You are here

Dr. Katherine Nathanson

Perfecting and Personalizing Risk Assessment with Dr. Katherine Nathanson

Dr. Katherine Nathanson shares her research on breast cancer gene mutations, the breakthrough CARRIERS study, and more
Tags: 

While breast cancer is not typically caused by inherited factors, as many as 10-15 percent of people diagnosed with breast cancer carry a known genetic mutation. The most well-known mutations are in the BRCA1 and BRCA2 genes. But these only account for 5-10 percent of inherited breast cancers, so what about the many other gene mutations that increase a person’s risk of breast cancer?

One major BCRF-supported cross-institution study, CARRIERS, found—among other results—that some mutations that had previously been linked to breast cancer were found not to increase the risk of disease.

What does this mean not only for genetic testing—but also how we should consider results? More significantly, what effect might this have on the personalization of risk?

BCRF investigator and cancer geneticist Dr. Katherine Nathanson was one of the study’s principal researchers. She is deputy director of the Abramson Cancer Center and the Pearl Basser Professor for BRCA-Related Research in the Perelman School of Medicine at the University of Pennsylvania. She not only runs a prominent research laboratory but also maintains a clinical practice. 


Read the transcript below:       

Chris Riback: Dr. Nathanson, thanks for joining. I appreciate your time.

Dr. Katherine Nathanson: Glad to be here.

Chris Riback: Let's start by understanding some of the background of hereditary breast cancer and the mutations. Most of us have heard of mutations in the BRCA1 and BRCA2 genes. But those aren't the only mutations, obviously, associated with breast cancer, are they?

Dr. Katherine Nathanson: No. So we think about breast cancer susceptibility as a spectrum of susceptibility. So, you have your high penetrance genes, in which mutations are associated with a relatively high risk. So usually BRCA1 and 2, for example, that's associated with a high lifetime risk. Then you have your genes in which mutations are associated with what we call a moderate lifetime risk, usually in the range of twofold. There is a gene that's sort of in between, which is called PALB2. But there is a gene that's associated with a moderate lifetime risk, which is CHEK2, ATM, and things like that. And there are a set of what we call normal variation. Sometimes those variants are associated with minimally increased risk, but when we put all those variants together, we get what's called a polygenic risk score, or lots of variants together to try to think about risk.

So the way that I talk about it, which hopefully is helpful, is everyone gets dealt a hand of cards. And sometimes when you get dealt a hand of cards, you get the ace of spades. You know what the ace of spades is. It's a really high card. It means something really important. That's your BRCA1 or 2. You have a moderate penetrance gene. You get dealt a hand of cards. Queen of clubs, it's helpful card if you're playing Bridge. But it's not the best card. It's not the greatest card, but it's there. And then you get everything else. Your six of diamonds, your eight of spades. And those are your normal variation. But when you put all those together, sometimes they can make a very powerful hand of cards.

And so if you think about the different genes in which there are mutations, you can think that everybody gets dealt a mixture. Sometimes you get different hands than others. Everybody gets dealt a mixture of those sort of hands. And some of them are high, some of them are moderate, and some of them are low combinations. This is the way I sort of think about an analogy to help people out.

Chris Riback: It's really helpful, and it makes me wonder. Which hand is the most common, and which hand is the most worrisome?

Dr. Katherine Nathanson: So the hand that's most common is probably the hand, I don't know if anyone plays Bridge. Right. The hand that you get the bunch of dreck [trash]. And it's like, you're just playing it defensive. That, in fact, in this sense is probably the better hand. Right? And the more unusual hand is when you get the high cards. And it's much more unusual to carry a BRCA1 and 2 mutation than it is to get the sort of regular hand of assorted not-so-great cards.

Chris Riback: And so, how would you characterize the extent to which those mutations, the extent to which they increase breast cancer risk or are associated with breast cancer risk, compared to the BRCA1 and BRCA2?

Dr. Katherine Nathanson: So like the queen of clubs in the analogy that we were just using.

Chris Riback: Yes.

Dr. Katherine Nathanson: So those increase your risk modestly. So, it depends on the gene, but instead of for example, you have to remember that any woman in the US has a 12 percent lifetime risk of breast cancer. So just to point that out, right? So, these increase your risk potentially to 20 to 25 percent. That's higher, and we can show in large studies significantly higher. But it's not nearly as high as a BRCA1 or 2 mutation. And in that context, actually, the other cards are much more important.

Chris Riback: Why is that?

Dr. Katherine Nathanson: It's because we talk about modifiers. So we know that in addition, it's a hand of cards. Right? So in addition to having that ace of spades, your other cards matter. But when you have something like a queen of clubs, the other cards matter more. Because it's really, there's not such a dominant, we say, mutation in a gene. The other things matter and affect how that works. The moderate penetrance genes more.

Chris Riback: The interplay among them.

Dr. Katherine Nathanson: Yes. Exactly.

Chris Riback: So earlier this year, you and others published a population study of more than 64,000 women in the New England Journal of Medicine. You analyzed the data from the CARRIERS study, cancer risk estimates related to susceptibility. What did you find, and why was it so groundbreaking?

Dr. Katherine Nathanson: So the reason it was groundbreaking was because it actually was a population-based study. So much of our estimates of risk actually don't come from people in the general population. They come from women who have a family history, who are collected on the basis of family history, or some of the recent studies from women who have had genetic testing. And usually you don't, maybe not these days, but [it] used to be you didn't just walk in and get genetic testing. Right? Anybody. And so they were biased estimates. Biased in terms of who was high-risk. So-

Chris Riback: And self-selection of who was going and getting the testing.

Dr. Katherine Nathanson: Exactly. Exactly. And so these were a number of, and I just want to emphasize, huge team of people worked on this project. Really would not have been possible without collaboration. The senior author was [BCRF investigator] Dr. Fergus Couch from Mayo Clinic. I'm on the executive committee after meeting every week for five years, but we got somewhere. But the people who ran the studies and the most important people who collected samples from patients, and the patients who contributed to those studies. So hugely important. You can't do it with any aspect of that.

So the point was that these were very large, mostly older women, who had breast cancer, who either actually were followed prospectively and then got breast cancer. So there is some cohorts of studies, something called the Nurses' Health Study, where they look at all the nurses and they follow them prospectively over a number of years. And then some of them get breast cancer, for example. And those studies are very helpful to understand, if you're in the general population, what's the risk of these mutations in someone who doesn't have a family history or doesn't present to a high-risk clinic. And so that's actually why it was particularly important. And not surprisingly, it showed that people with a family history have a higher risk of having mutation.

So, the idea was that we found that women with breast cancer from the population, about 5 percent of women overall, and mostly these were older women, had mutations. Whereas about 1.5 percent of women without breast cancer had these mutations. So it was significantly more frequent, but not high numbers, right, of women. And then we were able to identify what the risk associated with each of these mutations is, or mutations in each of these genes, I should say, more accurately is. And I think, actually, I'm going to tell you what I think is really important about this study, which I think is very hard to get from the news coverage.

Chris Riback: Yes please.

Dr. Katherine Nathanson: So I think that there are two parts that are very important. One is, this is 3,200 cases and 3,200 controls. Despite only five genes overall were significantly associated with breast cancer risk. So the number, actually, of genes wasn't all that high. Now there are some additional genes, if you looked at ER-negative and triple-negative disease, that were associated [with] disease. But the other part is that we routinely have women who are getting these huge gene panels for breast cancer susceptibility. And many of the genes on the panels were not associated with breast cancer susceptibility in this study. And I think the negative aspect of this is actually really, really important. We really-

Chris Riback: What do you mean by the negative aspect?

Dr. Katherine Nathanson: So for example, I'll give you a good example. So there's been a lot of discussion. Are mutations in a gene called MBN associated with breast cancer risk? And there's back, forth, is it associated with disease? Our study and the accompanying study BRIDGES, which was done in a European group, definitively showed no association with disease. So one of the things that is really important is that especially when you're looking at these women who are getting, because it's easy to do the testing, these big gene panels. Some of the genes who are basically ruled out as being associated with risk.

And that, actually, honestly, to me is the most important part of this. Because most of the ones that we associate with risk, we knew were associated with risk. But some of the other ones where some people said yes, some people said no, were more borderline. And we were able to really definitively say these are not associated with risk.

Chris Riback: And I would have to assume that in terms of levels of anxiety, intrusion of hair that may or may not have been needed, I would think. You correct me, please, if I have it wrong. That for years, probably, you had women who would get those results and there was uncertainty and stress, anxiety, and perhaps even more.

Dr. Katherine Nathanson: Right.

Chris Riback: And you were now kind of able to say, "We don't need to worry about that."

Dr. Katherine Nathanson: Exactly. And to me, that's actually in part the most important part of this. That we were able to really definitively show that they are not associated. I think it helps us tell us a little bit about who, although it doesn't really change the guidelines substantially about who should be tested. That's important too, like who should be tested. Women with ER-positive breast cancer over the age of 65 should not be tested. And those kinds of things, that's important too. I think there are another series, I'll say, of follow-up studies, some of which are out, some of which are not out.

So there was also a study that looked at the rate of mutations in Black women. And one that looked at the rates of compared within the CARRIERS [study] group, the rates of mutations in Black and white women. There were no differences, which is also very important to say. No, the rates, there was some variation within the genes. But overall, there was no significant difference between Black and white women. Very important to say. And we have some studies that are accepted, which are coming out soon, actually looking at the non-, what I call the lower cards, but the polygenic risk scores. And how the polygenic risk score looks in this dataset, how does it influence risk.

It's clearly much more important in, as I said, if you have a lower or a moderate penetrance gene than it is if you have a high penetrance gene. So those things are important. And looking at the rates of mutations in women over 65, very important mostly if you have triple-negative disease. Not so much if you don't. But to my mind, I think that part of the thing that people, for me, as someone who clinically sees patients who get huge panels, having really good negative data is really, really important. Because it helps tell people, we don't think this is causing your disease. And you don't need to have surgery, you don't need to have other kinds of interventions, because we think this is good.

Chris Riback: This has to significantly alter the way one thinks about the personalization of risk.

Dr. Katherine Nathanson: Right. And the personalization of risk, I think, is really, really interesting. I mean, lots of people are very interested in personalization of risk. And there's lots of ways to approach personalization of risk. So first of all, do you carry a high, moderate penetrance gene? What is your polygenic risk score? So we know that in women with a high polygenic risk score, their risk of breast cancer is as high as someone who has a moderate penetrance gene. If you have a low polygenic risk score, one of the things that they've thought about is decreasing mammographic screening in women who have low polygenic risk scores. Very hard in the US, particularly, to de-escalate things like that.

But it is definitely something that I think, in other countries, may happen. That if you have a low polygenic risk score, a low score, low risk of breast cancer, you may de-escalate or decrease screening. So you think, now we have genes. The genes and the variants sort of affect each other. So they're not independent. Mammographic density is a known risk factor. Obviously things that we also think about, smoking, alcohol, physical, all those kinds of things are risk factors. And even within genes, the different mutations don't all have the same risk. And so there's lots of ways that people are thinking about how can we better personalize risk. But I think the thing that we have to think about along with personalizing risk, is what are we going to do with that personalized risk? Do we feel comfortable, if we say someone's at a really low risk, saying, "Well, you need less mammography." I think that's something, as a...

Chris Riback: Yes. Very hard to do.

Dr. Katherine Nathanson: Very hard to do. These are complicated questions I don't have the answers for.

Chris Riback: Yes. Well the complicated questions, you understand, doctor, are your fault.

Dr. Katherine Nathanson: Right, exactly.

Chris Riback: If you had to participate in these studies, we have simple questions and very unclear answers. Instead, we've got a couple.

Dr. Katherine Nathanson: Well, yes. I mean, it's complicated questions.

Chris Riback: The complicated questions. You mentioned a moment ago some of the either currently ongoing or potentially beginning studies. And I believe, and you correct me, please, if I have this wrong. That one of them is one that you and your team are launching around a study of very high-risk women with early-onset breast cancer, to identify more about what drives the risks of breast cancer in these women. Do I have that right?

Dr. Katherine Nathanson: Yes. So we know that, even though we've done this large study, we know there are several things that are probably some rare, identified variants that are associated with disease. Now those could come in several flavors. So is it possible that there are additional breast cancer susceptibility genes that are rare and unidentified? Definitely possible. But you need to really look at very high-risk women who have been screened negative, and that's something that we're interested in, or doing, I should say. And I think that the other thing that we've also thought about is, there are some limitations to the way we currently look at the known genes.

Technological limitations. They're not state-of-the-art changes over time. I think that's really important. And so, are there variants in those genes or associated with those genes that might be associated with risk that we're missing? Which is definitely possible in women with breast cancer. And so, that's something that we're also interested in looking at. And we were starting to do something called whole genome sequencing. So not just looking at the genes, but looking at the sequences between the genes to see if there is variation that might potentially be associated with risk. There are also what we call rearrangements.

So we know, for example, in BRCA1, that if you think about it like a sentence, you can have misspellings in the sentence. But you can also have missing words or missing phrases. And so, we're very good at getting the misspellings. We're pretty good at getting some of the missing words. But sometimes we're not so good at getting the large phrases or rearrangements. Or if they're, for example, between the words or something like a comma. How far can I take this analogy, right? That we don't pick those up so well. So we're looking to see if some of the families have variants that are associated with that. Those are harder to prove as being functional, so that's one of the things that we think about.

Chris Riback: How did you come to focus your work so significantly on high-risk women? And I know you deal, as well, in other types of cancer. But in high-risk areas, is what I have taken.

Dr. Katherine Nathanson:  Oh sure. So I'm a geneticist. So I decided very early in my career, actually when I was in medical school, that I wanted to go into genetics. I was always fascinated by genetics growing up, and I thought it was really interesting. And I decided in medical school to go into genetics. So I'm actually double-trained in internal medicine and clinical genetics, which is wild, still. Pretty unusual. There's probably less than 50, maybe less than 100 people in the country.

Chris Riback: Wow.

Dr. Katherine Nathanson: And so, then when I decided to, I trained in genetics. I did research training. And so, I ended up doing cancer genetics for a variety of different reasons. And that's where I've sat ever since. But I really approach it or come to it from someone who's really trained as a geneticist, rather than an oncologist kind of perspective. So I've been doing it a long time. And my clinical practice actually deals with a lot of other hereditary cancers, as well. So that's-

Chris Riback: And there are two. So first, fascinating to me that even as a kid, genetics was what interested you, drove you. So, it was never going to be opera singing or—

Dr. Katherine Nathanson: Oh God, no. You wouldn't want that.

Chris Riback: So as well, there’s an intersectionality as I'm learning about you and reading about you, that you do. One is, you just talked about it. The translational medicine. You run a lab, but you also see patients. In addition, you don't focus, it seems, on just one type of hereditary cancer. You focus on several. Are there learnings? Is that just because you get bored easily? Or are there learnings and insights that you gain from?

Dr. Katherine Nathanson: So I would say, it isn't because I get. I think it's more historical. I sort of ended up doing lots of different things for various historical reasons. And I'm actually doing things that are a little different now, because I'm very interested in really trying to improve genetics generally. I think we really have to be prepared for genetics for everyone, and that's something that's really important to me. I have found in my science, however, that you never know what you're working on comes back to what you were working on before.

I think the biggest analogy, I worked on, actually, a study that actually has led to a drug, down the road, but for renal cancer. But part of the study, all of a sudden we found BRCA1 in this study of renal cancer. And it has to do with DNA damage, and it happened to be important there. And I thought, my God, I try to do whatever I do, and it sort of all comes back together. But it is important. I see neuroendocrine tumors, and one of the things that's really happening clinically to me is that people are getting these big panels. They get mutations in a gene called SDHA. SDHA is associated with pheochromocytomas and paragangliomas, but all the women that I see have breast cancer, and they got this big panel.

So we're now trying to figure out and make sure. They don't think it's associated with breast cancer, but we're doing some large-scale studies to make sure that I can confidently tell a woman who comes in with SDHA that that's not associated with the breast cancer risk. So I feel like it all links back together, because I can do that, because of all the work that I've done on breast cancer already. And I want to make sure that I'm counseling my patients appropriately. So I feel like it all seems to intersect and come back together on some level.

Chris Riback: Yes. It seems like it. And I would be remiss if I didn't ask you. What role has BCRF played in your research?

Dr. Katherine Nathanson: As I say, I will tell this to anybody. BCRF has been the most important thing to my research. I had a difficult early career for lots of reasons, but BCRF provided funding for me at a critical juncture in my career. I would not still be sitting here doing what I do if it wasn't for BCRF. I absolutely feel that that's true. I feel like they have been incredibly important over my career in terms of the support that they have provided. Like when I was an assistant professor, sort of struggling, and now, I mean, I don't think of myself as that much. But now I have a deputy director of a cancer center.

Chris Riback: Now you're a big poo-bah, yes.

Dr. Katherine Nathanson: And I wouldn't be there if I wasn't for BCRF.

Chris Riback: Well, we all benefit from the work that you do. Thank you for that. Thank you for taking the time today with me.

Dr. Katherine Nathanson: More than happy. It was great.

Donate By Check

If you prefer sending a donation by check, you can mail it to:

The Breast Cancer Research Foundation

60 East 56th St., 8th Floor

New York, NY 10022

Other Options

From paying tribute to a loved one to workplace giving, there are a variety of ways to support research.

Learn More