T. Huizinga

Four-month metacarpal bone mineral density loss predicts radiological joint damage progression after 1 year in patients with early rheumatoid arthritis: exploratory analyses from the IMPROVED study

Aim To assess whether in early (rheumatoid) arthritis (RA) patients, metacarpal bone mineral density (BMD) loss after 4u2005months predicts radiological progression after 1u2005year of antirheumatic treatment. Methods Metacarpal BMD was measured 4 monthly during the first year by digital X-ray radiogrammetry (DXR-BMD) in patients participating in the IMPROVED study, a clinical trial in 610 patients with recent onset RA (2010 criteria) or undifferentiated arthritis, treated according to a remission (disease activity score<1.6) steered strategy. With Sharp/van der Heijde progression ≥0.5 points after 1u2005year (yes/no) as dependent variable, univariate and multivariate logistic regression analyses were performed. Results Of 428 patients with DXR-BMD results and progression scores available, 28 (7%) had radiological progression after 1u2005year. Independent predictors for radiological progression were presence of baseline erosions (OR (95% CI) 6.5 (1.7 to 25)) and early DXR-BMD loss (OR (95% CI) 1.5 (1.1 to 2.0)). In 366 (86%) patients without baseline erosions, early DXR-BMD loss was the only independent predictor of progression (OR (95% CI) 2.0 (1.4 to 2.9)). Conclusions In early RA patients, metacarpal BMD loss after 4u2005months of treatment is an independent predictor of radiological progression after 1u2005year. In patients without baseline erosions, early metacarpal BMD loss is the main predictor of radiological progression.

OP0048 Survival in Early Rheumatoid Arthritis Patients After 10 Years of Targeted Treatment

Background Recent studies showed diverging results about mortality trends in patients with rheumatoid arthritis (RA). Objectives Our aim was to determine survival after 10 years of treat-to-target therapy in patients with early RA, to compare these survival rates with the general population and to define risk factors for mortality during the 10 years duration of the BeSt study. Methods The BeSt study enrolled 508 Dutch patients with early active RA (1987 criteria) who were randomized to: sequential monotherapy, step-up therapy, initial combination including either prednisone or infliximab. During 10 years, all patients were treated-to-target, aiming at a disease activity score (DAS) ≤2.4. Kaplan-Meier curves and the log-rank test were used to compare survival rates in the four treatment strategies. Standardized mortality ratios (SMR) were used to compare the BeSt population to the general Dutch population, matched by age, gender and calendar year. Cox regression analysis was used to calculate hazard ratios (HR) to determine baseline risk factors for increased mortality in the BeSt population. Time dependent Cox regression analysis was used to study DAS during follow-up as a risk factor for mortality. Results During 10 years, 72 of 508 patients died at a mean age of 75 years. No difference in survival was observed between the treatment strategies (p=0.805) (figure), with 16/126, 15/121, 21/133 and 20/128 deaths in arm 1 to 4, respectively. Based on the general Dutch population, 62 deaths were expected and 72 deaths occurred, resulting in an overall SMR of 1.16 (95% confidence interval, CI 0.92 – 1.46). Comparing the general population to each of the treatment strategies resulted in a SMR (95% CI) of 1.00 (0.61–1.64), 1.02 (0.61–1.69), 1.30 (0.85–1.99) and 1.32 (0.85–2.04) in arm 1 to 4, respectively. In the BeSt population, baseline age (HR 1.13, 95% CI 1.10-1.16), male gender (HR 1.78, 95% CI 1.06-2.99), smoking at baseline (HR 5.19, 95% CI 3.08-8.75) and health assessment questionnaire at baseline (HR 1.89, 95% CI 1.29-2.76) were associated with an increased risk of mortality. Randomization arm was not associated with an increased risk of mortality (arm 1 as reference category; arm 2 HR 0.99, 95% CI 0.49–2.00; arm 3 HR 1.27, 95% CI 0.66–2.44; arm 4 HR 1.25, 95% CI 0.65–2.41). DAS over time showed a trend (HR 1.23, 95% CI 0.96–1.58). Conclusions After 10 years of continued tight controlled treatment in patients with rheumatoid arthritis in the BeSt study, the survival rate was comparable to the general Dutch population, without significant differences between the treatment strategies. Higher age, male gender, smoking and worse functional ability were associated with an increased mortality risk within our study population, as in the general population. These results suggest that treatment targeted at DAS ≤2.4 prevents excess mortality during 10 years after diagnosis, and that the medication used in these strategies does not increase mortality. Disclosure of Interest I. Markusse: None declared, L. Dirven: None declared, H. Han: None declared, K. Ronday: None declared, I. Speyer: None declared, P. Kerstens: None declared, W. Lems: None declared, T. Huizinga: None declared, C. Allaart Grant/research support from: The study was designed by the investigators and supported by a government grant from the Dutch Insurance Companies, with additional funding from Schering-Plough B.V. and Janssen B.V. Data collection, trial management, data analysis and preparation of the manuscript were performed by the authors.

FRI0055 Good Adherence of Rheumatologists to A 10 Year Treat-To-Target Protocol for Patients with Recent-Onset Rheumatoid Arthritis (The Best Study)

Background Treat-to-target therapy (T2T) is recommended in the treatment of rheumatoid arthritis (RA), but appears not yet fully implemented in daily practice. Objectives To evaluate adherence to a T2T strategy and to determine the rheumatologists opinion about the study protocol steps, suppression of RA and performance of disease activity score (DAS). Methods During 10 year follow-up rheumatologists in 20 clinics treated 508 early RA patients enrolled in the BeSt study as part of their daily outpatient practice. The protocol required treatment adjustments based on 3-monthly DAS measurements, with treatment intensification aiming to achieve low disease activity (DAS ≤2.4) and tapering and discontinuation steps when the target was met and sustained. Rheumatologists filled in questionnaires about satisfaction with level of RA suppression, agreement with the next protocol step and the DAS as representation of disease activity. We checked adherence to the treatment protocol over time and investigated if questionnaire responses and hypothetical conditions (see results) where rheumatologists might disagree with the DAS were associated with protocol violations, using generalized estimating equations. Results Protocol adherence decreased over time from almost 100% in year 1 to ±60% per visit in year 10, with an average of 79% per visit over 10 years. Rheumatologists mostly agreed with DAS (80-90% of visits over time), and were satisfied with treatment steps in the protocol (75-90% over time) and the level of RA suppression (85-90% over time). Although even then the protocol was mostly followed, the odds for PV increased when the rheumatologist felt the DAS underestimated (Odds Ratio, OR 2.1, 95% confidence interval, CI 1.7-2.5) or overestimated (OR 2.0, 95% CI 1.7-2.3) true disease activity, when the rheumatologist disagreed with the study protocol (OR 2.5, 95% CI 2.1-2.9) or was dissatisfied with the level of RA suppression (OR 1.4, 95% CI 1.2-1.6). The odds for PV were also higher in some specific conditions: when swollen joint count (SJC) was ≤1 and erythrocyte sedimentation rate (ESR) was ≥28 (OR 1.3, 95% CI 1.1-1.5), or when the visual analogue scale of global health of the patient (VASpt) was ≥20 millimetre (mm) higher than the VAS of the physician (VASph) (OR 1.3, 95%CI 1.2-1.5). Other specified conditions as SJC ≤1 and tender joint count ≥2, or SJC ≤1 and VASpt ≥20 mm, or VASph ≥20 mm higher than the VASpt, did not increase the risk of PV. Conclusions Rheumatologists adhered well to the treat-to-target strategy of the BeSt protocol, although over time adherence decreased. This may be because in later years, physicians started taking part in the study who had not taken part in designing the protocol. Rheumatologists mostly agreed with how the DAS represented disease activity and with the treatment protocol steps, and were satisfied with the level of RA suppression. However, when they did not agree or were dissatisfied, protocol deviation was more likely. These results prove the feasibility of DAS targeted treatment, steering at low disease activity: rheumatologists feel the DAS performs well in estimating disease activity and are willing to comply with DAS steered therapy adjustments. Disclosure of Interest I. Markusse: None declared, L. Dirven: None declared, H. Han: None declared, K. Ronday: None declared, P. Kerstens: None declared, W. Lems: None declared, T. Huizinga: None declared, C. Allaart Grant/research support: The study was designed by the investigators and supported by a government grant from the Dutch Insurance Companies, with additional funding from Schering-Plough B.V. and Janssen B.V. Data collection, trial management, data analysis and preparation of the manuscript were performed by the authors. DOI 10.1136/annrheumdis-2014-eular.2905

FRI0155 Long-term cost-utility analysis of treatment strategies in patients with recent-onset rheumatoid arthritis – 5 year follow-up data from the best study

Objectives To evaluate 5-year costs and quality-adjusted life years (QALYs) of treatment strategies for patients with recent-onset active rheumatoid arthritis (RA). Methods Patients (n=508) were randomly allocated to one of the following four treatment strategies: 1. sequential monotherapy, 2. step-up combination therapy, 3. initial combination therapy with prednisone, or 4. initial combination therapy with infliximab. The current analysis includes the outcome measures that were significantly different in the 2-year analysis: study medication, QALYs based on the EQ5D and SF6D, and paid labour. Data were analyzed using multiple imputation, combined with ANOVA and net benefit analysis. In separate analyses, both the friction cost method and the human capital approach were used to value productivity. Results Treatment patterns continued to differ during the follow-up period. In the fifth year, the percentages of patients receiving infliximab in groups 1-4 were 20%, 6%, 13%, and 22%, respectively (P=0.002). Costs of study medication were largely determined by the use of infliximab and over the five-year period amounted to 14000, 5000, 9000, and 26000 euro (P<0.001). QALYs were significantly better in group 4 during the first two years (P<0.04). In later years, there were no significant differences between the groups. Total QALYs in groups 1-4 (ideally 5.0) amounted to 3.08, 2.98, 3.07, and 3.28 according to the EQ5D (P=0.06) and 3.26, 3.19, 3.24, and 3.34 according to the SF6D (P=0.16). Differences in paid labour decreased over time, but remained in favour of strategy 4 (P<0.06). Using the friction cost method to value productivity, the cost-utility ratio for group 4 against the next best alternative was estimated at €76,000 per QALY. Using the human capital approach to value productivity, the savings on paid labour in group 4 exceeded the higher medication costs in group 4. Conclusions Initial combination therapy with infliximab for patients with recent-onset active RA resulted in significantly better 2-year quality of life than the other strategies. After 2 years, the outcome differences between the strategies decreased, but differences in medication costs and paid labour remained. Using the friction cost method to value productivity, the higher costs of initial combination therapy with infliximab are generally considered too high. Giving full weight to the sustained productivity using the human capital approach, the higher costs of initial combination therapy with infliximab were more than compensated. Disclosure of Interest W. van den Hout Grant/Research support from: Dutch College of Health Insurances, Centocor Inc. and MSD (formerly Schering-Plough), N. Klarenbeek Grant/Research support from: Dutch College of Health Insurances, Centocor Inc. and MSD (formerly Schering-Plough), L. Dirven Grant/Research support from: Dutch College of Health Insurances, Centocor Inc. and MSD (formerly Schering-Plough), P. Kerstens Grant/Research support from: Dutch College of Health Insurances, Centocor Inc. and MSD (formerly Schering-Plough), T. Huizinga Grant/Research support from: Dutch College of Health Insurances, Centocor Inc. and MSD (formerly Schering-Plough), W. Lems Grant/Research support from: Dutch College of Health Insurances, Centocor Inc. and MSD (formerly Schering-Plough), C. Allaart Grant/Research support from: Dutch College of Health Insurances, Centocor Inc. and MSD (formerly Schering-Plough)

FRI0120 Prevalence and predictors of vertebral fractures after 5 years of disease activity steered treatment in patients with early active rheumatoid arthritis

Background Vertebral fractures (VFs) are more common in patients with rheumatoid arthritis (RA) compared to the general population. It is suggested that an appropriate control of disease, generally more effectively achieved with disease activity score (DAS)-steered treatment strategies, may prevent the development of vertebral fractures. Objectives To determine the prevalence of VFs after 5 years of DAS-steered treatment in patients with early active RA and to investigate the association of VFs with disease activity, functional ability and bone mineral density (BMD) over time. Methods Five-year radiographs of the lateral thoracic and lumbar spine of 275 patients in the BeSt study, a randomized trial comparing four treatment strategies, were available. Treatment adjustments were made every 3 months aiming at a DAS≤2.4. Vertebral fractures were assessed using the Genant method, with a fracture defined as loss of height reduction >20% in one vertebra. BMDs of the spine and hip were measured with dual energy X-ray absorptiometry. With linear mixed models, DAS and Health Assessment Questionnaire (HAQ) scores over 5 years were compared for patients with and without VFs. With GEE the association between BMD and VFs was determined. Results At baseline patients were on average 54 years old and most were female (67%), of whom 18% were postmenopausal. Mean DAS was 4.4 and mean HAQ score was 1.3. After 5 years of DAS steered treatment, VFs were observed in 41/275 patients (15%). No difference in prevalence was found when stratified for gender, prednisone use and menopausal status. Disease activity over time was higher in patients with VFs, with a mean difference of 0.20 (95% CI:0.05-0.36). HAQ scores were higher in patients with VFs, independent of disease activity, with a mean difference of 0.12 (95% CI:0.02-0.2). Although values were slightly lower over time in patients with vertebral fractures, mean BMDs in the spine and hip over time were not independently associated with VFs (OR 0.99, 95%CI:0.78-1.25 and 0.94, 95%CI:0.65-1.36, respectively). Higher age was independently associated with VFs (OR 1.06, 95%CI:1.02-1.09). Conclusions After 5 years of DAS-steered treatment, 15% of these RA patients had VFs. Higher age was associated with the presence of VFs, but mean BMDs in the hip and spine were not. VFs are associated with more disability, independent of disease activity. Patients with VFs have a higher disease activity over time, suggesting that VFs may be prevented by optimal disease activity suppression. Disclosure of Interest None Declared

THU0013 Interaction between two common HLA antigens defines a subset of individuals at a very high risk for ankylosing spondylitis

Background Susceptibility to spondyloarthritis (SpA) is largely genetically determined. To understand increasingly complex genetic associations, one can look for interaction between genetic risk factors. Objectives In this study, we investigated interaction between common HLA class I risk antigens in ankylosing spondylitis (AS) the most typical form of SpA. Methods In 155 patients with AS and 5584 controls, common HLA class I antigens were analyzed for association with AS. Biologic interaction was analyzed by investigating whether the effects of the risk factors combined departed from additivity. Results Apart from an association with HLA-B27, we found an association between HLA-B60 and AS (OR 2,0; 95%CI 1.3-2.9; p<0.001). This was confirmed in meta-analysis of published studies (OR 2.3; CI 1.8 – 2.8). While 21.3% of AS patients had both HLA-B27 and HLA-B60, this combination was found in only 0.4% of controls. With HLA-B27-/HLA-B60- AS patients as the reference, the relative risk (RR) for HLA-B27-/HLA-B60+ patients was 2.3 (CI 0.9-5.8). For HLA-B27+/HLA-B60- the RR was 66 (CI 40-111) but increased to 342 (CI 167- 703) in HLA-B27+/HLA-B60+ patients. For the interaction, the relative excess risk (RERI) was 310 (95% CI 130-490), the attributable proportion (AP) was 0.8 (CI 0.7-0.9), and the synergy index (S) 5.7 (CI 3.8 -8.5). The interaction was confirmed in an independent cohort. Conclusions We report a strong interaction between HLA-B60 and HLA-B27 in AS susceptibility. As a result, individuals with the HLA-B27+/HLA-B60+ genotype are at a very high risk of developing AS. Disclosure of Interest None Declared

Association between HLA class II genes and autoantibodies to cyclic citrullinated peptides (CCP) affects severity of rheumatoid arthritis

Objective. The functional role of HLA class II molecules in the pathogenesis of rheumatoid arthritis (RA) is unclear. HLA class II molecules are involved in the interaction between T and B lymphocytes required for long-lived B cell responses and generation of highaffinity IgG antibodies. We undertook this study to investigate the relationship between HLA class II gene polymorphisms and RA-specific IgG antibodies against cyclic citrullinated peptides (anti-CCP antibodies). Methods. High-resolution HLA–DR and DQ typing and anti–CCP-2 antibody testing were performed on 268 RA patients from the Early Arthritis Clinic cohort at the Department of Rheumatology of the Leiden University Medical Center. The presence of anti-CCP antibodies was analyzed in carriers of the different DR and DQ alleles. Disease progression was measured over a period of 4 years by scoring radiographs of the hands and feet using the Sharp/van der Heijde method. Results. Carriership of the individual alleles HLA–DRB1*0401, DRB1*1001, DQB1*0302, and DQB1*0501 was associated with the presence of antiCCP antibodies. Carriers of DQ–DR genotypes containing proposed RA susceptibility alleles were significantly more often anti-CCP antibody positive. Carriership of one or two HLA–DRB1 shared epitope (SE) alleles was significantly associated with production of anti-CCP antibodies (odds ratio [OR] 3.3, 95% confidence interval [95% CI] 1.8–6.0 and OR 13.3, 95% CI 4.6–40.4, respectively). An increased rate of joint destruction was observed in SE, anti-CCP patients (mean Sharp score 7.6 points per year) compared with that in SE, anti-CCP patients (2.4 points per year) (P 0.04), SE, anti-CCP patients (1.6 points per year) (P < 0.001), and SE, anti-CCP patients (1.6 points per year) (P < 0.001). Conclusion. HLA class II RA susceptibility alleles are associated with production of anti-CCP antibodies. Moreover, more severe disease progression is found in RA patients with both anti-CCP antibodies and SE alleles.

PS7:129 Synergetic b-cell immunomodulation with rituximab and belimumab combination treatment in severe, refractory sle

Background Neutrophil extracellular traps (NETs) are auto-antigenic DNA strands and potentially give rise to SLE-specific autoantibodies that can deposit in glomeruli. It has been shown that autoantibodies can induce NETs, contributing to the vicious circle of immune activation in SLE. We hypothesised that eliminating autoantibodies can lead to decreased NET induction and thereby ameliorating disease in SLE. Therefore, we designed a proof-of-concept study to eliminate autoantibodies and NET formation through synergetic B-cell immunomodulation with rituximab and belimumab (RTX+BLM) in severe refractory SLE. Methods We treated patients with severe, refractory SLE in a phase 2 study with RTX+BLM. The primary endpoint assessed reduction of pathogenic autoantibodies and NET induction at 24 weeks. Anti-dsDNA autoantibodies were measured and high sensitivity FACS was performed to assess B-cell subsets. NET induction was measured with 3D confocal immunofluorescence microscopy. Results We included 16 patients with severe, refractory SLE of whom 13 had a renal flare. At 24 weeks we observed significant reductions in anti-dsDNA autoantibodies (p=0.0004). CD19 +Bu2009cells were depleted throughout the study (p=0.0002) while plasma cells (PCs) temporarily decreased but returned at week 24 despite persistent depletion of transitional B-cells. Taken together with the observed reductions of autoantibodies and stable total IgG and protective antibodies, there is no reconstitution of autoreactive PCs. Further, we observed excessive NET induction in all patients at baseline which was reduced after 24 weeks (p=0.0006). In vitro studies elucidated this resulted in reduction of immune complexes by RTX+BLM. Importantly, the beneficial immunological effects translated to amelioration of clinical disease activity: SLEDAI decreased from a median of 18 to 2 (p<0.0001). Eleven out of 13 LN patients showed a response (5 complete renal responders). The response was achieved while tapering immunosuppressive medication. Treatment was generally well-tolerated. Conclusion The SynBiose study is the first to demonstrate that RTX+BLM ameliorated disease in patients with severe SLE in association with the reduction of pathogenic autoantibodies and immune complex-mediated NET induction. Therefore, RTX+BLM represents a novel treatment concept in SLE.

SAT0096 Window or no window: earlier is better when treating rheumatoid arthritis

Background Previous reports on a window of opportunity in rheumatoid arthritis (RA) may be related to the use of slow acting c(onventional) s(ynthetic) DMARDs. We hypothesised that onset of action of therapy might influence whether there is a window of opportunity or whether ‘earlier is just better’. Objectives To investigate the association between symptom duration at treatment onset and the achievement of sustained drug free remission (sDFR) in early RA patients initiating therapy including fast acting prednisone or infliximab, compared to patients initiating csDMARD monotherapy. Methods We analysed the shape (hyperbolic or linear) of the association between symptom duration and achievement of sDFR (DAS<1.6u2009and no DMARDs for ≥1u2009year) for patients initiating combination therapy in 3 cohorts: BeSt, IMPROVED study and METEOR. All patients had arthritis symptoms<2 years. In the BeSt study, early RA patients (1987 criteria) were randomised to 4 targeted treatment strategies: arm 1 and 2 started with csDMARD monotherapy, arm 3 with csDMARDs and tapered high dose prednisone and arm 4 with csDMARD and infliximab. Subsequent treatment adjustments aimed at DAS≤2.4. In the IMPROVED study early RA patients (2010 criteria) were treated with csDMARD and tapered high dose prednisone. Subsequent treatment adjustments aimed at DFR. METEOR is an international observational cohort including daily practice data from patients with a diagnosis and treatment of RA according to the rheumatologist. We selected patients who started on csDMARD monotherapy or a combination of csDMARD with prednisone or anti-TNF and at least 1.5u2009year follow-up. Missing DAS were imputed using last observation carried forward. We performed Cox regression analyses with as outcome sDFR and as predictor symptom duration and compared the likelihood ratio tests of a linear model and a model with inclusion of natural cubic spline functions (resulting in a hyperbola), to investigate which model was a better fit for the data. Results In BeSt (n=469), IMPROVED (n=421) and METEOR (n=1653) 54, 110 and 10 patients who started with csDMARD and prednisone or anti-TNF combination therapy, and 53 in BeSt and 15 in METEOR who started with csDMARD monotherapy (no monotherapy in IMPROVED) achieved sDFR. A hyperbolic model did not show a better fit for the data than a linear model (for combination therapy in BeSt p=0.743, IMPROVED p=0.337, METEOR p=0.608, for csDMARD monotherapy in BeSt p=0.609, in METEOR p=0.758). Figure 1 shows the linear association between symptom duration and sDFR per study. These results indicate that the earlier treatment is started, the higher the likelihood of achieving sDFR. However, the data do not suggest that a hyperbolic relationship between treatment onset and outcome sDFR fit the data better. Conclusions Our data suggest that there is no evidence for a window of opportunity in early RA in 3 cohorts. This was not related to use of fast acting combination therapy including prednisone or anti-TNF instead of slow acting csDMARD monotherapy nor was it dependent on strict treatment to target in a clinical trial. It remains that earlier treatment initiation is better when aiming to achieve sDFR. Disclosure of Interest J. van der Pol: None declared, S. A. Bergstra: None declared, N. Riyazi: None declared, Y. Goekoop-Ruiterman: None declared, A. Chopra: None declared, P. Kerstens: None declared, W. Lems: None declared, R. Tsonaka: None declared, T. Huizinga Consultant for: UCB, BMS, Pfizer, Roche, Sanofi-Aventis and Boeringher from outside the submitted work., C. Allaart Grant/research support from: The IMPROVED study was designed by the investigators and financially supported during the first year of follow-up by AbbVie. The BeSt study was supported by the Dutch College of Health Insurances, with an additional grant from Schering-Plough BV and Centocor Inc.

OP0176 Five Year Outcomes of Remission Steered Treatment Including Drug Tapering Strategies in Early Arthritis Patients

Background The window of opportunity hypothesis suggests that early initiation of treatment targeted at remission in early arthritis patients may avoid chronicity of inflammation and may result in early drug-free remission (DFR). Objectives To assess clinical outcomes of induction therapy followed by 5 years DAS-remission targeted therapy including rapid drug tapering strategies in early arthritis patients. Methods The IMPROVED study included 610 early rheumatoid arthritis (RA, 2010 criteria) or undifferentiated arthritis (UA) patients starting with methotrexate (MTX) and tapered high dose of prednisone. Patients in early DAS-remission (ER) (disease activity score (DAS)<1.6 after 4 months) stopped prednisone and if remission persisted at t=8 months also MTX. Patients not in ER (DAS≥1.6) were randomized to MTX+sulfasalazine+hydroxychloroquine+low dose prednisone (arm 1) or MTX+adalimumab (arm 2), 50 patients were not randomized and were treated “outside of protocol” (OP). Four monthly treatment adjustments aimed at DAS<1.6; if DAS<1.6 taper/stop medication and if DAS≥1.6 restart/intensify medication. Percentages of patients in (drug free) DAS-remission after 5 years were compared between RA and UA patients and the different treatment strategies. Logistic regression analysis was used to identify predictive factors of DAS-remission and DFR after 5 years. Results After 5 years mean±SD DAS decreased and functional ability improved in all treatment groups without differences between arm 1 and 2. 295/610 (48%) patients were in DAS-remission: 220/387 (57%) in the ER group, 31/83 (37%) in arm 1, 29/78 (37%) in arm 2 (p=0.768 arm 1 vs arm 2) and 15/50 (30%) in OP. 134/610 (22%) patients were in DFR: 105/387 (27%) in the ER group, 9/83 (11%) in arm 1, 12/78 (15%) in arm 2 (p=0.374 arm 1 vs arm 2) and 8/50 (16%) in OP. DAS-remission was achieved in similar percentages in RA and UA patients and autoantibody positive (+) vs negative (−) patients. More UA than RA patients were in DFR at year 5 (33% vs 19%, p<0.001), and the same was true for anti-citrullinated protein antibodies (ACPA)− vs ACPA+ patients (31% vs 15%, p<0.001) and rheumatoid factor (RF)− vs RF+ positive patients (28% vs 17%, p<0.001). Between anti-carbamylated antibodies (anti-CarP)+ and − patients there was no difference in DFR rates. Predictors of DAS-remission were age (OR 0.97 (95% CI 0.96–0.99)), symptom duration (0.99 (0.98–0.99)), baseline tender joint count (0.95 (0.90–1.00)) and achieving ER (1.95 (1.23–3.10)) (figure). Predictors of DFR were symptom duration (0.99 (0.97–1.00) and ER (2.67 (1.48–4.84)). ACPA+ was mostly associated with less DFR. Conclusions After 5 years of DAS-remission steered treatment, 48% of early RA and UA patients were in DAS-remission and 22% in DFR. Remission results were better for patients who achieved ER and who had a shorter symptom duration. Remission results were similar for RA and UA patients, randomization arms and autoantibody status, but more UA than RA patients and more autoantibody? than + patients were in DFR. Disclosure of Interest G. Akdemir: None declared, L. Heimans: None declared, R. J. Goekoop: None declared, M. van Oosterhout: None declared, J. B. Harbers: None declared, C. Bijkerk: None declared, G. M. Steup-Beekman: None declared, L. Lard: None declared, P. B. J. de Sonnaville: None declared, B. A. M. Grillet: None declared, T. W. J. Huizinga: None declared, C. F. Allaart Grant/research support from: Year 1 of the IMPROVED study was sponsored by Abbott.