Integrative genome-scale analysis of immune infiltration in esophageal carcinoma

To explore the molecular mechanism in the esophageal squamous carcinoma (ESCC) environment, we selected datasets of ESCC patients from The Cancer Genome Atlas (TCGA) (n = 78) and explored the infiltration condition of 24 immune cells in each sample. We assorted the microenvironment of ESCC into two Infiltration groups (I and II) and built a random forest classifier model. We showed traits of gene and clinicopathology in the tumor microenvironment (TME) phenotypes systematically. Infiltration I had low infiltration of immune cells and immunomodulators but relatively higher mutation load, while Infiltration II was enriched with cytotoxic T cells and immunosuppressive cells. The upregulation of several immune cytokines like IFN-γ, TNF-β, and PD-L1 was seen in Infiltration II. The infiltration group was an independent predictor of prognosis showed by Multivariable Cox analysis (Infiltration II vs. I, hazard ratio = 2.73, 95% confidence interval = 1.08-6.91, p = 0.03). All the results can be verified in datasets from the Gene Expression Omnibus database (GEO) and our institution (n = 98). Our results demonstrate a synthesis of the infiltration pattern of the immune in ESCC. We reveal the mechanism of TME, which may contribute to the progress of immunotherapy for patients with ESCC.