AN ECONOMIC WINDOW OF OPPORTUNITY FOR PATIENTS WITH EARLY RHEUMATOID ARTHRITIS: 5-YEAR COST-EFFECTIVENESS ANALYSIS OF THE CareRA TRIAL

Background: The CareRA trial showed that remission induction with methotrexate (MTX) and glucocorticoid (GC) bridging in a treat-to-target setting is cost-effective up to 2 years in early Rheumatoid Arthritis (eRA) patients. Objectives: To evaluate the long-term cost-effectiveness of treat-to-target treatments among recently diagnosed (<1 year), DMARD naïve patients with eRA using MTX and a step-down GC scheme (COBRA-Slim) compared to (a) the same combination with either sulphasalazine (COBRA-Classic) or leflunomide (COBRA-Avant-Garde) in high-risk patients and (b) MTX without GCs (Tight-Step-Up: TSU) in low-risk patients up to 5 years. Methods: We used data from the 2-year RCT CareRA trial and its 3-year observational follow-up, CareRA plus. Patients completing the 2-year visit of CareRA were eligible for participation in CareRA plus, in which patients were evaluated every 6 months till year 5. Healthcare costs considered in this piggyback economic analysis were rheumatology visits, RA-related medication (synthetic and biological DMARDs, GCs, and all recorded analgesics including paracetamol, non-steroidals, tramadol and opioids), hospital admissions, laboratory tests and radiographs occurring during the 5-year trial. All pricing is based on December 2021 rates. Total costs per resource were calculated by multiplying the number of resources by the cost unit price extracted from Belgian national websites. Total costs per patient were obtained by summing costs of all resources. Effectiveness was measured with DAS28-CRP and compared between the originally allocated treatment arms. An incremental cost-effectiveness ratio (ICER) was calculated by dividing the cost difference by the DAS28-CRP<2.6 remission difference per pair of treatment schemes. Multiple imputation was used to handle missing data and non-parametric bootstrapping with 25000 iterations of random sampling with replacement to calculate confidence intervals (95% CIs). Results: Of 322 eligible patients, 252 were included in CareRA plus, of which 203 completed the trial. Rates of disease control (DAS28-CRP<2.6) at year 5 in high-risk patients were 68%, 72% and 64% in the Classic, Slim and Avant-Garde group respectively (p=0.63) and related total costs were €11 358.39 (CI 7 776.84-14 939.93), €8 463.12 (CI 6 789.44-10 136.80), €11 752.47 (CI 7 705.11-15 799.82) respectively. In the low-risk population, 80% of patients in Slim and the TSU arm reached remission (DAS28-CRP<2.6) at year 5. While the costs were €6 332.55 (CI 3 607.63-9 057.48) for Slim, and €10 398.19 (CI 4165.95-16630.43) for TSU. In the high-risk group, Classic (ICER -€723.82) and Avant-Garde (ICER -€411.17) were more expensive and less effective compared to Slim. In the low-risk group, Slim was less expensive (Δ -€4 065.64) and equally effective as TSU. Figure 1 depicts how the different medication costs evolved during the 5-year follow-up. 22% of all patients were ever on bDMARDs. More specifically in 23% (16/69) of Classic, 21% (16/75) of Slim high-risk, 25% (15/59) of Avant-Garde, 17% (4/23) of Slim low-risk, and in 15% (4/26) of TSU patients. On average a first bDMARD was started later in the Slim arms, more specifically at week 69 for Classic, week 106 for Slim high-risk, week 97 for Avant-Garde, week 102 for Slim low-risk and week 76 for TSU. Conclusion: The combination of MTX with a GC bridging scheme (COBRA Slim) was more cost-effective (less expensive with comparable disease control) than more intensive step-down combination strategies or a conventional step-up approach 5 years after initial treatment. Over 5 years, around one-fifth of all patients, were in need of starting biological treatment and the transition to a bDMARD was later in COBRA Slim. These results point out the possibility of an early “economic” window of opportunity for diminishing costs long-term while still maintaining optimal disease control. Disclosure of Interests: None declared Citation: , volume 81, supplement 1, year 2022, page 23Session: From risk assesment to societal outcomes (Oral Presentations)