Circulation February 5, 2019 Issue

Dr Carolyn
Lam:               
Welcome to Circulation on the Run, your weekly podcast summary
and backstage pass to the journal and its editors. I'm Dr Carolyn
Lam, from National Heart Center and Duke National University of
Singapore.


Dr Greg
Hundley:            
And I'm Greg Hundley, associate editor for circulation from VCU
Health Systems in Richmond, Virginia.


Dr Carolyn
Lam:               
What does cardiac autoimmunity, glycemic control, and
cardiovascular disease risk and Type I diabetes have in common?
Well, you've got to wait for our feature discussion. This one's
such a hot one, don't you agree, Greg? We could hardly finish
talking.


Dr Greg
Hundley:            
Absolutely, and Myra, you're just going to love listening to her.


Dr Carolyn
Lam:               
Yep, but stay tuned. First, we're going to discuss a couple of
papers each. Greg.


Dr Greg
Hundley:            
Thanks Carolyn. So, the first paper I've got is from Professor
Van Rein at Leiden University Medical Center. And basically he's
getting at the issue of bleeding in patients with atrial
fibrillation. So this is a retrospective cohort that evaluates
different anticoagulation strategies for atrial fibrillation.
They examined 272,315 patients that had a median age of 75 years
and followed them longitudinally over time. These individuals
experience 31,459 major bleeding events, and what he did is he
evaluated whether they were not taking anticoagulant therapy,
whether they were on a vitamin K antagonist, a DOAC, antiplatelet
therapies, and then all combinations of the above, including
single, double and triple therapy.


                                               
What he observed is relative to taking a vitamin K antagonist
alone. The hazard ratios range from 1.13 to 3.73 in those that
were receiving dual antiplatelet therapy of vitamin K antagonist
plus antiplatelet therapy, a DOAC plus antiplatelet therapy, and
then of course triple therapy, which had that highest hazard
ratio.


Dr Carolyn
Lam:               
But were there particular combinations within these groups that
had particularly high bleeding risk?


Dr Greg
Hundley:            
Well, yeah, Carolyn. As we might expect, triple therapy was the
worst, but those that were receiving triple therapy, there were
two subgroups that were particularly susceptible to having a
bleeding episode. First, those that were greater than 90 years of
age, and second, those that had CHADS-VASc 2 scores greater than
six. Of course, these are very complicated patients, often
particularly that latter group. So there are clinical
implications. I mean, clearly, this isn't a randomized trial, but
what we should take away from this is that if we have one of
those two patient groups, age greater than 90, CHADS-VASc score
greater than six, that we ought to minimize the time that those
individuals are on that triple therapy.


Dr Carolyn
Lam:               
Talk about and bleeding, I've got a paper, and it's on the
performance of the ABC scores for assessing the risk of stroke
and systemic embolism or bleeding in patients with atrial
fibrillation. This is a study that actually looked at the
performance of these scores in an external cohort, which actually
hasn't really been done. Now, as a reminder, the ABC score is
actually the age biomarker clinical history stroke score, which
helps to estimate the risk of stroke or systemic embolism. The
ABC bleeding risk score incorporates biomarkers along with the
clinical variables to estimate the risk of bleeding.


                                               
All of these were tested in the ENGAGE AF-TIMI 48 trial, which
was that multinational randomized trial of the oral factor Xa
inhibitor edoxaban in patients with atrial fibrillation and a
CHADS-VASc 2 score of two and above. Now, this was from Dr Morrow
and the TIMI study group in the Brigham and Women's Hospital,
Harvard Medical School in Boston, Massachusetts. Basically what
they found was that the ABC stroke and ABC bleeding risk scores
performed well in stratifying the risk for stroke or systemic
embolic events or major bleeding in this multinational trial.


                                               
Compared to the CHADS-VASc score, the ABC stroke score provided
both correct upward and downward reclassification of the stroke
systemic embolism risk. Compared with the HAS-BLED score, the ABC
bleeding score resulted in a predominantly correct downward
reclassification of the bleeding risk.


Dr Greg
Hundley:            
So, this new ABC score, do we integrate it with HAS-BLED? Do we
integrate it with CHADS-VASc 2? How do we use this clinically?


Dr Carolyn
Lam:               
So first of all, there are some important remaining unanswered
questions, and this was really nicely discussed in an
accompanying editorial by Dr Hylek from Boston University School
of Medicine. Among this, first of all, the ABC scores need to be
validated in patients outside of a clinical trial. Remember, this
was a clinical trial cohort. Then there are questions about the
timing of measurements of the score, the different settings,
hospital and otherwise. Do these scores perform equally well
across different vascular beds and in diverse patient populations
at the same thresholds used?


                                               
So, all these things still need to be addressed. And really, in
Dr Hylek's words, the work has just begun.


Dr Greg
Hundley:            
This is an issue with the theme that might be bleeding, and I'm
going to talk about a study from Professor Huisman from Leiden
University again, and this is the RE-VERSE AD study. Again,
patients that are receiving dabigatran and that may have a GI
bleed or patients that are on this therapy and unexpectedly need
an emergent surgical procedure, this investigative team evaluated
the utility of idarucizumab on reversing that anticoagulant
dabigatran. So what did they do? They administered 2.5 milligrams
of idarucizumab twice separated by 15 minutes.


                                               
And again, the study population was uncontrolled GI bleeding or
those in need of an emergent procedure. The types of GI bleeds
that were involved in this study, a third were upper GI bleeds, a
third lower, and then a third, it was either unknown, or there
was a mixture of both upper GI or lower GI bleeding. So how do we
know that dabigatran is effective? We use a DTT time, and 98% of
those with an elevated diluted thrombin time had that reduced
after receiving these two twin 2.5 milligram doses at a time
point of four hours after administration.


Dr Carolyn
Lam:               
Okay, but were there any complications?


Dr Greg
Hundley:            
Yeah, there were. So first of all, something to think about is
that this is a high-risk group. In this study, 14.6% of the
cohort actually later died either from the bleeding or what have
you. Then another thing we need to be thinking about is when we
reversed this anticoagulant, do patients experience thrombotic
events? So what this group reported is 4.4% did within 30 days.
What were those? Myocardial infarction, deep venous thrombosis,
and subsequent PE. Then also at the 30-day time point, one
patient experienced an ischemic event.


                                               
Another question is once you've administered this, you've gone
through the procedure. You stopped the GI bleeding, or you've had
the surgery. In this particular study, 66% of those individuals
had restarted their DOAC. Those events occurred on top of that.
So, interesting information. Looking at administration of
idarucizumab, and we'll be using this I think frequently as DOACs
are used more frequently in the population, particularly
dabigatran, so some important data in guiding us on what we might
expect when we administer this therapy.


Dr Carolyn
Lam:               
I think going back to atrial fibrillation though, this is my
other selected paper, and it's actually results from the
GARFIELD-AF Registry. It's from Dr Bassand from University of
Besançon in France, and colleagues, and basically, they looked at
the early risks of death, stroke, systemic embolism and major
bleeding in patients with newly diagnosed atrial fibrillation in
the GARFIELD-AF Registry. They basically found that the rates of
all three major clinical events was significantly higher during
the first month than in the subsequent period set following up to
12 months.


                                               
The leading causes of early death were heart failure, sudden
death, acute coronary syndromes, infection or sepsis, and
respiratory failure.


Dr Greg
Hundley:            
So, what's the take-home message here?


Dr Carolyn
Lam:               
This is observational, so the key thing to understand here, it's
a registry. It's observational. We can't really tell chicken from
egg with regards to its newly diagnosed AF verses events, which
comes first, which causes what. But nonetheless, the increased
hazards of an early event and especially cardiovascular mortality
in these newly diagnosed AF patients really point to the
importance of comprehensive care for such patients and really
should alert physicians to detect warning signs of possible early
mortality in these newly diagnosed patients.


Dr Greg
Hundley:            
Very good, Carolyn.


Dr Carolyn
Lam:               
I think that wraps it up. Let's hop to our feature discussion,
shall we? I'm so super excited about today's feature paper
because it may explain that strong link between hyperglycemia and
cardiovascular disease in type one diabetes and all by revealing
a potential novel pathway that may have been hiding in plain
sight. And yes, I'm stealing the words of editorialists and our
associate editor, Dr Naveed Sattar from University of Glasgow,
and we're all so pleased to have with us the corresponding author
of today's feature paper, Dr Myra Lipes from Joslin Diabetes
Center in Boston, Massachusetts. Myra, start us off by telling us
a little bit about your study please.


Dr Myra
Lipes:                  
Sure. So we were interested in examining the role of whether
chronic hyperglycemia could trigger cardiac autoimmunity in type
one diabetes, because chronic hyperglycemia is associated with
subclinical myocardial damage, and we had actually previously
observed just unexpectedly in a young adult cohort that ...
Actually from Italy, where unexpectedly, we noticed that patients
with the poorest glycemic control expressed cardiac antibodies.
There's a lot of interesting people who are autoimmune-proned may
overreact to injury of certain tissues.


                                               
So, type one diabetes, it's a classical autoimmune disorder. So
we examined, really tested this hypothesis, in stored samples
from the DCCT/EDIC study, and this is a very landmark study where
patients were randomized to tight glycemic control, intensive
glycemic control. Then another group had just conventional
control, and this was done over an average of six and a half
years. So during this time, the samples were stored. Every year
samples were stored from participants, and this was quite a rich
data set that is publicly available. So we studied the
development of autoimmunity in two groups that had very distinct
separations of the A1C level.


                                               
We specifically excluded people who developed kidney disease or
cardiovascular disease events during the study. So this is a
cohort that had relatively recent onset type one diabetes.
They're relatively healthy, and again, groups were matched with
cardiovascular risk factors at the beginning and the end of this
DCCT period. And of course with our studies, we've also looked
genetically because your HLA immune response genes can influence
susceptibility to autoimmunity.


                                               
These patients were actually matched in HLA genotypes. So what we
found was that patients with poor glycemic control, there was
expression over time. You could see a time course relationship
between expression of antibodies over time on the levels of the
antibodies that were different in the two A1C groups. The number
of antibodies were different in that with the high group
expressing more antibodies, more different types of antibodies.
These are antibodies ... might say antibodies as like proteins in
the blood, and they're actually directed against parts of the
myocytes, the myofibrillar complex, and a major target is cardiac
myosin heavy chain.


                                               
We saw the different parts of the myosin heavy chain retarded,
and the presence of two or more antibodies, different types of
antibodies, different regions of the myosin to different
isoforms. Also, we saw antibodies, the troponin, troponin I. So
the number of antibodies with different ... with almost a
complete absence of antibodies in a tightly controlled group. I
might mention the A1C average was 6.5%, so this is a very tightly
controlled group whereas the poorly controlled group is at the
opposite extreme, the average A1C during DCCT. The mean updated
A1C was about 10%.


                                               
So, it was a very clean group, two different groups, and we could
see that the number of the types, the number over time, very
different in the two groups. In fact the profiles of these
antibodies were almost very similar to patients with Chagas
cardiomyopathy. That was our positive control group. Chagas
cardiomyopathy is possibility to be a form of chronic myocarditis
directed against cardiac myosin. So the profiles are almost
indistinguishable. So on one hand, you have relatively healthy
patients with type one before glycemic control, and that was very
unexpected that this would look pretty similar.


                                               
But very interestingly, and I might say unexpectedly, we saw ...
It was very clear that the people with the highest titers of
antibody and the most different types of antibodies, particularly
two or more, were subsequently ... We noticed that those patients
were at high risk for developing CVD events. And that's while the
number of events was slow, we noticed that all the patients, some
60%, had two or more antibodies and developed cardiovascular
events. Perhaps one more striking example is a single patient in
the study could die of cardiovascular death, had a positivity for
all five antibodies at highest titer.


                                               
Then we looked at coronary calcification just to measure
subclinical atherosclerosis. We noticed that the same numbers,
two or more, and also the same antibody specificities that were
the highest predictors of CVD events were also predictive of
coronary ... had detectable coronary calcification. In addition,
we looked at the levels trying to find mechanistically what could
explain the link between cardiac autoimmunity and an increased
risk for atherosclerosis. We looked at CRP, high sensitivity CRP
levels.


                                               
Again, these were measured about a decade after the antibody
samples were obtained, and we saw that the positivity for
multiple antibodies was also associated with markedly elevated
... subsequently elevated high sensitivity CRP levels with levels
of six versus something like 1.4 in a group with one or less
antibody. So these were very intriguing findings, suggesting a
role for autoimmune pathways as a susceptibility to
cardiovascular disease in type one diabetes.


Dr Greg
Hundley:            
Myra, that was absolutely incredible description of the study and
all the particulars of the findings. I wonder if I could ask both
you and Naveed, where do you see the next steps moving forward
with this research in the future? Number one. And number two, is
this in any way can be used to segregate patients that may need,
for example, really aggressive glucose control with an insulin
pump or something of that nature?


Naveed
Sattar:                 
I think we left this study as beautifully described as you see by
Dr Lipes. I think the context ... We looked at this from
editorial perspective ... is that most people don't realize if
you have a middle-aged person with type one, their hazard ratio
for cardiovascular risk is about somewhere between four to six
fold for men and women respectively, which is much higher than
type two. It's often thought that it's the area under the curve
for hyperglycemia. But what this paper throws up is actually
maybe there's another pathway, which we just didn't understand
that this wasn't a permanent autoimmunity closing subclinical
myocardial disease and inflammations.


                                               
But potentially, for me though, there's a saying in British that
one swallow does not make a summer. So, it would be nice for
other groups to replicate this. I think the findings are, as they
stand in isolation, fantastically well done. But it would be
lovely if other groups had accessible samples, and I knew of
several groups that have up towards tens of thousands of samples,
maybe even not 10,000. Certainly 10,000 or so plus or minus
samples for type and prospective outcomes to potentially validate
the findings and extend them.


                                               
And really, if the antibodies do help protect people at higher
risk in a meaningful way and improve beyond what we can already
do, then you're right. Absolutely. If we can pick up early people
who are going to have substantially higher risk, you would want
to potentially improve glycemic control, potentially pumps, CGM,
closed-loop systems or more intensive statins or lower blood
pressure targets or other types of antihyperglycemic agents,
which seem to be being tested in type one as well. So that's
really one example.


                                               
And for me, the other thing would be really nice is to pull up
any inflammation. Is this high systemic inflammation? Is it IL-6
level? Is it something else? What about troponin and BNP levels,
et cetera. I'd be interested to hear what Dr Lipes thinks and how
do you think to take it forward as well.


Dr Myra
Lipes:                  
So, this is something Dr Sattar said and I completely agree.
Actually, right now, we're looking at the DCCT cohort as a whole
for already. It's relatively small compared to the
population-based studies. But there's 1,400 patients, and the
subjects had CMR studies that were published in Circulation. So
we're going to actually study next whether we see CMR evidence of
systolic dysfunction and looking at the broader DCCT cohort. So,
those studies are underway. But of course the ultimate test would
be looking at if there were samples available from the Swedish
NDRs, Scottish registry.


                                               
I think it's something that's not often done prospectively. So
that would be incredibly exciting, and that's the important
thing. I'd say with type one diabetes, for screening for type one
diabetes, the use of autoantibodies and particularly two or more
different types of islet autoantibodies, and this is just putting
things in a broader context, is the entry criteria for type one
diabetes prevention trials and something cardiologists wouldn't
be aware of but this particular thing. So in decades, people,
researchers, in the field has spent decades optimizing islet
antibody assays.


                                               
So by analogy, it would be really important to standardize assays
so that they can be done in Sweden and Scotland and so that other
groups could confirm this, and I'm confident that this could be
done, since the setting up of our assays was really built on the
experience of people of developing standardized assays and
rigorous cutoff points for antibody positivity. So it would be
really important to work internationally to try to tap into this.


Dr Carolyn
Lam:               
Oh, my goodness. Myra, Naveed, these are such insightful
comments. I think as Greg said earlier, I think we could go on
forever discussing this paper, but I'm so sorry. Our time is up.
Before we go though, I must point all readers to look at figure
five of this marvelous paper. It puts together the whole schema
of how autoantibodies can play a role both in myocardial and
atherosclerotic cardiovascular disease and type one diabetes.


                                               
Thank you so much. Greg and I loved having you. Listeners, don't
forget to tune in again next week.


                                               
This program is copyright American Heart Association 2019.