Differentiation of soft tissue and bone sarcomas from benign lesions utilizing 18F-FDG PET/CT-derived parameters
Bo Chen, Hongbo Feng, Jinghui Xie, Chun Li, Yu Zhang, Shaowu Wang
Abstract
Abstract Background Accurate differentiation between malignant and benign changes in soft tissue and bone lesions is essential for the prevention of unnecessary biopsies and surgical resection. Nevertheless, it remains a challenge and a standard diagnosis modality is urgently needed. The objective of this study was to evaluate the usefulness of 18 F-fluorodeoxyglucose ( 18 F-FDG) PET/CT-derived parameters to differentiate soft tissue sarcoma (STS) and bone sarcoma (BS) from benign lesions. Methods Patients who had undergone pre-treatment 18 F-FDG PET/CT imaging and subsequent pathological diagnoses to confirm malignant (STS and BS, n = 37) and benign ( n = 33) soft tissue and bone lesions were retrospectively reviewed. The tumor size, PET and low-dose CT visual characteristics, maximum standardized uptake value (SUVmax), metabolic tumor volume (MTV), total lesion glycolysis (TLG), and heterogeneous factor (HF) of each lesion were measured. Univariate and multivariate logistic regression analyses were conducted to determine the significant risk factors to distinguish sarcoma from benign lesions. To establish a regression model based on independent risk factors, and the receiver operating characteristic curves (ROCs) of individual parameters and their combination were plotted and compared. Conventional imaging scans were re-analyzed, and the diagnostic performance compared with the regression model. Results Univariate analysis results revealed that tumor size, SUVmax, MTV, TLG, and HF of 18 F-FDG PET/CT imaging in the STS and BS group were all higher than in the benign lesions group (all P values were < 0.01). The differences in the visual characteristics between the two groups were also all statistically significant ( P < 0.05). However, the multivariate regression model only included SUVmax and HF as independent risk factors, for which the odds ratios were 1.135 (95%CI: 1.026 ~ 1.256, P = 0.014) and 7.869 (95%CI: 2.119 ~ 29.230, P = 0.002), respectively. The regression model was constructed using the following expression: Logit ( P ) = − 2.461 + 0.127SUVmax + 2.063HF. The area under the ROC was 0.860, which was higher than SUVmax (0.744) and HF (0.790). The diagnostic performance of the regression model was superior to those of individual parameters and conventional imaging. Conclusion The regression model including SUVmax and HF based on 18 F-FDG PET/CT imaging may be useful for differentiating STS and BS from benign lesions.