Comprehensive identification of a migrasomes-associated long non-coding RNA signature to predict the prognosis and treatment options in colon adenocarcinoma.

backgroundMigrasomes, recently discovered cellular substructures, may play a crucial role in cancer progression, treatment response, and prognosis. However, the prognostic value of migrasome-associated long non-coding RNAs (lncRNAs) in colon adenocarcinoma (COAD) remains unexplored.

methodsRNA-seq data from 459 COAD patients, including clinical characteristics and outcome information, were obtained from The Cancer Genome Atlas. A risk model was constructed through co-expression analysis of migrasome genes and lncRNAs, followed by Cox regression and least absolute shrinkage and selection operator analysis to identify prognostic lncRNAs. Functional enrichment analyses were performed to elucidate underlying biological mechanisms. Immune landscape characterization utilized ESTIMATE, CIBERSORT, Tumor Immune Estimation Resource (TIME), and single-sample Gene Set Enrichment Analysis (ssGSEA). Drug sensitivity analysis was conducted for select therapeutic agents.

resultsNine prognostic lncRNAs (AC010463.3, AL590483.4, AP005264.1, ZEB1-AS1, AC104088.1, PRKAR1B-AS2, AC009315.1, SUCLG2-AS1, and AC006111.2) were identified and incorporated into a risk model. Low-risk patients demonstrated significantly improved survival outcomes. The model exhibited independent prognostic capability, with AUCs of 0.783, 0.749, and 0.713 for one-, three-, and five-year survival, respectively, in the training cohort. High-risk patients displayed reduced overall survival and elevated tumor mutation burden. Additionally, these patients showed decreased sensitivity to therapeutic agents, including Oxaliplatin, Irinotecan, and 5-Fluorouracil.

conclusionOur novel migrasome-associated lncRNA signature demonstrates robust predictive capacity for both prognosis and chemotherapeutic sensitivity in COAD, potentially facilitating personalized treatment strategies and improved patient management.