多模态组学在晚期胃癌免疫微环境研究中的应用

Application of Multimodal Omics in Studying the Immune Microenvironment of Advanced Gastric Cancer

摘要: 胃癌是全球高发且高致死率的消化系统恶性肿瘤，晚期患者预后普遍不佳。随着免疫检查点抑制剂(ICIs)的临床应用，部分患者获益显著，但免疫微环境的高度异质性使疗效预测仍具挑战。近年来，多模态组学整合影像组学、基因组学、转录组学及蛋白质组学等多维度数据，为揭示胃癌免疫微环境的复杂性提供了新工具。CD8⁺ T细胞作为核心效应群体，其数量、空间分布和功能状态与免疫治疗响应密切相关。通过单细胞测序和空间转录组学，可解析其异质性及耗竭机制，为靶向干预提供依据。同时，PD-L1、MSI-H和TMB等分子标志物在免疫治疗预测中具有价值，但受限于检测一致性和时空异质性。人工智能与机器学习的应用，使影像与分子特征的融合分析成为可能，提高了预测模型的精度与可解释性。尽管仍存在标准化不足、样本量有限及临床转化障碍，多模态组学结合AI技术将在个体化免疫治疗中发挥重要作用，为改善晚期胃癌患者预后提供新方向。

Abstract: Gastric cancer is a common and highly lethal malignancy of the digestive system worldwide, with poor prognosis in advanced patients. With the clinical application of immune checkpoint inhibitors (ICIs), some patients have shown significant benefit, but the high heterogeneity of the immune microenvironment still makes efficacy prediction challenging. In recent years, multimodal omics integrating radiomics, genomics, transcriptomics, and proteomics has provided new tools to reveal the complexity of the gastric cancer immune microenvironment. CD8⁺ T cells, as key effector populations, are closely related to immunotherapy response through their quantity, spatial distribution, and functional status. Single-cell sequencing and spatial transcriptomics can resolve their heterogeneity and exhaustion mechanisms, providing a basis for targeted interventions. At the same time, molecular biomarkers such as PD-L1, MSI-H, and TMB have predictive value in immunotherapy, but are limited by inconsistency in detection and spatiotemporal heterogeneity. The application of artificial intelligence and machine learning makes it possible to integrate imaging and molecular features, improving the accuracy and interpretability of predictive models. Although limitations remain, such as lack of standardization, limited sample sizes, and barriers to clinical translation, multimodal omics combined with AI technology will play an important role in individualized immunotherapy and provide new directions for improving the prognosis of patients with advanced gastric cancer.