Premium
Risk for esophageal neoplasia in B arrett's esophagus patients with mucosal changes indefinite for dysplasia
Author(s) -
Horvath Bela,
Singh Prabhdeep,
Xie Hao,
Thota Prashanthi N,
Allende Daniela S,
Pai Rish K,
Patil Deepa T,
Plesec Thomas P,
Goldblum John R,
Liu Xiuli
Publication year - 2015
Publication title -
journal of gastroenterology and hepatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.214
H-Index - 130
eISSN - 1440-1746
pISSN - 0815-9319
DOI - 10.1111/jgh.12696
Subject(s) - medicine , dysplasia , esophagus , barrett's esophagus , gastroenterology , biopsy , univariate analysis , incidence (geometry) , population , esophageal disease , adenocarcinoma , cancer , multivariate analysis , physics , environmental health , optics
Background and Aim Patients with B arrett's esophagus ( BE ) are at increased risk for esophageal adenocarcinoma ( EAC ) and therefore require surveillance. Biopsies are classified as indefinite for dysplasia ( IND ) when the significance of epithelial abnormalities is uncertain due to inflammation or sampling. Our aim was to characterize the neoplastic risk of IND in BE patients and to identify predictors of neoplastic risk. Methods Our pathology database from 1992 to 2007 was searched for BE and IND . Progression rates were calculated and univariate analysis was performed to identify predictors for neoplasia progression in BE ‐ IND patients. Results Among 85 patients who had a follow‐up ( FU ) biopsy within 1 year, 11 (12.9%) patients had prevalent neoplasia (seven low‐grade dysplasia [ LGD ], two high‐grade dysplasia [ HGD ], and two EAC ). Among 82 patients who did not have prevalent neoplasia but had ≥ 1 year FU , 17 progressed to dysplasia (14 LGD , 3 HGD ) and 2 developed EAC during a mean FU period of 59 months. The incidence of neoplasia ( LGD , HGD , or EAC ) and advanced neoplasia ( HGD + EAC ) was 4.5 and 1.2 cases per 100 patient‐years, respectively. Longer length of BE and multi‐focal IND on index biopsy were associated with progression to neoplasia. Conclusion Patients with BE ‐ IND carry a significant risk of harboring prevalent dysplasia, but the risk of incident dysplasia is similar to the general BE population. The length of BE and the multifocal IND might tentatively help to identify a patient subpopulation at higher risk of neoplastic progression before more definitive data becomes available.