C. difficile colitis therapy is limited but effective

August 2002

Clostridium difficile is a gram-positive anaerobic bacillus that is a normal inhabitant of bowel microbial flora in 5% of adults. It is more commonly part of normal bowel flora in infants, as up to 50% will asymptomatically harbor C. difficile. This rate dramatically decreases after 12 months of life. C. difficile may, under certain circumstances, cause symptomatic disease.

C. difficile is not part of normal bowel flora in most children and adults because its growth is inhibited by other microbial microorganisms, such as Escherichia coli or Bacteroides. When the normal balance of the numerous microbial organisms inhabiting the human bowel is altered, an opportunity for C. difficile to invade the GI tract is created. The use of antibiotics, particularly oral and broad-spectrum, causes such an opportunity. The primary mode of transmission is fecal-oral. Transmission by contaminated fomites also is possible. C. difficile spores may remain viable for up to five months in the environment. C. difficile is present in soil and is ubiquitous in the environment. Stool carriage rates increase significantly in hospitalized patients, as up to 20% of hospitalized patients may asymptomatically harbor C. difficile. Transmission can occur within an institution and result in disease in other patients. Spores have been cultured from hospital flooring, bedding and toilets. Person-to-person transmission has also been documented in day care centers. Risk factors for C. difficile disease include antibiotic use, use of certain non-antibiotic drugs namely several chemotherapeutic agents (methotrexate, doxorubicin, cylcophosphamide or fluorouracil), tube feedings or other manipulations that compromise normal bowel flora.

The spectrum of infection and disease by C. difficile may include asymptomatic carriage or toxic megacolon and life-threatening colitis. Symptoms typically begin one to three weeks after the precipitating event, such as oral antibiotic use, but may occur up to six weeks later. Severe and life-threatening disease is more likely to occur in children with neutropenia, inflammatory bowel disease or in infants with Hirschsprung’s disease.

Infection with C. difficile and the development of disease depend upon several factors, such as the strain of the infecting organism and the presence of risk factors in the patient. Strains of C. difficile may differ in their virulence, as infection with some strains are not as likely to result in symptomatic disease as more highly toxigenic strains. Toxin production and clinical illness are absent in approximately 25% of those with C. difficile culture-positive stools. When infection does result in disease, most will experience mild-to-moderate watery diarrhea. Ten to 15% may progress to bloody diarrhea with a more severe inflammatory process. C. difficile produces two toxins – toxin A, an enterotoxin, and toxin B, a cytotoxin. These toxins produce an inflammatory response within the GI tract leading to fluid secretion, increases in capillary permeability and peristalsis, and potentially tissue necrosis, resulting in bloody diarrhea and possibly perforation and peritonitis. C. difficile toxins do not cause disease by direct tissue invasion, but bind to receptors on the colonic mucosa.

Nearly any antibiotic that disrupts the normal bowel flora may result in C. difficile infection and disease. Several antibiotics, however, are more commonly implicated: amoxicillin, amoxicillin-clavulanate (Augmentin, GlaxoSmithKline), second- and third-generation cephalosporins, and clindamycin. Broad-spectrum antibiotics are most commonly implicated, as they are more likely to upset normal bowel flora. Antifungals have also been implicated, and quite surprisingly, even metronidazole and vancomycin (the agents used to treat C. difficile colitis) have been implicated.

The potential seriousness of C. difficile disease is yet another reason clinicians should limit the use of antibiotics. It is important to consider that diarrhea following the use of an oral antibiotic does not always result from C. difficile, as about 20% of antibiotic-associated diarrhea is not due to C. difficile. Diarrhea following antibiotic use is more likely not a result of C. difficile infection and disease in outpatients, as compared with institutionalized patients treated with an antibiotic. Use of an antibiotic may also result in loose stools or diarrhea by altering normal bowel flora. Because C. difficile is more likely to be transmitted within the institutional setting, diarrhea following antibiotic use in an institutionalized patient is more likely the result of infection with C. difficile. An accurate diagnosis is important for treatment decisions for several reasons: not all C. difficile strains are virulent and produce toxin, C. difficile may normally reside in the bowel of some (without causing disease) and antibiotic use may result in bowel flora alteration and stool formation without C. difficile infection. Culturing for C. difficile in stool has a relatively low predictive value. Diagnosis by stool cytotoxicity assays for toxin presence is preferred.

Metronidazole or vancomycin

Specific pharmacotherapy treatments for C. difficile disease are limited; however, the treatments are effective. About 25% of patients will respond to discontinuing the offending antibiotic or drug only, without instituting specific therapy. When this is inadequate, metronidazole or vancomycin can be used. Metronidazole is well established as the treatment of choice.

The Hospital Infection Control Practices Advisory Committee of the CDC recommends limiting vancomycin use to second-line therapy to limit the development and spread of vancomycin-resistant enterococci (VRE) and vancomycin-resistant Staphylococcus aureus (VRSA). The AAP Red Book similarly recommends limiting vancomycin to patients who do not respond to treatment with metronidazole. Both metronidazole and vancomycin are equally effective in treating C. difficile disease, with efficacy rates shown by controlled clinical trials to be 80% to 100%. Controlled trials have been conducted primarily with adult subjects, although no significant reasons exist to believe that efficacy rates in children would likely be different. Treatment with metronidazole or vancomycin should be by the oral route, as oral administration results in locally high drug levels, and because C. difficile disease affects local physiologic functions. Metronidazole is not commercially available in a liquid dosage form, but recipes for a suspension are easily found. Vancomycin is commercially available in a liquid dosage form. When oral administration is difficult or not feasible, some have suggested the use of IV metronidazole, but only when used concomitantly with vancomycin given enterally (eg, catheter perfusion). Although metronidazole may be given intravenously, oral therapy is preferred, as evidence for its efficacy is greater with this route of administration. There is no role for the use of IV vancomycin. Vancomycin is not systemically absorbed to an appreciable extent when given orally; thus there is no role for monitoring blood concentration levels.

Some strains of C. difficile have been shown to be resistant to metronidazole, although this is uncommon (<5%). Vancomycin should be reserved for patients who do not respond to metronidazole (although this is unlikely), are allergic or intolerant to metronidazole (may cause an unpleasant metallic aftertaste and will interact with alcohol-containing medications) or are pregnant (first trimester). The use of metronidazole in pregnancy is controversial, for it is labeled a pregnancy category “B” drug and has been shown by several studies to be safe when used in second-third trimester pregnancy. Metronidazole is considered by the CDC to be contraindicated in first trimester pregnancy when used in the treatment of trichomoniasis. Improvement with metronidazole or vancomycin therapy should be seen by 48 hours; therapeutic failure should not be assessed until day 6 of therapy. Ten days of treatment is optimal.

Antimotility agents have no role in the treatment of C. difficile disease, as their use may promote toxin retention and ileus formation. Because treatment with metronidazole or vancomycin is so effective, alternative treatment decisions should be uncommon. However, options include bacitracin or cholestyramine (to bind toxin). Efficacy for neither of these therapies is well documented.

Relapses of C. difficile infection and disease are common. Relapse rates up to 33% within one to six weeks have been reported. Most relapses are reinfection and not failure of previous therapy. Numerous relapses may occur with some children, and up to 20 relapse episodes have been reported to occur. Relapses are generally no more severe than previous infections. Some evidence suggests that relapses are less likely if previous episodes were treated with supportive therapy only. Relapses may be due to the survival of spores over several months, with eventual germination and disease production. Metronidazole or vancomycin do not effectively reduce spore survival. Following therapy for C. difficile disease, it is beneficial to limit the use of oral antibiotics in general, in an attempt to hasten the development of normal bowel flora, which is likely to reduce relapses.

When relapses do occur, they should be treated as initial infections, with metronidazole the preferred therapy. No therapy has been shown superior by controlled trial, and relapses generally respond well to treatment. The difficulty in some patients may be preventing continuing relapses. Because of this, other therapies have been suggested, and some have been evaluated by controlled trials. The most studied are probiotics — nonpathogenic microbial organisms that restore normal bowel flora and balance. The results of controlled trials with probiotics, namely Lactobacillus GG (Culturelle, ConAgra) or Saccharomyces boulardi (not commercially available in the United States), have been published documenting the efficacy of these therapies. Most evidence has been shown with adult subjects, although limited evidence in children is also promising. A randomized, placebo-controlled trial in adult subjects compared Saccharomyces boulardii with metronidazole or vancomycin to antibiotic therapy alone. Rates of recurrence were reduced by nearly 50% with Saccharomyces boulardii therapy.

Not all probiotics may have similar efficacy, as success with their use depends upon the organism’s ability to survive in and adhere to the human bowel, as well as their ability to inhibit the growth of C. difficile. The efficacy of commonly available nutritional yogurt culture products is questionable.

Other therapies for the prevention of relapses include long-term metronidazole or vancomycin (eg, one to two months given intermittently) with or without cholestyramine, vancomycin plus rifampin, or IV immunoglobulin (in children with documented IgG anti-toxin A deficiency). The use of enemas containing stool from healthy subjects has also been suggested, although the practical considerations of this therapy are obvious. It is hard to overstate the importance of important infection control practices, namely hand washing, in preventing C. difficile infection, disease and relapse.

For more information:

• Fekety R. Guidelines for the diagnosis and management of Clostridium difficile-associated diarrhea and colitis. Amer J Gastroenterol. 1997;92:739-50.
• Teasley DG. Prospective randomized trial of metronidazole versus vancomycin for Clostridium-difficile-associated diarrhea and colitis. Lancet. 1983:1043-6.
• Hospital Infection Control Advisory Committee. Recommendations for preventing the spread of vancomycin-resistant enterococci. Amer J Infect Control. 1995;23:87-94.
• McFarland LV. A randomized placebo-controlled trial of Saccharomyces boulardii in combination with standard antibiotics for Clostridium difficile disease. JAMA. 1994;271:1913-8.
• ASHP therapeutic position statement on the preferential use of metronidazole for the treatment of Clostridium difficile-associated disease. Amer J Health-System Pharmacy. 1998;55:1407-11.
• Biller JA. Treatment of recurrent Clostridium difficile colitis with Lactobacillus GG. J Pediatr Gastroenterol Nutrition. 1995;21:224-6.
• Pruksananonda P. Multiple relapses of Clostridium difficile-associated diarrhea responding to an extended course of cholestyramine. Pediatr Infect Dis J. 1989;8:175-8.
• Pochapin M. The effect of probiotics on Clostridium difficile diarrhea. Amer J Gastroenterol. 2000;95 (Suppl 1):S11-S13.
• Leung DY. Treatment with intravenously administered gamma globulin of chronic relapsing colitis induced by Clostridium difficile toxin. J Pediatr. 1991;118:633-7.