The Immune Subtypes and Landscape of Sarcomas
Weng Weiwei,Yu Lin,Li Zhang,Tan Cong,Lv Jiaojie,Lao I. Weng,Hu Wenhuo,Deng Zhenzhong,Liu Zebing,Wang Jian,Xu Midie
DOI: https://doi.org/10.1186/s12865-022-00522-3
IF: 3
2022-01-01
BMC Immunology
Abstract:Background Considering the molecular heterogeneity of sarcomas and their immunologically quiet character, immunotherapy (e.g., immune checkpoint inhibitors) plays a viable role in only a subset of these tumors. This study aimed to determine the immune subtypes (IMSs) of sarcomas for selecting suitable patients from an extremely heterogeneous population. Results By performing consensus clustering analysis of the gene expression profiles of 538 patients with sarcomas in online databases, we stratified sarcomas into three IMSs characterized by different immune cell features, tumor mutational burdens (TMBs), gene mutations, and clinical outcomes. IMS1 showed an immune “hot” and immunosuppressive phenotype, the highest frequencies of CSMD3 mutation but the lowest frequencies of HMCN1 and LAMA2 mutations; these patients had the worst progression-free survival (PFS). IMS2 was defined by a high TMB and more gene mutations, but had the lowest frequency of MND1 mutations. IMS3 displayed the highest MDN1 expression level and an immune “cold” phenotype, these patients had the worst PFS. Each subtype was associated with different expression levels of immunogenic cell death modulators and immune checkpoints. Moreover, we applied graph learning-based dimensionality reduction to the immune landscape and identified significant intra-cluster heterogeneity within each IMS. Finally, we developed and validated an immune gene signature with good prognostic performance. Conclusions Our results provide a conceptual framework for understanding the immunological heterogeneity of sarcomas. The identification of immune-related subtypes may facilitate optimal selection of sarcoma patients who will respond to appropriate therapeutic strategies.