Establishment and validation of nomogram model for the diagnosis of AFP-negative hepatocellular carcinoma.

Abstract

As one of the most common malignant tumors in clinical practice, hepatocellular carcinoma (HCC) is a major threat to human health, where alpha-fetoprotein (AFP) is widely used for early screening and diagnoses. However, the level of AFP would not elevate in about 30-40% of HCC patients, which is clinically referred to as AFP-negative HCC, with small tumors at an early stage and atypical imaging features, making it difficult to distinguish benign from malignant by imaging alone. A total of 798 patients, with the majority being HBV-positive, were enrolled in the study and were randomized 2:1 to the training and validation groups. Univariate and multivariate binary logistic regression analyses were used to determine the ability of each parameter to predict HCC. A nomogram model was constructed based on the independent predictors. A unordered multicategorical logistic regression analyses showed that the age, TBIL, ALT, ALB, PT, GGT and GPR help identify non-hepatic disease, hepatitis, cirrhosis, and hepatocellular carcinoma. A multivariate logistic regression analyses showed that the gender, age, TBIL, GAR, and GPR were independent predictors for the diagnosis of AFP-negative HCC. And an efficient and reliable nomogram model (AUC=0.837) was constructed based on independent predictors. Serum parameters help reveal intrinsic differences between non-hepatic disease, hepatitis, cirrhosis, and HCC. The nomogram based on clinical and serum parameters could be used as a marker for the diagnosis of AFP-negative HCC, providing an objective basis for the early diagnosis and individualized treatment of hepatocellular carcinoma patients.