Study Focuses on Role of Microbes in Pancreatic Cancer

Get Permission

Although bacteria are predominant in the gastrointestinal tract, they also reside on and in other parts of the body, including some unexpected places, such as malignant tumors. There are numerous reports of this phenomenon, but most have not identified a functional role for the microbes. In the research lab of Florencia McAllister, MD, of The University of Texas MD Anderson Cancer Center, the impact of bacteria on the development of pancreatic cancer is becoming better understood. At the 2020 ASCO-SITC Clinical Immuno-Oncology Symposium, Dr. McAllister, Assistant Professor in the Departments of Clinical Cancer Prevention and Gastrointestinal Medical Oncology, described her findings.1

“If there is microbial crosstalk between the gut and the tumor, potentially we could intervene in the tumor by altering the gut microbiome.”
— Florencia McAllister, MD

Tweet this quote

The potential for bacteria’s impact should come as no surprise. “The human body houses as many bacteria cells as human cells; therefore, these bacterial cells are likely to be important,” said Dr. McAllister.

The scenario in pancreatic cancer is essentially one in which bacteria play indirect and direct roles as “antitumoral” microbes (increasing microbial diversity, enhancing immune activity) or indirect “protumoral” microbes (affecting enzyme activity and reducing sensitivity to gemcitabine, teetering in a protumorigenic/antitumorigenic balance). The fungal microbiome has also been shown to promote oncogenesis. “Most of the data have been published in the past 3 years, so this is actually a novel area,” Dr. McAllister commented.

Revealing Research

Recent preclinical studies have highlighted the role of the microbiome in the initiation and progression of pancreatic ductal adenocarcinoma. One early experiment found that cell lines grow differently when contaminated, and they responded differently to gemcitabine, leading researchers to examine whether bacteria could also exist in the tumors of patients. Ultimately, bacteria were indeed detected in pancreatic tumors, with the primary species being Gammaproteobacteria—a species that can affect enzyme activity such that sensitivity to gemcitabine is impaired.2 Geller et al tested 113 human pancreatic tumors and found that 86 (76%) were positive for bacteria, mainly Gammaproteobacteria. They concluded that intratumoral bacteria contribute to drug resistance in pancreatic cancers.

Pushalkar et al found that the cancerous pancreas harbors a markedly more abundant microbiome than a normal pancreas in both mice and humans and that select bacteria are differentially increased in the tumorous pancreas compared with the gut.3 In knockout mice with a genetic predisposition for developing pancreatic cancer, those raised in a germ-free environment did not develop premalignant lesions. Tumors did grow, however, after fecal transplant from a mouse with pancreatic cancer, although not after transplant from a normal mouse. Use of broad-spectrum antibiotics also delayed the growth of pancreatic tumors—but not in knockout mice that lacked activated T cells.

Such studies have not only shown that bacteria can live in the pancreas, but also that they have functional relevance in tumor initiation. To further define the role of intratumoral bacteria in pancreatic cancer, researchers are looking at long-term survivors after pancreatic tumor surgery. Genomic profiling has not distinguished them from less fortunate patients, suggesting other factors may be at play.

Dr. McAllister believes that treatment response is associated with the microenvironment (including the microbiome of the tumor) and the immune profile of the tumor. To study this, her team examined two pancreatic cancer cohorts—one at MD Anderson Cancer Center and the second at Johns Hopkins—employing 16S sequencing to determine the bacterial phylogenies.4 As has been previously shown, they found that the tumors of long-term survivors have a higher number of activated T cells, high neoantigen quality, and higher alpha (within-group) microbial diversity. For beta (between-group) diversity, they saw a certain microbial “clustering” for the long-term survivors from both cohorts, pointing to several common species that may be important in pancreatic tumors—a predictive microbial signature, she said.

Specifically, three species were present in more than 70% of the long-term survivors of both cohorts: Pseudoxanthomonas, Saccharopolyspora, and Streptomyces; Bacillus clausii was also highly represented in both cohorts. The combination of those four species yielded a sensitivity rate of 97% in predicting long-term survivorship of patients with pancreatic cancer who underwent resection.4

“We found that greater diversity in the tumors is better, and it correlates with immune activation, but what does that really mean, and can we intervene?” continued Dr. McAllister. “Although it’s not easy to put bacteria into the pancreas, we know there’s an interaction, or an anatomic connection, with the gut,” she said. “If there is microbial crosstalk between the gut and the tumor, potentially we could intervene in the tumor by altering the gut microbiome.”

In the next experiment, they did just that, showing that this microbial crosstalk allowed for fecal microbial transplant to modulate tumors in mice.4 Mice previously treated with antibiotics received stool from long-term survivors of pancreatic cancer, from patients with metastatic pancreatic cancer (short-term survivors), and from healthy controls and 2 weeks later were implanted with pancreatic tumors. The question was whether fecal transplant could change the gut microbiome, the tumor microbiome, and the immunoprofile of the mice, and whether this would ultimately affect the growth of those tumors.

Indeed, fecal transplant did change the gut microbiome, the tumor microbiome, and the immune activation profile in the tumors. Mice that received stool from patients with pancreatic cancer had the lowest number of activated T cells, more immunosuppressive cells, and the largest tumors. In comparison, the smallest tumors were seen in mice that received stool from the long-term survivors. To determine whether the effect might be mediated by CD8, the researchers neutralized CD8 and observed that the protective effect of the long-term survivor stool was lost. This finding suggested the immune response was, indeed, causing the antitumorigenic effect.

Next Steps

“This antitumoral effect seen in long-term survivors is likely mediated by CD8 T cells, and this needs to be further explored to see whether there is a metabolic component to that response,” Dr. McAllister suggested.

Dr. McAllister and colleagues are now determining whether these and other research findings can be used to develop biomarkers or to create a preventive approach. In one study, they are examining the gut, oral, and tumor microbiomes of patients at high risk for developing pancreatic cancer, intervening as appropriate in an attempt to prevent cancer. From a therapeutic perspective, they are planning their first fecal microbial transplant trial of patients with pancreatic cancer, seeking to replicate their preclinical findings in humans. 

DISCLOSURE: Dr. McAllister reported no conflicts of interest.


1. McAllister F: Role of intratumoral microbes in pancreatic cancer and interaction with gut microbes. 2020 ASCO-SITC Clinical Immuno-Oncology Symposium. General Session. Presented February 8, 2020.

2. Geller LT, et al: Potential role of intratumor bacteria in mediating tumor resistance to the chemotherapeutic drug gemcitabine. Science 357:1156-1160, 2017.

3. Pushalkar S, et al: The pancreatic cancer microbiome promotes oncogenesis by induction of innate and adaptive immune suppression. Cancer Discov 8:403-416, 2018.

4. Riquelme E, et al: Tumor microbiome diversity and composition influence pancreatic cancer outcomes. Cell 178:795-806, 2019.