• DMARDs (biologic)
• anti-tnf
• epidemiology
• outcomes research

Thank you for the interest in our publication based on data from the Danish DANBIO registry regarding 2061 patients who were eligible for a mandatory non-medical switch from originator to biosimilar etanercept in routine care.1 2 The marketing of biosimilars has changed the landscape of the biological drugs with the potential for huge cost reductions, most markedly if patients may be switched from an ongoing successful treatment with the expensive originator to a much cheaper biosimilar. Thus, up to 75% price reduction has been experienced in Denmark so far with no evidence of increased use of health resources.3 The outcomes of a non-medical switch of etanercept in the real-world setting are, however, largely unknown, since prevous publications on switching in routine care have included limited numbers of patients. The aim of the study was therefore to investigate the effectiveness of a large-scale, non-medical switch—including to characterise the patients who were not switched despite the national guideline, as well as those who switched back to the originator.

This study is an example of how observational studies constitute a valuable supplement to randomised trials and provide insight regarding the performance of a drug in large, unselected patient groups.4–6 In our publication, the strengths and limitations of the observational design are carefully discussed, which Cantini and Benucci1 largely ignore in their letter published recently in this journal. For example, the observed differences in the demographic and clinical characteristics of switchers compared with non-switchers illustrate that despite a national guideline, the clinical decision to switch a patient or not was associated with certain patient characteristics. Thus, patients with more comorbidities, higher disease activity and prior failed biological treatments were less likely to be switched. This important finding may reflect uncertainty among patients and rheumatologists on how to implement a newly introduced biosimilar in routine care, and it explains why the patients who maintained treatment with the originator drug had poorer retention to treatment compared with the patients who switched to the biosimilar. Furthermore, subjective negative expectations (the nocebo effect) may affect biosimilar treatment outcomes.7

The patients in the study suffered from rheumatoid arthritis, psoriatic arthritis and axial spondyloarthritis, diseases well known for fluctuations in inflammatory activity over time. Even patients in remission may experience flares. Therefore, a time window of 3 months and the use of patients as their own controls were chosen in the evaluation of disease activity prior to and following the switch. Increasing the time windows to 1 year as suggested by Cantini and Benucci1 would significantly have decreased the ability to attribute a flare to the switch.

A major point raised by Cantini and Benucci relates to the treatment regimen in the non-switchers. Their argumentation is not correct; as stated in the Results section, it was less than 15% that received 25 mg etanercept weekly, not 43% as claimed by Cantini and Benucci.

We fully agree that properly designed randomised controlled clinical switch studies such as the NOR-SWITCH study for infliximab are highly needed.8 It is reassuring that observational switch data on infliximab from DANBIO were in agreement with those reported in NOR-SWITCH.9 It is our opinion that our paper is a well-balanced contribution to the ongoing discussion on real-world effectiveness of biosimilar etanercept in patients with inflammatory arthritis. Randomised clinical trials provide strong evidence regarding treatment effects and safety of biosimilars in strictly selected patient groups with tight monitoring and short follow-up. On the other hand, observational studies allow us to explore outcomes in unselected patient cohorts representing the whole disease spectrum and—as discussed above—may also provide knowledge on how the biosimilars are implemented and perform in the real-world setting.6 At the end of the day, the decision on whether and how to carry out non-medical switching relies on evidence generated from both research methods.