Latest / 'Bone Up' / BoneUP Live @ ASBMR 2026 - Part 3

Transcript

BoneUP Live @ ASBMR 2026 - Part 3 · 'Bone Up'

0:00
Hello and welcome to Bone Up, the podcast all
0:04
about bones. I'm David Armstrong. Hi and I'm
0:07
Richie Abel and together we're going to bring
0:10
you the latest in bone research. So whether you
0:12
are a patient living with bone disease, a family
0:15
member, a healthcare professional or a scientist,
0:18
you've come to the right place. So if you want
0:20
to learn more about our infrastructure of calcified
0:24
collagen, this is Bone Up. Listeners, hello and
0:52
welcome back to another special episode of Bone
0:54
Up. We're here at the ASBMR conference in Boston
0:58
and we're going to bring you another flavour
0:59
of the talks, presentations and discussions that
1:02
are going on. We hope you enjoy these interviews.
1:05
We really enjoyed recording them. Listeners,
1:18
I'm really, really delighted to introduce our
1:21
next guest to you, Professor Emeritus Tony Keaveney
1:24
from the Departments of Mechanical and Bioengineering
1:27
at UC Berkeley and Founder and Chief Science
1:30
Officer at ON Diagnostics. Thank you for joining
1:33
us today. Thanks, Richie. Thanks for making the
1:36
time to come and speak to us, Tony. I'm very
1:39
busy, yes. You are! My pleasure to be here. Oh,
1:44
thank you. So we're always interested to hear
1:47
about the possibilities of new diagnostics and
1:50
new screening tests, because obviously the big
1:52
problem we have at the minute is we've got loads
1:54
of really good drugs for osteoporosis, but not
1:56
enough patients are getting those drugs. And
1:59
your company has some new tools and devices to
2:01
address that problem. So we developed a test
2:05
that we call biomechanical CT, or BCT for short,
2:09
and it is... An analysis of data in a CT scan.
2:14
So if you have an existing CT scan, say, of your
2:18
hip or spine that you got for any reason, we
2:22
can repurpose that scan and do a very detailed
2:25
analysis of your bone. And so the test measures
2:27
two things. First off, it measures bone mineral
2:30
density. So for the hip, you get the same bone
2:35
mineral density T -score as you would get from
2:37
DEXA. And then for the spine, because... DEXA
2:41
is rather limited for the spine because of the
2:45
degenerative changes that can occur in the posterior
2:47
elements as you get older. We use what's commonly
2:51
used in quantitative CT, and it's a measurement
2:54
of trabecular bone mineral density, and that's
2:57
a proper volumetric measure. Both of those come
3:00
with well -established cut points for identifying
3:03
osteoporosis. So if you're at the hip, your T
3:05
-score is minus 2 .5, you've got osteoporosis.
3:09
And then at the spine, if your trabecular bone
3:11
density is less than 80 milligrams per cc, you'd
3:14
be defined as having osteoporosis. So those are
3:16
very well -established guidelines. So we provide
3:19
that. On top of that, then, we do an engineering
3:23
structural analysis of your bones. We use a technique
3:27
called finite element analysis, which is pretty
3:30
much the gold standard structural analysis technique
3:33
in engineering. So cars, buildings, even microchips.
3:38
They're all designed for structural integrity
3:41
using finite element analysis. And what you do
3:45
is you create a virtual model of your structure.
3:49
In our case, it would be the proximal femur or
3:52
a vertebral body from the information in the
3:55
CT scan. And then we use this computational mechanics
3:59
technique to virtually stress test the bone.
4:02
And so we put forces on it and virtually... crush
4:05
it. We compress the spine to simulate a compressive
4:09
overload and then we load the hip to simulate
4:13
a sideways fall because that's the most common
4:15
configuration for hip fractures. Hip fractures
4:19
typically occur after a fall. So the test then
4:23
gives you a measurement of your breaking strength
4:25
of the femur for a sideways fall and of one of
4:31
your vertebrae for a compressive overload. And
4:33
if those strength measurements are less than
4:36
thresholds that we developed, then you'd have
4:39
what we call fragile bone strength, which is
4:41
equivalent to having osteoporosis. So overall,
4:44
then you're getting from the test, bone density
4:46
measurements, bone strength measurements, each
4:49
come with cut points to classify osteoporosis
4:51
or what we call fragile bone strength. And if
4:54
you're in any of those two categories, you would
4:56
be defined as being at high risk of fracture.
4:59
So that's how it's used clinically. It is all
5:03
cleared by the FDA, so it's indicated to be used
5:07
to assess fracture risk, identify osteoporosis,
5:11
or monitor treatment. And then in the United
5:14
States, it's also reimbursed by Medicare. So
5:18
it's in the same reimbursement category as DEXA.
5:21
So patients don't have to pay for anything as
5:24
a preventive service. It sounds it's supplying
5:27
that. big gap we have at the moment because I
5:30
can measure people's bone density and Richie
5:33
can maybe use some tools to measure their bone
5:36
quality but the one thing we can't do is actually
5:38
stress those bones. We can't push the patient
5:41
over to see if their hip is actually going to
5:43
break. In some ways that's what you're doing.
5:45
You're supplying that extra force. It's a real
5:48
world. Right. It's simulating the event that
5:54
is associated with fracture. Bone density is
5:58
what you'd call a surrogate for bone strength,
6:01
right? People want to know about your bone strength.
6:05
And so bone density is related to bone strength,
6:08
no question about that. But what the BCT analysis
6:11
does is it just provides our best engineering
6:15
estimate of what the strength is. And so because
6:22
of the way the cut points work... The test will
6:27
identify a subset of patients who are in the
6:32
low -density category, but not as bad as osteoporosis,
6:35
who will have the very low level of strength.
6:39
And that discrepancy occurs typically because
6:43
some folks have different distributions of trabecular
6:47
and cortical bone. They've got a weird shape.
6:49
They've got very small bones. The neck is at
6:52
an unusual angle. So there's structural features
6:54
of the bone that mean even if you're in the low
6:59
bone density category, not osteoporosis, you
7:02
still can have very weak bones, right? So we
7:05
have another test that we've just developed that's
7:07
not available for clinical use yet, where we
7:10
measure also from a hip scan muscle fat content
7:15
and muscle size, and we measure the soft tissue
7:19
padding around your hip. and how wide your hips
7:24
are and how big your bones are. And what we're
7:25
trying to do in that test is address the whole
7:28
biomechanics of hip fracture. So in theory, if
7:33
you fall and if your bone strength is less than
7:40
the magnitude of the impact force that occurs
7:43
when you fall, then you'll fracture. So if you
7:46
don't fall, you're not going to fracture. And
7:48
if the impact force... isn't enough to exceed
7:52
your bone strength, you won't fracture. So that's
7:55
very simple kind of etiology of a hip fracture.
8:00
And so what we're doing with this newer test
8:03
is the muscle measurements are associated with
8:06
risk of falling. The soft tissue padding around
8:10
your hip is cushioning, so people with a lot
8:12
of soft tissue around their hips are protected.
8:15
The size of your skeleton is... reflective of
8:19
the impact force the bigger you are the harder
8:21
you'll fall the taller you are the bigger your
8:23
bones are the harder you'll fall large bones
8:26
are actually associated with an increased risk
8:29
of hip fracture which is kind of you're thinking
8:32
how can that be but it's because of the role
8:35
on the impact force so we put all that together
8:37
and that gives us what we call a risk score so
8:41
that's the next generation that's coming up and
8:43
that is one where your body size etc come into
8:49
the picture with the bone strength measurement
8:51
by itself you're just looking at one part of
8:53
all of that right so the future is the future
8:56
is this next test you know it's putting all the
9:00
the other factors together that that play a role
9:03
in in the actual etiology of the fracture do
9:06
you think your technology might redefine the
9:09
whole of the bone field So at the minute, you
9:12
know, we talk about osteoporosis and everybody
9:15
understands it as this T -score minus 2 .5, but
9:17
we know a lot of people who fracture don't have
9:20
low BMD and vice versa. It feels like for a long
9:24
time we've needed a new definition of the disease,
9:27
and I know that who are working on something
9:28
at the minute, something about bone fragility
9:31
rather than just osteoporosis, and your technology
9:33
is going to be able to basically calculate if
9:36
you fall over, are you going to break? That's
9:38
like the ultimate. ultimate definition of bone
9:41
fragility yeah i mean of course it's just an
9:45
estimate of everything that's happening right
9:47
so not nothing the it's definitely filling out
9:53
filling out the assessment like right now fracture
9:56
risk assessment is done for example with fracs
9:59
so what you have there is you've got bmd and
10:02
then you put age onto that that's a huge factor
10:05
and then you have these other clinical risk factors
10:07
right that are associated with Basically frailty.
10:10
You know, if you've got rheumatoid arthritis,
10:12
if you're a heavy smoker, if you drink a lot,
10:14
all of those things are either affecting your
10:17
physiology or maybe, you know, reaction time,
10:20
all of those things. But they're very indirect
10:23
surrogates of the things that are associated
10:28
with the fracture. So, yes, we're measuring,
10:32
you know, physical traits. And it's not an AI
10:35
thing where it's just... pattern recognition
10:37
of the image. We're measuring something about
10:41
your muscle, the muscle fat content, the muscle
10:43
size. Those things are very easy to understand,
10:46
directly associated with elements of your physiological
10:51
function that are associated with hip fracture.
10:53
So yes, it's definitely getting you closer. closer
10:58
to where you want to go. But the reality is at
11:00
the end of the day, still, it's very hard to
11:02
predict who's going to fall or the way you're
11:04
going to fall. So hip fracture prediction is
11:07
always going to be imprecise. So even though
11:10
this new test has very good sensitivity, 80 %
11:13
sensitivity for predicting hip fracture within
11:16
the next five years, which is fantastic, but
11:19
about 20 or 30 % of people will. nonetheless
11:24
test positive. So it's not that you're saying,
11:27
oh, it's a super specific test. And that's because
11:32
to capture the folks who are fracturing, because
11:36
there's so much uncertainty about, for example,
11:38
falling, you end up with people testing positive
11:43
who would fracture if they fell, but they're
11:48
not going to fall. And you can't do that, right?
11:50
So that's the challenge. On a purely practical
11:53
level then, you're saying to me these are people
11:55
who've had CT scans already done for another
11:58
reason primarily. Are there people then that
12:01
you will recommend they have a CT scan done just
12:05
so that you can analyze it? Or is that part of
12:07
the package? Yeah, so the test works on both.
12:10
And right now, our strategy thinking about it
12:18
is let's at least go after the folks who already
12:21
have a CT. There's no radiation then for a new
12:25
-bound test at all. So all of that concern about
12:29
using CT scanning at the population level, that
12:32
doesn't apply. It's zero radiation because you've
12:34
already taken the CT. And it's very convenient,
12:37
right? And you don't have to go anywhere and
12:39
you don't have to wait three months or whatever
12:41
the line is. So that's that, right? There's no
12:45
question that there's elements of all of this
12:48
that are better than Dexa, right? Yes, I think
12:52
there are applications where it would be perfectly
12:55
fine. The new test works really well. So, you
13:00
know, would that justify going out and getting
13:04
a CT? Maybe for some people. But to start with,
13:08
let's focus on folks just because of the convenience
13:12
and the radiation concerns. And we'll see where
13:15
it goes. Today, for example, if I was having
13:21
a spine surgery, me personally, I'm not giving
13:25
medical advice, if I was having a spine surgery,
13:28
I would want a CT scan of my spine and I'd want
13:30
my test done. Because the bone quality issue
13:32
in the spine, it's super important for that application.
13:36
And if your spine surgery goes south, that's
13:39
a really bad... sequence of events that's going
13:42
to unfold. So I don't want to take any chance
13:44
on that. And Dex, it does not work well for the
13:46
spine, especially if I'm going to need surgery
13:48
for the spine. There's degenerative changes.
13:50
Dex is just not good. So absolutely in that case,
13:55
I would say get a CT scan and get BCT. So I think
13:58
there are definitely some clinical applications
13:59
right now that it's unequivocal. This is well
14:02
justified from a medical perspective and risk
14:08
and benefit. So you mentioned that it's been
14:12
FDA approved for Medicare. Once somebody has
14:15
the BCT scores and they've been diagnosed as
14:19
being fragile, is that enough to begin treatment
14:21
or do people have to go through older, more traditional
14:25
assessments? Yeah, so... The ISCD, International
14:29
Society for Clinical Densitometry, they have
14:33
practice guidelines. And so their practice guidelines
14:36
say, yes, if bone strength by finite element
14:38
analysis that's been validated, they recommend
14:43
that's a way to initiate therapeutic treatment.
14:46
So that's one society that's come out and said
14:47
that. Otherwise, most of the practice guidelines
14:53
don't even know we exist. So you're not going
14:56
to see it in the ACE guidelines if you go and
15:00
look there. But the feedback we have from people
15:04
who are using it clinically is absolutely. It's
15:06
on the basis of either of those people will treat.
15:12
That's amazing. That's really amazing. Where
15:16
did the idea come from originally? How did you
15:18
get started? Well, doing final analysis on...
15:25
Bones, people were doing that before I got around
15:30
to it. But I did my PhD at Cornell, and my PhD
15:36
was doing design of hip implants. And the tool
15:41
was finite element analysis. It was a whole computational
15:44
PhD project. So that's where I developed my expertise
15:47
in finite element modeling. And it was applied
15:51
to bone implant systems. So I was all about...
15:54
the implants then. But then I ended up doing
15:57
a postdoc in Boston with Toby Hayes, who had
16:01
a big lab focused on osteoporosis. And that's
16:05
where I learned about osteoporosis. He had actually,
16:08
some of his students had been doing final analysis
16:10
of bones. So when I got there, there had been
16:15
a little bit of earlier work done on that. But
16:17
then, and I actually focused most of my time
16:21
there on experiments. When I got my faculty job
16:25
at UC Berkeley, I had to set up my own lab. Then
16:29
I pulled everything together and said, okay,
16:30
let's get serious about the finite element modeling.
16:33
So we did very, very detailed finite element
16:35
modeling then. We were doing micro -CT -based
16:40
finite element models before there was micro
16:42
-CT. We used to embed pieces of trabecular bone
16:46
in acrylic. And layer by layer, grind them down
16:51
and take a picture at every single layer and
16:53
make up. So we were doing it. That was before
16:56
there was micro CT. And we would do final analysis
16:58
of that. So it was hardcore. There was a million
17:01
elements in the model. This is really high -end
17:03
computation. I was working with my colleagues
17:05
who could do this massively parallel computation.
17:09
But we also tried, as part of the research we
17:14
had, we would be breaking bones. And we would
17:16
do micro CT of whole things. Imagine doing that
17:19
thing for a whole now for T -Body. So it was
17:21
wacky stuff we were doing. But we also did some
17:25
clinical, we got some clinical CT scans taken.
17:27
And we did, because we were able to, we did finite
17:31
element modeling of those. And we found that
17:34
when you take the approach we currently use,
17:37
which is to make these models directly from the
17:40
image pixels into finite elements, so this so
17:43
-called voxel -based approach. We thought that
17:46
was a little crude because, you know, it's not
17:48
all curved side elements. We said, let's just
17:51
try it. Joyce Kiak at UCSF at the time had been
17:55
doing something like this. And it worked really
17:58
well. And so I was at a conference at NIH and
18:03
one of the medical directors from Eli Lilly was
18:07
there who was making Forteo, Eric Erickson. And
18:10
he came up to me afterwards and said, you know,
18:12
we want that. technology for our drug trials
18:15
you should start a company wow and so this is
18:19
2004 2005 and me as naive as you can imagine
18:24
said sure i'll start a company no idea what i
18:27
was getting into and so that was 20 years ago
18:29
over 20 years ago and so that's we we started
18:33
then very small obviously but it's just taken
18:36
such a long time to develop the technology Do
18:41
the clinical studies, the fracture outcome trials
18:43
to support everything. One of the things I didn't
18:45
mention in the description, one of the interesting
18:47
things we do also is when you do these quantitative
18:50
analysis of CT scans, you have to be careful
18:53
how you calibrate the scan. The grayscale values
18:56
in the scan itself aren't quite good enough to
18:58
be quantitative. So normally you patient, when
19:02
they're getting a quantitative CT, they lie on
19:05
a phantom. calibration chamber. And that's used
19:09
to standardize the grayscale values to calibrate
19:11
the scan. And that's a hassle, right? A lot of
19:14
places don't have these calibration phantoms.
19:16
They can deteriorate over time. It's just a hassle.
19:18
So we figured out a way to do this using your
19:22
internal tissues. So that was done along the
19:25
way. And that was driven by the need to try and
19:28
repurpose existing CT scans. We said, look, we
19:30
have to try and repurpose the unit because it's
19:32
going to be hard to get people to go out and
19:34
get new CT scans. Right off the bat, could we
19:37
do this? So that phantom miscalibration, that's
19:42
an important part of all of this too. Have you
19:44
been looking just at postmenopausal osteoporosis,
19:47
or is there a role, for example, in steroid -induced
19:50
osteoporosis and predicting fracture, which is
19:53
another big group of patients we see? Certainly
19:56
what happens with steroids is you can get...
20:02
Big changes in the cortical trabecular bone make
20:05
up. And that's perfect for a fine element analysis
20:07
because it doesn't care what's in the bone. It
20:11
just gets built into it. So we can certainly
20:15
analyze that. But the framework, you would still
20:18
at the end of the day, you'd say, okay, well,
20:19
is your bone very weak? So you don't have to
20:22
develop new criteria. Because it's all on the
20:28
same strength scale. It's all included in there.
20:30
Outcome, really, whether it's postmenopausal
20:32
or steroid -induced or caused by aromatase inhibitors,
20:35
the outcome is sort of broadly similar in terms
20:39
of what you can measure. Right. And what's nice,
20:42
too, we've used it a lot to assess drug treatments.
20:45
All the drug companies worked with us over the
20:47
years. And, yeah, you see some interesting things,
20:49
cortical changes, trabecular changes with drug
20:51
treatment. And, of course, DEXA doesn't pick
20:53
that up because by design it just averages everything.
20:55
So there are some, like, PTH is interesting.
20:58
It really boosts the trabecular bone, but the
21:00
cortical bone can get left behind for a little
21:02
bit. And, you know, so you get strange results
21:05
with DEXA because... If there's increased cortical
21:08
porosity, that cancels out. Yeah, it looks as
21:11
if the BMD is falling, even like at the hip,
21:14
for example. So the technique is very nice for
21:19
all of that. It's kind of agnostic, if you will,
21:21
to the patient's condition. The caveat, of course,
21:24
is that it's based on information in a CT scan,
21:27
and I can't tell you about the molecular makeup
21:31
of the collagen and the minerals. So if you've
21:33
got somebody and they've got a... a rare bone
21:36
disease where you know the collagen is messed
21:39
up or you've got some really weird stuff going
21:41
on there um you're not going to pick that up
21:43
so that you've always got but that's true if
21:45
decks are scanning and the way you think of it
21:47
is this is this is your bone strength assuming
21:49
your bone tissue is just normal yeah Increasingly,
21:55
we are going to be able to get that molecular
21:56
information by doing genetic tests and studying
21:58
mutations in the collagen maybe further down
22:00
the line. There could be loads of really nice
22:02
applications where you can program in bone material
22:05
properties based on the mutations. Yeah, I mean,
22:08
you could say... And we've actually done a couple
22:10
of studies like that. We're based on animal experiments
22:12
in that we said, OK, in this condition, osteogenesis
22:15
imperfecta or whatever, we think the tissue has
22:18
got about 50 % of its strength. So we can program
22:21
that in... to the final model. Yes, that can
22:24
be done. You know, so that's kind of an interesting
22:28
thought. We should talk about that. We've got
22:32
some good stuff in my lab, I think. Thank you
22:36
so much for coming to talk to us today and making
22:38
the time. Thank you. Thank you both. I appreciate
22:41
that. It's been a huge, huge pleasure. And yeah,
22:45
you're a Dublin man originally? Of course. Of
22:47
course. Dublin. Yes. You can't tell from my accent,
22:51
though, right? No, it's been a great pleasure,
22:56
and thanks for making the time. Because I know
22:58
we had to sort of work our way around you, but
23:00
we found a window to speak to you, so that's
23:01
great. It's my pleasure. I love talking. Tell
23:07
me all. Thanks, Tony. So our next guest is Olivia
23:14
White from Purdue University, and you're a PhD
23:19
candidate. Correct. Isn't that right? Congratulations.
23:22
Thank you, thank you. And you've been doing some
23:23
very fascinating and pertinent work on, well,
23:26
you tell us what it's about. It's about alcohol
23:29
and the effects of alcohol on bone? Yeah, so
23:32
right now my work is focused on evaluating the
23:34
outcomes, particularly skeletal and metabolic,
23:37
in a mouse model of type 2 diabetes and alcohol
23:40
use disorder. So both those diseases alone and
23:44
then in combination to see if when they come
23:46
together, are they double trouble or does it
23:49
really matter? And because we were talking a
23:52
little bit this earlier, so you have mice, and
23:54
these are mice that are particularly bred to
23:56
be prone to... Taking more alcohol, is that right?
24:00
Explain how we do that. Yeah, so we've actually
24:02
worked with mice in the past that they're called
24:04
cross -high alcohol preferring mice. And we were
24:07
able to use those mice because of a collaborator
24:09
named Dr. Graham. So he was very instrumental
24:12
in that work. And the way that those mice were
24:15
actually bred is that they're selectively bred
24:17
based on their alcohol consumption. So just depending
24:21
on which mice are drinking the most of the ethanol
24:24
that is provided, those are the mice that are
24:26
bred and they just do that for generations. to
24:29
actually acquire the high -alcohol -preferring
24:31
mice. But right now, the mice that we're focused
24:34
on working with in that kind of combined model
24:36
of diabetes and alcohol use disorder are just
24:38
standard C57s. How do you induce the diabetes
24:43
in those mice? Yeah, great question. So there's,
24:46
of course, a lot of ways to induce diabetes in
24:48
mice, but the way that we have found to be most
24:50
effective for specifically kind of a more type
24:53
2 diabetic phenotype, where you're looking for
24:56
obesity, insulin resistance, and hyperglycemia,
25:00
so high blood glucose, is we use a high -fat
25:03
diet, so 60 % fat diet, and we give that for
25:05
12 weeks, and then we give a single shot of STZ,
25:09
or streptozotazine, which actually will kind
25:11
of knock out a few of those pains. pancreatic
25:14
beta cells that are responsible for producing
25:17
insulin. And then we age them out for another
25:19
eight weeks. So you've selectively bred mice
25:23
to be alcoholic, and then you can give them a
25:26
diet which induces diabetes. So right now we're
25:30
just using the in -bred strain, but we have worked
25:32
with the selectively bred strain in the past.
25:35
What kind of experiments are you running? We
25:37
typically, right now, the outcomes that we're
25:40
looking for in these mice are measures of glucose
25:43
tolerance. So we'll do GTTs, glucose tolerance
25:46
testing. We'll do DEXA scans. So we kind of do
25:49
those longitudinal measures throughout the study.
25:52
And then once we actually have sacrificed the
25:54
mice and harvested them, we'll... typically collect
25:57
femora as well as tibia and we'll start doing
26:00
micro CT scanning and then we are a bone biology
26:03
mechanics lab so those mechanical outcomes are
26:05
really important to us so we'll typically move
26:08
into some form of bending whether it's three
26:10
or four point bending tests to failure to understand
26:12
kind of the whole bone mechanical outcomes and
26:15
then from there we can get into more bone quality
26:18
measures you know what's the composition of the
26:20
bone how is that changing as well as kind of
26:22
the micro scale mechanics. And you'll be able
26:24
to look at the effect of alcohol separately from
26:26
diabetes separately and then the combination
26:28
is that the thinking? Correct yeah so that's
26:30
actually what the work I'm presenting at ASBMR
26:32
is over is it's a model of alcohol use disorder
26:36
type 2 diabetes and then the combination of the
26:39
two and then of course we have a control group
26:41
that we're comparing back to. And it's a very
26:43
complex area because our understanding at the
26:45
clinic patients I see is that diabetes certainly
26:48
seems to increase the risk of fracture but it's
26:50
complex because it's not necessarily related
26:52
to BMD the way it would be normally. our understanding
26:56
is that people who have alcohol addiction certainly
26:58
seem to have an increased risk of fracture, but
27:00
how closely related to BMD that is, is also complex.
27:04
What sort of thing have you been finding? What's
27:06
been coming out of your work so far? Yeah, so,
27:09
so far, and again, this will be on my poster
27:12
tomorrow, what we've found is that in... terms
27:15
of BMD from those longitudinal DEXA scans, when
27:18
we actually got to the end of the study and did
27:20
that final scan, that combined group did have
27:23
the lowest bone mineral density measures out
27:26
of all of them. So even in our type 2 diabetes
27:29
model, we do see a little bit of a decline in
27:31
the BMD, but I would say even more so in the
27:33
alcohol use disorder. And then when you combine
27:35
the two together, it's even lower than either
27:38
disease state alone, which could imply that.
27:41
they're worse off when they are together. And
27:43
do you think it's that the diabetes and the alcohol
27:47
have completely different effects in combination?
27:50
Or do you think that the two conditions might
27:53
be acting through the same pathways to sort of
27:55
potentiate each other? That's a great question.
27:58
I honestly think it's a combo of the two because
28:01
they do have... pretty unique outcomes on their
28:05
own, but there are areas where they overlap.
28:07
So that's where my work is hopefully going to
28:10
go next is looking at one area of overlap that
28:12
they share, which is inflammation. So both of
28:14
those diseases can increase systemic inflammation
28:16
as well as kind of in the bone marrow. So that's
28:20
kind of my hypothesis is seeing if that's actually
28:23
what's driving that reduction in BND that we
28:26
see in that combination model. And it sounds
28:27
as if you're interested in the quality of the
28:29
bone and you talked about the resistance to bending
28:32
and that sort of thing. as well as just the BMD.
28:35
So that's certainly an interesting area. Yeah.
28:38
Yeah. And we have seen similar outcomes. You
28:40
know, when we do micro CT, we're looking at those
28:42
micro architectural outcomes of the bone and
28:44
the cortical and trabecular. And we see declines
28:46
due to each disease state alone in those outcomes.
28:50
But of course, that combined disease state also
28:52
does have the lowest measures out of all of them.
28:55
And we see that in... the microarchitectural
28:58
measures as well as the bending measures, but
29:01
still need to get into the quality assessment
29:02
of those bones. And I suppose, again, thinking
29:05
at the clinic, the next question I would ask
29:09
or my patient would ask is if I address my alcohol
29:12
addiction or if I improve my diabetes, does that
29:15
improve my BMD? Does it improve the quality of
29:19
my bone or does it reduce my risk of fracture?
29:21
Are you able to address that in your mouse model?
29:25
We could it isn't something that I've thought
29:28
of but it's a wonderful point. There are studies
29:30
other preclinical studies that have looked at
29:34
can bone recover whenever alcohol is taken away
29:36
in just that single disease state? And studies
29:39
have found that it can with time, but in the
29:42
combined disease state, we haven't approached
29:44
that yet. So that's an excellent question. Because
29:46
presumably you can't stop giving the mice alcohol
29:48
in a controlled way, which is harder with human
29:51
beings. Yes, absolutely, yeah. We've also kind
29:54
of thought about going into more of a pharmaceutical
29:56
treatment route. So like glucagon -like peptide
30:00
1 receptor agonists, those can, of course, improve
30:02
type 2 diabetes. outcomes but they've also found
30:05
that in certain studies they've been shown to
30:07
kind of reduce alcohol consumption so that's
30:09
always been a thought in the back of our heads
30:12
depending on how this work kind of comes to at
30:15
the end of the study is is that an option for
30:17
treatment is it something we could explore in
30:19
our model that's absolutely fantastic i i guess
30:22
again you can start to think about combined effects
30:24
what is the effect of a treatment What is the
30:27
effect of cutting down on the alcohol? And what
30:29
is the combined effect of the two of them? Because
30:31
I think we've discussed it before. Sometimes
30:33
if patients are taking a medicine, then they
30:35
maybe don't want to change their lifestyle as
30:37
well. They think, well, I'm taking the medicine
30:38
for it. I can just carry on. Exactly. Yeah. And
30:41
I mean, that's what makes it so complex. In mice,
30:44
we can control that. You know, we can isolate
30:46
every variable as best as we can and see how
30:48
they're kind of each impacting skeletal outcomes.
30:51
But how it actually translates to the clinic,
30:53
I think, is a lot more difficult. What I feel
30:56
is so important, so many big studies and so many
30:59
big pharmaceutical studies for understandable
31:01
reasons, they will exclude patients with alcohol
31:03
addiction. They will exclude patients who are
31:06
heavy smokers. They will exclude patients who
31:07
have uncontrolled diabetes. But then when I see
31:10
the patient at the clinic and when my primary
31:11
care colleagues listening to this podcast see
31:14
their patients, then the patients they see are
31:16
the ones who have all those other problems as
31:17
well. And it's hard to maybe apply some of the
31:20
bigger studies to a group of... patients who
31:22
are completely different so it's really great
31:24
that people are addressing you know the real
31:26
world in mice as it turns out. Yeah absolutely
31:30
and I mean to the point with the studies kind
31:33
of lacking this patient population that's something
31:36
I've struggled with when getting all of the info
31:38
to back this research up is a lot of studies
31:41
aren't really focusing on what are the actual
31:43
combined effects especially in a patient that's
31:45
kind of in that worse off disease state of the
31:47
two. So, yeah, trying to address that and keeping
31:50
that at the forefront of our minds is the patient
31:53
population and how we can help them with our
31:55
mice. So, listeners, we're in Boston. It's cheers
31:58
for mice. Yes. He's been waiting the whole podcast
32:02
to say that. You've got to get it in once. Thanks
32:05
very much. Yeah, thank you, guys. Thank you.
32:07
That was so beautifully explained. Thank you
32:09
for sharing your work with us. Thank you. Thanks
32:11
for having me. So, listeners, our next guest
32:18
is Gabrielle Dudek, who is a PhD. student from
32:24
University of Colorado Anschutz and we were very
32:28
interested to see your work or see your poster
32:31
yesterday and you're looking at a rare condition
32:34
called homocystinuria. Tell us a little just
32:37
about how that affects patients. Yeah so homocystinuria
32:41
is a disorder of sulfur metabolism so patients
32:45
end up with a lot of homocysteine residues they
32:49
can't properly process it. Patients have a deficiency
32:54
or an abnormality in the methionine pathway,
32:58
and so there are many variations to this disease.
33:00
It's on a big spectrum, but they can have a genetic
33:04
disorder of any enzyme along that methionine
33:07
pathway that helps break down methionine and
33:10
homocysteine down to cysteine. This causes a
33:13
lot of issues. On the non -bone side, it causes
33:17
intellectual disabilities, fatty liver disease.
33:20
Most patients, if they're not diagnosed as a
33:22
baby, they'll have lens dislocation by the age
33:25
of three. And if they're not caught by then,
33:28
then they can have a heart attack as young as
33:29
11 or 12 years old. But beyond that, on the bone
33:33
side of complications, they can have morphinoid
33:37
phenotype. They can also have scoliosis, low
33:40
bone mineral density, fracture, knobby joints,
33:44
especially like in the knees. So although this
33:47
disorder has primarily been looked at in the
33:49
metabolic side, there are a lot of consequences
33:52
to bone health. So it's a rare disorder. It's
33:55
one of these conditions that there's just a small
33:58
problem with one enzyme, homocysteine builds
34:02
up in the system over months and years. It can
34:05
cause damage to many different organs. And it
34:08
is one of those conditions, if you pick it up
34:10
when someone's a tiny baby, you can, I think
34:12
it's fairly easily treated at the moment. But
34:15
if it goes on untreated, then unfortunately these
34:17
complications affect different systems. It is
34:22
quite a rare disease. They say in the U .S. about
34:25
1 in 300 ,000. babies are born with this disorder
34:29
every year which is not many so the research
34:32
is quite lacking because there's not a lot of
34:34
people that have this disease but basically as
34:38
a newborn most babies are screened they'll look
34:41
at their methionine and homocysteine levels and
34:45
if they're elevated they'll get a hcu diagnosis
34:48
but the other issue is that this disease is on
34:51
such a spectrum that it isn't always full loss
34:54
of function of this enzyme. They might have partial
34:56
function, and so the disease either goes undetected
35:00
because it wasn't elevated in the first few days
35:03
of life, or they really don't have severe symptoms
35:06
until later in their adult life. So it's quite
35:09
complicated, very understudied. Right now, the
35:13
common clinical treatment for it is betaine,
35:16
and it does help on the metabolic side. They've
35:19
seen reductions in homocysteine and methionine
35:23
levels in the hippocampus and the liver. But
35:25
the problem that we've still seen, at least in
35:28
one of our mouse models, is that betaine is still
35:30
causing... not great bones, to say it lightly.
35:34
The bone structure is messed up. We're looking
35:37
greater into this, but what we've been looking
35:40
at with our collaborator is an alternative treatment
35:43
to see if we can progress the methionine and
35:47
homocysteine pathway without using betaine. And
35:52
as opposed to focusing on the metabolic part,
35:54
we want to focus on targeting that bone phenotype.
35:58
So this is the interesting bit, that there is
36:00
a fairly straightforward, a sort of a dietary
36:02
treatment at the moment, which can, it would
36:05
appear, make things a lot better, reduce the
36:08
homocysteine levels. that standard treatment
36:11
at the moment isn't sorting out the bone problem
36:14
for these patients? Yeah. And you're looking
36:15
at a different treatment which might not only
36:18
help the metabolic problem, but help the bone
36:21
problem as well? Yeah, so these patients get
36:23
put on a very strict low -protein diet, and they
36:26
get put on usually some kind of metabolic formula.
36:29
It's not very tasty, especially since these patients
36:33
are usually, if they're caught early diagnosed
36:36
in childhood, it's very hard to make a child.
36:39
eat low protein and food that doesn't taste good.
36:42
They're going to be at school and their friends
36:44
are going to have a pizza party and they're not
36:46
going to be able to participate for XYZ reason
36:48
of their disease. So it's really about increasing
36:52
their quality of life in a treatment that is...
36:57
more reasonable for these patients. So what we're
36:59
looking at is treating the patients with a formate
37:02
and taurine supplement. We haven't tried this
37:04
in humans yet, but we've tested it in a mouse
37:08
model and supplemented from the time of gestation
37:11
of the breeders to the birth and the wean. And
37:15
until euthanasia of the mice, they were treated
37:18
with formate and taurine supplemented water or
37:20
regular water. And we found that the formate
37:23
and taurine sublimation actually increased their
37:25
bone size, increased trabecular parameters, increased
37:28
cortical parameters such as like polar moment
37:31
of inertia, the bone length was increased. So
37:34
a lot of really promising results thus far. And
37:38
it's a very interesting approach actually treating
37:41
in utero, treating the developing fetus even
37:44
before they're born. Yeah, so a lot of people
37:47
think that part of HCU is very folate directed.
37:53
Beyond genetic inheritance, if a mother might
37:57
not be taking enough folate during her pregnancy,
37:59
that that also is related to this disease. It
38:03
does sound promising. Thank you for sharing with
38:05
us. It'd be lovely if we can catch up with you
38:07
again in future conferences and hear the updates
38:10
about this important disease. Yeah, I'm looking
38:11
forward to seeing you guys. Thank you so much.
38:13
Thank you. We hope you enjoyed this listener's
38:19
episode. We've got a lot more interviews to come,
38:22
so tune in for the next part. We'll get it to
38:25
you as soon as we can. We'll see you then.