For HIV-infected patients who have access to treatment, liver diseases are a major cause of morbidity and mortality. Hepatitis C is the most frequently encountered liver condition in this population. Both diseases allow a higher level of poisonous substances (toxins) normally produced by the bacteria present in the gut to enter the bloodstream. This leads to a chronic inflammatory state, which results in faster development of liver scars (fibrosis) and ultimately, end stage disease (cirrhosis). To prevent this from happening, the use of antibiotics has been attempted to reduce the quantity of gut flora in the hopes of lowering the amount of toxins produced. These trials have shown promising results, but the antibiotics studied had major side effects and were not designed for continuous use. Rifaximin is a non absorbable antibiotic with very few side effects. It is already used for long periods of time in cirrhotic patients to treat the effects of cirrhosis on the brain (encephalopathy). This project will try to determine if rifaximin, by reducing the level of toxins produced by the bacteria in the gut, can improve the evolution of liver fibrosis in HIV-infected patients with hepatitis C. In this pilot study, ten patients with HIV and HCV infection will be followed for one year. In addition, 10 patients with HCV mono infection will also be followed. Both populations will be included if they are starting on rifaximin, for its currently approved FDA indication (hepatic encephalopathy).

Many studies have already proved the deleterious effects of LPS on HIV and hepatic diseases evolution. There has never been a concerted effort to prevent the progression of liver disease in these patients. To date, the only treatment is initiation of antiretroviral therapy. Rifaximin could be an easy and well tolerated way to improve the outcome of liver disease in HIV-infected patients. We hypothesize that it could help to slow down the progression of liver disease at any stage in these patients. This is a pilot study. A total of twenty patients placed on rifaximin by their physician for a mild hepatic encephalopathy will be monitored over a period of one year. The evaluation of the fibrosis will be done through transient elastography every 3 months. Bacterial translocation will be evaluated through the dosing of soluble CD14. The safety of the prolonged use of rifaximin in patients will also be assessed.

Inclusion Criteria:

• Patients aged over 18 years old that can give an informed consent
• HIV-infected patients with hepatitis C associated liver disease demonstrated by a fibroscan score above 8 kiloPascals
• HCV infected patients with liver disease demonstrated by a fibroscan score above 8 kiloPascals
• Patients placed on rifaximin by their physician for a mild hepatic encephalopathy

Exclusion Criteria:

• Any patient unable to give informed consent.
• Patients on hepatitis C treatment
• Patients allergic to rifaximin or rifamycin
• Patients on any prolonged antibiotic treatment including patients on tuberculosis treatment.
• Patients with history of Clostridium difficile infection
• Uncontrolled HIV infection: CD4 less than 200 or detectable viral load Patients need to be on a stable ART regimen for at least one month.
• Patient on a HIV regimen including an unboosted protease inhibitor.
• Acute hepatitis of any cause.
• Child C cirrhosis
• Patients on dialysis
• Pregnant women or childbearing age women not accepting to use an effective contraceptive method Acceptable methods are double barrier methods (condom with spermicide jelly or diaphragm with spermicide), hormonal methods (oral contraceptives, patches or medroxyprogesterone acetate), or an intrauterine device with a documented failure rate of less than 1% per year. Females who have been surgically sterilized (e.g., hysterectomy or bilateral tubal ligation) or who are postmenopausal (total cessation of menses for >1 year) will not be considered "females of childbearing potential."
• Usual exclusion criteria for FibroScan (pregnancy, BMI over 40, ascites, pacemaker or defibrillator).