Introduction

Extraintestinal manifestations (EIMs) can occur in up to 47% of patients diagnosed with inflammatory bowel disease (IBD); hence the early detection of EIMs in IBD patients is crucial to prevent a disabling disease clinical course. However, there is a substantial variation in the reported time to diagnosis of EIMs due to a lack of a standardised approach. The clinical characteristics leading to the diagnosis were rarely specified, and the diagnostic accuracy could depend on the experience of the individual specialist. Conversely, care for IBD patients with EIMs should be multidisciplinary. We, therefore, evaluated the length of time to diagnose EIM through an integrated immune-mediated diseases (IMIDs) outpatient clinic compared to a conventional diagnosis and the impact of diagnostic delay on the disease course of IBD.

Aims & Methods

We conducted a single-center cohort observational study including IBD patients ≥18 yo presenting red flags for EIMs who underwent outpatient visits through a multidisciplinary IMIDs clinic between 2020-2022 and we compared this group to a historical cohort, with similar clinical features, who underwent conventional visits by individual referral specialists from 2017 to 2019. The primary outcome was to assess the diagnostic delay, defined as the period from the occurrence of red flags onset to diagnosis of EIM. Secondary endpoints included the rate of EIMs diagnosis in patients with red flags referred to IMIDs clinic and the impact of diagnostic delay on clinical outcomes, including therapeutic changes.

Results

We enrolled 176 IBD patients; 90 were referred to the IMIDs Clinic Group (IMID-G) and 86 to the Standard Outpatient Clinic group (SOC-G). Demographic details are summarised in Table 1.

Demographic characteristics	IMIDs Clinic Group (n=90)	Standard Outpatient Clinic Group (n=86)
Sex - Female - Male	55 (61.1%) 35 (38.9%)	53 (61.6%) 33 (38.4%)
UC patients - Proctitis - Left colitis - Pancolitis	41 (45.6%) 8 (19.5%) 16 (39%) 17 (41.5%)	38 (33.2%) 5 (13.1%) 18 (47.3%) 12 (31.6%)
CD patients	49 (54.4%)	48 (55.8%)
CD Montreal disease localization - Ileal - Colonic - Ileocolonic	19 (38.8%) 7 (14.3%) 23 (46.9%)	16 (33.3%) 7 (14.6%) 25 (52.1%)
CD Montreal disease behaviour - Inflammatory - Stricturing - Penetrating	15 (30.6%) 18 (36.7%) 16 (32.7%)	14 (29.2%) 21 (43.8%) 13 (27.1%)
Red flags for IMIDs - Chronic back pain - Peripheral arthralgia - Enthesitis - Dactilitis - Skin erythema - Skin ulcers	26 (28.9%) 24 (26.7%) 2 (2.2%) 7 (7.8%) 22 (24.4%) 9 (10%)	26 (30.2%) 31 (36%) - 3 (3.5%) 14 (16.3%) 12 (14.0%)

The average time to EIMs diagnosis was 2.6±1.7 months for IMID-G and 4.4±1.9 months for SOC-G, p<0.05. In both groups, the majority of patients referred red flags for spondyloarthritis (SpA). In IMID-G , chronic back pain was reported by 28.9% of patients and peripheral arthralgia by 26.7%, similarly in SOC-G , 30.2% had chronic back pain and 36% had peripheral arthralgia. At six months, 77 (85.5%) patients with red flags referring to the IMIDs clinic received diagnoses of EIMs, compared to 52 (60.5%) of those referred to the specialist of reference (p<0.05). Overall in IMID-G, 19 (21.1%) patients were diagnosed with peripheral arthritis and 28 (31.1%) with SpA, compared to 15 (17.4%) with peripheral arthritis and 21 (24.4%) with SpA in SOC-G. Survival curves were plotted to analyse the probability of diagnostic delay of EIMs between IMID-G and SOC-G and showed a significant reduction of diagnostic delay in IMID-G (log rank test, p=0.001). Additionally, in IMID-G, 42 patients (46.6%) changed therapy, and 13 (14.4%) added steroids in comparison to 27 (31.4%) and 11 (12.8%) respectively in the SOC-G cohort (p<0.05).

Conclusion

Attendance at a dedicated IMIDs clinic can improve the diagnostic process of EIMs in an IBD cohort and reduce the diagnostic delay. The most common EIMs were SpA and peripheral arthritis. This multidisciplinary approach can help to avoid diagnostic delays, choose the correct therapies, prevent complications, and improve clinical outcomes and hence the quality of life.

Introduction

Anemia is a common and often underdiagnosed condition in the elderly population. Digestive endoscopy plays a pivotal role in the etiological workup, particularly to rule out serious underlying causes such as malignancies or peptic lesions. However, the diagnostic yield of such procedures in routine clinical practice remains variable and underreported, especially in resource-limited settings.

Aims & Methods

To assess the diagnostic yield of digestive endoscopy in the investigation of anemia in elderly patients, with a particular focus on iron deficiency anemia (IDA) and vitamin B12 deficiency, and to identify specific endoscopic findings associated with these etiologies.
We conducted a retrospective descriptive study over a 6-month period. The study included all patients aged 65 years and older who underwent esophagogastroduodenoscopy (EGD) and/or colonoscopy as part of the diagnostic evaluation for anemia. Data collected included demographic information, clinical symptoms, laboratory parameters (including hemoglobin, ferritin, vitamin B12 levels), endoscopic findings, and histopathological results when available.

Results

A total of 252 patients were included in the analysis. Among them, 44 patients (17.5%) with a mean age of 71 years [68-82], had confirmed IDA, and six patients (2.4%) presented with vitamin B12 deficiency anemia, associated in all cases with histologically confirmed fundic atrophy.
In univariate analysis, IDA was statistically correlated with the presence of peptic esophagitis (p=0.04), gastric neoplasia (p=0.001), erosive gastropathy (p=0.02), and Helicobacter pylori gastritis (p=0.04). It was also weakly correlated with the presence of colorectal cancer (p=0.05).
Colonic angiodysplasias and colonic polyps were not associated with IDA in this cohort (p=0.1 and p=0.4, respectively).

Conclusion

Digestive endoscopy, particularly EGD, remains a cornerstone in the diagnostic algorithm of anemia in the elderly. It enables the identification of a wide range of underlying pathologies, from inflammatory lesions to malignancies, and guides appropriate therapeutic interventions. Early endoscopic evaluation should be encouraged in this population to prevent diagnostic delays and improve outcomes.

Introduction

Colorectal endoscopic submucosal dissection (ESD) is spreading worldwide, with increasing demand for training. Standardized skills assessment has yet to be defined to guarantee the quality and safety of the results obtained by trainees during their learning curve. ESGE recommends that the first ESD procedures should be supervised by an ESD-proficient endoscopist. The impact of a trainee's active participation on these initial supervised ESD results should be addressed.

Aims & Methods

Between 01/2017 and 01/2024, all consecutive colorectal ESD procedures were prospectively registered by 3 participating ESD training centers, and data were analyzed retrospectively. Trainee participation was tracked. A trainee was defined as having performed less than 50 ESD’s on humans, alone or under supervision. We compared the main procedural outcomes with respect to the presence or absence of a trainee involved in the colorectal ESD procedure. Multivariate analysis and propensity score matching (PSM) were performed.

Results

2462 consecutive colorectal ESD were included, of which 543 procedures (22%) involved the active participation of a trainee (15 trainers/15 trainees). Rectal lesions were more often treated with participation of a trainee (56% vs 31%, p<0.001 with-trainee vs trainer-alone respectively). There was no significant difference in en-bloc, R0 and curative resection rates whether a trainee participated or not (95.8%, 85.9% and 84.9% vs 95.1%, 88.2% and 84.7%, with-trainee vs trainer-alone respectively). Participation of a trainee resulted in a significant decrease in dissection speed (29.4 mm²/min vs 43.9 mm²/min, p<0.001 with-trainee vs trainer-alone respectively). Rates of perforation, delayed bleeding and surgery for complications were 8.6%, 5.9% and 0.9% vs 9%, 4.5% and 0.7% for with-trainee vs trainer-alone respectively, without significant differences. After PSM, the presence of a trainee significantly decreased the dissection speed (29.4 mm²/min vs 42.7 mm²/min, p<0.001 with-trainee vs trainer-alone respectively), increased the perforation rate (8.7% vs 5%, p=0.023 with-trainee vs trainer-alone respectively), but did not influence the curative resection rate (84.9% vs 88.8%, p=0.078 with-trainee vs trainer-alone respectively).

Conclusion

Active participation of a trainee during colorectal ESD did not affect the curative resection rate. However slower dissection speed and higher perforation rates, without increasing the need for surgery, were observed.

Disclosure

Nothing to disclose

Introduction

Repeated endoscopy is recommended to detect early-stage cancer or Barrett's oesophagus known as a precancerous lesion at the gastro-oesophageal junction (GOJ) in GERD patients. Transnasal endoscopy is considered suitable for this repeated endoscopy because of less discomfort in the throat compared to conventional (peroral) endoscopy. Thus, improvement of imaging technique at the GOJ by transnasal endoscopy is one of important issues.

Aims & Methods

To evaluate the usefulness of trasnasal endoscopy equipped with high-resolution imaging, blue laser imaging (BLI) bright and linked color imaging (LCI) for the detection short-segment Barrett’s oesophagus (SSBO). We analyzed 300 images at the GOJ acquired from 100 consecutive endoscopies performed at our medical check-up center. Each endoscopy includes white light imaging (WLI), BLI bright and LCI. Ten endoscopists independently interpreted the images on SSBO and endoscopic findings related Barrett’s oesophagus, including palisade vessels of squamo-columnar junction (SCJ) oral side (PV-O), palisade vessels of anal side (PV-A), squamous island (SI) and columnar island (CI). The quality of images for visualization was judged as: excellent visibility (score of 4), good visibility (3), fair visibility (2), or not detectable (1) were evaluated.

Results

The detection rates of each image were as follows (WLI, BLI bright and LCI respectively): SSBO 28.3%, 32.7% and 38.8%; PV-O 88.5%, 83.1% and 90.5%; PV-A 47.3%, 51.1% and 63.2%; SI 9.2%, 9.5% and 9.8%; CI 18.4%, 28.7% and 28.7%. The detection rates of LCI were significantly higher than that of WLI on SSBO, PV-O, PV-A, SI and CI, and that of BLI-bright on PV-A and PV-O. The mean visibility scores (± standard deviation) were as follows: SSBO 1.59 ± 1.02, 1.65 ± 1.03 and 1.86 ± 1.18; PV-O 3.23 ± 1.05, 3.00 ± 1.15 and 3.36 ± 1.01; PV-A 1.99 ± 1.18, 1.98 ± 1.13 and 2.40 ± 1.25; SI 1.18 ± 0.60, 1.21 ± 0.66 and 1.20 ± 0.63; CI 1.43 ± 0.95, 1.74 ± 1.19 and 1.84 ± 1.43. The visibility scores for LCI were significantly higher than that for WLI and BLI-bright on SSBO and all items of related findings (<0.001). Moreover, PV-A showed the strongest correlation with SSBO on visibility score among endoscopic findings related to SSBO (R=0.500; p<0.001).

Conclusion

These results suggest that LCI with transnasal endoscopy is quite effective for the observation GOJ and detecting SSBO.

Introduction

The incidence of early-onset colorectal cancer is increasing alarmingly worldwide, and countermeasures against this issue are urgently needed. Although previous studies suggested age-related molecular pathological changes, the association between the roles of Immunosocore using the ration of CD3+ and CD8+ T cells in anti-tumor immune responses and early-onset colorectal cancer and adenoma resected by endoscopy remains unclear. This study aimed to investigate the association of Immunosocore using the ratio of CD3+ and CD8+ T cells and clinicopathological characteristics mainly focused on early-onset tumors. We defined early-onset colorectal tumors as those that occurred at less than 50 years of age at diagnosis and later-onset as those that occurred at 50 years of age or older in this study.

Aims & Methods

The aim of this study was to investigate the association between Immunosocore and clinicopathological features of early-onset colorectal tumors. Utilizing a molecular pathological database of 169 colorectal cancer and adenoma cases, we analyzed 306 colorectal lesions including 214 early-onset tumors resected by endoscopic procedures including cold snare polypectomy, endoscopic mucosal resection, and endoscopic submucosal dissection from February 2017 to January 2020 at Nippon Medical School Hospital in Japan. After collecting formalin-fixed paraffin-embedded (FFPE) tissue blocks, CD3 and CD8 immunostaining was performed, and the number of CD3+ and CD8+ cells was quantified in each area of tumor, stromal, and normal. One to five fields of view (median 5 fields) were selected depending on the size of the lesion using a fluorescence microscope BZ-X800 (KEYENCE) at 200x magnification, and the average number of CD3+ and CD8+ cells was calculated by automatic counting. Counting stromal and normal cells was limited to those specimens for which the excised specimens were large enough to contain sufficient stromal and normal portions in addition to the tumor portion. Immunoscore was calculated by the ratio of CD8+/CD3+ positive T cells. Clinicopathologic characteristics were also evaluated retrospectively. Chi-square test was used to compare categorical variables, Spearman's test to examine correlation, Mann-Whitney U test to compare continuous variables, and Kruskal-Wallis test to compare the three groups.

Results

The tumor location was 126 (41%)/142 (47%)/37 (12%) in the proximal/ distal colon/rectum. The number of CD3+T cells decreased significantly with increasing tumor diameter (P<0.001). The number of CD3+T cells was predominantly lower in type Ip (protruding type) compared to type non-Ip (P<0.001). The distribution of Immunoscore measured in tumor area was significantly different among tumor locations (P<0.05).

Conclusion

Using endoscopically resected specimens, we suggest that the anti-tumor immune response may differ according to tumor size, gross type, and location. Further studies are warranted to stratify early-onset tumors requiring careful surveillance after endoscopic resection in terms of the anti-tumor immune response in T cells.

Introduction

Diagnosing and assessing Inflammatory Bowel Disease (IBD) requires invasive procedures such as endoscopy and intestinal biopsies. However, reliable and accurate non-invasive tools for assessing and monitoring inflammation in IBD patients are in an exciting phase of early development. Raman spectroscopy (RS), which provides molecular vibrational spectra, can characterise human biofluids and intestinal samples, offering the potential for non-invasive molecular assessment, characterisation and monitoring in IBD.

Aims & Methods

This study aims to explore the potential of RS applied to saliva, blood and intestinal biopsies from IBD patients for molecular characterisation and development of new biomarkers enabling a reliable and accurate non-invasive diagnosis of IBD, phenotyping between Crohn’s disease (CD) and Ulcerative Colitis (UC) and predicting mucosal healing (MH). IBD patients undergoing advanced endoscopy, i.e. ultra-high magnification endocytoscopy or probe-based confocal laser endomicroscopy, for disease assessment or surveillance and healthy controls were prospectively enrolled from two university hospitals in Cork, Ireland. Peripheral blood, saliva, and biopsies from the most representative bowel areas were collected. All matching samples were processed through a novel waveguide-enhanced RS technique, enabling sensitive detection of metabolites and other biochemical components. Additionally, MH from matching biopsies was assessed as RHI≤3 without lamina propria or epithelial neutrophils, NHI ≤1 and PHRI=0. RS data processing was performed using Matlab®. Spectral markers matched for saliva, blood and biopsies and differentiating IBD vs controls, CD vs UC and MH vs active inflammation were identified. A machine learning model based on linear discrimination analysis was used to assess the classification accuracy.

Results

A cohort of 16 IBD patients (median age 39, 11:5 M: F), comprising 9 UC and 7 CD, and 4 healthy controls (median age 43.5, 3:1 M: F) was enrolled, with matching saliva, blood, and biopsies. MH was found in 4 IBD patients. When comparing IBD patients with controls, differences in 3 wavenumbers were observed in all samples, corresponding to decreased tyrosine and lipids signals and increased carotenoids signal in IBD. Similarly, differences in spectral markers were found in CD vs UC, with increased lipids signal in CD and increased signals for carotenoids, nucleic acid, and amide proteins in UC. Interestingly, 5 matching differences in saliva, plasma and biopsies spectra were found in patients with MH compared to those with active inflammation, including an increase in nucleic acids and amide proteins and a decrease in phosphate, lipids, and carotenoids. The machine learning model showed promising accuracy (above 82%) of RS-aided biofluids analysis in diagnosing IBD and predicting phenotype and MH.

Conclusion

The combination of spectral markers from saliva and plasma, confirmed at tissue level, showed promise in accurately diagnosing IBD and distinguishing CD from UC. Notably, RS non-invasive biomarkers could predict MH with promising accuracy, identifying metabolites involved in IBD-associated inflammation, particularly lipids, carotenoids, and nucleic acids. Therefore, RS holds potential as a monitoring and predictive tool for inflammation and healing in IBD, with intriguing translational implications. Further studies are required to validate these preliminary findings and shed light on molecular mechanisms.

Introduction

The advancement of computer-aided detection systems using artificial intelligence (AI) opened a new scope in endoscopic image analysis. However, the deep-learning-based models developed on closed datasets are susceptible to unknown anomalies in complex clinical environments. In particular, the high false positive rate in polyp detection remains a major challenge in clinical practice. The frequent occurrence of false positives (FPs) leads to endoscopists' fatigue, distraction, and the need for refocusing. It costs additional time and effort to discriminate FPs from true positives (TPs).

Aims & Methods

We propose a generic and practical solution for FP suppression in real-world AI-assisted colonoscopy. We release the FPPD-13 dataset, which includes real-world cases of TPs (1300 samples) and 13 classes of FPs (100 samples/class) with a comprehensive taxonomy. A YOLOv5 polyp detector was trained on a private dataset collected in Zhongshan Hospital. It was employed to analyze real-world colonoscopy videos for the collection of TPs and FPs. Two endoscopists discriminated all the positive predictions into TPs and FPs, and a third senior endoscopist was consulted for controversial images. We further propose EndoBoost, a plug-and-play module for FP suppression during polyp detection. EndoBoost follows the formulation of anomaly detection (AD) and takes FPs as anomalies. Specifically, a normalizing flow is utilized for density estimation in the feature space and rejecting FPs during real-time inference. We develop a learnable image encoder to obtain an informative feature space for the AD task. The image encoder and the normalizing flow are jointly optimized to learn the TP manifold while the FP samples are also exploited through outlier exposure.

Results

	AP	AUC	Accuracy	Precision	Sensitivity	Specificity
EndoBoost	0.788±0.017	0.793±0.020	0.705±0.019	0.655±0.026	0.877±0.061	0.534±0.078
KDE	0.646±0.024	0.610±0.027	0.569±0.021	0.541±0.016	0.928±0.036	0.209±0.076
OC-SVM	0.733±0.011	0.726±0.004	0.627±0.037	0.588±0.036	0.886±0.071	0.368±0.145
MCD	0.700±0.009	0.730±0.004	0.657±0.009	0.604±0.011	0.915±0.036	0.398±0.051
PCA (10)	0.654±0.020	0.677±0.013	0.611±0.017	0.567±0.012	0.948±0.024	0.275±0.047
PCA (100)	0.690±0.013	0.721±0.005	0.657±0.008	0.605±0.008	0.902±0.030	0.412±0.035
AE (128)	0.647±0.031	0.675±0.025	0.605±0.022	0.561±0.014	0.960±0.013	0.249±0.044
VAE (128)	0.718±0.008	0.703±0.008	0.588±0.028	0.557±0.025	0.902±0.062	0.275±0.118
iForest	0.644±0.034	0.634±0.026	0.577±0.015	0.544±0.010	0.953±0.025	0.200±0.046

Extensive experiments demonstrate a promising false positive suppression in both retrospective and prospective validation. EndoBoost was validated and compared to other competitors on the proposed dataset. We first evaluated the performance of different AD models when only TPs were accessible during training. EndoBoost achieved the highest AP and AUC among all the AD models (Table). Data-efficiency experiments were conducted to explore how to make the most use of available FPs during training. EndoBoost reached the equivalent performance (AP: 0.965, AUC: 0.966) with the pervious fully supervised method using only 10% FPs. For robustness to unseen classes of false positives, EndoBoost outperformed others with the highest AP (0.781) and AUC (0.974). In the prospective study, we took three colonoscopic video clips with a total length of about 1 hour as the real-world deployment test set. With the assistance of EndoBoost, FPs were suppressed successfully during the withdrawal of the colonoscopy.

Conclusion

In this work, we present a practical solution of dataset and methodology for reducing FPs during AI-assisted colonoscopy. The FPPD-13 dataset and proposed EndoBoost demonstrate their effectiveness in supressing FPs during real-world colonoscopy. For future work, multi-center clinical trials using EndoBoost are being carried out.