Select Excerpts from the Interview

Track 2

DR LOVE: Can you discuss the rationale for the combined analysis of the adjuvant trastuzumab trials?

DR WOLMARK: Originally, the NSABP-B-31 trial was going to be an indication trial. The NCCTG-N9831 trial was scheduled to begin later as it really didn’t make sense to conduct the two trials concurrently in a population of patients with node-positive disease who account for only one quarter or perhaps even less of the patient spectrum. As it turned out, by the time all of the regulatory prerequisites were met, the two trials started almost concomitantly.

So there were only a few months between the time the NSABP-B-31 trial started and the time N9831 started. No one had envisioned that would happen; it certainly wasn’t planned. The rate of accrual for a quarter of the population with node-positive disease distributed between not only two competing studies but also a third, if you include the BCIRG 006 study (3.1), was less than ideal.

It became apparent that there was a need to really maintain focus on the goal of these trials, namely, to determine the efficacy of trastuzumab in women with breast cancer. If we were able to answer the question sooner, that was the right thing to do.

The combined analysis did not occur because someone decided, “We’re going to have a look at the data.” This was a very demanding and rigorous process for which a combined analysis plan was submitted to both the Cancer Therapy Evaluation Program and the FDA, as it required a number of changes, including our request to change the endpoint for analysis from overall survival to disease-free survival.

Once the combined analysis was approved, things moved very rapidly. The joint analysis plan was approved in January 2005, and the Data Monitoring Committee met during the third week of April. The first interim analysis indicated the number of requisite events had been surpassed. The requirement was for 355 events, and we actually had 395. The prerequisite for disclosure of the data was a p-value of 10^-3, and we had a p-value of 10^-13.

DR LOVE: Do you have a sense of how much earlier we received the results because of this combined analysis?

DR WOLMARK: We saved a considerable amount of time — probably two years. The NSABP study alone also crossed the boundaries with the proviso that we use disease-free survival as an endpoint, which was not the primary endpoint of the single trial analysis.

DR LOVE: Just to clarify, the NSABP-B-31 trial contained two arms, and the NCCTG-N9831 trial consisted of three arms. The two common arms were used in the combined analysis, correct?

DR WOLMARK: Precisely. The arms were so similar that not to combine them, I believe, would have been a disservice to women with breast cancer.

DR LOVE: When people first heard this analysis was going to be conducted, there were a lot of questions about whether or not this type of evaluation was appropriate. However, after the results revealed such large differences, the analysis was no longer questioned (3.2). What is your interpretation of what occurred?

DR WOLMARK: The methodology used for the combined analysis was absolutely solid. It was the right thing to do because, in essence, the trials were the same trial with some minor variations in the common arms. If the result had been considerably less impressive or only marginal, I still believe the data would have been a ref lection of what was actually happening.

Track 5

DR LOVE: There is some confusion amongst community-based oncologists regarding the issue of concurrent versus sequential trastuzumab/chemotherapy because the HERA study data demonstrate positive results in patients who received trastuzumab after chemotherapy (Piccart- Gebhart 2005; [3.3]). Essentially, there was no benefit in the sequential treatment arm of the NCCTG-N9831 trial. How do you interpret those findings?

DR WOLMARK: The only test of concomitant versus sequential therapy was in the NCCTG-N9831 trial. When you look at the curves in the comparisons of both treatment arms to the control, I do not believe that one can remain neutral. The concomitant arm had a hazard rate that fell in line with what we’re seeing in the other trials, including BCIRG 006 and the combined analysis. However, this is not true of the comparison between the control and sequential arm of N9831, which was associated with a hazard rate of 0.87 (p = 0.29).

The comparison of concomitant treatment with trastuzumab versus sequential treatment with trastuzumab was associated with a hazard rate of 0.64, which was significant (p = 0.01; [Perez 2005a]). It’s not inappropriate for a medical oncologist to look at those data and say they are more impressed with data from the concomitant use of trastuzumab.

Are the results from NCCTG-N9831 inconsistent with the HERA data? Not necessarily. If you look at the hazard in the taxane-treated population in the HERA trial, the least impressive hazard rate occurred in those patients. So the question is whether or not this occurred because they received more cycles of chemotherapy, which delayed the administration of trastuzumab.

The data are not necessarily inconsistent with one another. I think these trials all show there is clearly a benefit of trastuzumab when given concomitantly with chemotherapy, and there may be a benefit sequentially.

Track 6

DR LOVE: What are your thoughts on the issue of delayed trastuzumab in patients who have received chemotherapy in the recent past — six months to two years ago. We’ve already been sensitized to this issue through our experience with the aromatase inhibitors and the time course of disease recurrence. Do you believe it makes sense to look at the risk of recurrence in relation to the continuum of disease course and assume that risk might be decreased if you administer trastuzumab, even if it is delayed?

DR WOLMARK: I think a conditional probability might be a good thing to utilize as a guide to determine whether trastuzumab should or should not be used in a delayed fashion. Would you administer trastuzumab to a patient with five positive nodes, who has completed chemotherapy a year ago knowing or, more specifically, not knowing what the effect is going to be at that point? The majority of clinicians would make that decision on an individual basis, and I would also.

Track 7

DR LOVE: There has been a lot of attention on the distant disease-free survival curve, which was dramatically better with trastuzumab. What are your thoughts about these data?

DR WOLMARK: When you see a curve of distant disease-free survival where the investigational arm — namely, in those patients who have received trastuzumab — is flat, it’s dramatic, particularly so early on in follow-up (3.4). Does this mean that tumor cells have been eradicated? That’s the great hope.

For distant disease-free survival, we are seeing a difference of 90 percent versus 74 percent at four years in the combined analysis — an absolute difference of 16 percent. These results are very impressive. Have we crossed the threshold for opportunities to treat the disease? I think we have.

Track 10

DR LOVE: The initial data from BCIRG 006 have just been released. Do you think that a future data set from this study might show equivalent efficacy of TCH and ACTH?

DR WOLMARK: That is the great hope, but I don’t believe it is likely. We would have loved to have TCH show the same hazard rate as ACTH — that would have made everything much more simple. You would have a noncardiotoxic regimen that shows efficacy in the same range as an anthracycline- containing regimen, so that would have rapidly become the preferred regimen.

According to the press release (BCIRG 2005), the relative reduction in risk of relapse for ACTH was 51 percent and was 39 percent for TCH, so we can’t write eulogies for the ACTH regimen, and we ought not. On the other hand, one is certainly not eliminating TCH as a template to which bevacizumab may be added. It is not unreasonable to consider that as a possibility.

DR LOVE: It’s going to be interesting to see how oncologists and clinical investigators apply these findings to the clinical setting. Do you think it would be reasonable to consider using nonprotocol TCH based on the current data, particularly in the patient at lower risk or the older patient? The regimen clearly has efficacy.

DR WOLMARK: Medical oncologists are going to use the TCH regimen in a reasonable and logical way, as they should. The data that were disclosed relative to hazard rates from BCIRG 006 indicate that TCH is an effective regimen.

We would have liked for the hazard rate to have been the same as ACTH, but it wasn’t. We have seen a benefit when trastuzumab is added to the regimen, so if you have a patient who has cardiac compromise and you don’t wish to use doxorubicin, I certainly think that TCH is a reasonable alternative.

Track 11

DR LOVE: Can you comment on the issue of monitoring cardiac toxicity? A common question is how to treat the patient who has a drop in ejection fraction after receiving AC. Can you discuss what was observed in the NSABP trial and how that translates into clinical practice?

DR WOLMARK: With AC alone, we saw a significant proportion of patients with decreases in ejection fraction. For those patients, I think medical oncologists will be making their decisions, which are going to be mainly driven by risk of recurrence. The higher the risk, the greater the likelihood that the patient is going to be treated. People are going to be innovative in the way they interpret ejection fraction or the algorithm they use.

Track 12

DR LOVE: Can you talk about the NSABP-B-41 neoadjuvant trial, which is being designed for patients with HER2-positive tumors?

DR WOLMARK: We were all certainly focused on Aman Buzdar’s neoadjuvant study, which was a small trial indicating that paclitaxel followed by FEC with concomitant trastuzumab was associated with remarkable pCR rates (Buzdar 2005). This study caught our attention, so we would like to test that regimen in a larger clinical trial. That is what NSABP-B-41 is going to test: the Buzdar regimen compared to a more traditional sequential regimen of FEC followed by a taxane with trastuzumab.

DR LOVE: What is the endpoint of the study?

DR WOLMARK: The primary endpoint is pathologic complete response (pCR). We want to see if the pCR rate from the Buzdar study can be duplicated. Cardiac safety, of course, is another endpoint.

DR LOVE: What is the postsurgical therapy going to be in the B-41 study?

DR WOLMARK: Patients are going to receive trastuzumab for one year. Duplicating the Buzdar results will be a major step. Then — assuming that we do see that level of pCR — we will analyze the components of the regimen that led to the response. The Buzdar regimen was novel in that trastuzumab was administered concomitantly with both the taxane and the FEC.

DR LOVE: Can you discuss the NSABP-B-40 neoadjuvant study of three different arms of preoperative neoadjuvant chemotherapy?

DR WOLMARK: The novelty of B-40 is that we’re using pCR as an endpoint with an emphasis on developing a molecular taxonomy to determine whether or not we can characterize patients who obtain a pCR as a surrogate marker to measure outcome. There is a definite interest in tissue collection in the B-41 study.

Disease-free survival and overall survival are not endpoints for the NSABPB- 40 protocol. We view it as a new mechanism to test promising agents in the neoadjuvant setting, and we think it is an appropriate direction to pursue, particularly with the number of agents that are available and the limited resources, both from a support standpoint and a population standpoint.

Select publications