Introduction Mounting evidence implicates gut microbial dysbiosis in the pathogenesis of non-alcoholic steatohepatitis (NASH) by mechanisms including caloric salvage, lipopolysaccharide production, upregulation of proinflammatory cytokines, increased insulin resistance and consequent increases in body mass and hepatic steatosis. Rifaximin is a minimally-absorbed, gut-selective antibiotic with bactericidal activity against a broad spectrum of gut microbes, making it an attractive candidate therapy. We aimed to study the effect of Rifaximin on markers of hepatic inflammation, hepatic steatosis, hepatic and peripheral insulin sensitivity.

Methods Patients with biopsy-proven NASH, elevated aminotransferase values and no hepatic comorbidities were included in this open-label, randomised, cross-over study, all receiving 6 weeks of Rifaximin 400mg twice daily, before or after one of two 6 week observation periods on standard therapy. The primary endpoint was change in alanine aminotransferase (ALT) values after Rifaximin therapy. Secondary endpoints were change in percentage hepatic lipid assessed by hepatic proton magnetic resonance spectroscopy and change in hepatic and peripheral insulin sensitivity assessed by the hyperinsulinaemic euglycaemic clamp. Patients also had anthropometrics, serum biochemistry and cytokine profiling at each timepoint. Stool and urine were collected for subsequent analysis.

Results 15 patients, 13 male, 2 female, mean (SD) age 48(9.5) years were included. 7 had diabetes on oral hypoglycaemic medications and 8 did not have diabetes. After 6 weeks of therapy, there was no difference in ALT before, 69 (40)IU/L, and after, 71 (43)IU/L, treatment, p = 0.7. Hepatic lipid content was 23.3(12.8)% before and 26.5(15.9)% after Rifaximin, p = 0.16. Peripheral insulin sensitivity (Rd) was unchanged, 29.5 (6.5) to 29.4 (10.0) mmol/kg min, p = 0.91, hepatic insulin sensitivity (% suppression of endogenous glucose production) was unchanged (35.2(10.1)% to 31.2(13.2)%, p = 0.35). There were no significant differences in body mass index, waist and hip circumference, IL 1b, IL6, IL10, IL18, CD14, TNFa, Leptin, Resistin and Adiponectin values with treatment.

Conclusion Treatment with Rifaximin was not associated with changes in markers of hepatocellular damage, hepatic lipid content, cytokine profile or insulin sensitivity in patients with NASH.

Disclosure of Interest None Declared