Texture analysis on gadoxetic acid enhanced-MRI for predicting Ki-67 status in hepatocellular carcinoma: A prospective study

Abstract

ObjectiveTo investigate the value of whole-lesion texture analysis on preoperative gadoxetic acid enhanced magnetic resonance imaging (MRI) for predicting tumor Ki-67 status after curative resection in patients with hepatocellular carcinoma (HCC).MethodsThis study consisted of 89 consecutive patients with surgically confirmed HCC. Texture features were extracted from multiparametric MRI based on whole-lesion regions of interest. The Ki-67 status was immunohistochemical determined and classified into low Ki-67 (labeling index ≤15%) and high Ki-67 (labeling index >15%) groups. Least absolute shrinkage and selection operator (LASSO) and multivariate logistic regression were applied for generating the texture signature, clinical nomogram and combined nomogram. The discrimination power, calibration and clinical usefulness of the three models were evaluated accordingly. Recurrence-free survival (RFS) rates after curative hepatectomy were also compared between groups.ResultsA total of 13 texture features were selected to construct a texture signature for predicting Ki-67 status in HCC patients (C-index: 0.878, 95% confidence interval: 0.791−0.937). After incorporating texture signature to the clinical nomogram which included significant clinical variates (AFP, BCLC-stage, capsule integrity, tumor margin, enhancing capsule), the combined nomogram showed higher discrimination ability (C-index: 0.936vs. 0.795, P<0.001), good calibration (P>0.05 in Hosmer-Lemeshow test) and higher clinical usefulness by decision curve analysis. RFS rate was significantly lower in the high Ki-67 group compared with the low Ki-67 group after curative surgery (63.27%vs. 85.00%, P<0.05). ConclusionsTexture analysis on gadoxetic acid enhanced MRI can serve as a noninvasive approach to preoperatively predict Ki-67 status of HCC after curative resection. The combination of texture signature and clinical factors demonstrated the potential to further improve the prediction performance.

Radiomic Analysis of Gd-EOB-DTPA-enhanced MRI&nbsp;Predicts Ki-67 Expression in Hepatocellular Carcinoma

“…Among these features, 13 were rst-order statistics, 28 were texture features including gray level co-occurrence matrix (GLCM), gray level dependence matrix (GLDM), gray level run length matrix (GLRLM), gray level size zone matrix (GLSZM), and neighboring gray tone dependence matrix (NGTDM). Although some scholars have recently published articles on the same topic of using a radiomics model based on Gd-EOB-DTPA-enhanced MRI to predict Ki-67 expression in HCC [16,17], there are many differences in details compared with our study. In the study of Li et al [16], a single slice with the largest proportion of lesion was delineated, and the predictive performance of models were compared only by misclassi cation rate.…”

“…In our study, all slices covering the whole tumor were delineated, and, the predictive performance of different models were compared by AUC values, calibration curve and DCA. In the study of Ye et al [17] , a sum of texture signatures derived from AP , PVP , pre-contrast T1W and T2W images was used to predictive Ki-67 expression by multivariate logistical regression, and predictive performance of radiomics model derived from different phases were not be compared. Although, in the study of Ye et al [17], the C-index (AUC) of the combined model (AUC = 0.936) was approximately equivalent to that in our study----the AUC value of combined model was 0.922 in the training cohort in our study, the study of Ye et al incorporated a sum of texture signatures derived from multiple phase into one radiomics model, which was cumbersome in clinical practice.…”

“…In the study of Ye et al [17] , a sum of texture signatures derived from AP , PVP , pre-contrast T1W and T2W images was used to predictive Ki-67 expression by multivariate logistical regression, and predictive performance of radiomics model derived from different phases were not be compared. Although, in the study of Ye et al [17], the C-index (AUC) of the combined model (AUC = 0.936) was approximately equivalent to that in our study----the AUC value of combined model was 0.922 in the training cohort in our study, the study of Ye et al incorporated a sum of texture signatures derived from multiple phase into one radiomics model, which was cumbersome in clinical practice. Our study developed and compared predictive performance of radiomics models derived from different sequences and phases, including T2W, AP, PVP, and HBP images, then further validated the optimal model for preoperative prediction of Ki-67 expression in HCC, which obtained a good result and would be feasible for clinical practice.…”

“…Gadolinium ethoxybenzyl-diethylenetriamine pentaacetic acid (Gd-EOB-DTPA), which has characteristics of both a blood-pool agent and a hepatobiliary agent, is commonly used in clinical practice. Previous studies have applied texture analysis on Gd-EOB-DTPA-enhanced MRI to preoperatively predict Ki-67 expression in patients with HCC and indicated that the texture analysis was superior to subjective MRI characteristics determined by radiologists and obtained a good result in predicting Ki-67 expression [16,17]. Although previous studies were valuable, they have not compared predictive performance of radiomics models derived from different sequences and phases based on Gd-EOB-DTPA-enhanced MRI.…”

Radiomics in liver diseases: Current progress and future opportunities

“…Recently, two MRI‐based studies investigated radiomic features for HCC aggressiveness characterization, demonstrating the potential of radiomics as indicative biomarkers for HCC grade 24,84 . Regarding Ki‐67 level, Ye et al reported that radiomics analysis can evaluate the tumour Ki‐67 level preoperatively with good accuracy (C‐index: 0.936) in a prospective study 85 …”