To the Editors:

We appreciate the interest shown by Drs. Catrina and Klareskog in our recent article on the effect of DMARDs on the expression of OPG and RANKL in synovial tissue from patients with RA.

However, they have misquoted what we stated in our article with regard to their own results, which have been previously published (1, 2). We did not state that Catrina et al (1) demonstrated a relationship between changes in OPG expression and disease activity, because that was the finding in our study. In addition, we proposed several reasons for the difference in interpretation of the results, including patient population, disease duration, clinical response to therapy, and coprescription of methotrexate and corticosteroids in the study by Catrina et al. None of these potential explanations for the different interpretation of the results are addressed in the letter by Catrina and Klareskog. It should be noted that the results published by Catrina et al rely on 2 small groups of patients (9 patients receiving etanercept in group 1 and 9 patients receiving infliximab in group 2) with significant coprescription of DMARDs (7 of 9 in group 1 and 9 of 9 in group 2) and prednisolone (8 of 9 in group 1 and 5 of 9 in group 2). We still do not know the disease duration of the patient group they studied, whereas 64% of our patient group had early disease (duration <12 months). The response to TNF blockers in the study by Catrina et al was also modest, with mean DAS28 reductions from 5.9 to 3.8 in group 1 and 6.1 to 4.2 in group 2. This compares with our study where, as a result of DMARD treatment, the mean DAS28 decreased from 5.9 to 2.5, with 13 patients attaining a low disease activity state, defined by a DAS28 <2.6.

The study by Makrygiannakis et al (2) is difficult to compare with our study, because the intervention was intraarticular corticosteroids, the patient population was mixed (6 patients with RA, 2 with spondylarthropathies, and 5 with unclassified oligoarthritis), and the followup was only 2 weeks. We would therefore contend that our study was better able to determine whether the changes in OPG and RANKL with treatment were related to changes in disease activity (as we found) or directly related to a specific treatment such as TNF blockade (as concluded by Catrina et al).

Finally, we would not like the above discussion to detract from the conclusion of all studies that OPG, and to a lesser extent RANKL, are amenable to change with various therapeutic agents and that they are suitable potential therapeutic targets for the treatment of RA in the future.

Malcolm D. Smith MBBS, FRACP, PhD*, * Flinders Medical Centre, Adelaide, South Australia, Australia.