PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of jcoHomeThis ArticleSearchSubmitASCO JCO Homepage
 
J Clin Oncol. 2010 June 10; 28(17): 2817–2823.
Published online 2010 May 10. doi:  10.1200/JCO.2009.26.3988
PMCID: PMC2903316

Phase II Study of Cediranib, an Oral Pan–Vascular Endothelial Growth Factor Receptor Tyrosine Kinase Inhibitor, in Patients With Recurrent Glioblastoma

Abstract

Purpose

Glioblastoma is an incurable solid tumor characterized by increased expression of vascular endothelial growth factor (VEGF). We performed a phase II study of cediranib in patients with recurrent glioblastoma.

Methods

Cediranib, an oral pan-VEGF receptor tyrosine kinase inhibitor, was administered (45 mg/d) until progression or unacceptable toxicity to patients with recurrent glioblastoma. The primary end point was the proportion of patients alive and progression free at 6 months (APF6). We performed magnetic resonance imaging (MRI) and plasma and urinary biomarker evaluations at multiple time points.

Results

Thirty-one patients with recurrent glioblastoma were accrued. APF6 after cediranib was 25.8%. Radiographic partial responses were observed by MRI in 17 (56.7%) of 30 evaluable patients using three-dimensional measurements and in eight (27%) of 30 evaluable patients using two-dimensional measurements. For the 15 patients who entered the study taking corticosteroids, the dose was reduced (n = 10) or discontinued (n = 5). Toxicities were manageable. Grade 3/4 toxicities included hypertension (four of 31; 12.9%); diarrhea (two of 31; 6.4%); and fatigue (six of 31; 19.4%). Fifteen (48.4%) of 31 patients required at least one dose reduction and 15 patients required temporary drug interruptions due to toxicity. Drug interruptions were not associated with outcome. Changes in plasma placental growth factor, basic fibroblast growth factor, matrix metalloproteinase (MMP) -2, soluble VEGF receptor 1, stromal cell–derived factor-1α, and soluble Tek/Tie2 receptor and in urinary MMP-9/neutrophil gelatinase-associated lipocalin activity after cediranib were associated with radiographic response or survival.

Conclusion

Cediranib monotherapy for recurrent glioblastoma is associated with encouraging proportions of radiographic response, 6-month progression-free survival, and a steroid-sparing effect with manageable toxicity. We identified early changes in circulating molecules as potential biomarkers of response to cediranib. The efficacy of cediranib and the predictive value of these candidate biomarkers will be explored in prospective trials.

INTRODUCTION

Despite treatment with surgery, radiation, and chemotherapy almost all patients with glioblastoma experience recurrence and the median survival for most patients is fewer than 15 months. Therapy with conventional and experimental agents for recurrent glioblastoma is unsatisfactory and the proportion of these patients who are alive and progression free at 6 months (APF6) is 9% to 15%.

Increased vascular permeability leading to cerebral edema and microvascular proliferation are hallmarks of glioblastoma.14 This is due to high expression of proangiogenic cytokines, particularly of vascular endothelial growth factor (VEGF) and signaling via its endothelial tyrosine kinase receptor VEGF receptor 2 (VEGFR2).57 Levels of VEGF and its receptor correlate with the histologic grade of gliomas.8,9 We have previously shown that inhibiting the VEGF pathway normalizes the vasculature of gliomas in preclinical models and in patients and that this vascular normalization extends survival in preclinical murine orthotopic models of glioblastoma.1013 Thus, recurrent glioblastoma has emerged as an attractive setting in which to conduct clinical trials of novel anti-VEGF agents, such as monoclonal antibodies (bevacizumab; Avastin, Genentech, South San Francisco, CA) or tyrosine kinase inhibitors (TKI; eg, cediranib, Recentin, AZD2171, AstraZeneca Pharmaceuticals, Cheshire, United Kingdom).14

Cediranib is an orally available pan-VEGFR tyrosine kinase inhibitor with a half-life of 22 hours compatible with once daily dosing.15 Cediranib has a subnanomolar 50% inhibitory concentration for VEGF receptors with additional activity against platelet-derived growth factor ß and c-Kit. In a preliminary study in a subset of patients with recurrent glioblastoma, we observed that cediranib treatment normalizes tumor vasculature and alleviates edema.10 Herein, we report the final clinical efficacy, toxicity, and biomarker data on the entire cohort of patients treated on the first phase II study of cediranib in recurrent glioblastoma.

METHODS

Study Design

This phase II study of cediranib was approved by the local institutional review board (IRB) and was sponsored by the National Cancer Institute (NCI, NCT00305656). All patients signed an IRB-approved informed consent document before enrollment. The primary end point of this study was APF6, and secondary end points included radiographic response proportion, median overall survival (OS), and toxicity. Inclusion criteria for patients included pathologic diagnosis of glioblastoma; age ≥ 18 years; Karnofsky performance score ≥ 60; Mini-Mental Status Examination score ≥ 15; prior therapy with radiation; treatment with ≤ 2 chemotherapy regimens; recurrent glioblastoma by magnetic resonance imaging (MRI) or by tissue diagnosis; stable dose of corticosteroids for ≥ 5 days before the first baseline MRI scan; elapse of ≥ 3 months since completion of radiation; elapse of ≥ 3 weeks since completion of a non-nitrosourea chemotherapy; elapse of ≥ 6 weeks since completion of a nitrosourea-based chemotherapy; adequate bone marrow function (absolute neutrophil count ≥ 1,500/mcl; hemoglobin ≥ 8g/dL; platelet count ≥ 100,000/mcl); creatinine within institutional normal limit or creatinine clearance ≥ 60 mL/min/1.73 m2 for patients with creatinine more than institutional normal limits. Exclusion criteria included major surgery (including craniotomy) ≤ 4 weeks before the start of cediranib; concurrent use of anticoagulants; concurrent use of enzyme-inducing antiepileptic drugs; mean corrected QT interval more than 470 milliseconds or patients with a history of familial prolonged QT syndrome; ≥ 1 proteinuria on two consecutive urine dipstick assessments; pregnancy; history of uncontrolled hypertension or other serious medical illnesses including, but not limited to, unstable angina, arrhythmia, symptomatic congestive heart failure, active infection; infection with the human immunodeficiency virus; imaging (computed tomography or MRI) evidence of intratumoral or intracerebral hemorrhage deemed significant by the treating physician.

All patients were initially treated with cediranib 45 mg once each day. A cycle was defined as 28 days. A step-wise dose reduction scheme (Starting dose: 45 mg→ dose level −1: 30 mg→ dose level −2: 20 mg→ dose level −3: 10 mg) was utilized in patients who experienced dose-limiting toxicities. Patients were also allowed to temporarily interrupt cediranib for toxicity and resume the drug up to 14 days later. Algorithms for management of hypertension and diarrhea were followed when these toxicities were observed.

Treatment Response Evaluation

All patients were monitored by serial physical examinations, laboratory tests, and MRI scans. The MRI sequences included T1 pre-/postcontrast, T2, fluid-attenuated inversion recovery (FLAIR), diffusion weighted imaging, perfusion weighted imaging (dynamic susceptibility contrast) and dynamic contrast enhanced imaging. A schedule of the laboratory tests and MRI scans is enumerated in Appendix Table A1 (online only). Two baseline MRI scans were obtained 1 to 7 days before the first dose of cediranib followed by another MRI scan within 24 hours after the first dose of the medication then every month thereafter. The second baseline MRI scan (closer to the initiation of treatment) was used as the baseline for comparison of all subsequent studies. The postcontrast, T1-weighted MRI scans were assessed for response using a volumetric program by a central neuroradiologist who was blinded to patient identity and date of the scan. Scans were presented for review in a randomized sequence. The MRI scans were also assessed with two-dimensional measurements based on published criteria.16 Disease progression was defined according to Macdonald criteria. An independent radiologist from the Cancer Therapy Evaluation Program at the National Cancer Institute also confirmed radiographic responses in patients enrolled at the halfway point of the study. All toxicities were reported according to the National Cancer Institute Common Toxicity Criteria, version 3.

Circulating Biomarker Evaluations

Peripheral blood was obtained from all patients before therapy then 8 hours, 1 day, 9 days, 28 days, 56 days, 84 days, and 112 days thereafter to measure circulating proangiogenic and proinflammatory molecules and cells. Circulating progenitor cells were enumerated by flow cytometry using CD31, CD34, CD45, and CD133 as markers.17 Plasma analysis was carried out for circulating VEGF, placental growth factor (PlGF), sVEGFR1, basic fibroblast growth factor (bFGF), interleukin (IL) -1β, IL-6, IL-8, transforming growth factor α, matrix metalloproteinase (MMP) -2, and MMP-10 using multiplex enzyme-linked immunosorbent assay plates from Meso-Scale Discovery (Gaithersburg, MD) as well as soluble VEGFR2, stromal cell–derived factor-1α (SDF1α), angiopoietin 1 (Ang1), angiopoietin 2 (Ang2), and soluble Tek/Tie2 receptor (sTie2) from R&D System (Minneapolis, MN). Every sample was run in duplicate. Urine samples were obtained at similar time points as used for blood collection from the last 15 consecutive patients. Urinary MMP-2 (65kDa), MMP-9 (95kDa), and MMP-9/neutrophil gelatinase-associated lipocalin (NGAL) complex (125kDA) and activity were evaluated using gel zymography and were semi-quantitatively assessed by scoring from 1 (absent) to 9 (very strong).1820

Data and Statistical Analysis

Published historical outcomes in recurrent glioblastoma report an APF6 of 9% to 15%, median progression-free survival (PFS) of 54 to 63 days and median OS of 150 to 175 days.21,22 This phase II study was designed to detect an increase in APF6 from 10% to 25%.

Changes from baseline MRI parameters or circulating biomarkers were analyzed using the paired exact Wilcoxon test. P values were adjusted for multiple comparisons using Hommel's method.

Univariate analyses of PFS and OS with sex, age, Karnofsky performance status, baseline circulating, or urinary biomarkers and their early changes at 8 hours and 1 day were performed using a Wald test in the Cox proportional hazards model. Biomarker levels measured on quantitative scales were log-transformed and changes were calculated as ratios of on-study to baseline values. Analysis of the effect of drug interruptions on PFS and OS was performed using a Wald test in the time-dependent proportional hazards model, adjusting for the Vascular Normalization Index12 and using a sandwich estimator of variance23 to account for correlated data within patients.

Finally, we performed correlation analyses between all MRI-measured T1-contrast–enhanced tumor volumes and levels of plasma proteins and cell biomarkers at corresponding time points. This analysis of potential biomarkers of response and recurrence was based on a mixed-effects model, using the log-transformed biomarker level and a B-spline function of time in the fixed-effects model part and patient-specific linear function of time (including intercept) in the random-effect part.

RESULTS

Patient Characteristics

The study enrolled 31 patients with recurrent glioblastoma who had experienced prior treatment failure (Table 1). One patient who received only 18 doses of cediranib was included in the assessment of the OS, PFS, APF6, and toxicity, but excluded from other analyses. Eighteen of 31 patients had been treated by prior partial or total resection, 31 of 31 patients received prior radiation, and 29 of 31 patients had received prior temozolomide. Fifteen patients entered the study on dexamethasone with a median dose of 8 mg daily.

Table 1.
Patient Characteristics

Radiographic Tumor Response, Radiographic Disease Progression, and OS

All patients eventually experienced tumor progression and died except for one patient who remains alive having experienced disease progression after 26 months of cediranib therapy. The proportion of patients who achieved a partial radiographic response (> 50% reduction in contrast-enhancing volume) after treatment with cediranib was 56.7% using volumetric criteria and 27% using Macdonald criteria (Table 2). The APF6 was 25.8%, the median PFS was 117 days, and the median OS was 227 days (Table 2). Three patients were removed from the study by their treating physicians due to clinical progression without radiographic progressive disease. An independent review of MRI scans at the halfway point of the study confirmed all radiographic responses. Five patients demonstrated more than 25% increase in FLAIR dimensions 1 to 2 months before the observation of progressive disease on the postcontrast T1-weighted MRI sequences. None of these five patients was removed from the study due to clinical progression.

Table 2.
Efficacy of Cediranib in Recurrent Glioblastoma

Partial (over 50% volume reduction; n = 17) or minor responses (25% to 50% volume reduction; n = 6) correlated significantly with PFS (P < .05) but not with OS. Age was associated with a higher hazard of death (P = .027, Wald test), and Karnofsky performance status correlated with best radiographic responses after treatment (ρ = −0.51; P = .004).

Fifteen patients entered the study on a dose of corticosteroids. After cediranib treatment, the dose was reduced in 10 of 15 of these patients and corticosteroids were discontinued in five of 15 patients. Conversely, after discontinuation of cediranib 18 of 29 patients required either initiation of dexamethasone or a higher dose of dexamethasone. All cranial MRI sequences related to vasogenic cerebral edema (FLAIR, apparent diffusion coefficient, extracellular-extravascular volume fraction) demonstrated significant reductions after administration of cediranib and these changes persisted for at least 1 cycle (28 days, Appendix Table A1).

Safety and Tolerability

Two patients elected to stop the treatment due to fatigue. There were no other study terminations due to toxicity and there were no treatment-related deaths. There were no intratumoral or intracerebral hemorrhages observed during this study. The most common toxicities observed were hypertension, fatigue, and diarrhea. Grade 3/4 toxicities considered as possibly, probably, or definitely related to cediranib were observed in 21 (68%) of 31 of patients and are summarized in Table 3. Fifteen of 31 patients required at least one dose reduction while on the study treatment due to toxicity and 15 of 31 patients required a temporary drug interruption due to toxicity. The most common reasons for interruption were diarrhea (n = 3), hypertension (n = 2), proteinuria (n = 2), low thyroid stimulating hormone (n = 2), and hand-foot syndrome (n = 2). Twenty-seven of 31 patients treated with cediranib developed ≥ grade 1 hypertension after initiation of cediranib and 25 of 31 patients required medical treatment for hypertension. Drug interruptions had no significant association with mortality or disease progression (P > .8).

Table 3.
National Cancer Institute Common Toxicity Criteria Grade 3 or 4 Toxicities Possibly, Probably, or Definitely Related to Cediranib

The hazard of disease progression correlated inversely with diarrhea grade (P = .004, Wald test), but not with the hypertension grade (P = .18, Wald test). There were no significant correlations between these toxicities and OS.

Circulating Biomarker Analysis

In line with previous findings,10 biomarker kinetics after cediranib treatment in patients with recurrent glioblastoma were associated with immediate (by 8 hours) and persistent elevations in plasma of PlGF, SDF1α, and VEGF and more delayed decreases in soluble VEGFR2 (sVEGFR2; ie, by day 9; Table 4). In addition, we observed that cediranib treatment induced an immediate and persistent increase in MMP-10, a more delayed but persistent decrease in sTie2, and transient decreases in MMP-2 and Ang2 in plasma (P < .01; Table 4). The levels of VEGF, PlGF, and MMP-10 significantly decreased, and those of sVEGFR2 and sTie2 significantly increased after cediranib interruptions (ie, when measured within 2 weeks of the drug interruption). In this cohort, we detected no significant trends for the kinetics after treatment of circulating progenitor cells, or in plasma levels of bFGF, sVEGFR1, Ang1, IL-1β, IL-6, IL-8 or transforming growth factor α, and urinary MMP-2, MMP-9 or MMP-9/NGAL activity (Appendix Tables A2-A4 online only).

Table 4.
Plasma Cytokines (pg/mL) That Significantly Change After Cediranib Treatment in Patients With Recurrent Glioblastoma

The association between cediranib treatment outcome measures (OS, PFS) and biomarkers was explored for baseline levels as well as for early changes in these biomarkers. None of the biomarkers showed correlations with PFS or OS when evaluated at baseline. However, several dynamic biomarkers showed significant correlations with outcome. An increase in plasma MMP-2 at 8 hours after first administration of cediranib correlated with reduced PFS and OS (P < .05, Table 5). When measured at 1 day after treatment, an increase in urinary MMP-9/NGAL activity was associated with poor PFS (P < .01), and the extent of increase in PlGF and bFGF was significantly associated with longer OS (P < .05, Table 5). No other early biomarker changes correlated with OS or PFS (Appendix Table A5, online only).

Table 5.
Association Between Blood Angiogenic Biomarkers at Baseline, Their Changes at 8 Hours and at Day 1 After Cediranib Treatment With Radiographic Progression of Disease, and Mortality in Patients With Recurrent Glioblastoma

We also evaluated the correlation between biomarker changes at any time point during treatment and radiographic response in individual patients. A radiographic PR (ie, decreases of > 50% in enhancing tumor volume) was significantly associated with higher levels of plasma PlGF and IL-8 and lower levels of bFGF and sTie2 measured at the same time-point (P < .05). In addition, radiographic tumor progression (ie, increases of > 25% in enhancing tumor volume) was significantly correlated with increased levels of sVEGFR1, sTie2, and SDF1α (P < .05).

DISCUSSION

Bevacizumab—a humanized monoclonal antibody that specifically targets VEGF-A ligand—was approved by the US Food and Drug Administration as monotherapy for recurrent glioblastoma based on two phase II studies. In a noncomparative, randomized phase II trial of bevacizumab alone versus bevacizumab and irinotecan in patients with recurrent glioblastoma there were radiographic overall response rates of 28.2% and 37.8% and APF6 proportions of 42.6% and 50.3%, respectively.24 In another single-arm phase II study of bevacizumab alone followed by bevacizumab with irinotecan at progression in 48 patients with recurrent glioblastoma the APF6 was 29%.25

Based on the initial promising results with bevacizumab, several studies of oral agents that inhibit VEGF signaling have been conducted in the recurrent glioblastoma patient population with mixed results. A study of vatalanib (Novartis, Basel, Switzerland)—another oral pan-VEGFR TKI with additional activity against platelet-derived growth factor ß—with chemotherapy showed that fewer than 10% of the patients with recurrent glioblastoma achieved radiographic responses with a once-daily dosing schedule.26 Herein we report the first phase II trial of oral cediranib for recurrent glioblastoma. Potential advantages of cediranib relative to bevacizumab include oral bioavailability; a shorter half-life (22 hours v 21 days), which should allow more rapid clearance of drug in the event of serious toxicity such as hemorrhage; multiple tyrosine kinase targets and the ability to target intracellular VEGF receptors. We observed that cediranib treatment results in a radiographic response proportion, APF6 proportion, median PFS and median OS that compare favorably with data from historical controls.21 These data are also comparable to data obtained in phase II studies of bevacizumab in this patient population.27 The frequency of drug discontinuation due to toxicity was low and comparable to other anti-VEGF therapies. The safety profile of cediranib in patients with glioblastoma was acceptable, and there were no CNS hemorrhages or increased risk of thromboembolic complications.

Radiographic assessments of tumor response and progression to anti-VEGF therapies are challenging as these agents reduce permeability and, consequently, contrast leakage.28 Alternative radiographic methods are under investigation in order to more accurately define tumor response and progression in the setting of these agents. In this phase II trial progression of FLAIR signal abnormality was noted in five (16.6%) of 30 subjects before the observation of progressive disease on postcontrast T1-weighted sequences. This routinely acquired MRI sequence, as well as others including diffusion sequences, may therefore offer additional insight into disease progression in this patient population.29

The mechanism(s) of action of cediranib in patients with recurrent glioblastoma remains unclear. Cediranib treatment can transiently normalize the tumor vasculature and alleviate tumor-induced cerebral edema.10,30 Normalization of glioblastoma vessels may reduce tumor hypoxia and enhance sensitivity to concurrently administered cytotoxic therapies including ionizing radiation and chemotherapy. Thus, there is a strong rationale to test cediranib in combination with chemotherapy and radiation in patients with newly diagnosed glioblastoma. The antiedema effect and consequent reduction in corticosteroid use also has the potential to provide clinical benefit to patients with glioblastoma. In addition, as observed in preclinical models of glioblastoma treated with cediranib, edema alleviation may result in prolonged survival even without inhibition of tumor growth.11 Another potential antitumor mechanism could be targeting of the stem cell-like cancer cells in glioblastoma.

A major issue remains the heterogeneity in recurrent glioblastoma responses to cediranib, as observed for other anti-VEGF agents in various tumors.14 To date, there are no validated biomarkers of response to anti-VEGF therapy. Thus, identifying biomarkers that may predict benefit versus lack of benefit early during the treatment course is highly desirable.

We evaluated multiple plasma molecules and circulating cells that have been implicated in tumor angiogenesis.1 Biomarker kinetics were consistent with data on anti-VEGFR TKIs in our prior reports and others.10,31,32 In line with published literature, the baseline levels of any of these biomarkers did not appear to predict response.32,33 However, several of the biomarkers evaluated in our study (VEGF, PlGF, MMP-10, sVEGFR2, sTie2) changed significantly and reversibly after VEGF blockade. These are potential pharmacodynamic biomarkers, as similar changes have been reported for cediranib, vatalanib, and sunitinib in glioblastoma and other cancers.3439 Moreover, we observed significant correlations between several dynamic biomarkers (ie, the early change in plasma MMP-2, PlGF, sTie2, bFGF, and urinary MMP-9/NGAL activity) and radiographic responses and survival in recurrent glioblastoma after cediranib treatment. Increases in SDF1α, sVEGF1, and sTie2 were observed in patients at the time of glioblastoma progression after cediranib treatment. These observations are consistent with biomarker data from studies of sunitinib in hepatocellular carcinoma and bevacizumab in rectal cancer,39,40 and should be validated in preclinical and larger clinical studies.

Another important issue raised by some investigators is the potential of increasing the frequency of disease progression after interruption of anti-VEGF therapies, as seen in some mouse models.41,42 However, in this phase II study, there was no association of tumor progression with drug interruption.

In conclusion, cediranib monotherapy is active against recurrent glioblastomas and is associated with manageable toxicity. Further studies are warranted to confirm these results and to optimize the use of cediranib alone or in combination with cytotoxic therapies in patients with recurrent or newly diagnosed glioblastoma. Along these lines a randomized, three-arm, placebo-controlled, phase III trial in recurrent glioblastoma to test the efficacy of cediranib in this patient population has been initiated as well as studies of cediranib in combination with chemotherapy and radiation for patients with newly diagnosed glioblastoma.

Acknowledgment

We thank G. Gorospe, C. Koppel, S. Roberge, M. Zhu, M. Wang, W. Zhang, M. Foley, and O. Wu for outstanding technical assistance. We thank Dr Nicholas J. Patronas (NIH) for conducting the external radiographic review.

Appendix

Table A1.

MRI Parameters on Study Days –5, –1,1, 28, 56, and 112 Compared With Day –1 (pretreatment) Values in Individual Patients

ParameterDay –5
Day –1
Day 1
Day 28
Day 56
Day 112
MedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadj
CE-T1 volume, cm332.116.1-52.329< .001.03633.519.2-56.630NANA28.616.4-49.630< .001< .00113.18.1-35.130< .001< .00123.912.0-41.122< .001.02827.610.3-58.414.049.20
FLAIR volume, cm313680-16630.46.9012696-16230NANA14493-17030.90.907559-11230< .001< .0018962-11522< .001< .0019467-12514.078.24
ADC, mm2/sec1,2801,219-1,36830.14.141,2871,218-1,40630NANA1,2891,206-1,36130< .001< .0011,0971,023-1,19929< .001< .0011,060968-112020< .001< .0011,026984-1,17714< .001< .001
Vessel size6.465.99-7.3427< .001.0797.296.15-7.9530NANA6.645.80-7.5630.079< .0016.886.13-7.6729.441.07.296.16-7.9422.591.07.726.72-8.45141.01.0
CBV capillaries
    Large1.00.8-1.328< .001.0331.10.9-1.430NANA0.90.7-1.130< .001< .0010.80.7-1.029< .001< .0010.80.7-1.222.021.150.90.7-1.414.065.33
    Small0.70.5-0.828.44.940.70.5-0.830NANA0.60.5-0.730< .001< .0010.60.5-0.729.070.490.60.5-0.722.49.950.60.5-0.914.95.95
Permeability (min−1)0.0960.051-0.12328.15.330.0950.062-0.15530NANA0.0330.017-0.04429< .001< .0010.0060.002-0.01830< .001< .0010.0110.004-0.03122< .001< .0010.0190.005-0.03214< .001< .001
CBF capillaries
    Large0.9250.630-1.18028.69.810.9750.625-1.19030NANA0.7900.578-0.92530< .001< .0010.7600.540-1.03029< .001.0550.8800.535-1.01522.16.810.9350.560-1.30214.73.81
    Small0. 5700.445-0.70728.501.00.6150.415-0.79830NANA0.5550.410-0.62030< .001< .0010.5900.380-0.68029.501.00.5750.380-0.73522.46.920.6800.408-0.87814.501.0
MTT capillaries
    Large1.1551.110-1.29828.019.0631.2001.103-1.38030NANA1.1551.050-1.27530< .001.0421.0701.030-1.23029< .001< .0011.0550.952-1.22522.027.151.0400.975-1.14514.025.15
    Small1.1251.040-1.20028.71.711.1201.040-1.23030NANA1.0901.025-1.17030.20.401.0500.990-1.15029< .001.0821.0550.952-1.22522.027.0151.0400.975-1.14514.025.15

NOTE. Bold font indicates significant increases; italic font indicates significant decreases.

Abbreviations: MRI, magnetic resonance imaging; CE, contrast enhanced; FLAIR, fluid attenuated inversion recovery; ADC, apparent diffusion coefficient; CBV, cerebral blood volume; CBF, cerebral blood flow; MMT, median transit time; NA, not applicable.

Table A2.

Kinetics of Plasma Cytokines (pg/mL) That Do Not Change Significantly After Cediranib Treatment in the Patients With Recurrent Glioblastoma

BiomarkerPretreatment
8 Hours
Day 1
Day 9
Day 28
Day 56
Day 84
Day 112
MedianInterquartile RangeNo.P*PadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadj
Plasma Ang12,6581,496-4,67428NANA2,296968-3,63828.37.921,7781,298-6,35928.92.922,5781,496-5,15626.32.921,218627-2,52826.0079.0792,3651,404-3,88319.52.923,039997-3,60113.55.921,6961,426-4,15412.83.92
Plasma bFGF6831-10331NA*NA6938-10431.67*.966540-10731.79*.964930-9330.28*.964430-7930.13*.967945-11822.49*.967336-12617.96*.967539-12516.40.96
Plasma sVEGFR18355-10231NANA7770-8831.18.968671-10331.31.969164-10630.38.968470-10830.17.967968-9922.88.968374-10217.96.966957-10316.38.96
Plasma IL-1β0.310.05-0.9331NANA0.320.09-0.8631.16.940.350.08-1.0131.361.00.360.12-0.9830.421.00.420.14-0.6230.241.00.400.13-0.78221.01.00.310.02-0.7317.521.00.400.13-0.7816.191.0
Plasma IL-64.12.5-9.131NANA4.02.3-9.031.82.824.62.6-8.931.42.825.23.1-12.030.16.706.33.6-13.030.11.596.22.7-11.730.11.595.80.1-18.417.034.3110.31.4-15.716.0052.051
Plasma IL-84.22.4-9.331NANA4.32.2-7.531.22.784.52.8-8.631.78.785.13.4-8.830.42.787.54.1-10.531.012.0965.53.0-10.931.0066.0604.10.0-10.417.098.495.92.3-12.616.018.15
Plasma TNF-α22-426NANA22-426.601.022-426.961.022-426.861.022-325.831.022-417.781.022-412.621.022-512.271.0

NOTE. P values are shown with and without adjustment for multiple comparisons performed over time.

Abbreviations: Ang 1, angiopoietin 1; NA, not applicable; bFGF, basic fibroblast growth factor; sVEGFR2, soluble vascular endothelial growth factor receptor 2; IL, interleukin; TNF, tumor necrosis factor.

*P values are from the paired exact Wilcoxon test, unadjusted.
P values are from the paired exact Wilcoxon test, using Hommel's method.

Table A3.

CPCs (percent of peripheral blood mononuclear cells) After Cediranib Treatment in the Patients With Recurrent Glioblastoma

BiomarkerPretreatment
8 Hours
Day 1
Day 9
Day 28
Day 56
Day 84
Day 112
MedianInterquartile RangeNo.P*PadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadj
CPCs0.0800.055-0.12031NANA0.0800.050-0.11031.201.00.0800.051-0.10430.851.00.0550.050-0.09930.048.520.0700.042-0.12530.671.00.0700.060-0.18021.281.00.0680.048-0.13616.501.00.0800.070-0.13015.211.0

NOTE. P values are shown with and without adjustment for multiple comparisons performed over time.

Abbreviations: CPC, circulating progenitor cells; NA, not applicable.

*P values are from the paired exact Wilcoxon test, unadjusted.
P values are from the paired exact Wilcoxon test, adjusted using Hommel's method.

Table A4.

Urinary MMP Activity Levels After Cediranib Treatment in the Patients With Recurrent Glioblastoma

BiomarkerPretreatment
8 Hours
Day 1
Day 9
Day 28
Day 56
Day 84
Day 112
MedianInterquartile RangeNo.P*PadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadjMedianInterquartile RangeNo.PPadj
Urinary MMP-9/NGAL activity2.01.0-9.013NANA2.01.0-5.512.911.01.51.0-6.014.251.04.51.0-7.012.591.02.01.0-6.09.911.01.01.0-4.571.01.01.01.0-7.051.01.06.04.5-6.54.501.0
Urinary MMP-2 activity2.01.0-6.013NANA3.01.0-5.5121.01.03.01.0-6.514.751.03.51.5-7.0121.01.03.01.0-7.09.311.01.01.0-6.07.501.01.01.0-7.051.01.04.01.0-7.541.01.0
Urinary MMP-9 activity4.02.0-6.013NANA3.52.0-7.0121.01.04.53.5-6.014.381.04.03.5-7.012.971.04.03.0-6.09.961.06.04.0-6.57.381.04.02.0-6.051.01.06.05.0-6.54.381.0

NOTE. P values are shown with and without adjustment for multiple comparisons performed over time.

Abbreviations: MMP, matrix metalloproteinase; NGAL, neutrophil gelatinase-associated lipocalin; NA, not applicable.

*P values are from the paired exact Wilcoxon test, unadjusted.
P values are from the paired exact Wilcoxon test, adjusted using Hommel's method.

Table A5.

Association Between Blood Angiogenic and Inflammatory Biomarkers at Baseline, and Between Their Changes at 8 Hours and at Day 1 After Cediranib Treatment, and Radiographic Progression of Disease and Mortality in Patients With Recurrent Glioblastoma

BiomarkerPretreatment Measurement*
Change at 8 Hours*
Change at Day 1*
Progression
Mortality
Progression
Mortality
Progression
Mortality
HR95% CINo.PHR95% CINo.PHR95% CINo.PHR95% CINo.PHR95% CINo.PHR95% CINo.P
Plasma VEGF−18−64 to 8530.6350−28 to 21630.27−14−60 to 8730.70−36−70 to 3630.23−27−65 to 5530.41−50−79 to 1630.089
Plasma sVEGFR1−19−62 to 7530.6031−41 to 18930.5124−79 to 64030.81−47−89 to 15630.4384−43 to 48830.3145−49 to 31030.49
Plasma sVEGFR2−50−76 to 230.05115−41 to 12430.68−50−98 to 1,48530.70−82−99 to 39830.3286−75 to 1,26730.5489−80 to 1,65530.57
Plasma SDF1α81−22 to 32029.16−49−78 to 1829.12−75−100 to 1,54429.50−96−100 to 13229.1059−82 to 1,30129.68−42−95 to 57629.66
Plasma Ang1−3−35 to 4427.8736−6 to 9927.10−12−40 to 2926.51−10−38 to 3026.57−4−37 to 4626.85−3−35 to 4426.88
Plasma Ang26−49 to 12230.87−28−70 to 7730.48−45−93 to 32630.55−71−96 to 10530.1915−53 to 18230.76−56−83 to 1230.081
Plasma sTie228−50 to 22530.61−37−76 to 6630.35−89−100 to 39530.25−92−100 to 25430.1911−92 to 1,52630.94−56−97 to 53930.54
Plasma IL-1β−9−20 to 430.17−7−19 to 630.263−25 to 4230.84−12−20 to 5730.50−9−35 to 2830.5938−6 to 10330.12
Plasma IL-6−8−17 to 230.12−4−12 to 530.36−26−64 to 5230.40−27−69 to 7230.45−22−51 to 2330.29−18−29 to 9630.52
Plasma IL-8−2−13 to 1130.794−7 to 1730.45−48−79 to 3030.16−37−75 to 5430.31−20−61 to 6230.53−31−69 to 5430.37
Plasma TNF-α−8−37 to 3425.66−25−50 to 1325.17−29−77 to 12125.57−52−87 to 7725.28−32−80 to 13725.5513−71 to 33625.87
Plasma MMP-10−18−46 to 2530.35−12−42 to 3530.57−61−91 to 7930.22−69−92 to 2025.076−15−63 to 9830.715−50 to 11830.90
Urinary MMP-2 65 kDa−6−30 to 2512.670−26 to 35121.0−2−30 to 3711.9112−21 to 5811.51−3−29 to 3311.86−4−34 to 4111.85
Urinary MMP-9 92 kDa0−15 to 1712.961−14 to 1912.8813−16 to 5411.4239−13 to 12211.1329−15 to 9711.1710−19 to 4911.54
CPCs18−25 to 8730.4960−1 to 15730.054113−36 to 60730.20123−26 to 57330.143−49 to 10529.946−51 to 12929.89

Abbreviations: VEGF, vascular endothelial growth factor; sVEGFR, soluble VEGF receptor; SDF1α, stromal cell–derived factor-1α; Ang1, angiopoietin 1; Ang2, angiopoietin 2; sTie2, soluble Tek/Tie2 receptor; IL, interleukin; TNF, tumor necrosis factor; MMP, matrix metalloproteinase; CPC, circulating progenitor cell.

*P values are from the Wald test in a proportional hazards model.

Footnotes

Supported by Grants No. R21-CA117079, K24-CA125440, R01-CA129371, P01-CA80124, R01-CA115767, M01-RR-01066, P41-RR014075, R01-CA118764, and P01-CA455481 from the National Institutes of Health; by grants from the Federal Share/National Cancer Institute Proton Beam Program Income; by Grant No. 1UL1RR025758-01 from the Harvard Clinical and Translational Science Center, National Center for Research Resources; and by gifts from the Montesi Family Research Fund and the Simches Fund for Brain Tumor Research.

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Center for Research Resources or the National Institutes of Health.

Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.

Clinical trial information can be found for the following: NCT00305656.

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.

Employment or Leadership Position: Emmanuelle di Tomaso, Novartis (C); A. Gregory Sorensen, American College of Radiology Image Metrix (U) Consultant or Advisory Role: Tracy T. Batchelor, Acceleron (C), Exelixis (C), Imclone (C), EMD Serono (C), Schering-Plough (C); Marsha A. Moses, Predictive Bioscience (C); A. Gregory Sorensen, Genentech (C), Regeneron (C), Millennium Pharmaceuticals (C), AstraZenca (C), Mitsubishi Pharma (C), Merrimack Pharmaceuticals (C), Olea Medical (C), Siemens Medical Solutions (C), Lantheus (C), Biogen Idec (C), Epix Pharmaceuticals (C); Rakesh K. Jain, AstraZeneca (C), Dyax (C), Genzyme (U), Millennium Pharmaceuticals (C), Regeneron (C), Morphosys (C) Stock Ownership: Marsha A. Moses, Predictive Bioscience; Rakesh K. Jain, SynDevRex Scientific Honoraria: Tracy T. Batchelor, Roche, Schering-Plough; Rakesh K. Jain, Pfizer, Alnylam, Genzyme Research Funding: Tracy T. Batchelor, Millennium Pharmaceuticals, AstraZeneca, Schering-Plough; A. Gregory Sorensen, Siemens Medical Solutions, General Electric Health Care, GlaxoSmithKline, Novartis, Exelixis, Schering-Plough, AstraZeneca, Takeda Pharmaceutical, Millennium Pharmaceuticals; Patrick Y. Wen, AstraZeneca; Rakesh K. Jain, Dyax, AstraZeneca Expert Testimony: None Other Remuneration: None

AUTHOR CONTRIBUTIONS

Conception and design: Tracy T. Batchelor, Dan G. Duda, Emmanuelle di Tomaso, Marek Ancukiewicz, Jay S. Loeffler, Percy Ivy, A. Gregory Sorensen, Patrick Y. Wen, Rakesh K. Jain

Financial support: Tracy T. Batchelor, David N. Louis, A. Gregory Sorensen, Rakesh K. Jain

Administrative support: Tracy T. Batchelor, Dan G. Duda, Percy Ivy, A. Gregory Sorensen, Rakesh K. Jain

Provision of study materials or patients: Tracy T. Batchelor, Dan G. Duda, Emmanuelle di Tomaso, Scott R. Plotkin, April F. Eichler, Jan Drappatz, Fred H. Hochberg, David N. Louis, Kenneth S. Cohen, Alexis Exarhopoulos, Jay S. Loeffler, Marsha A. Moses, A. Gregory Sorensen, Patrick Y. Wen, Rakesh K. Jain

Collection and assembly of data: Tracy T. Batchelor, Dan G. Duda, Emmanuelle di Tomaso, Marek Ancukiewicz, Scott R. Plotkin, Elizabeth Gerstner, April F. Eichler, Jan Drappatz, Fred H. Hochberg, Thomas Benner, David N. Louis, Kenneth S. Cohen, Houng Chea, Alexis Exarhopoulos, Marsha A. Moses, Percy Ivy, A. Gregory Sorensen, Patrick Y. Wen, Rakesh K. Jain

Data analysis and interpretation: Tracy T. Batchelor, Dan G. Duda, Emmanuelle di Tomaso, Marek Ancukiewicz, Scott R. Plotkin, Elizabeth Gerstner, April F. Eichler, Jan Drappatz, Fred H. Hochberg, Thomas Benner, David N. Louis, Kenneth S. Cohen, Houng Chea, Alexis Exarhopoulos, Jay S. Loeffler, Marsha A. Moses, Percy Ivy, A. Gregory Sorensen, Patrick Y. Wen, Rakesh K. Jain

Manuscript writing: Tracy T. Batchelor, Dan G. Duda, Emmanuelle di Tomaso, Marek Ancukiewicz, Scott R. Plotkin, Elizabeth Gerstner, April F. Eichler, David N. Louis, Percy Ivy, A. Gregory Sorensen, Patrick Y. Wen, Rakesh K. Jain

Final approval of manuscript: Tracy T. Batchelor, Dan G. Duda, Emmanuelle di Tomaso, Marek Ancukiewicz, Scott R. Plotkin, Elizabeth Gerstner, April F. Eichler, Jan Drappatz, Fred H. Hochberg, Thomas Benner, David N. Louis, Kenneth S. Cohen, Houng Chea, Alexis Exarhopoulos, Jay S. Loeffler, Marsha A. Moses, Percy Ivy, A. Gregory Sorensen, Patrick Y. Wen, Rakesh K. Jain

REFERENCES

1. Jain RK, di Tomaso E, Duda DG, et al. Angiogenesis in brain tumours. Nat Rev Neurosci. 2007;8:610–622. [PubMed]
2. Plate KH, Mennel HD. Vascular morphology and angiogenesis in glial tumors. Exp Toxicol Pathol. 1995;47:89–94. [PubMed]
3. Rampling R, Cruickshank G, Lewis A, et al. Direct measurment of PO2 distribution and bioreductive enzymes in human malignant brain tumors. Int J Radiat Oncol Biol Phys. 1994;29:427–431. [PubMed]
4. Valk PE, Mathis CA, Prados MD, et al. Hypoxia in human gliomas: Demonstration by PET with fluorine-18-fluoromisonidazole. J Nucl Med. 1992;33:2133–2137. [PubMed]
5. Holash J, Maisonpierre PC, Compton D, et al. Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF. Science. 1999;284:1994–1998. [PubMed]
6. Shweiki D, Itin A, Soffer D, et al. Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature. 1992;359:843–845. [PubMed]
7. Millauer B, Shawver L, Plate KH, et al. Glioblastoma growth inhibited in vivo by a dominant-negative Flk-1 mutant. Nature. 1994;367:576–579. [PubMed]
8. Samoto K, Ikezaki K, Ono M, et al. Expression of vascular endothelial growth factor and its possible relation with neovascularization in human brain tumors. Cancer Res. 1995;55:1189–1193. [PubMed]
9. Schmidt NO, Westphal M, Hagel C, et al. Levels of vascular endothelial growth factor, hepatocyte growth factor/scatter factor and basic fibroblast growth factor in human gliomas and their relation to angiogenesis. Int J Cancer. 1999;84:10–18. [PubMed]
10. Batchelor TT, Sorensen AG, di Tomaso E, et al. AZD2171, a pan-VEGF receptor tyrosine kinase inhibitor, normalizes tumor vasculature and alleviates edema in glioblastoma patients. Cancer Cell. 2007;11:83–95. [PMC free article] [PubMed]
11. Kamoun WS, Ley CD, Farrar CT, et al. Edema control by cediranib, a VEGF targeted kinase inhibitor, prolongs survival despite persistent brain tumor growth in mice. J Clin Oncol. 2009:2542–2552. [PMC free article] [PubMed]
12. Sorensen AG, Batchelor TT, Zhang WT, et al. A “vascular normalization index” as potential mechanistic biomarker to predict survival after a single dose of cediranib in recurrent glioblastoma patients. Cancer Res. 2009;69:5296–5300. [PMC free article] [PubMed]
13. Winkler F, Kozin SV, Tong R, et al. Kinetics of vascular normalization by VEGFR2 blockade governs brain tumor response to radiation: Role of oxygenation, angiopoietin-1 and matrix metalloproteinases. Cancer Cell. 2004;6:553–563. [PubMed]
14. Jain RK, Duda DG, Clark JW, et al. Lessons from phase III clinical trials on anti-VEGF therapy for cancer. Nat Clin Pract Oncol. 2006;3:24–40. [PubMed]
15. Wedge SR, Kendrew J, Hennequin LF, et al. AZD2171: A highly potent, orally bioavailable, vascular endothelial growth factor receptor-2 tyrosine kinase inhibitor for the treatment of cancer. Cancer Res. 2005;65:4389–4400. [PubMed]
16. Macdonald DR, Cascino TL, Schold SC, Jr, et al. Response criteria for phase II studies of supratentorial malignant glioma. J Clin Oncol. 1990;8:1277–1280. [PubMed]
17. Duda DG, Cohen KS, Scadden DT, et al. A protocol for phenotypic detection and enumeration of circulating endothelial cells and circulating progenitor cells in human blood. Nat Protoc. 2007;2:805–810. [PMC free article] [PubMed]
18. Roy R, Louis G, Loughlin KR, et al. Tumor-specific urinary matrix metalloproteinase fingerprinting: Identification of high molecular weight urinary matrix metalloproteinase species. Clin Cancer Res. 2008;14:6610–6617. [PMC free article] [PubMed]
19. Smith ER, Manfredi M, Scott RM, et al. A recurrent craniopharyngioma illustrates the potential usefulness of urinary matrix metalloproteinases as noninvasive biomarkers: Case report. Neurosurgery. 2007;60:E1148–E1149. discussion E1149, 2007. [PubMed]
20. Smith ER, Zurakowski D, Saad A, et al. Urinary biomarkers predict brain tumor presence and response to therapy. Clin Cancer Res. 2008;14:2378–2386. [PubMed]
21. Wong ET, Hess KR, Gleason MJ, et al. Outcomes and prognostic factors in recurrent glioma patients enrolled onto phase II clinical trials. J Clin Oncol. 1999;17:2572–2578. [PubMed]
22. Ballman KV, Buckner JC, Brown PD, et al. The relationship between six-month progression-free survival and 12-month overall survival end points for phase II trials in patients with glioblastoma multiforme. Neuro Oncol. 2007;9:29–38. [PMC free article] [PubMed]
23. Lin DY, Wei LJ. The robust inference for the proportional hazard model. J Am Stat Assoc. 1989;84:1074–1078.
24. Friedman HS, Prados MD, Wen PY, et al. Bevacizumab along and in combination with irinotecan in recurrent glioblastoma. J Clin Oncol. 2009;27:4733–4740. [PubMed]
25. Kreisl TN, Kim L, Moore K, et al. Phase II trial of single-agent bevacizumab followed by bevacizumab plus irinotecan at tumor progression in recurrent glioblastoma. J Clin Oncol. 2009;27:740–745. [PMC free article] [PubMed]
26. Reardon D, Friedman H, Yung WKA, et al. A phase I/II trial of PTK787/ZK 222584 (PTK/ZK), a novel, oral angiogenesis inhibitor, in combination with either temozolomide or lomustine for patients with recurrent glioblastoma multiforme (GBM) Proc Am Soc Clin Oncol. 2004;23:110. abstr 1513.
27. Vredenburgh JJ, Desjardins A, Herndon JE, Jr, et al. Phase II trial of bevacizumab and irinotecan in recurrent malignant glioma. Clin Cancer Res. 2007;13:1253–1259. [PubMed]
28. Sorensen AG, Batchelor TT, Zhang W-T, et al. Response criteria for glioma. Nat Clin Pract Oncol. 2008;5:634–635. [PubMed]
29. Gerstner ER, Chen PJ, Wen PY, et al. Infiltrative patterns of glioblastoma spread detected via diffusion MRI after treatment with cediranib. Neuro Oncol. 2010;12:466–472. [PMC free article] [PubMed]
30. Jain RK. Normalization of tumor vasculature: An emerging concept in antiangiogenic therapy. Science. 2005;307:58–62. [PubMed]
31. Jain RK, Duda DG, Willett CG, et al. Biomarkers of response and resistance to antiangiogenic therapy. Nat Rev Clin Oncol. 2009;6:327–338. [PMC free article] [PubMed]
32. Dowlati A, Gray R, Sandler AB, et al. Cell adhesion molecules, vascular endothelial growth factor, and basic fibroblast growth factor in patients with non-small cell lung cancer treated with chemotherapy with or without bevacizumab–an Eastern Cooperative Oncology Group Study. Clin Cancer Res. 2008;14:1407–1412. [PubMed]
33. Jubb AM, Hurwitz HI, Bai W, et al. Impact of vascular endothelial growth factor-A expression, thrombospondin-2 expression, and microvessel density on the treatment effect of bevacizumab in metastatic colorectal cancer. J Clin Oncol. 2006;24:217–227. [PubMed]
34. Drevs J, Siegert P, Medinger M, et al. Phase I clinical study of AZD2171, an oral vascular endothelial growth factor signaling inhibitor, in patients with advanced solid tumors. J Clin Oncol. 2007;25:2993–2995. [PubMed]
35. Burstein HJ, Elias AD, Rugo HS, et al. Phase II study of sunitinib malate, an oral multitargeted tyrosine kinase inhibitor, in patients with metastatic breast cancer previously treated with an anthracycline and a taxane. J Clin Oncol. 2008;26:1810–1816. [PubMed]
36. Drevs J, Zirrgiebel U, Schmidt-Gersbach CI, et al. Soluble markers for the assessment of biological activity with PTK787/ZK 222584 (PTK/ZK), a vascular endothelial growth factor receptor (VEGFR) tyrosine kinase inhibitor in patients with advanced colorectal cancer from two phase I trials. Ann Oncol. 2005;16:558–565. [PubMed]
37. Motzer RJ, Michaelson MD, Redman BG, et al. Activity of SU11248, a multitargeted inhibitor of vascular endothelial growth factor receptor and platelet-derived growth factor receptor, in patients with metastatic renal cell carcinoma. J Clin Oncol. 2006;24:16–24. [PubMed]
38. Norden-Zfoni A, Desai J, Manola J, et al. Blood-based biomarkers of SU11248 activity and clinical outcome in patients with metastatic imatinib-resistant gastrointestinal stromal tumor. Clin Cancer Res. 2007;13:2643–2650. [PubMed]
39. Zhu AX, Sahani DV, Duda DG, et al. Efficacy, safety, and potential biomarkers of sunitinib monotherapy in advanced hepatocellular carcinoma: A phase II study. J Clin Oncol. 2009;27:3027–3035. [PMC free article] [PubMed]
40. Xu L, Duda DG, di Tomaso E, et al. Direct evidence that bevacizumab, an anti-VEGF antibody, up-regulates SDF1alpha, CXCR4, CXCL6, and neuropilin 1 in tumors from patients with rectal cancer. Cancer Res. 2009;69:7905–7910. [PMC free article] [PubMed]
41. Ebos JM, Lee CR, Cruz-Munoz W, et al. Accelerated metastasis after short-term treatment with a potent inhibitor of tumor angiogenesis. Cancer Cell. 2009;15:232–239. [PubMed]
42. Paez-Ribes M, Allen E, Hudock J, et al. Antiangiogenic therapy elicits malignant progression of tumors to increased local invasion and distant metastasis. Cancer Cell. 2009;15:220–231. [PMC free article] [PubMed]

Articles from Journal of Clinical Oncology are provided here courtesy of American Society of Clinical Oncology