Yardley: Understanding How to Sequence Therapies Will Be Key to Treat Breast Cancer

Denise Yardley, MD, is a medical oncologist with Tennessee Oncology, who specializes in breast cancer but does work in all adult malignancies. She is also affiliated with TriStar Health. Much of her research has focused on HER2 breast cancer, and she has been involved with trials with pretreated patients whose cancer progresses or who don’t achieve a complete response. Through her work with Tennessee Oncology, Yardley has published recently about the use of next-generation sequencing (NGS) in community oncology clinics, identifying discrepancies between real-world NGS data and that seen in studies.

The American Journal of Managed Care^® (AJMC^®): Next-generation sequencing, or NGS, has greatly advanced our understanding of cancer in a variety of diseases. Could you tell us a little bit about the role of NGS in terms of breast cancer?

Yardley: The role of NGS in breast cancer is still evolving. All of us have been privy to the data that NGS can tell us about cancers, about growth pathways, about resistance pathways—and how to utilize this information for actual treatment decisions based on NGS identified molecular drivers of cancer progression. I think this remains a bit of uncharted territory for breast cancer. When we look at the specific alterations identified or look for specific alterations, breast cancer as a whole isn't the most highly mutated cancer, as compared with some of the other solid tumors. So, breast cancer is coming in behind some of the established groundwork of NGS profiling of other solid tumors. Nonetheless, we're still trying to gain greater knowledge about these specific alterations that are associated with individual cancers and within specific breast cancer subtypes. In addition to providing prognostic information, the value of NGS is to identify candidate alterations for specific mutation driven treatment. Right now, it remains very promising but not yet ready for clinical use.

Breast cancer itself is so heterogeneous in its subtypes—whether it’s hormone receptor (HR)–positive, HER2-positive, or triple negative—that when you then add in some of the identified molecular alterations superimposed on germline mutations, the door certainly opens to identification of additional niche subtypes. NGS platforms may determine a change in the tumor from its primary cancer features and this may render identification of a genetic or phenotypic abnormality that may be a target for an available drug or render the patient eligible for a clinical trial directed toward that alteration. I think it's of paramount importance that we keep trying to chip away the ice on NGS in breast cancer to distinguish how it’s going to be used—whether just for prognostic or predictive factors or ultimately for making treatment decisions. It’s still a crucial part of very active research for breast cancer; trying to understand these molecular drivers through NGS tumor profiling really becomes informative.

Then, there are established alterations—drivers of disease progression, drivers of metastases, and drivers of resistance. I think we're still in the infancy of being able to make widespread decision changes based on the identification of some of these alterations, and this is a cornerstone of active research. To date there is a lack of data to show this information can guide treatment decisions. The presence of many of these alterations help us do a lot of prediction modeling in breast cancer—telling us whether these tumors have specific mutations that are driving their growth. But in terms of making this information a clinical practicality, I think it's not quite prime time yet. I think we're getting there, but we still have lots and lots of work to still do.

On a basic level, we're able to identify our standbys for making decisions—the presence or absence of the estrogen and/or progesterone receptors and the HER2 receptors. Trying to understand different phenotypes in breast cancer that may potentially make a patient eligible for clinical trial is also a role for NGS right now. We are trying to collect that data to see if these patients that are screened by NGS and found to have a specific alteration are enrolling in clinical basket cancer trials, to understand if an intervention, based on that detection, is going to result in a meaningful clinical output.

There have been many basket trials—we have several at our institution and they are present across the country. It's a little way from just treating breast cancer, but these basket trials enable breast cancer patients to enroll in trials that offer a chance at molecularly targeted agents outside their given indications, so as to not restrict the trial and investigational drug to only a specific solid tumor type. Since the drug is targeting a specific molecular alteration and not the tumor type, it could be offered to tumors bearing the specific mutation. However, it must be said that there are very few instances in which researchers have demonstrated the ability of a particular targeted drug therapy approved to target an aberration identified in a specific solid tumor, is effective when administered to another tumor type bearing that same mutation such as breast cancer. Using these molecularly targeted agents outside their indications frankly just does not necessarily improve outcomes relative to standard of care. While these trials suggest that actionable mutations can be identified, there is little evidence that just because a targeted therapy works in one cancer, that it will work in the same manner in another cancer, even if the molecular target is present in both tumor types.

We learned from some of these basket cancer trials that there are a few mutations that demonstrate a little bit more cross response activity. TRK is one of those mutations that treated, demonstrates a signal of benefit in multiple tumor types. And so, the hope is that we continue to chip away at our knowledge base with NGS profiling and breast cancer to see which group of drugs targeting a mutation or alteration or why is it that treating a TRK fusion protein mutation in breast cancer may be efficacious, where as another agent targeting a different mutation is not going to be beneficial?

I think it's of paramount importance that we continue to serially biopsy our patients with metastatic disease and do the NGS tumor profiling partnered with enrolling these patients in clinical trials so we can glean more information from these treatment strategies. All of these identified molecular alterations have an opportunity to have that home run with a specific intervention—an intervention that they wouldn't otherwise get prescribed, particularly when there is availability of a targeted drugapproved for that tumor alteration in another disease type.

I think it's an extremely exciting time because we're still able to treat those patients and treat someone on clinical trials that gain that information. We learn which molecules do have cross-response activity in all these different tumor types and which ones don't, and further open the doors, exploring “What are their modifiers?” Are there 2 targeted drugs that, when given together based on the NGS profile, may overcome a lack of therapeutic benefit seen when one is just prescribed giving a monotherapy with a given drug? The future seems primed for us to continue exploring combinations strategies as well as monotherapy.

While NGS in breast cancer developed a little bit later than with some of our other solid tumors, the receptor data has been of paramount importance in further subdividing breast cancer and selecting therapeutic targets for the ER/PR-positives, the HER2-positives, and the triple negatives….These are the ones that we can say stand the test of time. To these we can add the identification of the germline BRCA mutations and the evolving role of the PARP inhibitors. This is not as clear as in the identification of HER2 positivity that has clear level one evidence of utilizing that data for clinical decision making.

We’re still looking to see how identification of the ESR1 mutations are being used to make treatment selections in the HR-positive population. ESR1 identification has great potential. There are certainly several drugs in development that overcomes this resistance pathway as well as strategies exploring combinations of drugs. We see endocrine therapy, with the CDK4/6 also able to overcome the resistance mediated by ESR1. And then, we are looking at other drugs in development that are just coming out of the gate that may overcome the resistance mediated by ESR1 mutation. So, I think we're evolving in that clinical pathway and arena, but there's still so much work left to be done.

AJMC^®: I'd like to ask about a recent publication involving a pooled analysis of the MONALEESA trials—numbers 2, 3, and 7—on ribociclib, specifically on safety, and its efficacy with some dose reductions. To start us off, can you tell us a little bit about the moderate and severe adverse events that are associated with the use of CDK4/6 inhibitors, ribociclib included?

Yardley: CDK4/6 inhibitors come with some class side effects. Their impact on cell cycle progression and proliferation results in sustained inhibition and cell cycle arrest in the G1-S cell cycle that also impacts other organ systems. This impacts on actively proliferating cells such as the bone marrow which can be a significant target with daily cell turnover. So, for palbociclib and ribociclib, the predominant impact of the CDK 4/6 mediated cell cycle inhibition on the bone marrow, results in leukopenia, more specifically neutropenia and to a lesser degree on red blood cells and megakaryocytes manifested as anemia and thrombocytopenia. A clear-cut monitoring of white blood cell count, specifically neutropenia is required due to cell cycle inhibition in the G1-S in the bone marrow, resulting in decreased counts in these patients. I think what's really interesting is that we're all very comfortable with managing hematologic toxicities from chemotherapy drugs, that sometimes this is a little unexpected from our oral therapies, which are often incorrectly equated little to no toxicities. But that impact on the bone marrow and in rendering patients with neutropenia is certainly part of these CDK4/6 inhibitors, but certainly seen to a lesser degree with abemaciclib, attributed to the potential greater CDK 4 inhibition than CDK 6. Nonetheless, this hematologic toxicity responds to dose reductions and dose interruptions.

So, for palbociclib and ribociclib, the schedule of administration is for days 1 to 21, followed by a week of rest which allows the bone marrow to recover. It's interesting that patients aren't very symptomatic from this neutropenia and infections are infrequent: this appears to be much more asymptomatic and what many refer to as “paper toxicity.” We have patients return for lab checks, but it's infrequent that they're really encumbered by the same degree of complicated neutropenia, that we see in chemotherapy patients. Overall, patients are experiencing fewer toxicities with the CDK4/6 inhibitors than they experience while receiving a chemotherapeutic drug that has a high potential for hematologic toxicity. Abemaciclib is a little bit unique in that there is a lesser incidence of hematologic toxicity.

I think the other issue is most of these patients—but not all—come with prior chemotherapy exposure. So, their bone marrow has already experienced marrow suppressive insults from prior treatment, which may enhance sensitivity to another potential marrow proliferation altering drug. So hematologic toxicity is one that I think we see across the board, although to a much lesser degree from abemaciclib. There is a great comfort level managing this toxicity are just based on the years of managing neutropenia from the chemotherapy arena. And it's interesting that this is one toxicity patients obviously do not routinely complain about because they're not symptomatic or aware. There is also some mediated gastrointestinal toxicity, something not completely unexpected for an agent that must be taken orally. There is some nausea and much less frequently vomiting, diarrhea, and decreased appetite with diarrhea more common with abemaciclib. All are easily managed medically with supportive care drugs and dose modifications.

What I take away from my patients is that fatigue can be a real factor with the CDK4/6 inhibitors, and that's probably the most challenging side effect to manage. Dose reductions can have an impact, but I find this to be of less of a predictable benefit and may translate to no perceived real impact on improving fatigue. I find that dose reductions for fatigue are significantly less reliable than my algorithm for dose reductions to manage hematologic toxicity. So, improvement of fatigue is not that clear cut and not such a guarantee with dose modifications. Often, when I have a patient facing significant complaints of fatigue, I'll just give them a drug interruption or holiday and see if I can rescue them. They must manifest some degree of recovery before retreating with that specific drug, albeit often with an instituted dose reduction. But that experience of fatigue is challenging, because we don't have a lot of drugs that we can effectively prescribe as an intervention for the patient. We all should step back and look at other mediators of fatigue such as poor sleep or sleep quality due to the anti-estrogen effects and vasomotor instability that comes with the endocrine therapy. So, that also tends to often be overlooked, as a contributor to fatigue, when we're prescribing our CDK4/6 inhibitors in combination with endocrine therapy. A patient may have had challenging sleep cycles anyway, and now we've added an anti-estrogen and a CDK4/6 inhibitor, and it's really brought that to the forefront and manifested as daytime fatigue when it’s really more a reflection of poor sleep and associated and subsequent poor tolerance.

We try to have our patients stay as active as possible. Every once in a while, I've prescribed a stimulant, but it has been disappointing and not a clear true fix for that situation. But occasionally, I have a patient that seems to benefit from it. I always tell patients, it either works, or it doesn't, to try to combat drug induced fatigue.

From a patient standpoint, I know my patients are happy to not be on a chemotherapeutic agent as many of them have had that prior experience. I think there's a bias of equating oral therapy with non-toxic therapy, but we are trying to make sure patients are very clear that you can generate just as high a level of toxicity with oral targeted agents, as you can with intravenous chemotherapy. We must look at managing toxicities and balance this with quality of life, so my goal is to manage the potential for gastrointestinal toxicities with prophylactic supportive care drugs for diarrhea. The GI tract does acclimate, and it is important to make every effort to support the patient through this time period to be able to gain the median PFS improvement that we expect; we can typically manage the toxicities effectively with reducing the dose and introducing supportive care meds and give it our best shot….Some patients just frankly abandon it from the beginning; 40% require some dose reduction.

I always reassure patients about the importance of instituting dose reductions because I think patients have the mindset that they want to stick with the initial prescribed dose and tough out the toxicities in order to preserve the potential of benefit. It is important to inform patients that instituted dose reductions were part of the clinical and used to preserve the ability of patients to comply with prescribed therapy. You can say wholeheartedly with data generated from the MONALEESA trials, that when they looked at those patients that stayed on study treatment and looked at dose intensity, the 40% that had dose reductions, their benefit and median PFS were not impacted by those instituted dose reductions.

We do not want patients pushing their drug past toxicity—there’s not a reason to do that; there's still benefit with lower doses. It all goes back to how clinical trials and drug doses were developed….The study designs had goals of trying to establish the maximum tolerated drug dose and then taking it one step down….So, when we looked at the MONALEESA trials, those patients that had underwent mandated dose reductions due to toxicity, they achieved the same disease control benefits than patients who did not undergo dose reductions.

AJMC^®: Speaking on the HR-positive, HER2-negative patients with advanced or metastatic breast cancer, how many of these patients will receive a CDK4/6 inhibitor and then still develop disease progression? What studies do you think are needed or potentially upcoming to address this progression on CDK4/6 inhibitors?

Yardley: It’s that window of opportunity—how do we exploit the CDK4/6 inhibitors? We've all been enthralled with that median progression-free survival benefits as well as overall survival; it's obviously higher in the first-line setting, but still present in pretreated and later line therapy patients. I think all of us have a natural tendency to want to continue with a strategy and a backbone class of drug that has benefited the patient. There are studies that are looking at following a CDK4/6 inhibitor with another or different CDK4/6 inhibitor. I think palbociclib and ribociclib bear more similarities than abemaciclib, so if this is the initial CDK4/6 inhibitor, a switch to abemaciclib could be considered due to its unique structure, the differences in CDK4 inhibition, the dosing schedule having continuous exposure and on a daily basis, twice a day. Those studies are quite informative; we're learning can we partner it cautiously with some other targeted agents—there’s a triplet trial (TRINITI) with ribociclib, an endocrine therapy and the mTOR inhibitor, everolimus, that did show benefit in patient’s tumors that were previously exposed to a CDK4/6 inhibitor. By blocking at least one pathway of resistance with the mTOR inhibitor, prolonged disease stability and some disease regression was evident. This is certainly provocative as further blocking the hormone signaling pathway in breast cancer, resulted in extending the benefit of the CDK4/6 inhibitors. Explorations continue to identify and explore the drivers of that disease resistance. And does that driver mean no more CDK4/6 challenges, or do we switch out one CDK4/6 for a different CDK? There are additional CDK agents still in development—there’s a CDK2, CDK6, etc. Whether we'll be able to build that platform and line of CDK inhibitor therapy remains to be determined.

I look at multiple pathways to exploit in in those HR-positive patients….There’s testing for the PIK3CA mutations, for the small percentage, 30%, that may be eligible for consideration of the alpha pan inhibitor of the AKT-PI3 kinase pathway with alpelisib, currently only approved in combination with fulvestrant. So, you’re trying to build that deck of cards to throw at the HR-positive tumors. In doing so, hold out on fulvestrant, until the tumor has been assessed for a PI3 kinase mutation since that's the only endocrine agent partner approved for alpelisib, for HR+ HER2- tumors bearing a PIK3CA mutation. I prefer the aromatase inhibitors therapies, or tamoxifen, as partners with CDK4/6 inhibitors except for the combination of tamoxifen and ribociclib due to an interaction. We’re trying to figure out how to best sequence combination endocrine treatment strategies. What happens if you determine that there is a BRCA germline mutation in a patient with a HR-positive tumor? Where do you sequence your PARP inhibitor into the treatment algorithm? These are challenges we all face in making treatment decisions. We don't have a study that can enroll a patient through a varying treatment sequence and take them through every disease progression. It’s talked about but it's just not feasible. But you know, when I look at my germline BRCA mutation patients, and I look at the potential of sequencing a PARP or a CDK4/6 inhibitor, I go back to you trying to stay focused on data as the driver of those decisions. For ribociclib and abemaciclib, we have overall survival advantages in the metastatic setting and so that influences my decisions to prescribe my CDK4/6 inhibitor first-line and offer the PARP agent second-line or a PI3 kinase agent or even consider rolling out everolimus later line as an mTOR inhibitor. I think we're just more challenged now, but it’s a good challenge to have—sequencing a handful of targeted drivers that impact on that endocrine signaling pathway.

AJMC^®: If you could briefly tell us about the role of the androgen receptor in advanced breast cancer, I understand you're involved in in a phase 2 trial. Do you think this this receptor is going to be a key to managing breast cancer in the future?

Yardley: I tell you, there's been a lot of great effort for many years channeled on how to target the androgen receptor pathway in breast cancer. It’s prevalent in the HR-positive tumors as well as triple negative breast tumors and present in metastatic breast cancer. I think the challenge has been to try to figure out how to modulate estrogen and androgen mediated cell proliferation and in which breast cancer subgroup. The HR-positive/androgen receptor–positive subgroup is different than the triple negative androgen receptor–positive subgroup and then there are the differing aspects of the androgen receptor signaling pathways of the different agents used in the clinical trials. I think, even going back to the premise of how do we establish the criteria for defining androgen receptor positivity, using immunohistochemical staining? What's the threshold for positivity?

I was participating in a trial with enzalutamide evaluating it in metastatic breast cancer. It is an agent that targets the androgen receptor. In that trial, androgen receptor positivity used a threshold of 1% or greater as being considered androgen receptor positive. There was a companion diagnostic as a part of that trial with enzalutamide, and it did show about a 33% response rate in patients that were identified using the companion diagnostic, but it didn't meet the criteria to continue the exploration of enzalutamide in androgen receptor positive breast cancer. Another strategy in inhibition of the androgen synthesis is with an agent such as orteronel, which I recently published a paper on. It’s a17, 20 lyase, reversible hydroxylase inhibitor, a key enzyme in the androgen synthesis pathway that also impacts on estrogen synthesis.

So, we looked at measuring estradiol and the testosterone and could show a decline. But again, it’s belabored by the difficulty of identifying which assay is most accurate in measuring the hormones and with the parameters of trying to make a conclusion based on the peculiarities of the assays. Orteronel’s mechanism of action differed from the enzalutamide, but it did not show a clinical benefit that was meaningful enough to move forward with that drug. I think there’s still a lot of intrigue with targeting the androgen receptor pathway because those androgen receptor positive tumors, particularly in the triple negative population, are typically of a lower grade, more indolent, and less proliferative, a biology common to low grade HR–positive breast cancer tumors. Identifying this subpopulation tumor biology within the triple negative breast cancer population that might be sensitive to modulation with an anti-androgen agent would be meaningful because all these patients are currently treated with more toxic sequential chemotherapy or chemotherapy/immunotherapy strategies. If we can pull them out of this treatment algorithm, identify them confidently with validated criteria for establishing what we call androgen receptor positivity, and pair that with an efficacious drug targeting the androgen receptor signaling pathway, this would be a game changer. To date, most of these agents have shown various signals of activity but nothing worthy of moving forward.

For a randomized phase 3 clinical trial, I think targeting the insulin growth factor receptor pathway is still promising. There have been a variety of drugs in clinical development targeting different aspects of the insulin growth receptor pathway. The defined science is great on that pathway—we understand it, its presence and impact on the estrogen signaling pathway as well as cell proliferation, but we still lack that therapeutic intervention that will make the difference. So that downstream effect of targeting this pathway remains to be determined. We know for the HR-positive and androgen receptor–positive tumors, those pathways intersect and cross, but we still aren’t able to establish that pivotal inhibition that’s going to result in a significant enough benefit. To date, they’ve all kind of come up short in identifying a validated signal of activity in a huge pool of patients. So, trying to pick out those patients and with what aspect of interrupting the signaling pathway is most likely to yield benefit, is where the challenge remains.

AJMC^®: Do you have any closing thoughts on where the field of advanced breast cancer will be in the next 2 to 5 years?

Yardley: I think we’re at a point of really trying to figure out 2 avenues. One is now learning how to sequence these drugs appropriately. There are many trials, so there are strategies of doing doublets and triplets and even monotherapies. In trying to figure out how to best sequence these agents, drivers of tumor growth and progression, how we can extend the benefit of an agent, like the CDK4/6 inhibitors, with strategies of can we switch the endocrine partner and continue the CDK4/6, or add another modifier that may extend that benefit from the CDK4/6? This is an exciting platform, for at least the HR-positive population but explorations in the HR-positive, HER2+ disease are underway.

The second avenue is that now we’re cognizant of being able to improve on the toxicities of those initial agents on the market, yet we must capitalize on enhancing benefit. We’re moving more drugs into the early stage arena, but more drugs that are patient friendly, and more efficacious for patients. Yet the challenge really remains in the central nervous system (CNS) metastasis especially for both triple negative breast cancer and the HER2 positive tumors, and so developing more drugs that can effectively treat systemic disease but also treat CNS metastases. Patients have recurrent CNS events and are limited by the amount of radiation to be delivered partnered with an extremely limited arsenal of drugs that actually penetrate the CNS. And it’s truly still an unmet need for those patients; it’s so disabling. I have 2 patients right now with CNS metastases. We’ve had success controlling their systemic disease, totally quite quiescent, but their quality of daily life now is limited by the development of CNS metastases and the impact and toll this takes on the activities of daily living. The number of agents that may intervene meaningfully are only a handful, with limited success and benefit, in addition to radiation. I think just understanding that challenge, and really trying to find a work around remains challenging Is there a way that we could in the future even go as far as prevent, or do some type of prophylaxis for those patients? It’s just so devastating that the trigger usually is a largely irreversible neurologic deficit that brings it to our attention. And we know the recovery of that is just so limited for those patient populations.

As we can continue to define the challenges, I think we will continue to make tremendous progress in treating patients with breast cancer. Hopefully, knowledge obtained by the evidence of benefit of many agents in the advanced disease setting will continue to foster the roll out of these agents such as abemaciclib into the early adjuvant and neoadjuvant disease setting where it can potentially demonstrate benefit in a very high-risk group of patients. We are hoping to continue to catapult drugs known to have a significant role in metastatic breast cancer into earlier lines of therapy, including both neoadjuvant and adjuvant and in this disease setting, reduce the number of patients burdened by the development of metastatic disease.