Edited excerpt from the discussion:

DR SCHWARTZ: This patient is a 41-year-old attorney who, in October 2001, underwent a left-sided lumpectomy and sentinel lymph node dissection followed by a full axillary dissection for a one-centimeter, Grade III invasive ductal carcinoma. The tumor was ER/PR-positive and HER2-negative. The sentinel lymph node was positive by H&E, so this was a Stage II, T1b, N1 disease.

I saw this patient shortly before some of the adjuvant taxane data — TAC from BCIRG (Martin 2003) and dose-dense data with AC T (Citron 2003) — became available. I elected to treat her then with six cycles of adjuvant FEC 100 and postlumpectomy radiation therapy. She has been maintained thus far on tamoxifen for 36 months.

DR LOVE: What happened to her menstrual cycles?

DR SCHWARTZ: She initially did continue to menstruate, but it has now been approximately a year since her last period.

DR LOVE: Have you done blood work on her to assess her menopausal status?

DR SCHWARTZ: Yes, I have. I did check a recent serum estrogen level, which was 120 pg per milliliter. Her LH and FSH levels were consistent with a postmenopausal state, but as you know, some overlap occurs.

She fell into the range that could be postmenopausal, but she could also be in the luteal phase. So she was borderline by the parameters that I could check, except that she has not had any menses for nearly a year, and she also has had many of the complaints that young women on tamoxifen have. She is single, wants to get married and has gained a lot of weight on the tamoxifen. Hot flashes are also of concern for her.

DR LOVE: Bob, one of the most vexing situations we face right now in hormonal therapy is the patient who starts out premenopausal, receives chemotherapy and then at some point stops having menstrual periods. Can you talk about how you would have thought through this case?

DR CARLSON: The data today are quite convincing that the aromatase inhibitors should play a role as adjuvant hormonal therapy for postmenopausal women with ER-positive breast cancer. Precisely how to sequence or to incorporate those data into the premenopausal subset is much less clear. We do know that the aromatase inhibitors do not suppress circulating estrogen levels adequately in women with functioning ovaries, whether or not they have menstrual function.

Therefore, if you’re going to use an AI for a young woman, you have to be certain that she is postmenopausal, or I think she should be enrolled in one of the prospective trials evaluating the use of ovarian suppression and an AI in premenopausal women (3.1, 3.2).

We do know that a number of women stop having menstrual function or periods subsequent to cytotoxic chemotherapy, yet their ovaries continue to cycle.

A substantial proportion of women also stop having ovarian function with cytotoxic chemotherapy, at least over the short term, but on further follow-up, their ovarian function returns.

So even though this woman is not having menstrual periods and has an estradiol level that is technically postmenopausal, I would have some concerns that her menstrual function might resume. If she is crossed over to an aromatase inhibitor at this point in time, it would be important to serially monitor her estradiol levels to assure that she does, in fact, remain postmenopausal.

The current NCCN version of the breast cancer guideline was substantially modified last summer to include the option for premenopausal women at the time of diagnosis who become postmenopausal with cytotoxic chemotherapy or during initial tamoxifen therapy to cross over after the two- to three-year time point to an aromatase inhibitor (3.2).

This is to be done as long as the physician is confident that the woman has become postmenopausal and she is monitored during her aromatase inhibitor therapy to assure that she remains postmenopausal.

However, that recommendation is what we call the 2-B recommendation. This means that it is based upon low-level evidence, and substantial disagreement or differences in practice patterns existed among the panelists in terms of how they would handle that specific situation.

If you look closely at both the ITA and IES switching trials, neither of them allowed women who became postmenopausal while on tamoxifen to enroll (Boccardo 2005; Coombes 2004). Those trials were designed for women who were postmenopausal at the time of diagnosis.

DR LOVE: Can you talk a little bit more, Dr Schwartz, about your discussions with this woman? Was she aware of the issue of switching?

DR SCHWARTZ: She was aware of the aromatase inhibitors and that many people were switching over to them. She wasn’t pushing to do that. Her main issue was the side effects of the tamoxifen (3.3), more than the need to switch to an aromatase inhibitor. She knew that her risk of recurrence was not huge but was looking more for a better quality of life with a different medication.

DR LOVE: One of the things that is interesting about our Patterns of Care surveys is that sometimes we find large discrepancies between the practices of clinical investigators and community-based oncologists. One of the biggest gaps that we have seen deals with the issue of switching.

Based on our findings, over 95 percent of breast cancer research leaders generally switch postmenopausal patients on tamoxifen to an aromatase inhibitor at two to three years.

However, only 50 or 60 percent of physicians in community practice do the same. Of course, we use theoretical cases, but Bob, what is your approach for a patient who starts out postmenopausal?

DR CARLSON: For the woman who starts out postmenopausal, I find the decision fairly easy. I offer switching to almost all of those patients. I do that based on the philosophy we know from other studies, that women are willing to accept a lot of toxicity for a one- or two-percent improvement in their outcome. That is in the range of improvement in outcome that we see with switching to an aromatase inhibitor. So I offer switching to an AI to all of my patients who are postmenopausal at the time of diagnosis and have had two to three years of tamoxifen therapy.

In general, in terms of what to do with premenopausal women, I am continuing them on tamoxifen for five years and at that point am reassessing their ovarian function. If they are postmenopausal at that time, then I will switch them to letrozole, a la the MA17 trial (Goss 2003).

DR LOVE: If you are concerned about a patient potentially menstruating again, why not just put her on an LHRH agonist to be certain?

DR CARLSON: One could do that, and data from the metastatic setting show that the combination of an LHRH agonist and an aromatase inhibitor is an active regimen in premenopausal women with hormone receptor- positive metastatic breast cancer.

However, we simply lack the data in the adjuvant setting, and in the absence of convincing data, that is something we should do with great caution.

DR LOVE: Some of those data evaluating an LHRH agonist plus an aromatase inhibitor in the metastatic setting came from your work. Can you talk specifically about the study that you conducted?

DR CARLSON: A couple of Phase II trials have utilized an LHRH agonist to suppress ovarian function and then added an aromatase inhibitor for premenopausal women with hormone receptor-positive metastatic breast cancer. John Robertson has reported on this (Forward 2004), and we recently reported our Phase II trial at the San Antonio meeting last December (Carlson 2004; [3.4]). The regimen is well tolerated, as we would expect, and the rates of response are quite high. In both the UK trial and our trial, response rates are in the ballpark of about 80 percent, so it’s an active regimen. We don’t technically have a Phase III trial that compares them directly with other regimens; however, those Phase II rates of clinical benefit aren’t met by any other hormonal therapy we have available.

DR BUDD: Bob, for the patients whom you switch after two to three years, how long do you tell them you are going to give the AI?

DR CARLSON: That is a great question. One of the things we don’t know about the AIs is how long they should be administered. That is especially problematic for those women whom we are switching.

So far, the switching data are new enough that I have not had to deal with this issue. Hopefully, we’ll have some additional data about duration of therapy before we have to make that decision for a lot of patients. In this situation, for which we don’t have good data, my expectation is that I will continue them on the AI beyond the two- to three-year time point until we have some duration data from other trials.

However, I’m not sure that is the right thing to do. Based upon the trials we have, it would certainly be reasonable to stop the aromatase inhibitor on the five-year anniversary of hormonal therapy.

DR LOVE: Tom, my understanding is that the NSABP is considering a trial evaluating women who have been on hormone therapy for five years that has included an aromatase inhibitor. This might be a patient who switched at two to three years or started on the AI up front.

They are planning to randomly assign patients to five more years of an AI versus no more therapy. What are your thoughts about that kind of a strategy?

DR BUDD: This is an important question because we don’t know the optimal duration of these treatments. The optimal duration for tamoxifen has been pretty well worked out: Five years is better than two years and 10 years is not any better than five years. With the aromatase inhibitors, we don’t have that information.

Currently, patients who had been randomly assigned to the letrozole arm in the MA17 trial are being rerandomized. Half of these patients will continue on therapy for an additional five years, and the other half will not.

This may not answer the question of whether 15 years of total treatment is better than 10, but I think the issue is important enough that additional trials are needed.

Similarly, for tamoxifen, we can come up with biologic explanations as to why the treatment ceases to work after some period of time. At present, with aromatase inhibitors, we don’t have such a biologic explanation. I think the expectation is that continuing therapy would probably be helpful.

Select publications