Format

Send to

Choose Destination
Histol Histopathol. 2009 Jun;24(6):737-47. doi: 10.14670/HH-24.737.

Predictive values of clinical and pathological parameters for malignancy of gastrointestinal stromal tumors.

Author information

1
Department of Pathology, Zhongshan Hospital, Fudan University, Shanghai, China. houyingyong@hotmail.com

Abstract

Gastrointestinal stromal tumors (GISTs) possess a wide spectrum of biological properties, from indolent to highly aggressive. In this study, we evaluated a set of clinical and pathological parameters for their predicative values for malignancy of GISTs by retrospective reviews of tumor specimens and their relevant medical records from 840 patients. All GIST cases were first assigned as malignant if they met any of the following criteria: gross spreads, including liver metastassis and/or peritoneal dissemination, microscopic spreads, including lymph node metastasis, infiltrations to vascular, fat, nerve and muscularis mucosal tissues, or relapse. The remaining cases were recorded as biological behavior uncertain. This initial assignment revealed a set of five morphological features to be associated with malignancy. They were: mitotic counts greater than 10 per 50HPFs (P<0.0001), muscularis propria infiltration (P<0.0001), coagulative necrosis (P<0.0001), perivascular growth pattern (P=0.005), and severe nuclear atypia (P=0.014). Therefore, a new classification system, including criteria of 2 gross spreads, 5 microscopic spreads, and 5 histopathological parameters was developed. All the GIST cases were re-classified into a group of 485 malignant tumors, and a group of 355 nonmalignant tumors. Patient follow-up data revealed 5-year disease-free and overall survival rates as high as 99.3% and 100% for the nonmalignant group, but low rates of 43.9% and 59.7% for the malignant group. These results demonstrated a correlation of the new classification with clinical outcomes. Therefore, this set of 12 parameters has predictive values for malignancy of GISTs, and is potentially useful in the grading of the tumors.

PMID:
19337972
DOI:
10.14670/HH-24.737
[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Sercrisma International S.L.
Loading ...
Support Center