Clostridium difficile Infection Was Not Detected in Patients Who Received Rifaximin for Hepatic Encephalopathy in Community and University Practices

Abstract

Purpose: Clostridium difficile infection has been associated with several types of antibiotics, including fluoroquinolones. The incidence of C difficile infection related to the nonabsorbed antibiotic rifaximin is unknown. Rifaximin is approved for the treatment of travelers' diarrhea caused by noninvasive strains of Escherichia coli and is being investigated for the treatment of hepatic encephalopathy (HE). This retrospective study evaluated the incidence of C difficile infection in patients with cirrhosis who received rifaximin for treatment of HE in community and university gastroenterology practices. Methods: Medical charts for all patients diagnosed with cirrhosis who received oral rifaximin for the treatment of HE between January 2004 and May 2008 were reviewed; patients had been treated at 1 university practice and 3 community practices. Patients who developed diarrhea (>6 bowel movements/d) during treatment with rifaximin were included in the analyses. Stool samples had been analyzed by cytotoxin assay testing to determine if diarrheal symptoms (ie, markedly increased stool frequency and decreased stool viscosity) were due to C difficile infection. Results: Analyses included 212 patients with cirrhosis who received rifaximin for the treatment of HE. The mean dose of rifaximin was 1055 mg/d (range, 600–1600 mg/d) for a mean duration of 250 days (range, 180–385 d). Of the 212 patients who received rifaximin, 155 were treated in a university practice and 57 were treated in community practices; 97% of patients had been diagnosed with grade 2 or 3 HE. Eighteen patients (8%) developed diarrhea during rifaximin treatment; 13 were male and 5 were female (mean age, 52 y). Twelve of the 18 patients (67%) who developed diarrhea had received treatment in the university setting and the remaining 6 (33%) had received treatment in community practices. Stool cytotoxin test results were negative for C difficile in all 18 of these patients. Diarrheal symptoms resolved in all cases with standard therapy administered after stool analyses had excluded infection. Conclusion: These findings suggest that long-term treatment with rifaximin was not associated with the development of C difficile infection in patients with cirrhosis who received rifaximin for treatment of HE. Therefore, rifaximin does not appear to increase risk of C difficile infection. Further investigations of the incidence of C difficile infection in patients who receive rifaximin are warranted.