OBJECTIVES: In 2016, the French National Authority for Health (Haute Autorité de Santé, HAS) initiated an economic evaluation of biologic disease-modifying anti-rheumatic drugs (boDMARDs) for Rheumatoid Arthritis (RA). Our objective is to present the estimates of the comparative effectiveness of first-line boDMARDS, and their European Medicines Agency approved biosimilars (bsDMARDs), in terms of ACR (American College of Rheumatology) and EULAR (European League against Rheumatism) responses.

METHODS: A systematic literature review was conducted using eight databases, including MEDLINE and EMBASE, up to January 2017. Randomised controlled trials (RCTs) were sought of etanercept (ETN) and its biosimilar, infliximab and its biosimilars, abatacept, adalimumab, certolizumab pegol, golimumab and tocilizumab (TCZ), where data for licensed dose were extracted for each intervention as combination therapy with conventional synthetic (csDMARD) or as monotherapy. Two populations were considered: methotrexate (MTX)-naïve patients with severe active RA (DAS28≥5.1); and moderate-to-severe active RA (DAS28≥3.2) previously treated with csDMARDs. Outcomes were ACR response and EULAR response at 22-30 weeks follow-up. Network meta-analyses (NMA) were conducted using a Bayesian Markov Chain Monte Carlo approach based on a random effects model.

RESULTS: Forty-six RCTs met the eligibility criteria of the review. In the MTX-naïve population, MTX plus methylprednisolone was most likely, using median values, to achieve the best ACR response. In the csDMARD-experienced population, the greatest median effects for ACR and EULAR responses were associated with TCZ monotherapy, and the following combination therapies (plus MTX): boDMARD ETN, bsDMARD ETN and TCZ.

CONCLUSIONS: The MTX-naïve patients NMA demonstrated no superiority for combination boDMARDs over intensive csDMARDs. In csDMARDs-experienced patients, the NMA identified no statistically significant difference between the boDMARDs and bsDMARDs. The estimates of EULAR and ACR responses should be used as short-term effectiveness inputs in the RA economic model comparing the different therapeutic choices for RA treatment in France.