Send to

Choose Destination
J Cancer Res Ther. 2019;15(4):941-946. doi: 10.4103/jcrt.JCRT_52_19.

Establishment of a predictive model for short-term efficacy of transcatheter arterial chemoembolization treatment in hepatocellular carcinoma and its clinical application.

Author information

Department of Interventional Radiology, Beijing Friendship Hospital, Capital Medical University, Beijing, China.



This study aimed to establish a predictive model for the presence of transcatheter arterial chemoembolization (TACE) resistance in hepatocellular carcinoma (HCC) Barcelona Clinic Liver Cancer classification stage B or stage C4 (BCLC B/C) and further verify its accuracy, enabling clinicians to accurately predict the efficacy of TACE and propose individualized therapy to further optimize multidisciplinary team plans.

Materials and Methods:

A retrospective database review was performed, including 191 patients (39 females and 152 males; aged 50-76 years with a mean age of 55 ± 10 years) who received three consecutive TACE sessions for treating HCC (BCLC B/C) in 1 month apart, for a total of 3 months. After three TACE treatments, a total of 95 patients among the 191 patients showed TACE resistance, 112 cases were randomly selected to build the modeling group, and the remaining 79 cases formed a verification group. Some prognostic risk factors were obtained through clinical observation. Then, univariate and multivariate analyses were performed using the logistic proportional hazard regression model. Based on multivariate analysis results, a risk-index model was established and its effects on predicting the incidence of TACE resistance of those patients were evaluated.


Based on the results of the multivariate analysis, to show that were four-independent factors affecting a prognosis of patients with TACE resistance after three consecutive TACE treatment in 3 months, which were red blood cell (RBC), neutrophil count (NC), model for end-stage liver disease (MELD), and Apoprotein A1. The risk index model established according to the above factors was expressed as a predictive indice (PI), and PI = -3.79 + 4.916 × RBC - 1.547 × NC - 4.142 × MELD + 10.789 × ApoA1. The area under the receiver operating characteristic curves (AUROC) of PI of the modeling group was 0.986, which was significantly higher than that of each component index in the equation. The specificity of the modeling group was 86.3%, and the sensitivity was 70.4%, and 43 of 61 patients with PI ≤ 5.36 (70.5%) had a good outcome 3 months after consecutive TACE. From of the PI, among 51 patients with TACE resistance after consecutive TACE, 32 (62.7%) had a PI > 5.36, and only 19 patients were misidentified as having TACE resistance because of their PI > 5.36. The accuracy was 82.1%. The specificity of the validation group was 85.9%, and the sensitivity was 77.9%. The disease was under control in 29 of the 35 patients with PI ≤ 5.36 (82.9%) after consecutive TACE. According to PI, among the 44 patients with TACE resistence after consecutive TACE, 38 (86.4%) had PI > 5.36, and only 6 patients were misidentified as TACE resistance due to their PI > 5.36, with an accuracy of 87.3%, respectively.


According to the PI of this study, we investigated the risk factors and protective factors to estimate the presence of TACE resistance after three consecutive TACE treatment, so as it could help doctors to evaluate the patientet condition and choose more reasonable treatment methods.


Efficacy; hepatocellular carcinoma; the predictive index; transcatheter arterial chemoembolization resistance

Free full text

Supplemental Content

Full text links

Icon for Medknow Publications and Media Pvt Ltd
Loading ...
Support Center