Predictive Nomogram Model Of Portal Vein Tumor Thrombosis In Hepatocellular Carcinoma

Objective: The aim of this research is to ascertain the potential risks of portal vein tumor thrombosis(PVTT)in those with hepatocellular carcinoma(HCC),construct a clinical prognosis model,and create a nomogram.Methods: A retrospective investigation of 904 individuals suffering from hepatocellular carcinoma who were admitted to the Second Affiliated Hospital of Nanchang University between August 2019 and July 2022 was conducted.After the exclusion criteria were met,669 patients were eventually included in the study,206 of whom had portal vein tumor thrombosis.These patients were then randomly split into training cohorts(467)and validation cohorts(202).The training data set’s features are filtered by the least absolute shrinkage and selection operator(LASSO).Logistic regression is employed to construct a prediction model of the chosen features,and a nomogram is then drawn.Results: The c-index,receiver operating characteristic(ROC)curve,calibration chart,and decision curve were employed to assess the predictive model’s identification,calibration,and clinical efficacy.Key predictors of PVTT in HCC patients are Aspartate Aminotransferase(AST),Alkaline Phosphatase(ALP),Gamma-Glutamyl Transferase(GGT),Alpha-Fetoprotein(AFP)and Portal Hypertension.The training cohort’s c-index was 0.818,ranging from 0.777 to 0.859,while the validation cohort was 0.778,from 0.711 to 0.845.The training set’s area under the ROC curve(AUC)is 0.818,while the validation set’s AUC is 0.778.The calibration curve shows that the model is not over-fitted.The result of decision curve analysis(DCA)shows that the nomogram is effective in clinical practice.Conclusion: Based on clinical data,we established a prediction model of PVTT in HCC patients.