Microbiome-based stool diagnostics could enable population-wide screening of colorectal cancer (CRC) if specific microbiota or signatures could be clearly linked to the disease. A study in Nature Medicine addressed two challenges in cancer microbiome research: using quantitative microbiome profiling with ribosomal RNA amplicon sequencing, and identifying covariates that could obscure microbiota–CRC interactions.

First, a comprehensive metadata analysis of 589 patients referred for colonoscopy as part of the Leuven CRC Progression Microbiome (LCPM) study cohort was performed and variables associated with diagnosis groups (no colonic lesions, polyps, CRC) such as age, body mass index and calprotectin (an indicator of intestinal inflammation) levels were identified. Previous non-CRC tumours, high blood pressure and diabetes treatment affected group distribution. Analysis of 94 potential covariates showed no separation between diagnostic groups and no difference in microbial load; 24 covariates were linked to microbial variation, with 17 non-redundant covariates accounting for 6.8% of overall microbial composition variation. Stool moisture content had the greatest effect on microbial variation, followed by intestinal bowel disease and/or ulcerative colitis. Interestingly, the cancer diagnosis group was not identified as a covariate for microbial variation.