Abstract
BACKGROUND: Esophageal cancer (ESCA) carries a poor prognosis, and the exploration of mRNA vaccines for its treatment remains limited. This study aims to identify potential tumor antigens and characterize the immune landscape, thereby providing a foundation for developing mRNA vaccines against ESCA. METHODS: A total of 150 and 179 specimens were analyzed using The Cancer Genome Atlas (TCGA)-ESCA and GSE53625 datasets. Quantitative real-time polymerase chain reaction (qRT-PCR) were performed on cDNA microarrays to verify the transcriptional levels of potential antigens. The immune subtypes were delineated using consensus clustering and module eigengenes were calculated by weighted gene co-expression network analysis (WGCNA). RESULTS: We identified 5 tumor antigens with overexpression and mutation, which were associated with antigen-presentation and poor prognosis. A total of two subtypes (IS1 and IS2) were identified, and IS1 showed a better prognosis. The mutation count and tumor mutation burden were slightly higher, whereas immune checkpoint genes were lower in IS2. Compared with IS1, an increase in immune and stromal cell infiltration was associated with IS2, indicating that IS2 is immunologically "hot" and IS1 is immunologically "cold". Furthermore, ESCA patients' immune cell components were identified and survival outcome predictions were made by immune landscape construction. Immune hub genes, such as DKK1, could serve as biomarkers for predicting the prognosis and for vaccination. Finally, drug sensitivity analysis showed that IS1 patients might have higher sensitivity to TOP9 drugs with significant differences, which emphasized the importance of individualized treatment for ESCAs. CONCLUSIONS: We identified ANGPT2, CRIPT, GLA, LMNB1, and MARVELD3 as potential tumor antigens. Our study implies that IS2 phenotype might benefit from mRNA vaccination.