Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
We read with interest the article entitled “Three month treatment of reactive arthritis with azithromycin: a EULAR double blind, placebo controlled study”.1 The trial of Kvien et al suggests that weekly administration of azithromycin for 3 months is not efficacious in ameliorating the symptoms of reactive arthritis (ReA). Although this point seems clear, the authors then make a leap of faith and suggest that “this study does not support the prolonged use of antibiotics for the alleviation of ReA”. There are several problems with this generalisation.
As Kvien et al correctly point out, polymerase chain reaction technology has documented the presence of Chlamydia and other causative organisms in the synovial tissue of patients with ReA.2 This same technology has convincingly shown both in vitro and in vivo evidence of persistent metabolically active Chlamydia.2 The data on post-dysentery organisms have repeatedly demonstrated bacterial fragments,3 but viability has only been suggested in the case of Yersinia.4 This makes a strong argument for the use of antimicrobial agents in post-chlamydial ReA, yet both patients with post-venereal and post-dysentery ReA were included in this trial.
Previous therapeutic trials also suggest that post-chlamydial ReA is more susceptible to antimicrobial treatment than the post-dysentery form. A 1991 trial suggested that lymecycline was an effective treatment for post-chlamydial ReA, but not for the post-dysentery form.5 A subgroup analysis of post-chlamydial patients in another trial assessing ciprofloxacin showed a trend towards improvement.6 There were not enough post-chlamydial patients in the trial of Kvien et al for a meaningful analysis to be made.
We also question the treatment itself in their trial. A one-time dose of 1000 mg of azithromycin is approved for an acute Chlamydia infection; however, the proper dose for persistent infection has not been established. To our knowledge, 1000 mg weekly has never even been studied in vitro as a dose to treat persistent Chlamydia. In addition, persistent Chlamydia infections intermittently shed infectious elementary bodies, potentially evading weekly pulse antimicrobial treatment. It has also been demonstrated that the chronic treatment of Chlamydia trachomatis with azithromycin in vitro caused the Chlamydia temporarily to arrest in a persistent viable state.7 Lastly, it has not been established if 3 months of a single antimicrobial agent is successful at treating an obligate intracellular organism that exists in the form of a reticulate body. Other obligate intracellular organisms, such as Mycobacterium tuberculosis, require 9 months of combination antimicrobial treatment to ensure therapeutic response.
Kvien et al implied that their trial, along with previous trials, indicates a lack of efficacy of antibiotics in ReA. The antibiotics studied previously included tetracyclines, ciprofloxacin, and now azithromycin.1,5,6Chlamydia has demonstrated in vitro resistance to all of these antibiotics upon chronic administration.7,8 Further, ciprofloxacin has been shown to cause tendon based inflammation by potentiating interleukin 1β stimulated metalloproteinase-3 output in tendons.9 Is this then the proper antibiotic to choose in the treatment of an enthesophyte based inflammatory arthritis?
We have recently completed a trial assessing a 9 month course of a combination of doxycycline and rifampin versus doxycycline monotherapy.10 The results showed a rather dramatic response in the patients who received the combination. The chlamydial resistance that has been documented in vitro, was overcome when a combination of antibiotics were used.7 Ours was the first trial to assess a combination of antibiotics in this setting.
Do antibiotics work in ReA, specifically Chlamydia induced ReA? In our opinion, this question has not been answered. We believe studies of large groups of patients, with the appropriate antibiotics, in the right dose, used for the proper length of time, need to be conducted before this question can be answered.
We thank Carter et al1 for their valuable comments on our paper which reported the results of 3 months’ treatment of reactive arthritis (ReA) with azithromycin.2 The data from our study definitely did not support prolonged use of antibiotics for the alleviation of ReA, because no trend was found in favour of long term treatment. However, we do not disagree that the data from the study by Carter et al,1 and from other authors,3 may support long term treatment with antibiotics in patients with ReA induced by Chlamydiatrachomatis.
Such positive findings as have been reported seem to be restricted to this microbiological agent. We note that the study by Carter et al1 was performed in patients with chronic undifferentiated spondyloarthropathy without confirmed Chlamydia infection, but 9 of 30 patients had either a possible or probable preceding symptomatic Chlamydia infection.
We also agree that various arguments can be employed in the selection of the optimal antimicrobial agent in ReA. We chose azithromycin in our study because of its acceptable tolerability profile combined with a broad antimicrobial spectrum, as our study was designed to focus on all patients in whom ReA was a likely diagnosis—not just patients with Chlamydia induced ReA. Carter et al compared doxycycline 100 mg twice a day with doxycycline 100 mg twice a day + rifampicin 600 mg a day.1 The latter drug is most widely used for the treatment of tuberculosis. The safety of this combination should be clarified before recommendations are given for its wider use in ReA or undifferentiated spondyloarthritis.
We would also welcome an adequately powered trial confined to patients with Chlamydia induced arthritis, to clarify the efficacy or otherwise of long term treatment with antibiotics in this condition. However, in our opinion, such a trial will be difficult to perform, because of the logistic problems of recruiting large numbers of bacteriologically proven cases early in the course of their disease. For the present, therefore, clinicians must base their treatment on currently available data.