Introduction

The Maastricht VI/Florence Consensus recommends eradication of all Helicobacter pylori (Hp) infections. However, retreatment strategies after therapeutic failure remain inconsistent.

Aims & Methods

To evaluate the reasons and associated factors for non-retreatment after a failure of Hp eradication treatment.
Systematic, prospective, multicenter, non-interventional registry of clinical practice by European gastroenterologists on the management of Hp infection (Hp-EuReg). Patients who failed eradication treatment between 2013-2024 and were registered in AEG-REDCap were analysed and categorised into two groups: retreatment (control group) and non-retreatment. Reasons for no retreatment were also classified according to either the medical or the patient perspective. Multivariate logistic regression was conducted to identify factors associated with no retreatment from both the patient and medical perspectives, each compared with the control group.

Results

Overall, 6,904 patients were evaluated; 950 (14%) were not retreated: 41% due to a medical decision, 50% due to a patient decision, and 9% due to other reasons. The most frequent reasons for no retreatment were: previous poor tolerance, unclear indication, multiple prior eradication attempts, and previous non-compliance. From the medical perspective, factors significantly associated with no retreatment were: age ≥ 71 years (OR 1.49; 95% CI 1.04-2.13), previous non-compliance (OR 4.27; 95% CI 2.64-6.93), number of previous attempts (OR 2.07-9.25; 95% CI from 1.57-2.73 to 6.15-13.93) and previous treatment discontinuation due to adverse events (OR 2.06; 95% CI 1.19-3.56). From the patient perspective, the factors significantly associated with no retreatment included: being male (OR 1.34; 95% CI 1.10-1.63), having undergone two to four previous eradication attempts (OR 1.69-3.60; 95% CI from 1.16-2.44 to 1.75-7.41) and previous non-compliance (OR 10.04; 95% CI 7.37-13.68).

Factors associated with the lack of H. pylori retreatment after a therapeutic failure – Multivariate analysis
	Retreatment N=5,954	No retreatment Overall N=950	P value OR (95% CI)if applicable	No retreatment Medical perspective N=393	P value OR (95% CI)if applicable	No retreatment Patient perspective N=474	P value OR (95% CI)if applicable
Sex Male Female	2,100 (35.3%) 3,850 (64.7%)	386 (40.6%) 564 (59.4%)	0.000 1.30 (1.12-1.50)	147 (37.4%) 246 (62.6%)	0.116 -	206 (43.5%) 268 (56.5%)	0.004 1.34 (1.10-1.63)
Age 18-70 > 70	5,455 (91.6%) 475 (7.9%)	851 (89.5%) 96 (10.1%)	0.015 1.35 (1.06-1.72)	350 (89.1%) 40 (10.2%)	0.028 1.49 (1.04-2.13)	430 (90.7%) 44 (9.3%)	0.282 -
Number of previous eradication attempts One Two Three Four Five	1.054 (17.7%) 334 (5.6%) 115 (1.9%) 42 (0.7%) 28 (0.5%)	181 (19.1%) 105 (11.1%) 63 (6.6%) 27 (2.8%) 13 (1.4%)	p < 0.05 in all 1.57 (1.30-1.89) 2.82 (2.21-3.61) 5.05 (3.63-7.03) 6.00 (3.61-9.96) 4.04 (4.97-7.41)	82 (20.9%) 61 (15.5%) 39 (9.9%) 12 (3.1%) 5 (1.3%)	p < 0.05 in all 2.07 (1.57-2.73) 4.89 (3.54-6.74) 9.25 (6.15-13.93) 7.84 (3.96-15.53) 4.73 (1.75-12.77)	82 (17.3%) 26 (5.5%) 21 (4.4%) 10 (2.1%) 5 (1.1%)	p < 0.05 only in 2-4 attempts - 1.69 (1.16-2.44) 2.91 (1.78-4.76) 3.60 (1.75-7.41) -
Treatment compliance*	5,647 (94.8%)	739 (77.8%)	0.000 6.07 (4.97-7.41)	303 (77.1%)	0.000 4.27 (2.64-6.93)	358 (75.5%)	0.000 10.04 (7.37-13.68)
Interruption due to AE	183 (3.1%)	109 (11.5%)	0.053 -	68 (17.3%)	0.010 2.06 (1.19-3.56)	39 (8.2%)	0.000 0.34 (0.22-0.55)

OR: odds ratio. CI: confidence interval. AE: adverse events.
* Compliance: > 90% of drug intake.

Conclusion

A significant proportion (14%) of European patients are not retreated after Hp eradication failure. Key barriers include advanced age, a history of poor treatment adherence, and multiple prior eradication failures. The decision to avoid re-treatment was frequently driven by patient-related factors, although medical reasons also played a relevant role, highlighting the importance of shared decision-making and enhanced patient education.

References

1. Gisbert JP, Alcedo J, Amador J, Bujanda L, Calvet X, Castro-Fernández M, et al. V Spanish Consensus Conference on Helicobacter pylori infection treatment. Gastroenterol Hepatol Engl Ed. mayo de 2022;45(5):392-417.
2. Malfertheiner P, Megraud F, Rokkas T, Gisbert JP, Liou JM, Schulz C, et al. Management of Helicobacter pylori infection: the Maastricht VI/Florence consensus report. Gut. septiembre de 2022;71(9):1724-62.
3. Banks M, Graham D, Jansen M, Gotoda T, Coda S, Di Pietro M, et al. British Society of Gastroenterology guidelines on the diagnosis and management of patients at risk of gastric adenocarcinoma. Gut. septiembre de 2019;68(9):1545-75.

Disclosure

Olga P. Nyssen has served as a speaker or has received research funding from Allergan, Mayoly Spindler, Richen, Biocodex and Juvisé.
Javier P. Gisbert has served as speaker, consultant, and advisory member for or has received research funding from Mayoly, Allergan/Abbvie, Diasorin, Richen, Juvisé, Biocodex.
The remaining authors declare no conflicts of interest.

Introduction

Acute biliary pancreatitis (ABP) has been shown to be associated with increasing incidence and health-care costs, and possibly higher mortality than other etiologies. Despite the existence of various scoring systems, early and accurate risk stratification remains difficult. Due to heterogeneity of populations, focusing on one etiology might reveal important insights concerning predictors of outcome.

Aims & Methods

We conducted a retrospective, multicenter cohort study including all patients admitted with ABP between 2018 and 2021 at two hospitals. Epidemiological, clinical, and outcome data were analyzed. The main clinical outcome of interest was a hierarchical composite endpoint of five mutually exclusive clinical states: (i) death (90-day all-cause mortality), (ii) severe AP and/or admission to the intensive care unit (ICU), (iii) endoscopic, interventional radiological, or surgical interventions, (iv) moderately-severe AP, (v) mild AP. Univariable logistic regression identified parameters associated with worse outcomes, which were then tested in multivariable models. Emphasis was placed on early availability and clinical practicality of predictors. We aimed to stratify patients into outcome trajectories reflecting both health impact and healthcare resource utilization.

Results

Out of 569 total acute pancreatitis cases, 217 (38.1%) patients had ABP (mean age 65.9 ±17.4 years; median length of stay (LOS) 7 days, IQR 4;11) Mild, moderately-severe, and severe disease occurred in 62.2%, 30.0%, and 7.8% of patients. All-cause 90-day mortality was 2.3%. Several parameters were significantly (p<0.05) associated with worse outcomes, including BISAP and the Charlson comorbidity index (CCI) (p<0.001). Individually, BISAP and CCI classified patients reasonably, however, wide prognostic uncertainty persisted. When combining BISAP and CCI (BISAP-C), 5 discrete risk groups were identified, ranging from low (group 1, BISAP 0-2 & CCI 0-2) to very-high risk (group 5, BISAP 3-5 & CCI ≥6) (see Table 1). The prevalence of severe outcomes (i–iii) increased across these strata: 2.4% (Group 1), 9.5% (Group 2), 27.3% (Group 3), 60% (Group 4), and 62.5% (Group 5), with LOS also significantly longer in higher-risk groups (p<0.001). Mortality was 50% in group 5, while, importantly, 97.6% of patients in group 1 experienced mild-to-moderate AP without ICU admission, persistent organ dysfunction, surgical or endoscopic interventions, and no fatality, supporting potential early discharge. This also occurred in 90.5% of group 2. Notably, Groups 1 and 2 together comprised nearly 80% of the cohort.

Table 1. Outcome trajectories	Outcome groups*
	Group 5 Very-high	Group 4 High	Group 3 Mod.-high	Group 2 Mod.-low	Group 1 Low
BISAP CCI	3-5 ≥6	3-5 ≤5	0-2 ≥6	0-2 3-5	0-2 ≤2
Cases, n (% of cohort)	8 (4.0%)	5 (2.5%)	22 (10.9%)	84 (41.6%)	83 (41.1%)
LOS, median (IQR)	5 (3;10)	15 (9;106.5)	10 (6;14.25)	7 (4;11)	5 (3;9)
Mortality, 90d all-cause	4 (50%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)
Severe pancreatitis or ICU admission	0 (0%)	3 (60%)	6 (27.3%)	6 (7.1%)	2 (2.4%)
Intervention (endosc., IR, surgical)	1 (12.5%)	0 (0%)	0 (0%)	2 (2.4%)	0 (0%)
Moderately-severe pancreatitis	2 (25.0%)	2 (40%)	7 (31.8%)	18 (21.4%)	24 (28.9%)
Mild pancreatitis	1 (12.5%)	0 (0%)	9 (40.9%)	58 (69.0%)	57 (68.7%)

Conclusion

Comorbidity seems to play a major role concerning worse outcomes in ABP. The combination of CCI and BISAP (BISAP-C), both comprised of early-on and easily available parameters, offers improved early prognostic discrimination and may provide tailored, individualized management and healthcare resource utilization.

Introduction

This study investigates the dynamic mechanisms by which Helicobacter pylori (Hp) eradication therapy influences gut microbiota and metabolome. Current rescue regimens, specifically high-dose dual therapy (HDDT) combining vonoprazan (VPZ) and amoxicillin, lack thorough exploration regarding their effects on gut microbiota and host metabolism. Additionally, the timing of probiotic intervention remains poorly understood.

Aims & Methods

Utilizing multi-omics approaches, this study aims to clarify how probiotics administered at different times influence gut microbial composition, metabolic profiles, and their interaction in patients undergoing Hp rescue therapy, thereby providing a theoretical foundation for enhancing Hp eradication strategies through microbial-metabolic modulation.
From January to December 2024, 120 patients undergoing Hp rescue therapy were randomized into four groups (A–D). Group A received VPZ 20 mg bid + amoxicillin 0.75 g qid for two weeks, followed by Bacillus subtilis dual-strain enteric-coated capsules (2 tablets tid) for two weeks. Group B received VPZ 20 mg bid + amoxicillin 0.75 g qid + Bacillus subtilis capsules (2 tablets tid) for two weeks. Group C was treated with VPZ 20 mg bid + amoxicillin 0.75 g tid for two weeks. Group D received Bacillus subtilis capsules (2 tablets tid) for two weeks, followed by VPZ 20 mg bid + amoxicillin 0.75 g qid for two weeks. Twenty-five Hp-negative individuals served as controls. Fecal samples were collected at baseline, 4 weeks, and 12 weeks post-treatment. Gut microbiota diversity and composition were analyzed through metagenomic sequencing; untargeted metabolomics was conducted using LC-MS; and small intestinal bacterial overgrowth was assessed via hydrogen-methane breath testing.

Results

1) No significant differences in gut microbiota alpha (Shannon index) or beta diversity (Bray-Curtis) were observed between H. pylori (Hp)-infected and uninfected groups (P=0.65). Hp-positive individuals exhibited elevated Collinsella aerofaciens and reduced Roseburia faecis and Agathobacter spp. (P<0.05). Metabolomic analysis identified 886 differentially abundant metabolites (P<0.05, VIP>1, FC≠1) enriched in 45 KEGG pathways, with key biomarkers including deoxyribose 5-phosphate and 4-vinylcyclohexene dioxide (P<0.01). The prevalence of SIBO was higher in patients with H. pylori infection than in those without infection (P<0.05). 2) Post-treatment alpha/beta diversity remained stable across groups A–D (P>0.05). Metabolic pathway profiles diverged: Group A emphasized phenylpropanoid biosynthesis; Group B showed cytochrome P450 drug metabolism; Group C featured sphingolipid metabolism and necroptosis; Group D upregulated cofactor/flavonoid biosynthesis. No significant SIBO changes occurred post-treatment in any group. 3) At week 4, beta diversity separation (P=0.009) was driven by Bacillota, Blautia, Ruminococcus sp., and Phocaeicola sp., with metabolites like Fahfa(3:0/24:1) linked to inflammation. By week 12, Groups B and C displayed distinct alpha (Sobs) and beta diversity (P=0.011), distinguished by Escherichia coli and Faecalibacterium prausnitzii. Group C exhibited upregulated anti-inflammatory metabolites (e.g., luteolin) and mucosal repair markers. 4) Multi-omics integration revealed early metabolic compensation at week 4 (Bifidobacterium-mediated amino acid/bile acid homeostasis) and late-stage recovery (week 12) marked by Pseudomonadota proliferation (E. coli) and Faecalibacterium grausatizii-associated sulfated steroids/phosphosphingolipids, supporting immune regulation and mucosal repair.

Conclusion

Probiotic-augmented HDDT has minimal impact on gut microbiota diversity and small intestinal bacterial overgrowth but significantly enhances the probiotic-microbiota-metabolite axis, particularly in sustaining anti-inflammatory metabolites and immune regulation.

Disclosure

The authors have no conflicts of interest to declare.

Introduction

Tissue acoustic properties, such as sound speed, attenuation, and impedance, are closely associated with its mechanical characteristics. Acoustic microscopy, which employs high-frequency ultrasound ranging from several tens of MHz to several GHz, enables the quantitative evaluation of these acoustic properties. Understanding the acoustic properties of biological tissues may offer new insights into disease pathology. Furthermore, because clinical ultrasound images depend on tissue acoustic properties, understanding them could improve the interpretation of echogenic patterns and contribute to developing new diagnostic approaches. Although acoustic microscopy has been applied to cardiovascular and cutaneous tissues in medical research, studies focusing on the liver, particularly hepatic tumours, remain limited.

Aims & Methods

This study aimed to quantify sound speed in hepatocellular carcinoma (HCC)—the most common primary liver tumour—and to examine its relationship with histopathological features. Histopathological classification followed World Health Organization criteria, including histological grade (well, moderately, and poorly differentiated) and histological growth pattern (trabecular, pseudoglandular, and solid). Tissue sound speed was measured using a scanning acoustic microscope (AMS-50SI; Honda Electronics Co., Ltd.) with an 80 MHz transducer. Regions of interest were identified on haematoxylin and eosin-stained slides. Serial 10 μm-thick sections were prepared from formalin-fixed, paraffin-embedded tissue blocks and rehydrated for measurement. Differences in sound speed among histological categories were compared, and statistical analysis was conducted to assess significance. A p-value of < 0.05 was considered significant. This study was approved by the institutional review board, and written informed consent was obtained.

Results

Tissue sound speed was measured at 66 sites from 51 patients. Histological grading identified 22 well, 41 moderately, and 3 poorly differentiated cases. Median sound speeds were 1584.0 m/s, 1595.9 m/s, and 1550.1 m/s, respectively. Poorly differentiated HCC showed significantly lower sound speed than the other grades (p = 0.021). Growth patterns included 50 trabecular, 10 pseudoglandular, and 6 solid cases, with median sound speeds of 1595.9 m/s, 1586.0 m/s, and 1563.6 m/s, respectively. The trabecular pattern tended to exhibit higher sound speed than the solid pattern (p = 0.063). Tissues with abundant fat or glycogen (n = 16) had significantly lower sound speeds than those without (p = 0.002). After excluding these cases, which could act as potential confounding factors, sound speed in poorly differentiated HCC remained significantly lower (p = 0.016), and higher values in the trabecular pattern persisted. Additionally, the pseudoglandular pattern showed greater variability in sound speed than the trabecular pattern.

Conclusion

Variations in acoustic properties among histological subtypes of HCC may contribute to developing novel quantitative assessments in ultrasound imaging and pathological diagnosis.

Disclosure

Kazuto Kobayashi, a co-author, has reported a relationship with Honda Electronics Co., Ltd. that includes: employment.

Introduction

Autism Spectrum Disorder (ASD) is closely associated with gut-brain axis dysfunction[1]. Fecal microbiota transplantation (FMT) may alleviate gastrointestinal and behavioral symptoms in ASD children by remodeling gut microbiota[2], yet the key functional bacterial strains underlying FMT-mediated neurobehavioral regulation remain unidentified.

Aims & Methods

This study aims to systematically evaluate FMT efficacy through a cohort study, combined with metagenomic analysis to investigate the dynamic patterns of microbiota colonization in ASD children. We will thoroughly identify core bacterial strains and analyze the relationship between their colonization patterns and therapeutic outcomes. The children aged 0-18 years meeting the diagnostic criteria for ASD were included in this study. The clinical data of recipients was collected before FMT, alongside questionnaire surveys and fecal sample collection. Pre-FMT assessments were conducted by pediatric psychologists using standardized scales. Donor feces were extracted via standardized protocols, and 100 mL of donor microbiota suspension was administered into the jejunum or ileum via colonoscopy or nasojejunal tube once every other day, total three sessions. Post-FMT fecal samples were collected from recipients at 1, 2, 4, 12, and 24 weeks for metagenomic analysis at species and strain levels. At 12 and 24 weeks post-FMT, the same psychologists reevaluated recipients using assessment scales, while parents completed follow-up questionnaires to document clinical outcomes. Finally, correlation analysis was conducted between laboratory data and clinical outcomes.

Results

A total of age-matched healthy pediatric donors (86 fecal samples) and 52 ASD children (351 fecal samples) were included. Results revealed that 69.2% of ASD children exhibited gastrointestinal (GI) symptoms (e.g., constipation or diarrhea), 58.3% had one or more food protein allergies, and 78.5% experienced sleep disturbances (<8 hours/day). Post-FMT, GI symptoms improved most significantly: 80% of recipients showed at least one grade reduction in Gastrointestinal Symptom Rating Scale (GSRS) scores, while 60% demonstrated improvements in autism-related behavioral traits, including reduced emotional dysregulation/negativity and enhanced sensory responsiveness. Clinical improvements became more pronounced over time. Metagenomic analysis demonstrated a significant increase in gut microbiota diversity and structural remodeling in recipients, with microbial profiles converging toward donor compositions. Correlation analysis identified donor-derived Bacteroidetes and Bacillota phyla as closely associated with clinical improvements.

Conclusion

FMT can significantly and 'long-term' reconstruct the gut microbiota of children with autism, effectively alleviating concurrent symptoms such as gastrointestinal disorders. Successful colonization of Bacteroidetes and Bacillota from donor sources is closely related to clinical improvement.

References

1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. American Psychiatric Association; 2013.
2.Dossaji Z, Khattak A, Tun KM, Hsu M, Batra K, Hong AS. Efficacy of Fecal Microbiota Transplant on Behavioral and Gastrointestinal Symptoms in Pediatric Autism: A Systematic Review. Microorganisms. 2023 Mar 22;11(3):806.

Introduction

Common bile duct (CBD) stones are one of the most frequent indications for endoscopic retrograde cholangiopancreatography (ERCP). Endoscopic complete sphincterotomy (EST) followed by balloon or basket extraction is widely considered the gold standard technique for the treatment of CBD stones; being successful in more than 90% of cases with approximately 5% overall rate of adverse events [1]. Endoscopic papillary large balloon dilatation (EPLBD) is an effective technique in treating large common bile duct (CBD) stones. However, the efficacy and safety of EPLBD in the setting of periampullary diverticula (PAD) are prospectively lacking.

Aims & Methods

This study aimed to compare the effectiveness and short-term safety of EPLBD after limited sphincterotomy with those of EST for extracting large CBD stones in the setting of PAD prospectively. Forty-eight patients with large CBD stone(s) > 1 cm were recruited. All patients had a PAD. Before the procedure patients were evaluated by history taking, clinical examination full investigations including proper imaging studies. During the procedure, patients were randomized to either group I; patients treated with EPLBD after limited sphincterotomy, and group II; patients treated with EST. CBD clearance was defined as a single procedure (EPLBD vs EST), single session (no repeat ERCP), and non-assisted (no lithotripsy used) stone extraction. After the procedure patients were hospitalized with clinical, laboratory, and imaging assessment.

Results

Table 1: Univariate regression of the parameters affecting stone extraction

PAD: periampullary diverticulum			* significant as P value≤ 0.05
	Odds ratio	95% CI	P value​
The used technique	5.92	1.32 to 25.3	0.009∗
Stone (s) number	0.62	0.17 to 2.25	0.468
Stone (s) diameter (mm)	2.59	0.63 to 10.56	0.187
Type of PAD​	1.32	0.65 to 2.67	0.437

The overall stone extraction rate in the current study was only 77.3% (n=34). The efficacy of stone extraction was significantly higher in the EPLBD (n= 21,87.5%) group compared to EST (n= 13, 54.17%) group (p = 0.011). The procedure-related adverse events were lower in the EPLBD group compared to the EST group. Pancreatitis was significantly higher in the EST (n=14, 58.33%) group than in the EPLBD (n=3, 12.5%) group (p= 0.001). There were no perforations or case fatalities in either group. The predictors of stone extraction were analyzed using univariate regression (Table 1), and it is noticeable that the used technique was the independent predictor of success in stone extraction (Odds Ratio: 5.92, 95% Confidence Interval: 1.38 to 25.30, p-value = 0.009).

Conclusion

The current practice guidelines of the ESGE and ASGE suggest that for large CBD stones, an EST followed by a large balloon dilatation is preferable to EST alone because combining dilatation with EST improves duct clearance and reduces the need for lithotripsy without adding risk. Our results among PAD patients seem to echo this: the EPLBD approach in comparison to EST yielded higher stone clearance on first attempt even without using lithotripsy by the study design. Hence, based on the current study these existing recommendations can be extrapolated to the PAD population based on our study which is one of the first randomized trials in the specific PAD subgroup.

References

1. Zippi M, Traversa G, Pica R, De Felici I, Cassieri C, Marzano C, et al. Efficacy and safety of endoscopic retrograde cholangiopancreatography (ERCP) performed in patients with Periampullary duodenal diverticula (PAD). Clin Ter. 2014;165(4):e291-4.

Introduction

Common bile duct (CBD) dilation can be associated with a wide range of underlying conditions, from benign to significant pathology, including malignancy¹. A recently proposed score by Raza et al. aims to predict the likelihood of significant findings on endoscopic ultrasound (EUS) in patients with incidental CBD dilation², potentially optimizing its use. This study aims to validate this score in a cohort of patients who previously underwent EUS for incidental CBD dilation.

Aims & Methods

Patients with incidental CBD dilation (≥6mm or ≥8mm post-cholecystectomy) who underwent EUS and met the inclusion criteria were retrospectively analyzed. Patient demographics, clinical, laboratory and imaging data were collected to calculate the score: age ≥70 years (1 point), acute non-billiary type abdominal pain (2 points), previous endoscopic retrograde cholangiopancreatography (2 points), CBD diameter ≥15mm or ≥17mm if prior cholecystectomy (2 points). EUS findings were reviewed to identify outcomes.

Results

Sixty-five patients met the inclusion criteria for analysis, 39 females (60%) and 26 males (40%), mean age of 71.5 ± 8.5 years. Overall yield of EUS for significant findings was 21.5%, including 8 patients with choledocolithiasis (12.3%), 4 patients with ampullary adenoma (6.2%), 1 patient with a biliary mass (1.5%) and 1 patient with a pancreatic mass (1.5%). Among the cohort, 23 patients (35.4%) had a score of 0 and none had a significant finding. Thus, a score ≥1 had a sensitivity of 100% and specificity of 45.1%, reliably excluding significant findings. A score ≥3 had a sensitivity of 71.4% and a specificity of 94.1% for identifying meaningful pathology, enabling the identification of high-risk patients with greater precision.

Conclusion

The predictive score effectively stratifies patients with incidental CBD dilation, potentially reducing unnecessary EUS in low-risk individuals while prioritizing high-risk cases. This model enhances resource utilization and targeted care. Further research in larger cohorts may confirm its generalizability.

References

1. Kelly NM, Tharian B, Tham TCK. Natural history of unexplained biliary dilation. Gut 2011;60(Suppl 1):A32-42.
2. Raza, Z., Altayar, O., Mahmoud, T., Abusaleh, R., Ghazi, R., Early, D., Kushnir, V. M., Lang, G., Sloan, I., Hollander, T., Rajan, E., Storm, A. C., Abu Dayyeh, B. K., Chandrasekhara, V., & Das, K. K. (2024). Clinical predictors of significant findings on EUS for the evaluation of incidental common bile duct dilation.Gastrointestinal Endoscopy,100(4), 670–678.e1. https://doi.org/10.1016/j.gie.2024.04.2902