Colorectal cancer (CRC) is the third most frequent cancer worldwide and a leading cause of cancer mortality, with nearly 2 million new cases and over 900,000 deaths reported in 2020 [1,2]. Despite the well-established roles of genetic predispositions and environmental factors in colorectal tumorigenesis, mounting evidence suggests that perturbations of the intestinal microbiota critically influence the initiation and progression of CRC [3].

As an important constituent of the diverse gut microbial milieu, Bacteroides fragilis holds a dual role, functioning both as a predominant commensal and a potential opportunistic pathogen [4]. While non-toxigenic strains (NTBF) are considered to maintain intestinal homeostasis, the enterotoxigenic variant (ETBF) expresses the B. fragilis toxin (BFT), a metalloprotease capable of disrupting epithelial integrity and inducing chronic inflammation [5]. Notably, ETBF has been detected in both asymptomatic individuals and patients with gastrointestinal diseases, but a growing body of data implicates this pathobiont in the etiology of CRC, particularly through strain-specific attributes and host interactions that exceed the effects of general colonization [[6], [7], [8]]. Indeed, recent findings indicate that mere presence of ETBF is insufficient to explain CRC development, as its prevalence is comparable between CRC patients and healthy controls. This observation underscores the necessity to scrutinize strain heterogeneity, virulence gene profiles, and host susceptibility factors to elucidate the microbiome–cancer axis [9].

Central to ETBF pathogenicity is the production of BFT, a zinc-containing metalloprotease classified into at least three isotypes (bft-1, bft-2, and bft-3), each exhibiting distinct biological activity and pro-inflammatory potential [10,11]. bft gene carriage alone does not predict carcinogenic capacity; rather, isotype distribution and co-expression of accessory genes such as fpn (encoding fragipain), nanH (neuraminidase), and leuB (beta-isopropylmalate dehydrogenase) appear to contribute to ETBF's virulence and persistence in the colonic niche [[12], [13], [14], [15]]. Recent studies suggest that the predominance of certain bft isotypes, particularly bft-2, may serve as a putative biomarker for colorectal cancer (CRC) risk, although this requires further validation in diverse populations. For instance, fragipain (fpn) mediates activation of BFT precursors, while neuraminidase activity facilitates mucosal adhesion and immune evasion. Recent studies employing advanced molecular and phylogenomic analyses have shown that sequence diversity in these genes may correlate with pathogenic potential, clinical outcomes, and epidemiological patterns in CRC 16–18). However, the relationship between specific ETBF isotypes, their virulence repertoire, and CRC remains to be delineated with greater precision, as findings in different populations and study designs have yielded discordant results.

In light of the above, this study aims to investigate the distribution of ETBF virulence determinants—including bft isotypes, fpn, nanH, and leuB genes—in stool samples from patients with colorectal cancer unassociated matched healthy controls. By integrating molecular data with clinicopathological features, we seek to clarify the relevance of strain-specific factors in CRC risk, providing new insight into microbe–host dynamics that may inform targeted diagnostic or therapeutic strategies.