Metformin Discontinuation prior to FDG PET/CT: A Randomized Controlled Study to Compare 24- and 48-hour Bowel Activity

Purpose To investigate the relationship of 24- and 48-hour metformin discontinuation to bowel uptake of fluorine 18 fluorodeoxyglucose (FDG) on PET/CT scans. Materials and Methods Patients with diabetes who were treated with metformin and referred for FDG PET/CT were randomized to three equal groups based on duration of metformin discontinuation: 24 hours, 48 hours, and no discontinuation (control group). Two interpreters blinded to the study groups assessed FDG uptake in multiple segments of small and large bowel qualitatively and semiquantitatively by using maximum standardized uptake values (SUVsmax). Differences in age, sex, weight, dose of metformin, duration of metformin treatment, blood glucose levels, and FDG dose injected were assessed. Data were analyzed with analysis of variance when passing normality, and by nonparametric testing when not. Results Ninety study participants (62 male, 28 female; median age, 70 years) were enrolled from July 2010 through March 2012. There were no differences between study groups in weight, blood glucose levels 3 days prior to scanning, or normal organ uptake. Large bowel SUVmax was lower after 24 hours (4.10 ± 2.00 vs 5.42 ± 2.36; P = .020) and 48 hours (2.63 ± 0.88 vs 5.42 ± 2.36; P ˂ .001) of metformin discontinuation than for no discontinuation (control), and for 48 hours versus 24 hours of discontinuation (P = .0015). Small bowel SUVmax was lower after 24 hours (2.86 ± 0.67 vs 3.73 ± 1.08 control; P ˂ .001) and 48 hours (2.78 ± 0.73 vs 3.73 ± 1.08 control; P ˂ .001) of metformin discontinuation versus no metformin discontinuation, but not for 48 hours versus 24 hours of discontinuation (P = .57). Examination-day blood glucose levels increased after 48-hour withdrawal of metformin (8.41 mmol/L ± 2.86 vs 6.83 mmol/L ± 2.13 control; P = .002). Conclusion Metformin discontinuation for 48 hours prior to PET/CT was associated with lower accumulation of fluorodeoxyglucose in the bowel, compared to when there was no discontinuation (control group) or 24-hour discontinuation of metformin. © RSNA, 2018.