A prediction nomogram for hepatitis B virus-associated hepatocellular carcinoma

Background: The present study aimed to develop and validate a new nomogram for predicting the incidence of hepatocellular carcinoma (HCC) among chronic hepatitis B (CHB) patients receiving antiviral therapy from real-world data. Methods: The nomogram was established based on a real-world retrospective study of 764 patients with HBV from October 2008 to July 2020. A predictive model for the incidence of HCC was developed by multivariable Cox regression, and a nomogram was constructed. The predictive accuracy and discriminative ability of the nomogram were assessed by the concordance index (C-index), calibration curves, and decision curve analysis (DCA). Risk group stratification was performed to assess the predictive capacity of the nomogram. The nomogram was compared to three current commonly used predictive models. Results: A total of 764 patients with HBV were recruited for this study. Age, family history of HCC, alcohol consumption, and Aspartate aminotransferase-to-Platelet Ratio Index (APRI) were all independent risk predictors of HCC in CHB patients. The constructed nomogram had good discrimination with a C-index of 0.811. The calibration curve and DCA also proved the reliability and accuracy of the nomogram. Three risk groups (low, moderate, and high) with significantly different prognoses were identified (p < 0.001). The model's performance was significantly better than that of other risk models. Conclusions: The nomogram was superior in predicting HCC risk among CHB patients who received antiviral treatment. The model can be utilized in clinical practice to aid decision-making on the strategy of long-term HCC surveillance, especially for moderate- and high-risk patients.