Cancer Cell International (Jul 2021)
Prognostic value of tumor immune cell infiltration patterns in colon adenocarcinoma based on systematic bioinformatics analysis
Abstract
Abstract Background Although immunotherapy for colon cancer has made promising progress, only a few patients currently benefit from it. A recent study revealed that infiltrating immune cells are highly relevant to tumor prognosis and influence the expression of immune-related genes. However, the characterization of immune cell infiltration (ICI) has not yet been comprehensively analyzed and quantified in colon adenocarcinoma (COAD). Methods The multiomic data of COAD samples were downloaded from TCGA. ESTIMATE algorithm, ssGSEA method and CIBERSORT analysis were conducted to estimate the subpopulations of infiltrating immune cells. COAD subtypes based on ICI pattern were identified by consensus clustering then principal-component analysis was performed to obtain ICI scores to quantify the ICI patterns in individual tumors. Kaplan–Meier analysis was employed to validate prognostic value. Gene set enrichment analysis (GSEA) was applied for functional annotation. Finally, the mutation data was analyzed by employing “maftools” package. Results Three bioinformatics algorithms were used to evaluate the ICI patterns from 538 patients with COAD. Two ICI subtypes were determined using consensus clustering, and the ICI score was constructed by performing principal component analysis. Our findings showed that a higher ICI score often indicated a more advanced tumor and worse prognosis. The high-ICI score subgroup had a higher stromal score and more M0 macrophages but fewer plasma cells and decreased CD8 T cell infiltration. In addition, patients with high ICI scores had significantly higher expression levels of HAVCR2 and PCDC1LG2. Real-time polymerase chain reaction (PCR) was conducted to determine the prognostic significances of ICI-related genes. Conclusions In conclusion, ICI score may be considered as an original and useful indicator for independent prognostic prediction and individual immune-related therapy.
Keywords
