Exploring the Potential of Exosome-Related LncRNA Pairs as Predictors for Immune Microenvironment, Survival Outcome, and Microbiotain Landscape in Esophageal Squamous Cell Carcinoma

Accumulating studies have demonstrated the indispensable roles of exosomes and long non-coding RNAs (lncRNAs) in cancer progression and the tumor microenvironment (TME). However, the clinical relevance of exosome-related lncRNAs (ER-lncRNAs) in esophageal squamous cell carcinoma (ESCC) remains unclear. Three subtypes were identified by consensus clustering of 3459 valid ER-lncRNA pairs, of which subtype A is preferentially related to favorable prognosis, lower stromal and immune scores, and higher tumor purity scores. Higher immune cell infiltration, higher mRNA levels of immune checkpoints, higher stromal and immune scores, and lower tumor purity were found in subtype C, which presented a poor prognosis. We developed a prognostic risk score model based on 8 ER-lncRNA pairs in the GEO cohort using univariate Cox regression analysis and LASSO Cox regression analysis. Patients were divided into a high risk-score group and low risk-score group by the cut-off values of the 1-year ROC curves in the training set (GEO cohort) and the validation set (TCGA cohort). Receiver operating characteristic (ROC) curves, Decision curve analysis (DCA), clinical correlation analysis, and univariate and multivariate Cox regression all confirmed that the prognostic model has good predictive power and that the risk score can be used as an independent prognostic factor in different cohorts. By further analyzing the TME based on the risk model, higher immune cell infiltration and more active TME were found in the high-risk group, which presented a poor prognosis. Patients with high risk scores also exhibited higher mRNA levels of immune checkpoints and lower IC50 values, indicating that these patients may be more prone to profit from chemotherapy and immunotherapy. The top five most abundant microbial phyla in ESCC was also identified. The best ER-lncRNAs (AC082651.3, AP000487.1, PLA2G4E-AS1, C8orf49 and AL356056.2) were identified based on machine learning algorithms. Subsequently, the expression levels of the above ER-lncRNAs were analyzed by combining the GTEx and TCGA databases. In addition, qRT-PCR analysis based on clinical samples from our hospital showed a high degree of consistency. This study fills the gap of ER-lncRNA model in predicting the prognosis of patients with ESCC and the risk score-based risk stratification could facilitate the determination of therapeutic option to improve prognoses.