BMJ Open Gastroenterology最新文献

Objective: Endoscopic scoring of disease activity is essential in ulcerative colitis trials but is limited by inter-reader variability and operational complexity. We evaluated a novel central reading paradigm integrating two independently developed machine learning models with human adjudication for discordant cases ('2M+1H') to determine whether it is statistically non-inferior to the traditional two-reader-plus-adjudicator model.

Methods: A total of 150 full-length endoscopic videos were retrospectively scored using both the conventional central reading approach and the 2M+1H paradigm. Each machine learning model was developed independently using distinct algorithms and datasets. When the two model-generated scores disagreed, a board-certified gastroenterologist adjudicated the final score. The primary endpoint was agreement with the reference standard, measured by quadratic weighted kappa. Secondary endpoints included agreement on binary outcomes (endoscopic improvement and remission), reduction in human reads and evaluation of outcome variability due to human reader mix.

Results: The 2M+1H approach achieved a quadratic weighted kappa of 0.78 with the reference standard, meeting the prespecified threshold for non-inferiority. Agreement with the reference standard was 82.7% for endoscopic improvement and 89.3% for remission. Compared with the traditional method, the 2M+1H paradigm reduced human reads per video by 81%. Notably, 16% of cases in the human-only approach yielded different final scores depending on reader assignment.

Conclusion: The 2M+1H central reading paradigm provides statistically non-inferior accuracy with greater operational efficiency and potentially enhanced reproducibility. Prospective validation against clinical, biomarker and histological outcomes is warranted.

Objectives: Parasitic colitis is highly prevalent worldwide, may mimic inflammatory bowel disease (IBD) and is encountered by non-specialist physicians in high-income, non-endemic settings. In this context, patients may be at risk of misdiagnosis and poor outcomes. However, cases are not routinely reported, limiting the evidence base to individual case reports. We aimed to systematically describe and evaluate the diagnosis, clinical course and outcomes of affected patients whose cases have been reported in high-income settings.

Design: A systematic review of the literature using the Preferred Reporting Items for Systematic Reviews and Meta-analyses framework.

Data sources: PubMed and Scopus databases were searched for publications from between 1 January 2012 and 10 January 2025.

Eligibility criteria: Case reports of undifferentiated colitis presenting in high-income settings, where the aetiological agent was ultimately found to be parasitological, were eligible for inclusion. Key exclusion criteria included a history of travel to an endemic area within 4 weeks of presentation. No language restrictions were applied.

Data extraction and synthesis: Key components of each case report, encompassing patient presentation, diagnosis and management, were standardised as categorical descriptors. Key themes were identified, and a thematic synthesis approach was employed.

Results: 52 articles, describing 54 patients, were included in the final analysis: 33 cases of amoebiasis, 15 cases of strongyloidiasis and 6 cases of schistosomiasis. Misdiagnosis occurred in 37 out of 54 patients (69%), with 28 out of 37 (76%) of these misdiagnosed as IBD. Substantial harm was reported in 31 out of 54 (57%) patients, including death in seven patients. Major morbidity (defined as strongyloides hyperinfection syndrome, fulminant amoebic colitis, emergency surgery or sepsis) or death was associated with administration of corticosteroids in cases of strongyloidiasis and amoebiasis, occurring in 8 out of 9 (89%) and 1 out of 6 (17%) patients with strongyloidiasis who received/did not receive corticosteroids, and 10 out of 13 (77%) and 2 out of 20 (10%) patients with amoebiasis, respectively.

Conclusions: Parasitic colitis poses a diagnostic challenge in high-income settings. Misdiagnosis as IBD is commonly reported, with poor outcomes related to corticosteroid treatment. Awareness of parasitic colitis aetiologies, at-risk groups and diagnostic tests is essential among generalist clinicians assessing undifferentiated colitis to avoid poor outcomes.

Prospero registration number: CRD420251033374.

Serrated polyps are believed to play an important role in the development of colorectal cancer (CRC). They share molecular characteristics and proximal colonic location with postcolonoscopy colorectal cancer (PCCRC), indicating that they may give rise to a disproportionately high proportion of PCCRC. In this narrative literature review, we aimed to synthesise the current understanding of serrated polyps and their role in the development of CRC. Our review focuses on the prevalence of serrated lesions in various populations, risk of progression to advanced colorectal neoplasia and cancer, and the effect of serrated polyp detection on PCCRC rates. Serrated polyp detection is noted to be variable between studies based on the type of serrated polyps studied and the patient population in question. In addition, low-risk sessile serrated lesions (SSLs) do not seem to have a higher risk for progression to advanced neoplasia than conventional adenomas, while more advanced lesions (10 mm or larger, those with dysplasia and traditional serrated adenomas) do seem to have a higher malignant potential, particularly in combination with traditional adenomas. Finally, several studies demonstrate a decrease in PCCRC rates among endoscopists with higher serrated polyp detection metrics. We identified a need for standardising the definition of clinically significant serrated polyps for future monitoring and benchmarking of serrated polyp detection metrics, further research to determine the surveillance interval for patients with concomitant adenomas and serrated polyps, and widespread education and training for endoscopists and pathologists to improve detection, diagnosis and management of serrated lesions.

Objectives: We aimed to systematically review the real-world evidence (RWE) on the effectiveness and utilisation of advanced therapies for inflammatory bowel disease (IBD) in Middle Eastern populations.

Design: Systematic review.

Data sources: PubMed/MEDLINE, Embase, Scopus, Web of Science and the Cochrane Central Register of Controlled Trials were searched up to May 2025.

Eligibility criteria: Observational RWE studies investigating biologics or small molecules in Middle Eastern IBD patients (adult and paediatric) were included. Randomised controlled trials and case series with fewer than 10 patients were excluded. No language restrictions were applied.

Data extraction and synthesis: Data were independently extracted by two reviewers. Due to significant heterogeneity in study design, populations and outcome reporting, a narrative synthesis was performed.

Results: From 884 records, 23 studies were included, originating primarily from Saudi Arabia (n=8) and Iran (n=4). For anti-tumour necrosis factor (TNF) therapy, a Kuwaiti study of biologic-naive patients found 12-month endoscopic remission rates with infliximab of 56% for ulcerative colitis (UC) and 53% for Crohn's disease (CD), while a Saudi study reported higher odds of treatment failure with adalimumab versus infliximab (OR=26.91). Ustekinumab demonstrated strong efficacy, achieving 76.9% clinical remission at 52 weeks in a Saudi paediatric anti-TNF refractory IBD cohort and showing higher probability of effectiveness than vedolizumab in another Saudi study. In contrast, vedolizumab remission rates in advanced therapy-experienced UC patients were 89.3% with intensified dosing. Newer agents also showed promise; risankizumab induction led to 43.2% clinical remission in an Emirati CD cohort, while tofacitinib achieved clinical remission rates of 56.4% and 61.1% at 52 weeks in Lebanese and Iranian UC cohorts, respectively.

Conclusions: Advanced therapies for IBD appear to be effective in Middle Eastern cohorts; however, the available evidence is methodologically diverse, with substantial heterogeneity in study design, population characteristics and outcome reporting, which limits the ability to draw strong conclusions and highlights the need for further robust evaluation. Prospective, collaborative regional registries are imperative to address these gaps and inform local guidelines.

Prospero registration number: CRD420251083256.

Objective: The lack of a rapid, validated, consistent test for tracking disease activity in patients with inflammatory bowel disease (IBD) is currently a major challenge. Currently used biomarkers have notable disadvantages, such as the slow processing (faecal calprotectin) and the lack of specificity (bloodwork). White blood cell (WBC) subsets, also known as 'the differential', are commonly obtained in evaluating IBD patients, but there is minimal evidence on how these subsets relate to disease activity. Given the interplay between immune cells, it is possible that complex patterns in WBC subsets could be used to classify IBD activity. Machine learning (ML) could be used to reveal these changes. The aim of this study was to classify IBD activity via routine bloodwork results, using an ML approach.

Methods: 1458 bloodwork measurements from 108 IBD patients were included in this analysis. Disease activity was classified by physician's global assessment score. Four ML models were trained to classify active disease or remission based on routine bloodwork metrics (complete blood count, differential, albumin, erythrocyte sedimentation rate and C reactive protein).

Results: The optimal model, extreme gradient boosted decision trees, achieved a receiver operator characteristic area under the curve of 0.882. Feature analysis identified neutrophils, C reactive protein and albumin as consistently important contributors to the models. Additionally, no single individual biomarker was comparable to the ML model, and medications had only a minor impact on the ML model.

Conclusion: Classification of IBD activity can be augmented using ML analysis of commonly measured bloodwork parameters to help inform treatment plans and to improve IBD patient outcomes.

Pancreatic ductal adenocarcinoma (PDAC) remains one of the deadliest solid tumours, with 5-year survival consistently below 10% and only modest gains from conventional chemotherapy after first-line failure. Although oncogenic KRAS mutations dominate the genomic landscape, recent large-scale sequencing has revealed a series of less frequent but therapeutically actionable alterations. This review synthesises evidence from phase I-II trials published through April 2025. It demonstrates that targeting these lesions can yield outcomes that meet or exceed the benchmarks set by the NAPOLI-1 trial (liposomal irinotecan plus 5-fluorouracil and leucovorin), with a median overall survival of 6.2 months and progression-free survival of 3.1 months. Objective response rates reach 33% with adagrasib in KRAS G12C PDAC, 22% with olaparib maintenance in germline BRCA1/2 cancers, and over 50% with RET or NTRK inhibitors with fusion alterations; pembrolizumab produces durable benefit in the 1-3% of tumours that are MSI-H/dMMR. Emerging data highlight NRG1 fusions (overall response rate 42% with zenocutuzumab), HER2 amplification, MTAP deletion with PRMT5 dependency and variant-specific (MRTX1133) or pan-RAS (daraxonrasib) inhibitors as the next frontier. Toxicity profiles of targeted agents are generally favourable and often allow prolonged administration compared with cytotoxic regimens. Taken together, these advances represent a substantive therapeutic progress in PDAC over the past decades, even though they currently apply to a minority of patients. These findings underscore the necessity of comprehensive next-generation sequencing for every patient with advanced disease, enabling identification of rare, yet clinically meaningful, targets and moving PDAC management towards a precision-oncology paradigm.

Objectives: Higher liver fibrosis scores have been associated with all-cause and liver disease-related mortality. We investigated whether liver fibrosis scores were associated with all-cause, liver disease-related and non-liver disease-related mortality and liver cancer incidence in an occupational cohort.

Methods: The study included 43 870 Combined World Trade Center Rescue/Recovery Worker Cohort members who had blood drawn between 11/9/2001-31/12/2020. Aspartate aminotransferase to platelet ratio index (APRI), fibrosis-4 score (FIB-4), and non-alcoholic fatty liver disease fibrosis score (NFS) were calculated and categorised as low, intermediate or high scores per established cut-offs. Deaths and liver cancers were identified via National Death Index records and state cancer registries. Cox proportional hazards regression models estimated HRs and 95% CIs for each outcome in those with intermediate and high versus low fibrosis scores.

Results: By 31/12/2020, 1996 deaths, 81 liver disease-related deaths and 36 incident liver cancers occurred. Participants with intermediate or high APRI (HR=1.73, 95% CI 1.50 to 1.99 and HR=7.23, 95% CI 5.63 to 9.29), FIB-4 (HR=1.18, 95% CI 1.05 to 1.34 and HR=3.96, 95% CI 3.22 to 4.85) or NFS (HR=1.61, 95% CI 1.43 to 1.81 and HR=4.73, 95% CI 3.78 to 5.93) had increased risks of all-cause mortality vs those with low scores, controlling for potential confounders. Those with intermediate/high scores also had increased liver disease- and non-liver disease-related mortality. High APRI, FIB-4 and NFS, and intermediate APRI and NFS were associated with liver cancer.

Conclusions: Intermediate and high liver fibrosis scores predicted all-cause, liver disease-related and non-liver disease-related mortality, and high scores predicted liver cancer, in this population. Liver fibrosis scores may identify those at greater risk for poor health outcomes, even in a healthy occupational cohort.

Objective: Inflammatory bowel disease (IBD)-associated liver diseases (IBDALDs) are associated with hepatobiliary damage and malignancy, with diagnosis often delayed by heterogeneous presentation. We evaluated whether routinely collected biomarkers-at IBD diagnosis and during follow-up-can risk-stratify for IBDALD.

Methods: This observational retrospective longitudinal study included 1571 patients with IBD at University Hospital Southampton. Biomarkers including alanine aminotransferase (ALT), alkaline phosphatase (ALP) and erythrocyte sedimentation rate (ESR) (n=335 605 results) were summarised as patient-level medians within ±6 months of IBD diagnosis. Patients with pre-existing IBDALD were excluded. A 1:4 matched case-control design (age, sex, IBD subtype) was applied. Conditional logistic regression assessed associations with biomarkers (continuous values and binary-abnormal vs normal) and IBDALD. Longitudinal trends were evaluated using locally estimated scatterplot smoothing (LOESS) and linear mixed-effects models (LMMs).

Results: Median age of IBD diagnosis was 18.0 years, median follow-up 11.5 years. Thirty-five IBDALD cases were identified (27 post-IBD); median time to IBDALD was 4.5 years. At IBD diagnosis, cases had elevated ALT, ALP and ESR (p<0.01). In case-control matching, ALT (OR=1.04 per U/L; 95% CI 1.01 to 1.07; p=0.012), ALP (OR=1.01; 95% CI 1.00 to 1.02; p=0.014) and ESR (OR=1.05; 95% CI 1.00 to 1.09; p=0.034) were associated with IBDALD. Any abnormal ALT (OR=5.10; 95% CI 1.57 to 16.59; p=0.0068) and ALP (OR=15.33; 95% CI 1.87 to 125.77; p=0.0110) were strongly associated. LOESS plots and LMMs demonstrated distinct biomarker trajectories (ALT, ALP) preceding IBDALD.

Conclusion: Real-world biomarker data can support early risk stratification for IBDALD. Elevated ALT and ALP at IBD diagnosis and distinct longitudinal trajectories highlight the need for follow-up to biomarker normalisation, with persistent abnormalities prompting earlier hepatobiliary investigation to reduce diagnostic delay and improve outcomes.

Objectives: Long-term colorectal cancer (CRC) monitoring is common in Asia, but assessing colonoscopy effectiveness through adenoma detection rates (ADRs) remains uncommon. This study aimed to identify potential associations between ADR and risk of post-colonoscopy CRC (PCCRC).

Methods: This retrospective, longitudinal, observational study included participants (aged 40-74 years) with initial positive results and subsequent colonoscopy recruited from a community-based, multicentre, CRC screening programme from January 2012 to December 2022 in Tianjin, China. ADR was defined as the proportion of colonoscopies in which at least one adenoma was detected at the provider level. The primary outcome was PCCRC, defined as any CRC diagnosed 6 months after initial colonoscopy. ADR groups were categorised according to quartile distribution (<42%, 42%-55%, 55%-64% and >64%) or dichotomised as at or above vs below the median for statistical analyses. PCCRC incidence rates were expressed as the number of events per 1000 person-years.

Results: Among 9957 included participants, 116 PCCRC cases were detected in 33 881 person-years of follow-up. PCCRC incidence rates per 1000 person-years were 4.34, 3.76, 2.62 and 2.69 from the lowest to the highest ADR group, respectively. ADR was significantly inversely associated with PCCRC risk. Participants in the highest ADR group had 49% lower risk of PCCRC than those in the low ADR group (adjusted HR 0.51, 95% CI 0.29 to 0.88). Multivariate-adjusted restricted cubic spline analyses identified a linear dose-response relationship between ADR and PCCRC risk.

Conclusions: This study illustrates the relationship between endoscopist competence, as measured by ADR, and CRC risk after colonoscopy, which can guide assessment of monitoring programmes.

Objective: Faecal immunochemical test (FIT) is widely used in colorectal cancer (CRC) screening, but the impact of antithrombotic therapy on colonoscopy outcomes remains unclear. This study aimed to compare polyp detection rate (PDR) and prevalence of non-neoplastic findings in FIT-positive patients undergoing colonoscopy, stratified by antithrombotic therapy.

Methods: We conducted a retrospective, multicentre cohort study of 26 280 FIT-positive colonoscopies performed between 2016 and 2023 across seven gastroenterology centres in Israel. The FIT positivity threshold was 75 ng/mL. Patients were grouped by therapy: antiplatelets (n=1904), anticoagulants (n=262) or no antithrombotic treatment (n=24 079). Using crude and matched analyses, we assessed PDR, adenoma detection rate (ADR), CRC detection and diverticulosis prevalence.

Results: Patients on antithrombotics were older (mean 65.9±7 vs 61.0±8 years; p<0.001), more often men (65.2% vs 50.1%; p<0.001) and had higher comorbidity rates. Crude PDR was higher in antithrombotic users (55.5% vs 51.9%; p=0.001), but ADR was similar (28.2% vs 28.3%; p=0.950). Diverticulosis was more prevalent in the antithrombotic group (24.0% vs 17.4%; p<0.001). After matching for age and sex (n=2201 pairs), all previously significant associations were eliminated: PDR were no longer different (55.5% vs 53.1%; p=0.112), diverticulosis prevalence became identical (24.0% vs 24.3%; p=0.799) and normal colonoscopy rates were similar (27.5% vs 29.6%; p=0.123). ADRs remained comparable (28.2% vs 26.7%; p=0.269), and CRC detection rates were identical (2.0% vs 2.0%; p=0.998).

Conclusion: Demographic confounding, rather than antithrombotic therapy per se, explains the crude associations in colonoscopy outcomes. The matched analysis demonstrates that antithrombotics do not independently affect polyp detection, diverticulosis prevalence or adenoma detection when age and sex are controlled for. These findings support the continued use of antithrombotics during CRC screening, shifting clinical interpretation from medication-based to demographic-based risk stratification.