NEW YORK – Using sequence data for thousands of metastatic tumor samples spanning dozens of cancer types, a team led by investigators at the Netherlands Cancer Institute has identified microbial community members residing in those tumors. They include bacteria linked to immune features and treatment responses.
"Together, we generated a pan-cancer resource of the metastatic tumor microbiome that may contribute to advancing treatment strategies," co-senior and corresponding author Emile Voest, a researcher at the Netherlands Cancer Institute, and colleagues wrote in Cell on Tuesday.
For their study, Voest and colleagues performed whole-genome sequencing on 4,160 pre-treatment metastatic tumor biopsy samples linked to primary cancers in 26 tissue types and analyzed the cancer sequences in combination with clinical data and RNA sequencing-based transcriptome data.
In addition, the team turned to mapping-based metagenomic profiling with the PathSeq and Kraken2 tools — together with microbial genome assembly methods — to get a look at the microbial community composition, gene sequences, and bacterial diversity in the tumors.
Across 3,526 of the tumor biopsy samples, the researchers uncovered microbes from 165 genera, with nearly 70 percent representing species in anaerobic or facultative anaerobic genera.
Those results were shored up with targeted 16S ribosomal RNA sequencing on a subset of samples, while the metagenomic sequence set made it possible to put together more than 500 genomes of microbes found in the samples.
"As metastases are responsible for the majority of solid cancer deaths and are key targets for systemic anti-cancer therapies, there is an urgent need to understand how tumor-resident bacteria can reshape tumor biology, immune infiltration, and responsiveness to treatment," the authors explained.
With the help of bioinformatic analyses, the team explored relationships between the microbiome, metastatic tumor features, tissue type, immune cell composition, and other tumor microenvironment (TME) clues, identifying microbial community patterns that varied depending on the primary tumor type, site of metastasis, and environmental conditions at these sites.
"[W]e integrated metagenomics, genomics, and transcriptomics to profile the metastatic tumor microbiome and — at the same time — resolve physiological characteristics of the tumor and its TME," the authors explained, "to provide biological insight and guide the development of bacteria-oriented strategies to complement and enhance cancer (immuno)therapy."
The team saw distinct microbial clusters in groups of metastatic tumors classified as microsatellite instability (MSI)-high or microsatellite stable/MSI-low — differences that further coincided with distinct interferon-gamma immune signaling features.
Whereas anaerobic bacterial species turned up in metastatic tumors found in tissues with low exposure to oxygen (a so-called "hypoxic" environment), meanwhile, the researchers saw enhanced microbial diversity in tumor samples marked by specific immune cell compositions and immune signaling activity.
"We determined an association between a low-oxygen TME and the abundance of anaerobic bacteria in tumors," the authors reported, noting that the results "do not establish whether tumor-resident anaerobic bacteria play a causal role in the development of a hypoxic TME or whether their presence simply reflects suitable conditions for anaerobic colonization."
Intriguingly, their results also revealed metastatic tumor microbiome features that coincided with treatment outcomes, including poorer immune checkpoint blockade immunotherapy responses in non-small cell lung cancer patients whose metastatic tumors contained Fusobacterium bacteria.
Likewise, the team explored ties between bacterial diversity and immune evasion capabilities, while longitudinal profiling of a subset of tumors during immune checkpoint blockade immunotherapy highlighted bacterial genera such as Actinomyces, Bacteroidetes, Bifidobacterium, and Prevotella that get dialed down after immune activation prompted by such therapies.
"Our analyses show how intratumoral communities vary among anatomical sites, depend on the primary tumor type, associate with immune cell infiltration, and correlate with treatment responses, especially in the context of immunotherapy," the authors wrote.