An alternate hypothesis regarding radiologic damage to synovial tissue: Comment on the editorial by Kirwan
Article first published online: 4 NOV 2004
Copyright © 2004 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 50, Issue 11, pages 3735–3736, November 2004
How to Cite
Smith, M. D. (2004), An alternate hypothesis regarding radiologic damage to synovial tissue: Comment on the editorial by Kirwan. Arthritis & Rheumatism, 50: 3735–3736. doi: 10.1002/art.20753
- Issue published online: 4 NOV 2004
- Article first published online: 4 NOV 2004
To the Editor:
I read with interest the editorial by Kirwan (1) in relation to an accompanying article in Arthritis & Rheumatism. Dr. Kirwan uses the results reported by Molenaar et al (2), along with other selected references, to promote the hypothesis that clinical activity and radiologic joint destruction are mediated by different disease processes.
However, I believe that this hypothesis can be challenged by published studies which suggest that these two processes are closely linked, if not identical. Several studies have demonstrated a close association between the acute-phase response and clinical disease activity (3–5), and other studies demonstrate the association of cumulative acute-phase response with the radiologic progression of joint destruction (4–6). Joint inflammation has been associated with the occurrence of joint erosions in rheumatoid arthritis (RA) (5–8). Unfortunately, standard disease-modifying antirheumatic drug (DMARD) treatments have only a modest impact on joint damage in rheumatoid arthritis (RA). This is probably because most treatments do not result in the level of disease suppression that approximates the American College of Rheumatology (ACR) criteria for remission (9) or at least do not maintain it for a long enough period to achieve regression of joint damage (10–12). Our own results demonstrated that synovial membrane pathology could be returned to close to normal, and that this could be associated with retardation of radiologic damage (13, 14). We came to this conclusion upon examining the synovial membrane immunopathology in a selected group of RA patients in whom remission was achieved according to the ACR criteria and was maintained for a significant length of time.
A recent report by Aken et al (15) showed less radiologic progression in patients with early initiation of DMARD treatment than in patients who received delayed DMARD treatment. However, the rate of joint destruction during years 1–4 of the study did not differ between the two treatment groups. Why were the results in the early-treatment group not better in the long term? The answer may lie in the low remission rate for both treatment groups and the significant acute-phase response in both treatment groups up to 4 years, with similar percentages of change in erythrocyte sedimentation rate in both treatment groups over 4 years. In other words, in neither the early nor the delayed treatment group was the disease activity reduced sufficiently, or for a long enough period to favorably affect the rate of radiologic progression.
So what of the study by Molenaar et al (2), which Kirwan suggests supports the concept of different pathologic processes mediating clinical activity and radiologic progression. First, we are not informed in this report of the error rate in repeated scoring of radiographs of the hands and feet using the Sharp/van der Heijde method (16). What is the minimal detectable difference with this scoring method? When the results of this study are examined more closely, it appears that those patients whose RA remained in persistent remission over the 2-year observation period had a significantly better radiologic outcome over 1 and 2 years than did those with exacerbations of their disease activity over that period. It is questionable whether a change in the mean Sharp/van der Heijde score of 0.7 (1-year followup) or 1.1–1.2 (2-year followup) in the RA patients with persistent remission is actually greater than the error rate in the radiologic scoring technique. This study also showed that the area under the curve (AUC) for the Disease Activity Score (DAS) (17) was significantly higher in patients with relevant radiologic progression than in patients with low or no progression. In addition, the DAS AUC was a stronger predictor of radiologic progression than was the absence of persistent remission. Surely this is consistent with the hypothesis that lack of sustained disease activity suppression is related to radiologic suppression, again suggesting that the two pathologic processes are linked, if not identical. If the authors had provided other evidence of how completely disease activity was suppressed over the observation period, possibly by an AUC assessment of C-reactive protein, it would be possible to assess whether any apparent radiologic progression in the sustained-remission RA patient group (a small number of outliers in Figure ) was due to incomplete disease activity suppression.
In conclusion, the published evidence (including that referenced by Kirwan) does not necessarily support his hypothesis of different disease pathologies mediating disease activity and radiologic damage. The study results are equally consistent with the hypothesis that the pathology underlying joint inflammation and radiologic damage are linked if not identical. It may be that this reflects the inadequacies of standard DMARD treatments to completely suppress disease activity for long enough periods to favorably alter radiologic outcomes. Perhaps future studies of treatment with biologic agents will further address this issue, if such treatment aims at the goal of attaining and maintaining remission for a significant period of time, with restoration of synovial membrane pathology to close to that of normal synovium.
Malcolm D. Smith PhD*, * Flinders Medical Center and Repatriation General Hospital, Daw Park, South Australia.