|
Back to 2015 Annual Meeting Program
Inter-Observer Variability in the Interpretation of Endoscopic Mucosal Resection Specimens
Stephanie G. Worrell*1, Joshua a. Boys1, John G. Vallone1, Parakrama Chandrasoma1, Daniel S. OH1, Jeffrey a. Hagen1, Corey Johnson2, Michal J. Lada3, Christy M. Dunst4, Brian E. Louie2, Jeffrey H. Peters3, Thomas J. Watson3, Steven R. Demeester1
1Keck School of Medicine of Univeristy of Southern California, Los Angeles, CA; 2Surgery, Swedish, Seattle, WA; 3University of Rochester, Rochester, NY; 4Surgery, The Oregon Clinic, Portland, OR
Introduction: Endoscopic resection (ER) has revolutionized the therapy of superficial esophageal adenocarcinoma. Accurate pathologic evaluation of specimen is critical to determine the risk of lymph node metastases. The aim of this study was to assess the inter-observer variability in pathologic assessment of ER specimens.
Methods: A retrospective study was performed of ER specimens from 4 US institutions done for superficial esophageal adenocarcinoma. The original endoscopic resection slides were re-reviewed by two blinded expert (study) gastrointestinal pathologists with extensive experience interpreting ER specimens. None of the original reports were generated by either of the study pathologists. The original pathology report was compared to the findings by the study pathologists, and the interpretation of the study pathologists was compared to each other for the depth of tumor invasion, tumor grade and the presence of lymphovascular invasion (LVI).
Results: Twenty-five endoscopic resection specimens were re-evaluated. Review by the study pathologists differed in 12 of the 25 specimens (48%) for the depth of tumor invasion. In 10/12 cases (83%) the lesion was over-staged on the original pathology report compared to the review by the study pathologists. There was concordance between the study pathologists in 23/25 cases (92%) on the depth of invasion. The original interpretation of tumor grade differed from the study pathologists in 8/18 cases (44%) and in 3/16 cases (19%) for LVI. There was concordance between the expert pathologists in 20/25 cases (69%) for tumor grade and in 16/18 cases (89%) for LVI.
Conclusions: There was a high rate of discordance (48%) between expert pathologists and the original pathology assessment of the depth of tumor invasion in ER specimens. Most of the discrepancy was related to over-staging on the original interpretation, and typically this was related to mistaking the deep layer of a duplicated muscularis mucosa for the muscularis propria. Expert review of endoscopic resection specimens is recommended to ensure patients receive the appropriate treatment for a potentially curable superficial esophageal adenocarcinoma.
Back to 2015 Annual Meeting Program
|
