-
Translational Cancer Mechanisms and Therapy
BRAF and MEK Inhibitors Increase PD-1-PositiveMelanoma Cells
Leading to a PotentialLymphocyte-Independent Synergism
withAnti–PD-1 AntibodyMartina Sanlorenzo1,2,3, Igor Vujic4,5,
Arianna Floris1, MauroNovelli2, Loretta Gammaitoni6,Lidia Giraudo6,
Marco Macagno1,6, Valeria Leuci1,6, Ramona Rotolo1,6, Chiara
Donini1,Marco Basiric�o6, Pietro Quaglino2, Maria Teresa Fierro2,
Silvia Giordano1,7, Maria Sibilia3,Fabrizio Carnevale-Schianca6,
Massimo Aglietta1,6, and Dario Sangiolo1,6
Abstract
Purpose: BRAF and MEK inhibitors (BRAF/MEKi) favor
mel-anoma-infiltrating lymphocytes, providing the rationale
forcurrent combinatorial trials with anti–PD-1 antibody. A por-tion
of melanoma cells may express PD-1, and anti–PD-1antibody could
have a direct antitumor effect. Here, we explorewhether BRAF/MEKi
modulate rates of PD-1þ melanoma cells,supporting an
additional—lymphocyte-independent—basisfor their therapeutic
combination with anti–PD-1 antibody.
Experimental Design:With data mining and flow cytometry,we
assessed PD-1, PD-L1/2 expression on melanoma cell lines(CCLE, N ¼
61; validation cell lines, N ¼ 7) and melanomatumors (TCGA, N ¼
214). We explored in vitro how BRAF/MEKi affect rates of PD-1þ,
PD-L1/2þ melanoma cells, andcharacterized the proliferative and
putative stemness features ofPD-1þ melanoma cells. We tested the
functional lymphocyte-independent effect of anti–PD-1 antibody
alone and in com-
bination with BRAF/MEKi in vitro and in an in vivo
immuno-deficient murine model.
Results: PD-1 is consistently expressed on a small subset
ofmelanoma cells, but PD-1þ cells increase to relevant rates
duringBRAF/MEKi treatment [7.3% (5.6–14.2) vs. 1.5% (0.7–3.2),P ¼
0.0156; N ¼ 7], together with PD-L2þ melanoma cells[8.5% (0.0–63.0)
vs. 1.5% (0.2–43.3), P ¼ 0.0312; N ¼ 7]. PD-1þ cells proliferate
less than PD-1� cells (avg. 65% less; t¼ 7 days)and are
preferentially endowed with stemness features. In vivo, thedirect
anti-melanoma activity of PD-1 blockage as monotherapywas
negligible, but its association with BRAF/MEKi significantlydelayed
the development of drug resistance and tumor relapse.
Conclusions: BRAF/MEKi increase the rates of PD-1þ melano-ma
cells that may sustain tumor relapse, providing a
lymphocyte-independent rationale to explore combinatory strategies
withanti–PD-1 antibody. Clin Cancer Res; 24(14); 3377–85. �2018
AACR.
IntroductionMetastatic melanoma is still deadly, despite novel
immuno-
modulatory and protein kinase inhibitor therapies. In
preclinicalstudies, combinations of anti–PD-1 antibody and target
therapywith BRAF/MEK inhibitors (BRAF/MEKi) had synergistic
effects,explained by an increased number and activity of
tumor-infiltrat-
ing lymphocytes (1, 2). This increase of tumor-infiltrating
lym-phocytes following BRAF/MEKi treatment is well documented(3,
4), but the tumors may evade the immune system throughexpressionof
programmeddeath-receptor-ligand 1 (PD-L1) and2(PD-L2). These
ligands bind and activate the programmed death-receptor 1 (PD-1) on
T-lymphocytes and suppress the antitumorresponse (5), whereas its
blockage—by anti–PD-1 antibody—restores the antitumor effect.
However, it was suggested that PD-1 is "ectopically"
expressedalso on melanoma cells, and that its activation could
promotetumor growth (6, 7). The biological relevance of these
findings isstill not clear, but PD-1þ melanoma cell subsets were
found topreferentially express tumor-initiating determinants (6,
7). Suchputative cancer stem cells could contribute to the
development ofdrug resistance and tumor relapse (8–10), which is a
main issuefor patients treated with BRAF/MEKi (11–13). In fact,
after aninitial rapid anti-tumor response, most patients experience
dis-ease progression despite ongoing treatment (11–13).
Therefore,there is the need to elucidate the relevance of PD-1þ
melanomacells during BRAF/MEKi treatment, and to define
therapeuticapproaches, which could contrast the development of
resistanceto target therapies.
Here, we evaluate the "ectopic" melanoma-intrinsic
PD-1expression and show that PD-1þ and PD-L2þ melanoma cells
1Department of Oncology, University of Turin, Turin, Italy.
2Department ofMedical Sciences, Section of Dermatology, University
of Turin, Turin, Italy.3Institute of Cancer Research, Department of
Medicine I, Comprehensive CancerCenter, Medical University of
Vienna, Vienna, Austria. 4The RudolfstiftungHospital, Department of
Dermatology, Vienna, Austria. 5Department of Derma-tology, Medical
University of Vienna, Vienna, Austria. 6Division of
MedicalOncology—Experimental Cell Therapy, Candiolo Cancer
Institute, FPO—IRCCS,Candiolo, Turin, Italy. 7CancerMolecular
Biology, CandioloCancer Institute, FPO-IRCCS, Candiolo, Turin,
Italy.
Note: Supplementary data for this article are available at
Clinical CancerResearch Online
(http://clincancerres.aacrjournals.org/).
Corresponding Author: Martina Sanlorenzo, University of Turin,
Candiolo Can-cer Institute, FPO-IRCCS, Km 3,95, SP142, 10060
Candiolo (Turin), Italy. Phone:39 011 9933521; Fax: 39 011 9933522;
E-mail: [email protected]
doi: 10.1158/1078-0432.CCR-17-1914
�2018 American Association for Cancer Research.
ClinicalCancerResearch
www.aacrjournals.org 3377
on June 30, 2021. © 2018 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst April 12, 2018; DOI:
10.1158/1078-0432.CCR-17-1914
http://crossmark.crossref.org/dialog/?doi=10.1158/1078-0432.CCR-17-1914&domain=pdf&date_stamp=2018-7-13http://clincancerres.aacrjournals.org/
-
increase during BRAF/MEKi treatment, sensitizing tumor cells
todirect anti–PD-1 antibody effects, thus delaying the
developmentof resistance to target therapy.
Materials and MethodsCancer cell line encyclopedia
We extracted from the cancer cell line encyclopedia (CCLE)
themRNA expression values of the 61 available melanoma cell
lines(http://www.broadinstitute.org/ccle, access September
2015).
The Cancer Genome AtlasWedownloaded fromTheCancerGenomeAtlas
(TCGA)portal
the clinical data (clinical data, pathology report) and the
mRNAsequencing data (gene) of the 470 melanoma samples
included(https://portal.gdc.cancer.gov/, access September 2015).
ThemRNA sequencing data were available for 469 samples. Geneswith
FPKM values >0.1 were considered expressed (14). Wematched each
mRNA sequencing file with the correspondingclinical and histologic
data to divide the melanomas in primary,regional metastases and
distant metastases and to exclude fromfurther analyses all the
samples with histologic evidence ofimmune infiltrate.
Cell linesMelanoma cell lines SKMEL2, SKMEL5, SKMEL28 were
obtained from NCI-Frederick Cancer/DCTD Tumor Repositoryin 2011,
and A375 (ATCC CRL-1619) from the ATCC in2013. Cell line identity
was performed by the bank of originusing morphology, karyotyping,
and PCR-based approaches.Mycoplasma detection was performed after
cell thawing by theUniversal Mycoplasma Detection Kit ATCC
(ATCC-30-1012K)according to the manufacturer's instructions. Cell
lines for experi-mentswereobtained from theoriginal cryopreserved
golden stockand experiments performed immediately after and for no
longerthan 6 months, no further cell identification was
performed.SKMEL2, SKMEL5, and SKMEL 28 were maintained in
RPMI(Sigma Aldrich) supplemented with 10% FBS (Gibco BRL).Melanoma
cell line A375 cell was maintained in DMEM (GibcoBRL) with the
addition of 2mmol/L glutamine and 15% FBS(Gibco BRL). All the cells
were propagated at 37�Cunder 5%CO2.Cells were passaged and
harvested from flasks using Accutasesolution (Gibco BRL).
Patient-derived samplesPatient-derived melanoma cell cultures
(mMel2, mMel3,
mMel7, and mMel11) were generated from surgical biopsies
ofmetastatic/locally advancedmelanoma, before any systemic
treat-ment (December 2010–June 2012). All patients provided
consentunder institutional review board approved protocols.
Technical
procedures andmelanoma cell cultures were previously
described(10, 15). Mycoplasma detection was performed by
MycoplasmaPCR Detection Kit (Applied Biological Materials Inc.,
MICRO-TECH s.r.l.) according to the manufacturer's instructions.
Thetest was done after cell thawing/just before the
experimentexecution. All the experiments were performed on
patient-derived cell culture of not more than 24-week-old.
Generation of hOct4.eGFP transduced cell linesThe previously
described lentiviral vector (14) was transduced
in melanoma primary cells resuspended in fresh
KODMEM-F12(GibcoBRL)with 10%FBSadding virus-conditionedmediumat
adose of 400 ng P24/100,000 cells. The lentiviral vector
pRRL.sin.PPT.hOct4.eGFP.Wpre (LV-Oct4.eGFP) was obtained as
previ-ously described (15). Briefly, the hOct4-eGFP cassette
fromphOct4.eGFP1 vector (ref. 16; kind gift from Dr. Wei Cui,
IRDB,Imperial College, London) was cloned into the transfer
vectorpRRL.sin.PPT.hPGK.eGFP.Wpre (ref. 17; kindly provided byDr.
Elisa Vigna, IRCCS Candiolo/University of Turin, Italy) inplace of
the hPGK.eGFP cassette. After 16 hours, cells werewashedtwice and
grown for a minimum of 10 days to reach steady-stateeGFP expression
and to rule out pseudotransduction before flow-cytometry analysis.
Technical procedure including transductioncontrols were previously
described (10).
DrugsThe BRAF inhibitor dabrafenib (GSK2118436) and the MEK
inhibitor trametinib (GSK1120212) were purchased from
Sell-eckchem. The anti–PD-1 antibody is the inVivoMAb
anti-humanPD-1(CD279), Clone: J110 from Bio X Cell. The isotype
controlantibody is the inVivoMAb mouse IgG1 isotype control,
Clone:MOPC-21 from Bio X Cell. Drugs were used accordingly
toprevious reports (7, 18, 19).
Flow cytometryAnalyses of melanoma cells were performed using a
CyanADP
cytometer (BeckmanCoulter s.r.l.).
Thefluorochrome-conjugatedmonoclonal antibodies included anti-PD-L1
PE (clone MIH1);anti-PD-L2 APC (clone MIH18); anti–PD-1 APC (clone
MIH4)from BD Biosciences. The negative staining threshold was
estab-lished by the addition of an isotype-matched control
tube.
In vitro proliferation assay and CFSE stainingTo evaluate the
proliferation rate, cells had been labeled with
5(6)-Carboxyfluorescein diacetateN-succinimidyl ester (CFSE),for
which fluorescence intensity decreased by half at each celldivision
per kit protocol (Sigma-Aldrich). Briefly, the CFSE dyesolution was
prepared accordingly to the number of cells tostain and added to
the previously washed cell pellet. After a first15-minutes
incubation at 37�C, cells were washed once withculture medium added
with 10% heat-inactivated serum andincubated in culture medium
added with 10% heat-inactivatedserum for 30 minutes at 37�C. An
aliquot of these labeled andcounted cells was read on a Flow
Cytometry Cyan (Cyan ADP,Beckman Coulter s.r.l.) and analyzed using
Summit Software(Daki Cytomation, Heverlee, Belgium) to set the
baselinefluorescence level. The remaining cells were seeded in
cultureunder experimental conditions. After 4 and 7 days, the
reduc-tion in fluorescence was quantified by flow cytometry. In
case ofdrug treatments, treated cells were compared with
untreatedcells which were also labeled with the same dye.
Translational Relevance
BRAF and MEK inhibitors lead to increased rates of mela-noma
cells "ectopically" expressing PD-1, supporting a
lym-phocyte-independent antitumor effect of anti–PD-1 antibody.This
provides further rationale for BRAF and MEK inhibitors/anti–PD-1
antibody combination therapies in metastatic mel-anoma
patients.
Sanlorenzo et al.
Clin Cancer Res; 24(14) July 15, 2018 Clinical Cancer
Research3378
on June 30, 2021. © 2018 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst April 12, 2018; DOI:
10.1158/1078-0432.CCR-17-1914
http://www.broadinstitute.org/cclehttps://portal.gdc.cancer.gov/http://clincancerres.aacrjournals.org/
-
Figure 1.
Melanoma cells express PD-1. A, Box plots of PD-1, PD-L1, and
PD-L2 Affymetrix mRNA expression values of 61 melanoma cell lines
included in theCCLE. B, Bar graph of PD-1 expression values
compared with s100B in the same 61 melanoma cell lines. C, Box
plots representing levels of PD-1 in the61 melanoma cell lines and
in the 18 T-cell neoplasia cell lines included in CCLE. D, Flow
chart of the TCGA data analysis. Samples with histologicimmune
infiltrate were excluded from further analyses. E, Box plots of
PD-1, PD-L1, and PD-L2 mRNA expression values of 214 melanomas
without histologicevidence of immune infiltrate.
BRAF and MEK Inhibitors Increase PD-1–Positive Melanoma
www.aacrjournals.org Clin Cancer Res; 24(14) July 15, 2018
3379
on June 30, 2021. © 2018 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst April 12, 2018; DOI:
10.1158/1078-0432.CCR-17-1914
http://clincancerres.aacrjournals.org/
-
Sanlorenzo et al.
Clin Cancer Res; 24(14) July 15, 2018 Clinical Cancer
Research3380
on June 30, 2021. © 2018 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst April 12, 2018; DOI:
10.1158/1078-0432.CCR-17-1914
http://clincancerres.aacrjournals.org/
-
In vitro cell viability assayTo test the number of viable,
metabolically active cells after
treatment with BRAF/MEKi alone or in combination with anti–PD-1
antibody we used a method based on the quantitation ofATP present
(CellTiter-Glo Luminescent Cell Viability Assay,Promega Italia
s.r.l) according to the manufacturer's protocol.
In vivoNOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice were pur-
chased from Charles River Laboratories Italia s.r.l. (Calco -
Lecco,Italy). Mice were injected subcutaneously with 1,5 � 106
A375melanoma cells and randomly assigned to treatment
groups.Treatments were started when tumors became palpable
andcontinued for 40 days. Dabrafenib and trametinib were
admin-istrated by oral gavage 5 consecutive days a week at a dose
ofrespectively 600 and 4 mg. Anti–PD-1 antibody and
respectiveisotype control mAb were injected intraperitoneally (200
mg perinjection) three times a week. Mice were sacrificed when
themaintumor diameter reached 2 cm or massive ulceration occurred.
Allprocedures were performed accordingly Institutional
ReviewBoard–approved protocols.
Statistical analysesThe statistical analyses were performed
using Stata 12.0 statis-
tical software (Stata), and Prism7 (GraphPad Software, Inc.).
Allvariables were tested for normal distribution with the
Shapiro–Wilk test, and none of them was found normally
distributed.Comparisons between two independent non-normally
distribut-ed groups were performed using the nonparametric
Wilcoxonrank-sum test. Comparisons between matched groups were
per-formed with Wilcoxon signed rank test. Correlations
betweenvariables were tested with the Spearman's rank correlation
test.Differences in tumor volumes were statistically assessed
usingrepeating measurements two-way ANOVA followed by
Sidakcorrection and with two-tails. P values less than 0.05 were
con-sidered statistically significant.
ResultsMelanoma cells express low but consistent levels of
PD-1
To investigate PD-1 expressiononmelanoma cells, we analyzedtwo
datasets: the CCLE and TCGA.
All 61 CCLE melanoma cell lines expressed PD-1 withmRNA values
comparable to those of PD-L1 and PD-L2 (Affy-metrix mRNA values:
PD-1 4.20 (3.81–4.65), PD-L1 4.63(3.73–7.84), and PD-L2 3.73
(3.97–8.50; Fig. 1A). Averagemelanoma PD-1 values were about 40% of
those of theestablished melanoma antigen s100B [Affymetrix
mRNAvalues: 10.39 (3.28–13.80)], used as an internal control
puttingthe mRNA values into perspective (Fig. 1B), and around 95%
of
the average PD-1 expression found in 18 T-cell neoplasia
celllines included in the CCLE [Affymetrix mRNA values:
4.39(3.98–5.34); Fig. 1C].
From the 470 TCGA patient-derived melanomas, we
matchedgene-expression data with corresponding histologic reports
andwe excluded all the samples with histologic evidence of
immuneinfiltrate, as those would interfere with the assessment of
mela-noma-intrinsic PD-1 expression (Fig. 1D). PD-1 was expressedin
99.5% of the samples, with a median expression comparablewith PD-L1
and PD-L2 (Fig. 1E). We did not find significantdifferences when we
compared samples with (N ¼ 100)and without (N ¼ 114) stromal cells
[median FPKM values32.5 (0–1282.8), and 64.7 (0.4–1461.3)
respectively, P ¼0.0970]. Furthermore, we observed positive
correlations betweenPD-1 andPD-L1 (r¼ 0.66,P 0.05; N ¼7; Fig. 2C;
Supplementary Fig. S2B], we found a significantincrease of PD-L2þ
melanoma cells during BRAF/MEKi treat-ment [8.5% (0.0–63.0) vs.
1.5% (0.2–43.3), P ¼ 0.0312; N ¼7; Fig. 2D; Supplementary Fig.
S2C].
In BRAFV600 mutant cells, the combination of BRAF andMEK
inhibitors led to the highest percentage of PD-1þ cells,
Figure 2.BRAF/MEK inhibitors increase the rates of PD-1þ and
PD-L2þ tumor cells in BRAFV600 and NRASQ61 mutant melanomas. A,
Representative flow cytometry plotsof a BRAFV600mutantmelanoma cell
line (A375) treatedwith dabrafenib (1 mmol/L), trametinib (5
nmol/L), the combination of dabrafenibþ trametinib (1
mmol/Lþ5nmol/L), and fotemustine (50 mg/mL) for 96 hours. DAPI
staining was used to identify viable cells. PD-1, PD-L1, and PD-L2
plots were performed considering onlyviable cells. Rates of (B)
PD-1–, (C) PD-L1–, and (D) PD-L2–positive melanoma cells untreated
and after treatment with BRAF/MEKi [dabrafenib þ trametinib(1
mmol/Lþ 5 nmol/L) in BRAFV600mutant cell lines and trametinib (5
nmol/L) in the NRASQ61mutant cell line for 96 hours;N¼ 7]; � ,
P
-
and showed the greatest anti-tumor effect, whereas in theNRASQ61
mutant cell line this was observed with MEK inhib-itor alone
(Supplementary Fig. S3). Fotemustine efficientlykilled tumor cells,
but did not significantly change levels ofPD-1þ cells (Fig.
2E).
Rates of PD-1þ melanoma cells increase in a time- and
drug-dependent manner during BRAF/MEKi treatment
We explored whether the percentage of PD-1þ melanoma cellswas
influenced by time or drug exposure. Increasing the time
oftreatment with BRAF/MEKi, we observed a progressive increaseof
the percentage of PD-1þ cells. After 8 days of treatment,
viablePD-1þ melanoma cells increased to 31.8% on average
(15.0%–50.3%;N¼ 3; Fig. 2F and G]. Following BRAF/MEKi
withdrawal,the rate of PD-1þ melanoma cells returned back to the
loworiginal value (Fig. 2H and I; N ¼ 3).
PD-1þ melanoma cells are more quiescent and presentstemness
features
Since during BRAF/MEKi treatment PD-1þ cells reach signifi-cant
percentages among the viable tumor cell population, we
compared their proliferative capabilities with PD-1� cells.
Weused a carboxyfluorescein succinimidyl ester (CFSE)
dye-basedassay, where the dye decrease corresponds to higher
mitoticactivity and faster proliferation rate. Treatment with
BRAF/MEKidecreased the overall proliferation (Fig. 3A) with PD-1þ
cellsproliferating less than the PD-1� counterparts; on
average16.1% less after 96 hours (N ¼ 3), and 65% less after 7
days(N ¼ 2; Fig. 3B; Supplementary Fig. S4).
PD-1 was reported to be preferentially expressed on
putativemelanoma cancer stem cells (6). To test this, we used a
lentiviralvector carrying eGFP under the transcriptional control of
theOct4 stemness gene promoter (LV-Oct4.eGFP; Fig. 3C; ref.
15).This system visualizes putative cancer stem cells as
eGFPþ,based on their selective ability to activate the Oct4
promoter. Inthree LV-Oct4.eGFP transduced patient-derived cell
lines(mMel2-Oct4, mMel3-Oct4, mMel7-Oct4), BRAF/MEKi led tooverall
eGFPþ cell enrichment (on average 2.4-fold), suggestinga lower
sensitivity of eGFPþ melanoma cells to these drugs(Fig. 3D).
Moreover, eGFPþ putative cancer stem cells wereenriched among PD-1þ
cells compared with PD-1� cells (onaverage 1.6-fold; N ¼ 3; Fig.
3E).
Figure 3.
PD-1þmelanoma cells have reduced proliferative potential and
show stemness features.A,Melanoma cell proliferation after 96 hours
of treatment with BRAF/MEKi[dabrafenibþ trametinib (1 mmol/Lþ 5
nmol/L) in BRAFV600 mutant cell lines and trametinib (5 nmol/L) in
the NRASQ61 mutant cell line], compared with untreatedcontrols
measured with CFSE assay. N ¼ 3 (A375, SKMEL2, and SKMEL5). Average
values in red. No significant differences were found. B,
Proliferation ratesof PD-1þ melanoma cells compared with the PD-1�
counterparts after 96 hours (N ¼ 3) and 7 days (N ¼ 2) of BRAF/MEKi
treatment [dabrafenib þ trametinib(1 mmol/Lþ 5 nmol/L) in BRAFV600
mutant cell lines and trametinib (5 nmol/L) in the NRASQ61 mutant
cell line]. No significant differences were found. C,
Schematicrepresentation of the lentiviral vector used to transduce
melanoma cells; the eGFP expression is controlled by the promoter
regulatory element of the Oct4 gene(LV-Oct4. eGFP). D, Percentage
of eGFPþ cells at baseline and after 96 hours of BRAF/MEKi
treatment (dabrafenib þ trametinib; 1 mmol/L þ 5 nmol/L),
inpatient-derived cell lines transduced with the LV-Oct4.eGFP
vector (N ¼ 3). Average values and standard deviation. No
significant differences were found.E, Percentage of eGFPþ/PD-1þ
cells compared with eGFPþ/PD-1� cells in patient-derived cell lines
transduced with the LV-Oct4.eGFP vector (N ¼ 3). Averagevalues and
standard deviation. No significant differences were found.
Sanlorenzo et al.
Clin Cancer Res; 24(14) July 15, 2018 Clinical Cancer
Research3382
on June 30, 2021. © 2018 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst April 12, 2018; DOI:
10.1158/1078-0432.CCR-17-1914
http://clincancerres.aacrjournals.org/
-
Anti–PD-1 antibody prolongs the antitumor response
toBRAF/MEKi
Considering the hypothesis that PD-1 activation could lead
tomelanoma proliferation (7), we tested whether PD-1-blockagecould
have a direct anti-tumor effect.
In vitro, the sole use of anti–PD-1 antibody did not affect
cellviability (Supplementary Fig. S5). When we combined
anti–PD-1antibody with BRAF/MEKi, we observed only a trend toward
abetter anti-tumor effect compared with BRAF/MEKi alone
duringshort-term drug exposure (Supplementary Fig. S6). To test
thehypothesis that the subset of PD-1þ melanoma cells, which
arepreferentially endowed with stemness features, might
contributeto the development of BRAF/MEKi resistance, we set up an
in-vivolong-term experiment. We used non-obese diabetic/severe
com-bined immunodeficient (NOD-SCID)/interleukin 2
receptor[IL2r]gnull (NSG) mice bearing palpable subcutaneous
xenograftmelanoma. The treatmentwith anti–PD-1 antibody alone did
nothave any beneficial effect on tumor growth (N ¼ 6; Fig. 4A).
Onthe other side, when combined with BRAF/MEKi, anti–PD-1antibody
(N¼ 6) significantly prolonged the antitumor responseand
delayedmelanoma relapse compared to controls treated onlywith
BRAF/MEKi (N ¼ 6; P ¼ 0.0006).
DiscussionBRAF/MEK inhibitors (BRAF/MEKi) and anti–PD-1
antibody
combinations might be a therapeutic strategy for
metastaticmelanoma patients, and phase II and III clinical
trials(NCT02910700, NCT02224781, NCT02130466,
NCT02967692,NCT02858921) are currently recruiting. These trials are
basedon preclinical models which explain the synergism by
positiveBRAF/MEKi effects on T-cell recruitment, PD-L1 upregulation
ontumor cells and consequent enhancement of anti–PD-1 antibody
antitumor effect (1, 2). Our results point to a novel,
lymphocyte-independent, mechanism of action: BRAF/MEKi
treatmentleads to higher rates of viable melanoma cells expressing
PD-1and PD-L2, and therefore it could sensitize the tumor to a
directinhibitory effect of anti–PD-1 antibody.
"Ectopic" melanoma-intrinsic PD-1 expression and its
possiblerole in promoting tumor growth were proposed (7), but
thebiological relevance of thesefindings is still not clear. Such
subsetswere observed to have tumor-initiating properties; thus,
theycould contribute to the development of drug resistance.
We first chose an in vitro platform to characterize
melanoma-intrinsic PD-1 expression in normal conditions and
duringBRAF/MEKi treatment. We confirmed that melanoma cells
doexpress intrinsic PD-1, but at very low rates. Such low
percent-age of PD-1þ cells is unlikely to account for large
functionaleffects and indeed, treatment with anti–PD-1 antibody
aloneaffected neither tumor cell viability in-vitro, nor tumor
growthin immunodeficient xenograft models.
However, upon treatment with BRAF/MEKi, percentages ofPD-1þ
cells increased to relevant numbers in all tested mela-noma cell
lines, likely capable to enhance tumor proliferation ifactivated.
Furthermore, we found that BRAF/MEKi also upre-gulated the
PD-1-ligand PD-L2 on melanoma, therefore ajuxtacrine,
pro-proliferative PD-1-activation on melanoma isfeasible and
biologically plausible. This interaction wouldsupport melanoma
proliferation and thus counteract desiredanti-tumor effects of
BRAF/MEKi.
The increased percentage of PD-1þ melanoma cells duringtreatment
with BRAF/MEKi can be the result of a molecularmodulation of PD-1
protein expression, but also of a selectiveprocess of PD-1þ
melanoma cells less sensitive to target therapy.The linear kinetics
of PD-1 expression during BRAF/MEKi treat-ment, and its rapid
reversion upon drug withdrawal endorse the
Figure 4.
Anti–PD-1 antibody prolongs the antitumor response of BRAF/MEKi
in immunodeficient mice. A, Kinetics (mean � SEM) of A375 xenograft
growth in NSG micetreatedwith anti–PD-1 antibody (anti–PD-1mAb;
200mg three times aweek;N¼ 6) or isotype control antibody (200mg
three times aweek;N¼6). Treatment start ismarked by the
arrow.B,Kinetics (mean�SEM) ofA375 xenograft growth inNSGmice
treatedwithBRAF/MEKi (dabrafenibþ trametinib; respectively, 600
and4mg,five consecutive days a week) and anti–PD-1 antibody
(anti–PD-1 mAb; 200 mg three times a week; N ¼ 6) or BRAF/MEKi
(dabrafenib and trametinib) alone(N ¼ 6). Treatment start is marked
by the arrow. Statistical analysis was carried out using two-way
ANOVA followed by Sidak correction and with two-tails;� , P <
0.05; ���, P < 0.001.
BRAF and MEK Inhibitors Increase PD-1–Positive Melanoma
www.aacrjournals.org Clin Cancer Res; 24(14) July 15, 2018
3383
on June 30, 2021. © 2018 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst April 12, 2018; DOI:
10.1158/1078-0432.CCR-17-1914
http://clincancerres.aacrjournals.org/
-
hypothesis of a dynamic modulation. On the other hand,
ourfinding that the PD-1þ cell subsets are enriched with cells
withstemness features supports the idea of a higher resistance of
thosecells to target therapy treatment. The description of the
exactmechanism leading to the increase of PD-1þ melanoma
cellsduring treatment with BRAF/MEKi is beyond the scope of
thiswork. Instead, we focused our efforts to investigate the
biologicalrelevance of our findings.
In vitro, the addition of anti–PD-1 antibody to BRAF/MEKi,only
slightly improved their anti-tumor effects, and only whenused at
higher dose (100 mg/mL) than previously described inmelanoma (50
mg/mL; ref. 7). Considering the negligible in vitroeffect, we did
not further investigate possible off-target effects ofanti–PD-1
antibody. But the absence of any toxic or biologicaleffect in
melanoma cells treated exclusively with anti–PD-1 anti-body—where
PD-1 expression levels are very low (e.g.,
-
7. Kleffel S, Posch C, Barthel SR, Mueller H, Schlapbach C,
Guenova E, et al.Melanoma cell-Intrinsic PD-1 receptor functions
promote tumor growth.Cell 2015;162:1242–56.
8. Abdullah LN, Chow EK-H. Mechanisms of chemoresistance in
cancer stemcells. Clin Transl Med 2013;2:3.
9. Lee N, Barthel SR, Schatton T. Melanoma stem cells and
metastasis:mimicking hematopoietic cell trafficking? Lab Investig J
Tech MethodsPathol 2014;94:13–30.
10. Gammaitoni L, Giraudo L,MacagnoM, Leuci V,MesianoG, Rotolo
R, et al.Cytokine Induced Killer cells kill chemo-surviving
melanoma cancer stemcells. Clin Cancer Res 2017;23:2277–88.
11. Rizos H,Menzies AM, Pupo GM, CarlinoMS, Fung C, Hyman J, et
al. BRAFinhibitor resistance mechanisms in metastatic melanoma;
spectrum andclinical impact. Clin Cancer Res
2014;clincanres.3122.2013.
12. Sullivan RJ, Flaherty KT. Resistance to BRAF-targeted
therapy inmelanoma.Eur J Cancer Oxf Engl 1990 2013;49:1297–304.
13. Long GV, Eroglu Z, Infante J, Patel S, Daud A, Johnson DB,
et al. Long-termoutcomes in patients with BRAF V600–mutant
metastatic melanoma whoreceived dabrafenib combined with
trametinib. J Clin Oncol 2017;JCO.2017.74.1025.
14. Hugo W, Shi H, Sun L, Piva M, Song C, Kong X, et al.
Non-genomic andimmune evolution of melanoma acquiring MAPKi
resistance. Cell2015;162:1271–85.
15. Gammaitoni L, Giraudo L, Leuci V, Todorovic M, Mesiano G,
Picciotto F,et al. Effective activity of cytokine-induced killer
cells against autologousmetastatic melanoma including cells with
stemness features. Clin CancerRes 2013;19:4347–58.
16. Gerrard L, Rodgers L, Cui W. Differentiation of human
embryonicstem cells to neural lineages in adherent culture by
blockingbone morphogenetic protein signaling. STEM CELLS
2005;23:1234–41.
17. Follenzi A, Ailles LE, Bakovic S, Geuna M, Naldini L. Gene
transfer bylentiviral vectors is limited by nuclear translocation
and rescued by HIV-1pol sequences. Nat Genet 2000;25:217–22.
18. Boussemart L, Malka-Mahieu H, Girault I, Allard D,
Hemmingsson O,Tomasic G, et al. eIF4F is a nexus of resistance to
anti-BRAF and anti-MEKcancer therapies. Nature 2014;513:105–9.
19. Vujic I, Sanlorenzo M, Posch C, Esteve-Puig R. Metformin and
trametinibhave synergistic effects on cell viability and tumor
growth in NRASmutantcancer. Oncotarget 2015;6:969–78.
20. Long GV, Weber JS, Infante JR, Kim KB, Daud A, Gonzalez R,
et al. Overallsurvival and durable responses in patients with BRAF
V600—mutantmetastatic melanoma receiving dabrafenib combined with
trametinib.J Clin Oncol 2016;34:871–8.
21. Ascierto PA, Schadendorf D, Berking C, Agarwala SS, van
Herpen CM,Queirolo P, et al. MEK162 for patients with advanced
melanoma harbour-ingNRAS or Val600 BRAFmutations: a non-randomised,
open-label phase2 study. Lancet Oncol 2013;14:249–56.
22. DummerR, SchadendorfD, Ascierto
PA,AranceA,DutriauxC,DiGiacomoAM, et al. Binimetinib versus
dacarbazine in patients with advancedNRAS-mutantmelanoma (NEMO):
amulticentre, open-label, randomised, phase3 trial. Lancet Oncol
2017;18:435–445.
23. Qu�ereux G, Dr�eno B. Fotemustine for the treatment of
melanoma. ExpertOpin Pharmacother 2011;12:2891–904.
www.aacrjournals.org Clin Cancer Res; 24(14) July 15, 2018
3385
BRAF and MEK Inhibitors Increase PD-1–Positive Melanoma
on June 30, 2021. © 2018 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst April 12, 2018; DOI:
10.1158/1078-0432.CCR-17-1914
http://clincancerres.aacrjournals.org/
-
2018;24:3377-3385. Published OnlineFirst April 12, 2018.Clin
Cancer Res Martina Sanlorenzo, Igor Vujic, Arianna Floris, et
al.
PD-1 Antibody−Anti Leading to a Potential Lymphocyte-Independent
Synergism with
BRAF and MEK Inhibitors Increase PD-1-Positive Melanoma
Cells
Updated version
10.1158/1078-0432.CCR-17-1914doi:
Access the most recent version of this article at:
Material
Supplementary
http://clincancerres.aacrjournals.org/content/suppl/2018/04/10/1078-0432.CCR-17-1914.DC1
Access the most recent supplemental material at:
Cited articles
http://clincancerres.aacrjournals.org/content/24/14/3377.full#ref-list-1
This article cites 21 articles, 7 of which you can access for
free at:
E-mail alerts related to this article or journal.Sign up to
receive free email-alerts
Subscriptions
Reprints and
[email protected]
To order reprints of this article or to subscribe to the
journal, contact the AACR Publications Department at
Permissions
Rightslink site. Click on "Request Permissions" which will take
you to the Copyright Clearance Center's (CCC)
.http://clincancerres.aacrjournals.org/content/24/14/3377To
request permission to re-use all or part of this article, use this
link
on June 30, 2021. © 2018 American Association for Cancer
Research. clincancerres.aacrjournals.org Downloaded from
Published OnlineFirst April 12, 2018; DOI:
10.1158/1078-0432.CCR-17-1914
http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-17-1914http://clincancerres.aacrjournals.org/content/suppl/2018/04/10/1078-0432.CCR-17-1914.DC1http://clincancerres.aacrjournals.org/content/24/14/3377.full#ref-list-1http://clincancerres.aacrjournals.org/cgi/alertsmailto:[email protected]://clincancerres.aacrjournals.org/content/24/14/3377http://clincancerres.aacrjournals.org/
/ColorImageDict > /JPEG2000ColorACSImageDict >
/JPEG2000ColorImageDict > /AntiAliasGrayImages false
/CropGrayImages false /GrayImageMinResolution 200
/GrayImageMinResolutionPolicy /Warning /DownsampleGrayImages true
/GrayImageDownsampleType /Bicubic /GrayImageResolution 300
/GrayImageDepth -1 /GrayImageMinDownsampleDepth 2
/GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true
/GrayImageFilter /DCTEncode /AutoFilterGrayImages true
/GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict >
/GrayImageDict > /JPEG2000GrayACSImageDict >
/JPEG2000GrayImageDict > /AntiAliasMonoImages false
/CropMonoImages false /MonoImageMinResolution 600
/MonoImageMinResolutionPolicy /Warning /DownsampleMonoImages true
/MonoImageDownsampleType /Bicubic /MonoImageResolution 900
/MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000
/EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode
/MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None
] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false
/PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000
0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true
/PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ]
/PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier ()
/PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped
/False
/CreateJDFFile false /Description > /Namespace [ (Adobe)
(Common) (1.0) ] /OtherNamespaces [ > /FormElements false
/GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks
false /IncludeInteractive false /IncludeLayers false
/IncludeProfiles false /MarksOffset 18 /MarksWeight 0.250000
/MultimediaHandling /UseObjectSettings /Namespace [ (Adobe)
(CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /NA
/PageMarksFile /RomanDefault /PreserveEditing true
/UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling
/LeaveUntagged /UseDocumentBleed false >> > ]>>
setdistillerparams> setpagedevice