Selectivity of 18F-FLT and 18F-FDG for differentiating tumor from inflammation in a rodent model.

Increased glucose metabolism of inflammatory tissues is the main source of false-positive (18)F-FDG PET findings in oncology. It has been suggested that radiolabeled nucleosides might be more tumor specific. To test this hypothesis, we compared the biodistribution of 3'-deoxy-3'-(18)F-fluorothymidine (FLT) and (18)F-FDG in Wistar rats that bore tumors (C6 rat glioma in the right shoulder) and also had sterile inflammation in the left calf muscle (induced by injection of 0.1 mL of turpentine). Twenty-four hours after turpentine injection, the rats received an intravenous bolus (30 MBq) of either (18)F-FLT (n = 5) or (18)F-FDG (n = 5). Pretreatment of the animals with thymidine phosphorylase (>1,000 U/kg, intravenously) before injection of (18)F-FLT proved to be necessary to reduce the serum levels of endogenous thymidine and achieve satisfactory tumor uptake of radioactivity. Tumor-to-muscle ratios of (18)F-FDG at 2 h after injection (13.2 +/- 3.0) were higher than those of (18)F-FLT (3.8 +/- 1.3). (18)F-FDG showed high physiologic uptake in brain and heart, whereas (18)F-FLT was avidly taken up by bone marrow. (18)F-FDG accumulated in the inflamed muscle, with 4.8 +/- 1.2 times higher uptake in the affected thigh than in the contralateral healthy thigh, in contrast to (18)F-FLT, for which this ratio was not significantly different from unity (1.3 +/- 0.4). In (18)F-FDG PET images, both tumor and inflammation were visible, but (18)F-FLT PET showed only the tumor. Thus, the hypothesis that (18)F-FLT has a higher tumor specificity was confirmed in our animal model.