BACKGROUND: Borderline gastrointestinal stromal tumors (GISTs) are intermediate tumors between benign and malignant variants; however, the clinical and pathological features of borderline GISTs remain poorly defined. This study aimed to characterize GISTs and to identify a set of borderline criteria for practical use. METHODS: Medical records and specimens of 840 patients from 12 hospitals were retrospectively examined. Totally 485 and 76 patients with any of the parameters predictive of either malignant or benign tumors were excluded. The Kaplan-Meier method was used to calculate disease-free survival and overall survival rates. RESULTS: Among the remaining 279 borderline GIST patients, 223 were followed up for 1 to 31.48 years. Two patients developed local recurrence, and both were cured by subsequent operations alone. The 5-year disease-free survival and overall survival rates were 99% and 100%, respectively. Morphologically, borderline GISTs typically exhibited moderate cellularity, and subsets of them also showed moderate atypia, low mitotic activities, or large tumor size. According to the National Institutes of Health (NIH) consensus criteria, the risk levels of the 279 GISTs were classified to be very low to high. However, the disease-free survival rates were not significantly different among these risk groups (P = 0.681). CONCLUSIONS: The proposed borderline GIST criteria in the current study may complement the existing NIH criteria, based primarily on tumor size and mitotic count, in the evaluation of the biological behaviors of GISTs. Since a subset of borderline GISTs with high risk level showed favorable outcome, the introduction of the borderline GIST system may avoid overdiagnosis and over therapy.
BACKGROUND:Borderline gastrointestinal stromal tumors (GISTs) are intermediate tumors between benign and malignant variants; however, the clinical and pathological features of borderline GISTs remain poorly defined. This study aimed to characterize GISTs and to identify a set of borderline criteria for practical use. METHODS: Medical records and specimens of 840 patients from 12 hospitals were retrospectively examined. Totally 485 and 76 patients with any of the parameters predictive of either malignant or benign tumors were excluded. The Kaplan-Meier method was used to calculate disease-free survival and overall survival rates. RESULTS: Among the remaining 279 borderline GIST patients, 223 were followed up for 1 to 31.48 years. Two patients developed local recurrence, and both were cured by subsequent operations alone. The 5-year disease-free survival and overall survival rates were 99% and 100%, respectively. Morphologically, borderline GISTs typically exhibited moderate cellularity, and subsets of them also showed moderate atypia, low mitotic activities, or large tumor size. According to the National Institutes of Health (NIH) consensus criteria, the risk levels of the 279 GISTs were classified to be very low to high. However, the disease-free survival rates were not significantly different among these risk groups (P = 0.681). CONCLUSIONS: The proposed borderline GIST criteria in the current study may complement the existing NIH criteria, based primarily on tumor size and mitotic count, in the evaluation of the biological behaviors of GISTs. Since a subset of borderline GISTs with high risk level showed favorable outcome, the introduction of the borderline GIST system may avoid overdiagnosis and over therapy.