Herb Kressel: Hi.
This is Herb Kressel, Editor of Radiology and welcome to the October 2017 Radiology
podcast.
Today, I'm delighted to be joined by Professor Hildo Lamb who is Professor of Radiology at
the Leiden University in Holland, and who with his colleagues, actually in Leiden and
throughout the Netherlands, wrote a very stimulating paper entitled, "Is Hepatic Triglyceride Content
Associated with Aortic Pulse Wave Velocity and Carotid Intima-Media Thickness?
The Netherlands Epidemiology of Obesity Study."
Welcome, Dr. Lamb.
Thanks for joining us again.
Dr. Lamb: Yeah, thank you for another opportunity.
Dr. Kressel: It's a pleasure.
So, first thing's first.
What exactly is The Netherlands Epidemiology of Obesity Study?
Dr. Lamb: It's a large cohort study that we initiated in 2008, so now almost ten years
ago and we included more than 6,000, it's almost 7,000 patients in total and about one-third
of those patients also had imaging and in that imaging part, we had different randomizations.
One of the randomizations was to image the aorta and the cardiovascular part, but also
some of them were included in the arm to measure the carotid IMT, the intima-media thickness
and lots of other things like blood work, neurologic testing, and all kinds of things
that were not included in this paper, but for this particular topic we ended up with
almost 2,000 patients for the statistical analysis.
Dr. Kessel: Right.
So can you just share with us the specific study rationale?
Why did you target this particular content?
Why would one expect that hepatic triglyceride content should independently be associated
with pulse wave velocity and/or carotid artery intimal wall thickness?
Dr. Lamb: At first this seems a really strange association that you might think, "Well, what
has the liver to do with the cardiovascular system?"
It was actually based on experiments some time ago based on dietary interventions where
we observed sort of co-variants in lipid levels in the liver and normal cardiovascular function
and vessel stiffness and then we developed this concept in the end that we think that
all over in the body the process is more or less the same, like in atherosclerosis.
So first, you may have fat accumulation, then in response you develop inflammatory processes
and based on that you have collagen formation and scarring.
So what we thought maybe that the liver is the central organ in this process and that
the liver actually directs all these changes all over in the body.
So the idea was to study the link between fatty liver and other organs and in this case
we were interested in two separate processes - the vessel stiffness and as a separate process
atherosclerosis by measuring the intima-media thickness by ultrasound.
Dr. Kessel: So blame it all on the liver.
Dr. Lamb: Yeah, maybe that's a good thing to remember.
Dr. Kessel: The liver is the root of all of joy and problems.
Okay.
Very interesting.
So another thing that some of those listening and viewing the podcast may be curious about,
can you tell us a little more about pulse wave velocity alterations in the aorta and
what's the presumed mechanism of this in relation to atherosclerosis and to brain changes as
well?
Dr. Lamb: In general, it's a complicated process.
The name is also difficult, pulse wave velocity, but what it actually means is that we tried
to measure the propagation of the ampules by the heart ejection.
So we do not measure the flow or the flow volume.
It's just measuring how the blood propagates through the aorta and you can compare this
for example by water flowing through a pipe.
If the pipe is very stiff, the transportation time of the water from the pump to the distal
end of the tap is very fast, so the speed of the water is very high and if you have
a more elastic tube like the aorta and when you're young and healthy, then it will be
a little bit dampened and the speed of the fluid is decreased a little bit.
So high pulse wave velocity reflects a stiff pipe and in this case stiff aorta, and the
idea is that this is maybe related to atherosclerosis, but we are not sure if this is really the
cause or effect.
We think that this is related to the balance between the elastin in the vessel wall and
the collagen level and we also know that this ratio is different between the central part
of the arteries and the peripheral arteries.
For example, there is a difference between the central aorta and the peripheral vessels
like the carotid artery where we measured intima-media thickness.
Dr. Kessel: I see.
Very good.
Now, a little more about your study design.
What did you actually do to study your hypothesis and study the associations?
How did you choose the cohort?
Did you have any subjects with elevated triglycerides who were not obese, as this is primarily the
obesity study.
So how did you go about answering the question you asked?
Dr. Lamb: Well, the only major inclusion criteria was to be healthy or that you don't have
a real known cardiovascular disease or diabetic disease, but actually it was a quite open
study.
So also we included normal people with normal BMI and also we had people in the overweight
category between a BMI of 25 and 30, and obesity group with a BMI higher than 30.
And based on the results of blood measurements in the patients, the blood work, we afterwards
corrected the findings.
So for example, we also included the lipid work, the lipid levels, use of lipid lowering
medication, and we used all this data to detect confounding factors in our statistical model.
So actually everybody was welcome in the study and afterwards we were using the statistical
tricks and that is possible because it's a relatively large cohort.
Dr. Kessel: Yes.
So if you can tell us, briefly summarize your key findings.
Dr. Lamb: The key finding is best shown in Figure 2.
That is the easiest way to understand.
So actually what we found is that if you have higher triglyceride content in the liver,
so you have a fatty liver, that also the vessel stiffness or the pulse wave velocity is higher.
Also on the other end of the spectrum, we slowed that with a lower BMI and with low
liver fat, that also your pulse wave velocity is lower.
So this confirms the feeling we had from previous studies and small studies that in this large
cohort we indeed have an association or some sort of statistical link at least between
the fat level in the liver and vessel stiffness.
On the other hand, we also found relation between the carotid intima-media thickness
that was a little bit higher in the patients with higher liver fat.
Then it's always difficult to interpret these statistical numbers, so for the intima-media
thickness it means that in the real world if you have a ten-fold increase in your fat
level in the liver, then a 15 micrometer increase in your intima-media thickness happens.
So that's the way - how you have to interpret these results.
The same is true for the vessel stiffness, for the pulse wave velocity.
So if this fat level in the liver increases ten-fold, then you have about a 0.2 increase
in your meter per second speed of vessel stiffening.
So that's how you can look at these numbers.
Dr. Kessel: Good.
So can you kind of review how you identified this as independent of the other features
of the metabolic syndrome?
Dr. Lamb: Yeah, that's always nice of those larger cohort studies that you can use all
the information you have and run different models.
So what we normally do is start with the crude model, that's usually model 1, also in this
study.
Then we chose to use model 2 to check the confounding effects of general things like
age, sex, blood pressure, heart rate, but also alcohol use, smoking habits, use of antihypertensive
drugs, lipid-lowering drugs.
So that's model 2.
And then on top of that, we were more specifically interested in the factors determining the
metabolic syndrome.
That was model 3.
And in model 4 it was everything and also on top of that the visceral adipose tissue
and the total body fat.
And actually in that model 4 all the factors we could imagine are included there that could
confound the results and even in that situation, we had a significant correlation or association
between the liver fat and the pulse wave velocity and the liver fat and the intima-media thickness
of the carotid arteries.
Dr. Kessel: So are these models basically regression analyses or is it-
Dr. Lamb: No, mainly ANOVA linear analysis.
Dr. Kessel: Very good.
Now one thing I noticed is that you had a - at least it was termed - a high failure
rate in the acquisition of MR spectroscopy, which you used as the metric for hepatic fat,
and with approximately a third of the patients there wasn't usable data.
What do you think this was due to?
Not allocating sufficient time for the exam or some lack of training?
Dr. Lamb: That's an interesting question.
I think it's mainly due to the technical challenges you have to perform spectroscopy studies in
general.
In the liver it's one of the easiest variants of that.
In the heart, for example, it's much more complicated.
So we had a learning curve for the technicians.
That was one factor.
So in the beginning, they missed out a lot of scans because you have to perform it with
and without water suppression and sometimes we had two spectra with water suppression,
two spectra without water suppressions, so that was a problem in the beginning and it
was also a reason why we started developing newer techniques, so now in new studies we
prefer to use the imaging-based Dixon-like techniques where you have a very easy setup
and everybody can do the scan.
Dr. Kessel: I see.
Well, that certainly makes sense.
And you mentioned looking at the lipid-lowering drugs.
What was the effect of these when you kind of analyzed it in the regression models?
Dr. Lamb: Yeah, for the parameters we studied, that actually then - the difference between
model 1 and 2, there was an effect.
The significance of the model went down of the beta actually of the association, but
still it was statistically significant.
So there was an effect, but it did not explain the relation fully between the fatty liver
and the vessel stiffness and the IMT thickness.
Dr. Kessel: Okay.
So they sort of moved down the curve of the response curve if you will.
Dr. Lamb: Yeah.
Dr. Kessel: Okay.
And going forward, what do you think are the potential clinical and public health implications
of your findings other than everyone should control their weight?
Are there other implications?
Dr. Lamb: Yeah, there is one interesting pathophysiological finding that's a little bit more detailed,
but we found a difference between the IMT measurements association because there was
an effect of visceral fat.
Actually, when we included visceral fat in model 4, the relation between the liver fat
and IMT thickness was not significant anymore, but in the vessel stiffness, it didn't matter.
So the significance even got better.
So this may indicate that the pathophysiological process behind atherosclerosis is a little
bit different, different pathway, different factors than the vessel stiffening itself.
On the other hand, if you are very practical, we know from literature and we performed a
couple of those studies before, that there is still hope because if you do a diet, then
also the pulse wave velocity changes.
So it is still reversible if you are in a sort of sweet spot area.
Dr. Kessel: Well, thank you very much.
This is very exciting work and we look forward to seeing the results of your future manuscripts
and thank you so much for joining us today.
Dr. Lamb: Okay.
Thank you very much.
Dr. Kessel: You're welcome.
Bye bye.
Dr. Lamb: Bye bye.
No comments:
Post a Comment