Endoscopic Evaluation of the Gastroesophageal Junction and Diagnosis of Barrett's Esophagus.

Introduction
Barrett’s esophagus (BE) is a premalignant condition defined by the replacement of the native stratified squamous epithelium in the distal esophagus with columnar epithelium that often occurs as a result of chronic gastroesophageal reflux disease. The clinical burdens of BE and esophageal adenocarcinoma have markedly increased with lifestyle habits and decreased Helicobacter pylori infection rates [1]. In the USA, esophageal adenocarcinoma accounts for more than 50% of all esophageal cancers in white men, which is a striking increase from 1% to 2% 2 decades previously [2]. The prevalence of BE in the Asia-Pacific region, excluding Japan, is approximately 5%, which is lower than that in the USA [3–14]. BE rates in Australia, China, India, Korea, Malaysia, Singapore, Taiwan, and Japan are 1.8%, 1%, 9%, 0.8%, 2%, 1.7%, 0.06%, and >35%, respectively. Despite the increasing recognition of BE, the global approach to its diagnosis is highly variable. A major challenge to diagnosing BE is the lack of universally accepted criteria, including differences in the BE definition, differences in what is considered to comprise the gastroesophageal junction (GEJ), and whether intestinal metaplasia is a prerequisite for the diagnosis of BE. These inconsistencies not only confound epidemiologic comparisons and hinder research collaborations but also have direct clinical implications that influence patient surveillance and therapeutic decisions. This review provides a comprehensive update of the endoscopic diagnosis of BE by integrating insights from recent international consensus statements, advances in image-enhanced endoscopy (IEE), and evolving quality indicators (QIs).

Variability of the Endoscopic Landmark of the GEJ
One of the key challenges associated with the endoscopic identification of BE is the need to precisely determine the GEJ. Various definitions of the GEJ exist. Some guidelines have described BE as “a columnar-lined segment of the esophagus that is continuous with the stomach” or “an esophageal segment showing intestinal epithelialization extending from the stomach” [15, 16]. Because the presence of intestinal metaplasia is confirmed histologically by performing a biopsy, an endoscopic diagnosis of the columnar-lined esophagus, which reflects BE, must be established. However, this elucidates the major issue of defining the GEJ during endoscopy, which is a point of divergence between the Japanese Esophageal Society guidelines and those of Western countries.
In Japan, the distal limit of the palisade vessels in the lower esophagus is considered an accurate marker of the GEJ (Fig. 1a) [17]. Therefore, the Japanese Esophageal Society guidelines define the GEJ as either the endpoint of these longitudinal vessels or, when they are not clearly visible, the proximal end of the gastric folds. Using this framework, BE is diagnosed as the entire length of the columnar-lined mucosa that extends from the GEJ to the oral side, as defined by the palisade vessel boundary.
In contrast, in the USA, Europe, and Australia, palisade vessels are considered unreliable landmarks because they may not be readily visible unless the GEJ is fully distended with deep inhalation, and their visibility may be compromised by inflammation (Fig. 1b) [18]. Consequently, Western guidelines, including those created by the American College of Gastroenterology, American Gastroenterological Association, British Society of Gastroenterology, and Australian societies, advocate using the proximal edge of the gastric folds as the reference point for GEJ localization (Fig. 1c; Table 1) [15, 16, 19–22]. However, this alternative has limitations. For example, the position of the gastric folds can shift with insufflation, and the folds may flatten or disappear entirely under excessive air inflation. Therefore, it is recommended that the gastric fold margin should be used only under controlled insufflation, where fold edges remain clearly visible [18].
Our group previously demonstrated that, with adequate instruction, endoscopists in the USA were able to recognize the distal margin of the palisade vessels (Fig. 2) [23]. Therefore, we assessed the visibility of the GEJ in endoscopic practice in the USA [24]. When using the palisade vessels as the landmark, the GEJ was visible in 87.8% of patients; this rate increased to 97.5% when the gastric folds were used (p = 0.008). Notably, approximately 30% of patients had reflux esophagitis, which may have affected palisade vessel detection. Similar findings were reported by Schölvinck et al. [25], who reported that Western endoscopists could recognize palisade vessels under appropriate insufflation; however, some discrepancies remained. Because of the variability in the gastric fold position with air volume and the accumulating evidence that palisade vessels are identifiable in both Eastern and Western settings, using the palisade vessel margin as the reference point for the GEJ appears to offer greater diagnostic reproducibility.
The Asian Barrett Consortium surveyed endoscopists across Asia to determine which endoscopic landmarks they preferred for defining the GEJ [26]. The results showed a strong preference for the squamocolumnar junction (SCJ), whereas the palisade vessels and gastric folds were used by 36.7% and 19% of respondents, respectively. However, reliance on the SCJ is problematic, particularly in cases involving hiatal hernia, ulcers, or columnar-lined esophagus, because the SCJ and GEJ may not align under such conditions. This underscores the need for widespread education and an international consensus regarding accurate and consistent GEJ identification to improve diagnostic precision for BE.

Variability in the Endoscopic Diagnosis of BE
In 2006, Sharma et al. [18] proposed the Prague C & M criteria as a standardized approach to the endoscopic assessment of BE whereby the GEJ is defined as the proximal limit of the gastric folds. This classification system quantifies BE using the circumferential length and the maximal extent of the visible columnar epithelium above the GEJ. This method has shown robust interobserver and intraobserver agreement for long-segment BE. However, it is less reliable for very short segments, particularly when the BE length is less than 1 cm [21, 27].
The Asian Barrett Consortium evaluated the reproducibility of the Prague C & M classification among endoscopists in Asia using video-based assessment [28]. Although good agreement was observed with longer BE segments, detection accuracy for ultra-short-segment BE (USSBE) remained poor. This may be attributable to the variability of the gastric fold appearance, which can shift with the insufflation degree and the patient’s breathing status. When appropriate air insufflation is performed, the proximal folds often extend toward the SCJ, potentially obscuring short segments.
Norita et al. [29] conducted a prospective cohort study in Japan to assess the cancer risk associated with BE and reported annual progression rates of 0.47% and 0.31% for segments ≥2 cm and ≥3 cm, respectively, suggesting a relatively low risk profile among the Japanese population. However, the malignant potential of BE segments smaller than 1 cm remains uncertain [16]. Fukuda et al. [30] performed a retrospective cohort analysis of more than 9,000 patients and found that nearly 46% had BE; of these patients, more than one-third (36.4%) had USSBE. During a median follow-up period of approximately 4.5 years, only one case (0.0068%) of esophageal adenocarcinoma developed from USSBE, indicating a low carcinogenic risk among this subgroup [30]. These findings indicate that intensive surveillance may not be warranted for USSBE despite its frequent identification using gastric fold-based GEJ definitions.

Endoscopic Diagnosis of BE: Consensus from the Kyoto International Consensus Conference Report
The Kyoto International Consensus Conference addressed longstanding inconsistencies regarding the endoscopic diagnosis of BE and particularly focused on the conceptual definition of BE, the choice of anatomical landmarks for the GEJ, and the technical methods of enhancing diagnostic reliability [31].
The Kyoto International Consensus group proposed the following revised conceptual definition of BE: “BE is a condition in which a metaplastic columnar mucosa, predisposed to neoplasia, replaces the squamous epithelium of the distal esophagus.” Notably, this definition removes the requirement for intestinal metaplasia and no longer stipulates a minimum length of the columnar epithelium. The rationale stems from evidence that intestinal metaplasia may be inconsistently detected in biopsy samples.
The Kyoto International Consensus strongly supports using the distal end of the palisade vessels (DEPVs) as the preferred endoscopic landmark for the GEJ because of its anatomical fidelity and relative stability. Unlike the proximal end of gastric folds, which may shift with respiration, insufflation, or inflammation, the DEPV is anatomically located at the distal end of the lower esophageal sphincter and correlates well with histological and surgical findings.
The Kyoto International Consensus provided the following detailed guidance for visualizing the DEPV and proximal end of gastric folds effectively: use high-resolution white light imaging (WLI), with or without IEE, in both forward and retroflexed views under controlled air insufflation to identify the DEPV, and ensure minimal and standardized air insufflation to identify the proximal end of gastric folds because excessive insufflation can flatten the folds and excessive deflation may distort the anatomy or create pseudofolds.

Advanced Endoscopic Imaging for the Diagnosis of BE
WLI has been the primary modality for visualizing the GEJ and detecting BE. However, its limitations have become increasingly apparent. Conventional WLI often fails to distinguish subtle mucosal changes, especially in cases of USSBE and those involving inflammation or poor insufflation control. In particular, visualization of the DEPV is hindered in many cases when WLI is used alone [32].
WLI has traditionally served as the standard modality for evaluating the BE (Fig. 3a). More recently, IEE has revolutionized the diagnostic approach to BE. Among these technologies, IEE, narrow-band imaging, linked color imaging (Fig. 3b), blue laser imaging (Fig. 3c), and red dichromatic imaging are widely available.
A multicenter evaluation by Ono et al. [33] assessed DEPV visualization using multiple IEE modalities and demonstrated that red dichromatic imaging had the highest high-confidence detection rate (54.1%), linked color imaging provided significantly better reproducibility than narrow-band imaging and WLI, and narrow-band imaging and blue laser imaging had limited utility for GEJ landmark identification despite their usefulness for detecting vascular irregularities. These results underscore the importance of modality selection based on the diagnostic purpose, such as landmark recognition or dysplasia detection.
Recent advancements in artificial intelligence (AI), particularly in deep learning and machine learning, have demonstrated diagnostic capabilities that may exceed those of human experts in image analysis. The field of gastrointestinal endoscopy has experienced remarkable integration of AI, and interest in its application for BE and related neoplasia is increasing. Notably, Ali et al. [34] reported a novel AI-based system that reconstructs the esophageal anatomy in three dimensions using endoscopic images and classifies BE segments according to the Prague C & M criteria. Although this study was limited to a single-center experience, the findings highlighted the potential for AI to standardize diagnostic assessments and facilitate precise biopsies in areas with suspected dysplasia.

QIs in Endoscopic Surveillance of BE
Given the limitations in current clinical practice, recent attention has focused on defining and implementing QIs for BE endoscopy. In the context of BE, four main QIs have been proposed: (1) neoplasia detection rate (NDR), (2) Barrett’s inspection time, and (3) adherence to the Seattle biopsy protocol [35]. NDR is defined as the proportion of patients with histologically confirmed high-grade dysplasia or esophageal adenocarcinoma at index endoscopy. This metric parallels the adenoma detection rate used in colonoscopy. Reported NDRs range from 4.9% to 7.0% [36–38].
Longer inspection time of the BE segment has been shown to increase the detection of dysplasia. Studies suggest a minimum of 1 min per centimeter of BE length as a target inspection time [39, 40]. However, inspection time definitions vary – some measure from scope insertion to withdrawal, while others measure only the segment inspection time [38, 39]. Although current data support longer inspection, further randomized controlled trials are needed to validate this as a reliable QI.
The Seattle protocol recommends systematic 4-quadrant biopsies every 1–2 cm of the BE segment. By adhering to this protocol, the detection of LGD and high-grade dysplasia is significantly enhanced in comparison to targeted biopsies conducted independently [41–43]. Nevertheless, adherence rates remain low – only ∼49% globally [44], and even lower among general endoscopists and in regions such as Asia, where short-segment BE predominates and IEE is favored.

Conclusion
BE is a key precursor to esophageal adenocarcinoma; however, its diagnosis remains hampered by inconsistent definitions, technical limitations, and variable adherence to standards. The Kyoto International Consensus has provided clarity regarding anatomical landmarks and offered a pragmatic framework for redefining BE. Furthermore, emerging technologies such as IEE and AI have enabled visualization of previously elusive anatomical structures and reduced diagnostic variability.
To improve clinical outcomes, a more unified and evidence-based approach must be adopted. This includes the use of DEPV-based definitions of the GEJ when feasible and the application of IEE and AI technologies for real-time diagnostic guidance. Moreover, harmonization of surveillance strategies on a global scale is crucial to ensuring consistent care and effective cancer prevention. With continued collaboration among international experts and the integration of technological innovation, a highly standardized and accurate diagnostic procedure for BE is possible.

Conflict of Interest Statement
The author declares no conflicts of interest related to this article.

Funding Sources
The author declares no funding was obtained or used for this study.

Author Contributions
C.K. contributed to the conceptualization, review, and editing.