Preoperative Predictive Factors for Gastrointestinal Stromal Tumors: Analysis of 375 Surgically Resected Gastric Subepithelial Tumors

Research output: Contribution to journalArticlepeer-review

35 Scopus citations

Abstract

Background: Gastrointestinal stromal tumors (GISTs) and non-GIST subepithelial tumors (SETs) account for about 75 and 25 % of gastric hypoechoic SETs ≥2 cm, respectively. Therefore, identifying preoperative predictive factors for GISTs are required to refine surgical indications. Methods: We performed a retrospective review of 375 surgically resected gastric hypoechoic SETs ≥2 cm. Demographic data and tumor characteristics based on upper endoscopy and CT findings were compared between GIST and non-GIST SETs originating from muscularis propria layer (leiomyomas, Schwannomas, glomus tumors, and ectopic pancreas). Results: In cardia, leiomyomas were found twice more frequently than GISTs (63.6 versus 31.8 %). Perilesional lymph node enlargement (PLNE) was found only in patients with GIST or Schwannomas. Patients with GIST showed a significantly lower rate of PLNE than those with Schwannomas (3.5 versus 29.0 %). In multivariate analysis, tumor site outside cardia (odds ratio, 9.157), absence of PLNE (odds ratio, 11.519), old age, large tumor size, exophytic growth pattern, and ulceration or dimpling were identified as independent preoperative predictive factors for GISTs versus non-GIST SETs. Conclusions: The effort for preoperative pathologic diagnosis such as endosonography-guided tissue sampling might be positively considered for SETs at cardia and SETs with PLNE where the possibility of GIST is low.

Original languageEnglish
Pages (from-to)631-638
Number of pages8
JournalJournal of Gastrointestinal Surgery
Volume19
Issue number4
DOIs
StatePublished - Apr 2015

Keywords

  • Gastrointestinal stromal tumor
  • Stomach
  • Subepithelial tumor

Fingerprint

Dive into the research topics of 'Preoperative Predictive Factors for Gastrointestinal Stromal Tumors: Analysis of 375 Surgically Resected Gastric Subepithelial Tumors'. Together they form a unique fingerprint.

Cite this