INTERVIEW WITH DR. ROBERT W CARLSON

The recent very rapid evolution of endocrine therapy, and specifically Faslodex or fulvestrant was addressed by another presenter at this year's chemotherapy foundation meeting, Dr Robert Carlson, who also chairs the breast cancer committee of the National Comprehensive Cancer Network (NCCN), that produces what is widely viewed as the definitive oncology practice guidelines. Dr Carlson reviewed with me some of the recent deliberations of the NCCN's fascinating initiative in evidence-based medicine, but to begin, he commented on the recent emergence of fulvestrant, which is not yet available for nonprotocol use in the United States. In an era of targeted antitumor therapies, this fascinating agent is proving to have a unique mechanism of action.

DR. LOVE: The recent very rapid evolution of endocrine therapy, and specifically Faslodex or fulvestrant was addressed by another presenter at this year's chemotherapy foundation meeting, Dr Robert Carlson, who also chairs the breast cancer committee of the National Comprehensive Cancer Network (NCCN), that produces what is widely viewed as the definitive oncology practice guidelines. Dr Carlson reviewed with me some of the recent deliberations of the NCCN's fascinating initiative in evidence-based medicine, but to begin, he commented on the recent emergence of fulvestrant, which is not yet available for nonprotocol use in the United States. In an era of targeted antitumor therapies, this fascinating agent is proving to have a unique mechanism of action.

DR. CARLSON: There's a new class of hormonal agents known as the estrogen receptor down-regulators, that have a fundamentally different interaction with the breast cancer cell, than do the other hormonal agents. The estrogen receptor down-regulators, when they occupy the estrogen receptor "monomer," actually inhibit the dimerization receptor, and the dimerization of the receptor is necessary for activation of activation function 2. Not only do the estrogen receptor down-regulators inhibit the activation of the activation function 2, but also they result in the inactivation of the constituentively activated activation factor 1. And that results in estrogen receptor down-regulation. It results in the failure of the estrogen receptor to interact with the estrogen response elements in the nucleus, so that you get pretty much a complete failure of estrogen receptor responsive activation of transcription factors. And that results, in turn, in essentially complete antagonistic function on the estrogen receptor.

There are some good things about that, and that includes the failure of stimulation of the endometrium. And while the selective estrogen receptor modulators stimulate the endometrium and are associated with very low frequencies of endometrial carcinoma, nonetheless that does happen. And it looks like the class of drugs, the estrogen receptor down-regulators, will not have that effect.

Fulvestrant, which is the specific estrogen receptor down-regulator that has clinical experience, require intramuscular injection on a monthly basis. That can be viewed either as an advantage, because you can assure compliance, or a disadvantage. It's an intramuscular injection and requires visits to the healthcare practitioner to administer.
There have been several randomized trials looking at fulvestrant in postmenopausal women with metastatic hormone-responsive breast cancer. Those trials have been performed both in Europe and North America, and the results are quite consistent. The randomization was in women who were tamoxifen failures and randomized against anastrozole, a selective non-steroidal aromatase inhibitor. Anastrozole was chosen because of recent randomized studies in turn against tamoxifen that suggested the efficacy of anastrozole as superior as first-line therapy to tamoxifen. So, the hurdle that fulvestrant was put up against was pretty high.

The results of those trials demonstrate similar time to progression, versus anastrozole. One of the trials, the North American trial, in those women who did, in fact, achieve a response, the duration of response was longer, almost twice as long with fulvestrant as compared with anastrozole. That is a difficult comparison to do, however, because who responds is not a random event. So the direct comparison is fraught with methodologic hazards.

There is a methodology, however, that looks at probability of being in remission versus time, and that analysis tends not to be biased by whom responds and who doesn't respond. And in that analysis, fulvestrant in the North American trial also had superior time in response in comparison with anastrozole, and that difference was actually highly statistically significant.

DR. LOVE: That's interesting. I guess Ken Osborne presented that data for the first time at the 2000 San Antonio meeting. I don't remember seeing that variable. What was it called? Time in response, or what was that?

DR. CARLSON: It's a methodology called time in response. And that analysis has only recently been done. Steve Jones has taken the lead in overseeing that analysis. And that data was presented for the first time at the Lynn Sage meetings in Chicago last month and will be reported at the AACR meetings upcoming.

DR. LOVE: And that concept of a longer duration actually was what Ken Osborne had predicted, I guess, based on the laboratory work he'd done.

DR. CARLSON: That's absolutely correct. In the MCF-7 implanted tumor in nude mice in Dr. Osborne's laboratory, they did see, initially, estradiol- stimulated animal tumors that, after the estradiol stimulation was withdrawn, that in the control situation, the tumors had a plateau in growth and then eventually continued to grow. With tamoxifen, there was an initial decline in tumor mass and, eventually, stimulation of tumor growth. With fulvestrant, there was initial suppression of tumor growth that was greater than that achieved with tamoxifen and with a duration that was about twice as long as with tamoxifen. Eventually, though, even with fulvestrant, the MCF-7 cells did become resistant, but it took about twice as long as with tamoxifen.

DR. LOVE: What's your global take as a clinician in terms of when you look at these clinical trial data that are available right now? Do you have the general impression that fulvestrant is going to offer more benefit to patients than anastrozole in these tamoxifen-resistant patients?

DR. CARLSON: I think that question remains to be answered definitively. Certainly the data that we have available suggests that Fulvestrant is at least as active as anastrozole and, from the North American trial, there's a suggestion that it may be more active. The superiority in terms of duration of response, however, was not observed in the European trial. Whether that's because there really is no superiority or a result of the methodological differences between the trials remains to be seen. Certainly, by mechanism of action and what we see pre-clinically, I would expect it to have superior clinical activity, but, of course, the randomized trials are necessary to confirm that optimism.

DR. LOVE: And I guess the point that was brought up when these two trials were presented, you mentioned the methodologic difference. And the big one that I saw was that the North American trial was double-blinded, so that the patients were evaluated at the same time period, whereas the European trial, it was not double-blinded and, actually, the women on fulvestrant were getting evaluated more frequently. So, theoretically, they might be picked up to progress earlier. Does that seem like a valid observation?

DR. CARLSON: It's a totally valid observation. In the European trial, the women had to go to the healthcare provider to get their monthly fulvestrant injections in an unblinded fashion, while in the anastrozole arm, they were evaluated every three months. So we should expect a shorter time to progression in the fulvestrant arm, even if the two therapies are absolutely equivalent. In the North American trial, there was a placebo injection given on a monthly basis in those women who were treated with anastrozole and, of course, an anastrozole placebo for those who were treated with fulvestrant. While the women came in for injections every month, their periods of evaluation for time to progression were on an every-three-monthly basis.

DR. LOVE: So, I guess the idea is, if the patients come in more frequently, it's more likely that they're going to say, "Gee, I don't feel too good. I'm starting to have some bone pain," whatever, and then get worked up.

DR. CARLSON: That's right. And it also allows more frequent physical examinations of chest wall lesions or lymph node disease, liver size, and so forth. So there are just many more opportunities to decide that you want to look for evidence of progressive disease.

DR. LOVE: I'd like to dissect out a little bit about some of the different points that you made, and the first is, you refer to estrogen receptor down-regulators. "S." Is there any other one that's been identified right now, besides fulvestrant?

DR. CARLSON: Well, there are several estrogen receptor down-regulators that are in development. They include another AstraZeneca compound known as ICI 164384. There also is one known as EM 139 and RU 58668 that are in development. None of them, to my knowledge, have been subjected yet, however, to any meaningful prospective clinical trials to really try to define their true activity, though, in the clinic.

DR. LOVE: I remember when I first heard about this compound, fulvestrant. I think at that point, it was called ICI 182180 or whatever. But, it was a while back. It's been around for a while. I've been hearing about it for, seems like, maybe 10 years. And I think initially what I was hearing was that, it was a "pure anti-estrogen," almost the idea that it would be almost like a SERM, but only with antagonist effect. Now it seems like maybe, in the last few years, it's like, whoa, wait a second. That's not really what's going on. It's really getting rid of the receptor, which is very, very different.

DR. CARLSON: Well, I think we have learned over the years, information about the mechanism of action that makes it a more interesting compound. One of the difficulties with it, also, was that a lot of time was spent trying to formulate fulvestrant in an oral form, so that it could be taken as the other hormonal agents in an oral formulation. Eventually, though, I think it was decided that it just was not going to be possible, and so it was decided to move forward with it as an intramuscular injection. And, again, the data has been very encouraging with that formulation. So, part of the time that was required to develop the agent actually was time that was unsuccessfully devoted to trying to develop an oral form.

DR. LOVE: You also mentioned the fact that through a variety of mechanisms, basically - I guess the receptor disappears from the cell, in most cases?

DR. CARLSON: That's correct. By immunohistochemistry, estrogen receptor-positive cells that are exposed to fulvestrant actually lose the presence of the estrogen receptor. It's a true estrogen receptor down-regulation.

DR. LOVE: In a way, that could be interpreted as something from a negative perspective. Are we taking an ER-positive cancer and making it an ER-negative? Is that a correct approach, or not really?

DR. CARLSON: Well, one of the uncertainties that we have about fulvestrant is the sequencing of fulvestrant in relation to the other hormonal therapies. There has been concern that breast cancers treated with fulvestrant that lose the estrogen receptor may become hormone refractory for subsequent therapies. There have been limited studies done of patients previously with fulvestrant, that are then subsequently treated with either one of the selective non-steroidal aromatase inhibitors or with megestrol acetate. And in those studies, responses have been seen, objective responses in the range of about 20 percent, which is about what we would expect for a third-line hormonal therapy.
In pre-clinical systems, recent data would also suggest that the estrogen receptor does reappear in the cells after fulvestrant is withdrawn. So, some of the concerns about evolving the tumor into a hormone refractory state appear to be unfounded.

DR. LOVE: I guess that's the concept that I've been hearing about recently, that the genotype of the tumor really isn't changed. It's sort of a temporary change in the phenotype while the drug's available. Is that your concept right now?

DR. CARLSON: Yes. That appears to be exactly what happens. And, again, it does appear that there are hormone responses in the rates you would expect to third- and, presumably, to fourth-line hormonal therapies.

DR. LOVE: If you think about the whole paradigm of hormone therapy, in a way it hasn't really followed the antibiotic-resistant kind of a thing. We know that a woman who responds to hormonal therapy in metastatic disease, who then progresses, actually it's a predictor to respond again. So, it seems like somehow this responsiveness comes back. It's kind of a strange thing, but I guess that's something we've known for decades, really.

DR. CARLSON: Well, it's certainly an observation that's been recognized in the clinic for decades, and it's a very good thing, that we can have sequential hormonal responses.

DR. LOVE: Yeah. That's really an amazing thing. Now, a couple of other points in terms of some of the clinical issues here. You mentioned that these trials were done in postmenopausal women. Do we have any data in premenopausal women, and is there any reason to think it wouldn't work in a premenopausal woman?

DR. CARLSON: Well, first of all, there is no reason to believe that it wouldn't be effective in premenopausal women. The agent appears to be acceptably toxic in short-term trials for the premenopausal women. But we do not yet have long-term disease-oriented clinical trials in premenopausal women. And because of that, I would be very cautious about considering using this agent outside the confined of the clinical trial in premenopausal populations.

DR. LOVE: One of the things that was interesting about that - I guess it was really a safety study in patients about to undergo hysterectomy. Because, actually, that was presented at this meeting here in New York a year ago by Richard Elledge. That was the first time, I think, it was presented. And I guess these were in women who didn't have breast cancer. They were going to have hysterectomy and were studied short-term. And one of the things that was interesting was, at least in that short-term study, it didn't appear to have a negative effect on bone. And I guess we don't know, really, too much right now in terms of non-breast tissues and Faslodex. You mentioned the endometrium. And, actually, I think that some of those safety data looked pretty encouraging, at least early on.

DR. CARLSON: Right. In preclinical systems, it looks like fulvestrant has limited effects on bone marrow density and limited effects on serum lipids. The data in humans is quite early, and those questions are questions that will have to be addressed by subsequent clinical trials. Certainly, those questions are especially important if fulvestrant is moved into the adjuvant setting, as we certainly expect it to be, because those are the women who are likely to be long-term survivors and in whom bony events and cardiovascular events are of substantial concern.

DR. LOVE: Another issue in terms of side effects that's going to be of great interest, I think, is hot flashes. And I know, again looking back a couple of years ago when I started hearing about this from Tony Howell, there were some studies that suggested it doesn't cross the blood-brain barrier. At that point, there really weren't any clinical data, but there was sort of the cautious hope that, because of that, it wasn't going to cause hot flashes. Do we know anything more at this point?

DR. CARLSON: Well, we know that in the comparative trials with anastrozole, that the occurrence of hot flashes was in the range of about 20 percent in both treatment groups, both with anastrozole and with fulvestrant, so that it appears to be comparable in toxicity, at least as manifested by hot flashes, as anastrozole. The agent doesn't cross the blood-brain barrier, so there is hope that the occurrence of hot flashes will be lower than that with the selective estrogen receptor modulators that typically do cross the blood-brain barrier. But, again, I think we're going to have to see what happens with larger populations of women and, as we become more sophisticated in understanding what the true mechanism of hot flashes is.

DR. LOVE: One of the things that was interesting to me was this issue of the injection. I talked to nurses who administer the drug in these trials and also physicians who are involved with it, and actually, if you look at the numbers, it's interesting. The injection side effects, you can't tell the placebo from the actual fulvestrant patients. It looks like, other than the question of inconvenience, that the injection itself - and I guess it's a 5-cc injection, too, which there was a little concern about - but it seems like that's not going to turn out to be much of an issue in terms of local effects.

DR. CARLSON: It looks like it. The European and North American randomized trials did differ, not only in being unblinded in Europe and blinded in North America, but they also differed in terms of the number of injections.
In Europe the injection was given as a single 5-cc deep intramuscular injection. In the North American trial the study design had to be modified to use two different 2.5-cc injections on the same day, because nurses in North America are trained never to give more than 2.5 cc's in any single intramuscular injection.
If we look at the local toxicity from those injections, the local toxicity reported in the European trial is actually lower overall and per injection, than it was in the North American trial. I'm not sure if that's a result of twice as many injections, and so twice the opportunity to have discomfort in the North American trial, or whether that's related to differences in toxicity reporting between North American patients and physicians versus our European patients and colleagues.

DR. LOVE: Well, there are quite a few nurses who have done both, because there was another trial that was 5-cc versus tamoxifen. What I've heard from them is basically it's easier just to give one injection.

DR. CARLSON: Well, it certainly is easier. And the injections also need to be given fairly slowly because of volume.

DR. LOVE: Right. One of the things I like to ask when I interview research leaders is, how they take their patients and what their algorithms are. I think there's probably nobody in the country who's more tied into the concept of algorithms of breast cancer management than you are in terms of your chair of the Breast Cancer Committee of the NCCN, who does the practice guidelines.

DR. CARLSON: Yes.

DR. LOVE: An enviable task, I would say. (Laughter) I want to ask you, actually. I would like to talk a little bit about that and then, as we talk about it, see where you think fulvestrant's going to fit in, or maybe your group has already started to talk a little bit about it. But can you overview what the committee's been doing and how things have evolved since you started?

DR. CARLSON: Well, we're now in, I think, the fifth or sixth generation of breast cancer guidelines within the NCCN. And the panel has expanded from the initial six panel members, now into 20 or more panel members. That's been an intentional process to be more inclusive of the NCCN institutions, but also to try to bring in diversity of opinion and to expand the types of experts that we have on the panel to be more inclusive.

The guidelines started out as relatively simple algorithms that were intended to be quite broad and not very detailed to allow a lot of flexibility among the practitioners, and in a sense to allow the practitioner to justify with third-party carriers coverage for a broad spectrum in terms of treatment. I think we have maintained the goal of having broad treatment options available within the guideline, but the guidelines have become much, much more detailed or much more specific as we go to each of the different options in the guideline over the years.

I find it also quite fascinating that every year, as we go through the guideline, we find places where new evidence forces us to really change the guideline. And we're the only guideline in the NCCN that actually undergoes revision on an annual basis. And I think that tells us something about the quality and quantity of research in the area of breast cancer that's ongoing internationally.

DR. LOVE: Let's talk about your most recent deliberations. Would you say there's any other group that's approaching this in a comparable fashion?

DR. CARLSON: Well, there certainly are other groups, including the Canadians and the European Society of Medical Oncology and so forth, who are, or who have developed evidence-based breast cancer treatment guidelines. What really sets the NCCN guidelines apart, however, is that they're updated on an annual basis. To update these guidelines is a very time-consuming task and requires a pretty dedicated committee structure and staff support, I would add.

Within the NCCN, we use primarily an evidence-based system, but it's an evidence-based consensus that is reached by the panel. It differs from the evidence-based approach in that it doesn't require, necessarily, a very objective stereotypic scoring of each piece of evidence to decide how it fits in the hierarchy. The evidence-based consensus approach, rather, the experts on the panel identify the papers or the studies that are felt to be of highest quality, without necessarily assigning objective scoring systems to those papers. And the panel then uses those papers to make recommendations.

Using the evidence-based consensus approach also allows the panel to make recommendations in situations where evidence is lacking. And there are many situations where evidence is lacking, or at least high-quality evidence is lacking, in the treatment of breast cancer. And so this allows our guidelines to be much more complete than many of the other - actually, than any of the other guidelines for breast cancer treatment that have been developed.

DR. LOVE: Let's talk about metastatic disease, first. Can you talk a little bit, sort of overview that the algorithms were for metastatic disease when you first sat down at this most recent meeting and what are some of the ways that it changed?

DR. CARLSON: For those women who have systemic disease, the first branch point relates to whether or not the woman has hormone-responsive disease or not, and also whether or not the sites of disease are threatening or not.

For those women who have non-threatening sites of disease, that we would expect to be hormone-responsive, the guideline calls for the initial use of hormonal therapy. For those women who have hormone receptor-negative disease or who have threatening disease, such as liver disease with liver dysfunction, then the guideline calls for the initial use of chemotherapy. That sort of global outline has existed within the guideline for all the generations that we've gone through.

The current guideline for metastatic disease differs in two fundamental ways. One is that with the recent results, showing that the selective non-steroidal aromatase inhibitors as first-line hormonal therapy appear to be superior to tamoxifen. The use of aromatase inhibitors has been moved forward in the guideline to be a first-line option. Tamoxifen remains a first-line option. We have such a preponderance of data using tamoxifen in that setting that the panel felt that we did not want to remove tamoxifen as a first-line therapy, but the aromatase inhibitors have now moved forward to be an equivalent first-line option for postmenopausal women with hormone-responsive breast cancer.

DR. LOVE: Now, of the three currently available aromatase inhibitors, do the guidelines specify which, if any, of those to be used?

DR. CARLSON: We do only that in the non-steroidal aromatase inhibitors, as anastrozole and letrozole, are considered equivalent according to the guideline. The steroidal aromatase inhibitor, exemestane, is not included in the guideline as a first-line hormonal agent, not because it's not active in that situation. It's just that we don't have good comparative data yet available from large randomized studies that will help us. Those studies are ongoing, but until we have the results of those studies, we felt it was premature to move it into first-line. Another reason that it was not included in the first-line is that there is limited data looking at crossover responses from the non-steroidal to the steroidal aromatase inhibitors, anastrozole or letrozole to exemestane. Very reasonable rates of response are observed in that crossover, whether or not those crossover rates would exist going from exemestane to anastrozole or letrozole, to my knowledge, is not yet known. So, for all of those reasons, we did not include exemestane as a first-line hormonal agent felt to be equivalent to tamoxifen.

DR. LOVE: So, in the postmenopausal woman receiving hormonal therapy first-line, it would be either tamoxifen or anastrozole/letrozole. What about in the premenopausal woman getting hormonal therapy?

DR. CARLSON: In the premenopausal woman receiving hormonal therapy, tamoxifen remained the mainstay of hormonal therapy. The panel did add, this year, consideration for the use of one of the LH-RH agonist/antagonist in addition to tamoxifen, and that's based upon the recent META analysis published in premenopausal women with metastatic breast cancer that did demonstrate that there is some advantage in terms of time to progression and survival, adding an LH-RH agonist to tamoxifen in premenopausal women.

DR. LOVE: And what about an LH-RH agonist alone?

DR. CARLSON: We have not included that as first-line therapy in premenopausal women with breast cancer. It is an option for second- or third- or fourth-line therapy of premenopausal women, however.

DR. LOVE: That's interesting. Why not consider it as first-line therapy?

DR. CARLSON: Well, I think that most of us are hesitant to recommend ovarian ablation in premenopausal women. I think that some of that is historical, relating to the requirement for surgery or pelvic irradiation, historically. Some of it, I think, relates to the need for a monthly injection. Some of it, I think, relates to concerns about health economics. Tamoxifen has been considered now, for about two decades, the preferred first-line hormonal therapy for premenopausal women with metastatic breast cancer, and yet the available data would suggest that oophorectomy is probably equivalent. What is equivalent? And yet we prefer a pharmacologic intervention as opposed to a surgical or radiotherapeutic intervention. I think that that sort of thinking does bias the panel in terms of favoring a selective estrogen receptor modulator with or without ovarian ablation as opposed to ovarian ablation alone.

DR. LOVE: What about the issue - I'm hearing more and more about this - of using an LH-RH agonist plus an aromatase inhibitor. The concept that, as you mentioned, now we're starting to conceptualize at least the non-steroidal aromatase inhibitors as an alternative to tamoxifen as first-line therapy, and the next sort of leap you think about, as well. If a woman's on an LH-RH agonist, she essentially is postmenopausal. Was that discussed by the committee either in the first-, second- or third-line setting, or in any way?

DR. CARLSON: It was discussed briefly by the committee, but there is, for the most part, an almost, not complete, but an almost complete lack of evidence to base that sort of recommendation on. And because we have such high-level evidence for the other hormonal therapies that are available in premenopausal women, the panel thought it was premature to consider a LH-RH agonist/antagonist plus an aromatase inhibitor. I personally believe that that's a very important area for future research and would anticipate randomized trials, looking at LH-RH agonists with or without an aromatase inhibitor in premenopausal women. Those trials, I think, are needed. The superiority of the aromatase inhibitors in postmenopausal women does really suggest that that may be a very effective strategy in premenopausal women. But, again, the evidence to date is simply lacking to support that approach.

DR. LOVE: I think the pressure to develop clinical trial data on that question is going to increase a lot if the ATAC Trial, which is going to be presented next month for the first time, shows that either the Arimidex arm or the Arimidex-tamoxifen arm has any superiority over tamoxifen. And now we're talking about adjuvant, you know, quote "cure," or whatever. It seems like then the question's going to be, "Do I give my premenopausal woman who's still menstruating tamoxifen?" There's already a lot of interest in LH-RH agonists to start with in the adjuvant setting in a premenopausal woman. So, I'm kind of wondering if that's going to get to be a big topic of discussion based on what happens with ATAC.

DR. CARLSON: Well, I suspect that it will. We, actually, at Stanford are participating in a multi-institutional Phase II study that was investigator-initiated now, looking at goserelin plus anastrozole in premenopausal women with metastatic, hormone receptor-positive breast cancer as either first- or second-line hormonal therapy to try to get a sense of are the rates of response encouraging enough to warrant the initiation of a large Phase III trial?

DR. LOVE: That's interesting. Have you actually put patients on that study?

DR. CARLSON: Yes.

DR. LOVE: Any observations?

DR. CARLSON: Well, the study actually is accruing very well. We're accruing fast enough that we don't have objective response data yet, despite the fact that we have a number of women on the trial. I can tell you that I believe that we're seeing responses, but they have not been formally confirmed and/or they have not been maintained for the required protocol duration yet.

DR. LOVE: Now, these women are getting the two drugs for the first time. It's not women - because previously John Robertson had reported women who were progressing on goserelin and tamoxifen, where he stopped the tamoxifen and added in anastrozole, and he saw responses. These women in your Phase II trial are naïve to both drugs?

DR. CARLSON: Yes. They have to be premenopausal, have hormone receptor-positive breast cancer, with functioning ovaries, obviously, because they're premenopausal, and previously had no treatment with an LH-RH agonist or antagonist.

DR. LOVE: And I guess, also, we should, just for completeness, say that aromatase inhibitors in premenopausal women without ovarian suppression is not something that really should be done.

DR. CARLSON: That's absolutely correct. The major source of estrogen in the premenopausal woman is the ovary, not from conversion of adrenaline synthesized androgens to estrogen. And the aromatase inhibitors have no impact whatsoever on estrogen synthesis by the ovary. So, we would anticipate that the aromatase inhibitors would be inactive by themselves in premenopausal women.

DR. LOVE: I would like to go on and talk more about the chemotherapy algorithm in metastatic disease. But, before I do, since we brought up the ATAC trial, and I have to ask every single person I interview up until San Antonio for their predictions, any thoughts on what it's going to show?

DR. CARLSON: My prediction of the ATAC trial is that all three arms are going to be equivalent.

DR. LOVE: Wow!

DR. CARLSON: I think that tamoxifen is a very good drug. Its toxicity profile in the adjuvant setting is so favorable that I think it's going to be very hard to do better than tamoxifen. So, given that, my expectation is that the anastrozole arm will be equivalent and probably not superior to tamoxifen. The experience that we have historically, using combination hormonal therapies, suggests that you don't gain much, if anything, by combination hormonal therapy. So my expectation also is that the combination arm will not be superior to either tamoxifen or anastrozole alone.

DR. LOVE: And when you say that - because this is going to be a very early analysis and it could be that things might change over time - you think that's going to be the bottom line in this trial, period?

DR. CARLSON: I think that will be the bottom line of the trial. It's obviously going to be a very early analysis, and so we will need to see what happens with long-term follow-up. But recall that there are 9,000 women on the ATAC trial, and so if there are small differences very early, we should see them, because the patient numbers are so enormous. And if the differences are small enough that we can't see them with 9,000 women in the trial, they're probably not differences that are going to be biologically significant.

DR. LOVE: It's interesting, though. You said you thought that maybe the anastrozole arm would be the same efficacy-wise. If that turned out to be the case, do you think that maybe in terms of side effects and toxicity - one that seemed to start to come out in the advanced-disease setting was the lack of thrombosis - would actually end up making it a favorable choice?

DR. CARLSON: It very well may be a superior choice based upon toxicity, and one would predict that it actually would be superior based upon toxicity. There are tradeoffs there, though. While there may be lower frequency of thrombotic events and presumably there would be an absence of endometrial stimulation and so forth, the frequency of GI events is slightly higher with anastrozole, and arthralgias also were slightly higher with the aromatase inhibitors. So, it may be a tradeoff from woman to woman about which toxicities are of greatest concern for her. Overall, though, if they're therapeutically equivalent, my expectation is that anastrozole would be the agent to carry forward, primarily based upon toxicity experience.

DR. LOVE: And I guess the other thing would be the combination arm, the toxicity will be interesting. Because there, you might have an agonist effect on the bone from the tamoxifen. We don't know what anastrozole's going to do on the bone. So, again, that might be a situation where you saw equal efficacy, but maybe it would end up being more favorable clinically.

DR. CARLSON: That certainly is a possibility. But, again, my expectation in terms of efficacy is that the three arms will be equivalent.

DR. LOVE: If it does turn out that, either for efficacy reasons or side effect reasons, anastrozole alone or the combination looks more favorable clinically to you, do you think that it would be logical or rational to substitute, for example, letrozole, which is the other steroidal compound that seems to have a lot of similarities, or because of the fact that the adjuvant setting, you can't make that leap?

DR. CARLSON: Well, first of all, my expectation is that if the ATAC trial does show superiority to either anastrozole alone or tamoxifen plus anastrozole, that it's a large enough study. It's a very well designed and conducted study. So that my expectation would be that it will change practice patterns overnight.

Should those practice patterns include letrozole? The purist would say no, that letrozole is different than anastrozole and has subtle differences biochemically and structurally, obviously. Given that, the purist would say no, it should not be substituted. The other data would suggest that the two are so therapeutically equivalent, one to another, however, that I think it would not be totally outside of reason to utilize letrozole.

On the other hand, if the two are equivalent in the metastatic setting, which my interpretation of the data suggests they are, and you have proof that one of them is effective in the adjuvant setting but not the other, then it seems to me you'd lose little by going with the proven drug. And so that's where an evidence-based guideline would take us, and I think that that is what I would personally do in my own practice.

DR. LOVE: Hmm. Interesting. You know it's funny. There have been at least a dozen people, maybe 20, that I've asked that question to, and you're the first one who has actually - everybody else has at least said, "Okay. anastrozole is going to be superior," and a lot of debates about the combination arm. So, I'm gathering - are you predicting San Diego in the Super Bowl? (Laughter) I mean, we might as well. Let's get it all out right now. (Laughter)

DR. CARLSON: Well, Neil, if I'm right, it'll be remembered. (Laughter) If I'm wrong, people will probably forget.

DR. LOVE: Interesting. Incidentally, you mentioned the fact that you interpret the data as those two drugs being equivalent in the metastatic setting. What do you do in your own practice in terms of choosing between those two?

DR. CARLSON: I tend to use anastrozole personally. It was the first of the selective aromatase inhibitors available. It's one that I became very comfortable with, and we all do best and use best what we do or use all the time. So I personally use anastrozole almost exclusively in my practice. I have absolutely no criticism, however, of practitioners who prefer letrozole.

DR. LOVE: Let's go back to the algorithm again. I can just imagine. How long do these meetings take? Like, days or weeks, when you meet with your group?

DR. CARLSON: This last July we met for a day and a half. Individuals on the panel are assigned topics of discussion in advance, so that they can present a 10- or 15-minute sort of summary of the evidence within their specific area. Then the panel follows that up a discussion about the evidence and whether the evidence suggests that the guidelines should be modified or not.

We then, after we modify the guideline, the modified guideline gets distributed to the panel members again for their concern and further reflection after having some distance from the meeting. And then those criticisms or comments are collated and the guideline then modified again. Then the guideline then goes back out to the panel members, and so on. So, it's an iterative process that takes place over two to three months. But the panel meeting itself, the face-to-face meeting, is about a day and a half.

DR. LOVE: I'm going to say that I just admire the group. I mean, really, the work that's come out is incredible. Sometimes I wonder why you do it. (Laughing) Because I know you can't be getting paid enough for that.

DR. CARLSON: I don't get paid at all but the joy of the panel members that I have gotten to know and work with. It's just a superb group of clinicians and scientists on the panel. It's also been very gratifying to see the American Cancer Society and the NCCN modify the guideline for a lay version. It's nice to see patients come into my office with the guideline that I have, in part, been involved in developing, to talk with me about how they should be treated. And they've obviously done their homework and have read it. Of course, it makes my job easier, that they're using a guideline that I've been involved in to approach their decision-making process.

DR. LOVE: Well, I always look forward to seeing it. Now, let's get back to the metastatic algorithm. We probably won't even get to the adjuvant one this time. But one of the things you mentioned was the question of the bifurcation to go to either chemotherapy or hormonal therapy. You mentioned receptor, but also you used the term "life threatening." And there's a lot of debate about what that means, or visceral crisis, et cetera, et cetera. Do you try to define that or is it just sort of left out for the practitioner?

DR. CARLSON: We haven't tried to be real specific about defining what threatening disease is, in part because it's extraordinarily difficult to do and really is one that is based upon clinical judgment. It's based upon the individual woman in terms of her co-morbid conditions, her age, how symptomatic or not symptomatic she is, what her performance status is. All of those considerations get included in whether a woman has threatening disease or not. So, that's one of the situations in the guideline where we, at this point in time anyway, feel that it's too difficult to define.

Now, there are other places in the guideline where we initially were uncertain about how to define, also, and, as we have gained experience, we've become more confident in putting a number or a parameter that one would monitor in terms of making those decisions. But it's very, very difficult.

DR. LOVE: When you ask people clinically about this question, the one thing that I have heard, and I don't know whether you and your group would agree with this, is that the actual chance of response and the type of response you will get in an ER-positive or hormone-sensitive situation or patient, versus chemotherapy, is essentially the same, whether they have visceral mets, non-visceral mets. The difference is the time to response. Does your group sort of support that concept?

DR. CARLSON: Yes. That's why we prefer the use of initial hormonal therapies whenever possible. The toxicity profile of the hormonal therapies is just so much more favorable than the cytotoxic chemotherapies, that in situations where we would expect them to be therapeutically equivalent, the hormonal therapy should be preferred.

In those women who have threatening disease, however, the feeling is that a rapid response in necessary, and most of the women who would have threatening disease, I believe, would be those that would have hormone receptor-negative disease, as well. Or have sites of disease that are less likely to be hormone responsive, such as bulky liver disease of lymphangitic pulmonary disease, which are classically situations in which the tumors are more likely to be hormone refractory.

DR. LOVE: Hmm. That's interesting. Because I've heard people say liver mets that are ER-positive respond the same to hormones and chemotherapy. Would you agree with that statement in general, or not really?

DR. CARLSON: I would agree with that statement if the liver disease is limited.

DR. LOVE: Hmm. Interesting.

DR. CARLSON: I think that there are biases that come into that data in that women who have at their initial presentation for metastatic disease, bulky liver disease, tend to be a different subset of women, than who come to the table with first evidence of recurrent disease with limited hepatic disease. Those women who come with first diagnosis of metastatic disease with limited hepatic disease have typically been diagnosed because they have symptomatic disease elsewhere. For instance, bone or pleural effusion or chest wall recurrence, which would be sites of disease that would be classically hormone-responsive sites. And so I think that there are some biases in terms of the extent of liver disease at the time of presentation in terms of whether or not you would expect the patient to be hormone responsive or not.

DR. LOVE: Interesting. What about age? Does that enter into the decision of hormone therapy versus chemotherapy?

DR. CARLSON: I think it does in the extremes of age. Obviously, for the elderly or for those who have substantial co-morbid conditions, the hormonal therapies are definitely preferred. The elderly have a much higher probability of response to hormonal therapies, but they also tolerate the hormonal therapies so much better. For the very young - and "very young" in this context, I think, is under the age of 30 or 35 - it may be that cytotoxic chemotherapies would be preferred, because those women in those very young age groups may not respond as well to hormonal therapies.

DR. LOVE: Okay. Well, let's move over to the side of the algorithm that's the non-hormonal side in metastatic disease, and I guess the first thing I would think about is where does HER2 come in, HER2 assays and the evaluation there?

DR. CARLSON: This was a topic of heated discussion this year at the panel meeting and in multiple telephone calls and emails that followed.

DR. LOVE: I would imagine that could take a day and a half, just in and of itself. (Laughing)

DR. CARLSON: It probably will next year, as we try to sort out all of the issues. The guideline this year calls for HER2 testing of all breast cancers. And it did that in previous versions of the guideline, at least the most recent version. However, this time we're much more specific. We call for HER2 testing by IHC. And if the IHC result is 2+ by, presumably, the HercepTest™, then we call for FISH analysis.

Now, multiple of the panel members were concerned about that recommendation, in part because of the uncertainty about what do those results predict. We used a level of HER2 expression for prognostic purposes, trying to assess high-risk category of axillary lymph node-negative women who would benefit from adjuvant therapy. We use it in an attempt to predict which women's breast cancers would be particularly sensitive to anthracycline-based adjuvant therapy and, presumably, therefore, anthracycline-based chemotherapy in the metastatic setting. And we also use level of HER2 expression to predict for benefit from trastuzumab, or Herceptin therapy.

There is contradictory evidence in the literature, whether you should use IHC or FISH for all of those three settings, and so we're not terribly specific in the guideline primarily because of lack of definitive evidence about which test should actually be used for which decision point. We called for a FISH testing initially for the IHC 2+ primarily because in the metastatic setting, when you're looking for benefit or lack thereof from trastuzumab, it looks like FISH-positivity is by far the best predictor of responsiveness to trastuzumab, and those women who have IHC 3+ by the HercepTest are almost all FISH-positive, or their breast cancers are FISH-positive, while those who have IHC of zero or 1+ are almost always FISH-negative. And so the real issue is with the 2+ IHC by HercepTest are those women whose breast cancers, FISH overexpressed or not or FISH amplified or not, and, if they aren't, then the probability of response to trastuzumab is quite low.

DR. LOVE: Well, that algorithm, at least the first couple of parts of it, that I've heard a lot, which is if the IHC is 3+ they're her-positive; 2+ go to FISH. But the other thing that you hear a lot, or I've heard a lot, is zero to 1+, but very aggressive course, consider FISH. Did you all discuss that?

DR. CARLSON: We didn't. Part of the issue there is that we could not identify evidence that would suggest that those women who are IHC zero, but FISH-positive, in fact, do respond to trastuzumab. We know that those women who are 2+ by IHC and FISH-amplified, that they have a good chance of responding to trastuzumab. But for those that are zero or 1+ and FISH-amplified, I think the data is lacking in terms of whether or not that would define a subset that we would expect to respond to trastuzumab.

DR. LOVE: So, I guess one other thing, just for completeness, you said your recommendation is that HER2 testing should be done on all breast cancers. I'm assuming you're also talking about primary breast cancers at time of diagnosis or only when the patient has metastatic disease?

DR. CARLSON: The guideline calls for determination of HER2 level of expression for all breast cancers at the time of diagnosis and, for those women who have recurrent or metastatic breast cancer, in whom that information is not available, testing at that time.

DR. LOVE: I would assume, then, that basically what you're saying is every time you get breast cancer tissue, and there are times that it's obtained in the metastatic setting, too, that it should be HER2 tested.

DR. CARLSON: We would recommend that there be HER2 data available on the specific breast cancer. There is a high enough concordance between the HER2 testing from the primary breast cancer and the HER2 testing from the metastatic breast cancer, that it may not be necessary to re-test.

There also seems to be relatively little reason not to re-test, except for economics. So, certainly, doing IHC would be reasonable. Whether or not the expense of FISH analysis is warranted or not is perhaps a more difficult issue.

DR. LOVE: Did you get into the issue of when the HER2 test was done? And specifically, we've only really had the HercepTest and FDA approval the last couple of years. We've had patients who might have had testing four, five, six years ago with perhaps another technique. Did you address that issue?

DR. CARLSON: We have not struggled with that issue yet. It's clear that we will have to. And some of the difficulty there also relates to many of the studies that look at the importance of HER2 overexpression by IHC or amplification by FISH use different methodologies than are licensed. And yet if we're using those studies to justify recommendations in the guideline, but they're with methodologies that are different than we're recommending be performed, what does that tell us about the confidence we have in the evidence that's available. So, we have struggled with that and, at least at the current time, we believe that the evidence is insufficient to really push for one method of analysis over another in terms of this decision-making.

Now, I think that there is evidence that we should use the licensed tests in preference, in part because the reproducibility, especially of the HercepTest, is far superior to the reliability of the HER2 IHC's that are sort of home grown in pathology laboratories across the country.

DR. LOVE: You mentioned HER2 status in terms of chemotherapy, specifically anthracyclines. Do the guidelines deal with the issue of endocrine therapy?

DR. CARLSON: They do not at this point in time, except to say that given the current level of evidence that HER2 should not be used to decide whether or not to utilize hormonal therapies or not in any stage of disease, either the adjuvant or the metastatic setting.
This is obviously an issue that's going to have to be revisited because of the contradictory evidence that is currently available, but at least as the guidelines are now written, we do not look at HER2 level of expression or FISH amplification in making decisions about hormonal therapy.

DR. LOVE: Now, again, in that part of the algorithm, in the non-hormonal part of the algorithm for metastatic disease, what do you specify in terms - obviously, if the patient's HER2 negative, you're dealing with a chemotherapy algorithm. What about the HER2 positive patient?

DR. CARLSON: I'm sorry. The HER2 positive patient in which setting?

DR. LOVE: In metastatic disease. You've ruled out hormonal therapy. Now, obviously, if they're HER2 negative, there's a chemotherapy algorithm, what about the HER2 positive patient?

DR. CARLSON: In the metastatic setting where hormonal therapy is not selected as the initial therapy, the most recent generation of the guideline calls for the use either of single-agent trastuzumab or trastuzumab in combination with chemotherapy. And we are fairly inclusive in terms of the selection of the chemotherapeutic agent. The highest level evidence that we have available looks at trastuzumab in combination with paclitaxel, although the guidelines acknowledges in a footnote that there is non-randomized, therefore, lower level, evidence looking at the use of trastuzumab in combination with docetaxel, vinorelbine, and cisplatin and carboplatin. And so we do allow for the utilization in the guideline of trastuzumab in combination with any of those agents, although, again, the highest level evidence supports the use of paclitaxel in combination with trastuzumab.

DR. LOVE: That's interesting. So, essentially - and actually, I recall Cliff Hudis talking about his clinical management in this situation, what he says is, "Well, I know they're going to get Herceptin as first-line therapy," and then the question is whether they also need chemotherapy. And it sounds like that's what your guideline kind of says.

DR. CARLSON: That's right. And that's based upon the study reported by Chuck Vogel, that looked at trastuzumab as a single agent and found very good rates of response and long duration of response in those women who were IHC 3+ or IHC 2+ and FISH overexpressed, again with a single agent.

DR. LOVE: Just to track down that HER2 line a little bit more, what do you say about the patient who then progresses on trastuzumab plus, let's say, a taxane, in terms of whether or not to stop the trastuzumab?

DR. CARLSON: Neil, that's a great question, and it's one of those situations where the evidence is lacking. At least at this point in time, the guideline is silent on that issue, in terms of how long trastuzumab should be continued once it's initiated.

Home · Search