Although classic neuroscience teaching divides the brain into neurons and glia, it is increasingly clear that the extracellular environment, or stroma, plays a pivotal role in neurodevelopment and disease. The importance of stroma as a target of anticancer therapy has come to light in the era of antiangiogenic drugs. In a phase II trial, Vredenburgh et al¹ tested bevacizumab, a recombinant monoclonal antibody against vascular endothelial growth factor (VEGF), plus a conventional cytotoxic chemotherapy irinotecan to treat glioblastomas. His group noted a 57% response rate and a 46% 6-month progression free survival (PFS).¹ These results are striking when compared to treatment responses from conventional cytotoxic chemotherapies, which only have a 6% response rate and a 15% 6-month PFS.² The synergistic response from bevacizumab plus irinotecan results from vascular normalization. By sequestering VEGF, bevacizumab causes regression of hyperpermeable tumor vasculatures, which leads to increased tissue oxygenation, decreased interstitial pressure, and increased cytotoxic chemotherapy delivery into glioblastomas.³ In Vredenburgh’s study, the MRI was notable for resolution of gadolinium enhancement and decreased FLAIR signals.¹ In another trial using AZD2171, a pan-VEGF receptor tyrosine kinase inhibitor that blocks proliferation signaling to endothelial cells, dynamic susceptibility contrast and diffusion tensor MRI respectively demonstrated a drop in permeability constant K_trans and reappearance of white matter fibers when cerebral edema resolved.⁴ Both studies demonstrated antiangiogenic drug effect on vascular normalization.

Bevacizumab may impair our ability to visualize glioblastomas on MRI. This is because response assessment by Macdonald’s criteria depends on leakage of gadolinium from hypermeable tumor vasculatures. Bevacizumab plus irinotecan may decrease vascular permeability without affecting the underlying tumor. To clarify this issue, Chen et al⁵ performed positron emission tomography (PET) in glioblastoma patients using ¹⁸F-fluorothymidine (FLT), a radioactive nucleoside that is preferentially taken up by dividing tumor cells, before and after bevacizumab plus irinotecan treatment. They demonstrated a 47% metabolic response rate and a 65% 6-month survival, and their metabolic responders had a 3-fold increase in survival than nonresponders. The FLT-PET data support the notion that bevacizumab plus irinotecan has efficacy against recurrent glioblastomas.

Despite the advantages, bevacizumab plus irinotecan has potentially serious side effects. There is a thrombosis rate of 11%, and the rate for cerebral hemorrhage is uncertain. This is due to the small sample size in glioblastoma trials and other trials using bevacizumab for colon, breast, and lung malignancies specifically excluded patients with brain metastases. Another concern is rebound cerebral edema, which may occur 3 to 4 weeks after bevacizumab is stopped, and prophylactic dexamethasone may be necessary. Lastly, bevacizumab may cause reversible posterior leukoencephalopathy syndrome (RPLS), a peculiar disorder that occurs predominantly in the occipital brain.⁶ The exact mechanism for RPLS is unclear but it is related to low circulating VEGF levels.

Antiangiogenic treatment has demonstrated efficacy but it may not be enough to control glioblastomas. Because glioblastoma stem cells have angiogenesis-independent but highly invasive phenotype,⁷ it may be necessary to add anti-invasion drugs into the current multimodality treatment of glioblastomas. Here, the stroma may play just an important role for glioblastoma invasion as it is for angiogenesis.

Eric T. Wong, M.D.

Brain Tumor Center & Neuro-Oncology Unit

Beth Israel Deaconess Medical Center

330 Brookline Avenue

Boston, MA 02215

Tel: (617) 667-1665

Fax: (617) 667-1664

E-Mail: ewong@bidmc.harvard.edu