The prognostic model based on tumor-associated neutrophils contributes to the stromal landscape and influences metabolic reprogramming in colorectal cancer

Background The tumor microenvironment (TME) is highly complex and significantly influences cancer prognosis and drug sensitivity. Tumor-associated neutrophils (TANs) play a key role in the TME. In this study, we aimed to investigate the TANs-related markers in colorectal cancer (CRC) and develop an integrated signature for prognostic stratification. Methods The CRC single-cell RNA sequencing (scRNA-seq) data and RNA-seq data were obtained from TCGA and GEO. A risk score was calculated based on the 18 TAN-associated genes identified in CRC by scRNA-seq data and LASSO regression. Prognosis, stromal and immune infiltration landscape, metabolism, and treatment response were then investigated in the low- and high-risk score clusters using RNA-seq data. Results Patients with a High-risk score had a significantly worse survival outcome than those with a Low-risk score (p < 0.0001). The prognostic predictive potency of the risk score was validated in both the TCGA validation cohort (p < 0.0001) and the GEO cohort (p < 0.00015). The areas under the curves of 1-, 3-, and 5-year survival were 0.76, 0.74, and 0.70 in the TCGA training set; 0.78, 0.68, and 0.78 in the TCGA validation set; and 0.65, 0.64, and 0.62 in the GEO set. The risk score was related to T, N, and M stages. A prognostic nomogram was constructed, and the predictive accuracy was assessed by calibration curve analysis. Decision curve analysis showed the clinical utility of the nomogram. Furthermore, the High-risk score cluster was significantly associated with the levels of cancer-associated fibroblasts, as well as activity in the transforming growth factor-β and WNT pathway. In depth, the High-risk score cluster exhibited lower levels of amino acid, tricarboxylic acid, and nucleotide metabolism, as well as poorer responses to chemotherapeutic agents such as 5-fluorouracil. Conclusion This novel TAN subtype, based on 18 prognostic-related genes, could provide new insights into the prognostic stratification and treatment options for CRC.