Circling Back: Duration of Adjuvant Endocrine Therapy for Breast Cancer

January 16, 2014
Nancy E. Davidson, MD

Rachel C. Jankowitz, MD

In this Circling Back feature, authors of the review article “Adjuvant Endocrine Therapy for Breast Cancer: How Long Is Long Enough?” discuss some of the especially interesting points made in the commentaries.

Points Made in Commentaries on “Adjuvant Endocrine Therapy for Breast Cancer: How Long Is Long Enough?” Are Addressed by Authors Jankowitz and Davidson

In the December issue of ONCOLOGY, Rachel Jankowitz, MD, and Nancy Davidson, MD, of the University of Pittsburgh Cancer Institute and UPMC Cancer Center published a paper that addressed the very important question of how long adjuvant endocrine therapy should be continued in patients with early-stage breast cancer. As with all ONCOLOGY articles, Drs. Jankowitz and Davidson’s review was accompanied by two commentaries that presented different perspectives, different approaches to this question. Because the authors of these commentaries made some especially interesting points, we wanted to provide Dr. Jankowitz and Dr. Davidson with the chance to address some of these. So, I’m here today with both authors, and I’m looking forward to hearing what they have to say.

ONCOLOGY: Let’s start with the first commentary, which was by Jack Cuzick, PhD, head of the Centre for Epidemiology, Mathematics and Statistics at the Wolfson Institute of Preventive Medicine in London. This commentary focused on what Dr. Cuzick calls tamoxifen’s “carryover effect,” which he maintains had the result of skewing the results of the National Surgical Adjuvant Breast and Bowel Project B-14 trial and the Scottish trial, such that the data from these trials suggested that 10 years of tamoxifen provided little additional benefit compared with 5 years, while at the same time increasing side effects, when in fact that may not be the case. What do you think of Dr. Cuzick’s assertion that tamoxifen treatment has a carryover effect, and if you agree that this is correct, what implications does this have, both for trial design and for clinical practice?

NANCY DAVIDSON: I’m Nancy Davidson, and I’m going to tackle this question. I think Dr. Cuzick is correct, that it is well accepted that there is a carryover effect of adjuvant tamoxifen. This has been demonstrated in the clinical trials that have come together as part of the Early Breast Cancer Trialists’ Group, and we believe that this is absolutely true. This is good for patients, in the sense that, even if you stop therapy, you can continue to see benefits. It is a lot tougher for trialists, because it makes it harder to design a very clean clinical trial. I think that Dr. Cuzick might well be right that the carryover effect did complicate our interpretation of some of the initial trials of prolonged tamoxifen administration. However, I think the other part of it is, in retrospect, those trials were very small. In aggregate, the B-14 trial and the Scottish trial had a couple thousand patients, whereas the newer trials-the aTTom trial and the ATLAS trial-were considerably larger, and so I think that larger statistical power, because of the greater number of patients, also made it easier to see the effect of longer-term tamoxifen.

ONCOLOGY: Another point that Dr. Cuzick makes in his commentary is that a greater focus on distant recurrence as the primary outcome would permit an earlier but still reliable verdict on treatment efficacy. Do you think distant recurrence is a valid surrogate endpoint? Dr. Davidson, do you want to take that one?

NANCY DAVIDSON: I think that endpoints for adjuvant clinical trials in breast cancer are sometimes a little bit controversial. Surely the endpoint that we’ve used the most is survival, and that’s a pretty concrete endpoint. Everybody can agree on the importance of improved survival. It obviously takes a really long time, and in a disease like breast cancer, where the outcomes are very, very good these days, it can take a very large number of patients to go into a trial to see those kinds of benefits. I believe that many would say that it’s the gold standard, although we increasingly recognize that many breast cancer survivors do not die from their breast cancer, that their survival is actually affected more by other non–breast cancer–related illnesses, and it’s particularly true in some of these adjuvant hormonal therapy trials, which have frequently been done in older women, particularly in postmenopausal women.

Having said that, I think that I would be reluctant to have recurrence be the only endpoint for these trials, and I guess it’s true to say that it’s a valid surrogate endpoint, but I think attending to both of these endpoints is going to be increasingly important. I do think it is likely that we’re going to have less emphasis on all these things, because I think the fact that we’re going to see very few of these large adjuvant therapy trials in the future, the notion that we’re going to be able to test these interventions in thousands of women is not going to be realistic going forward, and so I think that part of what this implies is that we have to really be thinking about new clinical trial designs, trying to identify important biomarkers, important biological targets, identifying agents that might affect patients who have those targets, and we would hope in that way that we would be able to refine our clinical trial designs and be able to “trim them down” to make them more feasible in the current era.

ONCOLOGY: Thank you. The last point from Dr. Cuzick’s commentary that I wanted to bring up and get your reaction to is his argument that, based on the results of the Breast International Group 1-98 trial, less than 5 years of an aromatase inhibitor (AI) may be sufficient in certain low-risk patients. Now, Dr. Cuzick is certainly not saying that the BIG 1-98 data show conclusively that shorter aromatase inhibitor treatment is OK in certain patients, but would you agree with him that the data at least strongly suggest this? And do you think a trial of 2 years vs 5 years of an aromatase inhibitor in low-risk patients should be pursued, as he suggests?

NANCY DAVIDSON: The duration of adjuvant endocrine therapy was the whole purpose of our review. This remains a really interesting question. I guess I would start by looking back at our experience with tamoxifen, and there we have a series of clinical trials, which did inform the Oxford overview, which suggest in some that 2 years is better than 1, and 5 years is better than 2. I think that’s well worked out for the drug tamoxifen-and as we just talked about, it may well be that 10 years is better than 5. Therefore, there is certainly the feeling in our world that adjuvant endocrine therapy is going to have to be on the longer period of time than on the shorter period of time. Now, I cannot say with certainty whether the findings that we found with tamoxifen would translate into similar findings with an aromatase inhibitor. The BIG 1-98 trial included 5 years of some kind of endocrine therapy for all patients; no patient got less than 5 years of therapy. Patients got either 5 years of an aromatase inhibitor or 5 years of tamoxifen, or 5 years of various types of sequences, so I don’t personally think that we can take away from that trial that less than 5 years of an AI may be sufficient for certain low-risk patients. I think everybody got 5 years of something, and I have trouble with the assumption-or the suggestion-that Dr. Cuzick has taken from that.

Having said that, I also think that we’re at a point in the adjuvant therapy of many women with breast cancer, where the results are extremely good, and that it may be hard for us to improve on them in a meaningful way, and that one thing we might wish to do then is to see whether or not we can refine the way we give that therapy so that we can tailor it, so we can optimize it, and it is a rational question to ask whether or not shorter durations of an aromatase inhibitor would be as effective as a longer duration of an aromatase inhibitor in women. This would unfortunately require a very large trial, as we just discussed, because the event rate would be so low. One would hope that perhaps we would be able to identify some sort of biomarker that would allow us to predict for a woman her likelihood of outcome, and perhaps it would help us to refine the time that she would take her aromatase inhibitor.

ONCOLOGY: That kind of leads into the first question I wanted to ask about the second commentary, so let’s move on to that one now. Like Dr. Cuzick, the authors of that commentary, Phillip Blanchette, MD, and Kathleen Pritchard, MD, of the University of Toronto, also expressed the hope that some low-risk women might do OK with shorter-course endocrine treatment. However, rather than proposing a trial of 2 years vs 5 years to further explore this possibility, as Dr. Cuzick does, Drs. Blanchette and Pritchard suggest focusing on finding biomarkers that might be able to identify those patients for whom shorter-course aromatase inhibitor treatment would be sufficient. Dr. Jankowitz, let me ask you, which of these two approaches to establishing a sound basis for using shorter-course aromatase inhibitor treatment in at least some patients would you favor, and why?

RACHEL JANKOWITZ: Well, I think I’d have to agree with Dr. Davidson. She’s already pointed out the challenges, from the trialists’ perspective, of doing further large adjuvant studies, and it comes down to the length of time it would require to follow up those patients to answer that question definitively, so I definitely would favor biomarker development in terms of steering patient selection for therapy. I just think that it’s very clear that most likely when you look at the entire group of women being treated with these therapies, there’s going to be a trend for better outcomes with longer therapy. But I believe, and I agree with Dr. Pritchard, that certain women are lower risk, but it’s not always clearly defined by their immunohistochemistry, and so I would favor biomarker development in terms of identifying those low-risk women in a more timely fashion.

ONCOLOGY: Thank you. The last thing I was curious to hear your thoughts on is the treatment algorithm that Drs. Blanchette and Pritchard have proposed for guiding therapy in postmenopausal patients until more clinical trial data become available. Their algorithm would start all postmenopausal patients on 5 years of an aromatase inhibitor, then choose, based on a discussion with the patient of her particular risks and benefits-and any emerging data-either to continue the aromatase inhibitor for an additional year, with re-evaluation annually for possible continued treatment, or to follow the 5 years of aromatase inhibitor therapy with 5 years of tamoxifen. What do you think of this approach? Dr. Jankowitz, do you want to take this one?

RACHEL JANKOWITZ: Sure. I think we’re all facing this question daily in our office with our patients sitting in front of us, asking what to do when they’re completing their 5 years of an aromatase inhibitor, and it does need to be addressed in the meantime, despite the fact that we don’t have trial outcome data yet to definitively answer that question. I like their algorithm. The thing that’s attractive about it is, particularly when you have that patient who’s finishing her 5 years of an AI, and her bone density has declined, and she’s highly motivated to continue adjuvant endocrine therapy and maybe on the higher end of the risk spectrum-then you know, I personally sometimes do consider extending her endocrine therapy further, and it has crossed my mind to put that type of patient on tamoxifen, to be thinking about her bone density, because then you can avoid further bone loss, but it has to be clear that that’s not an entirely evidence-based approach at this time, but I think it is a reasonable algorithm.

ONCOLOGY: Thank you. And did you have any other remarks you wanted to make, Dr. Davidson?

NANCY DAVIDSON: This is a really tough and common question, as Dr. Jankowitz pointed out. We have a number of clinical trials that are going on right now that are trying to look at longer durations of aromatase inhibitors, and we can hope that we’ll hear the results of those in the next couple of years. In the interim, I would take a slightly different approach than what’s been proposed. First, I’m very open with patients who I do start on aromatase inhibitors-and it is true that’s most postmenopausal women; I’m open with them that when we start that, we’re looking at 5 years of therapy, that we hope that emerging data is going to tell us about longer durations, but right now the data are at 5 years, and so I put patients into that with the expectation that they’re going to stop therapy at 5 years. Now in some cases when they come to the 5-year boundary, they may come and say, “well, you know, but I feel fine,” and we know longer durations of tamoxifen are better, and so why wouldn’t this be true of an aromatase inhibitor? And so, on a case-by-case basis, I will sometimes extend aromatase inhibitor treatment for a year at a time, as Drs. Blanchette and Pritchard suggested, with a real full discussion about the pros and cons, the lack of information about the benefit of longer therapy, and the known downside of therapy, the bone loss that Dr. Jankowitz mentioned, as well as the achiness, and our concern, a little bit of concern that remains about long-term consequences of these drugs, because we simply haven’t given them for long enough to know them.

I’m less keen about the option of switching from aromatase inhibitor to tamoxifen. We have no information on doing that-with 5 years and 5 years. We certainly have information about doing it for shorter durations, as part of the BIG 1-98 trial, which did these kinds of sequences, but we don’t have this “5 years plus 5 years” approach, and so I have no information to suggest that there would be benefit by doing it, and I do know that the downsides of tamoxifen, namely endometrial cancer and clot formation, are elevated in the older postmenopausal woman. I worry in that particular setting that we might be in a position where we have a therapy of uncertain benefit but a known side effect profile that is prominent in the older patient population. I’m not excited about that as a strategy in my own practice. I think that just reflects lack of information we have on this, and how really good doctors can take a look at the same information, and the same information vacuums, and come to slightly different conclusions.

ONCOLOGY: Thank you both. This has been a great exploration of a very important question, and I’m sure our readers will enjoy hearing what you had to say.

NANCY DAVIDSON: Thank you for the chance to participate.