Article Text

Download PDFPDF

Ciprofloxacin v placebo for treatment of Yersinia enterocoliticatriggered reactive arthritis


Patients with yersinia triggered reactive arthritis were double blind randomly allocated to receive treatment with ciprofloxacin 500 mg twice daily orally or placebo during three months. The diagnosis was made by serology (specific IgA and IgG antibodies to yersinia outer membrane proteins (yops)), positive culture, and/or demonstration of Yersinia enterocolitica antigen in colon biopsy specimens. Patients were evaluated monthly during and after treatment up to 12 months. Of 18 patients enrolled, all could be evaluated for safety, 16 for efficacy. There was a tendency towards faster remission and relief of pain in those receiving ciprofloxacin. Y enterocolitica was eliminated from the gut associated lymphoid tissue in six of seven patients receiving ciprofloxacin compared with none of nine patients receiving placebo. Patients receiving placebo had more and prolonged circulating IgA antibodies against yops than patients treated with ciprofloxacin.

  • yersinia
  • arthritis
  • IgA
  • ciprofloxacin

Statistics from

Request Permissions

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.

Yersinia triggered arthritis is a complication of gastrointestinal infection by Yersinia enterocolitica; its incidence in endemic areas is calculated to be over 20%.1 Patients with yersinia triggered arthritis show persistent specific serum IgA antibodies against yersinia outer membrane proteins (yops)1-3 in contrast with patients with uncomplicated yersinia infection, which is the result of chronic stimulation of the gut associated lymphoid tissue by persistent virulent bacilli.1 Previous studies suggested that antibiotics could eliminate persistent bacilli with consequent disappearance of circulating IgA.2 These studies are not conclusive about the effect on the arthritis itself. We conducted a randomised, double blind, placebo controlled study on the clinical and microbiological efficacy of ciprofloxacin treatment in patients with proven yersinia triggered arthritis. Ciprofloxacin was chosen because of its high in vitro activity against Y enterocolitica.4

Patients and methods


Adults over 18 years of age with proven yersinia triggered arthritis for a duration of fewer than five years were included in the study after informed consent was given. Patients with rheumatic diseases, arthritis associated with other bacterial infections, rheumatic fever, psoriasis, Crohn's disease, ulcerative colitis, or lupus erythematodes were excluded, as were patients receiving antibiotics with a spectrum of activity similar to ciprofloxacin, or receiving corticosteroids, sulfasalazine, antacids, and theophylline derivatives.

Patients were randomly allocated to receive either ciprofloxacin 500 mg twice a day or placebo, orally during three months. Responses were assessed at 1, 2 (during), 3 (end), 4, 6, 8, 10, and 12 months after the start of treatment. Clinical response was assessed by the articular index score,5 the visual analogue pain scale (VAS), and the patient's impression. The articular index scored each joint for tenderness to pressure (scale 0–3). Additionally, joint swelling (scale 0–3) and pain at rest (scale 0–1) were scored. The VAS assessed pain intensity on a scale 0–10, where 0 = no pain and 10 = most intense pain. Each patient had a complete physical examination, radiographs of the thorax and lumbar vertebrae, and blood screening tests (haematology, HLA-B27, chemistry, C reactive protein (CRP)).


Yersinia infection was proved by demonstration of specific serum IgA and IgG antibodies against yops, positive culture of faeces or biopsy specimen, and/or demonstration ofY enterocoliticaantigens in intestinal biopsy specimens. Faeces, biopsy specimens (by sigmoidoscopy), and serum were taken at the start, at regular intervals, and after treatment. Faeces and biopsy specimens were cultured for Y enterocolitica according to standard methods.6 Biopsy specimens were also examined for histology and for the presence of yersinia bacilli by an immunofluorescence technique with monoclonal and monospecific antibodies to yersinia lipopolysaccharide, yadA and yopH6. Serum class-specific antibodies against Y enterocolitica were detected by immunoblotting.7For all subjects the following tests were negative: antistreptolysin, Rose-Waaler test, latex agglutination test, tests for antinuclear and anti-DNA antibodies, and tests for antibodies toSalmonella typhi, Salmonella paratyphi A and B,Campylobacter jejuni,Brucella abortus,Neisseria gonorrhoeae, andChlamydia trachomatis.



Eighteen patients were included in the study, of whom seven were randomly allocated to receive ciprofloxacin and 11 placebo. Two patients receiving placebo were excluded because of protocol violations. At the start of the study the two groups were comparable for duration of the disease, presence of the HLA-B27 antigen, articular index, VAS, ranges of sedimentation rate, and CRP (table 1 and fig 1). None of them had been treated with antibiotics for yersinia infection before the study. At the end point six of seven patients receiving ciprofloxacin were cured, with complete remission of the arthritis, compared with five of nine receiving placebo. Remission was accompanied by a decrease in the sedimentation rate and CRP in all patients. The articular index of patients receiving ciprofloxacin fell earlier than that of patients receiving placebo and remained lower at follow up. Ciprofloxacin had a positive effect on the disappearance of pain during active movement, measured by the VAS, which was not observed with placebo (fig 1). Additionally, a more rapid decline of the number of swollen joints and earlier relief of pain at rest were seen in patients receiving ciprofloxacin than in those receiving placebo.

Table 1

Patients' data, outcome of treatment

Figure 1

Improvement in the visual analogue pain scale (VAS) and joint tenderness in the patients receiving ciprofloxacin or placebo.

Ciprofloxacin was well tolerated in all patients and there were no significant side effects.



None of the patients had a positive culture at the start of the study—one receiving placebo became culture positive (faeces) at month 2, another receiving placebo at month 4.

Antigen detection

At the start Y enterocolitica antigens were found in all patients (fig 2).Yersinia bacilli were most frequently located deeply in the tunica propria, just below the basal membrane. At the end of treatment Y enterocolitica was found in three of seven patients receiving ciprofloxacin, two of them became negative at month 6, the other remained positive. After decoding it appeared that he was the patient with partial clinical cure. All nine patients receiving placebo appeared positive at the end of treatment, six patients were still positive at month 6, five at month 12. After decoding it appeared that they were two clinical failures, one partial clinical cure, and two cures. Persistence of antigen was linked with persistence of specific IgA in both groups.

Figure 2

Number of patients with yersinia antigens in biopsy specimens (antigen) and specific serum IgA and IgG antibodies against yersinia outer membrane proteins (yops) (total and yops E, D, N, H, and M) at the start, shortly after treatment (month 4), and during follow up (months 6 and 12).


The biopsy specimens from the colon taken at the start of the study showed a slightly inflammatory reaction, with lymphocytic infiltrates in six patients and no inflammation in the other 10 patients.


All patients had specific IgA and IgG antibodies at the start (fig2). IgA antibodies were directed to two or more yops, mainly yopD, yopN, and yopH. Immediately after ciprofloxacin treatment, specific IgA could not be demonstrated in five patients, one patient still had reactive bands to yopD, yopN, and yopH, one to yopD and yopN. During follow up four patients showed reactive bands to yopD, one of them also to yopN, this pattern changed slightly during the following six months. IgA antibodies persisted in eight patients receiving placebo, with two to five reactive bands to the various yops, including to yopN and yopM, which were not found after ciprofloxacin treatment. The IgG antibodies, numbers and patterns, showed a similar course.


The overall clinical outcome at the 12 months' end point was similar in both groups; in this way our study confirmed previous studies with comparable designs.8 ,9 Complicating factors for the evaluation of these studies are the diversity in patients, the experience that many patients with reactive arthritis may be cured spontaneously, and the knowledge that patients with long duration of the disease might have developed irreversible arthritic changes. One might easily conclude that antibiotics are of no value in the treatment of reactive arthritis. Looking in more detail at the results, there was clearly earlier relief of pain and faster remission of arthritis in those receiving ciprofloxacin. This paralleled also elimination ofY enterocolitica from the biopsy specimens. To our knowledge this is the first time that disappearance of the triggering organism could be associated with a favourable clinical outcome. By the second month disappearance of the organism was already clearly associated with the disappearance and decrease of IgA antibodies, whereas persistence of the organism in untreated and treated patients was associated with prolonged circulation of specific IgA antibodies. It was already known from earlier studies1 ,2 ,10 that patients with yersinia triggered arthritis have more antibodies, which remain for longer, than patients with uncomplicated yersinia infections.

The course of the antibody patterns in patients treated with ciprofloxacin resembles that of patients with uncomplicated yersinia infection, in whom only antibodies (mainly IgG) to yopD can be found after cure. These antibodies are demonstrable for many years.7 Patients treated with ciprofloxacin scarcely developed antibodies to yopH and did not show antibodies to yopM. These antibodies are produced at a late stage of the disease and reflect the more chronic phase. Apparently, ciprofloxacin interferes with antibody production by eliminating the trigger and thereby prevents development of a more chronic stage. Whether this was the primary reason for a better outcome is speculation, but the suggestion is attractive.

The behaviour of antibody patterns in the patients with untreated arthritis resembles that of experimental yersinia arthritis in rats. Prolonged circulation of IgA antibodies to all yops can be found in Lewis rats with persistent infections and arthritis,11 ,12fading of the immunoblot pattern against these yops was seen in brown Norway rats which did not develop arthritis.12 In experimental yersinia triggered arthritis in Lewis rats it was noted that the earlier ciprofloxacin treatment was started, the better were the results.13 This might also be the case in humans and might explain the positive effect in our patients with a disease duration of 1.9 years, which is much shorter than that (4.9 years) in the Finnish study,8 in which no effect was seen.

We realise that the β error of this trial was large and the power low because of the small number of patients. It was difficult to include sufficient patients, because rheumatologists who were already convinced that antibiotics are of benefit in yersinia triggered arthritis were not willing to deliver patients for a double blind study and rheumatologists who thought that antibiotics are of no value did not recruit patients. Most investigators in this field are faced with these problems,9 which make evaluation of studies difficult and conclusions hazardous. We think that faster remission of objective and subjective symptoms by antibiotic treatment is of benefit for every patient with yersinia triggered arthritis, though the ultimate outcome may not be influenced.