Tumor vessel biology in pediatric intracranial ependymoma

OBJECT: This study aimed to characterize the pediatric intracranial ependymoma vasculature in terms of angiogenic activity and maturation status so as to provide indications for the applicability of vessel-targeted therapy in cases of pediatric intracranial ependymoma.

METHODS: Tumor samples obtained in patients with ependymomas were immunohistochemically (double) stained for Ki 67/CD34, caspase 3a/CD34, vascular endothelial growth factor (VEGF)-A, -B, -C, -D, collagen Type IV, and smooth muscle actin to determine microvessel density, tumor and endothelial cell proliferation and apoptotic fraction, the relative expression of VEGF family members, and the coverage of the tumor endothelial cells by basal membrane and pericytes. Messenger RNA expression of angiopoietin-1 and -2 was analyzed by real-time reverse transcriptase polymerase chain reaction. These data were compared with those obtained in a glioblastoma series.

RESULTS: Despite a low endothelial cell turnover, the microvessel density of ependymomas was similar to that of glioblastomas. In ependymomas the expression of VEGF-A was within the range of the variable expression in glioblastomas. The staining intensities of VEGF-B, -C, and -D in ependymomas were significantly lower (p < 0.001). The expression of angiopoietin-1 was higher in ependymomas than in glioblastomas (p = 0.03), whereas angiopoietin-2 expression was similar. The coverage of tumor endothelial cells with basal membrane and pericytes was more complete in ependymomas (p = 0.009 and p = 0.022, respectively).

CONCLUSIONS: The ependymoma vasculature is relatively mature and has little angiogenic activity compared with malignant gliomas. Therefore, the window for vessel normalization as a therapeutic aim might be considered small. However, the status of the tumor vasculature may not be a reliable predictor of treatment effect. Therefore, possible benefits of antiangiogenic treatment cannot be excluded beforehand in patients with ependymomas.