research cediranib/azd2171 inhibits bone and brain...

Ther

CedPre

Juan

Abst

Intro

Metfor canto dist(1). AgrowtmicroVEGFbenefitientsshort-antiandecreahave bsuch pare sig

AuthorInstitutMouseStroke,

Note:Resear

CorresCenterRoom 14655; E

doi: 10

©2010

Cance8662

D

Published OnlineFirst October 19, 2010; DOI: 10.1158/0008-5472.CAN-10-1435
Published OnlineFirst on October 19, 2010 as 10.1158/0008-5472.CAN-10-1435
Canceresearch

apeutics, Targets, and Chemical Biology

iranib/AZD2171 Inhibits Bone and Brain Metastasis in a

R

clinical Model of Advanced Prostate Cancer

Juan Yin1, Luhua Zhang1, Jeeva Munasinghe2, R. Ilona Linnoila1, and Kathleen Kelly1

ractLate

treatmthat intherapgestedportanof cedand brrebralShort-establgrowthimal s

reclinicalnificant.

s' Affiliatione and 2LabImaging FaNIH, Bethe

Supplemench Online (h

ponding Aufor Cancer R068, Bethe-mail: kellyk

.1158/0008-

American A

r Res; 70(

ownloade

stage or aggressive cancers exhibit metastatic growth at multiple sites, and the characterization ofent response in various organs to drugs with potentially wide-ranging efficacy is needed. Tumor cellsduce angiogenesis are a common characteristic of metastatic disease, and clinically, antiangiogenicies have shown value in the setting of advanced cancer. However, recent preclinical studies have sug-that exposure to antiangiogenic drugs can increase tumor invasiveness and metastasis, making it im-t to determine which contexts antiangiogenic therapy is most appropriate. We describe here the effectsiranib, a receptor tyrosine kinase inhibitor, in a model of advanced prostate cancer metastatic to skeletonain. Treatment with cediranib decreased metastatic tumor burden in the brain and bone, decreased ce-vasogenic edema, and improved survival, despite increasing the invasive histology of brain metastases.duration cediranib treatment given at the time of tumor cell dissemination was sufficient to inhibit theishment and subsequent growth of bone metastases, although brain metastases were subject to reboundafter the discontinuation of cediranib. Distinct growth patterns at different organ sites in the same an-

howed that certain tumor microenvironments such as bone may be most amenable to interventions byascular endothelial growth factor (VEGF) therapies. In addition, anti-VEGF treatment may be of utility in
anti–v
decreasing the rapid growth of solid brain metastases and vasogenic edema in patients with advanced cancer,leading to reduced morbidity and associated clinical benefit. Cancer Res; 70(21); 8662–73. ©2010 AACR.

metasresponenvirogans.risks apies foBon

lung,othersing mwell-vpressecell pdepenconce

duction

astasis is the major cause of morbidity and mortalitycer patients. Advanced cancers frequently metastasizeant organs, and multiorgan metastasis is not unusualcommon feature of metastases is vascular endothelialh factor (VEGF)–initiated angiogenesis in the tumorenvironment (2). Antiangiogenic therapies targetingand VEGF receptors (VEGFR) have provided survivalts when combined with chemotherapy for treating pa-with late-stage disease, although the benefits are mostlylived (3). Recent preclinical studies have suggested thatgiogenic therapies increase tumor invasiveness andse overall survival (4, 5). Although the models thateen analyzed to date are limited, the implications of

investigations for future clinical trial designsThus, it is important to expand the range of

ture aare evare avtastastors thlittle idevelotases tThe

tasis odoes nhas esfectorstatic

s: 1Cell and Cancer Biology Branch, National Canceroratory of Functional and Molecular Imaging andcility, National Institute of Neurological Disorders andsda, Maryland

tary data for this article are available at Cancerttp://cancerres.aacrjournals.org/).

thor: Kathleen Kelly, Cell and Cancer Biology Branch,esearch, National Cancer Institute, 37 Convent Drive,sda, MD 20892. Phone: 301-435-4651; Fax: [email protected].

5472.CAN-10-1435

ssociation for Cancer Research.

21) November 1, 2010

Research. on September 16, 2cancerres.aacrjournals.org d from

tasis models and to analyze various parameters ofse to anti-VEGF therapies, including the tumor micro-nment context of metastatic growth in different or-This is a consideration for assessing organ-specificnd benefits and especially relevant in designing thera-r treating multiorgan metastasis.e and brain are common metastatic sites for breast,and prostate cancer, as well as melanoma among, and patients with such cancers are at risk for suffer-etastases in both sites simultaneously (6, 7). Brain is aascularized tissue, and many of the properties ex-d by tumors growing in the brain such as endothelialroliferation and increased vascular permeability aredent on VEGF (8). Thus far, most of the studiesrning the interactions of tumor cells and brain vascula-s well as the response of tumors to anti-VEGF therapyaluations of malignant gliomas (9, 10). Much less dataailable concerning the development of clonal brain me-es and their response to small molecule VEGFR2 inhibi-at penetrate the blood-brain barrier. Similarly, relativelys known about the vascular remodeling that occurs inping bone metastases and the response of such metas-o anti-VEGF treatment.DU145 line was originally isolated from a brain metas-f a prostate cancer patient, but the parental DU145ot form metastases in xenograft models. Previous worktablished that the introduction of an activated Ras ef-
mutant (RasV12G37) resulted in a gain of bone meta-capability following intracardiac inoculation (11). To
018. © 2010 American Association for Cancer

http://cancerres.aacrjournals.org/

improwere pisolatducesfactorat a h(AZD2VEGFof varirespoof mu

Mate

Cell cDU

pSFGnactivat

AnimAni

approcardiaas destion ahistolwhenbodywere r

ExperCed

dissolwaterby gar

Figuremorbiddesigninoculavehiclegroup),randomwas trewas treweightdifferenof tumo

Antiangiogenic Treatment for Metastases

www.a

D


ve metastatic efficiency, the DU145(RasV12G37) cellsassaged in vivo, and the DU145/RasB1 cell line was

ed from a bone metastasis (12). This cell line pro-high levels of VEGFA and platelet-derived growthB (PDGFB) and metastasizes to bone and brainigh frequency. In this study, we used cediranib171), a tyrosine kinase inhibitor directed againstR1/2 and PDGF receptor, to determine the efficacyous treatments and to characterize the organ-specificnses to antiangiogenesis regimens in the settingltiorgan metastasis.

rials and Methods

ulture145/RasB1 cells were infected with the retrovirus
esTGL, and positive cells were isolated by fluorescence-ed cell sorting (12).
mortatal des

r-bearing mice in each group. *, P < 0.05.

acrjournals.org

Research. on September 16, 2cancerres.aacrjournals.org ownloaded from

al studiesmal work was performed in accordance with a protocolved by NIH Animal Care and Use Committee. Intra-c inoculation and bioluminescent imaging (BLI) werecribed (12). See Supplementary Fig. S1 for the correla-nalyses between BLI and tumor burden determinedogically. For survival studies, mice were euthanizedone of the following situations applied: 10% loss ofweight, paralysis, or head tilting. All animal studiesepeated three times.

imental designiranib was provided by AstraZeneca. Cediranib wasved with 1% (w/v) aqueous polysorbate 80 in deionizedand given at 6 mg/kg body weight, a therapeutic dose,vaging mice daily. To study the effect of cediranib on
lity, morbidity, and tumor progression, an experimen-ign as shown in Fig. 1A was used. Each of the four final
1. Cediranib decreased mortality andity of tumor-bearing mice. A, experimentalfor cediranib treatment. Cells wereted at day 0. Mice were treated withor cediranib for the first 3 wk (n = 20 perand subsequently both groups werely divided into two groups. One groupated with vehicle, and the other groupated with cediranib (n = 9–10). B, bodychange of tumor-bearing mice int groups. *, P < 0.001. C, survival rate

Cancer Res; 70(21) November 1, 2010 8663



experition. Fpies, tinoculwereof hypand co

MagnT2 w

(16 slito deliresolu8; numQuant(TE/Tresoluslice pquiredcessedMATL

HistolBra

malinimmueach mone sejected30 minused: rtion kphomcollectcell prtifiedanalysviouslvesseltion odoubleand FFluoremeasudefinemass,tumoron thμm2),and smEDTA

DataDat

Softwrate was thesignif

Resu

CedirtumoIt h

bone mter intMoreoangiog(EGF)leptinbasedmineto anta smabloodmentainhibitreatmwarddetermventiomor cwere cfrom tThe

sure otumortreatmperiodgroupthis, c3 weeking migroupmice iwas nothe cogroupsociatmultipBy consympttasis i

CedirinhibCon

of tumbraincompaventio3 weektained(weekthe bu

Yin et al.

Cance8664

D


mental groups contained 9 or 10 animals at the initia-or combined cediranib and Zometa (Novartis) thera-he treatments were started 3 weeks after tumor cellation. To quantify osteolytic destruction, long bonesimaged with a Faxtron X-ray machine. The numberserintensive regions were counted for each long bonenfirmed with histology.

etic resonance imagingeighted axial slices, encompassing the whole brainces), were acquired using a fast spin echo sequenceneate anatomic details (field of view, 19.2 mm; in-planetion, 75 μm; TE/TR, 50 ms/3,000ms; echo train length,ber of averages, 8; slice thickness, 1 mm; 16 slices).itative T2 weighted images of nine 1-mm axial slicesR, 15/3,000 ms; number of echo images, 16; in-planetion, 150 μm; interslice gap, 1.5 mm), with the firstositioned 1.5-mm anterior to the Bregma, were ac-. Magnetic resonance imaging (MRI) data were pro-and analyzed using software routines written in

AB (Mathworks, Inc.).

ogy and immunohistochemistryin tissue was collected and fixed in 10% buffered for-for H&E staining or 4% paraformaldehyde for

nohistochemical staining. For morphometric analysis,ouse brain was coronally cut into four equal quarters;ction from each quarter was analyzed. Mice were in-with bromodeoxyuridine (BrdUrd; Sigma, 70 mg/kg i.p.)utes before euthanasia. The following antibodies wereat anti-BrdUrd (OBT), ApopTag in situ apoptosis detec-it (Millipore), CD34, and αSMA (Abcam). For histomor-etric analysis, bright field microscopic images wereed using an Axioplanmicroscopy system (Zeiss). Tumoroliferation (BrdUrd labeling) and apoptosis were quan-using AxioVision software (Zeiss). Histomorphometricis of CD34-stained vessels was performed based on pre-y described protocols with modifications (13). Bloodarea was measured at 200× magnification. Coregistra-f endothelial cells and pericytes was analyzed throughstaining with CD34 and αSMA followed by rhodamine-ITC-conjugated secondary antibodies, respectively.scent cells were counted at 400× magnification. Tore tumor invasiveness, the number of invasive edges,d as clusters of cells outside the contour of the tumorwas counted for each tumor and corrected for totalarea. Tumors were divided into four groups based

e total area of each tumor: extra large (>10 × 105

large (5–10 × 105 μm2), medium (1.5–4.9 × 105 μm2),all (<1.5 × 105 μm2). Bones were decalcified in 10%for 2 weeks before processing.

analysisa were analyzed using Prizm software (GraphPadare, Inc.) by repeated measures ANOVA. Survivalas analyzed by log-rank test. Data are expressed
mean ± SEM, with P < 0.05 considered statistically
icant.group(Fig. 2

r Res; 70(21) November 1, 2010


lts

anib improved mortality and morbidity ofr-bearing miceas been previously shown that DU145/RasB1 cells formetastasis and highly vascularized brain metastasis af-

racardiac inoculation into immunocompromised mice.ver, DU145/RasB1 cells secrete VEGFA, PDGFB, andenin, but not angiopoietin-2, epidermal growth factor, basic fibroblast growth factor, heparin-binding EGF,, or placental growth factor, as determined by ELISA-quantibody array (Raybiotec, Inc.; ref. 12). To deter-the physiologic response of brain and bone metastasisiangiogenic therapy, mice were treated with cediranib,ll molecule VEGFR antagonist that is permeable to the-brain barrier (9, 14). A graphic depicting the experi-l design is shown in Fig. 1A. To determine if cediranibts the growth of established brain and bone metastasis,ent group mice were given cediranib from week 3 on-when metastases can first be detected using BLI. Toine if cediranib prevents metastatic colonization, pre-n group mice were treated at the time of systemic tu-ell inoculation and thereafter for 3 weeks. Mice alsoontinuously treated in the prevention/treatment grouphe time of tumor cell inoculation.effects of cediranib treatment on morbidity, one mea-f which is body weight, and mortality were evaluated in-bearing mice. As shown in Fig. 1B, the mice in theent groups maintained their body weight for the study, whereas the mice in the control and the preventions began losing weight at ∼4 weeks. Consistent withediranib treatment, either continuously or starting afters, significantly improved the survival rate of tumor-bear-ce (Fig. 1C). Seventy percent of mice in the vehicle controldid not survive beyond 5 weeks, whereas 50% or more ofn the treatment groups were still alive after 7 weeks. Therestatistically significant difference between the survival ofntrol and prevention groups. Most mice in the controlwere euthanized as a result of neurologic disorders as-ed with edema, focal brain dysfunction secondary tole nodular tumors in the cerebrum and cerebellum.trast, cediranib-treated mice rarely showed neurologicoms. The percentage of animals that developed metas-s detailed in Supplementary Table S1.

anib decreased brain metastasis burden byiting solid tumor growthtinuous treatment with cediranib starting at the timeor cell dissemination resulted in significantly lessmetastatic burden after 3 weeks as measured by BLIred with nontreated animals (Fig. 2A, left). In the pre-n arm, which encompasses treatment for the firsts after inoculation, tumor growth inhibition was main-for the first week after discontinuation of treatment4), but 2 weeks after the withdrawal of cediranib (week 5),rden of brain metastasis was similar to the untreated
, suggesting that the inhibitory effects did not lastA, right). In the treatment arm, cediranib inhibited the
Cancer Research



growthhibito(weekgroupdistribmice das welperimemors abraininfiltraoped mtion wprotoction oTo

param

develoquentmorphsels incells ssity oftreatetreatmtheir cder (Fwere rTo

liferatdeoxying sta

Figuredevelopbrain mburden(left) andifferen0.05. Bimagestype (riT, tumoarrowsC, percsolid tuhistoloeach geach cnumbetypes oratio beincideneach group.


www.a

D


of established brain metastasis (Fig. 2A, right). The in-ry effect was obvious even after 1 week of treatment4). At the end point, the brains from each experimentalwere analyzed histologically in four coronal sections,uted equally throughout the brain area. The number ofisplaying tumors and the total number ofmice evaluated,l as the absolute number of tumors observed in each ex-ntal group, are indicated in Fig. 2C. Expansive solid tu-nd infiltrative tumors were observed in all regions of the(Fig. 2B). Solid tumors generally were larger in size thantive tumors (Fig. 4B). Vehicle-treated control mice devel-ore solid tumors than infiltrative tumors. This propor-as reversed in cediranib-treated mice. The treatmentol used for the prevention group also altered the propor-f tumor types toward more invasive tumors (Fig. 2C).
evaluate the effect of cediranib treatment on variouseters of tumor morphology, tumors were allowed to
serialtreate

acrjournals.org


p for 3 weeks after tumor cell inoculation, and subse-ly, mice were treated with cediranib for 1 week. Vesselology was analyzed using CD34 staining (Fig. 3A). Ves-both types of tumors seemed dilated, and endothelialeemed hypertrophic relative to normal brain. The den-blood vessels in solid tumors in control and cediranib-d mice is quantified in Fig. 3B. Following cediranibent, solid tumors lose blood vessel density mainly inenter while retaining a rim of vessels at the tumor bor-ig. 3A). By contrast, the vessels in infiltrative tumorselatively insensitive to cediranib (data not shown).determine the effects of drug treatment on tumor pro-ion and apoptosis, BrdUrd (Fig. 3A) and terminalnucleotidyl transferase-mediated dUTP nick end label-ining (data not shown) were separately performed on

2. Cediranib inhibited thement and progression ofetastasis. A, brain tumormeasured by BLI at week 3d weeks 4 and 5 (right) int groups (n = 9–10). *, P <, representative histologicof infiltrative (left) and solidght) brain metastasis.r. Scale bar, 100 μm. Yellowindicate tumor edges.entage of infiltrative andmor types quantified ongic sections of brain inroup. The numbers aboveolumn represent the totalr of specific histologicf brain metastases. Thelow the graph is thece of brain metastasis in

histologic sections from both control and cediranib-d mice. BrdUrd staining showed that the presence of




Figureproliferinfiltraticontrolin greein solid

Yin et al.

Cance8666

D


3. The effects of cediranib on tumor blood vessel density and tumor cell proliferation. A, CD34-positive endothelial cells and BrdUrd-positiveating cells in consecutive sections of brain metastases from control (left) and cediranib-treated (right) mice. Top, large solid tumors; bottom,ve tumors. Scale bar, 100 μm. B, blood vessel areas (left; n = 12) and BrdUrd-positive cells (right) were quantified in solid brain tumors fromand cediranib-treated groups. C, representative double-stained images for CD34-positive endothelial cells in red (left) and αSMA-positive pericytes
n (middle) in a solid tumor. Right, merged image. Scale bar, 20 μm. D, quantification of the ratio of αSMA- and CD34-positive blood vesselsand infiltrative tumors. *, P < 0.05.
r Res; 70(21) November 1, 2010 Cancer Research

Research. on September 16, 2018. © 2010 American Association for Cancercancerres.aacrjournals.org ownloaded from


dividinvesselthe cethe tuthe bogroupencesmors(Fig. 3mors win FigcantlysomenecrotBec

lackinlized bvesselpericycostaiimatetumorfor theOne ppericy

percentantlyafterwith αfollowFina

is thethe tuness.respecedgescediramors lvelopsmalleativelyand cof smathe grthe 1-

CedirCed

Figureaffectstumorsof largecontroltreatedindicate100 μmnumbetumorsabsolutquantifthe dat***, P < 0.001.


www.a

D


g cells correlated spatially with the pattern of blooddensity observed on CD34 staining. Solid tumors indiranib-treated group showed BrdUrd-labeled cells atmor border with significantly fewer BrdUrd+ cells indy of the tumor compared with tumors in the control(Fig. 3A). By contrast, there were no significant differ-in BrdUrd-labeled cells for the smaller infiltrative tu-when comparing control and cediranib-treated miceA). The quantification of BrdUrd+ cells in multiple tu-ith or without 1 week of cediranib treatment is shown

. 3B. The number of apoptotic cells was not signifi-increased in the cediranib-treated mice, althoughtumors from cediranib-treated animals displayedic centers (data not shown).ause VEGF withdrawal can cause angiogenic vesselsg pericytes to undergo regression whereas those stabi-y pericytes undergo survival, we investigated whethersurvival was correlated with coregistration of ate marker (15). Endothelial cells and pericytes werened for CD34 and αSMA, respectively (Fig. 3C). Approx-ly, 60% and 30% of vessels in solid and infiltratives, respectively, stained for αSMA (Fig. 3D). The reasonreduced staining in infiltrative vessels is not known.
ossibility is that a more elongated morphology oftes in infiltrative tumor vessels leads to the lower
tomaof ede

acrjournals.org


tage of sectioned vessels scored as positive. Impor-, the ratio of αSMA/CD34 did not change before andcediranib treatment (Fig. 3D). Therefore, associationSMA did not correlate with selective survival of vesselsing antiangiogenic treatment.lly, an important observation for antiangiogenic therapyphenotype of tumors affected by inhibition of VEGFR2 inmor vasculature resulting in increased tumor invasive-A quantitative analysis of tumors categorized witht to tumor area and the relative density of invasiveis shown in Fig. 4B for control and mice treated withnib. Although cediranib treatment leads to fewer tu-arger than 10 × 105 μm2, those large tumors that do de-have an increased number of invasive edges (Fig. 4A). Allr tumors have an invasive morphology and showed rel-constant numbers of invasive edges between control

ediranib-treated mice. Because the absolute numbersll tumors did not change with treatment, it seems thatowth of newly seeded tumors was not significant duringweek observation period.

anib treatment decreased vasogenic edemairanib has been shown to prevent edema in glioblas-

4. Cediranib treatmentthe invasiveness in larger. A, representative imagesr brain metastases from(left) and cediranib-(right) mice. Black arrowsinvasive edges. Scale bar,. B, quantification of ther of invasive edges forof various areas. Thee numbers of tumorsied are indicated abovea points. *, P < 0.05;

patients, and we hypothesized that the ameliorationma secondary to brain metastases was a factor in




the imaddrein thebetwemetasand edicant egroupsreimaand quof ede

the brno dismorphing thecedirain hist(Fig. 5alongintens

Figure(right foand bestain ofbar, 20Cediran

Yin et al.

Cance8668

D


proved survival of cediranib-treated mice (9, 16). Toss the effects of cediranib on vasogenic brain edemaDU145Ras/B1 model system, BLI and MRI were useden 3.5 and 4.5 weeks after the initiation of experimentaltasis to evaluate tumor burden, anatomic structures,ema. Tumor-bearing mice without evidence of signif-dema were randomized to control and cediranib-treated. After 7 days, five control and three treated mice wereged. As shown by a representative example in Fig. 5A
antified in Fig. 5B, the T2 maps showed accumulation To
0 μm. T, tumor. D, representative T2 scans (left two panels) and T2 maps (right tib was administrated after brain edema had developed. n = 3. Arrows indicate b



ain. The animals treated with cediranib, however, showedcernable difference in T2 values after 1 week, although theology scans showed an increase in tumor diameter dur-study period (Fig. 5A and B). The prevention of edema bynib was confirmed by Luxol fast blue staining for myelinologic sections from control and treated mouse brainsC). Demyelination, which is indicative of edema, was seenthe corpus callosum of control mice, matching the hyper-ive areas in the T2 maps (Fig. 5A).
test whether cediranib can ameliorate established
ma in the control mice along the white matter tracts of edema, three tumor-bearing mice with radiological

5. Cediranib inhibited the development and progression of brain edema in tumor-bearing mice. A, T2 scans (left four panels) and T2 mapsur panels) of brains in control (top) and cediranib-treated (bottom) mice. Mice were treated for 1 wk when brain metastases were obvious by BLIfore edema developed. Red arrows indicate brain metastasis. B, quantification of brain edema using T2 values from Fig. 4A. C, Luxol fast bluemouse brains show separation of myelin (red arrow) in control brain, whereas myelin remains intact in cediranib-treated brain (yellow arrow). Scale

wo panels) of brains pre- and post-cediranib treatment.rain metastasis.

Cancer Research



evidenContrwouldlished

out sirevea

Figure(left) an*, P < 0 repreof bloo the bo


www.a

D


ce of edema were treated with cediranib for 7 days.ol mice were not included in this arm, as they

.05; n = 9-10. C, CD34 stains of endothelial cells in bone metastases fromd vessel area on CD34-stained sections. ***, P < 0.001. Numbers indicate

not survive for an additional 7 days with estab-brain edema. Cediranib-treated mice survived with-

expanpholo

acrjournals.org


gnificant morbidity to the second scan date, whichled stabilization or reduction of edema, despite

sentative control and cediranib-treated mice. D, quantificationnes analyzed in each group.

6. Cediranib inhibited the development and progression of bone metastases. A, bone tumor burdens were measured by BLI at weeks 2 and 3d at weeks 4 and 5 (right; note scale differences). *, P < 0.05; n = 9-10. B, histomorphometric analysis of bone metastases on H&E-stained sections.

sion of the metastatic lesions as determined by mor-gy scans (Fig. 5D).




CedirduratThe

effects

gressiDU145

Figuren = 9-1*, P < 0

Yin et al.

Cance8670

D


anib inhibited bone metastasis even after shortion treatment

1. B, representative H&E orange G–stained sections of proximal tibia from each.05; **, P < 0.01; n = 9-11. D, brain tumor burden quantified by BLI on week 5. *



on of bone metastases. Intracardiac inoculation of/RasB1 cells leads to osteolytic bone metastases, and

re has been relatively little preclinical evaluation of theof antiangiogenic agents on the development and pro-

BLI of the long bones and spine was assessed in the fourgroups shown in Fig. 1A. The inhibition of bone metastatic

7. Cediranib had similar efficacy as Zometa in inhibiting bone metastases. A, bone tumor burden measured by BLI in each group. *, P < 0.05;
group. T, tumor. C, number of bone lesions quantified on X-rays.*, P < 0.01.
Cancer Research



growthmentdid noat weegrowttreatmfrom tBLI atwith hX-raysBon

the ef(Fig. 6densitmanyas thefromto totanib tVEG

osteoccedirateoclabone ibearincediramice, cFig. S2and femeterBisp

varietysionsnib dicediraapy agtion. Bof tumbinatiand apof bontases ithoseThe bshowein theanalysbitionallel Bno staden (F

Discu

Wetion aof hemtasizes

criticaextravcells. Tand boccurrunusutime oter thesurvivand dThe

facetemice wsive sobraintumorof thethe tucells acediralowinvesseland mPre

both tas welof gliosponsKim acancetumoroptionassessto ZDbenefiThe prdata sdifferevasculVes

selectbrainsion toRasB1explaioptingsive anot sebly ditreatmtastasaginginfiltrbevaciwith bto inh


www.a

D


was evident after 2 and 3 weeks of continuous treat-(Fig. 6A, left). In the prevention group, bone metastasist relapse at week 5, following withdrawal of cediranibk 3 (Fig. 6A, right). This contrasts with the renewedh of brain metastasis at this same time point. Also,ent of established bone metastasis either continuouslyheir initiation or starting at week 4 onward inhibitedleast 10-fold (Fig. 6A, right). The BLI data were confirmedistomorphometric analysis of bone sections (Fig. 6B) and(not shown).e metastases were stained for CD34 to determinefect of cediranib on vessel number and morphologyC and D). Nontreated mice displayed a uniform highy of CD34+ vessels within bone metastases, andvessels appeared dilated. The vessel density as welldilated appearance decreased in bone metastases

cediranib-treated animals. The vessel area relativeal tumor area decreased ∼6-fold as a result of cedir-reatment (Fig. 6D).FA binding to VEGFR1 has been shown to stimulatelast differentiation, migration, and activity (17). Becausenib has some inhibitory activity for VEGFR1 (14), the os-st number per millimeter normal bone or per tumornterface was quantified in non–tumor-bearing and tumor-g mice, respectively, either untreated for 4 weeks ornib-treated between weeks 3 and 4. In non–tumor-bearingediranib did not affect osteoclast number (Supplementary). Although cediranib-treated mice developed smallerwer bone metastases, the osteoclast number per milli-tumor bone interface was not affected.hosphonates, such as Zometa, are currently used for aof patients with metastatic bone lesions, including le-

originating from prostate cancer (18). Because cedira-d not seem to inhibit osteoclast activity, the effects ofnib and Zometa were directly compared as monother-ents and as potentially synergistic drugs in combina-LI revealed that continuous treatment from the timeor cell inoculation with cediranib, Zometa, and a com-on of both drugs inhibited tumor burden significantlyproximately equally (Fig. 7A). Histopathologic analysese sections showed that there were fewer bone metas-n cediranib-treated mice compared with controls, andmetastases that did develop were generally smaller.one metastases growing in the presence of Zometad metaphyseal sclerosis with tumor cells embeddedtrabecular bone (Fig. 7B). Consistent with the histologicis, Zometa treatment resulted in almost complete inhi-of radiologically evident osteolytic lesions (Fig. 7C). Par-LI of the brain at 5 weeks showed that Zometa alone hadtistically significant effect on the brain metastatic bur-ig. 7D).

ssion

investigated the efficacy of various cediranib preven-nd treatment regimens using an experimental model
atogenously disseminated prostate cancer that metas-predominantly to brain and bone. This system models
mor cgenesi

acrjournals.org


l steps in metastatic spread including dissemination,asation, and colonization of selected prostate cancerhe DU145/RasB1 model is specifically useful as brain

one are major sites of metastasis for many commonlying cancers, and co-occurrence at the two sites is notal (7, 8). Cediranib treatment starting either from thef DU145/RasB1 intracardiac inoculation or starting af-establishment of micrometastases led to a significant

al benefit that correlated with decreased tumor burdenecreased cerebral vasogenic edema.response of brain metastases to cediranib was multi-d and displayed distinct histologic features. Untreatedere more likely to develop rapidly growing and expan-lid brain metastases, which contributed to most of thetumor burden, whereas the occurrence of small invasives was more rare. Cediranib treatment led to regressionvessels in the center of large tumors. By contrast, neithermor vasculature of the invasive tumors nor the tumort the rim of the large expansive tumors regressed afternib treatment. The perivascular pattern of growth fol-g antiangiogenic treatment was characteristic ofco-option, which has been observed also for glioblastomaelanoma (19, 20).clinical models have implications for understandinghe mechanisms of treatment response and resistance,l as potential morbidity and mortality benefits. Outsideblastoma, there are limited examples analyzing the re-e of tumor cells in the brain to antiangiogenic treatment.nd colleagues described the response of MDA-231 breastr brain metastases to PTK 787, resulting in decreasedburden but no obvious survival benefit (21). Vessel co-was not specifically addressed. Leenders and colleaguesed the response of cerebral melanoma metastases6474, wherein a decreased tumor burden, no survivalt, and evidence of vessel co-option were observed (22).eclinical model presented here adds to the small body ofhowing that vessel co-option is a general response ofnt cell types to antiangiogenic treatment in the highlyarized brain microenvironment.sel co-option following treatment may represent strongive pressure on the propensity of tumor cells in theto colonize preexisting vessels as a result of their adhe-the vascular basement membrane (23). In the DU145/model, a morphologic difference was not apparent ton the sensitivity of endothelial cell survival in the co-versus solid tumors. The vasculature in both expan-

nd invasive tumors appeared dilated, and there didem to be a selective survival of vessels with demonstra-fferent pericyte coverage following antiangiogenicent. The lack of response of the perivascular brain me-es seen here is consistent with a limited amount of im-data from clinical trials suggesting increased

ative growth in advanced glioblastoma treated withzumab (10). Increasing long-term survival in patientsrain metastasis may require combination therapiesibit several functions including adhesion between tu-
ells and cerebral vessels, infiltrative growth, and angio-s. However, for palliative therapy, it seems likely that



cediraity assthe sitRec

icallymodelleadsformaanalysvealedtreatmOn thstaticcontinand cotreatmall surbreastof primcontracontrienviroantianjor sitewith M“metastancecoloniantianAn

the dito ced3 weedissemof bracantlycells inot su

alterncediraOne

cell nicdifferedifferethemsparento thethe brhave bthat cbonematopevidencontrotion ththe grbone eIn s

genicprostaducedTheseapprogenic

Discl

No p

Grant

IntraCancer

Theof pageaccorda

Refe1. Ch

gro2:5

2. Ellitum

3. LofueCa

4. EbKea p

5. Pamatan

6. Kinof

7. PaPS20

8. Ja

Yin et al.

Cance8672

D


nib treatment will be useful for mitigating the morbid-ociated with progressive brain metastasis, similar touation with advanced glioblastoma (9).ent attention has focused on observations from genet-engineered mouse tumor and experimental metastasiss showing that inhibition of the VEGF/VEGFR2 axisto increased tumor invasiveness and micrometastasistion (4, 5). Consistent with these studies, histologices of established DU145/RasB1 brain metastases re-increased invasive projections after 1 week of cediranibent, which were most apparent in the largest tumors.e other hand, increased survival and decreased meta-tumor burden in DU145/RasB1 inoculated mice treateduously with cediranib contrasts with the results of Eboslleagues (4). In the Ebos experimental design, sunitinibent accelerated metastatic burden and decreased over-vival in mice following introduction of MB-MDA-231cancer cells via tail vein injection or following removalary orthotopically grown tumors. It is likely that thesesting end points are in part a result of tumor-specificbutions. In addition, we suggest that the tumor micro-nment in various organs will be differently affected bygiogenic treatments. Specifically, it seems that the ma-of metastatic burden was the lungs in the Ebos studiesB-MDA-231. The lungs may be most susceptible to

static conditioning,” which is thought to reflect circum-s that increase tumor cell extravasation and initialzation, such as in response to stress/injury, includinggiogenic treatment (4, 24).interesting observation to emerge from our studies isfferent sensitivity of brain or bone resident tumor cellsiranib treatment and subsequent withdrawal. Afterks of cediranib treatment from the time of tumor cellination, cediranib withdrawal led to rebound growthin metastases, whereas bone metastases were signifi-inhibited. Thus, individual or micrometastatic tumor
n the bone compared with the brain seemed to either
Rece

. The biology of metastasis to a sanctuary site. Clin Cancer Res07;13:1656–62.in RK, di Tomaso E, Duda DG, Loeffler JS, Sorensen AG,

Ba20

9. Barecan11

10. Nabean

11. Yinpro75

12. JutioregCli

13. Laenrelvas



atively, were not able to emerge from dormancy afternib withdrawal.interpretation of these results is that tumor-initiatinghes in the brain and bonemicroenvironment are affectedntly by inhibition of the VEGF axis. The source of suchntial effects could be differences in the endothelial cellselves as well as the differentiated cells making up thechyma of the organ (25). Vascular niches that contributesurvival and establishment of glioma-initiating cells inain and leukemia-initiating cells in the bone marroween described (26, 27). An alternative interpretation isediranib directly disrupts osteogenic cells involved inremodeling through effects on VEGFR1-expressing he-oietic cells, such as osteoclasts. However, we found noce for functional effects of cediranib on osteoclasts inl or tumor-bearing mice. Thus, we favor the interpreta-at, at the dosage used in this study, cediranib influencesowth of bone metastasis via an effect on VEGFR2 andndothelium.ummary, our work provides evidence that antiangio-treatment not only inhibited the growth of aggressivete cancer metastases in bone and brain but also re-the morbidity and mortality of tumor-bearing mice.results support a value for clinical trials investigatingpriate combination therapies that include antiangio-drugs for treatment of patients with advanced cancer.

osure of Potential Conflicts of Interest

otential conflicts of interest were disclosed.

Support

mural Research Programs of NIH, National Cancer Institute, Center forResearch and the National Institute of Neurological Disorders and Stroke.costs of publication of this article were defrayed in part by the paymentcharges. This article must therefore be hereby marked advertisement innce with 18 U.S.C. Section 1734 solely to indicate this fact.

ived 04/21/2010; revised 07/26/2010; accepted 08/23/2010; published
rvive a VEGF signaling deficient microenvironment or, OnlineFirst 10/19/2010.
rencesambers AF, Groom AC, MacDonald IC. Dissemination andwth of cancer cells in metastatic sites. Nat Rev Cancer 2002;63–72.s LM, Hicklin DJ. VEGF-targeted therapy: mechanisms of anti-our activity. Nat Rev Cancer 2008;8:579–91.

ges S, Mazzone M, Hohensinner P, Carmeliet P. Silencing orling metastasis with VEGF inhibitors: antiangiogenesis revisited.ncer Cell 2009;15:167–70.os JM, Lee CR, Cruz-Munoz W, Bjarnason GA, Christensen JG,rbel RS. Accelerated metastasis after short-term treatment withotent inhibitor of tumor angiogenesis. Cancer Cell 2009;15:232–9.ez-Ribes M, Allen E, Hudock J, et al. Antiangiogenic therapy elicitslignant progression of tumors to increased local invasion and dis-t metastasis. Cancer Cell 2009;15:220–31.gsley LA, Fournier PG, Chirgwin JM, Guise TA. Molecular biologybone metastasis. Mol Cancer Ther 2007;6:2609–17.lmieri D, Chambers AF, Felding-Habermann B, Huang S, Steeg

tchelor TT. Angiogenesis in brain tumours. Nat Rev Neurosci07;8:610–22.tchelor TT, Sorensen AG, di Tomaso E, et al. AZD2171, a pan-VEGFeptor tyrosine kinase inhibitor, normalizes tumor vasculatured alleviates edema in glioblastoma patients. Cancer Cell 2007;:83–95.rayana A, Kelly P, Golfinos J, et al. Antiangiogenic therapy usingvacizumab in recurrent high-grade glioma: impact on local controld patient survival. J Neurosurg 2009;110:173–80.J, Pollock C, Tracy K, et al. Activation of the RalGEF/Ral pathwaymotes prostate cancer metastasis to bone. Mol Cell Biol 2007;27:38–50.anYin J, Tracy K, Zhang L, et al. Noninvasive imaging of the func-nal effects of anti-VEGF therapy on tumor cell extravasation andional blood volume in an experimental brain metastasis model.n Exp Metastasis 2009;26:403–14.vazza C, Carlo-Stella C, Giacomini A, et al. Human CD34+ cells
gineered to express membrane-bound tumor necrosis factor-ated apoptosis-inducing ligand target both tumor cells and tumorculature. Blood 2010;115:2231–40.
Cancer Research



14. Weten2 t200

15. BeBetum12

16. GetreCli

17. AldlialBio

18. Aabistion

19. Bethe

20. Ru

ofco

21. Kimgrome

22. Lecesio

23. Came

24. BidagCa

25. Airfun

26. Cabra


www.a

D


dge SR, Kendrew J, Hennequin LF, et al. AZD2171: a highly po-t, orally bioavailable, vascular endothelial growth factor receptor-yrosine kinase inhibitor for the treatment of cancer. Cancer Res5;65:4389–400.rgers G, Song S, Meyer-Morse N, Bergsland E, Hanahan D.nefits of targeting both pericytes and endothelial cells in theor vasculature with kinase inhibitors. J Clin Invest 2003;111:

87–95.rstner ER, Duda DG, di Tomaso E, et al. VEGF inhibitors in theatment of cerebral edema in patients with brain cancer. Nat Revn Oncol 2009;6:229–36.ridge SE, Lennard TW, Williams JR, Birch MA. Vascular endothe-growth factor receptors in osteoclast differentiation and function.chem Biophys Res Commun 2005;335:793–8.pro M, Abrahamsson PA, Body JJ, et al. Guidance on the use ofphosphonates in solid tumours: recommendations of an interna-al expert panel. Ann Oncol 2008;19:420–32.
rgers G, Hanahan D. Modes of resistance to anti-angiogenicrapy. Nat Rev Cancer 2008;8:592–603.benstein JL, Kim J, Ozawa T, et al. Anti-VEGF antibody treatment
27. Lacu11

acrjournals.org


glioblastoma prolongs survival but results in increased vascular-option. Neoplasia 2000;2:306–14.

LS, Huang S, Lu W, Lev DC, Price JE. Vascular endothelialwth factor expression promotes the growth of breast cancer braintastases in nude mice. Clin Exp Metastasis 2004;21:107–18.enders WP, Kusters B, Verrijp K, et al. Antiangiogenic therapy ofrebral melanoma metastases results in sustained tumor progres-n via vessel co-option. Clin Cancer Res 2004;10:6222–30.rbonell WS, Ansorge O, Sibson N, Muschel R. The vascular base-ntmembrane as “soil” in brainmetastasis. PLoSOne 2009;4:e5857.ard FC, Pierga JY, Vincent-Salomon A, PouponMF. A “class action”ainst themicroenvironment: do cancer cells cooperate inmetastasis?ncer Metastasis Rev 2008;27:5–10.d WC. Phenotypic heterogeneity of the endothelium: I. Structure,ction, and mechanisms. Circ Res 2007;100:158–73.labrese C, Poppleton H, Kocak M, et al. A perivascular niche forin tumor stem cells. Cancer Cell 2007;11:69–82.ne SW, Scadden DT, Gilliland DG. The leukemic stem cell niche:rrent concepts and therapeutic opportunities. Blood 2009;114:
50–7.



Published OnlineFirst October 19, 2010.Cancer Res Juan Juan Yin, Luhua Zhang, Jeeva Munasinghe, et al. Preclinical Model of Advanced Prostate CancerCediranib/AZD2171 Inhibits Bone and Brain Metastasis in a

Updated version

10.1158/0008-5472.CAN-10-1435doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2010/10/18/0008-5472.CAN-10-1435.DC1

Access the most recent supplemental material at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/early/2010/10/15/0008-5472.CAN-10-1435To request permission to re-use all or part of this article, use this link

Research. on September 16, 2018. © 2010 American Association for Cancercancerres.aacrjournals.org Downloaded from


http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-10-1435

http://cancerres.aacrjournals.org/content/suppl/2010/10/18/0008-5472.CAN-10-1435.DC1

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/early/2010/10/15/0008-5472.CAN-10-1435


research cediranib/azd2171 inhibits bone and brain...

Documents