The Horizon of Antiangiogenic Therapy for Colorectal Cancer
The Horizon of Antiangiogenic Therapy for Colorectal Cancer
Vascular endothelial growth factor (VEGF) plays a crucial role in the growth and metastatic spread of cancer. Bevacizumab (Avastin) is the first commercially available VEGF inhibitor, earning US Food and Drug Administration (FDA) approval in February 2004. In combination with fluorouracil (5-FU)-based chemotherapy, this agent significantly prolongs overall and progression-free survival of patients with metastatic colorectal cancer. This review details the emerging role of the drug, its unique side effects, and other practical considerations related to bevacizumab therapy. Ongoing trials attempting to define additional indications for bevacizumab as well as the development of other promising angiogenesis inhibitors are also reviewed.
After more than 30 years of research, antiangiogenesis therapy has become a clinical reality, representing one of the most exciting therapeutic advances in oncology. Of the numerous growth factor- receptor complexes that promote angiogenesis, the vascular endothelial growth factor (VEGF) pathway is of particular importance. Bevacizumab (Avastin), a monoclonal antibody that targets VEGF, is the first such agent to gain US Food and Drug Administration (FDA) approval after a landmark placebo-controlled phase III trial confirmed the efficacy of antiangiogenic therapy in metastatic colon cancer. The availability of bevacizumab as first-line therapy for metastatic colorectal cancer in combination with irinotecan (Camptosar) and fluorouracil (5-FU) has fostered a global effort to further develop this and other anti-VEGF therapies. Trials are either planned or under way to define the role of bevacizumab in the adjuvant setting, to develop additional bevacizumab- based combinations with other chemotherapy agents or novel targeted agents, and to optimize dosing. This article surveys the ongoing development of anti-VEGF-based therapies for colorectal cancer as well as the evaluation and management of side effects unique to this drug class. Role of VEGF in Regulation of Angiogenesis Angiogenesis is a complex process leading to the formation and maintenance of new blood vessels. It involves processing of the extracellu lar matrix as well as cell proliferation and organization, and is regulated by a large number of activating and inhibitory signals. VEGF is a highly specific and potent endothelial cell mitogen, the expression of which is induced primarily by hypoxia.[2,3] VEGF binding sites are present exclusively on vascular endothelium, including quiescent cells, suggesting a role in growth promotion and in survival of established blood vessels.[ 4] An additional and possibly clinically important VEGF activity consists of increasing vascular permeability. The two main VEGF receptors are designated fms-like tyrosine kinase, VEGFR-1 (Flt-1), and fetal liver kinase-1, kinase domain region, VEGFR-2 (Flk-1/KDR). VEGF in Colorectal Cancer Inadequate angiogenesis results in tumor necrosis and impairs metastatic potential. Similar to other cancers, colorectal adenocarcinomas exhibit an abnormally high level of VEGF mRNA and protein expression with an increased level of both Flt-1 and Flk-1/KDR receptors in adjacent vessels, consistent with a paracrine mechanism.[ 7] Serum levels of VEGF are increased in colorectal cancer and correlate with stage of disease. VEGF overexpression has been validated as a poor prognostic factor.[8-10] Tumor VEGF overexpression predicts for worse outcome in patients with resected stage II disease and elevated preoperative serum VEGF is a poor prognostic factor in both stage II and stage III disease.[11,12] Observational studies indicate that VEGF has an important role in hematogenous metastatic spread of human colon adenocarcinoma and establish a foundation for therapeutic research targeting VEGF and its receptors. Different approaches toward inhibition of VEGF-dependent angiogenesis include the use of monoclonal antibodies against VEGF or its receptors (VEGFR), small-molecule inhibitors of VEGFR-specific tyrosine kinase activity, ribozymes specifically cleaving VEGF/VEGFR mRNA, soluble VEGF receptors acting as a trap for the circulating factor, and antisense oligonucleotides of VEGF mRNA. The clinical development of some of these agents is summarized in Table 1. Existent inhibitors of the VEGF pathway exhibit very limited toxicity and can be combined safely with conventional chemotherapy. Bevacizumab: First Anti-VEGF Agent in Clinical Practice Bevacizumab is a recombinant humanized monoclonal antibody that is able to neutralize all biologically active isoforms of VEGF-A. In murine xenograft models, the anti-VEGF antibody was shown to inhibit the growth of metastatic tumors while it was devoid of cytotoxic activity on cell lines in vitro.[13,14] In a phase I study, no drug-specific grade 3 or 4 toxicities were observed at bevacizumab doses ranging from 0.1 to 10 mg/kg. More common adverse effects were infusion-related asthenia, headache, and fever. An elevation of systolic and diastolic blood pressure of 10 mm Hg on average was noted at higher dose levels. Two patients experienced serious hemorrhages within metastatic tumors. The 21-day half-life of bevacizumab with linear kinetics permits every-14-day dosing. If the bevacizumab dose and schedule is altered to 7.5 mg/kg every 3 weeks, pharmacokinetics and overall dose exposure are similar to the currently standard dosing of 5 mg/kg once every 2 weeks. Major Clinical Trials
- Phase II Study-Bevacizumab may be administered safely with either irinotecan-, oxaliplatin-, or 5-FU-based chemotherapies with no additive toxicity. As a treatment for metastatic colon cancer, bevacizumab has been combined with 5-FU, leucovorin, 5-FU/irinotecan, and 5-FU/oxaliplatin (Eloxatin) in a variety of phase II protocols with encouraging response rates, time to tumor progression, and median overall survival durations. To date, bevacizumab at 5 mg/kg every other week has been associated with better clinical outcomes than the 10-mg/kg dose when combined with chemotherapy. A three-arm randomized phase II trial compared the Roswell Park regimen of 5-FU/leucovorin alone (bolus 5-FU at 500 mg/m2 with 500 mg/m2 of leucovorin weekly for 6 weeks every 8 weeks) with either 5 or 10 mg/kg of bevacizumab every other week. The median time to disease progression was 5.2 months in the control arm, 9 months in the 5-mg/kg bevacizumab arm (P = .005), and 7.2 months in the 10-mg/kg arm. The decrease in the hazard of progression in the 10-mg/kg group was not statistically significant. An objective response was seen in 40% of patients who received 5 mg/kg of bevacizumab and 17% in the control group (P = .029 compared with the control arm), whereas the difference in response rates for the higher-dose bevacizumab arm did not reach statistical significance (24%, P = .434). Patients in the 5-mg/kg arm had an impressive although not statistically significant 21.5-month median survival compared with 13.8 months in the control arm, despite the fact that 61% of controls crossed over to single-agent bevacizumab therapy upon disease progression. Toxicities were more common in the experimental arms, including grade 3 and 4 events. These side effects are consistent with what is emerging as a bevacizumab-specific toxicity profile, summarized in Table 2. While bevacizumab did not worsen the usual 5-FU/leucovorin-related adverse events (such as gastrointestinal toxicities or myelosuppression), hypertension, thrombotic, and hemorrhagic events were more frequent. The major hemorrhagic event was mild epistaxis lasting less than 5 minutes. Thromboembolic complications included a grade 5 pulmonary embolism in one patient receiving bevacizumab at the 10-mg/kg dose. Notably, the study was partly confounded by imbalances in randomization (more women and slightly worse performance profiles were assigned to the experimental arms). However, both efficacy and toxicity data favored the lower dose of bevacizumab.
- Phase III Study-The results of AVF2107, a randomized, placebocontrolled phase III trial of bevacizumab in combination with irinotecan and bolus 5-FU/leucovorin (for details of this and other discussed regimens, refer to Figure 1) resulted in FDA approval of bevacizumab as first- line therapy for metastatic colorectal cancer in February 2004. Nine hundred and twenty-three patients with previously untreated metastatic colorectal cancer were assigned to receive either IFL (irinotecan, 5-FU, leucovorin), IFL with bevacizumab, or 5-FU/leucovorin (according to the Roswell Park schedule) with bevacizumab. The dose of bevacizumab was 5 mg/kg every 2 weeks. Notable exclusion criteria included clinically significant cardiovascular disease (encompassing myocardial infarction, stroke, unstable angina, class II-IV congestive heart failure within 1 year, dysrhythmias requiring therapy, uncontrolled hypertension, grade 2 or higher peripheral vascular disease), full-intensity anticoagulation (except for cardiac doses of aspirin), ascites, or proteinuria exceeding 500 mg/d. Median survival, the primary end point, reached statistical significance with outcomes of 20.3 vs 15.6 months (P < .001), in favor of the IFL/bevacizumab combination. Similarly, the progression-free survival (10.6 vs 6.2 months, P < .001) and response rate (44.8% vs 34.8%, P < .004) favored the experimental arm. The survival benefit of adding bevacizumab to IFL was evident in all study subgroups, including patients with advanced age and poor performance status. According to the trial design, once a planned preliminary safety analysis determined the feasibility of the IFL regimen with bevacizumab, accrual to the 5-FU/leucovorin-plus-bevacizumab arm ended. Nevertheless, among 110 evaluable patients, there was a provocative trend toward better survival in the 5-FU/leucovorin/bevacizumab arm (18.3 months) than in the IFL/placebo arm, and the difference in time to disease progression for these two groups reached statistical significance (8.8 vs 6.7 months, P = .03). In this large trial, further insight into the bevacizumab toxicity profile emerged. Side effects associated with the IFL-plus-bevacizumab arm are summarized in Table 2. A greater number of grade 3/4 adverse effects occurred in the IFL/bevacizumab arm (85% vs 74%), chiefly due to hypertension (11% vs 2.3%). Elevated blood pressure has been managed easily with single-agent antihypertensive therapy. The occurrence of gastrointestinal (GI) perforations in patients receiving bevacizumab (including one fatality) is very concerning and seems to represent a drug-specific toxicity. Ongoing efforts will attempt to identify subsets at risk for this side effect. Clinicians must consider GI perforation when patients deteriorate or develop abdominal symptomatology. Interestingly in this trial, no statistically significant increased incidence of severe hemorrhage, proteinuria, or venous thromboembolism was observed. However, a retrospective analysis of thromboembolic events revealed a higher rate of "any" arterial thrombotic event (3.3% vs 1.0%), myocardial infarction (1.5% vs 0.8%), and cerebrovascular accident (0.5% vs 0.0%) in the IFL-plus-bevacizumab arm compared with the IFL-plusplacebo arm. Similarly, study AVF2192g (which included elderly or less fit patients with untreated metastatic colorectal cancer) corroborated this finding with an arterial thrombosis rate of 10% in the 5-FU/ leucovorin-plus-bevacizumab arm vs 4.8% in the 5-FU/leucovorin-plusplacebo arm. Additional interest focused on trial participants who developed thromboembolism and required anticoagulation.[ 21] Fifty-three patients from the IFL/bevacizumab arm were treated concomitantly with full-dose warfarin for a median of 218 days. The incidence of grade 3/4 hemorrhage was actually slightly lower than that seen in anticoagulated patients from the placebo group (3.8% vs 6.7%, respectively). Therefore, full-intensity anticoagulation is not a contraindication to bevacizumab use. Consistent with other bevacizumab trials, the traditional toxicities of the IFL regimen (GI and myelosuppression) were not significantly augmented with the addition of bevacizumab.
- ECOG Study-A phase II study conducted by the Eastern Cooperative Oncology Group (ECOG's E2200) assessed the combination of IFL with bevacizumab at 10 mg/kg every 2 weeks as first-line therapy, primarily evaluating progression-free survival and response rates. Patients with a history of hemorrhage or thrombosis, or on chronic anticoagulant medication were excluded. Among 87 evaluable subjects, the major toxicities included grade 3 diarrhea (17%), grade 3/4 neutropenia (35%), any hemorrhages (54%; grade 1 in > 90%), and grade 3/4 thrombosis (10%). Neither hypertension (2.3% grade 3) nor proteinuria posed clinically significant problems. The overall response rate was 49% with 6% complete remissions (as defined by Response Evaluation Criteria in Solid Tumors [RECIST]), while an additional 38% of patients had stable disease. The median progressionfree survival was 10 months. These efficacy data are concordant with the report of AVF2107 by Hurwitz et al.
Based on the AVF2107 pivotal trial, the FDA approved bevacizumab for use in patients with previously untreated colorectal cancer in conjunction with 5-FU-based chemotherapy. Neither cardiovascular disease nor chronic anticoagulation are listed as contraindications in the package insert. Based on data derived from trials of bevacizumab in non-small-cell lung cancer (NSCLC), indicating a 9% risk of serious, even fatal pulmonary hemorrhage, the drug is contraindicated in patients with a recent history of hemoptysis. The safety and efficacy of bevacizumab in patients with central nervous system metastases have not been evaluated. A warning concerning congestive heart failure was included in the approval, although this adverse effect was observed mostly in the context of prior or concomitant anthracycline therapy in metastatic breast cancer studies. Notably, biweekly blood pressure evaluation and urine analysis via dipstick, with 24-hour urine collection in case of 2+ proteinuria, are recommended. Per the January 2005 amended package insert, vigilance for signs and symptoms of arterial thromboembolic events including angina, myocardial infarction, transient ischemic attack, and cerebrovascular accidents is warranted due to an estimated 4.4% overall risk of such events associated with bevacizumab use. Which Chemotherapy Regimen to Use?
The FDA approval of bevacizumab is open-ended with regard to choice of chemotherapy, advising simply that bevacizumab be combined with infusional 5-FU regimens. The IFL regimen used in the pivotal AVF2107 has been largely replaced by programs utilizing infusional forms of 5-FU in combination with irinotecan (FOLFIRI) or oxaliplatin (FOLFOX) due to their better efficacy and safety profile. These regimens and their clinical outcomes are summarized in Figure 1. The excellent survival outcomes observed with 5-FU/bevacizumab in AVF2107 as well as in the aforementioned phase II experiences raise questions about the relative contribution of irinotecan to first-line 5-FU/bevacizumab and the more general need to utilize chemotherapy doublets in combination with bevacizumab. A placebo-controlled, randomized study-AVF2192g-evaluated single- agent 5-FU/leucovorin with or without bevacizumab in 209 patients deemed ineligible for combination therapy using irinotecan or oxaliplatin due to age or poor performance status.[ 25] This approach proved useful, again yielding a progression-free survival of 9.2 months in the treatment arm vs 5.5 months in the control arm, although the overall survival difference (16.6 vs 12.9 months) was not statistically significant. These survival outcomes are comparable to those reported in studies of FOLFOX or FOLFIRI regimens in untreated general populations of patients with metastatic colorectal cancer.[ 26,27] The side-effect profile of this trial again supports the observation that bevacizumab does not worsen typical 5-FU-associated side effects but is associated with an increased incidence of hypertension, doubled rate of arterial thromboembolic events (10%), and 2% incidence of GI perforation. Bevacizumab with 5-FU, therefore, enables less fit patients to enjoy survival benefits similar to persons receiving other highly active new chemotherapeutic combinations such as FOLFIRI or FOLFOX. Trials evaluating bevacizumab in combination with either FOLFOX or capecitabine (Xeloda)-oxaliplatin doublets (XELOX, CAPEOX) in patients with previously untreated metastatic colorectal cancer are either ongoing or planned. Until mature outcome data are available from these trials, the general concept of selecting a regimen based on toxicity profile, patient comorbidity, and patient preference should govern the choice of chemotherapy to combine with bevacizumab. Treatment Duration: Maintenance vs Intermittent Therapy
The intriguing concept of continuing bevacizumab therapy with sequential non-cross-resistant chemotherapy regimens is being explored. Patients in the experimental arms of the pivotal AVF2107 trial were allowed to continue bevacizumab after disease progression in combination with other chemotherapy regimens (25% of patients received oxaliplatin). Some patients have received bevacizumab for up to 3 years. No late toxic events have been observed in association with chronic bevacizumab therapy of 1 or more years. The subset of patients who continued to receive bevacizumab in conjunction with second-line oxaliplatin attained a median survival of 25 months. The results of ECOG trial E3200 were announced at the American Society of Clinical Oncology (ASCO) gastrointestinal symposium in January 2005. This phase III study randomized 829 patients with metastatic colorectal carcinoma progressing on first-line, irinotecan-based therapy into treatment with FOLFOX4 alone or with the addition of bevacizumab at 10 mg/kg. A third arm of single-agent bevacizumab (10 mg/kg) was discontinued after planned interim analysis due to inferior efficacy. Participants were monitored for proteinuria, and if it exceeded 500 mg/24 hours, the dose of bevacizumab was adjusted to 5 mg/kg. A survival benefit (12.5 vs 10.7 months, P = .0024) favored the experimental arm in this patient population. The toxicity analysis in the E3200 trial is consistent with results of the other bevacizumab trials in metastatic colorectal cancer (Table 2). There was no significant increase in hematologic toxicity but somewhat higher rates of nausea and vomiting (20% vs 9%) and neuropathy (15% vs 9%). The incidence of grade 3 hemorrhage was 2% in the FOLFOX4/bevacizumab arm, compared to 0% in the control arm; rates of thrombosis were identical. The incidence of bowel perforation was approximately 1% with one fatal event. Three other deaths possibly associated with bevacizumab included pneumonitis, a possible pulmonary embolism, and a brain hemorrhage complicating deep-vein thrombosis-related anticoagulation. Preliminary data regarding the ac- tivity of bevacizumab plus cetuximab (Erbitux) with or without irinotecan as second- or third-line therapy for patients with irinotecan-refractory disease- the so-called BOND-2 trial- were also presented at the January 2005 ASCO gastrointestinal malignancy symposium. A response rate of 38% with a median time to disease progression of 8.5 months was attained in the irinotecan-containing arm. An encouraging 23% response rate and 6.9-month median time to disease progression was attained in the bevacizumabcetuximab- alone arm. Since the median time to disease progression with first-line bevacizumab therapy is consistently approaching 10 to 11 months, a practical issue of whether to continue therapy until disease progression or hold therapy after disease control has been attained (usually with 4-6 months of treatment), and retreat upon disease progression needs to be addressed. Such intermittent therapy has been demonstrated to be effective with regard to survival outcomes in the treatment of metastatic breast and non-small-cell lung cancer (NSCLC), and this strategy is already applied by oncologists treating those diseases. Comparable survival outcomes with intermittent 5-FU-based therapy vs uninterrupted therapy until disease progression also have been demonstrated in patients with metastatic colorectal cancer. Whether intermittent therapy can be applied to bevacizumab-based regimens without sacrificing survival outcomes is an important yet unstudied issue.