Edited excerpts from the discussion:

DR GOLDBERG: This 70-year-old woman presented with Stage II breast cancer. She had an ER/PR-positive, HER2-negative, 2.5- centimeter adenocarcinoma with two positive lymph nodes. She received adjuvant CAF chemotherapy and then went on to tamoxifen for five years. After completing tamoxifen, she was put on anastrozole. She’s been on anastrozole for about a year and she’s doing well.

DR LOVE: Gersh, can you discuss how you approach these patients who have completed five years of tamoxifen in terms of deciding whether to use an AI and which one?

DR LOCKER: If their tumors are node-positive, they should go on an AI, unless they have severe osteoporosis. I use letrozole because that was used in the original study, and I’m trying to be evidence based. I suspect the other AIs would be just as good. Data from Europe support anastrozole after five years of tamoxifen (Jakesz 2005a).

In a woman with node-negative disease, if there were any negative prognostic features, such as a large tumor, then I would switch to letrozole after tamoxifen. I do not routinely switch breast cancer patients with T1-B, ER-positive, node-negative tumors. In that group of women, tamoxifen is as much a chemopreventative as it is an adjuvant therapy, and I’m not sure whether the potential side effects, in terms of bone or the need for bisphosphonates, merit it. On the other hand, if the tumor is a T1-C/N0, I probably would start them on letrozole at five years.

DR LOVE: How would you manage a patient who had five or 10 positive nodes initially and completed adjuvant tamoxifen more than a year or two ago?

DR LOCKER: I have “sinned” and put patients with 20 positive nodes on letrozole five years after completing tamoxifen. I know I will burn in hell for this, but they had 20 nodes and I don’t care. For someone who’s node-negative, I won’t do it that late.

DR LOVE: Incidentally, would you start a patient with HER2-positive disease and 20 positive nodes on adjuvant trastuzumab five years later?

DR LOCKER: No, I would not. One of the things about HER2-positive disease is that the side effects are bad and can be bad early. A few patients will probably have a recurrence late, but I think the greatest concern is with those who have a recurrence earlier.

DR LOVE: Would you treat that same patient with adjuvant trastuzumab two years after diagnosis?

DR LOCKER: Probably.

DR LOVE: Aman, how do you think exemestane, letrozole and anastrozole compare with regard to serious toxicities?

DR BUZDAR: That’s an important issue we need to discuss with patients. Even though these studies did not compare one aromatase inhibitor head-on with another, the ATAC study, which has the longest follow-up, has not shown any increase in cardiovascular events with anastrozole, and substantial reduction was apparent in cerebrovascular events (Howell 2005), whereas with the letrozole in BIG-1-98, an increased risk of cerebrovascular accidents was apparent and also an increased risk of fatal myocardial infarcts (Thürlimann 2005b). These numbers are small but of some concern. The same thing was seen with the myocardial events in the exemestane study (Coombes 2004). I am concerned, but I discuss the data with the patient and use the aromatase inhibitor for the setting in which the most data are available.

We now have some of the data from the Austrian study, in which anastrozole was used after five years of tamoxifen (Jakesz 2005a). We see a similar proportional reduction in the risk of recurrence, which gives us one more piece of information that suggests we may be able to use anastrozole in this setting.

DR LOVE: At this time, what do you think is the most defensible aromatase inhibitor to use initially and after two to three years, Gersh?

DR LOCKER: Up front, it’s clearly anastrozole. Until we have five or six years of data in the BIG I-98 study, anastrozole is the drug that is most reasonable up front (5.1). After five years, letrozole should be used. I have no doubt that anastrozole will probably be just as good, but I act on the data that we have.

In terms of the switching studies, the IES data certainly support exemestane, while the German-Austrian and the Italian data support anastrozole (Coombes 2004; Jakesz 2005b; Boccardo 2005; [5.2]). I think either one is acceptable. I would probably use anastrozole, only because I have more experience with it in the adjuvant setting.

When the BIG I-98 trial reports on the two switching arms — and I’m praying they are not underpowered — then we will have an answer as to whether letrozole should be used after two years of tamoxifen. However, for now I go with the most mature studies that address the specific situations.

DR LOVE: Aman, do you agree with Gershon as to which AI should be used initially and when switching after two to three years?

DR BUZDAR: My feeling is that if you look at the efficacy, either up front or after two to three years, or even after five years, all these aromatase inhibitors show similar efficacy. Differences in safety exist, but you have to keep in mind that you are looking across the trials, not at a head-on comparison. One study is comparing anastrozole with exemestane, and that will provide head-on safety data.

However, I totally agree with Gershon that we have the most mature safety data for anastrozole, which has the longest followup, and the efficacy data of BIG-1-98 is almost a mirror image at the two-year follow-up. That study, at least, confirms the ATAC data, in which at two and a half years we saw a similar proportional reduction in risk of recurrence.

DR DESAI: When you do switch patients after two or three years of tamoxifen, patients ask whether a total of five years of endocrine therapy is enough. They want to continue therapy. What do we do about those patients?

DR LOCKER: The switching trials are problematic, because the implication is that if you switch, you stop therapy at five years. Does it make sense that we give five years of tamoxifen followed by five years of letrozole, but when we switch, we give only five years of therapy — two years of tamoxifen followed by three years of anastrozole or exemestane?

I don’t know what to do with those patients. I use the “20-node rule.” If they had a lot of positive nodes and I’m really worried, then they’re going to receive hormones until they die. If they are node-negative, I think you should just go by what the studies say.

DR BUZDAR: I totally disagree with Gershon on this point. If you look at the hazard ratios for patients who have 20 versus zero positive nodes, the risk of recurrence in the first two to five years becomes very high as the number of nodes increases. As time passes, the risk of recurrence for the two groups becomes very close. The patient who has survived five years disease free has a risk close to that of the patient who had maybe one or no positive nodes (Saphner 1996).

DR LOCKER: What about looking at receptor status within that subset? I know it’s a subset of a subset analysis, but the women who had 20 positive nodes and are receptor-positive are not the same as the women who had 20 positive nodes and are receptor-negative.

DR BUZDAR: The recurrence risk is regardless of the receptor status. Patients are at increased risk in the first several years, and after that, while the risk is there, the differences between a high number of positive nodes and negative nodes or a low number of positive nodes tend to become blurred.

DR LOCKER: Well, having said that, remember that for a woman with node-positive disease, from year five to year 10 after tamoxifen, the absolute yearly risk for recurrence is two percent per year.

DR LOVE: Do you agree with that, Aman?

DR BUZDAR: It is true that there is a risk and it is higher than the normal patient population, but my point is that when comparing the risk between 20 versus two positive nodes, the differences become very close.

DR DRAGON: My understanding in looking at the trial of letrozole after tamoxifen is that 20 percent of the recurrences occur after five years. How does that impact decision- making for node-negative patients, where you’re talking about a relatively small number of women who recur, and then reducing that by 40 percent? We’re talking about a one or two percent benefit for five years of letrozole, which is, by my calculation, about $14,000 per patient. So you’re treating 100 women at $14,000 per patient to reduce, perhaps to eliminate, one or two recurrences over the next five years. Is that a rational thought process?

DR LOVE: It is not uncommon to see adjuvant chemotherapy used under similar circumstances.

DR LOCKER: Other issues arise with the patients with node-negative disease in the MA17 trial. For example, the odd survival data, showing a statistically significant survival advantage in patients with nodepositive disease, but no survival advantage in women with node-negative disease. In fact, it’s trending the wrong way. I think Dr Dragon’s point is very well taken. In patients with node-negative disease, you have to be selective as to whom you treat with letrozole after tamoxifen.

DR SMITH: There seems to be a bit of a schism between those who would start patients on anastrozole or letrozole initially versus those who would give tamoxifen for two or three years and then switch the patient to an aromatase inhibitor.

DR BUZDAR: You bring up a very important point. At MD Anderson, we do not use tamoxifen up front on any patient who is postmenopausal. Right now, in 2005, I do not think we can say that any subset of postmenopausal patients with ER-positive disease should start on an anti-estrogen and switch to an aromatase inhibitor after two or five years. I think the data we have are very convincing that starting with an aromatase inhibitor is better. It reduces the risk of recurrence and the overall safety profile of the therapy is better. I think the question is whether a subset exists in which it is better to start on tamoxifen and then switch later. This is a research question and the studies are ongoing.

DR LOVE: Gersh, when deciding on an adjuvant therapy, how much of an issue is the risk of endometrial cancer and thrombosis?

DR LOCKER: I remember sitting with Aman when they presented the ATAC hysterectomy data, which was a curveball. It showed that seven percent of women on tamoxifen underwent a hysterectomy during five years of treatment, compared to a couple percent, at the most, for the patients taking anastrozole. I think, in terms of the switching versus starting, that is a number you cannot escape. One out of every 100 women who receives tamoxifen for two or two and a half years and then switches to anastrozole will recur and presumably die, who would not have if they were on anastrozole initially. What do you say to that one patient? I don’t think anybody can predict who that one woman is, such that you can start her on anastrozole and start everybody else on tamoxifen.

DR LOVE: Aman, do you think we will see a survival advantage for aromatase inhibitors over tamoxifen?

DR BUZDAR: I think that you will see a survival advantage once these studies mature. I think if you did a meta-analysis of the data from the current studies, you would see a significant survival advantage. I just want to remind some of you that initially when tamoxifen was being evaluated, only one or two trials showed a survival advantage.

It was the first Oxford meta-analysis that showed a dramatic reduction in the risk of death because you need a lot of events. In these postmenopausal patients with ER-positive disease, it takes much longer for the events that contribute to the survival advantage to develop. Major competing causes of death play a role and breast cancer becomes a secondary cause of death, so you need a large number of patients and longer follow-up.

DR STEINECKER: Should we be a little more attuned to following lipid profiles in patients on aromatase inhibitors?

DR BUZDAR: It would be reasonable — at least with letrozole and exemestane, for which we have data indicating that those agents change the lipid profile adversely in a sizable number of patients — to have baseline information and to re-evaluate the lipids several months down the line. If they are being affected adversely, then maybe an appropriate intervention should be made before any major event related to that effect develops.

I would like to come back to this issue of cost. Patients don’t come to us because they want us to manage their financial affairs. They are coming to receive the best treatment for their cancer. We need to tell them what we think is the best and most effective treatment and then if the cost is an issue, we need to help them. Some of the patients who bring up the issue of cost have a bag full of other, alternative “health medications” for which they may be paying twice as much per month as the difference in costs between these two drugs.

Select publications