Belén Agudo Castillo

Metabolic dysfunction-associated steatotic liver disease (MASLD) is increasingly recognised as a systemic disorder with clinically relevant extrahepatic manifestations. Pancreatic diseases have gained particular attention in this context, as they share key metabolic and inflammatory mechanisms with MASLD and may influence each other’s natural history and clinical outcomes. Beyond shared cardiometabolic risk factors, emerging evidence suggests that MASLD may actively contribute to pancreatic dysfunction and disease progression, supporting a bidirectional liver–pancreas relationship. Recognition of this association is essential to improve risk stratification, diagnostic assessment, and clinical management. This review summarises the current evidence linking MASLD with pancreatic disorders, focusing on acute and chronic pancreatitis, pancreatic cancer, and fatty pancreas. We examine shared pathophysiological pathways and discuss how these interactions may affect disease progression, nutritional status, diagnostic strategies, and therapeutic decision-making. We also address key clinical challenges, including alcohol assessment and MASLD–MetALD misclassification, highlight current knowledge gaps, and provide scenario-specific recommendations.

Background/Objectives: Accurate mediastinal and hilar lymph node (LN) assessment is central to lung cancer staging and treatment selection. Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) has become the standard minimally invasive procedure for mediastinal staging, enabling real-time identification, characterization and sampling of LNs. Although effective, this technique remains operator dependent, and subject to interobserver variability. This study aims to report the development of a Vision Transformer-based AI model for automatic LN detection and classification on EBUS images. Methods: A multicenter, multidevice, retrospective study, using EBUS images of three referral centers (Spain and Brazil). The dataset included images of patients with benign and malignant LNs. Bounding-box annotations were applied for LN localization, and LNs were classified as benign or malignant according to histopathologic reference standards. Detection performance was evaluated with mean average precision, at an intersection over union threshold of 0.5 (mAP50). Characterization performance was assessed using accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC). Results: The dataset included 6492 images from 27 patients, of whom 16 had malignant LNs. In the test set, the model achieved a mAP50 of 0.79 for LN detection. For malignant characterization, the model achieved an accuracy of 90.6%, sensitivity of 85.4%, and an AUROC of 0.938. Conclusions: The results of this study support the feasibility of AI-assisted EBUS interpretation for nodal staging workflows and raise the possibility that AI-assisted EBUS could enhance diagnostic performance, while reducing operator variability. Larger prospective studies are needed before clinical implementation.

Cholangiocarcinoma (CCa) is a complex malignancy of the biliary tract, classified as intrahepatic, perihilar, or distal. Digital single-operator cholangioscopy (D-SOC) enhances evaluation of biliary strictures, although it remains limited by suboptimal biopsy yield and technical constraints. Artificial intelligence (AI), particularly convolutional neural networks (CNNs), has emerged as a promising adjunct. However, performance across anatomical subtypes is not well defined. This study evaluated diagnostic performance of an AI-based model in detecting CCa lesions by location. A YOLOv8-based CNN was trained and validated using 315,993 D-SOC images from 183 patients across six international high-volume centers. Images were labeled as benign or malignant based on expert consensus. Frame-based analysis assessed diagnostic performance using macro-average F1-score, precision, and recall. Subgroup analysis explored anatomical site-specific performance. Among the included patients (mean age 66.1±12.1 years; 64.5% male), 43.2% had perihilar, 37.2% intrahepatic, and 19.7% distal biliary strictures. The model demonstrated high overall performance: F1 score 95.3%, precision 95.5%, and recall 95.1%. Site-specific analysis revealed an F1 score of 89.6% for distal strictures and 91.1% for perihilar strictures. Receiver operating characteristic curves showed areas under the curve of 0.980 for intrahepatic strictures and 0.990 for perihilar and distal strictures. This is the first study to demonstrate AI-based diagnostic performance across CCa topography using a large, multicenter D-SOC dataset. Although anatomical complexity affects detection, the model’s high precision and generalizability suggest potential for clinical utility. These results support application of an AI-enhanced algorithm decision algorithm for cholangioscopy.

Background/Objectives: Endoscopic ultrasound (EUS) is crucial for lymph node (LN) characterization, playing a key role in oncological staging and treatment guidance. EUS criteria for predicting malignancy are imprecise, and histologic diagnosis may have limitations. This multicenter study aimed to evaluate the effectiveness of a novel artificial intelligence (AI)–based system in predicting LN malignancy from EUS images. Methods: This multicenter study included EUS images from nine centers. Lesions were labeled (“malignant” or “benign”) and delimited with bounding boxes. Definitive diagnoses were based on cytology/biopsy or surgical specimens and, if negative, a minimum six-month clinical follow-up. A convolutional neural network (CNN) was developed using the YOLO (You Only Look Once) architecture, incorporating both detection and classification modules. Results: A total of 59,992 images from 82 EUS procedures were analyzed. The CNN distinguished malignant from benign lymph nodes with a sensitivity of 98.8% (95% CI: 98.5–99.2%), specificity of 99.0% (95% CI: 98.3–99.7%), and precision of 99.0% (95% CI: 98.4–99.7%). The negative and positive predictive values for malignancy were 98.8% and 99.0%, respectively. Overall diagnostic accuracy was 98.3% (95% CI: 97.6–99.1%). Conclusions: This is the first study evaluating the performance of deep learning systems for LN assessment using EUS imaging. Our AI-powered imaging model shows excellent detection and classification capabilities, emphasizing its potential to provide a valuable tool to refine LN evaluation with EUS, ultimately supporting more tailored, efficient patient care.

Introduction: Pancreatic exocrine insufficiency (PEI) results from a reduction in pancreatic secretion of enzymes, leading to malabsorption of nutrients, intestinal symptoms, nutritional deficiencies and related comorbidities. The diagnosis of pancreatic exocrine insufficiency should be based on digestive tests, mainly the coefficient of fat absorption (CFA), based on the quantification of 72 h fecal fat excretion (FFE). However, this test is rarely performed in clinical practice. Fecal elastase‐1 (FE‐1) is a simple and widely used alternative. This meta‐analysis evaluates the diagnostic accuracy of fecal elastase‐1 for the diagnosis of PEI diagnosed by CFA or 72h‐FFE.Methods: A systematic search of databases was performed to identify studies evaluating fecal elastase‐1 and CFA/FFE for the diagnosis of pancreatic exocrine insufficiency. Inclusion criteria required original studies with data on sensitivity, specificity and other diagnostic metrics. Two independent reviewers performed data extraction and quality assessment using the QUADAS‐2 tool. Pooled sensitivity, specificity, likelihood ratios and diagnostic odds ratio (DOR) were calculated and heterogeneity was assessed using I‐squared tests.Results: Thirteen studies with 888 patients were included. Fecal elastase‐1 at a cut‐off of 200 μg/g showed a pooled sensitivity and specificity of 0.94 and 0.69, respectively, with a DOR of 35.27. Lowering the cut‐off to 100 μg/g improved specificity to 0.82 but decreased sensitivity to 0.88. Subgroup analyses showed different diagnostic performance in different clinical contexts, with higher sensitivity in cystic fibrosis (0.98) and higher specificity in chronic pancreatitis (0.81). The positive and negative predictive values are limited in situations with low and high probability of pancreatic exocrine insufficiency, respectively.Conclusions: Fecal elastase‐1 is a sensitive and moderately specific diagnostic tool for pancreatic exocrine insufficiency and is suitable for initial screening in high‐risk populations. However, its moderate specificity requires careful interpretation in lower risk settings.

We read with interest the study by de Jong et al. [1], which demonstrated that indwelling double-pigtail stents significantly reduced the recurrence of pancreatic fluid collections (PFCs) compared with stent removal or migration (10.8% vs. 25.5%). These findings align with the meta-analysis by Hawa et al. [2], which similarly reported a reduction in PFC recurrence with long-term indwelling plastic stents; however, they stand in contrast to the randomized controlled trial by Chavan et al. [3], which found no significant benefit from maintaining plastic stents after large-caliber metal stent removal in patients with disconnected pancreatic duct syndrome (DPDS). Although the robust analysis and large cohort in the de Jong study are strengths, the median follow-up of 8.2 months for patients with an indwelling plastic stent may be insufficient to fully assess the long-term complications. Of note, the meta-analysis by Hawa et al. [2] reported an 8% adverse event rate with long-term indwelling plastic stents, including cases that required surgical intervention. The technical approach also warrants attention: placing two or more stents appears more effective, but raises questions about the optimal number to maximize benefits while minimizing risks. A critical issue is the clinical significance of PFC recurrence. The trial reported by Chavan et al. [3] showed that many recurrences were asymptomatic and did not require further intervention. Similarly, de Jong et al. [1] found that only 7.8% of patients required drainage of their recurrent PFCs. In light of these findings, future research should focus on: Future prospective studies that address these aspects could help in establishing evidence-based guidelines for the management of long-term stenting in patients with walled-off necrosis and DPDS. Article published online: 22 April 2025 © 2025. Thieme. All rights reserved. Georg Thieme Verlag KG Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany

We read with great interest the article “Correlation of endoscopic ultrasound-guided portal pressure gradient measurements with hepatic venous pressure gradient: a prospective study” by Martinez-Moreno et al. [1]. The study provides novel insights into the potential of the endoscopic ultrasound-guided portal pressure gradient (EUS-PPG) as an alternative to hepatic venous portal gradient (HVPG) for measuring portal pressure in cirrhotic patients, demonstrating a strong correlation (intraclass correlation coefficient of 0.82) between EUS-PPG and HVPG. However, significant discrepancies (≥5 mmHg) were found in four subjects, 75% of whom had metabolic dysfunction-associated steatotic liver disease (MASLD). In other words, almost half of MASLD patients exhibited notable portal pressure discrepancies between the two techniques. Previous studies have demonstrated a presinusoidal component of portal pressure in MASLD, attributed to sinusoidal structural and functional disturbances secondary to fatty liver features [2]. This presinusoidal pressure develops also in precirrhotic stages, potentially increasing portal hypertension, and is not fully captured by noninvasive tests or portal hemodynamics, as HVPG mainly reflects sinusoidal and post-sinusoidal pressures [3] [4]. Importantly, similarly to Martinez-Moreno et al., another study measuring portal pressure through direct portal vein puncture during transjugular intrahepatic portosystemic shunt (TIPS) placement found greater disagreement with HPVG measurements in patients with MASLD compared with those with hepatitis C virus [5]. These findings suggest that direct portal pressure measurements should be considered as an alternative to HVPG in patients without MASLD, given their excellent concordance. Further research is needed to assess the ability of portal hemodynamics, particularly EUS-PPG, to accurately capture presinusoidal portal pressure in MASLD. This approach would enhance the reliability of portal pressure assessments, facilitating the interpretation of results by elucidating the clinical significance of discrepancies between measurements. Consequently, it would refine the liver decompensation risk prediction and improve the management of MASLD patients. Article published online: 26 February 2025 © 2025. Thieme. All rights reserved. Georg Thieme Verlag KG Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany

Background: Several artificial intelligence systems based on large language models (LLMs) have been commercially developed, with recent interest in integrating them for clinical questions. Recent versions now include image analysis capacity, but their performance in gastroenterology remains untested. This study assesses ChatGPT-4’s performance in interpreting gastroenterology images. Methods: A total of 740 images from five procedures—capsule endoscopy (CE), device-assisted enteroscopy (DAE), endoscopic ultrasound (EUS), digital single-operator cholangioscopy (DSOC), and high-resolution anoscopy (HRA)—were included and analyzed by ChatGPT-4 using a predefined prompt for each. ChatGPT-4 predictions were compared to gold standard diagnoses. Statistical analyses included accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under the curve (AUC). Results: For CE, ChatGPT-4 demonstrated accuracies ranging from 50.0% to 90.0%, with AUCs of 0.50–0.90. For DAE, the model demonstrated an accuracy of 67.0% (AUC 0.670). For EUS, the system showed AUCs of 0.488 and 0.550 for the differentiation between pancreatic cystic and solid lesions, respectively. The LLM differentiated benign from malignant biliary strictures with an AUC of 0.550. For HRA, ChatGPT-4 showed an overall accuracy between 47.5% and 67.5%. Conclusions: ChatGPT-4 demonstrated suboptimal diagnostic accuracies for image interpretation across several gastroenterology techniques, highlighting the need for continuous improvement before clinical adoption.

Endocuff Vision (EV) is a disposable device designed to improve polyp detection. The primary aim of this study was to evaluate the effect of EV on adenoma detection rate (ADR) and mean number of adenomas per patient (MAP). This multicenter randomized controlled trial compared EV-assisted colonoscopy (EAC) and standard colonoscopy (SC). Patients were referred due to a positive fecal immunochemical test (FIT) in a bowel cancer screening program (BCSP), direct screening colonoscopy without prior FIT, surveillance colonoscopy, or family history of colorectal cancer. In total, 1437 patients (55.9% male; median age, 59 years) were randomized at eight Spanish university hospitals. No significant differences were found in either the ADR (EAC vs. SC, 55.8% vs. 54.3%; P = 0.576) or the MAP (1.60 vs. 1.35; P = 0.03). Compared with SC, EAC was not associated with a significant improvement in detection rates for advanced adenomas, sessile serrated lesions, or advanced sessile serrated lesions. There was no difference in cecal intubation rate but the successful ileal intubation rate, among patients in whom it was attempted, was lower with EAC (64.3% vs. 86.5% with SC; P = 0.001). EV did not improve either ADR or MAP. EV may increase difficulty of ileal intubation whenever it is attempted.

INTRODUCTION: Endoscopic ultrasound (EUS) allows for characterization and biopsy of pancreatic lesions. Pancreatic cystic neoplasms (PCN) include mucinous (M-PCN) and nonmucinous lesions (NM-PCN). Pancreatic ductal adenocarcinoma (P-DAC) is the commonest pancreatic solid lesion (PSL), followed by pancreatic neuroendocrine tumor (P-NET). Although EUS is preferred for pancreatic lesion evaluation, its diagnostic accuracy is suboptimal. This multicentric study aims to develop a convolutional neural network (CNN) for detecting and distinguishing PCN (namely M-PCN and NM-PCN) and PSL (particularly P-DAC and P-NET). METHODS: A CNN was developed with 378 EUS examinations from 4 international reference centers (Centro Hospitalar Universitário São João, Hospital Universitario Puerta de Hierro Majadahonda, New York University Hospitals, Hospital das Clínicas Faculdade de Medicina da Universidade de São Paulo). About 126.000 images were obtained—19.528 M-PCN, 8.175 NM-PCN, 64.286 P-DAC, 29.153 P-NET, and 4.858 normal pancreas images. A trinary CNN differentiated normal pancreas tissue from M-PCN and NM-PCN. A binary CNN distinguished P-DAC from P-NET. The total data set was divided into a training and testing data set (used for model's evaluation) in a 90/10% ratio. The model was evaluated through its sensitivity, specificity, positive and negative predictive values, and accuracy. RESULTS: The CNN had 99.1% accuracy for identifying normal pancreatic tissue, 99.0% and 99.8% for M-PCN and NM-PCN, respectively. P-DAC and P-NET were distinguished with 94.0% accuracy. DISCUSSION: Our group developed the first worldwide CNN capable of detecting and differentiating the commonest PCN and PSL in EUS images, using examinations from 4 centers in 2 continents, minimizing the impact of the demographic bias. Larger multicentric studies are needed for technology implementation.