Immune cell infiltration is associated with the prognosis of cancer. This study focused on the immune infiltration profiling and their association with survival outcome in nonsmall cell lung cancer (NSCLC). Research data were obtained from the Gene Expression Omnibus and The Cancer Genome Atlas databases. CIBERSORT algorithm was applied to assess the relative proportions of 22 kinds of immune cells. Log-rank test was performed to compare the survival outcome of patients with different proportions of immune cells. The estimated hazard ratios were presented with forest plot. Multivariate Cox regression analysis was conducted to estimate the adjusted associations between different types of infiltrating immune cells and survival prognosis controlling for other clinical features and confounders. With the CIBERSORT approach, we assessed the proportions of 22 infiltrating immune cells of 2050 cases with NSCLC. By conducting survival analysis, we found different survival outcomes among cases with different proportions of certain types of infiltrating immune cells. Among the cell subsets investigated, plasma cells (hazard ratio [HR] = 0.775, 95% confidence interval [CI]: 0.669-0.898) and regulatory T cells (HR = 1.258, 95% CI: 1.091-1.451) were associated with survival outcome of NSCLC patients controlling for other covariates. Subgroup analysis suggested a good consistency and robustness of our results. Our findings might provide useful information for prognosis prediction and cellular study in NSCLC.
Keywords: gene signature; immune infiltration; nonsmall cell lung cancer; survival.