Published in

American Society of Clinical Oncology, Journal of Clinical Oncology, 4_suppl(34), p. 39-39, 2016

DOI: 10.1200/jco.2016.34.4_suppl.39

Links

Tools

Export citation

Search in Google Scholar

The influence of tumor location, histology, and patient ethnicity on the accuracy of EUS in staging gastric and gastroesophageal junction (GEJ) cancers.

This paper was not found in any repository, but could be made available legally by the author.
This paper was not found in any repository, but could be made available legally by the author.

Full text: Unavailable

Red circle
Preprint: archiving forbidden
Orange circle
Postprint: archiving restricted
Red circle
Published version: archiving forbidden
Data provided by SHERPA/RoMEO

Abstract

39 Background: Endoscopic ultrasound (EUS) is a routine staging test performed in patients with gastric and GEJ cancers and is highly valuable in treatment planning. The reported accuracy varies from 40% to 90%. We aim to investigate the staging performance of EUS at our institution and hypothesize that EUS accuracy can be influenced by tumor location, histology, and patient ethnicity. Methods: We performed a retrospective analysis of patients diagnosed with gastric and GEJ cancer between January 2006 and June 2015 who had EUS proceeded directly by surgery. T and N staging by EUS were analyzed for their sensitivity, specificity, and positive (PPV) and negative predictive values using surgery as the gold standard. The following variables were evaluated: tumor histology (intestinal or diffuse type), location (GEJ, cardia, body, or antrum), and patient ethnicity (Asian versus non-Asian). Results: Over 46 patients who met the inclusion criteria, EUS had T and N accuracy of 47.8% and 58.7%, respectively (p <.05). EUS had high specificity (88%) but low sensitivity (38%) for N staging. In predicting T staging, EUS was most accurate when the tumor was located in the body (72%, n = 7) and least accurate in the cardia (33%, n = 6) with 46% (n = 13) in the GEJ and 45% (n = 20) in the antrum. EUS tended to upstage the T value when a tumor is located in the body or GEJ, and downstage it in the cardia or antrum. The PPV using EUS for T1, T2, T3, and T4 were 89%, 22%, 45%, and 57%, respectively. The PPV using EUS for N staging was 73%. EUS was more accurate for T in Asians (55%, n = 33) than in non-Asians (31%, n = 13), and in intestinal type (56%) than diffuse (36%). For N staging, it was most accurate in the body (86%) and least accurate in the GEJ (50%). It was also more accurate in diffuse tumors (79%) than intestinal (52%) and more accurate in non-Asians (77%) than in Asians (53%). Conclusions: EUS as a common tool in preoperative staging may be inaccurate when performed in community hospitals. The predictive rate appears to vary in Asian vs non-Asian patients and in intestinal vs diffuse tumors. It is most predictive when the tumor is T1 and located in the gastric body. A prospective study is needed to confirm these findings.