Letter to the Editor (Matters Arising)

2 downloads 0 Views 53KB Size Report
These are two separate trials with results presented side-by-side .... 7 Bellamy N, Carr A, Dougados M, Shea B, Wells G. Towards a definition of. ''difference'' in ...
Rheumatology Advance Access published June 27, 2007 Rheumatology 2007; 1 of 1

doi:10.1093/rheumatology/kem162

Letter to the Editor (Matters Arising) Response to Dr Gotzche and Dr Bjarnason SIR, Dr Gøtzche and Dr Bjarnason raise several questions regarding the design and reporting of our study. Their comments are largely based upon inaccurate assumptions and inferences regarding these protocols and data analyses. Specific responses are listed subsequently:

The authors have declared no conflicts of interest. CLIFTON O. BINGHAM Divisions of Rheumatology and Allergy, Johns Hopkins University, Baltimore, MD, USA Accepted 4 May 2007 Correspondence to: C. O. Bingham III, MD, Assistant Professor of Medicine, Divisions of Rheumatology and Allergy, Johns Hopkins University, 5200 Eastern Avenue, Mason F. Lord Building, Center Tower, Room 404, Baltimore, MD 21224. E-mail: [email protected]

1 US Department of Health and Human Services Guidance for Industry. E 9 Statistical principles for clinical trials, 1998. Available at: http://www.fda.gov/cder/guidance/ ICH_E9-fnl.PDF. Accessed 30 March 2007. 2 US Department of Health and Human Services Guidance for industry. E 10 Choice of control group and related issues in clinical trials, 2001. Available at: http:// www.fda.gov/cder/guidance/4155fnl.htm. Accessed 30 March 2007. 3 Birbara C, Ruoff G, Sheldon E et al. Efficacy and safety of rofecoxib 12.5 mg and celecoxib 200 mg in two similarly designed osteoarthritis studies. Curr Med Res Opin 2006;22:199–210. 4 Gibofsky A, Williams GW, McKenna F, Fort JG. Comparing the efficacy of cyclooxygenase 2-specific inhibitors in treating osteoarthritis: appropriate trial design considerations and results of a randomized, placebo-controlled trial. Arthritis Rheum 2003;48:3102–11. 5 Wittenberg RH, Schell E, Krehan G et al. First-dose analgesic effect of the cyclooxygenase-2 selective inhibitor lumiracoxib in osteoarthritis of the knee: a randomized, double-blind, placebo-controlled comparison with celecoxib [NCT00267215]. Arthritis Res Ther 2006;8:R35, Published online 16 January 2006; doi:10.1186/ar1854. 6 Piaggio G, Elbourne DR, Altman DG, Pocock SJ, Evans SJ. Reporting of noninferiority and equivalence randomized trials: an extension of the CONSORT statement. JAMA 2006;295:1152–60. 7 Bellamy N, Carr A, Dougados M, Shea B, Wells G. Towards a definition of ‘‘difference’’ in osteoarthritis. J Rheumatol 2001;28:427–30. 8 Leung AT, Malmstrom K, Gallacher AE et al. Efficacy and tolerability profile of etoricoxib in patients with osteoarthritis: a randomized, double-blind, placebo and active-comparator controlled 12-week efficacy trial. Curr Med Res Opin 2002;18:49–58. 9 Reginster J, Malmstrom K, Mehta A et al. Evaluation of the efficacy and safety of etoricoxib compared with naproxen in two, 138-week randomized studies of osteoarthritis patients. Ann Rheum Dis, published online 1 Dec 2006;doi:10.1136/ ard.2006.059162.

ß The Author 2007. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: [email protected]

Downloaded from http://rheumatology.oxfordjournals.org/ by guest on February 7, 2013

 These are two separate trials with results presented side-by-side in the manuscript, not data pooled from two studies. In no place in our manuscript do we refer to the primary efficacy or safety results as pooled, and the results are described and presented as being from Study 1 or from Study 2 throughout the manuscript.  Regarding the non-inferiority design, we state in the statistical methods subsection that the primary hypothesis was a non-inferiority hypothesis between the two active treatment groups, with the non-inferiority bounds defined. This design was pre-specified and filed with regulatory authorities. We acknowledge that it may have been clearer to also state that these were non-inferiority studies in the initial sentence of the study design subsection.  We disagree with the characterization that the use of noninferiority trials is limited only to scenarios in which the use of placebo is unethical, and that by extension, inclusion of a placebo arm in such a study is unethical. FDA guidance states that treatments can be compared without the objective of showing superiority, and that a placebo arm may be used as a measure of external validity [1, 2], and similar study designs are not uncommon in pain trials [3–5]. As this was the first time etoricoxib and celecoxib had been compared in a head-to-head trial, the inclusion of a placebo arm was to ensure assay sensitivity.  We appreciate that the study design was complex. Technically, there were two placebo groups in each trial, as stated in the study design subsection and shown in Figs 1 and 2. This is because the patients receiving placebo in Part I were switched to one of two active treatments in Part II. However, allocation during all periods of the study was determined at randomization. In other words, subjects were not randomized to one large placebo group in Part I and then re-randomized to an active group in Part II. For the efficacy and safety analyses in part I, the subjects in the two placebo groups were treated as one large placebo group, as they were receiving the same treatment in Part I.  The phrase ‘as effective as’ has been used commonly in describing the results of non-inferiority trials. We understand that the new CONSORT Extension Statement [6] recommends reporting non-inferiority trial results as ‘non-inferior’. We have noted this recommendation for future publications. Interestingly, these same guidelines also describe non-inferiority studies as ‘intended to show whether a treatment has at least as much efficacy’ as the comparator.  Our choice of a 10 mm change (on a 100 mm visual analogue scale) was based on the reported minimally clinically important differences in osteoarthritis [7], and this value has been used in several studies of osteoarthritis pain [4, 8, 9].

Finally, and most troubling, Dr Gøtzche and Dr Bjarnason suggest the description of these studies and the presentation of results were purposefully misleading and inaccurate. As noted above, this is based upon inaccurate assumptions and inferences made by Dr Gøtzche and Dr Bjarnason. These studies were conducted in concordance with international regulatory guidelines, the study designs and analyses were similar to other studies conducted in this area, the protocols were submitted to the FDA prior to study initiation, the data analysis plans were submitted to the FDA prior to freezing and unblinding of the database, the analyses were performed as pre-specified in the protocols and data analysis plans and the studies were listed at ClinicalTrials.gov. Thus, the claims made by Dr Gotzsche and Dr Bjarnason regarding these studies are inappropriate and without merit.