The cancer microbiome

Abstract

The discovery that tumours have their own specific microbiome, together with more understanding of the role of the gut microbiome, has great potential for cancer diagnostics and targeted therapies.

The cancer microbiome has only emerged as a term and research field during the past decade, even though the concept dates back to the late 19th century when William Coley first hypothesised that tumours might have a microbial origin and that bacteria could have potential for therapy (Coley, 1891). The recent upsurge in research, however, is not a rediscovery of Coley's musings but the result of better knowledge of tumourigenesis combined with next‐generation sequencing (NGS), which has enabled more detailed studies of the tumour microenvironment. NGS allows detection of low densities of microbes in tumours, often accounting for as little as 1% of the sample.

Bacteria and cancer

Indeed, it is becoming increasingly clear that microbes play a key role in cancer at two levels. First, the general microbiome, especially of the gut, determines a fundamental risk of developing cancer, at least solid tumours, through interactions with immune pathways and through mediating inflammatory processes. Gut microbiome dysbiosis has now been shown to increase the risk not just of colorectal cancers where it can play a direct role through contact, but also other cancer types in the body.

Secondly, bacteria in tumours themselves can play a role in changing the microenvironment, and interfering with host immunity or therapies such as immunotherapy or radiotherapy that are being applied. There is also increasing evidence that bacteria can indirectly influence metastasis, which is the main cause of death when secondary tumours spread and colonise other parts of the body.

… bacteria in tumours themselves can play a role in changing the microenvironment, and interfering with host immunity or therapies such as immunotherapy or radiotherapy…

These two strands—the role of the gut microbiome and the impact of bacteria on the tumours themselves—have been unravelled together. The immediate tumour environment itself had been thought to be sterile, despite the early insights from Coley and a few others at the time. Then, a series of papers during the past decade has provided evidence that intratumoural bacteria do exist and that they play a role in both development of the cancer and outcome of therapies.

Intratumoural bacteria

After various findings linking bacteria to specific tumours, a more comprehensive study of the tumour microbiome analysed 1,526 tumours and their adjacent normal tissues across seven cancer types, including breast, lung, ovary, pancreas, melanoma, bone, and brain (Nejman et al, 2020). The authors identified distinct microbiome compositions associated with each tumour type, with breast cancer having a particularly rich and diverse microbiome. The intratumour bacteria were found to be mostly intracellular and present in both cancer and surrounding immune cells. There was also a clear correlation between intratumour bacteria and their predicted functions with tumour types and subtypes, as well as lifestyle habits such as smoking status. Moreover, the authors also identified a link between the intratumoural bacteria and response to immunotherapy.

A year later, another comprehensive study looked at the four subtypes of breast cancer using microarray analysis (Banerjee et al, 2021). This yielded a signature of all pathogen types, including viruses, bacteria, fungi, and parasites, which was then correlated with clinical data to identify those microbes that might have prognostic potential. Perhaps more significantly, the study demonstrated correlations between the presence and absence of specific microbes in breast cancer subtypes and clinical outcomes. This led the authors to muse about building a map of the oncobiome of breast cancer subtypes that would provide insights into disease prognosis and that could enable precise therapeutic interventions.

Such studies of intratumoural bacteria have also been extended to other cancer types, including melanoma. One publication identified bacterial peptides exposed by melanoma cells that were at least potentially immunogenic, implying they could be targets for immunotherapy (Sepich‐Poore et al, 2021). At any rate, the work identified a new class of therapeutically relevant tumour antigens of non‐human origin. This has increased interest in using antibiotics to target these bacterial peptides. Indeed, antibiotics have already been used, often in conjunction with other therapies, because of their general stimulation of some apoptotic pathways and antiproliferation effects, both of which can impede tumour growth.

Antibiotics in cancer therapy

Fusobacterium is a genus of anaerobic, gram‐negative bacteria that is associated with colon cancer and oral tumours, and other human conditions, including periodontal diseases, Lemierre's syndrome and topical skin ulcers. Given its association with solid cancers there has been growing interest in targeting Fusobacterium specifically in conjunction with other therapies to enhance their efficacy, according to Matthew Meyerson at the Dana‐Farber Cancer Institute in Boston, USA. “Fundamentally, we still do not know whether the presence of Fusobacterium in cancers is in part a cause of the cancer, a consequence of the cancer, or a coincidence,” he said. But this does not mean it cannot be a useful target, and Meyerson cited an ongoing clinical trial in China to test the use of the antibiotic and antiprotozoal drug metronidazole in colorectal cancer (CRC) (Study of Oral Metronidazole on Postoperative Chemotherapy in Colorectal Cancer—Full Text View—ClinicalTrials.gov).

Given its association with solid cancers there has been growing interest in targeting Fusobacterium specifically in conjunction with other therapies to enhance their efficacy…

This trial is led by Jing‐yuan Fang at the Shanghai Jiao Tong University School of Medicine. The key motivation was the observation that Fusobacterium nucleatum was abundant in colorectal cancer tissues among patients whose cancer recurred after chemotherapy, and that it promoted resistance to the treatment. Metronidazole was chosen because it is effective against F. nucleatum. Fang pointed out that the trial was first impeded by the COVID‐19 pandemic and the extensive lockdowns in China but is now gathering pace. “We have enrolled more than half the patients and the enrolment is still ongoing, about one tenth of patients have completed the whole follow‐up,” he said. “We hope we will finish the enrolment in the next 3 months and it may take another 2 years to finish the trial.”

Fang is optimistic both that the trial will provide a firm evidence base for future treatment on account of its size, and that it will make a difference in a country with a large population where CRC is the third‐most diagnosed cancer. “Our trial is a well‐designed double‐blind, placebo‐controlled RCT of multi‐centres with large‐size numbers which means that the result will provide reliable evidence and influence the treatment strategy of CRC,” he commented. “The incidence of CRC continues to increase, which may lead to a sustained increase in the burden of CRC in the future. Our results may provide a cost‐effective alternative drug in treatment that we believe it will be welcomed by the government and National Healthcare Security Administration.”

Yet, as some studies have shown, antibodies should not be used unless carefully targeted, because they can also have a negative impact on anticancer treatments. One of those studies cited various findings that antibiotics can actually induce cancer by disrupting intestinal microbiota, which in turn promotes chronic inflammation and alters normal tissue metabolism (Gao et al, 2020). Treating cancer would therefore require more specific and targeted antibiotics to maximise impact on the tumour while minimising collateral damage on commensal or even beneficial bacteria. For that reason, there is growing interest in use of bacteriophages which, almost by definition, tend only to target the specific bacteria they have evolved to infect.

Treating cancer would therefore require more specific and targeted antibiotics to maximise impact on the tumour while minimising collateral damage on commensal or even beneficial bacteria.

“Bacteriophages might provide a species‐specific tool, and so might tsRNAs,” said Gilad Bachrach from the Institute of Biomedical and Oral Research, Hebrew University, in Jerusalem, Israel, referring to a study of tsRNAs, which examined the role of small noncoding RNAs (sRNAs) and host transfer RNA–derived small RNAs (tsRNAs) from human saliva (He et al, 2018). The point was that two such tsRNAs have highly similar sequences to some gram‐negative oral bacteria, including F. nucleatum. The study then showed that the presence of F. nucleatum triggered the release of these two tsRNAs by oral keratinocyte cells, which inhibited growth of the bacterium, probably through interference with bacterial protein biosynthesis. Yet, it did not affect growth of Streptococcus mitis, a healthy oral gram‐positive bacterium whose genome does not carry similar sequences to either of these two tsRNAs. In other words, these host‐derived tsRNAs were quite specific in their action, which may offer an alternative therapeutic approach to antibiotics or bacteriophages.

Bachrach also referred to more recent studies that suggested targeting this bacterium could enhance the effect of immunotherapy. One from Wuhan University in China showed that a combination of a specific phage and silver nanoparticles cleared F. nucleatum and activated the host immune system for suppression of CRC, partly by acting against the myeloid‐derived suppressor cells (MDSCs) that the tumour produces to retard the host's anticancer immune response (Dong et al, 2020). Another study, also at Wuhan University, found that inhibiting the growth of F. nucleatum in the gut microbiota helps enhance the effectiveness of first‐line chemotherapy treatments for CRC (Zheng, 2019). However, Bachrach suggested it would make sense to apply these treatments only after verifying from biopsies that the level of these bacteria was indeed elevated.

Crosstalk with gut bacteria

Various other studies have identified that the crosstalk between the gut microbiome and micro‐organisms, especially bacteria, in the tumour microenvironment plays a significant role in the development of tumours and ultimate outcomes. These findings are not confined to CRC or breast cancer where they showed up first, but also hold for a growing number of other solid cancers. A 2019 milestone study showed that the composition of the pancreatic adenocarcinoma microbiome, which engages in crosstalk with the gut microbiome, influences the host immune response as well as natural history of the disease. In particular, the authors performed faecal microbiota transplantation (FMT) from humans into mice and compared Short Term Survivors (STS), Long Term Survivors (LTS) and controls. “Yes, LTS has a unique signature, but we don't know yet if this is cause or consequence,” said lead author Florencia McAllister from the University of Texas, USA. “We are now conducting an FMT trial from healthy patients or survivors into PDAC patients going for surgery (NCT04975217),” she said, adding, “Other groups have confirmed the presence of bacteria but also fungi within pancreatic tumours.”

Apart from its use in research, FMT has also been proposed as a preventative therapy for gut dysbiosis in humans after successful use in ruminating animals such as cattle and sheep and some experimental treatments in humans suffering from Clostridioides difficile infection. But it is an expensive and complex operation to be applied routinely, with attendant risk of side effects, as well as being rather unpredictable. It is therefore unlikely to find routine adoption in the cancer clinic, indicated Ze'ev Ronai, Director of the National Cancer Institute‐Designated Cancer Centre at Sanford Burnham Prebys in California, USA. With reference to the question of when insights into the role of gut microbiota will enter the clinic, he replied, “I suspect it will take another few years. The recipe for bacterial strain will not do it, I am afraid, and faecal transplant is too messy. Hence there is the need to either identify a matter, that is specific metabolites or new prebiotics (meaning dietary fibre) that can mediate the needed changes in the gut microbiota.”

Ronai pointed out that it will be a complex task to refine this into more precise targeted and personalised therapies. “There are three major tasks we are trying to accomplish these days,” said Ronai. “First, to minimise the number of bacterial strains, in order to find out which of the 11 we reported on are the most important to achieve antitumour immunity and inhibit melanoma growth (Li et al, 2019). Second, to determine if we can break prebiotics down to simpler ingredients that can still elicit antitumour immunity, and third to find out mechanistically how altered gut microbiota can elicit antitumour immunity. All these are in the works, too early to report on.”

Potential implications for therapy

Indeed, further progress will be enhanced by identifying more clearly the underlying molecular mechanisms whereby antitumour resistance is mediated by bacteria and associated metabolites. “That remains to be established,” Ronai commented. “Some studies claim for the importance of dendritic cells within the intestine, which we have examined and ruled out, at least as a key driver. I believe it will be collection of different pathways including immune cells, intestinal epithelial cells and specified intestinal cells.”

This would potentially inform further research on characterising microbial profiles associated with different tumours, and understanding how this affects existing treatments, certainly radiotherapy, chemotherapy, and immune therapy. “Our work (Nejman et al, 2020), as well as work by many others, has demonstrated that bacteria are present in all types of solid tumours,” commented Ravid Straussman from the Weizmann Institute of Science, Israel. “There is also growing evidence that tumour bacteria can affect many hallmarks of cancer, including the response to many drugs. I envision that with time we will uncover more and more cases in which tumour bacteria can modulate the response to all types of anticancer therapy, including cytotoxic, targeted, immune and radiation therapy.”

“I envision that with time we will uncover more and more cases in which tumour bacteria can modulate the response to all types of anticancer therapy, including cytotoxic, targeted, immune and radiation therapy.”

Eventually, with more evidence available, research on the role of the microbiome in tumour genesis may help to target therapies more effectively on an individual basis, at least to reduce side effects that can do as much damage as the cancer itself. But there is also hope that a better understanding of how the microbiota affects the immune function and inflammatory processes could inform interventions that improve microbial health and thereby preventing cancers in the first place, or at least delay them until later in life.

EMBO reports (2023) 24: e57040