Targeting Anthracyclines in Early Breast Cancer: New Candidate Predictive Biomarkers Emerge

Home , Anthracycline

Oncogene (2010) 29, 5231–5240 & 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 www.nature.com/onc REVIEW Targeting anthracyclines in early breast cancer: new candidate predictive biomarkers emerge

AF Munro, DA Cameron and JMS Bartlett

Endocrine Cancer Group, Edinburgh Cancer Research Centre, Western General Hospital, Edinburgh, UK

The search for a predictive marker of sensitivity to (Minotti et al., 2004). Their discovery in the 1960s anthracycline-based chemotherapy has proven challen- and subsequent introduction into clinical practice led ging. Despite human epidermal growth factor receptor 2 to significant improvements in the treatment of some (HER2) being a strong prognostic marker in breast cancers, most particularly lymphomas, breast cancer cancer, the only therapies with which there is a recognized and sarcoma. More recently, they are increasingly used functional link to the HER2 oncogene are those directly in the treatment of non-metastatic gastric and lower targeting the molecule itself. Despite this, HER2 has been oesophageal cancers, breast, ovarian and lung cancers extensively assessed as a predictive marker in a variety of (Cortes-Funes and Coronado, 2007). Although effec- chemotherapy regimens including anthracyclines. Analysis tive, these drugs have serious side effects, most notably of anthracycline response in patients with HER2 ampli- cardiotoxicity, which is often under-reported because of fication has given conflicting results. This led to the it’s late onset, and is becoming more of a clinical issue as suggestion that HER2 amplification was acting as a treatments result in more long-term survivors of cancer, surrogate for the gene encoding topoisomerase IIa and as these drugs are increasingly offered to older (TOP2A), a direct cellular target of anthracyclines. patients (Eifel et al., 2001; Minotti et al., 2004). Two of Despite an attractive functional link between TOP2A the most commonly used anthracyclines are epirubicin and anthracyclines, published studies have failed to show and doxorubicin. Epirubicin is the 40-epimer of doxo- strong evidence of an interaction between TOP2A genetic rubicin, with a different spatial orientation of the aberrations and anthracycline response. A number of hydroxyl group at the C-40 position of doxorubicin other biomarkers have also been assessed for their role (reviewed in Cersosimo and Hong, 1986). This may in predicting anthracycline response, including TP53 contribute to the faster elimination rate and reduced (tumour protein 53) and BRCA1 (breast cancer 1, early toxicity of epirubicin, making it preferential, in some onset), together with an increasing emergence of gene chemotherapy regimens, to doxorubicin. expression profiling to produce predictive signatures of Anthracyclines are thought to exert their actions by response. Moreover, recent evidence has emerged from intercalation with DNA, generation of free radicals and presentations suggesting new candidate markers of crosslinking DNA to proteins (Minotti et al., 2004). response that warrant further investigation: Chr17CEP However, there remains insufficient data, even now, to duplication and tissue inhibitor of metalloproteases 1. This link their mode of action to one or more of the observed review will discuss research into HER2 and TOP2A as actions of anthracycline exposure. Intercalation of predictive markers of anthracycline response and will anthracyclines into DNA is thought to reduce the focus on current research into other possible candidate synthesis of both DNA and RNA (Minotti et al., predictive markers. 2004). Anthracyclines also catalyse the production of Oncogene (2010) 29, 5231–5240; doi:10.1038/onc.2010.286; formaldehyde through induction of cellular oxidative published online 2 August 2010 stress. Formaldehyde then acts to mediate covalent bonding of the anthracyclines to DNA (Taatjes et al., Keywords: HER2; TOP2A; anthracyclines; breast cancer 1999), causing both production of DNA adducts and covalently linking some proteins (including the DNA gyrase topoisomerase IIa (TOP2A)) to DNA-stabilizing double-strand breaks (Taatjes et al., 1999; Spencer et al., Anthracycline therapy 2003). As cells enter mitosis, the presence of these stabilized double-strand breaks is thought to trigger Anthracyclines are highly effective chemotherapeutic either mitotic catastrophe or apoptosis. In vivo associa- agents used in treating a broad range of malignancies tions between increased DNA damage and expression of targets such as TOP2A have been cited to support the Correspondence: Professor JMS Bartlett, Endocrine Cancer Group, proposed predictive role of TOP2A for anthracycline Edinburgh Cancer Research Centre, Western General Hospital, Crewe benefit in early breast cancer. A further key by-product Road South, Edinburgh EH4 2XR, UK. E-mail: [email protected] of the production of formaldehyde is an increased Received 30 March 2010; revised 11 June 2010; accepted 13 June 2010; intracellular concentration of free radicals; these may published online 2 August 2010 facilitate the actions of anthracyclines discussed briefly Targeting anthracyclines in early breast cancer AF Munro et al 5232 above or may, independently of these actions, alter tion, was associated with anthracycline sensitivity (Klijn membrane function or inhibit key intracellular processes et al., 1993). Evidence from published neoadjuvant (Minotti et al., 2004). With such broad-spectrum studies suggests that tumours with high grade or high activities, it might at first seem surprising that so much proliferative indices (measured by Ki67 and so on) effort has focussed on single gene alterations as respond preferentially to anthracyclines (Vincent- predictive biomarkers of anthracycline response. Salomon et al., 2004). However, data from large Genes that mediate cytoprotection from doxorubicin adjuvant trials, such as NEAT, do not show a marked belong to multiple pathways, including DNA repair, relationship between grade and benefit from the addi- RNA metabolism, chromatin remodelling, amino acid tion of anthracyclines to chemotherapy. More recent metabolism and heat shock response. Factors such as analyses (Bartlett et al., 2010) also fail to identify p53 and modulators of DNA repair and apoptosis have proliferation as a biomarker selecting patients for also been linked to anthracycline sensitivity (Fedier enhanced benefit from anthracycline-containing ther- et al., 2003; Small et al., 2003; Di Leo et al., 2007). apy. It would seem that simple measures of proliferation are unlikely to identify patient subgroups with enhanced benefit from adjuvant anthracycline therapy, although Anthracyclines in early breast cancer they may still indicate tumours with more general chemosensitivity. Two key adjuvant breast cancer trials that highlight the benefit of anthracycline containing polychemotherapy are the National Cancer Institute of Canada Clinical Trials Group Mammary 5 (MA.5) study and the HER2 and proliferation: mechanism or opportunism? National Epirubicin Adjuvant Trial (NEAT/BR9601) (Levine et al., 1998, 2005; Poole et al., 2006). Both the Given the suggested link between proliferation and MA.5 study, comparing cyclophosphamide–epirubicin– polychemotherapy response in general, and anthra- fluorouracil (CEF) with cyclophosphamide–methotrexate– cyclines in particular, the association between HER2 fluorouracil (CMF), and NEAT/BR9601, comparing and proliferation explains the early interest in HER2 as epirubicin (E-CMF) with CMF, reported a gain in overall a predictive biomarker of anthracycline response. HER2 survival of 7% after 5 years (Levine et al., 1998; Poole represents one of the most extensively studied biomar- et al., 2006) for patients treated with anthracyclines. kers in breast cancer. Overexpression and amplification Updated survival for the MA.5 study has shown a of HER2 is observed in approximately 20–25% of early persistent gain in overall survival of 4% after 10 years breast cancers eligible for chemotherapy (Slamon et al., (Levine et al., 2005). Although these trials have shown a 1989; Bartlett et al., 2001) and is commonly associated definite survival gain in patients treated with anthra- with increased tumour aggressiveness, increased risk of cyclines, the benefit, relative to the then conventional recurrence and reduced survival. HER2 amplification/ polychemotherapy (CMF), is modest and is accompanied overexpression, together with other HER family mem- by increased toxicity. In particular, outcomes in the bers, have been linked to increased proliferation in vivo control arms of these trials show that the majority of (Borg et al., 1991; Tovey et al., 2004). patients (470%) derive adequate disease control from non-anthracycline-containing therapies, challenging the need to treat ‘all comers’ with anthracyclines. This HER2 as a predictive biomarker for selection observation has driven attempts to identify those patient of dose-dense or high-dose chemotherapy subgroups for whom anthracyclines provide benefit ( for example, see Hutchins et al., 2005). Selection of patients HER2 was first associated with response to anthra- most likely to benefit from anthracycline therapy would cyclines in retrospective studies of trials assessing the also allow better use of current cytotoxics and reduce the benefit of anthracycline dose escalation or dose acce- risk of patients receiving toxicity with little or no efficacy. leration in which all patients received anthracyclines In addition, identifying biomarkers that can accurately with differing doses/schedules. The earliest and largest predict benefit from anthracyclines will highlight key of these studies, comprising almost 2000 patients, from resistance/susceptibility pathways that can then be the CALGB (Cancer and Leukemia Group B) 8541 exploited clinically to further increase efficacy. chemotherapy trial (Muss et al., 1994) suggested that women with tumours showing high HER2 expression (defined as tumours with X50% HER2-overexpressing Proliferation and tumour grade as biomarkers cells) experienced an increased benefit with increasing for chemotherapy/anthracycline therapy dose of adjuvant anthracyclines in terms of relapse-free survival when compared with women with low HER2- Early studies of anthracyclines suggested, as with many expressing tumours. Similarly, in the GONO-MIG-1 chemotherapy agents, that they are particularly effective (Del Mastro et al., 2005) study, the HER2-positive against actively proliferating cells (Klijn et al., 1993). subpopulation showed increased event-free survival and Over a decade ago, Pegram et al. (1997) showed overall survival when treated with dose-dense anthra- evidence suggesting that proliferation, but not human cyclines. Indeed, in studies of dose-dense chemotherapy, epidermal growth factor receptor 2 (HER2) amplifica- with large patient numbers, the interaction between

Oncogene Targeting anthracyclines in early breast cancer AF Munro et al 5233 HER2 overexpression and benefit from dose-dense significant evidence that in patients with HER2-positive scheduling of anthracyclines seems relatively robust, cancers there was greater benefit of anthracycline although none of the later trials evaluated similar size therapy when compared with non-anthracycline therapy populations to those studied in the CALGB trial and (with a significant ‘treatment by marker interaction’) no prospective validation or centralized meta-analysis (Pritchard et al., 2006). Although many previous trials has been performed. However, as none of these trials show a trend towards improved outcomes in the HER2- compared anthracyclines with a non-anthracycline positive population treated with anthracyclines (Paik control arm, it is difficult to directly extrapolate et al., 1998, 2000; Di Leo et al., 2001, 2002, 2004), these conclusions relating to HER2 as a predictive marker analyses did not show a statistically robust interaction for selecting patients specifically for higher-dose anthra- between HER2 and sensitivity to anthracyclines. Few of cyclines as opposed to higher overall doses of chemo- these studies were individually powered to detect such therapy. Such trials, when considering relationships an interaction, and hence many have interpreted this between biomarkers and outcome, provide weaker discrepancy as a factor of small sample sizes, suggesting evidence in favour of predictability of benefit from a that the analysis of larger, statistically powered studies specific agent (or class of agents) than studies that were required. In contrast, the earliest and the largest compare regimens with different chemotherapeutic study to date (Paik et al., 1998) failed to show a clear agents. It can be argued that it is logical to conclude treatment by marker interaction for HER2 in the from these studies that HER2 may indeed identify cases context of anthracyclines. Furthermore, a separate, requiring more aggressive treatment (with dose escala- prospectively powered, analysis of the UK NEAT/ tion), but there is only weak evidence that this effect is BR9601 (National Epirubicin Adjuvant Trial/BR9601 solely related to the anthracycline aspect of the trial) (Bartlett et al., 2010) showed no predictive effect of treatment regimen. HER2, and a meta-analysis of the Belgian (Di Leo), Recent models of tumour response to cyclical Canadian (MA.5), United Kingdom (NEAT), Scottish exposure to chemotherapy suggest that the benefit of (BR9601) and Danish (DBCG 89D) phase III anthra- multiple cycles of chemotherapy is partially offset by the cycline trials showed only borderline significant effects potential for tumour regrowth in the resting period for HER2 in selecting patients who benefit from between cycles (Norton, 2008). This model would treatment with anthracyclines with respect to recur- suggest that tumours with increased proliferation rates rence-free survival and no significant effect on overall would show increased risk of relapse after conventional survival (Di Leo et al., 2009; Bartlett et al., 2010). Both chemotherapy and benefit preferentially from accelera- analyses were independently performed and both were tion. Widespread evidence that proliferation is a key sufficiently powered to test the hypothesis that there was marker of chemotherapy ‘failure’ supports this hypo- significantly greater benefit from the addition of thesis (Tordai et al., 2008; Schmidt et al., 2009). This anthracyclines in patients whose tumours overexpressed hypothesis would suggest that the link between HER2 HER2. However, the NEAT/BR9601 trial did not amplification/overexpression and increased proliferation substantiate the effect that was suggested by the studies drives proportionally greater growth of HER2-positive of dose escalation or the modest effect observed in the tumour cells between treatment cycles than that of MA.5 trial. Results from the recent meta-analysis also HER2-negative cells. Therefore, suggesting more fre- suggest at best a modest effect of HER2 as a potential quent or ‘dose-dense’ scheduling of treatment may be predictive marker and show that a significant propor- advantageous for the treatment of HER2-positive tion of patients with HER2-negative tumours also disease, regardless of the treatment regimen. This is a benefit from anthracyclines (Di Leo et al., 2009). This hypothesis that remains to be fully validated in the study therefore also fails to fully substantiate a role of context of modern trials of accelerated chemotherapy. HER2 as a predictive biomarker to select patients likely to benefit from anthracycline therapy. Caution must be applied in interpreting different trial designs (the Belgian study included anthracycline dose HER2 as a marker of anthracycline sensitivity escalation) and when choosing methods of HER2 assessment. Furthermore, two studies on the interaction After publication of results (reviewed above) from early between HER2 and anthracycline benefit have shown anthracycline trials assessing HER2 in the context of internal inconsistencies between different detection dose escalation, further studies were performed to test methods (Di Leo et al., 2004; Pritchard et al., 2006). the hypothesis that HER2 acts as a predictive marker for Di Leo et al. (2001) tested a variety of commercially the identification of tumour response to anthracycline available HER2 antibodies highlighting discordance rather than CMF-based therapy. This concept was between methods; depending on the antibody used, the subsequently assessed in a number of trials with varying, percentage of tumours with HER2 overexpression although modest, sample sizes comparing anthracycline ranged from 8 to 18%, which clearly affects the analysis therapy to a non-anthracycline control arm with performed. Assessment using one immunohistochemis- inconclusive results (Paik et al., 1998, 2000; Di Leo try method showed a nonsignificant trend towards et al., 2001, 2002, 2004; Pritchard et al., 2008). increased survival in HER2-positive patients, whereas To date, in studies encompassing 43000 patients, another did not. Similarly, in the analysis of the MA.5 the MA.5 trial is the only trial to show statistically trial, a variety of methods such as immunohistochem-

Oncogene Targeting anthracyclines in early breast cancer AF Munro et al 5234 istry, fluorescence in situ hybridization and PCR were observed. This was a relatively small study group of used and again, despite substantial agreement between 430 patients, with 73 patients amplified for HER2 and detection methods, marked differences in the hazard 23 amplified for both HER2 and TOP2A. Therefore, ratios for the treatment interaction were apparent analysis of TOP2A aberrations in the presence of HER2 (Pritchard et al., 2006), and only for HER2 fluorescence amplifications, in a trial comparing three different in situ hybridization was it statistically significant (in the therapy regimens, resulted in the analysis being statis- absence of correction for multiple testing). tically underpowered. The SBG (Scandinavian Breast Individually, the earlier studies neither confirmed nor Group) trial 9401 analysed TOP2A in the context of refuted the role of HER2 as a marker of anthracycline HER2-amplified tumours (Tanner et al., 2006). SBG response. The main reasons for inconclusive results stem 9401 compared varying doses of tailored dose-escalating largely from studies that are not sufficiently powered FEC with standard FEC followed by CTCb (cylophos- to answer the question robustly. This, together with phamide, thiotepa, and carboplatin). Results showed a methodological inconsistencies, compromised the results significant treatment by marker interaction between of these studies. However, larger studies, including the co-amplification and high doses of FEC (hazard recent meta-analysis, also fail to show any consistent ratio 0.300, 95% confidence interval (CI) 0.11–0.83, predictive value of HER2 amplification or overexpres- P ¼ 0.020). However, with the lack of a non-anthra- sion with respect to anthracyclines (Di Leo et al., 2009; cycline control arm in the SBG 9401 trial, this response Bartlett et al., 2009a), in direct contrast to the early is only indicative of greater benefit from higher doses of studies of dose escalation. It would seem that although anthracycline-based chemotherapy, and no firm conclu- HER2 and proliferation may select for patients more sion can be made about a specific interaction with likely to respond to dose-dense therapy, there is now anthracyclines. One of the first trials to show a clear evidence that both HER2-positive and HER2- significant interaction between TOP2A and anthra- negative tumours can respond to anthracycline treat- cycline response was the DBCG 89-D trial (Knoop ment (Di Leo et al., 2009; Bartlett et al., 2009a, 2010). et al., 2005; Nielsen et al., 2008). This study analysed Therefore, on the basis of current evidence it would TOP2A aberrations in 1224 tumours with or without seem inappropriate to withhold anthracyclines from HER2 amplification. Similar to results from the Belgian HER2-negative patients. trial, the analysis showed TOP2A amplifications, but not deletions, to be predictive of outcome on anthracycline-based therapy. However, subsequent studies, including that in the NEAT/BR9601 trial, failed to TOP2A and markers of DNA damage confirm this finding (Bartlett et al., 2009a). Taken together, studies analysing the interaction between DNA damage and anthracyclines HER2/TOP2A genetic aberrations or TOP2A genetic After early inconclusive results for studies of HER2 as a aberrations have failed to either confirm or refute their biomarker, research focussed on the possibly that HER2 role as predictive markers of anthracycline response was acting as a surrogate marker for another gene. The (Knoop et al., 2005; Di Leo et al., 2009; O’Malley et al., discovery that TOP2A was in close proximity of HER2 2009; Bartlett et al., 2009a). Similar to results for HER2, on chromosome 17 gave rise to an increased interest in the recent meta-analysis of four phase III anthracycline TOP2A genetic aberrations and its overexpression with trials suggests that TOP2A amplified, deleted and non- respect to anthracycline response. As anthracyclines amplified cases may all gain greater benefit from directly target TOP2A among a range of other proteins, anthracycline treatment than CMF (Di Leo et al., this was a potentially more functionally relevant 2009; Bartlett et al., 2009a). approach. Data linking both TOP2A amplification Early in vitro studies of TOP2A gene aberrations and deletion to amplification of the HER2 gene suggested TOP2A amplification and deletion to be (Jarvinen et al., 1999; Jarvinen and Liu, 2003) gave associated with increased sensitivity and resistance to credence to this hypothesis. TOP2A inhibitors, respectively (Jarvinen et al., 2000b). Several studies have assessed the interaction between TOP2A may be linked to reduced protein expression TOP2A amplification or deletion and response to and this reduction in target levels could explain the anthracyclines (Di Leo et al., 2001; Knoop et al., 2005; relative resistance of TOP2A-deleted cells to anthracy- Slamon et al., 2005; O’Malley et al., 2006; Tanner et al., clines (Jarvinen et al., 2000a; Di Leo and Isola, 2003). 2006; Nielsen et al., 2008). Similar to analysis of HER2, Subsequent studies in patient samples have largely failed TOP2A has been assessed as a predictive marker of to confirm this finding, with a number of studies anthracycline response in both dose-escalation studies paradoxically reporting increased sensitivity in both and anthracycline versus non-anthracycline trials. One TOP2A amplified and deleted tumours (Knoop et al., of the first studies to assess TOP2A aberrations was a 2005; Di Leo et al., 2009; O’Malley et al., 2009). If Belgian study by Di Leo et al. (2001) comparing two correct, this challenges the mechanistic assumptions doses of anthracycline-based therapy (epirubicin and underlying current research that increased gene dosage cyclophosphamide) with CMF. Results suggested an of TOP2A (observed in amplified cases) is linked to advantage of both doses of epirubicin and cyclophos- increased response to anthracyclines. phamide over CMF in TOP2A-amplified patients but no In addition, a recent study has challenged the significant treatment by marker interactions were incidence of HER2 and TOP2A co-amplification

Oncogene Targeting anthracyclines in early breast cancer AF Munro et al 5235 reported by conventional fluorescence in situ hybridiza- patients receiving standard chemotherapy in the absence tion analysis (McArthur et al., 2009). Analysis of HER2 of trastuzumab, and who had polysomy of chromosome and TOP2A using representational oligonucleotide 17, benefited more than those with normal levels of microarray analysis suggests that TOP2A amplification chromosome 17. is a rare event in HER2-amplified tumours. Other This was recently confirmed in an analysis of BR9601, studies have recently reported that TOP2A amplification NEAT and MA.5 (Bartlett et al., 2009a, 2009b, 2010). is not, as previously suggested, strictly related to HER2 Analysis of Chr17CEP duplication in BR9601, NEAT amplification (Bhargava et al., 2005; Todorovic-Rakovic and MA.5 suggests chromosome 17 to be an indepen- et al., 2009). This contrasts with results from multiple dent predictive marker of anthracycline sensitivity. fluorescence in situ hybridization-based analyses (see Therefore, Chr17CEP duplication provides the first above). Further exploration of the true amplification rate unifying hypothesis of anthracycline response within for TOP2A in early breast cancer is warranted. In three trials for which the HER2 data are opposing (Di addition, the results of the Belgian TOP trial may provide Leo et al., 2004, 2009; Knoop et al., 2005; Pritchard additional insights, as they treated a series of patients with et al., 2006; Reinholz et al., 2007; Bartlett et al., 2008, large operable and locally advanced breast cancers with 2009a, b, 2010). Analysis of Chr17CEP in tumours epirubicin monotherapy preoperatively to test the hypo- recruited into the Belgian trial is near completion, which thesis that TOP2A could be a predictor of anthracycline would allow a meta-analysis of chromosome 17 in 2975 sensitivity, although the trial was stopped before the patients. The combined analysis of BR9601, NEAT, required patient numbers were reached and therefore may MA.5 and the Belgian trial is keenly awaited and has the be underpowered to address key questions. potential to determine whether Chr17CEP is a common In summary, evidence of a clear association between identifier of patients who may derive benefit from TOP2A and anthracycline response is not yet forth- anthracyclines in early breast cancer. If so, novel coming. Researchers may need to analyse further afield candidate genes related to Chr17CEP duplication may to identify pathways within breast cancer cells that are supply a mechanistic explanation of anthracycline associated with response to anthracyclines, and/or other benefit. Further validation in other tumour types, links to chromosome 17, on which both the TOP2A and including ovarian, gastric and lymphoma, may provide HER2 genes are located. additional supporting evidence.

Novel markers and DNA repair With the suggestion that both HER2 and TOP2A could Mechanisms, causes and consequences of polysomy be acting as surrogate markers, the question that remains is whether the true marker is functionally or Duplication of Chr17CEP does not, however, immedi- genetically linked to its proposed surrogates. Data are ately identify a functional link to a specific gene pathway. emerging that suggest that the true marker of anthra- The CEP probe targets a non-coding pericentromeric cycline sensitivity may have both a functional and a a-satellite repeat on chromosome 17 that can both genetic link to HER2 and TOP2A. Recent evidence has enumerate chromosome copy number (hence CEP) and suggested polysomy of chromosome 17 (more recently identify duplication of this region within the chromosome defined as duplication of the centromeric region (CEP)) (without chromosome duplication). What then are the as a predictive marker of anthracycline sensitivity potential mechanisms linked to anthracycline treatment (Bartlett et al., 2009a, 2010). Chr17CEP duplication that are identified by duplication of this genomic region? may imply a possible disruption of cell cycle dynamics, We have identified three key events for future investiga- DNA repair mechanisms and proliferation, thereby tion: the effect of increased gene copy numbers on providing a functional link to both HER2 and TOP2A. chromosome 17, the incidence of global genomic/chro- The effect of Chr17CEP duplication and the resultant mosomal instability and disruption of the cell cycle increased gene dosage of numerous chromosome 17 pathways responsible for maintaining normal cell division. genes pivotal to breast cancer may provide an indirect genetic link to the true marker of sensitivity. Chromosome 17

Chromosome 17 centromere duplication Chromosome 17 is the second most gene-dense chromosome in the human genome, containing many genes Evidence of chromosome 17 as a potential predictive central to breast cancer and DNA repair including marker of outcome was presented at the SABCS (San HER2, TOP2A, BRCA1 (breast cancer 1, early onset) Antonio Breast Cancer Symposium) in 2007 (Reinholz and TP53 (tumour protein 53) (Fraser et al., 2003; et al., 2007). Reinholz et al. (2007) tested the hypothesis McClendon and Osheroff, 2007; Pritchard et al., 2006; that chromosome 17 may be predictive of outcome on Olivier et al., 2008; Greenberg, 2009). Furthermore, the trastuzumab in 1888 patients recruited into the HER2 þ 17q12-q21 chromosomal region that incorporates HER2 N9831 intergroup adjuvant trastuzumab phase III is a gene-rich region of genetic instability that is pivotal clinical trial. Despite chromosome 17 not being pre- to tumour progression and poor prognosis in breast dictive of trastuzumab response, results suggested that cancer. Owing to the high level of coordinated genetic

Oncogene Targeting anthracyclines in early breast cancer AF Munro et al 5236 aberrations within this region, it may prove challenging predictive of poor clinical outcome (Carter et al., 2006). to pinpoint the gene(s) responsible for driving the Carter et al. (2006) identified a signature (CIN25) response to chemotherapeutic agents. HER2 is the most involving 25 genes that frequently correlate with recognized and studied gene within the region, with aneusomy and can predict survival in numerous amplification of HER2 being associated with amplifica- independent cohorts of various cancer types. The tions and deletions of neighbouring genes including CIN25 signature was derived from the top 25 ranked GRB7, RARA, THRA1, CDC6 and TOP2A (Arriola genes of a CIN70 signature that showed similar results et al., 2008). It has recently been shown that the patterns to the smaller gene set. The CIN70 signature contained of coordinated genetic aberrations are inconsistent genes such as TPX2, TOP2A, MAD2L1, CCNB1, between HER2-amplified tumours (Arriola et al., CCNB2, CDC20, PTTG1, ZWINT1, STK6 and 2008). However, Arriola et al. (2008) only considered CDC6; the genes known to be involved in processes 19 HER2-amplified tumours and did not consider such as DNA replication, chromosome segregation and genetic alterations in HER2 non-amplified tumours. cell cycle checkpoint control. If this gene signature can Previous to the work by Arriola et al. (2008), Orsetti accurately measure the level of CIN within tumours, it et al. (2004) assessed genomic gains and losses on offers an attractive method of characterizing tumours. chromosome 17 in 22 breast cancers irrespective of The genes that have been included in CIN25 highlight HER2 amplification status. Genomic gains and losses the importance of assessing genes that are functionally varied markedly between tumours with only 18% of involved in the cell cycle to pinpoint key genes that may tumours showing no genomic alterations. Strikingly, the be important in response to anthracyclines. A defect in majority of tumours showed regions of gains (and small genes associated with the mitotic spindle checkpoint is regions of losses) within the 17q22-q24 region, with only one of the mechanisms known to give rise to abnormal seven tumours showing no alteration within this region. cell division and aneuploidy. Mutations that disrupt the The 17q22-q24 region was first identified as a region of retinoblastoma pathways are a frequent event in genetic instability in breast cancer (Kallioniemi et al., tumours (reviewed in Nevins, 2001). Disruption of the 1994) and alterations are a frequent event in aneuploid retinoblastoma pathway causes a deregulation of the breast tumours (Ried et al., 1995). However, increased E2F family of transcription factors, which are important copy numbers at 17q23 are also a frequent event in a regulators of the cell cycle, and subsequently promotes variety of tumour types, including brain, lung, ovarian genomic instability (Hernando et al., 2004). MAD2 and bladder (Andersen et al., 2002). (mitotic arrest deficient protein 2) and BubR1 are mitotic checkpoint proteins that act synergistically to inhibit activation of the anaphase-promoting complex/ cyclosome. This subsequently leads to an accumulation Chromosome 17 alone or something more global? of securin (PTTG1) and cyclin B, thereby prolonging progression of mitosis to anaphase. Recent research has The detection of aneuploidy is relatively simple; how- shown expression of MAD2, a direct target of the E2F ever, the mechanisms involved in the development of family, to be associated with increased sensitivity to aneuploidy are more complex and potentially diverse. DNA-damaging agents (Cheung et al., 2005; Du et al., Chromosomal instability (CIN) is a frequent event in 2006; Fung et al., 2006). human cancer and is defined as an increased rate of chromosome aberrations. Cell cycle regulation is important in ensuring daughter cells receive one copy of chromosomes from parent cells. As cells progress Other potential or emerging markers through the cell cycle, they are regulated by cell cycle checkpoints that prevent progression if certain require- Tissue inhibitor of metalloproteases 1 (TIMP-1) is an ments are not met. The spindle assembly checkpoint emerging tumour biomarker that has a complex role in (SAC) is essential for ensuring cells do not enter cancer (reviewed in Jiang et al., 2002; Wurtz et al., anaphase until all chromosomes are aligned along the 2008). It has been shown to be elevated in a number of mitotic plate under bipolar tension. Defects in proteins human cancers, including breast, and has been asso- associated with the SAC enables cells to prematurely ciated with a poor clinical outcome (Ree et al., 1997; enter anaphase and can result in chromosome mis- Holten-Andersen et al., 2000; Schrohl et al., 2004). segregation, CIN and aneuploidy, a functional conse- HER2 expression has been associated with the expres- quence of CIN. Therefore, it is plausible to suggest that sion of metalloproteinases and contributes to an defects in the SAC, or CIN, may be responsible for the invasive phenotype, which may be linked to TIMP-1 association between chr17CEP and anthracycline sensi- expression (Pellikainen et al., 2004). Recent studies have tivity. Studies have shown that a functional SAC is shown that elevated levels of TIMP-1 in plasma or required for mitotic cell death induced by drugs serum may give rise to endocrine resistance in metastatic targeting topoisomerase, including doxorubicin. Cells breast cancer (Lipton et al., 2008). In vitro and in vivo with a compromised SAC escape from doxorubicin- studies have also suggested that a high level of TIMP-1 induced apoptosis. protects tumour cells against chemotherapy-induced A gene signature for chromosomal instability was apoptosis (Davidsen et al., 2006; Schrohl et al., 2006). recently published in Nature Genetics that claims to be Data presented by Ejlertsen et al. (2010) at SABCS

Oncogene Targeting anthracyclines in early breast cancer AF Munro et al 5237 (2008) suggested that TIMP-1 and TOP2A collectively 2003; John, 2003; McCullough et al., 2007; Kim and predict response to anthracycline therapy. Patients from Chen, 2008; Murphy and Moynahan, 2010). the DBCG 89-D were subdivided into a responsive Another approach is the use of gene expression group (TOP2A abnormal or TIMP-1 negative) and a profiling to predict response to anthracyclines (Biganzoli nonresponsive group (TOP2A normal and TIMP-1 et al., 2007; Bonnefoi et al., 2007). The use of gene positive). Patients in the ‘responsive’ group who were expression profiling enables the simultaneous analysis of treated with an anthracycline regimen (CEF) had hazard multiple genes involved in various pathways pivotal to ratios of 0.48 (95% CI 0.34–0.69) for relapse and 0.54 breast cancer. Although it has been proven that single (95% CI 0.38–0.77) for death when compared with genes can be independent predictors of outcome on patients receiving non-anthracycline regimen CMF. varying chemotherapy regimens, the effect of multiple Patients in the ‘nonresponsive’ group had hazard ratios genes should, in theory, provide a signature of response of 1.18 (95% CI 0.87–1.60) for relapse and 1.07 (95% CI more potent that a single gene alone. Despite several 0.78–1.47) for death. Although these are interesting studies claiming to have accurately produced a pre- data, there is no functional link between TOP2A and dictive profile, it seems that they require further TIMP-1; the rationale behind combining these markers validation before being used in the clinical setting. seems unclear. However, further validation of the TOP2A and TIMP-1 profile is warranted to determine whether the combination of these markers is more Conclusion beneficial than the individual markers alone. To date, research has centred on cancer cell-specific Numerous studies have tried to elucidate the role responses to therapy. Recent evidence suggests that of potential biomarkers of anthracycline sensitivity. stromal elements and immune elements within the Despite some success, results for the majority of markers tumour may mediate response to chemotherapy tested remain inconclusive. The main reasons for (Teschendorff et al., 2007; Finak et al., 2008; Ghir- inconsistent results seem to be because of both inghelli et al., 2009; Denkert et al., 2010). No evidence methodological problems and statistical underpowering. has yet emerged that these elements of tumour biology Another reason may be that the markers that are being are specifically related to response to anthracyclines but recurrently tested are not the true markers of anthra- further evidence may emerge in the future. cycline response. Therefore, researchers must remain As discussed above, chromosome 17 harbours numer- open to new emerging biomarkers and future studies ous genes central to breast cancer, and these include must be planned more carefully to enable adequate tumour suppressors TP53 and BRCA1. TP53 is involved statistical power for the biomarker in question. Emer- in cellular processes such as DNA repair and apoptosis. ging markers such as Chr17CEP duplication and TIMP- Owing to the differing drug regimens and tumour 1 provide new areas of research for which further subtypes tested, the analysis of TP53 mutations (and validation is required. Perhaps more importantly, such p53 protein expression) and anthracycline response has novel markers may also provide the potential for critical given conflicting results (Bidard et al., 2008). Although insights into the mechanisms underpinning response/ some reports suggest that TP53 mutations confer resistance to anthracyclines. Finding the true marker of resistance to anthracyclines (Aas et al., 1996; Rahko anthracycline response is beginning to look less like a et al., 2003), others have shown mutations to increase dream, and more like a reality. With the realization of sensitivity (Bertheau et al., 2002, 2009). Similarly, this goal comes the opportunity to identify novel BRCA1 has multiple roles in breast cancer including mechanisms of drug resistance and new targets for DNA damage repair, transcriptional regulation, and therapeutic intervention in early breast cancer. regulation of cell cycle checkpoints. The majority of studies have shown that absence of BRCA1 leads to hypersensitivity to DNA-damaging agents. Patients Conflict of interest with BRCA1 mutations have shown a significant gain from treatment with anthracycline therapy (Fedier et al., The authors declare no conflict of interest.

References

Aas T, Borresen AL, Geisler S, Smith-Sorensen B, Johnsen H, Bartlett J, Munro A, Dunn J, Hiller L, Jordan S, Twelves C et al. Varhaug JE et al. (1996). Specific P53 mutations are associated with (2009a). Chromosome 17 polysomy (Ch17) as a predictor of de novo resistance to doxorubicin in breast cancer patients. Nat Med anthracycline response: emerging evidence from the UK NEAT 2: 811–814. adjuvant breast cancer trial. Cancer Res 69: 45. Andersen CL, Monni O, Wagner U, Kononen J, Barlund M, Bucher C Bartlett JM, Desmedt C, Munro A, O’Malley FP, Larsimont D, Di et al. (2002). High-throughput copy number analysis of 17q23 in Leo A et al. (2009b). Chromosome 17 polysomy: a unifying 3520 tissue specimens by fluorescence in situ hybridization to tissue hypothesis underlying benefit from adjuvant anthracyclines? Cancer microarrays. Am J Pathol 161: 73–79. Res 69: 6059. Arriola E, Marchio C, Tan DSP, Drury SC, Lambros MB, Natrajan R Bartlett JMS, Going JJ, Mallon E, Watters AD, Reeves JR, Stanton et al. (2008). Genomic analysis of the HER2//TOP2A amplicon in PD et al. (2001). Evaluating HER2 amplification and overexpres- breast cancer and breast cancer cell lines. Lab Invest 88: 491–503. sion in breast cancer. J Pathol 195: 422–428.

Oncogene Targeting anthracyclines in early breast cancer AF Munro et al 5238 Bartlett JMS, Munro AF, Cameron DA, Thomas JS, Prescott RJ, Di Leo A, Isola J, Piette F, Ejlertsen B, Pritchard KI, Bartlett JM et al. Twelves C. (2008). Type I receptor tyrosine kinase profiles identify (2009). A meta-analysis of phase III trials evaluating the predictive patients with enhanced benefit from anthracyclines in the BR9601 value of HER2 and topoisomerase II alpha in early breast cancer adjuvant breast cancer chemotherapy trial. J Clin Oncol 26: 1–9. patients treated with CMF or anthracycline-based adjuvant therapy. Bartlett JMS, Munro AF, Dunn JA, McConkey C, Jordan S, Twelves Cancer Res 69: 705. CJ et al. (2010). Predictive markers of anthracycline benefit: a Di Leo A, Isola J. (2003). Topoisomerase IIalpha as a marker prospectively planned analysis of the UK National Epirubicin predicting the efficacy of anthracyclines in breast cancer: are we at Adjuvant Trial (NEAT/BR9601). Lancet Oncol 11: 266–274. the end of the beginning? Clin Breast Cancer 4: 179–186. Bertheau P, Plassa F, Espie´M, Turpin E, de Roquancourt A, Marty Di Leo A, Larsimont D, Gabcberg D, Jarvinen T, Beauduin M, M et al. (2002). Effect of mutated TP53 on response of advanced Vindevoghel A et al. (2001). HER-2 and topo-isomerase II{alpha} breast cancers to high-dose chemotherapy. Lancet 360: 852–854. as predictive markers in a population of node-positive breast cancer Bertheau P, Turpin E, Rickman D, Espie M, de Reynies A, Feugeas JP patients randomly treat. Ann Oncol 12: 1081–1089. et al. (2009). Exquisite sensitivity of TP53 mutant and basal breast Di Leo A, Tanner M, Desmedt C, Paesmans M, Cardoso F, Durbecq cancers to a dose-dense epirubicin-cyclophosphamide regimen. V et al. (2007). p-53 gene mutations as a predictive marker in a PLoS Med 4: e90. population of advanced breast cancer patients randomly treated Bhargava R, Lal P, Chen BY. (2005). HER-2/neu and topoisomerase with doxorubicin or docetaxel in the context of a phase III clinical II alpha gene amplification and protein expression in invasive breast trial. Ann Oncol 18: 997–1003. carcinomas—chromogenic in situ hybridization and immunohisto- Du Y, Yin F, Liu C, Hu S, Wang J, Xie H et al. (2006). Depression chemical analyses. Am J Clin Pathol 123: 889–895. of MAD2 inhibits apoptosis of gastric cancer cells by upregulating Bidard FC, Matthieu MC, Chollet P, Raoefils I, Abrial C, Domont J Bcl-2 and interfering mitochondrion pathway. Biochem Biophys et al. (2008). p53 status and efficacy of primary anthracyclines/ Res Commun 345: 1092–1098. alkylating agent-based regimen according to breast cancer mole- Eifel P, Axelson JA, Costa J, Crowley J, Curran WJ, Deshler A et al. cular classes. Ann Oncol 19: 1261–1265. (2001). National Institutes of Health Consensus Development Biganzoli L, Claudino WM, Pestrin M, Pozzessere D, Di Leo A. Conference statement: adjuvant therapy for breast cancer, (2007). Selection of chemotherapeutic drugs in adjuvant programs November 1-3, 2000. J Natl Cancer Inst 93: 979–989. based on molecular profiles: where do we stand? Crit Rev Oncol/ Ejlertsen B, Jensen MB, Nielsen KV, Balslev E, Rasmussen BB, Hematol 62: 1–8. Willemoe GL et al. (2010). HER2, TOP2A, and TIMP-1 and Bonnefoi H, Potti A, Delorenzi M, Mauriac L, Campone M, Tubiana- responsiveness to adjuvant anthracycline-containing chemotherapy Hulin M et al. (2007). Validation of gene signatures that predict the in high-risk breast cancer patients. J Clin Oncol 28: 984–990. response of breast cancer to neoadjuvant chemotherapy: a substudy Fedier A, Steiner RA, Schwarz VA, Lenherr L, Haller U, Fink D. of the EORTC 10994/BIG 00-01 clinical trial. Lancet Oncol 8: (2003). The effect of loss of Brca1 on the sensitivity to anticancer 1071–1078. agents in p53-deficient cells. Int J Oncol 22: 1169–1173. Borg A, Baldetorp B, Ferno M, Killander D, Olsson H, Sigurdsson H. Finak G, Bertos N, Pepin F, Sadekova S, Souleimanova M, Zhao H (1991). ERBB2 amplification in breast cancer with a high rate of et al. (2008). Stromal gene expression predicts clinical outcome in proliferation. Oncogene 6: 137–143. breast cancer. Nat Med 14: 518–527. Carter SL, Eklund AC, Kohane IS, Harris LN, Szallasi Z. (2006). A Fraser JA, Reeves JR, Stanton PD, Black DM, Going JJ, Cooke TG signature of chromosomal instability inferred from gene expression et al. (2003). A role for BRCA1 in sporadic breast cancer. Br J profiles predicts clinical outcome in multiple human cancers. Nat Cancer 88: 1263–1270. Genet 38: 1043–1048. Fung MKL, Cheung HW, Ling MT, Cheung ALM, Wong YC, Wang Cersosimo RJ, Hong WK. (1986). Epirubicin: a review of the X. (2006). Role of MEK/ERK pathway in the MAD2-mediated pharmacology, clinical activity, and adverse effects of an adriamycin cisplatin sensitivity in testicular germ cell tumour cells. B J Cancer analogue. J Clin Oncol 4: 425–439. 95: 475–484. Cheung HW, Jin DY, Ling MT, Wong YC, Wang Q, Tsao SW et al. Ghiringhelli F, Apetoh L, Tesniere A, Aymeric L, Ma YT, Ortiz C (2005). Mitotic arrest deficient 2 expression induces chemosensitiza- et al. (2009). Activation of the NLRP3 inflammasome in dendritic tion to a DNA-damaging agent, cisplatin, in nasopharyngeal cells induces IL-1 beta-dependent adaptive immunity against carcinoma cells. Cancer Res 65: 1450–1458. tumors. Nat Med 15: 1170–1199. Cortes-Funes H, Coronado C. (2007). Role of anthracyclines in the era Greenberg RA. (2009). Recognition of DNA double strand breaks by of targeted therapy. Cardiovasc Toxicol 7: 56–60. the BRCA1 tumor suppressor network. Chromosoma 117: 305–317. Davidsen ML, Wurtz SO, Romer MU, Sorensen NM, Johansen SK, Hernando E, Nahle Z, Juan G, Diaz-Rodriguez E, Alaminos M, Christensen IJ et al. (2006). TIMP-1 gene deficiency increases Hemann M et al. (2004). Rb inactivation promotes genomic tumour cell sensitivity to chemotherapy-induced apoptosis. Br J instability by uncoupling cell cycle progression from mitotic control. Cancer 95: 1114–1120. Nature 430: 797–802. Del Mastro L, Bruzzi P, Nicolo G, Cavazzini G, Contu A, D’Amico M Holten-Andersen MN, Stephens RW, Nielsen HJ, Murphy G, et al. (2005). HER2 expression and efficacy of dose-dense Christensen IJ, Stetler-Stevenson W et al. (2000). High preoperative anthracycline-containing adjuvant chemotherapy in breast cancer plasma tissue inhibitor of metalloproteinase-1 levels are associated patients. Br J Cancer 93: 7–14. with short survival of patients with colorectal cancer. Clin Cancer Denkert C, Loibl S, Noske A, Roller M, Muller BM, Komor M et al. Res 6: 4292–4299. (2010). Tumor-associated lymphocytes as an independent predictor Hutchins LF, Green SJ, Ravdin PM, Lew D, Martino S, Abeloff M of response to neoadjuvant chemotherapy in breast cancer. J Clin et al. (2005). Randomized, controlled trial of cyclophosphamide, Oncol 28: 105–113. methotrexate, and fluorouracil versus cyclophosphamide, doxor- Di Leo A, Chan S, Paesmans M, Friedrichs K, Pinter T, Cocquyt V ubicin, and fluorouracil with and without tamoxifen for high-risk, et al. (2004). HER-2/neu as a predictive marker in a population of node-negative breast cancer: treatment results of intergroup advanced breast cancer patients randomly treated either with single- protocol INT-0102. J Clin Oncol 23: 8313–8321. agent doxorubicin or single-agent docetaxel. Breast Cancer Res Jarvinen T, Liu ET. (2003). Topoisomerase IIalpha gene (TOP2A) Treat 86: 197–206. amplification and deletion in cancer–more common than antici- Di Leo A, Gancberg D, Larsimont D, Tanner M, Jarvinen T, Rouas G pated. Cytopathology 14: 309–313. et al. (2002). HER-2 amplification and topoisomerase ii{alpha} gene Jarvinen TA, Tanner M, Barlund M, Borg A, Isola J. (1999). aberrations as predictive markers in node-positive breast cancer Characterization of topoisomerase II alpha gene amplification patients randomly treated either with an an. Clin Cancer Res 8: and deletion in breast cancer. Genes Chromosomes Cancer 26: 1107–1116. 142–150.

Oncogene Targeting anthracyclines in early breast cancer AF Munro et al 5239 Jarvinen TAH, Tanner M, Rantanen V, Barlund M, Borg A, Grenman S Nielsen KV, Ejlertsen B, Moller S, Jorgensen JT, Knoop A, Knudsen et al. (2000a). Amplification and deletion of topoisomerase II alpha H et al. (2008). The value of TOP2A gene copy number variation as associate with ErbB-2 amplification and affect sensitivity to a biomarker in breast cancer: update of DBCG trial 89D. Acta topoisomerase II inhibitor doxorubicin in breast cancer. Am J Pathol Oncol 47: 725–734. 156: 839–847. Norton L. (2008). Cancer stem cells, self-feeding, and decremented Jarvinen TAH, Tanner M, Rantanen V, Barlund M, Borg A, Grenman exponential growth: theoretical and clinical implications. Breast S et al. (2000b). Amplification and deletion of topoisomerase Cancer 29: 27–36. ii{alpha} associate with erbb-2 amplification and affect sensitivity to O’Malley FP, Chia S, Tu D, Shepherd LE, Levine MN, Bramwell VH topoisomerase ii inhibitor doxorubicin in breast can. Am J Pathol et al. (2009). Topoisomerase II alpha and responsiveness of breast 156: 839–847. cancer to adjuvant chemotherapy. J Natl Cancer Inst 101: 644–650. Jiang Y, Goldberg ID, Shi YE. (2002). Complex roles of tissue O’Malley FP, Chia S, Tu D, Shepherd LE, Levine MN, Huntsman DG inhibitors of metalloproteinases in cancer. Oncogene 21: 2245–2252. et al. (2006). Prognostic and predictive value of topoisomerase II John RG. (2003). Impact of germline BRCA1 mutations and alpha in a randomized trial comparing CMF to CEF in overexpression of p53 on prognosis and response to treatment premenopausal women with node positive breast cancer (NCIC following breast carcinoma. Cancer 97: 527–536. CTG MA.5). ASCO Meeting Abst 24: 533. Kallioniemi A, Kallioniemi OP, Piper J, Tanner M, Stokke T, Chen L Olivier M, Petitjean A, Marcel V, Petre A, Mounawar M, Plymoth A et al. (1994). Detection and mapping of amplified DNA sequences in et al. (2008). Recent advances in p53 research: an interdisciplinary breast cancer by comparative genomic hybridization. Proc Natl perspective. Cancer Gene Ther 16: 1–12. Acad Sci USA 91: 2156–2160. Orsetti B, Nugoli M, Cervera N, Lasorsa L, Chuchana P, Ursule L Kim HT, Chen JJ. (2008). New players in the BRCA1-mediated DNA et al. (2004). Genomic and expression profiling of chromosome 17 in damage responsive pathway. Mol Cells 25: 457–461. breast cancer reveals complex patterns of alterations and novel Klijn JG, Berns PM, Bontenbal M, Foekens JA. (1993). Growth candidate genes. Cancer Res 64: 6453–6460. factors: clinical implications in breast cancer. Ann New York Acad Paik S, Bryant J, Park C, Fisher B, Tan-Chiu E, Hyams D et al. (1998). Sci 698: 85–101. cerB-2 and response to doxorubicin in patients with axillary lymph Knoop AS, Knudsen H, Balslev E, Rasmussen BB, Overgaard J, node-positive, hormone receptor-negative breast cancer. J Natl Nielsen KV et al. (2005). Retrospective analysis of topoisomerase iia Cancer Inst 90: 1361–1370. amplifications and deletions as predictive markers in primary breast Paik S, Bryant J, Tan-Chiu E, Yothers G, Park C, Wickerham DL cancer patients randomly assigned to cyclophosphamide, metho- et al. (2000). HER2 and choice of adjuvant chemotherapy for trexate, and fluorouracil or cyclophosphamide, epirubicin, and invasive breast cancer: national surgical adjuvant breast and bowel fluorouracil: Danish Breast Cancer Cooperative Group. J Clin project protocol b-15. J Natl Cancer Inst 92: 1991–1998. Oncol 23: 7483–7490. Pegram MD, Finn RS, Arzoo K, Beryt M, Pietras RJ, Slamon DJ. Levine MN, Bramwell VH, Pritchard KI, Norris BD, Shepherd LE, (1997). The effect of HER-2/neu overexpression on chemother- Abu-Zahra H et al. (1998). Randomized trial of intensive cyclopho- apeutic drug sensitivity in human breast and ovarian cancer cells. sphamide, epirubicin, and fluorouracil chemotherapy compared Oncogene 15: 537–547. with cyclophosphamide, methotrexate, and fluorouracil in Pellikainen JM, Ropponen KM, Kataja VV, Kellokoski JK, Eskelinen premenopausal women with node-positive breast cancer. National MJ, Kosma VM. (2004). Expression of matrix metalloproteinase Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 16: (MMP)-2 and MMP-9 in breast cancer with special reference to 2651–2658. activator protein-2, HER2, and prognosis. Clin Cancer Res 10: Levine MN, Pritchard KI, Bramwell VHC, Shepherd LE, Tu D, Paul 7621–7628. N. (2005). Randomized trial comparing cyclophosphamide, epir- Poole CJ, Earl HM, Hiller L, Dunn JA, Bathers S, Grieve RJ et al. ubicin, and fluorouracil with cyclophosphamide, methotrexate, and (2006). Epirubicin and cyclophosphamide, methotrexate, and fluorouracil in premenopausal women with node-positive breast fluorouracil as adjuvant therapy for early breast cancer. N Engl J cancer: update of National Cancer Institute of Canada Clinical Med 355: 1851–1862. Trials Group Trial MA5. J Clin Oncol 23: 5166–5170. Pritchard KI, Messersmith H, Elavathil L, Trudeau M, O’Malley F, Lipton A, Leitzel K, Chaudri-Ross HA, Evans DB, Ali SM, Demers L Dhesy-Thind B. (2008). HER-2 and topoisomerase ii as predictors et al. (2008). Serum TIMP-1 and response to the aromatase of response to chemotherapy. J Clin Oncol 26: 736–744. inhibitor letrozole versus tamoxifen in metastatic breast cancer. Pritchard KI, Shepherd LE, O’Malley FP, Andrulis IL, Tu D, J Clin Oncol 26: 2653–2658. Bramwell VH et al. (2006). HER2 and responsiveness of breast McArthur HL, Tan LK, Patil S, Wigler M, Hudis CA, Hicks J et al. cancer to adjuvant chemotherapy. N Engl J Med 354: 2103–2111. (2009). High resolution representational oligonucleotide microarray Rahko E, Blanco G, Soini Y, Bloigu R, Jukkola AA. (2003). Mutant analysis (ROMA) suggests that TOPO2 and HER2 co-amplification TP53 gene status is associated with a poor prognosis and is uncommon in human breast cancer. Cancer Res 69: 2023. anthracycline-resistance in breast cancer patients. Eur J Cancer McClendon AK, Osheroff N. (2007). DNA topoisomerase II, 39: 447–453. genotoxicity, and cancer. Mutat Res 623: 83–97. Ree AH, Florenes VA, Berg JP, Maelandsmo GM, Nesland JM, McCullough SD, Hu Y, Li R. (2007). BRCA1 in initiation, invasion, Fodstad O. (1997). High levels of messenger RNAs for tissue and metastasis of breast cancer: a perspective from the tumor inhibitors of metalloproteinases (TIMP-1 and TIMP-2) in primary microenvironment. Cancer Metast Biol Treat 11: 31–46. breast carcinomas are associated with development of distant Minotti G, Menna P, Salvatorelli E, Cairo G, Gianni L. (2004). metastases. Clin Cancer Res 3: 1623–1628. Anthracyclines: molecular advances and pharmacologic develop- Reinholz MM, Jenkins RB, Hillman D, Lingle WL, Davidson N, ments in antitumor activity and cardiotoxicity. Pharmacol Rev 56: Martino S et al. (2007). The clinical significance of polysomy 17 in 185–229. the HER2 þ N9831 intergroup adjuvant trastuzumab trial. Breast Murphy CG, Moynahan ME. (2010). BRCA gene structure and Cancer Res Treat 106: S11. function in tumor suppression a repair-centric perspective. Cancer Ried T, Just KE, Holtgreve-Grez H, Manoir S, Speicher MR, Schrock J 16: 39–47. E et al. (1995). Comparative genomic hybridization of formalin- Muss HB, Thor AD, Berry DA, Kute T, Liu ET, Koerner F et al. fixed, paraffin-embedded breast tumors reveals different patterns of (1994). c-erbB-2 expression and response to adjuvant therapy in chromsomal gains and losses in fibroadenomas and diploid and women with node-positive early breast cancer. N Engl J Med 330: aneuploid carcinomas. Cancer Res 55: 5415–5423. 1260–1266. Schmidt M, Hengstler JG, von Torne C, Koelbl H, Gehrmann MC. Nevins JR. (2001). The Rb/E2F pathway and cancer. Human Mol Gen (2009). Coordinates in the universe of node-negative breast cancer 10: 699–703. revisited. Cancer Res 69: 2695–2698.

Oncogene Targeting anthracyclines in early breast cancer AF Munro et al 5240 Schrohl AS, Holten-Andersen MN, Peters HA, Look MP, Meijer-van Tanner M, Isola J, Wiklund T, Erikstein B, Kellokumpu-Lehtinen P, Gelder ME, Klijn JGM et al. (2004). Tumor tissue levels of tissue Malmstrom P et al. (2006). Topoisomerase IIalpha gene amplifica- inhibitor of metalloproteinase-1 as a prognostic marker in primary tion predicts favorable treatment response to tailored and dose- breast cancer. Clin Cancer Res 10: 2289–2298. escalated anthracycline-based adjuvant chemotherapy in HER-2/ Schrohl AS, Meijer-van Gelder ME, Holten-Andersen MN, Christen- neu-amplified breast cancer: Scandinavian Breast Group Trial 9401. sen IJ, Look MP, Mouridsen HT et al. (2006). Primary tumor levels J Clin Oncol 24: 2428–2436. of tissue inhibitor of metalloproteinases-1 are predictive of Teschendorff AE, Miremadi A, Pinder SE, Ellis IO, Caldas C. (2007). resistance to chemotherapy in patients with metastatic breast An immune response gene expression module identifies a good cancer. Clin Cancer Res 12: 7054–7058. prognosis subtype in estrogen receptor negative breast cancer. Slamon D, Eiermann W, Robert N. (2005). Phase III randomized trial Genome Biol 8: R157. comparing doxorubicin and cyclophosphamide followed by doc- Todorovic-Rakovic N, Neskovic-Konstantinovic Z, Nikolic-Vuko- etaxel (AC-T) with doxorubicin and cyclophosphamide followed by savljevic D. (2009). Metastatic breast cancer survival according to docetaxel and trastuzumab (AC-TH) with docetaxel, carboplatin HER2 and Topo2a gene status. Dis Markers 26: 171–180. and trastuzumab (TCH) in HER2 positive early breast cancer Tordai A, Wang J, Andre F, Liedtke C, Yan K, Sotiriou C et al. patients: BCIRG 006 study. Breast Cancer Res Treat 94: S5. (2008). Evaluation of biological pathways involved in chemotherapy Slamon DJ, Godolphin W, Jones LA, Holt JA, Wong SG, Keith DE response in breast cancer. Breast Cancer Res 10: 37–46. et al. (1989). Studies of the HER-2/neu proto-oncogene in human Tovey SM, Witton CJ, Bartlett JMS, Stanton PD, Reeves JR, Cooke breast and ovarian cancer. Science 244: 707–712. TG. (2004). Outcome and human epidermal growth factor receptor Small GW, Somasundaram S, Moore DT, Shi YY, Orlowski RZ. (HER) 1-4 status in invasive breast carcinomas with proliferation (2003). Repression of mitogen-activated protein kinase (MAPK) indices evaluated by bromodeoxyuridine labelling. Breast Cancer phosphatase-1 by anthracyclines contributes to their antiapoptotic Res 6: R246–R251. activation of p44/42-MAPK. J Pharmacol Exp Ther 307: 861–869. Vincent-Salomon A, Rousseau A, Jouve M, Beuzeboc P, Sigal-Zafrani Spencer DM, Cutts SM, Kimura K, Gray PJ, Phillips DR. (2003). B, Fryneaux P et al. (2004). Proliferation markers predictive of the Detection of labile anthracycline-DNA adducts by real-time PCR. pathological response and disease outcome of patients with breast Oncology Res 13: 479–489. carcinomas treated by anthracycline-based preoperative chemo- Taatjes DJ, Fenick DJ, Koch TH. (1999). Nuclear targeting and therapy. Eur J Cancer 40: 1502–1508. nuclear retention of anthracycline formaldehyde conjugates im- Wurtz SO, Schrohl AO, Mouridsen HT, Brunner N. (2008). TIMP-1 plicates DNA covalent bonding in the cytotoxic mechanism of as a tumor marker in breast cancer—an update. Acta Oncologica 47: anthracyclines. Chem Res Toxicol 12: 588–596. 580–590.

Oncogene