ARTICLE
Auteur(s) : J-P Thiery, X Sastre-Garau,
B Vincent-Salomon, X Sigal-Zafrani, JY Pierga, C Decraene, JP
Meyniel, E Gravier, B Asselain, Y De Rycke, P Hupe, E Barillot, S
Ajaz, M Faraldo, MA Deugnier, M Glukhova, D Medina
Institut Curie, 26, rue d’Ulm, 75248 Paris Cedex 05
The number of cancers is increasing steadily in Western countries
in relation with aging and with changes in societal behaviors. The
conventional therapies, including surgery and radiotherapy, have
benefited from major technical advances. Chemotherapy has improved
with the discovery of more efficacious molecules, adjusted
schedules of administration and better drug combinations. Targeted
therapies are already contributing to longer relapse-free survival.
These targeted therapies include hormonotherapy, kinase inhibitors
and monoclonal antibodies to specific receptors. The therapeutic
protocols are based on more stringent clinico-histopathological
criteria. However, it is well known that only a fraction of
patients will respond to these therapies. The quality of response
can be evaluated more readily in the neo-adjuvant setting than in
adjuvant therapies following surgery. It is now clear that current
stratification methods are still relatively inadequate to define
precise prognosis and to predict response to treatment. Current
strategies are therefore aimed at establishing more accurate
methods for the stratification of cancers and for tailored
therapies. However, this goal is a major challenge still facing
numerous difficulties. Advanced oncogenomic approaches offer new
tools to improve stratification and to discover predictors of
response to treatment. Over the last few years, an increasing
number of publications have shown that small groups of identifier
genes could be used to determine the molecular status of tumours.
In this review we shall discuss some of the potential value of
these new findings in breast cancers, albeit stressing some crucial
issues to be solved. We shall also discuss whether the development
of preclinical models based on transgenic mice can provide new
insights into the molecular complexities of cancer focusing on
breast cancers.
Classification of breast cancers
Breast cancers comprise very heterogeneous diseases, which are
imperfectly described by histopathological and clinical parameters.
Although ductal invasive carcinomas is the most frequently
encountered histological type, other entities such as ductal in
situ carcinomas are becoming increasingly frequent in westernised
countries due, in part, to earlier detection of the disease. Ductal
carcinomas in situ encompasses different histological types
including comedo, cribriform and papillary. They can also be
stratified as low or high nuclear grade carcinoma. Almost 80% of
high-grade in situ carcinoma are characterized by overexpression of
HER2. Invasive lobular carcinomas is also a well-defined
histological entity; the lack of expression of E-cadherin is a
hallmark of lobular cancers as a result of mutations in a large
proportion of these tumours. Invasive carcinomas also include
histological variants such as mucinous, tubular, medullary and
papillary carcinoma. Most mucinous carcinomas have good outcome.
Medullary carcinomas are poorly differentiated ductal carcinoma,
which present frequently an unexpected favorable outcome, probably
related to a high sensitivity of the tumour to chemo- or
radio-therapy. These tumours are characterized by a high frequency
of p53 mutations and by the lack of oestrogen and progesterone
receptors. Interestingly, they exhibit a luminal-basal like mixed
phenotype. Micropapillary carcinomas have an unfavorable prognosis
often giving rise to microembolisms and lymph node metastasis.
Mixed phenotypes are often encountered; for instance, a large
number of ductal invasive carcinomas contain foci of in situ
carcinomas and in some cases lobular carcinomas. There is no
definitive data permitting one to establish tumor progression
scenarios for these different entities. Atypical hyperplasia can
give rise to ductal in situ carcinoma and ductal in situ carcinoma
can develop into ductal invasive carcinoma. The classical
distinction between ductal and lobular carcinoma does not provide
faithful indication for the cell types at the origin of the
carcinoma. Tumour heterogeneity is a hallmark of all tumours,
particularly of breast carcinomas. In addition to the cellular
heterogeneity of the carcinoma cells, tumours contain a variable
proportion of stromal cells including endothelial cells,
myofibroblasts, lymphocytes and macrophages. Staging and grading is
a critical step to evaluate the wide range in extension and
dedifferentiation of breast carcinoma diseases.
Staging and grading
Staging
The stage describes the extension of the tumour locally or at a
distance from the primary site. According to the International
Union against Cancer (UICC), staging should be based on clinical
characteristics at the time of diagnosis, including the size of the
tumour (T), the status of loco-regional lymph nodes (N) and the
presence of metastasis (M). Pathological characteristics of the
tumour, determined following surgery, provide additional and more
precise definition of the T, called pT, and status of lymph nodes,
pN. Each tumour site displays particular characteristics depending
on the pattern of local invasion and metastasis. A revised
classification was proposed by the American Joint Committee on
staging system for breast cancer [1].
- – stage I corresponds to T1N0M0 tumours (invasive tumour
measuring less or equal to 2 cm) with no lymph node
involvement and no metastasis ;
- – stage II includes from T1T2 N1 (ipsilateral lymph node
involvement) M0 tumors to T3 (over 5 cm) N0 M0
tumors ;
- – stage III is even more heterogenous, comprising
T0T1N2M0 and T4N0 or T4N1 or T4N2M0, and any T N3M0. T4 is a tumour
with extension to chest wall or skin; N2 corresponds to involvement
of ipsilateral axillary lymph nodes fixed or mated or of internal
mammary nodes and N3 includes ipsilateral infraclavicular lymph
node involvement in addition to axillary lymph node
metastasis ;
stage IV tumors include any T any N and M1 (for distant
metastasis).
Grading
The grade reflects the morphology and proliferative capacity of the
primary tumour. The microscopic analysis of tumour samples has two
major objectives: first, to establish a diagnosis of cancer and
determine the type of cancer; second, to produce a histoprognostic
index, which is used to determine the type of treatment. For breast
cancer, the histoprognostic index is based on analysis of three
criteria [2]. First, architecture, measuring the degree of
differentiation, is ranked 1 to 3; the score 3 describes tumours
with less than 10% glandular structure; second, anisokaryosis
(variation in size of the nucleus) is ranked 1 to 3; and third,
proliferation is assessed by the mitotic index (number of mitoses
per 10 microscopic fields). Tumours are defined as grade 1 when the
combined score is between 3 and 6, grade 2 for 6-7 and grade 3 for
8 or 9.
Classical diagnostic and prognostic markers
There are few classical markers routinely used in breast carcinoma.
The presence of oestrogen and progesterone receptors is determined
mostly by immunohistochemical methods. Tumours are classified as
positive if at least 1-10% of the carcinoma cells are labelled [3].
Approximately 80% of breast carcinomas are oestrogen receptor (ER)
positive. Of these, about 75% demonstrate hormone responsiveness.
Level of expression of HER2 is also critical since these patients
can potentially benefit from trastuzumab-based immunotherapy, in
conjunction with a cytotoxic drug. Between 15 an 25% of breast
carcinomas are HER positive, namely, ovexpressing HER2 often as a
result of gene amplification at the HER2 locus.
Micrometastasis
Carcinoma cells can disseminate through the lymph and blood vessels
associated with the tumour bed. Lymph node involvement is an
important aspect of staging. However, the routine examination of
lymph node does not involve the search for micrometastatic
invasion. There is an increasing interest to determine as early as
possible minimal distant dissemination of carcinoma cells. The
search for micrometastatic tumour cells in lymph node is now
benefiting from the sentinel lymph node technique [4]. This
technical approach is applied principally for T1 tumours as an
alternative procedure to initial axillary lymph node dissection.
Vital dye and/or radioactive colloid are injected in the
peritumoral space prior to surgery. The first lymph node(s)
draining the tumours (sentinel lymph node) is easily detected and
excised. The identification of tumour-free sentinel lymph node
avoids routine axillary dissection, reducing thus morbidity and
cost. The prognostic value of a few carcinoma cells detected only
by immunohistochemistry is still a matter of debate. Cancer cells
have also been detected in the blood of patients. Blood-born tumour
cells are more easily detected in patients carrying relatively
large primary tumour masses. The detection of these cells may be
particularly important as a surrogate marker to predict the
response to systemic treatments. A recent study has shown that the
search for circulating carcinoma cells in blood can be facilitated
by an automated detection of cytokeratin positive cells following
immunomagnetic enrichment using an anti Ep-CAM antibody. Patients
with more than 5 circulating carcinoma cells per 7.5 ml of blood
have shorter median and overall survival. Beneficial treatment is
suggested for a fraction of patients who showed less circulating
tumour cells following initiation of chemotherapy [5]. Although the
detection of carcinoma cells in blood is much less demanding and
better accepted by patients, it is clearly less sensitive than
detection in the bone marrow [6]. Rare tumour cells have been
detected in the bone marrow medulla, the only other accessible site
for most carcinoma. Many studies have confirmed that
immunocytochemical techniques involving anticytokeratin antibodies
provide relatively reproducible results [7]. The data obtained to
date clearly show that 5-25% of patients carrying T1 tumours (≤
2 cm) already have disseminated cancer cells in the bone
marrow (frequency of 1 per 106 mononucleated cells). The presence
of micrometastatic tumour cells provides a novel independent
prognostic indicator for recurrence and survival [8]. A European
consortium has recently analysed data pooled from 4703 patients and
confirmed the crucial importance of bone marrow micrometastasis for
prognosis in multivariate analysis [9]. Bone marrow culture may in
some instance reveal the presence of micrometastatic tumors cells
not detected by direct immunocytochemistry on bone marrow samples
[10]. Methods for enrichment of micrometastatic cells are urgently
needed to improve this diagnostic since these methods are
susceptible to artefacts resulting from the capture of
non-carcinomatous Ep-CAM positive cells [11]. Very interestingly,
CGH of chromosomes from micrometastatic cells isolated in M0
patients revealed many less alterations than those obtained from M1
patients, indicating the bone marrow micrometastasis can occur at
early stages of tumour progression challenging the dogma that
micrometastatic cells are derived from the most advanced primary
tumour foci [12].
High density molecular profiling
Genomic alterations
The genome of breast carcinoma is remarkably unstable, possibly as
a result of early dysfunction of DNA replication, repair and
recombination machineries. Numerous chromosomal aberrations have
already been extensively described by classical cytogenetic
approaches. The comparative genomic hybridisation technique, and
more recently the high density arrays, have revealed an
extraordinary complexity of genomic alterations. A provisional list
established in 2003 has described the frequency of loss and gain on
each chromosome [13]. Remarkably, low level gains on chromosomes
are more frequent than losses and amplification of loci. There is a
high proportion of loci that can be affected in both directions,
losses or gain. LOH and CGH studies are not providing overlapping
results, suggesting that LOH events are not describing the
behaviour of individual genes but rather variably large regions.
Poor outcome correlates with distinct patterns of alteration as
seen by LOH and CGH studies. Amplification of specific loci already
allows one to define subgroups of breast carcinomas. Aside from the
well described gene amplification at the HER2 locus, comprising 7
genes, other loci including CCND1, MDM2, MYC and EGFR have been
characterised. A recent study using fluorescent in situ
hybridisation on tissue arrays of more than 2000 breast carcinoma
specimens showed that co-amplifications are more prevalent than
previously described [14]. For instance, almost 30% of
CCND1-amplified tumours harbour other amplicons. More strikingly
CCND1 amplification was observed in 43% of HER2 amplified tumours
and in 56% of MDM2 amplified tumours. A CGH array with selected BAC
encoding major regions of interest in breast cancer was used to
screen a limited number of advanced breast carcinoma. This study
identified relatively frequent amplicons coding for 112 candidate
genes; out of these, 44 were validated [15]. Recently, a basal-like
phenotype was found with a subset of ductal invasive breast by
high-density arrays for molecular profiling of transcripts. This
phenotype had already been identified by immunohistochemical
characterisation of cytokeratins almost 20 years ago. The CGH
analysis of microdissected tumour cells from grade 3 basal
CK14-positive and negative tumours revealed that the majority of
basal-like tumours have significantly less genomic alterations than
the CK14-negative grade 3 tumours. Hierarchical clustering
identified a subgroup which contains 40% of basal-like tumours
which had a worse prognosis than the other basal-like tumours [16].
This study exemplifies the difficulty in stratifying breast
carcinoma even in the case of a relatively well defined molecular
entity by using only cytokeratins immunocytochemistry. Refined
genomic and transcriptomic approaches can detect heterogeneity in
an otherwise fairly homogenous ER, PR and HER2 negative group.
With the advent of new high density arrays which can scan the
genome at much higher definition, such as BAC arrays with more than
30,000 clones, long oligonucleotide and SNP arrays, one can expect
to see many more alterations. These new data will require new
software for signal analysis and precise determination of affected
loci. In this respect, an algorithm was developed to detect
breakpoints and outliers, and to assign a status to each loci from
array CGH data [17]. This software has also been adapted to carry
out the same analysis on SNP data.
Promising data will emerge from studies aimed at defining tumour
evolution in breast tumours. One crucial issue is to determine to
what extent ductal in situ gives rise to ductal invasive carcinoma
and lobular in situ carcinoma leads to lobular invasive carcinoma.
A similar issue concerns local regional relapses; to what extent
are they clonally derived from the primary tumour? One pilot study
based on a 2400 BAC clone CGH array showed that a majority of
synchronous lobular in situ and lobular invasive carcinoma are
clonally related [18].
Point mutations
Recently, a major effort has been deployed to sequence gene
candidates from breast tumour lines and breast carcinoma specimens.
The data are compiled and published regularly by the Sanger centre
(cosmic database; The Sanger Institute: catalogue of somatic
mutations). The p53 protein is mutated in 20-40% of breast cancers
(see http:oewww-p53.iarc.fr/index.html). Recent studies reveal that
PI3K is mutated in more than 25% of breast cancers. The mutations
are frequently found in the catalytic site. Pioneer studies with
limited number of samples could not show correlation with
anatomoclinical data [19-22]. Studies with larger number of
patients show correlation with the oestrogen receptor, lymph node
and HER2 status [23, 24]. In addition, mutation in PI3K and the
loss of PTEN are mutually exclusive [24]. Two activating mutations
in PI3K have been shown to transform normal mammary epithelial
cells suggesting that such mutations could contribute to tumour
progression [25]. Mutations are also relatively frequent in CDKN2A.
An extensive screen has been performed recently to search for
mutations in the kinase gene superfamily. This study, carried out
in a limited number of breast carcinoma, shows that only a few
tumours accumulated mutations in a large number of kinases while
most other tumours do not carry any mutations [26].
Transcriptomics
High density RNA profiling became possible with the advent of new
technological developments, including array spotters,
radio-labelled or fluorescent nucleotides, and phosphoro-imagers or
sensitive laser-based scanners. The first studies showed the great
utility of high density molecular profiling of tumours. The first
series of breast carcinomas analysed by the Stanford group who
pioneered cDNA arrays showed that distinct patterns could be
established for individual tumours and that tumours analysed before
and after chemotherapy resembled each other more than tumours
coming from other patients. Lymph node metastasis profiles were
also more closely related to their primary tumour profiles than to
those of other tumours [27]. Subsequent studies using the same
technology revealed a new molecular taxonomy for breast cancers.
One major ER negative cluster contains HER2 positive, basal-like
and normal breast-like tumours. The ER positive cluster can be
subdivided into three distinct luminal A, B, C subtypes [28]. Most
remarkably, the newly identified basal phenotype is associated with
shorter survival times similar to the amplified HER2 group. The ER
positive luminal B and C subtypes also showed poorer prognosis than
the luminal A subtype. These findings were confirmed in another
study showing that the luminal A and B, the normal-like, basal-like
and HER2 phenotypes were found in two independent sets of data with
similar prognostic values to the previous study. Interestingly, a
large fraction of the BRCA1 tumours exhibit a basal-like phenotype
[29]. Immunohistochemical approaches can be applied with a limited
number of markers to identify about 75% of the basal-like tumours.
This study showed that a subset of basal tumours exhibited an HER1
overexpression as compared to other tumour types. This simple
approach stratified ER and HER2 negative tumours using only
cytokeratins 5/6 and 17, HER1 and c-Kit. The relative frequency of
the different subtypes in a large group of specimens was 15% for
basal-like, 23% for HER2 positive and 40% for ER-positive tumours,
22% of the tumours could not be classified [30].
Surprisingly, RNA profiling studies of premalignant in situ and
invasive carcinoma revealed similar profiles, suggesting that
global gene alteration patterns are already acquired in atypical
ductal hyperplasia. Differences were, however, found between
different stages and subtle differences were found between in situ
and invasive forms [31].
RNA profiling can also be used to search for differential gene
expression in well defined histological entities such as lobular
and ductal invasive carcinoma. It can also provide the information
for the construction of class predictors, in the so-called
supervised classification analysis. Supervised classification based
on gene expression identified a limited list of genes that can
classify accurately lobular and ductal invasive carcinoma. Some of
the genes may indicate distinct molecular pathways for local
invasion [32]. RNA profiling was also used to define poor prognosis
gene signatures. A pioneering study identified a list of 70 genes
that can predict relapse within 5 years of diagnosis in patients
with node negative T1T2 tumours less than 55 years old [33]. In a
second study, lymph node negative and positive tumours were
analysed to evaluate the predictive power of the 70-gene signature.
This gene signature was found to be more powerful than prognosis
based on anatomical/clinical conventional criteria adopted in
consensus conferences in St-Gallen or at NIH [34]. A prognostic
score also could be given by a wound-response gene expression
signature, since wound response is a biological hallmark of tumour
progression. The integration of the 70 gene signature with the
wound signature in a decision tree improved significantly the
stratification of patients at high risk of metastasis [35].
A 17-gene pan-metastatic signature was found to be shared by
different types of adenocarcinomas, possibly suggesting that the
metastatic potential is encoded in the primary tumour and not by a
small subset of carcinoma cells undergoing a Darwinian type
selection throughout progression [36]. A similar conclusion was
reached by comparing RNA profiling of a limited number of primary
and matched metastatic breast cancer tumours [37]. An extensive
study was recently carried out on a large collection containing
mostly T1T2N0 tumours from patients who had not received adjuvant
chemotherapy. A 76-gene signature was identified with good
sensitivity but moderate specificity on a validation set. A 5.5
hazard ratio was obtained in multivariate analysis as compared to
2.6 for stage II and III versus stage I, demonstrating the
potential value of this new signature. This signature shared only
three genes in common with the van’t Veer signature [38].
Most of the studies so far have used different algorithmic and
biostatistical approaches to find a group of genes whose expression
profile predicts disease progression. Some studies are based on the
use of metagenes, i.e. a group of genes behaving similarly are
first identified to construct a decision tree in a Bayesian
approach. These studies, combining clinical and genomic data, allow
to establish probability predictions of lymph node status and
recurrences with a predictive accuracy of 90% [39, 40].
Considering the formidable heterogeneity of breast tumours, it
is not surprising that multiple gene expression prognostic
signatures have been found so far. There is an advantage to
establishing breast cancer gene signatures in clinically more
homogenous cohorts such as ER and age status, two well established
prognostic parameters. The van’t Veer cohort was analysed using
these criteria and new algorithms modifying the training set to
eliminate those patients that were not correctly classified during
a cross-validation procedure led to the definition of a new 50-gene
signature [41]. This gene signature was more clearly focusing on
one pathway than previous signatures. In this set of selected
genes, overexpression of the cell cycle associated genes were
clearly identifying the poor prognosis group. It is indeed a
valuable approach to determine signatures associated with a
potentially dominating pathway.
A signature related to p53 status was recently published and
outperformed the stratification established on p53 sequence data.
The 32-gene signature was able to identify two major groups of
patients defined as p53 wild type and p53 mutated. The two groups
contain a small proportion of patient whose p53 status did not fit
with their group status. However, these misclassified tumours were
most likely correctly assigned for their p53 functionality. For
instance, tumours with low wild type p53 expression may behave like
mutated p53 tumours [42]. Organ-specific metastasis is a long
debated issue since the pioneering work of Stephen Paget. The
molecular profiling of the MDA MB 231 pleural effusion metastatic
cell line, selected for its ability to uniquely metastasise to
bone, showed that a small set of genes was associated with organ
specificity [43]. This set of genes differs from those conferring a
general poor prognosis included in the original 70-gene signature
[33]. This list of genes has been tested on a cohort of breast
carcinoma showing the possibility of identifying the tumours which
will metastasise to bone. A similar study has been reported to
define metastasis to lung [44]. These signatures need to be
validated on a much larger cohort in order to determine whether
these organ-specific signatures remain valid for metastases
occurring at multiple sites.
Weaknesses in the transcriptomic approach
The rapidly increasing number of non-overlapping lists of genes
selected for prognostic purposes and for prediction of response to
treatment by different teams has already prompted several studies
to identify the origin of these discrepancies. These issues have
been discussed in a recent review [45]. The data from seven studies
comprising lung, breast, hepatocellular carcinoma, medulloblastoma,
non-Hodgkin’s lymphoma and acute lymphocytic leukaemia were
reanalysed by creating multiple random training sets to study the
stability of the molecular signatures and the proportion of
misclassification. The genes selected for prognosis are crucially
dependent on the choice of patients included in the training set.
Clearly, the proportion of misclassified patients in the validating
step decreases when the number of patients was increased in the
training set. Most of these studies could not prove that they
perform better than random [46]. Another study [47] showed that 50
patients is clearly a minimum for a training set to achieve some
significance, but a few hundreds are required to build a clinically
useful predictor. The 70-gene list for prognosis of breast cancer
metastasis [33] was also analysed independently to evaluate its
robustness. One important finding is that many genes are correlated
with survival but the differences in their correlation coefficients
are small and the correlation fluctuates strongly when the set of
patients is even partially modified, probably because of the high
heterogeneity of the disease [48]. The conclusion from these
studies is that gene signatures derived from high density
microarrays are not unique and not necessarily easily reproducible
from one platform to another platform [49]. However it is very
likely that the main cause of this lack of robustness is linked to
tumour heterogeneity and relatively poor quality of RNA preparation
in a fraction of the samples, due in part to inadequate collection
and preservation procedures. To circumvent this major difficulty, a
very large number of high-quality samples, selected on
histoprognostic and immunohistochemical criteria, are required to
diminish heterogeneity. Laser microdissection has been utilised by
several teams for such studies; however, this approach also suffers
from a number of drawbacks including the preparation of a
reasonable quantity of high quality RNA to avoid two amplification
steps. Better methods need to be developed for RNA preparation from
formalin-fixed paraffin-embedded specimens. Multiplex PCR may
overcome these difficulties, especially for the new major clinical
trials aimed at defining the best multiparametric histological and
molecular markers for prognosis or response to treatment [50].
Prediction of response to treatment
Gene classifiers
Molecular profiling is now thought to provide indicators which will
replace or complement the standard markers such as stage, grade and
HER2 and ER status. The surrogate markers used in the neo-adjuvant
setting for pathologic complete response, in comparison to partial
response, stable disease and tumour progression, have proven useful
to establish a limited list of gene predictors. Pathological
complete response is certainly correlated with lower risk of
relapse and death, but it is in no way a perfect surrogate for
cure. In reality, the response is rather a continuum than a very
discrete entity, which renders difficult supervised analyses [51].
In one study, a 74-gene predictor was shown to identify
non-responders with an overall accuracy of 78%, but recognized only
three out of seven complete responders. However this gene set was
established using a limited number of tumours in the training
cohort, possibly not including other genes that could identify
complete responders in the validating set [52].
The quality of response to paclitaxel followed by
5-fluorouracil, doxorubicin and cyclophosphamide chemotherapy was
evaluated using the molecular stratification described above. Very
interestingly, the basal-like and HER groups responded much better
than the luminal subtypes; the normal-like type had almost no
response. Noticeably, the gene predictors for the basal subgroup
were not overlapping with those predicting response in the HER2
group strongly suggesting different mechanisms mediating response
in the two ER negative tumor types [53].
The response to docetaxel was evaluated in a limited number of
core biopsy samples from breast cancer patients undergoing
neo-adjuvant therapy. A 92-gene predictor list was able to
classify, with 90% specificity and 85% sensitivity, in a
leave-one-out validation procedure [54, 55]. A complementary study
showed that residual tumour profiling was very similar in each case
and resembled that of the initially fully resistant tumours. These
results show that some specific transduction pathways could confer
sensitivity to docetaxel such as stress-related DNA damage and
apoptosis, while cell cycle arrest and survival confer resistance.
However, in another study with a small number of patients, no
specific gene expression profile was identified for response to
doxorubicin-cyclophosphamide or doxorubicin-doxetaxel, which
advocated for larger cohorts [56].
The search for predictors of response to treatment is currently
being studied in different laboratories. A collection of 60
ER-positive tumours was analysed to identify differentially
expressed genes between responders and non responders to tamoxifen
as a monotherapy following primary surgery. HOXB13 and IL17BR mRNA
levels, determined by semi-quantitative PCR, are sufficient to
predict outcome in an independent set of samples. Interestingly,
increased HOXB3 was observed in non-responding tumours. In vitro
constitutive expression of this gene confers motile and invasive
properties to the MCF10A mammary cell line. HOXB3 interferes in the
control of ER signalling by an unknown mechanism, as is the case
for EGFR and HER2 signalling, which are known to alter the response
to tamoxifen [57]. This important finding, however, was not
validated on an independent collection of tumours [58] stressing
the crucial importance of analysing very large and more homogenous
cohorts of tumours.
A 64-gene signature distinguishing good and poor prognosis was
established on a training set comprising node-negative and
node-positive patients who did or did not receive adjuvant therapy.
This set of genes was complemented by a risk factor score. The
training set showed that the patients could be subdivided in three
clusters; the first cluster contained mostly patients who did well
without treatment and the third cluster corresponded to patients
who did poorly with treatment but may benefit from other protocols.
However, the second cluster was not informative. It was not
identifying a group of patients who did poorly without treatment
and who, therefore, could have benefited from treatment. This study
was potentially aimed at determining which patients could escape
systemic chemotherapy and which patients could be treated with an
alternative therapeutic protocol to overcome failure from the
conventional treatment [59].
Resistance to trastuzumab is encountered relatively frequently;
however, the mechanism by which this resistance is acquired remains
unknown. One recent study has addressed this issue by establishing
a carcinoma cell line from a patient resistant to trastuzumab. This
cell line shares many characteristics of the primary tumour;
although it has an amplified HER2 locus, this cell line has a mixed
basal and luminal phenotype. This tumour type is, therefore,
atypical since the HER2 cluster is mostly of the luminal phenotype.
The lack of inhibition of AKT phosphorylation by trastuzumab is so
far the only detected alteration in signalling. However, PI3K
inhibitors have not been used in this study to determine whether
resistance to trastuzumab can be overcome [60]. Resistance may also
be acquired through a steric hindrance mechanism mediated by MUC4
at the cell surface. Diminished expression of MUC by RNA
interference resulted in increased binding to trastuzumab,
potentially abrogating resistance to this therapy [61].
Defining resistance
Multidrug resistance is a well know phenomenon applied to most
tumour tissues. Numerous studies have addressed mechanisms driving
this resistance. The RNA profiling of the 48 ABC transporters,
established by PCR in the NCI cell line collection, compared the
ability to respond to a panel of 1429 drugs [62] in a much better
correlated manner than a previous study based on expression profile
of 9000 transcripts [63]. A surprising result was that MDR1 (ABCB1)
overexpression potentiated the cytotoxic activity of some drugs
rather than resistance. This study opens a new strategy to overcome
drug resistance in a more rational way.
Murine models
Numerous transgenic murine models of breast carcinoma have now been
developed through the targeting of oncogenes, mostly using the MMTV
or WAP promoters. The two promoters are specifically expressed in
the luminal epithelium, but the MMTV promoter is also expressed in
some other epithelia and is expressed at an early stage in mammary
gland development, prior to the terminal differentiation into
secretory cells. A major effort has been devoted to classify
precisely the proliferative lesions [64]. Most tumours forming in
genetically engineered mice are morphologically distinct from
spontaneous MMTV or chemically induced tumours. Many of these
tumours in genetically engineered models are not closely related to
human breast tumours as they exhibit squamous metaplasia. However,
they have been extremely useful to assess the role of known
oncogenes. Tumours induced by each oncogene have a specific
morphological and molecular signature as revealed by a recent study
of KRAS2 expression signature in mouse and human lung cancers [65].
For instance, HER2 tumours are composed of solid sheets of
carcinoma cells without glandular differentiation. The c-MYC
expressing tumours have large cells with pleiomorphic nuclei with a
coarse chromatin and prominent nucleoli. RAS tumours form
papillary-like tumours resembling transitional cell carcinoma of
the bladder. The Ret 1 tumours form small crowded glands with large
pleiomorphic nuclei. HER2 and SV40Tag transgenes can produce ductal
carcinoma in situ of the comedo-type resembling human tumours.
Papillary carcinoma can be obtained with the cyclin D1 transgene.
The phenotype of multigenic transgene derived tumours is often
determined by the dominating oncogene such as c-MYC. Much care
should be paid to the genetic background of the mouse and different
phenotypes are obtained with MMTV or WAP promoters. Human and mouse
tumours differ notably with respect to their relative sensitivity
to hormones, their stroma, their capacity to metastasise and their
pattern of metastasis. Using terminal differentiation markers,
luminal myoepithelial and mesenchymal phenotypes have been
identified in a large variety of mouse tumors. Three types of
neoplasms have been described; simple carcinoma, complex carcinoma
possibly originating from a stem cell, and carcinoma undergoing an
epithelial-mesenchymal transition (EMT). Remarkably, an EMT
phenotype [66, 67] has been described in c-MYC, RAS and SV40 Tag
driven tumours [68]. Mammary epithelial cells expressing Met and
Myc can develop into tumours mixed luminal and myoepithelial, when
transplanted in the mammary fat pad suggesting that these tumours
arose from a bipotent progenitor [69].
Recently, the analysis of an epithelial cell line derived from
the mouse mammary gland taken at the mid gestation stage showed
remarkable epithelial cell plasticity. When deprived from EGF,
these mammary epithelial cells acquired a fibroblastic phenotype
and expressed characteristic markers of the basal phenotype such as
K5/14 and P-cadherin [70]. Their injection in vivo in the cleared
mammary fat pad clearly showed their capacity to produce luminal
cells. These findings indicating that these basal cells display
progenitor properties together with the demonstration that the
Wnt/b-catenin pathway is playing a crucial role in the maintenance
of progenitor cells in different epithelia prompted experiments to
target a truncated β-catenin (resulting in constitutive activation
of the pathway) into the basal myoepithelial layer. The
myoepithelial layer has not prompted as many studies as the luminal
layer and many less transgenic mice have been targeted to this
basal layer. As the myoepithelial layer is in direct contact with
the extracellular environment and interacts also directly with the
luminal layer, any alteration of these cells could direct
consequences for the luminal layer. Its disappearance in carcinomas
suggests a direct role in controlling invasive behaviour of DCIS
[71]. The truncated b-catenin transgene induced excessive lateral
branching and precocious lobulo-alveolar development of the mammary
gland at mid-gestation. Most interestingly, hyperplastic foci were
observed in the basal layer. These cells expressed basal
cytokeratins, but not smooth muscle a-actin, indicating their
undifferentiated state. Multiparous mice also exhibited squamous
carcinoma and most importantly invasive carcinoma with a strong
basal phenotype [72]. These transgenic mice potentially represent a
useful model to study breast carcinoma of the basal phenotypes.
Moreover the formation of undifferentiated basal tumors can be
interpreted as the amplification of a population of basal-type
mammary progenitors. The mammary gland is hypothesised to contain
one epithelial stem/progenitor cell in every 2000 cells [73].
Several studies have described attempts to isolate the
stem/progenitors cells in the mouse using various approaches [74].
These cells may be evidenced in vivo as a subpopulation of BrdU
long-term label retaining epithelial cells in the mouse and human
mammary tissues. Mammary epithelial cells belonging to the
so-called long-term label retaining cells are found to divide
asymmetrically and to retain their template strand. These cells
also self renew, thus they may represent the mammary stem cells
[75]. Other studies using surface markers have shown that Sca-1
positive epithelial cells from the mouse mammary gland had a much
higher regenerative potential in vivo than the Sca-1 negative cells
[76]. Very recently, two studies reported the isolation of a cell
population from the mouse mammary epithelium that is able, at the
clonal level, to give rise to the entire mammary gland if
transplanted in vivo. Interestingly, these progenitor cells were
characterized by high surface levels of integrins and cytoskeletal
markers of basal epithelial cells [77, 78].
Cancer stem cells
The presence of cancer stem cells has long been hypothesized in
solid tumours. Strong evidence was already provided in the
seventies through the analysis of teratocarcinomas [79]. The
enrichment of metastatic breast carcinoma cells derived from
pleural effusions using CD44 and CD24 as sorting criteria showed
that as low as 100 cells could form a malignant tumour [80]. This
pioneering study opened the road for new investigations in cancer
stem cells to further enrich and characterise the phenotype and
response to treatment. Breast cancer stem cell research is at very
early stages and many issues remain unsolved. The isolation of
stem/progenitor cells capable of self-renewing was successfully
achieved with a few primary high grade, ER positive tumours
specimens. As low as 100 tumour cells could form a tumour in SCID
mice. The phenotype of these cells was CD44-positive,
CD24-negative, Oct4-positive and connexin 43-negative [81]. An
important issue is whether these cells are enriched in the
so-called side population and whether this is related to the level
of expression of the ABCG2 transporter [82]. The phenotype of the
stem/precursor cells needs to be more accurately defined. In
addition, there may be several distinct types of progenitor cells
as has been already established for normal mammary gland [83, 84].
The fact that self-renewing progenitors were not identified from
aggressive ER negative tumours suggests that culture conditions may
not have been suitable. Alternatively, stem cells from different
types of breast cancers may express different phenotypic markers.
Clearly, conventional drugs are unlikely to efficiently eradicate
quiescent stem cells. In the same manner these cells may well be
resistant to radiotherapy and would then be responsible for local
or distant relapses. It would be intriguing to characterise such
cells in bone marrow micrometastasis and correlate their presence
with tumour progression in these patients.
Concluding remarks
Breast carcinomas comprise a very large set of remarkably
heterogeneous tumours. The conventional treatment of breast cancers
has made substantial progress over the last 20 years. The new
targeted therapies, although permitting longer-term survival, have
so far failed to cure metastatic diseases. At best, metastatic
cancer patients could benefit from protocols treating a chronic
disease. However, as it is already known from chronic myeloid
leukaemia patients treated with Gleevec™, resistance can be
acquired through specific mutations of the Abelson tyrosine kinase,
leading eventually to progression to an acute phase. The current
dogma is that one must treat cancer stem/progenitor cells in
addition to the actively proliferating cells. Defining such
stem/progenitor cells, which may be quite heterogeneous themselves,
requires more basic studies on normal stem/progenitor cells in
order to understand their phenotypes. Well designed transgenic
models may help refine our understanding of cancer stem/progenitor
cells.
Another major effort is to define better molecular markers,
which, in conjunction with the well established histoprognostic
markers, will permit tailored individual therapy. The major
challenge is indeed in the remarkable heterogeneity of breast
carcinoma. Progress has been made in unravelling this issue with
high-density arrays, landscaping the genome and the transcriptome
of breast tumours. Other high-density screening epigenetic
modifications, such as the methylome and posttranslational
modifications such as the phosphokinome, have considerable
potential to further define the molecular status of each tumour.
These combined studies may bring more robustness to the
transcriptomic data. They may also offer new potential targets for
therapies. The more conventional surrogate markers now routinely
used, such as bone micrometatases, must also be considered for
prognosis and for evaluation of response to therapy. This
formidable task which the research community now faces will provide
major benefits to breast cancer patients in the near future.
References
1 Singletary SE, Allred C, Ashley P,
Bassett LW, Berry D, Bland KI, et al. Revision
of the American Joint Committee on Cancer staging system for breast
cancer. J Clin Oncol 2002 ; 20 : 3628-36.
2 Rampaul RS, Pinder SE, Elston CW,
Ellis IO. Prognostic and predictive factors in primary breast
cancer and their role in patient management : The Nottingham
Breast Team. Eur J Surg Oncol 2001 ; 27 : 229-38.
3 Harvey JM, Clark GM, Osborne CK,
Allred DC. Estrogen receptor status by immunohistochemistry is
superior to the ligand-binding assay for predicting response to
adjuvant endocrine therapy in breast cancer. J Clin Oncol
1999 ; 17 : 1474-81.
4 Kim T, Giuliano AE, Lyman GH. Lymphatic mapping
and sentinel lymph node biopsy in early-stage breast carcinoma.
Cancer 2006 ; 106 : 4-16.
5 Cristofanilli M, Budd GT, Ellis MJ,
Stopeck A, Matera J, Miller MC, et al.
Circulating tumor cells, disease progression, and survival in
metastatic breast cancer. N Engl J Med 2004 ; 351 :
781.
6 Pierga JY, Bonneton C, Vincent-Salomon A, de
Crémoux P, Nos C, Blin N, et al. Clinical
significance of immunocytochemical detection of tumor cells using
digital microscopy in peripheral blood and bone marrow of breast
cancer patients. Clin Cancer Res 2004 ; 10 :
1392-400.
7 Pantel K, Cote RJ, Fodstad O. Detection and
clinical importance of micrometastatic disease. J Natl Cancer Inst
1999 ; 91 : 1113-24.
8 Braun S, Pantel K, Muller P, Janni W,
Hepp F, Kentenich CR, et al. Cytokeratin-positive
cells in the bone marrow and survival of patients with stage I, II,
or III breast cancer. N Engl J Med 2000 ; 342 :
525-33.
9 Braun S, Vogl FD, Naume B, Janni W,
Osborne MP, Coombes RC, et al. A pooled analysis of
bone marrow micrometastasis in breast cancer. N Engl J Med
2005 ; 353 : 793-802.
10 Pierga JY, Bonneton C, Magdelenat H,
Vincent-Salomon A, Nos C, Pouillart P, et al.
Clinical significance of proliferative potential of occult
metastatic cells in bone marrow of patients with breast cancer. Br
J Cancer 2003 ; 89 : 539-45.
11 Choesmel V, Anract P, Hoifodt H,
Thiery JP, Blin N. A relevant immunomagnetic assay to
detect and characterize epithelial cell adhesion molecule-positive
cells in bone marrow from patients with breast carcinoma :
immunomagnetic purification of micrometastases. Cancer 2004 ;
101 : 693-703.
12 Schmidt-Kittler O, Ragg T, Daskalakis A,
Granzow M, Ahr A, Blankenstein TJ, et al. From
latent disseminated cells to overt metastasis : genetic
analysis of systemic breast cancer progression. Proc Natl Acad Sci
USA 2003 ; 100 : 7737-42.
13 O’Connell P. Genetic and cytogenetic analyses of breast
cancer yield different perspectives of a complex disease. Breast
Cancer Res Treat 2003 ; 78 : 347-57.
14 Al-Kuraya K, Schraml P, Torhorst J,
Tapia C, Zaharieva B, Novotny H, et al.
Prognostic relevance of gene amplifications and coamplifications in
breast cancer. Cancer Res 2004 ; 64 : 8534-40.
15 Nessling M, Richter K, Schwaenen C,
Roerig P, Wrobel G, Wessendorf S, et al.
Candidate genes in breast cancer revealed by microarray-based
comparative genomic hybridization of archived tissue. Cancer Res
2005 ; 65 : 439-47.
16 Jones C, Ford E, Gillett C, Ryder K,
Merrett S, Reis-Filho JS, et al. Molecular
cytogenetic identification of subgroups of grade III invasive
ductal breast carcinomas with different clinical outcomes. Clin
Cancer Res 2004 ; 10 : 5988-97.
17 Hupe P, Stransky N, Thiéry JP,
Radvanyi F, Barillot E. Analysis of array CGH data :
from signal ratio to gain and loss of DNA regions. Bioinformatics
2004 ; 20 : 3413-22.
18 Hwang ES, DeVries S, Chew KL, Moore DH,
Kerlikowske K, Thor A, et al. Patterns of
chromosomal alterations in breast ductal carcinoma in situ. Clin
Cancer Res 2004 ; 10 : 5160-7.
19 Samuels Y, Velculescu VE. Oncogenic mutations of
PIK3CA in human cancers. Cell Cycle 2004 ; 3 :
1221-4.
20 Bachman KE, Argani P, Samuels Y,
Silliman N, Ptak J, Szabo S, et al. The PIK3CA
gene is mutated with high frequency in human breast cancers. Cancer
Biol Ther 2004 ; 3 : 772-5.
21 Campbell IG, Russell SE, Choong DY,
Montgomery KG, Ciavarella ML, Hooi CS, et al.
Mutation of the PIK3CA gene in ovarian and breast cancer. Cancer
Res 2004 ; 64 : 7678-81.
22 Lee JW, Soung YH, Kim SY, Lee HW,
Park WS, Nam SW, et al. PIK3CA gene is frequently
mutated in breast carcinomas and hepatocellular carcinomas.
Oncogene 2005 ; 24 : 1477-80.
23 Li SY, Rong M, Grieu F, Iacopetta B.
PIK3CA mutations in breast cancer are associated with poor outcome.
Breast Cancer Res Treat 2006 ; 96 : 91-5.
24 Saal LH, Holm K, Maurer M, Memeo L,
Su T, Wang X, et al. PIK3CA mutations correlate with
hormone receptors, node metastasis, and ERBB2, and are mutually
exclusive with PTEN loss in human breast carcinoma. Cancer Res
2005 ; 65 : 2554-9.
25 Isakoff SJ, Engelman JA, Irie HY, Luo J,
Brachmann SM, Pearline RV, et al. Breast
cancer-associated PIK3CA mutations are oncogenic in mammary
epithelial cells. Cancer Res 2005 ; 65 : 10992-1000.
26 Stephens P, Edkins S, Davies H,
Greenman C, Cox C, Hunter C, et al. A screen of
the complete protein kinase gene family identifies diverse patterns
of somatic mutations in human breast cancer. Nat Genet 2005 ;
37 : 590-2.
27 Perou CM, Sorlie T, Eisen MB, van de
Rijn M, Jeffrey SS, Rees CA, et al. Molecular
portraits of human breast tumours. Nature 2000 ; 406 :
747-52.
28 Sorlie T, Perou CM, Tibshirani R, Aas T,
Geisler S, Johnsen H, et al. Gene expression
patterns of breast carcinomas distinguish tumor subclasses with
clinical implications. Proc Natl Acad Sci USA 2001 ; 98 :
10869-74.
29 Sorlie T, Tibshirani R, Parker J,
Hastie T, Marron JS, Nobel A, et al. Repeated
observation of breast tumor subtypes in independent gene expression
data sets. Proc Natl Acad Sci USA 2003 ; 100 :
8418-23.
30 Nielsen TO, Hsu FD, Jensen K, Cheang M,
Karaca G, Hu Z, et al. Immunohistochemical and
clinical characterization of the basal-like subtype of invasive
breast carcinoma. Clin Cancer Res 2004 ; 10 :
5367-74.
31 Ma XJ, Salunga R, Tuggle JT, Gaudet J,
Enright E, McQuary P, et al. Gene expression
profiles of human breast cancer progression. Proc Natl Acad Sci USA
2003 ; 100 : 5974-9.
32 Korkola JE, DeVries S, Fridlyand J,
Hwang ES, Estep AL, Chen YY, et al.
Differentiation of lobular versus ductal breast carcinomas by
expression microarray analysis. Cancer Res 2003 ; 63 :
7167-75.
33 Van’t Veer LJ, Dai H, van de Vijver MJ,
He YD, Hart AA, Mao M, et al. Gene expression
profiling predicts clinical outcome of breast cancer. Nature
2002 ; 415 : 530-6.
34 Van de Vijver MJ, He YD, van’t Veer LJ,
Dai H, Hart AA, Voskuil DW, et al. A
gene-expression signature as a predictor of survival in breast
cancer. N Engl J Med 2002 ; 347 : 1999-2009.
35 Chang HY, Nuyten DS, Sneddon JB,
Hastie T, Tibshirani R, Sorlie T, et al.
Robustness, scalability, and integration of a wound-response gene
expression signature in predicting breast cancer survival. Proc
Natl Acad Sci USA 2005 ; 102 : 3738-43.
36 Ramaswamy S, Ross KN, Lander ES,
Golub TR. A molecular signature of metastasis in primary solid
tumors. Nat Genet 2003 ; 33 : 49-54.
37 Weigelt B, Glas AM, Wessels LF,
Witteveen AT, Peterse JL. van’t Veer LJ. Gene expression
profiles of primary breast tumors maintained in distant metastases.
Proc Natl Acad Sci USA 2003 ; 100 : 15901-5.
38 Wang Y, Klijn JG, Zhang Y, Sieuwerts AM,
Look MP, Yang F, et al. Gene-expression profiles to
predict distant metastasis of lymph-node-negative primary breast
cancer. Lancet 2005 ; 365 : 671-9.
39 Nevins JR, Huang ES, Dressman H,
Pittman J, Huang AT, West M. Towards integrated
clinico-genomic models for personalized medicine : combining
gene expression signatures and clinical factors in breast cancer
outcomes prediction. Hum Mol Genet 2003 ; (12 Spec Iss
2 : R153-7).
40 Huang E, Cheng SH, Dressman H, Pittman J,
Tsou MH, Horng CF, et al. Gene expression predictors
of breast cancer outcomes. Lancet 2003 ; 361 :
1590-6.
41 Dai H, van’t Veer L, Lamb J, He YD,
Mao M, Fine BM, et al. A cell proliferation
signature is a marker of extremely poor outcome in a subpopulation
of breast cancer patients. Cancer Res 2005 ; 65 :
4059-66.
42 Miller LD, Smeds J, George J, Vega VB,
Vergara L, Ploner A, et al. An expression signature
for p53 status in human breast cancer predicts mutation status,
transcriptional effects, and patient survival. Proc Natl Acad Sci
USA 2005 ; 102 : 13550-5.
43 Minn AJ, Kang Y, Serganova I, Gupta GP,
Giri DD, Doubrovin M, et al. Distinct organ-specific
metastatic potential of individual breast cancer cells and primary
tumors. J Clin Invest 2005 ; 115 : 44-55.
44 Minn AJ, Gupta GP, Siegel PM, Bos PD,
Shu W, Giri DD, et al. Genes that mediate breast
cancer metastasis to lung. Nature 2005 ; 436 :
518-24.
45 Brenton JD, Carey LA, Ahmed AA, Caldas C.
Molecular classification and molecular forecasting of breast
cancer : ready for clinical application? J Clin Oncol
2005 ; 23 : 7350-60.
46 Michiels S, Koscielny S, Hill C. Prediction of
cancer outcome with microarrays : a multiple random validation
strategy. Lancet 2005 ; 365 : 488-92.
47 Mukherjee S, Tamayo P, Rogers S,
Rifkin R, Engle A, Campbell C, et al.
Estimating data size requirements for classifying DNA micrarray
data. J Comp Biol 2003 ; 10 : 119-42.
48 Ein-Dor L, Kela I, Getz G, Givol D,
Domany E. Outcome signature genes in breast cancer : is
there a unique set? Bioinformatics 2005 ; 21 : 171-8.
49 Bloom G, Yang IV, Boulware D, Kwong KY,
Coppola D, Eschrich S, et al. Multi-platform,
multi-site, microarray-based human tumor classification. Am J
Pathol 2004 ; 164 : 9-16.
50 Van’t Veer LJ, Paik S, Hayes DF. Gene
expression profiling of breast cancer : a new tumor marker. J
Clin Oncol 2005 ; 23 : 1631-5.
51 Ellis M, Ballman K. Trawling for genes that predict
response to breast cancer adjuvant therapy. J Clin Oncol
2004 ; 22 : 2267-9.
52 Ayers M, Symmans WF, Stec J, Damokosh AI,
Clark E, Hess K, et al. Gene expression profiles
predict complete pathologic response to neoadjuvant paclitaxel and
fluorouracil, doxorubicin, and cyclophosphamide chemotherapy in
breast cancer. J Clin Oncol 2004 ; 22 : 2284-93.
53 Rouzier R, Perou CM, Symmans WF,
Ibrahim N, Cristofanilli M, Anderson K, et al.
Breast cancer molecular subtypes respond differently to
preoperative chemotherapy. Clin Cancer Res 2005 ; 11 :
5678-85.
54 Chang JC, Wooten EC, Tsimelzon A,
Hilsenbeck SG, Gutierrez MC, Elledge R, et al.
Gene expression profiling for the prediction of therapeutic
response to docetaxel in patients with breast cancer. Lancet
2003 ; 362 : 362-9.
55 Chang JC, Wooten EC, Tsimelzon A,
Hilsenbeck SG, Gutierrez MC, Tham YL, et al.
Patterns of resistance and incomplete response to docetaxel by gene
expression profiling in breast cancer patients. J Clin Oncol
2005 ; 23 : 1169-77.
56 Hannemann J, Oosterkamp HM, Bosch CA,
Velds A, Wessels LF, Loo C, et al. Changes in
gene expression associated with response to neoadjuvant
chemotherapy in breast cancer. J Clin Oncol 2005 ; 23 :
3331-42.
57 Ma XJ, Wang Z, Ryan PD, Isakoff SJ,
Barmettler A, Fuller A, et al. A two-gene expression
ratio predicts clinical outcome in breast cancer patients treated
with tamoxifen. Cancer Cell 2004 ; 5 : 607-16.
58 Reid JF, Lusa L, De Cecco L, Coradini D,
Veneroni S, Daidone MG, et al. Limits of predictive
models using microarray data for breast cancer clinical treatment
outcome. J Natl Cancer Inst 2005 ; 97 : 927-30.
59 Pawitan Y, Bjohle J, Amler L, Borg AL,
Egyhazi S, Hall P, et al. Gene expression profiling
spares early breast cancer patients from adjuvant therapy :
derived and validated in two population-based cohorts. Breast
Cancer Res 2005 ; 7 : R953-R964.
60 Tanner M, Kapanen AI, Junttila T,
Raheem O, Grenman S, Elo J, et al.
Characterization of a novel cell line established from a patient
with Herceptin-resistant breast cancer. Mol Cancer Ther 2004 ;
3 : 1585-92.
61 Nagy P, Friedlander E, Tanner M,
Kapanen AI, Carraway KL, Isola J, et al.
Decreased accessibility and lack of activation of ErbB2 in JIMT-1,
a herceptin-resistant, MUC4-expressing breast cancer cell line.
Cancer Res 2005 ; 65 : 473-82.
62 Szakacs G, Annereau JP, Lababidi S,
Shankavaram U, Arciello A, Bussey KJ, et al.
Predicting drug sensitivity and resistance : profiling ABC
transporter genes in cancer cells. Cancer Cell 2004 ; 6 :
129-37.
63 Scherf U, Ross DT, Waltham M, Smith LH,
Lee JK, Tanabe L, et al. A gene expression database
for the molecular pharmacology of cancer. Nat Genet 2000 ;
24 : 236-44.
64 Cardiff RD, Anver MR, Gusterson BA,
Hennighausen L, Jensen RA, Merino MJ, et al.
The mammary pathology of genetically engineered mice : the
consensus report and recommendations from the Annapolis meeting.
Oncogene 2000 ; 19 : 968-88.
65 Sweet-Cordero A, Mukherjee S, Subramanian A,
You H, Roix JJ, Ladd-Acosta C, et al. An
oncogenic KRAS2 expression signature identified by cross-species
gene-expression analysis. Nat Genet 2005 ; 37 :
48-55.
66 Thiéry JP. Epithelial-mesenchymal transitions in tumour
progression. Nat Rev Cancer 2002 ; 2 : 442-54.
67 Vincent-Salomon A, Thiery JP. Host microenvironment
in breast cancer development : epithelial-mesenchymal
transition in breast cancer development. Breast Cancer Res
2003 ; 5 : 101-6.
68 Mikaelian I, Blades N, Churchill GA,
Fancher K, Knowles BB, Eppig JT, et al.
Proteotypic classification of spontaneous and transgenic mammary
neoplasms. Breast Cancer Res 2004 ; 6 : R668-R679.
69 Welm AL, Kim S, Welm BE. Bishop JM. MET and
MYC cooperate in mammary tumorigenesis. Proc Natl Acad Sci USA
2005 ; 102 : 4324-9.
70 Deugnier MA, Faraldo MM, Janji B,
Rousselle P, Thiéry JP, Glukhova MA. EGF controls
the in vivo developmental potential of a mammary epithelial cell
line possessing progenitor properties. J Cell Biol 2002 ;
159 : 453-63.
71 Faraldo MM, Teulière J, Deugnier MA, Taddei de
la Hosseraye I, Thiéry JP, Glukhova M.
Myopepithelial cells in the control of mammary development and
tumorigenesis : data from genetically modified mice. J Mammary
Gland Biol Neoplasia 2005 ; 10 : 211-9.
72 Teulière J, Faraldo MM, Deugnier MA,
Shtutman M, Ben-Ze’ev A, Thiéry JP, et al.
Targeted activation of beta-catenin signaling in basal mammary
epithelial cells affects mammary development and leads to
hyperplasia. Development 2005 ; 132 : 267-77.
73 Smith GH, Medina D. A morphologically distinct
candidate for an epithelial stem cell in mouse mammary gland. J
Cell Sci 1988 ; 90 : 173-83.
74 Woodward WA, Chen MS, Behbod F, Rosen JM.
On mammary stem cells. J Cell Sci 2005 ; 118 :
3585-94.
75 Smith GH. Label-retaining epithelial cells in mouse
mammary gland divide asymmetrically and retain their template DNA
strands. Development 2005 ; 132 : 681-7.
76 Welm BE, Tepera SB, Venezia T,
Graubert TA, Rosen JM, Goodell MA. Sca-1(pos) cells
in the mouse mammary gland represent an enriched progenitor cell
population. Dev Biol 2002 ; 245 : 42-56.
77 Stingl J, Eirew P, Ricketson I,
Shackleton M, Vaillant F, Choi D, et al.
Purification and unique properties of mammary epithelial stem
cells. Nature 2006 ; 439 : 993-7.
78 Shackleton M, Vaillant F, Simpson KJ,
Stingl J, Smyth GK, Asselin-Labat ML, et al.
Generation of a functional mammary gland from a single stem cell.
Nature 2006 ; 439 : 84-8.
79 Reya T, Morrison SJ, Clarke MF,
Weissman IL. Stem cells, cancer, and cancer stem cells. Nature
2001 ; 414 : 105-11.
80 Al-Hajj M, Wicha MS, Benito-Hernandez A,
Morrison S, Clarke MF. Prospective identification of
tumorigenic breast cancer cells. Proc Natl Acad Sci USA 2003 ;
100 : 3983-8.
81 Ponti D, Costa A, Zaffaroni N, Pratesi G,
Petrangolini G, Coradini D, et al. Isolation and in
vitro propagation of tumorigenic breast cancer cells with
stem/progenitor cell properties. Cancer Res 2005 ; 65 :
5506-11.
82 Patrawala L, Calhoun T, Schneider-Broussard R,
Zhou J, Claypool K, Tang DG. Side population is
enriched in tumorigenic, stem-like cancer cells, whereas ABCG2+ and
ABCG2- cancer cells are similarly tumorigenic. Cancer Res
2005 ; 65 : 6207-19.
83 Clarke RB. Isolation and characterization of human
mammary stem cells. Cell Prolif 2005 ; 38 : 375-86.
84 Stingl J, Raouf A, Emerman JT, Eaves CJ.
Epithelial progenitors in the normal human mammary gland. J Mammary
Gland Biol Neoplasia 2005 ; 10 : 49-59.
* The Breast cancer group: B. Asselain, A.
Aurias, E. Barillot, F. Campana, P. de Crémoux, V. Diéras, O.
Delattre, A. Fourquet, M.-F. Poupon,F. Radvanyi, J.-Y. Pierga, L.
Mignot, R. Salmon, A. Salomon, B. Sigal-Zafrani, D. Stoppa Lyonnet,
A. Tardivon, F. Thibault, J.-P. Thiery, P. This.
|
Printable version
Version PDF
