Establishment and Significance of a Nomogram Prediction Model for the Risk of Hepatocellular Carcinoma in Budd-Chiari...

This study aimed to identify independent risk factors for hepatocellular carcinoma (HCC) in patients initially diagnosed with Budd-Chiari syndrome (BCS) and develop a nomogram for HCC risk assessment in these patients.

Retrospective analysis was conducted on clinical data from 631 newly diagnosed BCS patients (BCS group) and 50 BCS patients complicated with HCC (HCC group) admitted to the Interventional Radiology Department of the Affiliated Hospital of Xuzhou Medical University (Xuzhou, China) between October 2014 and October 2021. General data, clinical symptoms/signs, laboratory tests, imaging features, Child-Pugh classification, and Model for End-Stage Liver Disease score were analyzed. Univariate logistic regression screened risk factors (P<0.05 for multivariate inclusion), and independent risk factors were selected via backward selection (Akaike Information Criterion) to build the nomogram, validated by bootstrap method. Receiver operating characteristic (ROC) curve, calibration curve, decision curve analysis (DCA), and clinical impact curve evaluated the model.

Independent risk factors in the final model were disease duration [odds ratio (OR)=1.19, 95% CI=1.11-1.28], portal vein diameter (OR=140.29, 95% CI=31.63-622.22), and intrahepatic nodule formation (OR=5.03, 95% CI=2.42-10.44). Bootstrap validation showed the model's ROC area under the curve (AUC)=0.862 (95% CI=0.798-0.926), with cross-validation AUC=0.858 (95% CI=0.663-1.000, good discrimination). Calibration curves (model and internal validation) aligned with ideal status. DCA showed the nomogram had higher net benefit than extreme curves at 2-83% threshold probability. Clinical impact curve indicated threshold probability >60% identified HCC high-risk groups consistent with actual HCC occurrence.

The independent risk factors for HCC in patients initially diagnosed with BCS were disease duration, portal vein diameter and intrahepatic nodule formation. The developed nomogram model exhibited good discrimination, accuracy and clinical applicability for the prediction of HCC risk in patients with BCS. This study, for the first time, established a nomogram for predicting the risk of HCC in patients with BCS in a single-center cohort in China, which can provide a new tool for early screening.