Development of a prognostic risk model for colorectal cancer and association of the prognostic model with cancer stem cell and immune cell infiltration.

BACKGROUND

The development of a prognostic model for patients with colorectal cancer (CRC) can facilitate the assessment of patient survival and the effectiveness of clinical treatments. A reasonable prognostic model can provide a basis for individualized treatment, prognostic risk stratification, and subsequent therapy for CRC patients. The aim of our study was to construct a prognostic model for patients with CRC using sequencing data derived from The Cancer Genome Atlas (TCGA) database.

METHODS

Sequencing data of paracancerous tissues (n=51) and CRC samples (n=647) were downloaded from the TCGA database. Least absolute shrinkage and selection operator (LASSO) and Cox regression analyses were employed to identify prognostic factors. A restricted cubic spline (RCS) model was used to assess the nonlinear relationship between risk score and poor overall survival (OS). The Genomics of Drug Sensitivity in Cancer (GDSC) database was accessed to evaluate the correlation between the prognostic model's risk score and drug sensitivity. The single-sample gene set enrichment analysis (ssGSEA), estimate, and CIBERSORT algorithms were applied to quantify the association between prognostic genes and immune cell infiltration in CRC.

RESULTS

Our findings revealed that six genes, including Niemann-Pick C1-like 1 () [hazard ratio (HR) =1.53; 95% confidence interval (CI): 1.08-2.17; P=0.02], glucagon-like peptide 2 receptor () (HR =0.68; 95% CI: 0.48-0.97; P=0.04), solute carrier family 8 member A3 () (HR =0.67; 95% CI: 0.47-0.96; P=0.03), alpha-1-microglobulin/bikunin precursor () (HR =0.64; 95% CI: 0.45-0.91; P=0.01), single-pass membrane protein with coiled-coil domains 2 () (HR =0.68; 95% CI: 0.48-0.97; P=0.03), and tetratricopeptide repeat domain 16 () (HR =1.55; 95% CI: 1.09-2.20; P=0.02) function as independent prognostic factors for CRC. Based on these six genes, the developed prognostic assessment model identified a strong association between high risk score and poor OS (HR =2.43; 95% CI: 1.67-3.53; P<0.001) in patients with CRC. Furthermore, the analysis revealed a nonlinear relationship (P<0.001) between continuous variation in risk score and the risk of poor OS. Additionally, specific genes included in the prognostic model were found to be strongly associated with cancer stem cell and immune cell infiltration in CRC.

CONCLUSIONS

We developed a prognostic risk model incorporating a six-gene panel for patients with CRC. Our analysis revealed a nonlinear relationship between this prognostic model and OS in patients with CRC. A high risk score was associated with poor prognosis, indicating that the adverse outcomes observed in patients with CRC may be influenced by cancer stem cell and immune cell infiltration. Our model provides a promising predictive method for the prognosis of CRC patients, but it still needs to be validated in a larger sample size.