As newer, more potent, and broader spectrum antibiotics are introduced, we are being faced with the complications of these drugs. One of these, pseudomembranous enterocolitis (PMC), is caused by the toxins produced by the organism Clostridium difficile. C. difficile is a gram positive, anaerobic, spore-forming organism first isolated in 1935.
The first report describing classic pathologic lesions (found at autopsy) was published in 1893 by Finney. Because exposure to antibiotics is the leading (but not sole) cause of this disease, there were very few cases in the pre-antibiotic era. Beginning in the 1950s, the association with antibiotic use was made, leading some to call this antibiotic-associated colitis. From the 1950s until the 1970s, the prevailing thinking, due to stool culture and gram stain results, was that this disease was secondary to Staphylococcus aureus. This led to the name Staphylococcus enterocolitis, which was a common name until the association with clindamycin and the toxins produced by C. difficile was made by three studies, all published in 1974. (Hafiz SL. PhD Dissertation, University of Leeds, Leeds, UK, 1974; Ann Intern Med 1974;81:429; Am J Surg 1974;127:749.)
To develop pseudomembranous enterocolitis, three things must occur: disturbance of the normal colonic flora, colonization by C. difficile, toxin production, and toxin-mediated intestinal injury, and inflammation. First, the disturbance of the normal flora must occur. This is almost always a result of antibiotic exposure. Although clindamycin was the first to be described causing PMC, today the cephalosporins are the leading cause of this disease. Almost any antibiotic can cause the disease, but there are some exceptions: parenteral aminoglycosides, parenteral vancomycin, nitrofurantoin, fosfomycin, antifungal, antiviral, antiparasitics, and antimycobacterials (except rifampin). There are several reports showing that quinolones, especially the extended spectrum, are responsible for an increasing number of cases of PMC. (Emerg Infect Dis 2003;9:730.)
The second event that needs to occur is the patient must be exposed to the organism. Here's where the adage, “People get sicker in the hospital,” is supported. In the U.S., the intestinal carriage rate ranges up to three percent (Gastroenterology 1981;81:5), but after admission to the hospital and treatment with antibiotics, that number climbs to approximately 20 percent. (J Infect Dis 1994;19:181.) This organism has been cultured from the hands of hospital workers, bed rails, floors, bathrooms, and call buttons, hence the call for frequent hand-washing and glove use. Asymptomatic carriers rarely get sick, but they may contaminate their environment.
The third factor that needs to be present is the exposure to the toxins produced by the bacteria. C. difficile produces two exotoxins, called toxin A and toxin B. Toxin A is inflammatory and induces fluid secretion, leading to the enteritis and colitis seen in this disease. Toxin B is a strong cytotoxin but has very little if any of the effects seen with toxin A in animal studies. In clinical practice, however, there are reports of toxin A-negative/toxin B-positive strains causing diarrhea in humans. Risk factors for developing disease are age over 64, severity of other illnesses, recent surgery (colonic, pelvic, gastric), colonic obstruction, Crohn's disease, shock, burns, uremia, HUS, shigellosis, neonatal necrotizing enterocolitis, ischemic colitis, or cardiovascular disease. (Sleisenger & Fordtran's Gastro and Liver Dis, 2002, pp. 1914–1927.)
Asymptomatic or Life-Threatening
Infection with C. difficile may be completely asymptomatic or life-threatening. The incubation period is usually less than seven days, with median time being two days. The signs include diarrhea, which is described as frequent and loose, occasionally with mucous or occult blood (rarely with gross blood). Sometimes there is fever, leukocytosis, and crampy abdominal pain. Very rarely there are extraintestinal manifestations; septic arthritis, bacteremia, or splenic abscesses have been reported, but are uncommon. Interestingly, some very ill patients may not have diarrhea, and these patients will usually have abdominal pain and distention, fever, and elevated WBC.
Stool is positive for white blood cells in approximately 50 percent of cases. (J Clin Invest 1994;93:1257.) The patients, as expected, may be dehydrated with the usual electrolyte abnormalities. PMC is a protein-losing colitis, so one clue for the EP is a low serum albumin. Plain abdominal radiographs may reveal a dilated colon, greater than 7 cm, which is called toxic megacolon. Patients with this also may have a small bowel ileus, and air-fluid levels that mimic an obstruction. The patient with a toxic megacolon is the most ill, and it may perforate, causing death.
Treatment of PMC begins with discontinuing antibiotics, which cures 15% to 25%
Making the diagnosis begins with investigating the patient's medication history, obviously looking for current or recent antibiotic use. Any exposure to antibiotics within the previous three months should raise your suspicion for C. difficile colitis. Enzyme immunoassays to detect toxin antigens have become the most common way to diagnose PMC. This assay is fast (two to six hours), inexpensive, and highly specific. It detects either toxin A or B, but does have frequent false-negative results. A fresh stool should be sent to the lab, and refrigerated if it cannot be analyzed immediately. The gold standard is the tissue culture cytotoxic assay, but this takes 48 to 72 hours for results, is expensive, and is not readily available. Other methods include C. difficile culture, latex agglutination assay, and PCR, but these all have sufficient limitations that make them not useful in making the diagnosis. Endoscopy is reserved for high suspicion with negative laboratory results or when the situation requires a rapid diagnosis.
Treatment begins with discontinuing the offending antibiotic, which cures 15 to 25 percent of patients with PMC. (N Engl J Med 1994;300:257.) If that is not possible, switching to a drug which is less likely to cause the disease is the next option. For more seriously ill patients or those with multiple co-morbidities, antibiotic treatment is indicated. The most common regimens are metronidazole, 250-500 mg PO/IV QID for 10 to 14 days and vancomycin, 125 to 500 mg PO QID for 10 to 14 days. Vancomycin is reserved for metronidazole failures, pregnant women, severe cases, or children under 10.
Metronidazole is first-line due to its lower cost and the possibility of creating vancomycin-resistant bacteria. Antimotility agents should be withheld because they delay toxin clearance, and may precipitate an ileus or toxic megacolon. Unfortunately, with the advent of broader-spectrum antibiotics, we may be seeing more patients with this disease, so the astute EP should have a low suspicion for PMC in any patient with diarrhea who has been on antibiotics recently.