Zhang Xiaoxiang, Ding Dongxiao, Wang Dianqian, Qin Yunsheng
Health Science Center, Ningbo University, Ningbo, 315800, Zhejiang, China.
Department of Thoracic Surgery, The People's Hospital of Beilun District, Ningbo, 315800, Zhejiang, China.
Discov Oncol. 2025 Aug 5;16(1):1476. doi: 10.1007/s12672-025-03257-w.
Currently, liver hepatocellular carcinoma (LIHC) is characterized by high morbidity, rapid progression and early metastasis. Although many efforts have been made to improve the prognosis of LIHC, the situation is still dismal. Inability to initiate the process of programmed cell death (PCD) is closely associated with cancer progression, thus influencing patients' prognosis. In this study, our purpose was to construct PCD-related prognostic signature for LIHC patients.
The list of PCD-related genes was obtained from GSEA gene sets. The gene set associated with survival time and survival status was screened by weighted correlation network analysis (WGCNA). Via Cox regression test and LASSO Cox regression model, prognostic signature was established and was then externally validated by ICGC-LIRI-JP dataset and GSE14520 dataset. The immune infiltration status and immune function of the signature were analyzed by ESTIMATE algorithm and ssGSEA algorithm. TIDE score, IPS and immune checkpoints expression and IC50 value were utilized to predict chemosensitivity and immunotherapy response. Moreover, GSE91061 dataset and PRJEB23709 dataset were enrolled to verify the predictive efficacy on immunotherapy response.
A total of 89 genes correlated with survival time and survival status were screened out from 1249 PCD-related genes. Next, the prognostic signature consisting of GLA, CLTA, CHGA, ERP29, MAPK3, CDK5, NLE1, STYXL1 and SFN was constructed. And high-risk patients were related to an adverse prognosis in TCGA-LIHC cohort and ICGC-LIRI-JP cohort. The prognostic signature also showed moderate to high predictive accuracy and was an independent prognostic indicator for LIHC. In general, low-risk patients exhibited higher StromalScore, immune cell infiltration levels, IPS, IPS-PD1 blocker, IPS-CTLA4 blocker, immune checkpoints expression and HLA-related genes expression while lower TIDE score, which indicated low-risk group tended to profit from ICI treatment. Furthermore, responders to ICI treatment had a lower riskscore in GSE91061 cohort, which showed similar result with ours.
Our study developed a novel prognostic signature comprising of 9 PCD-related genes, which could stratify the risk and effectively predict the prognosis and the immunotherapy response of LIHC patients.
目前，肝细胞肝癌（LIHC）具有高发病率、进展迅速和早期转移的特点。尽管已经做出了许多努力来改善LIHC的预后，但情况仍然不容乐观。无法启动程序性细胞死亡（PCD）过程与癌症进展密切相关，从而影响患者的预后。在本研究中，我们的目的是构建LIHC患者的PCD相关预后特征。
从GSEA基因集中获取PCD相关基因列表。通过加权基因共表达网络分析（WGCNA）筛选与生存时间和生存状态相关的基因集。通过Cox回归检验和LASSO Cox回归模型建立预后特征，然后通过ICGC-LIRI-JP数据集和GSE14520数据集进行外部验证。通过ESTIMATE算法和ssGSEA算法分析该特征的免疫浸润状态和免疫功能。利用TIDE评分、IPS以及免疫检查点表达和IC50值来预测化疗敏感性和免疫治疗反应。此外，纳入GSE91061数据集和PRJEB23709数据集以验证对免疫治疗反应的预测效力。
从1249个PCD相关基因中筛选出89个与生存时间和生存状态相关的基因。接下来，构建了由GLA、CLTA、CHGA、ERP29、MAPK3、CDK5、NLE1、STYXL1和SFN组成的预后特征。高危患者在TCGA-LIHC队列和ICGC-LIRI-JP队列中与不良预后相关。该预后特征还显示出中等至高的预测准确性，并且是LIHC的独立预后指标。总体而言，低风险患者表现出较高的基质评分、免疫细胞浸润水平、IPS、IPS-PD1阻断剂、IPS-CTLA4阻断剂、免疫检查点表达和HLA相关基因表达，而TIDE评分较低，这表明低风险组倾向于从ICI治疗中获益。此外，在GSE91061队列中，ICI治疗的反应者具有较低的风险评分，这与我们的结果相似。
我们的研究开发了一种由9个PCD相关基因组成的新型预后特征，可对风险进行分层，并有效预测LIHC患者的预后和免疫治疗反应。
