Axitinib and/or bevacizumab with modified FOLFOX-6 as first-line therapy for metastatic colorectal cancer: A randomized phase 2 study


  • Medical writing support was provided by Giles Brooke, PhD, Joanna Bloom, PhD, and Joseph Ramcharan, PhD, of UBC Scientific Solutions and was funded by Pfizer Inc. We thank the study investigators, their staff, clinical trial personnel, and the patients who participated.



In this multicenter, open-label, randomized phase 2 trial, the authors evaluated the vascular endothelial growth factor receptor inhibitor axitinib, bevacizumab, or both in combination with chemotherapy as first-line treatment of metastatic colorectal cancer (mCRC).


Patients with previously untreated mCRC were randomized 1:1:1 to receive continuous axitinib 5 mg twice daily, bevacizumab 5 mg/kg every 2 weeks, or axitinib 5 mg twice daily plus bevacizumab 2 mg/kg every 2 weeks, each in combination with modified 5-fluorouracil/leucovorin/oxaliplatin (FOLFOX-6). The primary endpoint was the objective response rate (ORR).


In all, 126 patients were enrolled from August 2007 to September 2008. The ORR was numerically inferior in the axitinib arm (n = 42) versus the bevacizumab arm (n = 43; 28.6% vs 48.8%; 1-sided P = .97). Progression-free survival (PFS) (11.0 months vs 15.9 months; 1-sided P = .57) and overall survival (OS) (18.1 months vs 21.6 months; 1-sided P = .69) also were numerically inferior in the axitinib arm. Similarly, efficacy endpoints for the axitinib/bevacizumab arm (n = 41) were numerically inferior (ORR, 39%; PFS, 12.5 months; OS, 19.7 months). The patients who received axitinib had fewer treatment cycles compared with other arms. Common all-grade adverse events across all 3 treatment arms were fatigue, diarrhea, and nausea (all ≥49%). Hypertension and headache were more frequent in the patients who received axitinib. Patients in the bevacizumab arm had the longest treatment exposures and the highest rates of peripheral neuropathy.


Neither the addition of continuous axitinib nor the axitinib/bevacizumab combination to FOLFOX-6 improved ORR, PFS, or OS compared with bevacizumab as first-line treatment of mCRC. Cancer 2013;119:2555–2563. © 2013 American Cancer Society.


Colorectal cancer (CRC) is the third most common cause of cancer-related deaths in both men and women in the United States.[1] The prognosis is poor for patients who present with distant metastases (approximately 20% of those diagnosed), in whom the 5-year survival rate is approximately 12%.[1] Standard first-line chemotherapy for metastatic CRC (mCRC) consists of infusional 5-fluorouracil (5-FU)/leucovorin plus oxaliplatin (FOLFOX) or infusional 5-FU/leucovorin plus irinotecan (FOLFIRI).[2] First-line treatment with either regimen, followed by transfer to the alternate regimen upon progression, is associated with a median survival of 20.6 to 21.5 months.[3]

The addition of bevacizumab, a monoclonal antibody targeting vascular endothelial growth factor-A (VEGF-A) ligand, to chemotherapy reportedly prolongs the survival of patients with previously untreated mCRC compared with placebo plus chemotherapy.[4] In a pivotal phase 3 study, bevacizumab added to irinotecan, bolus fluorouracil, and leucovorin (IFL) significantly prolonged progression-free survival (PFS) (10.6 months vs 6.2 months; hazard ratio [HR], 0.54; P < .001) and overall survival (OS) (20.3 months vs 15.6 months; HR 0.66; P < .001) compared with IFL plus placebo.[4] Subsequently, it was shown that bevacizumab numerically improved survival when added to first-line FOLFIRI[5, 6] or FOLFOX.[7] Toxicities associated with bevacizumab include hypertension, proteinuria, hemorrhage, impaired surgical wound healing, infusion-related hypersensitivity reactions, thromboembolism, and gastrointestinal perforation.[8, 9]

Axitinib is a potent and selective second-generation inhibitor of VEGF receptor 1 (VEGFR1), VEGFR2, and VEGFR3 that blocks the VEGFRs at subnanomolar drug concentrations with minimal inhibition of other targets.[10] Axitinib is approved in the United States for the treatment of advanced renal cell carcinoma after failure of 1 prior systemic therapy,[11] having demonstrated superior efficacy compared with sorafenib in the second-line setting.[12] Simultaneous inhibition of the VEGF-A ligand and the VEGFRs may result in additive antitumor activity when axitinib is used in combination with bevacizumab in addition to chemotherapy.

In a phase 1 study, axitinib plus FOLFOX with or without bevacizumab was feasible and was devoid of pharmacokinetic interactions in patients with previously treated solid tumors, including mCRC.[13] Consequently, we conducted the current randomized phase 2 study to evaluate the safety and efficacy of axitinib in combination with the modified FOLFOX-6 regimen (referred to herein as FOLFOX) with or without bevacizumab in patients with previously untreated mCRC. The dosage of bevacizumab in combination with axitinib was chosen based on the results of the phase 1 study of axitinib in combination with chemotherapy for solid tumors, including mCRC.[13] The primary endpoint was the objective response rate (ORR), defined as the percentage of patients who achieved a complete response (CR) or a partial response (PR). Secondary endpoints included safety, duration of response, PFS, and OS.


Study Design

This was a multicenter, open-label, randomized phase 2 study conducted in the United States. A web-based, centralized randomization system was used to assign patients in a 1:1:1 ratio to receive axitinib plus FOLFOX (axitinib arm), bevacizumab plus FOLFOX (bevacizumab arm), or combined axitinib and bevacizumab plus FOLFOX (axitinib plus bevacizumab arm). Randomization, in block sizes of 3 patients, was stratified according to the receipt of prior adjuvant chemotherapy (yes vs no) and prior pelvic irradiation (yes vs no).

This study was conducted in accordance with the Declaration of Helsinki, the International Conference on Harmonization Guidelines on Good Clinical Practice, the study protocol, and applicable local regulatory requirements and laws. Study protocol, amendments, and informed consent forms were approved by an institutional review board/independent ethics committee. Written informed consent was provided by all participants before entering the study. The study is registered on as national clinical trial 00460603 (NCT00460603).


Eligible adult patients (aged ≥18 years) had previously untreated, histologically or cytologically documented, locally advanced CRC or mCRC. Patients who had received previous adjuvant chemotherapy were eligible if the last dose of adjuvant therapy was >12 months before enrollment. Patients also were required to have measurable disease according to the Response Evaluation Criteria in Solid Tumors (RECIST) (version 1.0)[14]; an Eastern Cooperative Oncology Group (ECOG) performance status 0 or 1; life expectancy ≥12 weeks; no evidence of pre-existing uncontrolled hypertension, ie, blood pressure (BP) >140/90 mm Hg (antihypertensive medications were permitted); and adequate liver function, renal function (serum creatinine ≤1.5 × upper limit of normal and ≤500 mg urinary protein/24 hours or dipstick <2+), and hematologic/bone marrow function (hemoglobin ≥9 g/dL, absolute neutrophil count ≥1500/mm3, platelet count ≥100,000/mm3).

Key exclusion criteria included second malignancy; metastases in the central nervous system; prior irradiation to ≥25% of the bone marrow; congestive heart failure or recent cardiovascular event; use of or anticipated need for known cytochrome P450 (CYP) 3A4 inhibitors (food or drugs) or drugs that induce CYP3A4 or CYP1A2; recent significant bleeding; clinically significant gastrointestinal abnormalities; known dihydropyridine dehydrogenase deficiency; or hypersensitivity to oxaliplatin, 5-FU, leucovorin, or recombinant human antibodies.

Drug Administration

Axitinib was administered orally with food at a starting dosage of 5 mg twice daily before administration of FOLFOX. After cycle 1, titration of the axitinib dose to 7 mg twice daily, and then up to 10 mg twice daily, was permitted if patients experienced no axitinib-related adverse events (AEs) of greater severity than Common Toxicity Criteria for Adverse Events (CTCAE)[15] grade 2 for consecutive 2-week periods, if their BP was ≤150/90 mm Hg, and if they were not receiving antihypertensive medication. Axitinib was continued at the same 5-mg twice daily dosage if a patient developed a grade 1 or 2 AE, but it was reduced to 3 mg twice daily and then to 2 mg twice daily, if necessary, in patients who developed grade 3 nonhematologic AEs or had systolic BP >150 mm Hg or diastolic BP >100 mm Hg while receiving maximal antihypertensive therapy. Axitinib was interrupted in patients who had grade 4 nonhematologic AEs, 2 readings of systolic BP >160 mm Hg or diastolic BP >105 mm Hg, or ≥2 g protein/24 hours; and axitinib was resumed at 1 dose level lower when AEs improved to grade ≤2, BP <150/100 mm Hg, or <2 g protein/24 hours.

Bevacizumab was administered intravenously at a dose of 5 mg/kg (bevacizumab arm) or 2 mg/kg (axitinib plus bevacizumab arm) once every 2 weeks immediately before chemotherapy or before axitinib for patients who were randomized to the combination arm. The initial dose of bevacizumab was administered as a 90-minute infusion; if no infusion-associated AEs occurred, then the second dose was administered as a 60-minute infusion, and each subsequent dose was administered as a 30-minute infusion. Bevacizumab was interrupted in patients with ≥2 g protein/24 hours or if they developed persistent or symptomatic hypertension. Bevacizumab was discontinued in patients who developed nephrotic syndrome, grade 4 hypertension, serious bleeding or thromboembolic events, or an infusion-related severe hypersensitivity reaction.

FOLFOX, administered in 2-week cycles, consisted of intravenous oxaliplatin 85 mg/m2 and intravenous leucovorin 400 mg/m2 infused concurrently over 120 minutes (through separate lines), followed by intravenous 5-FU 400 mg/m2 as a bolus injection, then 2400 mg/m2 as a continuous intravenous infusion over 46 to 48 hours. Study treatment was continued until patients developed disease progression or unacceptable toxicity. Dose modifications of both axitinib and bevacizumab in the axitinib plus bevacizumab arm were at the discretion of the investigators, according to their clinical judgment.


Tumors were assessed for response according to RECIST version 1.0[14] by computed tomography or magnetic resonance imaging at baseline and every 6 weeks; images were evaluated locally and were not centrally collected. To meet the criteria for stable disease, a minimum of 12 weeks was required for follow-up scans; patients who had stable disease for <12 weeks were classified as having an indeterminate response. Survival was monitored at least every 3 months after discontinuation of study treatment until ≥1 year after randomization of the last patient. Safety assessments included monitoring of AEs, physical examinations, and clinical laboratory tests. The severity of AEs was graded according to CTCAE version 3.0.[15] BP was measured at each clinic visit and twice daily by patients prior to axitinib dosing.

Statistical Analysis

Approximately 41 randomized patients per arm (N = 123) were required to determine whether the true response rate was ≤40% with an α value .10 versus a true response rate of ≥60% with a power of .90. Efficacy analyses were performed in the intent-to-treat population (all randomized patients). Exact 2-sided 95% confidence intervals (CIs) for the response rate of each treatment arm were calculated using a method based on the F distribution. PFS, OS, and duration of response were estimated using the Kaplan-Meier method.[16] Inferential statistics were calculated for the axitinib arm and the axitinib plus bevacizumab arm versus the bevacizumab arm, but not for the axitinib arm versus the axitinib plus bevacizumab arm. All patients who received at least 1 dose of study medication were included in the safety analysis.


Patient Characteristics

In total, 126 patients were randomized between August 2007 and September 2008 to receive, in addition to FOLFOX, axitinib (n = 42), bevacizumab (n = 43), or axitinib plus bevacizumab (n = 41) (Fig. 1). Of these, 123 patients received at least 1 dose of study treatment (axitinib, n = 39; bevacizumab, n = 43; axitinib plus bevacizumab, n = 41). Three patients in the axitinib arm were randomized but were not treated at the discretion of the investigator or patient. Patient demographics and baseline characteristics were generally similar among the study arms (Table 1).

Figure 1.

This is a Consolidated Standards of Reporting Trials (CONSORT) diagram illustrating patient disposition in the current study. FOLFOX indicates combined 5-fluorouracil, leucovorin, and oxaliplatin; ITT, intent-to-treat.

Table 1. Baseline Demographic and Disease Characteristics: Intent-to-Treat Population, N = 126
CharacteristicAxitinib and FOLFOX, n = 42Bevacizumab and FOLFOX, n = 43Axitinib, Bevacizumab, and FOLFOX, n = 41
  1. Abbreviations: ECOG, Eastern Cooperative Oncology Group; FOLFOX, combined 5-fluorouracil, leucovorin, and oxaliplatin; SD, standard deviation.

Sex: No. (%)   
Men25 (59.5)28 (65.1)26 (63.4)
Women17 (40.5)15 (34.9)15 (36.6)
Age: Median [range], y61 [43-79]64 [38-78]59 [30-75]
Race: No. (%)   
White35 (83.3)31 (72.1)40 (97.6)
Black4 (9.5)7 (16.3)1 (2.4)
Asian2 (4.8)2 (4.7)0
Other1 (2.4)3 (7)0
Weight: Mean ± SD, kg77.1 ± 18.276.9 ± 19.179.1 ± 16.9
ECOG performance status: No. (%)   
017 (40.5)20 (46.5)25 (61)
121 (50)20 (46.5)14 (34.1)
Not reported4 (9.5)3 (7)2 (4.9)
Prior radiation therapy: No. (%)2 (4.8)4 (9.3)2 (4.9)
Prior surgery: No. (%)39 (92.9)40 (93)40 (97.6)
Prior systemic therapy: No. (%)8 (19)7 (16.3)9 (22)

Treatment Exposure

Patients in the axitinib arm received fewer FOLFOX treatment cycles than those in the bevacizumab arm or the axitinib plus bevacizumab arm (Table 2). The median daily dose of axitinib and the median dose per cycle of bevacizumab were higher in the single-agent treatment arms than in the combined axitinib plus bevacizumab arm. In the axitinib arm, 4 patients (10.3%) had a dose titration to axitinib 7 mg twice daily during the study; in the axitinib plus bevacizumab arm, 4 patients (9.8%) had a dose titration to axitinib 7 mg twice daily, and 1 patient (2.4%) had a dose titration to axitinib 10 mg twice daily during the study.

Table 2. Patient Exposure to Study Drug: Safety Population, n = 123
VariableAxitinib and FOLFOX, n = 39Bevacizumab and FOLFOX, n = 43Axitinib, Bevacizumab, and FOLFOX, n = 41
  1. Abbreviation: FOLFOX, combined 5-fluorouracil, leucovorin, and oxaliplatin.

Exposure to study drug: Median [range], d   
 Axitinib92 [10-858] 178 [3-724]
 Bevacizumab 210 (14-1036]182 [14-728]
 Oxaliplatin98 [14-868]210 (14-1036]182 [14-728]
 Leucovorin98 [14-868]210 (14-1036]182 [14-728]
 5-fluorouracil98 [14-868]210 (14-1036]182 [14-728]
Daily axitinib dose: Median [range], mg9.7 [3.8-12.6] 9 [5.5-16.9]
Axitinib dosing interruption: No. (%)35 (89.7) 38 (92.7)
Axitinib dose reduction: No: (%)20 (51.3) 27 (65.9)
Bevacizumab dose per cycle: Median [range], mg/kg 5 [5-7.9]2 [2-5.2]


Objective responses occurred in 28.6% of axitinib-treated patients (1-sided P = .97 vs bevacizumab), 48.8% of bevacizumab-treated patients, and 39% of axitinib plus bevacizumab-treated patients (1-sided P = .82 vs bevacizumab) (Table 3). A CR was attained in 1 patient who was receiving axitinib and in 2 patients who were receiving bevacizumab. The median PFS was 11.0 months in patients receiving axitinib (1-sided P = .57 vs bevacizumab) compared with 15.9 months in patients receiving bevacizumab, and 12.5 months in those receiving axitinib plus bevacizumab (1-sided P = .87 vs bevacizumab) (Table 3, Fig. 2a). The median OS was 18.1 months in patients receiving axitinib (1-sided P = .69 vs bevacizumab) compared with 21.6 months in patients receiving bevacizumab and 19.7 months (1-sided P = .41 vs bevacizumab) in patients receiving axitinib plus bevacizumab (Table 3, Fig. 2b).

Figure 2.

(a) Progression-free survival and (b) overall survival are illustrated. FOLFOX indicates combined 5-fluorouracil, leucovorin, and oxaliplatin.

Table 3. Efficacy Measures: Intent-to-Treat Population, N = 126
MeasureAxitinib and FOLFOX, n = 42Bevacizumab and FOLFOX, n = 43Axitinib, Bevacizumab, and FOLFOX, n = 41
  1. Abbreviations: CI, confidence interval; FOLFOX, combined 5-fluorouracil, leucovorin, and oxaliplatin; HR, hazard ratio; NE, not estimable; ORR, objective response rate; OS, overall survival; PFS, progression-free survival.

  2. a

    P values were based on 1-sided Pearson chi-square tests; treatment comparison vs bevacizumab.

  3. b

    Patients who had stable disease for <12 weeks were classified as having an indeterminate response.

  4. c

    Patients without a follow-up scan were classified as missing.

  5. d

    Values were based on a Cox proportional hazards model stratified by receipt of prior adjuvant chemotherapy (yes vs no) and receipt of prior pelvic irradiation (yes vs no).

  6. e

    One-sided P values were calculated from log-rank tests stratified by receipt of prior adjuvant chemotherapy (yes vs no) and receipt of prior pelvic irradiation (yes vs no).

ORR: % [95% CI]28.6 [15.7-44.6]48.8 [33.3-64.5]39 [24.2-55.5]
Pa.97 .82
Response: No. (%)   
Complete response1 (2.4)2 (4.7)0
Partial response11 (26.2)19 (44.2)16 (39)
Stable disease8 (19)13 (30.2)11 (26.8)
Progressive disease3 (7.1)5 (11.6)3 (7.3)
Indeterminateb12 (28.6)3 (7)7 (17.1)
Missingc7 (16.7)1 (2.3)4 (9.8)
PFS: Median [95% CI], mo11 [7.4-24.6]15.9 [9.1-NE]12.5 [7.8-14]
HR for progression vs bevacizumab [95% CI]d1.08 [.47-2.45] 1.49 [.75-2.98]
Pe.57 .87
OS: Median [95% CI], mo18.1 [14.6-29.9]21.6 [16.2-29.5]19.7 [17.5-29]
HR for death vs bevacizumab [95% CI]d1.16 [.66-2.03] .94 [.54-1.65]
Pe.69 .41
One-year survival rate: % [95% CI]72.2 [55.8-83.4]79 [63.8-88.4]79.9 [64.1-89.3]

Safety and Tolerability

The most frequently reported treatment-emergent AEs in axitinib-treated patients were gastrointestinal disturbances (diarrhea, nausea, constipation, vomiting), fatigue, hypertension, and neutropenia (Table 4). Diarrhea, hypertension, insomnia, and dysphonia were less common in bevacizumab-treated patients (>10% difference between treatment arms); whereas neutropenia, peripheral neuropathy, abdominal pain, pyrexia, and depression were more common in that group. Hypothyroidism was reported in 17.9% of axitinib-treated patients compared with 4.7% of bevacizumab-treated patients and 17.1% of axitinib plus bevacizumab-treated patients; no grade ≥3 hypothyroidism was reported. In the axitinib arm, compared with the bevacizumab arm, patients had higher incidences of grade ≥3 hypertension, thrombocytopenia, nausea, and asthenia. Patients in the bevacizumab arm had a higher incidence of peripheral neuropathy than those treated with axitinib (51.2% vs 30.8%; all grades), possibly attributable to longer oxaliplatin treatment exposure. In addition, there were lower rates of grade ≥3 AEs with axitinib versus bevacizumab for diarrhea, vomiting, and dyspnea. Many AEs occurred at higher incidence in patients receiving axitinib plus bevacizumab than in those receiving either axitinib or bevacizumab as single agents. No cases of gastrointestinal perforation were reported in this study. More patients (56.1% [36.6% considered treatment-related]) experienced serious AEs in the axitinib plus bevacizumab arm than in either of the other 2 arms (axitinib arm, 38.5% [28.2% treatment-related]; bevacizumab arm, 39.5% [30.2% treatment-related]).

Table 4. Treatment-Emergent Adverse Events Occurring in ≥20% of Patients in Any Treatment Group: Safety Population, n = 123a
 No. of Patients (%)
EventAxitinib and FOLFOX, n = 39Bevacizumab and FOLFOX, n = 43Axitinib, Bevacizumab, and FOLFOX, n = 41
  1. Abbreviation: FOLFOX, combined 5-fluorouracil, leucovorin, and oxaliplatin.

  2. a

    Adverse events are reported as MedDRA version 15.0 preferred terms. Severity is graded according to the National Cancer Institute Common Toxicity Criteria for Adverse Events, version 3.0.

Serious adverse events15 (38.5)17 (39.5)23 (56.1)
Discontinuation because of adverse events      
Axitinib16 (41)  19 (46.3)
Bevacizumab  9 (20.9)14 (34.1)
Adverse eventsAll GradesGrade ≥3All GradesGrade ≥3All GradesGrade ≥3
Neutropenia15 (38.5)13 (33.3)21 (48.8)13 (30.2)19 (46.3)13 (31.7)
Thrombocytopenia11 (28.2)5 (12.8)12 (27.9)016 (39)2 (4.9)
Anemia8 (20.5)1 (2.6)11 (25.6)3 (7)12 (29.3)1 (2.4)
Diarrhea26 (66.7)3 (7.7)21 (48.8)6 (14)28 (68.3)6 (14.6)
Fatigue24 (61.5)7 (17.9)29 (67.4)6 (14)31 (75.6)12 (29.3)
Nausea23 (59)4 (10.3)23 (53.5)1 (2.3)24 (58.5)2 (4.9)
Hypertension16 (41)6 (15.4)8 (18.6)1 (2.3)27 (65.9)6 (14.6)
Constipation15 (38.5)1 (2.6)15 (34.9)018 (43.9)0
Vomiting13 (33.3)1 (2.6)13 (30.2)4 (9.3)17 (41.5)2 (4.9)
Peripheral neuropathy12 (30.8)4 (10.3)22 (51.2)7 (16.3)17 (41.5)2 (4.9)
Insomnia11 (28.2)07 (16.3)010 (24.4)1 (2.4)
Decreased appetite10 (25.6)1 (2.6)10 (23.3)1 (2.3)21 (51.2)5 (12.2)
Epistaxis10 (25.6)013 (30.2)012 (29.3)0
Abdominal pain9 (23.1)1 (2.6)15 (34.9)2 (4.7)15 (36.6)5 (12.2)
Asthenia9 (23.1)4 (10.3)6 (14)2 (4.7)5 (12.2)1 (2.4)
Dehydration9 (23.1)5 (12.8)10 (23.3)4 (9.3)15 (36.6)9 (22)
Peripheral sensory neuropathy9 (23.1)1 (2.6)7 (16.3)2 (4.7)8 (19.5)3 (7.3)
Dizziness8 (20.5)1 (2.6)6 (14)07 (17.1)1 (2.4)
Dysphonia8 (20.5)01 (2.3)08 (19.5)1 (2.4)
Headache8 (20.5)1 (2.6)6 (14)013 (31.7)3 (7.3)
Mucosal inflammation7 (17.9)1 (2.6)10 (23.3)2 (4.7)15 (36.6)4 (9.8)
Pyrexia3 (7.7)09 (20.9)08 (19.5)0
Decreased weight7 (17.9)1 (2.6)6 (14)017 (41.5)2 (4.9)
Proteinuria5 (12.8)1 (2.6)5 (11.6)012 (29.3)2 (4.9)
Upper abdominal pain2 (5.1)01 (2.3)012 (29.3)1 (2.4)
Dyspnea6 (15.4)1 (2.6)8 (18.6)4 (9.3)10 (24.4)3 (7.3)
Depression005 (11.6)09 (22)0
Stomatitis4 (10.3)1 (2.6)5 (11.6)09 (22)1 (2.4)

Approximately twice as many patients discontinued axitinib because of AEs (41%) than those who discontinued bevacizumab because of AEs (20.9%). There were no predominant toxicities leading to axitinib or bevacizumab discontinuation. In the axitinib arm, AEs leading to axitinib discontinuation were hypertension and pulmonary embolism (n = 2 each) and asthenia/pulmonary embolism, cerebral hemorrhage, cerebrovascular accident, deep vein thrombosis, diarrhea, disease progression/rectal abscess, epistaxis, fatigue, headache, peripheral neuropathy, sepsis, and unspecified reason (n = 1 each). In the bevacizumab arm, AEs leading to bevacizumab discontinuation were asthenia, device-related infection, pharyngolaryngeal dysesthesia, fatigue, jugular vein thrombosis, peripheral neuropathy, prostate cancer, proteinuria, and hemorrhagic shock (n = 1 each). Because bevacizumab dose reduction was not specified in the protocol, axitinib dose reductions because of treatment-related AEs were more frequent than bevacizumab dose reductions (axitinib arm, 28.2%; bevacizumab arm, 11.6%), whereas axitinib dose interruptions because of treatment-related AEs were less common than bevacizumab dose interruptions (axitinib arm, 46.2%; bevacizumab arm, 62.8%), as indicated in Table 5.

Table 5. Dose Reductions or Interruptions Because of Treatment-Related Adverse Events
 No. of Patients (%)
Study DrugAxitinib and FOLFOX, n = 39Bevacizumab and FOLFOX, n = 43Axitinib, Bevacizumab, and FOLFOX, n = 41
  1. Abbreviation: FOLFOX, combined 5-fluorouracil, leucovorin, and oxaliplatin.

Dose reduction11 (28.2) 8 (19.5)
Dose interruption18 (46.2) 23 (56.1)
Dose reduction 5 (11.6)2 (4.9)
Dose interruption 27 (62.8)26 (63.4)
Dose reduction13 (33.3)13 (30.2)15 (36.6)
Dose interruption18 (46.2)24 (55.8)25 (61)


In this randomized phase 2 study, axitinib plus modified FOLFOX-6 (ie, FOLFOX) did not improve efficacy outcomes compared with bevacizumab plus FOLFOX in patients with previously untreated mCRC (1-sided P = .97). ORR and median PFS and OS were numerically lower in the axitinib arm than in the bevacizumab arm. It is noteworthy that the median PFS with bevacizumab plus FOLFOX in the current study (15.9 months) was longer compared with that achieved using combinations of bevacizumab and oxaliplatin-containing chemotherapy reported in prior phase 3 studies in the first-line setting (9.4 months and 10.3 months, respectively).[7, 17] It is disappointing to report that the combination of axitinib plus bevacizumab also failed to improve ORR or survival compared with bevacizumab.

Efficacy endpoints and the safety profile of the bevacizumab arm in the current study are consistent with historic outcomes of bevacizumab in combination with oxaliplatin-based[7] or irinotecan-based chemotherapy for the first-line treatment of mCRC.[4, 18] The addition of axitinib, or of the combined antiangiogenic agents (axitinib plus bevacizumab), to FOLFOX appeared to be somewhat less tolerable than bevacizumab plus FOLFOX, with higher rates of hypertension, headache, and hypothyroidism. The increased rate of peripheral neuropathy in the bevacizumab arm may have been related to longer treatment duration and, thus, higher cumulative oxaliplatin exposure.

In this study, the median duration of treatment with axitinib plus FOLFOX was less than half that of bevacizumab plus FOLFOX, which may partially explain the differences in response rates and survival endpoints between the treatment arms. It is likely that the tolerability issues outlined above contributed to the shorter treatment duration with axitinib plus FOLFOX; however, the rates of some of the most troublesome side effects of fatigue, nausea, and vomiting were similar across all 3 treatment arms. Furthermore, only 7% of patients of who received axitinib plus FOLFOX had progressive disease as their best overall response. This suggests that our study also may have been limited by its open-label treatment design. Physician and/or patient bias may have contributed to the lower duration of exposure with axitinib and increased scrutiny of toxicities, and randomization to the arm without bevacizumab could have resulted in early discontinuation for patients with stable disease or limited toxicity.

Recently, it was demonstrated that the oral VEGFR tyrosine kinase inhibitor regorafenib improved OS and PFS compared with placebo in patients with advanced CRC who had received previous treatment with oxaliplatin and irinotecan-based chemotherapy.[19] Although those data support the utility of VEGFR inhibitors as monotherapy in the refractory setting, attempts at combining them with chemotherapy have been unsuccessful to date. In the HORIZON III study, there were no significant differences in PFS, OS, or ORR between cediranib and bevacizumab in patients who also received FOLFOX.[17] Similarly, a global phase 3 study of sunitinib plus FOLFIRI compared with placebo plus FOLFIRI for the first-line treatment of mCRC was stopped earlier than planned in June 2009, when the data monitoring committee established that sunitinib plus FOLFIRI did not significantly improve PFS, which was the primary study endpoint.[20] In the randomized phase 3 CONFIRM-1 study, vatalanib (PTK787/ZK222584) plus FOLFOX-4 did not improve PFS, OS, or ORR compared with placebo plus FOLFOX-4 in 1168 patients with previously untreated mCRC.[21]

Sunitinib and cediranib are multitargeted tyrosine kinase inhibitors with in vivo activity against other non-VEGFR kinases, such as platelet-derived growth factor receptor (sunitinib, cediranib), stem-cell factor receptor (sunitinib, cediranib), glial cell-line derived neurotrophic factor receptor (sunitinib), FMS-like tyrosine kinase-3 (sunitinib), and colony-stimulating factor-1 receptor (sunitinib).[22, 23] In the phase 3 studies described above,[17, 21] dose modifications were more frequent in the VEGFR arm; and, subsequently, exposures to the parent cytotoxic chemotherapy were decreased. In our study, in both the axitinib arm and the axitinib plus bevacizumab arm, over half of patients required dose reductions at some point during treatment. Although axitinib is a highly selective VEGFR inhibitor with less off-target kinase activity compared with other multitargeted tyrosine kinase inhibitors, our study adds to the body of evidence that adding this class of agents to FOLFOX creates tolerability issues and does not appear to improve tumor response or prolong survival compared with bevacizumab in the first-line treatment of mCRC.

It seemed advantageous that simultaneous inhibition of both the VEGF ligand and its receptor could provide more profound pathway inhibition, which hopefully would lead to improved clinical outcomes. However, the combination of axitinib and bevacizumab in our study did not appear to increase efficacy compared with bevacizumab alone in previously untreated patients receiving FOLFOX. There are several possible explanations for this, including the reduced dosage of bevacizumab in the axitinib plus bevacizumab arm (2 mg/kg vs 5 mg/kg bevacizumab alone); in the phase 1b study of this combination with FOLFOX, bevacizumab 5 mg/kg with axitinib 5 mg twice daily was deemed intolerable because of the dose-limiting toxicity of hypertension.[13] Even at the 2 mg/kg dose of bevacizumab plus axitinib 5 mg twice daily with FOLFOX in our study, the treatment exposures were shorter than with bevacizumab plus FOLFOX.

In summary, continuous administration of axitinib plus FOLFOX and the combination of axitinib plus bevacizumab plus FOLFOX appears to be somewhat less tolerable, with higher incidence of hypertension, headache, and hypothyroidism, than treatment with bevacizumab plus FOLFOX. It is plausible that tolerability in combination with chemotherapy may be improved when axitinib is administered on an intermittent schedule. The longest duration of treatment was observed in the bevacizumab arm, which may have contributed to the lower response rate in the axitinib arm. The lack of improvement in PFS and OS with axitinib compared with bevacizumab is consistent with other studies of VEGFR tyrosine kinase inhibitors in combination with oxaliplatin-based regimens in the first-line treatment of patients with mCRC. Unfortunately, simultaneous inhibition of the VEGFR and the VEGF ligand with the combination of axitinib and bevacizumab plus FOLFOX also did not improve ORR, PFS, or OS as compared with bevacizumab plus FOLFOX.


This study was sponsored by Pfizer Inc.


Dr. Cohn has received honoraria from speakers bureaus for Amgen, Novartis, and Genomic Health and has acted as a consultant to Pfizer and Genentech. Drs. Tarazi, Rosbrook, and Kim are employees of Pfizer Inc and own stock in Pfizer.