Pharmacology and Pharmacodynamics of Bevacizumab as Monotherapy or in Combination with Cytotoxic Therapy in Preclinical Studies

Angiogenesis, the development of the vascular system from endothelial cells, is essential during embryogenesis and occurs directly after the process vasculogenesis (1). Vasculogenesis is the development of the cardiovascular system whereby the endothelial cell precursors, the angioblasts, differentiate from hemangioblasts (2–4). Angiogenesis is required for further development of the vascular system in the embryo as well as for the growth and development of normal tissue, wound healing, and reproductive function in adults (5).

Angiogenesis also plays a pivotal role in several pathologic disorders, particularly tumorigenesis and metastases (6). Tumor angiogenesis is a complex process based on the concept that a tumor requires a vascular blood supply to grow beyond 1 to 2 mm (7). This blood supply is normally recruited from neighboring host, mature vasculature. Tumors that do not establish a neovascular supply may remain dormant for many years. The transition of a tumor from the “avascular” or “prevascular” phase to the “vascular phase” (increased growth and metastatic potential) is termed the “angiogenic switch” (8). The switch is believed to be stimulated by an increase in expression of proangiogenic factors, such as vascular endothelial growth factor (VEGF; vascular permeability factor), basic fibroblast growth factor, and transforming growth factor-β, and by a decrease in antiangiogenic factors, such as IFN-α or thrombospondin-1 (5, 9, 10). This shift in biochemical and molecular signals is thought to occur because tumors compress blood vessels, causing a decrease in oxygen (hypoxia) and vital nutrients. The angiogenic cascade leads to phenotypic changes in endothelial cells that line tumor blood vessels as well as other cell types, allowing the tumor to expand rapidly, invade surrounding tissues, and metastasize (8).

One of the most potent positive regulators of angiogenesis is VEGF (11). VEGF is a potent mitogen and survival factor for endothelial cells Fig. 1). It also has significant effects on vascular permeability and mediates the development of ascites. VEGF, also known as VEGF-A, is a member of the VEGF/platelet-derived growth factor gene family. Other family members include VEGF-B (12), VEGF-C (13), VEGF-D (14), and VEGF-E (15). Alternative splicing of the VEGF gene yields five isoforms, ranging from 121 to 206 amino acids (16). VEGF binds to two receptor tyrosine kinases: VEGF receptor (VEGFR)-1 (Flt-1) and VEGFR-2 (Flk-1/KDR). There is a third receptor, VEGFR-3 (Flt-4), which is mainly involved in the regulation of the lymphatic system (17). VEGFR-1 and VEGFR-2 are found predominantly on the surface of vascular endothelial cells and bind to VEGF with high affinity (16). Activation of VEGFR-2 leads to autophosphorylation and downstream signaling through pathways, such as phosphatidylinositol 3′-OH kinase/Akt, whereas autophosphorylation of VEGFR-1 is weak or undetectable in endothelial cells (18, 19). Studies show that activation of VEGFR-2 only induces angiogenesis and increased vascular permeability and mitogenesis (18, 19). It is thought that VEGFR-1 acts as a decoy receptor on endothelial cells, suppressing the availability of VEGF to VEGFR-2 (20). Other studies suggest that VEGFR-1 may also be important in hematopoiesis. Indeed, it has been suggested that VEGF and its receptors may control hematopoietic stem cell survival via an internal autocrine loop (21). VEGFR-1 is also thought to be involved in matrix metalloproteinase development and in the release of growth factors from endothelial cells (22, 23).

One mechanism contributing to tumor angiogenesis in some experimental models is the mobilization of endothelial progenitor cells from the bone marrow to peripheral circulation and subsequent homing of circulating endothelial progenitor cells to the tumor vasculature (24). VEGF is one of several factors that can induce endothelial progenitor cell recruitment to the peripheral circulation (25). Studies using VEGFR-1 blocking antibodies have suggested that VEGFR-1 has a key regulatory role of VEGFR-1 in the differentiation of endothelial progenitor cells from bone marrow progenitor cells to the tumor vasculature (26). [For further details of the role of VEGF in angiogenesis, see reviews by Ferrara (11, 27)].

Many studies show that VEGF is overexpressed in most human tumors (ref. 28; Table 1). Several studies have also reported a correlation between increased VEGF expression and tumor progression and/or a greater risk of recurrence in patients with specific cancers (29–31). Therefore, given its role in the growth and development of various cancers and its relatively low expression levels in healthy, adult tissues, VEGF is considered an attractive target for anticancer therapy. Importantly, anti-VEGF therapy offers significant advantages over conventional therapy, including direct access to the tumor vasculature, negating the need to penetrate the tumor itself. In addition, it was originally suggested that anti-VEGF therapy reduces the risk of drug resistance by targeting genetically stable endothelial cells rather than unstable tumor cells (32). However, more recent data suggest that resistance to antiangiogenic therapy remains an obstacle. Sweeney et al. (33) recently suggested that anatomic, biochemical, and histologic changes within the cancer and endothelial compartments may be mechanisms of resistance to antiangiogenic therapy. Studies are in progress to understand the possible mechanisms of resistance to antiangiogenic therapy.

Strategies being developed to target VEGF include monoclonal antibodies directed against VEGF, or its receptors VEGFR-1 and VEGFR-2, and small molecule tyrosine kinase inhibitors that act intracellularly to prevent autophosphorylation and activation of downstream growth-promoting signaling cascades.

This article reviews the pharmacology and pharmacodynamics of the humanized monoclonal antibody, bevacizumab, and/or its murine equivalent A4.6.1. Their activity against angiogenesis and tumorigenesis of primary tumors and metastases in human tumor cell lines xenografted in rodent models will be reported. Treatment modalities include single-agent and combination therapy with various cytotoxic compounds. The effect of bevacizumab and/or A4.6.1 on tumor vasculature and their pharmacologic safety profile will also be discussed.

Bevacizumab is derived from the murine VEGF monoclonal antibody A4.6.1 (34, 35). It is ∼93% human and 7% murine protein sequence, producing an agent with the same biochemical and pharmacologic properties as the parental antibody, but with reduced immunogenicity, and a longer biological half-life (34). Experimental studies show that bevacizumab and A4.6.1 neutralize all isoforms of human VEGF with a dissociation constant (K_d) of 1.1 and 0.8 nmol/L, respectively (refs. (34, 35) and data not shown). They also inhibit VEGF-induced proliferation of endothelial cells in vitro with ED₅₀ of 50 ± 5 and 48 ± 8 ng/mL, respectively. Both also block tumor angiogenesis and growth with almost identical potency and efficacy [percentage inhibition of human rhabdomyosarcoma cell line xenografted in nude mice 4 weeks after tumor cell implantation (5.0 mg/kg i.p. twice weekly); bevacizumab (95%) and A4.6.1 (93%); ref. (34)]. These studies show that bevacizumab and A4.6.1 are pharmacologically equivalent when tested in human cells, human tissue, or human VEGF isoforms (34).

Inhibition of Primary Tumor Growth. Numerous studies have examined the pharmacologic efficacy of bevacizumab and/or A4.6.1 in conjunction with tumor xenografts implanted into nude mice. These studies showed that anti-VEGF treatment results in 25% to 95% tumor growth inhibition compared with control mice in various tumor types (Table 2). Dose-dependent tumor inhibition was observed when bevacizumab or A4.6.1 was given against a range of tumor types irrespective of the route of administration and tumor location. In several studies, primary tumor growth inhibition was seen with early- and late-treatment onset (days 0-5 and >5 days after tumor cell implantation, respectively; refs. 36–38). Interestingly, one study suggested that, if anti-VEGF treatment was initiated concomitantly to tumor cell inoculation, growth inhibition exceeded 95%. However, if the tumor reached a certain size before initiation of treatment, inhibition was less complete. It was suggested that, in the rhabdomyosarcoma model, this phenomenon resulted from increased production of VEGF by the stroma of the host (because A4.6.1 does not neutralize murine VEGF). Therefore, the authors emphasized the need to block both tumor and host VEGF for maximum tumor growth inhibition (39).

In general, the studies shown in Table 2 focus on tumor growth inhibition as their primary pharmacodynamic end point. However, Rubenstein et al. (40) assessed survival in rats implanted with intracranial glioblastoma cells and showed that administration of A4.6.1 significantly prolonged survival as well as decreasing tumor vascularity and increasing tumor cell apoptosis (41). These data support the hypothesis that VEGF production is essential for glioblastoma angiogenesis, although experimental studies are ongoing to fully understand the potential role of anti-VEGF agents in the treatment of glioblastoma.

Kim et al. (36) showed that administration of bevacizumab or A4.6.1 to tumor cell lines has no effect on their proliferation rates. This confirms that the primary pharmacodynamic effect of bevacizumab or A4.6.1 is not the tumor cells per se but endothelial cell proliferation and, consequently, tumor blood supply (36). In conclusion, these monotherapy studies support the hypothesis that blocking VEGF and disrupting neovascularization leads to inhibition of tumor growth.

Inhibition of Metastases. As VEGF is essential for the development of a neovasculature at very early stages of tumorigenesis, it is thought to play a key role in the formation of tumor metastases (42).

Several experimental studies have examined the extent to which VEGF inhibitors prevent the growth of tumor cells at metastatic sites. Warren et al. (43) showed that administration of A4.6.1 in a colorectal cancer xenograft model caused a 90% reduction in primary tumor mass and a 10- to 18-fold reduction in the number of liver metastases compared with controls. In addition, they did not find the presence of blood vessels or expression of VEGFR-2 in the liver metastases of the A4.6.1-treated mice. A similar reduction in number and size of tumor metastases were observed for A4.6.1 treatment in conjunction with lung metastases 3 days after implantation of human prostate tumor cells in mice (37). When treatment was delayed until the primary tumors were well established, primary tumor growth and the progression of metastatic disease were still inhibited. Similarly, A4.6.1 prevented lung metastases from Wilms' tumors implanted into kidneys of nude mice (44). These studies clearly show that anti-VEGF therapy is a promising strategy for treating metastatic tumors, although it should be noted that they examined the activity of anti-VEGF therapy on micrometastatic disease. The effect of anti-VEGF therapy on well-established metastases has yet to be established, but it is likely that anti-VEGF therapy will have lower tumor-regressing effects than observed in the clinical setting (33).

Tumor Growth Inhibition with Combination Therapy. The studies described show that anti-VEGF therapy is an effective way to block primary tumor growth and metastases. However, if anti-VEGF therapy is combined with cytotoxic agents, such as chemotherapy or radiotherapy, because they act via different mechanisms of action, there may be additive antitumor activity in primary and metastatic tumor sites in the vascular phase. Combination with chemotherapy or radiotherapy also could eradicate smaller tumors in the prevascular phase that are dormant until angiogenic phenotypic changes occur.

Recently, Jain (45) proposed a further rationale for combining anti-VEGF therapy with cytotoxic compounds. One of the main challenges with cytotoxic therapies is gaining access to the interior of the tumor, because tumor vessels are structurally and functionally abnormal. An imbalance of proangiogenic and antiangiogenic factors produces chaotic, irregular, and leaky blood vessels, which are poorly structured and hyperpermeable and cause increased interstitial fluid pressure (IFP) in most tumors. These abnormalities, together with compression of the vessels by the tumor cells, result in impaired flow of fluid, macromolecules, and oxygen to the tumor. Tumor penetration by cytotoxic agents is consequently inhibited, and the hypoxic conditions within the tumor can limit the efficacy of chemotherapy and/or radiotherapy. Therefore, Jain proposed that if tumor vasculature could be “normalized” with anti-VEGF therapy it would facilitate better delivery of therapeutic agents to the tumor. This may result through apoptosis of tumor endothelial cells; a decrease in vessel diameter, density, and permeability; a subsequent decrease in IFP; and improved oxygenation (ref. 45; Fig. 2).

There is another important hypothesis concerning the feasibility of combining anti-VEGF inhibitors with chemotherapy and/or radiotherapy. Gorski et al. (46) suggested that radiotherapy increases VEGF expression, which may contribute to tumor resistance, and that the increase in VEGF could be the response of the tumor to radiation stress. Therefore, blocking the radiation-mediated increase in VEGF with anti-VEGF therapy could increase the destruction of tumor cells and produce additive antitumor effects.

In addition to their direct antitumor cell activity, studies have shown that chemotherapeutic agents, such as cyclophosphamide (47), paclitaxel (48), and doxorubicin (49), have antiangiogenic effects that are increased when lower doses are given more frequently. Klement et al. (50) suggested that low-dose chemotherapy has proapoptotic effects against endothelial cells; therefore, simultaneously targeting VEGF could amplify the antitumor effects of chemotherapy. Browder et al. (47) showed that metronomic chemotherapy was effective in treating tumor cells that had already developed resistance to the same chemotherapy. Therefore, combining anti-VEGF monoclonal antibodies with chemotherapeutic agents dosed “metronomically” or in an “antiangiogenic schedule” may result in additive antitumor activity.

Combination with Radiotherapy, Chemotherapy, and Other Targeted Agents. Lee et al. (51) examined the antitumor effects of A4.6.1 plus radiotherapy given in NCr/sed nu/nu mice with human glioblastoma multiforme or colon adenocarcinoma xenografts. They showed that combined treatment induced a tumor growth delay greater than additive in the glioblastoma cells and additive in adenocarcinoma tumors. In both cell lines, the enhanced tumor growth delay was independent of tumor oxygenation. The authors concluded that hypoxia plays a minor role in the antitumor activity of radiotherapy in combination with A4.6.1. Additional findings showed that A4.6.1 therapy alone significantly reduced IFP (74% in colon adenocarcinoma cells and 73% in glioblastoma cells) and hypoxia; however, this was associated with a major decrease in vascular density in the glioblastoma tumors only. The authors concluded that the lower IFP is probably a result of normalizing tumor vasculature and that the increase in tumor oxygenation is likely to be a result of the improved quality of the tumor vasculature. Interestingly, Tong et al. (52) recently showed that a drop in IFP leads to an induced hydrostatic pressure gradient across the tumor vasculature, which enhances penetration of large molecules in tumors.

A study by Wildiers et al. (53) supported the findings that A4.6.1 reduces microvessel density in nude mice with colon adenocarcinoma and also suggested that A4.6.1 improves intratumoral uptake of chemotherapy. These data support Jain's theory that although there is reduced microvessel density the vessels display normalized vascular functions when compared with untreated tumors, allowing improved delivery of blood-borne agents.

A4.6.1 or bevacizumab were examined in combination with four commonly used chemotherapeutic agents, doxorubicin, topotecan, paclitaxel, and docetaxel. Borgstrom et al. (54) examined A4.6.1 in combination with doxorubicin in three different human breast cancer cell lines implanted into nude mice. They showed that combination treatment resulted in a significant reduction in tumor mass compared with single-agent therapy (P < 0.001). Interestingly, the study showed that 5 mg/kg/wk doxorubucin reduced the growth rate of tumor cells but had no significant effect on angiogenesis.

A.4.6.1 was also examined in combination with low-dose topotecan in athymic mice with Wilms' tumor cells (55). Combination treatment significantly reduced tumor weight in treated mice compared with controls (1.81 g control, 0.42 g topotecan, 0.25 g A4.6.1, and 0.11 g combination therapy; P < 0.003) and significantly reduced metastases compared with controls (7 of 10 controls and 1 of 25 treated mice displayed lung metastases; P < 0.003). Tumor growth following cessation of treatment was greatest in topotecan only and slowest in combination-treated mice, suggesting that combination therapy may provide longer-lasting suppression of tumor cells. Similarly, Kim et al. (56) showed that topotecan in combination with A4.6.1 significantly suppressed neuroblastoma xenograft growth (0.08 g) compared with controls or A4.6.1-treated mice alone (1.18 g control and 0.53 g A4.6.1; P < 0.0007).

Paclitaxel in combination with bevacizumab resulted in a greater inhibition of tumor growth than either agent alone (98% growth inhibition; P = 0.024 bevacizumab versus P = 0.02 paclitaxel alone) when examined in the CWR22R androgen-independent xenograft model of prostate cancer (57). In two separate experiments, Hu et al. (58) showed that, when tested in an ovarian tumor model, combination therapy with bevacizumab and paclitaxel significantly reduced tumor growth compared with paclitaxel therapy alone (83.3% and 85.7% combination therapy and 58.5% and 59.5% paclitaxel therapy alone). Anti-VEGF therapy alone did not significantly decrease tumor burden, but histologic analysis revealed that VEGF neutralization enhanced paclitaxel-induced apoptosis. In this ovarian tumor model, ascites formation is frequently observed but did not occur in the combined treatment or anti-VEGF therapy alone group. Similarly, Mesiano et al. (38) found that VEGF plays an important role in malignant ascite formation but is not an essential regulator of peritoneal ovarian cancer. The mechanism by which anti-VEGF therapy inhibits ascites is not well understood; however, an increase in vascular permeability is believed to be pivotal for ascite formation (38).

Sweeney et al. (59) investigated the angiogenic properties of docetaxel in combination with bevacizumab. They showed that docetaxel inhibited endothelial cell proliferation in vitro in a dose-dependent fashion and angiogenesis in an in vivo Matrigel plug assay (59). They suggested that high levels of VEGF produced by the tumor and subsequent angiogenesis may protect endothelial cells from the antiangiogenic effects of docetaxel, whereas bevacizumab may reverse this effect. These data therefore provide a rationale for combining low doses of docetaxel with bevacizumab.

Finally, Hotz et al. (60) examined A4.6.1 in combination with a matrix metalloproteinase inhibitor in human pancreatic models. Combination treatment resulted in additive effects in only one of the two pancreatic tumor cell lines tested. Studies are examining A4.6.1 or bevacizumab with other targeted agents in different cell lines.

In summary, the preclinical combination experiments described identified superior antitumor effects compared with single-agent treatment. These data support the principle that combining antiangiogenic compounds with traditional therapies may result in increased efficacy. However, it should be noted that in all these studies anti-VEGF was given concomitantly with cytotoxic therapy, but our preclinical and clinical experience with biological agents suggests that schedule of administration will be a key factor in improving the efficacy of combination therapy (61). Studies are ongoing to assess the effects of scheduling in various xenograft models.

Correlation between Pharmacologic Effects of Tumor Growth Inhibition and Vascular Changes. Unlike normal vasculature, the microvessels of tumors are hyperpermeable to macromolecules, and the permeability and numbers of microvessels are increased by VEGF-induced angiogenesis. VEGF also increases permeability by enhancing the activity of the vascular-vacuolar organelle (62). Vascular-vacuolar organelle is found in the cytoplasm of endothelial cells and is thought to allow plasma proteins and other macromolecules to escape from circulation (63).

Several preclinical studies have examined the correlation between anti-VEGF therapy and vascular permeability. Yuan et al. (64) showed that A4.6.1 reduced tumor vascular permeability in human glioblastoma, colon adenocarcinoma, and melanoma xenografts implanted into mice and found that neutralization of VEGF causes a reduction in vessel diameter, eventually leading to vascular regression. Brasch et al. (65) also showed that anti-VEGF therapy reduces tumor microvascular permeability in athymic rats xenografted with human breast carcinoma. However, a recent study found that anti-VEGF therapy significantly reduced vessel density in mice implanted with human pancreatic carcinoma compared with controls (67.8 ± 10.6 versus 146.7 ± 10.0 cm/cm²), although vessel diameter and permeability were not significantly altered (66). These results support the hypothesis that new microvessels are formed through interactions between VEGF and endothelial cells; therefore, blocking VEGF should lead to a reduction in vessel diameter, thereby blocking the passage of blood and contributing to vascular regression.

In the past, tumor vasculature was assessed using histologic assays, which focused on the quantification of the microvascular density of stained tumor specimens. However, microvascular density assays are not practical because of the ethical and physical limitations of invasive procedures and the inherent anatomic and physiologic heterogeneity of tumors (a biopsy may not be representative of the entire tumor). Therefore, other, noninvasive techniques are being developed, including magnetic resonance imaging (MRI; ref. 65). MRI is an excellent technique for assessing the hypervascularity and hyperpermeability of tumor microvasculature because it is noninvasive, can be repeated frequently, and allows quantitative measurement.

Various experimental studies have used MRI to assess the pharmacodynamic effects of anti-VEGF therapy on the microvascular characteristics of different tumors. Brasch et al. (65) examined the effects of A4.6.1 on athymic rats with xenografted human breast carcinoma using MRI enhanced with a macromolecular contrast medium. They observed a significant decrease in microvascular permeability, specifically a 98% decrease in fractional leak rate (P < 0.005) and fractional reflux rate (P < 0.005) and a 97% decrease in surface area product (P < 0.001) compared with controls as early as 24 hours after administration of A4.6.1. Similarly, using dynamic contrast-enhanced MRI, Gossmann et al. (67) reported a significant decrease in tumor microvascular permeability in a glioblastoma multiforme mouse model 10 days after administration of A4.6.1. They also observed that rats with human ovarian cancer showed significantly lower MRI-assayed tumor microvascular permeabilities after 14 days of A4.6.1 versus control rats (68). These data suggest that MRI techniques could be used in clinical studies to assess the efficacy of anti-VEGF modalities. Indeed, a recent study showed that dynamic contrast-enhanced MRI may be a useful technique for assessing the response and dose of angiogenesis inhibitors for clinical development (69). Further studies are in progress.

Safety Profile of Bevacizumab in Animals. As A4.6.1 and bevacizumab do not bind endogeneous murine VEGF, no in vivo safety pharmacologic studies have been done in experimental murine models. However, the toxicity of anti-VEGF treatment has been examined in neonatal mice using two independent approaches: Cre-loxP-mediated gene targeting or administration of mFlt(1-3)-IgG, a soluble VEGFR chimeric protein (70). The findings showed that VEGF is essential for early postnatal life, but this requirement disappears after the fourth postnatal week. In a study focusing on lung development, systemic VEGF neutralization for 10 days in 4-, 8-, and 16-week-old mice with VEGF-Trap, an inhibitor of VEGF, reduced the number of capillaries in the tracheal mucosa by 39%, 28%, and 14%, respectively (71). The magnitude of the reduction decreased with age but was still significant at 16 weeks. More work is needed to investigate whether these changes in the lung are species specific and if they reflect differences between compounds targeting VEGF-A selectively and the multiligand blocking VEGF-Trap. Safety and toxicology studies using bevacizumab were also done in young adult cynomolgus monkeys (72), as cynomolgus VEGF is 99% homologous to human VEGF and bevacizumab is pharmacologically active in this species (73). Bevacizumab 2, 10, and 50 mg/kg/dose i.v. twice weekly for 4 or 13 weeks resulted in a dose-related increase in hypertrophied chondrocytes, subchondral bony plate formation, and inhibition of vascular invasion of the growth plate. In females receiving 10 and 50 mg/kg/dose, there was also decreased ovarian and uterine weights and an absence of corpora lutea. The physeal and ovarian changes were reversible when treatment ceased. The changes in both male and female animals were thought to reflect the pharmacologic consequences of VEGF inhibition in young and reproductively fertile animals. There was no evidence of pathologic changes, toxicity, or unexpected adverse effects even after prolonged treatment with high doses (up to 50 mg/kg/dose for 13 weeks) of bevacizumab.

The pharmacologic and pharmacodynamic activities of bevacizumab and/or A4.6.1 relevant to tumor growth were examined in numerous animal studies, which confirmed that anti-VEGF therapy has impressive antitumor activity in a broad range of primary and metastatic human tumors. Importantly, clinical findings have validated these studies. Willett et al. (74) recently showed that a single infusion of bevacizumab decreased tumor perfusion, vascular volume, microvascular density, IFP, and circulating endothelial and progenitor cells in patients with rectal carcinoma, suggesting that anti-VEGF therapy has a direct antivascular effect in human tumors.

The preclinical studies discussed also suggest that combining anti-VEGF agents, such as bevacizumab with chemotherapy and radiotherapy, may result in synergistic antitumor activity. One likely mechanism for the synergy between cytotoxic agents and antiangiogenic compounds is the normalization of the chaotic, tortuous blood vessels that are formed by the imbalance in proangiogenic and antiangiogenic factors released on tumor growth. This vascular normalization produced by anti-VEGF therapy is believed to decrease interstitial hypertension, allowing more efficacious delivery of large therapeutic molecules to solid tumors. Supporting data suggest that the most effective use of anti-VEGF agents is in combination with chemotherapy. These findings are being validated in the clinical setting. Notably, the efficacy of bevacizumab in combination with chemotherapy (irinotecan, 5-fluorouracil, and leucovorin) was shown recently in a phase III trial of patients with metastatic colorectal cancer (75). Patients receiving bevacizumab with chemotherapy lived ∼5 months longer compared with patients treated with chemotherapy alone (20.3 versus 15.6 months; P < 0.001). Similar increases were observed in progression-free survival (10.6 versus 6.24 months; P < 0.001), response rate (44.8% versus 34.8%; P = 0.0036), and duration of response (10.4 versus 7.1 months; P = 0.001). Overall, bevacizumab was generally well tolerated and there was no overlap with the toxicity profile of irinotecan, 5-fluorouracil, and leucovorin. Hypertension was significantly increased in the irinotecan, 5-fluorouracil, and leucovorin plus bevacizumab arm, but this was easily managed with oral medication. There was also a small risk of thromboembolic events in patients receiving irinotecan, 5-fluorouracil, and leucovorin plus bevacizumab; this was ∼2-fold higher with an estimated overall rate of 5% (75). Further research is needed to optimize its dose and schedule, manage the toxicities, and select patients most likely to benefit from therapy.

Various studies have also shown that MRI is an excellent technique for assessing the hypervascularity and permeability of tumor microvasculature. MRI could be effectively incorporated into clinical trials to assess the efficacy of anti-VEGF therapy. Finally, safety studies in healthy cynomolgus monkeys show that bevacizumab is not associated with any notable toxicities even after long-term administration at doses exceeding the proposed therapeutic dosages.

In summary, these experimental and preclinical findings show that VEGF inhibition is a rational strategy for the treatment of cancer. The first clinical evidence is the results from the phase III trial of bevacizumab in combination with irinotecan, 5-fluorouracil, and leucovorin in metastatic colorectal cancer patients. Further clinical investigations are ongoing to fully explore the potential of this and other anti-VEGF and angiogenic strategies in the treatment of solid and hematologic cancers.