Infiltración peritumoral en glioblastomas multiformes, estudio de la angiogénesis tumoral mediante correlación inmuno-histo-radiológica

Abstract

INTRODUCTION Glioblastoma multiforme (GBM) is the most aggressive and frequent primary malignant tumor of the central nervous system in adults. The histological heterogeneity of GBM is not limited to the tumor, but also extends to the peritumoral tissue. Actually, the tumor infiltration front, classically under-studied and without repercussion in the stage gradation and therapeutic decision making, seems to be a key place to understand GBM growth and a target for new immunological-based therapies. The peritumoral tissue is defined in Magnetic Resonance (MR) as the brain area around the tumor, hyperintense in T2WI sequences and without contrast enhancement. However, we need to define correlation between pathology and advanced MR techniques to better define this area. Overall survival (OS) in patients with GBM varies significantly. It is essential to be able to predict the evolution by independent prognostic factors that allow us to select the best therapy and increase survival. OBJECTIVES - To evaluate the peritumoral infiltration of high-grade tumors according to advanced MR imaging techniques: diffusion weighted imaging (DWI), diffusion tensor imaging (DTI), multivoxel spectroscopy and perfusion weighted imaging. - To assess the alteration in the peritumoral cerebral microvasculature with respect to the pericytes. To study the role of perycites in vascular malformation. - Search for reproducible radiological markers in tumor infiltration that can be correlated with the immunohistochemical alterations studied in the surgical specimen. - To know if radiological markers at the time of diagnosis can be correlated with the overall survival of GBM patients. MATERIAL AND METHODS In 70 patients with confirmed GBM, radiological study was performed using advanced MR techniques. Relative tumor and peritumoral values were obtained with respect to the contralateral white matter and applied to maps of apparent diffusion coefficient (ADC) in DWI, fractional anisotropy (FA) in DTI, multiple metabolite ratios and cerebral blood volume (CBV) in perfusion weigthed imaging. In 28 cases, the area of greatest radiological peritumoral infiltration was selected and peritumoral tissue samples were obtained during surgery for immunohistochemical study. The histological variables studied were: perivascular infiltration of GFAP tumor cells (glial fibrillary acidic protein), vascular malformation (VM) as tortuous vessels with glomeruloid appearance, presence of immunopositive NG2 cells (pericytes) in the vascular wall and infiltration of macrophage cells (immunopositive for CD68) in the peritumoral parenchyma. A statistical analysis of the radiological and histological variables was performed. RESULTS The advanced MR techniques are able to identify the peritumoral infiltration, by differentiating the tumor tissue from the peritumoral tissue significantly, except for the choline / creatine ratio (CHO / CR) which could nevertheless help us to delimit the infiltration front. Alterations of the cerebral microvasculature detected in the surgical specimen have shown significant correlation between histological markers of VM and GFAP and the NG2 marker or pericyte; therefore, we can postulate that these histological changes would be conditioned by the pericyte. The radiological markers that correlated with the alterations detected in the surgical specimen were: GFAP with tumoral and peritumoral relative (r) ADC values as infiltration markers, VM with tumoral rCBV and peritumoral CHO / CR ratio, NG2 with tumoral rCBV and peritumoral normalized values of lipids and CD68 with tumoral rCBV and peritumoral choline / N-acetylaspartate (CHO / NAA) ratio. Markers that turned out to be independent prognostic factors for survival were: from the radiological variables the peritumoral rADC values, the peritumoral CHO / NAA ratio and the normalized values of tumor lipids at 1.3ppm, and from the histological variables, the GFAP and the VM. Age also turned out to be an independent prognostic factor.