SABCS: Using Molecular Assays for Breast Cancer in the Clinic

Antonio Wolff, MD

Ahead of the annual San Antonio Breast Cancer Symposium held December 4–8 in San Antonio, Texas, we discuss the evolving role of molecular profiling in the diagnosis and treatment of breast cancer. We speak with Dr. Antonio Wolff, professor of oncology at the Kimmel Cancer Center at Johns Hopkins University, a breast cancer clinician and recently appointed associate editor of the Journal of Clinical Oncology. Dr. Wolff will be speaking during the "Practical Use of Molecular Profiling" session at this year's symposium.

-Interviewed by Anna Azvolinsky, PhD

Cancer Network: Dr. Wolff, as a practicing medical oncologist, what do you see as the most patient outcome-changing genetic or molecular test that is currently being used in practice?

Dr. Wolff: Thank you for the opportunity to speak to you today. It is actually quite interesting for those of us who take care of patients to realize how important attention to proper collection of tissue and a better understanding of even existing assays. We are beginning to see several molecular assays coming into the clinic. Some of them have already undergone some evidence of measure of prospective validation and others are still in development. It is quite interesting to see that many of them, especially those that are currently being used to determine whether a patient should be treated with adjuvant chemotherapy or not, especially patients who have estrogen-receptor (ER)-positive disease.

Many of these assays in a way are a reflection of proliferation that measure, in essence, grade and differentiation. We have known for a long time that patients with so-called high-grade tumors are patients who may have high-risk tumors and may be at higher-risk for recurrence, and I think we are now integrating that information with measures of estrogen-receptor status. Consequently, we have seen several assays which are now commercially available but not very well prospectively validated in terms of the clinical utility. But maybe these assays now can be used to identify which patients could potentially be treated or should be treated with adjuvant chemotherapy or not, in addition to anti-estrogen therapy for those with ER-positive disease.

Cancer Network: Could you go over the types of tests that are currently available for both diagnosis and to facilitate treatment decisions for breast cancer patients?

Dr. Wolff: One of the assays that is most commonly used today is the so-called 21-gene assay or Oncotype Dx. This is an assay that was prospectively developed and validated using retrospective samples from previous clinical trials. This one example appears to identify a subset of patients with ER-positive disease that would potentially benefit from the addition of chemotherapy to anti-estrogen, and who without the use of chemotherapy would have a greater risk of disease recurrence after 10 years. This is the assay that is most developed at this point.

There is another commercial assay right now called MammaPrint, which has undergone validation of its prognostic utility in terms of identifying among patients with ER-positive disease those who are at greater risk of recurrence. However, it is an assay that is still at this point undergoing clinical utility validation in terms of its ability to identify patients who will have a better outcome if a treatment decision is made on the basis of this specific assay. These are the two assays that are the most developed at this point in the United States.

However, one major limitation of many of these assays is that for the most part, they appear to provide in different ways, a measure of tumor proliferation, of the tumor grade. It appears that their utility seems to be limited, at this point, to the two-thirds of patients who are newly diagnosed who present with ER-positive disease. These assays at this point do not have any utility for patients with triple-negative disease or HER2-positive disease.

Cancer Network: Where is the field right now in terms of broader genomic profiling of patients' tumors? Is this still an analysis still largely confined to academic research centers and studies?

Dr. Wolff: I think it is fair at this point to say that these assays are considered to be investigational, and they are not yet ready for prime time. We are beginning to see a series of papers coming out from various groups such as the International Cancer Genome Consortium and also in the United States, the Cancer Genome Atlas, showing on multiple levels a better understanding of so-called "omics" of the tumor in terms of the transcriptome, in terms of the protein expression, in terms of DNA pathways, mutations, copy-number variations. We are beginning to understand that actually, breast cancer is a very complex disease. We do have some difficulty in applying some of this knowledge in the clinic today.

Today in the clinic, the assays we have are essentially the more traditional immunohistochemistry ones, and in the case of HER2, also in situ hybridization. But essentially, what we have are routine pathology measures, such as measure of expression of the estrogen or progesterone receptor, measure of expression or amplification of HER2 and measures of proliferation such as tumor grade, an assay for which we still have a lot of difficulty. For the most part, they provide us with information on the various phenotypes and allow clinicians in practice to make clinical decisions. At the same time, we do understand that many of the assays have a very important negative predictive value. For instance, tumors that are ER-negative, these are patients that we will not recommend adjuvant endocrine therapy, the same way that for patients who do not have HER2 tumors, we will not recommend HER2-therapy. But on the other hand, these assays are imperfect, at least the current phenotypes we have are imperfect from a positive predictive value standpoint in the sense that not all patients with ER-positive disease will benefit from anti-estrogen therapy, not all patients with HER-2 positive disease will benefit from anti-HER2 therapy. This is even more critical for the small number of patients, about 15% or so, with so-called triple-negative disease. We now understand that triple-negative disease is not just one type of phenotype but an umbrella within a larger umbrella of breast cancer in which we may have patients that have what is now described as the basal-like phenotype which is derived from DNA expression studies. We have tumors that have more of a BRCA1 mutation or "BRCAness" to them. We have tumors that appear to have even improved prognosis because not all patients with triple-negative disease have a high risk of recurrence. Some of these tumors may have a more immune profile. So all of that information is beginning to be dissected in laboratory studies where we are trying to describe the "omics" of all of these subtypes and generate information that at some point in the future will be ready for prime time, but we are not there yet.

Cancer Network: Do you have any advice for both academic and community oncologists on how to best use molecular testing to guide their diagnosis and treatment decisions now?

Dr. Wolff: I think this is a critical question. There are a few things that can be done from a more basic standpoint. What I think we have learned in the last couple of years is actually the critical nature of proper collection of tissue specimens. Even for assays which we thought were simple such as ER and HER2, it is critical that the time it takes from the moment the tissue is collected from the patient to the moment that it is properly handed and delivered to the pathology lab, this is the time we call cold ischemia where tissue necrosis can happen, it is important that that is kept to a minimum to make sure that the integrity of the tissue and all of its molecular processes are preserved, whether it is going to be for routine testing for ER or HER2 or whether it is going to be for future submission of these tissues for assays like the 21-gene assay that we discussed. So that information alone becomes critical for treatment decisions. I think we all should be, and must be very excited about the new studies that are taking place. I think we should, if we have the opportunity whether you are at an academic center or participating in clinical trials in the community, if you have access to some of these studies, I think it is incredibly important to join studies that are beginning to try to better understand the molecular profiling of tumors and also beginning to help clinicians to make decisions. But we are not at that point yet, where we can profile an individual tumor toward what we will call, in the future, precision medicine.

We are getting to a point where we are moving away from a more traditional qualitative, descriptive diagnosis that in the past has served us mostly for diagnostic purposes. We are gradually moving into a future where we are beginning to understand the role of molecular observations that can allow us to do a more quantitative, a more predictive assessment of what is actually the future risk of cancer recurrence for someone diagnosed with early-stage disease, and most important, if we make a treatment decision guided by these findings, whether we will, ultimately, improve patient outcome, which is the true meaning of clinical utility. So we are not there yet. But it has been impressive to see what has happened in the last 15 years, which truly builds on the investigations that have started years ago and are taking us to a moment where we can begin to individualize treatment decisions, not just for patient populations or groups of patients or patients participating in clinical trials, but increasingly for patients that we see in the clinical on a day-to-day basis.

Cancer Network: Thank you so much for joining us, Dr. Wolff.

Dr. Wolff: My pleasure. Thank you!