uva-dare (digital academic repository) acute respiratory ... · flammatory reaction.22 mice...

UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Acute respiratory tract infections and (athero)thrombotic disease

Keller, T.T.

Link to publication

Citation for published version (APA):Keller, T. T. (2006). Acute respiratory tract infections and (athero)thrombotic disease.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 29 Nov 2020

https://dare.uva.nl/personal/pure/en/publications/acute-respiratory-tract-infections-and-atherothrombotic-disease(587ffa39-a0fc-4e25-a0cc-315c1a3b8ad2).html

Acute respiratory tract infections and (athero)thrombotic disease, academic thesis, University of Amsterdam, Amsterdam, the Netherlands

Copyright © 2006, Tymen Keller, Amsterdam, the Netherlands.All rights reseved. No part of this thesis may be reproduced, stored in a retrieval system or transmitted in any form or by any means, without the permission of the author, or, when appropriate , of the publishers of the publications.

Coverdesign and lay-out: BOEM, Amsterdam, Alex Klerk, (www.boem.nu)

Print: Gildeprint BV, Enschede, the Netherlands

The printing of this thesis was financially supported by the Federatie van Nederlandse Trombosediensten, J.E. Jurriaanse Stichting, Jaques H. de Jong Stichting, Stichting Amstol, Stichting tot Steun Promovendi Vasculaire Geneeskunde, Universiteit van Amsterdam, ZonMw, Astellas, AstraZeneca, Boehringer Ingel-heim, Bristol-Myers Squibb, GlaxoSmithKline, Merck Sharp & Dohme, Novo Nordisk, Pfizer, Sanofi Aventis, Schering-Plough, Servier, and Wyeth.

Financial support by the Netherlands Heart Foundation is gratefully acknowledged.

ISBN: 90-9021309-0

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctoraan de Universiteit van Amsterdamop gezag van de Rector Magnificus

prof.mr. P.F. van der Heijdenten overstaan van een door het college voor promoties ingestelde

commissie, in het openbaar te verdedigen in de Aula der Universiteit

op woensdag 13 december 2006, te 14:00 uurdoor Tymen Tako Keller

geboren te Utrecht

PROMOTIECOMMISSIE

promotor: Prof.dr. H.R. Büller Prof.dr. M.M. Levi

co-promotor: Dr. D.P.M. Brandjes Dr. V.E.A. Gerdes

overige leden: Prof.dr. H. ten Cate Prof.dr. R.J.G. Peters Prof.dr. P.H. Reitsma Prof.dr. P. Speelman Dr. S. Florquin Dr. F.L.J. Visseren

Faculteit der Geneeskunde

Promotiecommissie

Table of Contents

S

7

19

45

59

75

87

99

109

121

129

143

155

167

189

197

205

209

TABLE OF CONTENTS

Chapter 1Introduction and outline of the thesis

Chapter 2Infection and inflammation and the coagulation system

Part I: Inflammation and coagulation as cardiovascular risk factors

Chapter 3Serum levels of mannose-binding lectin and the risk of future coronary artery disease in apparently healthy men and women

Chapter 4Serum levels of type II secretory phospholipase A2 and the risk of future coronary artery disease in apparently healthy men and women; the EPIC-Norfolk prospective population study

Chapter 5Tissue factor serum levels and the risk of future coronary artery disease in apparently healthy men and women; the EPIC-Norfolk prospective population study

Chapter 6Non-functional toll-like receptor 2 is not a risk factor for acute coronary heart disease

Part II: Infection and (athero)thrombotic disease

Chapter 7Serologic evidence of acute respiratory tract infection in patients with acute coronary syndrome

Chapter 8Influenza vaccine for preventing acute coronary syndromes

Chapter 9Influenza infection and the risk of acute pulmonary embolism

Chapter 10Selective expansion of influenza A virus specific T cells in unstable human atherosclerotic plaques

Part III: (Acute respiratory tract) infections and coagulation

Chapter 11Acute respiratory tract infections in elderly patients increase systemic levels of hemostatic proteins

Chapter 12Cytomegalovirus infection and the prothrombotic state in humans

Chapter 13Effects on coagulation and fibrinolysis induced by influenza in mice with a reduced capacity to gen-erate activated protein C and a deficiency in plasminogen activator inhibitor type-1

Summary

Samenvatting

List of publications

Dankwoord

Chapter

1IntroduCtIon and outlIne of the thesIs

Tymen Keller • Harry Büller • Marcel Levi

Department of Vascular MedicineAcademic Medical Center, Amsterdam, The Netherlands

This chapter is based on:

Acute respiratory tract infections and acute coronary syndromes. Keller TT, Mairuhu ATA, Gerdes VEA, Brandjes DPM, Peters RJG, van Gorp ECM. Nederlands Tijdschrift voor Geneeskunde 2005;149:1267-1272

Infections and endothelial cells. Keller TT, Mairuhu ATA, de Kruif MD, Klein SK, Gerdes VEA, ten Cate H, Brandjes DPM, Levi M, van Gorp ECM. Cardiovascular Research 2003;60:40-48

Chapter 1

8

INTRODuCTIONEpidemiology of venous thrombotic and cardiovascular disease

Deep vein thrombosis and pulmonary embolism, collectively known as venous thrombo-embolism, have an annual incidence of approximately 2-3 per 1000 people in Western societies. Nowadays venous thrombo-embolism is a treatable disorder in most cases, but pulmonary embolism remains a potential life threatening condition with mortal-ity rates between 1.5-7%.1-3 In addition, venous thrombosis and pulmonary embolism together account for a great burden of disease and considerable medical costs. Arterial cardiovascular diseases comprise all clinical manifestations of atherosclerotic disease and subsequent acute thrombotic complications, including ischemic heart disease, stroke and peripheral arterial disease. In the Netherlands, cardiovascular disease accounts for 33% of 140.000 yearly deaths. World-wide, 17 million people die from cardiovascular disease,4 7.2 million from coronary artery disease, and 5.5 million from stroke. In addi-tion, millions of people suffer from non-lethal coronary artery disease, cerebrovascular disease and peripheral arterial disease. Although in Westerns societies the percentage of patients with cardiovascular disease will decline due to new drugs and prevention programs, the rise in cardiovascular disease in developing countries will lead to increase in cardiovascular mortality and morbidity world wide. The World Health Organization estimates that 20.5 million people will die yearly of cardiovascular disease at the end of the forthcoming decennium.

Risk factors for venous thrombotic and cardiovascular disease

Venous thrombo-embolic disease and arterial cardiovascular disease are both multifacto-rial diseases caused by inherited and acquired risk factors, but with a different etiology. Inherited risk factors for venous thrombo-embolic disease comprise prothrombotic abnormalities of blood coagulation including deficiencies in antithrombin, protein C and protein S, the prothrombin G20210A mutation and Factor V Leiden; a mutation in the factor V gene which causes a resistance to activated protein C. Acquired risk fac-tors include age, malignancies, major surgery, pregnancy, the use of oral contraceptives and the presence of anti-phospholipid antibodies. However, although the majority of the patients with venous thrombotic disease have at least one of these risk factors, it is likely that other unidentified (hereditary) risk factors contributing to the risk of venous thrombo-embolic disease exist.5

The pathogenesis of arterial cardiovascular disease is caused by a combination of genetic, lifestyle and environmental conditions. Hyperlipidemia, smoking, diabetes, hypertension and obesity are established risk factors, but only approximately half of the acute cardio-vascular events can be explained by the presence of one or more of these risk factors.6

Recently found new cardiovascular risk factors, such as C-reactive protein and fibrinogen have gained much attention, however it is unknown whether these risk factors are causal or only indicators of increased risk. In analogy with venous thrombosis, it is likely that for cardiovascular disease also yet unidentified risk factors exist.For both venous thrombosis and cardiovascular disease evidence has emerged that respi-ratory tract infection poses an additional risk. Smeeth et al showed in a large cohort study

Introduction and outline of the thesis

9

including 7278 patients with a first deep vein thrombosis and 3755 with a first pulmonary embolism that the risk of developing venous thrombotic disease is 2-fold increased in the first two weeks after infection.7 Interestingly, also urinary tract infections increase the risk of venous thrombotic disease in this cohort. In a comparable study including 20,486 patients with a first acute myocardial infarction and 19,063 patients with a first stroke, it was shown that in the first three days after a respiratory tract infection the risk of acute myocardial infarction was 5-fold increased and the risk of stroke was 3-fold increased.8 Importantly, the number of chronic infections (the “pathogen burden”) is independently related to the risk of myocardial infarction.9 Furthermore, influenza vaccination protects against cardiovascular disease.10 These studies together suggest that infections play a pathogenic role in the development of cardiovascular disease.

Etiology of atherosclerosis

Hallmark of atherosclerosis is the accumulation of large-dense-lipoproteins and the pres-ence of inflammatory processes in the arterial wall. Atherogenesis is a complicated pro-cess in which numerous processes including cholesterol transport into and from the arte-rial wall, pro- and anti-oxidative pathways and pro- and anti-inflammatory actions play a role. Both innate and adaptive immunity has been implicated in atherogenesis.11 Recently, evidence has emerged concerning the role of mannose-binding lectin in atherogenesis. Mannose-binding lectin is an element of the complement cascade and plays an important role in the first line defense of the innate immune system against pathogenic micro-organisms.12 Mannose-binding lectin recognizes sugar patterns on the surface of many pathogens,13 phospholipids,14 immune complexes15 and apoptotic cells.16 In experimental models it was shown that the mannose-binding lectin-pathway is in-volved in ischemia-induced complement activation in mice.17 Consequently, in rats the ad-ministration of anti- mannose-binding lectin antibodies protects the heart from ischemia-reperfusion injury by reducing neutrophil infiltration and attenuating pro-inflammatory gene expression.18 However, studies examining the relation between coronary artery dis-ease risk and either serum levels of mannose-binding lectin or mannose-binding lectin genetic variants have reported equivocal results.19, 20 The atherogenicity of large-dense-li-poproteins depends on the duration of the residence of the lipoprotein in the arterial wall. A longer residence of large-dense-lipoproteins particles in the arterial wall increases the likelihood of oxidation. Secretory phopholipase A2 hydrolyses the outer layer of lipoproteins and increases the residence time of large-dense-lipoproteins in the arterial wall. Indeed, a small-scale hu-man study showed that high levels of secretory phopholipase A2 are associated with coro-nary artery disease.21 In addition, secretory phopholipase A2 has a pro-atherogenic effect by augmenting the formation of pro-inflammatory mediators. Further evidence for the role of the innate immune system in the development of coronary artery disease comes from studies with toll-like receptors. Toll-like receptors, remnants of the first line in host defense, recognize pathogen-associat-ed molecular patterns. However, because of molecular mimicry between these molecular patterns and oxidated phospholipids it has been suggested that mediators of the immune

Chapter 1

10

system mount an inflammatory reaction against newly exposed auto-antigens which are recognized as molecular patterns associated with pathogens.11 Indeed, ligation of toll-like receptor 2 and toll-like receptor 4 with certain oxidized phospholipids lead to a pro-in-flammatory reaction.22 Mice experiments showed that the absence of toll-like receptor 2 results in a reduction in atherosclerosis,23 and studies in human suggest a role of toll-like receptor 2 polymorphism in the risk cardiovascular disease.24

Etiology of coronary heart disease

The central role for inflammation in preclinical atherosclerosis continues in acute coro-nary heart disease.25 During acute ischemic events, the disturbed balance between pro- and anti-inflammatory mechanisms lead to atherosclerotic plaque weakening and subse-quent rupture. Indeed, dense plaque macrophage infiltration,26 plaque vascularity27 and intraplaque hemorrhage26 are associated with plaque rupture, suggesting possible causal links between plaque inflammation, hemorrhage and plaque instability. Studies with hu-man coronary atheroma obtained by atherectomy have shown that T cells are important during the onset of acute coronary syndromes, since the numbers of recently activated T cells are significantly increased in patients with unstable angina and acute myocardial infarction.28 At present, several antigens have been identified which could be responsible for the activation of T cells in atherosclerotic lesions. As discussed above these antigens include extra cellular matrix components and lipids prominent in atherosclerotic plaques, but also antigens from microbial origin, including pathogens that cause respiratory tract infections.29

In conclusion, arterial wall function and structure are modulated by interactions between injurious agents, blood vessel wall elements, hemostasis, monocytes and T lymphocytes. Invading mononuclear cells release enzymes that degrade collagen and elastin, thereby allowing cells to invade by disrupting matrix layers that otherwise stabilize developing plaque. Clot forming and inflammatory pathways then work in tandem to accelerate local macrophage and T-cell activation, which contributes to plaque erosion or rupture, form-ing a surface on which activated platelets may initiate thrombosis and subsequent acute atherothrombotic events.

Hemostasis, venous thrombosis and cardiovascular disease The relation between hypercoagulability and venous thrombosis has been postulated al-ready centuries ago. The well known Virchow triad (1856) defines three conditions lead-ing to thrombus formation: stasis of blood, changes in the vascular wall and changes in blood composition. Indeed, prothrombotic changes of circulating blood increase the risk for venous thrombosis. Deficiencies of protein C, protein S system or antithrombin are present in 5 to 10% of patients diagnosed with thrombosis.30

Homozygous carriers of Factor V leiden have a 80-fold increased risk of venous thrombo-embolism compared to non-carriers.31 Heterozygote carriers of factor V leiden or carriers of the prothrombin G20210A mutation, who have a 30% increase in prothrombin levels, have each a 4-fold increased risk.32 In addition, high levels of factor VIII, IX, XI, fibrino-gen, von Willebrand factor33 and thrombin activatable fibrinolysis inhibitor levels are all


11

found to be associated with higher risk for the development of venous thrombosis.34

Acute coronary heart disease and stroke often occurs as a result of thrombus formation on ruptured atherosclerotic plaques, leading to arterial occlusion.35 In addition, mural thrombosis in the vasculature of plaques has since long been recognized as important in the progression of atherosclerotic plaques.36 Therefore, coagulation activation and varia-tion in the fibrinolytic system have been proposed to increase the risk of coronary heart disease. However, evidence of hypercoagulability as a risk factor for coronary events is limited. Some evidence has emerged about the increased risk on coronary heart disease associated with high levels of von Willebrand factor,37 Factor VIII,38 tissue type plasmino-gen activator39 and plasminogen activator inhibitor type 1.40 A strong relation has been demonstrated for fibrinogen,41 however fibrinogen is an acute phase protein and the ob-served association could reflect the inflammatory response.Recently, several observations have strengthened the hypothesis that tissue factor, the main initiator of coagulation, plays an important role in the pathogenesis of cardiovas-cular disease. First, tissue factor is expressed extensively in atherosclerotic plaques,42 on the surface of ox-large-dense-lipoproteins-activated macrophages.43 Second, unstable ath-erosclerotic plaques contain more tissue factor than stable plaques,44 which was shown to be associated with increased plaque thrombogenicity.45 Third, acute myocardial infarc-tion and unstable angina are associated with increased circulating levels of tissue fac-tor-containing microparticles.46, 47 Finally, in people with acute myocardial infarction and unstable angina, serum tissue factor levels are increased compared to subjects with stable coronary heart disease or healthy controls.48, 49 However, it is unknown whether high levels of serum tissue factor predict future coronary heart disease in apparently healthy individuals. In conclusion, the associations between a prothrombotic state and venous thrombosis has been well established whereas this relation between a prothrombotic state and acute coronary heart disease remains less precisely defined.

Infection, inflammation and coagulation

Infection by various pathogens interacts with the vessel wall, endothelium, inflammatory and coagulation pathways.50 Disturbance of inflammatory and coagulation pathways are important for the pathogenesis of vascular disease. There is strong crosstalk between these two systems, whereby inflammation not only leads to activation of coagulation, but coagulation also markedly affects inflammatory activity.51 Pro-inflammatory media-tors can affect several coagulation mechanisms, and vice versa, activated coagulation proteases and physiological anticoagulants or components of the fibrinolytic system can modulate inflammation by specific cell receptors. The main links are via tissue factor, the protein C system, thrombin and the fibrinolytic system. In this crosstalk, the endothe-lium is playing a central role. In vitro it has been shown that infections with pathogens that cause respiratory tract infection, including influenza and cytomegalovirus, result in a prothrombotic state.52, 53

Chapter 1

12

OuTLINE OF THE THESISThis thesis provides a view on the spectrum of the infection induced conditions associ-ated with an increased risk of venous thrombosis and cardiovascular disease. The first part of this thesis contains studies on inflammation and coagulation as risk factors for cardiovascular disease (Part I). Chapter 3 and chapter 4 describe the relationship between components of the (innate) immunity, mannose-binding lectin and secretory phospholi-pase A2, and the risk of future coronary artery disease in apparently healthy individuals. In addition, in chapter 5 the risk of coronary heart disease is assessed in individuals with high levels of tissue factor, the main initiator of coagulation. Further studies on the role of innate immunity as risk factor for coronary heart disease are described in chapter 6, in which the role of toll-like receptor 2 in the risk of acute myocardial infarction and un-stable angina pectoris has been investigated. Part II explores the relation between infections and venous thrombosis and acute coro-


13

FIguRE 1

Schematic representation of the ‘cross-talk’ between coagulation and inflammation on rupture of atherosclerotic plaque. The expression of tissue factor on inflammatory cells and the exposure of tissue factor to blood result in thrombin generation. Activation of platelets occurs, both by thrombin and by exposure of collagen (and other subendothelial platelet-activating factors) to blood. Binding of tissue factor, thrombin, and other activated co-agulation proteases to specific receptors (PARs) on inflammatory cells induces the release of proinflammatory cytokines, subsequently further modulating coagulation and fibrinolysis. Inflammatory mechanisms by dashed arrows; Coagulation pathways are indicated by straight arrows.

nary heart disease. Chapter 7 describes the risk of acute coronary heart disease after acute respiratory tract infection. In addition, the association between influenza vaccination and risk of recurrent myocardial infarction is investigated. The current evidence of the latter is reviewed in a meta-analysis described in chapter 8. Whether patients with a pulmonary embolism show signs of recent influenza infection prior to the embolic event is studied in chapter 9. An etiological study is described in chapter 10, in which the T cell response to influenza is compared between T cells from unstable atherosclerotic plaques and pe-ripheral T cells. Part III addresses acute respiratory tract infections and coagulation. Chapter 11 de-scribes the effect of acute respiratory tract infections on levels of hemostatic proteins in elderly persons. These findings are further studied in chapter 12, in which cyto-megalovirus load in blood are correlated with levels hemostatic proteins. Chapter 13 describes the influenza-induced prothrombotic state in mice and explores the mecha-nisms of coagulation activation.

Chapter 1

14

Reference List 1. Douketis JD, Kearon C, Bates S et al. Risk of fatal pulmonary embolism in patients with treated venous thromboembolism. JAMA 1998;279:458-62. 2. Carson JL, Kelley MA, Duff A et al. The clinical course of pulmonary embolism. N Engl J Med 1992;326:1240- 5. 3. Goldhaber SZ, Visani L, De RM. Acute pulmonary embolism: clinical outcomes in the International Coopera- tive Pulmonary Embolism Registry (ICOPER). Lancet 1999;353:1386-9. 4. Murray CJ, Lopez AD. Mortality by cause for eight regions of the world: Global Burden of Disease Study. Lancet 1997;349:1269-76. 5. Heit JA, Phelps MA, Ward SA et al. Familial segregation of venous thromboembolism. J Thromb Haemost 2004;2:731-6. 6. Ridker PM. Novel risk factors and markers for coronary disease. Adv Intern Med 2000;45:391-418. 7. Smeeth L, Cook C, Thomas S et al. Risk of deep vein thrombosis and pulmonary embolism after acute infection in a community setting. Lancet 2006;367:1075-9. 8. Smeeth L, Thomas SL, Hall AJ et al. Risk of myocardial infarction and stroke after acute infection or vaccination. N Engl J Med 2004;351:2611-8. 9. Zhu J, Quyyumi AA, Norman JE et al. Effects of total pathogen burden on coronary artery disease risk and C-reactive protein levels. Am J Cardiol 2000;85:140-6. 10. Nichol KL, Nordin J, Mullooly J et al. Influenza vaccination and reduction in hospitalizations for cardiac disease and stroke among the elderly. N Engl J Med 2003;348:1322-32. 11. Binder CJ, Chang MK, Shaw PX et al. Innate and acquired immunity in atherogenesis. Nat Med 2002;8:1218- 26. 12. Gadjeva M, Takahashi K, Thiel S. Mannan-binding lectin--a soluble pattern recognition molecule. Mol Im- munol 2004;41:113-21. 13. Petersen SV, Thiel S, Jensenius JC. The mannan-binding lectin pathway of complement activation: biology and disease association. Mol Immunol 2001;38:133-49. 14. Kilpatrick DC. Phospholipid-binding activity of human mannan-binding lectin. Immunol Lett 1998;61:191- 5. 15. Roos A, Bouwman LH, van Gijlswijk-Janssen DJ et al. Human IgA activates the complement system via the mannan-binding lectin pathway. J Immunol 2001;167:2861-8. 16. Ogden CA, deCathelineau A, Hoffmann PR et al. C1q and mannose binding lectin engagement of cell surface calreticulin and CD91 initiates macropinocytosis and uptake of apoptotic cells. J Exp Med 2001;194:781- 95. 17. de Vries B, Walter SJ, Peutz-Kootstra CJ et al. The mannose-binding lectin-pathway is involved in comple ment activation in the course of renal ischemia-reperfusion injury. Am J Pathol 2004;165:1677-88. 18. Jordan JE, Montalto MC, Stahl GL. Inhibition of mannose-binding lectin reduces postischemic myocardial reperfusion injury. Circulation 2001;104:1413-8. 19. Madsen HO, Videm V, Svejgaard A et al. Association of mannose-binding-lectin deficiency with severe atherosclerosis. Lancet 1998;352:959-60. 20. Saevarsdottir S, Oskarsson OO, Aspelund T et al. Mannan binding lectin as an adjunct to risk assessment for myocardial infarction in individuals with enhanced risk. J Exp Med 2005;201:117-25. 21. Kugiyama K, Ota Y, Takazoe K et al. Circulating levels of secretory type II phospholipase A(2) predict coro nary events in patients with coronary artery disease. Circulation 1999;100:1280-4. 22. Walton KA, Cole AL, Yeh M et al. Specific phospholipid oxidation products inhibit ligand activation of toll- like receptors 4 and 2. Arterioscler Thromb Vasc Biol 2003;23:1197-203. 23. Mullick AE, Tobias PS, Curtiss LK. Modulation of atherosclerosis in mice by Toll-like receptor 2. J Clin Invest 2005. 24. Hamann L, Gomma A, Schroder NW et al. A frequent toll-like receptor (TLR)-2 polymorphism is a risk factor for coronary restenosis. J Mol Med 2005;83:478-85. 25. Lucas AR, Korol R, Pepine CJ. Inflammation in atherosclerosis: some thoughts about acute coronary syndromes. Circulation 2006;113:e728-e732. 26. Redgrave JN, Lovett JK, Gallagher PJ et al. Histological assessment of 526 symptomatic carotid plaques in relation to the nature and timing of ischemic symptoms: the Oxford plaque study. Circulation 2006;113:2320- 8. 27. Moreno PR, Purushothaman KR, Sirol M et al. Neovascularization in human atherosclerosis. Circulation 2006;113:2245-52. 28. van der Wal AC, Piek JJ, de Boer OJ et al. Recent activation of the plaque immune response in coronary lesions underlying acute coronary syndromes. Heart 1998;80:14-8. 29. de Boer OJ, van der Wal AC, Houtkamp MA et al. Unstable atherosclerotic plaques contain T-cells that respond to Chlamydia pneumoniae. Cardiovasc Res 2000;48:402-8. 30. Lensing AW, Prandoni P, Prins MH et al. Deep-vein thrombosis. Lancet 1999;353:479-85. 31. Rosendaal FR, Koster T, Vandenbroucke JP et al. High risk of thrombosis in patients homozygous for factor V Leiden (activated protein C resistance). Blood 1995;85:1504-8. 32. Emmerich J, Rosendaal FR, Cattaneo M et al. Combined effect of factor V Leiden and prothrombin 20210A


15

on the risk of venous thromboembolism--pooled analysis of 8 case-control studies including 2310 cases and 3204 controls. Study Group for Pooled-Analysis in Venous Thromboembolism. Thromb Haemost 2001;86:809-16. 33. Koster T, Blann AD, Briet E et al. Role of clotting factor VIII in effect of von Willebrand factor on occurrence of deep-vein thrombosis. Lancet 1995;345:152-5. 34. van Tilburg NH, Rosendaal FR, Bertina RM. Thrombin activatable fibrinolysis inhibitor and the risk for deep vein thrombosis. Blood 2000;95:2855-9. 35. Zaman AG, Helft G, Worthley SG et al. The role of plaque rupture and thrombosis in coronary artery disease. Atherosclerosis 2000;149:251-66. 36. Ross R. Atherosclerosis--an inflammatory disease. N Engl J Med 1999;340:115-26. 37. Morange PE, Simon C, Alessi MC et al. Endothelial cell markers and the risk of coronary heart disease: the Prospective Epidemiological Study of Myocardial Infarction (PRIME) study. Circulation 2004;109:1343-8. 38. Rumley A, Lowe GD, Sweetnam PM et al. Factor VIII, von Willebrand factor and the risk of major ischaemic heart disease in the Caerphilly Heart Study. Br J Haematol 1999;105:110-6. 39. Lowe GD, Danesh J, Lewington S et al. Tissue plasminogen activator antigen and coronary heart disease. Prospective study and meta-analysis. Eur Heart J 2004;25:252-9. 40. Juhan-Vague I, Pyke SD, Alessi MC et al. Fibrinolytic factors and the risk of myocardial infarction or sudden death in patients with angina pectoris. ECAT Study Group. European Concerted Action on Thrombosis and Disabilities. Circulation 1996;94:2057-63. 41. Danesh J, Lewington S, Thompson SG et al. Plasma fibrinogen level and the risk of major cardiovascular diseases and nonvascular mortality: an individual participant meta-analysis. JAMA 2005;294:1799-809. 42. Moreno PR, Bernardi VH, Lopez-Cuellar J et al. Macrophages, smooth muscle cells, and tissue factor in un stable angina. Implications for cell-mediated thrombogenicity in acute coronary syndromes. Circulation 1996;94:3090-7. 43. Hutter R, Valdiviezo C, Sauter BV et al. Caspase-3 and tissue factor expression in lipid-rich plaque macrophages: evidence for apoptosis as link between inflammation and atherothrombosis. Circulation 2004;109:2001-8. 44. Marmur JD, Thiruvikraman SV, Fyfe BS et al. Identification of active tissue factor in human coronary atheroma. Circulation 1996;94:1226-32. 45. Ardissino D, Merlini PA, Bauer KA et al. Thrombogenic potential of human coronary atherosclerotic plaques. Blood 2001;98:2726-9. 46. Cunningham MA, Romas P, Hutchinson P et al. Tissue factor and factor VIIa receptor/ligand interactions induce proinflammatory effects in macrophages. Blood 1999;94:3413-20. 47. Mallat Z, Benamer H, Hugel B et al. Elevated levels of shed membrane microparticles with procoagulant potential in the peripheral circulating blood of patients with acute coronary syndromes. Circulation 2000;101:841-3. 48. Suefuji H, Ogawa H, Yasue H et al. Increased plasma tissue factor levels in acute myocardial infarction. Am Heart J 1997;134:253-9. 49. Soejima H, Ogawa H, Yasue H et al. Heightened tissue factor associated with tissue factor pathway inhibitor and prognosis in patients with unstable angina. Circulation 1999;99:2908-13. 50. Keller TT, Mairuhu AT, de Kruif MD et al. Infections and endothelial cells. Cardiovasc Res 2003;60:40-8. 51. Levi M, Keller TT, van Gorp E et al. Infection and inflammation and the coagulation system. Cardiovasc Res 2003;60:26-39. 52. Bouwman JJ, Visseren FL, Bosch MC et al. Procoagulant and inflammatory response of virus-infected monocytes. Eur J Clin Invest 2002;32:759-66. 53. Pryzdial EL, Wright JF. Prothrombinase assembly on an enveloped virus: evidence that the cytomegalovirus surface contains procoagulant phospholipid. Blood 1994;84:3749-57.

Chapter 1

16

17


Chapter

2InfeCtIon and InflammatIon and the CoagulatIon system

Marcel Levi • Tymen Keller • Eric van gorp • Hugo ten Cate

Department of Vascular Medicine and Internal Medicine, Academic Medical Center, University of Amsterdam, the Netherlands

Department of Internal Medicine, Slotervaart Ziekenhuis, Amsterdam, the NetherlandsDepartment of Internal Medicine, Academic Hospital Maastricht, the Netherlands

Cardiovasc Res 2003;60:26-39

ABSTRACTSevere infection and inflammation almost invariably lead to hemostatic abnormalities, ranging from insignificant laboratory changes to severe disseminated intravascular co-agulation (DIC). Systemic inflammation results in activation of coagulation, due to tis-sue factor-mediated thrombin generation, downregulation of physiological anticoagulant mechanisms, and inhibition of fibrinolysis. Pro-inflammatory cytokines play a central role in the differential effects on the coagulation and fibrinolysis pathways. Vice versa, activation of the coagulation system may importantly affect inflammatory responses by direct and indirect mechanisms. Apart from the general coagulation response to inflam-mation associated with severe infection, specific infections may cause distinct features, such as hemorrhagic fever or thrombotic microangiopathy. The relevance of the cross-talk between inflammation and coagulation is underlined by the promising results in the treat-ment of severe systemic infection with modulators of coagulation and inflammation.

Chapter 2

20

INTRODuCTIONThe relevance of the interaction between coagulation and inflammation as a response to se-vere infection, in its most extreme form manifesting as disseminated intravascular coagula-tion (DIC) and multiple organ failure, is becoming increasingly clear.1,2 In recent years the various mechanisms that play an important role in this interaction have been elucidated and this knowledge has indeed been demonstrated to be applicable for the improvement of our understanding of the pathogenesis of severe infection or sepsis and, even more importantly, the clinical management of these patients.3 In this article the mechanisms that play a role in the interaction between inflammation and coagulation will be reviewed and specific features of infectious disease-mediated effects on the coagulation system will be highlighted. In addition, the cross-talk between activated coagulation and inflammatory mediators will be discussed.

INFLAMMATION-INDuCED COAguLATION MODIFICATIONIt has long been known that inflammation can lead to activation of the coagulation sys-tem. Acute inflammation, as a response to severe infection or trauma, results in a system-ic activation of the coagulation system.1,4 It was initially thought that this systemic activa-tion of coagulation was a result of direct activation of the contact system of coagulation by microorganisms or endotoxin. However, in the 1990s it became apparent that cytokines played a mediatory role in the activation of coagulation and subsequent fibrin deposition and that the point of impact on the coagulation system was rather the tissue factor-factor VIIa (‘extrinsic’) pathway than the contact system (‘intrinsic pathway’).5,6 Furthermore, the significance of impaired physiological anticoagulant pathways became increasingly clear.7 Lastly, it was shown that impaired fibrin removal by a suppressed fibrinolytic system con-tributed importantly to the microvascular deposition of fibrin. Vascular endothelial cells play a central role in all mechanisms that contribute to inflam-mation-induced activation of coagulation. Endothelial cells respond to the cytokines ex-pressed and released by activated leukocytes but can also release cytokines themselves.8 Furthermore, endothelial cells are able to express adhesion molecules and growth factors that may not only promote the inflammatory response further but also affect the coagula-tion response (Figure 1). However, it has recently become clear that, in addition to these mostly indirect effects of the endothelium, endothelial cells interfere directly with the initiation and regulation of fibrin formation and removal during severe infection.9,10

MECHANISMS OF INFLAMMATION-INDuCED COAguLATION AND FIBRINOLySIS ACTIvATION

Inflammation-induced coagulation activation is characterized by widespread intravascular fi-brin deposition, which appears to be a result of enhanced fibrin formation and impaired fibrin degradation.1,11 As outlined in the following paragraphs and schematically represented in fig-ure 1, enhanced fibrin formation is caused by tissue factor-mediated thrombin generation and simultaneously occurring depression of inhibitory mechanisms, such as the protein C and S system. The impairment of endogenous thrombolysis is mainly due to high circulating levels of PAI-1, the principal inhibitor of plasminogen activation. These derangements in coagulation and fibrinolysis are mediated by differential effects of various pro-inflammatory cytokines.5

Infection and inflammation and the coagulation system

21

(a) Tissue factor The initiating factors comprise the molecule tissue factor and the plasma protein fac-tor VII(a).12 Tissue factor is a membrane bound 4.5 kD protein, which is constitutively expressed on a number of cells throughout the body.13 These cells are mostly in tissues not in direct contact with blood, such as the adventitial layer of larger blood vessels. The subcutaneous tissue also contains substantial amounts of TF, and histologically TF ap-pears to be present in all blood tissue barriers.14 Upon expression at the cell surface, TF can interact with factor VII, either in its zymogen or activated form.15 Upon complexing, factor VII is activated, and the TF-factor VIIa complex catalyzes the conversion of both factors IX and X.16,17 Factors IXa and Xa enhance the activation of factors X and prothrom-bin respectively (Figure 1). In cells in contact with the blood circulation, TF is induced by the action of several compounds including cytokines, C reactive protein and advanced glycosilated endproducts.14 Inducible TF is predominantly expressed by monocytes and macrophages. The expression of TF on monocytes is markedly stimulated by the pres-ence of platelets and granulocytes in a P-selectin dependent reaction.6 This effect may be the result of nuclear factor kappa B (NFkB) activation induced by binding of activated platelets to neutrophils and mononuclear cells. This cellular interaction also markedly enhances the production of IL-1b, IL-8, MCP-1, and TNF-a.18

Under in vitro conditions, different cytokines such as tumor necrosis factor a (TNF-a), and interleukin (IL) 1, induce TF by vascular endothelial cells, but its relevance for in

Chapter 2

22

FIguRE 1: SCHEMATIC REPRESENTATION OF PATHOgENETIC PATHwAyS IN DIC

During systemic inflammatory response syndromes, both perturbed endothelial cells and activated mononuclear cells may produce proinflammatory cytokines that mediate coagulation activation. Activation of coagulation is initiated by tissue factor expression on activated mononuclear cells and endothelial cells. In addition, downregu-lation of physiological anticoagulant mechanisms and inhibition of fibrinolysis by endothelial cells will further promote intravascular fibrin deposition. PAI-1 indicates plasminogen activator inhibitor, type 1.

mononuclearcells

fibrin depos ition

tiss ue factorexpressi on

impairment ofanticoagulantmechanisms

P AI-1 mediatedinhibition offibrinolysispro-in fla mma tory

cy toki nes

intravasc ula rfibrin f ormatio n

insu ffici ent fibrinrem oval

mononuclearcells

fibrin depos ition

tiss ue factorexpressi on

impairment ofanticoagulantmechanisms

P AI-1 mediatedinhibition offibrinolysispro-infla mma tory

cy toki nes

intravasc ula rfibrin f ormatio n

insu ffici ent fibrinrem oval

vivo coagulation is uncertain.14,19 TF has also been localized on polymorphonuclear cells (PMN) in whole blood and ex vivo perfusion systems, suggesting an additional pool of “blood borne” TF,20 but it is unlikely that PMN actually synthesize TF in detectable quan-tities.6 Finally, TF and associated procoagulant activity in vitro has been detected on mi-crovesicles derived from predominantly platelets and granulocytes in patients with me-ningococcal sepsis.21 The localization on microvesicles may imply a highly efficient pool of procoagulant material assembling not only the initiating TF-factor VIIa complex but also the phospholipid surface facilitating the development of the tenase and prothrombi-nase complexes. In addition, a soluble form of TF is detectable in plasma from humans, and its levels are elevated in DIC.22 The significance of soluble TF is unknown. It is likely the result of membrane cleavage by metalloproteinases produced under inflammatory conditions, and it circulates in free as well as complexed forms. Whether soluble TF is catalytically active and involved in inducing coagulation remains to be investigated. In addition to its effect on the haemostatic system, tissue factor also has cell signalling and mitogenic effects. These effects are probably mediated by the protease-activated receptors (PAR’s) 1 and 2. Other examples of such effects are described below in the paragraph de-scribing the cross-talk between coagulation and inflammation. In humans with infection and an activated coagulation system, monocytes expressing enhanced levels of TF and procoagulant activity have been demonstrated. The tissue expression of TF appears to be confined to certain organs and vascular beds, but it is uncertain whether its expression is genetically controlled in an organ specific way.9

(b) The contact system

Whether in humans with severe infection the contact route plays any role remains un-known. Negatively charged substances including phospholipids and glycosaminoglycans, are a theroretically relevant source of contact activation. Recent studies suggest that intact cell surfaces mediate the assembly of contact factor proteins that may initiate this route of activation.23 The assembly of kininogens on a multiprotein receptor leads to controlled prekallikein activation. In the case of endothelial cells a complex between high molecular weight kininogen (HK) and prekallikrein or factor XI binds to a multiprotein receptor on these cells. This receptor consists of cytokeratin 1, uPAR and gC1qR.23 Cytokeratin 1 was also demonstrated on platelets and granulocytes, which may allow the interaction of HK with these cells as well. On endothelial cells PK activation leads to factor XII conversion amplifying PK activation. The activation of PK causes the cleavage of HK and liberation of bradykinin. Bradykinin induces tPA and this is one of the potential direct effects of the contact system on fibrinolysis.24, 25 Several lines of evidence support the important role of the contact system in activating fibrinolysis. Additional studies indicate that kininogens contain a peptide fragment that interferes with thrombin binding to platelets through preventing peptide liberation from PAR1 by thrombin.26 Taken together, a profibrinolytic and anticoagulant function of the contact system is more likely than a procoagulant role in vivo.Studies in patients with a clinical suspicion of DIC showed elevated levels of markers of activation of the contact system. Kaufmann et al measured elevated concentrations of


23

kininogen-a 1 anti-trypsin complexes,27 however studies with similar methods for coagula-tion activation markers have hardly ever detected elevated levels of contact activation.28 A study in patients with meningococcal septicemia showed a negative correlation between plasma factor XII levels and factor XIIa-C1 inhibitor complexes.29 Although this would suggest consumption of factor XII, and subsequent activation of factor XI downstream, a possible alternative explanation is a negative acute phase effect with reduced synthesis of factor XII, in conjunction with thrombin mediated activation of factor XI.30 Experimental studies suggested that blockade of the contact route in an E. coli sepsis model in primates with a monoclonal antibody against factor XIIa failed to protect ani-mals from the coagulation derangement, but diminished development of lethal hypoten-sion.31 This latter study provided reasonable support for the current view that the contact activation pathway does not contribute to coagulation activation. The contact route may play important roles in pro-inflammatory mechanisms related to vascular permeability, vascular proliferative processes (role of kininogen in smooth muscle cell proliferation) and stimulation of fibrinolysis.32

(c) Physiological anticoagulant pathways

Systemic activation of coagulation upon inflammation is counteracted by several mecha-nisms. First, coagulation inhibitors are present to slow down the coagulation mechanism (Figure 2). Soluble inhibitors constitute antithrombin (AT), proteins C and S, and tis-sue factor pathway inhibitor (TFPI).33 Antithrombin inhibits coagulation proteases by 1:1 complex formation, which, in the case of thrombin-antithrombin complexes, can be de-

FIguRE 2: POINT OF IMPACT OF THE THREE MAjOR PHySIOLOgICAL ANTICOAguLANT PATHwAyS

Antithrombin is the most important inhibitor of thrombin and factor Xa, activated protein C is able to degrade the essential cofactors Va and VIIIa, and TFPI inhibits the tissue factor/factor VIIa complex. Abbreviation: TFPI, tissue factor pathway inhibitor.

Chapter 2

24

TFPI

antithrombin

protein C system

tissue factor+ factor vII(a)

factor IXa(+ factor vIIIa)

factor Xa (+ factor va)

factor IIa (thrombin)

factor XIa

fibrinogen fibrin

tected by immunoassay in plasma from patients with DIC.22, 34 AT is thought to be one of the most important inhibitors of the activated coagulation system, and markedly lowered plasma levels are found in sepsis.35, 36 In the course of DIC the function of AT may be in-fluenced in several ways. First, a reduced absolute amount of the inhibitor may occur due to reduced protein synthesis on the one hand and clearance on the other hand. Clearance may be either in the form of protease-inhibitor complex, or due to proteolytic cleavage by proteolytic enzymes including granulocyte elastase.37 Second, the function of AT may be impaired due to the possibly reduced availability of glycosaminoglycans which may act as the physiological heparinlike cofactor of AT.38 Under the influence of cytokines the synthesis of glycosaminoglycans by endothelial cells may be reduced, impairing the in-hibitory potential of AT.39 In animal models of experimental bacteremia, an antithrombin concentrate prolonged survival and reduced the severity of DIC.40 This protective effect may have been distinct from the anticoagulant effect as such because in one of these mod-els, an active site blocked factor Xa preparation reduced the severity of DIC, but did not protect the animals from dying due to sepsis.41 The protective effect of antithrombin may have resulted from an anti- inflammatory effect. Infusion of recombinant antithrombin at very high concentrations (about tenfold higher than the basal plasma concentration) markedly reduced mortality due to lethal E. coli infusion, but also caused a significant reduction in the plasma levels of IL-6 and IL-8.42 In spite of significantly higher levels of TNF-a and lower levels of IL-10, the overall effect on DIC was apparently favourable because the reduction in fibrinogen level was less pronounced in the AT treated animals. An unexpected finding was that tPA levels, after an initial small rise in both groups, were markedly lower in the E. coli challenged baboons given antithrombin. As a result there was no incremement in the levels of plasmin-antiplasmin complexes (PAP), whereas the increase in PAI-1 levels was somewhat increased in AT treated animals. Since TNF-a is a critical mediator of fibrinolysis in primates given a low dose endotox-in,43 and given the elevated levels of TNF-a in baboons given antithrombin, a marked increase in tPA and PAP complexes would have been expected in the baboon model after AT infusion. The absence thereof suggests that TNF-a is not the single most important cytokine influencing fibrinolysis in sepsis. Potentially, the rise of PAI-1 sufficiently blunts fibrinolysis, but an alternative explanation may be a direct effect of antithrombin at tPA release by the endothelium, caused by AT binding to glycosaminoglycans at the endothe-lial surface. Given the net negative effect on fibrinolysis, the incomplete protection only achieved at very high doses of antithrombin, the therapeutic potential of antithrombin may be limited.Protein C and its cofactor protein S form a second line of defense. Thrombin binds to the endothelial cell membrane associated molecule thrombomodulin, and this complex converts protein C to its active form, protein Ca (PCa) (Figure 3).44 In addition, the throm-bin-thrombomodulin complex accelerates the conversion of the thrombin activatable fi-brinolytic inhibitor (TAFI).45 PCa inactivates factors Va and VIIIa by proteolytic cleavage, thus slowing down the coagulation cascade. Endothelial cells, primarily of large blood vessels, express an endothelial protein C receptor (EPCR), which augments the activation of protein C at the cell surface.44,46,47 Activated protein C has anti-inflammatory effects on


25

mononuclear cells and granulocytes, which may be distinct from its anticoagulant activity. Administration of PCa prevented thrombin induced thromboembolism in mice, mainly through its antithrombotic effect.48 An anti-inflammatory effect of PCa was demonstrated by Hancock et al,49 showing reduced TNF-a production in rats challenged with endotoxin. Activated PC reduced the severity of spinal cord injury in rats, probably due to inhibi-tion of leukocyte activity.50 In the latter model, soluble thrombomodulin had a similar anti-inflammatory effect, but this was also mediated by PCa. In a pulmonary inflamma-tory model induced by endotoxin, thrombomodulin also reduced vascular permeability through a protein C dependent mechanism.51

In contrast, defects in the protein C mechanism enhance the vulnerability to inflam-matory reactions and the ensuing activation of coagulation. In patient studies lowered levels of protein C and protein S are associated with increased mortality. Blockade of the activity of protein C by infusion of C4 binding protein turns a sublethal model of E. coli in baboons into a lethal model.52 Blockade of EPCR by a neutralizing monoclonal antibody also increased mortality in the E. coli baboon model.53 Vice versa, infusion of PC in the same model, protected against DIC and dying.54 Thus, it appears that the protein C mechanism is of great importance in the host defense against sepsis and coagulation activation. In situations, associated with DIC and systemic inflamma-tion, TNF-a and interleukin-1 may significantly down regulate the cellular expression

FIguRE 3: DySFuNCTION OF THE PROTEIN C SySTEM DuRINg DIC

Dysfunction of the protein C system in disseminated intravascular coagulation (DIC) is due to low levels of zy-mogen protein C, downregulation of thrombomodulin and the endothelial protein C receptor and low levels of free protein S due to acute phase-induced high levels of its binding protein (i.e., C4b-binding protein).

Chapter 2

26

1. physiological anticoagulation2. modulation of pro-inflammatory response

activated protein C

+ protein S

protein C

endothelium

thrombin

thrombomodulin

protein C receptor

of thrombomodulin, as suggested by cell culture experiments.55-57 The formation of thrombin may induce the shedding of EPCR by the endothelium, due to activation of metalloproteinases by thrombin. It is presently unknown whether thrombomodulin is also cleaved by similar mechanisms. A third inhibitory mechanism constitutes TFPI. This molecule, which exists in several pools either endothelial cell associated, or lipoprotein bound in plasma, inhibits the TF-factor VIIa complex by forming a quaternary complex in which factor Xa is the fourth component.58 Clinical studies in septic patients have not provided clues as to its importance, because in the majority of patients the levels of TFPI are not diminished as compared to normals.59 This may be explained by the lack of downregulatory effects of inflammatory mediators on cultured endothelial cells.60 The relevance of TFPI in sepsis-induced coagulation activation is illustrated by two lines of experimentation. First, depeletion of TFPI sensitized rabbits to DIC induced with tissue factor infu-sion sensitizes the animals to develop a more severe DIC.61 Second, TFPI infusion protected against the harmful effects of E. coli in primates. In this study, TFPI not only blocked DIC but all baboons challenged with lethal amounts of E. coli showed a marked improvement in vital functions and survived the experiment without apparent complications.62 The beneficial effects of TFPI may in part be due to its capacity to bind endotoxin and to interfere with its interaction with CD14,63 and to its attenuating effect on Il-6 generation. A recent study in volunteers confirmed the potential of TFPI to block the procoagulant pathway triggered by endotoxin.64 However, in this study in healthy subjects there was no reduction in cytokine levels, in contrast to the apparent anti-inflammatory effect of TFPI in the baboon study. There are several possible ex-planations for this discrepancy, which all relate to the marked differences between the endotoxin and E. coli sepsis models in which the effects of TFPI were investigated. In the beneficial effects of TFPI on survival, anti-inflammatory effects rather than anti-thrombotic activity may be prominent.62 In general, the presence of intact function and normal plasma levels of inhibitors appears to be important in the defense against DIC. It should be noted however, that there are no strong indications that individuals with congenital deficiencies of inhibi-tors would suffer a greater chance of developing DIC, but this issue remains to be explored. In addition, the influence of inhibitors in modifying the interaction between coagulation and inflammation deserves further attention.

(d) Fibrinolysis

In experimental models of severe infection fibrinolysis is activated, demonstrated by an initial activation of plasminogen activation, followed by a marked impairment caused by the release in blood of plasminogen activator inhibitor, type 1 (PAI-1).43,65,66 The latter inhibitor strongly inhibits fibrinolysis causing a net procoagulant situation. The molecu-lar basis is cytokine mediated activation of vascular endothelial cells; TNF-a and IL-1 decreased free tPA and increased PAI-1 production, TNF-a increased total uPA produc-tion in endothelial cells.67,68 Endotoxin and TNF-a stimulated PAI-1 production in liver, kidney, lung and adrenals of mice. The net procoagulant state is illustrated by a late rise


27

in fibrin breakdown fragments after E. coli challenge of baboons. Experimental data also indicate that the fibrinolytic mechanism is active in clearing fibrin from organs and circulation. Endotoxin induced fibrin formation in kidneys and adrenals was most dependent on a decrease in urokinase type plasminogen activator (uPA).69 PAI-1 knock-out mice challenged with endotoxin did not develop thrombi in the kidney in contrast to wildtype animals.68 Endotoxin administration to mice with a functionally inactive throm-bomodulin gene (TMProArg mutation) and defective protein C activator cofactor func-tion caused fibrin plugs in the pulmonary circulation, while wildtype animals did not develop macroscopic fibrin.70 This phenomenon proved to be temporary, with detectable thrombi at 4 hours after endotoxin, and disappearance of clots at 24 hours in animals sacrificed at that timepoint. These experiments demonstrate that fibrinolytic action is required to reduce the extent of intravascular fibrin formation.Fibrinolytic activity is markedly regulated by PAI-1, the principal inhibitor of this system. Recent studies have shown that a functional mutation in the PAI-1 gene, the 4G/5G polymorphism, not only influenced the plasma levels of PAI-1, but was also linked to clinical outcome of meningococcal septicemia. Patients with the 4G/4G genotype had significantly higher PAI-1 concentrations in plasma and an increased risk of death.71 Further investigations demonstrated that the PAI-1 polymorphism did not influence the risk of contracting meningitis as such, but probably increased the likelihood of develop-ing septic shock from meningococcal infection.72 These studies are the first evidence that genetically determined differences in the level of fibrinolysis influences the risk of developing complications of a gram-negative infection. In other clinical studies in cohorts of patients with DIC, high plasma levels of PAI-1 were one of the best predictors of mortality.73,74 These data suggest that activation of coagulation contributes to mortality in this situation, but as indicated earlier, the fact that PAI-1 is an acute phase protein, a higher plasma concentration may also be a marker of disease rather than a causal fac-tor. Interestingly, platelet a-granules contain large quantities of PAI-1 and release PAI-1 upon their activation. Since platelets become activated in case of severe inflammation and infection, this may further increase the levels of PAI-1 and contribute to the fibrino-lytic shut-off.

(e) Platelets

In addition to the mechanisms described above, platelets may play a role in the pathogenesis of inflammation-induced coagulation activation as well. Endotoxin can directly activate platelets and many pro-inflammatory cytokines are capable of inducing platelet activation through the sphingosine pathway and platelet activating factor (PAF) and its receptor. The importance of platelet activation for the activation of coagulation in severe infection is probably related to the provision of a suitable phospholipid sur-face on the activated platelet membrane for assembly of complexes of activated coagu-lation factors, such as the prothrombinase complex, consisting of activated factors X and V, prothrombin and calcium. The presence of such a surface may katalyze several-fold the generation of thrombin and render the coagulation system less susceptible to fluid-phase protease inhibitors.

Chapter 2

28

CyTOKINES AND OTHER MEDIATORy FACTORSActivation of blood coagulation requires a number of cofactors. Particularly, to develop coagulation activation, a number of interacting surfaces from cell remnants or intact cells, inflammatory mediators and coagulation proteins is required. The stimulus is pre-sented by the underlying disease and can thus be diverse: a bacterial cell compound such as endotoxin, a protease induced by a malignant cell type, tissue damage exposing catalytic surfaces and constitutively expressed TF, a substance that might directly acti-vated coagulation (fat, amniotic fluid) by unknown pathways, or others. Each of these triggers interact with other mediators: TF assembles on phospholipids, cells provide catalytic surfaces by exposure of phosphatidyl serine, cytokines interact with receptors and induce signalling pathways that induce TF and other proinflammatory components via the NFkB complex. To illustrate this situation, we will describe in greater detail the mechanisms that lead to endotoxin induced activation of the coagulation mechanism. This model provides the best studied model of the complex network of interactions that involves activated coagulation in a DIC like fashion.75 Endotoxin is the lipopolysaccha-ride compound from gram-negative bacteria inducing the sepsis syndrome and DIC. Gram-negative bacteria liberate endotoxins from their membrane which interact with cell surfaces by different pathways. In blood endotoxin directly binds to CD 14 at mono-cytes, and binds to endothelial cells after complexing with lipopolysaccharide binding protein (LBP) and the Toll-like receptor 4 complex.76 By these interactions, endotoxin induces a number of signalling pathways leading to the activation of the NFkB system, which starts the transcription of genes of proinflammatory cytokines, TF and others. These mechanisms are further reviewed elsewhere.Except for a direct positive effect of endotoxin on TF synthesis, the formation of the cyto-kines IL-1, TNF-a, and IL-6 stimulates TF formation.14 Ex vivo, this was directly demon-strated by showing that the production of TF mRNA was rapidly induced in blood cells after injection of endotoxin to human volunteers.77 In vitro, whole blood cells generate TF upon incubation with endotoxin or IL-1, TNF-a and IL-6. In vitro, cultured human endothelial cells also induce TF after incubation with TNF-a and IL-1, but their role in DIC remains unknown. Thus far, the only direct evidence for endothelial involvement in TF production has been the demonstration of TF expressing circulating endothelial cells in patients with sickle cell disease (a possible DIC associated condition).78 Studies in pri-mates have revealed the molecular mechanism underlying endotoxin induced activation of coagulation.65 After intravenous endotoxin challenge, a rapid production and liberation of proinflammatory cytokines is observed. Among those, TNF-a66 and Il-679 are impor-tant for inducing fibrinolytic and procoagulant changes in the blood. Specifically, a rapid increase in the plasma levels of tissue plasminogen activator (tPA) and the prothrombin fragment F1+2 as well as thrombin-antithrombin complexes (TAT) occurs. Fibrinolytic activity indicated by plasmin-antiplasmin complexes is impaired by a subsequent rise in plasminogen activator inhibitor 1 (PAI-1), and a net procoagulant state results.5 Studies in baboons challenged with lethal amounts of E. coli (a model of Gram-negative septice-mia) have underscored the importance of cytokines in mediating the sepsis syndrome, and demonstrated the prolonged proinflammatory course in which the actual formation


29

of fibrin is particularly evident at >24 hours after E. coli challenge.80 In fact most fibrin appears in blood, when thrombin generation is already declining, and evidence of endo-thelial cell activation becomes apparent indicated by elevated levels of soluble thrombo-modulin. Both in the chimpanzee endotoxin model and the E. coli model, TF is essential for inducing coagulation activity, and inhibition of the TF pathway abolishes clotting acti-vation.81 IL-6 is an important mediator of procoagulant effects, while TNF-a is particularly involved in the fibrinolytic response to endotoxin.82 Vice versa, inhibition of TF by TFPI reduced Il-6 in blood in the baboon model,83 suggesting that a more extensive crosstalk between inflammatory mediators and coagulation exists (see further). These effects may be more obvious in the sepsis model, than in the endotoxin model;84 Recent experiments in humans challenged with endotoxin did not reveal any effect of TFPI on IL-6 nor any of the other cytokines measured in blood.64 Thus the strength of the trigger may determine the degree of involvement of the cytokine network and its interactions with the coagula-tion system. Monocytes expressing TF directly bind factor VII(a), shed TF, or become as-sociated with the damaged vessel wall, a process probably occurring when macrophages expressing TF invade the vessel wall and play a procoagulant role in atherosclerosis.85 Circulating monocytes may trigger DIC, after interacting with platelets. Microvesicles may accelerate this process, and the complex interaction between cells, membrane frag-ments, soluble mediators and proteins may cause the full blown DIC syndrome. In the endotoxin model thrombin is an important feedback activator of the coagulation mecha-nism, because the thrombin specific inhibitor hirudin reduces the procoagulant response in human volunteers.86 The extent of catalysis and exhaustion of coagulation proteins and cellular elements is primarily determined by the strength of the triggers and the potential of the inhibitory mechanisms.

CROSS-TALK BETwEEN COAguLATION AND INFLAMMATIONCoagulation activation yields proteases that not only interact with coagulation protein zymogens, but also with specific cell receptors to induce signaling pathways. In par-ticular, protease interactions that affect inflammatory processes may be important in the development of DIC. Factor Xa, thrombin and the tissue factor–factor VIIa com-plex have each been shown to elicit pro-inflammatory activities.87,88 Fibrinogen/fibrin is important to the host defense mechanism, and probably has an additional role that is not directly related to clotting per se.89 Thrombin has been shown to induce a variety of non-coagulant effects, some of which may influence DIC by affecting cytokine levels in blood. It induces production of monocyte chemotactic protein-1 (MCP-1) and IL-6 in fibroblasts, epithelial cells and mononuclear cells in vitro.90 Endotoxin-stimulated whole blood produces significant IL-8, which has a procoagulant effect that is tissue factor- and thrombin-dependent.91 Thrombin also induces IL-6 and IL-8 production in endothelial cells. These effects of thrombin on cell activation are probably mediated by protease-ac-tivated receptors (PARs) 1, 3 and 4.92 PAR’s are cellular receptors for activated proteases that may contribute to intracellular signaling. Several studies have indicated that the tissue factor–factor VIIa complex also activates cells by binding to PARs, and it appears that PAR2 in particular is involved. Binding of the catalytic tissue factor–factor VIIa

Chapter 2

30

complex elicits a variety of inflammatory effects in mononuclear cells that may influence their contribution to coagulation activity.93 A direct indication of the relevance of these phenomena in vivo comes from a recent study showing that infusion of recombinant factor VIIa into healthy human volunteers causes a rise, albeit small, in plasma levels of IL-6 and IL-8.64 Although the plasma concentrations of factor VIIa in this experiment were much higher than endogenous factor VIIa concentrations in patients with sepsis, it is possible that the mechanism by which VIIa causes cytokine induction (direct or factor Xa/thrombin-mediated) is of physiologic importance. Another link between in-flammation and coagulation is formed by the Protein C system. Activated Protein C has anti-inflammatory effects on mononuclear cells and granulocytes, which may be distinct from its anticoagulant activity.94 The anti-inflammatory effect of activated Protein C was confirmed in vivo by the demonstration of reduced TNF-a production in rats challenged with endotoxin.49 In addition, we have recently shown that mice with a one-allele targeted deletion of the Protein C gene (heterozygous Protein C-deficient mice) not only develop more severe DIC and increased fibrin deposition upon systemic endotoxemia, but also have significantly higher circulating levels of pro-inflammatory cytokines compared with wild-type littermates.95 These experimental data are corroborated by the observations in the PROWESS trial that recombinant human Activated Protein C (drotrecogin alfa [acti-vated]) accelerated the decrease in levels of IL-6 in patients with severe sepsis.96 For high doses of antithrombin similar interactions with the inflammatory cascades were seen.42 Taken together, a number of coagulation proteases can induce pro-inflammatory media-tors that have procoagulant effects, which may amplify the cascade that leads to DIC. Ef-fects at the cellular level will be determined by the capacity of the coagulation inhibitors to inactivate these enzymes.

RELEvANCE OF THE CROSS-TALK BETwEEN INFLAMMATION AND COAguLATION FOR ORgAN FAILuRE

There is ample evidence that inflammatory activation in concert with microvascular thrombosis contributes to multiple organ failure in patients with severe infection and DIC. Firstly, extensive data has been reported on post-mortem findings of patients with DIC.97,98 These autopsy findings include diffuse bleeding at various sites, hemorraghic necrosis of tissue, microthrombi in small blood vessels and thrombi in mid-size and larger arteries and veins. The demonstration of ischemia and necrosis was invariably due to fibrin deposition in small and mid-size vessels of various organs.99 Importantly, the presence of these intravascular thrombi appears to be clearly and specifically related to the clinical dysfunction of the organ. Secondly, experimental animal studies of DIC show fibrin deposition in various organs. Experimental bacteremia or endotoxemia causes in-tra- and extravascular fibrin deposition in kidneys, lungs, liver brain and various other or-gans. Amelioration of the hemostatic defect by various interventions in these experimen-tal models appears to improve organ failure and, in some but not all cases, mortality.40,83 Lastly, clinical studies support the notion of coagulation as an important denominator of clinical outcome. DIC has shown to be an independent predictor of mortality in patients with sepsis and severe trauma.1,100


31

COAguLATION AND vASCuLAR COMPLICATIONS IN INFECTIOuS DISEASESApart from the generalized response upon systemic inflammation as discussed above, specific infections may result in thrombohemorrhagic syndromes, hemolytic uremic syndrome (HUS), thrombotic thrombocytopenic purpura (TTP) or vasculi-tis. Symptoms and signs may be dominated by bleeding, thrombosis, or both.1, 76,

101 Clinically overt infection-induced acvtivation of coagulation may occur in 30-50 percent of patients with Gram negative sepsis.102-105 Contrary to widely held belief, this may appear as common in patients with Gram positive sepsis as in those with Gram negative sepsis.106 Activation of the coagulation system has also been docu-mented for non-bacterial pathogens, i.e. viruses,107,108 protozoa (Malaria),109,110 fungi111 and spirochetes.112

(a) Thrombosis and bleeding

Viral and bacterial infections may result in an enhanced risk for local thromboembolic disease, i.e. deep venous thrombosis or pulmonary embolism. In a thrombo-embolic pre-vention study of low-dose subcutaneous standard heparin for hospitalized patients with infectious diseases, morbidity due to thromboembolic disease was significantly reduced in the heparin group compared to the group receiving no prophylaxis. There was however no beneficial effect of prophylaxis on mortality due to thromboembolic complications.113 In chronic viral diseases, such as CMV or HIV infection, the risk of thromboembolic complications is relatively low.114-116 Viral hemorrhagic fever is complicated by DIC in the most severe cases.117-121 DIC is not frequently encountered in other viral infections but has been reported in cases of infec-tion with Rotavirus,122 Varicella, Rubella, Rubeola and Influenza.123-129 TTP and HUS, trig-gered by a viral or bacterial infection130,131 frequently lead to bleeding symptoms, but also platelet and fibrin thrombi may be generated in various organs, leading to prominent symptoms with organ dysfunction. In specific infections, such as viral hemorrhagic fever, bleeding complications are prominent.117, 118 In other viral and bacterial infections associ-ated with TTP or HUS, bleeding also is often the prominent and presenting symptom (table 1).10,132,133

Chapter 2

32

TABLE 1

virus geographic distribution Source of infection

Dengue HF Southeast Asia, Mid/South America, China Mosquito

Chikungunya Southeast Asia Mosquito Ebola Zaire, Sudan Unknown Marburg HF Zimbabwe, Kenya, Uganda Unknown

Lassa fever West Africa Rodent Yellow fever South America, Africa Mosquito

Omsk HF Former Soviet Union Tick Hantaan Central Europe, former Soviet Union, Asia Rodent

(b) vasculitis

Bacterial and viral infections may result in a vasculitis-like syndrome with either bleeding manifestations or ischemic injury.134-136 Vasculitis is a well-documented phenomenon in CMV infection,137,138,138 occurring predominantly in the vasculature of the gastrointestinal tract where it causes colitis,139,140 the central nervous system where it causes cerebral infarction,141,142 and the skin where it results in petechiae, purpura papules, localized ulcers or a diffuse maculo-papular eruption.143 HIV (human immunodeficiency virus) infection may be accompanied by vasculitis syndromes, e.g. polyarteriitis nodosa, Henoch Schönlein purpura and leukocyto-clastic vasculitis.144-146 Hepatitis B and C infection may cause polyarteritis-like vasculitis.147,148 Parvovirus B19 has been suggested to be associated with vasculitis-like syndromes including Kawasaki disease, polyarteritis nodosa and Wegener’s granulomatosis.149-151

SPECIFIC FEATuRES OF THE PATHOgENESIS OF COAguLATION DISORDERS IN INFECTIOuS DISEASE

Much of our knowledge on the mechanisms that play a role in infection-associated activa-tion of coagulation comes from observations in clinical and experimental infectious dis-ease. Essentially, and as outlined in detail in the preceding paragraphs, several pathways appear to play a role, including tissue factor-mediated generation of thrombin, impair-ment of physiological anticoagulant mechanisms, and a depression of the fibrinolytic sys-tem, due to elevated levels of PAI-1. In addition, specific features of coagulation activation in patients with infectious disease may be appreciated:

(1) Endothelial cell activation: Endothelial cells may turn into a procoagulant state either by stimulation of cytokines in concert with circulating blood cells, such as lymphocytes or platelets, or by direct infection (viruses) of endothelial cells. The role of endothelial cells seems to be crucial in the development of shock and activation of coagulation.152,152,153,153 Endothelial cell injury is a common feature of viral infection and can alter hemostasis in a direct or indirect manner. The endothelial cell can be directly infected by different viruses,154,155 e.g. HSV, Adenovirus, Parainfluenzavirus, Poliovirus, Echovirus, Measles virus, Mumps virus, Cytomegalovirus, HTLV-1 and HIV.154-157 The ability to infect en-dothelial cells has also been demonstrated in hemorrhagic fevers due to Dengue virus, Marburg virus, Ebola virus, Hantaan virus, Lassa virus.158 Such infections may result in a procoagulant state, mainly by inducing tissue factor expression on the endothelial sur-face, probably mediated by cytokines such as IL-1, TNF-a, and IL-6.159-162 However not all viral infections affecting endothelial cells result in activation of coagulation, which may indicate that activation of endothelium is one factor in a multifactorial process. The change of the endothelial cell from a resting to a procoagulant state may be associ-ated with expression of endothelial surface adhesion molecules.160,163 These molecules, i.e. ICAM, VCAM, E Selectin and von Willebrand factor (vWF), play an essential role in the binding of leukocytes, resulting in a local inflammatory response, endothelial cell damage and subsequent plasma leakage and shock.164, 64 The finding of increased plasma concen-trations of these endothelial surface adhesion molecules is thought to be a reflection of the level of activation and perhaps damage of the endothelial cell.78,116,116


33

(2) Alterations in the protein C and S system: During sepsis the protein C/S system is down-regulated, as described earlier. Some viruses have the ability to induce specific changes in the coagulation inhibitory system. In the course of HIV infection the protein C/ S system may be impaired as a result of an acquired protein S deficiency, the pathogenesis of which is not yet clarified.165,166 Increased plasma concentrations of the C4b-binding protein, an acute phase protein which binds protein S, may result in decreased levels of free protein S. Antiphospholipid antibodies, which may be present in HIV patients, might interfere with the protein C/protein S complex and diminish its activity.167 In patients with Dengue hemorrhagic fever we also found decreased levels of both protein C and protein S activ-ity.168 As in Dengue, HIV can affect the endothelial cell; hypothetically this could lead to decreased production of protein S.

(3) Thrombocytopenia: Thrombocytopenia is seen in the course of many viral infections but is only occasionally serious enough to lead to hemostatic impairment and bleeding complications. It is assumed that thrombocytopenia is mainly immune-mediated.169 Al-ternative mechanisms are decreased thrombopoiesis, increased platelet consumption, or a combination of both.170 Direct interaction of the virus with platelets may lead to throm-bocytopenia and/or thrombocytopathy.171,172 Endothelial injury by the virus may lead to increased adherence and consumption of platelets.173 Viral infections that have been asso-ciated with thrombocytopenia are Mumps, Rubella,174 Rubeola, Varicella,175 disseminated Herpes simplex, CMV, infectious mononucleosis, Hantaan virus infection, Dengue hem-orrhagic fever, Crimean-congo hemorrhagic fever and Marburg hemorrhagic fever.174-179

INFLAMMATION-INDuCED COAguLATION AS A POINT OF IMPACT FOR (SuPPORTIvE) TREATMENT OF SEvERE INFECTION AND SEPSIS

Based on the assumption that defective physiological anticoagulant mechanisms play a pivotal role in the pathogenesis of coagulation derangement in sepsis, restoration of these pathways may be a logical approach in the (supportive) treatment of patients with severe sepsis. In fact, for all three anticoagulant pathways such treatment options are available or are at an advanced stage of development. Restoration of the antithrombin pathway may be achieved by administration of anti-thrombin concentrates. The use of antithrombin concentrates in patients with DIC has been studied relatively intensively. Most of these trials concern patients with sepsis and/or septic shock. All trials show some beneficial effect in terms of improvement of labo-ratory parameters, shortening of the duration of DIC, or even improvement in organ function. In the more recent clinical trials very high doses of antithrombin concentrate to attain supraphysiological plasma levels were used and the beneficial results in these trials seem to be more distinct.1, 180 Some trials showed a modest reduction in mortality in antithrombin-treated patients, however, the effect did never reach statistical signifi-cance. A large-scale multicenter randomized controlled trial to directly address this issue has very recently been completed.181 Preliminary results seem to indicate that there is no significant reduction in mortality of septic patients who were treated with antithrombin concentrate. A subgroup analysis indicates that patients not receiving concomitant hepa-

Chapter 2

34

rin may experience some benefit of antithrombin treatment but this hypothesis remains to be confirmed in a prospective study. A beneficial effect of recombinant human activated protein C was demonstrated in two randomized controlled trials. Firstly, in a dose-ranging clinical trial, 131 patients with sep-sis were enrolled. Included patients received activated protein C by continuous infusion at doses ranging from 12 μg/kg/hr to 30 μg/kg/hr, or placebo. In these patients a 40% reduction in the relative risk of mortality was shown, although this was not statistically significant (due to the size of the trial). Based on these encouraging results in the phase II trial, a large multicenter efficacy trial was performed.96 This trial was prematurely stopped at the second interim analysis because of a significant reduction in mortality in the ac-tivated protein C-treated patients. Mortality was 24.7% in the activated protein C group as compared with 30.8% in the placebo group (relative risk reduction 19.4 percent, 95% confidence interval, 6.6 to 30.5). The relative insufficiency of endogenous TFPI in sepsis and DIC may be overcome by the administration of pharmacological doses of recombinant TFPI. In a rat model of DIC the infusion of recombinant tissue factor pathway inhibitor (TFPI) immediately after endo-toxin administration significantly inhibited the consumption of coagulation factors and platelets.182 Furthermore a reduced number of fibrin thrombi was formed in liver, lungs, kidney and spleen. Similar effects were found in a rabbit model of DIC.61 In human endotoxemia, recombinant TFPI was shown to dose-dependently reduce thrombin gen-eration.64 Clinical phase II trials on the use of TFPI in patients with sepsis have yielded promising results but a large multicenter phase III trial turned out to be negative, al-though published results are not yet available.

CONCLuSIONA bi-directional relationship between coagulation and inflammation appears to play a pivotal role in the mechanisms leading to organ failure in patients with severe infection or sepsis. The endothelium plays a central role in all major pathways involved in the patho-genesis of hemostatic derangement during severe inflammation. Endothelial cells appear to be directly involved in the initiation and regulation of thrombin generation and the inhibition of fibrin removal. Pro-inflammatory cytokines are crucial in mediating these effects on endothelial cells, which themselves may also express cytokines, thereby ampli-fying the coagulative response. The interaction between inflammation and coagulation involves significant cross-talk between the systems. At all levels in the activated coagula-tion system in sepsis, natural anticoagulant pathways are defective and as a consequence incapable of containing the ongoing thrombin formation. Pharmacological restoration of these anticoagulant mechanisms may be a logical action in the (supportive) treatment of septic patients with coagulation abnormalities. Indeed, experimental and (initial) clinical studies show beneficial results of this type of treatment, not only confined to improve-ment of the coagulation status but also on clinically relevant outcome parameters such as organ function and survival.


35

Reference List 1. Levi M, De Jonge E, Van Der Poll T et al. Disseminated intravascular coagulation. Thrombosis and Haemo- stasis 1999;82:695-705. 2. Wheeler AP, Bernard GR. Treating patients with severe sepsis. New England Journal of Medicine 1999;340:207-14. 3. Levi M, Ten Cate H. New England Journal of Medicine 1999;341:586-92. 4. Esmon CT, Fukudome K, Mather T et al. Inflammation, sepsis, and coagulation. Haematologica 1999;84:254-9. 5. Levi M, Van Der Poll T, Ten Cate H et al. The cytokine-mediated imbalance between coagulant and anticoagulant mechanisms in sepsis and endotoxaemia. European Journal of Clinical Investigation 1997;27:3-9. 6. Osterud B. Tissue factor expression by monocytes: Regulation and pathophysiological roles. Blood Coagu- lation and Fibrinolysis 1998;9. 7. Taylor FB. Response of anticoagulant pathways in disseminated intravascular coagulation. Seminars in Thrombosis and Hemostasis 2001;27:619-31. 8. Ten Cate JW, Van Der Poll T, Levi M et al. Cytokines: Triggers of clinical thrombotic disease. Thrombosis and Haemostasis 1997;78:415-9. 9. Aird WC, Ulevitch R, Taylor F. Vascular bed-specific hemostasis:Role of endothelium in sepsis pathogen esis. Critical Care Medicine 2001;29. 10. Levi M, Ten Cate H, Van Der Poll T. Endothelium: Interface between coagulation and inflammation. Critical Care Medicine 2002;30. 11. Vallet B. Microthrombosis in sepsis. Minerva Anestesiologica 2001;67:298-301. 12. Ruf W, Edgington TS. Structural biology of tissue factor, the initiator of thrombogenesis in vivo. FASEB Journal 1994;8:385-90. 13. Mann KG, Van’t Veer C, Cawthern K et al. The role of the tissue factor pathway in initiation of coagulation. Blood Coagulation and Fibrinolysis 1998;9. 14. Camerer E, Kolstø AB, Prydz H. Cell biology of tissue factor, the principal initiator of blood coagulation. Thrombosis Research 1996;81:1-41. 15. Banner DW, D’Arcy A, ne C et al. The crystal structure of the complex of blood coagulation factor VIIa with soluble tissue factor. Nature 1996;380:41-6. 16. Ten Cate H, Bauer KA, Levi M et al. The activation of Factor X and prothrombin by recombinant factor VIIa in vivo is mediated by tissue factor. Journal of Clinical Investigation 1993;92:1207-12. 17. Bauer KA, Kass BL, Ten Cate H et al. Factor IX is activated in vivo by the tissue factor mechanism. Blood 1990;76:731-6. 18. Neumann FJ, Marx N, Gawaz M et al. Induction of cytokine expression in leukocytes by binding of thrombin-stimulated platelets. Circulation 1997;95:2387-94. 19. Edgington TS, Mackman N, Fan ST et al. Cellular immune and cytokine pathways resulting in tissue factor expression and relevance to septic shock. Nouvelle Revue Francaise d’Hematologie 1992;34. 20. Giesen PLA, Rauch U, Bohrmann B et al. Blood-borne tissue factor: Another view of thrombosis. Proceed- ings of the National Academy of Sciences of the United States of America 1999;96:2311-5. 21. Nieuwland R, Berckmans RJ, McGregor S et al. Cellular origin and procoagulant properties of microparticles in meningococcal sepsis. Blood 2000;95:930-5. 22. Koyama T, Nishida K, Ohdama S et al. Determination of plasma tissue factor antigen and its clinical significance. British Journal of Haematology 1994;87:343-7. 23. Schmaier AH, Røjkjær R, Shariat-Madar Z. Activation of the plasma kallikrein/kinin system on cells: A revised hypothesis. Thrombosis and Haemostasis 1999;82:226-33. 24. Smith D, Gilbert M, Owen WG. Tissue plasminogen activator release in vivo in response to vasoactive agents. Blood 1985;66:835-9. 25. Brown NJ, Nadeau JH, Vaughan DE. Selective stimulation of tissue-type plasminogen activator (t-PA) in vivo by infusion of bradykinin. Thrombosis and Haemostasis 1997;77:522-5. 26. Hasan AAK, Amenta S, Schmaier AH. Bradykinin and its metabolite, Arg-Pro-Pro-Gly-Phe, are selective inhibitors of thrombin-induced platelet activation. Circulation 1996;94:517-28. 27. Kaufman N, Page JD, Pixley RA et al. ?2-Macroglobulin-kallikrein complexes detect contact system activation in hereditary angioedema and human sepsis. Blood 1991;77:2660-7. 28. Nuijens JH, Huijbregts CCM, Eerenberg-Belmer AJM et al. Quantification of plasma factor XIIa-Cl-inhibitor and kallikrein-Cl-inhibitor complexes in sepsis. Blood 1988;72:1841-8. 29. Wuillemin WA, Fijnvandraat K, Derkx BHF et al. Activation of the intrinsic pathway of coagulation in children with meningococcal septic shock. Thrombosis and Haemostasis 1995;74:1436-41. 30. Minnema MC, Pajkrt D, Wuillemin WA et al. Activation of clotting factor XI without detectable contact activation in experimental human endotoxemia. Blood 1998;92:3294-301. 31. Pixley RA, De La Cadena R, Page JD et al. The contact system contributes to hypotension but not disseminated intravascular coagulation in lethal bacteremia. In vivo use of a monoclonal anti-factor XII antibody to block contact activation in baboons. Journal of Clinical Investigation 1993;91:61-8. 32. Levi M. Keep in contact: The role of the contact system in infection and sepsis. Critical Care Medicine 2000;28:3765-6. 33. Mann KG. Biochemistry and physiology of blood coagulation. Thrombosis and Haemostasis 1999;82:165-74.

Chapter 2

36

34. Kario K, Matsuo T, Kodama K et al. Imbalance between thrombin and plasmin activity in disseminated intravascular coagulation. Assessment by the thrombin-antithrombin-III complex/plasmin-alpha-2-antiplasmin complex ratio. Haemostasis 1992;22:179-86. 35. Mesters RM, Mannucci PM, Coppola R et al. Factor VIIa and antithrombin III activity during severe sepsis and septic shock in neutropenic patients. Blood 1996;88:881-6. 36. Alpert LI, Benisch B. Hemangioendothelioma of the liver associated with microangiopathic hemolytic anemia. Report of four cases. American Journal of Medicine 1970;48:624-8. 37. Jochum M, Lander S, Heimburger N et al. Effect of human granulocytic elastase on isolated human anti thrombin III. Hoppe-Seyler’s Zeitschrift fur Physiologische Chemie 1981;362:103-12. 38. Bourin M, Lindahl U. Glycosaminoglycans and the regulation of blood coagulation. Biochemical Journal 1993;289:313-30. 39. Kobayashi M, Shimada K, Ozawa T. Human recombinant interleukin-1- and tumor necrosis factor -mediated suppression of heparin-like compounds on cultured porcine aortic endothelial cells. Journal of Cellular Physiology 1990;144:383-90. 40. Kessler CM, Tang Z, Jacobs HM et al. The suprapharmacologic dosing of antithrombin concentrate for Sta phylococcus aureus-induced disseminated intravascular coagulation in guinea pigs: Substantial reduction in mortality and morbidity. Blood 1997;89:4393-401. 41. Taylor J, Chang ACK, Peer GT et al. DEGR-factor Xa blocks disseminated intravascular coagulation initiated by Escherichia coli without preventing shock or organ damage. Blood 1991;78:364-8. 42. Minnema MC, Chang ACK, Jansen PM et al. Recombinant human antithrombin III improves survival and attenuates inflammatory responses in baboons lethally challenged with Escherichia coli. Blood 2000;95:1117-23. 43. Biemond BJ, Levi M, Ten Cate H et al. Plasminogen activator and plasminogen activator inhibitor I release during experimental endotoxaemia in chimpanzees: Effect of interventions in the cytokine and coagulation cascades. Clinical Science 1995;88:587-94. 44. Esmon CT. The roles of protein C and thrombomodulin in the regulation of blood coagulation. Journal of Biological Chemistry 1989;264:4743-6. 45. Bouma BN, Meijers JCM. Fibrinolysis and the contact system: A role for factor XI in the down-regulation of fibrinolysis. Thrombosis and Haemostasis 1999;82:243-50. 46. Fukudome K, Esmon CT. Identification, cloning, and regulation of a novel endothelial cell protein C/activated protein C receptor. Journal of Biological Chemistry 1994;269:26486-91. 47. Laszik Z, Mitro A, Taylor J et al. Human protein C receptor is present primarily on endothelium of large blood vessels: Implications for the control of the protein C pathway. Circulation 1997;96:3633-40. 48. Gresele P, Momi S, Berrettini M et al. Activated human protein C prevents thrombin-induced thromboembolism in mice. Evidence that activated protein C reduces intravascular fibrin accumulation through the inhibition of additional thrombin generation. Journal of Clinical Investigation 1998;101:667-76. 49. Hancock WW, Tsuchida A, Hau H et al. The anticoagulants protein C and protein S display potent antiin- flammatory and immunosuppressive effects relevant to transplant biology and therapy. Transplantation Proceedings 1992;24:2302-3. 50. Taoka Y, Okajima K, Uchiba M et al. Activated protein C reduces the severity of compression-induced spinal cord injury in rats by inhibiting activation of leukocytes. Journal of Neuroscience 1998;18:1393-8. 51. Uchiba M, Okajima K, Murakami K et al. rhs-TM prevents ET-induced increase in pulmonary vascular permeability through protein C activation. American Journal of Physiology - Lung Cellular and Molecular Physiology 1997;273. 52. Taylor J, Dahlback B, Chang ACK et al. Role of free protein S and C4b binding protein in regulating the coagulant response to Escherichia coli. Blood 1995;86:2642-52. 53. Taylor J, Stearns-Kurosawa DJ, Kurosawa S et al. The endothelial cell proten C receptor aids in host defense against Escherichia coli sepsis. Blood 2000;95:1680-6. 54. Taylor J, Chang A, Esmon CT. Protein C prevents the coagulopathic and lethal effects of Escherichia coli infusion in the baboon. Journal of Clinical Investigation 1987;79:918-25. 55. Nawroth PP, Stern DM. Modulation of endothelial cell hemostatic properties by tumor necrosis factor. Journal of Experimental Medicine 1986;163:740-5. 56. Moore KL, Esmon CT, Esmon NL. Tumor necrosis factor leads to the internalization and degradation of thrombomodulin from the surface of bovine aortic endothelium cells in culture. Blood 1989;73:159-65. 57. Furlan M, Robles R, Galbusera M et al. Von Willebrand factor-cleaving protease in thrombotic thrombocytopenic purpura and the hemolytic-uremic syndrome. New England Journal of Medicine 1998;339:1578-84. 58. Broze J. Tissue factor pathway inhibitor and the revised theory of coagulation. Annual Review of Medicine 1995;46:103-12. 59. Novotny WF, Brown SG, Miletich JP et al. Plasma antigen levels of the lipoprotein-associated coagulation inhibitor in patient samples. Blood 1991;78:387-93. 60. Ameri A, Kuppuswamy MN, Basu S et al. Expression of tissue factor pathway inhibitor by cultured endothelial cells in response to inflammatory mediators. Blood 1992;79:3219-26. 61. Sandset PM, Warn-Cramer BJ, Rao LVM et al. Depletion of extrinsic pathway inhibitor (EPI) sensitizes rabbits to disseminated intravascular coagulation induced with tissue factor: Evidence supporting a physiologic role for EPI as a natural anticoagulant. Proceedings of the National Academy of Sciences of the United States of America 1991;88:708-12.


37

62. Randolph MM, White GL, Kosanke SD et al. Attenuation of tissue thrombosis and hemorrhage by ala-TFPI does not account for its protection against E. coli: A comparative study of treated and untreated non-sur- viving baboons challenged with LD100 E. coli. Thrombosis and Haemostasis 1998;79:1048-53. 63. Park CT, Creasey AA, Wright SD. Tissue factor pathway inhibitor blocks cellular effects of endotoxin by binding to endotoxin and interfering with transfer to CD14. Blood 1997;89:4268-74. 64. De Jonge E, Dekkers PEP, Creasey AA et al. Tissue factor pathway inhibitor dose-dependently inhibits coagulation activation without influencing the fibrinolytic and cytokine response during human endotoxemia. Blood 2000;95:1124-9. 65. Levi M, Ten Cate H, Bauer KA et al. Inhibition of endotoxin-induced activation of coagulation and fibrinolysis by pentoxifylline or by a monoclonal anti-tissue factor antibody in chimpanzees. Journal of Clinical Investigation 1994;93:114-20. 66. Biemond BJ, Friederich PW, Koschinsky ML et al. Apolipoprotein(a) attenuates endogenous fibrinolysis in the rabbit jugular vein thrombosis model in vivo. Circulation 1997;96:1612-5. 67. Schleef RR, Bevilacqua MP, Sawdey M et al. Cytokine activation of vascular endothelium. Effects on tissue- type plasminogen activator and type 1 plasminogen activator inhibitor. Journal of Biological Chemistry 1988;263:5797-803. 68. Sawdey MS, Loskutoff DJ. Regulation of murine type 1 plasminogen activator inhibitor gene expression in vivo. Tissue specificity and induction by lipopolysaccharide, tumor necrosis factor-a, and transforming growth factor-? Journal of Clinical Investigation 1991;88:1346-53. 69. Yamamoto K, Loskutoff DJ. Fibrin deposition in tissues from endotoxin-treated mice correlates with de creases in the expression of urokinase-type but not tissue-type plasminogen activator. Journal of Clinical Investigation 1996;97:2440-51. 70. Ten Cate H. Pathophysiology of disseminated intravascular coagulation in sepsis. Critical Care Medicine 2000;28. 71. Hermans PWM, Hibberd ML, Booy R et al. 4G/5G promoter polymorphism in the plasminogen-activator- inhibitor-1 gene and outcome of meningococcal disease. Lancet 1999;354:556-60. 72. Westendorp RGJ, Hottenga JJ, Slagboom PE. Variation in plasminogen-activator-inhibitor-1 gene and risk of meningococcal septic shock. Lancet 1999;354:561-3. 73. Brandtzaeg P, Joo GB, Brusletto B et al. Plasminogen activator inhibitor 1 and 2, alpha-2-antiplasmin, plasminogen, and endotoxin levels in systemic meningococcal disease. Thrombosis Research 1990;57:271-8. 74. Mesters RM, Florke N, Ostermann H et al. Increase of plasminogen activator inhibitor levels predicts out come of leukocytopenic patients with sepsis. Thrombosis and Haemostasis 1996;75:902-7. 75. Van Deventer SJH, Buller HR, Ten Cate JW et al. Experimental endotoxemia in humans: Analysis of cytokine release and coagulation, fibrinolytic, and complement pathways. Blood 1990;76:2520-6. 76. Aderem A, Ulevitch RJ. Toll-like receptors in the induction of the innate immune response. Nature 2000;406:782-7. 77. Franco RF, De Jonge E, Dekkers PEP et al. The in vivo kinetics of tissue factor messenger RNA expression during human endotoxemia: Relationship with activation of coagulation. Blood 2000;96:554-9. 78. Solovey A, Gui L, Key NS et al. Tissue factor expression by endothelial cells in sickle cell anemia. Journal of Clinical Investigation 1998;101:1899-904. 79. McClanahan TB, Ignasiak DP, Juneau P et al. Antithrombotic effects of BCH 2763, a new direct thrombin inhibitor, in a canine model of venous thrombosis. Journal of Thrombosis and Thrombolysis 1999;7:301-6. 80. Taylor J. Studies on the inflammatory-coagulant axis in the baboon response to E. coli: regulatory roles of proteins C, S, C4bBP and of inhibitors of tissue factor. Progress in clinical and biological research 1994;388:175-94. 81. Aguejouf O, Oualane FA, Inamo J et al. The arterial antithrombotic activity of thioxylosides in a rat model of laser-induced thrombosis. Seminars in Thrombosis and Hemostasis 1996;22:327-33. 82. Van der Poll T, Levi M, Hack CE et al. Elimination of interleukin 6 attenuates coagulation activation in experimental endotoxemia in chimpanzees. Journal of Experimental Medicine 1994;179:1253-9. 83. Creasey AA, Chang ACK, Feigen L et al. Tissue factor pathway inhibitor reduces mortality from Escherichia coli septic shock. Journal of Clinical Investigation 1993;91:2850-60. 84. Esmon CT. Introduction: Are natural anticoagulants candidates for modulating the inflammatory response to endotoxin? Blood 2000;95:1113-6. 85. Annex BH, Denning SM, Channon KM et al. Differential expression of tissue factor protein in directional atherectomy specimens from patients with stable and unstable coronary syndromes. Circulation 1995;91:619-22. 86. Pernerstorfer T, Hollenstein U, Hansen JB et al. Lepirudin blunts endotoxin-induced coagulation activation. Blood 2000;95:1729-34. 87. Altieri DC. Molecular cloning of effector cell protease receptor-1, a novel cell surface receptor for the protease factor Xa. Journal of Biological Chemistry 1994;269:3139-42. 88. Camerer E, Huang W, Coughlin SR. Tissue factor- and factor X-dependent activation of protease-activated receptor 2 by factor VIIa. Proceedings of the National Academy of Sciences of the United States of America 2000;97:5255-60. 89. Degen JL. Hemostatic factors and inflammatory disease. Thrombosis and Haemostasis 1999;82:858-64.

Chapter 2

38

90. Johnson K, Choi Y, DeGroot E et al. Potential mechanisms for a proinflammatory vascular cytokine response to coagulation activation. Journal of Immunology 1998;160:5130-5. 91. Johnson K, Aarden L, Choi Y et al. The proinflammatory cytokine response to coagulation and endotoxin in whole blood. Blood 1996;87:5051-60. 92. Kahn ML, Nakanishi-Matsui M, Shapiro MJ et al. Protease-activated receptors 1 and 4 mediate activation of human platelets by thrombin. Journal of Clinical Investigation 1999;103:879-87. 93. Cunningham MA, Romas P, Hutchinson P et al. Tissue factor and factor VIIa receptor/ligand interactions induce proinflammatory effects in macrophages. Blood 1999;94:3413-20. 94. Esmon CT, Xu J, Gu JM et al. Endothelial protein C receptor. Thrombosis and Haemostasis 1999;82:251-8. 95. Levi M, Dorffler-Melly J, Reitsma P et al. Aggravation of endotoxin-induced disseminated intravascular coagulation and cytokine activation in heterozygous protein-C-deficient mice. Blood 2003;101:4823-7. 96. Bernard GR, Vincent JL, Laterre PF et al. Efficacy and safety of recombinant human activated protein C for severe sepsis. New England Journal of Medicine 2001;344:699-709. 97. Robboy SJ, Major MC, Colman RW et al. Pathology of disseminated intravascular coagulation (DIC). Analysis of 26 cases. Human Pathology 1972;3:327-43. 98. Shimamura K, Oka K, Nakazawa M et al. Distribution patterns of microthrombi in disseminated intravascular coagulation. Archives of Pathology and Laboratory Medicine 1983;107:543-7. 99. Coalson JJ. Pathology of Sepsis, Septic Shock, and Multiple Organ Failure: Perspective on Sepsis and Septic Shock 1986;27-59. 100. Fourrier F, Chopin C, Goudemand J et al. Septic shock, multiple organ failure, and disseminated intravascular coagulation; Compared patterns of antithrombin III, protein C, and protein S deficiencies. Chest 1992;101:816-23. 101. Van Gorp ECM, Suharti C, Ten Cate H et al. Review: Infectious diseases and coagulation disorders. Journal of Infectious Diseases 1999;180:176-86. 102. Gando S, Kameue T, Nanzaki S et al. Disseminated intravascular coagulation is a frequent complication of systemic inflammatory response syndrome. Thrombosis and Haemostasis 1996;75:224-8. 103. Thijs LG, De Boer JP, De Groot MCM et al. Coagulation disorders in septic shock. Intensive Care Medicine 1993;19. 104. Baglin T. Disseminated intravascular coagulation: Diagnosis and treatment. British Medical Journal 1996;312:683-7. 105. Gando S, Nakanishi Y, Tedo I. Cytokines and plasminogen activator inhibitor-1 in posttrauma disseminated intravascular coagulation: Relationship to multiple organ dysfunction syndrome. Critical Care Medicine 1995;23:1835-42. 106. Bone RC. Gram-positive organisms and sepsis. Archives of Internal Medicine 1994;154:26-34. 107. Bhamarapravati N. Hemostatic defects in dengue hemorrhagic fever. Reviews of Infectious Diseases 1989;11 Suppl 4. 108. Heller MV, Marta RF, Sturk A et al. Early markers of blood coagulation and fibrinolysis activation in Argen- tine hemorrhagic fever. Thrombosis and Haemostasis 1995;73:368-73. 109. Mohanty D, Ghosh K, Nandwani SK et al. Fibrinolysis, inhibitors of blood coagulation, and monocyte derived coagulant activity in acute malaria. American Journal of Hematology 1997;54:23-9. 110. Chuttani K, Tischler MD, Pandian NG et al. Diagnosis of cardiac tamponade after cardiac surgery: Relative value of clinical, echocardiographic, and hemodynamic signs. American Heart Journal 1994;127:913-8. 111. Fera G, Semeraro N, De Mitrio V et al. Disseminated intravascular coagulation associated with disseminated cryptococcosis in a patient with acquired immunodeficiency syndrome. Infection 1993;21:171-3. 112. Schroder S, Spyridopoulos I, Konig J et al. Thrombotic thrombocytopenic purpura (TTP) associated with a Borrelia burgdorferi infection. American Journal of Hematology 1995;50:72-3. 113. Gardlund B. Randomised, controlled trial of low-dose heparin for prevention of fatal pulmonary embolism in patients with infectious diseases. The Heparin Prophylaxis Study Group. Lancet 1996;347:1357-61. 114. Laing RBS, Brettle RP, Leen CLS. Venous thrombosis in HIV infection. International Journal of STD and AIDS 1996;7:82-5. 115. Jenkins RE, Peters BS, Pinching AJ. Thromboembolic disease in AIDS is associated with cytomegalovirus disease. AIDS 1991;5:1540-2. 116. Drancourt M, George F, Brouqui P et al. Diagnosis of Mediterranean spotted fever by indirect immunofluorescence of Rickettsia conorii in circulating endothelial cells isolated with monoclonal antibody-coated immunomagnetic beads. Journal of Infectious Diseases 1992;166:660-3. 117. Hayes EB, Gubler DJ. Dengue and dengue hemorrhagic fever. Pediatric Infectious Disease Journal 1992;11:311-7. 118. Sumarmo, Wulur H, Jahja E. Clinical observations on virologically confirmed fatal dengue infections in Jakarta, Indonesia. Bulletin of the World Health Organization 1983;61:693-701. 119. Kuberski T, Rosen L, Reed. D et al. Clinical and laboratory observations of patients with primary and secondary dengue type 1 infections with hemorrhagic manifestations in Fiji. American Journal of Tropical Medicine and Hygiene 1977;26:775-83. 120. Egbring R, Slenczka W, Baltzer G. Clinical manifestations and mechanism of the hemorrhagic diathesis in Marburg viral disease. Marburg Virus Disease 1971;41-9.


39

121. Gear JSS, Cassel GA, Gear AJ. Outbreak of Marburg virus disease in Johannesburg. British Medical Journal 1975;4:489-93. 122. Limbos MAP, Lieberman JM. Disseminated intravascular coagulation associated with rotavirus gastroenteritis: Report of two cases. Clinical Infectious Diseases 1996;22:834-6. 123. McKay DG, Margaretten W. Disseminated intravascular coagulation in virus diseases. Archives of Internal Medicine 1967;120:129-52. 124. Linder M, ller-Berghaus G, Lasch HG et al. Virus infection and blood coagulation. Thrombosis et diathesis haemorrhagica 1970;23:1-11. 125. Talley NA, Assumpcao CA. Disseminated intravascular clotting complicating viral pneumonia due to influenza. Medical Journal of Australia 1971;2:763-6. 126. Whitaker AN, Graeme ER. Disseminated intravascular coagulation and acute renal failure in influenza A2 infection. Medical Journal of Australia 1974;2:196-201. 127. Settle H, Glueck HI. Disseminated intravascular coagulation associated with influenza. Ohio State Medical Journal 1975;71:541-3, 547. 128. Cumming AD, Thomson D, Davison AM et al. Significance of urinary C3 excretion in glomerulonephritis. Journal of Clinical Pathology 1976;29:601-7. 129. Clemens R, Pramoolsinsap C, Lorenz R et al. Activation of the coagulation cascade in severe falciparum malaria through the intrinsic pathway. British Journal of Haematology 1994;87:100-5. 130. Badesha PS, Saklayen MG. Hemolytic uremic syndrome as a presenting form of HIV infection. Nephron 1996;72:472-5. 131. Chu QD, Medeiros LJ, Fisher AE et al. Thrombotic thrombocytopenic purpura and HIV infection. Southern Medical Journal 1995;88:82-6. 132. Qadri SMH, Kayali S. Enterohemorrhagic Escherichia coli: A dangerous food-borne pathogen. Postgraduate Medicine 1998;103:179-87. 133. Laing RW, Teh C, Toh CH. Thrombotic thrombocytopenic purpura (TTP) complicating leptospirosis: A previously undescribed association. Journal of Clinical Pathology 1990;43:961-2. 134. Ackerman AB. Histologic Diagnosis of Inflammatory Skin Diseases 2nd Ed 1997. 135. Lie JT. Vasculitis associated with infectious agents. Current Opinion in Rheumatology 1996;8:26-9. 136. Guillevin L, Lhote F, rardi R. The spectrum and treatment of virus-associated vasculitides. Current Opinion in Rheumatology 1997;9:31-6. 137. Golden MP, Hammer SM, Wanke CA et al. Cytomegalovirus vasculitis. Case reports and review of the literature. Medicine 1994;73:246-55. 138. Ho DD, Rota TR, Andrews CA et al. Replication of human cytomegalovirus in endothelial cells. Journal of Infectious Diseases 1984;150:956-7. 139. Goodman MD, Porter DD. Cytomegalovirus vasculitis with fatal colonic hemorrhage. Archives of Pathology and Laboratory Medicine 1973;96:281-4. 140. Foucar E, Mukai K, Foucar K. Colon ulceration in lethal cytomegalovirus infection. American Journal of Clini- cal Pathology 1981;76:788-801. 141. Booss J, Dann PR, Winkler SR et al. Mechanisms of injury to the central nervous system following experimental cytomegalovirus infection. American Journal of Otolaryngology - Head and Neck Medicine and Surgery 1990;11:313-7. 142. Koeppen AH, Lansing LS, Peng SK et al. Central nervous system vasculitis in cytomegalovirus infection. Journal of the Neurological Sciences 1981;51:395-410. 143. Lin CS, Penha PD, Krishnan MN et al. Cytomegalic inclusion disease of the skin. Archives of Dermatology 1981;117:282-4. 144. Libman BS, Quismorio J, Stimmler MM. Polyarteritis nodosa-like vasculitis in human immunodeficiency virus infection. Journal of Rheumatology 1995;22:351-5. 145. Calabrese LH. Vasculitis and infection with the human immunodeficiency virus. Rheumatic Disease Clinics of North America 1991;17:131-47. 146. Gherardi R, Belec L, Mhiri C et al. The spectrum of vasculitis in human immunodeficiency virus-infected patients: A clinicopathologic evaluation. Arthritis and Rheumatism 1993;36:1164-74. 147. Sergent JS, Lockshin MD, Christian CL et al. Vasculitis with hepatitis B antigenemia: long term observations in nine patients. Medicine 1976;55:1-18. 148. Carson CW, Conn DL, Czaja AJ et al. Frequency and significance of antibodies to hepatitis C virus in polyarteritis nodosa. Journal of Rheumatology 1993;20:304-9. 149. Leruez-Ville M,Lauge A, Morinet F et al. Polyarteritis nodosa and parvovirus B19. Lancet 1994;344:263-4. 150. Nikkari S, Mertsola J, Korvenranta H et al. Wegener’s granulomatosis and parvovirus B19 infection. Arthritis and Rheumatism 1994;37:1707-8. 151. Yoto Y, Kudoh T, Haseyama K et al. Human parvovirus B19 infection in Kawasaki disease. Lancet 1994;344:58-9. 152. Stemerman MB, Colton C, Morell E. Perturbations of the endothelium. Progress in hemostasis and thrombosis 1984;7:289-324. 153. Kaiser L, Sparks J. Endothelial cells. Not just a cellophane wrapper. Archives of Internal Medicine 1987;147:569-73.

Chapter 2

40

154. Friedman HM, Wolfe J, Kefalides NA et al. Susceptibility of endothelial cells derived from different blood vessels to common viruses. In Vitro 1986;22:397-401. 155. Friedman HM. Infection of endothelial cells by common human viruses. Reviews of Infectious Diseases 1989;11 Suppl 4. 156. Hoxie JA, Matthews DM, Cines DB. Infection of human endothelial cells by human T-cell leukemia virus type I. Proceedings of the National Academy of Sciences of the United States of America 1984;81:7591-5. 157. Wiley CA, Schrier RD, Nelson JA. Cellular localization of human immunodeficiency virus infection within the brains of acquired immune deficiency syndrome patients. Proceedings of the National Academy of Sciences of the United States of America 1986;83:7089-93. 158. Butthep P, Bunyaratvej A, Bhamarapravati N. Dengue virus and endothelial cell: a related phenomenon to thrombocytopenia and granulocytopenia in dengue hemorrhagic fever. The Southeast Asian journal of tropical medicine and public health 1993;24 Suppl 1:246-9. 159. Van Dam-Mieras MCE, Muller AD, Van Hinsbergh VWM et al. The procoagulant response of cytomegalovirus infected endothelial cells. Thrombosis and Haemostasis 1992;68:364-70. 160. Etingin OR, Silverstein RL, Friedman HM et al. Viral activation of the coagulation cascade: Molecular inter actions at the surface of infected endothelial cells. Cell 1990;61:657-62. 161. Dudding L, Haskill S, Clark BD et al. Cytomegalovirus infection stimulates expression of monocyte-associated mediator genes. Journal of Immunology 1989;143:3343-52. 162. Smith PD, Saini SS, Raffeld M et al. Cytomegalovirus induction of tumor necrosis factor-a by human monocytes and mucosal macrophages. Journal of Clinical Investigation 1992;90:1642-8. 163. McEver RP. GMP-140: A receptor for neutrophils and monocytes on activated platelets and endothelium. Journal of Cellular Biochemistry 1991;45:156-61. 164. Takahashi M, Ikeda U, Masuyama JI et al. Monocyte-endothelial cell interaction induces expression of adhesion molecules on human umbilical cord endothelial cells. Cardiovasc Res 1996;32:422-9. 165. Sugerman RW, Church JA, Goldsmith JC et al. Acquired protein S deficiency in children infected with human immunodeficiency virus. Pediatric Infectious Disease Journal 1996;15:106-11. 166. Stahl CP, Wideman CS, Spira TJ et al. Protein S deficiency in men with long-term human immunodeficiency virus infection. Blood 1993;81:1801-7. 167. Hassell KL, Kressin DC, Neumann A et al. Correlation of antiphospholipid antibodies and protein S deficiency with thrombosis in HIV-infected men. Blood Coagulation and Fibrinolysis 1994;5:455-62. 168. Van Gorp ECM, Minnema MC, Suharti C et al. Activation of coagulation factor XI, without detectable contact activation in dengue haemorrhagic fever. British Journal of Haematology 2001;113:94-9. 169. Levin J. Bleeding with infectious diseases. Disorders of Hemostasis 1984;367-78. 170. Nakao S, Lai CJ, Young NS. Dengue virus, a flavivirus, propagates in human bone marrow progenitors and hematopoietic cell lines. Blood 1989;74:1235-40. 171. Halstead SB. Antibody, macrophages, dengue virus infection, shock, and hemorrhage: a pathogenetic cascade. Reviews of Infectious Diseases 1989;11 Suppl 4. 172. Terada H, Baldini M, Ebbe S et al. Interaction of influenza virus with blood platelets. Blood 1966;28:213-28. 173. Curwen KD, Gimbrone J, Handin RI. In vitro studies of thromboresistance. The role of prostacyclin (PGI2) in platelet adhesion to cultured normal and virally transformed human vascular endothelial cells. Laboratory Investigation 1980;42:366-74. 174. Morse EE, Zinkham WH, Jackson DP. Thrombocytopenic purpura following rubella infection in children and adults. Archives of Internal Medicine 1966;117:573-9. 175. Brook I. Disseminated varicella with pneumonia, meningoencephalitis, thrombocytopenia, and fatal intracranial hemorrhage. Southern Medical Journal 1979;72:756-7. 176. Myllyla G, Vaheri A, Vesikari T et al. Interaction between human blood platelets, viruses and antibodies. IV. Post-Rubella thrombocytopenic purpura and platelet aggregation by Rubella antigen-antibody interaction. Clinical and Experimental Immunology 1969;4:323-32. 177. Charkes ND. Purpuric chickenpox: Report of a case, review of the literature, and classification by clinical features. Ann Intern Med 1961;54:745-59. 178. Tobin J, Ten Bensel RW. Varicella with thrombocytopenia causing fatal intracerebral hemorrhage. American Journal of Diseases of Children 1972;124:577-8. 179. Chanarin I, Walford DM. Thrombocytopenic purpura in cytomegalovirus mononucleosis. Lancet 1973;2:238-9. 180. Fourrier F, Chopin C, Huart JJ et al. Double-blind, placebo-controlled trial of antithrombin III concentrates in septic shock with disseminated intravascular coagulation. Chest 1993;104:882-8. 181. Warren BL, Eid A, Singer P et al. High-dose antithrombin III in severe sepsis: A randomized controlled trial. Journal of the American Medical Association 2001;286:1869-78. 182. Elsayed YA, Nakagawa K, Kamikubo YI et al. Effects of recombinant human tissue factor pathway inhibitor on thrombus formation and its in vivo distribution in a rat DIC model. American Journal of Clinical Pathology 1996;106:574-83.


41

part

IInflammatIon and CoagulatIon as

CardIovasCular rIsk faCtors

Chapter

3serum levels of mannose-bIndIng leCtIn and

the rIsk of future Coronary artery dIsease In apparently healthy men and women

Tymen Keller • Sander van Leuven • Marijn Meuwese • Nicholas wareham • Robert Luben Erik Stroes • Erik Hack • Marcel Levi • Kay-Tee Khaw • Matthijs Boekholdt

Departments of Vascular Medicine and Cardiology, Academic Medical Center, Amsterdam, The NetherlandsMedical Research Council Epidemiology Unit and the Department of Public Health and Primary Care, Institute of Public Health,

University of Cambridge, Cambridge, United KingdomDepartment of Clinical Chemistry, VU Medical Center, Amsterdam, The Netherlands

Arterioscler Thromb Vasc Biol 2006;26:2345-50

ABSTRACTTo determine the association between serum levels of mannose-binding lectin (MBL) and the risk of future coronary artery disease (CAD) in apparently healthy men and women. We performed a prospective case-control study among apparently healthy men and wom-en nested in the EPIC-Norfolk cohort. Baseline concentrations of MBL were measured in serum samples of 946 patients who experienced a myocardial infarction or died of CAD during follow-up, and 1799 matched controls who remained free of CAD. Among men, median MBL levels were 1.63 ng/mL (IQR: 0.59 - 3.80) in cases and 1.20 ng/mL (IQR: 0.48 - 3.37) in controls. Among women, median MBL levels were 1.02 ng/mL (IQR: 0.43 - 2.95) in cases and 1.01 ng/mL (IQR: 0.43 - 2.94) in controls. After adjustment, the odds ratio in men for future CAD was 1.59 (95% confidence interval [CI]: 1.09-2.32; P for lin-earity = 0.01) for those in the highest quartile compared to those in the lowest quartile. In women no such relation was observed. Elevated levels of MBL are associated with an increased risk of future CAD in apparently healthy men but not in women. The sex differ-ence merits further exploration.

Chapter 3

46

INTRODuCTIONInflammation plays a major role in all phases of atherogenesis from plaque initiation to plaque rupture. Several inflammatory markers have been associated with an increased risk of atherosclerotic vascular disease, including CRP,1 SPLA2,2 and IL-6.3 In addition, markers of innate immunity have been shown to predict the development of CAD.4-7 However, prospective evidence for a causative role of inflammatory factors is still limited.8 Recently, evidence has emerged concerning the role of mannose-binding lectin (MBL) in the development of atherosclerosis. MBL is a part of the complement cascade and plays an important role in the first line of defense of the innate immune system against patho-genic micro-organisms.9, 10 MBL recognizes sugar patterns on the surface of many patho-gens11, phospholipids,12 immune complexes13 and apoptotic cells.14 In the circulation, MBL forms a complex with MBL-associated serine proteases (MASPs). This complex becomes enzymatically active and activates the classical complement route. This facilitates comple-ment-dependent opsonization and subsequent uptake and clearance by phagocytes.11 In addition, there are indications that MBL binds directly to granulocytes, monocytes and macrophages, which may stimulate the production of pro-inflammatory cytokines.15 Since innate immunity has been implicated in atherogenesis, MBL has been suggested to play a role in the formation of atherosclerotic plaque. However, studies examining the relations between either serum levels of MBL or MBL genetic variants associated with low serum levels of MBL and CAD risk have reported equivocal results.7, 16-21 22-24 At present, no con-clusive data are available about the relationship between serum MBL levels and future CAD risk in healthy individuals. We hypothesized that MBL levels are associated with an increased risk of future CAD in healthy individuals. We tested this hypothesis in a large prospective case-control study nested in the European Prospective Investigation into Can-cer in Norfolk (EPIC-Norfolk) prospective population study. We measured serum levels of MBL in apparently healthy men and women at baseline and assessed the development of CAD during a follow-up period of 6 years.

METHODSThe EPIC study (EPIC, European Prospective Investigation into Cancer and Nutrition), a collaborative study of nine countries in Europe, was designed to assess the determi-nants of cancer and other diseases. The EPIC-Norfolk cohort which is a part of the EPIC study has been described in detail previously.25 In brief, investigators recruited 25,663 men and women, all residents in Norfolk (United Kingdom), from general practices and performed a baseline survey between 1993 and 1997. Trained nosologists obtained vital status of the entire cohort based on death certificates of the United Kingdom Office of National Statistics and linkage of the unique National Health Service number linkage with the East Norfolk Health Authority (ENCORE) database, which identifies all hospi-tal contacts throughout England and Wales for Norfolk residents. Follow-up information was obtained up to January 2003, an average of about 6 years. The study was approved by the Norwich District Health Authority Ethics Committee and all participants gave in-formed consent. We performed a nested case-control study of coronary artery disease in the EPIC-Norfolk cohort. The design, methods and validation of this CAD case control

Serum levels of mannose-binding lectin and the risk of future coronary artery disease in apparently healthy men and women

47

study have been described in detail elsewhere.2 In brief, we identified apparently healthy participants from the cohort who developed a myocardial infarction or died of coronary heart disease. Two controls were matched to these cases based on sex, age (within 5 years), data of visit (within 3 months) and duration of follow-up time in the study. Controls were those who remained alive and did not had a myocardial infarction during follow-up. For this nested study we excluded participants from the EPIC-Norfolk cohort who reported a history of myocardial infarction or stroke at the baseline visit.

Biochemical analyses

Qualified staff collected blood from all participants at the baseline visit. The Department of Clinical Biochemistry, University of Cambridge processed the samples for assay. Serum levels of total cholesterol, high-density lipoprotein cholesterol (HDL-c), and triglycerides were measured on fresh samples with the RA 1000 (Bayer Diagnostics, Basingstoke, UK), and low-density lipoprotein cholesterol (LDL-c) levels were calculated with the Friedewald formula.26 In addition, serum samples were stored at –80˚C for future analyses. We mea-sured serum concentrations of C-reactive protein (CRP) with a sandwich-type enzyme linked immunosorbent assay (ELISA) as previously described,27 and related the results of this ELISA to a standard consisting of commercially available CRP (Behringwerke AG, Marburg, Germany). We measured serum MBL levels with a commercially available ELI-SA kit from Sanquin Research (Amsterdam, The Netherlands). We incubated the samples in mannan coated plates. After washing, we visualised the binding of MBL by incubation with biotinylated monoclonal antibody against MBL (CLB anti-MBL-1). Samples were ana-lyzed in random order and researchers and laboratory personnel were blinded as to case status of the samples.

Statistical analysis

MBL levels had a skewed distribution and were therefore log-transformed before being used in statistical analyses, but in tables we present untransformed medians and cor-responding interquartile ranges (IQR). MBL quartiles were based on the sex-specific dis-tribution among the controls. Sex-specific analyses were performed using sex-specific quartile cut-offs. In order to estimate the relative risk of future CAD, we calculated odds ratios (OR) and corresponding 95% confidence intervals (95%CI) per MBL quartile and a p-value for linearity across quartiles. ORs were calculated using conditional logistic re-gression, taking into account the matching for sex and age and enrollment time. ORs were adjusted for cardiovascular risk factors which were significantly related to MBL lev-els. In a second regression model, ORs were additionally adjusted for serum levels of CRP. The lowest quartile was used as the reference category.

Chapter 3

48

RESuLTSMBL levels in the study population

Table 1 shows the sex specific distribution of risk factors at baseline for cases and con-trols. As expected, levels of risk factors including blood pressure, total cholesterol, LDL-c, triglycerides, CRP, diabetes, cigarette smoking habit and BMI were higher and HDL-c lower in those who developed CAD during follow-up compared to those who remained free from CAD. Among men, median MBL levels were higher in cases when compared to controls. Among women, median MBL levels were similar in cases and controls.

TABLE 1: STuDy POPuLATION Men Controls Cases

(n = 958) (n = 486)

Age, year 64 ± 8 65 ± 8

Smoking – Current 8.3 (78) 16.6 (80)

– Former 61.7 (583) 58.0 (279)

– Never 30.1 (284) 25.4 (122)

Body mass index, kg/m2 26.4 ± 3.2 27.1 ± 3.4

Diabetes 2.3 (22) 7.2 (35)

Systolic blood pressure, mmHg 140 ± 17 144 ± 18

Diastolic blood pressure, mmHg 85 ± 11 86 ± 12

Total cholesterol, mmol/l 6.0 ± 1.0 6.3 ± 1.2

LDL cholesterol, mmol/l 3.9 ± 0.9 4.1 ± 1.0

HDL cholesterol, mmol/l 1.25 ± 0.34 1.16 ± 0.30

Triglycerides, mmol/l 1.7 (1.2 - 2.4) 2.0 (1.4 - 2.9)

C-reactive protein, mg/l 1.4 (0.7 – 2.7) 2.2 (1.0 - 4.5)

MBL, ng/ml 1.20 (0.48 - 3.37) 1.63 (0.59 - 3.80)

women

(n = 547) (n = 294)

Age, year 67 ± 7 67 ± 7

Smoking – current 7.0 (38) 15.1 (44)

– Former 36.3 (197) 37.1 (108)

– Never 56.6 (307) 47.8 (139)


Diabetes 1.3 (7) 5.4 (16)







C-reactive protein, mg/l 1.6 (0.7 - 3.7) 2.6 (1.1 - 5.1)

MBL, ng/ml 1.01 (0.43 - 2.94) 1.02 (0.43 - 2.95)

Data are presented as mean ± SD, % (n), or median (interquartile range). LDL indicates low-density lipoprotein; HDL indicates high-density lipoprotein; MBL indicates mannose binding lectin. Means, percentages and medians may be based on fewer observations than the indicated number of subjects.


49

Chapter 3

50

TABLE 2: CORRELATION BETwEEN RISK FACTORS AND MBL LEvELS

Quartile 1 2 3 4 P * R P †

Men

MBL range, ng/ml < 0.48 0.49- 1.19 1.20 - 3.37 > 3.38

Case / control 103 / 240 103 / 238 137 / 242 143 / 238

Age, year 64 ± 8 65 ± 8 64 ± 8 64 ± 8 0.7 0.04 0.9

Smoking - Current 12.9 (44) 7.5 (25) 10.2 (38) 13.5 (51)

- Former 60.7 (207) 65.3 (218) 57.2 (214) 59.2 (223) 0.9

- Never 26.4 (90) 27.2 (91) 32.6 (122) 27.3 (103)

Body mass index, kg/m2 26.7± 3.2 26.8± 3.2 26.7±3.2 26.6 ± 3.5 0.7 -0.01 0.8

Diabetes 2.9 (10) 3.2 (11) 5.8 (22) 3.7 (14) 0.2

Systolic blood pressure, mmHg 141 ± 17 143 ± 18 141 ± 18 141 ± 18 0.8 -0.01 0.7

Diastolic blood pressure, mmHg 85 ± 11 86 ± 12 85 ± 11 85 ± 11 0.3 -0.02 0.6

Total cholesterol, mmol/l 6.2 ± 1.1 6.2 ± 1.1 6.1 ± 1.1 6.1 ± 1.1 0.3 -0.03 0.2

LDL cholesterol, mmol/l 4.0 ± 1.0 4.0 ± 1.0 4.0 ± 0.9 4.0 ± 0.9 0.6 -0.02 0.5

HDL cholesterol, mmol/l 1.2 ± 0.4 1.2 ± 0.3 1.2 ± 0.3 1.2 ± 0.3 0.2 -0.03 0.2

Triglycerides, mmol/l 2.1 ± 1.0 2.2 ± 1.3 2.1 ± 1.2 2.0 ± 1.1 0.7 -0.01 0.6

C-reactive protein, mg/l 3.1 ± 5.1 2.7 ± 3.7 3.9 ± 5.9 3.9 ± 6.4 0.01 0.07 0.01

women

MBL range, ng/ml < 0.42 0.43 - 1.01 1.02 – 2.92 > 2.94

Case / control 72 / 134 74 / 141 74 / 135 74 / 137

Age, year 66 ± 8 67 ± 7 67 ± 7 66 ± 7 0.8 0.03 0.4

Smoking - Current 11.2 (23) 8.0 (17) 11.7 (24) 8.5 (18)

- Former 37.1 (76) 39.2 (83) 38.5 (79) 31.8 (67) 0.3

- Never 51.7 (106) 52.8 (112) 49.8 (102) 59.7 (126)

Body mass index, kg/m2 27.1 ± 4.6 27.0 ± 4.1 26.4 ± 4.0 26.2 ± 4.3 0.03 -0.07 0.04

Diabetes 1.9 (4) 2.3 (5) 2.9 (6) 3.8 (8) 0.5

Systolic blood pressure, mmHg 140 ± 19 140 ± 19 142 ± 19 140 ± 20 0.9 -0.01 0.8

Diastolic blood pressure, mmHg 83 ± 11 83 ± 11 84 ± 12 83 ± 12 0.6 0.00 1.0

Total cholesterol, mmol/l 6,8 ± 1.3 6.8 ± 1.2 6.7 ± 1.1 6.8 ± 1.2 0.6 0.01 0.8

LDL cholesterol, mmol/l 4.4 ± 1.2 4.3 ± 1.1 4.3 ± 1.0 4.4 ± 1.0 0.9 0.02 0.7

HDL cholesterol, mmol/l 1.6 ± 0.4 1.5 ± 0.4 1.5 ± 0.4 1.5± 0.4 0.6 -0.01 0.7


C-reactive protein, mg/l 3.5 ± 5.0 4.1 ± 6.5 4.4 ± 9.2 4.2 ± 6.5 0.4 0.03 0.4

Data are presented as mean ± SD per MBL quartile. MBL indicates mannose binding lectin; LDL indicates low-density lipoprotein; HDL indicates high-density lipoprotein; P* indicates p-value for linearity between the MBL quartile and risk factor levels; R indicates Pearsons’ correlation between log-transformed MBL levels and risk factor levels; P† indicates the corresponding P value. Because of their skewed distribution, triglycerides, CPR, and MBL levels were log-transformed before analysis as continuous variables.

MBL levels and other risk factors

Table 2 shows the relationship between cardiovascular risk factors and MBL quartiles. MBL levels were correlated with sex, HDL-c and CRP levels. These correlations were con-sistent over the MBL quartile. Among men, MBL levels were correlated with CRP levels. Among women, no such correlation with CRP was noted, but a significant correlation was seen with BMI. There was no significant relationship between MBL levels and other cardiovascular risk factor, including age, history of diabetes, smoking, blood pressure, LDL-c, and triglycerides.

MBL levels and the risk of future CAD

Table 3 presents the odds ratios for future CAD according to quartile of MBL among men and women (Table 3). In men, after adjustment for determinants of cardiovascular risk and MBL levels, the risk of future CAD was 1.55 (95% CI: 1.08 - 2.22) for men in the high-est MBL quartile compared to those in the lowest (P for linearity = 0.01). This remained significant after additional adjustment for CRP. Figure 1 shows odds ratios for future CAD in men stratified by quartiles of CRP and MBL. The relationship between MBL levels and future CAD was solely accounted for by men, whereas in women no such a relation was found (P for linearity = 0.2).


51

TABLE 3: RISK OF CAD AND MBL QuARTILES

MBL quartile 1 2 3 4 P linearity

Men

MBL range, ng/ml < 0.48 0.49- 1.19 1.20 - 3.37 > 3.38

Cases / controls 103 / 240 103 / 238 137 / 242 143 / 238

OR (95% CI) (1) 1.00 1.17 (0.80-1.70) 1.43 (1.01-2.04) 1.55 (1.08-2.22) 0.01

OR (95% CI) (2) 1.00 1.16 (0.79-1.71) 1.38 (0.95-2.00) 1.59 (1.09-2.32) 0.01

women

MBL range, ng/ml < 0.42 0.43 – 1.00 1.01 – 2.83 > 2.84

Cases / controls 71 / 134 74 / 141 74 / 135 74 / 137

OR (95% CI) (1) 1.00 0.91 (0.56-1.46) 1.08 (0.69-1.70) 0.98 (0.61 -1.56) 0.9

OR (95% CI) (2) 1.00 0.70 (0.40-1.21) 1.06 (0.62-1.79) 0.82 (0.47 -1.43) 0.8

Odds ratios for future CAD were calculated by conditional logistic regression with corresponding 95% confidence intervals per quartile. MBL quartiles were based on the sex-specific distribution among the controls. Because of their skewed distribution, CRP levels were log-trans-formed before analysis as continuous variables. In model 1 adjustment was done for determinant of cardiovascular risk and MBL levels, exclud-ing CRP (1). Model 2 was additionally adjusted for CRP (2).

DISCuSSIONIn this large prospective study among apparently healthy people, we showed that in men high serum levels of MBL are associated with an increased risk of future coronary artery disease. This relationship was independent of established cardiovascular risk factors. This observation suggests that MBL levels may reflect or contribute to a pathophysiological process relevant in the development of atherosclerosis. Recently, evidence emerged about a role of MBL in CAD. In experimental models it was shown that the MBL-pathway is involved in ischemia-induced complement activation in mice.28 Consequently, the administration of anti-MBL antibodies protects the heart from ischemia-reperfusion injury by reducing neutrophil infiltration and attenuating pro-inflammatory gene expression.29 In humans, it was recently shown that high MBL levels are associated with a high incidence of re-stenosis in patients with atherosclerotic disease of the carotid artery.18 Moreover, type I diabetics with a history of cardiovascular dis-ease had significantly elevated MBL levels when compared to type I diabetics without vascular complication.21 In our study, we were able to show an increased risk for CAD in apparently healthy men with high MBL levels. In women however, no such an association was seen. Sex differences with regard to MBL levels and cardiovascular risk have been reported before but the causes are presently unknown.18, 24 Endocrine status plays an important role in the regula-tion of MBL levels.30, 31 Hormonal differences and cardiovascular disease are an ongoing topic of research. Anti-inflammatory properties of estrogen have sparked much interest and may comprise complement activation. Interestingly, endothelial tissue from males may be more susceptible than that from females to the acute effects of complement activation.32 Differences in the expression of complement components in adipose tissue of men and women have also been observed.33 Furthermore, we observed significant differences in CRP levels among the MBL quartiles with highest CRP levels in the highest MBL quartile. High CRP and MBL levels could both be a sign of an inflammatory state. MBL concentrations are normally stable within one person, but can rise threefold during the acute phase response. Therefore, the differences in MBL levels between male cases and controls could be partially attributable to the differences in the inflammatory state of these individuals. However, the relationship between MBL levels and future CAD risk persisted upon additional adjustment for CRP.

Chapter 3

52

12

34

0

0 , 5

1

1 , 5

2

2 , 5

3

Adjusted OR for future CAD

C R P q u a r ti l e s

1

2 3

4

MBL quartiles

CRP quartiles

MBL quartiles

FIguRE 1: RISK OF CAD IN APPARENTLy HEALTHy MEN STRATIFIED By QuARTILES OF CRP AND MBL

Bars represent adjusted odds ratios for future CAD in apparently healthy men stratified by quartiles of CRP and MBL. Odds ratios were calculated by conditional logistic regression analysis. Adjustment was done for determinants of car-diovascular risk and MBL levels. Men in the lowest quartile for both CRP and MBL were used as the reference group.

Earlier studies regarding the role of MBL in atherogenesis in humans not conclusive. In one population-based study, DNA analysis in 434 apparently healthy American Indi-ans revealed that genotypes coding for diminished levels of MBL are predictive of CAD.7 However, measuring MBL protein levels may be more reliable than genotyping. MBL protein levels vary widely between individuals. This is partly influenced by variant MBL genotypes, but serum levels vary markedly between persons with an identical genotype.24 Depending on the antibody used however, different findings can be seen which may also contribute to the lack of consensus regarding the role of MBL in atherosclerotic vascular disease. In a nested case-control study in the Reykjavik cohort, Saevarsdottir showed that MBL levels are not associated with an increased risk of CAD, but in a sub-analysis high MBL levels were associated with a lower risk of CAD in diabetics.24 We observed several differences with our own study. Firstly, there is a difference in power, since we included twice as many patients in the study. Secondly, contrary to the former study we matched all cases for both age and sex and enrollment-time and subsequently we performed sex-spe-cific analysis. And finally, we measured also CRP levels which made it possible to correct for CRP levels during the statistical analysis.There are several mechanisms by which MBL can function atherogenic. MBL activates the complement system which has been implicated in atherogenesis and was recently shown to be associated with increased cardiovascular risk in patients with advanced ath-erosclerosis.4 Indeed, increased deposition of complement iC3b in ruptured and vulner-able plaques suggests a role for complement activation in acute coronary syndromes.34 Furthermore, endothelial oxidative stress, which plays a major role in atherogenesis, ac-tivates complement via the lectin complement pathway in human cell cultures. In addi-tion, in this experimental model, anti-MBL monoclonal antibodies inhibited MBL and C3 deposition after endothelial oxidative stress.35 MBL however, is not only involved in complement activation but also is a potent regulator of inflammatory pathways.36 Indeed, MBL was shown to enhance the production of chemokines by macrophages37 and may thereby enhance phagocyte recruitment to the subendothelium. Moreover, it was recently shown that MBL binds to leukocytes and can thereby modulate inflammation. Of note, this study suggested that such binding only occurred at extravascular sites in individuals possessing MBL genotypes conferring MBL sufficiency.38 This may help explain why high MBL levels can influence the atherosclerotic process in the subendothelium. These data support a role for MBL in the atherogenesis and possibly in plaque destabilization and the development of coronary events. A number of issues have to be taken into account when interpreting the results of the present study. Serum levels of MBL were determined in a single sample that was obtained at a nonuniform time of the day. Diurnal variation and variation over time could have affected the MBL concentrations, although such variation has been shown to be limited compared to other cardiovascular risk factors. In addition, we cannot rule out that sample storage at –80°C for 6 to 10 years may have affected the detection of MBL. However, both these limitations would lead to increased random mea-surement error, which leads to an underestimation of any relationship, and therefore do not negate our findings. Although the current study was not designed to establish wheth-er the relationship between MBL and CAD is causal, it is unlikely that differences in MBL


53

serum levels occurred as a consequence of cardiovascular events because individuals with symptomatic cardiovascular disease were excluded from our analysis. However, we can-not exclude the possibility that MBL concentration is a marker of advanced subclinical atherosclerosis. We conclude that in this nested case control study elevated serum concentrations of MBL are associated with an increased risk of future CAD in apparently healthy men. This rela-tionship was independent of other cardiovascular risk factors. These prospective data sup-port the hypothesis that the immune system plays an important role in the development of atherosclerosis and its prominent clinical manifestation: coronary artery disease. The lack of association in women warrants further investigation and may provide insights into possible explanations for the sex difference in coronary heart disease susceptibility.

ACKNOwLEDgEMENTSEPIC-Norfolk is supported by programmed grants from the Medical Research Council UK and Cancer Research UK with additional support from the European Union, Stroke Association, British Heart Foundation, Department of Health, Food Standards Agency and the Wellcome Trust. NW, RL and KTK designed the study. NW, RL, KTK, TK, MM and EH retrieved the data. TK and MB performed the statistical analysis. TK, SL and MB analysed and interpreted the data. TK and SL drafted the manuscript. MB, ES, ML, NW, RL and KTK revised the manuscript. We thank the participants, general practitioners and staff in EPIC-Norfolk. None of the authors have a conflict of interest.

Chapter 3

54

Reference List 1. Danesh J, Wheeler JG, Hirschfield GM et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med 2004;350:1387-97. 2. Boekholdt SM, Keller TT, Wareham NJ et al. Serum levels of type II secretory phospholipase A2 and the risk of future coronary artery disease in apparently healthy men and women: the EPIC-Norfolk Prospective Population Study. Arterioscler Thromb Vasc Biol 2005;25:839-46. 3. Ridker PM, Hennekens CH, Buring JE et al. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med 2000;342:836-43. 4. Speidl WS, Exner M, Amighi J et al. Complement component C5a predicts future cardiovascular events in patients with advanced atherosclerosis. Eur Heart J 2005;26:2294-9. 5. Szeplaki G, Prohaszka Z, Duba J et al. Association of high serum concentration of the third component of complement (C3) with pre-existing severe coronary artery disease and new vascular events in women. Atherosclerosis 2004;177:383-9. 6. Kostner KM, Fahti RB, Case C et al. Inflammation, complement activation and endothelial function in stable and unstable coronary artery disease. Clin Chim Acta 2005. 7. Best LG, Davidson M, North KE et al. Prospective analysis of mannose-binding lectin genotypes and coronary artery disease in American Indians: the Strong Heart Study. Circulation 2004;109:471-5. 8. Binder CJ, Chang MK, Shaw PX et al. Innate and acquired immunity in atherogenesis. Nat Med 2002;8:1218-26. 9. Gadjeva M, Takahashi K, Thiel S. Mannan-binding lectin--a soluble pattern recognition molecule. Mol Im- munol 2004;41:113-21. 10. Turner MW. Mannose-binding lectin: the pluripotent molecule of the innate immune system. Immunol Today 1996;17:532-40. 11. Petersen SV, Thiel S, Jensenius JC. The mannan-binding lectin pathway of complement activation: biology and disease association. Mol Immunol 2001;38:133-49. 12. Kilpatrick DC. Phospholipid-binding activity of human mannan-binding lectin. Immunol Lett 1998;61:191-5. 13. Roos A, Bouwman LH, van Gijlswijk-Janssen DJ et al. Human IgA activates the complement system via the mannan-binding lectin pathway. J Immunol 2001;167:2861-8. 14. Ogden CA, deCathelineau A, Hoffmann PR et al. C1q and mannose binding lectin engagement of cell surface calreticulin and CD91 initiates macropinocytosis and uptake of apoptotic cells. J Exp Med 2001;194:781-95. 15. Ghiran I, Barbashov SF, Klickstein LB et al. Complement receptor 1/CD35 is a receptor for mannan-binding lectin. J Exp Med 2000;192:1797-808. 16. Hegele RA, Ban MR, Anderson CM et al. Infection-susceptibility alleles of mannose-binding lectin are associated with increased carotid plaque area. J Investig Med 2000;48:198-202. 17. Madsen HO, Videm V, Svejgaard A et al. Association of mannose-binding-lectin deficiency with severe atherosclerosis. Lancet 1998;352:959-60. 18. Rugonfalvi-Kiss S, Dosa E, Madsen HO et al. High rate of early restenosis after carotid eversion endarterectomy in homozygous carriers of the normal mannose-binding lectin genotype. Stroke 2005;36:944-8. 19. Ohlenschlaeger T, Garred P, Madsen HO et al. Mannose-binding lectin variant alleles and the risk of arterial thrombosis in systemic lupus erythematosus. N Engl J Med 2004;351:260-7. 20. Dahl M, Tybjaerg-Hansen A, Schnohr P et al. A population-based study of morbidity and mortality in mannose-binding lectin deficiency. J Exp Med 2004;199:1391-9. 21. Hansen TK, Tarnow L, Thiel S et al. Association between mannose-binding lectin and vascular complications in type 1 diabetes. Diabetes 2004;53:1570-6. 22. Albert MA, Rifai N, Ridker PM. Plasma levels of cystatin-C and mannose binding protein are not associated with risk of developing systemic atherosclerosis. Vasc Med 2001;6:145-9. 23. Limnell V, Aittoniemi J, Vaarala O et al. Association of mannan-binding lectin deficiency with venous bypass graft occlusions in patients with coronary heart disease. Cardiology 2002;98:123-6. 24. Saevarsdottir S, Oskarsson OO, Aspelund T et al. Mannan binding lectin as an adjunct to risk assessment for myocardial infarction in individuals with enhanced risk. J Exp Med 2005;201:117-25. 25. Day N, Oakes S, Luben R et al. EPIC-Norfolk: study design and characteristics of the cohort. European Prospective Investigation of Cancer. Br J Cancer 1999;80 Suppl 1:95-103. 26. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499-502. 27. Bruins P, te VH, Yazdanbakhsh AP et al. Activation of the complement system during and after cardiopulmonary bypass surgery: postsurgery activation involves C-reactive protein and is associated with post operative arrhythmia. Circulation 1997;96:3542-8. 28. de VB, Walter SJ, Peutz-Kootstra CJ et al. The mannose-binding lectin-pathway is involved in complement activation in the course of renal ischemia-reperfusion injury. Am J Pathol 2004;165:1677-88. 29. Jordan JE, Montalto MC, Stahl GL. Inhibition of mannose-binding lectin reduces postischemic myocardial reperfusion injury. Circulation 2001;104:1413-8. 30. Riis AL, Hansen TK, Thiel S et al. Thyroid hormone increases mannazn-binding lectin levels. Eur J Endocrinol 2005;153:643-9. 31. Gravholt CH, Leth-Larsen R, Lauridsen AL et al. The effects of GH and hormone replacement therapy on


55

serum concentrations of mannan-binding lectin, surfactant protein D and vitamin D binding protein in Turner syndrome. Eur J Endocrinol 2004;150:355-62. 32. Lennon PF, Collard CD, Morrissey MA et al. Complement-induced endothelial dysfunction in rabbits: mechanisms, recovery, and gender differences. Am J Physiol 1996;270:H1924-H1932. 33. Gabrielsson BG, Johansson JM, Lonn M et al. High expression of complement components in omental adipose tissue in obese men. Obes Res 2003;11:699-708. 34. Oksjoki R, Jarva H, Kovanen PT et al. Association between complement factor H and proteoglycans in early human coronary atherosclerotic lesions: implications for local regulation of complement activation. Arterioscler Thromb Vasc Biol 2003;23:630-6. 35. Collard CD, Vakeva A, Morrissey MA et al. Complement activation after oxidative stress: role of the lectin complement pathway. Am J Pathol 2000;156:1549-56. 36. Jack DL, Jarvis GA, Booth CL et al. Mannose-binding lectin accelerates complement activation and increases serum killing of Neisseria meningitidis serogroup C. J Infect Dis 2001;184:836-45. 37. Nadesalingam J, Dodds AW, Reid KB et al. Mannose-binding lectin recognizes peptidoglycan via the Nacetyl glucosamine moiety, and inhibits ligand-induced proinflammatory effect and promotes chemokine production by macrophages. J Immunol 2005;175:1785-94. 38. Downing I, MacDonald SL, Turner ML et al. Detection of an autologous ligand for mannan-binding lectin on human B lymphocytes. Scand J Immunol 2005;62:507-14.

Chapter 3

56

57


Chapter

4serum levels of type II seCretory

phospholIpase a2 and the rIsk of future Coronary artery dIsease In apparently

healthy men and women; the epIC-norfolk prospeCtIve populatIon study

Matthijs Boekholdt • Tymen Keller • Nicholas wareham • Robert Luben • Sheila Bingham Nicholas Day • Manjinder Sandhu • wouter jukema • john Kastelein

Erik Hack • Kay-Tee Khaw

Departments of Cardiology and Vascular Medicine, Academic Medical Center, Amsterdam, The NetherlandsDepartment of Cardiology, Leiden University Medical Center, Leiden, The Netherlands

Sanquin Research at the CLB and Department of Clinical Chemistry, VU Medical Center, Amsterdam, The NetherlandsMedical Research Councils Epidemiology Unit and Dunn Nutrition Unit and Department of Public Health and Primary Care,

Institute of Public Health, University of Cambridge, Cambridge, United Kingdom

Arterioscler Thromb Vasc Biol 2005;25:839-46

ABSTRACTType II secretory phospholipase A2 (sPLA2) has been implicated in coronary artery dis-ease (CAD) because it plays a regulatory role in several pathophysiological processes relevant in atherosclerosis. We conducted a nested case-control analysis based on a pro-spective population study of 25,663 men and women aged 45-79 years followed up 1993-2003. Cases (n = 1105) were apparently healthy men and women who developed fatal or non-fatal CAD during follow-up. Controls (n = 2209) were matched by age, sex and en-rolment time. sPLA2 levels were significantly higher in cases than controls (9.5 ng/ml, IQR : 6.4-14.8 versus 8.3 ng/ml IQR 5.8-12.6, P < 0.0001). sPLA2 levels were higher in women than men (11.0 ng/ml, IQR 7.5-16.9 versus 7.7 ng/ml, IQR 5.5-11.3, p < 0.0001). Among both men and women, sPLA2 plasma levels significantly correlated with age, body mass index, systolic blood pressure, HDL-c levels, and C-reactive protein levels. Taking into account matching for sex and age and adjusting for BMI, smoking, diabetes, systolic blood pressure, LDL-c, and HDL-c, and CRP levels, the risk of future CAD was 1.34 (1.02-1.71, p = 0.02) for people in the highest sPLA2 quartile, compared to those in the lowest (P for linearity = 0.03). Elevated levels of sPLA2 were associated with an increased risk of future CAD in apparently healthy individuals. The observed association was continuous and independent of classical cardiovascular risk factors. The magnitude of the association was similar to that observed between CRP and CAD risk, and both as-sociations were independent.

Chapter 4

60

INTRODuCTIONInflammation is increasingly considered to play an important role in atherosclerosis.1 Several inflammatory plasma markers including C-reactive protein (CRP),2 interleukin-6 (IL-6),3 and interleukin-8 (IL-8),4 have been shown to predict future coronary artery disease (CAD) in apparently healthy individuals. The retention of low-density lipoprotein (LDL) particles in the arterial wall and their subsequent modification and oxidation could well be the first and key stimulus that elicits this inflammatory process.Secretory phospholipase A2 (sPLA2) is a member of a family of intracellular and secretory enzymes that can hydrolyse the sn-2 ester bond of phospholipids of cell membranes and lipoproteins.5 sPLA2 has been implicated in atherogenesis in several ways. First, treat-ment with sPLA2 modifies LDL lipoproteins such that they have a higher affinity for extracellular matrix proteins6 7,8 resulting in an increased retention of LDL particles in the arterial wall, which is an early hallmark of atherosclerotic lesion progression.9-11 Second, the sPLA2-mediated hydrolysis of phospholipids yields among others, lysophospholipids and free fatty acids such as arachidonic acid, known precursors of various proinflamma-tory mediators such as leukotrienes and prostaglandines.12 Third, sPLA2 has been shown to yield lipoproteins that are more susceptible to lipid peroxidation13 and to generate more bioactive phospholipids.14 Circulating levels of sPLA2 are higher in patients who had documented CAD than in controls.15,16 In addition, a small study has shown that CAD patients with high sPLA2 levels were at an increased risk of developing recurrent CAD events during follow-up.15 However, thus far, this relationship has not been examined in a large prospective study. We hypothesized that among apparently healthy men and women, high serum concen-trations of sPLA2 are associated with an increased risk of future CAD. We tested this hypothesis in a large prospective case-control study nested in the European Prospective Investigation into Cancer in Norfolk (EPIC-Norfolk) prospective population study.

METHODSWe performed a nested case-control study among participants of the EPIC-Norfolk (Euro-pean Prospective Investigation into Cancer and Nutrition) study, a prospective population study of 25,663 men and women aged between 45 and 79 years, resident in Norfolk, UK, who completed a baseline questionnaire survey and attended a clinic visit.17 Participants were recruited from age-sex registers of general practices in Norfolk as part of the ten-country collaborative EPIC study designed to investigate dietary and other determinants of cancer. Additional data were obtained in EPIC-Norfolk to enable the assessment of determinants of other diseases.The design and methods of the study have been described in detail.17 In short, eligible participants were recruited by mail. At the baseline survey between 1993 and 1997, par-ticipants completed a detailed health and lifestyle questionnaire. Blood was taken by ve-nepuncture into plain and citrate bottles. Blood samples were processed for assay at the Department of Clinical Biochemistry, University of Cambridge, or stored at –80˚C. All in-dividuals have been flagged for death certification at the UK Office of National Statistics, with vital status ascertained for the entire cohort. In addition, participants admitted to

Serum levels of type II secretory phospholipase A2 and the risk of future coronary artery disease in apparently healthy men and women; the EPIC-Norfolk prospective population study

61

hospital were identified using their unique National Health Service number by data link-age with ENCORE (East Norfolk Health Authority database), which identifies all hospital contacts throughout England and Wales for Norfolk residents. Participants were identified as having CAD during follow-up if they had a hospital admission and/or died with CAD as underlying cause. CAD was defined as codes 410-414 according to the International Classification of Diseases 9th revision. We report results with follow-up up to January 2003, an average of about 6 years. The study was approved by the Norwich District Health Authority Ethics Committee and all participants gave signed informed consent.

Participants

We have previously described a similarly designed nested case-control study.4,18 Since then, an extended follow-up has resulted in the identification of additional CAD cases allowing the present study to be larger. We excluded all individuals who reported a his-tory of heart attack or stroke at the baseline clinic visit. Cases were 1105 individuals who developed a fatal or non-fatal CAD during follow-up until November 2003. Controls were study participants who remained free of any cardiovascular disease during follow-up. We attempted to match two controls to each case by sex, age (within 5 years), and time of en-rolment (within 3 months). A total of 2209 controls could be matched to cases.


Serum levels of total cholesterol, HDL cholesterol, and triglycerides were measured on fresh samples with the RA 1000 (Bayer Diagnostics, Basingstoke, UK), and LDL cho-lesterol levels were calculated with the Friedewald formula.19 Serum concentrations of sPLA2 were measured with a sandwich-type ELISA as previously described.20 The mean intra-assay variation between duplicates was 9.2% and the lower detection limit was 0.4 ng/ml. Plasma concentrations of CRP were measured with a sandwich-type ELISA as previously described.21 Results were related to a standard consisting of commercially available CRP (Behringwerke AG, Marburg, Germany). The lower detection limit was 0.1 mg/l. Samples were analyzed in random order to avoid systemic bias. Researchers and laboratory personnel had no access to identifiable information, and could identify samples by number only.


Baseline characteristics were compared between cases and controls taking into account the matching between them. A mixed effect model was used for continuous variables and conditional logistic regression was used for categorical variables. Because triglycerides, CRP and sPLA2 levels had a skewed distribution, values were log-transformed before be-ing used as continuous variables in statistical analyses but in the tables we show untrans-formed medians and corresponding interquartile ranges (IQR). In order to determine relationships between sPLA2 and traditional cardiovascular risk factors, we calculated mean risk factor levels per sPLA2 quartile. Quartiles were based on the distribution in the controls. For sex-specific analyses, sex-specific quartiles were used and for pooled analy-ses, we used quartiles based on the sexes combined. In addition, Pearson’s correlation co-

Chapter 4

62

63

TABLE 1: BASELINE CHARACTERISTICS OF CAD CASES AND MATCHED CONTROLS IN EPIC-NORFOLK 1993-2003

Men Controls Cases P

(n = 1396) (n = 707)

Age, year 64 ± 8 64 ± 8 Matched

Smoking - Never 31.8 (439) 24.6 (173)

- Previous 59.8 (824) 59.8 (420) < 0.0001

- Current 8.4 (116) 15.5 (109)

Body mass index, kg/m2 26.3 ± 3.1 27.3 ± 3.5 < 0.0001

Diabetes 2.4 (33) 7.1 (50) < 0.0001

Systolic blood pressure, mmHg 139 ± 17 144 ± 18 < 0.0001

Diastolic blood pressure, mmHg 85 ± 11 87 ± 11 < 0.0001

Total cholesterol, mmol/l 6.1 ± 1.1 6.3 ± 1.1 < 0.0001

LDL cholesterol, mmol/l 4.0 ± 1.0 4.1 ± 1.0 < 0.0001

HDL cholesterol, mmol/l 1.25 ± 0.33 1.16 ± 0.31 < 0.0001

Triglycerides, mmol/l 1.7 (1.2 - 2.4) 2.0 (1.4 - 2.9) < 0.0001

C-reactive protein, mg/l 1.4 (0.7 - 2.9) 2.2 (1.0 - 4.5) < 0.0001

sPLA2, ng/ml 7.3 (5.3 - 10.6) 8.3 (5.8 - 12.5) < 0.0001

women

(n = 813) (n = 398)

Age, year 67 ± 7 67 ± 7 Matched

Smoking - Never 55.5 (447) 44.9 (176)

- Previous 36.7 (296) 39.8 (156) < 0.0001

- Current 7.8 (63) 15.3 (60)

Body mass index, kg/m2 26.3 ± 4.1 27.4 ± 4.5 < 0.0001

Diabetes 0.8 (7) 5.5 (22) < 0.0001

Systolic blood pressure, mmHg 139 ± 18 143 ± 19 < 0.0001

Diastolic blood pressure, mmHg 82 ± 11 85 ± 12 < 0.0001

Total cholesterol, mmol/l 6.6 ± 1.2 6.9 ± 1.3 < 0.0001

LDL cholesterol, mmol/l 4.3 ± 1.1 4.5 ± 1.1 0.001

HDL cholesterol, mmol/l 1.59 ± 0.4 1.45 ± 0.39 < 0.0001

Triglycerides, mmol/l 1.5 (1.1 - 2.2) 1.8 (1.3 - 2.6) < 0.0001

C-reactive protein, mg/l 1.6 (0.8 - 3.5) 2.6 (1.1 - 5.8) < 0.0001

sPLA2, ng/ml 10.4 (7.3 - 16.3) 12.3 (7.9 - 18.3) < 0.0001

Data are presented as mean ± SD, n (%), or median (interquartile range). LDL indicates low-density lipoprotein; HDL indicates high-density lipoprotein; sPLA2 indicates type II secretory phospholipase A2. Means, percentages and medians may be based on fewer observations than the indicated number of subjects.


64

efficients and corresponding p-values were calculated to assess the relationship between sPLA2 as a continuous variable and other continuous risk factors. Odds ratios (OR) and corresponding 95% confidence intervals (95%CI) as an estimate of the relative risk of incident CAD were calculated using conditional logistic regression analysis, which takes into account the matching for sex and age. The lowest sPLA2 quartile was used as the reference category. Odds ratios were adjusted for the following cardiovascular risk factors: body mass index, diabetes, systolic blood pressure, LDL-c, HDL-c, smoking (never, previ-ous, current). Odds ratios were also calculated after additional adjustment for CRP levels. The statistical interaction between sex and sPLA2 was calculated to assess the validity of pooling sexes. Statistical analyses were performed using SPSS software (version 12.0.1, Chicago, Illinois). A p-value < 0.05 was considered to indicate statistical significance.

RESuLTSBaseline characteristics of cases and controls are presented in table 1. Matching ensured that age and sex were comparable between cases and controls. As expected, individuals who developed CAD during follow-up were more likely than controls to smoke and have diabetes. Levels of total cholesterol, LDL-c, triglycerides, systolic and diastolic blood pres-sure, BMI, and CRP were significantly higher in cases than controls, and HDL-c levels were significantly lower. sPLA2 levels were higher in cases than controls both among men (8.3 ng/ml, IQR : 5.8-12.5 versus 7.3 ng/ml IQR 5.3-10.6, P < 0.0001), and women (12.3 ng/ml, IQR 7.9-18.3 versus 10.4 ng/ml, IQR 7.3-16.3, P < 0.0001)). Notably, sPLA2 levels were substantially higher in women than men (11.0 ng/ml, IQR 7.5-16.9 versus 7.7 ng/ml, IQR 5.5-11.3, P < 0.0001). Among both men and women, sPLA2 plasma levels significantly correlated with age, BMI, systolic blood pressure, HDL-c, and CRP levels. We conducted pooled analyses for men and women as no significant interaction was observed between sPLA2 levels and sex for CAD risk (P = 0.7). Using sPLA2 quartiles based on the distribution among controls, the risk of future CAD increased continuously (P for linearity = 0.001). Taking into account matching for sex and age and adjusting for BMI, smoking, diabetes, systolic blood pressure, LDL-c, and HDL-c, and CRP levels, the risk of future CAD was 1.34 (1.02-1.71, P = 0.02) for people in the highest sPLA2 quartile, compared to those in the lowest. The sex-specific analyses showed similar patterns. Both CRP and sPLA2 levels were independently related to the risk of CAD (figure 1).In order to assess the independency of the associations of CRP and sPLA2 with CAD risk, we calculated the increase in CAD risk that was associated with 1 SD increase of each. Taking into account the matching for sex and age and adjusting for BMI, systolic blood pressure, LDL-c, HDL-c, diabetes and smoking, 1 SD increase in CRP level was associated with an odds ratio of 1.18 (95%CI : 1.08-1.29); the equivalent value for sPLA2 was 1.19 (95%CI : 1.09-1.30). When both were entered in the conditional logistic regression model, both retained an independent and statistically significant association with CAD risk: OR = 1.12 (95%CI : 1.02-1.23) for CRP, and OR = 1.15 (95%CI : 1.05-1.26) for sPLA2. Thus, the relationship between serum levels of sPLA2 and CAD risk was not altered substantially upon additional adjustment for CRP levels.

Chapter 4

TABLE 2: DISTRIBuTION OF CARDIOvASCuLAR RISK FACTORS ACCORDINg TO SPLA2 QuARTILE IN EPIC-NORFOLK 1993-2003Quartile 1 2 3 4 P* R P†

Men

sPLA2 range, ng/ml < 5.3 5.3 - 7.29 7.3 - 10.59 > 10.6

Case / control 131 / 337 151 / 363 188 / 345 237 / 351

Age, year 63 ± 9 65 ± 8 65 ± 8 66 ± 8 < 0.0001 0.131 < 0.0001

Smoking - Never 35.8 (167) 31.0 (158) 26.9 (140) 25.1 (147)

- Previous 57.1 (266) 59.1 (301) 62.6 (326) 60.0 (351) < 0.0001

- Current 7.1 (33) 9.8 (50) 10.6 (55) 14.9 (87)

Body mass index, kg/m2 26.3 ± 3.1 26.4 ± 3.0 27.0 ± 3.4 26.9 ± 3.4 0.001 0.064 0.04

Diabetes 2.8 (13) 4.5 (23) 3.9 (21) 4.4 (26) 0.5

Systolic blood pressure, mmHg 141 ± 18 140 ± 18 141 ± 17 142 ± 19 0.2 0.050 0.02



LDL cholesterol, mmol/l 4.0 ± 0.9 4.0 ± 1.0 4.0 ± 1.0 4.0 ± 1.0 0.5 0.014 0.5

HDL cholesterol, mmol/l 1.26 ± 0.34 1.22 ± 0.31 1.22 ± 0.34 1.18 ± 0.33 0.001 -0.078 < 0.0001


C-reactive protein, mg/l 2.1 ± 3.2 2.5 ± 3.7 3.5 ± 6.9 5.4 ± 7.2 < 0.0001 0.292 < 0.0001

women

sPLA2 range, ng/ml < 7.3 7.3 - 10.39 10.4 - 16.29 > 16.3

Case / control 73 / 201 82 / 207 112 / 202 131 / 203

Age, year 66 ± 8 66 ± 7 67 ± 7 67 ± 7 0.04 0.083 0.004

Smoking - Never 57.1 (156) 55.6 (158) 47.7 (147) 48.6 (162)

- Former 34.4 (94) 36.3 (103) 39.3 (121) 40.2 (134) 0.1

- Current 8.4 (23) 8.1 (23) 13.0 (40) 11.1 (37)

Body mass index, kg/m2 25.8 ± 3.9 26.4 ± 4.1 26.5 ± 3.9 27.6 ± 4.8 < 0.0001 0.156 < 0.0001

Diabetes 1.5 (4) 1.4 (4) 2.2 (7) 4.2 (14) 0.07

Systolic blood pressure, mmHg 138 ± 18 140 ± 19 141 ± 19 141 ± 19 0.1 0.072 0.01



LDL cholesterol, mmol/l 4.2 ± 1.1 4.4 ± 1.1 4.4 ± 1.1 4.3 ± 1.1 0.07 -0.015 0.6

HDL cholesterol, mmol/l 1.58 ± 0.45 1.57 ± 0.44 1.54 ± 0.40 1.48 ± 0.37 0.02 -0.071 0.02

Triglycerides, mmol/l 1.8 ± 0.9 1.8 ± 1.0 2.0 ± 1.8 1.9 ± 1.1 0.1 0.007 0.8

C-reactive protein, mg/l 2.3 ± 3.2 3.1 ± 4.9 3.7 ± 4.6 6.8 ± 10.5 < 0.0001 0.321 < 0.0001

HRT - Never 77.0 (211) 76.7 (221) 77.1 (242) 81.1 (271)

- Previous 10.6 (29) 11.5 (33) 8.0 (25) 8.7 (29) 0.4

- Current 12.4 (34) 11.8 (34) 15.0 (47) 10.2 (34)

Data are presented as mean ± SD per sPLA2 quartile. LDL indicates low-density lipoprotein; HDL indicates high-density lipoprotein; sPLA2 indicates type II secretory phospholipase A2; HRT indicates hormone replacement therapy. P* indicates p-value for linearity between sPLA2 quartiles and risk factor levels; R indicates Pearson’s correlation between log-transformed sPLA2 levels and risk factor levels, and P† indicates the corresponding p-value. Because of their skewed distribution, triglycerides, CRP, and sPLA2 levels were log-transformed before analysis as continuous variables.


65

TABLE 3: RISK OF FuTuRE CAD PER SPLA2 QuARTILE FOR APPARENTLy HEALTHy MEN, wOMEN AND SEXES COMBINED

sPLA2 quartile 1 2 3 4 P linearity

Men + women

sPLA2 range, ng/ml < 5.8 5.8 - 8.29 8.3 - 12.69 > 12.7

Cases / controls 212 / 531 246 / 569 286 / 558 366 / 551

Model 1 1.00 1.02 1.2 1.48 0.001

(0.79-1.31) (0.93-1.55) (1.15-1.90)

Model 2 1.00 0.99 1.15 1.34 0.03

(0.77-1.28) (0.89-1.48) (1.02-1.71)

Men

sPLA2 range, ng/ml < 5.3 5.3 - 7.29 7.3 - 10.59 > 10.6

Cases / controls 131 / 337 151 / 363 188 / 345 237 / 351

Model 1 1.00 0.95 1.35 1.4 0.005

(0.70-1.30) (0.99-1.86) (1.03-1.89)

Model 2 1.00 0.94 1.31 1.27 0.04

(0.69-1.29) (0.96-1.80) (0.93-1.73)

women

sPLA2 range, ng/ml < 7.3 7.3 - 10.39 10.4 - 16.29 > 16.3

Cases / controls 73 / 201 82 / 207 112 / 202 131 / 203

Model 1 1.00 1.08 1.34 1.59 0.01

(0.70-1.65) (0.89-2.01) (1.06-2.39)

Model 2 1.00 1.05 1.44 1.44 0.06

(0.68-1.61) (0.82-1.89) (0.95-2.18)

Odds ratios for future CAD calculated by conditional logistic regression with corresponding 95% confidence intervals per quartile. Model 1 adjusted for BMI, diabetes, systolic blood pressure, LDL-c, HDL-c, smoking. Model 2 adjusted for the same variables and also CRP levels. For the pooled analysis, quartiles are based on the distribution among the controls for sexes combined. For the sex-specific analyses, quartiles were based on the sex-specific distribution among the controls. Because of their skewed distribution, CRP levels were log-transformed before analysis as continuous variables.

Chapter 4

66

DISCuSSIONIn this large prospective study among apparently healthy men and women, we observed that serum levels of sPLA2 are associated with an increased risk of future coronary artery disease. After adjustment for classical cardiovascular risk factors and CRP levels, people in the highest sPLA2 quartile had a 34% increased risk compared to those in the lowest quartile. This observation suggests that sPLA2 levels may reflect a pathophysiological process relevant in the development of atherosclerosis, that is different from that mir-rored by CRP levels.

sPLA2 and CRP as CAD risk markersOur observations are consistent with a small prospective study which showed that among patients with symptomatic CAD, those with high sPLA2 levels were at an increased risk of developing recurrent CAD events during follow-up.15 They are also consistent with two large prospective studies that investigated the association between a family member of sPLA2, lipoprotein-associated phospholipase A2 (Lp-PLA2), and CAD risk. Plasma levels of Lp-PLA2 were shown to predict future CAD in men with elevated LDL-c levels22 and in apparently healthy individuals.23

Both sPLA2 and CRP are acute phase proteins and as such might be regarded as surro-gate markers of vascular inflammation reflecting endothelial dysfunction. CRP has been established as plasma marker of CAD risk and only recently, other evidence is starting to accumulate that it may play an active role in atherogenesis as well.24 25 26 Conversely, substantial evidence already exists to support the causality of sPLA2 in a number of patho-physiological mechanisms potentially relevant in atherosclerosis but sound epidemiologi-cal evidence for a relationship with CAD risk is lacking. We observed that sPLA2 levels and CRP levels were strongly correlated both in men and in

FIguRE 1: RISK OF CAD IN APPARENTLy HEALTHy INDIvIDuALS STRATIFIED By TERTILES OF CRP

AND SPLA2 IN EPIC-NORFOLK 1993-2003

Bars represent fully adjusted odds ratios for future CAD in apparently healthy individuals stratified by tertiles of CRP and sPLA2. Odds ratios were calculated by conditional logistic regression, taking into account matching for age, sex, and enrolment time, and adjusting for smoking, diabetes, BMI, LDL-c, HDL-c and systolic blood pres-sure. People in the lowest tertile for both CRP and sPLA2 were used as the reference group.


67

0

0.5

1

1.5

2

Fully

adj

uste

d O

R fo

r fu

ture

CAD

CRP tertiles

sPLA2 tertiles3

21

1

2

3

women. Adjustment did attenuate the relationship between sPLA2 and CAD risk but both sPLA2 and CRP remained significant independent predictors of CAD. Second, we observed an inverse correlation between sPLA2 levels and HDL-c levels in men and women. In recent years, evidence is accumulating that HDL-c and inflammation are inversely related.27 For instance, hydrolysis of acute phase HDL particles by sPLA2 was 2- to 3-fold more rapid and intensive than of normal HDL.28 In addition, sPLA2 is known to cause increased catabolism of HDL particles under non-inflammatory circumstances as well,29 which may explain our observation. Finally, we observed that on average, women had substantially higher sPLA2 levels than men. This observation is consistent with the observation that women have higher CRP levels than men.30 The high levels of sPLA2 observed among women could not be ex-plained by the use of hormone replacement therapy, which may increase CRP levels. Several lines of biochemical evidence support the involvement of sPLA2 in atherogenesis. Transgenic mice expressing human sPLA2 develop more atherosclerosis.31 It is well-estab-lished that sPLA2 detrimentally affects lipid metabolism29,31-33 but this could not explain its effect on atherogenesis since it also occurred in transgenic mice kept on a low-fat diet.31

Subsequently, it was demonstrated that mice which did not express sPLA2 systemically but received bone marrow-derived cells expressing human sPLA2, also developed sig-nificantly more atherosclerosis.34 Thus, macrophage-expressed sPLA2 may also play an important role in atherogenesis. In humans, sPLA2 is highly expressed in atherosclerotic tissue and co-localizes with mono-cyte-derived macrophages.35-38 sPLA2 is expressed in response to a variety of inflamma-tory cytokines, including interleukin-1β, IL-6 and tumor necrosis factor-a.5,39,40 Vice versa, sPLA2 itself directly induces the expression of chemokines and adhesion molecules in microvascular endothelium.41 Thus, sPLA2 may play a role in the signaling pathways dur-ing inflammation but it also has direct atherogenic effects, possibly via the modification of the structure of lipoproteins. First, treatment of LDL lipoproteins with sPLA2 causes a substantial reduction of phosphatidylcholine in the surface monolayer of LDL particles resulting in smaller and denser LDL particles and altering the configuration of the apo-lipoprotein B molecule on the lipoprotein.7 This altered configuration may expose more arginine- and lysine-rich segments which can form strong interactions with glycosamino-glycans in the extracellular matrix6 explaining the higher affinity for extracellular matrix components of sPLA2-modified lipoproteins compared to control LDL.7,8 This increased affinity for extracellular matrix components results in increased retention of LDL particles in the arterial wall, an early marker of atherosclerotic lesion progression,9-11 possibly be-cause matrix-bound lipoproteins are more susceptible to form aggregated lipid droplets and vesicles in the vessel wall.42 Second, the sPLA2-mediated hydrolysis of the sn-2 ester bond in phospholipids liberates a number of biologically active agents, including non-es-terified fatty acids and lysophospholipids, which are precursors of various proinflammato-ry mediators including leukotrienes and prostaglandines.12 Third, sPLA2 may increase the susceptibility of lipoproteins to undergo lipid peroxidation13 yielding oxidized lipoproteins which may in turn enhance macrophage growth.43 In addition, sPLA2 may also generate more bioactive phospholipids which stimulate endothelial cells to bind monocytes.14 Finally, in addition to its potential role in the initiation and progression of atherosclerosis,

Chapter 4

68

sPLA2 may also have detrimental effects in the setting of ischemic events. Depositions of sPLA2 have been demonstrated in the necrotic center of infarcted human myocardium.44 Interestingly, sPLA2 was also found to be localized in myocardium adjacent to the border zone where cardiomyocytes do not show morphological signs of cell death.44 This sug-gests that sPLA2 may play a role in the enlargement of the necrotic myocardium during ischemia by enhancing damage in cells that are presumably only reversibly damaged by the ischemic challenge. Indeed, in vitro evidence suggests that sPLA2 can transform flip-flopped but viable cardiomyocytes into apoptotic/necrotic cells.45 This observation sug-gests that sPLA2 may enhance myocardial cell damage either by a direct cytotoxic effect or by enhancing the inflammatory response.

ConsiderationsCertain aspects of this study merit further consideration. First, levels of sPLA2 were de-termined in a serum sample that was stored at –80˚C so we cannot exclude some degree of protein degradation in these samples. However, this would introduce an increased ran-dom measurement error, which is likely to lead to an underestimation of any relationship, and therefore does not negate our findings. Second, CAD events were ascertained through death certification and hospital admission data, which is likely to lead both to underascertainment and to misclassification of cases. Previous validation studies in our cohort indicate high specificity of such case ascertain-ment.46 Again, case underascertainment and misclassification is likely to attenuate any relationships. Third, the current data do not allow us to study the causality of the relation-ship between sPLA2 and CAD. We cannot exclude the possibility that in this population study, sPLA2 concentration was a marker of subclinical atherosclerosis rather than an effector although this hypothesis would be not be consistent with a number of in vitro observations supporting a causal role of sPLA2 in lipoprotein modification. The independent relationship of sPLA2 with CAD is of clinical interest since inhibitors of sPLA2 activity are being developed. The results of the present analysis raise several issues concerning the potential effects of currently tested inhibitors of sPLA2 activity.5 First of all, we have studied apparently healthy individuals while sPLA2 inhibitors will be used, if ever, in people at high cardiovascular risk. Second, we have studied sPLA2 under physi-ological conditions in which sPLA2 activity and sPLA2 concentration are assumed to be strongly correlated. This contrasts with the reduced sPLA2 activity upon pharmacological inhibition. Taking these considerations into account, care is warranted when extrapolat-ing our finding that high sPLA2 concentrations are associated with a moderately reduced elevated risk of CAD in healthy individuals to the potential effects of sPLA2 inhibition on CAD risk in patients. The ongoing trials assessing the effect of sPLA2 inhibitors on sur-rogate markers of CAD as well as clinical endpoints will have to answer this question.

ConclusionElevated levels of sPLA2 are associated with an increased risk of future CAD in apparently healthy individuals. The observed relationship was continuous and was independent of classical cardiovascular risk factors. The magnitude of the relationship was similar to that


69

observed between CRP levels and CAD risk, and these relationships were independent. These prospective data support the hypothesis that sPLA2 plays a role in the pathogenesis of atherosclerosis or its major clinical manifestation, coronary artery disease. Inhibitors of sPLA2 activity may hold promise for the therapeutic future.

ACKNOwLEDgEMENTSWe thank the participants, general practitioners and staff in EPIC-Norfolk. EPIC-Norfolk is supported by programme grants from the Medical Research Council UK and Cancer Research UK and with additional support from the European Union, Stroke Associa-tion, British Heart Foundation, Department of Health, Food Standards Agency and the Wellcome Trust. JK is an established investigator (2000 D039) and WJ is an established clinical investigator (2001 D032) of the Netherlands Heart Foundation, The Hague, The Netherlands.

CONFLICTS OF INTEREST None

Chapter 4

70

References 1 Libby P. Inflammation in atherosclerosis. Nature 2002;420:868-74. 2 Ridker PM, Cushman M, Stampfer MJ et al. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med 1997;336:973-9. 3 Ridker PM, Hennekens CH, Buring JE et al. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med 2000;342:836-43. 4 Boekholdt SM, Peters RJ, Hack CE et al. IL-8 plasma concentrations and the risk of future coronary artery disease in apparently healthy men and women. The EPIC-Norfolk prospective population study. Arterioscler Thromb Vasc Biol 2004;24:1-7. 5 Hurt-Camejo E, Camejo G, Peilot H et al. Phospholipase A2 in vascular disease. Circ Res 2001;89:298- 304. 6 Camejo G, Hurt-Camejo E, Wiklund O et al. Association of apo B lipoproteins with arterial proteoglycans: pathological significance and molecular basis. Atherosclerosis 1998;139:205-22. 7 Sartipy P, Camejo G, Svensson L et al. Phospholipase A(2) modification of low density lipoproteins forms small high density particles with increased affinity for proteoglycans and glycosaminoglycans. J Biol Chem 1999;274:25913-20. 8 Hakala JK, Oorni K, Pentikainen MO et al. Lipolysis of LDL by human secretory phospholipase A(2) induces particle fusion and enhances the retention of LDL to human aortic proteoglycans. Arterioscler Thromb Vasc Biol 2001;21:1053-8. 9 Schwenke DC, Carew TE. Initiation of atherosclerotic lesions in cholesterol-fed rabbits. I. Focal increases in arterial LDL concentration precede development of fatty streak lesions. Arterioscler 1989;9:895-907. 10 Schwenke DC, Carew TE. Initiation of atherosclerotic lesions in cholesterol-fed rabbits. II. Selective retention of LDL vs. selective increases in LDL permeability in susceptible sites of arteries. Arterioscler 1989;9:908-18. 11 Collins Tozer E, Carew TE. Residence Time of Low-Density Lipoprotein in the Normal and Atherosclerotic Rabbit Aorta. Circ Res 1997;80:208-18. 12 Oestvang J, Bonnefont-Rousselot D, Ninio E et al. Modification of LDL with human secretory phospholipase A2 or sphingomyelinase promotes its arachidonic acid-releasing propensity. J Lipid Res 2004;45:831-8. 13 Neuzil J, Upston JM, Witting PK et al. Secretory phospholipase A2 and lipoprotein lipase enhance 15-lipoxy genase-induced enzymic and nonenzymic lipid peroxidation in low-density lipoproteins. Biochemistry 1998;37:9203-10. 14 Leitinger N, Watson AD, Hama SY et al. Role of group II secretory phospholipase A2 in atherosclerosis: 2. Potential involvement of biologically active oxidized phospholipids. Arterioscler Thromb Vasc Biol 1999;19:1291-8. 15 Kugiyama K, Ota Y, Takazoe K et al. Circulating levels of secretory type II phospholipase A2 predict coronary events in patients with coronary artery disease. Circulation 1999;100:1280-4. 16 Liu PY, Li YH, Tsai WC et al. Prognostic value and the changes of plasma levels of secretory type II phos pholipase A2 in patients with coronary artery disease undergoing percutaneous coronary intervention. Eur Heart J 2003;24:1824-32. 17 Day NE, Oakes S, Luben R et al. EPIC-Norfolk: study design and characteristics of the cohort. European Prospective Investigation of Cancer. Br J Cancer 1999;80 Suppl 1:95-103. 18 Boekholdt SM, Kuivenhoven JA, Wareham NJ et al. Plasma levels of cholesteryl ester transfer protein and the risk of future coronary artery disease in apparently healthy men and women; the prospective EPIC- Norfolk population study. Circulation 2004;110:1418-23 19 Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499-502. 20 Wolbink GJ, Schalkwijk C, Baars JW et al. Therapy with interleukin-2 induces the systemic release of phospholipase-A2. Cancer Immunol Immunother 1995;41:287-92. 21 Bruins P, te Velthuis H, Yazdanbakhsh AP et al. Activation of the complement system during and after cardiopulmonary bypass surgery: postsurgery activation involves C-reactive protein and is associated with post operative arrhythmia. Circulation 1997;96:3542-8. 22 Packard CJ, O’Reilly DS, Caslake MJ et al. Lipoprotein-associated phospholipase A2 as an independent predictor of coronary heart disease. N Engl J Med 2000;343:1148-55. 23 Ballantyne CM, Hoogeveen RC, Bang H et al. Lipoprotein-associated phospholipase A2, high-sensitivity C-reactive protein, and risk of incident coronary heart disease in middle-aged men and women in the Atherosclerosis Risk in Communities (ARIC) study. Circulation 2004;109:837-42. 24 Danenberg HD, Szalai AJ, Swaminathan RV et al. Increased thrombosis after arterial injury in human C-reactive protein-transgenic mice. Circulation 2003;108:512-5. 25 Zwaka TP, Hombach V, Torzewski J. C-reactive protein-mediated low density lipoprotein uptake by macrophages: implications for atherosclerosis. Circulation 2001;103:1194-7. 26 Bisoendial RJ, Kastelein JJ, Levels JH et al. Activation of inflammation and coagulation after infusion of C- reactive protein in humans. Circ Res 2005;96:714-716. 27 Wadham C, Albanese N, Roberts J et al. High-density lipoproteins neutralize C-reactive protein proinflam matory activity. Circulation 2004;109:2116-22.


71

28 Pruzanski W, Stefanski E, de Beer FC et al. Lipoproteins are substrates for human secretory group IIA phospholipase A2: preferential hydrolysis of acute phase HDL. J Lipid Res 1998;39:2150-60. 29 Tietge UJ, Maugeais C, Cain W et al. Overexpression of secretory phospholipase A(2) causes rapid catabolism and altered tissue uptake of high density lipoprotein cholesteryl ester and apolipoprotein A-I. J Biol Chem 2000;275:10077-84. 30 Rogowski O, Zeltser D, Shapira I et al. Gender difference in C-reactive protein concentrations in individuals with atherothrombotic risk factors and apparently healthy ones. Biomarkers 2004;9:85-92. 31 Ivandic B, Castellani LW, Wang XP et al. Role of group II secretory phospholipase A2 in atherosclerosis: 1. Increased atherogenesis and altered lipoproteins in transgenic mice expressing group IIa phospholipase A2. Arterioscler Thromb Vasc Biol 1999;19:1284-90. 32 de Beer FC, Connell PM, Yu J et al. HDL modification by secretory phospholipase A(2) promotes scavenger receptor class B type I interaction and accelerates HDL catabolism. J Lipid Res 2000;41:1849-57. 33 de Beer FC, de Beer MC, van der Westhuyzen DR et al. Secretory non-pancreatic phospholipase A2: influence on lipoprotein metabolism. J Lipid Res 1997;38:2232-9. 34 Webb NR, Bostrom MA, Szilvassy SJ et al. Macrophage-expressed group IIA secretory phospholipase A2 increases atherosclerotic lesion formation in LDL receptor-deficient mice. Arterioscler Thromb Vasc Biol 2003;23:263-8. 35 Hurt-Camejo E, Andersen S, Standal R et al. Localization of nonpancreatic secretory phospholipase A2 in normal and atherosclerotic arteries. Activity of the isolated enzyme on low-density lipoproteins. Arterioscler Thromb Vasc Biol 1997;17:300-9. 36 Elinder LS, Dumitrescu A, Larsson P et al. Expression of phospholipase A2 isoforms in human normal and atherosclerotic arterial wall. Arterioscler Thromb Vasc Biol 1997;17:2257-63. 37 Menschikowski M, Kasper M, Lattke P et al. Secretory group II phospholipase A2 in human atherosclerotic plaques. Atherosclerosis 1995;118:173-81. 38 Anthonsen MW, Stengel D, Hourton D et al. Mildly oxidized LDL induces expression of group IIa secretory phospholipase A(2) in human monocyte-derived macrophages. Arterioscler Thromb Vasc Biol 2000;20: 1276-82. 39 Akiba S, Hatazawa R, Ono K et al. Secretory phospholipase A2 mediates cooperative prostaglandin generation by growth factor and cytokine independently of preceding cytosolic phospholipase A2 expression in rat gastric epithelial cells. J Biol Chem 2001;276:21854-62. 40 van der Helm HA, Aarsman AJ, Janssen MJ et al. Regulation of the expression of group IIA and group V secretory phospholipases A2 in rat mesangial cells. Biochim Biophys Acta 2000;1484:215-24. 41 Beck GC, Yard BA, Schulte J et al. Secreted phospholipases A2 induce the expression of chemokines in microvascular endothelium. Biochem Biophys Res Comm 2004;300:731-7. 42 Oorni K, Pentikainen MO, Ala-Korpela M et al. Aggregation, fusion, and vesicle formation of modified low density lipoprotein particles: molecular mechanisms and effects on matrix interactions. J Lipid Res 2000;41:1703-14. 43 Kaneko K, Sakai M, Matsumura T et al. Group-II phospholipase A(2) enhances oxidized low density lipoprotein -induced macrophage growth through enhancement of GM-CSF release. Atherosclerosis 2000;153:37-46. 44 Nijmeijer R, Lagrand WK, Baidoshvili A et al. Secretory type II phospholipase A(2) binds to ischemic myocardium during myocardial infarction in humans. Cardiovasc Res 2002;53:138-46. 45 Nijmeijer R, Willemsen M, Meijer CJ et al. Type II secretory phospholipase A2 binds to ischemic flip-flopped cardiomyocytes and subsequently induces cell death. Am J Physiol Heart Circ Physiol 2003;285:H2218- H2224. 46 Boekholdt SM, Peters RJ, Day NE et al. Macrophage migration inhibitory factor and the risk of myocardial infarction or death due to coronary artery disease in adults without prior myocardial infarction or stroke: The EPIC-Norfolk prospective population study. Am J Med 2004;117:390-397

Chapter 4

72

73


Chapter

5tIssue faCtor serum levels and the rIsk of

future Coronary artery dIsease In apparently healthy men and women;

the epIC-norfolk prospeCtIve populatIon study

Tymen Keller • Daniel Choi • Henk te velthuis • victor gerdes • Nicholas warehamSheila Bingham • Robert Luben • Erik Hack • Pieter Reitsma • Marcel Levi

Kay-Tee Khaw • Matthijs Boekholdt

Departments of Vascular Medicine, Experimental Internal Medicine and Cardiology, Academic Medical Center, Amsterdam, The NetherlandsDepartment of Immunopathology, Sanquin Research, Amsterdam, The Netherlands

Medical Research Council Epidemiology, Dunn Nutrition Unit and the Department of Public Health and Primary Care, University of Cambridge, Cambridge, United Kingdom

Department of Clinical Chemistry, VU Medical Center, Amsterdam, The Netherlands

J Thromb Haemost 2006 (in press)

ABSTRACTTissue factor (TF) has been implicated in coronary artery disease (CAD). High levels of cir-culating TF are found in patients with acute atherothrombotic events. Whether high serum TF levels predict risk of future CAD independent of known risk factors remains unknown. We conducted a prospective case-control study nested in the EPIC-Norfolk population study. Cases (n=1037) were apparently healthy men and women, aged 45-79 years, who developed fatal or non-fatal CAD during follow-up. Controls (n=2005) were matched by age, sex and enrolment time. Serum TF levels were measured using high-affinity antibodies. In men, me-dian TF levels were not significant higher in cases than in controls (59.0 pg/ml range: 16.7-370.4 versus 54.9 pg/ml range: 16.2-452.4). In women, median TF levels were not significant higher in controls than in cases (73.4 pg/ml range: 16.7-492.3 versus 50.5 pg/ml range: 16.5-376.7). The prevalence of smoking was about double in the lowest compared to the highest TF quartile. Correcting for sex, age, BMI, smoking, diabetes, systolic blood pressure, LDL-c, and HDL-c, and CRP levels, the risk of future CAD was 1.05 (95% CI: 0.81-1.36) for people in the highest TF quartile, compared to those in the lowest (P value for linearity = 0.8). High levels of serum TF were not independently associated with an increased risk of future CAD in apparently healthy individuals.

Chapter 5

76

INTRODuCTIONCoronary artery disease (CAD) results from a chronic inflammatory disease of the vascular wall, of which the pathogenesis is still poorly understood. The evidence suggests that a com-bination of acquired risk factors and genetic predisposition initiate and maintain the inflam-matory process.1 Several inflammatory markers have been associated with an increased risk of CAD. And recently, tissue factor (TF), which activates coagulation2 and affects inflammation,3 has been implicated in the pathogenesis of CAD.4, 5

Several observations strengthen the hypothesis that TF plays an important role in the patho-genesis of CAD. First, TF is extensively expressed in atherosclerotic plaques,6 depending on ox-LDL-induced activation of macrophages.7 Second, unstable atherosclerotic plaques contain more TF than stable plaques,8 which was shown to be associated with increased plaque throm-bogenicity.9 Third, acute myocardial infarction (AMI) and unstable angina (UA) are associated with increased circulating levels of TF-containing microparticles which are capable of activat-ing coagulation and inflammation.3, 10 And finally, in people with AMI and UA, serum TF levels are increased compared to subjects with stable coronary heart disease or healthy controls.11, 12 These increased levels have been shown to be predictors of an unfavourable outcome in pa-tients with UA.13 Taken together these data indicate that among patients with CAD, increased serum TF levels may be associated with atherosclerotic plaque instability and increased risk of acute complications of CAD.A role for TF in the acute complications of CAD does not necessarily mean that a high level of TF expression or serum TF is also a risk factor for the initiation and/or progression of CAD. Although it has also been suggested that other risk factors for CAD interfere with TF expres-sion and thereby contribute to the role of TF in the development of the disease,14 it is unknown whether high levels of serum TF predict future CAD in apparently healthy individuals. In order to test this hypothesis, we performed a large prospective case-control study nested in the Euro-pean Prospective Investigation into Cancer in Norfolk (EPIC-Norfolk) prospective population study. We measured serum levels of TF in apparently healthy men and women and assessed the risk of future CAD during a follow-up period of 6 years.

METHODSWe performed a nested case-control study of coronary artery disease in the EPIC-Norfolk cohort. This prospective population study of 25,663 men and women aged between 45 and 79 years, as-sessed the determinants of cancer and other diseases. The design and methods of the study and of the nested CAD case control study have been described in detail elsewhere.15 16 At the baseline sur-vey between 1993 and 1997, qualified staff collected blood from all participants. The Department of Clinical Biochemistry, University of Cambridge, stored the samples at –80˚C or processed the sam-ples for assay. We identified participants as having CAD during follow-up through November 2003 if they had a hospital admission and/or died with CAD as underlying cause. Data were obtained by linkage with mortality registration of the UK Office of National Statistics and with ENCORE, the East Norfolk Health Authority database, which identifies all hospital contacts throughout England and Wales for Norfolk residents. We defined CAD as codes 410-414 according to the International Classification of Diseases, 9th revision. The study was approved by the Norwich District Health Authority Ethics Committee and all participants gave signed informed consent.

Tissue factor serum levels and the risk of future coronary artery disease in apparently healthy men and women; the EPIC-Norfolk prospective population study

77

Participants

Cases were individuals identified from the EPIC-Norfolk cohort who developed fatal or non-fatal CAD during follow-up through November 2003. Controls were selected from participants who re-mained free of any cardiovascular disease during follow-up. We attempted to match two controls to each case by sex, age (within 5 years) and time of enrolment (within 3 months) in the study. We excluded all individuals who reported a history of heart attack or stroke at the baseline clinic visit.


Serum levels of total cholesterol, HDL cholesterol, and triglycerides were measured on fresh samples with the RA 1000 (Bayer Diagnostics, Basingstoke, UK), and LDL cholesterol levels were calculated with the Friedewald formula.17 Serum concentrations of CRP were measured with a sandwich-type ELISA as previously described.18 Results were related to a standard con-sisting of commercially available CRP (Behringwerke AG, Marburg, Germany). Serum TF levels were measured with monoclonal anti-human TF antibodies from Sanquin Reagents (Amsterdam, The Netherlands), as recently described by van der Putten et al.19 They showed that TF levels measured with this ELISA are comparable between serum and plasma sam-ples.20 In brief, microtiterplates were coated with CLB/TF-5. Serum was diluted 5 fold in HPE buffer with 1% Triton X-100 and incubated with biotinylated CLB/TF-1 antibody. Thereafter, samples were incubated with poly-HRP conjugated Streptavidin (Sanquin Reagents). Colour development was done with tetramethylbenzidine (TMB) and the optical density was read on a microplate reader. Although van der Putten et al. used pmol/l to describe TF levels, in the present paper we present TF levels as pg/ml to enable a better comparison with previous reports. The lower detection limit of this assay was 3.4 pg/ml. The mean intra-assay variation between duplicates was 12%. Samples were analyzed in random order and researchers and laboratory personnel were blinded as to case status of the samples.


Because serum TF, triglycerides and CRP had a skewed distribution, values were log-transformed before being used as continuous variables in statistical analyses but in the tables we show un-transformed medians and corresponding interquartile ranges (IQR). In order to determine re-lationships between serum TF and traditional cardiovascular risk factors, we calculated mean risk factor levels per TF quartile. Quartiles were based on the distribution in the controls. For associations between TF quartiles and risk factor levels we used linear regression analysis for con-tinuous variables and chi-square tests for categorical variables. In addition, Pearson’s correlation coefficients and corresponding p-values were calculated to assess the relationship between TF as a continuous variable and other continuous risk factors. Odds ratios (OR) and corresponding 95% confidence intervals (95% CI) as an estimate of the relative risk of incident CAD were calcu-lated using conditional logistic regression, taking into account the matching for sex and age and enrolment time. The lowest quartile was used as the reference category. Odds ratios were adjusted for: (1) confounding factors based on table 2, (2) classical cardiovascular risk factors: age, body mass index (BMI), smoking (current), diabetes, systolic blood pressure, LDL-c, HDL-c, CRP and sex when appropriate. Statistical analyses were performed using SPSS software (version 12.0.2, Chicago, Illinois). A p-value < 0.05 was considered to indicate statistical significance.

Chapter 5

78

RESuLTSTF levels in the study population

Table 1 shows the sex specific distribution of risk factors at baseline between cases and controls. As expected, levels of risk factors including blood pressure, total cholesterol, LDL-c, triglycerides, CRP, diabetes, cigarette smoking habit and BMI were higher and HDL-c lower in those who developed CAD during follow-up compared to those who re-mained free from CAD. In cases and controls no significant different TF levels were seen. Among men, median TF level was 59.0 pg/ml (IQR: 16.7-370.4) in cases and 54.9 pg/ml (IQR: 16.2-452.4) in controls. Among women, median TF level as 50.5 pg/ml (IQR: 16.5-376.7) in cases and 73.4 pg/ml (IQR: 16.7-492.3) in controls.

TF and other risk factors

TF was most strongly related to cigarette smoking habit. In the total study population non smokers had higher levels of TF than former smokers and current smokers. In men who never

TABLE 1: BASELINE CHARACTERISTICS Controls Cases

Men, n 1276 659

Age, year 64 ± 8 65 ± 8

Smoking - Current 8.0 (102) 15.3 (101)

- Previous 59.7 (762) 60.1 (396)

- Never 31.0 (396) 23.8 (157)


Diabetes 2.1 (27) 7.3 (48)








TF, pg/ml 54.9 (16.2 - 452.4) 59.0 (16.7 - 370.4)

women, n 729 378

Age, year 67 ± 7 67 ± 7

Smoking - Current 7.8 (57) 15.3 (58)

- Previous 36.5 (266) 39.4 (149)

- Never 54.7 (399) 43.9 (166)


Diabetes 1.0 (6) 4.8 (15)








TF, pg/ml 73.4 (16.7 - 492.3) 50.5 (16.5 - 376.7)

Data are presented as mean ± SD, n (%) or median (interquartile range). LDL indicates low-density lipoprotein; HDL indicates high-density lipopro-tein; TF indicates tissue factor. Means, percentages and medians may be based on fewer observations than the indicated number of subjects.


79

smoked, median TF levels were 86.4 pg/ml (IQR: 20.5-855.9). In former smokers median TF levels were 51.1 pg/ml (IQR: 15.3-336.5), and in current smokers median TF levels were 30.3 pg/ml (IQR: 12.6-207) (P < 0.0001). This relation was also apparent in women. Median TF levels were 92.7 pg/ml (IQR: 21.2-583.0) in women who never smoked, 55.4 pg/ml (IQR: 14.4-412.7) in former smokers and 23.9 pg/ml (IQR: 13.1-89.1) in current smokers (P < 0.0001). Table 2 shows the relationship between cardiovascular disease risk factors and TF. The prevalence of current smokers was about double in people in the lowest compared to the highest TF quartile category for both men and women. In the lowest TF quartile 14.9% of the men smoked, compared to 8.1% in the highest TF quartile (P < 0.001). In addition, significantly more subjects in the high-est TF quartile had never smoked, compared to the lowest quartile (p<0.001). In men, besides total cholesterol and LDL-c levels, which were significantly decreased in high TF quartiles, no significant relationship was seen between TF levels and cardiovascular risk factors, including blood pressure, body mass index and history of diabetes. The total cholesterol and LDL-c levels showed no consistent trend over the TF quartiles with lower values in the middle two quartiles. In women no relationship was seen between TF levels and cardiovascular risk factors.

TABLE 2: DISTRIBuTION OF CARDIOvASCuLAR RISK FACTORS By TF QuARTILESQuartile 1 2 3 4 P* R P†

Men, case / control 164 / 320 161 / 319 186 / 318 148 / 319

TF range < 16.2 16.65 - 54.9 55.4 - 450.0 > 453.2

Age 65 ± 8 65 ± 8 64 ± 8 64 ± 8 0.8 -0.04 0.1

Smoking - Never 22.1 (107) 27.9 (134) 27.2 (137) 37.5 (175)

- Previous 62.4 (302) 59.8 (287) 63.9 (322) 29.3 (137) <0.01

- Current 14.9 (72) 10.8 (52) 8.1 (41) 8.1 (38)

Body mass index 26.4 ± 3.2 26.8 ± 3.1 26.8 ± 3.4 26.5 ± 3.1 0.2 -0.01 0.7

Diabetes 3.5 (17) 4.0 (19) 5.6 (28) 2.4 (11) 0.1

Systolic blood pressure 141 ± 19 141 ± 17 142 ± 18 139 ± 17 0.2 -0.39 0.1

Diastolic blood pressure 85 ± 12 85 ± 11 86 ± 11 85 ± 11 0.5 -0.002 0.9

Total cholesterol 6.3 ± 1.1 6.1 ± 1.1 6.1 ± 1.1 6.2 ± 1.1 0.01 -0.06 0.01

LDL cholestero 4.2 ± 1.0 4.0 ± 1.0 4.0 ± 0.9 4.0 ± 1.0 <0.01 -0.04 0.1

HDL cholesterol 1.22 ± 0.34 1.24 ± 0.34 1.21 ± 0.32 1.22 ± 0.32 0.9 0.01 0.6

Triglycerides 2.1 ± 1.1 2.0 ± 1.2 2.1 ± 1.2 2.0 ± 1.1 0.4 -0.03 0.2

C-reactive protein 3.6 ± 5.3 3.4 ± 5.1 3.6 ± 7.4 3.1 ± 5.0 0.5 -0.07 <0.01

women, case / control 96 / 184 114 / 181 83 / 182 85 / 182

TF range < 16.5 17.1 – 73.4 73.8 – 486.9 > 498.2

Age 67 ± 7 67 ± 7 67 ± 7 67 ± 7 1.0 -0.004 0.9

Smoking - Never 40.7 (114) 49.2 (145) 58.9 (156) 56.2 (150)

- Former 42.9 (120) 35.9 (106) 35.5 (94) 35.6 (95) <0.01

- Current 15.7 (44) 13.6 (40) 4.2 (11) 7.5 (20)

Body mass index 26.9 ± 4.4 26.8 ± 4.5 26.7 ± 4.4 26.2 ± 4.0 0.2 -0.07 0.03

Diabetes 2.1 (6) 2.7 (8) 2.6 (7) 2.2 (6) 1.0

Systolic blood pressure 139 ± 18 141 ± 21 140 ± 18 140 ± 18 0.9 -0.004 0.9

Diastolic blood pressur 83 ± 11 83 ± 12 83 ± 11 83 ± 11 0.9 0.01 0.9

Total cholesterol 6.7 ± 1.3 6.8 ± 1.2 6.7 ± 1.4 6.7 ± 1.2 1.0 -0.01 0.8

LDL cholesterol 4.4 ± 1.1 4.4 ± 1.1 4.3 ± 1.1 4.3 ± 1.0 0.7 -0.04 0.2

HDL cholesterol 1.51 ± 0.44 1.53 ± 0.39 1.53 ± 0.39 1.60 ± 0.42 0.1 0.08 0.01

Triglycerides 1.9 ± 1.1 1.9 ± 1.0 2.1 ± 1.9 1.9 ± 1.1 0.2 0.01 0.9

C-reactive protein 4.1 ± 6.3 4.1 ± 8.2 3.8 ± 5.0 4.3 ± 6.8 0.8 -0.01 0.9

Data are presented as mean ± SD per tissue factor (TF) quartile or percentage and numbers. LDL indicates low-density lipoprotein; HDL indicates high-density lipoprotein; P* indicates p value for linearity between TF quartiles and risk factor levels; R indicates Pearson’s correlation coefficients between log transformed TF levels and P† indicates the corresponding p-value. Because of their skewed distribution, triglycerides, CRP, and TF levels were log-transformed before analysis as continuous variables.

Chapter 5

80

TF levels and the risk of future CAD

Table 3 shows odds ratios for CAD according to quartile of TF in the whole group and sex stratified. After adjustment for age, BMI, diabetes, systolic blood pressure, LDL-c, and HDL-c, smoking, CRP levels and sex, the risk of future CAD was 1.06 (95% CI: 0.82-1.37) in the pooled analysis of men and women. There was no statistically significant linear trend in OR by increasing quartile of TF (P = 0.77). When men and women were analyzed separately, both crude and adjusted odds ratios showed similar results.

DISCuSSIONIn this large prospective study among apparently healthy men and women, we observed that high serum levels of TF are not independently associated with an increased risk of future CAD. Although TF has been recognized to be involved in the pathogenesis of CAD, no such a relation was demonstrated between serum TF levels at baseline and the development of CAD during follow-up. There are several explanations for the discrepancy between our findings and previous reports. First, the notion that TF plays a key role in the acute CAD event does not imply that the development of CAD is influenced by serum TF levels of healthy people, years before they have the disease. Second, many studies have

TABLE 3: RISK OF FuTuRE CAD ACCORDINg TO TF QuARTILE

TF quartile 1 2 3 4 P linearity

Men + women, cases / controls

265 / 514 278 / 489 262 / 502 232 / 500

TF range, pg/ml < 16.7 17.1 – 62.6 63.0 -466.2 > 470.3

OR unadjusted 1.00 1.10 (0.89-1.36) 1.01 (0.82-1.26) 0.90 (0.71-1.13) 0.3

OR adjusted (1) 1.00 1.18 (0.93-1.49) 1.06 (0.83-1.34) 1.00 (0.78-1.28) 0.8

OR adjusted (2) 1.00 1.18 (0.92-1.50) 1.11 (0.86-1.42) 1.05 (0.81-1.36) 0.8

Men, cases / controls 164 / 320 161 / 319 186 / 318 148 / 319

TF range, pg/ml < 16.2 16.7 – 54.9 55.4 - 450.0 > 453.2

OR unadjusted 1.00 0.98 (0.75-1.28) 1.14 (0.87-1.49) 0.91 (0.68-1.21) 0.8

OR adjusted (1) 1.00 1.02 (0.77-1.35) 1.19 (0.90-1.58) 0.97 (0.72-1.30) 0.9

OR adjusted (2) 1.00 1.09 (0.80-1.49) 1.20 (0.87-1.63) 1.03 (0.76-1.42) 0.7

women, cases / controls 96 / 184 114 / 181 83 / 182 85 / 182

TF range, pg/ml < 16.5 17.1 – 73.4 73.8 – 486.9 > 498.2

OR unadjusted 1.00 1.24 (0.87-1.75) 0.89 (0.62-1.27) 0.90 (0.61-1.31) 0.3

OR adjusted (1) 1.00 1.37 (0.92-2.02) 1.08 (0.72-1.62) 1.20 (0.72-1.62) 0.7

OR adjusted (2) 1.00 1.27 (0.85-1.91) 1.03 (0.68-1.60) 1.17 (0.76-1.81) 0.7

Odds ratios (OR) for future coronary artery disease (CAD) calculated by conditional logistic regression with corresponding 95% confidence inter-vals per TF quartile. (1) adjusted for confounding factors in our population. (2) adjusted for classical risk factors including age, BMI, diabetes, systolic blood pressure, LDL-c, HDL-c, smoking, and CRP levels, and sex in the combined analysis. For the pooled analysis, quartiles are based on the distribution among the controls for sexes combined. For the sex-specific analyses, quartiles were based on the sex-specific distribution among the controls.


81

focused on the role of TF that is present in atherosclerotic plaques, whereas we measured functional and intact TF in the circulation.20 The higher serum levels measured in pa-tients with acute coronary syndromes could reflect the disruption of an atherosclerotic plaque, when there is direct contact between circulating blood and the TF present in the atherosclerotic lesion. Circulating TF is a balance between excretion and immobilization of the molecule from activated monocytes, endothelial cells or subendothelial lineage from an inflamed site on a vessel. Although there is still an ongoing debate about the functional role of TF in the circulation,21, 22 it has recently been shown that circulating TF is (potential) active but is inhibited in vivo.19 In addition, our results suggest that high production on inflamed sites does not warrant high circulating levels. The amount of locally produced TF might be diluted in the high plasma volume or rapidly be cleared from the circulation by uptake by platelets or other cells. Third, in our study we measured TF antigen levels, whereas other studies have measured TF activity. In patients with acute coronary syndromes antigen levels are 20-30% higher than activity levels, and correlation coefficients between activity and antigen of approximately 0.5 have been described.19 Finally, TF activity is influenced by natural positive and negative feedback mechanisms, i.e. tissue factor pathway inhibitor (TFPI). It has been shown that TFPI is lower in patients with CAD,23 but it is still a pos-sibility that in apparently healthy people TFPI influences the relationship between serum TF and future CAD. In this study we could not confirm previously established associations between serum TF levels and risk factors for CAD.14, 24, 25 However, this could be a result of an intrinsic dif-ference between the healthy men and women included in our study and the populations included in other studies consisting also of persons with already established vascular dis-ease. It has been suggested before that the association between TF levels and risk factors could have been occurred secondary to TF release from already established atherosclerotic plaques.5 Interestingly, we found a strong inverse relation between smoking and TF levels. In our study, current smokers had lower TF levels than non smokers, and former smok-ers had TF levels in between the current and the non smokers. Furthermore, the quartile analysis showed that these findings were consistent over the different TF quartiles. The observational nature of our study makes it impossible to investigate the relation between smoking and TF any further. Previously, Sambola et al showed that smoking increases TF activity.14 In healthy volunteers current smokers had higher TF activity compared to non smoking controls. However, this study was limited by its small sample size and the base-line differences between the studied volunteers. In addition, direct comparison between both studies is difficult because TF activity was measured instead of TF antigen. Because TF antigen levels are not different in plasma and serum samples,20 it is unlikely that smoking-induced clotting ex vivo is responsible for the inverse relation found between smoking and TF in serum. Previously, some reports have suggested that gender effects positive and negative feedback mechanisms of TF.26 In line with these observations we observed a small (but not significant) difference in TF level between men and women. Only in controls such a trend was noted. Possibly, the different effect of aging on com-ponents of the TF pathway27 in combination with the difference in smoking habit could

Chapter 5

82

explain this trend found in our study. Certain aspects of the design of our study warrant caution. First, as described in the meth-ods section, CAD events were ascertained through death certification and hospital admis-sion data. This approach may have led to under ascertainment or misclassification. How-ever, validation studies in our cohort indicated high specificity of case ascertainment.28 Second, we measured levels of TF in a serum samples that were stored at –80˚C for several years. We cannot exclude some degree of protein degradation in these samples. Both limitations may have resulted in random error and misclassification which may have led to a loss of power to examine the relationship between TF levels and CAD risk. However, the TF levels measured showed a large range and random error is unlikely to produce spurious associations with risk factors such as smoking habit. Previous work in this cohort using the same study design has found strong relationships between several risk factors assayed from stored samples and cardiovascular risk.16, 28, 29

Our study indicates that elevated levels of serum TF are not associated with an increased risk of future CAD in apparently healthy individuals. These prospective data do not sup-port the hypothesis that serum TF plays a role in the pathogenesis of CAD.

ADDENDuMBecause this study is nested in the EPIC-Norfolk study this work could not have been done without the participation of many authors. Nicholas Wareham, Sheila Bingham, Robert Luben and Kay-Tee Khaw designed and conducted the EPIC-Norfolk study which is a large prospective cohort study as part of the established EPIC study (European Pro-spective Investigation into Cancer and Nutrition), a collaborative study of nine countries in Europe. Henk te Velthuis designed the tissue factor assay at Sanquin Research which was per-formed by Daniel Choi and Tymen Keller. Erik Hack, Marcel Levi, Pieter Reitsma, Victor Gerdes and Tymen Keller drafted the manuscript, revised it critically and provided im-portant intellectual content. Matthijs Boekholdt was responsible for the coronary artery disease study.

ACKNOwLEDgEMENTSWe thank the participants, general practitioners and staff in EPIC-Norfolk. EPIC-Norfolk is supported by programmed grants from the Medical Research Council UK and Cancer Research UK and with additional support from the European Union, Stroke Association, British Heart Foundation, Department of Health, Food Standards Agency and the Well-come Trust.


83

Reference List 1. Ross R. Atherosclerosis--an inflammatory disease. N Engl J Med 1999;340:115-26. 2. Nemerson Y. Tissue factor: then and now. Thromb Haemost 1995;74:180-4. 3. Cunningham MA, Romas P, Hutchinson P et al. Tissue factor and factor VIIa receptor/ligand interactions induce proinflammatory effects in macrophages. Blood 1999;94:3413-20. 4. Moons AH, Levi M, Peters RJ. Tissue factor and coronary artery disease. Cardiovasc Res 2002;53:313-25. 5. Steffel J, Luscher TF, Tanner FC. Tissue factor in cardiovascular diseases: molecular mechanisms and clinical implications. Circulation 2006;113:722-31. 6. Moreno PR, Bernardi VH, Lopez-Cuellar J et al. Macrophages, smooth muscle cells, and tissue factor in un stable angina. Implications for cell-mediated thrombogenicity in acute coronary syndromes. Circulation 1996;94:3090-7. 7. Hutter R, Valdiviezo C, Sauter BV et al. Caspase-3 and tissue factor expression in lipid-rich plaque macrophages: evidence for apoptosis as link between inflammation and atherothrombosis. Circulation 2004;109:2001-8. 8. Marmur JD, Thiruvikraman SV, Fyfe BS et al. Identification of active tissue factor in human coronary atheroma. Circulation 1996;94:1226-32. 9. Ardissino D, Merlini PA, Bauer KA et al. Thrombogenic potential of human coronary atherosclerotic plaques. Blood 2001;98:2726-9. 10. Mallat Z, Benamer H, Hugel B et al. Elevated levels of shed membrane microparticles with procoagulant potential in the peripheral circulating blood of patients with acute coronary syndromes. Circulation 2000;101:841-3. 11. Misumi K, Ogawa H, Yasue H et al. Comparison of plasma tissue factor levels in unstable and stable angina pectoris. Am J Cardiol 1998;81:22-6. 12. Suefuji H, Ogawa H, Yasue H et al. Increased plasma tissue factor levels in acute myocardial infarction. Am Heart J 1997;134:253-9. 13. Soejima H, Ogawa H, Yasue H et al. Heightened tissue factor associated with tissue factor pathway inhibitor and prognosis in patients with unstable angina. Circulation 1999;99:2908-13. 14. Sambola A, Osende J, Hathcock J et al. Role of risk factors in the modulation of tissue factor activity and blood thrombogenicity. Circulation 2003;107:973-7. 15. Day N, Oakes S, Luben R et al. EPIC-Norfolk: study design and characteristics of the cohort. European Prospective Investigation of Cancer. Br J Cancer 1999;80 Suppl 1:95-103. 16. Boekholdt SM, Keller TT, Wareham NJ et al. Serum levels of type II secretory phospholipase A2 and the risk of future coronary artery disease in apparently healthy men and women: the EPIC-Norfolk Prospective Population Study. Arterioscler Thromb Vasc Biol 2005;25:839-46. 17. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499-502. 18. Bruins P, te VH, Yazdanbakhsh AP et al. Activation of the complement system during and after cardiopulmonary bypass surgery: postsurgery activation involves C-reactive protein and is associated with post operative arrhythmia. Circulation 1997;96:3542-8. 19. van der Putten RF, te Velthuis HT, de Zwaan C et al. State and diagnostic value of plasma tissue factor in early-hospitalised patients with chest pain. Br J Haematol 2005;131:91-9. 20. van der Putten RF, te VH, Aarden LA et al. High-affinity antibodies in a new immunoassay for plasma tissue factor: reduction in apparent intra-individual variation. Clin Chem Lab Med 2005;43:1386-91. 21. Bogdanov VY, Balasubramanian V, Hathcock J et al. Alternatively spliced human tissue factor: a circulating, soluble, thrombogenic protein. Nat Med 2003;9:458-62. 22. Butenas S, Bouchard BA, Brummel-Ziedins KE et al. Tissue factor activity in whole blood. Blood 2005;105:2764-70. 23. Golino P, Ravera A, Ragni M et al. Involvement of tissue factor pathway inhibitor in the coronary circulation of patients with acute coronary syndromes. Circulation 2003;108:2864-9. 24. Felmeden DC, Spencer CG, Chung NA et al. Relation of thrombogenesis in systemic hypertension to angio genesis and endothelial damage/dysfunction (a substudy of the Anglo-Scandinavian Cardiac Outcomes Trial [ASCOT]). Am J Cardiol 2003;92:400-5. 25. Lim HS, Blann AD, Lip GY. Soluble CD40 ligand, soluble P-selectin, interleukin-6, and tissue factor in diabetes mellitus: relationships to cardiovascular disease and risk factor intervention. Circulation 2004;109:2524-8. 26. Thyzel E, Siegling S, Gotting C et al. Elevated TFPI levels in female hyperhomocysteinemic patients. Clin Biochem 2005;38:1038-40. 27. Ariens RA, Coppola R, Potenza I et al. The increase with age of the components of the tissue factor coagulation pathway is gender-dependent. Blood Coagul Fibrinolysis 1995;6:433-7. 28. Boekholdt SM, Peters RJ, Hack CE et al. IL-8 plasma concentrations and the risk of future coronary artery disease in apparently healthy men and women: the EPIC-Norfolk prospective population study. Arterioscler Thromb Vasc Biol 2004;24:1503-8. 29. Boekholdt SM, Kuivenhoven JA, Wareham NJ et al. Plasma levels of cholesteryl ester transfer protein and the risk of future coronary artery disease in apparently healthy men and women: the prospective EPIC (European Prospective Investigation into Cancer and nutrition)-Norfolk population study. Circulation 2004;110:1418-23.

Chapter 5

84


85

Chapter

6non-funCtIonal toll-lIke reCeptor 2

Is not a rIsk faCtor for aCute Coronary heart dIsease

Tymen Keller • Daniel Choi • Matthijs Boekholdt • victor gerdes • Michael Tanck Pieter Reitsma • Ron Peters • Marcel Levi

Departments of Vascular Medicine, and Cardiology, and Department of Clinical Epidemiology and Biostatistics, and Experimental Internal Medicine, Academic Medical Center, Amsterdam, The Netherlands

Submitted

ABSTRACTToll-like receptors play a key role in innate immunity and are implicated in coronary heart disease. The aim of this study was to determine whether this non-functional TLR2 vari-ant contribute to the risk of acute coronary heart disease. We performed a case-control study consisting of 376 consecutive patients admitted to the coronary care unit with acute coronary disease, and 332 control patients. In the case group there were 162 patients with acute myocardial infarction (MI) and 214 patients with unstable angina (UA); con-trols were patients attending the emergency department of the same hospital for various medical reasons. We screened for the TLR2 753 SNP by PCR analysis. Heterozygosity for TLR2 Arg753Gln was more frequent in controls than in cases (7.5 versus 5.4 %, P = 0.60). Compared with wild-type carriers, patients with the SNP753 variant had a non significant decreased risk of acute coronary heart disease (adjusted for baseline characteristics: 0.77 (95% CI 0.07 - 8.67, P = 0.83). We conclude that the non-functional TLR2 753 SNP variant is not clearly associated with an increased risk of acute coronary heart disease.

Chapter 6

88

BACKgROuNDToll-Like receptors (TLRs) are pattern recognition molecules involved in the innate im-munity.1 Both exogenous and endogenous ligands are recognized by TLRs. Upon binding, TLRs recruit adaptor molecules to their intracellular domain, leading to transcription of genes important for the innate immunity.2 To date, in humans a total of 10 TLRs have been described, each sensing a distinct repertoire of conserved microbial molecules.3

In addition to their role in immunity, TLR signaling pathways have been implicated in the pathogenesis of coronary heart disease. In mice, TLR deficiency result in reduced ath-erosclerosis and atherogenic lipid metabolism activate TLR.4, 5 In humans, atherosclerotic plaques contains high levels of TLR.6 Furthermore, genetic variants of TLR4, which is a receptor involved in the recognition of bacterial cell wall components, decrease risk of atherosclerosis.7, 8 However, the genetic association between TLR4 polymorphisms and coronary heart disease remains controversial.9-11 Recently, TLR2 has been linked with atherogenesis.12-14 In vitro studies have shown that TLR2 stimulation with peptidoglycan increased cytokine production in adventitial fibroblasts and that the use of a TLR2 ago-nist increased the expression of cytokines, enhanced neo-intima formation and increased atherosclerotic plaque size.13 Finally, TLR2 deficiency resulted in a reduction in athero-sclerosis in mice, whereas systemic exposure to a TLR2 ligand increased atherosclerotic lesions.12 In contrast to these experimental results, clinical evidence regarding the role of TLR2 in coronary heart disease is limited. A recent study showed that carriers of the non functional Arg753Gln variant had an increased risk of coronary re-stenosis after percuta-neous coronary intervention.15 Clinical studies have shown that this genetic variant of the TLR2 impairs the efficacy of receptor signaling and the capacity to elicit inflammation.16,

17 Because inflammation plays an important role in the development of coronary heart disease, non functional TLR2 may affect the risk of coronary heart disease. We hypothesize that the 753 SNP variant is associated with a decreased risk of acute coro-nary heart disease. We tested this hypothesis in a case-control study by comparing the 753 SNP prevalence in patients admitted with acute coronary heart disease and in control patients referred to the same hospital.

METHODSStudy population

The study population was recruited from the Academic Medical Center, Amsterdam, the Netherlands. Cases were consecutive male and female patients admitted to the Coronary Care Unit (CCU) with acute coronary heart disease. Acute coronary heart disease encom-passed the clinical diagnoses of acute myocardial infarction or unstable angina. Acute myocardial infarction was defined as the combination of ischemic symptoms plus signs of myocardial damage (new pathological Q waves on the ECG or elevated cardiac enzymes). Unstable angina (UA) was defined as chest pain occurring at rest with signs of myocardial ischemia on the ECG, but no elevation of cardiac enzymes. Controls were patients attend-ing the emergency department of the same hospital for various other medical reasons. Blood was collected during the admission period or hospital visit. Data on risk factors were obtained retrospectively from electronic patient files and completed by telephone

Non-functional toll-like receptor 2 is not a risk factor for acute coronary heart disease

89

interview if necessary. No differences between the data obtained from the electronic file and the questionnaire were observed. Therefore data obtained by both strategies were combined. The study protocol was reviewed and approved by the medical ethical review board, and all patients gave written informed consent.

genotyping

DNA was extracted from blood leukocytes, as described previously.18 Screening for the TLR2 Arg753Gln and Gln753Gln was performed using polymerase chain reaction (PCR) as previously described.16 Briefly, a 430 base pair fragment of the 4q arm of the genomic DNA, which included the mutation site, was replicated using forward and reverse primers as listed in table 1. Following PCR amplification, sequence analysis was performed using Pst 1 restriction enzyme. The sequence analysis was performed blinded to the phenotype of the patients. Statistics

Differences in age and BMI between cases and controls were tested using a t-test (with ad-justment for unequal variances) and values are presented as mean ± SD. Differences in cat-egorical cardiovascular risk factors were compared using a Fisher exact test and values are presented as percentages. Observed TLR2 SNP753 genotype frequencies in controls were compared with those expected assuming Hardy-Weinberg equilibrium and those observed in the cases using Fisher exact tests. Crude and adjusted odds ratios (OR) with 95% confi-dence interval (CI) were estimated for heterozygous (Arg/Gln) and homozygous (Gln/Gln) patients against wild type (Arg/Arg) patients as reference group using logistic regression. Adjustments were made for age and sex (model 1), and age, sex, BMI, smoking, hypercho-lesterolemia and medical history (model 2). This study with 376 cases and 332 controls had 80% power to detect an OR smaller than 0.35 or larger than 2.04, assuming 8% Gln carrier frequency in the controls (allele freq = 4.1%) and a dominant effect of the allele.

RESuLTSWe included 708 patients, 376 in the case group and 332 in the control group. The cases con-sisted of consecutive patients with acute myocardial infection (n=162) or unstable angina (n= 214). Controls were patients referred to the emergency department of the same hospital for

Chapter 6

90

TABLE 1: PRIMER SEQuENCES uSED FOR gENOTyPINg THE 753 SNP*

Primer Sequence 5’- 3’

Forward TATGGTCCAGGAGCTGGAGA

Reverse TGACATAAAGATCCCAACTAGACAA

* SNP indicates single nucleotide polymorphism

trauma, myalgia, erysipelas, cellulites or venous insufficiency. Table 2 shows the distribution of risk factors between cases and controls. Cases were older, more often males, and reported more risk factors for cardiovascular disease, including diabetes, hypertension, hypercholes-terolemia and a positive family history for cardiovascular disease. In addition, previous coro-nary heart disease and peripheral arterial disease were more prevalent in cases. In contrast, controls had a higher body mass index (BMI) and included more smokers.

The frequencies of heterozygous Arg/Gln and homozygous Gln/Gln patients in the con-trols were 7.5% (95% CI: 5.1-11.0) and 0.3 (95% CI: 0.1-1.9), respectively. In the cases, the frequency of heterozygous Arg/Gln patients was 5.4% (95% CI: 3.5-8.3) and of ho-mozygous Gln/Gln patients 0.3 (95% CI: 0.1-1.7). Between men and women there was no difference in distribution of the different genotypes (data not shown). In controls, the observed TLR2 SNP753 frequencies were in Hardy-Weinberg equilibrium (χ2 = 0.46, p=0.5). Although not significant (P = 0.6, Fisher’s Exact), the Arg/Gln genotype was more prevalent in the control group than in the cases. The crude odds ratios for patients car-rying the Arg/Gln or Gln/Gln genotypes were 0.7 (95% CI: 0.4-1.3), and 0.9 (95% CI: 0.1-14.3), respectively (Table 3). Adjustment for baseline characteristics (age, sex, BMI, smoking, hypercholesterolemia and history of cardiovascular disease) resulted in a lower, but non-significant (P = 0.83) OR for heterozygous patients of 0.77 (95% CI: 0.07 - 8.67). A subanalysis comparing cases only with controls without a history of cardiovascular dis-ease, gave similar results. The adjusted odds ratio for Arg/Gln patients was 0.48 (95% CI 0.19-1.18, P = 0.11) compared to wild-type Arg/Arg patients.

TABLE 2: CHARACTERISTICS OF STuDy gROuPSControls Case P-value

Patients, N 332 376

Age, y 60±16 65±12 <0.001

Male sex 37.7% 71.8% <0.001

Smoking 59.4% 41.8% <0.001

BMI* (kg/m2) 28.0 (±5.9) 26.5 (±3.6) <0.001

Diabetes 18.6% 29.2% 0.2

Hypertension 33.9% 44.4% 0.3

Hypercholerolemia 20.3% 38.9% 0.02

Atherosclerosis 3.9% 21.8% 0.01

Family history 45.8% 51.4% 0.6

TLR2 753 genotype

Arg/Arg 295 (92.2%) 332 (94.3%)

Arg/gln 24 (7.5%) 19 (5.4%)

gln/gln 1 (0.3%) 1 (0.3%)

* BMI = body mass index, # Atherosclerosis = history of coronary heart disease (angina pectoris, myocardial infarction, percutaneous coronary intervention/ coronary arterial bypass graft) or peripheral arterial disease.


91

DISCuSSIONIn this study the SNP753 variant in the TLR2 gene was not associated with a decreased risk of developing acute coronary heart disease. Carriers of the non-functional TLR2 SNP753 variant had an odds ratio of developing acute coronary heart disease of 0.77 (95% CI: 0.07 - 8.67, P = 0.83) compared to wild-type carriers. Our study results contrast with a previous study that described that the Arg753Gln increase the risk of in-stent re-stenosis.15 This first genetic association study with TLR2 polymorphisms and coronary heart disease showed in a Caucasian population who underwent successful percutaneous coronary interven-tion, that patients with the non functional TLR2 had a 3-fold increased risk of in-stent re-stenosis after 6 month follow-up (OR 3.0; 95% CI 1.1-8.7). The results of the study, seems to contradict with our study and previous studies on variants of the TLRs that described a decreased or unaffected risk of coronary heart disease in patients with non functional TLRs.8, 11 But the difference in pathophysiological mechanisms of in-stent re-stenosis and the development of coronary heart disease could be the reason for the different results. Whereas in-stent re-stenosis results of proliferation of vascular smooth muscles cells, this is a distinct process from the development of coronary heart disease. Certain aspects of our case control study merit further considerations. Although the role of the TLR2 in atherogenesis has been made plausible, the interpretation of this study is not straightforward. A central problem is the general lack of reproducibility of genetic association studies.19-21 Negative studies could be biased by genetic and environmental background or inadequate sample size. But, as demonstrated above, based on the fre-quency of the 753 SNP and the number of patients included, this study had adequate power. Nevertheless, determining which mutant genes are associated with coronary heart disease is very complex, because hundreds of genes are involved in the pathophysiology. In our study population stratification was unlikely because the control group consisted of patients admitted to the same hospital for other reasons than cardiovascular diseases and the 753 SNP variant was in linkage equilibrium both in the case group and the control group. This suggests that our results are probably not false negative. Another important issue was the prevalence of coronary heart disease in the control group. Although cases were more likely to have a history of cardiovascular disease than controls, also patients in the control group suffered from coronary heart disease. Therefore, in this study popu-lation the role of the TLR2 polymorphism in the development of acute coronary heart disease could be underestimated. Nevertheless, a sub-analysis in which controls with a

Chapter 6

92

TABLE 3: ODDS RATIOS (OR) AND 95% CONFIDENCE INTERvALS (CI) FOR TLR2 SNP753OR 95% CI P-value

Arg/gln crude 0.70 0.38 - 1.31 0.27

Model 1 0.66 0.33 - 1.31 0.23

Model 2 0.77 0.07 - 8.67 0.83

gln/gln crude 0.89 0.06 - 14.3 0.93

Model 1 0.83 0.05 - 14.2 0.90

Model 3 -- -- --

Model 1: Adjusted OR (95% CI) for age and sex. Model 2: Adjusted OR (95% CI) for age, sex, smoking, Hypercholerolemia and history of cardiovasclar disease.

history of cardiovascular disease were excluded did not produce different results. Interestingly, TLR2 seems to have a pluripotent receptor role. TLR2 recognizes bacte-rial cell wall components, in particular from gram positive bacteria.22 But recently it was also suggested that TLR2 expression is regulated by flow conditions.14 Dunzendorfer et al showed in that the expression of TLR2 on human aorta endothelial cells is upregulated when exposed to non-laminar flow. Furthermore, TLR2 seems to be involved in cell injury after myocardial infarction.23 Although it has been made likely that TLR2 is also involved in atherogenesis in mice12, 13 still limited evidence is available about the role of TLR2 in the development of coronary heart disease in humans.In summary, this study does not support an association between the non functional TLR2 variant and the risk of acute coronary heart disease. Future studies may need to focus on the pleiotropic role of the receptor and investigate down-stream adaptor proteins in the TLR2 pathway in the development of acute coronary heart disease.


93

Reference List 1. Aderem A, Ulevitch RJ. Toll-like receptors in the induction of the innate immune response. Nature 2000;406:782-7. 2. Akira S, Takeda K, Kaisho T. Toll-like receptors: critical proteins linking innate and acquired immunity. Nat Immunol 2001;2:675-80. 3. Beutler B. Inferences, questions and possibilities in Toll-like receptor signalling. Nature 2004;430:257- 63. 4. Michelsen KS, Wong MH, Shah PK et al. Lack of Toll-like receptor 4 or myeloid differentiation factor 88 reduces atherosclerosis and alters plaque phenotype in mice deficient in apolipoprotein E. Proc Natl Acad Sci U S A 2004;101:10679-84. 5. Xu XH, Shah PK, Faure E et al. Toll-like receptor-4 is expressed by macrophages in murine and human lipid-rich atherosclerotic plaques and upregulated by oxidized LDL. Circulation 2001;104:3103-8. 6. Edfeldt K, Swedenborg J, Hansson GK et al. Expression of toll-like receptors in human atherosclerotic lesions: a possible pathway for plaque activation. Circulation 2002;105:1158-61. 7. Ameziane N, Beillat T, Verpillat P et al. Association of the Toll-like receptor 4 gene Asp299Gly polymorphism with acute coronary events. Arterioscler Thromb Vasc Biol 2003;23:e61-e64. 8. Kiechl S, Lorenz E, Reindl M et al. Toll-like receptor 4 polymorphisms and atherogenesis. N Engl J Med 2002;347:185-92. 9. Morange PE, Tiret L, Saut N et al. TLR4/Asp299Gly, CD14/C-260T, plasma levels of the soluble receptor CD14 and the risk of coronary heart disease: The PRIME Study. Eur J Hum Genet 2004;12:1041-9. 10. Yang IA, Holloway JW, Ye S. TLR4 Asp299Gly polymorphism is not associated with coronary artery stenosis. Atherosclerosis 2003;170:187-90. 11. Edfeldt K, Bennet AM, Eriksson P et al. Association of hypo-responsive toll-like receptor 4 variants with risk of myocardial infarction. Eur Heart J 2004;25:1447-53. 12. Mullick AE, Tobias PS, Curtiss LK. Modulation of atherosclerosis in mice by Toll-like receptor 2. J Clin Invest 2005. 13. Schoneveld AH, Oude Nijhuis MM, van MB et al. Toll-like receptor 2 stimulation induces intimal hyperplasia and atherosclerotic lesion development. Cardiovasc Res 2005;66:162-9. 14. Dunzendorfer S, Lee HK, Tobias PS. Flow-dependent regulation of endothelial Toll-like receptor 2 expres sion through inhibition of SP1 activity. Circ Res 2004;95:684-91. 15. Hamann L, Gomma A, Schroder NW et al. A frequent toll-like receptor (TLR)-2 polymorphism is a risk factor for coronary restenosis. J Mol Med 2005;83:478-85. 16. Lorenz E, Mira JP, Cornish KL et al. A novel polymorphism in the toll-like receptor 2 gene and its potential association with staphylococcal infection. Infect Immun 2000;68:6398-401. 17. Kang TJ, Yeum CE, Kim BC et al. Differential production of interleukin-10 and interleukin-12 in mononuclear cells from leprosy patients with a Toll-like receptor 2 mutation. Immunology 2004;112:674-80. 18. Brooks-Wilson A, Marcil M, Clee SM et al. Mutations in ABC1 in Tangier disease and familial high-density lipoprotein deficiency. Nat Genet 1999;22:336-45. 19. Kathiresan S, Newton-Cheh C, Gerszten RE. On the interpretation of genetic association studies. Eur Heart J 2004;25:1378-81. 20. Ioannidis JP, Ntzani EE, Trikalinos TA et al. Replication validity of genetic association studies. Nat Genet 2001;29:306-9. 21. Peters RJ, Boekholdt SM. Gene polymorphisms and the risk of myocardial infarction--an emerging relation. N Engl J Med 2002;347:1963-5. 22. Knapp S, Wieland CW, van ‘, V et al. Toll-like receptor 2 plays a role in the early inflammatory response to murine pneumococcal pneumonia but does not contribute to antibacterial defense. J Immunol 2004;172:3132-8. 23. Shishido T, Nozaki N, Yamaguchi S et al. Toll-like receptor-2 modulates ventricular remodeling after myocardial infarction. Circulation 2003;108:2905-10.

Chapter 6

94


95

part

IIInfeCtIon and (athero)thrombotIC dIsease

Chapter

7serologIC evIdenCe of aCute respIratory

traCt InfeCtIon In patIents wIth aCute Coronary syndrome

Matthijs van wissen • Tymen Keller • geert weitjens • victor gerdes • Robert Bakker gerard van Doornum • Eric van gorp • Marcel Levi • Dees Brandjes • Harry Büller

Department of Vascular Medicine, Academic Medical Center, Amsterdam, The NetherlandsDepartment of Internal Medicine and Cardiology, Slotervaart Hospital, Amsterdam, The Netherlands

Department of Virology, Erasmus Medical Center, Rotterdam, The Netherlands

Submitted

ABSTRACTAcute respiratory infections have been associated with increased risk of acute ischaemic heart disease. This study was performed to further explore the association between re-spiratory tract infections, influenza vaccination and the occurrence of an acute coronary syndrome. In a prospective case-control study, we performed serologic tests to investigate whether infections with influenza A and B, respiratory syncytial virus, adenovirus and parainfluenza were more frequent amongst patients admitted to the coronary care unit with an acute coronary syndrome, compared to a control group of patients with stable coronary artery disease from the outpatient clinic. Furthermore, we compared symptoms of influenza-like illness (ILI) 2 weeks prior to inclusion in the study, using an ILI-score, and the number of persons vaccinated against influenza. Infections with Influenza A and B, respiratory syncytial virus and parainfluenza were more frequent in the control group compared to the case group, except for adenovirus. Patients with an acute coronary syndrome were less often vaccinated against influenza virus compared to controls (47% versus 71%, respectively; crude OR 0.41; 95% CI: 0.18-0.97). 11% of the cases had a posi-tive ILI score versus 7% in the control group (crude OR 0.99; 95% CI: 0.98-1.01). We did not find an association between serologically proven respiratory tract infection and acute coronary syndrome. Vaccination against influenza may protect against the develop-ment of an acute coronary syndrome. Due to small sample size, this study does not de-finitively rules out an association between respiratory tract infections and acute coronary syndrome. Therefore, large-scale studies are needed.

Chapter 7

100

INTRODuCTIONAtherosclerosis is considered to be an inflammatory disease.1 It has been shown that sys-temic inflammation and infections accelerate atherogenesis in animals.2 Inflammation does not only play a role in early atherogenesis, but also may be an important factor in the development of acute coronary artery disease.3 In humans, evidence is mounting that infections increase the risk of coronary artery disease (CAD). Studies have shown that there is an association between the incidence of atherosclerosis and the presence of her-pesvirus and cytomegalovirus4-7 and that inflammatory markers predict the outcome in acute vascular events.8-10 However, whether Chlamydia pneumoniae, herpes viruses and other infectious agents are really important factors in the pathogenesis of atherosclerosis and acute ischaemic events in humans is still controversial.11,12 Earlier, it was observed that during influenza epidemics, both mortality and hospitaliza-tions for cardiovascular and cerebrovascular causes are increased.13-18 Furthermore, in a case-control study, in which patients with an acute myocardial infarction and matched controls were compared, there was a statistically highly significant association of symp-toms of respiratory tract infection with the onset of myocardial infarction.19 Recently, acute respiratory tract infections have been associated with a temporary increase of the risk of acute ischaemic heart disease.20 The first three days after the clinical diagnosis of an acute respiratory infection has been made, there is a 5-fold increased risk of developing an myocardial infarction in elderly people (risk ratio 4.95; 95% CI: 4.43 to 5.53). Meier et al. also found that there was an association between respiratory tract infections and a first episode of acute myocardial infarction for a period of 10 days (OR 3.0; 95% CI: 2.1-4.4).21 If respiratory tract infections and especially influenza are related to acute coronary syn-dromes, it is conceivable that vaccination against influenza may protect against cardiovas-cular complications. Indeed, there are studies in which a protective effect of vaccination against hospitalization for coronary heart disease and cerebrovascular disease, as well as future myocardial infarction or ischaemic stroke could be demonstrated.20, 22-26

The purpose of this prospective case-control study was to further investigate whether acute respiratory infections are associated with an increased risk of acute coronary syndrome. Symptoms of influenza-like disease were registered and scored with an influenza-like disease score (ILI score). In addition, we performed serological tests to prove a respiratory virus infection in patients admitted to the CCU with an acute coronary syndrome. Finally, we determined whether vaccination against influenza decreases the risk of a future acute coronary syndrome.

PATIENTS AND METHODSStudy design

This prospective case-control study was carried out at the Coronary Care Unit (CCU) and outpatient clinic of the Slotervaart Hospital in Amsterdam from February 2004 until February 2005. Patients admitted to the CCU with a proven acute coronary syndrome, either myocardial infarction (MI) or unstable angina, were included in this study as cases. Myocardial infarction or unstable angina was considered proven when the definitions for measuring outcome in acute coronary syndromes of the American College of Cardiology

Serologic evidence of acute respiratory tract infection in patients with acute coronary syndrome

101

(ACC, 2001) were met.27 For the control group we included patients with a history of prov-en stable coronary artery disease, which visited the outpatient clinic of the department of Cardiology of the Slotervaart Hospital. Control patients were matched on age and sex. To minimize the seasonal influence on infectious diseases, we included control subjects within 7 days from inclusion of the case patient. Baseline characteristics were recorded, including vaccination against influenza prior to inclusion. The study was approved by the institutional ethical review board.

Exclusion criteria

Patients were excluded when they: were admitted to the CCU with stable angina, no signed or written informed consent was obtained, no blood was obtained within 72 hours after admission, received a coronary artery bypass graft (CABG) or a percutaneous coronary intervention (PCI) within 3 months of inclusion, acute coronary syndrome was provoked by cardiac intervention, did not speak Dutch, were not able to be reached by telephone and were not able to come to the hospital for a follow up visit. Since serological evidence for recent acute respiratory tract infection was the primary outcome of this study, patients were excluded when only unpaired blood samples were obtained.

Laboratory investigations

Blood was obtained from case patients within 72 hours of admission and from controls within 7 days of admission of their matched case patient. A second blood sample was collected at the follow-up visit 2 weeks after inclusion. For the assessment of virus in-fection in the blood samples, we used ELISA kits for IgA and IgG against influenza A and B, respiratory syncytial virus, adenovirus and parainfluenza (Serion ELISA classic; Institute Virion\Serion GmbH, Würzburg, Germany). A positive result for IgG against influenza A and B and RSV was considered when the antibody activity was >15 U/ml, for adenovirus >13 U/ml, for parainfluenza 1 >35 U/ml and for parainfluenza 2 >150 U/ml. For IgA this was >15 U/ml for influenza A and B and RSV, for adenovirus >14 U/ml and >25 U/ml for both parainfluenza 1 and 2. When IgG antibodies, but not the IgA antibodies in both samples were above cut-off values, the patient was considered to have the infection. Also, when the first sample was below and the second above the cut-off values, this was considered a positive result. When IgA antibodies but not IgG in any of the 2 samples were above cut-off values, the patient was considered not to have a recent infection.

Influenza-like illness (ILI) score

A questionnaire of symptoms of an acute respiratory tract infection was completed at baseline, including: acute onset of fever (temperature ≥ 37,8°C), subjective feeling of fever or chills, cough with or without sputum, rhinitis, headache and sore throat. Based on the results of previous studies on clinical signs and symptoms predicting influenza infec-tion, an influenza-like illness (ILI) score was developed from these symptoms.28,29 The ILI score was considered positive when there was an acute onset of fever (> 37.8°C) plus one of the above mentioned symptoms.

Chapter 7

102


Results are presented as means plus or minus standard deviation (SD) and in percent-ages. Crude and adjusted odds ratios (OR) with their corresponding 95% confidence in-tervals (95% CI) were calculated using logistic regression analysis, to indicate the as-sociation between respiratory tract infection and acute coronary syndrome and between vaccination against influenza and acute coronary syndrome. We adjusted the crude odds ratios for age, sex, smoking, diabetes, body mass index, previous stroke, previous myocar-dial infarction, prior angina, hypertension, dyslipidemia and family history of coronary artery disease.

RESuLTSWe included 188 patients, 82 cases and 106 controls. A total of 37 cases and 38 controls were excluded, because from these subjects only unpaired blood samples were obtained. Eventually, from 45 cases and 68 controls paired samples were available and these were included for final analysis.Baseline characteristics of both patient groups are shown in table 1. There was no difference in patient characteristics between the 45 cases included in the analysis and the 37 excluded cases.

Influenza A and B, RSV and parainfluenza infections were more frequent in the control group compared to the case group, except for adenovirus infections. The frequencies of infections in both patient groups and corresponding odds ratios are shown in table 2. There were less subjects vaccinated against influenza in the year prior to inclusion in the case group compared to the control group, 21 cases (47%) versus 48 controls (71%), respectively. The crude OR was 0.41 (95% CI: 0.18-0.97). After adjustment for risk factors and potential confounding factors, the OR was 2.24 (95% CI: 0.15-33.33). There were more subjects with a positive ILI score in the case group compared to the control group, 11% (n=57) versus 7% (n=5), respectively. The crude OR was 0.99 (95% CI: 0.98-1.01), after adjustment the OR was 1.07 (95% CI: 0.89-1.28).

TABLE 1: STuDy gROuP CHARACTERISTICS

Controls (n = 68) Cases (n = 45)

Sex (male) 45 (66.2%) 30 (66.7%)

Age (SD) 67 (12) 66 (13)

Previous MI 43 (63.2%) 7 (15.6%)

AP 27 (39.7%) 10 (22.2%)

TIA 7 (10.3%) 4 (8.9%)

Stroke 4 (5.9%) 2 (4.4%)

smoking 29 (42.6%) 20 (44.4%)

Family history of CAD 9 (13.2%) 9 (20.0%)

Hypertension 22 (32.4%) 20 (44.4%)

Dyslipidemia 34 (50.0%) 15 (33.3%)

DM 15 (22.1%) 10 (22.2%)

Influenza vaccination 48 (70.6%) 21 (46.7%)

Medication use

OAC 16 (35.6%) 7 (15.6%)

Anti-platelet 50 (73.5%) 16 (23.5%)

SD = Standard Deviation; TIA = transient ischaemic attack; MI = myocardial infarction; AP = angina pectoris; Vaccination: vaccinated against influenza in the year prior to current ´flu´ season. CAD = coronary artery disease. DM = diabetes mellitus; OAC = oral anticoagulation.


103

DISCuSSIONThe purpose of this prospective case-control study was to further investigate whether acute respiratory infections are associated with an increased risk of acute coronary syn-drome and whether vaccination against influenza reduces this risk. We did not observe a higher rate of patients with serologically proven respiratory tract infection among pa-tients with an acute coronary syndrome compared to patients with stable coronary artery disease. In addition, we observed a lower rate of vaccination against influenza among the patients with acute coronary syndrome. In previous studies an association between preceding symptoms of a respiratory tract infection and cardiovascular disease was reported.13,17-19,30-32 In some of these studies the number of included patients was limited and in none of the studies the clinical diagnosis of respiratory tract infection was confirmed with serologic tests or any other methods. As a consequence, no distinction could be made between proven respiratory tract infection and signs resembling respiratory tract infection caused by an acute coronary syndrome. In our study, the number of infections with influenza (and that of the other viruses we examined, except for adenovirus) was remarkably higher in the control group than in the case group, despite the fact that more people in the control group were vaccinated against influenza. This was an unexpected result. The wide confidence intervals of the odds ra-tio’s, a result of the limited sample size, indicate that this could be the result of chance. However, when we combined all viruses, the difference was statistically significant. The difference can not be explained by seasonal variation either, since control patients were included in the same week as the cases. These serologic results suggest that the risk of acute coronary syndrome in the weeks after a respiratory tract infection is considerably lower than suggested by previous studies.Several studies investigated the association between vaccination against influenza and risk of acute coronary syndrome. In a case-control study of patients with coronary artery disease, vaccination rate was lower among patients with a new myocardial infarction (OR 0.33, 95% CI: 0.13-0.82).24 Another study reported that patients who were admitted with primary car-diac arrest were less often vaccinated compared to a control group.33 However, this was a retrospective study, and therefore selection bias could not be ruled out. Nichol and al. showed that vaccination against influenza reduced the risk of hospitalization for cardiac disease (P < 0.001).22 A prospective randomized study showed that the risk of death and ischaemic events

Chapter 7

104

TABLE 2: FREQuENCy OF vIRuS INFECTIONS

Controls (n=68) Cases (n=45) Crude OR Adjusted OR

Influenza A 13 (19.1%) 4 (8.9%) 0.41 (0.13 - 1.36) 0.67 (0.03 - 18.07)

Influenza B 17 (25.0%) 6 (13.3%) 0.46 (0.17 - 1.28) 0.17 (0.01 - 3.42)

RSv 13 (19.1%) 2 (4.4%) 0.20 (0.04 - 0.92) 0.09 (0.004 - 2.17)

Adenovirus 11 (16.2%) 8 (17.8%) 1.12 (0.41 - 3.05) 0.09 (0.00 - 2.66)

Para influenza 9 (13.2%) 4 (8.9%) 0.64 (0.19 - 2.22) 1.00 (0.09 - 11.74)

Combined* 40 (58.8%) 17 (37.8%) 0.43 (0.20 - 0.92) 0.09 (0.01 - 0.71)

Number of infections with influenza A, B, RSV, adenovirus and para influenza, with percentages. Odds ratios (OR) with corresponding 95% confidence intervals (95% CI); OR are adjusted for adjusted for age, sex, smoking, diabetes, body mass index, previous stroke, previous myocardial infarction, prior angina, hypertension, dyslipidemia and family history of coronary artery disease. RSV = Respiratory Syncytial Virus; OR = odds ratio; * Combined: patients with a virus infection, regardless whether they had a single infection or multiple simultaneous infections

in patients suffering from infarction and post-angioplasty during flu season was reduced (relative risk 0.34, 95% CI: 0.17-0.71).26 However, Jackson et al. did not observe a reduction in the risk of recurrent events after vaccination.34 The results of our study regarding vaccination are in line with most of the previous studies (OR 0.41; 95% CI: 0.18-0.97) and vaccination may protect against future coronary syndrome. However, since the control patients were all individuals with proven coronary artery disease, 63% with a previous myocardial infarction, and only 16% of the cases had a history of myocardial infarction, it is likely that a higher pro-portion of patients in the control group had a reason to receive influenza vaccination during the period preceding this study. This may at least partially explain the observed difference and we stress that only well designed prospective studies can prove whether vaccination is protective for future ischaemic events or not.The results of the influenza-like illness questionnaire were not in line with the serological results. We developed the influenza-like illness score based on the results of previous studies on clinical signs and symptoms predicting influenza infection.28,29 There were no statistically relevant correlations between the ILI score and serology results (data not shown). The results indicate that these signs and symptoms are non-specific in patients with clinically manifest acute coronary syndromes. The use of a questionnaire like our ILI score may be useful for studies in the general population, with a low frequency of acute coronary syndrome and pul-monary embolism, but we question the use of the ILI questionnaire in this clinical setting. Also other aspects of our study require comment. A limitation of this study is that for the detection of respiratory virus infections we only used serum samples. A more precise reflec-tion of the presence of influenza infection in our patients would have been obtained when a throat swab, for direct isolation of viruses, had been obtained from all patients as well.35 The fact that we only used serum for the detection of viruses could lead to a possible underestima-tion of infection. However, this underestimation will be present in both patient groups, and therefore it is not likely that addition of throat swabs will alter the main conclusion.From a subset of patients unpaired samples were obtained. Because in case of an unpaired sample with an antibody activity above the cut-off value, one can not conclude whether the infection is acute, e.g. whether the antibody titers are increasing or decreasing. For final analyses, we therefore excluded patients from whom unpaired samples were obtained. Although patient characteristics were similar, this may have introduced bias and the ex-clusion of patients reduced the sample size. Due to this small sample size, the study has limited power. We regard this as a pilot study and are currently performing a prospective case-control study in which, besides serology, also throat swabs are systematically ob-tained. Based upon our findings, with a frequency of 37.8% infected subjects in the case and 58.8% in the control group (all viruses combined), we will need at least 88 cases and 175 matched controls to detect a significant difference between the study groups (power 0.90, statistically significant level at 0.05).In summary, we did not find an association between respiratory virus infection and an in-creased risk for cardiovascular disease. Furthermore, we found a lower rate of vaccination against influenza among patients with an acute coronary syndrome compared to patients with stable angina. However, our sample size was too small to draw definitive conclusions. Therefore, a study of larger scale is needed with optimal diagnostics.


105

References 1. Ross R. Atherosclerosis--an inflammatory disease. N Engl J Med. 1999;340:115-26. 2. Alber DG, Powell KL, Vallance P, Goodwin DA et al. Herpesvirus infection accelerates atherosclerosis in the apolipoprotein E-deficient mouse. Circulation. 2000;102:779-85. 3. Willerson JT. Systemic and local inflammation in patients with unstable atherosclerotic plaques. Prog Cardiovasc Dis. 2002;44:469-78. 4. Libby P, Egan D, Skarlatos S. Roles of infectious agents in atherosclerosis and restenosis: an assessment of the evidence and need for future research. Circulation. 1997;96: 4095-103. 5. Hendrix MG, Salimans MM, van Boven CP et al. High prevalence of latently present cytomegalovirus in arterial walls of patients suffering from grade III atherosclerosis. Am J Pathol. 1990;136:23-8. 6. Ridker PM, Hennekens CH, Buring JE et al. Baseline IgG antibody titers to Chlamydia pneumoniae, Helicobacter pylori, herpes simplex virus, and cytomegalovirus and the risk for cardiovascular disease in women. Ann Intern Med. 1999;131:573-7. 7. Ridker PM, Hennekens CH, Stampfer MJ et al. Prospective study of herpes simplex virus, cytomegalovirus, and the risk of future myocardial infarction and stroke. Circulation. 1998;98:2796-9. 8. Lindahl B, Toss H, Siegbahn A et al. Markers of myocardial damage and inflammation in relation to long-term mortality in unstable coronary artery disease. FRISC Study Group. Fragmin during Instability in Coronary Artery Disease. N Engl J Med. 2000;343:1139-47. 9. Danesh J, Wheeler JG, Hirschfield GM et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med. 2004;350:1387-97. 10. Keaney JF Jr, Vita JA. The value of inflammation for predicting unstable angina. N Engl J Med. 2002;347:55-7. 11. Ieven MM, Hoymans VY. Involvement of Chlamydia pneumoniae in atherosclerosis: more evidence for lack of evidence. J Clin Microbiol. 2005;43:19-24. 12. Andraws R, Berger JS, Brown DL. Effects of antibiotic therapy on outcomes of patients with coronary artery disease: a meta-analysis of randomized controlled trials. JAMA. 2005;293:2641-7. 13. Bainton D, Jones GR, Hole D. Inluenza and ischaemic heart disease--a possible trigger for acute myocardial infarction? Int J Epidemiol. 1978;7:231-9. 14. Eickhoff TC, Sherman IL, Serfling RE. Observations on excess mortality associated with epidemic influenza. JAMA. 1961;176:776-82. 15. Housworth J, Langmuir AD. Excess mortality from epidemic influenza, 1957-1966. Am J Epidemiol. 1974;100:40-8. 16. Alling DW, Blackwelder WC, Stuart-Harris CH. A study of excess mortality during influenza epidemics in the United States, 1968-1976. Am J Epidemiol. 1981;113:30-43. 17. Spencer FA, Goldberg RJ, Becker RC et al. Seasonal distribution of acute myocardial infarction in the second National Registry of Myocardial Infarction. J Am Coll Cardiol. 1998;31:1226-33. 18. Glezen WP, Payne AA, Snyder DN, Downs TD. Mortality and influenza. J Infect Dis. 1982;146:313-21. 19. Spodick DH, Flessas AP, Johnson MM. Association of acute respiratory symptoms with onset of acute myocardial infarction: prospective investigation of 150 consecutive patients and matched control patients. Am J Cardiol. 1984;53:481-2. 20. Smeeth L, Thomas SL, Hall AJ et al. Risk of myocardial infarction and stroke after acute infection or vaccination. N Engl J Med 2004;351:2611-8. 21. Meier CR, Jick SS, Derby LE et al. Acute respiratory-tract infections and risk of first-time acute myocardial infarction. Lancet. 1998;351:1467-71. 22. Nichol KL, Nordin J, Mullooly J et al. Influenza vaccination and reduction in hospitalizations for cardiac disease and stroke among the elderly. N Engl J Med. 2003;348:1322-32. 23. Nichol KL. Complications of influenza and benefits of vaccination. Vaccine. 1999;17 Suppl 1:S47-52. 24. Naghavi M, Barlas Z, Siadaty S et al. Association of influenza vaccination and reduced risk of recurrent myocardial infarction. Circulation. 2000;102:3039-45. 25. Lavallee P, Perchaud V, Gautier-Bertrand M et al. Association between influenza vaccination and reduced risk of brain infarction. Stroke. 2002;33:513-8. 26. Gurfinkel EP, de la Fuente RL. Two-year follow-up of the FLU Vaccination Acute Coronary Syndromes (FLU VACS) Registry. Tex Heart Inst J. 2004;31(1):28-32. 27. Cannon CP, Battler A, Brindis RG et al. American College of Cardiology key data elements and definitions for measuring the clinical management and outcomes of patients with acute coronary syn dromes. A report of the American College of Cardiology Task Force on Clinical Data Standards (Acute Coronary Syndromes Writing Committee). J Am Coll Cardiol. 2001;38:2114-30. 28. Monto AS, Gravenstein S, Elliott M et al. Clinical signs and symptoms predicting influenza infection. Arch Intern Med. 2000;160:3243-7. 29. Boivin G, Hardy I, Tellier G, Maziade J. Predicting influenza infections during epidemics with use of a clinical case definition. Clin Infect Dis. 2000;31:1166-9. 30. Marshall RJ, Scragg R, Bourke P. An analysis of the seasonal variation of coronary heart disease and respiratory disease mortality in New Zealand. Int J Epidemiol. 1988;17:325-31. 31. Pesonen E, Siitonen O. Acute myocardial infarction precipitated by infectious disease. Am Heart J. 1981;101:512-3.

Chapter 7

106


107

32. Mattila KJ. Viral and bacterial infections in patients with acute myocardial infarction. J Intern Med. 1989;225:293-6. 33. Siscovick DS, Raghunathan TE, Lin D et al. Influenza vaccination and the risk of primary cardiac arrest. Am J Epidemiol. 2000;152:674-7. 34. Jackson LA, Yu O, Heckbert SR et al. Influenza vaccination is not associated with a reduction in the risk of recurrent coronary events. Am J Epidemiol. 2002;156:634-40. 35. Covalciuc KA, Webb KH, Carlson CA. Comparison of four clinical specimen types for detection of influenza A and B viruses by optical immunoassay (FLU OIA test) and cell culture methods. J Clin Microbiol. 1999;37:3971-4.

Chapter

8Influenza vaCCIne for preventIng

aCute Coronary syndromes

Tymen Keller • viola weeda • Carlo van Dongen • Albert MairuhuDees Brandjes • Marcel Levi

Department of Vascular Medicine, Academic Medical Center, University of Amsterdam, the Netherlands Department of Internal Medicine, Slotervaart Ziekenhuis, Amsterdam, the Netherlands

Cochrane Database of Systematic Reviews 2004, Issue 4 (protocol)

ABSTRACTBackground: It has been suggested that influenza vaccination reduces the risk of acute coronary syndromes. However the evidence underlying this suggestion is scarce and the size of the benefit is unknown. Objectives: To assess the potential benefit of influenza vaccination for primary and secondary prevention of acute coronary syndromes. Search strategy: We searched the Cochrane Central Register of Controlled Trials (The Cochrane Library Issue 2, 2005), MEDLINE (From Januari 1996 to June 2005) and EMBASE (from Januari 1996 to June 2005) database. Furthermore, we search in June 2005 databases for recent or ongoing trials ((http://www.epi.bris.ac.uk/Cochrane/cardi.html) and “digital dissertations” (wwwlib.umi.com/dissertations), and reference lists of articles. Last we contacted pharmaceutical companies (MSD and Solvay) for non pub-lished trials on influenza vaccination. No language restrictions were applied. Studies were included when participants were people randomized to either influenza vaccination or placebo or no treatment, with or without a history of cardiovascular disease. Selection criteria: Randomized clinical trials of influenza vaccination with placebo or no treatment in primary or secondary prevention with outcome on acute coronary syndromes. Data collection & analysis: Data extracted and the assessment of quality was done with a predefined form by two review authors independently. We contacted investigators when data on the outcome were missing. Main results: We included two original randomized controlled trials. A total of 1030 par-ticipants received influenza vaccination or placebo. One study including 729 participants showed no significant effect of influenza vaccination on mortality. The other study was of moderate quality and quantified the potential benefit of influenza vaccination in the primary and secondary prevention of acute coronary syndromes. This randomized con-trolled trial showed that influenza vaccination was associated a significant decreased risk on cardiovascular death in the vaccinated patients compared to the controls (OR 0.34; 95% confidence interval 0.17-0.71). Reviewers’ conclusions: There is not enough data available to quantify the potential ben-efit of influenza vaccination on the risk of acute coronary syndromes.

Chapter 8

110

SyNOPSISVaccination against influenza has been associated with a reduction in the risk of acute coronary syndromes. In order to quantify the potential benefit of influenza vaccination on the risk of acute coronary syndromes, we searched all available randomized clinical trials. Two suitable randomized clinical trials were obtained. Studies were small and the methodological quality was moderate. There is not enough data available to quantify the potential benefit of influenza vaccination on the risk of acute coronary syndromes.

BACKgROuNDCardiovascular disease is a leading cause of morbidity and mortality. Annually 7.2 million people die from cardiovascular disease worldwide.1 Primary and secondary prevention together reduces the risk of cardiovascular disease but only half of the patients with acute cardiovascular disease remain free of recurrent events in the following five year. These data demonstrate that new treatment modalities and better prevention strategies are needed. Recently, epidemiologi-cal, experimental and clinical studies have demonstrated an association between infection and atherosclerosis. Therefore, the treatment and prevention of infection seems to be a potential strategy. Two large randomized controlled trials have shown that long-term treatment with antibiotics do not reduce the rate of acute coronary syndromes.2, 3 But despite these negative trials, evidence for the role of infection in cardiovascular disease continues to mount.4, 5 Acute respiratory tract infections, including influenza, are potential diseases associated with acute coronary syndromes.4 Smeeth et al showed in a large observational study that the risk of acute coronary syndromes is 5-fold increased in the first week after respiratory tract infections. Furthermore, several reports showed that influenza vaccination protects against acute coronary syndromes.6-8 Hospitalization for acute coronary syndromes is reduced with 19% in subjects who receive influenza vaccination.6 The beneficial effects of influenza vaccination are recently confirmed in a randomized controlled trial.9 This study, performed in patients with manifest coronary heart disease, showed a statistical significant reduction in mortality from acute coronary syndromes after 1 year follow-up. Because effective influenza vaccines are cheap and widely available, the use of influenza vaccination to prevent acute coronary syndromes is an appealing prevention strategy for those at risk. Whether influenza vaccination reduces the number of (recurrent) acute coronary syndromes remains not fully established. This systematic review is an attempt to quantify the size of the effect of influenza vaccination for people with and without cardiovascular disease in the prevention of acute coronary syndromes.

OBjECTIvESThe objective is to quantify the potential benefit and the potential harms of influenza vac-cination in the primary and secondary prevention of acute coronary syndromes.

CRITERIA FOR CONSIDERINg STuDIES FOR THIS REvIEwTypes of studies

Any randomized controlled trials comparing influenza vaccination with placebo or no treatment and with data on one of the outcomes were included. Studies with control vac-

Influenza vaccine for preventing acute coronary syndromes

111

cine or no intervention or comparing types, doses or schedules of influenza vaccine were included if placebo or control was added in the study.

Types of participants

People of all ages who were randomized to either influenza vaccination or placebo or no treatment were included. Participants were included if influenza vaccination was given as a routine prevention program, as primary prevention of myocardial infarction and as secondary prevention. Participants may had or had not a history of cardiovascular dis-ease (stable/unstable angina, myocardial infarction, stroke or peripheral arterial disease). Participants could be of either sex. Participants on medication were included.

Types of interventions

Influenza vaccination (inactivated whole virus, detergent-treated split products, or puri-fied hemaglutinin and neuramidase surface antigen formulation of the three influenza virus strains influenza A, B or C) administered by any route, at any dosage. Studies in which co-interventions are used were not rejected.

Types of outcome measures

The following clinical outcomes were considered during 1 year follow up: (1) Myocardial infarction or re-infarction; (2) Unstable angina; (3) Death from cardiovascular causes; (4) Any combination of the above mentioned outcomes. (5) All cause mortality SEARCH STRATEgy FOR IDENTIFICATION OF STuDIES

The search strategy of this review was developed in accordance with the Cochrane Heart Group guidelines. No language restriction was applied.All randomized controlled trials of influenza vaccination compared to placebo or no treat-ment with data on one of the outcomes were sought using the following methods:

(1) The Cochrane Central Register of Controlled Trials (CENTRAL) (June 2005) was searched using the following search terms:#1 INFLUENZA VACCINE #2 (influenza* near vaccin*) #3 (influenza* near immu-ni*) #4 (flu near immuni*) #5 (flu near vaccin*) #6 (#1 or #2 or #3 or #4 or #5) #7 CARDIOVASCULAR DISEASES #8 heart #9 coronary #10 cardiac #11 myocardial #12 cardiovascular #13 angina #14 (#7 or #8 or #9 or #10 or #11 or #12 or #13) #15 (#6 and #14) The search was updated by searching MEDLINE on Ovid (from 1996 to June 2005) and EMBASE on Ovid (from 1999 to June 2005) using a standard RCT filter (Dickersin 1994, Lefebvre 1996). Search term for the specific databases are listed in table 01 and 02.

(2) Websites for recent or ongoing trials including (http://www.epi.bris.ac.uk/Cochrane/cardi.html) and “digital dissertations” (wwwlib.umi.com/dissertations) were search in June 2005.(3) In addition, two large suppliers of influenza vaccine (MSD and Solvay) were contacted for information on published and unpublished relevant trials. Furthermore, extensive

Chapter 8

112

manual searching was done by one of the review authors for cross references in original articles and reviews. Primary authors of retrieved RCT not reporting data on one of the outcomes were contacted in order to find unpublished data.

METHODS OF THE REvIEwStudy selection Potential eligible trials were systematically selected from the search by two review authors independently. The studies were excluded if it could be concluded from the title and abstract that the study was not a randomized controlled trial. When a conclusion could not be made with certainty, the full text of the report was retrieved for further evaluation. If a trial was excluded after this point a record of the article and the reason for exclusion was kept. A predefined “in-out” form was used for evaluation. The full text was assessed by two reviewers independently.

Quality assessment A scoring system was used to assess the methodological quality in accordance with guide-lines in the Cochrane Handbook for Systematic Reviews of Interventions (version 4.2.5). Methodological quality was based on the method of randomization, the concealment of allocation, the blinding of outcome assessment, and intention-to-treat analysis. Based on these criteria the likelihood of bias was defined as low (quality A: all quality criteria are met), moderate (quality B: one or more of the quality criteria are only partly met) and high (quality C: one or more criteria are not met).

Data extraction Two reviewers independently extracted data based on a predefined review form. The follow-ing data were extracted systematically: (1) trial characteristics: design, duration, randomization (and method), allocation concealment (and method), blinding (outcome assessors), check-ing of blinding; (2) intervention: type of vaccination, method of vaccination; (3) participants: exclusion criteria, total number and number in comparison groups, gender/age, similarity of groups at baseline, withdrawals/losses to follow-up; (4) outcome: myocardial infarction or re-infarction, unstable angina, death from cardiovascular causes, death from any cause.

Resolution of differences

The included studies were independently assessed by two reviewers and in the instances that two review authors disagree over the grading, inclusion data extraction or method-ological quality, the opinion of a third reviewer was decisive.

Description of studies

We retrieved 392 references from our search strategy. Editorials, health care articles, case-series, case-control studies, cohort studies and duplicates were excluded (n=378). A total of 16 randomized controlled trials were identified but 7 trials were excluded because no influenza vaccination was tested. In addition 6 randomized controlled trials were excluded because no data on one of the outcomes was available. Two published random-


113

ized controlled trials were based on the same trial. Finally, two randomized controlled trial were included. As far as we know two additional trials are ongoing but no information on these trails could be obtained. Detailed description of study designs are not published yet and therefore the references were put into the “awaiting assessment” category.

Methodological quality of included studies

Two randomized controlled trials were included. Allsup et al, performed a placebo con-trolled randomized trial in which the benefit of influenza vaccination was assessed in people aged 65-74 years.10 This trial was performed in 20 general practitioner practices in Liverpool (United Kingdom). Primary outcome were pneumonia and influenza-like disease recorded by a general practitioner. Secondary outcomes were i.e. self reported influenza-like disease, hospitalization with respiratory tract infection and death from any causes. Randomization was done with a computer generated random allocation sequence. One member of the research team was aware of which participant received placebo. Participants were unaware of which intervention he or she had received. Data on the primary outcome was available of all randomized participants.

The FLUVAC study was the second trial which met all the inclusion criteria.9 The proto-col of this study has been published in detail previously.11 In fact, the FLUCAV study were two trials described as one trial. Patient admitted with acute myocardial infarction (MI) were randomized to receive either influenza vaccination or placebo and patients planned for percutaneouse coronary intervention (PCI) were also randomized to influenza vac-cination or placebo. This trial was performed in 6 healthcare centers in Argentina and in 2 sites in Buenos Aires City. The study randomized 200 patients with acute myocardial infarction and 101 patients planned for percutaneous coronary intervention to receive either influenza vaccination or placebo. A small number of lost to follow-up was reported, and two patients from the control group received influenza vaccination. The results from the FLUVAC study have been published in three different articles. The 6 month, 1 year and the two years follow-up data published separately. For this systematic review only the one-year follow-up data were used. According to the scoring system described in the Cochrane Handbook for Systematic Reviews of Interventions,12 there was a moderate likelihood of bias inherent in this trial. Although patients were randomization, the meth-od of randomizations and the concealment of allocation is not clear. We tried to contact the researchers but no additional information was obtained. Therefore, the likelihood of bias was defined as moderate (methodological quality B).

RESuLTSPrimary prevention

In the study performed by Allsup et al. the overall mortality was higher in the influenza vaccinated participants, however this difference was not statistical significant. Participants were aged 65-74 years for who a influenza vaccination is recommended (according to the UK Department of Health) A total of 3 patients died in the vaccinated participants (n=552) and 1 in the placebo group (177). The cause of death was cancer in all participants.

Chapter 8

114

Secondary prevention

In the FLUVAC study patients known with coronary heart disease were randomized. This small pilot study, consisted of two different randomized patient groups. The first group of patients had to undergo an elective percutaneous coronary intervention (PCI). The second group consisted of patients with an acute myocardial infarction (AMI) admitted to the coronary care unit of one of the participating centers. Both patients with ST-segment myocardial infarction or non-ST-segment myocardial infarction were included.In the PCI group 101 patients were randomized to receive either influenza vaccination (A/Moscow/10/99-like virus, A/New Caledonia/20/99 (H1N1)-like virus, and AB/Sichuan/379/99-like virus)(n=51) or a saline infusion as placebo (n=50). In the AMI group, 200 patients were randomized to receive either influenza vaccination (n= 100) or placebo (n = 100). Two patients assigned to the control group received influenza vaccina-tion. A total of 9 patients were lost to follow-up, of which 7 in the MI group (4 vaccinated and 3 non vaccinated) and 2 patients in the PCI group (both in the vaccinated group).There was no association between vaccination and the risk of acute myocardial infarction (OR 1.00 (95% CI: 0.33-3.00)). Cardiovascular death occurred in 9 vaccinated patients and in 26 non vaccinated patients. Therefore influenza vaccination was associated a significant decreased risk on cardiovascular death in the vaccinated patients compared to the controls (OR: 0.34; 95% CI: 0.17-0.71).

DISCuSSIONTwo different studies could be included in this systematic review of the literature of randomized controlled trials which study the relation between influenza vaccination and the risk of acute coronary syndromes. In a pooled analysis it was shown that influ-enza vaccination results in a decreased incidence of cardiovascular death. In patients with angina pectoris or myocardial infarction, influenza vaccination protects against cardiovascular death. However, the reliability of one of the trials was moderate because of the moderate methodological quality and therefore the results of the study is poten-tial biased. Interestingly, although the number of cardiovascular death were different between influenza vaccination participants and placebo treated controls, the incidence of acute myocardial infarction was the same. Because the effect of influenza vaccination on cardiovascular death was shown in patients with a history of cardiovascular disease, no conclusion can be drawn about the effect of influenza vaccination in primary prevention. With the limited evidence available it is not possible to quantify the beneficial effect of influenza vaccination in the prevention of acute coronary syndromes.

Observational studies have reported similar findings. In 2000 Naghavi and Siscovick reported independently that influenza vaccination was associated with a reduction in the risk of myocardial infarction in the subsequent influenza season.7, 8 Naghavi performed a case-control study in patients with a first time acute myocardial infarction. Controls were patients admitted to the same hospital without AMI or unstable angina. Influenza vaccination was associated with a decrease risk of recurrence in the subsequent influenza season (OR: 0.33, 95% CI: 0.13 - 0.82, P = 0.02). In a similar designed case-control study,


115

Siscovick included 342 patients with a first time AMI and matched 549 controls, identified using from the community by using random digit dialing. After adjustment for demo-graphic, clinical, and behavioural risk factors, influenza vaccination was associated with a reduced risk of recurrence (OR: 0.51, 95% CI: 0.33 - 0.79). Furthermore et al, reported that influenza vaccination was associated with a reduction in brain infarction.13 A total of 90 consecutive patients admitted to the hospital for brain infarction reported less influenza vaccination than 180 population-based controls, matched for age, sex, and district of resi-dency in Paris. He concluded that influenza vaccination protects against brain infarction (OR: of 0.50, 95% CI: 0.26 - 0.94; P = 0.03). And in another recent study the beneficial effect of influenza vaccination on stroke was also reported. Grau et al, interviewed case patients admitted with stroke and assessed vaccination status, risk factors, health-related behavior, and socioeconomic factors. Case patients reported less vaccination than the control patients randomly selected from the population. Therefore the authors concluded that vaccination was associated with a reduced risk of stroke (OR: 0.46, 95% CI: 0.28 - 0.77), which was only true for influenza vaccination and during the winter. All these studies point towards a protective effect of influenza vaccination against acute coronary syndromes and stroke. But the designs of these studies are prone to bias, such as the selec-tion of a subgroup of patients at low risk for stroke because of a healthier lifestyle. In contrast to these uniform results, two large cohort studies are recently published with differential data.6, 14 Nichol et al performed a large observational study in which two cohorts of community-dwelling members of three large managed-care organizations were studied during the 1998-1999 and 1999-2000 influenza seasons. He included 140.055 subjects in the 1998-1999 cohort and 146.328 in the 1999-2000 cohort. Of these subjects 55.5 percent and 59.7 percent were immunized, respectively. Vaccination against influenza was associated with a reduction in the risk of hospitalizations for cardiac disease (reduction of 19 percent during both seasons [P < 0.001]) and cerebrovascular disease (reduction of 16 percent during the 1998-1999 season [P < 0.02] and 23 percent during the 1999-2000 season [P < 0.001]). Nichol et al concluded that in elderly patients, vaccination against influenza is associated with reductions in the risk of hospitalization for acute coronary syndromes and stroke. However, one can not exclude that a certain selection bias of healthier people in the vaccinated group has led to an overestimation of the effect size. In contrast, based on a large cohort study Jackson et al, showed dif-ferent results.14 Jackson conducted a population-based inception cohort study of 1.378 Group Health Cooperative enrollees who survived a first myocardial infarction in 1992 through 1996. Recurrent coronary events, influenza vaccinations, and other covariates were identified by chart review and from administrative data systems. A total of 127 recurrent coronary events were identified during the median 2.3-year follow-up period. This study showed that influenza vaccination was not associated with a reduced risk of recurrent acute coronary syndromes during the corresponding period of November through October (adjusted hazard ratio (HR) 1.18, 95% CI: 0.79 - 1.75) or during the cor-responding periods of expected influenza activity (November through April) (HR 1.06, 95% CI: 0.63 - 1.78) or inactivity (May through October) (HR 1.34, 95% CI: 0.76 - 2.36).

Chapter 8

116

Differences in results between these studies could be explained by differences in study population. About half of the patients in the study by Jackson et al were younger than 65 years of age, and all of them had survived a myocardial infarction. Nichol et al included only patients older than 65 years and just a quarter of the patients were known with heart disease. Furthermore, differences in populations, the spectrum of outcomes and the outcome periods used in both studies could explain the differential results. Jackson et al looked at the six-month period from November through April. Nichols et al used a more specific outcome period based on influenza-surveillance data. In conclusion, based on the limited randomized controlled data found it is not possible to draw conclusions with respect to the protective potential of influenza vaccination against acute coronary syndromes. Whether influenza vaccination results in a decrease in acute coronary syndromes remains not known. Future randomized controlled studies for primary and secondary prevention are needed to assess the real benefit of influenza vaccination in preventing acute coronary syndromes.

REvIEwERS’ CONCLuSIONSImplications for practice. In this systematic review we tried to quantify the benefit effect of influenza vaccination in the primary and secondary prevention of acute coronary syn-dromes. Two suitable randomized controlled trials were obtained but the studies were small and the methodological quality of one trial was moderate. Therefore we conclude that it is not possible to draw conclusions on the beneficial or harmful effect of influenza vaccination in the prevention of acute coronary syndromes. Implications for research. Future randomized trials are needed to quantify the beneficial or harmful effect of influenza vaccination to prevent acute coronary syndromes. Cost-effect analysis may point towards those for whom influenza vaccination is most applicable.

ACKNOwLEDgEMENTSWe would like to thank Mrs. Theresa Moore form the Cochrane Heart Group for her help and support and Mrs. Margeret Burke form the Cochrane Heart Group for her advice on the search strategy.

POTENTIAL CONFLICT OF INTEREST None


117

Reference List

1. Lopez AD, Mathers CD, Ezzati M et al. Global and regional burden of disease and risk factors, 2001: systematic analysis of population health data. Lancet 2006;367:1747-57. 2. Cannon CP, Braunwald E, McCabe CH et al. Antibiotic treatment of Chlamydia pneumoniae after acute coronary syndrome. N Engl J Med 2005;352:1646-54. 3. Grayston JT, Kronmal RA, Jackson LA et al. Azithromycin for the secondary prevention of coronary events. N Engl J Med 2005;352:1637-45. 4. Smeeth L, Thomas SL, Hall AJ et al. Risk of myocardial infarction and stroke after acute infection or vaccination. N Engl J Med 2004;351:2611-8. 5. Madjid M, Aboshady I, Awan I et al. Influenza and cardiovascular disease: is there a causal relationship? Tex Heart Inst J 2004;31:4-13. 6. Nichol KL, Nordin J, Mullooly J et al. Influenza vaccination and reduction in hospitalizations for cardiac disease and stroke among the elderly. N Engl J Med 2003;348:1322-32. 7. Naghavi M, Barlas Z, Siadaty S et al. Association of influenza vaccination and reduced risk of recurrent myocardial infarction. Circulation 2000;102:3039-45. 8. Siscovick DS, Raghunathan TE, Lin D et al. Influenza vaccination and the risk of primary cardiac arrest. J Epidemiol 2000;152:674-7. 9. Gurfinkel EP, Leon dlF, Mendiz O et al. Flu vaccination in acute coronary syndromes and planned percutane ous coronary interventions (FLUVACS) Study. Eur Heart J 2004;25:25-31. 10. Allsup S, Haycox A, Regan M et al. Is influenza vaccination cost effective for healthy people between ages 65 and 74 years? A randomised controlled trial. Vaccine 2004;23:639-45. 11. Gurfinkel EP, de la Fuente RL, Mendiz O et al. Influenza vaccine pilot study in acute coronary syndromes and planned percutaneous coronary interventions: the FLU Vaccination Acute Coronary Syndromes (FLU VACS) Study. Circulation 2002;105:2143-7. 12. Higgins JPT, Greens S. Cochrane Handbook for Systematic Reviews of Interventions (version 4.2.5). The Cochrane Library 2005;2. 13. Lavallee P, Perchaud V, Gautier-Bertrand M et al. Association between influenza vaccination and reduced risk of brain infarction. Stroke 2002;33:513-8. 14. Jackson LA, Yu O, Heckbert SR et al. Influenza vaccination is not associated with a reduction in the risk of recurrent coronary events. Am J Epidemiol 2002;156:634-40.

Chapter 8

118

119


Chapter

9Influenza InfeCtIon and the rIsk of aCute

pulmonary embolIsm

Matthijs van wissen • Tymen Keller • Brechje Ronkes • victor gerdes • Hans ZaaijerEric van gorp • Dees Brandjes • Marcel Levi • Harry Büller

Department of Vascular Medicine, Academic Medical Center, Amsterdam, The Netherlands. Department of Internal Medicine, Slotervaart Hospital, Amsterdam, The Netherlands.

Department of Microbiology, Academic Medical Center, Amsterdam.

Submitted

ABSTRACTInfluenza infections have been associated with procoagulant changes. Whether influenza infections lead to an increased risk of pulmonary embolism remains to be established. We conducted a nested case control study in a large cohort of patients with a clinical suspicion of having pulmonary embolism. Blood samples were collected to investigate the presence of influenza A and B by complement fixation assay (CFA). We compared case patients, in which pulmonary embolism was proven (n=102), to controls, in which pulmonary embolism was excluded (n=395). Furthermore, we compared symptoms of influenza-like illness in both patient groups 2 weeks prior to inclusion in the study, using a questionnaire, and calculated a well established influenza-like illness (ILI) score. We calculated the risk (odds ratio and 95% confidence interval) on pulmonary embolism associated with influenza infection. The percentage of patients with influenza A was higher in the control group compared to the case group (4.3% versus 1.0%, respectively) with an odds ratio of 0.22 (95% CI: 0.03-1.67). Influenza B was not detectable in any of the cases and was found in 3 of the 395 controls (0.8%). The ILI score was positive in the vast majority of the cases (92.2%)) and also in the control group (83.0%), OR 2.4 (95% CI: 1.11-5.17; P < 0.05). We did not observe an association between the ILI score and proven influenza detection. In this clinical study, influenza infection was not associated with an increased risk of acute pulmonary embolism. The ILI score is non-specific in this clinical setting.

Chapter 9

122

INTRODuCTIONDeep vein thrombosis and pulmonary embolism, collectively known as venous thrombo-embolism (VTE), have an annual incidence of approximately 2-3 per 1000 people.1 Many risk factors of venous thromboembolism have been well established, including genetic predisposition, immobilisation, surgery, pregnancy, oral contraceptives and malignan-cies. However, there are patients who develop venous thromboembolism in the absence of one of these risk factors. There is growing evidence that acute infections are associated with an increased risk of developing a venous thromboembolic event. Emmerich and others found an association between positive antibody titers for Chlamydia pneumoniae and venous thromboembolism.2-4 Also, patients with HIV are at increased risk of developing venous thromboembolism.5 In a population-based case-control study, subjects with elevated IL-8 levels had an increased risk of venous thrombosis, which indicates the role of inflammation (potentially as a consequence of infection) in the patho-genesis of venous thrombosis.6 Recently, it has been proposed that acute urinary tract infection and respiratory tract infection increase the risk of venous thromboembolism.7 The risk was highest during the 2 weeks after an urinary tract infection, with an incidence ratio for deep vein thrombosis of 2.10 (95% CI: 1.56-2.82), and that for PE of 2.11 (95% CI: 1.38-3.23), and gradually falling over the subsequent months. The incidence ratio for deep vein thrombosis in patients with a respiratory tract infection was 1.91 (95% CI: 1.49-2.44). Possible diagnostic misclassification of an early presentation of PE as a respiratory infec-tion precluded a reliable estimate of the risk of PE after respiratory tract infection.Since long it has been recognized that venous thromboembolism is usually the result of either a procoagulant state of the blood, or dysfunction of the vessel wall, eventually in combination with reduced flow of the blood. It has been shown in vitro that influenza affects each of these factors. Visseren et al. demonstrated that monocytes and endothelial cells, that were incubated with influenza, are able to activate coagulation.8-9 There are also indications that influenza leads to a procoagulant state of the blood in vivo. Recently, we have shown that influenza infections in elderly people result in increased levels of prohemostatic proteins (unpublished data).There are few data on the possible association between acute respiratory tract infections caused by influenza virus and the occurrence of venous thromboembolism. We hypothesized that influenza infections lead to a pro-coagulant state of the blood, resulting in a higher risk of pulmonary embolism. Therefore, we assessed the frequency of influenza infection among patients with a clinical suspicion of acute pulmonary embolism. Furthermore, we established whether the presence of signs and symptoms of an acute influenza-like illness (ILI) are spe-cific in the clinical setting of patients suspect for having acute pulmonary embolism.

METHODSStudy design

We performed a nested case control study in the cohort of the ANTELOPE study. The ANTELOPE study was performed from May 1999 to April 2001 in three hospitals in the Netherlands, as described in more detail previously.10 In this study, in- and outpatients, with clinically suspected acute pulmonary embolism, were prospectively investigated. The diagnosis of pulmonary embo-lism was confirmed by a high probability perfusion-ventilation (V/Q) scan. From this cohort we

Influenza infection and the risk of acute pulmonary embolism

123

selected case patients, in whom pulmonary embolism was proven and we compared them to controls from the same cohort, in whom pulmonary embolism was excluded.

Laboratory investigations

Blood samples were taken on the day of inclusion in the ANTELOPE study. Follow-up samples were taken 2-3 weeks later. Influenza detection in serum samples was performed by a complement fixation assay (CFA). A dilution of 1:64 or more was considered to be highly suggestive for having influenza A and/or B and regarded as a positive test result.

Influenza-like illness (ILI)

At baseline we recorded symptoms of a respiratory tract infection, including: acute onset of fever (temperature ≥ 37,8 °C), subjective feeling of fever or chills, cough with or without spu-tum, rhinitis, headache and sore throat. From these symptoms an influenza-like illness (ILI) score was calculated, based on the recommendations of previous studies on clinical signs and symptoms predicting influenza infection.11,12 The ILI score was considered positive when there was an acute onset of fever (> 37.8°C) plus one of the above mentioned symptoms.


Results are presented as means plus or minus standard deviation (SD) and in percent-ages. Odds ratios (OR) and corresponding 95% confidence intervals (95% CI), to indicate the association between influenza infection and pulmonary embolism, were calculated using logistic regression analysis. Odds ratios were adjusted for age, chronic obstructive pulmonary disease (COPD) and asthma.

RESuLTSIn this study a total of 497 patients were included, of which 102 patients had an acute pulmonary embolism and in 395 patients pulmonary embolism was excluded. Baseline characteristics of the pulmonary embolism group and non-pulmonary embolism group are shown in table 1.

TABLE 1: CHARACTERISTICS OF PATIENT gROuPSControls(n = 395)

Cases(n = 102)

Sex (male) 139 (35.2%) 46 (45.1%)

Age (SD) 53 (17.6) 58 (17.1)

COPD 42 (10.6%) 5 (4.9%)

Asthma 38 (9.6%) 5 (4.9%)

Atherosclerotic disease 47 (11.9%) 10 (9.8%)

Malignancy 54 (13.7%) 23 (22.5%)

Immobilisation 42 (10.6%) 16 (15.7%)

Recent surgery 56 (14.2%) 20 (19.0%)

Recent trauma 4 (1.0%) 2 (2.0%)

Oral anticonceptives 47 (11.9%) 12 (11.8%)

History of vTE 35 (8.9%) 19 (18.6%)

Family history of CvD 40 (10.1%) 15 (14.7%)

Smoking 74 (18.7%) 11 (10.8%)

SD = standard deviation; COPD = Chronic Obstructive Pulmonary Disease; VTE = venous thromboembolism; CVD = Cardiovascular Disease;

Chapter 9

124

Number of influenza infections

The percentage of patients with a positive result of the CFA for influenza A was higher in the non-pulmonary embolism group (20/395 (4.3%)) compared to the pulmonary embolism group (1/102 (1.0%)). The odds ratio was 0.22 (95% CI: 0.03-1.67 table, 2). Influenza B was not detectable in any of the pulmonary embolism patients and in 3 of the 395 non-pulmonary embolism patients (0.8%). We did not calculate the odds ratio for influenza B, because only few patients with influenza B were found (0 in the case group, 3 in the control group).

Symptoms of respiratory tract infection (ILI)

The ILI score was positive in the vast majority of the pulmonary embolism patients (94/102 (92.2%)) This was also the case in the non-pulmonary embolism group (328/395 (83.0%). The odds ratio for the ILI score was 2.4 (95% CI: 1.11-5.17, P < 0.05). We did not observe a clear association between the ILI score and proven influenza detection by CFA in either group.

DISCuSSIONIn this case control study we assessed the frequency of influenza in a cohort of patients with a clinical suspicion for pulmonary embolism and demonstrate that influenza A infection is rare among patients with a proven pulmonary embolism, while it does occur significantly more often in the group of patients without pulmonary embolism. Symptoms of influenza, detected by the influenza-like illness score, were frequently present in both patient groups. Therefore, we could not establish an association between influenza, which causes respiratory tract infections, and an increased risk of pulmonary embolism. Our results indicate that influenza is not an important preceding factor for pulmonary embolism and that symptoms of influenza are highly non-specific in this clinical context. It could also indicate that influenza is only one of the possible pathogens causing an acute respiratory infection and that other viral pathogens, which we did not study, may play a role. It is not surprising that influenza infections do occur more often in the patients without pulmonary embolism. Respiratory tract infections can cause a clinical picture that may be very similar to that of a pulmonary embolism. In fact, when pulmonary embolism is excluded in a patient who was suspected to have this disease, respiratory tract infection is one of the more common alternative diagnoses.13 This could be an explanation for our finding that influenza is more common in patients in whom pulmonary embolism was excluded.Some aspects of the study require comment. First, for influenza detection we only used serum samples for the detection of viruses. Although the use of the complement binding

TABLE 2: ODDS RATIO FOR PuLMONARy EMBOLISM Odds ratios (OR)

Crude OR (95% CI) Adjusted OR (95% CI)

Influenza A 0.22 (0.03-1.67) 0.22 (0.03-1.72)

Influenza-like illness 2.40 (1.11-5.17) * 2.78 (1.27-6.05) *

The odds ratio (OR) for pulmonary embolism risk are adjusted for age, COPD (chronic obstructive pulmonary disease) and asthma. * P < 0.05.


125

assay makes false positive tests highly unlikely, for a more thorough detection, influenza genome detection in throat swabs or even nasal washes is preferred in combination with serum detection.14 However, nasal washes are considerably more costly and invasive compared to throat swabs. A more precise reflection of the presence of influenza infec-tion in our patients would have been obtained when a throat swab had been obtained in all patients as well. However, the possible underestimation of the number of influenza infection will be present in both patient groups, and it is not likely that the main conclu-sion would have been different when throat swabs would have been performed. Secondly, the differences in patient characteristics between the two groups may have influenced the outcome. Since the effects of these differences are in opposite directions, we do not expect that the net effect has major influence on the outcome of this study. Mean age in the patients with pulmonary embolism is slightly higher than in those without (58 vs. 53 years, respectively) and age is related with susceptibility for infection. The prevalence of asthma and COPD, which are also associated with susceptibility for influenza, was lower among the patients with pulmonary embolism. We adjusted the crude odds ratios for the following confounding factors: age, COPD and asthma. The adjusted odds ratios did not differ from the crude odds ratios. Since we compared our patients with pulmonary embolism to patients with symptoms of chest pain and/or dyspnea, and not to a control group of randomly selected individuals, we still can not exclude the possibility that influ-enza leads to an increased risk of venous thromboembolism. However, the low rate of the very common influenza infection in the pulmonary embolism group, 1 out of 102 (1.0%), indicates that if this would be the case, this risk is probably only marginally increased. A large scaled study is needed to solve this issue. We observed that the number of positive ILI scores was higher in the pulmonary embo-lism group than in the non-pulmonary embolism group, while the number of influenza infections was lower in the pulmonary embolism group. It shows that the symptoms of patients with a pulmonary embolism are very similar to those of patients with acute respiratory infection. An alternative explanation could be that patients in the pulmonary embolism group had a virus infection other than influenza, causing signs and symp-toms resembling those of influenza. Our results demonstrate that an ILI questionnaire, and thus symptoms of respiratory tract infection, can not be used to differentiate acute respiratory infection from pulmonary embolism in this clinical setting and has therefore limited utility. In fact, the rate of a positive ILI score was even higher among patients with a pulmonary embolism. ILI scores similar to our score are used in population based studies on respiratory tract infection and vaccination studies. Our results question the use of this score in studies in the hospital setting. This ILI score has previously not been validated in other cohorts.In conclusion, this study suggests that influenza infection is not an important risk factor for pulmonary embolism. Symptoms of influenza are very non-specific and often present in patients with a pulmonary embolism and can therefore not be used to differentiate between acute respiratory infection and pulmonary embolism in this clinical setting.

Chapter 9

126

References 1. Silverstein MD, Heit JA, Mohr DN et al. Trends in the incidence of deep vein thrombosis and pulmonary embolism: a 25-year population-based study. Arch Intern Med. 1998 Mar 23;158(6):585-93. 2. Emmerich, J. Infection and venous thrombosis. Pathophysiol Haemost Thromb, 2002; 32: 346-348 3. de Saint Martin L, Pasquier E, Betsou F et al.Chlamydia pneumoniae IgG serological status and venous thromboembolism: a cross-sectional hospital based study. Presse Med. 2004 Dec 4;33(21):1493-6. 4. Lozinguez O, Arnaud E, Belec L et al Demonstration of an association between Chlamydia pneumoniae infection and venous thromboembolic disease. Thromb Haemost. 2000 Jun;83(6):887-91. 5. Klein SK, Slim EJ, de Kruif MD et al. Is chronic HIV infection associated with venous thrombotic disease? A systematic review. Neth J Med. 2005 Apr; 63(4):129-36. 6. van Aken BE, Reitsma PH, Rosendaal FR. Interleukin 8 and venous thrombosis: evidence for a role of inflammation in thrombosis. Br J Haematol. 2002 Jan;116(1):173-7. 7. Smeeth L, Cook C, Thomas S et al. Risk of deep vein thrombosis and pulmonary embolism after acute infection in a community setting. Lancet. 2006 Apr 1; 367(9516): 1075-9. 8. Bouwman JJ, Visseren FL, Bosch MC et al. Procoagulant and inflammatory response of virus-infected monocytes. Eur J Clin Invest. 2002 Oct;32(10):759-66. 9. Visseren FL, Bouwman JJ, Bouter KP et al. Procoagulant activity of endothelial cells after infection with respiratory viruses. Thromb Haemost. 2000 Aug;84(2):319-24. 10. Ten Wolde M, Hagen PJ, Macgillavry MR et al. Advances in New Technologies Evaluating the Localization of Pulmonary Embolism Study Group. Non-invasive diagnostic work-up of patients with clinically suspected pulmonary embolism; results of a management study. J Thromb Haemost. 2004 Jul;2(7):1110-7. 11. Monto AS, Gravenstein S, Elliott M et al. Clinical signs and symptoms predicting influenza infection. Arch Intern Med. 2000 Nov 27;160(21):3243-7. 12. Boivin G, Hardy I, Tellier G et al. Predicting influenza infections during epidemics with use of a clinical case definition. Clin Infect Dis. 2000 Nov;31(5):1166-9. 13. Richman PB, Courtney DM, Friese J et al. Prevalence and significance of nonthromboembolic findings on chest computed tomography angiography performed to rule out pulmonary embolism: a multicenter study of 1,025 emergency department patients. Acad Emerg Med. 2004 Jun;11(6):642-7. 14. Covalciuc KA, Webb KH, Carlson CA. Comparison of four clinical specimen types for detection of influenza A and B viruses by optical immunoassay (FLU OIA test) and cell culture methods. J Clin Microbiol. 1999 Dec;37(12):3971-4.


127

Chapter

10seleCtIve expansIon of Influenza a vIrus

speCIfIC t Cells In unstable human atherosClerotIC plaques

Tymen Keller* • jelger van der Meer • Peter Teeling • Koen van der Sluijs • Mirza Iduguus Rimmelzaan • Marcel Levi • Allard van der wal • Onno de Boer

Departments of Vascular Medicine, and Pathology, and Experimental Internal medicine and Vascular Surgery, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands Department of Virology,

Erasmus Medical Center, Rotterdam, The Netherlands.

* TK and JvdM equally contributed to this study

Submitted

ABSTRACTEvidence is accumulating that infection with influenza A virus contributes to athero-thrombotic disease. Vaccination against influenza decreases the risk for atherosclerotic syndromes, indicating that inflammatory mechanisms may be involved. We hypoth-esized that influenza A virus specific T cells contribute to atherosclerotic plaque inflam-mation which mediates the onset of plaque rupture. T cell cultures were generated from atherosclerotic segments and peripheral blood of 30 patients with symptomatic carotid artery disease. The response of these cells to influenza A virus was analyzed. Selective outgrowth of intraplaque T cells was investigated by comparing the responses of plaque and peripheral T cells, after which the patients were categorized as high-, intermedi-ate- and no-responders. The presence of influenza A virus in the vesselfragments was evaluated by rtPCR. High responses of plaque derived T cells to influenza A virus were frequently observed. Five patients were categorized as high responders, 10 were interme-diate responders and 15 were no-responders. Influenza A virus could not be detected in the atherosclerotic plaques by PCR. Selective outgrowth of influenza A virus specific T cells occurs within human atherosclerotic plaques. Influenza virus derived antigens, or alternatively, mimicry antigens appear potential candidates for triggering or sustaining plaque inflammation which eventually lead to symptomatic plaque complications.

Chapter 10

130

INTRODuCTIONAtherosclerosis is a disease of the vesselwall with many features of a chronic inflam-matory disease. Atherosclerotic plaques contain variable amounts of macrophages, T cells and mast cells, and particularly lesions prone to develop complications (vulnerable plaques) contain large amounts of such cells.1 It is now widely accepted that T cells play an important role in atherosclerotic disease.2 Animal studies have shown that atheroscle-rotic lesions in T cell deficient apoE-/- mice are significantly smaller compared to immu-nocompetent littermates. After reconstitution with T cells the lesion size in these animals increases, showing that T cells contribute to atherosclerotic plaque growth, at least in this model.3 Studies from our group on human coronary atherectomy specimens have shown that T cells are important during the onset of acute coronary syndromes, since the numbers of recently activated T cells are significantly increased in patients with unstable angina and acute myocardial infarction.4, 5 At present, several antigens have been identi-fied that could be responsible for the activation of T cells in atherosclerotic lesions. These antigens include extracellular matrix components and lipids, which are indeed the major constituents of plaques, but also antigens from microbial origin, including C. pneu-moniae, periodontal pathogen P. gingivalis and Epstein Barr Virus.2, 6-9

Among the various diseases that have been linked with the onset of symptomatic athero-sclerotic disease, respiratory tract infections have attracted much attention.10 Many studies have shown that influenza increases the risk of acute coronary syndromes and stroke.11-13 Moreover, vaccination against influenza contributes to a decreased risk of myocardial infarction and stroke in the subsequent influenza season.14-16 In experimental studies a link between influenza A virus infection and atherosclerotic plaque inflammation has been proposed based on experiments showing that infection of apoE deficient mice with influenza virus resulted in an acute arterial wall inflammatory response,17 composed predominantly of macrophages and T lymphocytes, suggesting that influenza virus may trigger directly plaque inflammation. Still, the pathophysiological mechanisms that underlie the relationship with influenza virus infection and atherosclerotic syndromes are unknown. In particular, it is presently not known whether the micro environment of a human atherosclerotic plaque is indeed suitable for influenza virus driven T cell medi-ated inflammation. In the present study we investigated whether a local T cell response to influenza A virus could contribute to atherosclerotic plaque inflammation. To investigate this, we tested T cells isolated from culprit lesions of patients with symptomatic athero-sclerotic disease for their influenza virus specific response in vitro. Results were compared with the rate of influenza A virus specific responsiveness of peripheral blood derived T cells of the same patients. Furthermore, the occurrence of influenza A infection in the study patients, and the presence of influenza A virus in these lesions were investigated

METHODSPatient material

Human carotid endarterectomy specimens were collected at surgery from 30 consecu-tive patients with symptomatic carotid artery disease. A part of the plaque tissue was fixed in RNAlater (Ambion, Huntingdon, UK). The remaining tissue was used for the

Selective expansion of influenza A virus specific T cells in unstable human atherosclerotic plaques

131

isolation of T-lymphocytes. From the same patients heparinized blood was also col-lected. The study was approved by the local ethical committee and all patients gave informed consent.

Antigen presenting cells

Peripheral blood mononuclear cells (PBMC) were isolated from the freshly drawn blood of each patient by gradient centrifugation with Ficoll-paque (Pharmacia Biotech, Uppsala, Sweden). The obtained cells, to be used as antigen presenting cells (APC), were frozen and stored in liquid nitrogen until further use.

Culture of plaque and peripheral blood derived T cells

Primary T cell cultures were generated as described previously.9,18 Peripheral blood T lymphocytes from the same patients were cultured under the same conditions as plaque derived T cells, in order to generate control T cells cultures.

Influenza virus

Sucrose-gradient purified influenza A virus (H3N2) Resvir-9, a reassortant between A/Puerto Rico/8/34 (H1N1) and A/Nanchang/933/95 (H3N2), containing the hemag-glutinin (HA), neuraminidase (NA), and nucleoprotein (NP) of A/Nanchang/933/95, was used for the proliferation assays. Sucrose-gradient purified influenza B virus (B/Harbin/7/94) was used as control. The infectious virus titres were determined in cell culture using Madin-Darby canine kidney (MDCK) cells as indicator cells.19

T cell proliferation assays

T cell proliferation assays were performed as previously described.18 The MHC restric-tion of the response was tested by performing blocking experiments with monoclonal antibodies against HLA-DR and HLA-ABC (clone L243 and W6/32 respectively, both from the ATCC, Manass, VA, USA). Additional controls were performed using a differ-ent related pathogen (influenza B), and paraformaldehyde fixed influenza A virus. For the latter, influenza A virus was treated with paraformaldehyde (PFA (final concentra-tion 1%) for 20 minutes), followed by glycine treatment (final concentration 0.2 M) and dialysis against PBS.The stimulation index (SI) was calculated as the mean cpm of cultures in the presence of virus divided by the mean cpm of parallel cultures without virus. Differences between plaque and peripheral T cells were also determined by calculating the ratio between plaque and peripheral SI’s (SI-ratio). High responders were defined as those patients in which the SI- ratio was at least 5. An intermediate group of medium responders, in which the SI- ratio was between 2 and 5, and finally, no-responders, in which the SI-ratio was smaller than 2.

Detection of influenza A virus in atherosclerotic plaque tissue by real time rtPCR

RNA was isolated from frozen tissue using the Trizol method, cDNA synthesis was performed using a random hexamer cDNA synthesis kit (Applera, Foster City, CA) and

Chapter 10

132

amplified in a quantitative real-time PCR reaction20 (ABI PRISM 7700 Sequence Detector System). The following primers were used: 5’-GGACTGCAGCTGAGACGCT (sense); 5’-ATCCTGTTGTATATGAGGCCCAT (anti-sense) and 5’-CTCAGTTATTCTGCTGGTGCACTTGCC (5’-FAM labeled probe), recognizing a shared sequence in matrix proteins M1 and M2. Serial dilutions of a positive control (strain A/PR/8/34, 70.000 viral RNA cop-ies, EM counted) were also included. The detection limit with this protocol was 50 copies of viral genome. The integrity of the cDNA preparations was checked by conventional PCR with primers specific for human beta-actin: 5`-ACCCAACACTGTGCCCATCTA (sense) 5’-AGAAGCATTGCGGTGGACGA (anti-sense).

Serological Analysis

Serum samples, collected at time of surgery from 21 patients were available, and screened for evidence of a recent influenza A infection. Serum samples were screened for the pres-ence of antibodies against influenza A virus using the complement fixation (CF) assay, routinely performed at the department of Clinical Virology at our institute.


Differences between experimental conditions of the T cells were analyzed by ANOVA with post hoc Bonferroni test. P < 0.05 was considered statistically significant. Statistical analysis was performed using SPSS vs 12, (SPSS inc. Chicago, Illinois, USA).

FIguRE 1: STIMuLATION INDEX OF T CELLS

Stimulation index of plaque- (n) and peripheral blood derived T cells (®) from all 30 patients, sorted by SI-ratio. Five patients were classified as high responders, 10 as median responders and 15 were classified as non responders. The mean stimulation indices (± SEM) are presented. Significantly increased (*) or decreased (#) compared to the response of peripheral blood derived T cells (P<0.05).


133

RESuLTSResponsiveness of atherosclerotic plaque and peripheral blood derived T cells to influenza A virus

Figure 1 shows the T cell proliferation response to influenza A virus of plaque and peripheral blood-derived T cell cultures from all patients, sorted by SI-ratio. Five patients could be classified as high responders (SI-ratio ≥ 5), and 10 patients were classified as intermediate responders (2 ≤ SI-ratio < 5). In all these patients the response of plaque derived T cells was significantly increased compared to the peripheral blood T cells. Fifteen patients were categorized as non-responders (SI-ratio < 2). In three of these patients the response of plaque derived T cells was even significantly lower compared to peripheral blood T cells. The specificity of the influenza A virus-induced T cell response was further analyzed with different control experiments. To test if the influenza A virus induced prolifera-tion was MHC restricted, blocking experiments with antibodies against HLA-ABC and HLA-DR were performed. In figure 2 the results obtained for the high responders are illustrated. Strongest inhibition was observed with antibodies against HLA-DR, except in patient no.27 where the levels of inhibition with HLA-DR and HLA-ABC were similar. Similar results were observed for the intermediate responders, in all cases the response could be inhibited with antibodies against HLA-ABC and DR, with the strongest effect with anti-HLA-DR.

FIguRE 2: BLOCKINg STuDIES wITH ANTIBODIES

Blocking studies with anti HLA-DR and HLA-ABC antibodies. the results of the five high responders are illustrated. In all patients, the responses could be inhibited with anti HLA-DR and/or HLA-ABC, indicating that the observed responses were partly MHC class II and/or class I mediated, respectively. * P < 0.05; ** P < 0.005, as compared to cells incubated without the antibodies.

Chapter 10

134

We also tested whether antigen processing and presentation by the APC of virus proteins is required for eliciting a T cell response. Therefore, the responses of plaque derived T cells from 3 patients with a high SI of plaque derived T cells were tested with formalin fixed influenza A virus and compared with the responses of live virus. We did not observe any differences in terms of proliferation between live and inactivated virus (figure 2), indicative for a specific, MHC class II mediated immune response. The specificity of the response to influenza A virus was further verified by incubating the cells with a control virus, influenza-B. The cells that showed a high proliferative response to influenza A virus did not respond to influenza B virus (figure 3), showing that the response was indeed specific for influenza A

Detection of influenza A virus in atherosclerotic tissue

The presence of influenza A virus in all the atherosclerotic tissue samples was investigated using a sensitive realtime rtPCR method. The results of the four representative samples (patient no. 6, 13, 14 and 18, all high responders) are illustrated in figure 4, together with serial dilutions of the positive control. In none of the samples viral RNA could be detected.

Serology

To investigate whether the patients in our study group had a recent exposure to Influenza A virus, a complement fixation assay was performed. Serum samples, obtained at time of surgery from 21 patients were available. None of the patients showed signs of a recent infection with influenza A virus.

FIguRE 3: RESPONSE OF T CELLS TO INFLuENZA A, B AND FIXED INFLuENZA vIRuS

Response of plaque derived T cells to influenza A virus, formalin-fixed influenza A virus and influenza B vi-rus. T cells obtained from three patients were stimulated with influenza A virus, fixed influenza A virus and influenza B virus. No differences between influenza A virus and fixed influenza A virus, were observed. T cells did not respond to influenza B virus, illustrating the specificity of the response.


135

DISCuSSION In the present study we show that atherosclerotic plaque derived T lymphocytes from patients with symptomatic atherosclerotic disease frequently show high, specific respon-siveness to influenza A virus. Furthermore, we observed that these plaque derived T lymphocytes frequently show a much higher response, compared to T cells from the peripheral blood of the same patient. Five out of 30 patients were high responders with an SI-ratio of at least 5. Ten patients were intermediate responders with an SI-ratio between 2 and 5, and fifteen out of 30 patients were categorized as non responders. It appears that activation and expansion of influenza A virus specific T cells had occurred within the lesions of a subgroup of patients. However, we could not detect the presence influenza A virus in any of the symptomatic atherosclerotic lesions. The identification of inflammatory triggers in atherosclerotic lesions is important, because inflammation is the underlying cause of atherosclerotic plaque destabilization, and contributes to the onset of atherothrombotic events.21

Our results are especially interesting in the light of a recent publication by Naghavi et al, who showed that infection of apoE-/- mice with influenza A virus resulted in aggravated atherosclerotic plaque inflammation.17 This effect was limited to atherosclerotic lesions, no effects were observed on the normal vessel wall, and moreover, these authors could not demonstrate the presence of the virus in the affected lesions. Basically, we observed a similar situation: the atherosclerotic lesions contained influenza A virus specific T cells,

FIguRE 4: REAL TIME PCR

Real time PCR to investigate the presence influenza A virus RNA in atherosclerotic tissue. The results of four high responders (6, 13, 14 and 18) and the standard curve (70.000, 7000, 700 and 70 copies of viral RNA) of influenza A virus are illustrated. On the x-axis the cycle number during the PCR run is depicted, while on the Y-axis the fluorescence intensity, a measure for the amount of specific PCR product is illustrated. Similar results were observed for the other plaques, in none of the lesions viral RNA could be detected.

Chapter 10

136

but the virus itself could not be detected by rtPCR in the lesions, at least at the time of our analysis. These authors suggested that antigenic cross reactivity between influenza virus and plaque components could contribute to plaque inflammation.22 Apart from epitope mimicry between influenza virus and plaque components, it cannot be ruled out that influenza A virus had been present in the plaques, but disappeared, for example through eradication by the elicited inflammatory response. It is not clear whether influenza A virus is capable of infecting atherosclerotic plaques, since it is generally considered only to infect cells in the pulmonary tree, although virus genes have been detected in the peripheral blood of infected patients.23

We do not yet know which (cross reactive) antigens contributed to the activation of this subset of influenza A virus specific T cells. Theoretically, all influenza A virus derived proteins may provide antigens recognized by T cells. As we only observed high T cell responses in a subgroup of patients, it is tempting to speculate that these patients share a predisposition for the observed response of plaque T cells to influenza virus. Certain HLA-alleles might contribute more to the T cell response to influenza virus than others. Characterization of these patients for HLA-genotype may provide insight which antigens are responsible for the responses that we observed. One could argue that the presence of increased numbers of activated, antigen specific T lymphocytes in unstable atherosclerotic lesions is not the result of local, intraplaque proliferation but due to the recruitment of activated T lymphocytes from the peripheral blood. This seems unlikely, as proliferating T cells have been detected in atherosclerotic lesions.24 Furthermore, recently it was shown that unstable plaques from patients with acute coronary syndromes contain clonotypic expansions of specific T cell subsets, indic-ative for an antigen driven, intraplaque T cell mediated immune response.25

The isolation and subsequent in vitro culture of T cells from plaques or peripheral blood could have led to the selection of certain T cell subsets. To minimize this selection bias, T cells from the peripheral blood were first cultured under identical conditions as plaque derived T cells. Moreover, we did not observe any differences in terms of influenza A virus specific responses between peripheral blood derived T cells that were either freshly isolated or were subjected to our in vitro culture protocol (data not shown), which sug-gests that selection of T cells responsive to this pathogen did not occur in vitro.In this study, we showed that the atherosclerotic plaque derived T cells were indeed spe-cific for influenza A virus, and not activated through antigen independent mechanisms. First, we demonstrated that the plaque derived T cells showed no response to a related virus, influenza B. In addition, influenza A virus, inactivated with formaldehyde, was equally potent in stimulating T cells as live virus, suggesting that the stimulation was not fully dependent on protein synthesis and the endogenous route of antigen processing and presentation. This may indicate that at least CD4+ T cells may be involved. Indeed, the virus specific response of the polyclonal T cell populations could be inhibited with antibodies directed to MHC class II antigens. This is in concordance with the observation that activated T cells in atherosclerotic lesions are always CD4+.26 However, also an inhib-itory effect was observed with antibodies directed to MHC class I antigens. Therefore also CD8+ T cells may have contributed to the virus specific T cell responses.


137

In this study we used polyclonal T cell cultures containing both CD4+ and CD8+ T cells, so we are not informed about the functional properties of the antigen specific T cells at the clonal level. Properties like cytokine profiles, the antigen specificity and MHC restric-tion should be investigated in more detail, but for these studies it will be necessary to generate T cell clones specific for influenza A virus. These clones could also be used to investigate if epitope mimicry between viral proteins and plaque components plays a role in T cell mediated atherosclerotic plaque inflammation.Influenza virus is certainly not the only pathogen that triggers atherosclerotic plaque inflammation, and other antigenic stimuli are also likely involved. In vitro, responses of plaque derived T cells to lipids, heat shock proteins and other infectious pathogens have been reported.2 But, given the widespread occurrence of influenza within the total popu-lation, the contribution of influenza A virus induced plaque inflammation to the onset of acute cerebro- and vascular disease, may be substantial. In conclusion, the findings of this study suggest that influenza A virus may act as an important contributor to inflammation in plaques. Furthermore, this study suggests that mimicry between antigens from influenza virus and antigens present in atherosclerotic plaque could responsible for the outgrowth of these T cells in the lesions. Future research should focus on the influenza virus mimicking antigens in atherosclerotic plaques responsible for T cell activation.

Chapter 10

138

References 1. Ross R. Atherosclerosis--an inflammatory disease. N Engl J Med. 1999;340:115-126. 2. de Boer OJ, Becker AE, van der Wal AC. T lymphocytes in atherogenesis - functional aspects and antigenic repertoire. Cardiovasc Res. 2003;60:78-86. 3. Zhou X, Nicoletti A, Elhage R et al. Transfer of CD4(+) T cells aggravates atherosclerosis in immunodeficient apolipoprotein E knockout mice. Circulation. 2000;102:2919-2922. 4. van der Wal AC, Piek JJ, de Boer OJ et al. Recent Activation of the Plaque Immune Response in Coronary Lesions Underlying Acute Coronary Syndromes. Heart. 1998;80:14-18. 5. Hosono M, de Boer OJ, van der Wal AC et al. Increased expression of T cell activation markers (CD25, CD26, CD40L and CD69) in atherectomy specimens of patients with unstable angina and acute myocardial infarction. Atherosclerosis. 2003;168:73-80. 6. Stemme S, Faber B, Holm J, Wiklund O, Witztum JL, Hansson GK. T lymphocytes from human atherosclerotic plaques recognize oxidized low density lipoprotein. Proc Natl Acad Sci USA. 1995;92:3893-3897. 7. Mosorin M, Surcel HM, Laurila A et al. Detection of Chlamydia pneumoniae-reactive T lymphocytes in hu man atherosclerotic plaques of carotid artery. Arterioscler Thromb Vasc Biol 2000;20:1061-1067. 8. Choi J, Chung SW, Kim SJ. Establishment of Porphyromonas gingivalis-specific T-cell lines from atheroscl rosis patients. Oral Microbiol Immunol. 2001;16:316-318. 9. de Boer OJ, Teeling P, Idu MM et al. Epstein Barr virus specific T-cells generated from unstable human atherosclerotic lesions: Implications for plaque inflammation. Atherosclerosis. 2006;184:322-329. 10. Collins S. Excess mortality from causes other than influenza and pneumonia during influenza epidemics. Public Health Rep. 1932;47:2159-2180. 11. Pesonen E, Siitonen O. Acute myocardial infarction precipitated by infectious disease. Am Heart J. 1981;101:512-513. 12. Spodick DH, Flessas AP, Johnson MM. Association of acute respiratory symptoms with onset of acute myo cardial infarction: prospective investigation of 150 consecutive patients and matched control patients. Am J Cardiol. 1984;53:481-482. 13. Smeeth L, Thomas SL, Hall AJ et al. Risk of myocardial infarction and stroke after acute infection or vaccination. N Engl J Med. 2004;351:2611-2618. 14. Lavallee P, Perchaud V, Gautier-Bertrand M et al. Association between influenza vaccination and reduced risk of brain infarction. Stroke. 2002;33:513-518. 15. Naghavi M, Barlas Z, Siadaty S et al. Association of influenza vaccination and reduced risk of recurrent myocardial infarction. Circulation. 2000;102:3039-3045. 16. Gurfinkel EP, de la Fuente RL. Two-year follow-up of the FLU Vaccination Acute Coronary Syndromes (FLU VACS) Registry. Tex Heart Inst J. 2004;31:28-32. 17. Naghavi M, Wyde P, Litovsky S et al. Influenza Infection Exerts Prominent Inflammatory and Thrombotic Effects on the Atherosclerotic Plaques of Apolipoprotein E-Deficient Mice. Circulation. 2003;107:762-768. 18. de Boer OJ, van der Wal AC, Houtkamp MA et al. Unstable atherosclerotic plaques contain T-cells that respond to chlamydia pneumoniae. Cardiovasc Res. 2000;48:402-408. 19. Rimmelzwaan GF, Baars M, Claas EC et al. Comparison of RNA hybridization, hemagglutination assay, titration of infectious virus and immunofluorescence as methods for monitoring influenza virus replication in vitro. J Virol Methods. 1998;74:57-66. 20. van Elden LJ, Nijhuis M, Schipper P et al. Simultaneous detection of influenza viruses A and B using real- time quantitative PCR. J Clin Microbiol. 2001;39:196-200. 21. van der Wal AC, Becker AE. Atherosclerotic plaque rupture--pathologic basis of plaque stability and instability. Cardiovasc Res. 1999;41:334-344. 22. Madjid M, Aboshady I, Awan I et al. Influenza and cardiovascular disease: is there a causal relationship? Tex Heart Inst J. 2004;31:4-13. 23. Tsuruoka H, Xu H, Kuroda K et al Detection of influenza virus RNA in peripheral blood mononuclear cells of influenza patients. Jpn J Med Sci Biol. 1997;50:27-34. 24. Rekhter MD, Gordon D. Active proliferation of different cell types, including lymphocytes, in human atherosclerotic plaques. Am J Pathol. 1995;147:668-677. 25. De Palma R, Del Galdo F, Abbate G et al. Patients with acute coronary syndrome show oligoclonal T- cell recruitment within unstable plaque: evidence for a local, intracoronary immunologic mechanism. Circulation. 2006;113:640-646. 26. Buchner K, Henn V, Grafe M et al. CD40 ligand is selectively expressed on CD4+ T cells and platelets: implications for CD40-CD40L signalling in atherosclerosis. J Pathol. 2003;201:288-295.


139

part

III(aCute respIratory traCt) InfeCtIons

and CoagulatIon

Chapter

11aCute respIratory traCt InfeCtIons In

elderly patIents InCrease systemIC levels of hemostatIC proteIns

Tymen Keller • Matthijs van wissen • Albert Mairuhu • Thijs van Hees • Diederik wijffels victor gerdes • Philip Sutorius • joost Meijers • gerard van Doornum

Eric van gorp • Marcel Levi • Harry Büller • Dees Brandjes

Department of Vascular Medicine, Academic Medical Center, Amsterdam, The NetherlandsDepartment of Internal Medicine, Slotervaart Hospital, Amsterdam, The Netherlands

General Practitioners Office Overveen, Overveen, The NetherlandsDepartment of Virology, Erasmus Medical Center, University of Rotterdam, Rotterdam, The Netherlands

J Thrombosis Haemost (accepted in part)

ABSTRACT Acute respiratory tract infections increase the risk of acute ischemic heart disease and venous thrombosis. The aim of this study was to determine the effect of acute respira-tory tract infections on hemostatic proteins. We studied 255 elderly subjects during the winter. Blood samples were collected of all subjects before the start of the winter and dur-ing 54 episodes of flu-like disease. Different causative pathogens were found, including influenza A virus, influenza B virus, para influenza virus, respiratory syncytial virus and corona virus. In subjects with flu-like disease and those with proven viral infections von Willebrand factor and plasmin-a2-antiplasmin levels were increased at onset of the symp-toms. In the convalescent phase, two weeks after start of symptoms, von Willebrand factor and plasmin-a2-antiplasmin levels had returned to baseline. No difference was seen in hemostatic proteins between subjects with proven influenza, proven other viral infections and subjects with symptoms only. Common acute respiratory tract infections increase hemostatic proteins. The vascular point of impact seems to be the endothelium. This di-rects towards a possible link between acute respiratory tract infections, endothelium cell perturbation and the increased risk of acute ischemic heart disease.

Chapter 11

144

INTRODuCTIONAcute respiratory tract infections are associated with a temporarily increased risk of acute ischemic heart disease and venous thrombosis. In elderly subjects Acute respiratory tract infections causes an 5-fold increased risk of myocardial infarction in the first three days after start of symptoms (risk ratio 4.95; 95% CI: 4.43 - 5.53)1 and a 2-fold increase risk of venous thrombosis (2.10; 95% CI: 1.56–2.82).2 The precise underlying mechanism of this association is unclear, but hypothetically systemic inflammation upon infection may contribute to this.3 Transient change in local hemodynamic factors, coagulation activation and endothelial cell perturbation are potential culprits.4

Endothelial cell perturbation is associated with an increased risk of future isch-emic heart disease (IHD).5 Von Willebrand factor, a marker of endothelial cell per-turbation, mediates platelet adhesion and plays a role in thrombus formation. In-dividuals with plasma von Willebrand factor levels in the highest quartile, showed a 3-fold increase risk of IHD compared with those in the lowest quartile.6 Beside this long-term association, high von Willebrand factor levels are also related to a short-term increased risk of plaque rupture and subsequent thrombus formation.7 Other hemostatic markers – such as prothrombin fragment F1+2, plasmin-a2-an-tiplasmin complex and plasminogen activator inhibitor type (PAI)-1 – are also risk factors for IHD.9-11 Several cross-sectional reports showed increased levels of hemostatic proteins dur-ing symptoms of acute respiratory tract infection,12-16 however, these findings are dif-ficult to interpret since the diagnosis was not unequivocally established. Woodhouse et al demonstrated that an increase of fibrinogen and FVII was related to symptoms of acute respiratory tract infection.16 Furthermore, fibrinogen and FVII levels were related to neutrophil count and CRP, which suggests that an acute phase response was responsible for this change in hemostatic proteins. Whether viruses infecting the respiratory tract indeed perturb endothelial cells and thereby increase the risk of acute ischemic heart disease remains to be proven. The purpose of our investigation was to determine the effect of naturally occurring acute respiratory tract infection on hemostatic proteins in a prospective cohort study.

METHODS

Study population

Through postal invitations and subsequent phone calls, we asked men and women from a general practice in a small village (Overveen, the Netherlands) to participate in the study. Eligible patients consisted of those at risk for influenza, as defined by the crite-ria to participate in the national influenza vaccination campaign. The most important criteria are: individuals 65 years of age and older, or those with pulmonary or cardiac diseases, diabetes, renal insufficiency or immuno- compromised. We included subjects in October, and performed a follow-up during the winter. The Institutional Review Board approved the study and all subjects provided informed consent.

Acute Respiratory Tract Infections in Elderly Patients Increase Systemic Levels of Hemostatic Proteins

145

Study protocol

Of all subjects a detailed medical history including influenza vaccination status was obtained at entry and physical examination was performed. At the time of inclusion a blood sample was collected, immediately where after the follow-up period started. Trained trial nurses contacted all subjects by telephone twice during the winter. We instructed the subjects to immediately contact the investigators in case of flu-like symptoms.

Flu-like disease

Flu-like disease was defined based on the recommendations for the prediction of influenza in elder-ly patients.17 The acute onset of fever (body temperature > 37.8°C) plus at least 2 of the following flu-like symptoms was considered as flu-like disease: headache, myalgia, sore throat, and cough.

Specimen collection

Blood samples and nose swabs were obtained when the predefined criteria of flu-like symptoms were met. In the convalescence phase, 14 days after the beginning of symp-toms the specimen collection was repeated. Blood was collected in empty vacuum tubes and in tubes containing 0.106 M trisodium citrate. The tubes were centrifuged for 15 minutes at 2000g, plasma was pooled and then centrifuged again at 2000g for 5 min-utes. Subsequently, plasma was frozen and stored in aliquots of 0.5 ml at minus 80 °C.

Analysis of hemostatic proteins

Prothrombin fragment 1+2 (F1+2) and plasmin-a2-antiplasmin complexes (PAP) were measured by Enzyme-Linked Immuno Sorbent Assay (ELISA)-s from Dade Behring, Mar-burg, Germany, and plasminogen activator inhibitor type I (PAI-1) antigen was assayed with an ELISA from Hyphen BioMed, Andrésy, France. Von Willebrand factor antigen was also measured by immunoassay employing antibodies from Dako, Glostrup, Denmark. Results of von Willebrand factor are presented as percentages of normal pool plasma.

Serology

Presence of antibodies against respiratory pathogens was established using specific ELISA-s. For the detection of IgA antibodies in-house assays were used, for the detection of IgG or IgM antibodies tests purchased from Serion/Viron (Würzburg, Germany) were used. Samples were tested for IgA and IgG antibodies against influenza A virus, influen-za B virus, para influenza virus type 1 and type 2, respiratory syncytial virus, adenovirus, and specific IgM and IgG against Mycoplasma pneumoniae. We considered respiratory tract proven when there was a 4-fold elevation between the sample taken on the first day of flu-like symptoms and the sample taken in the convalescent phase two weeks later. virus isolation

Respiratory viruses were isolated by centrifuge-enhanced culture on human laryngeal carcinoma (Hep-2) cells, human embryonic lung fibroblasts (HEL) cells, tertiary monkey kidney (TMK) cells (Cynomolgus macaques), and human lung adenocarcinoma (A549) cells. Samples were tested for RSV, influenza viruses type A and B, Para influenza

Chapter 11

146

viruses 1, 2, 3, and 4, adenovirus, enterovirus, rhinovirus, and human metapneumovirus (HMPV) by routine diagnostic immunofluorescence (IF) assays 48 hr after inoculation. Cultures on HEp-2, HEL, and A-549 cells were observed for 14 days followed by an IF assay when cytopathic effect (CPE) was noted during this period of time. Cultures on tMK cells (using trypsin containing medium) were observed for development of CPE for 14 days followed by an IF assay for HMPV using guinea pig antisera.

Nucleic acid extraction and real time amplification

Total RNA and DNA was isolated with the total nucleic acid isolation kit (Roche Applied Science, the Netherlands), according to the recommendations of the manufacturer. Amplification of human metapneumovirus, human coronavirus NL63 and influenza virus was performed as described previously.18-21 The amplification of RSV type A and B, rhinovirus, coronavirus C43 and 229E is described elsewhere (manuscript in prepa-ration), using TaqMan EZ RT-PCR reagents (Applied Biosystems, Nieuwerkerk a/d IJssel, the Netherlands). Detection was performed on an ABI7700 detection system. All materials were spiked with an internal control, consisting of a fixed amount of Phocine Distemper Virus (PDV) to monitor for inhibition and/or loss of material.


Data are presented as mean ± standard deviation (SD), percentages and 95% confidence interval (CI), or median ± interquartile range (IQR). Change from baseline was calculated by subtracting values measured in the acute and convalescent phase from values mea-sured at baseline. Variables were compared using the Wilcoxon rank sum test. P values of < 0.05 were considered statistically significant. TABLE 1: PATIENTS CHARACTERISTICS AT BASELINE

Characteristics a flu-like disease (n = 53) Subjecs without disease (n=202)

Age, years b 67 ± 12 67 ± 11

Sex (male) 42 % (30-55) 46 % (39-53)

Influenza vaccination 74 % (60-84) 67 % (60-73)

Medical history

Myocardial infarction 8 % (3-18) 5 % (2-8)

Current angina pectoris 4 % (1-13) 2 % (1-5)

Cerebrovascular disease 2 % (0-1) 2 % (1-5)

PAD 4 % (1-13) 0 % (0-2)

Medication use

Aspirin 13 % (7-25) 4 % (2-7)

OAC 9 % (4-20) 2 % (1-5)

CvD risk factors

Smoking, current 25 % (15-38) 13 % (9-18)

Hypercholesterolemia 26 % (16-40) 9 % (6-14)

Hypertension 28 % (20-42) 16 % (12-22)

Diabetes 8 % (3-18) 3 % (1-6)

BMI > 30 17 % (9-29) 7 % (4-11)

Family history of CVD 26 % (16-40) 9 % (6-14)

a Data are presented as percentage and 95% confidential interval.b Age in years (mean ± standard deviation). PAD = Peripheral arterial disease, OAC = oral anticoagulation, CVD= cardiovascular disease, BMI = body mass index.


147

RESuLTS We included 255 different subjects in a cohort which was followed during the winter. The mean age of the total cohort was 67 years (±14) and 68% (95% CI: 62-74) of the subjects had received vaccination against influenza before inclusion in the study. A total of 53 subjects (21%, 95% CI: 16-26) reported flu-like symptoms which met our predefined criteria of flu-like disease. Because two subjects reported flu-like disease twice during follow-up, we counted 55 episodes of flu-like disease. Table 1 shows the characteristics of the subjects with and without flu-like disease.

Subjects with flu-like disease showed increases in von Willebrand factor antigen levels (Table 2). Median von Willebrand factor antigen levels changed from 151% (IQR: 114-189) at baseline to 197% (IQR: 162-247) at the start of the symptoms (P < 0.05). Thrombin generation was not increased during the infection, as evidenced by comparable levels of prothrombin fragment F1+2 at baseline and during the acute phase of disease. In subjects with flu-like disease fibrinolysis was activated, as demonstrated by an increase in me-dian plasmin-a2-antiplasmin levels from 343 μg/l (IQ range: 255-535) to 465 μg/l (range: 283-638, P < 0.05). Plasminogen-activator inhibitor-1 (PAI-1) levels did not changed. In the convalescent phase, two weeks after start of symptoms, von Willebrand factor and plasmin-a2-antiplasmin levels returned to baseline. In subjects with proven viral respira-tory tract infection, a similar activation pattern was seen. Different virusses were found by microbiological analysis. Causative pathogens were found in 26% (95% CI: 16-39) of

TABLE 2: HEMOSTATIC PROTEINS IN PATIENTS wITH RESPIRATORy TRACT INFECTIONS

All subjects Proven viral infection Proven influenza

N a 54 b 14 9

vwF (%) c

Baseline 151 (114-189) 175 (126-233) 148 (125-217)

Acute 197 (162-247) d 220 (189-284) d 208 (188-280) d

Conv. 147 (125-185) 157 (137-185) 149 (126-165)

F1+2 (nmol/l) c

Baseline 0.90 (0.66-1.11) 0.77 (0.29-0.90) 0.77 (0.32-1.09)

Acute 0.92 (0.65-1.29) 0.88 (0.47-0.94) 0.90 (0.50-0.96)

Conv. 1.01 (0.80-1.19) 0.99 (0.81-1,34) 0.91 (0.40-1.13)

PAI-1 (µg/l) c

Baseline 1.7 (0.6-3.4) 2.2 (0.6-2.6) 1.2 (0.5-2.6)

Acute 1.6 (0.7-3.4) 2.2 (0.6-3.4) 2.8 (1.0-3.4)

Conv. 1.7 (0.6-3.4) 0.8 (0.5-2.3) 1.7 (0.8-2.3)

PAP (U/ml) c

Baseline 343 (255-535) 345 (260-612) 344 (255-727)

Acute 465 (283-638) d 583 (342-719) d 583 (412-679)

Conv. 314 (237-559) 549 (314-708) 548 (343-743)

a Number of subjects. b Blood was collected in only 54 episodes due to logistic reasons. c All data are presented as median ± interquartile range. Levels of hemostatic proteins were measured ast baseline, in the acute and convales-cent phase of disease. d Significantly different (P < 0.05) from baseline as calculated by Wilcoxon rank sum test. Conv. = convalescent, VWF = von Willebrand factor, F1+2 = prothrombin fragment 1+2, PAI-1 = plasminogen activator inhibitor-1, PAP = plasmin-a2-antiplasmin.

Chapter 11

148

the cases. In patients with proven viral respiratory tract infection von Willebrand factor antigen levels showed an increase with 68% (IQR: 17-106) at the start of the symptoms (P < 0.05). Median plasmin-a2-antiplasmin levels increased with 136 μg/l (IQR: -18 – 287). However, this increase in plasmin-a2-antiplasmin did not reached statistical significance. In the convalescent phase, two weeks after start of symptoms, von Willebrand factor levels returned to baseline. We compared the levels of hemostatic proteins between subjects with an influenza infec-tion, those with proven other viral infections and those with clinical symptoms only. Figure 1 shows the laboratory findings as change from baseline in these subgroups of subjects. No statistical differences were observed in von Willebrand factor, F1+2, plas-min-a2-antiplasmin and PAI levels between those with proven influenza en other viral respiratory tract infections.

DISCuSSION

FIguRE 1: HEMOSTATIC PROTEINS

Comparison between subgroup of subjects with proven influenza (O), other viral infections (▲) or those with clinical symptoms only (▼). Pannels show (A) von Willebrand factor (VWF), (B) prothrombin frag-ments 1+2 (F1+2), (C) plasminogen activator inhibitor (PAI) -1 and (D) plasmin-a2-antiplasmin in the acute and convalescent phase of infection. Levels are presented as change from baseline with means and confidence intervals. No difference was seen between subgroup of subjects in hemostatic proteins.

General introduction and outline of the thesis

149

A B

C D

▼

▼

▼ ▼

▼

▼▼

▼

▲

▲ ▲ ▲

▲

▲ ▲ ▲

This study demonstrates that acute respiratory tract infections result in an increase of the hemostatic proteins von Willebrand factor and plasmin-a2-antiplasmin. Vascular point of impact seems to be the endothelium, as evidenced by an increase in von Willebrand factor during the acute phase of the disease. Because these proteins are known to be as-sociated with an increased risk of ischemic heart disease5, 6, 11 we hypothesize that patients at risk of ischemic heart disease could benefit from protection against acute respiratory tract infections. Our study is the first to prospectively show that hemostatic proteins are increased in patients with proven respiratory tract infections. Although others have reported similar changes in patients with flu-like symptoms, the cross-sectional nature of these studies prohibited a definitive conclusion as to whether respiratory tract infection were in fact the cause of these changes.12-16 In previous studies a seasonal patterns in factor VII, fibrinogen and platelets was shown.14, 15 Others have demonstrated that patients with flu-like symp-toms have increased fibrinogen and decreased free protein S.12, 13, 16 These studies were limited, however, because only self-reported clinical symptoms were used to establish the diagnosis of the infection and, in contrast to our study, the authors did not confirm the infection with objective analysis. This may have confounded the results since also non-infectious inflammatory diseases may present with fever in combination with cough. Unexpectedly, elevated levels of plasmin-a2-antiplasmin, indicative for increased fibrino-lysis, were not accompanied by an increase in PAI-1, the main inhibitor of fibrinolysis. There may be several explanations for this finding. First, whether chemokines involved in host-defence response to respiratory tract infection increase the expression of PAI-1 is unknown. PAI-1 is an acute-phase reactant whose release is influenced by cytokines, such as IL-1 and IL-6.22 But predominantly during viral infections the increase of IL-6 is accom-panied by an increase of interferon γ (INF-γ). 23 Although it is likely that INF-γ modulates the expression of PAI-1, in vitro results as well as in vivo results are contradictory.24-27 Second, the expression of PAI-1 by different micro-organisms and by different tissues is not uniform. In vitro studies investigating the effect of different causative micro-organ-isms of respiratory tract infection on PAI-1 levels have had variable results. Infection of vascular endothelial cells by C. pneumoniae has been shown to lead to the over expression of PAI-1,28 whereas infection of umbilical vein endothelial cells by herpes simplex virus (HSV) leads to a decrease in PAI-1.29 The influence of different pathogens on PAI-1 levels in vivo therefore remains unclear. Our study may indicate that if there is an effect of viral respiratory pathogens on PAI-1 levels, this increase is not likely to be major. The prospective nature of our study renders the effect of seasonal variation on mark-ers of coagulation and fibrinolysis quite unlikely. Although hemostatic proteins have a minimal seasonal variation, the fact that levels of hemostatic proteins returned to normal in the convalescent phase two weeks after the initiation of the disease episode, further negates the option that the observed increase is the result of this variation. Also, since we had baseline measurements before the winter season we were able to exclude that the increase in levels of hemostatic proteins was the result of pre-existent conditions. Our conclusions are limited by the high percentage of influenza-vaccinated subjects included in the study and the relatively small sample size. The high prevalence of

Chapter 11

150

influenza vaccination certainly dampened the effects of the infection because the main antigenicity of circulating influenza strains was covered by the vaccine. We can not rule out that full-blown influenza activates coagulation in some cases. Furthermore, the small number of patients limited significant assessment of more subtle influences on the mea-sured parameters.From the results obtained, we conclude that naturally occurring respiratory tract infec-tions in elderly human subjects result in endothelial cell perturbation, which is potential-ly associated with dysfunction. We hypothesize that the induced hemostatic changes may form a link between acute respiratory tract infections and acute atherothrombotic disease. The precise increment in risk still needs to be established in a prospective study.

ACKNOwLEDgMENTS We gratefully acknowledge the participants of the study and the staff of the General Practitioners Office Overveen. We are indebted to Ed Hull for reading and editing the manuscript and to Olga Ternede, Viola Weeda and Mechteld Vermeulen for assistance in patients care.


151

Reference List 1. Smeeth L, Thomas SL, Hall AJ et al. Risk of myocardial infarction and stroke after acute infection or vaccination. N Engl J Med 2004;351:2611-8. 2. Smeeth L, Cook C, Thomas S et al. Risk of deep vein thrombosis and pulmonary embolism after acute infection in a community setting. Lancet 2006;367:1075-9. 3. Libby P. Inflammation in atherosclerosis. Nature 2002;420:868-74. 4. Stone PH. Triggering myocardial infarction. N Engl J Med 2004;351:1716-8. 5. Yarnell J, McCrum E, Rumley A et al. Association of European population levels of thrombotic and inflammatory factors with risk of coronary heart disease: the MONICA Optional Haemostasis Study. Eur Heart J 2005;26:332-42. 6. Morange PE, Simon C, Alessi MC et al. Endothelial cell markers and the risk of coronary heart disease: the Prospective Epidemiological Study of Myocardial Infarction (PRIME) study. Circulation 2004;109:1343-8. 7. Lee KW, Lip GY, Tayebjee M et al. Circulating endothelial cells, von Willebrand factor, interleukin-6, and prognosis in patients with acute coronary syndromes. Blood 2005;105:526-32. 8. Koster T, Blann AD, Briet E et al. Role of clotting factor VIII in effect of von Willebrand factor on occurrence of deep-vein thrombosis. Lancet 1995;345:152-5. 9. Miller GJ, Bauer KA, Barzegar S et al. Increased activation of the haemostatic system in men at high risk of fatal coronary heart disease. Thromb Haemost 1996;75:767-71. 10. Kohler HP, Grant PJ. Plasminogen-activator inhibitor type 1 and coronary artery disease. N Engl J Med 2000;342:1792-801. 11. Cushman M, Lemaitre RN, Kuller LH et al. Fibrinolytic activation markers predict myocardial infarction in the elderly. The Cardiovascular Health Study. Arterioscler Thromb Vasc Biol 1999;19:493-8. 12. Horan JT, Francis CW, Falsey AR et al. Prothrombotic changes in hemostatic parameters and C-reactive protein in the elderly with winter acute respiratory tract infections. Thromb Haemost 2001;85:245-9. 13. Kaba NK, Francis CW, Hall WJ et al. Protein S declines during winter respiratory infections. J Thromb Hae most 2003;1:729-34. 14. Stout RW, Crawford V. Seasonal variations in fibrinogen concentrations among elderly people. Lancet 1991;338:9-13. 15. Crawford VL, McNerlan SE, Stout RW. Seasonal changes in platelets, fibrinogen and factor VII in elderly people. Age Ageing 2003;32:661-5. 16. Woodhouse PR, Khaw KT, Plummer M et al. Seasonal variations of plasma fibrinogen and factor VII activity in the elderly: winter infections and death from cardiovascular disease. Lancet 1994;343:435-9. 17. Govaert TM, Dinant GJ, Aretz K et al. The predictive value of influenza symptomatology in elderly people. Fam Pract 1998;15:16-22. 18. Fouchier RA, Hartwig NG, Bestebroer TM et al. A previously undescribed coronavirus associated with respiratory disease in humans. Proc Natl Acad Sci U S A 2004;101:6212-6. 19. Maertzdorf J, Wang CK, Brown JB et al. Real-time reverse transcriptase PCR assay for detection of human metapneumoviruses from all known genetic lineages. J Clin Microbiol 2004;42:981-6. 20. Ward CL, Dempsey MH, Ring CJ et al. Design and performance testing of quantitative real time PCR assays for influenza A and B viral load measurement. J Clin Virol 2004;29:179-88. 21. IJpma FF, Beekhuis D, Cotton MF et al. Human metapneumovirus infection in hospital referred South African children. J Med Virol 2004;73:486-93. 22. Levi M, Keller TT, van Gorp E et al. Infection and inflammation and the coagulation system. Cardiovasc Res 2003;60:26-39. 23. Hayden FG, Fritz R, Lobo MC et al. Local and systemic cytokine responses during experimental human influenza A virus infection. Relation to symptom formation and host defense. J Clin Invest 1998;101:643-9. 24. Kasza A, Kowanetz M, Poslednik K et al. Epidermal growth factor and pro-inflammatory cytokines regulate the expression of components of plasminogen activation system in U373-MG astrocytoma cells. Cytokine 2001;16:187-90. 25. Siren V, Immonen I, Cantell K et al. Alpha- and gamma-interferon inhibit plasminogen activator inhibitor-1 gene expression in human retinal pigment epithelial cells. Ophthalmic Res 1994;26:1-7. 26. de Metz J, Hack CE, Romijn JA et al. Interferon-gamma in healthy subjects: selective modulation of inflammatory mediators. Eur J Clin Invest 2001;31:536-43. 27. Gluszko P, Undas A, Amenta S et al. Administration of gamma interferon in human subjects decreases plasminogen activation and fibrinolysis without influencing C1 inhibitor. J Lab Clin Med 1994;123:232-40. 28. Dechend R, Maass M, Gieffers J et al. Chlamydia pneumoniae infection of vascular smooth muscle and endothelial cells activates NF-kappaB and induces tissue factor and PAI-1 expression: a potential link to accelerated arteriosclerosis. Circulation 1999;100:1369-73. 29. Bok RA, Jacob HS, Balla J et al. Herpes simplex virus decreases endothelial cell plasminogen activator inhibitor. Thromb Haemost 1993;69:253-8.

Chapter 11

152


153

Chapter

12CytomegalovIrus InfeCtIon and the

prothrombotIC state In humans

Tymen Keller • Matthijs van wissen • Rose Hoogeveen • victor gerdes • joost Meijers Ineke ten Berge • Marcel Levi • Frederieke Bemelman • Harry Büller

Department of Vascular Medicine and Nephrology and Internal Medicine, Academic Medical Center, Amsterdam, The NetherlandsDepartment of Internal Medicine, Slotervaart Hospital, Amsterdam, The Netherlands

Pilot study

ABSTRACTCytomegalovirus (CMV) has prothrombotic properties in vitro. However, whether CMV in-fection in humans result in a prothrombotic state is unknown. We conducted a prospective study in a cohort of renal transplant recipients. Patients were matched according to time after transplantation procedure and levels of hemostatic proteins and activation markers were compared between patients with and without active CMV. Furthermore, activation of coagulation was correlated with CMV load. Our data showed that patients with CMV had both activated coagulation and fibrinolysis, as reflected by significantly increased levels of prothrombin fragment 1+2, plasmin-a2-antiplasmin complexes, D-dimer and von Wil-lebrand Factor. No change in plasminogen activator inhibitor type-1 concentration, acti-vated protein C and global clotting assays, PT and aPTT was observed. In addition, levels of some of the hemostatic proteins were associated with CMV load. A significant association was observed between CMV load and levels of plasmin-a2-antiplasmin and D-dimer. We hypothesize that the coagulation activation and endothelial cell perturbation (as evidenced by increased levels of von Willebrand factor) caused by CMV may cause a prothrombotic state. Whether these changes are responsible for the observed association between CMV infection and vascular disease remains subject of further study.

Chapter 12

156

BACKgROuND Cytomegalovirus (CMV) is a common pathogen that infects people of all ages. CMV is a double-stranded enveloped DNA virus, member of the herpes family1 which may on entry destruct the infected cell or remain in a stage of latency for years. Infection with CMV is often asymptomatic but occcasionally, the infection may present with flu-like symptoms or as amononucleosis-like syndrome with or without symptoms and signs related to infected organs.2, 3 In immunecompromised hosts the infection is associated with serious morbidity and even mortality. CMV infections have been implicated in the pathogenesis of vascular disease.4 In the last decade experimental studies have linked CMV with the development of atherosclerosis. However, clinical studies showed contradictory results.4-5 Furthermore, CMV infections may be involved in venous thrombosis,4, 5 and CMV vas-culitis is a well documented disease.6 Although the pathofysiological mechanism of the association between CMV and vascular pathology is not clear, it has been suggested that endothelial cell perturbation and coagulation activation play an important role.7

Evidence of CMV-induced endothelial cell perturbation and coagulation activation mainly comes from in vitro experiments.4 Previous studies showed that the enhanced coagu-lability of CMV could be mediated by assembly of the prothrombinase complex on the CMV surface.8, 9 This observation suggests that the CMV surface contains the necessary pro-coagulant phospholipids for thrombin generation. Visseren et al. demonstrated that CMV infected endothelial cells increase tissue (TF) expression. They showed that the subsequent prothrombotic state is dependent on activation of coagulation factor VII.10 In addition, next to direct coagulation activation, also endothelial cell perturbation has been described.11 In addition, endothelial cells release von Willebrand factor upon infection with CMV12 and it was shown that CMV infections are associated with increased plasma levels of von willebrand factor in humans.13 However, it is not known whether CMV infection activates coagulation in vivo and which pathways are involved. Eventually, such knowledge might be useful for a better understanding of the mechanisms involved in the vascular effects of CMV and may potentially lead to better preventive or treatment strate-gies in patients infected with this virus. The purpose of this study was to determine whether CMV infection increases the pro-thrombotic state in vivo. In a prospective study in renal transplant recipients, we compared sequential markers of coagulation activation, fibrinolysis and endothelial cell perturbation in patients with CMV infection to those without active CMV infection. In addition, we determined the association between levels of hemostatic proteins and CMV load.

METHODSWe conducted a prospective study in a cohort of renal transplant recipients up to 6 months after renal transplantation. Before entry in the study, trial nurses registered patient characteristics, cardiovascular risk factors, the presence of cardiovascular or thrombotic disease, and current use of anticoagulant medication. All renal transplant recipients received standard therapy according to clinical guidelines, including immuno-suppressive therapy: CD25 monoclonal antibody, 20 mg on day 0 and day 4; predniso-lone (10 mg/day), cyclosporine (through level monitoring), and mycophenolate mofetil (2

Cytomegalovirus infection and the prothrombotic state in humans

157

g/day). Both renal transplant recipients with CMV infection in the past and those without (considered when CMV IgG was negative) were included in the study. Monitoring of CMV infection was performed following standard procedures. Trained personnel took blood samples every month and determined CMV specific antibodies and the number of CMV copies in blood. When CMV-DNA was detectable, samples were collected weekly. Quantitative CMV PCR was performed in EDTA whole-blood samples as described.14 For study purposes, we considered CMV screening positive when the result of the qCMV PCR was greater than 1000 CMV copies per mL, for both first CMV infections and CMV reactivation. Antiviral therapy was initiated when clinical symptoms and/or laboratory abnormalities such as pancytopenia or elevated liver enzymes occurred. We obtained written informed consent from all patients. The study was approved by the Medical Ethical Committee of the Academic Medical Center.

Hemostatic proteins

A total of 9 ml citrated blood was drawn, samples were centrifuged twice at 2000 x g for 20 minutes, and frozen at -80 ºC until assays were performed. Thrombin generation was assessed by measuring prothrombin fragment 1+2 (F1+2) by an enzyme-linked immu-nosorbent assay (ELISA) from Dade Behring (Marburg, Germany). Fibrinolysis was as-sessed by measuring plasminogen activator inhibitor type-I (PAI-1) (ELISA from Hyphen BioMed, Andrésy, France) and plasmin-a2-antiplasmin (PAP) complexes (ELISA from Dade Behring, Marburg, Germany). Degradation of fibrin was determined by measuring D-dimer levels (ELISA from Roche Diagnostics, Mannheim, Germany). Von Willebrand factor antigen (VWF) was measured by immunoassay employing antibodies from Dako, Glostrup, Denmark. Activated partial thromboplastin time (aPTT) and prothrombin time (PT) were determined according to standard methods. Resistance to activated protein C (APC) was determined as previously described.15 In brief, the endogenous thrombin po-tential (ETP)16 was determined in the presence and absence of activated protein C, divided by the same ratio of normal pooled plasma. Resistance to activated protein C increases the result of the ETP-based test.


Because CMV load and data on hemostasis had a skewed distribution we used medians and corresponding interquartile ranges and performed non-parametric statistical tests. We tested for statistical significant differences in baseline characteristics between patients with CMV compared to patients without CMV using the Wilcoxon rank sum test for numerical data and the Chi-squared test for categorical data. The Kruskal-Wallis test was used to test for time dependent differences of hemostatic data. We compared levels of hemostatic proteins and coagulation tests of patients with CMV infection to patients without CMV infection and tested for significancy using Mann-Whitney U test. Samples taken in the first 30 days after renal transplantation were excluded to minimise the effect of the surgical procedure on the outcome of interest. Finally, to determine the association between hemostatic proteins and CMV load we calculated the Spearman’s rank correlation coefficients and the corresponding P – values. P < 0.05 was considered statistically significant.

Chapter 12

158

RESuLTSA total of 70 patients were included in the study (median age 54 years, IQR 39 - 60), 67% of them were male. In 21 patients qCMV revealed CMV above the treshold of 1000 cop-ies/mL, while 49 patients did not had increased CMV load during follow-up. The baseline characteristics of patients with and those without CMV infection during follow-up are presented in table 1. A total of 31 patients (74%; 95% CI: 59 - 85) were CMV IgG positive in the group with CMV (re)activation during follow-up and 18 patients (64%; 95% CI: 46 - 79) were CMV IgG positive in those without CMV (re) activation. Although there were more males in the CMV infected group, compared to the group without CMV infection, this difference was not statistical significant (P = 0.7). Overall, no differences were seen between the groups in baseline characteristics, atherosclerotic risk factors, previous his-tory of cardiovascular disease, previous venous thromboembolism or medication use.

TABEL 1: BASELINE CHARACTERISTICS BEFORE INCLuSION IN THE STuDy OF PATIENTS wITH CMv AND wITHOuT CMv

No CMv CMv

n 28 42

Age, years (IQR) 53 (39 - 61) 54 (39 - 60)

Sex, male 18 (64%) 29 (69%)

CMv Igg positive 18 (64%) 31 (74%)

Cardiovascular disease

Myocardial infarction 2 (7%) 1 (2%)

Angina pectoris 4 (14%) 4 (10%)

Stroke 1 (4%) 2 (5%)

Peripheral arterial disease 2 (7%) 1 (2%)

venous thrombotic disease

Deep vein thrombosis 2 (7%) 2 (5%)

Pulmonary embolism 1 (4%) 0

Medication

Platelet inhibitors 3 (11%) 6 (14%)

Vitamin K antagonist 3 (11%) 1 (2%)

Risk factors cardiovascular disease

Diabetes mellitus 2 (7%) 6 (14%)

Hypertension 17 (61%) 28 (67%)

Dyslipidemia 4 (14%) 3 (7%)

Smoking 4 (14%) 3 (7%)

Patient characteristics (number and percentages), including previous history of cardiovascular disease and previous history venous thrombotic disease. A history of venous thrombotic disease was considered when a deep vein thrombosis or pulmonary embolism was proven before inclusion in the study.

TABEL 2: HEMOSTASIS AFTER RENAL TRANSPLANTATIONPeriod 1 2 3 P *

F1+2 (nmol/L) 1.6 (1.2 - 2.0) 1.3 (1.0-1.7) 1.2 (0.9-1.5) P = 0.02

PAP (µg/L) 728 (516 - 919) 489 (389-749) 479 (367-604) P < 0.01

PAI-1 (µg/L) 58 (31 - 71) 60 (41-77) 72 (48-82) P = 0.08

D-dimer (g/L) 1.0 (0.7 -1.4) 0.8 (0.3-1.1) 0.6 (0.3-1.2) P = 0.02

vwF (%) 305 (229 - 384) 224 (200-281) 186 (136-230) P < 0.01

APC ratio 1.7 (0.8-2.8) 1.2 (0.8-1.6) 1.0 (0.2-1.7) P = 0.16

PT (seconds) 11.8 (12.2 - 13.3) 12.2 (11.5-12.8) 11.9 (11.7-13.0) P = 0.18

aPTT (seconds) 25.3 (23.1 -27.4) 26.1 (23.6-28.0) 26.9 (24.0-27.7) P = 0.10

Medians with interquartile ranges (IQR) of prothrombin fragment 1+2 (F1+2), plasmin-a2-antiplasmin (PAP), plasminogen activator inhibitor type-1 (PAI-1), Von Willebrand factor (VWF), activated protein C (APC), prothrombin time (PT), activated partial thrombin time (aPTT); Period after renal transplantation (1 = first month, 2 = second and third month, 3 = fourth, fifth, and sixth month); * indicates P value for comparison between the periods (Kruskal-Wallis test).


159

Hemostatic changes during 6 months follow-up

To determine hemostatic changes after renal transplantation we collected blood samples in the 6 consecutive months after the procedure and stratified these samples to time after renal transplantation. Table 2 presents the levels of hemostatic proteins and coagulation tests one month after renal transplantation, in the second and third month and thereafter. Overall, high levels of hemostatic proteins were seen during the follow-up period. Von Willebrand factor levels were two to three times above normal ranges from pooled control plasma. During follow-up a significant decrease was observed in markers of thrombin generation and fibrinolytic activity. Specifically, levels of prothrombin fragment 1+2 (F1+2), plasmin generation (PAP-complex), D-dimer and Von Willebrand factor. Other hemostatic param-eters, including plasminogen activator inhibitor type-1, activated protein C, PT and the aPTT did not change significantly over time.

Chapter 12

160

FIguRE 1: HEMOSTASIS IN PATIENTS wITH CMv

Levels of hemostatic proteins and coagulation tests were measured two and three month after renal transplantation. Levels of prothrombin fragment 1+2 (F1+2) (A), plasmin-a2-antiplasmin (PAP) (C), plasminogen activator inhibitor type-1 (PAI-1) (B), D-dimer (D), von Willebrand factor (VWF) (E), activated protein C (APC) ratio (F) and PT (G) and aPTT (H) during CMV infection (n) compared to no CMV infection (®).Data are presented as means with standard error. * P < 0.05 compared to patients without CMV (Mann-Whitney U test).

A B

C D

E F

G H

Influence of CMv infection on hemostasis

The mean values of hemostatic proteins and coagulaton tests in the second and third month after transplantation are shown in Figure 1 for patients with CMV and patients without CMV. In patients with CMV higher F1+2, plasmin-a2-antiplasmin, D-dimer and von Willebrand factor levels were observed compared to patients without CMV infection. The level of F1+2 was 1.3 nmol/L (IQR: 1.2-1.7) in patients with CMV and 1.0 nmol/L (IQR: 0.9-1.4) in patients without CMV (P < 0.05). Median plasmin-a2-antiplasmin level was 639 μg/L (IQR: 433-780) in patients with CMV and 421 μg/L (IQR: 375-580) in pa-tients without CMV (p<0.05). Also, D-dimer level was increased in patients with CMV compared to patients without CMV (1.0 g/L; IQR: 0.6-1.3 vs 0.5 g/L; IQR: 0.3-0.9; P < 0.05). Finally, statistical significant difference was seen in von Willebrand factor. In CMV infected patients von Willebrand factor level was 256 % (IQR: 205-410) compared to 201 % (IQR: 164-235) in patients without CMV (P < 0.05). Between the patients with CMV and patients without CMV, no differences were seen in plasminogen activator inhibitor type-1, activated protein C, PT and aPTT.


161

FIguRE 2: ASSOCIATION BETwEEN CMv LOAD AND LEvELS OF HEMOSTATIC PROTEINS

Levels of hemostatic proteins von Willebrand factor (VWF) (A), prothrombin activation marker 1+2 (F1+2) (B), plasmin-a2-antiplasmin (PAP) (C), and D-dimer (D) plotted against levels of CMV. Both plasmin-a2-antiplasmin and D-dimer were significantly correlated to CMV load. * P < 0.05 (Spearman’s rank sum test).

CMv load is associated with levels of hemostatic proteins

We established whether high levels of CMV were associated with increased levels of he-mostatic proteins. Therefore, levels of hemostatic proteins as a function of CMV viral load are plotted in figure 2. Our data show that plasmin-a2-antiplasmin and D-dimer signifi-cantly correlated with CMV load. High CMV load was associated with high plasmin-a2-antiplasmin levels (R = 0.38, P < 0.05) and D-dimer (R = 0.35, P < 0.05). Also F1+2 and von Willebrand factor increased with higher CMV load, however this was not statistical significant. Levels of other hemostatic proteins and the global coagulation tests did not correlate with CMV load.

DISCuSSIONThe present study provides the first evidence that CMV infection is associated with a pro-thrombotic state in vivo. This was demonstrated by increased plasma levels of markers of coagulation, firbinolysis and endothelial cell perturbation in renal transplant recipients with CMV infection compared to those without. In addition we showed a positive correla-tion between CMV load and hemostatic proteins. The procoagulant effect of CMV has previously been suggestedbased on experimental studies.9 Our study results confirm these findings and show that CMV can affect hemo-stasis in humans in vivo. Our results suggest that CMV-infected patients have both acti-vation of coagulation and fibrinolysis. Theincreased D-dimer levels could be the result of either fibrin formation or increased endogenous fibrinolysis. However, the increased prothrombin fragment 1+2 levels suggest that at least partially the increased D-dimer levels were the result of coagulation activation, whereas the CMV-induced elevated levels of plasmin-α2-antiplasmin suggests an activated fibrinolytic system, presumably as a response to the activation of coagulation. The mechanism for the activation of coagulation and fibrinolysis remains to be eluci-dated. However, there is ample evidence that CMV affects endothelial cells, which may explain our findings.17 Perturbed endothelial cells are capable of enhanced tissue factor expression, which may lead to thrombin generation. Simultaneously, endothelial cells may release plasminogen activators, resulting in plasmin generation. Interestingly, the activation of fibrinolysis was not counterbalanced by an increase in the main inhibitor of plasminogen activation, i.e. plasminogen activator inhibitor type-1. The notion of endo-thelial cell perturbation is confirmed by our finding of strongly increased plasmalevels of von Willebrand factor, which may also be a prothrombotic factor by itself due to the associatyed increase in coagulation factor VIII. The risk of ischemic heart disease has also been associated with increased levels of hemostatic proteins.18 19 Individuals with high plasma von Willebrand factor levels have a 3-fold increase risk of ischemic heart disease compared with those with low levels.20 Our findings on increased levels of von Willebrand factor during CMV infection are in accordance with a previous study that showed a direct correlation between plasmalevels of von Willebrand factor and the pres-ence of a CMV infection in 39 patients with a recent renal transplantation.13 A number of issues have to be taken into account when interpreting the results of the present study. First, blood samples were collected at a non-uniform time of the day.

Chapter 12

162

Therefore, diurnal variation could have affected hemostatic proteins. However, this would introduce an increased random measurement error, which is likely to lead to an underestimation of any relationship, and therefore does not negate our findings. Second, the patients studied were all receiving immunomodulating therapy. Because inflammation and coagulation have a bi-directional relation,21 this could have decreased the hemostatic response of CMV. Of note, however, all patients in the analysis were treated similarly and no difference in medication use was detected between cases and controls. Lastly, because of the small sample size we could have missed subtle changes in hemostatic parameters and therefore underestimated the effects of CMV infection. The sample size of this study also prohibits to establish a link between changes in hemostatic proteins with clinical outcome. We conclude that CMV infection are associated with a procoagulant state in vivo. In addition, CMV is associated with endothelial cell perturba-tion. These data provide a possible pathophysiological link between CMV and vascular disorders. Whether these changes increase the risk of vascular disease should be subject of further investigation.

ACKNOwLEDgEMENTWe thank Ester Remmerswaal, Ineke Mollema and Sylvia ter Meulen for their dedicated help in patient care and sample handling.


163

Reference List 1. Cohen JI, Corey GR. Cytomegalovirus infection in the normal host. Medicine (Baltimore) 1985;64:100-14. 2. Landolfo S, Gariglio M, Gribaudo G et al. The human cytomegalovirus. Pharmacol Ther 2003;98:269-97. 3. Rowshani AT, Bemelman FJ, van Leeuwen EM et al. Clinical and immunologic aspects of cytomegalovirus infection in solid organ transplant recipients. Transplantation 2005;79:381-6. 4. Squizzato A, Gerdes VE, Buller HR. Effects of human cytomegalovirus infection on the coagulation system. Thromb Haemost 2005;93:403-10. 5. Youd P, Main J, Jackson E. Cytomegalovirus infection and thrombosis: a causative association? J Infect 2003;46:141-2. 6. Golden MP, Hammer SM, Wanke CA et al. Cytomegalovirus vasculitis. Case reports and review of the literature. Medicine (Baltimore) 1994;73:246-55. 7. Keller TT, Mairuhu AT, de Kruif MD et al. Infections and endothelial cells. Cardiovasc Res 2003;60:40-8. 8. Pryzdial EL, Wright JF. Prothrombinase assembly on an enveloped virus: evidence that the cytomegalovirus surface contains procoagulant phospholipid. Blood 1994;84:3749-57. 9. Sutherland MR, Raynor CM, Leenknegt H et al. Coagulation initiated on herpesviruses. Proc Natl Acad Sci U S A 1997;94:13510-4. 10. Visseren FL, Bouwman JJ, Bouter KP et al. Procoagulant activity of endothelial cells after infection with respiratory viruses. Thromb Haemost 2000;84:319-24. 11. Dam-Mieras MC, Muller AD, van Hinsbergh VW et al. The procoagulant response of cytomegalovirus infected endothelial cells. Thromb Haemost 1992;68:364-70. 12. Bruggeman CA, Debie WH, Muller AD et al. Cytomegalovirus alters the von Willebrand factor content in human endothelial cells. Thromb Haemost 1988;59:264-8. 13. Kas-Deelen AM, Harmsen MC, De Maar EF et al. Acute rejection before cytomegalovirus infection enhances von Willebrand factor and soluble VCAM-1 in blood. Kidney Int 2000;58:2533-42. 14. Rentenaar RJ, Gamadia LE, van DN et al. Development of virus-specific CD4(+) T cells during primary cytomegalovirus infection. J Clin Invest 2000;105:541-8. 15. Nicolaes GA, Thomassen MC, Tans G et al. Effect of activated protein C on thrombin generation and on the thrombin potential in plasma of normal and APC-resistant individuals. Blood Coagul Fibrinolysis 1997;8:28-38. 16. Hemker HC, Willems GM, Beguin S. A computer assisted method to obtain the prothrombin activation velocity in whole plasma independent of thrombin decay processes. Thromb Haemost 1986;56:9-17. 17. Levi M. CMV endothelitis as a factor in the pathogenesis of atherosclerosis. Cardiovasc Res 2001;50:432-3. 18. Feinbloom D, Bauer KA. Assessment of hemostatic risk factors in predicting arterial thrombotic events. Arterioscler Thromb Vasc Biol 2005;25:2043-53. 19. Yarnell J, McCrum E, Rumley A et al. Association of European population levels of thrombotic and inflammatory factors with risk of coronary heart disease: the MONICA Optional Haemostasis Study. Eur Heart J 2005;26:332-42. 20. Morange PE, Simon C, Alessi MC et al. Endothelial cell markers and the risk of coronary heart disease: the Prospective Epidemiological Study of Myocardial Infarction (PRIME) study. Circulation 2004;109:1343-8. 21. Levi M, van der Poll T, Buller HR. Bidirectional relation between inflammation and coagulation. Circulation 2004;109:2698-704.

Chapter 12

164


165

Chapter

13effeCts on CoagulatIon and fIbrInolysIs

InduCed by Influenza In mICe wIth a reduCed CapaCIty to generate aCtIvated

proteIn C and a defICIenCy In plasmInogen aCtIvator InhIbItor type-1

Tymen Keller • Koen van der Sluijs • Martijn de Kruif • victor gerdes • joost Meijers Sandrine Florquin • Tom van der Poll • Eric van gorp

Dees Brandjes • Harry Büller • Marcel Levi

Department of Vascular Medicine and Experimental Internal Medicine and Pathology, Academic Medical Center, Amsterdam, The Netherlands

Department of Internal Medicine, Slotervaart Hospital, Amsterdam, The Netherlands

Circulation Res 2006 (in press)

ABSTRACT Influenza infections increase the risk of diseases associated with a prothrombotic state such as venous thrombosis and atherothrombotic diseases. However, it is unclear whether influenza leads to a prothrombotic state in vivo. To determine whether influenza activates coagulation, we measured coagulation and fibrinolysis in influenza infected C57Bl/6 mice. We found that influenza increased thrombin generation, fibrin deposition and fibrinolysis. In addition, we used various anti- and prothrombotic models to study pathways involved in the influenza-induced prothrombotic state. A reduced capacity to generate activated protein C in TMpro/pro mice, increased throm-bin generation and fibrinolysis, whereas treatment with heparin decreased thrombin generation in influenza infected C57Bl/6 mice. Thrombin generation was not changed in hyperfibrinolytic mice, deficient in plasminogen activator inhibitor type-1 (PAI-1-/-), however, increased fibrin degradation was seen. Treatment with tranexamic acid reduced fibrinolysis, but thrombin generation was unchanged. We conclude that influenza infec-tion generates thrombin, increased by reduced levels of protein C and decreased by hepa-rin. The fibrinolytic system appears not to be important for thrombin generation. These findings suggest that influenza leads to a prothrombotic state by coagulation activation. Heparin treatment reduces the influenza induced prothrombotic state.

Chapter 13

168

INTRODuCTIONAcute respiratory tract infections are associated with an increased risk of venous thrombotic disease and ischemic heart disease. In the first two weeks after a respiratory tract infection the risk of venous thrombotic disease is increased 2-fold and the risk of acute ischemic heart disease is increased 5-fold.1, 2 Although the underlying mechanism of this association is unclear, systemic inflammation and coagulation activation are potential culprits.3 In recent years, it has been demonstrated that influenza virus is capable to modulate inflammation and activate coagulation in vitro.4, 5 Visseren et al. showed that endothelial cells incubated with influenza virus elicit interleukine-6 (IL-6). In addition, they showed that influenza-incubated monocytes are prothrombotic compared to non-infected control cells.6 It was suggested that the influenza-induced coagulation activation is tissue factor (TF) dependent. However, also increased plasminogen-activator inhibitor-1 (PAI-1) levels were shown. Increased PAI-1 levels might induce a prothrombotic state by inhibition of fibrinolysis. These experiments used in vitro techniques to study the prothrombotic effect of influenza. Until now we do not known if influenza infection results in a prothrombotic state in vivo, and it remains unknown which pathways are involved in the prothrombotic state. Understanding of the influenza-induced prothrombotic state might improve prevention of venous thrombotic disease and ischemic heart disease during influenza epidemics. Hallmarks of the inflammation induced prothrombotic state are coagulation activation and dysfunctional physiological anticoagulant pathways, most notably the protein C system.7 It has been shown that thrombomodulin (TM) is down regulated by cytokines.8 Because the activity of the protein C system is driven by TM,9 down regulation of the receptor is respon-sible for the inactivation the protein C system. In addition, PAI-1-mediated inhibition of fibrinolysis is another hallmark of inflammation-induced coagulation activation. In animal models it has been demonstrated that impaired fibrinolysis is due to high levels of the fibri-nolytic inhibitor plasminogen activator inhibitor, type 1 (PAI-1).10

The purposes of this study were to determine whether influenza results in a prothrombotic state, to quantify the role of the protein C system and fibrinolysis, and to test whether anti-thrombotic treatment could prevent the prothrombotic state. In an experimental study we measured coagulation and fibrinolysis in influenza infected C57Bl/6, prothrombotic TMpro/

pro and hyperfibrinolytic PAI-1-/- mice and mice treated with heparin or tranexamic acid. Our data show that influenza results in a prothrombotic state in vivo. In addition it was shown that reduced levels of protein C increase the prothrombotic state. Treatment with heparin reduced thrombin generation during influenza. Although the fibrinolytic system does not effect thrombin generation, prolonged increased levels of PAI-1 during influenza may be associated with a prothrombotic state as a result of impaired fibrinolysis.

MATERIALS AND METHODS Animals All experiments were approved by the Institutional Animal Care and Use Committee of the Academic Medical Center, Amsterdam, the Netherlands. We used female C57Bl/6 mice to study the natural history of influenza virus infection and the prothrombotic response in wild-type animals. To characterize coagulation activation during influenza we used female TMpro/pro

Effects on coagulation and fibrinolysis induced by influenza in mice with a reduced capacity to generate activated protein C and a deficiency in plasminogen activator inhibitor type-1

169

mice and female TM Wt mice as described recently by Weijer et al.11 TMpro/pro mice with a mutation in the gene for TM, specifically a single amino acid substitution (Glu404Pro), were kindly provided by Dr. R. D. Rosenberg (Massachusetts Institute of Technology, Cambridge). TMpro/pro mice were generated on a C57Bl/6 background, as previously described12 and TM Wt control mice were parallel bred second generation wild-type littermates. TMpro/pro mice have a decreased capacity to activate protein C by approximately 1000-fold and have an approxi-mately 100-fold decreased ability to bind thrombin.12 The role of fibrinolysis was studied using male PAI-1-/- mice, generated as previously described.13 Male PAI-1 wild type mice were used as control mice. PAI-1-/- mice and PAI-1 Wt mice used in the experiments were the product of 8 backcrosses to the C57Bl/6J genetic background. PAI-1-/- mice have a decreased abil-ity to inhibit plasmin generation and are thereby hyperfibrinolytic. Additional experiments were done in female C57Bl/6 mice treated with low molecular weight heparin (LMWH), tranexamic acid or phosphate-buffered saline (PBS). Daily subcutaneous injections with 5IU LMWH (Fraxiparine, GlaxoSmithKline) were given 3 days prior and during the experiments. Tranexamic acid (Cyklokapron, Pfizer) was administered orally at a dose of 150mg/Kg 3 days prior and during the study experiments. Study groups consisted of at least six mice per study group. All mice used in the experiments were at the age of 8-12 weeks.

Experimental infection

A previously described model of a non lethal influenza infection in mice was used.14 Briefly, isoflurane gas (2% + 2 L/min O2) anaesthetized mice were intranasally inoculated with influ-enza A (strain A/PR/8/34) or control inoculum in a final volume of 50 μl PBS. The mice were retained in a supine position, and breathing was monitored to assure complete inhalation of virus dose without immediate aspiration. Wild-type C57Bl/6 mice were inoculated with either 10 or 100 median tissue culture infective dose (TCID50) influenza or control buffer (PBS). Mice were sacrificed after intraperitoneal injection with 0.3 ml FFM (fentanyl citrate 0.079 mg/ml, fluanisone 2.5 mg/ml, midazolam 1.25 mg/ml in H20) on day 0, 2 or 4. TMpro/pro, TM Wt, PAI-1-/- and PAI-1 Wt mice were inoculated with 100 TCID50 influenza and sacrificed on day 0, 2, 4, 8 and 14. C57Bl/6 mice treated with LMWH, tranexamic acid or control buffer received 100 TCID50 influenza or PBS and were sacrificed after four days.

Assays

Blood was drawn from the intracardial cavity and anti-coagulated with sodium citrate (fi-nal concentration 6.4%). Plasma samples were centrifuged twice at 680g for 10 minutes, and frozen at −80 ºC until assays were performed. Thrombin generation was assessed by measuring thrombin-antithrombin complexes (TAT-c) with an enzyme-linked immu-nosorbent assay (ELISA) for the detection of these complexes in mice.15 PAI-1 levels were also measured with an ELISA, using antibodies directed at murine PAI-1.16 Degradation of fibrin in plasma was determined by measuring D-dimer levels with an ELISA, as previ-ously described.17 Lung were homogenized at 4°C in 4 volumes of sterile saline using a tissue homogenizer (Biospec Products, Bartelsville, OK). Homogenates were diluted 1:1 in lysis buffer (150mM NaCl, 15 mM Tris, 1 mM MgCl2, 1 mM CaCl2, 1% Triton X−100, 100 μg/mL pepstatin A, 100μg/mL leupeptin, and 100 μg/mL aprotinin) and incubated

Chapter 13

170

on ice for 30 minutes. Supernatants were centrifuged for 10 minutes at 3000 rpm and frozen at −20 ºC until assays were performed. Determination of D-dimer levels in lung homogenates was done with a sensitive ELISA from Diagnostica Stago (Asserachrom™). TNF-a and IFN-γ were measured in total-lung-lysate by Cytometric Bead Assay (BD Bio-sciences Pharmingen, San Diego, CA) as described previously.18 The detection level for these cytokines is 2.5 pg/ml, i.e. 25 pg/g lung tissue.

Immunohistochemical staining of lung tissue

Lungs were fixed in 10% formalin for 12 hours and embedded in paraffin in a routine fashion. Four-micrometer sections were stained with hematoxylin and eosin (H&E), in accordance with standard procedures. Granulocyte and fibrin staining were done as previ-ously described.19, 20 In brief, slides were deparaffinized, rehydrated and then digested by a solution of pepsin 0.25% (Sigma-Aldrich, St. Louis, MO) in 0.01 M HCl. After being rinsed, the sections were incubated in 10% normal goat serum (DAKO, Glostrup, Denmark). For granulocyte staining, sections were exposed to FITC-labeled anti-mouse Ly-6-G mAb (BD PharMingen, San Diego, CA). Secondary incubation was done with rabbit anti-FITC antibody (DAKO), with a biotinylated swine anti-rabbit antibody (DAKO). For the detection of fibrin, sections were incubated with human anti-mouse fibrin(-ogen) antibody (Ixell; Accurate Chemical & Scientific, Westbury, NY). For both stains a streptavidin-ABC solution (DAKO); 0.03% H2O2 and 3,3-diaminobenzidine tetrahydrochloride (Sigma) in 0.05 mol/L Tris (pH 7.6) was used as substrate. Fibrin slides were counterstained with methyl green (Sigma).


171

FIguRE 1: A THROugH F

Influenza induced changes in lungs. Histopathological analysis of lungs of influenza infected C57BL/6 mice. Lung sec-tions were stained with hematoxylin and eosin (HE)(A,B), Ly-6 for granulocytes (C,D) or fibrin (E,F). Compared to control mice (A,C,E), influenza infected mice (B,D,F) showed inflammation in the lungs (B), enhanced granulocyte influx (D) and fibrin deposition (F). Slides are representative of 6 mice, 4 days after influenza inoculation. Magnification, X 40 for HE and fibrin and X 200 for Ly-6. Mice inoculated with PBS (®) or influenza (n) and sacrificed on day 0, 2, 4 showed more granulocytes (G), and increased number of intravascular thrombi (H) four days after inoculation. General fibrin deposition was increased as established by determining the overall presence of fibrin in lung tissue by digital image analyzer (I). Data are presented as mean ± SE; n = 6 for each group. *P <0.05 versus control and **P <0.01 versus control.

Histopathological scoring

All histopathological analysis were performed in a blinded fashion for genetic back-ground and study group. The groups analyzed consisted of six randomly chosen mice. Images were obtained with an Olympus BX51 microscope (Olympus, Zoeterwoude, The Netherlands) using Olympus Plan objectives (4 x/0.13 numeric aperture [NA], 10 x/0.30 NA, 20 x/0.50 NA, and 40 x/0.85) and an Olympus DP70 camera. Images were acquired with Olympus DPController software and were processed with Adobe Photoshop CS (Adobe Systems, San Jose, CA). The level of inflammation was assessed by counting the number of granulocytes in ten randomly chosen non-overlapping fields (X 200 magni-fication). The extent of fibrin deposition in the lungs was assessed with two different scoring techniques. First, fibrin deposition was expressed as number of intravascular thrombi in ten randomly chosen non-overlapping fields (X 40 magnification) of lung tis-sue. Second, lung section were analyzed for fibrin using a digital image analyzer (Image pro-plus, Mediacybernetics, Germany). Area’s of 10 mm2 were analyzed for fibrin and results are expressed as a percentage of the analyzed tissue.

Determination of viral outgrowth Viral load was determined using real-time quantitative PCR as described.21 RNA extraction was done with 100 μl of lung-homogenates and Trizol reagent. RNA was resuspended in 10μl DEPC-treated water. cDNA synthesis was performed using 1μl of the RNA-suspen-sion and a random hexamer cDNA synthesis kit (Applera, Foster City, CA). A total of 5μl out of 25μl cDNA was used for amplification in a quantitative real-time PCR reaction (ABI PRISM 7000 Sequence Detector System). The viral load present in a sample was calculat-ed using standard curve of particle counted influenza virus (virus particles were counted by electron microscopy), included in every assay run. The following primers were used: 5’-GGACTGCAGCTGAGACGCT-3’ (forward); 5’-CATCCTGTTGTATATGAGGCCCAT-3’ (reverse) and 5’-CTCAGTTATTCTGCTGGTGCACTTGCC-3’ (5’-FAM labeled probe). The results of the real time PCR was corrected for the total RNA in the samples.


Results are presented as means plus or minus standard error of the mean (SE). Statistical analysis was performed using Mann-Whitney U test for comparisons between groups. A P−value of less than 0.05 was considered statistically significant.

Chapter 13

172


173

FIguRE 1: g THROugH I

RESuLTSNatural course of influenza infection in mice

We used a non-lethal mouse-model for influenza infection in wild-type C57Bl/6 mice to study prothrombotic changes during influenza. Previously we demonstrated that influ-enza virus replicates in the lungs, with peak levels at day four. Furthermore we showed that influenza is cleared from the lungs after two weeks.22 Influenza increased TNF-a and IFN-γ levels and induced interstitial inflammation in the lungs. Our data show that influ-enza resulted in granulocyte influx in the lung as can be seen from Figure 1G. In addition, influenza induced a prothrombotic state as evidenced by increased fibrin deposition. Fig-ure 1H shows the number of intravascular thrombi counted in lungs of influenza infected mice. A statistical significant increased number of intravascular thrombi was seen on day 4 compared to non-infected control mice (P < 0.01). Also, the measurements presented in Figure 1I demonstrate that overall fibrin deposition in lungs increased during influenza (P < 0.05).

Chapter 13

174

FIguRE 2: COAguLATION AND FIBRINOLySIS IN INFLuENZA INFECTED MICE

Plasma levels of TAT-c (A) and PAI-1 (B) in C57Bl/6 mice inoculated with control buffer (PBS), low dose influenza (Infl1) or high dose (Infl2) influenza. Mice were sacrificed on day 0, 2, and 4. Data are presented as mean ± SE. ** P <0.01 low dose versus control or high dose versus low dose

Influenza induced coagulation activation is dose-dependent. Our data show that influenza increased TAT-c concentration in blood with time. TAT-c levels increased with increasing influenza dose (Figure 2A). Also, PAI-1 levels signifi-cantly increased during influenza infection. Figure 2B presents PAI-1 levels in control mice and mice infected with low and high dose influenza. Increased levels of PAI-1 were found after 2 days. Also PAI-1 levels increased dependent on the influenza dose.

Influenza infection in TMpro/pro and PAI-1 -/- mice. To determine whether influenza replicates in TMpro/pro and PAI-1 -/- mice we measured influenza RNA in homogenized lung tissue. The real time quantitative PCR results pre-sented in Figure 3 show that influenza inoculum increased 100-fold after 4 days in all mice studied. After 4 days, however, no difference was seen between transgenic mice and control mice. To show influenza-induced inflammation in TMpro/pro and PAI-1-/- mice, we counted the number of granulocytes in lungs. We found that the number of granulocytes in influenza infected TMpro/pro mice, PAI-1-/- and control mice were not different (data not shown). To obtain more insight into the immune response of TMpro/pro and PAI-1-/- mice,


175

FIguRE 3: INFLuENZA LOAD IN TMPRO/PRO MICE AND PAI-1-/- MICE

TMpro/pro mice (A) and PAI-1-/- mice (B) were inoculated with influenza at baseline. Viral outgrowth in the lungs of control mice (°) and transgenic mice (l) was determined 4 days after inoculation. Viral load is expressed as log10 viral RNA copies per _g RNA. There was no difference in viral load between transgenic mice and wild type mice after 4 days.

we measured levels of TNF-a and IFN-γ. Our data covering the first two weeks after influ-enza inoculation show that influenza increased levels of TNF-a and IFN-γ. However, no differences were seen between transgenic and control mice (Table 2). Pro- and anticoagulant properties effects the prothrombotic state Having determined that influenza induced a prothrombotic state, we next investigated the role of pro- and anticoagulant pathways in the prothrombotic state. The mean levels of TAT-c, PAI-1 and D-dimer during the experiment are plotted in Figure 4 for influenza infected prothrombotic TMpro/pro mice and influenza infected TM Wt mice. These data show that influenza induced higher TAT-c levels and D-dimer levels in TMpro/pro mice than in Tm Wt mice. After two weeks differences in TAT-c levels between TMpro/pro mice and Tm Wt mice diminished. Influenza increased the PAI-1 concentration in blood of TMpro/pro mice and TM Wt mice, however, no difference was seen between these mice (Figure 4B). In the convalescent phase, two weeks after influenza inoculation, PAI-1 and D-dimer concentrations remained increased compared to baseline.

TAT-c (A), PAI-1 (B) and D-dimer (C) levels in plasma. TM pro/pro (l) and TM Wt mice (°) were inoculated with influenza and sacrificed on 0, 2, 4, 8 and 14 days after inoculation with influenza. Data are presented as mean ± SE. A significant increase in TAT-c and D-dimer was demonstrated in TM pro/pro mice compared to TM Wt mice. No difference in PAI-1 levels were seen. ** P <0.01 versus control..

Chapter 13

176

FIguRE 4: THROMBIN gENERATION AND FIBRIN DEgRADATION INCREASES IN INFLuENZA INFECTED TM PRO/PRO MICE

To assess whether thrombin generation results in increased fibrin deposition we counted the number of thrombi in lungs of in TMpro/pro mice and TM Wt mice. Figure 5 presents the number of thrombi in lungs of influenza infected mice. No difference were found in fibrin deposition between TMpro/pro mice and TM Wt mice. To establish these findings any further, D-dimer levels were measured in homogenized lungs tissue of these mice. Because highest levels of fibrin were expected four days after influenza inoculation, we tested the D-dimer levels in lung tissue on the fourth day after infection. We found no difference between TMpro/pro mice and TM Wt mice (data not shown).

We next tested whether we could reduce the influenza-induced prothrombotic state by treatment with low molecular weight heparin (LMWH). The mean levels of TAT-c, PAI-1 and D-dimer are plotted in Figure 6 for the non-infected C57Bl/6 mice and influenza infected C57Bl/6 mice treated with LWHW or control buffer (PBS). Our data show that the concentration of TAT-c in blood decreased in LMWH treated mice, however, little change was seen in PAI-1 and D-dimer levels.


177

FIguRE 5: LuNgS OF TMPRO/PRO MICE SHOw NO MORE FIBRIN DEPOSITION

TM Wt mice (®) and TMpro/pro mice (n) were inoculated with influenza and sacrificed on day 0, 4 and 8. Lung sections were stained for fibrin. Histopathological analysis showed no difference between TM Wt mice and TMpro/pro mice. Data are presented as mean ± SE.

TM Wt

The fibrinolytic system during influenza infection.

To obtain more insight into the role of the fibrinolytic system in the influenza-induced pro-thrombotic state, influenza infected PAI-1-/- and PAI-1 Wt mice were studied. Levels of TAT-c, PAI-1 and D-dimer during influenza are given in Figure 7. The results presented in Figure 7A show no increased TAT-c concentration in blood of PAI-1-/- mice compared to PAI-1 Wt mice. However, the results show that influenza induced higher D-dimer levels in PAI-1-/- mice than in PAI-1 Wt mice (Figure 7C). The PAI-1 concentration in blood remained high until 14 days

Chapter 13

178

FIguRE 6: TREATMENT wITH LMwH REDuCE THE PROTHROMBOTIC STATE

Plasma levels of TAT-c (A), PAI-1 (B) and D-dimer (C) in influenza inoculated C57BL/6 mice treated with PBS (n) or LMWH (n). A second control group of C57Bl/6 mice was inoculated with PBS and treated with PBS during the experiment (®). Mice were sacrificed 4 days after inoculation. Influenza increased TAT-c, PAI-1 and D-dimer levels. In influenza infected mice, treatment with LMWH reduced TAT-c levels compared to PBS. Data are presented as mean ± SE. ** P <0.01 influ-enza versus PBS or influenza + PBS versus influenza + LMWH.

after influenza inoculation compared to baseline. Figure 8 presents the number of thrombi in lungs of influenza infected in PAI-1-/- mice and PAI-1 Wt mice. No difference in thrombi was found in lungs of PAI-1-/- mice compared to PAI-1 Wt mice. In homogenized lungs tissue we measured D-dimer levels but also found no difference between PAI-1-/- mice and PAI-1 Wt mice (data not shown). To determine the role of the fibrinolytic system any further, we treated C57Bl/6 mice with tranexamic acid. Figure 9 presents levels of TAT-c, D-dimer and PAI-1 in control C57Bl/6 mice and C57Bl/6 mice treated with tranexamic acid or control buffer (PBS). Our data show that the concentration of D-dimer decreased in tranexamic acid treated mice, whereas TAT-c and PAI-1 levels were not effected.


179

Chapter 13

180

FIguRE 7: FIBRIN DEgRADATION INCREASE IN INFLuENZA INFECTED PAI-/-

MICE

TAT-c (A), PAI-1 (B) and D-dimer (C) levels in plasma. PAI-1-/- mice (l) and PAI-1 Wt mice (°) were inoculated with influenza and sacrificed on 0, 2, 4, 8 and 14 days after inoculation with influenza. Data are presented as mean ± SE. PAI-1 levels were undetectable in PAI-1-/- mice, but remained increased in control mice until 14 days after infection. D-dimer levels were significantly increased in PAI-1-/- compared to PAI-1 Wt mice. No differences in TAT-c levels were seen. ** P <0.01 versus control.


181

FIguRE 8

DISCuSSIONThe present study provides evidence that influenza results in a prothrombotic state in vivo. This is shown by increased thrombin generation and fibrin deposition in influenza infected C57BL/6 mice. It is also demonstrated that the protein C system plays an impor-tant role in the influenza-induced prothrombotic state. The reduced capacity to activate protein C in TMpro/pro mice increased thrombin generation and fibrin degradation. Prolonged down-regulation of fibrinolysis by high levels of PAI-1 result in a prothrom-botic state in the convalescent phase of influenza, as evidenced by sustained high levels of D-dimer. Administration of heparin decreased thrombin generation in C57BL/6 infected with influenza. Collectively, these data suggest that influenza induce imbalance between coagulation and fibrinolysis, which can be treated with heparin.

The present observations add to the notion that influenza virus elicits coagulation activa-tion in vitro.23, 24 Visseren et al. showed that influenza-incubated endothelial cells and monocytes have procoagulant properties.4, 5 They proposed that inflammation induced increased tissue factor (TF) expression and activation of Factor VII is causative for the prothrombotic state. However recently, also a TF/FVIIa independent pathway has been implicated in the procoagulant state during viral infections. Marsden et al. showed that fibrinogen-like protein 2 (Fgl2) plays an important role in the pathogenesis of fibrin deposition during viral hepatitis.25 Although Fgl2 has been shown able to cleave pro-thrombin into thrombin, data on the inflammation induced prothrombotic properties of Fgl2 are not uniform.26 We here demonstrate that thrombin generation during influenza infection is effected by the protein C system and treatment with LMWH. Previous stud-ies have characterized the inflammatory response in this model.14, 22, 27 Van der Sluijs et al. showed that influenza is replicates in the lungs with peak levels after four days. Furthermore, levels of cytokines including TFN-a and IFN-γ were increased during infec-

Lungs of PAI-/- mice show no more fibrin deposition. PAI-1 Wt mice (®) and PAI-1-/- mice (n) were inoculated with influenza and sacrificed on day 0, 4 and 8. Lung sections were stained for fibrin. Histopathological analysis showed no difference between PAI-1 Wt mice and PAI-1-/- mice. Data are presented as mean ± SE.

tion. We found comparable results and showed the accompanying prothrombotic state. Our data suggests that influenza infection lead to a prothrombotic state by activation of the coagulation cascade instead of direct cleavage of prothrombin into thrombin.

Some aspects of the study merit further consideration. First, we showed that influenza virus replicates in the lungs of the mice studied. But of note, we measured viral load by molecu-lar techniques which could have overestimate the total amount of vial influenza. Standard plaque assay would have detected only vial virus, but we do not expect that the conclusion of this study would change significantly. Second, although differences in TAT-c, PAI-1 and D-dimer levels were evident, we were unable to demonstrate differences in fibrin deposition

Chapter 13

182

FIguRE 9: TREATMENT wITH TRANEXAMIC ACID INCREASE FIBRINOLySIS

Plasma levels of TAT-c (A), PAI-1 (B) and D-dimer (C) in C57BL/6 mice inoculated with influenza and treated with PBS (n) or tranexamic acid (n). Control mice were inoculated with PBS and treated with PBS during the experiment (®). Mice were sacrificed 4 days after influenza inoculation. Influenza increased TAT-c, PAI-1 and D-dimer levels. In influenza infected mice, treatment with tranexamic acid decreased D-dimer levels compared to PBS. Data are presented as mean ± SE. ** P <0.01 influenza versus PBS or influenza + PBS versus influenza + tranexamic acid.

A

B

C

between prothrombotic and hyperfibrinolytic mice. There are several explanation for the latter, including the lack of specificity for fibrin/ fibrinogen in the assays used, the pre-exis-tent strong fibrinolytic system in mice28 and the relative insensitive pathological analysis. Previous studies used others techniques, such as electron microscopy29 and radiolabelled fibrin15 to assess for fibrin in micro vascular beds. However, these methods are not quanti-tative or might be insufficient discriminative for fibrin formation in small vessel. Although we were not able to demonstrate increased fibrin deposition in lungs of TMpro/pro and PAI deficient mice, we can not exclude that more fibrin is formed.Unexpectedly, although evidence indicates the role of both TM and PAI-1 activity in sep-sis,30 our results suggest that the anticoagulant properties of TM and PAI-1 seems not important for host defense against influenza. Previously, it has been shown that high lev-els of pro-inflammatory cytokines down regulate TM, which result in a defective protein C system.31 In addition, an intact protein C system has been shown to ameliorates the inflammatory response.32 In this influenza model however, we were unable to show that a reduced capacity to activate protein C modulates the inflammatory response. Previously Conway et al. clearly indicate that TM, and in particular its lectin domain, has anti-inflam-matory properties that can be dissected from the anticoagulant properties of TM.33 In this respect it is important to realize that the TMpro/pro mice used here have an intact TM lectin domain, which might explain the intact immune response in these mice. Although PAI-1 activity has been shown to predict lethality in patients with sepsis,34 and a polymorphism in the PAI-1 gene influences the risk of septic shock,35, 36 we here demonstrate that the complete absence of PAI-1 does not influence the inflammatory response during influenza infection. Recently, Renckens et al. showed that PAI-1-/- mice have increased inflammatory response upon tissue injury, however in a very time-limited manner only.37 Our data indicate that on the long-term plasmin generation in general has no role in host defense against influenza. Although we failed to show that a reduced capacity to activate protein C and PAI deficiency modulate the inflammatory response, these conclu-sion need comment. First, our study was designed to detect a significant difference in coagulant response and we could have missed subtle inflammatory changes. And second, studies from others have shown differential results of the inflammatory response upon infection with different pathogens with TMpro/pro mice11, 20 and PAI-/ mice.37, 38 With this reservation in mind, current murine data suggest that the anticoagulant properties of TM and PAI-1 seem not important for host defense against influenza.From the results obtained we conclude that influenza results in a prothrombotic state, dependent on coagulation activation and on coagulation regulation. This fact has not been established before and offers an explanation for the link between influenza infections and (athero) thrombotic disease. However, determining whether the influenza-induced prothrombotic state is clinically relevant for the development venous thrombosis and acute atherothrombotic events requires future study. In addition, future research may further elucidate the point of impact of influenza infection and more sensitive techniques for fibrin detection and the use of cell culture systems from different transgenic mice may provide additional insights. Finally we conclude that treatment with anti-thrombotic compounds decrease the influenza-induced prothrombotic state.


183

ACKNOwLEDgEMENTSThe authors thank Joost Daalhuisen and Marieke ten Brink for technical assistance.

Chapter 13

184

Reference List 1. Smeeth L, Thomas SL, Hall AJ et al. Risk of myocardial infarction and stroke after acute infection or vaccination. N Engl J Med 2004;351:2611-8. 2. Smeeth L, Cook C, Thomas S et al. Risk of deep vein thrombosis and pulmonary embolism after acute infection in a community setting. Lancet 2006;367:1075-9. 3. Stone PH. Triggering myocardial infarction. N Engl J Med 2004;351:1716-8. 4. Bouwman JJ, Visseren FL, Bosch MC et al. Procoagulant and inflammatory response of virus-infected monocytes. Eur J Clin Invest 2002;32:759-66. 5. Visseren FL, Bouwman JJ, Bouter KP et al. Procoagulant activity of endothelial cells after infection with respiratory viruses. Thromb Haemost 2000;84:319-24. 6. Visseren FL, Verkerk MS, Bouter KP et al. Interleukin-6 production by endothelial cells after infection with influenza virus and cytomegalovirus. J Lab Clin Med 1999;134:623-30. 7. Levi M, van der PT, Buller HR. Bidirectional relation between inflammation and coagulation. Circulation 2004;109:2698-704. 8. Moore KL, Esmon CT, Esmon NL. Tumor necrosis factor leads to the internalization and degradation of thrombomodulin from the surface of bovine aortic endothelial cells in culture. Blood 1989;73:159-65. 9. Esmon CT. The protein C pathway. Chest 2003;124:26S-32S. 10. Biemond BJ, Levi M, ten Cate H et al. Plasminogen activator and plasminogen activator inhibitor I release during experimental endotoxaemia in chimpanzees: effect of interventions in the cytokine and coagulation cascades. Clin Sci (Lond) 1995;88:587-94. 11. Weijer S, Wieland CW, Florquin S et al. A thrombomodulin mutation that impairs activated protein C generation results in uncontrolled lung inflammation during murine tuberculosis. Blood 2005;106:2761-8. 12. Weiler-Guettler H, Christie PD, Beeler DL et al. A targeted point mutation in thrombomodulin generates viable mice with a prethrombotic state. J Clin Invest 1998;101:1983-91. 13. Carmeliet P, Kieckens L, Schoonjans L et al. Plasminogen activator inhibitor-1 gene-deficient mice. I. Generation by homologous recombination and characterization. J Clin Invest 1993;92:2746-55. 14. van der Sluijs KF, van Elden L, Nijhuis M et al. Toll-like receptor 4 is not involved in host defense against respiratory tract infection with Sendai virus. Immunol Lett 2003;89:201-6. 15. Healy AM, Hancock WW, Christie PD et al. Intravascular coagulation activation in a murine model of throm bomodulin deficiency: effects of lesion size, age, and hypoxia on fibrin deposition. Blood 1998;92:4188-97. 16. Declerck PJ, Verstreken M, Collen D. Immunoassay of murine t-PA, u-PA and PAI-1 using monoclonal anti bodies raised in gene-inactivated mice. Thromb Haemost 1995;74:1305-9. 17. Elms MJ, Bunce IH, Bundesen PG et al. Measurement of crosslinked fibrin degradation products - an immunoassay using monoclonal antibodies. Thromb Haemost 1983;50:591-4. 18. Lauw FN, ten Hove T, Dekkers PE et al. Reduced Th1, but not Th2, cytokine production by lymphocytes after in vivo exposure of healthy subjects to endotoxin. Infect Immun 2000;68:1014-8. 19. Rijneveld AW, Levi M, Florquin S et al. Urokinase receptor is necessary for adequate host defense against pneumococcal pneumonia. J Immunol 2002;168:3507-11. 20. Rijneveld AW, Weijer S, Florquin S et al. Thrombomodulin mutant mice with a strongly reduced capacity to generate activated protein C have an unaltered pulmonary immune response to respiratory pathogens and lipopolysaccharide. Blood 2004;103:1702-9. 21. van Elden LJ, Nijhuis M, Schipper P et al. Simultaneous detection of influenza viruses A and B using real- time quantitative PCR. J Clin Microbiol 2001;39:196-200. 22. van der Sluijs KF, van Elden LJ, Xiao Y et al. IL-12 deficiency transiently improves viral clearance during the late phase of respiratory tract infection with influenza A virus in mice. Antiviral Res 2006;70:75-84. 23. Sutherland MR, Raynor CM, Leenknegt H et al. Coagulation initiated on herpesviruses. Proc Natl Acad Sci U S A 1997;94:13510-4. 24. Dam-Mieras MC, Muller AD, van Hinsbergh VW et al. The procoagulant response of cytomegalovirus infected endothelial cells. Thromb Haemost 1992;68:364-70. 25. Marsden PA, Ning Q, Fung LS et al. The Fgl2/fibroleukin prothrombinase contributes to immunologically mediated thrombosis in experimental and human viral hepatitis. J Clin Invest 2003;112:58-66. 26. Hancock WW, Szaba FM, Berggren KN et al. Intact type 1 immunity and immune-associated coagulative responses in mice lacking IFN gamma-inducible fibrinogen-like protein 2. Proc Natl Acad Sci U S A 2004;101:3005-10. 27. van der Sluijs KF, van Elden LJ, Arens R et al. Enhanced viral clearance in interleukin-18 gene-deficient mice after pulmonary infection with influenza A virus. Immunology 2005;114:112-20. 28. Carmeliet P, Stassen JM, Schoonjans L et al. Plasminogen activator inhibitor-1 gene-deficient mice. II. Effects on hemostasis, thrombosis, and thrombolysis. J Clin Invest 1993;92:2756-60. 29. Hirota-Kawadobora M, Kani S, Terasawa F et al. Functional analysis of recombinant Bbeta15C and Bbeta15A fibrinogens demonstrates that Bbeta15G residue plays important roles in FPB release and in lateral ag gregation of protofibrils. J Thromb Haemost 2005;3:983-90. 30. Levi M, Keller TT, van Gorp E et al. Infection and inflammation and the coagulation system. Cardiovasc Res 2003;60:26-39. 31. van der Poll T, de Jonge E, Levi M. Regulatory role of cytokines in disseminated intravascular coagulation.


185

Semin Thromb Hemost 2001;27:639-51. 32. Taylor FB, Jr., Chang A, Esmon CT et al. Protein C prevents the coagulopathic and lethal effects of Escherichia coli infusion in the baboon. J Clin Invest 1987;79:918-25. 33. Conway EM, Van de WM, Pollefeyt S et al. The lectin-like domain of thrombomodulin confers protection from neutrophil-mediated tissue damage by suppressing adhesion molecule expression via nuclear factor kappaB and mitogen-activated protein kinase pathways. J Exp Med 2002;196:565-77. 34. Mesters RM, Florke N, Ostermann H et al. Increase of plasminogen activator inhibitor levels predicts out come of leukocytopenic patients with sepsis. Thromb Haemost 1996;75:902-7. 35. Hermans PW, Hibberd ML, Booy R et al. 4G/5G promoter polymorphism in the plasminogen-activator-inhibi tor-1 gene and outcome of meningococcal disease. Meningococcal Research Group. Lancet 1999;354:556-60. 36. Westendorp RG, Hottenga JJ, Slagboom PE. Variation in plasminogen-activator-inhibitor-1 gene and risk of meningococcal septic shock. Lancet 1999;354:561-3. 37. Renckens R, Roelofs JJ, De Waard V et al. The role of plasminogen activator inhibitor type 1 in the inflammatory response to local tissue injury. J Thromb Haemost 2005;3:1018-25. 38. Rijneveld AW, Florquin S, Bresser P et al. Plasminogen activator inhibitor type-1 deficiency does not influence the outcome of murine pneumococcal pneumonia. Blood 2003;102:934-9. 31. van der Poll T, de Jonge E, Levi M. Regulatory role of cytokines in disseminated intravascular coagulation. Semin Thromb Hemost 2001;27:639-51. 32. Taylor FB, Jr., Chang A, Esmon CT et al. Protein C prevents the coagulopathic and lethal effects of Escherichia coli infusion in the baboon. J Clin Invest 1987;79:918-25. 33. Conway EM, Van de WM, Pollefeyt S et al. The lectin-like domain of thrombomodulin confers protection from neutrophil-mediated tissue damage by suppressing adhesion molecule expression via nuclear factor kappaB and mitogen-activated protein kinase pathways. J Exp Med 2002;196:565-77. 34. Mesters RM, Florke N, Ostermann H et al. Increase of plasminogen activator inhibitor levels predicts out come of leukocytopenic patients with sepsis. Thromb Haemost 1996;75:902-7. 35. Hermans PW, Hibberd ML, Booy R et al. 4G/5G promoter polymorphism in the plasminogen-activator-inhibi tor-1 gene and outcome of meningococcal disease. Meningococcal Research Group. Lancet 1999;354:556-60. 36. Westendorp RG, Hottenga JJ, Slagboom PE. Variation in plasminogen-activator-inhibitor-1 gene and risk of meningococcal septic shock. Lancet 1999;354:561-3. 37. Renckens R, Roelofs JJ, De Waard V et al. The role of plasminogen activator inhibitor type 1 in the inflammatory response to local tissue injury. J Thromb Haemost 2005;3:1018-25. 38. Rijneveld AW, Florquin S, Bresser P et al. Plasminogen activator inhibitor type-1 deficiency does not influence the outcome of murine pneumococcal pneumonia. Blood 2003;102:934-9.

Chapter 13

186


187

summary

190

Summary

191

SuMMARy The purpose of this thesis was to explore the infection and inflammation induced condi-tions associated with the increased risk of venous thrombosis and cardiovascular disease. The first part addresses the role of inflammation and coagulation and the risk factors for cardiovascular disease. The second part assesses the relation between infections and venous thrombosis and acute coronary heart disease. The third part of this thesis describes studies on the pathophysiology of infection-induced prothrombotic state.

Part 1: Inflammation and coagulation as cardiovascular risk factors

In the pathogenesis of atherosclerosis lipid deposition and inflammation in the arte-rial wall posses a central role.1 During the last decades numerous studies have been performed to elucidate the effect of inflammation on various lipoproteins (and visa versa) and the pro- and anti-inflammatory processes involved in cardiovascular diseases. Research showed that a disturbed balance between pro- and anti-inflammatory mecha-nisms and that pro- and anti-thrombotic mechanisms play a role in plaque weakening and subsequent plaque rupture leading to coronary heart disease. Although several crucial factors have been identified, the pathophysiology of atherogenesis and the subse-quent clinical diseases has not been elucidated. Recently, the innate immunity has been linked to the development of atherosclerosis and cardiovascular disease.2 Therefore, we explored the relationship between components of the (innate) immunity (mannose-bind-ing lectin and secretory phospholipase A2), and toll-like receptors and the risk of future coronary artery disease. In chapter 3 we assessed prospectively the relationship between mannose binding lectin and the risk of future coronary artery disease in apparently healthy men and women. Mannose binding lectin is a part of the complement cascade, playing an important role in the first line defense against pathogenic micro-organisms.3 We conducted a nested case-control study in a large cohort of healthy men and women. Cases were people in whom fatal or nonfatal coronary artery disease developed during follow-up. Controls were matched by age, sex, and enrolment time. Our results showed that high levels of mannose-binding lectin are associated with an increased risk of future coronary artery disease in healthy men, but not in women. Possible explanation for the sex differences with regard to mannose binding lectin levels and cardiovascular risk have been reported before and are possible related to differences in endocrine status and differences in the expression of complement components in adipose tissue between men and women. Another factor of innate immunity is secretory phospholipase A2. The protein secretory phospholipase A2 modifies lipoproteins in the arterial wall making the lipoprotein more readily interactive with proteoglycans. This prolongs the residence time of the lipoprotein in the arterial wall and increases the change of pro-atherogenic modification. In addition, secretory phospholipase A2 hydrolyses phospholipid precursors of pro-inflammatory mediators, and lipoproteins are more susceptible for oxidation when secretory phos-pholipase A2 is present. In chapter 4, we determined the relationship between secretory phospholipase A2 and the risk of future coronary artery disease in apparently healthy men and women. We showed in a nested case-control study in a large cohort of healthy

Summary

192

men and women that secretory phospholipase A2 levels were higher in cases than in controls and secretory phospholipase A2 plasma levels correlated with age, body mass index, systolic blood pressure, high-density lipoprotein cholesterol levels, and C-reactive protein levels. After adjustment, the risk of future coronary artery disease was increased for people in the highest secretory phospholipase A2 quartile, compared with those in the lowest. Thus, elevated levels of secretory phospholipase A2 were associated with an increased risk of future coronary artery disease in apparently healthy individuals. Hemostatis plays a key role in the development of cardiovascular disease. It has been shown that bleeding is important for plaque progression and subsequent rupture.4, 5 Thrombus formation on previous established atherosclerotic plaques is causative for acute coronary heart disease, and tissue factor, the main initiator of coagulation activa-tion has been implicated in coronary artery disease. Tissue factor expression is increased on plaques of patients with a high risk of rupture,6 and high levels of circulating tissue factor are found in patients with acute atherothrombotic events.7 In chapter 5 the risk of coronary heart disease is assessed in individuals with high levels of tissue factor, the main initiator of coagulation. From the data obtained we conclude that high levels of serum tissue factor were not independently associated with an increased risk of future coronary artery disease in apparently healthy individuals.Further study of disturbed innate immunity as risk factor for coronary heart disease was described in chapter 6. We determined whether polymorphism of the toll-like receptor 2 gene is related to an increased risk of developing acute coronary syndromes. Because toll-like receptor 2 has been suggested to be involved in atherogenesis, via microbial infec-tions or host-derived endogenous toll-like receptor 2 agonists, we expected a decreased risk of acute myocardial infarction and unstable angina pectoris in patients with polymor-phism in the toll-like receptor 2 gene. Compared with wild-type carriers, patients with the Arg753Gl toll-like receptor 2 genotype had no decreased risk of acute coronary heart disease. Thus we conclude that this polymorphism is not associated with a decreased risk of acute coronary heart disease.

Part 2: Infection and (athero-) thrombotic disease

Recently it has been shown that the number of chronic infections (the “pathogen bur-den”) is independently related to the risk of myocardial infarction8 and that acute respi-ratory tract infection temporary increase the risk on acute coronary heart disease.9, 10

Spodick et al. demonstrated an association between acute respiratory symptoms and the onset of acute coronary heart disease.11 To further explore these findings we performed a prospective observational study in a Dutch coronary care unit, described in chapter 7. We investigated whether patients with acute ischemic heart disease have more influenza infections prior to the ischemic event than stable control patients, using serologic tests and a questionnaire for influenza-like signs and symptoms. We were unable to prove an association between influenza infection and acute coronary heart disease, but interest-ingly, we found that influenza vaccination protects against the development of acute coro-nary heart disease. These results correspond to a previous report that showed a protective effect of influenza vaccination against coronary heart disease.12 To establish our findings

Summary

193

Summary

we searched for all available randomized controlled trials on influenza vaccination with data on coronary heart disease. In this study described in chapter 8 we found that insuf-ficient data from randomized controlled trials are published to prove that influenza vac-cination protects against coronary heart disease. To establish whether respiratory tract infection pose an additional risk on pulmonary embolism we performed a study in patients with pulmonary embolism and assessed the number of respiratory tract infections in the days prior to the event. This study is described in chapter 9. We were unable to show an association between influenza infec-tion and the occurrence of pulmonary embolism and we found that symptoms and signs of influenza are highly non-specific in this clinical setting. Thus we conclude that influ-enza infection is not associated with an increased risk of acute pulmonary embolism.To further investigate the role of influenza in the etiology of coronary heart disease, we studied the effects of influenza on inflammatory cells derived from unstable atheroscle-rotic plaques in vitro (chapter 10). Previous experimental studies showed that infection of apoE deficient mice with influenza resulted in an acute arterial wall inflammatory response,13 composed predominantly of macrophages and T lymphocytes, suggesting that influenza may trigger directly plaque inflammation. Therefore, we tested the hypothesis that influenza specific T cells contribute to the process of atherosclerotic plaque inflam-mation which may initiate plaque rupture. We found that atherosclerotic plaque derived T cells proliferate after exposure to influenza in vitro. These data suggest that T cells in human atherosclerotic plaques could be activated during influenza infection. Alternative, influenza derived antigens, or mimicry antigens, appear potential candidates for trigger-ing plaque inflammation, eventually leading to plaque rupture.

Part 3: (Acute respiratory tract) infections and coagulation

A prothrombotic state is associated with increased risk on vascular disease. Venous thrombosis is usually the result of either pro-coagulant alterations of the blood, or ves-sel wall, eventually in combination with reduced flow. Recently it has been shown that respiratory tract infections are associated with coagulation activation and alteration of the vessel wall.14, 15 The increased risk of respiratory tract infections on ischemic heart disease may be associated with increased levels of hemostatic proteins,16, 17 although, evidence of hyper-coagulability as a risk factor for coronary events is limited. It has been suggested that individuals with high plasma von Willebrand Factor, prothrombin fragment 1+2, D-dimer or plasmin-anti plasmin have increased risk of future ischemic heart disease.18-20 Previous experimental research have shown that respiratory tract infections are able to activate coagulation in vitro. We aimed to investigate the effects of acute respiratory tract infections on coagulation and fibrinolysis in humans. The purpose of the study, described in chapter 11, was to determine the effect of acute respiratory tract infections on hemo-static proteins in the elderly people. In a prospective study, a cohort of elderly people were followed during two consecutive winters. In case of flu-like symptoms we searched for causative pathogens and measured hemostatic proteins. Our data showed that common acute respiratory tract infections increase levels of hemostatic proteins, influencing in particular the endothelium. This finding points towards a possible link between acute

194

respiratory tract infections, endothelium cell perturbation and the increased risk of acute ischemic heart disease.The relation between viral infections and coagulation is further investigated in a study described in chapter 12 in which hemostasis was analyzed in patients with cytomegalo-virus infections. The results showed that cytomegalovirus infection is associated with levels of prothrombin fragment 1+2, plasmin-anti-plasmin, D-dimer and Von Willebrand Factor and we demonstrated that levels of these hemostatic proteins are correlated with the cytomegalovirus load. Thus, in vivo cytomegalovirus is associated with coagulation activation and a prothrombotic state. Furthermore, our data show that cytomegalovirus is associated with endothelial cell perturbation as evidenced by increased levels of |Von Willebrand factor. Whether these changes increase the risk of vascular disease remains subjects of further study.Finally, to determine mechanisms involved in the prothrombotic state we performed experimental studies in mice described in chapter 13. In mice infected with influenza, we showed activation of coagulation, increased thrombin generation, fibrin deposition and increased fibrinolysis. In addition, we used various anti- and prothrombotic mice to study pathways involved in the influenza-induced prothrombotic state. In TMpro/pro mice (with a reduced capacity to generate activated protein C) influenza increased thrombin generation and fibrinolysis. In influenza-infected anti-thrombotic mice treated with heparin, thrombin generation was reduced. In hyperfibrinolytic mice (plasminogen acti-vator inhibitor type-1-/-), thrombin generation was not changed, however, increased fibrin degradation was seen. Finally, treatment with tranexamic acid reduced fibrinolysis, but thrombin generation was not changed. We conclude that influenza generates thrombin, a process that can be increased by reduced levels of protein C and decreased by heparin. The fibrinolytic system appears not to affect thrombin generation in this model. Taken together, influenza may lead to a prothrombotic state by coagulation activation, which can be reversed by heparin treatment.

Summary

195

Summary

Reference List 1. Lucas AR, Korol R, Pepine CJ. Inflammation in atherosclerosis: some thoughts about acute coronary syn dromes. Circulation 2006;113:e728-e732. 2. Hansson GK, Libby P, Schonbeck U et al. Innate and adaptive immunity in the pathogenesis of atheroscle rosis. Circ Res 2002;91:281-91. 3. Gadjeva M, Takahashi K, Thiel S. Mannan-binding lectin-a soluble pattern recognition molecule. Mol Immunol 2004;41:113-21. 4. Kolodgie FD, Gold HK, Burke AP et al. Intraplaque hemorrhage and progression of coronary atheroma. N Engl J Med 2003;349:2316-25. 5. Takaya N, Yuan C, Chu B et al. Presence of intraplaque hemorrhage stimulates progression of carotid atherosclerotic plaques: a high-resolution magnetic resonance imaging study. Circulation 2005;111:2768-75. 6. Ardissino D, Merlini PA, Arlens R et al. Tissue factor in human coronary atherosclerotic plaques. Clin Chim Acta 2000;291:235-40. 7. Misumi K, Ogawa H, Yasue H et al. Comparison of plasma tissue factor levels in unstable and stable angina pectoris. Am J Cardiol 1998;81:22-6. 8. Zhu J, Quyyumi AA, Norman JE et al. Effects of total pathogen burden on coronary artery disease risk and C-reactive protein levels. Am J Cardiol 2000;85:140-6. 9. Smeeth L, Cook C, Thomas S et al. Risk of deep vein thrombosis and pulmonary embolism after acute infection in a community setting. Lancet 2006;367:1075-9. 10. Smeeth L, Thomas SL, Hall AJ et al. Risk of myocardial infarction and stroke after acute infection or vaccination. N Engl J Med 2004;351:2611-8. 11. Spodick DH, Flessas AP, Johnson MM. Association of acute respiratory symptoms with onset of acute myocardial infarction: prospective investigation of 150 consecutive patients and matched control patients. Am J Cardiol 1984;53:481-2. 12. Nichol KL, Nordin J, Mullooly J et al. Influenza vaccination and reduction in hospitalizations for cardiac disease and stroke among the elderly. N Engl J Med 2003;348:1322-32. 13. Naghavi M, Wyde P, Litovsky S et al. Influenza infection exerts prominent inflammatory and thrombotic effects on the atherosclerotic plaques of apolipoprotein E-deficient mice. Circulation 2003;107:762-8. 14. Bouwman JJ, Visseren FL, Bosch MC et al. Procoagulant and inflammatory response of virus-infected monocytes. Eur J Clin Invest 2002;32:759-66. 15. Visseren FL, Bouwman JJ, Bouter KP et al. Procoagulant activity of endothelial cells after infection with respiratory viruses. Thromb Haemost 2000;84:319-24. 16. Feinbloom D, Bauer KA. Assessment of hemostatic risk factors in predicting arterial thrombotic events. Arterioscler Thromb Vasc Biol 2005;25:2043-53. 17. Yarnell J, McCrum E, Rumley A et al. Association of European population levels of thrombotic and inflammatory factors with risk of coronary heart disease: the MONICA Optional Haemostasis Study. Eur Heart J 2005;26:332-42. 18. Morange PE, Simon C, Alessi MC et al. Endothelial cell markers and the risk of coronary heart disease: the Prospective Epidemiological Study of Myocardial Infarction (PRIME) study. Circulation 2004;109:1343-8. 19. Miller GJ, Bauer KA, Barzegar S et al. Increased activation of the haemostatic system in men at high risk of fatal coronary heart disease. Thromb Haemost 1996;75:767-71. 20. Cushman M, Lemaitre RN, Kuller LH et al. Fibrinolytic activation markers predict myocardial infarction in the elderly. The Cardiovascular Health Study. Arterioscler Thromb Vasc Biol 1999;19:493-8.

samenvattIng

Samenvatting

198

Samenvatting

199

SAMENvATTINgDit proefschrift beschrijft onderzoek naar de rol van infectie en ontsteking bij het ontstaan van veneuze en arteriële trombose. Het eerste deel beschrijft het effect dat ontsteking en bloedstolling hebben op het ontstaan van coronairlijden. Het tweede deel beschrijft de relatie tussen acute luchtweginfecties en veneuze trombose, atherosclerose en coronair-lijden. Het derde deel beschrijft studies naar de pathofysiologie van de door infectie geïn-duceerde protrombotische staat.

Deel 1: Ontsteking en bloedstolling als risico factoren voor coronairlijdenHart en vaatziekten zijn een uiting van atherosclerose, waardoor de toevoer van zuur-stof aan weefsel tekort schiet. Het werkelijke tekort schieten van atherosclerotische vaten wordt vaak getriggered door zogenaamde plaque-instabiliteit, waarbij (vaak door ontste-king) de atherosclerotische plaque die in de vaatwand is gevormd scheurt en trombusvor-ming plaats kan vinden. Centraal in de pathogenese van atherosclerose staat de afzetting van lipoproteïnen (vetachtige stoffen) en ontstekingsactiviteit in de vaatwand.1 In de laat-ste jaren is veel onderzoek gedaan naar het effect van oxidatie en ontsteking op lipoprote-inen (en vice versa) en het ontstaan van hart- en vaatziekten. Naast ontstekingsprocessen speelt ook de bloedstolling een rol bij de progressie van atherosclerotische plaques en de complicaties van atherosclerose. Ondanks de enorme toename van de pathofysiologische kennis van de atherogenese, blijven details van dit proces veelal nog onopgehelderd. Recent is een associatie aangetoond tussen de aangeboren, niet-specifieke afweer en het ontstaan van atherosclerose.2 Daarom onderzochten wij de relatie tussen mannose-bin-ding lectin en secretory phospholipase A2, beide component van de niet-specifieke afweer, en het risico op coronairlijden. Mannose-binding lectin is een eiwit dat specifiek bindt aan mannose en andere suikers en is onderdeel van de lectine route van het complement systeem.3 Dit systeem is betrokken bij de verdediging tegen micro-organismen. In een case-control studie, uitgevoerd in een cohort van gezonde mannen en vrouwen, onderzochten wij de relatie tussen mannose-binding lectin en het risico op coronairlijden. Personen die tijdens de follow-up coronair-lijden ontwikkelden werden vergeleken met personen zonder coronairlijden. Door mat-ching werd gecorrigeerd voor leeftijd, geslacht en de duur van de follow-up. Dit onderzoek toonde aan dat bij mannen in het hoogste mannose-binding lectin kwartiel het risico op coronairlijden is verhoogd. Bij vrouwen werd deze relatie niet gevonden. Een mogelijke verklaring ligt in de hormonale fluctuaties bij vrouwen die de expressie van componenten van het complement systeem kunnen beïnvloeden. Het eiwit secretory phospholipase A2 is pro-atherogeen door een aantal eigenschappen. Ten eerste modificeert het eiwit in de vaatwand lipoproteïnen waardoor de lipoproteï-nen beter aan proteoglycanen binden. Hierdoor neemt de tijd dat lipoproteïnen in de vaatwand verblijven toe, waardoor de kans op pro-atherogene modificatie wordt vergroot. Ten tweede hydrolyseert secretory phospholipase A2 voorlopers van ontstekingsmedia-toren. Ten derde worden lipoproteïnen makkelijker geoxideerd in de aanwezigheid van secretory phospholipase A2. In hoofdstuk 4, onderzochten wij de relatie tussen secretory phospholipase A2 en het risico op coronairlijden in een case-control studie. In mannen

en vrouwen was het risico op het ontwikkelen van coronairlijden verhoogd in de groep in het hoogste secretory phospholipase A2 kwartiel vergeleken met mensen in het laagste kwartiel. Deze relatie bleef significant na correctie voor klassieke risico factoren en C-reactive protein. Verhoogde secretory phospholipase A2 waarden in het bloed zijn dus geassocieerd met een verhoogd risico op coronairlijden in de toekomst. Hemostase speelt een belangrijke rol bij het ontstaan van hart en vaatziekten. Enerzijds zorgen bloedingen in de vaatwand voor progressie van atherosclerose, waardoor de kans op plaque ruptuur wordt vergroot.4,5 Anderzijds leidt stolselvorming op geruptureerde plaques tot de klinische manifestaties van atherosclerose. Bij het ontstaan van stolsels speelt tissue factor, de belangrijkste initiator van de bloedstolling, een belangrijke rol. Tis-sue factor wordt ook in verband gebracht met coronairlijden. Hoofdstuk 5 beschrijft een studie naar de relatie tussen verhoogde tissue factor waarden in het bloed en het ontstaan van coronairlijden. De resultaten toonden aan dat de aanwezigheid van tissue factor in het serum van mensen niet gecorreleerd is met het risico op coronairlijden. Hierbij dient te worden aangetekend dat de rol van meetbaar tissue factor in serum ter discussie staat en dat tissue factor aan het begin van de follow-up periode werd gemeten jaren voordat de ziekte manifest werd. Hoofdstuk 6 beschrijft een case-control studie naar de associatie tussen een polymor-fisme in het toll-like receptor 2 gen en een mogelijk verhoogd risico op het ontwikke-len van acute coronair syndromen. Toll-like receptor 2 zou een rol spelen bij het ont-staan van atherosclerose doordat infecties en endogene factoren via deze receptor de ontstekingsreactie in atherosclerotische plaques onderhouden. De hypothese was dat het Arg753Gl polymorfisme in het toll-like receptor 2 gen het risico op acute coronair syndromen zou verminderen. Echter, zowel in cases als in de controle patiënten was het aantal patiënten met toll-like receptor 2 Arg753Gl gelijk. Het bestudeerde polymor-fisme in het toll-like receptor 2 gen is dus niet geassocieerd met een veranderd risico op coronair syndromen.

Deel 2: Infecties en (athero-) trombotische aandoeningenDe opstelsom van het aantal chronische infecties dat iemand doormaakt (de “pathogen burden”) is geassocieerd met het ontstaan van coronairlijden.8 Daarnaast verhogen acute (luchtweg)infecties het risico op veneuze trombose en hart- en vaatziekten.9,10 Patiënten die in het ziekenhuis worden opgenomen wegens een acuut myocardinfarct hebben in de dagen voor opname vaker klachten die passen bij een acute luchtweginfectie dan contrôle patiënten. Om deze bevinding nader te onderzoeken voerden wij een prospectieve case-control studie uit beschreven in hoofdstuk 7, waarbij wij gebruik maakten van vragenlijs-ten en bloed onderzoek gericht op een mogelijke recent doorgemaakte influenza infectie. Wij konden in deze studie geen relatie aantonen tussen influenza en het ontstaan van acute coronair syndromen. Wel bleek influenza vaccinatie geassocieerd met een vermin-derd risico op acute coronair syndromen. De beschermende werking van influenza vacci-natie op het ontstaan van hart- en vaatziekten werd ook in eerder onderzoek aangetoond.12 Om deze bevinding te bevestigen werd systematisch naar alle beschikbare gerandomi-seerde studies gezocht die het effect van influenza vaccinatie op coronairlijden hebben

Samenvatting

200

onderzocht. Uit deze systematische review, beschreven in hoofdstuk 8, blijkt dat er tot op heden onvoldoende gerandomiseerde studies zijn gedaan om te bewijzen dat influenza vaccinatie ook daadwerkelijk beschermt tegen het optreden van coronairlijden. Hoofdstuk 9 beschrijft een studie waarin werd onderzocht of luchtweginfecties het risico op longembolieën vergroten. In een cohort patiënten verdacht van een longembolie, wer-den patiënten met en zonder een longembolie vergeleken. Gezocht werd naar een verschil tussen beide groepen in het voorkomen van een luchtweginfectie en serologisch bewijs voor een doorgemaakte luchtweginfectie. Een associatie tussen een luchtweginfectie en longembolieën werd niet aangetoond. Tevens bleken klachten van een luchtweginfectie niet van diagnostische waarde voor het inschatten van het risico op een longembolie. Om de rol van influenza in de pathogenese arterieel vaatlijden verder te onderzoeken, hebben wij het effect bestudeerd van influenza op ontstekingscellen uit atheroscleroti-sche plaques (hoofdstuk 10). Eerder onderzoek had aangetoond dat influenza in dierex-perimentele modellen effect had op de samenstelling van atherosclerotische plaques.13 Muizen die waren geïnfecteerd met influenza vertoonden een toename van macrofagen en T lymfocyten in de vaatwand. Wij onderzochten in een studie met humaan materiaal of T lymfocyten uit instabiele atherosclerotische plaques influenza specifiek zijn. In dit onderzoek toonden wij inderdaad aan dat deze T lymfocyten een specifieke proliferatie respons vertonen na incubatie met influenza. De bevindingen suggereren dat T lymfocy-ten geactiveerd kunnen raken tijdens een influenza infectie en zo kunnen bijdragen aan een verhoogd risico op atherosclerotische plaque ruptuur.

Deel 3: (Acute luchtweg) infecties en de bloedstollingVeneuze trombose wordt veroorzaakt door de trias van een veranderde vaatwand, ver-minderde bloeddoorstroming en bloedstollingsactivatie. In experimenteel onderzoek is aangetoond dat pathogenen die luchtweginfecties veroorzaken kunnen zorgen voor stol-lingsactivatie,14, 15 en dat luchtweg infectie geinduceerde stollingsactivatie mogelijk ook het verhoogde risico op hart- en vaatziekten zou kunnen verklaren.16, 17 Enkele studies toonden aan dat mensen met hoge plasma spiegels van eiwitten die betrokken zijn bij de bloedstolling een verhoogd risico hebben op ischemische hartziekten.18-20 De resultaten van de studies naar deze relatie zijn echter niet consistent. In hoofdstuk 11 wordt een studie beschreven waarin het effect van luchtweginfecties op de bloedstolling werd onderzocht bij verder gezonde personen. De resultaten toonden aan dat luchtweginfecties de bloedspiegels van eiwitten die betrokken zijn bij de stolling ver-hogen, waarbij tevens veranderingen in het endotheel (vaatwand) werden waargenomen. Deze bevindingen wijzen op endotheliale disfunctie als mogelijk mechanisme van het door luchtweginfecties veroorzaakte risico op hart- en vaatziekten. De relatie tussen luchtweginfectie en bloedstolling werd verder onderzocht in de studie beschreven in hoofdstuk 12. In deze studie werd de bloedstolling gemeten in patiënten met een (re)activatie van cytomegalovirus. De resultaten toonden aan dat de infectie geas-socieerd was met verhoogde spiegels van prothrombine fragmenten 1+2, plasmine-anti-plasmine, D-dimeren en Von Willebrand Factor in het bloed. Daarbij was de hoogte van deze eiwitten gecorreleerd aan de hoeveelheid cytomegalovirus in het bloed. Hieruit kun-

Samenvatting

201

nen wij concluderen dat cytomegalovirus geassocieerd is met stollingsactivatie en en-dotheliale dysfunctie. Of deze veranderingen ook het risico op vasculaire aandoeningen vergroten dient nader te worden onderzocht.Tot slot wordt in hoofstuk 13 experimenteel onderzoek beschreven naar de mechanis-men van de door influenza geïnduceerde stollingsactivatie. Dit onderzoek toonde aan dat influenza in muizen leidt tot stollingsactivatie, verhoogde trombine vorming en fibrine depositie en een toename van de fibrinolyse (afbraak van stolsels). Daarnaast bleek dat in muizen met minder geactiveerd proteine C (TMpro/pro) meer trombine werd gevormd en dat de fibrinolyse verder was toegenomen tijdens influenza. De behandeling met he-parine verminderde de door influenza geinduceerde stollingsactivatie. Muizen met een verhoogde fibrinolytische activiteit (plasminogen activator inhibitor type-1-/-) vertoonden geen veranderde trombine generatie. Als laatste leidde de behandeling met tranexamine-zuur, een fibrinolyse remmer, niet tot een verandering in de trombine generatie. Uit deze data kan worden geconcludeerd dat influenza leidt tot een protrombotische staat doordat de bloedstolling wordt geactiveerd. Deze influenza-geïnduceerde protrombotische staat kan door toediening van heparine worden verminderd.

Samenvatting

202

Referenties 1. Lucas AR, Korol R, Pepine CJ. Inflammation in atherosclerosis: some thoughts about acute coronary syndromes. Circulation 2006;113:e728-e732. 2. Hansson GK, Libby P, Schonbeck U et al. Innate and adaptive immunity in the pathogenesis of atherosclerosis. Circ Res 2002;91:281-91. 3. Gadjeva M, Takahashi K, Thiel S. Mannan-binding lectin-a soluble pattern recognition molecule. Mol Immunol 2004;41:113-21. 4. Kolodgie FD, Gold HK, Burke AP et al. Intraplaque hemorrhage and progression of coronary atheroma. N Engl J Med 2003;349:2316-25. 5. Takaya N, Yuan C, Chu B et al. Presence of intraplaque hemorrhage stimulates progression of carotid atherosclerotic plaques: a high-resolution magnetic resonance imaging study. Circulation 2005;111:2768-75. 6. Ardissino D, Merlini PA, Arlens R et al. Tissue factor in human coronary atherosclerotic plaques. Clin Chim Acta 2000;291:235-40. 7. Misumi K, Ogawa H, Yasue H et al. Comparison of plasma tissue factor levels in unstable and stable angina pectoris. Am J Cardiol 1998;81:22-6. 8. Zhu J, Quyyumi AA, Norman JE et al. Effects of total pathogen burden on coronary artery disease risk and C-reactive protein levels. Am J Cardiol 2000;85:140-6. 9. Smeeth L, Cook C, Thomas S et al. Risk of deep vein thrombosis and pulmonary embolism after acute infection in a community setting. Lancet 2006;367:1075-9. 10. Smeeth L, Thomas SL, Hall AJ et al. Risk of myocardial infarction and stroke after acute infection or vaccination. N Engl J Med 2004;351:2611-8. 11. Spodick DH, Flessas AP, Johnson MM. Association of acute respiratory symptoms with onset of acute myocardial infarction: prospective investigation of 150 consecutive patients and matched control patients. Am J Cardiol 1984;53:481-2. 12. Nichol KL, Nordin J, Mullooly J et al. Influenza vaccination and reduction in hospitalizations for cardiac disease and stroke among the elderly. N Engl J Med 2003;348:1322-32. 13. Naghavi M, Wyde P, Litovsky S et al. Influenza infection exerts prominent inflammatory and thrombotic effects on the atherosclerotic plaques of apolipoprotein E-deficient mice. Circulation 2003;107:762-8. 14. Bouwman JJ, Visseren FL, Bosch MC et al. Procoagulant and inflammatory response of virus-infected monocytes. Eur J Clin Invest 2002;32:759-66. 15. Visseren FL, Bouwman JJ, Bouter KP et al. Procoagulant activity of endothelial cells after infection with respiratory viruses. Thromb Haemost 2000;84:319-24. 16. Feinbloom D, Bauer KA. Assessment of hemostatic risk factors in predicting arterial thrombotic events. Arterioscler Thromb Vasc Biol 2005;25:2043-53. 17. Yarnell J, McCrum E, Rumley A et al. Association of European population levels of thrombotic and inflammatory factors with risk of coronary heart disease: the MONICA Optional Haemostasis Study. Eur Heart J 2005;26:332-42. 18. Morange PE, Simon C, Alessi MC et al. Endothelial cell markers and the risk of coronary heart disease: the Prospective Epidemiological Study of Myocardial Infarction (PRIME) study. Circulation 2004;109:1343-8. 19. Miller GJ, Bauer KA, Barzegar S et al. Increased activation of the haemostatic system in men at high risk of fatal coronary heart disease. Thromb Haemost 1996;75:767-71. 20. Cushman M, Lemaitre RN, Kuller LH et al. Fibrinolytic activation markers predict myocardial infarction in the elderly. The Cardiovascular Health Study. Arterioscler Thromb Vasc Biol 1999;19:493-8.

Samenvatting

203

Samenvatting

205

lIst ofpublICatIons


206

207

LIST OF PuBLICATIONS

Keller TT, Choi D, Te Velthuis H, Gerdes VE, Wareham NJ, Bingham SA, Luben R, Hack CE, Reitsma PH, Levi M, Khaw KT, Boekholdt SM. Tissue factor serum levels and the risk of future coronary artery disease in ap-parently healthy men and women; the EPIC-Norfolk prospective population study. J Thromb Haemost. 2006 (in press)

Keller TT, van Leuven SI, Meuwese MC, Wareham NJ, Luben R, Stroes ES, Hack CE, Levi M, Khaw KT, Boekholdt SM. Serum Levels of Mannose-Binding Lectin and the Risk of Future Coronary Artery Disease in Apparently Healthy Men and Women. Arterioscler Thromb Vasc Biol 2006 oct;26(10):2345-50.

Bisoendial R, Birjmohun R, Keller T, van Leuven S, Levels H, Levi M, Kastelein J, Stroes E. In vivo effects of C-reactive protein (CRP)-infusion into humans. Circ Res. 2005 Dec 9;97(12):e115-6.

Keller TT, Mairuhu AT, Gerdes VE, Brandjes DP, Peters RJ, van Gorp EC. Acute luchtweginfecties en acute coronaire syndromen. Ned Tijdschr Geneeskd. 2005 Jun 4;149(23):1267-72.

Boekholdt SM, Keller TT, Wareham NJ, Luben R, Bingham SA, Day NE, Sandhu MS, Jukema JW, Kastelein JJ, Hack CE, Khaw KT. Serum levels of type II secretory phospholipase A2 and the risk of future coronary artery disease in apparently healthy men and women: the EPIC-Norfolk Prospective population Study. Arterioscler Thromb Vasc Biol 2005 Apr;25(4):839-46. Epub 2005 Feb 3.

Klein SK, Slim EJ, de Kruif MD, Keller TT, ten Cate H, van Gorp EC, Brandjes DP. Is chronic HIV infection associ-ated with venous thrombotic disease? A systematic review. Neth J Med. 2005 Apr;63(4):129-36.

de Kruif MD, van Gorp EC, Keller TT, Ossewaarde JM, ten Cate H. Chlamydia pneumoniae infections in mouse models: relevance for atherosclerosis research. Cardiovasc Res. 2005 Feb 1;65(2):317-27.

Keller TT, Squizzato A, Weeda VB, Middeldorp S. Clopidogrel and aspirin versus aspirin for preventing cardio-vascular disease in those at high risk. (Protocol) Cochrane Database of Systematic Reviews 2005, Issue 1

Keller TT, Mairuhu ATA, Brandjes DPM, Levi M. Influenza vaccination to prevent myocardial infarction. (Pro-tocol) Cochrane Database of Systematic Reviews 2004, Issue 4

Levi M, Bijsterveld NR, Keller TT. Recombinant factor VIIa as an antidote for anticoagulant treatment. Semin Hematol. 2004 Jan;41(1 Suppl 1):65-9.

Keller TT, Mairuhu AT, de Kruif MD, Klein SK, Gerdes VE, ten Cate H, Brandjes DP, Levi M, van Gorp EC. Infec-tions and endothelial cells. Cardiovasc Res. 2003 Oct 15;60(1):40-8.

Levi M, Keller TT, van Gorp E, ten Cate H. Infection and inflammation and the coagulation system. Cardiovasc Res. 2003 Oct 15;60(1):26-39. Review.

Friederich PW, Henny CP, Messelink EJ, Geerdink MG, Keller T, Kurth KH, Buller HR, Levi M. Effect of recom-binant activated factor VII on perioperative blood loss in patients undergoing retropubic prostatectomy: a double-blind placebo-controlled randomised trial. Lancet. 2003 Jan 18;361(9353):201-5.

Friederich PW, Levi M, Bauer KA, Vlasuk GP, Rote WE, Breederveld D, Keller T, Spataro M, Barzegar S, Buller HR. Ability of recombinant factor VIIa to generate thrombin during inhibition of tissue factor in human sub-jects. Circulation. 2001 May 29;103(21):2555-9.

Friederich PW, Keller TT, Biemond BJ, Peters RJ, Hornberger W, Buller HR, Levi M. Successful attenuation of venous thrombus growth in rabbits after the administration of a novel oral thrombin inhibitor. Thromb Haemost. 2000 Nov;84(5):858-64.



209

dankwoord

210

Dankwoord

Ik had dit proefschrift niet kunnen schrijven zonder de directe hulp van de volgende personen.

Mijn promotoren Prof.dr. H.R. Büller en Prof.dr. M.M. Levi. Beste Harry, beste Marcel, vanaf de eerste dag dat ik gebogen over 12 New Zealand whites, betrokken was bij het onderzoek van de afdeling Vasculaire Geneeskunde, heb ik mij door jullie vertrouwen gesteund geweten. Nog steeds bewonder ik de vloeiende stijl waarmee jullie je door de arena van de wetenschap bewegen. Met recht ben ik trots op de geboden mogelijkheid om onder jullie supervisie aan mijn promotie te hebben mogen werken.

Mijn co-promotores Dr. D.P.M. Brandjes en Dr. V.E.A. Gerdes. Beste Dees en beste Victor, de vruchtbare samenwerking van de afdeling Vasculaire Geneeskunde met de Interne Geneeskunde van het Slotervaart Ziekenhuis is de basis geweest voor dit proefschrift. Jullie inbreng leidde tot succesvolle studies. Visie, geestdrift en bevlogenheid kenmerkt de een, geduld en doordachtzaamheid kenmerkt de ander. Ik dank jullie zeer!

Dr. E.C.M. van Gorp. Beste Eric jouw uitstekende netwerk is van groot belang geweest voor het realiseren van veel projecten beschreven in dit proefschrift. Daarnaast staan aan de basis van dit proefschrift twee artikelen die ik niet had kunnen schrijven zonder jouw visie en geduldige begeleiding.

Dr. J.C.M. Meijers. Beste Joost, als onmisbare kracht achter talloze projecten neem je een bijzondere positie in. Met je enorme kennis van de hemostase symboliseer jij de fundamenteel wetenschappelijke kant van dit proefschrift. Geen manuscript kwam niet rood van de correcties terug en iedere vrijpostige conclusie werd van kritisch commentaar voorzien. Helaas kon ik slechts weinig daarvan weerleggen!

Dr. A.T.A. Mairuhu. Beste Ronne, niet alleen zou de Overveense studie niet zijn uitgevoerd zonder jouw tomeloze inzet, ook in veel ander opzicht ben ik je dank verschuldigd. Je enthousi-asme en werklust zijn ongeëvenaard. De rallies over de A9 zullen mij altijd bijblijven. Hoewel ik nooit iets over snelheidovertredingen heb vernomen moeten de boetes talrijk zijn geweest.

Drs. M. van Wissen. Beste Matthijs, als mijn opvolger bij het illustere Influenza Gezelschap, wacht jou de grote taak om nu eens studies wel goed te laten lopen. Als jij ergens voor gaat wordt het ook goed gedaan: niet eerder werd de inclusie in de VIVID studie zo grondig aangepakt.

De leden van de promotiecommissie Prof.dr. H. ten Cate, Prof.dr. R.J.G. Peters, Prof.dr. P.H. Reitsma, Prof.dr. P. Speelman, Dr. S. Florquin en Dr. F.L.J Visseren dank ik voor het kritisch doornemen van het manuscript en voor hun bereidheid zitting te nemen in mijn promotiecommissie.

Op de afdeling Experimentele Inwendige Geneeskunde ben ik de volgende mensen dank verschuldigd voor de goede samenwerking: Prof.dr. T. van der Poll, Koen van der Sluijs,

211

Angelique Groot, Esther Vogels, Arnold Spek, Martijn de Kruif, Sjoukje Slofstra en Mark Dessing.

Op de afdeling Pathologie wil ik de volgende mensen danken: Onno de Boer, Allard van der Wal, Jelger van der Meer en Peter Teeling.

Van de afdeling Klinische Epidemiologie en Biostatistiek dank ik Prof.dr. A.H. Zwinderman, Prof.dr. P.M.M.Bossuyt, Barbara Hutten, Michael Tanck en Carlo van Dongen voor de goede adviezen.

Van de Immunologie-Reuma dank ik Prof.dr. R.J.M. ten Berge en van de Nierziekten Frederieke Bemelman voor de prettige samenwerking.

Hans Zaaijer van de Virologie dank ik voor de vele bepalingen die zijn uitgevoerd.

In het Slotervaart Ziekenhuis zijn Olga Ternede en Monique de Rijk onmisbaar geweest voor de patiëntenlogistiek en dank ik Rob Bakker voor de goede samenwerking.

Voor de samenwerking met het Erasmus Medisch Centrum wil ik Prof.dr. A.D. Osterhaus, Gerard Van Doornum en Guus Rimmelzwaan bedanken.

Op de afdeling Immunopathologie van Sanquin ben ik Prof.dr. C.E. Hack, Henk te Velthuis, Anke Eerenberg-Belmer en Gerard van Mierlo dank verschuldigd.

In Overveen dank ik Philip Sutorius en de medewerkers van de huisartspraktijk voor de prettige samenwerking.

Vanwege een (in)directe bijdrage aan dit proefschrift of vanwege de goede samenwerking op de afdeling dank ik de stafleden van de afdeling Vasculaire Geneeskunde, met in het bijzonder Prof.dr. J.J.P. Kastelein, de collega onderzoekers van de afdeling, de medewerk-ers van het Speciële Hemostase lab, de collega’s van de Experimentele Vasculaire Geneeskunde, de secretariële ondersteuning en de medewerkers van het trial bureau.

Secretariële ondersteuning werd verzorgd door Marianna Veendorp, Joyce Jansen, Agnes Vree en Coos Gosliga.

De lay-out van dit proefschrift werd verzorgd door Alex Klerk en Jasper van den Berg van Boem Ontwerpburo en Kees Visser zette de puntjes op de i.

Mijn broers Josbert en Hiske dank ik voor het aanvaarden van de taak als paranifm.

Ik dank mijn vader en moeder, Janchris en Ruth, voor de toegewijde zorg en de geboden mogelijkheden. Mijn dank hiervoor reikt verder dan een dankwoord beschrijven kan. Laura, jouw liefde is mijn inspiratie.

Dankwoord

uva-dare (digital academic repository) acute respiratory ... · flammatory reaction.22 mice...

Documents