{"title":"Morphologic Fault Lines in Distinguishing Gastric Cancer pTis From pT1a in Endoscopic Submucosal Dissection Specimens: An Interobserver Study.","authors":"Rika Omote, Kohei Okura, Eita Kumagai, Ryosuke Ichihara, Takayuki Kodama, Akira Ishikawa, Naoki Sano, Hiroyuki Abe, Minako Fujiwara, Hirokazu Taniguchi","doi":"10.1111/pin.70132","DOIUrl":null,"url":null,"abstract":"<p><p>The 16th edition of the Japanese Classification of Gastric Carcinoma newly distinguishes pTis from pT1a to improve international translatability with the WHO classification and UICC TNM system. However, the histologic boundary between intraepithelial neoplasia and lamina propria invasion remains difficult in routine practice. We retrospectively evaluated interobserver agreement in 50 gastric endoscopic submucosal dissection lesions randomly selected from 238 lesions previously diagnosed as intramucosal carcinoma under the 15th edition criteria. For each lesion, two representative histologic images were independently reviewed by 10 pathologists, who classified each lesion as pTis or pT1a according to predefined architectural criteria. Complete agreement was obtained in 21 lesions, including 10 unanimously diagnosed as pTis and 11 as pT1a, whereas 29 showed disagreement. Overall agreement was moderate, with a Fleiss' kappa value of 0.483. Discordant lesions involved fused or anastomosing glands, cribriform-like structures, crawling or lateral expansion-like patterns, and invasion mimics related to erosion, artifact, or mucosal remodeling. These findings indicate that distinguishing true invasive architecture from morphologic mimics is central to reproducible implementation of the revised classification.</p>","PeriodicalId":19806,"journal":{"name":"Pathology International","volume":"76 6","pages":"e70132"},"PeriodicalIF":1.9000,"publicationDate":"2026-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Pathology International","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1111/pin.70132","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"PATHOLOGY","Score":null,"Total":0}
引用次数: 0
Abstract
The 16th edition of the Japanese Classification of Gastric Carcinoma newly distinguishes pTis from pT1a to improve international translatability with the WHO classification and UICC TNM system. However, the histologic boundary between intraepithelial neoplasia and lamina propria invasion remains difficult in routine practice. We retrospectively evaluated interobserver agreement in 50 gastric endoscopic submucosal dissection lesions randomly selected from 238 lesions previously diagnosed as intramucosal carcinoma under the 15th edition criteria. For each lesion, two representative histologic images were independently reviewed by 10 pathologists, who classified each lesion as pTis or pT1a according to predefined architectural criteria. Complete agreement was obtained in 21 lesions, including 10 unanimously diagnosed as pTis and 11 as pT1a, whereas 29 showed disagreement. Overall agreement was moderate, with a Fleiss' kappa value of 0.483. Discordant lesions involved fused or anastomosing glands, cribriform-like structures, crawling or lateral expansion-like patterns, and invasion mimics related to erosion, artifact, or mucosal remodeling. These findings indicate that distinguishing true invasive architecture from morphologic mimics is central to reproducible implementation of the revised classification.
期刊介绍:
Pathology International is the official English journal of the Japanese Society of Pathology, publishing articles of excellence in human and experimental pathology. The Journal focuses on the morphological study of the disease process and/or mechanisms. For human pathology, morphological investigation receives priority but manuscripts describing the result of any ancillary methods (cellular, chemical, immunological and molecular biological) that complement the morphology are accepted. Manuscript on experimental pathology that approach pathologenesis or mechanisms of disease processes are expected to report on the data obtained from models using cellular, biochemical, molecular biological, animal, immunological or other methods in conjunction with morphology. Manuscripts that report data on laboratory medicine (clinical pathology) without significant morphological contribution are not accepted.