The microbiome refers to the sum of all of the gene sequences in a community of microbes. Rapidly improving, high-throughput DNA amplification and sequencing technologies are being used to characterize the microbial communities that reside on and in the human body, as well as their associations with disease. Our goal is to establish a bank of fecal specimens, suitable for microbiome and other assays, prospectively and cost effectively collected and stored from a general population of many thousands with essential covariate data and good ascertainment of cancer events. Toward that goal, we are pursuing a two-pronged strategy.
We have made the most progress on the first strategy, which is to identify associations of fecal microbiome metrics with malignancies and with cancer risk factors. Colorectal cancer (CRC) is a major focus. In a case-control study with data and specimens previously collected by others in DCEG, we found that CRC cases had reduced fecal microbiome alpha diversity, increased carriage of Fusobacterium and Porphyromonas taxa, and reduced abundance of Clostridia taxa. Metabolomic analysis of the same specimens uncovered 41 small molecules that differed between cases and controls, providing potentially improved diagnostic methods or insights on carcinogenesis. Metagenome analyses of associations with CRC and with fecal metabolites are in progress. Our study of twins, with collaborators at Washington University in St. Louis and the University of Southern California, found that 13 Hodgkin lymphoma survivors had less diversity of their fecal microbiome compared to their unaffected co-twin controls. Such lower diversity could be a consequence of the lymphoma or its treatment, or perhaps of a propensity to Hodgkin lymphoma as postulated by the “hygiene hypothesis”.
Our first strategy has also focused on endogenous estrogens and postmenopausal breast cancer. In 51 epidemiologists recruited from DCEG, our collaboration with the University of Maryland Medical School Institute of Genome Sciences (IGS) revealed that fecal microbiome alpha diversity was significantly lower with younger age, use of various prescription medications, and, in men and postmenopausal women, lower levels of systemic estrogens. The estrogen association is being pursued with fecal metagenome and meta-transcriptome analyses. Moreover, in a recently completed case-control study of newly diagnosed breast cancer in collaboration with IGS and Kaiser Permanente Colorado, we found that cancer cases had significantly lower alpha diversity compared to controls, and this was independent of estrogen levels.
The second strategy is to identify and resolve obstacles to obtaining microbiome-quality feces from existing cohorts, prepaid health plans, or screening populations. Our study of DCEG epidemiologists demonstrated excellent reproducibility of microbiome alpha and beta diversity metrics in stool self-collected with either of two devices containing RNAlater. In the same population we are evaluating whether used fecal immunochemical test (FIT) devices yield equivalent microbiome reproducibility. Feasibility of collecting and performing fecal microbiome analyses in a large cohort in Shanghai has been demonstrated in a study of 68 participants in a CRC screening program in collaboration with the Shanghai Center for Disease Control and BGI. Recent analysis has revealed a significant difference in the composition (beta diversity) of the fecal microbiota between people with colorectal adenoma and those with a completely normal colonoscopy.