Anticoagulants, Antiplatelets, and Thrombolytics, Second Edition (Methods in Molecular Biology 663)

ME T H O D S

IN

MO L E C U L A R BI O L O G Y

Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK

For other titles published in this series, go to www.springer.com/series/7651

TM

Anticoagulants, Antiplatelets, and Thrombolytics Second Edition

Edited by

Shaker A. Mousa Pharmaceutical Research Institute, Albany College of Pharmacy and Health Sciences, Rensselaer, NY, USA

Editor Shaker A. Mousa Albany College of Pharmacy and Health Sciences Pharmaceutical Research Institute One Discovery Drive Rensselaer, NY 12144 USA [email protected]

ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-60761-802-7 e-ISBN 978-1-60761-803-4 DOI 10.1007/978-1-60761-803-4 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2010929846 © Springer Science+Business Media, LLC 2003, 2010 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Humana Press is part of Springer Science+Business Media (www.springer.com)

Preface The past 2 decades have witnessed significant advances in the discovery and development of novel drugs to prevent and treat thromboembolic disorders, such as oral direct antiXa and anti-IIa (thrombin) antagonists, as well as oral antiplatelet ADP antagonists with rapid onset and offset. The introduction of direct oral factor Xa and thrombin inhibitors that do not require monitoring and have no significant food or drug interactions represents a significant advance that may lead to the replacement of oral warfarin and injectable heparin or low molecular weight heparin (LMWH) in some, but probably not all, indications. In addition, there has been concentrated effort aimed at identifying novel uses of traditional antithrombotic drugs as well as combinations of agents, such as more than one antiplatelet, or antiplatelet plus anticoagulant. These tremendous achievements have resulted in improved management of arterial and venous thromboembolic-associated disorders. Although the morbidity and mortality resulting from acute coronary disease has been reduced by more than 50% over the past 30 years, it is reasonable to anticipate further reductions of similar magnitude in the decade ahead. Advances in our understanding of the mechanisms of pathogenesis of venous thromboembolism (VTE), acute coronary syndromes, cerebral vascular ischemia, and diseases associated with thrombotic events have provided critical insight for the development of various therapeutic approaches to control these pathogenic events. The roles of plasmatic proteins, blood cells, vascular endothelium, and target organs in thrombogenesis are becoming more clear. Identification of endogenous inhibitors of thrombogenesis such as antithrombin III, tissue factor pathway inhibitor (TFPI), protein C, prostacyclin, nitric oxide, and physiologic activators of fibrinolysis has led to the development of both direct and indirect modalities to treat thrombosis. Knowledge of the proteases involved in thrombogenesis, as well as tissue factor, coagulation factors, adhesion molecules, and fibrinolytic inhibitors, has provided additional insight into the mechanisms by which thrombogenesis can be pharmacologically controlled. All of these novel strategies could not have happened without the utilization of key in vitro and in vivo clinically relevant experimental models for the screening and evaluation of these novel antiplatelets, anticoagulants, and thrombolytics (discussed in Chapters 1 and 2). Newly developed anti-Xa agents are characterized by high affinity and selectivity for Xa as compared to other serine proteases. In addition to their inhibitory effects on plasmatic coagulation processes, including thrombin generation, thrombin-mediated platelet reactions, and clot-bound pro-thrombinase complexes, there is evidence that some of these agents might interfere with receptor-mediated intracellular signaling events induced by factor Xa that regulate proliferation of vascular smooth muscle cells and other cells. The current outlook for anti-Xa agents is that they have the potential to become important prophylactic and treatment drugs for various venous thromboembolic disorders as well as adjuvants to other antithrombotic therapies in arterial thrombosis. Major advances in the development of oral anticoagulants are progressing very well, with the goal of developing safe and effective oral anticoagulants that do not require frequent monitoring or dose adjustment and that have minimal food/drug interactions.

v

vi

Preface

Vitamin K antagonist, with its inherent limitations of multiple food and drug interactions and frequent need for monitoring, remains the only oral anticoagulant currently approved for long-term secondary thromboprophylaxis in VTE. The oral direct thrombin inhibitor ximelagatran was withdrawn from the world market due to safety concerns. Newer anticoagulant drugs such as injectable pentasaccharides (e.g., idraparinux, SSR126517E), oral direct thrombin inhibitors (e.g., dabigatran), oral direct factor Xa inhibitors (e.g., rivaroxaban, apixaban, YM-150, DU-176b), and tissue factor/factor VIIa complex inhibitors are “tailor-made” to target specific pro-coagulant complexes and have the potential to greatly expand oral antithrombotic targets for both acute and long-term treatment of VTE, acute coronary syndromes, and prevention of stroke in atrial fibrillation patients (discussed in Chapter 5). The oral direct factor Xa inhibitor rivaroxaban represents a potentially attractive alternative to warfarin as it may enable simplified once daily dosing, appears to require no therapeutic monitoring, and has lower potential for drug interactions. At present, the safety and efficacy of rivaroxaban for the prophylaxis and treatment of VTE have been evaluated in phase-II and phase-III trials involving over 24,000 patients. In addition, rivaroxaban is currently being evaluated for the treatment of pulmonary embolism, secondary prevention after acute coronary syndromes, and the prevention of stroke and non-central nervous system embolism in patients with non-valvular atrial fibrillation. Several other oral direct anti-Xa inhibitors are in advanced clinical development, approved in Europe and under FDA review for the prevention and treatment of thromboembolic disorders (discussed in Chapter 6). Dabigatran is a novel oral direct reversible (fast onset and offset) thrombin inhibitor that binds to both free and clot-bound thrombin with a high affinity and specificity. Dabigatran has predictable and reproducible pharmacokinetics that are not affected by interactions with food. It is not metabolized by CYP450, does not induce nor inhibit CYP450, resulting in low potential for drug interactions, and does not require coagulation or platelet monitoring. The RE-NOVATE trial demonstrated that oral dabigatran etexilate at fixed doses is a well-tolerated alternative to injectable enoxaparin for the prevention of VTE after total knee replacement. The RELY trial demonstrated that oral dabigatran etexilate concurrently reduces both thrombotic and hemorrhagic events at two different doses (150 and 110 mg BID), exhibiting different and complimentary advantages over warfarin. At a dose of 150 mg BID, dabigatran had superior efficacy with similar bleeding, while at a dose of 110 mg BID, there was significantly less bleeding with similar efficacy in patients with atrial fibrillation at risk of stroke. Based on the accumulating clinical evidence, dabigatran represents the future of anticoagulation in the prevention and treatment of venous and arterial thrombosis alone and in conjunction with current antiplatelets and thrombolytics. Anti-platelet therapies remain a major focus in drug development. While aspirin is still considered the gold standard for antiplatelet therapy because of its high benefit-to-cost and benefit-to-risk ratios, ADP receptor antagonists, including ticlopidine, clopidogrel, and prasugrel, represent significant additions to aspirin in the management of different forms of arterial thromboembolic disorders (Chapter 7). Prasugrel is a novel thienopyridine that inhibits the platelet P2Y12 receptor and provides more rapid and consistent platelet inhibition than clopidogrel (Chapter 8). It is becoming clear, however, that there is variability in individual responses to antiplatelet agents such as clopidogrel, which may limit their widespread implementation. Various definitions of “non-responders” to antiplatelet therapy (i.e., aspirin resis-

Preface

vii

tance) tend to compound this issue. Aspirin resistance refers to aspirin-treated patients that are insensitive to aspirin treatment based on ex vivo tests of platelet activation and who experience recurrent cardiovascular disease. Estimates of aspirin resistance based on these criteria range from 20 to 80%, indicating that ex vivo tests are not an optimal tool for such assessments. In long-term aspirin-treated patients, there is evidence of low level but functionally relevant platelet thromboxane A2 formation, which was responsible for enhanced platelet activation in response to platelet agonists. These studies, however, did not fully exclude aspirin compliance, which could be a factor in such a phenomenon. Two trials performed in patients with coronary artery disease demonstrated that laboratory evidence of aspirin resistance was not detectable when aspirin compliance was accurately monitored. The same phenomenon was reported for other anti-platelet drugs such as clopidogrel. Given the multi-factorial nature of atherothrombosis, recurrence of cardiovascular events in aspirin-treated patients is not necessarily suggestive of drug failure. A cause-effect relationship between platelet insensitivity to aspirin and cardiovascular recurrence has not been defined overall because aspirin compliance has rarely been considered. Until such crucial information is taken into account, it would be prudent to take into consideration the distinction between “clinical resistance” to aspirin and resistance to taking the drug. To carefully define anti-platelet resistance, issues such as dose levels, standardized monitoring parameters, drug-drug interactions, and drug monitoring to document compliance should be addressed in future studies. The next decade should see considerable attention focused on the vascular endothelium, which occupies a strategic position at the interface between tissue and blood. The normal endothelium releases multiple antiplatelet, anti-inflammatory, thrombolytic and vasodilator molecules, such as prostacyclin and nitric oxide, which are potent inhibitors of platelet and monocyte activation and function as vasodilators. In addition, the normal endothelial surface expresses other protective molecules, including ecto-ADP, which degrades ADP, leading to inhibition of platelet aggregation; thrombomodulin, which activates protein C; and heparin-like molecules, which serve as cofactors for antithrombin III and heparin. The normal endothelium also secretes tissue plasminogen activator, which activates fibrinolysis. Insult or injury to the endothelium is accompanied by loss of these protective molecules and induction of expression of adhesive, pro-coagulant and pro-inflammatory molecules, vasoconstrictors, and mitogenic factors, leading to the development of thrombosis, smooth muscle cell migration and proliferation, and atherosclerosis. Hence, protective mechanisms of endothelial function represent new frontiers in the prevention and treatment of thromboembolic disorders that will have minimal effect on hemostasis. Improved understanding of the cell biology of plaque instability and endothelial hemostasis will promote a number of novel therapeutic strategies, including passivation of the endothelium, reduction of low-density lipoprotein (LDL) in the vessel wall (through decreasing serum LDL levels or accelerating reverse cholesterol transport), inhibition of LDL oxidation, thereby raising high density lipoprotein (HDL), and inhibition of inflammatory cytokine expression, as well as inhibition of thrombus formation upstream in the coagulation cascade or inhibition of activation of coagulation. The recognition that thrombotic disorders represent a syndrome rather than a disease is of crucial importance in the development of newer drugs. Either a poly-therapeutic approach with drug combinations or a drug with multiple actions will likely be more appropriate for the management of thrombotic disorders.

viii

Preface

This second edition of Anticoagulants, Antiplatelets, and Thrombolytics provides updates on various strategies in thrombosis, experimental models, and clinical and recent advances in the discovery and development of novel antithrombotics. Future directions in the coming decade should focus on the prevention of thromboembolic disorders and the protection of the vascular endothelium. Shaker A. Mousa

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v

Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xi

1.

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics . . . . . . . Shaker A. Mousa

1

2.

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics . . . . Shaker A. Mousa

29

3.

Heparin and Low-Molecular Weight Heparins in Thrombosis and Beyond . . . . 109 Shaker A. Mousa

4.

Laboratory Methods and Management of Patients with Heparin-Induced Thrombocytopenia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Margaret Prechel, Walter P. Jeske, and Jeanine M. Walenga

5.

Novel Anticoagulant Therapy: Principle and Practice . . . . . . . . . . . . . . . 157 Shaker A. Mousa

6.

Oral Direct Factor Xa Inhibitors, with Special Emphasis on Rivaroxaban . . . . . 181 Shaker A. Mousa

7.

Antiplatelet Therapies: Drug Interactions in the Management of Vascular Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Shaker A. Mousa

8.

Prasugrel: A Novel Platelet ADP P2Y12 Receptor Antagonist . . . . . . . . . . . 221 Shaker A. Mousa, Walter P. Jeske, and Jawed Fareed

9.

Antithrombotic Effects of Naturally Derived Products on Coagulation and Platelet Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Shaker A. Mousa

10.

Assessment of Anti-Metastatic Effects of Anticoagulant and Antiplatelet Agents Using Animal Models of Experimental Lung Metastasis . . . . . . . . . 241 Ali Amirkhosravi, Shaker A. Mousa, Mildred Amaya, Todd Meyer, Monica Davila, Theresa Robson, and John L. Francis

11.

Adhesion Molecules: Potential Therapeutic and Diagnostic Implications . . . . . 261 Shaker A. Mousa

12.

Pharmacogenomics in Thrombosis . . . . . . . . . . . . . . . . . . . . . . . . 277 Shaker A. Mousa

13.

Diagnosis and Management of Sickle Cell Disorders . . . . . . . . . . . . . . . 291 Shaker A. Mousa and Mohamad H. Qari

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309

ix

Contributors MILDRED AMAYA • Florida Hospital Center for Thrombosis Research, Orlando, FL, USA ALI AMIRKHOSRAVI • Florida Hospital Center for Thrombosis Research, Orlando, FL, USA MAMDOUH BAKHOS • Department of Thoracic and Cardiovascular Surgery, Stritch School of Medicine, Loyola University, Maywood, IL, USA MONICA DAVILA • Florida Hospital Center for Thrombosis Research, Orlando, FL, USA JAWED FAREED • Department of Pathology, Stritch School of Medicine, Loyola University, Maywood, IL, USA JOHN L. FRANCIS • Florida Hospital Center for Thrombosis Research, Orlando, FL, USA WALTER P. JESKE • Department of Thoracic & Cardiovascular Surgery, Stritch School of Medicine, Loyola University, Maywood, IL, USA TODD MEYER • Florida Hospital Center for Thrombosis Research, Orlando, FL, USA SHAKER A. MOUSA • Pharmaceutical Research Institute, Albany College of Pharmacy and Health Sciences, Rensselaer, NY, USA MARGARET PRECHEL • Department of Pathology, Stritch School of Medicine, Loyola University, Maywood, IL, USA MOHAMAD H. QARI • College of Medicine, King Abdul-Aziz University, Jeddah, Saudi Arabia THERESA ROBSON • Florida Hospital Center for Thrombosis Research, Orlando, FL, USA JEANINE M. WALENGA • Department of Thoracic & Cardiovascular Surgery, Stritch School of Medicine, Loyola University, Maywood, IL, USA

xi

Chapter 1 In Vitro Methods of Evaluating Antithrombotics and Thrombolytics Shaker A. Mousa Abstract Platelets play a crucial role in primary hemostasis by forming hemostatic plugs at sites of vascular injury. There is abundant evidence that platelets also play a pivotal role in the pathogenesis of arterial thrombotic disorders, including unstable angina (UA), myocardial infarction (MI), and stroke. The underlying pathophysiological mechanism of these processes has been recognized as the disruption or erosion of a vulnerable atherosclerotic plaque, leading to local platelet adhesion and subsequent formation of partially or completely occlusive platelet thrombi. A variety of methods have been used to assess platelet aggregation, blood coagulation, and the ex vivo and/or in vitro efficacy of platelet antagonists, anticoagulants, and thrombolytics. Key words: Platelets, aggregation, coagulation, thrombosis, in vitro models, adhesion.

1. Introduction The specific platelet surface receptors that support the initial adhesive interactions that ultimately lead to the formation of thrombi are determined by the local fluid dynamics of the vasculature and the extracellular matrix constituents exposed at the sites of vascular injury. Under high shear conditions, the adhesion of un-activated platelets to exposed sub-endothelial surfaces of atherosclerotic or injured vessels is mediated by binding of platelet glycoprotein (GP) Ib/IX/V complex to collagen and von Willebrand factor (vWF) presented on exposed vessel surfaces (1, 2). This primary adhesion to the matrix activates platelets, ultimately resulting in platelet aggregation, which is mediated predominantly by the binding of adhesive proteins, such as fibrinogen S.A. Mousa (ed.), Anticoagulants, Antiplatelets, and Thrombolytics, Methods in Molecular Biology 663, DOI 10.1007/978-1-60761-803-4_1, © Springer Science+Business Media, LLC 2003, 2010

1

2

Mousa

and vWF, to GPIIb/IIIa. Direct platelet aggregation in the bulk phase under conditions of abnormally elevated fluid shear stress, analogous to what occurs in atherosclerotic or constricted arterial vessels, may also be important (3). Shear-induced platelet aggregation is dependent upon the availability of vWF and the presence of GPIb/IX and GPIIb/IIIa on the platelet membrane. It has been postulated that under high shear stress conditions, the interaction of vWF with the GPIb/IX complex is the initial event leading to platelet activation and also triggers the binding of vWF to GPIIb/IIIa to induce platelet aggregate formation. The coagulation cascade consists of a complex network of interactions resulting in thrombin-mediated conversion of fibrinogen to fibrin, a major component of the thrombus. Initiation of the coagulation cascade occurs through the release of thromboplastin (tissue factor) and subsequent activation and conversion of factor VII into tissue factor/factor VIIa complex (“exogenous pathway”), or by the so-called contact activation pathway or “endogenous pathway,” which proceeds through factors XII, XI, and IX to the assembly of a tenase complex, consisting of activated factors VIII and IX and Ca2+ , on phospholipid membranes. Both the exogenous and endogenous tenase complex can activate factor X, which induces the formation of the prothrombinase complex, consisting of factor Xa, factor Va, and Ca2+ , on phospholipid surfaces. Assembly of the prothrombinase complex leads to the activation of thrombin, which cleaves fibrinogen to yield fibrin.

2. In Vitro Coagulation Tests 2.1. Blood Coagulation Tests 2.1.1. Purpose and Rational

Three coagulation tests, prothrombin time (PT), activated partial thromboplastin time (aPTT), and thrombin time (TT), can differentiate between exogenous and endogenous pathway effects and distinguish them from effects on fibrin formation. Typically, the influence of compounds on plasmatic blood coagulation is determined by measuring PT, aPTT, and TT ex vivo.

2.1.2. Procedure

Male Sprague-Dawley rats weighing 200–220 g are administered test compound, or vehicle as a control, through an oral, intraperitoneal, or intravenous route. After a period of time for absorption (adsorption time), animals are anesthetized by intravenous injection of sodium pentobarbital (60 mg/kg). The caudal caval vein is exposed by midline incision and 1.8 ml of blood is collected into

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

3

a plastic syringe containing 0.2 ml of 100 mM citrate buffer, pH 4.5 (Behring Werke, Marburg, DE). The sample is immediately agitated and then subjected to centrifugation in a plastic tube at 1,500×g for 10 min, after which plasma is collected and transferred to a clean plastic tube. Coagulation tests for TT, PT, and aPTT should be performed within 3 hours (h). In general, citrated plasma coagulates upon the addition of the appropriate compound (see below), and time to clot formation (coagulation time) is determined using a coagulometer (Schnittger and Gross, Amelung, Brake, DE). For the detailed laboratory diagnosis of bleeding disorders and assessment of blood coagulation, see Palmer (4) and Nilsson (5). PT. PT measures effects on the exogenous pathway of coagulation. Citrated plasma (0.1 ml) is incubated for 1 min at 37◦ C, R at which point 0.2 ml of human thromboplastin (Thromborel ; Behringwerke) is added. The coagulometer is started, and time to clot formation is recorded. aPTT. aPTT measures effects on the endogenous pathway of coagulation. Citrated plasma (0.1 ml) is mixed with 0.1 ml of R human placenta lipid extract (Pathrombin ; Behringwerke), and ◦ the mixture is incubated for 2 min at 37 C. Coagulation is initiated by the addition of 0.1 ml of 25 mM calcium chloride, at which point the coagulometer is started and time to clot formation is recorded. TT. TT measures effects on fibrin formation. Citrated plasma (0.1 ml) is mixed with 0.1 ml of diethyl barbiturate–citrate buffer, pH 7.6 (Behringwerke), and the mixture is incubated for 1 min at 37◦ C. Bovine test thrombin (0.1 ml) (30 IU/ml; Behringwerke) is added, at which point the coagulometer is started, and the time to clot formation is recorded. 2.2. Thrombelastography 2.2.1. Purpose and Rational

Thrombelastography (TEG) was developed by Hartert in 1948 (6). The thrombelastograph is a device that provides a continuous recording of the process of blood coagulation and subsequent clot retraction. Blood samples are transferred to cuvettes and maintained at 37◦ C. The cuvettes are set in motion around their vertical axes. Initially, a mirror suspended by a torsion wire in the plasma remains immobile as long as the plasma is fluid. There is a dynamic interplay between the cuvette and the mirror as fibrin forms, resulting in the transmission of motion within the cuvette to the mirror. The mirror will oscillate, the amplitude of which is governed by the specific mechanical properties of the clot, and reflect light onto a thermo-paper recording. Modern thrombelastographs translate the light recording into a digital signal that can be readily analyzed using a computer program.

4

Mousa

2.2.2. Procedure

TEG can be performed with whole blood or citrated plateletrich or platelet-poor plasma after re-calcification. Blood samples are obtained from test animals (i.e. Beagle dogs, 12–20 kg; rabbits, 1.7–2.5 kg; Wistar rats, 150–300 g; or humans). Test subjects receive the compound of interest by intravenous (iv), subcutaneous (sc) or oral administration. Ten- or twenty-min post dosing (iv or sc administration), or 60-, 90- or 180-min post dosing (oral administration), blood is collected. The blood samples are mixed with 3.8% trisodium citrate solution (1:9 citrate solution:blood) as an anticoagulant. Citrated whole blood is recalcified by the addition of 0.4 ml of isotonic calcium chloride. An aliquot (0.36 ml) of re-calcified whole blood is transferred to a pre-warmed cup of the thrombelastograph. After the apparatus has been correctly adjusted and the samples have been sealed with liquid paraffin to prevent drying, the start time for the procedure is noted, and the thrombelastogram is recorded for 2 h.

2.2.3. Evaluation

The following measurements are the standard variables of TEG: Reaction time (r). The time from sample placement in the cup until onset of clotting (defined as an amplitude of 1 mm). This represents the rate of initial fibrin formation. Clot formation time (k). The difference from the 1 mm r to 20 mm amplitude. k represents the time for a fixed degree of viscoelasticity to be achieved by the clot formation due to fibrin buildup and cross linking. Alpha angle (α ◦ ). Angle formed by the slope of the TEG tracing from the r to k value. It denotes the speed at which solid clot forms. Maximum amplitude (MA). Greatest amplitude on the TEG trace. MA represents the absolute strength of the fibrin clot and is a direct function of the maximum dynamic strength of fibrin and platelets. Clot strength (G; in dynes per square centimeter). It is defined by G = (5,000MA)/(96–MA). In tissue factor-modified TEG (7), clot strength is clearly a function of platelet concentration. Lysis 30, Lysis 60 (Ly30, Ly60). Reduction of amplitude relative to MA at 30 and 60 min after the time of MA. These parameters represent the influence of clot retraction and fibrinolysis. Readers are referred to a number of studies in which TEG has been instrumental in advancing the field of antithrombotics and thrombolytics. Most recently, TEG has been used to analyze the effects of a variety of stimuli on platelet/fibrin clot dynamics (8). In other work, Bhargava et al. (9) used TEG to compare the anticoagulant effects of a new potent heparin preparation and then commercially available heparin in vitro using citrated dog and human blood. Barabas et al. (10) used the fibrin plate

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

5

assay and TEG to assess the anti-fibrinolytic effects of synthetic thrombin inhibitors. Scherer et al. (11) described a short-time, endotoxin-induced rabbit model of hyper-coagulability for the study of the coagulation cascade, enabling the analysis of coagulation inhibitors and their therapeutic effects by a number of techniques, including TEG. In 1997, Khurana et al. (7) introduced tissue factor-modified TEG to study platelet GPIIb/IIIa function, establishing a quantitative assay of platelet function. Using this modification to TEG, Mousa et al. (8) identified two classes of GPIIb/IIIa antagonists, one with high binding affinity for resting and activated platelets and a slow platelet dissociation rate (class I) that exhibited potent inhibition of platelet function, and one with a fast platelet dissociation rate (class II). TEG was also used in a phase-II clinical trial to assess the efficacy of an oral platelet GPIIb/IIIa antagonist on platelet/fibrin clot dynamics (12). Zuckerman et al. (13) compared TEG with other common coagulation tests (fibrinogen, PT, aPTT, platelet count, and fibrin split products) and found a strong correlation between thrombelastographic variables and common laboratory tests. Moreover, TEG had a higher sensitivity for blood clotting anomalies. TEG also provides additional information on the hemostatic process. In contrast to most laboratory assays, in which the end point is the formation of the first fibrin strands, TEG measures the coagulation process from the initiation of clotting to the final stages of clot lysis and retraction. Another advantage of TEG is that it allows the use of whole non-anticoagulated blood without the influence of citrate or other anticoagulants. 2.3. Chandler Loop 2.3.1. Purpose and Rational

The Chandler loop technique measures the generation of in vitro thrombi in a moving column of blood (14). Thrombi generated in the Chandler device are morphologically similar to human thrombi formed in vivo (15), with platelet-rich upstream sections (“white heads”) that are relatively resistant to tissue plasminogen activator(t-PA)-mediated thrombolysis as compared to red blood cell-rich downstream components (“red tails”) (16).

2.3.2. Procedure

One millimeter of non-anticoagulated whole blood is drawn directly into a polyvinyl tube 25 cm in length with an internal diameter of 0.375 cm (1 mm=9.9-cm tubing). The two ends of the tube are then brought together and closed using an outside plastic collar. The circular tube is placed at the center of a turntable, tilted to an angle of 23º, and then rotated at 17 rpm. When the developing thrombus inside the tube becomes large enough to occlude the lumen, the blood column becomes static and begins to move on the table in the direction of rotation.

6

Mousa

Stringer et al. (16) used this method to determine the influence of an anti-plasminogen activator inhibitor (PAI)-1 antibody (CLB-2C8) on t-PA-induced lysis of Chandler thrombi in vitro. Investigators used citrated blood supplemented with 5.8 μM [125 I]-labeled fibrinogen prior to re-calcification. Thrombi generated in the Chandler loop were washed with isotonic saline and then cut transversally into an upstream (head) and a downstream (tail) part, and then each part was analyzed using a gamma counter to determine pre-values. The head and tail were then subjected to thrombolysis by the addition of 300 μl of phosphatebuffered saline (PBS) containing plasminogen (2 μM) and t-PA (0.9 nM). Over a period of 240 min, aliquots (10 μl) from each section were removed at 30, 60, 120, 180, and 240 min, and radioactivity was determined. Radioactivity at each time point as compared to the pre-value was then expressed as a percentage of clot lysis. In 1998, Van Giezen et al. (17) used this method to successfully differentiate the effects of an anti-PAI-1 polyclonal antibody (PRAP-1) on human and rat thrombi. 2.4. Platelet Aggregation and De-aggregation in Platelet-Rich Plasma or Washed Platelets (Born Method) 2.4.1. Purpose and Rational

Contact between non-activated platelets and exposed subendothelial tissue leads to adhesion through two main mechanisms: (1) at high shear rates, binding of sub-endothelial vWF to platelet GPIb–IX–V complex and (2) binding of collagen to integrin α2β1 and GPVI. Platelet adhesion initiates several processes, including shape changes, secretion, and the activation of GPIIb/IIIa ligand binding sites, resulting in the formation of platelet aggregates. Activation of GPIIb/IIIa is also achieved through receptor cross-signaling initiated by the binding of a number of agonists to G-protein-coupled receptors. To measure platelet aggregation, one of the following agonists is added to platelet-rich plasma (PRP) or washed platelets (WP): ADP, arachidonic acid (which is converted to thromboxane A2) or the thromboxane agonist U46619, collagen, thrombin or thrombin receptor-activating peptide (TRAP), serotonin, epinephrine, or platelet activating factor (PAF). Upon stirring, the formation of platelet aggregates is monitored photometrically as changes in optical density, typically for 4 min. This test was developed by Born (18, 19) and is used to quantitatively evaluate the effect of compounds on the induction of platelet aggregation in vitro or ex vivo. For in vitro studies, human PRP is the preferred starting material.

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

7

2.4.2. Procedure

For ex vivo assays, mice, rats, or guinea pigs of either sex receive test compound, or vehicle as a control, by oral, intraperitoneal (ip), or iv administration. At the end of absorption time, blood is collected by caval venipuncture under pentobarbital sodium anesthesia with xylazine (8 mg/kg intramuscular) premedication. From rabbits (Chinchilla, 3 kg in weight), blood is drawn by cardiopuncture under xylazine (20 mg/kg intramuscular) sedation. A control blood sample is collected before administration of the test compound, and the second sample is drawn at the end of absorption time. For in vitro assays of human blood, samples are collected from the vein of adult volunteers who have not received any medication during the 2 weeks prior to collection.

2.4.2.1. Preparation of PRP, Platelet-poor Plasma (PPP), and WP

The entire procedure is performed in plastic (polystyrene) tubes and carried out at room temperature. Freshly collected venous blood is anticoagulated with hirudin (1:9, hirudin:blood) or anticoagulant citrate dextrose (ACD) solution (1:9, ACD:blood) and then subjected to centrifugation at 170×g for 15 min to obtain PRP. The PRP supernatant is carefully removed, and the sample is subjected to centrifugation again at 1,500×g for 10 min to obtain PPP. PRP is diluted with PPP to a platelet count of 3×108 /ml before use in the aggregation assay. To obtain WP, 8.5 volumes of human blood are collected into 1.5 volumes of ACD and then subjected to centrifugation, as described for PRP. PRP is acidified to a pH of 6.5 by the addition of ACD (approximately 1 ml per 10 ml of PRP). Acidified PRP is subjected to centrifugation for 20 min at 430×g, and then the pellet is re-suspended to the original volume in Tyrode’s solution (120 mM NaCl, 2.6 mM KCl, 12 mM NaHCO3 , 0.39 mM NaH2 PO4 , 10 mM HEPES, 5.5 mM Glucose, and 0.35% albumin) and diluted to a platelet count of 3×108 /ml. The assay should be completed within 3 h of blood collection. For ex vivo assays, duplicate samples of PRP (320 μl) from drug-treated and vehicle control subjects (for rabbits, the control sample is taken before drug administration) are inserted into the aggregometer at 37◦ C under continuous magnetic stirring at 1,000 rpm. After the addition of 40 μl of physiological saline and 40 μl of aggregating agent, changes in optical density are monitored continuously at 697 nm. For in vitro assays, 40 μl of the test solution is added to 320 μl of PRP or WP from untreated subjects. The samples are inserted into the aggregometer and incubated at 37◦ C for 2 min under continuous magnetic stirring at 1,000 rpm. After the addition of 40 μl of aggregating agent, changes in optical density are monitored continuously at 697 nm for 4 min, or until aggregation values are constant. In cases of thrombin activation of PRP, glycine–proline–aspartate–proline (GPRP) peptide is added

8

Mousa

in order to prevent fibrin formation. To measure de-aggregation, experimental compounds are added to stimulated PRP at 70% or 100% of control aggregation, and samples are monitored for 10 min. De-aggregation is measured as the decrease in light transmission over time (20). 2.4.3. Evaluation 2.4.3.1. For In Vitro Assays

1. Percent inhibition of platelet aggregation is determined in each concentration group relative to the respective vehicle control. 2. IC50 values are determined from non-linear fitting of the concentration–effect relationship curve. IC50 is defined as the concentration of test drug that achieves half-maximal inhibition of aggregation. 3. Percent de-aggregation is determined 10 min after the addition of compound; IC50 is calculated from the concentration–effect curve 4. Statistical significance is evaluated by means of the unpaired Student’s t-test.

2.4.3.2. For Ex Vivo Assays

1. The mean values for aggregation in each dosage group are compared to the respective vehicle control group (for rabbits, the control is before drug administration). 2. ED50 values are determined from the dose–response curves. ED50 is defined as the dose of drug that achieves halfmaximal inhibition of aggregation in animals. 3. Statistical significance is evaluated by means of the Student’s t-test (paired for rabbits; unpaired for others).

2.4.4. Critical Assessment of the Method

This assay, introduced by Born in 1926, has become a standard method in the clinical diagnosis of platelet function disorders and aspirin intake. Furthermore, the method is widely used in the discovery of antiplatelet drugs. The advantages of the method include the ability to rapidly measure a functional parameter in intact human platelets. However, processing of platelets during the preparation of PRP, WP, or filtered platelets from whole blood can result in platelet activation and separation of large platelets.

2.4.5. Modifications of the Method

Several authors have described modifications of the Born assay. Breddin et al. (21) described the use of a rotating cuvette to measure spontaneous aggregation of platelets from vascular patients. Klose et al. (22) measured platelet aggregation under laminar flow conditions using a thermo-regulated cone-plate streaming chamber in which shear rates were continuously augmented and platelet aggregation was measured based on light transmission through a transillumination system. Marguerie et al. (23, 24)

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

9

developed a method of measuring two phases of platelet aggregation after gel filtration of a platelet suspension (see below). Lumley and Humphrey (25) described a method to measure platelet aggregation in whole blood (see below). Fratantoni and Poindexter (26) used a microtiter plate reader with specific modifications of sample agitation to measure platelet aggregation. A comparison of the 96-well microtiter plate method and conventional aggregometry revealed similar dose–response curves for thrombin, ADP, and arachidonic acid. Ammit and O’Neil (27) used a quantitative bioassay of platelet aggregation for rapid and selective measurement of PAF. 2.5. Platelet Aggregation After Gel Filtration 2.5.1. Purpose and Rational

Triggering of platelet activation by low concentrations of ADP, epinephrine, or serotonin, the so-called weak platelet agonists, in plasma- and fibrinogen-free platelet suspensions does not result in platelet aggregation unless exogenous fibrinogen is added. In contrast, platelet aggregation induced by thrombin, collagen, or prostaglandin endoperoxide, so-called strong agonists, is independent of exogenous fibrinogen because these substances induce the secretion of intracellular platelet ADP and fibrinogen. Analysis of platelet aggregation in gel-filtered platelet samples is carried out in cases when fibrinogen or vWF is needed in a defined concentration, or when plasma proteins could negatively interfere with the effects of compounds. The assay is used primarily to evaluate the influence of compounds on platelet integrin GPIIb/IIIa or other integrins or on platelet GPIb–IX–V.

2.5.2. Procedure 2.5.2.1. Preparation of Gel-Filtered Platelets

The entire procedure is performed in plastic (polystyrene) tubes at room temperature (24). Blood is drawn from healthy adult volunteers who have received no medication in the 2 weeks prior to collection. Venous blood (8.4 ml) is collected into 1.4 ml of ACD solution and subjected to centrifugation for 10 min at 120×g. PRP is carefully removed, the pH is adjusted to 6.5 with ACD solution, and the sample is subjected to centrifugation again at 285×g for 20 min. The resulting pellet is re-suspended in Tyrode’s buffer (approx. 500 μl of buffer/10 ml of PRP), and the platelet suspension is applied immediately to a Sepharose CL 2B column. Equilibration and elution (flow rate, 2 ml/min) are done with Tyrode’s buffer without hirudin and apyrase. Platelets are recovered in the void volume. The platelet suspension is adjusted to a final cell concentration of 4×108 /ml. Gel-filtered platelets (GFP) are kept at room temperature for 1 h before the assay is started.

10

Mousa

2.5.2.2. Experimental Course

For aggregation studies, GFP in Tyrode’s buffer are incubated with CaCl2 (final concentration, 0.5 mM) with or without fibrinogen (final concentration, 1 mg/ml) in polystyrene tubes. After 1 min, 20 μl of the test compound, or vehicle as a control, are added and the sample is incubated for an additional 2 min. Platelet agonist is added (20 μl), and changes in light transmission are recorded. The entire procedure is done under continuous magnetic stirring at 37◦ C (1,000 rpm) in the aggregometer. Samples with CaCl2 but without fibrinogen confirm that plasma proteins have been properly filtered out if neither spontaneous aggregation nor aggregation in the presence of weak agonists occurs; full aggregation of GFP in response to 10 μM ADP confirms that platelets are intact (i.e., minor pre-activation by gel filtration). Readers are referred to several references for a detailed methodology and evaluation of different agents by gel filtration platelet aggregation (23, 24, 28).

2.6. Platelet Aggregation in Whole Blood 2.6.1. Purpose and Rational

This method uses a whole blood platelet counter, which counts single platelets and does not require separation of platelets from other blood cell types. Platelet aggregation is induced in anticoagulated human whole blood samples by addition of the aggregating agents arachidonic acid or collagen. The number of platelets is determined in drug-treated and vehicle control samples and the percent inhibition of aggregation and IC50 values are calculated for each dosage group. This test system enables assessment of the effect of compounds on other blood cells, which can indirectly influence platelet aggregation. The method was originally described by Lumley and Humphrey (25) and Cardinal and Flower (29).

2.6.2. Procedure

The entire procedure is performed in plastic (polystyrene) tubes and is carried out at room temperature. Blood is drawn from healthy adult volunteers who have not received medication during the 2 weeks prior to collection. Venous blood (9 ml) is anticoagulated with 1 ml of sodium citrate and maintained in a closed tube at room temperature for 30–60 min until the start of the test. For aggregation analysis, 10 μl of compound, or vehicle as a control, are added to 480 μl of citrated blood. Samples in closed tubes are pre-incubated for 5 min in a 37◦ C water shaker bath with shaking (75 strokes/min). Aggregating agent (10 μl) is added and samples are incubated for another 10 min. The number of platelets (platelet count) is determined in aliquots of 10 μl immediately before and 10 min after the addition of aggregating

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

11

agent (initial platelet count and 10-min platelet count, respectively) using a hemocytometer. The following samples are prepared in duplicate: 1. Control aggregation/spontaneous aggregation: 480 ml blood + 20 ml vehicle. Samples with >20% spontaneous aggregation should not be used to test for induced aggregation. 2. Maximum aggregation: 480 ml blood + 10 ml vehicle + 10 ml aggregating agent. Represents the maximum induced aggregation rate. 3. Test substance aggregation: 480 ml blood + 10 ml test substance + 10 ml aggregating agent. 2.7. Platelet Microand Macro-Aggregation Using Laser 2.7.1. Purpose and Rational

A new platelet aggregometer (AG-10; Kowa, Japan) that uses a laser light-scattering beam was introduced in 1996 by Tohgi et al. (30). This technique is a highly sensitive method of studying platelet aggregation based on the measurement of mean radius or particle size, making it possible to record the kinetics of formation of micro- and macro-aggregates in real time. Sensitivity in measurement of spontaneous aggregation is higher than in routine light transmittance. Platelet aggregate size, derived from the total voltage of lightscatter intensity at 1.0-second (s) intervals over a 10-min period, can be divided into three ranges: small (diameter 9–25 μm), medium (26–50 μm), and large (>50 μm) aggregates. Using laser scatter aggregation, it was found that young smokers had an increased number of small platelet aggregates, which were undetectable using a conventional aggregometer and the turbidometric method (31). The light-scatter aggregometer can detect platelet aggregation in the small-aggregate size range after the addition of unfractionated heparin (UFH), and the aggregates are dissociated upon incubation with protamine sulfate. Analysis of platelet aggregation induced by 0.5 U/ml of UFH in 36 normal subjects with no history of heparin exposure revealed 13 subjects with a positive response in excess of 0.5 V of light intensity in the small-aggregate size range. In chronic hemodialysis patients who had used heparin regularly for many years, a positive response, that is, the detection of heparin-induced aggregates, was observed in 37 of 59 patients, an increase over normal subjects. Light intensity in the small-aggregate size range was also enhanced during heparinized dialysis. In patients with a positive heparin response, the intensity of aggregate formation after heparin was significantly higher than heparin non-responders. Using the same system, it

12

Mousa

has been shown that enhanced platelet aggregation response to heparin is not inhibited by aspirin or argatroban, but can be inhibited by anti-GPIIb/IIIa antibodies. Enhanced platelet aggregation during heparin infusion was observed without the addition of ADP or TRAP using laser aggregometry (32). Limitations. This technique cannot be applied to whole blood, but can be used with PRP, WP, or GFP. 2.8. Fibrinogen Receptor Binding 2.8.1. Purpose and Rational

This assay is used to evaluate the binding characteristics of drugs that target the fibrinogen receptor. A single concentration of radioligand (125 I-fibrinogen, 30–50 nM) is incubated with human GFP in the presence of increasing concentrations of a non-labeled test drug (0.1 nM–1 mM). If the test drug binds to the fibrinogen receptor, it will compete with the radiolabeled ligand for receptor binding sites. Generally, the higher the affinity of the test drug for the receptor, the lower the concentration required to compete for binding, and the more potent the test drug will be. Platelets are activated by 10 mM ADP to stimulate 125 I-fibrinogen binding to the GPIIb/IIIa receptor.

2.8.2. Procedure 2.8.2.1. Preparation of GFP

Blood is collected from a healthy volunteer (200 ml). An aliquot (8.4 ml) is mixed with 1.4 ml of ACD buffer in a polystyrol tube and then subjected to centrifugation at 1,000 rpm for 15 min. The resulting PRP is collected and an aliquot is removed for a platelet count. Ten ml of PRP are mixed with 1 ml of ACD buffer to yield ACD–PRP, pH ∼6.5, and then 5 ml portions of ACD– PRP are transferred to plastic tubes and subjected to centrifugation at 1,600 rpm for 20 min. The resulting supernatants are decanted, and each pellet is re-suspended in 500 μl of Tyrode buffer C. An aliquot is removed for a platelet count to determine platelet loss, and then the platelet suspension is transferred to a Sepharose column that has been pre-equilibrated with approximately 100 ml of degassed Tyrode buffer B (flow rate, 2 ml/min). The column is closed, and sample is eluted with degassed Tyrode buffer B (flow rate, 2 ml/min). The first platelets appear in the 18–20 min fractions and are collected thereafter for 10 min in a closed plastic cup. GFP are reconstituted at a density of 4×108 platelets/ml with Tyrode buffer B and maintained at room temperature until use.

2.8.2.2. Experimental Course

Each concentration of drug is tested in triplicate using No. 72708 Sarstedt tubes. The total volume of each test sample is 500 μl. The concentration of 125 I-fibrinogen is constant for all samples (10 μg/500 μl).

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

2.8.2.3. Competition Experiments

13

Competition reactions are carried out using a negative control (i.e., double distilled water), non-labeled fibrinogen, and increasing concentrations of test compound, as follows: • 100 μl 125 I-fibrinogen • 100 μl non-labeled fibrinogen or test drug (increasing concentrations of 10−10 to 10−3 M) • 50 μl ADP

2.8.2.4. Non-Specific Binding

Non-specific binding of 125 I-fibrinogen is defined as the level of binding of radioligand in the presence of 10−5 M non-labeled fibrinogen. The binding reaction is initiated by the addition of 250 μl of GFP (4×108 platelets/ml). The samples are incubated for 30 min at room temperature, and then 100 μl is transferred to a microcentrifuge tube containing 400 μl of a glucose solution. The tubes are centrifuged at 11,750 rpm for 2 min to separate bound 125 I-fibrinogen from free radioligand. The supernatant is carefully decanted and allowed to evaporate for approximately 30 min. The radioactivity of the platelet pellet is measured for 1 min using a gamma counter (efficiency in the range of 65.3%). Readers are referred to several excellent references for a detailed methodology and evaluation of various mechanisms and agents assessed using the fibrinogen binding assay (23, 24, 33, 34).

2.9. Euglobulin Clot Lysis Time 2.9.1. Purpose and Rational

Euglobulin lysis time is used as an indicator of the influence of compounds on fibrinolytic activity in rat blood and is based on the procedure of Gallimore et al. (35). The euglobulin fraction of plasma is separated from inhibitors of fibrinolysis by acid precipitation and centrifugation. Euglobulin predominantly consists of plasmin, plasminogen, plasminogen activator, and fibrinogen. Addition of thrombin to this fraction results in the formation of fibrin clots. The lysis time of these clots is related to the activity of activators of fibrinolysis (e.g., plasminogen activators). Thus, compounds that stimulate the release of t-PA from the vessel wall can be detected with this assay.

2.9.2. Procedure

Rats are anesthetized by ip injection of pentobarbital sodium (60 mg/kg) and placed on a heating pad (37◦ C). At the same time, the test solution, or vehicle as a control, is administered iv or ip. Twenty-five minutes later, the animals receive another ip injection of sodium pentobarbital (12 mg/kg) to maintain them in deep narcosis for 45 min.

2.9.2.1. Plasma Preparation

After the test compound is absorbed, the inferior caval vein is exposed by a midline excision and blood (1.8 ml) is drawn using a plastic syringe containing 0.2 ml of a 3.8% sodium citrate

14

Mousa

solution. The sample is thoroughly mixed, transferred to a plastic tube, and immediately immersed in ice. Plasma is obtained by centrifugation at 2,000×g for 10 min at 2◦ C. 2.9.2.2. Euglobulin Preparation

A 0.5 ml aliquot of plasma is added to 9.5 ml of ice-cold distilled water, and then the pH is brought to 5.3 by the addition of 0.13 ml of 1% acetic acid. The diluted plasma is kept on ice for 10 min and then the precipitated euglobulin fraction is collected by centrifugation at 2,000×g for 10 min at 2◦ C. The supernatant is discarded and the remaining fluid is removed by wicking away excess moisture with filter paper for 1 min. The euglobulin precipitate is dissolved in 1 ml of 0.12 M sodium acetate.

2.9.2.3. Euglobulin Lysis Assay

Aliquots (0.45 ml) of the euglobulin solution are transferred to test tubes, and 0.05 ml of thrombin (Test Thrombin, Behringwerke) (25 U/ml) is added. The tubes are transferred to a water bath at 37◦ C. The time interval between the addition of thrombin and complete lysis of the clots is recorded.

2.10. Flow Behavior of Erythrocytes 2.10.1. Purpose and Rational

The deformation of erythrocytes is an important rheological phenomenon in blood circulation (36). It allows the passage of normal red cells through capillaries with diameters smaller than that of the discoid cells and reduces the bulk viscosity of blood flowing in large vessels. In this test, the initial flow of filtrate is used as a measure of erythrocyte deformability. Prolonged filtration time can be attributed to two fundamental pathological phenomena: an increased rigidity of individual red cells and an increased tendency of the cells to aggregate. To simulate decreased red blood cell deformability, the erythrocytes are artificially modified by one (or a combination) of the following stress factors: • addition of calcium ions (increase in erythrocyte rigidity) • addition of lactic acid (decrease in pH value) • addition of 350–400 mmol NaCl (hyper-osmolarity) • storing the sample for at least 4 h (cellular ageing, depletion of ADP) This test is typically used to evaluate the effect of test compounds on the flow behavior of erythrocytes.

2.10.2. Procedure

Apparatus. Erythrocyte filtrometer model MF 4 (Fa. Myrenne, 52159 Roetgen, Germany) equipped with a membrane filter (Nuclepore Corp., Pleasanton, CA, USA) (pore diameter, 5– 10 μm; pore density, 4×105 pores/cm2 ). Ex vivo. Blood is collected from Beagle dogs (12–20 kg), rabbits (1.2–2.5 kg) or Wistar rats (150–300 g). Animals receive the test compound by oral, sc or iv administration 15, 60, 90, or 180 min before the withdrawal of blood.

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

15

In vitro. Following the addition of test compound, blood is incubated at 37◦ C for 5 or 30 min. Freshly collected venous blood is anticoagulated with K-EDTA (1 mg/ml) or heparin (5 IU/ml heparin sodium) and then subjected to centrifugation at 3,000 rpm for 7 min. The supernatant (plasma) and buffy coat are removed and discarded. The packed erythrocytes are resuspended in autologous plasma containing 0.25% human albumin, and the haematocrit is adjusted to 10%. Red blood cells are then subjected to one of a combination of stress factors mentioned above. A portion (2 ml) of the stressed suspension is applied to the filtrometer and initial flow rate is determined. The filtration curve is plotted automatically. 2.11. Filterability of Erythrocytes 2.11.1. Purpose and Rational

The single erythrocyte rigidometer (SER) measures the deformability of individual red blood cells by measuring passage time through a pore under constant shear stress. In this assay, the passage times of single erythrocytes through one pore in a synthetic membrane are determined (37–39). The pore in the membrane essentially represents a capillary with a defined diameter and length. The driving pressure is produced by constant shear stress. The passage of red blood cells is measured based on changes in electrical current. For example, a constant current of 50–200 nA is applied, and passage of an erythrocyte through the pore is recorded as an interruption in current. This test is used to detect compounds that improve the filterability, and thus deformability, of erythrocytes. To simulate decreased red blood cell deformability, the erythrocytes are artificially modified by one or a combination of the following stress factors: • calcium ions (increase erythrocyte rigidity) • lactic acid (decreases pH value) • 350–400 mmol NaCl (generates a state of hyper-osmolarity) • storage for at least 4 h (cellular ageing and depletion of ADP)

2.11.2. Procedure

Apparatus. Single erythrocyte rigidometer (Myrenne, 52159 Roetgen, Germany). Ex vivo. Blood is collected from Beagle dogs (12–20 kg), rabbits (1.2–2.5 kg), or Wistar rats (150–300 g), or from human. The subject receives the test compound by oral, sc, or iv administration 15, 60, 90, or 180 min before the withdrawal of blood. In vitro. Following the addition of the test compound, blood samples are incubated at 37ºC for 5 or 30 min. Blood samples are mixed with K-EDTA (1 mg/ml blood) or heparin (5 IE/ml heparin sodium) to prevent clotting and then subjected to centrifugation at 3,000 rpm for 7 min. The plasma

16

Mousa

and buffy coat are removed and discarded. The packed erythrocytes are re-suspended in filtered HEPES buffer containing 0.25% human albumin and the haematocrit value is adjusted to 100 ms) cause rheological occlusion. Untreated red blood cells serve as the control for this assay. 2.12. Erythrocyte Aggregation 2.12.1. Purpose and Rational

The aggregation of red blood cells into rouleaux and the transition of rouleaux into three-dimensional cell networks is a rheological parameter that decisively influences the flow behavior of blood, especially in disturbed microcirculation. In this test, an erythrocyte aggregometer is used to measure erythrocyte aggregation. The transparent measuring chamber in a cone-and-plate configuration is transilluminated by light of a defined wave length, and the intensity of the transmitted light, which is modified by the aggregation process, is recorded. The method has been used successfully to determine the effect of test compounds on erythrocyte aggregation (37, 40). Apparatus. SER (Fa. Myrenne, 52159 Roetgen, Germany). Ex vivo. Blood is collected from Beagle dogs (12–20 kg), rabbits (1.2–2.5 kg), or Wistar rats (150–300 g). The animals receive test compound by oral, sc, or iv administration 15, 60, 90, or 180 min before the withdrawal of blood. In vitro. Following the addition of test compound, the blood sample is incubated at 37◦ C for 5 or 30 min. Blood is obtained by venipuncture and mixed with K-EDTA (1 mg/ml) or heparin (5 IU/ml heparin sodium) to prevent clotting. Erythrocyte aggregation is determined in whole blood with a haematocrit of 40%. A portion (40 μl) of the blood is transferred to the measuring device and the red cells are dispersed at a shear rate of 600/s. After 20 s, flow is switched to stasis, and the extent of erythrocyte aggregation is determined photometrically.

3. In Vitro Models of Thrombosis A variety of methods have been used to assess the ex vivo and/or in vitro efficacy of platelet antagonists, including photometric aggregometry, whole blood electrical aggregometry, and particle counting, as described earlier. In photometric aggregometry, a sample is placed in a stirred cuvette in the optical light path

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

17

between a light source and a light detector. Aggregate formation is monitored by a decrease in turbidity, and the extent of aggregation is measured as percent maximal light transmission. The major disadvantage of this technique is that it cannot be applied to whole blood, since the presence of erythrocytes interferes with optical detection. Furthermore, it is insensitive to the formation of small aggregates. Particle counters are used to quantitate the size and the number of particles suspended in an electrolyte solution by monitoring the electrical current between two electrodes immersed in the solution. Aggregation in this system is quantitated by counting the number of platelets before and after stimulation and is usually expressed as a percentage of the initial count (41). The disadvantage of this technique is that it cannot distinguish platelets and platelet aggregates from other blood cells of the same size. Thus, one is limited to counting only a fraction of single platelets as well as aggregates that are much larger than erythrocytes and leukocytes. Electrical aggregometry allows the detection of platelet aggregates as they attach to electrodes immersed in a stirred cuvette containing whole blood or a platelet suspension. Attachment results in a decrease in conductance between the two electrodes, which can be quantitated in units of electrical resistance. A disadvantage of this method is that it is not sensitive enough to detect small aggregates (42). This section will discuss two complementary in vitro flow models of thrombosis that can be used to accurately quantify platelet aggregation in anticoagulated whole blood and evaluate the inhibitory effect of platelet antagonists: (1) a viscometric flow cytometric assay that measures direct shear-induced platelet aggregation in the bulk phase (43) and (2) a parallel-plate perfusion chamber assay coupled with a computerized videomicroscopy system that quantifies the adhesion and subsequent aggregation of human platelets in anticoagulated whole blood flowing over an immobilized substrate (i.e., collagen I) (43, 44). We also discuss a third in vitro flow assay described by Mousa et al. (44) in which surface-anchored platelets are pre-incubated with a GPIIb/IIIa antagonist, unbound drug is washed away, and then THP-1 monocytic cells are perfused into the system, enabling the characterization of agents with markedly distinct affinities and receptor-bound lifetimes. 3.1. Cone-and-Plate Viscometry Under Shear-Flow Cytometry 3.1.1. Purpose and Rational

The cone-and-plate viscometer is an in vitro flow model used to investigate the effects of bulk fluid shear stress on suspended cells. Anticoagulated whole blood (or isolated cell suspension) is placed

18

Mousa

between the two stainless steel plates of the viscometer. Rotation of the upper conical plate causes a well-defined and uniform shear stress to be applied to the entire fluid medium (1). The shear rate (γ ) in this system can be readily calculated from the cone angle and the speed of the cone using the following formula:   2πω γ = , 60θcp where γ is the shear rate per second, ω is the cone rotational rate in revolutions per minute (rev/min) and θ cp is the cone angle in radians. The cone angle is typically in the range of 0.3–1.0◦ . Shear stress, τ , is proportional to shear rate, γ , as shown by τ = μγ , where μ is the viscosity of the cell suspension (the viscosity of anticoagulated whole blood is approximately 0.04 cp at 37◦ C). Rotational viscometers are capable of generating shear stress in the range of ∼2 dyn/cm2 (venous level) to greater than 200 dyn/cm2 (stenotic arteries). 3.1.2. Procedure

Single platelets and platelet aggregates generated in blood upon shear exposure are differentiated from other blood cells by their characteristic forward-scatter and fluorescence profiles from flow cytometric analysis using fluorophore-conjugated platelet-specific antibodies (43). This technique requires no washing or centrifugation steps that could potentially induce artifacts due to platelet activation and enables the analysis of platelet function in the presence of other blood elements.

3.1.2.1. Isolation of Human Platelets

Venous blood is drawn by venipuncture into polypropylene syringes containing either sodium citrate (0.38% final concentration) or heparin (10 U/ml final concentration). Anticoagulated whole blood is subjected to centrifugation at 160×g for 15 min to obtain PRP.

3.1.2.2. Isolation of WP

PRP is subjected to centrifugation again at 1,100×g for 15 min in the presence of 2 μM PGE1. The platelet pellet is re-suspended in HEPES-Tyrode buffer containing 5 mM EGTA and 2 μM PGE1. Platelets are washed and collected by centrifugation (1,100×g for 10 min) and then re-suspended in HEPES-Tyrode buffer at a cell density of 2×108 cells/ml and maintained at room temperature for no longer than 4 h before use in aggregation/adhesion assays.

3.1.2.3. Experimental Course

The steps described in this section outline the procedure used to quantify platelet aggregation induced by shear stress in the bulk phase as well as the inhibitory effects of platelet antagonists. For a detailed description, see Konstantopolous et al. (43).

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

19

1. Incubate anticoagulated whole blood with platelet antagonist, or vehicle as a control, at 37◦ C for 10 min. 2. Place the sample (typically 500 μl) on the stationary plate of a cone-and-plate viscometer maintained at 37◦ C. 3. Remove a small aliquot (∼3 μl) from the pre-sheared blood sample, incubate in 1% formaldehyde in D-PBS (30 μl), and process as outlined below for sheared samples. 4. Expose the sample to well-defined shear levels (typically 4,000 s-1 in the absence of a platelet antagonist to induce significant platelet aggregation) in the presence or absence of a platelet antagonist for prescribed periods of time (typically 30–60 s). 5. Remove a small aliquot (∼3 μl) from the sheared sample and incubate immediately in 1% formaldehyde in D-PBS (30 μl). 6. Incubate the fixed blood samples (pre-sheared and sheared) with a saturating concentration of a fluorescent-labeled platelet-specific antibody, such as anti-GPIb (6D1)-FITC, for 30 min in the dark. 7. Dilute specimens with 2 ml of 1% formaldehyde, and analyze by flow cytometry 8. Flow cytometric analysis is used to distinguish platelets from other blood cells on the basis of their characteristic forwardscatter and fluorescence profiles (Fig. 1.1). Data acquisition is carried out on each sample for a defined period (usually 100 s), allowing equal volumes for the pre-sheared and sheared specimens to be achieved. Percent platelet aggregation is determined based on the disappearance of single platelets and increase in platelet aggregates using the following formula: % platelet aggregation = (1 − Ns /Nc × 100) where Ns represent the single platelet population of the sheared specimen and Nc represents the single platelet population of the pre-sheared specimen. By comparing the extent of platelet aggregation in the presence and absence of a platelet antagonist, antiplatelet effects can be readily determined. 3.2. Platelet Adhesion and Aggregation Under Dynamic Shear 3.2.1. Purpose and Rational

This section describes an in vitro flow model of platelet thrombus formation that can be used to evaluate the ex vivo and/or in vitro efficacy of platelet antagonists. Thrombus formation can be initiated by platelet adhesion from rapidly flowing blood

20

Mousa

Fig. 1.1. Quantification of shear-induced platelet aggregation by flow cytometry. Panel A corresponds to an un-sheared blood specimen. Panel B corresponds to a blood specimen that has been subjected to a pathologically high level of shear stress for 30 s. As can be seen in the Figure there are three distinct cell populations. The upper population consists of platelets and platelet aggregates. The “red blood cell–platelet” population corresponds to platelets associated with erythrocytes and leukocytes (see Evaluation, Comment 4). The white blood cell population consists of some leukocytes that have elevated levels of FITC autofluorescence. The left vertical line separates single platelets (≤ 4.5 μm in diameter) from platelet aggregates, whereas the right vertical line separates “small” from “large” platelet aggregates. The latter were defined to be larger than 10 μm in equivalent sphere diameter.

onto exposed sub-endothelial surfaces of injured vessels presenting collagen and vWF, resulting in platelet activation and aggregation. Konstantopolous et al. (43) described the use of a parallel-plate flow chamber that provides a controlled and welldefined flow environment (i.e., chamber geometry and flow rate through the chamber). Wall shear stress, τ w , assuming a Newtonian and incompressible fluid, can be calculated using the following formula: 6μQ τw = , wh 2 where Q is the volumetric flow rate, μ is the viscosity of the flowing fluid, h is the channel height, and w is the channel width.

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

21

A typical flow chamber consists of a transparent polycarbonate block, a gasket whose thickness determines the channel depth, and a glass coverslip coated with an extracellular matrix protein, such as type I fibrillar collagen. The apparatus is held together by vacuum. Shear stress is generated by flowing fluid (i.e., anticoagulated whole blood or isolated cell suspensions) through the chamber over the immobilized substrate under controlled kinematic conditions using a syringe pump. Recently, Mousa et al. (44) combined the parallel-plate flow chamber with a computerized epifluorescence videomicroscopy system, enabling separation and real-time visualization of adhesion and subsequent aggregation of human platelets in anticoagulated whole blood (or isolated platelet suspensions) flowing over an immobilized substrate. 3.2.2. Procedure 3.2.2.1. Preparation of Collagen-Coated Surfaces

1. Dissolve 500 mg of collagen type I from bovine achilles tendon into 200 ml of 0.5 M acetic acid, pH 2.8. 2. Homogenize for 3 h. 3. Centrifuge the homogenate at 200×g for 10 min, collect the supernatant, and then measure collagen concentration by modified Lowry analysis. 4. Coat glass coverslips with 200 μl of fibrillar collagen I suspension so that all but the first 10 mm of the slide is covered (coated area = 12.7 mm×23 mm) and then place coated coverslips in a humid environment at 37◦ C for 45 min. 5. Rinse the slides to remove excess collagen with 10 ml of warm (37◦ C) D-PBS and then assembly the flow chamber.

3.2.2.2. Platelet Perfusion

1. Add the fluorescent dye quinacrine dihydrochloride to anticoagulated whole blood samples at a final concentration of 10 μM immediately after blood collection. 2. Prior to the perfusion experiment, incubate blood with platelet antagonist, or vehicle as a control, at 37◦ C for 10 min. 3. Perfuse anticoagulated whole blood through the flow chamber for 1 min at wall shear rates of 100/s (typical of venous circulation) to 1,500/s (representative of partially constricted arteries) for prescribed periods of time (i.e., 1 min). Platelet–substrate interactions are monitored in real time using an inverted microscope equipped with an epifluorescence illumination apparatus and a silicon-intensified target video camera and recorded on videotape. The microscope stage and flow chamber are maintained at 37◦ C by a heating module and incubator enclosure during the experiment.

22

Mousa

4. Videotaped images are digitized and computer analyzed at 5, 15, and 60 s for each perfusion experiment. The number of adherent individual platelets in the microscopic field of view during the initial 15 s of flow is determined by image processing and used as the measurement of platelet adhesion that initiates platelet thrombus formation. The number of platelets in each individual thrombus is calculated as the total thrombus intensity (area×fluorescence intensity) divided by the average intensity of single platelets determined in the 5-s images. By comparing the extents of platelet aggregation in the presence and absence of a platelet antagonist, antiplatelet efficacy can be determined (Fig. 1.2). Along these lines, any potential inhibitory effects of a platelet antagonist on platelet adhesion can be readily assessed.

Fig. 1.2. Three-dimensional computer-generated representation of platelet adhesion and subsequent aggregation on collagen I/von Willebrand factor from normal heparinized blood perfused in the absence (control) or presence of a GPIIb/IIIa antagonist (XV454) at 37◦ C for 1 min at 1,500/s.

3.3. Cell Adhesion to Immobilized Platelets: Parallel-Plate Flow Chamber 3.3.1. Purpose and Rational

This section outlines an in vitro flow assay to distinguish agents with markedly distinct affinities and off-rates. In this assay, immobilized platelets are pretreated with a GPIIb/IIIa antagonist, and any unbound drug is washed away before the perfusion of monocytic THP-1 cells. Using this technique, Albulencia et al. (45) demonstrated that agents with slow platelet off-rates, such as XV454 (t1/2 of dissociation = 110 min; Kd = 1 nM) and abciximab (t1/2 of dissociation = 40 min; Kd = 9.0 nM), which are present predominantly as receptor-bound entities in plasma with little unbound agent, can effectively block platelet heterotypic interactions. In contrast, agents with relatively fast

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

23

platelet dissociation rates, such as orbofiban (t1/2 of dissociation = 0.2 min; Kd >110 nM), the antiplatelet efficacy of which depends on the plasma concentration of the active drug, do not exhibit such inhibitory effects (44). 3.3.2. Procedure 3.3.2.1. Preparation of 3-Aminopropyltriethoxysilane (APES)-Treated Glass Slides

1. Soak glass coverslips overnight in 70% nitric acid. 2. Wash coverslips with tap water for 4 h. 3. Dry coverslips by washing once with acetone, and then immerse in a 4% solution of APES in acetone for 2 min. 4. Repeat step 3, followed by a final rinse of the glass coverslips with acetone. 5. Wash coverslips 3 times with water and dry overnight.

3.3.2.2. Immobilization of Platelets on APES-Treated Glass Slides

1. Layer WP or PRP (2×108 cells/ml) on the surface of an APES-coated coverslip at a density of approximately 30 μl/cm2 .

3.3.2.3. Monocytic THP-1 Cell–Platelet Adhesion Assay

1. Assemble the platelet-coated coverslip in a parallel-plate flow chamber. Mount the chamber on the stage of an inverted microscope equipped with a CCD camera connected to a VCR and TV monitor.

2. Allow platelets to bind to the coverslip in a humid environment at 37◦ C for 30 min.

2. Perfuse antiplatelet antagonist at the desired concentration, or vehicle as a control, over surface-bound platelets and incubate for 10 min. The extent of platelet activation can be further modulated by chemical agonists such as thrombin (0.02–2 U/ml) during the 10-min incubation period. The microscope stage and flow chamber are maintained at 37◦ C with a heating module and incubator enclosure during the experiment. 3. In some experiments, unbound platelet antagonist is removed by a brief washing step (4 min) prior to the perfusion of the cells of interest over the platelet layer. Alternatively, platelet antagonist at the desired concentration is continuously maintained in the perfusion buffer during the entire course of the experiment. 4. Perfuse cells (i.e., THP-1 monocytic cells, leukocytes, tumor cells, protein-coated beads) over surface-bound platelets in the presence or absence of platelet antagonist (see above) at the desired flow rate for prescribed periods of time. Cell binding to immobilized platelets is monitored in real time and recorded on videotape. 5. Determine the extent of cell tethering, rolling, and stationary adhesion to immobilized platelets, as well as the average

24

Mousa

velocity of rolling cells. Comparison of the extent of cell tethering, rolling, and stationary adhesion to immobilized platelets in the presence and absence of platelet antagonist (Fig. 1.3) will provide a measure of antiplatelet activity.

Fig. 1.3. Phase-contrast photomicrograph of THP-1 cells (phase bright objects) attached to a layer of thrombin-treated platelets (phase dark objects) after THP-1 cell perfusion for 3 min at a shear stress level of 1.5 dyn/cm2 .

3.3.3. Evaluation

1. Low-speed centrifugation of whole blood results in the separation of platelets (top layer) from larger and denser cells such as leukocytes and erythrocytes (bottom layer). To minimize leukocyte contamination of PRP, slowly aspirate the uppermost 2/3 of the platelet layer after centrifugation. Also, certain rare platelet disorders such as Bernard–Soulier syndrome (BSS) are characterized by larger than normal platelets which must be isolated by allowing whole blood to separate by gravity for 2-h post-venipuncture. 2. The mechanical force most relevant to platelet-mediated thrombosis is shear stress. Normal time-averaged levels of venous and arterial shear stress range from 1–5 dyn/cm2 to 6–40 dyn/cm2 , respectively. However, fluid shear stress may reach levels well over 200 dyn/cm2 in small arteries and arterioles that are partially obstructed by atherosclerotic lesions or vascular spasm. The cone-and-plate viscometer and parallel-plate flow chamber are two of the most common devices used to simulate fluid mechanical shear stress in blood vessels. 3. Due to the high concentration of platelets and erythrocytes in whole blood, small aliquots (3 μl) of pre-sheared and post-sheared samples must be obtained and processed prior to the flow cytometric analysis. This will minimize artifacts

In Vitro Methods of Evaluating Antithrombotics and Thrombolytics

25

produced when platelets and erythrocytes pass through the light beam of a flow cytometer at the same time. 4. The “red blood cell–platelet” population in the flow cytometry histogram of Fig. 1.1 represents 3–5% of the displayed cells. A small fraction (5%) of this population likely represents leukocyte–platelet aggregates, based on analysis using an anti-CD45 monoclonal antibody. The remaining events correspond to platelet-associated erythrocytes. However, there is evidence that the majority of the latter population is an artifact generated by the simultaneous passage of a platelet and an erythrocyte through the light beam of the flow cytometer. That this population represents an artifact is supported by the observation that further dilution of pre-sheared and sheared blood specimens and/or reduction of the sample flow rate during the flow cytometric analysis results in a dramatic decrease in the “red blood cell–platelet” population. 5. Collagen density on the glass coverslip after rinsing with D-PBS can be determined by measuring the difference in average weight of 20 clean uncoated slides and 20 collagentreated slides. 6. Experiments are optimally monitored approximately 100– 200 μm downstream of the collagen/glass interface using a 60× FLUOR objective and a 1× projection lens, which gives a 3.2×104 μm2 field of view. Monitoring closer to the interface may yield non-reproducible results due to variations in the collagen layering in that region. Positions farther downstream are to be avoided as well so as to minimize the effects of upstream platelet adhesion and subsequent aggregation on the fluid dynamic environment and bulk platelet concentration. 7. The fluorescence intensity emitted by a single platelet can be determined by subtracting a digitized background image taken at the onset of perfusion, prior to platelet adhesion to the collagen I surface, from a subsequent image acquired 5 s after the initiation of platelet adhesion. The fluorescence intensity of a single platelet is represented by the mean gray level (black = 0; white = 255) of the platelet, obtained using image processing software (i.e., OPTIMAS; Agris-Schoen Vision Systems, Alexandria, VA, USA), multiplied by the corresponding area (total number of pixels) covered by the platelet. The intensity values for all single platelet events are averaged at the 5-s time point to arrive at an average single platelet intensity. 8. A single field of view (10×; 0.55 mm2 ) is monitored during the 3-min period of the experiment. At the conclusion

26

Mousa

of the experiment, five additional fields of view (0.55 mm2 ) are monitored for 15 s each. The following parameters can be quantified: (a) the total number of interacting cells per mm2 during the 3-min perfusion experiment; (b) the number of stationary interacting cells per mm2 after 3 min of shear flow; (c) the percentage of total interacting cells that are stationary after 3 min of shear flow; and (d) the average rolling velocity (μm/s) of interacting cells. The number of interacting cells/mm2 is determined manually by reviewing the videotapes. Stationary interacting cells/mm2 are those cells that move 90%) of dogs. It is also critical if one wishes to compare two or more drugs in this model and draw meaningful conclusions about drug effects. Inasmuch as the severity of the stenosis is an important component of the thrombogenic stimulus, comparable and uniform degrees of constriction between treatment groups are required, preferably those in which basal CBF is reduced between 10 and 25% and RH is abolished, or nearly so. It is important to apply these criteria when one is investigating a drug that possesses vasodilatory effects or one whose pharmacologic profile is not completely known. Without exhaustion of the vasodilatory reserve (as evidenced by abolition of RH), elimination of CFRs could result (at least partly) from coronary vasodilation. One difficulty in using RH or basal flow reduction immediately after placing a constrictor on the coronary artery is that CBF starts to decline quickly as platelets accumulate at the site of stenosis and intimal injury. Thus, one needs to assess the degree of flow reduction immediately after constricting the artery. Delaying this assessment will result in an overestimation of the stenosis severity due to accumulation of platelets on the vessel lining. Alternatively, the degree of stenosis can be ascertained by applying the constrictor before denuding the artery (see below), in which case the constrictor (or constrictors) needs to be removed and reapplied. After damaging and stenosing the coronary artery sufficiently, CBF starts declining immediately, reaching zero within 4–12 min, and remaining there until blood flow is restored by manually

36

Mousa

shaking loose the thrombus (“SL,” see Fig. 2.1, bottom). This is usually accomplished by either flicking the Lexan constrictor or sliding the constrictor up and down the artery to mechanically dislodge the thrombus. Spontaneous flow restorations occur under three circumstances: (1) non-severe conditions (i.e., minimal stenosis or de-endothelialization); (2) waning CFRs (which can occur as late as 30–45 mm after establishing CFRs); and/or (3) administration of a partially effective antithrombotic agent. Although the influence of blood pressure on the rate of formation of occlusive thrombi or their stability has not been studied systematically, one might predict that higher arterial pressures would increase the deceleration of CBF to zero by enhancing platelet aggregation through increased shear forces and increased delivery of platelets to the growing thrombus. Higher arterial pressure also might increase the propensity for spontaneous flow restorations before an occlusive thrombus is formed, due to greater stress on the nascent, unconsolidated thrombus. Several groups have examined histologically the coronary arteries harvested from dogs undergoing CFRs usually when CBF is declining or has ceased. Extensive intimal injury, including deendothelialization with adherent platelets and/or microthrombi, is consistently observed. Arteries harvested when CBF is zero invariably reveal a platelet-rich thrombus filling the stenotic segment (19, 21, 23, 28). These histological observations, coupled with the pattern of gradual, progressive declines in CBF and abrupt increases thereof (whether spontaneous or deliberate), provide further evidence that CFRs indeed are caused primarily by platelet thrombi, not vasoconstriction. Although the primary cause of CFRs is platelet aggregation, it is possible that local vasospasm and/or vasoconstriction downstream from the site of thrombosis are induced by vasoactive mediators released by activated and/or aggregating platelets. Experimental evidence supporting vasoconstriction just downstream from the stenosis during CFRs has been demonstrated (29). Further evidence for platelet-dependent thrombus formation in the etiology of CFRs is derived from the pharmacological profile of this model. In general, platelet-inhibitory agents consistently abolish or attenuate CFRs, whereas vasodilators (e.g., nitroglycerin, calcium entry blockers, and papaverine) affect them negligibly (30). Aspirin was the first described inhibitor of CFRs (21). However, in subsequent studies, its effects on CFRs were found to be variable and dose-dependent (22). Variability in the response to aspirin may be related to the severity of the stenosis, as further tightening of the constrictor after an effective dose of aspirin or ibuprofen usually restores CFRs. Prostacyclin, a powerful anti-aggregatory and potent coronary vasodilatory product of endothelial arachidonic acid

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

37

metabolism, is extremely efficacious and potent in abolishing CFRs. It is noteworthy that different drug classes can be compared, as evidenced by the wide range of percentages of responders (28). Advances in platelet physiology and pharmacology have identified a new class of antiplatelet agents that block the platelet membrane glycoprotein IIb/IIIa (GPIIb/IIIa) receptor and hence fibrinogen binding. Fibrinogen binding between platelets is an obligate event in aggregation and is initiated by blood-borne platelet agonists such as ADP, serotonin, thrombin, epinephrine, and collagen (31). The tripeptide sequence Arg–Gly–Asp (RGD), which occurs twice in the Aα-chain of fibrinogen, is believed to mediate, at least in part, the binding of fibrinogen to the GPIIb/IIIa complex. Early experimental results with GPIIb/IIIa antagonists in studies by Coller et al. (32), Bush (26), and Shebuski (25, 33) demonstrated that fibrinogen receptor antagonists are as effective as prostacyclin as anti-aggregatory and antithrombotic agents and do not possess the hemodynamic liabilities associated with prostaglandin-based compounds. Monoclonal antibodies directed against the platelet fibrinogen receptor (abciximab) are essentially irreversible, whereas RGD (tirofiban)- or KGD (eptifibatide)based fibrinogen receptor antagonists are reversible, their effects dissipating within hours after discontinuation of intravenous infusion. The prominence of platelet aggregation vis-à-vis coagulation mechanisms in the Folts model is evidenced by the lack of effect of heparin and thrombin inhibitors reported by most investigators (19, 22). However, heparin and MCI-9038, a thrombin inhibitor, were reported to abolish CFRs in about two-thirds of dogs with recently (30 min) established CFRs, but not in those extent after 3 h (34). The explanation for the differential effects of heparin is not immediately apparent. It may be related to the severity of the stenosis used. These apparently discrepant observations could be related to inhibition of thrombin-stimulated platelet activation and/or aggregation. An attractive feature of the Folts model is its amenability to dose-response studies. Unlike other models in which the thrombotic processes are dynamic, occurring over several minutes to hours, CFRs in the Folts model are repetitive and remarkably unchanging. In the many dogs that received either no intervention or vehicle 1 h after initiating CFRs, flow patterns remained unchanged for at least another hour (23). Thus, one can evaluate several doses of an investigational drug in a single dog. We and others have exploited this to determine potencies, an important basis of comparison between drugs with similar mechanisms of action, thus underscoring another feature of the model: its amenability to quantification of drug response. Two methods for quantifying drug effects in this model have been described.

38

Mousa

Aiken et al. (21) first described a four-point scoring scheme to assess and compare different doses or drugs, ranging from 0 (no effect on CFRs) to 3 (fully effective; complete abolition of CFRs). Intermediate scores of 1 and 2 are respectively applied when the CFR frequency was slowed (but occlusive thrombi still occurred) and when non-occlusive, spontaneously embolizing thrombi were observed. An advantage of this system is the provision of a single number for each evaluation period. A disadvantage is that agents that decrease systemic blood pressure (e.g., prostacyclin) will also decrease coronary perfusion pressure; the coronary flow pattern will be affected, making the scoring system somewhat more subjective. Another method of quantifying CFRs, described by Bush et al. (23), addresses the frequency, expressed on a per hour basis, and severity, based on the average nadir of CBF before a flow restoration. This system is less subjective, but it produces two values per evaluation period, and combinations of the two in an effort to provide a single parameter are awkward. In practical terms, both methods for quantifying CFRs described above provide similar answers. The important point for both is consistency in scoring. This end is best served by well-defined and communicated criteria. To date, the Folts model has been used only to evaluate antithrombotic drugs. No description of this model for the evaluation of thrombolytic drugs or adjunctive agents has been made. However, preliminary data reveal these thrombi to be resistant to doses of thrombolytic agents that lyse thrombi in other models (35). Of all the models described in this review, the thrombi in this model are probably the most platelet-rich and possess relatively less fibrin than, for example, the copper coil or wire models. However, it may be erroneous to conclude that these thrombi are devoid of fibrin, as the fibrinogen that links platelets during aggregation via the GPIIb/IIIa receptor is theoretically capable of undergoing fibrin formation. Several investigators have shown that the same combination of severe vessel narrowing and de-endothelialization results in CFRs in arteries other than the coronary. We have elicited CFRs in femoral arteries in anesthetized dogs with similar degrees of vessel narrowing and deliberate denudation of the artery (unpublished observation). Folts et al. (24) demonstrated that CFRs can be produced in conscious dogs with chronically implanted R Lexan coronary constrictors and flow probes. CFRs were prevented in the interim between implantation and acute study by the administration of aspirin. Al-Wathiqui (36) and Gallagher and co-workers (37) have demonstrated that progressive carotid or coronary arterial narrowing with ameroid constrictors will result in CFRs days to weeks after surgical implantation. These dogs apparently did not undergo deliberate vessel denudation at the

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

39

time of implantation. Perhaps focal inflammation developed in the intervening week(s) between the surgery and the development of CFRs in these animals. Alternatively, there was sufficient intimal vessel injury during implantation of the ameroid constrictors to induce development of CFRs at a later time. CFRs have also been elicited in the renal (38) and carotid (39) arteries of cynomolgus monkeys. Eidt et al. (40) showed that conscious dogs equipped with the same constrictors over segments of the LAD showing endothelial injury undergo repetitive CFRs in response to exercise, but not ventricular pacing. The frequency and severity of CFRs varied more in this model, and some CFRs were non-occlusive. CFRs of most dogs eventually deteriorated to persistent no- or lowflow states. Unlike the open-chest preparation, flow restorations observed in this model occurred spontaneously. Also, the severity of the stenosis produced was not as great as that produced by most practitioners of the Folts model, as reflected by the ability of CBF to increase above control levels initially during exercise. In summary, the Folts model of platelet-dependent thrombus formation is a well-established method to determine the pharmacology of antithrombotic agents. It represents an excellent choice for initial evaluation of antiplatelet activity in vivo, regardless of the artery used. Qualitatively, the thrombogenic stimuli in the Folts model and those responsible for unstable angina may be similar, since an involvement by platelets has clearly been demonstrated in the model and is strongly suspected clinically. It should be remembered, however, that flow restorations in the Folts model require vigorous shaking. In contrast, unstable angina is believed not to involve persistent, total thrombotic coronary occlusion. On the basis of the model’s pharmacological profile, the thrombi in this model also do not appear to resemble those usually responsible for acute myocardial infarction, as the former appear to be unresponsive to thrombolytic agents. The preliminary observations that either streptokinase (SK) or tissue plasminogen activator (t-PA) does not lyse thrombi in the Folts model contrast with the 50–75% response rate to thrombolytic therapy in patients with evolving myocardial infarction (35). However, it is tempting to speculate that the platelet-rich thrombi produced in this model are more like thrombi in those patients whose coronary arteries are not reopened by even early intervention and/or high doses of t-PA (41), and thus could represent a model of “thrombolytic-resistant” coronary thrombosis. In order to study new drugs for their antithrombotic potential in coronary arteries, Folts and Rowe (42) developed a model of periodic acute platelet thrombosis and CFRs in stenosed canine coronary arteries. Uchida described a similar model in 1975 (27). The model includes various aspects of unstable angina pectoris, i.e., critical stenosis, vascular damage, downstream vasospasm

40

Mousa

induced by vasoconstrictors released or generated by platelets. The cyclic variations in CBF are a result of acute platelet thrombi which may occlude the vessel but which either embolize spontaneously or can easily be embolized by shaking the constricting plastic cylinder. CFRs are not a result of vasospasm (24). Clinically, aspirin can reduce the morbidity and mortality of coronary thrombotic diseases but its effect is limited. Similarly, CFRs in the Folts model are abolished by aspirin but the effect can be reversed by increases in catecholamines and shear forces (43). As part of an expert meeting on animal models of thrombosis, a review of the Folts model has been published (44). In this section, five different protocols are described for the induction of coronary thrombosis. 2.1.1.1. Coronary Thrombosis Induced by Stenosis (Protocols 1–4)

The first four protocols are characterized by episodic, spontaneous decreases in CBF interrupted by restorations of blood flow. These alterations in CBF, or CFRs, are associated with transient platelet aggregation at the site of the coronary constriction and an abrupt increase in blood flow after embolization of platelet-rich thrombi. Damage to the vessel wall is achieved by placing a hemostatic clamp on the coronary artery. A fixed amount of stenosis is produced by an externally applied obstructive plastic cylinder at the damaged part of the vessel. In dogs, stenosis is critical; the reactive hyperemic response to a 10-second (s) occlusion is abolished (protocol 1). In pigs, stenosis is subcritical; partial reactive hyperemia remains (45). For some animals, particularly young dogs, damage of the vessel wall and stenosis are not sufficient to induce thrombotic cyclic flow variations. In these cases, additional activation of platelets by infusion of epinephrine (protocol 3) is required, leading to the formation of measurable thrombi. In protocol 4, thrombus formation is induced by subcritical stenosis without prior clamping of the artery and infusion of platelet activating factor (PAF), according to the model described by Apprill et al. (46). In addition to these protocols, coronary spasms induced by released platelet components can influence CBF. Therefore, this model includes the main pathological factors of unstable angina pectoris.

2.1.1.2. Coronary Thrombosis Induced by Electrical Stimulation (Protocol 5)

In this protocol, coronary thrombosis is induced by delivery of low amperage electrical current to the intimal surface of the artery, as described by Romson et al. (47). In contrast to the stenosis protocols, an occluding thrombosis is formed gradually without embolism after some hours. As a consequence of the time course, thrombi formed are of mixed type and contain more fibrin than platelet thrombi formed by critical stenosis.

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

41

2.1.2. Procedure 2.1.2.1. Protocol 1: Critical Stenosis

Dogs of either sex weighing 15–40 kg and at least 8 months of age are anesthetized with pentobarbital sodium (bolus of 30–40 mg/kg and then continuous infusion of approximately 0.1 mg/kg/min); respiration is maintained through a tracheal tube using a positive pressure respirator. The heart is exposed through a left thoracotomy at the fourth or fifth intercostal space; the pericard is opened and the left circumflex coronary artery (LCX) is exposed. An electromagnetic or Doppler flow probe is placed on the proximal part of the LCX to measure CBF. Distal to the flow probe, the vessel is squeezed with a 2-mm hemostatic clamp for 5 s. A small cylindrical plastic constrictor 2–4 mm in length and with an internal diameter of 1.2–1.8 mm (depending on the size of the LCX) is then placed around the artery at the site of the damage. Usually, the constrictor has to be changed several times (2–5 times) until the appropriate narrowing of the vessel is achieved and cyclic flow variations are observed. In the event of occlusion of the artery without spontaneous embolization of the formed thrombus, reflow is induced by shortly lifting the vessel with a thread placed beneath the stenotic site. Only dogs with regularly spaced CFRs of similar intensity within a pre-treatment phase of 60 min are used in these experiments. The test substance is administered by i.v. bolus injection or continuous infusion, or by intraduodenal application. CFRs are registered for 2–4×60 min and compared to pre-treatment values. Prior to testing, preparations for additional hemodynamic measurements are performed (see below).

2.1.2.2. Protocol 2: Subcritical Stenosis

Male castrated pigs (German landrace weighing 20–40 kg) are anesthetized with ketamine (2 mg/kg i.m.), metomidate (10 mg/kg i.p.), and xylazine (1–2 mg/kg i.m.). In order to maintain the stage of surgical anesthesia, animals receive a continuous i.v. infusion of 0.1–0.2 mg/kg/min pentobarbital sodium. Respiration is maintained through a tracheal tube using a positive pressure respirator. The heart is exposed through a left thoracotomy at the fourth and fifth intercostal space; the pericard is opened and the LAD is exposed. An electromagnetic or Doppler flow probe is placed on the proximal part of the LAD to measure CBF. Distal to the flow probe, the vessel is squeezed with a 1-mm hemostatic clamp for 5 s. A small cylindrical plastic constrictor 2 mm in length is then placed around the artery at the site of damage. Usually, the constrictor has to be changed several times until the appropriate narrowing of the vessel is achieved that produces CFRs. CFRs are similar to those in dogs; pigs, however, show a reactive hyperemic response. If embolization

42

Mousa

does not occur spontaneously, the formed thrombus is released by reducing blood flow by shortly lifting the vessel with forceps. Only pigs with regularly spaced CFRs of similar intensity within a pre-treatment phase of 60 min are used for the experiments. The test substance is administered by i.v. bolus injection or continuous infusion, or by intraduodenal application. CFRs are registered for 2×60 min and compared to pre-treatment values. 2.1.2.3. Protocol 3: Stenosis Plus Epinephrine Infusion

If protocol 1 does not lead to CFRs, additional epinephrine (0.2 μg/kg/min) is infused into a peripheral vein for 2×60 min (60 min before and 60 min after drug administration). CFRs are recorded, and the 60-min post-drug phase is compared to the 60-min pre-drug phase.

2.1.2.4. Protocol 4: Stenosis Plus PAF Infusion

The LCX is stenosed without prior mechanical wall injury. This preparation does not lead to thrombus formation (subcritical stenosis). For the induction of CFRs, PAF (C16-PAF, Bachem) (0.2 nmol/kg/min) is infused into one cannulated lateral branch of the coronary artery. After 30 min, PAF infusion is terminated and blood flow returns to a normal, continuous course. Thirty minutes later, the test substance is administered concomitantly with the initiation of a second PAF infusion for 30 min. CFRs are recorded and the drug treatment/second PAF phase is compared to the pre-drug/first PAF phase.

2.1.2.5. Protocol 5: Electrical Stimulation

The LCX is punctured distal to the flow probe with a chrome– vanadium–steel electrode (3 mm in length, 1 mm diameter). The electrode (anode) is placed in the vessel in contact with the intimal lining and connected over a teflon-coated wire to a 9-volt (V) battery, a potentiometer, and an amperemeter. A disc electrode (cathode) is secured to a subcutaneous thoracal muscle layer to complete the electrical circuit. The intima is stimulated with 150 μA for 6 hours (h). During this time, an occluding thrombosis is gradually formed. The test substance, or vehicle as a control, is administered either at the start of the electrical stimulation or 30 min after the start. The time until thrombotic occlusion of the vessel occurs and the thrombus size (wet weight measured immediately after removal at the end of the experiment) are determined. Prior to testing, preparations for additional hemodynamic measurements are performed (see below). For all protocols the following preparations and measurements are performed: 1. To measure peripheral arterial blood pressure (BP) [mm Hg], the right femoral artery is cannulated and connected to a Statham pressure transducer.

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

43

2. Left ventricular pressure (LVP) [mm Hg] is determined by inserting a micro tip catheter via the carotid artery retrogradely. 3. Left ventricular end-diastolic pressure (LVEDP) [mm Hg] is evaluated through sensitive amplification of the LVP. 4. Contractility (LV dp/dt max) [mm Hg/s] is determined from the initial slope of the LVP curve. 5. Heart rate [min–1] is determined from the pulsatile blood pressure curve. 6. The ECG is recorded in lead II. 7. Arterial pH and concentrations of blood gases are maintained at physiological levels by adjusting respiration and infusion of sodium bicarbonate. 8. Blood hematocrit values (37–40%) and number of erythrocytes are kept constant by infusion of oxypolygelatine in dogs and electrolyte solution in pigs. 9. Body temperature is monitored with a rectal thermistor probe and kept constant by placing the animals on a heated metal pad with automatic temperature regulation. 10. Template buccal mucosal bleeding time is carried out using the Simplate device. 11. At the end of the test, animals are sacrificed by an overdose of pentobarbital sodium. For detailed applications of the Folts model, see Folts (44), Folts and Rowe (42, 43), and Folts et al. (19, 24). 2.1.3. Evaluation

For all protocols, the mean maximal reduction of blood pressure (systolic/diastolic) [mm Hg] is determined.

2.1.3.1. Protocols 1–4

The following parameters are determined to quantify stenosisinduced coronary thrombosis: 1. Frequency of CFRs = cycle number per unit time 2. Magnitude of CFRs = cycle area [mm2 ] (total area of all CFRs per unit time, measured by planimetry) 3. Percent change in cycle number and cycle area after drug treatment is calculated relative to pre-treatment controls. 4. Statistical significance is assessed by the paired Student’s t-test.

2.1.3.2. Protocol 5

The following parameters are determined to quantify electrically induced coronary thrombosis: 1. Occlusion time [min] = time to zero blood flow. 2. Thrombus size [mg] = wet weight of the thrombus immediately after removal.

44

Mousa

3. Percent change in mean values for occlusion time and thrombus size in drug-treated groups is calculated relative to the control group. 4. Statistical significance is assessed by the non-paired Student’s t-test. 2.1.4. Critical Assessment of the Method

The stenosis (Folts) and electrical (Romson/Lucchesi) models of coronary thrombosis are widely used to study the role of mediators in the thrombotic process and the effect of new antithrombotic drugs. Bush and Patrick (28) provide an excellent review of the role of the endothelium in arterial thrombosis and the use of the Folts model to determine the effects of thrombosis inhibitors and mediators, i.e., thromboxane, prostacyclin, cyclooxygenase, serotonin, nitric oxide (NO) donors, and other vasodilators. The effect of an NO donor could be reversed by the NO scavenger oxyhemoglobin, which indicated that NO indeed mediates antithrombotic drug action (48). These coronary thrombosis models have recently been used to elucidate the mechanisms of action of several antithrombotic drugs, including the oral GPIIb/IIIa antagonist DMP 728 (49); the low molecular weight heparin (LMWH) enoxaparin (50), which, in contrast to unfractionated heparin, inhibited CFRs; the thrombin inhibitor PEG–hirudin (51); melagatran (52), an anti-P-selectin antibody (53); and activated protein C (54). The clinical relevance of the Folts model has been questioned because the model is very sensitive to antithrombotic compounds. However, in this model, lack of a reversal by epinephrine or an increase in degree of stenosis is able to differentiate any new drug from aspirin. Electrical coronary thrombosis is less sensitive (i.e., aspirin has no effect) and higher doses of some drugs are required. However, in principle, most drugs act in both models, if at all.

2.1.5. Modifications of the Method

Romson et al. (55) described a simple technique for the induction of coronary artery thrombosis in the conscious dog by delivery of low amperage electric current to the intimal surface of the artery. Benedict et al. (56) modified the electrical stimulation of thrombosis model by using two Doppler flow probes proximal and distal to the needle electrode in order to measure changes in blood flow velocity. The electrical current was stopped when a 50% increase in flow velocity was reached, at which point thrombosis occurred spontaneously. Using this model, the investigators demonstrated the importance of serotonin by measuring increased coronary sinus serotonin levels just prior to occlusion. Warltier et al. (57) described a canine model of thrombininduced coronary artery thrombosis to analyze the effects of intracoronary SK on regional myocardial blood flow, contractile function, and infarct size. Al-Wathiqui et al. (36) described

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

45

the induction of CFRs in the coronary, carotid, and femoral arteries of conscious chronically instrumented dogs. The Folts thrombosis model has also been applied to carotid arteries in monkeys. Coller et al. (58) induced CFRs in the carotid arteries of anesthetized cynomologus monkeys and demonstrated that they were abolished by the GPIIb/IIIa antibody abciximab. 2.2. Stenosis- and Mechanical Injury-Induced Arterial and Venous Thromboses (Harbauer Model) 2.2.1. Purpose and Rationale

Harbauer (59) first described a venous model of thrombosis induced by mechanical injury and stenosis of the jugular vein. In a modification of the technique, both arterial and venous thromboses are produced in rabbits by stenosis of the carotid artery and the jugular vein with simultaneous mechanical damage of the endothelium. This results in the activation of platelets and the coagulation system and leads to changes in the bloodstream pattern. As a consequence, occluding thrombi are formed and detected by blood flow measurements. The dominant role of platelets in this model is evidenced by the inhibitory effect of an antiplatelet serum in both types of vessels (60). The modified Harbauer model is used to evaluate the antithrombotic activity of compounds in an in vivo model of arterial and venous thromboses in which thrombus formation is highly dependent on platelet activation.

2.2.2. Procedure

Male Chinchilla rabbits weighing 3–4 kg receive test compound or vehicle as a control by oral, i.v., or i.p. administration. The first ligature (vein; for preparation, see below) is performed at the end of the absorption period (i.p., approximately 30 min; p.o., approximately 60 min; i.v., variable). Sixty-five minutes before stenosis, the animals are sedated by R i.m. injection of 8 mg/kg xylazine (Rompun ) and anesthetized by i.v. injection of 30–40 mg/kg pentobarbital sodium 5 min later. During the course of the test, anesthesia is maintained by continuous infusion of pentobarbital sodium (30–40 mg/kg/h) into one femoral vein. A Statham pressure transducer is placed into the right femoral artery for continuous measurement of blood pressure. Spontaneous respiration is maintained through a tracheal tube. One jugular vein and one carotid artery are exposed on opposite sides. Small branches of the vein are clamped to avoid blood flow around the vessel occlusion.

46

Mousa

Electromagnetic or Doppler flow probes are placed on the vein (directly central to the vein branching) and on the artery (as centered as possible). Blood flow [ml/min] is measured continuously. After blood flow reaches a steady state (approximately 15–30 min), a metal rod with a diameter of 1.3 mm is placed on the jugular vein (2 cm central to the vein branching) and a ligature is tightened. After 1 min, the rod is removed from the ligature. Immediately thereafter (approximately 1.5 min), the carotid artery is damaged by briefly squeezing it with forceps. A small plastic constricting cylinder (2 mm in length and 1.2 mm in diameter) is placed around the site of endothelial damage. Template bleeding time is measured at various time intervals before and after drug treatment (depending on the route of R administration) in the shaved inner ear using a Simplate device. Care is taken to select parts of the skin without large vessels. 2.2.3. Evaluation

1. Percent thrombus formation (thrombosis incidence) is determined as the number of occluded vessels (blood flow=0). 2. Percent inhibition of thrombosis is calculated in each dosage group relative to the respective vehicle controls. 3. Thrombosis incidence in the vehicle controls is set as 100%. 4. Statistical significance is assessed by means of the Fisherexact-test. 5. If initial values for blood flow do not significantly differ in the dosage and control groups, the area below the blood flow curve is measured by planimetry, and the mean value of each dosage group is compared to the control using the unpaired Student’s t-test. 6. Mean occlusion time [min] in the dosage and control groups are calculated and compared using the Students’s t-test. 7. The maximal change in systolic and diastolic blood pressure during the time period of stenosis as compared to the initial values before drug administration is determined. There is no standardized assessment score. For example, a reduction of systolic blood pressure by 30 mmHg and diastolic blood pressure by 20 mmHg is generally accepted as a strong reduction in blood pressure.

2.2.4. Critical Assessment of the Method

Two main factors of arterial thrombosis in human are essential components of this model: high-grade stenosis and vessel wall damage. In the absence of either, no thrombus is formed. The occlusive thrombus is formed fast and in a highly reproducible manner. In both vessels, thrombus formation is dependent on platelet function, as shown by the effects of antiplatelet serum.

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

47

Thus, jugular vein thrombosis in this model differs from stasisinduced deep vein thrombosis with prominent fibrin formation. On the other hand, occlusive thrombi are more stable than the pure platelet thrombi in the Folts model (see Section 2.1), as carotid blood flow cannot be restored by shaking the constrictor. The following antithrombotic drugs have been shown to be effective in this model: (1) antiplatelet drugs such as ticlopidine, prostacyclin/iloprost, NO donors (SNP, molsidomine), but not aspirin or thromboxane synthetase inhibitors; (2) anticoagulants such as hirudin, high-dose heparin, and warfarin; and (3) SK/t-PA (60, 61, and unpublished data). Drugs that simply lower blood pressure, such as hydralazine, clonidine, and prazosin have no effect on thrombus formation in this model. 2.2.5. Modifications of the Method

Bevilacqua et al. (61) applied this model to the rabbit carotid arteries and compared one artery before drug treatment with the contralateral artery after drug treatment. Heparin, the synthetic thrombin inhibitor FPRCH2 Cl, iloprost, and t-PA, but not aspirin, inhibited carotid occlusion in this model. Spokas and Wun (62) induced venous thrombosis in the vena cava of rabbits by vascular damage and stasis. Vascular wall damage was achieved by crushing the vessel with hemostat clamps. A segment of the vena cava was looped with two ligatures 2.5-cm apart, and then 2 h after ligation, the isolated venous sac was dissected and the clot was removed for determination of dry weight. Lyle et al. (63), in pursuit of an animal model that mimicked thrombotic re-occlusion and restenosis after successful coronary angioplasty in human, developed a model of angioplasty-induced injury in atherosclerotic rabbit femoral arteries. Acute 111 indiumlabeled platelet deposition and thrombosis were assessed 4 h after balloon injury in arteries subjected to prior endothelial damage (by air desiccation) and cholesterol supplementation (one month). The effects of inhibitors of FXa or platelet adhesion, heparin, and aspirin on platelet deposition were studied. Meng (64), Meng and Seuter (65), and Seuter et al. (66) described a method to induce arterial thrombosis in rats by chilling of the carotid artery (thrombosis induced by super cooling). Rats were anesthetized, and then the left carotid artery was exposed and occluded proximal by means of a small clamp. The artery was placed for 2 min into a metal groove that was cooled to –15◦ C. The vessel was then compressed using a weight of 200 g. In addition, a silver clip was fixed to the vessel distal to the injured area to produce disturbed and slow blood flow. After 4 min, the proximal clamp was removed and blood flow was reestablished in the injured artery. A similar model in the rabbit has also been developed, with slightly different conditions (chilling temperature of –12◦ C for 5 min, and a compression weight of 500 g). The wound is closed and the animal is allowed to recover from

48

Mousa

anesthesia. Antithrombotic compounds are administered in various doses at different time intervals before surgery. After 4 h, the animals receive heparin and are re-anesthetized. The lesioned carotid artery is removed and thrombus wet weight is immediately measured. 2.3. Electrically Induced Thrombosis

A novel technique for inducing arterial thrombosis was introduced by Salazar (67) in which anodal current was delivered to the intravascular lumen of a coronary artery in the dog via a stainless steel electrode. The electrode was positioned under fluoroscopic control, which somewhat complicated the procedure. Subsequently, Romson et al. (55) modified the procedure such that the electrode was placed directly into the coronary artery of an open-chest, anesthetized dog. This technique then allows one to produce a thrombus in the anesthetized animal or to close the chest after inserting the electrode and allow the animal to recover, after which thrombosis can be elicited later in the conscious animal. The advantage of this modification is that it allows induction of thrombus formation without the need for fluoroscopy. The stimulation electrode is constructed from a 25- or 26gauge stainless steel hypodermic needle tip, which is attached to a 30-gauge teflon-insulated silver-coated copper wire. Anodal current is delivered to the electrode via either a 9-V nickel–cadmium battery with the anode connected in series to a 250,000-ohm potentiometer or with a Grass stimulator connected to a Grass constant current unit and a stimulus isolation unit. The cathode in both cases is placed into a subcutaneous site completing the circuit. The anodal current can be adjusted to deliver 50–200 μA. Anodal stimulation results in focal endothelial disruption, which in turn induces platelet adhesion and aggregation at the damaged site. This process is then followed by further platelet aggregation and consolidation, with the growing thrombus entrapping red blood cells. A modification of the method of Romson et al. (55) involves placement onto the coronary artery of an external, adjustable occluder (68) to produce a fixed stenosis on the coronary artery. A flow probe to record CBF is placed on the proximal portion of the artery followed by the stimulation electrode, with the clamp being placed most distally (Fig. 2.2). The degree of stenosis can then be controlled by adjusting the clamp. The resulting stenosis is produced in an effort to mimic the human pathophysiology of atherosclerotic coronary artery disease, whereby thrombolytic therapy restores CBF through a coronary artery with residual narrowing due to atherosclerotic plaque formation. Another modification of the electrical stimulation model that merits discussion is described by Benedict et al. (56). They discontinued anodal current when mean distal coronary flow velocity (measured with Doppler flow meter) increased by approximately

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

49

Fig. 2.2. Model of coronary artery thrombosis in the dog. Electrical injury to the intimal surface of the artery leads to occlusive thrombus formation. The thrombus is formed in the presence of a flow-limiting stenosis induced by a Goldblatt clamp. Upon spontaneous occlusion, heparin is administered and the clot is aged for 1 h before initiating the t-PA infusion.

50%, reflecting disruption of normal axial flow by the growing thrombus. Occlusive thrombosis occurred within 1 h after stopping the current (2 h after starting the current). In these studies, coronary sinus plasma levels of serotonin, an index of intravascular platelet aggregation, were increased approximately 20-fold just before occlusive thrombus formation. The results of these studies agree with others in showing that either proximal flow velocity or electromagnetically measured CBF declines trivially over the majority of the time period in which the thrombus is growing. The largest declines in (volume) flow occur over a small and terminal fraction of the period between initial vessel perturbation and final occlusion. During that interval, coronary lumenal area decreases rapidly and to a critical degree, as platelets accrue at the growing thrombus. The studies by Benedict et al. (56) demonstrate that this final phase of thrombosis can occur independently of electrical stimulation. This variation of the model may be attractive to those who wish to produce occlusive thrombosis without continued electrical stimulation.

50

Mousa

Regardless of whether electrical stimulation is continued until occlusive thrombosis, there is another component to this model that has upside and downside potential, namely, the opportunity for coronary vasoconstriction to occur. Although the incidence of Prinzmetal’s angina is low, it is widely suspected that vasospasm superimposes on a primarily thrombotic event in unstable angina and myocardial infarction. In studies by Van der Giessen et al. (69), nifedipine was reported to increase the extent of CBF after plasmin-induced thrombolysis in a porcine model of electrically induced coronary thrombosis. In their model, the anodal stimulation was applied circumferentially to the exterior surface of the LAD, and an external constrictor was not used. Depending on the hypothesis being tested, the experimenter can leave intact or minimize this potential through the use or disuse of an external constrictor. As in the Folts and Gold coronary thrombo(ly)sis models, blood pressure must be taken into account or maintained within acceptable limits, since, in the presence of a critical stenosis, auto-regulation no longer exists. Under these conditions CBF is highly dependent on driving pressure (arterial pressure). Numerous experimental studies evaluating anticoagulants, antithrombotic, and/or thrombolytic drugs have been performed using this model. In the initial report by Romson et al. (55), the cyclooxygenase inhibitor ibuprofen was evaluated. Comparison of myocardial infarct size, thrombus weight, arrhythmia development, and scanning electron microscopy of drug-treated and control animals indicated that ibuprofen protected the conscious dog against the deleterious effects of coronary artery thrombosis. Subsequent studies in the same model and laboratory evaluated the antithrombotic potential of various TXA2 synthetase inhibitors, such as U 63557A, CGS 13080, OKY 1581, and dazoxiben. When the TXA2 synthetase inhibitors were administered before induction of the current, OKY 1581 (70) and CGS 13080 (71) reduced the incidence of coronary thrombosis, whereas U 63557A (72) and dazoxiben (73) were ineffective and partially effective, respectively. The differences in efficacy noted among the TXA2 synthetase inhibitors were ascribed to differences in potency and duration of action. Other investigators have used this model to study the prevention of original coronary thrombosis in the dog. Fitzgerald et al. (74) studied the TXA2 synthetase inhibitor U 63557A alone or in combination with L-636,499, an endoperoxide/thromboxane receptor antagonist. U 63557A alone did not prevent coronary thrombosis when administered before current application, whereas the combination of U 63557A and L-636,499 was highly effective. These data suggest that prostaglandin endoperoxides may modulate the effects of TXA2 synthetase inhibitors and that this response may be blocked by concurrent administration of

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

51

an endoperoxide/thromboxane receptor antagonist. The murine monoclonal antibody to platelet GPIIb/IIIa (7E3) was studied in this model for its ability to prevent thrombus formation. At a dose of 0.8 mg/kg i.v., the 7E3 monoclonal antibody completely prevented original thrombus formation (75). In addition to the evaluation of antithrombotic (i.e., antiplatelet) agents, the electrical injury model is useful for studying anticoagulant FXa inhibitors, such as YM 60628 (86), and thrombolytic drugs. When evaluating thrombolytic agents, the thrombus is allowed to form without drug intervention and then aged for various periods. Schumacher et al. (68) demonstrated that intracoronary SK was an effective thrombolytic drug in this model; the thrombolytic effectiveness being augmented by the concurrent administration of heparin and prostacyclin, or by a TXA2 synthetase inhibitor (68). In other studies reported by Shebuski et al. (76), the TXA2 receptor antagonist BM 13.177 hastened t-PA-induced thrombolysis and prevented acute thrombotic re-occlusion. Van der Giessen et al. (77) subsequently demonstrated that BM 13.177 prevented original thrombus formation in 75% of pigs undergoing electrical stimulation; aspirin was ineffective in this porcine model. These and other studies underscore the potential for adjunctive therapy to hasten thrombolysis and/or prevent re-occlusion, both contributing to greater salvage of ischemic myocardium. Like the copper coil model, the electrical stimulation model has been used to produce experimental myocardial infarction. Patterson et al. (78) have used this technique to produce coronary thrombosis in the LCX (which supplies blood flow to the posterior LV wall in dogs) in dogs with a previous anterior wall infarct to mimic sudden cardiac death that occurs in people during a second (recurrent) myocardial infarction or ischemic event. This model has also been modified to demonstrate the efficacy of adjuncts to thrombolytic therapy (79–82). In this case, the thrombus is allowed to extend until it completely occludes the vessel. Usually, the thrombus is allowed to stabilize, or “age,” to mimic the clinical setting in which a time lag exists between the thrombotic event and the pharmacological intervention. At the end of the stabilization period, thrombolytic agents such as t-PA or SK are administered in conjunction with the novel antithrombotic agent to lyse the thrombus and maintain vessel patency. The incidence and times of reperfusion and re-occlusion are the major endpoints. These studies have established that recombinant tick anticoagulant peptide (rTAP), a potent and selective FXa inhibitor derived from the soft tick (83), promotes rapid and prolonged reperfusion at doses that produce relatively minor elevations in PT, aPTT, and template bleeding time.

52

Mousa

2.3.1. Purpose and Rationale

The use of an electrical current to induce thrombosis in hamster and dog was described in the early 1950s by Lutz et al. (84) and Sawyer et al. (85, 86). In general, two different approaches are taken in this model. One method produces electrical damage by means of two externally applied hook-like electrodes (87, 88). The other method uses a needle electrode that is advanced through the walls of the blood vessel and positioned in the lumen; a second electrode is placed at a subcutaneous site to complete the circuit (55, 56, 67).

2.3.2. Procedure

Anaesthetized rats weighing 200–300 g are intubated and a femoral artery is cannulated for administration of test compound(s). One carotid artery is isolated from the surrounding tissue over a distance of 10–15 mm. A pair of rigid stainless steel wire hook-like electrodes with a working distance of 4 mm is positioned on the artery by means of a rack and pinion gear manipulator. The artery is raised slightly away from the surrounding tissue. Isolation of the electrodes is achieved by the insertion of a small piece of parafilm under the artery. Blood flow is measured with an ultrasonic Doppler flow meter (Transonic, Ithaca NY, USA); the flow probe (1RB) is placed proximal to the damaged area. Thrombus formation is induced in the carotid arteries by the application of an electrical current (350 V, DC, 2 mA) delivered by an electrical stimulator (Stoelting Co, Chicago, Cat. No 58040) for 5 min to the exterior surface of the artery.

2.3.3. Evaluation

1. Blood flow before and after induction of thrombus for 60 min. 2. Time to occlusion [min] = the time between onset of the electrical current and the time at which blood flow decreases to less than 0.3 ml/min. 3. Patency of the blood vessel over 30 min.

2.3.4. Critical Assessment of the Method

Thrombi formed by electrical induction are composed of densely packed platelets, with some red cells. Moreover, electrical injury causes extensive damage to intimal and sub-intimal layers. The endothelium is completely destroyed, and the damage extends to sub-endothelial structures, including smooth muscle cells. This deep damage could reduce sensitivity in terms of discriminating between drugs on the basis of their antithrombotic activity. However, Philp et al. (88) showed that unfractionated heparin completely blocks thrombus formation, whereas other antiplatelet agents exhibit differential antithrombotic actions. The investigators concluded that this relatively simple model of arterial thrombosis might prove to be a useful screening test for drugs with antithrombotic potential.

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

53

2.3.5. Modifications of the Method

In a modification of this model by Salazar (67), a stainless steel electrode is inserted into a coronary artery in the dog to deliver anodal current to the intravascular lumen. The electrode is positioned under fluoroscopic control, complicating the procedure somewhat. Romson et al. (55) described a further modification in which the electrode was placed directly into the coronary artery of open-chest anaesthetized dogs. Rote et al. (89, 90) applied the carotid thrombosis model to dogs. A calibrated electromagnetic flow meter was placed on each common carotid artery proximal to the point of insertion of an intravascular electrode and a mechanical constrictor. The external constrictor was adjusted with a screw until the pulsatile flow pattern was decreased by 25% without alteration in mean blood flow. Electrolytic injury to the intimal surface was accomplished with an intravascular electrode composed of a teflon-insulated silvercoated copper wire connected to the positive pole of a 9-V nickel– cadmium battery in series with a 250,000-ohm variable resistor. The cathode was connected to a subcutaneous site. Injury was initiated in the right carotid artery by application of a 150-μA continuous pulse anodal direct current to the intimal surface of the vessel for a maximum duration of 3 h, or for 30 min beyond the time of complete vessel occlusion, as determined by blood flow recordings. Upon completion of the study on the right carotid, the procedure was repeated on the left carotid artery after administration of test drug. Benedict et al. (56) introduced a procedure in which anodal current was discontinued when mean distal coronary flow velocity increased by approximately 50%, reflecting disruption of normal flow by the growing thrombus. An occlusive thrombosis occurred within 1 h after cessation of the electrical current. In this model, the final phase of thrombosis occurred independently of electrical injury. A ferret model of acute arterial thrombosis was developed by Schumacher et al. (91). A 10-min anodal electrical current of 1 mA was delivered to the external surface of the carotid artery while measuring carotid blood flow. This produced an occlusive thrombus in all vehicle treated ferrets within 41±3 min with an average weight of 8±1 mg. Thrombus weight was reduced by aspirin or a thromboxane receptor antagonist. Guarini (92) reported the formation of a completely occlusive thrombus in the common carotid artery of rats by applying an electrical current to the arterial wall (2 mA for 5 min) while simultaneously constricting the artery with a hemostatic clamp placed immediately downstream from the electrodes.

2.4. Ferric Chloride (FeCl3 )-Induced Thrombosis

The administration of a variety of chemicals either systemically or locally can result in damage to the endothelium with

54

Mousa

subsequent generation of a thrombus. Such compounds include ferric/ferrous chloride, fluorescein-labeled dextran and Rose Bengal. In models employing ferric (ferrous) chloride (93), the carotid artery of rats is isolated. A flow probe is placed proximal to the intended site of lesion and a 3-mm disc of filter paper which has been soaked in ferric/ferrous chloride (35-50%) is placed on the artery. The application of ferric (ferrous) chloride results in transmural vascular injury leading to the formation of occlusive thrombi. This injury is believed to be a result of lipid peroxidation catalyzed by the ferric (ferrous) chloride. Thrombus formation, measured as a decrease in blood flow through the vessel, typically occurs within 30 min. Microscopic analysis of the thrombi has shown them to be predominantly platelet-rich clots. This model has been used to study the antithrombotic effects of direct thrombin inhibitors (94–96) and heparins. Endothelial damage can also be induced by fluoresceinor fluorescein isothiocyanate (FITC)-conjugated compounds. A model has been described in which FITC–dextran is administered intravenously to mice. Thrombus formation is induced upon exposure of the microvessels of the ear to the light of a mercury lamp (excitation wavelength of 450–490 nm) (97). The endothelial damage induced in this model is believed to be a result of the generation of singlet molecular oxygen produced by energy transfer from the excited dye (98). Thrombus formation is measured using intravital fluorescence microscopy. This detection technique allows for a number of endpoints to be quantitated, including changes in luminal diameter due to thrombus formation, blood flow measurements, and extravasation of the FITC–dextran. This model offers the advantages of not requiring surgical manipulations, which can cause hemodynamic or inflammatory changes, allowing for repeated analysis of the same vessel segments over time, and being applicable to the study of both arteriolar and venular thromboses. The administration of Rose Bengal has been used similarly (99). 2.4.1. Purpose and Rationale

A variety of chemical agents have been used to induce thrombosis in animals. The use of topical FeCl3 as a thrombogenic stimulus in veins was described by Reimann-Hunziger (100). Kurz et al. (93) demonstrated that the thrombus produced with this method in the carotid arteries of rats is composed of platelets and red blood cells enmeshed in a fibrin network. This simple and reproducible test has been used for the evaluation of antithrombotic (101) and pro-fibrinolytic test compounds (102).

2.4.2. Procedure

Rats weighing 250–300 g are anaesthetized with Inactin (100 mg/kg) and a polyethylene catheter (PE-205) is inserted

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

55

into the trachea via a tracheotomy to facilitate breathing. Catheters are also placed in the femoral artery for blood sampling and measurement of arterial blood pressure and in the jugular vein for administration of test compounds. The right carotid artery is isolated and an ultrasonic Doppler flowprobe (probe 1RB, Transonic, Ithaca NY, USA) is placed on the vessel to measure blood flow. A small piece of parafilm “M” (American Can Co., Greenwich, CT, USA) is placed under the vessel to isolate it from surrounding tissue throughout the experiment. The test compound is administered by gavage or as an i.v. injection at a defined time prior to initiation of thrombus formation. Thrombus formation is induced by the application of a piece of filter paper (2 mm×5 mm) saturated with 25% FeCl3 to the carotid artery. The paper is allowed to remain on the vessel for 10 min and is then removed. Parameters (see below) are monitored for 60 min after the induction of thrombosis, after which the thrombus is removed and weighed. 2.4.3. Evaluation

1. Blood flow before and after induction of thrombus for 60 min 2. Time to occlusion [min]: the time between FeCl3 application and the point at which blood flow decreases to less than 0.3 ml/min 3. Thrombus weight after blotting the thrombus on filter paper

2.5. Thrombin-Induced Clot Formation in Rabbit Femoral or Canine Coronary Artery

Localized thrombosis can also be produced in rabbit peripheral blood vessels such as the femoral artery by injection of thrombin, calcium chloride and fresh blood via a side branch (103). Either femoral artery is isolated distal to the inguinal ligament and traumatized distally from the lateral circumflex artery by rubbing the artery with the jaws of forceps. An electromagnetic flow probe is placed distal to the lateral circumflex artery to monitor femoral artery blood flow (Fig. 2.3). The superficial epigastric artery is cannulated for induction of the thrombus and subsequent infusion of thrombolytic agents. Localized thrombi distal to the lateral circumflex artery with snares approximately 1-cm apart are induced by the sequential injection of thrombin, CaCl2 (1.25 mmol), and a volume of blood sufficient to distend the artery. After 30 min, the snares are released and femoral artery blood flow is monitored for 30 mm to confirm total obstruction of flow by the thrombus. These models are not appropriate for evaluating drugs for their ability to inhibit original thrombosis. However, the model is particularly appropriate for evaluating thrombolytic agents and adjunctive therapies for their ability to hasten and/or enhance lysis or prevent acute re-occlusion after discontinuing administration of a thrombolytic agent.

56

Mousa

Fig. 2.3. Rabbit model of femoral arterial thrombosis. A clot is introduced into an isolated segment of femoral artery by injection of thrombin, CaCl2 , and whole blood. After aging for 1 h, t-PA is infused. Reperfusion is assessed by restoration of blood flow.

2.5.1. Purpose and Rationale

A canine model of thrombin-induced clot formation was developed by Gold et al. (104) in which localized coronary thrombosis was produced in the LAD. This is a variation of the technique described by Collen et al. (105) who used radioactive fibrinogen to monitor the occurrence and extent of thrombolysis of rabbit jugular vein clots. The vessel was intentionally de-endothelialized by external compression with blunt forceps. Snare occluders were then placed proximal and distal to the damaged site, and thrombin (10 U) was injected into the isolated LAD segment in a small volume via a previously isolated side branch. Autologous blood (0.3–0.4 ml) mixed with calcium chloride (0.05 M) was also injected into the isolated LAD segment, producing a stasis-type red clot superimposed on an injured blood vessel. The snares were released 2–5 min later and total occlusion was confirmed by selective coronary angiography. This model of coronary artery thrombosis relies on the conversion of fibrinogen to fibrin by thrombin. The fibrin-rich thrombus contains platelets, but at no greater concentration than in a similar volume of whole blood. Once the thrombus is formed, it is allowed to age for 1–2 h, after which a thrombolytic agent can be administered to lyse the thrombus and restore blood flow.

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

57

2.5.2. Procedure

In the initial study described by Gold et al. (104), recombinant t-PA was characterized for its ability to lyse 2-h-old thrombi. Tissue plasminogen activator was infused at doses of 4.3, 10, and 25 μg/kg/min i.v. and resulted in reperfusion times of 40, 31, and 13 min, respectively. Thus, in this model of canine coronary thrombosis, t-PA exhibited dose-dependent coronary thrombolysis. It is also possible to study the effects of different doses of t-PA on parameters of systemic fibrinolytic activation, such as fibrinogen, plasminogen, and a2-antiplasmin, as well as to assess myocardial infarct size. For example, Kopia et al. (106) demonstrated that SK elicited dose-dependent thrombolysis in this model. Subsequently, Gold et al. (107, 108) modified the model to study not only reperfusion, but also acute re-occlusion. Clinically, re-occlusion is a persistent problem after effective coronary thrombolysis, which is reported to occur in 15–45% of patients (109). Thus, an animal model of coronary reperfusion and re-occlusion would be important from the standpoint of evaluating adjunctive therapies to t-PA to hasten and/or increase the response rate to thrombolysis as well as prevent acute re-occlusion. The model of thrombin-induced clot formation in the canine coronary artery was modified such that a controlled high-grade stenosis was produced with an external constrictor. Blood flow was monitored with an electromagnetic flow probe. In this model of clot formation with superimposed stenosis, reperfusion in response to t-PA occurs with subsequent re-occlusion. The monoclonal antibody against the human GPIIb/IIIa receptor developed by Coller et al. (110) and tested in combination with t-PA in the canine thrombosis model hastened t-PA-induced thrombolysis and prevented acute re-occlusion (111). These actions in vivo were accompanied by abolition of ADP-induced platelet aggregation and markedly prolonged bleeding time.

2.6. Laser-Induced Thrombosis

The physiologic responses to injury in the arterial and venous systems vary in part due to differences in blood flow conditions, leading to different clot compositions. This model of arterial thrombosis is based on the development of a platelet-rich thrombus following laser-mediated thermal injury to the vascular wall. This model was first described by Weichert and Breddin (112). In this model, an intestinal loop of an anesthetized rat is exposed through a hypogastric incision and spread on a microscope stage while being continuously irrigated with sterile physiologic saline. Vascular lesions are induced on small mesenteric arterioles with an argon laser beam (50 mW at microscope, 150-ms duration) directed through the optical path of the microscope. Exposure of the laser beam is controlled by means of a camera shutter. Laser shots are made every minute. Antithrombotic potency is evaluated

58

Mousa

in real time by microscopic evaluation of vascular occlusion. The number of laser injuries required to induce a thrombus with a length of at least 1.5 times the inner diameter of the vessel is taken as an endpoint. The antithrombotic activity of several thrombin inhibitors has been compared to unfractionated heparin using the laser-induced thrombosis model. Each inhibitor was administered intravenously via one of the tail veins and allowed to circulate for 5 min prior to the initiation of the laser-induced lesions. Saline-treated control rats required an average of three laser shots to reach an endpoint. Each thrombin inhibitor produced a dose-dependent antithrombotic effect in this model. In comparing the dose of each agent required to extend the endpoint to six laser shots, heparin was observed to be the most potent antithrombotic agent (0.08 μmol/kg), followed by Ac-(D)-Phe-Pro-boroArgOH (0.154 μmol/kg) and then hirudin (0.28 μmol/kg). Consistent with the results obtained with these agents in the rabbit jugular vein stasis thrombosis model, D-Me-Phe-Pro-Arg-H exhibited the weakest effects in the laser-induced thrombosis model (2 μmol/kg). 2.6.1. Purpose and Rationale

In this model, thrombus formation in rat or rabbit mesenteric arterioles or venules is induced by laser-mediated thermal injury to the vascular wall. The procedure can be performed in normal or pretreated (i.e., induced arteriosclerosis or adjuvant arthritis) animals. In this model, thrombus formation is mediated by a dual mechanism of platelet adhesion to the injured endothelial vessel wall and ADP-induced platelet aggregation. Most likely, ADP is released by erythrocytes that are lysed by the laser, based on the observation that erythrocyte hemoglobin strongly absorbs the frequencies of light emitted by the laser beam. A secondary aggregation stimulus following the release of ADP is mediated by the platelets themselves.

2.6.2. Procedure

2.6.2.1. Equipment

1. 4 W argon laser (Spectra Physics, Darmstadt, FRG) with a wave length of 514.5 nm; energy below the objective of 15 mW; duration of exposure, 1/30 or 1/15 s. 2. Microscope ICM 405, LD-Epipland 40/0.60 (Zeiss, Oberkochen, FRG) 3. Video camera (Sony, Trinicon tube) 4. Recorder (Sony, U-matic 3/4 ) 5. Videoanalyzer to determine blood flow velocity

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

2.6.2.2. Experimental Course

2.6.2.3. Standard Compounds

2.6.3. Evaluation

59

Male Sprague-Dawley, spontaneously hypertensive stroke-prone Wistar or Lewis rats with adjuvant-induced arthritis weighing 150–300 g are used. Alternatively, New Zealand rabbits with arteriosclerosis induced by cholesterol feeding for 3 months are used. Animals receive test compound by oral, i.v., i.p. or s.c. administration. Control animals are treated with vehicle alone. Prior to thrombus induction, the animals are pretreated by s.c. injection of 0.1 mg/kg atropine sulfate solution and anaesthetized by i.p. administration of 100 mg/kg ketamine hydrochloride and 4 mg/kg xylazine. Thrombus formation is induced 15, 30, 60, or 90 min postdosing. The procedure is carried out in arterioles or venules 13 ± 1 μm in diameter of the fat-free ileocaecal portion of the mesentery. During the procedure, the mesenterium is superfused with a physiological saline solution or degassed paraffin liquid (37◦ C). The ray of the argon laser is guided into the inverted ray path of the microscope by means of a ray adaptation and adjusting device. The frequency of injury is 1 per 2 min. The exposure time for a single laser shot is 1/30 or 1/15 s. The number of injuries necessary to induce a defined thrombus is recorded. All thrombi formed during the observation period with a minimum length of 13 μm or an area of at least 25 μm2 are evaluated. The procedure is photographed using a video system. • acetylsalicylic acid (10 mg/kg, per os) • pentoxifylline (10 mg/kg, per os) For a detailed description and evaluation of various agents and mechanisms, please refer to the following references: Arfors et al. (113); Herrmann (114); Seiffge and Kremer (115, 116); Seiffge and Weithmann (117); and Weichert (118). The number of laser shots required to produce a defined thrombus is determined. Mean values and SEM are calculated. Results are typically presented in graph form.

2.7. Photochemical Induced Thrombosis 2.7.1. Purpose and Rationale

In 1977, Rosenblum and Sabban (119) reported that ultraviolet light can produce platelet aggregation in cerebral microvessels of the mouse after intravascular administration of sodium fluorescein, and demonstrated that in contrast to heparin, aspirin and indomethacin prolonged the time to first platelet aggregation. A detailed study by Herrmann (114) demonstrated that scavengers of singlet oxygen, but not hydroxyl radicals, inhibited platelet

60

Mousa

aggregation induced by photochemical reaction. The investigators postulated that excitation of intravascular fluorescein results in the production of singlet oxygen, which damages endothelial cells and leads to platelet adhesion and aggregation. 2.7.2. Procedure

Studies are performed in mesenteric arteries 15–30 μm in diameter in anesthetized rats. After i.v. injection of 0.3 ml of fluorescein isothiocyanate–dextran 70 (FITC–dextran; 10%) (Sigma), arterioles are exposed to ultraviolet light (excitation, 490 nm; emission, 510 nm).

2.7.3. Evaluation

Thrombus formation is quantitated by determining the time between onset of excitation and appearance of the first platelet aggregate adhering to the vessel wall.

2.7.4. Critical Assessment of the Method

In contrast to other thrombosis induction methods, photochemically induced thrombosis is amenable to use in small animals. Thrombi are composed primarily of platelets, however, the primary target of the photochemical insult is endothelial cells through induced oxygen radical damage.

2.7.5. Modifications of the Method

Matsuno et al. (120) reported a method to induce thrombosis in the rat femoral artery by means of a photochemical reaction after injection of a fluorescent dye (Rose Bengal, 10 mg/kg i.v.) followed by transillumination with a filtered xenon lamp (wave length, 540 nm). Blood flow was monitored by a pulsed Doppler flow meter. Occlusion was achieved after approximately 5–6 min. Pretreatment with heparin prolonged the time required to interrupt the blood flow in a dose-dependent manner. This model has also been used to study the thrombolytic mechanisms of t-PA. For a comparative analysis of hirudin in various models, see Just et al. (45).

2.8. Foreign Surface-Induced Thrombosis

The presence of foreign materials in the circulation results in activation of the coagulation and platelet systems. A variety of prothrombotic surfaces have been used for the development of experimental animal thrombosis models. In contrast to many other thrombosis models, thrombosis induced by foreign surfaces does not presuppose endothelial damage.

2.8.1. Wire Coil-Induced Thrombosis 2.8.1.1. Purpose and Rationale

This classical method of producing thrombosis is based on the insertion of wire coils into the lumen of blood vessels. The model was first described by Stone and Lord (121) using the dog aorta and was further modified for use in arterial coronary vessels of opened-chest dogs. The formation of thrombotic material around

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

61

the coil is reproducible and can be easily standardized for pharmacological studies (48, 122, 123). The use of this model in venous vessels was described by Kumada et al. (124). Venous thrombosis is produced in rats by insertion of a stainless steel wire coil into the inferior caval vein. Platelets and plasmatic coagulation are activated on the wire coil. Thrombus formation on the wire is quantitated by measuring the protein content of the isolated thrombotic material. The kinetics of thrombus formation show an increase in weight and protein content within the first 30 min of insertion, followed by a period of steady state flux between thrombus formation and endogenous thrombolysis, leading to a level protein content of thrombi starting at 1 h and lasting up to 48 h after implantation. The incidence of thrombosis in untreated control animals in this model is 100%. The model is used to evaluate antithrombotic and thrombolytic properties of test compounds in an in vivo model of venous thrombosis in rats. 2.8.1.2. Procedure

2.8.1.3. Evaluation

Male Sprague-Dawley rats weighing 260–300 g receive test compound, or vehicle as a control, by oral, i.v. or i.p. administration. At the end of absorption (i.v., 1 min; i.p., 30 min; p.o., 60 min), the animals are anesthetized by i.p. injection of 1.3 g/kg of urethane. Through a midline incision the caudal caval vein is exposed R and a stainless steel wire coil (Zipperer , size 40; Zdarsky Erler KG, München) is inserted into the lumen of the vein just below the left renal vein branching by gently twisting the wire toward the iliac vein. The handle of the carrier is cut off so as to hold the back end of the wire at the vein wall. The incision is sutured and the animal is placed on its back on a heating pad (37◦ C). The wound is reopened after 2 h and the wire coil with the thrombus on it is carefully removed and rinsed with a 0.9% saline solution. The thrombotic material is dissolved in 2 ml of alkaline sodium carbonate solution (2% Na2 CO3 in 0.1 N NaOH) in a boiling water bath for 3 min. The protein content is determined in 100 μl aliquots by the colorimetric method of Lowry (Fig. 2.4). Thrombolysis. In addition to the procedure described above, a thrombolytic test solution is continuously infused through a polyethylene catheter inserted into the jugular vein. Ninety minutes after wire implantation, the test compound or the vehicle (control) is infused for up to 2.5 h. The wire coil is then removed and the protein content of the thrombus is determined. Using this model, Bernat et al. (125) demonstrated the fibrinolytic activity of urokinase and SK–human plasminogen complex. 1. Thrombosis incidence = number of animals in each dosage group that develop thrombi as compared to the vehicle control.

62

Mousa

Fig. 2.4. Schematic diagram of the canine femoral artery copper coil model of thrombolysis. A thrombogenic copper coil is advanced to either femoral artery via the left carotid artery. By virtue of the favorable anatomical angles of attachment, a hollow polyurethane catheter advanced down the left carotid artery nearly always enters the descending aorta, and with further advancement, either femoral artery without fluoroscopic guidance. A flexible, teflon-coated guidewire is then inserted through the hollow catheter and the latter is removed. A copper coil is then slipped over the guidewire and advanced to the femoral artery (see inset). Femoral artery flow velocity is measured directly and continuously with a Doppler flow probe placed just proximal to the thrombogenic coil and distal to a prominent sidebranch, which is left patent to dissipate any dead space between the coil and the next proximal sidebranch. Femoral artery blood flow declines progressively to total occlusion over the next 10–12 mm after coil insertion.

2. The mean protein content [mg] of thrombotic material in each dosage group as compared to the vehicle control is determined. Percent change in protein content is calculated relative to control. 3. Statistical significance is assessed by means of the unpaired Student’s t-test. 2.8.2. Eversion Graft-Induced Thrombosis 2.8.2.1. Purpose and Rationale

The eversion graft model of thrombosis in the rabbit artery was first described by Hergrueter et al. (126) and later modified by

In Vivo Models for the Evaluation of Antithrombotics and Thrombolytics

63

Jang et al. (127, 128) and Gold et al. (15). A 4- to 6-mm segment of the rabbit femoral or the dog left circumflex artery is excised, everted, and then re-implanted into the vessel by end-toend anastomosis. After restoration of blood flow, a platelet-rich occlusive thrombus forms rapidly leading to complete occlusion of the vessel. The rabbit model described here uses a carotid graft inserted into the femoral graft to avoid vasoconstriction, which often occurs in the inverted femoral segments. 2.8.2.2. Procedure

2.8.2.3. Evaluation

In anaesthetized New Zealand white rabbits, the right carotid artery is exposed. After double ligation, a 3-mm segment of the artery is excised, everted, and immersed in pre-warmed (37◦ C) isotonic saline. The right femoral artery is exposed and occluded by means of a double occluder (2-cm distance). The femoral artery is transected and the everted graft from the carotid artery is inserted by end-to-end anastomosis using 12 sutures and 9-0 nylon (Prolene; Ethicon, Norderstedt, Germany) under a surgical microscope (Wild M650; Leitz, Heerbrugg, Switzerland). Perfusion of the graft is measured by means of an ultrasonic flow meter (Model T106; Transonic, Ithaca, NY, USA). The flow probe is positioned 2 cm distal from the graft. After a stabilization period of 15 min, the test substance is administered i.v. through the catheterized right jugular vein. Ten minutes after administration of the test compound, the vessel clamps are released and blood flow is monitored by the flow meter for 120 min. Arterial blood is collected from the left carotid artery at baseline (immediately before administration of test compound), and 10, 60, and 120 min after administration. 1. Time until occlusion = time between restoration of vessel blood flow and occlusion of the vessel, as indicated by a flow of less than 3.0 ml/min. 2. Patency = time during which perfusion of the graft is measured relative to an observation period of 120 min after administration of test compound. 3. Time until occlusion and patency are expressed as median and inter-quartile range/2 (IQR/2). Significant differences (P3 times ULN) was lower in the dabigatran groups (1.5–3.1%) than in the enoxaparin group (7.4%). There were no cases of clinically relevant thrombocytopenia. The authors concluded that dabigatran started in the early post-operative period was effective and safe across a wide range of doses. In addition, the frequency and extent of severe hepatic abnormalities were lower than those observed with ximelagatran. Dabigatran is currently undergoing extensive phase-III evaluation for VTE prevention, treatment, and secondary thromboprophylaxis through the RE-VOLUTION program. A randomized, double-blind, non-inferiority trial was conducted comparing dabigatran etexilate to enoxaparin for prevention of VTE after total hip replacement (72). Patients (3,494 total) undergoing total hip replacement were randomized into treatment for 28–35 days with dabigatran etexilate 220 mg (n=1157) or 150 mg (n=1174) once daily, starting with a halfdose 1–4 h after surgery, or SC enoxaparin (40 mg) once daily (n=1162), starting the evening before surgery. The primary efficacy outcome was the composite of total VTE (venographic or symptomatic) and death from all causes during treatment. On the basis of the absolute difference in rates of VTE with enoxaparin versus placebo, the non-inferiority margin for the difference in rates of thromboembolism was defined as 7.7%. Both doses of dabigatran were non-inferior as compared to enoxaparin. There was no significant difference in major bleeding rates with either dose of dabigatran etexilate as compared with enoxaparin (72). The frequency of increased liver enzyme concentrations and acute coronary events during the study did not differ significantly between the groups. The study concluded that oral dabigatran etexilate was as effective as enoxaparin in reducing the risk of VTE after total hip replacement surgery, with a similar safety profile (72). 4.3.3. TGN-167

TGN-167 (TRI-50c-04) is an oral thrombin inhibitor being developed by Trigen Holdings for the potential treatment of

174

Mousa

thrombosis. A controlled release formulation of the drug is also being developed with Eurand for long-term treatment of thrombosis. The compound produces a marked increase in thrombin clotting time, with minimal effects on aPTT. A double-blind, phase-I, dose-escalation study with 20 volunteers showed the drug to be well tolerated, with no significant adverse events reported (73). At 600 mg, all dosed subjects achieved effective anticoagulant activity in vitro. Trigen is planning to continue TGN-167 into phase-II studies.

5. Conclusions The antithrombotic management of VTE will undergo significant changes in the next 5-10 years. Limitations of existing parenteral and oral anticoagulants has led to the development of new agents designed to target specific procoagulant complexes in the coagulation pathway, inhibiting coagulation initiation, coagulation propagation, or thrombin activity. With respect to efficacy, during acute treatment of VTE, newer antithrombotic agents must exhibit at least non-inferiority in a methodologically sound study as compared to the existing parenteral agent of choice, LMWH, and the emerging agent fondaparinux. This is true particularly in high-risk venous thromboses, such as ileofemoral VTE, PE, or VTE associated with cancer. For long-term VTE treatment, there is a need to improve upon existing oral anticoagulants, namely, VKAs. Target-selective oral agents must exhibit an improved safety profile (especially as it pertains to major or clinically significant bleeding), ease of use, and tolerability as compared to VKAs. If successful, emerging oral anticoagulants could negate the traditional distinction of acute versus long-term treatment of VTE, as they could potentially be used throughout the spectrum of disease without the need for overlap with parenteral therapies (52). Lastly, any new long-term anticoagulant must be safely tolerated in combination with antiplatelet agents, as an increasingly aging population will be prone to arterial as well as venous thromboembolic disease. Cost considerations are also important, especially from a populational perspective. Newer agents should, in theory, fulfill the following requirements of an ideal anticoagulant: a rapid onset with predictable response characteristics, predictable pharmacokinetics, pharmacodynamics with low plasma protein binding, no required monitoring, a half-life that provides both safety and ease of use (particularly during temporary withdrawal), lack of food or drug interactions, an excellent safety profile (particularly with respect to immune-mediated thrombocytopenia, hepatotoxicity, and

Novel Anticoagulant Therapy: Principle and Practice

175

potential for thrombotic rebound phenomenon), and reversibility or availability of an antidote. In addition, oral agents with predictable intestinal absorption/bioavailability used in a simple, fixed-dose once or twice daily regimen and for which compliance can be monitored would be even more advantageous. At this time, drugs at the most advanced stage of development with respect to VTE management include the parenteral indirect FXa inhibitor idraparinux and biotinylated idraparinux, the oral DTI dabigatran, and the oral selective direct FXa inhibitors rivaroxaban and apixaban. Whether there are inherent advantages in blocking initial thrombin formation via the prothrombinase complex early in the coagulation system or blocking thrombin directly and preventing feedback amplification is still a matter of debate, as is the notion of whether there is any clinically meaningful effect of small-molecule DTIs that target both clot-bound and free thrombin. Long-term clinical data with respect to efficacy of anti-Xa inhibitors will be available shortly, while long-term data are currently available on the efficacy of direct thrombin inhibition. The lessons from ximelagatran reveal the importance of long-term safety data in different patient populations. Ximelagatran had shown significant potential as a possible replacement to warfarin therapy, but was withdrawn because of potential liver toxicity. In contrast, dabigatran appears to have a better safety profile and has recently entered a phase-III randomized clinical trial for AF. Oral direct FXa inhibitors (rivaroxaban, apixaban, and others) may prove to be more potent and safe. Selective inhibitors of specific coagulation factors involved in the initiation and propagation of the coagulation cascade (FIXa, FVIIa, circulating TF) are at an early stage of development. Additional new agents in clinical development include NAPc2, protein C derivatives, and soluble thrombomodulin.

References 1. Anderson, F.A., Jr., and Spencer, F.A. (2003) Risk factors for venous thromboembolism Circulation 107, I9–16. 2. Kearon, C. (2003) Natural history of venous thromboembolism Circulation 107, I22–30. 3. Warkentin, T.E., Levine, M.N., Hirsh, J., Horsewood, P., Roberts, R.S., Gent, M., and Kelton, J.G. (1995) Heparin-induced thrombocytopenia in patients treated with lowmolecular-weight heparin or unfractionated heparin N Engl J Med 332, 1330–5. 4. Spyropoulos, A.C. (2006) Managing oral anticoagulation requires expert experience and clinical evidence J Thromb Thromboly 21, 91–4.

5. Bates, S.M., and Ginsberg, J.S. (1997) Anticoagulants in pregnancy: fetal effects Baillieres Clin Obstet Gynaecol 11, 479–88. 6. Hutten, B.A., Prins, M.H., Gent, M., Ginsberg, J., Tijssen, J.G., and Buller, H.R. (2000) Incidence of recurrent thromboembolic and bleeding complications among patients with venous thromboembolism in relation to both malignancy and achieved international normalized ratio: a retrospective analysis J Clin Oncol 18, 3078–83. 7. Palareti, G., Legnani, C., Lee, A., Manotti, C., Hirsh, J., D’Angelo, A., Pengo, V., Moia, M., and Coccheri, S. (2000) A comparison of the safety and efficacy of oral anticoagulation

176

8.

9.

10.

11.

12. 13.

14.

15. 16. 17.

18.

19.

Mousa for the treatment of venous thromboembolic disease in patients with or without malignancy Thromb Haemost 84, 805–10. Bates, S., Greer, I., Hirsch, J., and Ginsberg, J. (2004) Use of antithrombotic agents during pregnancy. The seventh ACCP conference on antithrombotic and thrombolytic therapy Chest 126, 627S–44S. Greer, I.A., and Nelson-Piercy, C. (2005) Low-molecular-weight heparins for thromboprophylaxis and treatment of venous thromboembolism in pregnancy: a systematic review of safety and efficacy Blood 106, 401–7. Lee, A.Y., Levine, M.N., Baker, R.I., Bowden, C., Kakkar, A.K., Prins, M., Rickles, F.R., Julian, J.A., Haley, S., Kovacs, M.J., and Gent, M. (2003) Low-molecular-weight heparin versus a coumarin for the prevention of recurrent venous thromboembolism in patients with cancer N Engl J Med 349, 146–53. Weitz, J.I., Hirsh, J., and Samama, M. (2004) New anticoagulant drugs. The seventh ACCP conference on antithrombotic and thrombolytic therapy Chest 126, 265S– 86S. Dahl, O.E. (1999) Mechanisms of hypercoagulability Thromb Haemost 82, 902–6. Fernandez, P.M., Patierno, S.R., and Rickles, F.R. (2004) Tissue factor and fibrin in tumor angiogenesis Semin Thromb Hemost 30, 31–44. Girard, T.J., and Nicholson, N.S. (2001) The role of tissue factor/factor VIIa in the pathophysiology of acute thrombotic formation Curr Opin Pharmacol 1, 159–63. Hirsh, J., O’Donnell, M., and Weitz, J.I. (2005) New anticoagulants Blood 105, 453–63. Weitz, J.I., and Bates, S.M. (2005) New anticoagulants J Thromb Haemost 3, 1843–53. Mousa, S.A., and Mohamed, S. (2004) Inhibition of endothelial cell tube formation by the low molecular weight heparin, tinzaparin, is mediated by tissue factor pathway inhibitor Thromb Haemost 92, 627–33. Lee, A., Agnelli, G., Buller, H., Ginsberg, J., Heit, J., Rote, W., Vlasuk, G., Costantini, L., Julian, J., Comp, P., van Der Meer, J., Piovella, F., Raskob, G., and Gent, M. (2001) Dose-response study of recombinant factor VIIa/tissue factor inhibitor recombinant nematode anticoagulant protein c2 in prevention of postoperative venous thromboembolism in patients undergoing total knee replacement Circulation 104, 74–8. Giugliano, R.P., Wiviott, S.D., Stone, P.H., Simon, D.I., Schweiger, M.J., Bouchard,

20.

21.

22.

23.

24.

25.

26. 27.

28.

29.

A., Leesar, M.A., Goulder, M.A., Deitcher, S.R., McCabe, C.H., and Braunwald, E. (2007) Recombinant nematode anticoagulant protein c2 in patients with non-STsegment elevation acute coronary syndrome: the ANTHEM-TIMI-32 trial J Am Coll Cardiol 49, 2398–407. Moons, A.H., Peters, R.J., Bijsterveld, N.R., Piek, J.J., Prins, M.H., Vlasuk, G.P., Rote, W.E., and Buller, H.R. (2003) Recombinant nematode anticoagulant protein c2, an inhibitor of the tissue factor/factor VIIa complex, in patients undergoing elective coronary angioplasty J Am Coll Cardiol 41, 2147–53. Edgington, T.S., Mackman, N., Brand, K., and Ruf, W. (1991) The structural biology of expression and function of tissue factor Thromb Haemost 66, 67–79. Arnold, C.S., Parker, C., Upshaw, R., Prydz, H., Chand, P., Kotian, P., Bantia, S., and Babu, Y.S. (2006) The antithrombotic and anti-inflammatory effects of BCX-3607, a small molecule tissue factor/factor VIIa inhibitor Thromb Res 117, 343–9. Lazarus, R.A., Olivero, A.G., Eigenbrot, C., and Kirchhofer, D. (2004) Inhibitors of tissue factor. Factor VIIa for anticoagulant therapy Curr Med Chem 11, 2275–90. Buckman, B.O., Chou, Y.L., McCarrick, M., Liang, A., Lentz, D., Mohan, R., Morrissey, M.M., Shaw, K.J., Trinh, L., and Light, D.R. (2005) Solid-phase synthesis of naphthylamidines as factor VIIa/tissue factor inhibitors Bioorg Med Chem Lett 15, 2249–52. Suleymanov, O.D., Szalony, J.A., Salyers, A.K., LaChance, R.M., Parlow, J.J., South, M.S., Wood, R.S., and Nicholson, N.S. (2003) Pharmacological interruption of acute thrombus formation with minimal hemorrhagic complications by a small molecule tissue factor/factor VIIa inhibitor: comparison to factor Xa and thrombin inhibition in a nonhuman primate thrombosis model J Pharmacol Exp Ther 306, 1115–21. Bajaj, M.S., and Bajaj, S.P. (1997) Tissue factor pathway inhibitor: potential therapeutic applications Thromb Haemost 78, 471–7. Mousa, S.A., Fareed, J., Iqbal, O., and Kaiser, B. (2004) Tissue factor pathway inhibitor in thrombosis and beyond Methods Mol Med 93, 133–55. Sandset, P.M., Abildgaard, U., and Larsen, M.L. (1988) Heparin induces release of extrinsic coagulation pathway inhibitor (EPI) Thromb Res 50, 803–13. Amirkhosravi, A., Meyer, T., Amaya, M., Davila, M., Mousa, S.A., Robson, T., and

Novel Anticoagulant Therapy: Principle and Practice

30.

31.

32.

33.

34.

35.

36.

37.

38.

Francis, J.L. (2007) The role of tissue factor pathway inhibitor in tumor growth and metastasis Semin Thromb Hemost 33, 643–52. Mousa, S.A., Linhardt, R., Francis, J.L., and Amirkhosravi, A. (2006) Anti-metastatic effect of a non-anticoagulant low-molecularweight heparin versus the standard lowmolecular-weight heparin, enoxaparin Thromb Haemost 96, 816–21. Sierko, E., Wojtukiewicz, M.Z., and Kisiel, W. (2007) The role of tissue factor pathway inhibitor-2 in cancer biology Semin Thromb Hemost 33, 653–9. Biemond, B.J., Friederich, P.W., Levi, M., Vlasuk, G.P., Buller, H.R., and ten Cate, J.W. (1996) Comparison of sustained antithrombotic effects of inhibitors of thrombin and factor Xa in experimental thrombosis Circulation 93, 153–60. Buller, H.R., Davidson, B.L., Decousus, H., Gallus, A., Gent, M., Piovella, F., Prins, M.H., Raskob, G., Segers, A.E., Cariou, R., Leeuwenkamp, O., and Lensing, A.W. (2004) Fondaparinux or enoxaparin for the initial treatment of symptomatic deep venous thrombosis: a randomized trial Ann Intern Med 140, 867–73. Buller, H.R., Davidson, B.L., Decousus, H., Gallus, A., Gent, M., Piovella, F., Prins, M.H., Raskob, G., van den Berg-Segers, A.E., Cariou, R., Leeuwenkamp, O., and Lensing, A.W. (2003) Subcutaneous fondaparinux versus intravenous unfractionated heparin in the initial treatment of pulmonary embolism N Engl J Med 349, 1695–702. Herbert, J.M., Herault, J.P., Bernat, A., van Amsterdam, R.G., Lormeau, J.C., Petitou, M., van Boeckel, C., Hoffmann, P., and Meuleman, D.G. (1998) Biochemical and pharmacological properties of SANORG 34006, a potent and long-acting synthetic pentasaccharide Blood 91, 4197–205. Reiter, M., Bucek, R.A., Koca, N., Heger, J., and Minar, E. (2003) Idraparinux and liver enzymes: observations from the PERSIST trial Blood Coagul Fibrinolysis 14, 61–5. Buller, H.R., Cohen, A.T., Davidson, B., Decousus, H., Gallus, A.S., Gent, M., Pillion, G., Piovella, F., Prins, M.H., and Raskob, G.E. (2007) Idraparinux versus standard therapy for venous thromboembolic disease N Engl J Med 357, 1094–104. Alexander, J.H., Dyke, C.K., Yang, H., Becker, R.C., Hasselblad, V., Zillman, L.A., Kleiman, N.S., Hochman, J.S., Berger, P.B., Cohen, E.A., Lincoff, A.M., Saint-Jacques, H., Chetcuti, S., Burton, J.R., Buergler, J.M., Spence, F.P., Shimoto, Y., Robertson,

39.

40.

41.

42.

43.

44.

45.

46.

177

T.L., Kunitada, S., Bovill, E.G., Armstrong, P.W., and Harrington, R.A. (2004) Initial experience with factor-Xa inhibition in percutaneous coronary intervention: the XaNADU-PCI Pilot J Thromb Haemost 2, 234–41. Alexander, J.H., Yang, H., Becker, R.C., Kodama, K., Goodman, S., Dyke, C.K., Kleiman, N.S., Hochman, J.S., Berger, P.B., Cohen, E.A., Lincoff, A.M., Burton, J.R., Bovill, E.G., Kawai, C., Armstrong, P.W., and Harrington, R.A. (2005) First experience with direct, selective factor Xa inhibition in patients with non-ST-elevation acute coronary syndromes: results of the XaNADUACS trial J Thromb Haemost 3, 439–47. Lassen, M., Davidson, B., Gallus, A., Pineo, G., Ansell, J., and Deitchman, D.B.a.A. (2003) A phase II randomized, doubleblind, five-arm, parallel group, dose–response study of a new oral directly-acting factor Xa inhibitor, razaxaban, foe the prevention of deep vein thrombosis in knee replacement surgery Blood 102, Abstract 41. Lassen, M.R., Davidson, B.L., Gallus, A., Pineo, G., Ansell, J., and Deitchman, D. (2007) The efficacy and safety of apixaban, an oral, direct factor Xa inhibitor, as thromboprophylaxis in patients following total knee replacement J Thromb Haemost 5, 2368–75. Perzborn, E., Strassburger, J., Wilmen, A., Pohlmann, J., Roehrig, S., Schlemmer, K.H., and Straub, A. (2005) In vitro and in vivo studies of the novel antithrombotic agent BAY 59-7939–an oral, direct Factor Xa inhibitor J Thromb Haemost 3, 514–21. Kubitza, D., Becka, M., Voith, B., Zuehlsdorf, M., and Wensing, G. (2005) Safety, pharmacodynamics, and pharmacokinetics of single doses of BAY 59-7939, an oral, direct factor Xa inhibitor Clin Pharmacol Ther 78, 412–21. Eriksson, B.I., Borris, L., Dahl, O.E., Haas, S., Huisman, M.V., Kakkar, A.K., Misselwitz, F., and Kalebo, P. (2006) Oral, direct Factor Xa inhibition with BAY 59-7939 for the prevention of venous thromboembolism after total hip replacement J Thromb Haemost 4, 121–8. Turpie, A.G., Fisher, W.D., Bauer, K.A., Kwong, L.M., Irwin, M.W., Kalebo, P., Misselwitz, F., and Gent, M. (2005) BAY 597939: an oral, direct factor Xa inhibitor for the prevention of venous thromboembolism in patients after total knee replacement. A phase II dose-ranging study J Thromb Haemost 3, 2479–86. Laux, V., Perzborn, E., Kubitza, D., and Misselwitz, F. (2007) Preclinical and clinical

178

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

Mousa characteristics of rivaroxaban: a novel, oral, direct factor Xa inhibitor Semin Thromb Hemost 33, 515–23. Claxton, A.J., Cramer, J., and Pierce, C. (2001) A systematic review of the associations between dose regimens and medication compliance Clin Ther 23, 1296–310. Richter, A., Anton, S.E., Koch, P., and Dennett, S.L. (2003) The impact of reducing dose frequency on health outcomes Clin Ther 25, 2307–35, Discussion 6. Jacobson, B.C., Ferris, T.G., Shea, T.L., Mahlis, E.M., Lee, T.H., and Wang, T.C. (2003) Who is using chronic acid suppression therapy and why? Am J Gastroenterol 98, 51–8. Eriksson, B.I., Turpie, A.G., Lassen, M.R., Prins, M.H., Agnelli, G., Kalebo, P., Gaillard, M.L., and Meems, L. (2007) A dose escalation study of YM150, an oral direct factor Xa inhibitor, in the prevention of venous thromboembolism in elective primary hip replacement surgery J Thromb Haemost 5, 1660–5. Eriksson, B.I., Turpie, A., Lassen, M., Prins, M., Agnelli, G., Gaillard, M., and Meems, B. (2005) YM150, an oral direct factor Xa inhibitor, as prophylaxis for venous thromboembolism in patients with elective primary hip replacement surgery. A dose escalation study Bood 106, 1865. Fukuda, F., Honda, Y., Matsumoto, C., Sugiyama, N., Matsushita, T., Yanada, M., Morishima, Y., and Shibano, T. (2005) Impact of antithrombin deficiency on efficiencies of DU-176b, a novel orally active direct factor Xa inhibitor, and antithrombin dependent anticoagulants, fondaparinux, and heparin Blood 106, 1874. Agnelli, G., Haas, S., Krueger, K., Bedding, A., and Brandt, J. (2005) A phase II study of the safety and efficacy of a novel oral fXa inhibitor (LY517717) for the prevention of venous thromboembolism following TKR or THR Blood 106, 85a. Rothlein, R., Shen, J., Naser, N., Gohimukkula, D., Caligan, T., Andrews, R., Schmidt, A., Rose, E., and Mjalli, A. (2005) TTP889, a novel orally active partial inhibitor of FIXa inhibits clotting in two A/V shunt models without prolonged bleeding Blood 106, 1886. Schumacher, W.A., Seiler, S.E., Steinbacher, T.E., Stewart, A.B., Bostwick, J.S., Hartl, K.S., Liu, E.C., and Ogletree, M.L. (2007) Antithrombotic and hemostatic effects of a small molecule factor XIa inhibitor in rats Eur J Pharmacol 570, 167–74. Lundblad, R.L., Bradshaw, R.A., Gabriel, D., Ortel, T.L., Lawson, J., and Mann, K.G.

57.

58.

59.

60.

61. 62.

63.

64.

65.

(2004) A review of the therapeutic uses of thrombin Thromb Haemost 91, 851–60. Baughman, R.A., Kapoor, S.C., Agarwal, R.K., Kisicki, J., Catella-Lawson, F., and FitzGerald, G.A. (1998) Oral delivery of anticoagulant doses of heparin. A randomized, double-blind, controlled study in humans Circulation 98, 1610–5. Hull, D., Kakkar, A., Marder, V., Pineo, G.F., Goldberg, M., and Raskob, G. (2001) PROTECT trial: oral SNAC-heparin vs enoxaparin for preventing venous thromboembolism following total hip replacement. Blood 100, 148–9a. Kim, S.K., Lee, D.Y., Lee, E., Lee, Y.K., Kim, C.Y., Moon, H.T., and Byun, Y. (2007) Absorption study of deoxycholic acid-heparin conjugate as a new form of oral anticoagulant J Control Release 120, 4–10. Mousa, S.A., Zhang, F., Aljada, A., Chaturvedi, S., Takieddin, M., Zhang, H., Chi, L., Castelli, M.C., Friedman, K., Goldberg, M.M., and Linhardt, R.J. (2007) Pharmacokinetics and pharmacodynamics of oral heparin solid dosage form in healthy human subjects J Clin Pharmacol 47, 1508–20. Kelton, J.G. (2005) The pathophysiology of heparin-induced thrombocytopenia: biological basis for treatment Chest 127, 9–20S. Weitz, J.I., Hudoba, M., Massel, D., Maraganore, J., and Hirsh, J. (1990) Clotbound thrombin is protected from inhibition by heparin-antithrombin III but is susceptible to inactivation by antithrombin IIIindependent inhibitors J Clin Invest 86, 385–91. Mattsson, C., Menschik-Lundin, A., Nylander, S., Gyzander, E., and Deinum, J. (2001) Effect of different types of thrombin inhibitors on thrombin/thrombomodulin modulated activation of protein C in vitro Thromb Res 104, 475–86. Eriksson, B.I., Agnelli, G., Cohen, A.T., Dahl, O.E., Lassen, M.R., Mouret, P., Rosencher, N., Kalebo, P., Panfilov, S., Eskilson, C., Andersson, M., and Freij, A. (2003) The direct thrombin inhibitor melagatran followed by oral ximelagatran compared with enoxaparin for the prevention of venous thromboembolism after total hip or knee replacement: the EXPRESS study J Thromb Haemost 1, 2490–6. Eriksson, B.I., Agnelli, G., Cohen, A.T., Dahl, O.E., Mouret, P., Rosencher, N., Eskilson, C., Nylander, I., Frison, L., and Ogren, M. (2003) Direct thrombin inhibitor melagatran followed by oral ximelagatran in comparison with enoxaparin for prevention of venous thromboembolism after total hip

Novel Anticoagulant Therapy: Principle and Practice

66.

67.

68.

69.

or knee replacement Thromb Haemost 89, 288–96. Fiessinger, J.N., Huisman, M.V., Davidson, B.L., Bounameaux, H., Francis, C.W., Eriksson, H., Lundstrom, T., Berkowitz, S.D., Nystrom, P., Thorsen, M., and Ginsberg, J.S. (2005) Ximelagatran vs low-molecularweight heparin and warfarin for the treatment of deep vein thrombosis: a randomized trial JAMA 293, 681–9. Francis, C.W., Berkowitz, S.D., Comp, P.C., Lieberman, J.R., Ginsberg, J.S., Paiement, G., Peters, G.R., Roth, A.W., McElhattan, J., and Colwell, C.W., Jr. (2003) Comparison of ximelagatran with warfarin for the prevention of venous thromboembolism after total knee replacement N Engl J Med 349, 1703–12. Francis, C.W., Davidson, B.L., Berkowitz, S.D., Lotke, P.A., Ginsberg, J.S., Lieberman, J.R., Webster, A.K., Whipple, J.P., Peters, G.R., and Colwell, C.W., Jr. (2002) Ximelagatran versus warfarin for the prevention of venous thromboembolism after total knee arthroplasty. A randomized, double-blind trial Ann Intern Med 137, 648–55. Schulman, S., Wahlander, K., Lundstrom, T., Clason, S.B., and Eriksson, H. (2003) Secondary prevention of venous thromboembolism with the oral direct thrombin

70.

71.

72.

73.

179

inhibitor ximelagatran N Engl J Med 349, 1713–21. Stangier, J., Rathgen, K., Gansser, D., Kohlbrenner, V., and Stassen, J. (2001) Pharmacokinetics of BIBR 953 ZW, a novel low molecular weight direct thrombin inhibitor in healthy volunteers. Abstract Thromb Haemost 86, OC2347. Eriksson, B.I., Dahl, O.E., Buller, H.R., Hettiarachchi, R., Rosencher, N., Bravo, M.L., Ahnfelt, L., Piovella, F., Stangier, J., Kalebo, P., and Reilly, P. (2005) A new oral direct thrombin inhibitor, dabigatran etexilate, compared with enoxaparin for prevention of thromboembolic events following total hip or knee replacement: the BISTRO II randomized trial J Thromb Haemost 3, 103–11. Eriksson, B.I., Dahl, O.E., Rosencher, N., Kurth, A.A., van Dijk, C.N., Frostick, S.P., Prins, M.H., Hettiarachchi, R., Hantel, S., Schnee, J., and Buller, H.R. (2007) Dabigatran etexilate versus enoxaparin for prevention of venous thromboembolism after total hip replacement: a randomised, doubleblind, non-inferiority trial Lancet 370, 949–56. Coombe, S., Allen, G., and Kennedy, A. (2005) A phase I double-blind, ascending dose study of an oral synthetic direct thrombin inhibitor, TGN167 Blood 106, 530a.

Chapter 6 Oral Direct Factor Xa Inhibitors, with Special Emphasis on Rivaroxaban Shaker A. Mousa Abstract Rivaroxaban is a small-molecule, direct factor Xa inhibitor that is under investigation for the prevention and treatment of venous and arterial thrombosis. To date, oral anticoagulants have been limited largely to vitamin K antagonists. Despite their remarkable benefits, vitamin K antagonists are limited by their narrow therapeutic window, the existence of multiple food and drug interactions, and the need for frequent monitoring and dose-adjustment. Rivaroxaban represents a potentially attractive alternative to warfarin, as it could enable simplified once-daily dosing, requires no therapeutic monitoring, and has a lower potential for drug interactions. At present, the safety and efficacy of rivaroxaban for the prophylaxis and treatment of venous thromboembolism has been evaluated in phase-II and phase-III trials involving over 24,000 patients. Rivaroxaban is also being evaluated for the treatment of pulmonary embolism, secondary prevention after acute coronary syndromes, and the prevention of stroke and non-central nervous system embolism in patients with non-valvular atrial fibrillation. The need for new oral anticoagulants, the development and pharmacology of rivaroxaban, results of completed studies of rivaroxaban, and details of ongoing phase-II and phase-III trials with rivaroxaban are the subjects of this chapter. Key words: Oral anticoagulant, rivaroxaban, BAY 59-7939, factor Xa inhibitor, venous thromboembolism.

1. Introduction Arterial thrombosis, venous thrombosis, and subsequent thromboembolism account for significant morbidity and mortality worldwide. In the United States, more than 200,000 patients develop venous thromboembolism (VTE) every year, and 30% of these patients die within 30 days (1). Despite administration of current prophylaxis, 5–20% of all hip replacement surgeries S.A. Mousa (ed.), Anticoagulants, Antiplatelets, and Thrombolytics, Methods in Molecular Biology 663, DOI 10.1007/978-1-60761-803-4_6, © Springer Science+Business Media, LLC 2003, 2010

181

182

Mousa

are complicated by VTE (2). Furthermore, deep venous thrombosis and pulmonary embolism are associated with a significant economic burden due to the costs of acute care as well as longterm costs associated with recurrent VTE and post-thrombotic syndrome (3). A problem growing at a faster rate is the increasing burden imposed by atrial fibrillation(AF)-associated thromboembolism. AF leads to stasis of blood in the atria and formation of thrombi that can leave the heart and embolize any vascular bed, most seriously the cerebral circulation, leading to stroke. At present, 2.3 million Americans have AF, including 10% of all patients 80 years and older. By 2050, an estimated 5.6 million Americans will have AF, increasing the risk of stroke fivefold, and ultimately accounting for 15–20% of all strokes in the United States (4, 5). Furthermore, patients who have an AF-associated stroke are twice as likely to remain bedridden as other stroke victims (6). While there are several efficacious intravenous and subcutaneous alternatives for acute anticoagulation with a favorable balance of effectiveness and safety in the setting of acute coronary syndromes, deep venous thrombosis, or pulmonary embolism, there are few medications available for chronic, oral anticoagulation in AF. Oral anticoagulation thus far has been limited to vitamin K antagonists (VKAs), principally warfarin sodium. 1.1. Pharmacology of Warfarin

Warfarin, first described by Karl Paul Link in 1940, is the current treatment of choice for the prevention of thromboembolism in patients with AF (7, 8). By interfering with the cyclic interconversion of vitamin K and vitamin K 2,3-epoxide, warfarin impairs the γ-carboxylation of vitamin K-dependent proteins, including important serine proteases in the coagulation cascade that require vitamin K for their biologic activity (Factors II, VII, IX, and X). Warfarin is highly effective in preventing thromboembolic events in patients with AF. In 29 randomized trials involving more than 28,000 patients, warfarin reduced the risk of stroke by 64% (8). Furthermore, warfarin was associated with a 26% (95% CI, 3– 43%) reduction in all-cause mortality in randomized controlled trials when compared to no anticoagulation therapy in patients with AF (8). However, the benefits of warfarin in patients well controlled with the agent might overestimate benefits compared to the effects seen in warfarin-naive patients (9). Despite its effectiveness in preventing stroke and non-central nervous system embolism, warfarin has significant limitations. It has a slow onset of action, narrow therapeutic window, and requires frequent monitoring due to marked inter-individual variation in drug metabolism, as well as multiple drug and food interactions. In fact, patients taking warfarin spend nearly a third of the time outside their target INR window (10). Unfortunately, patients who spend more than 10% of their time outside of their target INR window are more likely to suffer an ischemic stroke and

Oral Direct Factor Xa Inhibitors

183

mortality (10). Patients with the highest risk of stroke, i.e., elderly individuals with evidence of heart failure, hypertension, diabetes, or prior stroke, have the highest rates of major hemorrhage, with a discontinuation rate up to 26% in the first year of treatment (11). This is likely due to difficulty in maintaining patients within a therapeutic window. Warfarin is a racemic compound, in which the S-warfarin enantiomer is a fivefold more potent vitamin K antagonist than the R-enantiomer (12). Each enantiomer exhibits unique interactions depending on an individual’s genetic background and concomitant medications. Polymorphisms in cytochrome P450 C29 (the enzyme responsible for metabolism of the S-isomer) have been associated with impaired hydroxylation of S-warfarin, leading to low-warfarin dose requirements (13). S-warfarin clearance is preferentially impaired by metronidazole and trimethoprim– sulfamethoxazole, while other medications such as amiodarone inhibit both R- and S-warfarin clearance (14). As a result, predicting the anticoagulation response with drug interactions is extremely complex. Warfarin dosing is further complicated by dramatic interindividual variability in warfarin metabolism. While the mean daily dose is 4.57 mg, more than 5% of patients require a daily dose greater than 10 mg (15). Polymorphisms in the vitamin K receptor gene (VKORC1) have been associated with elevated warfarin dose requirements (16). In addition to difficulties with dosage and frequent and expensive monitoring, warfarin is also associated with significant adverse effects. Long-term warfarin use is associated with a 25% increased risk of osteoporotic fractures (17). Patients on warfarin are at increased risk of life-threatening bleeding, including intracranial hemorrhage, which affected 1.8% of patients older than 75 in the Stroke Prevention in Atrial Fibrillation II (SPAF II) study (18). It has also been observed that the risk of bleeding is highest in the first year of warfarin therapy (9). Finally, as with most therapeutics, there is the following risk/benefit paradox with respect to warfarin-associated bleeding: patients at highest risk of stroke, according to both age and comorbidities, are also the patients with the highest risk of major, life-threatening hemorrhage (11, 19). 1.2. Investigating Potential Alternatives to Warfarin

Due to the limitations and risks associated with warfarin, less than half of eligible patients are ultimately treated with warfarin for stroke prophylaxis (20, 21). Despite these limitations, warfarin has remained the therapy of choice for prevention of thromboembolism in patients with AF since it became clinically available in 1954. Recently, randomized trials have investigated alternatives to warfarin therapy. The Stroke Prevention using an Oral Thrombin Inhibitor in Atrial Fibrillation (SPORTIF) trials compared the use of ximelagatran, a competitive inhibitor of human α-thrombin,

184

Mousa

with dose-adjusted warfarin (22, 23). At the time, ximelagatran represented a favorable alternative to warfarin as it had a predictable anticoagulant profile with rapid oral absorption and a rapid onset of action. Despite promising early results, a review of 6,948 patients treated with ximelagatran revealed transient elevations in alanine aminotransferase greater than 3 times the normal upper limit (ALT > 3×ULN) in 7.9% of patients treated with ximelagatran versus 1.2% in the comparator group (24). In addition to three cases of fatal hepatotoxicity, the clinical trial results also showed a troubling trend toward increased nonfatal myocardial infarction in patients treated with ximelagatran, especially in those in whom it was recently discontinued. As a result, the US Food and Drug Administration did not approve ximelagatran for the prevention of stroke and non-central nervous system (CNS) embolism in patients with AF (25). It was allowed on the market in Europe, but was rapidly pulled by the manufacturer when a fatal case of hepatic failure was attributed to the drug. While warfarin reduces the risk of stroke by 45% (95% CI, 29–57%) compared to antiplatelet monotherapy with aspirin, it is also associated with a 70% increased risk of bleeding (26). The ACTIVE W study was a randomized trial designed to determine whether dual antiplatelet therapy with aspirin and the thienophyridine clopidogrel was non-inferior to adjusted dose warfarin therapy (9). The trial was terminated early due to clear superiority of oral anticoagulation with warfarin. 1.3. Safer and More Efficacious Anticoagulation

The ideal oral anticoagulant would have a wide therapeutic window, rapid onset of action, minimal food and drug interactions, a short half-life allowing for quick termination in the event of bleeding, a readily available antidote or reversal agent, and clear efficacy in large trials without adverse effects. When considering candidates for potential new oral anticoagulants, attention must be paid to the three temporal aspects of hemostasis, including (1) initiation, (2) amplification, and (3) termination (27). The prototypical anticoagulant would target the amplification phase without interfering with initiation or termination, in order to allow some hemostasis in the event of tissue injury. Activated factor Xa (FXa) is central to the coagulation cascade and is the cornerstone of serine protease activity amplification. FX is a vitamin K-dependent serine protease synthesized in the liver that can be activated by either the intrinsic or the extrinsic clotting cascade. Binding of FXa to activated FV in the presence of calcium on a phospholipid bilayer results in formation of the prothrombinase complex. FXa catalyzes the conversion of prothrombin (Factor II) to thrombin (Factor IIa) and is the rate-limiting step in thrombin generation (Fig. 6.1) (28). One molecule of FXa can catalyze the formation of over a thousand molecules of thrombin (29). Thrombin potentiates clot formation by

Oral Direct Factor Xa Inhibitors

Intrinsic pathway

185

Extrinsic pathway

FXa

FXa

+ AT Prothrombin

Thrombin

FX

AT

Fibrin clot Fibrinogen

Fig. 6.1. Rivaroxaban selectively inhibits FXa. Schematic representation of the mechanism of inhibition of FXa by rivaroxaban

up-regulating its own production through feedback activation of factors V, VII, VIII, and XI, and inducing platelet activation (30, 31). Therefore, inhibition of FXa represents a potentially more efficacious anticoagulation strategy than targeting all of the vitamin K-dependent clotting factors. Inhibition of FXa can occur through direct binding to the FXa active site or indirectly through interaction with antithrombin. Direct FXa inhibitors have an advantage because they can bind both free FXa and FXa within the prothrombinase complex, and therefore penetrate the active thrombus to limit further thrombin generation.

2. Clinical Pharmacology of Rivaroxaban

Rivaroxaban is an oxazolidinone derivative that binds to the active site of FXa, leading to potent and selective inhibition of FXa (32). In animal models of both venous stasis and thrombosis, oral rivaroxaban inhibited FXa activity, leading to reduced thrombus formation and extension (33, 34). Rivaroxaban inhibits FXa activity in a dose-dependent manner, accompanied by prolongation of prothrombin time (PT) (Fig. 6.2). Phase-I data has shown that 15 mg of rivaroxaban decreases FXa activity by 35% and increases PT 1.4-fold over baseline values (35). The observed prolongation in PT correlated strongly with plasma rivaroxaban concentration (r = 0.935), with little inter-individual variability (36, 37). Thus, therapeutic monitoring might be possible through determination of PT when necessary. In phase-I studies in healthy male volunteers, a single 30-mg dose of rivaroxaban inhibited thrombin generation for greater than 24 h (38). Finally, rivaroxaban has no direct effect on platelet aggregation.

186

Mousa

40

PT Model

Prothrombin time (s)

30

20

10

0 0

100

200 300 400 Plasma concentration of BAY 59-7939 (µg/l)

500

600

Fig. 6.2. Prothrombin time correlates strongly with plasma concentration of rivaroxaban (r = 0.958). This figure reprinted with kind permission from Springer Science and Business Media (39).

2.1. Pharmacokinetics

The pharmacokinetic profile of rivaroxaban is consistent with rapid oral absorption and 80% bioavailability (36). The time to peak plasma concentration ranges from 2.5 to 4 h. After multiple doses, the drug half-life is 5–9 h in healthy volunteers, and 9–13 h in the elderly (mean age 65) (39). There are no major active circulating metabolites of rivaroxaban. A portion of the drug is excreted in the urine (1/3), and the remainder (2/3) is metabolized by the liver (36). Drug elimination demonstrates first-order kinetics and is impaired with advancing age, renal insufficiency, and in the presence of strong cytochrome P450 3A4 inhibitors (such as ketoconazole, macrolide antibiotics such as clarithromycin, and many protease inhibitors). In a phase-I study of patients with renal insufficiency, subjects with severe renal impairment (creatinine clearance 120 kg) demonstrated no change in peak serum concentrations in those >120 kg, but did show mild elevation (24%) in those ≤50 kg. This minor elevation was associated with a small (15%) increase in PT, which was not considered clinically significant (41). Therefore, no dose adjustment is required for sex or body weight when dosing rivaroxaban.

Oral Direct Factor Xa Inhibitors

187

While the available phase-I data have been largely limited to healthy Caucasian males, studies examining rivaroxaban in raceand ethnicity-specific populations are ongoing. 2.2. Pharmacodynamic Interactions

Rivaroxaban absorption is improved when taken with food; however, the drug can be administered on an empty stomach. The pharmacokinetics of the drug are unaffected by ranitidine or alteration of gastric pH with antacids in healthy male volunteers (42). Many patients who will require oral anticoagulation for AF or VTE prophylaxis often have risk factors for or documented coronary artery disease and require aspirin therapy. Rivaroxaban is being studied in patients with acute coronary syndromes, a patient population treated with aspirin in addition to other antiplatelet agents. In a randomized two-way crossover study (with healthy male subjects), antiplatelet therapy with aspirin did not alter the pharmacokinetics or pharmacodynamics of rivaroxaban (as determined by bleeding and PTs). Furthermore, platelet aggregometry studies were unaffected by rivaroxaban (43). A similar phase-I, two-way crossover study demonstrated no clinically significant interaction between naproxen and rivaroxaban in healthy subjects (35). Since rivaroxaban might have a role in stroke prophylaxis in those with AF, 20 mg of rivaroxaban was co-administered in 20 healthy male volunteers who also received 0.375 mg of digoxin. Drug exposure was not significantly different between patients receiving rivaroxaban alone and those who received rivaroxaban and digoxin. Based upon these results, there appears to be no apparent interaction between rivaroxaban and digoxin, suggesting that they can be prescribed together (44). Bridging with low-molecular weight heparins (LMWHs) is common in patients receiving chronic oral anticoagulants. When given together with rivaroxaban, enoxaparin resulted in additive inhibition of FXa activity and prolongation of bleeding times; however, coadministration of LMWH and rivaroxaban has been demonstrated (45). Overall, compared to the VKAs and other cardiovascular medications such as amiodarone, rivaroxaban has relatively low potential for substantial pharmacodynamic interactions, allowing for a wide range of concomitant pharmacotherapy.

2.3. Toxicity and Adverse Effects

Therapeutic anticoagulation always carries an attendant risk of bleeding, either due to errors in dosing and administration, occult pathology such as gastric ulceration, unrecognized bleeding diatheses, or urgent and emergency medical procedures. Therefore, there is great interest in and need for neutralizing agents in the event of significant bleeding. To address this concern, investigators explored the use of recombinant activated factor VII (rFVIIa) as a partial reversal agent for rivaroxaban. In a rat model of mesenteric hemorrhage, rFVIIa was administered after

188

Mousa

high-dose rivaroxaban (2 mg/kg). In the presence of high-dose, supratherapeutic rivaroxaban, 400 mcg/kg of rFVIIa reduced bleeding times by nearly 50% (46). rFVIIa also partially reversed the prolongation of PT and partially restored total thrombin activity, without affecting rivaroxaban-dependent FXa inhibition. Therefore, rFVIIa might be of use as an intravenous antidote for major bleeding in patients taking rivaroxaban. While not yet R investigated, factor VIII inhibitor bypassing activity (FEIBA ), a vapor-heated lyophilized powder for reconstitution, also represents a potential alternative for serious and/or life-threatening bleeding with rivaroxaban. In 1,102 patients, rivaroxaban did not affect electrocardiographic parameters, including the QTc (47). This is important, as many patients with AF who might be candidates for rivaroxaban therapy will be taking antiarrhythmic drugs that are known to prolong the QT interval, including dofetilide and sotalol. Given the hepatotoxicity observed with ximelagatran (even though ximelagatran is a direct thrombin inhibitor, not a FXa inhibitor), particular attention has been focused on liver function surveillance in patients receiving rivaroxaban. In the recent Oral Direct Factor Xa Inhibitor BAY 59-7939 in Patients with Acute Symptomatic Deep-Vein Thrombosis (ODIXa-DVT) trial, there was no evidence of hepatotoxicity with long-term (3 months) administration of rivaroxaban for treatment of deep venous thrombosis (DVT) (48). In the first 3 weeks of this trial, patients randomized to rivaroxaban had a lower incidence of ALT >3×ULN (1.9–4.3% versus 21.6%) as compared to those receiving enoxaparin. After 3 weeks however, the incidence of ALT >3×ULN was similar in both groups [1.9% (95% CI 0.8–3.6) versus 0.9% (95% CI 0.0–4.8)]. Rivaroxaban was stopped early in three patients due to abnormal liver function tests (two of these patients died, one from fulminant hepatitis B and one from carcinoma with hepatic metastasis) (48). In a pooled analysis of 1,343 patients randomized in phase-II studies of rivaroxaban for the prevention of post-operative VTE, there was no difference in the incidence of ALT >3×ULN between rivaroxaban and enoxaparin (3.8–6.0% versus 7.7%) (47). While there is no evidence of increased hepatotoxicity with rivaroxaban in multiple phase-II studies or in early reports of phase-III studies of VTE prophylaxis, more long-term data are needed.

3. Clinical Indications As evidenced by the significant clinical and economic burden imposed by VTE, including DVT and pulmonary embolism (PE), in both medical and surgical patients, the rising incidence of AF in

Oral Direct Factor Xa Inhibitors

189

the rapidly expanding elderly population, and the global impact of ischemic heart disease, there are many potential patient populations and indications for novel, safe, and effective oral anticoagulants such as rivaroxaban. Accordingly, a broad network of clinical trials has been designed to evaluate the safety and efficacy of rivaroxaban in patients at risk for arterial and venous thrombosis. In the sections to follow, we will review completed, ongoing, and planned clinical trials of rivaroxaban, according to clinical indication.

4. Post-operative Thromboprophylaxis

Proof of principle for the use of rivaroxaban in the prevention of VTE was first demonstrated in a phase-IIa study of 642 patients undergoing total hip replacement (ODIXa-HIP). In this open-label, dose-escalation (12-fold dose range) study, patients were randomized in a 3:1 ratio to rivaroxaban (2.5, 5, 10, 20, and 30 mg twice daily, or 30 mg once daily) or enoxaparin (40 mg daily) (49). Patients were treated until mandatory bilateral venography was performed 5–9 days after surgery. The primary efficacy endpoint (DVT, PE, or all-cause mortality) was not different in those treated with rivaroxaban compared to enoxaparin (10.2–23.8% vs. 16.8%). There was a dose-dependent reduction in major VTE (defined as proximal DVT, PE, VTE-related death, Table 6.1) with rivaroxaban, which was accompanied by a dose-dependent (0–10.8%) increase in the incidence of major post-operative bleeding (P2 g/dl fall in hemoglobin, or re-operation). The incidence of major bleeding ranged from 0.9 to 7% with rivaroxaban as compared to 1.7% with enoxaparin. This dose-dependent risk of bleeding remained significant, even after adjustment for age, gender, and study-specific bleeding rates. The majority of major bleeding episodes, as expected, were confined to the operative site. There was no significant difference in the risk of major bleeding between enoxaparin and rivaroxaban (total daily dose of 5–20 mg). Accordingly, this dose range was suggested as the optimal dosing for the prevention of VTE after orthopedic surgery. While ODIXa-KNEE and ODIXa-HIP2 compared twicedaily dosing of rivaroxaban, a randomized double-blind, doubledummy, active-comparator controlled trial of 873 patients examined once-daily rivaroxaban (over an eightfold dose range comprised of 5, 10, 20, 30, or 40 mg) compared to enoxaparin 40 mg daily for the prevention of VTE after total hip replacement (ODIXa-OD-HIP) (51). As in prior trials, patients were treated for 5–9 days after surgery and then underwent mandatory bilateral venography. In this once-daily dosing trial, there was a trend to significance in the dose-dependent reduction of VTE in patients treated with rivaroxaban. Perhaps of greater relevance was the statistically significant dose-dependent reduction in major VTE (proximal DVT, PE, or death). There was a dosedependent increase in the risk of major bleeding, however, the strength of the relationship was less than previously observed in the twice-daily dosing studies (2.3–5.1%, P = 0.039). Rivaroxaban was well tolerated in this trial, and no dose arm was stopped due to safety concerns. Based on the comparable efficacy between doses and the increase in major bleeding from 0.7 to 4.3% in the 10 mg as compared to 20 mg groups, the authors recommended 10 mg as the daily dose for future phase-III VTE prevention trials. All of these early trials led to the phase-III Regulation of Coagulation in Major Orthopaedic Surgery Deducing the Risk of

0.9

Major bleeding

1.3

2.4

2.1

3.1

∗ P-value for dosing trend; enox, enoxparin qd

3

10

30

7.0 70 years, or those given fibrinolytic therapy. Based on this study, it was concluded that routine use of clopidogrel plus aspirin in patients with acute MI safely reduces mortality and major vascular events in a large range of patients, including those aged >70 years.

2.2.4. The CHARISMA Study

The CHARISMA (Clopidogrel for High Artherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance) trial was a 28-month trial evaluating antiplatelet combination therapy versus aspirin monotherapy for both primary and secondary prevention of atherothrombotic events in 15,603 stable patients (41). All patients were categorized as high risk for atherothrombotic events and were randomized to clopidogrel 75 mg/day plus low-dose aspirin (75–162 mg/day) or low-dose aspirin plus placebo. The efficacy end point was a composite of MI, stroke, or death from cardiovascular disease. Patients were categorized into two subgroups: a symptomatic subgroup (12,153 patients), composed of those with documented cardiovascular disease (remote MI, stroke or symptomatic PAD) and an asymptomatic group (3,284 patients), who were enrolled with multiple atherothrombotic risk factors but without established atherosclerosis. The primary safety end point was any event of severe bleeding based on GUSTO (Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries) criteria. Overall, treatment was discontinued in 20.4% of patients in the clopidogrel group and in 18.2% of the placebo group (P< 0.001) due to adverse events. The primary end point occurred in 6.8% of the clopidogrel group and in 7.3% of the placebo group, which was not significant (P< 0.22). Severe bleeding occurred in 1.7% of the clopidogrel group and in 1.3% of the

Antiplatelet Therapies: Drug Interactions in the Management of Vascular Disorders

211

placebo group (P< 0.09). The relative risk reduction in the symptomatic group received clopidogrel (6.9%) versus that placebo (7.9%) was 0.88 (P = 0.046). The results of the CHARISMA study indicated that only in the selected symptomatic group was there a suggestion of risk reduction that outweighed the risk of bleeding, and in the asymptomatic group with multiple risk factors, there was a suggestion of harm. Clopidogrel plus aspirin was not more effective than aspirin alone in reducing the rate of death from MI, stroke, or death from cardiovascular disease in stable patients in this long-term study. 2.2.5. The CURE and PCI-CURE Studies

The CURE (Clopidogrel in Unstable Angina to Prevent Recurrent Events) trial evaluated the benefits and risks of clopidogrel plus aspirin versus aspirin alone in patients with non-ST elevation ACS (42). A total of 12,562 patients were included in the trial; 2,072 underwent CABG intervention, 2,658 underwent a PCI, and 2,658 were managed medically. Patients were included in the study if they had symptoms indicative of ACS within the preceding 24 h without ST-segment elevations, supporting evidence of ischemia from their most recent electrocardiogram, and elevated concentrations of cardiac enzymes (including troponin) that were at least twice the reference range. All patients received aspirin 75–325 mg/day and were then randomized to receive clopidogrel 75 mg/day with a loading of 300 mg or placebo for 3–12 months. The primary outcome was a composite of cardiovascular death, MI, or stroke among patients who underwent CABG, PCI, or medical therapy. Overall, 10.6% of patients in the clopidogrel group and 12.5% of patients in the placebo group experienced one of the primary outcomes. This occurred with a relative risk of 0.84 (P = 0.001). For patients who underwent CABG treatment, primary outcomes occurred in 16.2% of the placebo group versus 14.5% in the clopidogrel group. Benefits were seen mainly in those patients who had received combination therapy before the procedure. For patients undergoing CABG, there was no significant trend of lifethreatening bleeding with clopidogrel, the use of which was confined to within 5 days of CABG surgery. According to the CURE trial, the benefits of starting clopidogrel with aspirin in non-ST MI outweighed the hemorrhagic risk, even in patients treated by CABG. Clopidogrel-treated patients also experienced reduced in-hospital refractory ischemia, recurrent angina, and heart failure. The CURE trial concluded that clopidogrel is beneficial in ACS patients whether or not they undergo revascularization. With regard to maximizing the clinical benefit and minimizing the hemorrhagic risk associated with clopidogrel and CABG, the results of the CURE trial suggested initiating dual therapy upon presentation and stopping clopidogrel 5 days before the CABG procedure.

212

Mousa

The PCI-CURE (Percutaneous Coronary InterventionClopidogrel in Unstable angina to Prevent Recurrent Events) trial (42), published in 2001, was a sub-analysis of the CURE trial in which 2,658 patients who were included in the CURE study and had undergone PCI were studied. Primary outcomes were cardiovascular death, MI, or urgent revascularization within 30 days of the intervention. PCI patients were given clopidogrel or placebo in combination with aspirin for 4 weeks prior to surgery. They resumed drug treatment post-surgically and were assessed for long-term effects up to 1 year later. There was a significant difference in cardiovascular death and MI between the two groups: 12.6% in the placebo group and 8.8% in the clopidogrel group experienced primary outcomes (P = 0.002). PCI-CURE investigators concluded that compared to placebo, patients with non-ST elevation MI treated with clopidogrel plus aspirin and PCI had a reduced risk of cardiovascular death and MI by about a third. Long-term therapy was also associated with a lower rate of cardiovascular death, MI, or the need for revascularization (P = 0.03). There was no significant difference in major, but not life-threatening, bleeding with clopidogrel. The results of the PCI-CURE study supported the use of combination antiplatelet therapy with clopidogrel plus aspirin in patients undergoing PCI with ACS and non-ST-segment elevation MI. It was further suggested that long-term combination therapy is beneficial in reducing major cardiovascular events. 2.2.6. The MATCH Study

In the MATCH trial, 7,599 high-risk patients with cerebrovascular disease were treated with clopidogrel or clopidogrel plus aspirin and assessed for vascular events (43). Patients admitted into this trial were stable and receiving clopidogrel (75 mg/day), and were randomized to either placebo or low-dose aspirin (75 mg/day). Patients had to have experienced an ischemic stroke or transient ischemic attack (TIA) within the past 3 months, and they had to have one or more of the following four risk factors: previous MI, angina pectoris, diabetes mellitus, and symptoms of PAD within the past 3 years. The primary end point for the MATCH trial was a composite of ischemic stroke, MI, vascular death, and rehospitalization for an acute ischemic event (angina pectoris, worsening PAD symptoms, or TIA). Secondary outcomes included death from any type of stroke. The MATCH trial failed to show any significant difference in risk reduction between the clopidogrel plus aspirin group versus the clopidogrel plus placebo group. There was no statistical significance in any of the primary end points between the two arms that indicated a reduction in rates of ischemic stroke, MI, vascular death, or rehospitalization for any acute ischemic event. In addition, there was a significantly higher rate of major bleeding in the combination therapy group (P< 0.0001). The risk was

Antiplatelet Therapies: Drug Interactions in the Management of Vascular Disorders

213

1.36% higher in the combination arm as compared to the clopidogrel monotherapy arm. Thus, there was no significant reduction in vascular events by adding aspirin to clopidogrel for high-risk patients with ischemic stroke or TIA. 2.2.7. The CARESS Study

In the CARESS (Clopidogrel and Aspirin for Reduction of Emboli in Symptomatic Carotid Stenosis) trial, Markus et al. (44) evaluated combination therapy in cerebrovascular risk patients. All patients were recently symptomatic, with ≥50% carotid stenosis with micro-embolic signals (MES), as detected by transcranial Doppler ultrasound. Patients were randomized to clopidogrel (loading dose of 300 mg on day 1 followed by 75 mg/day) or placebo. All patients also received aspirin 75 mg/day for the duration of the study. Patients were all >18 years of age, had >50% carotid stenosis, and had experienced ipsilateral carotid territory TIA or stroke within the preceding 3 months. The concomitant use of anticoagulants, thrombolytic agents, analgesics (other than acetaminophen or opioids), and any additional antiplatelet agents was prohibited during the course of the study. The CARESS trial assessed the proportion of patients who were MES-positive on a 1-h recording conducted on day 7 of the trial. A reduction in the MES value would correspond to decreased markers of platelet and thrombus emboli in the ipsilateral middle cerebral artery, leading to a lower risk of recurrent strokes in this patient population. Safety end points were recorded as any adverse or cerebrovascular events, such as TIA, ischemic stroke, or cerebral hemorrhage. Bleeding events were divided into three categories: lifethreatening, major, or minor bleeding. Of the 107 randomized patients who had an MES value >1, 43.8% of patients in the combination group versus 72.7% in the placebo group had a positive MES reading on day 7, favoring combination therapy. There was a relative risk reduction of 39.8% for those patients who received clopidogrel plus aspirin versus aspirin alone. This was statistically significant (P = 0.0046), with no further increases in bleeding events. The CARESS study concluded that in “actively embolizing” patients with recently symptomatic carotid stenosis (>50%), the combination of clopidogrel plus aspirin therapy is more effective than aspirin alone in reducing asymptomatic embolization, with a relative risk reduction of 40% in 7 days. Active treatment response was seen with a similar magnitude of effect on day 2. These findings were in contrast to the results of the MATCH study; however, there are some differences to be noted between the two patient populations. The MATCH study data were based on all types of ischemic stroke, including small-vessel disease, which has the lowest risk of early recurrent stroke because it is not a process caused by embolism from an atherosclerotic plaque (45). This may be one of the reasons why interpretation of the

214

Mousa

MATCH trial data at 18 months is not consistent with that of the CARESS trial after 7 days. The MATCH trial also included patients who were experiencing events several weeks after their acute phase. This group of patients is at the highest risk for recurrent stroke. In contrast, the CARESS study was consistent with reducing recurrent stroke in patients with large-vessel atherosclerotic stroke in the acute phase. The difference in the power of these two studies should also be noted. The CARESS study analyzed 107 patients versus 7,276 patients analyzed in the MATCH trial. Upon direct comparison, the MATCH trial had a greater power value than the CARESS study, which could also influence the observed outcomes. 2.2.8. Summary: Combination Aspirin and Clopidogrel Therapy

There is considerable clinical evidence supporting the use of combination antiplatelet therapy with clopidogrel and aspirin. The CLARITY-TIMI and the COMMIT trials both studied patients suffering from acute MI with ST-segment elevations or newly developed left-bundle branch block (39, 40). These two trials demonstrated greater efficacy when patients were given clopidogrel plus aspirin versus aspirin alone. Death, reinfarction, and stroke were all reduced with the combination of antiplatelets, with no significant difference in major bleeding compared to aspirin alone. The CURE trial was conducted to investigate the reduction of risks in patients undergoing vascular interventions. Again, patients who received clopidogrel and aspirin before surgery experienced a significant reduction in the primary end points. These results were supported by the PCI-CURE study (42). Asymptomatic patients who were considered to be at high risk for atherothrombotic events did not benefit from combination antiplatelet therapy in the CHARISMA study (41). Clopidogrel plus aspirin is not recommended for the prevention of atherothrombotic events in stable patients who have multiple risk factors alone. The MATCH and CARESS trials raised much controversy due to their conflicting results. The CARESS trial observed patients who experienced ipsilateral carotid territory TIA/stroke only (43–45). Furthermore, observations in the MATCH trial were recorded up to 18 months, whereas the CARESS results were recorded after 7 days. One issue that deserves further discussion is the duration of therapy. There is conflicting evidence from the MATCH and CARESS trials as to the optimal duration of antiplatelet therapy in cerebrovascular disease. For coronary artery disease, the CHARISMA trial failed to show a benefit of long-term clopidogrel in the overall trial population, although the 80% of patients with clinically evident atherothrombosis experienced a modest reduction of the primary endpoint, and emerging data with drug-eluting stents suggest that dual antiplatelet therapy may

Antiplatelet Therapies: Drug Interactions in the Management of Vascular Disorders

215

be required even beyond 1 year. Clearly, additional studies are needed to evaluate optimal antiplatelet therapy combinations and duration of therapies to permit maximal benefit with the minimum of harm in patients with cardiovascular disease.

3. Expert Opinion and Future Directions

Platelet aggregation plays a key role in the pathogenesis of coronary thrombosis, and pharmacologic inhibition of platelet function forms the cornerstone of treatment for ACS. Cardiovascular treatment with clopidogrel versus aspirin has been shown to be beneficial, but with adequate precautions (46). When clopidogrel and aspirin were each studied in monotherapy, clopidogrel proved to be superior to aspirin in reducing the risk of ischemic events (47). When the two were combined in the first trial to show effectiveness versus aspirin monotherapy, the combination was associated with a high risk of bleeding (46). Only when investigators began to observe carefully specific patient populations did combination therapy begin to demonstrate promising results (e.g., the PCI-CURE study). Clopidogrel has been shown to be beneficial in the initial treatment and secondary prevention of ACS. Compared with aspirin alone, dual antiplatelet therapy with aspirin plus clopidogrel reduces the risk of vessel thrombosis and recurrent ischemic events in patients undergoing PCI and is a useful adjunct to coronary artery stenting. In CABG surgery, combination aspirin and clopidogrel therapy initiated immediately postoperatively improves bypass graft patency. The search for ADP receptor antagonists with a rapid onset and short half-life is ongoing. Such agents would allow for urgent surgical procedures and would overcome the current limitations of clopidogrel. Furthermore, antiplatelet agents that work via collagen receptors, serotonin receptors, or thrombin receptors may have additional value and the ability to complement current antiplatelet therapy. Several rapid-onset and rapid-offset reversible ADP antagonists are currently in clinical development (35, 48). AZD-6140 is a reversible oral P2Y12 receptor antagonist that has been studied in ACS patients in comparison to clopidogrel in the DISPERSE-2 (Dose Confirmation Study Assessing Antiplatelet Effects of AZD6140 versus Clopidogrel in Non-ST-Segment Elevation MI) study. AZD-6140 exhibited greater mean inhibition of platelet aggregation than a standard regimen of clopidogrel in ACS patients. In addition, AZD-6140 further suppressed platelet aggregation in clopidogrel-pretreated patients (48).

216

Mousa

One could extrapolate that reversible ADP receptor antagonists would have greater clinical benefits based on ex vivo studies showing that their ability to inhibit platelet aggregation is superior to other drugs such as clopidogrel. However, as demonstrated by the case of oral GPIIb/IIIa inhibitors, promising laboratory results do not necessarily translate into successful outcomes in the clinical setting. Despite consistent evidence of substantial inhibition of platelet aggregation by oral GPIIb/IIIa inhibitors, this class of drugs provided no clinical benefit in phase-III trials and, in fact, was harmful. The connection between ex vivo inhibition of platelet aggregation and clinical benefit of platelet P2Y12 antagonists is also not straightforward. A link between inflammatory status and clinical benefit from antiplatelet agents continues to emerge and highlights the fact that biomarkers beyond ex vivo platelet aggregation might better predict the clinical benefit of antiplatelet agents that reduce platelet activation. Ongoing trials hold promise for determining the appropriate targets for maximizing antiplatelet efficacy, but the current lack of a proven ex vivo assay that correlates with clinical outcomes hampers clinical investigation, particularly in light of the expense involved in conducting these mega trials. ADP receptor blockers require 3–7 days to reach maximum inhibition of platelet aggregation. When investigators began to assess the use of a loading dose with clopidogrel, therapeutic results occurred more rapidly (33, 44). There was a rapid onset of platelet aggregate inhibition, with an antithrombotic effect observed within 90 min. The optimal type of antiplatelet therapy for patients who will undergo surgery would be one with a rapid onset and a relatively short elimination half-life, allowing for a once- or twice-daily regimen. A significant impediment to the development and clinical application of antiplatelet therapies is the prohibitive cost of clinical trials and the potential risk involved in attaining clinical superiority and safety over existing regimens. The recruitment and study of a large number of patients for a prolonged period of time may be necessary to demonstrate a modest clinical benefit. Thus, the prohibitive cost of clinical trials demonstrating significant differences to current antiplatelet therapies might limit progress in advancing new antiplatelet targets.

References 1. Jamieson, D.G., Parekh, A., and Ezekowitz, M.D. (2005) Review of antiplatelet therapy in secondary prevention of cerebrovascular events: a need for direct compar-

isons between antiplatelet agents J Cardiovasc Pharmacol Ther 10, 153–61. 2. Collaboration, A.T. (1994) Collaborative overview of randomised trials of antiplatelet

Antiplatelet Therapies: Drug Interactions in the Management of Vascular Disorders

3.

4.

5.

6.

7.

8. 9. 10. 11.

12.

13.

14.

15.

therapy – I: prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. BMJ 308, 81–106. Fitzgerald, D.J., Roy, L., Catella, F., and FitzGerald, G.A. (1986) Platelet activation in unstable coronary disease N Eng J Med 315, 983–9. Fuster, V. (1981) Role of platelets in the development of atherosclerotic disease and possible interference with platelet inhibitor drugs Scand J Haematol Suppl 38, 1–38. Hamm, C.W., Lorenz, R.L., Bleifeld, W., Kupper, W., Wober, W., and Weber, P.C. (1987) Biochemical evidence of platelet activation in patients with persistent unstable angina J Am Coll Cardiol 10, 998–06. Mousa, S., and Bennett, J.S. (1996) Platelets in health and disease: platelet GPIIb/IIIa structure and function; recent advances in antiplatelet therapy Drugs Future 21, 1141–54. Mousa, S., and Guigliano, R. Antiplatelet therapies: platelet GPIIb/IIIa antagonists and beyond. In: Sasahara, A., ed. New Therapeutic Agents in Thrombosis. New York: Marcel Dekker;2003:341–7. Born, G.V., Gorog, P., and Begent, N.A. (1983) The biologic background to some therapeutic uses of aspirin Am J Med 74, 2–9. Mousa, S.A. (2003) Antiplatelet therapies: platelet GPIIb/IIIa antagonists and beyond Curr Pharm Des 9, 2317–22. Hart, R.G., and Harrison, M.J. (1996) Aspirin wars: the optimal dose of aspirin prevent stroke Stroke 27, 585–7. Hennekens, C.H., Sechenova, O., Hollar, D., and Serebruany, V.L. (2006) Dose of aspirin in the treatment and prevention of cardiovascular disease: current and future directions J Cardiovasc Pharmacol Ther 11, 170–6. Haynes, R.B., Sandler, R.S., Larson, E.B., Pater, J.L., and Yatsu, F.M. (1992) A critical appraisal of ticlopidine, a new antiplatelet agent. Effectiveness and clinical indications for prophylaxis of atherosclerotic events Arch Intern Med 152, 1376–80. Love, B.B., Biller, J., and Gent, M. (1998) Adverse haematological effects of ticlopidine. Prevention, recognition and management Drug Saf 19, 89–98. Creager, M.A. (1998) Results of the CAPRIE trial: efficacy and safety of clopidogrel. Clopidogrel versus aspirin in patients at risk of ischaemic events Vasc Med 3, 257–60. Moussa, I., Oetgen, M., Roubin, G., Colombo, A., Wang, X., Iyer, S., Maida, R., Collins, M., Kreps, E., and Moses,

16.

17.

18.

19.

20.

21.

22.

23.

24. 25.

26.

217

J.W. (1999) Effectiveness of clopidogrel and aspirin versus ticlopidine and aspirin in preventing stent thrombosis after coronary stent implantation Circulation 99, 2364–6. Hankey, G.J., Sudlow, C.L., and Dunbabin, D.W. (2000) Thienopyridine derivatives (ticlopidine, clopidogrel) versus aspirin for preventing stroke and other serious vascular events in high vascular risk patients Cochrane Database Syst Rev CD001246. Bednar, M.M., Quilley, J., Russell, S.R., Fuller, S.P., Booth, C., Howard, D., and Gross, C.E. (1996) The effect of oral antiplatelet agents on tissue plasminogen activator-mediated thrombolysis in a rabbit model of thromboembolic stroke Neurosurgery 39, 352–9. Nicolau, D.P., Tessier, P.R., and Nightingale, C.H. (1999) Beneficial effect of combination antiplatelet therapy on the development of experimental Staphylococcus aureus endocarditis Int J Antimicrob Agents 11, 159–61. Lenz, T.L., and Hilleman, D.E. (2000) Aggrenox: a fixed-dose combination of aspirin and dipyridamole Ann Pharmacother 34, 1283–90. Steinhubl, S.R., Berger, P.B., Mann, J.T., 3rd, Fry, E.T., DeLago, A., Wilmer, C., and Topol, E.J. (2002) Early and sustained dual oral antiplatelet therapy following percutaneous coronary intervention: a randomized controlled trial JAMA 288, 2411–20. Bertrand, M.E., Rupprecht, H.J., Urban, P., and Gershlick, A.H. (2000) Double-blind study of the safety of clopidogrel with and without a loading dose in combination with aspirin compared with ticlopidine in combination with aspirin after coronary stenting : the clopidogrel aspirin stent international cooperative study (CLASSICS) Circulation 102, 624–9. Mousa, S.A. (1999) Antiplatelet therapies: recent advances in the development of platelet glycoprotein IIb/IIIa antagonists Curr Interv Cardiol Rep 1, 243–52. Mousa, S.A. (1999) Antiplatelet therapies: from aspirin to GPIIb/IIIa-receptor antagonists and beyond Drug Discov Today 4, 552–61. Phillips, D.R., Charo, I.F., and Scarborough, R.M. (1991) GPIIb-IIIa: the responsive integrin Cell 65, 359–62. Pytela, R., Pierschbacher, M.D., Ginsberg, M.H., Plow, E.F., and Ruoslahti, E. (1986) Platelet membrane glycoprotein IIb/IIIa: member of a family of Arg-Gly-Asp – specific adhesion receptors Science 231, 1559–62. The EPIC Investigators (1994) Use of a monoclonal antibody directed against the

218

27.

28.

29.

30.

31.

32.

Mousa platelet glycoprotein IIb/IIIa receptor in high-risk coronary angioplasty. N Engl J Med 330, 956–61. Fitchett, D.H., Langer, A., Armstrong, P.W., Tan, M., Mendelsohn, A., and Goodman, S.G. (2006) Randomized evaluation of the efficacy of enoxaparin versus unfractionated heparin in high-risk patients with non-STsegment elevation acute coronary syndromes receiving the glycoprotein IIb/IIIa inhibitor eptifibatide. Long-term results of the Integrilin and Enoxaparin Randomized Assessment of Acute Coronary Syndrome Treatment (INTERACT) trial Am Heart J 151, 373–9. Hayes, R., Chesebro, J.H., Fuster, V., Dangas, G., Fallon, J.T., Sharma, S.K., Coller, B.S., Badimon, L., Marmur, J.D., and Badimon, J.J. (2000) Antithrombotic effects of abciximab Am J Cardiol 85, 1167–72. Januzzi, J.L., Jr., Snapinn, S.M., DiBattiste, P.M., Jang, I.K., and Theroux, P. (2002) Benefits and safety of tirofiban among acute coronary syndrome patients with mild to moderate renal insufficiency: results from the Platelet Receptor Inhibition in Ischemic Syndrome Management in Patients Limited by Unstable Signs and Symptoms (PRISMPLUS) trial Circulation 105, 2361–6. Lincoff, A.M., Califf, R.M., Van de Werf, F., Willerson, J.T., White, H.D., Armstrong, P.W., Guetta, V., Gibler, W.B., Hochman, J.S., Bode, C., Vahanian, A., Steg, P.G., Ardissino, D., Savonitto, S., Bar, F., Sadowski, Z., Betriu, A., Booth, J.E., Wolski, K., Waller, M., and Topol, E.J. (2002) Mortality at 1 year with combination platelet glycoprotein IIb/IIIa inhibition and reduceddose fibrinolytic therapy vs conventional \hbox{fibrinolytic} therapy for acute myocardial infarction: GUSTO V randomized trial JAMA 288, 2130–5. Neumann, F.J., Kastrati, A., Schmitt, C., Blasini, R., Hadamitzky, M., Mehilli, J., Gawaz, M., Schleef, M., Seyfarth, M., Dirschinger, J., and Schomig, A. (2000) Effect of glycoprotein IIb/IIIa receptor blockade with abciximab on clinical and angiographic restenosis rate after the placement of coronary stents following acute myocardial infarction J Am Coll Cardiol 35, 915–21. Anderson, J.L., Adams, C.D., Antman, E.M., Bridges, C.R., Califf, R.M., Casey, D.E., Jr., Chavey, W.E., 2nd, Fesmire, F.M., Hochman, J.S., Levin, T.N., Lincoff, A.M., Peterson, E.D., Theroux, P., Wenger, N.K., Wright, R.S., Smith, S.C., Jr., Jacobs, A.K., Halperin, J.L., Hunt, S.A., Krumholz, H.M.,

33.

34.

35.

36.

37.

Kushner, F.G., Lytle, B.W., Nishimura, R., Ornato, J.P., Page, R.L., and Riegel, B. (2007) ACC/AHA 2007 guidelines for the management of patients with unstable angina/non-ST-elevation myocardial infarction: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 2002 Guidelines for the Management of Patients With Unstable Angina/Non-ST-Elevation Myocardial Infarction) developed in collaboration with the American College of Emergency Physicians, the Society for Cardiovascular Angiography and Interventions, and the Society of Thoracic Surgeons endorsed by the American Association of Cardiovascular and Pulmonary Rehabilitation and the Society for Academic Emergency Medicine J Am Coll Cardiol 50, e1–157. Cadroy, Y., Bossavy, J.P., Thalamas, C., Sagnard, L., Sakariassen, K., and Boneu, B. (2000) Early potent antithrombotic effect with combined aspirin and a loading dose of clopidogrel on experimental arterial thrombogenesis in humans Circulation 101, 2823–8. Coma-Canella, I., Velasco, A., and Castano, S. (2005) Prevalence of aspirin resistance measured by PFA-100 Int J Cardiol 101, 71–6. Wiviott, S.D., Braunwald, E., McCabe, C.H., Montalescot, G., Ruzyllo, W., Gottlieb, S., Neumann, F.J., Ardissino, D., De Servi, S., Murphy, S.A., Riesmeyer, J., Weerakkody, G., Gibson, C.M., and Antman, E.M. (2007) Prasugrel versus clopidogrel in patients with acute coronary syndromes N Eng J Med 357, 2001–15. Angiolillo, D.J., Shoemaker, S.B., Desai, B., Yuan, H., Charlton, R.K., Bernardo, E., Zenni, M.M., Guzman, L.A., Bass, T.A., and Costa, M.A. (2007) Randomized comparison of a high clopidogrel maintenance dose in patients with diabetes mellitus and coronary artery disease: results of the Optimizing Antiplatelet Therapy in Diabetes Mellitus (OPTIMUS) study Circulation 115, 708–16. Spertus, J.A., Kettelkamp, R., Vance, C., Decker, C., Jones, P.G., Rumsfeld, J.S., Messenger, J.C., Khanal, S., Peterson, E.D., Bach, R.G., Krumholz, H.M., and Cohen, D.J. (2006) Prevalence, predictors, and outcomes of premature discontinuation of thienopyridine therapy after drug-eluting stent placement: results from the PREMIER registry Circulation 113, 2803–9.

Antiplatelet Therapies: Drug Interactions in the Management of Vascular Disorders 38. Smith, S.C., Jr., Feldman, T.E., Hirshfeld, J.W., Jr., Jacobs, A.K., Kern, M.J., King, S.B., 3rd, Morrison, D.A., O Neill, W.W., Schaff, H.V., Whitlow, P.L., Williams, D.O., Antman, E.M., Adams, C.D., Anderson, J.L., Faxon, D.P., Fuster, V., Halperin, J.L., Hiratzka, L.F., Hunt, S.A., Nishimura, R., Ornato, J.P., Page, R.L., and Riegel, B. (2006) ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/SCAI Writing Committee to Update the 2001 Guidelines for Percutaneous Coronary Intervention) J Am Coll Cardiol 47, e1–121. 39. Sabatine, M.S., Cannon, C.P., Gibson, C.M., Lopez-Sendon, J.L., Montalescot, G., Theroux, P., Claeys, M.J., Cools, F., Hill, K.A., Skene, A.M., McCabe, C.H., and Braunwald, E. (2005) Addition of clopidogrel to aspirin and fibrinolytic therapy for myocardial infarction with ST-segment elevation N Eng J Med 352, 1179–89. 40. Chen, Z.M., Jiang, L.X., Chen, Y.P., Xie, J.X., Pan, H.C., Peto, R., Collins, R., and Liu, L.S. (2005) Addition of clopidogrel to aspirin in 45,852 patients with acute myocardial infarction: randomised placebocontrolled trial Lancet 366, 1607–21. 41. Bhatt, D.L., Fox, K.A., Hacke, W., Berger, P.B., Black, H.R., Boden, W.E., Cacoub, P., Cohen, E.A., Creager, M.A., Easton, J.D., Flather, M.D., Haffner, S.M., Hamm, C.W., Hankey, G.J., Johnston, S.C., Mak, K.H., Mas, J.L., Montalescot, G., Pearson, T.A., Steg, P.G., Steinhubl, S.R., Weber, M.A., Brennan, D.M., Fabry-Ribaudo, L., Booth, J., and Topol, E.J. (2006) Clopidogrel and aspirin versus aspirin alone for the prevention of atherothrombotic events N Eng J Med 354, 1706–17. 42. Mehta, S.R., Yusuf, S., Peters, R.J., Bertrand, M.E., Lewis, B.S., Natarajan, M.K., Malmberg, K., Rupprecht, H., Zhao, F., Chrolavi-

43.

44.

45.

46.

47.

48.

219

cius, S., Copland, I., and Fox, K.A. (2001) Effects of pretreatment with clopidogrel and aspirin followed by long-term therapy in patients undergoing percutaneous coronary intervention: the PCI-CURE study Lancet 358, 527–33. Diener, H.C., Bogousslavsky, J., Brass, L.M., Cimminiello, C., Csiba, L., Kaste, M., Leys, D., Matias-Guiu, J., and Rupprecht, H.J. (2004) Aspirin and clopidogrel compared with clopidogrel alone after recent ischaemic stroke or transient ischaemic attack in high-risk patients (MATCH): randomised, double-blind, placebo-controlled trial Lancet 364, 331–7. Markus, H.S., Droste, D.W., Kaps, M., Larrue, V., Lees, K.R., Siebler, M., and Ringelstein, E.B. (2005) Dual antiplatelet therapy with clopidogrel and aspirin in symptomatic carotid stenosis evaluated using doppler embolic signal detection: the Clopidogrel and Aspirin for Reduction of Emboli in Symptomatic Carotid Stenosis (CARESS) trial Circulation 111, 2233–40. Rothwell, P.M. (2004) Lessons from MATCH for future randomised trials in secondary prevention of stroke Lancet 364, 305–7. Hirsh, J., and Bhatt, D.L. (2004) Comparative benefits of clopidogrel and aspirin in high-risk patient populations: lessons from the CAPRIE and CURE studies Arch Intern Med 164, 2106–10. CAPRIE Steering Committee (1996) A randomised, blinded, trial of clopidogrel versus aspirin in patients at risk of ischaemic events (CAPRIE). Lancet 348, 1329–39. Storey, R.F., Husted, S., Harrington, R.A., Heptinstall, S., Wilcox, R.G., Peters, G., Wickens, M., Emanuelsson, H., Gurbel, P., Grande, P., and Cannon, C.P. (2007) Inhibition of platelet aggregation by AZD6140, a reversible oral P2Y12 receptor antagonist, compared with clopidogrel in patients with acute coronary syndromes J Am Coll Cardiol 50, 1852–6.

Chapter 8 Prasugrel: A Novel Platelet ADP P2Y12 Receptor Antagonist Shaker A. Mousa, Walter P. Jeske, and Jawed Fareed Abstract Novel adenosine diphosphate (ADP) P2Y12 antagonists such as prasugrel, ticagrelor, cangrelor, and elinogrel are in various phases of clinical development. These ADP P2Y12 antagonists have advantages over clopidogrel ranging from faster onset to greater and less variable inhibition of platelet function. Novel ADP P2Y12 antagonists are under investigation to determine whether their use can result in improved antiplatelet activity, faster onset of action, and/or greater antithrombotic effects than clopidogrel without an unacceptable increase in hemorrhagic or other side effects. Prasugrel (CS-747; LY640315), a novel third-generation oral thienopyridine, is a specific, irreversible antagonist of the platelet ADP P2Y12 receptor. Pre-clinical and early phase clinical studies have shown prasugrel to be characterized by more potent antiplatelet effects, lower inter-individual variability in platelet response, and faster onset of activity compared to clopidogrel. Recent findings from large-scale phase-III testing show prasugrel to be more efficacious in preventing ischemic events in acute coronary syndrome patients undergoing percutaneous coronary intervention (PCI); however, this is achieved at the expense of an increased risk of bleeding. Prasugrel provides more rapid and consistent platelet inhibition than clopidogrel. Key words: Antiplatelet, acute coronary syndrome, antithrombotic, percutaneous coronary intervention, platelets, thienopyridines, thrombosis, antiplatelet combinations.

1. Introduction Platelets are the principle effectors of cellular hemostasis and key mediators in the pathogenesis of thrombosis. A variety of membrane receptors determine platelet reactivity with numerous agonists and adhesive proteins, and, therefore, represent key targets for the development of antiplatelet drug therapies. In this regard, several rapid-onset and rapid-offset reversible ADP antagonists are in clinical development, including reversible oral and rapid acting intravenous (1, 2) P2Y12 receptor antagonists (Table 8.1). S.A. Mousa (ed.), Anticoagulants, Antiplatelets, and Thrombolytics, Methods in Molecular Biology 663, DOI 10.1007/978-1-60761-803-4_8, © Springer Science+Business Media, LLC 2003, 2010

221

222

Mousa, Jeske, and Fareed

Table 8.1 Promising new antiplatelet P2Y12 inhibitor therapies in development Agent

Potential advantages (clopidogrel, ticlopidine)

over

current

agents

Prasugrel

Faster onset, greater antiplatelet effect, less variable response (FDA approved)

AZD6140

Reversible, faster onset and offset, greater antiplatelet effect

Cangrelor (IV)

Reversible, shorter half-life, faster onset, greater antiplatelet effect, less variable response

Novel inhibitors of platelet adhesion in early development target the von Willebrand (vWF)-GPIb/IX and collagen/GPVI interactions. Since platelet aggregation also plays such a critical role in the pathogenesis of arterial thrombosis, more potent agents that interfere with platelet aggregation via other pathways (e.g., the thrombin receptor) are also under clinical investigation (3) The major limitation to treatment with multiple antiplatelet agents is the increased bleeding risk associated with the enhanced antiplatelet effect. This is exemplified by the clinical conundrum in patients with acute coronary syndrome (ACS) who may need to undergo coronary artery bypass graft (CABG) surgery. Aspirin and clopidogrel irreversibly inhibit platelet function, with maximal antiplatelet effect occurring after 3–5 days of treatment. The increased risk of procedural bleeding arising from dual aspirin and clopidogrel administration immediately prior to CABG surgery raises the question of whether clopidogrel should be routinely given to patients presenting with ACS. In light of the current recommendation to discontinue clopidogrel at least 5 days prior to elective CABG surgery, the emergency physician is likely to avoid clopidogrel in anticipation that the patient may require urgent cardiac surgery. Delaying clopidogrel therapy until coronary revascularization has been performed would, however, deprive patients of the early clinical benefits of the drug. These limitations might be solved with the availability of rapid-onset and rapid-offset ADP antagonists. Furthermore, it is becoming clear that there is variability in individual responses to clopidogrel, with reported rates of inadequate antiplatelet response ranging between 4 and 30% (4). Reasons for this include genetic variables (polymorphisms of the P2Y12 receptor or CYP3A4 pathway), up-regulation of alternative pathways of platelet activation, and greater baseline pre-treatment platelet reactivity, as well as extrinsic mechanisms such as patient non-compliance and drug–drug interactions involving CYP3A4 (4). Hence, a clinical need exists for superior antiplatelet agents.

Prasugrel: A Novel Platelet ADP P2Y12 Receptor Antagonist

223

2. Prasugrel 2.1. Pharmacokinetics

Prasugrel is a pro-drug requiring activation by the hepatic CYP system. As it is rapidly absorbed, prasugrel has a faster onset of action compared to clopidogrel, with peak concentrations of active metabolite seen at 30 min (5). Prasugrel’s active metabolite is predominantly renally excreted (approximately 70%) and has a mean elimination half-life of 3.7 h. Important metabolic differences between clopidogrel and prasugrel result in a higher concentration of active metabolite with prasugrel. Approximately 85% of clopidogrel is hydrolyzed by esterases to an inactive carboxylic acid derivative, leaving 15% of the pro-drug to be metabolized in a two-step CYP process into active metabolite. Prasugrel, by contrast, is rapidly hydrolyzed by carboxyesterases and then metabolized in a single, CYP-dependent step that uses primarily isotypes CYP3A4 and CPY2B6 (6), which translates into improved inhibition of platelet aggregation on a mg/kg basis compared with clopidogrel. Current evidence from phase-I studies demonstrates that prasugrel, when compared to clopidogrel, provides greater inhibition of platelet aggregation with more rapid onset and less nonresponsiveness, most likely due to more efficient generation of the prasugrel active metabolite. In one study, 68 healthy subjects not taking aspirin received either a 300-mg loading dose of clopidogrel (then 75 mg daily) or a 60-mg loading dose of prasugrel (then 10 mg daily), followed by the alternate therapy after a 2week washout (7). The peak inhibitory effect on platelet aggregation was greater with prasugrel (mean inhibition, 79% versus 35%, P