Bleeding Definitions in Clinical Trials

Thanks to Monica Lewinsky’s boy-toy, we can now say with confidence that the meaning of the word(s) “bleeding definitions” depends on what the meaning of the word(s) ‘bleeding definitions’ is.  Sound complicated?  I feel your pain.   If you ask Michelle’s boy-toy he would clear it up for you in three words:  “It’s Bush’s Fault”.  But as with everything else he says, it’s untruthful, and more to the point, totally unhelpful.


Truth be told, it’s Obama’s fault that you don’t understand why bleeding definitions are so complicated.  And clarification on this topic comes along once in a blue moon.  But guess what?  Today’s your lucky day!  BOLT’s Anticoagulants Thought Leader Panel (btw… contact me if you want to be alerted when our current panel publishes in a couple of weeks.  It’s REALLY juicy!) has some words of wisdom that just might help.  So here’s what the Panel had to say (these WOWs are from a report that we published about a year ago):



  • We preface our section on bleeding by saying that safety trumps efficacy and that is why our Panel focuses so much on bleeding.  HOWEVER… the reason why Panelists are using anticoagulants is that they prevent strokes.  Every Panelist can relate a story about a patient who was taken off their anticoagulant due to bleeding – who then had a stroke.  The point:  There is no win-win situation, and the physician will balance the risk of each adverse event for each patient.


“I think as cardiologists we want to reduce bleeding but we do recognize that there is a little bit of price to pay.  For example, if I have someone who is bruising easily when they are taking aspirin and clopidogrel after a drug-coated stent I don’t stop the clopidogrel.  But if I have someone who keeps getting admitted with repeated GI bleeds, then I have to rethink the whole equation there.  What do you do? You have to balance the stent thrombosis and the bleeding part.  If there is a way to treat the GI bleeding with the procedure I try to do that because I know that the risk of stent thrombosis is still high.  I have a patient, for example, that needed to be on Coumadin for atrial fibrillation but he kept having GI bleeds and he had angioectasias that we couldn’t treat in the small bowel and we just had him on aspirin and then he had a major stroke and it was a sad outcome but he was being admitted frequently and getting transfused because we couldn’t stop his bleeding.  I think the dynamic in the outpatient is a little complex.  I think we are all willing to accept nuisance bleeding, as long as it doesn’t significantly impair patient compliance with therapies.  But we are not willing to accept major bleeding that could be life-threatening or lead to admissions and that is where the crux is.  As far as the orthopedist they typically give these drugs for such a short duration they don’t see much bleeding.  Obviously they don’t want to have bleeding like if you have a knee replacement they don’t want to have a hemarthrosis develop.  That is huge.  But I think the bleeding concerns with them are different and less common because the patients aren’t taking the drugs for so long”.



“I just had one panelist that I just got off the phone with him and he said I had the guy kept GI bleeding, kept GI bleeding, kept GI bleeding and I took him off enoxaparin and he had a stroke. There is no winning in this one. So how are you going to walk that tight rope? It is the same as we do with Coumadin nowadays.  There are folks who are well indicated.  I have got a dear lady who has about a CHADS score of 7 or so.  She has had a TIA before, she is diabetic, hypertensive, and old.  She doesn’t have heart failure but she is only CHADS score of 4 and every time we give her the least little bit of Coumadin or even aspirin she has a two unit bleed.  Of course she has been scoped and she has just got diffuse telangiectasia all over her GI tract.  They can’t cauterize all these things and to eliminate it would be basically taking out all of her guts.  And so we say okay you have got a certain moderately high stroke risk and our advice to you is to get your transfusions every once in a while and she said I just don’t like doing that and I would rather take my chances.  So she is taking her chances.  That is a rare individual”.




  • From the orthopedic surgeon’s perspective there are basically three types of major bleeds.  Critical organ bleeds (stage 2 bleeds) involve the central nervous system including the spinal tract, the kidneys or the interperitoneal bleeds which can be massive and life-threatening.  GI bleeds are also very common because many patients undergoing major surgery, be it major abdominal surgery or major orthopedic surgery, often have post-operative stress syndrome.  Many of them go into a catabolic state for a couple of days, and having anticoagulants on board increases the risk of GI bleeds.  Finally, about half of patients who have knee or hip replacement have hematomas either proximally or distally.  These hematomas are quite common, and it is always difficult to know exactly when this is due to an anticoagulant drug and when it is due to the surgical procedure itself.


“It is in a critical organ and this might be life-threatening.  A critical organ is the central nervous system including the spinal tract and also kidneys and also the interperitoneal bleed which can be massive and life-threatening.  So these are the critical organ bleeds.  They are stage 2 major bleeds.  Then we still look at the bleeds in the surgical area and we also discuss the rather common GI bleeds.  Of course we as investigators we are very much focusing on bleeds in the critical areas but these bleeds are not all the bleeds to assess.  I would say at least half of our patients undergo either hip or knee replacement, especially knee replacement patients, they have rather clear hematoma in the knee area, proximally or distally.  And some of them have quite substantial hematoma.  Be it LMW heparin or other study drugs.  So these hematomas are quite common, and it is always difficult to know exactly when this is due to an anticoagulant drug and when it is due to the surgical procedure itself.  There is a great individual variation between different patients to tended to bleed.  So the GI bleeds are also very common because many patients undergoing major surgery, be it major abdominal surgery or major orthopedic surgery that have often the post-operative stress syndrome.  Many of them go into a catabolic state for a couple of days.  There is a tendency of stress ulcers.  And if they have anticoagulants on board this might of course increase possible bleeds from this area”.



“Intracranial mainly and a smattering of others; GI bleeds.  It is difficult to have massive blood loss with that.  It can occur.  You transfuse them up and figure out why they…they have a bleeding ulcer or a diverticulum or some such.  Bleeding into the urinary tract is almost always a sign of some pathology and it is a blessing to find because you catch that bladder cancer before it gets too far.  But it is the intracranial ones that are the major ones”.




  • By choice of the anticoagulant manufacturers, “major bleeding”, “significant bleeding” and “clinically relevant bleeding” are ambiguous terms.  Each chooses a definition that minimizes the significance  of their drug’s increased bleeding vs. its comparator.  Companies like The Medicines Company suggest that even mild hematomas are associated with worse outcomes.  Others say that bleeds are only significant, or “major bleeds” if they require multiple transfusions or they cause a significant hemodynamic shift.


“There is absolutely a lot of wordsmithing.  I probably can’t settle it for you.  I know that people are trying to get people to agree on it.  There are three camps.  There is the GUSTO camp, there is the TIMI camp and there is the Medicine’s company, and then there are other people that have used various other things.  All of it is bad.  But I think what remains to be seen is how mild can you go before it is not bad.  The Medicine’s Company, they seem to think that even significant hematomas are associated with worse outcomes.  And then in other bleeding definitions to really be significant you need to be transfused several units and have some hemodynamic shifts before you call it major bleeding.  I don’t think this has worked out.  I think confusion and suspicion are the appropriate ways to feel about it because there is a fair amount to be gained by being sort of unclear about this as you advance a new drug”.




  • Bleeding is not as severe of a problem to the emergency room physician or the physician managing a-fib as it is to the orthopedic surgeon or the cath jockey.  The exception to this statement is an intracranial bleed.


“In the emergency department we rarely have bleeds.  Later on is where the bleeding occurs.  It occurs in the cath lob or thereafter.  The big one I worry about is the intracranial hemorrhage.  That is the bad one. That is my biggest concern.  Do you have any procedures to be prophylactic about the intracranial bleeds? No, just blood pressure control.  And when you have them what are your options for management? If there are patients on heparin we give protamine.  If they are on Coumadin we give fresh frozen.  If the patient is on a Xa or Lovenox or something like that we just stop the drug and wait for it to wear off and sometimes give protamine too but that is about it.  Unless they are on antiplatelets we give platelets sometimes give protamine too.  Sometimes we give platelets too, ReoPro”.




  • The key point is that every practicing physician knows that bleeding in a clinical trial does not predict bleeding in the real world, because while patients included in trials always get therapy according to the protocol, in the real world the decision to actually treat hinges on the physician’s expert judgment when he evaluates the patient.  Those who are thin, cachectic, old, have cancer or other comorbidities will NOT be anticoagulated the same way as other patients, so the overall bleed rates in the real world are always lower than what are seen in the trials.  They also know that they are consciously trading off the increased risk of bleeding with the increased risk of an ischemic event.  There is no free lunch.


“I think that to some extent it is because we have overestimated the importance of bleeding.  Our fascination with bleeding arose to some extent out of analysis from REPLACE and some of the Medicine’s Company’s own trials.  One of their goals was to put bleeding on the map and they did.  Then the next thing that happened were retrospective looks at all the clinical trials and bleeding was bad and bleeding was predictive.  But when you look at that registries it doesn’t seem to be as predictive.  It is possible that the impact of bleeding is less important in real life than we think in clinical trials.  That is a possibility as well.  Do you buy that? Do I buy it?  I think there is something to that.  I think that part of the deal is this in real life I look at 90-year-old lady and I say I am not giving you anything.  I am just not doing it.  Or I look at somebody and I can say from the head of the bed you look like somebody who is going to bleed and so I am not as aggressive.  In a clinical trial they are in and they get it and not only do they get it but they get whatever the dose the research nurse said to give them.  That is very helpful.  Give me some of your intuition.  I am not sure if this is easy to do to distill your intuition down to some words but when you are at the head of the bed…? So really thin people I think run a risk of bleeding.  Really old people or people who come in and just aren’t straightforward, they make me nervous about bleeding.  For instance, you don’t exactly know what is wrong with them.  It might be this and it might be that.  You are worried about a lot of different things.  They have a lot of comorbidities.  Maybe they have some cancer.  Things like that make them at a high-risk for bleeding.  People who are in the hospital a lot and malnourished are at a high-risk for bleeding.  I enroll people in clinical trials too, and I can tell you I am only looking at the entry criteria.  I don’t say this is a cachectic man and I am not putting him in; they get in.  Whereas at the head of the bed you said there is a cachectic man and I am going to be really careful.  I think because there is a financial incentive and so people enroll, people that you might have been less aggressive with otherwise”.



“Bleeding has become an area of intense focus at least in cardiology in the last year or two.  So as cardiologists we are very concerned about bleeding.  I think bleeding that occurs during the hospitalization for ACS or something else obviously gets our attention more because we are seeing the patients every day and those are things that usually relay to the short-term therapies.  What happens with bleeding at home is that it is a whole variety of different things.  It could be nuisance bleeding and leading someone to stop their drug or take it less frequently.  It can be a major bleed that leads to someone being admitted to the hospital with GI bleed requiring a lot of transfusions.  We see that sometimes with Coumadin.  It could be a spontaneous intracranial hemorrhage that kills someone at home.  I mean you see that whole variety.




  • The need for transfusion is a measure of a significant bleeding event because it is thought by some groups to be associated with detrimental inflammatory responses.


“Here at UK (Kentucky) we also think that bleeding and transfusion may be associated with inflammatory response that is detrimental”.




  • Even if they are not associated with worsening long term outcomes, hematomas do lengthen hospital stays, and that has significant negative pharmacoeconomic consequences.  It is very difficult to control the bleeding in morbidly obese patients who have hematomas.


“I am a hospital geek and even a hematoma lengthens stay.  My patients in the hospital are more and more often morbidly obese and their hematomas tend to be “Holy cow, I can’t control the bleeding” kind of hematomas.  So I am very, very cautious about bleeding.  I don’t like it.  I believe that even hematomas are a problem.  I think that in patients that have a lot of comorbidities, I believe you have got to reverse some of these things in the setting of hematomas”.




  • Antiplatelet agents are contraindicated once a patient bleeds on an antithrombotic.  Thus the problem with bleeding on an antithrombotic is that patients will not be treatable in the future with the antiplatelets and thus will not receive the benefit of this evidence-based medicine.  The first time a patient bleeds in the community is the last time they will see one of our new branded anticoagulants.  They are going back onto warfarin.


“But Sunil Rao at Duke has done a lot of this.  Sunil and his group think that the problem with bleeding is not so much the bleeding, but the fact that once you bleed no one gives you antiplatelets anymore.  They stop all of your medicines and then you don’t receive the benefit of the evidence-based medicines”.



“We want to reduce their pill count.  We want to reduce the amount of monitoring we want to do.  But the thing about it is 10 years ago or five years ago I thought just get a long-acting drug, and heck everybody bleeds to death with warfarin anyway, all you have got to do is beat warfarin and that can’t be too hard.  But you know what it has turned out to be relatively hard to beat warfarin because of bleeding.  Because you can’t tell when somebody does start to bleed the bleeding is more serious because you can’t reverse the drug.  I think these trials will in the long run have some trouble being better than warfarin.  We have to watch as they go through because they are going to look for noninferiority with wide confidence intervals.  They are going to do the same thing they did in the ACUITY trial; 25% noninferiority boundary.  And then we will start to see them used in the wrong people and we will start to see some bleeding.  This is what TRITON taught us or REAFFIRM taught us.  You can’t get more intensive antiplatelet and antithrombotic therapy without more bleeding.  In the outpatient setting the first time you have somebody go off and have a major head bleed on one of these drugs you are going to think twice.  Really in good Coumadin clinics we have very few really huge problems because if they have a GI bleed we stop the drug.  We reverse the drug and some FFP and kaboom there we are and we are treating it.  Head bleeds are the only thing that you really have trouble with.  That is my sense”.

Categories: antiplatelets, Antithrombotics, Cardiovascular, DVT/PE, SPAF
Tags: , , , , , , , , , , , , , , , , , , , , , ,
Bookmark the permalink.

Post a comment

Comments on are monitored.

You may use these HTML tags and attributes:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>