Abstract
BACKGROUND: Bile duct invasion is a relatively rare event and is not well characterised in hepatocellular carcinoma (HCC). It remains very difficult to diagnose HCC with bile duct tumour thrombus (BDTT) before surgery. Increasing evidence has revealed that inflammation plays a critical role in tumorigenesis. This study aimed to develop nomograms based on systemic and hepatic inflammation markers to predict microscopic BDTT (micro-BDTT) before surgery in HCC. METHODS: A total of 723 HCC patients who underwent hepatectomy as initial therapy between January 2012 and June 2020 were included in the study. Logistic regression analysis was used to identify independent risk factors for micro-BDTT. The nomograms were constructed using significant predictors, including α-fetoprotein (AFP), alkaline phosphatase (ALP), direct bilirubin (DB), prognostic nutritional index (PNI), and γ-glutamyl transferase (γ-GT)/alanine aminotransferase (ALT). The prediction accuracies of the nomograms were evaluated using the area under the receiver operating characteristic (ROC) curve. RESULTS: AFP, ALP, DB, PNI, and γ-GT/ALT were independent risk factors for predicting micro-BDTT (P = 0.036, P = 0.004, P = 0.013, P = 0.012, and P = 0.006, respectively), which were assembled into the nomograms. The area under the ROC curve of the nomograms combining PNI and γ-GT/ALT for predicting micro-BDTT was 0.804 (95% confidence interval [CI]: 0.730-0.878). The sensitivity and specificity values when used in predicting micro-BDTT before surgery were 0.739 (95% CI: 0.612-0.866) and 0.781 (95% CI: 0.750-0.813), respectively. CONCLUSIONS: The nomogram based on combining systemic and hepatic inflammation markers is suitable for predicting micro-BDTT before surgery in HCC patients, leading to a rational therapeutic choice for HCC.