V. Stouten

OP0172 A Cost-Effectiveness Analysis of Different Intensive Combination Therapies for Early Rheumatoid Arthritis: 1 Year Results of The Carera Trial

Background The (Care in Early RA) CareRA trial compared different treatment strategies in patients with early Rheumatoid Arthritis (ERA) with or without markers of poor prognosis. No significant differences were shown between the treatment arms in terms of remission rate and functionality at 16 and 52 weeks. The differential cost-effectiveness of these therapeutic strategies could aid in determining the optimal approach for ERA. Objectives To compare the cost-effectiveness of different therapeutic strategies used in the CareRA trial during the first treatment year in patients with early Rheumatoid Arthritis. Methods The CareRA trial is a 2-year investigator-initiated randomized pragmatic superiority trial comparing remission induction regimens in a treat-to-target approach. DMARD naïve ERA patients were stratified into a high- or low-risk group based on prognostic markers. High-risk patients were randomized to a COBRA Classic (Methotrexate (MTX) + Sulphasalazine + prednisone step-down from 60mg), COBRA Slim (MTX + prednisone step-down from 30mg) or COBRA Avant-Garde (MTX + Leflunomide + prednisone step-down from 30mg) scheme. Low-risk patients were randomized to MTX tight step-up (MTX-TSU) or COBRA-Slim. The direct costs for each patient in the trial were calculated by summing the cost of a rheumatology consultation, per protocol CareRA study medication and RA-related medication such as biologicals and extra glucocorticoids not stipulated in the protocol. The incremental cost-effectiveness ratio (ICER) was calculated using these costs and the treatment efficacy as observed in CareRA. Efficacy parameters investigated were the proportion of patients in remission (DAS28(CRP) <2.6) and the proportion with a clinically meaningful improvement of Health Assessment Questionnaire (HAQ change >0.22) at week 52. Results 349 out of 379 patients had sufficient cost information to be evaluated at week 52. Remission was achieved in 65,6% Classic, 61.7% Slim (high-risk) and 64.1% Avant-Garde patients; and in 56.8% MTX-TSU and 70.0% Slim (low-risk) patients. A clinically meaningful HAQ response was reached in 66.7% Classic, 70.2% Slim (high-risk) and 75.6% Avant-Garde patients; and in 58.3% MTX-TSU and 60.0% Slim (low-risk) patients. Mean ±SD total costs during year one were €1044.3 ± €871.0 €859.7 ±€431.3 and €1267.7 ±€632.7 for the Classic, Slim (high-risk) and Avant-Garde respectively; and €880.6 ±€917.4 and €867.5 ±€364.0 for MTX-TSU and Slim (low-risk) patients. In MTX-TSU patients, the high average cost ±SD of €268.6 ±€876.5 was remarkable. Cost-effectiveness analysis in the high-risk arm showed a higher ICER for Classic (€47.3/1% remission) and Avant-Garde (€170.0/1% remission) compared to Slim (high-risk); In the low-risk arm, the cost-effectiveness of MTX-TSU was shown to be comparable to Slim (low-risk), with an ICER of €-0.99/1% remission. Conclusions Cobra Slim, a combination of MTX with a moderately dosed glucocorticoid remission induction showed favorable cost-effectiveness in both ERA patients with or without poor prognosis. Disclosure of Interest None declared

OP0301 Two year cost-effectiveness analysis of the carera trial in early ra: a piggy back study

Background Rheumatoid arthritis (RA) causes high individual, medical and societal costs. EULAR guidelines suggest treating early, intensively and to target using disease modifying anti-rheumatic drugs (DMARDs), preferably with initial glucocorticoid (GC) bridging. COBRA slim, a combination of methotrexate (MTX) with a moderate dose prednisone step down bridge scheme showed a positive efficacy/tolerability balance in the Care in early RA (CareRA) trial. COBRA Slim in comparison to DMARD combination therapy with GC bridging, has the necessary intensity to induce remission, but with a lower risk of severe discomfort or adverse events, decreasing the early need for biologic (b)DMARDs. Objectives Perform an economic evaluation on the 2 year pragmatic randomised CareRA trial. Methods Patients with early RA (≤1 year) naïve to DMARDs were randomised to monotherapy or synthetic (cs)DMARD combination with or without GC bridging, after risk stratification based on classical prognostic markers. Clinical and patient-reported data were collected at each visit (≥10 times in 2 years). Direct costs of visits and RA medication (systemic GCs, cs and bDMARDs) over 2 years were calculated for each patient from each of the 5 treatment arms (table 1). For cost-effectiveness analysis, benefits were expressed as the proportion of patients with DAS28CRP<2.6 at year 2. Missing data was imputed per item with expectation maximisation. For cost-utility analysis, utilities were calculated using a validated mapping algorithm reconstructing EQ-5D scores based on age, sex, HAQ and pain scores at relevant study visits. Quality-adjusted life years (QALYs) encapsulating the impact of treatment on a patient’s length of life and health-related quality of life, were calculated as the time-weighted average of all available EQ-5D scores (area under the curve). Incremental cost-effectiveness ratios (ICERs) from each strategy were calculated. ICERs compare the additional costs a strategy imposes over another with the additional benefits it delivers. Means and medians based methods were calculated with confidence intervals via bootstrapping. Results From the initial CareRA cohort (n=379),cost/benefit data of 326 patients was used for a 2 year economic analysis. The mayor driver of direct costs was bDMARDs (57%>87% of total costs). Number of consultations were comparable (±11) across all treatment strategies. The cost-effectiveness analysis in the high risk population showed a higher ICER for COBRA Avant Garde (mean €198.65/1%, median €78.41/1%) and a dominated ICER for COBRA Classic (mean €-181.40/1%, median €-35.01/1%) compared to the Slim. In the low risk arm, ICERs for COBRA Slim compared to Tight Step Up (TSU) were €46.75/1% (mean) and €43.64/1% (median). Cost-utility analysis in the high risk arm showed an incremental cost of €1 469.36 for an increased utility of 0.012555 QALYS for COBRA Classic compared to COBRA Slim, resulting in an ICER of €117 033.85/QALY. The ICER of COBRA Avant Garde vs COBRA Slim was €69 329.19/QALY. In the low risk arm, the comparison of COBRA Slim to TSU yields an ICER of €1 342.78 per QALY.Abstract OP0301 – Table 1 Conclusions COBRA Slim which consists of an initial combination of MTX and a moderate dosed GC remission induction scheme has a favourable cost-effective and cost-utility profile for patients with early RA independent of their prognostic factors. Disclosure of Interest S. Pazmino: None declared, R. Westhovens: None declared, J. Joly: None declared, V. Stouten: None declared, D. De Cock: None declared, K. Van der Elst: None declared, P. Verschueren Grant/research support from: Unrestricted Pfizer Chair in the management of early rheumatoid arthritis

FRI0726 Long term use of analgesics and nsaids in early ra: lessons from the carera study

Background One might consider pain accompanying musculoskeletal conditions as a separate illness entity deserving specific treatment. A subgroup of early Rheumatoid Arthritis (RA) patients has remaining pain despite adequate disease control and this might be reflected in the use of analgesics and NSAIDs. Objectives To investigate the usage of analgesics and antiphlogistics prospectively in the pragmatic randomized controlled CareRA trial and describe the users of such drugs taking into account body mass index (BMI), VAS pain and DAS28CRP. Methods This study utilized data from the CareRA trial, a 2-year prospective investigator-initiated multicentre pragmatic RCT in patients with early RA (≤1 year) comparing different early intensive treat to target strategies, including glucocorticoid (GC) bridging, aiming for remission in all participants. All concomitant medication for each patient was recorded, including: name, dose, frequency/timing, continuous/intermittent use, start/end date and indication (possibly/definitely RA-related or other). Meaningful sub-classifications were made (analgesics, cox2-selective and non-selective NSAIDs). We defined two subgroups with prolonged analgesic/NSAID use till w104: the pre-induction group starting before/at baseline (BL) and the post-induction group starting w28-w52, after GC step-down. Results From the CareRA cohort (n=379), 90% were on analgesics/NSAIDs at some time during the study and 84% had already started before inclusion (mean 24 weeks). In general, we observed a decreasing trend in the use of analgesics/NSAIDs, with a slight increase after GC step-down. Overall, 266 patients used non-selective and 131 cox2-selective NSAIDs, 154 paracetamol and 85 opioids. Of the total population, 15, 5% (53 patients) started long-term analgesics/NSAIDs before/at BL (pre-induction group) and 13% (45 patients) from w28-w52 (post-induction group). A detailed distribution of on-demand/daily analgesic/NSAID use is provided in the table below.Table 1 Group of long-term users NSAIDs Paracetamol Opioids Combined therapy Non-selective Selective COX2 Pre-induction (n=53) 26% (17%/d) 11% (4%/d) 45% (8%/d) 9% (6%/d) 9% (2%/d) Post-induction (n=45) 33% (16%/d) 20% (7%/d) 33% (4%/d) 13% (7%/d) 0% Proportionally more patients were using NSAIDs in the post-induction group and paracetamol in the pre-induction group. Mean DAS28CRP (BL-w104) in the pre-induction group was 3.11 (±0.70) and mean VAS pain 32.7 (±16.3). In this group 62.3% had sustained low disease activity while taking continuous analgesics/NSAIDs. Results in the post-induction group were comparable. Analgesic/NSAID use was not significantly associated with mean (BL-w104) BMI, DAS28CRP or VAS pain. Conclusions Analgesics/NSAID intake in early RA was high (90%) and remarkably about 30% continued using these drugs, despite overall low disease activity. This high consumption might be explained by the ambiguous nature of arthritis-related pain and the lack of differentiation between nociceptive and non-nociceptive pain. Therefore pain management in early RA deserves more attention. References Verschueren, P. et al. ffectiveness of MTX with step-down glucocorticoid remission induction (COBRA Slim) vs other intensive treatment strategies for early RA in a treat-to-target approach: 1-year results of CareRA, a randomised pragmatic open-label superiority trial. Ann Rheum Dis. Disclosure of Interest None declared

AB0995 Does Bmi Increase in Patients with Early Rheumatoid Arthritis When Treated with Short-Term Glucocorticoid Remission Induction Schemes?

Background Glucocorticoids (GCs) can rapidly reduce disease activity in many chronic inflammatory diseases. Numerous adverse events have been attributed to GCs. One of the most commonly mentioned and feared side effects is an increase in weight and BMI, especially with chronic GC use. However, data remains scarce about the impact of short term glucocorticoid remission induction schemes on BMI in patients with Early RA (ERA). Objectives To explore the BMI evolution of patients with ERA treated with short term glucocorticoid remission induction schemes in the (Care in Early RA) CareRA trial. Methods The CareRA trial is a RCT comparing remission induction regimens in a treat-to-target approach. DMARD naïve ERA patients were stratified into a high- or low-risk arm based on classical prognostic markers. High-risk patients were randomized to COBRA Classic (Methotrexate (MTX) + Sulphasalazine + prednisone step-down from 60mg), COBRA Slim (MTX + prednisone step-down from 30mg) or COBRA Avant-Garde (MTX + Leflunomide + prednisone step-down from 30mg) scheme. Low-risk patients were randomized to MTX tight step-up (MTX-TSU) without oral GCs or COBRA Slim. Prednisone was tapered down over 6 weeks to 7.5mg in Classic and 5mg in the other COBRA arms. Oral GCs were discontinued from Week 28 and stopped at week 34. Demographics including BMI and disease parameters were routinely registered. For this analysis, only patients with all BMI parameters on all study visits during the first 52 treatment weeks were included. BMI change per group was tested with a paired t-test. Differences in BMI change between groups were tested with a students t-test, Spearmans rho was used for correlation between variables of interest and BMI change. Finally, an adjusted linear regression model further explored the effect of treatment choice on BMI change. Results 281 Of 379 patients were included. Figure 1 describes the evolution of BMI over 52 weeks for the different treatment strategies. Only patients on Cobra Classic had an increase in BMI after 52 weeks (p=0.003). In the High-Risk arm, change in BMI over 52 treatment weeks was larger in Classic versus Slim (High-Risk) (p<0.039) and in Classic versus Avant-Garde (p<0.001). In the Low-Risk arm, no difference between groups was found. No correlations were detected between change in BMI and cumulative GC dose, DAS28(CRP) change nor HAQ change. In the High-Risk arm, linear regression revealed a statistically significant effect of treatment strategy on BMI change, even when corrected for age, gender, GC cumulative dose, DAS28(CRP) change and HAQ change (p=0.014). No effect of treatment strategy on BMI change was seen in the Low-risk group in a similar linear regression model. Conclusions In general, BMI changes were small with any treatment strategy. A statistically significant increase in BMI over 52 weeks was only seen with a remission induction scheme starting at a high GC dose in patients with poor prognosis ERA. No difference in BMI change over 52 weeks was seen between a remission induction scheme starting at a moderate GC dose and a treatment strategy without oral GCs in patients with good prognosis ERA. Disclosure of Interest None declared