Background: Statins, widely prescribed for hypercholesterolemia, have demonstrated potential antineoplastic properties in preclinical studies. Despite growing interest in their oncologic effects, the role of statin therapy within curative treatment of esophageal cancer remains unexplored. This study aimed to evaluate the impact of statin use on pathologic complete response (pCR) rate, disease-free survival (DFS), and overall survival (OS) in patients undergoing neoadjuvant chemo(radio)therapy followed by esophagectomy. Methods: All consecutive patients with esophageal or gastroesophageal junction cancer who underwent esophagectomy following neoadjuvant therapy between March 1994 and September 2013 were retrospectively analyzed using a prospectively maintained database. Baseline demographic and clinical variables were compared between statin users and non-users. Results: A total of 463 patients were included, of whom 90 (19.4%) were statin users at diagnosis. Neoadjuvant chemotherapy (CT) was administered in 88 patients (19%) and chemoradiotherapy (CRT) in 375 patients (81%). pCR (ypT0N0M0) was achieved in 85 patients (18%), with no statistically significant difference between statin users and non-users (22.2% vs. 17.4%, P=0.29). Median DFS (45 vs. 40 months, P=0.25) and OS (44 vs. 42 months, P=0.28) were also not significantly different between the two groups. However, a non-significant trend toward improved DFS was identified in patients with esophageal adenocarcinoma receiving lipophilic statin therapy. Conclusions: In this cohort, statin use was not associated with improved pathologic response or survival outcomes following neoadjuvant therapy for esophageal cancer. These findings do not support modification or discontinuation of statin therapy in this patient population.

Background:Although endoscopic resection (ER) is recommended as first-choice treatment for early esophageal neoplasia, patients with esophageal varices are considered a high-risk group owing to an increased risk of bleeding. This systematic review aimed to evaluate the effectiveness and safety of ER in this specific patient category. Methods:We searched for studies reporting on clinical outcomes of ER in the presence of esophageal varices, irrespective of study design or follow-up time. End points included the incidence of prophylactic measures to reduce the risk of variceal hemorrhage, radical and curative resection rates, and adverse events. Results:After screening 2371 studies, 42 studies (including our own unpublished cohort) with a total of 186 patients were included in this systematic review. Endoscopic band ligation (72/186; 39%) and endoscopic injection sclerotherapy (22/186; 12%) were the prophylactic measures most widely adopted to eradicate varices prior to ER. Other frequently described prophylactic measures included direct varix coagulation during ER (18/186; 10%) and the placement of a transjugular intrahepatic portosystemic shunt prior to ER (9/186; 5%). While the radical and curative resection rates were high (86% and 72%, respectively), the periprocedural and delayed bleeding risks were reported to be relatively low (6% and 3%, respectively). In all studies, no procedure-related mortality was observed. Conclusions:ER appeared to be a safe and effective treatment option in selected patients with concurrent early esophageal neoplasia and esophageal varices, provided that a tailored approach of adequate prophylactic measures to prevent bleeding is applied.

Background and Aims Re-cellularization via electroporation therapy (ReCET) is an endoscopic procedure that uses electroporation to induce cellular apoptosis and subsequent re-epithelization in the duodenum. ReCET has demonstrated a positive effect on glycaemia. In this study, a single ReCET procedure was combined with a glucagon-like peptide-1 receptor agonist (GLP-1RA) with the goal of eliminating insulin treatment in patients with type 2 diabetes (T2D). At 12 months 86% of patients remained off insulin. Here we report the longer-term efficacy and safety of ReCET combined with GLP-1RA over 24 months. Methods This was a first-in-human study in patients with T2D on basal insulin (28-75 years of age; body mass index, 24-40 kg/m2; hemoglobin A1c [HbA1c], ≤64 mmol/mol; C-peptide, ≥0.2 nmol/L). All of the patients underwent ReCET, after which insulin was discontinued and semaglutide (GLP-1RA) was initiated. The follow-up was extended to 24 months for initial responders (off insulin and HbA1c ≤58 mmol/mol at 12 months). Primary end points were number of responders (efficacy) and number of adverse events, serious adverse events, and hypoglycemic events (safety). Secondary end points included glycemic, metabolic, and cardiovascular parameters, and treatment satisfaction scores. Results Fourteen patients underwent endoscopic ReCET. Eleven initial responders completed the 24-month follow-up, of which 1 withdrew consent after 18 months. All patients remained off insulin with adequate glycemic control. No procedure-related serious adverse events or severe hypoglycemic events were observed during 24 months of follow-up. Homeostatic model assessment of insulin resistance and liver fat percentage improved significantly compared with baseline. Conclusions These results suggest that duodenal ReCET combined with semaglutide is durable, allowing insulin elimination up to 24 months, and appears safe. Glycemic control was maintained, and metabolic health improved during the 24-month follow-up period.

Background and Aims Endoscopic eradication therapy (EET) for Barrett’s esophagus (BE) with high-grade dysplasia (HGD) aims to prevent progression to life-limiting cancer. However, HGD patients with a short life expectancy realize limited gains from this approach owing to competing mortality. Clinicians are poorly equipped to counsel such patients, because there are little data regarding survival in untreated HGD. We aimed to evaluate existing literature and describe a case series of patients with untreated HGD. Methods We systematically reviewed Medline, Embase, and Cochrane Library, selecting studies describing the natural history of untreated HGD in BE. The primary outcomes were symptomatic esophageal adenocarcinoma (EAC) and EAC-related death. In addition, cases in which HGD was left untreated were retrospectively identified in the Netherlands. We assessed the time until progression to clinically evident EAC. Results A total of 3229 studies were identified, of which 3 were included. In 1 study, progression from HGD to clinically evident EAC occurred in 4 subjects after a median of 34 months. The remaining 2 cases progressed to clinically evident EAC after 70 and 115 months. In our previously unreported case series, 11 Dutch patients with flat HGD (n = 3) or HGD in a visible abnormality (n = 8) were included. Four of these 11 patients progressed to clinically evident EAC after a median 52 months (range, 17-78 months). Conclusions The lag-time between the diagnosis of HGD and progression to clinically evident EAC varied from 1.5 to 10 years. EET for BE with HGD in patients with less than 3 years of life expectancy seems unlikely to be beneficial. These results may guide management decisions for patients with BE. (Netherlands Trial Registry, NL7039; NL-OMON29089.)

PRE-ADOPTION:Endoscopists performing upper gastrointestinal (UGI) endoscopy should train to recognize the appearance of early esophageal cancer and its precursor lesions, for Barrett neoplasia and for squamous neoplasia, to improve detection rates.Every endoscopist should have achieved competency in UGI endoscopy before commencing training in optical diagnosis in the UGI tract, and this requires personal experience of 300 UGI endoscopies and meeting the ESGE quality measures for UGI endoscopy. TRAINING:Endoscopists should undertake interactive training courses to improve diagnostic accuracy for esophageal preneoplastic and neoplastic lesions.Endoscopists should train with the Barrett's Oesophagus-Related Neoplasia (BORN) module to improve detection and delineation for early Barrett's esophagus-related neoplasia.Endoscopists should train in using the Prague classification to describe the circumferential and maximal extent of Barrett epithelium above the gastroesophageal junction.Endoscopists should train in using the Paris classification to describe suspected neoplastic lesions within both Barrett and squamous epithelium.Endoscopists should train in using virtual chromoendoscopy to assess both mucosal and vascular patterns in potentially neoplastic lesions in Barrett epithelium as being regular or irregular.When using acetic acid chromoendoscopy for Barrett's esophagus inspection, endoscopists should be familiar with the appearance of neoplastic lesions by assessing loss of aceto-whitening and mucosal surface patterns.Endoscopists should train in using the Japanese Esophageal Society (JES) classification to describe esophageal squamous cell lesions, to estimate the likelihood of and degree of dysplasia, and in the case of squamous cell cancer, the invasion depth.

Background and study aims:Histological analysis of regular duodenal biopsies to study morphologic changes after duodenal ablation for type 2 diabetes (T2D) and metabolic syndrome is hampered by variability in tissue orientation. We designed an optimized tissue acquisition protocol using duodenal cold snare resections to create tissue microarrays (TMAs) and to allow for single-cell RNA sequencing (scRNA-seq). Patients and methods:The open-label DIRECT study included patients undergoing an upper gastrointestinal interventional endoscopy for non-duodenal indications. All underwent one ot two single-piece duodenal cold snare resections. Endpoints were safety, adequate histological orientation of specimen and TMA, and tissue dissociation quality for scRNA-seq. The optimized tissue acquisition protocol was validated in a duodenal ablation study, EMINENT-2. Results:In DIRECT, nine patients were included in whom a total of 16 cold snare resections were obtained. No severe adverse events (SAEs) occurred. Eighty percent of specimens and corresponding TMAs showed optimal tissue orientation. Further improvement was achieved by reducing tissue damage during endoscopic retrieval and improving histologic evaluation by eliminating ink use and pinning the tissue on cork. High-quality tissue dissociation scores for scRNA-seq were achieved in 13 of 18 samples (72%). In EMINENT-2, 38 cold snares were obtained without SAEs, histopathologic analysis showed good orientation in all samples, and dissociation scores for scRNA-seq were qualified in 35/38 (92%) samples. Conclusions:Duodenal cold snare resection is safe and can provide high-quality tissue for optimally oriented TMAs and high-quality tissue dissociation scores for scRNA-seq (Clinicaltrials.gov, NCT06333093, NCT05984238). This approach will allow mechanistic studies about the effects of duodenal ablation on metabolic syndrome and T2D.

BACKGROUND & AIMS:Training deep learning systems in endoscopy generally requires vast datasets of annotated images, which are often scarce and costly to obtain. Foundation models are pretrained on large, diverse datasets and can be applied across a wide range of tasks with minimal additional fine-tuning. For endoscopy, foundation models require datasets of general endoscopic images. Yet, datasets for developing such models remain limited. In this study, we present GastroNet-5M, a dataset comprising 4,820,653 endoscopic images of ∼500,000 procedures. METHODS:GastroNet-5M consists of anonymized general endoscopic images captured in 8 Dutch hospitals between 2012 and 2020. Using self-supervised learning, GastroNet-5M was used to develop a foundation model for subsequent downstream endoscopic artificial intelligence (AI) applications. We compared our GastroNet-5M foundation model with publicly available endoscopic foundation models and state-of-the-art nonfoundation models across 17 endoscopic AI applications throughout the gastrointestinal tract. Outcome measures were classification and segmentation accuracy, data efficiency, and robustness to data heterogeneity. RESULTS:GastroNet-5M-pretrained models outperformed all other models in accuracy for nearly all classification and segmentation tasks. Furthermore, GastroNet-5M-pretrained models required significantly less application-specific training data for satisfactory model performance and displayed more robust performance when models were exposed to data heterogeneity such as imagery from different endoscope manufacturers. CONCLUSIONS:This study presents GastroNet-5M, a dataset of ∼5 million endoscopic images. Pretraining endoscopic deep learning systems with GastroNet-5M improves diagnostic accuracy, reduces the need for scarce application-specific endoscopic imagery and annotations, and increases their robustness to the inevitable data heterogeneity in clinical practice. This may significantly accelerate development and implementation of endoscopic AI systems. GastroNet-5M is publicly available for scientific use.

Background:Focal cryoballoon ablation (FCBA) is a relatively new modality for treatment of Barrett esophagus (BE)-related neoplasia. This study evaluated the efficacy and safety of FCBA for BE. Methods:Patients with BE segments (Prague classification C≤2M≤5) with dysplasia or early cancer were eligible for inclusion. Following endoscopic resection of visible lesions, FCBA was performed at 3-month intervals until complete eradication of BE (maximum five sessions). After ≥2 FCBA sessions, add-on treatment was allowed. Follow-up endoscopy was scheduled at 6 months and annually thereafter. Outcomes were complete eradication of endoscopically visible BE (CE-BE), intestinal metaplasia (CE-IM), and dysplasia (CE-D), durability of treatment response, and adverse events. Results:107 patients (mean age 65 years, 91 males, median BE C0M2) were included. Endoscopic resection was performed at entry in 65% (69/107). Patients received a median of 2 FCBA treatments. Add-on treatment was performed in 40% (43/107), mainly APC for small remaining islands (38%; 41/107). CE-BE and CE-D were achieved in 94% (101/107; 95%CI 90%–98%) and CE-IM in 91% (97/107; 95%CI 85%–95%), per intention-to-treat analysis. In per-protocol analysis, CE-BE and CE-D was achieved in 100% (101/101; 95%CI 100%–100%), and CE-IM in 96% (97/101; 95%CI 92%–99%). After a median follow-up of 18 months, 96% (97/101; 95%CI 92%–99%) remained free of endoscopically visible BE. Esophageal stricture was the most common adverse event, in 13% (13/101; 95%CI 6%–20%). Conclusion:FCBA was highly effective in selected patients with BE of limited length, although the potential risk for stricture formation warrants further research.

Background:Post-endoscopic resection (ER) management of high risk T1 esophageal adenocarcinoma (EAC) is debated, with conflicting reports on lymph node metastasis (LNM). We aimed to assess outcomes following radical ER for high risk T1 EAC. Methods:We identified patients who underwent radical ER (tumor-negative deep margin) of high risk T1 EAC, followed by surgery or endoscopic surveillance, between 2008 and 2019 across 11 international centers. Results:106 patients (86 men; mean [SD] age, 70 [11] years) were included. Of these, 26 (age, 64 [11] years) underwent additional surgery, with residual T1 EAC found in five patients (19%) and LNM in two (8%). After a median [IQR] follow-up of 47 [32–79] months, 2/26 patients (8%) developed LNM/distant metastasis, with one EAC-related death (4%), one unrelated death (4%), and four patients lost to follow-up (15%). Of the 80 patients (age, 71 [9] years) who entered endoscopic surveillance, 5/80 (6%) developed LNM/distant metastasis, with four EAC-related deaths (5%) over 46 (IQR 25–59) months follow-up; there were 15 unrelated deaths (19%), and 10 patients lost to follow-up (13%). The overall rates (95%CI) were: LNM, 6% (2%–12%); LNM/distant metastasis, 7% (3%–13%); EAC-related mortality, 5% (2%–11%); overall mortality, 20% (95%CI 13–29). Conclusion:Our findings present low rates of LNM after radical ER of high risk T1 EAC, consistent with other endoscopy-focused studies. Post-surgical patients are still at risk for metastasis and disease-specific mortality. These results suggest that endoscopic surveillance is suitable for selected cases, but further prospective studies are needed to refine patient selection and confirm optimal outcomes.

BACKGROUND : Timely detection of neoplasia in Barrett's esophagus (BE) remains challenging. While computer-aided detection (CADe) systems have been developed to assist endoscopists, their effectiveness depends heavily on the quality of the endoscopic procedure. This study introduces a novel computer-aided quality (CAQ) system for BE, evaluating its stand-alone performance and integration with a CADe system. METHOD : The CAQ system was developed using 7,463 images from 359 BE patients. It assesses objective quality parameters (e. g., blurriness, illumination) and subjective parameters (mucosal cleanliness, esophageal expansion) and can exclude low-quality images when integrated with a CADe system.To evaluate CAQ stand-alone performance, the Endoscopic Image Quality test set, consisting of 647 images from 51 BE patients across 8 hospitals, was labeled for objective and subjective quality. To assess the benefit of the CAQ system as a preprocessing filter of a CADe system, the Barrett CADe test set was developed. It consisted of 956 video frames from 62 neoplastic patients and 557 frames from 35 non-dysplastic patients, in 12 Barrett referral centers. RESULTS : As stand-alone tool, the CAQ system achieved Cohen's Kappa scores of 0.73, 0.91, and 0.89 for objective quality, mucosal cleanliness, and esophageal expansion, comparable to inter-annotator scores of 0.73, 0.93, and 0.83. As preprocessing filter, the CAQ system improved CADe sensitivity from 82 % to 90 % and AUC from 87 % to 91 %, while maintaining specificity at 75 %. CONCLUSION : This study presents the first CAQ system for automated quality control in BE. The system effectively distinguishes poorly from well-visualized mucosa and enhances neoplasia detection when integrated with CADe.

Introduction Endoscopic duodenal mucosal resurfacing (DMR) induces mucosal regeneration through hydrothermal ablation. Studies have shown that DMR improves hyperglycaemia in patients with type 2 diabetes. The INSPIRE study (n=16) aimed to eliminate insulin by combining DMR with a glucagon-like peptide-1 receptor agonist (GLP-1RA). At 18 months, 53% of patients were off insulin with adequate glycaemic control. This retrospective analysis evaluates long-term effects of DMR and GLP-1RA.Research design and methods Fourteen patients gave consent for the prolonged follow-up and data on glycaemic control, medication and DMR satisfaction were obtained via general practitioners. Primary endpoint was the number of patients off insulin with glycaemic control (haemoglobin A1c (HbA1c) ≤58 mmol/mol (7.5%)) at 4 years post DMR. Secondary endpoints were changes in HbA1c, fasting plasma glucose (FPG), weight, body mass index (BMI), use of glucose-lowering medications, treatment satisfaction and willingness to undergo DMR retreatment.Results Nine out of 14 patients (64%) were off insulin at 4 years. Of these, five patients (36% of study population) were adequately controlled. HbA1c and FPG did not change significantly, whereas BMI decreased at 3 and 4 years. Among the responders at 12 months, there was 80% durability at 4 years. Patients rated the DMR procedure 9 out of 10 and 86% were willing to undergo DMR retreatment.Conclusions Despite a stable number of patients remaining off insulin, a decrease in patients with adequate glycaemic control was observed. This suggests a slightly fading long-term DMR effect on glycaemia. High patient satisfaction and willingness for DMR retreatment indicate a promising avenue for further research.

BACKGROUND:Artificial intelligence (AI) systems in endoscopy are predominantly developed and tested using high-quality imagery from expert centers. However, their performance may be different when applied in clinical practice, partly due to the diversity in post-processing enhancement settings used in endoscopy units. We evaluated the impact of post-processing enhancement settings on AI performance and tested specific data augmentation strategies to mitigate performance loss. METHODS:We used a computer-aided detection (CADe) system for Barrett's neoplasia (6223 images, 906 patients) and a computer-aided diagnosis (CADx) system for colorectal polyps (3288 images, 969 patients), both trained on datasets acquired with Olympus equipment and with limited variability in enhancement settings. The CAD systems were then tested across a wide range of test sets, which comprised the same images, but displayed with different enhancement settings. Both CAD systems were then retrained using image enhancement-based data augmentation. The performance of the adjusted CAD systems was evaluated on the same test sets. RESULTS:Both systems displayed substantial performance variability over a range of enhancement settings (CADe: 83 %-92 % sensitivity, 84 %-91 % specificity; CADx: 78 %-85 % sensitivity, 45 %-63 % specificity). After retraining, variability in sensitivity and specificity was reduced to 2 % (P < 0.001) and 1 %, respectively (P = 0.003) for CADe, and 2 % (P = 0.03) and 8 %, respectively (P = 0.19) for CADx. CONCLUSION:The performance of endoscopic AI systems can vary substantially depending on post-processing enhancement settings of the endoscopy unit. Specific data augmentation can mitigate this performance loss.

Background and aimsIn this pilot study we evaluated performance of a recently developed computer-aided detection (CADe) system for Barrett’s neoplasia during live endoscopic procedures.Methods15 patients with and 15 without a visible lesion were included in this study. A CAD assisted workflow was employed that included: a slow pullback video recording of the entire Barrett’s segment with live CADe assistance, followed by CADe assisted level-based video recordings every 2cm of the Barrett’s segment. Outcomes were per patient and per level diagnostic accuracy of the CAD assisted workflow, where the primary outcome was per patient in-vivo CADe sensitivity.ResultsIn the per patient analyses, the CADe system detected all visible lesions (sensitivity 100%). Per patient CADe specificity was 53%. Per-level sensitivity and specificity of the CADe assisted workflow were 100% and 73%, respectively.ConclusionIn this pilot study, the CADe system detected all potentially neoplastic lesions in Barrett's esophagus comparable to an expert endoscopist. Continued refinement of the system may improve specificity. External validation in larger multicenter studies is planned

Background: To evaluate the proportion of patients with residual neoplasia after endoscopic resection (ER) for Barrett’s neoplasia with confirmed tumor-positive vertical resection margin (R1v). Methods: Retrospective cohort study including all patients treated with ER for Barrett’s neoplasia with histologically documented R1v since 2008 in the Dutch Barrett Expert Centers. R1v was defined as cancer cells touching the vertical resection margin and Rx as not assessable margins. Reassessment of R1v specimen was performed by experienced pathologists until consensus was reached regarding vertical margins. Results: 101/110 included patients had macroscopically complete resections (T1a n=17, T1b n=84), of which 99/101 (98%) ER specimens were reassessed. Reassessment confirmed R1v in 74 (75%) patients and found Rx in 16% and R0 in 9%. Presence of residual neoplasia could be assessed in 66/74 patients during endoscopic reassessment (n=52) and/or in the surgical resection specimen (n=14), of whom 33/66 (50%)had residual neoplasia. Residual neoplasia detected during endoscopy was always endoscopically visible and biopsies from a normal appearing ER-scar did not detect additional neoplasia. Twenty-five patients with no residual neoplasia during endoscopic reassessment underwent endoscopic follow-up for median 37 months(IQR 12-50), in which 4 developed a local recurrence(16.0%), all detected as visible abnormalities. Conclusions: Histological evaluation of ER margins appears challenging as 75% of documented R1v cases were confirmed during reassessment. After ER with R1v, 50% of the patients had no residual neoplasia. Endoscopic reassessment 8-12 weeks after ER seems accurately able to detect residual neoplasia and help to determine the most appropriate strategy for patients with R1v.

[This corrects the article DOI: 10.1016/j.jhepr.2019.10.006.].

Aims Although endoscopic resection (ER) is recommended as first-choice treatment for early esophageal neoplasia, patients with esophageal varices are considered a high risk group due to an increased bleeding risk. However, since most of these patients are precluded from major esophageal surgery due to portal hypertension, endoscopic therapy may be the only treatment option in these patients. This retrospective, multicenter study aimed to evaluate the efficacy and safety of endoscopic therapy of early esophageal neoplasia in this specific patient category.