Background: Esophageal cancer (EC) is a prevalent malignancy with heterogeneous outcomes. This study explores the significance of anoikis-related long non-coding RNAs (lncRNAs) in EC, aiming to unravel their molecular roles and clinical implications.
Methods: Transcriptome and clinical data were obtained from TCGA database for EC samples. We identified anoikis-related genes and lncRNAs by Pearson correlation analysis. The risk score model hinged on prognostic lncRNAs filtered from multiple steps. Risk scores were calculated using the derived formula, and categorized patients into low- and high-risk groups. Model robustness was assessed through Kaplan-Meier (KM) survival analysis and Receiver Operating Characteristic (ROC) curve, with clinical utility achieved via a constructed nomogram. We also explored the interplay between the risk score and immune cell infiltration, and investigated drug sensitivity.
Results: We identified 2365 anoikis-related lncRNAs through co-expression analysis, including 1415 significant lncRNAs differentially expressed between normal and tumor samples. A risk score model was constructed from ten prognostic lncRNAs. The risk score model effectively stratified patients based on the median score, and its predictive capacity was validated through KM survival, ROC curve analyses, and the external GSE53622 dataset. The nomogram provided a practical tool for individualized prognosis evaluation. We unveiled significant correlations between specific immune cell subsets and the risk score. Eosinophils and common lymphoid progenitors exhibited positive associations, while endothelial cells and myeloid dendritic cells showed negative correlations. Drug sensitivity analysis revealed potential sensitive drugs for EC treatment that aligned with the risk subgroups.
Conclusion: This study established an anoikis-related lncRNAs risk score model that may predict the prognosis, immune infiltration, and drug sensitivity in EC, in hope of facilitating tailored patient management.
