Abstract
Background: Telomeres, consisting of TTAGGG repeats and six shelterin proteins, have been extensively studied in oncogenesis due to their crucial role in maintaining genomic stability. Despite these advances, the prognostic significance of telomere-related genes (TRGs) in esophageal squamous cell carcinoma (ESCC) remains poorly characterized. Methods: Transcriptomic data, single-cell RNA sequencing (scRNA-seq) data, and clinical information of ESCC patients were obtained from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. TRGs were retrieved from the TelNet database. Core TRGs were identified using three machine learning algorithms: least absolute shrinkage and selection operator (LASSO) regression, support vector machine (SVM), and random forest (RF). Based on prognosis-related TRGs, unsupervised clustering was performed to classify ESCC patients into two distinct molecular subtypes, and a prognostic risk model was subsequently constructed. Following risk stratification, survival analysis, immune infiltration analysis, drug sensitivity analysis, and molecular docking were conducted to further evaluate the potential clinical value of the risk model in ESCC. In addition, the expression patterns and intercellular communication of the model genes were examined using single-cell data. Finally, the differential expression of the core genes was validated by quantitative real-time PCR (qRT-PCR). Results: We developed a prognostic risk prediction model using five independent prognostic TRGs. The model demonstrated excellent performance in predicting patient outcomes and effectively revealed the heterogeneity of the tumor immune microenvironment (TIME). Drug sensitivity analysis indicated that ribociclib was more effective in the high-risk group. Single-cell analysis revealed the distribution of the five genes across cellular subpopulations. The qRT-PCR results further validated the differential expression of the core genes. Conclusion: Through a systematic analysis of TRGs expression patterns, we successfully developed a prognostic prediction model that effectively captures the heterogeneity in survival outcomes, TIME characteristics, and drug sensitivity among patients with ESCC.