Metformin discontinuation less than 72 h is suboptimal for F-18 FDG PET/CT interpretation of the bowel

1. Tissue accumulation of the antihyperglycaemic agent metformin (dimethylbiguanide) was examined after oral administration to the normal and streptozotocin (STZ) diabetic mouse. 2. Metformin (50 mg/kg body weight containing 14C-metformin 25 microCi/kg body weight), which is stable and not metabolized, resulted in maximum plasma concentrations at 0.5 h which declined to 1000 mumol/kg wet weight at 0.5-2 h, but declined to < 2% of maximum by 24 h. 4. Stomach, colon, salivary gland, kidney and liver accumulated metformin more than two-fold, and concentrations of the drug in heart and skeletal (gastrocnemius) muscle were greater than plasma concentrations on some occasions up to 8 h. 5. In a separate study, i.v.-administered metformin was selectively accumulated by tissues of the small intestine. Thus, retention of metformin by tissues of the small intestine may represent a deep compartment for the drug.

This prospective and bi-centric study was conducted in order to determine the impact of antidiabetic treatments (AD) on (18)F-FDG bowel uptake in type 2 diabetic patients. Fifty-five patients with previously diagnosed and treated type 2 diabetes mellitus (group 1) were divided in two subgroups: AD treatment including metformin (n=32; group 1a) and AD treatment excluding metformin (n=23; group 1b). The 95 patients without diabetes mellitus made up controls (group 2). (18)F-FDG uptake in small intestine and colon was visually graded and semi-quantitatively measured using the maximum standardized uptake value. (18)F-FDG bowel uptake was significantly increased in AD patients (group 1) as compared to controls (group 2) (p<0.001). Bowel uptake was significantly higher in AD patients including metformin (group 1a) as compared to AD patients excluding metformin (group 1b) (p<0.01), whose bowel uptake was not significantly different from controls (group 2). A metformin treatment was predictive of an increased bowel uptake in the small intestine (odds ratio OR=16.9, p<0.0001) and in the colon (OR=95.3, p<0.0001), independently of the other factors considered in the multivariate analysis. Bowel uptake pattern in the patients treated with metformin was typically intense, diffuse and continuous along the bowel, strongly predominant in the colon, in both the digestive wall and lumen. This study emphasizes that metformin significantly increases (18)F-FDG uptake in colon and, to a lesser extent, in small intestine. It raises the question of stopping metformin treatment before an (18)F-FDG PET/CT scan is performed for intra-abdominal neoplasic lesion assessment.