Tipaldi Marcello Andrea, Ronconi Edoardo, Lucertini Elena, Krokidis Miltiadis, Zerunian Marta, Polidori Tiziano, Begini Paola, Marignani Massimo, Mazzuca Federica, Caruso Damiano, Rossi Michele, Laghi Andrea
Department of Surgical and Medical Sciences and Translational Medicine, Sapienza-University of Rome, 00189 Rome, Italy.
Department of Radiology, Sant'Andrea University of Hospital La Sapienza, 00189 Rome, Italy.
Diagnostics (Basel). 2021 May 26;11(6):956. doi: 10.3390/diagnostics11060956.
(1) Introduction and Aim: The aim of this study is to investigate the prognostic value, in terms of response and survival, of CT-based radiomics features for patients with HCC undergoing drug-eluting beads transarterial chemoembolization (DEB-TACE). (2) Materials and Methods: Pre-treatment CT examinations of 50 patients with HCC, treated with DEB-TACE were manually segmented to obtain the tumor volumetric region of interest, extracting radiomics features with TexRAD. Response to therapy evaluation was performed basing on post-procedural CT examination compared to pre-procedural CT, using modified RECIST criteria for HCC. The prognostic value of texture analysis was evaluated, investigating the correlation between radiomics features, response to therapy and overall survival. Three models based on texture and clinical variables and a combination of them were finally built; (3) Results: Entropy, skewness, MPP and kurtosis showed a significant correlation with complete response (CR) to TACE (all < 0.001). A predictive model to identify patients with a high and low probability of CR was evaluated with an ROC curve, with an AUC of 0.733 ( < 0.001). The three models built for survival prediction yielded an HR of 2.19 (95% CI: 2.03-2.35) using texture features, of 1.7 (95% CI: 1.54-1.9) using clinical data and of 4.61 (95% CI: 4.24-5.01) combining both radiomics and clinical data (all < 0.0001). (4) Conclusion: Texture analysis based on pre-treatment CT examination is associated with response to therapy and survival in patients with HCC undergoing DEB-TACE, especially if combined with clinical data.
(1) 引言与目的：本研究旨在探讨基于CT的影像组学特征对接受载药微球经动脉化疗栓塞术（DEB-TACE）的肝癌患者在反应和生存方面的预后价值。(2) 材料与方法：对50例接受DEB-TACE治疗的肝癌患者的治疗前CT检查进行手动分割，以获取肿瘤体积感兴趣区域，使用TexRAD提取影像组学特征。根据改良的肝癌RECIST标准，通过将术后CT检查与术前CT检查进行比较来评估治疗反应。评估纹理分析的预后价值，研究影像组学特征、治疗反应与总生存之间的相关性。最终构建了基于纹理和临床变量以及它们组合的三种模型；(3) 结果：熵、偏度、平均像素间距（MPP）和峰度与TACE的完全缓解（CR）显著相关（均P<0.001）。通过ROC曲线评估了一个用于识别CR高概率和低概率患者的预测模型，AUC为0.733（P<0.001）。为生存预测构建的三种模型，使用纹理特征时HR为2.19（95%CI：2.03 - 2.35），使用临床数据时为1.7（95%CI：1.54 - 1.9），结合影像组学和临床数据时为4.61（95%CI：4.24 - 5.01）（均P<0.0001）。(4) 结论：基于治疗前CT检查的纹理分析与接受DEB-TACE的肝癌患者的治疗反应和生存相关，特别是与临床数据相结合时。
