SAT0595 DIRECT COMPARISON OF CERTOLIZUMAB PEGOL, ABATACEPT AND BIOSIMILAR INFLIXIMAB IN PATIENTS WITH RHEUMATOID ARTHRITIS TREATED IN ROUTINE CARE. OBSERVATIONAL DATA FROM THE DANISH DANBIO REGISTRY ANALYZED LIKE A RANDOMIZED CLINICAL TRIAL

Background Nationwide Danish guidelines regarding RA patients (pts) initiating bDMARDs are issued approximately annually. For bio-naive pts on concomitant methotrexate (MTX) the recommended drugs (year, recommended compliance) were: certolizumab pegol(CTZ)(2013-2014, 80%)/abatacept(ABA)(2014-2015, 80%)/biosimilar infliximab(CT-P13)(2015-2016, 50%). We hypothesized that the guidelines could be perceived as a surrogate randomization tool where calendar-time rather than patient-specific factors defined choice of bDMARD. Objectives To assess compliance to guidelines (justifying the assumption of surrogate randomization) and compare effectiveness of CTZ/ABA/CT-P13 in Danish RA patients treated according to guidelines. Methods Observational cohort study analyzed like a randomized clinical trial (RCT, intention-to-treat). RA patients were identified in DANBIO. Compliance in each calendar period was defined as number of pts adherent to guideline/numbers of all bio-naive pts initiating bDMARD+MTX. Outcomes were DAS28-remission-rates (at 6 and 12-months) and one-year retention rates, compared across treatments (confounder-adjusted multivariable logistic and Cox regression analyses). Results Compliance to guidelines was 70%/65%/59%, and 776 patients were included (CTZ/ABA/CT-P13: 336/215/225). Baseline characteristics were similar across drugs. DAS28 remission-rates after 6 and 12 months were: 37%/34%/44% and 37%/33%/36%, respectively. Adjusted odd ratios for achieving DAS28 remission were at 6 months: 0.96(95% CI: 0.6;1.5) for ABA and 1.38(0.9;2.1) for CT-P13; 12 months: 0.74(0.5;1.2) and 0.96(0.6;1.5), CTZ reference drug(figure). The adjusted hazard ratios for withdrawal were 1.16(95% CI: 0.84;1.60) for ABA and 0.83 for CT-P13(0.59;1.17)(table). Variables are median (IQR) unless otherwise mentioned. *age, gender, DAS28, HAQ, smoking, CCI Abbreviations: CCI: Charlson Comorbidity Index: CI: confidence intervals, DAS28: Disease Activity Score of 28 joints, HAQ: Health assessment questionnaire Conclusion Compliance to guidelines was high. Direct comparison showed that remission and retention rates were highest for CT-P13, intermediate for certolizumab and lowest for abatacept. Results should, however, be interpreted with caution due to wide CIs and risk of residual confounding. Acknowledgement Thanks to Laura Kocher from Karolinska Institutet, Stockholm, Sweden for help with the figure Disclosure of Interests Kathrine Gron Grant/research support from: BMS, Bente Glintborg Grant/research support from: Biogen, Pfizer, AbbVie, Mette Norgaard: None declared, Frank Mehnert: None declared, Mikkel Ostergaard Grant/research support from: Abbvie, Celgene, Centocor, Merck, Novartis, Consultant for: Abbvie, BMS, Boehringer-Ingelheim, Celgene, Eli Lilly, Hospira, Janssen, Merck, Novartis, Novo, Orion, Pfizer, Regeneron, Roche, and UCB, Speakers bureau: Abbvie, BMS, Boehringer-Ingelheim, Celgene, Eli Lilly, Hospira, Janssen, Merck, Novartis, Novo, Orion, Pfizer, Regeneron, Roche, and UCB, Lene Dreyer Consultant for: MSD, UCB and Janssen Pharmaceuticals, Speakers bureau: MSD, UCB and Janssen Pharmaceuticals, Speakers bureau: UCB, MSD, Eli Lilly and Janssen Pharmaceuticals., Niels Steen Krogh: None declared, Jakob B. Bjorner Employee of: Optum, Merete L. Hetland Grant/research support from: BMS, MSD, AbbVie, Roche, Novartis, Biogen, Pfizer, Consultant for: Eli Lilly, Speakers bureau: Orion Pharma, Biogen, Pfizer, CellTrion, Merck, Samsung Bioepis