Establishment and validation of a predictive model for HBV-positive hepatocellular carcinoma / 中华检验医学杂志

ABSTRACT

Objective:The risk factors for hepatocellular carcinoma (HCC) in hepatitis B virus (HBV)-positive cirrhosis patients were screened based on commonly used laboratory indexes for the purpose of establishing a predictive model and the prediction efficacy of established model was validated in a validation patient cohort.Methods:The clinical data of 661 male patients with HBV-positive cirrhosis (cirrhosis group) and 694 male patients with HBV-positive HCC admitted to the First Hospital of China Medical University from 2010 to 2020 were retrospectively analyzed, age and complete blood count,liver function index (aspartate transaminase/alanine transaminase,glutamine transpeptidase,total protein, prealbumin, total bile acid, total bilirubin,direct bilirubin,cholinesterase), HBV markers, alpha-fetoprotein (AFP), fibrinogen,calcium were compared between the two groups. Multivariate Logistic regression was used to analyze the independent risk factors of HCC. The prediction model of high risk HCC ( P<0.05) was constructed and validated by receiver operating characteristic (ROC) curve and calibration curve. Results:There was significant difference in complete blood count, liver function index, HBV core antibody, HBV core antibody IgM, alpha-fetoprotein, fibrinogen, calcium between the two groups ( P<0.05). Multivariate analysis showed that calcium ( OR=35.770,95% CI 13.39-99.304),HBV core antibody ( OR=0.878,95% CI 0.816-0.944), AFP ( OR=1.002, 95% CI 1.001-1.003), fibrinogen ( OR=1.369, 95% CI 1.202-1.564) were the independent risk factors for HCC ( P<0.05), and were used for the nomogram. The AUC of the nomogram was 0.750 (95% CI 0.720-0.781) and the AUC of the validation group was 0.752 (95% CI 0.705-0.798). Conclusions:Based on calcium, hepatitis B virus core antibody, AFP, fibrinogen, a nomogram of the HCC is established and verified by ROC curve, which could be used to predict the risk of HBV-positive HCC.