Imaging proliferation in brain tumors with 18F-FLT PET: comparison with 18F-FDG.

3'-Deoxy-3'-(18)F-fluorothymidine ((18)F-FLT) is a recently developed PET tracer to image tumor cell proliferation. We characterized (18)F-FLT PET of brain gliomas and compared (18)F-FLT with (18)F-FDG PET in side-by-side studies of the same patients. METHODS: Twenty-five patients with newly diagnosed or previously treated glioma underwent PET with (18)F-FLT and (18)F-FDG on consecutive days. Three stable patients in long-term remission were included as negative control subjects. Tracer kinetics in normal brain and tumor were measured. Uptake of (18)F-FLT and (18)F-FDG was quantified by the standardized uptake value (SUV) and the tumor-to-normal tissue (T/N) ratio. The accuracy of (18)F-FLT and (18)F-FDG PET in evaluating newly diagnosed and recurrent gliomas was compared. More than half of the patients underwent resection after the PET study and correlations between PET uptake and the Ki-67 proliferation index were examined. Patients were monitored for a mean of 15.4 mo (range, 12-20 mo). The predictive power of PET for tumor progression and survival was analyzed using Kaplan-Meier statistics. RESULTS: (18)F-FLT uptake in tumors was rapid, peaking at 5-10 min after injection and remaining stable up to 75 min. Hence, a 30-min scan beginning at 5 min after injection was sufficient for imaging. (18)F-FLT visualized all high-grade (grade III or IV) tumors. Grade II tumor did not show appreciable (18)F-FLT uptake and neither did the stable lesions. The absolute uptake of (18)F-FLT was low (maximum-pixel SUV [SUV(max)], 1.33) but image contrast was better than with (18)F-FDG (T/N ratio, 3.85 vs. 1.49). (18)F-FDG PET studies were negative in 5 patients with recurrent high-grade glioma who subsequently suffered tumor progression within 1-3 mo. (18)F-FLT SUV(max) correlated more strongly with Ki-67 index (r = 0.84; P < 0.0001) than (18)F-FDG SUV(max) (r = 0.51; P = 0.07). (18)F-FLT uptake also had more significant predictive power with respect to tumor progression and survival (P = 0.0005 and P = 0.001, respectively). CONCLUSION: Thirty-minute (18)F-FLT PET 5 min after injection was more sensitive than (18)F-FDG to image recurrent high-grade tumors, correlated better with Ki-67 values, and was a more powerful predictor of tumor progression and survival. Thus, (18)F-FLT appears to be a promising tracer as a surrogate marker of proliferation in high-grade gliomas.