Carriers of colibactin-producing E. coli have a three-fold increased risk of developing colorectal cancer

A study by Japanese oncologists and gastroenterologists was published in eGastroenterology: in people with familial adenomatous polyposis (FAP), E. coli carrying the pks islet and producing the genotoxin colibactin were searched for in polyps samples. It turned out that in patients who had already been diagnosed with colorectal cancer, the presence of such bacteria in polyps was three times more common than in patients without a history of cancer. This supports the idea that individual microorganisms of the microbiome can accelerate carcinogenesis in the genetically vulnerable colon.

Background of the study

Familial adenomatous polyposis (FAP) is a hereditary syndrome due to APC mutations, in which the colon is strewn with adenomas and the risk of colorectal cancer (CRC) is extremely high already at a young age. Even with active endoscopic monitoring and drug prevention, some patients progress rapidly, which prompts the search for additional "accelerators" of carcinogenesis, including among the components of the intestinal microbiome.

One such candidate has long been considered to be colibactin, a genotoxin of Escherichia coli strains carrying the pks islet. In models and clinical series, it causes DNA damage, characteristic cell cycle delays, and forms a recognizable mutational signature SBS88/ID18 in colon epithelium; in some patients with sporadic CRC, the presence of this signature distinguishes a specific molecular subtype of the tumor. This makes pks+ E. coli a potential risk modifier, especially where the genetic background is already “prepared” for tumor growth.

However, data on hereditary syndromes have been fragmentary: most studies have examined sporadic CRC rather than precancerous conditions such as FAP; they have often used stool samples or “mixed” tissues, making it difficult to link the bacteria specifically to polyps; and tissue markers of DNA damage (e.g., γ-H2AX) and inflammation have rarely been assessed in parallel to get closer to the mechanism. Therefore, the question remains whether pks+ E. coli is more common in polyps from patients with FAP and whether it is associated with clinical features of higher oncological risk.

A new study in eGastroenterology addresses this gap: in a cohort of patients with FAP, mostly without prior colorectal surgery, the researchers looked for pks+ E. coli in polyps and correlated its presence with CRC history and tissue markers of DNA damage/inflammation. This design allows us to assess not only the prevalence of a putative risk factor, but also its biological footprint in the very target of carcinogenesis—colonic adenomas in FAP.

Context: Why colibactin and FAP

Colibactin-producing E. coli (pks+ E. coli ) have already been found in ~67% of patients with sporadic colorectal cancer and about 21% of healthy people; in experiments, the toxin causes DNA damage (γ-H2AX), cell cycle arrest, and accelerates tumorigenesis. In FAP, a hereditary condition associated with an APC mutation, the gut is littered with adenomas, and any “addition” to the risk is especially critical. The new work studies not after gut removal, but in patients with a preserved colon - that is, in the most “natural” environment for the microbiota.

How it was done

From January 2018 to August 2019, polyp and mucosal samples were collected from 75 patients with FAP during endoscopy and tested for pks+ E. coli. In parallel, clinical factors were assessed and immunohistochemistry was performed for DNA damage (γ-H2AX) and inflammation markers (IL-6, IL-1β). Patients who did not undergo colorectal surgery were compared separately to exclude the effects of surgery on microbiota.

Main results

In non-operated patients with FAP, carriage of pks+ E. coli in polyps was significantly more common in those who had previously had colon cancer: odds ratio 3.25 (95% CI 1.34-7.91). In polyps with pks+ bacteria, γ-H2AX (a sign of DNA damage) was more strongly stained, and IL-6 tended to increase; IL-1β did not change significantly. In smokers, pks+ E. coli was more common, while gender, age and alcohol did not show a significant association. It is noteworthy that in patients after colon surgery, pks+ bacteria were not detected in polyps - an indirect hint at how much the operation changes the microbial "field".

What is important to remember (in two steps)

• There is a link, but causality is not proven: the study is associative and designed to generate a hypothesis. Large multicenter cohorts and longitudinal observations are needed.
• The biomarkers of colibactin “footprint” are clear: γ-H2AX and the inflammatory signal (IL-6) were increased in pks+ polyps - mechanistically, this fits into the picture of colibactin-induced genomic instability.

Why this matters for patients with hereditary risk

FAP is a rare but serious condition: adenomas appear in dozens and hundreds, and the risk of cancer is high already at a young age. If part of this risk is “fueled” by a specific bacterium, new levers of prevention appear. In the study, the authors emphasize that pks+ E. coli was not associated with the “density” of polyps (the severity of FAP) - that is, it is more likely that we are talking about a qualitative accelerator of carcinogenesis, and not just a “satellite” of multiple adenomas.

What could this mean in practice (hypothetically for now)

• Microbial risk screening: searching for pks+ E. coli in biopsies/stool as part of surveillance of patients with FAP.
• Spot prevention of microbiota: targeting colibactin (bacteriophages, selective antibiotics, probiotics/postbiotics) - only after clinical trials.
• Response markers: monitoring γ-H2AX, IL-6 as indicators of microbial-induced stress during interventions.
• Behavioural factors: Smoking cessation appears particularly relevant given the higher incidence of pks+ in smokers.

Limitations that the authors themselves honestly stated

A small sample and a single center limit statistical power; not all lifestyle factors (e.g. diet) are taken into account; biased selection of polyps is possible; some patients did not have genetic verification due to legal restrictions. The authors separately note that confirmation is needed on external cohorts and a search for the colibactin "signature" in the mutational profile (SBS88) - this would help move from the association to more confident conclusions about the contribution of the toxin.

What's next?

A logical next step is multicenter studies before/after interventions (polypectomy, microbiota sanitization), integration of microbiome testing with clinical and molecular markers, and testing whether eradication of pks+ E. coli reduces the real risk of cancer in people with FAP. If the hypothesis is confirmed, we will have a rare example of how a specific microbial factor can be targeted for cancer prevention in a hereditary syndrome.

Source: Ishikawa H., Aoki R., Mutoh M., et al. Contribution of colibactin-producing Escherichia coli to colonic carcinogenesis. eGastroenterology. 2025;3(2):e100177. https://doi.org/10.1136/egastro-2024-100177