Non-alcoholic fatty liver disease (NAFLD) includes benign hepatic simple steatosis (SS) and steatohepatitis (NASH), which is characterised by inflammation leading to fibrosis and cirrhosis. NAFLD is the hepatic manifestation of the metabolic syndrome, and the prevalence is 74-98% in morbidly obese individuals undergoing bariatric surgery. Although steatosis improves post bariatric surgery, hepatic inflammation and fibrosis do not consistently improve. Alterations of the human gut flora (intestinal microbiota; IM) may play a role. One mechanism linking IM to obesity, insulin resistance (IR), and NAFLD is through translocation of bacterial lipopolisaccharide (LPS=endotoxin) into the blood stream (=endotoxemia), causing chronic inflammation. Morbidly obese subjects have different IM compared to lean controls, and the IM structure is significantly altered after bariatric surgery, probably due to a combination of anatomic changes, diet, and weight loss. For example, the ratio of Firmicutes/Bacteroidetes may be lower in obese subjects compared to lean controls and lower numbers of Faecalibacterium prausnitzii were reported in some obese subjects before bariatric surgery, which increased 3 months post-surgery. This is of interest since, in animal studies, low abundance of F. prausnitzii, a butyrate producing bacterium, is associated with increased intestinal permeability, endotoxemia, and inflammation. To our knowledge, only two studies are available describing IM in patients pre and post bariatric surgery, and no data have been published on the relationship between IM and NAFLD in these patients.

Study Design: A. Cross-sectional study: Sixty patients with morbid obesity undergoing bariatric surgery diagnosed with NAFLD on liver biopsy (30 SS, 30 NASH). Main hypothesis: The ratio of Firmicutes/Bacteroidetes is higher in stool samples from morbidly obese subjects with NASH compared to SS. Other differences in IM composition exist. Objective: to compare bacterial dynamics using Illumina technology to assess the IM. The relative abundance of the dominant fecal microorganisms (including Firmicutes, Archaea, Bacteroides, Bifidobacteria, Mollicutes, Enterobacteriaceae, Clostridia clusters, F. prausnitzii, Roseburia, and Lactobacilli) will also be assessed by real-time PCR. Sub-hypotheses: In NASH compared to SS, there will be: a) lower fecal butyrate concentration; b) higher endotoxin and other inflammatory markers (TNF-α, IL-6) in plasma. Potential covariates assessed: small intestinal bacterial overgrowth (SIBO), measured by glucose hydrogen breath test (GHBT), which can contribute to endotoxemia and inflammation; IR, diabetes status, lipid profile, plasma vitamin E, liver enzymes, anthropometry, food intake, physical activity and environmental factors.

B. Prospective cohort study: Patients undergoing bariatric surgery with either SS or NASH (up to 60 of them recruited from Part A) will be followed prospectively over 12 months to assess changes in the IM and liver histology. Goal is to have 60 subjects who complete the study with a 2nd liver biopsy. Main Hypothesis: In morbidly obese patients with NAFLD (SS or NASH), changes in IM post bariatric surgery will be associated with changes in liver histology. Specifically, an increased number of F. prausnitzii in feces with be associated with improvement in liver histology while a reduction will be associated with deterioration of liver histology. Objective: To correlate changes in liver histology (NAFLD activity score [NAS], inflammation, fibrosis, steatosis) between 0 and 12 months with changes in F. prausnitzii. Other changes of the fecal IM community structure, fecal short chain fatty acids (including butyrate), plasma endotoxin, inflammatory markers (TNF-α, IL-6) and SIBO will also be assessed, in addition to diet, activity, weight change, improvement of diabetes and plasma vitamin E. Secondary hypotheses: Increased number of F. prausnitzii in feces will be associated with increased fecal butyrate, lower serum endotoxin and lower inflammatory markers (TNF-α, IL-6) in plasma.

Significance: In humans with morbid obesity and NAFLD undergoing bariatric surgery, very little data are available on IM and its metabolic effect and contribution to NAFLD. These studies will add more information regarding the role of IM and its effect on potential mechanisms contributing to NAFLD. It will also provide us with pilot data for future intervention studies assessing the potential use of pre- or probiotics for NAFLD in morbidly obese subjects in the setting of bariatric surgery.

Inclusion Criteria:

• BMI>40 kg/m2 or BMI>35-40 kg/m2 with severe weight loss comorbidities
• Male or female, equal or over 18 years of age
• Alcohol consumption is leass than 20 g/d

Exclusion Criteria:

• No diagnosis of NAFLD
• Having liver disease of other etiology
• Having advance liver disease
• Having abnormal coagulation or other reason contraindicating a Liver Biopsy
• On medication known to precipitate steatohepatitis 6 months prior to entry
• On regular intake of non-steroidal anti-inflammatory drugs, prebiotics, probiotics and antibiotics, ursodeoxycholic or any experimental drug in the 3 months prior to study entry
• Having type-1 diabetes, chronic gastrointestinal diseases, previous gastrointestinal surgery modifying the anatomy (prior to bariatric surgery)
• Smoking
• Pregnancy or Breastfeeding

NAFLD is the hepatic manifestation of the metabolic syndrome, and the prevalence is 74-98% in morbidly obese individuals. The aim of this study is to examine the role of intestinal microbiota (IM) in non-alcoholic fatty liver disease (NAFLD) in morbidly obese patients undergoing Roux-en-Y gastric bypass surgery. Alterations of the human gut flora (intestinal microbiota) will be determined before and after surgery in realtion with the change of liver histology.