Edited excerpts from the discussion:

DR STEINEKER: This woman was 76 years old and presented a year ago in July 2004 with a Grade III DCIS in the left breast, for which she had a modified radical mastectomy.

In February 2005, she was diagnosed with an infiltrating ductal carcinoma, and a right modified radical mastectomy was performed, showing a Grade III ER/PR-negative, HER2- positive tumor. The disease was staged as T2/N0/M0, and her primary tumor was 3.5 centimeters.

I recommended the CALGB protocol 49909, with chemotherapy and trastuzumab. She was randomly assigned to Arm B, which was AC followed by paclitaxel followed by trastuzumab. That trial is a little bit of a challenge because the investigators want to check the FISH results, so the patient doesn’t find out until they’re pretty far into the protocol whether or not they’re going to receive the trastuzumab.

She was treated with AC and had a lot of trouble with that. I have not seen colitis before, but she had diarrhea to the point where we had to hospitalize her for a week. She did have a preexisting history of diverticulosis but did not have any particular problem with that. According to the protocol, we had to reduce her doses, and she completed four cycles of AC. She then started on paclitaxel. At the time I wrote this up, she was into her ninth week, with a fair amount of colitis and diarrhea despite Imodium®. In fact, we had to hold the paclitaxel for one week to let her recover. Her blood counts were always good.

The dilemma is that, according to the protocol, after the AC she needed a MUGA. Compared to before the chemotherapy, her MUGA had dropped from 66 to 50 percent, and the protocol had a cutoff of 15 percentage points. She was therefore excluded from consideration for trastuzumab, no matter what happened to the ejection fractions.

Out of curiosity, though, I repeated her MUGA one month after the AC and her MUGA showed 71 percent left ventricular ejection fraction, but to be careful, I did a 2-D Echo, and this showed 45 percent. Anyway, according to the protocol, she can’t receive trastuzumab. She was at relatively low risk, because she was lymph node-negative.

DR LOVE: Eric, what do we know about variations in ejection fractions and 2-D Echos?

DR WINER: I don’t pretend to be an expert in terms of the cardiac toxicity of these agents. Clearly, measurements of cardiac function can bounce around, and as Barbara said earlier, getting MUGAs doesn’t prevent heart failure, although it gives us some clues about who may be at greater risk. I would be concerned about giving her trastuzumab, given the fact that in the study the only data we have in terms of cardiac toxicity are from patients whose ejection fractions didn’t drop 15-plus percent. Although you had that reassuring second MUGA, you have an Echo that gives you a very different result. So I think there is reason to believe that her ejection fraction has dropped with the anthracycline.

The issue of age is hard to deal with, because on one hand, one doesn’t want to discriminate against older women. On the other hand, you want to include that information appropriately in your decision- making. Given the fact that events in HER2-positive patients do tend to be earlier rather than later, it is likely that her life expectancy will be such that she will not die before she has a recurrence, if she’s destined to have a recurrence. On the other hand, this issue of toxicity is a big one, and toxicity with one agent often correlates with toxicity with another. I would just be quite concerned.

I think the more interesting situation would be if she were 45 years old or if she were 76 and had 10 positive lymph nodes. In those situations, I would probably cautiously and with a lot of discussion think about using trastuzumab.

DR LOVE: Kevin, how are you approaching the issue of trastuzumab for patients with node-negative tumors?

DR FOX: NCCTG-N9831 did include node-negative patients, although not many. My feeling is that if they meet the criteria for entry onto the study, they should be entitled to the benefits of the therapy, if there are no extenuating circumstances.

With respect to your patient, something came to mind that I think everybody probably has seen by now. When the NSABP did their detailed cardiac analysis of the first 1,000 patients, this grid was produced that has made the rounds, which perhaps you’ve seen, which tries to identify the patients who are at the highest risk for developing trastuzumab-related cardiac problems. In that grid, they positioned patients based on their post-AC ejection fraction, and if someone had fallen into the range of 50 to 54 percent, from another number, if they had not fallen more than 15 points, they were still eligible to go on, and they did.

If they fell to an ejection fraction range between 50 and 54 percent and were over the age of 50, in that small group of patients, of which there were 47, the cardiac event rate was 20 percent. That was the subset that stood out as being uniquely susceptible to the ill effects of trastuzumab. Your patient is elderly and showed a substantial drop in ejection fraction from AC alone. That alone would raise a red flag, that she has the potential, even under the best of circumstances, for trouble later on. So I, too, would be inherently reluctant to use trastuzumab.

However, since we’re confessing things today, I have had two younger patients on clinical trials whose ejection fraction dropped as yours did, unacceptably, and were denied trastuzumab on the clinical trial. These were people who, getting back to the other case, were enrolled in 2003 and now are presenting with normalized ejection fractions and the same question that your patient asked.

DR WINER: It would be a more interesting dilemma if she had a number of positive lymph nodes or if she were much younger. In those situations, I think that I would approach this situation a little bit less stringently. In this case, I think that I’d fall back on the guidelines in the trial and probably not use trastuzumab.

DR STEINECKER: This lady had so much trouble, she was so happy to hear she wouldn’t be getting the trastuzumab once a week for the next year that she was smiling from ear to ear — for better or worse.

DR WINER: The CALGB has had this ongoing trial that is slowly accruing older women, comparing capecitabine with either CMF or AC — “dealer’s choice.” Presumably, most patients will get AC. This study is of women over the age of 65 and now allows women to receive trastuzumab after the completion of chemotherapy. If I were seeing this woman today, I might well encourage her to enroll in the trial with the idea that after the completion of chemotherapy, with the knowledge of HERA, we would use trastuzumab as a single agent at that point.

DR REEVES: I’m concerned about subclinical heart disease with trastuzumab. Should we be considering substituting epirubicin for doxorubicin? Or should we be considering dexrazoxane to protect people, if it’s going to exclude our use of trastuzumab, which we know now makes such a difference in patients with HER2-positive disease?

DR WINER: This is a good question. The results of the BCIRG trial imply that we’re not ready to get rid of an anthracycline yet, and I think an important question is whether we can make the anthracycline less toxic. The CALGB had a study that enrolled all of two patients about six years ago, evaluating dexrazoxane and a variety of other questions in the locally advanced setting with trastuzumab. MD Anderson has been using epirubicin concurrently with trastuzumab. Other, less toxic anthracyclines, the liposomal preparations, both Doxil® and a drug that’s commonly called D-99, have yet to be approved in the United States, and may or may not ever be approved. I believe it’s an important question and, of course, if you could get the benefit from an anthracycline with less concern about cardiac toxicity, so much the better. Outside of a trial, I wouldn’t be in a rush to give dexrazoxane, but I think it’s an important research question.

Everybody was hoping that the BCIRG study would make this simple — that TCH would be better, we wouldn’t be giving anthracyclines any more and we’d be moving on to a new era, in which preclinical assays would predict how patients were going to do on a routine basis. I don’t think we’re quite there.

DR LOVE: It never works out that easily, does it?

DR FOX: I don’t want this to come out wrong, but how many people were really put in harm’s way, clinically, by the cardiotoxicity (8.1)? I’m not diminishing the fact that there were a couple of cardiac deaths. In the HERA trial, the rate of congestive heart failure for those who received sequential therapy was remarkably low, and maybe we ought not allow this to be more of a concern than it merits. This was a large clinical trial where far fewer people came into harm’s way from cardiac toxicity than they did from recurrence and death from breast cancer had they not received trastuzumab.

With stringent monitoring, if we all just adhere consistently to the protocol criteria and we pay close attention to this post-AC ejection fraction, which appears maybe to be quite important, I don’t think we’ll have too much trouble, and we won’t need to use more expensive therapies like epirubicin and dexrazoxane to offset the problem.

DR LOVE: Eric, what kind of clinical scenario in terms of risk for cardiovascular disease would have you considering TCH right now, and how much of a clinical history would make you say, even in a node-positive patient, “I’m not going to use trastuzumab”?

DR WINER: As long as a patient had an acceptable ejection fraction level, there isn’t any cardiac risk factor that would push me in the direction of not using trastuzumab in that setting.

I think where this becomes an issue is with a patient with a 1.3-centimeter ER/PR-negative, node-negative tumor, who was just barely eligible both for HERA and for the Intergroup trial, or the patient who’s got a slightly larger ER-positive tumor, who might have been eligible for HERA but wasn’t eligible for the Intergroup trial. In those situations, there’s good reason to think that the relative risk reduction with trastuzumab will be the same across the board, but it’s not relative risk reduction that should push us to decide to give a therapy. It’s the absolute reduction. That’s where you’ve got to really think about the toxicity issues.

DR LOVE: So what about a patient with normal ejection fraction but a history of a couple of MIs and hypertension?

DR WINER: In the setting of a normal ejection fraction in somebody who’s got a relatively high risk of recurrence — node-positive disease — I would probably use trastuzumab in those patients who were eligible for the trial.

DR LOVE: Which chemotherapy?

DR WINER: Would I give TCH or AC followed by paclitaxel? The problem is that I think TCH is a pretty toxic regimen, and I’m not sure in which patients I would use it.

I suppose I would use it in a younger woman who I thought could tolerate the other toxicities, other than the cardiac issues, who either has a borderline ejection fraction — and occasionally, you’ll come across somebody, a 38- or 42-year-old woman, who has an ejection fraction of 48 percent — or someone, perhaps, who’s had Hodgkin’s disease before and had mantle irradiation, and I’m more concerned about cardiac issues. At the moment, I’m not quite sure for whom I would use it.

DR FOX: With these questions in mind, there was actually a case that’s come up just in the last week. I treated a patient in 1989 for a T2/N1, ER-negative right breast cancer. She received six cycles of CAF, 360 mg/m² of doxorubicin and has been fine. She has a normal ejection fraction currently but has a contralateral breast cancer that is 2.5 centimeters, ER-negative, HER2-positive, node-negative. I’m not going to give her anthracyclines. Would I administer TCH? It looks like maybe it’s my default position, as much as my enthusiasm for doing it has gone down.

DR WINER: I think Kevin’s case is perhaps the best of all of them in terms of the situation in which I would administer TCH, which is for someone who can’t receive an anthracycline. It’s not even an issue. You’re not going to give it to this woman who’s had 360 mg/m² of doxorubicin. At the moment, this is someone to whom I absolutely would give TCH.

DR FOX: Now, had I treated her two years later, when the patterns of care were changing a bit, I probably would have given her AC, and then it would be 240 mg/m². Would that make you feel any differently?

DR WINER: I think I’d still give her TCH today.

DR LOVE: Would you have given her TCH before the press release came out from the BCIRG?

DR WINER: I probably would have because, at a minimum, you can’t give her an anthracycline, and it’s a regimen that we know has been given to a large group of women, and so there’s at least a toxicity experience that’s been amassed.

DR LOVE: Would you be surprised if it turns out that TCH is not as effective as the anthracycline-containing regimen?

DR WINER: Right now, we don’t know that it isn’t as effective. But if it turns out to be the case, I think the real question will be which patients need an anthracycline in the HER2-positive setting and which don’t.

It doesn’t surprise me that a subgroup of patients exists who benefit from an anthracycline, given all of the data that suggest that the benefits of anthracycline- based regimens, compared to CMF, are largely confined to women with HER2-positive disease, albeit retrospective — but convincing — analyses.

We need to spend time focusing on molecular predictors of resistance to trastuzumab. As a medical community and a breast cancer research community, it’s almost shameful that we haven’t figured this out, in spite of the fact that we’ve been using trastuzumab in the metastatic setting for seven or eight years.

Much of the problem relates to the fact that it just hasn’t been in our practice approach to perform biopsies at the time of relapse. If we had 100 patients who had been on trastuzumab and experienced disease progression, had biopsies, and we had interrogated the tissue, we might have some clues. But we do have clues based on preclinical data.

DR LOVE: The C-Myc data done by Soonmyung Paik was able to separate out a group that had a greater than 90 percent chance of remaining disease free, as opposed to a group whose chance of remaining disease free was in the 60s. It reminds me of some of the work he’s done with Genomic Health, which has some teeth to it in terms of decision-making. Maybe we’ll have trials that’ll focus on the people who we think have high relapse rates.

DR WINER: Right, and those are the patients for whom you might consider strategies such as other HER2-directed therapies in place of trastuzumab or therapies in addition to trastuzumab.

DR LOVE: Kevin, what were your thoughts about Edith Perez’s NCCTG data suggesting an advantage to concurrent versus sequential chemo-trastuzumab?

DR FOX: She did, indeed, draw that conclusion, but I think she was quick to point out that at that point in time, the number of recurrences was very small. So it’s too early to conclude that we know the worth or lack of worth of sequential versus concurrent treatment. We just don’t know.

DR LOVE: There is a lot of confusion about this exact point, because the HERA study, which is seemingly a similar sequential strategy, showed a 50 percent reduction in relapse rate.

DR WINER: I think the HERA results are impressive and stand on their own without a lot of difficulty. It is quite possible that concurrent may be better than sequential, but we don’t know that at the moment. The only reason we know anything from N9831 about sequential versus concurrent therapy is that when the DSMV met and decided to release the data about trastuzumab, as a practice management question in terms of what to tell doctors whose patients were on the trial, they asked to look at those two arms, so that they could give doctors a sense of what to do for those patients who had been treated on the trial and didn’t receive trastuzumab.

While there is a statistically significant difference between the concurrent and sequential arms on Edith’s trial, and the sequential arm wasn’t significantly better than no trastuzumab, it did not meet any boundary in terms of early stopping. I think that we just need more data.

DR STEINECKER: So if Edith’s study shows no benefit for that sequence, then the timing of one year, two years, is shot.

DR WINER: We know there’s benefit from HERA. The risk reduction in HERA was similar to what we’ve seen in all of the studies. All of these studies — other than that one arm in N9831 — have shown that the use of trastuzumab either with or following chemotherapy reduces the risk of disease recurrence by about half, and the results are shockingly consistent.

DR LOVE: At the NSABP meeting, people were talking about the fact that most of the HERA patients did not receive taxanes. Could that be confounding this? If you’re going to give the patient a taxane, isn’t it going to bump up the efficacy and, in some way, dampen what you’d see with trastuzumab?

DR WINER: Perhaps. What has been stated incorrectly is that most of the patients in HERA didn’t receive an anthracycline. In fact, 92 or 94 percent of patients in HERA received an anthracycline, and I think about a third of them received a taxane.

It’s very difficult to compare across trials. I think there’s a signal from Edith’s trial. I’m not saying it should be ignored, but I would not conclude, based on her data, particularly considering HERA, that trastuzumab after chemotherapy is ineffective. That would be an inappropriate conclusion. I think an appropriate conclusion is that — particularly if you’re giving a taxane — sequential therapy isn’t as good as concurrent, but we have to see.

DR LOVE: The schedule of how the paclitaxel was administered originally was weekly, and then every three weeks. Does that make a difference?

DR WINER: In N9831, it was administered as a weekly regimen. In the NSABP trial, it was initially administered every three weeks, and then the study was amended and it was allowed to be administered weekly. I don’t know whether it matters or not. Certainly the suggestion exists that weekly paclitaxel is better than every three-week paclitaxel. Whether that matters in the context of HER2-positive disease and with concurrent or sequential trastuzumab, we don’t know. If I were going to use it, I would tend to use the weekly schedule as was done in the studies.

Select publications