Regional heterogeneity in the DNA content of human gliomas.

Abstract

Regional heterogeneity causes significant errors in the histologic classification and grading of gliomas, but little is known about its implications for other modalities. Attempts to predict glioma behavior using flow cytometry (FCM) have yielded contradictory results, possibly due to regional heterogeneity in DNA content, a recognized phenomenon that has not been evaluated systematically. The authors used FCM to analyze the DNA content of 353 regions from 18 resected human gliomas. Five to 60 regions were sampled from each tumor, and the topographic relationships of ploidy, proliferative activity (S-phase fraction [SPF]), and histologic features were established. Most tumors demonstrated heterogeneity among regions in the number and relative sizes of aneuploid populations and/or proliferative activity. The degree of heterogeneity increased with tumor grade. Heterogeneity in histologic features, ploidy, and proliferative activity appeared to vary independently, except for a significant association between the presence of mitotic figures and the SPF of individual regions (P < 0.0001). The clustering of regions with similar percentages of aneuploid cells supported the concept of local clonal expansion. Gliomas express significant regional heterogeneity in ploidy and proliferative activity and this will have an adverse effect on the usefulness of their analysis. However, the independent variation of ploidy, proliferative activity, and histologic features suggests that the use of multiple analyses may allow more accurate prediction of glioma patient survival. Regional heterogeneity appears be a fundamental property of malignant gliomas; systematic studies to determine its effects on the diagnostic usefulness of other new methods used to evaluate gliomas are indicated.