|
You are here: Home: BCU 6|2002: Daniel F Hayes, MD
 |
|
|
|
 |
Daniel
F Hayes, MD |
 |
|
Medical Director, Breast Evaluation
Center
Clinical Director, Breast Oncology Program
University of Michigan Comprehensive Cancer
Center |
|
 |
 |
|
Edited comments by Dr Hayes
Understanding the complexities of the estrogen
receptor
What we have learned about the biology of the estrogen receptor
in the last five years is mind-boggling — how it works and
interacts with the other growth factor pathways, like EGFR and HER2,
and how those interactions manifest clinically as well as their
complexity. Every time I think I understand this system, results
from another clinical trial tell me, “Nope, that’s not
the answer. It’s completely different.” The staggering
complexity of this disease makes it so hard to treat.
We have learned so much in terms of endocrine therapy. We now
know about the complexity of estrogen receptor ligands (i.e., estrogen,
tamoxifen, raloxifene), which change the conformation of the receptor
so that it is prone to phosphorylation. This phosphorylation, which
occurs through the peptide growth factor signaling pathway, dimerizes
estrogen receptors that then bind to the promoter of estrogen-sensitive
genes and signal for coactivators and corepressors.
We have always known about estrogen receptor alpha, and now we
have identified estrogen receptor beta. We recognize that the estrogen
receptor actually can bind to a different part of the DNA that does
not have an estrogen response element, called AP-1. We are beginning
to understand why tamoxifen has this interesting duality —
antiestrogenic in some cells and estrogenic in others. Understanding
the biology of the estrogen receptor may help to explain why five
years of tamoxifen might be optimal, why serial hormone therapies
work and why hormone withdrawal elicits a response.
Mechanism of action of the SERMs
Like estrogen, all of the SERMs (tamoxifen, toremifene, raloxifene,
droloxifene, idoxifene) bind to the estrogen receptor. They all
induce phosphorylation, dimerization and binding to the estrogen
response element (ERE) in the promoter of the specific genes. However,
they then signal for different coactivators and corepressors in
the cell.
The response of a specific cell to a specific ligand depends on
a number of things, such as the amount of estrogen receptor alpha
and beta and the types of coactivators and corepressors present.
The response may be even related to the genes that are “turned
on” in one cell compared to another. The cells are primed
to see these ligands as either estrogens or antiestrogens. Therefore,
even though the ligands fundamentally do the same things —
induce phosphorylation, dimerization, binding to ERE — they
can have very different effects.
Can we design new SERMs that are antiestrogens in one place and
estrogens in another? I personally do not believe so, because this
is just too complex to fully understand. Then again, 15 years ago
I said that I would not spend any more time on endocrine therapy,
because I could not believe you could squeeze any more effect out
of the estrogen receptor. I thought we had gotten all we could with
tamoxifen — I was absolutely wrong. So, I am willing to have
smart people like Craig Jordan prove me wrong about the SERMs.
Mechanism of action for fulvestrant
Another endocrine therapy is fulvestrant. Unlike the SERMs, which
induce some biological response, fulvestrant binds to the estrogen
receptor and completely shuts the system down. It prevents estrogen
receptor phosphorylation, dimerization and binding to the ERE. Fulvestrant,
a rationally designed drug, is truly an antiestrogen. It looks like
it will be a step forward. Fulvestrant may be as active as the aromatase
inhibitors, another class of rationally designed drugs.
Sequencing of endocrine therapy
We now have several options for endocrine therapy. The issues
are how, when and in what order we should use these agents. As they
are making it to the clinic, I get phone calls from my colleagues
asking, “What order do I use these in?” I do not think
we know the answer. The challenge for the cooperative groups and
pharmaceutical companies is to conduct trials evaluating sequential
and combination endocrine therapies.
I believe we will find that different subgroups of patients will
respond differently to individual endocrine therapies. Just as we
use ER status to decide who will receive endocrine therapy, in the
future we may use the progesterone receptor, HER1, 2, 3 and 4, or
some of the coactivators and corepressors. These markers may indicate
which patients should receive tamoxifen, an aromatase inhibitor
or fulvestrant. We are a long way away, but I think we will see
it happen.
Hormone dependence versus hormone sensitivity
Why multiple endocrine therapies work sequentially is one of the
conundrums of hormonal therapy in breast cancer. In the last five
years, the concept of hormone dependence, rather than hormone sensitivity,
has jelled in my mind. A cancer may start out as either hormone-dependent
or hormone-independent. The cancer may remain hormone-dependent
for many years, but become resistant to specific endocrine therapies
that have different mechanisms of action.
For example, a patient’s hormone-dependent cancer may initially
be sensitive to tamoxifen. The cancer may later become resistant
to tamoxifen, but may respond to another endocrine therapy like
an aromatase inhibitor. So, the tamoxifenresistant cancer is still
hormone-dependent, and the next endocrine therapy will work. When
hormone-dependent cancers become resistant, they are resistant to
specific drugs.
Mechanisms of resistance to the SERMs
When a hormone-dependent cancer becomes resistant to a SERM, we
are not sure of the exact resistance mechanism. One possibility
is that cells begin to upregulate HER2, an epidermal growth factor
receptor, resulting in constitutive phosphorylation, dimerization
and activation of the estrogen receptor. Then, the ligand has no
effect, because the estrogen receptor is already activated. Another
resistance mechanism might be the mutation of the estrogen receptor
so it becomes hypersensitive to individual ligands. If that is the
case, ligand-based therapy (i.e., the SERMs) might suddenly start
acting like estrogen; whereas, ligand-annihilating or ligand-depleting
therapy (i.e., oophorectomy, LHRH agonists and the aromatase inhibitors)
might still be effective. Even with the upregulated HER2 hypothesis,
it is possible that phosphorylation makes the receptor hypersensitive
to the ligand. In that case again, ligand depletion might be ideal.
Fulvestrant, on the other hand, is a ligand that binds to the
estrogen receptor and prevents downstream signaling. There is a
constant turnover in the estrogen receptor, but the receptor is
completely inactivated by fulvestrant because it cannot dimerize.
Mechanisms of resistance to the aromatase inhibitors
Since the aromatase inhibitors block the peripheral conversion
of DHEA and testosterone to estradiol, then, in theory, resistance
should not develop because they are somatic enzymes in the fat that
are not prone to the genetic instability of cancer cells. But, we
know that resistance does develop.
One possible explanation is that the cancer cells themselves mutate
and produce an abnormal aromatase that converts DHEA and testosterone
into estradiol. This is still speculative.
Another potential mechanism of resistance is that the cells ultimately
become hormone-independent. This is analogous to a car that runs
on gasoline and is retrofit with a solar panel. If the estrogen
receptor is the gasoline tank and estrogen is the gasoline, the
car can still run without gasoline (estrogen) through solar power.
Likewise, another factor may drive the cancer cells that then become
hormone-independent.
A third possible explanation is that the still hormone-dependent
cells become hypersensitive to small amounts of estrogen. If this
were the case, fulvestrant might work when the cells became resistant
to the aromatase inhibitors. Although not yet published, Kent Osborne
has been discussing the results from the trials comparing fulvestrant
to anastrozole.
In the US trial, the duration of response was longer with fulvestrant
than anastrozole. His explanation for this difference in the duration
of response is that anastrozole may reduce estrogen levels by 99%,
and the estrogen receptors then become hypersensitive to that one
percent of estrogen. Fulvestrant simply does not let the estrogen
get to the estrogen receptor.
There are a variety of possible mechanisms for the resistance
to the aromatase inhibitors, and there may be others that we are
not aware of yet. It is important to understand these mechanisms,
because they may dictate how we use these drugs in the next five
years.
Evolution of estrogen independence in breast
cancer
Are all breast cancers initially estrogen receptor-positive with
estrogen receptornegative cancers evolving from those? Or, are there
fundamentally two kinds of breast cancers — estrogen-dependent
and estrogen-independent?
I believe there are fundamentally two types of cells that become
malignant. Studies, by Craig Allred and others, tell us that not
every epithelial cell in the normal breast is hormone-dependent.
We do not know which stem cell is responsible for the development
of any epithelial cancer or breast cancer, but another area of active
research is trying to identify the epithelial stem cell that becomes
cancer.
We will probably find that even before we identify a cancer, there
are cells dedicated to becoming cancer that are either estrogen-independent
or -dependent. Ultimately, those patients with estrogen-dependent
cancer, whom we do not cure, develop estrogen-independent cancer.
But that is probably a late event. The estrogen receptor as an oncogene
Clearly, breast cancer is related to estrogen and the female endocrine
system, but we do not understand it entirely. People have called
the estrogen receptor an oncogene. Are there fundamental defects
in the estrogen receptor that lead to the oncogenic process? In
contrast to HER2, we have not found the classic oncogenic steps
such as amplifications or activating mutations. To my knowledge,
you cannot transfect normal cells with multiple copies of estrogen
receptor and make them cancerous.
It is probably a secondary phenomenon with other things in the
cell producing genomic instability. In the right milieu, changes
in the expression of coactivators result from some upstream change.
Those, then, result in downstream effects that, in and of themselves,
are not oncogenic, but set the cell up to be more responsive to
external stimuli, like estrogen.
Mechanism of action for high-dose estrogen
We are beginning to understand the mechanism of action for high-dose
estrogen. It has always been counterintuitive that the treatment
of choice for breast cancer, prior to tamoxifen and chemotherapy,
was pharmacologic doses of estrogenic-like therapies, such as DES.
Rob Nicholson’s data demonstrate a biphasic response to
pharmacologic doses of estrogen in MCF-7 cells. Without estrogen,
these cells do not grow because they are hormone-dependent. With
modest doses of estrogen, they grow quite nicely. At high doses
of estrogen, they quit growing again. This is consistent with the
clinical observation.
If those cells are preconditioned with different concentrations
of estrogen, there is a similar biphasic response that is shifted
to the right or left in regards to the estrogen concentration. The
cell may have different coactivators and corepressors under one
estrogenic condition. When the hormonal milieu is changed, the cells
reset their coactivators.
In terms of clinical practice, we may learn that patients on hormone
replacement therapy might have a different hormonal milieu when
they are diagnosed with breast cancer. We might want to treat those
patients differently. This concept is not ready for prime time in
2002, but it may be in 2010. In the meantime, we are beginning to
understand the molecular basis of hormone dependence, treatment
and resistance.
Adjuvant aromatase inhibitors
All of us are very enthusiastic about the potential for the aromatase
inhibitors. However, I think we need to be very cautious about overinterpreting
the ATAC trial data and implementing the aromatase inhibitors in
the adjuvant setting.
Why be enthusiastic? Because of preclinical data and data in the
metastatic setting, we believe the aromatase inhibitors are at least
as effective and probably more effective than tamoxifen. The ATAC
trial data fit our bias.
Why be cautious? The downside, I think, are the potential complications
associated with these drugs. The obvious one is osteoporosis.
The ATAC trial is not the only study comparing an aromatase inhibitor
to tamoxifen. There are at least two other trials that are similar
in design. There are also two other trials in which women receiving
five years of tamoxifen are then randomized to an aromatase inhibitor
or placebo. Before we routinely offer all postmenopausal patients
an adjuvant aromatase inhibitor, we need to see those data as well
as more mature ATAC trial data. On the other hand, I am already
using adjuvant aromatase inhibitors for the occasional patient with
a contraindication to tamoxifen — a history of deep venous
thrombosis, stroke/TIA or a tamoxifen allergy.
Select publications
|
