The clinical value of F-18 FDG PET/CT in differentiating malignant from benign lesions in pneumoconiosis patients

Objectives We reviewed PET/CT findings of pneumoconiosis and determined the ability of PET/CT to differentiate lung cancer from progressive massive fibrosis (PMF), and metastatic lymph nodes (LNs) from underlying reactive LN hyperplasia. Methods This was a retrospective study of patients with pneumoconiosis and suspected lung cancer. Maximum standardized uptake value (SUVmax), long- and short-axis diameters (D-L and D-S), ratio of D-L to D-S (D-L/S), and Hounsfield unit (HU) from the lung mass and mediastinal LNs were measured. The cutoff values of each parameter were obtained by ROC analysis, and we evaluated the diagnostic sensitivity. Results Forty-nine pneumoconiosis patients were included. Eighty-three lung masses were detected, of which 42 were confirmed as lung cancer (23 squamous cell carcinomas, 12 adenocarcinomas, and 7 small cell carcinomas) and 41 were PMF. There were significant differences between lung cancer and PMF in terms of SUVmax, D-S, D-L/S, and HU (all p < 0.05). The sensitivity, specificity, and accuracy for diagnosis of lung cancer were 81.0%, 73.2%, and 77.1%, respectively, with an SUVmax cutoff value of 7.4; and 92.8%, 87.8%, and 90.4%, respectively, with a HU cutoff value of 45.5. Among the 40 LNs with available pathological results, 7 were metastatic. Metastatic LNs showed higher SUVmax, larger D-S, and lower HU than benign lesions (all p < 0.05). The sensitivity, specificity, and accuracy for predicting metastatic LNs by PET/CT were 85.7%, 93.9%, and 92.5%, respectively. Conclusion By applying PET and CT parameters in combination, the accuracy for differentiating malignant from benign lesions could be increased. PET/CT can play a central role in the discrimination of lung cancer and PMF.