HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:
Guest Access | Sign In via User Name/Password

This Article Abstract Full Text (PDF) Free Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Article Archive Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (28) Citing Articles via Google Scholar Google Scholar Articles by Verstraete, M. Articles by Verhaeghe, R. Search for Related Content PubMed PubMed Citation Articles by Verstraete, M. Articles by Verhaeghe, R.

A European View on the North American Fifth Consensus on Antithrombotic Therapy^*

Marc Verstraete, MD; Colin R. M. Prentice, MD; Michel Samama, MD; Raymond Verhaeghe, MD and on behalf of a European Working Group

^* From the Center for Molecular and Vascular Biology (Drs. Verstraete and Verhaeghe), Katholieke Universiteit Leuven, Belgium; Division of Medicine (Dr. Prentice), University of Leeds, United Kingdom; and the Department of Hématologie-Biologie (Dr. Samana), Hôtel-Dieu de Paris, France. A complete list of participants is located in the ^Appendix.

Correspondence to: Marc Verstraete, MD, Center for Molecular and Vascular Biology, Katholieke Universiteit Leuven, Campus Gasthuisberg, O&N, Herestraat 49, 3000 Leuven, Belgium

	Abstract

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
An American-Canadian group of experts have, in the November 1998 issue of CHEST, published for the fifth time their recommendations for antithrombotic therapy. This remarkable consensus document was the result of an extensive review of the literature by an interdisciplinary group. Considering the impact of this document on medical practice, also outside North America, a group of European experts reviewed in detail the fifth report, particularly the sections on clinical indications of antithrombotic treatment. The aim was not to indicate the many areas of agreement and to quote literature that has become available since publication of the last consensus documents, but rather to refer to the gray zones of uncertainty and limited number of divergent opinions.

Key Words: antithrombotic agents • coumarin drugs • heparin • warfarin

	Introduction

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
The American College of Chest Physicians (ACCP) published in a supplement to the November 1998 issue of CHEST the conclusions of the Fifth Consensus Conference on Antithrombotic Therapy.¹ This is a comprehensive document based on an exhaustive review of the literature as of April 1998, written by an interdisciplinary but essentially American-Canadian Group of about 90 experts.In this latest evolving consensus, the grading system, already used in the 1995 document, was modified to emphasize more strongly the concept of clinically important difference, and to build the benefit-to-risk ratio of a particular treatment into the grading system.

The North American initiative as well as the guidelines proposed for American practice were well received and can, in general, be recommended to European clinicians. At a discussion of this latest ACCP report, in Paris on January 21, 1999, French clinicians expressed some divergent views from the North American consensus on a selected number of indications for antithrombotic therapy. In the present report, the American-Canadian recommendations are critically discussed by a group of European experts in an effort to reach gradually a transatlantic if not a global consensus. The European participants in this endeavor are listed in the to this article.

The grading of the level of evidence of clinical trials is a milestone of progress, in that objective comparisons can now be made. One could make an argument that an A₁ recommendation can be based on a single trial comprising a large, randomized, and placebo-controlled study. Alternatively, the case could be made that A₁ grading should only be given when a second trial, virtually identical in design and treatment comparison, confirms the results of the first. This question needs to be resolved. As one could anticipate, the lower the grade, the greater was the divergence of opinion among our experts.

The European group has taken the opportunity to update the text in topics for which it believed that important publications have appeared since April 1998, the closing date of the North American Consensus. In general, only references that confirm or modify a recommendation have been included. It must be appreciated that the range of drugs and other treatments used in some situations is different from those used in Europe, owing to different customs and drug regulatory requirements.

	Use of Antithrombotic Agents During Pregnancy

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
Although there is an increasing willingness among clinicians to use anticoagulants during pregnancy, there is a paucity of evidence about the efficacy and safety of the ever-expanding number of antithrombotic agents for the prevention and treatment of maternal and fetal thrombotic disorders. The lack of controlled trials leads to low levels of evidence of the recommendations (mostly C₂) in the chapter on the use of antithrombotic agents during pregnancy of the North American document.² The European experts face the same limitation; therefore, their comments have to be taken as updated suggestions.

Main differences between the North American and European experts concern prophylaxis in pregnant women with thrombophilia or mechanical heart valves. Low-molecular-weight (LMW) heparins are used more often to treat pregnant women with thrombophilia in Europe than in North America. Oral anticoagulants are still used by European cardiologists to treat pregnant women with mechanical heart valves. LMW heparins are sometimes considered to be more expensive than unfractionated heparin (UH), but their advantages have to be taken into account, and cost varies from country to country.

Epidemiology
The exact incidence of venous thromboembolism during pregnancy is not well documented, but compared with nonpregnant fertile women < 35 years of age, it is at least 10 times higher.³ In addition to the risk factors in pregnant women mentioned in the North American report, prothrombin G20210A polymorphism and probably hyperhomocysteinemia should be added to the list of thrombophilias.⁴

Prevention of Venous Thromboembolism in Pregnant Patients at Risk
During Pregnancy: The statement that prophylaxis with 5,000 IU of subcutaneous UH bid is effective is not fully shared by European experts. A dose of LMW heparin with a peak anticoagulant effect corresponding to 0.2 to 0.45 IU of anti-factor Xa per milliliter of plasma 3 h after subcutaneous injection seems to be effective and is also more practical for the patient (C₁).⁵ LMW heparins are more often used in Europe in pregnant women because of simplicity, comfort, good cutaneous tolerance and supposed decreased risk of heparin-induced thrombocytopenia and osteoporosis. Their safety during pregnancy has been recently reviewed.⁶ The limited available evidence suggests that for the prevention of recurrent venous thromboembolism in women with a previous history of venous thrombosis, LMW heparin 50 IU of anti-factor Xa per kilogram (early pregnancy weight) subcutaneously once daily is generally adequate and safe. This dose is usually associated with a plasma peak of around 0.25 to 0.35 IU of anti-factor Xa per milliliter of plasma.⁵ However, there is concern that these doses of UH and LMW heparin may not be sufficient in very high-risk situations, for example, pregnant women with a previous history of venous thromboembolism associated with antiphospholipid antibodies or with severe thrombophilic defects, such as antithrombin III deficiency. The doses of LMW heparin required during the third trimester are those in nonpregnant women of similar weight.^{7 8} A measurement of anti-factor Xa activity as well as weight-adjusted doses have been recommended (C₂) in Europe in order to avoid insufficient anticoagulation.

In pregnant women with a history of previous venous thromboembolism secondary to a transient risk factor, clinical surveillance is proposed by the North American consensus group. By contrast, European clinicians prefer a pharmacological prophylaxis from the beginning of pregnancy or at least during the last trimester (C₂). In these patients, screening for thrombophilia is compulsory. In the North American consensus document, the possibility of starting prophylaxis at a later stage in pregnancy is not considered.

For pregnant women with previous history of idiopathic thromboembolism or with thrombophilia but without previous personal thromboembolism, the North American consensus document lists three options: clinical surveillance, 5,000 IU UH bid, or adjusted-dose heparin to produce plasma levels of 0.1 to 0.2 IU of anti-factor Xa per milliliter of plasma throughout pregnancy.² In the North American consensus, no difference is made among the genetic defects. The European clinicians consider that antithrombin III deficiency carries a higher thrombotic risk than protein C or S deficiency. The heterozygous factor V Leiden mutation and prothrombin polymorphism are usually associated with the lowest thrombotic risk. Accordingly, clinical surveillance is appropriate only for the latter two conditions in asymptomatic women without previous history of thromboembolism. Although there is no definite evidence, it may be prudent to use daily prophylaxis with LMW heparin in pregnant women from symptomatic kindreds with protein C or with antithrombin III deficiency, even in the absence of a previous thromboembolic episode (C₂). In protein S deficiency, the same prophylaxis might be recommended in pregnant women with previous thrombosis only.

	Postpartum: European experts agree on the prophylaxis with UH heparin or LMW heparin overlapped and followed by warfarin for 4 to 6 weeks postpartum (C2)

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
Treatment of Venous Thromboembolism in Pregnant Women
Although a full dose of UH by IV infusion for 5 to 10 days followed by an injection of the full dose after 9 to 12 h until term is the treatment of choice in the North American document,² its duration at full doses is not accepted by all European experts. Some of them recommend a full therapeutic dose for 1 to 3 months, followed by a lower dose to limit the risk of osteoporosis. Full doses of heparin until delivery might be a lengthy treatment when the venous thrombosis occurs at the beginning of pregnancy. The severity of the thrombotic episode, its evolution during treatment, and its cause should probably also be considered in determining the duration of heparin treatment.

LMW heparin as an alternative to UH has been proposed and seems to offer some advantages,⁹ but there are no prospective clinical trials allowing general recommendations. Whether it should be given twice or once a day is still a matter of debate. Rare reports have used the second alternative.^{9 10} There is an agreement that the peak anti-factor Xa activity should be tested at least during the first month of LMW heparin treatment, then every month or trimester. However, the therapeutic range is uncertain,⁶ and standardization of anti-factor Xa activity measurement is essential. At present, LMW heparins, if used, should be administered bid to treat pregnant women with venous thromboembolism (C₂). The dose should be adjusted to peak anti-factor Xa concentrations according to the instructions provided by the manufacturers of the LMW heparin used for nonpregnant patients.

Pregnant Women With Mechanical Heart Valves
As indicated in the North American consensus document,² there are two approaches to anticoagulant therapy for pregnant women with mechanical heart valves. The first is to use UH throughout pregnancy, administered subcutaneously bid in doses adjusted to keep the mid-interval activated partial thromboplastin time (APTT) in the therapeutic range (at least twice the control), or an anti-factor Xa heparin concentration of 0.35 to 0.70 IU/mL plasma. However, some European experts are in favor of giving exclusively LMW heparin throughout pregnancy because of its comfort and better tolerance, although no firm data are available on the use of LMW heparin in this situation. It is difficult to make firm recommendations, except that all experts in North America and in Europe insist on the importance of planning the treatment together with the expectant couple.

The second approach is to use heparin until the 13th week, to switch to warfarin until the middle of the third trimester, and then to restart heparin treatment until delivery. This approach is encouraged by some experts because they consider it to be more effective than the first one.¹⁰

Pregnant Women With Antiphospholipid Antibodies
The North American recommendations are based on a high level of evidence (A₁), and are generally accepted by the European experts, who agree with the combined treatment of heparin and aspirin. LMW heparin is used almost universally in Europe.^{11 12 13}

	Prevention of Venous Thromboembolism

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
In general, this chapter¹⁴ is received as a well-documented overview on antithrombotic prophylaxis, but a number of qualifications should be made. A general concern is how to translate data from studies with surrogate end points (eg, venography) into recommendations for daily practice in which clinical efficacy must prevail (eg, fatal and nonfatal pulmonary embolism and symptomatic deep vein thrombosis).¹⁵ Some studies using clinical end points cast some doubt on the figures usually given for the frequency of postoperative pulmonary embolism, and question the need of perioperative antithrombotic prophylaxis for all patients.¹⁶ However, most European experts agree with the North American consensus that aspirin is less effective than heparin as a prophylactic agent.

A further concern is that clinical thrombotic end points are nowadays much rarer than those recorded in historical literature. For instance Fender et al¹⁷ carried out a retrospective analysis of > 2,000 consecutive operations for hip replacement in the Trent region of the United Kingdom. There were only four definite fatal pulmonary emboli, an incidence of 0.2%. The near impossibility of carrying out controlled clinical trials in this situation is readily seen in the context of trials of equivalence, where a 0.5% difference is required between two treatments before equivalence is discarded.

A first comment is on the classification of level of risk. The North American document uses four levels of risk,¹⁴ ) whereas other consensus documents in Europe use only three. In principle, a fourth level should allow the prophylactic regimen to be better tailored to the patient’s needs. However, if LMW heparin is the preferred agent, as in many European centers, it is not clear if each risk level will indeed receive a different regimen of LMW heparin. Furthermore, the text remains to some extent vague as to the category in which an individual patient has to be classified, eg, with one vs multiple additional risk factors. Few trials indeed studied homogeneous groups of patients except for total hip and knee replacement. The reader gets the impression that all patients with inherited thrombophilia automatically fall into the highest risk category, whereas some differentiation may be justified.

Antithrombotic regimens to prevent venous thromboembolism are detailed in the North American consensus report,¹⁴ ) but the list is not complete, inasmuch as in Europe a wider range of LMW heparins are licensed (eg, boxol, parnaparin, reviparin). In Table 1 of the North American document,¹⁴ ) the four levels of risk are not given; thus, it is difficult to relate the treatment regimens to those levels of risk given in Table 4 of the corresponding chapter of the document.¹⁴ ) The recommendations do not detail the LMW heparin regimens either; they suggest starting antithrombotic prevention 12 to 24 h after surgery in total hip replacement, which is clearly not standard European practice.

Because of its limited use, the coverage of dextran is brief. In renal insufficiency, it is only the 10% dextran-40 solution that should be used with caution. Rare anaphylactic responses can largely be avoided if the hapten principle dextran 1 is used.

It has been suggested in the North American consensus that aspirin is less effective than heparin as a prophylactic agent.¹⁴ However, a recent very large randomized trial of aspirin vs placebo in orthopedic patients has shown a 58% reduction in mortality in the aspirin-treated group compared with placebo.¹⁸ Heparin was not randomized in this trial and did not further reduce thromboembolism, but increased the amount of postoperative bleeding.

The evidence presented on the efficacy of elastic stockings in general surgery is meager and does not justify an A₁ recommendation.

A hot issue is the duration of prophylaxis. The North American document¹⁴ ) lists the studies with prolonged LMW heparin prophylaxis, and points correctly to a significant risk of venographic thrombosis during the first postoperative month, even in patients with a normal venography at discharge. It is true that the incidence of symptomatic thromboembolism was low, and the document therefore calls for additional trials on out-of-hospital prophylaxis in which objectively documented symptomatic venous thromboembolism is the primary efficacy end point. For some, this seems like going back to the future, because most of this chapter is based on studies with surrogate end points. In addition, it may raise an ethical issue in some European countries that use prolonged prophylaxis, because a low risk of fatal pulmonary embolism still represents a not insignificant number of lives in view of the scale of orthopedic surgery. Others believe that the question is not settled, and trying to reach a consensus now may appear premature.

Sections that eventually might be added are discussions on contraindications of pharmacological prophylaxis because of a too-high bleeding risk and on conditions for which there are at present little or no data, eg, laparoscopic surgery, arthroscopy, and application of plaster casts. These issues occasionally lead to medicolegal discussion. The last North American recommendation addresses the risk of epidural or spinal hematoma with LMW heparin in patients who receive regional anesthesia or analgesia.¹⁴ ) Is caution only warranted with LMW heparin and not with other antithrombotic drugs?

The available data pertaining to very heterogenous medical groups do not support an A₁ recommendation, although a recent publication is encouraging.¹⁹

Regarding orthopedic surgery, perhaps insufficient attention has been given to publications using clinical end points, and some question the degree of benefit of heparin to reduce fatal pulmonary embolism in orthopedic patients.²⁰ Further doubt on the A₂ recommendation of heparin antithrombotic prophylaxis for hip fracture surgery is raised by the Cochrane database.²¹ Here, the conclusion is made that insufficient trials of good quality, using clinical end points, have been conducted to allow definite recommendations in favor of antithrombotic prophylaxis to be made.

	Antithrombotic Therapy for Venous Thromboembolic Disease

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
The North American recommendations on the antithrombotic therapy of venous thromboembolism raise no fundamental disagreement.

A few points deserve attention to convert the recommendations into practical guidelines for daily use. For instance, the advice to prolong the APTT to a range that corresponds to a certain plasma heparin concentration (which in turn depends on the assay used) is universally valid. Conversion into a heparin nomogram for daily practice depends on the locally used APTT reagent, as explained in the chapter on heparin in the North American document. The substitution of LMW heparin for UH eliminates the problem of monitoring and meets with general support. European practice is probably a step ahead of the US Food and Drug Administration regulation: preparations other than enoxaparin are admitted as well, and pulmonary embolism is accepted as an indication for LMW heparin in a number of European countries. European doctors also have a larger choice of coumarin derivatives: phenprocoumon and acenocoumarol are prescribed as well as warfarin.

There is some concern about the North American recommendation to continue with a treatment dose of LMW heparin should oral anticoagulation be either inconvenient or contraindicated.²² There is no evidence to support this recommendation (although this treatment has been prescribed during pregnancy). A few unquoted studies tested lower doses of LMW heparin after the initial phase, but the data are still too scarce to justify a grade A₁ recommendation.^{23 24 25}

Divergent views persist among Europeans experts on the optimal duration of therapy for certain categories of patients. For instance, one view holds that patients with congenital or acquired thrombophilia may need long-term to indefinite anticoagulant treatment after a first episode of thromboembolism if this event was not precipitated by an additional transient risk factor. The opposite view is that clinical trials should assess the cost-benefit ratio before treating patients indefinitely, even those with recurrent venous thromboembolism or a continuing risk factor, because the risk of recurrence varies with the thrombophilic state. In a trial that compared 6 months to 4 years of anticoagulant treatment after a second episode of venous thromboembolism, the benefit of long-term treatment was offset by an increased incidence of major bleeding.²⁶ Thus, the current North American recommendation on the duration of treatment may also be a workable consensus in Europe.

Thrombolytic treatment evokes divergent comments. In Europe, its use in deep vein thrombosis is not generally advocated, except possibly in some cases of massive iliofemoral thrombosis. A prospective study found no difference in outcome between iliofemoral thrombosis and more distal thrombosis; therefore, the usefulness of thrombolysis in the former group may be questioned.²⁷ On the other hand, a more liberal use in pulmonary embolism has been advocated by some clinicians, eg, in patients with echocardiographic signs of right ventricular overload.²⁸ The ACCP opinion²² ) that thrombolysis is highly individualized (with regard to clinicians as well as to patients) also prevails in Europe.

	Antithrombotic Therapy in Atrial Fibrillation

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
The guidelines published by the North American group comprehensively review the evidence from trials and risk stratification for atrial fibrillation (AF).²⁹ Overall, the European experts agree with the recommendations, but the following items need to be addressed.

First, the evidence from epidemiologic data suggests that paroxysmal AF has an intermediate risk of stroke between chronic AF and sinus rhythm. However, trial data suggest that paroxysmal AF confers a similar risk of stroke to chronic AF.^{30 31} It would be reasonable to assume that the risk of stroke in paroxysmal AF is related to the frequency and duration of paroxysms, and it is more likely that a patient with one short paroxysm of AF per year is at less risk compared to someone with long paroxysms of AF occurring every day: we may therefore be dealing with a wide spectrum of thromboembolic risk. Important risk factors are age of the patient, influence of underlying pathology, such as structural heart disease, and hypertension. Another difficulty arises because many patients have asymptomatic or "silent" paroxysms. The precise risk to benefit ratio for thromboprophylaxis in patients with paroxysmal AF therefore remains unclear, although the current guidelines (including the North American guidelines) suggest treating patients with paroxysmal AF similar to patients with chronic AF.

In patients with very infrequent episodes of paroxysmal AF, the benefits of warfarin therapy may be significantly offset by its associated inconvenience and morbidity, and aspirin therapy may be a safer and reasonably effective option, at least for those patients aged < 75 years with no structural heart disease.^{32 33} However, in patients aged > 75 years who have underlying structural heart disease and frequent paroxysms of AF or in patients with prior transient ischemic accidents or minor stroke, warfarin therapy is recommended until clearer risk stratification strategies are available for such patients.^{34 35} Care is to be taken in the very elderly to balance the risk of anticoagulation, such as hemorrhage and noncompliance, with the possible benefit.

The North American guidelines²⁹ ) summarize risk stratification schemes from the Atrial Fibrillation Investigators Study³⁰ and the Stroke Prevention in Atrial Fibrillation-III trial.^{34 36} Their own recommendations²⁹ ) appear to be a synthesis of both, and the role of echocardiography in risk stratification is not precise.

The North American recommendations should clearly emphasize that risk stratification can be performed using clinical guidelines in the majority of cases, with only a little refinement from echocardiography, as evident from recent reports.^{37 38} The meta-analysis from the AF Investigators of 1,066 patients from three clinical trials also suggested that echocardiography refined clinical risk stratification, and that only moderate-to-severe left ventricular systolic dysfunction via two-dimensional transthoracic echocardiography independently predicted stroke risk.³¹ Contrary to previous reports, isolated left atrial dilatation on the echocardiogram was not an independent risk factor for increased stroke risk on multivariate analysis. Echocardiography is therefore useful in the small group of patients who have a low risk of stroke according to clinical risk factors. Only very rarely would clinicians need to proceed to transesophageal echocardiography to assist risk stratification, but in the Stroke Prevention in Atrial Fibrillation-III transesophageal echocardiography substudy, stroke and thromboembolism were correlated with dense spontaneous echocardiographic contrast, left atrial appendage thrombus, and complex aortic plaques.³⁹

The European experts are in agreement with current recommendations of a minimum 3 weeks of warfarin therapy before cardioversion and 4 weeks after cardioversion.

	Antithrombotic Therapy in Valvular Heart Disease

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
The guidelines published by the North American group comprehensively review the evidence for antithrombotic therapy in valvular heart disease.⁴⁰ Overall, the recommendations can be supported, but a small number of qualifications should be made. In the North American text reviewing evidence and recommendations, increased left atrial size is stated as a risk factor for stroke, especially in the presence of AF. The meta-analysis mentioned above, in 1,066 patients, found that only moderate-to-severe left ventricular systolic dysfunction via two-dimensional transthoracic echocardiography independently predicted stroke risk in nonvalvular AF. Contrary to previous reports, isolated left atrial dilatation on the echocardiogram was not an independent risk factor for increased stroke risk on multivariate analysis.³¹

Whether left atrial dilatation increases stroke risk in mitral valve disease needs to be clearly defined. Mitral annular calcification has not been identified on multivariate analysis as an independent risk factor for stroke and thromboembolism in AF,³¹ although the North American recommendations for anticoagulation in older subjects (ie, age > 75 years) seems reasonable.²⁹ ) The distinction between the relative stroke risk of mitral stenosis and mitral regurgitation is less clear. Mitral stenosis is associated with a high risk of stroke and systemic thromboembolism, especially if AF is present. However, there is some evidence that mitral regurgitation of sufficient severity may prevent the development of left atrial spontaneous echocardiographic contrast on transesophageal echocardiography and intracardiac thrombi, suggesting a protective effect of the regurgitant jet against thrombus formation (thrombogenesis).^{41 42} Whether this translates into significant clinical differences with respect to antithrombotic therapy still needs some clarification.

	Antithrombotic Therapy in Patients With Mechanical and Biological Prosthetic Heart Valves

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
Making recommendations on antithrombotic treatment in patients with prosthetic heart valves⁴³ is a difficult subject, as acknowledged in the first paragraph by the North American Consensus Group.⁴⁴ However, it is noted that the issue has not been addressed in terms of basic concepts. There is too much emphasis on reported rates of embolism, without discussion of the mechanisms involved in these predominantly cerebrovascular events. No mention is made of the prevention of valve thrombosis, the primary purpose of anticoagulation in these patients. Valve thrombosis is a much more serious and life-threatening event than most embolic events, although it occurs less frequently. Among patients receiving anticoagulation therapy, the incidence of valve thrombosis varies from 0 to 0.5%/yr in the aortic position and from 0 to 3%/yr in the mitral position.^{45 46 47} Moreover, valve thrombosis rates give a much better indication of thrombogenicity of a particular prosthesis than embolic rates, which are heavily influenced by patient risk factors, eg, the presence of coronary artery, aortic or carotid artery disease, hypertension, diabetes, increased left ventricular mass, or cigarette smoking,⁴⁸ a problem not discussed in the North American document. Unless the prevalence of these risk factors is known in a series of prosthetic valve patients, it is difficult to interpret data on embolism. The North American document assumes that all embolic events originate from the prosthesis, despite the fact that many other pathophysiologic mechanisms lead to cerebral infarction.⁴⁷

The North American experts discuss primarily the problems associated with the St. Jude valve, by far the most common prosthetic valve used in the United States. However, in many other countries, the St. Jude valve is not the most commonly used valve. The article devotes little space to discussing tilting disk and caged-ball valves. It is assumed in this article that all prosthetic heart valves within a broad design category (for example, bileaflet or tilting disk) are equivalent in terms of thromboembolic risk. Examination of the prosthetic heart valve literature shows that this is not the case. Two obvious examples spring to mind. The Medtronic parallel bileaflet valve (Medtronic; Minneapolis, MN) was withdrawn before clinical trials were completed because of an unacceptably high incidence of valve thrombosis. It cannot therefore be assumed that all bileaflet valves are of equal thrombogenicity. Similarly, among tilting disk valves, there is a wide range of thrombogenicity. The Medtronic Hall valve has the lowest thrombogenicity on the basis of valve thrombosis rates, whereas the Lillehei-Kaster/Omniscience series of valves has the highest thrombogenicity. In the United Kingdom, the Omniscience valve was found to have an unacceptably high incidence of valve thrombosis when patients were treated with anticoagulants at the level recommended by the North American Consensus.⁴⁸

Prominence is given to the article of Stein et al⁴⁹ on the optimal international normalized ratio (INR), despite the facts that this article was based on an analysis of the literature in which most studies were uncontrolled and retrospective, that definitions of thromboembolic and bleeding events differed among studies, and that the sensitivity of North American thromboplastins used was not reported, shortcomings that invalidate the conclusions.

The article of Horstkotte et al⁵⁰ is quoted on the basis that it was a prospective series. However, although the data may have been collected prospectively, the conversion to INR from the Quick test used in Germany during the period of observation was performed retrospectively, based on the assumption that each anticoagulant clinic had used the same thromboplastin reagent throughout a 12-year period. Furthermore, the separation of the patients into three groups was based on an average INR during a 12-year period, which seems unrealistic, especially when one takes into account the number of events, both thromboembolic and hemorrhagic, that occur while the patient is outside the therapeutic range.

In the very short paragraph on tilting disk valves, it is stated that Cannegieter and associates⁵¹ showed trends that suggested that an INR of 3.0 to 3.9 was optimal for tilting disk valves. The particular tilting disk valves in this series were predominantly Bjork-Shiley valves of different generations, and it has been shown that data from one type of tilting disk valve (particularly older-generation tilting disk valves) cannot be extrapolated to other tilting disk valves.⁴⁸ The same reference in the North American Consensus is also used to support a statement that thromboembolic rates are higher with caged ball and caged disk valves. In fact, the number of caged ball and caged disk valves in this series was extremely small, so that it was impossible to demonstrate a significant difference between these types of prostheses and the other prostheses.

In general, the authors⁴³ are too ready to recommend aspirin in addition to warfarin for risk factors for which there is no evidence of benefit, while appearing to diminish the importance of the increased risk of bleeding. The benefit of dipyridamole cannot be ignored, as is shown by two recent meta-analyses.^{52 53}

Patients with mechanical prosthetic heart valves who suffer systemic embolism despite adequate anticoagulation should first be investigated for a possible source of embolism before aspirin is simply added to their medication. Some European experts favor dipyridamole compared with low-dose aspirin in view of a lower bleeding risk. The relative efficacy of dipyridamole over aspirin in this indication is unclear.

	Antithrombotic Agents in Coronary Artery Disease

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
This chapter was generally well received by the European reviewers.⁵⁴ The statement is made that new trials of LMW heparins have been published since the fifth consensus document. This is unavoidable in periodic guidelines updates. Reference has also been made to the direct thrombin inhibitors, desirudin, and to long-term treatment with oral glycoprotein IIb/IIIa inhibitors.

Concerning the dosage of aspirin, the difference between Europe and United States arises from the formulation of low-dose aspirin. In the United States, the low-dose tablet contains 81 mg of acetylsalicylic acid, whereas in Europe, it is 75 mg. In patients with unstable coronary syndromes, ie, evolving myocardial infarction without persistent ST-segment elevation and unstable angina, 75 mg is an adequate, well-documented lowest dose. Also in chronic coronary artery disease, a dose of oral aspirin, 75 mg, is definitely the best documented dose concerning efficacy as well as safety. On the basis of the International Study on Infarct Survival-II results, aspirin, 150 mg, might be the lowest effective dose during the first month after myocardial infarction, either with or without initial thrombolysis.⁵⁵

Regarding the effects of IV glycoprotein IIb/IIIa receptor blockers, it should be emphasized that effects have been obtained in patients with unstable angina, treated or not with invasive procedures. So far, a short-term treatment with an IV glycoprotein IIb/IIIa inhibitor agent in patients without early coronary angiography and revascularization has been associated with differences in long-term gains. Limited efficacy in a noninvasive setting was clearly demonstrated in the PURSUIT trial (Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppression Using Integrilin Therapy), which is the largest ever performed in this setting.⁵⁶ The Platelet Receptor Inhibition for Ischemic Syndrome Management trial with tirofiban was conducted in patients with more severe unstable angina and resulted in a significant reduction of death and myocardial infarction at 7 days, 30 days, and at 6 months, with demonstrable effects in the noninvasive as well as invasive setting.⁵⁷ Therefore, the North American recommendation for the use of IV glycoprotein IIb/IIIa inhibitors should be modified to widen their indication. They are presently recommended for unstable angina of sufficient severity to warrant coronary care unit admission. The reason for coronary care unit admission is variable among hospitals and among countries, and there is no scientific basis for such a recommendation. The European reviewers propose that IV glycoprotein IIb/IIIa receptor inhibitors can be considered for patients hospitalized with unstable coronary syndromes and with refractory angina, together with conventional treatment with aspirin and heparin or LMW heparin, especially if early coronary angiography and revascularization procedures are planned or elevated troponin concentrations are present.⁵⁸

In the recommendations of the treatment of unstable angina with anticoagulants, European investigators suggest that, so far, two LMW heparins, enoxaparin and dalteparin, can be considered to be the first choice rather than IV UH infusion (A₁ level).^{59 60 61} The comparison between enoxaparin and UH shows that the effects of enoxaparin are equivalent or better than IV UH. There are also fewer side effects with LMW heparins, especially thrombocytopenia. The time has come to recommend those LMW heparins, tested and proved superior to UH in this clinical setting, as the first choice in the unstable coronary syndrome.^{59 60 61}

Another questionable issue is the recommendation to maintain heparin treatment for at least 48 h in unstable coronary syndromes. In studies on heparin in addition to aspirin in unstable coronary syndromes, all the placebo-controlled trials of UH or LMW heparin have used at least 5 days of treatment. This recommendation is not refuted by other trials comparing 2 to 3 days treatment with LMW heparin vs UH.^{59 60} In these trials, it is completely unknown whether the outcome of such a short-term treatment would be any better than placebo for more serious events, such as myocardial infarction and death. Thus, based on available evidence, the European reviewers propose that patients hospitalized with unstable angina or non-Q-wave myocardial infarction should, in addition to aspirin, be treated with LMW-heparin subcutaneously bid. Although the ideal treatment based on clinical trials is for 5 days or until a coronary procedure, if one is performed earlier, in practice, many patients are discharged from hospital after 2 to 3 days. If a coronary procedure is planned within the next few days or weeks, the LMW heparin treatment should be maintained until the procedure. As an alternative, conventional IV UH therapy can be used, which might be an advantage if the coronary procedure is planned within the initial 24 to 48 h after admission.

Regarding the prevention of deep vein thrombosis in myocardial infarction patients, it is stated in the North American Consensus document that there are no trials of IV low-dose heparin vs no heparin in myocardial infarction patients receiving aspirin and thrombolytic therapy.⁵⁴ ) Much of the thrombotic prophylactic data are indeed old and based on the surrogate end points of labeled fibrinogen scanning or venography, and do not relate to real-life events. One can therefore question the strong recommendation⁵⁴ ) that every patient with acute myocardial infarction should receive low-dose heparin therapy until ambulation.

Regarding patients with acute myocardial infarction, where no thrombolytic therapy has been given, the North American Consensus Group recommends low-dose UH treatment, 7,500 IU subcutaneously q12h, in addition to aspirin.⁵⁴ ) There is very little evidence that treatment over and above standard aspirin, eg, subcutaneous heparin, is effective. Some of these patients will belong to the unstable coronary syndromes, which include non-Q-wave myocardial infarctions. Thus, in this subgroup, there is a clear indication on the basis of numerous trials for full-dose LMW heparin subcutaneously bid in addition to aspirin for 5 days.

In the primary prevention of thromboembolism in asymptomatic individuals > 50 years old and with at least one major atherothrombotic risk factor, aspirin, 75 mg qd, is also considered a reasonable proposition in Europe, but this should be strengthened by more clinical evidence. This recommendation is also based on the Hypertension Optimal Treatment trial consisting of 20,000 hypertensive individuals.⁶² A daily dose of aspirin, 75 mg, reduced myocardial infarction by 30% without increased risk of stroke. However, the primary aim should be to reduce the risk factors, such as raised cholesterol, high BP, or smoking habit. Concerning the use of low-intensity warfarin and its combination with aspirin for primary coronary prevention, the available evidence is fairly sparse at present.

	Coronary Thrombolysis

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
The European reviewers agree with the North American recommendations on the use of thrombolytic drugs in acute myocardial infarction,⁶³ but have some reservations that the numerous subgroups described may somewhat confuse the practicing physician or busy cardiologist. The recommendation to start with IV ß-blockade in the acute phase of myocardial infarction in all patients without contraindications is based on the International Study on Infarct Survival-II and Metoprolol in Acute Myocardial Infarction trials, in which the overall very low mortality rate renders the interpretation difficult.^{64 65} In one trial, IV ß-blockade was even harmful.⁶⁶ One also wonders whether the use of angiotensin-converting enzyme inhibitors would not be limited to patients with congestive heart failure or hypertension, rather than recommended for all patients with myocardial infarction.

In general, there is a good correlation between the coronary patency rates at 90 min after start of thrombolytic treatment and 30-day mortality, except for the Reteplase vs Alteplase Patency Investigation During the Acute Myocardial Infarction-II trial,⁶⁷ in which no difference in mortality was found for patients with either grade 2 or 3 angiographic coronary flow as defined by the study Thrombolysis in Myocardial Infarction.

In discussing the equivalence or absence of superiority in direct comparisons among thrombolytic agents, an introductory paragraph on the concept of equivalence would be desirable, and the upper boundary for equivalence or noninferiority needs to be specified. In a conventional equivalence trial, the starting point is the belief that treatments are identical, and the objective is to confirm this. There is no general agreement about what is acceptable as equivalence in the effect of treatments in myocardial infarction. In the first trial of thrombolytic agents in which equivalence was the declared end point (International Joint Efficacy Comparison of Thrombolytics),⁶⁸ equivalence was defined as the fatality rates of the two treatments being within 1%. On this basis, streptokinase and reteplase were equivalent. However, in the Continuous Infusion vs Double-Bolus Administration of Alteplase study⁶⁹ which compared two regimens of alteplase, equivalence was defined as the total mortality in the two groups being within 0.4%. In as much as a higher mortality was observed with the new, double-bolus regimen, equivalence could not be claimed (upper 95% boundary of equivalence, 0.49%, which exceeded the prespecified upper limit of 0.40%). The recent Assessment of the Safety and Efficacy of a New Thrombolytic Agent-II⁷⁰ and Intravenous mTPA for Treatment of Infarcting Myocardium Early-II⁷¹ trials were also equivalence or noninferiority trials that used a 1% absolute or 14% relative difference⁷⁰ and a ratio of 1.196 as upper limits⁷¹ of equivalence for 30-day mortality. There are clearly statistical problems here that need to be resolved in guidelines.

The chapter could also have included a final paragraph on the generality of the results of clinical trials, in which the mortality rate is so much lower than those observed in complete registers of all heart attack admissions.

	Antithrombotic Therapy in Patients With Saphenous Vein and Internal Mammary Artery Bypass Grafts

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
In the North American document, the recommended dosage of aspirin in patients with coronary bypass graft is 325 mg/d started within 6 h of surgery (A₁ level).⁷² It is stated that aspirin, 100 mg/d, was not consistently effective, referring to two small trials, one of which was negative and the other equivocal. However, the results in this context of a large trial have been ignored.⁷² ) In this trial, aspirin 50 mg tid, and aspirin 50 mg tid plus dipyridamole, were compared with placebo.⁷³ Occlusion rates were 14% (aspirin), 13% (aspirin plus dipyridamole), and 18% (placebo). The difference between aspirin and placebo approached statistical significance in a ratio estimate analysis, whereas multivariate analysis revealed a significant benefit (odds ratio, 0.73; 95% confidence intervals, 0.54 to 0.99). Furthermore, meta-analyses performed by the Antiplatelet Trialist’s Collaboration in these and other high-risk patients provided no evidence that low-dose aspirin is less effective than high-dose aspirin.⁷⁴

For these reasons, the recommended dose of aspirin after coronary artery bypass surgery could be lowered to 75 mg qd. Aspirin is recommended for circa 1 year, as longer-term trials demonstrating benefit in terms of patency have not been conducted. Its continued use after 1 year in patients with saphenous vein to coronary artery grafts is justified on the grounds of the beneficial effect of aspirin in coronary artery disease. There is no unequivocal support for the use of coumarin anticoagulation, although the trials were too low-powered to exclude benefit in all patients and circumstances.⁷⁵

Oral anticoagulants rather than aspirin, alone or in combination with aspirin, should be considered after coronary artery bypass surgery in patients in whom oral anticoagulants are simultaneously indicated, for example because of heart valve replacement or AF.

Recommending ticlopidine in patients who are allergic to aspirin is one option. Since the publication of the Clopidogrel vs Aspirin in Patients at Risk of Ischaemic Events study,⁷⁶ clopidogrel instead of ticlopidin should now be recommended. Warfarin is another alternative, considering that several trials showed no differences comparing oral anticoagulants to either aspirin or aspirin plus dipyridamole.

Antithrombotic drugs showed no benefit with regard to patency of internal mammary artery (IMA) grafts, mainly due to their low occlusion rates. It is likely that this observation was biased by selection of patients and coronary arteries to be grafted. In an analysis of vein and IMA grafts,⁷⁷ location of distal anastomosis and internal diameter of the grafted coronary artery were identified as independent predictors of graft occlusion rather than graft material (vein or IMA). Thus, apart from coronary heart diseases, antithrombotic drug therapy, preferably aspirin, should be recommended after IMA grafting.

	Antithrombotic Therapy in Peripheral Arterial Occlusive Disease

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
Full agreement exists with the recommendations made by the North American consensus group,⁷⁸ but a few qualifications are required.

First, it is reemphasized that the main aims of antithrombotic therapy are to prevent cardiovascular events and to retard local progression of the disease. These aims are not well demonstrated in patients with peripheral arterial occlusive disease. Particularly, the data on aspirin are poor in this category of patients: no single study with clinical end points yields significant results, and the reduction of vascular mortality, nonfatal stroke, and myocardial infarction was only 15.7% (not significant) in the Antiplatelet Trialist’s meta-analysis.⁷⁴ One suggestion is to split this category: for patients with known cardiovascular or cerebrovascular disease (those who survived a myocardial infarction or stroke), the recommendation could remain at an A₁ level; for the rest, a level of B₂ is more appropriate. In addition, the data on aspirin are more scanty than that on ticlopidine, which received a B₂ level recommendation. Not all the available data with ticlopidine are referred to in the consensus text, and it may well be that ticlopidine has better credentials than aspirin. Therefore, patients who do not tolerate aspirin or who experience a major cardiovascular event despite aspirin intake should be changed to ticlopidine, or preferably to clopidogrel, which has similar clinical efficacy but a better safety profile.⁷⁹ Of course, cost-benefit is another aspect that deserves further attention, and this applies as well for clopidogrel as for ticlopidine.

Intra-arterial thrombolysis for acute leg ischemia is practiced in a number of European centers with alteplase (recombinant tissue plasminogen activator [rtPA]) as well as with urokinase. The Surgery vs Thrombolysis for Ischemia of the Lower Extremity trial⁸⁰ found no difference in efficacy and safety between rtPA and urokinase, but the summary table only retains urokinase as a B₂ recommendation, which most probably corresponds to the US practice.

Newer data on prostaglandins and cilostazol published in the last 2 years will presumably be discussed in the next North American consensus conference. Although prostaglandins do have antiplatelet effects, their use in end-stage leg arterial disease is not intended to compete with aspirin. Defibrotide is an oligonucleotide endowed with antithrombotic properties and tested in Italy, but its effectiveness needs to be confirmed in further trials. Regarding trials with pentoxifylline, the end point was intermittent claudication, not thrombotic events.

	Antithrombotic and Thrombolytic Therapy for Ischemic Stroke

TOP Abstract Introduction Use of Antithrombotic Agents... Postpartum: European experts... Prevention of Venous... Antithrombotic Therapy for... Antithrombotic Therapy in Atrial... Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic Agents in... Coronary Thrombolysis Antithrombotic Therapy in... Antithrombotic Therapy in... Antithrombotic and Thrombolytic... Antithrombotic Therapy in... Antithrombotic Therapy in... Expression of Benefit Need for Further Clinical... Appendix 1 References

 
This chapter is an impressive document, but is clearly written from a US perspective, with overwhelming emphasis on thrombolytic therapy using alteplase (rtPA).⁸¹ For European readers for whom rtPA use in acute stroke is not approved or even prohibited (eg, in France), the discussion may be less relevant.

Acute Treatment of Ischemic Stroke
A discussion on the heterogeneity of ischemic stroke would have probably been a better starting point for this chapter. Subtyping ischemic stroke is crucial to patient management. Which clinician can seriously admit that the benefits and risks of alteplase (rtPA) are similar in a young person with embolism from the heart to the middle cerebral artery, as in a frail hypertensive octogenerian with a lacunar infarct? A post hoc analysis on