Combined use of anti-ErbB monoclonal antibodies and erlotinib enhances antibody-dependent ...

14 Pages·2013·1.36 MB·English

Cavazzonietal.MolecularCancer2012,11:91
http://www.molecular-cancer.com/content/11/1/91
RESEARCH Open Access
Combined use of anti-ErbB monoclonal antibodies
and erlotinib enhances antibody-dependent
cellular cytotoxicity of wild-type erlotinib-sensitive
NSCLC cell lines
Andrea Cavazzoni1†, Roberta R Alfieri1*†, Daniele Cretella1, Francesca Saccani1, Luca Ampollini2, Maricla Galetti1,3,
Federico Quaini1, Gallia Graiani1, Denise Madeddu1, Paola Mozzoni1,3, Elena Galvani1, Silvia La Monica1,
Mara Bonelli1, Claudia Fumarola1, Antonio Mutti1, Paolo Carbognani2, Marcello Tiseo4, Elisabetta Barocelli5,
Pier Giorgio Petronini1 and Andrea Ardizzoni4
Abstract
Background: The epidermal growth factor receptor (EGFR) is an established target for anti-cancer treatmentin
different tumour types. Two different strategies have been exploredto inhibit this pivotal molecule inepithelial
cancer development: small molecules TKIs and monoclonal antibodies. ErbB/HER-targetingby monoclonal
antibodies such as cetuximab and trastuzumab or tyrosine-kinase inhibitors as gefitinib or erlotinib has been proven
effective inthe treatment of advanced NSCLC.
Results: Inthis study weexploredthe potential of combining either erlotinib with cetuximab or trastuzumab to
improve theefficacy of EGFR targetedtherapy in EGFR wild-type NSCLCcell lines. Erlotinib treatmentwas observed
to increase EGFR and/or HER2 expression at theplasma membrane levelonly inNSCLCcell lines sensitive to the
drug inducing protein stabilization. The combined treatment had marginaleffecton cell proliferation but markedly
increased antibody-dependent, NKmediated, cytotoxicity in vitro. Moreover, intheCalu-3 xenograft model,the
combination significantly inhibited tumour growth when compared with erlotinib and cetuximab alone.
Conclusion: Our results indicate that erlotinib increases surface expression ofEGFR and/or HER2 only inEGFR-TKI
sensitive NSCLC cell lines and,in turns, leads to increased susceptibility to ADCC both in vitro and in a xenograft
models. The combinationof erlotinib with monoclonal antibodies represents a potential strategy to improve the
treatment ofwild-type EGFR NSCLC patients sensitive to erlotinib.
Keywords: Lung cancer, EGFR, Erlotinib, Cetuximab, ADCC
Background intracellular C-terminal domain with tyrosine kinase
The epidermal growth factor receptor (EGFR, ErbB1, activity [1]. Upon specific binding of EGF-like ligands to
HER1) is the prototypic member of the ErbB family of the extracellular domain, ErbB receptors dimerize, either
receptor tyrosine kinases (TKs), which further consists as homo- or heterodimers, and undergo autophosphory-
ofErbB2-4(HER2-4).TheErbBreceptorsshareasimilar lation at specific tyrosine residues within the intracellu-
protein structure, consisting of an extracellular ligand lar domain. The phosphorylated tyrosines serve as
binding domain, a single transmembrane domain and an docking sites for adapter molecules, such as Grb2 and
the p85 subunit of PI3K, which activate a complex
downstream network. The activated signaling pathways,
*Correspondence:[email protected]
†Equalcontributors including the Ras/MAPK, Akt/mTOR kinase and STAT
1DepartmentofClinicalandExperimentalMedicine,UniversityofParma, cascades, in turn regulate transcription factors and other
Parma,Italy proteins involved in cell proliferation, survival, motility
Fulllistofauthorinformationisavailableattheendofthearticle
©2012Cavazzonietal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreative
CommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,and
reproductioninanymedium,providedtheoriginalworkisproperlycited.
Cavazzonietal.MolecularCancer2012,11:91 Page2of14
http://www.molecular-cancer.com/content/11/1/91
and differentiation [2]. Two main strategies targeting Several preclinical studies on cell lines from different
ErbB receptors have been developed: small-molecule tumour types, indicated that the association between
inhibitors of the tyrosine kinase domain (EGFR tyrosine EGFR/HER2 mAbs withTKIs displays an increased effi-
kinase inhibitors [TKIs], such as erlotinib and gefitinib), cacy [15].
and monoclonal antibodies (such as cetuximab, anti- In this study we explored the potential of combining
EGFR and trastuzumab, anti-HER2), directed against the erlotinib with either cetuximab or trastuzumab in order
extracellular domain, which inhibit phosphorylation/ to improve the efficacy of EGFR targeted therapy in
activation and promote internalization. EGFR and HER2 EGFR wild-type sensitive NSCLC cell lines. Our results
are overexpressed in 40-80% and 25-30%, respectively, of indicate that EGFR-TKI increases surface expression of
non-small cell lung cancer (NSCLC) patients and their EGFR and/or HER2 only in erlotinib sensitive NSCLC
overexpression has been frequently correlated with a cell lines and, in turns, leads to increased susceptibility
poorprognosis[3,4]. toADCC both invitro andinxenograft models.
Erlotinib is an effective treatment for NSCLC patients
and has been registered as a second and third-line treat- Results
mentofNSCLCregardlessofEGFR mutationstatus [5]. DifferentialeffectsoferlotinibonEGFRandHER2
Gefitinib has been registered for the therapy of expressioninsensitiveandresistantNSCLCcelllines
advanced NSCLC harbouring activating EGFR mutations Firstly,weevaluatedtheeffectoferlotinibontotalEGFR
in the tyrosine kinase domain, the most frequent being and HER2 protein levels in sensitive NSCLC cell lines
L858R in exon 21 and Del (746–750) in exon 19 [6]. (Calu-3, H322 and H292 cell lines carrying wild-type
Although mutations in EGFR are useful predictors for EGFR; PC9 and HCC827 carrying EGFR E746-A750del
the activity of EGFR-TKI, they cannot be used as the mutation) and in resistant cell lines (A549, H1299,
only criterion to determine who should receive anti- H1703 and Calu-1 intrinsically resistant carrying wild-
EGFR therapy and it is becoming increasingly clear that type EGFR; HCC827GR5 with MET amplification as
even patients with EGFR wild-type can benefit from mechanism of acquired resistance to TKI) [16]. As
EGFR-TKI [5,7,8]. shown in Figure 1A, erlotinib induced accumulation of
Cetuximab is a chimeric IgG1 monoclonal antibody EGFR protein in Calu-3 and H322 cells while HER2
(mAb) that blocks ligand binding to EGFR, leading to a accumulated in H322, H292, PC9 and HCC827 cells in a
decrease in receptor dimerization, autophosphorylation, dose-dependent manner. The EGFR/Actin and HER2/
and activation of signaling pathways [9]. In addition the Actin ratios obtained after treatment at 1 μM or 10 nM
binding of cetuximab initiates EGFR internalization erlotinib were calculated and values expressed as fold
and degradation which leads to signal termination. differences versus control (Figure 1B). In contrast, EGFR
Moreover, unlike EGFR-TKIs, cetuximab can induce and HER2 protein accumulation was not observed in
antibody dependent cellular cytotoxicity (ADCC) any cancer cell line with intrinsic resistance to EGFR
activity, an important immunologic antitumour effect. inhibitors until the concentration of 10 μM. Indeed the
Cetuximab in combination with chemotherapy has ratios EGFR/Actin or HER2/Actin were similar or even
been approved by the FDA for the treatment of meta- lowerthanthosecalculatedinuntreatedcells(Figure1C)
static colorectal cancer and of locally advanced head and similar results were obtained with gefitinib (not
and neck cancer. shown). A representative Western blotting of resistant
Two randomized phase III trials in NSCLC patients, H1299celllineisreportedinFigure1D.
evaluating cetuximab in addition to first-line chemo- The different effect of TKIs on HER2 expression be-
therapy, showed a small benefit in overall survival for tween sensitive and resistant NSCLC cell lines was con-
the experimental treatment, which was considered in- firmed in the HCC827 parental and in the HCC827GR5
sufficient by the EMA for marketing approval [10,11]. resistantclone treated for 48hwith gefitinib(Figure 1E).
However, a subgroup analysis of the FLEX phase III
trial recently demonstrated a larger survival benefit ErlotinibincreasesthecellsurfaceexpressionofEGFR
from the experimental treatment in patients with high andHER2inerlotinibsensitiveNSCLCcelllines
immunohistochemical EGFR expression [12]. EGFR and HER2 expression on the plasma membrane
Trastuzumab, registered for the treatment of HER2 was quantified by flow cytometry in sensitive EGFR
positive breast cancer, has also been tested in phase II wild-type NSCLC cell lines Calu-3, H322 and H292 after
trialsasa singleagentandincombinationwithcytotoxic exposure to 1 μM erlotinib for 24 h. The drug enhanced
chemotherapy for patients with NSCLC. These trials surface expression, calculated as molecules of equivalent
have not yet produced any convincing evidence of an soluble fluorophore, of EGFR in Calu-3 (Figure 2A) and
improved antitumour activity by adding trastuzumab to H322 (Figure 2C, 2D) and of HER2 in H292 (Figure 2B)
standard chemotherapy inNSCLC[13,14]. and H322 (Figure 2C, 2D) cell lines. In H322 cell line,
Cavazzonietal.MolecularCancer2012,11:91 Page3of14
http://www.molecular-cancer.com/content/11/1/91
Figure1(Seelegendonnextpage.)
Cavazzonietal.MolecularCancer2012,11:91 Page4of14
http://www.molecular-cancer.com/content/11/1/91
(Seefigureonpreviouspage.)
Figure1ErlotinibinducesEGFRandHER2proteinaccumulationonlyinsensitiveNSCLCcelllines.(A)Calu-3,H322,H292,PC9and
HCC827celllinesweretreatedwiththeindicatedconcentrationsoferlotinibfor48h.Attheendofthedrugtreatmentcelllysateswere
immunoblottedtodetecttheindicatedproteins.Theimmunoreactivespotswerequantifiedbydensitometricanalysis,ratiosofEGFR/Actinand
HER2/Actinwerecalculatedat1μMerlotinibforCalu-3H322andH292or10nMforPC9andHCC827andvaluesareexpressedasfoldincrease
versuscontrol(B).(C)HCC827GR5,A549,H1299,H1703,Calu-1celllinesweretreatedwith1μMerlotinibfor48handattheendoftreatment
celllysateswereimmunoblottedtodetecttheindicatedproteins.Theimmunoreactivespotswerequantifiedbydensitometricanalysis,ratiosof
EGFR/ActinandHER2/Actinwerecalculatedandvaluesareexpressedasfoldincreaseversuscontrol.(D)RepresentativeWesternblottingof
resistantH1299celllineexposedtoincreasedconcentrationoferlotinib.(E)HCC827parentalcelllineandHCC827GR5resistantclonewere
treatedwiththeindicateddosesofgefitinibandprocessedasabove.Theresultsarefromrepresentativeexperiments.Eachexperiment,repeated
threetimes,yieldedsimilarresults.
the increase in EGFR and HER2 surface expression was expression of EGFR or HER2 on sensitive NSCLC cells,
dose and time dependent (Figure 2C, 2D). Western blot leading to an increase of mAbs binding to cancer cell
analysis of isolated cell surface membrane proteins (inset surface.
Figure 2A) confirmed the increase of EGFR in erlotinib
treated Calu-3 cells. ErlotinibinducesEGFRproteinstabilization
Exploiting the ability of cetuximab and trastuzumab to The possibility that the higher EGFR level observed in
bind EGFR and HER2, we used these mAbs as primary Calu-3 cells exposed to erlotinib was due to protein
antibodies for flow cytometry analysis. By this approach, stabilization or increased synthesis was then explored.
as shown in Figure 3, we confirmed that the surface As shown in Figure 4A, EGFR level increased after 2 h
density of cetuximab and trastuzumab-binding sites, re- of erlotinib treatment and reached a plateau after 24 h.
spectively, on Calu-3 (Figure 3A), H322 (3B) and H292 Furthermore, the maximum level was maintained during
(3C) cells were increased after 1 μM erlotinib treatment. time in the presence of the drug. However, after 48 h of
Theseresultssuggestthaterlotinib enhancedcellsurface erlotinib removal, EGFR expression was reduced to level
Figure2EGFRandHER2increaseattheplasma-membranelevel.Calu-3(A)andH292(B)celllinesweretreatedwith1μMerlotinibfor
24h,H322celllinewastreatedwithincreasingconcentrationoferlotinib(C)orwith1μMerlotinibfortheindicatedperiodoftime(D).Atthe
endofthetreatment,cellsurfaceexpressionofEGFRand/orHER2wereevaluatedbyflowcytometryandthequantificationisreportedas
MoleculesofEquivalentFluorophore[MEF]orasfoldincreaseversusuntreatedcontrolcells(D).InsetFigure2A:WesternblotanalysisofEGFR
proteinmembranelevelinCalu-3aftertreatmentwith1μMerlotinibfor24h.Wholecellswerelabeledwithbiotinandmembranebound
proteinswerepulleddownwithneutrAvidinbeads.Theresultsarefromrepresentativeexperiments.Eachexperiment,repeatedthreetimes,
yieldedsimilarresults.
Cavazzonietal.MolecularCancer2012,11:91 Page5of14
http://www.molecular-cancer.com/content/11/1/91
Figure3Erlotinibinducestheincreaseofcetuximabandtrastuzumabbindingsites.Calu-3,H322andH292celllinesweretreatedwith
erlotinibfor24h.Bindingsiteswereassessedbyflowcytometryusingcetuximab(Calu-3,H322)andtrastuzumab(H292)asprimaryantibodies
followedbyPE-anti-humanIgGexposure.BindingsitesquantificationisreportedasMoleculesofEquivalentFluorophore[MEF].TheresultsinA,
B,Carefromrepresentativeexperiments.Eachexperiment,repeatedthreetimes,yieldedsimilarresults.
comparable to untreated cells (Figure 4B). Calu-3 were mechanisms such as protein stabilization. Moreover, we
also treated with erlotinib in the presence of specific analyzed EGFR transcript level by real time PCR after
inhibitors of mRNA (Actynomicin D) and protein erlotinib treatment (Figure 4D). Erlotinib did not affect
(Cycloheximide) synthesis. As shown in Figure 4C, the EGFR mRNA levelwhen compared tountreatedcells.
erlotinib- induced EGFR protein increase was neither With the aim to clarify why the increased level of
influenced by Actynomicin D nor Cycloheximide treat- EGFR was induced only in sensitive cells, we then tested
ment indicating that the higher level of EGFR after erlo- the effect of EGFR inhibitors (gefitinib, erlotinib, cetuxi-
tinib treatment could be ascribed to post-transcriptional mab, lapatinib) and of inhibitors of MAPK and PI3K/
Figure4ErlotinibinducesEGFRproteinaccumulationthroughproteinstabilization.(A)Calu-3cellsweretreatedfortheindicatedperiod
oftimewith0.5μMerlotinib.AttheendofdrugtreatmentscelllysateswereimmunoblottedtodetectEGFRprotein.(B)Calu-3cellswere
treatedwith0.5μMerlotinibfor24hthenthedrugwasremovedanddrug-freemediumwasaddedforfurther24and48h.Then,celllysates
wereimmunoblottedtodetecttheEGFRproteinlevels.(C)Calu-3cellsweretreatedfor24hwitherlotinib0.5μM,intheabsence/presenceof
0.1μg/mlactynomicinDand2μg/mlcyclohexymide.Attheendoftheexperiment,celllysateswereimmunoblottedtodetecttheindicated
proteins.Theimmunoreactivespotswerequantifiedbydensitometricanalysis,ratiosofEGFR/Actinwerecalculatedandvaluesareexpressedas
foldincreaseversuscontrol.(D)Calu-3cellswereexposedto0.5μMerlotinibfortheindicatedperiodoftimeandtheEGFRmRNAwasdetected
byRT-PCR.Themeanvaluesoftwoindependentmeasurements(±SD)areshown.(E)EGFR,p-P70S6K,p-P44/42andP44/42weredetectedby
WesternblottinginCalu-3cellsuntreatedortreatedfor24hwith1μMgefitinib,erlotinibandlapatinib,10μg/mlcetuximab,10μMU0126,1
μMNVP-BKM-120andNVP-BYL-719and100nMRAD001.Theresultsarefromrepresentativeexperiments.Eachexperiment,repeatedthreetimes,
yieldedsimilarresults.
Cavazzonietal.MolecularCancer2012,11:91 Page6of14
http://www.molecular-cancer.com/content/11/1/91
AKT/mTOR signaling transduction pathways on EGFR (Figure 5C) cells were resistant to both the single regi-
accumulation in Calu-3 cell line. Gefitinib, erlotinib, mens. Comparing the experimental combination points
lapatinib significantly inhibited the phosphorylation of with that expected by the Bliss criterion, an additive
p70S6Kand p44/42 and induced a significant increase in effect was observed only in the Calu-3 cells. In fact, in
EGFR protein level (Figure 4E). The MEK inhibitor the H322 cells we failed to observe any improvement
U0126 strongly enhanced EGFR expression, in contrast treating cells with the combined treatment and H1299
noincreaseintheEGFRlevelwasobservedafterincuba- remainedresistant.
tion with the inhibitors of PI3K/AKT/mTOR pathway Moreover, cell death, evaluated by morphological ana-
tested (NVP-BKM-120 and NVP-BYL-719 PI3K inhibi- lysis, caspase-3 activation and cleavage, was negligible
tors andRAD001mTORinhibitor). under any of the tested treatments at all the time points
analyzed (not shown) suggesting that the combined
Effectsoferlotinibandcetuximabcombinedtreatment erlotinib-cetuximab treatment exerted a cytostatic and
onNSCLCcellgrowthandantibody-dependent not acytotoxic effect.
cell-mediatedcytotoxicity Since the engagement of immune component system
We then investigated the effect of targeting EGFR by is one of the main mechanisms of the activity of specific
both the TKI erlotinib and the mAb cetuximab in a cell mAbs directed to ErbB family members in vivo, we
viability assay (Figure 5). We treated Calu-3, H322 and examined whether erlotinib could enhance cetuximab-
H1299 cells with erlotinib, cetuximab (doses ranged or trastuzumab-mediated ADCC by NK cells. As shown
from 1 to 50 μg/ml) or the combination based on the respectively in Figure 6 A-B cetuximab-dependent cyto-
schedule erlotinib 24 h followed by the combination of toxicity in the presence of IL-2 activated NK cells was
erlotinib with cetuximab for 72 h. As expected Calu-3 higher in Calu-3 and H322 cells previously treated with
(Figure 5A) and H322 (Figure 5B) cells were responsive erlotinib compared with cells treated with cetuximab
to erlotinib and cetuximab treatment, whereas H1299 alone. Similarly, trastuzumab-dependent cytotoxicity was
Figure5EffectoferlotinibandcetuximabcombinationoncellviabilityinNSCLCcelllines.Calu-3,H322sensitivecells(A,B)andH1299
resistantcells(C)wereexposedtotheindicatedconcentrationsoferlotinibfor96horcetuximabfor72handtoerlotinibfor24hfollowedby
thecombinationoferlotinibandcetuximabfor72h.Afterthetreatments,cellviabilitywasassessedbyMTTassay.Dataareexpressedaspercent
inhibitionofcellviabilityversuscontrolcellsandaremeanvaluesofthreeseparateexperiments(**P<0.01***P<0.001).
Cavazzonietal.MolecularCancer2012,11:91 Page7of14
http://www.molecular-cancer.com/content/11/1/91
Figure6Erlotinibpotentiatesantibodydependentcellcytotoxicity.TheindicatedhumanNSCLCcelllinesweretreatedwith1μMerlotinib
for24h.Afterthetreatmentwitherlotinib,10μg/mlCetuximab(A,B,E)orTrastuzumab(C,D)wereaddedtocancercellsseededwith
100U/mlIL-2activated-NKcellsattheratioof1:25and1:50.After4hLDHreleasewasdeterminedasdescribedinMethodssection.Theresults
arefromrepresentativeexperiments.Theexperiment,repeatedthreetimes,yieldedsimilarresults(**P<0.01***P<0.001).
higher inH322andH292cells (Figure 6C-D) previously EffectoferlotinibandcetuximabonCalu-3xenografts
treated with erlotinib compared with cells treated with To extend our results in vivo, we tested the combination
trastuzumabalone. of erlotinib with cetuximab in a Calu-3 xenograft model
Onthecontrary,thecombinationoferlotinibwithcetux- (Figure 7). Whentumourswerewell established (14 days
imabdidnotsignificantlymodifythemAbdependentcyto- post-injection, average volume of 300 mm3) mice
toxicityinH1299resistantcancercells. were randomized into four treatment groups receiving
Figure7AntitumouractivityoferlotinibandcetuximabonCalu-3xenografts.Calu-3cellsweresuspendedinmatrigelandsterilePBS(1:1)
andimplanteds.c.(rightflank)onfemaleBALB/c-Nudemice.Tumourswereallowedtoestablishgrowthafterimplantationfor14days,andthe
treatmentsstartedwhentumoursreachedanaveragevolumeof300mm3.Vehicle,erlotinib(25mg/Kg,orally5days/week),cetuximab(2mg/Kg
i.p.twiceweekly),orerlotinibpluscetuximabwereadministeredforthedurationofthestudy.Dataareexpressedaspercentchangeintumour
volume±SEMof6micepergroup.(**p<0.01,***p<0.001vscontrol;#p<0.05,###p<0.001vserlotinib;§p<0.05vscetuximab;two-wayANOVA
followedbyBonferronipost-test).
Cavazzonietal.MolecularCancer2012,11:91 Page8of14
http://www.molecular-cancer.com/content/11/1/91
erlotinib alone, cetuximab alone, the combination, or treated mice. Interestingly, cetuximab clearly resulted in
vehicles as described in the Methods section. Drug dense inflammatory periglandular infiltrates mostly com-
treatments were well tolerated, and no signs of tox- posed oflymphocytes (Figure8C-3). Thus,thereal impact
icity were detected during the study. The treatment of treatment on tumour mass within the nodules was
with either erlotinib or cetuximab as single agent delayed assessed by the morphometric analysis of tissue compos-
tumourgrowth.However,thesignificanceofthetreatment ition. By this quantitative approach, in agreement with
versus the control was observed only with cetuximab as gross anatomic measurements, we documented that the
single agent or in combination. Interestingly, the treat- combination of erlotinib with cetuximab was the most ef-
ment with the combination of erlotinib plus cetuximab fectivetreatmentontumourgrowthinhibition(Figure8D).
significantly inhibited tumour growth when compared to This contention was further supported by the
boththesingleagenttreatments. immunofluorescence analysis of Ki67 labelling on tumour
The histologic analysis of tumour samples showed that tissues at the end of the experimental protocol (Figure9).
thesubcutaneousinjectionofCalu-3strikinglyreproduced Erlotinib was able to reduce proliferation of neoplastic
within four weeks the morphological features of human cytokeratinpos cells only in association with cetuximab
adenocarcinoma(Figures8A,8B1-4,8C-1).Neoplasticepi- whereas cetuximab had a negativeimpact on cycling cells
thelialcellsclearlyexpressedcytokeratin(Figure8C-2)and alsoas individualagent.TheTUNEL assayindicatedthat,
wereorganizedinsecretoryglandssurroundedbycellular- according with in vitro data, apoptosis was not a signifi-
ized collagen as evidenced by Masson’s trichrome staining cant ongoing cellularevent implicated inthe effect ofdif-
(Figure 8C-4). Regressive phenomena and changes in size ferenttreatments.
ofneoplasticglandstogetherwithintensestromalreaction We have calculated that 0.026+/−0.016% neoplastic
were observed in histologic samples of tumours from cells were undergoing apoptosis in untreated tumours.
Figure8Hystologicalanalysisoftumours.A:SelectedexamplesofH&Estainedsectionsoftheentiresubcutaneousxenograftedtumour
inducedbyCalu-3injectioninuntreated(C)anderlotinib(Erl),cetuximab(Cet)orerlotinib+cetuximab(Erl+Cet)treatedBALB/cnudemice
(scalebars:1mm).HighermagnificationofthesamesamplesareshownoncorrespondingpanelsinB(scalebars:500μm).C:representative
morphologicaldetailsofthecontrolneoplasticepithelium(1,H&E)expressingcytokeratin(2,*brown-immunoperoxidase)thatalsodepictsthe
epidermis(arrowhead).Thepresenceofinflammatoryinterstitialcellsinacetuximabtreatedtumour(3,H&E)andtheintensecollagendeposition
(bluish)surroundingneoplasticglands(purple)inaErl+Cettreatedtumour(4,Masson’strichrome)areshown(scalebars:100μm).D:Bargraphs
illustratingthequantitativemeasurementsofneoplastic,inflammatorycellsandstromalcompartmentscomposingthetumours.(*p<0.05,
**p<0.01,vscontrol;#p<0.05,##p<0.01vserlotinib;§p<0.05vscetuximab).
Cavazzonietal.MolecularCancer2012,11:91 Page9of14
http://www.molecular-cancer.com/content/11/1/91
Figure9Immunohistochemicalanalysisofcellularproliferation.ImmunofluorescenceimagesofKi67(green)nuclearlabellingofcytokeratin
(CK,red)positiveneoplasticcellsinsectionsofxenograftedtumoursfromanuntreated(A)andErl+Cettreated(B)BALB/cnudemouse.C:bar
graphillustratingtheeffectofthedifferenttreatmentsonthepercentageofcycling(Ki67pos)neoplasticcellswithinthetumour.CTRL:untreated,
ERL:erlotinib,CET:cetuximab,COMB:erlotinib+cetuximab.*p<0.01vsCTRL.
Similar low numbers were obtained after Erlotinib or preclinical studies in several cell lines originating from
Cetuximab single treatment whereas Erl+Cet increased different tumour types [15]. In EGFR wild-type H292
the amount of TUNEL positive neoplastic cells although and A549 NSCLC cell lines, the combination of either
reachingarateof0.12+/−0.03%.However,we cannotex- gefitinib or erlotinib with cetuximab was reported to en-
clude that apoptotic cell death may have contributed to hance growth inhibition in comparison to single treat-
the positive effect of tumor shrinkage at earlier times ment, particularly inthe H292 gefitinib sensitive cell line
afterdrug administration. [17]. In the A549 cell line, expressing both EGFR and
Thus, these experimental observations suggest that HER2, the combination of gefitinib with trastuzumab
targeting EGFR by the combination of small molecules significantly inhibited cell growth and proliferation [18].
and antibodies increases the in vitro and in vivo anti- In Calu-3 xenograft models, the combined treatment of
proliferative activity of both individual agents and erlotinib and pertuzumab showed an enhanced antitu-
seems to be a potent therapeutic strategy against mouractivity[19].
NSCLC. A correlation between cetuximab efficacy and EGFR
expression has been reported in preclinical studies [20]
Discussion and recently confirmed in clinical trials. Thus, the phase
The potential for dual-agent molecular targeting of the III FLEX study involving patients with advanced NSCLC
ErbB family, has been clearly demonstrated in pre- showed a strong correlation between high tumour EGFR
clinical models and confirmed on the clinical setting for overexpression and the efficacy of adding cetuximab to
HER2-targeting agents in breast cancer. However, little platinum basedfirst-line chemotherapy[12].
is known about this therapeutic strategy for different ThecombinationofaTKIandamAbwasexploredasa
targetsinothertumour types. potentialstrategytoovercomeacquiredresistancetofirst-
In our current study we demonstrated that the generation EGFR-TKIs. Kim and colleagues demonstrated
combination of erlotinib with cetuximab or trastuzumab that the combination of lapatinib with cetuximab over-
may enhance the antitumour activity of EGFR-TKI in came gefitinib resistance due to the secondary T790M
NSCLC cell lines harbouring wild-type EGFR and in mutation in NSCLC by inducing enhanced cytotoxicity
xenograftmodels. bothinvitroandinvivo[21].Furthermore,theassociation
The efficacy of the association between an EGFR/ of cetuximab with afatinib has been shown to be effective
HER2 mAbs with TKIs has been documented in toovercomeT790M-mediateddrugresistance[22].
Cavazzonietal.MolecularCancer2012,11:91 Page10of14
http://www.molecular-cancer.com/content/11/1/91
However, the combination of erlotinib with cetuxi- The reason why EGFR inhibitors such as gefitinib,
mab did not lead to a significant radiological response erlotinib or lapatinib induce EGFR accumulation only in
in NSCLC patients with clinically defined acquired sensitive cells could be ascribed to their ability to inhibit
resistance to erlotinib indicating that such strategy is signal transduction pathways downstream EGFR. The
not sufficient to overcome acquired resistance to erlo- constitutive activation ofsignalingpathwaysdownstream
tinib[23]. of EGFR (i.e. presence of RAS mutations) is indeed a
The mechanisms leading to an enhanced activity of recognized mechanism of resistance against reversible
combining a TKI with a monoclonal antibody have EGFR-TKIs [29]. The inhibition of the MAPK pathway
been ascribed, in other cancer cell models, either to a might represent a link between EGFR inhibition and
more efficient inhibition of TK receptors [17] or to EGFR accumulation since U0126, a well known MEK1/2
an increased targeted receptors on plasma membrane inhibitor, induced EGFR accumulation in Calu-3 cells,
induced by TKIs [24,25]. Scaltriti et al. showed that while none of PI3K/AKT/mTOR inhibitors tested was
lapatinib enhanced the effects of trastuzumab by in- effective. A correlation between MAPK pathway and
ducing HER-2 stabilization and accumulation at the protein degradation by the ubiquitin system was
cell surface of breast cancer cell lines [24], and described for the pro-apoptotic BH3-only protein BIM,
Mimura et al. reported that lapatinib induced accumu- indeed in the absence of MAPK activation, BIM protein
lation of HER-2 and EGFR on esophageal cancer cell accumulated in the cell promoting activation of apop-
lines evoking trastuzumab- and cetuximab- mediated totic celldeath[30].
ADCC[25]. Considering that EGFR TKIs, in particular erlotinib,
ADCC, one of the killing mechanism of the immune demonstratedtobeeffectiveonlyinasmallpercentage
system mediated by Natural Killer cells, plays a pivotal of NSCLC patients not harboring EGFR mutations,
role in the anti-cancer effects exerted by mAbs. There- our preclinical results could support clinical trials on
fore, increasing the ADCC activity is an important the combinations of erlotinib and cetuximab or trastu-
objective in the development of novel therapeutic zumab aiming to improve treatment efficacy. Although
approaches. the addition of cetuximab to erlotinib is insufficient to
It has been recently demonstrated that the EGFR inhi- overcome erlotinib resistance in EGFR-driven lung
bitors gefitinib and erlotinib enhance the susceptibility adenocarcinoma [23], the clinical potential of dual-
to NK cell mediated lysis of A549, NCI-H23 and SW- agent molecular targeting of the EGFR in patients with
900 lung cancer cell lines [26] by the induction of EGFR wild-type tumours remains to be elucidated and
ULBP1 (a ligand of the NK cell activation receptor may represents an interesting research area to be
NKG2D). These data indicate that EGFR blockade could pursued.
not be the only mechanism of action of EGFR inhibitors
in vivo. The efficacy of these inhibitors in lung cancer Conclusions
may be at least in part mediated by increased suscepti- In this study we explored the potential of combining
bility to NK activity. Moreover, cetuximab serves as a erlotinib with cetuximab or trastuzumab in improving
potent stimulus for NK functions including INF-gamma the efficacy of EGFR targeted therapy in EGFR wild-type
production [27] and is also associated with a comple- erlotinib-sensitive NSCLC cell lines. Our results indicate
ment–mediatedimmuneresponse[28]. that erlotinib, through ERK inhibition, increases surface
We here demonstrated that erlotinib induces an accu- expressionofEGFR and/orHER2only inerlotinibsensi-
mulation of EGFR and/or HER2 protein at the plasma tive NSCLC cell lines and in turn leads to increased sus-
membrane level only in TKI sensitive NSCLC cell lines ceptibility to ADCC both in vitro and in xenografts
whereas, in resistant cells (both, intrinsic or MET models.
amplification-mediated acquired resistance), this en- These data prompt future adequate clinical trials that
hancement was not observed. The anti-tumour effect of will give the ultimate proof of the utility of this com-
drug combination was more evident in ADCC experi- bined treatment for the care of NSCLC patients carrying
ments compared with cell viability experiments. In the EGFR-wildtype thataresensitivetoTKIs.
Calu-3 xenograft model, the combined treatment
resulted in a lower rate of tumour growth, suggesting Methods
the involvement of NK activity as a determinant factor Cellculture
to improve the efficacy of the combined treatment. ThehumanNSCLCcelllinesH322,H292,Calu-3,H1299,
Moreover, regressive phenomena and changes in size of A549, H1703 and Calu-1 were obtained from American
neoplastic glands together with intense stromal reaction Type Culture Collection (Manassas, VA, USA) and were
were observed in histologic samples of tumours from cultured as recommended. The PC9, HCC827 and
micetreatedwithcetuximabalone orthe combination. HCC827GR5 cell lines were kindly provided by Dr P.