Combination Therapy with Bispeciﬁc Antibodies and PD-1 ......conjugation between a dimer of anti-X Fab and a monomer of anti-CD3 scFv via the fused DDD2 and AD2 peptides. B, CD8þ

Therapeutics, Targets, and Chemical Biology

Combination Therapy with Bispecific Antibodiesand PD-1 Blockade Enhances the AntitumorPotency of T CellsChien-Hsing Chang1,2, Yang Wang1, Rongxiu Li1, Diane L. Rossi1, Donglin Liu1,2,Edmund A. Rossi1, Thomas M. Cardillo1, and David M. Goldenberg1,2

Abstract

The DOCK-AND-LOCK (DNL) method is a platform tech-nology that combines recombinant engineering and site-spe-cific conjugation to create multispecific, multivalent antibodiesof defined composition with retained bioactivity. We haveapplied DNL to generate a novel class of trivalent bispecificantibodies (bsAb), each comprising an anti-CD3 scFv covalent-ly conjugated to a stabilized dimer of different antitumor Fabs.Here, we report the further characterization of two such con-structs, (E1)-3s and (14)-3s, which activate T cells and targetTrop-2– and CEACAM5-expressing cancer cells, respectively.(E1)-3s and (14)-3s, in the presence of human T cells, killedtarget cells grown as monolayers at subnanomolar concentra-tions, with a similar potency observed for drug-resistant cells.Antitumor efficacy was demonstrated for (E1)-3s coadminis-tered with human peripheral blood mononuclear cells (PBMC)

in NOD/SCID mice harboring xenografts of MDA-MB-231,a triple-negative breast cancer line constitutively expressingTrop-2 and PD-L1. Growth inhibition was observed followingtreatment with (E1)-3s or (14)-3s combined with humanPBMC in 3D spheroids generated from target cell lines to mimicthe in vivo behavior and microenvironment of these tumors.Moreover, addition of an antagonistic anti–PD-1 antibodyincreased cell death in 3D spheroids and extended survival ofMDA-MB-231-bearing mice. These preclinical results emphasizethe potential of combining T-cell–redirecting bsAbs with antago-nists or agonists that mitigate T-cell inhibition within the tumormicroenvironment to improve immunotherapy of solid cancersin patients. They also support the use of 3D spheroids as apredictive alternative to in vivo models for evaluating T-cell func-tions. Cancer Res; 77(19); 5384–94. �2017 AACR.

IntroductionT cells play an undisputed role in cancer immunotherapy.

T-cell–orchestrated antitumor immunity requires the recogni-tion of the antigenic peptide/major histocompatibility com-plex by the T-cell antigen receptor (TCR) and the subsequentengagement of costimulatory receptors to sustain T-cell acti-vation, which is negatively regulated by coinhibitory receptors.The increased knowledge of T-cell activation and modulationhas resulted in four current approaches to harness T-cell–mediated immune responses against malignant tumors; name-ly, (i) adoptive transfer of autologous T cells that exhibitantigen-specific, antitumor activity (1); (ii) infusion of hostT cells genetically engineered to express either TCR with alteredspecificity, or chimeric antigen receptors (CAR) with antibody-like recognition of a certain tumor (2); (iii) the exploration ofimmune checkpoint blockade to promote tumor regression

by overcoming inhibition and increasing survival of tumor-reactive T cells (3); and (iv) redirection of T cells by T-cellactivating and tumor-targeting, bispecific antibodies (bsAb) todestroy cancers (4). Moreover, ongoing preclinical and clinicalstudies to evaluate combinatorial strategies for further unleash-ing the antitumor immunity of T cells have yielded promisingresults, as exemplified by combinations of immune checkpointinhibitors (5–7), combinations of immune checkpoint inhi-bitors either with immune checkpoint activators (8–10) orwith genetically modified T cells (11), combinations ofT-cell–redirecting bsAbs with PD-L1 antagonists (12) or CD28agonists (13), and others (14).

The numerous bsAbs reported to date vary in formats, mole-cules they target, potential applications, as well as stages ofdevelopment (4, 15–17). For example, bsAbs generated to redirectimmune cells for killing cancer, besides their diverse tumortargets, can be Fc-bearing (12, 18, 19) or Fc-lacking (20–22);bivalent (20, 22), trivalent (18, 20,21, 23), or tetravalent (24–26);and activating T cells (18, 21, 22) or NK cells (23, 27).

We have previously described the construction of a novelclass of trivalent, Fc-lacking, T-cell–redirecting bsAbs (21), eachcomprising a single chain variable fragment (scFv) of Okt3(anti-CD3) site-specifically linked via the DOCK-AND-LOCK(DNL) method to a stabilized dimer of different antitumorFab's. These DNL conjugates are designated (X)-3s, where thecodes (X) and 3s denote the bivalent antitumor Fab dimerand the monovalent anti-CD3 scFv, respectively, as depictedin Fig. 1A. They have been shown to mediate the formation ofimmunological synapses between T cells and cognate target

1Immunomedics, Inc., Morris Plains, New Jersey. 2IBC Pharmaceuticals, Inc.,Morris Plains, New Jersey.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Corresponding Author: Chien-Hsing Chang, Immunomedics, Inc., 300 TheAmerican Road, Morris Plains, NJ, 07950. Phone: 973-531-9108; Fax: 973-605-8282; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-16-3431

�2017 American Association for Cancer Research.

CancerResearch

Cancer Res; 77(19) October 1, 20175384

on August 27, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst August 17, 2017; DOI: 10.1158/0008-5472.CAN-16-3431

http://crossmark.crossref.org/dialog/?doi=10.1158/0008-5472.CAN-16-3431&domain=pdf&date_stamp=2017-9-19

http://cancerres.aacrjournals.org/

cells, induce T-cell activation and proliferation in the presenceof target cells, kill target cells with subnanomolar IC50 whencocultured with T cells in vitro, and inhibit growth of humantumor xenografts in NOD/SCID mice reconstituted withhuman peripheral blood mononuclear cells (PBMC; ref. 21).We have also demonstrated in a follow-on study with (E1)-3s, arepresentative (X)-3s with specificity for Trop-2–expressingepithelial cancer cells, that the addition of interferon-a (IFN-a)further enhanced the potency of redirected T cells in vitrowithout a significant increase in cytokine production, and thatthe combination of (E1)-3s with peginterferon alfa-2a moreeffectively delayed the growth of Trop-2–expressing NCI-N87human gastric cancer xenografts in vivo than single treatmentswith either (E1)-3s or peginterferon alfa-2a (28). In the currentstudy, we provide evidence that the combination of a novelPD-1–blocking antibody (cPD-1) with (E1)-3s or the CEACAM5-targeting (14)-3s could potentiate the antitumor activity of redir-ected T cells against target human cancer cells grown in vitro asmonolayer cultures or three-dimensional (3D) multicellulartumor spheroids (MCTS; ref. 29), or in vivo as xenografts inNOD-SCID mice.

Materials and MethodsCell lines, antibodies, and reagents

Human cancer cell lines of breast (MDA-MB-231, HCC1954,HCC38, and BT-20), colon (HT-29, LS 174T, LoVo and COLO205), and T lymphocytes (Jurkat), were purchased from theATCC. Whereas MDA-MB-231, HCC38, HT-29, LS 174T, LoVoand COLO 205 were authenticated by short tandem repeat(STR) profiling in 2013 or 2014; HCC1954 (acquired in 2014),Jurkat (acquired in 2014), and BT-20 (acquired in 2010) havenot been tested by STR profiling, but were never in culturefor more than 50 passages or 6 months between collectionor thawing and use in the described experiments. Each cellline was maintained according to the recommendations ofATCC and routinely tested for Mycoplasma using MycoAlertMycoplasma Detection Kit (Lonza). For monolayer cultures,cells were grown in RPMI-1640 medium supplemented withL-glutamine (2 mmoles/L), 10% FBS, and 1% penicillin–strep-tomycin (Life Technology) at 37�C in a humidified atmosphereof 5% CO2. The establishment of MDA-MB-231-S120, a sub-line of MDA-MB-231 resistant to SN-38, has been describedpreviously (30). The bispecific (E1)-3s and (14)-3s were producedas reported (21).

Sublines of MDA-MB-231–overexpressing Trop-2 were gener-ated as follows. The cDNA of Trop-2 (GenBank: X77754.1) wassynthesized, inserted into the HindIII-EcoRI sites of pcDNA3.1(Thermo Fisher Scientific), and worked up to obtain the correctlyassembled vector, which was transfected into MDA-MB-231 cellswith lipofectamine using the Lipofectamine 2000 DNA Transfec-tion Reagent Protocol (Thermo Fisher Scientific). At 72 hourspost-transfection, cells were placed in fresh medium containing1,000 mg/mL of G418 (Geneticin; Thermo Fisher Scientific) for7 days, then serially diluted and dispensed into a 96-well plate.G418-resistant colonies were identified, picked, expanded in6-well plates, stained with AF647-conjugated hRS7, and analyzedon a FACSCanto flow cytometer (BD Biosciences), from whichseven clones with a similar or enhanced expression of Trop-2relative to the parental MDA-MB-231 were selected for furtherexperiments.

Figure 1.

Assembly of (X)-3s by the DNL method and the in vitro cytotoxicitydetermined with the MTS assay. A, Schematic showing the site-specificconjugation between a dimer of anti-X Fab and a monomer of anti-CD3 scFvvia the fused DDD2 and AD2 peptides. B, CD8þ T cells combined with breastcancer cells of MDA-MB-231, HCC38, HCC1954, or BT-20 at a ratio of 6 to 1were treated with (E1)-3s at different concentrations for 48 hours. Theresulting cell viability relative to that of the untreated sample (set as 100%)was plotted against the log concentration of (E1)-3s to obtain thecorresponding dose–response curve and IC50 as shown. C, The dose–responsecurve and IC50 obtained for (14)-3s in COLO 205, HT-29, LoVo, and LS 174T.

Combination of PD-1 Blockade and T-cell–Redirecting bsAbs

www.aacrjournals.org Cancer Res; 77(19) October 1, 2017 5385




SpESF-X10-2D1, a PD-1–overexpressing clone, was selected inmedium containing methotrexate (0.2 mmol/L) after transfectionof SpESF-X10 (31) with PD-1-pdHL2, which was constructed asfollows. The cDNA of PD-1 was PCR-amplified from HumanPD-1/PDCD1 Gene cDNA ORF Clone (AcroBiosystems) usingPD-1/XbaI left primer (TCTAGACACAGGACCTCACCATGCA-GATCCCACAGGCGCC) and PD-1/EagI right primer (CGGCC-GTCAGAGGGGCCAAGAGCAGTGTCC). The resulting amplimerwas cloned into pGemT plasmid, fromwhich the sequence of theXbaI/EagI insert was confirmed and ligated with the XbaI/EagIfragment of IFNa2b-DDD2-pdHL2 (32).

Generation of 5G9.G1.B11 and cPD-1The PD-1–blocking mAb and its chimeric counterpart were

generated as follows. BALB/c mice were immunized with recom-binant human PD-1–Fc fusion protein (AB Biosciences), resultingin the isolation of a positive clone (5G9) by hybridoma technol-ogy. To ensuremonoclonality, 5G9was subcloned twice, yielding5G9.G1.B11. Subsequently, VK and VH sequences of 5G9.G1.B11were determined, combined with their respective human kappaand human IgG1 constant domains, and expressed in SpESF-X10to obtain cPD-1.

Isolation of PBMCsBuffy coats of healthy donors were purchased from the NJ

Blood Center. PBMCs were obtained by density gradient cen-trifugation on Ficoll-Paque Premium (GE Healthcare Bios-ciences) following the manufacturer's protocol with minormodifications. Briefly, the buffy coat was diluted 3-fold withsterile PBS, and 35 mL of this suspension was slowly pouredover 15 mL separation medium. After 30 minutes of centrifu-gation at 1,700 rpm (Sorvall RT Plus) with brakes turned off,the interphase was moved to a new reaction tube, washed twicewith PBS, and centrifuged for 10 minutes at 1,500 rpm. Theresulting cell pellets were suspended in complete RPMI-1640medium, adjusted to a concentration of 1 � 107/mL, and useddirectly. Alternatively, the suspended PBMCs were adjusted to aconcentration between 2 and 6 � 107/mL, mixed with an equalvolume of 2 x freezing medium (20% DMSO and 80% FBS),from which 1-mL aliquots were dispensed into cryogenic vialsand stored at �80�C until needed.

Isolation of CD8þ T cells from whole bloodWhole blood of healthy donors was added to Uni-SepMAXI

tubes (Novamed) and fractionated by centrifugation. The buffycoat was collected, washed 2 to 3 times with PBS and centrifugedat 250 � g to further remove platelets. The resulting PBMCs werecounted and 1 � 108 cells were pelleted for isolation of CD8þ Tcells by negative selection using the kit and procedure of MiltenyiBiotech.

Cytotoxicity assays for tumor cells grown in monolayerAdherent tumor cells were dissociated with Trypsin-EDTA

(0.25%), washed 1 time with complete medium, counted, com-bined with CD8þ T cells at an effector to target ratio of 6:1, andcentrifuged at 400 � g. The resulting pellet was suspended inmedium such that the addition of 100 mL would provide 1� 104

tumor cells and6�104T cells (effector-to-target ratio of 6) in eachwell of a 96-well plate. (E1)-3s or (14)-3s, starting at 6 nmol/mL,was 10-fold serially diluted in the medium specified for thetarget tumor cell line. After adding 200 mL of each dilution to

each well, the plate was incubated for 24 to 72 hours. At theend of incubation, media were removed from wells and replacedwith fresh, warm complete media to flush any nonadherentcells remaining in the wells. Media containing MTS reagentwere then introduced and OD490 of the plate was read whensufficient color appeared in the untreated wells. The assay wasdone in triplicate.

Formation of MCTS for cytotoxicity evaluationNinty-six–well plates were coated with 80 mL of 1% agarose

(Sigma) per well, which was then filled with 10,000 tumor cellsdetached from monolayer in 200 mL culture medium. After4 days, the single spheroid formed in each well was transferredto non–tissue-culture–treated 6-well plates, with each wellcontaining up to 6 spheroids in 2 mL culture medium andtreated with testing agents for 2 to 3 days as indicated. At theend of the treatment, propidium iodide (PI) was added to staindead cells and the fluorescent images were taken with a micro-scope camera.

Quantitation of surface antigensExpression of Trop-2 or CEACAM5 on the cell surface was

determined by flow cytometry (33). Briefly, cells were harvestedwith Accutase Cell Detachment Solution (BD Biosciences) andassayed for Trop-2 or CEACAM5 expression using QuantiBRITEPE beads (BD Biosciences) and a PE-conjugated anti–Trop-2 oranti-CEACAM5 antibody (eBiosciences,) following the manufac-turer's instructions. Data were acquired on a FACSCanto flowcytometer (BDBiosciences) with CellQuest Pro software. Stainingwas analyzed with Flowjo software (Tree Star).

In vivo studiesFemale NOD/SCID mice, 5-week-old, were purchased from

Charles River Laboratories. All animal studies were approved bythe Rutgers School of Biomedical and Health Sciences Institu-tional Animal Care and Use Committee. Mice were injectedsubcutaneously with a mixture of MDA-MB-231 tumor cells (5� 106) and purified human T cells (2.5� 106) combined with anequal volume of Matrigel. Therapy began one hour later. Treat-ment regimens, dosages, and number of animals are described inthe figure legend. Animals were monitored daily for signs oftumor out-growth. Once tumors appeared, they were measuredtwiceweekly. Tumor volumewas determinedbymeasurements intwodimensions using calipers,with volumesdefined as:L�w2/2,where L is the longest dimension of the tumor and w the shortest.

Statistical analysis for the tumor growth data was based onAUC (area under curve) and survival time. Profiles of individualtumor growth were obtained through linear curve modeling. AnF-test was employed to determine equality of variance betweengroups before statistical analysis of growth curves. A two-tailedt test was used to assess statistical significance between all thevarious treatment groups and controls except for the salinecontrol, in which a one-tailed t test is used. As a consequence ofincompleteness of some of the growth curves (due to deaths),statistical comparisons of AUC were only performed up to thetime at which the first animal within a group was sacrificed.Survival was analyzed using Kaplan–Meier plots (log-rankanalysis), using the Prism GraphPad Software (v6.05) pur-chased from Advanced Graphics Software. Survival surrogate

Chang et al.

Cancer Res; 77(19) October 1, 2017 Cancer Research5386




endpoint was the time for tumor progression to 1.0 cm3.Significance was considered at P � 0.05.

ResultsCharacterization of 5G9.G1.B11, an antagonistic murineanti–PD-1 monoclonal antibody, and its chimericcounterpart (cPD-1)

The 5G9.G1.B11 mAb was purified by Protein A to homo-geneity, as shown by SE-HPLC (Supplementary Fig. S1A) andSDS-PAGE (Supplementary Fig. S1B). The reactivity of 5G9.G1.B11 for PD-1 was confirmed by SE-HPLC with its binding torecombinant PD-1-His (Supplementary Fig. S1C), by ELISAwith its binding to either PD-1-His or PD-1-Fc (SupplementaryFig. S1D), and by flow cytometry with its binding to PD-1-expressed on activated Jurkat T cells (Supplementary Fig. S2).Importantly, the blocking activity of 5G9.G1.B11 was demon-strated by a dose-dependent increase of IL-2 secreted by T cellsin a mixed lymphocyte assay (Supplementary Fig. S3). cPD-1, achimeric version of 5G9.G1.B11, comprising the VH and VK of5G9.G1.B11 and human Fc of IgG1, was generated, and themAb from the lead clone (2G9) was purified and shown to behomogeneous by SE-HPLC (Supplementary Fig. S4A) and SDS-PAGE (Supplementary Fig. S4B). The binding of cPD-1 torecombinant PD-1-His was demonstrated by ELISA (Supple-mentary Fig. S4C; EC50 ¼ 49 � 16 pmol/L based on threeseparate experiments), and further confirmed by flow cytome-try (Supplementary Fig. S5A) with SpESF-X10-2D1, a subline ofSpESF-X10 (34) transfected to overexpress PD-1 (Supplemen-tary Fig. S5B). The blocking activity of cPD-1 was comparablewith that of 5G9.G1.B11 or EH12.2H7 (a commercially avail-able mouse anti-human PD-1 blocking mAb), as demonstratedby inhibiting the binding of biotinylated PD-1 (B�-PD-1) to theendogenous PD-L1 expressed on MDA-MB-231 (Supplemen-tary Table S1). In this experiment using flow cytometry, the

median fluorescence intensity (MFI) determined for the bind-ing of biotinylated PD-1 (B�-PD-1) to PD-L1 on MDA-MB-231cells with phycoerythrin-labeled streptavidin (PE-SA) decreasedwith the addition of increasing amounts cPD-1, 5G9.G1.B11, orEH12.2H7, and was observed at a similar level for each of thethree tested concentrations (1.5, 15, and 150 mg/mL), furthersupporting the antagonistic properties of cPD-1.

Potent killing of target cells grown in monolayerThe cytolytic activity of (E1)-3s via engagement of human T cells

with Trop-2–expressing tumor cells has been demonstrated in vitrofor Capan-1 (pancreatic cancer) and NCI-N87 (gastric cancer) todisplay IC50 (the concentration to achieve 50% lysis) of 29 and0.85 pmol/L, respectively (21). In addition, the IC50 of (14)-3s wasdetermined to be 2 pmol/L for LS 174T (a human colonic cancercell line). We now report that (E1)-3s alsomediated potent killing,via T cells at an effector-to-target ratio of 6, of four breast cancer celllines (MDA-MB-231, HCC1954, HCC38, and BT-20), all of whichexpress endogenous Trop-2 on the cell surface (SupplementaryTable S2); as well as four Trop-2-transfected sublines of MDA-MB-231 (designated as 231-C13, 231-C29, 231-C36, and 231-C39),each having a higher expression of Trop-2 than the parental MDA-MB-231 (Supplementary Table S3). Based on the 48-hours viabil-ity assays using MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxy-methoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium)], the IC50 of(E1)-3s was 4.4 � 1.5 pmol/L (n ¼ 4) for MDA-MB-231 andbetween 1 and 10 pmol/L for the other cell lines (Fig. 1B andSupplementary Table S3). In addition, the IC50 of (E1)-3s deter-mined for the SN-38-resistant MDA-MB-231-S120 (33) from the24-hours assaywas 11.5 pmol/L, similar to the IC50 of 12.4 pmol/Ldetermined for MDA-MB-231 (Supplementary Table S4), indicat-ing the overexpression of ABCG2, a multidrug-resistant trans-porter belonging to the ATP-binding cassette sub-family G, inMDA-MB-231-S120 has little impact on the antitumor effect ofT cells redirected by (E1)-3s.

Figure 2.

Effect of (E1)-3s and T cells on single spheroids of MDA-MB-231. A, Notable death of MDA-MB-231 cells in the single spheroids, revealed as red fluorescenceof PI, was only evident in the presence of both T cells [6 � 106; labeled with green fluorescent (CFSE) and (E1)-3s (100 pmo/L)]. B, Killing of cells in MDA-MB-231spheroids by (E1)-3s (100 pmol/L) was detectable with 1.5 � 106 T cells. All images were taken at 24 hours after adding T cells, (E1)-3s, or both, to spheroids.






For colonic cancer cell lines (Fig. 1C), (14)-3s was confirmedto display a subnanomolar IC50 (19 � 13 pmol/L; n ¼ 2) inLS 174T, while showing a similar activity in LoVo (IC50 ¼ 19 �17 pmol/L; n ¼ 2), but a much reduced strength in eitherCOLO 205 (IC50 ¼ 671 � 311 pmol/L; n ¼ 2) or HT-29 (IC50 >3,000 pmol/L; n ¼ 2). Because the surface expression ofCEACAM5 is relatively high on LS 174T or LoVo cells (bindingsites for hMN-14 per cell in excess of 20,000; SupplementaryTable S2) and considerably lower on COLO 205 cells (bindingsites for hMN-14 per cell about 7,000 to 8,000; Supplemen-tary Table S2), these results reflect a likely correlation betweenCEACAM5 expression and (14)-3s potency, as observed fortwo other bsAbs, CEA TCB (18) and MEDI-565 (34) thatredirect T cells to kill CEACAM5-positive colorectal cancer celllines. However, the much higher IC50 observed for HT-29 alsosuggests that the expression levels of CEACAM5 might not bethe only determinant for the potency of (14)-3s.

Cytolysis of MCTS by bsAb-redirected T cellsSingle spheroids of MDA-MB-231 were grown for 5 days in

96-well plates coated with agarose, transferred to 6-wellplates, and evaluated for cytolysis by T cells [labeled intra-

cellularly with carboxyfluorescein succinimidyl ester (CFSE)]in the presence of (E1)-3s at 100 pmol/L. As shown in Fig. 2A,cells in spheroids were effectively killed at 24 hours only inthe presence of both (E1)-3s and T cells. Under these condi-tions, cell death was detectable at 24 hours with 1.5 � 106 ofT cells (Fig. 2B) and most, if not all, cells in the spheroidstreated with the highest concentration of T cells (6 � 106)were dead at 48 hours, but not the surrounding T cells (datanot shown). With PBMCs as effector cells and after a 48-hoursincubation, (14)-3s at 100 pmol/L likewise killed the spher-oids of HT-29 (Fig. 3A) and LS 174T (Fig. 3B). Similarcytotoxicity in 3D spheroids of HT-29 and LS 174T was alsodemonstrated for (14)-3s using purified T cells, as shownSupplementary Fig. S6.

Blockade of PD-1 by cPD-1 enhanced lysis of target cellsgrown in monolayer or as MCTS

The constitutive expression of PD-L1 on MDA-MB-231(Supplementary Fig. S7 and ref. 35) and its Trop-2–overex-pressing sublines (Supplementary Table S3) suggests theywould be suitable for evaluating the potential of cPD-1 toenhance the potency of T cells by blocking the interaction of

Figure 3.

Effect of (14)-3s and PBMCs on singlespheroids of HT-29 or LS 174T.Prominent cell death was observed insingle spheroids of HT-29 (A) or LS 174T (B) with the addition of (14)-3s (100pmol/L) and PBMC (thawed overnightfrom a frozen stock), but not in thethree controls [spheroid-only;spheroid and PBMCs; spheroid and(14)-3s]. Each well contained 2 mL ofmedium with 6 � 106 PBMCs.

Chang et al.





PD-1 with PD-L1. The results shown in Supplementary TableS4 indicate that a 2- to 3-fold decrease of IC50 was observed forMDA-MB-231 and MDA-MB-231-S120 with the addition of 10to 15 mg/mL of cPD-1 to the monolayer culture in which CD8þ

T cells were present at an effector-to-target ratio of 6, whereasthe concentrations of (E1)-3s were varied. Moreover, when thecorresponding EC50 (half maximal effective concentration)values were analyzed with the Prism software for each pairedsamples, the enhanced potency by cPD-1 was highly significant(P < 0.0001; Supplementary Table S4). Further evidence forcPD-1 to enhance the potency of T cells redirected by (E1)-3swas provided by the demonstration of more dead cells in thespheroids of MDA-MB-231 treated with both (E1)-3s and cPD-1 than those treated with only (E1)-3s, using PBMCs (Fig. 4).

The effect of cPD-1 also was investigated for HCC1954,which can be induced by IFN-g to express PD-L1, as shownby flow cytometry of individual cells obtained from monolayerculture (Supplementary Fig. S8A) or by immunostaining ofclustered cells in a spheroid (Supplementary Fig. S8B). Themethod to evaluate the response of HCC1954 spheroids,stimulated with IFN-g or not, to combined (E1)-3s and PBMCs,is illustrated in Fig. 5A. Whereas the addition of cPD-1 clearlyincreased the extent of cell death in HCC1954 spheroidspretreated with IFN-g (Fig. 5B) or not (Fig. 5C), PD-1 blockadewith cPD-1 evidently resulted in more T-cell–mediated deathin pretreated (Fig. 5D, bottom) than in untreated samples (Fig.5D, top). Using T cells purified from PBMCs, the addition ofcPD-1 to (E1)-3s also induced more cell death in the spheroidsformed from IFN-g–stimulated MDA-MB-231, when comparedwith (E1)-3s without the addition of cPD-1 (Supplementary

Fig. S9A). The purified T cells also induced (E1)-3s–mediatedcell death in the spheroids formed from MDA-MB-231, but theeffect of cPD-1 was not as prominent as that observed in thespheroids formed from IFN-g–stimulated MDA-MB-231 (Sup-plementary Fig. S9B and S9C). These results thus provide abasis for using the 3D spheroids to study T-cell functions as analternative to humanized murine models. The presence of PD-1on T cells isolated from PBMCs as well as on T cells in thePBMC population was shown by flow cytometry in Supple-mentary Fig. S10.

Combination of (E1)-3s and human T cells with cPD-1improved the prolonged survival of NOD/SCID micebearing MDA-MB-231 xenografts

Because MDA-MB-231 (a triple-negative human breast can-cer cell line) is shown to express PD-L1 and that in vitrotreatment with (E1)-3s and cPD-1 effects a significant increasein T-cell killing of target cells mediated by (E1)-3s, an animalmodel was established to determine whether these resultswould translate to an in vivo setting. The study consisted offive groups: three treatment groups, one untreated controlgroup, and one out-growth control group, with 5 NOD/SCIDmice in each group. To begin the therapy, the out-growthcontrol group was injected with MDA-MB-231 cells only, andthe other four groups were each injected subcutaneously with amixture of MDA-MB-231 and human T cells. One hour later,(E1)-3s, cPD-1, or a combination of (E1)-3s and cPD-1 wasadministered to each of the three treatment groups intrave-nously (i.v.), whereas saline was given to the untreated controlgroup as well as the out-growth control group.

Figure 4.

Effect of cPD-1 and PBMCs on single spheroids of MDA-MB-231 treated with (E1)-3s. Cells were grown for 4 days to form spheroids and transferred to 6-wellplates by a first operator. The addition of PBMCs (2 � 105 in 2 mL), (E1)-3s (10 pmol/L), and cPD-1 (20 mg/mL) to respective wells containing thespheroids was performed by a second operator and blinded to the first operator. After incubation for 72 hours, PI was added, photos taken, and thecontent of each well disclosed. Enhanced cell death by cPD-1 was indicated by the higher fluorescent intensity of Exp-5 than that of Exp-4.






Untreated control mice injected with the mixture of tumorcells and T cells produced tumors that progressed at a slowerrate when AUC was compared with that of outgrowth controlanimals injected with only MDA-MB-231 cells (Fig. 6A; P ¼0.0378), suggesting that these T cells may have some activityon their own to slow tumor growth and improve survival(Fig. 6B; P ¼ 0.0031). However, treatment with only cPD-1was not significant to provide an antitumor benefit to theanimals and was no different from untreated control mice.Conversely, both (E1)-3s alone and (E1)-3s plus cPD-1 sig-nificantly slowed tumor growth in these mice (P < 0.0121;AUC), with a resulting significant improvement in overallsurvival compared with controls (Fig. 6B; P < 0.0031). Impor-tantly, the addition of cPD-1 therapy with (E1)-3s did pro-vide an even greater antitumor effect compared with micetreated with only (E1)-3s (P ¼ 0.0121; AUC). At the lastassessable time-point for both groups (day 42), mice treatedwith (E1)-3s plus cPD-1 had tumors that were 2.3-fold smallerin volume than in those treated with only (E1)-3s (0.432 �0.264 cm3 vs. 0.979 � 0.126 cm3; P ¼ 0.0031). The combi-nation therapy likewise provided a significant survival benefit

compared with (E1)-3s monotherapy (median survivaltime ¼ 49 days vs. 42 days, respectively; P ¼ 0.008). Thesedata suggest that in the absence of (E1)-3s, T cells alone mayexhibit some antitumor activity in this model, but it was solow that the addition of a checkpoint inhibitor was notenough to produce an improved immune response. On theother hand, the redirection of these T cells by (E1)-3s toMDA-MB-231 would better activate them, and the additionof cPD-1 would block PD-L1–mediated inhibitory signaling ofthese same T cells, resulting in an overall significantly improvedtherapeutic effect.

DiscussionSalient features of (X)-3s, a novel class of T-cell–redirectingbsAbs

BsAbs capable of redirecting T cells to cancers and therebyactivating such T cells to kill malignant tumors have shownpromise in both preclinical and clinical studies. However, theirtherapeutic benefits on late-stage patients with solid cancersremain to be established. Previously, we reported highly effective

Figure 5.

Effect of cPD-1 and PBMCs on singlespheroids of HCC1954 treated with(E1)-3s. A, Schematic showing themethod to evaluate the response ofHCC1954 spheroids, stimulated with orwithout IFN-g , to (E1)-3s and PBMCs;representative results shown forHCC1954 stimulated with and withoutIFN-g in B and C, respectively. D,Blockade of PD-1 by cPD-1 elicitedmore T-cell-mediated death inHCC1954 spheroids pretreated withIFN-g (bottom image) than untreated(top image).

Chang et al.





T-cell–redirected therapy of Trop-2–expressing pancreatic andgastric tumors in xenograft models using (E1)-3s, and noted thatthe relatively low level of cytokine release induced by (E1)-3swould be a potential advantage over other types of T-cell–redir-ected bsAbs, in particular, BiTE (28).

In the current study, we explored the potential utility of (E1)-3s for therapy of Trop-2–expressing breast cancers, includingtriple-negative breast cancer (TNBC), as well as that of (14)-3sfor CEACAM5-expressing colonic cancers. Both (E1)-3s and(14)-3s were highly potent for their target tumor cells, exhibit-ing largely subnanomolar IC50 against diverse cell lines grownin monolayer cultures. Moreover, the addition of cPD-1, a

novel checkpoint inhibitor to PD-1, could significantly en-hance (E1)-3s–mediated T-cell killing of MDA-MB-231 cells,which constitutively express PD-L1 (35). Additional evidencesupporting the therapeutic benefits of PD-1 blockade withcPD-1 was highlighted in treatments of NOD/SCID mice bear-ing MDA-MB-231 xenografts with a combination of cPD-1,(E1)-3s, and human T cells, which in comparison with treat-ments without the addition of cPD-1, significantly inhibitedtumor growth, as assessed by AUC (P ¼ 0.0121), and improvedthe median survival time by one week (P ¼ 0.008). Althoughthe in vivo data were obtained from only one xenograft model,appeared to be modest, and might not be observed in other

Figure 6.

Therapeutic efficacy of (E1)-3s plus cPD1 in a PD-L1-expressing TNBC tumor-xenograft disease model and potential combination therapy to enhance antitumoractivity of bsAb-redirected T cells. A, Mean tumor growth curves of MDA-MB-231 xenografts for treated and control animals. (E1)-3s (red arrows) wasadministered as 47 mg i.v. injections (administered daily for 5 days). cPD1 (gray arrows) was administered as 500 mg i.p. injections twice weekly for 4 weeks.Statistical comparisons were made on AUC data as described in Materials and Methods. B, Survival curves generated for each treatment group. Mice weredeemed to have succumbed to disease progression once the tumor volume exceeded 1.0 cm3 in size. Kaplan–Meier plots were analyzed for significanceas described in Materials and Methods. n.a., not applicable.






tumor models, these encouraging results were demonstratedsemi-quantitatively for (E1)-3s and (14)-3s with the cognate3D spheroids, which have been widely recognized to betterreflect the physiology and microenvironment of tumor cellsin vivo, and are being used increasingly in cancer research tocomplement monolayer cultures for drug testing, as exempli-fied by the pharmacological evaluation of T-cell–redirecting,anti-EpCAM x anti-CD3, calutumomab (36) and other antitu-mor agents, including antibody–drug conjugates (37) and amonospecific anti-EGFR antibody (38).

Comparison with other T-cell–redirecting bsAbsAddition of a PD-L1-blocking antibody toHER2-TDB, a T-cell–

engaging andHER2-targeting bsAb produced in bacteria using theknob-into-hole approach (39), was reported to enhance thetherapeutic outcomes of hCD3 transgenicmice carrying syngeneicCT26-HER2 tumors (12).Despite thedifference fromHER-TDB inmolecular components and target tumor antigen, (E1)-3s displaysantitumor potency in Trop-2-expressing TNBC cells grown inmonolayer cultures with low picomolar IC50 comparable withHER2-TDB in HER2-expressing cancer cells, and like HER2-TDB,was effective in killing target cancer cells resistant to establisheddrug and antibody–drug conjugates. Moreover, the potentialbenefits of combination therapy with (E1)-3s and anti–PD-1 foraugmenting T-cell immunity was demonstrated in a humanizedmousemodel and further shown in 3D spheroids to complementthe in vivo results.

More recently, several examples of Fc-bearing bsAbs generatedby CrossMab technology (40) have been reported, including theT-cell-redirecting CEA TCB (18), which as (14)-3s, binds mono-

valently to CD3 and bivalently to the membrane proximaldomain of human CEACAM5. Despite apparent differences inconstituent antibodies and respective architecture, the resem-blance of CEA TCB to (14)-3s in target specificity and valency isnoteworthy. Yet a comparison of the EC50 determined undersimilar conditions in LS 174T for CEATCB (1,410pmol/L; ref. 41)with that of (14)-3s (2 to 28 pmol/L; ref. 21 and this study)indicates at least a 50-fold increase in potency for (14)-3s. Theanti-CEACAM5 component of (14)-3s is derived from hMN-14(42),whichbinds to a regionof theA3B3domain in the vicinity ofthe cell membrane. Because the effectiveness of a T-cell–redirect-ing bsAb could be inversely correlated with the distance of itscognate epitope to the target cell surface (43), whether the higherpotency of (14)-3s might result from the closer binding ofhMN-14 to the cell surface of LS 174T than that of PR1A3 (44)used in CEA TCB, remains to be investigated.

Cancer immunotherapy using bsAb-redirected T cells versusCAR-engineered T cells

Whereas T cells can be redirected to kill cancer cells usingnovel CARs or modified TCRs, with impressive results againstB-cell malignancies as reported in clinical studies for acutelymphoblastic leukemia (45) and relapsed refractory chroniclymphocytic lymphoma (46), the alternative approach to redir-ecting T cells with bsAbs for cancer therapy is believed to bemore amenable to commercial development and has resultedin two regulatory approvals, catumaxomab in 2009 for thetreatment of malignant ascites (47) and blinatumomab in 2014for the treatment of B-precursor acute lymphoblastic leukemia(48). Similar to the continuing evolution of novel CAR-T cells

Figure 7.

Schematic depicting the stimulatory and inhibitory interplays induced by T-cell–redirecting (X)-3s upon linking effector T cells to target tumor cells. Theligation of CD3 on effector T cells with the cognate cell-surface antigen on tumor cells by (X)-3s triggers the activation of T cells, which releaseperforin and granzyme B to kill target tumor cells, secrete IFN-g to induce the expression of PD-L1 and IDO on target tumor cells, and upregulate PD-1. Whereasblockade of PD-1/PD-L1 interaction may augment the antitumor immunity of T cells, it can also enhance A2A adenosine receptor expression on T cells,leading to inhibition of T cells by adenosine produced from CD73-expression tumor cells. Understanding the relative contribution of each of these processesto the destruction and protection of tumor should lead to more effective combination immunotherapy of cancer with T-cell–redirecting bsAbs.

Chang et al.





with the goal to improve safety and retain efficacy (49), numer-ous efforts are underway to develop optimal bsAbs that wouldredirect T cells to kill cancer cells with picomolar potency andIgG-like serum half-life, among which HER2-TDB appears to bea promising lead (12). Encouragingly, (E1)-3s is comparable toHER2-TDB with regard to several in vitro properties, includingthe enhanced potency in combination with antagonistic anti–PD-1 or anti–PD-L1 antibodies. However, recent reports haverevealed the blockade of PD-1 could enhance A2A adenosinereceptor (A2AR) expression on tumor-infiltrating CD8þ T cells,making them more susceptible to suppression by adenosineproduced by CD73 on cancer cells (50). Thus, combiningantagonists of A2AR with checkpoint inhibitors of the PD-1/PD-L1 pathway should further augment the antitumor activityof bsAb-redirected T cells.

In conclusion, the T-cell–redirecting bsAbs generated with theDNLmethod,which are distinguishable by their highpotency andlow production of PD-L1–stimulating IFN-g , offer a promisingnew format to improve the combination immunotherapy of solidcancers with antagonists or agonists that alleviate T-cell inhibitionwithin the tumor microenvironment. The current study alsoprovides a basis for using 3D spheroids as an alternative to invivo models for evaluating T-cell functions, with future experi-ments to address the effect of exhausted T cells, which are thepredominant phenotype present in the tumor stroma. Althoughthe cytolytic efficiency of (14)-3s, (E1)-3s, and other T-cell–redirecting bsAbs evidently requires the presence of cognateantigens on target cells, their ultimate potency in a certain cancermay depend on additional factors, such as the mutational status,epitope specificity and density on surface, and expression ofintrinsic and induced immune evasion molecules, all of whichremain to be explored (depicted in Fig. 7).

Disclosure of Potential Conflicts of InterestD.M. Goldenberg has ownership interest (including patents) in Immuno-

medics, Inc. No other potential conflicts of interest were disclosed by theauthors.

Authors' ContributionsConception and design: C.-H. Chang, E.A. Rossi, D.M. GoldenbergDevelopment of methodology: C.-H. Chang, D. Liu, E.A. RossiAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): Y. Wang, R. Li, D.L. Rossi, D. Liu, T.M. CardilloAnalysis and interpretation of data (e.g., statistical analysis, biostati-stics, computational analysis): C.-H. Chang, Y. Wang, D.L. Rossi, D. Liu,E.A. Rossi, T.M. Cardillo, D.M. GoldenbergWriting, review, and/or revision of the manuscript: C.-H. Chang, Y. Wang,D. Liu, E.A. Rossi, T.M. Cardillo, D.M. GoldenbergAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): D.M. GoldenbergStudy supervision: C.-H. Chang, D.M. Goldenberg

AcknowledgmentsWe thank John Kopinski, Diana Chereches, and Ali Mostafa for excellent

technical assistance.

Grant SupportThis work was supported by Immunomedics, Inc. There was no external

financial or grant support received.The costs of publication of this article were defrayed in part by the

payment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

ReceivedDecember 19, 2016; revisedMay 24, 2017; accepted August 4, 2017;published OnlineFirst August 17, 2017.

References1. Rosenberg SA, Restifo NP. Adoptive cell transfer as personalized immu-

notherapy for human cancers. Science 2015;348:62–8.2. Sharpe M, Mount N. Genetically modified T cells in cancer therapy:

opportunities and challenges. Dis Model Mech 2015;8:337–50.3. Sharma P, Allison JP. The future of immune checkpoint therapy. Science

2015;348:56–61.4. Del Bano J, Chames P, Baty D, Kerfelec B. Taking up cancer immuno-

therapy challenges: bispecific antibodies, the path forward? Antibodies2016;5:1–23.

5. Carlino MS, Long GV. Ipilimumab combined with nivolumab: a standardof care for the treatment of advanced melanoma? Clin Cancer Res2016;22:3992–8.

6. Sakuishi K, Apeto L, Sullivan JM, Blazar BR, Kuchroo VK, Anderson AC.Targeting Tim-3 and PD-1 pathways to reverse T cell exhaustion and restoreanti-tumor immunity. J Exp Med 2010;207:2187–94.

7. Woo SR, Turnis ME, Goldberg MV, Bankoti J, Selby M, Nirschl CJ, et al.Immune inhibitory molecules LAG-3 and PD-1 synergistically regulate T-cell function to promote tumoral immune escape. Cancer Res 2011;72:917–27.

8. Chen S, Lee LF, Fisher TS, Jessen B, Elliott M, EveringW, et al. Combinationof 4-1BB agonist and PD-1 antagonist promotes antitumor effector/mem-ory CD8 T cells in a poorly immunogenic tumor model. Cancer ImmunolRes 2015;3:149–60.

9. Guo Z, Cheng D, Xia Z, Luan M, Wu L, Wang G, et al. Combined TIM-3blockade and CD137 activation affords the long-term protection in amurine model of ovarian cancer. J Transl Med 2013;11:215.

10. Lu L, Xu X, Zhang B, Zhang R, Ji H, Wang X. Combined PD-1 blockade andGITR triggering induce a potent antitumor immunity in murine cancermodels and synergizes with chemotherapeutic drugs. J Transl Med2014;12:36.

11. John LB, DevaudC,DuongCP, YongCS, Beavis PA, HaynesNM, et al. Anti-PD-1 antibody therapy potently enhances the eradication of establishedtumors by gene-modified T cells. Clin Cancer Res 2013;19:5636–46.

12. Junttila TT, Li J, Johnston J, Hristopoulos M, Clark R, Ellerman D, et al.Antitumor efficacy of a bispecific antibody that targets HER2 and activatesT cells. Cancer Res 2014;74:5561–71.

13. Laszlo GS, Gudgeon CJ, Harrington KH, Walter RB. T-cell ligandsmodulate the cytolytic activity of the CD33/CD3 BiTE antibody construct,AMG 330. Blood Cancer J 2015;5:e340.

14. Vilgelm AE, Johnson DB, Richmond A. Combinatorial approach to cancerimmunotherapy: strength in number. J Leukoc Biol 2016;100:275–90.

15. Fan G, Wang Z, Hao M, Li J. Bispecific antibodies and their applications.J Hematol Oncol 2015;8:130.

16. Kontermann RE, Brinkmann U. Bispecific antibodies. Drug Discov Today2015;20:838–47.

17. Spiess C, Zhai Q, Carter PJ. Alternative formats and therapeutic applica-tions for bispecific antibodies. Mol Immunol 2015;67:95–106.

18. Bacac M, Fauti T, Sam J, Colombetti S, Weinzierl T, Quaret D, et al. A novelcarcinoembryonic antigen T-cell bispecific antibody (CEA TCB) for thetreatment of solid tumors. Clin Cancer Res 2016;22:3286–97.

19. Heiss MM, Strohlein MA, Jager M, Kimmig R, Burges A, Schoberth A, et al.Immunotherapy of malignant ascites with trifunctional antibodies. Int JCancer 2005;117:435–43.

20. Moore GL, Bautista C, Pong E, Nguyen D-H T, Jacinto J, Eivazi A, et al. Anovel bispecific antibody format enables simultaneous bivalent andmonovalent co-engagement of distinct target antigens. mAbs 2011;3:546–57.

21. Rossi DL, Rossi EA, Cardillo TM, Goldenberg DM, Chang CH. A newclass of bispecific antibodies to redirect T cells for cancer immunotherapy.MAbs 2014;6:381–91.






22. Wolf E, Hofmeister R, Kufer P, Schlereth B, Baeuerle PA. BiTEs: bispecificantibody constructs with unique anti-tumor activity. Drug Discov Today2005;10:1237–44.

23. Shahied LS, Tang Y, Alpaugh RK, Somer R, Greenspon D, Weiner LM.Bispecific minibodies targeting HER2/neu and CD16 exhibit improvedtumor lysis when placed in a divalent tumor antigen binding format. J BiolChem 2004;279:53907–14.

24. Asano R, Ikoma K, Sone Y, Kawaguchi H, Taki S, Hayashi H, et al. Highlyenhanced cytotoxicity of a dimeric bispecific diabody, the hEx3 tetrabody.J Biol Chem 2010;285:20844–9.

25. McAleese F, Eser M. RECRUIT-TandAbs: harnessing the immune system tokill cancer cells. Future Oncol 2012;8:687–95.

26. Reusch U, Harrington KH, Gudgeon CJ, Fucek I, Ellwanger K, Weichel M,et al. Characterization of CD33/CD3 tetravalent bispecific tandem diabo-dies (TandAbs) for the treatment of acute myeloid leukemia. Clin CancerRes 2016;22:5829–38.

27. Rozan C, Comillon A, Petiard C, Chartier M, Behar G, Boix C, et al. Single-domain antibody-based and linker-free bispecific antibodies targetingFcgRIII induces potent antitumor activity without recruiting regulatoryT cells. Mol Cancer Ther 2013;12:1481–91.

28. Rossi EA, Rossi DL, Cardillo TM,ChangCH,GoldenbergDM. Redirected T-cell killing of solid cancers targeted with an anti-CD3/Trop-2-bispepcifcantibody is enhanced in combination with interferon-a. Mol Cancer Ther2014;13:2341–51.

29. Friedrich J, Seidel C, Ebner R, Kunz-Schughart LA. Spheroid-baseddrug screen: considerations and practical approach. Nat Protoc 2009;4:309–24.

30. Chang CH, Wang Y, Zalath M, Liu D, Cardillo TM, Goldenberg DM.Combining ABCG2 inhibitors with IMMU-132, an anti-Trop-2 antibodyconjugate of SN-38, overcomes resistance to SN-38 in breast and gastriccancers. Mol Cancer Ther 2016;15:1910–9.

31. Rossi DL, Rossi EA, Goldenberg DM, Chang CH. A new mammalianhost cell with enhanced survival enables completely serum-free devel-opment of high-level protein production cell lines. Biotechnol Prog2011;27:766–75.

32. Rossi EA,GoldenbergDM,Cardillo TM, Stein R, ChangCH.CD20-targetedtetrameric interferon-alpha, a novel and potent immunocytokine for thetherapy of B-cell lymphomas. Blood 2009;114:3864–71.

33. Brockhoff G, Hofstaedter F, Knuechel R. Flow Cytometric Detection andquantitation of the epidermal growth factor receptor in comparison toscatchard analysis in human bladder carcinoma cell lines. Cytometry1994;17:75–83.

34. Oberst MD, Fuhrmann S, Mulgrew K, Amann M, Cheng L, LutterbueseP, et al. CEA/CD3 bispecific antibody MEDI-565/AMG 211 activation ofT cells and subsequent killing of human tumors is independent ofmutations commonly found in colorectal adenocarcinomas. MAbs2014;6:1571–84.

35. Mittendorf EA, Philips AV, Meric-Bernstam F, Qiao N, Wu Y, Harrington S,et al. PD-L1 expression in triple-negative breast cancer. Cancer ImmunolRes 2014;2:361–70.

36. Hirschhaeuser F,Walenta S,Mueller-KlieserW. Efficacy of catumaxomab intumor spheroid killing is mediated by its trifunctional mode of action.Cancer Immunol Immunother 2010;59:1675–84.

37. Sapra P, Darmelin M, DiJoseph J, Marquette K, Geles KG, Golas J, et al.Long-term tumor regression induced by an antibody-drug conjugate thattargets 5T4, an oncofetal antigen expressed on tumor-initiating cells. MolCancer Ther 2013;12:38–47.

38. Hoffmann TK, Schirlau K, Sonkoly E, Brandau S, Lang S, Pivarcsi A, et al. Anovel mechanism for anti-EGFR antibody action involves chemokine-mediated leukocytes infiltration. Int J Cancer 2009;124:2589–96.

39. Spiess C, Merchant M, Huang A, Zheng Z, Yang NY, Peng J, et al.Bispecific antibodies with natural architecture produced by co-culture ofbacterial expressing two distinct half-antibodies. Nat Biotechnol 2013;31:753–8.

40. Klein C, Schaefer W, Regula JT. The use of CrossMAb technology for thegeneration of bi- and multispecific antibodies. MAbs 2016;8:1010–20.

41. Lehmann S, Perera R, GrimmHP, Sam J, Colombetti S, Fauti T, et al. In vivofluorescence imaging of the activity of CEA TCB, a novel T-cell bispecificantibody, reveals highly specific tumor targeting and fast induction ofT-cell–mediated tumor killing. Clin Cancer Res 2016;22:4417–27.

42. Blumenthal RD, HansenHJ, Goldenberg DM. In vitro and in vivo anticancerefficacy of unconjugated humanized anti-CEAmonoclonal antibodies. Br JCancer 2008;99:837–8.

43. Bluemel C, Hausmann S, Fluhr P, SriskandarajahM, StallcupWB, BaeuerlePA, et al. Epitope distance to the target cell membrane and antigen sizedetermine the potency of T cell-mediated lysis by BiTE antibodies specificfor a large melanoma surface antigen. Cancer Immunol Immunother2010;59:1197–1209.

44. Durbin H, Young S, Stewart LM, Wrba F, Rowan AJ, Snary D, et al. Anepitope on carcinoembryonic antigen defined by the clinically relevantantibody PR1A3. Proc Natl Acad Sci U S A 1994;91:4313–7.

45. Grupp SA, Kalos M, Barrett D, Aplenc R, Porter DL, Rheingold SR, et al.Chimeric antigen receptor-modified T cells for acute lymphoid leukemia.N Engl J Med 2013;368:1509–18.

46. Porter DL, Hwang WT, Frey NV, Lacey SF, Shaw PA, Loren AW, et al.Chimeric antigen receptor T cells persist and induce sustained remissions inrelapsed refractory chronic lymphocytic leukemia. Sci Transl Med 2015;7:303ra139.

47. Heiss MM,Murawa P, Koralewski P, Kutarska E, Kolesnik OO, IvanchenkoVV, et al. The trifunctional antibody catumaxomab for the treatment ofmalignant ascites due to epithelial cancer: results of a prospective ran-domized phase II/III trial. Int J Cancer 2010;127:2209–21.

48. Przepiorka D, Ko CW, Deisseroth A, Yancey CL, Candau-Chacon R, ChiuHJ, et al. FDA Approval: Blinatumomab. Clin Cancer Res 2015;21:4035–9.

49. Wu CY, Rupp LJ, Roybal KT, Lim WA. Synthetic biology approaches toengineer T cells. Curr Opin Immunol 2015;35:123–30.

50. Beavis PA, Milenkovski N, Henderson MA, John LB, Allard B, Loi S, et al.Adenosine receptor 2A blockade increases the efficacy of anti–PD-1through enhanced antitumor T-cell responses. Cancer Immunol Res2016;3:506–17.


Chang et al.




2017;77:5384-5394. Published OnlineFirst August 17, 2017.Cancer Res Chien-Hsing Chang, Yang Wang, Rongxiu Li, et al. Blockade Enhances the Antitumor Potency of T CellsCombination Therapy with Bispecific Antibodies and PD-1

Updated version

10.1158/0008-5472.CAN-16-3431doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2017/08/16/0008-5472.CAN-16-3431.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/77/19/5384.full#ref-list-1

This article cites 50 articles, 22 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/77/19/5384.full#related-urls

This article has been cited by 6 HighWire-hosted articles. Access the articles 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://cancerres.aacrjournals.org/content/77/19/5384To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-16-3431

http://cancerres.aacrjournals.org/content/suppl/2017/08/16/0008-5472.CAN-16-3431.DC1

http://cancerres.aacrjournals.org/content/77/19/5384.full#ref-list-1

http://cancerres.aacrjournals.org/content/77/19/5384.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/77/19/5384


Documents

Combination Therapy with Bispeciﬁc Antibodies and PD-1 ......conjugation between a dimer of anti-X Fab and a monomer of anti-CD3 scFv via the fused DDD2 and AD2 peptides. B, CD8þ