bolero-3_2014

8/10/2019 BOLERO-3_2014

1/12580 www.thelancet.com/oncology Vol 15 May 2014

Articles

Everolimus for women with trastuzumab-resistant,

HER2-positive, advanced breast cancer (BOLERO-3):a randomised, double-blind, placebo-controlled phase 3 trial

Fabrice Andr, Ruth ORegan, Mustafa Ozguroglu, Masakazu Toi, Binghe Xu, Guy Jerusalem, Norikazu Masuda, Sharon Wilks, Francis Arena,

Claudine Isaacs, Yoon-Sim Yap, Zsuzsanna Papai, Istvan Lang, Anne Armstrong, Guillermo Lerzo, Michelle White, Kunwei Shen, Jennifer Li tton,

David Chen, Yufen Zhang, Shyanne Ali, Tetiana Taran, Luca Gianni

SummaryBackground Disease progression in patients with HER2-positive breast cancer receiving trastuzumab might beassociated with activation of the PI3K/Akt/mTOR intracellular signalling pathway. We aimed to assess whether theaddition of the mTOR inhibitor everolimus to trastuzumab might restore sensitivity to trastuzumab.

MethodsIn this randomised, double-blind, placebo-controlled, phase 3 trial, we recruited women with HER2-positive,trastuzumab-resistant, advanced breast carcinoma who had previously received taxane therapy. Eligible patients wererandomly assigned (1:1) using a central patient screening and randomisation system to daily everolimus (5 mg/day)plus weekly trastuzumab (2 mg/kg) and vinorelbine (25 mg/m) or to placebo plus trastuzumab plus vinorelbine, in3-week cycles, stratified by previous lapatinib use. The primary endpoint was progression-free survival (PFS) by localassessment in the intention-to-treat population. We report the final analysis for PFS; overall survival follow-up is stillin progress. This trial is registered with ClinicalTrials.gov, number NCT01007942.

FindingsBetween Oct 26, 2009, and May 23, 2012, 569 patients were randomly assigned to everolimus (n=284) or placebo(n=285). Median follow-up at the time of analysis was 202 months (IQR 150271). Median PFS was 700 months(95% CI 674818) with everolimus and 578 months (549690) with placebo (hazard ratio 078 [95% CI 065095];p=00067). The most common grade 34 adverse events were neutropenia (204 [73%] of 280 patients in the everolimusgroup vs 175 [62%] of 282 patients in the placebo group), leucopenia (106 [38%] vs 82 [29%]), anaemia (53 [19%] vs

17 [6%]), febrile neutropenia (44 [16%]vs

ten [4%]), stomatitis (37 [13%]vs

four [1%]), and fatigue (34 [12%]vs

11 [4%]).Serious adverse events were reported in 117 (42%) patients in the everolimus group and 55 (20%) in the placebo group;two on-treatment deaths due to adverse events occurred in each group.

InterpretationThe addition of everolimus to trastuzumab plus vinorelbine significantly prolongs PFS in patients withtrastuzumab-resistant and taxane-pretreated, HER2-positive, advanced breast cancer. The clinical benefit should beconsidered in the context of the adverse event profile in this population.

FundingNovartis Pharmaceuticals Corporation.

IntroductionHER2 is expressed or amplified in about 25% of breastcancers, and is associated with tumour aggressiveness,resulting in recurrent disease or shorter overall survival.1

However, these tumours are highly sensitive totrastuzumab, an anti-HER2 antibody that has beenshown to be effective in the management of advancedbreast cancer.2 Despite this effectiveness, metastaticdisease in most patients who have received trastuzumabwill eventually progress because of development of de-novoor secondary resistance.24Several treatment optionsfor patients with disease progression on trastuzumabhave been developed, including trastuzumab emtan-sine, lapatinib, or continued trastuzumab in combi-nation with chemotherapy or with pertuzumab.510Although these agents have different mechanisms ofaction, these approaches are all means of targetingHER2; whether or not they address the underlyingmechanism of trastuzumab resistance is unknown.

Trastuzumab resistance is thought to be mediatedoften via the PI3K/Akt/mTOR pathway. In preclinicalmodels, acquisition of HER2 resistance has resulted inloss of the tumour suppressor gene PTEN, which

antagonises the PI3K cascade, or causes mutations inPIK3CA, the catalytic subunit of PI3K,1114 leading toactivation of mTOR. mTOR in turn regulatestranscription and translation by phosphorylatingdownstream proteins, including the 40S ribosomalprotein S6 (pS6).15Inhibition of mTOR has been shownto sensitise HER2-positive, PTEN-deficient breasttumour xenografts to trastuzumab, so is an attractivetarget for treatment of patients with trastuzumab-resistant, advanced breast cancer.16

Everolimus is an mTOR inhibitor that has been shownto potentiate activity of trastuzumab and vinorelbine.17,18In phase 1 studies, the combination of chemotherapy,trastuzumab, and everolimus was associated with 44%of patients with trastuzumab-resistant breast cancer and

Lancet Oncol2014; 15: 58091

Published Online

April 15, 2014

http://dx.doi.org/10.1016/

S1470-2045(14)70138-X

See Commentpage 541

Department of Medical

Oncology, INSERM Unit U981,

Universit Paris Sud, Institut

Gustave Roussy, Villejuif,

France(Prof F Andr MD);

Winship Cancer Institute of

Emory University, Atlanta,

GA, USA(Prof R ORegan MD);

Cerrahpasa Medical Faculty,

Department of Medicine,

Division of Medical Oncology,

Istanbul University, Istanbul,

Turkey(Prof M Ozguroglu MD);

Kyoto University, Sakyo-ku,

Kyoto, Japan(Prof M Toi MD);

Department of MedicalOncology, Cancer Hospital

and Institute, Chinese

Academy of Medical Sciences

and Peking Union Medical

College, Chaoyang District,

Beijing, China(Prof B Xu MD);

Centre Hospitalier Universitaire

du Sart-Tilman, Domaine

Universitaire B35, Lige,

Belgium(Prof G Jerusalem MD);

NHO Osaka National Hospital,

Chuou-ku, Osaka, Japan

(N Masuda MD); Cancer Care

Centers of South Texas,

San Antonio, TX, USA

(S Wilks MD); NYU Langone

Arena Oncology, Lake Success,NY, USA(F Arena MD);

Lombardi Comprehensive

Cancer Center, Georgetown

University, Washington, DC,

USA(Prof C Isaacs MD); National

Cancer Centre Singapore,

Singapore(Y-S Yap MD);

Oncology Department, Military

Hospital, Budapest,

Hungary (Z Papai MD);

Orszagos Onkologiai Intezet,

Budapest, Hungary

(Prof I Lang MD); Christie

Hospital NHS Foundation

Trust, Manchester, UK

(A Armstrong MD); Sanatorio de

la Providencia, Buenos
http://crossmark.crossref.org/dialog/?doi=10.1016/S1470-2045(14)70138-X&domain=pdf

8/10/2019 BOLERO-3_2014

2/12

Articles

www.thelancet.com/oncology Vol 15 May 2014 581

55% in patients with taxane-resistant and trastuzumab-

resistant tumours achieving an overall response.

19,20

The combination of everolimus and trastuzumab in aphase 2 study was associated with 34% of patients withtrastuzumab-resistant tumours having a clinical benefit(defined as complete response plus partial response plusstable disease for 24 weeks).21

On the basis of these initial findings, the objective ofthis phase 3 trial was to compare the effi cacy and safety ofeverolimus and placebo when combined with trastuzumaband vinorelbine in patients with trastuzumab-resistant,HER2-positive, advanced breast cancer.

MethodsStudy design and participants

This study was an international, randomised, double-blind, placebo-controlled, phase 3 trial done at149 centres in 21 countries. Eligible patients hadtrastuzumab-resistant, HER2-positive, advanced breastcancer, and had received previous chemotherapy,including a taxane. Trastuzumab resistance was definedas recurrence during or within 12 months of adjuvanttreatment or progression during or within 4 weeks oftreatment for advanced disease. Patients had to be aged18 years or older, have histologically or cytologicallyconfirmed invasive breast carcinoma with locallyrecurrent or radiological evidence of metastatic disease,an Eastern Cooperative Oncology Group performancestatus (ECOG PS) of 2 or less, documentation of anegative pregnancy test, left ventricular ejectionfraction assessment showing a value greater than thelower limit of normal within 4 weeks beforerandomisation, and adequate organ and haematologicalfunctions. Exclusion criteria were more than threeprevious lines of chemotherapy for advanced disease,previous treatment with vinca alkaloids or mTORinhibitors, concomitant immunosuppressive agents orchronic corticosteroids, or radiotherapy to 25% or moreof bone marrow within 4 weeks before randomisation;previous lapatinib was allowed. Patients with stable andpreviously treated CNS metastases were eligibleprovided all treatment was completed 8 weeks or more

before randomisation. Patients with impairment ofgastrointestinal function, including active ulceration,peripheral neuropathy of grade 2 or worse atrandomisation, active cardiac disease or history ofcardiac dysfunction, known HIV infection, activebleeding diathesis, or on oral vitamin K medication(except low-dose warfarin or aspirin), clinicallysignificant third space fluid accumulation, and otherconcurrent or uncontrolled medical disorders wereexcluded. Patients with any malignancy within the past5 years, with the exception of adequately treated in-situcarcinoma of the cervix, or basal-cell or squamous-cellcarcinoma or non-melanoma skin cancer, were alsoexcluded. Women who were breast-feeding or pregnantwere excluded.

All patients provided written informed consent. The

study was done in accordance with the principles of GoodClinical Practice and the Declaration of Helsinki, andapproved by an independent ethics committee at each site.A steering committee supervised the conduct of the study,and an independent data and safety monitoring committeedid regular safety reviews.

Randomisation and maskingInvestigators enrolled patients who were then randomlyassigned 1:1 to receive either everolimus or placebo usinga central patient screening and randomisation system.Randomisation was stratified according to previouslapatinib use. An interactive voice and web responsesystem was used to gather screening information and

allocate treatment (Covance/Phase Forward/OracleCorporation, Redwood City, CA, USA). Patients andinvestigators (including local radiologists) were unawareof the assigned treatments from time of randomisationuntil time of final overall survival analysis. Prematureunblinding of study drug assignment was only allowedin case of emergency. Identity of experimental treatmentswas concealed by use of everolimus and placebo thatwere identical in packaging, labelling, appearance, andadministration schedule.

Aires, Argentina(G Lerzo MD);

Monash Medical Center,

Moorabbin Hospital,

Bentleigh East, VIC, Australia

(M White MD); Cabrini

Brighton Hospital, Brighton,

VIC, Australia(M White);

Comprehensive Breast Health

Center, RuiJin Hospital, School

of Medicine, Shanghai JiaoTong

University, Shanghai, China

(Prof K Shen MD); University

of Texas MD Anderson Cancer

Center, Houston, TX, USA

(J Litton MD); Oncology

Global Development,

Novartis Pharmaceuticals

Corporation, East Hanover,

NJ, USA(D Chen PhD,

Y Zhang PhD, S Ali, T Taran MD);

and Ospedale San Raffaele

Scientific Institute, Milan, Italy

(L Gianni MD)

Correspondence to:

Dr Luca Gianni, Department of

Medical Oncology, Ospedale

San Raffaele Scientific Institute,

Milan, Italy

[email protected]

Figure : Trial profile

731 assessed for eligibility

4 did not receive treatment

280 received treatment(safety set)

247 discontinued study 27 adverse event 193 disease progression 17 withdrew consent or

lost to follow-up 5 new cancer therapy

2 protocol deviation/ administrative problem 3 died

254 discontinued study 14 adverse event 221 disease progression 14 withdrew consent or

lost to follow-up 1 new cancer therapy

2 protocol deviation/ abnormal test result 2 died

284 allocated to everolimus +vinorelbine + trastuzumab(full analysis set)

285 allocated to placebo +vinorelbine + trastuzumab(full analysis set)

569 randomly assigned

162 did not meet entry criteriaor declined to participate

3 did not receive treatment

282 received treatment(safety set)

33 treatment ongoing 28 treatment ongoing

8/10/2019 BOLERO-3_2014

3/12

Articles

582 www.thelancet.com/oncology Vol 15 May 2014

Procedures

Vinorelbine and trastuzumab were administered weeklyin 3-week cycles to all patients. Vinorelbine was givenintravenously at 25 mg/m and trastuzumab was givenintravenously at 2 mg/kg after a 4 mg/kg loading dose onday 1 of cycle one (unless the last dose of previoustrastuzumab was within 4 weeks). 5 mg/day of oraleverolimus (Novartis Pharmaceuticals Corporation, EastHanover, NJ, USA) or matching placebo was given daily.The everolimus dose was established in a phase 1 study19in combination with trastuzumab and vinorelbine.Treatment continued until disease progression, un-acceptable toxicity, or withdrawal of patient consent. Doseadjustments were permitted if a patient was unable totolerate the protocol-specified dose. Two levels of dose

reduction of everolimus and placebo were permitted inthe study: 25 mg daily and 25 mg every other day. Anystudy medication that was interrupted for more than4 weeks consecutively could not be reintroduced. Patientswho discontinued one or more study drugs for any reasonbesides progression were allowed to continue the other

drugs, and followed the same assessment schedule until

progression. After treatment discontinuation, patientswere followed up for overall survival. Patients whodiscontinued study treatment and had not progressedwere followed up to assess tumour status, ECOG PS, andquality of life every 6 weeks until disease progression.Study discontinuation could occur because of withdrawalof consent, adverse events, abnormal laboratory values ortest procedure results, protocol violation, loss to follow-up, administrative problems, death, new cancer therapy,or disease progression.

A full physical examination, neurological examination,and assessment of ECOG PS and vital signs were donefor each patient during the screening period, on day 1 ofevery treatment cycle, and at the treatment discon-

tinuation visit. Haematology and biochemistry wereassessed at baseline and weekly thereafter. Tumourassessment included CT or MRI of the chest, abdomen,and pelvis at baseline and every 6 weeks until diseaseprogression.

Adverse events were monitored continuously usingNational Cancer Institute Common Terminology Criteriafor Adverse Events version 3.0. Assessments of leftventricular ejection fraction were done at baseline, every

Everolimus group

(n=284)

Placebo group

(n=285)

Women 284 (100%) 285 (100%)

Age, years

Median (range) 545 (3081) 540 (2577)

8/10/2019 BOLERO-3_2014

4/12

Articles


12 weeks (or more frequently if clinically indicated), andat time of treatment discontinuation (if the assessmenthad not been done within 2 weeks before the end-of-treatment visit). After treatment discontinuation, survivalinformation was obtained by telephone or at clinic visitsevery 3 months, and will continue until at least 388 deathshave been recorded.

Archival formalin-fixed, paraffi n-embedded (FFPE)tissue samples (80% from primary and 20% frommetastatic lesions) were obtained from trial participantswho had consented to the exploratory biomarker study. Allspecimens were stained with haematoxylin and eosin andassessed for the presence of tumour. Protein expressionwas assessed by immunohistochemistry using antibodiesD57.2.2.E and 138G6 from Cell Signaling Technology(Danvers, MA, USA) for pS6 and PTEN, respectively. Boththe staining intensity (0, 1+, 2+, 3+) and proportion of

positive tumour cells at each intensity and subcellularcompartment (cytoplasm for both proteins) were assessedand reported as histoscores. For mutational analysis ofPIK3CA gene exons 9 and 20 by Sanger sequencing,genomic DNA was extracted from FFPE tissue andamplified by PCR (one amplicon for exon 9; two ampliconsfor exon 20). PCR products were sequenced bidirectionallyusing the Applied Biosystems 3730 Genetic Analyzer (LifeTechnologies, Foster City, CA, USA).

OutcomesThe primary endpoint was locally assessed progression-free survival (PFS), defined as time from randomisation totime of first radiologically documented tumour progressionor death according to Response Evaluation Criteria In

Solid Tumors (RECIST version 1.0). The key secondaryendpoint was overall survival. Other secondary endpointswere the proportion of patients with an objective response(complete response plus partial response), the proportionof patients who achieved clinical benefit (completeresponse, partial response, or stable disease for 24 weeks),and safety. Additional secondary endpoints were time todeterioration of ECOG PS, change in quality of life scoresover time, duration of objective response, time to objectiveresponse, and pharmacokinetics of everolimus,vinorelbine, and trastuzumab. We also did severalpreplanned analyses of specific subgroups of patients.

A RECIST complete or partial response requiredconfirmation 4 weeks or more after initial observation. Toaccount for discrepancies between investigator assessmentand retrospective independent central radiology review, ablinded adjudication was done. Specifically, all cases that

were thought to be PFS events by either local or centralreviewers, but not both, and cases in which the date of thePFS event or censoring differed between local and centralreviews, were adjudicated. For this additional analysis,radiological images and relevant physical examinationfindings were reviewed retrospectively in a blindedmanner by an independent adjudication committeeconsisting of an expert radiologist and clinician.

Statistical analysisWe did effi cacy analyses on an intention-to-treat basisusing the full analysis set, which comprised all randomisedpatients. The safety population included all patients whoreceived at least one dose of any of the study drugs, andhad undergone at least one post-baseline assessment. We

Figure : Kaplan-Meier estimates of locally assessed progression-free survival in the full analysis set

Patients were stratified by previous lapatinib use. Symbols represent censoring events. PFS=progression-free survival.

100

80

60

40

20

0

0 6 12 18 24 30 36 42 48 54 60 66 72 78 84 90 96 102 108 114 120 126 132 138 144 150 156 162 168 174

Progression-freesurvival(%)

Time (weeks)

Number at risk

Everolimus

Placebo

284 259 233 200 161 126 98 78 54 40 35 26 18 14 14 9 5 4 2 2 1 1 1 1 1 1 1 1 1 0

285 253 202 177 138 109 85 64 49 38 26 23 19 16 12 10 7 4 3 3 1 1 1 0 0 0 0 0 0 0

EverolimusPlacebo

n/N

196/284219/285

Median PFS (95% CI)

700 (674818)578 (549690)

Everolimus

Placebo

HR=078 (95% CI 065095)Log-rank p=00067

8/10/2019 BOLERO-3_2014

5/12

Articles


analysed PFS using a stratified log-rank test. For the final

analysis, 417 PFS events were required to detect a hazardratio (HR) of 0727 with 90% power using a log-rank testand a two-look group-sequential Lan-DeMets design withan OBrian-Fleming-type boundary at a one-sidedcumulative 25% level of significance. This reduction inthe HR represented an absolute improvement in PFS of375%, corresponding to 225 months (assuming amedian PFS of 6 months in the placebo group). An interimanalysis was planned when 251 PFS events had occurred.Assuming 10% loss to follow-up and a 1:1 randomisation, asample size of 572 patients was needed. We estimated thedistribution of PFS according to the Kaplan-Meier method.We assumed that overall survival in eligible patients wouldbe about 18 months in the placebo group and 24 months in

the everolimus group.We have presented the proportions of patients who

achieved an overall response and clinical benefit bytreatment group with 95% CIs, and compared with theCochran-Mantel-Haenszel test. Exploratory analysis ofsubgroups was done using the multivariate Coxproportional hazards model, adjusted for baselinecharacteristics. We used descriptive statistics to summarise

adverse events and laboratory abnormalities. We used SAS

version 9.3.1 for the statistical analyses.This trial is registered with ClinicalTrials.gov, numberNCT01007942.

Role of the funding sourceThe study was designed by the investigators and funder.The funder provided study drugs, and participated inregulatory and ethics approval, safety monitoring, datacollection, and statistical analyses. All authors had fullaccess to study data for interpretation and analysis.No authors were paid to write this report, and they hadfinal responsibility for the decision to submit forpublication.

Results569 women were enrolled and randomly assigned toreceive everolimus (n=284) or placebo (n=285) betweenOct 26, 2009, and May 23, 2012, and were included in theintention-to-treat analysis (figure 1). Four patients in theeverolimus group did not receive treatment (two patientswere randomised in error, one withdrew consent, and onedid not have study treatment due to protocol deviation)and three patients in the placebo group did not receivetreatment (two patients deteriorated before treatment andone did not have study treatment due to protocoldeviation), and were not included in the safety analysis.Patient demographics and baseline disease characteristicswere well balanced between treatment groups (table 1).Overall, 439 (77%) of 569 patients had visceral metastasesand 267 (47%) had oestrogen-receptor-negative disease.110 (39%) of 284 patients in the everolimus group hadthree or more metastatic sites compared with 123 (43%)of 285 patients in the placebo group. Randomly assignedtreatment in the trial was first line in the metastaticsetting in 46 (16%) patients in the everolimus group and46 (16%) in the placebo group. The mean total number ofprevious regimens was 34 (SD 17) in the everolimusgroup and 34 (18) in the placebo group, including amean of 16 (085) and 16 (090) trastuzumab regimens,respectively. Most patients had received trastuzumab inthe metastatic setting; 78 (27%) patients in the everolimus

group and 64 (22%) in the placebo group had receivedtrastuzumab in the adjuvant or neoadjuvant setting only.Median time between the last dose of adjuvant orneoadjuvant trastuzumab and randomisation was39 months (IQR 10127) in the everolimus group and46 months (11103) in the placebo group. 135 (48%)patients in the everolimus group and 148 (52%) in theplacebo group had received trastuzumab for more than1 year. Of note, 124 (44%) patients in the everolimus groupand 137 (48%) in the placebo group had received two ormore trastuzumab-containing regimens. In accordancewith the protocol, all but one patient (in the everolimusgroup) had received a previous taxane, and 76 (27%)patients in the everolimus group and 76 (27%) in theplacebo group had received previous lapatinib. Median

Figure : Forest plot analysis of locally assessed progression-free survival

HR (95% CI) presented for progression-free survival in all subgroups, according to baseline characteristics.

ECOG PS=Eastern Cooperative Oncology Group performance status. ER=oestrogen receptor. HR=hazard ratio.

PR=progesterone receptor. *Each treatment interaction was assessed using a multivariate Cox proportional

model, which included treatment, covariate, and treatment-covariate interaction terms. Interaction term was

kept in the model only if corresponding p values were less than 005. Defined as use of lapatinib for 6 weeks ormore before randomisation.

n HR (95% CI)

All

Age

8/10/2019 BOLERO-3_2014

6/12

Articles


follow-up at the time of this analysis was 202 months(IQR 150271).

An interim analysis was done after 311 PFS events hadoccurred; the boundary to reject the hypothesis and stopthe trial at this point was Z=2350 (p

8/10/2019 BOLERO-3_2014

7/12

Articles


vinorelbine in 272 (97%) and 248 (88%) in the everolimusand placebo groups, respectively.

Grade 34 adverse events occurring in greater than orequal to 5% more patients in the everolimus group thanin the placebo group were neutropenia, stomatitis,anaemia, leucopenia, fatigue, and febrile neutropenia(table 4). Non-infectious pneumonitis was reported in28 (10%) of 280 patients in the everolimus group and12 (4%) of 282 patients in the placebo group; most ofthese events were grade 12, with six cases of grade 34non-infectious pneumonitis reported in the everolimusgroup and five cases reported in the placebo group.

Serious adverse events were noted in 117 (42%) of280 patients in the everolimus group and 55 (20%) of282 patients in the placebo group. The most frequentlyreported serious adverse events in the everolimus group

were febrile neutropenia in 29 (10%) patients (vs four[1%] patients in the placebo group), pyrexia in 13 (5%)patients (vsfive [2%]), neutropenia in 11 (4%) patients (vsthree [1%]), anaemia in ten (4%) patients (vs two [1%]),and stomatitis in nine (3%) patients (vs one [

8/10/2019 BOLERO-3_2014

8/12

Articles


Everolimus group (n=280) Placebo group (n=282)

Grade 12 Grade 3 Grade 4 Grade 12 Grade 3 Grade 4

Neutropenia 24 (9%) 98 (35%) 106 (38%) 22 (8%) 90 (32%) 85 (30%)

Stomatitis 138 (49%) 37 (13%) 0 74 (26%) 4 (1%) 0

Anaemia 85 (30%) 47 (17%) 6 (2%) 66 (23%) 16 (6%) 1 (

8/10/2019 BOLERO-3_2014

9/12

Articles


other to acute respiratory distress syndrome; in the placebogroup, one was due to acute respiratory failure and theother to pneumonia.

DiscussionThe present study shows that the addition of everolimus totrastuzumab and vinorelbine significantly improved PFSin patients with heavily pretreated HER2-positive,advanced breast cancer that had progressed duringtreatment with trastuzumab. We note that the Kaplan-Meier curves of PFS begin to separate early after initiation

of treatment, possibly as a result of a reversal oftrastuzumab resistance in a subset of patients. Onepossible explanation for this result is that activation of thePI3K/Akt/mTOR pathway, due to HER2 escape mecha-nisms, is inhibited.1115However, the addition of everolimusto a backbone of trastuzumab and vinorelbine wasassociated with a higher incidence of adverse events thanwas the addition of placebo, which should be consideredwhen assessing the clinical benefit of this regimen.

In BOLERO-3, trastuzumab plus vinorelbine combi-nation was chosen as the backbone treatment. Thiscombination has been assessed in two phase 3 randomisedtrials and has been recommended in treatment guidelinesfor first-line therapy of advanced breast cancer,8,22,23but hasnot been fully investigated as second-line or third-line

treatment (panel). In the current trial, the control grouphad a PFS of 578 months (95% CI 549690), suggestingthat trastuzumab plus vinorelbine might be considered asan appropriate backbone for trastuzumab-resistant, HER2-positive breast cancer.

Median relative dose intensity for vinorelbine in this trialwas 064 in the everolimus group compared with 073 inthe placebo group. The need to decrease the dose intensityof backbone chemotherapy is common in randomisedtrials of kinase inhibitors added to chemotherapy, havingbeen reported in clinical trials testing lapatinib, sunitinib,

and sorafenib in breast cancer.7,24,25 Regardless, we reportimproved outcomes with everolimus treatment even whendose reductions were made. Study discontinuation due toadverse events occurred in 10% of patients in theeverolimus group, and fewer patients discontinuedtreatment due to disease progression in the everolimusgroup than in the placebo group. Everolimus did not affectthe dose intensity of trastuzumab. The dose of everolimusused in this study was 5 mglower than that used forother populations and indicationsand was chosen on thebasis of the results of a phase 1 dose escalation trialincluding everolimus in combination with vinorelbine andtrastuzumab.19

Patients who received everolimus had a higherincidence of adverse events than did those in the placebo

Everolimus group (n=280) Placebo group (n=282)

Grade 12 Grade 3 Grade 4 Grade 12 Grade 3 Grade 4

(Continued from previous page)

Exertional dyspnoea 7 (3%) 1 (

8/10/2019 BOLERO-3_2014

10/12

Articles


group. Non-haematological adverse events that occurred

in the everolimus group with greater frequency than inthe placebo group were consistent with those reportedin previous trials of the drug and included stomatitis,fatigue, diarrhoea, pyrexia, and decreased appetite andweight.26,27We noted a higher incidence of haematologicaladverse events, including neutropenia, anaemia,leucopenia, thrombocytopenia, and febrile neutropenia,in the everolimus group. Although non-infectiouspneumonitis was reported in this trial, the incidencewas somewhat lower than that seen in BOLERO-2,28which explored the combination of everolimus andexemestane.

In the present study, the effi cacy of everolimus wasmore pronounced in patients with oestrogen-receptor-

negative breast cancer than in those with oestrogen-receptor-positive cancer, similar to previous studies withtrastuzumab-containing therapy.29 These findingssuggest that dual inhibition might not be suffi cient toprovide clinical benefit in patients with oestrogen-receptor-positive or HER2-positive cancers.30Substantialcross-talk exists between HER2 and oestrogen-receptorpathways such that inhibition of HER2 alone increasessignalling through the oestrogen receptor, which affectseffi cacy of HER2-directed agents. Thus, the combinationof everolimus and trastuzumab could be enhanced if theoestrogen receptor is inhibited concomitantly.30,31Furtherstudies should assess whether patients with oestrogen-receptor and HER2-dual positive breast cancer wouldbenefit from a combination of endocrine therapy,trastuzumab, and mTOR inhibition. Other futureanalyses could include assessment of the benefit ofeverolimus-containing therapy according to treatmenthistory, including the number of lines of previoustherapy and adjuvant versus first-line, second-line, orthird-line trastuzumab in the metastatic setting.Although drug development on the basis of biomarkerstatus is preferred, BOLERO-3 was done in a populationdefined only by HER2-positive status because of theabsence of solid previous evidence for other biomarkersin phase 1/2 trials. However, since mTOR is thought tobe activated via the PI3K/PTEN/Akt pathway, we explored

whether alteration of this pathway correlated withsensitivity to everolimus.14,3235Low PTEN expression andhigh pS6 seemed to be significantly associated withhigher sensitivity to everolimus. These data suggest thateverolimus was able to counter PI3K/Akt/mTORpathway activation in trastuzumab-resistant breastcancer, and that identification of patients who mightderive the greatest benefit from everolimus-containingtherapy is plausible. Everolimus provided less benefit topatients with oestrogen-receptor-positive disease and noevidence of PI3K/PTEN/Akt/mTOR pathway activation.Although biomarkers of the PI3K/Akt/mTOR pathwayhave been assessed in previous studies, these studieswere done in different populations of patients so do notcorrespond to the results described here.26,36

The role of everolimus in heavily pretreated patients

with advanced breast cancer should be considered in thecontext of newer agents, such as pertuzumab andtrastuzumab emtansine, which have recently showngood effi cacy in this population.9,37These agents are beingstudied for adjuvant treatment of breast cancer and forfirst-line treatment of advanced breast cancer. As a result,there remains an unmet need for treatment-experiencedpatients with advanced breast cancer. However, very fewpatients in our trial had received previous treatment with

Panel: Research in context

Systematic review

We did a PubMed search with the terms trastuzumab,

advanced breast cancer, and metastatic breast cancer.We restricted the review to English language original reports

of prospective clinical trials, and reviewed abstracts to

identify trials of trastuzumab-resistant disease. Some

significant advances have been made in the management of

trastuzumab resistance in recent years, so our review

included all trials published up until the end of 2013.

19 relevant trials reporting progression-free survival (PFS)

were identified. Some trials reported time to progression

rather than PFS; however, we have included those data with

PFS in this systematic review (see appendix for a

chronological list of all publications included). BOLERO-3 is

the first randomised trial testing everolimus in this setting

and, beyond this, testing a therapy that did not target the

HER family. Median PFS for all regimens tested ranged from

51 days to 142 months. In view of the range of treatments

tested, we also looked at PFS in trials, including single-agent,

HER2-directed therapy, a targeted therapy plus

chemotherapy, and trastuzumab-containing therapy. For

single-agent HER2 regimens, PFS ranged from 81 to

153 weeks for lapatinib and 46 to 142 months for

trastuzumab emtansine. PFS ranged from 41 to 94 months

for targeted therapy plus chemotherapy, and 51 days to

92 months for trastuzumab-containing therapy. Only the

phase 1/2 trial of everolimus in combination with

trastuzumab and paclitaxel comprised a triple regimen of

chemotherapy, trastuzumab, and a targeted agent.20

Interpretation

We found a high degree of heterogeneity in study design and

patient populations, making cross-trial comparisons diffi cult,

and they should be interpreted with caution. However, the

median PFS of 70 months reported here in the everolimus

group compares favourably with results from previous studies

of trastuzumab-resistant breast cancer in heavily pretreated

patients. In fact, the placebo group containing trastuzumab

and vinorelbine did well compared with other trastuzumab

plus chemotherapy regimens. Therefore, these results suggest

that mTOR inhibition with everolimus might offer a new

option for trastuzumab-resistant, advanced breast cancer,

although further work is necessary to fully delineate the

riskbenefit profile of such regimens.

8/10/2019 BOLERO-3_2014

11/12

Articles


pertuzumab or trastuzumab emtansine. Furthermore,

our results suggest that everolimus might be particularlyeffective in patients with oestrogen-receptor-negativedisease, and in those with low PTEN and high pS6expression. Validation of the results in these biomarkerpopulations could lead to an important role foreverolimus in these groups.

In conclusion, to the best of our knowledge,everolimus is the first non-HER2-targeted therapy toaddress the underlying mechanism of trastuzumabresistance. Most other therapies studied in this settinghave focused on continued HER2 inhibition.59The results of this study support the concept that, atleast in part, resistance to trastuzumab is sustained byaltered intracellular signalling, and justify that future

strategies could include the combination of everolimuswith therapies such as trastuzumab emtansine andpertuzumab. However, as with all therapies in thissetting, everolimus is associated with side-effects, so theriskbenefit profile of treatment with everolimus shouldbe considered.

Contributors

FA, GJ, YZ, and TT designed the study. FA and AA were responsible forpatient accrual. SA was responsible for trial management, and AA forclinical care. FA, ROR, MO, MT, BX, GJ, NM, SW, FA, CI, Y-SY, ZP, IL,AA, GL, MW, KS, JL, YZ, SA, TT, and LG collected the data. YZ was thetrial statistician. BX, GJ, SW, Y-SY, YZ, and TT analysed the data, andFA, BX, GJ, NM, Y-SY, IL, MW, YZ, SA, and TT interpreted the data.DC was responsible for the biomarker data generation, analysis, andinterpretation. All authors drafted and revised the manuscript, and had

final review and approval of the manuscript.

Declaration of interests

FA is a consultant for Novartis Pharmaceuticals Corporation, and hasreceived honorarium from Novartis. ROR is an adviser and consultantfor Novartis Pharmaceuticals Corporation, and has received researchsupport from Genentech and Novartis. MO has received honoraria forexpert testimony from Novartis Pharmaceuticals Corporation and Sanofi.MT has received research funding, honoraria, and consultancy fees fromNovartis Pharmaceuticals Corporation. GJ is an advisory board memberand consultant for Novartis Pharmaceuticals Corporation, and hasreceived lecture fees, a research grant, and a travel grant from Novartis.FA has received a study grant for this trial and has been paid as expertwitness by Novartis Pharmaceuticals Corporation. CI has receivedhonoraria from Novartis Pharmaceuticals Corporation. Y-SY has receivedhonoraria and a travel grant from Novartis Pharmaceuticals Corporation.AA has received travel grants and honoraria from Roche. MW is anadvisory board member for Novartis Australia. JL is institutionalprincipal investigator for BOLERO studies done by NovartisPharmaceuticals Corporation (no compensation received). DC is anemployee and stock holder of Novartis Pharmaceuticals Corporation.YZ is an employee of Novartis Pharmaceuticals Corporation. SA is anemployee of Novartis Pharmaceuticals Corporation, and receives stockawards from Novartis. TT is an employee and stock holder of NovartisPharmaceuticals Corporation. LG is a consultant or advisory boardmember for Roche, Genentech, GlaxoSmithKline, NovartisPharmaceuticals Corporation, Pfizer, Boehringer lngelheim, Celgene,and Tahio. BX, NM, SW, ZP, IL, GL, and KS declare that they have nocompeting interests.

Acknowledgments

We thank all the patients, investigators, and their institutions for thetime and effort put into this study. Douglas Robinson did the statisticalanalysis of biomarker data. Editorial support was provided bySusan DePetris and James Street of Phase Five Communications Inc.

(New York, NY, USA), and was supported by Novartis PharmaceuticalsCorporation.

References1 Slamon DJ, Clark GM, Wong SG, et al. Human breast cancer:

correlation of relapse and survival with amplification of theHER-2/neu oncogene. Science1987; 235:17782.

2 Slamon DJ, Leyland-Jones B, Shak S, et al. Use ofchemotherapy plus a monoclonal antibody against HER2 formetastatic breast cancer that overexpresses HER2. N Engl J Med2001; 344:78392.

3 Esteva FJ, Valero V, Booser D, et al. Phase II study of weeklydocetaxel and trastuzumab for patients with HER-2-overexpressingmetastatic breast cancer.J Clin Oncol 2002; 20:180008.

4 Montemurro F, Donadio M, Clavarezza M, et al. Outcome ofpatients with HER2-positive advanced breast cancer progressingduring trastuzumab-based therapy. Oncologist 2006; 11:31824.

5 Bartsch R, Wenzel C, Altorjai G, et al. Capecitabine andtrastuzumab in heavily pretreated metastatic breast cancer.J Clin Oncol2007; 25:385358.

6 Blackwell KL, Burstein HJ, Storniolo AM, et al. Randomizedstudy of lapatinib alone or in combination with trastuzumab inwomen with ErbB2-positive, trastuzumab-refractory metastaticbreast cancer.J Clin Oncol 2010; 28:112430.

7 Geyer CE, Forster J, Lindquist D, et al. Lapatinib plus capecitabinefor HER2-positive advanced breast cancer. N Engl J Med 2006;355:273343.

8 National Comprehensive Cancer Network. NCCN clinicalpractice guidelines in oncology (NCCN Guidelines) for breastcancer, V.2.2014. http://www.nccn.org/professionals/physician_gls/pdf/breast.pdf (accessed March 31, 2014).

9 Verma S, Miles D, Gianni L, et al. Trastuzumab emtansine forHER2-positive advanced breast cancer. N Engl J Med 2012;367:178391.

10 Baselga J, Gelmon KA, Verma S, et al. Phase II trial of pertuzumaband trastuzumab in patients with human epidermal growth factorreceptor 2-positive metastatic breast cancer that progressed duringprior trastuzumab therapy.J Clin Oncol2010; 28:113844.

11 Berns K, Horlings HM, Hennessy BT, et al. A functional geneticapproach identifies the PI3K pathway as a major determinant of

trastuzumab resistance in breast cancer. Cancer Cell2007;12:395402.

12 Nagata Y, Lan KH, Zhou X, et al. PTEN activation contributes totumor inhibition by trastuzumab, and loss of PTEN predictstrastuzumab resistance in patients. Cancer Cell 2004; 6:11727.

13 Pandolfi PP. Breast cancerloss of PTEN predicts resistance totreatment. N Engl J Med 2004; 351:233738.

14 Wang L, Zhang Q, Zhang J, et al. PI3K pathway activation resultsin low effi cacy of both trastuzumab and lapatinib. BMC Cancer 2011;11:248.

15 Steelman LS, Chappell WH, Abrams SL, et al. Roles of the Raf/MEK/ERK and PI3K/PTEN/Akt/mTOR pathways in controllinggrowth and sensitivity to therapy-implications for cancer and aging.Aging2011; 3:192222.

16 Lu CH, Wyszomierski SL, Tseng LM, et al. Preclinical testingof clinically applicable strategies for overcoming trastuzumabresistance caused by PTEN deficiency. Clin Cancer Res2007;13:588388.

17 Mondesire WH, Jian W, Zhang H, et al. Targeting mammaliantarget of rapamycin synergistically enhances chemotherapy-induced cytotoxicity in breast cancer cells. Clin Cancer Res 2004;10:703142.

18 Zhu Y, Zhang X, Liu Y, et al. Antitumor effect of the mTOR inhibitoreverolimus in combination with trastuzumab on human breastcancer stem cells in vitro and in vivo. Tumour Biol 2012; 33:134962.

19 Jerusalem G, Fasolo A, Dieras V, et al. Phase I trial of oral mTORinhibitor everolimus in combination with trastuzumab andvinorelbine in pre-treated patients with HER2-overexpressingmetastatic breast cancer. Breast Cancer Res Treat 2011; 125:44755.

20 Andre F, Campone M, ORegan R, et al. Phase I study of everolimusplus weekly paclitaxel and trastuzumab in patients with metastaticbreast cancer pretreated with trastuzumab.J Clin Oncol2010;28:511015.

21 Morrow PK, Wulf GM, Ensor J, et al. Phase I/II study oftrastuzumab in combination with everolimus (RAD001) in patientswith HER2-overexpressing metastatic breast cancer who progressedon trastuzumab-based therapy.J Clin Oncol 2011; 29:312632.

8/10/2019 BOLERO-3_2014

12/12

Articles

22 Andersson M, Lidbrink E, Bjerre K, et al. Phase III randomizedstudy comparing docetaxel plus trastuzumab with vinorelbine plus

trastuzumab as first-line therapy of metastatic or locally advancedhuman epidermal growth factor receptor 2-positive breast cancer:the HERNATA study.J Clin Oncol 2011; 29:26471.

23 Burstein HJ, Keshaviah A, Baron AD, et al. Trastuzumab plusvinorelbine or taxane chemotherapy for HER2-overexpressingmetastatic breast cancer: the trastuzumab and vinorelbine ortaxane study. Cancer 2007; 110:96572.

24 Baselga J, Segalla JG, Roche H, et al. Sorafenib in combination withcapecitabine: an oral regimen for patients with HER2-negative locallyadvanced or metastatic breast cancer.J Clin Oncol 2012; 30:148491.

25 Mayer EL, Dhakil S, Patel T, et al. SABRE-B: an evaluation ofpaclitaxel and bevacizumab with or without sunitinib as first-linetreatment of metastatic breast cancer. Ann Oncol 2010; 21:237076.

26 Baselga J, Campone M, Piccart M, et al. Everolimus inpostmenopausal hormone-receptor-positive advanced breastcancer. N Engl J Med 2012; 366:52029.

27 Motzer RJ, Escudier B, Oudard S, et al. Phase 3 trial of everolimusfor metastatic renal cell carcinoma: final results and analysis ofprognostic factors. Cancer 2010; 116:425665.

28 Yardley DA, Noguchi S, Pritchard KI, et al. Everolimus plusexemestane in postmenopausal patients with HR+ breast cancer:BOLERO-2 final progression-free survival analysis. Adv Ther2013;30:87084.

29 Bhargava R, Dabbs DJ, Beriwal S, et al. Semiquantitative hormonereceptor level influences response to trastuzumab-containingneoadjuvant chemotherapy in HER2-positive breast cancer.Mod Pathol2011; 24:36774.

30 Nahta R, ORegan RM. Therapeutic implications of estrogenreceptor signaling in HER2-positive breast cancers.Breast Cancer Res Treat2012; 135:3948.

31 Viale G, Regan MM, DellOrto P, et al; BIG 1-98 Collaborativeand International Breast Cancer Study Groups. Which patients

benefit most from adjuvant aromatase inhibitors? Results using acomposite measure of prognostic risk in the BIG 1-98 randomizedtrial. Ann Oncol2010; 22:220107.

32 Chandarlapaty S, Sakr RA, Giri D, et al. Frequent mutationalactivation of the PI3K-AKT pathway in trastuzumab-resistantbreast cancer. Clin Cancer Res2012; 18:678491.

33 Esteva FJ, Guo H, Zhang S, et al. PTEN, PIK3CA, p-AKT, andp-p70S6K status: association with trastuzumab response andsurvival in patients with HER2-positive metastatic breast cancer.Am J Pathol 2010; 177:164756.

34 Perez EA, Dueck AC, McCullough AE, et al. Impact of PTENprotein expression on benefit from adjuvant trastuzumab inearly-stage human epidermal growth factor receptor 2-positivebreast cancer in the North Central Cancer Treatment Group N9831Trial.J Clin Oncol 2013; 31:211522.

35 Razis E, Bobos M, Kotoula V, et al. Evaluation of the associationof PIK3CA mutations and PTEN loss with effi cacy of trastuzumabtherapy in metastatic breast cancer. Breast Cancer Res Treat 2011;

128:44756.36 Bachelot T, Bourgier C, Cropet C, et al. Randomized phase II trial

of everolimus in combination with tamoxifen in patients withhormone receptor-positive, human epidermal growth factor2-negative metastatic breast cancer with prior exposure toaromatase inhibitors: a GINEO study.J Clin Oncol2012;30:271824.

37 Baselga J, Corts J, Kim S-B, et al, for the CLEOPATRAStudy Group. Pertuzumab plus trastuzumab plus docetaxelfor metastatic breast cancer. N Engl J Med2012; 366:10919.

bolero-3_2014

Documents