ARTICLE IN PRESS

THE BREAST

0960-9776/$ - se

doi:10.1016/j.br

�CorrespondE-mail addr

Christos.sotirio

(M.J. Piccart).

The Breast 16 (2007) S10–S16

www.elsevier.com/locate/breast

Original article

The evolution of treatment strategies: Aiming at the target

Phuong Dinha, Christos Sotirioub, Martine J. Piccarta,�

aDepartment of Medical Oncology, Institut Jules Bordet, Universite Libre de Bruxelles (U.L.B), 121 Blvd de Waterloo, 1000 Brussels, BelgiumbTranslational Research and Functional Genomics Unit, Fonds de la Recherche Scientifique (F.N.R.S), Institut Jules Bordet,

Universite Libre de Bruxelles (U.L.B), 121 Blvd de Waterloo, 1000 Brussels, Belgium

Abstract

Historically, the selection of adjuvant systemic therapy in early breast cancer has relied on risk assessment embodied by the TNM

classification. Since the 2005 International St. Gallen Consensus, treatment selection now involves firstly identifying critical targets and

then using risk to assess the trade-off between anticipated toxicity and efficacy. This evolution in treatment strategies began with the

identification of the estrogen receptor, and culminated with the HER2 receptor, with recent astounding success in several adjuvant

trials.

Newer technologies including gene expression profiles and micrometastases tracking bear exciting potential in refining the treatment

strategies further. Alongside the progress in developing agents that target different molecules across the whole breast cancer population,

these newer technologies aim to tailor adjuvant treatment further by identifying breast cancer subgroups that may benefit most from

being targeted with specific therapy, by defining molecular subtypes, recognizing chemo-sensitivity and resistance, identifying at-risk gene

signatures, and by detecting stem cells capable of generating metastases.

This paper will review this evolution of treatment strategies, from the lessons learnt from the past, to the exciting promise of tailored

therapy of the future.

r 2007 Elsevier Ltd. All rights reserved.

Keywords: Breast cancer; Target; Treatment; Stem cell; Gene signatures

Introduction

Historically, the selection of adjuvant systemic therapyin early breast cancer has relied on risk assessmentincorporating both patient-related and tumor-relatedprognostic factors. These prognostic factors serve tocharacterize the background level of risk of relapse againstwhich the benefits and burdens of adjuvant therapies areweighed .1 Patient-related factors include age, menopausalstatus and co-morbidities. Tumor-related factors includelymph node involvement, tumor size, tumor grade andestrogen receptor (ER) status. More recently, HER2 statushas been added to this group of prognostic factors.

To assist with decisions in treatment selection, theseprognostic factors, with varying emphases are incorporated

e front matter r 2007 Elsevier Ltd. All rights reserved.

east.2007.07.032

ing author. Tel.: +322 541 34 06; fax: +322 538 08 58.

esses: phuong.dinh@bordet.be (P. Dinh),

u@bordet.be (C. Sotiriou), martine.piccart@bordet.be

into various treatment guidelines including the USNational Institutes of Health consensus (NIH)2 and theInternational St. Gallen Expert Consensus.3 These factors,though robust as prognostic markers, in fact, are weaker inpredicting for treatment responsiveness, apart from ERand HER2 status. This is because ER and HER2 positivityactually reflect molecular targets for specific therapyincluding tamoxifen and aromatase inhibitors for ER,and trastuzumab for HER2.The 9th International St. Gallen Breast Cancer Con-

gress,3 held in 2005, land marked a fundamental shift in theway that breast cancer patients are recommended forvarious anti-cancer treatments. Moving away from factorswith prognostic potential, ER status became the primaryconsideration for treatment choice, based on its predictivepotential for endocrine therapy. No longer considered aspart of the risk categories, it became the discriminating toolfor firstly deciding between treatment modalities—couldendocrine therapy be used alone—then using risk to assesstrade-off between anticipated efficacy and toxicity. This

ARTICLE IN PRESSP. Dinh et al. / The Breast 16 (2007) S10–S16 S11

principle of firstly selecting the target has since beenapplied to the targeted treatment of HER2 positive patientswith trastuzumab with astounding success.

This paper will review the evolution in the strategies fortreatment selection, with the gradual shift from riskprediction to the determination of tumor responsiveness.It will describe how the lessons learnt from adjuvantendocrine therapy particularly with tamoxifen have lead tothe recent success with adjuvant trastuzumab as well asdiscuss the exciting promise of newer technologies includ-ing gene expression profiling and stem cell tracking infurther tailoring adjuvant therapy.

Shifting from risk of relapse to tumor targets and

responsiveness as ‘‘driving’’ forces in adjuvant treatment

selection

Fig. 1 nicely summarizes the historical evolution ofadjuvant treatment strategies.

Two 57-year-old patients, with markedly differenttumors, are considered for adjuvant therapy selection inthe years 1990–2005, 2005–2010 and 2010–2015. In the firstperiod, with risk of relapse as the major driver of treatmentselection, the woman on the left is given the mostaggressive chemotherapy regimen.

In the second period, with endocrine responsivenessdriving the treatment choice, it is the woman on the rightwho is offered the ‘‘strongest’’ chemotherapy.

Speculation on how this evolution might continue leadsus to imagine further refinement of treatment selection inthe year 2010 and beyond, using molecular fingerprints inthe tumor as well as information on circulating tumor cells.

Fig. 1. Evolution of treatment stra

The estrogen receptor—the first critical target

Amongst breast cancer cells, the first critical target to beidentified was the ER. Tamoxifen is a selective modulatorof ER that competitively inhibits the binding of estrodioland in doing so, disrupts a series of mechanisms thatregulate cellular replication4 and proliferation. Tamoxifenhas been used in clinical trials since the 1970s, with initialnon-selective prescription across the whole breast cancerpopulation. In these unselected patients, tamoxifen canproduce responses of up to 30%.5,6

With time, the benefits of tamoxifen, in terms ofrecurrence and survival, was observed to be significantlyhigher in patients who were selected for treatmentaccording to their ER status. If selectively used in ERpositive breast cancers, responses of greater than 80%5,6

have been observed. In light of this, the recommendationfor routine ER testing emerged in the early 1990s tofacilitate tailored adjuvant therapy.More recently, the same results have been reported

by The Early Breast Trialists’ Collaborative Group(EBCTCG)7 in their 15-year meta-analysis of adjuvanttherapy. In ER-positive tumors, the use of tamoxifenresulted in absolute improvements in the 10-year survivalof 12.6% for node-positive patients and 5.3% for node-negative patients, independent of patient’s age, menopausalstatus, progesterone receptor status, and the use ofadjuvant chemotherapy. In the ER-poor tumors, thetamoxifen trials showed no benefit in either recurrencerisk or survival.The predictive potential of ER status in early breast

cancer, however, does not appear to be confined to

tegies: ‘‘Aiming at the target’’.

ARTICLE IN PRESSP. Dinh et al. / The Breast 16 (2007) S10–S16S12

endocrine therapy alone. There is now accumulatingevidence that the benefits of chemotherapy also differaccording to ER status, defined by histopathology. Muchof this evidence comes from the International BreastCancer Group (IBSCG), who was the first to detect largerchemotherapy benefit in ER-negative or low-ER breastcancer compared with ER-intermediate or ER-rich diseasein both the IBSCG Trial IX and the IBSCG Trial VII.

The IBSCG Trial IX8 with 1669 post-menopausal node-negative patients, showed that CMF followed by tamoxifensignificantly improved DFS compared with tamoxifenalone (RR ¼ 0.80; 95% CI 0.64–1.00) but this benefitwas seen exclusively in the ER-negative group (RR ¼ 0.52,95% CI 0.34–0.79) with no treatment difference in the ER-positive group (RR ¼ 0.99; 95% CI 0.75–1.30), after amedian follow-up of 6 years.

The IBSCG trial VII,9 with 1212 post-menopausal nodepositive patients, reported a detrimental effect of lateinitiation of CMF at 9, 12 and 15 months concurrent withtamoxifen that had been initiated 9 months earlier, seenexclusively in patients with ER-negative tumors, after amedian follow-up of 10 years. Those with ER-positivetumors benefited from the addition of chemotherapytogether with tamoxifen irrespective of the timing andduration of the chemotherapy.

Four other adjuvant studies have also provided addi-tional evidence that ER-negative and ER-positive breastcancers represent at least two different diseases withappreciable differences in treatment response.

The first of these was the long-term findings of theNSABP-B20 trial,10 where the benefit of adding CMF totamoxifen, in post-menopausal women with node-negativedisease, was shown to be inversely related to thequantitative concentration of ER.

The combined retrospective review of 3 CALGB/intergroup trials,11 namely the trial 8541 (comparing threeCAF regimens), the trial 9344 (incorporating paclitaxel insequence to AC) and the trial 9741 (examining dose-dense)also showed a greater incremental benefit of the experi-mental treatment regimen in ER-negative subgroups.

Similar results were also observed with neo-adjuvantchemotherapy in the MD Anderson analysis,12 where ER-negative tumors exhibited four times higher pCR rates thanER-positive tumors.

With the above results demonstrating the importance ofthe ER in predicting response to endocrine therapy as wellas chemotherapy, treatment guidelines have naturallyevolved to reflect this. The 2005 International St. GallenExpert Consensus3 represented this fundamental shift inthe treatment allocation paradigm, where ER, predictingfor endocrine responsiveness became the primary consid-eration for treatment, above risk itself.

Many lessons have, therefore, been learnt from theexperience with the ER (the critical target) especially in itspredictive potential for treatment responsiveness. Beyondits initial identification, the recognition of target sub-groups, based on ER status, enabled appropriate selection

of patients with the best chance of responding to endocrinetherapy only.

The features of the critical target lessons learnt from the ER

experience

In the future design of clinical trials, we must learn fromthe lessons derived from the invaluable experience withER—the first critical target in adjuvant treatment of earlybreast cancer.

1.

The target must be identified, and its critical role intumorigenesis be well understood.

2.

The target must be measured by quantitative methods,because tumor responsiveness may vary with differentlevels of target positivity.

3.

The duration of the target therapy must be determined,to maximize clinical benefit and minimize exposure toside effects.

4.

Modulators of the target must be understood, includingphysiology, dosing and toxicity profiles.

More recently, with these important lessons in mind, theadjuvant trastuzumab trials were designed and conductedin HER2 positive patients only. By selecting for a specificsubgroup of breast cancer patients, the clinical benefitobserved was not only overwhelmingly significant but alsohighly reproducible from one trial to the other.

The HER2 receptor—the success with trastuzumab trials

The HER2 receptor is a 185 kd transmembrane glyco-protein that has an important role in mediating cell growth,differentiation and survival. Its amplification or over-expression occurs in 25–30% of breast cancer patients andis associated with an aggressive tumor phenotype andreduced survival rate.13,14

In the 1980s, trastuzumab, a humanized monoclonalantibody directed against the extracellular domain of theHER2 receptor, was developed. Although approved for usein the metastatic breast cancer since 1998, the results of 5adjuvant clinical trials were only reported in 2005. Despitedifferences in study design and short follow-up periods, the5 trials, HERA,15 the combined North American trialsNSABP-B31 and, NCCTG/N9831,16 BCIRG 00617 andFinHER,18 showed consistently remarkable results withreduction of the recurrence rate between 39% and 52%and mortality by 30%.Of note, the results of HERA15 comparing 1-year

trastuzumab to observation, at 23-month median follow-up, showed an unadjusted hazard ratio (HR) of 0.64 (95%CI 0.54–0.76; po0.0001), which corresponded to anabsolute disease free survival (DFS) benefit of 6.3%.This degree of benefit in early breast cancer is the largest

reported since the introduction of tamoxifen in hormonereceptor positive disease. The resounding success of thesetrials can be attributed to the selected treatment withtrastuzumab in only HER2 positive patients. Conceivably,

ARTICLE IN PRESSP. Dinh et al. / The Breast 16 (2007) S10–S16 S13

if conducted in unselected patients, the magnitude ofbenefit in these trials would undoubtedly be diluted, orperhaps, become non-apparent.

Even with the unfettered success of these trials, selectionof HER2 patients for trastuzumab therapy can be furtherrefined. Indeed, new promising biomarkers of adjuvanttrastuzumab benefit have emerged with the progress intranslational oncology, and these may serve to define theHER2 patient population into even smaller subgroups.Amongst these, c-myc has been shown to be co-amplified in30% of patients with exquisite sensitivity to trastuzumab.19

On the other hand, resistance to trastuzumab appearslinked to phosphatase and tensin homologue (PTEN)deleted on chromosome (10), which is lost or reduced inexpression in 40% of HER2-overexpressing breast can-cers,20,21 and also to p95HER2 (a truncated HER2receptor lacking the external domain), which is seen inapproximately 10%.22 Although exciting in their potential,the clinical utility of these markers will need to be testedprospectively in large trials such as ALTTO (adjuvantLapatinib and/or Trastuzumab treatment optimization)before being adopted to clinical practice.

Future direction

Adjuvant systemic therapy for early breast cancer hashistorically been focused on tumor burden, embodiedwithin the TNM classification. While this has served us wellup until now, the future of adjuvant therapy will bedependent on our understanding of tumor biology.

The identification of new molecular targets will bearpromise to the development of new therapies whilst theenhanced understanding of other suspected tumor targetssuch as VEGF, PDGF, IGF, proteosome UPA/PAIIalongside already known targets of ER, PR and HER2,will make possible the combination of multiple targetedtherapies.

However, simply knowing that the target exists is notsufficient to tailor adjuvant therapy, and newer technolo-gies are also aimed at identifying breast cancer subgroupsthat are most likely to respond to the targeted therapy.

These newer technologies include gene expression profil-ing and micrometastases tracking. From the initial break-through with molecular subtypes and prognostic gene-expression signatures, gene expression profiling has alsoundergone a significant evolution towards predictingtreatment response, similar to that of the ER experience.

This evolution towards firstly identifying the target isfurther reinforced by the major advances in micrometas-tasis tracking, and both are briefly discussed below.

Gene expression profiling

Micro-array technology, by allowing for a simultaneousstudy of the expression of multiple genes, has lead to thediscovery of different molecular subtypes, with distinct

clinical characteristics capable of providing useful prog-nostic information.With the seminal study by Perou et al.23 at least 4

subtypes were identified, based on the expression of 500‘‘intrinsic’’ genes. The first two subtypes included the basal-like (expressing cytokeratins 5, 6, 17, laminin and fatty-acidBP7) characterized by low or absent expression of ER andER-related genes and the HER2+ (expressing genes in theerbb7 amplicon such as GRB7). Amongst the ER positivetumors, 2 subtypes were identified; luminal A (expressingcytokeratin 8, 18 and other breast luminal genes) andluminal B (expressing a lower level of breast luminalgenes).Not only did they differ in their gene expressions

particularly with ER and ER-related genes, these mole-cular subtypes differed significantly in their clinicalbehaviors and treatment responsiveness. The basal-likeand HER2+ subtypes were likely to be more aggressiveincluding a higher proportion of TP53 mutations,24,25 anda markedly higher likelihood of being grade III (po0.0001,and P ¼ 0.0002) than luminal A tumors. However, despitea poorer prognosis, they tended to respond better tochemotherapy as demonstrated by a higher pathologiccomplete response with neo-adjuvant chemotherapy.26 Onthe other hand, fewer than 20% of luminal A subtype hadmutations in TP53,24 and these tumors were often grade I.They tended to be more sensitive to endocrine therapy witha better clinical outcome.Apart from this novel molecular classification, micro-

array technology has also resulted in various geneexpression signatures, also bearing prognostic potential.The Netherlands Cancer Institute (NKI) was the firstgroup to report a 70-gene prognostic signature (Mamma-printR), using the Agilent platform, in 78 fresh frozenbreast cancer samples from patients younger than 55 yearsof age and systemically untreated node-negative disease.27

Subsequently validated on a larger set of 295 patients fromthe same institution, including both node-negative andnode-positive disease treated and untreated,28 they showedthat the 70-gene signature was the strongest predictor fordistant metastases-free survival, independent of adjuvanttreatment, tumor size, histological grade and age. TheRotterdam group, using the Affymetrix platform, lateridentified a 76-gene prognostic signature in 115 untreatednode-negative breast cancer patients, with 80 patients beingER positive.29 Subsequently validated in 171 patients, the76-gene signature also predicted well for distant metas-tases-free survival. Recently, these two signatures wereindependently validated by the TRANSBIG consortium,showing that both are better in predicting clinical outcomecompared to other risk assessment tools including that ofAdjuvantOnline, the Nottingham prognostic index and theInternational St. Gallen expert consensus.30,31

Several other gene expression signatures with powerfulprognostic information have also been reported, includingthe OncotypeDxTM,32 the genomic grade,33 the wound-response signature,34 the wild/mutant p53 signature35 and

ARTICLE IN PRESSP. Dinh et al. / The Breast 16 (2007) S10–S16S14

the ‘‘invasiveness’’ gene signature.36 Interestingly, all ofthese signatures, despite the disparity in their gene lists,carry similar information regarding prognostication mainlydriven by proliferation.37–39 Additionally, they seem to bevery useful for determining the risk of recurrence inthe ER+subgroup and less informative for ER- andERBB2+disease.39

Although improving breast cancer prognostication is ofcritical importance for better identifying those patients notneeding treatment, it is also critically important to knowwhich therapy will benefit the individual patient. Therefore,several investigators have applied microarray technologyto try and identify gene expression signatures that couldpredict drug sensitivity in breast cancer.

In predicting for endocrine therapy resistance, severalstudies have been performed; including the study with 44genes by Jansen et al.40 and the one using the expressionratio of two genes—homeobox B13 and IL17BR, whichcould predict for disease survival with 80% accuracy.41–43

In the context of chemotherapy, few studies have beenreported so far because these studies ideally requireprospective sample collection in the context of a well-designed clinical trial. Nonetheless, several groups haveidentified genes associated with response to chemother-apy.44–51 In fact, the best studies looking at chemotherapyresponse have been those conducted in the neoadjuvantsetting, examining pathological response rate (pCR).

By looking at responders versus non-responders in 89patients treated with neoadjuvant paclitaxel and doxor-ubicin for locally advanced breast cancer, Gianni et al.52

found 86 genes that correlated to pCR, which was morelikely to occur with a higher expression of proliferation-related genes (including CDC20, E2F1, MYBL2, TO-PO2A) and immune-related genes (including MCP1,CD68, CTSB, CD18, ILT-2, CD3z, FasL, HLA.DPB1)and with a lower expression of estrogen (ER)-related genes(including ER, PR, SCUBE2, GATA3).

Similarly, the MD Anderson group53 also developed a30-probe set of pharmacogenomic predictor for pCR in 82patients treated with neoadjuvant weekly paclitaxel fol-lowed by fluorouracil/doxorubicin/cyclophosphamide(T/FAC) with high sensitivity and negative predictivevalue, compared to clinical variable-based (ER, age, grade)predictors.

Recently, using a ‘‘hypothesis-driven’’ approach, Duke’sresearchers identified several expression patterns associatedwith the deregulation of a variety of oncogenic pathwaysthat could predict response to different therapeutic agentstargeting specific deregulated pathways.54 The same group,using publicly available drug sensitivity data derived fromin vitro experiments, developed multiple classifiers ofresponse to a variety of chemotherapy drugs and showedthat a combination of these classifiers accurately predictedresponse to preoperative multi-drug regimen treatmentsderived from two breast cancer studies.55

Although the primary goal of all these studies was toidentify gene expression profiles predicting response to a

given regimen, they have provided crucial new insights intobreast cancer biology and mechanisms of resistance. As aconsequence, new therapeutic targets, including the micro-tubule-associated protein, tau (MAPT), associated withpaclitaxel resistance, have been discovered.Strategies for treatment allocation have thus clearly

evolved with the advent of microarray technology,generating the molecular subtype classification and thevarious gene expression signatures. Whilst their prognosticpotential can help select the right type of patient, theirpredictive potential can also help to select the right type oftreatment regimen.As exciting as their potential, it is still equally important

to validate the true clinical utility of gene expressionprofiles in large randomised prospective trials beforeimplementation into routine practice. It is howeverimportant to start designing these trials for each of thekey molecular breast cancer subtypes.

Tracking micrometastases

A provocative finding in the validation studies of the 70-gene and 76-gene signatures by TRANSBIG is the time-dependency phenomenon.32,56 Although both signaturesappeared strong predictors of the development of distantmetastases within 5 years, their prognostic ability decreasedwith longer periods of follow-up. This phenomenonsuggests that the molecular mechanisms involved in thedevelopment of early and late distant metastases may, infact, be different, and call for the need to study further themetastatic process including circulating tumor cells (CTC),as well as bone-marrow micrometastases (DTC).In this metastatic process, the critical step occurs when

cancer cells enter the blood circulation and home to adistant site, and up until recently, this step was believed tooccur late in the tumor progression process. However,observations by the Klein group have suggested thatthe dissemination of tumor cells, indeed, is an earlyevent.57,58In these studies, disseminated cancer cells wereisolated from the bone marrow of breast cancer patientsusing micro-manipulation and after global amplification oftheir genomic DNA, a comprehensive analysis of theirnumerical chromosomal aberrations was performed. Bycomparing the data from disseminated cancer cells withprimary tumors and, in the context of clinical staging,several interesting results were observed.Firstly, breast cancer cells from bone marrow repre-

sented early stages of the genomic progression of theprimary tumors. Secondly, at the stage of minimal residualcancer, disseminated cancer cells displayed very hetero-geneous chromosomal aberrations. Thirdly, at the stagethat manifested metastasis, clonal expansion of a specificgenotype displaying stable genetic aberrations becameevident. Fourthly, the genotype of single disseminatedcancer cells from bone marrow predicted, with highaccuracy, the clinical stage and clinical outcome of thepatient.

ARTICLE IN PRESSP. Dinh et al. / The Breast 16 (2007) S10–S16 S15

Indeed, several studies using immunocytochemistry todetect bone marrow DTCs at the time of diagnosis, foundtheir presence was associated with poor outcome. Thesefindings were nicely confirmed in a meta-analysis involvingmore than 4500 patients with early breast cancer59.Similarly, using Q-RT-PCR to detect Cytokeratin-19(CK-19) mRNA-positive circulating tumor cells (CTCs)in the peripheral blood before adjuvant treatment wasfound to be an independent prognostic factor for earlydisease recurrence and decreased survival in patients withaxillary node-negative breast cancer.60

These intriguing results strongly suggest a parallelevolution model of breast cancer progression, in whichthe local and systemic disease progress in parallel andindependently. It would, therefore, be important not onlyto detect but also to directly analyze systemically spreadcancer cells in order to identify novel therapeutic targets.This may lead to the development of novel adjuvanttherapeutics based both on the primary tumor as well asmicrometastatic cells.

Concluding remarks

Breast cancer is a complex heterogeneous disease, andthe complex evolution of adjuvant treatment strategiesillustrates this well. Established prognostic markers havehistorically reflected clinical and histopathological featuresof the tumor. Since the 2005 St. Gallens InternationalExpert Consensus as well as the advent of micro-arraytechnology, newer prognostic markers have evolved toreflect tumor biology.

Beginning with the ER, and culminating with the HER2receptor, the future of adjuvant therapy will be in theidentification of targets. Whether the targets are molecu-larly based such as ER and HER2, or micrometastasisbased, both clinical trials and subsequent treatment shouldbe tailored to select appropriate breast cancer subgroups,and not empirically applied to the whole breast cancerpopulation.

Gene expression profiling, DTCs and CTCs bear greatpromise as new tools for dissecting the complexity of breastcancer and providing the platform for tailored adjuvanttherapy, but still require prospective validation beforeroutine clinical use.

As we continue to evolve from empirical oncology totailored therapy, cross-talk and cooperation between keygroups will be critical including patients, clinicians,scientists, governments and industry. The financial, scien-tific and ethical burdens must be discussed and shared inorder to facilitate an evolution towards an ultimate cure.

References

1. Gelber RD, et al. Tailoring adjuvant treatments for the individual

breast cancer patient. The Breast 2003;12:558–68.

2. Eifel P, et al. National Institutes of Health Consensus Development

Conference Statement: adjuvant therapy for breast cancer, November

1–3, 2000. J Natl Cancer Inst 2001;93(13):979–89.

3. Goldhirsh A, et al. Meeting highlights: international expert consensus

on the primary therapy of early breast cancer. Ann Oncol 2005;

16(10):1569–83.

4. Jordan VC, Dowse LJ. Tamoxifen as an anti-tumor agent: effect on

oestrogen binding. J Endocrinol 1976;68:297–303.

5. Buzdar AU, Hortobagyi G. Update on endocrine therapy for breast

cancer. Clin Cancer Res 1998;4:527–34.

6. Ravdin PM, et al. Prognostic significance of progesterone receptor

levels in estrogen receptor-positive patients with metastatic breast

cancer treated with tamoxifen: results of a prospective Southwestern

Oncology Group study. J Clin Oncol 1992;10:1284–91.

7. Early Breast Cancer Trialists’ Collaborative Group. Effects of

chemotherapy and hormonal therapy for early breast cancer on

recurrence and 15-year survival: an overview of the randomised trials.

Lancet 2005;265:1687–717.

8. International Breast Cancer Study Group. Endocrine responsiveness

and tailoring adjuvant therapy for postmenopausal lymph node-

negative breast cancer: a randomized trial. J Natl Cancer Inst 2002;

94:1054–65.

9. International Breast Cancer Study Group. Effectiveness of adjuvant

chemotherapy in combination with tamoxifen for node-positive

postmenopausal breast cancer patients. J Clin Oncol 1997;15:1385–94.

10. Fisher B, Jeong J-H, Bryant, et al. Treatment of lymph-node negative,

estrogen-receptor-positive breast cancer: long term findings from

National Surgical Adjuvant Breast and Bowel Project randomized

clinical trials. Lancet 2004;364:858–68.

11. Berry DA, et al. Effects of improvements in chemotherapy on disease-

free and overall survival of estrogen-receptor negative, node-positive

breast cancer: 20-year experience of the CALGB and US Breast

Intergroup. Proc. SABCS. 2004;88(1) (Abstract 29).

12. Buzdar AU, et al. Pathological complete response to chemotherapy is

related to hormone receptor status. Proc SABCS 2003;82(1) Abstract

302.

13. Slamon DJ, et al. Human breast cancer: correlation of relapse and

survival with amplification of the HER-2/neu oncogene. Science

1987;235:177–82.

14. Slamon DJ, et al. Studies of the HER-2/neu proto-oncogene in human

breast and ovarian cancer. Science 1989;244:707–12.

15. Piccart-Gebhart MJ, et al. Trastuzumab after adjuvant chemo-

therapy in HER2-positive breast cancer. N Engl J Med 2005;353:

1659–72.

16. Romond EH, et al. Trastuzumab plus adjuvant chemotherapy for

operable HER2-positive breast cancer. N Engl J Med 2005;353:

1673–84.

17. Slamon D, et al. Phase III randomized trial comparing doxorubicin

and cyclophosphamide followed by docetaxel (AC T) with doxor-

ubicin and cyclophosphamide followed by docetaxel and trastuzumab

(AC TH) with docetaxel, carboplatin and trastuzumab (TCH) in

HER2 positive early breast cancer patients: BCIRG 006 study. Breast

Cancer Res Treat 2005;94(suppl 1):S5 Abstract 1.

18. Joensuu H, et al. Trastuzumab in combination with docetaxel or

vinorelbine as adjuvant treatment of breast cancer: the FinHer trial.

Breast Cancer Res Treat 2005;94(Suppl 1) Abstract 2.

19. C. Kim, et al., Trastuzumab sensitivity of breast cancer with co-

amplification of HER2 and c-MYC suggests pro-apoptotic function

of dysregulated c-MYC in vivo, San Antonio Breast Cancer

Symposium 2005; Abstract 46.

20. Nagata Y, et al. PTEN activation contributes to tumour inhibition by

trastuzumab, and loss of PTEN predicts trastuzumab resistance in

patients. Cancer Cell 2004;6:117–27.

21. Fujita T, et al. PTEN activity could be a predictive marker for

trastuzumab in the treatment of erbB2-overexpressing breast cancer.

Br J Cancer 2006;94(2):247–52.

22. Molina MA, et al. Trastuzumab (herceptin), a humanized anti-HER2

receptor monoclonal antibody, inhibits basal and activated HER2

ARTICLE IN PRESSP. Dinh et al. / The Breast 16 (2007) S10–S16S16

ectodomain cleavage in breast cancer cells. Cancer Res. 2001;

61(12):4744–9.

23. Perou CM, et al. Molecular portraits of human breast tumors. Nature

2000;406:747–52.

24. Sorlie T, et al. Gene expression patterns of breast carcinomas

distinguish tumor subclasses with clinical implications. Proc. Natl.

Acad. Sci. USA 2001;98:10869–74.

25. Carey LA, et al. Race and the poor prognosis basal-like breast cancer

(BBC) phenotype in the population-based Carolina Breast Cancer

Study. J. Clin. Oncol. 2004;suppl (abstract 9510).

26. R. Rouzier, et al., Basal and luminal types of breast cancer defined by

gene expression patterns respond differently to neoadjuvant che-

motherapy. San Antonio Breast Cancer Symposium. San Antonio,

TX 2004; abstract 1026.

27. van’t Veer LJ, et al. Gene expression profiling predicts clinical

outcome of breast cancer. Nature 2002;415(6871):530–6.

28. van de Vijver MJ, et al. A gene-expression signature as a predictor of

survival in breast cancer. Nature 2002;347(25):1999–2009.

29. Wang Y, et al. Gene-expression profiles to predict distant metastases

of lymph-node-negative primary breast cancer. Lancet 2005;

365(9460):671–9.

30. Buyse M, Loi S, van’t Veer L, et al. Validation and clinical utility of a

70-gene prognostic signature for women with node-negative breast

cancer. J. Natl. Cancer Inst. 2006;98(17):1183–92.

31. C. Desmedt, et al., Strong time-dependency of the 76-gene prognostic

signature for node-negative breast cancer patients in the TRANSBIG

multi-centre independent validation series. Late breaking abstract,

Fifth European Breast Cancer Conference, 2006.

32. Paik S, et al. A multigene assay to predict recurrence of tamoxifen-

treated, node-negative breast cancer. N. Engl. J. Med. 2004;351:

2817–26.

33. Sotiriou C, et al. Gene expression profiling in breast cancer:

understanding the molecular basis of histologic grade to improve

prognosis. J. Natl. Cancer Inst. 2006;98(4):262–72.

34. Chang HY, et al. Gene expression signature of fibroblast serum

response predicts human cancer progression: similarities between

tumors and wounds. PLoS Biol. 2004;2(2):E7.

35. Miller LD, et al. An expression signature for p53 status in human

breast cancer predicts mutation status, transcriptional effects, and

patient survival. Proc. Natl. Acad. Sci. USA 2005;102:13550–5.

36. Liu R, et al. The prognostic role of a gene signature from tumorigenic

breast-cancer cells. N. Engl. J. Med. 2007;356:217–26.

37. Loi S, et al. Definition of clinically distinct molecular subtypes in

estrogen receptor-positive breast carcinomas through genomic grade.

J. Clin. Oncol. 2007;25:1239–46.

38. Fan C, et al. Concordance among gene-expression-based predictors

for breast cancer. N. Eng. J. Med. 2006;355:560–9.

39. Sotiriou C, et al. Comprehensive analysis integrating both clinico-

pathological and gene expression data in more than 1500 samples:

proliferation captured by gene expression grade index appears to be

the strongest prognostic factor in breast cancer (BC). Proc. Am. Soc.

Clin. Oncol. 2006;24 (abstr 507).

40. Jansen MP, et al. Molecular classification of tamoxifen-resistant

breast carcinomas by gene expression profiling. J. Clin. Info. 2005;23:

732–40.

41. Ma XJ, et al. A two-gene expression ratio predicts clinical outcome in

breast cancer patients treated with tamoxifen. Cancer Cell 2004;5:

607–16.

42. Ma XJ, et al. The HOXB13:IL17BR expression index is a pro-

gnostic factor in early-stage breast cancer. J. Clin. Oncol. 2006;24:

4611–9.

43. Jansen MP, et al. HOXB13-to-IL17BR expression ratio is related with

tumor aggressiveness and response to tamoxifen of recurrent breast

cancer: a retrospective study. J. Clin. Oncol. 2007;25:662–8.

44. Folgueira MA, et al. Gene expression profile associated with response

to doxorubicin-based therapy in breast cancer. Clin. Cancer Res.

2005;11:7434–43.

45. Hannemann J, et al. Changes in gene expression associated with

response to neoadjuvant chemotherapy in breast cancer. J. Clin.

Oncol. 2005;23:3331–42.

46. Chang JC, et al. Gene expression profiling for the prediction of

therapeutic response to docetaxel in patients with breast cancer.

Lancet 2003;362:362–9.

47. Chang JC, et al. Patterns of resistance and incomplete response

to docetaxel by gene expression profiling in breast cancer patients.

J. Clin. Oncol. 2005;23:1169–77.

48. Iwao-Koizumi K, et al. Prediction of docetaxel response in human

breast cancer by gene expression profiling. J. Clin. Oncol. 2005;23:

422–31.

49. Bertucci F, et al. Gene expression profiling for molecular character-

ization of inflammatory breast cancer and prediction of response to

chemotherapy. Cancer Res. 2004;64:8558–65.

50. Andre F, et al. DNA arrays as predictors of efficacy of adjuvant/

neoadjuvant chemotherapy in breast cancer patients: current data

and issues on study design. Biochim. Biophys. Acta. 2006;1766:

197–204.

51. Rouzier R, et al. Microtubule-associated protein tau: a marker of

paclitaxel sensitivity in breast cancer. Proc. Natl. Acad. Sci. USA

2005;102:8315–20.

52. Gianni L, et al. Gene expression profiles in paraffin-embedded core

biopsy tissue predict response to chemotherapy in women with locally

advanced breast cancer. J. Clin. Oncol. 2005;23:7265–77.

53. Hess KR, et al. Pharmacogenomic predictor of sensitivity to

preoperative chemotherapy with paclitaxel and fluorouracil, doxor-

ubicin, and cyclophosphamide in breast cancer. J. Clin. Oncol. 2006;

24:4236–44.

54. Bild A, et al. Oncogenic pathway signatures in human cancers as a

guide to targeted therapies. Nature 2006;439:353–7.

55. Potti A, et al. Genomic signatures to guide the use of chemother-

apeutics. Nat. Med. 2006;12:1294–300.

56. Desmedt C, et al. Strong time-dependency of the 76-gene prognostic

signature for node-negative breast cancer patients in the TRANSBIG

multi-centre independent validation series. Clin. Cancer Res. 2007;

13(11):3207–14.

57. Schmidt-Kittler O, et al. From disseminated cells to overt metastasis:

genetic analysis of systemic breast cancer progression. Proc Natl Acad

Sci USA 2003;100(13):7737–42.

58. Klein CA, et al. Systemic cancer progression and tumor dormancy:

mathematical models meet single cell genomics. Cell Cycle Aug

2006;5(16):1788–98 August 15, 2006.

59. Braun S, et al. A pooled analysis of bone marrow micrometastases in

breast cancer. N Engl J Med 2005;353:793–802.

60. Xenidis N, et al. Predictive and prognostic value of peripheral blood

cytokeratin-19 mRNA-positive cells detected by real-time polymerase

chain reaction in node-negative breast cancer patients. J Clin Oncol

2006;24:3756–62.