Anastomotic leakage is a major complication following colorectal surgery leading to peritonitis, complications, and mortality. Akkermansia muciniphila has shown beneficial effects on the gut barrier function. Whether A. muciniphila reduces peritonitis and mortality during colonic leakage is unknown. Whether A. muciniphila can directly modulate the expression of genes in the colonic mucosa in humans has never been studied. We investigated the effects of a pretreatment (14 days) with live A. muciniphila prior to surgical colonic perforation on peritonitis, mortality, and wound healing. We used mice with an inducible intestinal-epithelial-cell-specific deletion of MyD88 (IEC-MyD88 KO) to investigate the role of the innate immune system in this context. In a proof-of-concept pilot study, healthy humans were exposed to A. muciniphila for 2 h and colonic biopsies taken before and after colonic instillation for transcriptomic analysis. Seven days after colonic perforation, A.-muciniphila-treated mice had significantly lower mortality and severity of peritonitis. This effect was associated with significant improvements of wound histological healing scores, higher production of IL22, but no changes in the mucus layer thickness or genes involved in cell renewal, proliferation, or differentiation. All these effects were abolished in IEC-MyD88 KO mice. Finally, human subjects exposed to A. muciniphila exhibited an increased level of the bacterium at the mucus level 2 h after instillation and significant changes in the expression of different genes involved in the regulation of cell cycling, gene transcription, immunity, and inflammation in their colonic mucosa. A. muciniphila improves wound healing during transmural colonic wall defect through mechanisms possibly involving IL22 signaling and requiring MyD88 in the intestinal cells. In healthy humans, colonic administration of A. muciniphila is well tolerated and changes the expression of genes involved in the immune pathways.

Hepatocellular cancer (HCC) is one of the main reasons for liver transplantation (LT). Biomarkers, such as alpha-foetoprotein (AFP) and Des-gamma-carboxy-prothrombin (DCP), can be helpful in defining the recurrence risk post LT. This study aims to evaluate the association between the intensity of DCP immunohistochemical labelling and serum DCP levels in patients undergoing LT for HCC. We carried out a prospective monocentric study including patients who all underwent LT for cirrhosis between 2016 and 2018 and all fell under the Milan criteria. The accepted diagnostic criteria for HCC were contrast-enhanced imaging and histology. Thirty-nine patients were followed for a median of 21 months, with HCC lesions categorized into negative, focally positive, and diffusely positive groups based on DCP immunohistochemistry. The serum DCP levels were significantly higher in the positive groups (258 mAU/mL for the focally and 257 mAU/mL for the diffusely positive) than in the negative group (48 mAU/mL) (p = 0.005) at diagnosis and at the time of liver transplantation (220 mAU/mL for the diffuse positive group). Microvascular invasion (58.8% vs. 19.0% for the diffusely positive and negative groups, respectively, p < 0.001) and lesion size (20 mm in the diffusely labelled group versus 12 mm in the other groups, p = 0.002) were significantly correlated with DCP labelling. Late recurrence occurred only in the positive groups; in the negative group, it occurred within the first 3 months after transplantation. DCP labelling in liver lesions correlates with serum levels and a more aggressive tumour profile. Further investigation is needed to determine if highly DCP-labelled tumours allow for the better selection of high-risk patients before LT.

Villous morphology is routinely used to stratify risk in conventional colorectal adenomas, yet it remains largely interpreted as a visual histological descriptor. This review proposes a different view: villous architecture is a graded, spatially organized epithelial state that links premalignant tissue remodeling to malignant potential. Across histological, molecular, and microenvironmental studies, villous-containing adenomas show convergent features of WNT-driven stem-like expansion, YAP-associated plasticity, fetal-regenerative or fetal-metaplastic transcriptional programs, mucin remodeling, stromal activation, immune attenuation, and genomic instability. These features appear to arise through interacting routes rather than through a single deterministic mutation. This framework helps explain why villous proportion predicts risk imperfectly: it captures a composite biological state but does not resolve its individual components. Quantitative morphometry, spatial transcriptomics, multiplex imaging, and region-resolved genomics could therefore move villous architecture from an observer-dependent category to a continuous, biologically interpretable marker of early malignant transition. Such a shift may improve adenoma classification, refine surveillance strategies, and identify high-risk lesions before conventional thresholds for advanced histology are reached.