A Clinically Relevant Gene Signature in Triple

Negative and Basal-like Breast Cancer

Achim Rody; Thomas Karn; Cornelia Liedtke; Lajos Pusztai; Eugen Ruckhaeberle; Lars Hanker; Regine
Gaetje; Christine Solbach; Andre Ahr; Dirk Metzler; Marcus Schmidt; Volkmar Müller; Uwe Holtrich; Manfred
Kaufmann
Posted: 02/10/2012; Breast Cancer Res. 2011;13(5):R97 © 2011 BioMed Central, Ltd.
Abstract and Introduction
Abstract
Introduction: Current prognostic gene expression profiles for breast cancer mainly reflect proliferation status
and are most useful in ER-positive cancers. Triple negative breast cancers (TNBC) are clinically heterogeneous
and prognostic markers and biology-based therapies are needed to better treat this disease.
Methods: We assembled Affymetrix gene expression data for 579 TNBC and performed unsupervised analysis
to define metagenes that distinguish molecular subsets within TNBC. We used n = 394 cases for discovery and
n = 185 cases for validation. Sixteen metagenes emerged that identified basal-like, apocrine and claudin-low
molecular subtypes, or reflected various non-neoplastic cell populations, including immune cells, blood,
adipocytes, stroma, angiogenesis and inflammation within the cancer. The expressions of these metagenes were
correlated with survival and multivariate analysis was performed, including routine clinical and pathological
variables.
Results: Seventy-three percent of TNBC displayed basal-like molecular subtype that correlated with high
histological grade and younger age. Survival of basal-like TNBC was not different from non basal-like TNBC.
High expression of immune cell metagenes was associated with good and high expression of inflammation and
angiogenesis-related metagenes were associated with poor prognosis. A ratio of high B-cell and low IL-8
metagenes identified 32% of TNBC with good prognosis (hazard ratio (HR) 0.37, 95% CI 0.22 to 0.61; P <
0.001) and was the only significant predictor in multivariate analysis including routine clinicopathological
variables.
Conclusions: We describe a ratio of high B-cell presence and low IL-8 activity as a powerful new prognostic
marker for TNBC. Inhibition of the IL-8 pathway also represents an attractive novel therapeutic target for this
disease.
Introduction
Different molecular subtypes of breast cancer have been described.[1] The most profound effects on gene
expression profiles in breast cancer are related to estrogen (ER), and proliferation status, and to a lesser extent
to Human Epidermal Growth Factor Receptor 2 (HER2) status. Not surprisingly, molecular classification and
current prognostic signatures mainly reflect these molecular features.[2] However, substantial clinical and
molecular heterogeneity remains within current molecular subsets, particularly among ER, progesterone (PgR)
and HER2 receptor negative (that is, triple negative breast cancers, TNBC[3]). Furthermore the relationship
between clinically defined TNBC and the gene expression profile-based basal-like breast cancer subtype
(BLBC)[4] is not fully defined.[5] Some authors use these two terms synonymously given the substantial overlap
between the two definitions.[6,7] However, immunohistochemical and molecular profiling studies have shown
that only a subset of TNBC express the combination of basal cell markers (for example, CK5 and CK14) that is
required for the molecular definition of this disease.[5] The prognostic significance and therapeutic implications
of molecular heterogeneity within TNBC remains to be established. From a clinical point of view, further
understanding of TNBC is important because better prognostic markers and new treatments are needed.[8]
The goal of this analysis was to assemble all currently available TNBC gene expression datasets generated on
Affymetrix gene chips and search for molecular structures in the data to define gene expression-based subsets
within TNBC. We defined metagenes as the average expression of groups of highly co-expressed genes in the
data without considering any clinical outcome variable. These metagenes identified several molecular subsets
within TNBC, some with good prognosis even in the absence of systemic therapy. Our results also suggest
possible new therapeutic strategies for TNBC. This study represents the largest attempt to define clinically
important molecular subsets within TNBC.[9]
Materials and Methods
Discussion
It has been suggested that TNBC represent a group of several molecularly[3] and clinically[41,42] distinct disease
subtypes. We used gene expression data of a cohort of 394 TNBC to identify molecular subsets within this
tumor type. The definition of TNBC was based on gene expression data which is not the standard definition
used in the clinic. This might be a caveat but holds the promise that samples erroneously characterized as
receptor-negative by immunohistochemistry do not introduce noise into our analysis. We identified 16
metagenes associated with several distinct biological processes that showed variable expression across TNBC
(Table 2). Some of the metagenes seem to point to the distinct origins of these cancers.[43,44] These include the
basal-like,[4] the apocrine,[18,19] and the claudin-low[28,29] subtypes of TNBC. Other metagenes were related to
non-neoplastic cellular constituents of the tumor microenvironment including stroma,[26,27] blood cell[30] and
adipocytes,[4] as well as signatures for angiogenesis[23,34] and inflammation.[31–33] Five metagenes appear to reflect
the variable presence of immune cells and may contribute to the clinical behavior of the cancer[4,20–25,27,45] (Table
2).
We used an unsupervised class discovery approach to first identify the main molecular subtypes within the data
and then assess the prognostic differences between the molecular subsets. Interestingly, when we performed an
independent supervised analysis that compared TNBC cases with or without recurrence, we also identified IL-8
as the top ranked gene associated with poor prognosis (Additional file 1, Supplementary Figure S13 and
Additional file 2, Supplementary Table S8). However, gene signatures obtained through supervised analysis
were not superior to the molecular structure based prognostic predictions in validation (Additional file 1,
Supplementary Figure S14). In addition, the biological interpretation of the empirically derived prognostic
signature is more difficult than the interpretation of metagenes. In summary, we performed the largest
unsupervised analysis of pooled gene expression data from TNBC. We describe a new prognostic signature for
these cancers that identify about one-third of TNBC as relatively low risk for recurrence. These cancers are
characterized by high B-cell and low IL-8 metagene expression and have about 84% recurrence-free survival at
five-years. Whereas, this may not be sufficiently high to forego adjuvant chemotherapy, these observations pave
the way to develop a clinically useful multivariate prognostic model for TNBC. A combined, prognostic score,
including clinical variables, such as nodal status and perhaps tumor size, and molecular variables, such as
optimized B-cell and IL-8 metagenes (measured by an RT-PCR or array-based method), may identify patients
with very low risk of recurrence even with ER-, PgR- and HER2-negative breast cancer. Equally important, the
prognostic importance of B-cells and the negative impact of IL-8 suggest potential novel therapeutic strategies
for TNBC that can be tested in the clinic.[31,32] It could allow the selection of those patients who could profit
most from novel immune stimulating drugs like anti-CTLA-4 antibodies that have shown promise in melanoma.
[60,61]
IL8 could also directly increase the survival of breast cancer stem cells after chemotherapy,[62] which can be
blocked with IL8 directed drugs.[63] Such an effect might explain the triple negative paradox with high relapse
rates despite a good initial response to chemotherapy.
Conclusions
In the largest and most comprehensive analysis of all available gene expression data in TNBC, we first
identified structures in the molecular data without considering any clinical outcome. Subsequently, these
molecular phenotypes were correlated with survival in multivariate analysis, including routine clinical and
pathological variables. Our most important observation is that a high B-cell presence and low IL-8 activity
identifies a good prognosis group, even in the absence of systemic therapy, among TNBC. These observations
directly point to therapeutic interventions, such as the inhibition of the IL-8 pathway and activation of the
immune system in the tumor microenvironment that could benefit patients with this disease.
References

Triple-negative Breast Cancer: Disease Entity or

Title of Convenience?
Lisa Carey; Eric Winer; Giuseppe Viale; David Cameron; Luca Gianni
Posted: 01/17/2011; Nat Rev Clin Onc. 2010;7(12):683-692. © 2010 Nature Publishing Group
Abstract and Introduction
Abstract
This Review outlines the understanding and management of triple-negative breast cancer (TNBC). TNBC
shares morphological and genetic abnormalities with basal-like breast cancer (BLBC), a subgroup of breast
cancer defined by gene-expression profiling. However, TNBC and BLBC tumors are heterogeneous and overlap
is incomplete. Breast cancers found in BRCA1 mutation carriers are also frequently triple negative and basal
like. TNBC and BLBC occur most frequently in young women, especially African Americans, and tend to
exhibit aggressive, metastatic behavior. These tumors respond to conventional chemotherapy but relapse more
frequently than hormone receptor-positive, luminal subtypes and have a worse prognosis. New systemic
therapies are urgently needed as most patients with TNBC and/or BLBC relapse with distant metastases, and
hormonal therapies and HER2-targeted agents are ineffective in this group of tumors. Poly (ADP-ribose)
polymerase inhibitors, angiogenesis inhibitors, EGFR-targeted agents, and src kinase and mTOR inhibitors are
among the therapeutic agents being actively investigated in clinical trials in patients with TNBC and/or
BRCA1-associated tumors. Increased understanding of the genetic abnormalities involved in the pathogenesis
of TNBC, BLBC and BRCA1-associated tumors is opening up new therapeutic possibilities for these hard-to-
treat breast cancers.
Introduction
In April 2009, the Fondazione Michelangelo and the Fondazione IRCCS Istituto Nazionale dei Tumori
organized the ninth in a series of seminars on recent advances and controversies in breast cancer. This seminar
was designed to update practicing clinicians on 'triple-negative' breast cancer (TNBC)—that is the subgroup of
tumors that do not express clinically significant levels of the estrogen receptor (ER), progesterone receptor
(PgR) and HER2. In particular, the seminar addressed the question of whether TNBC is a distinct pathological
subtype of breast cancer or a pragmatic category for determining eligibility for clinical trials and guiding
individual patient treatment. In addition, emerging treatment strategies were discussed because patients with
TNBC are not eligible for hormonal therapies or HER2-targeted agents and, therefore, have no established
systemic therapy options other than chemotherapy.
What is TNBC?
TNBC is a diagnosis of exclusion, based on lack of ER and PgR expression by immunohistochemistry (IHC),
and of HER2 overexpression or gene amplification by IHC or in situ hybridization, respectively.[1] The
frequency of this cluster of findings therefore depends to some extent on the methods and thresholds used
(which have evolved over time), but generally 10–24% of invasive breast cancers fall into this category.[2–5]
TNBCs are typically high-grade tumors, although low-grade tumors do occur.[2,6–10] Furthermore, although most
TNBCs are of ductal (not otherwise specified) type, other rarer tumors (for example, metaplastic, medullary and
adenoid cystic tumors) may be triple negative and these may have a better prognosis than the usually poor one
of TNBCs in general.[1,7,11] As described in further detail below, molecular-profiling studies suggest that although
most TNBCs (about 70%) fall into the basal-like subtype, the remainder consist of a variety of molecular
subtypes that are biologically distinct.[12] These findings suggest that TNBC is not a single disease but includes
additional tumor types, making it a heterogeneous entity.
Apart from lacking ER, PgR and HER2 expression, TNBCs are associated with a number of other pathological
features (Table 1). TNBCs frequently express basal cytokeratins (particularly cytokeratin 5, 14 and 17) and the
EGFR HER1,[2,7–10,13–15] features that are associated with a poor prognosis in breast cancer.[2,16–18] Compared with
other tumor types, TNBCs are also more likely to express myoepithelial markers, such as caveolins (CAV) 1
and 2,[8] c-kit,[10,15] and P-cadherin,[7] and less likely to express epithelial markers, such as E-cadherin.[7] They
also have high expression of genes associated with proliferation (Ki67 and TOP2A),[2] high levels of cyclin E,
and low levels of cyclin D1.[19,20] About 50% of TNBCs have aberrant p53 as detected by immunostaining[7,10] or
mutational analysis,[13] and/or expression of the p53 homolog p63.[2,21] Many of these features are also associated
with a poor prognosis in breast cancer.[2,19,20]
[ CLOSE WINDOW ]
Table 1. Pathological features of triple-negative and basal-like breast cancers
Incidence relative to
Pathological features Frequency in TNBC Frequency in BLBC* other breast tumor
types
Triple negative (ER−, PgR−,
100% (by definition) 80–100%28,38–40,67 –
HER2−)
Basal-like by array 71%12 – –
Increased in both TNBC
Grade 3 66–91%2,7–10 58–100%9,15,17,29,38,40,55
and BLBC
Increased in both TNBC
Ductal NOS histology 74–85%7,8,10 86–91%40,55
and BLBC
Elevated mitotic count,
78% sharply
high apoptotic rate,
demarcated borders;
geographic or central tumor
46% lymphocytic Found in 61–100% of cases;15,38,40
necrosis or fibrosis, Typical features of
infiltrate; 51% central 76% high mitotic count, 51%
pushing margin of BLBC
fibrotic zone,10 necrosis, 34% pushing border17
invasion, stromal
pushing margin,7
lymphocytic response,
metaplasia7
metaplasia
Basal cytokeratins (5, 14,
42–88%2,7–10,13–15 61–100%‡,40,55 –
17)
Increased in BLBC
(expressed in
EGFR (HER1) 13–52%2,7,8,10 15–72%17,28,34,39,45,55
approximately 8–11%
of non-BLBC cases)28
Vimentin 94%15 45–100%39,40,47 Increased in BLBC
93% versus 44% in
P-cadherin 83%49 –
non-TNBC7
Increased in BLBC
(expressed in
Expression of KIT 38%10,15 2–32%17,28,32,55
approximately 11% of
non-BLBC cases)28
Increased in BLBC
compared with luminal
Nuclear accumulation of A tumors, but similar
48–70%7,8,10 29–75%15,17,22,28,34,38,39,45
p53 or TP53 gene mutation frequency to HER2-
enriched and luminal B
tumors
 More frequent Rb loss
Rb1 LOH and/or decreased than HER2-enriched
– Approximately 72%46
expression (7%) and luminal types
(0%)
Positive correlation
Increased cyclin E between cyclin E and – –
TNBCs19
Low cyclin D1 50%8 – –
Increased versus other
67% showed strong (3+) staining;46
Increased p16 – intrinsic types (<12.5%
approximately 71% positive
3+ staining)46
Expression of α-basic- Increased versus other
– Expressed in approximately 50%33
crystallin intrinsic types
Approximately 21% of BLBC
52–96% of tumors
express CAV1 (90% of
expressing CAV1 were
CAV1 expressed in metaplastic carcinomas);140 CAV2
Increased CAV1 and CAV2 basal-like140 (versus 9%
63%; CAV2 in 32%8 associated with BLBCs;141
of CAV1-negative
approximately 29% coexpress
tumors)
CAV1 and CAV2
AR positivity less
13% expressed AR
Decreased frequency of AR Low AR expression found in frequent than in luminal
(versus 73% of non-
expression BLBCs142 tumors142 and associated
TNBCs)7
with better prognosis143
*
BLBC—variously defined (that is not restricted to tumors defined on gene-expression profiling). ‡100%
because some series defined BLBCs as ER-negative or HER2-negative tumors that were positive for one or
more basal cytokeratins (5, 6, 14 or 17). Abbreviations: AR, androgen receptor; BLBC, basal-like breast cancer;
CAV, caveolin; ER, estrogen receptor; LOH, loss of heterozygosity; NOS, not otherwise specified; PgR,
progesterone receptor; TNBC, triple-negative breast cancer.
Relationship between TNBC and BLBC
In 2000, Perou et al.[22] classified breast cancers into molecularly distinct (or 'intrinsic') subtypes, based on gene-
expression profiles. Four subtypes were initially identified: normal-like (a category which remains uncertain
since in some cases this could be due to contamination of samples with normal tissue), luminal (now divided
into two groups: luminal A and luminal B),[23,24] HER2-enriched, and basal-like. More recently, the claudin-low
subtype was added to these subtypes; however, this entity requires further research.[25] The findings of Perou et
al.[22] have been confirmed and replicated in other breast cancer series (and breast cancer cell lines)[26] using both
RNA-based techniques[23–27] and IHC.[28] Whether these intrinsic subtypes reflect different cells of origin or
different differentiation pathways is the subject of some debate. However, they clearly help explain differences
in the behavior of tumors and their response to treatment despite apparent morphological similarity.[24]
As the name suggests, the BLBC subtype expresses genes normally expressed in cells of the normal, non-
luminal (basal), myoepithelial layer of the breast duct and lobular system,[29–31] including cytokeratins 5, 14 and
17, smooth muscle actin, EGFR, P-cadherin, and CAV1 and CAV2.[22,23,31–36] These basal-like features are found
in 2–20% of all invasive breast cancers.[9,16,17,22,27,28,35,37,38] Most BLBCs are triple negative; however, about 5% of
ER-positive tumors and between 6% and 35% of HER2-positive tumors show a basal-like gene-expression
profile.[27,39] Basal-like features have also been demonstrated in the in situ component of invasive BLBCs.[15]
BLBCs share many other features that are also associated with TNBCs (Table 1). Typically, they are grade 3,
ductal carcinomas with a high mitotic count, high apoptotic rate, geographic or central tumor necrosis (and/or
fibrosis), a pushing border of invasion, and a stromal lymphocytic response.[15,17,38,40] Several of these
characteristics, including hormone-receptor negativity, high grade, medullary-like histology, lymphocytic
infiltrate, TP53 mutations, EGFR expression, HER2 negativity, characteristic copy number aberrations,[41] and
X-chromosome inactivation patterns,[42] have been termed 'BRCAness' because of their similarity to tumors that
develop in BRCA1 mutation carriers.[43] BLBCs express proliferation-associated genes (for example, Ki67,
PCNA and cyclin A) at high levels,[17,32,44] and up to 75% of BLBCs show TP53 gene mutations and/or nuclear
accumulation.[22,34,38,39,43–45] Expression of p63 is also upregulated in BLBCs.[15,39] Decreased expression and loss
of heterozygosity (LOH) of the tumor-suppressor gene Rb1 has also been described in BLBCs, leading to
upregulation of the cyclin-dependent kinase inhibitor p16, and a worse prognosis than those without Rb1 LOH.
[46]
The small heat-shock protein α-basic-crystallin is expressed in about half of all BLBCs and has been
associated with poor survival;[33] overexpression of α-basic-crystallin transforms immortalized human mammary
epithelial cells in vitro, via the MEK/ERK pathway, and increases cell migration and invasion.[33]

Some tumors with a basal-like phenotype might have acquired the basal-like characteristics through a process of
transdifferentiation, called epithelial-to-mesenchymal transition (EMT).[47] This process, which also occurs
during normal embryogenesis, is marked by changes in the expression of a range of proteins, including
upregulation of the mesenchymal marker, vimentin.[48,49] EMT is seen more frequently in BLBCs[40,47] and the
resulting acquisition of myoepithelial features by epithelial cells of luminal origin is associated with increased
aggressiveness, invasive and metastatic potential, and a poor prognosis.[50,51] Alternatively, some of these
characteristics may be attributed to the claudin-low subtype, that comprises only 5–10% of breast cancers, but is
also triple negative on clinical assays for ER, PgR and HER2.[25] This subgroup is also notable for expression of
cell–cell junction proteins and for the possession of stem-cell signatures and EMT characteristics with
upregulation of mesenchymal markers.[52,53] It has been suggested that vimentin-expressing breast carcinomas
may derive from progenitor (stem) cells that have the potential to differentiate into either luminal or
myoepithelial cells.[35,44,54] This possibility is supported by the frequent coexpression of luminal and basal
cytokeratins in BLBCs in tumors of some series.[40,55]
Although most BLBCs have a ductal histology, other clinicopathologic tumor types occur, notably metaplastic
(sarcomatoid) carcinoma,[17,56] medullary,[17,35,57] and inflammatory breast cancers.[58,59] The gene-expression
profile of inflammatory breast cancer, a morphologically and clinically distinct form of aggressive breast
cancer, is particularly intriguing because all intrinsic subtypes have been described in inflammatory breast
cancer, but they also have a distinct gene-expression profile (compared with non-inflammatory breast cancers)
that is independent of the intrinsic expression profile.[58,59]
Metaplastic carcinomas are rare tumors that seem to have distinct genetic (as well as morphologic) differences
when compared with other BLBCs, including activating mutations in the PI3K pathway, increased genomic
instability, and high expression of stem-cell and EMT markers.[53] The majority (67–95%) of metaplastic
carcinomas cluster near or within the BLBC subtype,[17,55] and metaplastic features within ductal carcinomas are
associated with the BLBC subtype. Recent studies suggest that metaplastic carcinomas are generally triple
negative and often overexpress EGFR, but have different gene copy-number alterations and more frequently
have PI3K pathway aberrations compared with other BLBCs.[53,56] Metaplastic carcinomas often have features
that distinguish them from BLBC and are more similar to the claudin-low subtype, these features include gene-
expression patterns consistent with EMT, such as downregulation of cell-adhesion molecules, as well as high
expression of stem-cell markers.[53]
Other evidence of genetic heterogeneity within the BLBC subtype comes from gene-expression profiling[10] and
comparative genomic hybridization,[60] although the significance of many variants within the BLBC subtype has
yet to be determined. Heterogeneity within the BLBC subtype may account for the sometimes conflicting data
from different studies. For example, in one study, there were more changes in DNA copy number in BLBCs
than non-BLBCs,[45] but in another study, BLBCs had fewer changes in DNA copy number than non-BLBCs.[38]
However, in general, BLBCs are thought to have more widespread copy-number aberrations than non-BLBCs,
suggesting greater genomic instability.[61–64] It is possible that this instability is the underlying cause of the
heterogeneity of BLBCs, although as more subtypes are delineated some of what seems to be molecular
heterogeneity within BLBC may turn out to be as yet undetermined distinct subtypes, such as the claudin-low
group of tumors.
The terms TNBC and BLBC are frequently used interchangeably. However, a diagnosis of TNBC depends on
the lack of detection of the expression of only three genes using primarily protein-based assays (IHC), whereas
rigorous definition of BLBC depends on assessment of mRNA expression from around 500 genes.[64] Moreover,
around 10–35% of TNBCs are not basal-like according to gene-expression profiling,[8,64] and only 40–80% of
TNBCs express basal markers at the protein level.[7,9,10,13–15] Furthermore, up to 45% of BLBCs are not triple
negative (that is they express ER, PgR and/or HER2).[8,64,65,67] A recent study used the Prediction Analysis of
Microarray (PAM50) method of gene-expression profiling to examine the molecular subtypes of 142 TNBCs
from patients participating in two clinical trials.[12] This revealed that approximately 70% were BLBC; the
remainder were a variety of identifiable non-BLBC tumors.[12] These findings indicate that TNBC and BLBC
should be regarded as distinct but overlapping categories of breast cancer,[1,64,65,68] and that the use of TNBC as a
surrogate for BLBC will result in a small but important level of misclassification (Figure 1).

Figure 1. Shared features of triple-negative, basal-like and BRCA1-associated breast cancers. Abbreviations:
AR, androgen receptor, BLBC, basal-like breast cancer; ER, estrogen receptor; PgR, progesterone receptor;
TNBC, triple-negative breast cancer.
TNBC and the BRCA1 Pathway
The majority of tumors in BRCA1 mutation carriers are triple negative and show morphological and
immunohistochemical similarities to BLBCs.[27,38,69–71] So close is the association that the detection of BLBC
features may help with the selection of patients for BRCA1 screening.[72] By contrast, the frequency of basal-
marker expression in tumors from carriers of BRCA2 germline mutations does not differ significantly from
sporadic tumors[71] and the gene-expression profiles are distinct from BLBC.[73] Although most BLBCs are
sporadic and do not arise in the setting of germline BRCA1 mutation, sporadic BLBCs and breast cancers
arising in carriers of germline BRCA1 mutations share a number of common features ('BRCAness', as described
previously), which suggest that BRCA1 dysfunction may have a role in all BLBCs.[43]
Considerable effort has been expended to identify if sporadic BLBCs have BRCA1 dysfunction or deficiency
because of the integral role of BRCA1 in many cellular pathways and because the presence of BRCA1
abnormalities has therapeutic implications. BRCA1 dysfunction or deficiency results in defects in the repair of
DNA double-strand breaks by homologous recombination. This results in genomic instability,[43] and
susceptibility to chemotherapy and other agents which target DNA repair by poly (ADP-ribose) polymerase
(PARP) inhibition.[74,75] Recent data support the view that sporadic BLBCs have defective DNA repair similar to
that seen in tumors in BRCA1 mutation carriers.[76] Sporadic BLBCs can have reduced BRCA1 protein
expression,[77] which is possibly caused by reduced transcription as a result of LOH and/or BRCA1 promoter
methylation (especially in metaplastic carcinomas).[43,78,79] Alternatively, low levels of BRCA1 protein may be
caused by high levels of ID4, a negative regulator of BRCA1, which is found in BLBCs.[78] If BRCA1 itself is
not aberrantly expressed in sporadic BLBC, it is possible that abnormalities of other members of the BRCA1
pathway lead to functional BRCA1 loss.
Identification of BLBC in Clinical Practice
Gene-expression profiling is the 'gold standard' for identification of BLBCs, but this technique is not yet widely
available or suitable for large numbers of clinical specimens. A 50-gene expression-profiling assay, the Breast
Bioclassifier™ (University Genomics Inc., Saint Louis, MO, USA and ARUP Laboratories, Salt Lake City, UT,
USA), can determine the intrinsic breast cancer subtype using formalin-fixed, paraffin-embedded samples, and
is expected to be launched in 2010.[26,80] In the meantime, TNBCs will continue to be identified in routine
clinical practice, but whether additional basal markers should be evaluated remains a matter for debate. Most
studies have found that evaluation of basal markers provides additional prognostic information;[7,9,16,28] however,
there is currently no consensus on the optimal markers for defining BLBCs. One study compared different IHC
definitions, using only a limited panel of immunohistochemical markers, and found that the percentage of
tumors defined as BLBCs varied widely (from 14% to 40% for invasive ductal carcinomas and from 19% to
76% for high-grade tumors).[81] Most researchers support testing for (at least) cytokeratin 5, cytokeratin 6 and
EGFR to define a basal-like phenotype.[9,28] In addition, many restrict their definition of BLBC to tumors that
are ER negative and HER2 negative (with or without a PgR-negative phenotype). It is pertinent to note that
most clinical trials in this therapeutic area currently enroll patients with TNBC, based on the routine acquisition
of these markers and the established nature and therapeutic limitations of triple-negative disease. Evaluation of
basal markers is not usually required prospectively, although this may change in the future as more targeted
agents are evaluated and as more sophisticated methods of risk prediction become established.
Clinical Features of TNBC and BLBC
TNBCs and BLBCs occur more frequently in younger patients (<50 years old) than in those over 50 years
old[3,29,36] and generally behave aggressively. In studies conducted in the USA, TNBCs occurred 2–3 times more
frequently in African Americans (up to 47% of breast cancers in these patients) than in other racial groups.
[3,4,82,83]
By comparison, the percentage of TNBCs was 55% in West-African patients (27% BLBCs),[84] 31% in
Korean patients (15% BLBC),[55] 18% in Chinese patients,[85] 16% in UK patients[7] and Taiwanese patients,[86]
and 8% in Japanese patients.[87] Racial and geographic differences in incidence may partly account for the
relatively favorable prognosis of breast cancer in Japanese women and the unfavorable prognosis in African-
American women. However, African-American women with TNBCs are also more likely than white women to
present with larger tumors of higher grade, mitotic activity, tumor necrosis, and apoptotic index.[88] Their tumors
also have different levels of expression of several cell-cycle and apoptotic-related proteins, such as p16, p53,
cyclin E, Bcl-2 and cyclin D.[88]
TNBCs and BLBCs are more frequent in women with abdominal obesity,[89] although this did not account for
the increased risk in African Americans in one study.[83] Abdominal obesity is a risk factor for type II diabetes
and, interestingly, the antidiabetic drug metformin has been shown to inhibit proliferation and induce apoptosis
of TNBC cells in preclinical studies, in addition to inducing other beneficial biological changes in breast cancer
of various types.[90] There is also evidence that metformin may enhance the efficacy of chemotherapy in patients
with breast cancer (of unspecified molecular subtype)[91] and trials of adjuvant metformin (not restricted to
TNBC) are planned[92] or ongoing (ClinicalTrials.gov identifier: NCT00909506). Population-based studies
suggest that BLBC might have a unique risk-factor pattern, with young age at first birth, multiparity, and high
waist:hip ratio as risks, and a particularly strong protective effect from breastfeeding.[89] Oral contraceptive use
has also been associated with increased risk of TNBC in women ≤40 years of age.[93]
TNBC and BLBCs present more often as interval cancers than other subtypes do, with a relatively large primary
tumor and a low incidence of lymph-node positivity in relation to tumor size.[55,94] The tumors show distinctive
ultrasound, mammogram and MRI imaging features, consistent with their morphological characteristics.[95]
TNBC and BLBCs have a characteristic pattern of early relapse, primarily involving the viscera.[96] Certain
tissue-specific expression profiles, such as the 'lung metastasis gene-expression signature',[97] and expression of
individual markers such as the myoepithelial marker fascin,[37] are observed more frequently in BLBCs than
other tumor subtypes, although the role these have in tumor behavior is as yet uncertain. Central nervous system
metastases also frequently occur with TNBCs, which may be a manifestation of generally poor disease control
with current therapies or reflect a predilection of TNBCs for central nervous system involvement.[96]
Numerous studies have demonstrated a worse prognosis for TNBCs and BLBCs when compared with luminal
and hormone receptor-positive tumors,[9,16,17,23,24,32,35,55,67] although not all investigators have found the prognosis
to be worse than other hormone receptor-negative tumors (a notable example being the HER2-enriched subtype
before the advent of HER2-targeted treatment).[11,55,85,98] In general, patients with BLBC tend to have a higher
recurrence rate after diagnosis, a shorter disease-free interval, rapid progression from distant recurrence to
death, and shorter overall survival; despite poorer distant metastasis-free rates, local recurrence rates are only
slightly increased compared with other tumor types,[99] with the exception of luminal A tumors.[100] These
findings suggest that TNBCs and BLBCs are radiation sensitive, which is consistent with the view that basal-
like (and BRCA1-associated) tumors are responsive to DNA-damaging agents.[75]
Treatment Strategies for TNBC and BLBC
Chemotherapy
The overall poor prognosis of patients with TNBC and/or BLBC and their tendency to relapse with distant
metastases indicate a definite need for effective systemic therapies for this disease. Chemotherapy is currently
the mainstay of systemic treatment for these patients because hormonal and HER2-directed therapies are not
effective.
Despite historically poor outcomes in TNBC and/or BLBC, chemotherapy is effective, and improvements in
chemotherapy are likely to preferentially benefit this subtype of breast cancer because of rapid proliferation
rates and defects in DNA repair.[66,101] Neoadjuvant studies indicate that TNBCs and BLBCs respond well to
anthracycline-based or anthracycline and taxane-based chemotherapy.[66,102] Several studies have shown higher
rates of pathologic complete response (pCR) to such treatments in TNBC, BLBC and HER2-enriched subgroups
compared with luminal tumors.[66,102] Unfortunately, despite higher response rates, relapse rates are high in
patients who do not achieve a pCR, resulting in a worse overall survival in the BLBC and HER2-enriched
groups than in patients with luminal tumors.[102]
Responsiveness to chemotherapy may be determined differently in different intrinsic subtypes. For example,
TOP2A gene amplification might be predictive for anthracycline therapy efficacy in HER2-enriched tumors.[103]
In vitro studies also show different patterns of gene expression in response to cytotoxics in the different breast
cancer cell types. On exposure to doxorubicin and 5-fluorouracil, luminal cell lines repressed a large number of
cell cycle-regulated and proliferation-associated genes, whereas BLBC cell lines repressed genes that were
involved in differentiation.[104] These studies highlight the challenges of moving to an era in which drugs are
used rationally based on accurate, validated predictive markers.
Despite high proliferation rates and TP53 mutation rates, TNBC seems not to show increased sensitivity to
taxanes (compared with non-TNBC) in the metastatic setting. In one study in patients with advanced disease,
the response rate and time to treatment failure following paclitaxel monotherapy was not different in patients
with TNBC versus other tumor types; however, overall survival was reduced in patients with TNBC.[82]
Retrospective data suggest that regimens that include antimetabolites (for example, cyclophosphamide,
methotrexate and 5-fluorouracil [CMF]) may be more effective in TNBC than in other tumor types (although
not necessarily more effective than anthracyclines and taxanes).[8] In a neoadjuvant study, TNBCs had the
highest response rates to CMF compared with other breast cancers (65% for TNBC versus 51% for HER2-
positive tumors and 40% for luminal tumors).[105] In one adjuvant study, patients with BLBC responded better
with adjuvant CMF than without, but no such benefit was seen for patients with non-BLBCs.[37] Another study,
the MA.5 trial, showed overall superiority for adjuvant cyclophosphamide, epirubicin and 5-fluorouracil (CEF)
compared with CMF; however, CMF seemed to outperform CEF in patients with BLBC determined by IHC
profile.[106] Although interesting, these data require prospective confirmation before subtype should be used to
choose specific chemotherapy regimens.
Mounting preclinical evidence suggests that tumors with a dysfunctional BRCA1 pathway have increased
sensitivity to DNA crosslinking agents such as platinum compounds.[107] Clinical data are scant and mainly
retrospective but also suggest that BRCA1-associated breast cancers, TNBCs, and BLBCs are sensitive to
platinum agents. In one retrospective study of neoadjuvant therapy for breast cancer in BRCA1-mutation
carriers, 10 of 12 (83%) patients treated with cisplatin monotherapy had a pCR compared with <25% of the 90
patients treated with other regimens.[108] Sirohi et al.[109] studied patients treated with platinum-based
chemotherapy in the neoadjuvant, adjuvant and advanced-disease setting. Although 5-year disease-free survival
and overall survival were worse for patients with TNBCs compared with other tumor types, the neoadjuvant
response rates were higher (88% versus 51%). Response rates for patients with advanced disease were also
higher for patients with TNBCs (41% versus 31%), progression-free survival (PFS) was improved, and there
was a trend for superior survival in patients with TNBC.[109] However, in another retrospective study of patients
with metastatic disease who received taxane plus platinum chemotherapy, response rates were similar in patients
with TNBC and non-TNBC, and patients with TNBC had a shorter survival.[110] One small prospective study
showed a 64% (18 out of 28 patients) overall response rate and a modest 21% (6 of 28) pCR rate in TNBC
patients treated with neoadjuvant cisplatin monotherapy.[111]
Whether platinum compounds are more effective than conventional agents in patients with TNBCs, BLBCs
and/or BRCA1 mutations, remains to be determined but ongoing trials should answer this question within the
next few years (Supplementary Table 1 online). These include head-to-head comparative trials of a platin versus
a taxane in the neoadjuvant setting (NCT00432172) and advanced-disease setting (NCT00532727 and
NCT00321633), and a trial in which the addition of carboplatin to taxane-based therapy is evaluated
(NCT00589238). Several other nonrandomized phase II trials of platinum monotherapies and/or combinations
with other chemotherapeutic agents are ongoing (Supplementary Table 1 online).
Other cytotoxic drugs being specifically evaluated in patients with TNBC and/or BLBC include trabectedin, a
natural product isolated from sea sponges that binds to the minor groove of DNA. A large, nonrandomized
phase II trial that assessed trabectedin in three separate cohorts—patients with TNBC, patients with HER2-
positive disease, and patients who are BRCA1 or BRCA2 mutation carriers—failed to reveal trabectedin
efficacy as a single agent in TNBC, and that arm was closed.[112] Ixabepilone, an epothilone B analog that binds
to β-tubulin causing microtubule stabilization and mitotic arrest,[113] is also being studied specifically in TNBC.
Retrospective data from 443 patients with advanced TNBC treated in a randomized phase III trial showed
superior response rates and PFS in patients who received ixabepilone plus capecitabine compared with
capecitabine alone, with a trend towards improved overall survival for patients treated with the two-drug
combination.[114] On the basis of these data, two large adjuvant studies have been initiated in which ixabepilone
is compared directly with a taxane in patients with early-stage TNBC (NCT00630032 and NCT00789581). At
this time, it remains unclear whether TNBCs and BLBCs are generally sensitive to cytotoxic therapy or if there
are cytotoxic-specific approaches that will prove to have superior responses.
PARP Inhibitors
PARPs are a family of nuclear enzymes involved in the detection and repair of DNA damage.[74] PARPs are
critical enzymes in cell proliferation and are differentially upregulated in many cancers including TNBC and
BRCA1-associated and BRCA2-associated tumors.[115] Loss of BRCA1 or BRCA2 is thought to lead to
increased dependence on PARP for DNA repair, and preclinical data indicate that BRCA1 and BRCA2 defective
cells are markedly sensitive to PARP inhibitors.[116,117] In addition, PARP inhibitors enhance the effects of a range
of DNA-damaging agents, including radiation and cytotoxic therapies such as cisplatin, etoposide, and
temozolomide.[115]
In oncology, PARP inhibitors were initially investigated primarily for their chemosensitizing effects. In 2006,
data from phase I and II trials of AG-014699 (Cancer Research UK, London, UK) in combination with
temozolomide showed that PARP could be inhibited in tumor cells and peripheral blood mononuclear cells.
[118,119]
There was also evidence of encouraging antitumor activity in patients with metastatic melanoma, but
increased hematologic toxic effects necessitated a dose reduction in temozolomide when coadministered with
AG-014699.[119] A phase I/II trial is ongoing in which patients with advanced breast or ovarian cancer, known to
carry BRCA1 or BRCA2 mutations, are treated with AG-014699 monotherapy (NCT00664781).
Several other PARP inhibitors are in phase II/III development, many focusing on patients with TNBC or BRCA
mutations (Supplementary Table 2 online). The most clinical data are available for the drugs iniparib and
olaparib. No dose-limiting toxic effects were encountered with iniparib alone or in combination with cytotoxic
therapies in phase I studies,[120,121] and promising preliminary results have been reported from a randomized
phase II study in patients with TNBC, in which iniparib in combination with gemcitabine plus carboplatin
improved the clinical benefit rate, median PFS and overall survival compared with gemcitabine plus carboplatin
alone, without increasing the frequency of adverse events.[122] Iniparib is now being evaluated in a phase III trial
in combination with gemcitabine and carboplatin in patients with previously-treated metastatic TNBC
(NCT00938652; Supplementary Table 1 online).
Encouraging results have also been reported for olaparib. In a phase I clinical trial, 60 patients were treated with
olaparib; objective antitumor activity was reported only in BRCA1 or BRCA2 mutation carriers (22 of 60
patients) with ovarian, breast, or prostate cancer.[123] In phase II, olaparib was active as a single agent in
confirmed BRCA1 or BRCA2 mutation carriers with advanced refractory breast cancer, with a response rate of
41% (11 of 27 patients) observed for the 400 mg twice-daily dose (the maximum-tolerated dose) and 22% (6 of
27 patients) for the 100 mg twice-daily dose.[124] Toxic effects were mainly mild in severity (grade 1–2) and
included fatigue, nausea, vomiting and anemia.[124] Olaparib is currently being studied in TNBC and in
combination with a range of different cytotoxic and other agents; however, a trial of single-agent olaparib in
BRCA1-associated or BRCA2-associated breast and ovarian cancer as well as TNBC and sporadic ovarian
cancer did not find activity in the TNBC subset.[125] A small study of another oral PARP inhibitor, veliparib, in
combination with temozolamide in unselected breast cancer showed that activity was limited to BRCA-
mutation carriers.[126]
Angiogenesis Inhibitors
The angiogenesis inhibitor bevacizumab, which targets VEGF, is being actively investigated in patients with
TNBC following positive results from the E2100[127] and AVADO[128] trials of bevacizumab in combination with
a taxane in patients with metastatic breast cancer. The majority of patients in the E2100 trial were HER2-
negative and subgroup analysis showed that patients with ER-negative or PgR-negative disease (who were
largely triple negative) benefitted considerably from the addition of bevacizumab to paclitaxel therapy.[127]
Similarly, in the AVADO trial, the subgroup of patients with TNBC benefited from the addition of bevacizumab
to docetaxel.[128] However, at this time, there are little prospective data and it is unclear if patients with TNBC or
BLBC should be selected for treatment with antiangiogenic drugs. Preliminary data from one phase II study of
neoadjuvant cisplatin plus bevacizumab in TNBC found an overall response rate of 63%, and a 15% pCR,
which was not notably different from historical expectations.[129] Even less data are available for small-molecule
angiogenesis inhibitors. One phase II study of sunitinib found a 15% response rate in a heavily pretreated subset
of patients with TNBC.[130]
Phase II trials of bevacizumab in combination with different cytotoxic agents are ongoing in patients with
advanced disease and in the neoadjuvant setting, including a randomized phase II trial of neoadjuvant paclitaxel
(with or without carboplatin) followed by dose-dense doxorubicin and cyclophosphamide, with or without
bevacizumab (CALGB-40603; Supplementary Table 1 online). In addition, a large adjuvant trial has started, in
which patients with operable invasive TNBC are randomized to standard chemotherapy, with or without 1 year
of adjuvant bevacizumab treatment (BEATRICE; Supplementary Table 1 online).
 EGFR Inhibitors
TNBC and BLBCs express EGFR in up to 72% of cases,[2,28,34,39,55] although activating mutations are rare.[131,132]
As a result of this consistent finding, EGFR-targeted agents (small molecule and monoclonal antibodies) are
being evaluated in clinical trials in TNBC and/or BLBC. These include ongoing randomized studies of
ixabepilone with or without cetuximab (NCT00633464), and of cisplatin with or without cetuximab
(NCT00463788). Data so far suggest that EGFR inhibitors have low efficacy in patients with TNBC when used
alone but may improve the efficacy of other agents. In one small study, 12 patients with metastatic TNBC were
treated with a taxane in combination with cetuximab; nine of the 11 assessable patients responded although
three patients developed brain metastasis during treatment.[133] Another study, TBCRC 001 (Supplementary
Table 1 online), randomly assigned patients with metastatic TNBC to cetuximab either alone or with
carboplatin.[134] The cetuximab-alone arm revealed few responses to single-agent therapy (response rate 6%);
however, 13 patients (18%) responded to cetuximab plus carboplatin, and the clinical benefit rate (partial
response or stable disease for >6 months) was 31% in this pretreated population.[134] A subset analysis from
another study showed a higher response rate to irinotecan, carboplatin and cetuximab than to irinotecan and
carboplatin alone in patients with TNBC, although the addition of the EGFR inhibitor did not change the PFS.
[135]

Src Inhibitors
Activity and expression of the tyrosine kinase c-src are frequently increased in breast cancers,[136] and this is
associated with increased motility and invasiveness.[137] Preclinical data indicate that BLBCs are particularly
sensitive to inhibition of scr kinase.[138] However, the antitumor activity of the dual abl/src kinase inhibitor
dasatinib, was modest (<5% responses) when given as monotherapy to heavily pretreated patients with TNBC.
[139]
Trials of dasatinib and other dual inhibitors (bosutinib and saracatinib) alone or in combination with
chemotherapy, are ongoing, although apart from one trial of dasatinib monotherapy (NCT00817531), these are
not specifically in patients with TNBC.
Other Strategies
The range of genetic abnormalities seen in TNBCs and BLBCs (not all described in this Review) has opened the
door to other therapeutic strategies, many of which are being evaluated in clinical trials. These include agents
targeting kinases, including mTOR, androgen receptor, TGF-β, and the TRAIL receptor. It is still too early to
assess the clinical efficacy of these strategies in TNBC and BLBC.
Conclusions
TNBC is neither a single disease entity nor a title of convenience, and much more remains to be learned about
this intriguing group of diseases. TNBC comprises the majority of BLBC, an important association given our
expanding understanding of the biology of this subtype. Although conventional chemotherapy has demonstrated
efficacy in the treatment of the tumors, this remains a subtype with a poor prognosis and for which we have no
known targeted agents. The most promising avenues of investigation include targeting DNA repair with PARP
inhibitors, antiangiogenic agents, and identifying unique genetic or environmental risk factors that may provide
a basis for prevention.
Sidebar 1
Key Points
 Triple-negative breast cancer (TNBC), a subgroup that lacks expression of hormone receptors and
HER2, overlaps with basal-like breast cancer (BLBC), a subgroup that expresses cytokeratins and other non-
luminal (basal) genes
 Breast cancers occurring in patients with germline BRCA1 mutations are often triple negative and basal
like, and BRCA1 defects or deficiency may be involved in sporadic TNBC and BLBC
 Although heterogeneous, TNBCs and BLBCs typically occur in younger women, and are associated
with a range of adverse biological features including high grade, high mitotic count and p53 positivity
 Although responsive to chemotherapy, TNBCs and BLBCs tend to relapse and metastasize early and
have a worse prognosis than other tumor subtypes
 There are no specific therapies for patients with TNBC or BLBC; new treatments under investigation
include novel cytotoxics, poly (ADP-ribose) polymerase inhibitors, angiogenesis inhibitors, EGFR-targeted
agents, and src kinase inhibitors
Sidebar 2
Review Criteria
Information for this Review was obtained during preparation for the seminar held in April 2009 by reviewing
the literature (including abstracts from major oncology conferences) and by performing PubMed literature
searches. A single definitive literature search using predefined terms and start and end dates was not performed.
Owing to the wealth of data available, only English language publications were included and publications after
2000 were given priority. Further information was obtained by searching the NIH Clinical Trials database for
clinical trials involving specific agents (for example, PARP inhibitors), and for any trials involving patients with
TNBC, BLBC or BRCA-associated tumors.

Proof of concept for iniparib plus chemotherapy in metastatic triple-negative breast cancer
25.01.11
Category: Scientific News
Adding iniparib to chemotherapy improves the clinical benefit and survival

Triple-negative breast cancers have inherent defects in DNA repair, making this type a rational target for
therapy based on poly(adenosine diphosphate–ribose) polymerase (PARP) inhibition.
A group of USA researchers led by Dr Joyce O'Shaughnessy of the Baylor Charles Sammons Cancer Center in
Dallas conducted an open-label, phase II study, funded by BiPar Sciences, to compare the efficacy and safety of
gemcitabine and carboplatin with or without iniparib, a small molecule with PARP-inhibitory activity, in
patients with metastatic triple-negative breast cancer. A total of 123 patients were randomly assigned to receive
gemcitabine (1000 mg/m2) and carboplatin (at a dose equivalent to an area under the concentration–time curve
of 2) on days 1 and 8 — with or without iniparib (at a dose of 5.6 mg per kilogram of body weight) on days 1,
4, 8, and 11 — every 21 days. Primary end points were the rate of clinical benefit (the rate of objective response
and the rate of stable disease for ≥6 months) and safety. Additional end points included the rate of objective
response, progression-free survival, and overall survival.
The addition of iniparib to gemcitabine and carboplatin improved the rate of clinical benefit from 34% to 56%
(P=0.01) and the rate of overall response from 32% to 52% (P=0.02). The addition of iniparib also prolonged
the median progression-free survival from 3.6 months to 5.9 months (P=0.01) and the median overall survival
from 7.7 months to 12.3 months (P=0.01). The most frequent grade 3 or 4 adverse events in either treatment
group included neutropenia, thrombocytopenia, anemia, fatigue or asthenia, leukopenia, and increased alanine
aminotransferase level. No significant difference was seen between the two groups in the rate of adverse events.
The results largely matched those Dr O'Shaughnessy firstly presented at the San Antonio Breast Cancer
Symposium in 2009.
The study results are published in the 20 January issue of The New England Journal of Medicine and the
authors concluded that the addition of iniparib to chemotherapy improved the clinical benefit and survival of
patients with metastatic triple-negative breast cancer without significantly increased toxic effects. On the basis
of these results, a phase III trial adequately powered to evaluate overall survival and progression-free survival is
being conducted.
Despite the promise of PARP inhibition for triple-negative tumors, an accompanying editorial by Dr Lisa A
Carey and Dr Norman E Sharpless of the Lineberger Comprehensive Cancer Center at the University of North
Carolina, Chapel Hill, brought several notes of caution. The trial included only 123 patients, gemcitabine–
carboplatin regimen is unconventional, and there were imbalances at baseline in prognostically-important
characteristics favoring the iniparib group, including number of metastatic sites, the editorialists noted.
The researchers didn't indicate what proportion of women included in the trial carried a BRCA mutation.
Without such data, it can not be concluded whether the benefit from the PARP inhibitor accrued to all triple-
negative tumors equally or whether the benefit preferentially accrued to a subgroup of BRCA-deficient tumors,
with less effect in those without the deficiency, Drs Carey and Sharpless further observed in their editorial.
The fact that nearly all patients eventually had disease progression while receiving iniparib plus chemotherapy
suggests an acquired resistance to iniparib, Dr O'Shaughnessy and co-authors noted in the paper. The
editorialists cautioned that PARP1-deficient mouse models have shown diet-induced obesity and insulin
resistance, so there may be as-yet-unknown "on-target" toxic effects. They also added that the risk of secondary
cancer from DNA-repair inhibition needs to be considered carefully if these agents are used for longer periods
in healthier patients.
Despite these limitations and the possible bias introduced by investigator assessment of the endpoints, both the
editorialists and the researchers concluded that the results provide proof of concept for iniparib with the
chemotherapy combination studied in triple-negative breast cancer.

Differential Response to Neoadjuvant

Chemotherapy Among 7 Triple-Negative Breast
Cancer Molecular Subtypes
Clin. Cancer Res. 2013 Sep 23;[EPub Ahead of Print], H Masuda, KA Baggerly, Y Wang, Y Zhang, AM
Gonzalez-Angulo, F Meric-Bernstam, V Valero, BD Lehmann, JA Pietenpol, GN Hortobagyi, WF Symmans,
NT Ueno

TAKE-HOME MESSAGE
MD Anderson’s study of 146 TNBC cases with gene expression microarray results enabled these findings
relating response to neoadjuvant therapy to subtype.

ABSTRACT
Purpose: The clinical relevancy of the 7-subtype classification of triple-negative breast cancer (TNBC)
reported by Lehmann and colleagues is unknown. We investigated the clinical relevancy of TNBC
heterogeneity by determining pathologic complete response (pCR) rates after neoadjuvant chemotherapy, based
on TNBC subtypes.
Experimental Design: We revalidated the Lehmann and colleagues experiments using Affymetrix CEL files
from public datasets. We applied these methods to 146 patients with TNBC with gene expression microarrays
obtained from June 2000 to March 2010 at our institution. Of those, 130 had received standard neoadjuvant
chemotherapy and had evaluable pathologic response data. We classified the TNBC samples by subtype and
then correlated subtype and pCR status using Fisher exact test and a logistic regression model. We also assessed
survival and compared the subtypes with PAM50 intrinsic subtypes and residual cancer burden (RCB) index.
Results: TNBC subtype and pCR status were significantly associated (P = 0.04379). The basal-like 1 (BL1)
subtype had the highest pCR rate (52%); basal-like 2 (BL2) and luminal androgen receptor had the lowest (0%
and 10%, respectively). TNBC subtype was an independent predictor of pCR status (P = 0.022) by a likelihood
ratio test. The subtypes better predicted pCR status than did the PAM50 intrinsic subtypes (basal-like vs. non
basal-like).
Conclusions: Classifying TNBC by 7 subtypes predicts high versus low pCR rate. We confirm the clinical
relevancy of the 7 subtypes of TNBC. We need to prospectively validate whether the pCR rate differences
translate into long-term outcome differences. The 7-subtype classification may spur innovative personalized
medicine strategies for patients with TNBC.

PARP Inhibitors and Triple-Negative Breast Cancer

Eric P. Winer, MD; J. Dirk Iglehart, MD Posted: 11/02/2009
Triple-Negative Breast Cancer
Triple-negative breast cancer, by definition, is estrogen- and progesterone-receptor negative and HER2 "normal" or not amplified. Triple-
negative breast cancer accounts for 10%-15% of the 192,370 cases of breast cancer diagnosed in the United States each year, which means
that approximately 20,000-25,000 American women are diagnosed annually. [1,2] Outside of the United States and Western Europe, triple-
negative disease may account for a substantially higher proportion of breast cancer cases. [3,4] Moreover, triple-negative breast cancer exacts a
disproportionate toll in terms of breast cancer mortality. [1] Far more than 10%-15% of all deaths from breast cancer are attributable to triple-
negative disease.
Pathophysiology
Many triple-negative cancers share a set of common histopathologic and genetic features. Triple-negative cancers typically fall within the
basal-like cluster of breast cancers on microarray studies. [1] Indeed, over 90% of all triple-negative cancers are basal-like, and most basal-like
cancers are also triple negative. Triple-negative tumors tend to be high grade, have pushing borders, and a scant or absent in-situ component.
Triple-negative tumors are often described as interval cancers; that is, they are identified between routine screening examinations. Unlike
other forms of breast cancer, the usual hormonal risk factors do not appear to increase a woman's risk of developing triple-negative breast
cancer. These cancers are more commonly found in African American women, [1] particularly premenopausal African American women,[5] as
well as in individuals who harbor a deleterious mutation in the BRCA1 gene.[1] Triple-negative tumors are found less frequently in older
women, and they seem to be associated with obesity. [6]
It is not fully appreciated to what extent the sporadic triple-negative cancers share the molecular make-up of BRCA1-associated cancers.[7]
Some studies have suggested that there is a loss of BRCA1 function in these tumors,[8] but there is substantial ongoing debate about the role of
BRCA1 in sporadic triple-negative cancers. It is known that both BRCA1-associated cancers and sporadic triple-negative tumors share a high
degree of genomic instability, implying an impaired ability to repair DNA damage. Both BRCA1-associated cancers and sporadic triple-
negative tumors are usually high grade and commonly have p53 tumor suppressor gene mutations. [7] Many have overexpression of the
epidermal growth factor receptor and express the CK5/6 species of cytokeratin (linking them to normal duct basal cells). [8] They are also quite
sensitive to chemotherapy, perhaps more sensitive than the average breast tumor. [9-12] Nevertheless, systemic recurrences after treatment,
including adjuvant chemotherapy, can occur and can be strikingly resistant to standard chemotherapy. Because of the triple-negative nature of
the disease, treatments such as hormonal therapy and HER2-directed therapy are ineffective. New and improved therapies are urgently
needed.
PARP Inhibitors
The poly (ADP-ribose) polymerase (PARP) inhibitors appear to be among the most promising treatments under investigation for BRCA-
associated cancers and sporadic triple-negative disease. There are at least 5 PARP inhibitors that are in clinical trials, and 2 of these agents,
BSI-201 (BiPar Pharmaceuticals)[13] and olaparib (Astra Zeneca)[14] have been evaluated in phase 2 trials in women with metastatic breast
cancer.
Early Data
The results of a randomized trial of carboplatin/gemcitabine +/- BSI 201 were reported by Dr. Joyce O'Shaughnessy at the American Society
of Clinical Oncology in June 2009.[15] The study included 123 women with triple-negative metastatic breast cancer who had received 0-2 prior
chemotherapy regimens for their metastatic disease. Sixty-two women were randomized to chemotherapy alone and 61 received
chemotherapy plus BSI-201 5.6 mg/kg intravenously on days 1, 4, 8, and 11 of every 21-day cycle. The treatment groups appeared to be
reasonably well balanced from the standpoint of performance status, sites of disease, and prior therapy. The overall response rate was 16% in
the group of patients treated with chemotherapy alone and 48% among those who received chemotherapy plus BSI-201 (P = .002). There was
also a striking improvement in both progression-free survival (PFS) and overall survival (OS) in the group of patients who received BSI-201
(PFS 3.3 months vs 6.3 months, P < .0001; OS 5.7 months vs 9.2 months, P = .0005). Remarkably, the addition of BSI-201 did not appear to
be associated with additional hematologic or nonhematologic toxicity.
The results are highly promising, but the study was small, and confirmation in a phase 3 trial is critical. A phase 3 trial testing the
carboplatin/gemcitabine combination with and without BSI-201 is underway and is accruing briskly. [16] At the time of progression, the study
allows cross-over from the chemotherapy-alone arm to the combination of chemotherapy and BSI-201. While the inclusion of a cross-over
can be problematic, particularly if OS is a study endpoint, it will increase the accrual rate.
Proposed Mechanisms
How does the PARP inhibitor add to the effect of chemotherapy in this setting? PARP expression is thought to be increased both in triple-
negative breast cancer and in breast cancer cells exposed to chemotherapy. Preclinical data further suggest that the combination of BSI-201
and chemotherapy is more effective than chemotherapy alone.[15] The presumption is that carboplatin and gemcitabine lead to DNA damage,
and repair of that damage is inhibited in the presence of BSI-201. As a consequence, cytotoxicity is increased, leading to higher overall
response rates and more prolonged PFS and OS. Assuming this strategy proves to be effective in the phase 3 setting, there will be many
unanswered questions. Perhaps the most critical issue is whether the addition of a PARP inhibitor to chemotherapy will be effective for
patients with other breast cancer subtypes or, for that matter, individuals with other malignancies. While it is possible that PARP inhibitors act
in a nonspecific manner to augment the effectiveness of chemotherapy, it is likely that the benefits of PARP inhibition depend on an inherent
deficiency in DNA damage repair (such as is found in triple-negative breast cancer).
Hereditary Breast Cancers
The PARP inhibitors are of particular interest in patients with BRCA1 (many of whom also have triple-negative cancers) and BRCA2
mutations. Both BRCA1 and BRCA2 are intimately involved in the repair of DNA damage. In particular, they play an essential role in
homologous recombination. In women with BRCA-associated breast cancer, the tumor characteristically loses the wild-type allele of BRCA1
or BRCA2 and is left with a nonfunctioning BRCA1 or 2 protein. In this setting, a second pathway of DNA repair known as base excision
repair, which is dependent on the function of PARP, allows the cancer cell to recover from DNA damage. In the absence of BRCA function,
the addition of a PARP inhibitor results in what is termed "synthetic lethality". [17-19] In a previous phase 1 trial, olaparib demonstrated activity
in individuals with advanced solid tumors who had BRCA mutations.[19] In addition, olaparib was quite active in women with BRCA1 or
BRCA2 mutations and advanced ovarian cancer. [20]
In a second presentation at the American Society of Clinical Oncology, Dr. Andrew Tutt and colleagues presented the results of a phase 2
single-arm study of olaparib in patients with metastatic breast cancer who also harbored a deleterious mutation in BRCA1 or BRCA2.[21] A
total of 54 patients were included in the trial. The first 27 received oral olaparib on a continuous daily schedule at a dose of 400 mg twice
daily. The second 27 received olaparib at 100 mg twice daily. The entire patient population included 33 patients with mutations in BRCA1, 20
patients with BRCA2 mutations, and 1 patient who had a mutation in both genes. More than half of the patients had triple-negative disease
and had received a median of 3 prior chemotherapy regimens. The overall response rate was 41% among patients treated at 400 mg twice
daily and 22% in the population treated with the lower dose. The median PFS in the cohort treated at 400 mg twice daily was 5.7 (range, 4.6-
7.4) months. In general, the treatment was well tolerated, although side effects were somewhat more prominent in women treated with the
higher dose. Although the higher dose appeared to be somewhat more active, a formal comparison is not possible in this small,
nonrandomized experience. Of note, the extent of prior treatment did not influence the likelihood of obtaining a response.
Discussion
The overall results indicate that olaparib has activity in women with BRCA mutations who have metastatic breast cancer. Additional trials are
needed to characterize the treatment outcomes more fully and to determine which patients derive the greatest benefit. At this time, a major
question is whether olaparib will have activity in women with triple-negative breast cancer but without a BRCA1 mutation. Ongoing and
planned trials are combining olaparib with chemotherapy in the hope of developing treatment regimens that are more effective than olaparib
alone.
In summary, the PARP inhibitors are promising agents for the treatment of triple-negative breast cancer and breast cancer (whether triple
negative or not) that arises in women with BRCA mutations. Ongoing trials are addressing a wide variety of clinical questions in an effort to
understand better the role of the PARP inhibitors in patients with breast cancer. It seems likely, based on the preliminary evidence available to
date, that either BSI-201 and/or olaparib will ultimately obtain US Food and Drug Administration approval, but additional research is
warranted at this time. The planned research needs to answer a variety of clinical questions and, at same time, address the many unresolved
and complex translational research questions.

British Journal of Cancer (2010) 103, 249–255. doi:10.1038/sj.bjc.6605735 www.bjcancer.com

Published online 15 June 2010
Significance of E-cadherin expression in triple-negative breast cancer
S Kashiwagi1, M Yashiro1,2, T Takashima1, S Nomura1, S Noda1, H Kawajiri1, T Ishikawa1, K Wakasa3 and
K Hirakawa1
Purpose:
Triple-negative breast cancer (TNBC), a subtype of breast cancer that is oestrogen receptor (ER) negative,
progesterone receptor (PR) negative, and human epidermal growth factor receptor 2 (HER2) negative, has a
poor prognosis. Although a correlation between E-cadherin expression level and outcome has been
demonstrated among all types of breast cancer, little is known about the significance of E-cadherin expression
levels in TNBC.
Methods:
A total of 574 patients who had undergone a resection of a primary breast cancer except for invasive lobular
carcinomas were enrolled in this study. Expressions of ER, PR, HER2, and E-cadherin were assessed by
immunohistochemistry. We examined the association between TNBC and other clinicopathological variables
and evaluated the significance of the E-cadherin expression.
Results:
Among the 574 breast cancer cases, 123 (21.4%) revealed a triple-negative phenotype. Patients with TNBC
experienced more frequent lymph node metastasis (P=0.024) and a poorer prognosis (P<0.001) in comparison
with non-TNBC patients. Triple-negative breast cancer was an independent prognostic factor. Reduced levels of
E-cadherin were observed in 238 (41.5%) of the 574 breast cancer cases. E-cadherin reduction was significantly
frequent in cases of TNBC (P<0.001) and lymph node metastasis (P=0.032). Furthermore, in the 123 TNBC
cases, the prognosis of patients with an E-cadherin-negative expression was significantly worse than that of E-
cadherin-positive patients (P=0.0265), especially for those in clinical stage II (P=0.002). A multivariate logistic
regression analysis showed a reduction of the E-cadherin expression to be an independent prognostic factor
(P=0.046).