Abstract 726: Efficacy and Safety of Etanercept 100 mg in Patients with Rheumatoid Arthritis who are Suboptimal Responders to Etanercept 50 mg
Authors:
M. E. Weinblatt, M. H. Schiff, E. M. Ruderman, C. O. Bingham, III, J. Li, J. Louie, D E. Furst
Background
In RA, certain TNF inhibitors (i.e. infliximab) have been shown to have increased efficacy at higher doses in patients who do not respond adequately to lower doses. In addition, some disorders (i.e. psoriasis) respond better to higher doses of etanercept than those typically used in RA. However, whether additional efficacy can be gained by increasing the dose of etanercept in RA has not been systematically evaluated.
Methods
Patients were eligible for enrollment if they had a suboptimal response (defined as five or more refractory swollen and tender joints of the 28 used in the DAS28 assessment) despite etanercept 50 mg weekly and at least 15 mg of methotrexate per week for at least six months. Eligible patients were randomized to receive either etanercept 100 mg per week or etanercept 50 mg per week in a 4:1 design. For the first 12 weeks after randomization, patients received drug in a double-blind fashion. At 12 weeks, subjects entered an open-label phase in which non-responders to etanercept 50 mg per week were crossed over into the higher dose group and non-responders to etanercept 100 mg per week were discontinued from the study. The primary outcome measure was achievement of a EULAR good or moderate response as 12 weeks. Secondary outcomes were ACR 20, 50, and 70 percent responses, change in DAS28 score, and change in individual ACR core set measures at 12 weeks. Safety assessments were obtained throughout and serum concentration of etanercept and anti-etanercept antibodies were measured at week 12.
Results
Among the 200 subjects who received at least one dose of study drug, 187 completed 12 weeks of follow-up and 102 completed the entire 24 weeks. There was differential loss to follow-up in the etanercept 100 mg group for the 24 week outcome, as 75 of the 160 subjects randomized to the higher etanercept dose discontinued at 12 weeks due to lack of response.
Due to the comparatively smaller sample size in the control group, there were some baseline differences in demographics and RA characteristics between the two treatment groups. On average, the etanercept 100 mg group contained a greater proportion of women (84% vs. 70%), a lower proportion of rheumatoid factor positive participants (61% vs. 73%), and the mean age was slightly greater (53 ± 12 years vs. 51 ± 14 years) compared to the 50 mg group. However, the racial distribution (85% Caucasian), mean RA disease duration (approximately 10 ± 8 years), mean baseline methotrexate dose (approximately 17.5 mg per week), and baseline numbers of swollen and tender joints were comparable between groups.
At the end of the double-blind period, 46% of subjects in the etanercept 100 mg group met the primary endpoint compared to 35% in the lower dose continuation group (p=0.286). In addition to the primary outcome, higher dose of etanercept was not efficacious in any of the secondary outcomes. Higher etanercept levels, on average, were seen in subject on the higher dose and no patients were found to have anti-etanercept antibodies. Serious adverse events were higher in the etanercept 100 mg group. Serious infections were statistically similar between the two treatment groups, although numerically there were 3 serious infections in the etanercept 100 mg group vs. none in the 50 mg group.
Conclusions
Doubling the dose of etanercept in the face of non-response to standard dosing was not associated with clinical improvement. At the same time, adverse events may be increased.
Editorial Comment
Increasing the dose of etanercept in the face of inadequate response has undoubtedly been tried off-label in clinical practice, but has not been a widespread phenomenon. This study should effectively eliminate the practice. Why there appears to be a dose response to other TNF inhibitors in RA may be related to differences in mechanism between monoclonal antibody and decoy receptor TNF inhibitors. The finding that etanercept levels were increased in the serum, and that increases in etanercept levels did not predict response, suggests that the reason for a lack of sufficient efficacy was not related to the pharmacodynamics of the drug. Despite an only 10% difference in efficacy (and an even smaller effect size), the risk of serious infection seemed to be a little higher for the higher dose etanercept group, suggesting that the mechanisms of benefit vs. harm may not operate on the same scale of magnitude.
