Begin main content

Triple Conventional Synthetic Disease-Modifying Anti-Rheumatic Drugs for the Management of Rheumatoid Arthritis

Last updated: March 26, 2019
Project Number: RC1087-000
Product Line: Rapid Response
Research Type: Drug
Report Type: Summary with Critical Appraisal
Result type: Report

Question

    What is the cost-effectiveness of triple conventional synthetic disease-modifying anti-rheumatic drugs compared with other pharmacologic options for the management of rheumatoid arthritis?

Key Message

Two relevant publications comprising cost-effectiveness analyses were identified; one involved patients with rheumatoid arthritis (RA) that could not be adequately controlled by methotrexate (MTX) monotherapy and the other was relevant to patients with early aggressive RA. One evaluation found that in patients with MTX monotherapy-resistant RA, triple conventional synthetic disease-modifying anti-rheumatic drug therapy (consisting of a combination of MTX, sulfasalazine, and hydroxychloroquine) was cost-effective relative to a combination of etanercept and MTX (ETN-MTX combination therapy) over 24 weeks or 48 weeks. Under the best-case scenario, the projected incremental cost-effectiveness ratio (ICER) for ETN-MTX combination therapy over triple therapy was $137,000 per quality adjusted life-years (QALYs) gained over a lifetime horizon of 50 years. For the ICER to fall below an assumed willingness-to-pay threshold of $100,000 per QALYs gained, the price of ETN would have to fall by two-thirds. The second evaluation found that triple therapy offered immediately was likely to be cost-effective relative to immediate ETN-MTX combination therapy, step-up ETN-MTX combination therapy, and step-up triple therapy in patients with early aggressive RA. The ICER for immediate ETN-MTX therapy compared with triple therapy was $12.5 million per QALY gained. The ICER for the best case scenario was reported as $5.6 million per QALY gained while $14 million per QALY was the ICER for the worst case scenario. The results from the two evaluations are not comparable as they involve different populations, models, inputs, and assumptions.