Antiangiogenic Agents in Combination with Chemotherapy in Patients with Advanced Non-Small Cell Lung Cancer

Most patients with non-small cell lung cancer (NSCLC) present with advanced disease requiring systemic chemotherapy. Treatment with the antiangiogenic agent bevacizumab in combination with standard platinum-based doublet chemotherapy has been shown to improve outcomes in patients with advanced NSCLC. Several multitargeted antiangiogenic tyrosine kinase inhibitors (e.g., sorafenib, sunitinib, cediranib, vandetanib, BIBF 1120, pazopanib, and axitinib) are also being evaluated in combination with standard chemotherapy. Here we review current clinical data with combination therapy involving antiangiogenic agents and cytotoxic chemotherapy in patients with advanced NSCLC.


INTRODUCTION
Lung cancer remains the leading cause of cancer-related deaths in the United States, with approximately 219,440 new cases and 159,390 deaths expected in 2009 (1). Each year more deaths result from lung cancer than from breast, colorectal, and prostate cancers combined (2). Approximately 85% of all lung cancer cases are categorized as non-small cell lung cancer (NSCLC), and most patients present with advanced disease at the time of diagnosis (1,3). The standard of care for patients with advanced disease is platinum-based doublet chemotherapy (4). Adding a third cytotoxic agent to the regimen increases toxicity and does not provide additional clinical benefits (4). The Eastern Cooperative Oncology Group (ECOG) conducted a large (N = 1,207) randomized study that compared four platinum-based doublet chemotherapy regimens in patients with NSCLC (5). None of the regimens was found to yield superior efficacy, though fewer episodes of toxicity were noted with the combination of carboplatin and paclitaxel (5). The median survival in this study was 8 months (5). Although there is a survival benefit with improved quality of life when chemotherapy is given to patients with advanced NSCLC, it appears that an efficacy plateau is reached when conventional chemotherapy is used alone. Angiogenesis is the growth of new microvessels from preexisting vasculature, a process that involves a fine balance of proangiogenic and antiangiogenic factors and coordination between multiple cell types such as macrophages, endothelial cells, and pericytes (6)(7)(8)(9). Angiogenesis is necessary for cancer cells to proliferate beyond microscopic size and to metastasize (10). The vasculature associated with pathologic angiogenesis is abnormal in structure and function; it is characterized by tortuous, dilated, saccular vessels that are poorly organized and hyperpermeable (6)(7)(8). These vascular abnormalities lead to an abnormal tumor microenvironment with interstitial hypertension, hypoxia, and acidosis; this, in turn, increases the production of vascular endothelial growth factor (VEGF) and decreases the effectiveness of cytotoxic chemotherapy (11,12).
The vascular endothelial growth factor plays a key role in regulation, both in normal and cancer cells, promoting endothelial cell migration and proliferation necessary for angiogenesis. VEGF is over expressed in a majority of malignant tumors, including NSCLC (12)(13)(14)(15), and elevated blood levels of VEGF are associated with tumor aggressiveness and a poor prognosis (13). Three VEGF receptors (VEGFR) have been identified: VEGFR1, VEGFR2, and VEGFR3. The biologic effects of VEGF are mediated by VEGFR1 and VEGFR2; VEGFR2 is believed to play the primary role in activating endothelial cells. VEGFR3 is associated primarily with lymphatic vessel growth (12,14). Other growth factors, such as platelet-derived growth factor (PDGF) and fibroblast growth factor (FGF), also play key roles in promoting angiogenesis (16,17).
Bevacizumab is a monoclonal antibody that targets circulating VEGF and inhibits VEGF binding to VEGFRs, thereby preventing its proangiogenic activity (18). In 2006, bevacizumab was approved by the US Food and Drug Administration for the first-line treatment of patients with advanced nonsquamous NSCLC in combination with carboplatin and paclitaxel (19). However, resistance often develops, and only approximately 50% of patients are actually eligible for bevacizumab treatment. Several mechanisms have been proposed that may account for the additive or synergistic activity of antiangiogenic agents and cytotoxic The retrospective analyses from E4599 revealed that OS was not significantly improved with bevacizumab in women (20). However, OS with or without bevacizumab was higher in women than in men, though this difference did not reach statistical significance (20). There was no difference in OS in patients of >70 years of age, but they did have a higher degree of reported toxicity (38 addition, many other clinical studies are currently recruiting patients and will evaluate first-line bevacizumab in combination with chemotherapy and/or pemetrexed (Table 2).

Second-line Treatment
Bevacizumab has also been studied as second-line therapy (

Safety
The toxicities associated with bevacizumab may be directly related to its mechanism of action. Hypertension, which occurs frequently, may be due to the decreased synthesis  S. V. Ulahannan and J. R. Brahmer  (47).
The ATLAS trial of maintenance bevacizumab and erlotinib (N = 598) (48), the PASSPORT trial of bevacizumab with first-or second-line chemotherapy (N = 106) (49), and the BeTa trial of bevacizumab with erlotinib in the second-line setting (N = 37) (50) all included patients with treated brain metastases, with some receiving therapeutic anticoagulation. Central nervous system (CNS) hemorrhages were reported in three patients participating in ATLAS, and five patients in ATLAS and three patients in PASSPORTexperienced pulmonary hemorrhages (48,49). These data indicate that patients with treated brain metastases and patients receiving therapeutic anticoagulation may be treated with bevacizumab.
The most common AEs in AVAiL were hematologic and related to GI, with a similar incidence in the three treatment arms (low-dose bevacizumab, high-dose bevacizumab, and placebo) (51). Adverse events that occurred at a higher frequency with bevacizumab included hypertension (7% and 9% vs. 2%), proteinuria (2% and 3% vs. 0%), and bleeding (4% and 5% vs. 2%). Hemoptysis was reported in 0.5% and 1.2% of patients in the low-and high-dose bevacizumab arms and in 1.3% of patients in the placebo arm (51). Serious AEs were reported in 39%, 45%, and 36% of patients, respectively (51). Despite the fact that 9% of the study population was receiving therapeutic anticoagulation, no pulmonary hemorrhage was reported in the initial publication of the trial or the final safety analysis (33,51 Results from E4599 have suggested a longer OS in patients with hypertension (15.9 months vs. 11.5 months without hypertension) and improved PFS with the onset of hypertension during bevacizumab treatment (7.0 months vs. 5.5 months), although these results did not reach statistical significance (52). Similar findings regarding the relationship between bevacizumab-associated hypertension and improved survival have been reported in the CALGB 90206 trial involving patients with metastatic renal cell carcinoma (53). However, further investigation of this association is warranted. A recent presentation of a large study of approximately 5,900 patients across six placebo-controlled, phase III studies of bevacizumab showed hypertension arising during treatment did not predict improvement in PFS or OS (54).
A retrospective evaluation of risk factors associated with severe pulmonary hemorrhage in patients treated with carboplatin/paclitaxel plus bevacizumab suggests that baseline tumor cavitation was the only risk factor for early-onset pulmonary hemorrhage. Central tumor location was not predictive of risk (55).

SMALL-MOLECULE, ANTIANGIOGENIC TYROSINE KINASE INHIBITORS (TKIs)
The proangiogenic activity of VEGF is dependent on signaling through its cognate receptors (i.e., the VEGFRs); thus, blocking these receptors is another antiangiogenic strategy (56). The receptors can be inhibited using small-molecule TKIs, which compete with adenosine triphosphate (ATP) for the active site of the tyrosine kinase (TK) domain and block receptor activation (56). Many TKIs that inhibit VEGFR also inhibit other key pathways involved in angiogenesis, including FGF and PDGF and their respective receptors (56). It has been suggested that because of some redundancy in proangiogenic signaling, both the FGF and PDGF pathways may play a role in the development of resistance to VEGF blockade (57)(58)(59)(60). Thus, by targeting multiple pathways, these agents may have the potential to overcome resistance to agents directed against only VEGF, such as bevacizumab (61,62).

Sorafenib
Sorafenib is an oral multi-kinase inhibitor that targets tumor growth, survival, and angiogenesis, by inhibiting VEGFR2, VEGFR3, and PDGF receptor (PDGFR) TKs (63). It also targets the Raf kinases, key signaling molecules downstream of Ras that transmit proliferative and cell survival signals (63).  S. V. Ulahannan and J. R. Brahmer  (Table 3) (66,67). In one phase I/II trial, carboplatin/paclitaxel in combination with sorafenib was evaluated in patients with advanced NSCLC (66). Among 39 evaluable patients, nine achieved a PR, and one achieved a CR; however, all these patients had melanoma. Median PFS for patients without melanoma was 104 days. The drug-related AEs were similar to those reported with single-agent sorafenib and included rash, hand-foot syndrome, and GI side effects (66,74). Based on these results, the randomized phase III ESCAPE trial was initiated. The ESCAPE trial enrolled 926 patients with advanced NSCLC who received carboplatin/paclitaxel with or without sorafenib as first-line therapy (67). There were no significant differences between the treatment arms in RR (24% vs. 27%), PFS (4.6 months vs. 5.4 months), or OS (10.7 months vs. 10.6 months), and as a result the trial was stopped early. There was a higher rate of drug-related infection in patients who received sorafenib than in those who received placebo (6.5% vs. 2.2%; p = .002). The grade 5 toxicity was observed more frequently in patients who received sorafenib versus those who received chemotherapy alone (14 patients vs. 4 patients; p < .001). In a subset analysis, shorter survival times were observed in patients with squamous cell histology who received sorafenib plus chemotherapy compared with those who received chemotherapy alone, though this observation was not statistically significant (67). Another ongoing large phase III trial, NExUS (NCT00449033), is evaluating gemcitabine/cisplatin with or without sorafenib as first-line therapy in patients with NSCLC (Table 3), and other studies are assessing second-line sorafenib monotherapy (Table 4). (75,76). Sunitinib has shown single-agent activity in a multicenter phase II trial as second-or third-line therapy in patients with advanced NSCLC, when administered according to a schedule of 4 weeks with treatment followed by 2 weeks without treatment, at a starting dose of 50 mg/day (76). The trial resulted in a RR of 11%, PFS of 12 weeks, and OS of 23.4 weeks. These findings are comparable with those of currently approved agents in this treatment setting. The most common grade 3/4 nonhematologic AEs included fatigue/asthenia (29%), pain/myalgia (17%), dyspnea (11%), and nausea/vomiting (10%). The grade 3/4 hematologic AEs included lymphopenia (25%), thrombocytopenia (5%), and neutropenia (5%) (76) In a phase I study, sunitinib was combined with cisplatin and gemcitabine as first-line therapy in patients with advanced NSCLC (78). The combination resulted in a manageable toxicity profile and PRs were observed in five of 24 patients. A second phase I study evaluated the combination of sunitinib and docetaxel in 50 patients with advanced solid tumors, including 18 patients with NSCLC (79). PR was observed in three patients, SD was observed in 12 patients, and AEs were manageable. Ongoing clinical trials are further evaluating the benefits of adding sunitinib to standard therapies and as maintenance after first-line therapy (Table 4).

Cediranib
Cediranib is a highly potent and selective inhibitor of the VEGF pathway with activity against all three VEGFRs, PDGFRs, and c-kit (80). In phase I studies, cediranib has demonstrated antitumor activity as a single agent with a manageable toxicity profile. The most frequently reported AEs include diarrhea, fatigue, dysphonia, and hypertension (81,82). Cediranib at doses of 20, 30, and 45 mg was evaluated in a phase I study in combination with four chemotherapy regimens (FOLFOX, irinotecan, docetaxel, and pemetrexed) in 46 heavily pretreated patients with advanced solid tumors (83). Of 35 patients who were evaluable for toxicity, grade 3/4 AEs observed across all four arms included fatigue, di-arrhea, hand-foot syndrome, neutropenic fever, and hypertension (83). In another phase I study (N = 15), cediranib 30 mg and 45 mg was combined with standard doses of cisplatin/gemcitabine in patients with advanced NSCLC (84). The combination was associated with increased toxicity compared with chemotherapy alone (84). All 12 of the evaluable patients showed some degree of initial tumor shrinkage and four achieved PRs (84). In another phase I trial (N = 20), once-daily cediranib 30 mg and 45 mg in combination with carboplatin/paclitaxel administered every 3 weeks was evaluated as first-line treatment in patients with advanced NSCLC (85). Patients receiving anticoagulation were eligible, but patients with a history of hemoptysis or bleeding were excluded. Adverse events included fatigue, myalgia, hypertension, GI toxicities, and neutropenia (85). Progression-free survival was reported in nine patients, and all but one patient showed some evidence of tumor shrinkage; median TTP was 7.6 months (85). Antitumor activity was observed at both dose levels, but with no indication of a dose effect (85).  S. V. Ulahannan and J. R. Brahmer BR .24 is a randomized, double-blind, placebo-controlled phase II/III trial of cediranib 30 mg in combination with carboplatin/paclitaxel as first-line treatment in patients with NSCLC (86). This trial was not continued into phase III as there appeared to be excessive toxicity, although evidence of clinical activity was observed (86,87). BR.29 will compare cediranib 20 mg in combination with carboplatin/paclitaxel with chemotherapy alone as first-line treatment in patients with NSCLC (Table 4).
An ongoing phase II trial is evaluating cediranib in combination with pemetrexed for second-or third-line treatment of advanced NSCLC (Table 3) (68). This study consists of two cohorts: patients who have not received prior bevacizumab (cohort A) and patients who have received prior bevacizumab (cohort B). In a preliminary analysis of the first 31 evaluable patients, the confirmed RR was 16% (10% in cohort A and 25% in cohort B) and the disease control rate was 71% (74% in cohort A and 67% in cohort B). The grade 3/4 nonhematologic AEs included fatigue (21%), diarrhea (9%), anorexia (6%), hypertension (3%), cardiac ischemia (3%), bronchopleural fistula (3%), and esophagitis (3%). The grade 3/4 neutropenia was reported in 21% of patients and febrile neutropenia occurred in 3% of patients (68).

Vandetanib
Vandetanib is a small-molecule inhibitor that blocks both the VEGFR and EGFR pathways, although it is more specific for the VEGFR pathway (88,89). It is also a potent inhibitor of RET receptor TK activity (72). A number of studies have evaluated vandetanib in combination with chemotherapy (Table 3) (69)(70)(71)(72)(73). Response rates were increased in the combination arms compared to chemotherapy-only arms. In addition, most of these studies demonstrated a prolongation in PFS, but no improvement in OS. In a double-blind randomized, phase 3 trial (ZODIAC) the combination of vandetanib and docetaxel was evaluated as a second-line treatment in patients with advanced NSCLC (N = 1,391). The addition of vandetanib improved PFS (4.0 months in the Vandetanib group vs. 3.2 months in the placebo group) but no significant improvement in OS was reported (72). ZEIST, a randomized phase III trial (N = 1,240) in patients with advanced, previously treated NSCLC, demonstrated that singleagent vandetanib and erlotinib had equivalent efficacy by PFS (HR, 0.98; p = .721) and OS (HR, 1.01; p = .830), but that vandetanib was associated with a higher incidence of toxicity (90). Vandetanib is also being evaluated in the phase III trial (ZEPHYR) in patients with advanced NSCLC who have progressed after treatment with chemotherapy and an EGFR TKI, but preliminary results indicate that the trial did not reach its primary end point of OS (91). Regulatory submissions for vandetanib in patients with NSCLC have also been withdrawn (91).

BIBF 1120
BIBF 1120 is a TKI targeting VEGFR1, VEGFR2, VEGFR3, FGFR1, FGFR2, FGFR3, and PDGFR, with the potential to inhibit proangiogenic signaling pathways in vascular endothelial cells, pericytes, and smooth-muscle cells (92). BIBF 1120 was evaluated as a single agent in a phase II trial in 73 patients with advanced NSCLC who had received one to two prior chemotherapeutic regimens (93). Patients were randomly assigned to receive twice-daily BIBF 1120, 150 mg (N = 37) or 250 mg (N = 36). Median PFS was 1.6 months in the overall population (N = 73) and 2.9 months in patients with an ECOG performance status (PS) 0 to 1 (N = 57), while median OS was 22 weeks in the overall population and 38 weeks in patients with PS 0 to 1. One PR was achieved, and the SD rate was 48% for all patients and 59% in patients with a PS of 0 to 1. The grade 3/4 toxicities included reversible alanine aminotransferase (ALT) elevations (9.6%), diarrhea (9.6%), nausea (8.2%), fatigue (5.5%), and vomiting (4.1%) (93). BIBF 1120 was not associated with a high frequency of hypertension, which is commonly reported with other VEGF inhibitors (46,93).
Results from phase I studies confirmed the feasibility of combining BIBF 1120 with chemotherapy (94,95). In a phase I trial (N = 25), BIBF 1120 in combination with carboplatin/paclitaxel was evaluated in chemo-naive patients with advanced NSCLC (94). The maximum tolerated dose (MTD) of BIBF 1120 was 200 mg twice daily, and no clinically relevant changes to carboplatin/paclitaxel pharmacokinetic parameters were observed (94). In another phase I doseescalation study, BIBF 1120 plus pemetrexed (500 mg/m 2 ) was administered to 26 patients with NSCLC who had received prior first-line, platinum-based chemotherapy (95). The MTD of BIBF 1120, in combination with standard-dose pemetrexed, was 200 mg twice daily, with no clinically relevant effects of BIBF 1120 pharmacokinetics observed in combination with pemetrexed (95). During the first treatment cycle, dose-limiting toxicities (DLTs) (all grade 3) occurred in seven patients for all doses and included transaminase elevations, fatigue, confusion, anorexia, and GI disorders (95). Among the 26 evaluable patients, one patient achieved a CR and 13 (50%) patients had SD. Based on these data, two randomized phase III studies are under way to evaluate BIBF 1120 in combination with docetaxel or pemetrexed for patients with advanced NSCLC after failure of first-line therapy (Table 4).

Pazopanib
Pazopanib is a multitargeted TKI that blocks VEGFR1, VEGFR2, VEGFR3, PDGFR, and c-kit (96). Preclinical studies indicate that pazopanib is effective in inhibiting angiogenesis (97). Pazopanib has been used as neoadjuvant monotherapy in patients with early-stage NSCLC (98). Of a total of 35 patients, three achieved a PR. The grade 3 toxicities were observed in five patients and included ALT elevations, hypertension, dyspnea, pneumonia, urinary tract infection, rash, increase in blood potassium, and lymphopenia. One patient experienced grade 4 bilateral pulmonary emboli 11 days after surgery (98). In a separate exploratory analysis of cytokines and angiogenic factors (C/AFs) in the serum of patients with early-stage NSCLC who received preoperative treatment with pazopanib, significant changes in eight C/AFs were reported (99). In particular, plasma levels of VEGFR2 (p < .0001) and placental growth factor (PIGF; p < .0001) were significantly Antiangiogenic Agents in Non-Small Cell Lung Cancer  decreased after pazopanib treatment. There was also a correlation between serum levels of VEGFR2 and tumor shrinkage (p < .05), suggesting its potential for use as a predictive marker of response (99). The efficacy and tolerability of pazopanib in advanced NSCLC, either alone or in combination with chemotherapy, is being evaluated in numerous clinical trials (Table 4).

Axitinib
Axitinib is a potent small-molecule TKI of VEGFR1, VEGFR2, VEGFR3, PDGR, and c-kit (100). In a phase I trial in 36 patients with advanced solid tumors, single-agent axitinib demonstrated antitumor activity in multiple tumor types, including NSCLC (101). The most common toxicities of any grade were hypertension (61%), fatigue (28%), nausea (19%), and diarrhea (17%) (101). Based on these data, a phase II study was conducted in 32 patients with advanced NSCLC (102). Nine patients had not received prior chemotherapy for metastatic disease and 23 patients received ≥1 prior regimen (102). Response rate was 9% and median PFS was 4.9 months in the overall study population and 9.2 months in treatmentnaive patients. Median OS was 14.8 months in the overall population and 14.8 months in patients who received axitinib as first-line therapy. One-year survival rates were 57% and 78%, respectively. Of the grade 1 or 2 AEs that occurred in at least 15% of patients, those that also occurred at grade 3 severity included fatigue, hypertension, diarrhea, and vomiting (102).
Axitinib was combined with standard paclitaxel/ carboplatin and gemcitabine/cisplatin chemotherapy in a phase I trial of 47 patients with NSCLC and other solid tumors (103). Response rate in the paclitaxel/carboplatin cohort was 29%, and in the gemcitabine/cisplatin cohort RR was 26% (103). The dose-limiting toxicities included fatigue, proteinuria, and rashes (103). A subcohort analysis of patients with squamous cell histology showed that axitinib plus paclitaxel/carboplatin was well tolerated, with no evidence of grade ≥3 hemoptysis (103). Trials with single-agent axitinib and combination therapy with axitinib in advanced NSCLC are ongoing (Tables 3 and 4).

CONCLUSIONS
The availability of treatment options for patients with advanced NSCLC has expanded. As seen in the E4599 trial of bevacizumab, therapy aimed at blocking angiogenesis can be effective when combined with standard doublet chemotherapy in patients with NSCLC. While a large number of patients were initially excluded from bevacizumab treatment because of safety concerns, some of these patients are now eligible, including patients with brain metastases and those receiving anticoagulation therapy. Patients