Select Excerpts from the Interview*

 Track 2

DR LOVE: Your Adjuvant! Online model provides important estimates on the projected benefits of adjuvant therapy. How are you going to include trastuzumab in the model?

DR RAVDIN: There really are three questions that need to be addressed to incorporate this data. The first question is, what are the real efficacy numbers?

What we’ve seen in the combined analysis of the North American data, as projected for three years of follow-up, is a proportional risk reduction of about 53 percent for relapse and about 33 percent for death for those patients who received trastuzumab (Romond 2005; [2.1]). I believe those numbers are still probably unstable in terms of the fact that they represent the impact of the very earliest relapses.

The second question is, how should we use this information? It potentially will impact how the model is used not only for patients with HER2-positive disease but also those with HER2-negative disease. The estimates in the population with HER2-negative disease will have modestly better risk than if it had included the patients with HER2-positive disease.

The third question addresses the toxicity aspect of trastuzumab therapy, that is, how safe is the therapy over a long period of time? We have three years of data at this point, so whether there’s any late cardiac toxicity from the therapy has not been thoroughly addressed. The initial part of the data looks hopeful because most of the toxicity occurs during therapy.

 Track 4

DR LOVE: Do you believe that trastuzumab should now be offered and utilized in the adjuvant setting?

DR RAVDIN: With three trials showing consistent effects, trastuzumab is ready to be used in the adjuvant setting. All of the trials show great consistency. Although the Intergroup and the NSABP trials were combined for analysis (2.1), they were also split out in the presentations so you could see the effects in each one of the trials, and the results were virtually identical (Romond 2005; [2.2]).

The data from the European study — although trastuzumab was administered in a slightly different way — were consistent with trastuzumab having powerful adjuvant effects (Piccart-Gebhart 2005). So I believe we have as good evidence for the early effectiveness of trastuzumab as we have for any other agent right now in breast cancer.

DR LOVE: What about the use of adjuvant trastuzumab in patients with node-negative disease?

DR RAVDIN: The issues for determining when to use trastuzumab are going to be the same as those faced when deciding to use chemotherapy or hormonal therapy. They are going to be a balance between benefit and risk. We have a spectacular early benefit; however, we still need much more information about risks. It’s the patients with Stage I node-negative disease for whom it’s quite possible that the benefit-to-risk ratio will not actually point favorably in the direction of the use of trastuzumab.

 Track 5

DR LOVE: What information do we now have regarding cardiac function and trastuzumab?

DR RAVDIN: In the analysis of the NSABP-B-31 trial, the cumulative hazard of developing congestive heart failure was four percent at three years of follow-up. Most of that four percent occurred within the year of trastuzumab treatment. There were some that occurred afterwards; however, it didn’t appear to be accelerating (Romond 2005; [2.3]).

What we know from anthracycline-based regimens is that they cause an excess incidence of congestive heart failure of somewhere around 0.5 to one percent. I wouldn’t be surprised if trastuzumab doubled that to a level of maybe one to two percent.

If that’s the case, then there may be some patients for whom that risk will outweigh the potential benefits. Although it is also important to note that in the Intergroup and NSABP trials with over 3,000 patients, there has been only one cardiac death so far, which was in the control arm (Romond 2005b).

 Track 6

DR LOVE: In your own practice, will you start trastuzumab during the taxane portion of chemotherapy or wait until the chemotherapy is completed?

DR RAVDIN: I will be starting trastuzumab concurrently with paclitaxel. There really are two pieces of information that address this issue, and they’re somewhat contradictory. One of them is in the Intergroup study. It had an arm that was not included in the NSABP-B-31 study — a sequential arm — in which all of the chemotherapy was administered before trastuzumab was started (Romond 2005a). In this arm, the proportional risk reduction was a lot less obvious and spectacular than with concurrent treatment.

The European data actually point in a somewhat opposite direction. The patients in the European trial completed their chemotherapy before trastuzumab was started, and those patients seemed to gain more than a 40 percent proportional benefit (Piccart- Gebhart 2005; [2.4]). In that case, administering chemotherapy and trastuzumab sequentially seemed to work quite well.

At any rate, there didn’t seem to be any additional risks to administering them concurrently, at least with the safety data we have now. The benefit certainly looked a little bit better — actually, a lot better — in the Intergroup study (Perez 2005). I believe most people will be starting trastuzumab at the same time they’re starting paclitaxel.

 Track 7

DR LOVE: At this point, all of the trastuzumab data are for a total duration of one year. Is that what you recommend?

DR RAVDIN: Yes, and the Europeans are actually taking a look at the duration question. In the European trial, one arm has no trastuzumab, the second arm has one year of trastuzumab after chemotherapy and the third arm has two years of trastuzumab after chemotherapy (Piccart-Gebhart 2005). Because the data at this point addresses one year of trastuzumab, I believe that’s the appropriate length of time.

DR LOVE: Another very common question is the issue of what to do for patients with HER2-positive tumors who were diagnosed six months, one year or three years ago — the impact of delayed implementation of trastuzumab.

DR RAVDIN: In the Intergroup trial, they’re supplying trastuzumab to the control group of patients who want to cross over out to one year of follow-up. I believe there are theoretical arguments that one year is somewhat an arbitrary length of follow-up. Actually, the peak in relapses occurs in patients at approximately two to three years. So I could see a rationale for treating even out beyond a year, particularly for high-risk patients with multiple nodes, although that, of course, is going beyond the data that we have at hand and is somewhat speculative. I can’t see any reason to treat patients who are many years out and have had a HER2-positive tumor.

 Track 8

DR LOVE: How are you planning to approach cardiac monitoring in patients receiving adjuvant trastuzumab?

DR RAVDIN: Both of the North American studies utilized exactly the same cardiac monitoring plan, and it was very thorough. Patients were only eligible if they had a normal left ventricular ejection fraction and no cardiac history. After the first cycles of the AC times four — the anthracycline part of the therapy — they had another left ventricular ejection fraction measured (Romond 2005).

At the end of the anthracycline portion of the regimen, approximately four percent of the patients had left ventricular ejection fractions that had fallen below the normal limits (Perez 2004), so they did not receive trastuzumab (Perez 2005). Patients who went on to receive trastuzumab were a very carefully selected group.

During treatment with trastuzumab, patients underwent a MUGA scan every three months. During the year of treatment, 20 percent of patients had to stop trastuzumab because of an asymptomatic drop in left ventricular ejection fraction or because of developing symptomatic problems (Romond 2005a).

The predictors of developing symptomatic problems were age and a relatively low ejection fraction. For patients who were older than 50 years of age and had a left ventricular ejection fraction after the AC times four that was between 50 and 54 percent, their risk of developing congestive heart failure was 20 percent.

Conversely, for the patients who were younger than 50 years of age and had left ventricular ejection fractions of greater than 65 percent after the AC times four, their risk of congestive heart failure during therapy was 0.6 percent.

It’s clear that there are at-risk populations that need to be identified and monitored during such treatment (Romond 2005; [2.5]).

 Track 10

DR LOVE: Would you comment on the study that was done in Canada to validate the Adjuvant! Online model?

DR RAVDIN: Adjuvant! Online is an evidence-based model, but there are a few studies that have been done to validate it. One of them utilized the excellent population-based database in British Columbia (Olivotto 2005). They have very good records across the province, capturing most of the breast cancer cases and detailed information about them. They sent us the actual patient and tumor descriptions. Based on that information, we generated projections of 10-year survivals for those patients.

We put 4,000 cases into the model. The results of our projections were then sent back to British Columbia, and they matched them with the outcomes. We were very close to the actual numbers, on average, within about one percent (2.6).

One of the areas of inaccuracy was in patients less than 35 years of age, where we debated about whether or not, based on the literature and some of our own analyses, to assign those patients a little bit worse prognosis. The British Columbia analysis would suggest that we should have (Olivotto 2005). Our analysis overall was strongest for breast cancer-specific mortality. Relapse was very strong, but not as strong as mortality. I still feel that mortality is the best endpoint for the model. They felt the model was well validated across all the different subsets and endpoints. It was presented last year at ASCO, and on April 20^th, it was published in the Journal of Clinical Oncology (Olivotto 2005).

 Track 12

DR LOVE: Let’s talk about adjuvant endocrine therapy. Do you think there’s any justification for utilizing five years of adjuvant tamoxifen at this point in postmenopausal women?

DR RAVDIN: I believe very few people are being started on tamoxifen with the intention of receiving five years of tamoxifen and then switching to an aromatase inhibitor. The problem with initially starting on tamoxifen is that strategies that originally start with an aromatase inhibitor will have lower recurrence rates than those starting with tamoxifen.

For instance, if you start a patient on tamoxifen, you’re conceding that she is going to do worse than she would have done on an aromatase inhibitor. Then you have to feel that when you switch her, the curves will then recross.

In other words, the aromatase inhibitor will be so much more effective if delivered later that it will catch up and overtake the group that did not receive the aromatase inhibitor from the beginning. That is possible, theoretically, because tamoxifen and the aromatase inhibitors — since they have somewhat different mechanisms of action — will actually be somewhat noncross resistant.

Therefore, a strategy that employs both agents might provide the most benefit. But that’s a theoretical consideration against the very real fact that we know if you start with an aromatase inhibitor, the patients do better.

DR LOVE: Our CME group did a patterns of care study looking at whether physicians switch postmenopausal women to an aromatase inhibitor after two to three years of tamoxifen. A surprising number of community oncologists continue tamoxifen for five years, whereas breast cancer clinical investigators tend to switch to an aromatase inhibitor after two to three years of tamoxifen. Who’s right?

DR RAVDIN: At this point, it’s not clear who’s right, because it’s not clear which strategy is really better. My money is on the idea that switching is the right thing to do, because we know that the patients whose therapy is switched at two to three years are doing better than the patients who continue on tamoxifen for five years.

I believe much of the benefit of adjuvant tamoxifen actually occurs in the first two years. There’s a little bit of additional benefit for continuing it out to five years. Continuing it beyond that doesn’t apparently provide any additional benefit, at least in NSABP-B-14 (Fisher 2001).

So with time, the benefit of tamoxifen attenuates. If you’re switching at two to three years, what you’re doing is you’re switching away from a drug that’s beginning to be less and less effective. And it looks like you’re obtaining a very effective impact from the aromatase inhibitor administered at that point.

DR LOVE: In the patient who’s going to switch from tamoxifen during the first five years of therapy, how long are you going to use the aromatase inhibitor?

DR RAVDIN: The clinical trial data we have right now use two to three years of an aromatase inhibitor after tamoxifen. There are some trials extending aromatase inhibitor therapy out beyond five years of therapy. MA17 is looking to see whether or not there might be additional benefit. My guess is there will be additional benefit because the aromatase inhibitors, I believe, are not agonists-antagonists like tamoxifen.

Tamoxifen can be reinterpreted by cells, after molecular changes, to be a stimulant, whereas I can’t see how the absence of estrogen could ever be reinterpreted as a stimulant. I would suspect the aromatase inhibitors would be harder to develop resistance to and therefore might be superior in the adjuvant hormonal therapy of breast cancer.

DR LOVE: What about the patient who’s been on tamoxifen for six months, one year or one and a half years? Do you wait until she is at the two-year or three-year point to switch her therapy?

DR RAVDIN: Actually, I wait. At this point I have no patients who have just started tamoxifen. For those patients who have been on tamoxifen for a year, I wait until the two- to three-year point before switching.

 Track 14

DR LOVE: What about hormonal therapy in the premenopausal patient with an ER-positive tumor? There are a number of major international trials looking at this (2.7). How do you approach it right now in a clinical setting? The question a lot of people have is in regard to the patient with many positive nodes or the patient with many positive nodes and an ER-positive, HER2-positive tumor.

DR RAVDIN: The data, of course, are insufficient to decide or we wouldn’t be conducting these trials. Nonetheless, off study, particularly in patients less than 35 years of age, looking at the trial data, it’s apparent that combined hormonal therapy looks like it’s better than single-modality hormonal therapy. Specifically, the kind of combined therapy that’s been looked at up until this point in that scenario is ovarian suppression with an LHRH analog plus tamoxifen.

I actually like that strategy more than the strategy of using such ovarian suppression plus an aromatase inhibitor. The reason why I like that strategy better is because tamoxifen is blocking the estrogen receptor and you’re lowering the estrogen level.

If you’re using both ovarian suppression and also the aromatase inhibitor, you’re using two things to reduce estrogen, and you’re not really attacking two different parts of the chain. Whether that is the right way to think will be, I believe, demonstrated by the ongoing trials, which are comparing the two strategies.

 Track 17

DR LOVE: The 2000 overview (EBCTCG 2005) was published recently. Is there anything in this publication that you think is important for clinicians in practice to know about?

DR RAVDIN: Actually, there’s nothing in it that I believe is strikingly different from the overview that was published in 1998. I believe what is important about the overview is it shows that we’re not delaying recurrence and death, but rather adjuvant therapy seems to be preventing it.

In other words, the relapse-free survival and the overall survival advantages that accrue early in the patients’ clinical course actually are maintained out to 15 years (EBCTCG 2005). That is really good news, because there don’t appear to be late toxicities eroding those advantages. I believe that’s one thing the overview says that’s powerfully important.

There are a lot of data in the overview about anthracyclines and some of the anthracycline-based regimens. What made the headlines is if you use a good chemotherapy regimen plus a good hormonal regimen, you reduce your risk of mortality by about 50 percent (2.8). So that’s something we all knew, but the publication caused a special notice.

Select publications