18F-FDG PET/CT Predicts the Prognosis of Patients with Hepatocellular Carcinoma Undergoing Liver Transplantation

Introduction: In addition to radical resection, liver transplantation (LTx) is an effective treatment for hepatocellular carcinoma (HCC). However, tumor recurrence limits the efficacy of LTx in some patients. This study investigated the role of F-18-fludeoxyglucose (F-18-FDG) positron emission tomography/computed tomography (PET/CT) in predicting the prognosis of patients with HCC after LTx. Methods: A total of 278 consecutive patients with HCC who underwent pre-LTx PET/CT were divided into derivation (n = 178) and temporal validation (n = 100) cohorts and evaluated for PET/CT values, immunohistochemical (IHC) findings, and DNA sequencing of tumor tissues. Results: Patients with post-LTx recurrence exhibited significantly higher tumor maximum standardized uptake values (SUVmax) in pre-LTx PET/CT scans. Receiver operating characteristic curve analyses identified the tumor SUVmax to liver SUVmax ratio (T-SUVmax/L-SUVmax) as the strongest predictor of post-LTx recurrence, with an optimal cutoff value of 1.43. Kaplan-Meier analyses demonstrated that a T-SUVmax/L-SUVmax >1.43 was associated with a shorter time to recurrence (TTR) and overall survival (OS) in both cohorts (p < 0.001 for both). Multivariate Cox regression analyses confirmed that T-SUVmax/L-SUVmax >1.43 was an independent risk factor for tumor recurrence in both cohorts. IHC revealed that T-SUVmax/L-SUVmax >1.43 correlated with higher Ki-67 and CK19 expression. DNA sequencing indicated that tumors with T-SUVmax/L-SUVmax >1.43 had more mutations and a higher TMB. Furthermore, T-SUVmax/L-SUVmax >1.43 was significantly associated with mutations in TP53, EPPK1, MDM4, SLAMF7, SDHC, B4GALT3, RXRG, and FCGR family genes, as well as TP53 and PI3K signaling-related alterations. Conclusions: The preoperative T-SUVmax/L-SUVmax is a potential predictor of tumor recurrence in patients with HCC following LTx. Its use improves candidate selection and post-LTx management. (c) 2025 The Author(s).