Transcriptomic analysis of patients with colorectal cancer in Ireland.

191 Background: Early onset colorectal cancer (EOCRC), diagnosed in patients under 50 years, is rising globally. The underlying biological mechanisms of EOCRC, compared with late-onset CRC (LOCRC), remain to be elucidated. Transcriptomic profiling enables a comprehensive exploration of regulatory networks, molecular subtypes, immune landscape, and metabolic rewiring across age-defined CRC cohorts. Methods: Bulk RNA sequencing was performed on matched tumour and adjacent normal tissues from microsatellite stable EOCRC (n=19) and LOCRC (n=28) patients. Genes with FPKM ≥1 in ≥50% of samples were retained (n=6,726). Principal component analysis (PCA) and clustering were used to assess global expression patterns. Further, consensus molecular subtypes (CMS) were assigned using the classifieRc application. Immune cell composition was inferred using the MCP-counter deconvolution framework. Metabolic reprogramming was assessed by projecting transcriptomic data onto metabolic networks using Metabolic Reaction Enrichment Analysis (MaREA). Results: Principal component analysis showed clear separation of tumour and adjacent normal tissue. CMS distribution was similar between EOCRC and LOCRC, with CMS1 the most frequent in both cohorts. Notably, all CMS1 cases were microsatellite stable (MSS), highlighting a potential therapeutic gap as this immune active subtype is currently excluded from immunotherapy eligibility. Immune deconvolution confirmed greater infiltration in CMS1 but revealed no significant difference between the two cohorts. Metabolic profiling identified EOCRC specific vulnerabilities, including reduced glutamate cysteine ligase activity and cystine uptake, decreased nucleotide synthesis, and reduced pyruvate dehydrogenase complex function, consistent with glycolytic dependence. These features suggest heightened EOCRC sensitivity to oxidative stress and potential metabolic intervention opportunities. Conclusions: Transcriptomic profiling reveals both shared and unique features of EOCRC and LOCRC. While CMS distribution and immune landscapes are broadly conserved, EOCRC demonstrates unique metabolic vulnerabilities and transcriptomic separation from normal tissue. These findings underscore the need for refined molecular stratification, particularly for MSS CMS1 patients, and highlight metabolic interventions as an underexplored avenue in CRC treatment.