DR LOVE: The use of Peter Ravdin’s Adjuvant! program and some of the other computer models is a recent phenomenon. Interestingly, it now looks like more than half of all oncologists either have used or are using these models. Joyce, do you use a computer model in making clinical decisions (Figures 14-16)?

DR O’SHAUGHNESSY: I sometimes use the Ravdin model. I’m surprised to see that over half of the oncologists who responded have used these models — that’s higher than I would have expected.

DR LOVE: You can see that in addition to providing patients with risk estimates, the most common use is treatment decisions in node-negative cases. What do you think that means?

DR O’SHAUGHNESSY: I believe it reflects that clinicians are struggling with treatment decisions, particularly in the patients with receptor-positive, nodenegative breast cancer. I use Ravdin’s model to give patients a quantitative estimate of their risk based on an authoritative source rather than just my opinion, particularly in these gray areas.

DR LOVE: Bob, do you have any additional thoughts on these numbers or the use of these models?

DR CARLSON: I am really pleased that the percentage of practitioners actually using computer-based models is as high as 60 percent. My expectation is that the number is rapidly increasing and it seems that people who have used these models in practice use them quite frequently. What made the use of Adjuvant! widespread was distribution — a diskette version is available at no cost to medical oncologists across the country.

I have found that it is difficult to convince practitioners to try these models; however, when they do, I believe that they see the power of the numbers and how the presentation of absolute benefits to the patient can make decision-making an easier and much more objective process.

DR LOVE: I believe these models have changed the culture of medical oncology especially with regard to the way the benefits of adjuvant therapy are being presented in terms of absolute versus relative risk.

DR CARLSON: I think you’re right. Historically, most of us tried to present both the relative and the absolute benefits of therapy. But when it boils down to decision-making, patients really don’t care about relative benefits. They want to know the absolute numbers. Having a model like Adjuvant! allows us to estimate absolute benefit much faster and more reliably.

DR O’SHAUGHNESSY: I believe that’s true, particularly in the low-risk scenario. In the node-positive, higherrisk situation, I still use the relative risk, but when it comes down to one, two or three points, I always use absolute. While too many variables exist for these models to be totally germane to one patient, I believe as a general rule they are very positive.

DR LOVE: Do you use the models yourself?

DR CARLSON: I use these models for every patient who comes to me for a discussion of adjuvant therapy. For the past two years I have printed out the results and I usually give them to the patient. I love the Adjuvant! model because it helps me avoid biases. Many factors influence how physicians think about a specific patient — personality type, type of relationship that is established, referral source — these models totally remove those from the equation.

DR LOVE: We presented a case in this survey (Figure 17) in which we asked doctors how they would estimate the patient’s 10-year risk of relapse and mortality and then we calculated the numbers using the Adjuvant! model. It’s interesting how closely they match.

DR CARLSON: That surprises me too. It makes me wonder if the use of these models has resulted in more education and now physicians are able to estimate the absolute risk of relapse and of death more accurately.

It would be fascinating to see this table broken into two parts: one evaluating the 60 percent of the practitioners who said that they have used the computerbased models and the other showing the 40 percent who have not used them. My prediction is that we would see more accuracy and consistency in the group that uses these models than we do in the group that does not.

DR LOVE: We actually did compare those groups and, surprisingly, we did not find many differences. Joyce, how would you estimate the risk of relapse in this patient?

DR O’SHAUGHNESSY: Without adjuvant therapy, I believe this patient has approximately a 12 percent risk. Adjuvant therapy would cut that risk in half; however, she would receive little benefit from chemotherapy — possibly zero benefit.

DR LOVE: The physicians who responded, and Adjuvant!, gave higher estimates of risk. Are they overestimating the risk?

DR O’SHAUGHNESSY: I fine tune risk a bit more. The size difference in Adjuvant! is between a 1.1-centimeter tumor and a 1.9-centimeter tumor. I estimate that a one-centimeter tumor has a 10 percent risk of relapse and a 12-millimeter tumor has a 12 percent risk of relapse.

DR LOVE: Regarding the question of chemotherapy (Figures 18 and 19), I am interested in your perspective on what physicians are doing in practice — especially the number that are using dose-dense AC. Also what would be your choices for chemotherapy in this situation at varying ages?

DR CARLSON: I am surprised by how many physicians are using AC alone, especially for a very young woman with what I would view as a substantial risk of relapse. I would have thought people would be more aggressive, either adding a taxane or considering dosedense adjuvant therapy. While all of these therapies offer proportional risk reductions, the addition of a taxane or dose-dense therapy does increase that risk reduction.

For me it would be a 50-50 split between AC for four cycles every three weeks and dose-dense AC followed by paclitaxel. I have not used dose-dense AC without paclitaxel in part because I believe if the risk is great enough to warrant the use of dose-dense therapy, then it is presumably great enough to add the taxane. I am cautious because we don’t have any prospective randomized data evaluating the utility of dose-dense AC by itself.

As the age of the patient increases, obviously the magnitude of absolute benefit decreases; therefore, I think as women grow older, fewer are willing to accept chemotherapy.

However, if you compare AC every three weeks to AC followed by paclitaxel in a dose-dense fashion, the magnitude of benefit that you achieve by using AC is at least matched by the addition of the taxane and the dose-dense application of it.

I think that most women who are willing to accept AC for the absolute improvement in benefit would also be willing to accept the addition of a taxane in a dosedense fashion.

In my practice a 65-year-old woman would receive chemotherapy because the benefits are relatively substantial. I would offer her two alternatives — AC or dose-dense AC followed by paclitaxel. For an 85-year-old woman, however, I think few of us would be anxious to give her AC because of toxicity concerns.

DR LOVE: What about at age 75?

DR CARLSON: At age seventy-five it is a tough discussion and is influenced by comorbidities that may or may not exist, the woman’s philosophy of life and how much toxicity she would be willing to tolerate for a very modest, but presumably real, gain.

DR LOVE: Joyce, which chemotherapy would you use in this clinical scenario?

DR O’SHAUGHNESSY: I generally don’t use chemotherapy in postmenopausal patients with ER-positive, node-negative disease unless bad prognostic factors exist, such as a tumor that is HER2- positive, has a high proliferative fraction or is Grade III.

DR LOVE: When you use chemotherapy in a node-negative patient, which regimen do you use?

DR O’SHAUGHNESSY: I generally use six cycles of FAC. In the younger patient, age 55 or younger, I discuss TAC with the patient.

DR LOVE: Why TAC as opposed to AC/ docetaxel?

DR O’SHAUGHNESSY: When I use TAC in that situation, I sometimes use only four cycles. AC/docetaxel is six months of therapy and if a patient is eligible for that regimen, then I usually enroll her in our clinical trial. I tend to use AC/docetaxel in patients with highervolume breast cancer, either T2 or nodepositive. For patients age 65 and older, I generally don’t use chemotherapy in this scenario.

DR LOVE: How would you treat the same patient if the tumor was HER2-positive (Figure 20)?

DR O’SHAUGHNESSY: I would use an anthracycline-based regimen for the patient with HER2-positive breast cancer.

DR CARLSON: It appears as though physicians are much more likely to use a taxane or dose-dense therapy for patients with HER2-positive disease. I think that is likely due to the perception that women with HER2-overexpressed breast cancer have a higher probability of recurrence.

DR LOVE: Do you agree with that?

DR CARLSON: I think it is probably true, but the available studies that have examined the issue retrospectively are somewhat contradictory. If we evaluate this in one univaried analysis, HER2 is definitely prognostic. But after you correct for lymph-node status, tumor size, degree of differentiation and so on, whether or not it remains independently prognostic, I am not yet convinced.

DR LOVE: In your own practice, do you tend to push your patients toward taxanes a little bit more if their tumor is HER2-positive?

DR CARLSON: I do, and I tell them what I just told you. I also provide an Adjuvant! estimate and explain that the estimate does not incorporate the level of HER2-overexpression. If their tumor is HER2-overexpressed, the estimates from Adjuvant! in terms of outcome are high and, therefore, the estimates for benefit from therapy are probably slightly low.

DR LOVE: The next case (Figure 21) is a woman in average health with a 1.2- centimeter, ER-positive, HER2-negative tumor and three positive nodes. This is another situation in which we asked physicians to estimate risk of relapse, and here we see some disparity at least in terms of the baseline risk of recurrence. I would have guessed it would be a little higher. Do you think a 34 percent risk of relapse sounds right for this patient?

DR CARLSON: That is what I would tell my patients because I quite literally use Adjuvant! for every adjuvant patient I see.

DR LOVE: It suggests that maybe a prognosis of node-positive disease is not as bad as we thought.

DR CARLSON: I think what it tells us is that axillary lymph node status in isolation is not as strongly predictive as we would have expected.

In women with positive lymph nodes, the degree of differentiation, hormonereceptor status, size of the tumor and so forth are also independently prognostic and perhaps have a greater influence on the ultimate prognosis than we previously gave them credit for.

DR LOVE: That’s interesting. When you use Adjuvant! within the node-positive population do you notice significant shifts based on grade and tumor size? I have never played around with the model in that respect.

DR CARLSON: Yes, you do see shifts and it is not difficult to find patients who are node-negative who have a risk of recurrence that is greater than some of the node-positive subsets.

I expect as these models are used more and more over the next few years we will see a culture shift and won’t be stratifying patients in our brains or in our practices based on nodal status as much as we will base it on a risk estimate.

DR LOVE: Joyce, which chemotherapy regimen would you use for a patient like this with three positive nodes (Figure 22)?

DR O’SHAUGHNESSY: My standard treatment in this case would be AC followed by docetaxel. I offer the same regimen to patients ages 70 and over, but I usually begin the docetaxel at 75 mg/m2 rather than 100 mg/m2. If they do well, then I increase the docetaxel to 85 mg/m2.

One study I’m intrigued with is a trial that Denise Yardley and Skip Burris are planning, of TAC versus dose-dense chemotherapy starting with docetaxel at 100 mg/m2 every two weeks for four cycles followed by AC every two weeks for four cycles. Filgrastim is used in this trial.

DR LOVE: Wow! Has docetaxel every two weeks been evaluated?

DR O’SHAUGHNESSY: Yes. George Raptis published a paper in Anti-Cancer Drugs evaluating preoperative docetaxel 100 mg/m² every two weeks with a good PCR rate demonstrating feasibility. However, docetaxel 100 mg/m² every two weeks following AC is not feasible because of skin toxicity, but you can administer the docetaxel first. This results in some skin toxicity, but it’s at the tolerable level.

DR LOVE: What strikes me about the use of chemotherapy for this type of patient with node-positive disease is the amount of dose-dense AC T being used and a reasonable amount of docetaxel given in different ways.

DR CARLSON: It is interesting to me how quickly dose-dense AC T has become adopted and how widely it is used. I view that as good news because it looks like we’re doing a good job of translating research findings into community practice in a timely fashion.

DR LOVE: Joyce, what are your thoughts about the dose-dense data?

DR O’SHAUGHNESSY: I’m the principal investigator of the US Oncology trial of AC followed by docetaxel versus AC followed by docetaxel/capecitabine, so I spend a lot of my day cheerleading for that study. Therefore, I’m biased.

I’ve used dose-dense AC followed by paclitaxel, but I tend to use it only in patients whose disease is not high enough risk to enroll in our clinical trial. I have a patient who is a 40-yearold nurse with a one-centimeter, Grade III, Ki-67, 80 to 90 percent, ER-/PRand HER2-negative tumor and negative nodes.

While that’s not a “good” cancer, she was not eligible for our trial because the tumor was not staged as T1C. If her tumor had been 1.1 centimeter, she would have been eligible, but she wasn’t eligible and I wasn’t going to treat her with AC or FAC. She has a “bad” cancer, so I gave her dose-dense AC followed by paclitaxel.

I have another patient with multiple myeloma in addition to breast cancer. I plan to use hematopoetic growth factor support, so I’m treating her with dosedense AC followed by paclitaxel.

So, yes, I do use the dose-dense regimen for selected patients.

DR LOVE: Bob, would you present AC alone or CMF to a patient in this situation as options?

DR CARLSON: If a woman who has three positive lymph nodes has a risk of recurrence that is substantial enough in an absolute sense to warrant the use of chemotherapy, then the addition of the taxane or a dose-dense regimen or both provides an advantage that is at least equivalent to the addition of AC or CMF.

If a woman is willing to accept AC for four cycles for the absolute benefit she will derive, then she is almost certainly going to accept another four cycles of treatment with a taxane.

DR LOVE: Do you ever use TAC off protocol?

DR CARLSON: No, I don’t.

DR LOVE: It’s interesting that people almost always use growth factors when they use TAC. It seems that the message has gotten through.

DR CARLSON: I think that is probably one of the reasons why the TAC regimen is not more widely used. People initially didn’t use growth factors and had bad experiences.

DR LOVE: Joyce, if this same patient were HER2-positive, I assume that wouldn’t change your treatment plan with chemotherapy. Is that correct (Figure 23)?

DR O’SHAUGHNESSY: Correct.

DR LOVE: We asked these clinicians how many times a month they start a breast cancer patient on a taxane-containing adjuvant regimen, the response was about twice a week (Figure 24). A significant percentage of clinicians reported they use AC followed by docetaxel for adjuvant taxane therapy (Figure 25). Is that your practice?

DR O’SHAUGHNESSY: In a nonprotocol setting, my standard regimen is AC followed by docetaxel.

DR LOVE: What has been your experience with nanoparticle paclitaxel?

DR O’SHAUGHNESSY: I have treated approximately seven patients with this agent and I’ve found it’s extremely well tolerated, particularly at the 100 mg/m² dose. In the Phase II trial with 125 mg/m², I had two patients who experienced either significant fatigue or some neuropathy with this higher dose. I like the 100 mg/m² dose because I see very little myelosuppression or fatigue and I can’t recall any patients experiencing peripheral neuropathy.

DR LOVE: I assume you don’t premedicate patients receiving nanoparticle paclitaxel with the taxane.

DR O’SHAUGHNESSY: That is correct; I don’t believe weekly dexamethasone is good for patients — it tires them and has a crash effect. Avoiding the premedication may be one of the reasons why we don’t see significant side effects with nanoparticle paclitaxel.

DR LOVE: Would you use nanoparticle paclitaxel in the adjuvant setting off protocol?

DR O’SHAUGHNESSY: Without data, I am not generally willing to substitute nanoparticle paclitaxel for docetaxel off protocol in the adjuvant setting. In a middle-aged patient with numerous comorbidities whom I can’t give TAC because of the risk of febrile neutropenia and other complications, I would consider every three-week nanoparticle paclitaxel because it would be more tolerable than docetaxel 100 mg/m² and I don’t feel dose-dense therapy is right for every patient.

We are considering an adjuvant trial comparing dose-dense AC followed by dose-dense paclitaxel versus dose-dense AC followed by dose-dense nanoparticle paclitaxel.

I generally use docetaxel 100 mg/m² in the adjuvant setting. If I had a patient who had been treated for early breast cancer and had a recurrence with a small, solitary lung metastasis that was biopsy-positive — a pseudoadjuvant setting — I would still use docetaxel.

I have two patients who have experienced long-term complete responses after six doses of docetaxel at 100 mg/m2; however, in patients who clearly have metastatic disease, I am moving toward using nanoparticle paclitaxel if it is available.

DR LOVE: How do you think nanoparticle paclitaxel compares with paclitaxel and docetaxel in terms of efficacy?

DR O’SHAUGHNESSY: I believe nanoparticle paclitaxel 260 mg/m² is superior to paclitaxel 175 mg/m² in terms of response rate and time to progression. The data in the pivotal trial of nanoparticle paclitaxel in anthracyclinepretreated patients basically shows it to be as efficacious as docetaxel in terms of response rates.

DR LOVE: When you use adjuvant docetaxel, what dose do you use and when do you use growth factors (Figure 26)?

DR O’SHAUGHNESSY: I generally use 100 mg/m² of docetaxel. If I start any lower, I begin with 85 mg/m², particularly in patients who are older, frail or have multiple comorbid conditions. I have started as low as 75 mg/m², although that is rare. I only use growth factors if a patient requires it during AC treatment and then I use pegfilgrastim while continuing the AC. I never use filgrastim.

DR LOVE: Can you comment on the recent data regarding the incidence of febrile neutropenia at 100 mg/m² dose?

DR O’SHAUGHNESSY: In a large trial of docetaxel 100 mg/m² with over 700 patients, the 19 percent rate of febrile neutropenia was reduced to one percent with pegfilgrastim. I believe it included patients with metastatic disease. We use AC followed by docetaxel in the adjuvant setting and less than 10 percent, maybe eight or nine percent, of our patients experience febrile neutropenia.

DR LOVE: Why is febrile neutropenia more common in metastatic disease?

DR O’SHAUGHNESSY: A number of possibilities exist. With liver metastases, even if the patient’s liver function studies are normal, we still wonder about their metabolism. Also, if the patient has received other chemotherapeutic agents, the integrity of the mucous membranes in the gut may not be 100 percent. Also, we know that as a patient’s performance status declines, drug metabolism and excretion may not be as robust and if a patient is symptomatic from their breast cancer, their nutritional status may not be optimal.

DR LOVE: If you could reduce the rate of febrile neutropenia from eight or nine percent to one percent in the adjuvant setting, why not use growth factors?

DR O’SHAUGHNESSY: Whenever we treat a patient, we want to have an evidence-based reason to do so and we have little data on these agents in the adjuvant setting.

We have a good deal of experience in using hematopoietic growth factors in metastatic patients, but, unfortunately, their life spans are too short to determine the long-term effects of these agents.

In the adjuvant setting, most of these patients will be cured, but they’re receiving alkylators, anthracyclines and many receive radiation therapy and the long-term risk of leukemia is unknown. Filgrastim was used in the NSABP studies B-22 and B-25 and elevated risks of leukemia occurred, but they were also using high doses of cyclophosphamide. In addition, it doesn’t make economic sense to use an agent in all patients that is not needed 90 percent of the time.

Select publications