A cost-effectiveness analysis of treatment options for patients with methotrexate-resistant rheumatoid arthritis

Objective. Recently, new treatment options for rheumatoid arthritis (RA) pa tients with an inadequate response to methotrexate (MTX) have become availa ble. Given the wide variability in efficacy and costs among these different treatment options, we sought to determine their cost-effectiveness (CE) in order to guide policy in different cost-constrained settings. Methods. We performed a CE analysis comparing 6 treatment options for patie nts with MTX-resistant RA: 1) etanercept + MTX, 2) etanercept monotherapy, 3) cyclosporine + MTX, 4) triple therapy (hydroxychloroquine, sulfasalazine , and MTX), 5) continuation of MTX monotherapy, and 6) no second-line agent . A decision model was used with a time horizon of 6 months. We used 2 meas ures of effectiveness based on published clinical trial data: the American College of Rheumatology 20% response criteria (ACR 20); and a weighted aver age of proportions of patients achieving responses of ACR 70, ACR 50, and A CR 20 (ACR 70 weighted response [ACR 70WR]). Incremental CE ratios were cal culated as the additional cost per patient achieving either outcome, compar ed with the next least expensive option. To help interpret CE relative to t hese RA-specific outcomes, we conducted a separate, "reference" CE analysis of MTX use in MTX-naive RA patients, using the same outcomes. Results. In our reference analysis, MTX therapy for MTX-naive RA cost $1,10 0 per ACR 20 outcome and $1,500 per ACR 70WR, compared with no second-line agent. In our base-case analysis with either outcome, MTX continuation, cyc losporine + MTX, and etanercept monotherapy cost more, but either were not more efficacious or had a higher incremental CE ratio than the next most ex pensive option (i.e., they were dominated). Therefore, these options were n ot cost-effective. The least expensive option, triple therapy, cost 1.3 tim es more per patient with ACR 20 outcome ($1,500/ACR 20) and 2.1 times more per ACR 70WR ($3,100/ACR 70WR) than MTX therapy for MTX-naive RA. The most efficacious option, the combination of etanercept and MTX, cost 38 times mo re per patient with ACR 20 outcome ($4,600/ACR 20) and 23 times more per AC R 70WR ($34,800/ACR 70WR) than MTX therapy for MTX-naive RA. Overall, the r esults of extensive sensitivity analyses did not substantially affect these results. Conclusion. Our analysis indicates that if 15 mg/week MTX is cost-effective for achieving ACR 20 or ACR 70WR in MTX-naive RA over a 6-month period, th en most likely so is triple therapy in MTX-resistant RA. Whether etanercept + MTX is cost-effective depends on whether $34,800/ACR 70WR (or $42,600/AC R 20) over a 6-month period is considered acceptable.