Hepatic and pancreatic fat as imaging biomarkers of metabolic syndrome

Abstract

ObjectiveTo objectively evaluate hepatic and pancreatic involvement in metabolic syndrome through magnetic resonance imaging (MRI) biomarkers. Material and methodsFrom an initial retrospective sample of 407 patients diagnosed with metabolic syndrome studied by MRI in a single center during a 2-year period, 154 were excluded because of a lack of clinical and/or laboratory data, pancreatic abnormalities, or inadequate quality of MRI studies. To measure hepatic and pancreatic fat, we used chemical shift imaging (in-phase and out-of-phase), measuring the fat fraction (%) in regions of interest in the pancreas and liver. Associations between the fat fraction and selected clinical and laboratory variables were assessed with beta regression models. ResultsIn the end, 253 patients were included. The hepatic fat fraction was 4.9% and the pancreatic fat fraction was 7.9%. We found no significant associations between the hepatic fat fraction and any of the clinical or laboratory variables. However, the pancreatic fat fraction was positively associated with age (OR=1.025, p<0.001) and baseline glucose (OR=1.005, p<0.001). Patients with diabetes had higher values of pancreatic fat fraction (OR=2.64, p=0.038). Pancreatic fat fraction and hepatic fat fraction were positively associated (OR=69.44, p<0.001). ConclusionsPancreatic steatosis can be considered a marker of metabolic syndrome and diabetes. Quantitative MRI enables the diagnosis and grading of fatty pancreas through simple chemical shift techniques.