|
Background
During the last fifteen years
there has been considerable innovation and development of techniques which can
be used to study cell biology and behaviour. These including immunocytochemistry,
monoclonal antibody technology, molecular biology and biomedical engineering
resulting in advanced flow and image cytometric systems. The volume of research
results emerging from basic scientists, pathologists and clinicians relative to
these techniques is ever increasing. This generation of data demonstrates the
immense interest in the potential of these methods to provide useful information
of biological and clinical importance. Breast cancer is one area that has
received such attention because of its frequency in western countries and its
exhibition of a wide spectrum of clinical behaviour. An acceptance that
conservation techniques of surgery and radiotherapy are safe alternatives to
mastectomy and developments of chemotherapeutic schedules have now provided the
patient, surgeon and oncologist with realistic choices of treatment. The
question now being asked by many groups is whether this choice can be influenced
appropriately by a knowledge of molecular processes which are present or have
occurred and which may influence tumour cell.
Monoclonal
and Polyclonal antibodies
A host of both polyclonal and
monoclonal and monoclonal antibodies have been produced to breast epithelial
cells, their cell products, tumours or to common cell determinants which have
associations with breast disease. The expression of antigens identified by these
antibodies can be studied either by immunocytochemistry on tissue sections or
cytology preparations, by enzyme linked immunozorbant assay (ELIZA) of body
fluids or soluble cell fractions and by flow cytometry of single cell or cell
fraction preparations. Antigens which vary in expression or appear at specific
times during normal cell growth, function, differentiation, proliferation or
neoplasia are of particular interest.
Antibodies to the following
molecules have stimulated the most interest in breast cancer research:
Epithelial
Mucins
The primary functional role of
breast epithelial cells is to produce milk during lactation. The lipid rich
droplets are surrounded by a membrane derived from the epithelial cell apical
membrane. This membrane is rich in carbohydrates on the external and, uniquely,
on the internal cytoplasmic side. A highly immunogenic mucin component of the
membrane has stimulated considerable investigation. Initially polyclonal
antibodies to delipidated human milk fat globule membrane (HMFGM) 
were produced. Following production of polyclonal antiserum many groups have
produced monoclonal antisera to such high molecular weight glycoprotein
mucins(see reviews
normally expressed on apical membrane of human breast and other epithelial
cells. The varying reactivity of some of these antibodies gave the initial
impression that there may be a family of such mucins present in glandular
epithelial cells and led to a variety of names such as epithelial membrane
antigen (EMA),
the 'MAM' series 'polymorphic epithelial mucin' (PEM)
and episialin
being used to describe the mucin recognised by particular antibodies.
Interlaboratory collaborative studies
have shown that one highly immunogenic high molecular weight glycoprotein (over
400 KD) carries most of the epitopes recognised by these antibodies. This mucin
is now known as MUC 1
and it is a member of the heterogeneous group, of at least eight highly
glycosylated mucin proteins (MUC 1-8), which form the major component of mucus.
They have a central threadlike non-globular protein core with highly
glycosylated and unglycosylated regions. MUC 1 is a transmembrane glycoprotein
with a large cytoplasmic tail which interacts with the actin containing
microfilaments. The core protein is made up of tandem repeats of 20 amino acids
and reacts with many of the antibodies raised to HMFGM and breast cancer cells.
Carcinomas show a difference in glycosylation of the core protein from normal
cells which could result in exposure of some epitopes in malignancy. The
function MUC 1 and other mucins is not clear but they are thought to have roles
in cell protection or lubrication, maintenance of viscosity in secretions,
regulation of cell growth and cellular recognition.
They are thought to facilitate tumour protection from immune attack, tumour
growth and metastasis.
Expression of other mucins such
as MUC 2, 3 and 8 is seen in a proportion of breast carcinomas.
Other antisera to lower molecular weight blood group
and oncofetal antigens
have also been produced through immunisation with HMFGM.
Growth
factors and their receptors
Peptide growth factors
Normal and malignant breast
epithelial cells co-express a number of epidermal growth factor (EGF) related
peptides including EGF, transforming growth factor alpha (TGF- ,
amphiregulin (AR), and cripto-1 (CR-1).
The frequency and level of expression of TGFa, AR, and CR-1 are higher in breast
cancer lines. Some of these peptides can function as autocrine and/or juxtacrine
growth factors in mammary epithelial cells and they are regulated by hormones
such as oestrogen receptor. The lack of expression in some normal and malignant
mammary epithelial cells suggests that some of these peptides may be involved in
regulating other aspects of cell behaviour such as differentiation as well as
proliferation.
Nine different classes of
tyrosine kinase growth factor receptors have been identified. They differ with
respect to the structure of their extra-cellular ligand binding and
intracellular kinase domains and the nature of their activating ligands.
The type 1 family includes epidermal growth factor receptor (EGFR, c-erbB-1),
c-erbB-2 (HER2, neu), c-erbB-3 (HER3) and c-erbB-4 (HER4) receptors, all four of
which are expressed in breast cancer.
Epidermal
Growth Factor and Epidermal Growth Factor Receptor
Epidermal growth factor
(EGF) is
a powerful polypeptide growth factor which is essential in the development of
mammary glands in mice.
It has been shown to influence, and in some situations be necessary, for the
growth of normal mammary epithelium,
breast cancer cell lines
and other cell lines.
The mitogenic effect of EGF is mediated through binding with a specific membrane
receptor - epidermal growth factor receptor (EGFR).
It is a 170 KD transmembrane glycoprotein with a heavily glycosylated external
domain responsible for ligand binding, a single transmembrane spanning sequence
and an internal domain with tyrosine kinase activity.
Activation of the receptor induces cell division usually synergistically with
other growth regulatory signals. The protein sequence for EGFR has been
determined and its gene cloned. The gene and receptor show close similarity to
two oncogenes and their oncoproteins, namely V-erbB-2 and C-erbB-2
EGFRs have been found in a variety of animal and human tissues, are particularly
elevated in squamous tumours of the skin
and have been identified in a proportion of breast cancers.
Regulation of EGFR is not clearly understood but the level of expression in
breast cells can be altered by EGF, TGF-a and steroid hormones.
There is a striking inverse relationship between EGFR and oestrogen receptor
(ER) expression.
EGFR may be measured in breast
tumours by radioligand binding assay of membrane fractions and by
immunohistological staining of tumour sections. The reported frequency of
expression varies between 15 - 60% of tumours. There is a relationship to tumour
size
which could influence the frequency in individual studies. Clinical interest in
EGFR has been further stimulated by the demonstration of an association between
EGFR expression and poor.
EGFR also appears to have an influence on the processes of tumour invasion and
dissemination and which has led to speculation that it may be a suitable target
for antimetastatic therapy.
Transforming
growth factors alpha and beta
The epidermal growth factor
family of growth regulating peptides also includes transforming growth factor
alpha (TGF-)
a related single chain polypeptide
which can stimulated growth by binding to and activating the EGF receptor.
In normal breast epithelial cell lines and some breast cancer cell lines the
production of TGF-
is controlled in part by oestrogen which stimulates TGF-
synthesis and secretion.
It is secreted by all tumour cell lines including breast and clinical and
experimental studies have demonstrated that TGF-
is an important modulator of malignant progression of mammary epithelial cells
in breast cancer.
Transforming growth factor beta
(TGF-ß) is an unrelated two chain polypeptide which is a member of a complex
structurally related family of growth and differentiation factors.
The various forms of TGF-ß bind to a set of three structurally and functionally
distinct cell surface receptors.
TGF-ß has a growth inhibitory effect on epithelial cells, including mammary
epithelium.
Some squamous cell carcinoma cell lines are not inhibited by TGF-ß and it has
been suggested that TGF-ß may be involved in regulation of normal epithelial
cell growth through negative feedback, carcinomas having altered growth due to
their lack of response.
The evidence that both TGF-
and TGF-ß are produced in situ has led to speculation that there is an
autocrine growth loop, involving TGF-a and modulated by TGF-ß, influencing cell
proliferation in normal and malignant epithelial tissues.
c-erbB-2
The
proto-oncogene c-erb-2 (also
known as neu or HER-2) encodes a 190 kD transmembrane glycoprotein similar in
structure to the epidermal growth factor receptor.
c-erbB-2 is a distinct gene but is related to the c-erbB-1 gene (epidermal
growth factor receptor) and v-erb-B oncogene (avian erythrobastosis virus, AEV).
Other oncogenes of AEV include the homologue c-erb-A gene, a steroid receptor
gene encoding a nuclear receptor for thyroid hormone.
In humans both c-erb-A and c-erbB-2 are located on chromosome 17q 21 - 22.
The extracellular domains of c-erbB-2 protein and EGFR are 40% identical in
sequence and both possess two regions rich in cysteine residues which may be
responsible for stabilisation of their three dimensional structure and ability
to bind ligands. Monoclonal antibodies which bind to and down-regulate mutant
c-erbB-2 receptor cancers inhibit tumour cell growth in vitro and in vivo
and over-expression of the normal c-erbB-2 protein in NIH 3T3 cells leads to
transformation. Antibodies to natural human c-erbB-2 have been shown to inhibit
the growth of breast cancer derived cell line SKBR-3 which expresses high levels
of the protein.
These observations imply an important role for c-erbB-2 in at least a subset of
human breast cancers.
The c-erbB-2 gene has been found
to be amplified in 15 - 20% of invasive human breast carcinomas and gene
amplification of over 3 fold appears characteristically to be associated with
c-erbB-2 gene protein localisation on tumour cell membranes. This localisation
can be detected by immunocytochemical techniques.
High frequency of gene amplification of around 50% have been found in ductal
carcinoma in situ (DCIS)
and of over 90% in Paget's disease of the nipple.
In DCIS there is an association with the large cell comedo subtype.
There has been increasing interest in the role of c-erbB-2 oncogene in breast
cancer, particularly its relationship to prognosis.
Over-expression of c-erbB-2 oncogene is now generally accepted to correlate with
poor prognosis in both primary operable and advanced breast cancer patients
and is associated with poor differentiation.
Our knowledge of the function of
c-erbB-2 oncoprotein is rudimentary. The similarities to EGFR and its persistent
over-expression in a significant proportion of breast carcinomas with poorer
prognosis imply an important growth regulatory role. This is further supported
by the observation that monoclonal antibodies raised against the extracellular
domain
have exerted an anti-tumour effect on mutant neu transformed NIH 3T3 cells and
on a human breast tumour derived cell line. In addition it is known that EGFR
expression is associated with poorer prognosis
and one might postulate that both EGFR and c-erbB-2 oncoprotein are both
components of a mechanism responsible for breast tumour development or
progression. One group
has demonstrated that c-erbB-2 oncoprotein can act as a substrate for EGFR
tyrosine kinase and it has recently been demonstrated that a combination of
expression of EGFR and c-erbB-2 more efficiently transforms cells than either
protein alone.
A possible hypothesis of their role in neoplasia or tumour progression is that
binding of ligand to an increasing number of receptors leads to an elevated in
phosphokinase activity which would promote cell replication.
c-erbB-2 protein is found in over
90% of cases of Paget's disease of the nipple demonstrated by membrane staining
using immunohistochemistry.
Apart from a growth stimulatory effect, the molecule may play an important role
in motility of tumour cells by the activity of a motility factor, which acts as
a specific ligand for the neu-protein.
This motility factor is believed to induce chemotaxis of neu-overexpressing
breast cancer and may lead to an increased metastatic potential of
overexpressing breast tumours. Also in Paget's disease of the breast, a motility
factor secreted by epidermal keratinocytes may attract the overexpressing
Paget's cells by chemotaxis and leads to invasion of the epidermis by the tumour
cells.
Breast
cancer associated genes
The development and progression
of a malignant phenotype of human tumours is related to abnormalities of
structure or activity of proto-oncogenes
and/or mutation of tumour suppresser genes such as p53.
Many cellular oncogenes have been found to be activated in breast cancer. Of
these, c-erbB-2 (see above), c-myc and ras have excited the most interest. A
variety of other oncogenes including BRCA1, BRCA2, AD1 and Retinoblastoma have
also been implicated in the genesis or progression of breast cancer and their
products are being actively investigated at present but their relationships to
clinical variables is not yet clear. A greater understanding of the consequences
of such genetic changes and their functions may influence treatment and
assessment of prognosis in the future.
p53
p53 is a tumour suppresser gene
located on chromosome 17p. Mutations of p53 are the commonest molecular
abnormality found in human solid tumours and are found in a high proportion of
breast cancers.
Normal function of p53 is regulated post-translationally and could be influenced
by phosphorylation state, sub cellular localisation and interaction with any
number of cellular proteins.
The range of functions of p53 are cell type specific and appear to be directly
related to the ability of p53 to act a specific transcriptional activator. The
role that transcriptional repression plays in the function of wild type p53 is
less clear. It is possible that p53 has a more direct activity in DNA regulation
and repair.
Numerous roles are described, but in particular p53 appears to have a central
role in cell cycle control after exposure to DNA damage.
Wild type p53 is a negative regulator of cell growth, it is thought by forming
homodimers around DNA, allowing DNA repair to occur before cell division, or if
repair does not occur then inducing cell death through apoptosis. p53 has been
described as the guardian of the genome.
Mutation of p53 is believed to
result in more stable forms of the protein which form ineffective dimers around
wild type p53 and lead to a failure of growth regulation.
Most documented mutations result from a single amino-acid substitution with 50%
of mutations occurring between exons 5-8, which are highly conserved during
evolution.
Mutations are mainly missense type and their frequency and distribution vary
amongst different types of cancer.
p53 mutation is associated with more aggressive biological phenotypes of tumours
and poorer prognosis in breast cancer patients.
Because of its role in regulation of apoptosis and response to DNA damage, p53
status could act as a predictive marker for a response to hormone and
chemotherapy.
ras
The ras family of genes
c-rasH, c-rasK and N-ras are closely related.
They encode GTP binding proteins which act as intracellular messengers involved
in transmitting signals from activated growth factor receptors to the nucleus.
The most widely studied ras protein is the c-rasH p21 oncoprotein which has
sequence homology with the G-protein alpha subunit involved in adenylate cyclase
activation. Single or small numbers of amino acid point mutations of ras genes
can induce cell transformation and have been found in approximately 15% of human
carcinomas
and ras mRNA is over-expressed in a variety of tumours including breast.
The mutations result in gene products deficient in GTPase reactivity and hence
may influence proliferation control.
c-rasH p21 protein expression has
been studied in human breast tumours by immunohistological staining. There are
conflicting results with some groups finding increase p21 expression in
carcinomas
and pre-malignant lesions
and others finding similar expression in benign and malignant lesions.
At present one must conclude that ras genes may play a role in development of
some human breast cancers.
c-myc
c-myc is one of a number of
cellular and viral oncogenes (myc, myb, fos, p53) which code for nuclear
proteins which appear to have a role in embryogenesis and proliferation.
c-myc amplification has been found in up to 30% of breast cancers.
It is a 62 KD protein product found in the nucleus during the G0 to G1 phase of
the cell cycle.
In breast cancer an association has been found between c-myc protein expression
and the histological grade of the tumour suggesting a relationship with tumour
differentiation.
No association has been found with prognosis.
Cell
Cycle Assessment
Many groups have now shown that
an estimate of the proliferative activity of a tumour through whatever means can
give prognostic
and therapeutic
information (see below). It must be borne in mind that true assessment of the
proliferative rate of a tumour can only be determined by combining measurements
of the growth fraction and the cell cycle time through sequential sampling. A
single measurement in time of the growth fraction of a tumour should be regarded
as an index of proliferation only.
The growth fraction can be
assessed by counting the number of cells in mitoses (mitotic index)
but this requires high quality tissue sections. Recently techniques of labelling
cells which are active in the cell cycle have provided alternatives. Cells in
the synthesis phase of the cycle (S phase) will take up thymidine or analogues
of thymidine such as 5-Bromodeoxyuridine (BRDU). Incorporated molecules can be
identified by prior radiolabelling, or in the case of BRDU by
immunocytochemistry or flow cytometry using anti-BRDU antibody. Counts of the
proportion of labelled cells will given an estimate of the S-phase fraction
(SPF).
Two other growth fraction markers
have emerged. The monoclonal antibody Ki 67 was raised against a Hodgkin's
disease cell line and identifies cells active in the cell cycle.
The antigen is of unknown structure and is highly labile. It can be used to
estimate a 'proliferative index' in breast cancers which can provide prognostic
information.
The antibody MIB1, was raised by the same group to portions of the Ki67
molecule, and has the benefit that it recognises a stable part of the Ki67
molecule that can be detected in formalin fixed paraffin embedded tissues
allowing retrospective studies which have also confirmed the prognostic
significance of growth fraction assessment in this fashion.
Antibodies are also available to
other cell to cell cycle related proteins including a 36 kD nuclear protein
named Proliferating Cell Nuclear Antigen (PCNA) or Cyclin which appears in the
cell nucleus in late S phase.
This molecule is an auxiliary protein of the DNA polymerase delta enzyme
and can be identified in standard histological sections.
Potentially it can provide information analogous but not directly comparable
with flow cytometric estimation of SPF.
Other
molecules of interest
The reactivity of numerous
additional antibodies has been investigated in breast cancer. This include
metalloproteases, intermediate filament proteins, basement membrane components,
CEA, alpha lactalbumin, caseins, blood group antigens and many others.
Many of these reagents have not found wide acceptance as routine tests of
clinical importance but some such as Cathepsin D have resulted in some
controversy and others such as p-glycoprotein have potential therapeutic
importance.
Cathepsin
D
The Cathepsins D, B and L are
acidic lysosomal proteinases which are involved in intracellular protein
turnover. Increased levels of Cathepsin D identified by cytosol radio-immuno
assay have been demonstrated in breast cancer and shown to have an association
with indicators of tumour aggression such as large size, high histological grade
and lymph node positivity.
More recently immunohistological studies have demonstrated that Cathepsin D can
be identified not only in breast cancer cells, but also frequently in
accompanying stromal tissue and particularly in infiltrating macrophages.
Some of these studies have failed to demonstrate an independent prognostic
effect,
whilst others have shown a prognostic effect only for stromal macrophage
reactivity.
This evidence suggests that the prognostic effect of Cathepsin D is an
epiphenomenon related particularly to associated inflammation and stromal
macrophage infiltration.
P-glycoprotein
A proportion of tumours of many
types will develop resistance to a variety of chemotherapeutic agents. This
multidrug resistant phenotype is associated with expression of a 170 kD membrane
glycoprotein (P-glycoprotein) which acts as an energy dependent pump removing
certain families of chemotherapeutic drugs.
Antibodies to P-glycoprotein have
been produced which can be used to identify expression in tumour samples. Few
studies have been performed in human breast cancer but development of the MDR
phenotype appears to be a late phenomenon96 and may therefore be of limited
clinical value.
Morphometry
and Cytometry
Objective measurements of the
shape, arrangement and tinctorial characteristics of breast tumour cell
populations can be made using a variety of simple morphometric principles or
more complex computer assisted morphometric,
image cytometric
and flow cytometric equipment.
The type of morphometric information that can be obtained includes cell and
nuclear size and shape, cellularity, mitotic frequency, nuclear chromatin and
nucleolar texture. Use of fluorescent or visible stoichiometric DNA stain allows
measurement of nuclear DNA content by flow cytometry and with image cytometry.
DNA content can be used to assess abnormalities of ploidy
and to estimate the SPF.
Many of the variables measured by these techniques have been shown to provide
prognostic information, albeit of varying importance, in breast cancer patients.
Practical
Applications of Molecular Markers
Recognition
of malignancy
Following the discovery of
monoclonal antibody technology many groups have attempted to identify tumour
specific antigens. This search has been largely fruitless. It is perhaps not
surprising that such molecules are either elusive or do not exist. The
neoplastic cell is a product of a normal cell through a process of
transformation and is likely to retain most of its basic cellular and molecular
structure. Novel antigens generated by this process would most probably lead to
cell elimination by the host immune response.
Some groups have claimed that
certain oncogene products may be increased in malignancy and in preneoplastic
processes and could be used as recognition of these processes.
Some of these claims have not been substantiated by others trying to reproduce
the results.
These techniques used such as immunocytochemistry are interpreted by subjective
methods and can be difficult to quantify and this may explain some of the
conflicts; at present it certainly precludes the use of such reagents in routine
diagnostic situations. Finally this approach, unless combined with other
molecular investigations, identifies specific individual events which only occur
in a proportion of tumours and could not therefore be applied to a routine
system requiring involving all tumours.
However, gross disturbances in
the tumour cell DNA or genome are very unlikely to occur in normal or non
neoplastic cells. Identification of a highly aneuploid stem line by image of
flow cytometry, or detection of high levels of amplification of oncogenes such
as c-erbB-2 by molecular techniques or immunocytochemistry could be used to
identify malignancy in a subset of tumours. These tumours are usually easily
recognised by pathologists in histological sections or cytological preparations
as they tend to exhibit characteristic morphological features of malignancy.
There is far greater need for accurate identification of pre-neoplastic and
malignant tumours at an early stage of development where detection methods and
cytological and histological criteria are less well established or subtle.
At present the only technique
which has shown potential as a routine system to aid diagnosis is image
cytometry. By combined assessment of nuclear morphology, DNA content and
possibly chromatin texture relatively acceptable levels of sensitivity and
specificity for diagnosis of malignancy can be obtained.
Further development of preparative techniques, system hardware and feature
selection criteria are imminent. This technology has the potential to supersede
traditional cytology in diagnosis of breast disease.
Assessment
of Prognosis
Many of the biological processes
or effects which can be examined by the above techniques are related to the
prognosis of a given tumour.
The mechanisms for some associations have an apparently simple basis. For
example, a measure of proliferative activity can give an indication of tumour
doubling time. Other associations are less clearly understood but may relate to
differentiation or events occurring during tumour promotion or progression to a
more anaplastic state.
Clinicians and scientists are now
beginning to appreciate the ability of traditional prognostic factors in breast
cancer, such as tumour size, lymph node stage, histological grade and
histological type, and are able to predict the likely prognosis of a given
individual at presentation.
When combined in a prognostic index such prognostic factors can give a highly
accurate assessment of likely prognosis.
It would be fortunate indeed if a
single molecular event could offer analogous information. This would provide a
relatively simple objective method of assessing prognosis. One must bear in mind
that traditional factors are dependent on a host of variables including the time
a tumour has been present (size and lymph node stage), differentiation (grade
and type), proliferation (grade) and metastatic potential (lymph node stage). In
a study using a series of multivariate analyses designed to establish its
independent prognostic value in comparison with traditional factors (size, stage
and grade), c-erbB-2 protein expression achieved significance only when included
with the time related variables of tumour size and lymph node stage. When the
powerful tumour related prognostic factor, histological grade, was introduced
into the analysis the independent significant of c-erbB-2 protein expression was
lost.
Similar results have been found with EGFR expression
and using antibodies to epithelial mucin.
Such studies do show that if accurate information about tumour grade is not
available such molecular information can provide analogous, although less
powerful, information which could act as a substitute for histological grade.
It is possible to discuss some
reasons for this lack of power by again using c-erbB-2 amplification as an
example, although it is difficult to speculate without precise knowledge of its
function. c-erbB-2 gene amplification is found in only a small proportion of
tumours and for this reason alone it is perhaps not surprising that it fails to
provide prognostic information of a magnitude similar to histological grade. It
has been suggested that amplification and over-expression of certain genes may
be reflected in tumour cell morphology
which has been borne out partly by evidence that c-erbB-2 amplification is
related to large cell morphology, particularly in ductal carcinoma in situ.
Histological grading is assessed by combining the appearance of various
morphological features and mitotic figure frequency.
Thus it provides a summation of a variety of tumour variables. One could
extrapolate further from the above tentative evidence and suggest that
histological grade gives an overview of various molecular events affecting
morphological appearance. It is unlikely therefore that a single molecular event
could compete with histological grade in such a statistical multivariate
analysis. The future of clinical application of molecular markers of prognosis
will be in combination, providing information analogous to histological grade.
Histological grade has been
criticised for its subjective nature and lack of reproducibility in some
centres.
When carried out by enthusiastic pathologists good correlation can be achieved.
Use of guidelines and introduction of semi-quantitative assessment of the
components could improve consistency further. The most important component of
histological grade is the assessment of mitotic frequency. Objective
measurements of tumour cell proliferation such as percentage of cells in
mitosis, thymidine or BRDU labelling index, S phase fraction measurement and Ki
67 labelling index
have been shown to provide powerful prognostic information. Again the
information is analogous to histological grade but also provides a tumour
dependent prognostic variable which can be measured in an objective fashion. All
have some drawbacks; mitotic frequency is time consuming to perform, thymidine
or BRDU labelling index require in vivo or in vitro incorporation and subsequent
measurement of labelling by flow cytometric analysis or assessment of
immunocytochemical preparation and S phase fraction requires flow cytometry
equipment. However, Ki 67 labelling index can be assessed on paraffin sections
using the MIB 1 antibody. It is difficult to justify some of these methods in a
cost conscious Health Service environment when histological grade and tumour
type can be assessed rapidly on a routine histological tissue section, and
provide extremely powerful information about prognosis.
Response
to treatment
There is a need to develop
accurate methods of predicting the response to primary local treatment to
identify at least three groups; those at high risk of distant relapse, those who
will be cured by local modalities and those who will respond to systemic
therapy. Use of a prognostic index using traditional prognostic factors
or one incorporating morphometric or molecular variables
can help identify a group of patients who have an extremely good prognosis,
which is not significantly different from the expected survival of the
non-breast cancer bearing female population. These patients clearly have a very
low risk of significant local or distant recurrence and use of systemic forms of
treatment, which may carry a hazard in themselves, can not be justified as a
routine policy. Similarly a group of patients with a grave prognosis and a high
risk of distant recurrence can also be identified. Use of adjuvant or secondary
forms of systemic therapy can easily be justified in these patients. The value
of hormone receptor assessment in the latter group is discussed in the last
section of the handout.
In addition to the use of
prognostic indices there is evidence that response to chemotherapy can be
predicted in patients with breast cancer through measurement of the
proliferative activity of the tumour. It is widely accepted that tumours with a
very high proliferative rate such as acute leukaemias, high grade lymphomas and
germ cell tumours can respond dramatically to chemotherapy schedules. Similar
although less dramatic behaviour has been reported in breast cancer. A
relationship has been shown between S phase fraction (SPF) and tumour response
in patients with stage II - IIIa disease.
Tumours with a low SPF (<5%) had a response rate of 46%. Those with an
intermediate SPF (5 - 10%) had a response rate of 84% and those with a high SPF
(> 10%) all responded. There was considerable overlap between the groups but
these results are encouraging and supportive data has emerged from thymidine
labelling (TLI) studies on patients receiving adjuvant chemotherapy. Long term
follow up has shown that patients with a high TLI had delayed recurrence in
contrast to patients with a low tumour TLI in whom no benefit was observed. It
is likely that use of such information about growth fraction if combined with
other data such as hormone receptor status and histological information (for
example histological type) could become an important method of assessment of
breast cancer patients.
Detection
of secondary events
Although many markers of tumour
differentiation exist currently there is no widely accepted marker present on
the primary tumour cell which can indicate metastatic potential. Binding of
tumour cells by Helix Pomatia lectin has been found to show an association with
lymph node stage
but the mechanism of this relationship is unclear and the relationship needs
further evaluation. Information about the potential or in reality the existence
of lymph node or distant metastasis can only be obtained reliably through tissue
biopsy. Clinical examination, detection of highly elevated levels of certain
markers in serum and imaging techniques can be used to detect metastatic disease
in more advanced stages.
Deposits over 2 cm in size can
sometimes be visualised by imaging using radiolabelled targeting monoclonal
antibodies
but this method is not sensitive enough to detect small early deposits.
Antibodies to epithelial antigens
such as cytokeratins and epithelial mucins can also be used to detect
micrometastases in excised lymph nodes
and bone marrow aspirate samples
by immunocytochemical staining. This increases the detection rate of metastatic
disease from levels found by routine examination. The long term significance of
such findings is still debated but there is evidence from the few published
studies that patients with micrometastatic disease have a higher chance of
subsequent overt recurrence.
Conclusion
The relatively recent
developments described above are dramatic. They provide mechanisms and evidence
to increase our understanding of the biology and behaviour of breast cancer and
have served to stimulate and bring together scientists, pathologists,
oncologists and surgeons working on this common condition. At present the direct
applications to laboratory and clinical practice are limited but there is no
doubt that such information will eventually be applied more rigorously to the
clinical setting resulting in a greater awareness of their potential in the
management of the individual patient.
|
