Antibiotics treatment of reactive arthritis: a systematic review of the literature

Background: Reactive arthritis is a common disease, with a similar incidence to rheumatoid arthritis (RA). Around 15% of patients develop chronic arthritis (symptom duration more than 6 months). There is no clear consensus among rheumatologists on whether to treat ReA with antibiotics. We have systematically reviewed the evidence for antibiotic treatment in reactive arthritis (ReA). The last clinical review of antibiotic treatment in ReA was published in 2006. Methods: A search of PubMed 1950‐October 2008 was performed using the keywords “antibiotic reactive arthritis”. Further searches substituted “antibiotic” with 7 generic antibiotic names. These searches were repeated using the Embase and Cochrane databases, National Library of Health resources, Ovid [All Evidence Based Medicine reviews: Cochrane DSR, ACP Journal Club, DARE, and Cochrane Central register of controlled trials (CCTR)]. Guidelines from national health organisations in the UK and USA were searched. Clinical trials of antibiotic therapy in reactive arthritis were identified and scored using the JADAD score for methodological quality. Studies were excluded if the patient group had no definite prior history of infection. Primary outcomes included clinical examination findings, patient reported symptoms and laboratory indices of inflammation. Results: The search yielded 11 randomised controlled trials (RCTs) and 1 prospective controlled trial. Eight RCTs were judged to be of good methodological quality (JADAD score ≥3). Three studies were excluded because the patient group had no prior history of infection. The duration of ReA ranged from 3 months to 6 years. Two studies of short‐term antibiotic therapy in acute ReA were of poor methodological quality (JADAD score=1). No difference in outcome measures was found between groups. There were two studies of long‐term antibiotic therapy in acute ReA: one study found significantly reduced CRP in the group treated with lymecycline, and one follow‐up study found that complete recovery was significantly greater in the group treated with ciprofloxacin. There were 5 studies looking at longterm antibiotic therapy in chronic ReA. Time to clinical remission, global VAS score and VAS pain were significantly reduced in one study population treated with azithromycin with synovectomy compared with synovectomy alone. The other studies found no significant improvement in Ritchie articular index, remission at 3,6 and 12 months, ACR remission score, swollen and tender joint counts, VAS pain, patient VAS global health assessment, VAS morning stiffness, physician assessment of treatment success, symptom score, CRP/ESR, or serology. Conclusions: No new evidence exists for short‐term antibiotics in acute ReA. There is conflicting evidence on the effects of long‐term antibiotics in acute ReA on prognosis. Although there is a lack of evidence, some authors advocate the use of prolonged antibiotics in C.trachomatis‐triggered ReA.