Evaluation of a diagnostic 18F-FDG PET/CT strategy for differentiating benign from malignant retroperitoneal soft-tissue masses.

ABSTRACT

AIM: To investigate the optimal combined 2-[18F]-fluoro-2-deoxy-d-glucose (FDG) positron-emission tomography (PET)/computed tomography (CT) diagnostic criteria for distinguishing between benign and malignant retroperitoneal soft-tissue masses (RPMs). MATERIALS AND METHODS: A total of 74 patients (MF=3440; age, 53±13.2 years) who underwent FDG PET/CT for the initial work-up of RPMs were included. The maximum standardised uptake value (SUVmax), tumour size, presence of fat or calcifications and separated hypermetabolic lesions were included as PET/CT diagnostic parameters. Receiver-operating characteristic (ROC) curves were used to compare the diagnostic performance. RESULTS: The final pathological diagnoses included 52 malignant and 22 benign tumours. High SUVmax (>4.8) and large size (>13 cm) favoured malignancy, and yielded a diagnostic accuracy and AUC of 64.9%, 0.820±0.059, and 68.9%, 0.738±0.061, respectively. In a subgroup of RPMs with a fat component, both SUVmax and size were significantly different between benign and malignant RPM, which yielded a diagnostic accuracy and AUC of 91%, 0.977±0.024 (cut-off, 1.9 cm) and 87.9%, 0.865±0.072 (cut-off, 13 cm), respectively. In a subgroup without a fat component, only SUVmax was significantly different with an accuracy of 90.2% and AUC of 0.919±0.043. The optimal diagnostic flow by combining SUVmax and tumour size after dividing patients into two groups according to the presence of fat showed a sensitivity of 90.4%, a specificity of 95.5%, and an accuracy of 91.9%. CONCLUSIONS: The combination of SUVmax and size according to the presence of a fat component may be the optimal PET/CT diagnostic criteria for distinguishing benign and malignant RPMs.