Introduction

PolyDeep is a computer-aided diagnosis system designed for both the detection and characterization (CADe/x) of colorectal lesions during colonoscopy. It has been evaluated in three clinical trials focused on polyp detection.

Aims & Methods

We aimed to compare the diagnostic accuracy of endoscopists and PolyDeep for the optical diagnosis of colorectal lesions (i.e. neoplasia and adenoma) during colonoscopy. We conducted a post-hoc analysis within the three PolyDeep Advance trials, including all lesions detected during AI-assisted colonoscopies with confirmed histology. Endoscopists used the NICE classification system for the optical diagnosis of neoplasia and adenoma, while PolyDeep classified lesions based on the probability of neoplasia (i.e., adenomas and serrated lesions) using a 60% probability threshold. We calculated diagnostic accuracy using 2×2 confusion matrices, which included sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). We also calculated the area under the receiver operating characteristic curve (AUC) to assess overall performance in the optical diagnosis of colorectal lesions. We compared sensitivity and specificity using the McNemar test and assessed the overall diagnostic performance for the optical diagnosis of neoplasia and adenoma using the chi-square homogeneity areas test.

Results

Among the 1240 participants eligible for analysis, we detected a total of 2915 lesions. We excluded 1304 lesions identified during non-AI-assisted colonoscopies and 60 lesions due to the lack of confirmed histology. Consequently, we included 1551 lesions in the final analysis, comprising 1,003 adenomas (64.7%), 424 serrated lesions (27.3%), and 124 lesions with other histology (8.0%). Endoscopists classified as NICE II 1,070 (69.0%) lesions and as NICE I 481 (31.0%) lesions. PolyDeep identified as neoplastic lesions 938 (60.4%), as non-neoplastic lesions 575 (37.1%), and as undefined 38 (2.5%). In the characterization of neoplastic lesions, the sensitivity of the endoscopists and PolyDeep was 81.4% and 68.7% (p < 0.0001), while the specificity was 64.2% and 56.4% (p < 0.05) respectively. The endoscopists achieved a PPV of 85.9% compared to 81.2% of PolyDeep, while the NPV was 56.3% and 39.7%. Regarding overall diagnostic performance, endoscopists reached an AUC of 0.72 (95% CI 0.70–0.75) versus 0.62 (95% CI 0.59–0.65) for PolyDeep (p < 0.0001). In the characterization of adenomatous lesions, endoscopists reached a sensitivity of 84.6% compared to 70.2% for PolyDeep (p < 0.0001), and a specificity of 59.7% in comparison to 53.2% (p < 0.05) respectively. The PPV was 79.3% for endoscopists and 73.6% for PolyDeep, while the NPV was 68.0% and 49.0%. For overall diagnostic accuracy in adenoma characterization, the AUC was 0.72 (95% CI 0.69–0.74) for endoscopists and 0.61 (95% CI 0.59–0.64) for PolyDeep (p < 0.0001).

Conclusion

The diagnostic performance of endoscopists was superior to that of PolyDeep in the optical characterization of colonic lesions.

Introduction

Gastrointestinal endoscopy advancements have improved diagnostics and treatment but also increased the risk of adverse events, potentially affecting the psychological well-being of practitioners. This study aims to assess the psychological state experienced by Portuguese gastroenterologists following endoscopic adverse events in endoscopic procedures.

Aims & Methods

A nationwide online survey was conducted between November 2024 and January 2025, with invitations sent to all registered Portuguese gastroenterologists (n=709). Psychological state was assessed using validated scales assessing psychological distress (DASS-21), professional quality of life (ProQOL-5), self-compassion (SCS-12), and shame feelings (OAS2). Results are presented as median (Mdn) and interquartile range (IQR).

Results

Seventy gastroenterologists participated in the study (Mdn age 41 years; experience 11 years; both IQR 11), with a 9.87% participation rate. Nearly all (n=69) had encountered adverse events, with most reporting 1–5 incidents. None sought psychological support after the adverse events, 6 faced legal repercussions, and 3 discontinued specific techniques. DASS-21 had a Mdn of 6.00 (IQR 12.00) for stress (mild if <10), 4.00 (IQR 8.00) for anxiety (mild if <6), and 4.00 (IQR 6.00) for depression (mild if <10). ProQOL-5’s total score Mdn was 41.00 (IQR 6.50) for compassion satisfaction, 25.5 (IQR 7.00) for burnout, and 24.00 (IQR 6.25) for traumatic stress, all within the moderate range (23–41). OAS2 shame score was 16.00 (IQR 7.30). Self-compassion showed a Mdn of 3.46 (IQR 1.00; moderate if 2.3–3.5). Gastroenterologists with legal repercussions had higher compassion satisfaction (52.50 vs. 33.33; p = 0.025). Those who discontinued procedures reported higher anxiety levels (56.17 vs. 31.88; p = 0.026), but no other significant differences were found.

Conclusion

This is the first published study assessing the psychological state of adverse events on gastroenterologists. While most reported mild to moderate distress, legal repercussions were associated with higher compassion satisfaction, and stopping procedures correlated with higher anxiety. Findings suggest a need for psychological support after endoscopic adverse events, and for research on their impact on gastroenterologists.

Introduction

EUS-guided hepaticogastrostomy (EUS-HGS) is a novel technique that allows internal drainage of the biliary system via the stomach. This approach may be of value in patients with duodenal obstruction, surgically altered anatomy, or in hilar strictures with inadequate drainage of the left hepatic duct by other techniques (1, 2). However, this procedure has a high technical complexity and until recently no dedicated stents were available exposing the patient to adverse events such as the risk of biliary leakage and stent migration (3). A partially covered self-expanding metallic stent (pcSEMS) with anti-migration flaps may overcome these risks. So far, the available data on the use of this specific stent is limited, and prospective studies are particularly missing.

Aims & Methods

Aim of this study was to assess the safety and feasibility of EUS-HGS in patients with an inoperable malignant biliary obstruction, with the use of this dedicated stent. This prospective single-center study included patients with inoperable malignant biliary obstruction that underwent an EUS-HGS. Patients were ineligible in case of uncontrolled coagulopathy, tumor infiltration in the gastric wall, intervening blood vessels or extensive amount of ascites. Procedure was performed with a 10cmx10mm pcSEMS (Giobor stent, Taewoong Medical). Primary outcome was safety, defined as rate of procedural related adverse events (AEs) and mortality <30 days. Technical and clinical success rate were evaluated as secondary endpoints, defined as successful deployment of the stent in the intended location and at least 50% decrease of bilirubin <14 days.

Results

Between January 2022 and July 2024, 25 patients (14 female [56%], median age 70 years [IQR 64-76]) were included. Reason to perform EUS-HGS included duodenal obstruction (n=12), altered anatomy (n=6) or failed drainage of the left hepatic duct by ERCP (n=7). In 21/25 patients EUS-HGS was technically successful (84%). In one patient the procedure was aborted due to a self-limiting bleeding after puncture of the bile duct, in one patient due to unsuccessful cannulation of the central bile duct and in two patients the cholangiogram showed previously unidentified intrahepatic strictures that precluded adequate drainage by EUS-HGS. No severe periprocedural AEs occurred. Three patients (14%) were re-admitted with presumed cholangitis <30 days after the procedure; two with undrained right-sided bile ducts for which a percutaneous drain was placed, and one patient due to obstruction of the bile duct branches by the covered part of the stent, for which an endoscopic re-intervention was performed replacing the Giobor stent by a fully covered and uncovered SEMS. Five patients died within 30 days due to fulminant disease progression. None of these patients experienced procedural related AEs. Clinical success was achieved in 18/19 (95%) patients with technical success and follow-up of at least two weeks. Six out of the 18 patients with clinical success developed recurrent biliary obstruction after a mean of 105 days (SD 80). Obstruction was caused by hyperplasia at the uncovered portion of the stent (n=3) and due to sludge obstructing the stent (n=3). Successful re-intervention was performed in all patients.

Conclusion

This prospective study shows that EUS-HGS with a dedicated partially covered stent is feasible and safe. Tissue hyperplasia in the uncovered part of the stent and sludge obstruction may compromise long-term stent patency. Larger, comparative studies addressing stent design and timing of EUS-HGS are eagerly awaited.

References

1. Pawa, S., et al., American Society for Gastrointestinal Endoscopy guideline on the role of therapeutic EUS in the management of biliary tract disorders: summary and recommendations. Gastrointest Endosc, 2024.
2. van der Merwe, S.W., et al., Therapeutic endoscopic ultrasound: European Society of Gastrointestinal Endoscopy (ESGE) Guideline. Endoscopy, 2022. 54(2): p. 185-205.
3. Binda, C., et al., Efficacy and safety of endoscopic ultrasound-guided hepaticogastrostomy: a meta-regression analysis. Endoscopy, 2024. 56(9): p. 694-705

Clinical Case Summary

We report a case of a 71-year-old male, former heavy smoker with obesity and obstructive sleep apnea, admitted with headache, fatigue, anorexia, and persistent cough. Initial blood tests showed hemoglobin 12.7 g/dL and elevated C-reactive protein (145.4 mg/L). Chest CT revealed a spiculated lingular nodule (37×28 mm) with pulmonary and mediastinal lymphadenopathy. Bronchoscopy confirmed non-small cell lung carcinoma (adenosquamous subtype). Staging PET-CT demonstrated widespread metastases (lungs, liver, adrenals, bones, peritoneum, muscles), and brain MRI identified cerebral involvement.
During hospitalization, the patient developed melena with hemoglobin dropping to 7.1 g/dL. Upper GI endoscopy was negative. Ileocolonoscopy revealed multiple clots and a friable, infiltrative lesion in the terminal ileum (25 cm from ileocecal valve) with active low-grade bleeding. Hemostasis was successfully achieved using argon plasma coagulation (APC).
The final diagnosis was mid-gastrointestinal bleeding due to small bowel metastasis from lung cancer. Given the extent of disease, palliative care and oncology follow-up were arranged. No further bleeding episodes occurred post-intervention.
This case highlights the diagnostic challenge of small bowel metastases as a rare cause of GI bleeding. It emphasizes the role of endoscopic therapy as an effective, minimally invasive palliative strategy to control bleeding and improve patient quality of life, avoiding surgical intervention in advanced oncologic settings.
Points toiscuss:
• Recognition of small bowel metastasis in patients with known malignancy presenting with GI bleeding.
• The utility of ileocolonoscopy in diagnosing mid-GI bleeding sources.
• Endoscopic management (APC) as a palliative tool in metastatic disease.
Illustrative Material:
• Ileocolonoscopy image of terminal ileum metastatic lesion.
• PET-CT scan showing metastatic spread.

Introduction

Eosinophilic esophagitis (EoE) is a chronic inflammatory disease characterized by esophageal dysfunction and eosinophilic inflammation of the esophageal mucosa. In this study, we investigated the bacterial flora in saliva and esophageal mucus in patients with EoE and examined the relationship between EoE disease activity and mucosal cytokine expression.

Aims & Methods

Cases consisted of patients with active and inactive EoE (A-EoE and I-EoE, respectively). A-EoE was defined as a peak eosinophil count > 15/high-power field, according to the 2025 consensus guidelines. Saliva samples were collected from the patients before the endoscopic examination. Brushing samples were collected from the distal esophagus of patients with EoE during endoscopic procedures. The degree of EoE inflammation was assessed using the EoE endoscopic reference score (EREFS). The samples were profiled using the Illumina MiSeq platform. The V3–V4 regions of the 16S rRNA gene (460 bp) were amplified by using tailed PCR.

Results

Fifty-nine patients were enrolled, including 8 with I-EoE, 17 with A-EoE, and 28 non-EoE controls. Major bacterial genera, such as Streptococcus, Prevotella, Veillonella, and Alloprevotera, were detected in both the oral cavity and esophagus. Compared with the control group, the active EoE group had significantly more Prevotella spp. in the saliva and esophageal mucosa. Conversely, significantly fewer Neisseria spp. were found in the saliva and Streptococcus spp. in the esophageal mucosa of patients with active EoE. The EREFS of EoE and Streptococcus were inversely correlated.

Conclusion

In this study, we clarified the characteristics of the bacterial flora in the saliva and esophageal mucosa of patients with eosinophilic esophagitis.

Introduction

Spontaneous bacterial peritonitis (SBP) is a well-recognized decompensation that significantly affects the clinical outcome of patients with liver cirrhosis. The aim of this study is to assess the factors associated with short-term mortality in this patient population.

Aims & Methods

We conducted a retrospective study (January 2022 to December 2024) on all cirrhotic patients with SBP, admitted in the Gastroenterology Department of “St. Spiridon” County Clinical Emergency Hospital Iasi, a tertiary referral center. The SBP cases were identified based on elevated ascitic fluid absolute PMN count > 250 cells/mm³, with or without positive ascitic fluid bacterial culture.

Results

Out of the 114 cirrhotic patients diagnosed with SBP for the first time, 39 (34,2%) were female and 75 (65,8%) were male, with a mean age of 57 years. Most patients had advanced liver disease according to the Child-Pugh classification: 98 patients (86%) were classified as Child C, and 16 patients (14%) as Child B. The mean MELD-Na score was 23.3 points. The overall in-hospital mortality rate was 17.5% (20 patients). According to the multivariable analysis, several factors were significantly associated with in-hospital mortality, such as: renal dysfunction (OR 8.4118 [95% CI 2.9190 to 24.2403]; P=0.0001), serum bilirubin higher than 3 mg/dl (OR 4.8571 [95% CI 1.7445 to 13.5216]; P=0.0025), serum sodium lower than 130 mmol/l (OR 4.3750 [95% CI 1.5973 to 11.9833]; P=0.0041), international normalized ration (INR) higher than 2 (OR 4.5222 [95% CI 1.6471 to 12.4161]; P=0.0034).

Conclusion

Renal dysfunction, elevated INR, hyperbilirubinemia, and hyponatremia appear to be significant factors associated with the worsening of prognosis in cirrhotic patients with SBP, influencing in-hospital mortality.

Introduction

Recent studies have revealed that liquid-liquid phase separation (LLPS) plays a critical role in tumorigenesis. Myotilin (MYOT), a key component of actin cross-linking protein complexes, is involved in regulating myofibril assembly and maintaining the structural integrity of Z-discs in muscle cells. Intriguingly, MYOT contains intrinsically disordered regions (IDRs) and has been implicated in the migration of gastric cancer. However, whether MYOT undergoes LLPS and whether this process contributes to gastric cancer progression remains unexplored.

Aims & Methods

We aimed to investigate whether MYOT undergoes LLPS and its potential role in gastric cancer progression.
Differential gene expression analysis was performed on gastric cancer samples from our sequencing data and validated using public databases. The Disprot, Database of Protein Disorder was used to predict the IDRs of MYOT. Fluorescence recovery after photobleaching (FRAP) and cellular immunofluorescence were explored to assess MYOT’s ability to undergo LLPS. Cell proliferation, apoptosis and migration assay were conducted to investigate the potential role of MYOT in gastric cancer. Actin polymerization assays were performed to evaluate the qualitative effects of purified MYOT on actin polymerization.

Results

Differential gene expression analysis revealed that MYOT was significantly downregulated in gastric cancer tissues compared to adjacent normal tissues, a finding further confirmed by RT-qPCR. In the prediction of protein structure, it can be found that MYOT contains extensive IDRs. FRAP analysis demonstrated rapid fluorescence recovery in GFP-MYOT droplets within minutes, suggesting dynamic LLPS behavior. Similarly, purified GFP-MYOT exhibited fusion behaviors, further supporting its capacity for LLPS in vitro. Overexpression of MYOT did not significantly affect gastric cancer cell proliferation or apoptosis, but reduced cell migration. Actin polymerization failed to occur both in the GFP-MYOT and GFP control samples.

Conclusion

Our findings demonstrate that MYOT suppresses gastric cancer cell migration and purified GFP-MYOT does not directly promote on actin polymerization, implicating its potential role as a tumor suppressor. Additionally, MYOT undergoes LLPS both in vitro and in vivo, as evidenced by FRAP. We propose that MYOT may exert its anti-tumor effects through LLPS-mediated mechanisms, offering a novel perspective on its function in gastric cancer progression.