De-escalation of Cancer Treatment: When Is Less Really More?

This transcript has been edited for clarity.

Kathy D. Miller, MD: Hi. I'm Dr Kathy Miller, professor and associate director of clinical research at the Indiana University Simon Cancer Center. Today I'm joined by Dr Eric Winer, chief of the Division of Breast Oncology at the Dana-Farber Cancer Institute in Boston. Thank you for coming in, Eric.

Eric P. Winer, MD: Thank you for having me.

Miller: We have been hearing a lot about de-escalating therapy. Let's start with the basics. When people talk about de-escalation, what are they talking about?

Winer: Fundamentally, it's about doing less rather than doing more. It could be cutting back on the number of chemotherapy drugs that are being used or eliminating chemotherapy. It could be shortening the duration of therapy, or it could be simplifying a regimen in some other way. Of course, there have been trials focusing on de-escalation in breast cancer surgery and in radiation oncology for quite some time.

Miller: Those are great examples. You could argue that sentinel node biopsy was a huge de-escalation success.

Winer: Surgical de-escalation absolutely is. And in terms of time, hypofractionated radiation therapy is a de-escalation.

Miller: TAILORx would be another recent example—not necessarily changing the therapy but getting it to the people who need it and not burdening those who don't benefit.

Winer: This is a consequence of us understanding far more now than we did before, that one size does not fit all. Thirty years ago we treated breast cancer as if it was all the same. We've been slowly chipping away at that, and now I think everyone buys into the idea that patients may need very different treatments, even patients with the same subtype of disease.

Potential Areas for De-escalation

Miller: We're not quite at the point where we can truly individualize, but what are the areas that you think we're ready to begin de-escalating?

Winer: Some people don't like the term "de-escalated" and like the term "optimizing." But I'd argue that we've already de-escalated within the setting of HER2+ breast cancer. We've studied less intensive regimens.

In the APT study, led by my colleague Sara Tolaney, we looked at patients with node-negative, smallish tumors (essentially stage I disease, although patients with tumors up to 3 cm were included) receiving only paclitaxel and trastuzumab—no anthracycline, no carboplatin, nothing else. As people know, the results were pristine. It is important to recognize that it was a phase 2 trial. The weakness of this phase 2 trial is that we could not say that it was better than nothing. Some of these patients might have done well with absolutely no treatment. When you have a result with a very high distant recurrence-free survival, it's hard to imagine that you are going to do a lot better by adding more drugs.

Miller: At the San Antonio Breast Cancer Symposium (SABCS) meeting, we saw the updated results of the APHINITY trial. About one third of the patients enrolled in the APHINITY trial were node negative, many with tumor sizes that would have made them eligible for the APT regimen. Yet they were subjected to sequential anthracycline, taxane, dual HER2-targeted therapy. That is a very big difference for that group of patients.

Winer: It is a big difference. I argued strongly against having those patients in the APHINITY trial. Hindsight is always 20/20; in retrospect, it appears that the trial was actually weaker because of the inclusion of those patients.

A Failure of De-escalation?

Miller: I want to talk to you about the ATEMPT trial, which could be interpreted in different ways, maybe even as a failure of further de-escalating. The hope was that substituting trastuzumab emtansine (T-DM1) would preserve the pristine results of the APT regimen but be less toxic. Toxicity was different in my reading of the results but not necessarily less.

Winer: I agree with you completely. The ATEMPT trial was randomized to look at toxicity. The single T-DM1 arm did extremely well, and at 3 years, there are two distant recurrences and two local regional recurrences. So, from an efficacy standpoint, it's fine. But we were specifically trying to see if there would be less toxicity. The toxicities differed. For example, T-DM1 does not cause alopecia, but we were not able to capture that terribly well in the way we assess toxicity. The toxicities overall with a year of T-DM1 versus 3 months of paclitaxel were not so very different, which is one of the reasons why we are pushing hard to do another ATEMPT or an ATEMPT 2.0, where we give a much shorter duration of T-DM1 followed by straight trastuzumab.

Miller: I was sitting with my partner during that presentation at SABCS and we both immediately looked at each other and said, "It really should be 12 weeks of T-DM1." Essentially T-DM1 should be used just to deliver the chemotherapy for the same duration as the 12 weeks of paclitaxel, and then it should be switched to trastuzumab so that you don't have that ongoing toxicity.

Winer: We'll have to see where it winds up. We do not have approval yet to do such a study. We think we will. But it'll either be 12 weeks or 18 weeks, or somewhere less than 6 months, because it's really after the 6-month point that there tends to be more drop-off, presumably related to toxicity. I personally would like to keep it shorter than that, but we will see what happens. In terms of the results, paclitaxel and trastuzumab are still the standard at this point. The follow-up of 3 years is still a pretty short follow-up. It was not so dramatically better in terms of toxicity; in fact, we said that it really was not better. The only patient I would think about using T-DM1 in who has stage I or II HER2+ breast cancer is someone comfortable with the idea of therapy but very much fearful of the side effects with paclitaxel and trastuzumab—for example, a concert pianist who is terrified of neuropathy. But those patients are few and far between.

When Is De-escalation an Option?

Miller: Part of the discussion we've been having gets to some of the underlying tension and concern about de-escalating. There is an inherent risk here that we're giving people in some way less therapy than we had given them before and you could lose benefit. How do we make the decisions about what to de-escalate and how much and how far? How do we do this in a way that gets us there but is safe?

Winer: Initially, it's much easier to de-escalate in patients who have the earliest-stage disease because the risks are simply lower. If there is a small decrease in the overall effectiveness in absolute terms, it's going to be a very small difference. It becomes harder when patients have somewhat more advanced disease. Within our cooperative group system, we are going to have a trial in patients with stage II and III HER2+ breast cancer where everyone will receive paclitaxel, trastuzumab, and pertuzumab. For women who have a pathologic complete response, we're going to follow them with trastuzumab and pertuzumab only. That is a de-escalation trial. We are using pathologic complete response as a very powerful biomarker for the patient. You could argue that it may put some at risk. I am pretty comfortable with pathologic complete response as a very strong biomarker, but we have to demonstrate this in a trial.

Miller: Individual perceptions of where risk is are often very different. Is the risk more in the toxicities and bad things that come with our therapy or is it more in the disease?

Winer: We have to think about toxicity broadly. It's not just nausea and neutropenia; it's also time lost from work. It's the ability to do all of the things you want to do with your life in the subsequent years. Studies have demonstrated that some proportion of women who received chemotherapy are still not totally themselves a number of years later. It's a matter of weighing risks and benefits. Doctors, and more importantly, patients, put different weights on risks and toxicities.

Overtreatment Is Common

Miller: Getting even more precise in our estimates of risk of the disease becomes an important component of this because we, as a medical community, have gotten comfortable over the years with overtreatment. That is how we improved with those incremental gains when we could not segregate patients. We've gotten very comfortable overtreating people. I'm not sure our patients have ever been that comfortable.

Winer: I agree with you. Our patients want two things: They want to do well from the standpoint of the cancer, but they also don't want therapies that are going to beat them up. It's a fine line that one has to walk. But we know enough at this point that, in certain settings, this de-escalation approach makes a great deal of sense.

The other area is in ER+ breast cancer. Twenty years ago, we used to routinely give almost every patient with ER+ breast cancer whose tumor was greater than 1 cm a combination of hormonal therapy and chemotherapy. For a whole range of reasons, including tests like the recurrence score and studies like TAILORx, we now know that a large proportion of patients don't need chemotherapy.

The one place where I have some difficulty with de-escalation now is still in triple-negative disease—certainly stage II and III triple-negative disease. We're still in a place there where the outcomes are not terribly favorable. I can imagine people putting forward strategies to de-escalate therapy there, but we need to understand the biology a little more first.

Miller: I see this through the lens of the testes cancer history, and I think you articulated that perfectly. We don't typically start thinking about and being comfortable with de-escalation until we have an all-comers population that is over that 90%, or more, mark in survival. Then it becomes both impossible and impractical to think about improving that even more by adding more and more to everybody. We start to focus more on the acute and long-term toxicities and consider what is not needed and how to get there. In triple-negative, we're not at that level yet.

Winer: I agree. One of the challenges when people have a prognosis that is so good (ie, 90%-95% of them are remaining disease-free) is that it becomes extraordinarily difficult to do randomized trials. You are really talking about noninferiority trials, and they become huge and unwieldy. That is where, very cautiously, there is almost certainly a rationale for single-arm trials. But those single-arm trials only work if very few recurrences are being expected. It's much harder to know how a single arm with a 75% disease-free survival would perform versus some historical control.

Miller: They only work if the response is really excellent—and if it works. If your historical control suggests 95% and you come in at 92%, had it been a 10,000-patient noninferiority trial, that may be within your margin. But that is a difference that would have people questioning whether that is really the same or if we lost something.

Winer: With APT, one of the things that helped people feel comfortable is that we were looking, if I remember correctly, to exclude > 9% recurrence rate, and we would declare success if it were > 95%. Not only was it > 95%, but it was statistically significantly superior to 95%. That made people feel much more comfortable than they would have felt otherwise.

Miller: We will look forward to much more successes in de-escalating or "right-sizing" therapy for our patients. Eric, thank you for coming in and talking about this today.

Kathy D. Miller, MD, is associate director of clinical research and co-director of the breast cancer program at the Melvin and Bren Simon Cancer Center at Indiana University. Her career has combined both laboratory and clinical research in breast cancer.

Eric P. Winer, MD, is chief of the Division of Breast Oncology at the Dana-Farber Cancer Institute in Boston. He has designed and conducted clinical trials that have changed clinical practice and paved the way toward more personalized treatment of patients with breast cancer.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube

Comments

Commenting is limited to medical professionals. To comment please Log-in.

Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.