NEW YORK – Gut microbial communities appear to shift in individuals with pancreatic cancer, according to new research by investigators in Germany, South Korea, and other international sites.
Using a combination of shotgun metagenomic sequencing and targeted 16S ribosomal RNA sequencing, the researchers profiled microbiome features found in stool, saliva, tumor, and/or matched normal samples collected before and after treatment from 25 Spanish individuals diagnosed with early-stage pancreatic ductal adenocarcinoma (PDAC) and another 32 patients with advanced forms of the disease.
Their findings, published in the journal Gut on Tuesday, suggested that microbiome features found in stool samples were more informative than saliva microbiome analyses, when it came to predicting PDAC risk. The group highlighted several microbial species that were overrepresented in gut microbial communities from the PDAC patients compared to samples from 29 individuals with chronic pancreatitis, a condition linked to enhanced pancreatic cancer risk, and to samples from 50 unaffected control individuals.
"Taken together, our results indicate that non-invasive, robust, and specific fecal microbiota-based screening for the early detection of PDAC is feasible," wrote co-corresponding authors Peer Bork, Nuria Malats, and their colleagues. Bork is a structural and computational biology researcher affiliated with the European Molecular Biology Laboratory, the University of Wurzburg, Yonsei University, and the Max Delbruck Centre for Molecular Medicine, while Malats is affiliated with the Spanish National Cancer Research Centre and CIBERONC.
Along with altered levels of certain Bifidobacterium, Lactobacillus, and Bacteroides species in tumor samples relative to normal tissue samples, verified using fluorescence in situ hybridization on a subset of samples, the team identified several gut microbial species with higher- or lower-than-usual levels in the PDAC patients relative to control individuals or pancreatitis patients.
Such microbiome shifts did not appear to be pancreatic cancer stage-specific, the researchers explained. With a machine learning analysis of the gut microbiome features, they could detect pancreatic cancer cases with roughly 84 percent accuracy.
That classification got a further boost, reaching 94 percent accuracy, when they incorporated information on a blood marker for pancreatic cancer — results the team verified using samples from another 44 individuals with PDAC and 32 unaffected controls from a prior German case-control study and in a set of almost 5,800 more samples from individuals diagnosed with a wide range of health conditions.
"[T]he described fecal microbiome signatures enabled robust metagenomic classifiers for PDAC detection at high disease specificity, complementary to existing markers, and with potential towards cost-effective PDAC screening and monitoring," the authors wrote.
Even so, they cautioned that the current findings "are strictly observational," but provide rationale for further analyzing fecal microbiome patterns in future efforts to find noninvasive strategies for early pancreatic cancer detection.
In a related commentary article in Gut, University of Florida researchers Rachel Newsome and Christian Jobin noted that "these findings have limited clinical value due to the cross-sectional nature of the study."
"[T]he predictive markers will need to be tested using a prospective cohort before reaching a conclusion on their clinical impact," Newsome and Jobin wrote.