P211 Indirect measurement of central sensitisation as a predictor of response to different drug treatments in patients with inflammatory arthritis in a UK clinical cohort

Abstract Background Rheumatoid arthritis (RA) is a common, painful and disabling condition. Despite effective treatment of inflammation, many patients continue to experience ongoing pain. Central sensitisation could explain this apparent discordance, but at current is not routinely assessed in clinical practice. Previous studies have used the non-inflammatory components of Disease Activity Score-28 (DAS28-CRP) to derive novel, indirect indicators of central sensitisation. The aim of this study was to investigate whether DAS28-derived indices could be used to predict response to drug therapy in a clinical dataset. Methods This retrospective observational cohort study of patients with RA used NHS data collected prospectively from February 2016 to June 2018 from the Rheumatology Assessment Database in Oxford (Rhadio). The study received necessary ethical clearance from the NHS Health Research Authority (IRAS number: 246029). Baseline clinical and demographic data were extracted the following potential indirect indicators of centrally mediated pain were calculated: tender-swollen joint count difference (TJC-SJC difference), swollen to tender joint count ratio (STR), and DAS28-P, a measure of the non-inflammatory proportion of DAS28-CRP. The DAS28-P measure used herein includes CRP rather than ESR and is multiplied by 100 for ease of interpretation. Individual disease outcomes at follow-up were calculated according to European League Against Rheumatism (EULAR) criteria and multivariate logistic regression models including age, gender, treatment duration and drug regimens were generated to assess the relationship between potential baseline pain indicators and EULAR outcome at follow-up. Results We included data from 2,176 patients with RA.The majority were female (71.4%) with a median age of 64.8 (IQR= 53.6 to 73.3) years. The largest proportion of patients were classified as being in clinical remission at baseline, though there was considerable heterogeneity in the symptomatic profiles of patients within any given disease activity classification. 872 patients had valid follow-up data at ≥ three months post-baseline visit. Patients undergoing treatment escalation had significantly decreased DAS28-P scores at ≥ three months follow-up compared to baseline, while no significant change was observed in patients on stable therapy. Regression analyses showed that the three indirect indices of central sensitisation were positive predictors of treatment outcome. A unit increase in any one of TJC-SJC difference, STR or DAS28-P at baseline significantly increased the odds of a patient achieving at least a moderate EULAR response at follow-up, even when confounding factors were controlled for (DAS28-P: OR = 1.062 (95% CI: 1.050, 1.075) p < 0.0005; STR: OR = 2.746 (95% CI: 1.466, 5.146) p = 0.002; TJC-SJC difference: OR = 1.101 (95% CI: 1.055, 1.150) p < 0.0005). Conclusion The overall DAS28-CRP score is sensitive to fluctuations of different disease components including inflammation, pain and general health. Novel DAS28-CRP-derived indices can highlight the frequently observed discordance between the inflammatory and non-inflammatory components of the DAS28-CRP score. Disclosures S. Santos-Paulo None. A.R. Segerdahl None. S. Hawley None. R.A. Luqmani None. I. Tracey None. A. Soni None.