Acquisition and validation of four painful subtypes of colon adenocarcinoma and prognostic analysis

Abstract

Background: Colon adenocarcinoma (COAD) is the most common type of colorectal cancer. Pain is a multidimensional unpleasant experience and various molecular and cellular pathways are implicated in pain signaling. Nevertheless, the exploration of pain-related genes related to colon adenocarcinoma is not clear yet. Methods: In this study, the pathways enriched for pain-related genes were analyzed by Metascape. Then, we obtained pain subtypes versus classical subtypes and explored the link between the two. Next, marker genes for different pain subtypes were identified, the enrichment pathways were explored and these marker genes were used to validate the pain subtypes. We then performed an investigation of survival differences between pain subtypes by selecting specific top pathways in each subtype, calculating top pathway scores, and calculating pathway differences by heatmap and Kruskal test. Finally, we predicted the response of different pain subtypes to immunotherapy. Results: A total of 146 pain-related genes were enrolled in this study and we finally obtained 4 painful subtypes and 4 stable subtypes. The marker genes for subtypes were validated by The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) datasets and found to have a worse prognosis for CS1. The genes of CS1, CS2, CS3 and CS4 markers were mainly enriched in the pathways of Focal adhesion, Human T cell leukemia virus1 infection, Metabolic pathway, and Pertussis, respectively. CS1 and CS4 are more immunogenic. Moreover, CS1 is more sensitive to treatment with CTLA4 inhibitors, CS4 is sensitive to treatment with PD-1 inhibitors. Conclusions: Our study's identification of four pain subtypes of COAD provides new ideas for personalised therapy for patients with COAD.