Skip to main content
NCBI home page
United European Gastroenterology Journal logoLink to United European Gastroenterology Journal
. 2026 Oct 26;14(8):e70309. doi: 10.1002/ueg2.70309

Optimising Expert Pathology Referral in Barrett's Oesophagus Dysplasia Through p53‐Guided Triage: A Nationwide Real‐World Cohort Study

Ylva A Weeda 1,2, Myrtle J van der Wel 1,3, Lodewijk A Brosens 4, Ariadne H A G Ooms 5,6, Hans Halfwerk 1, Maartje T P M Houben 7, Onno J de Boer 1, Michail Doukas 6, Fiebo ten Kate 3, Gursah Kats‐Ugurlu 8, Jaap van der Laan 9, Ineke van Lijnschoten 10, Lindsey Oudijk 6,11, Rachel S van der Post 12, Mihaela G Raicu 13, Wim Timens 8, Bert Timmer 8, Mike Visser 14, Jacques J G Bergman 7, Lucas Duits 7, Sybren L Meijer 1,2,✉
PMCID: PMC13711024  PMID: 42935139

ABSTRACT

Background

Histopathological interpretation of Barrett's oesophagus‐associated dysplasia remains variable, and although guidelines recommend expert pathology review, specialist capacity is increasingly constrained. We assessed the real‐world diagnostic impact of national expert review and tested whether p53‐IHC identifies subgroups in which review is high‐ or low‐yield.

Methods

The Dutch Oesophageal Pathology Panel reviewed consecutive BE cases submitted from 39 hospitals (2015–2024). Digitised haematoxylin and eosin‐stained and p53‐IHC slides were assessed by experienced gastrointestinal pathologists. Panel diagnoses required ≥ 75% agreement or consensus meeting review. Referring and panel diagnoses were compared at the patient level, with biopsy‐level analyses assessing diagnostic shifts according to referring‐centre p53‐IHC status (wild‐type, aberrant or equivocal).

Results

The cohort included 1146 consecutive patients and 2187 biopsy levels, each assessed by a median of 12 panel pathologists. Expert review changed 46.9% of referring diagnoses (downgrading 24.7% and upgrading 22.2%). Indefinite for dysplasia (IND) was the most unstable diagnostic category, with 73.4% of cases reclassified. Among 1731 biopsy levels with reported p53‐IHC at referral, p53‐IHC strongly stratified review yield. Low‐grade or high‐grade dysplasia (LGD/HGD) with aberrant p53‐IHC was confirmed in 91.7% of biopsy levels. In contrast, LGD/HGD with wild‐type or equivocal p53‐IHC showed substantially higher downgrade rates (53.6% and 32.8%), with odds ratios for diagnostic adjustment of 12.7 (95% CI: 8.2–19.7) and 5.4 (95% CI: 2.9–10.0) versus aberrant p53‐IHC.

Conclusion

Expert review substantially altered BE dysplasia diagnoses, supporting current guideline recommendations. P53‐IHC markedly stratified review yield: IND and dysplasia with wild‐type or equivocal p53‐IHC remain high‐priority categories, whereas aberrant p53‐IHC dysplasia was rarely reclassified. A p53‐guided triage strategy could help optimise the use of expert pathology capacity, but prospective validation is needed before clinical implementation.

Keywords: Barrett's oesophagus, biomarker‐guided triage, dysplasia, expert pathology review, p53 immunohistochemistry, resource utilization

Key Summary

  • Summarise the established knowledge on this subject:

    • ◦

      Barrett's oesophagus‐associated dysplasia shows substantial interobserver variability, leading to inconsistent patient management

    • ◦

      Guidelines recommend expert pathology review for IND and dysplasia, but specialist gastrointestinal pathology capacity is increasingly constrained

    • ◦

      p53‐IHC is an established adjunct in BE dysplasia diagnostics, but its potential role in guiding referral allocation has not been evaluated at a national scale

  • What are the significant and/or new findings of this study?

    • ◦

      In this largest BE expert review series to date (1146 cases from 39 hospitals), expert review altered nearly half of referring diagnoses, supporting current guideline recommendations

    • ◦

      p53‐IHC markedly stratified review yield: IND and dysplasia with wild‐type or equivocal p53‐IHC were frequently reclassified, whereas dysplasia with aberrant p53‐IHC was rarely altered

    • ◦

      These findings support a potential role for p53‐IHC in guiding expert review allocation, with the greatest potential to reduce low‐yield referrals and concentrate expert pathology resources where diagnostic uncertainty is highest

    • ◦

      Prospective validation, including assessment of diagnostic safety and clinical outcomes, is needed before clinical implementation.

graphic file with name UEG2-14-e70309-g005.webp

1. Introduction

Barrett's oesophagus (BE) is a premalignant condition in which the distal squamous mucosa is replaced by metaplastic columnar epithelium, in response to chronic gastroesophageal reflux [1, 2]. Progression through metaplasia, stages of dysplasia and eventually oesophageal adenocarcinoma (EAC) underpins surveillance and treatment strategies [3, 4, 5]. For this reason, accurate risk stratification at the point of biopsy assessment is central to clinical management and decision‐making [5, 6].

The current diagnostic gold standard remains haematoxylin and eosin (H&E)‐based histological assessment according to the modified Vienna classification [7]. Nonetheless, distinguishing true dysplasia from reflux‐associated regenerative atypia and inflammation remains susceptible to interobserver variation [8, 9, 10]. This uncertainty has direct clinical consequences: overdiagnosis may expose patients to unnecessary surveillance or treatment, whereas underdiagnosis may delay intervention. For this reason, national and international guidelines recommend expert pathology review for dysplastic BE [11, 12].

Expert review improves diagnostic reliability but is also a constrained healthcare resource. Specialist gastrointestinal pathologists are scarce, second opinions can delay management, and growing surveillance programs will further increase pathology workload. A practical question therefore emerges: can objective ancillary biomarkers identify which biopsies would benefit most from expert review and which are unlikely to change after centralized assessment?

p53 immunohistochemistry (p53‐IHC) is the leading candidate biomarker for this purpose. Aberrant p53 immunolabelling is associated with TP53 pathway disruption and increased risk of neoplastic progression, and p53‐IHC can improve diagnostic reproducibility in BE dysplasia [13, 14, 15, 16, 17, 18]. European and British guidelines recognize p53‐IHC as an adjunct, but its role in routine workflows remains unsettled because definitions of aberrancy vary and evidence for clinical benefit has been inconsistent [5, 12, 19].

The Dutch Oesophageal Pathology Panel (DEPP) provides a unique real‐world setting in which to address this key question towards rationally designed surveillance programs. The DEPP is a national digital pathology review panel comprising trained, benchmarked gastrointestinal pathologists from Barrett expert centres, established to provide consistent expert review of BE‐associated dysplasia across the Netherlands [20, 21]. Here, we analysed 1146 consecutive patient cases from 39 hospitals over a 10‐year period. Our primary aim was to quantify the impact of expert review of patient‐level diagnosis. Our second aim was to test whether p53‐IHC status at referral can guide efficient allocation of expert review by identifying diagnostic subgroups with high or low likelihood of clinically relevant revision.

2. Materials and Methods

2.1. Patients and Samples

This study included 1146 consecutive revision requests assessed by the DEPP from 39 hospitals in the Netherlands between February 2015 and December 2024. Patients underwent esophagogastroduodenoscopy for BE screening and surveillance. Referring centres were accredited pathology departments participating in external quality assurance programs. Cases were submitted primarily for diagnostic uncertainty, most commonly suspected indefinite for dysplasia (IND) or low‐grade dysplasia (LGD), either as whole patient cases or selected biopsy levels. Biopsies deemed irrelevant for establishing a definitive case‐level diagnosis, such as most NDBE or squamous samples, were generally not submitted.

Selected histopathological slides were sent to the central pathology laboratory at the Amsterdam University Medical Centre. Slides were digitized at 20x objective, initially using an Olympus (Olympus, Tokyo, Japan) scanner, and subsequently a Philips IntelliSite Ultra Fast (Philips, Eindhoven, The Netherlands) scanner. The digitized images were assessed for adequate quality according to standard digital workflow protocols and uploaded to a digital slide viewer (initially Digital Slidebox 4.5, Leica Biosystems and later Cirdan PathXL Tutor) to provide access for all participating panel pathologists.

2.2. Expert Panel Review

The DEPP currently consists of 13 active gastrointestinal pathologists from nine leading BE expert centres with 5–30 years of experience. Across the study period, 22 pathologists participated. Prior to becoming panel members, pathologists completed mandatory training using benchmark criteria to promote consistent BE dysplasia interpretation [20, 21].

Histopathological identification of dysplasia was assessed according to the modified Vienna criteria for gastrointestinal neoplasms and grouped as non‐dysplastic BE (NDBE), IND, LGD and high‐grade dysplasia (HGD). Cases reported as suspicious for adenocarcinoma were included in the HGD category [7]. Referring hospital diagnoses were extracted from the original pathology reports, based on H&E, and p53‐IHC when available. Panel pathologists independently reviewed cases and were blinded to one another's assessments. Panel diagnoses required agreement among ≥ 75% of participating pathologists, with a minimum of four raters. In cases reviewed by three pathologists, unanimous agreement was required. Cases not meeting these thresholds were discussed during a monthly online consensus meeting. The meeting leader (SM) presented the individual assessments, after which the participating pathologists discussed the case and reached consensus on the final diagnosis for patient management.

2.3. p53‐IHC Triage Analysis

To assess whether p53‐IHC could guide efficient review allocation, we analysed all biopsy levels within this cohort for which referring‐centre p53‐IHC status was available, allowing diagnostic shifts between referring‐centre and panel diagnosis to be assessed across p53‐IHC categories. Biopsy levels without a recorded p53‐IHC result were excluded.

Referring‐centre p53‐IHC status was classified from the original pathology report as wild‐type, aberrant, or equivocal. Aberrant patterns included overexpression, null‐pattern staining, or double‐clone patterns. Wild‐type staining was defined as weak to moderate heterogeneous nuclear expression. Overexpression was defined as strong nuclear staining in at least one fully involved crypt, or equivalent, depending on the plane of section, with staining intensity exceeding that of the basal squamous layer or adjacent epithelium when present. Null pattern was defined as complete absence of nuclear immunolabelling. DC referred to the coexistence of NM and OE patterns within the same biopsy. Reports describing p53‐IHC as unclear or ambiguous were classified as ‘equivocal’. Detailed classification criteria and p53‐IHC usage within the DEPP panel have been described previously [22].

2.4. Statistical Analysis

Agreement between referring hospital and panel diagnoses was summarized using proportions of unchanged diagnoses, upgrades and downgrades. To quantify the yield of review by p53‐IHC subgroup, we calculated the proportion of diagnostic adjustments after panel review for aberrant, wild‐type, and equivocal p53‐IHC categories. Odds ratios (OR) with 95% confidence intervals (CIs) were derived from contingency tables using aberrant p53‐IHC as the reference category. Visualizations and computations were performed using Python (version 3.7; Python Software Foundation, Wilmington, DE, USA) and RStudio (version V.4.4.3; Copyright 2025 The R Foundation for Statistical Computing).

3. Results

3.1. Nationwide Expert Review Cohort

The DEPP reviewed 1146 consecutive patient cases submitted from 39 teaching and community hospitals. Cases included a median of two endoscopic biopsy levels (range: 1–8), yielding 2187 evaluated endoscopic biopsy levels. Each case was evaluated by a median of 12 DEPP pathologists (range: 3–20), and each pathologist assessed a median of 548.5 (range: 80–1143) cases. Of the 1146 cases reviewed, 406 (35.4%) required consensus meeting discussion to establish the final diagnosis (Figure 1). At referral, patient‐level diagnoses were NDBE in 35 cases (3.1%), IND in 425 (37.1%), LGD in 592 (51.7%) and HGD in 94 (8.2%). After DEPP review, final diagnoses were NDBE in 226 cases (19.7%), IND in 188 (16.4%), LGD in 542 (47.3%), and HGD in 190 (16.6%). Overall, the panel altered 46.9% of referring diagnoses, downgrading 24.7% and upgrading 22.2% (Figure 2).

FIGURE 1.

FIGURE 1

Open in a new tab

Study design and national expert review workflow. The DEPP reviewed 1146 consecutive patient cases from 39 hospitals, comprising 2187 biopsy levels with H&E and, when available, p53‐IHC. Referring hospital diagnoses were compared with final panel diagnoses assigned by at least 75% agreement or by consensus meeting.

FIGURE 2.

FIGURE 2

Open in a new tab

Patient‐level diagnostic shifts after expert panel review. The contingency table shows referring and panel diagnoses across 1146 patient cases. The diagonal represents unchanged diagnoses, while off‐diagonal cells represent downgrades or upgrades. Overall, 46.9% of diagnoses were altered, including downgrading in 24.7% and upgrading in 22.2%. Column‐wise percentages are also displayed, representing the ratio of alterations within each referring hospital diagnostic category. Percentages may not add up to 100% due to rounding.

3.2. Diagnostic Impact by Referral Category

IND was the most unstable referral diagnosis, and among 425 IND cases, 73.4% were altered by panel review: 33.2% were downgraded to NDBE, while 40.2% were upgraded to LGD (32.0%) or HGD (8.2%). LGD diagnoses were changed in 34.3% of 592 cases: downgraded to NDBE in 9.3%, to IND in 11.7%, and upgraded to HGD in 13.3%. Among 94 HGD cases, 19.2% were changed, mostly to LGD (16.0%) and less often to IND (3.2%). NDBE diagnoses were the most stable, with 14.3% of 35 cases altered, including upgrades to IND in 8.6% and LGD in 5.7% (Figures 2 and 3a–b).

FIGURE 3.

FIGURE 3

Open in a new tab

(a) Distribution and direction of patient‐level diagnostic changes. Sankey (a) and stacked bar (b) visualizations show how referring hospital diagnoses align with or shift after panel review, highlighting the high reclassification rate for IND and the lower change rates for NDBE and HGD. The exact numbers and percentages have been omitted in these graphs for clarity, but they reflect the values and column‐wise percentages shown in Figure 2.

These findings confirm the substantial real‐world impact of expert review, but also show that review yield varies markedly by the diagnostic category. In particular, the high reclassification rate for IND supports continued expert review of this ambiguous category, whereas many LGD and HGD changes occurred within the dysplasia spectrum and may not always alter immediate management in the same way as downgrades to NDBE. The results for 2187 individual endoscopic biopsy levels can be found in the Supporting Information S1 (Figure S1 in Supporting Information S1).

3.3. p53‐IHC Identifies High‐Yield and Low‐Yield Review Groups

For the p53‐IHC subgroup analysis, 456 biopsy levels were excluded due to absent or unreported p53‐IHC status, leaving 1731 levels for analysis: 820 with wild‐type p53‐IHC, 727 with aberrant p53‐IHC, and 184 with equivocal p53‐IHC (Figure 4).

FIGURE 4.

FIGURE 4

Open in a new tab

This schematic diagram graphically displays the study workflow and p53‐IHC subgroup analysis. Endoscopic biopsy levels without p53‐IHC or unreported p53‐IHC status were excluded (n = 456), leaving a total of 1731 biopsy levels: 820 wild‐type, 727 aberrant, and 184 equivocal. Referring hospital sign‐outs are shown for each p53‐IHC subset.

P53‐IHC status strongly modified the direction and yield of expert review. NDBE with wild‐type p53‐IHC remained unchanged in 297 of 343 biopsy levels (86.6%); 13.4% were upgraded to IND (7.3%), LGD (5.8%) or HGD (0.3%). Referral IND was unstable across p53‐IHC categories. IND with wild‐type p53‐IHC was most often downgraded to NDBE (176 of 324; 54.3%), whereas IND with aberrant p53‐IHC was most often upgraded to LGD or HGD (101 of 160; 63.1%). IND with equivocal p53‐IHC was also frequently reclassified (74 of 96; 77.1%) with both downgrades (31.2%) and upgrades (45.9%) (Figure 5a–c).

FIGURE 5.

FIGURE 5

Open in a new tab

Biopsy‐level diagnostic shifts stratified by referring hospital p53‐IHC status. Heat maps show referring hospital versus panel diagnoses (Vienna classification) within each p53‐IHC subgroup as reported by the referring hospital: wild‐type (a), aberrant (b), and equivocal (c). Percentages are column‐wise and represent alteration rates within each referring diagnostic category. Darker shades the colour scale reflects these percentages, with darker shades indicating higher percentages. This enables clear visualization of the frequency and predominant direction of diagnostic shifts across p53‐IHC subgroups. Percentages may not add up to 100% due to rounding.

The clearest triage signal emerged in biopsies referred to as LGD or HGD. LGD/HGD with aberrant p53‐IHC was confirmed by the panel in 505 of 551 biopsy levels (91.7%), leaving only 46 biopsy levels (8.3%) with a diagnostic adjustment (Figure 5b). In contrast, LGD/HGD with wild‐type p53‐IHC and equivocal p53‐IHC had substantially higher downgrade rates of 53.6% and 32.8%, corresponding to ORs of 12.7 (95% CI 8.2–19.7) and 5.4 (95% CI 2.9–10.0), respectively, compared with aberrant p53‐IHC (Figure 5a–c).

From a resource‐utilization perspective, these data suggest that expert review might have been low yield for 29.2% of eligible biopsy levels (505 of 1731) and 36.3% of referred patient cases (419 of 1146): cases referred as dysplasia with at least one aberrant p53‐IHC biopsy level and no subsequent diagnostic change after expert review. Conversely, referral IND and LGD/HGD with wild‐type or equivocal p53‐IHC represented higher‐yield categories in which expert review frequently changed diagnosis (Figure 6).

FIGURE 6.

FIGURE 6

Open in a new tab

p53‐IHC as a triage marker for expert review. Sankey diagrams show biopsy‐level transitions for referral IND (a) and LGD/HGD (b). IND with wild‐type p53‐IHC was commonly downgraded, whereas IND with aberrant p53‐IHC was commonly upgraded. LGD/HGD with aberrant p53‐IHC was largely confirmed after expert review, supporting its potential role as a lower‐yield review category. EQ = equivocal; WT = wild‐type.

To assess the potential impact of missing p53‐IHC data at referral, diagnostic shifts for the 456 excluded biopsy levels are presented separately in the Supporting Information S1: (Figure S2).

4. Discussion

In this nationwide, real‐world cohort of 1146 consecutive BE patient cases reviewed by a central expert pathology panel, review changed nearly half of the referring diagnoses. These results reinforce the clinical value of expert review for BE dysplasia, particularly for IND. When p53‐IHC status was combined with referral diagnosis, it identified subgroups in which review was either highly informative or comparatively low yield. The main implementation‐relevant finding is that referral of LGD/HGD with aberrant p53‐IHC was rarely altered after review. In contrast, IND or LGD/HGD with wild‐type p53‐IHC, and LGD/HGD with equivocal p53‐IHC were frequently reclassified. This supports a biomarker‐guided triage model in which expert capacity is concentrated where diagnostic uncertainty is highest.

The 46.9% overall change rate confirms that expert pathology review remains an important safeguard in BE‐associated dysplasia. The most clinically relevant downgrades were from IND or LGD/HGD to NDBE (17.1%), because these may alter surveillance intensity, referral pathways, and consideration of endoscopic therapy. Upgrades from IND to LGD/HGD (14.9%) also matter because they clarify the risk and can accelerate treatment planning. Earlier studies reported high rates of LGD downgrading after expert review and helped establish current guideline recommendations [4, 23, 24]. Our cohort differs in scale, era, and setting: it reflects post‐guideline national practice across 39 hospitals and showed lower LGD downgrading (21.0%) than older Dutch cohorts. This may reflect improved diagnostic awareness, DEPP‐led training and increased p53‐IHC use. Consistent with this, we observed a temporal shift from predominantly LGD to predominantly IND referrals over the study period (Supporting Information S1: Figure S3). This suggests that as diagnostic practice evolves, review strategies should be reassessed accordingly rather than applied uniformly over time.

p53‐IHC is usually discussed as a diagnostic adjunct, but these data suggest an additional role: as a biomarker for allocating expert pathology capacity. IND remained the highest‐priority category, with diagnoses altered in 73.4% of patient‐level cases and 75.8% of biopsy levels. LGD/HGD with wild‐type or equivocal p53‐IHC likewise showed frequent diagnostic adjustment, warranting continued prioritisation for expert review. In contrast, 91.7% of LGD/HGD biopsy levels with aberrant p53‐IHC, expert review did not change the diagnosis. Our findings support a more targeted approach to expert review of BE dysplasia. Universal review treats all dysplasia labels as equally uncertain, whereas our data demonstrate that the need for confirmation differs substantially: diagnostic uncertainty is concentrated in IND and dysplasia with wild‐type or equivocal p53‐IHC, while aberrant p53‐IHC dysplasia has a lower incremental yield from expert confirmation. LGD/HGD with aberrant p53‐IHC could therefore be considered as a potential lower‐priority review category when morphology is convincing and immunolabelling technically reliable. Omitting mandatory expert review for dysplasia with aberrant p53‐IHC would have affected 29.2% (505/1731) of eligible biopsy levels and 36.3% (419/1146) of referred patient cases without changing the final panel outcome. These findings provide a rationale for a p53‐guided triage strategy, in which expert review is prioritised for cases with the greatest likelihood of diagnostic reclassification.

Our data reinforce the importance of expert pathology review in BE‐associated dysplasia, while suggesting that the need for specialist assessment may differ according to diagnostic category and p53‐IHC status. This may have practical implications for the organisation of expert reviews. Although expert review is advocated in international guidelines, its organisation and accessibility vary across healthcare systems. In the Netherlands, expert review is standardised through a centralised panel, but establishing and maintaining such a system requires substantial specialist capacity and organisational resources and may not be feasible in all healthcare settings. In contrast, p53‐IHC is a relatively accessible, inexpensive and scalable ancillary technique with reproducible interpretation that could be incorporated into routine diagnostic workflows. A p53‐guided triage approach could therefore help prioritise cases in which expert review is most likely to alter the diagnosis, thereby concentrating limited specialist capacity where it is most needed rather than replacing expert review.

Our multicentre cohort included referrals from 39 hospitals, capturing considerable variation in routine pathology practice. Nevertheless, the magnitude of the observed triage effect may differ between healthcare systems because of differences in referral practices, diagnostic thresholds, and the use and interpretation of p53‐IHC. Diagnostic assessment followed the Vienna classification, providing a common international framework for BE assessment. The threshold for diagnosing IND, however, rather than NDBE, or LGD, varies considerably between pathologists worldwide, even within expert settings. This variability may be further influenced by local diagnostic practices and the use of ancillary p53‐IHC. Thus, differences in diagnostic thresholds should be considered when evaluating the applicability of p53‐guided triage across healthcare systems.

A recent multicentre cohort using the same p53‐IHC criteria demonstrated that aberrant p53 expression was strongly associated with progression risk, regardless of histopathological diagnosis [25]. These findings support a prognostic role for p53‐IHC beyond morphological classification and are consistent with our findings that p53‐IHC may have a complementary role in prioritising cases for expert histopathological review. At this stage, this p53‐guided triage approach should be considered implementation‐informing rather than practice‐changing, as prospective validation is still needed to confirm diagnostic safety and assess its applicability across different healthcare settings.

This study has some limitations. First, this was a consultation cohort enriched for diagnostically difficult cases, and findings should not be extrapolated to an unselected BE surveillance population. Second, p53‐IHC was not performed or reported for 456 of 2187 biopsy levels (20.9%), and these were excluded from the p53‐IHC subgroup analysis. Most of these biopsy levels were referred to as NDBE (62.7%, 286/456), while 12.3% (56/456) were referred to as IND and 25.0% (114/456) as LGD/HGD. The reasons for not performing or reporting p53‐IHC were unknown, and we can therefore not determine whether the availability of p53‐IHC was associated with case complexity or diagnostic uncertainty at referral. Consequently, selection bias in the p53‐IHC subgroup analysis cannot be excluded. Third, the analysis used p53‐IHC status reported by referring hospitals rather than central re‐staining of all biopsies, reflecting real‐world practice but allowing variation in staining and interpretation across centres. In our previous analysis of the overall referred cohort, 8.6% of biopsy levels reported as aberrant p53 by referring hospitals were classified as wild‐type by the expert panel, demonstrating that discordant p53 classification occurs in routine practice [22] In the present study, among the biopsy levels referred to LGD/HGD with aberrant p53‐IHC that remained LGD/HGD after expert review, 95.6% were confirmed as aberrant p53 by the expert panel, while only 1.2% were classified as wild‐type. This suggests that p53 misclassification is unlikely to have substantially contributed to the low‐risk subgroup. Nevertheless, internationally harmonised criteria for overexpression, null pattern, double‐clone pattern, and focal aberrancy are needed before triage pathways can be standardized [19, 22, 26]. Fourth, although aberrant p53‐IHC dysplasia was rarely altered, the 8.3% adjustment rate can be impactful and drive overdiagnosis. Triage should therefore not mean automatic bypass of review in every case. Poor‐quality staining, discordance between morphology and p53‐IHC, tangential biopsy orientation, marked inflammation, very focal abnormalities, or uncertainty in local interpretation should still trigger expert review. Finally, this study did not correlate referring or panel diagnoses with clinical outcomes. Future studies should therefore assess whether p53‐guided review triage preserves cancer‐prevention outcomes while improving efficiency.

In summary, expert review substantially alters BE dysplasia diagnoses in real‐world practice, but p53‐IHC can identify where review is most likely to change the outcome. IND and dysplasia with wild‐type or equivocal p53‐IHC should remain high‐priority categories. Dysplasia with aberrant p53‐IHC appears to have a lower yield for mandatory review and may be suitable for selective triage after prospective validation.

Author Contributions

Ylva A. Weeda: conceptualization, data curation, formal analysis, investigation, methodology, project administration, software, validation, visualization, writing – original draft. Myrtle J. van der Wel: conceptualization, data curation, methodology, project administration, resources, supervision, writing – review and editing. Lodewijk A. Brosens: resources, writing – review and editing. Ariadne H. A. G. Ooms: resources, writing – review and editing. Hans Halfwerk: data curation, resources, writing – review and editing. Maartje T. P. M. Houben: writing – review and editing. Onno J. de Boer: data curation, resources, writing – review and editing. Michail Doukas: conceptualization, data curation, methodology, project administration, resources, supervision, writing – review and editing. Fiebo ten Kate: resources, writing – review and editing. Gursah Kats‐Ugurlu: resources, writing – review and editing. Jaap van der Laan: resources, writing – review and editing. Ineke van Lijnschoten: resources, writing – review and editing. Lindsey Oudijk: resources, writing – review and editing. Rachel S. van der Post: resources, writing – review and editing. Mihaela G. Raicu: resources, writing – review and editing. Wim Timens: resources, writing – review and editing. Bert Timmer: resources, writing – review and editing. Mike Visser: resources, writing – review and editing. Jacques J. G. Bergman: writing – review and editing. Lucas Duits: writing – review and editing. Sybren L. Meijer: conceptualization, data curation, funding acquisition, methodology, project administration, resources, supervision, validation, visualization, writing – original draft.

Funding

This work was supported by grants from the ‘Koningin Wilhelmina Fonds’ (KWF 150591) and ‘Maag Lever Darm Stichting’ (MLDS WO 21‐25). The funding agencies had no role in the design of the study, in the collection, analyses or interpretation of data, in the writing of the manuscript or in the decision to publish the results.

Ethics Statement

The study was approved by the Ethics Committee Amsterdam (approval number: 2025.0121).

Consent

Patient consent was waived because of the retrospective design nature of this study and the use of pseudo‐anonymised data.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting Information S1

UEG2-14-e70309-s001.docx (466.9KB, docx)

Acknowledgements

During the preparation of this manuscript, the author(s) used AI‐assisted technology to improve language clarity, grammar, and readability and to support programing tasks related to data visualization. All figures were based on author‐derived data and code, with AI assistance limited to coding support. The figures were not AI‐generated. No AI tools were used for data interpretation or scientific decision‐making. The author(s) critically reviewed and revised all content and take full responsibility for the accuracy and integrity of the work.

Data Availability Statement

The datasets generated and/or analysed during the current study contain sensitive patient information and are not publicly available. Access to pseudonymized data may be granted following a formal data access request and approval in accordance with institutional policies, informed consent, and applicable data protection regulations.

References

  • 1. Shaheen N. J., Falk G. W., Iyer P. G., and Gerson L. B., “ACG Clinical Guideline: Diagnosis and Management of Barrett's Esophagus,” American Journal of Gastroenterology 111, no. 1 (2016): 30–50, 10.1038/ajg.2015.322. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Liu W., Hahn H., Odze R. D., and Goyal R. K., “Metaplastic Esophageal Columnar Epithelium Without Goblet Cells Shows DNA Content Abnormalities Similar to Goblet Cell‐Containing Epithelium,” American Journal of Gastroenterology 104, no. 4 (2009): 816–824, 10.1038/ajg.2009.85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Schoofs N., Bisschops R., and Prenen H., “Progression of Barrett's Esophagus Toward Esophageal Adenocarcinoma: An Overview,” Annals of Gastroenterology 30, no. 1 (2017): 1–6, 10.20524/aog.2016.0091. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Duits L. C., Phoa K. N., Curvers W. L., et al., “Barrett's Oesophagus Patients With Low‐Grade Dysplasia Can Be Accurately Risk‐Stratified After Histological Review by an Expert Pathology Panel,” Gut 64, no. 5 (2015): 700–706, 10.1136/gutjnl-2014-307278. [DOI] [PubMed] [Google Scholar]
  • 5. Weusten B., Bisschops R., Dinis‐Ribeiro M., et al., “Diagnosis and Management of Barrett Esophagus: European Society of Gastrointestinal Endoscopy (ESGE) Guideline,” Endoscopy 55, no. 12 (2023): 1124–1146, 10.1055/a-2176-2440. [DOI] [PubMed] [Google Scholar]
  • 6. Critchley‐Thorne R. J., Duits L. C., Prichard J. W., et al., “A Tissue Systems Pathology Assay for High‐Risk Barrett's Esophagus,” Cancer Epidemiology, Biomarkers & Prevention 25, no. 6 (2016): 958–968, 10.1158/1055-9965.epi-15-1164. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Schlemper R. J., Riddell R. H., Kato Y., et al., “The Vienna Classification of Gastrointestinal Epithelial Neoplasia,” Gut 47, no. 2 (2000): 251–255, 10.1136/gut.47.2.251. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Kerkhof M., van Dekken H., Steyerberg E. W., et al., “Grading of Dysplasia in Barrett's Oesophagus: Substantial Interobserver Variation Between General and Gastrointestinal Pathologists,” Histopathology 50, no. 7 (2007): 920–927, 10.1111/j.1365-2559.2007.02706.x. [DOI] [PubMed] [Google Scholar]
  • 9. Montgomery E., Bronner M. P., Goldblum J. R., et al., “Reproducibility of the Diagnosis of Dysplasia in Barrett Esophagus: A Reaffirmation,” Human Pathology 32, no. 4 (2001): 368–378, 10.1053/hupa.2001.23510. [DOI] [PubMed] [Google Scholar]
  • 10. Vennalaganti P., Kanakadandi V., Goldblum J. R., et al., “Discordance Among Pathologists in the United States and Europe in Diagnosis of Low‐Grade Dysplasia for Patients With Barrett's Esophagus,” Gastroenterology 152, no. 3 (2017): 564–570.e4, 10.1053/j.gastro.2016.10.041. [DOI] [PubMed] [Google Scholar]
  • 11. Wani S., Rubenstein J. H., Vieth M., and Bergman J., “Diagnosis and Management of Low‐Grade Dysplasia in Barrett's Esophagus: Expert Review From the Clinical Practice Updates Committee of the American Gastroenterological Association,” Gastroenterology 151, no. 5 (2016): 822–835, 10.1053/j.gastro.2016.09.040. [DOI] [PubMed] [Google Scholar]
  • 12. Fitzgerald R. C., di Pietro M., Ragunath K., et al., “British Society of Gastroenterology Guidelines on the Diagnosis and Management of Barrett's Oesophagus,” Gut 63, no. 1 (2014): 7–42, 10.1136/gutjnl-2013-305372. [DOI] [PubMed] [Google Scholar]
  • 13. van der Wel M. J., Duits L. C., Pouw R. E., et al., “Improved Diagnostic Stratification of Digitised Barrett's Oesophagus Biopsies by p53 Immunohistochemical Staining,” Histopathology 72, no. 6 (2018): 1015–1023, 10.1111/his.13462. [DOI] [PubMed] [Google Scholar]
  • 14. Skacel M., Petras R. E., Rybicki L. A., et al., “p53 Expression in Low Grade Dysplasia in Barrett's Esophagus: Correlation With Interobserver Agreement and Disease Progression,” American Journal of Gastroenterology 97, no. 10 (2002): 2508–2513, 10.1111/j.1572-0241.2002.06032.x. [DOI] [PubMed] [Google Scholar]
  • 15. Kaye P. V., Haider S. A., Ilyas M., et al., “Barrett's Dysplasia and the Vienna Classification: Reproducibility, Prediction of Progression and Impact of Consensus Reporting and p53 Immunohistochemistry,” Histopathology 54, no. 6 (2009): 699–712, 10.1111/j.1365-2559.2009.03288.x. [DOI] [PubMed] [Google Scholar]
  • 16. Kastelein F., Biermann K., Steyerberg E. W., et al., “Aberrant p53 Protein Expression is Associated With an Increased Risk of Neoplastic Progression in Patients With Barrett's Oesophagus,” Gut 62, no. 12 (2013): 1676–1683, 10.1136/gutjnl-2012-303594. [DOI] [PubMed] [Google Scholar]
  • 17. Redston M., Noffsinger A., Kim A., et al., “Abnormal TP53 Predicts Risk of Progression in Patients With Barrett's Esophagus Regardless of a Diagnosis of Dysplasia,” Gastroenterology 162, no. 2 (2022): 468–481, 10.1053/j.gastro.2021.10.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Tokuyama M., Geisler D., Deitrick C., et al., “Use of p53 Immunohistochemistry in Conjunction With Routine Histology Improves Risk Stratification of Patients With Barrett's Oesophagus During Routine Clinical Care,” Histopathology 77, no. 3 (2020): 481–491, 10.1111/his.14143. [DOI] [PubMed] [Google Scholar]
  • 19. Srivastava A., Appelman H., Goldsmith J. D., Davison J. M., Hart J., and Krasinskas A. M., “The Use of Ancillary Stains in the Diagnosis of Barrett Esophagus and Barrett Esophagus‐Associated Dysplasia: Recommendations From the Rodger C. Haggitt Gastrointestinal Pathology Society,” American Journal of Surgical Pathology 41, no. 5 (2017): e8–e21, 10.1097/pas.0000000000000819. [DOI] [PubMed] [Google Scholar]
  • 20. van der Wel M. J., Duits L. C., Klaver E., et al., “Development of Benchmark Quality Criteria for Assessing Whole‐Endoscopy Barrett's Esophagus Biopsy Cases,” United European Gastroenterology Journal 6, no. 6 (2018): 830–837, 10.1177/2050640618764710. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Klaver E., van der Wel M., Duits L., et al., “Performance of Gastrointestinal Pathologists Within a National Digital Review Panel for Barrett's Oesophagus in the Netherlands: Results of 80 Prospective Biopsy Reviews,” Journal of Clinical Pathology 74, no. 1 (2021): 48–52, 10.1136/jclinpath-2020-206511. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Weeda Y. A., Issa S. M., Halfwerk H., et al., “Diagnostic Impact and Reproducibility of p53 Immunohistochemistry in Barrett's Oesophagus: Results of the Dutch Esophageal Pathology Panel (DEPP),” Histopathology (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Curvers W. L., ten Kate F. J., Krishnadath K. K., et al., “Low‐Grade Dysplasia in Barrett's Esophagus: Overdiagnosed and Underestimated,” American Journal of Gastroenterology 105, no. 7 (2010): 1523–1530, 10.1038/ajg.2010.171. [DOI] [PubMed] [Google Scholar]
  • 24. Huhta H., Melkko J., Kuopio T., Karttunen T. J., and Helminen O., “Risk of Progression in Barrett's Esophagus Based on Diagnoses of General and Gastrointestinal Pathologists. A Retrospective Case‐Control Study From Northern and Central Finland,” Scandinavian Journal of Gastroenterology 57, no. 9 (2022): 1024–1029, 10.1080/00365521.2022.2063033. [DOI] [PubMed] [Google Scholar]
  • 25. Zellenrath P. A., Provoost N., Jansen J. C. M., et al., “Aberrant p53 Expression, a Stronger Risk Factor for Neoplastic Progression in Barrett's esophagus Than Confirmed Low Grade Dysplasia: A Large Multicenter Prospective Cohort Study,” Endoscopy 58, no. 9 (2026): 937–946, 10.1055/a-2863-0566. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Chornenkyy Y., Vyas M., and Deshpande V., “The Future is Now: Advancing p53 Immunohistochemistry in Barrett's Oesophagus and Its Implication for the Everyday Pathologist,” Histopathology (2025). [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information S1

UEG2-14-e70309-s001.docx (466.9KB, docx)

Data Availability Statement

The datasets generated and/or analysed during the current study contain sensitive patient information and are not publicly available. Access to pseudonymized data may be granted following a formal data access request and approval in accordance with institutional policies, informed consent, and applicable data protection regulations.

Articles from United European Gastroenterology Journal are provided here courtesy of Wiley

RESOURCES