Journal of Crohn's and Colitis

BACKGROUND & OBJECTIVE Tofacitinib, a Janus kinase (JAK) inhibitor, is approved for ulcerative colitis (UC). We investigated baseline and longitudinal transcriptomics and proteomics in intestinal biopsies, peripheral blood, and serum from UC patients starting tofacitinib to characterise treatment-associated effects and identify which molecular compartment harbours the most predictive response-associated biomarkers. METHODS In this prospective, open-label study, patients with moderately to severely active UC started tofacitinib 10 mg BID for eight weeks. Clinical, biochemical, endoscopic, and histological data were collected at baseline and week 8. The primary outcome, histo-endoscopic mucosal improvement (HEMI), was defined as endoscopic Mayo subscore [≤]1 with histologic remission (Robarts Histopathology Index [≤]3, no mucosal neutrophils). Mucosal biopsies and peripheral blood were collected at baseline and week 8 for bulk RNA-sequencing (RNA-seq); biopsies and serum underwent proteomic profiling using Olink Explore 384 (Cardiometabolic, Inflammation, Neurology, Oncology panels). RESULTS In total, 127 paired multi-omic samples were collected longitudinally from 40 UC patients: 15 (38%) achieved HEMI at week 8 (responders), 13 (32.5%) were late responders, and 12 (30%) were non-responders. Pretreatment mucosal biopsies from future responders showed elevated interferon, IL-6-family, IL-2/common-{gamma}-chain, and T-cell-receptor signalling gene expression, which declined on treatment only in responders. A responder-specific JAK-STAT-dependent mechanism spanned interferon, cytokine-receptor, T-cell-receptor, and NF-{kappa}B signalling, with concordant gene- and protein-level reductions. Compared with a public infliximab dataset, JAK-STAT/interferon downregulation was shared and response-independent, while non-canonical NF-{kappa}B downregulation was tofacitinib-specific. In blood, shared pathway labels reflected distinct genes from mucosal effectors. CONCLUSION Response to tofacitinib depends on a JAK-tractable mucosal pathway absent in non-responders, suggesting a pretreatment biomarker for predicting treatment response.

Background There are currently no available biomarkers that predict the clinical outcome of ulcerative colitis (UC) patients after anti-TNF withdrawal. Here we developed a semi-supervised machine learning framework on RNA-seq data derived from rectal biopsies to predict relapse following anti-TNF discontinuation. Methods We included UC patients (n=136) enrolled in the prospective, randomized controlled BIOSTOP trial. Patients were assigned to a withdrawal group (stopped treatment immediately) or a maintenance group (stopped treatment after 2 years). A machine learning framework was trained on gene expression data from mucosal biopsies collected at baseline from the withdrawal group and validated in the maintenance group using baseline samples (test cohort) and 2-year follow-up samples (interference cohort). Results A semi-supervised machine learning model, combining a shared UMAP manifold representation with LabelSpreading on a targeted 38-gene panel achieved an AUC of 0.794, a sensitivity of 90.3%, and a specificity of 73.1% on the test cohort, and an AUC of 0.711, a sensitivity of 82.9%, and a specificity of 42.3% on the inference cohort. Applying Shapley Additive exPlanations (SHAP) we found that IL1B, TYROBP, REG1B, S100A8, and TNF showed the highest predictive importance for relapse risk, whereas MUC2, HMGCS2, and CDKN1A displayed the highest predictive importance for sustained remission. Conclusion The high sensitivity makes the model a promising screening tool to prevent withdrawal-related relapse, though its moderate specificity underscores the need for multi-center validation in larger, more diverse cohorts before clinical deployment.

Using an analysis pipeline that combines machine learning-based feature selection with classical statistical modelling, we aimed to predict incident inflammatory bowel disease (IBD) and to identify candidate proteins associated with disease development. We used UK Biobank data in an observational cohort study of 52,331 participants with proteomics data and without a history of IBD. The cohort was split into an 80/20 train/test dataset, and three separate extreme gradient boosting models were trained on the training dataset. Using Shapley values, we ranked the predictive importance of 2,912 proteins for incident IBD, Crohn's disease (CD), and ulcerative colitis (UC). The top 20 proteins per outcome served as features in multivariate Cox models predicting 350 incident cases over 14 years. All three proteomics Cox models showed moderate discrimination in the test dataset, with the UC model performing the best (C-index: 0.68), followed by IBD (C-index: 0.65), and CD (C-index: 0.60). All proteomics models outperformed clinical models. Smoking remained the only significant clinical predictor in the combined proteomics + clinical model. 48 unique proteins were identified across the three models, with 16 reaching statistical significance (p< 0.05). Machine learning-based feature selection with statistical survival modelling identifies plasma proteins that predict incident IBD and highlight biologically plausible, potentially targetable pathways in the preclinical phase of IBD.

Background and aimsCrohns disease (CD) is a highly heterogeneous disease. Patients with similar characteristics at baseline may end up on very different therapeutic and surgical paths. We aimed to identify disease-course trajectory phenotypes from a mix of clinical, treatment, surgical and comorbidity variables, and evaluate whether information already available at diagnosis could anticipate which trajectory a patient would follow. MethodsWe analysed a retrospective cohort of 98 unique CD patients enrolled at the IRCCS Ca Granda Policlinico Hospital, Milan. Factor analysis of mixed data (FAMD) was applied to 30 disease-course variables spanning demographic and Montreal descriptors, extra-intestinal manifestation burden, individual advanced-therapy drug exposures, surgical burden and comorbidity/symptom burden; clusters were then ranked by treatment burden using a line-weighted therapy score built separately from drug line positions and therapeutic-class weights. Clustering itself was carried out in the retained FAMD coordinate space, and cluster robustness was checked with silhouette metrics and bootstrap Jaccard stability. Baseline-associated variables were assessed separately, while advanced-therapy episodes were reconstructed from the therapy timeline. ResultsThree trajectory phenotypes emerged: C1 (n=56), C2 (n=31) and C3 (n=11). C3 was characterised by younger age at diagnosis, a heavier extra-intestinal manifestation burden, more intensive and prolonged therapy, greater biologic exposure and a higher cumulative number of surgeries, with comorbidity burden rising across the gradient. Montreal behaviour and comorbidity burden emerged as the strongest baseline correlate of trajectory membership, ahead of age at diagnosis and extra-intestinal manifestation burden; penetrating (B3) behaviour, notably, was actually more common in C1 than in C3. A model built on the diagnosis-available variables independently linked to trajectory membership (age at diagnosis, disease location and Montreal behaviour) showed moderate discrimination for the C3 trajectory (area under the curve, AUC=0.80, 95% CI 0.67-0.93). A sensitivity analysis that also added extra-intestinal manifestation count and comorbidity burden, whose timing relative to diagnosis we could not confirm, pushed discrimination to AUC=0.90. ConclusionsFAMD-space clustering identified stable, clinically interpretable CD-course trajectory phenotypes, including a small, high-burden C3 subgroup. This C3 trajectory, small, early-onset and treatment-refractory, shares clinical features that overlap with the difficult-to-treat CD phenotype, though C3 was derived by unsupervised clustering rather than validated against that frameworks own criteria; variables available at or near diagnosis showed exploratory discrimination for the high-burden C3 trajectory, though external validation and careful timing verification are needed before any clinical use.

Background: Inflammatory bowel disease (IBD) has been linked to a two- to eightfold increase in colorectal cancer (CRC) risk. However, particular genes that bridge intestinal dysbiosis and oncogenic transcriptional reprogramming remain unclear. The Hippo/YAP pathway is a major regulator of intestinal epithelial homeostasis whose effector YAP is hyperactivated in CRC, but its transcriptional engagement in IBD mucosa and its link to specific cancer-risk genes has not been thoroughly mapped. Methods: Bulk RNA-seq data were integrated from two independent cohorts: GSE235236 (IBD colon mucosa; UC/CD vs. healthy controls; n = 56) and TCGA-COAD/READ (CRC tumors stratified by YAP activity; n = 592). Differential expression analysis was conducted separately within each cohort using Welch's t-test with Benjamini-Hochberg FDR correction. YAP transcriptional activity was estimated per sample as the mean row-wise Z-score of seven canonical YAP target genes, and TCGA tumors were stratified into Hippo-High (Q4) and Hippo-Low (Q1) groups. Genes showing consistent dysregulation across both cohorts were identified and ranked to define a Dysbiosis-Hippo Response (DHR) module, which was further organized into functional gene archetypes. Results: IBD mucosa showed 2,848 significantly dysregulated genes (FDR < 0.05). Comparison with the CRC cohort identified 120 genes showing concordant dysregulation, forming the DHR module. These genes were significantly enriched in TNF-alpha signaling via NF-kB (adjusted p = 0.006), inflammatory response (adjusted p = 0.013), and Myogenesis (adjusted p = 0.013). DHR gene classification revealed three functional archetypes: (i) innate immune activation (SPI1, OSM, IL11, LILRA2, SLC11A1) reflecting convergent myeloid reprogramming; (ii) epithelial-mesenchymal transition (TWIST1, SNAI1, PRRX2, MMP9) priming invasive potential before malignancy; and (iii) tumor microenvironment immune regulation (TN-FRSF4, TNFRSF18, PDCD1, VASN, SPON2) shaping immune tolerance. Interestingly, the canonical YAP target genes (CTGF, CYR61, ANKRD1, AMOTL2, and FSTL1), sharply upregulated in Hippo-High CRC, were essentially unchanged in IBD mucosa. This signifies that the shared DHR programme operates through non-canonical YAP effectors. And yet, YAP target scores are positively correlated with DHR module expression across IBD samples (Spearman rho = 0.726, p = 2.5 x 10-10). Conclusions: Our findings define a three-archetype DHR module that links gut dysbiosis to YAP-driven oncogenesis at single-gene resolution. The programme shared between IBD and CRC is driven by innate immune activation, EMT priming, and remodeling of the immune microenvironment, not by the canonical CTGF/CYR61 YAP target axis, which appears only at the CRC stage. This gene-level framework provides tractable mechanistic targets for chemoprevention and biomarker development at the inflammation-to-cancer transition.

Background and Aims: As a microbiome/diet-derived vitamin, queuine (q) is a micronutrient that eukaryotes acquire from microbiota or diet. Vitamin q, the free base form of queuosine (Q), is used for Q-modification of transfer RNA (tRNA). Although tRNA-Q-modification in regulating eukaryotic cell physiology is recognized, the relationship between Q-tRNA-modification and the human microbiome in health and inflammatory bowel diseases (IBD) remains unknown. Which bacteria produce q and how the q-producing bacteria contributing to human health and IBD are not explored. Here, we hypothesize that disrupted microbial Q-biosynthesis contributes to the dysbiosis of human IBD. We performed comprehensive multi-omics integrative analyses using publicly available datasets. Specifically, we re-analyzed publicly available shotgun metagenomic, single-cell RNA sequencing (scRNA-seq), metabolic datasets, and completed meta-analysis. Methods: To systematically investigate microbial Q-biosynthesis in healthy individuals and IBD patients, we performed a systematic secondary analysis by integrating 9 publicly available metagenomic datasets, 15 scRNA-seq datasets and one metabolic dataset published prior to 2026. We present a novel secondary bioinformatic reanalysis of sequence outputs mapped to the latest microbial taxonomy, functional pathways and microbial Q-related genes. Longitudinal study of shotgun metagenomics across the human lifespan (3146 samples) enabled taxonomic and functional profiling. Results: We first identified key q-producing bacteria (e.g., Bacteroides and Alistipes) in healthy populations. We then found that patients with IBD had significantly reducation of q-producing bacteria and a coordinated downregulation of the Q-related pathways, including Q, preQ0, de novo guanosine ribonucleotide biosynthesis and de novo guanosine deoxyribonucleotide biosynthesis pathways. This functional impairment was primarily driven by reduced contributions from Bacteroides, Alistipes and Akkermansia. The longitudinal stability of the q-producing bacteria community across the human lifespan (1638 samples). We observed a significant negative correlation between host age and the abundance of q-producing bacteria in patients with UC. Moreover, Queuine tRNA-ribosyltransferase 1 (QTRT1) expression in human intestinal epithelial cells was significantly reduced in patients with IBD compared to healthy controls. Conclusions: Our study used multi-omics datasets to define a novel functional dysbiosis characterized by a reduction of microbial Q-biosynthesis in patients with IBD. Vitamin q and its tRNA modification is lesser-recognized in human IBD. Understanding the cellular and organismal mechanisms of this microbiome-dependent micronutrient will advance the prevention of IBD and improve the quality of life of patients with IBD.

Background Endoscopic resection is used for the management of dysplasia in ulcerative colitis (UC), but a subset of patients develop subsequent advanced neoplasia (AN), i.e., high-grade dysplasia and/or colorectal cancer, despite apparently complete resection with histologically clear margins. Current risk stratification relies predominantly on histopathological and procedural features, which does not capture underlying genomic field cancerization of morphologically normal mucosa. Objective To determine whether genomic alterations within histologically non-dysplastic resection margins are associated with subsequent neoplastic progression following endoscopic resection of UC-associated dysplasia. Design We analysed a retrospective cohort of 51 UC patients, who underwent endoscopic resection of dysplastic lesions at a tertiary referral centre, with a median follow-up of 1947 days. 11 of these had local progression. We performed low-coverage whole genome sequencing (lcWGS) of 56 dysplastic lesions and 91 adjacent histologically non-dysplastic margins. Somatic copy number alteration (CNA) burden and phylogenetic analysis were correlated with progression to AN . Results Histologically non-dysplastic margins from lesions that subsequently progressed to AN demonstrated significantly higher CNA burden compared with non-progressors. In multivariate analysis high margin CNA burden was independently associated with local progression, whereas histologically clear resection margins and en bloc excisions were not. Conclusion Histological clearance does not necessarily indicate clearance of the cancerised field in UC-associated dysplasia. Genomic detection of field cancerisation within histologically non-dysplastic resection margins identifies patients at increased risk of subsequent neoplasia and provides a potential biomarker for biologically informed post-resection surveillance.

Background and aimsPerianal fistulizing Crohns disease (pCD) affects 20% of patients with Crohns disease (CD) and severely impacts quality of life. Current therapies fail to provide sustained relief, and the molecular underpinnings of pCD remain poorly understood. This study aimed to elucidate the molecular landscape of perianal fistulas through multiomic profiling. MethodsPaired fistula-tract and adjacent rectal mucosal biopsies were collected from 63 patients (48 with pCD, 15 with cryptoglandular fistula [CPTGL]). Longitudinal sampling generated 101 unique molecular profiles, comprising of both RNA sequencing (RNA-seq) and 16S rRNA sequencing from both tissue sites, followed by integrative multiomics and gene network analyses. ResultsUnsupervised clustering revealed three patient clusters primarily defined by host gene expression, with minimal contribution from microbial profiles. Fistulae in clusters 1 and 2 showed strong immune activation and epithelial-mesenchymal transition (EMT). In contrast, cluster 3 fistulae displayed keratinization and metabolic reprogramming resembling psoriatic skin, together with reduced JAK-STAT signalling. Cluster 3 was enriched for patients classified as TOpClass:2a, who are more suitable for surgical repair (p = 0.02). Conversely, cluster 2, characterized by rectal keratinization and EMT in both fistula and rectum, showed the highest MRI inflammatory-mass score (p = 0.02) and a greater risk of subsequent ileostomy (Kaplan-Meier; p = 0.008). An independent RNAseq dataset validated the keratinization signature in a subset of pCD fistulae. ConclusionIntegrated multiomic analysis identified distinct molecular subtypes of pCD with surgical and therapeutic relevance. These findings refine the molecular understanding of pCD and support a precision medicine approach. What You Need to Know?O_ST_ABSBACKGROUND AND CONTEXTC_ST_ABSO_LIPerianal fistulas affect 1 in 5 Crohns disease patients and have a severe negative impact on the quality of life of the patients. C_LIO_LIMost advanced IBD therapies are not efficacious for perianal Crohns disease. C_LIO_LISurgical repair is not suitable for all patients. C_LI NEW FINDINGSO_LIThis study is the first to delineate molecularly defined patient subtypes in perianal fistulizing Crohns disease. It further characterizes a psoriasis-like keratinization program in a subset of fistulas. Finally, it identifies molecular features and histological indicators that may explain-and potentially help predict-favorable surgical outcomes in selected patients C_LI LIMITATIONSO_LIFunctional and mechanistic studies will be required to further dissect the drivers and dynamics of the epithelial remodeling identified. C_LI CLINICAL RESEARCH RELEVANCEO_LIRaises the possibility of refining of existing clinical stratification using molecular markers for improved therapeutic management and outcome. C_LI BASIC RESEARCH RELEVANCEO_LIProposes keratinization as an opposing molecular process to EMT within the perianal fistula tract. C_LIO_LIIdentifies robust gene signatures associated with EMT and keratinization in the fistula for further experimental and clinical studies. C_LI

Background and aims. Inflammatory Bowel Disease (IBD) is a complex immune-mediated condition with a strong genetic contribution. The genetics of IBD are highly complex, with inflammation, autoimmunity, and mucosal immunity pathways involved in its pathogenicity. GenePy is a gene-level disease burden metric. The integration of clinical and genetic data and the application of machine learning methods can aid patient stratification for personalised care. Methods. Lists of consensus IBD pathway genes were curated. GenePy scores were computed from whole exome sequencing data. A K-Means algorithm was optimised using clustering quality and stability metrics on the GenePy scores of genes in each IBD pathway, and the results replicated in three independent IBD cohorts (SPARC IBD, MSCCR IBD, and Southampton IBD). Chi-square tests assessed the association between cluster assignment and phenotypic features, and were multiple test corrected via Benjamini-Hochberg FDR. Results. A strong association was replicated for NOD signalling genes driving IBD subtype, Crohns disease behaviour, and IBD surgery across all cohorts, as well as granuloma in Southampton IBD. This was mainly driven by NOD2. Similar genetic clustering patterns were observed across all cohorts, with additional phenotypic enrichment in clusters driven by NLRP1 (surgery in MSCCR IBD), and nominally significant enrichment for disease behaviour linked to BIRC2 (B2 in Southampton IBD and B2B3 in SPARC IBD). Conclusions. This analysis identified a common genetic architecture across NOD signalling genes across IBD cohorts, replicated a signal for NOD2 as a monogenic driver of disease with a strong phenotypic enrichment, and identified other putative genetic drivers of IBD, such as the NLRP1 gene.

Background and Aims: Fibrostenotic stricture is a severe complication of Crohn's disease (CD) that affect up to 50% of patients over their lifetime, causing irreversible narrowing of the intestinal lumen and bowel obstruction, but the molecular programs that distinguish stenotic from non-stenotic inflammation and the relative contributions of fibrosis and smooth-muscle remodeling remain incompletely characterized. We aimed to dissect bowel-wall compartment-specific and histology-defined remodeling programs in fibrostenotic CD. Methods: Matched full-thickness terminal ileal tissues from 4 patients with fibrostenotic CD were sampled from stenotic stricture (SS), non-stenotic inflamed bowel (NS), and non-inflamed bowel (NI). Histomorphology-guided NanoString GeoMx whole-transcriptome profiling was performed across mucosa, muscularis mucosae, submucosa, and muscularis propria. GeoMx data were processed by standard software provided by GeoMx platform. DESeq2 models included patient identity as a blocking factor, and the analysis was performed in 74 histologically valid regions of interest (ROIs) representing full spectrum of CD histopathologies. DEGs were defined as adjusted P < .05 and |log2FC| >= 1.0. Results: Out of 18,677 genes analyzed in each of the ROIs, direct SS-versus-NS comparison identified 0 significant DEGs in mucosa, 18 in muscularis mucosae, 43 in submucosa, and 0 in muscularis propria. NS-versus-NI submucosa showed 132 DEGs (131 increased, 1 decreased), dominated by humoral/plasma-cell genes with a modest stromal-remodeling component. SS-versus-NI submucosa showed 144 DEGs (138 increased, 6 decreased) with coordinated contractile, extracellular-matrix/remodeling and cell-matrix adhesion programs. Muscularized versus fibrotic SS submucosa showed 219 DEGs, with marked enrichment of contractile and remodeling genes in muscularized regions. Excluding 5 follicle-containing submucosal ROIs markedly altered the NS comparison but not the core SS fibromuscular program. Exploratory follicular-niche analyses showed that lymphoid follicle-rich (LFR) submucosa was lymphoid/B-cell enriched relative to lymphoid follicle-null (LFN) areas; SS-LFR nevertheless acquired a stronger fibroblast/ECM program relative to NS-LFR, while CCL19 was the only significant LFR-versus-LFN DEG within SS. Conclusions: Fibrostenotic CD is characterized by spatially concentrated submucosal fibromuscular remodeled inflammatory program. Non-stenotic submucosa is heterogeneous and may already show regional stromal remodeling, whereas established stenosis displays a broader and more robust ECM-contractile-adhesion phenotype with molecularly distinct fibrotic and muscularized niches.

Inflammatory bowel disease (IBD) is a heterogeneous chronic disease driven by dysregulated mucosal immunity and impaired epithelial barrier function. Although biologics have improved disease management, they are frequently associated with systemic immunosuppression and adverse effects, highlighting the need for localized therapeutic strategies that both control inflammation and promote tissue repair. Here, we developed a protein bispecific termed 7A2-IgG4-IL22, composed of a human IgG4-Fc domain displaying an antagonistic anti-human MAdCAM-1 single chain (sc)-Fv and a human interleukin (IL-)22. The anti-MAdCAM-1 scFv retained the functional activity of the parental monoclonal antibody, inhibiting T cell activation, expansion and differentiation from naive precursors. Blockade of the MAdCAM-1 signaling axis also reduced production of pro-inflammatory cytokines relevant to IBD pathogenesis, including IFN{gamma} and TNF. On the epithelial side, the IL-22 cargo induces robust signaling in epithelial cells, promoting the expression of IL-22 response genes associated with antimicrobial defense, mucosal homeostasis, as well as IL-10 and CXCL1 expression. This effect contributes to immune cell trafficking to the intestinal mucosa. Together, this bispecific provides a localized dual-mechanism strategy for restoring intestinal immune homeostasis.

Background and Aims: Fibroblast-neutrophil interactions are implicated in ulcerative colitis (UC) treatment resistance, but whether fibroblasts modify neutrophil inflammatory activity remains unclear. We investigated responses to fibroblast exposure, stromal depletion, and 5{beta}1-directed treatment. Methods: We integrated human and mouse colonic single-cell RNA sequencing (scRNA-seq), spatial protein profiling (CODEX), fibroblast-neutrophil coculture, and dextran sulfate sodium (DSS) mouse colitis with fibroblast activation protein (FAP) ablation or 5{beta}1-directed treatment. Readouts included cellular representation and protein fluorescence via flow cytometry, neutrophil extracellular trap (NET)-associated elastase activity, RNA programs and epithelial responses using scRNA-seq. Results: Inflamed UC tissue contained higher fractions of inflammatory fibroblasts and oncostatin M (OSM)- and CXCR4-associated neutrophils. Inflammatory fibroblasts had higher FAP-5{beta}1 module scores, and proximity to FAP-high fibroblasts was associated with higher neutrophil OSM and CXCR4 fluorescence. UC fibroblasts increased neutrophil OSM, CXCR4 and myeloperoxidase (MPO) fluorescence and NET-associated elastase activity relative to control fibroblasts, while 5{beta}1-directed treatment attenuated these responses. Both mouse FAP ablation and 5{beta}1 blockade in DSS colitis reduced histologic inflammation and overall neutrophil frequency. Ablation broadly reduced recovered neutrophil representation, whereas blockade increased OSM-positive and PADI4-positive percentages within neutrophils despite lower MPO fluorescence within these subsets. Both interventions were associated with lower epithelial chemokine scores but distinct absorptive and mucus/secretory responses. Conclusions: Fibroblast exposure modifies neutrophil inflammatory phenotype and effector-associated activity, identifying candidate therapeutic pathways in UC. Evaluating these responses alongside neutrophil representation and tissue inflammation could inform pharmacodynamic assessment of stromal-directed therapies.

BACKGROUND & AIMS: Inflammatory bowel disease (IBD) is often diagnosed months to years after symptoms begin, and diagnostic delay is associated with complications. We tested whether longitudinal machine learning using electronic health record (EHR) data can identify incident IBD before diagnosis, and quantified lead time and deployment burden. METHODS: We studied adults with an IBD-compatible symptom, sign, or finding at UCSF (2012-2024). Incident cases were confirmed by structured filtering and validated large language model note review. We trained a state space model (Mamba) and comparators on pre-diagnostic EHR data, excluding diagnostic-workup features, and assessed discrimination at 6- and 12-month landmarks. We externally tested Mamba without recalibration at ZSFG, a safety-net hospital serving a demographically and socioeconomically distinct population, and simulated diagnosis-code-only deployment at UCSF. RESULTS: Six months after presentation, among patients still undiagnosed (67% of UCSF test cases), Mamba achieved an AUROC of 0.969 (95% CI, 0.928-0.999) at UCSF and 0.965 (0.952-0.976) at ZSFG. Median lead time was 20 months at UCSF and 15 months at ZSFG. Among future cases with an initially normal colonoscopy (N=32) or a histological finding of acute inflammation without chronicity (N=63), the model flagged 84% and 89% of these patients, respectively, before those procedures, indicating the potential for pre-diagnostic detection. Among 358,886 symptomatic patients not used for training (60 incident cases), the diagnosis-code-only model achieved an AUROC of 0.926; reviewing 20 alerts/month identified 34 cases (PPV, 1.09%; 64-fold above baseline). CONCLUSIONS: Longitudinal machine learning can identify increased IBD risk months to years before diagnosis and may define a population for prospective studies of early IBD biology. Prospective evaluation should determine whether earlier identification reduces diagnostic delay.

Objective Vedolizumab (VDZ) is effective in ulcerative colitis (UC), but its onset of action may be relatively slow during induction therapy. VSL#3TM, a high-potency multi-strain probiotic, may provide synergistic effects through microbiota modulation and immune regulation. This preliminary real-world study aimed to evaluate the efficacy and safety of VSL#3TM combined with VDZ in patients with UC. Methods Clinical data were retrospectively collected from patients with active UC who received VSL#3TM combined with VDZ for at least 12 weeks at the Second Affiliated Hospital of Zhejiang University School of Medicine in China between January 2023 and December 2024. The primary endpoints were clinical response rates at weeks 6 and 12. Secondary endpoints included clinical remission, changes in inflammatory bowel disease questionnaire (IBDQ) scores, safety assessment. Results Using PRO2 criteria, clinical response rates were 82.4% (14/17) at 6 weeks and 100.0% (17/17) at week 12, with clinical remission in 58.8% (10/17) at week 12. By Full Mayo Score, clinical response and remission at week 12 were 68.8% (11/16) and 50.0% (8/16), respectively. These response rates appeared numerically higher than those reported in published historical VDZ monotherapy studies, although direct comparisons are limited by the single-arm design. Mean IBDQ score improved from 157.3 at baseline to 174.5 at week 12 (p<0.001). The combination was well-tolerated with no serious adverse events. Fatigue, borborygmus, and arthralgia were reported in 1/17 (5.9%), 2/17 (11.8%), and 1/17 (5.9%) patients, respectively. Notably, all 3 patients with baseline history of Clostridioides difficile (CDI) positivity tested negative for both toxin and antigen at 12 weeks; one toxin-positive patient had received anti-CDI antibiotic therapy. Conclusions This preliminary real-world study suggests that VSL#3TM may enhance early clinical outcomes in patients receiving VDZ for active UC. The observed benefits may be related to complementary effects on the gut microbiota and intestinal immune responses, although these mechanisms were not directly assessed. Further larger prospective randomized controlled trials are warranted to confirm these findings.

Background and study aims: Endoscopic assessment of ulcerative colitis (UC) is central to clinical decision-making however remains subjective and limited in characterisation of disease distribution. Artificial intelligence (AI) enables analysis of entire endoscopic examination, rather than relying on selected views. We aimed to develop and validate a deep learning model for automated assessment of UC severity from full-length endoscopic videos, introducing spatial representation of inflammation (Continuous Disease Score, CDS), and assess histological correlation. Patients and methods: Full-length endoscopy videos from adult patients with UC undergoing colonoscopy or flexible sigmoidoscopy were analysed; isolated proctitis was excluded. Videos were segmented and annotated using Mayo Endoscopic Score (MES) and Ulcerative Colitis Endoscopic Index of Activity (UCEIS). A deep learning model was trained for frame-level quality control and severity prediction, enabling analysis of full-length videos. Performance was evaluated using quadratic weighted kappa (QWK) and Cohen's kappa, with patient-level separation between datasets. CDS was derived from UCEIS predictions to quantify cumulative inflammatory burden, spatial extent of disease and histological prediction. Results: A total of 67 videos from 59 patients were included. The model demonstrated agreement for remission classification (MES=0 {kappa} 0.76; UCEIS[&le;]1 {kappa} 0.84). CDS enabled quantification of inflammatory burden and revealed spatial heterogeneity not reflected in categorical scores. Agreement with histology was strong (AUROC 0.84-0.87). Conclusions: AI-based analysis enables automated assessment of UC activity from full-length endoscopic videos, including remission detection and estimation of histological healing. CDS provides continuous characterisation of inflammatory burden and disease extent beyond conventional categorical scores, with potential to support more standardised assessment in clinical trials and practice.

Ulcerative colitis (UC) is characterised by chronic colonic inflammation with marked heterogeneity in disease severity and therapeutic outcomes. Here, we define a spatially organised, polyfunctional cytotoxic CD4 T-cell state associated with mucosal inflammation and adverse therapeutic outcomes in UC. Integrating ex vivo T-cell receptor stimulation with multi-cohort bulk and single-cell transcriptomics and multiparameter flow cytometry, we show that GZMB CD4 T cells are preferentially enriched in inflamed UC mucosa, but not peripheral blood, and co-express cytotoxic molecules, Th1- and Th17-associated cytokines and chemokines, and immunoregulatory receptors. Single-cell analyses implicate inflammatory cytokine and antigen-presentation signals in the acquisition or maintenance of this state. High-resolution spatial profiling localised this programme predominantly to Th17 cells, which were preferentially enriched within multicellular inflammatory and tertiary lymphoid structure-associated niches. Across independent patient cohorts, a transcriptional signature derived from this state increased with endoscopic disease severity and was associated with reduced response to anti-TNF and anti-IL-12/23p40 therapies. Adoptive transfer of Gzma/Gzmb-deficient rather than wild-type CD4 T cells into Rag2-deficient recipient mice markedly attenuated experimental colitis and abrogated the polyfunctional cytokine phenotype, demonstrating that granzyme-dependent effector activity is a key mechanism driving CD4+ T-cell-mediated intestinal inflammation. Finally, human host-microbiome analysis linked this programme to intestinal dysbiosis, while transfer of dysbiotic microbiota promoted the emergence of a corresponding state in vivo. Collectively, these findings define a microbiota-responsive, spatially organised polyfunctional cytotoxic CD4 T-cell programme that contributes to intestinal inflammation and is associated with disease severity and treatment resistance in UC.

In patients with Crohns disease, the noninflamed intestinal epithelium exhibits inflammation-like transcriptional signatures that persist even during clinical remission. However, it is unclear whether this signature is due to sustained environmental triggers or to the epitheliums intrinsic immunological adaptions. To disentangle environmental and cell-intrinsic effects on disease signatures, we used noninflamed intestinal tissue biopsies from patients with active and inactive Crohns disease, as well as control donors, to generate donor-matched organoids enriched for intestinal epithelial cells. We collected single-cell RNA-seq and single-cell ATAC-seq data from intestinal tissues, and from their matched organoids before and after stimulation with luminal metabolites of patients. This approach allowed us to distinguish between the effects of extrinsic factors and the intrinsic alterations that persist when cells are removed from their local environment. We found that organoid cells retained epithelial-autonomous disease phenotypes, even when derived from patients in remission, whereas signatures from tissue cells showed evidence of immune- and microbial-epithelial crosstalk. In contrast to healthy control organoids, organoids from Crohns disease patients exhibited selective activation of disease-associated transcriptional changes and environment-responsive inflammatory chromatin remodeling after exposure to patient-derived luminal metabolites. Overall, our study suggests that disease-associated epithelial states are shaped by both cell-intrinsic dysregulation and microenvironmental cues.

The gut microbiome can influence cancer development, but its contribution to cancer predisposition in Li-Fraumeni Syndrome (LFS) remains poorly understood. Here, we investigated whether germline mutant p53 alters the intestinal environment using Trp53R172H/+ mice. Tumour-naive mutant mice exhibited increased intestinal permeability and a pro-inflammatory colonic phenotype, including elevated inflammatory cytokines and enhanced NF-{kappa}B signalling. Untargeted metabolomic profiling revealed a striking alteration in bile acid metabolism, with targeted analysis identifying increases in multiple primary and microbiota-derived secondary bile acids. Notably, these changes occurred despite no detectable differences in overall microbial community composition, suggesting altered host-microbial metabolic output in the absence of broad compositional change. Several faecal bile acids were positively associated with distal tumour burden, with the microbiota-derived secondary bile acid deoxycholic acid (DCA) showing the strongest relationship. Microbiota depletion reduced distal tumour growth in mutant mice and attenuated intestinal barrier dysfunction, colonic inflammation and the altered faecal bile acid profile, including DCA. Conversely, transfer of cell-free faecal filtrate from mutant donors to WT recipients induced colonic inflammation and enhanced distal tumour growth. Together, these findings identify a mutant p53-gut axis in which germline mutant p53 establishes a microbiota-shaped metabolic and inflammatory intestinal environment capable of promoting tumour growth beyond the gut.

The aim of this study was to determine whether the frequency of metabolic dysfunction-associated steatotic liver disease (MASLD) and the severity profile of inflammatory bowel disease (IBD) differ across body mass index (BMI) categories. A retrospective analysis was performed using 400 inpatient medical records from 2023-2024 with verified ulcerative colitis or Crohn's disease and sufficient information on anthropometry, hepatic status, and intestinal disease activity. The cohort comprised 118 patients with normal BMI, 176 with overweight, and 106 with obesity. MASLD status was reconstructed from routine clinical documentation and required hepatic steatosis in the appropriate cardiometabolic context. Ultrasonography was the principal source of evidence for steatosis; laboratory findings, non-invasive indices, and elastography were considered supportive when available. Ulcerative colitis activity was assessed with the Mayo Score and Crohn's disease activity with the Crohn's Disease Activity Index; disease-specific results were harmonized into mild, moderate, and severe categories for pooled comparisons. MASLD was identified in 181 of 400 patients (45.3%). Its frequency increased from 20.3% in the normal-BMI group to 49.4% in the overweight group and 66.0% in the obesity group (chi2=49.29; p<0.001). The largest absolute increase, 29.1 percentage points, occurred between normal BMI and overweight. The proportion of severe IBD rose in the same order, from 6.8% to 14.8% and 32.1% across the three BMI strata (chi2=39.45; p<0.001). In unadjusted pairwise comparisons, the odds of MASLD were higher with overweight than with normal BMI (OR=3.83; 95% CI 2.24-6.55) and higher with obesity than with normal BMI (OR=7.62; 95% CI 4.17-13.91); obesity also differed from overweight (OR=1.99; 95% CI 1.21-3.28). These findings demonstrate an ordered BMI-related pattern in which MASLD becomes more frequent and severe IBD represents a larger share of cases. However, MASLD was still documented in approximately one fifth of patients with normal BMI, indicating that BMI is useful for clinical prioritization but not for excluding steatotic liver disease. Because the study was retrospective, liver assessment reflected routine care, and the effect estimates were not adjusted for treatment exposure, disease duration, or other metabolic covariates, the observed relationships should be interpreted as associations rather than causal effects.

CD28 costimulation contributes to pathogenic T cell responses in inflammatory bowel disease (IBD), but current B7-directed blockade also limits CTLA-4 signaling. Using a sensitive NanoBiT split-luciferase screening platform, we identified and optimized CA-23, a small molecule antagonist that directly binds human and mouse CD28 without measurable binding to CD80, CD86, or CTLA-4. CA-23 inhibited CD28-B7 engagement and CD28-dependent T cell activation without agonist activity in human whole blood and peripheral blood mononuclear cells. CA-23 achieved exposure in the colon and mesenteric lymph nodes and reduced disease severity, histologic injury, and pathogenic Th1 and Th17 responses in a T cell transfer model of colitis. In PBMCs from donors with ulcerative colitis or Crohns disease, CA-23 suppressed inflammatory cytokine production and T cell activation to a degree matching or exceeding Abatacept. In human intestinal epithelial-PBMC co-cultures, CA-23 preserved Treg suppressive activity and epithelial barrier integrity, whereas Abatacept reduced Treg function. CA-23 did not alter CD80 or CD86 expression on autologous antigen-presenting cells and showed no substantial off-target activity in the tested selectivity panel. These findings support direct CD28 antagonism as a mechanistically differentiated alternative to B7-directed co-stimulation blockade for suppressing pathogenic T cell responses in preclinical models of IBD. One Sentence SummaryA CD28-selective small molecule blocks pathogenic T cell activation and preserves Treg function unlike Abatacept in IBD models.