Background:
Evolution of individualized radiosurgical therapeutic methods for brain metastasis as an ominous prognostic finding may encourage a more extensive application of neuroimaging in patients with extracerebral cancer. The aim of the present study was to investigate the added value of brain-included 18 F FDG PET/CT acquisition protocol based on primary cancer type and clinical indication.
 Methods:
A retrospective review was performed on 3945 PET/CT reports of patients with extra-cerebral cancer underwent brain-included PET/CT study. Cerebral lesions suggestive of brain metastasis were subsequently verified by MRI, MRI+MRS, surgical pathology and a 1-year clinical formal follow up. The detection rate of new brain metastasis and related impact on disease status were then investigated in each cancer type based on clinical indication.
 Results:
Of a total 3933 eligible patients, 44 (1.12%) were finally verified to have new cerebral metastasis. The most common primary sources were lung cancer (19/385, 4.93%), cancer of unknown primary (CUP) (5/168, 2.97%) and breast cancer (8/468, 1.71%). The lowest detection rate of new cerebral metastasis was found in lymphoma (2/859 [0.23%]). The most common clinical indications were initial staging (17/44, 43.1%) and restaging (19/44, 36.4%). Change in disease status occurred in 12 out of 44 patients (27.3%), more frequently occurred in lung cancer (n=4), in all indications and breast (n=3) cancers at restaging (n=7, 43.8%).
 Conclusion:
PET/CT acquisition protocol study may be best optimized based on the type of primary cancer and timing of evaluation. Brain-included field of view may be recommended for lung cancer regardless the clinical indication, cancer of unknown primary and breast cancer at restaging. Skull-based protocol is best suited for lymphoma.