Sci-Reykjavik2007.org

Discussion

Adalimumab, administered with or without methotrexate, improved signs and symptoms of disease in children with juvenile rheumatoid arthritis. Disease flares were significantly less frequent among children receiving adalimumab than among those receiving placebo. Disease flare was defined conservatively in this study, with a relatively small degree of worsening exceeding the threshold for a flare. Therefore, many patients met the criteria for a disease flare while showing substantial improvement as compared with baseline disease activity. The ACR Pedi scores at week 48, which defined all patients who had disease flares as having no response, were significantly greater for patients treated with adalimumab than for those receiving placebo, both among patients receiving methotrexate and among all patients regardless of whether they received methotrexate, but not among patients not receiving methotrexate. Of patients treated with adalimumab, 30% not receiving methotrexate and 42% receiving methotrexate had an ACR Pedi 90 response, a measure that we believe has not previously been included in an efficacy trial of juvenile rheumatoid arthritis (Table 3). During 2 further years of adalimumab treatment in the open-label extension phase, ACR Pedi responses were sustained, including achievement of ACR Pedi 100 responses by 40% of the patients.

The study was not statistically powered to detect differences between patients receiving and those not receiving methotrexate; however, the proportions of patients with ACR Pedi 30, 50, 70, or 90 responses were somewhat higher among those receiving adalimumab in combination with methotrexate than among those receiving adalimumab without methotrexate. Although these results are consistent with findings of studies in adult patients with rheumatoid arthritis,8,14 any differences between patients in the two strata (those receiving and those not receiving methotrexate) are difficult to interpret, because the patients underwent randomization within each stratum but not across strata. Eleven of 86 patients not receiving methotrexate withdrew from the study before undergoing randomization, because of lack of efficacy of adalimumab.

A small percentage of patients withdrew because of adverse events. The most frequently reported adverse events were infections and injection-site reactions. Serious adverse events considered possibly drug-related by the investigator occurred in 14 patients, 7 of whom had serious infections. No deaths, opportunistic infections, malignant conditions, demyelinating diseases, or lupuslike reactions occurred in this study. The size of the study population and the length of the study were not sufficient to determine the risks of rare adverse events.

Approximately 16% of the patients had at least one positive test for anti-adalimumab antibody during the study. This percentage is greater than the 5% observed during clinical trials of adult patients with rheumatoid arthritis.14 There were no increases in discontinuations of the study drug or adverse events associated with the presence of anti-adalimumab antibody. Positive anti-adalimumab antibody tests were less frequent among patients receiving concomitant methotrexate than among those receiving adalimumab monotherapy, a finding consistent with the findings of trials in adult patients with rheumatoid arthritis.

Conducting placebo-controlled trials in pediatric populations requires special consideration of the ethical issues associated with denying active treatment during a double-blind phase. At the time the adalimumab trial was designed, two other trials of TNF antagonists in pediatric patients were ongoing. Our trial of adalimumab was designed after considering both parallel and randomized approaches to withdrawal, in consultation with the Food and Drug Administration (FDA). The FDA and the study designers agreed that the blinded, randomized medication-withdrawal design provided acceptable scientific rigor while minimizing exposure to placebo in this population of children with severe, active juvenile rheumatoid arthritis and that the primary end point should be a comparison, in the no-methotrexate stratum, between the percentages of patients with disease flares in the group receiving adalimumab and the group receiving placebo during the double-blind phase.

Although the double-blind phase, and thus the primary end-point analysis, was conducted in patients who had a response, the open-label lead-in approach is generalizable to clinical practice, given that treatment decisions are made in an open-label setting after a reasonable trial with a new active treatment. If a patient does not have a response to an active treatment after a period of time (16 weeks or so, as in the current trial), a physician will probably consider other treatment options. For patients who do have a response, treatment will be continued.

Adalimumab, alone or in combination with methotrexate, appears to be an efficacious option for the treatment of children with polyarticular juvenile rheumatoid arthritis. Responses were sustained through 2 years of continued treatment.

Supported by a research grant from Abbott Laboratories.