Select Excerpts from the Discussion

Tracks 1, 2

DR LOVE: Chuck, can you summarize your data with regard to the use of growth factor support during the administration of adjuvant chemotherapy?

DR VOGEL: In our study, docetaxel was chosen as a representative regimen that could cause somewhere around a 20 percent risk of febrile neutropenia at 100 mg/m². All three endpoints — febrile neutropenia, febrile neutropenia-related hospitalization and use of anti-infectives — showed dramatic improvement with the addition of pegfilgrastim (Vogel 2005).

Most people would agree with the new NCCN guidelines (Lyman 2005) stating that prophylactic growth factors should be used for patients with greater than 20 percent risk of febrile neutropenia. The use of prophylactic growth factors should also be considered in the intermediate-risk group, 10 to 20 percent. Patients at low risk should not receive growth factors.

AC followed by docetaxel, AT and TAC (Martin 2005) all have very high febrile neutropenia rates, and prophylactic growth factors should be strongly considered with these regimens. AC is considered an intermediate-risk regimen, as is docetaxel/capecitabine.

FAC, FEC and TC are regimens associated with borderline to low febrile neutropenia rates. Certainly, dose densification of any of these would be a reason to use prophylactic pegfilgrastim, as would the avoidance of dose reductions and delays.

A third reason to consider it would be the risk factors that may cause patients to be at risk for febrile neutropenia.

DR BURSTEIN: I believe most of us would have a hard time consenting to a regimen associated with a one in five chance of a patient being hospitalized with febrile neutropenia compared to one that wasn’t, simply for the administration of prophylaxis. So I don’t find a problem with the recommendation for prophylactic treatment at 15 to 20 percent risk.

The problem is that we as a community haven’t defined an acceptable level of febrile neutropenia. For instance, with nausea and vomiting, we all agree the desired goal is zero, so we liberally use prophylaxis.

For cancer pain, the goal is zero, so we liberally use pain medicine. We haven’t said what we’re willing to tolerate in the way of febrile neutropenia risk.

The only other anecdote I can offer is that as I administer AC every three weeks for patients destined to receive adjuvant trastuzumab, I’m struck by how many patients end up having dose delays and tweaks.

It’s clearly more toxic than using dose-dense AC followed by paclitaxel with growth factor support.

This hasn’t caused me to use G-CSF prophylactically in these settings, but it is impressive how predictable and clockwork-like every two-week AC with growth factor support is compared to other treatments.

I believe if you asked patients whether they would take a growth factor for a four percent decrease in their chance of febrile neutropenia, they’d all say yes. Whether that is cost-effective is a totally different question.

Track 4

DR LOVE: Cliff, what’s your take on the adjuvant trial data that Steve Jones presented at San Antonio comparing docetaxel/cyclophosphamide to AC?

DR HUDIS: We vary in our acceptance of new regimens based on the clinical endpoint bar they cross. Sometimes disease-free survival is absolutely fine. In other settings, people go ballistic if you don’t have overall survival data as well. Here is a setting in which, at the second analysis, we have an improvement in disease-free survival and we still don’t have an overall survival advantage.

I recall that when these data were presented for the first time at ASCO a couple of years ago, with widely separated curves, we were told that the trial would never be statistically significant because it was underpowered. So this result came as a bit of a surprise at San Antonio.

Having said all that, I have a bias. If I were a user of AC, I’d have a hard time not justifying TC. If nothing else, it’s no more acutely toxic, by the randomized comparison, and it certainly should eliminate the small but meaningful long-term risk of cardiac toxicity. It will be interesting to see the long-term leukemia risk without the anthracycline.

DR RAVDIN: The hazard ratio for recurrence shows a 24 percent proportional advantage in survival for docetaxel/cyclophosphamide, which is as big a step as we usually take in our clinical trials, and it shows a 36 percent improvement in disease-free survival. I believe the improvement in overall survival is real, and the correct interpretation isn’t that it doesn’t show a survival advantage but that it’s underpowered to show a 24 percent advantage.

DR WINER: It’s one study, not multiple studies, and it comes on the heels of the negative ECOG trial of AC versus AT (Goldstein 2005). I have a difficult time reconciling those two trials. If, in fact, the substitution of docetaxel for cyclophosphamide wasn’t better, I find it certainly not inconceivable but a little funny that it’s better than an anthracycline.

TC is a fine regimen to use, but I don’t believe that it has to be the standard regimen to replace AC at the moment. I haven’t chosen to use it as a standard regimen other than for patients for whom I don’t want to administer an anthracycline.

DR HUDIS: This makes a point that there’s no evidence that TC is inferior to AC, and it may well be safer in the long term. So I would feel little risk in substituting docetaxel for doxorubicin. I never use AC alone, so it’s easy for me to say that. In my hands, everybody who receives AC also receives paclitaxel.

DR HAYES: If someone called me and said, “I’m going to use TC instead of AC,” I would say, “I believe that’s a perfectly fine regimen.”

Tracks 5, 6

DR LOVE: Cliff, ER status and response to chemotherapy have become a controversial issue. What’s your viewpoint?

DR HUDIS: Don Berry started the discussion of the impact of ER status on chemotherapy outcomes in the modern era by performing an unplanned retrospective analysis of CALGB trials on the basis of ER status (Berry 2006).

He initially presented his three-study analysis at San Antonio in 2004, and compared the high-dose every four-week CAF regimen to the standard AC arm of CALGB-9344. He then studied the AC paclitaxel arm of 9344 against the standard arm of the dose-dense 9741 trial.

For patients with ER-negative disease, the hazard for disease-free survival was significantly improved with each one of these steps — better CAF, addition of paclitaxel, dose-dense scheduling. Adding up the overall impact for ER-negative breast cancer, we see a profound chemotherapy effect.

In the subset of patients with ER-positive disease, the difference in each one of these steps was not statistically significant, but they were always favorable.

The point estimate for benefit is half the size for the patients with ER-positive disease compared to those with ER-negative disease. It is likely that it is still favorable, although the confidence interval does not exclude the possibility of no benefit at all.

To some degree, this has been wildly overinterpreted as suggesting that chemotherapy doesn’t work in patients with ER-positive disease. It simply doesn’t say that. It says that the magnitude of the benefit is likely to be much smaller than for those with ER-poor disease.

The important point is that when people say that the addition of dose-dense scheduling in 9741 doesn’t yield much among patients with ER-positive disease, they’re really not comparing apples to apples when they then look at the TAC-FAC data (Martin 2005).

The TAC-FAC trial demonstrated hazard rates for risk reductions, which looked about the same in the ER-positives and the ER-negatives. The FAC control arm, of course, includes no paclitaxel or docetaxel.

You can’t say that each individual step is or is not significant vis-à-vis another separate randomized trial. You can’t compare these regimens head to head.

If you were to argue that you know to utilize TAC instead of dose-dense ACpaclitaxel in a patient with ER-positive, node-positive disease, then you’re presuming to know the results of NSABP-B-38.

I would argue that there is equipoise on this question and that either regimen is entirely appropriate for patients with ER-positive disease.

Tracks 9, 11, 12

DR LOVE: Eric, what are your thoughts about the influence of ER status on the effects of chemotherapy, particularly with the newer adjuvant regimens?

DR WINER: It’s very hard for me to get more excited about TAC as opposed to dose-dense AC followed by paclitaxel. I believe the bottom line is that if you take all patients with ER-positive breast cancer, the benefits of chemotherapy are dramatically less than in patients with ER-negative disease.

Almost certainly, some groups of women with ER-positive breast cancer derive no benefit and others probably derive every bit as much benefit as the ER-negative group. It’s not going to be chemotherapy-agent-specific, particularly when we get down to the level of taxanes.

DR OSBORNE: This is such an important question because 60 percent of all patients have ER-positive/PR-positive disease. Will anyone conduct a randomized trial of chemotherapy versus no chemotherapy or endocrine therapy alone versus the addition of chemotherapy in that subgroup?

DR HAYES: The patients with node-negative, ER-positive disease in the TAILORx, or ECOG-PACCT-1, study will all be profiled by the Oncotype DX^TM assay. Those patients with a good recurrence score of 11 or lower will receive hormone therapy only.

Those with a high recurrence score of 25 and higher will all receive hormone therapy and chemotherapy of “dealer’s choice.” Those in the intermediate group will be randomly assigned to receive chemotherapy or not (investigator’s choice). They then will all receive hormone therapy, also at the investigator’s choice.

DR SLEDGE: One of the practical implications of this discussion is that it is almost impossible to sort all this out in any clinically reasonable time frame during a patient encounter. It would be wonderful to have strategies to facilitate this because there’s no way that anybody in the community has enough time for these kinds of conversations with the average patient.

Select publications