Bcu5 supplement

Bridging the Gap between Research and Patient Care C O N T E N T S
Certified
w w w . b r e a s t c a n c e r u p d a t e . c o m
DR GEORGE SLEDGE SUPPLEMENT
D R N E I L L O V E : Welcome to Breast Cancer Update. This is Dr Neil Love. The last San
Antonio Breast Cancer Symposium contained a wealth of new clinical research
information that is of immediate practical relevance to oncologic practice. One of the first
presentations, by Dr George Sledge, addressed the crucial but poorly understood cardiac
dysfunction that has been observed in breast cancer patients receiving trastuzumab. I met
with Dr Sledge to discuss this and other emerging data sets in breast cancer, and he began
by discussing the background to ECOG protocol 2198, a pilot adjuvant trial in node-
positive, HER2-positive patients.
D R G E O R G E S L E D G E : If you look at patients who are receiving Herceptin, there are
certain things that we know about Herceptin-related cardiotoxicity. First, we have an
enormous amount of data that suggests if you give Herceptin along with an anthracycline
that significantly increases the rate of congestive cardiomyopathy. But, in the initial
Genentech pivotal trial, there was also a group of patients who got Taxol plus Herceptin
typically after they had received a prior anthracycline-based chemotherapy. Those
patients suffered a smaller, but still real, incidence of cardiac events including occasional
cases of congestive heart failure.
So, a reasonable question, moving Herceptin from the metastatic setting into the adjuvantsetting is – How can we do that safely? How can we do that, particularly in a group ofpatients who we frequently give anthracyclines to? The common sense is thatanthracyclines may, in fact, work better in HER2-positive patients. So, ECOG 2198, when it was developed in 1998, was developed to address these concerns.
The way the trial was set up was that patients who had lymph node-positive breast cancerand who were HER2-positive by immunohistochemistry received Adriamycin andCytoxan following Taxol and Herceptin, rather than before Taxol and Herceptin. The ideaat the time was a simple one. If the patient got all of her Herceptin prior to receivingAdriamycin and Cytoxan, then the concern of the interaction of Herceptin andAdriamycin would not exist and, therefore we’d see a lower rate of congestivecardiomyopathy. As it turns out, like many hypotheses in medicine, this one was absolutely wrong. It wasabsolutely wrong for a fairly straightforward reason. What we’ve learned aboutHerceptin, basically in the last year as a result of the work of Dr Brian Leyland-Jones andhis colleagues, is that Herceptin has a considerably longer half-life than we originallythought. So while we had designed the trial so that patients would be off Herceptin forthree weeks before receiving their anthracycline, in fact, what we now know is that threeweeks is not a sufficient clearance period to get rid of all the Herceptin from a patient’sserum.
D R L O V E : This study was randomized, correct?
D R S L E D G E : Yes. It was randomized, but only in a sense that all of the patients in the
trial received Taxol and Herceptin up front. But the randomization was either to receive
or not receive Herceptin after the Adriamycin and the Cytoxan.
But in terms of what I’m talking about, and that’s to say the overlap of Herceptin andAdriamycin, everyone on the study started receiving Adriamycin three weeks after theirlast dose of Herceptin. Based on what we know about the PK of this agent, it is likely thatthose patients still had circulating Herceptin when they received an anthracycline. D R L O V E : But the other thing in terms of that study was, since both arms got Herceptin it
seems like it might have been more helpful to have an arm that didn’t get Herceptin in
order to compare cardiac toxicity.
D R S L E D G E : Well, perhaps. I don’t think that would be unreasonable, though of course,
when we designed the trial in 1998, we didn’t know about the current PK data. Having
said that, I find the results of the trial fairly reassuring and comforting. That’s to say, if
we look overall at all of the patients who received Taxol and Herceptin followed by four
cycles of Adriamycin and Cytoxan, there was a relatively low incidence of congestive
heart failure.
D R L O V E : But I guess my question was, I couldn’t decipher whether or not the incidence
that you saw was actually a baseline that maybe had nothing to do with the Herceptin.
D R S L E D G E : Yes. Our statistical assumption in designing this trial was that the baseline
rate of congestive heart failure was somewhere around one percent, or perhaps ever so
slightly less than one percent because that’s been trial data over the years. We saw a 1.7
percent rate of congestive heart failure for all of the patients entered into the trial. And,
two of the four patients who had congestive heart failure had it in the context of a
myocardial infarction, which presumably is not due to Herceptin. That left us with two
patients out of more than 200 patients entered into the trial who had bona fide congestive
heart failure, which presumably was related to their therapy. That’s, in essence, a one-
percent incidence of congestive heart failure that’s probably appropriately attributable to
the drugs.
D R L O V E : So, what would be your overall take on this in terms of what it means to a
clinician in practice?
D R S L E D G E : Well, I think first off, this is, in fact, the first and, as far as I know, the only
data where we’ve actually looked specifically at what Taxol and Herceptin in the absence
of Adriamycin does to cardiac function. The answer is, it probably does a little. There are
a few patients who had drops in their left ventricular ejection fraction. But it doesn’t do
very much. These patients who had drops in left ventricular ejection fraction had their
LVEFs bounce back by and large, despite the fact that they went onto Adriamycin and
Cytoxan. So, I think that’s pretty reassuring for the practicing physician.
I think the clinician, as always, needs to practice the art of medicine. There are probablypatients who one would want to avoid four cycles of AC and four cycles of Taxol andHerceptin regardless of what order. Certainly, if you look at the patients who ran intotrouble on our trial, virtually all of them had some significant predisposition for cardiacdisease.
D R L O V E : But right now, in terms of treating patients off protocol, the last time I talked to
you, you were recommending strongly against using Herceptin.
D R S L E D G E : That’s absolutely correct. I still am strongly against using it in the adjuvant
setting.
D R L O V E : What about considering Herceptin in the metastatic disease setting? Are there
any situations where, because of prior cardiac disease, etc. that the clinician should not be
using Herceptin?
D R S L E D G E : Well, I think the metastatic setting is quite a different setting, and we’re
literally talking about a life-or-death setting for patients with HER2-positive metastatic
breast cancer. The natural history of a HER2-positive, metastatic breast cancer, both in the
ECOG data set and in the Genentech pivotal trial, is that with standard chemotherapy,
these patients live, on average, about 17 months. That is just a miserable survival.
Herceptin clearly improves survival. It clearly improves survival even for the patients inthe pivotal trial who received an anthracycline plus Herceptin. Even in a group of patientswhere a quarter of them developed a cardiac event, there was still a significantimprovement in survival. This argues to me that, allowing for appropriate safetyconcerns, I think that for the vast majority of patients it’s reasonable to treat them withHerceptin if they have HER2-positive breast cancer.
D R L O V E : In terms a clinical algorithm that you use, do you routinely use Herceptin as
part of your first-line therapy in HER2-positive patients?
D R S L E D G E : Routinely. Yes.
D R L O V E : Because one of the things I’ve heard people talk about is, well, I’m going to use
Herceptin. The question is whether I use it by itself or with chemotherapy. Is that, again,
the kind of algorithm you’ve used?
D R S L E D G E : Well, I think that’s a separate question, which is – Do we give Herceptin
alone or do we give Herceptin in combination with chemotherapy? There are definitely
patients who I will treat with Herceptin alone. There are definitely patients I’ll use
Herceptin plus chemotherapy. Certainly, in the patient who has impaired performance
status or where there are concerns about co-morbidities related to the chemotherapy, it’s
certainly reasonable and appropriate to give that patient Herceptin. The patient who has
what appears to be relatively indolent, small-volume disease, it certainly seems rational to
give that patient single-agent Herceptin. I don’t think one’s lost a whole lot.
D R L O V E : Any take on what happens in the community in that regard in terms of
managing HER2-positive patients? Do you think that most oncologists in practice are
using Herceptin, either alone or with chemotherapy, for first-line therapy?
D R S L E D G E : My sense is that the majority are using Herceptin in combination with
chemotherapy for first-line therapy, though I certainly run into physicians and see
patients who are receiving single-agent Herceptin. Over the last couple of years, certainly
there has been a general trend towards using Herceptin more and more up front in the
metastatic setting, rather than later on.
D R L O V E : What about defining HER2 status? What’s your algorithm for that?
D R S L E D G E : Well, typically, the patient, of course, already comes to me with HER2
testing. If we look at the data that’s been generated from the trials available to us, to my
mind it looks as if patients who are 2+ by immunohistochemistry frequently are HER2-
negative when they’re tested by FISH. Both in the ECOG data set related to the trial I
presented this morning, as well as several other data sets, if you’re 2+, probably only
about a quarter of those patients are positive by FISH. So, certainly, for those patients, I
pretty much routinely do FISH testing.
If the patient is a 3+ by immunohistochemistry and it’s from a lab where I trust thepathologist, I think the data is suggestive that being 3+ by immunohistochemistry andbeing FISH-positive probably give one fairly similar results in terms of clinical outcomeand likelihood of response. So, I’m comfortable with either a 3+ by immunohistochem-istry or FISH positivity in that setting.
D R L O V E : What about the patient who’s zero or 1+ by IHC. We know that some of those
patients are going to be FISH-positive. Do you test any of those patients?
D R S L E D G E : I do, not regularly, but certainly in some cases. I view HER2-positive breast
cancer as a breast cancer that has a fairly specific phenotype, and that’s to say it’s more
likely to be a steroid receptor-negative disease. It’s somewhat more likely to occur in
younger women, yet somewhat more likely to be associated with early relapse after initial
diagnosis. So, if I have a patient who fits that phenotype, and yet has, say, a two- or
three-year-old immunohistochemistry test, saying that the patient was negative, I’m quite
happy re-testing that patient, and typically I’m going to re-test that patient with FISH.
D R L O V E : And if they’re positive by FISH?
D R S L E D G E : Then I’m going to treat them. Now, I would hasten to add that we really
don’t have any clinical data on that group of patients, the patients who are immunohisto-
chemistry-negative, but FISH-positive. But I think what we know about the biology is
fairly straightforward. If you look at frozen sections where there are no questions about
fixation artifacts that we have to deal with for paraffin-embedded sections, it is
astonishingly rare to have cell-surface overexpression of the HER2 receptor without
having amplification at the nuclear level for HER2. So it, to my mind, only makes sense
that the FISH-positive patients are going to be the patients who are going to have ample
amounts of receptor on their surface that we’ll be attacking with the antibody.
D R L O V E : Any thoughts about some of the data that’s come out in the last year in terms
of quality control both IHC, coming from the NSABP and Intergroup suggesting that
some patients out there who are labeled as 3+ really aren’t? And then, also, more recently
I heard from Edith Perez, a few patients are labeled as FISH-positive, who aren't. Any
thoughts on that?
D R S L E D G E : Well, we clearly lack quality control for both FISH and immunohistochem-
istry. Clearly, there are laboratories that have done it for a long time, have done it well,
and that you can trust. Clearly, there are laboratories that you probably can’t. I think this
is highly analogous to the situation with estrogen receptor, which I’m sure, since you
come from an era similar to mine, remember a period during the mid to late ‘70s, whereone simply couldn’t trust estrogen receptor values in many cases. And I suspect thatHER2 testing in the year 2001 is very similar to estrogen receptor testing, say in the year1975 or 1976. We desperately need improvements in quality control. D R L O V E : That’s a great analogy because when I think about it, again, one of the
similarities is the relatively benign toxicity profile of Herceptin, as you see with hormonal
therapy. I get concerned about people not being treated who maybe really are HER2-
positive. It’s a shame to see somebody treated who shouldn’t be treated, but it was the
same kind of thinking in terms of ER assays. And there were a lot of women over the
years that truly had ER-positive tumors and didn’t get treated.
D R S L E D G E : That, of course, turns out to be astonishingly important in the adjuvant
setting. One wonders how many patients around the world have died as a result of bad
estrogen receptor testing and, if adjuvant Herceptin ends up having a similar benefit to
adjuvant tamoxifen, we will run into exactly the same problem. I think it’s a real concern.
D R L O V E : When I saw you downstairs at the San Antonio symposium, you told me that
you were really excited, and this is one of the best ones you’ve seen in years, or the most
exciting information?
D R S L E D G E : Yes. This was an absolutely fascinating morning for the San Antonio Breast
Cancer Conference and I think it would match up with any morning that I’ve seen in
recent years in any breast cancer group of presentations.
We had Trevor Powles giving us the adjuvant clodronate trial, a multi-national trialinvolving some 1,000 patients who were randomized to receive or not receive thebisphosphonate, clodronate. That trial is showing a survival benefit for patients receivingadjuvant clodronate, albeit a marginal survival benefit. We had a much larger and very impressive trial, the ATAC trial, in the adjuvant setting,that suggested for the first time that adjuvant aromatase therapy might prove superior toadjuvant hormonal therapy with tamoxifen, albeit with very early follow-up. Andfascinatingly, that the combination of the two was no better than adjuvant tamoxifen.
There’s some fascinating biology there that I think remains to be teased out over the nextfew years.
The ATAC trial was fascinating from a number of other standpoints. First, there was amarked reduction in the development of contralateral breast cancers for patients receivingArimidex compared to tamoxifen — not an entirely expected finding, but a fascinatingone.
D R L O V E : Particularly compared to tamoxifen.
D R S L E D G E : Yes, particularly compared to tamoxifen, which we already know reduces
the incidence of contralateral breast cancers by up to half in some trials. This was a
significant additional benefit on top of tamoxifen.
And, indeed, I think, in looking at those results, one has to wonder, long-term, what theresults of the STAR trial will mean to us when they become available in the next fewyears.
D R L O V E : I was thinking about that, too. As a matter of fact, I asked a couple of the
people involved in the STAR trial what they thought about that, and I think they’re kind
of processing it. But you mean the question of whether or not comparing tamoxifen and
raloxifene is going to maybe be a moot point someday?
D R S L E D G E : If we compare two SERMs – and I think the expectation of many of us
looking at that trial, is that the differences in that trial will relate more to toxicity than
efficacy. If that totally unwarranted prediction turns out to be the case – and we now
have evidence that aromatase inhibitors are superior to tamoxifen – then, to a certain
extent, the results of the STAR trial may be less relevant in the long run than I think many
of us had hoped when the trial was initiated.
D R L O V E : Of course, it’s a long way from seeing some contralateral data to a prevention
trial, probably five years, at least.
D R S L E D G E : Indeed. Though I would hasten to add that the justification for doing the
breast cancer prevention trial in the first place was the data in terms of contralateral breast
cancers, and that the relative risk reduction seen in the adjuvant trials for contralateral
breast cancers have been pretty much identical to the benefits seen in the prevention
setting. My suspicion is that one is translatable to the other, though, again, we don’t have
data, as you point out.
D R L O V E : Yes. That contralateral data was absolutely shocking, particularly when you
see the shape of the curves.
D R S L E D G E : It was striking and immediate, and huge in terms of the difference.
D R L O V E : Even if it had the same effect, later proved, five years, to have the same effect.
The other issue that we saw with tamoxifen is people just don’t want to use it, I think
some of it’s even psychological, because of endometrial cancer and thrombosis. That just
doesn’t fit into the mindset of prevention. Of course, we’re not seeing that, at least at this
point with anastrozole.
D R S L E D G E : In the ATAC trial there appeared to be a lower rate of deep venous
thromboses and there appeared to be a lower rate of endometrial cancer. There appeared
to be a lower rate of some quality-of-life things, such as hot flashes. I think, from a
toxicity standpoint, the aromatase inhibitors may prove to be better than tamoxifen.
Now, the major qualifier here with the ATAC trial, of course, is that we have very brieffollow-up, a median, as I recall, of somewhere around 29 months. So, we’re going to needa lot longer follow-up to make any definitive recommendations based on it. But, certainly,it represents a fascinating new avenue in terms of our therapy, not only in the adjuvantsetting, but also in the chemoprevention setting.
I suppose one possible synthesis of this morning’s meeting is that down the road, we’regoing to be giving patients aromatase inhibitors and then having to give thembisphosphonates in addition, not only for their potential benefit in terms of reducingrecurrence, but also perhaps to prevent the fractures they get with the aromatase inhibitor.
D R L O V E : Anything else happening at this meeting or papers being presented, data that’s
coming out, or even anything in the last few months, in terms of breast cancer, that’s got
you excited?
D R S L E D G E : Well, I think there are several other interesting presentations at this meeting.
Perhaps the most interesting one was a presentation by the NSABP that looked at the
breast cancer prevention trial. Follow-up of that trial.
D R L O V E : That was interesting.
D R S L E D G E : The data that had been presented to us in the past regarding the P-1 trial
was data that primarily looked at invasive cancers and non-invasive cancers. The NSABP
has now gone back and looked at pre-malignant changes in the breast, some of which are
quite benign, such as cystic disease and some of which are clearly on the road to breast
cancer, such as hyperplasias both typical and atypical, and fibroadenomas.
What is seen in this analysis of the P-1 trial is that there appears to be a fairly generalizedreduction in pre-malignant changes of the breast for women receiving tamoxifen as achemopreventative agent. What was particularly striking to me, however, was the age-relatedness of this reduction. And, again, this represents a fascinating piece of biology.
There was a huge reduction in terms of these pre-malignant events and, indeed, perhapsnon-malignant or non-pre-malignant events in the younger women and little or noreduction in many of the categories for the older women. I find this an absolutelywonderful finding. If we look from a toxicity standpoint at tamoxifen as a chemopreventative agent, it’s fairlyclear that the greatest degree of toxicity occurs in older women. Problems such as deepvenous thrombosis, thromboembolic events in general and endometrial cancer primarilyrepresent problems for older women. So, using a chemopreventative agent such astamoxifen in a younger woman, combining both the toxicity data plus the data that weheard this morning, now becomes a much more appealing thing.
D R L O V E : That was an amazing presentation. Any thoughts about why there was such a
difference between the age groups?
DR SLEDGE: Well, to a certain extent, it represents the fact, I suspect, that younger
women’s breasts are in constant motion compared to an older woman’s breasts. We
certainly see higher frequencies, for instance, of fibrocystic changes, of cysts shrinking and
swelling related to periods in younger women. And certainly, problems such as cyclic
mastalgia are virtually entirely problems of younger women as opposed to older women.
So, perhaps not surprisingly, those younger women are the women who get more breast
biopsies for symptomatic problems. But there’s something probably a little bit deeper going
on here, as well. Something that may reflect much more basic underlying biologic changes
where something of major importance occurs when a woman goes through menopause.
D R L O V E : To me, there were two aspects of the presentation. One was that the benign
breast disease changes, like cysts and even things that we don’t even associate with
increasing the risk of breast cancer, were decreased again, mainly in younger women.
D R S L E D G E : Indeed.
D R L O V E : The other was these pre-malignant, even atypical hyperplasia being decreased
again by tamoxifen. It also made me think a little bit about the debate that went on when
the P-1 study first came out about is this “true prevention” or just treating early disease?
Now we’re seeing less atypical hyperplasia, less pre-malignant change. I would assume,
biologically, the implication is that maybe this is true prevention.
D R S L E D G E : Absolutely. I think that is the only read of this data. It’s probably also
worth pointing out that, while we think we have some handle now on pre-malignant
changes, we pretty clearly don’t have a very good handle on it. And while we have
thought of breast cancer as, in essence, typical hyperplasia followed by atypical
hyperplasia followed by DCIS followed by invasive cancer. In fact, that fairly straight
route which we’ve mapped out in our minds may not be quite so straight, and that there
may be a number of tributaries coming into the main stream. It may be that the more of
these tributaries you shut down, the less a flood heads in the direction of breast cancer.
That may well be one implication of what we heard this morning.
D R L O V E : The other thing that happened this morning that was interesting was when
Nancy Davidson updated her trial, the Intergroup trial, of adjuvant ovarian ablation-
tamoxifen. Maybe you’ve seen that update before, but I found that some of the newer data
was kind of interesting. I think the whole thing kind of begins to start to integrate
together.
D R S L E D G E : Yeah. It begins to tell a pretty good story all the way around. Part of the
story it tells us is the continuing importance of hormonal manipulations in breast cancer
for American oncologists, in particular. I think our European colleagues have always had
a different view of this. But American oncologists, for a long time, I think, felt that
hormonal therapy was of secondary importance in the treatment of breast cancer, and that
what was really important was getting in good chemotherapy drugs in good, tough,
strong doses. It’s absolutely clear from today, as it’s been clear to many of us for several
years, that the kinder, gentler approach aimed at the biology of the tumor is very
important. And that the better we get at shutting off the estrogen-receptor pathway, the
better the advantage our patients have in terms of disease-free and overall survival.
D R L O V E : Getting back to Nancy’s presentation, any overall take-away that you have
from this update?
D R S L E D G E : Well, if we look at this trial, which was designed in the late 1980s and
randomized patients with node-positive, estrogen-receptor-positive, premenopausal
breast cancer to receive either CAF, a standard chemotherapy at the time, CAF followed
by Zoladex, or CAF followed by Zoladex and tamoxifen. This trial was designed at the
time where many thought that tamoxifen would have no advantage in premenopausal
patients. So, the obvious fourth arm, which was CAF-T, was not incorporated.
Allowing for that qualification, what these data suggest is that Zoladex probably has itsgreatest effect in the youngest group of women, who are, say, under the age of 40. That,looking at an unplanned retrospective subset analysis, which, of course, is always fraughtwith danger, the suggestion of that analysis would be that shutting off a women’s ovariesin the setting where chemotherapy has not already shut if off, for a young woman, isprobably a good thing.
D R L O V E : Is that something you’ve been doing?
D R S L E D G E : Well, I’ve been privy to this data through ECOG for the last year. I will say
that in my own practice, I have tended to offer more ovarian ablation to younger women,
certainly, than I was doing three or four years ago.
D R L O V E : It’s easy to just pick out the younger women. But isn’t the real issue whether
or not they’re still premenopausal endocrinologically regardless of their age? If you have a
38-year-old woman who stops having her period after chemotherapy then she’s
postmenopausal. But is she’s 44 and is still having her periods, I mean, what’s the age
really have to do with it other than a statistical thing?
D R S L E D G E : Well, it partly is statistical. Partly, presumably, is the fact that for many
women – well, for a much higher proportion of women who are older and closer to their
natural menopause, chemotherapy is much more likely to shut down their ovaries. Even
for the women whose ovaries are not shut down immediately by chemotherapy, if they’re
closer to their natural menopause, they’re probably more likely to become estrogen-
deprived on the basis of natural progression through menopause in a short time. So, I
think that’s what underlies the statistics. But I think your point is extremely well taken.
If you are a premenopausal woman and if we’re asking the question – What’s most likely
to benefit that woman? What’s most likely to prevent her cancer from coming back? The
simple answer is that the best that we can do to deprive her tumor of estrogen, the more
likely we are to see real benefit. And, again, we’re most likely to see the greatest benefit in
that group of women who have the most estrogen for the longest period of time. That’s
the younger women.
D R L O V E : Well, the other thing that was really a neat curve, and I don’t know maybe
she’s showed that before, but I don’t remember ever seeing it was just the CAF treatment
alone arm – those who became menopausal from CAF and those who didn’t. It was a
major difference. Had you seen that before?
D R S L E D G E : I had not seen that before even though those data have been presented at
the ECOG group meeting. There’s a couple of different possibilities to that. One is the
obvious one, which is that estrogen deprivation is good for you and that being
menopausal is a sign of that.
The other possibility is that this may be confounded by some other effect. For instance, ifthere are different circulating levels of Cytoxan, Adriamycin and Fluorouracil, differentpharmacokinetics for those drugs in two different patients. So that one patient hasrelatively higher circulating dose levels of the agent inducing the menopause and theother patient has relatively lower dose levels, well, menopausal status may be a markerfor the pharmacokinetics, rather than menopause causing the benefit itself. That’s apossibility. I would certainly prefer the one you mentioned to, in essence, what’s an intra-patient dose intensity argument.
D R L O V E : Let me give you a clinical question. You have a patient who’s got five or eight
positive nodes; she’s strongly ER-positive. She’s 38. She’s gotten her ACT.
D R S L E D G E : Mm-hmm.
D R L O V E : Even though she’s ER-positive.
D R S L E D G E : Well, why don’t we say that she’s received some combination
chemotherapy.
D R L O V E : She’s gotten her chemotherapy and she’s just having absolutely normal
menstrual periods.
D R S L E D G E : Yes.
D R L O V E : So, now what do you do?
D R S L E D G E : I’m likely to recommend to that patient that she undergo some ovarian
ablation procedure, be it surgical or medical.
D R L O V E : Okay. She says, fine. Then what? Anything else?
D R S L E D G E : Well, I don’t think it’s unreasonable to put that patient on tamoxifen, as
well. Certainly, the group of patients who did the best overall in the overall trial analysis
were the patients who received chemotherapy plus Zoladex plus tamoxifen. So, I would
feel comfortable putting patients on all three.
Now, the related issues that, of course, come up, is that, if one goes out of one’s way toinduce a premature menopause in that patient, then I think one has responsibilities to thatpatient in terms of trying to maintain the patient’s general health, such as looking at thepatient’s serum lipids and trying to do things to improve the patient’s bone mineraldensity status. So, I think, if one is going to offer the patient this approach, one needs tooffer the entire package.
D R L O V E : If, though, after you process the ATAC trial for more than four hours and are
now starting to use anastrozole as adjuvant therapy, do you now look at this woman and
say, well, she’s postmenopausal?
D R S L E D G E : Well, I’m uncomfortable doing that. The ATAC trial was a trial in
postmenopausal women. I think, again, along the lines of paying one’s dues, one needs to
re-investigate the same hypothesis in premenopausal women. I think it’s a hypothesis
well worth investigating. I think if we were to look at the question – Does an LHRH
agonist need to have an aromatase inhibitor added to it to get maximum benefit? Framing
the question that way, I think it’s a very reasonable and testable hypothesis, and certainlyone that we need to test.
D R L O V E : Anything else happen at this meeting? Any new things in the last couple of
months, or new trials that are starting that are getting you excited?
D R S L E D G E : Well, in the Eastern Cooperative Oncology Group, in the very near future,
we’ll be starting a trial that I think points us in a different direction. My colleague, Dr
Kathy Miller at Indiana University will be chairing this trial through ECOG. The trial will
be taking patients who have front-line, metastatic breast cancer, and randomizing them to
either receive weekly Taxol, which is, of course, a commonly used therapy, or to receive
weekly Taxol plus a recombinant humanized monoclonal antibody to vascular endothelial
growth factor. This is the first large trial in the setting of breast cancer that will be looking
at an anti-angiogenic agent in a front-line regimen. Hopefully, this will give us a fairly
good idea in the near future as to whether or not these anti-angiogenic agents are going to
offer something significant.
Now, I’ll also mention that in refractory disease, the recent completion of a trial ofcapecitabine plus or minus the anti-VEGF monoclonal antibody. So, in the very next fewyears, we’re going to be, in addition to hormonal therapy and in addition to HER2-directed therapy, we’re going to be able to see whether or not another molecularly-targeted biologic therapy – that’s to say, anti-angiogenic therapy – will add somethingsignificant. That, I think, is another very exciting direction for us.
D R L O V E : What’s the name of the anti-VEGF?
D R S L E D G E : The trade name is Avastin.
D R L O V E : You would think since the protocols have finished accrual, that the results will
be available fairly soon?
D R S L E D G E : Well, this was a trial in anthracycline- and taxane-refractory patients
receiving capecitabine. So, certainly one would expect, no later than a year or so from
now, relatively mature data on a relatively large number of patients.
D R L O V E : What was the rationale to then go look at Taxol in addition to that?
D R S L E D G E : Well, the rationale for Taxol, aside from moving from a refractory setting to
a front-line setting, is based on something that we’ve learned about the taxanes in recent
years. That’s to say that in a number of preclinical models, taxanes are, in fact, pretty
good anti-angiogenic agents themselves. They are fairly good at killing vascular
endothelial cells in vitro. They affect capillary tubule formation. They affect capillary
migration in preclinical models. In a number of preclinical models, when one treats a
human tumor with a taxane, one sees decreased micro-vessel density. So, in fact, to a
certain extent, we may have been using anti-angiogenic therapy for some time by giving
patients our standard chemotherapy agents, such as the taxanes.
The hope is that by combining an agent that has both anti-cancer and anti-endothelial cellactivity with another anti-angiogenic agent, one would get an additive or synergisticeffect. Indeed, in some preclinical model systems, there have been synergistic effectsagainst endothelial cells when one combines a taxane with anti-VEGF antibodies.
D R L O V E : How about with capecitabine? I know at one point you had raised the
question whether capecitabine is somehow tied into anti-angiogenesis?
D R S L E D G E : Capecitabine represents a similar issue. Thymidine phosphorylase, which is
the enzyme that metabolizes capecitabine to fluorouracil intra-tumorally, is also a pro-
angiogenic agent. There is some intriguing preclinical data that suggests that capecitabine
may itself have some anti-angiogenic activity. So, again, certainly the potential there
exists that by combining a chemotherapeutic agent with an anti-angiogenic agent, to a
certain extent, we’re combining two anti-angiogenic agents.
D R L O V E : I’ve seen some of the data that Kathy presented in terms of – I guess it’s a
phase II study. There actually have been real partial responses with this agent.
D R S L E D G E : Yes, not only partial responses, but we’ve also seen, in fact, a complete
response.
D R L O V E : Really?
D R S L E D G E : To anti-VEGF monoclonal antibody.
DR ANTHONY HOWELL SUPPLEMENT
D R N E I L L O V E : Dr Sledge commented on one of the most discussed new data sets in
breast cancer clinical research — the initial results of the ATAC adjuvant trial, comparing
anastrozole to tamoxifen to the combination. I met with Dr Tony Howell, a faculty
member of the recent Miami Breast Cancer Conference, and a key investigator in this
historic study. Dr Howell has been a pioneer in endocrine therapy of breast cancer and
later on in our conversation, he provided an update on fulvestrant, a fascinating new
addition to the hormonal armamentarium. But he began our conversation by tracing the
background to the development of the ATAC study.
D R A N T H O N Y H O W E L L : Clearly the idea of anastrozole for adjuvant therapy came in
because it was being successful for advanced disease. It actually came through
extraordinarily quickly because there wasn’t any phase II study. It went straight through
from phase I to phase III in second-line advanced disease. And, then clearly an adjuvant
trial had to be designed. And it was actually designed by Michael Baum and, I think,
Joan Houghton, on the back of an envelope. It was a three-arm study, obviously –
anastrozole, tamoxifen and the combination.
A lot of people complained about the combination because there was no data on this atall, and no rationale really. But despite the complaints, the trial went ahead as a three-armstudy. As far as the combination arm is concerned, to my mind, what we found wasentirely predictable. In animal models, in the immature rat uterus assay, for example, ifyou have a low estrogen environment, tamoxifen stimulates uterine growth. If you have ahigh estrogen environment, tamoxifen acts as an anti-estrogen. So, why wouldn’t thathappen in patients? And I guess my suggestion is that we’ve got tamoxifen working in alow-estrogen environment. Under those circumstances, it’s slightly more likely to be anagonist than an antagonist. Another reason, which is quite possible, but we won’t know until we get the receptordata, is the fact that tamoxifen does not seem as active in low estrogen receptor conditionsas anastrozole or letrozole, at least from the Matt Ellis data. If you look at the responserate according to the Allred score in his data, you see that tamoxifen only works with anAllred score above three. Whereas, in fact, letrozole works with an Allred score of 2, 3,and 4, as well as the higher scores. So, that may be another reason why tamoxifen is not aseffective as anastrozole. Anastrozole may be working at low estrogen concentrations, andclearly, the data that will come out of the biological sub-study within ATAC are going tobe extremely interesting.
D R L O V E : But this thing about aromatase inhibitors working with low levels of estrogen
better than tamoxifen, does that have anything to do with why the combination arm
didn’t work as well?
D R H O W E L L : I think that is the likely explanation. Let’s get it straight. The combination
arm is very similar to tamoxifen. What we’re just talking about is a small reduction in
activity compared with tamoxifen, which is definitely nowhere near significance. But you
have to explain why it is slightly worse than tamoxifen, and of course, you have to
explain why it is definitely worse than anastrozole. And so the slight worsening
compared to tamoxifen, is to my mind, likely to be due to the fact that tamoxifen is more
likely to be an agonist in a low estrogen environment. Whereas, then you have to explain
why tamoxifen and the combination are not as good as anastrozole, and that may be
because anastrozole seems to – or letrozole, at least – but the aromatase inhibitors seem to
work, for example, in HER2-positive tumors, and they work in tumors with low levels ofestrogen receptor from the Matt Ellis data. Although, clearly, we need to confirm thosedata that Matt produced, because they are extremely interesting and, if true, give us theexplanation of why one set of drugs, the aromatase inhibitors, are better than another setof drugs, the SERMs.
D R L O V E : I assume you’re going to be looking at HER2 status in the ATAC trial?
D R H O W E L L : Mitch Dowsett is heading up the biological sub-committee of the ATAC
trial, and he will be looking at that.
D R L O V E : Will you be able to look at Allred scores in terms of ER?
D R H O W E L L : We will look at ER, and Mitch will be doing that. I think he does the Allred
score. I’m not totally convinced that the Allred score is the right way to look at it. We, in
Europe, tend to look at the numbers of cells positive, rather than their intensity, although
there are also groups in Europe who look at intensity particularly the Cardiff and the
Edinburgh groups. So, there’s a variation in how receptor is looked at in Europe as there
is a variation across the United States. I agree with Kent Osborne when he says that the
variation is extraordinarily worrying.
D R L O V E : Yeah. Particularly as we get better and better hormonal therapy, it gets more
and more worrying.
D R H O W E L L : We are just analyzing our own data on 900 women who have had adjuvant
tamoxifen or not. And we find that the low estrogen receptor, in terms of percentage of
cells, naught to 20 percent, that we see responsiveness, an effect of tamoxifen in the
adjuvant situation, in a low-receptor environment, as did Craig Allred. So, that’s a little
bit against the Matt Ellis data. But clearly, we need to look at measuring low levels of
estrogen receptor, because my gut feeling is that, if any receptor is there, it’s a positive.
And clearly, you can miss a very low positive quite easily. Patients are not being put on
adjuvant therapy, who should be being put on adjuvant therapy.
D R L O V E : I’ve heard it estimated that there may be 20 percent of breast cancer patients
being diagnosed as having, in one way or another, a false-negative ER. That’s pretty
scary.
D R H O W E L L : I think that’s pretty high. But in the U.K., we have a quality control system,
which is pretty rigorous. We have 200 laboratories in a quality control system, but there
are laboratories that are not in the system, and they can just do what they like. In any
small breast unit in the country, the pathologist with no experience can give an answer to
an ER assay, which may be correct or may be incorrect. It’s a big, big worry.
D R L O V E : I guess from my take as a clinician, as I look at this debate about ER, I see two
things. First is the actual performance of the test. There’s a lot of data out there, and you
were just talking about some of it, questioning whether or not the test is getting done
correctly. Secondly is the interpretation of the test. Even here at this meeting, when we
polled the audience, we also did a survey of oncologists and surgeons before the meetingand asked – How do you define ER-positivity? The most common answer was if it wasabove a level set by the lab. We’ve looked around and labs will set a level anywhere from10-20 percent of cells. Yet a lot of other people – and I don’t know what your take is – willconsider adjuvant hormonal therapy if there’s any ER. Is that the way you approach it? D R H O W E L L : Yes. I think that was quite an interesting answer that most of your people
coming to the conference would take the level that was decided by the lab. Then you
have to ask – How does the lab come by a level? I guess they might look at Allred’s paper
or they might look at other papers. There aren’t very many out there. There really aren’t
many out there on which to gauge a level.
So if they take the lab level, they’re taking something, very often, out of the air. And Ithink it will come down to, if you’re certain about it, any ER-positive is positive. This wassaid years ago by an investigator in London, Charles Coombes, who published showingthat any positive was indicative of response in advanced breast cancer. So, that thenmakes it very difficult, because one cell positive, is it positive or is it a false-positive? Is ita bit of dirt? Clearly, we need to know the answer to that question, and we actually don’tat the moment. We need – and we will get – much more investigation. For example, the ATAC trial. We’llhave receptor measurements in that, and it is, as you well know, in vogue to havebiological information about every trial that we do, and clearly we will be able to answerthe question in the future. But, at the moment, there’s a high degree of uncertainty.
D R L O V E : There still are women in the ATAC trial, who are either ER-unknown or ER-
negative. Is that correct?
D R H O W E L L : That’s extremely embarrassing. There are women in the ATAC trial, who
are ER-negative, about eight percent. And there are women in the ATAC trial who are still
ER-unknown. In fact, we know the ER status is positive, ER+ and/or PR+ in over 80
percent of the women in the ATAC trial. So that is reassuring. If it were lower than that,
one would be worried about the veracity of the results in the ATAC trial.
In fact, the positivity rate in the ATAC trial was lower than that, particularly becausewhen the trial was started there were many centers in the U.K. that actually did notmeasure receptors on the patients going into the trial, because the protocol allowedreceptor unknown. Clearly, that was a mistake. What happened was that AstraZenecawent back to all these labs and got the blocks from most of the patients and measuredestrogen receptor, so that over 80 percent positive is reasonably reassuring.
D R L O V E : The interesting thing was that, even if you include the negatives and the
unknown, there was still a statistical benefit in terms of disease-free survival.
D R H O W E L L : Yes. That’s right. There is a benefit in the intention-to-treat analysis of all
patients, which is about 18-percent reduction by anastrozole compared to tamoxifen. But,
of course, that goes up to a 23-percent reduction as the hazard rate when you just take the
ER-positive population. It’s interesting in the contralateral breast, you can predict how
many contralateral breast cancers there would be in 3,000 women in one arm. There
would be 60, if you didn’t treat the women. You can see how many you get withtamoxifen, which is 30, which is halved, and that’s what we expect with contralateraltumors. In the anastrozole arm there are only nine, so that’s about an 80-percent reductionin tumors in the anastrozole arm, which is really, really exciting. If that is translated intopreventative therapy that augurs well for the prevention of breast cancer in the future.
D R L O V E : In fact, that’s about to go into trial as I understand with the IBIS II study?
D R H O W E L L : The IBIS II trial is, at the moment, a three-arm preventive study to follow
on from IBIS – so-called IBIS I. IBIS stands for International Breast Cancer Intervention
Study, and the IBIS I study has 7,000 patients. The IBIS I study is like the other three
tamoxifen trials – a comparison between placebo and tamoxifen. We have 7,000 women in
the study. It closed to accrual over a year ago, and it’s analyzed at the moment.
The IBIS II study will be in postmenopausal patients because it’s anastrozole versussomething. When we get the feel from our own investigators and from what opinionleaders think, we will then design the final IBIS II protocol. Now, it may be a three-armstudy, which is anastrozole versus tamoxifen versus placebo, or it may be anastrozoleversus placebo, or anastrozole versus tamoxifen. We, at the moment, are in some doubt.
But we have geared up the study to start in September-October, and clearly there’senormous enthusiasm for this study because we’ve had investigators around the worldwho want to buy in, including some of the people who are already in the Americancooperative groups.
D R L O V E : Has there been any talk about possibly having an arm with anastrozole and a
bisphosphonate?
D R H O W E L L : That’s a very important question. The real worry, as far as anastrozole is
concerned, is twofold. One, is its effect on bone, and you’ll know from the ATAC trial,
there was an increase in fracture rate. It was probably not as big an increase as it looks
because in the control arm they were given tamoxifen, and tamoxifen prevented some
fractures. But we estimate that half of the fracture rate in the ATAC trial is due to the
negative effects of anastrozole. So, clearly, we need to have a bisphosphonate in IBIS II,
and that also is under discussion at the moment, because there are some who wish to give
everybody a bisphosphonate, and there are others who say, well, that’s over-treatment of
a large part of the population.
D R L O V E : Why can’t you just do bone density measurements?
D R H O W E L L : There is, at the moment, a bone sub-protocol, which has it in the first 1,000
patients have sequential bone density measurements. Then, the remainder of the patients
should have bone density at the start, to know whether they are low and, if they are low
at the start, at the moment the thinking is that they should go in the study, but they
should be on bisphosphonates, as well. And we think that we need to work out the
bisphosphonate issue and, in the end, I think it will come down to the fact that most
patients will need serial bone density measurements.
D R L O V E : I’ve heard some people say, looking at the side effects and toxicity profile in
ATAC, that if the efficacy had been exactly the same as tamoxifen, it still might be worthusing anastrozole.
D R H O W E L L : I haven’t heard that said, and that is extremely interesting, and I would
agree with you. The thromboembolic complications are much lower. Endometrial cancer
is much lower, and so that would make it an attractive drug from that point of view. The
real problem is how to deal with the bone problems. But you could even say that that is a
positive because these are postmenopausal women who might develop bone problems
even if they haven’t developed breast cancer. So, this nicely focuses on bone, which is a
problem for a large proportion of postmenopausal women. So, it’s almost like you could
see the bone “problem” as an advantage.
D R L O V E : Also, in terms of anastrozole for adjuvant therapy, another unexpected benefit
of having to focus on bone is, if the women do end up getting bisphosphonates, you
might end up with a benefit in terms of metastatic disease. Trevor just presented his
information, which is now the third clodronate study. What did you think about those
data showing fewer bone mets and actually better survival in women who got clodronate?
D R H O W E L L : That is very interesting. The clodronate data are also difficult, but to hear
those data from Trevor at the San Antonio meeting presented in exactly the same session
as the ATAC results makes it really interesting. In the same session, you have the very
good result from ATAC and also you have a solution to the problem of bone in the
anastrozole arm of the ATAC trial.
D R L O V E : Is there any reason to think that other bisphosphonates would be different than
clodronate?
D R H O W E L L : No. It’s only a matter of degree. The bisphosphonate that can be given
weekly may be useful. But I spoke to Ingo Diel last night, and his view is that we have the
most information about clodronate for women with breast problems, therefore, clodronate
may be the drug to go with even though it’s an old drug and even though it’s been
superceded by alendronate, for example. Because we have so much data on clodronate
and it has a pretty low side-effect profile, it may be appropriate to use that
bisphosphonate for the IBIS II trial.
I guess my feeling is, given the Ingo Diel data and Trevor Powles’ data, plus the ATACdata, we might treat all elderly women with ER-positive breast cancer with a combinationof clodronate and anastrozole. That would make the treatment pretty safe, I guess, andthen we may have an additional effect of clodronate on survival as those two trials haveshown.
D R L O V E : Of course, clodronate’s not available commercially in the United States. So,
would you substitute another bisphosphonate?
D R H O W E L L : My feeling is, and I’ve spoken to the osteoporosis people on this and they
don’t think there’s any qualitative difference between the bisphosphonates; that it is a
quantitative problem. The real problem with bisphosphonates is obviously the toxicity,
particularly GI toxicity, and we need to find the best bisphosphonate to have low GI
toxicity. That’s my view.
D R L O V E : Can you talk a little bit about where things are right now with Faslodex as we
move into the endocrine era?
D R H O W E L L : Faslodex is going to be shortly accepted by the FDA. And all additional
treatments for advanced breast cancer are useful. As you well know, we give sequential
endocrine agents, and here you have another one that is a highly potent, estrogen receptor
downregulator. In second-line therapy, it is equivalent to aromatase inhibitors. It is
equivalent to our best drug. So, we have another best drug.
D R L O V E : With the current information we have right now, where do you see the place of
Faslodex?
D R H O W E L L : I think the first-line treatment for advanced disease in postmenopausal
women is an aromatase inhibitor, if they haven’t already had one. Then I see Faslodex
coming in after that, at the moment. I guess it probably doesn’t matter which order you
give them in, but we have more data on aromatase inhibitors, and you have to treat
patients on the basis of data. So, I’ve given an aromatase inhibitor followed by Faslodex.
D R L O V E : How much of a problem is the injection?
D R H O W E L L : I think it’s a non-problem. In fact, it could be seen as an advantage that
women don’t have to take pills every day. As you know, there haven’t been huge
problems in terms of injection site reactions at all. Even the Japanese don’t have problems
and they have smaller buttocks. So, I see that as possibly an advantage to the patient
compared to taking pills. I guess it might be an advantage to the doctor in the United
States because you get paid for giving injections, I understand. But that doesn’t happen in
the rest of the world. But no, it’s just a non-problem. And women say they don’t mind
having an injection if they’re getting an active compound. On the whole, it doesn’t
concern most of them.
D R L O V E : Do you generally think that giving one 5-cc injection is better than two 2.5-cc
injections?
D R H O W E L L : I don’t see why you should give two. There is no difference in the pain, or
there’s very little pain anyway. Why give two compared to one? In Europe, every
woman’s just had one injection. In fact, in the two trials, the one done in the rest of the
world and the one done in the U.S., there were more complaints of injection site reactions
in American women compared with the rest of the world. Now, that may be of greater
sensitivity or greater reporting in American women, but it may be due to the fact that they
had two injections.
D R L O V E : I’ve heard people say that an aromatase inhibitor and Faslodex are going to be
equivalent options, and maybe other issues, like whether an injection or pill would be
better for that individual woman – is that your approach?
D R H O W E L L : That’s great, isn’t it? You have a choice. Those treatments are clearly, pretty
equivalent, and the woman can make the choice. If cost is not an issue, I guess both of
these drugs are going to be more costly than tamoxifen.
D R L O V E : Now, when you say, in terms of first-line therapy, do you make a difference
whether or not the woman’s had prior tamoxifen? For example, in the adjuvant setting.
D R H O W E L L : No. It doesn’t seem to matter. We know that those drugs are equally active
in women whether they have had previous adjuvant tamoxifen or previous tamoxifen for
advanced disease. So, that’s not an issue. I guess if a woman hasn’t had adjuvant
tamoxifen and gets relapse, she’s ER-positive, you’re not going to use tamoxifen in that
situation. You’re going to use an aromatase inhibitor as first-line treatment. If expense is
no option, you might well use Faslodex as a second-line treatment, and maybe you’ll keep
tamoxifen right towards the end, with high-dose estrogens, which are making a
comeback.
D R L O V E : Can you capsulize what we know right now about the estrogen-estrogen
receptor interaction? What happens when you give a drug like tamoxifen and what
happens when you give a drug like Faslodex?
D R H O W E L L : Well, we know that the effects on the ER with Faslodex are different than
tamoxifen. Tamoxifen causes dimerization of the receptor binding to the estrogen
response element, and then activation of AF-1, but inactivation of AF-2, and partial
estrogen-agonist activity and partial antagonist activity depending on the cell and the
gene promoter context.
Faslodex is different, and there’s a lot of data to suggest that it’s different, in that, first ofall, it probably causes some dimerization, but it increases the turnover of receptors. Sosome receptor goes out of the nucleus. And the turnover time of the receptor is greatlyreduced. The second thing is that when it does bind and cause dimerization and thereceptor gets to the estrogen response element it inactivates AF-1 and AF-2. So, it totallyswitches off the receptor.
So, on those two mechanisms, it has a different effect than tamoxifen. And this is seen inthe preoperative studies where you give Faslodex and look over a period of two to threeweeks, what happens to immunostaining for estrogen receptor. Whereas, with tamoxifen,after two weeks, you can stain almost as much receptor as you had in the pre-treatmentsample. In the Faslodex sample there’s virtually no receptor there to be stained. So, thatindicates that there’s a totally different mechanism of action. D R L O V E : Any biologic reason to think that it might be a better drug than an aromatase
inhibitor?
D R H O W E L L : Yes, because you’re blocking the receptor continuously with Faslodex, and
there’s no receptor there for any estradiol, which is around, to stimulate the tumor.
Whereas, what Arimidex is doing, obviously, is lowering the serum estradiol, but there’s
still estradiol around, which could potentially stimulate the tumor. We know from the
work of Dick Santen, particularly, which is being reproduced by Mitch Dowsett, that,
looking at MCF-7 cells, those cells sort of hunt the prevailing estrogen concentrations. If
you take MCF-7 cells and put them in vitro and do a dose response curve to estrogen, you
find that they are maximally stimulated by 10–9 molar estradiol. But then if you deprive
those cells of estrogen for a month, you find that the dose response curve shifts to the left,
dramatically. That then the cells are sensitive to 10–15 molar estradiol.
D R L O V E : Wow! So, they increase their sensitivity?
D R H O W E L L : They increase their sensitivity quite dramatically. That’s a very, very
important experiment that Dick has done. So, you’ve got a woman on Arimidex and
you’ve lowered the estradiol to virtually undetectable levels, say from 50 picomolar to 5
or 2 picomolar. But if Dick’s data are transferable into humans – and I don’t see any
reason why not – there is the potential for the tumor to adapt to the lower prevailing
estradiol concentration, and so therefore, to circumvent the effect of Arimidex.
D R L O V E : Is this a genetic adaptation?
D R H O W E L L : Nobody knows, to be quite honest, the mechanism of it. My own
hypothesis is that there are clones of cells within the MCF-7 cell milieu, which actually
have the ability to respond to such low levels of estradiol. But it may be some change
within co-activators for example. Susan Fuqua has published a very, very interesting
paper in Cancer Research where she showed that an estrogen receptor with a mutation at
amino acid 303 in the hinge region, between the D part of the receptor and the E part of
the receptor, the hormone binding part. If you have that mutation, the tumor is sensitive,
and she’s shown this in pre-malignant lesions that the tumor is sensitive to exceedingly
low concentrations of estradiol. It seems that the co-activators c-activate at a much lower
concentration, in essence steroid receptor co-activator 1 causes stimulation at a much
lower concentration of estradiol.
So, that’s information concerning the fact that that is now actually being shown in pre-malignant lesions in women. Dick Santen’s hypothesis, or data, if you like, may indicatethat that’s what’s happening in tumors in women, and that may mean that there is anability of tumors to get around the very low estradiol concentrations of Arimidex and fortumors to be stimulated by those low concentrations. Whereas, theoretically, ICI 182,780should completely switch off receptors forever more. Actually, of course, we know thatthat doesn’t happen, because even in the phase II study, where we had fairly well selectedpatients, the median duration of response was only 24 months, and those tumors actuallydid become resistant.
The in vitro data would suggest that that resistance is not due to stimulation by ICI182,780. And those tumors don’t become stimulated either by ICI 182,780 or lowconcentrations of estradiol, and nobody knows how tumors become resistant ICI 182,780.
I think we have to see the data from the randomized trials. The Santen and the Fuqua data are tremendously important for the way that we look atendocrine therapy in the future. Because what Santen then did was, in the estrogendeprived MCF-7 cells that were maximally stimulated by 10–15 molar estradiol, he thengave those cells 10–10 molar estradiol, and they apoptosed. They were completelyeliminated. So that adding back – because 10–10 molar was a lot of estrogen for thoseparticular cells that were maximally stimulated by 10–15. And so that is very, veryimportant for the future of endocrine therapy because we may add back to women whoare failing on Arimidex or we may add back even before that failing, some estrogen. Wemay cause tumor depression by adding back small doses of estrogen. We’ve done a studywhere we’ve added back more estradiol in women who’ve had four endocrine therapies.
We’ve added back Stilbestrol in large doses.
D R L O V E : Like the old days?
D R H O W E L L : Like the old days. But it sort of tests Santen’s hypothesis. Adding back
estradiol in this group of 30 patients that we studied showed, after four endocrine
therapies – they may have had oophorectomy, they may have had tamoxifen, they may
have had a couple of aromatase inhibitors and then they have estradiol – and a third of
those women responded for a median duration of one year. So, this thinking that you
could re-challenge with estradiol, is important. We have not thought out endocrine
therapy sufficiently, and that we may alternate with aromatase inhibitors, estrogens;
there’s maybe a lot more mileage in endocrine therapy than we actually know.
There is also a preoperative trial being organized by the EORTC, in Europe, which issimply one injection of Faslodex after the diagnosis of breast cancer has been made versusan injection of placebo. And, of course, the Faslodex injection being present for over onemonth would cover the operative period in terms of anti-estrogenicity, with a potent anti-estrogen with estrogen receptor levels lowered. And if there is anything in the hypothesisfrom Bernie Fisher and others, that something happens around the peri-operative periodthis should be able to pick that up, because the aim is to put 3,000 women into that study.
And then the patients can have any adjuvant therapy that they want, because that shouldall come out in the wash with 3,000 patients.
D R L O V E : I assume that the ER would be determined before the Faslodex?
D R H O W E L L : No.
D R L O V E : Oh, really?
D R H O W E L L : No, because you want to give the drug as soon as you’ve done that true-cut
biopsy. And waiting around for ER estimations – which, in some centers, may take up to
a week or two weeks – would mitigate the effect of giving the drug. So, no. The protocol
is that you see the women, you do the true-cut biopsy, and as soon as you’ve got the
result of the true-cut biopsy, then that is invasive cancer. Then you randomize her straight
away.
DR CRAIG HENDERSON SUPPLEMENT
D R N E I L L O V E : 14 years of producing the Breast Cancer Update series has left me with
countless fond memories of interviewing brilliant and creative people like George Sledge
and Tony Howell. Our next speaker always provides an intriguing and usually
unpredictable trip through the maze of breast cancer biology and clinical research. Craig
Henderson seems to thrive on innovative ideas, and as usual, after meeting with him
during the Miami Breast Cancer Conference, I came away scratching my head in deep
thought. Dr Henderson began this sojourn by addressing what he considers the most
important recent development in the field.
D R C R A I G H E N D E R S O N : I think that the most exciting and challenging issue in this last
year was the presentation in San Antonio of the ATAC trial. Now, I can’t really say that it’s
affected my practice yet, except that I have to discuss that with every patient. I don’t
mean that in a negative sense. Clearly, they are important data, but I don’t think we know
quite what to do with it yet.
First of all, it’s dramatic, the advantage of Arimidex over tamoxifen. Secondly, the trial isso big that it’s very credible, and it’s hard not to believe that that’s correct. On the otherhand, there clearly is not enough follow-up to say anything about survival. Frequently,the end points and the outcomes of studies are not what we think they’re going to be.
Oftentimes we don’t understand the mechanisms quite as well as we think.
Let me give you an example. First of all early in my career in the ‘80s, I wrote, “tamoxifenis not a good adjuvant. After all, we had to kill all the cells.” That was the era in which wewere beginning to do high-dose chemotherapy. The idea of something that justsuppressed growth didn’t make sense. Of course, our thinking has changed a lot. Thenwe said, well, if it does work and we’re just suppressing the tumor, we obviously have togive it for a lifetime. But then we get five years versus ten years, and clearly, there’s not anadvantage. In fact, it’s kind of weird that ten is worse than five years. Well, then,sometimes the science follows.
Now we have good data that we’ve heard from people like Kent Osborne and CraigJordan and so on. We know that in some situations tamoxifen can actually stimulatetumor growth. And we know that tamoxifen does a lot of things other than just its effecton breast cancer. You can have all the overviews and everything in the world — it’s hardfrom clinical and laboratory data together to get the whole picture. So, all of a sudden, wehave this explosion of interesting information, starting about a decade ago when we gotthe new generation of aromatase inhibitors. So it wasn’t too surprising that Arimidex wasbetter than tamoxifen, but I have to say that I was betting that the two would be betterthan either one.
D R L O V E : I think a lot of people were.
D R H E N D E R S O N : Yeah. I suspect that Kent Osborne wasn’t, and that he was betting the
other way around. I wish we all had to put our money down before the answer or
something where we were identifiable before we saw the results. That’s kind of fun,
when you see the results. So, given all of that, I asked myself, should I wait until I have
the survival data, or do I go for it now?
Then the other thing that’s clearly important is bone. So, I say, well, am I going to give abisphosphonate at the same time? The data on suppression of bone metastases, or thedelay of bone metastases, using bisphosphonates is pretty compelling.
Then the other thing that I found kind of fascinating – I’ve been thinking to myself, wedon’t need three aromatase inhibitors. But one of the things that’s emerging is they maynot all be the same. I guess it’s kind of like an extension now of what I said about theATAC being exciting. So, I tell my patients about the ATAC trial. I think you’re obligated to. The results are justtoo important. I tend to start out with tamoxifen, yet I am prepared to change at anytime. I think maybe the other thing that’s important is that there are patients, for onereason or another, whom you can’t give tamoxifen to. I’m very comfortable switching inthose patients, if I need to, and I wouldn’t have been as comfortable six months ago.
D R L O V E : So, you see a patient right now who’s 60 years old, who’s got five positive
nodes, and she’s ER-positive. How would you shade your presentation of that? What
exactly would you say to the woman?
D R H E N D E R S O N : Well, first of all, 60-year-old woman with an estrogen receptor-positive
tumor is always going to get hormone therapy. In fact, probably unlike a lot of people, I
would certainly discuss chemotherapy with her, but I would give her hormone therapy. I
would tell her the results of the ATAC trial and tell her that, at this point, because of the
issues about survival – and for a 60-year-old woman, bone is very important. People
sometimes forget that if you take all 60-year-old women, the most common cause of death
is cardiovascular disease.
D R L O V E : But not in a woman with five positive nodes.
D R H E N D E R S O N : No. But the point still is that for all women osteoporosis is second. I
found that surprising when I learned that. I didn’t realize that the mortality from
osteoporosis was quite that high. Then third is breast cancer. And five positive nodes – I
never write off patients. I have too many five-positive-nodes that I’ve been seeing for 20
years. I still see patients that come to me from the years I was at the Farber.
D R L O V E : I’m not saying write it off. I’m just saying her health risk is breast cancer.
D R H E N D E R S O N : I wouldn’t disagree with that at all. There’s no question that her
biggest risk factor is her positive nodes. But I think you have to treat the whole patient,
and I think you do have to be concerned about bone in a 60-year-old woman, and I think
medical oncologists have to be.
D R L O V E : Incidentally, are you saying you wouldn’t give that woman chemotherapy?
D R H E N D E R S O N : No. I probably wouldn’t. I would talk with her about what we know
about chemotherapy, but I think that for a 60-year-old woman – now, let’s just say that her
ER was pretty weak. I honestly believe that the importance of having accurate ERs isrelated to whether you give chemotherapy or not, rather than whether you give tamoxifenor not. There’s a lot of talk here at this conference about the bottom end. I personallythink and I can’t prove this so, we’re getting into the area of opinion, but I think there’s alot of inferential data that would suggest that as the estrogen receptor level increases, itshould reflect a higher and higher percentage of the cells that are hormone-sensitive.
Now, I don’t know that we know that for sure. I’d be interested in, for example,somebody like Craig Allred’s opinion on that. But I think he would probably say thatthat’s probably true. We know, as we were discussing yesterday, that patients with even relatively smallamounts of estrogen receptor will have a surprisingly large response. In other words, thatthere will be clear benefit even with small receptor values. What I’m saying is that withdecisions about chemotherapy, I still feel that although chemotherapy will kill a receptor-positive cell, it makes sense to me, intuitively, that if the receptor-positive cell has beentreated, or affected by hormone therapy, then you don’t need to kill it twice. So, thatmeans that the reason that you’re adding chemotherapy is because you have cells that arenot responsive to the hormone therapy. If you’ve got a tumor that is predominantlyhormone-sensitive, then the added benefit of chemotherapy is going to be very small.
Now, we know, also – not just following that argument, but adding one other factor -–that the older the woman is, the smaller the benefit from chemotherapy and then thesmaller the added benefit from chemotherapy. The effects of chemotherapy correlatebetter with age than they do with menopausal status.
D R L O V E : I wanted to pick up on 10 things that you said, but let me just try to think of a
couple of them. One of the things I hadn’t thought about, if you think about Craig
Allred’s data that he presented again at this meeting, where he showed it was almost like
there was this effect that, if you have any ER, you see an effect of adjuvant tamoxifen, and
it really doesn’t change very much. It’s sort of like you go over the hump, and then you
just see this effect.
And Peter Ravdin brought up the question of, “if breast cancer’s a clonal disease, why isit that you would have some ER-positive cells, the rest would not be? Why are youresponding?” And you brought up the thought of maybe there was some interactionbetween the ER-positive cells and the other cells. Any thoughts about that wholediscussion? D R H E N D E R S O N : Well, I have a couple of thoughts about the discussion yesterday. One
is that I think Craig Allred is on the right track in the sense that I do believe that we need
to have reproducible, accurate measures that quantify the amount of ER that’s there. The
point that he was making yesterday is that we do not have those measures. They’re not in
place with any of the immunohistochemical assays that are available. We did have that 20
years ago, but we don’t now.
The second point is if you want to make a correlation between the effect of therapy andthe quantitative amount of a marker, you have to have two things. First you have to havean accurate assessment of the marker and second you have to have a sensitive end point.
Now, what is the most sensitive end point when you’re studying breast cancer? It is not response, and it’s not response duration. It’s not even survival in patients withmetastatic breast cancer. The most sensitive end point is disease-free survival in womenwith early breast cancer. Anything you do in metastatic breast cancer is a crude kind ofthing, because it’s just so complicated. So many things happen. The patients are sovariable. It’s so hard to measure the disease. There are so many things that affectsurvival. So, the most sensitive marker we have for response is disease-free survival.
So, you ask yourself, do we have any evidence, any data, any data set, in which we have1,000 to 3,000 patients where we can say these are reliable estimates of the size of the ERand also have reliable information on disease-free survival? We do have the latter. Wehave lots of trials, tens of thousands of patients, where the disease-free survival has beenvery, very carefully measured. We have small series, where we have probably veryaccurate estimates using the ligand-binding assays, but we don’t have both of thosetogether. So, the question is one that we can only raise in theory. The evidence does not exist because we don’t meet those two criteria – a sensitive endpoint in the clinic and a reproducible, accurate estimate of the size of the marker, in thiscase estrogen receptor in the laboratory. Those don’t exist.
D R L O V E : You said that you wouldn’t be offering chemotherapy to this 60-year-old
woman who’s ER-positive. Does that, in general, mean that for postmenopausal women
who are ER-positive, you’re not using chemotherapy?
D R H E N D E R S O N : That’s correct. And that’s always sort of been my position. However, I
do want to make sure that there are two footnotes to that. One is that I always go through
all of the evidence. I present the evidence from the overviews as being probably the most
reliable. Number two is that I always indicate to the patients that I see that I think it is
reasonable to get chemotherapy, and that I think it’s reasonable to make either choice.
The benefit, however, is very, very modest. Let’s just take the woman you described.
She’s in a 60 to 70-year-old category. The effects on survival are going to be about one-
half to one-third the effects of chemotherapy on premenopausal women.
Then, if you take the women who are receptor-positive, the effects are going to be evensmaller than that. It’s very difficult to get precise numbers, because the most reliablesource is the overview. Of course, as women get older and older, we have fewer andfewer observations.
D R L O V E : But the reason that there’s less of an effect is because the baseline relapse rate
is being decreased by hormonal therapy. Is that your take?
D R H E N D E R S O N : No.
D R L O V E : What’s the reason?
D R H E N D E R S O N : I don’t think we know all of the reasons. I can speculate. And, again,
it’s important to understand that this is speculation. A lot of this is stuff I’ve just been
talking about this last year and not much before that, because I’ve been trying to
understand what happened in what people refer to as the taxane trial. Why weren’t there
more effects from taxanes in the receptor-positive women? What we do know, first of all, with considerable certainty, is that each year that a womangets older, there is less effect of chemotherapy. So, for example, in a young woman, thereduction in the odds of death is going to be pushing 30 percent or slightly higher. Andeach decade that gets smaller, so that by the time you get to a woman 60 to 70 years old,the reduction in the odds of death has fallen to about nine percent.
Interestingly, tamoxifen of course goes in exactly the opposite pattern, going from youngto old. Now, the first time I saw these data – these were generated originally by RichardPeto – I said, “Well, that’s just an effect of chemotherapy on the ovaries.” We didn’t reallyhave menopausal data that was as good as the age data. Age of human beings all overthe world is accurate, by and large, and menopausal data is very inaccurate for a numberof reasons, which most logisticians readily understand. But I don’t think they evenunderstand fully how inaccurate it is. But we know that the menopause doesn’t at amoment in time, and we know that, different studies use different definitions and so on.
But I finally got him to do an analysis on the relationship between both tamoxifen andchemotherapy and menopause. And it wasn’t neat in the same way that age was. Age,every time, it’s consistent. You see this pattern. And, in fact, there’s a highly significantlinear effect. But you don’t see it with menopause.
So, he first presented this to me, probably, in the late ‘80s, maybe with the 1990 overview.
In fact, it may have gone back to ’85, but I just rejected it out of hand as this just doesn’tmake sense to me. For 15 years, I’ve been trying to figure out how to explain this. But it’sdefinitely a real phenomenon. And it’s not simply explained by menopausal status, forthe reasons I just gave you. Then we come along to this taxane trial, and the thing that’sinteresting is that in the ER-negative patients, adding Taxol is very beneficial, whetheryou’re premenopausal or postmenopausal. So, if you were a 60-year-old woman with fivepositive nodes or ER-negative, I would have no problem giving her chemotherapy. Butshe was ER-positive. So, why don’t we see an effect? In fact, my first reaction – and this is not published. Nobody’s really looked at it – but myfirst reaction was it was because we were using five years of tamoxifen. I thought, if youuse longer and better hormone therapy, then there’s less for the chemotherapy to add.
D R L O V E : It’s harder to see it.
D R H E N D E R S O N : That’s right.
D R L O V E : Fewer events.
D R H E N D E R S O N : What I did is I went back to the overview, and I took all of the trials
where women had gotten one year of tamoxifen plus or minus chemotherapy. I then put
all the trials together that had gotten two years plus or minus chemotherapy, then five
years plus or minus chemotherapy. My hypothesis was that if you got only one year of
tamoxifen, the chemotherapy would have a greater benefit than if you’d gotten five years
of tamoxifen. That was my first hypothesis. Wrong. It didn’t work out. I couldn’t show
that. And using the overview data, of course, I had large numbers. So, that wasn’t the
explanation.
Then I thought we know that among women ages 50 to 59 in the overviews, 20 percent ofthose women are actually menstruating. They’re truly premenopausal. So, I thought,well, maybe that’s part of it. It’s kind of mucking it up. We’re seeing a premenopausaleffect in women that were classified as postmenopausal. But then, all of a sudden we have – well, not all of a sudden – we have these data thatwere presented by Kathy Albain last year at ASCO. I think it’s a very important study.
This is the Intergroup study, and they randomized ER-positive women to tamoxifen ortamoxifen plus CAF.
Now, those women, unlike the ones in the overviews, were all postmenopausal. In otherwords, they had to have stopped menstruating. So, unlike the NSABP studies, therewasn’t a question. They defined it not by age, but by menopausal status. And secondly,the effect is enormous. So, then I went back again to the overview data – and I haven’t finished this analysis.
This was very preliminary – and some of it was stimulated discussions with KathyAlbain, among others, so I don’t want to take all the credit for this. But if you look at theincidence of amenorrhea and I think almost all people will agree at this point that inpremenopausal women, that there is a relationship between the development ofamenorrhea and the effectiveness of chemotherapy. I won’t review all that, but there’s anabundance of data. So, then, the next thing you say, if that’s the case, then you would expect that thoseregimens which cause the most amenorrhea would be the regimens that would be mosteffective in at least premenopausal women. So, if you go back and you look, you see thatthe most effective regimens, the highest incidence of amenorrhea actually comes withCMF. But, also, if you look at the Canadian trial using CEF, in fact that has given some ofthe most dramatic and the largest actual benefits from adjuvant chemotherapy of any trialthat’s been done thus far. But in that particular trial, they used cyclophosphamide in thesame way we do with CMF; that is, daily oral cyclophosphamide times 14 days. Andthat’s exactly what Kathy Albain did in her trial. She used classic CAF. So, then I went back to the overview — and it’s hard to sort these things out. And, as Isaid, I’m still working on it. But, I see that the trials that used the classic CAF, CEF andCMF, using just exactly like it was in the original Bonnadonna trial, actually seemed to bethe most effective trials that we have.
Then I went back, also, and look at work that people have done – there’s a paper in JCO,where they looked at the incidence of amenorrhea with different regimens, and it’s veryinteresting. CA times four is one of the lowest. This leads me to the possibility that, infact, maybe one of the reasons that we didn’t see an effect of adding four cycles of Taxol iswe were giving four injections, and that this prolonged suppression somehow may be acritical factor. So, I come around to think that a prolonged cyclophosphamide may be avery important issue. Also the duration itself may be important, but we have that six-month duration with the taxane trial. Well, now, how does this relate to the question youasked? D R L O V E : Whatever that might have been.
D R H E N D E R S O N : Well, you were asking about a 60-year-old woman.
D R L O V E : Exactly. Right.
D R H E N D E R S O N : So, I’ve been puzzling for years because all I’ve been talking about
now in these last few minutes is the effect on the ovary. But somehow we have his idea,
at least I had it, rather naïvely, that once a woman stops menstruating the ovary somehow
flops over and dies. It’s a simplistic notion, but I don’t think I’m alone in dismissing the
ovary.
When you hear discussions, for example, on aromatase inhibitors, they routinely startabout something about the fact that androstenedione is produced by the adrenal gland,and then that goes to various fat tissues and then the breast tissues themselves and isconverted with aromatase to estrogens. That’s the usual lecture. Nobody mentionsanything about the ovary there.
So, a couple of things now, I’ve been pondering this last year. And I talked about some ofthis at ASCO last year, and I’ve given further lectures since on it. But first thing I did is Iwent back to the literature on postmenopausal ovarian function. I looked up all thepapers that I could find. The first thing you see is that in fact, when a woman has her last menstrual period, herovaries don’t stop functioning. We knew this, in a way, before I looked up these papers,because we know that women frequently will have, during the perimenopausal period,short menses. They’ll go for three or four months without one, and then maybe haveanother one, and then they’ll go maybe six months. But it’s interesting how that goes onfor quite a while. But when you actually look at the formal studies, you find that, in fact,there is quite a bit of estrogen production and that it tails off. In fact, one of the beststudies indicated that the average time that the ovary continues to produce estrogen isfour years, with outliers as far as 10 years. Now, the other thing, of course, is that you also have testosterone and androstenedionebeing produced. In fact, as women age, if anything, there is more and more production,rather than less.
D R L O V E : From the ovary.
D R H E N D E R S O N : From the ovary. And so, when those aromatases are working, they’re
not working just on – in fact, probably half of the androgens in a postmenopausal woman
come from the ovary, and the other half from the adrenal gland. It was interesting, when I
started my career, we always did adrenalectomies, and the surgeons always took the
ovaries at the same time. Of course, these were postmenopausal women. But I thought
that was stupid. Why should they go down, it’s a bigger operation? A Little extra fee? I
was kind of cynical about it. Now I look at it from this point of view, and I have to say, by
the way, that some of the best responses I’ve had in my entire career were those older
adrenalectomy patients. I had patients that were in remission for 10 years or more after
adrenalectomy.
D R L O V E : Hmm.
D R H E N D E R S O N : It was a funny time, and we forget. The other group that have had the
longest remissions of my career — these are, now, metastatic disease patients — are
patients who had ovarian ablation. Anybody that’s been practicing for a long time has
always been aware of that. Okay. So, the ovary is not unimportant in the
postmenopausal woman.
Then I look to see, well, are there any papers that have measured the effects ofchemotherapy on androgens, for example, or estrogens in postmenopausal women? Infact, there are some data, very, very few, but they do suggest that chemotherapy reducesboth estrogens and androgens even in postmenopausal women. As I said, this is very,very skimpy data, because it doesn’t fit with anybody’s hypothesis. There’s another little pearl that people have forgotten. The first person to teach me wasNissen-Meyer. Remember Nissen-Meyer, the 6-day cyclophosphamide? He’s still alive,by the way. He’s about 90. Wonderful guy.
D R L O V E : I remember reading his papers.
D R H E N D E R S O N : Well, he’s a very, very thoughtful scholar of the biology of breast
disease. He did one of the earliest ovarian ablation trials — oophorectomy trials — in
Norway back in actually, I think, his trial was in the early ‘50s. This was before he did
chemotherapy. But they did an ovarian ablation trial. And they included
postmenopausal women. In fact, most of those early trials had both pre- and
postmenopausal women. After all, there wasn’t anything else to do, but, remember, the
first adjuvant trials were all either surgical or radiation removal of the ovaries.
I’ve always thought that was kind of silly to include postmenopausal women. Andalmost everybody else says that that was silly. When you look at the effects, you don’t seeany effects. So, I asked him one day, “Why did youe include postmenopausal womenbecause it doesn’t make sense”? He said, “Oh, postmenopausal women will respond tooophorectomy.” They will? That was strange. But he said, “There’s one caveat. The onlyones that respond are those who are in their sixties. The ones who are just a few yearsafter the menopause don’t respond.” Well, two comments on that, that are kind of interesting. The first is that a group ofsurgeons at the University of Oregon did an ovarian ablation, or oophorectomy series inthe mid ‘80s, and they found exactly the same thing that Nissen-Meyer had told me yearsbefore. That is, ovarian ablation does work in postmenopausal women, but it’s only in theolder postmenopausal women in which you get an effect. That was strange.
The other thing – and, again, these are old data, long before we had ER and so on. Peoplehave forgotten this, because, remember, there was a time when we had to figure out, whowe are going to treat with hormone therapies, and we didn’t have estrogen receptors orprogesterone receptors to help us, so we had to use the clinical characteristics. I publishedthis in the New England Journal in, I think, ’81, so it’s in the literature, where I looked atall of these trials and added all this stuff up. And, in fact, you will find that using all thevarious hormone therapies that are appropriate and using the right ones, estrogens, high- dose estrogens, in postmenopausal women — this also included tamoxifen data, andusing ovarian ablation in younger women. But, you look at all of these patterns — weused a lot of androgens in those days — it turns out that you got the best responders,where those people who were premenopausal and very postmenopausal. There was a blipin there in the perimenopausal years where no kind of endocrine therapy worked verywell. And so that sort of was just another little piece of the puzzle that seems to fit in. So, where is all this going? I have a hunch — in fact, it’s more than a hunch — ahypothesis that I think we have to begin to test now, that, in fact, what we’re seeing inboth pre- and postmenopausal women when we give chemotherapy, if they are receptor-positive, is an effect that is mediated through the ovary. Now, I don’t want to oversimplify this because let’s say that you’ve got a breast cancerthat has lots of ER-negative and lots of ER-positive. I imagine that there’s some additiveeffect. And so the way I practice, at least at this point, is that, if a woman has a wild andwooly tumor, which you can tell by looking in the microscope. It’s high grade. She has,let’s say, a very weak ER — and I always try to get what I can with the ER. I look at thetumor under the microscope with the pathologist. Everything makes me think that thiswoman is just marginally receptor-positive, and she were 60 with five positive nodes, Iwould consider giving both hormone therapy and chemotherapy.
On the other hand, if she’s strongly ER-positive, has a well-differentiated tumor, and let’ssay she’s both ER- and PR-positive even with five positive nodes, I think the addedbenefit of the chemotherapy is going to be so small. She needs to be told that this mayprolong your life by a few additional months, but probably is not going to prolong yourlife by years, the way the hormone therapy will.
D R L O V E : So, what you’re saying then is that when you look again, for example, at the
CA-Taxol study in the ER-positive, postmenopausal woman, the reason you’re not seeing
any benefit of adding the taxane is you’ve already given them tamoxifen. So, you’re just
adding another hormonal intervention by giving them the chemotherapy?
D R H E N D E R S O N : Well, what I’m saying is that I think in the case of that study –
remember, we had patients with 15 positive nodes and patients who were ER-weakly-
positive as well as strongly positive. But I was suggesting that the CA times four and the
Taxol — we don’t know for sure, well with the CA times four, we do know. It doesn’t
cause a lot of amenorrhea. My guess is that if we get at the data, which we probably will
one of these days — but we did not collect it, by the way, in CALGB-9344 in the
Intergroup study. But if we can get at that data, I think we will find that probably the
taxanes do not give as much ovarian suppression as, let’s say 14 days of
cyclophosphamide do.
Now, that’s a hypothesis, but my hypothesis is that neither CA times four nor Taxol timesfour represent optimal treatment if you want to get ovarian suppression. And what I’vebeen saying is that I think ovarian suppression is important in receptor-positive patientswhether they’re premenopausal or postmenopausal.
D R L O V E : How do you explain what’s going on in the 50- to 60-year-olds?
D R H E N D E R S O N : As opposed to the 60- to 70-year-olds? Oh, I think that applies to
women of all ages. In other words, what’s happening is, during the period from 50 to 60,
a woman’s ovary is changing. In fact, it’s changing probably from the age of 40 to 60. A
woman’s ovary is constantly changing, and the mix of estrogen and androgen is changing
throughout that period of time She’s gradually getting decreases in estrogen, increases in
androgen, and that in any given woman, it’s not a smooth process. Again, reflected by
these funny perimenopausal patterns. It’s actually been documented in estrogen levels.
And there’s also a lot of variation from woman to woman, which makes it very difficult to
interpret data.
D R L O V E : So, for example, you mentioned the fact that the trials of ovarian ablation don’t
show an effect in that age group, and they do in the older and younger. Why do you
think that is?
D R H E N D E R S O N : I’m not certain why that is. I just can’t answer that question. I link it to
this earlier observation that in the perimenopausal women, endocrine therapy doesn’t
work very well. But, nonetheless, I can’t explain why it doesn’t work in that
perimenopausal group. That’s an old observation.
Over the years, I’ve always been trying to understand why higher doses of estrogen seemto be suppressive. Most of the animal models don’t really answer that question. I have tosay this meeting is the first time where I’ve seen really elegant data — presented by KentOsborne and Tony Howell yesterday — to understand the dose response curve, and themost elegant data to convince me that there really is a dose response curve to estrogen.
One of the things, of course, as you saw yesterday, was that we saw several different doseresponse curves. If you were to lump all those settings together, you would have a hardertime seeing the dose response, simply because, depending on what your prior estrogenstatus is, you have a different dose response. In other words, if you’ve been estrogendeprived, it takes less estrogen to stimulate the tumor. And then you get a fall off withhigher doses and tumor suppression at many different doses than if you haven’t beenestrogen deprived. So, maybe those observations yesterday — I haven’t had enough timeto kind of think about it more — but maybe they somehow relate in part to the questionyou’ve raised. I can’t give you a neat hypothesis at this point. But I found those verycompelling and interesting data yesterday.

Source: http://www.breastcancerupdate.com/pdf/BCU200205_sup.pdf