Enbrel, Actemra May Be Best Monotherapies for csDMARD-intolerant RA

Meta-analysis found differences in effectiveness but not in harm across biologics

Diana Swift October 11, 2016

Action Points

A meta-analysis in Seminars in Arthritis & Rheumatism has cautiously concluded that etanercept (Enbrel) or tocilizumab (Actemra) may be the optimal choices for most patients with rheumatoid arthritis unsuitable for conventional standard disease-modifying anti-rheumatic drugs (csDMARDs) who require biological monotherapy.

The study, led by Simon Tarp, PhD, of Bispebjerg og Frederiksberg Hospital in Copenhagen, Denmark, analyzed 28 randomized trials, published from 1998 to 2015 and including a total of 79 unique arms and 8,602 RA patients. Treatments across trials varied as follows: abatacept (two trials, 350 patients); adalimumab (six trials, 1,928 patients); anakinra (one trial; 472 patients); certolizumab pegol (two trials, 421 patients); etanercept (five trials, 2,047 patients); golimumab (four trials, 1,279 patients); infliximab (one trial, 58 patients); rituximab (one trial; 80 patients); and tocilizumab (six trials, 1,967 patients).

Trial design varied from comparing a biological agent in monotherapy with placebo or with methotrexate to comparing two biological agents in monotherapy. Eight trials included DMARD-naïve patients and 20 included DMARD-inadequate responder patients. The primary efficacy endpoint was an American College of Rheumatology response of 50% (ACR50).

"This study suggests that there are potential differences in effectiveness but not in harm among biological agents used in monotherapy in patients with RA," Tarp and associates wrote. Anakinra and infliximab excepted, most of the biologics were statistically significantly more likely than placebo to achieve ACR50 (P<0.05). And compared with for placebo, withdrawals due to adverse events were not statistically significantly higher among any of the drugs (P>0.068).

The researchers found that while tocilizumab was superior to a higher number of agents than etanercept, no statistically significant difference between the two agents emerged in any analysis. For infliximab, no statistically significant differences were found compared with placebo, other biological agents in monotherapy, or methotrexate, but the confidence in the estimates was substantially limited as only 44 patients received infliximab monotherapy.

In other findings, etanercept and rituximab were found to be superior to anakinra (P=0.018 and P=0.049, respectively); and tocilizumab was superior to adalimumab (P=0.0082), anakinra (P=0.0083), certolizumab (P=0.037), and golimumab (P=0.049).

No differences emerged among etanercept, tocilizumab, and rituximab (P>0.52), the authors reported. "However, because rituximab was evaluated in just 40 patients, our confidence in the estimates is limited."

When only DMARD-inadequate responder trials were included, the same statistical pattern was evident. Etanercept and tocilizumab were superior to abatacept, and at the recommended doses, both etanercept and tocilizumab were superior to adalimumab and certolizumab.

Last year a Danish study reported acceptable 2-year remission rates with biologic monotherapy in csDMARD-intolerant patients.

"Our findings are relevant to inform decisions for RA patients in whom csDMARD therapy is inappropriate," Tarp and colleagues wrote. "Evidence from randomized trials suggests that most biological agents are effective as monotherapy. However, given our limited confidence in the estimates, including the possibility of bias, it is appropriate to strongly weight patients' preferences and values in the final treatment choice."

Concurring with these caveats, Petros Efthimiou, MD, of Weil Cornell Medical College in New York City, told MedPage Today, "I think the only safe conclusion from this study is that monotherapy with most biologic agents is effective. This is important, because many patients cannot take methotrexate because of contraindications, or elect not to take methotrexate because of the need for behavioral modifications such as abstinence from alcohol or concerns about toxicity and infertility."

He added, however, that study limitations and the potential for inherent bias made comparison of biologic DMARDs' efficacy as monotherapy problematic.

Among these limitations, the authors noted the 17-year publication span of the studies, with the changing characteristics of eligible patients. In addition, the participants across studies were heterogeneous, and the target population of patients unsuitable for csDMARDs was under-represented. Furthermore, only one head-to-head trial was identified, so future head-to head studies of biologic monotherapy would likely have an important impact on the estimates of efficacy.

The analysis also gave priority to 6-month data, using shorter or longer time points when necessary. "Comparisons among studies across different time points could potentially limit the interpretation of our results," Tarp et al wrote. "Further, whether our results can be extrapolated to long-term efficacy and safety is not clear."

There is also a potential for bias in the imputation method for discontinued patients in evaluating ACR50, the team added. In addition, an ACR50 improvement may be a suboptimal endpoint for response, since a patient can have an ACR50 response but still have moderate or high disease.

This study was supported by an unrestricted grant from Roche, Denmark.

Several study authors disclosed having various financial relationships with industry.