Microbiome confounders and quantitative profiling challenge predicted microbial targets in colorectal cancer...

Area of Science:

• Microbiome research
• Cancer biology
• Gastroenterology

Background:

• Cancer microbiome research has advanced, but confounders and absolute quantification are underutilized.
• This underutilization raises concerns about potentially spurious associations in studies.

Purpose of the Study:

• To investigate the fecal microbiota in colorectal cancer (CRC) patients across different stages.
• To identify reliable microbial biomarkers for CRC by controlling for significant covariates.
• To assess the impact of confounders on previously identified CRC-associated microbes.

Main Methods:

• Quantitative microbiome profiling using 16S ribosomal RNA amplicon sequencing.
• Analysis of fecal microbiota from 589 CRC patients and comparison with 15 published studies (4,439 total).
• Rigorous control for confounders including transit time, fecal calprotectin, and body mass index.

Main Results:

• Transit time, fecal calprotectin, and BMI were primary microbial covariates, explaining more variance than CRC stage.
• Established CRC targets like Fusobacterium nucleatum showed no significant association with CRC groups after covariate control.
• Anaerococcus vaginalis, Dialister pneumosintes, Parvimonas micra, Peptostreptococcus anaerobius, Porphyromonas asaccharolytica, and Prevotella intermedia showed robust associations.
• Individuals without colonic lesions but meeting colonoscopy criteria were enriched for the Bacteroides2 enterotype, indicating challenges in defining healthy controls.

Conclusions:

• Quantitative microbiome profiling and rigorous covariate control are essential for accurate biomarker identification in CRC studies.
• Previously assumed CRC biomarkers may be spurious due to unaddressed confounding factors.
• Specific bacteria like A. vaginalis and P. intermedia show promise as potential CRC biomarkers.