Systematic Drug Screening Identifies Tractable Targeted Combination Therapies in Triple-Negative Breast Cancer

Home , Elesclomol

Published OnlineFirst November 21, 2016; DOI: 10.1158/0008-5472.CAN-16-1901 Cancer Therapeutics, Targets, and Chemical Biology Research

Systematic Drug Screening Identiﬁes Tractable Targeted Combination Therapies in Triple- Negative Breast Cancer Vikram B. Wali1,2, Casey G. Langdon3, Matthew A. Held3, James T. Platt1, Gauri A. Patwardhan1, Anton Safonov1, Bilge Aktas1, Lajos Pusztai1,2, David F. Stern2,3, and Christos Hatzis1,2

Abstract

Triple-negative breast cancer (TNBC) remains an aggressive way. Combination therapies deemed immediately tractable to disease without effective targeted therapies. In this study, we translation included ABT-263/crizotinib, ABT-263/paclitaxel, addressed this challenge by testing 128 FDA-approved or paclitaxel/JQ1, ABT-263/XL-184, and paclitaxel/nutlin-3, all of investigational drugs as either single agents or in 768 pairwise which exhibited synergistic antiproliferative and apoptotic drug combinations in TNBC cell lines to identify synergistic activity in multiple TNBC backgrounds. Mechanistic investiga- combinations tractable to clinical translation. Medium- tions of the ABT-263/crizotinib combination offering a throughput results were scrutinized and extensively analyzed potentially rapid path to clinic demonstrated RTK blockade, for sensitivity patterns, synergy, anticancer activity, and were inhibition of mitogenic signaling, and proapoptotic signal validated in low-throughput experiments. Principal compo- induction in basal and mesenchymal stem–like TNBC. Our nent analysis revealed that a fraction of all upregulated or ﬁndings provide preclinical proof of concept for several com- downregulated genes of a particular targeted pathway could bination treatments of TNBC, which offer near-term prospects partly explain cell sensitivity toward agents targeting that path- for clinical translation. Cancer Res; 77(2); 1–13. 2016 AACR.

Introduction To help generate genetic predictors of drug response, the Cancer Cell Line Encyclopedia (CCLE) study compared pharmacological The prevailing standard treatment for triple-negative breast profiles of 24 anticancer drugs across 479 cancer cell lines, cancer (TNBC) is cytotoxic chemotherapy, but more than 70% including 56 breast cell lines (5), but only few predictors emerged, of the patients have a partial response only and have significantly including SLFN11 expression that predicted sensitivity to irino- higher relapse and mortality rates compared with non-TNBC tecan and topotecan. Genome-wide shRNA screen in 77 breast patients (1). Non-TNBC patients greatly benefit from targeted cancer cell lines coupled with genomic and proteomic data therapies, such as trastuzumab and tyrosine kinase inhibitors, but revealed that basal A cells are preferentially sensitive to effective targeted therapies are not available for TNBC. TNBCs are CAND1-NEDD8 depletion and TNBC patients may benefit from transcriptionally distinct compared with other breast cancer sub- NEDD8 inhibitors (6). A screen testing 130 drugs in 639 cell lines types and often have active EGFR, high activity of the PI3K from different cancers reported that single gene–drug associations pathway, frequent functional loss of RB pathway activity, and a rarely explain drug sensitivities (7). Drug screen in lymphoma and very high rate of TP53 mutations (2). Yet, EGFR inhibitors have melanoma have discovered multiple combinations for lympho- not proven effective in clinical trials, and synthetic lethality ma and mutant BRAF melanomas that could be selected for strategies targeting DNA repair pathway deficiency have failed to further clinical evaluation (8–10). Although broad combinatorial improve cure rates in TNBCs (3, 4). screens are crucial to identify and prioritize effective drug combinations for a thorough investigation prior to clinical evaluation, only few such screens have been reported for cancers. 1Department of Internal Medicine, Section of Medical Oncology, Yale School of To address these limitations, we undertook a comprehensive Medicine, Yale University, New Haven, Connecticut. 2Yale Cancer Center, New medium-throughput drug combination screen in TNBC cell lines 3 Haven, Connecticut. Department of Pathology, Yale School of Medicine, Yale with different and overlapping genetic backgrounds to identify University, New Haven, Connecticut. novel effective combination therapies for TNBC. In order to fast- Note: Supplementary data for this article are available at Cancer Research track clinical testing of potentially promising combinations, we Online (http://cancerres.aacrjournals.org/). systematically assessed the impact of 128 single agents in combi- Corresponding Authors: Vikram B. Wali, Yale University School of Medicine, nations with each of six drugs already approved by the FDA, Yale Cancer Center, 333 Cedar Street, New Haven, CT 06510. Phone: 203-737- resulting in 768 pairwise drug combinations covering a wide 6491; Fax: 203-785-7232; E-mail: [email protected]; Christos Hatzis, E-mail: range of targets and processes implicated in cancer biology. We [email protected] report anticancer activity of drug combinations based on overall doi: 10.1158/0008-5472.CAN-16-1901 growth inhibition and synergistic drug effects in six TNBC cell 2016 American Association for Cancer Research. lines. Finally, we used transcriptional profiles, protein expression

www.aacrjournals.org OF1

Wali et al.

and activity, and mutational data to characterize the drug-sensi- were performed as described previously (13). See Supplementary tivity patterns and mechanisms of action of the most promising Data for details. drug combinations. Results Materials and Methods Growth inhibitory effect of drugs as single agents Cell lines and selected drug panels Paclitaxel (microtubule inhibitor), erlotinib (EGFR inhibitor), TNBC cell lines (BT-20, MDA-MB-231, MDA-MB-468, BT-549, everolimus (mTOR inhibitor), vismodegib (SMO/Hedgehog MDA-MB-436, HCC-38, HCC-70, HCC1500, and MDA-MB-157) pathway inhibitor), XL-184 (VEGFR2/MET inhibitor), and crizo- harboring genetic abnormalities commonly observed in TNBC tinib (ALK/MET inhibitor) are the six FDA-approved drugs (panel patients were purchased from the ATCC, where all cell lines were A drugs) that we used in the MTS along with 128 panel B drugs in authenticated by short-tandem repeat profiling, karyotyping, basal-like (MDA-MB-468, HCC-38, and BT-20), mesenchymal morphology, and cytochrome C oxidase I testing. Cell lines were stem-like (MDA-MB-231 and MDA-MB-436), and mesenchymal obtained between 2013 and 2015, and used at passages 3 to 9, and (BT-549) TNBC cell lines (14). The dose response of panel A drugs cultured less than 3 months after resuscitation. For details, see in cell lines was first characterized in low-throughput screening Supplementary Data. experiments (Supplementary Fig. S1) to select a single dose of each drug. Single dose was selected as a highest concentration that Medium-throughput screen protocol was still IC50 across all cell lines and was also well tolerated On the day of the experiment, cells were trypsinized, harvested, clinically (Supplementary Table S1). This single dose was com- counted using Countess (Invitrogen) and deposited into 384-well bined with five doses each of 128 panel B drugs (Supplementary ViewPlate-384F microtiter plates (Perkin Elmer) at 750 cells/well Table S2) for MTS. Drugs were tested singly or as combination in 16-mL respective medium using a multidrop dispenser treatments in 384-well plates to assess growth inhibition and (Thermo) and allowed to attach overnight (Fig. 1). The following superadditivity as described in Materials and Methods (Fig. 1). day, single-dose 5Â stocks of the six panel A drugs were prepared Growth inhibition of TNBC cells by a single dose of each of the six in medium, and 4 mL of respective drug-containing medium was panel A drugs in the MTS format was moderate (Fig. 2A), as < added to each well (4 plates/drug) using the multidrop dispenser. expected for these doses ( IC50). The growth inhibitory effect of For panel B drug addition, a PlateMate Plus automated instru- the 128 secondary drugs (panel B), estimated by the area under ment (MatrixTechCorp) was used for pin transfer of 20-nL drug the dose–response curve (AUC) for each drug, indicated cell volume from 1,000Â drug stock plates into 384-well microtiter growth suppression by most drugs in all cell lines (Fig. 2B). TNBC cell plates. For details, see Supplementary Data. cells were more sensitive to genotoxins (GTOX) and apoptosis regulators (APOP), as indicated by higher AUC (red color) of most single agents in these drug categories (Fig. 3A and Supplementary Synergy assessment Fig. S2A). Single agents that suppressed growth with AUC > 75% Synergy was assessed as the deviation from the Bliss non- across TNBC cells clearly show enrichment in GTOX and APOP interaction model (11) as well as by Chou and Talalay isobolo- categories (Fig. 3B and Supplementary Fig. S2B). Basal-like sub- gram analysis (12). Best combinations screened by medium- type of TNBC reportedly exhibit higher expression of cell-cycle throughput screen (MTS) were validated in low-throughput for- and DNA damage response genes and a higher sensitivity to DNA- mat. See Supplementary Data for details. damaging agents (14), while we have reported resistance of mesenchymal stem-like (MSL) cells to EGFR inhibitors, including Flow cytometry lapatinib, and BEZ-235, a PI3K–mTOR inhibitor (15). Indeed, Â 6 Cells were plated at 1 10 cells/well in 100-mm culture plates basal like-1 HCC-38 and MDA-MB-468 cells were most sensitive, and allowed to adhere overnight. Cell lines were exposed to while MSL MDA-MB-436 and MDA-MB-231 were least sensitive respective control or treatment media for 24 hours. Cells were to these drugs. P53, a prime regulator of apoptosis, is mutated in stained using the BD Pharmingen Apoptosis Detection Kit II most TNBCs (16). All TNBC cell lines in our screen carried a TP53 according to the manufacturer's protocol (BD Biosciences). See mutation and, interestingly, were highly sensitive to APOP drugs Supplementary Data for details. (Fig. 3B and Supplementary Fig. S2B). Specifically, APOP drugs YM155, bortezomib, and carfilzomib displayed most potent Reverse phase protein arrays anticancer activity as single agents across all TNBC lines, as Cells were plated in 100-mm culture plates (4 plates/cell line) indicated by higher AUC (red color) in Fig. 3A. YM155 is a that were exposed to DMSO control, 1 mmol/L crizotinib, 1 mmol/L selective suppressant of survivin, a member of the inhibitor of ABT-263, or their combination. The experiment was run in dupli- antiapoptosis (IAP) family that has higher nuclear activity in cate. After 24 hours of treatment exposure, cells were collected and TNBC compared with other subtypes (17). Interestingly, MDA- washed with PBS and 48 cell pellets corresponding to all samples MB-231 TNBC cells were more sensitive to YM155-induced were shipped to RPPA (reverse phase protein array) Core Facility at growth inhibition than SKBR3 or MCF7 non-TNBC cell lines the MD Anderson Cancer Center (Houston, TX) for analysis. (18). AKT inhibitors KP372-1, dactinomycin, digoxin, and trip- tolide were a few other highly potent drugs that similarly inhibited Western blot and immunoprecipitation growth of all TNBCs. Whole protein lysates were prepared, and protein concentrations were determined by the Bradford assay. Phospho-tyrosine Growth inhibitory effect of pairwise drug combinations immunoprecipitations and standard blotting procedures for gel The inhibitory effect of 128 drugs individually and in combi- electrophoresis and immunoblotting using PVDF membranes nation with paclitaxel, erlotinib, vismodegib, everolimus,

OF2 Cancer Res; 77(2) January 15, 2017 Cancer Research

Novel Combination Therapies in Triple-Negative Breast Cancer

Diagram of medium-throughput drug screen in TNBC

Day 1: Cell deposition

750 Cells/well in 384-well plates CellsCells

Day 2: Drug treatment

Panel A Panel B 6 FDA-Approved Drugs X 128 Experimental drugs (Single dose each) (5 doses each)

Day 5: Read assay signal output

Figure 1. Diagram of medium-throughput drug screen in TNBC.

CellTiter-Glo

Luminescence

Superadditivity Output luminescence Signal by plate reader Inhibition Drug A + B (actual) Quality control Drug A + B (Theoretical Bliss Sum) Drug B Data analysis Drug A Growth curves (To estimate Growth Inhibition & Superadditivity) Drug B concentration crizotinib, or XL-184 was assessed by the AUC for each dose– FLT3, FLT4, and ALK, but express MET, KIT, EGFR, ERBB3, response curve and is shown across cell lines in the heat maps MST1R, and AXL. AXL expression is variable in each TNBC in Fig. 3A and Supplementary Fig. S2A. More than 90% of the category,whileKIT,EGFR,ERBB3,andMST1Raredownregu- combinations tested exert antiproliferative activity while the lated in MSL and mesenchymal cell lines, which may contribute remaining have either no effect on cell growth or a slight to their lower sensitivity to drugs. Interestingly, MET is proproliferative effect (Supplementary Fig. S3). Global com- expressed by all TNBC categories. parison between cell line sensitivities in this screen reveals Combinations with erlotinib were particularly effective in BT- overall lower sensitivity of MSL MDA-MB-436 and MDA-MB- 20 and MDA-MB-468 cell lines, while MSL MDA-MB-231 and 231 versus basal-like TNBC cell lines (Supplementary Fig. S2A MDA-MB-436 cells were the least sensitive (Fig. 3A and 4B). We and S2B). Because mesenchymal and MSL TNBC cells often have previously reported that MSL cells downregulate EGFR and have downregulated RTK expression, we examined the expres- become erlotinib resistant (15). MDA-MB-468 cells express the sion of RTKs in the CCLE portal (http://www.broadinstitute. highest levels of EGFR followed by BT-20 (data not shown), which org/ccle) and observed that TNBC cell lines express very low may partly explain their higher sensitivity to drug combinations levels of many RTK genes, RET, TEK, PDGFRA, PDGFRB, FLT1, involving the EGFR inhibitor erlotinib. The effects of drug

www.aacrjournals.org Cancer Res; 77(2) January 15, 2017 OF3

Wali et al.

A Growth-inhibitory effect of each drug from panel A BT-20 BT-549 HCC-38 100806040200 100806040200 100806040200

CrEr Ev Px Vs XL Cr Er Ev Px Vs XL Cr Er Ev Px Vs XL

MDA-MB-231 MDA-MB-436 MDA-MB-468 100806040200 100806040200 100806040200 Percent inhibition

Figure 2. Effect of drugs from panels A and B on the growth of TNBC cell lines. A, Percent growth inhibition of TNBC cell lines by single dose of each of the six panel A drugs (Cr, crizotinib, 1 mmol/L; CrEr Ev Px Vs XL Cr Er Ev Px Vs XL Cr Er Ev Px Vs XL Er, erlotinib, 3 mmol/L; Ev, everolimus, 30 nmol/L; Px, paclitaxel, 3 nmol/L; Vs, vismodegib, 10 mmol/L; XL, XL-184, 3 B Growth-inhibitory effect of each drug from panel B mmol/L) from the medium-throughput drug screen. B, Growth inhibition by BT-20 BT-549 HCC-38 128 panel B drugs as single treatments (2.4 nmol/L–10 mmol/L) in TNBC cell lines; vertical bars indicate the area 0 0 0 under the individual dose–response curve (AUC) for each drug. 60 8 40 0 2 20 40 60 8 0 2040608 0 0

–20 1 50 100 –20 150100 –20 150100

MDA-MB-231 MDA-MB-436 MDA-MB-468 80 80 80 0 0 0 6 6 of each drug from panel B 0 0 4 Area under dose-response curve 20 4 20 40 6 0 0 020

–20 1 50 100 –20 150100 –20 150100 Individual drugs from panel B

combinations grouped by categories and visualized in the chord basal-like HCC-38 and MDA-MB-468 being most sensitive and plot of Fig. 3B also suggest drugs in the GTOX and APOP MSL MDA-MB-436 cells least sensitive, but displayed similar categories as most potent suppressors of TNBC cell growth, while sensitivity to APOP drug combinations. However, combinations drugs in the glucose/lipid metabolism (GLM) category have the with GLM were least effective and therefore absent from the chord lowest potency overall (Fig. 3B and Supplementary Fig. S2B). Cell plot for drug combinations with AUC > 75% (Fig. 3B; Supple- lines displayed differential sensitivity toward GTOX drugs with mentary Fig. S2B).

OF4 Cancer Res; 77(2) January 15, 2017 Cancer Research

Novel Combination Therapies in Triple-Negative Breast Cancer

Cell line sensitivities to single agents and drug combinations Chord plots depicting cell line AB (AUC in each dose–response curves) sensitivities to drug categories

Single agents

HCC.38 MDA.MB.231

MDA.MB.436

BT.549

MDA.MB.468

BT.20

RTK

PI3K/AKT

APOP

Other

NRTK

CCR GTOX GLM CKM

Drug combinations

HCC-38 MDA-MB-231

MDA-MB-436

BT-549

MDA-MB-468

BT-20

RTK

PI3K.AKT

Other APOP

NRTK

GTOX CCR CKM

Figure 3. Patterns of activity of drugs grouped in different categories across cell lines. A, Heatmap summarizing the results of MTS by the AUC for each dose–response curve for single agents or pairwise drug combinations. AUC gives a single estimate of growth inhibition over multiple doses for each curve, with higher red color intensity indicating higher AUC. The first column set indicates response of six cell lines to 128 single agents (panel B) as rows, arranged in categories according to the drug's target or mechanism of action. The subsequent six column sets indicate growth inhibitory response when single agents are combined with either paclitaxel, everolimus, vismodegib, everolimus, crizotinib, or XL-184 (panel A), respectively, in the six TNBC cell lines. B, Chord plots display the drugs in each category that suppressed growth with AUC > 75% as single agents (top) or drug combinations (bottom) for each of the cell lines. In each chord plot, the upper semicircle reflects the relative sensitivities of cell lines toward various drug categories, while the lower semicircle reflects the relative efficacy of drug categories; circumference length indicates the number of drugs achieving AUC > 75% in that category.

Association between drug sensitivities of cell lines and performed unsupervised principal component analysis (PCA) to expression of drug targets segregate the cell lines according to the extent of upregulation or To explore potential links between the drug sensitivities of cell downregulation of all targeted pathways (Supplementary Fig. lines and their genetic background, we compared the overall S5A). Interestingly, pathways targeted by drugs in the RTK and response of cell lines to drugs with expression levels obtained NRTK categories appear to be highly upregulated in HCC-38 cells from the CCLE portal (5) of the 170 genes that are direct or (Supplementary Fig. S5A), and these drugs appear to be largely indirect targets of the 128 drugs included in the MTS (Supple- ineffective against these cells, either when used alone or in mentary Fig. S4). Although sensitivities differed between cell combination with other drugs (Supplementary Fig. S5B). Lack lines, the expression patterns of drug targets appeared broadly of upregulation of the same pathways in BT-549 cells appears similar across cell lines (Supplementary Fig. S4). We evaluated to be associated with sensitivity to these agents. Upregulation these effects more systematically at the cellular pathway level by of GLM targets is associated with greater sensitivity to GLM grouping the proteins targeted by drugs in each of the 9 categories drugs in BT-20, while downregulation of RTK genes was associ- into "targeted pathways" and assessing the expression of genes in ated with reduced sensitivity to RTK, NRTK and cytokine mod- each targeted pathway across cell lines (Supplementary Table S3). ulator (CKM) drugs in MDA-MB-436 cells (Supplementary Speciﬁcally, the degree of upregulation or downregulation of a Fig. S5A and S5B). Accounting for drug-to-drug variability over targeted pathway was determined for each cell line as the per- drugs within the same category, the observed patterns suggest that centage of the pathway genes found in the upper or lower tertiles upregulation of the targeted pathway as determined by a high of expression relative to all CCLE breast cancer cell lines. We then proportion of highly expressed targeted genes, is generally but not

www.aacrjournals.org Cancer Res; 77(2) January 15, 2017 OF5

Wali et al.

always associated with sensitivity to the targeted treatment in signaling as two major cellular signaling epicenters critical to these TNBC cell lines. Collectively, these results suggest that the survival and proliferation of TNBC cells. Alternatively, RTKs fraction of all upregulated or downregulated genes of a particular could also be activated as part of resistance mechanisms targeted pathway as a surrogate index of overall pathway activity in TNBCs. ABT-263, a BCL2 inhibitor (apoptosis inducer), may partly explain cell sensitivity toward agents targeting that synergized with RTKs erlotinib and crizotinib. Erlotinib also pathway. synergized with 5-FU, an antimetabolite commonly adminis- tered systemically in many cancers. Apoptosis inducers, ABT- Top synergistic combinations 263, Nutlin-3 (MDM2 inhibitor), STA-4783 (oxidative stress Drug combinations where two agents at lower doses act syn- inducer), and JQ1 (bromodomain inhibitor) also synergized ergistically are of particular interest owing to their potential to with paclitaxel, which is a part of combined chemotherapy improve efficacy and moderate the toxicity associated with high regimen with cyclophosphamide and adriamycin in breast doses of single agents in the clinic. We used the Bliss model of cancer. Nutlin-3 inhibits the interaction between MDM2 synergy (19, 20) to assess superadditivity. Sham combinations in and P53, stabilizing P53 and inducing apoptosis, and has which the same drug was present in both drug panels were shown potent preclinical activity in several cancers (22). STA- evaluated in the same MTS protocol as controls and yielded 4783 or elesclomol triggers apoptosis in cancer cells by inducing almost zero superadditivity, as expected (Supplementary Fig. oxidative stress and is being tested in trials for various can- S6A). Most drug combinations tested were antiproliferative, but cers (23, 24). JQ1 is a bromodomain (BRD) inhibitor. BRD only a minority displayed superadditive effects (Supplementary induced oncogenic signaling appears to be pervasive in all Fig. S6B). We selected top synergistic combinations as those breast cancer subtypes (25); disrupting BRD function in bas- having superadditivity (area between combination and Bliss al-like breast cancer suppressed tumorigenesis (26), and TNBCs dose–response curves) 10%, inhibitory potential (area above appear to be preferentially sensitive to BRD inhibition (27). We the curve) 75%, estimable EC10 concentrations for the combi- have recently reported anticancer activity of JQ1 in drug com- nation and theoretical Bliss curves, maximum growth inhibition binations in vitro and in vivo (28). Interestingly, JQ1 was also by drug combination 80%, and superadditivity observed for at found to synergize with RTK inhibitors, erlotinib, XL-184, and least 3 consecutive doses of the panel B drug used in the combi- crizotinib. nation. These top combinations are indicated by green lines in the YM155, bortezomib, carfilzomib (APOP), KP372-1, and trip- circular synergy plots in Fig. 4A (listed in Supplementary Table tolide resulted in complete growth inhibition of all TNBC cell S4), while gray lines indicate second best combinations where lines at low nanomolar concentrations as single agents and also in superadditivity was observed for at least 2 consecutive panel B combinations. The lowest two or three doses of these drugs drug concentrations and had maximum growth inhibition already achieved 100% growth inhibition, but the data points 50%. These represent novel synergistic drug combinations that were inadequate to reliably assess synergy. Nevertheless, our can be potentially effective in TNBC and therefore merit further observations recognize and emphasize the potential of these investigation. Summary of the top combinations (green lines) drugs to benefit TNBC patients when given alone or in combi- by panel A drugs and cell lines indicates higher number of nation with other agents. synergistic combinations in MDA-MB-468 and BT-20, frequently containing erlotinib or XL-184 panel A drugs (Fig. 4B). Isobolo- Validation of MTS results in low-throughput experiments gram-based synergy analyses (12, 21) of representative top A large number of variables can affect high/medium- combinations also demonstrated synergistic growth inhibitory throughput assays, and lack of standardized protocols adds to effects of combinations (Fig. 4C). Combination index (CI) the likelihood of discordant and misleading results (29). Our values were less than 1, indicating a high level of synergism MTS was conducted in duplicate plates following rigorous that resulted from combined treatment (Supplementary Table quality control and data filtering steps as described in Materials S5). The dose reduction index (DRI) confirmed a high level of and Methods. Yet, to assess the reproducibility of MTS results, synergism that resulted from combined treatment, e.g., the dose we retested two of the top synergistic combinations, ABT-263 of nutlin-3 or paclitaxel could be respectively reduced 34- and with crizotinib or paclitaxel, in 96-well format across all cell 2-fold and still produce the same antiproliferative effect in lines and compared with the medium-throughput results HCC-38 cells if combined together (Supplementary Table S5). (Fig. 5A). Paclitaxel (3 nmol/L) or 1 mmol/L crizotinib inhib- As the cell lines used represent distinct mutational and expres- ited <50% of cell growth, and 2.4 nmol/L–10 mmol/L ABT-263 sion backgrounds, it is not surprising that none of the drug when given alone generated a sigmoidal dose–response curve combinations was highly synergistic across all cell lines. How- in sensitive cell lines or a flatter curve in resistant cell lines ever, it is worth noting that top combinations indicated in the similarly in both formats. Also, superadditive inhibitory effects circular plots may also be synergistic in multiple cell lines, of combinations were observed, as indicated by the dose- albeit to a lesser degree but falling below the threshold for top response curve for the combination (black line) being above hits in circular plots. Combinations of the NFkB inhibitor Bay- the theoretical Bliss sum (green line) in same cell lines in MTS 11-7082 and erlotinib, and WEE1 kinase inhibitor MK-1775 as well as low-throughput experiments. Broadly, growth inhi- and everolimus had mild synergistic activity in all TNBC lines. bition and synergy patterns observed with two combinations Superadditivity between Bay-11-7082 and erlotinib was high across cell lines in 96-well plate low-throughput experiments enough to be indicated by gray lines in BT-549, MDA-MB-436, with 4 replicates per group were similar to MTS results with 2 and MDA-MB-468 circular plots (Fig. 4). replicates per group, validating medium-throughput findings. A striking observation was that most synergistic combina- The entire MTS dataset used for analysis is provided as Sup- tions contained drugs inducing apoptosis (APOP) or receptor plementary Table S6, while the low-throughput validation tyrosine kinases (RTK) or both, suggesting apoptotic and RTK dataset is provided as Supplementary Table S7.

OF6 Cancer Res; 77(2) January 15, 2017 Cancer Research

Novel Combination Therapies in Triple-Negative Breast Cancer

A BT-20 BT-549 HCC-38

2 1

1 i 8

e l

e 1 c e i

b 9 0

l b 0

i n

n c i

4 9 a n

i b

0 t

n 9

i 7

t i

e 2 7 i i 299 4

i l r

r e

1 0

a -

r 5 t 9 b i - 7

7 2 n

b x

0 l b

3 i 5

n r l

e u

l 3 - e

0 m l 3 c

1 i 2

e 3

3 m 1

3 8

n 1

u n e

3 x

in e e

s i e

1 y 1

1 n 5 u

t 2 i e

1 s a 3

3 1

5 2 u r

l e 8

5 3 7 r 4

r 8 i t 3 l

- l o r s x i

0 a 1

l o - 1

8 c

5 e

a 2 3

c t n

1 o

m 2

8 t

-03 m

8 o

7 y 3 5 0 a 8

o - c 2 p r

o i

2 i

e a 3

y o a

z y r c l

P - 9 n

l - 7 D

z - - 2 0 t

o 5

g 0

1 o

7 u

3 n

y a 2 8 o

d D i u

d z

2 n 2 a

X i 5 -

o p l u

a 2 i m c a 3 i l r

i e

V o T i

0 2 b

c u

N B -

P l f

0 1 D 8 5 e

a b

r l 8 i

H l o i

t 7 o

B c 2

i n

- u C

T BX 5

PD I- a

B V X - -

M B f P a I D 5

C T

1 t

V o

L i T

B V

d 5

N e 9 P 2

V S

P r

B n

- 7 i

b T u

p o

o a 8

N T i S

c a

D P d b

B K N

x r 2 P 1 7

r - 8

n H

a l

9 i A

s 1 V

M p

o c

i N

n -

C a - r P

i L

a l 7 9 C o I o r

G h

8 c i

o A t C

M v H N t

n M

t N o

r N

m B

e t 1 2 d D u

3 p o

a K

p T

C i

a a L l e

c M

G a - - n

c N

l l x

- r e i

f 7 o I

c 0 o 9 i

D h -

G m 9

- t M a

c K

F m M

7 v t 8 T T p

o G

o 8

o y

D e G a

B 8 3

l e t M o 3

b G

o D

i H

m 7

1 m

a v

r R c 0

G A B

F e 7 0

l T e

e -

6 c

o - -

i G

o 7 o a a

e D 7

d y

B I

t l

i 3

i 2

D i

- i 8

l H D

c 8

C I D n

s F

5 r

m 1

R a

a 1 o

- R b

o 7

8 B 2 76 l 6 t

- D e 6

t B a

0 p 7 g

p D

D 7 3 2 B

I i - t

- i - 3 2

l I O 8

0 m n

o 5 6 I

t B n 6 c 5

M a

1 r 8

D B 2 7 e

r B

a 2 7

2 n

5 x - M 1

M 7 2 K

m Z C

K P - I c

a 7

5 2 a - I

C 2 C e -

O e

h A c

0 B B

M C 1 D t B

o e

e D

G R 4 C 5

K o S

n Z 5

i M C

4 i Z

5 3 e p

S m C

C o 5

c 3

A z C

Y 6 8 n

x 1 o A p o

e R i y

i 1 8

3 a 7 i r

o z

T n e

l p R

H 7

M d

i 2

4 i M 2 n a

F e n

c a

i - 2 4 p

r i o

t Y 2

p -

r Y

P o d 0 e a

i A

a t

o 0 i y

l A

4 T p

a r T 4 m u

M p

l X T

t o

i 4 d

y T

b f

l 4

S i

e 3 6

h S

l r

- x r 6

t X

x O 3 x

i O

x 3 a o T

n o e a

i r l

r F l r

l c u F O l t

t 8 T

o c

r O P f T

e o

u i 3 P

u t o r

l t

v 6 8

m a N i

T a

n T

a 3

e b G

n 6 b

C0 T 3

i t

c 3

f 0 O -

l a O - N

n t

o t 4

e C G i n

t l

U i c

o 2

e t l

o a

N 9 r u t s t l

p X u i r

n i

U u

o i s

o 4 v

- - v

c N

T a N u

o a

N n

e i

n - n

i V

O L r i y

-g f

in l f o

x l

p M e 4 o e 4

- V

a 2

i 8

i e

0 y N 2

z N r 9

T 1 x

3 d 9

4 4 -

r 1 - 4

2 - 8

N 1

L 1 de

p N

0 -

4 L

o 8 L

M 2 4

G A

u 0

l M

f 1 M

0 R

r 0

1 n

i 0

T 4 3

- - 3

1 a t 8

R 1

X s 3

n L

a i

6 B L

t 8

0 A M 8

H a

t 0 M

C 0 o

s 1

t 1

N 1

A A A 1

va 4 U

lin r - -4 o

P e t X X n

b A B

le B

t m H u

s P H

r P n

o E

r n

O l i n

o i

e u

e n el el z s

i e b b o

o b O l C

a m f O i

u t m

l i E g

g c E n

- a i

e z

p D

P D l e y

i n

D l

x i ne

o b bi

e Everolimus Erlotinib Everolimus Erlotinib a Everolimus Erlotinib p

d P t a - i P t

a c i

1 e i ec

- D m

r D

D o f

P t e

M M n

1 i

4 C

o 1 f

0 1

t - -

A e

- D D

M P P

A CM M M t C b a

i t s

i m l

o M r

filz O

Car ib L

at ib t

n m i

i t l o m

r a

t z o

O l

s i z

a rf il v f

r L a r o

t C a

A C 2 4

2 9

b P

ezomi G 2

t 4

r 2

o 9

2 ib

B 4 b

F m i

P o m

G Bortez ortezo

n B i

l e

r u

e C

e n o z

a t i l i g s o

263 R - e

n T o z a i t B l g s i A o

R ABT-263 ABT-263

Simvastatin

Dapagliflozin M

m i v a s a t i t

L 17-DMAG n 17-DMAG 17-DMAG

B o

t i

i b

M e

t f

m i n

B o

G u t

i n i

WZ4002 b WZ4002 WZ4002

i s

t a t

Vismodegib Crizotinib C Paclitaxel Crizotinib Paclitaxel Crizotinib I

1 T V

3 iv a -

7 a nd

V 4 ntinib etanib

a C

nd Y e T

t -

an 3

ib 8 7

T S T 9 - i

3 v

2 u a 8 n

4 n 7 i

2 tin ti

ib E nib

T a

iv v i

a r

n e

tin n

ib E

S K

i S

m S

v o u

v n

r r e

a a it n

I i

s f

K n

e m

ta n ib

i a

t b t i T n

P i

H n

A b

6 m

T P

B 6 P

5 a H t

o H

7 i A n

s A 5 J

u 2 b 6 R

6 N

t 6

i 6 K

n 5 5

i 7

b 7

P i 5 n R 5

D J 2

2 h

1 i P

7 P b

K D R

C D

t 3 N

o 1

i I 0 1

P 7 7 7 F

h 3

4 3 I

N ib 0

1 0

4 7

3 7

R t

P 4

- O 4 o 1

7 O F T

SI r O 3

- S 9

I S

- 4

C 9 I I

3 - -

0 F 9

Y 1 9

K 6 0

K T

0 -

3 K T

6 5 6

- 7

3 N

F N

N 3

8 V

V - V K

S T 7

P P 2 P

7 t

8 -

- - a

3 T

f T T

a t

2 A S

A A t

v L i

2 L

c E R

E t E

i L a 8 a

r 6 a 6 6

e a p

8 t p 8 8

t a

n XL-184 Paclitaxel p Vismodegib XL-184 a Vismodegib XL-184 4

4 4 y

I c

z a t ti m

i k n

N t n

i S F i n I

a F i b

b n

i y o

t b T F o h i

n r

r i

o r e

i e

I a

b r t

n J

t i i

t e T

i m n

o N

t h Er n E

r i

i r

i b

i e

n a V

K r

b b l

i l t o E m

b o e

o t

n r

t mu i e i P

l V r i n

b D

h o n

t D i

F e

t i b i i

n o

b i b

o r

n m -

i a v i

D -

t i v

b n i o

u t 3 i

o 5

t e D P i o

r D n

r i

v -

n n

D a F

n i

I a 3 i

i i b

t f

y b b s

in o e -

9 s

D i

5 a

o n C

n a

6 a i C o S 7 t b I P

E i

y t i

b r n

s i 3 x g 3

3 r

n y l t i

o i z

a a i i I

o z

i K b E

C K r

3 C o

t x b

C T 2

o y x

i r

H / t

T n

K r /

i p H

i t i

e A -

e H i t

i n i n

z r T r

/ b n h

a -

h K E A

y i

K a B

C o - t

8 z r

t B i 5 b

b T p 5 E K T

m t n 1 O

a M 4

V H M t

T i 4 h

n a S

4 z n s 2

e - 2 B S

i m B

B a S 6

b t 8

5 G

t 8 a

m S

9 I -

4 i s

i I - 0 5

M 4 o B

n 0 7

B u

G A

7 3 A

t 2 x

u 2 A - e

i 5

a 2 9

B r W H

a 5 6

b S W

i Z

r 8

A 4

u Z 3 n

4 l

I a

o - A

0 H A D - 8

T r

B x 6

7 N D - 8 J t

f 2

2 e A a 2

i 0

e Z n

5 y Z 0 - g

W K 9

J E

T -

y T

9 r 8 7

Z a T D d

4 8 7 1 J

o D 9

6 9

A t

i - e 9 e e 1

D 8 r 9 Q

b R

J 2 x

R x - E 2 1 - 2

o 3

M m

x Z 0 3 4 m 8

1 y

i 4

M - 9 Q a

n x 1

8 7 M a 1 M 5

M c

D 4

G N 8

- 5 s

9 S p y t s

1 p 4

9 h

K N

r K 4 K i

2 4 K

- S K

c i a r O

- V y

A K H

a 7

8 a r

4 3 7 l o 2

- o h G

P O

p G o P

V G m

- -

P P

- P

J 1 m

N 2 5 y

l l

r 2

2 G

P 2 l

G G i i

3 o

t 3 o P

a i 4 K S

S l m

M 2 3 4 i d S

P B y

m 5

a T

i B 2 P

B 7 y

2 p

N q

7 i m

g 7 L

P Z D

0 Z N c R 7 7 o D - D

r Z K

K o

u c

R 0

O K u

5 X

N X EZ R 6

V 0 e

S V

L E

u X

E V

P T

1 T o 2 T T o D 2 T i

P T

6 T

2 m i i q

X i i C n e

r 6 s

i U 6

C i 1 n C 6 b

6 6 n i

b s r A

b i A 8

x - U

i m -

- li t

Z n e

Z h

r i h r

i r ipi

u - i

E t N l

n s r

e 1

o 9 1 P-

p s 9 1

2 p o

9 y -

i o 1 o i y

o i

- e

4 o - -

l V

1 V r E i

- i

- p p 0

m 7

2 o

n 4

y m 0 0 o

0 n 2 0

c o 2

e 7 U c t

o i 0 a n

2 7 t 2

o t

2 f fa e

t t

2 t e 9 4 e 2

e 9 i 6 2

x 9 3

h r

6 a

r a

5 6 o

o g o p

3 4 9 3 9

o n a

m o

i n 2

l i a 4 6

n 5

n 41 g 4 9 7

l 9 5

v rn d

o n 9 r

l 5 u l a

d 5 6 l 3 3 v

or d

l i

5 i i t 1 n

t 5 u

a y t 1 3

3 i

d 9

46 3 d a d

n m a

E m 5

i 9 a 9

e a r

r i

t i 4 c

m oq

e 0 e

e e

z b b n g

4 n

2 g n

g i

i i

i n

54 9 i

n l u i

2 i

ib n

b n u

5 e e

e i

e n e

MDA-MB-231 MDA-MB-436 MDA-MB-468

2 2 2 l 1

8 1 8 i 8

e 1 l

e 9

c 9

0 0 0 b

l b

9 ne n

i n b

9 c i

2 i

4 i n

a i t l 9

7 t 2

n 4

t 9 i l

r r

- a

i -

- 2 4

n r

2 bi

5 l

b e 2

e r

i lb

0 l r n

2 5 e

l a

u e 1 3

i 1 3

1 0

m e

x m 3 e

c 3

e r 1

t m

e 0 u n e

3 x i in

8 3 7

1 o s

3 i

1 13

1 s 1 u 5 x 3

u r

0 a

s e

re u 3 7

t 3 a

n u 5

0 a -

8 -

3 - t 3 n

5 n

l a

2 5

3 5 a 03

2 7 0 5 i c t a n

o i r

i o

a y 3

8 e 0 a

o o c 2 i 8 o

l - 5 2 i

y t y y rc m

8 - 2 0 r l

p z

c 5 l

z -

e i 2 -

3 r D 2

2 o 5

- z n

n -

l n 2

n l 8 8

c D

u 2

o i X

a - 8 a

i a X i

u 8 5 5 p

m u

g P o

i y X o c D l o

7 e

- l

d 2 D o

u 2

2 u

o P l

o -

c n o

i D 8 e f

5 8 a 1 8 l

b 9 M n

D t 7 2 o

P n

B B

u V

B a

B p 7 B V -

i V

i - - T C 8

f B

V l T 1

I o T m

d i

l P

a 8 d B 9

T 7 M V

I n

a B

r 2

S - N i P - 7 i

1 P b

f H

P o e I

o S t 5

V in

T d o P r

- r b

o N

9 l r 2

c N o a

i -

7 -1 n H i t

P S L i 1

n I

t i

A N

N d b -

H i

a r

p N -

2 i e L

c a B l

i r I

1 5 N c i i r

A o

p t

g K

C - T N

L i r r

h -

C M x

o N I i l p e

a K

A p

a o 9 T n C

D t

c M a c a

o a

K p e o M G x

M T t v 8 o r n

t o 9

e i D

h C

i G E

x t M p

r a

o m v

m 3 8

9 a e C

c G t c

M G o

T a l a

v o G Da

e a

n 8 c 7 0 G 3 o t t

y e

i l n

h F -

E m

a i

G 3 7 m 7

d l e o c 0

E F

T o m -

a C o

B D

e c 7 8 7

l F 0 o

- D

o - n

x 1 D i

I o 8 p t G

R H

i i D i

e D 6 l B

T m

87 o D R 1

D c

B 3 2 -

l 6

a -

e o

M 1

R E

7 D H I

5 6 t B e p x 2

x 3

6 i -

in i

B D B

l I

t B 2 7

e 6

2 m 5

- 3 -

- r

I a 7 7

t o 6 I 8 B 2

n 5

1 - e

M 8

x M - i 7

K 7

7 M e a B

C 2 p B K I 2

7 - i

P K D C

- 0 l

c 7

I C 5 O

M 0

- m B

8 D

a Z

e C

B 5 8

t 3

B e C 5

D x Z e C C

A n

1 S 1 o 8

n C 5 3 P

5 2

Z a

i O R A e

2 7 p C

5 a 1 8

3 c R

e M 2 z

o A

4 B 7

i n R 1 8 e

a - 2 B

i Y M

t 2 e

T 4

7 S t

d A 0 - 2

i M

e 2 Y

a 4 o

4 e

i 0

- po A

p 2 T

r m Y

p 4

e 4

o o

A 0 e T

S n

6 z

i 4 i

u x h 4 n T

X e i l

t T

a S

T 3

z 6

f 4 a r

n x

p l e

i o

S F ap 3 ri

r e

6 i

r c n a

a X x i l i

O r i

o F

3 e P o

a d c

i t T o

u l

X m

F p

l P

T o

f c a

T u

o t

l O

3 P

h f

r b

o u a i t

t l 3

h f

T O b - o i t

0 a O

r T

G n T 3

b o -

C i O o

t r

3 T tl n

T u s a

- o

N O i

r l

t l

o u i

n u

8 t i

s l

U i

c v G

r r

l f

3 u o i i N u

t 8 a l r T

6 v n r

u 3 G i i N g n

T i

- a

N v 6 lf n o C

D a 4 i - e f

t V

n N l

y i 2

f N a 4

x e t e

l 9 U

N 2 t

s o

e 4 4 s N

U 9 o

8 - e o

2 t

e 1 N 4

s n

d 9 N -

u i s

l o

4 r o L

2 - N n

c - i

1 L r u M

N V l

0 D

L g M

A 0 -

M 3 x 0 D

1 o

n 3

t A

e 1

x L a 30 d

8 t

1 8 - L s

M 1

0 2

4 a 8

L d 1 M

4 v

- 0

r 8 1

1 2 1

M 1 A

o 0 4

0 t

1 4 - A 1

4 4

A A 1

X - -

4 1

- B R X n

X H A B -

n P n

B i R H l

t P

u a

H A

r P in

l s e

a n

i e C i

v l

n b r

i a e t

l O o

e m t b

n O

i b E

O v m

m E r m o

o E f e A

n n

i e

e i

b n n e

M e i

n P ta l b

i i u

Everolimus Erlotinib b Everolimus Erlotinib c r Everolimus Erlotinib ta e

n P i

a i e

C c

P it n

e D e

ec l D t ru

t D

s C i

r i

o l

1 f

- i l

M D g -1

1 P a

- p D

n a

D i M P

P o C l

f M 2

M l

4 C g 2

C a

p 2

L a 4

P b

i i b

m m i r

b o o M

f m

i lz t o

om rfi e lz

M fi

rfilz M Ca ar n

a i C

C m r

o n G

f o t e z

M i

l L

g i

s o

R L

b t

i a t s ib m i l

o r m

Bortezomib Bortez O Bortezo

t n i a t G t a i s r l st a O v

m i

S G

4 2 2 9 2 4 F

ABT-263 ABT-263 P ABT-263

Bosutinib 429242 PF

i m v a s

a t i t

17-DMAG 17-DMAG n 17-DMAG

I P 1 S i 3 m - 7 v

4 a

s a t

t i

B o

s u

n i

WZ4002 WZ4002 b WZ4002

B T

o -

3 s 8

u 7 t

i n i b

Paclitaxel Crizotinib Paclitaxel Crizotinib C Paclitaxel Crizotinib I

V 1 V

a 3 a n V -

d a 7 nd

etan nde 4 eta

E ib tan nib f

a ib C v

I r P

1 z

- 7

T Y

iv T T -

an T 3 iv

t iv 8 a 7 n

I in a

m ib nt tin

a in ib

C t i

i b

n Y

b T

S f

u a S

n S v u

i i

t r

J n

in u e i

n n N i t

b i K i

t z n

K i

E in b

i ib R

i b r K

e i m

t S n o

o T S a

r r

I S o a i P I

fe o n ra

F i

N n ra fe

i I T

4 b m fe n

3 n ib

1 i t b R

P i

H N

i P 6

A K

P P H

R H i

F 6 A n

- 6 A h J 6

5 R

i 6

7 7 6 5

3 5 6 K

P P 7

2 5 o

2 5

D 7 r P i

2 2

N 5

S I

8 1 D

T P I

7 2 4 t

i D 1 T 3

3 b 7

T 1 P i

t 0

t 3

o 7 i 7

c 0

96 r 3 4 P

K O 7

- 0

I F 4 O

K 4

S I 7

- S

I K O 4 S

- 5 1

y 9 I

S 7

3 -

k 0

P 9

6 9 I- 0

F I 2 6

N 9 S

n 6

- 0 N V t h

T 8

6 a

i P N V 7 b

t r

a 3

i Vismodegib XL-184 - Vismodegib XL-184 V Vismodegib XL-184 P i

t T c S

o 2

m L A P -

r T 2

a E a - A

8 T L

V t

p t

6 y E6

i L a

t A

e a i

8 k p

c a E

i t 4 m

b p a 8

i 6 In

F n S t 4

a 8 T

u i i

o b y t 4 F n r r

5 i i

k i

a n a

r F b o b

- e i m f

i I

n b r

t e o

t n

E i o e

n n r e

r t

r e

n i

i r i

t i E

D n b

b a

i l e

b t

o E i n i

m r

i ni i m b

l t

t i g l

i o o V rl

D n e b

x u

r t

o 5

i t e i

y x D

o b r n

t -

H i

v D n i r o

n n b

i b

D t i e v T u o

p i b

D n

n n i

r v

- D

a - t

i P i

y 1

i n i

b b n

n s D t

g I

m a

8 i

o i e r

3 C a n

I b

D s

a E

6 i

n 3 E

r x n

n i

t I

e C a

r i s b x

y i

H e y

n i

e i K s

e g

h z C a t

t r

l i

t i

C h

r h t

o o A

T n

/ ox i

o b E t

a t t z y r E

/ i

9 x

K x C i

O i n

H t -

p o b K

u H a

z O

i O S 1

3 T

n i C i

T H l t

B - n o b

a t

T 8 T

6 i i

i H

5 r n

n t

H H N b

9 B -

M y 6

4 - in 5 i

3 1

B A b

T -

B 2 i M

i n 4

S d J 5

8 6

J b H

NT a M

6 8 B

I e 2

- r 4 S 64

K B m r

0 o

A 7 B y I 8

2 o

o S -

J J B

J W 2 d

x 5 i

1 A 0

I x 7

Z 8 6 1

n x

- K Q

Q B

4 y W

A r 2

A 8 5

e 7 0 1

y Z

8 D

M M J - o l

8 c

W a A 4

4 5 2

1 2 c Z Z

4 D 1 o

h -

T - 0 A t 8 h

9 4 e S

S Z 2 x

N 8 N D M

D 8 y

7 l - T -

e o 0

1 l

R 9 9 a Z 2 8 D 8 9

c o

- -

T 4

V V -

- l e r 7 N

O S 2 2 m 2 0

O R 9 a 9 D h

r 4 3 8 9

M a o e -

o R m

t 7

7 9 2

9 O 1 V 4 e 5 q 3

P a

l s - M l

p m o

K 2 i

5 4 2

u M -

- a 3 5 1 g

P 4 s

K g

i P i p r

P K

a i l

L r L

7 p

G o K

p 5

P K s 1 i i 4

P 7 p o

o o n

P a

N - m G r

i N K

D D

n G P q

e G i 4 R P e

2 p o 7 e

R a -

m l G L

B S 3 m P r m

R i G

R 2 T

U B

r - o 7 S

E m

E 3

Z U

2 y r l e

T B m

D R

X Z X 7 T

i S i l

Z 2 m

T 3

i V

T i

E T

K l

T X y V T

V D

Vi T

t y

o c l 2 y m

U n t h O

o 7

0 r 2

i E

i 2 D K

1 i

h 2

o u E i

1 2 i

t E

c A b C 7 O

c K y

7 i mu l

i i c

h 0 e

n e b r

6 7 r

Z p

i A y 6 n e i 1 C 2 2 o 2 b

p n o

2 - s u

i p

n s Z 6 i

i s

2 0 i C n

i c

o 2 i i 4 Z

2 6 7 o

g p n i 4

9 6 c

p V 1 - n r

p o

i n

g 5 - 5 -

r 4 2 o n

n s

m V o

a i

modegi

f -

6 i 9

0 r -

a - a f 1

t 2 2

6 a n i

r - - f 0 6

t 9

t t

u 0

s t

2 g

i a n

e t 1 a 2

v o 0

i o 2 0

o e

9 d 9 v p

5 9 3 e

m p 6

d 6

o 5 0 i r

m 9

r 9

n v 3

i 6 r

l 3 9

l 4 i n l n 9

a a

4 3

r 5 n n a 5

9 4 i r

a 6

i d i 4 3 n i m

9 5

i t

a d 9

i 5 4

1 r t

i 0 n

9 a z t m i

4 3 n

m z 9

b 1

i a

b a

e 3

z e

1 b

i e

9 n

e i 3 n

g n n

i n

n 9

i i

7 l n

i e

e b 7

b e

B C

MDA-MB-231 7 BT-549 MDA-MB-231 Paclitaxel us 6 im V 3 rol is 30 ve m E od eg 5 ib 2 4 ib 20 n ti lo r 3 E X L . 1 8 1 2 4 10 Paclitaxel (nmol/L) Crizotinib ( m mol/L) b i 1 n Paclitaxel (nmol/L) i

z i 0 r 0 0 C 0 0.5 1 1.5 2 0 0.3 0.6 0 0.2 0.4

M ABT-263 (mmol/L) JQ1 (mmol/L) Digoxin (mmol/L) D A - M 0 B 2 - - T 40 4 6 B 8 MDA-MB-231 HCC-38 MDA-MB-468 M 6 D A 30 6 -M B -4 3 6 9 4 MD -54 A-M BT 20 4 B-2 31 HCC-38 2 10 2 Paclitaxel (nmol/L) Paclitaxel (nmol/L) Erlotinib ( m mol/L) 0 0 0 0 0.51 1.5 2 0 2 46 0369 ABT-263 (mmol/L) Nutlin-3 (mmol/L) Thioridazine (mmol/L)

www.aacrjournals.org Cancer Res; 77(2) January 15, 2017 OF7

Wali et al.

MDA-MB-231 MDA-MB-468 HCC-38 BT-20 BT-549 MDA-MB-436 A 384-well format Growth inhibition (%) inhibition Growth -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 Medium-throughput 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 Crizotinib + ABT-263 Crizotinib + 96-well format Low-throughput Growth inhibition (%) inhibition Growth -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 Growth inhibition (%) inhibition Growth 384-well format -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 Medium-throughput 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 Paclitaxel + ABT-263 Paclitaxel + 96-well format Low-throughput Growth inhibition (%) inhibition Growth -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10 ABT-263 (µmol/L) ABT-263 (µmol/L) ABT-263 (µmol/L) ABT-263 (µmol/L) ABT-263 (µmol/L) ABT-263 (µmol/L)

B MDA-MB-157 HCC-70 HCC-1500 MDA-MB-157 HCC-70 HCC-1500 ABT-263 ABT-263 Paclitaxel + Crizotinib + Growth inhibition (%) inhibition Growth Growth Inhibition (%) Inhibition Growth -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 -50 0 50 100 0.001 0.1 1 10 0.001 0.1 1 10 0.001 0.1 1 10 0.001 0.1 1 10 0.001-50 0 500.1 100 1 10 0.001 0.1 1 10 ABT-263 (µmol/L) ABT-263 (µmol/L) ABT-263 (µmol/L) ABT-263 (µmol/L) ABT-263 (µmol/L) ABT-263 (µmol/L)

Figure 5. Validation of MTS results in low-throughput format. A, Experiments with crizotinib-ABT-263 and paclitaxel-ABT-263 were repeated across cell lines in low- throughput 96-well format (4 replicates/group) as opposed to 384-well medium-throughput format (2 replicates/group). Each graph shows growth response curves for individual drugs and their combination. Growth inhibition by a single dose of panel A drugs (1 mmol/L crizotinib or 3 nmol/L paclitaxel) is indicated by broken red line, while the dose–response curves for ABT-263 (panel B drug) alone, theoretical bliss (additive effect), and drug combinations are indicated by blue, green, and black lines, respectively. B, The two drug combinations were tested in additional basal (HCC-70 and HCC-1500) and mesenchymal stem-like (MDA-MB-157) cell lines in 96-well plate format.

These drug combinations were also tested in additional Mechanistic effects of ABT-263 and crizotinib mesenchymal and stem-like (MDA-MB-157) and basal (HCC- To assess the biological mechanism of crizotinib/ABT-263, we 1500 and HCC-70) cell lines. ABT-263 elicited dose-dependent looked at the mRNA expression of their targets in CCLE. Targets growth inhibition alone and in combination with paclitaxel or of ABT-263 are BCL-xL and BCL2, while crizotinib inhibits activity crizotinib in all these cell lines. Synergy was observed with of anaplastic lymphoma kinase (ALK), MET and AXL and is paclitaxel-ABT-263 in all these cell lines and with crizotinib- approved by the FDA for treatment of ALK-positive non–small ABT-263 in HCC-1500 cells (Fig. 5B). Intriguingly, XL-184, cell lung cancer. However, ALK expression is generally low and another MET inhibitor, exhibited very similar response to cri- ALK rearrangement or mutations are very rare in breast cancer, and zotinib when combined with ABT-263 across all cell lines none of the cell lines expressed ALK or had any ALK gene (Supplementary Fig. S7A). aberration. Interestingly, MDA-MB-231 cells express higher levels

Figure 4. Synergistic drug combinations against TNBC cells. A, Circular plots show the six drugs inside the inner circle (panel A) that were combined with 128 drugs (panel B) shown in the periphery grouped in drug categories. Green lines link the best, while gray lines link the second best synergistic combinations, as described in Results. B, Top synergistic combinations are summarized in a chord plot by cell line and panel A drug. C, Isobologram analysis depicting pharmacological

interaction between two drugs at the IC50 (except Nutlin-3, where IC10 was used) level of individual drugs and their combination. The straight line in each isobologram joins the individual IC50 doses of two drugs on the x- and y-axes. The data point in each isobologram, corresponding to IC50 dose of drugs given in combination, lies below the line indicating synergism.

OF8 Cancer Res; 77(2) January 15, 2017 Cancer Research

Novel Combination Therapies in Triple-Negative Breast Cancer

Table 1. Proteins altered by crizotinib/ABT-263 assessed by RPPA Crizotinib ABT-263 Combination (Protein FC relative to control) (Protein FC relative to control) (Protein FC relative to control) HCC38 DM Histone H3 (1.11) DM Histone H3 (2.09) DM Histone H3 (6.62) Histone H3 (1.18) Histone H3 (2.26) Histone H3 (4.81) PAI1 (1.19) PAI1 (1.45) PAI1 (1.64) PAR (1.01) PAR (4.35) PAR (5.93) Collagen IV (1.3) Collagen IV (1.86) Collagen IV (1.88) Cleaved Caspase-7 (1.03) Cleaved Caspase-7 (1.58) Cleaved Caspase-7 (1.74) Cleaved Caspase-3 (1.04) Cleaved Caspase-3 (1.23) Cleaved Caspase-3 (1.39) p27 Kip1 (1.03) p27 Kip1 (1.03) p27 Kip1 (1.18) Bcl-xL (0.99) Bcl-xL (0.91) Bcl-xL (0.91) p44/42 MAPK (0.99) p44/42 MAPK (0.85) p44/42 MAPK (0.77) p38 p180_pY182 (0.95) p38 p180_pY182 (0.65) p38 p180_pY182 (0.67) LDHA (1.04) LDHA (0.47) LDHA (0.51) MDA-MB-468 PAI1 (1.41) PAI1 (1.45) PAI1 (1.58) PAR (1.29) PAR (1.33) PAR (1.39) Collagen IV (1.05) Collagen IV (1.08) Collagen IV (1.29) NDRG1_pT346 (1.55) NDRG1_pT346 (1.54) NDRG1_pT346 (1.57) Bcl-xL (1.06) Bcl-xL (0.91) Bcl-xL (0.88) MAPK_pT202_Y204 (0.91) MAPK_pT202_Y204 (0.82) MAPK_pT202_Y204 (0.80) EGFR_pY1068 (0.84) EGFR_pY1068 (0.92) EGFR_pY1068 (0.87) LDHA (1.05) LDHA (0.79) LDHA (0.78) B7H4 (0.83) B7H4 (0.92) B7H4 (0.87) X14.3.3.zeta (0.77) X14.3.3.zeta (0.98) X14.3.3.zeta (0.83) MDA-MB-231 Collagen IV (1.16) Collagen IV (1.11) Collagen IV (1.33) PAR (1.05) PAR (1.21) PAR (1.15) S6_pS235_S236 (0.95) S6_pS235_S236 (0.75) S6_pS235_S236 (0.73) S6_pS240_S244 (0.87) S6_pS240_S244 (0.73) S6_pS240_S244 (0.76) Bcl-xL (0.95) Bcl-xL (0.82) Bcl-xL (0.91) LDHA (0.95) LDHA (0.73) LDHA (0.83) B7H4 (0.94) B7H4 (0.77) B7H4 (0.75) Bad_pS112 (0.97) Bad_pS112 (0.88) Bad_pS112 (0.89) BT-20 DM Histone H3 (1.04) DM Histone H3 (1.63) DM Histone H3 (1.82) Histone H3 (0.98) Histone H3 (1.67) Histone H3 (1.65) Collagen IV (1.07) Collagen IV (1.07) Collagen IV (1.59) PAR (1.08) PAR (3.34) PAR (2.91) Mcl1 (1.00) Mcl1 (1.74) Mcl1 (1.74) PDL1 (1.02) PDL1 (1.70) PDL1 (1.64) p21 (0.95) p21 (1.28) p21 (1.2) Bcl-xL (1.00) Bcl-xL (0.78) Bcl-xL (0.81) LDHA (1.10) LDHA (0.64) LDHA (0.55) SDHA (0.99) SDHA (0.54) SDHA (0.57) NDRG1_pT346 (0.98) NDRG1_pT346 (0.56) NDRG1_pT346 (0.60) MAPK_pT202_Y204 (1.18) MAPK_pT202_Y204 (0.67) MAPK_pT202_Y204 (0.62) BT-549 DM Histone H3 (1.23) DM Histone H3 (1.05) DM Histone H3 (1.23) Histone H3 (1.28) Histone H3 (1.28) Histone H3 (1.28) Mcl-1 (1.31) Mcl-1 (1.31) Mcl-1 (1.31) PAR (1.39) PAR (1.39) PAR (1.39) S6_pS235_S236 (0.95) S6_pS235_S236 (1.23) S6_pS235_S236 (1.18) S6_pS240_S244 (0.92) S6_pS240_S244 (1.08) S6_pS240_S244 (1.13) Bcl-xL (1.00) Bcl-xL (0.92) Bcl-xL (0.95) LDHA (0.688) LDHA (0.688) LDHA (0.68) MDA-MB-436 Mcl1 (1.10) Mcl1 (1.46) Mcl1 (1.50) DUSP4 (0.99) DUSP4 (1.45) DUSP4 (1.433) Histone H3 (1.01) Histone H3 (1.36) Histone H3 (1.28) DM Histone H3 (1.04) DM Histone H3 (1.33) DM Histone H3 (1.23) PAI1 (1.09) PAI1 (1.20) PAI1 (1.23) PAR (0.91) PAR (1.34) PAR (1.20) Collagen IV (0.96) Collagen IV (1.07) Collagen IV (1.1) Bcl-xL (0.97) Bcl-xL (1.04) Bcl-xL (0.97) NOTE: TNBC cell lines were exposed to 1 mmol/L crizotinib or 1 mmol/L ABT-263 alone or their combination in duplicate for 24 hours, and the protein extracts were assayed for the expression of basal and activated levels of 291 proteins at the RPPA Core Facility at the MD Anderson Cancer Center (Houston, TX). Top altered proteins are listed and their fold change (FC) compared with respective vehicle-treated controls shown in parentheses. Blue indicates proteins upregulated while red indicates proteins downregulated by the drug combination. Interesting candidate proteins are represented in bold.

www.aacrjournals.org Cancer Res; 77(2) January 15, 2017 OF9

Wali et al.

24 h 1 h MDA-MB-436 MDA-MB-231 A Phosphorylated MET and AXL C MDA-MB-231 MDA-MB-436 V Cr A Co V Cr A Co µ IP: pY100 (2 g) V Cr A Co V Cr A Co IgG BCL-xL MET Cleaved Casp-3

AXL PARP

Total MET and AXL Cleaved PARP MDA-MB-231 MDA-MB-436 pEGFR V Cr A Co V Cr A Co MET EGFR pBAD AXL BAD GAPDH GAPDH

3 h 24 h Mechanism of Crizotinib and ABT-263 B MDA-MB-231 MDA-MB-436 MDA-MB-231 MDA-MB-436 EGFR D HGF GAS6 HGF Dimer EGF EGFR-MET AXL MET VCrACoVCrACoVCrACoVCrACo Cell membrane AXL Shc Grb2 Grb2 SOS Crizotinib PI3K phospho-AKT RAS

AKT AKT RAF p53 BAD/ BIM phospho-ERK1/2 MEK mTOR Cleaved Caspase-3

MAPK BCL-xL/ ERK1/2 Cleaved BCL2 PARP

Actin ABT-263 Apoptosis Cell Proliferation and Survival

Figure 6. Mechanism of action of ABT-263 and crizotinib. MDA-MB-231 cells in which crizotinib-ABT-263 elicited highest synergistic response and MDA-MB-436 cells that were least sensitive were treated with either vehicle (V), 1 mmol/L crizotinib (Cr), 1 mmol/L ABT-263 (A), or their combination (Co) for 1, 3, or 24 hours. A, To determine phosphorylated MET and AXL levels, phosphorylated proteins were immunoprecipitated with antiphospho-tyrosine (P-Tyr-100), using IgG as control, and then immunoblotted with anti-MET and anti-AXL antibodies. Total MET and AXL were directly immunoblotted using respective antibodies, using GAPDH as internal control. B, Effect of treatments on the downstream activated (phosphorylated) levels of AKT and ERK1/2 were determined at 3 hours and 24 hours, using actin as internal control. C, To determine the effects of drug treatments on apoptotic markers and EGFR activity, relative levels of BCL-xL, cleaved caspase-3 and PARP, total and phosphorylated BAD (pBAD), and EGFR were determined by immunoblotting after 24 hours treatment exposure. D, Proposed mechanism of action of the crizotinib-ABT-263 combination. Crizotinib inhibits MET activity and MET-dependent RTK activity and downstream signaling. In contrast, ABT-263 inhibits BCL-xL and BCL2, thereby enabling proapoptotic proteins to trigger apoptosis indicated by cleavage of caspase-3 and PARP. When combined, crizotinib and ABT-263 act synergistically resulting in enhanced inhibition of TNBC cell growth and apoptosis.

of BCL-xL, MET, and AXL mRNA and are most sensitive, while poly(ADP-ribose) polymer (PAR), and slightly decreased BCL-xL, MDA-MB-436 cells with the lowest expression of BCL-xL are least while crizotinib decreased activated levels of phosphorylated sensitive to the drug combination. However, expression of BCL- EGFR or MAP kinases in different cell lines, and their combination xL, AXL, and of about 26 direct and indirect targets of these two produced enhanced alterations in these proteins. Intriguingly, drugs does not explain sensitivities of all cell lines (Supplemen- however, combined treatment with ABT-263 and crizotinib pre- tary Fig. S7B). dominantly altered the levels of histone H3, dimethyl histone H3, To directly assess the mechanistic effects, we measured the basal plasminogen activator inhibitor 1 (PAI1), Collagen IV, and lactate and activated levels of 291 signaling proteins, including the direct dehydrogenase A (LDHA) across most TNBC cell lines in addition drug targets, before and after crizotinib and ABT-263 treatment by to alterations in phosphorylated kinases in the MAPK pathway reverse phase protein array (RPPA) analysis. Table 1 describes the (p44/42, pT202/204, p180/182 p38, p235/236, and p240/244 RPPA results as fold changes of the most affected proteins in each S6), BCL-xL, PAR, and cleaved caspases. Whether these effects on cell line, as compared with respective vehicle-treated control cells. H3, PAI1, collagen IV and LDHA are bystander changes caused Proteins with fold changes differing in multiple cell lines but with by ABT-263-crizotinib, or if they play a critical role in TNBC same directionality are also listed in this table. ABT-263 treatment growth needs further investigation. Highest fold changes in pro- increased the expression of cleaved-caspase-7, cleaved caspase-3, teins were observed in HCC-38 cells that were most sensitive to

OF10 Cancer Res; 77(2) January 15, 2017 Cancer Research

Novel Combination Therapies in Triple-Negative Breast Cancer

this treatment (Table 1). Overall, crizotinib alone modestly induces apoptosis and exerts synergistic anticancer activity in altered protein expression, ABT-263 alone produced relatively combination with the MET inhibitor crizotinib that suppresses greater fold changes, while the combination produced the highest MET kinase activity in TNBCs, as summarized in Fig. 6D. fold changes in the above proteins. In contrast to more sensitive HCC38, MDA-MB-468, MDA-MB-231, and BT-20 cell lines in which the MAPK pathway was suppressed, no decrease in Discussion pathway activity was seen in BT-549 or MDA-MB-436 cells that We chose six FDA-approved drugs and combined them with were least sensitive. These results suggest a proapoptotic program 128 agents to identify highly synergistic and effective combina- turned on by ABT-263 and inhibition of mitogenic signaling by tion treatments in TNBC. This is a novel guided medium-through- crizotinib blockade of RTKs resulting in growth inhibition and put drug screening design that allows to identify effective combi- apoptosis of cells. nations that can be translated quickly to the clinic. The broadly The effect of ABT-263/crizotinib on programmed cell death was selected drug panel allowed us to systematically evaluate associa- further corroborated by Annexin V labeling. Exposure to crizoti- tions of cell line sensitivities with drug categories, and integrating nib across cell lines resulted in maximum apoptotic/dead cells in genomic, transcriptomic, and proteomic data to provide addi- BT-20 (21%) while ABT-263 alone resulted in maximum cell tional biological insights for the effective combinations. This death in HCC-38 (73%) as indicated by the sum of two upper could enable making informed predictions as to which drug quadrants in each dot plot in Supplementary Fig. S8. However, combinations are likely to have the highest impact in clinical combined treatment with ABT-263 and crizotinib resulted in trials. We discovered specific promising drug combinations that maximum cell death in HCC-38 cells (82%). Apoptosis was could be tested in clinical trials owing to their potent synergistic greater in HCC-38, MDA-MB-468, and MDA-MB-231 cell lines, anticancer activity in culture. consistent with their higher sensitivity in growth assays, while BT- The cost and time needed to develop a new chemical entity are 549 and MDA-MB-436 were least sensitive to drug-induced apo- often prohibitive (30). Therefore, repurposing approved and ptosis (Supplementary Fig. S8). partially developed drugs for cancer creates an opportunity to To confirm RPPA and flow-cytometry results, we performed rapidly advance to patients better drug therapies by capitalizing immunoblotting using sensitive MDA-MB-231 and relatively on existing data and experience (31, 32). Furthermore, rationally resistant MDA-MB-436 cell lines. We first looked at the effect of selecting and testing such drugs in large combination screens is crizotinib on the total and activated levels of its targets, MET and critical to find novel effective therapies, but such screens have been AXL, when given alone or in combination with ABT-263. Crizo- rather limited (8). We present the first guided medium-through- tinib blocked MET phosphorylation in both cell lines, when given put combination screen for TNBC to test efficacy and synergy of alone or in combination, but had no effect on the phosphorylated 768 unique drug combinations. AXL levels after 1-hour treatment exposure (Fig. 6A). Basal MET One of the effective synergistic drug combinations identified and AXL levels were both higher in MDA-MB-231 cells than in in this screen was ABT-263 (BCL-xL/BCL2 inhibitor) and cri- MDA-MB-436. Total MET levels remained unaltered, and total zotinib. Targeting the BCL2 apoptotic pathway has been pro- AXL was slightly increased by crizotinib, while ABT-263 had no posed as a therapeutic strategy in TNBC (16, 33). ABT-263 or effect on the total or active levels of AXL or MET. Next, we navitoclax was well tolerated and showed efficacy in chronic determined the effect of this combination on downstream AKT lymphocytic leukaemia (CLL) in phase I and II trials (34, 35), and MAPK activity. Consistent with the RPPA results, there was but thrombocytopenia was the dose limiting toxicity. Recently, little or no decrease in the phosphorylated levels of AKT and ABT-199, a more specific BCL2 inhibitor, was granted priority ERK1/2 upon drug combination treatment in these cell lines (Fig. review by the FDA for CLL. However, we found ABT-263 to be 6B). It is noteworthy that although MDA-MB-436 cells express considerably more potent than ABT-199 in inhibiting TNBC lower levels of AKT than MDA-MB-231, they express conspicu- cell growth. If combined with standard cytotoxics or targeted ously higher levels of phosphorylated (active) AKT, indicating a therapy at lower doses, ABT-263 may enhance sensitivity of hyperactive AKT signaling in this cell line, which may contribute solid tumors and prevent high-dose toxicity. to its lower sensitivity to the drug combination. Indeed, MDA-MB- ABT-263 synergized with crizotinib (MET/AXL inhibitor). We 436 cells were more sensitive than MDA-MB-231 to the AKT identified MET as one of the top 681 overexpressed genes in inhibitor KP372-1 that resulted in 100% growth inhibition of TNBCs versus non-TNBCs in three independent breast cancer MDA-MB-436 cells in nanomolar doses when given alone or in clinical datasets. RTKs including MET and EGFR have been sug- combination in our screen. On the other hand, we looked at the gested as potential targets in triple-negative breast and in other effect of ABT-263 on caspase-3 and PARP cleavage. As shown cancers (36–38). However, targeting RTKs with EGFR inhibitors in in Fig. 6C, ABT-263 induces cleavage of caspase-3 and PARP, TNBC has shown limited clinical efficacy. MET interaction with which is further enhanced by combination treatment with crizo- the AXL and EGFR family has been suggested as a possible tinib, confirming apoptosis. BCL-xL protein expression was higher resistance mechanism (39). Although AXL was not overexpressed, in MDA-MB-231 than in MDA-MB-436 cells, but neither the levels MET expression was clearly high in TNBC subtype in TCGA of BCL-xL nor its proapoptotic counterpart BAD changed upon samples (Supplementary Fig. S9). Similarly, most TNBC cell lines drug treatment in 24 hours. Interestingly though, we observed a from CCLE express MET but only a minority express AXL. Crizo- slight decline in BCL-xL levels after 72 hours of ABT-263 treatment tinib and ABT-263 displayed synergistic and potent antiprolifera- in additional experiments (data not shown). Among the two MSL tive and apoptotic activity in both MSL and basal TNBCs. The cell lines, MDA-MB-436 cells express very low levels of basal and same combination also demonstrated anticancer activity in addi- active levels of EGFR. Although drug treatment alone had no effect tional TNBC cell lines from each category. We evaluated the results on EGFR, combined treatment slightly decreased phosphorylated of another MET inhibitor XL-184 in combination with ABT-263 EGFR levels in MDA-MB-231 cells. Taken together, ABT-263 from our screen, and found strikingly similar response with

www.aacrjournals.org Cancer Res; 77(2) January 15, 2017 OF11

Wali et al.

crizotinib/ABT-263 in all cell lines (Supplementary Fig. S7A). regulated genes rather than the expression of all or specific genes ABT-263 has also been shown to synergize with other RTK of a targeted pathway. Results were validated systematically in inhibitors (40). ABT-263 with the EGFR inhibitor AZD9291 will addition to initial QC steps and internal controls to assess synergy, be tested in clinical trial (ClinicalTrials.gov study NCT02520778) critical to reliably interpret MTS results. The dataset generated in in EGFR inhibitor-resistant lung cancer. Because TNBCs do not this study is made publicly available to help translational respond well to EGFR inhibitors in clinic, some additional inter- researchers by providing a roadmap of effective or sub-effective, esting combinations that emerge from this study are erlotinib with synergistic or sub-synergistic drug combinations in TNBC, and to 5-fluorouracil or ABT-263 that could be tested. possibly help train predictors of effective drug combinations from YM155, Bortezomib, and carfilzomib showed remarkable anti- single agent data. Our results revealed distinct patterns between cancer activity alone and also, in combinations across all cell lines TNBC backgrounds and cell line sensitivities, and proposed tested. A recently concluded phase II randomized study of YM155 several promising novel drug combinations including ABT-263- and docetaxel advocates rational selection of breast cancer sub- crizotinib that could be tested in the clinic. All TNBC types express population with higher survivin activity (41), while LCL161, MET unlike most other RTKs, and therefore, targeting MET with another inhibitor of IAP, has shown promising efficacy with crizotinib may achieve broader activity across TNBCs. Combining paclitaxel in a phase II, neoadjuvant, randomized study in TNBC crizotinib with ABT-263 at lower doses may provide therapeutic patients (ClinicalTrials.gov study NCT01617668). Bortezomib benefits in TNBC patients while avoiding high-dose toxicity and carfilzomib are exciting protease inhibitors recently approved associated with ABT-263 monotherapy. by the FDA for multiple myeloma and are in clinical trials for various cancers, including breast (42, 43). Disclosure of Potential Conflicts of Interest ABT-263, nutlin-3, and JQ1 combined synergistically with No potential conflicts of interest were disclosed. paclitaxel in our screen. ABT-263-paclitaxel combination has demonstrated preclinical efficacy in NSCLC, ovarian, and prostate cancer (44–46). Phase I study of ABT-263 in combination with Authors' Contributions carboplatin/paclitaxel revealed modest antitumor activity in solid Conception and design: V.B. Wali, L. Pusztai, D.F. Stern, C. Hatzis tumors; however, only one breast cancer patient participated in Development of methodology: V.B. Wali, C.G. Langdon, M.A. Held, C. Hatzis this study (47); hence, ABT-263-paclitaxel combination may be Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): V.B. Wali, C.G. Langdon, M.A. Held, G.A. Patwardhan worth exploring in TNBC patients. Nutlin-3 synergizes with Analysis and interpretation of data (e.g., statistical analysis, biostatistics, cisplatin to induce p53-dependent tumor cell apoptosis in NSCLC computational analysis): V.B. Wali, J.T. Platt, G.A. Patwardhan, A. Safonov, (48). However, nutlins alone at higher doses have off-target B. Aktas, L. Pusztai, C. Hatzis effects, and therefore are being studied in combination with other Writing, review, and/or revision of the manuscript: V.B. Wali, C.G. Langdon, drugs (48). Thus, ABT-463, nutlin-3 or its derivatives, and JQ1 J.T. Platt, A. Safonov, L. Pusztai, D.F. Stern, C. Hatzis emerge as obvious candidates to test in combination with pac- Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): V.B. Wali, J.T. Platt, L. Pusztai litaxel in clinic. Because paclitaxel is an already integral compo- Study supervision: V.B. Wali, L. Pusztai, C. Hatzis nent of standard chemotherapy in TNBC, such trials can be designed and conducted faster and, if successful, these new drugs can be directly introduced to the existing therapeutic regimen. Grant Support Our study design tested FDA-approved drugs with selected This work was supported by an award from the Breast Cancer Research anticancer agents in a medium-throughput screen whose output Foundation to L. Pusztai. The costs of publication of this article were defrayed in part by the payment of raw signal was scrutinized for systematic errors, filtered, and page charges. This article must therefore be hereby marked advertisement in extensively analyzed. Overall, MSL cells were relatively less sen- accordance with 18 U.S.C. Section 1734 solely to indicate this fact. sitive, while genotoxins, apoptosis and cell-cycle regulators were more effective drug classes in TNBC. Drug sensitivities could be Received July 19, 2016; revised October 17, 2016; accepted November 4, explained better by the proportion of upregulated and down- 2016; published OnlineFirst November 21, 2016.

References 1. Liedtke C, Mazouni C, Hess KR, AndreF,TordaiA,MejiaJA,etal. 6. Marcotte R, Sayad A, Brown KR, Sanchez-Garcia F, Reimand J, Haider M, Response to neoadjuvant therapy and long-term survival in et al. Functional genomic landscape of human breast cancer drivers, patients with triple-negative breast cancer. J Clin Oncol 2008;26: vulnerabilities, and resistance. Cell 2016;164:293–309. 1275–81. 7. Garnett MJ, Edelman EJ, Heidorn SJ, Greenman CD, Dastur A, Lau KW, 2. Koboldt DC, Fulton RS, McLellan MD, Schmidt H, Kalicki-Veizer J, et al. Systematic identification of genomic markers of drug sensitivity in McMichael JF, et al. Comprehensive molecular portraits of human cancer cells. Nature 2012;483:570–5. breast tumours. Nature 2012;490:61–70. 8. Mathews Griner LA, Guha R, Shinn P, Young RM, Keller JM, Liu D, et al. 3. Anders CK, Winer EP, Ford JM, Dent R, Silver DP, Sledge GW, et al. Poly High-throughput combinatorial screening identifies drugs that cooperate (ADP-Ribose) polymerase inhibition: "targeted" therapy for triple-negative with ibrutinib to kill activated B-cell-like diffuse large B-cell lymphoma breast cancer. Clin Cancer Res 2010;16:4702–10. cells. Proc Natl Acad Sci U S A 2014;111:2349–54. 4. Carey LA, Rugo HS, Marcom PK, Mayer EL, Esteva FJ, Ma CX, et al. TBCRC 9. Errico A. Screening: high-throughput screen identifies a roadmap for 001: randomized phase II study of cetuximab in combination with carbo- combination drug trials. Nat Rev Clin Oncol 2014;11:124. platin in stage IV triple-negative breast cancer. J Clin Oncol 2012;30: 10. Held MA, Langdon CG, Platt JT, Graham-Steed T, Liu Z, Chakraborty A, 2615–23. et al. Genotype-selective combination therapies for melanoma identified 5. Barretina J, Caponigro G, Stransky N, Venkatesan K, Margolin AA, Kim S, by high-throughput drug screening. Cancer Discov 2013;3:52–67. et al. The Cancer Cell Line Encyclopedia enables predictive modelling of 11. Greco WR, Bravo G, Parsons JC. The search for synergy: a critical review anticancer drug sensitivity. Nature 2012;483:603–7. from a response surface perspective. Pharmacol Rev 1995;47:331–85.

OF12 Cancer Res; 77(2) January 15, 2017 Cancer Research

Novel Combination Therapies in Triple-Negative Breast Cancer

12. Chou T-C. Drug combination studies and their synergy quantification 31. Weir SJ, DeGennaro LJ, Austin CP. Repurposing approved and abandoned using the Chou-Talalay method. Cancer Res 2010;70:440–6. drugs for the treatment and prevention of cancer through public-private 13. Langdon CG, Held MA, Platt JT, Meeth K, Iyidogan P, Mamillapalli R, et al. partnership. Cancer Res 2012;72:1055–8. The broad-spectrum receptor tyrosine kinase inhibitor dovitinib suppresses 32. Huang R, Southall N, Wang Y, Yasgar A, Shinn P, Jadhav A, et al. The NCGC growth of BRAF-mutant melanoma cells in combination with other sig- pharmaceutical collection: a comprehensive resource of clinically naling pathway inhibitors. Pigment Cell Melanoma Res 2015;28:417–30. approved drugs enabling repurposing and chemical genomics. Sci Transl 14. Lehmann BD, Bauer JA, Chen X, Sanders ME, Chakravarthy AB, Shyr Y, et al. Med 2011;3:80ps16. Identification of human triple-negative breast cancer subtypes and pre- 33. Goodwin CM, Rossanese OW, Olejniczak ET, Fesik SW. Myeloid cell clinical models for selection of targeted therapies. J Clin Invest 2011;121: leukemia-1 is an important apoptotic survival factor in triple-negative 2750–67. breast cancer. Cell Death Differ 2015;22:2098–106. 15. Colavito SA, Zou MR, Qin Y, Nguyen DX, Stern DF. Significance of glioma- 34. Gandhi L, Camidge DR, Ribeiro de Oliveira M, Bonomi P, Gandara D, associated oncogene homolog 1 (GLI1)expression in claudin-low breast Khaira D, et al. Phase I study of Navitoclax (ABT-263), a novel Bcl-2 family cancer and crosstalk with the nuclear factor kappa-light-chain-enhancer of inhibitor, in patients with small-cell lung cancer and other solid tumors. activated B cells (NFkB) pathway. Breast Cancer Res [Internet] 2014;16. J Clin Oncol 2011;29:909–16. Available from: http://dx.doi.org/10.1186/s13058-014-0444-4 35. Rudin CM, Hann CL, Garon EB, Ribeiro de Oliveira M, Bonomi PD, 16. O'Toole SA, Beith JM, Millar EKA, West R, McLean A, Cazet A, et al. Camidge DR, et al. Phase II study of single-agent navitoclax (ABT-263) Therapeutic targets in triple negative breast cancer. J Clin Pathol and biomarker correlates in patients with relapsed small cell lung cancer. 2013;66:530–42. Clin Cancer Res 2012;18:3163–9. 17. Sarti M, Pinton S, Limoni C, Carbone GM, Pagani O, Cavalli F, et al. 36. Kim YJ, Choi J-S, Seo J, Song J-Y, Lee SE, Kwon MJ, et al. MET is a potential Differential expression of testin and survivin in breast cancer subtypes. target for use in combination therapy with EGFR inhibition in triple- Oncol Rep 2013;30:824–32. negative/basal-like breast cancer. Int J Cancer 2014;134:2424–36. 18. Cheng SM, Chang YC, Liu CY, Lee JYC, Chan HH, Kuo CW, et al. YM155 37. Dietrich MF, Yan SX, Schiller JH. Response to Crizotinib/Erlotinib Com- down-regulates survivin and XIAP, modulates autophagy and induces bination in a Patient with a Primary EGFR-Mutant Adenocarcinoma and a autophagy-dependent DNA damage in breast cancer cells. Br J Pharmacol Primary c-met-Amplified Adenocarcinoma of the Lung. J Thorac Oncol 2015;172:214–34. 2015;10:e23–5. 19. Tan X, Hu L, Luquette LJ3rd, Gao G, Liu Y, Qu H, et al. Systematic 38. Gaule PB, Crown J, O'Donovan N, Duffy MJ. cMET in triple-negative breast identification of synergistic drug pairs targeting HIV. Nat Biotechnol cancer: is it a therapeutic target for this subset of breast cancer patients? 2012;30:1125–30. Expert Opin Ther Targets 2014;18:999–1009. 20. Fitzgerald JB, Schoeberl B, Nielsen UB, Sorger PK. Systems biology and 39. Meyer AS, Miller MA, Gertler FB, Lauffenburger DA. The receptor AXL combination therapy in the quest for clinical efficacy. Nat Chem Biol diversifies EGFR signaling and limits the response to EGFR-targeted inhi- 2006;2:458–66. bitors in triple-negative breast cancer cells. Sci Signal 2013;6:ra66. 21. Wali VB, Sylvester PW. Synergistic antiproliferative effects of gamma- 40. Hata AN, Niederst MJ, Archibald HL, Gomez-Caraballo M, Siddiqui FM, tocotrienol and statin treatment on mammary tumor cells. Lipids Mulvey HE, et al. Tumor cells can follow distinct evolutionary paths to 2007;42:1113–23. become resistant to epidermal growth factor receptor inhibition. Nat Med; 22. Khoo KH, Hoe KK, Verma CS, Lane DP. Drugging the p53 pathway: 2016;22:262–9. understanding the route to clinical efficacy. Nat Rev Drug Discov 2014;13: 41. Clemens MR, Gladkov OA, Gartner E, Vladimirov V, Crown J, Steinberg J, 217–36. et al. Phase II, multicenter, open-label, randomized study of YM155 plus 23. Hedley D, Shamas-Din A, Chow S, Sanfelice D, Schuh AC, Brandwein JM, docetaxel as first-line treatment in patients with HER2-negative metastatic et al. A phase I study of elesclomol sodium in patients with acute myeloid breast cancer. Breast Cancer Res Treat 2015;149:171–9. leukemia. Leuk Lymphoma 2016;57:2437–40. 42. Falchook GS, Wheler JJ, Naing A, Jackson EF, Janku F, Hong D, et al. 24. O'Day SJ, Eggermont AMM, Chiarion-Sileni V, Kefford R, Grob JJ, Mortier Targeting hypoxia-inducible factor-1a (HIF-1a) in combination with L, et al. Final results of phase III SYMMETRY study: randomized, double- antiangiogenic therapy: a phase I trial of bortezomib plus bevacizumab. blind trial of elesclomol plus paclitaxel versus paclitaxel alone as treatment Oncotarget 2014;5:10280–92. for chemotherapy-naive patients with advanced melanoma. J Clin Oncol 43. Papadopoulos KP, Burris HA3rd, Gordon M, Lee P, Sausville EA, Rosen PJ, 2013;31:1211–8. et al. A phase I/II study of carfilzomib 2-10-min infusion in patients with 25. Fu L-L, Tian M, Li X, Li J-J, Huang J, Ouyang L, et al. Inhibition of BET advanced solid tumors. Cancer Chemother Pharmacol 2013;72:861–8. bromodomains as a therapeutic strategy for cancer drug discovery. Onco- 44. Stamelos VA, Robinson E, Redman CW, Richardson A. Navitoclax aug- target 2015;6:5501–16. ments the activity of carboplatin and paclitaxel combinations in ovarian 26. Shi J, Wang Y, Zeng L, Wu Y, Deng J, Zhang Q, et al. Disrupting the cancer cells. Gynecol Oncol 2013;128:377–82. interaction of BRD4 with diacetylated Twist suppresses tumorigenesis in 45. Wang C, Huang S-B, Yang M-C, Lin Y-T, Chu I-H, Shen Y-N, et al. basal-like breast cancer. Cancer Cell 2014;25:210–25. Combining paclitaxel with ABT-263 has a synergistic effect on paclitaxel 27. Shu S, Lin CY, He HH, Witwicki RM, Tabassum DP, Roberts JM, et al. resistant prostate cancer cells. PLoS One 2015;10:e0120913. Response and resistance to BET bromodomain inhibitors in triple-negative 46. Tan N, Malek M, Zha J, Yue P, Kassees R, Berry L, et al. Navitoclax enhances breast cancer. Nature 2016;529:413–7. the efficacy of taxanes in non-small cell lung cancer models. Clin Cancer 28. Langdon CG, Wiedemann N, Held MA, Mamillapalli R, Iyidogan P, Res 2011;17:1394–404. Theodosakis N, et al. SMAC mimetic Debio 1143 synergizes with taxanes, 47. Vlahovic G, Karantza V, Wang D, Cosgrove D, Rudersdorf N, Yang J, et al. A topoisomerase inhibitors and bromodomain inhibitors to impede growth phase I safety and pharmacokinetic study of ABT-263 in combination with of lung adenocarcinoma cells. Oncotarget 2015;6:37410–25. carboplatin/paclitaxel in the treatment of patients with solid tumors. Invest 29. Hatzis C, Bedard PL, Birkbak NJ, Beck AH, Aerts HJWL, Stem DF, et al. New Drugs 2014;32:976–84. Enhancing reproducibility in cancer drug screening: how do we move 48. Deben C, Wouters A, Op de Beeck K, van Den Bossche J, Jacobs J, forward? Cancer Res 2014;74:4016–23. Zwaenepoel K, et al. The MDM2-inhibitor Nutlin-3 synergizes with cis- 30. Scannell JW, Blanckley A, Boldon H, Warrington B. Diagnosing the decline platin to induce p53 dependent tumor cell apoptosis in non-small cell lung in pharmaceutical R&D efficiency. Nat Rev Drug Discov 2012;11:191–200. cancer. Oncotarget 2015;6:22666–79.

www.aacrjournals.org Cancer Res; 77(2) January 15, 2017 OF13

Systematic Drug Screening Identifies Tractable Targeted Combination Therapies in Triple-Negative Breast Cancer

Vikram B. Wali, Casey G. Langdon, Matthew A. Held, et al.

Cancer Res Published OnlineFirst November 21, 2016.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-16-1901

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2016/11/19/0008-5472.CAN-16-1901.DC1

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2017/01/09/0008-5472.CAN-16-1901. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.