Bacterium Blamed for Antibiotic-Associated Colitis

GRAZ, Austria, Dec. 6 -- The cause of a form of hemorrhagic colitis associated with treatment with penicillin and its derivatives has been pinned down, according to researchers here.

The culprit is Klebsiella oxytoca, a Gram-negative bacterium, reported Christoph Hgenauer, M.D., of the Medical University of Graz, and colleagues, in the Dec. 7 issue of the New England Journal of Medicine.

Colitis is a known complication of antibiotic treatment, but the cause is generally an overgrowth of toxin-producing Clostridium difficile in the colon, said Dr. Hgenauer.

When C. difficile is absent and yet patients develop bloody diarrhea, the condition -- first described in 1978 -- has become known as antibiotic-associated hemorrhagic colitis, the researchers noted.

Until now, no cause was known and in some cases patients were misdiagnosed as having ischemic colitis with spontaneous resolution, an error that "has consequences for the care of such patients, who could be harmed by unnecessary treatment with drugs," the researchers said.

Dr. Hgenauer and colleagues conducted a study of 22 consecutive patients who had suspected antibiotic-associated colitis but were negative for C. difficile. Patients had a diagnostic colonoscopy and those who were diagnosed with antibiotic-associated hemorrhagic colitis had stool samples cultured for K. oxytoca.

Of the 22, six were considered to have antibiotic-associated hemorrhagic colitis and testing showed that five of those had an overgrowth of K. oxytoca, although no other common enteric pathogens were found.

All five had been taking penicillin for such complaints as tonsillitis and sinusitis and two were also taking non-steroidal anti-inflammatory drugs. When the antibiotics were stopped, all five recovered completely.

In contrast, when the researchers tested 385 healthy volunteers, only six (1.6%) were positive for the bacterium.

The supernatant liquid from the cell cultures of the five, Dr. Hgenauer and colleagues said, was toxic in cell cultures, as indicated by cell rounding and cell death, although supernatant from known non-toxic strains of the bacterium had no effect.

Finally, Dr. Hgenauer and colleagues said, they attempted to replicate the disease in animals, following Koch's postulates for linking an organism to a disease laid down in the 19^th century.

Using the strain of K. oxytoca isolated from one of the patients, the researchers tested its ability to colonize rats and produce disease. The animals were divided into six groups:

• 11 received the antibiotic combination amoxicillin-clavulanate (sold under various brand names) as well as the strain of K. oxytoca.
• Eight got only amoxicillin-clavulanate.
• 10 were treated with K. oxytoca only.
• 18 animals were given amoxicillin-clavulanate, the nonsteroidal anti-inflammatory drug Indocin (indomethacin), and K. oxytoca.
• 11 got amoxicillin-clavulanate and Indocin only.
• A dozen animals were given Indocin only.

Only when the animals got antibiotics and bacteria together -- groups one and four -- did the researchers see bacterial colonization of the colon and subsequent colitis. Strains isolated from the animals were identical to the original toxic strain.

Because Klebsiella, which expresses a beta-lactamase, is nearly always resistant to penicillins, antibiotic-associated hemorrhagic colitis is particularly seen during therapy with penicillin derivatives.

The addition of Indocin tended to produce a more severe colitis, Dr. Hgenauer and colleagues said, but the difference was not significant.

"Our study suggests that toxigenic K. oxytoca should be considered in the differential diagnosis of potential intestinal pathogens," the researchers concluded. Stopping antibiotics and NSAIDs and treating the symptoms of the colitis will probably be sufficient therapy.