Joyce O’Shaughnessy, MD
Co-Director, Breast Cancer Research, Baylor-Sammons Cancer Center US Oncology

Edited comments by Dr O’Shaughnessy

Adjuvant/neoadjuvant trials incorporating capecitabine/ docetaxel (XT)

The ongoing US Oncology adjuvant clinical trial is for patients with nodepositive or high-risk, node-negative breast cancer. Patients are randomized to AC x 4 cycles, followed by four cycles of docetaxel with or without capecitabine x 4 cycles. There are two other trials evaluating the XT combination: the proposed replacement study for NSABP-B-27 and the MD Anderson ongoing neoadjuvant trial in which patients are randomized to weekly paclitaxel for three months versus the XT combination with both arms followed by FEC.

In a nonprotocol setting, my standard adjuvant approach is AC followed by docetaxel. I believe duration is important, especially for patients at high risk — patients with large tumors or positive nodes, particularly those with macrometastases in the nodes. When treating women at very high risk, I stick with the data as much as possible and use the CALGB FAC regimen at 600/60/600 mg/m2 followed by docetaxel at 100 mg/m2.

The data from San Antonio comparing three different dose levels of docetaxel — 100 mg/m2 versus 75 mg/m2 versus 60 mg/m2 — resulted in response rates of 36 percent versus 23 percent versus 22 percent, respectively. That’s a 50 percent improvement in response rate, which may translate into improved disease-free and overall survival. If I had breast cancer and was treated with docetaxel, I would want 100 mg/m2.

Utilization of microarray profiles to predict pathologic response

In their neoadjuvant study, MD Anderson will utilize microarray technology to perform gene transcription profiling on fine needle aspirates. They will then correlate the differential gene expression with pathologic responses. Dr Pusztai has compared core biopsies to fine needle aspirations (FNAs) and found the microarrays perform equally well. If we can define a microarray signature that predicts a pathologic complete response — and apparently there’s been some recent success in doing that — that would be very exciting.

We will work very closely with Dr Pusztai to repeat that trial with a different set of chemotherapy drugs which we anticipate to be FEC followed by capecitabine/docetaxel as neoadjuvant therapy. We want to see if we can correlate the microarray profiles on the primary tumor — also collected through fine needle aspiration — to determine the sensitivity and specificity for a microarray pattern predicting whether a patient will have a pathologic complete response or not.

Evaluating molecular markers in clinical trials of capecitabine

There are ongoing clinical trials evaluating protein and mRNA of thymidine phosphorylase (TP), dihydropyrimidine dehydrogenase ([DPD], which is the key catabolic enzyme for 5-FU), and thymidylate synthase ([TS], which is the target enzyme for 5-FU).

Preclinical and xenograph data demonstrate that the likelihood of responding to capecitabine is related to the ratio of TP to DPD. In animal models, the higher the TP level and the lower the DPD level, the more likely there will be a response to capecitabine. I think capecitabine may be one of the first chemotherapy agents we’re able to select based on protein and mRNAs.

Impact of CALGB-9741 dose-dense data on clinical practice

CALGB-9741 was a well-conducted trial, and with the hazard ratio for recurrence of 0.5 at four years, I believe the data will hold up. Dose-dense scheduling resulted in an 82 percent, four-year, disease-free survival rate.

That is exactly the same as the three-year, disease-free survival rate achieved with TAC when compared to FAC. Dose density has become an option for patients and completing therapy more quickly may be beneficial. However, there are no data to prove that a dose-dense regimen is superior to TAC or to AC followed by docetaxel.

In the adjuvant setting, if you are going to administer AC followed by paclitaxel, you should probably use the dose-dense regimen. However, we need to know whether it’s true that by cycling agents every two weeks, we produce more cell kill, as the Norton-Simon hypothesis suggests.

Marjorie Green’s data from MD Anderson, comparing weekly paclitaxel preoperatively to every-three-week paclitaxel — both followed by FAC — showed a doubling of the pathologic CR rate with weekly paclitaxel. Many oncologists believe that part or perhaps all of the benefit seen in CALGB-9741 could be explained by administering paclitaxel alone every two weeks, so it’s not at all clear that AC should be given every two weeks.

In my practice, I have not yet incorporated dose-dense chemotherapy because I believe duration is important. I want to cure women, and I’m going to administer an agent for whatever duration I believe is necessary.

Impact of CALGB-9741 dose-dense data on clinical trial design

We would like to improve upon the CALGB-9741 data, and our clinical trial of AC followed by docetaxel versus capecitabine/docetaxel may result in the next big advance in adjuvant therapy. The US Oncology Breast Committee discussed giving docetaxel every two weeks, but there’s no real rationale for doing that. It has a long half-life and capecitabine is already dose-dense because it’s administered for two weeks followed by one week off. We didn’t feel the need to manipulate that aspect of the clinical trial.

We considered administering AC every two weeks, but we don’t have any data that it is more beneficial than an every-three-week schedule. There was a neoadjuvant trial presented in San Antonio by Euler and colleagues evaluating three cycles of epirubicin/cyclophosphamide 120 mg/m2 and 600 mg/m2, respectively, randomizing between administering the combination every three weeks versus every two weeks. The pathologic CR rate in the every-three-week group was 9.5 percent compared to 3.9 percent in the every-two-week group. With only 262 patients, it’s not a definitive study, but it certainly doesn’t suggest that giving EC every two weeks is better.

Our US Oncology Breast Committee concluded that we would watch accrual to our trial, and doctors will “vote with their feet.” Currently we’re accruing between 65 and 75 patients a month, and we haven’t seen a big drop-off since the San Antonio meeting.

Dose-dense scheduling of docetaxel

Approximately one-third of practitioners are using dose density in their practice. Interestingly, a question was posed at a meeting with community oncologists: “ Would you use dose-dense AC followed by a taxane every two weeks?” Onethird responded affirmatively. When asked: “What taxane would you use?” approximately 55 percent indicated docetaxel and 45 percent chose paclitaxel.

It is premature to use docetaxel every two weeks without more safety and efficacy data. You can administer docetaxel 100 mg/m2 by itself and it’s relatively safe. It’s really not feasible to administer docetaxel 100 mg/m2 every two weeks following AC because there’s too much skin toxicity. I don’t know what you would use for dose-dense therapy in the neoadjuvant or adjuvant setting because we just don’t have enough data.

Phase III trial of nanoparticle paclitaxel versus paclitaxel

I’m closely watching the randomized Phase III trial of paclitaxel 175 mg/m2 every three weeks versus ABI-007 260 mg/m2 every three weeks. ABI-007, or nanopaclitaxel, is an albumin-formulated paclitaxel. It’s cremophor-free, so you can raise the dose safely, and it’s very well-tolerated.

In our Phase II trial of weekly ABI-007 100 mg/m2, three weeks on, one week off, in taxane-refractory patients, we’re seeing some very impressive responses and extremely good tolerability. If that’s a positive trial, as it may well be, it will be a very important advance because you won’t need steroids. In my experience, the neurotoxicity with 100 mg/m2 is negligible.

CALGB trial: Adjuvant capecitabine versus CA or CMF in elderly women

We did a small, randomized Phase II trial comparing intravenous CMF and fulldose capecitabine as front-line therapy in elderly patients in the metastatic setting. The response rate with capecitabine was 30 percent compared to 16 percent with intravenous CMF.

In a randomized Phase II trial of patients pretreated with anthracycline, comparing paclitaxel 175 mg/m2 every three weeks to full-dose capecitabine, 1,250 mg/m2 BID, two weeks on, one week off, the response with the capecitabine was 36 percent compared to 26 percent with paclitaxel. The confidence intervals were widely overlapping, so we couldn’t conclude that capecitabine is superior, but what you can say from these two studies is that it’s certainly unlikely that capecitabine is worse than CMF or paclitaxel.

It’s interesting how quickly capecitabine has moved to trials in the adjuvant setting. In women over age 65, 75 percent have ER/PR-positive breast cancers. I think the role of chemotherapy in that group of patients is sufficiently unknown. For women over 70, in particular, there are so few patients in the overview analysis in that age group that I think it’s very reasonable to compare capecitabine to AC or CMF. I’d be a little less comfortable with it in a younger patient population, only because the overview has clearly shown that polychemotherapy is superior to monotherapy.

Fulvestrant in the treatment of postmenopausal ER/PR-positive metastatic disease

I’ve used a fair amount of fulvestrant, and it’s very well-tolerated. We’ve had some very nice responses to fulvestrant, including one of my patients who was on the original clinical trial of fulvestrant versus anastrozole. She was on fulvestrant for three and a half years and now she’s on anastrozole. The injections have not been an issue for patients, and most women are very grateful that the side-effect profile is close to nil. I think fulvestrant probably crosses the blood-brain barrier and patients do have hot flashes on it, but in general, they’re quite mild.

I am a little disquieted by the fact that it can take three to five months to reach a steady state with fulvestrant. A patient with rapidly progressing disease may not benefit from fulvestrant, but fortunately most women with hormoneresponsive breast cancer have relatively indolent disease. I’m very interested in the clinical trial in which they are loading fulvestrant 500 mg every two weeks for a couple of doses and then reducing it to 250 mg monthly. That makes sense to me, so I’ve been trying to load it a little by giving it every three weeks for several injections in an attempt to raise the levels more quickly.

Sequencing hormonal therapies in the front-line treatment of metastatic breast cancer

When I see a postmenopausal patient who has relapsed on adjuvant tamoxifen, I tend to use an aromatase inhibitor followed by fulvestrant when the disease progresses. In the frontline metastatic trials of aromatase inhibitors versus tamoxifen, data demonstrates that regardless of whether you administer an aromatase inhibitor after tamoxifen or tamoxifen after an aromatase inhibitor, there’s a 40 to 50 percent clinical benefit rate. In the second-line setting, if you administer fulvestrant after an aromatase inhibitor or an aromatase inhibitor after fulvestrant, there’s approximately a 33 percent clinical benefit rate. These are all small trials and most of them are not randomized, but they show that any of these regimens can be effective and there’s no mandatory sequence for these agents.

Sequential single-agent versus combination chemotherapy for metastatic disease

George Sledge’s Phase III trial of single-agent doxorubicin, paclitaxel versus the combination of doxorubicin/paclitaxel as front-line chemotherapy for metastatic breast cancer failed to show a survival benefit for the combination. It’s difficult to demonstrate a survival advantage in front-line metastatic disease because, according to the MD Anderson series, these patients live an average of four years. What you do early in their treatment may never be reflected in a survival advantage because they have many other opportunities for treatment down the line.

In chemotherapy-naïve patients with metastatic disease, I generally use capecitabine/docetaxel (XT). There’s no evidence that administering an anthracycline after a taxane harms the patient in any way. I eventually use an anthracycline; I just don’t feel compelled to use it up front. The decision to use single-agent taxane or single-agent capecitabine or the combination for frontline therapy depends on factors such as the patient’s presentation and the extent of her disease.

As we begin later-line therapy, when patients become more symptomatic and heavily tumor-burdened, and their life expectancy is shortening, a very reasonable argument can be made for better palliation and maybe even better survival with a well-tolerated combination regimen.

Combination chemotherapy for late-line therapy in patients with hormone refractory, HER2-negative disease

For late-line therapy in patients with hormone refractory, HER2-negative disease, I prefer a well-tolerated combination. I love the combination of capecitabine and vinorelbine. It’s the nonalopecia approach to excellent palliative care, and I’ve used it often.

Combination therapy with chemotherapy and antiVEGFs

In the trial comparing capecitabine with or without bevacizumab in patients with metastatic disease, the response rate was significantly improved with the combination, but the primary endpoint did not improve. It almost seemed like the VEGF antibody, bevacizumab, had an initial antiangiogenic effect that increased the response rates, which were not durable.

This suggests that perhaps, in late-line cancers, there are other proangiogenic substances that may take over. It also suggests the late-line patient population may not be the ideal place to study angiogenesis inhibitors.

The ongoing randomized study of paclitaxel with or without bevacizumab is an earlier-line trial in metastatic disease. Studies like this need to be done because we can’t assume the response will be the same in early- versus late-stage disease, and we may eventually have to take it into the micrometastatic setting.

Dr O’Shaughnessy’s viewpoint on the management of the primary lesion

I probably would not do anything further in terms of local therapy to the breast at that point. I have a patient exactly like this in my practice. She presented with bone and lung metastases but had a locally advanced breast cancer and a mass in her other breast. When we reduced the locally advanced tumor to a manageable size, she had a salvage mastectomy because we were concerned it might cause a local control problem.

The contralateral breast still has a small nodule, which we are following closely; we will do a salvage mastectomy if it starts growing. While I think surgeons would be inclined to excise the lesion in this case, medical oncologists view it as just another indicator lesion.

Dr O’Shaughnessy’s viewpoint

This patient could potentially be treated with a single cytotoxic agent, and I believe the best choice would be either paclitaxel, docetaxel or capecitabine. However, she’s very symptomatic and better response rates are achieved with combination chemotherapy, so I would treat her with capecitabine/docetaxel. If the combination is well-tolerated, I continue administering it. If the combination becomes too toxic, I discontinue docetaxel and utilize capecitabine monotherapy.

The problem with combination therapy is you don’t know whether the patient responded to the combination or one of the single agents. I’ve had excellent responses with this regimen, and the responses have been very durable on the continued capecitabine monotherapy. If the patient then progress on singleagent capecitabine, I usually stop the capecitabine and go to another therapy — either single-agent vinorelbine or gemcitabine/carboplatin late-line.

Combination chemotherapy has a much higher response rate than either single agent alone, but if the patient was relatively asymptomatic and had a small volume of disease, you don’t really gain anything by using the combination regimen. In these patients, giving single agents sequentially is perfectly reasonable.

Select publications

Home · Search