Clinical Case Summary

Immune checkpoint inhibitor (ICI) related gastrointestinal (GI) toxicities are associated with increased incidence due to their rising use in cancer management.
We report a unique case of a late-onset immune-related enteritis in a 66-year-old man with metastatic melanoma, initially receiving dual ICI therapy (Ipilimumab – anti-CTLA4 and Nivolumab – anti-PD-1 agent). Аpproximately 18 months after being placed on Nivolumab maintenance monotherapy he presented with abdominal pain, peripheral edema and watery diarrhea up to 10 times a day. Infectious causes of diarrhea were excluded, including Cl.difficile, along with negative serology for CMV, EBV and HSV. Endoscopic evaluation revealed multiple deep ulcerations along the duodenum, proximal jejunum and terminal ileum with no involvement of the stomach and colon. Histological analysis was indicative of acute chronic small bowel inflammation with positive CMV-PCR and no signs of neoplasms or specific granulomas. Patient was placed on parenteral nutrition and peroral Valacyclovir with no significant improvement. Systemic steroid therapy was then initiated followed by a reduction in bowel movements, confirming the immune-mediated course of the disease.
After one-month follow up patient reported of persisting grade 1 diarrhea with elevated calprotectin levels while slowly tapering steroid dose. Control gastroscopy revealed remaining multiple ulcers at the descending duodenal part and jejunum. Second line therapy of Infliximab was then initiated, achieving complete response after three infusions. Corticosteroids were eventually slowly tapered off and Nivolumab was permanently ceased.
Ruling out all other possible causes, the diagnosis of late-onset immune-mediated enteritis was made. Although anti-PD-1 related GI toxicities seem to be less frequent with typical onset within three months after initiation, long-term surveillance and therefore prompt recognition is encouraged to avoid life-threatening complications.

References

1. CT showing diffuse enteritis
2. Endoscopic findings in the descending duodenal part and terminal ileum
3. Histopathological image
4. Follow up gastroscopy findings

Introduction

Endoscopic depth diagnosis is essential for the treatment decision of Barrett's esophageal adenocarcinoma (BEA). However, it is sometimes difficult to diagnose due to the anatomy of the esophagogastric junction, such as the difficulty of luminal distension. For inconclusive cT1a or cT1b lesions, we introduced staging ESD, allowing for histologic diagnosis for depth invasion. However, the clinical feasibility and significance of staging ESD remain unclear.

Aims & Methods

In this study, we evaluated the performance of endoscopic depth diagnosis of cT1 BEA and the appropriate treatment strategy.
This study included cT1N0M0 BEA treated with ESD or surgery from January 2015 to June 2024. The cases with poorly differentiated component on target biopsy were excluded. Endoscopic depth diagnosis was based on macroscopic appearance (such as the following findings suggestive of cT1b: polypoid, mixed macroscopic type, thickness, hardness, severe redness, and ulceration). EUS was added at the discretion of the endoscopist. We made the treatment decision based on our multidisciplinary conference, and the final diagnosis was classified as conclusive cT1a or cT1b, and inconclusive cT1a or cT1b. We performed ESD for conclusive cT1a, and surgery for conclusive cT1b. And we performed staging ESD for inconclusive cT1a or cT1b. Curative resection was defined as pT1a with R0 resection, and without lymphovascular invasion or poorly differentiated component.

Results

This study included 54 lesions in 52 patients. The median age was 66 years (43-92). There were 42 male patients. Siewert type [type I/type II] were 12 and 42, respectively. Background Barrett’s esophagus [SSBE/LSBE] were 36 and 16, respectively. The median size was 20mm (5-50). Major macroscopic appearances [0-I/0-IIa/0-IIb/0-IIc] were 11, 19, 1, and 23, respectively. Histology on targeted biopsy [tub1/tub2] were 41, 13, respectively. 39 lesions were diagnosed by white light endoscopy alone, and 15 lesions were comprehensively diagnosed by additional EUS. 41 lesions were diagnosed as cT1a and 13 lesions were diagnosed as cT1b. ESD and surgery were performed for 47 and 7 lesions, respectively. Pathologic depths [T1a/T1b/T2] were 42, 11, and 1, respectively. The sensitivity, specificity, and positive predictive value of endoscopic diagnosis of T1b were 25.0%, 76.2%, and 23.1%, respectively. 10 of 54 (18.5%) T1a lesions were overdiagnosed. Staging ESD was performed for 20 lesions, and all of them achieved en bloc resection without adverse events. R0 resection was achieved in 13 of 14 (92.9%) cT1a and 5 of 6 (83.3%) cT1b, respectively. Curative resection was achieved in 8 of 14 (57.1%) cT1a and 4 of 6 (60.0%) cT1b, respectively. 3 of 54 (5.6%) T1a were diagnosed as conclusive cT1b, and underwent surgery, leading to overtreatment.

Conclusion

The diagnostic performance of endoscopic depth diagnosis is insufficient, leading to overdiagnosis. Staging ESD for inconclusive cT1 BEA is technically feasible and can be an option to avoid overtreatment.

Introduction

Increased intestinal permeability, mediated by zonulin, has been implicated in the pathogenesis of insulin resistance, dyslipidemia, and non-alcoholic fatty liver disease (NAFLD). Small intestinal bacterial overgrowth (SIBO) may further contribute to metabolic dysfunction through endotoxin-induced inflammation.

Aims & Methods

The aim was to determine zonulin concentrations in patients with dyslipidemia and NAFLD and analyze their relationships with metabolic parameters and the prevalence of small intestinal bacterial overgrowth (SIBO). A total of 342 patients with dyslipidemia were enrolled; NAFLD was confirmed in 152 of them based on ultrasound and steatometry. Fecal zonulin levels were measured using ELISA; SIBO was assessed by hydrogen breath test. Lipid profiles, high-sensitivity C-reactive protein (hsCRP), and HDL-C levels were also analyzed.

Results

Zonulin levels in patients with NAFLD (115±2.76 ng/g) exceeded the normal range (≤107 ng/g), compared to 87±0.28 ng/g in the dyslipidemia group without NAFLD (p<0.05). SIBO was present in 45% of dyslipidemic patients and 53.2% of those with NAFLD. A positive correlation was found between zonulin and hsCRP (r=0.65) and triglycerides (r=0.58), and a negative correlation between SIBO and HDL-C (r=–0.78) in the dyslipidemia group.

Conclusion

Elevated zonulin and increased intestinal permeability are associated with higher cardiometabolic risk in patients with NAFLD and dyslipidemia. Both zonulin and SIBO may serve as potential therapeutic targets for preventing progression of NAFLD and related metabolic disturbances.

Introduction

Gilbert's Syndrome (GS) is a benign polymorphism of the UGT1A1 gene, leading to decreased enzyme activity in the liver and mildly increased levels of unconjugated bilirubin in blood (= hyperbilirubinemia), that is affecting 5-12% of the worldwide population. A mild hyperbilirubinemia (HBIL) is associated with reduced chronic inflammation, a lower risk for metabolic and cardiovascular disorders, and lower levels of reactive oxygen species (ROS). Despite these associations, HBIL effects on the intestine, as the primary site of bilirubin excretion, and its interactions with the gut microbiota remain unclear. We propose that chronic HBIL redirects bilirubin excretion through the intestinal epithelium that affects microbe-host interactions in the colon.

Aims & Methods

Germ-free (GF) and specific-pathogen-free (SPF) C57BL/6N wildtype (WT) mice with induced HBIL and transgenic mice with the human UGT1A1*28 polymorphism (TG) were investigated, the latter one mimicking chronic HBIL. WT mice with induced HBIL received intraperitoneal bilirubin injections for 12-14 days. Bilirubin levels, in the intestine, gut content and serum were analyzed using the fluorescent protein UnaG, which specifically binds to unconjugated bilirubin, and microbiota composition was analyzed by 16S rRNA gene sequencing. To characterize HBIL-associated antioxidative function, colonic adenoma cell lines (Ptk-6, HCT116, HT-29-MTX-E12) and small intestinal organoids were incubated with different levels of bilirubin and the generation of ROS and oxidative stress markers measured.

Results

In general, mice of the induced HBIL groups exhibited higher bilirubin levels in the colonic mucosa, indicating "intestinal HBIL" with increased concentrations of mucosal bilirubin being detected in SPF HBIL > GF HBIL mice (SPF HBIL = 12.34 nM/mg, GF HBIL = 9.9 nM/mg; SPF control = 0.5951 nM/mg; GF control = 8.21 nM/mg), as well as a lower hepatic Ugt1a1 expression. Intracellular uptake of bilirubin was confirmed when different human and murine colonic epithelial cell lines were treated with bilirubin in vitro. Intestinal HBIL led to altered gut microbiota compositions, characterized by decreased abundances of operational taxonomic units (OTUs) associated with Gram-negative taxa (e.g., Prevotellaceae) and increased levels of OTUs associated with Gram-positive taxa (e.g., Lachnospiraceae) under intestinal HBIL conditions. Gut microbiota changes were more pronounced in TG (=chronic HBIL) compared to induced HBIL mice. In vitro experiments showed decreased detection of ROS and markers of cellular oxidative stress in a HCT116 reporter cell line dependent on bilirubin presence (Nfk-⁠β, Nrf2 and Hif1α). Small intestinal organoids from TG mice showed an altered proliferation pattern in comparison to WT organoids.

Conclusion

The results suggest an altered bilirubin excretion through the intestinal epithelium under mild HBIL that alters gut microbiota composition. These effects were more pronounced under chronic HBIL conditions, suggesting a sustained effect of GS-like HBIL on the gut microbiota.

Introduction

Endoscopic submucosal dissection (ESD) for early upper gastrointestinal (GI) tumors creates large mucosal defects with risks of bleeding, perforation, and strictures, particularly in the esophagus where post-ESD stricture rates reach clinically significant levels. Current closure methods are inadequate for managing esophageal defects due to anatomical complexity. We present an endoscopy-integrated platform for real-time in situ nanofiber deposition during upper GI ESD, leveraging blow spinning technology to address these limitations.

Aims & Methods

This study aimed to
(1) engineer a coaxial nozzle system compatible with standard upper GI endoscopes (working channel ≤3.7 mm),
(2) optimize nanofiber deposition parameters for esophageal wound repair, and
(3) validate therapeutic efficacy in preclinical models.
A simplified in vitro setup was used to optimize polymer composition (polycaprolactone/gelatin blends), gas pressure (0.2-0.5 MPa), and blowing distance (10-20 cm). A murine skin wound model was employed to validate nanofiber adhesion and repair properties. For endoscopic integration, a dual-layer coaxial nozzle was designed using computational fluid dynamics (CFD), featuring an inner polymer delivery channel (0.5 mm inner diameter) and an outer airflow channel (1.2 mm inner diameter) to generate nanofibers via controlled airflow. The nozzle was fabricated using stereolithography 3D printing with biocompatible resins and integrated into a compact endoscopic system. Functional testing was performed in a transparent esophageal phantom under simulated physiological flushing, followed by in vivo validation in a porcine esophageal ESD model comparing nanofiber-treated wounds to controls.

Results

The 3D-printed coaxial nozzle (3.0-mm outer diameter) demonstrated full compatibility with standard upper GI endoscopes. In simulated esophageal conditions, the platform achieved uniform nanofiber deposition (fiber diameter: 15.2±3.6 μm) with robust adhesion unaffected by luminal flushing. In the porcine esophageal ESD model, immediate nanofiber application provided complete wound coverage, reduced perioperative bleeding, and prevented stricture formation. Histological analysis at 4 weeks revealed markedly reduced fibrosis and enhanced mucosal regeneration in treated groups compared to controls. The system operated seamlessly within the standard ESD workflow without procedural interference.

Conclusion

This study develops the first endoscopy-compatible platform for real-time in situ nanofiber deposition, demonstrating preclinical efficacy in addressing post-ESD complications in the upper gastrointestinal tract. By synergizing mechanical barrier formation with bioactive healing modulation, the technology reduces stricture risks and promotes mucosal repair in porcine esophageal ESD models. While current findings underscore its translational promise, further optimization of biomaterial formulations and rigorous human trials are essential steps toward clinical adoption in therapeutic upper GI endoscopy.

Introduction

Anorectal dysfunction affects approximately 8% of the global population.¹ Whilst conservative management can manage mild symptoms, surgery is often indicated in patients with structural abnormalities and severe symptoms. A Magnetic Resonance Imaging Defecating Proctogram (MRDP) forms a key part of patients’ pre-surgical investigations, visualising the anterior, middle and posterior compartments of the pelvis, and provides a dynamic image of the pelvic floor during defecation. It has been criticised as it is often performed in the supine position, which does not reflect normal physiological positioning for defecation.² However, little is known about how patients themselves feel about how this test reflects their usual defecatory process.

Aims & Methods

As part of a service evaluation, between September 2024 and April 2025, all patients who underwent an MRDP at Imperial College Healthcare NHS Trust in London, United Kingdom, were asked to complete a 21-item questionnaire exploring their experiences, prior to, during and after their MRDP. Five-point Likert Scales (1 - Strongly Disagree, 5 - Strongly Agree) were used for each question, except two concluding questions on overall experience, which were answered using free-text boxes. Patients with insufficient cognitive function or comprehension of English to complete the questionnaires were excluded.

Results

Thirty-eight patients responded to the questionnaire, and after excluding those with missing data, 35 completed questionnaires were included in the analyses (30[86%] female, 28[80%] 50-79 age category). Reasons for referral to MRDP included FI (15/35 [43%]) and/or constipation (13/35 [37%]) and/or POP (9/35 [26%]). Although 32(91%) strongly agreed that their privacy was respected, and none strongly agreed they felt embarrassed, just 12(34%) strongly agreed that they felt ‘comfortable during the test’. Less than a third of patients (10[29%]) agreed that they were able to ‘defecate as they do at home’ and only a quarter (9[26%]) felt that defecating lying supine was easy. Whilst 16(46%) patients felt that the contrast mixture inserted gave them a ‘normal urge to open their bowels’, only 4(12%) reported that the contrast mixture was ‘similar to their own stool’. Nevertheless, 23(66%) patients agreed that the experience matched what they would expect, and most patients (29[83%]) would have the test again if it were necessary. Patients aged 70-79 were associated with disagreeing with the statement that they could ‘defecate as they would at home’ (neutral/disagree/strongly disagree vs. agree/strong agree: 8/8[100%] vs 0.0[0%], p = 0.011).

Conclusion

Whilst patients reported being able to tolerate the MRDP, the large majority reported that defecating during the investigation did not reflect their usual defecatory process. Contrary to expectation, it appears that patients’ main complaints were not due to embarrassment, but rather patient position and environment. Whilst patient experiences are highly institution and staff dependent, understanding patients’ perspectives during an MRDP is useful for optimisation of pre-investigation counselling and expectation setting. Implementing post-MRDP questionnaires, or pro-active clinician exploration of the extent to which patients felt able to replicate their symptoms could facilitate clinical interpretation of MRDP results. Moving towards evacuation testing that aligns more closely with a patient's normal defecatory habits is likely to yield results that better reflect symptom profiles and therefore provide better insights into specific surgical targets responsible for symptom generation.

References

1. Jaafari H, Houghton LA, West RM, et al. The national prevalence of disorders of gut brain interaction in the United Kingdom in comparison to their worldwide prevalence: Results from the Rome foundation global epidemiology study. Neurogastroenterology & Motility. 2023; 35:e14574. doi:10.1111/nmo.14574
2. Pilkington, S.A., Nugent, K.P., Brenner, J., Harris, S., Clarke, A., Lamparelli, M., Thomas, C. and Tarver, D. (2012), Barium proctography vs magnetic resonance proctography for pelvic floor disorders: a comparative study. Colorectal Disease, 14: 1224-1230. https://doi.org/10.1111/j.1463-1318.2012.02945.x

Introduction

A grey zone known as the borderline resectable (BR) category, which lies between resectable and unresectable, has been proposed for perihilar cholangiocarcinoma; however, there is no international consensus on this classification. BR can be considered from various perspectives, including local tumor factors, surgical risk, long-term prognosis, liver reserve capacity, and systemic condition.

Aims & Methods

The study's aim is to define the borderline resectable (BR) of perihilar cholangiocarcinoma based on surgical procedure and preoperative liver function. We examined 1) the comparison of surgical outcomes between extended surgery and standard surgery and 2) the risk of postoperative mortality from liver failure in patients who underwent portal vein embolization.
Study 1) Surgical cases from January 2011 to December 2023 were divided into two groups: the Extended surgery group: left or right trisectionectomy(L3/R3), concomitant hepatic artery resection (AR), hepatopancreaticoduodenectomy (HPD), and the other standard surgery groups. Study 2) The risk of postoperative liver failure in patients who underwent resection after portal vein embolization (PVE) between January 2001 and December 2022 (including HPD) was evaluated

Results

Study 1) A total of 181 patients (59 in the extended surgery group and 122 in the standard surgery group) were included in the study. There were significant differences in operation time (median 798/700 minutes, p<0.01) and blood loss (2000 ml/1295 ml, p<0.01) in the extended surgery group compared to the standard surgery group, but no significant differences in postoperative complication rate (Clavien Dindo 3a or higher: 79%/67%) or in-hospital mortality rate (5%/5%). The 5-year overall survival rate was 34% in the extended surgery group and 50% in the standard surgery group, which was not significantly different by the log-rank test. Study 2: The incidence of liver failure (%) and mortality (%) by surgical technique were R (37/5.7), L3 (30/5), and R3 (54/7.7). ICGK x proportion of remnant liver volume（ICGKF）<0.05 before PVE, operation time >660 minutes, and blood loss >1900 ml were identified as risk factors in univariate and multivariate analyses, and the mortality rate for patients with these three factors was high at 25%.

Conclusion

Patients who required extended surgery with poor pre-PVE liver function (ICGKF<0.05) should be considered BR in perihilar cholangiocarcinoma.

References

1: Nooijen LE, de Boer MT, Braat AE, Dewulf M, den Dulk M, Hagendoorn J, Hoogwater FJH, Lam HD, Molenaar Q, Neumann U, Porte RJ, Swijnenburg RJ, Zonderhuis B, Kazemier G, Klümpen HJ, van Gulik T, Groot Koerkamp B, Erdmann JI.
National consensus on a new resectability classification for perihilar cholangiocarcinoma - A modified Delphi method. Eur J Surg Oncol. 2025 Feb;51(2):107117. PMID: 37880001.
2: Kuriyama N, Usui M, Gyoten K, Hayasaki A, Fujii T, Iizawa Y, Kato H, Murata Y, Tanemura A, Kishiwada M, Sakurai H, Mizuno S, Isaji S.
Neoadjuvant chemotherapy followed by curative-intent surgery for perihilar cholangiocarcinoma based on its anatomical resectability classification and lymph node status. BMC Cancer. 2020 May 11;20(1):405. doi: PMID: 32393197;.
3: Matsuyama R, Mori R, Ota Y, Homma Y, Yabusita Y, Hiratani S, Murakami T, Sawada Y, Miyake K, Shimizu Y, Kumamoto T, Endo I.
Impact of Gemcitabine Plus S1 Neoadjuvant Chemotherapy on Borderline Resectable Perihilar Cholangiocarcinoma.
Ann Surg Oncol. 2022 Apr;29(4):2393-2405. doi: 10.1245/s10434-021-11206-4. Epub
2022 Jan 7. PMID: 34994885.