Hepatic and pancreatic fat as imaging biomarkers of metabolic syndrome.

OBJECTIVE: To objectively evaluate hepatic and pancreatic involvement in metabolic syndrome through magnetic resonance imaging (MRI) biomarkers. MATERIAL AND METHODS: From 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. RESULTS: In 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). CONCLUSIONS: Pancreatic 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.