The Angiogenic Switch

Angiogenesis is the process by which new blood vessels sprout from existing blood vessels. Although angiogenesis is essential in normal vascular development, it can also be seen in many pathologic processes, including tumorigenesis. New blood vessels develop in response to a hypoxic environment and the increased demand for oxygen and nutrients. However, in pathologic conditions, this neovascularization is persistent and does not resolve by the reestablishment of adequate vascular perfusion. In this setting, there is a transition from a physiologic prevascular environment to a pathologic vascular setting. This model of tumor angiogenesis was first introduced by Dr Judah Folkman in 1971 and is commonly referred to as the angiogenic switch ( Fig. 1 ).

Angiogenesis in high-grade gliomas. Tumor growth begins with malignant cells inhabiting regions adjacent to normal blood vessels ( A ). As the number of tumor cells increase at a rapid rate, as occurs in high-grade gliomas, the cells outgrow their previous blood supply and become hypoxic and necrotic ( B ). The hypoxic environment stimulates new blood vessel development, which in turn promotes further tumor growth ( C ). The blood vessels are depicted in red, the tumor cells are yellow, normal cells are blue, mitotic cells are green, and necrosis is demonstrated with purple triangles.

( Data from Bergers G, Benjamin L. Tumorigenesis and the angiogenic switch. Nat Rev Cancer 2003;3(6):401–10.)

The angiogenic switch provides 2 distinct advantages to tumor cells. The first and perhaps more significant benefit is the direct blood supply afforded to tumors by this neovascularization. In general, tumors that have access to a sufficient blood supply will continue to have exponential growth. As described by Dr Folkman, there is a highly interdependent relationship between tumor cells and endothelial cells within the capillaries of a neoplasm, such that their rates of growth are contingent on each other. The second advantage of neovascularization is the ability for tumor vessels to indirectly support malignant cells in an environment called the vascular niche. This niche consists of small nests of dormant neoplastic cells. Neighboring endothelial cells supports these quiescent cells, yet their dormancy maintains a resistance to treatments, such as radiation and chemotherapeutic agents. Thus, neovascularization, which occurs in response to the angiogenic switch, provides tumors with a distinct growth advantage and proliferative autonomy when compared with normal cells.

The mechanism by which this angiogenic switch occurs is a complex series of events. As tumors enlarge and compress neighboring blood vessels, there is a decrease in blood supply leading to hypoxia. Tumors growing to a size as small as 1 to 2 mm can outgrow the local blood supply, triggering an angiogenic process. The decreased partial pressure of oxygen tension in the tissue induces an upregulation of a transcription factor, hypoxia-inducible factor (HIF). During times of normal oxygenation, HIF is kept at low levels by ubiquitin proteasome-dependent protein degradation, mediated by various tumor suppressor genes. In cases of tumorigenesis, however, there is a loss of tumor suppressor function and HIF levels remain elevated. This condition results in increased expression of proangiogenic factors, such as VEGF, basic fibroblast growth factor (FGF), and transforming growth factor-β, and a concomitant decrease in antiangiogenic factors, including interferon-α and thrombospondin-1. The result is the stimulus for the angiogenic switch, which allows the growth of new capillaries and consequently the continued growth of a malignant tumor.

Vascular Endothelial Growth Factor

VEGF has been shown to play a key role in the regulation of both normal and abnormal angiogenesis and it is commonly overexpressed in solid tumors. VEGF is a member of the VEGF/platelet-derived growth factor gene family. It binds to 2 tyrosine kinase receptors, VEGF receptor-1 (VEGFR-1 or Flt-1) and VEGF receptor-2 (VEGFR-2 or Flk-1/KDR). Both of these receptors predominantly occupy the surface of vascular endothelial cells. When VEGF binds to the VEGFR-2 receptor, it signals downstream pathways, such as the phosphatidylinositol 3′OH kinase/AKT, to induce angiogenesis, vascular permeability, and mitogensis. In contrast, the binding to the VEGFR-1 receptor is thought to have negative feedback on VEGFR-2 signaling, monocyte migration, and endothelial cell secretion of proteases and growth factors.

In one study, levels of VEGF and other proangiogenic factors were quantified by enzyme-linked immunosorbent assay in patients with primary and recurrent malignant gliomas. Twelve patients with primary GBM, 26 patients with recurrent GBM, and 7 patients with recurrent anaplastic astrocytoma underwent surgery for tumor resection and placement of an Ommaya reservoir into the resection cavity. Intracavitary fluid was drawn from the Ommaya reservoir between 2 to 12 weeks, before receiving chemotherapeutic treatment, and VEGF levels within the fluid were measured. The VEGF levels in the plasma of patients were also evaluated and compared with the plasma levels in 23 healthy controls. The VEGF levels in the plasma of patients were found to be higher than that of the controls ( P = .04). When comparing the plasma and intracavitary VEGF levels in patients, the intracavitary levels proved to be higher. Finally, the VEGF levels were highest in the intracavitary fluid of patients with recurrent GBM, followed by patients with primary GBM. Intracavitary levels were lowest in patients with recurrent anaplastic astrocytomas. The study also showed a trend toward longer survival with lower intracavitary VEGF levels in the patients with recurrent glioblastomas. As a result, more therapies are aimed at inhibiting the activity of VEGF and, thus, arresting angiogenesis, ( Fig. 2 ).

Angiogenic switch and antiangiogenic theory. Tumor growth begins with a dormant nest of tumor cells, which slowly proliferate. As the tumor grows, it secretes proangiogenic factors and induces neovascularization. Tumor growth will continue as long as there is a sufficient network of blood vessels to support this growth. This process is the angiogenic switch. Antiangiogenic agents aim to inhibit the proangiogenic factors.

( Data from Bergers G, Benjamin L. Tumorigenesis and the angiogenic switch. Nat Rev Cancer 2003;3(6):401–10.)

Avastin

One of the more recognized antiangiogenic drugs available is Avastin, a recombinant monoclonal IgG ₁ antibody derived from the murine VEGF monoclonal antibody. The protein sequence is composed of 7% murine VEGF-binding residues incorporated into a human IgG framework constituting the other 93% of the protein sequence. Experimental studies with Avastin show neutralization of all isoforms of human VEGF. Treatment with Avastin results in a dose-dependent inhibition of tumor growth and a reduction in tumor density, diameter, and permeability. For instance, patients who received greater than or equal to 0.3 mg/kg of intravenous Avastin demonstrate nondetectable serum levels of VEGF.

In addition to neutralizing VEGF, Avastin has also been found to return tumor vasculature to a more physiologic state. In nonmalignant blood vessels, there is a delicate balance between proangiogenic and antiangiogenic signaling mechanisms. This balance allows for the development of proper support structures, such as pericytes and basement membranes. In tumors, however, the proangiogenic state is favored, resulting in aberrant blood vessels, irregular basement membranes, and discontinuous endothelial lining. In effect, there is a leaky and unorganized vascular network with hyperpermeable membranes, which ultimately increases pressure in the interstitial space within tumors. As a result, there is impaired delivery of oxygen and cytotoxic agents from the blood to the tumor. Moreover, the hypoxic environment may impair the effects of any chemotherapy and radiation that is delivered to tumor cells. By restoring the function of abnormal tumor vessels, Avastin indirectly promotes the delivery of cytotoxic agents to tumor cells while reducing leakage into the interstitial space.

As alluded to earlier, anti-VEGF treatment may also enhance the effects of radiation in addition to chemotherapy. Gorski and colleagues determined a dose-dependent increase in VEGF levels in tumors treated with ionizing radiation both in vitro and in vivo using mouse models. In the in vitro model, VEGF levels were initially 3 times higher in tumors concomitantly irradiated than in those that were not, and levels remained elevated for 14 days ( P = .032). The in vivo mouse model further demonstrated this synergy, showing no inhibition of tumor growth with anti-VEGF therapy alone, 68.8% inhibition of tumor growth with radiation alone, and 83.4% inhibition of tumor growth with a combined therapy ( P = .046).

Results in Other Malignancies

Angiogenesis is a common pathologic process in several types of malignancies. Similarly, the use of Avastin has been applied to the treatment of other, nonglial neoplastic processes. In 2004, the drug was FDA approved as a first-line agent with 5-florouracil chemotherapy for the treatment of colorectal cancer. In 2006, Avastin was approved as a second-line treatment with 5-fluorouracil/leucovorin/oxaliplatin -4 for colorectal cancer and as a first-line treatment of patients with unresectable or metastatic nonsquamous, non–small cell lung cancer in combination with carboplatin and paclitaxel. The approval for colorectal and lung cancer was based on randomized clinical trials that showed a statistically significant improvement in overall survival. In 2008, the FDA approved Avastin, in conjunction with paclitaxel, as a first-line treatment of metastatic breast cancer because of a single clinical trial in an accelerated approval process. However, in 2010, the FDA began efforts to revoke their approval of the drug for the treatment of breast cancer after data from 5 randomized clinical trials failed to show improvement in overall survival or quality of life.

Glioblastoma

The development of agents targeting VEGF for the treatment of GBM has also been under investigation. Initial trials with Avastin showed promising response rates, and the drug was thought to be a breakthrough in the treatment of glioblastoma. In 2009, Avastin was granted accelerated FDA approval as a single-agent therapy for patients with recurrent glioblastoma. A phase II, noncomparative, multicenter trial (AVF3708 g) evaluated the efficacy of Avastin alone and in combination with irinotecan in patients with recurrent glioblastoma. All patients included in this study had histologically confirmed glioblastoma at either their first or second recurrence previously treated with radiotherapy and temozolomide. Patients were stratified by Karnofsky score and by first or second relapse. They received Avastin 10 mg/kg intravenously every other week and were observed for 6 months. Outcomes were compared with historical data with patients with 6-month PFS receiving either salvage therapy or irinotecan.

The 2 primary endpoints in this study were 6-month PFS and objective response rates. The 6-month PFS was defined as the percentage of patients who were alive and progression free at the end of the 24-week period. The objective response rate was defined as either a complete or partial response seen on MRI taken at least 4 weeks apart. This study showed a 6-month PFS of 42.6% in the Avastin alone group and 50.3% in the Avastin and irinotecan group. Additionally, there was no reported investigator-determined clinical progression. These results were significantly superior to the historical controls.

Based on the results of this trial, Genentech applied for accelerated approval of Avastin for monotherapy of recurrent glioblastomas. When the FDA reviewed this study, it reanalyzed the data using an exact 6-month cutoff instead of the 5.52-month cutoff used in the published study, which changed the PFS to 36.0%. The FDA did not include the results of the Avastin and irinotecan arm because that could have confounded the results. Additionally, the FDA excluded the objective response rate data from the trial because they determined that the characteristic histology of pseudopalisading necrosis of glioblastomas gives the tumor an irregular configuration that cannot be accurately measured on MRI. The FDA refused to grant accelerated approval based on these results alone until a confirmatory study with randomized controls rather than historical controls proved Avastin’s efficacy.

A single-arm single-site study was performed on patients with histologically confirmed recurrent glioblastoma previously treated with radiotherapy and temozolomide. In the study, patients received Avastin, 10 mg/kg every 14 days on a 28-day cycle. Patients who were noted to have progression of tumor growth during the study were asked to participate in a companion trial with the addition of irinotecan. MRI and positron emission tomography (PET) scans were performed at treatment onset and then again after 4 weeks. The primary end point in this study was PFS at 6 months, which was found to be 29%, with a median PFS of 16 weeks. Based on the Macdonald criteria, the overall response rate was 35%. In assessing the fludeoxyglucose F 18 uptake measured by PET scan 4 weeks after the start of treatment, this uptake was diminished in 49%, equal to the baseline in 37% and increased compared with the baseline in 14% of patients. In this study, Avastin was generally well tolerated.

Adverse Events

Inhibition of VEGF by Avastin occurs throughout the body and is not specific to tumor angiogenesis. Its effects on normal vascular function and angiogenesis have lead to reports of several serious adverse events. Perhaps the most commonly reported and often the most devastating complication is intracranial hemorrhage. Severe hemorrhage is estimated to occur 5 times more frequently in patients treated with Avastin than those receiving standard chemotherapy. Wound infection and wound healing are additional potential complication because the drug may preclude adequate blood for wound healing. A large cohort study analyzed the incidence of wound complications in patients undergoing abdominal surgery for colon cancer and found it to be directly related to the interval between surgery and the initiation of treatment.

In the initial AVF3708 g trial of Avastin therapy in patients with glioblastomas, 98.8% of the patients experienced adverse events, with 26.2% of patients in the Avastin arm experiencing serious complications. The most common findings were fatigue (45.2%), headache (36.9%), and hypertension (29.8%). A total of 46.4% of the patients in the Avastin-only arm of the study experienced grade 3 or higher adverse events. The most common of these were hypertension (8.3%), convulsion (6%), and fatigue (3.6%). A total of 18 patients discontinued Avastin during this study because of the adverse events, which included cerebral hemorrhage, fatigue, seizure, myocardial infarction, reversible posterior leukoencephalopathy, infection, gastrointestinal perforation, and others. In the National Cancer Institute 06-C-0064E trial, the most frequent adverse events were thromboembolism (12.5%), hypertension (12.5%), hypophosphatemia (6%), and thrombocytopenia (6%).

Results in Other Malignancies

Angiogenesis is a common pathologic process in several types of malignancies. Similarly, the use of Avastin has been applied to the treatment of other, nonglial neoplastic processes. In 2004, the drug was FDA approved as a first-line agent with 5-florouracil chemotherapy for the treatment of colorectal cancer. In 2006, Avastin was approved as a second-line treatment with 5-fluorouracil/leucovorin/oxaliplatin -4 for colorectal cancer and as a first-line treatment of patients with unresectable or metastatic nonsquamous, non–small cell lung cancer in combination with carboplatin and paclitaxel. The approval for colorectal and lung cancer was based on randomized clinical trials that showed a statistically significant improvement in overall survival. In 2008, the FDA approved Avastin, in conjunction with paclitaxel, as a first-line treatment of metastatic breast cancer because of a single clinical trial in an accelerated approval process. However, in 2010, the FDA began efforts to revoke their approval of the drug for the treatment of breast cancer after data from 5 randomized clinical trials failed to show improvement in overall survival or quality of life.

Glioblastoma

The development of agents targeting VEGF for the treatment of GBM has also been under investigation. Initial trials with Avastin showed promising response rates, and the drug was thought to be a breakthrough in the treatment of glioblastoma. In 2009, Avastin was granted accelerated FDA approval as a single-agent therapy for patients with recurrent glioblastoma. A phase II, noncomparative, multicenter trial (AVF3708 g) evaluated the efficacy of Avastin alone and in combination with irinotecan in patients with recurrent glioblastoma. All patients included in this study had histologically confirmed glioblastoma at either their first or second recurrence previously treated with radiotherapy and temozolomide. Patients were stratified by Karnofsky score and by first or second relapse. They received Avastin 10 mg/kg intravenously every other week and were observed for 6 months. Outcomes were compared with historical data with patients with 6-month PFS receiving either salvage therapy or irinotecan.

The 2 primary endpoints in this study were 6-month PFS and objective response rates. The 6-month PFS was defined as the percentage of patients who were alive and progression free at the end of the 24-week period. The objective response rate was defined as either a complete or partial response seen on MRI taken at least 4 weeks apart. This study showed a 6-month PFS of 42.6% in the Avastin alone group and 50.3% in the Avastin and irinotecan group. Additionally, there was no reported investigator-determined clinical progression. These results were significantly superior to the historical controls.

Based on the results of this trial, Genentech applied for accelerated approval of Avastin for monotherapy of recurrent glioblastomas. When the FDA reviewed this study, it reanalyzed the data using an exact 6-month cutoff instead of the 5.52-month cutoff used in the published study, which changed the PFS to 36.0%. The FDA did not include the results of the Avastin and irinotecan arm because that could have confounded the results. Additionally, the FDA excluded the objective response rate data from the trial because they determined that the characteristic histology of pseudopalisading necrosis of glioblastomas gives the tumor an irregular configuration that cannot be accurately measured on MRI. The FDA refused to grant accelerated approval based on these results alone until a confirmatory study with randomized controls rather than historical controls proved Avastin’s efficacy.

A single-arm single-site study was performed on patients with histologically confirmed recurrent glioblastoma previously treated with radiotherapy and temozolomide. In the study, patients received Avastin, 10 mg/kg every 14 days on a 28-day cycle. Patients who were noted to have progression of tumor growth during the study were asked to participate in a companion trial with the addition of irinotecan. MRI and positron emission tomography (PET) scans were performed at treatment onset and then again after 4 weeks. The primary end point in this study was PFS at 6 months, which was found to be 29%, with a median PFS of 16 weeks. Based on the Macdonald criteria, the overall response rate was 35%. In assessing the fludeoxyglucose F 18 uptake measured by PET scan 4 weeks after the start of treatment, this uptake was diminished in 49%, equal to the baseline in 37% and increased compared with the baseline in 14% of patients. In this study, Avastin was generally well tolerated.

Adverse Events

Inhibition of VEGF by Avastin occurs throughout the body and is not specific to tumor angiogenesis. Its effects on normal vascular function and angiogenesis have lead to reports of several serious adverse events. Perhaps the most commonly reported and often the most devastating complication is intracranial hemorrhage. Severe hemorrhage is estimated to occur 5 times more frequently in patients treated with Avastin than those receiving standard chemotherapy. Wound infection and wound healing are additional potential complication because the drug may preclude adequate blood for wound healing. A large cohort study analyzed the incidence of wound complications in patients undergoing abdominal surgery for colon cancer and found it to be directly related to the interval between surgery and the initiation of treatment.

In the initial AVF3708 g trial of Avastin therapy in patients with glioblastomas, 98.8% of the patients experienced adverse events, with 26.2% of patients in the Avastin arm experiencing serious complications. The most common findings were fatigue (45.2%), headache (36.9%), and hypertension (29.8%). A total of 46.4% of the patients in the Avastin-only arm of the study experienced grade 3 or higher adverse events. The most common of these were hypertension (8.3%), convulsion (6%), and fatigue (3.6%). A total of 18 patients discontinued Avastin during this study because of the adverse events, which included cerebral hemorrhage, fatigue, seizure, myocardial infarction, reversible posterior leukoencephalopathy, infection, gastrointestinal perforation, and others. In the National Cancer Institute 06-C-0064E trial, the most frequent adverse events were thromboembolism (12.5%), hypertension (12.5%), hypophosphatemia (6%), and thrombocytopenia (6%).