Effects of caloric restriction and bariatric surgery on non-alcoholic fatty liver disease

Non-alcoholic fatty liver disease (NAFLD) represents the hepatic manifestation of the metabolic syndrome. It encompasses a broad spectrum of severity, from steatosis through to inflammation (non-alcoholic steatohepatitis) and subsequent progressive liver fibrosis. It is rapidly becoming the leading cause of cirrhosis and liver transplantation worldwide and, even in advanced disease, liver biochemistry is often normal. No pharmaceutical agents are currently licensed to treat NAFLD; its clinical management is focused on weight loss (via reduced energy intake and exercise-induced increased energy expenditure), and addressing metabolic comorbidities and liver disease risk factors. Bariatric surgery is recognised as a highly effective treatment for obesity and type 2 diabetes (T2D). Additionally, it appears to usually improve NAFLD, although it is not currently recommended in the UK’s clinical guidelines for NAFLD management. In contrast to T2D, where improved postoperative glycaemic control is expected, cases of postoperative liver dysfunction have been reported when advanced NAFLD was present at the time of surgery. The mechanisms underlying changes in NAFLD after bariatric surgery and the optimal surgical intervention for its treatment are unknown. In order to evaluate current clinical practice and the views of key stakeholders in NAFLD management, I developed parallel surveys for patients, bariatric surgeons and hepatologists. These identified that most patients would consider bariatric surgery as an intervention for the combined treatment of their obesity and NAFLD, but that this option is usually not discussed by health professionals. Most centres did not have defined protocols for NAFLD staging and monitoring in patients undergoing bariatric surgery, and surgeons rarely used non-invasive biomarkers to predict NAFLD severity and guide clinical care. Hepatologists identified bariatric surgery’s omission from NAFLD guidelines as the major barrier to the wider implementation of bariatric surgery as a treatment for NAFLD. Advanced fibrotic NAFLD is associated with adverse clinical outcomes such as liver and cardiovascular morbidity and mortality. Since isolated liver blood tests poorly predict liver histology, non-invasive biomarkers have previously been developed to improve diagnostic accuracy. However, they have undergone limited validation in patients with severe obesity and NAFLD. Therefore, I performed a comprehensive assessment of the performance of a range of existing non-invasive biomarkers and found that these performed relatively poorly in this patient population. Machine learning approaches were then used to develop and validate a novel model that incorporated routine clinical parameters and offered improved diagnostic accuracy for advanced fibrotic NAFLD prediction. Clinical outcomes after conventional medical or bariatric surgical treatment of NAFLD have not previously been compared. Therefore, a retrospective cohort study was designed for this purpose. Reduced mortality and greater improvements in a range of key clinical targets in NAFLD management were found after management with bariatric surgery. Patients undergoing bariatric surgery are often prescribed preoperative low energy diets. These are designed to reduce liver volume and improve surgical visualisation, but it is unclear to what extent these may also improve NAFLD. A prospective study was performed to characterise the effects of a routine preoperative diet in non-diabetic females with severe obesity. Significant improvements in liver and metabolic health outcome measures were observed but these remained elevated relative to non-obese control participants. 5-10% total body weight loss is often recommended for patients with NAFLD, but it is unclear whether this is sufficient for people with severe obesity. I randomised such individuals to undergo either sleeve gastrectomy (SG) or Roux-en-Y gastric bypass (RYGB) surgery and studied them after matched 20% postoperative weight loss. Whilst postoperative improvements were observed, these were still insufficient to reverse NAFLD relative to non-obese control participants. Preliminary results show normalisation of liver steatosis 1 year after surgery. Similar changes occurred after SG and RYGB surgery and I conclude that these interventions represent promising interventions for the combined treatment of obesity and NAFLD.