To the Editor:

The article by Hetland and colleagues reported the results of a nationwide longitudinal observation of tumor necrosis factor α (TNFα)–blocking therapy for rheumatoid arthritis (RA) in Denmark (Hetland ML, Christensen IJ, Tarp U, Dreyer L, Hansen A, Hansen IT, et al. Direct comparison of treatment responses, remission rates, and drug adherence in patients with rheumatoid arthritis treated with adalimumab, etanercept, or infliximab: results from eight years of surveillance of clinical practice in the nationwide Danish DANBIO registry. Arthritis Rheum 2010;62:22–32.) While the data set used for these analyses is comprehensive and provides important insights, we suggest that some of the authors' conclusions may not be fully supported by their data.

The authors acknowledge that significantly fewer patients treated with etanercept or adalimumab required concomitant prednisolone and methotrexate therapy (P = < 0.0003 and < 0.0001, respectively); however, the implicit confounding influence of this more aggressive therapy is not discussed. As the article states that all patients were treated with “routine care,” it is likely that those patients requiring substantially more aggressive care may have had significantly more severe and/or difficult-to-manage disease. Moreover, the patients treated with infliximab appeared to have been treated unsuccessfully with significantly more disease-modifying antirheumatic drugs prior to initiating anti-TNFα therapy (P = 0.0044), again introducing the likelihood of more difficult-to-manage disease. Although it is true that there were no significant differences in disease duration or disease activity among the groups at baseline, the groups appeared to have had substantively different disease courses. In fact, it may be reasonable to posit that the clinicians in the study chose to use infliximab more frequently in their most difficult-to-manage cases, possibly because of the flexibility in dosing and infusion frequency with this agent relative to the fixed doses of the injection agents. Despite the attempt to statistically control for confounders, if the infliximab-treated patients represent a subpopulation of RA patients fundamentally different from the rest of the study group, then the logistic regression analysis may not yield interpretable results.

Finally, it is interesting that quantitatively, the largest number of patients had been treated with infliximab. While this may merely represent wider availability of this agent in Denmark, it is also plausible that Danish rheumatologists may reserve the use of the injection agents for their patients with RA that is the easiest to manage. Although a prospective controlled trial would never be practical on the scale of this study, it may be premature to conclude on the basis of these data that the remission rate for infliximab is inferior to that of other TNFα antagonists.

Serene Francis*, Md Joel A. Block MD*, * Rush University Medical Center, Chicago, IL.