Colorectal Cancer Subtyping With Microbiome—A Game Changer for Personalized Medicine?
Colorectal cancer (CRC) was one of earliest solid tumors to be molecularly characterized, and the cascade of genetic events involved in the initiation and progression of CRC are well established.1 Advances in “omics” platforms encompassing genomic, epigenetic, transcriptomic, and immunologic data have led to an international effort to consolidate consensus molecular subtypes (CMS) of CRC in 2015,2 including micro-satellite instability (MSI)–immune (CMS1), canonical (CMS2), metabolic (CMS3), and mesenchymal (CMS4). Although CMS has proven to be a robust classification in capturing the tumor biology of CRC, conflicting results have been reported for CMS as predictor of therapy response.3 This implies that CMS only partially reflects tumoral features that affect drug response, and alternative elements, such as tumor heterogeneity and micro-environment, are potential complicating factors for CRC prognostication. Research over the past decade has established the gut microbiome as an indispensable component of tumor microenvironment that co-evolves with CRC development.4 CRCassociated dysbiosis is causative in tumorigenesis5 and can predict patient survival.6,7 Emerging preclinical evidence has also suggested gut microbiota as potential modifiers of drug response,8,9 highlighting gut microbes as confounding factors that might limit the utility of molecularbased signatures such as CMS. In this issue of Gastroenterology, Mouradov et al10 seek to establish a novel microbiotabased stratification strategy for CRC, its correlation with molecular features of CRC, and its prognostic implications. In this study, Mouradov et al enrolled 2 independent cohorts with a total of 716 CRC patients and performed bacterial 16S rRNA gene sequencing of tumors and normal colon mucosa in 2 independent cohorts. In parallel, molecular characterization of CRC tumors, such as MSI, DNA methylation (CpG island methylator phenotype [CIMP]), mutation signature, chromosome instability (CIN), and CMS, was conducted to integrate genomic, epi-genetic, and meta-genomic features in these patients. Previous studies that attempted to subgroup CRC patients either lack detailed clinico-molecular information4 or have limited cohort size.11 As expected, Mouradov et al identified the enrichment of pathogenic microbes together with depletion of protective commensal bacteria. Taking the top 30 bacteria enriched in CRC and an equal number of depleted species, tumor microbiomes are clustered into major onco-microbial community subtypes (OCS). OCS1 is enriched