Expert Gastrointestinal Pathologists Research Articles

Preinjection With Ligation-Assisted Endoscopic Mucosal Resection for Barrett's Dysplasia and Early Esophageal Adenocarcinoma: Characteristic Histological Features of the Depth of Resection.

Endoscopic mucosal resection (EMR) is an established technique for the diagnosis and treatment of high-grade dysplasia (HGD) and early esophageal adenocarcinoma (EAC) in Barrett's esophagus. Submucosal preinjection is not universally used or generally recommended when performing routine ligation-assisted EMR. Prior studies, however, have demonstrated evidence of at least superficial muscle injury on ligation-assisted EMR without submucosal injection. There are limited published data supporting any potential benefit of submucosal preinjection. Our aim was to review this technique and determine the rate of any degree of muscle injury in patients with Barrett's HGD and EAC treated with submucosal preinjection before ligation-assisted EMR. Patients undergoing submucosal preinjection before ligation-assisted EMR for Barrett's esophagus at a single institution between 2012 and 2016 were identified. Data were collected regarding patient demographics and medical history, endoscopy and histopathology findings, adverse events, and subsequent outcomes. All EMR specimens were reviewed by an expert gastrointestinal pathologist. One hundred fifty consecutive EMR procedures were performed on 70 patients. Of 70 patients, 85.7% of patients were men, with a median age of 68 years. EAC was identified in 75 specimens (50%) and HGD in 44 specimens (29.3%). Deep resection margins were clear of adenocarcinoma in all specimens. Muscularis propria was not identified in any of the 150 specimens. There were no cases of post-EMR perforation. Preinjection before ligation-assisted EMR achieved complete excision with histologically clear margins, without histological evidence of any inadvertent muscularis propria.

Gastric Carcinoma in Autoimmune Gastritis: A Histopathologic and Molecular Study

Patients with autoimmune gastritis (AIG) have a 13-fold risk of developing type-1 neuroendocrine tumors, whereas the risk for gastric adenocarcinoma is still uncertain. Here we describe the clinicopathologic and molecular features of a series of gastric carcinomas (GC) arising in the context of AIG. A total of 26 AIG-associated GC specimens were collected from 4 Italian Institutions. Immunohistochemistry for MUC1, MUC2, MUC5AC, MUC6, CDX2, HER2, PD-L1, CLDN18, mismatch repair (MMR) proteins, and p53 and EBV-encoded RNA (EBER) in situ hybridization were performed. Histologic and immunohistochemical features were jointly reviewed by 5 expert gastrointestinal pathologists. Next-generation sequencing analysis (TrueSight Oncology 500, Illumina) of 523 cancer-related genes was performed on 19 cases. Most tumors were diagnosed as pT1 (52%) and they were located in the corpus/fundus (58%) and associated with operative link for gastritis assessment stage II gastritis (80.8%), absence of parietal cells, complete intestinal metaplasia, and enterochromaffin-like-cell micronodular hyperplasia. Only 4 (15.4%) GCs were diagnosed during follow-up for AIG. The following histotypes were identified: 20 (77%) adenocarcinomas; 3 (11%) mixed neuroendocrine–non-neuroendocrine neoplasms, and 2 (8%) high-grade solid adenocarcinomas with focal neuroendocrine component, 1 (4%) adenocarcinoma with an amphicrine component. Overall, 7 cases (27%) showed MMR deficiency, 3 (12%) were positive (score 3+) for HER2, 6 (23%) were CLDN18 positive, and 11 (42%) had PD-L1 combined positive score ≥ 10. EBER was negative in all cases. Molecular analysis revealed 5/19 (26%) microsatellite instability (MSI) cases and 7 (37%) tumor mutational burden (TMB) high. The most frequently altered genes were TP53 (8/19, 42%), RNF43 (7/19, 37%), ERBB2 (7/19, 37% [2 amplified and 5 mutated cases]), ARID1A (6/19, 32%), and PIK3CA (4/19, 21%). In summary, AIG-associated GCs are often diagnosed at low stage in patients with longstanding misrecognized severe AIG; they often display a neuroendocrine component or differentiation, have relatively higher rates of MMR deficiency, and TMB high.

DOP35 Spatial transcriptomics reveals cellular niches associated with histological inflammation in postoperative Crohn’s Disease

Abstract Background Postoperative Crohn’s disease (CD) recurs in ~70% of patients ≤1 year following ileocolonic resection (ICR), as observed endoscopically. Current understanding of molecular determinants underlying histological intestinal inflammation in patients with postoperative CD recurrence is minimal. To address this issue, we used advanced spatial transcriptomic analyses on intestinal biopsy samples with matched histological and endoscopic assessments from neoterminal ileum, anastomosis, and colon of patients with CD who had undergone colonoscopy following ICR. Methods Formalin-fixed paraffin-embedded (FFPE) tissue biopsies from neoterminal ileum, anastomosis, and colon were prospectively collected from a 15-patient cross-sectional cohort with CD who underwent routine endoscopic assessment following ICR. Spatial transcriptomics was performed on samples using 10X Genomics Visium CytAssist Spatial Gene Expression for FFPE V2.0. Annotation of tissue regions of histologically active inflammation and classification of histologically inactive (Robarts Histopathology Index ≤3 with subscores of 0 for neutrophils in the lamina propria and epithelium) or active disease were performed by an expert gastrointestinal pathologist. R programming language was used for data analysis (Figure). Results Fourteen of the 43 (32.6%) FFPE biopsies had histological evidence of inflammation. Spatial transcriptomics clustering revealed a distinct transcriptional signature by the segregation of 4 clusters ("inflamed clusters") composed mostly (≥69%) of cells from samples defined by histologically active disease. Tissue regions identified as inflamed by expert pathologist annotation contributed to >46% of inflamed clusters’ composition and <15% of remaining clusters. Cell type decomposition and cellular niche analysis revealed immune (e.g., myeloid, T cells, and B cells) and epithelial cells in close proximity sharing a similar microenvironment in 3 inflamed clusters while the 4th cluster was composed of >90% epithelial cells. Differentially expressed genes from the inflamed clusters were similarly expressed across intestinal segments including inflamed anastomotic biopsies and were enriched for signalling pathways involving immune cells recruitment/activation, inflammatory signalling, extracellular matrix organisation, and oxidative stress response. Conclusion This is the first study to explore spatial transcriptomic signatures associated with histological inflammation in patients with postoperative CD recurrence. Our analysis revealed a spatial signature of cellular niches and signalling pathways arising upon histological inflammation, providing insight into therapeutic targets and biomarkers across anatomic locations and the anastomosis.

P412 Faecal Calprotectin as a predictor of endoscopic and histologic remission in patients with Ulcerative Colitis

Abstract Background Faecal calprotectin (FCP) is a reliable surrogate marker for disease activity in ulcerative colitis (UC). However, there are no consensus cut-off values for histoendoscopic remission. The aim of the study is to correlate FCP with Mayo Endoscopic Score (MES) and histological disease activity (Geboes score) of UC patients in clinical remission. Methods We performed a prospective study including adult UC patients at the McGill IBD Center between 2013 and 2017. Patients in clinical remission (partial Mayo score ≤2) or disease relapse, undergoing endoscopy for disease activity or dysplasia surveillance, were enrolled. Before bowel preparation, FCP was collected. MES was documented during colonoscopy. Biopsies were taken throughout the colon and histological activity was assessed using Geboes score (GS) by a blinded expert gastrointestinal pathologist. Results The study included 253 patients, of which 117 (46%) were male with a mean age of 38.2 years (standard deviation [SD] ± 24.8). 46% of the patients had pancolitis at the time of colonoscopy. 5-aminosalicylate were used by 68.4% of patients while 20.9% were on immunomodulators, 19% on biologics and 0.4% on small molecules as treatment. Regarding endoscopic correlation, it showed that an FCP ≥ 200 μg/g predicts MES > 1 despite clinical remission, yielding sensitivity of 59% and specificity of 74%. A FCP ≥ 123 predicts MES > 0 despite clinical remission (58% sensitivity and 70% specificity), also aiding to differentiate MES 0 from MES 1 (61% sensitivity and 70% specificity). As to histologic correlation, a FCP ≥ 80 μg/g identified histological disease activity using Geboes criteria (GS score > 3.1) in patients with clinical remission, yielding 64.7% sensitivity, 58.7% specificity and 77% positive predictive value (Figure 1). When using a GS score threshold of > 2, a cut-off value of FCP ≥ 50 μg/g, appears to be the most clinically relevant to identify patients in clinical remission who continue to have active histologic inflammation, with an 49.6% sensitivity, 41.6% specificity and 40% positive predict value. Results are summarized in Table 1. Conclusion FCP correlates very well with endoscopic and histologic disease activity and can be used as a surrogate marker for disease activity. Our study prospectively demonstrated the optimal cut-off values helping to discriminate endoscopic healing and histologic remission.

P357 Automated assessment of histological disease activity in Ulcerative Colitis using a deep learning algorithm

Abstract Background Assessment of histological disease activity in patients with ulcerative colitis (UC) is becoming more important. Therefore, the DCA score was developed for use in clinical practice. Distribution (D), chronicity (C) and activity (A) are each scored on a three-point scale with distribution being dependent on the percentage of biopsy area affected by inflammatory changes. A major problem for implementation of any score is interobserver heterogeneity. To overcome this hurdle, we developed a deep learning algorithm to automatically compute the DCA score. Methods A retrospective cohort was created consisting of 117 adult patients with ulcerative colitis from which biopsies were taken during a colonoscopy performed in the setting of routine clinical evaluation (training set of 537 slides). From 75 cohort patients, an additional set of biopsies was available (test set of 299 slides). An expert gastrointestinal pathologist performed the DCA score. The model was trained on H&E-stained whole slide images scanned at 40x (0.243-μm pixel) resolution. Training was performed in a weakly supervised multiple-instance learning setup, using slide-level labels without region of interest annotations. The features from the convolutional neural network-based model trained at 2.5x, 5x, and 10x magnification were incorporated using a random forest classifier for final prediction. Model accuracy was evaluated using the test set. Results The algorithm differentiates between normal and inflamed with an accuracy of 90.9% (Table 1). Model performance from features aggregated across all magnification levels (i.e., 2.5x, 5x, and 10x) was superior to features from individual levels. This can be explained with the help of GradCam heatmaps (Figure 1), where we show that features learned at different magnification levels correspond to different aspects of histological changes in UC. The multiple instance approach did not appear to detect crypt abscesses, which may have resulted in a relatively poor performance in differentiating A1 from A2. The model accuracy for the 9-class DCA was 64.8%. Conclusion The algorithm first appears to assess architectural distortions at lower magnification and then the presence of cellular elements like neutrophils at higher magnification. This closely mirrors the approach of pathologists. The present model automates the scoring process and by supporting the daily diagnostic practice highlights the potential of deep learning technologies in assessment of histological disease activity in UC.

WATS 3D : An Interobserver Study of Barrett's Esophagus-Associated Dysplasia Among Gastrointestinal Pathologists.

Wide-area transepithelial sampling with 3-dimensional computer-assisted analysis (WATS 3D ) has been shown to increase the detection rate of dysplasia (and intestinal metaplasia) in patients with Barrett's esophagus (BE). The purpose of this study was to evaluate the interobserver variability and accuracy of diagnosing BE-associated dysplasia in WATS 3D specimens among gastrointestinal (GI) pathologists without prior experience with this technology. Five GI pathologists underwent a 4-hour in-person (at microscope) and virtual training session and then evaluated digital images of discrete cellular foci from 60 WATS 3D cases with BE (20 nondysplastic BE [NDBE], 20 low-grade dysplasia [LGD], and 20 high-grade dysplasia/esophageal adenocarcinoma [HGD/EAC]). Each case consisted of 1 hematoxylin and eosin-stained image (cell block), and 1 liquid cytology or papanicolaou-stained smear image (120 images in total). The overall kappa value among the 5 study pathologists was excellent (overall kappa = 0.93; kappa = 0.93 and 0.97 for cell block and smear specimens, respectively). There were no significant differences noted in kappa values in interpretation of the cell block vs smear specimens or in any of the individual diagnostic categories when the latter were evaluated separately. Furthermore, agreement was perfect (100%) regarding detection of neoplasia (either LGD, HGD, or EAC). Diagnoses were made with complete confidence in 91% of instances. We conclude that GI pathologists, without any prior experience in interpretation of WATS 3D specimens, can undergo a short training session and then diagnose these specimens with a very high level of accuracy and reproducibility.

Multidisciplinary Inflammatory Bowel Disease Conference: The Impact of the Expert Pathologist on Patient Care.

Multidisciplinary teams (MDT) aid the diagnosis and management of patients with inflammatory bowel disease (IBD) and improve patient outcomes. The direct impact of a gastrointestinal expert pathologist on MDT care of IBD patients is unknown. A retrospective chart review was conducted evaluating all cases (N = 289) discussed at the IBD MDT conference at Carilion Roanoke Memorial Hospital from June 1, 2013, through December 31, 2019. Cases were discussed between 1 and 6 times at the conference. Data collected included demographics, diagnosis before and after conference, reason for diagnostic change, endoscopy findings, medications, surgeries, and clinical follow-up. Approximately 15% to 42% of patients had a change in diagnosis after the first 3 conferences. The majority of diagnostic changes after the first (84%), second (73%), and third (67%) conferences were due to expert pathologist interpretation. Indeterminate colitis was the most frequently changed diagnosis, and Crohn's disease was the most common new diagnosis after conference. Among patients with a diagnostic change, 28.6% to 38.5% of patients had a change in their IBD medication regimen, and 7.7% to 10.9% had a surgical intervention after the first 2 conferences. Approximately 54.2% to 60% of patients reported clinical improvement or remission within 6 months of the first 3 conferences. The majority of diagnostic changes made at the multidisciplinary IBD conference were due to histopathologic re-interpretation. A change in diagnosis at times led to significant modifications in medical or surgical management. An expert gastrointestinal pathologist is an essential MDT member for IBD management.

Endoscopic features of low-grade dysplastic Barrett's.

Background and study aims Barrett's esophagus (BE) with low-grade dysplasia (LGD) is considered usually endoscopically invisible and the endoscopic features are not well described. This study aimed to: 1) evaluate the frequency of visible BE-LGD; 2) compare rates of BE-LGD detection in the community versus a Barrett's referral unit (BRU); and 3) evaluate the endoscopic features of BE-LGD. Patients and methods This was a retrospective analysis of a prospectively observed cohort of 497 patients referred to a BRU with dysplastic BE between 2008 and 2022. BE-LGD was defined as confirmation of LGD by expert gastrointestinal pathologist(s). Endoscopy reports, images and histology reports were reviewed to evaluate the frequency of endoscopically identifiable BE-LGD and their endoscopic features. Results A total of 135 patients (27.2%) had confirmed BE-LGD, of whom 15 (11.1%) had visible LGD identified in the community. After BRU assessment, visible LGD was detected in 68 patients (50.4%). Three phenotypes were observed: (A) Non-visible LGD; (B) Elevated (Paris 0-IIa) lesions; and (C) Flat (Paris 0-IIb) lesions with abnormal mucosal and/or vascular patterns with clear demarcation from regular flat BE. The majority (64.7%) of visible LGD was flat lesions with abnormal mucosal and vascular patterns. Endoscopic detection of BE-LGD increased over time (38.7% (2009-2012) vs. 54.3% (2018-2022)). Conclusions In this cohort, 50.4% of true BE-LGD was endoscopically visible, with increased recognition endoscopically over time and a higher rate of visible LGD detected at a BRU when compared with the community. BRU assessment of BE-LGD remains crucial; however, improving endoscopy surveillance quality in the community is equally important.

Three pathologic criteria for reproducible diagnosis of colonic sessile serrated lesion versus hyperplastic polyp

Colonic SSLs are thought to predispose to ∼30% of colonic adenocarcinomas. This increased risk, compared to benign HPs, makes their distinction vitally important. However, no gold standard exists to differentiate them, and wide observer variability is reported. To better distinguish these polyps, we investigated 94 serrated polyps (53 SSLs and 41 HPs)using an easy-to-apply pathologic scoring system that combines, for the first time, three established distinguishing features: polyp morphology, location, and size. As an additional novel approach, polyp size was assessed by serrated biopsy number compared to endoscopic size. RNA expression profiling served as an additional biomarker. The considerable morphologic overlap across serrated polyps was quantitated for the first time. Interobserver variability was assessed by 8 expert gastrointestinal pathologists. By ROC analysis, polyp size by biopsy number performed best, followed by polyp location and morphology (areas under the curves [AUCs] = 85.9%, 81.2%, and 65.9%, respectively). Optimal discrimination combined all 3 features (AUC = 92.9%). For polyp size, the biopsy number proved superior to endoscopic size (AUC = 85.9% versus 55.2%, P=.001). Interobserver variability analysis yielded the highest reported Fleiss and Kappa statistics (0.879) and percent agreement (96.8%), showing great promise toward improved diagnosis. The proposed 3-criteria pathologic system, combining size by biopsy number, location, and morphology, yields an improved, easy-to-use, and highly reproducible diagnostic approach for differentiating SSLs and HPs.

MACHINE LEARNING-BASED PREDICTION OF GEBOES SCORE AND HISTOLOGIC IMPROVEMENT AND REMISSION THRESHOLDS IN ULCERATIVE COLITIS

Abstract BACKGROUND Histology is emerging as a key therapeutic endpoint for ulcerative colitis driven by associations between histologic response and long-term outcomes. However, existing scoring systems are subjective and consequently have variable inter- and intra-reader variability. Geboes scoring is a well-established system for ulcerative colitis histologic assessment that has previously been used to define thresholds for histo-endoscopic mucosal improvement (Geboes Score ≤3.1, together with Mayo score 0 or 1) and histologic remission (Geboes Score <2). Here we report the first machine learning (ML)-based prediction of the Geboes Score, and Geboes Score-derived thresholds of histologic improvement and remission, directly from whole slide images (WSI) of hematoxylin and eosin (H&E)-stained mucosal biopsies. METHODS 3,148 WSI were scored by three expert gastrointestinal pathologists and the median consensus score was used to determine the Geboes score for each slide as ground truth. ML models were trained on median consensus scores to predict the Geboes score and subscores for each slide. Model performance vs. pathologist median consensus score was measured using accuracy and the F1 score, which accounts for both false positive and false negative errors. RESULTS The ML-based model performance, measured against median consensus scores of three pathologists, showed strong performance at predicting overall Geboes Score, with a quadratic kappa of 0.89. The model was also able to predict both histologic improvement and histologic remission with high accuracy. For predicting histological improvement as defined by a Geboes Score of ≤3.1, the model showed accuracy of 0.92 and F1 score of 0.92 (Figure 1). For predicting histological remission as defined by a Geboes Score of < 2, the model showed accuracy of 0.91 and F1 score of 0.89 (Figure 2). CONCLUSIONS We report a ML-based approach for predicting Geboes score and Geboes score-based key thresholds of histologic improvement and histologic remission. Model predictions show high accuracy compared to median consensus pathologist scores. This approach may enable standardized, reproducible and accurate prediction of these clinically relevant thresholds to better measure histologic disease activity and treatment response in clinical trials.

Multicenter Randomized Controlled Trial of Surveillance Versus Endoscopic Therapy for Barrett’s Esophagus With Low-grade Dysplasia: The SURVENT Trial: Study Rationale, Methodology, Innovation, and Implications

Multicenter Randomized Controlled Trial of Surveillance Versus Endoscopic Therapy for Barrett’s Esophagus With Low-grade Dysplasia: The SURVENT Trial: Study Rationale, Methodology, Innovation, and Implications

Development and Validation of a Digital Analysis Method to Quantify CD3-immunostained T Lymphocytes in Whole Slide Images of Crohn's Disease Biopsies.

The T-lymphocyte-mediated inflammation in Crohn's disease can be assessed by quantifying CD3-positive T-lymphocyte counts in colonic sections. We developed and validated a process to reliably quantify immunohistochemical marker-positive cells in a high-throughput setting using whole slide images (WSIs) of CD3-immunostained colonic and ileal tissue sections. In regions of interest (ROIs) and/or whole tissue sections of 40 WSIs from 36 patients with Crohn's disease, CD3-positive cells were quantified by an expert gastrointestinal pathologist (gold standard) and by image analysis algorithms developed with software from 3 independent vendors. Semiautomated quantification of CD3-positive cell counts estimated in 1 ROI per section were accurate when compared with manual analysis (Pearson correlation coefficient, 0.877 to 0.925). Biological variability was acceptable in digitally determined CD3-positive cell measures between 2 to 5 ROIs annotated on the same tissue section (coefficient of variation <25%). Results from computer-aided analysis of CD3-positive T lymphocytes in a whole tissue section and the average of results from 2 to 5 ROIs per tissue section lacked reliability (overestimation or underestimation and systematic bias), suggesting that absolute quantification of CD3-positive T lymphocytes in a whole tissue section may be more accurate. Semiautomated image analysis in WSIs demonstrated reproducible CD3-positive cell measures across 3 independent algorithms. A computer-aided digital image analysis method was developed and validated to quantify CD3-positive T lymphocytes in colonic and ileal biopsy sections from patients with Crohn's disease. Results support consideration of this digital analysis method for use in future Crohn's disease clinical studies.

The utility of P53 immunohistochemistry in the diagnosis of Barrett's oesophagus with indefinite for dysplasia.

AimsBarrett's oesophagus with indefinite for dysplasia (BE‐IND) is a subjective diagnosis with a low interobserver agreement (IOA) among pathologists and uncertain clinical implications. This study aimed to assess the utility of p53 immunohistochemistry (p53‐IHC) in assessing BE‐IND specimens.Methods and resultsArchive endoscopic biopsies with a BE‐IND diagnosis from two academic centres were analysed. First, haematoxylin and eosin‐stained slides (H&E) were reviewed by four expert gastrointestinal (GI) pathologists allocated into two groups (A and B). After a washout period of at least 8 weeks, H&E slides were reassessed side‐to‐side with p53‐IHC available. We compared the rate of changed diagnosis and the IOA for all BE grades before and after p53‐IHC. We included 216 BE‐IND specimens from 185 patients, 44.0 and 32.9% of which were confirmed after H&E slide revision by groups A and B, respectively. More than half the cases were reclassified to a non‐dysplastic BE (NDBE), while 5.6% of cases in group A and 7.4% in group B were reclassified to definite dysplasia. The IOA for NDBE, BE‐IND, low‐grade dysplasia (LGD) and high‐grade dysplasia (HGD)/intramucosal cancer (IMC) was 0.31, 0.21, −0.03 and −0.02, respectively. Use of p53‐IHC led to a >40% reduction in BE‐IND diagnoses (P < 0.001) and increased IOA for all BE grades [κ = 0.46 (NDBE), 0.26 (BE‐IND), 0.49 (LGD), 0.35 (HGD/IMC)]. An aberrant p53‐IHC pattern significantly increased the likelihood of reclassifying BE‐IND to definite dysplasia (odds ratio = 44.3, 95% confidence interval = 18.8–113.0).ConclusionP53‐IHC reduces the rate of BE‐IND diagnoses and improves the IOA among pathologists when reporting BE with equivocal epithelial changes.

Risk of progression in Barrett’s esophagus based on diagnoses of general and gastrointestinal pathologists. A retrospective case-control study from Northern and Central Finland

Background Esophageal adenocarcinoma (EAC) is the sixth leading cause of cancer-related death worldwide. It develops through Barrett’s metaplasia – dysplasia sequence. However, the effectiveness of endoscopic surveillance is limited, since diagnosis of low-grade dysplasia (LGD) is known to be challenging for pathologists. Our aim was to compare the risk of Barrett’s progression based on diagnoses of general and expert gastrointestinal (GI) pathologists in a population-based cohort. Methods A total of 60 patients with non-dysplastic metaplasia (BE) or LGD progressing to high grade dysplasia (HGD) or EAC during follow-up could be identified in the population. For comparison, series representing non-progressive BE (n = 56) and LGD cases (n = 54), matched for age, gender, and length of follow-up were collected. All available original HE stained slides (n = 292) were blindly re-evaluated by two experienced GI pathologists and patient groups of progressive non-progressive BE and LGD were formed according to revised diagnoses. Results Original diagnosis for each sample was changed in 25% of BE, 59% of LGD, and 33% of HGD diagnoses. Of the original LGD diagnoses, 53% were downgraded to BE or indefinite for dysplasia (ID). Of LGD diagnoses made by an expert GI pathologist, 61% were in the progressive LGD group, whereas only 42% of general pathologists’ LGD diagnoses were in the progressive LGD group. Conclusion Based on this retrospective case-control study, LGD is strongly over-diagnosed among general pathologists. LGD diagnosed by expert GI pathologists predicts progressive disease. Recommendation for consensus diagnosis by expert GI pathologists is justified also in the Finnish population-based setting.

S440 Clinical Characteristics and Endoscopic Features in Patients With Esophageal Graft versus Host Disease

Introduction: Graft versus host disease (GVHD) is an immunologic disorder most commonly following hematopoietic stem cell transplantation (HSCT) that frequently involves the gastrointestinal tract but not the esophagus. The aim of this study was to describe the clinical characteristics and endoscopic features in patients with esophageal GVHD. Methods: In this retrospective cohort study, we identified all patients with esophageal GVHD at Mayo Clinic between January 1, 2000 and June 3, 2021. Patients were included if they had a histologic diagnosis of GVHD from esophageal biopsies interpreted by an expert gastrointestinal pathologist (C.E.H). Retrospective chart review was performed to gather data on clinical characteristics and endoscopic features. Numerical and categorical data were reported as median (range) and proportions. Results: Forty-three patients with esophageal GVHD were included with median age 59 (5-75) years and 29 (67%) were men (Table 1). Forty-2 (98%) had received HSCT and 1 liver transplant. The median time from transplant to esophageal GVHD was 3 (0-39) months. Twenty-5 (58%) patients had known GVHD prior to esophageal involvement. Other organ involvement with GVHD was common at esophageal GVHD diagnosis, most commonly involving skin (72%). Clinical symptoms at esophageal GVHD diagnosis included nausea/vomiting (63%), anorexia (51%), diarrhea (44%), weight loss >5% (37%), and dysphagia (26%). Odynophagia (5%) and heartburn (5%) were less common. EGD showed no abnormalities in the esophagus in 13 (30%) patients. Where esophageal endoscopic abnormalities were present, 20 (67%) were diffuse and 10 (33%) were focal, and included erosive esophagitis in 10 (23%) patients, ulceration 7 (16%), desquamation 6 (14%), rings/furrows 6 (14%), and erythema/edema 5 (12%) (Figure 1). Concurrent endoscopic abnormalities were present in the stomach in 24 (56%) patients, duodenum 16 (37%), and colon 15 (35%). All patients were treated with steroids and 16 (37%) had recurrent gastrointestinal symptoms with steroid discontinuation. Twenty-one (49%) patients had a follow-up EGD at median 4.5 (0-29) months with 11 (52%) having persistent esophageal GVHD and 3 (14%) requiring dilation. Conclusion: As esophageal GVHD is associated with gut and other organ involvement and clinical and/or endoscopic findings are non-specific or absent, there should be a low threshold for biopsy when esophageal GVHD is suspected in the presence of other gastrointestinal involvement.Figure 1.: Endoscopic findings seen with esophageal graft versus host disease include A) erosive esophagitis, B) ulceration, C) desquamation, and D) rings.Table 1.: Clinical characteristics in patients with esophageal graft versus host disease.

P137 Preliminary validation of a multi-stage machine learning algorithm to assess histological inflammation in inflammatory bowel disease

Abstract Background The histologic assessment of inflammatory bowel disease (IBD) relies on qualitative grading methods. Although widely accepted, these instruments are time consuming, require specialized training, and suffer from inter-rater disagreement. For these reasons there is a need for more consistent and less biased methods to assess IBD histology. Methods The algorithm was initially developed using hematoxylin and eosin (H&amp;E) stained whole slide images of colon biopsies (238 ulcerative colitis [UC], 30 Crohn’s Disease [CD], and 28 endoscopically normal adjacent [ENA]). The first two stages implement convolutional neural networks (CNN) which segment 11 key anatomical features (Figure 1). The third stage extracts the features and models them for prediction. Results The first stage of the algorithm was validated on an independent test dataset by calculating the intersection-over-union (IoU) for the ground truth and prediction masks, resulting in a value of 0.97. A preliminary validation for stage 2 was performed by randomly selecting 30 unique biopsy sections from the test dataset and applying a 150um x 150um counting frame. An expert gastrointestinal pathologist confirmed correct cell identification by the algorithm for three of the primary inflammatory cell types: plasma cells, eosinophils, and neutrophils which resulted in a sensitivity/specificity of 0.76/0.99, 0.78/1.00, and 1.00/0.98 respectively. The final stage predicts RHI grades which could be directly compared to pathologist reads (Figure 2, 3, and 4). Conclusion This is the first study to demonstrate the value of machine learning to assess histologic activity in IBD. These methods lay the foundations for future work, and we believe stages 1 and 2 can be explored independently to statistically characterize the histologic changes of IBD, enabling the improvement of preexisting grading systems.

Prevalence of Barrett's Esophagus in Female Patients With Scleroderma.

Systemic sclerosis or scleroderma (SSc) is a chronic autoimmune disease that renders the esophagus prone to significant gastroesophageal reflux due to impaired esophageal clearance and reduced lower esophageal sphincter pressure. The reported prevalence of Barrett's esophagus (BE) in women with SSc varies from 2% to 37% and is derived from older studies with small sample sizes. We aimed to assess the prevalence of BE in a large cohort of women with SSc. Women with SSc referred from the Mayo Clinic Arizona Rheumatology Clinic who completed esophagogastroduodenoscopy between 2002 and 2020 were included. Demographic and high-resolution manometry data were evaluated. The diagnosis of scleroderma was confirmed by an expert rheumatologist. The BE diagnosis was confirmed by an expert gastrointestinal pathologist. There were 235 women with SSc who underwent EGD. High-resolution manometry (HRM) was completed in 172 patients. Women with SSc with BE were significantly more likely to have scleroderma esophagus (absent contractility with hypotensive lower esophageal sphincter) on HRM than women with SSc without BE (P = 0.018). There were 30 patients with SSc (12.8%) with histologically proven BE. Dysplasia was found in 13 (43.3%): 4 with indefinite, 7 with low grade, and 2 with adenocarcinoma. The incidence of any dysplasia was 5.3% per year (0.9% per year for adenocarcinoma). This the largest study on prevalence of BE in women with SSc, yielding a prevalence of 12.8%. Women with SSc with BE were significantly more likely to have absent contractility with hypotensive lower esophageal sphincter findings on HRM. The high prevalence and incidence of dysplasia found suggest that women with SSc should be included in the screening recommendations for BE.

Neutrophil to lymphocyte ratio and risk of neoplastic progression in patients with Barrett's esophagus.

Patient's with Barrett's esophagus (BE) are at risk of progression to esophageal adenocarcinoma (EAC). Neutrophil to lymphocyte ratio (NLR) was found to be a predictor of poor prognosis in patients with EAC; however, its performance in premalignant esophageal lesions is vague. We aimed to evaluate the utility of NLR as a predictor of histologic progression in patients with BE. METHODS : A prospective cohort of patients with proven BE in a tertiary referral center was retrospectively analyzed. All biopsies were reviewed by an expert gastrointestinal pathologist. The discriminatory capacity of NLR was evaluated by area under the receiver operating characteristic (AUC) curve analysis and Cox regression analysis. 324 patients (mean age 62.3 years, 241 [74.4 %] males) were included in the final analysis. Overall, 13 patients demonstrated histologic progression to neoplasia over a mean follow-up of 3.7 years (progression risk 1.0 % per year). The AUC of NLR for progression to high grade dysplasia (HGD) or EAC was 0.88 (95 % confidence interval [CI] 0.83 - 0.96), and baseline NLR was associated with a 3-fold increase of progression to HGD and EAC during follow-up (hazard ratio [HR] 3.2, 95 %CI 1.5 - 5.8; P < 0.001). Notably, in a subgroup analysis of patients with nondysplastic BE (NDBE) at presentation, NLR was also a risk factor for histologic progression (HR 2.4, 95 %CI 1.7 - 3.4; P < 0.001). NLR predicted histologic progression in patients with BE. Patients with NDBE and NLR above 2.4 can be considered for specific surveillance programs with shorter intervals between sessions.

S0383 Prevalence of Barrett’s Esophagus in Female Scleroderma Patients

INTRODUCTION: Scleroderma renders the esophagus prone to significant gastroesophageal reflux due to impairment in esophageal clearance (aperistalsis in the lower third of the esophagus) and reduced lower esophageal sphincter (LES) pressure. Scleroderma mainly affects women, whereas Barrett’s esophagus (BE) predominantly affects men. The reported prevalence of BE in women with scleroderma varies widely from 6% to 37%, and is mainly derived from older small studies. We aimed to assess the true prevalence of BE in a large cohort of women with scleroderma. METHODS: Clinical data were collected from two prospectively maintained databases at a single academic tertiary referral center. One database consisted of scleroderma patients referred for high-resolution esophageal manometry (HREM). The other database consists of all patients with BE and associated clinical and endoscopic data. The diagnosis of scleroderma was confirmed by a rheumatologist with expertise in scleroderma according to the 2013 American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) criteria for systemic sclerosis/scleroderma. Inclusion criteria included female sex, confirmed scleroderma, and a diagnosis of Barrett’s esophagus confirmed by an expert gastrointestinal pathologist. Prevalence data were tabulated for each database and in aggregate. RESULTS: There were 286 women with scleroderma who underwent HREM (Table 1). There were 17 patients with histologically proven Barrett’s esophagus, representing 5.9% of the group. Of these Barrett’s patients, 11 had non-dysplastic Barrett’s esophagus, 2 had indefinite dysplasia, 3 had low grade dysplasia (LGD), and 1 had esophageal adenocarcinoma (EAC) (Table 2). Of 616 women with Barrett’s esophagus, 33 had scleroderma (5 diffuse, 25 limited and 3 unclassified). CONCLUSION: In this large cohort of female patients, the prevalence of BE in patients with scleroderma presenting for esophageal evaluation was 5.9% and the prevalence of scleroderma in women with Barrett’s esophagus was 5.4%. Although uncommon, both LGD and EAC were found in this population. BE is a prevalent complication of scleroderma and screening for its presence should be considered.Table 1

Best techniques for endoscopic ampullectomy

Endoscopic ampullectomy has proven its safety and efficacy in multiple studies, making it the favorable option for the management of endoscopic ampullectomy. In this review, we plan to review the basic steps every endoscopist should be familiar with before undertaking endoscopic ampullectomy. In this review, we plan to discuss the indications, preresection evaluation process including endoscopic ultrasound, endoscopic ampullectomy techniques with side-viewing endoscope and endoscopist experience/comfort with the management of the endoscopic ampullectomy complications. Things of utmost interest include reviewing all the ampullary biopsy specimens by an expert gastrointestinal pathologist, careful preresection staging, en bloc resection, familiarity with tools and techniques of endoscopic ampullectomy, postresection adverse events & management and postprocedure surveillance.