Breast Cancer: Extended Adjuvant Endocrine Therapy

Transcript:Adam M. Brufsky, MD, PhD: What do we do about extended adjuvant therapy now? We have, on the one hand, all this data with tamoxifen and a meta-analysis that shows benefit, and now we have a bunch of trials that just came out. We had MA.17R—that one of us will talk about at some point—and now we have randomized data from extended adjuvant therapy, 5 versus 10 years, that have come out at the San Antonio Breast Cancer Symposium. Do you want to talk about B-42 to start us off, Kim?

Kimberly L. Blackwell, MD: Basically, across all of the studies, I think we should focus on the 5 versus 10 studies first. B-42 is a study where everyone had to have at least 2 years of an aromatase inhibitor (AI), if I remember correctly, and they were randomized—assuming that they were free of disease at 5 years—to either continuing the aromatase inhibitor or placebo. And what we really saw was a small but meaningful benefit. I think that the P-value may or may not have reached statistical significance. You probably know the numbers better than I do, but at least what we saw was a tiny benefit for the addition of an adjuvant aromatase inhibitor. And the benefit, in my mind, is mostly in prevention. I’ll say this theme quite a bit, which is that we’re creating a whole population of breast cancer survivors who are at risk of a secondary breast cancer. And if they have residual breast tissue, I think in all of the studies that we’re going to talk about today, one of the major drivers for that very small difference in breast cancer-free survival is really the prevention of a secondary breast cancer. So, at least in my mind, for the 5- versus 10-year—as you said, meta-analysis—whether you’re using tamoxifen or an aromatase inhibitor, the extended use needs to be discussed with patients at 5 years.

Adam M. Brufsky, MD, PhD: Right. Aditya, someone walks into your door with bilateral mastectomy—ER-positive breast cancer 5 years ago—and she says, “Oh, I have such terrible hot flashes, I’m so miserable. Dr Bardia, do I need really to continue?” This is postmenopausal, not premenopausal. She drove all the way up from Raleigh to Massachusetts General Hospital to talk to you. So, what do you do? What do you tell her?

Aditya Bardia, MD, MPH: I think it’s interesting. If we think about the MA.17R and NSABP B-42, these are trials where essentially you can use the trial to justify either decision. If you are a believer in extended adjuvant therapy AI, you can look at MA.17R and NSABP B-42 and say, “Oh, there was a significant difference.” It was statistically significant in MA.17R and borderline significant in B-42, but you can say there was a 3% difference, which would be clinically meaningful for my patient, so we should do it. But, on the other hand, you can look at the same data and make a case that we should not be doing it because it was not statistically significant, that 3% is not that important. Here, we are subjecting patients to a standard AI therapy for 5 years, concerned about hot flashes, as you were saying, or osteopenia. I think it’s a situation where it’s a bit of “art of medicine,” depending on the particular risk that the patient has, the benefit that you would see the adjuvant AI would add, and the potential risk in terms of osteopenia; the personal interest, how well they tolerated 5 years of AI. Then, all of these factors need to be taken into account to make that decision. So, it’s a long answer to your question because it’s pretty complicated.

Adam M. Brufsky, MD, PhD: So, I’m just curious what you guys say. Mark, are you guys worried that there wasn’t statistical significance in B-42? Does it bother you?

Mark E. Robson, MD: No, not particularly because I think worshipping at the altar of the 0.05 sometimes doesn’t make sense. The question is, do you really think that this was a spurious result, particularly given that you got other data from other places suggesting that this works? I’m not so worried about the level of significance. I am more worried about what Aditya says, which is that the benefits are tiny, just like we were saying with chemotherapy in the discordant groups, that we’re looking at the genomic assays in. And aromatase inhibitors are not much fun for a lot of people. We haven’t even talked about the premenopausal women who want to get pregnant or something like that. And so, you have to have a default setting when you’re talking to the patient. You probably can’t just throw it in front of them and say, “You decide.” You have to have some version of what you think should be done. And, given the data, I think the default setting is probably for extended therapy. But you also have to acknowledge that the benefit is small, and that if somebody decides they want to walk away from that, they’re not necessarily foregoing a huge benefit.

Adam M. Brufsky, MD, PhD: You guys agree, Kim?

Kimberly L. Blackwell, MD: Yes. We just finished talking about genomic assays. I think this is where, from a biology standpoint, I’m a little confused as to what to do. If I sent an Oncotype DX test 5 years ago and it’s low risk, the other interpretation of a low recurrence score is that’s a very endocrine, hormonally driven cancer. And so, those low risks, we don’t really have that integrated into B-42 or MA.17. We have to be a little careful about, “Oh, it was low, therefore you won’t benefit,” when, in fact, those might be the very women we see every day. I teach my fellows why we see people at 6 years. That’s when the breast cancer comes back because we’ve stopped the endocrine therapy at 5 years. Honestly, I just saw 2 patients in clinic yesterday—tamoxifen for 5 years and 2 years later, lung, sternal, and bone metastases. I know that’s a bias, but that’s what we see. We don’t see the women that go on and live forever off endocrine therapy. We see the women who go off of it, it comes back, and they’re exquisitely sensitive. So, I guess my 2 take-home points about extended adjuvant endocrine therapy are that I personally separate from hot flashes, bone loss, and all of that. At least we didn’t see a survival detriment. I just want to be clear. I didn’t see a signal. It was not powered to show. But that’s one thing. And then, the second thing is I would be a little cautious about using the 21-gene predictor to predict who might benefit from extended adjuvant therapy or not. Because, for me, the low score means you have an exquisitely hormone-sensitive breast cancer, and that’s where the prediction of the 21-gene for extended adjuvant falls short.

Adam M. Brufsky, MD, PhD: It’s actually approved by Medicare, the Breast Cancer Index (BCI), which is that H/I ratio. You guys use that? Anybody use that? Because Dennis Sgroi, who’s at Massachusetts General Hospital, helped develop it.

Aditya Bardia, MD, MPH: So, we’ve started using it because we’re faced with the decision—as you were saying, you have a patient who comes in clinic, has completed 5 years of an AI therapy—should we consider 10 years of AI versus not? And if the patient is having some side effect, we just have to be careful about making the recommendation of 5 additional years. In that situation, the BCI can be helpful because at least we have some data that can support the decision of using extended adjuvant therapy.

Kimberly L. Blackwell, MD: I found it helpful. When you look at the results in the extended setting, if I remember correctly, it’s like 5% risk versus 13% risk. That, I think, for the patient who’s not tolerating the therapy, can be very helpful. The patient who’s not tolerating, you actually want them to stop the endocrine therapy at 5 years. I think that can be helpful to the patient in deciding to stop it. So, yes, I think it’s a good assay, and especially given the very small separation of the curves at B-42, we need something to help us either make it wider or at that 5-year point.

Adam M. Brufsky, MD, PhD: NSABP is now thinking of actually actively planning to look at the tissue blocks from B-42 and doing a very similar trial to the bake-off that was done with TransATAC. So, hopefully, we’re going to have some data in the not-too-distant future to help us figure this out once and for all, which I think will be really good.

Mark E. Robson, MD: Do you keep doing bone protection if you continue for a long time?

Adam M. Brufsky, MD, PhD: Well, there you go. That’s the next question. I was going to ask that and I will talk about it for a minute. This is advanced breast cancer, we’ve got to get to metastatic at some point. But I would use it, again, at 3 to 5 years of some sort of bone protection. What do you do after 5 years? Because, the longer we use it, the more chance of atypical femoral fracture. That’s a great question. Because when you look at MA.17, the biggest side effect was osteoporosis. That was the real trade-off. If you’re doing 15 years of an AI, I think the osteoporosis rates were fairly substantial.

Kimberly L. Blackwell, MD: Remember, MA.17 was long before the denosumab prophylaxis studies, so I think the reality is the majority of bone loss with aromatase inhibitor occurs within the first 2 years of the estrogen deprivation on an AI. So, I’m actually stopping the bone protective agent at 5 years and following that, I’m a once-a-year bone density person.

Adam M. Brufsky, MD, PhD: They allow that? Insurance allows that?

Kimberly L. Blackwell, MD: Yes.

Adam M. Brufsky, MD, PhD: I don’t know what it is in your guys’ part of the country, but in western Pennsylvania, we can only get it every 2 or 3 years. They won’t pay for it.

Kimberly L. Blackwell, MD: I still pull out those ASCO Guidelines for bone health, that are now 10 years old, that say women on AI are at high risk and should have annual bone mineral density. It has been good enough.

Transcript Edited for Clarity