For the past decade, PET with (18)F-fluoro-ethyl-tyrosine ((18)F-FET) has been used in the evaluation of patients with primary brain tumors (PBTs), but so far series have reported only a limited number of patients. The purpose of this systematic review and metaanalysis was to assess the diagnostic performance of (18)F-FET PET in patients with suspicion of PBT. We examined studies published in the literature using MEDLINE and EMBASE databases. Inclusion criteria were use of (18)F-FET PET for initial assessment of patients with a newly diagnosed brain lesion; patients who had no radiotherapy, surgery, or chemotherapy before (18)F-FET PET; and use of histology as a gold standard. Metaanalysis was performed on a per-patient basis. We secondarily performed receiver-operating-characteristic analysis of pooled patients to determine tumor-to-background ratio (TBR) of (18)F-FET uptake and best diagnostic value. Thirteen studies totaling 462 patients were included. For the diagnosis of PBT, (18)F-FET PET demonstrated a pooled sensitivity of 0.82 (95% confidence interval [CI], 0.74-0.88), specificity of 0.76 (95% CI, 0.44-0.92), area under the curve of 0.84 (95% CI, 0.80-0.87), positive likelihood ratio of 3.4 (95% CI, 1.2-9.5), and negative likelihood ratio of 0.24 (95% CI, 0.14-0.39). Receiver-operating-characteristic analysis indicated that a mean TBR threshold of at least 1.6 and a maximum TBR of at least 2.1 had the best diagnostic value for differentiating PBTs from nontumoral lesions. (18)F-FET PET demonstrated excellent performance for diagnosing PBTs. Strict standardization of PET acquisition protocols and prospective, multicenter studies investigating the added value over current MRI are now needed to establish (18)F-FET PET as a highly relevant tool for patient management.
