This article is more than 5 years old. Information may no longer be current.
Gut microbiome markers show promise for colorectal cancer diagnosis
Certain strains of gut microbiota are associated with colorectal cancer, and could be used as a non-invasive, accurate and cost-saving diagnostic test, according to new research published in Gut.
“Although previous studies have found associations between the gut microbiome and colorectal cancer, there was little agreement between the results. To our knowledge, our study is the first attempt to gather existing microbial marker gene data and reprocess it uniformly,” Manasi S. Shah, PhD, of the University of Texas School of Public Health, said in a press release. “Despite differences in cohort demographics, laboratory protocols and downstream analysis, which are known to influence microbiome outcomes, it is encouraging that we found bacterial signals that are consistently and significantly associated with colorectal cancer.”
Manasi Shah, PhD
Shah and colleagues performed a meta-analysis of nine prior studies published from 2012 to 2016, which used varying laboratory methods to evaluate associations between CRC and the gut microbiome. Seeking to identify a common set of diagnostic microbial markers, they evaluated raw 16S rRNA gene sequence data from 509 stool samples (79 colorectal adenoma, 195 CRC and 235 controls) sourced from ethnically diverse cohorts.
They also compared two different bioinformatics pipelines for processing the data: a commonly used closed-reference operational taxonomic unit assignment approach, and an alternate, novel strain-specific method.
Specific microbial strains and species associated with CRC included Parvimonas micra ATCC 33270, Streptococcus anginosus and several members of the Proteobacteria phylum. These were “frequently and significantly increased” in stool samples from CRC patients compared with controls across studies.
Compared with control classification, classifying CRC using these microbial markers yielded an area under the receiver operator characteristic (AUROC) curve of 76.6% with the closed reference bioinformatics pipeline and 80.3% using the strain-specific pipeline. When combining the microbial markers with clinical features, classification accuracies exceeded 80%, with AUROCs of 83.3% and 91.3% for each respective bioinformatics pipeline.
“Microbiome studies offer tremendous potential for advancing diagnostics and therapeutics. For some diseases and conditions, identifying microbiome-based associations has been relatively simple. For the majority, however, the research community is still sifting through a lot of data and a lot of noise as we seek to understand how the microbiome contributes to disease onset, progression and our ability to provide rapid, accurate diagnoses,” Emily B. Hollister, PhD, assistant professor of pathology and immunology at Baylor College of Medicine, said in the press release. “The identification of consistent patterns across studies of the colorectal cancer-associated microbiome represents a big step forward in these efforts.” – by Adam Leitenberger
Disclosures: Shah reports she worked as a consultant for Second Genome during the study, and several other researchers were employed by Second Genome during the study and hold stock options.
Read more about