abt-165,adualvariabledomainimmunoglobulin (dvd-ig ... · vegf and vegf receptor (vegfr) tyrosine...

Large Molecule Therapeutics

ABT-165, a Dual Variable Domain Immunoglobulin(DVD-Ig) Targeting DLL4 and VEGF,Demonstrates Superior Efficacy and FavorableSafety Profiles in Preclinical ModelsYingchun Li1, Jonathan A. Hickson1, Dominic J. Ambrosi2, Deanna L. Haasch1,Kelly D. Foster-Duke1, Lucia J. Eaton2, Enrico L. DiGiammarino3, Sanjay C. Panchal3,Fang Jiang1, Sarah R. Mudd4, Catherine Zhang5, Surekha S. Akella6,Wenqing Gao7,Sherry L. Ralston8, Louie Naumovski9, Jijie Gu2, and Susan E. Morgan-Lappe1

Abstract

Antiangiogenic therapy is a clinically validated modality incancer treatment. To date, all approved antiangiogenic drugsprimarily inhibit the VEGF pathway. Delta-like ligand 4 (DLL4)has been identified as a potential drug target in VEGF-independent angiogenesis and tumor-initiating cell (TIC)survival. A dual-specific biologic targeting both VEGF and DLL4could be an attractive strategy to improve the effectiveness of anti-VEGF therapy. ABT-165 was uniquely engineered using a propri-etary dual-variable domain immunoglobulin (DVD-Ig) technol-ogy based on its ability to bind and inhibit both DLL4 and VEGF.In vivo, ABT-165 induced significant tumor growth inhibitioncompared with either parental antibody treatment alone, due,in part, to the disruption of functional tumor vasculature. In

combination with chemotherapy agents, ABT-165 also inducedgreater antitumor response and outperformed anti-VEGF treat-ment. ABT-165 displayed nonlinear pharmacokinetic profiles incynomolgusmonkeys,with anapparent terminal half-life>5daysat a target saturation dose. In a GLP monkey toxicity study,ABT-165 was well-tolerated at doses up to 200 mg/kg withnon-adverse treatment–related histopathology findings limitedto the liver and thymus. In summary, ABT-165 represents a novelantiangiogenic strategy that potently inhibits both DLL4 andVEGF, demonstrating favorable in vivo efficacy, pharmacokinetic,and safety profiles in preclinical models. Given these preclinicalattributes, ABT-165 has progressed to a phase I study. Mol CancerTher; 17(5); 1039–50. �2018 AACR.

IntroductionAngiogenesis is one of the integral hallmarks of cancer (1).

Multiple pathways orchestrate angiogenesis, among which theVEGF and VEGF receptor (VEGFR) tyrosine kinase pathway havebeen heavily studied, leading to the development of the first-generation antiangiogenic drugs targeting these pathways (2).However, these therapies exhibit only modest overall survivalbenefit when combined with chemotherapy. In addition,

development of intrinsic and acquired VEGF-independent resis-tance mechanisms limits their effectiveness (2, 3), therebyhighlighting the need for targeting additional VEGF-independentpathways critical for tumor growth, angiogenesis, and survivalmaintenance.

After the FDA approval of the first anti-VEGF agent bevacizu-mab in 2004 (4), the Delta-like ligand 4 (DLL4)/Notch pathwaywas discovered as another pivotal signaling node in regulatingtumor angiogenesis (5, 6). The essential role of DLL4 in angio-genesis was initially revealed in knock-out studies in which micemissing only a single copy of theDll4 gene resulted in embryoniclethality due to major defects in vascular development (7–9).Notably,DLL4 andVEGFare theonly angiogenic factors known tohave this haplo-insufficiency phenotype, highlighting the criticalrequirement of both genes for vascular function (10, 11). DLL4 isupregulated on tumor vasculature relative to the endothelium ofadjacent normal tissues (12–14), and endothelial expressionlevels were found to be inversely correlated with survival ofbreast and ovarian cancer patients treated with anti-VEGF therapy(15, 16). To investigate the role of DLL4 on tumor growth, DLL4function was blocked in experimental tumor models with eitheranti-DLL4 neutralizing antibodies or a soluble form ofDLL4 protein that interferes with Notch receptor activation. DLL4blockade results in the inhibition of tumor growth acrossmultiplesolid tumor types, and the enhancement of antitumor activity byseveral anticancer agents (5, 6, 17, 18). Importantly, these studies

1Oncology Discovery, AbbVie Inc., North Chicago, Illinois. 2AbbVie BioresearchCenter, Worcester, Massachusetts. 3Global Protein Sciences, AbbVie Inc., NorthChicago, Illinois. 4Translational Imaging, AbbVie Inc., North Chicago, Illinois.5Drug Metabolism and Pharmacokinetics - Bioanalysis, AbbVie Biotherapeutics,Redwood City, California. 6Preclinical Safety, AbbVie Biotherapeutics, RedwoodCity, California. 7Drug Metabolism and Pharmacokinetics, AbbVie Inc., NorthChicago, Illinois. 8Preclinical Safety, AbbVie Inc., North Chicago, Illinois. 9Oncol-ogy Early Development, AbbVie Inc., Redwood City, California.

Note: Supplementary data for this article are available at Molecular CancerTherapeutics Online (http://mct.aacrjournals.org/).

Y. Li and J.A. Hickson contributed equally to this article.

Corresponding Author: Susan E. Morgan-Lappe, Oncology Discovery, AbbVieInc., Building AP10-1, 1 North Waukegan Road, North Chicago, IL 60064. Phone:847-937-6432; Fax: 847-935-5165; E-mail: [email protected]

doi: 10.1158/1535-7163.MCT-17-0800

�2018 American Association for Cancer Research.

MolecularCancerTherapeutics

www.aacrjournals.org 1039

on February 29, 2020. © 2018 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst March 28, 2018; DOI: 10.1158/1535-7163.MCT-17-0800

http://crossmark.crossref.org/dialog/?doi=10.1158/1535-7163.MCT-17-0800&domain=pdf&date_stamp=2018-4-11

http://mct.aacrjournals.org/

also showed that DLL4 inhibition can be effective in tumors thatare resistant to VEGF inhibition (5, 6). Of note, DLL4 blockadeleads to an apparent increase of tumor vascular density, which isthe opposite of VEGF inhibition. However, these tumor bloodvessels were found to be nonfunctional and therefore could notsupport tumor growth (5, 6). Notably, DLL4 is also implicated incontributing to the chemoresistance of tumor-initiating cells(TIC) or cancer stem cells (19). DLL4 mAb treatment delaystumor recurrence postchemotherapy in a variety of primarychemoresistant patient-derived xenograft (PDX)models that havean enriched TIC population (19, 20). In serial transplantationexperiments involving PDX mouse models, DLL4 mAbs reducedthe portion of cancer stem cells in colorectal, breast, and pancre-atic cancers (19–21). In addition, DLL4mAb treatmentwas foundto regulate gene expression involved in the epithelial-to-mesen-chymal transition (EMT), a phenotype closely linked to stem cellmaintenance and metastasis (20, 22). In light of these findings,DLL4 is an attractive therapeutic target and several anti-DLL4mAbs have advanced into clinical trials as potential new antican-cer therapeutics (23, 24).

However, inhibiting both DLL4 and VEGFmay provide amoreeffective and durable antitumor response relative to blockingeither axis alone given the targeting of key VEGF-independentand -dependent angiogenesis pathways, as well as TICs (25–28).An additional benefit for dual targeting includes a potentialmitigation of safety liabilities associated withDLL4mAb-inducedendothelial cell hyperproliferation that is VEGF-dependent (29).To support such a combination strategy, we have engineered apotent DLL4/VEGF bispecific, ABT-165, based on a proprietarydual-variable-domain immunoglobulin (DVD-Ig) technology(30). DVD-Ig is dual-specific, tetravalent immunoglobulin G(IgG)-like and has been used to successfully engineer bispecificmolecules with therapeutic efficacy in preclinical disease modelsfor rheumatoid arthritis (30) and age-related macular degenera-tion (31). ABT-165 is capable of simultaneous bivalent bindingand inhibition of DLL4 and VEGF, and exhibits favorable pre-clinical in vivo efficacy and safety profiles, thereby serving as anattractive novel anticancer agent.

Materials and MethodsReagents and cell lines

Recombinant human, cynomolgus monkey, and mouse DLL4extracellular domains (ECD), human and cynomolgus monkeyVEGF165 proteins, and biotinylated human DLL4 ECD and sulfo-tag–labeled human VEGF were produced by AbbVie. Recombi-nant mouse VEGF164, human Notch1-Fc and human VEGFR-2-Fcwere purchased from R&D Systems. Paclitaxel, irinotecan, andgemcitabine were purchased from Hospira. Temozolomide waspurchased from Merck.

The anti-DLL4 mAbs h1A11.1 (human IgG1/k L234A andL235A), ABT-487 (human IgG1/l L234A and L235A), andanti-VEGF mAb AB014 (human IgG1/k) were produced by Abb-Vie. To generate anti-DLL4/anti-VEGF DVD-Igs, the variabledomain sequences from the humanized anti-DLL4 and anti-VEGFmAbs were used to design the VH and VL domains of anti-DLL4/anti-VEGF DVD-Ig molecules, according to the DVD-Ig technol-ogy (30). ABT-165 was the lead DVD-Ig based on the in vitropotency, physicochemical properties, in vivo activity, and phar-macokinetic properties. The Fab and F(ab0)2 fragments ofABT-165 were generated by AbbVie.

Human DLL4/293HEK-stable cell line was generated by trans-fecting 293HEK with an expression vector encoding the full-length human DLL4 cDNA and maintained in DMEM supple-mented with 10% FBS and 800 mg/mL G418. Notch reporter cellline was generated by cotransducing EA.hy926 (32) with lentivi-rus containing the RBP-Jk firefly luciferase Notch reporter andRenilla luciferase control reporter (SABiosciences) and main-tained in DMEM supplemented with 10% FBS and 1 mg/mLpuromycin. Human umbilical vein endothelial cells (HUVEC)were from Lonza and maintained in EBM-2 basal medium sup-plemented with EGM-2 SingleQuots (Lonza). U87-MG andSW-48 cancer cell lines were purchased from ATCC. HT-29 waspurchased from NCI (Bethesda, MD). PA0123 was purchasedfrom The Jackson Laboratory. SUM149PT was purchased fromAsterand Bioscience. MDA-MB-231-luc-D3H2LN was purchasedfromCaliper Life Sciences. HT-29 cells were cultured inMcCoy 5amodified medium supplemented with 10% FBS. U87-MG,SW-48, SUM149PT, and PA0123 cells were cultured in DMEMsupplemented with 10% FBS. MDA-MB-231-luc-D3H2LN cellswere cultured inminimum essential medium supplemented with10% FBS, 2 mmol/L glutamine, 0.1 mmol/L nonessential aminoacids, and 1 mmol/L sodium pyruvate. All cells were maintainedat 37�C in 5%CO2 and 95% relative humidity. The cell lines wereexpanded upon receipt and then cryopreserved at an early passagenumber in liquid nitrogen. Cells were then cultured for fewer than3months after resuscitation and were expanded for nomore than5 passages when used for in vivo studies. Cell lines were charac-terized at their original sources and used without additionalauthentication, with the exception of U87-MG, SW-48, andHT-29 that were verified with additional short tandem repeatauthentication by AbbVie using the GenePrint 10 System fromPromega. The cell lines were routinely tested to confirm theabsence of mycoplasma by the MycoAlert Kit from Lonza.

Binding assay using surface plasmon resonance technologyThe simultaneous binding studies and binding kinetics mea-

surements of ABT-165 and parental mAbs against purifiedrecombinant DLL4 or VEGF proteins were analyzed by surfaceplasmon resonance–based measurements at 25�C using BiacoreT100/T200 instruments (GE Healthcare). An anti-Fc capturemethod was used for these binding studies. Data wereprocessed and fitted globally to a 1:1 bindingmodel using BiacoreT100/T200 Evaluation software to determine the binding kineticrate constants, ka (M�1s�1) and kd (s�1), and the equilibriumdissociation constant KD (M). Details can be found in the Sup-plementary Information.

Ligand and receptor binding competition ELISA assaysELISA-based immunoassays were used to measure DLL4 and

Notch1, and also VEGF and VEGFR2 binding competitions.Details can be found in the Supplementary Information.

Inhibition of cellular DLL4-mediated Notch activation incoculture bioassay

EA.hy926 Notch reporter cells resuspended in DMEM supple-mented with 10% FBS were preseeded in 96-well tissue cultureplates for 24 hours. Human DLL4-expressing 293HEK cells resus-pended in DMEMwith 10% FBS were mixed with a test antibodyor DVD-Ig in serial dilutions for 30 minutes. The mixtures werethen added to the preseeded EA.hy926Notch reporter cells. Plateswere incubated at 37�C with 5% CO2 for an additional 24 hours.

Li et al.

Mol Cancer Ther; 17(5) May 2018 Molecular Cancer Therapeutics1040




Firefly and Renilla luciferase assays were performed according tothe manufacturer's protocol (Dual-Glo Luciferase Assay System,Promega). Luminescence was read on a plate reader (Victor,PerkinElmer) and data generated were normalized to the Renillaluciferase luminescence signal.

Inhibition of VEGF-mediated HUVEC proliferationHUVECs were plated at 5,000 cells/well in 96-well plates in

maintenance medium. The following day, the medium wasreplaced with EBM-2 basal medium (Lonza) with 0.1% BSA. Testantibodies and DVD-Ig were serially diluted in EBM-2 with 0.1%BSA and added to the cells together with recombinant VEGF(50 ng/mL). The plates were incubated at 37�C with 5% CO2

for 72 hours. Viable cells were measured using CellTiter-Glo(Promega) according to the manufacturer's protocol.

Tumor xenograft model studiesFemale SCID mice (Charles River Laboratories) were inoculat-

ed with 1 � 106 cells subcutaneously in the right flank in a 50:50mixture of cells in growthmediumandMatrigel (BDBiosciences).Tumors were allowed to establish to an approximate volume of200mm3, as determined by calipermeasurement (V¼ L�W2/2).Animals were sizematched into treatment groups andwere dosedwith either biologic or cytotoxic agents as described in the figurelegends. All treatment groups consisted of 10 mice per group.Tumor growth inhibition (TGI) was calculated as the greatesttreatment response using the following equation: TGImax ¼ (1 �mean tumor volume of the treated group/mean tumor volume ofthe vehicle control group) � 100. Tumor growth delay wasdetermined as the percentage increase of the median time periodfor the treatment group to reach an arbitrary tumor volume of1,000mm3 relative to the vehicle control group. Tumor regressionis defined as tumors that dropped below 50% of their startingvolume for at least three consecutive measurements. Statisticalsignificance within the experiments for TGImax and TGD wereperformed by Student t test and Wilcoxon rank sum test, respec-tively. All animal studies were conducted in accordance with theguidelines andprotocols established and approvedby the internalInstitutional Animal Care and Use Committee of AbbVie.

Tumor vascular labeling and assessmentThree mice bearing U87-MG xenograft tumors from each treat-

ment group were used. Fluorescein isothiocyanate–labeled lyco-persicon esculentum lectin (100 mg of lectin/FITC, Vector Labo-ratories) was injected through tail vein and allowed to circulate for10–15 minutes prior to euthanasia. Following dissection, tumorswere immersed into liquid nitrogen, fixed with 4% paraformal-dehyde for 60minutes, and cut as 10-mm sections. Tumor sectionswere stained with CD31/PE (BD Biosciences) and covered withanti-fade mounting medium containing 40, 6-diamidino-2-phenylindole (DAPI). Images were capturedwith a Zeiss AxioPhot2 fluorescent microscope (Carl Zeiss) using AxioVision 4.6 soft-ware. Vessels were examined by dual labeling of lectin/FITC andCD31/PEandvessel densitywasquantifiedat200�magnification.For each tumor, 2–4 sections were examined and 4–6 highlyvascularized areas (0.4 mm2/each) from each section were ran-domly selected for image acquisition and analysis.

Imaging studiesFor bioluminescence imaging, luciferin was dissolved in PBS

and administered intraperitoneally at 150 mg/kg. Animals were

then anesthetized (2% isoflurane) and images acquired usingan IVIS Spectrum. Luciferase activity was measured as photosper second for designated regions-of-interest using Living Imaging4.4 software.

For a dynamic contrast enhanced–MRI (DCE-MRI), animalswere imaged at baseline and at days 2, 7, and 14 posttreatment.Group sizes ranged from 6 to 7 animals, depending on the timepoint. Animals were given 0.2 mmol/kg Magnevist (Bayer) as anintravenous bolus dose after beginning the MRI sequence. Imag-ing was performed on a 4.7T Bruker Pharmascan (Billerica) usingDCE_FLASH sequence (33). An image-derived vascular inputfunction was obtained using the transverse sinus based on themethods as described in ref. 34. Ktrans values were determined onthe basis of vascular and tissue concentrations of the contrastagent (35).

Pharmacokinetic studiesThe pharmacokinetics of ABT-165 was evaluated either in

single-dose pharmacokinetic studies or as part of the repeat-dosetoxicologic studies. The concentrations of ABT-165 in cynomol-gus monkey serum samples were determined using an electro-chemiluminescent (ECL) immunoassay on MSD platform (MesoScale Discovery). Briefly, biotinylated recombinant human DLL4protein used as the capture reagent was incubated for 1 hour onblocked streptavidin-coated MSD plates (Meso Scale Discovery).The assay was carried out in 1% final serum concentration anddetection was achieved with sulfo-tag–labeled VEGF. Standardcurve fitting and data evaluation was performed using XLfit4software. Pharmacokinetic parameters for each animal were cal-culated using WinNonlin software Version 5.3 (Pharsight Cor-poration) by noncompartmental analysis with the NCA Model201 and linear trapezoidal method.

Toxicology and safety pharmacology studiesToxicity of ABT-165 was evaluated in a non-Good Labora-

tory Practice (GLP) dose tolerability study and a GLP 1-monthtoxicity study. In the non-GLP study (once weekly dosing for3 weeks; 4 total doses), ABT-165 was administered via intra-venous bolus injection to two female cynomolgus monkeys/group at doses of 10, 30, or 100 mg/kg. Standard toxicologyendpoints including histopathology were evaluated. In theGLP 1-month toxicity study, ABT-165 was administered tocynomolgus monkeys (0 and 200 mg/kg/dose: 6 animals/sex/group; 10 and 50 mg/kg/dose: 4 animals/sex/group) viaintravenous infusion (30 minutes) once weekly for 4 weeks(a total of 5 doses), followed by a 12-week recovery period(0 and 200 mg/kg/dose groups only; 2 animals/sex/group).The GLP study included detailed assessments of clinical signs,body weights, food consumption, ophthalmology, electrocar-diography, ABT-165 serum concentrations and toxicokinetics,ABT-165 cerebrospinal fluid concentrations, clinical pathology(hematology, coagulation, clinical chemistry, and urinalysis),and anatomic pathology (organ weights, gross necropsy eva-luations, and histopathology). Antidrug antibodies (ADA)were not evaluated in the GLP study as there was no evidenceof a decrease in exposure suggesting the presence of ADA. In astand-alone GLP safety pharmacology study, ABT-165 wasadministered intravenously to conscious cynomolgus monkeysat single doses of 10, 50, and 200 mg/kg. Animals weremonitored for cardiovascular function (heart rate, blood pres-sure, and electrocardiograms).

Dual Inhibition of DLL4 and VEGF By a DVD-Ig

www.aacrjournals.org Mol Cancer Ther; 17(5) May 2018 1041




ResultsIdentification of ABT-165, a DLL4/VEGF dual variable domainimmunoglobulin

Using proprietary dual variable domain (DVD-Ig) technology,ABT-165 was identified by the testing of multiple candidateDVD-Ig molecules in a series of physicochemical, pharmacoki-netic, and functional potency evaluations. On the basis of pre-ferred expression, solubility and stability properties, and in vitroand in vivo potencies coupled with mouse exposure, ABT-165emerged as the lead candidate. ABT-165 contains the variabledomain of a humanized anti-DLL4 mAb (h1A11.1) at the outerposition of the DVD-Ig molecule, and the variable domain of ahumanized anti-VEGFmAb (AB014) at the inner position, joinedby a short heavy-chain linker (ASTKGP) and a long light-chainlinker (TVAAPSVFIFPP) (Fig. 1A). To reduce potential effectorfunction-induced toxicity within normal tissue that express DLL4or VEGF, the Fc domain was engineered to contain two pointmutations at the lower hinge region (L234A, L235A) to signifi-cantly attenuate binding to Fcg I and Fcg II receptors and com-plement component C1q (36).

The simultaneous dual-binding properties of ABT-165 wereconfirmed using surface plasmon resonance technology as dem-onstratedby the ability of ABT-165 tobeboundby saturated levelsof both VEGF and DLL4 antigens (Fig. 1B). These results dem-onstrate that ABT-165 can bind simultaneously to both targets,

and binding of one target does not hinder binding to the other. Assummarized in Supplementary Table S1, ABT-165 bindsto human DLL4 extracellular domain (ECD) with a KD of8.7 nmol/L, and to human VEGF protein with a KD of0.9 nmol/L, which are within comparable ranges relative to theparentalmAbs.With regards to cross-species binding interactions,ABT-165 cross-reacts with mouse DLL4, but not mouse VEGF.However, ABT-165 binds to both cynomolgus monkey DLL4 andVEGF at a similar binding affinity to that of the human counter-parts, which enables pharmacokinetic and safety assessments.

Dual target inhibitory activities of ABT-165The functional potency of ABT-165 was evaluated in ligand-

receptor competition ELISA and cell-based assays for both DLL4and VEGF (Fig. 2A and B; Supplementary Table S2). ABT-165blocked the interaction of DLL4 with Notch1, and VEGF withVEGFR-2, with IC50 values of 0.37 nmol/L and 5.9 nmol/L,respectively, comparable with the parental mAbs. This blockadeof ligand–receptor binding translates into potent inhibition ofDLL4-mediated Notch-1 signaling (IC50 ¼ 12 nmol/L) similar tothe parental anti-DLL4 mAb (IC50 ¼ 9.0 nmol/L), as measuredusing DLL4-expressing cells to activate a Notch1-expressing cellline engineered to express a luciferase reporter gene driven by aNotch-responsive element. In addition, ABT-165 inhibited VEGF-stimulated HUVEC proliferation (IC50 ¼ 3.6 nmol/L), compara-ble with anti-VEGF mAb activity (IC50 ¼ 1.6 nmol/L). These data

A

DLL4/VEGF DVD-Ig, ABT-165DLL4 mAb VEGF mAb

CHVH2VH1

CLVL2VL1

Linkers

5’

5’

3’

3’

B VEGF Binding DLL4 Binding

DLL4 mAb

ABT-165VEGF mAb

–200

0

20

40

Res

pons

e (R

U)

Time (s)

60

80

0 200 400 600 800 1,000 1,200 1,400

Figure 1.

Generation of ABT-165 with dualbinding activity to DLL4 and VEGF.A, Schematic diagram of ABT-165DVD-Ig with an anti-DLL4 and ananti-VEGF variable domain linked intandem in heavy and light chain.B, ABT-165, DLL4 mAb, and VEGF mAbwere captured on biosensor chips andtested for binding to recombinantVEGF. After the binding of VEGFbecame saturated, recombinant DLL4protein was injected resulting in asecond binding signal for ABT-165.

Li et al.





demonstrate that ABT-165 retains in vitro binding and functionalpotencies against both targets.

In the presence of recombinant VEGF (�1.2 nmol/L),the anti-DLL4 potency of ABT-165 is enhanced by greater than10-fold as measured in the DLL4-mediated Notch activation

bioassay (Fig. 2C; Supplementary Table S3). Notably, thisincreased potency is unique to the DLL4/VEGF DVD-Ig as theparental mAb combinations induced similar inhibition ofNotch activation whether in the presence or absence of recom-binant VEGF (Fig. 2C).

Inhibition of VEGF/VEGFR2 bindingInhibition of DLL4/Notch1 binding

Inhibition of HUVEC proliferationInhibition of Notch activation

A

B

C VEGFABT-165 + VEGFDLL4 mAb + VEGF mAb +

Concentration (nmol/L)

Abs.

450

0.01 0.1 1 10 10

01,0

000.00

0.25

0.50

0.75

1.00

1.25

0.1 1 10 100

0.00

0.25

0.50

0.75

1.00

1.25

1.50


Abs.

450


Fold

not

ch a

ctiv

atio

n

0.001 0.0

1 0.1 1 10 100

1,000

0

50

100

150

200

250

300

350

0.1 1 10 1000

20

40

60

80

100

120


% V

iabl

e ce

lls


Fold

not

ch a

ctiv

atio

n

0.01 0.1 1 10 10

01,0

000

50

100

150

200

250

300

350

-VEGF+VEGFNegative control


Fold

not

ch a

ctiv

atio

n

0.01 0.1 1 10 10

01,0

000

50

100

150

200

250

300

350

-VEGF+VEGFNegative control

Figure 2.

ABT-165 DVD-Ig inhibits the function of DLL4 and VEGF. A, ABT-165 blocks the interaction between human DLL4 and Notch1 and the interaction betweenhuman VEGF and VEGFR2 in competition ELISAs. B, ABT-165 potently inhibits DLL4 cell–mediated Notch activation in a coculture bioassay and inhibitsVEGF-stimulated HUVEC proliferation. C, VEGF potentiates the activity of ABT-165. The activities of ABT-165 and the combination of DLL4 and VEGF mAbmixture were analyzed in the Notch activation bioassay in the presence or absence of recombinant VEGF (6 nmol/L). A nonrelevant recombinant protein, BSG2,was used as the negative control.






Tumor growth inhibition and angiogenesis regulation byABT-165

To investigate antitumor efficacy of ABT-165 in vivo, we focusedon human cancer xenograft models dependent on human VEGFas ABT-165 binds human but not mouse VEGF. To this end,U87-MG human glioblastoma and SW-48 human colon cancerxenograft models were treated with ABT-165, negative control,anti-VEGF or anti-DLL4 mAbs, or the combination of the VEGFandDLL4mAbs at an equivalentmolar dose. As a result, ABT-165treatment significantly increased tumor growth inhibition andprolonged tumor growth delay compared with either anti-DLL4mAb or anti-VEGF mAb monotherapy (Fig. 3A and B; Supple-mentary Table S4). In each study, the antitumor activity ofABT-165 was equivalent to the combination of the two mAbs.

Consistent with the expected mechanism of action, histologicexaminationwithin the viable regions of glioblastoma (U87-MG)tumors following a single dose of ABT-165 revealed significantchanges in tumor vasculature at day 7 (Fig. 3C). Specifically,ABT-165 treatment resulted in a decrease in overall tumor vesselperfusion similar to anti-DLL4 mAb as indicated by reduction inan intravascular tracer lectin/FITC. The staining of the endothelialcell marker (CD31/PE) revealed that ABT-165 did not inducesignificant change in tumor vascular density when comparedwithanti-VEGF treatment (lower vascular density compared with thenegative control treatment). Nonfunctional vasculature as a resultof ABT-165 treatment is highlighted by the lack of lectin stainingin CD31-positive tumor vessels. Changes in tumor vasculaturefunction by ABT-165 in the U87-MG model were further con-firmed by DCE-MRI, which is a quantitative imaging techniqueused to assess changes directly at the tumor. A decrease in Ktrans,a measure of perfusion and permeability, was observed two daysafter a single dose of 6.7 mg/kg ABT-165 (Fig. 3D). Thesemechanistic data demonstrate early changes in tumor vascularfunction upon ABT-165 treatment that correlate with tumorgrowth inhibition.

ABT-165 improves efficacy of chemotherapy, and outperformsanti-VEGF treatment in combinationwith chemotherapy agents

We further investigated whether ABT-165 could be broadlycompatible with standard-of-care (SoC) chemotherapy agentswith diverse mechanisms of action to drive more durable anti-tumor responses. ABT-165 in combination with temozolomide(SoC for gliomas) induced significant prolonged tumor stasis andincreased tumor growth delay in the U87-MG glioblastomaxenograft model compared with either monotherapy alone(Fig. 4A; Supplementary Table S4). Similarly, the combinationof ABT-165 with irinotecan (SoC for colorectal cancer) inducedsuperior efficacy compared with irinotecan alone in the HT-29colon cancer xenograft model (Fig. 4B; Supplementary Table S4).Using a PDXmodel of pancreatic origin, PA0123, the addition ofABT-165 to gemcitabine (SoC for pancreatic cancer) significantlyimproved the tumor growth inhibition compared with gemcita-bine alone (Fig. 4C; Supplementary Table S4). Furthermore, usingan orthotopic breast cancer model MDA-MB-231-luc, the addi-tion of ABT-165 with paclitaxel (SoC for breast cancer) inducedtumor regressions during treatment with significantly improvedtumor growth inhibition and tumor growth delay compared withpaclitaxel alone (Fig. 4D; Supplementary Table S4). Of note,bioluminescence imaging revealed that the combination ofABT-165 with paclitaxel also significantly reduced spontaneousmetastasis to both the lung and lymph nodes (Fig. 4E).

The antitumor efficacy of ABT-165with SoC chemotherapywasalso compared with anti-VEGF mAb plus chemotherapy. In theSUM149PT triple-negative breast cancer xenograft model, addi-tion of paclitaxel to either ABT-165 or anti-VEGF mAb signifi-cantly increased the antitumor efficacy compared with eithermonotherapy (Fig. 5A; Supplementary Table S4). The combina-tion of ABT-165 and paclitaxel demonstrated a 70% tumorregression rate compared with 0% for all other treatments.Importantly, compared with VEGF mAb plus paclitaxel, ABT-165plus paclitaxel induced significant improvement in both tumorgrowth inhibition and tumor growth delay. In the SW-48 coloncancer xenograft model, addition of irinotecan to ABT-165 sig-nificantly increased the antitumor efficacy compared with eithermonotherapy (Fig. 5B; Supplementary Table S4). However, addi-tion of anti-VEGFmAb to irinotecan did not significantly improveefficacy beyond monotherapy. Furthermore, compared withVEGF mAb plus irinotecan, ABT-165 in combination with irino-tecan significantly improved both tumor growth inhibition andtumor growth delay.

ABT-165 exhibits acceptable pharmacokinetic profiles incynomolgus monkeys

Pharmacokinetic properties of ABT-165 were assessed incynomolgus monkeys following single or weekly intravenousdosing regimens (Table 1; Supplementary Fig. S1). ABT-165exhibited nonlinear pharmacokinetic, indicating contributionsof target-mediated disposition (TMD) to total serum clearance.Nonlinear elimination is likely due to saturation of binding tocell membrane DLL4 followed by internalization and degra-dation of ABT-165. At higher doses (�10 mg/kg), the pharma-cokinetic profiles in cynomolgus monkeys approached linearrange and were similar to those of typical mAbs, with lowserum clearance (�5.1 mL/day/kg) and small volumes ofdistribution (Vss< 40 mL/kg). The apparent terminal half-lifewas >5 days. Notably, upon weekly dosing in cynomolgusmonkeys at 10 mg/kg, the steady-state Ctrough levels weremaintained at >140 mg/mL (week 3) and the overall pharma-cokinetic profile approached linear range (SupplementaryFig. S1). There was no loss of exposure following repeateddosing in the monkey toxicology study, which indicates theabsence of antidrug antibody (ADA) response.

ABT-165 is well tolerated in cynomolgus monkeysThe comparable binding and potency of ABT-165 against both

human and cynomolgus monkey DLL4 and VEGF supported theselection of the cynomolgus monkey species for toxicology andsafety pharmacology assessments.

ABT-165-was well tolerated in the repeat-dose toxicity studiesat doses up to 200 mg/kg/dose. Histopathologic findings in theGLP study included non-adverse effects in the liver and thymus at�10 mg/kg/dose (liver: sinusoidal dilation, centrilobular atro-phy, and periportal hepatocellular hypertrophy; thymus: revers-ible decrease in cortical lymphocytes). These findings were con-sidered nonadverse due to the nature and generally low magni-tude of these changes, and the absence of significant associatedalterations in the clinicopathologic data. Of note, these findingsdiffer from a previous GLP toxicity study in cynomolgusmonkeysusing an anti-DLL4 mAb (ABT-487; Supplementary Table S5),which was shown to induce increased interstitial cellularity of theheart that was potentially considered to be related to the mech-anism of action targeting endothelial cells.

Li et al.





Figure 3.

ABT-165 inhibits tumor growth and decreases functional tumor angiogenesis. A and B, In vivo efficacy was evaluated in glioblastoma (U87-MG) and coloncancer (SW-48) xenografts. A 5 mg/kg dose of DLL4 mAb or VEGF mAb, alone or in combination, or an equimolar 6.7 mg/kg dose of ABT-165 wasadministered intraperitoneally weekly for four weeks (black arrows). Tumor growth was measured by calipers, and mean tumor volumes with SEs arepresented (n ¼ 10 per group). C, U87-MG xenograft tumors were sectioned seven days postintraperitoneal administration of a single 5 mg/kg dose of mAbs or6.7 mg/kg dose of ABT-165. Within viable regions of treated tumors, lectin perfusion (FITC, green) or anti-CD31 (PE, red) staining was performed.Colocalization for both lectin/FITC and CD31/PE labeling is depicted in orange (Merge panel). Nuclei were stained with DAPI (blue). Percent positive functionalvessels (lectin staining), and vascular density quantitation (CD31 staining) are graphed below (� , P values <0.05 compared with control group). D, U87-MGxenograft tumors were imaged by DCE-MRI after the animals were dosed with a single intraperitoneal dose of mAbs (5 mg/kg) or ABT-165 (6.7 mg/kg)(�� , P values <0.01 compared with vehicle control group).






In the safety pharmacology study in conscious cynomolgusmonkeys, ABT-165 had no cardiovascular effects at a dose of10 mg/kg. At 50 and 200 mg/kg, ABT-165 produced an increasein mean arterial pressure (�20 mm Hg) that is consistent withantiangiogenic activity, and a decrease in heart rate (�15 bpm).

DiscussionSolid tumor growth depends on the development of neovas-

culature and therefore, an effective antiangiogenesis strategy may

offer significant antitumor benefits. Although the concept of"antiangiogenesis" as a form of cancer therapy was first put forthby Judah Folkman in 1971 (37), nearly three decades passedbefore the development of the first antiangiogenic drug, bevaci-zumab, followed by approvals of several other drugs mainlytargeting the same pathway (2, 4). As angiogenesis is a highlycomplex process that involves several pathways and factors,maximizing the potential of antiangiogenesis for cancer therapyis likely to require amulti-pronged approach.Dual targetingof theDLL4/Notch and VEGF/VEGFR pathways may represent one such

Figure 4.

Combination of ABT-165 with standard-of-care chemotherapy agents improves antitumor efficacy. A–D, ABT-165 was administered intraperitoneally weeklyat 6.7 mg/kg (black arrows). A single 5 mg/kg dose of temozolomide (TMZ) was administered intraperitoneally, irinotecan was given intraperitoneally at 60 mg/kgevery 3 days for 4 doses, gemcitabine was given intraperitoneally at 100 mg/kg every 3 days for 4 doses, or paclitaxel was given intraperitoneally at 25 mg/kgevery four days for 3 doses, respectively (pink arrows). Tumor growth was measured by calipers, and mean tumor volumes with SEs presented (n¼ 10 per group).E, In the orthotopic MDA-MAB-231-luciferase model, spontaneous metastasis to the lung and lymph nodes were measured noninvasively by bioluminescentimaging (n ¼ 10 per group).

Li et al.





opportunity as DLL4 also plays a distinct pivotal role in tumorangiogenesis through both VEGF-dependent and -independentpathways, as well as in cancer stem cell maintenance (25, 38).

A bispecific approach represents an attractive combinationstrategy to efficiently inhibit both targets, and currently two otherDLL4/VEGF bispecific biologics in addition toABT-165 have beenrecently reported, HD105 (39) and OMP-305B83 (40). HD105was engineered by fusing a DLL4-targeting single-chain variablefragment (scFv) to the C-terminal of an anti-VEGF mAb. The invivo efficacy ofHD105 could not be directly determined asHD105lacks mouse DLL4 cross-reactivity. Rather, a surrogate bispecificwithmouseDLL4bindingwas used toperform in vivo evaluations,and the pharmacokinetics and safety profiles of HD105 remainunknown. OMP-305B83 consists of a common light chain andtwo different heavy chains with point mutations in the CH3domain to drive heterodimer formation. This design makesOMP-305B83 monovalent for binding to both targets (i.e.,anti-DLL4 activity in one arm and anti-VEGF activity in the otherarm) (40).

Here, we describe the development of ABT-165, an IgG-likeDLL4/VEGF bispecific molecule using a proprietary DVD-Ig plat-form. ABT-165 carries two anti-DLL4 variable domains and twoanti-VEGF variable domains with bivalent binding to both targets(a tetravalent bispecific). Notably, we have found that anti-DLL4bivalency is more potent than monovalency in blocking thefunction of DLL4 in a cellular functional assay in vitro (Supple-mentary Fig. S2). Also, ABT-165 cross-reacts with mouse DLL4,which enables the direct assessment of its efficacy in a range ofhuman tumor xenograft murine models. In vivo, ABT-165 signif-icantly augmented antitumor effects compared with either anti-DLL4 or anti-VEGF mAb alone. Depending on the xenografttumor model and dosing schedule, we observed tumor growthinhibition ranging from approximately 50%–80% by ABT-165monotherapy (Supplementary Table S4), acknowledging the lackofmouseVEGF cross-reactivitymayunderestimate the full activityof DLL4 and VEGF dual inhibition in these models. ABT-165 canalso be combinedwith chemotherapeutics of diversemechanismsof action to enhance antitumor response, with activity that is also

Figure 5.

ABT-165 is superior to VEGF mAb in combination with chemotherapy. Tumor growth inhibition by ABT-165 in combination with paclitaxel (A) or irinotecan (B)was assessed versus VEGF mAb plus chemotherapy. In the SUM149PT model, a 5 mg/kg dose of VEGF mAb or an equimolar 6.7 mg/kg dose of ABT-165 wasadministered intraperitoneally weekly (black arrows), while paclitaxel was administered intraperitoneally at 25 mg/kg every four days for 12 doses (pink arrows).In SW-48, a 10 mg/kg dose of VEGF mAb or an equimolar 13.3 mg/kg dose of ABT-165 was administered intraperitoneally weekly (black arrows), whileirinotecan was given intraperitoneally at 60 mg/kg every 3 days for 4 doses (pink arrows). Tumor growth was measured by calipers, and mean tumorvolumes with SEs presented (n ¼ 10 per group).

Table 1. Mean pharmacokinetic parameters of ABT-165 after single IV bolus administration in monkeys (n ¼ 2 per dose group)

Dose (mg/kg) AUC (mg*hr/mL) Cmax (mg/mL) Vss (mL/kg) CL (mL/day/kg) t1/2 (days) MRT (days)

1 2.6 30.3 35.9 10.5 2.8 3.93 9.7 92.8 33.9 7.5 4.2 4.610 47.0 395.1 35.9 5.1 4.8 7.130 207.0 1344.4 27.0 3.7 5.1 7.8

NOTE: Units: Dose (mg/kg); Cmax (denotes the first observed concentrations post IV dosing, mg/mL); Vss (volume of distribution at steady state, mL/kg); t1/2 (half life,days; harmonic mean); AUC (area under the curve from 0 to infinity, mg*hr/mL); CL (serum clearance, mL/day/kg); MRT (mean residence time, days).






superior to that of anti-VEGF plus chemotherapy. Vascular per-fusion and DCE-MRI imaging studies confirmed that ABT-165effectively reduced functional tumor angiogenesis, consistentwith its mechanism of action. As DLL4 has also been implicatedinmaintaining tumor initiating cell growth and survival (19–21),evaluating ABT-165 for its ability to inhibit cancer stem cellfunction is currently under investigation. Recent advances havealso suggested that angiogenesis and immunity may have com-plementary mechanisms at the level of the tumor microenviron-ment (41), and indeed ongoing clinical trials combining anti-angiogenic agents with immune checkpoint inhibitorsmay revealsome of these potential synergies (42, 43).

Interestingly, in a DLL4 cellular assay that measures the activityof cell-bound DLL4 to activate the endogenous Notch pathway,ABT-165 exhibited enhanced anti-DLL4 activity in the presence ofVEGF, a phenomenon not observed with the parental mAbcombinations. The underlying molecular mechanism mayinvolve VEGF dimer-induced cross-linking of the DVD-Ig, whichcould in part contribute to enhanced DLL4 protein downregula-tion. To support this hypothesis, we have observed in vitro thatVEGF can promote, in a dose-dependent manner, the activity ofABT-165 to downregulate DLL4 protein (Supplementary Fig. S3).It has been reported that VEGF levels are elevated at local tumorsites than the peripheral blood (44). However, it remains to bedetermined whether the VEGF-enhanced inhibition of DLL4 byABT-165 will contribute to more robust antitumor activity in vivowhere higher concentrations of localized VEGF are present at thetumor site. As ABT-165 only neutralizes human VEGF, but notmouse VEGF, an equivalent mouse VEGF/DLL4 cross-reactivesurrogate DVD-Ig will be required to further test this hypothesisin VEGF-high xenograft tumor models by inhibiting VEGFexpressed from both human tumors and mouse stroma.

The cross-reactivity of ABT-165 to cynomolgus monkey DLL4and VEGF allowed cynomolgus monkey to be a relevant toxicol-ogy species to examine the safety profile. We found that ABT-165waswell tolerated inmonkeys even at doses as high as 200mg/kg,with nonadverse histopathologic findings noted only in the liverand thymus. Of note, DLL4 inhibition alone is known to induceendothelial cell proliferation that could be VEGF-dependent(6, 39). Therefore, we hypothesized that dual DLL4 and VEGFinhibitionmaymitigate some normal tissue toxicity derived fromDLL4 inhibition alone by potentially restoring the balance ofDLL4 and VEGF pathways thereby, preventing proliferation ofendothelial cells. In support of our hypothesis, we found thatABT-165 did not induce increased interstitial cellularity in cyno-molgus monkey heart, a histopathologic finding that wasobserved when monkeys were dosed with an anti-DLL4 mAb(ABT-487) that has comparable in vitro and in vivo anti-DLL4potency. Recently, a VEGF trap was also reported to reversethe vascular changes in mouse heart and liver caused by anti-DLL4 mAb treatment (29) and interestingly, the monovalentDLL4/VEGF bispecific OMP-305B83 also reported a potentialimprovement of heart safety profile in cynomolgusmonkeys (40).

VEGF mAb administration is known to induce the accumula-tion of total plasma VEGF in patients (45), and we saw a similareffect on both total (ABT-165-bound and free) VEGF and solubleDLL4 by ABT-165 in cynomolgus monkeys (SupplementaryFig. S4). The increase of both VEGF and soluble DLL4 uponABT-165 administration is likely a result of their binding toABT-165 in circulation, which lends value as a surrogate markerfor target binding by ABT-165 in vivo.

In summary, ABT-165 is a novel DVD-Ig that potently inhibitsboth DLL4 and VEGF, demonstrating favorable in vivo efficacy,pharmacokinetic, and safety profiles in preclinical models. If anacceptable therapeutic window canbe achieved in cancer patients,ABT-165 may be an attractive next-generation antiangiogenicagent armed with multiple mechanisms including targeting can-cer stem cells. The preclinical attributes of ABT-165 in providingsuperiority to anti-VEGF or anti-DLL4 treatments provide scien-tific rationale for investigating its potential clinical utility for thetreatment of solid tumors. Indeed, ABT-165 is currently beingevaluated in a phase I clinical trial (NCT01946074) in cancerpatients, and to date has induced partial responses in severalheavily pre-treated ovarian cancer patients (46). Mechanistic-based toxicities expected for antiangiogenic agents, such as sys-temic hypertension, pulmonary hypertension, and gastrointesti-nal perforation have been noted in the phase I study. Furtherinvestigation in a controlled randomized clinical trial is needed toestablish the efficacy and safety of ABT-165 compared with ananti-VEGF agent.

Disclosure of Potential Conflicts of InterestY. Li has ownership interest (including patents) in AbbVie. J.A. Hickson has

ownership interest (including patents) in Abbvie. D.L. Haasch has ownershipinterest (including patents) in stock. F. Jiang is a senior scientist at AbbVie.L. Naumovski has ownership interest (including patents) in AbbVie. S.EMorgan-Lappe has ownership interest (including patents) in AbbVie. Nopotential conflicts of interest were disclosed by the other authors.

Authors' ContributionsConception and design: Y. Li, J.A. Hickson, S.R. Mudd, L. Naumovski, J. Gu,S.E. Morgan-LappeDevelopment of methodology: Y. Li, J.A. Hickson, D.J. Ambrosi, L.J. Eaton,E.L. Digiammarino, F. Jiang, S.R. Mudd, C. ZhangAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J.A. Hickson, D.J. Ambrosi, D.L. Haasch, L.J. Eaton,E.L. Digiammarino, S.C. Panchal, S.R. Mudd, S.S. AkellaAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): Y. Li, J.A. Hickson, D.J. Ambrosi, D.L. Haasch,E.L. Digiammarino, S.C. Panchal, F. Jiang, S.R. Mudd, S.S. Akella, W. Gao,L. Naumovski, J. Gu, S.E. Morgan-LappeWriting, review, and/or revision of the manuscript: Y. Li, J.A. Hickson,D.J. Ambrosi, D.L. Haasch, E.L. Digiammarino, F. Jiang, S.R. Mudd, S.S. Akella,W. Gao, S.L. Ralston, L. Naumovski, J. Gu, S.E. Morgan-LappeAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): J.A. Hickson, W. GaoStudy supervision: Y. Li, J.A. Hickson, S.S. Akella, J. Gu, S.E. Morgan-LappeOther (provided preclinical in vivo efficacy data): K.D. Foster-Duke

AcknowledgmentsThe authors thank Wen Zhen Gu, Evelyn McKeegan, and Suhasini Iyer for

their dedicated support for this program. We also thank Karl Walter, JohnHarlan, Eric Hebert, Marc Lake, and Ramesh Iyer for generating critical reagentsrequired for this work; Lorenzo Benatuil, Chung-Ming Hsieh, and CristinaHarris for contributions to antibody generation and DVD-Ig engineering; LoraTucker for the HUVEC assay; Scott Mittelstadt and James Lynch for the safetypharmacology study; Joann Palma, Jerry Clarin, and Sally Schlessinger for theirsupport in biomarker studies; Martin Voorbach for help with the DCE-MRI;Thomas McGonigal and Erwin Boghaert for their support in in vivo pharma-cology; and Monica Motwani and Daniel Afar for their support in the ongoingABT-165 clinical trial.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received August 21, 2017; revised December 28, 2017; accepted March 8,2018; published first March 28, 2018.

Li et al.





References1. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell

2011;144:646–74.2. Jayson GC, Kerbel R, Ellis LM, Harris AL. Antiangiogenic therapy in

oncology: current status and future directions. Lancet 2016;388:518–29.3. Kerbel RS. Tumor angiogenesis. N Engl J Med 2008;358:2039–49.4. Ferrara N, Hillan KJ, Gerber HP, Novotny W. Discovery and development

of bevacizumab, an anti-VEGF antibody for treating cancer. Nat Rev DrugDiscov 2004;3:391–400.

5. Noguera-Troise I,Daly C, PapadopoulosNJ, Coetzee S, BolandP,GaleNW,et al. Blockade of Dll4 inhibits tumour growth by promoting non-pro-ductive angiogenesis. Nature 2006;444:1032–7.

6. Ridgway J, Zhang G, Wu Y, Stawicki S, Liang WC, Chanthery Y, et al.Inhibition of Dll4 signalling inhibits tumour growth by deregulatingangiogenesis. Nature 2006;444:1083–7.

7. Gale NW, Dominguez MG, Noguera I, Pan L, Hughes V, Valenzuela DM,et al. Haploinsufficiency of delta-like 4 ligand results in embryonic lethalitydue to major defects in arterial and vascular development. Proc Natl AcadSci U S A 2004;101:15949–54.

8. Krebs LT, Shutter JR, Tanigaki K, Honjo T, Stark KL, Gridley T. Haploin-sufficient lethality and formation of arteriovenousmalformations inNotchpathway mutants. Genes Dev 2004;18:2469–73.

9. Duarte A, Hirashima M, Benedito R, Trindade A, Diniz P, Bekman E, et al.Dosage-sensitive requirement for mouse Dll4 in artery development.Genes Dev 2004;18:2474–8.

10. Carmeliet P, Ferreira V, Breier G, Pollefeyt S, Kieckens L, Gertsenstein M,et al. Abnormal blood vessel development and lethality in embryos lackinga single VEGF allele. Nature 1996;380:435–9.

11. Ferrara N, Carver-Moore K, Chen H, Dowd M, Lu L, O'Shea KS, et al.Heterozygous embryonic lethality induced by targeted inactivation of theVEGF gene. Nature 1996;380:439–42.

12. Jubb AM, Browning L, Campo L, Turley H, Steers G, Thurston G, et al.Expression of vascular Notch ligands Delta-like 4 and Jagged-1 in glio-blastoma. Histopathology 2012;60:740–7.

13. JubbAM, Soilleux EJ, TurleyH, Steers G, Parker A, Low I, et al. Expression ofvascular notch ligand delta-like 4 and inflammatory markers in breastcancer. Am J Pathol 2010;176:2019–28.

14. Jubb AM, Turley H, Moeller HC, Steers G, Han C, Li JL, et al. Expression ofdelta-like ligand 4 (Dll4) and markers of hypoxia in colon cancer. Br JCancer 2009;101:1749–57.

15. HuW, LuC,DongHH,Huang J, ShenDY, StoneRL, et al. Biological roles ofthe Delta family Notch ligand Dll4 in tumor and endothelial cells inovarian cancer. Cancer Res 2011;71:6030–9.

16. JubbAM,Miller KD, RugoHS,Harris AL, ChenD, Reimann JD, et al. Impactof exploratory biomarkers on the treatment effect of bevacizumab inmetastatic breast cancer. Clin Cancer Res 2011;17:372–81.

17. Liu SK, Bham SA, Fokas E, Beech J, Im J, Cho S, et al. Delta-like ligand4-notch blockade and tumor radiation response. J Natl Cancer Inst2011;103:1778–98.

18. Scehnet JS, Jiang W, Kumar SR, Krasnoperov V, Trindade A, Benedito R,et al. Inhibition of Dll4-mediated signaling induces proliferation ofimmature vessels and results in poor tissue perfusion. Blood 2007;109:4753–60.

19. Hoey T, YenWC,Axelrod F, Basi J, Donigian L,Dylla S, et al. DLL4 blockadeinhibits tumor growth and reduces tumor-initiating cell frequency. CellStem Cell 2009;5:168–77.

20. YenWC, FischerMM,HynesM,Wu J, KimE, Beviglia L, et al. Anti-DLL4 hasbroad spectrum activity in pancreatic cancer dependent on targeting DLL4-Notch signaling in both tumor and vasculature cells. Clin Cancer Res 2012;18:5374–86.

21. FischerM, YenWC, Kapoun AM,WangM,O'YoungG, Lewicki J, et al. Anti-DLL4 inhibits growth and reduces tumor-initiating cell frequency incolorectal tumors with oncogenic KRAS mutations. Cancer Res 2011;71:1520–5.

22. Mani SA, Guo W, Liao MJ, Eaton EN, Ayyanan A, Zhou AY, et al. Theepithelial-mesenchymal transition generates cells with properties of stemcells. Cell 2008;133:704–15.

23. Chiorean EG, LoRusso P, Strother RM, Diamond JR, Younger A,Messersmith WA, et al. A phase I first-in-human study of enoticumab(REGN421), a fully human delta-like ligand 4 (Dll4) monoclonal

antibody in patients with advanced solid tumors. Clin Cancer Res2015;21:2695–703.

24. SmithDC, Eisenberg PD,ManikhasG, ChughR,GubensMA, Stagg RJ, et al.A phase I dose escalation and expansion study of the anticancer stem cellagent demcizumab (anti-DLL4) in patients with previously treated solidtumors. Clin Cancer Res 2014;20:6295–303.

25. Gurney A, Hoey T. Anti-DLL4, a cancer therapeutic with multiple mechan-isms of action. Vasc Cell 2011;3:18.

26. Li JL, Harris AL. Crosstalk of VEGF and Notch pathways in tumourangiogenesis: therapeutic implications. Front Biosci 2009;14:3094–110.

27. Miles KM, Seshadri M, Ciamporcero E, Adelaiye R, Gillard B, Sotomayor P,et al. Dll4 blockade potentiates the anti-tumor effects of VEGF inhibitionin renal cell carcinoma patient-derived xenografts. PLoS One 2014;9:e112371.

28. Huang J, HuW, Hu L, Previs RA, Dalton HJ, Yang XY, et al. Dll4 inhibitionplus aflibercept markedly reduces ovarian tumor growth. Mol Cancer Ther2016;15:1344–52.

29. Kuhnert F, Chen G, Coetzee S, Thambi N, Hickey C, Shan J, et al. Dll4blockade in stromal cells mediates antitumor effects in preclinical modelsof ovarian cancer. Cancer Res 2015;75:4086–96.

30. WuC, YingH,Grinnell C, Bryant S,MillerR,Clabbers A, et al. Simultaneoustargeting of multiple disease mediators by a dual-variable-domain immu-noglobulin. Nat Biotechnol 2007;25:1290–7.

31. Ding K, Eaton L, BowleyD, RieserM,ChangQ,HarrisMC, et al. Generationand characterization of ABBV642, a dual variable domain immunoglob-ulin molecule (DVD-Ig) that potently neutralizes VEGF and PDGF-BB andis designed for the treatment of exudative age-related macular degenera-tion. mAbs 2017;9:269–84.

32. Edgell CJ, McDonald CC, Graham JB. Permanent cell line expressinghuman factor VIII-related antigen established by hybridization. Proc NatlAcad Sci U S A 1983;80:3734–7.

33. Zhang X, Pagel MD, Baker AF, Gillies RJ. Reproducibility of magneticresonance perfusion imaging. PLoS One 2014;9:e89797.

34. Checkley D, Tessier JJ, Kendrew J, Waterton JC, Wedge SR. Use of dynamiccontrast-enhanced MRI to evaluate acute treatment with ZD6474, a VEGFsignalling inhibitor, in PC-3 prostate tumours. Br J Cancer 2003;89:1889–95.

35. Tofts PS, Brix G, Buckley DL, Evelhoch JL, Henderson E, Knopp MV, et al.Estimating kinetic parameters from dynamic contrast-enhanced T(1)-weighted MRI of a diffusable tracer: standardized quantities and symbols.J Magn Reson Imaging 1999;10:223–32.

36. Hezareh M, Hessell AJ, Jensen RC, van de Winkel JG, Parren PW. Effectorfunction activities of a panel of mutants of a broadly neutralizingantibody against human immunodeficiency virus type 1. J Virol 2001;75:12161–8.

37. Folkman J. Tumor angiogenesis: therapeutic implications. N Engl J Med1971;285:1182–6.

38. Kuhnert F, Kirshner JR, Thurston G. Dll4-Notch signaling as a therapeutictarget in tumor angiogenesis. Vasc Cell 2011;3:20.

39. Lee D, Kim D, Choi YB, Kang K, Sung ES, Ahn JH, et al. Simultaneousblockade of VEGF and Dll4 by HD105, a bispecific antibody, inhibitstumor progression and angiogenesis. mAbs 2016;8:892–904.

40. Yen WC, Fischer MM, Argast G, Wallace B, Wang M, Meisner R, et al. Dualtargeting of the DLL4 and VEGF pathways with a bispecific monoclonalantibody inhibits tumor growth and reduces cancer stem cell frequency[abstract]. In: Proceedings of the AACR-NCI-EORTC International Con-ference; 2015Nov 5–9; Boston,MA. Philadelphia (PA): AACR;Mol CancerTher 2015;14(12 Suppl 2):Abstract nr C164.

41. Tartour E, Pere H, Maillere B, Terme M, Merillon N, Taieb J, et al. Angio-genesis and immunity: a bidirectional link potentially relevant for themonitoring of antiangiogenic therapy and the development of noveltherapeutic combination with immunotherapy. Cancer Metastasis Rev2011;30:83–95.

42. Buonerba C, Di Lorenzo G, Sonpavde G. Combination therapy for met-astatic renal cell carcinoma. Ann Transl Med 2016;4:100.

43. Ott PA, Hodi FS, Buchbinder EI. Inhibition of Immune checkpoints andvascular endothelial growth factor as combination therapy for metastaticmelanoma: an overview of rationale, preclinical evidence, and initialclinical data. Front Oncol 2015;5:202.






44. Kut C, Mac Gabhann F, Popel AS. Where is VEGF in the body? A meta-analysis of VEGF distribution in cancer. Br J Cancer 2007;97:978–85.

45. Jayson GC, Zweit J, Jackson A, Mulatero C, Julyan P, Ranson M, et al.Molecular imaging and biological evaluation of HuMV833 anti-VEGFantibody: implications for trial design of antiangiogenic antibodies. J NatlCancer Inst 2002;94:1484–93.

46. Gordon MS, Nemunaitis JJ, Ramanathan RK, Sledge GW, Wang L,Fang X, et al. Phase 1, open-label, dose-escalation and expansionstudy of ABT-165, a dual variable domain immunoglobulin(DVD-Ig) targeting both DLL4 and VEGF, in patients (pts)with advanced solid tumors. J Clin Oncol 34:15s, 2016 (suppl;abstr 2507).


Li et al.




2018;17:1039-1050. Published OnlineFirst March 28, 2018.Mol Cancer Ther Yingchun Li, Jonathan A. Hickson, Dominic J. Ambrosi, et al. Favorable Safety Profiles in Preclinical ModelsTargeting DLL4 and VEGF, Demonstrates Superior Efficacy and ABT-165, a Dual Variable Domain Immunoglobulin (DVD-Ig)

Updated version

10.1158/1535-7163.MCT-17-0800doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2018/03/28/1535-7163.MCT-17-0800.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/17/5/1039.full#ref-list-1

This article cites 45 articles, 14 of which you can access for free at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://mct.aacrjournals.org/content/17/5/1039To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-17-0800

http://mct.aacrjournals.org/content/suppl/2018/03/28/1535-7163.MCT-17-0800.DC1

http://mct.aacrjournals.org/content/17/5/1039.full#ref-list-1

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

mailto:[email protected]

http://mct.aacrjournals.org/content/17/5/1039


abt-165,adualvariabledomainimmunoglobulin (dvd-ig ... · vegf and vegf receptor (vegfr) tyrosine...

Documents