Expert GI Pathologists Research Articles

Histopathological Evaluation of Pouch Neoplasia in IBD and Familial Adenomatous Polyposis.

IPAA is often required for patients with ulcerative colitis or familial adenomatous polyposis after colectomy. This procedure reduces but does not completely eliminate the risk of neoplasia. This study focuses on the histopathology of neoplasia in the ileal pouch, rectal cuff, and anal transition zone. We performed a MEDLINE search for English-language studies published between 1981 and 2022 using the PubMed search engine. The terms "ileal pouch-anal anastomosis," "pouchitis," "pouch dysplasia," "pouch lymphoma," "pouch squamous cell carcinoma," "pouch adenocarcinoma," "pouch neoplasia," "dysplasia of rectal cuff," and "colitis-associated dysplasia" were used. Human studies of neoplasia occurring in the pouch and para-pouch were selected, and the full text was reviewed. Comparisons were made within and across studies, with key concepts selected for inclusion in this article. Neoplasia in the pouch is a rare complication in patients with IPAA. Annual endoscopic surveillance is recommended for familial adenomatous polyposis patients and ulcerative colitis patients with a history of prior dysplasia or carcinoma. In familial adenomatous polyposis, dysplastic polyps of the pouch are visible and readily amenable to endoscopic removal; however, glandular dysplasia in the setting of ulcerative colitis may be invisible on endoscopy. Therefore, random biopsies and adequate tissue sampling of the pouch and rectal cuff are recommended in this setting. The histological diagnosis of IBD-associated dysplasia can be challenging and should be confirmed by at least 1 expert GI pathologist. See video from the symposium.

P939 Ozanimod-Exposed Patients with Ulcerative Colitis Undergoing Total Colectomy Exhibit Unique Histologic Changes in Lymph Nodes

Abstract Background Ozanimod regulates lymphocyte egress from the spleen and lymph nodes into the systemic circulation. The histologic changes which occur in the lymph nodes of patients on these therapies is unknown. We describe histologic changes in lymph nodes from patients with UC who were treated with ozanimod and underwent subsequent colectomy due to treatment-refractory disease. Methods This retrospective study included patients with UC undergoing total colectomy for treatment-refractory disease who were treated with ozanimod within the 2 months prior to colectomy and a cohort of patients with UC undergoing colectomy who did not have ozanimod exposure. Histology of the lymph nodes from the mesentery of colectomy specimens were reviewed blindly by a hematophatologist (JXC) and an expert GI pathologist (CRW). Histopathological parameters were scored and demographic and clinical data were recorded. Results Of forty-five patients treated with ozanimod at our center, 6 (13%) required surgery for treatment-refractory disease. Colectomy specimen data were available for 5 patients (1 patient had surgery at an outside center). Lymph node specimens from 6 control patients with UC who had colectomy were also examined. Demographic and clinical data are summarized in Table 1. The duration of ozanimod therapy prior to surgery ranged from 6 to 24 weeks. Lymph nodes from patients treated with ozanimod had significantly greater extent of sinus dilatation than control patients (1.5 vs 0.67, p=0.022), they also had significantly more sinus histiocytosis (2.67 vs 1.67, p=0.03) (Figure 1). There was a tendency towards more Castleman-like angiotrophic hyperplasia (1.5 vs 1, p=0.34), plasma cell infiltration (1.5 vs 0.67, p=0.26) and subcortical interfollicular expansion (1.5 vs 0.83, p=0.1) in ozanimod treated patients. Patients treated with ozanimod and controls showed similar amounts of monomorphic lymphoid cells (1.3 vs 1.5, p=0.7), follicular hyperplasia (1 vs 1.1, p=0.8) and vascular proliferation (1.5 vs 1.5, p=1). Conclusion This study identifies unique histologic changes in the lymph nodes of patients with UC treated with ozanimod. These changes could be in keeping with the known mechanism of action of ozanimod and suggests that other inflammatory pathways were driving their disease activity in these patients. The significance of the changes, particularly the Castleman-like features, needs to be further investigated. Table 1: Demographic and clinical data of patients treated with ozanimod and control patients Figure 1: Representative images of lymph nodes from patients with UC treated with ozanimod and control patients showing significant sinus dilatation in patients receiving ozanimod

Seattle Protocol Is More Effective in Detection of Dysplasia Compared to Technology-Assisted Targeted Biopsies in Patients with Barrett's Esophagus.

With the development of narrow-band imaging (NBI) in the endoscopic evaluation of patients with Barrett's esophagus (BE), the role of random biopsies according to the Seattle protocol (SP) has been questioned. We aim to compare the utility of advanced imaging to SP in patients with BE. A prospective cohort of patients with proven BE was retrospectively analyzed. All biopsies were reviewed by an expert GI pathologist. Advanced imaging was tandemly used with SP in each endoscopic procedure. A total of 155 out of 340 patients (45.5%) with BE were diagnosed with dysplasia during a median follow-up of 4.7 years (IQR 3.4-6.1 years) and were part of the statistical analysis. A total of 82 patients had a diagnosis of dysplasia at presentation, whereas 84 patients developed dysplasia during follow up. A total of 67 out of 82 patients with dysplasia at presentation (81.7%), and 65 out of 84 patients that were diagnosed with dysplasia during follow-up (77.4%) were diagnosed using SP. In addition, whereas all the events of EAC were diagnosed using targeted biopsies, 57.1% of the events of HGD and 86.3% of LGD were diagnosed using SP. Our findings demonstrate the significance of SP in the detection of low- and high-grade dysplasia in patients with BE. SP should remain the mainstay of endoscopic surveillance in this population.

S410 Algorithm Training and Independent Test Set Performance for a Molecular Non-Endoscopic Test for Detection of Esophageal Adenocarcinoma and Barrett’s Esophagus in Multicenter Cohorts

Introduction: Sedated endoscopy for Barrett’s esophagus (BE) and esophageal adenocarcinoma (EAC) detection is invasive and expensive. Non-endoscopic BE/EAC detection tools have been guideline-endorsed to facilitate higher patient participation at lower cost. We previously described a promising panel of 5 methylated DNA markers (MDMs) assayed on esophageal specimens obtained by a sponge-on-a-string (SOS) cell collection device in phase II studies. We aimed to train an algorithm (establishing marker cut offs, to adjudicate samples as positive/negative) using a final MDM panel followed by testing in an independent sample set. Methods: Algorithm training samples (N=352) were prospectively collected from patients seen at 6 US medical centers. Test samples (N=125) were obtained from an independent, NIH-funded study conducted at 3 US medical centers. Cases had columnar metaplasia with intestinal metaplasia; controls had no endoscopic evidence of BE. Histology was reviewed by expert GI pathologists. The SOS device (25 mm, 10 ppi) was swallowed and withdrawn after 6-8 minutes followed by criterion standard endoscopy within 24 hours. DNA was extracted and bisulfite treated. Five MDMs were blindly assayed using the long probe quantitative amplified signal method. The algorithm was set using cross-validated logistic regression. The locked algorithm was applied to assay results from the test set. Results: Baseline characteristics of patients in training and test sets were comparable (Table). The final panel included 3 MDMs (NDRG4, VAV3, and ZNF682) and a reference marker (B3GALT6). Overall sensitivity for BE/EAC detection in the training set was 81% (95% CI 76-88%) at 90% (85-94%) specificity. Overall BE/EAC sensitivity in the test set was 88% (78-94%) at 84% (70-93%) specificity. Sensitivity for HGD and EAC was 100% in the training and test sets. Sensitivity for short segment NDBE in the test set was 63% (38-84%). Areas under the receiver operating characteristic (AUROC) curve for BE/EAC detection were 0.92 (95% CI 0.89-0.95) and 0.94 (0.90-0.98) in the training and test sets, respectively (Figure). The algorithm was not influenced by age, sex, or smoking history. 97% of participants in the training set and 85% in the test set successfully swallowed the SOS device, which was well tolerated and safe. Conclusion: A 3-MDM panel for BE/EAC detection demonstrated excellent sensitivity for high risk BE cases in multi-center case control training and test sets.Figure 1.: Area under the receiver operating characteristics curves for a 3-marker methylated DNA panel assayed from cytology specimen extracted DNA in training and test sets. Table 1. - Baseline characteristics of patients and performance characteristics of SOS test (overall sensitivity and specificity, and stratified by BE dysplasia grade) in training and test sets Variable Training Set (N=198 controls,154 cases) Test Set(N= 44 controls,81 cases) P value (comparing training and test sets) Control Case Control Case Mean (SD) age 55 (13) 65 (10) 52 (15) 65 (11) 0.312 Male Sex (%) 102 (52) 119 (77) 17 (39) 64 (79) 0.992 Mean (SD) BMI 29 (7) 30 (6) 30 (7) 30 (6) 0.283 Ever Smokers % 77 (39) 87 (56) 18 (41) 47 (58) 0.733 Mean *SD) BE length, cm - 4 (3) - 5 (3) 0.070 Long segment BE, N, (%) - 97 (63) - 56 (69) 0.426 Short segment BE, N (%) - 57 (37) - 25 (31) BE dysplasia grade, N (%) EAC - 12 (8) - 2 (2) HGD - 18 (12) - 11 (14) LGD - 7 (4) - 10 (12) IND - 18 (12) - 14 (17) NDBE (long segment) - 57 (37) - 25 (31) NDBE (short segment) - 42 (27) - 19 (24) Training Set% Positive (95% CI) Test Set% Positive (95% CI) Overall 81% (76-88%) 88 (78-94) Dysplasia grade EAC 100 (100-100) 100 (16-100) HGD 100 (100-100) 100 (72-100) LGD 71 (0-100) 90 (55-100) IND 74 (0-100) 93 (66-100%) NDBE (long segment) 91 (73-100) 96 (80-100%) NDBE (short segment) 61 (25-100) 63 (38-84%) Control (No BE) 10 (2-21) 16 (7-30)

S3695 Collagenous Gastritis: A Rare Cause of Severe Iron Deficiency Anemia in a Basic Military Trainee

Introduction: Collagenous gastritis is a rare inflammatory disease characterized by deposition of a thick subepithelial band of collagen and associated inflammatory infiltrate. This disorder has a reported prevalence of 13 per 100,000 esophagogastroduodenoscopies (EGDs) with a female predominance. Two distinct clinical phenotypes have been described: adult-onset characterized by diffuse gastrointestinal involvement and diarrhea, and pediatric-onset manifesting with abdominal pain and anemia. We present an interesting case of pediatric-phenotype collagenous gastritis to highlight this poorly-recognized disorder. Case Description/Methods: A 19-year-old male basic military trainee was referred for evaluation of incidentally discovered iron deficiency anemia. He denied any symptoms, overt blood loss, or NSAIDs. Physical exam was unremarkable. Labs revealed hemoglobin of 9.3 g/dL, iron saturation 7% and ferritin 15 ng/mL. The patient underwent EGD and colonoscopy for further evaluation. EGD revealed a diffusely nodular appearance of the gastric mucosa with areas with suspected atrophy under narrow-band imaging (Figures 1A, B). Colonoscopy was normal. Biopsies returned with a non-specific chronic gastritis without H. pylori. Abdominal computed tomography and additional work up for autoimmune gastritis, fecal H. pylori antigen, syphilis, and heavy metals toxins were all normal. Decision was made to repeat EGD for additional tissue sampling, including cold snared samples. Repeat gastric biopsies with expert GI pathologist review revealed collagenous gastritis, characterized by subepithelial deposition of a thick (greater than 10 µm) collagen band and associated inflammatory infiltrate (Figures 1C, D). Patient was lost to follow up due to disqualification for military service. Discussion: Collagenous gastritis is an exceedingly rare heterogeneous disease process of poorly understood causes and pathogenesis, with autoimmune conditions, medication effects, and infections theorized to be responsible. Numerous treatments have been reported with variable effect, including antisecretory agents, corticosteroids, immunomodulators, and hypoallergenic diets, along with micronutrient supplementation. Due to its rarity, a high level of suspicion is required by gastroenterologists and pathologists based on clinical and endoscopic findings. Our case helps to bring awareness to collagenous gastritis, and emphasizes the potential value for repeat tissue sampling with expert pathologist review.Figure 1.: A: Diffusely nodular appearance of the gastric mucosa with areas. B: Narrow-band imaging reveals areas of suspected atrophy. C: Multiple fragments of gastric mucosa with a prominent lymphoplasmacytic infiltrate within the lamina propria and occasional foci of subepithelial collagen deposition (red arrows). [Hematoxylin & Eosin, 50× magnification]. D: On higher magnification a focus of subepithelial collagen deposition may be seen associated with epithelial injury indicated by loss of foveolar mucin and detachment of the surface epithelium. [Hematoxylin & Eosin, 200× magnification].

Not all low grade dysplasia in Barrett's oesophagus is the same: using specific histological criteria in predicting progression to neoplasia.

Barrett's oesophagus with low grade dysplasia (LGD) is a risk factor for progression to high grade dysplasia (HGD) and oesophageal adenocarcinoma (OAC); however, only a subgroup of LGD will progress. We used a combination of specific histological criteria to identify patients with LGD who are more likely to progress to HGD or OAC. LGD slides from 38 patients within the progressor group (PG) and 17 patients from the non-progressor group (NPG) were obtained and reviewed by two expert GI pathologists, to be stratified by the same four specific histological variables identified by Ten Kate etal.: loss of surface maturation, mucin depletion, nuclear enlargement, and increase of mitosis. After review of LGD slides by two expert GI pathologists, 27 suitable patients were identified. Of these 27 patients there was a higher proportion of patients from the PG with all four specific criteria reported, compared to the NPG: 14 (78%) vs 3 (33%) p=0.0394. Patients with all four specific criteria were more likely to progress compared to those who had one or less specific criteria reported (OR 7, 95% CI 1.1848-41.3585, p=0.032). A combination of ≥2 or ≥3 specific histological criteria was not prognostic. Patients with a combination of all four specific histological criteria (loss of surface maturation, mucin depletion, nuclear enlargement, and increase of mitosis) were associated with greater progression from LGD to HGD or OAC in Barrett's oesophagus.

P584 Mucosal Eosinophilia is an Independent Predictor of Vedolizumab Effectiveness in Ulcerative Colitis

Abstract Background Ulcerative Colitis (UC) is characterized by intestinal inflammation with chronic inflammatory features. Along with the typical neutrophil-predominant inflammation, there has been growing understanding of the role of eosinophils in UC and the potential therapeutic benefit of Vedolizumab (VDZ) in those cases. VDZ blocks gastrointestinal homing of T lymphocytes and eosinophil recruitment to the intestinal mucosa via blocking α4β7 which plays a major role in eosinophils localization to the intestinal mucosa. We aimed to assess whether mucosal eosinophilia can predict response to VDZ. Methods A retrospective study in a single IBD center. Active UC patients who were initiated on VDZ were identified. All patients had at least one documented encounter of VDZ infusion, colonoscopy prior to initiation of VDZ and afterwards with mucosal biopsies at both time points. The endoscopic activity was scored using Mayo endoscopic subscore (MES) with remission as MES 0 or 1. Clinical response was assessed using Physician’s Global Assessment (PGA) as no response, partial or complete clinical response and clinical remission. Histology was reported by an expert GI pathologist and the response was assessed as none, partial or complete. Mucosal eosinophilia was graded at initial and follow-up biopsies as not conspicuous vs. conspicuous with predominance of eosinophils (>100 eosinophils in at least 2 hpf). Eosinophil activity defined as presence of at least 1 of the 3 histologic features: eosinophilic microabscesses, eosinophilic cryptitis, or degranulated eosinophils Results 107 UC patients were included in the study with a mean age of 44.9 years, majority were males (N = 61, 55.5%), Caucasians (N = 96, 87.3%) and with pancolonic location (N = 100, 90.9%). Mean disease duration was 11.3 yrs (SD, 8.6). 55.5% of patients had prior aTNF exposure. Clinical remission was observed in (N = 63, 57.3%) and endoscopic remission (N = 73, 66.4%) after initiating VDZ. Complete histological response was observed in (N=50, 46.7%). Baseline characteristics based on endoscopic remission is shown where male sex, disease location and the proportion of patients with conspicuous eosinophils with abnormal activity in mucosal biopsies as well as no improvement in eosinophil activity compared to pre-treatment biopsies is greater among those not in endoscopic remission. In a stepwise backward logistic regression model including sex and disease location, improvement in abnormal eosinophilic activity was independently associated with increased odds of endoscopic remission after initiating vedolizumab (OR 4.3; CI, 1.1 - 16.4, p = 0.03). Conclusion In ulcerative colitis, change in mucosal eosinophilia after initiating VDZ was an independent predictor of endoscopic remission with VDZ.

ID: 3524488 UTILIZATION OF A DEEP LEARNING ARTIFICIAL INTELLIGENCE MODEL IN THE HISTOLOGIC DIAGNOSIS OF DYSPLASIA IN BARRETT’S ESOPHAGUS

The risk of malignant transformation to esophageal adenocarcinoma (EAC) in Barrett’s esophagus (BE) increases with development of dysplasia. However, the histologic diagnosis of dysplasia, particularly low-grade dysplasia (LGD), is challenging leading to lack of interobserver agreement amongst pathologists with upto 85% of all community diagnosed LGD being downstaged upon expert GI pathologist review. This may result in unnecessary therapeutic procedures with associated adverse effects. Hence, there is a critical need to improve diagnostic capabilities to identify dysplastic BE. We aimed to leverage artificial intelligence techniques and utilize a deep learning model to diagnose dysplasia in BE histologic slides.

The Utility of P53 Immunohistochemistry in the Diagnosis of Barrett's Esophagus with Indefinite for Dysplasia

Background: Barrett's esophagus 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: Archive endoscopic biopsies with a BE-IND diagnosis from two academic centers were analyzed. Firstly, hematoxylin and eosin-stained slides (H&E) were reviewed by four expert GI pathologists allocated into two groups (A and B). After a wash-out period of at least eight weeks, H&E slides were re-assessed 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. Findings: We included 216 BE-IND specimens from 185 patients, of which 44.0% and 32.9% were confirmed after H&E slide revision by Groups A and B, respectively. Over half of the cases were downgraded to a non-dysplastic BE (NDBE), while 5.6% of cases in Group A and 7.4% in Group B were upgraded to 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<.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%CI:18.8-113.0). Interpretation: P53-IHC reduces the rate of BE-IND diagnoses and improves the IOA among pathologists when reporting BE with equivocal epithelial changes. Funding: This study received infrastructure support from the NIHR Cambridge Biomedical Research Centre. Declaration of Interest: None to declare. Ethical Approval: An ethics committee approval for this study was granted at both institutions (Cambridge University: 7 LREC01/149; Mayo Clinic: 9-000514).

S0423 “WATS” the Incremental Yield of Wide-Area Transepithelial Sampling for Dysplasia Detection in Barrett’s Esophagus? A Systematic Review and Meta-Analysis

INTRODUCTION: Seattle protocol forceps biopsies (FB) are currently recommended for surveillance in Barrett’s esophagus (BE). Dysplasia detection is limited by sampling error. Wide-area transepithelial sampling (WATS) utilizes computer-aided diagnosis of esophageal brushings to increase dysplasia detection. We assessed the incremental yield of WATS compared to FB and limitations of published data in a systematic review and meta-analysis. METHODS: We queried multiple scientific databases for studies utilizing WATS with FB from 2000–2020. We utilized the arcsine transformation with a fixed-effects meta-analysis of single arm proportions for analysis given the small number of studies. Primary outcome was the incremental yield of WATS over FB for dysplastic BE (IND/LGD + HGD + EAC) as a composite. Secondary outcomes were incremental yield of WATS for HGD/EAC, and the rate of reconfirmation of incremental WATS dysplasia on subsequent FB. RESULTS: We included 6 full-length manuscripts (from 2011 to 2020), studying 37,468 patients (6 included data for dysplastic BE, and 4 for HGD/EAC), Study characteristics are shown in the Table. All WATS samples were read by pathologists at CDx Diagnostics (Suffern, NY). BE histology was confirmed by 2 expert GI pathologists in only 2 (33%) studies. Meta-analysis of 6 studies demonstrated that FB diagnosed dysplasia in 5.8% (95% CI: 5.1%–6.6%) of cases, while WATS added an incremental yield of 4.5% (95% CI: 3.9%–5.2%; Figure 1). Meta-analysis of 4 studies demonstrated that FB diagnosed HGD/EAC in 0.6% (95% CI 0.4%–0.9%) of patients, while WATS added an incremental yield of 0.4% (95% CI: 0.2%–0.6%; Figure 2). Notably, WATS was negative in 1%–45% of cases where FB identified dysplasia. Only 1 study reported reconfirmation of WATS dysplasia on subsequent endoscopy: only 7 of 23 (30%) WATS dysplasia (+) patients had dysplasia on subsequent FB. In 3 studies the incremental yield of WATS dysplasia was reported in several cases with FB histology negative for intestinal metaplasia (85% of cases). No endoscopic data was available in these cases. CONCLUSION: WATS appears to increase the yield of dysplastic BE and HGD/EAC as an adjunct to FB. However, in most studies, this incremental yield is not reconfirmed by FB, WATS reads are done only by CDx pathologists and FB dysplasia is not confirmed by expert GI pathologists. Long-term follow up is required to ascertain dysplasia outcomes in patients with dysplasia based solely on WATS.Figure 1Figure 2Table 1

S2919 Atypical Presentation odslrf Cryptosporidiosis in an Immunocompetent Female

INTRODUCTION: Cryptosporidiosis is an opportunistic diarrheal infection caused by Cryptosporidium, an intracellular protozoan, prevalent mostly in immunocompromised individuals. We report a rare case of extensive cryptosporidiosis of the upper gastrointestinal (UGI) tract in an immunocompetent individual. CASE DESCRIPTION/METHODS: A 77 year-old female with history of cholecystectomy presented to the clinic with 2-months history of nausea and chronic epigastric pain. Her symptoms were aggravated by spicy foods and not relieved with omeprazole and sucralfate. She did not have any diarrhea, fever or chills. She was a retired nurse and denied any sick contacts, contact with farm animals or a travel history. She denied any animal bites. She consumed only bottled water for drinking. Her workup was unremarkable for anemia or leukocytosis and normal liver enzymes. She had mild delayed gastric emptying and a positive glucose breath test for small bowel bacterial overgrowth (SIBO). She received Reglan and Doxycycline with no significant improvement. Her esophagogastroduodenoscopy (EGD) showed esophagitis and antral erosions. Biopsy report by local pathologist showed esophagitis, chronic inactive gastritis and duodenitis. A secondary review by our GI pathologist confirmed prior findings but also reported cryptosporidium in the esophageal, gastric and duodenal biospies. Giemsa stain was negative for helicobacter pylori. Her HIV antibody was negative and IgA levels were normal. She was treated with a 3-day course of Nitazoxanide 500 mg twice daily with improvement in her symptoms. A repeat EGD after 4 months showed no inflammation and absence of cryptosporidium suggesting successful eradication. DISCUSSION: Cryptosporidiosis presents commonly as secretory diarrhea, malaise and cramping abdominal pain with or without biliary involvement (10–30%). Asymptomatic cases have been reported in immunocompetent hosts especially in children. Fecal leukocytes are rare unless co-infected with another enteric pathogen. We report an atypical presentation of non-diarrheal cryptosporidiosis associated with biopsy-proven inflammation in UGI tract that responded to Nitazoxanide. The patient likely contracted the parasite from an untraceable contaminated water source. Clinically, a high level of suspicion should be exercised in patients with refractory UGI symptoms to look for infectious causes. It is also important to consider an expert GI pathologist’s review of the histopathology in cases with an unclear diagnosis.Figure 1.: A. Low power photomicrograph of the stomach demonstrates only mild reactive change.Figure 2.: B. At high power, numerous organisms similar to those seen in B are noted in adherent mucus.Figure 3.: C. High power image of duodenum highlights numerous small (1-2 micron) basophilic “beads” seemingly adherent to the brush border.

Tumor size and the risk for lymph node metastases in T1 esophageal adenocarcinoma.

332 Background: Adjuvant therapy after endoscopic resection (ER) of T1 EAC in non-surgical candidates is largely based on the risk of LNM. Risk factors for LNM in T1 EAC are not clearly defined. Our aim is to evaluate risk factors for LNM in T1 EAC patients following esophagectomy or ER with ≥ 5 years of follow-up. Methods: This is a retrospective analysis at a large tertiary referral center. Our pathology database identified patients who underwent esophagectomy or ER with ≥ 5 years follow-up, with histologically proven T1 EAC from 2010-2017. Patients were excluded if they (a) received chemoradiation prior to esophagectomy or before/after ER (b) had any other primary cancer treated within the previous 5 years. Specimens were reviewed by an expert GI pathologist for accuracy. Results: Of 80 patients [85% males], 61 (76%) underwent esophagectomy and 19 (24%) underwent ER. Twelve (15%) developed LNM per study criteria. Tumor size was significantly (p-value 0.014) associated with risk of LNM (Table). No other factors including lymphovascular invasion, differentiation on pathology, macroscopic appearance, infiltration growth pattern, or tumor distance from the gastroesophageal junction were significant risk factors for LNM. Conclusions: In T1 EAC, tumor size appears to be a significant risk factor for LNM at five years following surgical or endoscopic resection. Adjuvant therapy should be considered in patients with larger tumor size. [Table: see text]

2932 Histological Diagnosis of Pancreatic Serous Cystadenomas Using a 22G Core Biopsy Needle: A Case Series

INTRODUCTION: Pancreatic Serous Cystadenoma or Pancreatic Cystic Serous Neoplasm (SCN) are a type of benign pancreatic tumor that account for 1-2% of all pancreatic masses. They are most often benign, with very rare malignant potential. SCN's will often have a cystic or honeycomb appearance on CT scan. In the past, the mainstay of diagnosing SCN has been via endoscopy ultrasound (EUS) with fine needle aspiration (FNA) to assess fluid for CEA levels and also for cytological analysis. However fluid can be difficult to obtain and often there is not enough cellularity for cytology. In this case series of 7 patients we performed fine needle biopsy (FNB) of the cell walls of pancreatic cysts as a technique for histological review. METHODS: We retrospectively reviewed medical records of patients who underwent EUS for FNA for fluid sampling of pancreatic cysts but due to microcystic nature a decision was made to proceed with FNB to obtain tissue for histology. FNB was done using 22 G FNB needle with slow stylet pull technique. The tissue was placed in formalin for histological analysis by an expert GI pathologist. RESULTS: A total of seven patients were evaluated. The mean age was 75 years old and 5 of the patients were female. All patients had one cyst that was sampled. The location of the cyst was on the head in 1 of the patients and body or tail of the other 6 patients. The mean cyst size was 31.1 mm (range 14-40 mm). There were no post procedure complications. All seven samples were adequate for histological analysis. Six out of the seven biopsy samples were consistent with serous cystadenomas. One samples showed benign pancreatic acinar tissue and stroma without malignancy. CONCLUSION: Management of pancreatic SCNs can be challenging as no follow up is recommended if the diagnosis is confirmed but lots of times a definite diagnosis cannot be made based on endosonographic and CT/MRI findings. These patients can be potentially subject to repeat imaging for surveillance. Here we show that pursuing FNB instead of FNA for microcystic lesions can yield enough tissue for histological diagnosis of SCNs and can alter management and avoid further imaging.

Persistent confirmed low-grade dysplasia in Barrett's esophagus is a risk factor for progression to high-grade dysplasia and adenocarcinoma in a US Veterans cohort.

Patients with Barrett's esophagus (BE) and low-grade dysplasia (LGD) are at increased risk of esophageal adenocarcinoma (EAC), although many regress to nondysplastic BE. This has significant clinical importance for patients being considered for endoscopic eradication therapy. Our aim is to determine the risk for progression in patients with confirmed persistent LGD. We performed a single-center retrospective cohort study of patients with BE and confirmed LGD between 2006 and 2016. Confirmed LGD was defined as LGD diagnosed by consensus conference with an expert GI pathologist or review by an expert GI pathologist and persistence as LGD present on subsequent endoscopic biopsy. The primary outcome was the incidence rate of HGD (high-grade dysplasia)/EAC. Secondary outcomes included risk factors for dysplastic progression. Risk factors for progression were assessed using univariate and multivariate analysis with logistic regression. Of 69 patients (mean age 65.2 years) with confirmed LGD were included. In total, 16 of 69 patients (23.2%) with LGD developed HGD/EAC during a median follow-up of 3.74 years (IQR, 1.24-5.45). For persistent confirmed LGD, the rate was 6.44 (95% confidence interval (CI), 2.61-13.40) compared to 2.61 cases per 100 patient-years (95% CI, 0.83-6.30) for nonpersistent LGD. Persistent LGD was found in only 29% of patients. Persistent LGD was an independent risk factor for the development of HGD/EAC (OR 4.18; [95% CI, 1.03-17.1]). Persistent confirmed LGD, present in only 1/3 of patients, was an independent risk factor for the development of HGD/EAC. Persistence LGD may be useful in decision making regarding the management of BE.

Endoscopic submucosal dissection using a novel endoscopic articulating knife for clinical staging of early esophageal neoplasia.

114 Background: Clinical staging of early esophageal neoplasia traditionally involves histological confirmation and imaging with endoscopic ultrasound (EUS), CT, and PET which have low sensitivity and specificity for staging esophageal cancer (EC). Endoscopic submucosal dissection (ESD) therapy is traditionally used for treatment, but not diagnosis as it is felt to be technically challenging and have a high risk for complications. We applied a new articulating endoscopic knife that permits safe ESD (ESD-CC) to evaluate early neoplasia with potential curative resection. Methods: We performed a retrospective study of patients undergoing ESD to stage or treat suspected early EC (cT1-T2). Clinical stage was done by EUS, CT, and PET. Two expert GI pathologists reviewed all histology. Lesions were examined with high resolution white light endoscopy and narrow band imaging. ESD was done with 1:200,000 epinephrine and methylene blue dye injection for lifting and staining the submucosal space. A 5mm, scissors-like articulated knife was used to perform ESD and hemostasis. Complications during post-ESD observation or follow-up were recorded. Results: A total of 35 patients who underwent ESD-CC were included with median age 70 (IQR 12), 26 males (74%), and followed for a median 3.4 months (IQR 6.4). This group consisted of 32 potential adenocarcinomas and 3 squamous cell cancers. The clinical Pre-ESD diagnoses were cT1 EC (24, 69%) and suspected EC in Barrett’s esophagus (BE) (11, 31%). The cT1 EC cases had ESD staged at least T1b (5, 21%), T1a (11, 46%), EC in situ (1, 4%), and dysplastic BE (7, 29%). The suspected EC cases had ESD staged at least T1b (1, 9%), T1a (2, 18%), and DBE (8, 73%). ESD-CC up-staged 4 (11%), down-staged 10 (29%), and confirmed prior diagnosis in 21 (60%). No complications including bleeding, perforation, or stricture formation regardless of size of ESD, age of patient, or co-morbidities. Conclusions: Staging of early esophageal cancer can be improved using ESD with an articulating knife, without increase in complications. ESD may be used as a staging modality in early esophageal cancer.

Lessons Learned From Introduction of Chromo-Colonoscopy into an Academic GI Practice: A Review of the Early Experience

Introduction: Chromocolonoscopy has become an increasingly recognized method to detect dysplasia in patients with inflammatory bowel disease (IBD). IBD patients with risk factors for dysplasia are offered chromocolonoscopy as part of their screening/surveillance colonoscopy. A retrospective review of our initial experience with a case control comparison of procedure performance was conducted.2790 Figure 1 No Caption available.Methods: A retrospective study was completed at Mayo Clinic in Arizona comparing a cohort of IBD patients undergoing chromocolonoscopy to historical controls with IBD undergoing screening, surveillance or dysplasia evaluation by the same 5 expert colonoscopists. We had 37 cases from 3/2015-6/2017 and 134 controls from 1/2012-2/2015. Of the 37 cases, 10 were for Lynch Syndrome or colorectal neoplasia alone, leaving 27 IBD cases for analysis. Indications in the 27 IBD patients were: known dysplasia evaluation, screening or surveillance in high risk patients. Pancolonic application of indigo carmine, 2g/liter, was used. All dysplasia was confirmed by expert GI pathologists. Results: The results show that the mean withdrawal times were 29.6 min. (SE 8.44 min) vs 19.2 min (SE 8.79 min), P The yield of targeted versus random biopsies for dysplasia and adenomatous polyps were compared in patients who underwent chromocolonoscopy (Table1). Three patients had dysplasia detected on random biopsies. Conclusion: This data suggests that targeted biopsies are valuable, but do not find all instances of dysplasia and may not be a replacement for random biopsies at this time. Until further refinements in the procedure are developed, we recommend taking random biopsies in addition to targeted biopsies in patients getting chromocolonoscopy. In addition, chromocolonoscopy in IBD takes longer than routine colonoscopy and this should be considered when scheduling these procedures.

Does Multifocal Dysplasia in Barrett’s Esophagus Increase the Risk of Progression to Esophageal Adenocarcinoma?: Presidential Poster Award

Introduction: Barrett’s Esophagus (BE), is thought to be secondary to gastroesophageal reflux disease and progresses to esophageal adenocarcinoma (EAC) in a stepwise manner. Current guidelines utilize dysplasia to guide management, but there is conflicting evidence on the extent of dysplasia for risk stratification. Here, we evaluated the effect of multifocal dysplasia on the risk for EAC and hypothesized that multifocality in low-grade dysplasia (LGD) and high-grade dysplasia (HGD) are risk factors for progression to EAC. Methods: This was a retrospective study of BE patients with, or with suspected, dysplasia presenting to an academic medical center for evaluation from 2006 to 2016. Information collected include: patient demographics; BE length; histology, including grade and presence of multifocal dysplasia; disease progression; PPI use; need for esophagectomy; and length of follow-up. All pathology specimens were read by expert GI pathologists as: no dysplasia (ND), indefinite for dysplasia (IND), LGD, or HGD. LGD or HGD was further categorized as unifocal or multifocal. Results: Of 305 BE patients with, or with suspected, dysplasia, 51 were excluded (33 ND, 18 IND). 254 BE patients were included in the final analysis: 112 LGD (unifocal=90, multifocal=22) and 142 HGD (unifocal=108, multifocal=34). In total, twenty-five patients with dysplastic BE (13 unifocal, 12 multifocal) developed EAC; all were diagnosed with intramucosal adenocarcinoma with either T1a or T1b lesions. Multifocality significantly increased the risk for progression to EAC (Table 2: RR 3.26, p=0.004), with the effect most pronounced in the subgroup with multifocal LGD as compared to those with unifocal LGD (RR 8.18, p=0.013). In patients with HGD, multifocality did not increase the risk for subsequent EAC (RR 2.31, p=0.079). Lastly, multifocal dysplasia associated with long segment BE (LSBE) significantly increased the risk for EAC (Table 3: RR 3.09, p=0.011), and HGD patients with LSBE were more at-risk for progression to EAC than that of LGD patients (RR 4.33, p=0.007). Conclusion: Multifocal dysplasia significantly increases the risk of progressing from BE to EAC and is especially important in BE patients with LGD. Identification of the extent of dysplasia is also essential in patients with LSBE, particularly those diagnosed with HGD. If confirmed, incorporating dysplasia multifocality into the BE management algorithm should be considered.335_A Figure 1 No Caption available.335_B Figure 2 No Caption available.335_C Figure 3. Short Segment (SS) BE is defined as <3 cm in length, while Long Segment (LS) BE is defined as > or = to 3 cm in length

An Evaluation of the Quality of Esophageal Biopsies in Patients with Eosinophilic Esophagitis

Introduction: Eosinophilic esophagitis (EoE) is a clinicopathologic diagnosis based on the presence of ≥15 eosinophils per high power field (hpf) in the epithelium. Additional changes include microabscesses, lamina propria (LP) papillae elongation, basal zone hyperplasia, intercellular edema, and LP fibrosis. Recent studies have shown the importance of subepithelial fibrosis as this correlates with clinical and endoscopic findings of EoE, disease duration and response to therapy. We sought to determine the adequacy of esophageal biopsies, defined as the presence of LP, among patients with EoE. Methods: This was a single-center, retrospective, observational study at a tertiary referral center. The pathology database was interrogated to identify pathology records from 2014-2015 that contained ≥15 eosinophils/hpf. All records were reviewed by one expert GI pathologist for the presence of lamina propria. We investigated factors that could influence acquisition of deeper esophageal biopsies including type of sedation, trainee involvement, procedure duration and seniority of attending endoscopist. Results: We reviewed 119 pathology records that were contributed by 109 patients. Within the pathology records, 244 esophageal biopsies were reviewed and classified by location (proximal, distal, and other; note that a pathology record could contribute multiple location biopsies). Only 98 biopsies had the presence of LP (40%). Breakdown according to location: proximal biopsies 25/66 (38%); distal biopsies 34/90 (38%); and other biopsies 39/88 (44%) had the presence of LP. Univariable generalized linear mixed models were used to estimate the odds of a higher number of LP biopsies. Patient demographics, trainee involvement, seniority of attending endoscopist, sedation used, and procedure time were not associated with odds of a higher number of LP biopsies. When analyzed by location of biopsies, senior attending physicians (defined as > 3 years in practice) were 3.74 (95% CI: 1.45 - 9.65) times as likely to obtain a positive distal biopsy with LP when compared to junior attendings (p = 0.01). Conclusion: The majority of biopsies did not contain LP. Senior attendings were more likely to obtain quality distal biopsies. Duration of procedure, sedation type, and trainee involvement did not affect depth of biopsy. As esophageal biopsy remains the gold standard for monitoring disease activity, quality of mucosal sampling is essential to providing accurate information to guide clinical management.395_A Figure 1. Odds of a higher count of lamina propria biopsies395_B Figure 2. Odds of a positive LP on distal biopsy

Expression of MUC-2 and CDX-2 in Mucosal Brushings Accurately Predicts Barrettʼs Esophagus (BE): 1 Step Closer to a Non-Invasive BE Screening Test

Introduction: Purpose: To evaluate MUC-2 and CDX-2 expression as biomarkers for Barrett's Esophagus (BE) using esophageal brushing and quantitative reverse transcriptase polymerase chain reaction (RT-qPCR). Background: Esophageal adenocarcinoma (EAC) starts as BE which is a direct consequence of gastroesophageal reflux (GERD). EAC usually presents with advanced disease in the United States and carries a poor prognosis. More than half of EAC patients report minimal or no GERD symptoms prior to presenting with EAC. If dysplastic BE or early EAC are detected, treatments can cure the disease. Sensitive and specific biomarkers for BE are needed to noninvasively detect dysplastic BE and early EAC in this population. Mucin 2(MUC-2) is secreted by intestinal goblet cells and CDX-2 is a caudal-type homeobox protein specific for intestinal epithelial cells. Both proteins are associated with esophageal intestinal metaplasia (BE). Methods: 40 patients had upper endoscopy (EGD) performed with esophageal brushing, 37 men and 7 women with a mean age of 62.5 years. Brushed samples were assayed for messenger RNA (mRNA) expression of MUC-2 and CDX-2 by RT-qPCR. A house keeping gene, beta-actin served as the reference gene. Delta delta cycle threshold (delta delta Ct) values were compared to histology read by a single expert GI pathologist blinded to the RT-qPCR data (gold standard). All patients had a prior diagnosis of BE or high-risk for BE. Histology demonstrated 17 patients with BE and 23 without. Delta delta Ct values were compared between the two groups. A lower delta delta Ct correlates with higher mRNA expression. Results: Delta delta Ct values for MUC-2 and CDX-2 expression were significantly lower for BE patients compared to those patients without BE (P value<0.001). The Receiver Operating Curve (ROC) for MUC-2 and CDX-2 measurements showed an area under the curve (AUC; % test accuracy) of 0.982 (98.2%) and 0.881 (88.1%),respectively. Conclusion: mRNA expression of MUC-2 and CDX-2 by RT-qPCR are promising biomarkers with high accuracy for identifying BE. Future studies are needed to evaluate these biomarkers using noninvasive cell retrieval methods (Cytosponge and balloon capsule).

Ex vivo experimental study on the Thulium laser system: new horizons for interventional endoscopy (with videos).

Background and study aims The Thulium laser system (TLS) is an emerging interventional tool adopted in many surgical specialties. Its 2.0-μm wavelength allows precise coagulation (0.2 – 0.4 mm in depth) and cutting, limiting the possibilities of collateral injuries. We tested the impact of the TLS for gastric endoscopic submucosal dissection (ESD) and per oral endoscopic myotomy (POEM) ex vivo in pigs. Materials and methods Ex vivo porcine stomach and esophagus models underwent 2 POEMs, and 3 ESDs (mean diameter 3.5 cm) with TLS using a 272-µm and a 365-µm thick optical fibers. Both continuous and pulsed laser emission were evaluated. Subsequent histopathological analysis was performed by an expert GI pathologist on the whole porcine models. Results Complete POEMs and gastric ESDs were successfully performed in all cases in 30 to 70 and 15 to 20 minutes. Both optical fibers were equally effective and precise. The best power output for mucosal incision was 25 to 30 W during ESD and 25 W for POEM using continuous laser emission. During submucosal dissection and tunneling the favorite power output was 20 W and 15 to 20 W, respectively, operating in continuous mode. No transmural perforation occurred throughout the operations and histopathology confirmed the absence of accidental muscular layer damage. Conclusions The TLS stands out as a precise and manageable instrument in ex vivo models. This technique appears to be a promising tool for advanced interventional endoscopy.