肿瘤的细胞和分子异质性是癌症免疫治疗的主要障碍。在这里，我们使用系统生物学方法从肺癌转录组学中得出肿瘤微环境（TME）异质性主要来源的特征。我们证明，这个被我们称为 iHet 的特征在不同的癌症中是保守的，并且与抗肿瘤免疫相关。通过分析单细胞和空间转录组数据，我们追溯了解释 iHet 特征的变异性的细胞起源。最后，我们证明 iHet 对癌症免疫治疗具有预测价值，可以通过解开抗癌免疫反应的三个主要决定因素来进一步改进：免疫细胞的活性、免疫浸润或排斥以及癌细胞异质性。这项工作展示了如何整合转录组学数据来得出 TME 表型异质性的整体表示，并预测其在免疫检查点阻断剂免疫治疗期间的展开和命运。© 2024 作者。

The cellular and molecular heterogeneity of tumors is a major obstacle to cancer immunotherapy. Here, we use a systems biology approach to derive a signature of the main sources of heterogeneity in the tumor microenvironment (TME) from lung cancer transcriptomics. We demonstrate that this signature, which we called iHet, is conserved in different cancers and associated with antitumor immunity. Through analysis of single-cell and spatial transcriptomics data, we trace back the cellular origin of the variability explaining the iHet signature. Finally, we demonstrate that iHet has predictive value for cancer immunotherapy, which can be further improved by disentangling three major determinants of anticancer immune responses: activity of immune cells, immune infiltration or exclusion, and cancer-cell foreignness. This work shows how transcriptomics data can be integrated to derive a holistic representation of the phenotypic heterogeneity of the TME and to predict its unfolding and fate during immunotherapy with immune checkpoint blockers.© 2024 The Authors.