Johann S. de Bono, MBChB, PhD, MSc; Howard I. Scher, MD


October 08, 2012

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


Johann S. de Bono, MBChB, PhD, MSc: Hello, I am Johann de Bono, Professor of Experimental Cancer Medicine at the Institute of Cancer Research and Honorary Consultant Medical Oncologist at the Royal Marsden Hospital in Sutton in the United Kingdom. We are at the European Society for Medical Oncology (ESMO) meeting in Vienna. Welcome to this edition of Medscape Oncology Insights on Prostate Cancer. Joining me today is Howard Scher, Chief of the Genitourinary Oncology Service at the Sidney Kimmel Center for Urologic and Prostate Cancers at Memorial Sloan-Kettering in New York. Howard, it's a pleasure to speak with you. It has been exciting at ESMO. Of the data presented here, what strikes you as being most noteworthy, most exciting?

"Riches We Have Never Seen Before"

Howard I. Scher, MD: The field has changed dramatically in the past 3 years. We have seen 5 drugs hit survival endpoints in phase 3 trials. Before that, the last trial to achieve this endpoint was in 2004, which led to the approval of docetaxel. Now, we have riches that we have never seen before, and we are seeing the next generation of compounds that will have to develop a path to regulatory approval. That is much more difficult than it was previously, and we have had the privilege of working together on several of these compounds, notably the androgen-receptor signaling inhibitors: abiraterone, which was approved 2 years ago, and most recently enzalutamide.

We have established some important paradigms, for example, that these tumors are not hormone refractory. It was particularly noteworthy that these agents improved survival in patients who had not only failed hormones but also had failed chemotherapy. At ESMO, we have heard about 2 new compounds that target the androgen receptor (ODM-201[1] and ARN-509[2,3]) both of which have very promising activity. But how these agents will go forward is going to be a challenge, particularly with these newer agents already approved.

Dr. de Bono: So, we have 4 or 5 new agents now, with some having similar mechanisms of action: docetaxel, cabazitaxel, maybe abiraterone, and enzalutamide. The challenge now is sequence combinations. What happens pre-chemotherapy? What happens with docetaxel? How do you think that is going to pan out? And what should we do next?

Better Trial Designs for Better Results

Dr. Scher: We have to think very carefully about the trials we design. Previously when we started working with abiraterone acetate and enzalutamide, there was no standard of care at the time those trials were designed, so in essence those trials were run against placebo. Obviously you would like to move the hormonal therapies and some of the more promising biologics earlier in the course of the disease, but what is the control arm for those studies? Will you be able to show a survival benefit recognizing that there are, fortunately, therapies for the patients who do not respond on these trials after they have progressed? We are seeing what in trial terminology we call "contamination," which is a benefit to patients but may in fact blunt the survival benefit of an active compound. So, now we are focusing on intermediate endpoints, short of survival, that in themselves can be demonstrated to be an approval clinical benefit endpoint, as well as on the development of surrogate endpoints, which will allow us to predict the clinical benefit downstream.

Dr. de Bono: It is a key issue. Abiraterone and enzalutamide both have significant activity, both pre- and post-chemotherapy, and perhaps more activity pre-chemotherapy. It is likely that abiraterone will be used more pre-chemotherapy in the near future. With these agents moving forward, do you think that how we use chemotherapy is going to change?

Still a Place for Cytotoxics

Dr. Scher: There are clearly patients who do not respond either to enzalutamide or abiraterone. The question will be, what is their response to cytotoxics? From most of the experimental data, there are cells that will not respond to hormonal manipulation present at the time the tumor is diagnosed. In fact, a French study was presented at this meeting that essentially combined hormones in chemotherapy vs hormones alone; it did not show a benefit. Yet we know that later in the disease when the tumor is growing and resistant, the same agent prolongs life. So, we are going to have to think differently in how we design our trials. If you are designing a combination trial vs a sequential trial, you may have to focus on downstream endpoints -- for example, preventing symptoms, preventing skeletal-related events, or other endpoints that are clearly beneficial to patients and can be objectively measured but will require different eligibility criteria. These events fortunately do not occur that often, so it is going to be harder to show the impact of the treatment.

Dr. de Bono: That French study was a relatively small study and looked at those types upfront first-line. Other studies that will look at this question will be challenged by the crossover issue because, presumably, part of the negativity of that trial on overall survival was the crossover issue. That will be an issue for all of these compounds, won't it?

Dr. Scher: Right.

Dr. de Bono: Do you think we are currently stuck with overall survival endpoints and studying these drugs later?

Overall Survival: The Endpoint We're Stuck With?

Dr. Scher: We are always challenged by the RECIST [Response Evaluation Criteria In Solid Tumors] criteria. Prostate cancer does not spread in discrete nodules that can be measured objectively, and how to assess bone has remained very problematic. There was a very important presentation from Chuck Ryan.[4] He looked at the prechemotherapy Cougar 302 trial,[5]which used a consensus-developed endpoint as a biomarker and showed a very strong association with survival. In addition, he showed very importantly that that endpoint could be reproducibly measured in 15 countries. In the past we knew that there was a lot of interpretation to scans, and it has been very challenging to put in the database a description of a scan that says "it's worse," or just "progression" without some quantifiable objective measurement. Clearly that is not all the information we can get from a bone scan, but it was something that was reproducible, objective, and associated with survival. If we can show that in other trials, where there is survival benefit, then we will be closer to having a potential regulatory endpoint short of survival.

Dr. de Bono: My concern is, for example, with other agents that have different mechanisms of action, such as cabozantinib. Can we extrapolate from the abiraterone data in the Cougar 302 study to other compounds? I guess we can't. But a strong association between progression-free survival, radiographic progression-free survival, and overall survival is encouraging, but there is obviously still a lot of work to do to move this forward.

Steroids: No One-Size-Fits-All

Can we discuss the steroid issue? Obviously if abiraterone goes into the prechemotherapy space, we are going to have to give it with steroids. We have some concerns about steroids and the benefit they can give some patients that we have shown in the past, but eventually the steroids could be detrimental. Can you comment on that and the presentation you had today?

Dr. Scher: We looked at the enzalutamide data and at the survival of patients who were on steroids when they started the protocol,[6]and we have seen over time that there can be changes in the androgen receptor such that agents that are initially beneficial can become agonists and stimulate growth. To our surprise we found that in both the placebo-treated as well as the enzalutamide-treated patients, those who were on steroids had an inferior survival. This was more of a hypothesis-generating observation. We did our best to correct for prognostic factors, counting for the drug effects for the stratification factors on entry in terms of pain, for example, and we still saw that in this population, steroids appear to be detrimental.

On the other hand, we have also seen that steroids can be given as treatment. Going back 2-3 decades, we have used glucocorticoids. Ian Tannock was the first to show this subjectively. In my early days at Sloan-Kettering, I happened to be visiting him in Toronto when he was just putting together one of the first objective quality-of-life studies to show a clinical benefit -- but not a clear survival benefit because that was not the question he was asking. He demonstrated that patients benefited from steroids, and that became the basis for the control arm of what was later the mitoxantrone plus prednisone vs prednisone alone trial that established for the first time that a systemic chemotherapy can provide the benefit of pain relief.[7]This study led to the approval of mitoxantrone for that indication.

Necessary to Characterize Disease Over Time

Dr. de Bono: Are we going to monitor for androgen receptor mutations that are activated by prednisone or other compounds, or will it be more complicated than that?

Dr. Scher: The simple answer is that it is always more complicated, as you know. We need to pay more attention to the previous treatments that a patient has received. If we look back at some of our phase 3 studies, we would generally count hormones, how many different hormonal regimens that a patient received. As you have seen in the laboratory, if you take a series of prostate cancer cells and allow them to grow under different stimulation, depending on whether it is low-dose antiandrogens or low-dose chemotherapy, the tumors that emerge on the other end will be very different. Not only must we consider this from a clinical point of view at entry, but we must also think about it in terms of what is happening biologically. Although we are interested in characterizing primary prostate cancer when a patient is diagnosed, the tumor that is progressing in bone that is causing symptoms is likely to be very different from the primary. It becomes imperative that we have the tools to characterize the disease at the time we are trying to make a treatment decision. Fortunately, we are working together as part of a large "Stand Up to Cancer" program to do just that.

Dr. de Bono: So, the message is that steroids can be helpful to some patients and may become detrimental at some point later on as the cancer adapts.

Dr. Scher: Right. There also are side effects from the steroids, particularly in patients receiving them long term -- for example, thinning of the skin and muscle weakness. The benefit can be very dramatic. On the other hand, if the patient is taking steroids for a longer period of time, there are downsides.

Dr. de Bono: If abiraterone goes prechemotherapy (and that is possible in the next year or so on the basis of the 302 data[5]), we will have abiraterone, prednisone, and docetaxel prednisone. It is likely that these patients are going to be receiving a lot of prednisone for a long period of time. Do you have any recommendations as to how we manage the prednisone in this regimen in view of our concerns? Do we try to stop it eventually?

Dr. Scher: When we were doing the trials in the postchemotherapy population, we would try to stop the prednisone in many patients, but in some we simply couldn't. These patients become steroid dependent and go through a period of steroid withdrawal when they become more fatigued, and in some cases, pain can recur. It's not a simple one-size-fits-all approach. If you looked at the patients on the trials who were receiving steroids, they were generally sicker than the patients who were not on steroids. Nevertheless, in both scenarios, the steroids appeared to be detrimental to survival later on.

A Tisket, a Tasket, a Biomarker Basket?

Dr. de Bono: We are very excited about biomarkers involving circulating tumor cell studies. How do you think this is panning out? Where are we going with circulating tumor cells?

Dr. Scher: Since the original paper showing that circulating tumor cells are prognostic, both pre-treatment and post-treatment, in patients treated with taxanes, we have independently shown, with a number of other agents, a similar relationship. We were able to embed the circulating tumor cell question in 2 large phase 3 trials, the first being the Cougar 301 trial[8]and the second being AFFIRM.[9]We are analyzing the data. We made a preliminary presentation at the American Society of Clinical Oncology meeting about a year and a half ago suggesting a strong association and potential as a surrogate biomarker. We are in the process of finishing up that analysis to see how strong that relationship can be. The important issue is that we took the time to embed the biomarker question in the phase 3 trial. It is now clear that unless we study these biomarkers, (whether it is imaging, the immune response, circulating tumor cells, or circulating DNA) and embed the question in the phrase 3 trials, we will never know whether these are potential surrogates for survival. If we are limited to survival-based trials, the field will have a lot of trouble moving forward and many active drugs may not make it to the clinic.

Dr. de Bono: I strongly concur. These data are very promising, but we may need a whole multiplex set of biomarkers together.

Dr. Scher: As our colleagues in the agency said, we are looking to develop a biomarker basket. The idea that one particular biomarker will explain all of survival is a pipe dream.

Cabozantinib: Dramatic Relief From Pain

Dr. de Bono: Looking at other presentations at ESMO, we have seen some data on cabozantinib,[10] an interesting drug that hits multiple kinases. What is your personal opinion on the data that we have seen so far?

Dr. Scher: I have a colleague in the pharmaceutical industry who had moved to Exelixis (South San Francisco, California), the sponsor of the program, who called me and said, "You need to see this." What was particularly interesting about the first observation of that effect (the almost complete resolution of changes on bone scan) was that it occurred in May, and within 6-8 months, there was a flurry of activity, which not only confirmed that the bone scan changes occur but that these changes were associated with very dramatic pain relief. We have used this compound now for about a year, and we have seen pain relief to a degree that I have not seen before; that alone was an impetus to suggest to the sponsor a dedicated pain trial, which is currently ongoing. The data you presented[11] showing that the drug can have similar benefits at a lower dose is obviously very important because there is a downside to the agent, particularly at higher doses. The field is doing the proper thing by taking an observation of a change in bone scan pain relief and doing the appropriate studies to show what that means in the phase 3 setting. That is the only way we are going to get definitive answers.

Dr. de Bono: We are very impressed with the pain relief, but we have also seen tumor necrosis on the multiparametric diffusion-weighted MRI, which convinces me that this drug is not only hitting the microenvironment (which it clearly is) but also tumor cells. This drug shows a lot of promise.

"There Are No Shortcuts"

We have also seen the lenalidomide MAINSAIL study[11] that Dan Petrylak presented at this meeting, which was a negative study. What did we learn from that study?

Dr. Scher: Virtually every taxane combination that has been tested has not been shown to be beneficial, and 2 were shown to be inferior. We reported on the ASCENT trial[12] with high-dose vitamin D, and some very important lessons can be learned from that trial. One is that there are no shortcuts, and it becomes imperative to have very strong data, particularly now that we have so many options. We need to make sure that before we embark on a large phase 3 study that there is a very strong scientific underpinning. That trial went from phase 1 to phase 3, and virtually all the strategies that have been used to modulate vascularity (with the exception of one which has not gone to phase 3) have not shown a benefit.

Dr. de Bono: You have worked a lot on biomarkers, and you have been a leader in the study of pain biomarkers. We have discussed imaging biomarkers. You know our data on this immune response on whole blood. What is your take on that?

Dr. Scher: The real lesson is that there is more to treating cancer than just slowing growth. We know that cancer adapts to different parts of the body. We have known that prostate cancer, in particular, is bone-tropic, and a unique interaction occurs within the bone, which is like a very nice neighborhood in which the cancer cell can survive. We have seen these effects with cabozantinib and dasatinib. We know that targeting bone stroma can be beneficial. We have seen it with the bisphosphonates and the radionucleotides, and you can see these benefits without necessarily seeing declines in prostate-specific antigen levels. Similarly with the immune modulators -- sipuleucel-T has shown a survival benefit, PROSTVAC-VF is now in phase 3, and ipilimumab is in phase 3, and we will be reporting those data shortly. We are learning that targeting what have been called the "hallmarks of cancer" can be beneficial to patients. What we haven't had is a way to understand a patient's immune competence. We have talked about it; we think about patients being healthier vs sicker. David Almo's observation was interesting: If you look at the mononuclear cells and do a global gene expression looking for genes that are associated with survival, they are predominantly immune predictors. We are starting to realize that immune competence, still to be determined, is an important factor in a patient's outcome, and this will ultimately be a factor that we have to control for in our clinical trials.

Dr. de Bono: Howard, it's such an exciting time for prostate cancer medicine. Thanks for all you do. Thank you for joining us for this edition of Medscape Oncology Insights. This is Johann de Bono reporting from ESMO 2012 in Vienna.