Table3_Risk model of hepatocellular carcinoma based on cuproptosis-related genes.DOCX

Background: Owing to the heterogeneity displayed by hepatocellular carcinoma (HCC) and the complexity of tumor microenvironment (TME), it is noted that the long-term effectiveness of the cancer therapy poses a severe clinical challenge. Hence, it is essential to categorize and alter the treatment intervention decisions for these tumors.
Materials and methods: “ConsensusClusterPlus” tool was used for developing a secure molecular classification system that was based on the cuproptosis-linked gene expression. Furthermore, all clinical properties, pathway characteristics, genomic changes, and immune characteristics of different cell types involved in the immune pathways were also assessed. Univariate Cox regression and the least absolute shrinkage and selection operator (Lasso) analyses were used for designing the prognostic risk model associated with cuproptosis.
Results: Three cuproptosis-linked subtypes (clust1, clust2, and clust3) were detected. Out of these, Clust3 showed the worst prognosis, followed by clust2, while Clust1 showed the best prognosis. Three subtypes had significantly different enrichment in pathways related to Tricarboxylic Acid (TCA) cycle, cell cycle, and cell senescence (p < 0.01). The clust3 subtype with poor prognosis had a low “ImmuneScore” and low immune cell infiltration, and the three subtypes had significant differences in the antigen processing and presentation pathway of the macrophages. Clust1 had a low TIDE score and was sensitive to immunotherapy. Then, according to the prognosis-related genes of cuproptosis, a prognosis risk model related to cuproptosis was constructed, containing seven genes (KIF2C, PTTG1, CENPM, CDC20, CYP2C9, SFN, and CFHR3). “High” group had a higher TIDE score compared to the TIDE score value shown by the “Low” group, which benefited less from immunotherapy, whereas the “High” group patients were more sensitive to the conventional drugs. Finally, the prognosis risk model related to cuproptosis was combined with clinical pathological characteristics to further improve the prognostic model and survival prediction.
Conclusion: Three new molecular subgroups based on cuproptosis-linked genes were revealed, and a cuproptosis-related prognostic risk model comprising seven genes was established in this study, which could assist in predicting the prognosis and identifying the patients benefit from immunotherapy.

背景：鉴于肝细胞癌（hepatocellular carcinoma, HCC）的异质性及肿瘤微环境（tumor microenvironment, TME）的复杂性，现有研究表明癌症治疗的长期有效性面临严峻的临床挑战。因此，对这类肿瘤进行分类并优化治疗干预决策至关重要。

材料与方法：本研究借助"ConsensusClusterPlus"工具，以铜死亡（cuproptosis）相关基因表达谱为基础构建可靠的分子分类体系。此外，本研究还对免疫通路中不同细胞类型的各项临床特征、通路特征、基因组改变及免疫特征进行了系统评估。通过单变量Cox回归分析与最小绝对收缩和选择算子（least absolute shrinkage and selection operator, Lasso）分析，构建了铜死亡相关的预后风险模型。

结果：本研究共鉴定出3个铜死亡相关亚型（clust1、clust2及clust3）。其中，clust3的预后最差，clust2次之，clust1的预后最佳。3个亚型在三羧酸循环（Tricarboxylic Acid, TCA）、细胞周期及细胞衰老相关通路的富集水平存在显著差异（p < 0.01）。预后较差的clust3亚型表现为较低的免疫评分（ImmuneScore）与免疫细胞浸润程度，且3个亚型在巨噬细胞抗原加工呈递通路中存在显著差异。clust1的TIDE评分较低，对免疫治疗更为敏感。随后，基于铜死亡预后相关基因，本研究构建了包含7个基因（KIF2C、PTTG1、CENPM、CDC20、CYP2C9、SFN及CFHR3）的铜死亡相关预后风险模型。"高风险组"的TIDE评分高于"低风险组"，其从免疫治疗中获益更少，但对常规化疗药物更为敏感。最后，将铜死亡相关预后风险模型与临床病理特征相结合，进一步优化了预后模型与生存预测效能。

结论：本研究揭示了3个基于铜死亡相关基因的新型分子亚型，并构建了包含7个基因的铜死亡相关预后风险模型，该模型可辅助预测患者预后并筛选可从免疫治疗中获益的人群。