J. Rodrigues

THU0365 Do Patterns of Joint Swelling or Tenderness in Rheumatoid Arthritis Patients Impact Disease Activity Outcomes and Pain? Implications for Clinical Practice

Objectives This analysis aimed to describe the pattern of specific joint involvement (tender and/or swollen) pre- and post-TNFi treatment and the impact of specific joint pattern involvement on composite score outcomes and pain. Methods BioTRAC is an ongoing, prospective registry of patients initiating treatment for RA, AS, or PsA with infliximab (IFX) or golimumab (GLM). In this analysis, RA patients included those treated with IFX between 2002-2014 or with GLM between 2010-2014. Based on joint involvement 7 groups were created: shoulder(s), elbow(s), metacarpophalangeal (MCP(s)), wrist(s), proximal interphalangeal (PIP(s)), knee(s), and thumb(s). The impact of specific joints on disease activity indices and pain was assessed with the independent-samples t-test; linear regression produced adjusted estimates. Results A total of 1030 RA patients were included with 5177 assessments. At baseline, MCP(s) (84.8%) and wrist(s) (66.1%) were the most commonly swollen joints. Tenderness was most frequent at baseline in these two joint types (81.1% and 70.9% of patients, respectively). Swelling/tenderness rates in all joint groups were significantly lower (p<0.001) among patients enrolled in 2010-2013 vs. those enrolled in 2002-2005; no significant differences, however, were observed in joint involvement pattern. Swelling and tenderness in all joint groups were associated with significantly (P<0.001) higher pain. Upon adjusting for age, gender and the total number of swollen (SJC28) or tender (TJC28) joints, swollen shoulder(s) and knee(s), and tender shoulder(s) and elbow(s) had the biggest impact on pain. Swollen MCP(s), knee(s) and thumb(s) had the greatest impact on DAS28, while for CDAI and SDAI swollen thumb(s) and swollen thumb(s) and knee(s), respectively, showed the highest association. Tender wrist(s), shoulder(s), and knee(s) showed the highest association with DAS28, while tender MCP(s) had the greatest impact on CDAI and SDAI. However, all indices were significantly higher among cases with swollen thumb(s) (unstandardized coefficient (B): BDAS28=0.25, P=0.006; BCDAI=2.09, P=0.001; BSDAI=2.66, P=0.001). Conclusions Although joint swelling/tenderness documented at anti-TNF initiation has decreased over time, the profile of affected joints has remained stable. Swelling/tenderness in specific joint groups was differentially associated with pain, with larger joints having the greatest impact. Furthermore, differences were observed in levels of disease activity based on the type of affected joint which could be attributed to their impact on patient global assessment. These results suggest that location of joint involvement, in addition to the number of affected joints, has an independent impact on pain. Disclosure of Interest R. Arendse Consultant for: Janssen, J. Kelsall Consultant for: Janssen, A. Avina-Zubieta Consultant for: Janssen, P. Baer Consultant for: Janssen, J. Rodrigues Consultant for: Janssen, A. Jovaisas Consultant for: Janssen, I. Fortin Consultant for: Janssen, M. Sheriff Consultant for: Janssen, M. Khraishi Consultant for: Janssen, E. Rampakakis: None declared, J. Sampalis: None declared, F. Nantel Employee of: Janssen, M. Shawi Employee of: Janssen, C. Tkaczyk Employee of: Janssen, S. Otawa Employee of: Janssen, A. Lehman Employee of: Janssen

SAT0062 What is the Effect of TNF Inhibitors on Employment Status in Rheumatoid Arthritis Patients and what are the Predictors of Progression to Unemployment

Objectives The aim of this analysis was to evaluate the prevalence of unemployment due to work disability in RA patients initiating treatment with infliximab (IFX) or golimumab (GLM) and to identify determinants of disability. Methods BioTRAC is an ongoing, prospective registry of patients initiating treatment for RA, AS, or PsA with IFX or GLM as first biologics. Data were obtained from RA patients treated with IFX (2002-2014) or GLM (2010-2014). Between employment group differences were assessed for statistical significance with the independent samples t-test or the chi-square. Time to employment and time to unemployment were assessed with the Kaplan-Meier (KM) estimator of the survival function. Cox regression was used to identify predictors of time to unemployment. Results A total of 581 RA patients were included; 374 (64.4%) employed and 207 (35.6%) unemployed due to disability. The following baseline parameters were associated with significantly increased likelihood of being unemployed due to disability: female vs. male gender (40.1% vs. 27.6%; P=0.006), earlier enrolment period (2002-05 vs. 2006-09 vs. 2010-14: 49.3% vs. 30.5% vs. 22.4%; P<0.001), insurance type (provincial vs. private vs. both: 54.9% vs. 23.8% vs. 20.0%; P<0.001), older age (P=0.033), and increased disease activity as evidenced by the higher DAS28 (P<0.001), SJC (P<0.001), TJC (P<0.001), HAQ (P<0.001), MDGA (P<0.001), PtGA (P<0.001), CDAI (P<0.001), SDAI (P<0.001), pain (P<0.001), and ESR (P<0.001). Among disabled patients, 10.1% were able to return to work upon treatment with TNF a mean KM-based duration of 119.5 months from baseline; whereas 6.4% of employed patients became disabled (2002-05 vs. 2006-09 vs. 2010-14: 7.0% vs. 10.1% vs. 1.7%; P=0.021) with a mean time to unemployment of 113.4 months. Multivariate survival analysis showed that, upon adjusting for enrolment period, higher baseline HAQ [HR (95%CI): 3.59 (1.64, 7.87), P=0.001], and higher baseline SJC [HR (95%CI): 1.09 (1.02, 1.16), P=0.011] were significant predictors of unemployment due to disability. Conclusions A significant proportion of RA patients are unemployed due to disability. At anti-TNF initiation, work disability was associated with higher disease activity, female gender, earlier enrolment period, and provincial insurance. Increased HAQ and higher SJC were significant predictors of progression to unemployment. Anti-TNF treatment was effective in enabling a considerable portion of disabled patients to return to employment. Disclosure of Interest A. Chow: None declared, W. Bensen Consultant for: Janssen, R. Arendse Consultant for: Janssen, E. Keystone Consultant for: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, Speakers bureau: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, P. Baer Consultant for: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, Speakers bureau: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, J. Kelsall Consultant for: Janssen, W. Olszynski: None declared, J. Rodrigues: None declared, A. Avina-Zubieta: None declared, M. Baker: None declared, W. Olszynski: None declared, W. Bensen Consultant for: Janssen, P. Baer Consultant for: Janssen, D. Choquette Consultant for: AbbVie, Amgen, Celgene, BMS Canada, Janssen, Pfizer, S. Kapur: None declared, A. Jaroszynska: None declared, J. Sampalis Shareholder of: JSS Medical Research, Employee of: JSS Medical Research, D. Choquette Consultant for: AbbVie, Amgen, Celgene, BMS Canada, Janssen, Pfizer, E. Rampakakis Employee of: JSS Medical Research, S. Kapur: None declared, J. Stewart: None declared, C. Tkaczyk Employee of: Janssen, J. Sampalis Shareholder of: JSS Medical Research, Employee of: JSS Medical Research, M. Shawi Employee of: Janssen, E. Rampakakis Employee of: JSS Medical Research, A. Lehman Employee of: Janssen, F. Nantel Employee of: Janssen, S. Otawa Employee of: Janssen, C. Tkaczyk Employee of: Janssen, A. Lehman Employee of: Janssen

SAT0090 Exploring The Das: What is the Level of Agreement in the Classification of Remission and Low Disease Activity Among the Various Versions of the Disease Activity Score (DAS) and Their Correlation? An Analysis from a Prospective, Observational Registry

Background Two versions of DAS28 are available, DAS28-4 comprising 4 variables [tender and swollen joint counts, acute phase reactant (APR), and patient global assessment] and DAS28-3 where patient global has been omitted. Despite the difference between DAS28-4 and DAS28-3 thresholds for remission and low disease activity (LDA) are the same. The APR used to calculate DAS may be either ESR or CRP. Objectives This analysis describes the agreement between these four possible indices, DAS28-4-ESR, DAS28-4-CRP, DAS28-3-ESR and DAS28-3-CRP and compares them in terms of classifying remission and LDA in a real-world, routine clinical care setting. Methods BioTRAC is a prospective registry of patients initiating treatment with infliximab or golimumab. In this analysis, data from RA patients who were treated with infliximab between 2002-2014 or with golimumab between 2010-2014 and had available information in all indices were used. The definitions for remission were: DAS28-3/4 <2.6; LDA was defined as: DAS28-3/4 <3.2. Correlation between different indices was assessed with Pearsons correlation coefficient (r) and classification agreement was assessed with Cronbachs alpha (CA) and the kappa statistic. Results 869 RA patients who had 3,517 complete assessments were included in the analysis. Non-remission was classified by all indices in 61.4% of cases, while remission was achieved in one (5.9%), two (10.3%), three (5.3%), or all four (17.2%) indices. Similarly, non-LDA was classified by all indices in 46.1% of cases, while LDA was achieved in one (6.2%), two (10.9%), three (5.4%), or all four (31.3%) indices. Overall, a strong linear positive correlation (r>0.8) was observed between all indices. When looking at the internal consistency in terms of classifying disease state, the CA was 0.905 for remission and 0.923 for LDA suggesting an overall high internal consistency. However, when looking at individual inter-item correlations, agreement between indices was variable with DAS28-3CRP and DAS28-4CRP showing highest correlation and DAS28-3-ESR and DAS28-4-CRP showing lowest correlation. When comparing DAS28-4-ESR with DAS28-3-ESR, the latter categorized 16.5% of DAS28-4-ESR remission cases as non-remission and 3.0% of DAS28-4-ESR non-remission cases as remission. With respect to LDA, DAS28-3-ESR categorized 9.1% of DAS28-4-ESR LDA cases as non-LDA and 4.7% of DAS28-4-ESR non-LDA cases as LDA. Similar results were observed with DAS28CRP. Conclusions The results of this analysis show that, despite being highly correlated, variability exists in the classification of remission and LDA by the various DAS indices. Decision making based on disease state achieved may vary significantly based on type of APR used in the DAS index. Disclosure of Interest W. Bensen Consultant for: Janssen, E. Keystone Consultant for: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, Speakers bureau: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, P. Baer Consultant for: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, Speakers bureau: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, J. Rodrigues: None declared, A. Avina-Zubieta: None declared, W. Olszynski: None declared, D. Choquette Consultant for: AbbVie, Amgen, Celgene, BMS Canada, Janssen, Pfizer, S. Kapur: None declared, J. Sampalis Shareholder of: JSS Medical Research, Employee of: JSS Medical Research, E. Rampakakis Employee of: JSS Medical Research, C. Tkaczyk Employee of: Janssen, M. Shawi Employee of: Janssen, A. Lehman Employee of: Janssen, F. Nantel Employee of: Janssen, S. Otawa Employee of: Janssen

FRI0036 What is the Level of Agreement Between Disease Activity Indices and Response Criteria Among Rheumatoid Arthritis Patients Treated with TNF Inhibitors

Background Several standardized response criteria and disease activity indices are used to assess treatment efficacy in rheumatoid arthritis (RA). These measures comprise different types and number of variables resulting in different weighting of individual variables within each of them. Objectives The aim of this analysis was to compare the performance of ACR, SDAI major and minor, and HAQ response criteria and to determine their level of agreement with the DAS28, SDAI, and CDAI definitions of low disease activity (LDA) and remission in RA patients treated with infliximab (IFX) or golimumab (GLM) in a real-world, Canadian, clinical practice setting. Methods BioTRAC is an ongoing, prospective registry of patients initiating treatment for RA, AS, or PsA with IFX or GLM. The analysis was based on RA patients treated with IFX between 2002-2014 or GLM between 2010-2014. Response was assessed with ACR20, ACR50, ACR70, HAQ improvement of 0.22 and 0.5, SDAI major (≥22) and minor improvement (≥10). Disease state was assessed with DAS28, SDAI, and CDAI definitions of LDA (<3.2, ≤11, ≤10, respectively) and remission (<2.6, ≤3.3, ≤2.8, respectively). The level of agreement was assessed with the proportion of concordant pairs over the total number of cases in each cross-tabulation and the Kappa statistic. Results A total of 830 RA patients with 4,100 available observations were included. The criteria for each definition of response/disease state were met for the following proportion of cases: ACR20 (66.4%), ACR50 (44.5%), ACR70 (26.4%), ΔHAQ≥0.22 (65.5%), ΔHAQ≥0.5 (53.4%), SDAI major improvement (55.8%), SDAI minor improvement (80.8%), DAS28 remission (29.4%), CDAI remission (20.4%), SDAI remission (21.8%), CDAI LDA (57.5%), SDAI LDA (58.1%), and DAS28-ESR LDA (46.0%). Statistically significant (Kappa P<0.001) associations were observed for all combinations of variables examined. Overall, the ACR response criteria performed better than the HAQ and SDAI response criteria in their agreement with LDA and remission. In general, higher levels of response in all three measures (ACR20 vs. ACR50 vs. ACR70; ΔHAQ≥0.22 vs. ΔHAQ≥0.5; SDAI minor vs. major) showed better agreement with LDA and remission. The highest level of agreement between response criteria and disease state was observed between the strictest definitions, namely between ACR70 and CDAI/SDAI remission; whereas, SDAI minor improvement showed the lowest level of agreement with remission, irrespective of definition. Conclusions This analysis showed that significant variation exists in the agreement between the various efficacy outcome measures. Thus, the choice of outcome measure used to make treatment decisions could have a significant impact on patient management. Disclosure of Interest E. Keystone Consultant for: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, Speakers bureau: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, P. Baer Consultant for: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, Speakers bureau: AbbVie, Amgen, BMS, Janssen, Pfizer, Roche, A. Avina-Zubieta: None declared, A. Jaroszynska: None declared, J. Rodrigues: None declared, R. Arendse Consultant for: Janssen, D. Sholter: None declared, M. Starr Consultant for: Janssen, A. Masetto: None declared, J. Sampalis Shareholder of: JSS Medical Research, Employee of: JSS Medical Research, E. Rampakakis Employee of: JSS Medical Research, C. Tkaczyk Employee of: Janssen, M. Shawi Employee of: Janssen, A. Lehman Employee of: Janssen, F. Nantel Employee of: Janssen, S. Otawa Employee of: Janssen

SAT0336 Change over Time in the Profile of Ankylosing Spondylitis Patients Treated with Infliximab in A REAL World Routine Care

Objectives The objective of this study was to describe and compare over time the demographics and disease parameters at infliximab treatment initiation and to assess the effectiveness of treatment at 6 and 12 months in real life cohort AS patients. Methods BioTRAC is an ongoing, prospective registry of patients initiating treatment for rheumatoid arthritis (RA), ankylosing spondylitis (AS), or psoriatic arthritis (PsA) with infliximab or golimumab as first biologics or after having been treated with a biologic for less than six months. People with AS treated with infliximab who were enrolled between 2002 and 2013 were included in this analysis (N=303) and stratified to two groups (2005-2007: n=135; 2008-2013: n=168) based on the year of enrolment in the registry. Results Patient demographics were comparable in the two cohorts with a mean (SD) age of 45.72 (11.74) years and the majority being males (62.4%). A significant change in the geographic distribution of patients enrolled in the BioTRAC registry was observed (P=0.001) and more patients with provincial coverage were enrolled in 2008-2013 compared to 2005-2007 (P=0.012). A trend towards earlier initiation of infliximab was observed in more recent years as indicated by the shorter disease duration (11.12 vs. 8.24 years; P=0.013) and the lower number of prior traditional DMARDs used (0.83 vs. 0.59; P=0.078). Furthermore, overall, patients recruited in 2008-2013 had lower disease activity compared to those enrolled in 2005-2007. ESR (29.96 vs. 19.91 mm/hr; P<0.001), physician global assessment (MDGA; 6.99 vs. 6.26; P=0.001) were significantly lower in the 2008-2013 cohort while a statistical trend was observed in morning stiffness (78.96 vs. 70.11 minutes; P=0.064) and ASDAS (3.90 vs. 3.70; P=0.103). Treatment for 6 months resulted in a greater proportion of patients in the 2008-2013 cohort achieving inactive disease (ASDAS<1.3) without reaching statistical significance (20.7% vs. 34.9%; P=0.140). Conclusions The results of this analysis show that the profile of the AS patient population in the BioTRAC registry has changed over time towards lower disease activity and earlier initiation in the patient management process. These results may reflect differences in patient management over time or may be related to earlier access to care. Irrespective of enrolment period, 6-month treatment with infliximab was effective in reducing disease activity in AS patients. Disclosure of Interest D. Choquette: None declared, M. Starr: None declared, M. Khraishi: None declared, W. Bensen: None declared, S. Shaikh: None declared, J. Rodrigues: None declared, D. Sholter: None declared, M. Sheriff: None declared, J. Vaillancourt: None declared, J. Sampalis: None declared, A. Lehman Employee of: Janssen, S. Otawa Employee of: Janssen, F. Nantel Employee of: Janssen, M. Shawi Employee of: Janssen DOI 10.1136/annrheumdis-2014-eular.2375

AB0302 Profile of Joint Involvement over Time in Rheumatoid Arthritis and Psoriatic Arthritis Patients Treated with Anti-TNF in A Real-World Setting

Background Unlike rheumatoid arthritis (RA), the pattern of joint involvement in psoriatic arthritis (PsA) is usually asymmetric. Furthermore, PsA may demonstrate oligoarthritis or polyarthritis, while RA usually manifests in multiple joints. Objectives To describe the most commonly affected joints in patients with RA and PsA at baseline and after 6 months (mos) of treatment with infliximab (IFX) in a clinical practice setting. Methods BioTRAC is an ongoing, prospective registry of patients initiating treatment for RA, ankylosing spondylitis, or PsA with IFX as first biologics or after having been treated with a biologic for <6 mos. RA patients treated between 2002-2012 and PsA patients treated with IFX between 2005-2012 were included. Based on the 28-joint involvement 7 groups were created: shoulder(s), elbow(s), metacarpophalangeal (MCP(s)), wrist(s), proximal interphalangeal (PIP(s)), knee(s), and thumb(s). The impact of treatment on joint swelling/tenderness was assessed with the McNemar test while the Chi-square test was used to compare the affected joints between disease groups. Results 832 RA patients (mean age: 55.8 yrs disease duration: 10.2 yrs) and 92 PsA patients (age: 48.7 yrs; disease duration: 6.8 yrs) were included. At baseline, mean DAS28, SJC28 and TJC28 in RA vs. PsA patients were 5.8 vs. 4.1 (P<0.001), 10.7 vs. 4.0 (P<0.001), and 12.6 vs. 5.9 (<0.001), respectively. Among RA patients, the joints most commonly swollen at baseline were MCPs (86.8% of patients), wrists (70.5%), and PIPs (53.2%). Knees, thumbs, elbows and shoulders were swollen in 42.3%, 33.7%, 30.5%, and 16.7% of patients, respectively (Figure 1A). With respect to tenderness, MCPs were tender in 83.8% of patients, wrists in 75.3%, shoulders in 57.8%, knees in 54.8%, PIPs in 55.3%, thumbs in 38.8%, and elbows in 41.0% (Figure 1B). Statistically significant differences were observed between RA and PsA patients both in the frequency of joint swelling/tenderness, which were lower in PsA, and the profile of affected joints. Among PsA patients, MCPs, wrists, and knees were the joints most commonly swollen, affected in 57.6%, 34.8%, and 31.5% of patients, respectively; MCPs, knees, and wrists were the joints most commonly tender (63.0%, 43.5%, and 42.4% of patients, respectively). Upon 6 mos of treatment, significant improvement in swelling/tenderness in all the most commonly affected joints in both RA and PsA patients was observed. The joints most resistant to treatment, still remaining affected at 6 mos, were MCPs in both patient groups. Figure 1. Types of swollen (A) and tender (B) joints at anti-TNF treatment initiation in RA and Psa Conclusions Significant differences exist in both the frequency and the profile of swollen and tender joints in RA and PsA patients although both diseases shared the MCPs as the joint most commonly affected and most resistant to treatment. Treatment with IFX for 6 mos resulted in a significant reduction in the 28-swollen and tender joint counts in both RA and PsA patients. Disclosure of Interest A. Jovaisas: None declared, M. Starr: None declared, D. Choquette: None declared, M. Zummer: None declared, R. Arendse: None declared, D. Sholter: None declared, R. Faraawi: None declared, J. Rodrigues: None declared, S. Kapur: None declared, E. Rampakakis: None declared, J. Sampalis: None declared, F. Nantel Employee of: Janssen, A. Lehman Employee of: Janssen, S. Otawa Employee of: Janssen, M. Shawi Employee of: Janssen DOI 10.1136/annrheumdis-2014-eular.2385

AB1044 Are There Gender Specific Differences in Patient Reported Outcomes at Initiation of Golimumab Treatment in Rheumatoid Arthritis

Background The prevalence of rheumatoid arthritis (RA) is 2-4 times higher in women compared to men. Furthermore, RA incidence in women increases from the age of menarche peaking around menopause, while it is rare in men younger than 45 years (1). Several studies have shown that treatment outcomes are worse in women (2). Objectives This analysis examined gender-specific differences with respect to disease parameters at initiation of the first anti-TNF agent for RA treatment in a Canadian routine clinical practice setting. Methods BioTRAC is an ongoing, prospective registry of patients initiating treatment for RA, AS, or PsA with infliximab or golimumab as first biologics or after having been treated with a biologic for less than six months. In this analysis, data were assessed from RA patients treated with golimumab subcutaneous as a first biologic who were enrolled between 2010 and 2012. Results 121 RA patients were included with mean (SD) disease duration of 9.3 (9.6) years. Eighty-nine patients (73.6%) were female. In the overall population, rheumatoid factor positivity was observed in 67% of patients and 17.5% were smokers, without any significant differences between genders. Patient reported disease parameters differed significantly between genders. Despite the younger age (56.6 vs. 62.0 years; P=0.051), female patients reported significantly higher pain (56.6 vs. 44.5 mm; P=0.045), patient global assessment (PtGA: 6.0 vs. 4.7 cm; P=0.034), tender joint count (10.2 vs. 6.8; P=0.022), and functional disability (HAQ-DI: 1.43 vs. 1.06; P=0.024). In addition, a statistical trend towards higher morning stiffness in female patients was observed (55.7 vs. 38.9; P=0.063). However, physician assessment of global disease activity (MDGA: 5.7 vs. 5.5; P=0.581), SJC (7.9 vs. 8.5; P=0.641), DAS28-ESR (5.3 vs. 4.7; P=0.086), CDAI (29.7 vs. 24.5; P=0.120) and SDAI (32.1 vs. 30.2; P=0.675) were statistically comparable between genders. Conclusions Objective measures (SJC, CRP/ESR), MDGA and composite outcomes were similar for male and female patients at golimumab initiation, with the exception of TJC being higher in women. Patient reported outcomes (PROs: pain, PtGA, HAQ-DI), however, were worse at baseline for female patients at biologic treatment initiation. These findings are similar to our previous research on patients initiating anti-TNF IV therapy (3). Overall, the results may suggest gender bias in the interpretation and use of PROs during the treatment decision making process in Canadian RA patients. References Wilder RL. J Rheumatol Suppl. 1996 Mar;44:10-2. Forslind K. et al. Ann Rheum Dis. 2007 Jan;66(1):46-52. Bensen W. et al. J Rheumatol Suppl. 2013 Jun:40:1020-1. Disclosure of Interest D. Sholter: None declared, W. Bensen: None declared, D. Choquette: None declared, I. Fortin: None declared, R. Arendse: None declared, J. Kelsall: None declared, M. Sheriff: None declared, R. Faraawi: None declared, J. Rodrigues: None declared, M. Zummer: None declared, S. Dixit: None declared, M. Starr: None declared, E. Rampakakis: None declared, J. Sampalis: None declared, F. Nantel Employee of: Janssen Inc Canada, M. Shawi Employee of: Janssen Inc Canada, S. Otawa Employee of: Janssen Inc Canada, A. Lehman Employee of: Janssen Inc Canada DOI 10.1136/annrheumdis-2014-eular.3979

THU0247 What is More Predictive of Achieving Remission at 12 Months: the Percentage of Baseline Improvement or the Actual Disease State Achieved?

Background The aim of rheumatoid arthritis (RA) treatment is to optimize symptom control and, when possible, achieve sustained remission. Therefore, identification of clinical signs predicting future remission is valuable to clinical decision making. One question faced by clinicians is whether the achievement of a lower disease activity value or a higher rate of change of disease activity is indicative of better future disease outcomes. Objectives To determine whether change in disease activity measures or the actual values achieved at 6 months were more predictive of remission at 12 months in RA patients treated with infliximab (IFX) in a real-world, clinical practice setting. Methods BioTRAC is an ongoing, prospective registry of patients initiating treatment for RA with IFX or golimumab as first biologics or after having been treated with a biologic for <6 months. Eligible people for this study included RA patients treated with IFX enrolled between 2002-2012 with available 12-month information on remission. Multivariate logistic regression models with the parametric Wald statistic and the log-likelihood ratio were used to assess the independent contribution of the actual value and the change at 6 months in predicting 12-month remission as defined by DAS28 (<2.6), SDAI (≤3.3) and CDAI (≤2.8) criteria. These two statistics assess the extent of contribution of an individual predictor to an outcome of interest - higher values signify greater contribution - and can be used to compare the contribution of different predictors in a standardized fashion. Results 436 patients were included with mean age of 56.1 yrs and disease duration of 10.4 yrs. With respect to 12-month DAS28 remission, a stronger association was observed with the actual DAS28 score compared to the percent improvement in DAS28 at 6 months. The Wald statistic for the percent change and actual value of DAS28 at 6 months was 5.38 and 46.88, respectively, while the change in log-likelihood was 4.98 (P=0.026) and 61.64 (P<0.001), respectively, indicating that the actual DAS value achieved is significantly more predictive of remission when compared to percent change in DAS from baseline. For SDAI remission at 12 months, the respective Wald values for percent change and actual value at 6 months were 0.075 and 18.28 and log-likelihood changes were 0.07 (P=0.788) and 24.08 (P<0.001). For CDAI remission at 12 months, the Wald statistic was 0.01 and 34.42 for 6 month percent change and actual value, respectively, and change in log-likelihood was 0.01 (P=0.934) and 34.23 (P<0.001). Similar results were obtained when predicting low disease activity at 12 months. Conclusions These results demonstrate that the actual disease outcome value achieved at 6 months is a stronger predictor of remission at 12 months than the percent change in disease activity. These findings suggest that the treatment target in a real-world setting should be set as specific endpoints and not as change over time. Disclosure of Interest : None declared DOI 10.1136/annrheumdis-2014-eular.4443

AB0234 Differential Relative Contribution of Individual Components on DAS28 over Time. an Analysis from the Prospective, Observational Registry, Biotrac

Background DAS28 is an important outcome for clinical research and practice assisting with therapeutic decisions. The main contributors to DAS28 are joint tenderness and acute-phase reactants. A simulation analysis showed that, due to its logarithmic transformation in the DAS28 formula, the ESR contribution is greater in the lower than in the higher DAS28 range. Objectives This analysis assessed the relative contribution of individual DAS28 components and examined its clinimetric properties in rheumatoid arthritis (RA) patients treated with infliximab in a Canadian real-world setting. Methods BioTRAC is an ongoing, prospective registry of patients initiating treatment for RA, ankylosing spondylitis (AS), or psoriatic arthritis (PsA) with infliximab or golimumab as first biologics or after treatment with a biologic for <6 months (M). RA patients treated with infliximab between 2002-2012 and with ≤60M of follow-up were included. The association between treatment duration and parameter improvement was assessed using linear regression. Slope correlation was assessed with the Pearsons correlation coefficient. Results 832 patients evaluated over 4,002 visits were included. Longer treatment duration was associated with significant (P<0.001) improvements in DAS28, TJC28, SJC28, PtGA, ESR, and CRP. Correlation analysis of the rate of change over time showed a high correlation (0.7-0.9) of DAS28 with TJC28, SJC28, and PtGA but low correlation with ESR (r=0.418) and CRP (r=0.411). Overall, the relative contribution of TJC28, SJC28, PtGA, and ESR in DAS28-ESR was 22%, 9%, 12%, and 57%, respectively. For DAS28-CRP, the relative TJC28, SJC28, PtGA, and CRP contributions were 25%, 10%, 12%, and 20%. Over 60M of treatment, the mean relative contribution of TJC28 (M0:31%, M60:17%), SJC28 (M0:15%, M60:5%), and PtGA (M0:15%, M60:9%) significantly (P<0.001) decreased whereas the weight of ESR contribution increased (M0:39%, M60:69%). Similar results were obtained with DAS28-CRP although the CRP contribution was lower compared to ESR. Increased DAS28-ESR was associated with higher relative contributions (per unit of DAS28-ESR increase) of TJC28 [parameter estimate (B) =5.3], SJC28 (B=2.1), and PtGA (B=0.7) but lower ESR contribution (B=-8.1). Similarly, increased DAS28-CRP was associated with lower relative CRP contribution (B=-2.0). Conclusions This analysis shows that TJC28 and acute-phase reactants have a greater weight than SJC28 and PtGA within DAS28. Furthermore, the relative contribution of acute-phase reactants is greater with lower DAS28, due to their logarithmic nature. These findings suggest that biologic variability and variability in laboratory techniques may have significant impact on classifying remission or DAS28 changes among patients with low DAS28 and on therapeutic plan changes. Disclosure of Interest D. Choquette: None declared, D. Sholter: None declared, I. Fortin: None declared, M. Starr: None declared, C. Thorne: None declared, M. Baker: None declared, R. Arendse: None declared, P. Baer: None declared, M. Zummer: None declared, J. Rodrigues: None declared, M. Sheriff: None declared, E. Rampakakis: None declared, J. Sampalis: None declared, F. Nantel Employee of: Janssen, A. Lehman Employee of: Janssen, S. Otawa Employee of: Janssen, M. Shawi Employee of: Janssen DOI 10.1136/annrheumdis-2014-eular.2379

SAT0083 Does Specific Joint Involvement in Rheumatoid Arthritis Patients PREDICT Patient Reported Outcomes? Implications for Clinical Practice

Background Assessment of functional (dis)ability in rheumatoid arthritis (RA) is subject to patient judgment when appraising their ability to do daily activities. Objectives The aim of this analysis was to describe the impact of specific joint involvement on patient reported outcomes (PROs) - functional activity, pain and patient global assessment of disease activity (PtGA) - and to identify joints most resistant to treatment over time. Methods Biologic Treatment Registry Across Canada (BioTRAC) is an ongoing, prospective, registry of patients initiating treatment for RA, AS, or PsA with infliximab or golimumab as first biologics or after having been treated with a biologic for less than six months. In this analysis, data were included from RA patients treated with infliximab between 2002 and 2012 or with golimumab between 2010 and 2012. Based on joint involvement seven groups were created: shoulder(s), elbow(s), metacarpophalangeal (MCP), wrist(s), proximal interphalangeal (PIP), knee(s), and thumb(s). The impact of specific joints on patient outcomes was assessed with the independent-samples t-test. Linear regression was used in order to produce adjusted estimates. Results A total of 935 RA patients with 4,854 assessments were included. Swelling, tenderness, and swelling and/or tenderness in all joint groups were associated with significantly (P<0.001) higher HAQ-DI, PtGA, and pain. Upon adjusting for age, gender and the total number of swollen (SJC28) and tender (TJC28) joints, swelling in all joint groups but the thumb(s) had a significant impact on PtGA, pain and HAQ. Similarly, tenderness in all joints but PIP had a significant impact on these parameters. Overall, swelling and/or tenderness at specific joints - shoulders, wrists, knees and elbows - had the greatest impact on HAQ-DI, PtGA, and pain (P<0.001). Swollen and/or tender PIP(s) did not have a significant effect on any PRO. At baseline, the MCP joint(s) and the wrist(s) were the most commonly swollen (86.4% and 67.9% of patients, respectively) or tender (82.9% and 73.1%, respectively) joints. Upon 12 months of treatment, the MCP joints were the joints most resistant to treatment, still remaining affected. Conclusions Significant variability in PROs exists based on the presence of swelling and/or tenderness in specific joint groups. Swelling/tenderness of shoulders, wrists, knees and elbows were the main drivers of HAQ-DI, PtGA and pain. The results have important implications for the achievement of disease remission and suggest that the joint type in addition to the number of affected joints has unique impact on PROs. Disclosure of Interest R. Arendse: None declared, W. Bensen: None declared, A. Chow: None declared, J. Rodrigues: None declared, S. Dixit: None declared, D. Sholter: None declared, P. Baer: None declared, M. Baker: None declared, D. Choquette: None declared, I. Fortin: None declared, A. Jovaisas: None declared, E. Rampakakis: None declared, J. Sampalis: None declared, F. Nantel Employee of: Janssen Inc Canada, S. Otawa Employee of: Janssen Inc Canada, M. Shawi Employee of: Janssen Inc Canada, A. Lehman Employee of: Janssen Inc Canada DOI 10.1136/annrheumdis-2014-eular.5185