2025+AGA临床实践指南：Barrett's食管的监测

AGA Clinical Practice Guideline on Surveillance of Barrett's Esophagus

Sachin Wani,1,* Margaret J. Zhou,2,* Tarek Sawas,3,* Joel H. Rubenstein,4,5,6,*

Swathi Eluri,7 David A. Leiman,8 Shahnaz Sultan,9,10 Siddharth Singh,11 John Inadomi,12 Aaron P. Thrift,13 David A. Katzka,14 and Perica Davitkov15,16

$^{1}$ Division of Gastroenterology and Hepatology, University of Colorado Anschutz Medical Campus, Aurora, Colorado;$^{2}$ Division of Gastroenterology and Hepatology, Department of Medicine, Stanford University School of Medicine, Stanford,California; $^{3}$ Division of Digestive and Liver Disease, University of Texas Southwestern, Dallas, Texas; $^{4}$ Center for Clinical Management Research, LTC Charles S. Kettles Veterans Affairs Medical Center, Ann Arbor, Michigan; $^{5}$ Barrett's Esophagus Program, Division of Gastroenterology, University of Michigan Medical School, Ann Arbor, Michigan; $^{6}$ Cancer Control and Population Sciences Program, Rogel Cancer Center, University of Michigan Medical School, Ann Arbor, Michigan; $^{7}$ Division of Gastroenterology and Hepatology, Mayo Clinic Florida, Jacksonville, Florida; $^{8}$ Division of Gastroenterology and Duke Clinical Research Unit, Duke University School of Medicine, Durham, North Carolina; $^{9}$ Division of Gastroenterology,Hepatology, and Nutrition, University of Minnesota, Minneapolis, Minnesota, 10° Veterans Affairs Healthcare System,Minneapolis,Minnesota; 11'Division of Gastroenterology, University of California San Diego, La Jolla, California;$^{12}$ Department of Internal Medicine, The University of Utah School of Medicine, Salt Lake City, Utah; $^{13}$ Section of Epidemiology and Population Sciences, Department of Medicine, Baylor College of Medicine, Houston, Texas; $^{14}$ Division of Gastroenterology and Hepatology, Columbia University, New York, New York; $^{15}$ Department of Medicine, Case Western Reserve University, Cleveland, Ohio; and $^{16}$ Division of Gastroenterology, Veterans Affairs Northeast Ohio Healthcare System, Cleveland, Ohio

BACKGROUND & AIMS: Barrett's esophagus (BE) is the only identifiable precursor to esophageal adenocarcinoma(EAC). Endoscopic surveillance has been proposed for early detection of BE-related neoplasia and reducing EAC mor-tality. This clinical practice guideline aims to inform clini-cians and patients by providing evidence-based practice recommendations for surveillance in patients with BE.METHODS: The Grading of Recommendations Assessment,Development and Evaluation framework was used to assess evidence and make recommendations. The panel prioritized clinical questions and outcomes according to their impor-tance for clinicians and patients, conducted an evidence review, and used the Evidence-to-Decision Framework to develop recommendations regarding the role of endoscopic surveillance in patients with BE. The clinical domains addressed included:(1) overall role of endoscopic surveillance,(2) surveillance in patients with columnar-lined esophagus <1 cm, (3) optimal imaging modalities, (4)adjunctive sampling techniques, (5) the utility of bio-markers in risk-stratification, (6) chemopreventive strate-gies, and (7) antireflux procedures in the prevention of progression in patients with BE. Clinical recommendations were based on the balance between the desirable and un-desirable effects, patient values, costs, and health equity considerations. RESULTS: The panel agreed on 8 recommen-dations. Based on the available evidence, the panel provided a conditional recommendation in favor of surveillance for

patients with nondysplastic BE. In patients with columnar-lined esophagus <1 cm , a conditional recommendation was made against endoscopic surveillance. The panel made a strong recommendation in favor of a combination of high-definition white light endoscopy and chromoendoscopy compared with white light endoscopy alone. The panel made no recommen-dation on the use of enhanced sampling techniques, such as wide-area transepithelial sampling to enhance neoplasia detection and biomarkers such as p53 and TissueCypher to predict progression in BE. The panel provided a conditional recommendation for the use of daily proton pump inhibitor therapy compared with no therapy and compared with antire-flux surgery to prevent progression in BE. CONCLUSIONS: This document provides a comprehensive outline on the role of surveillance in patients with BE. Key implementation state-ments included in this document stress the importance of a high-quality endoscopy examination, sampling using a structured biopsy protocol, and confirming the diagnosis of BE-related neoplasia by an expert pathologist. This document also provides guidance on surveillance intervals and man-agement of patients with BE-related low-grade dysplasia and indefinite for dysplasia. Providers should engage in shared decision making based on patient preferences. Limitations and gaps in the evidence are highlighted to guide future research opportunities. GUIDELINE ENDORSEMENT:This guideline is endorsed by the Canadian Association of Gastroenterology.

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1185

Keywords: Barrett's Esophagus; Esophageal Adenocarcinoma;Surveillance; Chromoendoscopy; Biomarkers; Chemopreven-tion; Proton Pump Inhibitor; Antireflux Surgery.

Executive Summary

Esophageal adenocarcinoma (EAC) is a lethal cancer with increasing incidence and sobering mortality rates over the past several decades. ${ }^{1}$ The incidence of EAC rose 5-fold from the 1970s to the 2010s, and adenocarcinoma now represents the most common form of esophageal cancer in Western populations, with a dismal overall 5-year survival rate of approximately 20%. ${ }^{2}$ Survival from all but the earliest stage of EAC remains poor. ${ }^{3}$ Barrett's esophagus(BE) is the only identifiable precursor lesion for EAC and characterized by replacement of squamous epithelium in the distal esophagus by metaplastic specialized columnar epithelium in response to chronic gastroesophageal reflux-induced injury and inflammation. ${ }^{4,5}$ Recent evidence sug-gests that most cases of EAC arise from BE or from intestinalized metaplastic precursor cells in the gastric cardia. ${ }^{4,6}$ BE is believed to progress to EAC in a stepwise and probabilistic fashion through steps of low-grade dysplasia (LGD), then high-grade dysplasia (HGD), before developing into invasive EAC. Given that the prognosis of EAC is strongly related to stage at diagnosis, endoscopic surveillance in patients with BE has been proposed with the goal of early detection of BE-related neoplasia. These patients can then be referred for endoscopic eradication therapy (EET), reducing the morbidity and mortality related to esophagectomy and ultimately preventing EAC mortality. ${ }^{7-10}$

# Objective of the Review and Guideline

The American Gastroenterological Association (AGA)developed this clinical practice guideline to provide evidence-based recommendations for endoscopic surveil-lance in patients with BE. Importantly, the role of EET was addressed in a recent AGA guideline ${ }^{10}$ and screening for BE and EAC will be addressed in a future guideline document.This document addresses the overall role and effectiveness of endoscopic surveillance, endoscopic surveillance in pa-tients with columnar-lined esophagus <1 cm, optimal im-aging strategies, adjunctive sampling techniques to improve detection, biomarkers and other risk stratification strategies to guide surveillance, chemopremvention and antireflux procedures for prevention of progression in pa-tients with BE.

Target Audience

The target audience for these guidelines includes pri-mary care, internal medicine, family medicine, gastroen-terology, oncology, and surgery health care providers;patients; and policy makers. These recommendations are not intended to be used as the standard of care, but as a guide to facilitate management of patients with BE and

related neoplasia. ${ }^{10}$ Each recommendation in this guideline is accompanied by implementation considerations or qualifying remarks that should be considered an integral part of the recommendation statement and should not be omitted. Although no single recommendation can encom-pass every individual circumstance and context, it can be used to address the benefits and harms of treatments and support the processes of shared decision making based on patients' values and preferences.

Methods

Overview

This document represents the official recommendations of the AGA. These recommendations were developed using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) framework for diagnostic tests and stra-tegies and adheres to best practices in guideline development,as outlined by the National Academy of Medicine (Supplementary Tables 1 and 2). ${ }^{11}$

# Organization and Panel Composition

The guideline panel included 14 members with clinical expertise in BE and methodologic expertise (1 senior and 3 junior guideline methodologists). Members of the AGA guide-line committee helped review all the synthesized evidence,contributed to discussion, and helped develop the clinical de-cision support tool. A librarian assisted with designing and executing the relevant literature searches.

# Management of Conflict of Interest and Guideline Funding

Panel members disclosed all potential conflicts of interest.Conflicts were managed according to AGA policies, the Na-tional Academy of Medicine, and Guidelines International Network standards. ${ }^{12-14}$ Development of this guideline was wholly funded by the AGA Institute with no support from the industry. A full list of conflicts can be accessed at the AGA's National Office in Bethesda, MD.

* Authors share co-first authorship.

Abbreviations used in this paper: AGA, American Gastroenterological Association; Al, artificial intelligence; BE, Barrett's esophagus; CD, crypt dysplasia; CE, chromendoscopy; DDR, dysplasia detection rate; EAC,esophageal adenocarcinoma; EGG, esophagogastroudodenoscopy; EtD,Evidence to Decision; EET, endoscopic eradication therapy; FN, false negative; FP, false positive; GERD, gastroesophageal reflux disease;GRADE; Grading of Recommendations Assessment, Development, and Evaluation; HGD, high-grade dysplasia; HR, hazard ratio; IHC, immuno-histochemistry; IND, indefinite for dysplasia; LGD, low-grade dysplasia;MA, meta-analysis; NBI, narrow-band imaging; NDBE, nondysplastic Barrett's esophagus; NDR, neoplasia detection rate; NPV, negative pre-dictive value; OR, odds ratio; PEEC, post-endoscopy esophageal adenocarcinoma; PEEN, post-endoscopy esophageal neoplasia; PICO.Population, Intervention, Comparison, Outcome; PPI, proton pump in-hibitor; PPV, positive predictive value; RCT, randomized controlled trial;RR, relative risk; SAE, serious adverse events; SR, systematic review; TN,true negative; TP, true positive; VCE, virtual chromendoscopy; WATS-3D, wide-area transepithelial sampling; WLE, white light endoscopy.

 Most current article

©2025 by the AGA Institute.

https://doi.org/10.1053/j.gastro.2025.09.012

1186 Wani et al

Gastroenterology Vol.169, Iss. 6

Scope

The guideline panel and evidence review team formulated clinically relevant questions on endoscopic surveillance stra-tegies and therapies for BE and related neoplasia. The most recent comprehensive position paper by the AGA on BE was published in 2011, and included guidance on screening, sur-veillance, biomarkers, and endoscopic therapy.15 Since then,the AGA has published Clinical Practice Updates on the man-agement of BE with LGD, ${ }^{16}$ endoscopic submucosal dissection(including outside the setting of BE), ${ }^{17}$ endoscopic treatment of neoplastic BE, ${ }^{18}$ screening and surveillance, ${ }^{19}$ and most recently, clinical practice guideline on EET of BE and related neoplasia using the GRADE framework. ${ }^{10}$

Formulation of Clinical Questions and Determining Outcomes of Interest

Through an iterative process, the guideline panel outlined focused clinical questions related to endoscopic surveillance in patients with BE that were then used to define patients,intervention, comparator, and outcomes(PICO) and used to inform the study inclusion and exclusion criteria and guided the literature search. The AGA Governing Board approved the final set of questions(Table 1).

Search Strategy

Using an a priori protocol, an information specialist helped search for recently published systematic reviews and meta-analyses(SR/MA) that used a comprehensive search strategy(PubMed, Embase, and Cochrane Library). We updated the searches to January 2025. When no pre-existing SR/MA meeting our inclusion criteria was identified, a new compre-hensive search was conducted on the following databases:Embase, MEDLINE, Cochrane, and PubMed. The search terms used, and the final strategy can be found in the Supplementary Material (Supplementary Table 3). References from published studies and prior guidelines were searched to identify any missing relevant studies. Furthermore, content experts aided in the identification of potentially relevant ongoing studies.

Study Selection, Data Collection, and Analysis

<BeginOfPara>Using Rayray bibliographic software, ${ }^{20}$ 2 experts inde-pendently screened each title and full-text review of the eligible studies to reach consensus and any conflicts were resolved by the senior methodologist (see Supplementary Figure 1 for the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Flow Diagram). ${ }^{21}$ In summary, we prioritized randomized controlled trials (RCTs) and observa-tional studies with control arms (when RCT data were not available or sparse). Data were extracted from each study and MAs were conducted when possible using the generic inverse variance method of weighting and applied the random-effects model (or fixed-effects model when $\leq 3$ studies were included)using either STATA software, version 14.2 (StataCorp, College Station, TX, 2015) or Review Manager RevMan software,version 5.3 for comparative studies (Copenhagen, Denmark,The Cochrane Collaboration, 2014), and OpenMeta analyst for statistical analyses of single arm studies (OpenMetaAnalyst, B.

C. Wallace, I. J. Dahabreh, T. A. Trikalinos, J. Lau, P Trow, and C.H, Schmid). Statistical heterogeneity was assessed using the $I^{2}$ index. For risk of bias assessment of diagnostic accuracy studies, we used the revised QUADAS-2 tool ${ }^{22}$ for diagnostic accuracy studies, the Newcastle Ottawa tool ${ }^{23}$ for observa-tional studies and Cochrane risk-of-bias tool for RCTs.

Certainty of the Evidence

We used the GRADE approach to assess the certainty of evidence using the software GradePro Guideline Development Tool (https://gradepro.org). The GRADE approach considers factors, such as study design, population studied, risk of bias,inconsistency, indirectness, imprecision, and risk of publica-tion bias, to rate the certainty of evidence as high, moderate,low, or very low (Supplementary Tables 1 and 2). ${ }^{24}$ The results of certainty assessment are reported in evidence profiles available in for all the interventions.

Development of Recommendations

The process of translation of evidence into guideline rec-ommendations followed the GRADE Evidence-to-Decision (EtD) framework and consensus was achieved by discussion during virtual meetings of the committee. ${ }^{25}$ The EtD frame-work considers the certainty of evidence, balance of benefits and harm, patient values and preferences, feasibility, accept-ability, equity, and resource use. ${ }^{25}$ All EtD tables are presented in the document. The interpretation of strength of recom-mendations is summarized in Supplementary Table 2. In sit-uations when the recommendation is only supported with very low certainty for the benefits and very low certainty for po-tential harms, the guideline panel put a higher value on risk avoidance.

Document Review

The guideline underwent external peer review and public comments, and the final document was revised to address pertinent comments.

Recommendations

As outlined, a summary of the recommendations and the accompanying implementations is pro-vided in Table 2. A clinical decision support tool is pre-sented in Figure 1.

Endoscopic Surveillance

In patients with NDBE, endoscopic surveillance may allow for early detection of BE-related neoplasia (early EAC or HGD) and endoscopic eradication with the goal of reducing morbidity and mortality associated with esoph-actomy and EAC-related mortality. ${ }^{7-10}$ The panel deemed the following outcomes as critical: EAC-specific mortality,all-cause mortality, detection of early stage EAC and HGD,and complications from diagnostic esophagogas-troduodenoscopy (EGD). EAC diagnosis was an important outcome (Table 1).

Table 1.Population, Intervention, Comparison, Outcome Questions

|Focused question|Patients|Intervention|Comparator|Outcomes| |---|---|---|---|---| |1. Should patients with BE without dysplasia undergo endoscopic surveillance?|Adult patients with NDBE|Endoscopic surveillance|No surveillance|Benefits: Reduction in EAC-related mortality (critical) Reduction in all-cause mortality (critical) Earlier stage of EAC detection (critical) Harms: Bleeding Perforation SAEs| |2. Should patients with columnar-lined esophagus <1 cm with intestinal metaplasia undergo endoscopic surveillance?|Adult patients with columnar-lined esophagus <1 cm|Endoscopic surveillance|No surveillance|Benefits: Progression to EAC (critical) Progression to HGD/EAC (critical) Reduction in EAC mortality Harms: Bleeding SAEs| |3. What is the optimal imaging strategy for patients with BE undergoing endoscopic surveillance?|Adult patients with BE undergoing surveillance|CE (standard or virtual) plus high-definition WLE|WLE|Benefits: Increased yield of neoplasia (dysplasia and EAC) detection (critical) Reduction in rates of PEEC and PEEN Improved diagnostic characteristics for neoplasia detection (sensitivity and specificity) Reduction in the number of biopsies required Time to complete endoscopy Harms: Bleeding Perforation AEs related to dye-based chromoendoscopy Serious adverse events| |4. What is the role of adjunctive sam- pling techniques in patients under- going surveillance endoscopy?|Adult patients with BE undergoing surveillance|Structured biopsy protocol plus WATS-3D sampling|Structured biopsy protocol|Benefits: Increased yield of neoplasia (dysplasia and EAC, excluding CD or IND) detection (critical) Reduction in rates of PEEC and PEEN Prediction of development of dysplasia or progression Harms: Bleeding Perforation SAEs|

S3NT30IN

1188 Wani et al

Table 1.Continu ed

|Focused question|Patients|Intervention|Comparator|Outcomes| |---|---|---|---|---| |5 and 6. In patients with BE undergoing surveillance endoscopy, is the use of biomarkers superior to grade of dysplasia in predic- tion of progression?|Adult patients with BE undergoing surveillance endoscopy (stratified by NDBE, IND/LGD)|Combination of p53 staining with histology Combination of TissueCypher with or without histology|Standard histopathology|Benefits: Diagnostic characteristics (sensitivity, specificity, TP,TN, FP, and FN) (critical outcome) Improved prediction of progression to HGD/EAC (critical outcome) Proportion of cases with change in management (EET, change in frequency of surveillance endoscopy)| |7. What is the role of chemopreventive strategies in prevention of progres- sion in patients with BE?|Adult patients with BE|Once-daily PPI PPI plus aspirin Twice-daily PPI therapy|No PPI therapy Daily PPI therapy|Benefits: Reduction in progression to EAC (critical) Reduction in progression to HGD/EAC (critical) Reduction in eAC mortality Harms: AEs related to PPI therapy AEs related to aspirin therapy| |8. What is the role of antireflux proced- ures in the prevention of progression in patients with BE?|Adult patients with BE|Antireflux procedures|PPI therapy|Benefits: Reduction in progression to EAC (critical) Reduction in progression to HGD/EAC (critical) Reduction in EAC mortality Harms: AEs related to PPI therapy AEs related to antiref lux surgery|

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1189

What is the role of endoscopic surveillance in patients with nondysplastic Barrett's esophagus?

Recommendation 1: In patients with NDBE, the AGA suggests performing endoscopic surveillance compared with no surveillance.(Conditional recommendation, low certainty of evidence)

Implementation Considerations

● Data supporting optimal surveillance intervals is limited but most experts on the panel advise a high-quality endoscopy every 3 years. Surveillance in-tervals may be extended to every 5 years in patients at lower risk of progression, for instance, those with short-segment BE(<3 cm).

● Discontinuation of surveillance endoscopy in patients with NDBE should be considered based on age and medical comorbidities.

Summary of the Evidence

Evidence informing this recommendation is derived from an RCT, comparative cohort studies, and case-control studies. A prior SR/MA of observational studies (1996-2017) evaluated the association of BE surveillance on EAC-related mortality, overall mortality, and stage at diag-nosis. ${ }^{26}$ We updated the search (2017 through January 2025) and identified 832 studies, of which 36 underwent full-text review, but no new relevant comparative cohort studies were identified (Supplementary Table 3). The only available new evidence was a recently published RCT, the Barrett's Oesophagus Surveillance vs Endoscopy at Need Study (BOSS trial). ${ }^{27}$ The SR/MA by Codipilly et al ${ }^{26}$included 4 cohort studies with the largest study by El-Serag et al ${ }^{28}$ comparing 209 patients diagnosed with EAC during surveillance vs 215 patients with EAC with prior BE diag-nosed by nonsurveillance endoscopy. The BOSS trial con-ducted at 109 centers in the United Kingdom, randomized 1733 patients to surveillance endoscopy every 2 years and 1719 patients were offered "at need" endoscopy only for evaluation of symptoms. For the primary outcome of overall survival (including cancers aside from EAC), median follow-up was 12.8 years. Mean age at time of randomiza-tion was 63 years, 71\% of patients were male, and 56\% of patients had long-segment BE.

Harms of surveillance endoscopy were assessed using 2 large national databases and the BOSS trial. Kim et al ${ }^{29}$ retro-spectively evaluated 387,647 patients in Korea who under-went upper endoscopy and reported serious complications within 30 days of procedure including bleeding, perforation,and cardiopulmonary events. Wang et al ${ }^{30}$ evaluated more than 7 million upper endoscopies in the United States and assessed for infections within 7 days of endoscopy.

Benefits

For EAC-specific mortality, the SR/MA by Codipilly et $al^{26}$ found a pooled relative risk (RR) of 0.60 (95% CI,

0.50-0.71) corresponding to 231 fewer EAC deaths per 1000 (from 289 fewer to 168 fewer) (Table 3). Overall survival data came from the BOSS trial, which randomized 3453 patients to surveillance every 2 years vs symptom-triggered endoscopy, with a median follow-up of 12.8 years. The primary outcome, overall survival, showed no significant difference between groups (hazard ratio [HR],0.95; 95\% CI, 0.82-1.10), with an absolute difference of 9 fewer deaths per 1000. Supporting evidence from Codipilly et al showed a modest mortality benefit (HR, 0.75; 95\% CI,0.59-0.94), which persisted after adjusting for lead-time bias (HR, 0.85; 95\% CI, 0.75-0.95).

Data for the critical outcome of overall survival were derived from the BOSS trial. All-cause mortality in the sur-veillance group was 19.2%(333 deaths in 1733 patients) vs 20.7% in the"at need" endoscopy group(356 deaths in 1719 patients)(HR, 0.95; 95% CI, 0.82-1.10). The absolute risk reduction in the group undergoing scheduled endoscopic sur-veillance was 9 fewer per 1000(from 34 fewer to 18 more)(Table 3). As supportive evidence, for the outcome of all-cause mortality, the SR/MA by Codipilly et al ${ }^{26}$ examined 3 studies with comparative groups and found an HR of 0.75(95% CI,0.59-0.94). The SR/MA also performed sensitivity analyses attempting to adjust for lead-time bias for the outcome of all-cause mortality, which also found an attenuated reduction in mortality in the group undergoing endoscopic surveillance vs no surveillance(HR, 0.85; 95% CI, 0.75-0.95). Data for the important outcome of stage of EAC diagnosis, early-stage EAC and HGD detection was higher in the scheduled surveillance group, 3.3%(58 of 1733 patients) vs 1.2%(20 of 1719 pa-tients) in the at-need surveillance group(RR, 2.82; 95% CI,1.73-4.56) with 21 more cases of early-stage EAC/HGD detected per 1000(from 8 more to 41 more). Additional evi-dence from Codipilly et al also showed cohorts with prior BE diagnosis were significantly more likely to present with earlier stage EAC (stage 0/l) compared with cohorts without a prior diagnosis of BE (RR, 5.52; 95% CI, 3.7-8.24; $l^{2}=96 \%$ )(Table 3).

Harms

Harms of surveillance endoscopy were assessed using 2 large databases. Assessing for adverse events within 30 days after endoscopy, Kim et al ${ }^{29}$ reported low rates of serious complications: bleeding (7.9 per 10,000 persons), perforation(0.4 per 10,000 persons), cerebrovascular accidents (2.8 per 10,000 persons), acute myocardial infarction (2.8 per 10,000 persons), and congestive heart failure (1.5 per 10,000 per-sons). In a US study of more than 7 million endoscopies, Wang et al ${ }^{30}$ found a 7-day post-endoscopy infection rate after of 30 per 10,000 individuals, with respiratory infections in 0.3% of patients and bacteremia in 0.07% of patients. The BOSS trial reported serious adverse events (SAEs) in 8 patients (0.46%)in scheduled surveillance and 7 patients (0.41%) in the "at need" endoscopy group. ${ }^{27}$

Certainty of the Evidence

The overall certainty of evidence for benefits and harms was low (Table 3). For EAC-related mortality, evidence

1190 Wani et al

Gastroenterology Vol.169, Iss.6

Table 2.Summary of Recommendations and Implementation Considerations

|||Strength of recommendation|Certainty of evidence| |---|---|---|---| |Endoscopic Surveillance|||| |1. In patients with NDBE, the AGA suggests performing endoscopic sur-veillance compared to no surveillance.|||| |Implementation considerations for recommendation 1:|||| |● Data supporting optimal surveillance intervals is limited but most experts on the panel advise a high-quality endoscopy every 3 yr. Surveillance intervals may be extended to every 5 yr in patients at lower risk of progression, for instance those with short-segment BE (<3 cm). ● Discontinuation of surveillance endoscopy in patients with NDBE should be considered based on age and comorbidities.|||| |2. In patients with columnar-lined esophagus <1 cm with intestinal meta-plasia, the AGA suggests against surveillance endoscopy.|||| |plasia, the AGA suggests against surveillance endoscopy.|||| |3. In patients undergoing surveillance endoscopy for BE, the AGA recommends using a combination of high-definition WLE plus CE compared WLE alone.|||| |Implementation considerations for recommendation 3: ● Among the CE modalities that meet optimal perf omance characteristics, the choice of CE modality (virtual or dye-based CE) should be based on endoscopist and center expertise.|||| |● CE-directed biopsies should be used as an adjunct to sampling using a structured biopsy protocol.|||| |● A structured biopsy protocol includes targeted biopsies f rom any visible lesions and random 4-quadrant biopsies every 2 cm if no history of dysplasia and every 1 cm if there is a history of dysplasia.|||| |General Implementation Considerations for Endoscopic Surveillance:|||| |● Endoscopic evaluation in patients with suspected or confirmed BE should meet the requirements of a high-quality endoscopic examination.|||| |● Refer patients with BE-related neoplasia, including patients diagnosed with LGD and IND to high-volume endoscopists with expertise in EET, pathologists with expertise in BE neoplasia and access to multidisciplinary care.|||| |● Histologic diagnosis of BE-related dysplasia or early cancer should be confirmed by an expert pathologist.|||| |● The diagnosis of BE and IND and LGD should be conf imbed by a repeat upper endoscopy by an expert endoscopist within 6 mo on high-dose acid suppressive therapy primarily to rule out prevalent HGD or EAC.|||| |● In patients with confirmed LGD choosing surveillance, the panel continues high-dose acid suppressive therapy and performs an upper endoscopy at 6-mo intervals for 1 yr, then annually, until there is a change in histologic grade of dysplasia.|||| |● In patients with IND, the panel performs annual endoscopy until there is a change in histologic grade of dysplasia.|||| |● In patients with IND confirmed by expert pathology review, the expert panel does not perform EET.|||| |4. In patients undergoing surveillance endoscopy for BE, the AGA makes no recommendation for or against the use of WATS-3D as an adjunctive sampling technique to a structured biopsy protocol.||||

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1191

Table 2.Continuited

|||Strength of recommendation|Certainty of evidence| |---|---|---|---| |Recommendations|||| |7. In adult patients with BE, the AGA suggests the use of daily PPI therapy compared with no PPI therapy for the prevention of neoplastic progression of BE.||Conditional|Low| |Implementation considerations for recommendation 7:|||| |●In patients with BE, counsel tobacco cessation and weight loss if overweight or obese.|8. In patients with BE, the AGA suggests use of PPIs over surgery for the prevention of neoplastic progression to HGD or EAC.|Conditional|Low|

 Figure 1. Clinical decision support tool for surveillance in Barrett's esophagus.

S3NT30IN

1192 Wanli et al

Table 3.Grading of Recommendations Assessment, Development and Evaluation Evidence Profile for Population, Intervention, Comparison, Outcome Question: Role of Endoscopic Surveillance in Patients With Nondysplastic Barrett's Esophagus

|Study|Study design|Risk of bias|Certainty assessment|Certainty assessment|Other considerations|Patients, n (%)|Patients, n (%)|Effect|Effect|Certainty|Importance||| |---|---|---|---|---|---|---|---|---|---|---|---|---|---| |Study|Study design|Risk of bias|Inconsistency|Indirectness|Other considerations|Imprecision|Considerations|Surveillance|Not surveillance|Certainty|Importance|No (95% CI)|| |Time to death (from all causes) (median follow-up 12.8 yr) 1 study|Randomized trials|Not serious|Not serious|Not serious|Not serious|Not serious|Very serious|None|333/1733 (19.2)|356/1719 (20.7)|HR, 0.95 (0.82-1.10)|9 fewer per 1000 (from 34 fewer to 18 more)|| |EAC diagnosis 1 study|Randomized trials|Not serious|Not serious|Not serious|Not serious|Not serious|Very seriousb|None|40/1733 (2.3)|31/1719 (1.8)|RR, 1.28 (0.80-2.04)|5 more per 1000 (0.80-2.04)|Low b| |Early-stage EAC and HGD detection rates 1 study|Randomized trials|Randomized serious|Not serious|Not serious|Seriousc|Seriousb|None|58/1733 (3.3)|20/1719 (1.2)|RR, 2.82 (1.73-4.56)|21 more per 1000 (1.73-4.56)|21 more per 1000 (from 8 more to 41 more)|| |EAC-related mortality from NRS 4 studies|Nonrandomized studies|Nonrandomized serious|Not serious|Not serious|Not serious|Not serious|None|None|335/677 (49.5) (28.5)|3834/13,465 (0.57-0.94)|RR, 0.73 (28.5)|77 fewer per 1000 (0.57-0.94)||

Table 3. Continued

|Study|Study design|Risk of bias|Risk of bias|Certainty assessment|Certainty assessment|Other considerations|Patients, n (%)|Patients, n (%)|Effect|Effect|Certainty|November Importance|| |---|---|---|---|---|---|---|---|---|---|---|---|---|---| |Study|Study design|Risk of bias|Inconsistency|Indirectness|Imprecision|Other considerations|Considerations|Other considerations|No Surveillance|No (95% CI)|Certainty|November Importance|Relative (95% CI)| |EAC-related mortality (NRS cohort studies with well-defined surveillance) 4 studies|Nonrandomized studies|Serious|Not serious|Not serious|Not serious|Not serious|Non|101/282 (35.8)|144/249 (57.8)|RR,0.60 (0.50-0.71)|231 fewer per 1000 (from 289 fewer to 168 fewer)|000 Very low|CRITICAL| |Complications from diagnostic EGD, data from large cohort study with 387,647 patients 1 study|Nonrandomized studies|Not serious|Not serious|Not serious|Not serious|Not serious|Not serious|None|Retrorespective, observational cohort study with total of 387,647 patients who underwent EGD Bleeding rate was 7.9/10,000 people Perforations 0.4/10,000 people CVA,AMI,and CHF 2.8/10,000,4.4/10,000, and 1.5/10,000 people, respectively|||CRITICAL Low||

AMI,acute myocardial infarction;CHF,congestive heart failure;CVA,cerebrovascular accident;NRS,norrandomized studies.

aCI crosses the minimal important difference threshold of 1%-1.4%; from clinically significant mortality reduction to clinically significant increase in mortality.

bVery small event number and CI crossing from clinically no important cancer diagnosis to clinically important cancer diagnosis.

cSurrogate outcome, for patient important outcome, such as decrease in EAC mortality.

dLead and length time biases:3 studies in this group had data adjusting for lead time bias and no studies adjusted for length time bias. Sensitivity analysis after accounting for lead-time bias resulted in a substantial attenuation in mortality benefit(HR, 0.85; 95% $Cl, 0.75-0.95)$ .

Lead and length time biases:3 studies in this group had data adjusting for lead time bias and no studies adjusted for length time bias. Sensitivity analysis after accounting for lead-time bias resulted in a substantial CI,0.78-2.07).

Gastroenterology Vol. 169, Iss. 6

came from nonrandomized studies with very low certainty due to serious risk of bias and no adjustment for length time bias with only 3 studies adjusting for lead-time bias.For overall mortality and early stage EAC/HGD, evidence from a single RCT and supporting observational data was low certainty due to serious or very serious imprecision.The effect estimates crossed 3 minimal important differ-ence thresholds (minimal important difference, 1%-1.4%)ranging from clinically significant mortality reduction to clinically significant increase in mortality

Discussion

The guideline panel acknowledged several factors that ultimately impacted the evidence to decision process for this recommendation (Supplementary Table 4). Firstly,despite the limitations in the evidence (low certainty), the panel made the judgment that there are likely benefits from surveillance, and those benefits are likely small to moder-ate. Although the BOSS trial is the first and only RCT of surveillance of BE and contributed important evidence to address the question, it did not report directly on EAC-specific mortality. The study was likely underpowered to detect small differences in all-cause mortality and would have been underpowered for the outcome of EAC-specific mortality if it had been ascertained, as the study planned for a mortality rate of 1.25\% per year and HR of 1.3 without surveillance, translating roughly to an absolute expected difference of 0.29\% per year. Although 5-year mortality from EAC is $79 \%, \quad { }^{31}$ patients with BE are 10 times as likely to die from non-EAC causes than EAC. ${ }^{32}$ A recent SR/MA that assessed EAC and non-EAC mortality risk in patients with BE showed all-cause mortality was elevated in pa-tients with BE compared with population controls (pooled standardized mortality ratio, 1.24; 95\% CI, 1.01-1.53)driven in part by increased EAC mortality risk (standard-ized mortality ratio, 8.98; 95\% CI, 5.12-15.77). Patients with BE were 10 times more likely to die from noncancer etiologies than EAC (RR, 10.71; 95\% CI, 5.98-19.16). ${ }^{32}$

Observational studies suggest a progression rate of NDBE to EAC of 0.6% per year and a RR with surveillance for EAC mortality of 0.6, 26,33-35 which would translate to a substantially smaller expected absolute difference in mor-tality of 0.19% per year with surveillance compared with no surveillance. The BOSS trial observed 356 deaths in 1719 patients over a median of 12.8 years in the "at need"arm for a mortality rate of roughly 1.6% per year. A dif-ference of 0.19% per year with surveillance as expected based on observational studies would roughly equate to an HR of 0.88 for overall mortality, which is within the CIs of the effect observed in the BOSS trial. Furthermore, in the BOSS trial, 59% of subjects in the control arm underwent upper endoscopies "at need," including for subjectively worsening symptoms of gastroesophageal reflux disease(GERD); the median intervals between endoscopies were only slightly longer than those assigned to surveillance(25.7 months vs 24.8 months), limiting the ability to detect differences between surveillance and no surveillance. In addition, nearly 25% of patients at study inception did not

have intestinal metlasia and more than one-half of the EACs were diagnosed in the surveillance arm with T stage >1, which is far greater than observed in prospective high-quality surveillance programs,36 limiting the generaliz-ability of these results. It is unclear how many surveillance endoscopies adhered to the basic tenets of a high-quality endoscopic examination. ${ }^{36}$ For these reasons, the panel relied heavily on observational studies acknowledging the limitations (ie, potential for selection effects for whom endoscopy was performed, lead-time and length-time bias,and potential unmeasured confounders).

Overall, surveillance likely provides small to moderate benefit, the harms were judged to be trivial, and surveillance was expected to be acceptable to patients and feasible to implement. Costs of surveillance were judged to be moder-ate. Although the BOSS trial suggested surveillance was not cost-effective, it did not account for the contamination described above and in the Markov analysis modeling extension of the trial beyond the observed period, it assumed there was no mortality benefit from shifting stages with earlier diagnosis. ${ }^{37}$ A prior SR of cost-effectiveness analyses regarding surveillance strategies that included EET for dysplasia, found that each of the identified studies concluded that surveillance was cost-effective. ${ }^{38-40}$ The panel concluded that the evidence supports a conditional recom-mendation for surveillance in patients with NDBE, recog-nizing variability in patient values and preferences.

Implementation Considerations for

Recommendation 1

Endoscopic surveillance intervals. Although the evidence likely supports endoscopic surveillance for NDBE,data guiding optimal surveillance intervals are limited. Early recommendations (every 2-3 years) were based on expert opinion and small case series of patients undergoing sur-veillance. ${ }^{41}$ NCTs and few empirical studies have evalu-ated interval timing. Some early cohort studies (predating the availability of high-definition endoscopes and virtual chromoenodoscopy, and unknown quality of biopsy sam-pling) suggested dysplasia detection declined with longer intervals between endoscopies. In studies totaling 21 pa-tients who progressed to cancer, dysplasia was detected in 54% at endoscopies performed 2 years before cancer detection, but only in 18%,20%, and 25% at 3,4, and 5 years,respectively. ${ }^{42-44}$ One case-control study comparing pa-tients with GERD who died from EAC or gastric cardia adenocarcinoma compared with controls with GERD found a protective effect only with a prior endoscopy performed at 2 or 3 years; however, none of the controls spared from cancer death had undergone esophagectomy, raising concerns about the biological mechanism explaining the observed association. ${ }^{45}$ A cohort study of patients undergoing close surveillance with Seattle protocol biopsies demonstrated that somatic chromosomal alterations increased dramati-cally between 24 and 48 months before progression to cancer, along with co-selection of large regions of gains or losses in chromosomes and chromosome instability, which occur with or before evidence of histologic dysplasia. ${ }^{46,47}$

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1195

The risk of neoplastic progression from NDBE has since been evaluated by numerous studies, and the observed inci-dence of progression to cancer depends on study design(pop-ulation-based vs active surveillance cohort), duration of follow-up, definition of BE, and whether patients with neoplasia diag-nosed within the first year are excluded.33,34 A modeling study indicated that the study designs would converge to an incidence of 6.4 per 1000 patient-years with long enough follow-up.35 Guidelines have primarily relied on cost-effectiveness analyses,which are calibrated to these observed progression rates for suggesting surveillance intervals. Cost-effectiveness analyses that have included strategies of endoscopic therapy dysplastic BE have identified optimal surveillance intervals of every 2-5 years in NDBE, depending on the setting(United States with shorter intervals than United Kingdom, where the willingness-to-pay threshold is lower) and sex(shorter in-tervals in men than women).38-40 Based on the modeling of male individuals with BE in the United States and the cohort studies above, the expert panel endorses endoscopic surveil-lance intervals of every 3 years in patients with NDBE.40

The effectiveness of surveillance is dependent on a high-quality endoscopic examination. Studies have shown a high prevalence of missed dysplasia and cancer particularly on index examination, with pooled data showing a 17% rate of postendoscopy esophageal adenocarcinoma (PEEC) and postendoscopy esophageal neoplasia (PEEN) in patients with BE undergoing surveillance. 1,48 Prior guidelines and these recent observations have fueled the discussion on repeating an upper endoscopy in patients with NDBE within the first year of index endoscopy to reduce the rates of PEEC and PEEN. However, early repeat examinations may lead to overuse of resources and associated costs. Thus, the preferred approach is to ensure high-quality index examination with appropriate sampling (biopsy protocols) as detailed in sub-sequent sections. However, if clinical circumstances preclude this optimal endoscopic assessment of BE, repeat endoscopy within the first year of diagnosis of NDBE may be considered.

Risk stratification. Although experts generally endorse a 3-year surveillance interval in NDBE, there are several clinical variables that might be used to stratify risk of progression and guide surveillance intervals. The risk factor with the greatest evidence is length of BE,49 and other guidelines have also recommended surveillance every 5 years in patients with short segment BE(<3 cm).50-52 One SR/MA reported that increasing BE length per centimeter was asso-ciated with an increased risk of progression to HGD/EAC(odds ratio[OR], 1.25; 95% CI, 1.16-1.36).53 Another MA of 1979 patients with BE length<3 cm and 2118 patients with BE length of≥3 cm demonstrated that the annual risk of progression to EAC was significantly lower for shorter seg-ments than for long-segment BE(0.06% vs 0.31%; OR,0.25;95% CI,0.11-0.56).49 The risk of progression to HGD or EAC doubles for every 4 cm increase in length.53 However, there is no direct evidence indicating short-segment BE is associated with a longer dwell time in the dysplastic or intramucosal cancer state and extending surveillance intervals may reduce surveillance effectiveness, despite lower progression risk.

Smoking is associated with approximately 1.5 times higher odds of progression. ${}^{53}$ Male sex also confers higher

<EndOfPara>risk, with the 2 MAs ${ }^{53,54}$ reporting a 2-fold increase, although a third MA with individual-level data found a weaker associ-ation (HR, 1.44). ${ }^{55}$ The Progression in Barrett's Esophagus score was developed by combining known risk factors: male sex (9 points), smoking (5 points), and BE length (1 point/cm), plus including confirmed LGD (11 points). ${ }^{56}$ A score $<11$predicted low risk of progression to HGD or EAC of 0.13% per year over mean follow-up of 6 years in the development cohort, and 0.2% per year over a median follow-up of 7.5 years in a validation study. ${ }^{57}$ It should be noted that the point estimates used in Progression in Barrett's Esophagus for the effects of male sex and smoking (HR, 3.0 and 1.8, respectively)were stronger than identified in the MA above, and weaker for length (doubling every 6 cm). Additional studies are needed to understand the role of age (advancing age is associated with increased risk of progression, but with uncertain effect size)and body mass index (increased risk of progression; OR, 1.06 per increments of $5 ~kg / m^{2}$ ). ${ }^{58}$ Although family history has been associated with BE and with EAC, ${ }^{59}$ there are very limited data regarding the risk of family history on progres-sion from NDBE. ${ }^{60}$

The panel acknowledged recent studies evaluating the capsule-sponge test as a nonendoscopic tool to address the current limitations of endoscopic surveillance. ${ }^{61-63}$ In a recent prospective study of 910 BEs, 489 (54\%) were classified as low risk (clinical and capsule-sponge biomarkers were negative), 283 (31\%) as moderate risk (negative for capsule-sponge biomarkers, positive for clinical biomarkers, ie, age,sex, and BE length) and 138 (15\%) as high risk (p53 abnor-mality or glandular atypia regardless of clinical biomarkers, or both). ${ }^{61}$ Patients with both atypia and aberrant p53 had the highest risk of HGD/EAC (RR, 135.8; 95\% CI, 32.7-564),whereas the prevalence of HGD/EAC in the low-risk group was 0.4\% (95\% CI, 0.1-1.6) with a negative predictive value(NPV) for any dysplasia or EAC of 97.8\% (95\% CI, 95.9-98.8).This strategy has the potential to decrease the endoscopic burden and improve risk stratification, although further research is needed to define its role in BE surveillance.

Discontinuation of surveillance. As patients age and develop comorbidities, the benefits of surveillance decline due to increase procedural risks and competing mortality, partic-ularly from cardiovascular disease. ${ }^{64,65}$ Based on a survey of gastroenterologists, most support discontinuing surveillance of NDBE based on age and comorbidities. ${ }^{66}$ Although no RCTs or prospective studies have been performed to guide the timing of discontinuation, cost-effectiveness analysis using 3 models suggested optimal discontinuation ages ranging from 65 to 83 years, depending on sex and comorbidities. ${ }^{67}$ Optimal discon-tinuation was at younger ages in women compared with men due to the lower risk of progression, despite the longer life expectancy. The European Society of Gastrointestinal Endos-copy recommended cessation of endoscopic surveillance in individuals at age 75 years in the absence of a history of dysplasia. ${ }^{50}$ We propose initiating the conversation about discontinuation of surveillance at the earliest potential penul-timate EGD (eg, age 75 years in a man without comorbidities),introducing the concept of diminishing benefit and rising risk with continued surveillance with advancing age. The age at which to discontinue should be based on patient life expectancy

1196 Wani et al

Gastroenterology Vol. 169, Iss. 6

(due to comorbidities) and might also be influenced by risk of $$progression(\text{ length of BE});\text{ fitness to undergo repeat endos-}$$copy; and ability to tolerate endoscopic, surgical, or oncologic therapies for esophageal neoplasia.

What is the role of endoscopic surveillance in patients with columnar-lined metaplasia without neoplasia?

Recommendation 2: In patients with columnar-lined esophagus < 1 cm with intestinal metaplasia without neoplasia, the AGA suggests against endoscopic surveillance. (Conditional recommendation, very low certainty of evidence)

Columnar-lined esophagus $<1 ~cm$ is a finding that has been reported in approximately $15 \%$ of the population un-dergoing upper endoscopy. ${ }^{68}$ Observational studies have re-ported an up to $44 \%$ prevalence rate of intestinal metaplasia in patients with columnar-lined esophagus $<1 ~cm .69,70$ To address surveillance in this population, the panel identified the following critical outcomes: progression to HGD/EAC,progression to EAC, and endoscopic complications.

Summary of the Evidence

Evidence informing the recommendation was derived from 6 single-arm observational cohort studies. No existing SR/MA evaluating this question was identified, so a de novo literature search (through March 1, 2025) was conducted(Supplementary Table 3). This search identified 1014 studies, of which 74 underwent full-text screening and 6 rele-vant studies were included in the final analysis. We could not address the critical outcome of progression to EAC alone, as not all studies reported disaggregate outcomes for HGD and EAC.No study addressed the patient-important outcome of EAC-related mortality. The largest study was by Anaparthy et al ${ }^{71}$$(408$ patients, 5 US tertiary care centers) included patients with<1 cm BE without dysplasia (excluding patients who devel-oped HGD or EAC within 1 year of BE diagnosis to exclude prevalent neoplasia) who were followed for a mean of 5.2 years. The BE cohort was 88% men, age mean age was 59.2 years,and 93% of the cohort were White. Incident HGD/EAC

developed in 5 patients with an annual risk of HGD/EAC of$0.23 \%$ per year. Data regarding the harms of surveillance endoscopy were derived largely from 2 large national database studies, as previously summarized in PICO question 1.29,30

Benefits and Harms

For the critical outcome of progression to HGD/EAC, we performed an MA and found a pooled incidence of 0.1 per 100 patient-years (Figure 2). There were no data on progression to EAC or EAC-specific mortality. Data on SAEs associated with endoscopy, were previously summarized in PICO ques-tion 1 on the use of surveillance endoscopy in NDBE(Table 3).

Certainty of the Evidence

The overall certainty across the critical outcomes was low(Table 4). For the outcome of progression to HGD/EAC, the evidence was very low, as we relied on data from single-arm,nonrandomized studies with serious risk of bias and indi-rectness, as these studies consisted only of studies without a comparator arm. It is also unclear whether patients under-went endoscopic surveillance under a structured protocol or whether outcomes were based on endoscopy performed as needed. AEs associated with surveillance endoscopy were assessed from observational cohort studies with low certainty of evidence due to serious risk of bias. One cost-effectiveness study was used in the evidence to decision making. ${ }^{72}$ No study addressing patient values and preferences were identified.

Discussion

The guideline panel made a recommendation against routine endoscopic surveillance in patients with columnar-lined esophagus $<1 ~cm$ with intestinal metaplasia without neoplasia. The panel considered several factors, including the low risk of progression to HGD/EAC and minimal benefit from surveillance (Supplementary Table 5). ${ }^{73-75}$ Although intesti-nal metaplasia is common in patients with columnar-lined esophagus $<1 ~cm$ (reported an up to 44\% of patients), the panel weighed the minimal benefit against the burden of lifelong surveillance, including costs, procedural risks, strain on endoscopy resources, and patient anxiety. ${ }^{69,70}$ In addition,the inconsistent terminology and description of this finding at

![Study\nEvents Time\nIncidence Rate\nRate\n95%-CI\n0.0022\n0.0009; 0.0054]\n0\n119\n0.0000\n0.0003; 0.0672]\n0\n0\n0\nItskoviz\n0.0001; 0.0214]\n0\n0\n472\n0.0000\n0.0001; 0.0169]\n0\n0\n0\n3\n799\n0.0038\n0.0012; 0.0116]\n0\n0\n985\n0.0000\n0.0081]\n0\n0.0016[0.0008; 0.0032]\n0.0015[0.0004; 0.0058]\n0\nHeterogeneity: $l^{2}=0.0\%,\tau^{2}=0.0598, p=0.9917$ \n0.002 0.004 0.006 0.008 0.01\n](2/aj0YEe4oUhYkBdfpda0fYP/figured06ce479f7cb45609bc408c3ed88061b_0_16.png) Figure 2. Forest plots for progression to HGD or EAC in patients with columnar-lined esophagus <1 cm with intestinal metaplasia.

Table 4.Grading of Recommendations Assessment, Development and Evaluation Evidence Profile for Population, Intervention, Comparison, Outcome Question: Role of Endoscopic Surveillance in Patients With Columnar-Lined Esophagus <span class="node is a group of Table 4. Grading of Recommendations Assessment, Development and Evaluation Evidence Profile for Population, Intervention, Comparison, Outcome Question: Role of

|Study|Study design|Certainty assessment|Certainty assessment|Imprecision|Other considerations|Impact|Certainty|Importance| |---|---|---|---|---|---|---|---|---| |Study|Study design|Risk of bias|Inconsistency|Imprecision|Other considerations|Impact|Certainty|Importance| |Progression to HGD/EAC 6 studies|Nonrandomized studies|Serious ${}^{a}$|Not serious|Not serious|Nerious ${}^{b}$|Not serious|No comparative data from RCT or nonrandomized studies were found that informs of the benefits or harms of surveillance vs no surveillance in patients with CLE IM $<1$ cm. We identified only single-arm studies reporting on natural progression of CLE IM $<1$ cm to HGD/EAC in patients who underwent follow-up EGDs. The pooled incidence was 0.1 per 100 patient-years.|Very Low $^{a,b}$| |Complications from diagnostic EGD, data from large cohort study with 387,647 patients 1 study|Nonrandomized studies|Not serious|Not serious|Not serious|Not serious|Not serious|Retrospective, observational cohort study with total of 387,647 patients who underwent EGD Bleeding rate was 7.9/10,000 people RVA, AMI, and CHF 2.8/ 10,000, 4.4/10,000, and 1.5/10,000 people, respectively|Low|

AMI, acute myocardial infarction; CHF, congestive heart failure; CLE, columnar-lined esophagus; CVA, cerebrovascular accident; IM, intestinal metaplasia.

aSuspected selection bias, given small studies and limited follow-up time.

bAll the studies are single-arm studies on natural progression of the patients with IM in CLE<1 cm and no real intervention.

1198 Wani et al

Gastroenterology Vol. 169, Iss. 6

endoscopy(such as esophagogastric junction intestinal metaplaspy or irregular Z-line) and the substantial interob-server variability in endoscopic diagnosis, add to the uncer-tainty of defining this population. ${ }^{76}$ A recent decision modeling analysis evaluated the cost-effectiveness of endoscopic sur-veillance (no surveillance, 1-time endoscopy at 3 years,endoscopy every 3 years and endoscopy every 5 years) in patients with a diagnosis of esophagogastric junction intestinal metaplasia. ${ }^{72}$ Surveillance strategies were assessed based on annual cancer progression incidence rates (0.01%, 0.05%,0.12%, and 0.22%). At the lowest progression rates, either no surveillance (annual incidence rate of 0.01%) or 1-time endoscopy (annual incidence rate of 0.05% or 0.12%) can be considered; however, more data are required on progression rates to identify the optimal strategy. The guideline panel acknowledged the very low certainty evidence outlining po-tential for missed prevalent neoplasia in patients with columnar-lined esophagus < 1 cm. ${ }^{69}$ Therefore, a critical component in the decision making regarding endoscopic sur-veillance in these patients is whether a high-quality endo-scopic examination was performed. If the quality of the index examination is in question, a repeat 1-time high-quality ex-amination can be performed to rule out prevalent neoplasia.Based on the above uncertainties, patients who place a higher value on the potential benefit of endoscopic surveillance in preventing EAC progression, especially in patients who did not undergo a high-quality index endoscopic examination and place a lower value on the risks of endoscopy, can reasonably choose to undergo a 1-time repeat surveillance endoscopy. Future studies addressing the role of risk stratification and identification of patients with columnar-lined esophagus<1 cm at high risk of progression to HGD/EAC are needed.

What is the optimal imaging strategy for Barrett's esophagus patients undergoing endoscopic surveillance? Should adult patients with Barrett's esophagus undergo surveillance endoscopy using high-definition white-light endoscopy plus chromoendoscopy vs white-light endoscopy alone?

Recommendation 3: In patients undergoing surveillance endoscopy for BE, the AGA recommends using a combination of high-definition WLE plus CE compared with WLE alone.(Strong recommendation, moderate certainty of evidence)

Implementation Considerations

● Among the CE modalities that meet optimal perfor-mance characteristics, the choice of CE modality(virtual or dye-based CE) should be based on endo-scopist and center expertise.

● CE-directed biopsies should be used as an adjunct to sampling using a structured biopsy protocol.

- A structured biopsy protocol includes targeted bi-opsies from any visible lesions and random 4-quadrant biopsies every 2 cm if no history of dysplasia and every 1 cm if there is a history of dysplasia.

Several advanced imaging techniques, such as dye-based chromoendoscopy(CE), for example, using acetic acid, or virtual chromoendoscopy(VCE), have been developed to increase detection of dysplasia and early EAC and to identify lesions that may be missed by high-definition white light endoscopy(WLE) and sampling using a structured biopsy protocol. Narrow-band imaging(NBI), the most widely used VCE, enhances mucosal and vascular patterns using filtered blue light without dye.Acetic acid CE highlights neoplastic areas through dif-ferential whitening. For this question, the critical outcome was incremental HGD/cancer detection during surveillance(Table 1).

Summary of the Evidence

Evidence comparing CE(including dye-based and VCE) plus WLE vs WLE alone was informed by RCTs. We identified a 2019 SR/MA published by the American Society for Gastrointestinal Endoscopy Standards of Practice Committee ${}^{77}$ that included 12 RCTs and 1 study was removed because it was an unpublished conference abstract before the year 2020. An updated search iden-tified 1033 studies, yielding 1 additional RCT ${}^{78}$ (Supplementary Table 3). The final MA included 12 RCTs(crossover or tandem design): 6 studies used dye-based CE, ${}^{78-83}$ 5 studies used VCE, ${}^{84-88}$ and 1 study used both. ${}^{89}$ All studies were from Western countries. Baseline pathology varied among studies, including HGD/EAC(3 studies) ${}^{80,84,89}$ ; all degrees of dysplasia, including NDBE(6 studies) ${}^{78,81-83,85,888}$ ; all degrees of dysplasia, including NDBE, except EAC(1 study), ${}^{79}$ BE with dysplasia(1 study), ${}^{87}$ and LGD only(1 study). ${}^{86}$

Benefits

An MA of 11 RCTs(including 795 patients in the CE plus WLE group and 795 patients in the WLE alone group)demonstrated a higher detection rate of HGD/EAC compared with WLE alone(RR, 1.2; 95% CI, 1.03-1.4; I2=0%)(Supplementary Figure 2) with an absolute increase of 33 more cases per 1000 patients(from 5 more to 67 more).For all dysplasia(LGD/HGD/EAC), based on 10 RCTs including 817 participants in each arm, CE plus WLE detected 368 of 817 dysplasia cases compared with 311 of 817 with WLE alone(RR, 1.16; 95% CI, 1.07-1.27; I2=0%)] detecting 61 more cases per 1000 patients(27 more to 103 more)(Supplementary Figure 2, Supplementary Tables 6 and 7).

Subgroup analyses were also performed of studies examining dye-based CE or VCE plus WLE. An MA of 5 RCTs(323 patients in each arm) comparing dye-based CE plus WLE with WLE alone demonstrated a similar point esti-mate for the detection of HGD/EAC as in the overall anal-ysis(RR, 1.19; 95% CI, 0.96-1.46; I2= 0%) detecting 12 more cases per 1000 patient (3 fewer to 30 more). For all dysplasia (LGD/HGD/EAC), dye-based CE plus WLE compared with WLE alone (6 RCTs including 443 partici-pants in each arm) detected 53 more cases per 1000 pa-tients (6 more to 108 more) (RR, 1.19; 95% CI, 1.02-1.39;

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1199

$I^{2}=28.4\%$ ). Comparing VCE plus WLE with WLE alone, an MA of 6 RCTs(472 patients in each arm demonstrated a similar point estimate(RR, 1.22; 95% CI, 0.97-1.52; $I^{2}=$ 0\%) to the overall estimate, detecting 52 more cases per 1000 patient(7 fewer to 123 more). For all dysplasia detection(4 RCTs including 374 participants in each arm)80 more cases per 1000 patient were detected(95% CI, 5 fewer to 186 more) with an RR of 1.16(95% CI, 0.99-1.37; $I^{2}=0\%$ )(Supplementary Figures 2 and 3, Supplementary Table 6).

Given the potential for selection bias by including studies with enriched populations (patients with an established diagnosis of HGD/EAC), a subgroup analysis of 7 RCTs(564 patients in each arm) was performed ${ }^{78,79,81-83,85,88}$that included patients undergoing endoscopy for NDBE and BE-related neoplasia and reflects the general BE surveillance population more accurately. CE plus WLE improved detec-tion of HGD/EAC compared with WLE alone (RR, 1.55;95\% CI, 1.05-2.29; $I^{2}=0 \%$ ) and overall dysplasia detection(RR, 1.17; 95\% CI, 0.95-1.46; $I^{2}=0 \%$ ). For the outcome of number of biopsies required to diagnose dysplasia, 1 crossover RCT (123 patients with BE) compared NBI-targeted biopsies of abnormal mucosa with high-definition WLE using the Seattle biopsy protocol. ${ }^{88}$ Both WLE and NBI detected 92\% of patients with intestinal metaplasia(similar rates of detection), but NBI required fewer biopsies per patient (3.6 vs 7.6; $P<0001$ ). NBI detected a higher proportion of areas with dysplasia (30\% vs 21\%; P=.01),and all 100\% (all 14) of HGD/EAC cases but only $29 \%$ of LGD cases. All areas that harbored HGD and EAC had an irregular mucosal or vascular pattern. We found no studies specif-ically addressing the role of CE in reducing rates of PEEC and PEEN.

Harms

CE is generally safe and well-tolerated, with no signifi-cant adverse events reported. Risks include allergic re-actions to dyes, although this is rare, and theoretical concerns about methylene blue, getting photosensitized by white light, potentially leading to oxidative DNA damage,particularly in Barrett's mucosa. ${ }^{90}$ Finally, CE may slightly increase biopsy numbers and procedure time, but the overall risk of harm is considered very low, particularly with VCE. VCE adds no extra cost as it is integrated into most endoscopes. A single cost-effectiveness study showed that dye-based CE using acetic acid in high-risk patients with BE was more cost effective than random biopsy sampling. ${ }^{91}$

Certainty in Evidence of Effects

The overall certainty of evidence was moderate(Supplementary Tables 6 and 7). For the outcome of detection of HGD/EAC, there was no risk of bias or incon-sistency, but we considered the indirectness of the outcome due to lack of longitudinal follow-up (to determine long-term impact of CE) and lack of a true gold standard.

However, these limitations are likely to apply equally to both CE and WLE. Certainty was rated down for impreci-sion, as CIs crossed the minimal important difference threshold, thus the overall certainty was moderate. The panel also discussed potential selection bias due to inclu-sion of enriched populations with BE-related neoplasia referred for EET. This was mitigated by performing a sensitivity analysis. Finally, the panel also acknowledged that all studies were conducted at expert centers largely by expert endoscopists and the assessment of diagnostic and incremental yields of dysplasia among nonexpert centers may lead to different estimates.

Discussion

Several advanced imaging techniques have been devel-oped to improve dysplasia and early EAC detection, beyond high-definition-WLE and structured biopsy protocol. CE-directed examination may be conducted using dye-based approaches that requires the application of various topical dyes or solutions to the BE mucosa or VCE that uses light filters within the endoscope to achieve enhancement of the mucosal surface abnormalities. Among the VCE platforms, NBI is the most widely used in clinical practice. It uses blue light with narrow-band filters that enables detailed imaging of the mucosal and vascular surface pat-terns with a high level of resolution and contrast.92 Other VCE platforms use proprietary post-image acquisition processing technology to modify the white-light image to enhance the superficial mucosal and vascular patterns. ${}^{77}$ Among the various dye-based CE approaches, acetic acid CE is the most widely studied and used in clinical practice.Acetic acid CE enhances mucosal surface patterns by contrast staining-initial application results in an initial whitening of the BE segment and areas of neoplasia lose this whitening more rapidly than NDBE epithelium. ${ }^{50}$

The performance characteristics of CE in BE surveil-lance patients have been reported on a per-lesion analysis and not per-patient analysis. The American Society for Gastrointestinal Endoscopy Preservation and Incorporation of Valuable Endoscopic Innovations document has provided acceptable performance thresholds for advanced imaging techniques in BE surveillance.93 NBI (sensitivity, 94.2%;NPV,97.5%; specificity, 94.4%) and acetic acid CE (sensi-tivity,96.6%; NPV, 98.3%; specificity, 84.6%) have been shown to meet acceptable performance thresholds.93,94 An international group of experts has developed and validated an NBI classification system for detection of dysplasia and EAC in patients with BE using NBI images of NDBE, HGD,and EAC to characterize regular and irregular mucosal and vascular patterns visible by NBI.95 The overall sensitivity and specificity for these criteria to detect HGD/EAC in selected still images was 80% and 88% and improved to 91\% and 93\%, respectively, when the assessment of images was made with high confidence. The panel suggests the use of this classification system in routine clinical practice acknowledging the learning curves associated with these advanced imaging techniques. Training tools, such as the

1200 Wani et al

Gastroenterology Vol.169, Iss. 6

Barrett's Oesophagus-Related Neoplasia web-based training course, which is an available validated training course focused on improving early neoplasia detection and can be accessed at www.iwgco.net, www.ueg.eu, or www.best-academia.eu.

Implementation Considerations for Recommendation 3

Choice of chromoendoscopy. Given the similar(in-cremental) yield of dysplasia detection using VCE and dye-based CE, ${ }^{77}$ the guideline makes no recommendation on CE

modality(VCE over dye-based CE, type of VCE platform or type of dye)(Supplementary Figures 2 and 3). The choice should depend on endoscopist and institutional expertise.All contemporary endoscopy systems include VCE plat-forms and circumvent several issues related to dye-based CE, such as the need for additional equipment(dye and catheters) and time, difficulty achieving complete and uniform coating of the mucosal surface, and inability to detect superficial vascular patterns. This question was designed to address the adjunctive role of CE to high-definition WLE and not as a substitute for a structured sampling protocol, as areas that do not appear suspicious

 Figure 3.(A) Sampling of BE using a structured biopsy protocol.(B) Prague classification system to define the length of the Barrett's segment.(C) Paris classification system to define any visible lesion within the Barrett's segment.

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1201

on CE may still harbor dysplasia. ${ }^{88}$ A meta-regression analysis in an SR/MA showed that the overall number of biopsies strongly correlated with the diagnostic yield of dysplasia. ${ }^{97}$ Thus, CE-directed biopsies should not be used as a substitute for sampling using a structured biopsy protocol.

The role of artificial intelligence (AI) in improving detection of dysplasia and early EAC detection is evolving.AI platforms using computer-aided detection and computer-aided diagnosis systems have shown promise in identifying neoplastic areas, and potentially quantitating the comprehensiveness of endoscopic examinations. 98-104 Preliminary studies have also demonstrated that these platforms can outperform endoscopist in both BE images and videos ${ }^{105}$ and be used during live endoscopic pro-cedures. ${ }^{99,100}$ Using a benchmarking test set, the computer-aided detection system was superior to endo-scopists in detecting neoplasia (90% vs 74%; OR, 3.75) for images and 91% vs 67% (OR, 11.68) for videos and non-inferior to BE experts. A recent pilot study also showed that computer-aided diagnosis assistance increased char-acterization performance of BE-related neoplasia by gen-eral endoscopists to the level of expert endoscopists. ${ }^{106}$Although it is anticipated that these AI platforms will impact the current approach to BE surveillance, the clin-ical utility of AI in routine practice requires validation in RCTs. ${ }^{107}$

Structured biopsy protocol. After careful initial examination, patients with BE undergoing surveillance endoscopy should be sampled using a structured biopsy protocol, such as the Seattle biopsy protocol, to minimize sampling bias. The Seattle biopsy protocol includes tar-geted biopsies of all visible or suspected lesions plus 4-quadrant random biopsies every 1-2 cm throughout the Barrett's segment extending from the squamocolumnar junction to the lower esophageal sphincter(Figure 3A). ${}^{108}$ Specimens should be placed in separate jars labeled by distance from the incisors. The goal of sampling with the Seattle protocol, although potentially time intensive, is to help increase the accuracy and yield of detection of dysplasia and EAC in patients with BE undergoing surveillance endoscopy. ${ }^{97,109,110}$ Multiple observational studies support the routine use of a structured biopsy protocol. Pooled data from 2 studies(506 patients) showed a higher rate (19.1%) of dysplasia detection with Seattle protocol compared with those undergoing nonprotocolized biopsies (2.6%; RR, 6.27;95% CI, 2.75-14.33) in patients with BE undergoing surveillance. ${ }^{111}$ In a UK study (222 patients) a system-atic 4-quadrant biopsy protocol was associated with a higher proportion of dysplasia detection (LGD/HDG/EAC) (73% vs 27%) compared with solely targeted bi-opsy with WLE. ${ }^{109}$ A retrospective study found that systematic 4-quadrant biopsies outperformed NBI-targeted biopsies with 57.1% of HGD and 86.3% of LGD events diagnosed using biopsies from protocolized 4-quadrant biopsies. ${ }^{112}$

Limitations of the Seattle biopsy protocol include sub-optimal adherence rates (49%). 110,113,114 An important factor associated with low adherence is longer BE segment length with the likelihood of nonadherence increasing by 31% for every 1 cm of BE in the GIQuIC (Gastrointestinal Quality Improvement Consortium) national quality bench-marking registry. ${ }^{115,116}$ Also, this approach samples a fraction of the entire Barrett's segment, is time-consuming,and is associated with considerable costs. Despite these limitations, the Seattle biopsy protocol currently remains the best-studied biopsy protocol for BE and a cornerstone of a high-quality BE examination. In sum, in the context of existing observational data, patients with BE undergoing screening or surveillance endoscopy should be sampled using a structured biopsy protocol (ie, the Seattle biopsy protocol).

General Implementation Considerations for Endoscopic Surveillance

● Endoscopic evaluation in patients with suspected or confirmed BE should meet the requirements of a high-quality endoscopic examination.

● Refer patients with BE-related neoplasia, including patients diagnosed with LGD and indefinite for dysplasia (IND) to high-volume endoscopists with expertise in EET, pathologists with expertise in BE neoplasia and access to multidisciplinary care.

● Histologic diagnosis of BE-related dysplasia or early cancer should be confirmed by an expert pathologist.

●The diagnosis of BE and IND and LGD should be confirmed by a repeat endoscopy by an expert endoscopist within 6 months on high-dose acid sup-pressive therapy primarily to rule out prevalent HGD or EAC.

● In patients with confirmed LGD choosing surveillance,the panel continues high-dose acid suppressive ther-apy and performs an upper endoscopy at 6-month intervals for 1 year, then annually, until there is a change in histologic grade of dysplasia.

● In patients with IND, the panel performs annual endoscopy until there is a change in histologic grade of dysplasia.

●In patients with IND confirmed by expert pathology review, the expert panel does not perform EET.

High-Quality Endoscopic Examination

In patients with known or suspected BE, it is important to perform a meticulous high-quality endoscopic examina-tion with appropriate sampling for accurate disease staging and minimize the risk of missed dysplasia or neoplasia. The

1202 Wani et al

Gastroenterology Vol.169, Iss. 6

AGA Clinical Practice Update for a high-quality upper endoscopy examination117 suggests preprocessing, intra-procedure, and postprocedure best clinical practices. These best clinical practices as they relates to patients with BE are highlighted below and in Table 5.

Preprocedure. Indications, benefits, and risks should be discussed in detail as part of informed consent. Opti-mization of acid suppressive therapy before endoscopic surveillance can help improve interpretability of results and antithrombotic management should follow published guidelines.118

Intraprocedure. Prior guidelines ${}^{50}$ have endorsed a 10-step approach to a high-quality endoscopic examination in BE.119 The first steps rely on achieving and documenting adequate mucosal visualization, followed by identifying the esophageal landmarks. A critical step is careful and systematic examination of the Barrett's segment using high-definition WLE and CE (virtual or dye-based) to enhance identification of subtle mucosal lesions and vascular pattern abnormalities.Although data are limited, a single study showed that average inspection time of more than 1 minute for every 1 cm of Bar-rett's mucosa was associated with higher detection of suspi-cious lesions (54% vs 13.3%). ${}^{120}$ A European society guidelines recommend a procedure time of $\geq 7$ minutes for upper endoscopy and inspection time of $\geq 1$ min/cm of the circumferential extent of the BE mucosa. ${ }^{121}$

For documentation, use of a standardized reporting system is recommended. Prague classification is optimal to describe the circumferential and maximal extent of the columnar mucosa and location of the proximal gastric folds and diaphragmatic hiatus. Size and location of any islands proximal to the maximal extent of the BE

segment should be documented. The Prague classifica-tion system(Figure 3B) has been studied in endo-scopists with varying degrees of experience122 and in different settings123 with excellent reliability（coeffi-cient of 0.94-0.95 in videos BE segments of ≥1 cm).124 Size，location，and macroscopic appearance of any visible lesion using the Paris classification (Figure 3C) is best to describe superficial neoplasia.125 Representative images of visible lesions in patients with BE are high-lighted in Figure 4. In addition to photo documentation of routine landmarks, any suspicious lesions or findings would benefit from more rigorous photo documentation with annotation inclusion of descriptive details to help inform future management. Finally, any presence or absence of erosive esophagitis should be graded and reported using the Los Angeles classification system. In the presence of active Los Angeles classification grade C or D esophagitis, BE surveillance endoscopy should be repeated at least 6 weeks after optimizing antireflux therapy with random biopsies obtained on the follow-up examination after healing of active esophagitis.However, a detailed inspection and targeted biopsies of any suspicious lesions are recommended, even in the presence of severe reflux esophagitis, to prevent delays in diagnosing dysplasia and malignancy. Although the potential for overalling dysplasia (especially LGD) in the setting of active inflammation exists, this should not preclude obtaining biopsies, as expert pathologists have been shown to be able to distinguish inflammation from true LGD.126 When such samples are obtained, docu-mentation should include the presence and severity of the esophagitis visualized. Treatment for 8 weeks was

Table 5.Key Considerations for a High-Quality Endoscopic Examination in Barrett's Esophagus

|Key consideration|Description| |---|---| |Preprocedure|Informed consent: Discuss indications, benefits, and potential risks of the procedure in detail. Acid suppressive therapy: Optimize therapy before BE surveillance to minimize interpretation challenges due to active esophagitis. Antithrombotic management: Follow published guidelines for periprocedural management of antithrombotic agents.| |Intrapprocedure|Visualization techniques: Use a distal attachment cap, appropriate insufflation/desufflation, and mucosal cleansing agents and washing to enhance mucosal visibility. Landmark identification: Accurately document the top of the squamocolumnar junction(both maximal and circumferential extent), gastroesophageal junction, and diaphragmatic hiatus. Systematic inspection: Conduct multiple pull-throughs using high-definition WLE and CE(virtual or dye-based) including retroflexion and inspection of the distal esophagus, gastroesophageal junction and gastric cardia, spend adequate time inspecting the Barrett's mucosa to improve detection of Barrett's-related neoplasia. Standardized reporting: Use the Prague classification to describe BE segment extent and length; Paris classification for any visible lesions; Los Angeles classification to describe presence of esophagitis. Photodocumentation: Capture routine landmarks and mark suspicious lesions with annotations and descriptive details. Biopsy protocol: Use the Seattle protocol for systematic sampling.| |Postprocedure|Provide follow-up recommendations and timing to resume antithrombotics. Document pending pathology review for further guidance. Ensure surveillance intervals align with guidelines.|

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1203

 Figure 4. Representative images of visible lesions in patients with BE.

recommended, as this has been the typical duration of most trials of proton pump inhibitors(PPIs) for the healing of esophagitis and has been recommended in prior guidelines. ${ }^{127-129}$ The panel noted that a relook endoscopy (to document healing of significant esopha-gitis and to assess for any persistent features of ma-lignancy) is likely only needed for those with Los Angeles Classification grades C and D esophagitis(which means a significant portion of mucosal surface is destroyed). ${ }^{130}$ Follow-up EGD may reveal underlying BE in up to 10%-12% of patients. ${ }^{131-133}$ Only after completion of all the above described steps, Seattle

protocol by should be used for sampling of the BE segment.

Postprocedure. Instructions should be provided regarding any follow-up based on endoscopic findings, as well as the timing to resume antithrombotic agents, if applicable. Most patients suspected of harboring BE on index endoscopy are not confirmed as such on histology. ${ }^{134}$ Therefore, endoscopists should refrain from recommending surveillance endoscopy in the report of the index endoscopy until pathology results are available. It is advisable to have a recall system in place for patients with BE under-going surveillance.

1204 Wani et al

Gastroenterology Vol. 169, Iss. 6

Surveillance in Patients With Barrett's Esophagus and Low-Grade Dysplasia

As outlined in the AGA Guideline on EET for BE, patients with dysplastic BE and EAC should be referred to high-volume endoscopists with expertise in endoscopic exami-nation and resection, and expert pathologists. ${ }^{10}$ Up to 27%of patients with BE with LGD and no documented visible lesion referred from community settings to expert endo-scopists are found to have a visible lesion that requires endoscopic resection. ${ }^{135}$ In addition, systematic sampling and expert pathology review led to an upgrade from LGD to HGD in up to 26% of patients and even to EAC in 7%-11%of patients. including some cases with advanced EAC not amenable to EET. ${ }^{135-138}$ Regenerative changes from un-controlled reflux can mimic dysplasia, leading to possible overdagnoses of LGD or IND. ${ }^{139}$ Studies showed LGD may regress with effective reflux control. ${ }^{140,141}$ Therefore, pa-tients should undergo repeat EGR by an expert endoscopist while on maximum acid suppression (PPI twice daily, 30-45 minutes before meals for at least 2 months) with resection of any visible lesion to rule out prevalent HGD or EAC.

As per the AGA Guideline on EET, patients with BE with confirmed LGD should engage in shared decision making regarding EET vs continued surveillance. ${ }^{10}$ Expert pathol-ogy review shows that LGD progresses to HGD/EAC at 5.7%annually, warranting closer surveillance than in NDBE. ${ }^{142}$ If EET is not pursued, cost-effectiveness models support surveillance every 6 months for 1 year, then annually. 40 Those surveillance EGGs should ideally be performed by expert endoscopists with targeted tissue sampling of any visible lesions or endoscopic resection and random 4- quadrant biopsies every $1 ~cm$ . If findings revert to NDBE,3-year intervals can be resumed.

Management of Indefinite for Dysplasia

BE and IND is found in up to 8.4% of patients with BE.143 A 2020 SR/MA (8 studies, 1441 patients) found a pooled incidents of HGD/EAC of 1.5 per 100 person-years (95% CI, 1.0-2.0) with modest between study heteroge-neity $\left(I^{2}=56.5\%\right)$ . The pooled incidence of EAC alone was 0.6 per 100 person-years (95% CI, 0.1-1.1) with consid-erable between study heterogeneity $\left(I^{2}=89\%\right)^{143}$ A recent UK cohort study(88 patients) from 2 identified 102 biopsies with BE and IND: endoscopic follow-up was per-formed in 88%(n= 78) patients and 12 of 78 progressed to LGD (15%, 2.6 per 100 person-years), 6 of 78 progressed to HGD (7.7%, 1.3 per 100 person-years), and 6 of 78 progressed to EAC (7.7%, 1.3 per 100 person-years). ${}^{144}$ Predictors of progression included longer BE segment,multifocal and persistent IND. Another multicenter cohort study(242 patients) found that after a median follow-up of 1.5 years, 184 patients (76%) had no evidence of dysplasia,prevalent neoplasia was identified in 23 patients (9.5%; 20 LGD, 2 HGD, 1 EAC) and 35 patients (14.5%) developed incident neoplasia (27 LGD, 5 HGD, 3 EAC). ${}^{145}$

Management of BE and IND poses similar challenges to those with LGD, including variable progression rates, and interobserver variability among pathologists in differentiating

between true dysplasia and inflammatory changes or reactive atypia. ${}^{146}$ In addition, a high proportion of pa-tients with BE and IND do not demonstrate any dysplasia on subsequent endoscopy especially when performed on high-dose PPI therapy. ${}^{144}$ Given the uncertainties in the true risk of neoplastic progression and the reduction in risk of progression to HGD/EAC using EET, EET is not recommended in patients with confirmed BE and IND.However, these patients should also be evaluated at expert centers and undergo repeat endoscopy on twice-daily PPI therapy within at least 6 months with repeat sampling using the Seattle biopsy protocol and resection of any visible lesions to confirm the diagnosis of IND and rule out prevalent HGD or EAC. ${ }^{144}$ Patients with confirmed BE and IND should undergo surveillance in 1 year and annually thereafter until there is a change in histologic grade of dysplasia. The role of risk-stratification tools,including the use of biomarkers to more accurately prognosticate the finding of BE and IND, is discussed in another section of this guideline.

What is the role of adjunctive sampling techniques in patients with Barrett's esophagus undergoing surveillance endoscopy? Should adult patients with Barrett's esophagus undergoing surveillance endoscopy be sampled using structured biopsy protocol plus wide-area transepithelial sampling vs sampling using a structured biopsy protocol alone?

Recommendation 4: In patients undergoing surveillance endoscopy for BE, the AGA makes no recommendation for or against the use of WATS-3D as an adjunctive sampling technique to a structured biopsy protocol.(Knowledge gap)

Given the limitations of current endoscopic surveillance strategies, as highlighted above, the use of techniques to improve detection of BE-related neoplasia in patients un-dergoing surveillance endoscopy has received much attention in recent years. Although advanced imaging techniques focus on increasing detection of dysplasia and early EAC by improving visualization of neoplastic areas,wide-area transepithelial sampling(WATS-3D) attempts to improve detection of intestinal metaplasia and dysplasia by increasing the surface area sampled. ${}^{77}$ We evaluated the role of WATS-3D as an adjunct to the Seattle biopsy pro-tocol. The critical outcomes that informed the benefits for this PICO question were:(1) yield of HGD/EAC and(2)yield of neoplasia(defined as LGD, HGD, or EAC)(Table 1).Other patient-important outcomes were reduction in rates of PEEC and PEEN, prediction of disease progression(to HGD and/or EAC), and adverse events related to WATS-3D sampling.

Summary of the Evidence

Evidence for the use of WATS-3D as an adjunct to the Seattle biopsy protocol comes from RCTs, cross-sectional studies, and observational studies. No direct comparative evidence with longitudinal follow-up was found. A prior

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1205

SR/MA included 7 studies(5 cross-sectional studies, 1 retrospective cohort study, and 1 RCT) and evaluated the incremental yield of dysplasia detected by WATS-3D in conjunction with forceps biopsy compared with forceps biopsy alone. ${}^{147}$ The composite outcome of dysplasia in this prior SR/MA included IND, LGD, HGD, and EAC. An updated search(January 2021-2025) identified 4 additional studies to include in our analysis(2 cross-sectional studies, 1 RCT,and 1 retrospective cohort study) (Supplementary Table 3). ${}^{148-151}$ We excluded studies examining use of WATS-3D for the purpose of surveillance after EET.Another SR/MA, also published in 2022, did not yield additional data. ${}^{152}$

Adjunctive yield of neoplasia(LGD, HGD, or EAC) was defined as the proportion of cases of neoplasia identified by WATS-3D plus Seattle biopsy protocol divided by the pro-portion of cases identified by Seattle biopsy protocol alone(Table 1). In contrast to the previously published SR/MA,we excluded crypt dysplasia(CD) or IND in our outcome due to their low risk of progression and diagnostic vari-ability. ${ }^{50}$ Thus, we excluded the following 2 studies:(1)Johanson et al. ${ }^{153}$ because the dysplasia outcome included CD, and (2) Agha et al. ${ }^{154}$ because this study reported dysplasia only for the WATS-3D/Seattle protocol biopsy-discordant cases and we could not obtain adjunctive yield rates. For assessing progression to HGD/EAC among cases of BE with NDBE or LGD, we used data from Shaheen et al. ${ }^{155}$ which described progression rates based on WATS-3D diagnosis and compared this with natural history data available in the literature. Historical progression rates published in a prior SR for natural progression of LGD was used as a comparator for the outcome of progression. ${ }^{142}$

We included 9 studies in our final analysis-4 studies including screening and/or surveillance populations and 5 studies including populations enriched for patients with a history of BE-related neoplasia or who were potentially referred for EET. Subgroup analyses were conducted for each. Demographic characteristics among the screening/vsurveillance cohorts were similar: 39%-78% were male and mean age was 56-61 years. In the enriched cohorts, 53%-82% were male and mean age was 63-68 years. The largest enriched population study(Trindade et al ${ }^{151}$ ) included 8471 patients; 1.1% had LGD, 0.13% had HGD, and 0.26% had EAC. In this study, 74% of patients had short-segment BE.151 Demeester et al ${ }^{149}$ performed an RCT of 1002 patients (786 for BE screening and 118 for BE surveillance) comparing WATS-3D as stand-alone sampling method vs Seattle biopsy protocol with most endoscopies performed for screening.Demographic characteristics included: 66% were female,mean age was 57 years, 89% of patients were White, and 21% had short-segment BE. ${ }^{149}$ Progression to HGD/EAC data came from a retrospective cohort study of 4545 pa-tients with NDBE, CD, or LGD who had both WATS-3D and forceps biopsy and had at least 1 follow-up endoscopy with WATS-3D at least 12 months later. Their primary outcome was progression to HGD/EAC on subsequent for-ceps biopsy. ${ }^{155}$ The mean follow-up was 2 years, included 50% of patients were female, mean age of 62 years, and 51%had short-segment BE.

Benefits

A pooled analysis of 9 studies(33,132 patients) showed that WATS-3D plus Seattle biopsy protocol detected 2 more cases of HGD/EAC per 1000 persons(1 more to 4 more)compared with Seattle protocol biopsy alone(RR, 1.61;95% CI, 1.25-2.08; I2= 18%)(Figure 5). In the subgroup analysis for only the screening/surveillance populations(4 studies), incremental detection of HGD/EAC with WATS-3D plus Seattle biopsy was 1 more per 1000 patients(0-2 more), RR of 1.35(95% CI, 0.92-1.99). For the enriched population, the RR was 1.75(95% CI, 1.18-2.58).We also did a sensitivity analysis for the surveillance population only, but including the studies with enriched population(total of 3 studies) and there the RR was 1.68(95% CI, 1.17-2.42). For the outcome, detection of any neoplasia(LGD/HGD/EAC), MA of 6 studies(19,901 pa-tients), WATS-3D plus Seattle biopsy protocol detected 2 more per 1000 persons(0 more to 4 more) compared with Seattle protocol biopsy alone(RR, 1.36; 95% CI, 1.14-1.64;I2=0%)(Figure 5, Table 6, Supplementary Table 8). In the subgroup analysis for screening or surveillance populations(3 studies), the incremental yield of any neoplasia using WATS-3D plus Seattle protocol biopsies was 3 more per 1000 patients(0-7 more) with an RR of 1.42(95% CI,1.00-2.00), and in the 3 studies in the subgroup analysis of the enriched population the RR was 1.33(95% CI, 1.06-1.68).

We then used data from the retrospective cohort study by Shaheen et al ${ }^{155}$ of 4545 patients and compared this indirectly with previously described natural history data on progression of LGD to HGD/EAC. In this study, 5 of 43 patients(11.6%) with LGD diagnosed by WATS-3D pro-gressed to HGD/EAC identified on forceps biopsy in the 2-year follow-up period (5.79 per 100 person-years). In comparison, prior MAs have reported the following pro-gression rates to HGD/EAC: 1.7 per 100 person-years ${}^{142}$ (LGD diagnosed on forceps biopsy) and 0.8 to 13.4 per 100 person-years ${}^{16}$ (LGD with expert pathology confirmation).In addition, in the Shaheen et al study, 28 of 43 patients with LGD diagnosed by WATS-3D (65.1%) had regression to NDBE on subsequent endoscopy. A recent abstract using data from WATS-3D BE registry showed that progression rates to HGD/EAC in patients who were WATS-3D-positive and forceps biopsy-negative were similar to patients who were WATS-3D-negative and forceps biopsy-positive(NDBE: 0.09% vs 0.08%, IND/crypt dysplasia: 1.46% vs 1.59%, and LGD: 5.97% vs 4.8%). ${ }^{156}$ These data suggest that clinical management decisions can be based on WATS-3D results even when forceps biopsies fail to confirm the diagnosis. We found no studies specifically addressing the role of WATS-3D in reducing rates of PEEC and PEEN.

Harms

Among the 9 included studies, no SAEs were reported.One perforation occurred in the WATS-3D group(Demeester et al ${ }^{149}$ ) in a patient on chronic steroids post-hiatal hernia repair/fundoplication for GERD. There were no details whether the perforation was related to the

1206 Wani et al

Gastroenterology Vol.169,Iss.6

A Detection of HGD/EAC

|Study or Subgroup|Experimental|Experimental|Control|Control|Risk Ratio|Risk Ratio|Risk Ratio|Risk Ratio|| |---|---|---|---|---|---|---|---|---|---| |Study or Subgroup|Events|Total|Events|Total|Weight|M-H,Random,95%|CI|M-H,Random|55%| |1.1.1 Screening and/or Surveillance|||||||||| |Gross 2019|1|4203|0|4203|0.6%|3.00[0.12,73.62]|||| |Shaheen 2024|35|6829|30|6829|19.1%|1.17[0.72,1.90]|||| |Smith 2019|23|12899|13|12899|11.5%|1.77[0.90,3.49]|||| |Zhao 2022|1|109|1|109|0.8%|1.00[0.06,15.79]|||| |Subtotal (95 % CI)||24040|24040|32.0%|1.35[0.92,1.99]||||| |Total events|60|44|||||||| |Heterogeneity:Tau 2|3|1.24|$C^{2}$ ;|3.04[0.74; 2|$P=0\%$||||| |Test for overall effect: Z = 1.52(P = 0.13)|||||||||| |1.1.2 Surveillance on enriched or population with known prior dysplasia|||||||||| |Anandasapathy 2011|9|151|8|151|6.8%|1.13[0.45,2.84]|||| |Raphael 2019|11|138|10|138|8.3%|1.10[0.48,2.51]|||| |Trindade 2023|44|8471|23|8471|18.2%|1.91[1.16,3.16]|||| |van Munster 2023|51|172|33|172|25.9%|1.55[0.05,2.27]|||| |Vennalaganti 2018|30|160|7|160|8.9%|4.29[1.94,9.47]|||| |Subtotal (95 % CI)||9092|9092|68.0%|1.75[1.18,2.58]||||| |Total events|145|81|||||||| |Heterogeneity.Tau 2|3|6.7.7.6|4.5|4.0(0.11); 12|47%||||| |Test for overall effect: Z = 2.80(P = 0.005)|||||||||| |Total(95 % CI)||33132|33132|100.0%|1.61[1.25,2.08]||||| |Total events|205|125|||||||| |Heterogeneity:Tau 2|0.03; Chi ${}^{2}$ = 9.81, df = 8 (P = 0.28); ${}^{I}=18\%$ Test for overall effect: Z = 3.67(P = 0.0002)|||||||||

B Detection of LGD/HGD/EAC

|Study or Subgroup|WATS-3D + FB|WATS-3D + FB|FB|FB|Risk Ratio|Risk Ratio|Risk Ratio|Risk Ratio| |---|---|---|---|---|---|---|---|---| |Study or Subgroup|Events|Total|Events|Total|Weight|M-H,Random,95% CI|M-H,Random|W3| |1.2.1 Screening and/or Surveillance||||||||| |Gross 2019|49|4203|26|4203|15.1%|1.88[1.17,3.03]||| |Shaheen 2024|49|6829|41|6829|19.7%|1.20[0.79,1.81]||| |Zhao 2022|6|109|6|109|2.8%|1.00[0.33,3.00]||| |Subtotal(95% CI)||11141||11141|37.6%|1.42[1.00,2.00]||| |Total events||73||||||| |Heterogeneity:Tau2=0.02;Ch $^{2}$=2.43,df=2(P=0.30);P=18%||||||||| |Test for overall effect: $Z=1.97(P=0.05)$||||||||| |1.2.2 Surveillance on enriched population with known prior dysplasia||||||||| |Anandasapathy 2011|54|151|38|151|27.8%|1.42[1.00,2.01]||| |Raphrael 2019|25|138|21|138|12.0%|1.19[0.70,2.02]||| |Trindade 2023|59|8471|45|8471|22.6%|1.31[0.89,1.93]||| |Subtotal(95% CI)|8760|8760|8760|62.4%|1.33[1.06,1.68]|||| |Total events|138|104||||||| |Heterogeneity:Tau2=0.00;Ch $^{2}$=0.01;df=2(P=0.85);P=0%||||||||| |Test for overall effect: $Z=2.43(P=0.02)$||||||||| |Total(95% CI)|19901|19901|100.0%|1.36[1.14,1.64]||||| |Total events|242|177||||||| |Heterogeneity:Tau2=0.00;Ch $^{2}$=2.84,df=5(P=0.73);P=0%||||||||| |Test for overall effect: $Z=3.31(P=0.0009)$||||||||| |Test for subgroup differences: $Chi^{2}=0.08,df=1(P=0.78),P=0$ %|||||||||

Figure 5. Forest plot of incremental neoplasia detection using WATS-3D plus structured biopsy protocol compared with structured biopsy protocol alone.(A) Detection of HGD/EAC.(B) Detection of LGD/HGD/EAC.

surgery or upper endoscopy. WATS-3D added a mean of 4.5-4.8 minutes to procedure time in addition to Seattle biopsy protocol. ${}^{148,157}$ The panel acknowledged that this was not relevant if this was associated with an incremental yield in neoplasia detection. The cost-effectiveness of WATS-3D sampling as an adjunct to the Seattle biopsy protocol in patients with BE undergoing surveillance has

not been assessed. Other potential harms include the downstream effects of potential false-positive(FP) results for dysplasia(where the diagnosis of dysplasia is not confirmed on follow-up endoscopy). Prior studies on can-cer screening have suggested psychosocial harms or nega-tive impacts on screening adherence associated with FP test results. ${ }^{158,159}$ Additional potential harms could include

November 2025

Table 6. Grading of Recommendations Assessment, Development and Evaluation Evidence Profile for Population, Intervention, Comparison, Outcome Question: Role of Adjunctive Sampling Techniques (Wide-Area Transepithelial Sampling) in Patients With BE Undergoing Surveillance: Neoplasia Detection Rates Between Wide-Area Transepithelial Sampling Plus Structed Biopsy Protocol vs Structured Biopsy Protocol Alone

|Study|Study design|Risk of bias|Certainty assessment|Certainty assessment|Other considerations|Patients, n(%)|Patients, n(%)|Effect|Effect|Certainty|Importance|| |---|---|---|---|---|---|---|---|---|---|---|---|---| |Study|Study design|Risk of bias|Inconsistency|Indirectness|Other considerations|Imprecision|Considerations|WATS- 3D+FB|FB|Certainty|Importance|| |HGD/EAC detection (diagnostic yield; additional cases detected using WATS-3D as add-on test) 8 studies|Nonrandomized studies|Serious|Not seriousa|Not serious|Not serious|Not serious|None|138/ 22,548 (0.6)|82/22,548 (0.4)|RR,1.61 (1.25-2.08)|2 more per 1000 (from 1 more to 4 more)|0000 Very Low| |LGD/HGD/EAC detection (diagnostic yields||||||||||||| |LGD/HGD/EAC detection (diagnostic yields||||||||||||| |as add-on test)||||||||||||| |6 studies|Nonrandomized studies|Serious|Not serious|Not serious|Not serious|Nonserious|None|19901 (1.1)|177/19,901 (0.9)|RR,1.18 (0.96-1.45)|2 more per 1000 (from 1000 from Very Low 4 more)|000CRITICAL|

Harms

1 study

Randomized trials

Not serious

Not serious

Not serious

Extremely serious

None

In the RCT there was only 1 SAE in the WATS-3D group(perforation).

Cross-sectional studies with adjunctive diagnostic yield without referent

standard test were used.Thus,no downstream harms from FP testing

are available. Less concern for FN, given the adjunctive role of the test.

FB,forceps biopsy.

${}^{a}$ Limited follow-up to assess outcomes related to increased dysplasia detection on WATS and lack of confirmation.

bIn the RCT, there was just 1 event and very wide Cl.

1208 Wani et al

Gastroenterology Vol.169, Iss. 6

unnecessary use of endoscopy for closer surveillance or adverse events after EET for treatment of FP dysplasia(including bleeding, stricture, or perforation).10

Certainty of the Evidence

The overall certainty in the evidence for critical out-comes was very low(Table 6, Supplementary Table 8). For detection of HGD/EAC or neoplasia(LGD/HGD/EAC), we rated down for risk of bias and imprecision because the included cross-sectional studies reported adjunctive diag-nostic yield without reference standard tests(eg, confir-matory endoscopic biopsy), there was limited follow-up, and 5 of the 9 studies included study populations enriched for patients with dysplasia. The risk of false-negative(FN)results was considered less concerning due to the test's adjunctive role. Certainty for adverse events was very low due to imprecision and lack of data on potential harms from FP results.

Discussion

Given the limitations of current endoscopic surveillance strategies, techniques like WATS-3D have gained attention to improve detection of BE-related neoplasia. WATS-3D at-tempts to improve detection by sampling a broader surface area with an abrasive brush is passed through the channel of the endoscope to sample deeper layers of the glandular Barrett's epithelium across areas with columnar-lined mu-cosa. The sample acquired by this brush contains dis-aggregated clumps of structurally intact tissue, the sample is then smeared on a slide, yielding a tissue specimen that is up to 150 um in thickness, unlike a typical forceps biopsy slide in which tissue sectioning produces samples that are only 3-to 5-μm thick.77 This is followed by analysis using a neural network algorithm and computer analysis that results in generation of 3D images of the sampled BE. These images are scanned and high-risk features are flagged for pathologists trained in interpretation of WATS-3D samples to review and provide a final diagnosis. ${}^{50,160}$

The guideline authors considered several factors when providing a recommendation for this PICO question.Several included studies were conducted using an enriched population of patients with BE (patients with an estab-lished diagnosis of BE-related neoplasia referred for EET).However, the real clinical relevance of these adjunctive sampling techniques is best addressed in patients with BE undergoing surveillance endoscopy without known dysplasia. Both the absolute and adjunctive effects were attenuated when the impact of WATS-3D was restricted to a surveillance population. Another limitation is that the incremental benefit in dysplasia detection was not confirmed in subsequent sampling in most studies and the longitudinal follow-up is limited. The increase in dysplasia detection reported on WATS-3D is also largely driven by higher detection rates of LGD. It is not clear that the very small absolute increase in LGD detection is worth the substantial cost. There are no studies assessing the cost-effectiveness of WATS-3D as an adjunctive sampling tech-nique and patient preferences and values have not been

addressed. Future studies need to validate results reporting the natural history of dysplasia detected on WATS-3D sampling alone and demonstrate that these results can be reproduced by nonindustry pathologists, as well as repro-ducible in community settings with nonexpert endo-scopists, improving the generalizability in routine clinical practice. A recent pilot study showed that gastrointestinal pathologists without any prior experience can interpret WATS-3D specimens with a high level of accuracy and reproducibility after a short training session. ${}^{161}$ The role of WATS-3D sampling in patients undergoing screening for BE will be addressed in a separate document. Finally, it should be noted that the guideline panel does not recommend against the use of WATS-3D and recognizes the importance of incremental dysplasia and EAC detection in select cases over histopathologic assessment using Seattle biopsy pro-tocol alone. Ongoing prospective trials will address the knowledge gaps discussed above, which will impact the recommendation provided for the use of WATS-3D in routine endoscopic surveillance in patients with BE.

Based on the available literature and as recommended by CDx Diagnostics, WATS-3D sampling should not be used as a substitutive sampling technique to a structured biopsy protocol (Seattle biopsy protocol). The use of WATS-3D sampling technique alone vs the Seattle biopsy protocol in patients with BE undergoing surveillance has not been assessed and is currently being assessed in an ongoing multicenter RCT that will compare the diagnostic yield of LGD/HGD/EAC in patients with BE undergoing surveillance(ClinicalTrials.gov, Number: NCT05530343). This study will also provide the evidence needed to assess the role of WATS-3D as an adjunct to the Seattle biopsy protocol in a nonenriched surveillance population.

Due to the lack of confirmed outcomes, most experts agree that patients diagnosed with dysplasia only by WATS-3D should undergo repeat endoscopy with surveil-lance within 3-6 months using a structured biopsy protocol and endoscopic resection of any visible lesions. The man-agement of patients diagnosed with HGD/EAC based solely on WATS-3D sampling(without confirmation on repeat endoscopic sampling) should be individualized with consideration for EET and the risks and benefits of EET should be adequately discussed with patients. With WATS-3D,160,162 pathologists report CD on WATS-3D samples when dysplasia is detected only in crypts and not in the surface epithelium.155 Although previous studies suggest that the biologic and molecular properties of CD may be similar to LGD, clinical outcomes data associated with this diagnosis of CD are limited but appear to have intermediate risk between NDBE and LGD.155 Based on these diagnostic uncertainties, the experts on the panel advise that patients diagnosed with CD,IND, and LGD based solely on WATS-3D should undergo endoscopy by an expert endoscopist on a high-dose acid suppressive regimen within 6 months.

Biomarkers

A biomarker, as defined by the World Health Organi-zation, is any substance, structure, or process that can be

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1209

measured in the body or its products and influence or predict the incidence of outcome or disease. ${ }^{163}$ Histologic grading of dysplasia during endoscopy is currently the primary biomarker used for risk stratification and man-agement of BE based on its stepwise progression from LGD to HGD to mucosal EAC and finally, invasive EAC. This process is driven by a series of genetic and epigenetic al-terations, ${ }^{164}$ leading to chromosomal instability and even-tually carcinogenesis. However, dysplasia as a biomarker has significant limitations. These include the existence of patchy and focal distribution of neoplasia within the BE segment and that a structured biopsy protocol samples only a fraction of the entire BE segment. This coupled with the lack of adherence to biopsy protocols adds to the risk of sampling errors. ${ }^{116}$ In addition, the high degree of inter-observer variability among pathologists for detecting dysplasia decreases the impact and the efficiency of relying on a histologic diagnosis of dysplasia. ${ }^{50,165-167}$ Finally,recent data suggests that a substantial proportion of EACs develop more rapidly through an alternate pathway of ge-netic and epigenetic alterations, followed by genome doubling and then catastrophically acquiring genomic instability, oncogene amplification, and EAC. This pathway may explain, in part, EACs not detected through endoscopic surveillance programs. ${ }^{164,168,169}$

In this guideline, the panel evaluated the role of commercially available biomarkers for the neoplastic pro-gression of BE in the United States. The outcomes of in-terest were the overall performance characteristics (sensitivity, specificity, true positive [TP], FP, true negative [TN] and FN, positive predictive value (PPV), and NPV) and biomarkers as predictors for progression among patients with BE. Similar to previous AGA guidelines on the role of biomarkers in inflammatory bowel disease, 170,171 the guidelines panel considered the downstream consequences on important patient outcomes corresponding to each possible outcome of the diagnostic test. Patients with TP results are identified as patients at increased risk of pro-gression and would be considered for endoscopic surveil-lance at shorter intervals or undergo EET. Patients with TN results are at lower risk of progression and might undergo endoscopic surveillance at longer intervals. However, pa-tients with FN results may be falsely reassured and may undergo endoscopic surveillance at longer intervals or no surveillance increasing the risk of undetected progression to EAC that is not amenable to EET. Patients with FP results may undergo unnecessary endoscopies and EET, increasing patient anxiety and resource utilization.

What is the role of p53 assessment in patients with Barrett's esophagus undergoing endoscopic surveillance? Is p53 assessment superior to the grade of dysplasia in predicting progression in patients with Barrett's esophagus?

<BeginOfPara>Recommendation 5: In patients diagnosed with NDBE,BE with IND, or BE with LGD, the AGA makes no recommendation for or against the routine use of p53 assessment as an adjunct test to histopathology to predict progression to HGD or EAC.(Knowledge gap)

p53 is an important tumor suppressor gene and alteration in function is a key event during progression of BE to EAC. The detection of p53 abnormalities in BE by immunostaining is the most widely investigated biomarker that can be used as an adjunct to aid in the diagnosis of dysplasia and as a biomarker for risk stratification in BE. Aberrant expression of p53 pro-tein is evidence of alteration in p53 function (either over-expression or absent expression). ${ }^{50}$ Abnormalities in p53 expression can be detected by p53 immunostaining of formalin-fixed, paraffin-embedded BE tissue or sequencing.Although sequencing provides an objective evaluation of p53 status and can be combined with other genomic biomarkers,this approach is time-consuming, expensive, and requires special instruments and DNA isolation. However, p53 immunohistochemistry (IHC) staining has the largest body of evidence, is faster, relatively inexpensive and can be per-formed at any pathology laboratory that has the technical capability of performing it.

Summary of the Evidence

Evidence informing the recommendation for the use of p53 as an adjunct test to standard dysplasia assessment of dysplasia in BE surveillance comes from both observational cohort and case-control studies. We identified a 2017 SR/MA (14 studies, 7 cohort, 7 case-control) by Snyder et al ${ }^{172}$ that evaluated the risk of progression to HGD/EAC associ-ated with aberrant p53 expression. An updated search(August 2017 to January 2025) identified a total of 581 studies, of which 29 were eligible for inclusion(Supplementary Table 3). We selected studies that assessed p53 IHC as an adjunct test to histopathology assessment of dysplasia/EAC on surveillance esophageal biopsies and re-ported subsequent progression to HGD/EAC, in patients with baseline NDBE,IND, or LGD and excluded those studies with follow-up of <1 year. The 29 studies included 15 case-control studies, 13 cohort studies, and 1 study with a case-control cohort and a prospective validation cohort. All studies were from Western countries. Baseline pathology with respect to IND, LGD, and NDBE varied and the mean duration of follow-up was 4.4 years(range, 2.3-6.2 years).There was considerable variability between studies in how aberrant p53 expression was assessed and scored strictly overexpression of the p53 protein on IHC (13 studies) or as absent or overexpression of the p53 protein (13 studies) and the definition of aberrant p53 expression was not available for the remaining 3 studies. The threshold for over-expression varied, with some studies using a percentage of positive cells in the epithelium or the glands, and others relied on qualitative assessments by pathologists. In addi-tion, there was inconsistency in antibody selection and staining protocols. All of these differences may have contributed to the significant variation in diagnostic per-formance of aberrant p53 expression between studies.

Benefits

The critical outcomes that informed the benefits for this PICO question were predicting progression to HGD/EAC

1210 Wani et al

Gastroenterology Vol. 169, Iss. 6

(based on cohort studies) and diagnostic test characteris-tics(Table 1). The pooled proportion of patients with aberrant p53 expression based on the combined baseline pathology of NDBE, IND, and LGD was 20%(95% CI, 14%-27%; $I^{2}=90\%$ ). The pooled rate of progression among patients with BE undergoing endoscopic surveillance with aberrant p53 expression to HGD/EAC was 8 per 100 person-years(95% CI, 6-11; $I^{2}=26.5\%$ ). In contrast, the pooled rate of progression to HGD/EAC among patients without aberrant p53 expression was 0.3 per 100 person-years(95% CI, 0.1-0.6; $I^{2}=26.6\%$ ). Compared with pa-tients without aberrant p53 expression, patients with aberrant p53 had significantly higher risk for progression to HGD/EAC(RR, 10.17; 95% CI, 6.89-15; $I^{2}=12\%$ )(Figure 6, Table 7).

The pooled proportion of aberrant p53 expression in patients with NDBE was 5%(95% CI, 1%-11%; I2=83%).The pooled rate of progression to HGD/EAC among patients with NDBE with aberrant p53 was 7 per 100 person-years(95% CI, 4-11 per 100 person-years; $I^{2}=0\%$ ) compared with 0.5 per 100 person-years(95% CI, 0-2.2 per 100 person-years; $I^{2}=90\%$ ) without aberrant p53 expression.Compared with NDBE without aberrant p53, patients with NDBE with aberrant p53 had significantly higher risk for progression to HGD/EAC(RR, 4.8; 95% CI, 3.12-7.38; I2=92\%). Using standard annual risk of progression of 0.6%among patients with NDBE, ${}^{173}$ p53 testing resulted in an absolute increase of 23 additional cases of HGD/EAC per 1000 individuals tested-years(95% CI, 13-38 more cases).

The pooled proportion of aberrant p53 expression in patients BE with LGD was 52% (95% CI, 34%-70%; $I^{2}=$ 79.4%). The pooled rate of progression into HGD/EAC among patients with BE and LGD with aberrant p53 expression undergoing endoscopic surveillance was 11 per 100 person-years (95\% CI, 4-20 per 100 person-years; $I^{2}=$ 61.9\%) compared with 1 per 100 person-years (95\% CI,0-5 per 100 person-years; $I^{2}=55.9\%)$ without aberrant p53 expression. Compared with patients with BE and LGD without aberrant p53, patients with LGD with aberrant p53 had significantly higher risk for progression to HGD/EAC(RR, 2.53; 95\% CI, 1.56-4.1; $I^{2}=0\%)$ . Using standard annual risk of progression of 1.7% among patients with LGD, 142 p53 testing resulted in an absolute increase of 6 additional cases of HGD/EAC per 1000 individuals tested(95\% CI, 1-13 more cases) (Figure 6 and Table 7).

The pooled proportion of aberrant p53 expression in patients with BE and IND was 28\% $\left(95\% CI, 16\%-42\%\right.$ ; $I^{2}=81.6\%\}$ . The pooled rate of progression to HGD/EAC among patients with BE and IND with aberrant p53 expression undergoing endoscopic surveillance was 6 per 100 person-years $(95\% Cl, 0-15$ per 100 person-years; $I^{2}=$ $73\%)$ compared with 1 per 100 person-years $(95\% Cl, 0-2$ per 100 person-years; $I^{2}=0\%)$ without aberrant p53 expression. Compared with patients with BE and IND without aberrant p53 expression, patients with BE and IND with aberrant p53 expression had a significantly higher risk for progression to HGD/EAC(RR, 4.99; 95\% CI, 2.98-8.35; $I^{2}=0\%$ )]. Using standard annual risk of progression of

$1.5\%$ among patients with IND, ${}^{143}$ p53 testing resulted in an absolute increase of 20 additional cases of HGD/EAC detected per 1000 individuals tested(95% CI, 10-36 more cases)(Figure 6 and Table 7).

Harms

No studies specifically addressing harms associated with p53 testing were identified but potential concerns were: unnecessary surveillance endoscopies at shorter in-tervals or EET for FP results and extending surveillance intervals for FN results. In an MA of 10 studies, the pooled sensitivity was 0.42(95\% CI, 0.30-0.54) and the pooled specificity was 0.96(95\% CI, 0.90-0.99)(Table 7 and Figure 7). Assuming a 0.6\% risk for progression in patients with NDBE, for 1 year of follow-up, for every 3 per 1000 TP test results there were 40 FP results and for every 954 TN results there were 3 FN test results. The calculated PPV was 5.96%, and NPV was 99.96% for 1 year of follow-up(Table 7). Furthermore, a subgroup analysis among pa-tients with BE and IND and LGD was performed(IND only[6 studies], LGD and IND together[1 study] and LGD only[5 studies]]. The pooled sensitivity for BE and IND/LGD was 0.77(95\% CI, 0.62-0.87) and specificity was 0.76(95\% CI, 0.66-0.85). For patients with BE and IND, the pooled sensitivity was 0.72(95\% CI, 0.47-0.88), and the pooled specificity was 0.79(95\% CI 0.60-0.91). Using a risk for progression of 1.5%, the test accuracy per 1000 individuals was: TP 11, FP 207, TN 778, and FN 4 for calculated PPV of 4.2% and NPV of 99.5%. Similarly, for LGD only, the pooled sensitivity was 0.85(95\% CI, 0.66-0.94), and the pooled specificity was 0.64(95\% CI, 0.55-0.73). Using a 1.73% progression rate, the test accuracy per 1000 individuals was: TP 15, FP 315, TN 668, and FN 2 for calculated PPV of 4.5%, and NPV of 99.7%(Table 7 and Figure 7). The author panel had extensive discussions regarding the downstream consequences of FP results:unnecessary endoscopies and possible EET and potential adverse without any added benefit, potential financial and psychological burden, and unnecessary health care costs and resource overutilization. The FN results could have the potential for false reassurance, leading to extending sur-veillance intervals, and placing patients at risk for disease progression and poor overall outcomes. The author panel also recognized the significant variability among patholo-gists in the interpretation of these results, the lack of standardized criteria that may contribute to inconsistent reporting adding to the diagnostic uncertainty complicating clinical decision making.

Certainty in Evidence of Effects

The overall certainty in the evidence was very low(Table 7, Supplementary Table 9). Using the QUADAS-2 tool, there were issues in multiple domains leading to very serious risk of bias: (1) patient selection: case-control study design was used in more than a one-half of the studies and not all the studies stratified results based on baseline histology;(2) index diagnostic test(p53): there

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1211

 Figure 6.Forest plots for pooled outcome of progression to HGD/EAC in patients with BE with aberrant p53 expression compared with those without.(A)BE with NDBE,IND, and LG.(B)Nondysplastic BE.(C)BE with LGD.(D)BE with IND.

1212 Wani et al

Gastroenterology Vol. 169, Iss. 6

Table 7.Grading of Recommendations Assessment, Development and Evaluation Evidence Profile for Population,Intervention, Comparison, Outcome Question: Role of Biomarkers(p53 Staining) in Predicting Progression in Patients With Barrett's Esophagus Undergoing Surveillance Endoscopy

|Sample site|NDBE|IND|LGD| |---|---|---|---| |Sensitivity (95% CI)a|0.42(0.30-0.54)|0.72(0.47-0.88)|0.85(0.66-0.94)| |Specificity (95% CI)|0.96(0.90-0.99)|0.79(0.60-0.91)|0.64(0.55-0.73)| |Assumed risk of progression to EAC over 1 yr,%|0.6|1.5|1.73| |Outcome, effect per 1000 patients tested|||| |TPs (patients with aberrant P53 who progress to EAC)|3(2-3)|11(7-13)|15(12-16)| |FNs (patients with normal p53, but progress to EAC)|3(3-4)|4(2-8)|2(1-5)| |No. of studies|9|7|7| |No. of patients|686|271|271| |Certainty of evidence|000 Very LOWb,c|000 Very LOWb,c|000 Very LOWb,c| |TNs (patients normal p53 and did not progress to EAC)|954(895-984)|778(581-887)|668(609-708)| |FPs (patients with aberrant p53, but did not progress to EAC)|40(10-99)|207(98-404)|315(275-374)| |No. of studies|9|7|7| |No. of patients|2170|706|707| |Certainty of evidence|000 Very LOWb,c|000 Very LOWb,c|000 Very LOWb,c|

NOTE. This table is based on applying the sensitivity and specificity estimates to calculate TPs, FPs, TNs, and FNs in a hypothetical population of 1000 individuals.

aSensitivity and specificity were pooled from studies that report on combine HGD/EAC outcome.

bQUADUS 2 tool was used for assessing risk of bias and there were issues in multiple domains:(1) patient selection: case-control study design was used in more than one-half of the studies stratify for baseline pathology;(2)index test: there were concerns regarding applicability if the index test because there was significant variability in inter-pretation of the test results;(3) reference test was standard biopsy and progression to EAC, but the follow-up period and therefore interpretation varies between the studies.

cSerious inconsistency, high $\ell^{2}$.

${}^{g}$ The FP range crosses the clinical threshold of 20%(200/1000).

were concerns regarding applicability of the index test; and(3) reference test was standard biopsy and progression to HGD/EAC with variable follow-up. Furthermore, there was significant heterogeneity largely driven by significant variability in performing and interpreting p53 testing,imprecision(FP results crossed the clinical threshold of 20% and low event rate for number of patients progressing to HGD/EAC) and finally, indirectness. There was a lack of studies reporting on how p53 results impacts decision making in clinical practice. Studies addressing the cost-effectiveness of p53 assessment are limited and with methodological issues. ${}^{174,175}$

Discussion

The guideline panel makes no recommendation in favor of or against the use of p53 testing and the role of p53 for risk stratification among patients with BE as a knowledge gap. There are several issues with the routine use of p53 IHC in clinical practice, including the subjec-tive nature in the interpretation of p53, the need for interpretive skill, and no consistent scoring system.Different criteria have been reported. Redston et al176 have reported criteria that have undergone internal validation with a large longitudinal cohort and have been validated by sequencing. These criteria and others have

suggested that abnormal p53 IHC stain in a single-positive crypt is a sensitive and specific marker of dysplasia, however, the reproducibility of this system by other pathologists is uncertain. ${}^{176,177}$ In arriving at this recommendation, the guideline panel heavily weighed the following factors: limited prospective validation data,suboptimal diagnostic characteristics, and the subjective nature of this test. In particular, the panel was uncertain how the results of routine p53 IHC in NDBE should be used to guide management. Patients with aberrant p53 may not have a great enough risk of progression to warrant EET, and patients with normal p53 may not have enough risk to warrant infrequent surveillance or cessation of surveillance.

Multiple studies have suggested that p53 IHC can be used in confirming the diagnosis of dysplasia and enhancing the interobserver agreement among patholo-gists, a major issue in the evaluation of patients with BE related neoplasia. ${}^{178-181}$ Studies have shown that the use of p53 IHC significantly improved diagnostic concordance for dysplasia among pathologists. ${}^{167,181}$ A study that included 51 pathologists from more than 20 countries investigated BE concordance rates and pathologist features predictive of diagnostic concordance using 55 digitized biopsies across the diagnostic spectrum, before and after viewing matched p53 labeling. Addition of matched p53-labeled slides

November 2025

$$\text{AGA Clinical Practice Guideline on Surveillance of BE 1213}$$

 Figure 7. Forest plots for pooled sensitivity and specificity of p53 assessment among patients with BE undergoing sur-veillance endoscopy.(A) NDBE, sensitivity and specificity.(B)BE with IND/LGD combined, sensitivity and specificity.(C)BE with IND, sensitivity and specificity.(D) BE with LGD,sensitivity and specificity.

$\text{improved diagnostic concordance(NDBE: 78.8\% vs 83.8\%,}$IND: 22.9% vs 23.5%, LGD: 41.6% vs 46%, HGD: 71.1% vs 70.7\%, and LGD/HGD: 77.5\% vs 79.3\%) and reduced the number of major overinterpretations and under-interpretations. ${}^{167}$ In addition, p53 IHC may also help clarify the presence or absence of dysplasia in cases initially diagnosed as IND by reducing the number of pa-tients diagnosed with IND(>50% reclassified as NDBE). ${}^{146,181}$ However, progressors to HGD/EAC include patients with histologic LGD but normal p53, so it is not clear that p53 should be used to alter the histologic interpretation of LGD, which still carries independent information.

What is the role of TissueCypher testing in patients with Barrett's esophagus undergoing endoscopic surveillance?Is the TissueCypher test superior to the grade of dysplasia in predicting progression in patients with Barrett's esophagus?

Recommendation 6: In patients diagnosed with NDBE, BE with IND or BE with LGD, the AGA makes no recommendation for or against the routine use of TissueCypher testing as an adjunct test to histopathology.(Knowledge gap)

TissueCypher, a tissue systems pathology assay, is an innovative approach that uses systems biology, viewing tissue as a system comprising multiple compartments that can be analyzed quantitatively for genetic, immunologic,vascular, and morphologic features relevant to progression from BE to EAC. ${}^{164}$

Summary of the Evidence

Evidence informing the recommendation for the use of TissueCypher(also known as Tissue Systems Pathology-9 test, Castle Biosciences, Friendswood, TX) as an adjunct test during BE surveillance comes from 5 studies(1 cohort and 4 case-control studies). We iden-tified 2 individual-level data SR/MAs by Iyer et al ${}^{182}$ and Davison et al ${}^{183}$ that evaluated the risk of progression to HGD/EAC based on TissueCypher risk class(high, inter-mediate, and low). Both MAs included the same 4 case-control and 1 cohort study by Frei et al ${}^{184}$ (the latter included patients from the screening cohort for the Surveillance vs Radiofrequency Ablation[SURF] trial). An updated literature search(August 2017 to January 2025)(Supplementary Table 3) found no additional studies.Baseline pathology in these studies were the following:NDBE(1 study), ${}^{185}$ LGD(1 study), ${}^{184}$ and a combination of NDBE/IND/LGD(3 studies). ${}^{186-188}$ The training set cases and controls used to train the algorithm were excluded from both MAs. Four studies assessed the utility of TissueCypher in predicting incident HGD/EAC(≥1 year after a BE diagnosis) and 1 study assessed the utility of TissueCypher in predicting prevalent HGD/EAC(within 1 year of BE diagnosis). TissueCypher risk class was divided into high, intermediate, and low.

1214 Wani et al

Gastroenterology Vol. 169, Iss. 6

Benefits

The critical outcomes were:(1) predicting progression to HGD/EAC and(2) test performance (Tables 1 and 8,Supplementary Table 9). The MA by Iyer et al ${ }^{182}$ analyzed individual-level data from 552 patients with BE who were tested with TissueCypher (152 progressed to incident HGD/EAC and 400 nonprogressors). Baseline pathology included NDBE (n = 472), IND (n = 32), and LGD (n = 48).In the overall cohort, (1) comparing high-risk vs intermediate/low-risk, TissueCypher had a sensitivity of 38% and specificity of 94\%(OR, 6; 95\% CI, 2.99-12) for progression to HGD/EAC and (2) comparing high/inter-mediate risk vs low risk, TissueCypher had a sensitivity of 55% and specificity of 82\%(OR, 1.58; 95\% CI, 0.8-3.12) for progression to HGD/EAC. In the NDBE subgroup of 472 patients (112 progressors, 360 nonprogressors) (1) high-risk TissueCypher had a sensitivity of 37\% and specificity of 96\%(OR, 13.55; 95\% CI, 4.90-37.43) for progression to HGD/EAC and (2) high/intermediate-risk compared with low-risk category, TissueCypher had a sensitivity of 52%and specificity of 85\%(OR, 95\% CI, 3.73-9.52) for pro-gression to HGD/EAC. In the LGD subgroup of 48 patients(31 progressors, 17 nonprogressors), high/intermediate risk, TissueCypher had a sensitivity of 71\% and a specificity of 35\%(OR, 1.33; 95\% CI, 0.38-4.71; P=.65) for pro-gression to HGD/EAC. In the IND subgroup of 32 patients (9 progressors, 23 nonprogressors), high/intermediate risk, TissueCypher had a sensitivity of 33\% and a specificity of 74\%,(OR, 1.42; 95\% CI, 0.7-2.57; P=.68) for progression

to HDG/EAC(Table 8). Data were not available for high risk only among the subgroup of patients with LGD and IND. In a separate study from Davison et al ${ }^{183}$ (40 prevalent pro-gressors, and 509 nonprogressors), TissueCypher had a sensitivity of 77.5% in detecting missed prevalent cases of HDG/EAC. Patients who scored high or intermediate risk were more likely to harbor prevalent HDG/EAC (HR, 30.9;95% CI, 13.4-87.6 and HR, 7.6; 95% CI, 2.9-20.8,respectively).

Harms

No studies directly addressed harms associated with the use of TissueCypher testing. As described above for high/intermediate risk TissueCypher, results compared with the low-risk category in patients with NDBE yielded a sensi-tivity of 52% and specificity of 85%. Using the 0.6% HGD/EAC progression rates per year, among 1000 NDBE there were 3 TP results, 3 FN results, 845 TN results, and 149 FP results(Table 8). Calculated PPV and NPV over 5 years was 9.8% and 98.3%, respectively. Lastly, for IND(n=48)and LGD(n= 32), the numbers of participants in the studies were very low, making the pooled sensitivity and specificity estimates very imprecise. Similar to the discus-sion regarding the use of p53 testing, the author panel discussed the following downstream consequences of FP results; unnecessary endoscopies and possible EET and potential AEs without any added benefit, potential financial and psychological burden, and unnecessary health care costs and resource overutilization. The FNs could result in

Table 8. Grading of Recommendations Assessment, Development and Evaluation Evidence Profile for Population,Intervention, Comparison, Outcome Question: Role of Biomarkers (TissueCypher) in Predicting Progression in Patients With Barrett's Esophagus Undergoing Surveillance Endoscopy

|Sample site|NDBE|IND|LGD| |---|---|---|---| |Sensitivity of intermediate or high-risk score (95% CI)|0.52 (0.43-0.61)|0.66 (0.58-0.74)|| |Specificity of low-risk score (95% CI)|0.85 (0.78-0.92)|0.76 (0.69-0.83)|| |Assumed risk of progression to EAC over 1 yr, %|0.6|1.5|1.73| |Outcome,effer per 1000 patients tested|||| |TPs (patients with int/high TC score who progress to EAC)|3 (3-4)|10 (9-11)|11 (10-13)| |FNs (patients with low TC score, but progress to EAC)|3 (2-3)|5 (4-6)|6 (4-7)| |No. of studies|5||5| |No. of patients|112||40| |Certainty of evidence|LOW|④000|Very LOWa,b| |TNs (patients low TC score and did not progress to EAC)|845 (775-914)|749 (680-818)|747 (678-816)| |FPs (patients with int/high TC score, but did not progress to EAC)|149 (80-219)|240 (170-310)|236 (167-305)| |No. of studies|5|5|5| |No. of patients|360|40|| |Certainty of evidence|④000 Very LOWa,b|④000 Very LOWa,b||

NOTE. This table is based on applying the sensitivity and specificity estimates to calculate TPs, FPs, TNs, and FNs in a hypothetical population of 1000 individuals.

aQUADUS 2 tool was used for assessing risk of bias and there were issues in multiple domains:(1) patient selection: case-control study design was used in more than one-half of the studies, not all of the studies stratify for baseline pathology;(2)index test: there were concerns regarding applicability if the index test because there was significant variability in inter-pretation of the test results;(3) reference test was standard biopsy and progression to EAC, but the follow-up period and therefore interpretation varies between the studies.

${}^{b}$ The FP range crosses the clinical threshold of 20%(200/1000).

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1215

the potential for false reassurance, leading to extending surveillance intervals, placing patients at risk for disease progression and poor overall outcomes.

Certainty in Evidence of Effects

The overall certainty in the evidence for TissueCypher was very low (Table 8, Supplementary Table 9). Using the QUADAS-2 tool, there were issues in multiple domains leading to very serious risk of bias and very serious imprecision (the number of patients with IND and LGD was very small). Many of the same limitations to the use of p53 staining were identified with the use of TissueCypher testing (Table 8). The author panel also acknowledged the lack of robust studies on patient preferences, cost-effectiveness, and how TissueCypher results should impact decision making in clinical practice. ${ }^{175,189}$

Discussion

The guideline panel made no recommendation for or against the use of TissueCypher testing and identified the role of TissueCypher testing for risk stratification among patients with BE as a knowledge gap. TissueCypher is a tissue systems pathology assay that can be analyzed quantitatively for genetic, immunologic, vascular, and morphologic features relevant to progression from BE to EAC. ${ }^{164}$ This test is performed on formalin-fixed, paraffin-embedded biopsy specimens that demonstrate BE and uses a spatialomics-based, multiplexed fluorescent imaging platform to automatically and objectively quantify 9 protein-based biomarkers, nuclear morphology, and tissue architecture. The quantitative image analysis is linked to a risk-prediction algorithm that integrates 15 quantitative image analysis features to produce a risk score ranging from 0 to 10, which then classifies patients into high, in-termediate, and low risk for progression to HGD/EAC within 5 years. ${ }^{186,187,190,191}$ TissueCypher has the advan-tage of being an objective test using a locked, automated assay algorithm, but the panel could not recommend its routine use due to suboptimal diagnostic performance, high cost, and lack of prospective validation data. Its test char-acteristics are similar to p53 when combining high and intermediate risk score, as suggested by the company(albeit without direct comparison with p53 within the same cohorts). Ongoing studies, including a multicenter RCT, will help determine whether TissueCypher can accurately pre-dict progression in patients with BE and LGD and in turn benefit from EET. ${ }^{192}$

<BeginOfPara>Although the panel could not make a recommendation for routine use of these biomarkers for risk stratification in patients with BE undergoing surveillance, it should be noted that the panel does not recommend against the use of these biomarkers, given that these tests may be better in some instances than histopathologic diagnosis and may provide benefit in a subset of patients with BE. Future research needs to focus on validation of biomarkers in prospective trials (ideally RCTs); cost-effectiveness analysis of routine use; automated assessment of p53 expression to reduce subjectivity; and patient values and preferences,

<EndOfPara>and assess critical end points of EAC incidence and mor-tality and ultimately define the cohort likely to benefit the most from these tests. Several of these knowledge gaps are currently being assessed in prospective trials, results of which will impact the recommendations provided in this document. In addition, there are several other novel bio-markers (single panel) that are being investigated currently using biopsy specimens or brushings to improve risk stratification. ${ }^{193,194}$ Studies performing a careful com-parison between the available risk-stratification strategies will be needed. Finally, a judgment must be made for acceptance of an FN biomarker test (if at all) that could increase the risk of missed cancer.

What Do Other Guidelines Say?

These recommendations are consistent with the 2022 American College of Gastroenterology guidelines on diag-nosis and management of BE. ${ }^{50}$ The British Society of Gastroenterology guidelines state that biomarker panels cannot yet be recommended as routine care until evidence from RCTs is available. However, these guidelines suggest that p53 immunostain may improve the diagnostic repro-ducibility of a diagnosis of dysplasia in BE and should be considered as an adjunct to routine clinical diagnosis. ${ }^{51}$Similarly, the European Society of Gastrointestinal Endos-copy guidelines do not recommend the routine use of mo-lecular biomarkers in patients with no evidence of dysplasia. Those guidelines recommend the use of p53 IHC to support reproducibility of dysplasia diagnosis and aid the assessment of atypia of uncertain significance. ${ }^{52}$

Prevention of Progression:Chemoprevention and Antireflux Surgery

What is the role of chemopreventive strategies in the prevention of neoplastic progression in patients with Barrett's esophagus?

Recommendation 7: In adult patients with BE, the AGA suggests the use of daily PPI therapy compared with no PPI therapy for the prevention of neoplastic progression of BE.(Conditional recommendation, low certainty)

Implementation Consideration

● In patients with BE, counsel tobacco cessation and weight loss if overweight or obese.

Summary of the Evidence

Evidence informing the recommendation for the use of PPI therapy vs no PPI to prevent progression to HGD/EAC in patients with BE is derived from RCTs, observational cohort, and case-control studies. A 2020 SR/MA (12 studies, 6 cohort, 6 case-control) evaluated the risk of progression to HGD/EAC in patients on vs off PPI. ${ }^{195}$

1216 Wani et al

Gastroenterology Vol. 169, Iss. 6

An updated search (October 1, 2020 to January 2025) (Supplementary Table 3) identified 355 studies, but no relevant studies were identified for further analysis. The prior SR/MA extracted ORs from individual studies, which used multivariate analyses to control for various con-founders including age, sex, race, smoking, and BE length.There was significant heterogeneity among the studies$\left(I^{2}=78\%\right)$ .

The 2 largest cohort studies included in the prior SR/MA were by Krishnamoorthi et al ${ }^{196}$ and Gaddam et al. ${ }^{197}$ Krishnamoorthi et al $\left(12,373\right.$ patients in a UK research database) analyzed the incidence of EAC among patients on vs off PPIs, adjusting for age, gender, smoking, body mass index, hiatal hernia, type 2 diabetes, and use of various medications. The mean age was 63 years and 63% were male. Absence of dysplasia at study inception was not identified definitively in this study; however, inclusion of patients who had not had an upper endoscopy within 3 years before study date was used as an indicator of absence of dysplasia. Gaddam et al (3635 patients with BE) followed patients undergoing screening/surveillance to assess risk of HGD/EAC adjusting for sex, age, race, BE length,histamine-2 receptor antagonist use, smoking history, and aspirin or nonsteroidal anti-inflammatory drug use. Mean age was 61 years, 88% were male, and mean BE length was 3.5 cm.

Indirect evidence as supportive data informing the recommendations for the use of PPI was derived from the AspECT trial, a multicenter (United Kingdom and Canada)RCT examining twice-daily PPI vs once-daily PPI to prevent progression to HGD/EAC, and PPI with aspirin vs PPI alone ${ }^{198}$ using a $2 \times 2$ factorial design to randomize pa-tients with BE into high-dose (esomeprazole 40 mg twice daily) or low-dose PPI (esomeprazole 20 mg once daily)with or without aspirin (300 mg/325 mg daily) for at least 8 years. Mean age was 59 years, 80\% were male, and mean BE length was 4 cm. Median follow-up was 8.9 years(interquartile range, 8.2-9.8 years), making up 20,095 patient-years of follow-up. The primary composite end point was time to HGD,EAC, or all-cause mortality. In total,2557 patients were recruited between 2005 and 2009,including 1281 in the high-dose PPI group (704 without aspirin, 577 with aspirin) and 1,276 in the low-dose PPI group (705 without aspirin, 571 with aspirin).

Evidence on harms associated with PPI use are derived from 2 RCTS: AsPECT trial, COMPASS, and MANAGE.199,200 In the COMPASS trial, 17,598 participants with stable atherosclerotic disease were enrolled in a double-blind,placebo-controlled trial with a 3 $\times$ 2 factorial design and randomized to receive low-dose rivaroxaban with aspirin, high-dose rivaroxaban alone, or aspirin alone. Among par-ticipants not already taking a PPI(64% of study cohort),individuals were randomized to receive pantoprazole 40 ng or placebo once daily. Although the study's primary outcome was rate of cardiovascular disease, it also assessed safety outcomes, including PPI adverse effects(pneumonia,Clostridioides difficile infection, non-C difficile enteric in-fections, fracture, gastric atrophy, chronic kidney disease,

<EndOfPara>and dementia. The mean age was 67.6 years, 78% were male, and 60% were White European. The median follow-up was 3 years(interquartile range 2.5-3.6 years).

The MANAGE trial, ${ }^{200}$ which used a similar factorial design randomized 556 of 1754 patients to receive dabi-gatran vs placebo, and all patients not taking PPI were also randomized to operrazole 20 mg once daily vs placebo.Although study results regarding PPI outcomes have not yet been published in a peer-reviewed journal, preliminary findings were made available through communication with the study authors. Patients were followed for a mean of 17 months to assess major upper gastrointestinal events and PPI adverse events, including C difficile-associated diar-rhea, diarrhea, community-acquired pneumonia, and fracture.

Benefits

The critical outcomes were: (1) progression to HGD/EAC and (2) progression to EAC (Tables 1 and Supplementary Tables 10 and 11). Other patient-important outcomes were EAC mortality. In the SR/MA by Chen et $^{1195}$ daily PPI use reduced the risk of progression to HGD/EAC compared with no PPI use, with 84 fewer per 1000 (from 44 fewer to 112 fewer) and a pooled OR of 0.47(95% CI, 0.32-0.71).

Indirect data from the AspECT trial comparing high-dose(twice daily) PPI vs low-dose(daily) PPI found 13 fewer HGD/EAC per 1000 (ranging from 9 more to 29 fewer); HGD/EAC occurred in 84 (6.6%) of the twice-daily PPI group vs 100 (7.9%) of the once-daily PPI group (RR,0.84; 95\% CI, 0.63-1.11). For progression to EAC alone, the absolute risk reduction was 1 fewer per 1000 (from 16 more to 12 fewer), but only indirect evidence was available because the comparison was twice-daily PPI with once-daily rather than PPI with no PPI (AspECT). Comparing twice-daily PPI vs daily PPI, EAC occurred in 40 patients(3.1%) in the high-dose PPI group vs 41 patients(3.2%) in the low-dose PPI group (HR, 0.97; 95\% CI, 0.63-1.50)(Supplementary Table 10).

For EAC-related mortality also using the AspECT trial,high-dose PPI vs low-dose PPI was associated with fewer deaths: 8 events (0.6%) vs 12 events (0.9%), respectively(HR, 0.65; 95% CI, 0.27-1.57). Lastly, the AspECT trial also evaluated aspirin plus PPI vs PPI alone: HGD/EAC occurred in 65 of 1138 patients (5.7%) who received aspirin vs 93 of 1142 patients (8.1%) who did not receive aspirin. EAC occurred in 31 patients (2.7%) who received aspirin vs 30 patients (2.6%) who did not receive aspirin (HR, 1.0; 95% CI, 0.62-1.58). EAC-related mortality was not analyzed for this comparison due to too few events.

Harms

The patient-important outcome included SAEs related to PPI use. In the AsPECT trial, SAEs attributed to esomepra-tone were reported using the Common Terminology Criteria for Adverse Events. Thirteen events across both low-dose (4 SAE) and high-dose (9 SAE) PPI groups were attributed to

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1217

the drug, corresponding to a low annual SAE rate of 0.06%.We also evaluated SAEs potentially associated with PPI use in several observational studies. ${}^{199,201-207}$ We restricted our evaluation to select SAEs based on the perceived frequency with which patients may raise related concerns, such as in the case of dementia, and on availability of evidence(Supplementary Figure 4).

Pooled data from the COMPASS and MANAGE trials for PPI compared with placebo showed minimal risks: an in-crease of 1 case of C difficile infection per 1000 (from 0 fewer to 3 more), 1 more pneumonia per 1000 (from 4 fewer to 7 more), and 5 fewer to 4 more fractures per 1000. Comparing the PPI vs placebo groups, the RRs of C difficile infection, fracture, and pneumonia were 2.48 (95% CI, 0.8-7.44), 0.98 (95% CI, 0.81-1.18), and 1.03 (95% CI, 0.88-1.20), respectively (Supplementary Figure 4). The panel acknowledged that both the COMPASS and MANAGE trials had relatively short follow-up (3 years in COMPASS and 1.4 years in MANAGE) limiting conclusions about long-term AEs, which are particularly relevant, given the prolonged use required in patients with BE. A recent meta-analysis of RCTs found no association of cardiovascular events with PPI therapy. ${ }^{208}$

Certainty of the Evidence

The overall certainty in the evidence across the critical outcomes was low (Supplementary Tables 10 and 11).Despite having higher certainty in harms, given the low certainty in the benefits of PPI, the overall certainty in the evidence was low. For progression to HGD/EAC, the evi-dence was very low due to inconsistency, as there were studies demonstrating significant benefit with PPIs,whereas some studies did not show any benefit ( $I^{2}=78 \%$ ).Although most of the studies were case-control in design with concern for residual cofounding, the pooled OR in the prior SR/MA used multivariable-adjusted estimates, and thus, we did not further rate down for risk of bias. As there was direct high-quality evidence for PPI vs no PPI, we largely relied on data evaluating the question of twice-daily PPI vs once-daily PPI and rated down for indirectness on the level of comparison. For adverse events associated with PPI use, we used data from high-quality RCTs that had limited follow-up time, with overall moderate certainty of evidence. Studies addressing the cost-effectiveness of che-mopreventive strategies and patient values or preferences using PPI therapy were not identified.

Discussion

PPIs are commonly prescribed in patients with BE to control GERD symptoms, heal erosive esophagitis, and as a chemopreventive agent to reduce progression to dysplasia and EAC. ${ }^{50}$ High-dose twice-daily PPI therapy is standardly prescribed in BE-related neoplasia patients undergoing EET, details are highlighted in the AGA EET Guideline document. ${ }^{10}$ Current available evidence from observational studies suggests that PPIs may reduce the risk of neoplastic

progression to HGD/EAC. In addition, histamine receptor antagonists have shown no significant benefit. ${ }^{209}$

The panel recognized the potential adverse effects of PPIs that have been reported, but many of these studies are low-quality, retrospective, observational studies limited by confounding and other biases and, most importantly, do not establish causality. ${ }^{210,211}$ In most studies, biases can account for the small but statistically significant associations and these associations have no plausible biologic mechanism of action. Given the rarity of adverse events, the absolute risk to patients is small. ${ }^{211}$ It is recommended that patients with BE are prescribed at least once-daily PPI therapy to reduce the risk of neoplastic progression. Control of reflux symptoms along with healing/prevention of erosive esophagitis should also be accounted for in the dose of prescribed PPI ther-apy. The AGA published a recent Clinical Practice Update on long-term PPI therapy use that focuses on ensuring that the patient has a strong indication for long-term use;when indicated, counseling the patient regarding the rationale, the issues related to published literature on adverse events, and that benefits outweigh the potential risks; using the lowest effective dose; and for patients concerned about a specific adverse event, consider describing the absolute excess risk of the adverse event per patient-year and personalizing the discussion of risk. ${ }^{211}$

Epidemiologic evidence from observational studies and biomarker-based preclinical studies have shown that aspirin and nonsteroidal anti-inflammatory drugs may reduce the risk of EAC by inhibiting the cyclooxygenase pathway. ${ }^{212-217}$ The AspECT chemoprevention study demonstrated that high-dose PPI was superior to low-dose PPI for lengthening the time to reach the combined end point of death from any cause, HGD, or EAC (time ratio,1.27; 95% CI, 1.01-1.58; P=.038). Censoring those with the use of concurrent nonsteroidal anti-inflammatory drugs, aspirin was superior to no aspirin (time ratio,1.29; 95% CI, 1.01-1.66; P=.04). Finally, combining high-dose PPI with aspirin had the strongest effect compared with low-dose PPI without aspirin (time ratio, 1.59; 95\% CI,1.14-2.23; P=.006). Because of the small effect size, low event rates, and unclear impact on EAC incidence (the benefit was largely driven from a reduction in all-cause mortality), the panel made no recommendation regarding the use of aspirin or high-dose PPI therapy to prevent neoplastic progression in patients with BE. Long-term data on adverse events related to high-dose PPI are limited.Thus, a decision to prescribe high-dose PPI therapy to prevent neoplastic progression needs to account for the above factors (net benefit vs harms) along with patient values and preferences. The guideline authors acknowledge that the clinical profile of several patients with BE may require daily aspirin for cardio-protection. Evidence from the AspECT trial showed that participants taking aspirin were more likely to have adverse events than those who were not (RR, 1.17; 95\% CI, 1.02-1.35) and may have been attenuated by the use of high-dose PPI therapy. ${ }^{198,218}$

1218 Wani et al

Gastroenterology Vol.169, Iss. 6

The guideline panel also acknowledged the growing body of literature describing the efficacy of potassium-competitive acid blockers in the management of patients with GERD and discussed in a recent AGA Clinical Practice Update.219 The benefits of potassium-competitive acid blockers over PPIs include more rapid onset with initial dosing, no premeal dosing requirement, less variability in pharmacodynamic effects related to CYP2C19 metabolism,higher potency, and longer duration of effects; however,long-term safety data are lacking, and their role in BE for reflux symptoms, healing of erosive esophagitis and reduction in neoplastic progression in patients undergoing EET, is not known, and finally these medications are costly.Extrapolating the available data from patients with GERD,potassium-competitive acid blockers may be considered in patients with BE who continue to have symptoms or erosive esophagitis on twice-daily PPI therapy.

Implementation Consideration for Recommendation 7

In addition to acid suppression with PPIs, preventing neoplastic progression also includes behavioral modifica-tion. Tobacco use is both reflux-inducing as well as muta-genic,220 increasing the odds of developing BE by 67%. Many risk-prediction models, therefore, include smoking history as a factor when evaluating eligibility for screening for BE.50 Furthermore, tobacco use also has been identified as a risk factor in BE progression to HGD as well as EAC, with an HR of progression that is nearly double among tobacco users compared with nonusers. Likewise, tobacco cessation may reduce BE-associated risks. Previous data demonstrated that among smokers who had quit for>20 years, there was no increased risk of BE.222 As a result, smoking cessation should be encouraged among patients undergoing BE sur-veillance. Although quantifiable evidence for risk reduction with tobacco cessation among patients with BE is lacking,encouraging such behavior has pleiotropic health benefits and plausible direct advantages for BE specifically. Recent GERD guidelines suggest avoidance of tobacco among those with GERD.130 Likewise, screening for tobacco use and providing tobacco cessation counseling for all patients is an accepted quality measure and was recently adopted within a gastroenterology core measure set.223

Central adiposity is another important modifiable risk factor for BE. Obesity increases GERD risk24 and weight loss is recommended for overweight and obese individuals to control GERD symptoms.130 Although body mass index alone is imperfect for measuring risks for BE, 225 markers of insulin resistance are associated with increased risk for progression to EAC.226 A recent SR/MA showed in patients with NDBE and/or LGD, increasing body mass index was associated with an increased risk of progression to HGD and EAC in a dose-dependent manner. 58 Given that behavioral modification can lead to weight loss and symptom improvement in GERD, 227 it is appropriate to encourage weight loss among overweight or obese patients

with BE to reduce symptomatic GERD and reduce the risk of progression.

What is the role of antireflux surgery compared with medical management using acid suppressive therapy in the prevention of progression in patients with Barrett's esophagus?

Recommendation 8: In patients with BE, the AGA suggests use of PPIs over surgery for the prevention of neoplastic progression to HGD or EAC.(Conditional recommendation, low certainty of evidence)

The critical outcomes that informed the benefits for this PICO question were:(1) progression to EAC and(2) pro-gression to HGD/EAC(Table 1 and Supplementary Table 12). Other patient-important outcomes we evalu-ated included EAC mortality.

Summary of the Evidence

Evidence informing the recommendation came from 5 observational cohort studies and 1 RCT. A prior SR/MA (4 cohort results, 1 RCT) by Wilson et al ${ }^{228}$ evaluated the risk of progression to HGD/EAC in patients who underwent antireflux surgery vs medical management with PPI/histamine-2 receptor antagonist (search through February 2021). Disease progression was defined as progression of intestinal metaplasia to LGD, HGD, or EAC or by progres-sion of LGD to HGD and EAC $\left(I^{2}=0\%\right)$ . We updated the analysis with 1 additional cohort study by Akerstrm et al ${ }^{229}$ that reported on neoplastic progression to HGD/EAC in patients with BE.

A single RCT, included in the aforementioned SR/MA,compared the impact of surgical vs medical management on progression in BE. ${}^{230}$ This RCT (101 patients) randomized 43 patients to receive medical treatment (initially histamine-2 receptor antagonists, and omeprazole from 1992 onward) and 58 patients to undergo antireflux surgery(primarily Nissen fundoplication and Collis Nissen proced-ure in 2 patients). Median age was 50 years (medical ther-apy group) vs 43 years (surgical treatment group), 71% of patients were male, and median follow-up was 5 years. The largest Nordic population-based observational study by Akerström et al ${}^{229}$ analyzed 33,939 patients with BE from multiple national registries. Of the cohort, 542 patients(1.6%) underwent antireflux surgery (fundoplication) and 33,397 patients underwent medical management. In the antireflux surgery cohort, 67% were men, mean age was 64 years, and mean follow-up was 4.9 years. In the medical management group, 69% were men, mean age was 53 years,and mean follow-up was 11 years. Although no formal SR/MA was conducted, indirect evidence regarding harms was obtained from a prior SR/MA examining outcomes after antireflux surgery vs medical management of GERD. ${}^{231,232}$

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1219

Benefits

For progression to EAC, pooled data from 6 studies (5 studies identified from the SR/MA and the cohort study by Akerström et al, showed that antireflux surgery vs medical management showed no clear benefit. ${ }^{228,229}$ Of 33,528 pa-tients who underwent medical management, 438 (1.3%)progressed to EAC vs 14 of 765 patients (1.8%) who un-derwent antireflux surgery (RR, 1.33; 95\% CI, 0.29-6.16); 4 more EAC cases per 1000 in the medical management group (from 9 fewer to 67 more) (Supplementary Figure 5).The largest cohort showed that antireflux surgery did not reduce the risk of EAC among patients with BE followed up for up to 32 years (adjusted HR, 1.9; 95\% CI, 1.1-3.5). ${ }^{229}$

To assess for progression to HGD/EAC, data from the same 6 studies were used. ${ }^{228,229}$ Of 33,528 patients who underwent medical management, 443 (1.3%) progressed to HGD/EAC vs 19 of 765 patients (2.5%) who underwent antireflux surgery (RR, 0.86; 95% CI, 0.30-2.41)(Supplementary Figure 5). The absolute risk reduction was 2 fewer per 1000 (from 9 fewer to 19 more). There was no available evidence for the impact of antireflux surgery vs medical management in patients with BE on EAC mortality.

Harms

To evaluate SAEs related to antireflux surgery and PPIs,we evaluated data from an existing review focusing on the most common major adverse events after fundoplication, as well as rates of SAEs in existing RCTs comparing antireflux surgery vs medical management for the treatment of acid reflux or BE. Adverse events related to PPI therapy are reviewed in the Prevention of Progression: Chemopreven-tion and Antireflux Surgery section.

A comprehensive qualitative review by Yadlapati et al ${ }^{233}$ summarized short and long-term outcomes of laparoscopic fundoplication for GERD, drawing from large population-based cohorts and randomized trials. In a Swedish cohort of 2655 patients, the 30-day all-cause mortality rate was 0.1\% and 4.1\% had early complica-tions (within 30 days of surgery), which included infection(1.1\%), bleeding (0.9\%), and esophageal perforation(0.9\%). ${ }^{234}$ Reflux occurred in 17.7\%, and 16.4\% of the cohort required secondary antireflux surgery. An older review study evaluating outcomes after 10,489 primary fundoplication procedures found similar complication rates, including mortality rate of 0.08\%, perforation in 0.8\%, and wound infection in $0.1 \% .{ }^{235}$ The mortality esti-mate of $0.1 \%-0.2 \%$ in the qualitative review is further supported by a more recent population-based cohort study of 5 Nordic countries, which evaluated 26,193 patients who underwent primary laparoscopic antireflux surgery and found 90-day postoperative mortality of $0.13 \%$ . ${ }^{236}$

We also evaluated adverse events from multiple RCTs comparing antireflux surgery vs medical management for the treatment of acid reflux or BE. No deaths within 30 days of surgery were reported in the available RCTs.230,237-240 The SOPRAN study randomized 310 patients with chronic GERD to PPI vs open antireflux surgery. ${ }^{239}$ with an extended follow up to 14 years, found similar rate of SAEs between both groups

(antireflux surgery: 0.1\% per patient-year; PPI: 0.1\% per pa-tient-year). ${ }^{241}$ The LOTUS RCT randomized 554 patients with chronic GERD to acid suppression vs laparoscopic antireflux surgery. ${ }^{238}$ At 5-year follow-up of 372 patients, SAEs were reported by 28.6\% of patients who underwent antireflux sur-gery vs $24.1 \%$ of the PPI group. At 7-year follow-up, SAEs remained similar between the antireflux surgery $(0.2 \%$ per patient-year) and PPI $(0.2 \%$ per patient-year) groups. ${ }^{241}$

Certainty of the Evidence

The overall certainty in the evidence was low(Supplementary Table 12). For progression to HGD/EAC and EAC the evidence was low. We largely relied on data from nonrandomized studies included in an SR/MA. There was a concern for selection bias because of lack of consecutive patient inclusion and concerns about baseline comparability of the groups. However, the larger studies were without concerns, so we did not rate down for risk of bias. Studies addressing cost-effectiveness and patient values or prefer-ences for these 2 strategies to reduce neoplastic progression in patients with BE were not identified.

Discussion

Surgical antireflux procedures for preventing neoplastic progression are appealing, given that they restore the gastroesophageal anatomic and physiological reflux barrier and reduce gastric contents(acid and bile) from reaching the esophagus.242,243 In addition, antireflux surgery is highly effective in reducing esophageal acid exposure,reducing gastroesophageal reflux episodes, healing esoph-agitis, and reducing symptoms associated with reflux. However, the existing evidence does not consistently show superiority of antireflux surgery over medical management(typically using PPIs) in reducing neoplastic progression in patients with BE. Most of the studies that helped inform this recommendation were from single centers with small sample size, short follow-up, moderate heterogeneity, and potential biases due to patients with more severe disease more likely to undergo antireflux surgery, and confounding factors, such as BE length. In addition, other risk factors,such as tobacco smoking and body mass index-factors that could influence the decision to proceed with antireflux surgery-were not available in all studies.229 Data on medication use were not available for most studies and compliance was not routinely monitored. The failure rate of antireflux surgery is 20%-30% and recurrence of GERD symptoms at 5 years is reported in approximately 15% of patients who have undergone antireflux surgery.234,238,244 An SR/MA of antireflux surgery in patients with GERD re-ported that 28% randomized to surgical intervention still reported PPI use.231 The decision to recommend PPIs over antireflux surgery was also driven by the overall low risk of progression in patients with NDBE, making it difficult to justify the risks associated with antireflux surgery high-lighted above. The role of antireflux surgery in patients with BE-related neoplasia undergoing EET is highlighted in the AGA EET Guidelines document. ${ }^{10}$ Finally, future studies need to better define whether there are patients with BE

1220 Wani et al

Gastroenterology Vol.169, Iss.6

Table 9.Quality Measures and Other Proposed Quality Metrics in Barrett's Eophagus Surveillance

|Quality metrics|Type of metric|Rationale| |---|---|---| |Sampling technique using the Seattle biopsy protocol|Process|Adequate sampling increases neoplasia detection compared with random biopsy sampling| |Patients with NDBE undergo surveillance endoscopy no sooner than 3 yr|Process|Reduction in overutilization of endoscopy| |Proposed quality metrics||| |Performing surveillance endoscopy using a combination of high-definition WLE with CE (dye-based or virtual CE)|Process|Combination of high-definition WLE and CE associated with higher neoplasia rates compared with WLE alone| |NDR defined by percent detection of dysplasia on index endoscopy for screening for BE|Process|Reflects the overall quality of the endoscopic examination| |Barrett's inspection time that suggests inspection time of 1 min/cm of circumferential BE|Process|Reflects the quality of the endoscopic examination and may lead to increased NDR| |Documentation of the extent of suspected or confirmed BE using the Prague criteria|Process|Consistent reporting that facilitates evidence- based decision making, improves communication and patient monitoring| |Tracking and monitoring rates of PEEC and PEEN|Outcome|Reflects the overall performance of endoscopy as most cases of PEEC and PEEN are due to missed lesions|

who are most likely to benefit from antireflux procedures to reduce the risk of neoplastic progression.

Quality Assurance in Barrett's Esophagus Surveillance

This section outlines proposed quality assurance frame-work to improve outcomes for BE surveillance in the context of value-based and quality focused health care(Table 9).These can be measured at an individual endoscopist level or across a facility, institution or system level by comparing performance to a benchmark. In this framework, non-adherence usually reflects suboptimal care. In this section, the term quality measures refers to metrics that are fully specified,tested, and validated and in quality payment programs, the term quality indicators refers to metrics derived from guide-line documents, best practice statements, and consensus recommendations that do not necessarily have a specified numerator and denominator nor are they included in quality payment programs. ${}^{245,246}$ Two quality measures that have received the most attention are frequency of surveillance endoscopies at appropriate intervals, that is, avoidance of overuse by repeating a surveillance endoscopy no sooner than 3-5 years in patients with NDBE and sampling the BE segment using the Seattle biopsy protocol. ${}^{232,247}$ These measures have been developed as fully specified quality measures according to the established AGA conception to implementation pathway(https://gastro.org/practice-resources/quality-and-performance-measures/), which requires a strong recommendation based on moderate- or high-quality evi-dence. ${}^{245}$ Several observational studies, including data from a national benchmarking registry (GIuulC Registry) have demonstrated suboptimal adherence and wide variability among endoscopists and participating.

This section provides an overview of other proposed quality metrics in BE surveillance. Similar to adenoma

detection rate, a quality metric used to measure colorectal cancer screening can are and associated with the highest penetrance in clinical gastroenterology practice, neoplasia detection rate (NDR) has been proposed to monitor quality in BE screening and surveillance. NDR is defined as the prevalence of HGD or EAC (numerator) within BE during the index screening endoscopy (denominator). A meta-analysis reported a pooled HGD/EAC prevalence of 7%(95 CI, 4%-10%) and proposed 4% as the NDR threshold on screening endoscopy. ${ }^{252}$ Rates of dysplasia detection rate (DDR), defined as LGD or HGD, were reported using the GIQUC Registry among patients undergoing surveil-lance of NDBE and recent studies have also highlighted the wide variability among endoscopists and participating sites for DDR in patients with NDBE.115,249

Similar to the phenomenon of postcolonoscopy colorectal cancer, there is a growing body of literature describing BE-associated HGD and EAC before the next recommended endoscopic evaluation after an endoscopy that was ostensibly negative for HGD/EAC.253 To address this issue in BE endos-copy quality, an AGA Clinical Practice Update and an interna-tional expert panel introduced the concepts of PEEC and PEEN.PEEC is the preferred term for EAC and PEEN is the preferred term for HGD or EAC detected before the next recommended surveillance endoscopy in a patient with prior NDBE, provided it occurs between 6 months and 3 years after a screening or surveillance endoscopy.111,253 It is believed that most cases of PEEC/PEEN may be attributed to missed HGD/EAC, and the remaining cases may be attributed to rapidly progressive neoplasia due to accelerated pathways of neoplasia.111,254 Several factors may contribute to missed neoplasia at endos-copy, including suboptimal adherence to the Seattle biopsy protocol, limited mucosal sampling, inadequate time spent inspecting the BE segment, limited knowledge in the recogni-tion of subtle findings of early neoplasia, and interobserver

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1221

variability among pathologists for the histologic classification of dysplasia. ${ }^{1,111,253}$ Observational studies, including a recent population-based cohort study, have provided contemporary estimates of PEEC and PEEN. ${ }^{1,48}$ An SR/MA reported that the proportion of EACs that were PEEC was $21 \%$ (95% CI, $13 \%$ -31\%) and the proportion of HGD/EAC that were PEEN was $26 \%$ (95% CI, $19 \%$ -34\%); outcomes of PEEC/PEEN defined as HGD/EAC detected within the first year after an index endoscopy that demonstrated NDBE, IND, or LGD. ${ }^{48}$ Similarly, a recent population-based study conducted in Denmark, Finland, and Sweden (the Nordic Barrett's Esoph-agus Study) that included 20,588 patients with newly diag-nosed BE and showed that 23.5\% of EACs were categorized as PEEC and 17.2\% of HGD/EACs were categorized as PEEN. ${ }^{1}$ Several sensitivity analyses that varied time intervals for occurrence of PEEC/PEEN demonstrated similar results. ${ }^{1}$ In addition, a time-trend analysis demonstrated rising incidence rates for PEEC/PEEN. ${ }^{1}$ These findings suggest suboptimal performance among current screening and surveillance prac-tices and provide improvement opportunities for early neoplasia detection in BE. The International Expert Panel on PEEC and PEEN and the AGA Clinical Practice Update sug-gested that services and individual endoscopists review PEEC/PEEN cases periodically and identify areas for improvement,provided a construct for categorizing PEEC/PEEN cases ac-cording to their most plausible explanations and best practice advice to reduce PEEC/PEEN. ${ }^{111}$ Recent data from the Nordic Barrett's Esophagus Study demonstrated an overall inverse association between facility NDR and the risk of PEEC/PEEN adding to the validity of NDR and PEEC/PEEN as potential quality measures in BE. ${ }^{255}$

<BeginOfPara>Although a number of additional quality metrics specific to BE have been proposed, incorporation into value-based care plans has been limited for several reasons, including lack of formal and validated testing. Several of these proposed

<EndOfPara>quality metrics do not have established thresholds that are adjusted for risk factors. Most of these proposed quality metrics are based on weak evidence or consensus expert opinions and most importantly, not tied to clinical outcomes.Monitoring of BE-specific quality metrics, such as use of the Seattle biopsy protocol, Barrett's inspection time of at least 1 min/cm of columnar-lined epithelium, appropriate endo-scopic surveillance intervals, and monitoring of NDR/DDR must correlate with important clinical outcomes, such as PEEC and, ideally, EAC mortality, as well as drive performance improvement to be considered a high-value quality mea-sure.246 other challenges include operationalizing these,especially NDR/DDR and PEEC/PEEN, given the variable in-dividual number of endoscopies performed for BE per endoscopist and the overall low incidence of EAC. Future ef-forts will need to focus on how these proposed metrics could be calculated in an automated fashion and establish minimum standards using risk-adjusted strategies.

Knowledge Gaps

This evidence review identified several important knowledge gaps that need to be addressed in future studies and highlighted in Table 10.

Plans for Updating

Considerable resources are expended for the develop-ment of guidelines, and keeping guidelines up to date is a challenging process. Future update of this guideline will depend on the availability of new evidence on the existing and new interventions. We hope to incorporate the ad-vances in the technological platforms and models of guideline development in the future updates without duplication or reproduction of the current guideline document.

Table 10.Knowledge Gaps

Knowledge gaps

The role of surveillance needs to be assessed in study designs that minimize contamination and is adequately powered to assess differences in EAC mortality.

Future studies need to identify risk-stratification tools to better guide endoscopic surveillance intervals, risk of progression, and discontinuation of surveillance.

Future studies need to address the role of biomarker risk stratification using nonendoscopic cell collection devices to guide endoscopic surveillance for early cancer detection and efficient resource allocation.

The role of Al in enhancing dysplasia and early cancer detection among patients with BE undergoing endoscopic surveillance examinations needs to be defined(ideally in RCTs).

The role of advanced sampling techniques, such as WATS-3D as an adjunctive and substitutive technique to a structured biopsy protocol needs to be addressed in future RCTs.

<BeginOfPara>Validation of biomarkers needs to be performed in prospective trials(ideally in RCTs) that assess critical end points of EAC incidence and mortality.

Studies on clinical utility of p53 (using consistent terminology, analysis, and interpretation) as a diagnostic tool to improve interrater reliability for dysplasia and as a prognostic tool with longer follow-up (3 years/5 years) for predicting progression to neoplasia.

The role of potassium-competitive acid blockers in patients with BE needs to be evaluated in future studies for the outcomes of reflux control,healing of erosive esophagitis, reduction in neoplastic progression and outcomes in patients undergoing EET.

Future research needs to define patients with BE who are most likely to benefit from antireflux procedures to prevent neoplastic progression.

1222 Wani et al

Gastroenterology Vol.169, Iss. 6

Supplementary Material

Note: To access the supplementary material accompanying this article, visit the online version of Gastroenterology at www.gastrojournal.org, and at https://doi.org/10.1053/j.gastro.2025.09.012.

References

1.Wani S, Holmberg D, Santoni G, et al. Magnitude and time-trends of post-endoscopy esophageal adenocar-cinoma and post-endoscopy esophageal neoplasia in a population-based cohort study: the Nordic Barrett's Esophagus Study. Gastroenterology 2023;165:909-919.e13.

2.Rodriguez GM, DePuy D, Aljehani M, et al. Trends in epidemiology of esophageal cancer in the US, 1975-2018. JAMA Netw Open 2023;6:e2329497.

3.National Cancer Institute. Surveillance, Epidemiology and End Results(SEER) Program. Cancer Statistics Explorer Network: Adenocarcinoma of the Esophagus Survival.Available at: https://seer.cancer.gov/statistics-network/explorer/application.html?site=600$data_type=4&graph_type=5&compareBy= age_range&chk_age_range_1= 1&series= 9&sex= 1&race= 1&stage=101&advopt_precision=1=advopt_show_ci= on&hdn_view=0.Accessed January 16,2023.

4.Curtius K, Rubenstein JH, Chak A, et al. Computational modelling suggests that Barrett's oesophagus may be the precursor of all oesophageal adenocarcinomas. Gut 2020;70:1435-1440.

5.Black EL, Ococks E, Devonshire G, et al. Understanding the malignant potential of gastric metaplasia of the oesophagus and its relevance to Barrett's oesophagus surveillance: individual-level data analysis. Gut 2024;73:729-740.

6.Nowicki-Osuch K, Zhuang L, Jammula S, et al. Molec-ular phenotyping reveals the identity of Barrett's esophagus and its malignant transition. Science 2021;373:760-767.

7.Pech O, Bollsweiler E, Manner H, et al. Comparison between endoscopic and surgical resection of mucosal esophageal adenocarcinoma in Barrett's esophagus at two high-volume centers. Ann Surg 2011;254:67-72.

8.Schembre DB, Huang JL, Lin OS, et al. Treatment of Barrett's esophagus with early neoplasia: a comparison of endoscopic therapy and esophagectomy. Gastro-intest Endosc 2008;67:595-601.

9.Zehetner J, DeMeester SR, Hagen JA, et al. Endoscopic resection and ablation versus esophagectomy for high-grade dysplasia and intramuscosal adenocarcinoma.J Thorac Cardiovasc Surg 2011;141:39-47.

10.Rubenstein JH, Sawas T, Wani S, et al. AGA clinical practice guideline on endoscopic eradication therapy of Barrett's esophagus and related neoplasia. Gastroen-terology 2024;166:1020-1055.

11.Schunemann HJ, Oxman AD, Brozek J, et al. Grading quality of evidence and strength of recommendations for diagnostic tests and strategies. BMJ 2008;336:1106-1110.

12.Graham R, Mancher M, Wolman DM, et al. Institute of Medicine Committee on standards for developing trustworthy clinical practice guidelines: clinical practice guidelines we can trust. National Academies Press,2011.

13.Schunemann HJ, Al-Ansary LA, Forland F, et al.Guidelines International Network: principles for disclo-sure of interests and management of conflicts in guidelines. Ann Intern Med 2015;163:548-553.

14.Schunemann HJ, Wiercioch W, Etxeandia I, et al.Guidelines 2.0: systematic development of a compre-hensive checklist for a successful guideline enterprise.CMAJ 2014;186:E123-E142.

15.Spechler SJ, Sharma P, Souza RF, et al. American Gastroenterological Association medical position statement on the management of Barrett's esophagus.Gastroenterology 2011;140:1084-1091.

16.Wani S, Rubenstein JH, Vieth M, et al. Diagnosis and management of low-grade dysplasia in Barrett's esophagus: expert review from the Clinical Practice Updates Committee of the American Gastroenterolog-ical Association. Gastroenterology 2016;151:822-835.

17.Draganov PV, Wang AY, Othman MO, et al. AGA Insti-tute clinical practice update: endoscopic submucosal dissection in the United States. Clin Gastroenterol Hepatol 2019;17:16-25.e1.

18.Sharma P, Shaheen NJ, Katzka D, et al. AGA Clinical Practice Update on endoscopic treatment of Barrett's esophagus with dysplasia and/or early cancer: expert review. Gastroenterology 2020;158:760-769.

19.Muthusamy VR, Wani S, Gyawali CP, et al. AGA clinical practice update on new technology and innovation for surveillance and screening in Barrett's esophagus:expert review. Clin Gastroenterol Hepatol 2022;20:2696-2706.e1.

20.Ouzzani M, Hammady H, Fedorowicz Z, et al.Rayyan-a web and mobile app for systematic reviews.Syst Rev 2016;5:210.

21.Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BSM J 2021;372:n71.

22.Whiting PF, Rutjes AW, Westwood ME, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011;155:529-536.

23.Wells GAS B, O'Connell D, Peterson J, et al. The Newcastle-Ottawa Scale(NOS) for assessing the quality of non-randomized studies in meta-analyses. Ottawa Hospi-tal Research Institute. Available at: https://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. Accessed September 4,2025.

24.Guyatt G, Oxman AD, Akl EA, et al. GRADE guidelines:1. Introduction-GRADE evidence profiles and sum-mary of findings tables. J Clin Epidemiol 2011;64:383-394.

25.Alonso-Coello P, Schünemann HJ, Moberg J, et al.GRADE Evidence to Decision(EtD) frameworks: a sys-tematic and transparent approach to making well informed healthcare choices. 1: Introduction. BMJ 2016;353:i2016.

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1223

26. Codipilly DC, Chandar AK, Singh S, et al. The effect of endoscopic surveillance in patients with Barrett's esophagus: a systematic review and meta-analysis.Gastroenterology 2018;154:2068-2086.e5.

27.Old O, Jankowski J, Attwood S, et al. Barrett's Oesophagus Surveillance Versus Endoscopy at Need Study (BOSS): a randomized controlled trial. Gastroen-terology 2025;169:1232-1242.

28.El-Serag HB, Naik AD, Duan Z, et al. Surveillance endoscopy is associated with improved outcomes of oesophageal adenocarcinoma detected in patients with Barrett's oesophagus. Gut 2016;65:1252-1260.

29.Kim HI, Yoon JY, Kwak MS, et al. Gastrointestinal and nongastrointestinal complications of esophagus-droluodenoscopy and colonoscopy in the real world: a nationwide standard cohort using the common data model database. Gut Liver 2021;15:569-578.

30.Wang P, Xu T, Ngamruenghong P, et al. Rates of infection after colonoscopy and asophogasogans-troduodenoscopy in ambulatory surgery centres in the USA. Gut 2018;67:1626-1636.

31.Surveillance, Epidemiology, and Eng Results Program.Cancer stat facts: esophageal cancer. Available at: https://seer.cancer.gov/stafacts/html/esoph.html.Accessed October 1, 2024.

32.Sawas T, Jones AR, Alsawas R, et al. Overall and cause-specific mortality in patients with Barrett's esophagus: a systematic review and 2025. https://doi.org/10.14309/ajg.000000000003456.

33.Sikkema M, de Jonge PJ, Steyerberg EW, et al. Risk of esophageal adenocarcinoma and mortality in patients with Barrett's esophagus: a systematic review and meta-analysis. Clin Gastroenterol. Hepatol 2010;8:235-244; quiz e32.

34.Desai TK, Krishnan K, Samala N, et al.The incidence of oesophageal adenocarcinoma in non-dysplastic Bar-rett's oesophagus: a meta-analysis. Gut 2012;61:970-976.

35.Kroep S, Lansdorp-Vogelaar I, Rubenstein JH, et al. An accurate cancer incidence in Barrett's esophagus: a best estimate using published data and modeling.Gastroenterology 2015;149:577-585.e4; quiz e14-e15.

36.Klaver E, Bureo Gonzalez A, Mostafavi N, et al. Barrett's esophagus surveillance in a prospective Dutch multi-center community-based cohort of 985 patients dem-onstrates low risk of neoplastic progression. United European Gastroenterol J 2021;9:929-937.

37.Deidda M, Old O, Jankowski J, et al. Cost-effectiveness of regular surveillance versus endoscopy at need for patients with Barrett's esophagus: economic evaluation alongside the Barrett's Oesophagus Surveillance Study(BOSS) randomized controlled trial. Gastroenterology 2025;169:1243-1251.

38.Rubenstein JH, Indomi JM. Cost-effectiveness of screening, surveillance, and endoscopic eradication ther-apies for managing the burden of esophageal adenocar-cinoma. Gastrointest Endosc Clin N Am 2021;31:77-90.

39.Kastelein F, van Olphen S, Steyberberg EW, et al. Sur-veillance in patients with long-segment Barrett's

oesophagus: a cost-effectiveness analysis. Gut 2015;64:864-871.

40.Omidvari A-H, Ali A, Hazelton WD, et al. Optimizing man-agement of patients with Barrett's esophagus and low-grade or no dysplasia based on comparative modeling.Clin Gastroenterol Hepatol 2020;18:1961-1969.

41.Sampliner RE. Practice guidelines on the diagnosis,surveillance, and therapy of Barrett's esophagus. The Practice Parameters Committee of the American Col-lege of Gastroenterology. Am J Gastroenterol 1998;93:1028-1032.

42.van Sandick JW, van Lanschot JJ, Kuiken BW, et al.Impact of endoscopic biopsy surveillance of Barrett's oesophagus on pathological stage and clinical outcome of Barrett's carcinoma. Gut 1998;43:216-222.

43.Ferguson MK, Durkin A. Long-term survival after esophagectomy for Barrett's adenocarcinoma in endo-scopically surveyed and nonsurveyed patients.J Gastrointest Surg 2002;6:29-35; discussion 36.

44. Rubenstein JH, Sonnenberg A, Davis J, et al. Effect of a prior endoscopy on outcomes of esophageal adeno-carcinoma among United States veterans. Gastrointest Endosc 2008;68:849-855.

45.Kearney DJ, Crump C, Maynard C, et al. A case-control study of endoscopy and mortality from adenocarcinoma of the esophagus or gastric cardia in persons with GERD. Gastrointest Endosc 2003;57:823-829.

46.Li X, Galipeu PC, Paulson TG, et al. Temporal and spatial evolution of somatic chromosomal alterations: a case-conort study of Barrett's esophagus. Cancer Prev Res(Phila) 2014;7:114-127.

47.Killcoyne S, Fitzgerald RC. Evolution and progression of Barrett's oesophagus to oesophageal cancer. Nat Rev Cancer 2021;21:731-741.

48.Sawas T, Majzoub AM, Haddad J, et al. Magnitude and time-trend analysis of postendoscopy esophageal adenocarcinoma: a systematic review and meta-anal-ysis. Clin Gastroenterol Hepatol 2022;20:e31-e50.

49.Chandrasekar VT, Hamade N, Desai M, et al. Signifi-cantly lower annual rates of neoplastic progression in short- compared to long-segment non-dysplastic Bar-rett's esophagus: a systematic review and meta-anal-ysis. Endoscopy 2019;51:665-672.

50.Shaheen NJ, Falk GW, Iyer PG, et al. Diagnosis and management of Barrett's esophagus: an updated ACG guideline. Am J Gastroenterol 2022;117:559-587.

51.Fitzgerald RC, di Pietro M, Ragunath K, et al. British Society of Gastroenterology guidelines on the diagnosis and management of Barrett's oesophagus. Gut 2014;63:7-42.

52.Weusten B, Bisschops R, Dinis-Ribeiro M, et al. Diag-nosis and management of Barrett esophagus: European Society of Gastroointerinal Endoscopy (ESGE) Guide-line. Endoscopy 2023;55:1124-1146.

53.Krishnamoorthi R, Singh S, Ragunathan K, et al. Factors associated with progression of Barrett's esophagus: a systematic review and meta-analysis. Clin Gastroenterol Hepatol 2018;16:1046-1055.e8.

54.Stephanie M, Nour H, de Sa Ines M, et al. Gender dif-ferences in Barrett's esophagus and progression of

1224 Wani et al

Gastroenterology Vol.169, Iss. 6

disease: a systematic review and meta-analysis. Dis Esophagus 2022;35:10.

55.Zellenrath PA, van Tilburg L, Pouw RE, et al. Neoplastic progression risk in females with Barrett's esophagus: a systematic review and meta-analysis of individual pa-tient data. Clin Gastroenterol Hepatol 2025;23:225-235.e8.

56.Parasa S, Vennalaganti S, Gaddam S, et al. Develop-ment and validation of a model to determine risk of progression of Barrett's esophagus to neoplasia.Gastroenterology 2018;154:1282-1289.e2.

57.Kunzmann AT, Thrift AP, Johnston BT, et al. External validation of a model to determine risk of progression of Barrett's oesophagus to neoplasia. Aliment Pharmacol Ther 2019;49:1274-1281.

58.Ko MT, Thomas T, Holden E, et al. The association between obesity and malignant progression of Barrett's esophagus: a systematic review and dose-response meta-analysis.Clin. Gastroenterol Hepatol 2025;23:726-738.e28.

59.Peters Y, van Grinsven E, Sierema PD. Systematic review with meta-analysis: the effects of family history on the risk of Barrett's esophagus and esophageal adenocarcinoma. Aliment Pharmacol Ther 2021;54:868-879.

60.Thiruvengadam SK, Tieu AH, Luber B, et al. Risk factors for progression of Barrett's esophagus to high grade dysplasia and esophageal adenocarcinoma. Sci Rep 2020;10:4899.

61.Tan WK, Ross-Innes CS, Somerset T, et al. Biomarker risk stratification with capsule sponge in the surveillance of Barrett's oesophagus: prospective evaluation of UK real-world implementation. Lancet 2025;406:271-282.

22. Landy R, Kilcoyne S, Tang C, et al. Real-world imple-mentation of non-endoscopic triage testing for Barrett's oesophagus during COVID-19. QJM 2023;116:659-666.

63.Chien S, Glen P, Penman I, et al. Oesophageal cell collection device and biomarker testing to identify high-risk Barrett's patients requiring endoscopic investiga-tion. Br J Surg 2024;111.

64. Erichsen R, Horvath-Puho E, Lund JL, et al. Mortality and cardiovascular diseases risk in patients with Barrett's oesophagus: a population-based nationwide cohort study.Aliment Pharmacol Ther 2017;45:973-982.

65.Cook MB, Coburn SB, Lam JR, et al. Cancer incidence and mortality risks in a large US Barrett's oesophagus cohort. Gut 2018;67:418-529.

66.Ozanne EM, Silver ER, Saini SD, et al. Surveillance cessation for Barrett's esophagus: a survey of gastro-enterologists. Am J Gastroenterol 2021;116:1730-1733.

67.Omidvari A-H, Hazelton WD, Lauren BN, et al.The optimal age to stop endoscopic surveillance of patients with Barrett's esophagus based on sex and comorbid-ity: a comparative cost-effectiveness analysis. Gastro-enterology 2021;161:487-494.e4.

68.Wallner B, Sylvan A, Stenling R, et al. The esophageal Z-line appearance correlates to the prevalence of intesti-nal metaplasia. Scand J Gastroenterol 2000;35:17-22.

69.Wani S, Williams JL, Falk GW, et al. Analysis of the GlQuIc nationwide quality registry reveals unnecessary

surveillance endoscopies in patients with normal and irregular Z-lines. Am J Gastroenterol 2020;115:1869-1878.

70.Dickman R, Levi Z, Vilkin A, et al. Predictors of specialized intestinal metaplasia in patients with an incidental irregular Z line. Eur J Gastroenterol Hepatol 2010;22:135-138.

71.Anaprothy R, Gaddam S, Kanakadandi V, et al. Asso-ciation between length of Barrett's esophagus and risk of high-grade dysplasia or adenocarcinoma in patients without dysplasia. Clin Gastroenterol Hepatol 2013;11:1430-1436.

72.Yoon JY,Lim F,Shah SC,et al.Endoscopic surveillance of intestinal metaplasia of the esophagogastric junction:a decision modeling analysis. Am J Gastroenterol 2024;119:1289-1297.

73.Pohl H, Pech O, Arash H, et al. Length of Barrett's oesophagus and cancer risk: implications from a large sample of patients with early oesophageal adenocarci-noma. Gut 2016;65:196-201.

74.Russin M, Harbrecht M, Mishra A, et al. Detection of Barrett's esophagus prior to development of esopha-geal and esophagogastric junction adenocarcinoma.Clin Gastroenterol Hepatol 2025;23:1917-1924.e3.

75. Rubenstein JH, Burns J, Arasim ME, et al. Validation of tools using the electronic health record for predicting Barrett's esophagus with high-grade dysplasia. Clin Gastroenterol Hepatol. Published online April 30,2025.https://doi.org/10.1016/j.cgh.2025.03.005.

76.Guda NM, Partington S, Vakil N. Inter- and intra-observer variability in the measurement of length at endoscopy: Implications for the measurement of Bar-rett's esophagus. Gastrointest Endosc 2004;59:655-658.

77.ASGE Standards of Practice Committee, Qumseya B,Sultan S, Bain P,et al.ASGE guideline on screening and surveillance of Barrett's esophagus. Gastrointest Endosc 2019;90:335-359.e2.

78.Longcroft-Wheaton G, Fogg C, Chedgy F, et al.A feasibility trial of acetic acid-targeted biopsies versus nontargeted quadratic biopsies during Bar-rett's surveillance: the ABBA trial. Endoscopy 2020;52:29-36.

79.Canto Ml, Setrakian S, Willis J, et al. Methylene blue-directed biopsies improve detection of intestinal meta-plasia and dysplasia in Barrett's esophagus. Gastro-intest endostc 2000;51:560-568.

80.Gossner L, Pech O, May A, et al. Comparison of methylene blue-directed biopsies and four-quadrant biopsies in the detection of high-grade intraepithelial neoplasia and early cancer in Barrett's oesophagus.Dig Liver Dis 2006;38:724-729.

81.Horwhat JD, Maydonovitch CL, Ramos F,et al.A randomized or comparison of methylene blue-directed biopsy versus conventional four-quadrant biopsy for the detection of intestinal metaplasia and dysplasia in patients with long-segment Barrett's esophagus. Am J Gastroenterol 2000;35:17-22.

82.Ragunath K, Krasner N, Raman VS, et al. A randomized,prospective cross-over trial comparing methylene blue-

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1225

directed biopsy and conventional random biopsy for detecting intestinal metaplasia and dysplasia in Bar-rett's esophagus. Endoscopy 2003;35:998-1003.

83.Wo JM, Ray MB, Mayfield-Stokes S, et al. Comparison of methylene blue-directed biopsies and conventional biopsies in the detection of intestinal metaplasia and dysplasia in Barrett's esophagus: a preliminary study.Gastrointest Endosc 2001;54:294-301.

84.Curvers WL, Alvarez Herrero L, Wallace MB, et al.Endoscopic tri-modal imaging is more effective than standard endoscopy in identifying early-stage neoplasia in Barrett's esophagus. Gastroenterology 2010;139:1106-1114.

85.Curvers WL, Singh R, Song LM, et al. Endoscopic tri-modal imaging for detection of early neoplasia in Bar-rett's oesophagus: a multi-centre feasibility study using high-resolution endoscopy, autofluorescence imaging and narrow band imaging incorporated in one endos-copy system. Gut 2008;57:167-172.

86.Curvers WL, van Vilsteren FG, Baak LC, et al. Endo-scopic trimodal imaging versus standard video endos-copy for detection of early Barrett's neoplasia: a multicenter, randomized, crossover study in general practice. Gastrointest Endosc 2011;73:195-203.

87.Wolfsen HC, Crook JE, Krishna M, et al. Prospective,controlled tandem endoscopy study of narrow band imaging for dysplasia detection in Barrett's esophagus.Gastroenterology 2008;135:24-31.

88.Sharma P, Hawes RH, Bansal A, et al. Standard endoscopy with random biopsies versus narrow band imaging targeted biopsies in Barrett's oesophagus: a prospective, international, randomised controlled trial.Gut 2013;62:15-21.

89.Kara MA, Peters FP, Rosmolen WD, et al. High-resolu-tion endoscopy plus chromoendoscopy or narrow-band imaging in Barrett's esophagus: a prospective ran-domized crossover study. Endoscopy 2005;37:929-936.

90.Olliver JR, Wild CP, Sahay P, et al. Chromoendoscopy with methylene blue and associated DNA damage in Barrett's oesophagus. Lancet 2003;362:373-374.

91.Bhandari P, Kandaswamy P, Cowlishaw D, et al. Acetic acid-enhanced chromoendoscopy is more cost-effective than protocol-guided biopsies in a high-risk Barrett's population. Dis Esophagus 2012;25:386-392.

92.Wani S, Sharma P. Endoscopic surface imaging of Barrett's esophagus: an optimistic view. Gastroenter-ology 2007;133:11-13.

93.Sharma P, Savides TJ, Canto MI, et al. The American So-ciety for Gastrointestinal Endoscopy PIVI(Preservation and Incorporation of Valuable Endoscopy IPIV(Preservation and Incorporation) in aging in Barrett's esophagus. Gastrointest Endosc 2012;76:252-254.

94.Committee AT, Thosani N, Abu Dayeh BK, et al. ASGE Technology Committee systematic review and meta-analysis assessing the ASGE Preservation and Incorpora-tion of Valuable Endoscopic Innovations thresholds for adopting real-time imaging-assisted endoscopic targeted biopsy during endoscopic surveillance of Barrett's esoph-gaus. Gastrointest Endosc 2016;83:684-698.e7.

95.Sharma P, Bergman JJ, Goda K, et al. Development and Validation of a classification system to identify high-grade dysplasia and esophageal adenocarcinoma in Barrett's esophagus using narrow-band imaging.Gastroenterology 2016;150:591-598.

96. Bergman J, de Groof AJ, Pech O, et al. An interactive web-based educational tool improves detection and delineation of Barrett's esophagus-related neoplasia.Gastroenterology 2019;156:1299-1308.e3.

97. Qumseya BJ, Wang H, Badie N, et al. Advanced imag-ing technologies increase detection of dysplasia and neoplasia in patients with Barrett's esophagus: a meta-analysis and systematic review. Clin Gastroenterol Hepatol 2013;11:1562-1570.e1-e2.

98.de Groof AJ, Struyvenberg MR, van der Putten J, et al.Deep-learning system detects neoplasia in patients with Barrett's Esophagus with higher accuracy than endo-scopists in a multistep training and validation study with benchmarking. Gastroenterology 2020;158:915-929.e4.

99.de Groof AJ,Struyenberg MR,Fockens KN,et al.Deep learning algorithm detection of Barrett's neoplasia with high accuracy during live endoscopic procedures: a pilot study(with video). Gastrointest Endosc 2020;91:1242-1250.

100.Ebigbo A, Mendel R, Probst A, et al. Real-time use of artificial intelligence in the evaluation of cancer in Bar-rett's oesophagus. Gut 2020;69:615-616.

101. Hashimoto R, Requa J, Dao T, et al. Artificial intelligence using convolutional neural networks for real-time detection of early esophageal neoplasia in Barrett's esophagus(with video). Gastrointest Endosc 2020;91:1264-1271.e1.

102.Struyvenberg MR, de Groof AJ, van der Putten J, et al.A computer-assisted algorithm for narrow-band imag-ing-based tissue characterization in Barrett's esoph-agus. Gastrointest Endosc 2021;93:89-98.

103.Faghani S, Codipilly DC, David V, et al. Development of a deep learning model for the histologic diagnosis of dysplasia in Barrett's esophagus. Gastrointest Endosc 2022;96:918-925.e3.

104.Chen D, Wu L, Li Y, et al. Comparing blind spots of unsedated ultrafine, sedated, and unsedated conven-tional gastroscopy with and without artificial intelli-gence: a prospective, single-blind, 3-parallel-group,randomized, single-center trial. Gastrointest Endosc 2020;91:332-339.e3.

105.Fockens KN, Jukema JB, Jong MR, et al. The use of a real-time computer-aided detection system for visible lesions in the Barrett's esophagus during live endoscopic proced-ures: a pilot study(with video). Gastrointest Endosc 2024;100:527-531.e3.

106.Jukema JB, Kusters CHJ, Jong MR, et al. Computer-aided diagnosis improves characterization of Barrett's neoplasia by general endoscopists(with video). Gas-trointest Endosc 2024;100:616-625.e8.

107.Force AAT, Parasa S, Berzin T, et al. Consensus state-ments on the current landscape of artificial intelligence applications in endoscopy, addressing roadblocks, and advancing artificial intelligence in gastroenterology.Gastrointest Endosc 2025;101:2-9.e1.

1226 Wani et al

Gastroenterology Vol.169, Iss. 6

108. Levine DS, Haggitt RC, Blount PL, et al. An endoscopic biopsy protocol can differentiate high-grade dysplasia from early adenocarcinoma in Barrett's esophagus.Gastroenterology 1993;105:40-50.

109.Nachiappan A, Ragunath K, Card T, et al. Diagnosing dysplasia in Barrett's oesophagus still requires Seattle protocol biops in the era of modern video endoscopy:results from a tertiary centre Barrett's dysplasia data-base. Scand J Gastroenterol 2020;55:9-13.

110.Roumans CAM, van der Bogt RD, Nieboer D, et al. Clinical consequences of nonadherence to Barrett's esophagus surveillance recommendations: a Multicenter prospective cohort study. Dis Esophagus 2023;36:doac113.

111.Wani S, Yadlapati R, Singh S, et al. Post-endoscopy esophageal neoplasia in Barrett's esophagus: consensus statements from an international expert panel. Gastroenterology 2022;162:366-372.

112.Peleg N, Ollech JE, Shamah S, et al. Seattle protocol is more effective in detection of dysplasia compared to technology-assisted targeted biopsies in patients with Barrett's esophagus. J Clin Med 2023;12:2544.

111.Roumans CAM, van der Bogt RD, Steyerberg EW,et al.Adherence to recommendations of Barrett's esophagus surveillance guidelines: a systematic review and meta-analysis. Endoscopy 2020;52:17-28.

114.Abrams JA, Kapel RC, Lindberg GM, et Adherence to biopsy guidelines for Barrett's esophagus surveillance in the community setting in the United States. Clin Gas-troenterol Hepatol 2009;7:736-742; quiz 710.

115.Kolb JM, Davis C, Williams JL, et al. Wide variability in dysplasia detection rate and adherence to sealpro-tocol and surveillance recommendations in Barrett's esophagus: a population-based analysis using the GluQIC National Quality Benchmarking Registry. Am J Gastroenterol 2023;118:900-904.

116.Wani S, Williams JL, Komanduri S, et al. Endoscopists systematically undersample patients with long-segment Barrett's esophagus: an analysis of biopsy sampling practices from a quality improvement registry. Gastro-intest Endosc 2019;90:732-741.e3.

117.Nagula S, Parasa S, Laine L, et al. AGA clinical practice update on high-quality upper endoscopy: expert review.Clin Gastroenterol Hepatol 2024;22:933-943.

118.Anderson MA, Ben-Menachem T, Gan SI, et al. Man-agement of antithrombotic agents for endoscopic pro-cedures. Gastrointest Endosc 2009;70:1060-1070.

119.Kolb JM, Wani S. Endoscopic eradication therapy for Barrett's oesophagus: state of the art. Curr Opin Gas-troenterol 2020;36:351-358.

120.Gupta N, Gaddam S, Wani SB, et al. Longer inspection time is associated with increased detection of high-grade dysplasia and esophageal adenocarcinoma in Barrett's esophagus. Gastrointest Endosc 2012;76:531-538.

121.Bisschops R, Areia M, Coron E, et al. Performance measures for upper gastrointestinal endoscopy: a Eu-ropean Society of Gastrointestinal Endoscopy (ESGE)Quality Improvement Initiative. Endoscopy 2016;48:843-864.

122.Vahabzadeh B, Seetharam AB, Cook MB, et al. Valida-tion of the Prague C & M criteria for the endoscopic grading of Barrett's esophagus by gastroenterology trainees: a multicenter study. Gastrointest Endosc 2012;75:236-241.

123.Alvarez Herrero L., Curvers WL, van Vilsteren FG, et al.Validation of the Prague C&M classification of Barrett's esophagus in clinical practice. Endoscopy 2013;45:876-882.

124.Sharma P, Dent J, Armstrong D, et al.The development and validation of an endoscopic grading system for Barrett's esophagus: the Prague C & M criteria.Gastroenterology 2006;131:1392-1399.

125.Update on the Paris Classification of superficial neoplastic lesions in the digestive tract. Endoscopy 2005;37:570-578.

126.Curvers WL, ten Kate FJ, Krishnadath KK, et al. Low-grade dysplasia in Barrett's esophagus: overdiagnosed and underestimated. Am J Gastroenterol 2010;105:1523-1530.

127.Castell DO, Kahrilas PJ, Richter JE, et al.Esomeprazole(40 mg) compared with lansoprazole(30 mg) in the treatment of erosive esophagitis. Am J Gastroenterol 2002;97:575-583.

128.ASGE Standard of Practice Committee, Muthusamy VR,Lightdale JR, Acosta RD, et al. The role of endoscopy in the management of GERD. Gastrointest Endosc 2015;81:1305-1310.

129. Richter JE, Kahrillas PJ, Sontag SJ, et al. Comparing lansoprazole and omeprazole in onset of heartburn re-lief: results of a randomized, controlled trial in erosive esophagitis patients. Am J Gastroenterol 2001;96:3089-3098.

130.Katz PO, Dunbar KB, Schnoll-Sussman FH, et al. ACG clinical guideline for the diagnosis and management of gastroesophageal reflux disease. Am J Gastroenterol 2022;117:27-56.

131. Hanna S, Rastogi A, Weston AP, et al. Detection of Barrett's esophagus after endoscopic healing of erosive esophagitis. Am J Gastroenterol 2006;101:1416-1420.

132.Modiano N, Gerson LB. Risk factors for the detection of Barrett's esophagus in patients with erosive esophagi-tis. Gastrointest Endosc 2009;69:1014-1020.

133.Rodriguez S, Mattek N, Lieberman D, et al. Barrett's esophagus on repeat endoscopy: should we look more than once? Am J Gastroenterol 2008;103:1892-1897.

134. Rubenstein JH, Mattek N, Eisen G. Age- and sex-specific yield of Barrett's esophagus by endoscopy indication. Gastrointest Endosc 2010;71:21-27.

135.Davis C, Fuller A, Katzka D, et al. High proportions of newly detected visible lesions and pathology grade change among patients with Barrett's esophagus referred to expert centers. Dig Dis Sci 2023;68:3584-3595.

136.Nieuwenhuis EA, van Munster SN, Curvers WL, et al.Impact of expert center endoscopic assessment of confirmed low grade dysplasia in Barrett's esophagus diagnosed in community hospitals. Endoscopy 2022;54:936-944.

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1227

137.Tsoi EH, Mahindra P, Cameron G, et al. Barrett's esophagus with low-grade dysplasia: high rate of upstaging at Barrett's esophagus referral units suggests progression rates may be overestimated. Gastrointest Endosc 2021;94:902-908.

138.Noordzij IC, Van Loon van de Ende MCM, Curvers WL,et al. Dysplasia in random biopsies from Barrett's sur-veillance is an important marker for more severe pa-thology. Dig Dis Sci 2021;66:1957-1964.

139.Bellizzi AM, Odze RD. Histopathology of Barrett's esophagus: a review for the practicing gastroenterollo-gist. Techniq Gastrointesl Endosc 2010;12:69-81.

140.Tolone S, Limongelli P, Romano M, et al. The patterns of reflux can affect regression of non-dysplastic and low-grade dysplastic Barrett's esophagus after medical and surgical treatment: a prospective case-control study. Surg Endosc 2015;29:648-657.

141.Rossi M, Barreca M, de Bortoli N, et al. Efficacy of Nissen fundoplication versus medical therapy in the regression of low-grade dysplasia in patients with Bar-rett esophagus: a prospective study. Ann Surg 2006;243:58-63.

142.Singh S, Manickam P, Amin AV, et al. Incidence of esoph-ageal adenocarcinoma in Barrett's esophagus with low-grade dysplasia: a systematic review and meta-analysis.Gastrointest Endosc 2014;79:897-909.e4.

143.Krishnamoorthi R, Mohan BP, Jayaraj M, et al. Risk of progression in Barrett's esophagus indefinite for dysplasia: a systematic review and meta-analysis.Gastrointest Endosc 2020;91:3-10.e3.

144.Kopczynska M, Ratcliffe E, Yalamanchil H, et al. Bar-rett's oesophagus with indefinite for dysplasia shows high rates of prevalent and incident neoplasia in a UK multicentre cohort. J Clin Pathol 2023;76:847-854.

145.Phillips R, Januszewicz W, Pilonis ND, et al. The risk of neoplasia in patients with Barrett's esophagus indefinite for dysplasia: a multicenter cohort study. Gastrointest Endosc 2021;94:263-270.e2.

146.Januszewicz W, Pilonis ND, Sawas T, et al. The utility of P53 immunohistochemistry in the diagnosis of Barrett's oesophagus with indefinite for dysplasia. Histopatholo-gy 2022;80:1081-1090.

147.Codipilly DC, Krishna Chandar A, Wang KK, et al. Wide-area transepithelial sampling for dysplasia detection in Barrett's esophagus: a systematic review and meta-analysis. Gastrointest Endosc 2022;95:51-59.e7.

148.van Munster SN, Leclercq P, Haidry R, et al. Wide-area transepithelial sampling with computer-assisted anal-ysis to detect high grade dysplasia and cancer in Bar-rett's esophagus: a multicenter randomized study.Endoscopy 2023;55:303-310.

149.DeMeester S, Smith C, Severson P, et al. Multicenter randomized controlled trial comparing forceps biopsy sampling with wide-area transepithelial sampling brush for detecting intestinal metaplasia and dysplasia during routine upper endoscopy. Gastrointest Endosc 2022;95:1101-1110.e2.

150.Zhao CL, Hossein-Zadeh Z, Dabiri B, et al. The concordance between wide-area transepithelial

sampling with computer-assisted 3-dimensional anal-ysis(WATS-3D) and standard endoscope biopsy in the detection of Barrett's esophagus and esophageal dysplasia. Ann Diagn Pathol 2022;60:151982.

151. Trindade AJ, Odze RD, Smith MS, et al. Benefit of adjunctive wide-area transepithelial sampling with 3-dimensional computer-assisted analysis plus forceps biopsy based on Barrett's esophagus segment length.Gastrointest Endosc 2023;98:316-325.

152.Qumseya B, Bukannan A, Rosasco R, et al. Surveillance of Barrett's esophagus using wide-area transepithelial sampling: systematic review and meta-analysis. Endosc Int Open 2022;10:E394-E402.

153. Johanson JF, Frakes J, Eisen D, et al. Computer-assisted analysis of abrasive transepithelial brush bi-opsies increases the effectiveness of esophageal screening: a multicenter prospective clinical trial by the EndoCDx Collaborative Group. Dig Dis Sci 2011;56:767-772.

154.Agha YH, Srinivasan S, Hyder J, et al. WATS3D versus forceps biopsy in screening for Barrett's esophagus:experience in community endoscopy centers. Ann Gastroenterol 2021;34:164-168.

155. Shaheen NJ, Smith MS, Odze RD. Progression of Barrett's esophagus, crypt dysplasia, and low-grade dysplasia diagnosed by wide-area transepithelial sampling with 3-dimensional computer-assisted analysis: a retrospective analysis. Gastrointest Endosc 2022;95:410-418.e1.

156.Shaheen N, Smith M, Kaul V, et al. Progression rates of Barrett's esophagus and dysplasia detected by WATS-3D are similar to those diagnosed by forceps biopsies.Gastroenterology 2025;169:S-224-S-225.

157.Vennalaganti PR, Kaul V, Wang KK, et al. Increased detection of Barrett's esophagus-associated neoplasia using wide-area trans-epithelial sampling: a multicenter,prospective, randomized trial. Gastrointest Endosc 2018;87:348-355.

158.Miglioretti DL, Abraham L, Sprague BL, et al. Association between false-positive results and return to screening mammography in the Breast Cancer Surveillance Con-sortium Cohort. Ann Intern Med 2024;177:1297-1307.

159.Rasmussen JF, Sierersma V, Malmqvist J, et al. Psycho-social consequences of false positives in the Danish Lung Cancer CT Screening Trial: a nested matched cohort study. BMJ Open 2020;10:e034682.

160. Shaheen NJ, Odze RD, Singer ME, et al. Adjunctive use of wide-area transepithelial sampling-3D in patients with symptomatic gastroesophageal reflux increases detec-tion of Barrett's esophagus and dysplasia. Am J Gas-troenterol 2024;119:1990-2001.

161.Patil DT, Goldblum JR, Lauwers G,et al. WATS 3D: an interobserver study of Barrett's esophagus-associated dysplasia among gastrointestinal pathologists. Clin Transl Gastroenterol 2024;15:e00661.

162.Schlemper RJ, Riddell RH, Kato Y, et al. The Vienna classification of gastrointestinal epithelial neoplasia. Gut 2000;47:251-255.

163.Strimbu K, Tavel JA. What are biomarkers? Curr Opin HIV AIDS 2010;5:463-466.

1228 Wani et al

Gastroenterology Vol.169, Iss. 6

164.Souza RF, Spechler SJ. Advances in biomarkers for risk stratification in Barrett's esophagus. Gastrointest Endosc Clin N Am 2021;31:105-115.

165.Wani S, Mathur SC, Curvers WL, et al. Greater inter-observer agreement by endoscopic mucosal resection than biopsy samples in Barrett's dysplasia. Clin Gas-troenterol Hepatol 2010;8:783-788.

166.Vennalaganti P, Kanakadandi V, Goldblum JR, et al.Discordance among pathologists in the United States and Europe in diagnosis of low-grade dysplasia for patients with Barrett's esophagus. Gastroenterology 2017;152:564-570.e4.

167.van der Wel MJ, Coleman HG, Bergman J, et al. His-topathologist features predictive of diagnostic concor-dance at expert level among a large international sample of pathologists diagnosing Barrett's dysplasia using digital pathology. Gut 2020;69:811-822.

168.Stachler MD, Taylor-Weiner A, Peng S, et al. Paired exome analysis of Barrett's esophagus and adenocar-cinoma. Nat Genet 2015;47:1047-1055.

169.Nones K, Waddell N, Wayne T, et al. Genomic catas-trophes frequently arise in esophageal adenocarcinoma and drive tumorigenesis. Nat Commun 2014;5:5224.

170.Singh S,Ananthakrishnan AN, Nguyen NH, et al. AGA clinical practice guideline on the role of biomarkers for the management of ulcerative colitis. Gastroenterology 2023;164:344-372.

171.Ananthakrishnan AN, Adler J, Chachu KA, et al. AGA clinical practice guideline on the role of biomarkers for the management of Crohn's disease. Gastroenterology 2023;165:1367-1399.

172.Snyder P, Dunbar K, Cipher DJ, et al. Aberrant p53 Immunostaining in Barrett's esophagus predicts neoplastic progression: systematic review and meta-analyses. Dig Dis Sci 2019;64:1089-1097.

173.Peters Y, Honing J, Kievit W, et al. Incidence of pro-gression of persistent nondysplastic Barrett's esoph-agus to malignancy. Clin Gastroenterol Hepatol 2019;17:869-877.e5.

174.Menon S, Norman R, Iyer PG,et al. Stratification of Barrett's esophagus surveillance based on p53 immunohistochem-istry: a cost-effectiveness analysis by an international collaborative group. Endoscopy 2024;56:727-736.

175.van den Berg D, Ovidmi AH, Duits L, et al. The cost-effectiveness of risk-based management of Barrett's esophagus patients using TissueCypher or p53 bio-markers: a microsimulation study. Dis Esophagus 2025;38(Suppl 1):doaf061.325.

176.Redston M, Noffersinger, Kim A, et al. Abnormal TP53 predicts risk of progression in patients with Barrett's Esophagus regardless of a diagnosis of dysplasia.Gastroenterology 2022;162:468-481.

177.Tomaszewski KJ, Neyaz A, Sauder K, et al. Defining an abnormal p53 immunohistochemical stain in Barrett's oesophagus-related dysplasia: a single-positive crypt is a sensitive and specific marker of dysplasia. Histopa-thology 2023:82:555-566.

178.Skacel M, Petras RE, Rybicki LA, et al. p53 expression in low grade dysplasia in Barrett's esophagus: correlation

<EndOfPara>with interobserver agreement and disease progression.Am J Gastroenterol 2002;97:2508-2513.

179. Kaye PV, Haider SA, llyas M, et al. Barrett's dysplasia and the Vienna classification: reproducibility, prediction of progression and impact of consensus reporting and p53 immunohistochemistry. Histopathology 2009;54:699-712.

180. Kaye PV, Ilyas M, Soomro I, et al. Dysplasia in Barrett's oesophagus: p53 immunostaining is more reproducible than haematoxylin and eosin diagnosis and improves overall reliability, while grading is poorly reproducible.Histopathology 2016;69:431-440.

181.van der Wel MJ, Duits LC, Pouw RE, et al. Improved diagnostic stratification of digitised Barrett's oesoph-agus is biopsies by p53 immunohistochemical staining.Histopathology 2018;72:1015-1023.

182.18yer PG, Codipilly DC, Chandar AK, et al. Prediction of progression in Barrett's Esophagus using a tissue sys-tems pathology test: a pooled analysis of international multicenter studies. Clin Gastroenterol Hepatol 2022;20:2772-2779.e8.

183.Davison JM, Goldblum JR, Duits LC, et al. A tissue systems pathology test outperforms the standard-of-care variables in predicting progression in patients with Barrett's esophagus. Clin Transl Gastroenterol 2023;14:e00631.

184.Frei NF, Khoshiwal AM, Konte K, et al. Tissue systems pathology test objectively risk stratifies Barrett's esophagus patients with low-grade dysplasia. Am J Gastroenterol 2021;116:675-682.

18.Frei NF, Konte K, Bossart EA, et al. Independent vali-dation of a tissue systems pathology assay to predict future progression in nondysplastic Barrett's esoph-agus: a spatial-temporal analysis. Clin Transl Gastro-enterol 2020;11:e00244.

186.Critchley-Thorne RJ, Davison JM, Prichard JW, et al.A tissue systems pathology test detects abnormalities associated with prevalent high-grade dysplasia and esophageal cancer in Barrett's esophagus. Cancer Epidemiol Biomarkers Prev 2017;26:240-248.

187.Critchley-Thorne RJ, Duits LC, Prichard JW, et al.A tissue systems pathology assay for high-risk Barrett's esophagus. Cancer Epidemiol Biomarkers Prev 2016;25:958-968.

188.Davison JM, Goldblum J, Grewal US, et al. Independent blinded validation of a tissue systems pathology test to predict progression in patients with Barrett's esoph-agus. Am J Gastroenterol 2020;115:843-852.

189.Leggett C, Das A, Kamboj A, et al. Risk-aligned management guided by the tissue systems pathology test can improve health outcomes in Barrett's esophagus and reduce healthcare-associated costs.Am J Gastroenterol 2024;119:S486-S487.

190.Khoshiwal AM, Frei NF, Pouw RE, et al. The tissue systems pathology test outperforms pathology review in risk stratifying patients with low-grade dysplasia.Gastroenterology 2023;165:1168-1179.e6.

191.Prichard JW, Davison JM, Campbell BB, et al. Tissue CypherTM: a systems biology approach to anatomic pathology.J Pathol Inform 2015;6:48.

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1229

192.Wani S, Souza RF, Durkalski VL, et al. Multicenter ran-domized controlled trial of surveillance versus endo-scopic therapy for Barrett's Esophagus with low-grade dysplasia: the SURVENT Trial: study rationale, meth-odology, innovation, and implications. Gastroenterology 2022;163:556-562.e4.

193.Douville C, Moinova HR, Thota PN, et al. Massively parallel sequencing of esophageal brushings enables an aneuploidy-based classification of patients with Bar-rett's esophagus. Gastroenterology 2021;160:2043-2054.e2.

194. Laun SE, Kann L, Braun J, et al. Spatiotemporal study of a risk-stratification epigenetic-based biomarker assay in patients with Barrett esophagus. Am J Gastroenterol 2025;120:1285-1295.

195.Chen Y, Sun C, Wu Y,et al. Do proton pump inhibitors prevent Barrett's esophagus progression to high-grade dysplasia and esophageal adenocarcinoma? An upda-ted meta-analysis. J Cancer Res Clin Oncol 2021;147:2681-2691.

196.Krishnamoorthi R, Borah B, Heien H, et al. Rates and predictors of progression to esophageal carcinoma in a large population-based Barrett's esophagus cohort.Gastrointest Endosc 2016;84:40-46.e7.

197.Gaddam S, Thota PN, Vennalaganti P, et al. Proton pump inhibitor use but not statin use is associated with decreased risk for high-grade dysplasia and esophageal adenocarcinoma in patients with Barrett's esophagus:results from a large, multicenter cohort study. Gastro-enterology 2014;146:S-123.

198.Jankowski JAZ, de Caestecker J, Love SB, et al. Eso-meprazole and aspirin in Barrett’s oesophagus(AsPect): a randomised factorial trial. Lancet 2018;392:400-408.

199.MoaydyE, Pikelboom JW, Bosch J, et al. Safety of proton pump inhibitors based on a large, multi-year,randomized trial of patients receiving rivaroxaban or aspirin. Gastroenterology 2019;157:682-691.e2.

200.Devereaux PJ, Duceppe E, Guyatt G, et al. Dabigatran in patients with myocardial injury after non-cardiac surgery(MANAGE): an international, randomised, placebo-controlled trial. Lancet 2018;391:2325-2334.

201. Laheij RJ, Sturkenboom MC, Hassing RJ, et al. Risk of community-acquired pneumonia and use of gastric acid-suppressive drugs. JAMA 2004;292:1955-1960.

202.Yang YX, Lewis JD, Epstein S, et al. Long-term proton pump inhibitor therapy and risk of hip fracture. JAMA 2006;296:2947-2953.

203.Leonard J, Marshall JK, Moayyedi P. Systematic review of the risk of enteric infection in patients taking acid suppression. Am J Gastroenterol 2007;102:2047-2956;quiz 2057.

204.Lazarus B, Chen Y, Wilson FP, et al. Proton pump in-hibitor use and the risk of chronic kidney disease. JAMA Intern Med 2016;176:238-246.

205.Gomm W, von Holt K, Thomé F, et al. Association of proton pump inhibitors with risk of dementia: a phar-macoepidemiological claims data analysis. JAMA Neu-rol 2016;73:410-416.

206.Mehta RS, Kochar B, Zhou Z, et al. Association of proton pump inhibitor use with incident dementia and cognitive decline in older adults: a prospective cohort study. Gastroenterology 2023;165:564-572.e1.

207.Salvo EM, Ferko NC, Cash SB, et al. Umbrella review of 42 systematic reviews with meta-analyses: the safety of proton pump inhibitors. Aliment Pharmacol Ther 2021;54:129-143.

208.Mosholder AD, Hada M, Leishear K. Proton-pump in-hibitors and cardiovascular adverse events: a meta-analysis of randomized controlled trials. Am J Gastro-enterol 2025;120:362-369.

209.Singh S, Garg SK, Singh PP, et al. Acid-suppressive medications and risk of oesophageal adenocarcinoma in patients with Barrett's oesophagus: a systematic re-view and meta-analysis. Gut 2014;63:1229-1237.

210.Targownik LE, Fisher DA, Saini SD. AGA clinical prac-tice update on de-prescribing of proton pump in-hibitors: expert review. Gastroenterology 2022;162:1334-1342.

211.Kamboj AK, Patel DA, Yadlapati R. Long-term proton pump inhibitor use: review of indications and special considerations. Clin Gastroenterol Hepatol 2024;22:1373-1376.

212.Falk GW,Buttar NS, Foster NR,et al. A combination of esomeprazole and aspirin reduces tissue concentra-tions of prostaglandin E(2) in patients with Barrett's esophagus. Gastroenterology 2012;143:917-926.e1.

213. Buas MF, He Q, Johnson LG, et al. Germline variation in inflammation-related pathways and risk of Barrett's oesophagus and oesophageal adenocarcinoma. Gut 2017;66:1739-1747.

214. Hashemi Gordel N, Najafi M, Salehi E, et al. Cyclo-oxygenase-2 in cancer: a review. J Cell Physiol 2019;234:5683-5699.

215.Corley DA, Kerlikowski S, Verma R, et al. Protective association of aspirin/NSAIDs and esophageal cancer: a systematic review and meta-analysis. Gastroenterology 2003;124:47-56.

216.Liao LM, Vaughan TL, Corley DA, et al. Nonsteroidal anti-inflammatory drug use reduces risk of adenocarci-nomas of the esophagus and esophagogastric junction in a pooled analysis. Gastroenterology 2012;142:442-452.e5; quiz e22-e23.

217.Gammon MD, Terry MB, Arber N, et al. Nonsteroidal anti-inflammatory drug use associated with reduced incidence of adenocarcinomas of the esophagus and gastric cardia that overexpress cyclin D1: a population-based study. Cancer Epidemiol Biomarkers Prev 2004;13:34-39.

218.di Pietro M, Trudgill NJ, Vasileiou M, et al. National Institute for Health and Care Excellence(NICE) guidance on monitoring and management of Barrett's oesoph-agus and stage I oesophageal adenocarcinoma. Gut 2024;73:897-909.

219.Patel A, Laine L, Moayyedi P, et al. AGA clinical practice update on integrating potassium-competitive acid blockers into clinical practice: expert review. Gastro-enterology 2024;167:1228-1238.

1230 Wani et al

Gastroenterology Vol.169, Iss. 6

220.Cook MB, Shaheen NJ, Anderson LA, et al. Cigarette smoking increases risk of Barrett's esophagus: an anal-ysis of the Barrett's and Esophageal Adenocarcinoma Consortium. Gastroenterology 2012;142:744-753.

221.Coleman HG, Bhat S, Johnston BT, et al. Tobacco smoking increases the risk of high-grade dysplasia and cancer among patients with Barrett’s esophagus.Gastroenterology 2012;142:233-240.

222.Balasubramanian G, Gupta N, Giacchino M, et al. Cigarette smoking is a modifiable risk factor for Barrett's oesophag.United. European Gastroenterol J 2013;1:430-437.

223.Core Quality Measures Collaborative. Consensus Core Set: Gastroenterology. Available at: https://p4qm.org/sites/default/files/2023-06/CQMC-Gastroenterology-Core-Set-v4.0.pdf. Accessed September 4,2025.

224.Jacobson BC, Somers SC, Fuchs CS, et al. Body-mass index and symptoms of gastroesophageal reflux in women. N Engl J Med 2006;354:2340-2348.

225.Cook MB,Greenwood DC,Hardie LJ,et al.A systematic review and meta-analysis of the risk of increasing adiposity on Barrett's esophagus. Am J Gastroenterol 2008;103:292-300.

226.Duggan C, Onstad L, Hardikar S, et al. Association be-tween markers of obesity and progression fr. Bar-rett's esophagus to esophageal adenocarcinoma. Clin Gastroenterol Hepatol 2013;11:934-943.

227.Yadlapati R, Pandolfino JE, Alexeeva O, et al. The Reflux Improvement and Monitoring(TRIM) program is asso-ciated with symptom improvement and weight reduc-tion for patients with obesity and gastroesophageal reflux disease. Am J Gastroenterol 2018;113:23-30.

228.Wilson H, Mocanu V, Sun W, et al. Fundoplication is su-perior to medical therapy for Barrett's esophagus disease regression and progression: a systematic review and meta-analysis. Surg Endosc 2022;36:2554-2563.

229.Akerstrom JH, Santoni G, von Euler Chelpin M, et al.Antireflux Surgery versus antireflux medication and risk of esophageal adenocarcinoma in patients with Barrett's esophagus. Gastroenterology 2024;166:132-138.e3.

230.Parrilla P, Martinez de Haro LF, Ortiz A, et al. Long-term results of a randomized prospective study comparing medical and surgical treatment of Barrett's esophagus. Ann Surg 2003;237:291-298.

231.McKinley SK, Dirks RC, Walsh D, et al. Surgical treat-ment of GERD: systematic review and meta-analysis.Surg Endosc 2021;35:4095-4123.

232.Yadlapati R, Early D, lyer PG, et al. Quality indicators for upper Gl endoscopy. Am J Gastroenterol 2025;120:290-312.

233.Yadlapati R, Hungness ES, Pandolfino JE. Complica-tions of antireflux surgery. Am J Gastroenterol 2018;113:1137-1147.

234.Maret-Ouda J, Wahlin K, El-Serag HB,et al. Association between laparoscopic antireflux surgery and recurrence of gastroesophageal reflux. JAMA 2017;318:939-946.

235.Carlson MA, Frantzides CT. Complications and results of primary minimally invasive antireflux procedures: a review of 10,735 reported cases. J Am Coll Surg 2001;193:428-439.

236.Yanes M, Santoni G, Maret-Ouda J,et al. Mortality,reop-eration, and hospital stay within 90 days of primary and secondary antireflux surgery in a population-based multi-national study. Gastroenterology 2021;160:2283-2290.

237.Anvari M,Allen C,Marshall J,et al.A randomized controlled trial of laparoscopic Nissen fundoplication versus proton pump inhibitors for the treatment of patients with chronic gastroesophageal reflux disease(GERD): 3-year out-comes. Surg Endosc 2011;25:2547-2554.

238.Galmiche JP, Hatlebakk J, Attwood S, et al. Laparo-scopic antireflux surgery vs esomeprazole treatment for chronic GERD: the LOTUS randomized clinical trial.JAMA 2011;305:1969-1977.

239. Lundell L, Miettinen P, Myrvold HE, et al. Comparison of outcomes twelve years after antireflux surgery or omepra-zole maintenance therapy for reflux esophagitis. Clin Gas-troenterol Hepatol 2009;7:1292-1298; quiz 1260.

240.Spechler SJ, Lee E, Ahnen D, et al. Long-term outcome of medical and surgical therapies for gastroesophageal reflux disease: follow-up of a randomized controlled trial. JAMA 2001;285:2331-2338.

241.Attwood SE,Eli C,Galmiche JP,et al. Long-term safety of proton pump inhibitor therapy assessed under controlled, randomised clinical trial conditions: data from the SOPRAN and LOTUS studies. Aliment Phar-macol Ther 2015;41:1162-1174.

242.Maret-Oud a J, Markar SR, Lagergren J. Gastroesoph-ageal reflux disease: a review. JAMA 2020;324:2536-2547.

243.Souza RF. Fron in reflux esophagitis to esophageal adenocarcinoma. Dig Dis 2016;34:483-490.

244.Csendes A,Orellana O,Cuneo N,et al.Long-term(15-year)objective evaluation of 150 patients after laparoscopic Nissen fundoplication. Surgery 2019;166:886-894.

245.Adams MA,Allen JI, Saini SD.Translating best practices to meaningful quality measures: from measure conceptualization to implementation. Clin Gastroenterol Hepatol 2019;17:805-808.

246.Sheth SG,Maratt JK,Newberry C,et al.AGA Institute quality indicator development and uses. Clin Gastro-enterol Hepatol 2023;21:1399-1402.

247.Sharma P, Katzka DA, Gupta N, et al. Quality indicators for the management of Barrett's esophagus, dysplasia,and esophageal adenocarcinoma: international consensus recommendations from the American Gastroenterological Association Symposium. Gastro-enterology 2015;149:1599-1606.

248.Wani S, Williams JL, Komanduri S, et al. Over-utilization of repeat upper endoscopy in patients with non-dysplastic vertiang's esophagus: a quality registry study. Am J Gastroenterol 2019;114:1256-1264.

249.Wani S, Williams JL, Komanduri S, et al. Time trends in adherence to surveillance intervals and biopsy protocol among patients with Barrett's esophagus. Gastroen-terology 2020;158:770-772.e2.

250.Tavakkoli A, Appelman HD, Beer DG, et al. Use of appropriate surveillance for patients with nondysplastic Barrett's esophagus. Clin Gastroenterol Hepatol 2018;16:862-869.e3.

November 2025

AGA Clinical Practice Guideline on Surveillance of BE 1231

251.El-Serag HB, Duan Z, Hinojosa-Lindsey M, et al. Prac-tice patterns of surveillance endoscopy in a Veterans Affairs database of 29,504 patients with Barrett's esophagus. Gastrointest Endosc 2012;76:743-755.

252.Parasa S, Desai M, Vittal A, et al. Estimating neoplasia detection rate(NDR) in patients with Barrett's oesophagus based on index endoscopy: a systematic review and meta-analysis. Gut 2019;68:2122-2128.

253.Wani S, Gyawali CP, Katzka DA. AGA clinical practice update on reducing rates of post-endoscopy esopha-geal adenocarcinoma: commentary. Gastroenterology 2020;159:1533-1537.

254.Sawas T, Killcoyne S, Iyer PG, et al. Identification of prognostic phenotypes of esophageal adenocarcinoma in 2 independent cohorts. Gastroenterology 2018;155:1720-1728.e4.

255.Wani S, Santoni G, Holmberg D, et al. Neoplasia detection rate and risk of post-endoscopy esophageal adenocarcinoma and neoplasia in a population-based cohort study. Clin Gastroenterol Hepatol. Published online July 25,2025.https://doi.org/10.1016/j.cgh.2025.07.020.

Author names in bold designate shared co-first authorship.

Correspondence

Address correspondence to: Chair, Clinical Guidelines Committee, American Gastroenterological Association, National Office, 4930 Del Ray Avenue,Bethesda, Maryland 20814. e-mail: clinicalpractice@gastro.org.

Acknowledgments

The authors would like to thank Ms Caitlin Bakker, Archer Library at University of Regina, Regina Saktanchewan, for designing and conducting the literature searches and Melanie Stephens-Lyman, Garrett Adams, and Krisie Gardner for their administrative support.

Conflicts of interest

These authors disclose the following: Joel H. Rubenstein has received research funding from Lucid Diagnostics. David A. Katzka has received funding for consultation for Medtronic. Sachin Wani has received funding for consultation to Exact Sciences, Lucid, Enterotracker, Castle Biosciences,and Cyted and research support from Lucid, CDx Diagnostics, Cyted, and Exact Sciences. Siddharth Singh has received research funding from Pfizer.Margaret J. Zhou has received funding for consultation to Capsovision, Inc and Neptune Medical. The remaining authors disclose no conflicts.

Funding

Sachin Wani was supported by the National Institute of Health(1U01DK129191, 5U01CA271867-02) and the Katy O. and Paul M. Rady Endowed Chair in Esophageal Cancer Research. Joel H. Rubenstein was funded in part by the National Institutes of Health(U01DK129191) and Department of Defense(W81XWH2010898). Swathi Eluri was funded in part by the National Institutes of Health(5K23DK131338).