基于肝细胞癌经导管动脉栓塞化疗应答基因预测预后的基因标签评分研究

肝癌电子杂志 ›› 2025, Vol. 12 ›› Issue (3): 7-18.

基于肝细胞癌经导管动脉栓塞化疗应答基因预测预后的基因标签评分研究

闫东¹, 韩山山², 曹家玮¹, 许飞¹, 李槐^3,*

1.国家癌症中心/国家肿瘤临床医学研究中心/中国医学科学院肿瘤医院介入治疗科,北京 100021;
2.北京朝阳急诊抢救中心综合外科,北京 100021;
3.福建医科大学附属厦门弘爱医院介入治疗科,福建厦门 361006

收稿日期:2022-08-24 出版日期:2025-09-30 发布日期:2025-11-03
通讯作者: ^*李槐,E-mail：lihuai1956@hotmail.com
基金资助:
临床与转化医学基金（2021-I2M-C&T-067）

Study on gene signatureabel score for predicting prognosis based on TACE response genes in hepatocellular carcinoma

Yan Dong¹, Han Shanshan², Cao Jiawei¹, Xu Fei¹, Li Huai^3,*

1. Interventional Therapy Department, National Cancer Center/National Cancer Clinical Medicine Research Center/Cancer Hospital of Chinese Academy of Medical Sciences, Beijing 100021, China;
2. General Surgery, Beijing Chaoyang Emergency Rescue Center, Beijing 100021, China;
3. Department of Interventional Therapy, Fujian Medical University Xiamen Humanity Hospital, Xiamen 361006, Fujian, China

Received:2022-08-24 Online:2025-09-30 Published:2025-11-03
Contact: ^* Li Huai, E-mail: lihuai1956@hotmail.com

摘要/Abstract

摘要： 目的: 预测肝细胞癌（HCC）经导管动脉栓塞化疗（TACE）应答是临床必要需求。本研究基于TACE应答基因开发预后基因标签评分来预测TACE应答及预后。
方法: 在GSE104580数据集中使用差异表达分析和加权基因共表达网络分析（WGCNA）识别TACE应答基因;在GSE14520数据集中使用最小绝对收缩选择算子（LASSO）-Cox风险回归模型分析预后相关应答基因并构建基因标签评分,并在癌症基因组图谱数据库和人类蛋白质图谱中验证。使用CIBERSORT算法分析TACE应答与无应答的免疫浸润丰度;同时分析与36种免疫检查点基因表达关系。
结果: 在GSE104580数据集中共识别276个差异表达基因（校正P均< 0.05）;在WGCNA中,识别出模块7（基因数=846）和模块8（基因数=127）与TACE应答相关。LASSO-Cox风险回归模型发现CTSO、CLGN、RTP4基因与患者预后独立相关（均P<0.05）。基因标签评分预测患者1、3、5年死亡率的曲线下面积（AUC）分别为0.812（0.748~0.965）、0.785（0.687~0.845）、0.755（0.697~0.838）。多发肿瘤、TNM分期、基因标签评分与患者预后独立相关（均P<0.05）。列线图预测患者1、3、5年总生存率的AUC分别为0.729（0.455~0.915）、0.753（0.651~0.915）、0.727（0.616~0.821）。TACE应答和无应答患者间存在多种免疫细胞丰度差异（P<0.05）。CTSO、CLGN和RTP4基因与多种免疫细胞间和免疫检查点基因表达存在相关性（均P<0.05）。
结论: 基于CTSO、CLGN、RTP4基因标签评分能预测HCC患者TACE应答和预后。基因标签评分结合临床病理参数构建的列线图有助于研究结果的临床转化。

关键词: 肝细胞癌, 经导管动脉栓塞化疗, 基因标签评分, 列线图, 预测模型

Abstract: Objective: Predicting the response to transarterial chemoembolization (TACE) in hepatocellular carcinoma (HCC) is a clinical necessity. This study developed a prognostic gene signature score based on TACE-responsive genes to predict TACE response and prognosis.
Methods: In the GSE104580 dataset, differential expression analysis and weighted gene co‐expression network analysis (WGCNA) were used to identify TACE-responsive genes. In the GSE14520 dataset, Least Absolute Shrinkage Selection Operator (LASSO) - Cox regression model was used to analyze prognosis-related response genes and construct gene signature scores, and then validated in the The Cancer Genome Atlas database and the Human Protein Atlas. The CIBERSORT algorithm was used to analyze the abundance of immune infiltration between TACE responders and non-responders; meanwhile, the relationship with the gene expression of 36 immune checkpoints was analyzed. The CIBERSORT algorithm was used to analyze the immune infiltration of TACE response and non-response; meanwhile, the relationship with the gene expression of 36 immune checkpoints was analyzed.
Results: A total of 276 differentially expressed genes were identified in the GSE104580 dataset (all adj.P<0.05); in WGCNA, module 7 (number of genes = 846) and module 8 (number of genes = 127) were identified to be associated with TACE responses. LASSO-Cox regression model found that CTSO, CLGN and RTP4 genes were independently correlated with patient prognosis (all P<0.05). The area under the curve (AUC) of the gene signature score for predicting 1-, 3-, and 5-years death rates was 0.812(0.748-0.965)、0.785(0.687-0.845), and 0.755(0.697-0.838), respectively. Multiple tumors, TNM stage and gene signature score were independently correlated with the prognosis of patients (all P<0.05). The AUC of Nomogram for predicting 1-, 3-, and 5-year overall survival rates were 0.729 (0.455-0.915), 0.753 (0.651-0.915), and 0.727 (0.616-0.821), respectively. There were differences in the abundance of various immune cells between TACE-responsive and no-responsive patients (P<0.05). CTSO, CLGN and RTP4 genes were correlated with the expression of various immune cells and immune checkpoints (all P<0.05).
Conclusion: Gene signature scores based on CTSO, CLGN and RTP4 genes can predict TACE response and prognosis in HCC patients. The Nomogram constructed by the combination of gene signature scores and clinicopathological parameters is helpful for the clinical translation of research results.

Key words: Hepatocellular carcinoma, Transarterial chemoembolization, Gene signature score, Nomogram, Predictive model

闫东, 韩山山, 曹家玮, 许飞, 李槐. 基于肝细胞癌经导管动脉栓塞化疗应答基因预测预后的基因标签评分研究[J]. 肝癌电子杂志, 2025, 12(3): 7-18.

Yan Dong, Han Shanshan, Cao Jiawei, Xu Fei, Li Huai. Study on gene signatureabel score for predicting prognosis based on TACE response genes in hepatocellular carcinoma[J]. Electronic Journal of Liver Tumor, 2025, 12(3): 7-18.

参考文献 19

[1]	KONYN P, AHMED A, KIM D.Current epidemiology in hepatocellular carcinoma[J] . Expert Rev Gastroenterol Hepatol, 2021, 15(11): 1295-1307.
[2]	RAOUL J L, GILABERT M, PIANA G.How to define transarterial chemoembolization failure or refractoriness: a European perspective[J] . Liver Cancer, 2014, 3(2):119-124.
[3]	FORNER A, REIG M, BRUIX J.Hepatocellular carcinoma[J] . Lancet, 2018, 391(10127): 1301-1314.
[4]	CHANG Y, JEONG S W, JANG J Y, et al.Recent updates of transarterial chemoembolilzation in hepatocellular carcinoma[J] . Int J Mol Sci, 2020, 21(21): 8165.
[5]	LI C, WEN T F, YAN L N, LU W S, et al.Liver resection versus liver resection plus TACE for patients with hepatocellular carcinoma beyond Milan criteria[J] . J Surg Res, 2017, 209: 8-16.
[6]	POON R T, FAN S T, TSANG F H, et al.Locoregional therapies for hepatocellular carcinoma: a critical review from the surgeon's perspective[J] . Ann Surg, 2002, 235(4): 466-486.
[7]	ZHANG B, TANG B, GAO J, et al.A hypoxia-related signature for clinically predicting diagnosis, prognosis and immune microenvironment of hepatocellular carcinoma patients[J] . J Transl Med, 2020, 18(1): 342.
[8]	LIU G M, ZENG H D, ZHANG C Y, et al.Identification of a six-gene signature predicting overall survival for hepatocellular carcinoma[J] . Cancer Cell Int, 2019, 19: 138.
[9]	LONG J, ZHANG L, WAN X, et al.A four-gene-based prognostic model predicts overall survival in patients with hepatocellular carcinoma[J] . J Cell Mol Med, 2018, 22(12): 5928-5938.
[10]	LANGFELDER P, HORVATH S.WGCNA: an R package for weighted correlation network analysis[J] . BMC Bioinformatics, 2008, 9: 559.
[11]	UHLÉN M, FAGERBERG L, HALLSTRÖM B M, et al. Tissue-based map of the human proteome[J] . Science, 2015, 347(6220): 1260419.
[12]	LIANG J Y, WANG D S, LIN H C, et al.A novel ferroptosis-related gene signature for overall survival prediction in patients with hepatocellular carcinoma[J] . Int J Biol Sci, 2020, 16(13): 2430-2441.
[13]	SHI J Y, WANG X, DING G Y, et al.Exploring prognostic indicators in the pathological images of hepatocellular carcinoma based on deep learning[J] . Gut, 2021, 70(5): 951-961.
[14]	FENG J, DAI W, MAO Y, et al.Simvastatin re-sensitizes hepatocellular carcinoma cells to sorafenib by inhibiting HIF-1α/PPAR-γ/PKM2-mediated glycolysis[J] . J Exp Clin Cancer Res, 2020, 39(1): 24.
[15]	MA H Y, YAMAMOTO G, XU J, et al.IL-17 signaling in steatotic hepatocytes and macrophages promotes hepatocellular carcinoma in alcohol-related liver disease[J] . J Hepatol, 2020, 72(5): 946-959.
[16]	MCELIGOT A J, POYNOR V, SHARMA R, et al.Logistic LASSO regression for dietary intakes and breast cancer[J] . Nutrients, 2020, 12(9): 2652.
[17]	LAURIN M, HUBER J, PELLETIER A, et al.Rac-specific guanine nucleotide exchange factor DOCK1 is a critical regulator of HER2-mediated breast cancer metastasis[J] . Proc Natl Acad Sci U S A, 2013, 110(18): 7434-7439.
[18]	XU N, WU Y P, KE Z B, et al.Identification of key DNA methylation-driven genes in prostate adenocarcinoma: an integrative analysis of TCGA methylation data[J] . J Transl Med, 2019, 17(1): 311.
[19]	LLOVET J M, MONTAL R, SIA D, et al.Molecular therapies and precision medicine for hepatocellular carcinoma[J] . Nat Rev Clin Oncol, 2018, 15(10): 599-616.