- Email: [email protected]
Final discussion and concluding comments
THROMBOSIS RESEARCH, Supplement TV; 185-190, 1983 0049-3848/83 $3.00 + .OO Printed in the USA. Copyright (c) 1983 Pergamon Press Ltd. All rights reserved.
FINAL DISCUSSION AND
CONCLUDING
COMMENTS
Dr. McGregor - You will remember that yesterday afternoon Dr. Packham put a very poignant question to us. "In real life, we spend much effort establishing the optimal dose of a drug. Why", she asked, "is it that in all these therapeutic trials we use a single dose of medication?* She received two answers, each very revealing. The first was that it would cost too much to prepare and administer a personalized dose. The second was to the effect that when you commit 18 million dollars and several years of your time to study a drug on a large population of patients you try for the biggest dose possible without toxic effects. But I submit that there is a third, understated reason, and it is this. Not only are we uncertain whether the drugs we are talking about will work, we usually do not even know the mechanism of the disease which we wish to influence by our therapy. So g our drugs work we are not at all sure how they work. How then, could we choose an optional dose? We are using therapy before we understand mechanism. Now there is a good precedence for this. For example, when in 1848 Dr. Snow deduced that the cause of an epidemic of cholera in Lamberth was due to the drinking water, he did not try to establish the mechanism of the disease, he simply advocated removing the handle from the pump which supplied the water (1). The subsidence of the epidemic supplied a substantial clue as to mechanism. And this is what we too have been doing. We have been using therapeutic intervention involving a large population to explore mechanism. Thus, the original hypothesis of the anturane trial was, if this therapy reduces the incidence of reinfarction or death, then this-supports the concept that platelet adhesion is an essential step inxnfarction or sudden death (2). There clearly can be a time when it is justifiable to test a remedy before understanding the mechanisms of the disease we are curing. But as applied to large populations it is a crude and expensive research tool. As a non-pharmacological visitor to this conference, it has occurred to me that we have discussed drugs and their actions in great detail. And I have learned a great deal about them. But I have heard very little about the exact logic behind their use. It is clear we wish to prevent or retard atherosclerosis. But are we trying to treat the platelet? Or the vascular endothelium? Or interfere elsewhere in the clotting mechanism? Or inhibit prostaglandin related coronary spasm?
The Commonwealth Fund, New York, N.Y., 1936.
1.
Frost WH.
2.
Sulfinpyrazone in the prevention of sudden death infarction. N. Engl. J. Med. 302:250-256, 1980.
Snow on Cholera.
185
after
myocardial
186
FINAL DISCUSSION
SUPPI.
IV,
1983
There has been discussion today of what therapeutic trial we should try next, what model we should use, and what drug combination. If this were my field of research, I would want to understand mechanisms a little better before trying out more therapies on a large scale. And if I were a pharmaceutical company, this is where I would put my money. I would go very slowly on further therapeutic trials until our understanding of mechanisms had made substantial advances. A fraction of the vast cost of these trials would go a long way if it were spent on the basic research necessary to understand these disease processes. The model that Dr. Hirsh proposes for study, namely the ability of a drug to stop clotting on an artificial heart valve or on an arteriovenous fistula, is a good one. But I do not need to remind you that if successful, you have a drug which is proven good for only one purpose, namely to prevent clotting on these prostheses. It is not going to tell you what drugs will prevent the development of atherosclerosis in a healthy vessel, or the development of thrombosis in a diseased one. Our understanding of mechanism has not gone far enough to allow the extrapolations. Just one more point that I think we should be aware of as we consider the next post-infarction secondary prevention trial; with good post-infarction care, the mortality is soon going to be so low that we will require a vast number of cases to prove the benefit of any:intervention. In the era in which patients had the era which we are only just leaving, returned to normal life without any special infarcts and then slowly patients with life precautions, one could anticipate, if one excludes limiting non-ischemic cardiovascular disease and patients over the age of 65, a mortality of around 8-122 in the first year after discharge from hospital. Today, by a Today, however, we can and are doing much better than this. those patients series of measures which I will not recount here, we identify we believe to be at risk and subject a very considerable proportion of them By these to bypass surgery or to special antiarrhythymic drug medication. at the Royal Victoria means in our own Post-Infarction Follow-Up Clinic Hospital, Dr. Allan Sniderman and his colleagues have reduced post-infarction mortality in a comparable population to slightly under 3% in the first year hospital is an Clearly, after hospital discharge. from now on, this And I suspect that, within inappropriate place for a post-infarction trial. the next 5 years, these figures will be the norm. Dr. Hirsh - The arterial-venous shunt model is useful to study the My second relationship between drug levels and their antithrombotic effects. comment relates to Dr. Taylor’s fourth temptation, that is, the temptation to prejudge and issue which was done with the seven day rule in the infarction study. 1 do not know of any good sulfinpyrazone myocardial evidence that it does take seven days for sulfinpyrazone to have an effect. Mr. Taylor - These rules have not made any effect on the analysis of the results. In fact we involved in the Canadian Cooperative Stroke Study that 7 days back into the if you put the events that happened in the first analysis the results are even more strongly significant in favour of aspirin, and I understand it doesn’t make any difference to the ART Trial either. issue.
Dr. Hirsh - Yes,
I agree,
but one takes
big
risks
when one prejudges
the
Suppl.
IV, 1983
FINAL DISCUSSION
187
Dr. McGregor - I must agree with Mr. Taylor. It does not matter in the “seven day rule” was introduced because of a whether the slightest If it is misunderstanding at the time of the action of Anturan or not. decided beforehand that the drug will only be considered to have taken effect seven days after its administration this can in no way destroy the validity One makes the same assumption for drug and placebo and, as of the results. The Mr. Taylor has pointed out, it has no measureable affect on the outcome. This is surely not worst thing that we lose is seven days of active therapy. one of “the seven deadly sins” . I do not understand the fuss that is being made about it. Dr. Hirsh - I have been asked to summarize the lessons learned from the the clinical trials and the current state of knowledge of the mode of action of the drugs which have been tested. I would like -to consider these under The only antiplatelet agent that has been timing and combinations. dose, evaluated in terms of dosage in clinical trials is aspirin. Aspirin ha s shown to be effective in preventing venous thrombosis after major knee surgery in a dose of 3.5 grams per day but it was ineffective when used in a dose of 900 mg per day. It is possible that its effectiveness in very high doses is due to an independent effect of salicylate. Aspirin has also been shown to be effective in preventing venous thrombosis in patients with arterial-venous shunts when used in a dose of 165 mg per day. Although one cannot extrapolate these results to patients with ischemic heart disease c’r transient cerebral ischemia, it does at least tell us that when aspirin is given in a dose which inhibits the synthesis of thromboxane AZ, it is antithrombotic in patients with arterial-venous shunts. To consider timing, aspirin was effective when given once per day to patients with arterial-shunts and it was effective when it was given four times a day in patients undergoing major knee surgery. There is considerable evidence that aspirin augments the effect of dipyridamole in normalizing platelet survival in patients who have increased platelet turnover. As yet there have been no studies in man which have demonstrated that the combination of aspirin and dipyridamole is better than either aspirin or dipyridamole alone. Although there are theoretical reasons why aspirin and sulfinpyrazone might be incompatible as antithrombotic agents , there is no evidence that the inhibitory effect of either aspirin or sulfinpyrazone is impaired by using a combination of both drugs. If anything, there is some evidence that aspirin and sulfinpyrazone are more effective than aspirin alone in patients with transient cerebral ischemia. Dr. Fuster - I would like to ask five minutes to defend a point. When testing platelet inhibitors in a clinical setup it is essential to understand the mechanisms and the natural history of the disease process that we are trying to treat. Three mistakes have been made in the last few years when we measure clinical trials. First, the mechanism of the disease was not looked at critically before the trials were started. Seven trials have been published in the literature stating that platelet inhibitors do not protect occlusion of saphenous vein coronary bypass vein grafting. We looked at this issue in an animal model in 1976-78 and found that thrombotic occlusion in the saphenous vein grafts occurs within the first few hours of the operation. Thus, if you plan to commence a drug regime to prevent saphenous vein bypass occlusion, you must start pr e-opera tively, not 4 or 5 days after surgery, as was the case in most of these trials. The recent trial that we finished in collaboration with Cheseboro at the Mayo Clinic which was successful, was based upon what we learned in the animal studies, and I would say that in
188
saphenous
FINAL DISCUSSION
vein bypass
surgery,
we already
supp 1. IV, 1983
have the drugs are effective.
The second mistake is related to coronary artery disease. As you know, the trials or most of the trials have been equivocal in terms of the results in which the endpoints measured have been reinfarction and sudden death. The problem here is lack of sensitivity. These were patients all with myocardial infarction who were followed for one to three years and the endpoints, although important clinically, were insensitive because as the comment made before, only 6%-8X of the patients developed a reinfarction in the two years of follow-up. You need many, many patients to determine a successful drug intervention when you are dealing with a very insensitive endpoint. To me, the approach to coronary disease from the thrombotic standpoint, is trying to demonstrate that platelet inhibitors prevent the progression of the coronary thrombotic process by using an objective measurement such as long-term follow-up with angiograpby. We started such a study four years ago, and it will be finished in three years. have an angiogram and Br ie fly , patients five years later they have a second angiogram. All the patients have already entered in the study and we now are starting to do the second angiogram. I think they will be able to answer a crucial question, **Do we effect coronary anatomy with platelet inhibitors?“. We will also look, of course, at the clinical endpoints, sudden death and myocardial infarction. Now in terms of coronary artery disease, I feel that the disease progresses in two ways. One is by mural thrombi which become organized and develop into atherosclerotic plaques which grow rapidly. The second mechanism is through the development of a platelet layer upon a de-endothelialized and the disease surface progresses slowly through the mechanism described by Drs. Ross and Harker, i.e. platelet derived growth factor. It is unlikely that the second mechanism will be prevented by dipyridamole, ASA or sulfinpryazone, since in the pig model, these drugs do not prevent platelet deposition onto denuded endothelium. What these drugs prevent is platelet aggregation which may be important in thrombosis. And I think in this regard perhaps the new drugs about which we have heard described in the last day or so might be meaningful, such as ticlopidine or nafazatrom. Perhaps these drugs might prevent the formation of this platelet layer and retard the growth of the But I am concerned that these drugs may be too atherosclerotic plug. We will prevent powerful and promote bleeding in the patients. atherosclerotic disease but we will cause other problems, one of them being bleeding because we inhibit platelet function completely thus also preventing platelet adhesion. Perhaps these drugs or future drugs might play a role in the acute stages. For example, one of the problems which we are facing today is the patients undergoing angioplasty of the coronary as cardiologists, There is a re-stenosis rate of 30%. We have some evidence in the arteries. is pig model, that one layer of platelets at the time of angioplasty sufficient to promote intimal thickening which may lead to re-stenosis. go we probably need a very powerful drug at the time of angioplasty to block ASA or the platelet Perhaps dipyridamole, completely involvement. sulfinpyrazone may be useful in preventing the thrombotic process that comes afterwards, but not the platelet adhesion that occurs during the procedure. The third problem is related to is the use of platelet inhibitors alone The rate of the incidence of in patients with prosthetic heart valves. thromboembolism is very high if you give only platelet inhibitors to a mechanical valve, particularly in the mitral position. We have recently learned from a trial which has been completed at the Mayo Clinic, that with mechanical prosthetic valves, the optimal antithrombotic regime is the use of
suppl.
xv,
1983
FINAL DISCUSSION
189
and that the platelet inhibitor should coumadin with a platelet inhibitor, a significant bleeding problem occurs. not be aspirin. If it is aspirin, inhibitors that we should not use platelet emphasize The se points We should probably develop new platelet inhibitors for a indiscriminantly. specific problem which other platelet inhibitors are not able to solve, and research in the future should go in this direction. a Dr. Harker - This meeting has achieved importance by facilitating significant dialogue among those interested in antithrombotic therapy, industry, basic scientists and clinicians. The difficulties of developing new clinically useful antithrombotic therapy is much more complicated at present than it appeared several years ago because of the negative results reported for a number of important, well-designed controlled trials involving Thus, we come to this patients with presumed arterial thrombotic disease. The complexities meeting more seasoned and willing to listen and share. associated with anithrombotic agents in the management of thrombotic vascular occlusive disease have grown progressively as those groups interested in antithrombotic therapy have pursued their own unique approaches and goals. Since none of the individual objectives have been fully achieved by such endeavors, it is time for a “mid-course correctionVt. The factors in the clinical equation describing thrombogenesis and its prevention are many and their relative importance variable. Three major problems relate to the available clinical models: 1) There is a low frequency of events in all clinical disease states associated with thrombosis. Consequently , large and expensive trials are necessary. 2) There is continued uncertainty regarding the precise processes of thrombosis involved in the pathogenesis of the various vascular occlusive clinical models. 3) The currently available antithrombotic agents that may be selected for clinical trials are of modest potency and unclear mechanism of action, or have potent specific pharmacologic effects without significant anti thrombotic potency. Consequently, the selection of drug dosage, regimen and clinical model have not been based on solid, objective evidence. It is not surprising, therefore, that negative trials have appeared without clarifying any of the issues related to pathogenesis, drug or regimen. The resultant confusion is substantial. Solutions to these problems require the co-operative efforts of industry, basic science and clinical medicine. Despite the apparent setback of several recent negative trials with respect to arterial thrombosis, there has been significant progress in a number of areas. the design, completion and analysis of controlled Firstly, clinical trials have improved greatly in the past several years. The group at McMaster has uniquely contributed to trial design in this field. Secondly, industry has been active in preparing pharmacologic agents with specific mechanisms of action. These drugs are potentially very helpful in defining biological effects. a specific thromboxane For example, dazoxiben, synthetase inhibitor has shown weak antithrombotic effects despite its potency as a specific thromboxane synthetase inhibitor. These results are surprising in view of the postulated importance of thromboxane A2 in the activation of platelets and thrombogenesis. Thirdly, rapid developments in chemical synthesis, Cell and molecular biology have accelerated greatly the rate at which pharmacologic agents can be synthesized and characterized. Fourthly , application of objective intermediate endpoints provide a useful Strategy to answer efficiently in patients questions regarding drug efficacy and dose regimen. For example, quantitative angiography of peripheral and coronary arteries or digital subtraction angiography of carotid arteries
190
supp 1. IV, 1983
FINAL DISCUSSION
allow for objective documentation arterial disease progression.
of
drug
effects
on
native
or
induced
_ Some important areas to be addressed in this field include: 1) clarification of the pathogenesis of various clinical thrombotic disorders; 2) the development of appropriate thrombotic animal models to stimulate more closely human disease; and 3) expand the limited resources available to carry out control clinical trials. Close co-operation among basic scientists, clinicians, industry and governmental health agencies will greatly enhance our capacity to improve the care of patients with vascular thrombo-occlusive disease. Dr. Barnett - The first point that I would like to make is that in doing clinical trials to determine the efficacy of a drug or a strategy in altering disease, one can have a number of endpoints. On the other hand, the bottom line of any clinical trial is the evaluation of the effect of the treatment strategy on the disease process. The assessment of a physiological reaction or a pharmacological endpoint may be achieved without altering the clinical course of the condition whatsoever. Thus, a trial of drug efficacy that does not have clinical endpoints as its ultimate goal is not truly a clinical trial. It is quite reasonably to hypothesize about dosage on the basis of drug It cannot be effect on some physiological or biological parameters. concluded from this that the disease process will ever be responsive or Nowhere in present clinical situations is this more apt than unresponsive. A recent letter to the editor of Stroke states that “it is in drug dosage. concluded that the aspirin dosage used in the large control studies are far in excess of those that are needed”. The reason for this was because a number of platelet studies were done and to alter these particular platelet functions one needed a very small dose of aspirin. The huge giant step is you need only to use this then taken by concluding that to prevent stroke, Without a clinical trial which compares this small dose of drug. hypothetical dosage with the dosage that was used in the other trials, such concentrations are verging on the absurd. At a recent International Symposium on Cerebral Bypass Surgery, a number that cerebral blood flow could be of papers were given which indicated temporal-to-middle cerebral artery improved by performing a superficial The conclusion reached by a number of these observers was that one bypass. could therefore conclude that strokes would be prevented by doing this operation in stroke-threatened patients afflicated with narrowing of arteries The laboratory inside the head or occlusion of arteries in the neck. The clinical trial was not observation was regarded as the endpoing. This is an extrapolation from a non-clinical considered as essential. endpoint that is unjustified. An erroneous assumption is made in many clinical situations that we know In very few circumstances at the bedside is the natural history of disease. We know what the probability is for 100 or 1000 patients and this accurate. very often it is only when we conduct the clinical trial that we realize how Sometimes all that many variations there are in the expressions of disease. comes out of a clinical trial is a negative benefit from a therapeutic strategy .
FINAL DISCUSSION AND
CONCLUDING
COMMENTS
Dr. McGregor - You will remember that yesterday afternoon Dr. Packham put a very poignant question to us. "In real life, we spend much effort establishing the optimal dose of a drug. Why", she asked, "is it that in all these therapeutic trials we use a single dose of medication?* She received two answers, each very revealing. The first was that it would cost too much to prepare and administer a personalized dose. The second was to the effect that when you commit 18 million dollars and several years of your time to study a drug on a large population of patients you try for the biggest dose possible without toxic effects. But I submit that there is a third, understated reason, and it is this. Not only are we uncertain whether the drugs we are talking about will work, we usually do not even know the mechanism of the disease which we wish to influence by our therapy. So g our drugs work we are not at all sure how they work. How then, could we choose an optional dose? We are using therapy before we understand mechanism. Now there is a good precedence for this. For example, when in 1848 Dr. Snow deduced that the cause of an epidemic of cholera in Lamberth was due to the drinking water, he did not try to establish the mechanism of the disease, he simply advocated removing the handle from the pump which supplied the water (1). The subsidence of the epidemic supplied a substantial clue as to mechanism. And this is what we too have been doing. We have been using therapeutic intervention involving a large population to explore mechanism. Thus, the original hypothesis of the anturane trial was, if this therapy reduces the incidence of reinfarction or death, then this-supports the concept that platelet adhesion is an essential step inxnfarction or sudden death (2). There clearly can be a time when it is justifiable to test a remedy before understanding the mechanisms of the disease we are curing. But as applied to large populations it is a crude and expensive research tool. As a non-pharmacological visitor to this conference, it has occurred to me that we have discussed drugs and their actions in great detail. And I have learned a great deal about them. But I have heard very little about the exact logic behind their use. It is clear we wish to prevent or retard atherosclerosis. But are we trying to treat the platelet? Or the vascular endothelium? Or interfere elsewhere in the clotting mechanism? Or inhibit prostaglandin related coronary spasm?
The Commonwealth Fund, New York, N.Y., 1936.
1.
Frost WH.
2.
Sulfinpyrazone in the prevention of sudden death infarction. N. Engl. J. Med. 302:250-256, 1980.
Snow on Cholera.
185
after
myocardial
186
FINAL DISCUSSION
SUPPI.
IV,
1983
There has been discussion today of what therapeutic trial we should try next, what model we should use, and what drug combination. If this were my field of research, I would want to understand mechanisms a little better before trying out more therapies on a large scale. And if I were a pharmaceutical company, this is where I would put my money. I would go very slowly on further therapeutic trials until our understanding of mechanisms had made substantial advances. A fraction of the vast cost of these trials would go a long way if it were spent on the basic research necessary to understand these disease processes. The model that Dr. Hirsh proposes for study, namely the ability of a drug to stop clotting on an artificial heart valve or on an arteriovenous fistula, is a good one. But I do not need to remind you that if successful, you have a drug which is proven good for only one purpose, namely to prevent clotting on these prostheses. It is not going to tell you what drugs will prevent the development of atherosclerosis in a healthy vessel, or the development of thrombosis in a diseased one. Our understanding of mechanism has not gone far enough to allow the extrapolations. Just one more point that I think we should be aware of as we consider the next post-infarction secondary prevention trial; with good post-infarction care, the mortality is soon going to be so low that we will require a vast number of cases to prove the benefit of any:intervention. In the era in which patients had the era which we are only just leaving, returned to normal life without any special infarcts and then slowly patients with life precautions, one could anticipate, if one excludes limiting non-ischemic cardiovascular disease and patients over the age of 65, a mortality of around 8-122 in the first year after discharge from hospital. Today, by a Today, however, we can and are doing much better than this. those patients series of measures which I will not recount here, we identify we believe to be at risk and subject a very considerable proportion of them By these to bypass surgery or to special antiarrhythymic drug medication. at the Royal Victoria means in our own Post-Infarction Follow-Up Clinic Hospital, Dr. Allan Sniderman and his colleagues have reduced post-infarction mortality in a comparable population to slightly under 3% in the first year hospital is an Clearly, after hospital discharge. from now on, this And I suspect that, within inappropriate place for a post-infarction trial. the next 5 years, these figures will be the norm. Dr. Hirsh - The arterial-venous shunt model is useful to study the My second relationship between drug levels and their antithrombotic effects. comment relates to Dr. Taylor’s fourth temptation, that is, the temptation to prejudge and issue which was done with the seven day rule in the infarction study. 1 do not know of any good sulfinpyrazone myocardial evidence that it does take seven days for sulfinpyrazone to have an effect. Mr. Taylor - These rules have not made any effect on the analysis of the results. In fact we involved in the Canadian Cooperative Stroke Study that 7 days back into the if you put the events that happened in the first analysis the results are even more strongly significant in favour of aspirin, and I understand it doesn’t make any difference to the ART Trial either. issue.
Dr. Hirsh - Yes,
I agree,
but one takes
big
risks
when one prejudges
the
Suppl.
IV, 1983
FINAL DISCUSSION
187
Dr. McGregor - I must agree with Mr. Taylor. It does not matter in the “seven day rule” was introduced because of a whether the slightest If it is misunderstanding at the time of the action of Anturan or not. decided beforehand that the drug will only be considered to have taken effect seven days after its administration this can in no way destroy the validity One makes the same assumption for drug and placebo and, as of the results. The Mr. Taylor has pointed out, it has no measureable affect on the outcome. This is surely not worst thing that we lose is seven days of active therapy. one of “the seven deadly sins” . I do not understand the fuss that is being made about it. Dr. Hirsh - I have been asked to summarize the lessons learned from the the clinical trials and the current state of knowledge of the mode of action of the drugs which have been tested. I would like -to consider these under The only antiplatelet agent that has been timing and combinations. dose, evaluated in terms of dosage in clinical trials is aspirin. Aspirin ha s shown to be effective in preventing venous thrombosis after major knee surgery in a dose of 3.5 grams per day but it was ineffective when used in a dose of 900 mg per day. It is possible that its effectiveness in very high doses is due to an independent effect of salicylate. Aspirin has also been shown to be effective in preventing venous thrombosis in patients with arterial-venous shunts when used in a dose of 165 mg per day. Although one cannot extrapolate these results to patients with ischemic heart disease c’r transient cerebral ischemia, it does at least tell us that when aspirin is given in a dose which inhibits the synthesis of thromboxane AZ, it is antithrombotic in patients with arterial-venous shunts. To consider timing, aspirin was effective when given once per day to patients with arterial-shunts and it was effective when it was given four times a day in patients undergoing major knee surgery. There is considerable evidence that aspirin augments the effect of dipyridamole in normalizing platelet survival in patients who have increased platelet turnover. As yet there have been no studies in man which have demonstrated that the combination of aspirin and dipyridamole is better than either aspirin or dipyridamole alone. Although there are theoretical reasons why aspirin and sulfinpyrazone might be incompatible as antithrombotic agents , there is no evidence that the inhibitory effect of either aspirin or sulfinpyrazone is impaired by using a combination of both drugs. If anything, there is some evidence that aspirin and sulfinpyrazone are more effective than aspirin alone in patients with transient cerebral ischemia. Dr. Fuster - I would like to ask five minutes to defend a point. When testing platelet inhibitors in a clinical setup it is essential to understand the mechanisms and the natural history of the disease process that we are trying to treat. Three mistakes have been made in the last few years when we measure clinical trials. First, the mechanism of the disease was not looked at critically before the trials were started. Seven trials have been published in the literature stating that platelet inhibitors do not protect occlusion of saphenous vein coronary bypass vein grafting. We looked at this issue in an animal model in 1976-78 and found that thrombotic occlusion in the saphenous vein grafts occurs within the first few hours of the operation. Thus, if you plan to commence a drug regime to prevent saphenous vein bypass occlusion, you must start pr e-opera tively, not 4 or 5 days after surgery, as was the case in most of these trials. The recent trial that we finished in collaboration with Cheseboro at the Mayo Clinic which was successful, was based upon what we learned in the animal studies, and I would say that in
188
saphenous
FINAL DISCUSSION
vein bypass
surgery,
we already
supp 1. IV, 1983
have the drugs are effective.
The second mistake is related to coronary artery disease. As you know, the trials or most of the trials have been equivocal in terms of the results in which the endpoints measured have been reinfarction and sudden death. The problem here is lack of sensitivity. These were patients all with myocardial infarction who were followed for one to three years and the endpoints, although important clinically, were insensitive because as the comment made before, only 6%-8X of the patients developed a reinfarction in the two years of follow-up. You need many, many patients to determine a successful drug intervention when you are dealing with a very insensitive endpoint. To me, the approach to coronary disease from the thrombotic standpoint, is trying to demonstrate that platelet inhibitors prevent the progression of the coronary thrombotic process by using an objective measurement such as long-term follow-up with angiograpby. We started such a study four years ago, and it will be finished in three years. have an angiogram and Br ie fly , patients five years later they have a second angiogram. All the patients have already entered in the study and we now are starting to do the second angiogram. I think they will be able to answer a crucial question, **Do we effect coronary anatomy with platelet inhibitors?“. We will also look, of course, at the clinical endpoints, sudden death and myocardial infarction. Now in terms of coronary artery disease, I feel that the disease progresses in two ways. One is by mural thrombi which become organized and develop into atherosclerotic plaques which grow rapidly. The second mechanism is through the development of a platelet layer upon a de-endothelialized and the disease surface progresses slowly through the mechanism described by Drs. Ross and Harker, i.e. platelet derived growth factor. It is unlikely that the second mechanism will be prevented by dipyridamole, ASA or sulfinpryazone, since in the pig model, these drugs do not prevent platelet deposition onto denuded endothelium. What these drugs prevent is platelet aggregation which may be important in thrombosis. And I think in this regard perhaps the new drugs about which we have heard described in the last day or so might be meaningful, such as ticlopidine or nafazatrom. Perhaps these drugs might prevent the formation of this platelet layer and retard the growth of the But I am concerned that these drugs may be too atherosclerotic plug. We will prevent powerful and promote bleeding in the patients. atherosclerotic disease but we will cause other problems, one of them being bleeding because we inhibit platelet function completely thus also preventing platelet adhesion. Perhaps these drugs or future drugs might play a role in the acute stages. For example, one of the problems which we are facing today is the patients undergoing angioplasty of the coronary as cardiologists, There is a re-stenosis rate of 30%. We have some evidence in the arteries. is pig model, that one layer of platelets at the time of angioplasty sufficient to promote intimal thickening which may lead to re-stenosis. go we probably need a very powerful drug at the time of angioplasty to block ASA or the platelet Perhaps dipyridamole, completely involvement. sulfinpyrazone may be useful in preventing the thrombotic process that comes afterwards, but not the platelet adhesion that occurs during the procedure. The third problem is related to is the use of platelet inhibitors alone The rate of the incidence of in patients with prosthetic heart valves. thromboembolism is very high if you give only platelet inhibitors to a mechanical valve, particularly in the mitral position. We have recently learned from a trial which has been completed at the Mayo Clinic, that with mechanical prosthetic valves, the optimal antithrombotic regime is the use of
suppl.
xv,
1983
FINAL DISCUSSION
189
and that the platelet inhibitor should coumadin with a platelet inhibitor, a significant bleeding problem occurs. not be aspirin. If it is aspirin, inhibitors that we should not use platelet emphasize The se points We should probably develop new platelet inhibitors for a indiscriminantly. specific problem which other platelet inhibitors are not able to solve, and research in the future should go in this direction. a Dr. Harker - This meeting has achieved importance by facilitating significant dialogue among those interested in antithrombotic therapy, industry, basic scientists and clinicians. The difficulties of developing new clinically useful antithrombotic therapy is much more complicated at present than it appeared several years ago because of the negative results reported for a number of important, well-designed controlled trials involving Thus, we come to this patients with presumed arterial thrombotic disease. The complexities meeting more seasoned and willing to listen and share. associated with anithrombotic agents in the management of thrombotic vascular occlusive disease have grown progressively as those groups interested in antithrombotic therapy have pursued their own unique approaches and goals. Since none of the individual objectives have been fully achieved by such endeavors, it is time for a “mid-course correctionVt. The factors in the clinical equation describing thrombogenesis and its prevention are many and their relative importance variable. Three major problems relate to the available clinical models: 1) There is a low frequency of events in all clinical disease states associated with thrombosis. Consequently , large and expensive trials are necessary. 2) There is continued uncertainty regarding the precise processes of thrombosis involved in the pathogenesis of the various vascular occlusive clinical models. 3) The currently available antithrombotic agents that may be selected for clinical trials are of modest potency and unclear mechanism of action, or have potent specific pharmacologic effects without significant anti thrombotic potency. Consequently, the selection of drug dosage, regimen and clinical model have not been based on solid, objective evidence. It is not surprising, therefore, that negative trials have appeared without clarifying any of the issues related to pathogenesis, drug or regimen. The resultant confusion is substantial. Solutions to these problems require the co-operative efforts of industry, basic science and clinical medicine. Despite the apparent setback of several recent negative trials with respect to arterial thrombosis, there has been significant progress in a number of areas. the design, completion and analysis of controlled Firstly, clinical trials have improved greatly in the past several years. The group at McMaster has uniquely contributed to trial design in this field. Secondly, industry has been active in preparing pharmacologic agents with specific mechanisms of action. These drugs are potentially very helpful in defining biological effects. a specific thromboxane For example, dazoxiben, synthetase inhibitor has shown weak antithrombotic effects despite its potency as a specific thromboxane synthetase inhibitor. These results are surprising in view of the postulated importance of thromboxane A2 in the activation of platelets and thrombogenesis. Thirdly, rapid developments in chemical synthesis, Cell and molecular biology have accelerated greatly the rate at which pharmacologic agents can be synthesized and characterized. Fourthly , application of objective intermediate endpoints provide a useful Strategy to answer efficiently in patients questions regarding drug efficacy and dose regimen. For example, quantitative angiography of peripheral and coronary arteries or digital subtraction angiography of carotid arteries
190
supp 1. IV, 1983
FINAL DISCUSSION
allow for objective documentation arterial disease progression.
of
drug
effects
on
native
or
induced
_ Some important areas to be addressed in this field include: 1) clarification of the pathogenesis of various clinical thrombotic disorders; 2) the development of appropriate thrombotic animal models to stimulate more closely human disease; and 3) expand the limited resources available to carry out control clinical trials. Close co-operation among basic scientists, clinicians, industry and governmental health agencies will greatly enhance our capacity to improve the care of patients with vascular thrombo-occlusive disease. Dr. Barnett - The first point that I would like to make is that in doing clinical trials to determine the efficacy of a drug or a strategy in altering disease, one can have a number of endpoints. On the other hand, the bottom line of any clinical trial is the evaluation of the effect of the treatment strategy on the disease process. The assessment of a physiological reaction or a pharmacological endpoint may be achieved without altering the clinical course of the condition whatsoever. Thus, a trial of drug efficacy that does not have clinical endpoints as its ultimate goal is not truly a clinical trial. It is quite reasonably to hypothesize about dosage on the basis of drug It cannot be effect on some physiological or biological parameters. concluded from this that the disease process will ever be responsive or Nowhere in present clinical situations is this more apt than unresponsive. A recent letter to the editor of Stroke states that “it is in drug dosage. concluded that the aspirin dosage used in the large control studies are far in excess of those that are needed”. The reason for this was because a number of platelet studies were done and to alter these particular platelet functions one needed a very small dose of aspirin. The huge giant step is you need only to use this then taken by concluding that to prevent stroke, Without a clinical trial which compares this small dose of drug. hypothetical dosage with the dosage that was used in the other trials, such concentrations are verging on the absurd. At a recent International Symposium on Cerebral Bypass Surgery, a number that cerebral blood flow could be of papers were given which indicated temporal-to-middle cerebral artery improved by performing a superficial The conclusion reached by a number of these observers was that one bypass. could therefore conclude that strokes would be prevented by doing this operation in stroke-threatened patients afflicated with narrowing of arteries The laboratory inside the head or occlusion of arteries in the neck. The clinical trial was not observation was regarded as the endpoing. This is an extrapolation from a non-clinical considered as essential. endpoint that is unjustified. An erroneous assumption is made in many clinical situations that we know In very few circumstances at the bedside is the natural history of disease. We know what the probability is for 100 or 1000 patients and this accurate. very often it is only when we conduct the clinical trial that we realize how Sometimes all that many variations there are in the expressions of disease. comes out of a clinical trial is a negative benefit from a therapeutic strategy .