Continuous flushing of the bladder in rodents reduces artifacts and improves quantification in molecular imaging

In this study, we evaluated the partial volume effect (PVE) of 2-deoxy-2-[18F]fluoro-d-glucose (18F-FDG) tracer accumulation in the bladder on the positron emission tomographic (PET) image quantification in mice and rats suffering from inflammatory bowel disease. To improve the accuracy, we implemented continuous bladder flushing procedures. Female mice and rats were scanned using microPET/computed tomography (CT) at baseline and after induction of acute colitis by injecting 2,4,6-trinitrobenzene sulfonic acid (TNBS) intrarectally. During the scans, the bladder was continuously flushed in one group, whereas in the other group, no bladder flushing was performed. As a means of in vivo and ex vivo validation of the inflammation, animals also underwent colonoscopy and were sacrificed for gamma counting (subpopulation) and to score the colonic damage both micro- and macroscopically as well as biochemically. At baseline, the microPET signal in the colon of both mice and rats was significantly higher in the nonflushed group compared to the flushed group, caused by the PVE of tracer activity in the bladder. Hence, the colonoscopy and postmortem analyses showed no significant differences at baseline between the flushed and nonflushed animals. TNBS induced significant colonic inflammation, as revealed by colonoscopic and postmortem scores, which was not detected by microPET in the mice without bladder flushing, again because of spillover of bladder activity in the colonic area. MicroPET in bladder-flushed animals did reveal a significant increase in 18F-FDG uptake. Correlations between microPET and colonoscopy, macroscopy, microscopy, and myeloperoxidase yielded higher Spearman rho values in mice with continuously flushed bladders during imaging. Comparable, although somewhat less pronounced, results were shown in the rat. Continuous bladder flushing reduced image artifacts and is mandatory for accurate image quantification in the pelvic region for both mice and rats. We designed and validated experimental protocols to facilitate such.