Dear Sir,

I would like to thank you for showing interest in our work. We have reviewed the letter you wrote to us in response to our latest article entitled “An indirect comparison of etanercept, infliximab, and adalumimab for psoriatic arthritis: mixed treatment comparison using placebo as common comparator” which was published in the Journal of Clinical Rheumatology and available on line from 21 June 2011.

As was stated in the original article, all the studies included in the Mixed Treatment Comparison (MTC)-based analysis had ACR20 as a primary endpoint. It is under this pretext that the study of Mease et al. published by The Lancet was excluded [1]. The study of Mease et al. employed a PsARC score as opposed to a primary endpoint of ACR20. We acknowledge that a mistake occurred within our references that may have caused confusion. However, from the data presented in Table 2, it is clear that these scores were obtained from the impact 2 study conducted by Antoni et al. [2, 3]. We are aware that the protocol’s design gave us the opportunity of switching from one type of therapy to another after a 16-week period. We, however, chose to continue using the data from the impact 2 study for a further 8 weeks. These data were examined and reported, and the possibility of an early “escape” at week 16 was taken into account.

With regard to the ACR20 analysis, we feel that comparing ACR20 results in terms of the percentage of patients showing an ACR20 response would not allow for a direct comparison. It is, therefore, not possible to draw statistically reliable results based on that methodology. On the other hand, we were able to produce an indirect probabilistic comparison by using an MTC-based approach of pooling placebo data. Furthermore, we feel that this was explained comprehensively in our article.

We reported the data from the study of Genovese et al. [4] concerning the use of adalimumab for the treatment of psoriatic arthritis. Additionally, the heterogeneity of the study’s duration (12 weeks) was considered although it was found to not alter the final results. The study’s duration acted as a variable, nevertheless, this disparity was resolved by MTC-based analysis. As our primary endpoint remained ACR20 regardless of the study’s duration, we reject your claim that our study is invalid. Furthermore, the ADEPT study consisted of a larger sample population of patients treated with adalimumab and a follow-up of 24 weeks [4]. This ensured that the adalimumab data were consistent and comparable to the other randomized controlled trials included in this meta-analysis.

We recognize that there was a composition error resulting in transposed data between the articles. We insist that this did not, however, affect the results as these were taken directly from their corresponding studies. These results may also be confirmed by running the same data in WinBUGS.

In conclusion, we acknowledge that there are two mistakes in our aforementioned article. These are located in our reference section and in our column compilation for Table 2. We wish to reiterate that the methodology applied, the data used for the meta-analysis, and the reproducibility of the results strongly confirm the results we reported in our meta-analysis.

Kindest regards,

Prof. Alberto Migliore

Prof. Bruno Laganà

Dr. Serena Broccoli

Dr. Emanuele Bizzi