Robert W Carlson, MD
Professor of Medicine, Division of Oncology and Stanford Medical Informatics, Stanford University Medical Center

Edited comments by Dr Carlson

Case discussion: 65-year-old woman with recurrent ER-positive, HER2-negative breast cancer

History

This patient was initially diagnosed with breast cancer about a decade ago. She had ER-positive, HER2-negative, node-positive disease for which she received adjuvant CMF chemotherapy followed by tamoxifen.

Several years following the completion of tamoxifen, she experienced a biopsydocumented pulmonary recurrence. She had mediastinal lymphadenopathy with some tracheal compression evidenced by both chest X-ray and CT scan. She was active, although somewhat limited by dyspnea and cough, especially upon exertion. Her Karnofsky performance status was 70 to 80 percent. The symptoms had been evolving over several months.

Since the recurrence was quite symptomatic, she was treated initially with chemotherapy. We administered a three-hour paclitaxel infusion every three weeks, and she had a good response.

Follow-up

She had symptomatic improvement quite rapidly in just four to six weeks. Ultimately, we were able to document the improvement radiographically, and we continued the taxane for about six months. She had achieved a radiographic complete response and was experiencing fatigue, alopecia and some neurotoxicity.

At that time her disease was not progressing. We discontinued the taxane and started her on anastrozole, which was maintained for a couple of years, and she continued to do well. Then she redeveloped a pulmonary recurrence that was detected radiographically. Her disease was persistently mediastinal, and she was having minimal coughing at that time.

We elected to continue hormonal therapy and switched her to fulvestrant in a single 5-mL injection. Her cough has improved, and serial chest X-rays and CT scans have documented continued improvement in her mediastinal disease. She has been on fulvestrant for almost one year.

Discussion

If her disease progresses on fulvestrant and she doesn’t have substantial endorgan dysfunction, I would consider another endocrine maneuver; megestrol acetate or ethinyl estradiol would be possibilities. Exemestane would also be a reasonable choice. Exemestane’s toxicity profile is probably somewhat more acceptable than either megestrol acetate or ethinyl estradiol. The crossover rates of response from the nonsteroidal aromatase inhibitors to exemestane are quite modest.

When we have exhausted all hormonal therapy options, sequential single-agent vinorelbine or capecitabine would probably be my next choice for her. I would likely choose capecitabine over vinorelbine, but I think that either would be reasonable. Ultimately, she is almost certainly going to be treated with both agents.

Patients prefer the convenience of oral therapy, and the response rates for capecitabine are arguably equivalent to vinorelbine. Most patients tolerate vinorelbine, but it requires weekly administration, and a small number of women experience profound asthenia.

We do not start capecitabine at the FDA-approved dose. We typically use capecitabine at 2,000 mg/m2 per day (total daily dose) divided in two doses for two weeks on and one week off. Most women tolerate that dose well for several cycles. The development of the hand-foot syndrome is a problem that ultimately may require either a dose reduction or prolongation of the one-week interval off therapy to two or sometimes even three weeks.

Sequential single-agent versus combination chemotherapy

The available data support the use of either combination chemotherapy or sequential single agents in patients with metastatic disease. Typically, my practice is to use sequential single agents because there is generally less toxicity and, arguably, equivalent efficacy.

This is one situation in which the NCCN guidelines have changed dramatically in the last year. Previously, the guidelines called for combination chemotherapy in women with first relapse. They recommended the use of either CMF or CAF or a single-agent taxane and then, following failure, crossover to whatever had not been given.

The current NCCN guidelines acknowledge the uncertainty that combination chemotherapy is really superior. Now, we have a list of preferred single agents. We have a list of preferred combinations, and an acknowledgement that there is inadequate data to support a dogmatic statement about whether sequential single agents or combination chemotherapy should be used. We also list a series of other active agents that could also be considered in the sequence.

It’s generally accepted that response rates are higher with combination chemotherapy. The duration of ultimate disease control, however, is not clearly superior for combination chemotherapy compared to sequential single agents.

The patient who is more ill when beginning therapy may be less able to tolerate aggressive combination chemotherapy; however, that’s precisely the patient who needs aggressive combination chemotherapy. I think it’s a situation in which clinical judgment is important, and it’s crucial to involve the patient in the decision-making process.

Chemotherapy followed by hormonal therapy

The practice of initially treating patients who have ER-positive breast cancer with chemotherapy and then switching to hormonal therapy is not addressed in the NCCN guidelines; however, it’s a common strategy that makes sense.

In general, the NCCN guidelines classify women into two groups: those who should be given endocrine therapy until they sequence through all of them or develop organ impairment and those who should be given chemotherapy until they have exhausted all of the reasonable chemotherapy options.

The guidelines do, however, recommend that women who have substantial organ dysfunction — even those with hormone receptor-positive disease — be treated initially with cytotoxic chemotherapy.

Fulvestrant in ER/PR-positive, postmenopausal patients with metastatic disease

Fulvestrant binds with the estrogen-receptor monomer in the cytoplasm and prevents the dimerization of the estrogen receptor, which is required for exertion of its maximal activity. Lack of estrogen-receptor dimerization results in accelerated degradation of the ER-fulvestrant complex. Ultimately, there is a loss of estrogen receptors within the cells.

The estrogen receptor is continually regenerated, so continued exposure to fulvestrant is required. After fulvestrant is discontinued, the estrogen receptor will, with time, reappear in cells. The fact that we see subsequent hormonal responses is convincing biological or clinical evidence that the estrogen receptors do reappear.

Injection site reactions and hot flashes are the only side effects that I’ve observed in patients receiving fulvestrant. There may be something about the administration technique for fulvestrant that can affect the pain that is infrequently experienced. If the injection is inadvertently given subcutaneously into fat, it’s more painful than if it’s given intramuscularly. It may be that many of the women who have pain with the injection are not actually receiving true intramuscular injections; this is more likely to occur in women who are obese.

Sequencing of hormonal therapies

Women with breast cancer who fail on tamoxifen can clearly respond to fulvestrant, and the rate of response is equivalent to that seen with anastrozole. Also, in women with disease that has failed anastrozole who are then crossed over to fulvestrant, the rate of clinical benefit is substantial and in the range of about 40 percent. Patients who are crossed over from fulvestrant to aromatase inhibitors also show response rates around 40 percent.

Surprisingly, the magnitude of benefit from fulvestrant does not predict whether the cancer will respond to a subsequent hormonal maneuver. One rule of thumb in the past has been that the magnitude and duration of response to the most recent hormonal therapy predicts for the likelihood of response for subsequent hormonal therapies. A small retrospective study suggests that may not be the case with fulvestrant.

Novel hormonal therapy combinations

There is an increasing body of preclinical evidence suggesting that breast cancers that become resistant to tamoxifen or fulvestrant have upregulation of epidermal growth factor receptor (EGFR) and HER2 expression. As those endocrine-sensitive cells become endocrine-resistant and the EGFR and HER2 upregulate, some of the sensitivity to the endocrine agents may return if those cells are exposed to EGFR inhibitors.

There are series of trials being conducted to evaluate the role of fulvestrant or other hormonal agents in combination with gefitinib. ECOG is initiating a Phase II randomized trial comparing fulvestrant/gefitinib to anastrozole/gefitinib.

Trastuzumab in combination with fulvestrant would be a very interesting study. There is evidence that HER2-overexpressing tumors are relatively more hormoneresistant. There’s a lot of cross talk between those pathways, and that study would be similar to the studies looking at gefitinib plus fulvestrant. We can extend that thinking even further and look at the utilization of hormonal therapies in combination with an EGFR inhibitor and a HER2 inhibitor.

The combination of an aromatase inhibitor and fulvestrant is of some interest, but the difficulty with such a study is that fulvestrant eliminates the estrogen receptor. Theoretically, if the estrogen receptor is eliminated, then the cells shouldn’t care how much estrogen is present.

ATAC trial results

Many of us were surprised that anastrozole alone was superior to tamoxifen in the ATAC trial. The difference in contralateral breast cancers was remarkable, with a 75 percent risk reduction for anastrozole compared to what we would expect with placebo.

I’m not totally surprised that the differences were seen so soon; presumably most of the breast cancers prevented from being diagnosed in those women were pre-existing breast cancers. Relatively few women in the ATAC trial actually received cytotoxic chemotherapy, so the contralateral breast cancers should have already been present.

The ATAC trial is an important trial biologically, demonstrating that the aromatase inhibitors likely have a very important role in early breast cancer. The differences seen between tamoxifen and anastrozole, especially if they’re maintained with further follow-up, are substantial and clinically significant.

Implications of the ATAC trial on clinical practice

As a result of the ATAC trial, my practice pattern changed overnight. I am not treating all of my patients with anastrozole, but I am certainly discussing the results of the ATAC trial and the pros and cons for tamoxifen and anastrozole. I’m using shared decision-making with patients to determine which of the agents they prefer. The recent update at the San Antonio meeting in December 2002 confirmed that practice.

Generally, I recommend anastrozole; however, there are other factors to consider that would sway me one way or another. Obviously, in women with an absolute or relative contraindication to tamoxifen, it’s a very easy decision. Conversely, there are patients who may have relative contraindications to anastrozole.

The major relative contraindication is severe osteoporosis. The bone mineral density loss associated with the aromatase inhibitors is a concern. Presumably, we can blunt that effect using bisphosphonates, so it is unlikely to be a major problem.

The patient’s nodal status does not make a great deal of difference to me in terms of hormonal therapy recommendations. I look at the patient’s HER2 status, and it does shade my thinking a bit. There is some data, although somewhat contradictory, that HER2-overexpressing tumors may be relatively resistant to tamoxifen.

Likewise, there is data suggesting that both letrozole and anastrozole maintain antitumor activity in HER2-overexpressing tumors. I think it would be reasonable to consider anastrozole, in preference to tamoxifen, for patients with tumors that have an IHC score of 2+ or 3+.

NCCN Practice Guidelines: Adjuvant hormonal therapy

In response to the initial presentation of the ATAC results, the NCCN guidelines were modified. Tamoxifen was maintained as the recommended adjuvant therapy in the text of the guidelines. There is, however, a footnote to the guidelines stating that anastrozole should be considered as an alternative to tamoxifen. The guidelines recommend a discussion between the physician and the patient regarding tamoxifen and anastrozole as adjuvant therapy.

The guidelines state that anastrozole may, in fact, be superior to tamoxifen, but we need to recognize there is short follow-up with adjuvant anastrozole relative to very long follow-up with tamoxifen. Because of that, it’s difficult to be dogmatic.

To some extent, it depends on the woman — is she someone who is an early adaptor of a new therapy, or is she someone who is more conservative in terms of adopting new technology or new therapies?

ASCO Technology Assessment regarding adjuvant aromatase inhibitors

The ASCO Technology Assessment is a superb document, but it needs to be viewed for exactly what it is. A technology assessment looks at a given therapy, attempts to decide whether that therapy has utility in a given clinical situation and determines what the preponderance of data is within that clinical situation. The ASCO Technology Assessment, in both the first and second versions, states that tamoxifen remains the standard adjuvant therapy to which other therapies should be compared.

Interestingly, several members of the ASCO Technology Panel also sit on the NCCN Practice Guidelines Panel. When the NCCN Practice Guidelines Panel looked at this issue, there was no major dissension in considering anastrozole as an option. The difference between groups occurred because of the different processes.

The ASCO Technology Assessment is strictly evidence-based, and it cannot go beyond the evidence. So, there are no extrapolations beyond five years of anastrozole or the 47 months of follow-up.

In the NCCN Practice Guidelines process, we use a methodology called evidence-based consensus. We establish recommendations based on evidence, but we are also able to use expert consensus in situations where the evidence is lacking. Obviously, 10-year data with adjuvant anastrozole are lacking, but we can come up with expectations about what might happen and make recommendations that extrapolate into the unknown.

The NCCN Practice Guidelines are patient-focused, and they look at the various therapies that are available from a patient’s perspective. The NCCN Guidelines Panel believes that women should consider the use of anastrozole, although we don’t say it should necessarily be used in preference to tamoxifen.

Nonprotocol use of adjuvant letrozole or exemestane

I’m not using either letrozole or exemestane in the adjuvant, nonprotocol setting, primarily because all the data we have is with anastrozole. I await, with a great deal of interest, the results of the many ongoing adjuvant trials evaluating letrozole and exemestane. Until we have that data, I think it’s premature to consider they are equal or even superior to anastrozole in the adjuvant setting.

CALGB-49907: Adjuvant chemotherapy trial in elderly women

I anticipate we will participate in Hyman Muss’ trial comparing capecitabine to CA or CMF in the adjuvant setting. It’s an excellent study that will compare a well-tolerated single agent to two generally well-tolerated combinations.

The difficulty with that study is two-fold. First, the magnitude of benefit from cytotoxic chemotherapy in older women is not clear, and if there are benefits, they are likely to be small. Second, it’s going to be very difficult to detect differences between the combination arm and the single-agent arm if, in fact, they exist, because they are likely to be so small.

There is likely to be a significant difference in quality of life, side effects and tolerability for the three regimens. Which of those regimens — AC or CMF or single-agent capecitabine — will ultimately be better tolerated is relatively unpredictable. The hand-foot syndrome associated with capecitabine is a problem for many women, and whether they will find it more acceptable than the nausea, vomiting, alopecia, etc., associated with CMF and AC, we’ll have to wait and see.

All women are concerned about alopecia as a side effect, but many women would find alopecia an acceptable side effect for the benefits associated with cytotoxic chemotherapy. However, I’m equally confident that no woman looks forward to experiencing alopecia. We underestimate how much of an issue alopecia is for women; it is very commonly the most feared side effect.

Women who have nausea and vomiting realize it’s something they can experience in the privacy of their home. The same applies to myelosuppression; women may have problems with fever and neutropenia, but when they walk out of their home nobody realizes they have myelosuppression. Women who have alopecia feel very uncomfortable when they walk outside their home. Many are quite uncomfortable with their wigs and don’t feel they look good. Alopecia is the toxicity women complain to me about the most.

Oral therapy is always advantageous for the compliant patient. One of the difficulties is assuring that the patients are compliant with their therapy. The oral regimens are especially beneficial in situations where accessibility to a medical oncologist is limited — in rural and underserved communities.

Adjuvant dose-dense chemotherapy

My expectation is that we will see dose-dense chemotherapy added to the NCCN Practice Guidelines. The real question is: How will it be weighted within the guidelines? If one looks at the duration of follow-up and magnitude of risk reduction with dose-dense chemotherapy compared to what is achieved with anastrozole, they are almost superimposable.

I have started using dose-dense therapy in selected patients. As we translate the results of dose-dense therapy into practice, one needs to be careful about who is selected for such therapies. Initially, I tend to use it more in the younger, highrisk patient. As I become more familiar with it, my indications will expand. We also need be cautious not to jump to the conclusion that dose-dense therapy will be superior when long-term data are available. I’m hopeful that it will be, I expect it to be, but we’re going to have to continually look at that data.

Select publications

Home · Search