,30% ADA, 42.5% ETA) were enrolled in a double-blind, placebo-controlled, phase 2 study to evaluate three different intramuscular doses of TNF-K (90, 180, 360 mcg) and two immunization schedules (D0 and 28 or D0, 7 and 28). Immune responses were evaluated through titration of anti-TNFα and anti-KLH Abs. Clinical response was evaluated by the ACR and EULAR core set response. Results No related serious adverse event has been reported. Few minor transient local and systemic reactions have been recorded following immunization. Anti-TNFα Abs were induced in 50%, 75% and 91% of patients at 90 mcg, 180 mcg and 360 mcg, respectively. 100% of patients with three injections of 180 or 360 mcg had immunogenic response against 67% in two injections groups. Among the 21 patients who developed anti-TNF Abs, 48% present a moderate to good response according to EULAR score as opposed to only 31% of the 16 patients without Abs. A mean decrease of -14% of the C reactive protein level is measured in patients with Abs while in patients without Abs, the mean CRP level increased by 5%The mean decrease of the DAS28 score is -0.8 points in patients with Abs while it is only -0.5 in patients without Abs. Conclusions Active immunization with TNFα kinoid to induce a polyclonal, self-anti-TNFα antibody response is safe and immunogenic. A clinical and biological response were observed in some patients included in this preliminary phase 2 study and further studies are needed to confirm this new approach in RA. Disclosure of Interest P. Durez: None Declared, P. Miranda: None Declared, A. Toncheva: None Declared, A. Berman: None Declared, O. L. Rillo: None Declared, Y. Boutsen: None Declared, T. Kehler: None Declared, E. Mociran: None Declared, L. Soto Saez: None Declared, B. Fautrel: None Declared, X. Mariette: None Declared, P. Sokalov: None Declared, E. Lucero: None Declared, T. Vlak: None Declared, S. Grazio: None Declared, K. Mastrovic: None Declared, R. Chirieac: None Declared, G. Grouard-Vogel Employee of: Neovacs, O. Dhellin Employee of: Neovacs, S. Ouary Employee of: Neovacs, B. Fanget Employee of: Neovacs, P. Vandepapelière Employee of: Neovacs, M.-C. Boissier Consultant for: Neovacs

Background CT-P13 was developed as a biosimilar product to infliximab (Remicade®), a chimeric monoclonal antibody approved in the European Union in 1999 for the treatment of rheumatoid arthritis (RA), ankylosing spondylitis (AS), Crohn’s disease, ulcerative colitis, psoriasis, and psoriatic arthritis. Objectives To compare the pharmacokinetic (PK) profile of CT-P13 with that of infliximab at steady state in terms of area under the concentration-time curve over a dosing interval (AUCτ) and observed maximum serum concentration (Cmax,ss), and evaluate the efficacy and overall safety of both treatments in patients with AS. Methods Two hundred fifty patients with active AS were randomized 1:1 to receive either CT-P13 or infliximab (5 mg/kg, 2-hour IV infusion per dose) at weeks 0, 2, and 6 (dose-loading phase) and at weeks 14, 22, and 30 (maintenance phase). Ratios of geometric means of primary PK parameters (AUCτ and Cmax,ss) from the 2 treatment arms between weeks 22 and 30 were subjected to ANCOVA analysis at 90% confidence intervals (CIs). Efficacy measures (including ASAS20 and ASAS40), and safety parameters (including the incidence of adverse events [AEs]) were also evaluated. This report presents PK, efficacy, and safety results up to week 30 (as approved by the European Medicines Agency). Results The mean (% CV) AUCτ was 34855.45 (34.3%) μg-h/mL and 34688.71 (45.4%) μg-h/mL in the CT-P13 and infliximab arms, respectively. The mean (% CV) Cmax,ss was 153.52 (27.6%) μg/mL and 150.39 (26.9%) μg/mL in the CT-P13 and infliximab arms, respectively. The ratio (%) between the geometric means of the AUCτ and Cmax,ss values in the CT-P13 and infliximab arms were 104.1% (90% CI 93.9% to 115.4%) and 101.5% (90% CI 94.6% to 108.9%), respectively, between weeks 22 and 30, indicating PK equivalence in terms of AUCτ and Cmax,ss. Secondary parameters at week 30 were also comparable, including ASAS20 and ASAS40 response rates (70.5% for CT-P13 vs 72.4% for infliximab and 51.8% vs 47.4%, respectively). AEs considered by the investigators to be related to study treatment were reported in 57 (44.5%) patients and 58 (47.5%) patients in the CT-P13 and infliximab arms, respectively. Related AEs due to infection were reported for 24/128 (18.8%) patients and 22/122 (18.0%) patients in the CT-P13 and infliximab treatment groups, respectively. AEs due to infusion reactions considered related to study drug were reported in 5 patients in the CT-P13 arm, and 6 patients in the infliximab arm. Tuberculosis was reported in 2 patients in the CT-P13 arm and in 1 patient in the infliximab arm. Conclusions CT-P13 and infliximab are equivalent in terms of AUCτ and Cmax,ss in patients with AS. In addition, CT-P13 was well tolerated, with an efficacy and safety profile comparable to that of infliximab up to week 30. Disclosure of Interest W. Park: None Declared, P. Hrycaj: None Declared, V. Kovalenko: None Declared, P. Miranda: None Declared, S. Gutierrez-Ureña: None Declared, Y. Lee: None Declared, M. Lim: None Declared, C. Ahn: None Declared, H. Kim Employee of: Celltrion, D. Yoo: None Declared, J. Braun: None Declared

To examine physical function, health-related quality of life(HRQOL) and work productivity in patients enrolled from Latin American countries (Argentina,Chile and Mexico) in Phase III clinical trials for golimumab(GLM) in rheumatoid arthritis(RA). Active RA patients not previously treated with methotrexate(MTX)(GO-BEFORE,N=637) or with inadequate response to MTX(GO-FORWARD,N=444) were randomized to SC GLM(50 or 100mg)+MTX or PBO+MTX, q4wks.At wk24(GO-FORWARD) or wk52(GO-BEFORE), PBO+MTX group switched to GLM 50mg+MTX. Physical function was assessed using HAQ(0-3). HRQOL was assessed using SF-36 PCS(0-100) and SF-36 MCS(0-100). Impact of disease on work productivity was assessed using a productivity VAS (0-10). Clinically meaningful improvement was defined as improvement of ≥0.25 point in HAQ, or ≥5 points in SF-36 PCS and MCS. At baseline, both MTX naive (N=96) and MTX experienced (N=56) RA patients enrolled in Latin American region experienced moderate to severe physical disability (mean HAQ score of 1.60 to 1.75) and impaired HRQOL (mean PCS of 30.0 to 30.3 and mean MCS of 9.4 to 42.6). The impact of RA on productivity was severe (mean VAS score of 6.3-6.7). Patients treated with GLM (50 or 100 mg)+MTX had significantly greater mean improvement than PBO+MTX group in HAQ (0.87 vs. 0.56, p=0.01), PCS (12.44 vs. 6.93, p<0.01) and work productivity (-3.69 vs. -2.25, p<0.01) at wk 24, and greater proportions of patients in GLM+MTX group than PBO+MTX achieved clinically meaningful improvement in HAQ (84.04% vs. 65.45%, p=0.01), PCS (74.47% vs. 49.09%, p<0.01) and MCS (45.74% vs. 41.82%, p=0.73). Similar results were observed in patients who were MTX naïve and experienced although the magnitudes of improvements were greater in MTX naïve patients than MTX-experienced patients. The improvements were sustained over wk52 and 104. MTX naïve and MTX experienced RA patients from Latin America treated with GLM demonstrated improved physical function, HRQOL, and work productivity.

Background CT-P13 was developed as a biosimilar product to infliximab (Remicade®), a chimeric monoclonal antibody approved in the European Union in 1999 for the treatment of rheumatoid arthritis (RA), ankylosing spondylitis, Crohn’s disease, ulcerative colitis, psoriasis, and psoriatic arthritis. Objectives To compare the clinical efficacy and overall safety of CT-P13 with those of infliximab in patients with active RA. Methods Six hundred six patients with active RA despite previous treatment with disease-modifying anti-rheumatic drugs including methotrexate were randomized 1:1 to receive either CT-P13 or infliximab (3 mg/kg, 2-hour IV infusion per dose) plus methotrexate and folic acid. The treatment period consisted of a dose-loading phase (weeks 0, 2, and 6) and a maintenance phase (weeks 14, 22, and 30). The primary endpoint was the proportion of patients achieving 20% improvement in ACR20 at week 30. The exact binomial test with 95% confidence intervals (CIs) for ACR20 within a margin of ±15% was used to define equivalence between the 2 treatments. Secondary efficacy and safety endpoints (including ACR50/70; frequency of adverse events [AEs]) were also evaluated. This report presents results up to study week 30 (as approved by the European Medicines Agency). Results At week 30, ACR20 response rates were 60.9% for CT-P13 vs 58.6% for infliximab (2% difference; 95% CI -6% to 10%) in the intent-to-treat population, and 73.4% vs 69.7% (4% difference; 95% CI: -4% to 12%) in the per-protocol population. Outcomes for the secondary efficacy and safety endpoints were also comparable as follows. ACR50 rates in the per-protocol population were 42.3% vs 40.6% (2% difference; 95% CI -7% to 10%), and ACR70 rates were 20.2% vs 17.9% (2% difference; 95% CI -5% to 9%) in the CT-P13 and infliximab arms, respectively, all indicating equivalence in clinical efficacy at week 30. AEs considered by the investigators to be related to study treatment were reported for 106 (35.2%) patients and 108 (35.9%) patients in the CT-P13 and infliximab arms, respectively. Related AEs due to infection were reported in 46 (15.3%) patients and 51 (16.9%) patients in the CT-P13 and infliximab arms, respectively. 15 (5%) CT-P13-treated and 17 (6%) infliximab-treated patients experienced at least 1 infusion reaction. Tuberculosis was reported in 3 patients in the CT-P13 arm and in 1 patient in the infliximab arm. Conclusions CT-P13 and infliximab are equivalent in terms of ACR20 in patients with RA. In addition, CT-P13 was well tolerated, with an efficacy and safety profile comparable to that of infliximab up to week 30. Disclosure of Interest D. Yoo: None Declared, P. Miranda: None Declared, M. Piotrowski: None Declared, E. Ramiterre: None Declared, V. Kovalenko: None Declared, N. Prodanovic: None Declared, M. Tee: None Declared, S. Gutierrez-Ureña: None Declared, R. Jimenez: None Declared, O. Zamani: None Declared, S. Lee: None Declared, H. Kim Employee of: Celltrion, W. Park: None Declared, U. Müller-Ladner: None Declared

Objective: We compared the efficacy of etoricoxib 30 mg to placebo and ibuprofen 2400 mg for the treatment of osteoarthritis (OA) of the hip and knee.Design: In this 12-week, randomized, double-blind, placebo- and active-comparator-control led trial, 548 patients (median age 63 years) with OA of the hip or knee were randomized to receive placebo, etoricoxib 30 mg q.d., or ibuprofen 800 mg t.i.d. Demonstration of etoricoxib's efficacy vs placebo and comparison of its efficacy to ibuprofen were assessed using three co-primary endpoints: Western Ontario and McMaster's University Osteoarthritis Index (WOMAC) Pain Subscale (WOMAC-PS); WOMAC Physical Function Subscale (WOMAC-PFS); and Patient Global Assessment of Disease Status (PGADS). Each primary endpoint utilizes a 0-100 mm visual analog scale. To demonstrate comparable efficacy of etoricoxib vs ibuprofen, the 95% confidence intervals (Cls) for the difference in the least squares (LS) mean change over 12 weeks for all three co-primary endpoints had to fall within 10 mm. Safety and tolerability data were collected throughout the study.Results: Mean baseline values for the three co-primary endpoints ranged from 62.52 to 70.14 mm. Both etoricoxib and ibuprofen demonstrated superior (P <= 0.002) efficacy for all primary endpoints. The LS mean (mm) changes (95% Cl) over 12 weeks for etoricoxib and ibuprofen, respectively, compared to placebo were given as follows: WOMAC-PS: -11.66 (-16.31, -7.01) and -7.62 (-12.30, -2.94); WOMAC-PFS: -10.15 (-14.74, -5.57) and -7.23 (-11.85, -2.61); PGADS: -11.65 (-16.81, -6.50) and -8.11 (-13.30, -2.92). The efficacy of etoricoxib 30 mg was comparable to ibuprofen 2400 mg. All treatments were similarly well tolerated.Conclusion: Treatment with etoricoxib 30 mg q.d. provides superior efficacy vs placebo and comparable clinical efficacy vs ibuprofen 2400 mg (800 mg t.i.d.) for the treatment of OA of the hip and knee. (c) 2007 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.