assessed by receiver-operating-characteristic curve analyses using histology or clinical course as a reference. Results: In patients with newly diagnosed cerebral lesions, the highest accuracy (77%) to detect neoplastic tissue (19/26 patients) was obtained when the maximum TBR was 1.7 or greater (area under the curve, 0.80 ± 0.09; sensitivity, 79%; specificity, 71%; positive predictive value, 88%; P 5 0.02). For diagnosing tumor progression or recurrence, the highest accuracy (82%) was obtained when curve patterns 2 or 3 were present (area under the curve, 0.80 ± 0.11; sensitivity, 75%; specificity, 90%; positive predictive value, 90%; P 5 0.02). During chemotherapy, ad ecrease of TBRs was associated with a stable clinical course, and in 2 patients PET detected residual tumor after presumably complete tumor resection. Conclusion: Our findings suggest that 18 F-FET PET

BACKGROUND: Experience regarding the use of dynamic O-(2-[18F]-fluoroethyl)-L-tyrosine (18F-FET) PET in children and adolescents with brain tumors is limited. METHODS: Sixty-nine 18F-FET PET scans of 49 patients (median age, 13 years; range, 1-18 years) were analyzed retrospectively. Patients had been referred for: (A) assessment of newly diagnosed cerebral lesions (26 scans in 26 patients), (B) diagnosing tumor progression/recurrence (24 scans in 18 patients), (C) monitoring of chemotherapy effects (8 scans in 4 patients), and (D) the detection of residual tumor tissue after resection (11 scans in 10 patients). Maximum and mean tumor/brain ratios (TBRmax/mean) of 18F-FET uptake were determined (20-40 min p.i.) and time-activity curves were generated and assigned to one of the following patterns: (1) constantly increasing uptake, (2) uptake peaking at a midway point (>20-40 min) followed by a plateau, and (3) uptake peaking early (≤20 min) followed by a constant descent. The diagnostic values of TBRs and kinetic parameters to detect neoplastic tissue or diagnose tumor progression/recurrence were assessed using ROC analyses. Diagnoses were confirmed histologically and/or by clinical course. RESULTS: In patients with newly diagnosed cerebral lesions, highest accuracy (77%) to detect neoplastic tissue (7 of 26 patients) was obtained when TBRmax was >1.7 (AUC, 0.80 ± 0.09; sensitivity, 79%; specificity, 71%, PPV, 88%; P = 0.02). For diagnosing tumor progression/recurrence, highest accuracy (82%) was obtained when curve patterns 2 or 3 were present (AUC, 0.80 ± 0.11; sensitivity, 75%; specificity, 90%, PPV, 90%; P = 0.02). During chemotherapy, a decrease of TBRs was associated with a stable clinical course at least for 6 months. In patients after complete tumor resection (2 of 10 patients), 18F-FET PET detected metabolically active tumor (TBRmax ≥ 1.7). CONCLUSIONS: Our findings suggest that 18F-FET PET can add valuable information for clinical decision-making in pediatric brain tumor patients.