Understanding risks and predictive factors behind Breast Cancer


There are some things you can do to manage your personal health and the risk of developing cancer in your lifetime. But, wherever you go, you’ll take your DNA with you.


Dr Erika Spaeth joins us to discuss those mutations and how they combine into a measurable impact on cancer risk and prognosis. Plus, the development a new model to assess that multi-factoral risk with a greater deal of certainty to concerned patients for what their genetic fortune holds.


Read the original research: https://doi.org/10.1007/s10549-022-06834-7


Image Source: Adobe Stock Images / Blue Planet Studio





The following transcript is automatically generated.



00:00:04 Will Mountford 

Hello. I’m will, welcome to researchpod. 

00:00:07 Will Mountford 

There are some things you can do to manage your personal health and the risk of developing cancer in your lifetime. Age, anatomy, location, and lifestyle hold tremendous sway. 

00:00:18 Will Mountford 

But wherever you go, you’ll take your DNA with you, along with the little mutations and variations that make you. 

00:00:26 Will Mountford 

Erica Spaith joins us to discuss those variations and how over a lifetime their cumulative contributions to disease stack up into meaningful, measurable impact on cancer risk, which can ultimately impact a patient’s screening and risk reduction options available to them based on current guidelines. 

00:00:43 Will Mountford 

Her team’s work at genetic technologies has led to the development of a new single model to assess that combination of clinical and genetic risk together, which delivers a greater deal of certainty to a concerned patient for what their genetic fortune holds. 

00:01:01 Will Mountford 

Joining me to talk about the work is Doctor Erika Spaeth. Erika, Hello. 

00:01:05 Dr Erika Spaeth 

Hello, thank you for having me. 

00:01:08 Will Mountford 

Thanks for coming. Could you tell us a bit about your work, about the position that you hold and everything that’s led us to the interview that we’re having today? 

00:01:15 Dr Erika Spaeth 

Yes. So I work for a company called Genetic Technologies. I’m the director of clinical and Scientific Affairs and our focus is on disease risk assessment, so. 

00:01:28 Dr Erika Spaeth 

We look at a series of risk factors, both clinical and genetic. We put them together in an algorithm and we predict. 

00:01:36 Dr Erika Spaeth 

People’s risk of developing certain. 

00:01:38 Dr Erika Spaeth 

Uses and what we’re talking about today is breast cancer risk and how we assess women’s chances of developing breast cancer and then what they can do as a result. 

00:01:49 Dr Erika Spaeth 

Of that risk. 

00:01:50 Will Mountford 

Don’t forgive me for saying that algorithmic healthcare modelling sounds like a very 2023 kind of idea to have. How long have you been working on this and what’s led to the paper that we’re discussing? 

00:02:03 Dr Erika Spaeth 

Yes, there’s quite a bit of work that’s gone into this. The company has been working on breast cancer risk assessment since 2011. The scientists and epidemiologists and biostatisticians have been working on modeling long before us as well. So we haven’t developed this concept, but we are part of the game. 

00:02:24 Dr Erika Spaeth 

Now and we’ve gone through several different iterations of this algorithm and it started off as one model and slowly developed over the years to the current format. 

00:02:37 Will Mountford 

What is it about the work that has kept you with the company for as long as you’ve been there to now be Director of Scientific Affairs and what is it about this place in healthcare research specifically that connects with you on a personal level? 

00:02:51 Dr Erika Spaeth 

So my background is in cancer biology. I did. 

00:02:55 Dr Erika Spaeth 

My doctoral and postdoctoral studies. 

00:02:57 Dr Erika Spaeth 

Learning about the tumor microenvironment and studying why the tumor microenvironment. 

00:03:03 Dr Erika Spaeth 

In some adults is well, it’s protective and it never turns into a tumor microenvironment. So the person stays healthy. But in other adults, the normal cells end up supporting tumor growth and what flips that switch. So that’s where my background is. And when I learned a little bit about. 

00:03:23 Dr Erika Spaeth 

Polygenic risk, which is one of the risk factors that our model incorporates here at genetic technologies. 

00:03:31 Dr Erika Spaeth 

It really turned on a light bulb in my head and from my background in the tumor microenvironment, I started thinking about polygenic risk and one of. 

00:03:41 Dr Erika Spaeth 

The many factors cancer is a multifactorial disease, so there’s no one thing that can drive cancer. But it’s this combination of things and some of these risk factors that we’re looking at. 

00:03:51 Dr Erika Spaeth 

Particularly from the genetic landscape in this polygenic. 

00:03:55 Dr Erika Spaeth 

Risk. Maybe some of these baseline markers that you’re born with, so you can’t change them and maybe it predisposes some people to a higher risk of developing certain cancers and that’s really of interest to me. 

00:04:10 Will Mountford 

Could you tell me what that polygenic risk is to kind of break it down to get everybody up to speed to start? 

00:04:16 Dr Erika Spaeth 

Polygenic risk is. 

00:04:18 Dr Erika Spaeth 

A combination of genetic risk factors, what we call them as single nucleotide polymorphisms, which is a a big word, and we can shorten that and call them snips. 

00:04:31 Dr Erika Spaeth 

Most of them are in between genes. 

00:04:35 Dr Erika Spaeth 

They are situated throughout the genome and every adult has these markers, so the majority of adults DNA is 99.9% identical. The differences between all of us are these snap. 

00:04:52 Dr Erika Spaeth 

Ups and these snips can help us define a person by their genetic ancestry, but they can also be used to help us associate certain disease risks. It’s different from hereditary genetics and something called BRC A1 or B RCA two when you have. 

00:05:12 Dr Erika Spaeth 

A so-called mutation or a pathogenic variant in a gene like a BRC one, you have one specific point that’s different. 

00:05:23 Dr Erika Spaeth 

And it causes the gene to break and so the gene can’t function and it can’t create the. 

00:05:30 Dr Erika Spaeth 

Appropriate downstream protein. That’s the landscape of hereditary genetics. The way we understand it, if you have a significant family history, then you might get genetic testing to look for one of these mutations or pathogenic variants. 

00:05:45 Dr Erika Spaeth 

Polygenic risk is very different from that we aren’t looking at specific genes. 

00:05:51 Dr Erika Spaeth 

Most of these snips are actually located in between the genes, and they don’t have a known functional consequence. The polygenic risk is really just an association with the disease and one single snip that we look at. 

00:06:11 Dr Erika Spaeth 

On its own is not very relevant. It has such a small amount of risk associated with the single location. 

00:06:20 Dr Erika Spaeth 

Now the value of polygenic risk is the Poly part, meaning many. So we put many of these snips together and in some adults that risk can incrementally build up. And that is the value of this risk factor being included with clinical risk factors in this. 

00:06:40 Dr Erika Spaeth 

Algorithm that we started off talk. 

00:06:42 Dr Erika Spaeth 

Thing about. 

00:06:43 Will Mountford 

To draw a possibly inexact metaphor, it’s the difference between a wheel falling off of your car with a known risk factor like BRC one, just one or two screws falling out here and there over the course of you know, running it for 5-10 years and then suddenly everything falls off like a clown show. 

00:07:01 Dr Erika Spaeth 

Exactly. That’s a perfect metaphor. 

00:07:04 Will Mountford 

Well, seeing as I’ve just made it up, I’ll. 

00:07:05 Will Mountford 

Have to remember it for later. 

00:07:11 Will Mountford 

Sticking with breast cancer for now to kind of get a lay of the land for what is out there at the moment in terms of risk forecasting and modelling for cancer risk and breast cancer specifically, what are some of the tools and frameworks that? 

00:07:26 Will Mountford 

Have come before that people are using that any clinician listening to this might be familiar with. 

00:07:31 Will Mountford 

What’s come before? And then here’s what we are doing to move things forwards aft. 

00:07:36 Dr Erika Spaeth 

Yes. So there are some risk models that currently exist in the clinical space. One of them is called the Gale model or the breast cancer risk assessment test. Another model is called the Tyler Cusip or the Ibis model. 

00:07:51 Dr Erika Spaeth 

There’s a model called the can risk model, which was formerly the body Asia. So there are quite a few different models that. 

00:07:58 Dr Erika Spaeth 

Exist that clinicians. 

00:08:00 Dr Erika Spaeth 

Depending on the patient type. So some patients with extreme family histories where they have very suspicious cancer running through multiple generations, one model might be used over another. The majority of women don’t actually have a family history at all, or might have one distant cousin. 

00:08:20 Dr Erika Spaeth 

For an aunt and then a more basic general population model can be used. So there are quite a few different models that are used depending on the scenario. 

00:08:30 Will Mountford 

So for the model that we are talking about with that polygenic consideration, am I right in saying it is the risk model to set stand for anything and brisk is the catchy name for it? 

00:08:40 Dr Erika Spaeth 

It is the name that we’ve put in the academic manuscripts, so the official name from a commercial perspective is gene type, because genetic technologies has created this test and it’s called gene type. But in our academic papers. 

00:08:54 Dr Erika Spaeth 

The model is. 

00:08:55 Dr Erika Spaeth 

Called Brisk, which stands for breast cancer risk. 

00:08:59 Will Mountford 

So the differences about this being that polygenic difference, how are samples being gathered and put into it to I guess go through the validation and calibration of setting this tool up? Is it just a matter of going back to past patient data or new patients who are coming in to participate in clinics? 

00:09:19 Will Mountford 

Some of the operations and logistics behind getting this. 

00:09:23 Will Mountford 

Up and running. 

00:09:24 Dr Erika Spaeth 

Yes. So we built the model and then cross validated the model using large cohort data. So these are large collections of longitudinal data. So the patients were enrolled into a study. We have a lot of their baseline metrics as well as their DNA. 

00:09:44 Dr Erika Spaeth 

So we have the genetic information, the clinical information and then these studies follow the patient over a series of time. And so we’re able to capitalize on data like that by looking at the patients, applying our model at a time zero when all the patients are healthy. 

00:10:02 Dr Erika Spaeth 

And then we look at them over a period of time and what our risk model and any other risk model does is it predicts a woman. 

00:10:09 Dr Erika Spaeth 

‘S chance of. 

00:10:09 Dr Erika Spaeth 

Developing breast cancer over a five year period of time. 

00:10:14 Dr Erika Spaeth 

So we provide A5 year prediction and then we actually observe these women over 5 years and we see how right we were of the women that we put in a high risk category, how many of them actually did go on to develop breast. 

00:10:29 Dr Erika Spaeth 

Answer the cohorts we used. One is the UK Biobank which has over 200,000 women in it. Another one is called the Nurses Health study, which is a smaller nested case control cohort with a few thousand women. So we used two different groups of women and in both of those groups we were able to compare. 

00:10:50 Dr Erika Spaeth 

Our model against a gold standard, one of them being the Gale model and the other being the Ibis model. So. 

00:10:57 Dr Erika Spaeth 

Those were the 2. 

00:10:58 Dr Erika Spaeth 

Gold standard models that we compared to. 

00:11:01 Dr Erika Spaeth 

2 compared to those models, we look at a few different statistical metrics, something called discrimination, and that’s the ability of the model to differentiate between affected and unaffected or cases and controls. And in both data sets, our model was able to better discriminate. 

00:11:23 Dr Erika Spaeth 

Then the traditional gold standard model. 

00:11:27 Dr Erika Spaeth 

Another metric we look at is called calibration, and again we look at calibration by looking at all of these women at a time zero. We predict their risk and then we observe that risk and we want to see that our predicted risk actually aligns with the observed risk. 

00:11:47 Dr Erika Spaeth 

And if we see a Nice 45 degree line, if you plot that on an X&Y axis that is a well calibrated model, because what we are predicting is actually being observed. 

00:12:00 Dr Erika Spaeth 

And then another metric we look at is called net reclassification. What we do with net reclassification is we classify women based on a model and here we use the gold standard model. So either the Gale model or the Ibis model, we classify the women into risk categories. Now, clinically, there are actually thresholds. 

00:12:22 Dr Erika Spaeth 

There are risk thresholds that identify a woman as at increased risk or average risk, so we use those clinically actionable thresholds to classify women with the traditional model. 

00:12:34 Dr Erika Spaeth 

And then we take those same women and classify them with our model and what we show is some women we move up in category up a risk. Some women may move down a risk and how well are we appropriately recategorizing. 

00:12:49 Dr Erika Spaeth 

Again, what we show there is that we move more women who end up being affected, meaning they developed breast cancer. We’re moving more of them into the increased risk category, which is a good thing. It means we are better able to identify. 

00:13:04 Dr Erika Spaeth 

By them and we’re moving more women who are unaffected into the lower risk categories. 

00:13:10 Will Mountford 

Imagine getting a letter in the post to say due to statistical error, you are now at a higher risk of disease or we have, you know, take the weight off your shoulders, actually reduce your risk. Is that anything that goes to patients or does just stay within the systems? 

00:13:27 Dr Erika Spaeth 

So for these cross validation studies that stays within the systems right now, so anybody that’s enrolled in a study and given researchers access to the data, a lot of that data never actually goes out to the specific. 

00:13:41 Dr Erika Spaeth 

But where we do communicate with patients is the doctors that currently order the tests now. And so doctors do order the tests, order the risk assessment and we do have reports that go out to the patients identifying where their risk is sitting. 

00:13:55 Will Mountford 

This all puts risk in a position to be at least non inferior to the other model. This is that right? 

00:14:01 Dr Erika Spaeth 

Correct, exactly. And that was the point of this cross validation to show that certainly it’s non inferior and in fact outperforms the traditional models. 

00:14:12 Will Mountford 

Was that something that was in the scope of the design, or would you have to come up with a bigger study cohort or a bigger design to truly capture the extent of that superiority? 

00:14:22 Dr Erika Spaeth 

No, I think the numbers that we have in, in this UK Biobank data set are quite high. 200,000 women is quite a large number in these data. We looked over a five year period of time. We’ve actually since gone back and looked over a 10 year period of time as well with that same data. 

00:14:39 Dr Erika Spaeth 

Set and the model performs just as well, so it’s working over a longer period of time, and the women that have subsequently gone on to develop breast cancer were still predicted to go on to develop breast cancer, which is a good thing. 

00:14:53 Will Mountford 

Is there a plan for that to be put out to healthcare systems, to doctors, and start bringing it into the process? 

00:15:00 Dr Erika Spaeth 

That is what we are working on because even though these initial data. 

00:15:05 Dr Erika Spaeth 

We’re used to just show at least equivalency, and now we’ve shown that they are better than the models and theoretically you would think they should be used clinically. We’ve done extra work on the side as well to show the economic benefit from clinical implementation, which is an important aspect because doctors are busy. 

00:15:25 Dr Erika Spaeth 

They don’t have a lot of time on their hands and we want to try to give them a tool that can. 

00:15:31 Dr Erika Spaeth 

Bring benefit to them and their patients. 

00:15:34 Dr Erika Spaeth 

And so in these modeling that we did, we did show that finding more of these at risk, women were able to identify more cancers earlier, which is a good thing. Having an earlier stage of cancer improves outcomes. 

00:15:55 Dr Erika Spaeth 

What’s also interesting but underutilized in clinical practice right now is there are risk reducing medications that exist for breast cancer. 

00:16:03 Dr Erika Spaeth 

These medications aren’t new, and the clinical trials to back up the efficacy are 25 years old. There’s nothing novel about these risk reducing medications that exist, but they are severely underutilized in clinical practice. 

00:16:20 Dr Erika Spaeth 

Part of our efforts is not only to identify women that can benefit from more screening and supplemental screening, which is one of the actions that a clinician can take. 

00:16:31 Dr Erika Spaeth 

As a result of a risk assessment, but they can also identify the women that have significant risk, where that medication, the benefit of the medication, outweighs any potential harmful side effects of the medication. Most of these drugs are all generics already tamoxifen, Raloxifene, aromatase inhibitors. These are common drugs that breast cancer. 

00:16:52 Dr Erika Spaeth 

Patients go. 

00:16:53 Dr Erika Spaeth 

On so, it’s not like a vitamin, they are more significant drugs. The studies, the clinical prevention trial studies have shown between a 30 and a 65% risk reduction in women on the drugs compared to placebo, which is amazing. If you can prevent half of all breast cancers. 

00:17:14 Dr Erika Spaeth 

Are occurring. Why aren’t we putting more? 

00:17:16 Dr Erika Spaeth 

Women on it. 

00:17:17 Dr Erika Spaeth 

And this is 1 area that genetic technologies would like to help support, and we’re hoping by developing a risk model that is more accurate and can identify more women that are truly at increased risk, maybe that would actually make clinicians and their patients feel more comfortable taking a medication that, yes. 

00:17:37 Dr Erika Spaeth 

The medication does have side effects, hot flashes inducing menopausal symptoms. That’s not pleasant, but if you’re a truly high risk of developing breast cancer, it’s something worth considering. 

00:17:52 Will Mountford 

Is there any leading publication or partnerships to name drop? 

00:17:56 Will Mountford 

At this point. 

00:17:58 Dr Erika Spaeth 

We do have quite a few different working relationships from both the research side of things as well as developing partnerships. We are in the process of trying to show the economic value to get payers on board, which is a big step particularly in the US. 

00:18:17 Dr Erika Spaeth 

To get payer coverage. 

00:18:19 Dr Erika Spaeth 

Right now this test is a self pay test which does limit its ability to reach the general population. Yes, exactly. So we do have some partnerships with larger institutions where we are rolling out this model to show utility and that clinicians do change the way they. 

00:18:26 Will Mountford 

Let’s hurdle. 

00:18:39 Dr Erika Spaeth 

Treat their patients based on this different risk. 

00:18:43 Dr Erika Spaeth 

And as a result of that, we hopefully we’ll be publishing something in the next 12 to 18 months showing those data, but also gaining coverage. So this can actually be rolled out because this test, even though it’s a wonderful test in terms of understanding your own DNA from the you know you can talk about. 

00:19:04 Dr Erika Spaeth 

Personalized medicine and here we are looking at a person specific DNA markers that are unique to them and the combination of their polygenic risk is unique to that. 

00:19:15 Dr Erika Spaeth 

But in terms of a population health application, this is a population level test. It’s meant for all women. It’s meant to stratify all women to identify that subset that could actually benefit from more screening or the risk reducing medication or more frequent. 

00:19:34 Dr Erika Spaeth 

Screening or in younger women, earlier screening, maybe they shouldn’t be waiting till that standard age when mammograms. 

00:19:41 Dr Erika Spaeth 

Start. Maybe they need it a little sooner, so there are so many decisions that could be made based on risk and that’s what we’re looking to show. There is value at a population level and it really does need to be implemented population wide to see the benefit. 

00:19:59 Will Mountford 

You mentioned in your very introduction that genetic technologies is working on a couple of other avenues of research, including cardiology and other cancers. Does the logic, the rationale that maybe not the exact algorithm but the the framework behind the brisk model makes sense when applied to other? 

00:20:18 Dr Erika Spaeth 

It does 100%. It’s all about the combination of this polygenic risk with clinical risk factors and together we get a better picture of that person’s risk of developing that disease. And so we are really excited about some of the other cancers. Colorectal cancer is 1 where we have very good. 

00:20:38 Dr Erika Spaeth 

Stratification results, and there’s some other cancers that have fewer screening options right now. So colorectal cancer, there are good screening options that exist. Pancreatic cancer and ovarian cancer, for example, don’t have great screening option. 

00:20:53 Dr Erika Spaeth 

And so to risk stratify and to be able to identify people that could benefit from the screening options that do exist. Since there’s not population wide screening programs, it’s beneficial to try to capture those at risk adults earlier. And so that’s something we’re really excited about when we look at our. 

00:21:14 Dr Erika Spaeth 

The multiple cancer risk assessments that we are continuing. 

00:21:17 Dr Erika Spaeth 

To work on. 

00:21:24 Will Mountford 

And these tests, as you mentioned, are clinically validated for use like the breast cancer model, which to return to for a second. Can you tell me why risk assessment is such a valuable tool in clinical practice anyway? 

00:21:37 Dr Erika Spaeth 

It’s important to really understand where a woman is sitting, because not all women fit into the general population yet we have general population screening, and so most women are funneled into a single category. 

00:21:51 Dr Erika Spaeth 

Right now we really only pull women out if they have a suspicious family history of cancer, but there is more to risk than just the family history. 

00:22:01 Dr Erika Spaeth 

And that’s a really important point. And I think the second point in the context of breast cancer is breast density and particularly in the US with this federal mandate that’s going into effect in a few months here in the US where all imaging centers across all 50 states have to have. 

00:22:22 Dr Erika Spaeth 

Breast density notification letters going out. 

00:22:26 Dr Erika Spaeth 

These notification letters tell a woman and their provider that the woman is at increased risk of breast cancer because of dense breast tissue. Almost half of all women are going to be getting these letters because dense breast tissue is very common. 

00:22:42 Dr Erika Spaeth 

But that’s it. They get a letter. 

00:22:44 Dr Erika Spaeth 

And then what? 

00:22:46 Dr Erika Spaeth 

There aren’t any specific downstream actions and so to risk assess on that population is actually a very valuable tool. 

00:22:57 Dr Erika Spaeth 

A lot of imaging clinics already do provide risk assessment, but not all of them because it’s not necessarily standard across all clinical practices and so applying a risk assessment to help make it more of a standard of care. 

00:23:12 Dr Erika Spaeth 

So clinicians aren’t in the dark, so patients also aren’t in the dark and saying, well, I got this. 

00:23:18 Dr Erika Spaeth 

Letter. I don’t know what. 

00:23:19 Dr Erika Spaeth 

It means not sure what. 

00:23:20 Dr Erika Spaeth 

To do. 

00:23:21 Dr Erika Spaeth 

I guess I’ll just wait till my next mammogram. 

00:23:25 Dr Erika Spaeth 

Or maybe a risk assessment is appropriate, and if they’re at high risk, maybe they can have extra discussions. And if they’re at lower risk or within average risk, then yes, maybe just waiting till that next mammogram is sufficient and they can be slightly reassured. So there is value in just aiding in decision making. Discussions between clinicians and their patients. 

00:23:46 Dr Erika Spaeth 

By adding an additional tool like a risk assessment. 

00:23:49 Will Mountford 

If anyone listens to this, wants to know more about your research specifically or about genetic technologies as a company. 

00:23:56 Will Mountford 

Where can they find that if they want to? 

00:23:58 Will Mountford 

Know more? 

00:23:58 Dr Erika Spaeth 

Our website is genetype.com, so that’s a good place to start and our contact information is on there and we would be happy to get in touch. And then on the research articles, I believe my e-mail is associated with the author list on their research articles. So getting in contact with our authors. 

00:24:18 Dr Erika Spaeth 

Correctly is also something that’s possible. 

00:24:21 Will Mountford 

Doctor Spaeth, thank you so much for. 

00:24:22 Will Mountford 

Your time today. 

00:24:23 Dr Erika Spaeth 

Thank you. 


Leave a Reply

Your email address will not be published.

Researchpod Let's Talk

Share This

Copy Link to Clipboard