Background: COmbinatie therapie Bij Reumatoide Artiritis (COBRA)-light therapy (methotrexate and initially 30 mg/day prednisolone) has proven to be non-inferior to COBRA therapy (methotrexate, sulfasalazine and initially 60 mg/day prednisolone) in the first year of treatment of early rheumatoid arthritis (RA) patients.

Objectives: This study assessed changes in bone mineral density (BMD) after four years in early RA patients initially randomised to one year of COBRA or COBRA-light therapy.

Methods: In the open-label, randomised, non-inferiority trial patients were assigned to COBRA or COBRA-light therapy. After one year, treatment was at the discretion of the treating rheumatologists. BMD in g/cm² was measured at baseline, after one, two and four years at total hip, femoral neck, and lumbar spine with dual-energy X-ray absorptiometry (DXA).

Results: Of the 164 original patients, 154 could be assessed after a follow-up of four years (range 34 to 74 months); 68% were female; mean (SD) age at follow-up 52¹³ years. In the COBRA-light group, 11% of the patients used bisphosphonates after four years; the mean cumulative prednisolone dosage was 2.6 g (inner quartiles:1.9; 5.9) and 49% of the patients had minimal disease activity (DAS44 <1.6). In the COBRA group, these numbers were 10%, 3.2 g (2.5; 6.2) and 49%, respectively. At the lumbar spine, both groups showed no significant decline in BMD over four years and no difference between treatment groups in BMD change (table 1). At the hips, 1.7% to 3.7% BMD loss over four years was found with slightly but significantly more loss in the COBRA-light group (table 1).

Outcomes are mean (SD) unless stated otherwise. BMD in g/cm².+Significant difference between COBRA-light and COBRA on average over time. *Adjusted for bisphosphonate usage (yes vs no). **Adjusted for bisphosphonate usage (yes vs no), cumulative prednisolone usage, age, gender and disease activity based on DAS44 (DAS44 <1.6 in remission vs DAS44 ≥1.6 not). CI, confidence interval; GEE, Generalised Estimating Equations; T4, measurement after four years; SD, Standard Deviation.

Conclusions: In modern treat-to-target management of RA, including bone surveillance, a high starting dose of prednisolone, either 30 or 60 mg/day, was not associated with a dramatically increased bone loss at the lumbar spine, and minor losses at the hip over four years.

Disclosure of Interest: M. Lucassen: None declared, M. ter Wee: None declared, D. den Uyl: None declared, N. Konijn: None declared, M. Nurmohamed Speakers bureau: Janssen, Roche, MSD, Pfizer, Eli Lilly, BMS and Abbvie, D. van Schaardenburg: None declared, P. Kerstens: None declared, I. Bultink Speakers bureau: Lilly Netherlands, MSD, Amgen BV, UCB Pharma BV, Sanofi Genzyme BV, L. Van Tuyl: None declared, M. Boers Consultant for: Pfizer, Union Chimique Belge and Teva, W. Lems Grant/research support from: Pfizer, Speakers bureau: Pfizer, Abbvie and Roche

DOI: 10.1136/annrheumdis-2018-eular.2826