DRF 19: Cardiovascular Outcomes for Diabetes Drugs: Lessons Learned and Path Forward


(calm music) – Research Forum. Thanks for being here today. We have an all star panel today to tell you about cardiovascular outcomes for diabetes drugs. One housekeeping issue
before we get started. Next week’s research forum
is canceled, due to ACC. So for those of you going to Orlando, I hope you have a wonderful and safe trip and great conference. And with that I will turn
it over to Dr. Granger who apparently doesn’t have a chair, so will be modeling good cardiovascular risk prevention practices
by standing the whole time. I hope that’s okay with you. – [Dr. Granger] Try to have good posture. I’ll get some walking;
I’ll get some steps. – I have a microphone. Welcome and we look forward
to a panel discussion and engaging you all too in what I think is a really important topic. But before I get into that, maybe Matt, can you introduce yourselves and about your interest in this topic? – Sure. Matt Rowe, cardiologist
here and I’ve been helping to direct a think tank
series for the last few years and I have a particular
interest in clinical trials, especially now in diabetes
and cardiovascular outcomes trials, which is a really growing area so I was really pleased
to be part of this. I think there’s a fantastic think thank. Probably the best one
that I’ve been at ever. – Thanks. – It’s also the best
one I’ve ever been at. It’s the only one I’ve ever been at. (laughing) And I’m a cardiologists
and I’m interested in cardiometabolic health, both clinically and from a research perspective. – Eric Peterson. I’m interested in diabetes and cardiovascular trials
and implementation. – There we go. Perfect. – I’m Jennifer Green. I’m an endocrinologist. I joined the DCRI faculty in 2008 a the start of the TECOS trial, which is one of the CV
diabetes outcomes trials that we’ll be discussing today. – Great. The topic is cardiovascular
disease and diabetes and drug development for patients with diabetes and
cardiovascular disease. And it’s a really important topic. Probably most of you kind
of have a sense of this, but Rob Kalif gave a not surprisingly inspirational dinner
lecture at our meeting, at this meeting in the end
of January, this think tank, and he made the key points
including telling anecdotes about the FDA experience,
which he’s not been able to do until recently, because he
had a one year moratorium on talking about these things. But in addition to that,
he highlighted the fact that for the first year since
the flu epidemic in 1917, life expectancy decreased
this year in the U.S. and two of the main contributors to that are the opioid epidemic, which is actually particularly
important in North Carolina; I think New Hanover is like the number one county in the United States for opioid related deaths;
and the diabetes epidemic, or cardiometabolic. A fact that now we are seeing the consequences of obesity and diabetes impacting cardiovascular
health in a way that’s manifesting a shorter life expectancy. So it’s a big deal and it’s something that’s incredibly important. The history for Duke in
this particular topic goes back at least to 1997
when the first think tank, one of the first think
tanks was actually … It was 97 and it was on this issue of cardiovascular outcomes
in patients with diabetes and the effect of these drugs. And back then we knew very little. There were a couple of early trials that provided some information, but the main conclusion
of that think tank was that we need more research in this area. And then there was another one in 2004 and another one in 2007 and the one in 2007 actually … By this point, this was the time when Steve Nissen had done some work using some data that was
not particularly high quality, but available through the FDA, to highlight the fact that
the drug rosiglitazone was associated with cardiovascular risk. And then this think tank asked the question, is it good enough, when we develop drugs, to
look at glucose control or should we really be getting beyond that and looking at cardiovascular outcomes. I think that meeting
actually did have some role in prompting the FDA to
say there needs to be a new paradigm about these drugs, because there really is uncertainty about the impact of these
drugs on cardiovascular outcome and, in fact, for patients
with type II diabetes the main adverse outcome
that these patients have is cardiovascular: cardiovascular death, myocardial infarction, stroke are the mot important outcomes,
therefore these drugs, it’s really important
to understand the effect of these drugs on those outcomes and that led to this guidance that’s shown on this
slide, this 2008 guidance. earlier that year. You guys can’t see this can you? But these are some of the … (audience members speaking
indistinguishably) Go ahead and sit down for minute. I’m gonna take about five
minutes to go through this, because I do think it’s
important information. People in the back may not
be able to read this either, so I’ll read some of it. It’s a pretty short document. I think it’s a total of six pages of text and it really makes
four or five key points that are the basis of the think tank, because the think tank was really designed to ask the questions, since 2008, over a 10 year
period, what have we learned based on the result of this guidance and what should be done differently now. So those were really the
questions that we brought together the typical think tank
interdisciplinary group including something like
20 people from the FDA and people from NIH, from funding agencies including commercial insurance and CMS as well as academic and industry leaders. This was the guidance. The recommendation was
to establish the safety of new antidiabetic therapy
and to treat type II diabetes. Sponsors should demonstrate
that the therapy will not result in an unacceptable increase
in cardiovascular risk. That’s important. The intent here was not to prompt sponsors to show that there was benefit. It was to show that there was not unacceptable increase in cardiovascular risk and ensure that a new therapy does not increase cardiovascular
risk to an unacceptable extent. The development program should
include the following … And this is interesting. The first recommendation was … And this was partly based on the fact that this data that Steven Nissen reviewed included data on, for
example, myocardial infarction with some uncertainty as to whether these were real
myocardial infractions. They weren’t really very
well defined or adjudicated. So the first recommendation was one that we liked,
from the perspective of the DCRI clinical events
classification group and that is that there be a formal clinical events classification strategy and that the outcomes of
interest needed to include cardiovascular death, myocardial
infarction and stroke. That was what was required and then they could
include hospitalization for acute coronary syndrome, urgent revascularization and possibly other end points. And we’ll bet back to that, because there’s some
important experience here. They also acknowledge the fact that when there are a
series of trials done, that a meta-analysis
technique could be used to incorporate all the
information on a new drug to establish whether or
not there was a signal of harm that the trial should include
patients at higher risk of cardiovascular events, to ensure that both there were
enough events occurring and that the results would be applicable to the population for whom
we have the greatest concern, which is those at higher
cardiovascular risk. And then there also was a specific comment that the trials needed to be long enough with a minimum of two years of followup to understand what the
longer term effect was of these drugs. This is a topic that we covered, to some extent, in the think tank. It’s a challenge, because
most clinical trials don’t include the long term follow up that our patients get exposed
to when treated for a lifetime of type II diabetes and are subjected,
therefore, to potential risk. That a meta-analysis should be performed including all of the available data and then the statistics were interesting and somewhat arbitrary at the time that there were these two boundaries that were created; a boundary of 1.8 in order
to get the drug approved provisionally so that one would have to exclude with 95% confidence a greater than 1.8 fold increase risk of cardiovascular outcomes and then once that was done, one
would also need to exclude a 1.3 fold increased
risk, as a second step. That could either be done right away, in the initial
development program before approval or it could be done in a postapproval required cardiovascular outcome trial. And the way the guidance was written, it could even be done in one trial. So you could do one trial and have a data safety monitoring
board that would decide after there were about 90 or 100 events, which is the number needed to
exclude the 1.8 fold margin and then keep that trial going until the 660 or so events that
would typically be needed to assure that the 1.3
fold margin was excluded. So I’ll just briefly mention this and we’ll get the panel up. So what happened? That was 2008 and then what happened
over that 10 year period? We spent a fair amount of
time going through this, but it was pretty
interesting what happened. The first thing that happened
is industry was pretty upset, because they were gonna have to spend, Jennifer told me, like
$400 million per trial. We can do it more cheaply, but most of these programs,
it’s something like, I think it’s more typically been … We heard about something
canagliflozin program was like a billion dollars,
I think something like that. – [Jennifer] That sounds fairly expensive. One of the trials that
we ran at the university was a little under $250 million. It depends. – [Dr. Granger] A lot of
money went into these, so industry was not happy they
were going to have to spend all this money before they
could sell these drugs to lots of people have type II diabetes. So the initial response, even
some of the academic leaders, I think, were a little bit irritated that this barrier was in place, but then over time people became, especially once they were
drugs shown to be beneficial then people said, “Well,
this was a great thing. “Now we actually have some
drugs that are beneficial.” So there were at least 15 large programs of diabetes drugs performed,
many billions of dollars spent. Maybe $10 billion. And spent, you could argue for what’s not a terribly exciting objective and that’s to prove that these new drugs, which are very expensive are not worse than placebo. Doesn’t sound like a very exciting thing to be involved with, to prove something is
not worse than placebo for really serious problems occurring. And then what happened
is none of these trials showed increased cardiovascular risk. Every one of them has shown less than a 1.8 margin for increased risk. So one could say okay, zero out of, I don’t know how many have
actually been reported out, 12 or 14 or something, in terms of the programs, but you said 10. You think 10? So what does none out of 11 mean? Does that mean that this will never happen? No, it doesn’t mean it will never happen, but it certainly suggests that most of these drugs, in
spite of the initial concern, don’t have cardiovascular risk, but some did show other
risks like heart failure and liver toxicity and other things, which one could argue
were seen largely because the trials were large enough and because the adjudication committees were in place to be looking
at cardiovascular outcomes in a fairly systematic way. And then these four drugs that Neha will go through
a little bit, I think, showed cardiovascular risk reduction, statistically significant improvements in cardiovascular outcome and two of those have been
approved with labeling to acknowledge that they
improve cardiovascular outcomes. But then finally, the final kicker is none of these are
being used in practice. So huge amount of money went into showing these drugs were not harmful, some of them were shown to be beneficial and then those aren’t
being used in practice so we talk some about why. We finally have something to
improve cardiovascular outcomes from one of the largest
cardiovascular diseases in the world and they’re not being used. Why is that and how can that be addressed? So that’s a little bit of background. Now I’ll get the panel up. And actually Neha maybe
you can take us through this next couple of slides. – [Man] Wanna do the slides first? – [Dr. Granger] Maybe, maybe one more. Sorry. Getting a little exercise. – I will very quickly, just to kinda get everybody on the same
page of what these drugs are, I’m just gonna quickly
outline the four trials that have been done that were positive, that showed benefit and that
everybody will talk about. The first one was EMPA-REG OUTCOME. This was published in 2015 and it was with the drug empagiflozin,
which is an SGL2 inhibitor and I think of these; I’m sure Jennifer can give a much better explanation; I think of these drugs like a
cardiologist thinks of them, so very simply. SGLT2 inhibitors give you sweet pee. You pee out your glucose and this was the only one
of the positive trials that only looked at a secondary
prevention population. Everybody had a history
of cardiovascular disease and they also obviously
had type II diabetes and what they found was
over a mean of 3.1 years there was a 14% risk reduction for MACE, 38% risk reduction for
cardiovascular death and nobody saw this coming. This was the first of
these trials to come out that was positive and
nobody had saw this coming and the curve split very
early, so the thought was that for this SGLT2
inhibitor this might be kind of a hemodynamic
effect, maybe it led to some diaeresis in these patients who had subclinical volume overload. And along those lines, there
was a 35% risk reduction in heart failure hospitalization. Again, a huge risk
reduction in heart failure. The down side to this
drug is that it can cause pelvic infections, because
you’re peeing out sweet pee. For LEADER that was liraglutide. That came out in 2016. That looked at both a primary and a secondary prevention population, but a majority of the people
had cardiovascular disease. Sorry, I should say this is a GLP agonist. A GLP1 agonist is an incretin hormone and again I think of this very simply. That’s a hormone from the gut that does things to make your sugar better and I’ll keep it at that, because it does a lot of different things. There was a 13% risk reduction for MACE, 22% reduction for cardiovascular death, though the MI and stroke and
heart failure risk reductions were not significant. There you can potentially
get a lot of GI upset; that may be part of why it
helps people lose weight; and some gallstones. Semaglutide came out the same year. That is another GLP1 agonist. As opposed to liraglutide,
which is a daily shot, empagliflozin is a pill
that you take every day. Liraglutide is a shot
that you take every day. Semaglutide was a once weekly shot and that was also done in a primary and secondary prevention population, also showed a MACE reduction of 26% and a hefty stroke reduction of 39%. And this is in the same class and you see a stroke reduction here
and not with liraglutide, so it’s kind of interesting we’re seeing some of these differences within a class. The issue there was retinopathy. And then the final one was
CANVAS, which was Canagliflozin, which included two trials within it, one a general cardiovascular outcome trial and one a renal program,
a renal outcome trial. Again, looked at people with both primary and secondary prevention, but here, starting to move more towards a primary prevention population. A majority of people had not had primary cardiovascular disease. Again we see a 14% MACE risk reduction, 33% heart failure reduction, but here; as opposed to with empagliflozin, this is another SGL2 inhibitor;
people’s toes come off. Just the toes though. Not from the knee. It’s really toe and metatarsal amputation was twofold greater with
Canagliflozin than with placebo. So that’s just a quick outline
of the four positive trials and this is just an outline
to basically emphasize what Dr. Granger has said, which is the space has gotten really
crowded in the last few years. This outlines the trials
that have been done, that have preliminary results
and that are ongoing in green. And it think Dr. Granger’s
right that if he were to guess, probably somewhere around $10 billion have been spent on these programs. That doesn’t include R&D and
everything that went before it and it leaves some of the
data that we have at Duke. And we can get at this,
but of patients at Duke who have diabetes and
cardiovascular disease, 1.7% of them are on a GLP1 agonist and 0.2% are on an SGL2 inhibitor, at least as of the end of 2016. So it’s basically nobody. So this just gives you a sense
of some of that perspective. – Perfect. Just leave that up and now
let’s get the panel back. Try one more time. And if people have questions, please ask questions as we go through. John has a question, I know. You will soon. The final preamble is Jennifer. I’d like for Jennifer just to come and … Including for many of you in the room who’ve been involved
in trials at the DCRI, what the DCRI involvement
has been with these trials. – The DCRI involvement
has been fairly extensive for many of these trials. As you may know, the TECOS
trial was actually the trial that was first designed to satisfy the requirements of the FDA guidance and that, essentially, was a
full service DCRI run trial. We shared academic
coleadership with Oxford, but really essentially supervised the entire conduct of the study, U.S. site management,
supervision rest of the world, CEE, safety, publications
ongoing, et cetera. The EXSCEL trial, which is more towards the middle here, had a similar operational and
thought leadership construct, although I think we may not have had quite the same extent of responsibilities, it was very close to what we had in TECOS and then the ongoing HARMONY trial, we have thought leadership,
U.S. site management, we’re involved with the
statistical reporting and analysis and we also have an EHR
study that’s ongoing. That’s a really special
and exciting feature of what we’re doing in that trial. – [Dr. Granger] Great. – Jennifer, obviously when
we work on big trials, we have academic
collaborators around the world who we work with and traditionally our collaborators with ACS trials, or coronary disease trials,
but we have some new groups who we’ve been working
with on diabetes trials, such as the DTU group, so maybe you can just briefly describe that group, just so everyone can understand
that we have to collaborate and we need to collaborate
to do these types of trials and what the relationship has been like with the new collaborators we
work with on these studies. – I think what we’ve seen over the years, there was some evolution of, I think, what you traditionally
considered your ACS copartners, into this space. I think everyone’s becoming interested. So everyone’s a potential partner, but this collaboration with
the Oxford Diabetes Trials Unit group was something new that was established with TECOS and then continued into
the EXSCEL program. There is a large group that
are interested in diabetes and various aspects and
types of diabetes trials and DTU is one of those. We have an upcoming study with which we’ll be collaborating with a somewhat different group at Oxford, but it’s been a very
fruitful relationship. – Good, okay. Well let’s start then and … So again, there were
those two main questions that I’ll just put up again
here for the audience. What have we learned and
what can we do better? And Eric, let’s start
with your perspective on, no simple answers here, but
both your personal perspective and your sense from the meeting. It was kind of an experiment, right? Here a regulatory authority kind of made these somewhat arbitrary boundaries, created the
cost of billions of dollars and it resulted in these
four drugs, so far, showing some benefit and a
lot of drugs showing no harm. Was it the right thing to do and what’s your sense about
what should happen now? Should we abandon it or modify it or what’s your advice to the
new commissioner of the FDA? – I guess I would start
by saying a little bit of the billion number has
to be put into perspective. If you think about yearly sales, for example the drug
that we studied in TECOS, their yearly sales are into the $5 billion range. So every year they’re cranking
in that, in terms of revenue. So the ability to spend
some proportion of money to study a drug to make sure
that it’s safe and effective in that setting may not be that wrong, if you wanna take it that way. – Sorry, I was just gonna say, but isn’t that fascinating that the drug that has no cardiovascular
benefit is the one that is having billions
of dollars of sales and yet these other drugs
that cost the same amount and they’re incredibly expensive, but they cost the same amount
as the DPP4 inhibitors, they’re just not getting used at all. – It’s the Jennifer Green effect. – It’s because you were part of the trial. – It creates and opportunity. – We’ll get back to that topic. We can do it now also, but I
do think the question about how do we get these
things better implemented is a critical one that
you can either address now or we get back to it. – We’ll get to that, the
implementation, in just a second. The first question, I didn’t mean to say about which drug was being used and I was kidding about
Jennifer, but the realities are that these drugs are used in
a lot of people worldwide. Diabetes is only going up,
in terms of it’s frequency, so yes, it’s probably reasonable that we make sure that these drugs … And there were safety signals
at the time that it came out. Whether they were confirmed
or not was not clear. Had we not done large outcome studies, we would have never identified the fact that there is significant variance across the class of drugs, in terms of which ones are more effective versus not and lo and behold we have found several that are not only showing benefit for patients with diabetes, but may open up entire new fields. These drugs may have usefulness, as was mentioned by Neha and others, in the fields of kidney
disease in a way to prevent progression of renal failure,
heart failure progression and creation and even in other fields, maybe even as a cardiovascular
preventative agent. These are all the things that
people are investigating now that they honestly would
have never probably done had it not been for these
large outcome studies. So I think the net on this
is a pretty positive story. I think it should be written up as story of where regulation actually had a positive effect for a change. – I think that reflects the group. Don’t you guys think that most
people at the think tank … Actually Mary Parks, who’s the
head of the metabolism group at the FDA, she described
this as being … For a couple of years she
was getting a lot of heat and very few people were supporting her, but then more recently
the conventional wisdom is this has been a good thing for the field and probably even for the
pharmaceutical industry, the fact that we now have drugs which have been proven
to have bigger benefits. – Would we say the same
thing if we just confirmed the cardiovascular safety
of all these drugs? Because the reason these
guidances were put in place was to confirm the safety and
I’ve been on two FDA panels; not in diabetes, but in other fields, some more related than others; where we’ve gotten asked
this same question. We have hundreds of patients
that confirm the efficacy. There’s some safety signal, some potential
cardiovascular safety signal. Should we be doing 10,000 patient cardiovascular safety studies in all drugs in obesity for example, is the
place where this has come up? And I do think Eric, to your point, some of it does depend on
the prevalence of the disease and the major morbidities
of the disease we’re study, obviously, and common diseases where there are a lot of
drugs that are gonna be used for long periods of time
in millions of people, it’s quite feasible and maybe reasonable to say we should confirm safety. And it’s not just a question
for cardiovascular safety. There’s which kinds of
cancer does a drug cause and how carefully should
we adjudicate that and you can go down the list. As cardiologists, we all think cardiovascular safety’s
the only important thing, but there are other kinds of safety we should be worried about as well. – It’s a great point, isn’t it? And this, in fact, was
brought up in especially, I think, some of the
pharmaceutical sponsors asked this question– – Can make one quick before we– – Can I just finish. (audience member speaking
indistinguishably) Let me just finish. Let me just finish and then your … That one of the reasons
that there’s a pushback against this is that it’s not being fair, is that we’re being capricious and why aren’t hypertension drugs, why aren’t lipid drugs,
there are lots of other drugs where you can say the same thing. So I think it’s an important issue. – Before we get too f– – [Audience Member] The
microvascular benefits. – Yes. (audience member speaking
indistinguishably) – [Man] And what percentage is this. – That wasn’t my point,
but I will speak to that. – [Audience Member] Already,
in context, useless. Obviously now we’ve seen
their positive ones. I think the prelim to this
was really around the glucose, the microvascular benefits,
which as cardiologists– – Ask about heart failure. – Before we get too far down this road of trying to decide
whether or not these trials have any value is I’m gonna
make it my personal mission to change the slide that Neha shows, because what that leaves out are the trials that have actually resulted in discontinuation of drug development, because of signal of harm. For example, muraglitazar,
which was a dual PPAR agent which was designed specifically,
both to lower glucose and improve various lipid parameters with the intent of reducing
cardiovascular risk actually had its development discontinued, because of signs of cardiovascular harm and unexpected safety
related related signals like GI hemorrhage. So that drug didn’t make it to market and was an important trial to perform, for the safety of our patients. And then there was another
drug in an emerging class– – Aleglitazar too. – But there is a drug in a different class that more recently was
taken off the market too, so that needs to be on the slide, because it’s lost in the conversation, but the microvascular complications. It’s almost as if these
drugs we’re talking about, because they don’t reduce
cardiovascular risk, are worthless. So that’s sort of, I think,
the point Rob is trying to make is that these drugs are approved because they lower glucose and, in fact, lowering of glucose and A1C
is a pretty good surrogate for reduction of
microvascular complications. That’s been proven in multiple
well conducted trials. – Maybe you could explain what those microvascular
complications are. – Retinopathy, risk of blindness, renal disease, nerve damage in the feet. These are nontrivial complications. They’re not death, but they’re things that our patients would like to avoid. – The other thing from
a policy perspective, and this is interesting,
you might argue that and even the sponsors at the meeting somewhat tried to do this,
but made the argument that this regulation had
ultranegative effects in that inhibited innovation
and prevents industry from investing in these drugs, but yet you kind of have
trouble to believe that, because, in fact, 15 or
more drugs have been tested. 17 or 18 ,maybe if you take the negatives, that have been tested and brought forward through this same pathway. That’s a lot of drugs, almost many more than a
lot of different fields, so in part reflects great pipelines and other things that are out there, but it also doesn’t seem
to reflect that the hurdles of getting past this were high enough that they prevented
companies from doing it. Nowadays actually what you’re hearing is there are things
that are preventing it, but it may be the fact that other drugs are having such success that there’s concern, can I do even better than some of the results we’re seeing. So it’s inhibition by superiority, rather than, necessarily by regulation. – One thing that you brought up, Neha, was the cost of these drugs and the one that’s being used most has no cardiovascular benefit, but costs the same as the others. So what’s the average monthly
cost for these new drugs, $200 a month, $300 a month? – Oh, you’re at the BA system. I think it was something around 500. – Are they covered by
most insurance plans? – You have to get a prior
auth for most of them, So thats not necessarily for any of us who have tried to do it. – What’s the copayment most
patients who have insurance would have to pay for one of these drugs? – 25 to 50 dollars. – Yeah, I think it’s about the same. – [Audience Member] Pretty cheap, so– – For the copay. (several people talking at once) – For the provider, that
takes time and energy that we are all under the gun for. – Don’t you want your
patients to be alive? – Yes. (laughing) – I’m telling you that the point being when you introduce these
steps, there’s inertia on the part of the provider to figure out how to get them over that inertia, through education and so forth, but all these drugs are branded. Non of them are generic yet, right? – No. – So there’s gonna be a copayment cost and there’s a preauth
side and those things influence the prescription
use of the drugs. We can do a lot better
than we’re doing right now when we recognize that … I’ve seen a number of my patients refuse to take the drugs,
because the copayment that they (speaking to
softly to understand). So there are those issues that are in play that we need to work on. – So GoodRx at Costco for example, Empagliflozin is $449
dollars with the discount. So they’re expensive. – That’s out of pocket. – It’s not necessarily out of pocket, because that’s … – But that’s a total cost. – Total cost, right. – Before we quite get there, the other question was raised earlier was is this something we
should do for other drugs. The lipid drugs don’t
have the requirement, for example, on them, but yet pretty much almost
every one of those drugs has gone through a longitudinal evaluation in tens of thousands of patients. So in essence, to prove a benefit
to get market penetration, most of them are gonna
have that data on it. Two, we focused in on
cardiovascular disease and we didn’t focus in on other
conditions and that’s true. It does affect, in part,
how we collect the data during the study itself, but if you study tens of thousands of patients in a trial and are looking at what happens to them, you often pick up the signals even if you’re not
perfectly tuned to do that. For example, even the amputation thing was an unexpected finding. We didn’t collect a lot
of detailed information in the first studies on
those types of things, but we were able to get a signal that there was some concern there. So even though imperfect,
it’s pretty darn good to follow that people and
having randomized data in comparison to a standard
of care is pretty good. Should we do it for other
large common conditions? Well maybe. It’s interesting. I don’t think people,
again, initially anticipated seeing this big of a
differential effect on outcome with these drugs and quote,
unquote off target effects that we did not anticipate, why there was cardiovascular benefits, for example the empagliflozin
or other of the agents. So would we anticipate being
differences in other fields? It has, in part, been required or seen in certain cases with
NSAIDs, as one example, nonsteroidal antiinflammatory
agents you get. So there’s been studies that have started to look
at those kind of conditions. Also obesity seems primed for it, because of the potential
both benefits or risk associated with those
agents are pretty darn high, but agents that are gonna
be taken that are common, that are gonna be taken
for a long time period, both are practical to do
them and maybe are important, because there usually is the money associated with long term use. – [Audience Member] So if
I have two diabetes drugs and I’m gonna design the next trial to look at the cardiovascular
benefit of this drug, what’s it compared to? It should be active comparative trial now. What is the comparative, empagliflozin or one of the other drugs? Jennifer, I’ve asked you that question. – I don’t think it necessarily has to be one of the drugs with
demonstrated benefit, but I think definitely at
least a drug with known effects as best as we can determine them and potentially a neutral effect wouldn’t be an unreasonable comparator. – But it’s an active comparator, it’s no longer against placebo, right? – I think it depends. It’s going to depend on what you allow their other care to be. If you decide that you’re gonna
enroll a patient population that now has an indication
for drugs like empagliflozin to reduce cardiovascular risk, you could permit that
therapy in the back ground. It doesn’t necessarily
have to be your comparator. – [Audience Member] Used in
less than 1% of patients. – Right. – Right now, as we go forward
it will hopefully increase, but it was very relevant in EXSCEL. So in EXSCEL and think Rob can probably speak to this better, but in EXSCEL the control arm, which was placebo on the
background of standard therapy, had a significantly higher rate
of empagliflozin new starts than the exenatide arm and
that probably contributed; and I don’t know if that
study has been completed yet, but we were looking at that
drop in therapy question; might have contributed to why we didn’t see a difference in MACE. – The point being is that these trials take three to five years
to run and complete, so what you start with as a comparator may actually be obsolete by the time you’re
finished with the study. We have to keep that in mind. So I think it pays to be a student and also the FDAs never been
in favor of adaptive designs, but looking forward in the future, I think that this might be an area where designing a smart trial that could reflect changes in practice during the mid course of the trial may be something that
should be considered. – And there are a few questions here Matt. Maybe I’ll push this back to you. The question is what’s the role of the FDA in mandating what should
the comparator be, because we still have to go back and say the main reason the
guidance there is assure that these drugs are not
causing cardiovascular harm, so should there be the option about whether to use a
active comparator with a drug proven to safer, should
that be a requirement. So that’s a regulatory question. Then the clinical question is, is there there equipoise still? Can you randomize someone to not be on one of these drugs and the answer, probably right now is yes, because only 1% of
people are getting them, so you probably still could now, but maybe not a couple of years from now. – [Audience Member] If you
goal is to show safety, you wouldn’t want to randomize a drug that has no cardiovascular
benefit against empagliflozin, because a drug that has no harm signal might look worse than empagliflozin. – You could be strategic about it and decide which comparison
was most important to you. You could have various
degrees of permissiveness regarding background therapies. It would depend, I think, on the patient population enrolled. I don’t think we can prevent
patients with indications for particular drugs from getting them, within the context of a clinical trial. I think that’s kind of the bottom line. – But I think the regulatory question is a really important
one and it depends on what type of label the company’s seeking. The FDA will always defer to
what the standard of care is. So for example, when we’re looking at more novel P2Y122
inhibitors, they obviously had to be compared to clopidogrel right, because clopidogrel was the
only one that was in use, but it was a recommended therapy that was needed in the post ACS setting so those trials had to
go against clopidogrel. And if you wanted to show superiority and have that be your label claim, you had to use what was the best available active comparator, but the FDA generally stays out of it until it gets to the type of label that
the company’s hoping for. So if you have the next
generation diabetes therapy, you want to show you’re better than empa and that’s gonna be in your label them empa’s going to be
your active comparator. So the FDA, just so everyone realizes, they don’t set the rigid standards. It’s more they have a
discussion with the company of what they’re seeking in their labeling and then you design the trial based upon the positive result
resulting in the label change that the company would hope to get. – The only thing that’s tricky here, is we’ve already seen in
the linagliptin program, that the initial CVOT was designed to be compared sulfonylurea and then, in fact, the company
was required to initiate a second trial in which their
drug was compared to placebo. So they’re doing two. – Yeah, but it was the second trial. They did an actual midcourse
change the first trial. – Right, but the original thinking was that they’d be able to
use an active comparator, which I think is really going to provide a lot of valuable information, but again, that wasn’t considered acceptable at the end of the day by the FDA. – To this discussion
including about lipids, for lipids now you have to have a trial looking at cardiovascular outcome otherwise clinicians wouldn’t use them in and payers wouldn’t pay and you’d hope we’d get to
that point with diabetes drugs, where the marketplace and the patients and the clinicians would
say, “You have to do this “or else you’re not gonna be able to “sell a drug for $500 a month.” – The only thing that I
would like to bring up, as an endocrinologist,
is it would be wonderful if when people were
diagnosed with diabetes they could be started on one drug shown to demonstrate
cardiovascular risk reduction in addition to glucose lowering and have their diabetes
controlled forever on it, but that’s just not the
way that it happens. So very likely people are gonna end up on combinations of many
drugs and I personally don’t have a problem with someone with an indication being on a drug that reduces cardiovascular
risk in addition to a drug that has demonstrated
cardiovascular safety. I don’t think that’s unacceptable. So I think there is a role for drugs that lower glucose, that
don’t necessarily also reduce cardiovascular risk independent of that. – I think also the other concern here; and Chris there was a
slide from the meeting, I don’t know if we have; of
the current diabetes guidelines and it had multiple colors. It was like the one
that had all these boxes and it’s really hard to make any sense out of what the recommended
diabetes treatments are right, especially from that slide,
if you remember that. If we don’t have it, maybe we
can share that with everybody. Someone once said that ACS stood for acute confusional state when there are a lot of drugs being
developed for the same thing. I looked at that slide
as a cardiologist going, I don’t know what the heck I’m gonna do to treat a diabetes patient, when the ADA guidelines
have multiple stacked layers of five or six drugs for patients. It was just very confusing. – We’ll get back to this,
but it’s one of the reasons I think, as cardiologists, we need to be working with diabetologists to come up with pathways, because it is gonna be complicated. It’s not gonna be incredibly simple. Neha, can you summarize Because there were a few things, I think, that we agreed on in the think tank, that could be changed to be improved and there were a lot of
questions that were raised that we need more information on and we’ll bring Rob Mentz
into the discussion too. Can you summarize, what did people agree? – Not much. (laughing) Again, this is the only one I’ve been to. It’s basically people
are sitting around this huge U shaped table and
it’s like 50 or 60 people and I’m at end and 50
or 60 people are talking and nobody’s agreeing on anything, but really interesting points and a few things did get agreed upon. I will. I’m just giving the full background. (laughing) one thing was that, in
terms of the endpoint, the composite endpoint that
the FDA is kind of mandating; they mandated at least MACE,
maybe you could consider some other outcomes; I
think everybody agreed that hospitalization for unstable angina is relatively useless and
should not be included in the composite endpoint and also that heart
failure hospitalization and renal disease progression
probably should be included, at least for the drugs
where it makes sense, but that that’s something that should be seen more regularly. So that was one of the things that, probably the only thing
that was truly decided upon. There was definitely a lot of discussion about what the comparator should be and there was no decision or consensus about what the comparator should be, but there was a lot of discussion about active comparators and maybe things should be moving that direction. And another point was this idea of when all of these trials
have shown safety, do we still need to the 1.8
safety margin at this point or can we just move to 1.3 and allow marketing to happen and move to 1.3 with the promise that this non inferiority trial with the 1.3 margin would happen. There wasn’t complete
consensus about that, but a lot of people were
kind of leaning towards that. – Can I add one? The only other thing that I can say that was universally agreed upon was that randomization is key and that observational analyses and comparative effective
analyses are helpful, but cannot give you an answer that you could use to change
guidelines or change practice. And remember, who was it
that got up and started … Was it Hertzel Gerstein or someone who’s like, “Randomization”? He was like a cheerleader,
he stood up and I mean he was right. Randomization is key and you have to do that right, to have a randomized trials
to answer the question and you can’t get the proper answer from an observational
non-randomized analysis. – This slide that I just
put up behind you again, it illustrates, I think,
that there are a lot of important outcomes and there’s
really variable effects on these outcomes, both in the drugs and maybe even in some of the drugs in same class. Some drugs have caused
increase in heart failure and some drugs have caused
a remarkable decrease in heart failure. And Rob, you thought a lot about this. Is it real heart failure
that they’re preventing or is it just because
they’re getting rid of fluid or how do we deal with
this heart failure issue and is it important? – I think the short answer’s
yes it is important. I like the comment about the
importance of having it as, potentially, within the composite or a key secondary endpoint. The landscape has changed a lot. Initially there was so much
concern when SABER came out that oh my gosh these DPP4s are causing all this heart failure then, aside from alogliptin,
you really do see that TECOS showing a very
safe hazard ratio of one. So I think getting at this idea that it’s not just a class effect, there are differences within the class, heart failure is important. And it comes up now, as
we’re taking care of patients and for the first time over
the past couple of weeks I’m prescribing empaglofloxin for patients that have heart failure and diabetes and I think the key piece is that we have to have
those definitive studies. So there are now large
heart failure studies that are looking at empaglifloxin regardless of diabetes status. It’s called the EMPEROR program and that’ll be the definitive answer. Right now the U.S. heart
failure guidelines, they say start with metformin and they don’t really discuss
these CV outcome trials, whereas the European ones,
they actually do already say it’s reasonable to consider
using empagloflozin, since some patients in those
studies did have heart failure, but it’s not well care drive, so it’s a checkbox on the case
report form of heart failure. We don’t know as well about REF versus PEF and reduced versus preserved EF. And them I’m gonna have one
self serving comment as well. Jennifer and I are leading this diabetes cardiovascular disease symposium in April, where Chris and many others
are gonna be talking at. It’s April 7th at the JB Duke and I think it’s really gonna extend
some of the conversation you guys have already led and hopefully will be another important piece as we think about how Duke
can position ourselves in really getting out the message of how to use these drugs, implementation. And one final comment around this. I wanna get your guys thoughts on, you mentioned yes, randomized data are kind of the linchpin of this, but with CVD-REAL and
others, observational and real world evidence, I
think, have a piece in this that can be complimentary, but I think as we look at these data, I think you talk nicely
with sustained and others. This is a much smaller study. The effects sizes almost
seem too good to be true, compared to the more
adequately powered ones. So this idea of, as an institution, we’re trying to advance
science in this space. We’re gonna be leading clinical trials, but understanding the role
of real world evidence and how we fit in that and place
it in the important caveats. – Yeah, I’d say you’re right. It’s the spectrum and it
does have an important role. I would add that you made
a very important point. These drugs are now being tested in populations that don’t have diabetes. Five years ago did we ever
believe that would happen? That to me is the most remarkable thing, but I think Eric probably
can speak better than anybody about the cycle of evidence
in different types, but I’m amazed now that we’re doing trials of diabetes therapies in
patients who don’t have diabetes. It’s great. I love that. – It makes the discussion about how expensive these trials are a little bit harder to swallow. If there are opportunities
and new directions for these agents, it could
be financially beneficial. – Raj had a question. – [Raj] Just going from the
private practice standpoint, I face this problem every single day and who’s responsibility it is. If I try to cross the line and and say, “No, you need this drug. “This has been shown”, when the endocrinologist
with whom the person is not following and the
primary care is not following. We had these issues with cardiac surgeons and cardiologists and we have a heart team and should we have an endocard team now to make decisions about this thing, because this is really what
is the cost is one issue and the other major issue is
am I stepping on someone’s toes and we get really nasty phone calls and in private this is very important. In academia, if you lose a patient or 10 patients or 15
patients doesn’t matter. – Perfect segue. Perfect segue. Jennifer, said differently, all this work has been billions of dollars spent. We finally have something to address one most important causes of death an disability in the world and Duke hospital doesn’t
have them on formulary. How can we fix this? – I know. I’m arguing with individuals
in my own division about the value of this information and the importance of
incorporating these drugs into the care of patients with established cardiovascular disease. I sort of feel like we’re just starting to push the boulder up the mountain and I think it’s all of
our responsibilities. There will be pushback. There will be pushback every day, but we just need to try
to keep moving the needle. – Chris, can you send
the link to this video to the chair of the P&T. (laughing) – [Audience Member] Can
you comment on what effect all these drugs have on glucose? One of the things that I’ve
been told about these drugs is that they don’t do
a whole lot to glucose and maybe also in glucose
in people without diabetes, because that has a big influence, I think, on peoples’ willingness to prescribe them. One of the concerns I have
about prescribing these drugs is that their endocrinologist manages all their other diabetes drugs and I’m worried I’m gonna
make them hypoglycemic by putting them on something like this. – That is a great question. The good thing about all of these agents is let’s say if you
gave them by themselves or with a drug like metformin that doesn’t cause hypoglycemia, none of them will make your
blood sugar drop too low. It just doesn’t happen. The mechanism of action is such that they’re kind of smarter drugs, they won’t drop your blood sugar too low, or they really shouldn’t. Every now and again you
hear something strange. So these are drugs that
are really pretty safe, from a hypoglycemia perspective, however, if you have a patient
who is already on insulin or sulfonylurea and you add any drug, the risk of hypoglycemia goes up. That’s part of their
care equation already. You have to be very cognizant of that and you will need to talk with whoever is managing their diabetes, unfortunately, to decide if it’s okay to do that. But these were all pretty
potent glucose lowering agents, especially semaglutide really
brings blood sugar down a lot and it’s not really clear if
the cardiovascular benefit is necessarily related to that,
it could be a piece of it. But as far as glucose control overall, they’re really very, very effective. – I think this raises a good point. As providers, the thing
that drives our behavior, we showed this over an over again, the first thing that drives it is do no harm or non-maleficence. We’re more worried about giving a drug that can cause harm and
hypoglycemia is the harm that we’re worried about. We not so worried about hyperglycemia, because these patients are
unlikely to go into DKA unless something crazy happens, but we’re more worried
about the hypo episode that could lead to death. I think that’s the mentality
that we have to overcome. Most drugs that are new that’s
the thing that do no harm, people are worried about
prescribing the drug and cause bleeding for
a new antithrombotic or causing hypoglycemia. That limits our use. So we really need to think,
in our implementation studies, about the physician mindset
of when and why to start as well as how to interact
with the other specialties to overcome some of these barriers. I think that’s a really key point, that suddenly we don’t even think about, that influences all of our behavior. (audience member speaking
indistinguishably) – We have a couple of minutes left. I’m the moderator. I’m the moderator. And I’m gonna give you the last word Eric, because I’m gonna ask you. But I hope this has been
interesting for people. One of the things that the think tank, in today’s session, has highlighted is there’s a huge amount of opportunity to better understand and do
more research in this area, including around heart
failure, renal progression, understanding combinations of these drugs, but probably nothing more important than how do we get things
implemented more effectively and make sure that patients
actually are achieving the benefits, not just
of these treatments, but of ways to prevent
cardiovascular outcome related to diabetes and
we’ve got a lot of talent here on the panel addressing that. Eric, can you summarize what do you think the DCRI should be doing? – I think it’s doing it actually. If you wanna take it to the extreme, a few weeks ago or last week it was whenever you did the talk
on atrial fibrillation and we summarized
incredible circles of data from very early studies that we were doing to all the different therapies
and how we tested out and proved they’re safe and effective to ultimately now the implementation work has gone a long way to
changing how and what we do with atrial fibrillation and is preventing strokes
across this country. We should be very proud. With diabetes, it may even be broader, because we went back to
actually the formative stages of the regulation that created this field, we were a part of and helped drive. Now we carried out some
of the earlier studies and continued to carry
out the large studies that show the benefits
of these agents or not and now we’re just at the beginning of that final and very important step of how do we get these
drugs more commonly used and look at all the issues
that we have in other fields, but take the lessons learned
from those other fields and now apply them to diabetes. – Final questions or comments? – I’ll say that one
thing is that we do have a whitepaper that will be coming, as we do after each on,
and Neha has promised to have it submitted within four months of the finish of the think
tank and the whitepaper is an enduring material
from every think tank that hopefully everybody
will get a chance to see. I’m sorry to interject, because I know Marty had a comment, but I hope you’re already
working on writing that. – Yes. – [Marty] Maybe this is a
question for Dr. Green and (microphone cutting out)
the challenge they have is not as much the specialist
as it is the primary care. Who is using the combo therapies and is it endocrinologists
and cardiologists they should be focused on? Does it start there or does
it start in primary care? – One point that we haven’t mentioned is that it’s really
important to implementation is getting this information
into the guidelines in a simplified and user friendly way and in a way that is
palatable to primary care, because that is the setting in which the vast majority of diabetes care occurs. So yes, I think that needs
to be a key component of whatever we end up doing. – I’d like to thank you
all for participation and our great panel and have a good day. (applauding) (upbeat music)

Leave a Reply

Your email address will not be published. Required fields are marked *