A New Nomogram for Mortality Prediction Within 5 Years in Portal Vein Thrombosis due to Hepatitis B‐Related Cirrhosis

ABSTRACT

This study aims to develop a nomogram model intended for predicting long‐term mortality in patients suffering from hepatitis B‐related cirrhosis associated with Portal vein thrombosis (PVT). A total of 961 patients were recruited from January 2015 to December 2023, forming the training cohort. We analysed independent risk factors affecting 5‐year mortality through COX regression, from which we constructed a nomogram. To assess the nomogram's utility, we employed the receiver operating characteristic (ROC) curve, C‐index, calibration curves, and decision curve analysis (DCA). The independent prognostic factors retained in the final model for patients with hepatitis B‐related cirrhosis complicated by PVT included age (HR = 1.09, 95% CI: 1.06–1.11; p  < 0.001), leukocyte count (WBC) (HR = 1.09, 95% CI: 1.03–1.17; p  = 0.002), and the neutrophil‐to‐lymphocyte ratio (NLR) (HR = 1.29, 95% CI: 1.22–1.35; p  < 0.001), along with serum sodium (HR = 0.94, 95% CI: 0.90–0.98; p  = 0.005). The AUROC values for the nomogram were 0.906 (95% CI: 0.891–0.921) in the training cohort and 0.910 (95% CI: 0.885–0.927) in the validation cohort, demonstrating improved performance and discriminatory ability compared to current models, including CTP and MELD scores. Age, leukocyte count (WBC), NLR, and serum sodium emerged as independent risk factors impacting prognosis for patients with hepatitis B cirrhosis and PVT within 5 years. The nomogram developed in this research holds significant potential for accurately predicting the outcomes of these patients within 5 years.