Role of the Microbiome in Malignancy

Abstract

The conceptual underpinning of carcinogenesis has been strongly influenced by an expanded understanding of the human microbiome. Malignancy risks in diverse organs have been uniquely tied to aspects of the resident microbiota in different organs and systems including the colon, lungs, pancreas, ovaries, uterine cervix, and stomach; other organs are increasingly linked to maladaptive aspects of the microbiome as well. In this way, the maladaptive microbiome may be termed an oncobiome. Microbe-driven inflammation, anti-inflammation, and mucosal protection failure, as well as diet-induced microbiome derangement are all mechanisms that influence malignancy risk. Therefore, they also offer potential avenues of diagnostic and therapeutic intervention to modify malignancy risk, and to perhaps interrupt progression toward cancer in different sites. Each of these mechanisms will be explored using colorectal malignancy as a prototype condition to demonstrate the microbiome's role in carcinogenesis.

Knowledge of how the microbiome impacts diverse areas of medicine has recently exploded. In addition to the role of individual organisms, there has been an increasing emphasis on how the entirety of the microbiome contributes to both health and disease. This is largely secondary to the advent, expansion, and affordability of genomic analytic tools that can characterize the composition and function of the human microbiome. These tools include 16s rRNA sequencing, PathoChip^® (Agilent Technologies, Santa Clara, CA, USA) sequencing, immunohistochemistry, and others that represent detailed and high throughput methods of microbiome analysis. Data from each of these analytic techniques are enhancing our understanding of how health and disease is enabled or deranged by the specific composition and function of the multiple human microbiomes.

One exciting area of inquiry relates to understanding how different microbial species interface with the genesis of malignancy. The goal is to discover how microbiome-driven malignancy pathogenesis can lead to prevention, characterization, and treatment of various oncologic processes. Accordingly, the microbiome has been linked in various ways to the development of colorectal, pulmonary, pancreatic, ovarian, cervical, gastric, and other malignancies.^1–5 Clinical practice will be increasingly informed and modified based on newly discovered aspects of the complex relation between the nascent or established tumor and the host microbiome. With regard to malignancy, the microbiome is often termed an oncobiome. This review explores how the host microbiome or specific components directly or indirectly impacts the growth, spread, and treatability of cancers commonly treated by surgeons.

How the Microbiome Contributes to Carcinogenesis

Individual microbes play a role in carcinogenesis. The International Association for Cancer Registries (IARC) identifies 11 microbes (Table 1) that have been identified and studied in great detail as causal organisms of disease.⁶ These organisms are human T-cell leukemia virus 1 (HTLV-1), human papillomavirus (HPV), hepatitis B and hepatitis C virus (HBV and HCV), human immunodeficiency virus (HIV), Epstein-Barr virus (EBV), human herpesvirus 8 (HHV-8), Helicobacter pylori, Opisthorchis viverrini, Clonorchis sinensis, and Schistosoma haematobium. Greater understanding of the host–microbe relation has led to meaningful prevention and treatment measures for individual malignancies. For example, our knowledge of Helicobacter pylori and its contribution to gastric mucosa-associated lymphoid tissue lymphomas has converted what was previously a surgical disease to a subset of malignancies that can be treated primarily with antibiotic therapy. Undoubtedly, more microbes than those listed by the IARC will be shown to contribute to carcinogenesis. Certain bacteria have been identified as having immunomodulatory qualities, although they are not directly linked with development of cancer. These organisms may not be directly responsible alone for the development of cancer but may play a crucial and potentially modifiable role in promoting tumorigenesis.

Table 1.

Individual Microbes and Associated Malignancy

Microbe	Disease
Virus
Hepatitis B virus	Hepatocellular carcinoma
Hepatitis C virus	Hepatocellular carcinoma
Human papilloma virus	Cervical cancer, skin cancer, penile cancer, anal cancer, head and neck
Epstein-Barr virus	Burkitt lymphoma, Hodgkin's lymphoma, nasopharyngeal carcinoma
Human T-cell leukemia virus-1	Adult T-cell leukemia
Human immunodeficiency virus	Post-transplant lymphoproliferative disease, lymphoma
Human herpes virus-8	Kaposi's sarcoma
Helminths
Schistosoma haematobium	Burkitt's lymphoma, Hodgkin's lymphoma, nasopharyngeal carcinoma
Opisthorchis viverrini	Cholangiocarcinoma
Clonorchis sinesis	Cholangiocarcinoma
Bacteria
Helicobacter pylori	Gastric cancer, mucosa-associated lymphoid tissue lymphoma, oral cancer

This table pairs microbes and their associated disease, highlighting the breadth of microbes that are causally linked with malignancy.

There are many proposed mechanisms that may underpin how the microbiome contributes to cancer development throughout the body. Indeed, the microbiome's effect are not limited to gastrointestinal malignancies. Inflammation is well-known to contribute to carcinogenesis, and the microbiome can influence inflammation in a multitude of manners. For example, in colorectal cancer (CRC), chronic inflammation has been durably associated with malignancy in patients with inflammatory bowel disease, and the correlation of the microbiome and inflammation has been shown to impact the development of CRC.^7,8

Inflammation (for example, in ulcerative colitis or Crohn's disease) may trigger alterations in bacterial metabolism, and activate toll-like receptor (TLR)-initiated cytokine signaling, as well as NOD-like receptors (NLRs) involved in the host immune response.^9–11 Subsequent alterations in microbial metabolites also control the phenotypes of tumor somatic mutations and can alter the effectiveness of immune checkpoint inhibitors, at least in murine models.^9,10,12 Additionally, cytokine signaling is altered by the presence of various microbes and can lead to malignancy or failure of malignancy surveillance. Anti-inflammatory cytokines can be suppressed, such as in the case of interleukin (IL)-17 and IL-23 expression, or microbial alterations may lead to increasing secretion of proinflammatory cytokines such as IL-1, IL-6, and tumor necrosis factor-α.¹³ Therefore, both anti-inflammation and inflammation may be maladaptive with regard to carcinogenesis, and both extremes are shaped by microbial activity with the host immune system across endothelial or epithelial surfaces.

Furthermore, microbes play a key role in the mucosal protection of normal epithelium, which is critical in the prevention of carcinogenesis.^14–16 One way in which microbes contribute to this process is through production of compounds from short-chain fatty acids (SCFAs). Short-chain fatty acids are involved in downregulating pro-inflammatory immune factors in multiple ways. When broken down by bacteria in the gastrointestinal tract, SCFAs are involved in downregulating proinflammatory cytokines in the gut through inhibition of histone deacetylases.¹⁷ Byproducts of SCFA breakdown such as butyrate also play a role in modulating T-regulatory cells in the mucosa, creating an anti-inflammatory environment that diminishes tumorigenesis.¹⁷ When the mucosal layer breaks down, these normally protective mechanisms are derailed, and carcinogenesis can occur. An example of this linkage is seen in red meat consumption. Heme in the gastrointestinal tract can be broken down by an abundance of bacteria, leading to hydrogen sulfide production and nitric oxide-induced tissue injury. These toxic byproducts have been shown to cause DNA damage and promote carcinogenesis along the gastrointestinal tract, particularly in CRC.¹⁴

Besides red meat, a high-fat/low-fiber diet (Western diet) is a known risk factor for cancer, in part at least to diet-induced alterations of the normally balanced and diverse microbiome. For example, a Western diet can lead to enrichment of the microbiome with collagenase-producing bacteria, such as Enterococcus faecalis, Pseudomonas aeruginosa, and Serratia marcescens, which in animal studies can promote colorectal tumor formation.^18,19 These organisms appear to preferentially colonize human tumor tissue, and under laboratory conditions can promote a metastatic phenotype in CRC cells.²⁰ These observations challenge the prior notion that metastatic behavior was uniquely linked to tumor cell features and was devoid of external influence.

Beyond inflammation and barrier breakdown, mechanisms of microbial contributions to carcinogenesis include dysbiosis and immune dysregulation among others (Fig. 1). Alterations in composition or activity of commensal bacteria in healthy individuals may predispose to carcinogenesis.²¹ It may also follow that the presence of early cancer-promoting processes within cells may themselves create alterations in the microbiome, which could accelerate carcinogenesis.²² Additionally, cancers at various stages and levels of differentiation are characterized by changes in the microbiome.^2,23 These changes influence the rate of surgical site infection, nutritional deficiency, and even disease recurrence. The vast number of ways which the microbiome may influence carcinogenesis are beyond the scope of this review; however, it seems clear that the microbiome may influence the progression of malignancy.

FIG. 1.

This figure demonstrates the complex interplay of microbiome elements with different aspects of the immune system that may maladaptively lead to malignancy genesis. This dynamic is focused on colorectal cancer but has broad applicability to epithelial malignancies in general.

Colorectal Cancer and the Microbiome

Although evidence for a critical role of the microbiome in an increasing number of malignancies continues to generate interest and enthusiasm, the most extensively studied relation between a cancer and the microbiome is that of CRC. Indeed, the microbiome has been implicated along the entire progression of CRC including cancer initiation, surgical treatment, and response to adjuvant therapies. A recent review and an investigative study assessed the impact of appendectomy on the gastrointestinal tract microbiome.^24,25 Altered gut microbiome composition was noted with enrichment of maladaptive bacteria and reductions in beneficial species, which are elements that create a dysbiome. Moreover, in the population-based longitudinal study, CRC risk substantially increased after appendectomy.²⁵ Although there are likely a number of other factors at work with respect to CRC tumorigenesis, it is intriguing to consider that one of the most common operations performed worldwide may lead to unanticipated dysbiosis. Accordingly, these observations may drive re-evaluation of CRC screening recommendations for patients who have undergone appendectomy.

First, gut bacteria can play a role in CRC initiation. One example of this role is as part of the driver-passenger bacteria model.²⁶ Driver bacteria such as Escherichia coli and Bacteroides fragilis promote inflammation and have been shown to have a significant impact on the development of CRC.^27,28 Enzymes produced by Escherichia coli and Bacteroides fragilis increase the production of reactive oxygen species and create DNA damage in normal colonic epithelium. These bacteria promote a microenvironment that is amenable to the growth of passenger bacteria such as Fusobacterium nucleatum. These passenger bacteria promote further inflammation through myeloid cell invasion and may contribute to the progression of CRC from early to more advanced stages.²⁹

Beyond tumorigenesis, microbiota have been found to play a role in the local recurrence and metastasis of cancer, especially CRC. Recurrence after surgery may be related to shed tumor cells within the gastrointestinal tract. After primary resection, tumor cells that are shed can be influenced by the microbiome to invade locally as well as metastasize.¹⁸ Specifically, an imbalance of commensal organisms can promote metastasis in CRC, and dysbiosis is well characterized by a decrease in microbiome diversity. Escherichia coli-mediated secretion of substances such as cathepsin K, and increased levels of lipopolysaccharide (LPS) has been shown to lead to increases in metastasis.³⁰ This untoward effect is mediated through the action of TLR4, which decreases innate immune response by polarizing macrophages to M2 differentiation.³⁰ Xu et al.³¹ have recently demonstrated that Fusobacterium nucleatum similarly causes M2 differentiation via the miRNA-1322/CCL20 pathway. Differentiated M2 macrophages secrete transforming growth factor-β, IL-10, and decrease the innate immune response to tumor cells. The impaired surveillance of tumor cells thereby promotes metastasis.³¹

Because the Western diet is implicated in tumorigenesis by creating a dysbiome, it is also reasonable to examine the influence of the dysbiome on tumor recurrence after resection. In a murine model of CTC resection, when organisms promoted by a Western diet (Enterococcus faecalis, Pseudomonas aeruginosa, and Serratia marcescens) predominate on anastomotic tissue, they appear to induce tumor recurrence.¹⁸ Strikingly, these same organisms appear to preferentially colonize human tumor tissue, and under laboratory conditions, can promote a metastatic phenotype in CRC cells.²⁰ The mechanisms by which these microbes promote recurrence is likely multifactorial.

First, these organisms, specifically, Enterococcus faecalis can activate the urokinase → pro-urokinase → plasminogen pathway, a pathway that is well-known to promote tumorigenesis, and perhaps by extension, recurrence.^20,32 Relatedly, activation of certain genomic, transcriptomic, or metabolomic pathways portends a poor prognosis related to metastasis.²² Second, Enterococcus faecalis through both collagenase-dependent and collagenase-independent pathways may directly promote cancer cell invasion and migration.²⁰ Third, these specific bacteria can impair the anastomotic healing, resulting in porosity that can enable extraluminal escape of intraluminal cancer cells.³³ Together, this body of literature suggests a microbial role for the development of recurrence, and presents an intriguing target for future therapeutic endeavor. For example, multiple studies have demonstrated that mechanical and oral antibiotic bowel preparation leads to a decrease in local recurrence of CRC. Patients who underwent surgery for locally advanced CRC after oral and mechanical bowel preparation had increased five-year survival compared with patients who did not undergo bowel preparation (76.3% vs. 64.2%, respectively).^34,35

Finally, microbiota may influence the selection of adjuvant therapies prior to, or after, surgical resection. In a small study of 84 patients at a single institution, patients with rectal cancer with microbiomes predominantly characterized by Dorea and Anaerostipes bacteria were more likely to respond to neoadjuvant chemoradiotherapy.³⁶ These bacteria produce butyrate that, as previously described, leads to a decrease in inflammation, whereas patients with Coriobacteriaceae and Fusobacterium were present in greater numbers in non-responders.³⁶ Response to treatment as well as associated drug toxicity in patients with CRC may be influenced by the composition of microbiota.^36,37 Although data are limited, these findings suggest a future role of microbiome characterization as an adjunct to CRC treatment and suggests potential therapeutic applications.

Future Directions

How the human microbiome interfaces with different cancers may vary widely based on the specific malignancy and the environment within which the organ–microbiome dynamic occurs. Moreover, there is wide variation in how microbiome samples are collected, processed, and characterized.³⁸ In CRC, fecal samples can be obtained non-invasively and easily, whereas in lung cancer the microbiome can only be accurately sampled reliably via bronchoalveolar lavage or via transbronchial biopsy, both relatively invasive procedures.¹³ Alternatively, and perhaps in the near future, advances in sampling technology may be able to detect cancer via characteristic alterations in the host microbiome. The long-awaited technology that can readily identify bacterial, fungal, or viral DNA or RNA in blood samples is emerging. In a related fashion, cell-free DNA detection has been used to identify metastatic melanoma.³⁹ Similar approaches can be potentially used to serially map changes in sampleable microbiota to detect early malignancy. Opportunities abound for standardizing the collection and characterization of the microbiome in a consistent and patient-centered manner.

As the field progresses, we may be able to determine the appropriate timing of surgery better by characterizing changes in the microbiome, especially during pre-habilitation for re-sectional therapy. Pre- and post-operative interventions may be shaped by the predominance—or under-representation—of certain bacteria, potentially tailoring antibiotic intestinal preparation, dietary intake, and ultimately, deliberate, and adaptive microbiome manipulation to support healing as well as tumor suppression. It remains unclear how to leverage microbiome alteration in conjunction with non-operative therapies that may ultimately mitigate against needing surgical therapy for certain malignant processes. Each of these above approaches relies on augmenting new knowledge discoveries to enable patient-focused precision medicine.

Conclusions

The microbiome may be a key contributor that impacts the diagnosis and treatment of malignancy. Similar to our current understanding of the normal microbiome during health, the deranged microbiome—or dysbiome—and its influence of cancer diagnosis, prognosis, treatment, and surveillance is undergoing explosive growth. Colorectal cancer serves as an exemplar of how the microbiome and malignancy are related and drives the surgeon to consider additional elements of both pre-operative preparation and post-operative management that are not solely related to isolated tumor biology.

Authors' Contributions

Conceptualization: Hyman. Data research and interpretation: all authors. Formal analysis: all authors. Data curation: Godley, Shogan. Writing–original draft: Godley, Shogan. Writing–review and editing: Hyman. All authors approved the final version.

Funding Information

No funding was received.

Author Disclosure Statement

No competing financial interests exist.