Diagnostic and prognostic values of 2-[F-18]FDG PET/CT in resectable thymic epithelial tumour

Objectives We aimed to evaluate the diagnostic ability for the prediction of histologic grades and prognostic values on recurrence and death of pretreatment 2-[F-18]FDG PET/CT in patients with resectable thymic epithelial tumours (TETs). Methods One hundred and fourteen patients with TETs who underwent pretreatment 2-[F-18]FDG PET/CT between 2012 and 2018 were retrospectively evaluated. TETs were classified into three histologic subtypes: low-risk thymoma (LRT, WHO classification A/AB/B1), high-risk thymoma (HRT, B2/B3), and thymic carcinoma (TC). Area under the receiver operating characteristics curve (AUC) was used to assess the diagnostic performance of PET/CT variables (maximum standardised uptake value [SUVmax], metabolic tumour volume [MTV], total lesion glycolysis [TLG], maximum diameter). Cox proportional hazards models were built using PET/CT and clinical variables. Results The tumours included 52 LRT, 33 HRT, and 29 TC. SUVmax showed good diagnostic ability for differentiating HRT/TC from LRT (AUC 0.84, 95% confidence interval [CI] 0.76 - 0.92) and excellent ability for differentiating TC from LRT/HRT (AUC 0.94, 95% CI 0.90 - 0.98), with significantly higher values than MTV, TLG, and maximum diameter. With an optimal cut-off value of 6.4, the sensitivity, specificity, and accuracy for differentiating TC from LRT/HRT were 69%, 96%, and 89%, respectively. In the multivariable Cox proportional hazards analyses for freedom-from-recurrence, SUVmax was an independent prognostic factor (p < 0.001), whereas MTV and TLG were not. SUVmax was a significant predictor for overall survival in conjunction with clinical stage and resection margin. Conclusion SUVmax showed excellent diagnostic performance for prediction of TC and significant prognostic value in terms of recurrence and survival.