Optimal processing of ESD specimens to avoid pathological artifacts.

ABSTRACT

BACKGROUND: En bloc endoscopic submucosal dissection (ESD) has been recently introduced as a treatment for precancerous/neoplastic gastrointestinal conditions. The aim of the present study was histological assessment of en bloc ESD specimens. METHODS: Fifty-three ESD specimens were positioned over a cellulose acetate support (40 specimens; 12 from the upper gastrointestinal tract and 28 from the lower gastrointestinal tract) or pinned with nails on polystyrene or cork (13 specimens; 7 from the upper gastrointestinal tract and 6 from the lower gastrointestinal tract). We cut consecutive 2 mm-thick sections stained with hematoxylin and eosin. From the first and the last sections, we obtained a second slide, after a 180° rotation and re-embedding. The quality of ESD samples was scored as inadequate, suboptimal and adequate, based on the amount of crushing, shearing and stretching artifacts that were scored from 0 (absent) to 2 (diffuse or maximum). From the sum of these we obtained a global artifact score (GAS). RESULTS: Removed lesions were adenocarcinoma (5 cases), neuroendocrine tumor (NET) G1 (1 case), premalignant conditions, including adenomatous polyps (41 cases) and hyperplastic lesions (6 cases). A positive deep surgical margin was found in 8/53 cases (15%) high- and low-grade dysplastic glands were detected in 5 cases, low-grade adenocarcinoma in 2, and NET cells in 1. Dysplastic glands were detected in the lateral surgical margins of 12 ESD specimens (23%). Among the ESD specimens positioned on the cellulose acetate support, apart from the modifications due to electrocoagulation, 2 (5%) showed shearing modifications. In the group of ESD specimens fixed with nails, 5 (38%) showed shearing, 10 (77%) crushing artifacts, 11 (85%) stretching and 11 (85%) multiple holes caused by the nails. On the basis of these data all histological specimens from ESD on cellulose acetate were adequate (GAS 0-1).However, in the group of ESD fixed with nails, 1 was adequate (GAS 0), 11 suboptimal (GAS 2-5) and 1 inadequate (GAS 6). CONCLUSIONS: Specific devices including cellulose support and adequate sampling blocks can be helpful to perform accurate histological assessment of ESD specimens after en bloc ESD for precancerous/neoplastic gastrointestinal lesions, with complete analysis of the status of the margins and the entirely en bloc evaluation of the lesion.