A Look Into the Future of Breast Cancer: A Conversation With Joyce A. O'Shaughnessy, MD

The 10th International Congress on the Future of Breast Cancer was held in Coronado, California, August 4-7, 2011. The 4-day, interactive program was developed and chaired by leading breast cancer researcher Joyce A. O’Shaughnessy, MD, a medical oncologist with Texas Oncology-Baylor Charles A. Sammons Cancer Center in Dallas.

Our editors had a unique opportunity to sit down with Dr O’Shaughnessy to discuss the goals of the meeting—namely, to increase our understanding of the biologic subtypes of breast cancer—and to get her perspective on how the hypotheses and data are evolving in this dynamic field.

OBTN: This 2011 meeting marks the tenth year of The International Congress on the Future of Breast Cancer. The 10th anniversary is typically associated with tin or aluminum, signifying pliability and durability—the ability to bend, but not break. Looking back on the past 10 years, there have certainly been a few bends in the road. Can you talk a little about how the field of breast cancer research—and this meeting specifically—has evolved over the past decade?

Dr O’Shaughnessy: As I remember back to the very first meeting, I think the main thing I recall is that Carlos L. Arteaga, MD, gave a very good talk about the role of the HER (ERBB) family in breast cancer and the interconnectedness between the HER receptors and downstream proteins. There was an incredible sense of excitement at that time about us being able to start to tease out resistance mechanisms to therapy and really have a handle on what some of the deadly pathways were for patients who had aggressive breast cancer.

That’s the kind of information that I try to bring to this meeting—the cutting edge, new mechanistic, pathogenetic mechanisms—because that’s what physicians are so eager to learn about. This area has just grown over the years as we’ve become increasingly mechanistic in our understanding of the various subtypes of breast cancer. So, this meeting focuses primarily on the various biologic subtypes of breast cancer; our understanding of these subtypes has increased significantly since we first started meeting 10 years ago.

The meeting draws an international and an interdisciplinary audience. How does this year’s program manage to address this level of diversity in the attendees?

We really focus more on what all breast cancer practitioners need around the world: a better biologic understanding of the disease, better diagnostics, and better therapeutics. So, we don’t focus on the clinical needs of the international audience, per se. The informational needs in breast cancer are universal, that is to really understand this disease and make better therapeutics. We have experts from around the world who are as interested in breast cancer as people in the United States, and so they come to this meeting because it’s a way to get a focused update on all the important clinical information, and to see how the hypotheses and data are evolving with regard to the patients they’re taking care of right now.

You participated in the FDA hearings on Avastin. In a nutshell, the real issue seemed to be that the drug showed improvement in progression-free survival but was unable to demonstrate a benefit in overall survival. In your opinion, what is the most meaningful endpoint to demonstrate an unequivocal benefit to the patient? How much improvement/benefit is enough? What should be the standard going forward?

The front-line Avastin trials in metastatic breast cancer—3 large, randomized trials—all met their primary endpoint of progression-free survival. They did not meet the endpoint of overall survival, but that wasn’t their primary endpoint. Nor were these trials really powered for survival in terms of picking up small benefits. So, I think that the primary endpoint for approval for Avastin and other first-line therapies should be progression-free survival.

I think that if you have a statistically significant finding and you meet your primary endpoint, the drug should be approved. Then it’s up to the oncology community to decide whether the magnitude of benefit is large enough to warrant the risks versus the benefits. I think Avastin in first-line therapy for metastatic breast cancer is very safe, as physicians are now accustomed to avoiding Avastin use in patients who have the possibility of complications (ie, patients with recent surgery, diverticulitis, any kind of active wound in the body, or with baseline uncontrolled hypertension).

“It’s all hands on deck in the worldwide race to understand this disease. The Cancer Genome Atlas Project for breast cancer will be completed by the end of this year. So, we will soon have a full catalog of all of the mutations in breast cancer by subtype.”

Avastin remains my standard of care for aggressive metastatic breast cancer that requires the use of 2 agents for optimal control of the disease. The choice is using either 2 chemotherapy agents or using Taxol with Avastin. The Taxol/Avastin combination allows you to continue the treatment for metastatic breast cancer without having to decrease the dose or stop one of the drugs. So, you’re able to continue with doublet therapy for a longer period of time to improve progression-free survival, as opposed to using a doublet of chemotherapy that you have to reduce the dose or stop one agent over time, thereby losing the effectiveness of the 2 agents in combination. I believe Taxol/Avastin is an excellent treatment for patients with aggressive breast cancer.

In the triple-negative subset of patients—many of whom are quite refractory to chemotherapy— where we have fewer options, I especially think that this is the area where Avastin fills its greatest unmet medical need.

It must have been pretty powerful and emotional to hear all the patient testimony at the hearings. How do you reconcile the widely varying opinions on the contributions of Avastin in breast cancer treatment? How do you anticipate this will net out from the varying stakeholders’ perspectives: the FDA, the patient, the insurer, and the community oncologist?

I don’t know. To tell you the truth, I don’t think a lot about that. I really think about what I need to do in my practice to take care of patients. I think that the good news is that after the FDA’s Notice of Opportunity for Hearing, Medicare said that it would continue to pay for the drug. That was fabulous news. All I want is access. I feel strongly that patients and physicians should continue to have access to this agent. Because Avastin is approved for so many other cancers and approved in Europe for 2 indications as first-line therapy with Taxol and with capecitabine, there’s plenty of information out there to guide the practicing doctor.

I firmly believe in the confirmatory trial that will look at VEGF-A levels as a predictive biomarker. I think that’s very important. But I think we have to focus on having the few agents like Avastin that are non-cross-resistant with chemotherapy available for patients who have very limited options.

We hear a lot about personalized medicine— finding specific subsets of patients who respond to particular drugs. Consider the differences between Herceptin and Avastin. Herceptin uses HER2 as a prognostic marker; Avastin came to market without a biomarker signature, but some suggest perhaps there is one that would explain why some people seem to be “super responders.” Should all new products be required to have this type of biomarker signature?

I don’t think so. I think it’s very difficult to define these signatures given the complexity of the cell. I think the recent information on pertuzumab in the CLEOPATRA (CLinical Evaluation Of Pertuzumab And TRAstuzumab) trial is a good example. This trial evaluated taxane/Herceptin with or without the antibody pertuzumab also against HER2. Although I’ve not yet seen the data, I’ve heard that it’s a very positive trial. Even so, we still don’t know which of the HER2-positive cancers will benefit the most from pertuzumab. It may take us years to figure that out. Meanwhile, there’s apparently a very nice incremental benefit. I think going into the adjuvant setting with a mixed population of triple-negatives and indolent ER-positive breast cancer is a thing of the past, and we really can’t be doing that kind of “all-comer” trials any longer.

Even though I think we need to do some selection, I don’t think it will be feasible to come up with the “ALK story” for each and every drug we’re developing. Because as we’ve heard from this meeting, a lot of what determines sensitivity or resistance is systems biology. We may not be able to really know all of the operative pathways and microenvironment issues, but at least if we can get some understanding of the biology of these tumors and—importantly—not enroll subsets of patients in clinical trials who we think have very little possibility of benefit. That’s a way of selecting too—rather than a positive selection, a negative selection by not enrolling those who may not benefit. I think that this is something we will strive to do, but I think it will be very challenging to identify very narrow molecular subsets. There will be some successes, no doubt, but I think for the most part it’s going to come down to more broadly defined biological subsets.

In light of the data you presented at ASCO on iniparib, PARP inhibitors may not be the magic bullet people were expecting. What can we learn from these negative results? Where do you see the class of PARP inhibitors going?

Well, it turns out that iniparib is not a PARP inhibitor at the physiologic concentrations that are achieved in women. The mechanism of action is under intensive study. Of course, there had been a lot of work done on this mechanism from the time it was first discovered, and I’ve been told about some of the new hypotheses coming out, and they’re very interesting. So, I think that the iniparib story will be interesting once again, in terms of its mechanism of action at physiologic concentrations. If you give enough of it, you do inhibit PARP-1. But the levels that are achieved with a dose of 5.6 mg per kilogram, you’re not inhibiting PARP-1. That’s really clear.

In terms of crossover in the setting of a trial that has survival as an endpoint, I think we’re back to being more cautious about that. With the phase II results being so robustly positive, even with a crossover, I think people felt that it was reasonable to go ahead with a crossover in the phase III study. With survival as a primary endpoint, you can muddy the waters with crossover.

Sanofi has gone back to phase I to continue to increase the dose of iniparib because the drug has such a low toxicity profile. It may be that they will get to a point where they’re able to inhibit PARP-1 with higher doses, but at the current physiologic concentrations, the drug is biologically active. It does cause DNA damage, and there is a case of an excellent response in BRCA2—mutated pancreatic cancer. We need to better understand its mechanism of action, and to parse triple-negative breast cancers into biologic subsets because there’s a signal of activity in the phase III trial.

Although the iniparib phase III trial didn’t meet its primary endpoint, this negative trial tells us that focusing on treatments for triplenegative breast cancers is still too broad. We need to get more granular in our focus with DNA-damaging agents like iniparib regarding which triple-negative patients benefit. So, in terms of lessons learned, I think going forward we need to better understand the biologic subsets of triple-negative breast cancer and to select subtypes of triple-negative breast cancer for future trials.

“

I also think that doing serial biopsies of metastatic disease— real-time monitoring of disease—will be important over the next 5 years. For example, monitoring circulating tumor cell changes over time and redoing biopsies at multiple time points in metastatic disease for patients with longer natural histories will become increasingly important.”

Moving forward with research on iniparib, how will future studies deal with the issue of heterogeneity in triple-negative breast cancer tumors? Would testing for BRCA mutation status be required?

Because iniparib is not a PARP inhibitor, the BRCA story becomes less relevant. One could consider doing a study of basal breast cancers, for example, or basal triple-negatives versus non-basal triple-negatives if your hypothesis was about DNA-damaging agents and DNA repair. However, I think that what we’ll find is that investigators will be looking for specific mutations such as loss of PTEN or activating mutations in phosphoinositide-3 kinase (PI3K) to determine which patients to enroll in studies. For PI3K pathway inhibitors, investigators will be looking for mutations that lead to RAS/ MAP kinase activation, for example. So, I think there will increasingly be selection for pathway activation and microenvironment cytokines. I think what we’re going to see are many smaller, randomized phase II hypothesis-generating studies over the next 5 years.

How do you see the landscape of breast cancer treatment changing over the next 10 years? What trends in breast cancer research seem really exciting to you? Are there new areas that are just beginning to be explored that sound very promising even in the early stages?

In one of the presentations at the meeting, Matthew Ellis, MD, PhD, described the concept of using “genome-forward” approaches to patient selection, as opposed to doing studies where we collect tissue and then retrospectively look at the molecular characteristics of the patients who responded versus not. I think that this is an exciting concept that we can apply in areas where we need better patient selection tools. However, with regard to some of these targeted therapeutics, we do need to make our very best hypotheses about which patients are likely to benefit and build that into the trial. You may also evaluate and treat a broader population, but you have to have a prospective hypothesis within your proof-ofconcept trial looking at particular subsets. I think the genome-forward concept is one of the things we will be focusing on in the future.

I also think that doing serial biopsies of metastatic disease—real-time monitoring of disease—will be important over the next 5 years. For example, monitoring circulating tumor cell changes over time and redoing biopsies at multiple time points in metastatic disease for patients with longer natural histories will become increasingly important.

Exclusive interview with

Dr Joyce A. O’Shaughnessy

I think that the future is very bright for HER2-positive disease, with all the new agents available, and I think that there will be a deeper understanding of the subtypes of HER2-positive disease, possibly through the evaluation of phosphoprotein signaling in the cell.

For ER-positive disease, I think we will gain a better understanding of the biology—the molecular pathogenesis—with regard to which of the indolent cancers just can’t die. Patients with these cancers may need prolonged endocrine therapy. We’ll have a better understanding of the molecular abnormalities of the aggressive ER-positive tumors, in terms of which amplicons and deletions are driving the cancers, and where the dominant pathway abnormalities are. The BOLERO-2 trial with everolimus is a good step in this direction. We also have the entinostat data that were mentioned at this meeting that may be very promising for the more aggressive ER-positive cancers.

The triple-negative breast cancers remain our greatest challenge because we still have a relative dearth of sound, clinically testable hypotheses at this time. I think we’re going to see some breakthroughs soon with excellent hypotheses emerging, so I am optimistic. It’s all hands on deck in the worldwide race to understand this disease. The Cancer Genome Atlas Project for breast cancer will be completed by the end of this year. So, we will soon have a full catalog of all of the mutations in breast cancer by subtype. I think these data will be very important in subtyping triple-negative disease and will spur a lot of hypothesisgenerating studies that will ultimately lead to some breakthrough therapies.