Pos0068 Flare and Drug-Free Remission in Rheumatoid Arthritis - Clinical Predictors From the Bio-Flare Study

Background: Rheumatoid arthritis (RA) flares remain poorly understood at a mechanistic level and have been difficult to study because of their sporadic nature. BIO-FLARE (BIOlogical Factors that Limit sustAined Remission in rhEumatoid arthritis; ISRCTN16371380) is an experimental medicine programme designed to scientifically study flares in RA, in which patients in remission stopped all DMARDs. We anticipated that 50% would develop disease flares over 6 months, providing matched flare and remission (control) groups to study. Objectives: Here we describe the main clinical characteristics and outcomes from the BIO-FLARE study, identify predictors of flare among routine baseline clinical parameters, and generate an exploratory clinical model to predict risk of flare. Methods: BIO-FLARE was a multi-centre, prospective, experimental medicine study 1. Eligible RA patients in remission (DAS28-CRP Table 1). Univariate analyses were performed to assess strength of association between each candidate variable and time-to-flare. A set of predictive variables was formally selected through penalised Cox models. Predictive modelling was carried out using Cox proportional hazards methodology, and final model performance assessed by internal validation using bootstrapping. Results: 121 participants were recruited between September 2018 and December 2020. Baseline demographics are shown in Table 1. Mean (SD) age was 64.1 (11.9) years, 60.3% were female, and median (IQR) disease duration was 6.3 (4.5–12.3) years. 56.3% were RF positive, 66.7% were ACPA positive, 56.6% were double positive. 83.5% were on treatment with methotrexate (monotherapy or combination). At baseline, mean (SD) DAS28-CRP was 1.61 (0.32), 61.2% fulfilled ACR/EULAR Boolean remission criteria, and 84.9% fulfilled SDAI remission criteria. Over the 24-week study period, 52.3% had confirmed flare (95% CI 43.0–61.7), with median time-to-flare of 63 days (IQR 41–96 days, range 13–155 days). Mean (SD) DAS28-CRP at flare was 3.81 (0.78). Hazard ratios and confidence intervals for each baseline variable from univariate time-to-flare analyses are shown in Table 2. The variable selection process identified female sex, baseline methotrexate use, RF, and ACPA as significant predictors of flare. These variables were used to build a predictive model providing estimated risk of flare for each individual by a given timepoint, computed as 1- S0(t)exp(PI), where S0(t) is the baseline survival function (0.67 at 6 months), and PI is the prognostic index (propensity to flare). PI can be calculated as (-0.56 x sex) + (1.06 x methotrexate use) + (0.04 x √(RF+0.1)) + (0.56/(ACPA+0.1)) – (1.87/√(ACPA+0.1)). The model had acceptable classification performance (C index 0.709) and was well-calibrated (calibration slope 1.00). Conclusion: Approximately half of RA patients in remission on csDMARDs experienced flare within 6 months of stopping treatment, with median time-to-flare of 9 weeks. Female sex, baseline methotrexate use, RF and ACPA were predictors of flare. Our predictive model, based on routine clinical parameters only, provides an estimate of risk of flare at the individual level over time. To our knowledge, no comparable clinical model has been described previously in this context. Work is ongoing in BIO-FLARE to identify immune biomarkers to strengthen our model, and to provide wider insights into the underlying biology of flare and remission in RA. REFERENCES: 1 Rayner et al. BMC Rheumatology. 2021. Acknowledgements: NIL. Disclosure of Interests: Fiona Rayner: None declared, Shaun Hiu: None declared, Andrew Melville MRC GSK EMINENT PhD studentship, Theophile Bigirumurame: None declared, Amy E Anderson: None declared, Sean Kerrigan Honoraria from Vifor., Andrew McGucken: None declared, Bernard Dyke: None declared, Jonathan Prichard: None declared, Mohadeseh Shojaei Shahrokhabadi: None declared, Catharien Hilkens Research funding from GSK., Iain B. Mc Innes Honoraria or research support from Abbvie, Janssen, Novartis, Eli Lilly, Astra Zeneca, GSK, BMS, Moonlake, Evelo, Causeway THerapeutics, Cabaletta, Roche, Pfizer and Compugen., Wan-Fai Ng: None declared, Carl S Goodyear: None declared, Dawn Teare: None declared, Andrew Filer: None declared, Stefan Siebert Speaker or consulting fees from AbbVie, Amgen, AstraZeneca, Eli Lilly, GSK, Janssen, UCB., Institutional research grants from Amgen (previously Celgene), Boehringer Ingelheim, Bristol Myers Squibb, Eli Lilly, GSK, Janssen and UCB., Karim Raza Personal fees for lectures/ consultancy from Abbvie and Sanofi., Research grant support from Bristol Myers Squibb and Sanofi., Arthur Pratt: None declared, Kenneth F Baker Consulting fees from Modern Biosciences., Research support from Genentech and clinical improvement funding from Pfizer., John Isaacs Speaker fees from AbbVie., Consulting fees from Anaptys Bio, Annexon Biosciences, AstraZeneca, BMS, Cyxone AB, Eli Lilly, Galapagos NV, Gilead Sciences Ltd, GSK, Istesso Ltd, Janssen, Kenko International, Kira Biotech, Ono Pharma, Pfizer, Revelo Biotherapeutics, Roche and Sanofi., Research grants from Pfizer, Janssen and GSK.