Inside the GENOME

A look at breast cancer gene expression assays with Dr. Adam Brufsky

July 12, 2021 Myriad Oncology Season 1 Episode 22
Inside the GENOME
A look at breast cancer gene expression assays with Dr. Adam Brufsky
Show Notes Transcript

In this episode, Dr. Slavin welcomes medical oncologist and researcher, Adam Brufsky, MD, PhD, FACP.  He is a professor of medicine at the University of Pittsburgh School of Medicine and co-director of the Comprehensive Breast Cancer Center. Dr. Brufsky has been an innovator involved with breast cancer research for many years. Together, they discuss breast cancer gene expression assays and the role it plays in treating breast cancer. 



0:00:11.2 Thomas Slavin: I am Dr. Thomas Slavin, Senior Vice President of Medical Affairs for Myriad Oncology. Welcome to the Inside the GENOME. Hi everyone, and welcome back to the podcast. Today, I have Dr. Adam Brufsky. He is a professor at UPMC, and I've invited him here today to talk about gene expression assays for breast cancer. And if you have no idea what I am talking about, you hopefully will by the end of the podcast today. So thank you so much for coming on, Dr. Brufsky. I'd love for you just to tell the audience a little bit about yourself and what you do in medicine, your expertise as a clinician and a little bit about your practice.


0:00:51.1 Adam Brufsky: Sure. So I'm a Professor of Medicine and a Co-Director of the Comprehensive Breast Cancer Center. I'm a medical oncologist. And for many, many years since really my earliest training, I was always involved in how human genetics and clinical medicine interact. And actually a very, very, very long time ago, I cloned the first mutation, splicing mutations in osteogenesis imperfecta. You guys have a genetics audience, I might as well tell you this stuff. And then also as a post-doc, I cloned the first mutation in [0:01:18.7] ____, the first human mutation.


0:01:20.7 TS: Wow.


0:01:21.4 AB: And so after training at Harvard, I came here about almost 25 years ago, 24-25 years ago. And initially, it was to do a clinical research on prostate cancer. And what happened was they needed someone to do breast, and I decided to start doing breast cancer. So I build a big breast program and do a lot of clinical research and also basically kind of translational research. I'm very interested obviously in molecular assays and the applications of human genetics, of RNA profiling, other molecular features both in the tumor, as well as in the tumor micro-environment, the kind of like understanding breast cancer and developing new therapies. It's kind of what I do.


0:02:00.1 TS: Yeah.


0:02:00.9 AB: So I've had a long history of doing that.


0:02:01.8 TS: Yeah, that's great. And what does your average week look like from a patient perspective? 


0:02:06.0 AB: So I have two days... I usually... Well, it used to be one day, but it's anywhere... It's two days now, where I basically have a dry lab stuff that I do, I have a few post-docs and fellows that I work with to try to analyze gene sets and develop new assays. And during those days, obviously, I also have a lot of administrative stuff that I do. For about three days a week, I'm in clinic. My clinic is almost exclusively breast cancer of all stages, from Stage DCIS, which is the earliest stage of breast cancer, all the way up through advanced metastatic disease. That's generally how my week goes. So usually, Wednesday, Thursday, Friday, I'm in clinic.


0:02:42.8 TS: Yeah, you sound like a busy, busy bee.


0:02:45.8 AB: It is, very.


0:02:46.6 TS: I know you because we have shared interest in genomic profile assays, and so there's this whole field that looks at when women get breast cancer, looks at the actual breast cancer itself, at the molecular features of the breast cancer. And there's many different types of assays that are trying to help understand what is this particular woman's risk for a recurrence of that particular breast cancer, and also what is the benefit of chemo prevention. So can you use a chemotherapy to prevent the recurrence? And we don't wanna unnecessarily give chemotherapy, and so a big portion of this is if we provide a treatment, will it be effective? 


0:03:31.2 TS: And what we hope that these breast prognostic assays do, to a large extent, is help understand if a woman gets chemotherapy, what kind of benefit would she be expected to get from that chemotherapy, in addition to a main question, which is, yeah, does this woman need chemotherapy or not? And then if she doesn't get chemotherapy, what would we expect this particular patient's course to be? And so, these assays have been now available for some time. They're becoming increasingly common and now strongly recommended in all sorts of guidelines for treating patients with breast cancer. And you are definitely an expert on this topic, and I wanted you to just talk about your experience with these types of breast prognostic assays, what you think about them, where you think the field is going and things.


0:04:21.5 AB: Sure. Well, let me start from the beginning. I, as a junior assistant professor, this is now 1998, 1999, had the idea that I would develop a genomic assay, a predictor from the actual tumor itself to decide who develops lymph nodes that are involved. That was kind of what I wanted to do. I actually had the first... I had a grant, I had the first grant the Komen Foundation ever gave for it.


0:04:47.3 TS: Oh, wow.


0:04:48.0 AB: And so the thing is that, I've been very interested in this for many, many years. The whole concept behind this is that the intrinsic biology of the tumor decides the ultimate prognosis of the patient, as well as, the characteristics of what's gonna happen, and the response of this to therapies. And that's gonna be partially right because I think there's this whole other, we're not... And we'll talk about this kind of where we think the field is going, but there's all these host factors that we haven't even talked about it.


0:05:16.6 AB: But getting back to the tumor itself, there were a number of predictors that were developed. The first one... Going back even further, so in the mid-1990s, it became pretty clear that you can do reverse transcriptase PCR, that is what measure RNA levels in plasma pretty well, and they used to do that for... And they still do that for RNA load for HIV. It's really kind of where they got this. And so then a lot of us had the idea of trying to quantify RNA out of paraffin embedded tissue, and it was really, really difficult. I tried it, I was a senior, well I was basically, Assistant Professor, with maybe one or post-docs. It was really hard to do, but you know, [0:05:54.4] ____ figured it out, they figured out how to do it and how to make it reproducible. And the idea is that at least Genomic Health initially had a 21-gene assay that could basically tell us of someone's prognosis.


0:06:07.2 AB: They went back to retrospectively, all these databases they had for the NSABP, which is the National Surgical Breast and Bowel Adjuvant Project that had been going on for 30 years already, and they've been accumulating tissue blocks in that trial and they use this 21-gene assay to prognosticate the patients.


0:06:25.6 AB: And so that happened around 2003-2004. And what happened is that all of us got really excited and we went to the insurers and said, "Let us use it." And they said, "Why?" They said, "You're not really telling us anything." And so what happened is that then they said, "Okay, let's look at a trial called NSABP 20, which was a trial where they had women who got hormonal therapy with or without chemo, like Tamoxifen with or without chemo at the time, whatever the chemo was at that time, it was I think, CMF, it was a really old chemo, and they were able to say the people who had high risk, that is a score on the 21 gene score of over 31. They had benefit from chemo, and if you got a score under 18, you didn't have any benefit from chemo.


0:07:06.7 AB: And so that was really important, and suddenly the insurer said, "Okay, you now have a biomarker for telling us whether people get chemo or not." And so for that reason, then they approved it. The issue with the initial genomic health assay was that you could pretty much... Was just looking at RNA from things that we could measure very easily with immunohistochemistry, like the estrogen receptor or K-67 progesterone receptor or HER2. And so we wanted to do more, and there was a bunch of assays that were developed. There's one from Agendia, which is a 70 gene assay, which didn't have any of those genes in it, and that actually could predict whether you were high risk or low risk. So a lot of us started using that assay because it gave us, we felt more information.


0:07:52.4 TS: Yeah. It seems like it brought in more cell cycle progression factors.


0:07:55.1 AB: Yeah. We just felt that it brought in more genes. Interesting thing it didn't share... It shared one gene called CD68. That was shared between all of them. And CD68 turned out to be a really fascinating gene, we'll get to at the end 'cause it hints at kinda where all of this is going. But it turns out that we started using that. The problem with the 70 gene assay is that it was very binary. It wasn't linear, it wouldn't give us an exact number. You couldn't tell a woman... You just could tell her, "You're a high risk of recurrence, you made need chemo," or "you're low risk of recurrence and you don't." It didn't tell the precise number of their recurrence number. So then we now have the 12 gene assay, which is made by Myriad... Well, it's actually made by a company, made by a bunch of people in Europe and kinda sub-licensed to Myriad.


0:08:37.1 AB: The 12-gene assay does a lot of different things. The 12-gene assay gives you a score, but it does it on a continuous variable. And it also tends to use the size of the cancer and the number of lymph nodes that are involved. And so that brings you an entire picture because it turns out the size of the cancer and the number lymph nodes involved turn out to be very important for recurrence between years five to 15. Where a lot of these molecular features that this 21 gene assay was doing, the 70 gene assay was doing, really were more for recurrence in year zero to five. Although there's data now out about nine years for this that's still... That still works for those assays. So the bottom line is that for a lot of patients, I've actually moved to the 12-gene assay because these women want a precise number, because what they're doing when they come to us as a medical oncologist is that, well, "What's gonna happen to me?" is the first question. "Am I gonna die or not?" Number two, "To not die, what do I need to do?" Those are the two big questions they ask all breast medical oncologists. And so these genomic tests, at least to a first approximation can help us out.


0:09:43.4 AB: So a woman will come with an estrogen receptor positive, early stage breast cancer that has no more than three nodes, we will run one of these assays on her, and we'll build it into the entire equation. Kind of ask someone what their tolerance of chemo, do they really want it or are they busy, do they not wanna loose their hair? You get a sense in talking to women what they really wanna do. And then you weave in these assays into your recommendation to the woman. It really has changed everything. I'm old enough to remember when we didn't have any of this stuff that we would just take the pathology report and kinda make an educated guess based on that. These molecular assays do a much better job of letting us know who should need chemo and what their prognosis is, then [0:10:26.0] ____.


0:10:26.0 TS: Yeah, and really keep a lot of women off chemotherapy that would not have gotten any sort of benefit anyways from chemotherapy.


0:10:32.9 AB: Yeah, that's correct. It's about half, it's about half the women that we normally would give them chemo to, we don't have anymore. That's great because it's not only losing your hair, it's not really temporarily being sick. The potentially long-term side effects to chemo that we just don't know about. 'Cause we've only been giving chemo for 20 years or so, really 20-30 years of breast cancer and since the '90s. And I think that it's important that we understand that these are not totally benign things. Yeah, if you need it to prevent recurrence, you don't want your breast cancer to come back. So we give you chemo for that in addition to hormone therapy. But on the other hand, when we can avoid it, at least in my opinion, I try to.


0:11:11.0 TS: Yeah, that's great. And it seems like the three parts of this are in the beginning, do you need chemotherapy? Yes, no. What's the kind of benefit that can be expected from the chemotherapy? And then, where some of these assays are now also starting to be used is trying to figure out if extended endocrine therapy should be used for women. I just... What are your thoughts on that? 


0:11:30.9 AB: Right. That's really important too, because again, these therapies aren't benign. Giving a woman extended endocrine therapy can do a bunch of things. It can raise her cholesterol, it can lower her bone density. These are things, it can kind of mess a little bit with her... It's called chemo brain and hormonal brain. These are subtle and they're not common, thank goodness. But on the other hand, there are things that you have to take into account. It's not just preventing breast cancer. It's preventing breast cancer, but preventing it in someone who remains a healthy person and a functioning person. I think that you got... I've been doing this long enough now to know that it's not just... Preventing breast cancer isn't the be all and end all. Everything about a woman and her life that you gotta take into account.


0:12:16.2 AB: And I think that this whole idea of giving women more hormonal therapy is not... It does prevent a certain amount of recurrences. Absolutely. And it may, we're not sure yet, it may improve survival, there's no benefit yet. We do know it improves progression free survival, we don't know if it improves overall survival yet. So because of that, it's important to have some sort of assay. A molecular assay that we can get around the guess work.


0:12:38.5 AB: And so there are assays that do that now. There's one that I think is a 5-gene assay. Actually, it's a 5 or 7-gene assay from Biotheranostics. Theirs again. We can use this 12-gene assay from Myriad to help us do that. These are very helpful to us. I mean again, the problem is that these are so new that they often don't get paid. Although I think right now, Medicare, at a minimum and most private insurances are now reimbursing it. So we'll see how that goes. It really depends on your part of the country, but most people are reimbursing these now.


0:13:08.2 TS: Yeah. Walk the audience through a little bit of how you approach a patient. You're working someone up for breast cancer, how do you use the assays, have these conversations with your patient? 


0:13:18.2 AB: You know, so what will happen is that someone will come to see me, she'll either... Generally... Let's talk about someone who's post-op. We're not doing like the role what we call neoadjuvant therapy or this therapy before surgery, let's talk about someone post-op. So she'll come and she'll have a pathology report, it'll let me know the size of the cancer, the estrogen positivity, progesterone positivity, HER2, K-67. It'll let me know how many nodes, usually it's a [0:13:38.5] ____ central or lymph node dissection. How many lymph nodes are involved if any. '


0:13:45.2 AB: And so from there, we'll talk to the woman, get a sense of kinda what's going on, and I'll explain to her what her pathology is, what I think is going on, and if it's really benign looking, we have ways of deciding this, it's what they call in our group the McGee score. If the McGee score is really low, which is based on the pathology, immunohistochemistry and the Nottingham score, if that's really low, then a woman will not have one of the genomic tests, we'll just give her hormonal therapy. But if there's a chance she's gonna need chemo, we'll do one of these genomic tests and it usually takes two to three weeks to come back on average. And when they do come back, if it comes back low risk, then we'll put her just on hormonal therapy alone and probably radiation. If it comes back high risk, we'll talk about various chemos, then radiation, then hormonal therapy. That's generally how it works.


0:14:32.8 TS: Yeah, and how do you think the patients are? I mean when you see them, are they receptive to using these types of genetic...


0:14:39.0 AB: Very, they're extraordinarily receptive because they wanna avoid chemo if they can.


0:14:42.3 TS: Right.


0:14:42.6 AB: I mean they'll do it. Most women are gonna say... I mean, I have occasional women who say, "No chemo, no matter what." And those women, I'm not gonna do the test because I know they're never gonna want chemo.


0:14:51.5 TS: Right.


0:14:52.1 AB: But I think that the vast majority of women that I see wanna know. And I think that... And I always tell people it's better to do this right than fast, you have time. Because I always tell people, "What we do in the next three to four weeks isn't gonna determine what's gonna happen to you tomorrow. It's gonna determine what happens to you in the next five to ten years." And I think that. So that's why I encourage patience, I encourage a total work-up like this, I encourage taking their time to wait when the genomic tests are appropriate, I encourage people to take the time to kinda wait for them to come back before we make any decisions.


0:15:25.3 TS: Yeah. Excellent, and are you starting to use them more in the neoadjuvant setting, so prior to cancer? 


0:15:32.5 AB: Yeah, so one of the ideas is that...


0:15:34.1 TS: Prior to surgery, I should say.


0:15:35.3 AB: Yeah, we're learning... I mean, it's interesting. We've gone through this whole evolution. So again, I've been part of it, I participated in a lot of these clinical trials that are almost 20-30 years old already. Where we did neoadjuvant therapy on women and found that there wasn't a detriment to give the chemo before surgery. Now, is it any better to give the chemo before surgery? We're not sure. A certain percentage of women will develop what's called a pathological complete response, that's no cancer left after surgery, those women have a spectacular survival, over 90%.


0:16:04.5 TS: Yeah.


0:16:06.0 AB: And really the question is what do we do with these women afterwards? So we started giving everybody neoadjuvant therapy for a while, and we realized that certain women just weren't responding that well, and it wasn't really kind of affecting their overall survival, and those women tended to be estrogen-receptor positive breast cancers. We have a lot of factors that we use when we look at a core biopsy from somebody, whether we think that chemotherapy is gonna help or not. And we're not the only group. Groups around the country, and the world, are trying to figure out who is gonna benefit from neoadjuvant chemo? Those with a profound shrink of their cancer. I think we're starting to believe that some of these genomic assays can help us in that way.


0:16:45.3 TS: Yeah.


0:16:45.7 AB: That in other words, if you have a high risk genomic assay, it's likely you're gonna have a really high response to the chemotherapy. And so for that reason, some of us are determined to use genomic assays. Now, a lot of us will try to do it based on the characteristics of the profile of the core biopsy when some of us can do it, and when we have a way of doing it with what we call the McGee score. So we don't use these molecular tests neoadjuvantly but a lot of people around the country do, and I think it's a reasonable idea, it's a good use of the...


[overlapping conversation]


0:17:14.2 TS: Yeah. It does seem like it's shaping up a little bit, like you brought up, I mean if there's high molecular score, the tumor looks very aggressive, at least on the molecular side, those are the ones that if they're gonna respond to chemotherapy, they tend to in the neoadjuvant setting better. And then if there's a low molecular score and they don't look as quite as aggressive, those people are probably... Seem to be better candidates to go directly to surgery or if anything, maybe new adjuvant endocrine therapy.


0:17:41.8 AB: Yeah.


0:17:42.7 TS: Which is being evaluated, so... Yeah, thank you. No one knows better than you. What do you think this whole field... Where would you see this field going in 10 years from now? 


0:17:51.7 AB: Yeah, I'll tell you. It's really cool. I think what we're starting to learn just at the edges of all of these tests and some other things that we're doing, is that the immune system is extraordinarily important in the response to adjuvant therapy and what you do. I think to me and prognosis and survival. And I think that we're just beginning to scratch the surface, this is really the next quarter century in this business. Both not only in breast cancer, but in all cancers, and I think that we're trying to develop immune biomarkers, this is where the host comes in now, so [0:18:21.9] ____ 50 years, trying to figure out the cancer itself and its characteristics at a molecular level, and I think what we really haven't done for a while is look at the host. I've been very interested in this because I've been very interested in bisphosphonates in breast cancer for probably my entire career. And what bisphosphonates really do, they don't really affect the cancer per se, they affect the soil.


0:18:42.8 TS: Yeah.


0:18:42.9 AB: They affect whole micro environment that the cancer lives in. I think there's some recent things that have just started to come out that we're starting to think about. Like for example, we know that in certain women with breast cancer, both ER positive, ER negative, they have a survival... If they get bisphosphonates for up to five years. Be it Zoledronic acid intravenous or even oral bisphosphonates clodronate, they have an improved progression-free and overall survival. And so why is that? It's not the tumor, it turns out, then there probably is some manipulation of the micro-environment, and we would love to develop genomic assays to predict who benefits from that. And so that's one thing that people are doing. Another thing people are doing is that we now have CDK4/6 inhibitors in the metastatic setting, which is a big advance, and now we're being introduced into the adjuvant setting. There's a trial called monarchE that was recently announced at ESMO a few months ago, and there's a clear disease re-survival benefit to... In certain subsets of very high risk patients, there is a clear survival benefit if you give a CDK4/6 inhibitor with endocrine therapy. So what we would love to do is somehow come up with a genomic predictor for benefit from the CDK4. Right? 


0:19:49.4 TS: Mm-hmm.


0:19:49.7 AB: That way you would know upfront not only whether you need chemo, but whether you need hormonal therapy, whether you need a bisphosphonate, whether you need CDK4/6. All-in one, I'd love it all to be one test, and guys like me are dreaming about that. I mean obviously, it's probably never gonna be one test for a variety of reasons, but it may not even work for certain things, we may not find the predictor, genome predictor for these things, but on the other hand, it would be really cool to be able to find a genome predictor for all these other things, and I've been pushing a lot of people recently to think about that.


0:20:20.6 AB: Again, it's all gonna come back. All of this stuff that we're doing is gonna come back to the immune system. I'll tell you one in particular CD-68. Okay, is the one shared gene between the 70-gene assay and the 21-gene assay that turns out to be a very important gene in immunity on macrophages, it's really interesting where this is gonna go. People are now presenting immune signatures and prognosis at a lot of our meetings, how to integrate those into the existing therapy, the existing molecular assays we have. These are all the questions that a lot of us are gonna be kinda thinking about over the next 5 to 10 years. And then again, we're all getting old. We're gonna leave this to the next generation, we'll set the stage for these guys. I've been doing this a long time. And I'd love to do it forever, but everybody kinda reaches retirement. So I mean I'm not retiring tomorrow, but everybody reaches retirement at some point. And we just have to set the stage for the young generation, the generation in front of us to kinda take these and run with them. And I think that's kinda where this is gonna go.


0:21:21.8 TS: Yeah, and it just seems overwhelming that you can get so much extra information by digging deep into the cancers. And then on top of everything that we're talking about here, then it also brings in targeted therapeutics and everything which...


0:21:33.2 AB: Correct. Absolutely.


0:21:35.0 TS: Breast cancer has been a little slow to warm up to for various reasons, and I'm just wondering how you think that's gonna play out in the future too. Do you think we're gonna [0:21:44.7] ____ all breast cancers fairly soon? 


0:21:45.6 AB: We... Breast cancer, by the way, had the first real targeted therapy we had trastuzumab. People tend to forget that. Before all the targeted stuff that we have in the lung and in melanoma and all these other tumors, pancreas now, prostate, we had breast. We had trastuzumab for breast. Tamoxifen target the estrogen receptor too, that's a targeted therapy too. We've had that for almost 50 years. So the thing is that we've been a little bit slower to use molecular biology to find targets, absolutely. But now we have a lot of targets. We have the estrogen receptor whether you are ESR mutated or not. So if your ESR mutated, you do better with fulvestrant than aromatase inhibitor. We have PI3 kinase mutations, we now have alpelisib for that. We now and 3 to 5 percent of patients have HER2 mutations. HER2 tyrosine kinase is likely you are likely gonna have activity with these people. Even you and I actually published with a fellow a couple of years ago, estrogen receptor amplification. Now we published a case report, we had a woman with ER amplification and we gave her high dose estrogen based on an animal model, that someone else had published and she had a dramatic response for a year and a half.


0:22:50.0 TS: Yeah, that's great.


0:22:51.1 AB: So we have a lot of targets that are really coming out, in breast now. And we also have PARP inhibitors like we have for ovarian cancers.


0:22:58.9 TS: Right and immunotherapies. Yeah.


0:23:00.7 AB: Exactly. So you can now have that. And so we have all of that stuff and then on top of it, we're starting to see these signatures that will decide who can benefit from immunotherapy, checkpoint inhibitors. And we're gonna probably start applying that to breast cancer also. Again, these molecular assays are just gonna expand dramatically. My hope is that we can have one global asset that encompasses it all, but that's unlikely because of obviously the competition from all of the different pharma that make them. But on the other hand, I think that the idea that we're gonna be able to individualize therapy based on the molecular aspects, not only of the tumor, but of the host herself or himself in the case of other cancers. I think it's gonna be a really fascinating thing to watch for the next 15 to 20 years.


0:23:47.0 TS: Yeah, absolutely. It's gonna be exciting. Exciting times ahead. I really, really appreciate you coming on the podcast. You have an incredibly deep knowledge base and your experience is very evident when you casually talk about the timelines here. So thank you so much for coming on. I hope now at the end of this, our listening population really does have a good handle of different types of gene expression assays and where this whole field is going. So thank you so much from the bottom of my heart.


0:24:16.2 AB: You're welcome. Thank you very much for having me, this has been fun.