Clostridium difficile infection (CDI) is a healthcare associated infection of growing importance that occurs after prolonged hospitalization and exposure to broad-spectrum antibiotics [1,2]. CDI has a wide range of clinical manifestations, including asymptomatic infection, mild diarrhea, fulminant pseudomembranous colitis (PMC), and even death [3,4]. The incidence and severity of CDI in Korea, is increased in recent years [5]. According to the national wide study, the total incidence of CDI in Korea was 1.7 cases/1,000 adult admissions in 2004, and 2.7/1,000 adult admissions in 2008 (P=0.028) [6]. Most clinical practice guidelines recommend oral metronidazole as the initial therapy for the treatment of mild to moderate CDI, then oral vancomycin reserved for severe CDI, or intolerance to or failure to respond to metronidazole treatment [7,8]. However, in a real world clinical setting with severe CDI, the question to whom to use oral vancomycin is still a challenging one, due to the risk of emergence of the vancomycin-resistant enterococci and the high cost of vancomycin [9,10]. There is an issue about the over-enrollment of severe CDI with current criteria which suggests potential efficacy of less active agents for these “severe” patients [11] (Table 1): for example, a patient aged over 60 years old presenting with a fever (body temperature >38.3°C) will be classified as having severe CDI regardless of any co-morbidity. In actual clinical practice, oral metronidazole is prescribed to many patients who would be formally classified as presenting with severe CDI due to its tolerability and low cost. Therefore, we compared the clinical efficacy (clinical cure, recurrence, and 30-day mortality) of oral metronidazole with that of oral vancomycin in patients with severe CDI, and evaluated the risk factors associated with treatment failure.

The Society for Healthcare Epidemiology of America/Infectious Diseases Society of America guideline and the American Gastroenterology Association recommend oral vancomycin as the first-line agent for the treatment of severe CDI patients [7,13].

Oral metronidazole treatment is generally considered to be less effective than oral vancomycin treatment, because cases of elderly patients with severe comorbidities are increasing, and because the use of broad-spectrum antibiotics (new fluoroquinolones and cephalosporins) is also increasing [14–16]. Metronidazole resistance, the emergence of the BI/NAP1/027 (ribotype 027) strain of C. difficile, and poor concentration of oral metronidazole in colon are also potential causes of the decreasing efficacy of oral metronidazole.

The emergence of the hypervirulent BI/NAP1/027 (ribotype 027) strain of C. difficile is associated with more severe disease due to increased toxin production. In 2009, Tae et al. [17] reported the first case of C. difficile polymerase chain reaction (PCR) ribotype 027 infection refractory to oral metronidazole therapy in Korea. However, the proportion of this hypervirulent strain is low [18]. Kim et al. [19] performed a prospective study that found that binary toxin-producing C. difficile infection, including ribotype 027, was not common in Korea and that these isolates were highly susceptible to metronidazole and vancomycin.

Metronidazole-resistant strains of C. difficile are rare. In 1999, for the first time, Wong et al. [20] detected a metronidazole-resistant strain of C. difficile in a patient with CDI. In Korea, Kim et al. [21] reported that no metronidazole resistance was detected. Metronidazole resistance does not appear to be increasing [22]. Therefore, the emergence of the BI/NAP1/027 strain and metronidazole resistance does not have a critical influence on the efficacy of oral metronidazole treatment in severe CDI patients in Korea.

The antibiotic concentration in the colon is very important for the treatment for CDI. The difference in treatment effect for CDI between oral metronidazole and oral vancomycin is due to the difference in the absorption rate in the colon. When oral metronidazole is administered, it is absorbed rapidly and almost completely, with only 6%–15% of the drug metabolites excreted in the stool. This concentration rate of metronidazole in stool reflects the rate of its secretion in the colon. Once treatment of CDI is initiated, the stool concentration rate of metronidazole decreases rapidly from 9.3 μg/g in watery stools to 1.2 μg/g in formed stools [23]. In asymptomatic C. difficile carriers, metronidazole is not detected in stool [24]. In contrast, oral vancomycin is poorly absorbed; thus, fecal concentrations are maintained throughout treatment. Johnson et al. [24] reported that the fecal concentration rate after oral vancomycin administration (125 mg administered 4 times daily) is generally >1,000 μg/g. Oral metronidazole treatment is reported to fail, because the concentration of metronidazole in colon was slowly increased but was not maintained at a consistent rate in colon [10,22,25]. If a patient shows no clinical response to oral metronidazole, clinicians should consider changing medications (e.g., oral vancomycin or intravenous metronidazole). In our study, all treatment failure patients receiving oral metronidazole achieved a clinical cure after switching to oral vancomycin or intravenous metronidazole.

Especially, patients with severe complicated CDI, with symptoms including hypotension, shock, and/or megacolon, do not have enough time to wait until metronidazole treatment will take effect and a clinical cure will be achieved [7]. Thus, severe complicated CDI patients should be treated in accordance with current clinical practice guidelines, based on high dose vancomycin (500 mg 4 times daily) [7].

In our study, H2 antagonist treatment and fever (body temperature >38.3°C) were related to treatment failure. There are some reports supporting the association between H2 antagonist therapy and CDI treatment failure. The mechanism of how H2 antagonist exposure contributes to an increased risk of CDI treatment failure is not yet clear. However, it is strongly suspected that the effect of H2 antagonist in suppressing gastric acid secretion aids the survival of C. difficile [26–28]. Dial et al. [29] reported that the use of acid-suppressive therapy is associated with an increased risk of CDI. However, Jung et al. [30] did not find an association between the use of H2 antagonist and metronidazole treatment failure. Further studies on the role of acid-suppressive therapy in CDI treatment appear to be necessary.

This study had some limitations: it was retrospective, taking place in a single tertiary medical center. The number of patients was relatively small. In addition, the hospital used a toxin assay for diagnosis, which is less sensitive than diagnosis by C. difficile PCR and C. difficile culture.

In conclusion, we found that in treating severe CDI patients, oral metronidazole was an effective treatment in addition to oral vancomycin. Therefore, oral metronidazole can be viewed as a valid clinical option for treating not only patients with mild to moderate CDI but also those with severe CDI.