Community-Acquired Clostridium difficile Colitis

Emerg Med 39(7):37, 2007

By Sarah Perloff, DO, and David Horn, MD, FACP

The authors present a case of C. difficile-associated diarrhea that illustrates its recent emergence in a new patient population.

Clostridium difficile has long been associated with pseudomembranous colitis. Described traditionally as a nosocomial infection, it has been a disease entity seen primarily in elderly or debilitated patients with recent exposure to antibiotics and recent or current hospitalization. However, there is now an increasing incidence of C. difficile-associated diarrhea emerging in a previously unlikely patient population—young people living in a community setting with few or no comorbidities. We present here a case of community-acquired C. difficile-associated diarrhea and review the predisposing factors for the disease and the treatment options available.

Patient Presentation

A 24-year-old single man who works as a packaging engineer and volunteer EMT and has no significant medical history presents to the emergency department complaining of a four-week history of brown, watery, nonbloody diarrhea, nausea, abdominal cramping, anorexia, headache, fatigue, and fevers up to 101°F. These symptoms and signs began 12 days into a 14-day trip to Beijing and Yunong in southern China in September 2005. Prior to his trip he had received vaccines for hepatitis A and B, tetanus, and typhoid. Throughout the trip he ate local foods but drank only bottled or boiled water.

His activities abroad included a camping trip in the jungle, a visit to friends who lived in a small village with poor sanitation and plumbing, and a tour of the city of Beijing. Neither of his friends (one a resident of China, the other an American) became ill.

The patient returned home two days into his illness, noting 10 to 12 episodes of diarrhea daily. Over the subsequent four weeks, his nausea and fever resolved, but the diarrhea, abdominal cramping, and anorexia persisted. The frequency of the diarrhea decreased to four or five episodes daily but continued to limit his daily activities. On presentation, the patient reports having lost 10 pounds.

The review of systems is negative for fever, chills, sweats, hematochezia, melena, ill contacts, weakness, rash, lightheadedness, paresthesias, cough, chest pain, shortness of breath, dysuria, or hematuria. His last antibiotic exposure was more than 10 years ago.

The patient is not taking any prescribed or over-the-counter medications. As a child, he had what he describes as an allergic reaction to sulfonamide drugs. He has no known family history of gastrointestinal disorders. He is a heterosexual who lives with two friends in an urban home without pets. Although not currently sexually active, he reports 100% use of condoms with previous sexual contacts. He says he does not smoke, drink alcohol, or use illicit substances.

Since becoming a volunteer EMT in 2001, he has no history of needlesticks. His last routine purified protein derivative test in 2004 was negative. Previous travel included backpacking through Europe, also in 2004, and trips to London in 1998 and 2001, during and after which he remained healthy.

On presentation, his vital signs are: heart rate, 61; blood pressure, 110/70; respiratory rate, 12; temperature, 98.3ºF. He does not have orthostatic hypotension.

Physical examination reveals a well-appearing young man with loose-fitting pants, white sclera, pink conjunctiva, a normal oropharynx, and good dentition. The cardiac exam is normal, and his lungs are clear on auscultation. His abdomen is soft, nontender, and nondistended, with good bowel signs and no bruits, masses, or hepatosplenomegaly. There is no lymphadenopathy, rash, or skin lesions.

After a stool sample is sent for culture and testing for ova and parasites, the patient is started empirically on nitazoxanide 500 mg orally twice daily for three days for suspected giardiasis or cryptosporidiosis. No improvement results.

He returns for follow-up three weeks later with increasingly frequent diarrhea despite completing his antibiotic course. Initial complete blood count, electrolyte levels, blood urea nitrogen/creatinine, and liver function tests are all normal. His stool culture grew normal flora and was negative for Salmonella, Shigella, Campylobacter, Giardia lamblia antigen, Entamoeba histolytica antigen, Cryptosporidium antigen, Vibrio, and Yersinia. A repeat stool is sent for culture and testing for ova or parasites, Microsporidium, Isospora, Cyclospora, and C. difficile. Ultimately, the stool is positive for C. difficile antigen and toxin B (toxin A was negative), so the patient is treated with a 14-day course of metronidazole, which results in resolution of his symptoms.

DISCUSSION

Clostridium difficile is a spore-forming, anaerobic, gram-positive rod. The spores are widespread in the environment, especially in places like hospitals, nurseries, and bathrooms, with acquisition of infection believed to be due to fecal-oral transmission. Pathogenic strains of C. difficile produce exotoxins called toxin A and toxin B, which bind to receptors on intestinal epithelial cells, triggering production of prostaglandins, leukotrienes, histamine, and interleukins (IL)-1 and -8. This cytokine release produces an acute inflammatory infiltrate of neutrophils and monocytes within the intestinal mucosa, resulting in necrosis of the colonic brush border with sloughing of these cells and subsequent massive fluid secretion into the lumen of the colon.

Although only toxin A has been demonstrated to cause an inflammatory diarrhea in animal models, both toxins have been shown to damage the colonic epithelium in humans by inactivating the intracellular Rho proteins that regulate actin filaments, ultimately destroying the cell’s cytoskeleton. The resulting clinical picture of acute watery diarrhea with lower abdominal pain, fever, and leukocytosis is what we know classically as antibiotic-associated or pseudomembranous colitis.

Traditionally, this disease is seen in patients who have recently received antibiotics, particularly elderly, hospitalized patients, who often have concomitant exposure to proton pump inhibitors. Specific antibiotics that have been implicated in this population as major factors in promoting C. difficile-associated disease include clindamycin in the 1970s, cephalosporins in the 1980s, and fluoroquinolones in the 1990s.

RISING INCIDENCE

Since the turn of the century, there has been a rising incidence of severe C. difficile-associated disease. Investigations of hospital outbreaks in Canada, Great Britain, the Netherlands, and the United States have revealed the emergence of a previously uncommon, highly toxigenic strain of C. difficile. This strain has been demonstrated by restriction fragment-length polymorphism analysis and pulsed field gel electrophoresis studies to have two distinct findings that contribute to its pathogenicity.

First, it produces a third toxin type called a binary toxin, which is encoded at a separate locus on the genome, distinct from the genes encoding toxins A and B. This toxin appears to be associated with more severe disease, including toxic megacolon, leukemoid reactions, severe sepsis, perforation, need for colectomy, and death. However, this association has yet to be proven in large prospective trials.

Second, there is an 18-base pair deletion within the pathogenicity locus in the tcdC gene, which is a putative negative regulator of toxin A and B production. The result is unchecked production of the two toxins.

This emerging strain of C. difficile has been shown to produce disease not only in the hospital setting in traditional at-risk patients, but also in patients previously considered not at risk for the disease. There have been a number of case reports of severe C. difficile-associated diarrhea in otherwise healthy patients with minimal or no exposure to either the health care setting or antibiotics in Pennsylvania, New Jersey, Ohio, and New Hampshire. As reported in Morbidity and Mortality Weekly Report, epidemiologic studies from these states have shown that although the incidence rates for both patients from the community and peripartum women were approximately equivalent to the national average, the strict definitions employed to exclude hospital-acquired disease and the nature of the study (voluntary reporting) may have yielded an inaccurately low incidence.

NEW RISK FACTOR

In addition, a new risk factor for hospital-acquired disease is emerging—gastric acid suppression (see table below). A recent case-control study on the influence of gastric acid suppression in adults on the rates of C. difficile infection was examined. The data were extracted from the United Kingdom General Practice Research Database spanning a 10-year period, with each case having 10 matched controls. Cases were defined as being community-acquired if the patient had not been hospitalized in the year prior to diagnosis. After controlling for comorbidities, age, gastrointestinal diseases, and sex, conditional logistic regression analysis revealed that the use of proton pump inhibitors conferred a relative risk of 2.9, followed by a relative risk of 2 with the use of H₂-receptor antagonists versus matched controls.

The purported role of gastric acid suppression is that of impaired host defense against ingestion of
C. difficile spores, which have been shown in animal models to transform into the vegetative state within one hour of ingestion, largely within the small intestine. There currently are no data available regarding the effect discontinuing gastric acid suppression may have on clinical cure.

DIAGNOSTIC  TESTING

Currently, although there are a number of diagnostic modalities for determining the presence of C. difficile infection, there is no method for determining which strain of C. difficile (binary toxin-producing versus a historic strain) is causing the disease. The gold standard diagnostic test is the cytotoxicity assay that detects cytopathic effects on a cell culture medium inoculated with stool filtrate. As a result of disruption of the cytoskeleton, cells in the medium develop a rounded appearance when exposed to one of the toxins. Although the sensitivity and specificity of this test are very high (sensitivity, 92% to 100%; specificity, 99%), it is an uncommonly employed diagnostic modality because of its high cost and the time it takes (two to three days) to complete the test.

Instead, the diagnosis is more commonly made by one of a variety of enzyme immunoassays (EIAs), which detect production of toxin A, toxin A and B, and/or the common C. difficile antigen (glutamate dehydrogenase). The sensitivity of EIA testing increases with assays looking for both toxin A and B. Sensitivity for toxin A alone is 65% to 88%, with a specificity of 65% to 100%, but when a combined EIA for toxins A and B is used, sensitivity increases to 66% to 96%, with a specificity of 93% to 100%. The presence of C. difficile antigen without toxin production may indicate a carrier state rather than disease. There is no commercially available test at this time to detect the binary toxin of the emerging epidemic strain of C. difficile.

Endoscopy may detect the presence of pseudomembranes in a subset of patients with C. difficile-associated diarrhea. However, it is not required for the diagnosis because of its low sensitivity (only about 50% of patients with C. difficile-associated diarrhea have pseudomembranes). Other findings on endoscopy, though none is pathognomonic, include erythema, edema, friability, and nonspecific colitis with small ulcerations.

TREATMENT OPTIONS

Once C. difficile-associated diarrhea has been diagnosed, treatment should begin promptly, starting with cessation of ongoing antibiotics as early as possible. Local treatment of the infected colonic mucosa should be initiated using either oral metronidazole or oral vancomycin. Although only oral vancomycin is approved by the Food and Drug Administration for use in C. difficile-associated diarrhea, oral metronidazole is also a proven effective treatment, with an 89% clinical cure rate by the end of a 10-day course. In clinical practice, oral metronidazole is commonly dosed at 500 mg three times daily. Intravenous metronidazole has been demonstrated to achieve bactericidal fecal concentrations in patients with active C. difficile-associated diarrhea, but its use has not been proved or disproved as effective therapy in vivo.

Oral vancomycin 125 mg orally four times daily is highly efficacious, with a reported clinical cure rate
of 90%. However, it is significantly more expensive than oral metronidazole and carries a similar relapse rate (12% vs 5%).

The Infectious Diseases Society of America (ISDA) is in the process of updating its guidelines for the treatment of C. difficile colitis, so some recommendations may change in the future.

Newer therapies continue to be evaluated in clinical trials as future treatment options. In a recent randomized, prospective, double-blind study, nitazoxanide—a nitrothiazolide that had previously been shown in low concentrations to inhibit the growth of C. difficile on culture—was found to be noninferior to metronidazole with regard to resolution of clinical symptoms at seven days, adverse effects, and 31-day relapse rates. However, each study group ultimately had fewer patients than was calculated as necessary to achieve 80% power in evaluating the null hypothesis. Therefore, larger studies need to be performed to validate these findings and open up another avenue for treatment of C. difficile-associated diarrhea.

Another potential treatment option being investigated is the use of tolevemer, a soluble anionic polymer that binds toxins A and B, which has been shown in animal studies of C. difficile to produce a lasting cure. A multicenter, randomized, double-blind phase 2 trial published in Clinical Infectious Diseases in 2006, comparing oral vancomycin to tolevemer in the treatment of mild to moderate C. difficile-associated disease (3 to 12 stools/day) in adults, found that a 14-day oral course of tolevemer 2 grams three times daily is noninferior to a 10-day course of oral vancomycin 125 mg four times daily with regard to clinical resolution of disease. Recurrence rates were not statistically significant, but there was a trend toward lower recurrence rates in the tolevemer-treated group.

Of note, patients treated with tolevemer were statistically more likely to develop hypokalemia than those in the vancomycin-treated group. Because this presents an increased risk for arrhythmias in patients with underlying cardiac conditions (a common patient population receiving hospital care and therefore at higher risk for developing C. difficile-associated disease), it is possible that the use of this drug will be somewhat limited if it is ultimately approved for use in the treatment of this disease.

Other toxin-binding resins have been proposed as adjunctive therapy for C. difficile-associated disease. One such drug—cholestyramine, which is a bile acid-binding resin approved as a cholesterol-lowering agent—has been shown in animal models to bind cytotoxin; however, in human studies, there was no effect on the amount of C. difficile toxin excreted. In addition, cholestyramine binds to orally administered vancomycin, which may lead to a decrease in the efficacy of this therapy. As such, the use of these anion-exchange resins may have limited application in the treatment of C. difficile-associated disease.

Lastly, studies are beginning to address prevention of C. difficile-associated disease with probiotics initiated concurrently with antimicrobials for other infections. Specifically, Saccharomyces boulardii and various lactobacilli have been studied. A meta-analysis published in American Journal of Gastroenterology in 2006 looked at prophylactic probiotic use for C. difficile-associated diarrhea; it evaluated six studies, each utilizing different probiotic strains, and found that overall there was a statistically significant protective effect with the use of these agents. However, within these studies, only S. boulardii was shown to reduce C. difficile recurrence rates. Given the outcomes of this and other studies, there might be potential for disease prevention with the concomitant initiation of antimicrobials and probiotics.

BOTTOM LINE

As demonstrated by our case report, with the emergence of a highly toxigenic strain of C. difficile affecting patients without obvious risk factors for disease, any patient presenting with severe diarrhea should be suspected of having C. difficile infection and appropriately tested with an EIA to detect toxin production. On diagnosis, prompt treatment with either oral metronidazole or oral vancomycin should begin. Despite equivalent efficacy, metronidazole may be preferred therapy in the IDSA guidelines due to its lower cost and the risk of creating vancomycin-resistant organisms. Clinical response should be apparent by six days into a course of therapy. Newer agents, including nitazoxanide and tolevemer, are being investigated, but their places in the treatment pathway have yet to be defined.

back to top

Suggested Reading Aslam S and Musher DM: An update on diagnosis, treatment, and prevention of Clostridium difficile-associated disease. Gastroenterol Clin N Am 35(2):315, 2006. Chernac E, et al.: Severe Clostridium difficile-associated disease in populations previously at low risk—four states. MMWR Morb Mortal Wkly Rep 54(47):1201, 2005. Dial S, et al.: Use of gastric acid-suppressive agents and the risk of community-acquired Clostridium difficile-associated disease. JAMA 294(23):2989, 2005. Gerding DN, et al.: Clostridium difficile-associated diarrhea and colitis. Infect Control Hosp Epidemiol 16(8):459, 1995. Mogg GA, et al.: Randomized controlled trial of colestipol in antibiotic-associated colitis. Br J Surg 69(3):137, 1982. Musher DM, et al.: Nitazoxanide for the treatment of Clostridium difficile colitis. Clin Infect Dis 43(4):421, 2006. Pantosti A, et al.: Comparison of the in vitro activities of teicoplanin and vancomycin against Clostridium difficile and their interactions with cholestyramine. Antimicrob Agents Chemother 28(6):847, 1985. Poutanen SM and Simor AE. Clostridium difficile-associated diarrhea in adults. CMAJ 171(1):51, 2004. Taylor NS and Bartlett JG: Binding of Clostridium difficile cytotoxin and vancomycin by anion-exchange resins. J Infect Dis 141(1):92, 1980.

Printable version of this article

E-mail this article