Background ACTION (NCT02109666) is an ongoing study to provide prospective, real-world data on IV abatacept (ABA) retention in RA. Prognostic factors for IV ABA retention were previously identified in a 12-month interim analysis of patients (pts) enrolled May 2008–Jan 2011 who failed ≥1 prior biologic DMARD.1 Objectives To compare rates and identify predictors of IV ABA retention by treatment line (biologic naïve vs biologic failure) in a 12-month interim analysis. Methods ACTION is a 2-year, observational study of pts with moderate-to-severe RA who initiated IV ABA in Europe and Canada. Enrolment periods: May 2008–Dec 2010 (biologic naïve/failed prior biologics); Sept 2010–Dec 2013 (biologic naïve); Oct 2011–Dec 2013 (failed prior biologics). Data cut-off was May 2015. Crude retention rates were estimated by Kaplan–Meier analysis and compared using a log-rank test. Known risk factors and factors with significance in univariate models (p≤0.20) and no collinearity were entered into a multivariate Cox proportional hazards regression model. Factors with p≤0.10 after backward selection were retained. Efficacy assessments at 12 months included EULAR response and DAS28 (CRP) score. Results 2364 pts were enrolled and 2350 were evaluable: 674 (28.7%) were biologic naïve and 1676 (71.3%) had failed ≥1 prior biologic: 728/1676 (43.4%) failed 1 biologic, 948/1676 (56.6%) ≥2 biologics; 1605/1676 (95.8%) had failed anti-TNFs. Biologic-naïve pts had similar baseline characteristics vs biologic-failure pts, except for shorter disease duration, fewer erosions in combination with RF or anti-citrullinated protein antibody (ACPA) positivity, prior conventional DMARDs failed and ABA monotherapy, and more co-morbidities (COPD, cardiac disorders and neoplasms). Crude retention rates over 12 months were statistically higher in biologic-naïve (78.1%; 95% CI 74.7, 81.2) vs biologic-failure pts (69.9%; 95% CI 67.6, 72.1; p<0.001). Predictors of higher retention for biologic-naïve pts are shown (Figure). Neoplasms and psychiatric disorders were associated with lower retention and need investigation. In both treatment lines, RF/ACPA double positivity was associated with higher retention vs double negativity after stratifying for erosion status, but was significant only in pts with erosive disease at baseline. A numerically greater proportion of biologic-naïve vs biologic-failure pts had a good/moderate EULAR response (83.8 vs 73.3%) and DAS28 (CRP) <2.6 (44.5 vs 35.0%); unadjusted. Conclusions These data confirm the high levels of IV abatacept retention in the real world, particularly when used in earlier lines of treatment. Poor prognostic factors in RA, including RF/ACPA double positivity combined with erosions at baseline, positively impact retention, independently of treatment line. References Nüßlein HG, et al. BMC Musculoskeletal Disorders 2015;16:176. Disclosure of Interest R. Alten Grant/research support from: Bristol-Myers Squibb, Consultant for: Bristol-Myers Squibb, Speakers bureau: Bristol-Myers Squibb, H. Nüßlein Consultant for: Bristol-Myers Squibb, AbbVie, Celgene, Chugai, UCB, Pfizer, MSD, Novartis, Roche, Speakers bureau: Bristol-Myers Squibb, AbbVie, Celgene, Chugai, UCB, Pfizer, MSD, Novartis, Roche, M. Galeazzi: None declared, H. M. Lorenz Consultant for: AbbVie, Bristol-Myers Squibb, Roche-Chugai, UCB, MSD, GSK, Sobi, Medac, Novartis, Janssen-Cilag, AstraZeneca, Pfizer, Actelion, Speakers bureau: AbbVie, Bristol-Myers Squibb, Roche-Chugai, UCB, MSD, GSK, Sobi, Medac, Novartis, Janssen-Cilag, AstraZeneca, Pfizer, Actelion, X. Mariette Grant/research support from: Biogen, GSK, Pfizer, Speakers bureau: Bristol-Myers Squibb, GSK, Pfizer, sanofi, UCB, A. Cantagrel Grant/research support from: UCB, Pfizer, Consultant for: AbbVie, Bristol-Myers Squibb, Chugai, Lilly, MSD, Novartis, Pfizer, Roche, M. Chartier Consultant for: Bristol-Myers Squibb, Y. Elbez Employee of: Bristol-Myers Squibb, C. Rauch Shareholder of: Bristol-Myers Squibb, Employee of: Bristol-Myers Squibb, M. Le Bars Shareholder of: Bristol-Myers Squibb, Employee of: Bristol-Myers Squibb