Mechanisms of evasion to antiangiogenic therapy in glioblastoma.

Recognition of the role of vascular endothelial growth factor (VEGF) in developing the rich vascularity of glioblastomas, which contributes to their growth and resistance to radiation, surgery, and chemotherapy, has led to clinical trials of agents targeting VEGF or VEGF receptors 1 and 2. VEGF has been targeted with the mouse anti-human VEGF antibody avastin (bevacizumab), which has undergone phase II clinical trials in glioma patients. VEGF receptors have been targeted in phase II clinical trials in glioma patients using the tyrosine kinase inhibitor AZD2171 and the protein kinase C-b inhibitor enzastaurin. Encouraging results in phase II clinical trials studying bevacizumab treatment of recurrent glioblastomas led to the May 6, 2009, accelerated Food and Drug Administration (FDA) approval of bevacizumab for recurrent glioblastoma treatment, making bevacizumab just the third FDA-approved treatment for glioblastoma in nearly 4 decades and leading to a Radiation Therapy Oncology Group–sponsored phase III clinical trial that will soon begin in primary glioblastomas comparing radiation and temozolomide, a DNA-damaging agent that is the existing standard of care in glioblastoma chemotherapy, meaning that bevacizumab may soon become part of the standard treatment regimen for both primary and recurrent glioblastomas. Unfortunately, 43% of recurrent glioblastomas failed to respond to bevacizumab in the phase II clinical trials, and despite initial responsiveness, other glioblastomas frequently grow during bevacizumab treatment. Preclinical studies have suggested that, unlike the gene mutations that cause tumor cell resistance to DNA-damaging chemotherapy, tumor cells become resistant to antivascular agents by evasive mechanisms, ie, adaptive nongenetic mechanisms like transcriptional upregulation that allow tumor cells to find alternative ways of sustaining tumor growth while the antivascular target remains inhibited. Evasion to antiangiogenic therapy differs from classic chemotherapy resistance in that these evasive mechanisms reflect transcriptional changes that are generated more readily than the DNA gene mutations that characterize traditional chemotherapy resistance; these responses may occur to some extent in all treated tumors, with only tumors with the greatest transcriptional changes exhibiting tumor growth consistent with evasion to antiangiogenic therapy. At our institution, 15% of patients with recurrent glioblastomas treated with bevacizumab have experienced radiographic progression during treatment after initially responding to bevacizumab, leading to another surgical resection, consistent with acquired bevacizumab evasion. Half of these bevacizumabevasive glioblastomas (BEGs) have exhibited nonenhancing infiltrative regrowth and half have shown nodular-enhancing regrowth (Figure 1), suggesting the possibility of at least 2 different bevacizumab evasion mechanisms causing glioblastoma regrowth. Unfortunately, regardless of radiographic appearance, BEGs lack any therapeutic option and are invariably rapidly fatal. To identify mediators of these 2 subtypes of bevacizumab evasion, we performed a gene expression analysis and immunohistochemistry comparing BEGs with their paired primary tumors.