Comparative diagnostic performance of ¹⁸F-FLT and ¹⁸F-FDG PET for differentiating primary CNS lymphoma from...

This study aimed to compare the diagnostic performance of 3'-deoxy-3'-¹⁸F-fluorothymidine (FLT) positron emission tomography (PET) with that of ¹⁸F-fluorodeoxyglucose (FDG) PET for differentiating primary central nervous system lymphoma (PCNSL) from glioblastoma, isocitrate dehydrogenase (IDH)-wildtype, in patients who underwent both examinations before treatment.

We retrospectively evaluated 58 patients with histologically confirmed tumors: 40 with glioblastoma, IDH-wildtype, and 18 with PCNSL. Maximum standardized uptake values of the tumors were normalized to the mean standardized uptake values of the contralateral normal-appearing brain to calculate the tumor-to-normal brain (T/N) ratio. The paired areas under the receiver operating characteristic curves (AUCs) were compared.

FLT PET findings were positive for all tumors, whereas FDG PET findings were positive in 17 of 18 PCNSLs (94%) and 25 of 40 glioblastomas (63%). The mean FLT T/N ratio was 30.14 ± 15.02 for PCNSLs and 12.80 ± 7.37 for glioblastomas (P < 0.001), and the mean FDG T/N ratios were 3.01 ± 1.45 and 1.50 ± 1.22, respectively (P < 0.001). The AUC was 0.885 (95% confidence interval [CI], 0.794-0.976) for FLT and 0.826 (95% CI, 0.713-0.940) for FDG. Using cutoff values of 20.43 for FLT and 1.96 for FDG, the sensitivity was 77.8% and 77.8%, whereas specificity was 90.0% and 80.0%, respectively. The paired AUCs did not differ significantly (P = 0.380).

Both FLT and FDG PET differentiated PCNSL from glioblastoma, IDH-wildtype, with no significant difference between the paired AUCs. Minimal physiological FLT uptake in the normal brain enabled conspicuous tumor visualization. FLT PET may therefore serve as a useful complementary imaging approach in selected patients for whom FDG PET is less suitable.