Accuracy of pathologic interpretation of Barrett's metaplasia and dysplasia by community hospital general pathologists

M. Alikhan, O. W. Cummings, T. Ulbright, D. Rex

Research output: Contribution to journalArticle

1 Scopus citations

Abstract

Correct interpretation of Barrett's metaplasia is essential as histology guides subsequent surveillance and/or surgical intervention. Aim: To study the accuracy of pathologic interpretation of Barrett's metaplasia in clinical practice by 20 community hospital general pathologists. Methods: We identified 5 histologic slides representing different types of columnar epithelium from the esophagus, each chosen because it had a straight-forward histologic interpretation. We then submitted these slides to blinded review by 20 randomly selected general pathologists in community practice in Indiana. There were 3 cases of Barrett's metaplasia (1 with no dysplasia, 1 with low-grade dysplasia (LGD), and 1 with high-grade dysplasia (HGD), and 2 cases of gastric metaplasia (1 fundic-type and 1 cardia-type). Results: Barrett's metaplasia was correctly identified in 95% of 60 readings (as Barrett's metaplasia in 82% and with the correct descriptor "specialized columnar epithelium" (SCE) or "intestinal metaplasia" in another 13%). Of these 57 correct readings of Barrett's metaplasia, the presence or absence of dysplasia was commented upon in 38(67%) of the readings. The overall presence of dysplasia was correctly identified in 71%(27 of 38 readings). HGD was correctly identified in 28%(5 of 18 readings) and misinterpreted as invasive adenocarcinoma in 22%(4 of 18 readings) and LGD in 28%(5 of 18 readings). The term "moderate dysplasia" was inappropriately used in the remaining 22%(4 of 18 readings). LGD was correctly identified in 35%(7 of 20 readings) and misinterpreted as HGD in 25%(5 of 20 readings) and no dysplasia in 25%(5 of 20 readings). The term "moderate dysplasia" was inappropriately used in the remaining 15%(3 of 20 readings). Barrett's metaplasia with no dysplasia was correctly identified in 37%(7 of 19 readings) and misinterpreted as LGD in 37% (7 of 19 readings) and invasive adenocarcinoma in 5%(1 of 19 readings). Gastric metaplasia was correctly identified in 65% of 40 readings and misinterpreted as Barrett's metaplasia in the remaining 35%. Of the 26 correct readings of gastric metaplasia, the diagnosis of Barrett's metaplasia was still given in 31%(8 of 26 readings). Summary: 1) Barrett's metaplasia was identified in over 90% of cases and the presence or absence of dysplasia was provided in about two-thirds of these cases. 2) When present, dysplasia was recognized over 70% of the time but the degree of dysplasia was often misinterpreted. 3) Gastric metaplasia was often misinterpreted as Barrett's metaplasia. Conclusion: Community pathologists should be educated to refrain from using the term Barrett's metaplasia in the absence of SCE. The presence and degree of dysplasia in Barrett's metaplasia, interpreted by community pathologists, is frequently unreliable and should be commonly subjected to review.

Original languageEnglish (US)
Pages (from-to)AB64
JournalGastrointestinal endoscopy
Volume47
Issue number4
StatePublished - Dec 1 1998

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Gastroenterology

Fingerprint Dive into the research topics of 'Accuracy of pathologic interpretation of Barrett's metaplasia and dysplasia by community hospital general pathologists'. Together they form a unique fingerprint.

  • Cite this