1.The predictive value of a nomogram model based on aspartate aminotransferase-platelet ratio index for hepatocellular carcinoma recurrence after radiofrequency ablation
Yaxiang JI ; Jing XI ; Chunyan LIU ; Ping WU ; Xiaolan ZHANG ; Qian SONG
Journal of Interventional Radiology 2024;33(1):38-43
Objective To investigate the relationship between aspartate aminotransferase-platelet ratio index(APRI)and hepatocellular carcinoma(HCC)recurrence after radiofrequency ablation(RFA),and to construct a nomogram model for predicting the prognosis.Methods The clinical data of a total of 204 patients,whose initial diagnosis was HCC and received RFA at the Wujin Hospital Affiliated to Jiangsu University of China between January 2017 and December 2020,were retrospectively analyzed.The optimal cut-off value of APRI was determined using receiver operating characteristic(ROC)curve.Kaplan-Meier curves were plotted to estimate the recurrence-free survival(RFS)of high-APRI group patients and low-APRI group patients.The independent predictors of HCC recurrence after RFA were identified by using univariate and multivariate Cox regression analysis,and significant variables were selected to construct a nomogram model.The predictive ability of the nomogram model for HCC recurrence was evaluated by the consistency index(C-index)and calibration curves.Results The incidence of HCC recurrence after RFA was 57.4%(117/204),the optimal cut-off value of APRI for predicting HCC recurrence was 0.501,and the area under curve(AUC)value was 0.678(95%CI=0.603-0.752).High-APRI group(≥0.501)had 121 patients and low-APRI group(<0.501)had 83 patients.High APRI index was significantly correlated with low RFS(χ2=12.929,P<0.01).The univariate and multivariate Cox regression analysis revealed that the number of tumors(HR=1.541,95%CI=1.039-2.286,P=0.031),maximum tumor diameter(HR=1.461,95%CI=1.011-2.112,P=0.044),serum AFP level(HR=2.286,95%CI=1.576-3.318,P<0.01)and APRI index(HR=1.873,95%CI=1.257-2.790,P=0.002)were the independent risk factors for HCC recurrence.Based on the above four variables,a nomogram model for predicting HCC recurrence after RFA was constructed,the C-index was 0.769(95%CI=0.676-0.862),and the AUC values for 1-,2-,and 3-year RFS prediction were 0.707,0.719,and 0.707,respectively.The calibration curves showed that a good consistency existed between the predicted probability and actual probability.Conclusion The nomogram model based on APRI and tumor biological characteristics has an excellent predictive ability for HCC recurrence after RFA.(J Intervent Radiol,2024,32:38-43)