Jie Yamin, Wu Jianing, An Dongxue, Li Man, He Hongjiang, Wang Duo, Gu Anxin, E Mingyan
Department of Radiation Oncology, The Fourth Affiliated Hospital of Harbin Medical University, Harbin, China.
Department of Radiation Oncology, Harbin Medical University Cancer Hospital, Harbin, China.
Front Pharmacol. 2023 Jan 16;14:1099927. doi: 10.3389/fphar.2023.1099927. eCollection 2023.
Although the role of tumor microenvironment in lung adenocarcinoma (LUAD) has been explored in a number of studies, the value of TME-related signatures in immunotherapy has not been comprehensively characterized. Consensus clustering was conducted to characterize TME-based molecular subtypes using transcription data of LUAD samples. The biological pathways and immune microenvironment were assessed by CIBERSORT, ESTIMATE, and gene set enrichment analysis. A TME-related risk model was established through the algorithms of least absolute shrinkage and selection operator (Lasso) and stepwise Akaike information criterion (stepAIC). Four TME-based molecular subtypes including C1, C2, C3, and C4 were identified, and they showed distinct overall survival, genomic characteristics, DNA methylation pattern, immune microenvironment, and biological pathways. C1 had the worst prognosis and high tumor proliferation rate. C3 and C4 had higher enrichment of anti-tumor signatures compared to C1 and C2. C4 had evidently low enrichment of epithelial-mesenchymal transition (EMT) signature and tumor proliferation rate. C3 was predicted to be more sensitive to immunotherapy compared with other subtypes. C1 is more sensitive to chemotherapy drugs, including Docetaxel, Vinorelbine and Cisplatin, while C3 is more sensitive to Paclitaxel. A five-gene risk model was constructed, which showed a favorable performance in three independent datasets. Low-risk group showed a longer overall survival, more infiltrated immune cells, and higher response to immunotherapy than high-risk group. This study comprehensively characterized the molecular features of LUAD patients based on TME-related signatures, demonstrating the potential of TME-based signatures in exploring the mechanisms of LUAD development. The TME-related risk model was of clinical value to predict LUAD prognosis and guide immunotherapy.
尽管肿瘤微环境在肺腺癌（LUAD）中的作用已在多项研究中得到探索，但肿瘤微环境相关特征在免疫治疗中的价值尚未得到全面表征。利用LUAD样本的转录数据进行共识聚类，以表征基于肿瘤微环境的分子亚型。通过CIBERSORT、ESTIMATE和基因集富集分析评估生物学途径和免疫微环境。通过最小绝对收缩和选择算子（Lasso）算法和逐步赤池信息准则（stepAIC）建立了一个与肿瘤微环境相关的风险模型。确定了包括C1、C2、C3和C4在内的四种基于肿瘤微环境的分子亚型，它们表现出不同的总生存期、基因组特征、DNA甲基化模式、免疫微环境和生物学途径。C1的预后最差，肿瘤增殖率高。与C1和C2相比，C3和C4具有更高的抗肿瘤特征富集。C4的上皮-间质转化（EMT）特征和肿瘤增殖率明显较低。与其他亚型相比，C3预计对免疫治疗更敏感。C1对包括多西他赛、长春瑞滨和顺铂在内的化疗药物更敏感，而C3对紫杉醇更敏感。构建了一个五基因风险模型，该模型在三个独立数据集中表现良好。低风险组比高风险组显示出更长的总生存期、更多浸润的免疫细胞和对免疫治疗更高的反应。本研究基于肿瘤微环境相关特征全面表征了LUAD患者的分子特征，证明了基于肿瘤微环境的特征在探索LUAD发生机制方面的潜力。肿瘤微环境相关风险模型对预测LUAD预后和指导免疫治疗具有临床价值。
