Background The clinical value of alpha-fetoprotein (AFP) in patients with AFP-negative (< 20 ng/ml) hepatocellular carcinoma (HCC) who underwent curative resection remained controversial. Aims To investigate clinical relevance and prognostic effect of preoperative serum AFP level in this subgroup. Methods A total of 1879 patients with AFP-negative HCC who underwent curative resection were included in the study. Overall survival (OS) and disease-free survival (DFS) rate were displayed by Kaplan–Meier method and compared by log-rank test. Multivariate cox proportional hazard regression analysis was used to identify the independent prognostic factors. The prognostic predictive performance was analyzed by time-dependent areas under receiver operating characteristic curve (AUC). Results Even in AFP-negative HCC, patients with high preoperative serum AFP level tended to have multiple tumor (P < 0.001), poorer differentiation of tumor cell (P < 0.001), presence of satellite nodules (P < 0.001), and MVI (P = 0.002). Kaplan–Meier analysis showed the adverse impact of AFP level on prognosis, especially for DFS. Multivariate analysis identified AFP as the independent unfavorable factor for OS and DFS (P < 0.001 for both). Time-dependent AUC analysis showed that the combination with AFP could improve the prognostic predictive performance of 8th AJCC and BCLC staging system. Conclusions AFP was still the surrogate of aggressive behavior of HCC and independent prognostic factor for patients with AFP-negative HCC underwent curative resection. Even combining with such a low level of AFP could significantly improve the predictive performance of conventional staging system.

Objectives: The purpose of this study was to develop and validate a radiomics nomogram for predicting thymidylate synthase (TYMS) status in hepatocellular carcinoma (HCC) by using Gd-DTPA contrast enhanced MRI. Methods: We retrospectively enrolled 147 consecutive patients with surgically confirmed HCC and randomly allocated to training and validation set (7:3). The TYMS status was immunohistochemical determined and classified into low TYMS (positive cells ≤ 25%) and high TYMS (positive cells > 25%) groups. Radiomics features were extracted from the arterial phases and portal venous phase of Gd-DTPA contrast enhanced MRI. Least absolute shrinkage and selection operator (LASSO) were applied for generating the Rad score. Clinical data and MRI findings were assessed to build a clinical model. Rad score combined with clinical features was used to construct radiomics nomogram. Results: A total of 2260 features were extracted and reduced to 7 features as the most important discriminators to build the Rad score. InAFP was identified as the only independent clinical factors for TYMS status. The radiomics nomogram showed good discrimination in training (AUC, 0.759; 95% CI 0.665–0.838) and validation set (AUC, 0.739; 95% CI 0.585–0.860), and showed better discrimination capability (P < 0.05) compared with clinical model in training (AUC, 0.656; 95% CI 0.555–0.746) and validation set (AUC, 0.622; 95% CI 0.463–0.764). Conclusions: The radiomics nomogram shows favorable predictive efficacy for TYMS status in HCC, which might be helpful for the personalized treatment of HCC.