| Purpose:To develop a CT-based radiomics signature and assess its ability for predicting the early recurrence(≤1 year)and DFS of hepatocellular carcinoma(HCC).Methods:In the first part of the study,A total of 215 HCC patients underwent partial hepatectomy were enrolled in this retrospective study and all patients were followed up at least 1 year.Radiomics features were extracted from arterial-and portal venous phase CT images,and a radiomics signature was built by least absolute shrinkage and selection operator(LASSO)logistic regression model.Pre-and post-operative clincopathological factors associated with early recurrence were evaluated.A radiomics signature,pre-and post-clinical model,pre-and post-combined model were built,and the area under the curve(AUC)of operating characteristics(ROC)was used to explore their performance to discriminate early recurrence.While,In the second part of the study,A total of 234 patients underwent partial hepatectomy with HCC pathlogically confirmed were enrolled.All patients were followed up at least 6 moths.Radiomics features were still extracted from both arterial-and portal venous phase images as the first part of study,and a radiomics signature was also built by LASSO Cox regression model.Both pre-and post-operative clincopathological factors associated with DFS were evaluated.A clincopathological model and combined model were built.The C-index was used to assessed the Model discrimination.Results:In the first part of the study,twenty radiomics features were chosen from 300 candidate features to build a radiomics signature that was significantly associated with early recurrence(P<0.001)and presented good performance in the discrimination of early recurrence alone with an AUC of 0.817(95%CI:0.758-0.866),sensitivity of 0.794 and specificity of 0.699.The AUCs of the clinical and combined pre-operative model were 0.781(95%CI:0.719-0.834)and 0.836(95%CI:0.779-0.883),respectively,with the sensitivity being 0.784 and 0.824,and the specificity being 0.619 and 0.708.Adding a radiomics signature into conventional clinical variables can significantly improve the accuracy of the pre-operative model in predicting early recurrence(P=0.01).While,the AUCs of the clincopathological and combined model were 0.850(95%CI:0.795-0.895)and 0.893(95%CI:0.844-0.931),respectively,with the sensitivity being 0.775 and 0.823,and the specificity being 0.788 and 0.805.The radiomics signature can also significantly improve the performance of the clincopathological model in prediciton of early recurrence(P =0.043).In the second part of the study,six radiomics features were chosen to build a radiomics signature that was still significantly associated with DFS(HR=1.941;95%CI:1.215-3.101;P=0.005).C-index of clincopathological and combined model in estimation of DFS were 0.752,0.757,respectively,and C-index of the two model in estimation of 3-,6-,12-,36-,60-moths DFS were 0.827,0.799,0.783,0.742,0.724 and 0.828,0.803,0.790,0.747,0.729.Conclusions:The radiomics signature was a significant predictor for early recurrence in HCC.Incorporating radiomics signature into conventional clinical factors performed better for estimation of early recurrence than clincopathological variables alone.Although the the radiomics signature failed to improve the clincopathological variables in estimation of DFS,it was still a independent prognostic factor.Thus,the radiomics signature can help better stratify patients for surgery and enables clinicians to select optimal treatment strategies and an individualized monitoring protocol to improve clinical outcomes. |
PDF Full Text Request