Establishment of a clinical model based on vessels encapsulating tumour clusters that could efficiently predict recurrence of patients with hepatocellular carcinoma after curative hepatectomy.

According to previous studies, vessels encapsulating tumour clusters (VETC) could promote metastasis of hepatocellular carcinoma (HCC) in a manner independent from epithelial-mesenchymal transition (EMT). However, the prognostic significance of VETC among patients undergoing curative hepatectomy has not been fully explored. This study was performed to assess the prognostic significance of VETC among patients with HCC undergoing curative hepatectomy. A total of 81 patients were included in this study. A predictive model based on VETC was established, then this model was compared with the American Joint Committee on Cancer, Tumor Node Metastasis (AJCC TNM) stage and Barcelona Clinic Liver Cancer (BCLC) system. It was revealed by multivariate Cox regression analysis that a high neutrophil-to-lymphocyte ratio (NLR) [p=0.013, hazard ratio (HR)=6.175, 95% confidence interval (CI) 1.468-25.977], number of tumours (p<0.001, HR=4.119, 95% CI 1.886-8.995) and VETC positivity (p=0.010, HR=2.440, 95% CI 1.235-4.821) were independent predictive factors for disease-free survival (DFS). Additionally, by Kaplan-Meier analysis, we revealed that VETC positivity was associated with worse DFS (p=0.018). The clinical predictive model combining the NLR, number of tumours, and VETC was compared with AJCC TNM stage and BCLC classification system by performing time-dependent receiver operating curve (td-ROC) analysis, revealing that the clinical predictive model was superior to AJCC TNM stage and BCLC system at different timepoints. Additionally, we demonstrated that the clinical model could well predict DFS by plotting calibration curves. VETC could be utilised as an efficient prognostic factor for HCC and the clinical predictive model combining the NLR, number of tumours, and VETC was superior to AJCC TNM stage and BCLC system in predicting cancer recurrence.