Frontiers in Immunology (Aug 2024)
Establishment of a chemokine-based prognostic model and identification of CXCL10+ M1 macrophages as predictors of neoadjuvant therapy efficacy in colorectal cancer
Abstract
BackgroundAlthough neoadjuvant therapy has brought numerous benefits to patients, not all patients can benefit from it. Chemokines play a crucial role in the tumor microenvironment and are closely associated with the prognosis and treatment of colorectal cancer. Therefore, constructing a prognostic model based on chemokines will help risk stratification and providing a reference for the personalized treatment.MethodsEmploying LASSO-Cox predictive modeling, a chemokine-based prognostic model was formulated, harnessing the data from TCGA and GEO databases. Then, our exploration focused on the correlation between the chemokine signature and elements such as the immune landscape, somatic mutations, copy number variations, and drug sensitivity. CXCL10+M1 macrophages identified via scRNA-seq. Monocle2 showed cell pseudotime trajectories, CellChat characterized intercellular communication. CytoTRACE analyzed neoadjuvant therapy stemness, SCENIC detected cell type-specific regulation. Lastly, validation was performed through multiplex immunofluorescence experiments.ResultsA model based on 15 chemokines was constructed and validated. High-risk scores correlated with poorer prognosis and advanced TNM and clinical stages. Individuals presenting elevated risk scores demonstrated an increased propensity towards the development of chemotherapy resistance. Subsequent scRNA-seq data analysis indicated that patients with higher presence of CXCL10+ M1 macrophages in tumor tissues are more likely to benefit from neoadjuvant therapy.ConclusionWe developed a chemokine-based prognostic model by integrating both single-cell and bulk RNA-seq data. Furthermore, we revealed epithelial cell heterogeneity in neoadjuvant outcomes and identified CXCL10+ M1 macrophages as potential therapy response predictors. These findings could significantly contribute to risk stratification and serve as a key guide for the advancement of personalized therapeutic approaches.
Keywords
