Fine Needle Aspiration of Gastrointestinal Stromal Tumors.

Publication/Presentation Date

1-1-2001

Abstract

OBJECTIVE: Gastrointestinal stromal tumors (GISTs) are uncommon mesenchymal tumors of the gastrointestinal tract. Fine needle aspiration (FNA) is one option for diagnosing GISTs before surgery. This study was designed to evaluate the clinical utility of FNA in the diagnosis of GISTs.

STUDY DESIGN: FNAs from 19 GISTs originating in the stomach, small bowel and colon obtained from 1988 to 1998 were studied. Immunocytochemistry was performed on 12 cases. The GISTs were classified as benign, borderline and malignant, according to location, size, mitotic activity and clinical outcome.

RESULTS: Benign (three) and borderline (five) GISTs were all spindle cell type; malignant GISTs included five spindle cell type and six epithelioid type. Most smears contained abundant cellular material. Benign and borderline GISTs of spindle cell type tended to have cells arranged in tightly cohesive clusters, while malignant GISTs were more likely to exhibit loosely cohesive groups with many single cells, occasional nuclear pleomorphism, hyperchromasia and irregular nuclear contours. Epithelioid-type GISTs mimicked adenocarcinoma. Mitoses were seldom observed in either type. CD117 (KIT protein product) was demonstrated by immunocytochemistry in 9 cases, CD34 in 11, desmin in 3, S-100 protein in 2 and smooth muscle actin in 6 cases.

CONCLUSION: FNA can be used to diagnose GISTs as spindle cell and epithelioid types, but cytomorphology alone cannot be used to assess malignant potential. Immunocytochemical staining for CD117 is helpful in confirming the diagnosis. Care must be taken to differentiate epithelioid-type GISTs from adenocarcinoma.

Volume

45

Issue

1

First Page

9

Last Page

17

ISSN

0001-5547

Disciplines

Medical Pathology | Pathology

PubMedID

11213513

Department(s)

Department of Pathology and Laboratory Medicine, Pathology Laboratory Medicine Faculty

Document Type

Article

Link to Full Text

Share

COinS