Background

Obesity is associated with neuroinflammation and non-alcoholic steatohepatitis (NASH). While bariatric surgery is known to improve NASH and cognitive impairment, the linkage between liver-brain inflammatory axis and weight loss is understudied. Inflammation can alter glucose transport and metabolism. We used glucose analogue 18F-fluorodeoxyglucose (FDG) and total-body, dynamic positron emission tomography (PET) to study the effect of weight loss from RYGB on the liver-brain axis. Patients undergoing RYGB with liver biopsies had pre- and 6-month post-operative labs, dynamic FDG-PET, and MR-Elastography (MRE, fibrosis) and Proton Density Fat Fraction (MR-PDFF, steatosis) imaging studies performed. Liver biopsies were scored according to established NASH-CRN criteria. Tracer kinetic modeling was used to provide quantitative measures of the rate of FDG transport from blood to tissue (K1) and FDG net influx rate (Ki) that assesses glucose metabolism. Pre/post-operative mean values were compared by paired t-tests and Pearson’s correlation coefficients between parameters were calculated. Three of four patients had NASH based on pathology scores. Significant changes in BMI, HDL, insulin, HOMA-IR for insulin resistance, and hepatic steatosis were found at 6-months postoperatively (Table 1). Worse baseline NAS scores correlated with greater reduction in fibrosis. Though liver K1 (previously found to be associated with NASH severity) did not markedly change postoperatively, brain Ki significantly increased, suggesting RYGB induced increases in brain glucose metabolism. Non-invasive imaging detected post-RYGB steatosis improvement. Patients with NASH trended to have post-RYGB changes in brain glucose metabolism, possibly representing altered neuroinflammation. Further studies are needed to correlate imaging findings with cognitive clinical relevance.