Not to Diss Denosumab, but More Data Needed in Breast Cancer

Kathy D. Miller, MD; George W. Sledge, MD


February 22, 2016

This feature requires the newest version of Flash. You can download it here.

Kathy D. Miller, MD: Hi. I'm Kathy Miller, professor of medicine at the Indiana University School of Medicine in Indianapolis. Welcome to Medscape Oncology Insights and the 2015 San Antonio Breast Cancer Symposium.

The role of bisphosphonates in adjuvant therapy for patients with breast cancer has been hotly debated for the past several years. At this meeting, we have results of one of the first trials looking at denosumab in the adjuvant setting. To give us some perspective on these data and make sense of all the controversy, I am joined by Dr George Sledge, professor of medicine and chief of the Division of Oncology at Stanford University in California. Welcome back, George.

George W. Sledge, MD: It is a delight to be here, Kathy.

Adjuvant Bisphosphonates: So Many Data, Yet No Consensus

Dr Miller: First, let's start with the basics. Why was there interest in bisphosphonates in the adjuvant setting in the first place?

Dr Sledge: This is a very old story. It dates back 20 years to trials that were done in mice that showed that the use of bisphosphonates could actually prevent the development of bone metastasis. After the large body of data emerged for bisphosphonates in the metastatic setting showing that they had clear benefits in terms of preventing skeletal-related events, it seemed logical, on the basis of preclinical data, to move this into the adjuvant setting. This has now been done with bisphosphonates in well over 30 trials of varying size and varying quality.

Dr Miller: I remember that some of those first trials had very different results.

Dr Sledge: Yes, indeed.

Dr Miller: Some were positive, but not just in preventing bone metastasis. Some of those early trials suggested that bisphosphonates would prevent soft tissue and organ metastases.

Dr Sledge: They did, and this is the problem with small trials. Small trials can lead you astray fairly quickly. Now that we have a much larger database, the suggestion is that those nonskeletal events are probably a minimal part of the story.

Dr Miller: So this is a breast cancer story. Whenever we have 30 trials—some large, some small—with disparate results, we go to Oxford.

Dr Sledge: We go to Oxford.

Dr Miller: And the meta-analysis.[1]

Dr Sledge: Yes, indeed.

Dr Miller: Does that help bring some order to this chaos?

Dr Sledge: I believe so. If one looks at these studies individually, they are all over the map. They vary in size, they vary in which bisphosphonate was used, they vary in terms of the duration of bisphosphonate use, and not surprisingly, one gets highly variable results.

These individual trials probably missed what may turn out to be the most important signal with bisphosphonates, which is menopausal status. The fairly clear suggestion from the meta-analysis—but also from a number of the individual trials—is that this [adjuvant bisphosphonate use] is, in essence, ineffective in premenopausal patients but relatively effective in postmenopausal patients.[1] That certainly does not come out in most of the individual trials, which were not designed to look at menopausal status.

There has been less than a universal jump onto the bisphosphonate bandwagon.

Dr Miller: That brings us to another problem. The meta-analysis, if I recall correctly, had a roughly 3% or 4% improvement in the postmenopausal group. That is a similar improvement to what we have seen with changes in hormone therapy and changes in chemotherapy. It feels to me that there has been less than a universal jump onto the bisphosphonate bandwagon.

Dr Sledge: I think that is absolutely the case. Certainly, the guideline committees have been relatively slow in embracing this. Physicians, as a group, have been slow in embracing this. Some of us, of course, have used these agents whenever we had a patient who is on an aromatase inhibitor and has a little bit of bone thinning, but it is certainly not something that we have done universally.

Dr Miller: I have a bunch of questions for you. Why do you think there has been this different response to a similar magnitude of benefit? You have a clinical practice. Have you jumped on the bandwagon? Are you using this in all of your postmenopausal patients?

Dr Sledge: I have not, yet. Though, I must say that after this meeting I am probably more likely to. Can we start with the biology? I think the biology here is kind of interesting.

Bisphosphonates Appear to Act Directly on Osteoclasts

Dr Miller: If you could explain the biology to me I would be grateful, because I have not heard a great explanation.

Dr Sledge: I don't know that I can explain it well, but we actually have some preclinical data now. It is quite fascinating because the preclinical data are, in fact, very recent, just from the past year or two from a very good Australian bone group that took an MDA-MB-231 model of bone metastasis.[2] For our listeners, this is an estrogen receptor–negative breast cancer model, a triple-negative breast cancer model.

Dr Miller: One huge extrapolation that we have to make, besides animal-to-human, is the totally different disease phenotype.

Dr Sledge: Totally different disease phenotype. And, of course, most bone metastases are estrogen receptor–positive, so again, that is a qualifier.

There is something about menopause that allows cancer cells to grow in bone.

But the study was quite interesting. This was a study where you do an intracardiac injection and then the cancer cells go to the bone. But here is what is interesting: If you inject this into your regular old premenopausal mouse, the cancer cells go to the bone and they don't divide. But if you then do an ovariectomy on these mice, the bone metastases take off. There is something about menopause that allows cancer cells to grow in bone. Then, in the same model system, if you give a bisphosphonate, you are able to ablate the development of bone metastases.

Now, it is a big extrapolation to go from a mouse to a human. Bone metabolism differs and the cancer cells differ, but at least there is the suggestion from preclinical modeling that this may be important.

Why is this the case? What does it suggest? Well, the fact that the bisphosphonates ablate this [the development of metastases ] after it [cancer] occurs suggests that they are having a direct effect on osteoclasts, that somewhere in the endosteal plate where the bone metastatic niche is felt to be, you are preventing the development or progression of cancer cells. That is as much biology as I think we have right now, but it is interesting biology.

Clinical Dilemma: Which Drug and for How Long?

Dr Miller: So take me to your clinic. You mentioned that you have not jumped firmly on board, but the meta-analysis data are as compelling as several other meta-analyses data[3,4] that have had people jumping firmly on board. You just gave us some biology that is at least consistent with the clinical data. What has prevented you from fully embracing this?

Dr Sledge: What has prevented me has been my confusion about how we should actually use this.

Dr Miller: I was hoping that you would get me un-confused, because I have a similar problem. The studies are all different drugs, all different doses, all different intensities. If I were going to do it, I am not sure that I would know how.

None of these studies have actually compared doses and schedules.

Dr Sledge: Should I use clodronate for a year or should I use zoledronic acid for 5 years?

Dr Miller: And should your zoledronic acid be every 3 months, every 6 months, more frequently in the beginning, less frequently afterwards?

Dr Sledge: Indeed. And none of these studies have actually compared doses and schedules, even for zoledronic acid, which has been studied most commonly. I think the simple answer is that we don't know. Because of that, I think the medical oncologist in me—who has used a lot of zoledronic acid over the years, primarily in the metastatic setting—is concerned about the side effects of zoledronic acid, particularly osteonecrosis of the jaw. Although it has been a rare side effect in these adjuvant trials, it does occur.[5] It occurs at a low rate, but if it does occur, it is a devastating side effect for an otherwise healthy patient. That has been my concern, and I suspect that it is a concern of many oncologists.

Denosumab to the Rescue?

Dr Miller: So perhaps at this meeting, denosumab will come to the rescue?

Dr Sledge: Indeed.

Dr Miller: At least in terms of confusion about which agent, what dose, and what schedule. Which takes us to the ABCSG-18 trial.[6] This is not the first time that our Austrian colleagues have come to our rescue when we were confused.

Dr Sledge: Indeed. Our Austrian colleagues have previously done a very fine zoledronic acid trial in the adjuvant setting, which was a positive trial.[7] Now they come to us with denosumab given twice yearly.[8] This is a drug given by a simple subcutaneous injection. The results, while they are not nearly as long in follow-up and not nearly as robust statistically, look very similar to the overview results with zoledronic acid and the other bisphosphonates. That is, similar improvement was seen in terms of disease-free survival in the adjuvant setting, and the time points were tracking with what we saw with zoledronic acid and the other bisphosphonates in the meta-analysis. That is very reassuring.

Beyond that, of course, we have a large body of metastatic data comparing the two drugs.[9,10,11] To summarize that data, we would say that denosumab appears to be at least modestly more effective than zoledronic acid in the metastatic setting. From a toxicity standpoint—and I think this is what is relatively exciting about Michael Gnant's presentation at this meeting—there was essentially no toxicity.

Dr Miller: And it is certainly easier to put into practice. Subcutaneous injection is always easier than intravenous (IV), especially in our adjuvant patients who may not have needed IV therapy at all or are otherwise done with their IV therapy. Renal function is not really a concern, nor are the flu-like issues. We do need to think about osteonecrosis of the jaw, though.

Dr Sledge: We do, and if you look at this agent given on a frequent basis in the metastatic setting, we certainly see osteonecrosis of the jaw with denosumab. However, none was seen in this trial, and I suspect that this may be a function of simply using it twice yearly rather than using it on a monthly basis.

But Does Denosumab Increase Survival?

Dr Miller: A criticism of the Austrian trial, which I have already heard folks talking about, is that it was stopped early on the basis of a decrease in fracture risk. It is a good thing for our patients to prevent fractures, but that is going to hamper the ability to have longer-term follow-up for their primary endpoint, which is: What happens to the breast cancer?

This is a curiosity of the trial: The primary endpoint, in fact, was not disease-free survival.

Dr Sledge: Yes. This is a curiosity of the trial: The primary endpoint, in fact, was not disease-free survival. The primary endpoint was bone loss and bone fracture rate, so it met its primary endpoint and met it in spades: There was a 50% reduction in terms of bone fracture rates. The data-monitoring committee—like many data-monitoring committees—felt that trials live and die by their primary endpoint and we have met our primary endpoint. It is an ethical issue to try to reduce the bone fracture rate for a patient who is on an aromatase inhibitor and is postmenopausal. Therefore, we feel that we should be able to offer this to patients.

Starting in 2016, patients in this trial who are on the placebo arm will be allowed to switch over to denosumab. What does that mean in terms of the disease-free survival outcome? It means that everything past this point, in essence, will be contaminated.

Dr Miller: Complicated, at least.

Dr Sledge: Complicated, certainly. The P values that we have for disease-free survival, at this point, are right on the borderline—my memory of it is .05112 or something along those lines—so we are talking about a borderline statistical significance in terms of disease-free survival. If we want to look at a long-term outcome, such as overall survival, will we have the power or strength to see that in a single study? Perhaps that means we need a meta-analysis.

Dr Miller: We may need to turn back to Oxford in a few years because there are several other studies looking at denosumab in the adjuvant setting.

Dr Sledge: Indeed.

Dr. Miller: Unfortunately, that may take our simple world of one drug, one dose, one schedule and mess it up again because the trials are looking at it in different ways.

Dr Sledge: We will undoubtedly see differences in terms of dose, duration, frequency, and follow-up, of course, in the trials themselves. Having said that, if these agents are at least comparable in the metastatic setting—and denosumab is perhaps slightly better—I think there is a reasonable likelihood that we would see something similar in the metastatic setting. The fact that we would see it with less clinical toxicity, though the financial toxicity may be real, is a good thing.

What Happens Now?

Dr Miller: Many of the folks who tune in to our program are practicing oncologists. We have very busy clinic schedules. This is a very common disease and perhaps the most common subset of our breast cancer population. Come Monday morning, when you are back in Stanford and faced with a 63-year-old postmenopausal woman starting her adjuvant aromatase inhibitor, will you be giving her denosumab on the basis of these data?

Dr Sledge: I am certainly going to discuss it with her. I am going to discuss bone agents with her. The patient may or may not have insurance that allows her to receive denosumab, which is still an expensive drug. On the other hand, she might be able to get zoledronic acid. I am going to have to talk with her about that, those financial toxicities. I am also going to have to look at the patient's disease itself.

It is relatively difficult for me to imagine, for instance, using it in a patient with a 6-mm strongly estrogen and progesterone receptor–positive tumor and a low Oncotype DX recurrence score, which we know from our recent analysis of the ECOG TAILORx trial[12] has a 99.3% distant disease-free survival at 5 years.

Dr Miller: Suppose she is 55 years old and has been postmenopausal for 3-4 years, but she has a 4-cm node-positive tumor and her recurrence score is 45.

I don't think this is something that I would use in a blanket fashion in my personal practice.

Dr Sledge: Well, this is the art of medicine. This is now defined as a high risk for recurrence, and the several percentage points in terms of disease-free survival and perhaps overall survival are going to be valuable to such a patient.

We are still going to need to individualize. I don't think this is something that I would use in a blanket fashion in my personal practice.

Dr Miller: I have a couple of other clinical questions for you. Suppose that the same patient with the high-risk tumor has an estrogen receptor–negative tumor. The ABCSG data[7] were limited to patients with estrogen receptor–positive tumors starting an aromatase inhibitor, but she does have a very high risk for recurrence. The other bisphosphonate trials were not all limited to the estrogen receptor–positive space. Are you willing to make that extrapolation at this point?

Dr Sledge: We don't have the data. Where we don't have the data, it is hard to say for sure. Personally, I'd feel reasonably comfortable with that because I don't think that the mechanism of action has anything to do with the estrogen receptor positivity of the cancer.

Dr Miller: No, but the predilection for bone versus other sites of disease is a bit different.

Dr Sledge: Indeed. If we look at the Oxford meta-analysis,[1] they did test for heterogeneity to see if there was any benefit on the basis of subtype, and there was no benefit on the basis of steroid receptor status. That certainly implies that if the primary action is on the osteoclast and not on the breast cancer cell, it is probably independent of steroid receptor status. If it is true for zoledronic acid, I would expect it to be true for denosumab as well.

Dr Miller: George, thank you for joining us again. It seems like the bandwagon is picking up steam.

Dr Sledge: Perhaps.

Dr Miller: We will have you back again to see if that comes true. Thank you for joining us. This is Dr Kathy Miller from the San Antonio Breast Cancer Symposium.


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.