60 mm/hour. Sulindac was stopped, and she was given meclofenamate sodium monohydrate, 100 mg 4 times daily.
On November 25, 1987, she had slight swelling and pain of her right ankle, but all other joints were normal. She had no gastrointestinal symptoms, and the ESR was 10 mm/hour. She was reevaluated on January 13, 1988, at which time she was totally asymptomatic, and joint examination results were normal. The ESR was 4 mm/hour. She has since not taken the meclofenamate for more than 1 year and has had no recurrence of symptoms.
The characteristics of reactive arthritis after enteric infection include the onset of arthritis predominantly during the second or third week after the onset of diarrhea. The arthritis is asymmetric, additive or migratory, predominates in the lower extremities, and may involve the lower back or sacroiliac joints; fever is present in most cases (8). Laboratory findings include an elevated ESR and inflammatory synovial fluid with negative results of cultures (8). Our patient's symptoms met all of the above features. In addition, the patient's stools were loaded with B hominis. After treatment with metronidazole, the diarrhea resolved, stool test results were negative for B horninis, and the arthritis subsequently resolved on treatment with nonsteroidal antiinflammatory agents. The ESR became normal. Despite stopping treatment, symptoms have not recurred for over 1 year.
Reactive arthritis has been reported in association with a number of enteric bacterial pathogens including Shigella pexneri, Salmonella species, Yersinia enrerocolitica, Campylobacter jejuni, and Brucella species (8). These reactive enteric-arthritic syndromes demonstrate an HLA-B27 negative linked predisposition, except for that caused by Brucella species (8). Our patient was negative for HLA-B27.
B hominis, an intestinal protozoan, has been increasingly implicated as a pathogen in diarrheal disease. We are not aware of any report of reactive arthritis after enteric infection with B hominis and believe this to be the first such report. The relative numbers of the organism are associated with the severity of symptoms (9). When 2-5 organisms are present per preparation, the result is reported as rare; 1 organism per 5-10 microscopic fields is reported as few; 1 organism per 2-3 microscopic fields to 1-2 organisms per microscopic field are reported as moderate, and "several" organisms in every microscopic field are reported as many (9). The specimen from our patient was described as being "loaded" with B hominis (per microscopic field), which was more than "several." Whether the numbers of the organism represent a predisposing factor in reactive arthritis is not known.
Our patient had reactive inflammatory polyarthritis after enteric infection with B hominis. The disease resolved on treatment with metronidazole and nonsteroidal antiinflammatory drugs. An absolute cause and effect cannot be proven because other enteric pathogens are not present in the stool when the reactive arthritis appears, and because B hominis is not a usual major pathogen. Nonetheless, B hominis should be considered as a possible pathogenic agent in patients with enteritis and arthritis. The diagnosis is easily made by stool examination, and the treatment may be curative.