June 1, 2025 • 37 mins
In this episode, we sit down with Dr. Michael Liebman, a leading voice in personalized medicine, to explore how data, disease modeling, and global collaboration are reshaping women’s health. From breast cancer to menopause, Dr. Liebman explains why current approaches fall short—and how more accurate, individualized risk profiles can unlock safer, smarter care.
Discover what it will take to move beyond “one-size-fits-most” healthcare and what his groundbreaking research could mean for the future of women everywhere.
Learn more about Dr Liebman's work: https://ipqanalytics.com/
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:01):
Welcome to Sage Women, the podcast for midlife ladies
navigating perimenopause, menopause, and healthy aging.
I'm Melanie White, and each weekI talk with experts and women,
sharing real stories, practical insights, and smart tips to
thrive through this stage of life and beyond.
It's Melanie White here on the Sage Women podcast, and I'm
really pleased to be introducingtoday's guest, Doctor Michael
(00:23):
Liebman. Michael, welcome and thanks for
being here. Thank you.
Good to be joining you and especially as we start on our
collaboration. Very excited to talk about that
collaboration and I guess in this episode for people
listening, we'll be talking about a really innovative
approach that is going to changethe face of women's healthcare.
(00:47):
Well, we hope. We hope to, Michael, Could you?
Could you? Give the listeners a bit of
background about you please. Sure, I'm actually trained as a
theoretical chemist, but I tell people I'm a reformed chemist.
So I have a mixed background. I've been in academics for a
(01:08):
number of years and then I went into industry back when
bioinformatics was starting out.I was with Amoco, the oil
company and headed up bioinformatics.
And then from Amoco we launched the DNA diagnostics company and
(01:30):
I worked on the original her twonew tests that's used in cancer
before there were any drugs for it to try to figure out what to
do with the test. And from there I went to pharma.
I was at Wyeth instead of bioinformatics, and then Roche,
as global head of computational biology and genomics, went back
(01:55):
to academics for a few years, University of Pennsylvania.
And then I was director of a breast Cancer Center sponsored
by the US Department of Defense jointly with Walter Reed, where
we did everything from surgery, pathology, tissue banking, all
(02:16):
the way back to genomics, proteomics and informatics, and
then launched the group now IPQ Analytics, and now our nonprofit
as well. Wow, there's some big words in
there and a a lot of history andexperience.
I know I'm thinking for the layperson listening to this
(02:36):
what? What is the simple definition of
something like informatics? That's changed over the years.
It, it's when I started in bioinformatics, to be honest
with you, I wrote one of the very first papers using the word
and I bet something very different from what it is now.
(02:57):
But obviously it has to do with information.
And I, I actually have a figure that's been shared a number of
times about the difference between data, information,
knowledge and clinical utility and how they grow at very
(03:18):
different rates. And so information is really
clean data and in my terminology, but it's not
knowledge, you still haven't interpreted it.
It's just a clean set of data that you can start to work with.
That that's a great definition. Thank you.
(03:38):
And it just makes me think aboutcoming from Biological Sciences
and Biostatistics, I start thinking about the clean data
and then all of the things that you can make it say depending on
what your interest is. Yeah, there's a lot of sayings
about that. Yeah, Yeah, right.
And so then it's it's what do wedo with this data and how do we
(04:01):
use it for good? And it sounds like you've done a
lot of work and some pioneering work in the Women's Health
space. Can you talk a little bit about
that? Yeah, I've, I've been
interested. Well, actually it started back
when I was working on the original her two test, which
obviously we were looking at forbreast cancer.
And it evolved over the years and came from recognizing,
(04:25):
although I'm not trained in any clinical areas that we really
didn't understand disease. But we had a lot of technologies
we were trying to develop and apply to it, whether it's
biomarkers or diagnostics or something like that.
And I actually do what I call systems modeling.
(04:45):
And so I'm interested in lookingat how big is the system, not
just the question you're asking,because we tend as humans to
reduce the complexity to something that makes it simpler
to address. But that means we frequently are
leaving out very critical pieces.
And so was I was looking at how to use things like the her two
(05:11):
new test and recognizing that wereally didn't understand the
disease. That became something that I
became very interested in, in following up on and looking at
it from a lot of different perspectives because we have to
recognize that the patient has aperspective.
(05:31):
Even the patient's caretaker hasa perspective.
But then there's the form of perspective, there's a
physician's perspective, the regulatory agents perspective,
device perspective, diagnostics,and they're all different
perspectives on the same thing. But what they aren't really
(05:51):
tending to do is integrate very effectively across all of those
views. You know, I, I use an analogy
that it's like having a house and having everyone looking into
the house through different windows telling you what they
see and no one really recognizing that they're looking
(06:12):
at a house. And so I became interested in
trying to see if we could build a house and then let everyone
have their view. So what we've tried to do is
sort of invert that data collection perspective to look
at the size of the problem, not necessarily expecting that we
(06:33):
can solve the whole problem, butwe'd like to make the complexity
of the problem much more transparent.
So if what you are doing doesn'twork, we have an idea of why and
what the next steps are and can build it logically from that.
So I think what you're saying isyou're taking a very holistic
(06:54):
approach, looking at a problem from all angles and the complete
size and journey of the problem,just not the symptoms or one
person's point of view or experience, so that you can
really see what's going on and and what led somebody to that
point. Well, and, and we like to enable
then people who have those specific viewpoints to also see
(07:15):
that these other factors could be very critical, even if
they're not normally typically thinking about them.
But it gives them a way to startto incorporate that and expand
their view in a fairly logical way.
Not suddenly disruptive per SE, but adding on even incrementally
(07:38):
to broaden their views. I'm wondering if your breast
cancer risk model is an example of that?
Well, it is, you know, it's LED us to look at a number of
different things. So we build models, we use a lot
of different kinds of technology.
(08:01):
What I focus on is not the technology side, but it's what
is the problem that needs to be addressed.
And so we want to start looking at what's the question that not
even the question that's being asked, but what is the question
that should be asked. And so from running the breast
(08:21):
Cancer Center, what I learned was how to work with clinicians,
breast surgeons and and oncologists to gain their
confidence so that why we could do what they were asking us to
do to contribute. I needed to know what they
really didn't know and what werethe questions they wouldn't stop
(08:46):
to ask because that wasn't theirtraining, nor did they think
that they were easy answers or easy ways to get answers.
And so the way I describe it is physicians have to be
operational. They don't have the luxury of
being strategic in general. It's a generalization.
(09:08):
We have the opportunity to be more strategic work to find out
what root cause issues may existand whether they can be
addressed. And then having identified them,
then we look to figure out what's the right technology.
So we don't, we, we use a lot ofdifferent technologies as a
result, but we're really problemfocused.
(09:30):
And there's a there's a saying by Edward Deming, who was an
economist and is, if you don't know what the question is,
you'll never find the answer. And and that's really a guiding
point for how we operate. And so on a path to identifying
breast cancer risks, what might be some of the things that you'd
(09:53):
be looking at? Well, what we're looking at is
the breast is somewhat unique because it's one of the organs
that undergoes a lot of developmental change post
delivery and throughout a woman's lifetime.
And as a result, what we're trying to identify is where are
(10:13):
critical events taking place that may be years before a
breast cancer develops, but our actual early triggers or setting
the path for how that breast cancer may develop.
Because we have been very successful if we look at the
statistics in extending lifetimeof a breast cancer patient.
(10:39):
So the life expectancy after a diagnosis because of all the
effort on treatment has been fairly successful in extending
that life cycle or lifetime. But what we haven't had any
impact on is the incidence of breast cancer.
(10:59):
And so while survival is going up, long term survival is going
up, mortality is going down, incidence has continued to rise.
What we're interested in is understanding that whole process
to understand why incidence is going up and whether there may
be ways to actually intercede much earlier than what we're
(11:23):
doing now, which is the diagnosis as an early stage of
the disease. But could we actually even start
to look at preventing the disease if we know what those
causes are? And so that's how we're trying
to build models of how women progress over their lifetime,
what are the factors of their personal development that are
(11:46):
unique to them, which could be clinical history, it could be
lifestyle, it could be diet, allthese factors.
And then what are external factors that may be important
exposure to toxins, environmental perturbations,
things of that nature. And so we're trying to build
models that allow us to encapsulate all of that data and
(12:11):
then see how women might be identified as higher risk much,
much earlier. And it's more than just
genomics. We're looking, we include
genomics, but it's not purely A genomics effort because we know
genomics is involved, but it's not necessarily the sole driver
(12:35):
in any of this. In fact, most breast cancer when
we talk about it, only 10% of itis typically associated directly
with genomics, 90% is associatedwith PERI and postmenopausal
transition. And that leads us to that
conversation of menopause and obviously breast cancer is one
(12:57):
of the things that is discussed at that time, including many
other things such as neurodivergence and osteoporosis
and all sorts of things where what are we looking at in terms
of understanding the perimenopause to postmenopause
journey? Well, I, I, I use, I, I use the
Rumsfeld term that we're dealingwith the unknown unknowns in
(13:21):
that domain. I think at least from all the
women I've ever spoken with, none of them are satisfied with
the information they're able to obtain from their gynecologists
or obstetricians. And it's not the fault of their
clinicians, but it's not actually taught very much in
(13:41):
medical school. And and that comes about because
we haven't had the data, we haven't had the research, we
haven't had that as an emphasis on understanding it.
It's just always been there and expected to be there.
But if we look at something likeperimenopause, which is anywhere
on average 7 to 10 years in length, it's a very complex
(14:06):
process, that transition and it's going to impact every other
disease that a woman may have inthat 7 to 10 year period.
So understanding it just as a process and how it affects an
individual is important to manage her entire healthcare
(14:26):
profile, not just the transitioninto menopause.
Menopause is actually just a state transition.
It means you haven't had a period for 12 months, but
perimenopause is that process upuntil then.
And if we look at it again from a systems view where typically
perimenopause, it can start muchearlier, but typically it starts
(14:50):
somewhere in the early 40s and and menopause is typically
around 51 on average. Now what that means with given
how lifetimes have increased over the years, a woman is going
to spend almost half of her lifeeither perimenopausal or
(15:11):
postmenopausal. And so not understanding that is
really detrimental to having thebest life experience and
managing everything else that the woman will experience,
whether it's disease or just quality of life issues.
And so that's that's how we've started to try to build profile
(15:36):
like that. I can see that it's a complex
and convoluted area to be working in because every woman's
going to have her own history upuntil that point.
And there's the genetic side of things and then there's the
environmental side of things, and then there's the lifestyle
side of things, all of which aregoing to influence this,
(15:57):
influence the woman's journey through that stage of her life.
And I'm thinking particularly ofthe, the, the data that exists
showing the average age of menopause and how that
difference different differs in different racial groups.
And then for smokers, they go through menopause 2 years
earlier. So there's obviously a lot of
things that are influencing whena woman goes through menopause
(16:20):
and what she goes through. Right.
And, and, you know, in, in general, in our modeling, one of
the key elements that we focus on is that disease or a
condition is a process, not a state.
And so that means we need to understand that women will go
(16:42):
through perimenopause differently.
We suspect that there are probably several common paths
that they may take. Now, that's not going to be an
absolute identical path for every woman.
There'll always be some nuances,but there will be some
commonalities. We think that can be observed,
but we don't have the data to measure that and observe that.
(17:05):
But what that means then is whatwomen tend to do, because as I
said before, they're not gettingthe information they really
would like or the answers to questions they have.
They turn to their friends and networks of friends.
And I mean, it's a very natural way to try to get information.
(17:26):
But as I always tell groups thatI talk to, your hot flashes and
your friends, hot flashes are not the same.
You may both have hot flashes, but they're not the same.
And you know you have it on the there's been identified about 62
different symptoms that are associated with perimenopause.
(17:49):
Not everyone goes through them, of course, but knowing which
ones you're experiencing are going to also impact something
like your hot flashes. And we treat those even as
comorbidities or other conditions along with your hot
flashes, just as an example, because hot flashes affect about
(18:12):
80% of all women. So it's a common experience the
to show you what what we mean. It's very interesting, isn't it,
that the different pathways to, to getting to that state.
And as you say, it's a we, we need to look at the process.
It's not just the state. And I'm also thinking about
(18:34):
treatment now because what you're saying if we have
different pathways to get to hotflashes, for example, that means
that the intervention or the treatment is probably going to
be different too. Is that right?
We would expect that. So this is really personalizing
medicine. This is what personalized
medicine means, at least the waywe think about it.
(18:57):
It requires having accurate, an accurate understanding of the
disease. And, you know, when we're
looking at how women are going through perimenopause, we are
actually mapping out in our models all of the critical
hormonal transitions that take place during development, which
(19:17):
means we are going back at leastto menarche.
And we're doing a very detailed profile of pregnancy histories
because these are the most hormonally disruptive or
variation, you know, periods in a woman's lifetime.
(19:38):
And we need that kind of information to know what her
journey is going to be like. The challenge we have is that
data doesn't exist, and we've scoured a number of
international databases at leastespecially at the level we're
building models because we're looking at, say, for instance,
(20:00):
the the idea that we know that something called null parity is
a risk for breast cancer. Most people think null parity
means you never had a child, butit actually only means that you
never had a live birth. And when we look at menarche and
we understand from the data thatthe age of menarche is going
(20:25):
down, it's decreasing globally and the age of first pregnancy
is going up, that means the period of time in between them
is increasing. That's a period of no parity for
every woman. And so the question is, does
that actually it correlates with, but is it causal for what
(20:48):
we're seeing with the incidence increase of breast cancer?
We don't know. We don't have the data to study
it. And that that's the kind of data
we're trying to now gather through the collaboration we're
setting up with you to work withwomen who are interested in
contributing data, contributing some and learning more about
(21:11):
themselves. Because not only are they going
to have individualized perimenopausal transition, but
that should also indicate they're going to have
individualized risk for breast cancer, for cardiovascular
disease and osteoporosis. And they're not all going to
have the same risk profiles. Yeah, it's, it's such
(21:35):
interesting work, Michael, and I'm really pleased that we're
collaborating to start collecting data.
And one thing that came to mind as you're talking about that is
how important it is to get good quality data and uniform data.
There's so many. It's, it's not just a matter of
collecting different sorts of data from different sorts of
(21:56):
people and trying to cobble it together because that's not
meaningful and it's apples and oranges.
You really need to have a unified collection method and
something that's very aligned and and ubiquitous in a way to
to get that quality that you need.
Well, and, and that's actually part of our research is how to
(22:16):
establish what are the criteria for the data elements you're
collecting because we that we work in other areas besides
Women's Health. Women's Health is the focus of
our nonprofit, and we've seen that challenge of trying to put
data together from different databases and understanding from
(22:40):
work that I've done in the past that different centers may call
something or define something differently.
And So what we're finding nowadays with the push for big
data, which everyone is looking at so they can apply things like
artificial intelligence and machine learning, the data
(23:02):
scientists, the data engineers who aren't familiar with the
nuances of how the clinical practice takes place, are
putting data together based on data labels.
And data labels aren't adequate.You have to actually understand
how those things were done. And you know, the example that I
(23:24):
use is looking at breast cancer.One of the areas that's very
under under served so to speak, is the area of triple negative
breast cancer. And you would think what triple
negative breast cancer means is that you in testing the patient
(23:46):
you find out that their estrogenreceptor level, their
progesterone receptor level and their her 2 levels are all below
the threshold. So they're considered negative
for all of those. That's why they call it triple
negative. But what you also find out when
you work with different centers is what test was administered in
(24:09):
one center may differ from the test administered in another
center, or how they set the threshold for negative may be
different. And the reality is a patient
who's diagnosed triple negative in one center may not be
diagnosed triple negative in another center.
And that presents its own challenge for how to manage the
(24:33):
patient, but that also propagates into drug
development. If you look at the clinical
trials for new drugs that are targeting triple negative, you
find out that even in the in those clinical trials, that kind
of detail is and rigor is not necessarily being addressed.
And there could easily be many drugs being developed today that
(24:57):
because of the lack of rigor anddata collection, are not
actually being evaluated or shown where they could be
effective. And you, you end up with,
through lack of, of standardization, you bring in
heterogeneity into patient populations and it washes out
(25:19):
the effect of the drug. So you don't see that part of
the population where it might work.
And so it, it's not just that's designation and and the I CD10
code or something like that, butit goes through the entire
healthcare ecosystem. And it really dilutes what's
(25:42):
happening and it and it masks things.
Yeah. So then that's there.
Our collaboration is really important because your goal is
to collect data globally for a study that's going to start to
really look at a more personalized approach for
Women's Health. Yeah, yeah.
So again, we've extracted methods.
(26:08):
We borrow methods from other disciplines, as I said, because
we're focusing problem solving. But then as we look at the
methods, we see what needs to beadapted or expanded.
So one of the methods we use is called a knowledge graph.
It's a form of artificial, It's in that area of artificial
(26:29):
intelligence. But we've expanded the the
approach in a novel way to look at a lot more rigor so that it's
not just a data warehouse, but something that can be made
actionable in the clinic. But in doing that, these kinds
of nuances, for instance of knowing where the test, what
(26:51):
kind of test was done, how the threshold was set, these are
very detailed characteristics that we are building into the
knowledge graph. And then when we build the study
or the survey used in our study,we carry that over and apply
that to ask the questions the correct way to be able to know
(27:15):
what those unique differences are.
And that that that is very critical to being able to
actually use the data we would like in the ways we would like
to, to build models. The other kind of question that
is important, I pointed out thatthere's 62 symptoms that are
(27:36):
identified or associated with perimenopause.
And many people, many women havehad surveys or participated in
surveys from different organizations or or government
policy, whatever, where they have a checklist to check off
what they've actually experienced.
(27:59):
And of course, these are also unique to individuals.
But it's not enough if we're using our approach that says we
need to understand the process, to just have a checklist of what
you experienced. What we're asking is when did
you experience it? In what order did you experience
(28:21):
it, and what did you experience at the same time?
And those are the factors that will help identify an
individual's journey, not just the general characteristics.
And that's the kind of data thatwe build into the model and then
into the surveys or studies thatwe're doing.
(28:43):
It's such exciting work, Michael.
I can see that this is going to really be a game changer for
Women's Health and for healthcare generally.
It's going to lead for more personalized medication and
other treatments. It's going to help us to
understand contraindications better for medications and other
treatments. And is 1.
It feels like A1 size fits most at at the moment.
(29:06):
Well, yeah, I mean, but what youwhat one of the things we would
like it to be able to do is be able to provide it for a woman
some guidance to know if what she's experienced experiencing
is part of that normal transition or not, rather than
(29:29):
not knowing and and not having agood resource to talk to to find
out. A common a common occurrence
during perimenopause are some cardiac events, small events.
But instead of running to the cardiologist who may not even
anticipate as part of perimenopause, if you know that
(29:51):
this kind of event actually occurs in your transition,
you're going to feel more comfortable with that and not
not immediately panic, so to speak, that having, you know, a
much more serious condition or if you're experiencing something
that isn't necessarily part of that transition, then you know,
(30:14):
it is something that needs additional attention.
So giving the individual, givingeach woman more control over her
own experience and knowing how to interpret it.
Yeah. And it's such an important point
that you make about that, the knowing and the sense of control
(30:36):
around that and sense of confidence or, or greater
certainty that comes with that. Because too often we're speaking
to women who don't even realize that they've been in
perimenopause for the past 2-3 or four years, which means
they've missed an opportunity for early intervention, for one.
And it means that they've been struggling with symptoms and
(30:57):
thinking that they're going mad too, and then going to doctors
or other healthcare practitioners and not getting a
diagnosis. So there's this fear coming up
around what's actually happeningto me.
And I've heard women say, I wonder if I'm getting early
onset dementia, for example, it's a real concern.
So to have to have your work available, to have that data
(31:22):
collection, to understand your own unique risk profile, your
own opportunities for early intervention, that is really
important and powerful. And we're so pleased to be
partnering with you to share this data collection, study and
initiative. With the people, yeah, right.
(31:43):
We, we think it, it's well the, the goal is running it through
the non profit. And the reason we're doing that
is having been involved in commercial operations and pharma
and investment small groups and so on.
The, the need to commercialize something can sometimes redirect
(32:09):
how the study is done by keepingit in our nonprofit world, we
believe we're doing it the way we think it should be done and,
and not trying to take certain shortcuts or, or directions that
that may not give us the comprehensive knowledge we're
(32:30):
trying to get out of this. So that's one of the reasons
we're we've moved it into the nonprofit.
We have other aspects of Women'sHealth that we're tying into
this, of course, because we're also looking at hypertensive
disorders of pregnancy and preeclampsia on their own, but
(32:51):
then also how they impact even perimenopause much later on.
So none of these things we thinkare isolated.
It's really a continuum of looking at normal development
and how these presentations are influenced from different
factors. You know, we, we actually in the
(33:14):
US have been running some publicforums for women in association
with some of the professional women sororities in particular.
And it's just to allow women to come in and ask questions and
you know, they're they're there aren't really those
opportunities. I know you're through your
(33:36):
organization providing coaching and and that's a real key factor
that provides some comfort I'm sure to a lot of women being
able to rely find a reliable source of resource.
But you know, we think this kindof information, we're collecting
it in the US, we're in discussion, collecting it in
(34:00):
Africa and in Chile as well. And so we think this kind of
information and being able to docomparative analysis as well as
going to at least open up new doors.
I never tell anyone we're going to solve the problem, but we're
going to make things much more transparent than they are.
(34:22):
Yeah, it's really, really powerful work.
I love that you're exploring different countries and
different populations and aimingfor that broader transparency.
And I mean, certainly in Australia, we realize that there
are underserved populations here, remote populations,
regional and rural groups calledwomen, culturally and
(34:43):
linguistically diverse. And we know nothing about their
experiences and their life history and how that might
influence their journey into chronic disease or not in later
lives. So we're really excited to be
getting this work out and we'll be launching this in June for
anybody that's interested in partnering with us to share the
(35:04):
the work and get women engaged to contribute information and
help you with your data collection and study.
That's that's the goal for us. And I'd love to finish on this,
Michael, with understanding whatwhat your goal is for this year
with with this work. Well, we've done preliminary
studies using UK Biobank, the US, all of US studies and so on.
(35:31):
But, but we need to actually fill the gaps and the gaps are
the data we're trying to collect.
So what we'd like to be able to do and, and one of the the ways
we've set this up using it as anonline anonymous first pass data
collection is exactly to addressthe fact that you could be
(35:54):
rural, you could be you, you canaddress it on a telephone, a
phone and answer the questions. So being able to link in from
anywhere. It's not something that requires
an academic setting or going into your doctor's office.
In fact, it doesn't require anything that you should not be
(36:16):
able to address on your own without looking up records.
But our goal then is to to get alarge enough population that we
can show the feasibility and then we really have a much
deeper set of questions we'd like to ask, but that will be
(36:38):
much more involved. This first pass is what we want
to accomplish this year, collecting this first set of
data. We look forward to helping you
with that work, Michael, and launching this in Australia in
June. It's really exciting and I will
probably come back to you later in the end for another update on
(36:59):
how things are going and what you've learned.
And thank you for your time today.
Thank you, Looking forward to continuing to work together.
It's amazing. We're really thankful for the
opportunity to partner with you.
Welcome to Sage Women, the podcast for midlife ladies
navigating perimenopause, menopause, and healthy aging.
I'm Melanie White, and each weekI talk with experts and women,
sharing real stories, practical insights, and smart tips to
thrive through this stage of life and beyond.
It's Melanie White here on the Sage Women podcast, and I'm
really pleased to be introducingtoday's guest, Doctor Michael
(00:23):
Liebman. Michael, welcome and thanks for
being here. Thank you.
Good to be joining you and especially as we start on our
collaboration. Very excited to talk about that
collaboration and I guess in this episode for people
listening, we'll be talking about a really innovative
approach that is going to changethe face of women's healthcare.
(00:47):
Well, we hope. We hope to, Michael, Could you?
Could you? Give the listeners a bit of
background about you please. Sure, I'm actually trained as a
theoretical chemist, but I tell people I'm a reformed chemist.
So I have a mixed background. I've been in academics for a
(01:08):
number of years and then I went into industry back when
bioinformatics was starting out.I was with Amoco, the oil
company and headed up bioinformatics.
And then from Amoco we launched the DNA diagnostics company and
(01:30):
I worked on the original her twonew tests that's used in cancer
before there were any drugs for it to try to figure out what to
do with the test. And from there I went to pharma.
I was at Wyeth instead of bioinformatics, and then Roche,
as global head of computational biology and genomics, went back
(01:55):
to academics for a few years, University of Pennsylvania.
And then I was director of a breast Cancer Center sponsored
by the US Department of Defense jointly with Walter Reed, where
we did everything from surgery, pathology, tissue banking, all
(02:16):
the way back to genomics, proteomics and informatics, and
then launched the group now IPQ Analytics, and now our nonprofit
as well. Wow, there's some big words in
there and a a lot of history andexperience.
I know I'm thinking for the layperson listening to this
(02:36):
what? What is the simple definition of
something like informatics? That's changed over the years.
It, it's when I started in bioinformatics, to be honest
with you, I wrote one of the very first papers using the word
and I bet something very different from what it is now.
(02:57):
But obviously it has to do with information.
And I, I actually have a figure that's been shared a number of
times about the difference between data, information,
knowledge and clinical utility and how they grow at very
(03:18):
different rates. And so information is really
clean data and in my terminology, but it's not
knowledge, you still haven't interpreted it.
It's just a clean set of data that you can start to work with.
That that's a great definition. Thank you.
(03:38):
And it just makes me think aboutcoming from Biological Sciences
and Biostatistics, I start thinking about the clean data
and then all of the things that you can make it say depending on
what your interest is. Yeah, there's a lot of sayings
about that. Yeah, Yeah, right.
And so then it's it's what do wedo with this data and how do we
(04:01):
use it for good? And it sounds like you've done a
lot of work and some pioneering work in the Women's Health
space. Can you talk a little bit about
that? Yeah, I've, I've been
interested. Well, actually it started back
when I was working on the original her two test, which
obviously we were looking at forbreast cancer.
And it evolved over the years and came from recognizing,
(04:25):
although I'm not trained in any clinical areas that we really
didn't understand disease. But we had a lot of technologies
we were trying to develop and apply to it, whether it's
biomarkers or diagnostics or something like that.
And I actually do what I call systems modeling.
(04:45):
And so I'm interested in lookingat how big is the system, not
just the question you're asking,because we tend as humans to
reduce the complexity to something that makes it simpler
to address. But that means we frequently are
leaving out very critical pieces.
And so was I was looking at how to use things like the her two
(05:11):
new test and recognizing that wereally didn't understand the
disease. That became something that I
became very interested in, in following up on and looking at
it from a lot of different perspectives because we have to
recognize that the patient has aperspective.
(05:31):
Even the patient's caretaker hasa perspective.
But then there's the form of perspective, there's a
physician's perspective, the regulatory agents perspective,
device perspective, diagnostics,and they're all different
perspectives on the same thing. But what they aren't really
(05:51):
tending to do is integrate very effectively across all of those
views. You know, I, I use an analogy
that it's like having a house and having everyone looking into
the house through different windows telling you what they
see and no one really recognizing that they're looking
(06:12):
at a house. And so I became interested in
trying to see if we could build a house and then let everyone
have their view. So what we've tried to do is
sort of invert that data collection perspective to look
at the size of the problem, not necessarily expecting that we
(06:33):
can solve the whole problem, butwe'd like to make the complexity
of the problem much more transparent.
So if what you are doing doesn'twork, we have an idea of why and
what the next steps are and can build it logically from that.
So I think what you're saying isyou're taking a very holistic
(06:54):
approach, looking at a problem from all angles and the complete
size and journey of the problem,just not the symptoms or one
person's point of view or experience, so that you can
really see what's going on and and what led somebody to that
point. Well, and, and we like to enable
then people who have those specific viewpoints to also see
(07:15):
that these other factors could be very critical, even if
they're not normally typically thinking about them.
But it gives them a way to startto incorporate that and expand
their view in a fairly logical way.
Not suddenly disruptive per SE, but adding on even incrementally
(07:38):
to broaden their views. I'm wondering if your breast
cancer risk model is an example of that?
Well, it is, you know, it's LED us to look at a number of
different things. So we build models, we use a lot
of different kinds of technology.
(08:01):
What I focus on is not the technology side, but it's what
is the problem that needs to be addressed.
And so we want to start looking at what's the question that not
even the question that's being asked, but what is the question
that should be asked. And so from running the breast
(08:21):
Cancer Center, what I learned was how to work with clinicians,
breast surgeons and and oncologists to gain their
confidence so that why we could do what they were asking us to
do to contribute. I needed to know what they
really didn't know and what werethe questions they wouldn't stop
(08:46):
to ask because that wasn't theirtraining, nor did they think
that they were easy answers or easy ways to get answers.
And so the way I describe it is physicians have to be
operational. They don't have the luxury of
being strategic in general. It's a generalization.
(09:08):
We have the opportunity to be more strategic work to find out
what root cause issues may existand whether they can be
addressed. And then having identified them,
then we look to figure out what's the right technology.
So we don't, we, we use a lot ofdifferent technologies as a
result, but we're really problemfocused.
(09:30):
And there's a there's a saying by Edward Deming, who was an
economist and is, if you don't know what the question is,
you'll never find the answer. And and that's really a guiding
point for how we operate. And so on a path to identifying
breast cancer risks, what might be some of the things that you'd
(09:53):
be looking at? Well, what we're looking at is
the breast is somewhat unique because it's one of the organs
that undergoes a lot of developmental change post
delivery and throughout a woman's lifetime.
And as a result, what we're trying to identify is where are
(10:13):
critical events taking place that may be years before a
breast cancer develops, but our actual early triggers or setting
the path for how that breast cancer may develop.
Because we have been very successful if we look at the
statistics in extending lifetimeof a breast cancer patient.
(10:39):
So the life expectancy after a diagnosis because of all the
effort on treatment has been fairly successful in extending
that life cycle or lifetime. But what we haven't had any
impact on is the incidence of breast cancer.
(10:59):
And so while survival is going up, long term survival is going
up, mortality is going down, incidence has continued to rise.
What we're interested in is understanding that whole process
to understand why incidence is going up and whether there may
be ways to actually intercede much earlier than what we're
(11:23):
doing now, which is the diagnosis as an early stage of
the disease. But could we actually even start
to look at preventing the disease if we know what those
causes are? And so that's how we're trying
to build models of how women progress over their lifetime,
what are the factors of their personal development that are
(11:46):
unique to them, which could be clinical history, it could be
lifestyle, it could be diet, allthese factors.
And then what are external factors that may be important
exposure to toxins, environmental perturbations,
things of that nature. And so we're trying to build
models that allow us to encapsulate all of that data and
(12:11):
then see how women might be identified as higher risk much,
much earlier. And it's more than just
genomics. We're looking, we include
genomics, but it's not purely A genomics effort because we know
genomics is involved, but it's not necessarily the sole driver
(12:35):
in any of this. In fact, most breast cancer when
we talk about it, only 10% of itis typically associated directly
with genomics, 90% is associatedwith PERI and postmenopausal
transition. And that leads us to that
conversation of menopause and obviously breast cancer is one
(12:57):
of the things that is discussed at that time, including many
other things such as neurodivergence and osteoporosis
and all sorts of things where what are we looking at in terms
of understanding the perimenopause to postmenopause
journey? Well, I, I, I use, I, I use the
Rumsfeld term that we're dealingwith the unknown unknowns in
(13:21):
that domain. I think at least from all the
women I've ever spoken with, none of them are satisfied with
the information they're able to obtain from their gynecologists
or obstetricians. And it's not the fault of their
clinicians, but it's not actually taught very much in
(13:41):
medical school. And and that comes about because
we haven't had the data, we haven't had the research, we
haven't had that as an emphasis on understanding it.
It's just always been there and expected to be there.
But if we look at something likeperimenopause, which is anywhere
on average 7 to 10 years in length, it's a very complex
(14:06):
process, that transition and it's going to impact every other
disease that a woman may have inthat 7 to 10 year period.
So understanding it just as a process and how it affects an
individual is important to manage her entire healthcare
(14:26):
profile, not just the transitioninto menopause.
Menopause is actually just a state transition.
It means you haven't had a period for 12 months, but
perimenopause is that process upuntil then.
And if we look at it again from a systems view where typically
perimenopause, it can start muchearlier, but typically it starts
(14:50):
somewhere in the early 40s and and menopause is typically
around 51 on average. Now what that means with given
how lifetimes have increased over the years, a woman is going
to spend almost half of her lifeeither perimenopausal or
(15:11):
postmenopausal. And so not understanding that is
really detrimental to having thebest life experience and
managing everything else that the woman will experience,
whether it's disease or just quality of life issues.
And so that's that's how we've started to try to build profile
(15:36):
like that. I can see that it's a complex
and convoluted area to be working in because every woman's
going to have her own history upuntil that point.
And there's the genetic side of things and then there's the
environmental side of things, and then there's the lifestyle
side of things, all of which aregoing to influence this,
(15:57):
influence the woman's journey through that stage of her life.
And I'm thinking particularly ofthe, the, the data that exists
showing the average age of menopause and how that
difference different differs in different racial groups.
And then for smokers, they go through menopause 2 years
earlier. So there's obviously a lot of
things that are influencing whena woman goes through menopause
(16:20):
and what she goes through. Right.
And, and, you know, in, in general, in our modeling, one of
the key elements that we focus on is that disease or a
condition is a process, not a state.
And so that means we need to understand that women will go
(16:42):
through perimenopause differently.
We suspect that there are probably several common paths
that they may take. Now, that's not going to be an
absolute identical path for every woman.
There'll always be some nuances,but there will be some
commonalities. We think that can be observed,
but we don't have the data to measure that and observe that.
(17:05):
But what that means then is whatwomen tend to do, because as I
said before, they're not gettingthe information they really
would like or the answers to questions they have.
They turn to their friends and networks of friends.
And I mean, it's a very natural way to try to get information.
(17:26):
But as I always tell groups thatI talk to, your hot flashes and
your friends, hot flashes are not the same.
You may both have hot flashes, but they're not the same.
And you know you have it on the there's been identified about 62
different symptoms that are associated with perimenopause.
(17:49):
Not everyone goes through them, of course, but knowing which
ones you're experiencing are going to also impact something
like your hot flashes. And we treat those even as
comorbidities or other conditions along with your hot
flashes, just as an example, because hot flashes affect about
(18:12):
80% of all women. So it's a common experience the
to show you what what we mean. It's very interesting, isn't it,
that the different pathways to, to getting to that state.
And as you say, it's a we, we need to look at the process.
It's not just the state. And I'm also thinking about
(18:34):
treatment now because what you're saying if we have
different pathways to get to hotflashes, for example, that means
that the intervention or the treatment is probably going to
be different too. Is that right?
We would expect that. So this is really personalizing
medicine. This is what personalized
medicine means, at least the waywe think about it.
(18:57):
It requires having accurate, an accurate understanding of the
disease. And, you know, when we're
looking at how women are going through perimenopause, we are
actually mapping out in our models all of the critical
hormonal transitions that take place during development, which
(19:17):
means we are going back at leastto menarche.
And we're doing a very detailed profile of pregnancy histories
because these are the most hormonally disruptive or
variation, you know, periods in a woman's lifetime.
(19:38):
And we need that kind of information to know what her
journey is going to be like. The challenge we have is that
data doesn't exist, and we've scoured a number of
international databases at leastespecially at the level we're
building models because we're looking at, say, for instance,
(20:00):
the the idea that we know that something called null parity is
a risk for breast cancer. Most people think null parity
means you never had a child, butit actually only means that you
never had a live birth. And when we look at menarche and
we understand from the data thatthe age of menarche is going
(20:25):
down, it's decreasing globally and the age of first pregnancy
is going up, that means the period of time in between them
is increasing. That's a period of no parity for
every woman. And so the question is, does
that actually it correlates with, but is it causal for what
(20:48):
we're seeing with the incidence increase of breast cancer?
We don't know. We don't have the data to study
it. And that that's the kind of data
we're trying to now gather through the collaboration we're
setting up with you to work withwomen who are interested in
contributing data, contributing some and learning more about
(21:11):
themselves. Because not only are they going
to have individualized perimenopausal transition, but
that should also indicate they're going to have
individualized risk for breast cancer, for cardiovascular
disease and osteoporosis. And they're not all going to
have the same risk profiles. Yeah, it's, it's such
(21:35):
interesting work, Michael, and I'm really pleased that we're
collaborating to start collecting data.
And one thing that came to mind as you're talking about that is
how important it is to get good quality data and uniform data.
There's so many. It's, it's not just a matter of
collecting different sorts of data from different sorts of
(21:56):
people and trying to cobble it together because that's not
meaningful and it's apples and oranges.
You really need to have a unified collection method and
something that's very aligned and and ubiquitous in a way to
to get that quality that you need.
Well, and, and that's actually part of our research is how to
(22:16):
establish what are the criteria for the data elements you're
collecting because we that we work in other areas besides
Women's Health. Women's Health is the focus of
our nonprofit, and we've seen that challenge of trying to put
data together from different databases and understanding from
(22:40):
work that I've done in the past that different centers may call
something or define something differently.
And So what we're finding nowadays with the push for big
data, which everyone is looking at so they can apply things like
artificial intelligence and machine learning, the data
(23:02):
scientists, the data engineers who aren't familiar with the
nuances of how the clinical practice takes place, are
putting data together based on data labels.
And data labels aren't adequate.You have to actually understand
how those things were done. And you know, the example that I
(23:24):
use is looking at breast cancer.One of the areas that's very
under under served so to speak, is the area of triple negative
breast cancer. And you would think what triple
negative breast cancer means is that you in testing the patient
(23:46):
you find out that their estrogenreceptor level, their
progesterone receptor level and their her 2 levels are all below
the threshold. So they're considered negative
for all of those. That's why they call it triple
negative. But what you also find out when
you work with different centers is what test was administered in
(24:09):
one center may differ from the test administered in another
center, or how they set the threshold for negative may be
different. And the reality is a patient
who's diagnosed triple negative in one center may not be
diagnosed triple negative in another center.
And that presents its own challenge for how to manage the
(24:33):
patient, but that also propagates into drug
development. If you look at the clinical
trials for new drugs that are targeting triple negative, you
find out that even in the in those clinical trials, that kind
of detail is and rigor is not necessarily being addressed.
And there could easily be many drugs being developed today that
(24:57):
because of the lack of rigor anddata collection, are not
actually being evaluated or shown where they could be
effective. And you, you end up with,
through lack of, of standardization, you bring in
heterogeneity into patient populations and it washes out
(25:19):
the effect of the drug. So you don't see that part of
the population where it might work.
And so it, it's not just that's designation and and the I CD10
code or something like that, butit goes through the entire
healthcare ecosystem. And it really dilutes what's
(25:42):
happening and it and it masks things.
Yeah. So then that's there.
Our collaboration is really important because your goal is
to collect data globally for a study that's going to start to
really look at a more personalized approach for
Women's Health. Yeah, yeah.
So again, we've extracted methods.
(26:08):
We borrow methods from other disciplines, as I said, because
we're focusing problem solving. But then as we look at the
methods, we see what needs to beadapted or expanded.
So one of the methods we use is called a knowledge graph.
It's a form of artificial, It's in that area of artificial
(26:29):
intelligence. But we've expanded the the
approach in a novel way to look at a lot more rigor so that it's
not just a data warehouse, but something that can be made
actionable in the clinic. But in doing that, these kinds
of nuances, for instance of knowing where the test, what
(26:51):
kind of test was done, how the threshold was set, these are
very detailed characteristics that we are building into the
knowledge graph. And then when we build the study
or the survey used in our study,we carry that over and apply
that to ask the questions the correct way to be able to know
(27:15):
what those unique differences are.
And that that that is very critical to being able to
actually use the data we would like in the ways we would like
to, to build models. The other kind of question that
is important, I pointed out thatthere's 62 symptoms that are
(27:36):
identified or associated with perimenopause.
And many people, many women havehad surveys or participated in
surveys from different organizations or or government
policy, whatever, where they have a checklist to check off
what they've actually experienced.
(27:59):
And of course, these are also unique to individuals.
But it's not enough if we're using our approach that says we
need to understand the process, to just have a checklist of what
you experienced. What we're asking is when did
you experience it? In what order did you experience
(28:21):
it, and what did you experience at the same time?
And those are the factors that will help identify an
individual's journey, not just the general characteristics.
And that's the kind of data thatwe build into the model and then
into the surveys or studies thatwe're doing.
(28:43):
It's such exciting work, Michael.
I can see that this is going to really be a game changer for
Women's Health and for healthcare generally.
It's going to lead for more personalized medication and
other treatments. It's going to help us to
understand contraindications better for medications and other
treatments. And is 1.
It feels like A1 size fits most at at the moment.
(29:06):
Well, yeah, I mean, but what youwhat one of the things we would
like it to be able to do is be able to provide it for a woman
some guidance to know if what she's experienced experiencing
is part of that normal transition or not, rather than
(29:29):
not knowing and and not having agood resource to talk to to find
out. A common a common occurrence
during perimenopause are some cardiac events, small events.
But instead of running to the cardiologist who may not even
anticipate as part of perimenopause, if you know that
(29:51):
this kind of event actually occurs in your transition,
you're going to feel more comfortable with that and not
not immediately panic, so to speak, that having, you know, a
much more serious condition or if you're experiencing something
that isn't necessarily part of that transition, then you know,
(30:14):
it is something that needs additional attention.
So giving the individual, givingeach woman more control over her
own experience and knowing how to interpret it.
Yeah. And it's such an important point
that you make about that, the knowing and the sense of control
(30:36):
around that and sense of confidence or, or greater
certainty that comes with that. Because too often we're speaking
to women who don't even realize that they've been in
perimenopause for the past 2-3 or four years, which means
they've missed an opportunity for early intervention, for one.
And it means that they've been struggling with symptoms and
(30:57):
thinking that they're going mad too, and then going to doctors
or other healthcare practitioners and not getting a
diagnosis. So there's this fear coming up
around what's actually happeningto me.
And I've heard women say, I wonder if I'm getting early
onset dementia, for example, it's a real concern.
So to have to have your work available, to have that data
(31:22):
collection, to understand your own unique risk profile, your
own opportunities for early intervention, that is really
important and powerful. And we're so pleased to be
partnering with you to share this data collection, study and
initiative. With the people, yeah, right.
(31:43):
We, we think it, it's well the, the goal is running it through
the non profit. And the reason we're doing that
is having been involved in commercial operations and pharma
and investment small groups and so on.
The, the need to commercialize something can sometimes redirect
(32:09):
how the study is done by keepingit in our nonprofit world, we
believe we're doing it the way we think it should be done and,
and not trying to take certain shortcuts or, or directions that
that may not give us the comprehensive knowledge we're
(32:30):
trying to get out of this. So that's one of the reasons
we're we've moved it into the nonprofit.
We have other aspects of Women'sHealth that we're tying into
this, of course, because we're also looking at hypertensive
disorders of pregnancy and preeclampsia on their own, but
(32:51):
then also how they impact even perimenopause much later on.
So none of these things we thinkare isolated.
It's really a continuum of looking at normal development
and how these presentations are influenced from different
factors. You know, we, we actually in the
(33:14):
US have been running some publicforums for women in association
with some of the professional women sororities in particular.
And it's just to allow women to come in and ask questions and
you know, they're they're there aren't really those
opportunities. I know you're through your
(33:36):
organization providing coaching and and that's a real key factor
that provides some comfort I'm sure to a lot of women being
able to rely find a reliable source of resource.
But you know, we think this kindof information, we're collecting
it in the US, we're in discussion, collecting it in
(34:00):
Africa and in Chile as well. And so we think this kind of
information and being able to docomparative analysis as well as
going to at least open up new doors.
I never tell anyone we're going to solve the problem, but we're
going to make things much more transparent than they are.
(34:22):
Yeah, it's really, really powerful work.
I love that you're exploring different countries and
different populations and aimingfor that broader transparency.
And I mean, certainly in Australia, we realize that there
are underserved populations here, remote populations,
regional and rural groups calledwomen, culturally and
(34:43):
linguistically diverse. And we know nothing about their
experiences and their life history and how that might
influence their journey into chronic disease or not in later
lives. So we're really excited to be
getting this work out and we'll be launching this in June for
anybody that's interested in partnering with us to share the
(35:04):
the work and get women engaged to contribute information and
help you with your data collection and study.
That's that's the goal for us. And I'd love to finish on this,
Michael, with understanding whatwhat your goal is for this year
with with this work. Well, we've done preliminary
studies using UK Biobank, the US, all of US studies and so on.
(35:31):
But, but we need to actually fill the gaps and the gaps are
the data we're trying to collect.
So what we'd like to be able to do and, and one of the the ways
we've set this up using it as anonline anonymous first pass data
collection is exactly to addressthe fact that you could be
(35:54):
rural, you could be you, you canaddress it on a telephone, a
phone and answer the questions. So being able to link in from
anywhere. It's not something that requires
an academic setting or going into your doctor's office.
In fact, it doesn't require anything that you should not be
(36:16):
able to address on your own without looking up records.
But our goal then is to to get alarge enough population that we
can show the feasibility and then we really have a much
deeper set of questions we'd like to ask, but that will be
(36:38):
much more involved. This first pass is what we want
to accomplish this year, collecting this first set of
data. We look forward to helping you
with that work, Michael, and launching this in Australia in
June. It's really exciting and I will
probably come back to you later in the end for another update on
(36:59):
how things are going and what you've learned.
And thank you for your time today.
Thank you, Looking forward to continuing to work together.
It's amazing. We're really thankful for the
opportunity to partner with you.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Special Summer Offer: Exclusively on Apple Podcasts, try our Dateline Premium subscription completely free for one month! With Dateline Premium, you get every episode ad-free plus exclusive bonus content.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.