Cardiologists are concerned about bleeding associated with the use of
increasingly potent antiplatelet therapy used to prevent ischemia-related
events.
Carl J. Pepine, MD, chief medical editor of Cardiology
Today, moderated a round table discussion that included Stephen
Wiviott, MD, associate physician at Brigham and Women’s Hospital and
assistant professor of medicine at Harvard Medical School, Sunil Rao,
MD, assistant professor of medicine at Duke University Medical Center, and
Roxana Mehran, MD, associate professor of medicine at Columbia
University Medical Center, during the
American Heart
Association’s Scientific Sessions in November in Orlando, Fla. The
panel discussed bleeding associated with medications for acute coronary
syndromes.
In part one of this two-part round table, the group considered the
optimal antiplatelet strategy, classifications proposed relative to bleeding in
and around interventional procedures, the relationship between bleeding and CV
outcomes and existing study data. Part two will appear in next month’s
issue of Cardiology Today.
Carl J. Pepine, MD: The question on the table is what to do about
bleeding related to the treatment of patients undergoing percutaneous coronary
intervention for either acute or chronic syndromes? These patients are treated
with multiple drugs that interfere with platelet aggregation, what do you think
is the optimal antiplatelet strategy? How does this impact ischemic and
bleeding events?
|
Stephen Wiviott, MD, participated in the round
table discussion in November.
Source: Christie’s
Photographic Studios |
Stephen Wiviott, MD: Over the last several years, we have
recognized the importance of both ischemia and bleeding. In the area of
antiplatelet therapy, beginning with aspirin and then extending out to novel
compounds with more intense antiplatelet therapy, there have been increases in
bleeding associated with the reductions of ischemia. I think it’s not
straightforward to answer what the ideal antiplatelet therapy is as a general
statement, but with the goal of reducing both important ischemic events while
maintaining adequate safety it needs to be tailored for both individual
patients and individual situations.
Sunil Rao, MD: I certainly agree with what Steve said, but there
are a couple of issues here. No. 1 is why has bleeding all of the sudden become
the topic of discussion at these major meetings and around the clinical tables?
The second issue is the window during which the bleeding may occur, both
acutely as well as chronically. Because of the development of these powerful
agents and combination therapies along with the evolution of devices and
smaller catheter profiles, the reality is most patients after PCI do extremely
well with respect to ischemic events. What has started to bother
interventionalists — and complaints we hear from patients — are
bleeding complications post-procedure. There’s the acute risk, and then
chronically we get a lot of these strategies, like the trial that Steve’s
been involved with, that involve therapies that are taken for a year, two
years, maybe even longer. So, there is a chronic bleeding risk that becomes
manifest. I think the words that Steve used are very appropriate, which is
important ischemic event, and I would translate that to important bleeding
event because all of these drugs are going to increase bleeding risk. The
question is what are the events that are important, both to physicians and to
patients, and that may be different actually. What patients complain about may
be different from what we think of as important bleeding events.
Roxana Mehran, MD: You both make very important points, but I
think the bottom line today is tailored therapy for patients presenting in
different clinical scenarios. The days of thinking of bleeding as just a
nuisance are over, and it is now clearly established that bleeding is as
important as ischemic complications when we treat patients with antiplatelet
agents.
|
Carl J. Pepine
|
Pepine: Steve, we’re using this term ‘bleeding’
generically, like everybody understands what it is. Can you tell us some of the
classifications that have been proposed relative to bleeding in and around
interventional procedures?
Wiviott: We have used the TIMI bleeding criteria, developed
several years ago in the time of thrombolytic therapy, and in many ways
identify the worst bleeding episodes. TIMI major bleeding is bleeding
that’s associated with either intracranial hemorrhage or a hemoglobin drop
of more than 5 g/dL associated with a clinical syndrome. TIMI minor bleeding,
and perhaps minor isn’t exactly the right term in terms of the emotional
impact, is a hemoglobin drop of 3 g/dL or more. Below that, we previously
termed minimal bleeds, but within that group of bleeding events, it’s a
broad array of different things from patients who have enough bleeding that
they have to stop their medications, to events that are seemingly trivial such
as bruising one’s arm when bumping into a table. Our bleeding definitions
have been sort of the tip of the iceberg, and in the lytic era also came the
GUSTO bleeding definitions. GUSTO bleeding definitions were more clinically
based as opposed to laboratory based. For instance, GUSTO severe bleeding would
be things like intracranial hemorrhage or also bleeding resulting in a
hemodynamic compromise that required multiple transfusions or other types of
interventions. More recently, there has been a move toward more sensitive, more
inclusive bleeding definitions. Sunil can correct me if I’m wrong, but the
ACUITY definitions have been sort of the standard for these more inclusive
definitions, and they contain things even in the more major categories, like
large hematomas or even modest-sized skin-related bleeds, in addition to the
higher level bleeding. So certainly, as you might expect, with these different
approaches you get vastly different estimates of how frequently bleeding occurs
and also how likely the value to the patient in terms of quality of life and
clinical outcomes differs depending on the severity of the bleeding scale used.
Rao: TIMI and GUSTO certainly served as the foundation, and much
of these subsequent bleeding definitions that have been developed build on
those. They include elements from those and changed them a little bit; for
example, OASIS. What’s happened over time, as Steve said, is that there
has been this sort of gradual broadening of what we consider to be a major
bleed, for better or for worse. You say we have been able to reconstruct all
these definitions. Now, the problem of course is that I’m not sure anyone
in clinical practice actually looks at a patient and says, ‘Well,
you’ve suffered a GUSTO severe bleed or a GUSTO moderate bleed.’ The
challenge becomes, how do you translate that type of safety data into a
clinical realm? Perhaps other scales may be important here. There has been a
scale developed that has ‘nuisance bleeding’ as part of it. This is
not the greatest title because in reality nuisance bleeding is bleeding that
happens to somebody else. But it involves these skin ecchymoses and kinds of
things that patients complain about. In at least one single-center study, 10%
of the patients who suffered a nuisance bleed stopped taking their medication.
There seems to be some clinical impact of these more minor or less severe types
of bleeding complications that may have relevance in the clinical world, and so
we’re expanding this outside the ACS realm, for example, to AF.
Steve’s group did a phase-2 trial in AF, where they used a specific
bleeding definition called ‘bleeding requiring medical attention.’
Wiviott: Right, this was the ATLAS study, and the bleeding
requiring medical attention definition is one that separates TIMI minimal
(<3 g Hb) into those that seem to have an immediate clinical impact and
those that do not.
Mehran: The trouble we have today is that when we talk about
bleeding we have no way of comparing agents across different trials because
each trial has used different definitions for major bleeding, and there is no
real consensus in which bleeding is important and should be considered major.
There is no question that the most widely used definition, TIMI bleeding, is
just the tip of the iceberg. There must be an effort to standardize these
definitions so we are all on the same page when comparing agents and their
effects on bleeding.
Rao: But that’s an important definition because that
actually, in the title itself, says that this is an important bleed to the
patient. They actually required medical attention for this particular bleeding
definition. That may actually be the future of bleeding definitions.
Pepine: Even today we’re starting to realize that patients who
bleed don’t have good outcomes. Not from the bleeding, per se, but from
the CV outcomes. Do you want to comment on that?
Wiviott: These are complicated relationships. When you’re in
a trial or in a clinical setting of comparing endpoints as it relates to
bleeding, certainly there are terrible bleeds, like intracranial bleeds. There
are bleeds that result in exsanguination, there are bleeds that result in
catastrophic events that cause people to die.
Rao: The reality is that these catastrophic bleeds are rare.
Wiviott: Yes, those are quite rare. There are also bleeds that
result in stopping medications that have the potential to help patients in
terms of reducing ischemic events. For instance, patients with stents who have
bleeding events stop their clopidogrel and then have a stent thrombosis. That
bleed indirectly led to that ischemic event. There is certainly a relationship
between bleeding and adverse outcomes. It’s also the case, though, that
people who do bleed are patients who are frailer, older, and at higher risk for
other bad things happening to them. This is what makes the relationship between
bleeding and ischemic outcomes or death difficult; the same people who are at
risk for bleeding are at risk for MI and death, but it may not be a result of
the bleeding.
|
Sunil
Rao
|
Rao: Yes, I would actually agree with that. The agent that has
the least amount of bleeding is placebo, with the exception of the OASIS 6
trial. Fundamentally, we have to think about why we are treating these patients
with these agents. The reason we’re treating them is because we want to
reduce ischemic events. Now, doing that while minimizing bleeding risk is
ultimately where we want to get to. I completely agree with Steve. A lot of
what we’re picking up here are associations which have been consistent
over all the studies. Is it really causal? Probably not. In the rare instances
in which you have intracranial hemorrhages or perhaps, frankly, fatal bleeding,
they are probably causal. So why do we have the association? A large part of it
is what Steve said, these are sicker patients in general. But there are other
mechanisms that may be involved here. We talked about cessation of medical
therapy or lack of adherence to evidence-based therapy in patients who bleed.
That’s probably part of it, but the one thing that has been interesting
over the last couple of years is that there are certain therapies where an
association between a reduction in bleeding risk appears to be also associated
with an improvement in survival. I’m choosing my words carefully here
because, again, it’s really hard to draw causal inferences. But there are
strategies that appear to reduce bleeding risk that may also improve survival.
When you have multiple strategies, including using the radial approach, you can
reduce bleeding risk by 60%. It seems to me that bleeding reduction should be a
goal whether or not it’s associated with improved survival and if you have
options to do that, then perhaps that should be the way to go.
Wiviott: One thing worth adding in this area is that, in general,
the strategies that have shown reduction in bleeding and improvements in
mortality have been neutral with respect to ischemic events. The difficulty
comes with most agents where you get a more potent effect on platelets or on
anticoagulation, which reduces ischemic events at the expense of bleeding, in
which case these balances of trying to minimize bleeding are not so clearly
associated with improved survival. It is when you can keep ischemia constant
while reducing bleeding or vice versa, improve ischemia without increasing
bleeding; that’s the Holy Grail, and we’ve seen that with the
antithrombins, but from the antiplatelet standpoint, we haven’t seen that
yet. That remains a goal; to alter the balance in one direction without tipping
it the other way.
Rao: So what’s your read on the ticagrelor data?
Wiviott: In the PLATO trial, which is what we’re referring
to as comparison of ticagrelor, a more potent antiplatelet agent, with a
standard-dose clopidogrel in the primary bleeding endpoint, a composite of
non-CABG-related bleeding and CABG-related bleeding, there was a neutral
effect. When you look at the non-CABG-related bleeding, 95% of the patients in
the trial didn’t have CABG. There was the expected 25% to 30% increase in
bleeding. In fact, it was exactly the same change in TIMI major bleeding as was
observed in the TRITON TIMI 38 trial of prasugrel, which is often pointed to as
a drug that is substantially increasing bleeding. Ticagrelor is a reversible
agent, so when stopped prior to CABG, the drug effect wears off faster and it
would be expected to have less bleeding. In 95% of the people in PLATO, those
not undergoing CABG, it did just what you’d expect; it reduced ischemia,
increased bleeding and, we haven’t seen this yet, but, unless the survival
benefit is wholly contained within that CABG population and something different
is happening here with regard to mortality, just that, bleeding isn’t
present. That’s actually, in some senses, the best example in my mind of a
drug that increases bleeding and improves mortality.
Pepine: Then you have the opposite with dabigatran, where you see
better outcomes all across the board, including a reduction in stroke and
bleeding. Is that just a case where the comparator is bad?
Rao: It’s interesting isn’t it? We like to beat up on
warfarin, but as far as I can tell, this is the first, perhaps viable,
competitor to an agent that we all hope to replace someday. There are a couple
of issues here. It’s a different patient population, it’s a different
setting and it’s a different comparator. We talked a little bit about how
we’ve become so focused on bleeding only recently in the world of ACS, but
the reality is we’ve known about bleeding with warfarin for a long time.
There are validated risk models for bleeding and for stroke with warfarin
trials. We’ve actually been able to define something we can’t get
defined with antiplatelet therapy, which is, what is that therapeutic window?
With the new agent, dabigatran, it certainly does seem to push everything in
the right direction. It seems to reduce stroke, reduce bleeding risk. It may
have to do with the fact that with warfarin, the therapeutic range fluctuates a
little bit more than perhaps a reliable agent, like dabigatran.
Pepine: So let’s talk about some specifics. You have a patient
with ACS. You anticipate that the patient is likely to need a stent.
What’s your order for that patient?
Rao: I’m very aggressive with early loading with
thienopyridine.
Pepine: Before aspirin?
Rao: They get aspirin en route. They get oxygen. About 70% of our
patients with ACS get unfractionated heparin and about 30% get
low-molecular-weight heparin. That decision is usually left to the ED. But all
of them get pretreated with thienopyridine prior to coming to the cath lab.
That’s been a relatively recent change. Our use of IIb/IIIa inhibitors has
gone substantially down and our median time from presentation to cath for a
patient with non-STEMI and ACS is still about 20 to 21 hours. They’ve had
a full 24 hours of thienopyridine, dual-antiplatelet therapy on board before
coming to the cath lab. Once they’re in the cath lab, the antithrombin
choice then becomes the roulette wheel where whoever happens to be doing the
intervention will give the antithrombin of choice. Some of us will switch them
to bivalirudin. Some will continue the heparin. Some will still use IIb/IIIa in
that setting, although most of us don’t. My default approach is the radial
approach. The only thing that’s uniform is that the patients are getting
loaded with clopidogrel early.
Wiviott: With what dose of clopidogrel?
Rao: That’s a good question. We have adopted a 600-mg
loading dose. Not only is that off-label, but as a general strategy does not
appear to help the sort of broad population of patients undergoing PCI unless
they are getting high-dose aspirin as well.
Mehran: We have an aggressive approach as well. All patients with
ACS receive a loading dose of 600 mg of clopidogrel and 325 mg of
acetylsalicylic acid regardless of whether they were on chronic therapy at
home.
Part two of this round table will appear in next month’s issue.
