LAUSR

Introduction Colorectal tumours are composed of mixed tissues, including tumoral and stromal cell types. The functional communication between these cell entities defines the biology of the tumours and eventually influences disease progression and patients’ prognosis. Recently, the first bona fide consensus molecular subtypes of colorectal cancer have been defined, each of them showing a specific immune-cell composition. CMS1 comprises hypermutant tumours with better prognosis associated to a higher immune infiltration. CMS2 tumours are characterised by a high CIN and strong WNT/MYC pathways activation. CMS3 have activated pathways related to metabolism. CMS4 tumours are more stromatic and showed the worst survival. In this work, we aim to assess the immune cell infiltration in the tumour as a prognosis indicator for colorectal cancer as well as to understand the molecular biology behind this crosstalk, while stratifying tumours by molecular subtype. Material and methods Our discovery series includes 100 paired normal and stage II MSS tumour samples. For validation purposes, public GSE39582 and TCGA series have been used. Using gene expression data, tumours have been classified into CMS subtypes using CMSclassifier R package. We were estimating tumour immune cell infiltration (R library MCPcounter) and calculating the immunephenoscores. GSEA has been performed to make a functional enrichment analysis. To assess the utility as prognosis biomarkers of genes expressed in the immune compartment, a cox model were fitted. Results and discussions We corroborate that CMS1 and CMS4 subtypes showed higher levels of immune and stromal infiltration than CMS2-3 subtypes. Several CD molecules have been found to be under-expressed in tumour vs. normal tissues that have a protective effect. Interestingly, when stratifying by subtypes, they showed the better accuracy discriminating between good and poor prognosis patients in the CMS4 subtype. Significant functions associated with higher expression of those molecules are IGA production, IL12 pathway, TCR pathway and PD1 signalling among others; thus suggesting an immune modulation of the microenvironment that leads to a better prognosis. Conclusion The CMS4 subtype has usually higher risk of relapse comparing to other subtypes. However, segregating the group using specific immune biomarkers is able to identify patients with the significantly better survival.