VEGF inhibition and metastasis: possible implications for antiangiogenic therapy

Since the 2004 approval of bevacizumab, a humanized monoclonal antibody that binds vascular endothelial growth factor (VEGF), antiangiogenic therapy has been used to treat selected cancers, including colon, non-small cell lung, and breast cancers. In addition to bevacizumab, the FDA has approved two other antiangiogenic therapies, sorafenib and sunitinib, both of which are tyrosine kinase inhibitors that target VEGF receptors, particularly VEGFR2. However, even with the approval of these three drugs, antiangiogenic therapy has not yet had the impact some have hoped for. Many clinical benefits are short-lived; while numerous trials have shown an increase in survival in patients treated with antiangiogenic therapy, the increase for many was a matter of months.¹ While any improvement in overall survival should be regarded as an accomplishment, it is important to understand why such clinical improvements are sometimes fleeting so that newer therapies can result in more enduring benefits.

One possible explanation for the varying nature of these improvements is a possible link between antiangiogenic therapy and increased metastasis. In 2000, Rubenstein et al., using rat models of orthotopically grown human glioblastoma, showed that animals treated with an anti-VEGF antibody had a 23-fold increase in satellite tumor area than rats that had not received antiangiogenic therapy.² Further, in a recent issue of Cancer Cell (Volume 15, 2009), both Pàez-Ribes et al. and Ebos et al. demonstrated that different forms of antiangiogenic therapy can lead to increased metastasis in multiple tumor types.^3,4

Ebos et al. first showed SCID mice that received short-term treatment with sunitinib (120 mg/kg/daily for 7 days) either prior to or after i.v. innoculation of MeWo, (human melanoma) or 231/LM2-4^LUC+ (human metastatic breast cancer cells expressing luciferase) had increased metastasis, as measured by immunohistochemical staining for both cell lines as well as bioluminescence for the 231/LM2-4^LUC+-innoculated mice.³ These increases in metastasis and tumor burden corresponded with shortened survival compared to control mice that received no antiangiogenic therapy. Similarly, mice with orthotopic 231/LM2-4^LUC+ tumors that received short-term sunitinib treatment prior to primary tumor removal also had increased metastasis compared to control mice. It is important to note that the authors only investigated metastasis and did not examine tumor phenotype. These data support the observation that under certain conditions, antiangiogenic therapy may lead to increased metastasis. However, mice with preestablished orthotopic MeWo, 231/LM2-4^LUC+, or B16 (mouse melanoma) tumors had significant tumor growth inhibition after sustained treatment with sunitinib (60mg/kg/daily), showing the benefits of antiangiogenic therapy.³

These observations are supported by the research of Pàez-Ribes et al., who showed that different antiangiogenic treatments may also lead to a more invasive phenotype in mice with pancreatic neuroendocrine (PNET) cancer or glioblastoma.⁴ After one week of treatment with DC101, a function-blocking anti-VEGFR2 antibody, RIP1-Tag2 mice had reduced tumor volume and vasculature but had a more invasive phenotype compared to the control mice, as determined by histological imaging and immunofluorescence. The mice treated with DC101 for one week had a 54% incidence of widely invasive tumors, while control mice had an incidence of 6%. After four weeks of DC101 treatment, the mice had a 62.5% incidence of widely invasive carcinomas. Even three weeks after the termination of treatment, mice that had been treated with DC101 had a 10-fold higher incidence of widely invasive tumors. This invasive tumor phenotype translated to increased distant metastases, with DC101 treated mice having a 4-fold higher incidence of lymph node metastasis. The invasive tumor phenotype and increase in metastasis were also seen in mice with PNET or orthotopic glioblastoma after other methods of disruption of the VEGF pathway, either by continuous sunitinib treatment or by tumor-specific deletion of VEGF-A in a ß-VEGF-KO background.⁴

These recent studies by Ebos et al. and Pàez-Ribes et al. complement each other well, together demonstrating that different antiangiogenic therapies targeting the VEGF pathway may lead to increased metastasis in some tumor types.^3,4 While a link between antiangiogenic treatment and tumor invasiveness and metastasis may help explain why antiangiogenic therapy has varying clinical benefits, it must be noted that other studies have shown inhibition of the VEGF pathway to reduce metastasis,^5,6 and large clinical trials involving many different antiangiogenic treatments have not resulted in increased observed metastasis.⁷ Though a relationship between inhibition of angiogenesis and increased metastasis may seem to complicate the field of cancer treatment, it may also provide opportunities for research to better understand tumor angiogenesis and to make the clinical improvements of antiangiogenic therapy more enduring. Pàez-Ribes et al. suggest that hypoxia may play a role in inducing the invasive tumor phenotype, though the mechanism leading to increased metastasis has not been fully elucidated.⁴ Many molecules have been linked to increased invasiveness, including HIF1-alpha and one of its targets, Met.⁸ Perhaps further research into the relationship between antiangiogenic therapy and metastasis will provide additional potential drug targets, resulting in adjuvant therapies that can enhance the clinical benefits of antiangiogenic treatment, continuing to develop the late Judah Folkman's vision of angiogenesis inhibition as a powerful weapon in the fight against cancer.

Abbreviations

FDA

Food and Drug Administration

VEGF

vascular endothelial growth factor

VEGFR2

VEGF-receptor 2

SCID

severe combined immunodeficiency

PNET

pancreatic neuroendocrine tumor