A Novel TGF-β-Related Signature for Predicting Prognosis, Tumor Microenvironment, and Therapeutic Response in Colorectal Cancer
Baorui Tao,Chenhe Yi,Yue Ma,Yitong Li,Bo Zhang,Yan Geng,Zhenmei Chen,Xiaochen Ma,Jinhong Chen
DOI: https://doi.org/10.1007/s10528-023-10591-7
Abstract:The transforming growth factor beta (TGF-β) signaling plays a critical role in immune evasion and tumor progression. However, its modulatory influences on prognosis, tumor microenvironment (TME), and therapeutic efficacy remain unknown in colorectal cancer (CRC). We summarized TGF-β-related genes and comprehensively estimated their expression pattern in 2142 CRC samples from 9 datasets. Two distinct cluster patterns were divided and biological characteristics of each pattern were further analyzed. Then, to quantify the TGF-β cluster pattern of individual CRC patient, we generated the TGF-β score (TGFBscore) model based on TGF-β cluster pattern-relevant differentially expressed genes (DEGs). Subsequently, we conducted correlation analysis for TGFBscore and clinical prognosis, consensus molecular subtypes (CMSs), TME characteristics, liver metastasis, drug response, and immunotherapeutic efficacy in CRC. We illustrated transcriptional and genetic alterations of TGF-β-relevant genes, which were closely linked with carcinogenic pathways. We identified two different TGF-β cluster patterns, characterized by a high and a low TGFBscore. The TGFBscore-high group was significantly linked with worse patient survival, epithelial-mesenchymal transition (EMT) activation, liver metastasis tendency, and the infiltration of immunosuppressive cells (regulatory T cells [Tregs], M2 macrophages, cancer-associated fibroblasts [CAFs], and myeloid-derived suppressor cells [MDSCs]), while the TGFBscore-low group was linked with a survival advantage, epithelial phenotype, early CRC staging, and the infiltration of immune-activated cells (B cell, CD4 T cell, natural killer T [NKT] cell, and T helper 1 [Th1] cell). In terms of predicting drug response, TGFBscore negatively correlated (sensitive to TGFBscore-high group) with drugs targeting PI3K/mTOR, JNK and p38, RTK signaling pathways, and positively correlated (sensitive to TGFBscore-low group) with drugs targeting EGFR signaling pathway. Also, TGFBscore could predict the efficacy of different anti-tumor therapies. TGFBscore-low patients might benefit more from anti-PDL1 immunotherapy, adjuvant chemotherapy (ACT), and ERBB targeted therapy, whereas TGFBscore-high patients might benefit more from antiangiogenic targeted therapy. Our study constructed a novel TGF-β scoring model that could predict prognosis, liver metastasis tendency, and TME characteristics for CRC patients. More importantly, this work emphasizes the potential clinical utility of TGFBscore in evaluating the efficacy of chemotherapy, targeted therapy, and immunotherapy, guiding individualized precision treatment in CRC.