To the Editor:

We thank Drs. Francis and Block for their interest in our study comparing the effect of 3 TNFα inhibitors (adalimumab, etanercept, and infliximab) in clinical use and for their reflections regarding the results of our investigations. Based on a nationwide longitudinal observation of 2,326 patients with RA treated in Denmark, we calculated the treatment responses, disease remission rates, and treatment adherence rates after correction for differences in sex, age, disease duration, seropositivity, Disease Activity Score in 28 joints (DAS28), treatment with concomitant methotrexate (MTX) and prednisolone, the Health Assessment Questionnaire score at baseline, and the study center. We reported that infliximab had the lowest treatment responses, disease remission rates, and treatment adherence rates. Adalimumab had the highest treatment responses and remission rates, whereas etanercept had the longest drug survival rates.

Surprisingly few studies (and no head-to-head randomized clinical trials [RCTs]) have attempted to compare the efficacy of the individual TNF inhibitors. With a lack of comparative RCTs, the observational cohorts of real-life patients represent the best available data. Francis and Block suggest that the infliximab-treated patients might represent a subpopulation of RA patients who are fundamentally different from the rest of the study group, characterized by more severe, and more difficult-to-manage, disease. There is, however, significant evidence that this is not the case: in Denmark, treatment with biologic agents is reimbursed by the public health system, and only the hospital departments of rheumatology, not the private practitioners, can prescribe and administer biologic treatments. Infliximab has been considered the cheapest of the TNF inhibitors, and for financial reasons some hospitals (corresponding to 37% of the patients in our study) have issued choice-of-drug recommendations, in favor of infliximab as the preferred biologic drug. In these areas, infliximab is prescribed as the first biologic treament in >80% of patients, and subcutaneous treatments are reserved for patients who, for example, cannot tolerate concomitant treatment with MTX.

In the Danish treatment strategy, the use of concomitant MTX is not associated with more severe RA (as Francis and Block suggest), but merely reflects compliance with the evidence from several RCTs of a superior radiographic outcome in patients receiving TNF inhibitors in combination with MTX, as compared with patients receiving TNF inhibitors as monotherapy. As a result, it is likely that patients who do not take concomitant MTX may experience various comorbid conditions such as pulmonary or hepatic diseases that contraindicate the use of MTX. Based on this, one would expect the patients treated with etanercept and adalimumab to be more difficult to treat. In the preparation of our report, we performed a subanalysis that showed similar treatment effects and treatment adherence rates in patients treated in hospitals applying the recommendations of infliximab as the first-line biologic drug versus those treated according to the physician's free choice of drug.

Until 2003, infliximab was more prevalent as the first TNF inhibitor because of a shortage of etanercept, and because adalimumab was not yet on the market. The patients who were treated before 2003 had more longstanding and treatment-resistant disease, as judged by the use of concomitant prednisolone as well as the number of disease-modifying antirheumatic drugs prescribed prior to initiation of the TNF inhibitor. However, as stated in our report, we performed a subanalysis, in which patients who were treated soon after the earliest TNF inhibitors were marketed were excluded, and we found similar results with regard to differences in efficacy and adherence to therapy between the 3 drugs. An additional subanalysis, in which all patients treated with prednisolone were excluded, also showed results similar to those published.

Our findings are supported by those of other research groups. A study of patients in the Dutch Rheumatoid Arthritis Monitoring registry showed larger DAS28 improvements in adalimumab- and etanercept-treated patients than in infliximab-treated patients (1), and discontinuation of treatment was significantly higher in infliximab-treated patients as compared with the patients receiving the subcutaneous treatments. Similarly, a Swiss study showed poorer drug survival for infliximab therapy (2), and a French study showed poorer drug survival for infliximab therapy as compared with etanercept therapy (3).

Thus, evidence of differences in efficacy and drug treatment adherence between individual TNF inhibitors is emerging from several cohorts of RA patients treated during routine care.

Merete Lund Hetland MD, PhD*, Louise Linde MD*, Ib Jarle Christensen MSc†, Lene Dreyer MD, PhD†, Ulrik Tarp MD, DMSc‡, Annette Hansen MD§, Ib Tønder Hansen MD, PhD¶, Gina Kollerup MD, PhD**, Hanne M. Lindegaard MD, PhD††, Uta Engling Poulsen MD‡‡, Annette Schlemmer MD§§, Dorte Vendelbo Jensen MD¶¶, Signe Jensen MD11, Gisela Hostenkamp MScEcon12, Mikkel Østergaard MD, PhD, DMSc13, * DANBIO Registry, and Copenhagen University Hospitals, Hvidovre and Glostrup, Denmark, † Copenhagen University, Hospital Rigshospitalet, Copenhagen, Denmark, ‡ Aarhus University Hospital, Aarhus, Denmark, § DANBIO Registry and Copenhagen University Hospital, Gentofte, Denmark, ¶ Viborg Hospital, Viborg, Denmark, ** Copenhagen University Hospital, Frederiksberg, Denmark, †† Odense University Hospital, Odense, Denmark, ‡‡ University of Southern Denmark, Graasten, Denmark, §§ Aarhus University Hospital, Aalborg, Denmark, ¶¶ Hørsholm Hospital, Hørsholm, Denmark, 11 Copenhagen University Hospital, Gentofte, Denmark, 12 Institute for Public Health, Odense, Denmark, 13 Copenhagen University Hospitals, Hvidovre, Glostrup, and Gentofte, Denmark.