Gastrojejunal anastomotic (GJA) ulcers are a common complication after Roux-en-Y gastric bypass (RYGB) and may be refractory to medical therapy. Impaired mucosal perfusion is suspected to contribute to non-healing ulcers, but no method exists to directly assess ischemia. We evaluated the feasibility of real-time endoscopic tissue oxygen saturation (StO2) imaging to quantify mucosal perfusion and its association with ulcer healing. In this prospective study, 51 adults were included: 18 with normal gastrointestinal anatomy, 26 with RYGB anatomy without complications, and a subgroup of 7 RYGB patients with confirmed GJA ulcers. StO2 was measured in the esophagus, stomach/pouch, duodenum/jejunum, and anastomotic sites using an endoscopic StO2 imaging system. Ulcer patients received maximal medical therapy and were reassessed by follow-up endoscopy. StO2 values were compared across regions and between healed and non-healed ulcers. In normal anatomy, esophageal StO2 was lower than gastric and duodenal values, while stomach and duodenum were similar. In RYGB patients without complications, StO2 values were comparable across the pouch, anastomosis, and Roux limb. Among ulcers, healed sites had significantly higher StO2 than non-healed sites (p = 0.002). Healed ulcers exhibited StO2 near baseline anastomotic levels, whereas non-healed ulcers demonstrated markedly reduced StO2. Endoscopic StO2 imaging is feasible and enables quantitative assessment of mucosal perfusion in both normal and post-RYGB anatomy. Low StO2 at GJA ulcers is associated with poor healing, highlighting its potential to identify ischemic ulcers that may benefit from early surgical intervention. However, given the small sample size, particularly in the ulcer subgroup, these findings should be considered preliminary and interpreted with caution. Larger studies are needed to validate StO2 thresholds and define its role in clinical decision-making after RYGB.

Background and Aim:Despite advances in non-invasive methods, liver biopsy remains essential for diagnosis, prognosis, and treatment planning, and this study aimed to evaluate trends in its indications and pathological findings. Materials and Methods:A total of 1533 adult liver biopsies at a tertiary referral hospital (2000-2022) were retrospectively reviewed for indications and diagnoses. Results:The annual number of liver biopsies performed remained stable over the last two decades (730 vs. 803; p=0.06). The mean patient age was 43±15 years, with 52% being female. The leading indications were abnormal liver function tests (53.4%), fibrosis/cirrhosis (40.9%), treatment follow-up (2.7%), hepatomegaly (1.6%), and fever of unknown origin (0.9%). Between 2000-2011 and 2012-2022, the indications for liver biopsy changed significantly: biopsies for fibrosis/cirrhosis increased, whereas those for evaluation of treatment response and disease monitoring and for hepatomegaly decreased (p=0.01, p<0.001, and p=0.02, respectively). In the post-COVID-19 period, abnormal liver function tests became the predominant indication (42.2% vs. 69.2%), whereas fibrosis/cirrhosis decreased (54.2% vs. 28%). The most frequent pathological diagnosis was chronic hepatitis B virus (HBV) infection (38.6%), followed by chronic hepatitis C virus (HCV) (11.5%), metabolic dysfunction-associated steatohepatitis (MASH) (8.5%), autoimmune hepatitis (AIH) (7.7%), and primary biliary cholangitis (PBC) (4.2%). Across decades, MASH (6.4% vs. 10.3%; p=0.008), AIH (4% vs. 11.1%), and PBC (1.1% vs. 7%) increased significantly (p<0.001), while chronic HCV declined (17.3% vs. 6.1%; p<0.001). Chronic HBV remained stable (38.2% vs. 38.9%). Conclusion:Over the past 22 years, the annual number of liver biopsies remained stable, with abnormal liver function tests as the leading indication, a decline in fibrosis/cirrhosis evaluation, chronic hepatitis B as the most common diagnosis, and a gradual rise in autoimmune liver diseases and MASH.

Background and aims The use of large language models (LLMs), such as ChatGPT, is rapidly expanding in clinical medicine but data in gastroenterology are limited. We aimed to assess the prevalence and patterns of LLM use among European gastroenterologists, evaluate perceived benefits and risks and identify barriers and priorities for future integration. Methods A Europe-wide online survey was conducted between June and August 2025. A 24-item questionnaire was distributed via national gastroenterology societies and social media. Questions on specific applications, reliability and clinical impact were shown only to respondents with prior LLM experience. Results A total of 966 gastroenterologists from 34 European countries responded, 55.3% reported using LLMs, mainly weekly (54.4%) or daily (20.2%). Common uses included education (59.2%), clinical decision support (58.2%) and scientific writing (39.1%). Among academic users, summarisation (80.1%) and grammar refinement (74.4%) were most frequent. Most users rated LLMs as moderately or highly reliable (81.0%), and 84.3% believed they improve outcomes. Nearly two-thirds (64.5%) felt LLMs can support endoscopy practice. Inaccuracies were reported by 42.7% and were associated with lower trust (rho=-0.40, p<0.001). Although only 7.8% had formal training, 90.0% supported structured education and 85.2% favoured curricular integration. Most respondents would recommend LLM use (91.4%) and expect broader adoption within 5 years (93.4%). Reported barriers included institutional regulations (45.7%), cost (36.1%), complexity (29.4%) and time constraints (27.9%). Conclusion LLMs are increasingly integrated into clinical and academic workflows in European gastroenterology. Structured training, specialty-specific development and clear governance are essential for safe implementation.

Bariatric procedures, including endoscopic sleeve gastroplasty (ESG) and laparoscopic sleeve gastrectomy (LSG), are established weight loss interventions. Recent advances in endoscopic technology enable real-time measurement of mucosal tissue oxygen saturation (StO2) in the gastrointestinal tract, providing insight into tissue perfusion. This pilot study aimed to establish baseline StO2 values across gastrointestinal regions in uncomplicated post-LSG patients and to investigate changes in gastric StO2 pre- and post-ESG and their relationship with weight loss outcomes. This prospective, single-center pilot study enrolled 14 adults (6 LSG, 8 ESG). In the LSG group, StO2 was measured in the esophagus, proximal sleeve, distal sleeve, and duodenum. In the ESG group, gastric mucosal StO2 was recorded immediately before and after the procedure. Demographics, clinical history, and one-month post-ESG weight loss data were also collected. Paired t-tests compared StO2 values, and multivariable linear regression assessed the association between StO2 change and weight loss. In LSG patients, the proximal gastric sleeve showed higher StO2 (82

BACKGROUND & AIMS:Evidence on optimal bismuth and antibiotic dosing is limited. We aimed to analyze the effectiveness of bismuth-containing quadruple regimens according to drug dosage. METHODS:This was an analysis of an international, prospective, non-interventional registry on the management of Helicobacter pylori infection by European gastroenterologists (European Registry on Helicobacter pylori Management [Hp-EuReg]). Patients receiving empirical bismuth-containing quadruple therapies in any treatment line and registered from 2013 to 2024 were included. Effectiveness was assessed using a modified intention-to-treat (mITT) analysis. RESULTS:A total of 10,767 first-line records were analyzed. The clarithromycin-amoxicillin-bismuth (CAB) scheme was >90% effective with standard doses of amoxicillin and clarithromycin (2000 mg/d and 1000 mg/d), regardless of bismuth dose. The bismuth-metronidazole-tetracycline (BMTc) regimen achieved >90% effectiveness with tetracycline ≥1500 mg/d, metronidazole 1500 mg/d, and bismuth ≥480 mg/d. The amoxicillin-metronidazole-bismuth (AMB) regimen reported ≈90% mITT with amoxicillin 2000 mg/d, metronidazole ≥1000 mg/d, and bismuth 480 mg/d. The clarithromycin-metronidazole-bismuth (CMB) showed ≥90% mITT with metronidazole ≥800 mg/d, without improvement from increasing bismuth dose. In the amoxicillin-levofloxacin-bismuth (ALB) regimen, mITT was >90%, with no benefit from increasing levofloxacin >500 mg/d. A total of 2952 second- to sixth-line treatments were analyzed, all showing mITT <90% (ALB, 83%; BMTc, 79%; CAB, 88%; MDB, 63%; AMB, 88%), with no significant differences with different bismuth doses. CONCLUSIONS:Increasing antibiotic or bismuth doses above the standard did not improve effectiveness of H pylori treatment. Adherence, treatment length, and proton pump inhibitor co-therapy may be more impactful than escalation of antibiotics/bismuth doses. CLINICALTRIALS:gov, Number: NCT02328131.

Background Advances in artificial intelligence have brought renewed attention to tools that can work with the large amount of written information generated in clinical practice. Among these, large language models (LLMs) stand out for their ability to interpret and generate medical text in a flexible, context-aware way. Gastroenterology, like many specialties, produces a wide range of narrative and semi-structured data, and this has encouraged researchers to explore how LLMs might help clinicians manage everyday tasks. Recent studies have examined their potential contributions to patient education, communication between care teams and patients, decision support, and routine documentation, reflecting a growing interest in how these systems might fit into real-world clinical workflows. Objective This scoping review aimed to map current applications of LLMs in gastroenterology clinical practice, including subspecialty focus, study designs, model types, and reported outcomes. Method Following PRISMA guidelines, a systematic search was conducted in PubMed, Scopus, and Web of Science for studies published between January 2022 and August 2025. Eligible studies included original research assessing LLM applications in gastroenterology clinical practice. Data were extracted on subspecialty, application domain, LLM type, data source, and outcomes. We employed thematic analysis to address our primary research question. Results 73 out of 2895 studies identified in the initial search met the inclusion criteria. Six subspecialties and six application domains emerged from our review. Hepatology (20/73 studies, 27.3%) and endoscopy (17/73 studies, 23.2%) were the most represented subspecialties. The most frequently investigated application domains were patient education and communication (38 studies) and decision support and clinical guidance (24 studies). Most studies were simulation-based or literature-based cases, although an increasing number have used real-world clinical data, particularly in recent years. The majority evaluated general-purpose models such as GPT-3.5 and GPT-4, with some incorporating retrieval augmentation or fine-tuning. Reported outcomes varied by application domain and included measures of accuracy, concordance, completeness, relevance, safety, reliability, usability, user satisfaction, efficiency, time savings, and educational value. Commonly described limitations included variable reliability, incomplete responses, and challenges in generalizing from simulated to clinical settings. Conclusions Research on LLMs in gastroenterology has expanded across multiple subspecialties and application domains. Current evidence is primarily based on simulation studies, with limited but growing evaluation using real-world clinical data. Further work is needed to assess performance in prospective and applied clinical contexts.

BACKGROUND:The Maastricht VI/Florence consensus recommends the eradication of all Helicobacter pylori (H. pylori) infections. However, retreatment strategies remain inconsistent. AIM:To evaluate the reasons for and factors associated with non-retreatment after H. pylori eradication treatment failure. METHODS:A multicenter, prospective registry (European registry on H. pylori management) was used to evaluate the decisions and outcomes of H. pylori management by European gastroenterologists. Patients with at least one eradication failure between 2013 and 2024 were included and classified into retreatment (control) and non-retreatment. The reasons for non-retreatment were categorized as medical- or patient-related. Multivariate logistic regression identified factors associated with non-retreatment from each perspective. RESULTS:Of the 6904 patients with eradication failure, 950 (14%) were not retreated: 41% due to medical decisions, 50% due to patient decisions, and 9% due to other reasons. Frequent reasons for non-retreatment included previous poor tolerance, noncompliance, unclear indications, and multiple eradication attempts. From a medical perspective, non-retreatment was associated with age ≥ 71 years [odds ratio (OR) = 1.49; 95% confidence interval (CI): 1.04-2.13], previous noncompliance (OR = 4.27; 95%CI: 2.64-6.93), treatment discontinuation due to adverse events (OR = 2.06; 95%CI: 1.19-3.56), and number of previous attempts (OR range: 2.07-9.25). From the patient perspective, the associated factors included male sex (OR = 1.34; 95%CI: 1.10-1.63), two to four previous eradication attempts (OR range: 1.69-3.60) and previous noncompliance (OR = 10.04; 95%CI: 7.37-13.68). CONCLUSION:Up to 14% of European patients are not retreated after H. pylori eradication failure. Key barriers include advanced age, previous noncompliance and prior eradication failures. The decision to avoid retreatment was often patient-driven, although medical reasons were also relevant, highlighting the importance of shared decision-making and improved patient education.

Background: While large language models (LLMs) have demonstrated proficiency in medical examinations, their comparative performance on European gastroenterology assessments remains underexplored, particularly regarding architectural differences between reasoning and non-reasoning models. This study benchmarks five state-of-the-art LLMs-DeepSeek-R1, ChatGPT-o1, ChatGPT-4o, Gemini-1.5-Pro, and Llama-3.1-405B (All versions January 2025)-using 203 board-style questions from validated ESEGH preparation materials. Methods: Questions from two commercial ESEGH preparation banks were administered five times per model using standardized prompts. Accuracy, consistency, and domain-specific performance across clinical, diagnostic, and therapeutic questions were analyzed. Four practicing gastroenterologists validated human performance under uniform conditions. Results: ChatGPT-o1 achieved the highest overall accuracy at 84.0% (95% CI: 81.8-86.3), followed closely by ChatGPT-4o (81.7%), DeepSeek-R1 (79.0%), and Llama-3.1-405B (77.2%), while Gemini-1.5-Pro significantly underperformed with 68.5% accuracy (difference vs. ChatGPT-o1: 15.5 percentage points, 95% CI: 11.9 to 19.1, p < 0.01). Although all models exhibited high internal consistency ≥98.4% average agreement across repeated attempts, with 94.6-98.0% of questions answered identically in all five attempts), greater consistency did not necessarily correspond to higher accuracy. Domain-specific analysis revealed that diagnostic questions were answered most accurately, whereas clinical examination questions posed considerable challenges. Topic analysis demonstrated that questions on small intestine disorders were answered with the highest accuracy, in contrast to the lower performance observed in bariatric and pancreatic disorders. Notably, reasoning models, which employed explicit chain-of-thought strategies, outperformed non-reasoning counterparts (81.5% vs. 75.8%, difference: 5.7 percentage points, 95% CI: 3.4 to 8.0, p < 0.001), particularly on therapy questions and complex bait-and-switch formats. Practicing gastroenterologists achieved substantially lower accuracy (mean: 50.9%, range: 37.9-69.0%) compared to all LLMs. All models exceeded the current ESEGH passing threshold of 61.5%, with the top four models surpassing this benchmark by 15.7-22.5 percentage points. Conclusions: This benchmarking study demonstrates that current LLMs, particularly those with reasoning architectures, achieve high accuracy on European gastroenterology board-style questions. However, significant performance gaps in specific domains highlight limitations that must be addressed before clinical application. These findings provide a baseline for evaluating LLM capabilities in European medical contexts.

Gastrointestinal bleeding is a common and potentially life-threatening condition that requires prompt and effective management. Ankaferd blood stopper (ABS) is a plant-derived topical hemostatic agent that has recently attracted attention as a potential adjunctive tool in endoscopic hemostasis. This narrative review summarizes the available evidence regarding its mechanisms of action and clinical applications in gastrointestinal bleeding. Current data suggest that ABS may provide immediate bleeding control, particularly in refractory bleeding where conventional methods are insufficient or challenging. Reported clinical use indicates variceal and nonvariceal bleeding, malignancy-related hemorrhage, radiation-induced injury, and post-procedural bleeding. However, the available evidence is largely derived from case reports and small observational studies, limiting definitive conclusions regarding efficacy and safety. Overall, ABS appears to be a promising adjunctive or rescue option in selected clinical scenarios. Further well-designed prospective and comparative studies are required to better define its role within current endoscopic hemostatic strategies.

Background: Colorectal cancer (CRC) remains a significant global health burden, with early detection and intervention crucial for improving patient outcomes. This study aims to develop and evaluate a novel proof-of-concept ensemble framework combining transformer-based language models and decision tree-based models for early-stage CRC screening, diagnosis, and prognosis. Methods: The ensemble framework consists of four key components: (1) GastroGPT, a transformer-based language model for extracting relevant data points from patient histories; (2) a decision tree-based model for assessing CRC risk and recommending colonoscopy; (3) GastroGPT for extracting data points from early CRC patients' histories; and (4) a suite of decision tree-based models for predicting survival outcomes in early-stage CRC patients. The study employed a retrospective, observational, methodological design using simulated patient cases. Results: GastroGPT demonstrated high accuracy in extracting relevant data points from patient histories. The decision tree-based model for CRC risk assessment achieved an area under the receiver operating characteristic curve (AUC-ROC) of 0.85 (95% CI: 0.78-0.92) in predicting the need for colonoscopy. The decision tree-based models for survival prediction showed strong performance, with C-indices ranging from 0.71 to 0.75 for overall survival and disease-free survival at 24, 36, and 48 months. Conclusions: The novel ensemble framework demonstrates promising performance in early-stage CRC screening, diagnosis, and prognosis. Further research is needed to validate the models using larger, real-world datasets and to assess their clinical utility in prospective studies.

BACKGROUND:Surgery and percutaneous radiological methods [puncture, aspiration, injection, re-aspiration (PAIR)] are the current invasive treatment strategies for patients with hepatic hydatid cyst (HHC). Biliary leak is a common complication in patients who underwent these treatments of HHC. Bile leak should be treated effectively as uncontrolled biliary fistula may lead to life-treating conditions such as severe cholangitis, intraabdominal abscesses and septicemia. Endoscopic retrograde cholangiopancreatography (ERCP) has become the main treatment of post-interventional biliary fistula. AIM:To evaluate the efficacy and safety of ERCP in the management of biliary fistula following HHC-related surgery or PAIR. METHODS:We evaluated data of patients who developed bile leakage following HHC-related interventions from endemic area during the period of March 2017 and February 2025. We included 88 patients (50 female, 57%) with a median age of 33 years (range: 8-83 years) at the time of ERCP. Bile leak occurred following surgery in 72 (82%) patients and after PAIR in 16 (18%) patients. Low-grade leakage (< 400 mL/day) was identified in 46 (52%) patients. RESULTS:Initial mode of ERCP was endoscopic sphincterotomy (ES) with biliary drainage (plastic stent or nasobiliary drain) in 73 (83%) patients and ES alone in remaining 15 (17%) patients. Six patients who initially treated by ES alone had persistent fistula and underwent repeat ERCP with stent placement. ERCP type (ES + biliary stenting) and fistula flow rate (< 400 mL/day) were significantly associated with 20-days complete closure of the fistula [P = 0.020; odds ratio (OR) = 5.27, 95% confidence interval (95%CI): 1.30-21.37] and (P = 0.008; OR = 3.43, 95%CI: 1.37-8.55), respectively. ERCP-related complications were mild pancreatitis in 5 (5.9%) patients and minor bleeding in 4 (4.7%) patients and mild-moderate cholangitis in 4 (4.7%) patients. CONCLUSION:This case based-study from endemic area demonstrates that ERCP is highly effective and safe for managing bile leakage following both surgery and PAIR. ES + biliary stenting seems better mode of ERCP procedure.

Background/Aims: Inflammatory bowel diseases (IBDs) are increasingly prevalent and challenging globally. Data regarding IBD frequency and severity between Europe and Asia are needed. The aim was to investigate the trend of IBD in Türkiye during the last 3 decades. Materials and Methods: The study was conducted retrospectively at 14 centers in Türkiye between June 1993 and March 2024. Results: Over 30 years, 4308 patients, of whom 2507 (58.2%) had ulcerative colitis (UC) and 1717 (39.9%) had Crohn’s Disease (CD), were included in the study. The overall median age at the onset of IBD was 34.43 (25.28-45.64) years; the age at onset of IBD was younger in CD compared to UC (32.72 vs. 35.52 years respectively, P < .001). The peak age onset range was 28-32 years in CD, whereas 23-27 years in UC. Overall, 2526 (58.6%) patients were male. The most common location was left-sided colitis in UC patients (45.1%), followed by extensive colitis (31.2%), and proctitis (23.7%), while ileal involvement in CD patients (45.2%), afterward ileocolonic (409%), and colonic (13.5%). Both illnesses are becoming increasingly prevalent. The UC/CD ratio tends to decrease over time. During the study period, 1577 (36.6%) patients received biologic treatment. During the study period, 418 (24.3%) underwent resective surgery for CD and 88 (3.5%) total colectomy for UC; the major abdominal surgery has declined over time. Conclusion: The frequency and characteristic features of IBD in Türkiye appear to be between Europe and Asia. Over time, while the usage of biologic therapy and the rate of CD have increased, the frequency of surgery has decreased. Cite this article as: Durak MB, Cagin YF, Balkan A, et al. Three-decade analysis of inflammatory bowel disease in Türkiye: A multicenter study (1993-2024). Turk J Gastroenterol. 2025;36(12):822-833.

Background/Objectives: Acute pancreatitis (AP) is an illness that requires prompt diagnosis and treatment since it has the potential to become life-threatening. The American College of Gastroenterology 2024 (ACG24) guidelines offer a framework for diagnosis, severity, and treatment criteria. To assess Google Gemini application of ACG24 guidelines to Medical Information Mart for Intensive Care-III AP cases for risk, nutrition, and complication management. Methods: This observational cross-sectional study was based on 512 patients with AP who were treated in the Medical Information Mart for Intensive Care-III database from 2001 to 2012. The study compared the efficiency of Gemini in relation to the ACG24 guidelines in the three main areas of risk stratification, enteral nutrition timing, and necrotizing pancreatitis management. Enteral nutrition, according to the ACG24 guidelines, should be started within 48 h for patients who are capable, and antibiotics should only be used for confirmed infected necrosis. Results: The study included 512 patients who were divided into two groups: 213 patients with mild pancreatitis (41.6%) and 299 patients with severe pancreatitis (58.4%). The model achieved 85% accuracy for mild cases and 82% accuracy for severe cases of pancreatitis. The Acute Physiology and Chronic Health Evaluation II and Ranson scores matched the predictions of Gemini for both mild cases (p = 0.28 and p = 0.33, respectively) and severe cases (p = 0.31 and p = 0.27, respectively). The recommendations for early enteral nutrition and delayed feeding in mild cases were correct for 78% of patients, but the system suggested oral intake prematurely in 8% of severe cases. The antibiotic guideline compliance reached 82% among 156 patients with necrotizing pancreatitis, and the procedure for draining infected necrosis was correct 85% of the time. Conclusions: The Gemini model achieved 78-85% accuracy in determining pancreatitis severity and adherence to treatment guidelines but showed lower accuracy in nutrition timing compared to other parameters. Core Tip: This study evaluated the Google Gemini model in applying the American College of Gastroenterology 2024 guidelines for acute pancreatitis across 512 Medical Information Mart for Intensive Care-III cases. Results demonstrated 85% accuracy in severity classification, precise prediction of Acute Physiology and Chronic Health Evaluation II and Ranson scores, and 78-85% compliance with nutritional and necrotizing pancreatitis management guidelines. These findings suggest that artificial intelligence-based clinical decision support systems can provide rapid, consistent, and guideline-concordant recommendations, which are particularly valuable in settings with limited specialist expertise.

This review highlights gender gaps in training, career, and work-life balance in gastroenterology. Stereotypes and biases toward women's abilities and commitment to their careers can influence evaluations, advancement in gastroenterology training, and career progression. The findings indicate that lack of or limited access to mentorship and sponsorship, as well as support networks, can hinder the professional development of women. Moreover, results indicate that we must improve work-life balance measures, for example offering flexible working hours and compensation and support for women during pregnancy, after childbirth, and motherhood. However, reports on such equity measures are scarce, and we lack scientific evidence of their impact. This review concludes that to reduce gender gaps and make a positive impact, we need educational and promotional programs and monitoring of their outcomes.