Diagnostic performance of 18F-fluorodeoxyglucose PET/computed tomography in primary tumor localization and its impact on management in patients with metastases of unknown origin: a cross-sectional study

AIM: Metastases of unknown origin (MUO) remain a diagnostic challenge, with identification of the primary tumor closely tied to treatment planning and survival. This study aimed to evaluate the diagnostic performance of 18F-fluorodeoxyglucose PET/computed tomography (18F-FDG PET/CT) in detecting the primary tumor in MUO patients. METHOD: This cross-sectional study included 73 patients with MUO who underwent 18F-FDG PET/CT at the Department of Clinical Oncology and Nuclear Medicine, Assiut University Hospitals, between November 2023 and December 2025. PET/CT findings were correlated with histopathological or clinical-radiological follow-up to classify results as true/false positive/negative. RESULT: Median age was 64 years (37 males, 36 females). PET/CT suggested a primary tumor site in 76.7% (56/73) of patients. PET/CT correctly localized the primary in 50 patients (true positive), was falsely positive in 6, falsely negative in 2, and correctly excluded a detectable primary in 15 (true negative). Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 96.15, 71.4, 89.3, 88.2, and ~89%, respectively. Lung was the most common identified primary (20%), followed by lymphoma and ovarian cancer (10% each). PET/CT altered tumor, node, metastasis staging or diagnosis in 10 patients (13.7%): 3 (4.1%) were upstaged to M1b, 2 (2.7%) were downstaged to M0, and 5 (6.8%) were reclassified as nonmalignant based on PET/CT findings. CONCLUSION: 18F-FDG PET/CT is a high-yield, single-step imaging tool for evaluating MUO, showing a primary tumor detection rate and diagnostic accuracy favorable relative to previously published series, and significantly influencing clinical management strategies. Given the modest sample size and heterogeneous reference standard, these findings should be regarded as preliminary pending validation in larger, multicenter cohorts.