Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
Molecular
Cancer
Research
Signal Transduction
Intrinsic Resistance to Cixutumumab Is Conferred
by Distinct Isoforms of the Insulin Receptor
Amelie Forest1, Michael Amatulli1, Dale L. Ludwig1, Christopher B. Damoci1, Ying Wang1,
Colleen A. Burns1, Gregory P. Donoho2, Nina Zanella3, Heinz H. Fiebig3, Marie C. Prewett1,
David Surguladze1, James T. DeLigio1, Peter J. Houghton4, Malcolm A. Smith5, and
Ruslan Novosiadly1
Abstract
Despite a recent shift away from anti–insulin-like growth factor
I receptor (IGF-IR) therapy, this target has been identified as a key
player in the resistance mechanisms to various conventional and
targeted agents, emphasizing its value as a therapy, provided that
it is used in the right patient population. Molecular markers
predictive of antitumor activity of IGF-IR inhibitors remain largely
unidentified. The aim of this study is to evaluate the impact of
insulin receptor (IR) isoforms on the antitumor efficacy of cixutumumab, a humanized mAb against IGF-IR, and to correlate
their expression with therapeutic outcome. The data demonstrate
that expression of total IR rather than individual IR isoforms
inversely correlates with single-agent cixutumumab efficacy in
pediatric solid tumor models in vivo. Total IR, IR-A, and IR-B
expression adversely affects the outcome of cixutumumab in
combination with chemotherapy in patient-derived xenograft
models of lung adenocarcinoma. IR-A overexpression in tumor
cells confers complete resistance to cixutumumab in vitro and
in vivo, whereas IR-B results in a partial resistance. Resistance in IRB–overexpressing cells is fully reversed by anti–IGF-II antibodies,
suggesting that IGF-II is a driver of cixutumumab resistance in this
setting. The present study links IR isoforms, IGF-II, and cixutumumab efficacy mechanistically and identifies total IR as a biomarker predictive of intrinsic resistance to anti–IGF-IR antibody.
Introduction
formation induced by viral and cellular oncogenes (4). IGF-IR
upregulation was observed in a variety of tumor types, including
prostate, breast, colon, and lung cancer and melanoma (5, 6).
Moreover, the IGF-IR pathway has also been implicated in the
development of resistance to other antitumor modalities, including radiotherapy, chemotherapeutic agents, and targeted therapies (1). Therefore, targeting the IGF-IR pathway represents an
attractive strategy for the treatment of various tumor types.
Over the last decade, a number of mAb and small-molecule
tyrosine kinase inhibitors (TKI) directed against IGF-IR have made
their way into clinical trials (www.clinicaltrials.gov). Anti–IGF-IR
mAbs, including cixutumumab, a fully human mAb against IGFIR, are currently the most clinically advanced molecules. Despite
very promising results in preclinical and early phase clinical
studies, results from phase III trials have failed to meet expectations (7). It is important to note, that although no significant
clinical benefit was observed in the intention-to-treat population, a distinct subset of patients seems to benefit from IGF-IR
targeting (8–11). Elucidating molecular markers predictive of
antitumor efficacy of anti–IGF-IR therapy, however, is an important and ongoing challenge.
Somatic genetic aberrations are frequently the major determinants of oncogenic and pharmacologic dependence in cancer
(12, 13). In most tumors, however, IGF-IR pathway is not altered
genetically suggesting that additional nongenomic factors may
mediate sensitivity or resistance to IGF-IR–targeted therapies.
Intrinsic or acquired resistance to targeted agents frequently results
from the activation of alternative receptor tyrosine kinases (RTK),
including ERBB, MET, FGFR, and AXL family members (14–18).
For more than two decades, the insulin-like growth factor (IGF)
system, which includes receptors (IGF-IR, IGF-IIR, and insulin
receptor), ligands (IGF-I and IGF-II), and high-affinity IGF-binding proteins (IGFBP1-6), has been studied with great interest in
cancer biology. Although this highly regulated pathway plays a
crucial role in the normal development and growth of tissues, its
deregulation contributes to tumor initiation, proliferation, and
survival (1). Elevated circulating IGF-I levels have been associated
with increased cancer risk (2). Conversely, individuals with genetic disorders resulting in low circulating levels of IGF-I and IGF-II
are resistant to cancer development (3). In vitro studies have
demonstrated the importance of functional IGF-IR for cell trans-
1
Eli Lilly and Company, New York, New York. 2Eli Lilly and Company,
Indianapolis, Indiana. 3Oncotest, Freiburg, Germany. 4Nationwide
Children's Hospital, Columbus, Ohio. 5Cancer Therapy Evaluation Program, NCI, Bethesda, Maryland.
Note: Supplementary data for this article are available at Molecular Cancer
Research Online (http://mcr.aacrjournals.org/).
Current address for P.J. Houghton: Greehey Children's Cancer Research Institute, University of Texas Health Sciences Center, San Antonio, Texas.
Corresponding Author: Ruslan Novosiadly, Eli Lilly and Company, 450 East 29th
Street, 11th Floor, New York, NY 10016. Phone: 646-638-6392; E-mail:
[email protected]
doi: 10.1158/1541-7786.MCR-15-0279
2015 American Association for Cancer Research.
Implications: This study identifies total IR as a biomarker predictive
of primary resistance to IGF-IR antibodies and provides a rationale
for new clinical trials enriched for patients whose tumors display
low IR expression. Mol Cancer Res; 13(12); 1615–26. 2015 AACR.
www.aacrjournals.org
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
1615
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
Forest et al.
Insulin receptor (INSR or IR), which shares up to 70% homology
with IGF-IR and is commonly expressed in neoplasms and tumor
cells, might be implicated in the resistance to anti–IGF-IR therapy.
Alternative splicing of INSR transcript results in two isoforms, IR-A
and IR-B, which differ by the exclusion of exon 11 encoding 12
amino acids (19). Although IR-B isoform binds primarily insulin,
IR-A is capable of binding both insulin and IGF-II (20). IGF-II
upregulation has been reported in numerous tumor types (5, 6)
and frequently results from the loss of imprinting (LOI) of the IGF2
gene (21). Additionally, inactivating mutations or loss of heterozygosity of the gene encoding IGF-IIR, thought to act as a scavenger
for IGF-II, can also contribute to increased IGF-II bioavailability
(22, 23). This provides yet another alternate route for IGF-II
signaling via the IR and results in mitogenic and antiapoptotic
signals in tumors. Deregulated IGF-II expression in tumors and the
ability of this ligand to signal through the IR-A in addition to the
IGF-IR suggest that endogenous IR expression may be an important
determinant of sensitivity to IGF-IR mAbs.
In the present study, we provide evidence that IR, irrespective of
isoform type, mediates primary resistance to IGF-IR–targeted therapy and can be used as a potential biomarker for patient selection.
(TCGA) Data Portal. mRNA expression levels were estimated by
RSEM (http://deweylab.biostat.wisc.edu/rsem/) and then RNASeq by Expectation-Maximization (RSEM) expression estimates
were normalized to set the upper quartile count at 300. IGF-IR
(combined expression of two major IGF-IR isoforms, uc010bon.2
and uc002bul.2), IR-A (uc002mge.1), and IR-B (uc002mgd.1)
expression levels were calculated for 6,943 malignant samples
representing 21 tumor types.
Materials and Methods
Generation of stable cell lines that overexpress insulin receptor
isoforms
Plasmids with cDNA encoding the IR-A and IR-B sequence
were generated at Eli Lilly. cDNAs encoding IR isoforms
were amplified with 50 TATACTAGTATGGCCACCGGGGGAAGGAG- 30 and 50 -TAATCTAGACTAAGAAGGATTGGACCGAGGC- 30 primers, digested with Spe-I and Xba-I and ligated into
pLVX-IRES-puro HIV-1-based lentiviral expression vector
(Clontech). Lentiviruses were engineered by transfecting plasmids into HEK-293T cells using reagents and protocols provided with the Lenti-X Lentiviral Expression Systems (Clontech). A549, NCI-H1299, and MCF-7 cells were transduced
with the respective virus for 48 hours and then cultured in the
presence of puromycin (1 mg/mL) for selection of stably
transduced A549 (A549-Mock, A549-IR-A, and A549-IR-B),
NCI-H1299 (NCI-H1299-Mock, NCI-H1299-IR-A, and NCIH1299-IR-B), and MCF-7 (MCF-7-Mock, MCF-7-IR-A, and
MCF-7-IR-B) variants.
Materials
Chemicals and materials were obtained from the following
sources: CHAPS (Affymetrix); Cisplatin (Medac GmbH); Linsitinib (OSI-906; Selleckchem); Matrigel (BD Bioscience); Ham's F12
Nutrient Mix (F-12 HAM), RPMI 1640 medium, Improved Minimum Essential Medium (IMEM), NuPage 4% to 12% Bis-Tris,
iBlot Gel Transfer Stacks Nitrocellulose, and puromycin (Invitrogen); plasmids Lv-105-IGF-I, Lv-105-IGF-II, and Lv-105-IGF-IR
(GeneCopoeia); recombinant human IGF-I and IGF-II (R&D
Systems); Complete Protease Inhibitor Cocktail, PhosSTOP Phosphatase Inhibitor Cocktail, and recombinant human insulin
(Roche); Xba-I, Spe-I, Pierce Protein Assay, and Spectra Multicolor
Broad Range Protein Ladder (Thermo Scientific); CellTiter-Glo
Luminescent Cell Viability Assay (Promega); pemetrexed disodium heptahydrate (Eli Lilly).
Antibodies
The following antibodies were purchased from commercial
sources as indicated: mouse mAbs against Akt (#2920) and
p42/p44 MAPK (Erk1/2; #9107), rabbit mAbs against phosphoAktS473 (#4060) and phospho-IGF-IR beta Y1135/1136/Insulin
Receptor betaY1150/1151 (#3024), rabbit polyclonal antibody
against phospho-p42/p44 MAPKT202/Y204 (Erk1/2; #9101; Cell
Signaling Technology); mouse mAb against IGF-IR (#MS-641-P)
and Insulin Receptor (#MS-632-P; Thermo Fisher Scientific); rabbit
polyclonal Insulin Receptor (#sc-711) and (#sc-7953; Santa Cruz
Biotechnology); mouse mAb against IGF-II (MAB292) and goat F
(ab0 )2 anti-mouse IgG-phycoerythrin (#F0102B; R&D Systems);
goat polyclonal anti-mouse IRDye 680 conjugated (#926-32220)
and anti-rabbit IRDye 800 conjugated (#926-32211; LI-COR
Biosciences). Cixutumumab was supplied by Eli Lilly and Company. Control human IgG was purchased from Equitech-Bio Inc.
IGF-IR, IR-A, and IR-B mRNA analysis in The Cancer Genome
Atlas data sets
Receptor isoform expression was quantified using RNA-seq v2
(level 3) data downloaded from The Cancer Genome Atlas
1616 Mol Cancer Res; 13(12) December 2015
Patient specimens
Snap-frozen non–small cell lung cancer (NSCLC) and colorectal carcinoma tumor samples were provided by CureLine, Indivumed, and Tissue Solutions. Specimens were collected in compliance with all applicable regulations.
Cell cultures
Cell lines used in this study were obtained from the ATCC.
A549, NCI-H1299, and MCF-7 cells were cultured in F-12 HAM,
RPMI 1640, and IMEM medium, respectively, supplemented with
1 GlutaMAX supplement (Life Technologies) and 10% FBS at
37 C in 5% CO2 atmosphere and 95% humidity.
Colony formation assay
Colony formation assay was performed as described by Ulanet
and colleagues (24) with some minor modifications. Tested on
control (T/C) values below and above 50% were considered
sensitive and resistant, respectively.
Animal models
Experimental studies in xenograft models of pediatric solid
tumors were conducted through Pediatric Preclinical Testing
Program (PPTP) and have been described previously (25). In the
in vivo studies including xenograft A549 model with overexpression of IR-A and IR-B isoforms and patient-derived xenograft
(PDX) models of lung adenocarcinoma, 5- to 8-week-old female
nude mice (nu/nu) were used (The Jackson Laboratory; Harlan
Laboratories). To evaluate pharmacodynamics effects of cixutumumab, C57B1/6 male mice were used (Charles River Laboratories). Animals were maintained under barrier conditions, and
experiments were performed according to the institutional protocols (Eli Lilly, Oncotest) in compliance with the NIH Guide for
the Care and Use of Laboratory Animals and German Animal
Molecular Cancer Research
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
IR Confers Resistance to IGF-IR Antibody
Welfare Act (Tierschutzgesetz). Mice were kept on a 12-hour light/
dark cycle with access to a standard laboratory chow diet and fresh
water ad libitum.
To study the antitumor efficacy of cixutumumab in xenograft
A549 model with overexpression of IR isoforms, mice (n ¼ 30
for each variant) were subcutaneously injected with 2 107
A549-Mock, A549-IR-A, or A549-IR-B cells resuspended in
100% Matrigel. When tumors reached approximately 250 to
300 mm3, animals were randomized into two treatment groups
(0.9% USP Saline and cixutumumab; n ¼ 12 for each group).
USP Saline [10 mL/g of body weight (BW)] and cixutumumab
(60 mg/kg BW) were administered i.p. twice weekly for 5 weeks.
To evaluate the efficacy of cixutumumab in combination with
chemotherapy, nine PDX models of lung adenocarcinoma were
employed. LXFA-289, LXFA-297, LXFA-526, LXFA-623, LXFA629, LXFA-737, LXFA-749, LXFA-923, and LXFA-983 models
were derived from tumor specimens of patients treated at the
University Hospital in Freiburg, Germany, and directly implanted into nude mice according to the procedure established at
Oncotest GmbH. Additional information on the PDX models
including clinical annotations and mutational status of genes
commonly mutated in lung adenocarcinoma is provided in
Supplementary Table S2. The aforementioned models were
used to evaluate the antitumor activity of cixutumumab in
monotherapy and in combination with cisplatin and pemetrexed, a current standard of care in patients with metastatic
nonsquamous NSCLC. Mice were randomized into four treatment group (6–7 animals per group) when tumors reach 50 to
250 mm3 (control, cixutumumab, pemetrexed/cisplatin, cixutumumab/pemetrexed/cisplatin) hIgG and cixutumumab were
administered i.p. at 40 mg/kg of BW three times a week for a
duration of 3 to 4 weeks. Pemetrexed and its vehicle [0.9%
(w/v) NaCl] were given i.p. at 100 mg/kg of BW, daily except
weekends, for 2 to 4 weeks. Cisplatin (3.2 mg/kg BW) and its
vehicle (PBS) were administered subcutaneously once weekly
for 3 weeks.
Tumor size was measured twice weekly via calipers. Antitumor
activity was evaluated as maximum tumor volume inhibition
compared with the vehicle control group (optimal T/C values
calculated based on median values). Prior to euthanasia, tumors
were carefully excised and immediately snap-frozen in liquid
nitrogen for further analysis.
RNA extraction and cDNA preparation
Total RNA was isolated from snap-frozen tumor tissue using the
TissueLyser with stainless steel beads (5 mm) and the AllPrep
DNA/RNA Mini kit from QIAGEN or the MagMAX 96 Total RNA
isolation kit from Life Technologies. RNA concentration was
determined spectrophotometrically with the OD260 (260/280 >
1.9) and purity was verified with the Agilent Bioanalyzer. RNA was
reverse-transcribed with a high-capacity cDNA reverse transcription kit (Life Technologies) and random primers according
to the manufacturer's instructions. cDNA was then diluted 1:5
(20 ng/mL) for qPCR use.
PCR primers and probe sets
Commercial TaqMan Gene Expression assays were purchased
from Life Technologies for quantification of total IR (INSR,
Assay ID: Hs00961554_m1), IGF-IR (IGF-IR, Assay ID:
Hs00609566_m1), IGF-I (IGF-I, Assay ID: Hs01547656_m1),
IGF-II (IGF-II, Assay ID: Hs01005963_m1), and EGFR (EGFR,
www.aacrjournals.org
Assay ID: Hs01076078_m1). Full-length mRNA transcript
sequences for IR-A (NM_001079817) and IR-B (NM_000208)
isoforms were retrieved from the NCBI Reference Sequence
database. The IR-B TaqMan Gene Expression assay was
designed using the Custom TaqMan Assay Design Tool from
Life Technologies. Program input specified exon 11 inclusion;
final positioning of the probe spanned the exon 10/11 junction.
All probes incorporated a minor groove binding moiety and
were labeled with a fluorescent dye (FAM) for detection and
a nonfluorescent quencher. Custom-made primers/probe
sequence for IR-B includes: probe sequence: 50 -TCCCCAGAAAAACCTC-30 , forward primer: 50 -CCTGCACAACGTGGTTTTCG-30 and reverse primer: 50 -CGGCACCAGTGCCTGAA30 . To confirm specificity of the IR-B probe, spiking experiments
were performed using plasmids containing cDNA encoding
IR-A or IR-B. It was not possible to identify completely
specific TaqMan probes to IR-A. IR-A expression levels were,
therefore, calculated by subtracting IR-B values from the total
IR values.
Gene expression analysis
qPCR was performed in a 96-well format on 100 ng of cDNA
(20 ng/mL) with 1 mL of TaqMan Gene Expression Assay (20)
and 10 mL of TaqMan Gene Expression Master Mix (2) in a
final volume of 20 mL. Each sample was tested in triplicate. All
assays were run on Applied Biosystems 7500 Fast detection
system using standard setting (10 minutes incubation at 95 C,
followed by 40 cycles at 95 C for 15 seconds and 60 C for 1
minute). Absolute values (mRNA copies/ng of cDNA) were
derived from a standard curve generated with a serial dilution
of plasmid containing the cloned sequence of the target gene.
Each target standard curve demonstrated linearity with regression coefficients (r2 values) above 0.997. All assays had efficiency between 88% and 100% with a standard curve slope
between 3.3 and 3.6. Data were extracted with the 7500
Software V2.0.5 (Life Technologies).
Statistical analysis
Results are expressed as the mean SEM. Student t test, oneway ANOVA, two-factor ANOVA, ANOVA on repeated measurement (RM-ANOVA) followed by Tukey post-test and mixed
model were used for statistical analysis. Mixed model analysis
for repeated measures was used for the evaluation of the statistical
significance of tumor inhibition. The term mixed model refers to
the use of both random and fixed effects in the same analysis,
which is ideal for analyzing "unbalanced" data sets (due to death
of mice, etc.).
Results
IGF-IR, IR-A, and IR-B expression in human malignancies
To understand the prevalence of IGF-IR, IR-A, and IR-B
mRNA expression in patient tumor samples, we evaluated
mRNA levels of IGF-IR and IR isoforms using TCGA RNA-seq
data from 6,943 samples representing 21 tumor types. These
results suggest that expression of all three receptors significantly
varies between tumor types and individual tumors (Fig. 1A).
High levels of IGF-IR expression were observed in breast,
ovarian, prostate, head and neck, and squamous lung cancer
and melanoma. IR-A expression was observed in virtually all
tumor types, and was particularly high in clear cell renal cell
Mol Cancer Res; 13(12) December 2015
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
1617
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
Forest et al.
A
40,000
IGF-IR
IRA
IRB
35,000
Normalized count
30,000
25,000
20,000
15,000
10,000
5,000
Thyroid
Sarcoma
Renal
papillary
Prostate
EG
FR
Renal
Chromo
Pancreatic
Ovarian
Melanoma
SCCL
Lung adeno
L.G.glioma
HCC
H&N
GBM
Endometrial
-B
B
CRC
Cervical
ccRCC
Breast
Bladder
AML
0
4,000
2,000
Copies/ng cDNA
1,000
800
600
400
200
Lung
adenocarcinoma
SC C L
IR
-A
IR
IG
FIR
EG
FR
-B
IR
-A
IR
IR
IG
F-
EG
FR
-B
IR
-A
IR
IG
F-
IR
0
CRC
Figure 1.
A, IGF-IR, IR-A, and IR-B mRNA expression in major tumor indications according to the TCGA RNA-seq V2 data set (n ¼ 6,943). Expression levels were
estimated by RSEM followed by normalization of RSEM expression estimates to set the upper quartile count at 300. Green, combined expression of two
major IGF-IR isoforms (uc010bon.2 and uc002bul.2); red, IR-A (uc002mge.1); blue, IR-B (uc002mgd.1). Represented are individual and median values.
Abbreviations: acute myeloid leukemia (AML), clear cell renal cell carcinoma (ccRCC), colorectal carcinoma (CRC), glioblastoma (GBM), head and neck
cancer (H&N), hepatocellular carcinoma (HCC), low-grade glioma (L.G. glioma), lung adenocarcinoma (Lung adeno), squamous cell carcinoma of the
lung (SCCL), chromophobe renal cell carcinoma (Renal Chromo). B, IGF-IR, IR-A, IR-B, and EGFR mRNA expression in tumor samples obtained from
patients with lung adenocarcinoma (n ¼ 30), squamous cell carcinoma of the lung (n ¼ 39), and colon carcinoma (n ¼ 19). Represented are individual
and median values.
carcinoma. Many tumor samples also displayed significant IR-B
expression, with highest levels observed in clear cell renal cell
carcinoma and hepatocellular carcinoma. We further verified
RNA-seq data by a more accurate qPCR analysis of clinical
NSCLC and colon cancer samples. Median levels of IGF-IR
transcripts were 392, 680, and 328 copies/ng cDNA in lung
adenocarcinoma, squamous cell carcinoma of the lung, and
colorectal carcinoma, respectively (Fig. 1B). Median levels of
1618 Mol Cancer Res; 13(12) December 2015
IR-A and IR-B mRNA were 230 and 120 copies/ng cDNA in lung
adenocarcinoma, 256 and 106 copies/ng cDNA in squamous
cell carcinoma of the lung, and 328 and 85 copies/ng cDNA in
colorectal carcinoma. Furthermore, in colorectal carcinoma
samples IR-A and IR-B expression was several fold higher
compared with expression levels of EGFR, a clinically validated
target in this tumor type. Taken together, these data indicate
that both IR isoforms are broadly expressed in human tumors.
Molecular Cancer Research
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
IR Confers Resistance to IGF-IR Antibody
B
4.0 × 10 4
IR
IGF-IR
4
r = 0.4342
P = 0.0103
3 × 104
2 × 104
1 × 104
0
2.0 × 10 4
0
100
50
150
T/C%
1.0 × 10 4
BT
KT-29
KT-14
KT-12
KT-10
SK K T 1 1
-N -13
EP
EW-1
EW5
CH TC 8
LA -71
2
Rh58
Rh10
R 28
Rh h30
3
Rh0R
Rh41
BT 18
BT-45
BT-50
GB -36
BTM2
D639
D445
5
NBNB- 6
S
NB-17 D
7
NB-16 1
9
NB -EB 1
-1 c1
6
OS43
OS-1
O -2
OSS-9
OS-33
-3
1
0
IGF-IR protein levels
ECL signal
3.0 × 10
4 × 104
IR protein levels
A
4 × 104
r = –0.3333
P = 0.0541
3 × 104
2 × 104
1 × 104
0
0
50
100
T/C%
150
Figure 2.
Insulin receptor (IR) protein levels negatively correlate with antitumor efficacy of cixutumumab monotherapy in preclinical models of pediatric solid tumors in vivo.
A, total IR and IGF-IR protein levels in tumor lysates were measured by Meso Scale Discovery Electrochemiluminescence assay. B, correlation analysis of IR and
IGF-IR protein levels and antitumor efficacy of cixutumumab expressed as T/C% values. Correlation coefficient (Pearson r) and P value (P) are indicated.
Total IR expression inversely correlates with antitumor efficacy
of cixutumumab monotherapy in preclinical models of
pediatric solid tumors
As previously demonstrated (25), single agent cixutumumab
exhibited antitumor activity in vivo (T/C < 50%) in 16 of 34
xenograft models of pediatric solid tumors (rhabdoid, Ewing's
sarcoma, rhabdomyosarcoma, glioblastoma, neuroblastoma,
and osteosarcoma) tested through PPTP.
It is important to state that cixutumumab exhibits equipotent
binding to both human and mouse IGF-IR and is therefore
suitable for xenograft studies in mice. We observed elevated levels
of growth hormone, IGF-I, and insulin in the circulation of mice
treated with cixutumumab (Supplementary Fig. S1). Therefore,
mouse models properly mimic the pharmacodynamic changes
seen in humans (Eli Lilly, data on file; refs. 10, 11).
To understand if baseline IR and IGF-IR expression in tumor
tissue is predictive of cixutumumab efficacy in vivo, tissue samples
collected from individual tumor models were analyzed for IR and
IGF-IR by ECL assay and qPCR. The amount of IR and IGF-IR
protein and mRNA varied significantly across different tumor
models (Fig. 2). However, a significant correlation was observed
between mRNA and protein levels of both IR and IGF-IR (r ¼ 0.82
and 0.91, respectively, P < 0.0001) within individual tumor
samples. The impact of the two receptors on the outcome of
cixutumumab therapy, though, differed greatly. Although IGF-IR
expression levels tended to directly correlate with sensitivity to the
antibody, high IR was associated with higher T/C values indicative
of de novo or intrinsic resistance (Figs. 2 and 3A and B). Expression
of IR isoforms, IR-A and IR-B, was also evaluated by qPCR
analysis. Although IR-A was the predominant isoform, ubiquitously and robustly expressed in various pediatric solid tumors
(range: 0–406; mean, 157 copies/ng cDNA), expression of IR-B
was rather weak (range: 0–210; mean, 12 copies/ng cDNA) and
restricted to fewer tumor models. In contrast to total IR, however,
expression of either IR isoform failed to correlate with cixutumumab efficacy suggesting that both IR isoforms can potentially
contribute to cixutumumab resistance (Fig. 3C and D).
www.aacrjournals.org
It is known that IGF-I primarily stimulates IGF-IR, whereas IGFII is capable of activating both IGF-IR and IR-A. We aimed to
elucidate which of the two ligands is expressed in pediatric solid
tumors. As demonstrated by qPCR, IGF-I expression was barely
detectable in most samples (range: 0–1,100; mean, 74 copies/ng
cDNA), whereas IGF-II was expressed at markedly high levels
(range: 0–107,723; mean, 23,657 copies/ng cDNA) in a significant number of tumor models (Supplementary Table S1). It is
therefore plausible that in cixutumumab-treated tumors with
high IR expression, IGF-II may be capable of promoting tumorigenesis, bypassing the IGF-IR.
Total IR, IR-A, and IR-B expression correlates with antitumor
efficacy of cixutumumab in combination with cisplatin/
pemetrexed in PDX models of NSCLC
Clinical data indicate that IGF-IR antibodies display rather
weak single agent activity in epithelial tumors, and in most trials
they were used in combination with other antitumor agents
including chemotherapy. We therefore tried to understand if
baseline IGF-IR and IR expression was associated with the antitumor efficacy of cixutumumab used in combination with pemetrexed and cisplatin, which is a standard of care in NSCLC with
nonsquamous histology. When combined with chemotherapy,
cixutumumab displayed significantly improved efficacy over chemotherapy alone in 3 out of 9 patient-derived models of lung
adenocarcinoma (Fig. 4; Supplementary Table S2). To identify
predictors of cixutumumab sensitivity or resistance, individual
lung adenocarcinoma models were analyzed for baseline levels of
total IGF-IR, total IR, IR-A, IR-B, IGF-I, and IGF-II expression by
qPCR (Fig. 5A and Supplementary Fig. S2). In contrast to pediatric
solid tumors, IGF-IR expression was not predictive of the efficacy
of the IGF-IR mAb in lung adenocarcinoma models. Expression of
total IR and, more importantly, IR-A and IR-B was associated with
the outcome of cixutumumab therapy (Fig. 5B). Of note,
although IR-A was the predominant isoform (range: 12–301;
mean, 89 copies/ng cDNA), IR-B expression (range: 19–123;
mean, 61 copies/ng cDNA) was significantly higher in lung
Mol Cancer Res; 13(12) December 2015
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
1619
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
Forest et al.
B
700
r = 0.3242
600
P = 0.0702
IGF-IR copies/ng cDNA
Total IR copies/ng cDNA
A
500
400
300
200
100
0
3,000
r = - 0.4317
P = 0.0136
2,000
1,000
0
0
50
100
150
0
50
T/C %
150
T/C %
500
r = 0.2903
400
P = 0.1070
D
IR-B copies/ng cDNA
IR-A copies/ng cDNA
C
100
300
200
100
0
240
r = 0.3822
210
P = 0.2461
50
40
30
20
10
0
0
50
100
150
T/C %
0
50
100
150
T/C%
Figure 3.
Expression of total IR rather than individual isoforms correlates with antitumor efficacy of cixutumumab monotherapy in preclinical models of pediatric solid
tumors. Correlation analysis of total IR (A), total IGF-IR (B), IR-A (C), and IR-B (D) mRNA expression and antitumor efficacy of cixutumumab expressed
as T/C% values. Correlation coefficient (Pearson r) and P value (P) are indicated.
adenocarcinoma compared with the aforementioned pediatric
solid tumors. In addition, IGF-I expression was negligible while
IGF-II was expressed (range: 1–2,302; mean, 490 copies/ng
cDNA) in most lung adenocarcinoma models at various levels
(Supplementary Fig. S2). These results further support our initial
observation suggesting that (1) IGF-II rather than IGF-I is
expressed in most tumors and (2) both IR isoforms seem to
contribute to de novo cixutumumab resistance.
Two insulin receptor isoforms confer resistance to
cixutumumab in NSCLC and breast cancer models
To understand if there is a mechanistic link between individual
IR isoforms and cixutumumab resistance, we developed stably
transduced variants of NSCLC and breast cancer cell lines (A549,
NCI-H1299, and MCF-7) that overexpress either IR-A or IR-B.
A549, NCI-H1299, and MCF-7 cells display sensitivity to cixutumumab when tested in vitro in anchorage-independent colony
formation assay or anchorage-dependent cell viability assay.
Furthermore, A549 cells exhibit sensitivity to cixutumumab in
nude mice. IR expression was markedly increased in IR-A and IRB-overexpressing A549, NCI-H1299, and MCF-7 cells as demonstrated by qPCR, Western blot, and/or flow cytometry. Cell surface
IGF-IR expression, however, did not differ significantly between
A549-Mock and A549-IR-overexpressing variants (Supplementary Fig. S3A). Furthermore, A549-Mock cells were responsive to
stimulation with IGF-I, IGF-II, and, to a lesser extent, insulin as
1620 Mol Cancer Res; 13(12) December 2015
exemplified by increased phosphorylation of Akt and/or IGF-IR/
IR. ERK1/2 phosphorylation was not significantly altered in
mock-transduced A549 cells. Although IGF-I–induced signal
transduction was not very different in mock- and IR-overexpressing cells, insulin potently stimulated downstream signaling in
both IR-overexpressing variants as demonstrated by increased
phosphorylation of IGF-IR/IR, Akt, and ERK1/2. In addition,
treatment of A549-IR-A and A549-IR-B cells with IGF-II also
resulted in increased phosphorylation of IGF-IR/IR and ERK1/
2, which was barely seen in A549-Mock cells (Supplementary Fig.
S3B). These results are in agreement with previously published
data indicating that both insulin and IGF-II are capable of binding
to IR, although IGF-II affinity for IR-B is several fold lower
compared with IR-A (26, 27).
The efficacy of cixutumumab was then tested in A549 and NCIH1299 variants using colony formation assay (Fig. 6A and B). As
expected, in A549-Mock and NCI-H1299-Mock variants, the
antibody inhibited colony formation by 70% to 80% (T/C
19.5% and 32.8%, respectively). However, IR-A overexpression
fully abolished this effect (T/C 93.8% and 76.8% for A549 and
NCI-H1299 cells, respectively). Ectopic expression of IR-B
resulted in a partial cixutumumab resistance (T/C 45.4% and
57.6% for A549 and NCI-H1299 cells, respectively). Moreover,
the efficacy of cixutumumab was tested in MCF-7 variants under
anchorage-dependent conditions. In line with soft-agar assay
data, the mock-transfected line exhibited cixutumumab
Molecular Cancer Research
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
IR Confers Resistance to IGF-IR Antibody
1,000
*
500
0
11 14 18 21 25 28
0
Days aer treatment iniaon
400
200
0
1,500
3
7
0
LXFA 297
3
Tumor volume, mm
3
Tumor volume, mm
400
200
0
0
3
7
10
3
7
14
17
Days aer treatment iniaon
21
10 14 17 21 24 28
7
11 14 18 21 25 28
Ctrl
Cix.
Cis. / Pem.
Cis. / Pem. / Cix.
1,200
1,000
800
600
400
200
0
0
600
400
200
0
2
6
9
13
16
20
Days aer treatment iniaon
2
6
9
13 16 20 23 27 29
Days aer treatment iniaon
Ctrl
Cix.
Cis. / Pem.
Cis. / Pem. / Cix.
800
0
4
LXFA 289
LXFA 749
1,000
Ctrl
Cix.
Cis. / Pem.
Cis. / Pem. / Cix.
600
0
Days aer treatment iniaon
Days aer treatment iniaon
800
*
1,400
500
10 14 17 21 24 29
1,000
200
Days aer treatment iniaon
1,000
0
0
10 14 17 21 24 28
Ctrl
Cix.
Cis. / Pem.
Cis. / Pem. / Cix.
3
Tumor volume, mm
3
Tumor volume, mm
600
400
LXFA 983
2,000
Ctrl
Cix.
Cis. / Pem.
Cis. / Pem. / Cix.
800
7
600
Days aer treatment iniaon
LXFA 526
1,000
3
Ctrl
Cix.
Cis. / Pem.
Cis. / Pem. / Cix.
800
0
3
7
200
Tumor volume, mm
4
400
3
0
600
Tumor volume, mm
0
800
LXFA 623
1,000
Ctrl
Cix.
Cis. / Pem.
Cis. / Pem. / Cix.
3
3
Tumor volume, mm
3
Tumor volume, mm
1,500
LXFA 629
1,000
Ctrl
Cix.
Cis. / Pem.
Cis. / Pem. / Cix.
Tumor volume, mm
LXFA 923
2,000
22
1,600
1,400
1,200
1,000
800
600
400
200
0
LXFA 737
Ctrl
Cix.
Cis. / Pem.
Cis. / Pem. / Cix.
0
3
7
10
14
17
21
Days aer treatment iniaon
Figure 4.
Growth curves of patient-derived lung adenocarcinoma xenograft tumors in response to cixutumumab, cisplatin/pemetrexed, or cixutumumab/cisplatin/
pemetrexed. For the evaluation of the statistical significance of tumor inhibition, the mixed model analysis for repeated measures was used. For comparisons
between the triple combination therapies and the respective monotherapy or dual therapy, the test results are given for the day on which the optimal T/C value
of the triple combination therapy group was recorded. Statistically significant difference between cisplatin/pemetrexed or cixutumumab/cisplatin/
pemetrexed groups (P 0.05).
sensitivity (maximum inhibition 60.4%), while the IR-A overexpressing cells were completely resistant (maximum inhibition
0%) and the IR-B overexpressing cells were partially resistant
(maximum inhibition 26.9%; Supplementary Fig. S4). To corroborate these findings in vivo, mice bearing A549-Mock, A549-IRA, or A549-IR-B xenograft tumors were treated with cixutumumab
for up to 33 days (Fig. 6C). A549-IR-A tumors grew more rapidly
compared with A549-Mock and A549-IR-B xenografts (RMANOVA followed by Tukey multiple comparison test; P <
0.01). More importantly, treatment with cixutumumab resulted
in a statistically significant (P ¼ 0.0014) tumor growth inhibition
in A549-Mock model only. Consistent with our in vitro findings,
overexpression of IR-A and, to a lesser extent, IR-B led to cixutumumab resistance.
IGF-II mediates resistance to cixutumumab in NSCLC cells with
IR-B overexpression
A previous study has shown that IGF-II acting though IR-A
could be accountable for the limited efficacy of anti–IGF-IR
therapy (24). To understand if IR-B–mediated resistance to
cixutumumab is driven by IGF-II, we examined the effect of
www.aacrjournals.org
cixutumumab in the presence or absence of neutralizing
anti–IGF-II mAb (clone 75015) in A549-IR-B cells. Both
cixutumumab and anti–IGF-II mAb only partially inhibited
clonogenic potential of A549 cells with ectopic IR-B overexpression. Combination of the two antibodies, however,
completely reversed drug-resistant phenotype suggesting that
IGF-II is implicated in the development of cixutumumab
resistance in tumor cells with of IR-B overexpression (Fig.
7A). We also confirmed that treatment with IGF-II at the
concentration capable of binding to IR-B (>20 nmol/L; ref. 28)
resulted in marked phosphorylation of IR, Akt, and ERK1/2 in
cixutumumab-pretreated A549-IR-B cells (Fig. 7B). A similar
approach was employed with linsitinib, a dual IGF-IR/IR
inhibitor. In contrast to cixutumumab, linsitinib abrogated
IGF-II-mediated Akt and ERK1/2 phosphorylation in A549IR-B cells (Fig. 7C). Furthermore, linsitinib was equipotent in
inhibiting A549 and MCF-7 tumor cells irrespective of insulin
receptor isoform expression (Fig. 7D). Collectively, these data
indicate that IGF-II is capable of transducing downstream
signal via IR-B in tumor cells with antibody-mediated blockade of IGF-IR.
Mol Cancer Res; 13(12) December 2015
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
1621
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
Forest et al.
500
IR copies/ng cDNA
IGF-IR copies/ng cDNA
2,500
2,000
1,500
1,000
500
300
200
100
2,000
1,500
1,000
500
0
20
40
60
80
T/C % (Cis./Pem./Cix.)
FA
LX
300
200
100
100
r = 0.8107
P = 0.0080
400
300
200
100
0
400
0
0
0
20
40
60
80 100
T/C % (Cis./Pem./Cix.)
r = 0.8127
P = 0.0078
500
IR copies/ng cDNA
r = 0.2879
P = 0.4525
2,500
LX - 9
FA 2 3
LX - 6
FA 2 9
LX - 6
FA 2 3
LX - 5
FA 2 6
LX - 9
FA 8 3
LX - 2
FA 8 9
LX - 2
FA 97
LX - 7
FA 49
-7
37
0
IR-B copies/ng cDNA
IGF-IR copies/ng cDNA
IR-A copies/ng cDNA
B
IR-A
IR-B
400
0
LX
FA
LX - 9
FA 2 3
LX - 6
FA 29
LX - 6
FA 2 3
LX - 5
FA 26
LX - 9
FA 83
LX - 2
FA 8 9
LX - 2
FA 9 7
LX - 7
FA 49
-7
37
A
0
r = 0.7609
P = 0.0173
150
100
50
0
0
Discussion
IGF-IR mAbs have been tested in multiple clinical trials (www.
clinicaltrials.gov). Although overall results in intention-to-treat
population were generally negative, in some clinical trials subsets
of patients seem to benefit from IGF-IR therapy (8–11). Robust
molecular markers predictive of the efficacy of this class of agents
remain largely unidentified.
The objective of our study was to explore the role of the insulin
receptor as a potential negative predictive biomarker for IGF-IR–
targeted therapy. To our knowledge, this study is the first report
demonstrating the adverse impact of both IR isoforms on the
efficacy of IGF-IR mAb, cixutumumab.
The role of insulin signaling in cancer biology has recently
received a considerable amount of attention. Based on epidemiologic studies, obesity and type II diabetes were found to be
associated with increased cancer risk and mortality (29). Compensatory hyperinsulinemia and/or insulin therapy utilized to
treat type II diabetes can activate the IR signaling pathway in
tumor cells leading to enhanced tumorigenesis (30). Moreover, a
number of studies described the role of IR in different tumor types
1622 Mol Cancer Res; 13(12) December 2015
20
40
60
80 100
T/C % (Cis./Pem./Cix.)
Figure 5.
A, total IGF-IR, IR-A, and IR-B mRNA
expression in patient-derived
xenograft models of lung
adenocarcinoma (n ¼ 3 per tumor
model). Represented are mean values
SEM. B, correlation analysis of total
IGF-IR, total IR, IR-A, and IR-B mRNA
expression and antitumor efficacy of
cixutumumab/cisplatin/pemetrexed
expressed as T/C% values. Correlation
coefficient (Pearson r) and P value (P)
are indicated.
20
40
60
80 100
T/C % (Cis./Pem./Cix.)
(28). A compensatory mechanism between IGF-IR and closely
related IR has been demonstrated: genetic, shRNA-mediated, or
pharmacologic inactivation of IGF-IR can result in IR upregulation (31–34). In line with these observations, we found that
baseline IR expression in tumors could negatively affect the
efficacy of anti–IGF-IR therapy. In our proof-of-concept experiments, we demonstrated that ectopic IR expression is indeed
sufficient to alleviate the effect of cixutumumab in otherwise
sensitive cell lines. We also analyzed IR expression levels in a
large panel of tumor models and noted a negative correlation
between IR and the efficacy of cixutumumab.
In previously published studies, the role of individual IR isoforms as a potential mechanism of intrinsic resistance to IGF-IR
targeting had not been dissected, although studies in Ewing's
sarcoma cell lines with almost exclusive expression of IR-A
(100% of total IR) support a role of this isoform in the development of drug resistance (34). The role of IR-B in this process,
however, remained poorly understood.
In normal tissue, the two IR isoforms mediate very distinct
functions. IR-B, which primarily is expressed in the liver, muscle, and fat tissue, regulates metabolic functions, whereas IR-A,
Molecular Cancer Research
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
IR Confers Resistance to IGF-IR Antibody
% of Colony formaon
150
125
150
A549
hlgG
Cix.
% of Colony formaon
A
100
75
50
25
0
Mock
IR-A
A549-Mock
B
100
50
0
IR-B
NCI-H1299
hlgG
Cix.
Mock
IR-A
A549-IR-A
IR-B
A549-IR-B
hIgG
Cix.
1,500
1,000
*
500
0
1
2,500
Control
Cix.
5 8 12 15 19 22 26 29 33
Days aer treatment iniaon
2,000
A549-IR-A
1,500
1,000
500
0
1
2,000
Control
Cix.
3
A549-Mock
Tumor volume, mm
Tumor volume, mm
3
Control
Cix.
3
2,000
Tumor volume, mm
C
5
8 12 15 19 22 26
Days aer treatment iniaon
A549-IR-B
1,500
1,000
500
0
1
5 8 12 15 19 22 26 29 33
Days aer treatment iniaon
Figure 6.
Two IR isoforms confer resistance to cixutumumab in vitro and in vivo. A, clonogenic assay with A549 and NCI-H1299 cells, with or without IR-A or IR-B
overexpression, in the presence or absence of cixutumumab. Represented are mean values SEM, expressed as percentage of colony formation relative to
hIgG-treated cells. , P < 0.05; , P < 0.01 (two tail Student t test). B, representative images of clonogenic assay in A549 cells. C, antitumor efficacy of cixutumumab in
mice bearing xenograft A549 tumors that overexpress empty vector, IR-A, or IR-B models. RM-ANOVA was used to compare tumor growth between
treatment groups. P values of 0.05 indicate statistical significance ( ) of tumor inhibition.
highly expressed in various tissues during prenatal life, mediates
proliferative and antiapoptotic effects (20). It is currently
believed that IR-A rather than IR-B is implicated in tumor
growth because it is frequently upregulated in tumor tissue
(28) and has a greater affinity for IGF-II, a growth factor also
known to be markedly overexpressed in some tumors as a result
of LOI (35). However, several lines of evidence also indicate that
expression levels of the B isoform in tumor cells are quite
significant, representing up to 50% of the total IR pool in breast,
lung, and colorectal tumors (28). We therefore aimed to understand the roles of the two IR isoforms in the resistance to IGF-IR
mAbs. Our results demonstrate that ectopic expression of either
IR-A or IR-B in tumor cells resulted in the resistance to cixutumumab in vitro and in vivo, with IR-A inducing a stronger effect
compared with IR-B. Expression of either isoform indicated an
inverse effect on the antitumor activity of IGF-IR mAb in
patient-derived models of lung adenocarcinoma, further supporting the role of the "metabolic" B isoform as an important
determinant of cixutumumab efficacy.
Earlier publications reported different ligand specificity for the
two IR isoforms. Insulin is known to bind to both IR-A and IR-B.
www.aacrjournals.org
IGF-II binds to IR-A with an affinity similar to insulin and seems
to be also capable of binding to IR-B with several fold lower
affinity (26, 36). Our in vitro experiments showed that treatment
with physiologic concentrations of IGF-II induced phosphorylation of IR in both A549 IR-overexpressing cell lines, with
downstream PI3K/Akt and MAPK pathway activation. This effect
was more prominent in A549-IR-A cells, but was also quite
substantial in A549-IR-B cells. Moreover, neutralization of
IGF-II in culture media conditioned by A549-IR-B cells was able
to restore tumor sensitivity to cixutumumab, suggesting that
IR-B-driven resistance to anti–IGF-IR mAb is indeed mediated by
IGF-II. Interestingly, A549-IR-B xenografts exhibit sensitivity to
cixutumumab initially but became fully resistant to the antibody
4 weeks after treatment initiation. Given that in mice IGF-II
expression levels are negligible after birth (35), it is tempting to
speculate that IR-B-mediated resistance to IGF-IR mAb can only
be observed if large amounts of IGF-II are produced within IR-Bexpressing tumor tissue. Because IGF-IR and IR can form hybrid
receptors (1), it is also possible that resistance mediated by both
IR isoforms is at least in part caused by IGF-II-mediated signaling
via hybrid receptors, as a result of ligand binding to the IR arm.
Mol Cancer Res; 13(12) December 2015
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
1623
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
Forest et al.
A
% of Colony formaon
B
hIgG
Cix.
An-IGFII
Cix. + an-IGFII
150
125
Phospho-IRbY1150/51/IGF-IRbY1135/36
60 kDa
a
75
Cix.
97 kDa
a
100
hlgG
50
abc
25
Phospho-AktS473
60 kDa
Akt
44 kDa
42 kDa
Phospho-p42/p44 MAPKT202/Y204
44 kDa
42 kDa
0
IGF-II, 25 nmol/L
C
D
p42/p44 MAPK
–
+
–
+
MCF-7
100
97 kDa
Phospho-IRbY1150/51/IGF-IRbY1135/36
60 kDa
Phospho-AktS473
% Inhibion
75
50
25
60 kDa
Akt
44 kDa
42 kDa
Phospho-p42/p44 MAPKT202/Y204
44 kDa
42 kDa
p42/p44 MAPK
–
–
+ – – – + + +
– 0.01 0.1 1 0.01 0.1 1
2
Mock
IRA
IRB
Log (Linsinib)(mmol/L)
A549
100
75
% Inhibion
IGF-II, 25 nmol/L
Linsinib (mmol/L)
0
–2
0
–25
50
25
2
0
–2
0
–25
Log (Linsinib)(mmol/L)
Figure 7.
Insulin-like growth factor II (IGF-II) mediates resistance to cixutumumab in tumor cells overexpressing IR-B. Clonogenic assay in A549 cells that overexpress IR-B (A)
in the presence or absence of cixutumumab and/or anti–IGF-II mAb. Represented are mean values SEM, expressed as percentage of colony formation
relative to control treated cells. Statistical significance was determined by one-way ANOVA followed by Tukey post-hoc analysis. Statistically significant difference
versus control treated cells (a), cixutumumab-treated cells (b), and anti–IGF-II-treated cells (c). B and C, Western blot analysis of signal transduction in
A549-IR-B cells treated with IGF-II (25 nmol/L), in the presence or absence of cixutumumab and linsitinib. Serum-starved cells were pretreated with cixutumumab
(100 nmol/L) or linsitinib (0.01, 0.1, or 1 mmol/L) for 15 minutes and 2 hours, respectively, followed by incubation with rhIGF-II (25 nmol/L) for 10 minutes.
Y1150/51
Y1135/36
/IGF-IRb
,
Proteins (15–20 mg) extracted from cultured cells were size-fractionated by SDS-PAGE and immunoblotted with anti–phospho-IRb
S473
T202/Y204
anti–phospho-Akt
, and anti–phospho-p42/p44 MAPK
antibodies. Total level of proteins was demonstrated by immunoblotting with antibodies
directed against total Akt and p42/p44 MAPK. D, MCF-7 and A549 cells with or without IR-A or IRB overexpression were treated with increasing concentrations
of linsitinib (0.00015–10 mmol/L) for 72 hours, and tumor cell viability was quantified by CellTiter-Glo assay. The results are expressed as the percentage of
inhibition of tumor cell viability.
However, because anti–IGF-IR antibodies, including cixutumumab, have been shown to be very proficient at receptor internalization and degradation (37), hybrid receptors are likely
effectively neutralized, and therefore not present in abundance
on tumor cells following treatment with anti–IGF-IR mAbs.
More than any other component of the IGF-IR signaling pathway, IGF-II is frequently deregulated in cancer. It is important to
note that mouse may not be the most appropriate species to
model tumor resistance to cixutumumab in vivo. In contrast to
humans, mice have very low levels of circulating IGF-II. However, in human tumors, IGF-II is suggested to act in an autocrine/
paracrine rather than endocrine manner. Loss of IGF2 gene
imprinting and IGF-IIR downregulation are two fairly common
phenomena that increase the bioavailability of IGF-II in the
vicinity of IGF-IR and IR on tumor cells (38). Moreover, IGF2
gene is markedly overexpressed in colorectal tumors, and recent
studies reported a substantial prevalence of IGF2 amplification
in this tumor type (39).
1624 Mol Cancer Res; 13(12) December 2015
IGF-II targeting may thus represent an interesting alternative
to the anti–IGF-IR blockade, with the advantage of reducing
signaling via IR. Of note, some pharmaceutical companies have
already initiated the development of therapeutic mAbs against
IGFs. BI-836845 and MEDI-573 are fully human mAbs currently in preclinical and clinical development, respectively,
which neutralize both IGF-I and IGF-II and are efficacious in
experimental tumor models (40, 41). In a phase I trial, MEDI573 demonstrated acceptable safety profile and tolerability
with no significant perturbations of glucose homeostasis. The
antitumor activity of MEDI-573 was modest (0% responses,
33% stable disease). These safety and efficacy findings should
be interpreted with caution as no changes in circulating growth
hormone were observed suggesting that negative feedback
regulation of growth hormone by IGFs was not altered during
treatment with MEDI-573 (42).
IGF-IR TKIs are known to block both IGF-IR and IR tyrosine
kinase activity (43), and IGF-II/IR-mediated tumor growth
Molecular Cancer Research
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
IR Confers Resistance to IGF-IR Antibody
could be suppressed by this class of agents, as demonstrated by
the present study. However, the antitumor efficacy of IGF-IR TKIs
may require an optimal exposure and tissue distribution, which
might be difficult to achieve in patients due to dose limiting
toxicities associated with IR inhibition and development of
insulin resistance and hyperglycemia in up to 40% of treated
patients (43).
Based on the current findings, we argue that discovery of
suitable biomarkers to guide patient selection should become a
priority to enable successful IGF-IR-targeted therapy. Previous
work based on expression profiling in sarcoma and neuroblastoma identified potential predictors of intrinsic and acquired
resistance to IGF-IR inhibition. Higher levels of IGF-I, IGF-II,
and IGF-IR were linked to a better response to IGF-IR TKI,
whereas IGFBP-3 and IGFBP-6 overexpression were observed in
resistant models (44). Similar studies in breast and colorectal
cancer indicate that IGF-IR expression together with adaptor
proteins IRS-1 and IRS-2 or IGF-II expression could be linked to
sensitivity to IGF-IR mAb treatment (45, 46). In addition,
tumor cells can overcome IGF-IR inhibition via an alternative
RTK such as EGFR (44, 47). Our work further extends previously disclosed data by demonstrating that that insulin receptor
mediates resistance to IGF-IR antibodies used in monotherapy
or in combination with cytotoxic agents in a number of various
indications including lung adenocarcinoma and breast carcinoma. This study also provides ample evidence that both IR
isoforms play an important role in the sensitivity to IGF-IRtargeted therapies, sheds a new light on the role of IR-B in
tumor biology, and provides a scientific rationale for a clinical
trial with a preselection of patients based on the IR expression
levels of the tumor.
Furthermore, although many pharmaceutical companies discontinued their IGF-IR programs in oncology, there is growing
evidence that the IGF-IR pathway is implicated in the development of resistance to novel targeted agents (ALK inhibitors, nextgeneration EGFR TKIs; refs. 48, 49). It is therefore possible that
there might be a need to reinvigorate the development of IGF-IR
antibodies as more targeted agents will be approved for treatment
of human malignancies.
Disclosure of Potential Conflicts of Interest
G.P. Donoho has ownership interest (including patents) in Eli Lilly & Co.
No potential conflicts of interest were disclosed by the other authors.
Authors' Contributions
Conception and design: A. Forest, D.L. Ludwig, M.A. Smith, R. Novosiadly
Development of methodology: A. Forest, Y. Wang, J.T. DeLigio, R. Novosiadly
Acquisition of data (provided animals, acquired and managed patients,
provided facilities, etc.): A. Forest, M. Amatulli, C.B. Damoci, Y. Wang,
C.A. Burns, N. Zanella, H.H. Fiebig, M.C. Prewett, D. Surguladze, J.T. DeLigio,
P.J. Houghton
Analysis and interpretation of data (e.g., statistical analysis, biostatistics,
computational analysis): A. Forest, Y. Wang, N. Zanella, H.H. Fiebig,
M.C. Prewett, D. Surguladze, J.T. DeLigio, R. Novosiadly
Writing, review, and/or revision of the manuscript: A. Forest, M. Amatulli,
D.L. Ludwig, Y. Wang, G.P. Donoho, N. Zanella, M.C. Prewett, D. Surguladze,
J.T. DeLigio, P.J. Houghton, M.A. Smith, R. Novosiadly
Administrative, technical, or material support (i.e., reporting or organizing
data, constructing databases): A. Forest, C.B. Damoci, G.P. Donoho, N. Zanella
Study supervision: G.P. Donoho, N. Zanella, H.H. Fiebig, M.C. Prewett,
M.A. Smith, R. Novosiadly
Acknowledgments
The authors apologize to those colleagues whose publications were not cited
owing to space limitations.
Grant Support
This work was funded in part by the NCI grants (NO1-CM-42216 and
CA165995).
The costs of publication of this article were defrayed in part by the payment of
page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received June 19, 2015; revised August 4, 2015; accepted August 5, 2015;
published OnlineFirst August 11, 2015.
References
1. Pollak M. The insulin and insulin-like growth factor receptor family in
neoplasia: an update. Nat Rev Cancer 2012;12:159–69.
2. Renehan AG, Zwahlen M, Minder C, O'Dwyer ST, Shalet SM, Egger M.
Insulin-like growth factor (IGF)-I, IGF binding protein-3, and cancer
risk: systematic review and meta-regression analysis. Lancet 2004;363:
1346–53.
3. Guevara-Aguirre J, Balasubramanian P, Guevara-Aguirre M, Wei M, Madia
F, Cheng CW, et al. Growth hormone receptor deficiency is associated with
a major reduction in pro-aging signaling, cancer, and diabetes in humans.
Sci Transl Med 2011;3:70ra13.
4. Baserga R. The IGF-I receptor in cancer research. Exp Cell Res 1999;253:1–6.
5. Khandwala HM, McCutcheon IE, Flyvbjerg A, Friend KE. The effects of
insulin-like growth factors on tumorigenesis and neoplastic growth.
Endocr Rev 2000;21:215–44.
6. Samani AA, Yakar S, LeRoith D, Brodt P. The role of the IGF system in cancer
growth and metastasis: overview and recent insights. Endocr Rev 2007;28:
20–47.
7. Yee D. Insulin-like growth factor receptor inhibitors: baby or the bathwater? J Natl Cancer Inst 2012;104:975–81.
8. Naing A, LoRusso P, Fu S, Hong DS, Anderson P, Benjamin RS, et al. Insulin
growth factor-receptor (IGF-1R) antibody cixutumumab combined with
the mTOR inhibitor temsirolimus in patients with refractory Ewing's
sarcoma family tumors. Clin Cancer Res 2012;18:2625–31.
9. Weigel B, Malempati S, Reid JM, Voss SD, Cho SY, Chen HX, et al. Phase 2
trial of cixutumumab in children, adolescents, and young adults with
www.aacrjournals.org
10.
11.
12.
13.
14.
15.
16.
refractory solid tumors: a report from the Children's Oncology Group.
Pediatr Blood Cancer 2014;61:452–6.
Rajan A, Carter CA, Berman A, Cao L, Kelly RJ, Thomas A, et al. Cixutumumab for patients with recurrent or refractory advanced thymic epithelial
tumours: a multicentre, open-label, phase 2 trial. Lancet Oncol 2014;15:
191–200.
Schwartz GK, Tap WD, Qin LX, Livingston MB, Undevia SD, Chmielowski
B, et al. Cixutumumab and temsirolimus for patients with bone and softtissue sarcoma: a multicentre, open-label, phase 2 trial. Lancet Oncol
2013;14:371–82.
Weinstein IB. Cancer. Addiction to oncogenes–the Achilles heel of cancer.
Science 2002;297:63–4.
Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA Jr., Kinzler
KW. Cancer genome landscapes. Science 2013;339:1546–58.
Yonesaka K, Zejnullahu K, Okamoto I, Satoh T, Cappuzzo F, Souglakos
J, et al. Activation of ERBB2 signaling causes resistance to the EGFRdirected therapeutic antibody cetuximab. Sci Transl Med 2011;3:
99ra86.
Engelman JA, Zejnullahu K, Mitsudomi T, Song Y, Hyland C, Park JO, et al.
MET amplification leads to gefitinib resistance in lung cancer by activating
ERBB3 signaling. Science 2007;316:1039–43.
Harbinski F, Craig VJ, Sanghavi S, Jeffery D, Liu L, Sheppard KA, et al.
Rescue screens with secreted proteins reveal compensatory potential of
receptor tyrosine kinases in driving cancer growth. Cancer Discov 2012;
2:948–59.
Mol Cancer Res; 13(12) December 2015
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
1625
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
Forest et al.
17. Wilson TR, Fridlyand J, Yan Y, Penuel E, Burton L, Chan E, et al. Widespread
potential for growth-factor-driven resistance to anticancer kinase inhibitors. Nature 2012;487:505–9.
18. Zhang Z, Lee JC, Lin L, Olivas V, Au V, LaFramboise T, et al. Activation of the
AXL kinase causes resistance to EGFR-targeted therapy in lung cancer. Nat
Genet 2012;44:852–60.
19. Mosthaf L, Grako K, Dull TJ, Coussens L, Ullrich A, McClain DA. Functionally distinct insulin receptors generated by tissue-specific alternative
splicing. EMBO J 1990;9:2409–13.
20. Belfiore A, Malaguarnera R. Insulin receptor and cancer. Endocr Relat
Cancer 2011;18:R125–47.
21. Cui H, Horon IL, Ohlsson R, Hamilton SR, Feinberg AP. Loss of imprinting
in normal tissue of colorectal cancer patients with microsatellite instability.
Nat Med 1998;4:1276–80.
22. Ghosh P, Dahms NM, Kornfeld S. Mannose 6-phosphate receptors: new
twists in the tale. Nat Rev Mol Cell Biol 2003;4:202–12.
23. Scott CD, Firth SM. The role of the M6P/IGF-II receptor in cancer: tumor
suppression or garbage disposal? Horm Metab Res 2004;36:261–71.
24. Ulanet DB, Ludwig DL, Kahn CR, Hanahan D. Insulin receptor functionally enhances multistage tumor progression and conveys intrinsic
resistance to IGF-1R targeted therapy. Proc Natl Acad Sci U S A 2010;
107:10791–8.
25. Houghton PJ, Morton CL, Gorlick R, Kolb EA, Keir ST, Reynolds CP, et al.
Initial testing of a monoclonal antibody (IMC-A12) against IGF-1R by the
Pediatric Preclinical Testing Program. Pediatr Blood Cancer 2010;54:921–6.
26. Benyoucef S, Surinya KH, Hadaschik D, Siddle K. Characterization of
insulin/IGF hybrid receptors: contributions of the insulin receptor L2 and
Fn1 domains and the alternatively spliced exon 11 sequence to ligand
binding and receptor activation. Biochem J 2007;403:603–13.
27. Frasca F, Pandini G, Scalia P, Sciacca L, Mineo R, Costantino A, et al. Insulin
receptor isoform A, a newly recognized, high-affinity insulin-like growth
factor II receptor in fetal and cancer cells. Mol Cell Biol 1999;19:3278–88.
28. Belfiore A, Frasca F, Pandini G, Sciacca L, Vigneri R. Insulin receptor
isoforms and insulin receptor/insulin-like growth factor receptor hybrids
in physiology and disease. Endocr Rev 2009;30:586–623.
29. Ferguson RD, Gallagher EJ, Scheinman EJ, Damouni R, LeRoith D. The
epidemiology and molecular mechanisms linking obesity, diabetes, and
cancer. Vitam Horm 2013;93:51–98.
30. Novosyadlyy R, Lann DE, Vijayakumar A, Rowzee A, Lazzarino DA, Fierz Y,
et al. Insulin-mediated acceleration of breast cancer development and
progression in a nonobese model of type 2 diabetes. Cancer Res 2010;70:
741–51.
31. Zhang H, Pelzer AM, Kiang DT, Yee D. Down-regulation of type I insulinlike growth factor receptor increases sensitivity of breast cancer cells to
insulin. Cancer Res 2007;67:391–7.
32. Buck E, Gokhale PC, Koujak S, Brown E, Eyzaguirre A, Tao N, et al.
Compensatory insulin receptor (IR) activation on inhibition of insulinlike growth factor-1 receptor (IGF-1R): rationale for cotargeting IGF-1R and
IR in cancer. Mol Cancer Ther 2010;9:2652–64.
33. Dinchuk JE, Cao C, Huang F, Reeves KA, Wang J, Myers F, et al. Insulin
receptor (IR) pathway hyperactivity in IGF-IR null cells and suppression of
downstream growth signaling using the dual IGF-IR/IR inhibitor, BMS754807. Endocrinology 2010;151:4123–32.
34. Garofalo C, Manara MC, Nicoletti G, Marino MT, Lollini PL, Astolfi A, et al.
Efficacy of and resistance to anti-IGF-1R therapies in Ewing's sarcoma is
dependent on insulin receptor signaling. Oncogene 2011;30:2730–40.
1626 Mol Cancer Res; 13(12) December 2015
35. Bergman D, Halje M, Nordin M, Engstrom W. Insulin-like growth factor 2
in development and disease: a mini-review. Gerontology 2013;59:240–9.
36. Denley A, Bonython ER, Booker GW, Cosgrove LJ, Forbes BE, Ward CW,
et al. Structural determinants for high-affinity binding of insulin-like
growth factor II to insulin receptor (IR)-A, the exon 11 minus isoform of
the IR. Mol Endocrinol 2004;18:2502–12.
37. Burtrum D, Zhu Z, Lu D, Anderson DM, Prewett M, Pereira DS, et al. A fully
human monoclonal antibody to the insulin-like growth factor I receptor
blocks ligand-dependent signaling and inhibits human tumor growth in
vivo. Cancer Res 2003;63:8912–21.
38. Ferguson-Smith AC. Genomic imprinting: the emergence of an epigenetic
paradigm. Nat Rev Genet 2011;12:565–75.
39. Seshagiri S, Stawiski EW, Durinck S, Modrusan Z, Storm EE, Conboy CB,
et al. Recurrent R-spondin fusions in colon cancer. Nature 2012;488:
660–4.
40. Gao J, Chesebrough JW, Cartlidge SA, Ricketts SA, Incognito L, Veldman-Jones M, et al. Dual IGF-I/II-neutralizing antibody MEDI-573
potently inhibits IGF signaling and tumor growth. Cancer Res 2011;71:
1029–40.
41. Friedbichler K, Hofmann MH, Kroez M, Ostermann E, Lamche HR,
Koessl C, et al. Pharmacodynamic and antineoplastic activity of BI
836845, a fully human IGF ligand neutralizing antibody, and mechanistic rationale for combination with rapamycin. Mol Cancer Ther
2014;13:399–409.
42. Haluska P, Menefee M, Plimack ER, Rosenberg J, Northfelt D, LaVallee T,
et al. Phase I dose-escalation study of MEDI-573, a bispecific, antiligand
monoclonal antibody against IGFI and IGFII, in patients with advanced
solid tumors. Clin Cancer Res 2014;20:4747–57.
43. Puzanov I, Lindsay CR, Goff LW, Sosman JA, Gilbert J, Berlin J, et al. A phase
I study of continuous oral dosing of OSI-906, a dual inhibitor of insulinlike growth factor-1 and insulin receptors in patients with advanced solid
tumors. Clin Cancer Res 2015;21:701–11.
44. Huang F, Greer A, Hurlburt W, Han X, Hafezi R, Wittenberg GM, et al. The
mechanisms of differential sensitivity to an insulin-like growth factor-1
receptor inhibitor (BMS-536924) and rationale for combining with EGFR/
HER2 inhibitors. Cancer Res 2009;69:161–70.
45. Mukohara T, Shimada H, Ogasawara N, Wanikawa R, Shimomura M,
Nakatsura T, et al. Sensitivity of breast cancer cell lines to the novel
insulin-like growth factor-1 receptor (IGF-1R) inhibitor NVP-AEW541
is dependent on the level of IRS-1 expression. Cancer Lett 2009;282:
14–24.
46. Zha J, O'Brien C, Savage H, Huw LY, Zhong F, Berry L, et al. Molecular
predictors of response to a humanized anti-insulin-like growth factor-I
receptor monoclonal antibody in breast and colorectal cancer. Mol Cancer
Ther 2009;8:2110–21.
47. Haluska P, Carboni JM, TenEyck C, Attar RM, Hou X, Yu C, et al.
HER receptor signaling confers resistance to the insulin-like growth
factor-I receptor inhibitor, BMS-536924. Mol Cancer Ther 2008;7:
2589–98.
48. Lovly CM, McDonald NT, Chen H, Ortiz-Cuaran S, Heukamp LC, Yan Y,
et al. Rationale for co-targeting IGF-1R and ALK in ALK fusion-positive lung
cancer. Nat Med 2014;20:1027–34.
49. Cortot AB, Repellin CE, Shimamura T, Capelletti M, Zejnullahu K, Ercan D,
et al. Resistance to irreversible EGF receptor tyrosine kinase inhibitors
through a multistep mechanism involving the IGF1R pathway. Cancer Res
2013;73:834–43.
Molecular Cancer Research
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.
Published OnlineFirst August 11, 2015; DOI: 10.1158/1541-7786.MCR-15-0279
Intrinsic Resistance to Cixutumumab Is Conferred by Distinct
Isoforms of the Insulin Receptor
Amelie Forest, Michael Amatulli, Dale L. Ludwig, et al.
Mol Cancer Res 2015;13:1615-1626. Published OnlineFirst August 11, 2015.
Updated version
Supplementary
Material
Access the most recent version of this article at:
doi:10.1158/1541-7786.MCR-15-0279
Access the most recent supplemental material at:
http://mcr.aacrjournals.org/content/suppl/2015/08/13/1541-7786.MCR-15-0279.DC1
Cited articles
This article cites 49 articles, 24 of which you can access for free at:
http://mcr.aacrjournals.org/content/13/12/1615.full.html#ref-list-1
E-mail alerts
Sign up to receive free email-alerts related to this article or journal.
Reprints and
Subscriptions
Permissions
To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at
[email protected].
To request permission to re-use all or part of this article, contact the AACR Publications Department at
[email protected].
Downloaded from mcr.aacrjournals.org on June 18, 2017. © 2015 American Association for Cancer Research.