university of groningen individual approach towards ... · university of groningen individual...

University of Groningen

Individual approach towards optimal oral anticoagulationVeeger, Nicolaas Johannes Gerardus Maria

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2010

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Veeger, N. J. G. M. (2010). Individual approach towards optimal oral anticoagulation. Groningen: s.n.

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

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 01-10-2020

https://www.rug.nl/research/portal/en/publications/individual-approach-towards-optimal-oral-anticoagulation(d3a6896a-cb04-4fdc-9996-04ce0fbac7da).html

https://www.rug.nl/research/portal/en/persons/nic-veeger(85d39ec2-0193-40a6-befa-f77a2a6ac186).html

https://www.rug.nl/research/portal/en/publications/individual-approach-towards-optimal-oral-anticoagulation(d3a6896a-cb04-4fdc-9996-04ce0fbac7da).html

Individual approach

towards

optimal oral anticoagulation

door

Nic Veeger

Financial support for the publication of this theses by the Trial Coordination

Center Groningen and the Department of Hematology of the University

Medical Center Groningen, Bielderman International Clinical Research for

Medical Devices, Federatie van Nederlandse Trombosediensten and Centrum

Oost Manuele therapie en Fysiotherapie Groningen is gratefully acknowledged.

Veeger, N.J.G.M.

Individual approach towards optimal oral anticoagulation

Thesis, University of Groningen, with summary in Dutch

© Copyright 2010 N.J.G.M. Veeger

ISBN 978-90-367-4184-2

Printed by: Facilitair Bedrijf RUG, Groningen

All rights reserved. No part of this publication may be reproduced, stored in a retrieval

system, or transmitted, in any form or by any means – electronic, mechanical, photocopy,

recording or otherwise - without prior written permission of the author.

RIJKSUNIVERSITEIT GRONINGEN

Individual approach

towards

optimal oral anticoagulation

Proefschrift

ter verkrijging van het doctoraat in de

Medische Wetenschappen

aan de Rijksuniversiteit Groningen

op gezag van de

Rector Magnificus, dr. F. Zwarts,

in het openbaar te verdedigen op

woensdag 17 februari 2010

om 14.45 uur

door

Nicolaas Johannes Gerardus Maria Veeger

geboren op 13 juni 1961

te Zwolle

Promotores: Prof. dr. J.C. Kluin-Nelemans

Prof. dr. J.L. Hillege

Beoordelingscommissie: Prof. dr. H.R. Büller

Prof. dr. H.J.G.M. Crijns

Prof. dr. M.H. Prins

Paranimfen: Arne Hemkes

Mirjam Veeger

To Jan Veeger

Jan van der Meer

TABLE OF CONTENTS

Chapter 1 9

General introduction and outline of this thesis

Chapter 2 23

Individual Time Within Target Range in Patients Treated with Vitamin

K Antagonists; Main Determinant of Quality of Anticoagulation and

Predictor of Clinical Outcome A Retrospective Study in 2300

Consecutive Patients with Venous Thromboembolism

British Journal of Haematology 2005, 128, 513–519

Chapter 3 41

Early Detection of Patients with a Poor Response to Vitamin K

Antagonists; The Clinical Impact of Individual Time Within Target

Range in Patients with Heart Disease

Journal of Thrombosis and Haemostasis 2006; 4: 1625–7

Chapter 4 59

Minor Bleeds Alert for Subsequent Major Bleeding in Patients Using

Vitamin K Antagonists

Submitted

Chapter 5 75

Quality of Anticoagulation Associated with Late Occurrence of Cardiac

Events after Coronary-Artery-Bypass-Graft Surgery in Patients Treated

with Vitamin K Antagonists; Fourteen Year Follow-up Results of

CABADAS

Submitted

Chapter 6 93

Vitamin K Antagonists or Aspirin Combined with Dipyridamole Not

Superior to Aspirin Alone; Fourteen Year Follow-up Results of

CABADAS

Submitted

Chapter 7 111

Excellent Long-Term Clinical Outcome after Coronary Artery Bypass Surgery

Using Three Pedicled Arterial Grafts in Patients with Three-Vessel Disease

Ann Thorac Surg 2008;85:508-12

Chapter 8 125

Summary, Discussion and Future Perspectives

Appendix I Popular Summary in Dutch 139

Appendix II Dankwoord 149

Chapter 1

- 9 -

Chapter 1

General Introduction and

Outline of this Thesis

General introduction

- 10 -

GENERAL INTRODUCTION In a variety of diseases, the use of antithrombotic agents and thrombolytic therapy for primary and secondary prevention of both arterial and venous thromboembolism have significantly influenced morbidity and mortality. This thesis addresses the use of vitamin K antagonists (VKA) in five of these diseases, i.e. deep vein thrombosis, pulmonary embolism, atrial fibrillation, heart valve replacement and coronary artery disease. The focus lies on the ability to achieve an optimal level of anticoagulation, in which arterial or venous thromboembolic events are prevented, without an unacceptable risk for major bleeding. Furthermore, in patients with coronary artery disease undergoing coronary-artery-bypass-grafting surgery, the choice of aspirin over vitamin K antagonists is re-evaluated. Optimal level of anticoagulation Vitamin K antagonists are used for a variety of indications, with different optimal ranges for different indications. Also, specific risk profiles (e.g. high risk for major bleeding) are weighted in selecting the optimal range for an individual patient. Currently, there is international consensus regarding the target range of INR (International Normalized Ratio) 2.0 to 3.0 as optimal for most diseases [1]. A higher target range might be more effective in secondary prevention after an acute myocardial infarction. However, this will be at the cost of an increased risk of major bleeding [2,3]. For prosthetic heart valves, the optimum depends on the type and position of the valve, and the presence of additional risk factors indicating a higher target range of INR 2.5 to 3.5 [4].

In the Netherlands, the target range is higher than internationally recommended (INR 2.5 to 3.5 and 3.0 to 4.0), with an additional wider therapeutic range (INR 2.0 to 3.5 and 2.5 to 4.0) to avoid particularly under-anticoagulation. These ranges are advocated by the Dutch Federation of Thrombosis Services, which unites a nationwide network of 61 specialized anticoagulation clinics responsible for the VKA management of all outpatients. In these specialized clinics, experienced physicians and dosing assistants are involved in optimizing VKA therapy, using computerized dosing programs and protocols for dose adjustments in case of under- and over-anticoagulation. In addition, patient education and awareness towards factors influencing the quality of VKA therapy are considered. Furthermore, patients are

Chapter 1

- 11 -

trained to perform self-management, which is done under guidance of the clinic. The quality of care that is achieved is evaluated on a yearly basis by the Dutch

Federation of Thrombosis Services, and presented in an annual report. In 2007, VKA therapy was managed in approximately 360000 patients. Over 5.5 million INR assessments were performed, with an median of 20 INRs per patient per year. The vitamin K antagonists used by these clinics were predominantly acenocoumarol (approximately 80%) and phenprocoumon (20%). When considering the different indications for which vitamin K antagonists were prescribed, 15% were venous and 85% arterial in origin, of which atrial fibrillation was the most frequent one, namely in 52% of all patients.

The quality indicators are the percentage of patients with an INR within the therapeutic range and the occurrence of major bleeds. For the percentage of patients within the therapeutic range, the threshold is set to 70% for long-term patients in the lower intensity group and 67% in higher intensity group. As presented in the annual report of 2007, this was achieved by 95% of the 61 anticoagulation clinics. Regarding major bleeding, annual incidences as observed in each clinic was compared with a reference incidence of 2% per year. In 2007, in 93% of the clinics the annual incidence did not exceed 2%. Furthermore, in only 10 of 61 clinics the annual incidence of major bleeding was above 1.5% (range 1.6 to 3.2%). Interacting factors in dose response relationship Next to the importance of the optimal range, in VKA therapy the ability to achieve a stable level of anticoagulation within this range is at least equally important. The dose-response relationship of vitamin K antagonists can be characterized by a high variability. Many environmental factors are known to play an important role in the achieved level of anticoagulation under VKA therapy, as well as genetic factors.

In reference to the genetic factors, both an increased sensitivity as well as an increased resistance to vitamin K antagonists are associated with CYP2C9 and VKORC1 genetic polymorphisms leading to a modified dose response relationship. Clinical outcome is more frequently compromised in these patients [5-9]. Environmental factors are concomitant drugs or herbal medicines, dietary intake of vitamin K, co-morbidity, alcohol intake as well as age and changes in activity level. These factors can result in an increase or a decrease of the effect of vitamin K antagonists [1,5,10-14]. Furthermore, the half-life time of the type of VKA also plays a


- 12 -

role in the achieved stability in the level of anticoagulation. With phenprocoumon and warfarin, with longer half-life times, a more stable level is achieved than with acenocoumarol [11,15]. However, due to this longer half-life, it is more difficult to counteract anticoagulation in case of high INRs or minor bleeding, as compared to acenocoumarol, for which a one-day interruption (with vitamin K only for higher INRs) results in a swift reduction of the level of anticoagulation. Finally, also compliance to VKA therapy regimens has been identified as an important factor of inter- and intra-individual variability [16,17].

To summarize, there are many factors known to influence the dose-response relationship of vitamin K antagonists. However, as Wittkowsky and Devine argued [18], still a large percentage of the instability observed can not be attributed to these known risk factors. In this respect, the occurrence of temporary over- and under-anticoagulation cannot be predicted nor prevented beforehand, and must be monitored closely. Possibly, the occurrence of such instability already early after initiation of VKA therapy, i.e. within the first month, could be used to identify those patients at higher risk for unstable anticoagulation resulting in a higher risk for both thromboembolic complications and major bleeding. INR monitoring During the course of VKA therapy, close monitoring of the achieved level of anticoagulation is essential to prevent both under- and over-anticoagulation. VKA therapy is most commonly monitored by using the prothrombin time (PT) test, were changes reflect the reduction of vitamin K-dependent procoagulant clotting factors II, VII and X. During the initial phase of VKA therapy, the PT mainly reflects the reduction of factor VII. Thereafter, factors II and X reduction contribute to prolongation of the PT [1]. Using the international normalized ratio (INR), PT test results are standardized and results comparable [19]. For this, the International Sensitivity Index (ISI) is used. It indicates the relative sensitivity of the thromboplastin compared to an international reference thromboplastin. Because these ISI values are based on stable anticoagulant doses for at least 6 weeks, the INR values are somewhat less reliable during the initial treatment phase. Time within target range With the given variability in dose response relationship and the strongly related clinical

Chapter 1

- 13 -

outcome, it is of great importance to use a valid methodology to measure and thereby safeguard the quality of VKA therapy. This quality can be expressed as the time within the target range (TTR). In a wide range of studies, the association between the TTR and clinical outcome has been established. Next to the time within the target range, this also holds for the two counterparts of TTR, i.e. the time spent below and above the target range as measures of relative under- and over-anticoagulation [20-24].

The TTR can be calculated in different ways, and as the different methodologies result in different TTR values, this is of importance when evaluating the quality of VKA therapy. TTR can be expressed as a percentage of the INR measurements within the target range, either compared to all measurements performed or only at a certain moment in time in a cross-sectional approach [25]. Also, the actual number of days spent within the target range can be used as the basis of the TTR, either using linear interpolation as proposed by Rosendaal or other estimation methods for the days between INR measurements (e.g. equidivision method) [26-28]. All methodologies have their own advantages and disadvantages. Rates based upon the number of INR measurements are simple to calculate, but are potentially biased by an increased frequency of measuring during a period of inadequate anticoagulation. The cross-sectional approach considers individual patients, but only at a certain moment in time. Linear interpolation, which includes time, can be used to evaluate the association between the level of anticoagulation and the occurrences of thromboembolic and bleeding complications (i.e. INR-specific incidence rates) [20]. In this method, a linear change of INR between consecutive measurements is assumed. Especially with longer intervals and extreme INR changes this might lead to a somewhat biased estimate of the TTR.

Recommending either one of these methods is not easy. In fact, each has its limitations and a choice may be based on local applicability [28]. However, when considering the use of TTR as a quality measurement in individual patient care, an approach that considers continuous changes in time is more appropriate. In this setting, the linear interpolation method as proposed by Rosendaal et al. is the method of choice [27]. Although the method has been advocated for its applicability in assessing INR-related incidences rates at a group level, it also makes it possible to assess the level of anticoagulation during the course of VKA therapy in an individual patient.


- 14 -

Evaluating individual course of INR through time During monitoring, frequent dose adjustments are made to achieve a level of anticoagulation within target range. When using computerized algorithms, these dose changes are based on an empiric dose-INR relationship. The effectiveness and safety of such computer-assisted dosage systems have been demonstrated and are comparable to experienced medical staff dosage [29]. However, considering the many factors influencing the achieved level of anticoagulation, a more individual multi-causal approach would do more justice to the complex inter- and intra-individual variability, and may lead to an increased quality of care. Although a more model-based approach with multiple factors as proposed by Pasterkamp et al. looks promising [30], the effectiveness of such algorithms must be evaluated in clinical trials.

When evaluating the achieved level of anticoagulation during VKA therapy, it is also important to realize that there still is a gap between guideline target ranges and acceptance of instability of response in daily practice. This is illustrated by a study by Palareti et al., where treating physicians considered many patients as stable despite a large amount of time spent above or below the target range, and thereby accepting INR values outside the target range as appropriate [11]. This observation is in line with the reported TTR values ranging from 50% to 75%, indicating that many patients are outside the target range for a considerable amount of time [1,20,21,26,31-35].

As part of the individual approach to monitoring, patients who are outside the target range for a considerable amount of time should be identified as early as possible. This is especially of importance as improving the patient’s awareness to the specific problems of VKA therapy can be a major determinant in increasing compliance and stability of the level of anticoagulation. Individual coaching of patients with regard to the proper use of vitamin K antagonists resulted in an increase in stability from 39% to 62% of time within the therapeutic range in a study [11]. In this respect, patient self-testing or self-management have shown to lead to an increase in the quality of VKA therapy [36-38]. However, not all patients are able or willing to perform all tasks required for a safe self-testing or self-management program.

Chapter 1

- 15 -

Aspirin versus vitamin K antagonists for secondary prophylaxis after coronary-artery-bypass-graft surgery Over the recent years, in the Netherlands the number of patients with atrial fibrillation requiring VKA therapy is increasing, and the number of patients with a myocardial infarction and arterial vascular surgery is steadily decreasing over the last years (Annual report 2007 Dutch Federation of Thrombosis Services). The latter is in line with the recommendations for antithrombotic therapy for coronary-artery-bypass-graft surgery (CABG) patients, as recently re-established in the 8th edition of the American College of Chest Physicians evidence-Based Clinical Practice Guidelines [39]. For all patients with coronary artery disease undergoing CABG, indefinite use of aspirin is recommended. Vitamin K antagonists should only be considered if concomitant conditions, e.g. heart valve replacement are present. In those instances, vitamin K antagonists combined with aspirin is suggested [39].

Several prospective controlled clinical trials have evaluated efficacy and safety of antithrombotic drug therapies for secondary prophylaxis after coronary-artery-bypass-graft surgery to prevent vein graft closure [40]. This resulted in a high level of evidence, identifying aspirin as standard of care [39,41]. To maximize the effect of anticoagulant treatment it must be initiated already early after surgery [42-44]. In this respect, current guidelines stipulate the initiation of aspirin within 48 hours after surgery with the aim to be continued indefinitely [42,43]. When started more than 48 hours after surgery, efficacy of aspirin is hampered and early graft occlusion more frequently observed [45].

One of the important clinical trials that evaluated efficacy and safety of antithrombotic drug therapies was CABADAS (CABADAS = prevention of Coronary Artery Bypass graft occlusion by Aspirin, Dipyridamole and Acenocoumarol /phenprocoumon Study). CABADAS evaluated the effect of aspirin, aspirin plus dipyridamole and vitamin K antagonists on one-year graft patency and early clinical outcome. The CABADAS study concluded that there was no convincing evidence that the addition of dipyridamole to a low dose of aspirin improves one-year vein-graft patency after coronary artery bypass grafting. Oral anticoagulants, compared with aspirin, provided no benefit [46].

However, in many of the clinical trials that contributed to the evidence of aspirin as primary choice for secondary prophylaxis after CABG, data on long-term clinical outcome are limited. As in CABADAS, mainly early effects on graft patency


- 16 -

and short- and mid-term clinical outcome are available, and long-term outcome could provide further evidence for the use of aspirin as secondary prophylaxis after CABG. Coronary-artery-bypass-graft surgery CABADAS studied graft patency after coronary artery bypass graft surgery. Since its introduction in the 1960s, CABG is now one of the most common operations performed in the world. It has proven to consistently relieve angina and improve the quality of life in symptomatic patients. Both saphenous vein and arterial grafts are used, in which arterial grafts have a late patency rate of 90% at 10 years after surgery, compared to approximately 50% for vein grafts. And although optimization of peri-operative (e.g. platelet inhibitors) and long-term medical treatment (e.g. more aggressive statin use) will improve vein graft patency rates, the current recommendation is to use the left internal mammary artery (IMA) as primary choice for revascularization of the left anterior descending artery because of its excellent late patency rate [41].

In line with the superiority of a single IMA over the saphenous vein graft [47-49], the benefit of bilateral IMA grafting can be expected. There are several non-randomized comparative studies confirming this hypothesis [50-52], but randomized trials on this issue are lacking. A major drawback for arterial grafting is the increased operative difficulty, with increased operative time and a higher risk of wound complications, also reflected by the limited number of institutions performing this type of surgery. Furthermore, considering the excellent outcome of single IMA grafting during the first 10 years, comparative studies should be performed with extended follow-up over more than 10 years. This is even more the case for total arterial grafting in patients with three-vessel disease. Whether three pedicled arterial grafts can further improve especially the long-term clinical outcome is still unknown, although the mid-term results of previous studies were promising [53,54].

Chapter 1

- 17 -

OUTLINE OF THIS THESIS We addressed the use of vitamin K antagonists in deep vein thrombosis, pulmonary embolism, atrial fibrillation, heart valve replacement and coronary artery disease.

In evaluating the effectiveness of vitamin K antagonists, we focused on the individual responses to VKA, i.e. the individual time within the INR target range, the determination of an inadequate level of anticoagulation and its impact on both thromboembolism and bleeding (major and minor) complications (chapter 2 and 3). In addition, the predictability of such an inadequate level of anticoagulation was assessed. Especially the importance of the initial 30-day response to VKA was evaluated (chapter 3). Next, we evaluated whether minor bleeds could predict major bleeding, the latter causing serious morbidity and mortality (chapter 4).

Long-term effectiveness of VKA therapy was evaluated in patients with coronary artery disease in whom bypass surgery was performed using vein grafts (with or without an arterial graft). These patients were from the CABADAS extended 14-year follow-up study. Next to a formal comparison of vitamin K antagonists, aspirin and aspirin plus dipyridamole (chapter 5), in the patients treated with vitamin K antagonists the impact of differences in quality of anticoagulation during the first year of therapy on long-term clinical outcome was evaluated (chapter 6). Finally, the findings of this cohort were put into the perspective of arterial vs. vein grafting. The long-term benefit of total arterial grafting was assessed in patients with 3-vessel coronary artery disease, and compared with historical data. In an additional analysis, these patients were also compared with a subgroup of patients from CABADAS, formally comparing arterial vs. vein grafting in patients with 3-vessel disease (chapter 7).


- 18 -

References 1. Ansell J, Hirsh J, Hylek E, Jacobson A, Crowther M, Palareti G. Pharmacology and management

of the vitamin K antagonists: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008; 133: 160S-198S.

2. Anand SS, Yusuf S. Oral anticoagulant therapy in patients with coronary artery disease: a meta-analysis. JAMA 1999; 282: 2058-2067.

3. Husted S, Emanuelsson H, Heptinstall S, Sandset PM, Wickens M, Peters G. Pharmacodynamics, pharmacokinetics, and safety of the oral reversible P2Y12 antagonist AZD6140 with aspirin in patients with atherosclerosis: a double-blind comparison to clopidogrel with aspirin. Eur Heart J 2006; 27: 1038-1047.

4. Salem DN, O'Gara PT, Madias C, Pauker SG. Valvular and structural heart disease: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008; 133: 593S-629S.

5. Holbrook AM, Pereira JA, Labiris R, McDonald H, Douketis JD, Crowther M, Wells PS. Systematic overview of warfarin and its drug and food interactions. Arch Intern Med 2005; 165: 1095-1106.

6. Wilkinson GR. Drug metabolism and variability among patients in drug response. N Engl J Med 2005; 352: 2211-2221.

7. Kucher N, Connolly S, Beckman JA, Cheng LH, Tsilimingras KV, Fanikos J, Goldhaber SZ. International normalized ratio increase before warfarin-associated hemorrhage: brief and subtle. Arch Intern Med 2004; 164: 2176-2179.

8. Schwarz UI, Ritchie MD, Bradford Y, Li C, Dudek SM, Frye-Anderson A, Kim RB, Roden DM, Stein CM. Genetic determinants of response to warfarin during initial anticoagulation. N Engl J Med 2008; 358: 999-1008.

9. Sconce EA, Khan TI, Wynne HA, Avery P, Monkhouse L, King BP, Wood P, Kesteven P, Daly AK, Kamali F. The impact of CYP2C9 and VKORC1 genetic polymorphism and patient characteristics upon warfarin dose requirements: proposal for a new dosing regimen. Blood 2005; 106: 2329-2333.

10. Wittkowsky AK. Dietary supplements, herbs and oral anticoagulants: the nature of the evidence. J Thromb Thrombolysis 2008; 25: 72-77.

11. Palareti G, Legnani C, Guazzaloca G, Lelia V, Cosmi B, Lunghi B, Marchetti G, Poli D, Pengo V. Risks factors for highly unstable response to oral anticoagulation: a case-control study. Br J Haematol 2005; 129: 72-78.

12. Wittkowsky AK, Whitely KS, Devine EB, Nutescu E. Effect of age on international normalized ratio at the time of major bleeding in patients treated with warfarin. Pharmacotherapy 2004; 24: 600-605.

13. Penning-van Beest FJ, van ME, Rosendaal FR, Stricker BH. Characteristics of anticoagulant therapy and comorbidity related to overanticoagulation. Thromb Haemost 2001; 86: 569-574.

14. Wittkowsky AK. Factors associated with INR elevation and bleeding complications during Warfarin therapy. Arch Intern Med 2005; 165: 703.

15. Fihn SD, Gadisseur AA, Pasterkamp E, van der Meer FJ, Breukink-Engbers WG, Geven-Boere LM, van Meegen E, de Vries-Goldschmeding H, Antheunissen-Anneveld I, van't Hoff AR,

Chapter 1

- 19 -

Harderman D, Smink M, Rosendaal FR. Comparison of control and stability of oral anticoagulant therapy using acenocoumarol versus phenprocoumon. Thromb Haemost 2003; 90: 260-266.

16. van der Meer FJ, Briet E, Vandenbroucke JP, Sramek DI, Versluijs MH, Rosendaal FR. The role of compliance as a cause of instability in oral anticoagulant therapy. Br J Haematol 1997; 98: 893-900.

17. McCormick D, Gurwitz JH, Goldberg RJ, Becker R, Tate JP, Elwell A, Radford MJ. Prevalence and quality of warfarin use for patients with atrial fibrillation in the long-term care setting. Arch Intern Med 2001; 161: 2458-2463.

18. Wittkowsky AK, Devine EB. Frequency and causes of overanticoagulation and underanticoagulation in patients treated with warfarin. Pharmacotherapy 2004; 24: 1311-1316.

19. Kirkwood TB. Calibration of reference thromboplastins and standardisation of the prothrombin time ratio. Thromb Haemost 1983; 49: 238-244.

20. Cannegieter SC, Rosendaal FR, Wintzen AR, van der Meer FJ, Vandenbroucke JP, Briet E. Optimal oral anticoagulant therapy in patients with mechanical heart valves. N Engl J Med 1995; 333: 11-17.

21. The European Atrial Fibrillation Trial Study Group. Optimal Oral Anticoagulant Therapy in Patients with Nonrheumatic Atrial Fibrillation and Recent Cerebral Ischemia. N Engl J Med 1995; 333: 5-10.

22. Forfar JC. Prediction of hemorrhage during long-term oral coumarin anticoagulation by excessive prothrombin ratio. Am Heart J 1982; 103: 445-446.

23. Palareti G, Leali N, Coccheri S, Poggi M, Manotti C, D'Angelo A, Pengo V, Erba N, Moia M, Ciavarella N, Devoto G, Berrettini M, Musolesi S. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Italian Study on Complications of Oral Anticoagulant Therapy. Lancet 1996; 348: 423-428.

24. Palareti G, Manotti C, DAngelo A, Pengo V, Erba N, Moia M, Ciavarella N, Devoto G, Berrettini M, Leali N, Poggi M, Legnani C, Musolesi S, Coccheri S. Thrombotic events during oral anticoagulant treatment: results of the inception-cohort, prospective, collaborative ISCOAT study: ISCOAT study group (Italian Study on Complications of Oral Anticoagulant Therapy). Thromb Haemost 1997; 78: 1438-1443.

25. Loeliger EA. Laboratory control, optimal therapeutic ranges and therapeutic quality control in oral anticoagulation. Acta Haematol 1985; 74: 125-131.

26. Rosendaal FR, Cannegieter SC, van der Meer FJ, Briet E. A method to determine the optimal intensity of oral anticoagulant therapy. Thromb Haemost 1993; 69: 236-239.

27. Hutten BA, Prins MH, Redekop WK, Tijssen JG, Heisterkamp SH, Buller HR. Comparison of three methods to assess therapeutic quality control of treatment with vitamin K antagonists. Thromb Haemost 1999; 82: 1260-1263.

28. Schmitt L, Speckman J, Ansell J. Quality assessment of anticoagulation dose management: comparative evaluation of measures of time-in-therapeutic range. J Thromb Thrombolysis 2003; 15: 213-216.

29. Poller L, Keown M, Ibrahim S, Lowe G, Moia M, Turpie AG, Roberts C, van den Besselaar AM, van der Meer FJ, Tripodi A, Palareti G, Shiach C, Bryan S, Samama M, Burgess-Wilson M, Heagerty A, Maccallum P, Wright D, Jespersen J. An international multicenter randomized study of computer-assisted oral anticoagulant dosage vs. medical staff dosage. J Thromb Haemost 2008; 6: 935-943.


- 20 -

30. Pasterkamp E, Kruithof CJ, van der Meer FJ, Rosendaal FR, Vanderschoot JP. A model-based algorithm for the monitoring of long-term anticoagulation therapy. J Thromb Haemost 2005; 3: 915-921.

31. Azar AJ, Cannegieter SC, Deckers JW, Briet E, van Bergen PF, Jonker JJ, Rosendaal FR. Optimal intensity of oral anticoagulant therapy after myocardial infarction. J Am Coll Cardiol 1996; 27: 1349-1355.

32. Rosendaal FR. The Scylla and Charybdis of oral anticoagulant treatment. N Engl J Med 1996; 335: 587-589.

33. Hirsh J, Fuster V, Ansell J, Halperin JL. American Heart Association/American College of Cardiology Foundation guide to warfarin therapy. Circulation 2003; 107: 1692-1711.

34. Petersen P, Grind M, Adler J. Ximelagatran versus warfarin for stroke prevention in patients with nonvalvular atrial fibrillation: SPORTIF II: a dose-guiding, tolerability, and safety study. Journal of the American College of Cardiology 2003; 41: 1445-1451.

35. Palareti G, Legnani C, Cosmi B, Guazzaloca G, Cini M, Mattarozzi S. Poor anticoagulation quality in the first 3 months after unprovoked venous thromboembolism is a risk factor for long-term recurrence. J Thromb Haemost 2005; 3: 955-961.

36. Heneghan C, Alonso-Coello P, Garcia-Alamino JM, Perera R, Meats E, Glasziou P. Self-monitoring of oral anticoagulation: a systematic review and meta-analysis. Lancet 2006; 367: 404-411.

37. Christensen TD, Johnsen SP, Hjortdal VE, Hasenkam JM. Self-management of oral anticoagulant therapy: a systematic review and meta-analysis. Int J Cardiol 2007; 118: 54-61.

38. Siebenhofer A, Berghold A, Sawicki PT. Systematic review of studies of self-management of oral anticoagulation. Thromb Haemost 2004; 91: 225-232.

39. Becker RC, Meade TW, Berger PB, Ezekowitz M, O'Connor CM, Vorchheimer DA, Guyatt GH, Mark DB, Harrington RA. The primary and secondary prevention of coronary artery disease: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008; 133: 776S-814S.

40. Antiplatelet Trialists' Collaboration. Collaborative overview of randomised trials of antiplatelet therapy--II: Maintenance of vascular graft or arterial patency by antiplatelet therapy. BMJ 1994; 308: 159-168.

41. Eagle KA, Guyton RA, Davidoff R, Edwards FH, Ewy GA, Gardner TJ, Hart JC, Herrmann HC, Hillis LD, Hutter AM, Jr., Lytle BW, Marlow RA, Nugent WC, Orszulak TA, Antman EM, Smith SC, Jr., Alpert JS, Anderson JL, Faxon DP, Fuster V, Gibbons RJ, Gregoratos G, Halperin JL, Hiratzka LF, Hunt SA, Jacobs AK, Ornato JP. ACC/AHA 2004 guideline update for coronary artery bypass graft surgery: summary article. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1999 Guidelines for Coronary Artery Bypass Graft Surgery). J Am Coll Cardiol 2004; 44: e213-e310.

42. Chesebro JH, Fuster V, Elveback LR, Clements IP, Smith HC, Holmes DR, Jr., Bardsley WT, Pluth JR, Wallace RB, Puga FJ, . Effect of dipyridamole and aspirin on late vein-graft patency after coronary bypass operations. N Engl J Med 1984; 310: 209-214.

43. Lorenz RL, Schacky CV, Weber M, Meister W, Kotzur J, Reichardt B, Theisen K, Weber PC. Improved aortocoronary bypass patency by low-dose aspirin (100 mg daily). Effects on platelet aggregation and thromboxane formation. Lancet 1984; 1: 1261-1264.

Chapter 1

- 21 -

44. Mangano DT. Aspirin and mortality from coronary bypass surgery. N Engl J Med 2002; 347: 1309-1317.

45. Sharma GV, Khuri SF, Josa M, Folland ED, Parisi AF. The effect of antiplatelet therapy on saphenous vein coronary artery bypass graft patency. Circulation 1983; 68: II218-II221.

46. van der Meer J, Hillege HL, Kootstra GJ, Ascoop CA, Mulder BJ, Pfisterer M, van Gilst WH, Lie KI. Prevention of one-year vein-graft occlusion after aortocoronary-bypass surgery: a comparison of low-dose aspirin, low-dose aspirin plus dipyridamole, and oral anticoagulants. The CABADAS Research Group of the Interuniversity Cardiology Institute of The Netherlands. Lancet 1993; 342: 257-264.

47. Sergeant PT, Blackstone EH, Meyns BP. Does arterial revascularization decrease the risk of infarction after coronary artery bypass grafting? Ann Thorac Surg 1998; 66: 1-10.

48. Loop FD, Lytle BW, Cosgrove DM, Stewart RW, Goormastic M, Williams GW, Golding LA, Gill CC, Taylor PC, Sheldon WC, . Influence of the internal-mammary-artery graft on 10-year survival and other cardiac events. N Engl J Med 1986; 314: 1-6.

49. Cameron A, Kemp HG, Jr., Green GE. Bypass surgery with the internal mammary artery graft: 15 year follow-up. Circulation 1986; 74: III30-III36.

50. Endo M, Nishida H, Tomizawa Y, Kasanuki H. Benefit of bilateral over single internal mammary artery grafts for multiple coronary artery bypass grafting. Circulation 2001; 104: 2164-2170.

51. Berreklouw E, Rademakers PP, Koster JM, van Leur L., van der Wielen BJ, Westers P. Better ischemic event-free survival after two internal thoracic artery grafts: 13 years of follow-up. Ann Thorac Surg 2001; 72: 1535-1541.

52. Lytle BW, Blackstone EH, Sabik JF, Houghtaling P, Loop FD, Cosgrove DM. The effect of bilateral internal thoracic artery grafting on survival during 20 postoperative years. Ann Thorac Surg 2004; 78: 2005-2012.

53. Bergsma TM, Grandjean JG, Voors AA, Boonstra PW, den Heyer P, Ebels T. Low recurrence of angina pectoris after coronary artery bypass graft surgery with bilateral internal thoracic and right gastroepiploic arteries. Circulation 1998; 97: 2402-2405.

54. Hirose H, Amano A, Takanashi S, Takahashi A. Coronary artery bypass grafting using the gastroepiploic artery in 1,000 patients. Ann Thorac Surg 2002; 73: 1371-1379.

Chapter 2

- 23 -

Chapter 2

Individual Time Within Target Range in

Patients Treated with Vitamin K Antagonists; Main Determinant of Quality of Anticoagulation

and Predictor of Clinical Outcome A Retrospective Study in 2300 Consecutive

Patients with Venous Thromboembolism

Nic J.G.M. Veeger

Margriet Piersma-Wichers

Jan G.P. Tijssen

Hans L. Hillege

Jan van der Meer

British Journal of Haematology 2005, 128, 513–519

Quality of anticoagulation using ITTR

- 24 -

SUMMARY The efficacy and safety of vitamin K antagonists (VKA) are related to the actual level of anticoagulation (given as the international normalized ratio, INR). It is often difficult to maintain an optimal INR over time. We assessed the clinical impact of the individual time spent within INR target range (ITTR) in 2304 consecutive patients with venous thromboembolism.

Annual incidences of recurrent thromboembolism and major bleeding were 6.2% and 2.8% respectively. The relative risk (RR) of thromboembolism was 4.5 (95% confidence interval (CI) 3.1-6.6, P < 0.001) at INR < 2.0, for major bleeding it was 6.4 (2.5-16.1, P < 0.001) at INR > 5.0, compared with INR 2.0-3.0. Patients with ITTR < 45% were at higher risk than those with ITTR > 65% (RR 2.8, 1.9-4.3, P < 0.001), while no difference was demonstrated comparing ITTR 45-65% and ITTR > 65% (RR 1.2, 0.7-1.8, P = 0.54). Annual incidences of recurrent thromboembolism were 16.0%, 4.9% and 4.6% at ITTR < 45%, 45-60% and > 65% respectively. For major bleeding these were 8.7%, 2.1% and 1.9% respectively. ITTR < 37% during the first 30 treatment days was highly predictive for the total treatment time ITTR < 45% (RR 24.2, 13.5-43.1, P<0.001).

In conclusion, ITTR can be used to identify patients on VKA at risk of recurrent thromboembolism or major bleeding. Since the 30-days ITTR is highly predictive for total treatment ITTR, these patients can be identified soon after start of treatment.

Chapter 2

- 25 -

INTRODUCTION Anticoagulant therapy in patients with deep vein thrombosis (DVT) and pulmonary embolism (PE) consists of unfractionated or low-molecular-weight heparin for at least 5 days and vitamin K antagonist (VKA) treatment for three to 6 months [1-5].

The efficacy and safety of VKA treatment are related to the level of anticoagulation, as expressed by the international normalized ratio (INR) of the prothrombin time. Low INR values are associated with an increased risk of recurrent venous thromboembolism, high INR values are associated with major bleeding. Consensus regarding the optimal intensity of anticoagulation in these patients is a therapeutic range of INR 2.0 to 3.0 [4-7].

Efficacy and safety also depend on the actual proportion of treatment time spent within the therapeutic range [5,8]. It is difficult to maintain the therapeutic range throughout the course of treatment. This is mainly because of wide inter- and intra-individual variation in the effect of VKA [2,4,5,9] leading to under- and over-anticoagulation for 25-40% of treatment time [5,10-14].

We performed a retrospective study in a large cohort of consecutive patients with DVT or PE treated with VKA, to assess the impact of the achieved intensity of anticoagulation and individual time within therapeutic range (ITTR) on clinical outcome. METHODS Patients and study design A cohort of 2304 patients with DVT or PE was retrospectively studied [15]. The cohort consisted of consecutive patients referred to a specialized anticoagulant clinic in The Netherlands for control of VKA therapy, from 1 January 1995 to 1 February 1998. The patients were followed from the first INR measurement taken at this clinic after the start of treatment until the end of treatment or the end of the study.


- 26 -

Data collection Patient characteristics Patient characteristics were collected from the patient records as these were on file at the anticoagulant clinic. Baseline data included age, sex, co-morbidity, prescribed VKA, concomitant drugs and indication for VKA therapy. Anticoagulation data In all patients, the level of anticoagulation, i.e. INR, was measured once every 3-4 weeks, or more frequently when appropriate. Dose adjustments were performed using a nomogram based automated system, with physician control. At INR values less than 2.0 the dose was increased and at INR 3.5-5.0 it was tapered, with predefined dose steps. In addition, treatment was interrupted for 1 day at INR > 5.0, and vitamin K was given at INR > 10.0. The next control of INR was performed within 3 days.

The target range was INR 2.0-3.5. This range, slightly wider than the internationally accepted therapeutic range (INR 2.0-3.0), is applied by the Dutch anticoagulant clinics to avoid, in particular, under-anticoagulation.

All data regarding anticoagulant therapy were stored in a computerized registration system. Prothrombin times, dates of measurements, type of VKA and dosage schedules were extracted from this system. Adverse events The physicians of the anticoagulant clinic registered all clinically relevant events. Information regarding recurrent thromboembolic events, major and minor bleeds, hospital admissions, treatment interruptions or cessation, and death during the treatment period was collected.

All reported events were adjudicated by an experienced haematologist, who was not informed about the level of anticoagulation in each patient, to ensure a blinded classification. Clinical outcome The primary endpoint contained recurrent thromboembolism and major bleeding (composite endpoint). Recurrent thromboembolism was defined as DVT, PE, or thromboembolism at other sites, demonstrated by objective diagnostic techniques. Compression ultrasonography and venography were used to confirm deep vein

Chapter 2

- 27 -

thrombosis, lung perfusion and ventilation scanning, spiral CT scanning and pulmonary angiography when PE was suspected. Major bleeding was defined as an overt bleed leading to transfusion, hospitalization and/or death, as well as retroperitoneal, intracranial or intra-ocular bleeding. All overt bleeds not classified as major were considered minor bleeds. Death was classified in accordance with its reported most likely cause, i.e. recurrent thromboembolism, major bleeding or other cause. Statistical analysis Intensity of anticoagulation Based on actual INR values, the day-to-day INR values were calculated. We used a linear estimation method, as proposed by Rosendaal et al. [13] with an adjustment of estimated values towards the next actual INR value (see Appendix).

INR-specific annual incidences were calculated for recurrent thromboembolism, major bleeding and the composite of recurrent thromboembolism and major bleeding. In each of these calculations, the post-event follow-up time was not included. Ninety-five per cent confidence intervals (95% CI) around the annual incidences were calculated under the Poisson distribution assumption [13,16].

In addition, to assess the efficacy and safety of different anticoagulation intensities, adjusted rate ratios were calculated using multivariable Poisson regression analysis [17,18].

From the estimated day-to-day levels of anticoagulation, the percentage of time within the target range of INR 2.0-3.5 was calculated for each individual patient (ITTR). Again, post-event follow-up time and INR values were not included in the calculation of the ITTR.

The effect of ITTR on recurrent thromboembolism and major bleeding was evaluated by survival analysis, comparing quartiles of ITTR. Survival was estimated using the Kaplan-Meier method, and Cox proportional hazard regression analysis was used to obtain adjusted hazard ratios.

In the multivariable Cox and Poisson regression analysis, clinically relevant covariates that were significant at a p-level of 0.20 in univariate analysis were included in the model. A backward elimination strategy, with selected covariates and interaction terms, was used to achieve the most suitable model to estimate rate- and hazard-ratios [17,18].


- 28 -

For all analyses, commercially available computer software (Statistical Analysis System version 6.12, SAS Institute, Cary, NC) was used. Reported P-values are all two-sided. A P-value < 0.05 was considered statistically significant. RESULTS Study patients From 1 January 1995 to 1 February 1998, 2304 consecutive patients with DVT or PE were included in the study. The mean age of the patients was 62 years and 58% were female. They received VKA therapy for DVT in 69% of cases and for PE in 31%. Malignancy was present in 10% of patients. Concomitant and potentially interactive drugs were used for long-term in 4% of patients, and for a short-time in 24%. Anticoagulation monitoring Anticoagulation monitoring consisted of 46397 prothrombin time measurements over a total treatment time of 1441 patient-years. Median patient follow-up (90% central range) was 4.1 months (0.2-29.2). VKA therapy was still ongoing in 33% of the patients at the end of follow-up. The median number of INR assessments per patient was 13 (2-70) and the median interval between two assessments 10 days (3-28). The VKA of choice was acenocoumarol, with 98% of the patients using this drug.

Figure 1 shows the percentage of patients below, within and above the INR target range on a daily basis, with day 1 as the first day of treatment in all patients. On average, the daily percentage of patients within the target range of INR 2.0-3.5 was 65%.

The mean ITTR, i.e. the time spent within the target range for each individual patient, was 63%. Patients were below this range for 11% and above this range for 26% of treatment time.

Overall, the mean intensity of anticoagulation achieved was INR 3.12 (95% CI, 3.09-3.16).

In individual patients, on average, 67% of INR measurements resulted in dose adjustments (increase 37%, reduction 30%). In patients with ITTR < 45%, the dose was changed after 81% of INR measurements (35% and 46%), compared with 71% in patients with ITTR 45-65% (38% and 33%) and 60% in patients with ITTR > 65% (38% and 22%).

Chapter 2

- 29 -

Number of days on vitamin K antagonist therapy

0 100 200 300 400 500 600 700 800 900

Patie

nts b

elow

, with

in a

nd a

bove

INR

targ

et ra

nge

(%)

10

20

30

40

50

60

70

80

90

100

(2304) (1304) (784) (541) (418) (332) (265) (211) (163) (114)

Figure 1 Achieved day-to-day level of anticoagulation, expressed as the daily percentage of

patients below (grey), within (white) and above (black) the therapeutic range (INR 2.0-3.5) during the course of VKA therapy. The number of patients on treatment are shown between brackets.

Clinical outcome A total of 132 endpoint events were observed in 121 of 2304 patients (5.3%) (Table 1). These consisted of 91 episodes of recurrent thromboembolism in 85 patients (3.7%) and 41 episodes of major bleeding in 40 patients (1.7%).

A single event occurred in 111 patients, 76 experienced recurrent thromboembolism and 35 had major bleeding. Nine patients had two events (recurrences of thromboembolism, n = 4; major bleedings, n = 1; recurrence of thromboembolism followed by major bleeding, n = 2; major bleeding followed by recurrence of thromboembolism, n = 2) and one patient had three events (recurrences of thromboembolism). In addition to 41 major bleeds, 572 minor bleeds were reported in 376 patients (16%).

Events were fatal in eight patients, seven of whom had a major bleed and one had a recurrent thromboembolic event. All were treated for DVT, mean age was 81 years and 88% were females. None of them had malignancy. Only one patient had


- 30 -

received a concomitant potentially interactive drug for a short time (antibiotic), which was stopped more than 1 week prior to the fatal event. Mean percentage of times below, within and above INR target range were 14%, 45% and 41% respectively. In four patients ITTR was < 45%, in two patients it was 45-65% and two patients had an ITTR > 65%.

Overall, 107 patients (4.6%) died during follow up. Table 1 Recurrent venous thromboembolic events and major bleeds during VKA therapy.

TYPE OF EVENT Number of events

Number of patients with event * (N=2304)

Annual incidence

(%)

Recurrent thromboembolism Deep vein thrombosis Pulmonary embolism Ischaemic strokes (one fatal) Retinal artery thrombosis

91 61 24 5 1

85 (3.7%) 57 22 5 1

6.2

Major bleed Intracranial (three fatal) Gastro-intestinal (four fatal) Other major bleed

41 5 10 26

40 (1.7%) 5 10 25

2.8

Composite endpoint 132 121 (5.3%) 8.9 * First occurring event taken into account

Annual incidences of recurrent thromboembolism and major bleeding were 6.2% and 2.8% respectively. For the composite endpoint, the annual incidence was 8.9%.

Figure 2 presents INR-specific annual incidences of recurrent thromboembolism, major bleeding and composite events. It shows the typical U-shape curve of annual incidences of the composite events at different intensities of anticoagulation. The relative risk of recurrent thromboembolism was 4.5 (95% CI, 3.1-6.6, P < 0.001) at INR < 2.0, compared with INR 2.0-3.0. Above INR 3.0, a risk reduction was observed. For major bleeding, the relative risk was 6.4 (95% CI, 2.5-16.1, P < 0.001) at INR > 5.0. At lower INR levels the relative risks were not significant, all compared with INR 2.0-3.0.

Chapter 2

- 31 -

Recurrent thromboembolism

0

10

20

30

40

Major bleeding

Ann

ual i

ncid

ence

(%)

0

10

20

30

40

Composite of recurrent thromboembolism and major bleeding

Intensity of anticoagulation

INR 1.0 to 2.0 INR 2.0 to 3.0 INR 3.0 to 4.0 INR 4.0 to 5.0 INR > 5.00

10

20

30

40

4.5 (3.1-6.6) 1.0 0.6 (0.3-0.97) 0.5 (0.3-0.9) 0.4 (0.1-1.5)

0.4 (0.1-2.6) 1.0 1.2 (0.7-2.1) 1.3 (0.4-4.1) 6.4 (2.5-16.1)

3.7 (2.4-5.6) 1.0 0.7 (0.5-1.0) 0.8 (0.4-1.3) 2.0 (0.7-5.8)

Figure 2 INR-specific annual incidences (IR) of recurrent thromboembolism, bleeding and the composite events, with 95% confidence intervals. Underlined values are relative risks (95% CI).

Figure 3 shows the effect of ITTR on the composite endpoint, adjusted for age, sex and malignancy. In this analysis, the ITTR was categorised into quartiles. The two highest quartiles (ITTR 65-80% and 80-100%) were combined, as these showed comparable risks. Patients with an ITTR < 45%, i.e. the lowest quartile, showed a significantly higher risk of the composite endpoint, as compared with patients with an ITTR of 65-100% (hazard ratio 2.8, 95% CI, 1.9-4.3, P < 0.001). Patients with an ITTR of 45-65% did not differ from patients with an ITTR of 65-100% (hazard ratio 1.2, 95% CI, 0.7-1.8, P = 0.54).

When minor bleeding was included in the composite endpoint, the hazard ratios were 3.2 (95% CI, 2.6-3.9, P < 0.001) in patients with ITTR < 45%, and 0.9


- 32 -

Follow-up time in months

0 6 12 18 24 30

Free

from

thro

mbo

embo

lism

or m

ajor b

leed

ing

(%)

0

70

75

80

85

90

95

100

ITTR < 45%ITTR 45-65%ITTR > 65%

P < 0.001

(95% CI, 0.7-1.1, P = 0.33) in patients with ITTR of 45-65%, both compared with patients with an ITTR of 65-100%. Figure 3 Freedom from recurrent thromboembolism and major bleeding in patients according

to the percentage time within the therapeutic range (ITTR). RR ITTR<45% = 2.8 (95% CI 1.9-4.3, P < 0.001) and RR ITTR45 to 65% = 1.2 (95% CI 0.7-1.8, P = 0.54), compared with ITTR > 65%.

Absolute annual incidences of the composite endpoint were 25.0% for ITTR < 45%, 6.7% for ITTR 45-65% and 6.5% for ITTR > 65%. For recurrent thromboembolism, the absolute annual incidences were 16.6%, 4.9% and 4.6% respectively. For major bleeding, these were 8.7%, 2.1% and 1.9% respectively. When considering the percentages of time spent outside the target range, the ratio of time spent below and above the target range was approximately 1:2 in all three groups (0.45, 0.45 and 0.52 in patients with ITTR < 45%, 45-65% and > 65%). As malignancy, age and sex were associated with an adverse clinical outcome, a multivariable analysis was performed to adjust for these potential confounders. Although malignancy was more frequently observed in the ITTR < 45% subgroup (15% vs. 8% in each of the two other ITTR subgroups), multivariable analysis did not show interaction between malignancy and ITTR. Both ITTR and malignancy were strong independent predictors of clinical outcome (Figure 4).

Concomitant medication consisted of a wide variety of drugs for long-term (antiarrhythmics, antihypertensive drugs, lipid-lowering drugs and cytostatics) or

Chapter 2

- 33 -

Relative risks with 95% confidence intervals

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

Indication DVT

Male sex

Age > 55 years

Malignancy

ITTR 45 to 65%ITTR < 45%

short-term (antibiotics, NSAIDs, immunosuppressive drugs and psychotropic drugs) treatment.

Figure 4 Relative risks of the composite of recurrent thromboembolism and major bleeding,

for time within therapeutic range (ITTR) (reference group > 65%), malignancy, age, sex and indication for VKA (reference group pulmonary embolism). DVT, deep vein thrombosis.

Long-term drugs were used in 5% of the patients with adverse outcome vs. 4% of event-free patients (P = 0.65), short-term drugs in 21% vs. 24% (P = 0.66). Furthermore, 3% of patients in the ITTR < 45% subgroup used long-term concomitant drugs, compared with 6% and 4% in ITTR 45-65% and > 65% respectively.

Early ITTR as predictor of a low overall ITTR Since the distribution of INR values above, within and below the target range were constant over time, we evaluated whether a low overall ITTR (< 45%) could be predicted from the achieved level of anticoagulation during the first 30 days of treatment.

All patients were categorized into quartiles based on their ITTR for the first 30 days. Figure 5 shows the chance (95% CI) to find an overall ITTR < 45% for each quartile of patients. It was increased approximately 25-fold in patients with a 30-day ITTR less than 37% (median ITTR 20%), compared with patients with a 30-day ITTR of more than 90% (median ITTR 97%).


- 34 -

Percentage of time within target range during first 30 days20406080100

Relat

ive

risk

for p

oor o

vera

ll IT

TR

0

10

20

30

40

Figure 5 The 30-day ITTR as predictor of a low ITTR (< 45%) during the total treatment

period. Presented are the odds ratios with 95% confidence intervals, per quartile of 30-day ITTR, indicated by median values.

DISCUSSION In spite of optimal conditions for monitoring VKA treatment by a Dutch anticoagulant clinic, INR values were within the target range for, on average, 63% of treatment time. This finding is in agreement with the results of previous studies [8]. The distribution of proportions of time above, within and below the target range was fairly consistent over time. Hence, the observed ITTR of 63% could not be attributed to poor anticoagulation early after the start of VKA treatment in these patients who were treated for a relatively short period. Furthermore, also insufficient monitoring was not likely to be the cause of poor anticoagulation.

To assess the impact of variations in the quality of anticoagulation in individual patients, we analysed clinical outcome related to the ITTR. An ITTR < 45%, observed in a quarter of the patients, was associated with an annual incidence of the composite of recurrent thromboembolism and major bleeding that amounted to 25.0%, as compared with 6.7% in patients with an ITTR of 45-65% and 6.5% in patients with an ITTR > 65%. In each of the ITTR subgroups, approximately two-thirds of the endpoints were recurrences of thromboembolism, while approximately one-third of the time outside the target range was spent below the target range. The risk of recurrent thromboembolism was related to the total time at which INR values were

Chapter 2

- 35 -

below the target range in individual patients, rather than incidental low INR values. The same applied to the risk of major bleeding, although this was shown to be lower. This difference may be explained by a wider safety range of INR, considering the applied target range (2.0-3.5) and INR values associated with an increased risk of recurrent thromboembolism (< 2.0) and major bleeding (> 5.0). One might speculate that inherited or acquired thrombophilic disorders in patients with venous thromboembolism protect against major bleeding, whereas these patients are more prone to recurrence of venous thromboembolism for the same reason. It should be noted that fatal outcome was observed more frequently in patients with a major bleed (seven of 41, 17.1%) than in patients who experienced recurrent thromboembolism (one of 91, 1.1%).

As indicated by the annual incidences of recurrent thromboembolism and major bleeding at different INR ranges, our data was consistent with the consensus of an optimal target range of INR 2.0-3.0. A wider INR target range, as applied by the Dutch anticoagulant clinics and evaluated in this study (INR 2.0-3.5 instead of 2.0-3.0) will be more feasible in practice, because it can be more easily achieved and maintained. Moreover, as the chance of INR values below 2.0 is reduced, the risk of recurrent thromboembolism will probably be lower, whereas the somewhat higher upper limit of the INR target range will hardly influence the risk of major bleeding.

To improve the safety of prolonged VKA treatment, a lower INR target range has been suggested. Considering the results of our study, an increased risk of recurrent thromboembolism is likely, while the aimed benefit of a lower risk of major bleeding will be small. This is supported by two recent studies in patients with venous thromboembolism, who received prolonged VKA treatment [19,20]. Lower intensity VKA treatment (INR target range 1.5-2.0) was shown to be more effective than placebo in the first study [19], but was less effective than conventional intensity treatment (INR target range 2.0-3.0) in the second study [20]. The risk of major bleeding in both studies was comparable.

It is emphasized that the patients reported by Ridker et al. [19] and Kearon et al. [20] differed from the patients in our study. We analysed all patients with either a first episode or recurrence of venous thromboembolism from the start of VKA treatment. Patients in the previous studies had been pretreated with VKA for at least 3 months before study enrolment. Strict eligibility criteria were applied, including the absence of contraindications and the presence of other indications for prolonged


- 36 -

treatment. Patients were probably excluded if they had experienced recurrent thromboembolism or major bleeding during pre-treatment. These differences may have influenced the reported risk estimates.

Efforts to improve the efficacy and safety of VKA treatment in patients with venous thromboembolism should be directed at the subgroup with an ITTR < 45%. The clinical outcome in our patients who spent more than 45% of the treatment time within the target range did not significantly improve with increasing ITTR. This subgroup could be identified soon after the start of treatment, as a low ITTR calculated over the first 30 days was shown to be predictive for a low ITTR value over the total treatment period. With the data available in our study, it was not possible to clarify the consistently poor anticoagulation in this subgroup. The interaction of concomitant drugs with VKA or insufficient monitoring were not probable causes, but non-compliance and co-morbidity may have contributed. However, even with an extensive prognostic model explaining poor anticoagulation, it remains questionable whether the quality of VKA anticoagulation can be improved in these patients.

We are aware of the limitations of our study, because of its retrospective design. Even when objective techniques are used, it is often difficult to establish recurrence of venous thromboembolism. Furthermore, to distinguish recurrent DVT from the postphlebitic syndrome remains another challenge. In this respect, misclassification and, consequently, an overestimation of the recurrence rate cannot be fully ruled out.

However, it is unlikely that the clinical outcome related to ITTR has been influenced because (1) there is no evidence that poor anticoagulation and postphlebitic syndrome are correlated, and (2) misclassification would have equally affected the ITTR subgroups.

We included all patients with either a first episode or recurrence of DVT or PE, irrespective of the a priori risk of recurrent thromboembolism and major bleeding. Moreover, our study represents the daily practice of treatment with VKA in the setting of optimal monitoring by an anticoagulant clinic. One might speculate whether clinical outcome can be improved. It is more likely that clinical outcome would be worse if the infrastructure for optimal monitoring was not available and hence more patients probably would have an ITTR of less than 45%.

In conclusion, we established the known relationship between the risk of recurrent thromboembolism and major bleeding, and the level of oral anticoagulation. It was demonstrated that the risk of these events depends on the individual time

Chapter 2

- 37 -

within the target range, rather than incidental INR values outside the target range. The highest event rates were observed in those patients with an ITTR < 45%, despite intensive INR monitoring and adjustment of the dose of VKA accordingly. These patients can be identified early during treatment, as the ITTR of the first 30 days was highly predictive for the ITTR of the total treatment period. It remains questionable whether ITTR and clinical outcome can be improved, especially in this subgroup. APPENDIX In the estimation of the in-between values of INR, an adjustment factor was applied to modify the linearly estimated values [adjustment factor = (INR next actual – INR linear estimate)/2]. Using the adjustment factor, the estimated values of INRs were corrected towards the next actual INR value. By doing so, a change of dosage regimen following an undesirable level of anticoagulation was taken into account when estimating the in-between values of INR. To incorporate a delayed response of the anticoagulation level to a change of dosage, no adjustment was made on each first estimated INR value after an actual value.

If treatment was interrupted for more than 8 weeks, INRs were not estimated over that period, since the assumption of linearity was no longer judged as valid [11].


- 38 -

References 1. van Beek EJR, Buller HR, ten Cate JW. Epidemiology of venous thromboembolism. In: Tooke

JE, Lowe GD (editors). A textbook of vascular medicine. London: Arnold, 1996; 471-488. 2. Hirsh J. Oral anticoagulant drugs. N Engl J Med 1991; 324: 1865-1875. 3. Hyers TM, Agnelli G, Hull RD, Morris TA, Samama M, Tapson V, Weg JG. Antithrombotic

therapy for venous thromboembolic disease. Chest 2001; 119: 176S-193S. 4. Hirsh J, Dalen J, Anderson DR, Poller L, Bussey H, Ansell J, Deykin D. Oral anticoagulants:

mechanism of action, clinical effectiveness, and optimal therapeutic range. Chest 2001; 119: 8S-21S.

5. Hirsh J, Fuster V, Ansell J, Halperin JL. American Heart Association/American College of Cardiology Foundation guide to warfarin therapy. Circulation 2003; 107: 1692-1711.

6. Hull R, Hirsh J, Jay R, Carter C, England C, Gent M, Turpie AG, McLoughlin D, Dodd P, Thomas M, Raskob G, Ockelford P. Different intensities of oral anticoagulant therapy in the treatment of proximal-vein thrombosis. N Engl J Med 1982; 307: 1676-1681.

7. Schulman S, Lockner D. Relationship between thromboembolic complications and intensity of treatment during long-term prophylaxis with oral anticoagulants following DVT. Thromb Haemost 1985; 53: 137-140.

8. Ansell J, Hirsh J, Dalen J, Bussey H, Anderson D, Poller L, Jacobson A, Deykin D, Matchar D. Managing oral anticoagulant therapy. Chest 2001; 119: 22S-38S.

9. Levine MN, Raskob G, Landefeld S, Kearon C. Hemorrhagic complications of anticoagulant treatment. Chest 2001; 119: 108S-121S.






15. Sackett DL, Hayes RB, Guyatt GH, Tugwell P. Clinical Epidemiology. Boston: Little, Brown and Company, 1991; 173-185.

16. Rosner B. Fundamentals of biostatistics. Belmont: Wadsworth Publishing Company, 1995. 17. Kleinbaum DG, Kupper LL, Muller KE, Nizam A. Applied regression analysis and other multivariable

methods. Pacific Grove: Brooks/Cole Publishing Company, 1998. 18. Rothman KJ, Greenland S. Modern epidemiology. Philadelphia: Lippincott-Raven Publishers, 1998. 19. Ridker PM, Goldhaber SZ, Danielson E, Rosenberg Y, Eby CS, Deitcher SR, Cushman M, Moll

S, Kessler CM, Elliott CG, Paulson R, Wong T, Bauer KA, Schwartz BA, Miletich JP,

Chapter 2

- 39 -

Bounameaux H, Glynn RJ. Long-term, low-intensity warfarin therapy for the prevention of recurrent venous thromboembolism. N Engl J Med 2003; 348: 1425-1434.

20. Kearon C, Ginsberg JS, Kovacs MJ, Anderson DR, Wells P, Julian JA, Mackinnon B, Weitz JI, Crowther MA, Dolan S, Turpie AG, Geerts W, Solymoss S, Van Nguyen P, Demers C, Kahn SR, Kassis J, Rodger M, Hambleton J, Gent M. Comparison of low-intensity warfarin therapy with conventional-intensity warfarin therapy for long-term prevention of recurrent venous thromboembolism. N Engl J Med 2003; 349: 631-639.

Chapter 3

- 41 -

Chapter 3

Early Detection of Patients with a Poor Response to Vitamin K Antagonists;

The Clinical Impact of Individual Time Within Target Range

in Patients with Heart Disease

Nic J.G.M. Veeger


Hans L. Hillege

Harry J.G.M. Crijns

Jan van der Meer

Journal of Thrombosis and Haemostasis 2006; 4: 1625–7 (published in part).

Early detection of poor anticoagulation; clinical impact of ITTR

- 42 -

SUMMARY We recently demonstrated that in patients with venous thromboembolism, inadequate anticoagulation, i.e. reduced efficacy and safety was mainly restricted to a subgroup rather than being randomly distributed over time in all patients. Moreover, the response to VKA in terms of individual time within target range (ITTR) was predictable from the ITTR for the first 30 days of treatment. The objective of this study was to evaluated whether in a large cohort of 4454 consecutive patients with heart disease, inadequate anticoagulation was also restricted to a subgroup and whether these patients could be identified early during VKA treatment.

Mean ITTRs were 39%, 42% and 44% in patients with myocardial infarction (MI), atrial fibrillation (AF) and a prosthetic heart valve (PHV) respectively. Both thromboembolism and major bleeding occurred significantly most frequent in the patients with the lowest ITTRs (lowest quartile). Compared to patients with higher ITTRs, relative risks of thromboembolism were 6.4 (95% CI, 3.4-2.0; P < 0.001), 2.7 (1.4-5.2; P = 0.002) and 3.1 (1.2-8.8; P < 0.02) in MI, AF and PHV respectively. For major bleeding, these were 2.3 (0.5-11.3; P = 0.32), 3.2 (1.7-6.0; P < 0.001) and 2.6 (1.1-6.1; P = 0.03) respectively.

The chance to find a poor overall ITTR was 8.2, 7.7 and 15.2-fold higher in patients from the lowest quartile, compared to the highest quartile of the first 30-days ITTR in MI, AF and PHV respectively.

In conclusion, ITTR is a reliable measure of anticoagulation in individual patients. It enabled us to early identify a subgroup of poorly anticoagulated patients, who revealed a compromised clinical outcome.

Chapter 3

- 43 -

INTRODUCTION Vitamin K antagonists (VKA) are widely used for the primary and secondary prevention of thromboembolic events in a variety of heart diseases [1,2]. The therapeutic window of VKA is narrow and therefore requires close monitoring of the achieved intensity of anticoagulation, as expressed by the international normalized ratio (INR) of the prothrombin time. INR values above the optimal intensity range are associated with an increased risk of bleeding, whereas efficacy is reduced at values below this range. It is difficult to maintain the optimal levels of anticoagulation throughout the course of treatment. Due to wide inter- and intra-individual variations in the effect of VKA [1,3-6], INR values within the target range are achieved for only 60 to 75% of total treatment time in study populations [1,7-12]. Furthermore, a reduced time within the target range is associated with an adverse clinical outcome [4].

We recently reported on the individual time within the target range (ITTR) to evaluate VKA treatment in individual patients with deep vein thrombosis or pulmonary embolism. Adequate prophylaxis was obtained at ITTR values above a threshold of 45%. Both recurrent thromboembolism and major bleeding were strongly associated with an ITTR below 45%, which was observed in 25% of the patients that can be regarded as poor responders to VKA therapy. As the ITTR during the first 30 days of VKA therapy was strongly associated with the overall ITTR, these patients could be detected early [13].

In the present study, we determined the ITTR in a large group of consecutive patients who were treated with VKA for either myocardial infarction, atrial fibrillation or a prosthetic heart valve. We evaluated whether inadequate anticoagulation was a random phenomenon in all patients, or was restricted to a subgroup of patients. In addition, we assessed to what extent a reduced ITTR was related to a compromised clinical outcome. DESIGN AND METHODS Patients and study design A cohort of 4454 patients who were treated with VKA for either atrial fibrillation, myocardial infarction or a prosthetic heart valve was retrospectively studied [14]. The


- 44 -

cohort consisted of consecutive patients referred to a specialized anticoagulant clinic in The Netherlands for control of VKA therapy from 1 January 1995 to 1 February 1998. The patients were followed from the first INR measurement by this clinic after discharge from the hospital until the end of the study. Data collection Patient characteristics Patient characteristics were collected from the patient records as these were on file at the anticoagulant clinic. Baseline data included age, sex, co-morbidity, concomitant drugs and indication for VKA therapy. Anticoagulation data The target range, as recommended by the Dutch Federation of Thrombosis Services, was INR 2.5-3.5 in patients with atrial fibrillation and INR 3.0-4.0 in patients with myocardial infarction. In patients with a prosthetic heart valve, up to 1 January 1997 the target range was INR 3.5-4.8. From 1 January 1997 onwards, the target range was lowered to INR 3.0-4.0. In all patients, VKA dosing was performed using a nomogram based automated system, with physician control. The level of anticoagulation was measured once every 3 to 4 weeks, or more frequently when appropriate. In the initial phase, control also was more frequently performed, i.e. within 3 to 5 days. The dose was increased with predefined dose steps at INR values below the target range, and tapered at INR exceeding the target range but less than 6.4. In addition, treatment was interrupted for one day at INR values above 6.4 and vitamin K was given at INR > 10.0. The next control was then performed within 3 days. The coumarin derivative predominantly used was acenocoumarol.

All data regarding anticoagulant therapy were stored in a computerized registration system. Prothrombin times, dates of measurements and dosage schedules were extracted from this system. Adverse events The physicians of the anticoagulant clinic registered all clinically relevant events. Information regarding thromboembolic events, major and minor bleedings, hospital admissions, treatment interruptions or cessation, and death during the treatment period was collected.

Chapter 3

- 45 -

All reported events were adjudicated by an experienced haematologist, who was not informed about the level of anticoagulation to assure a blinded classification. Clinical outcome Clinical outcome was assessed by the occurrence of thromboembolic events and major bleeding. Thromboembolic events were classified as myocardial infarction, ischemic stroke, transient ischemic attack (TIA), other arterial thrombosis, deep vein thrombosis, pulmonary embolism, or venous thromboembolism at another site, demonstrated by objective diagnostic techniques.

Myocardial infarction was diagnosed by the presence of new pathological Q-waves on the ECG in combination with CPK and CPK-MB serum levels that exceeded the upper limits of normal ranges (troponin assay was not yet available). Ischemic stroke was defined as the onset of rapidly developed symptoms and signs of loss of cerebral function lasting at least 24 hours and had an apparent vascular cause, as demonstrated by computed tomography or magnetic resonant imaging. If a cerebral event completely recovered within 24 hours without cerebral lesions at scanning, it was classified as TIA. Compression ultrasonography and venography were used to confirm deep vein thrombosis, lung perfusion and ventilation scanning, spiral CT scanning and pulmonary angiography when pulmonary embolism was suspected.

Major bleeding was defined as a clinically overt bleed leading to transfusion, hospitalization and/or death, as well as retroperitoneal, intracranial or intra-ocular bleeding [5]. An overt bleed not classified as major was considered minor. Death was classified in accordance with its reported most likely cause, i.e. thromboembolism, major bleeding or another cause. Statistical analysis Intensity of anticoagulation Based on actual INR values, the day-to-day INR values were calculated. We used a linear estimation method, as proposed by Rosendaal et al. [10] with an adjustment of estimated values towards the next actual INR value [13]. If treatment was interrupted for more than 8 weeks, INRs were not estimated over that period, since the assumption of linearity was no longer judged as valid [8]. From the estimated day-to-day levels of anticoagulation, the percentage of time within the predefined target range was calculated for each individual patient (ITTR). The post-event data were not


- 46 -

included in the calculation of ITTR. Incidence rates were calculated for thromboembolism and major bleeding.

Ninety-five percent confidence intervals (95% CI) around the incidence rates were calculated under the Poisson distribution assumption [10,15]. In these calculations, the post-event follow-up time was not included.

The association of ITTR with thromboembolic events and major bleeding was assessed by multivariable Cox proportional hazard regression analysis, comparing patients with different ITTR, categorised into quartiles. In the multivariable analyses, a priori defined covariates were included (and interaction terms) to estimate adjusted relative risks for the primary variable of interest (ITTR) [16,17]. Overall survival was depicted using the Kaplan-Meier method.

In addition, we evaluated whether the initial response to anticoagulant therapy was associated with a poor overall ITTR. We calculated the initial response from the first 30 days INR values and categorized the patients into quartiles based on their 30-days ITTR. A poor overall ITTR was defined as the lowest quartile of overall ITTR. Using multivariable logistic regression, relative risks of poor overall ITTR for the quartiles of 30-days ITTR were calculated. Patients treated for less than 30 days were not included in this analysis.

Reported P-values are two-sided and a P-value < 0.05 was considered statistically significant. For all analyses, commercially available computer software (Statistical Analysis System version 8.2, SAS Institute, Cary, NC) was used. RESULTS Patients’ characteristics and anticoagulation monitoring From 1 January 1995 to 1 February 1998, 4454 consecutive patients were included in the study; 1012 with myocardial infarction, 2614 with atrial fibrillation and 828 patients with a prosthetic heart valve.

In patients with a myocardial infarction, total treatment time was 1054 patient-years, during which 26504 INR measurements were performed. In atrial fibrillation patients, treatment time was 2666 patient-years with 71581 INR measurements and in patients with a prosthetic heart valve 1521 patient-years with 45280 INR measurements.

Chapter 3

- 47 -

With 95% in the patients with a myocardial infarction, 98% in atrial fibrillation patients and 91% in patients with a prosthetic heart valve, acenocoumarol was predominantly used. The other coumarin derivative used was phenprocoumon. Aspirin was combined with coumarin in 9%, 3% and 1% of patients respectively. Table 1 Characteristics of patients and anticoagulation data.

Myocardial infarction N=1012

Atrial fibrillation N=2614

Prosthetic heart valve

N=828

Patients’ characteristics Age, years * Male gender, % Diabetes mellitus, % Malignancy, % Concomitant drugs cardiovascular, % long-term non-cardiovascular, %

short-term non-cardiovascular, %

68 ± 12

70 11 2

61 3 39

73 ± 11

51 10 3

47 8 40

65 ± 14

54 10 1

47 4 52

Anticoagulation data * Follow-up time, months Number of INRs Time between INRs, days INR ITTR, %

12.5 ± 11.4

26 ± 22 15 ± 9

3.4 ± 0.7 38.7 ± 22.5

12.2 ± 10.7

27 ± 22 14 ± 9

3.3 ± 0.7 42.3 ± 20.7

22.0 ± 13.1

55 ± 31 13 ± 5

3.5 ± 0.5 43.6 ± 18.5

Cardiovascular drugs included beta-blockers, Ca antagonists, nitrates, ACE-inhibitors, other anti-hypertensive drugs, anti-arrhythmic drugs, lipid lowering drugs, aspirin and diuretics. Non-cardiovascular drugs were drugs that potentially interact with vitamin K antagonists, including oral contraceptives, hormone replacement therapy, cytostatics and anti-epileptic drugs. Short-term non-cardiovascular drugs included incidental use of NSAIDs, a course of antibiotics, immunosuppressive drugs and sedatives. * mean ± sd

The characteristics of the patients are summarized in Table 1. The percentage

of males was the highest in the myocardial infarction cohort (70% vs. 51% and 54 % in patients with atrial fibrillation and a prosthetic heart valve respectively), as was the use of cardiovascular drugs (61% vs. 51% and 54% respectively). The patients in the atrial fibrillation cohort were older (73 years vs. 68 and 65 respectively). Mean INR was only slightly different (INR 3.4, 3.3 and 3.5, in patients with myocardial infarction, atrial fibrillation and a prosthetic heart valve respectively), as well as mean ITTR


- 48 -

(39%, 42% and 44% respectively), despite the differences in target range. Clinical outcome The overall occurrence of thromboembolic events and major bleeding in each of the three cohorts is depicted in Figure 1. Patients with a myocardial infarction experienced more thromboembolic events (at 30 months follow-up 7.4%) than patients with atrial fibrillation (3.8%) or patients with a prosthetic heart valve (3.5%), whereas the frequency of major bleeding was similar in the three cohorts (3.1%, 3.5% and 4.5% at 30 months respectively). Figure 1 Thromboembolism and major bleeding.

The incidence rate of thromboembolism was 4.0 per 100 person-years (95% CI, 2.9-5.4) in patients with myocardial infarction, 1.7 (95% CI, 1.2-2.3) in patients with atrial fibrillation and 1.4 (95% CI, 0.9-2.1) in patients with a prosthetic heart valve. With regard to major bleeding, the incidence rate was 0.9 per 100 person-years (95% CI, 0.5-1.7) in patients with myocardial infarction, as compared to 1.6 (95% CI, 1.2-2.2) in patients with atrial fibrillation and 1.7 (95% CI, 1.1-2.5) in patients with prosthetic heart valve.

Myocardial infarction and ischemic stroke were the most frequent thromboembolic events. In the myocardial infarction cohort these represented 93% and 7% respectively; in the atrial fibrillation cohort 38% and 51%; and in the prosthetic heart valve cohort 38% and 48%. Further details about the type of thromboembolic event and major bleeding are presented in Table 2.

Fatal outcome was similar in the three cohorts, but the causative type of event was different. In the myocardial infarction cohort, five events were fatal (0.5%) of

Time in months

0 6 12 18 24 30

Tim

e-re

lated

occ

urre

nce

(%)

0

5

10

Time in months

0 6 12 18 24 300

5

10

Atrial fibrillationMyocardial infarctionProsthetic heart valve

Thromboembolism Major bleeding

Chapter 3

- 49 -

which four were recurrent myocardial infarctions and one an intracranial bleed. In the atrial fibrillation cohort fourteen fatal events had occurred (0.5%), four myocardial infarctions and ten major bleeds (five intracranial bleeds). Three patients with a prosthetic heart valve had a fatal event (0.4%), all myocardial infarctions. Table 2 Thromboembolic events and major bleeds during vitamin K antagonist therapy.

TYPE OF EVENT *

Myocardial infarction N=1012

Atrial fibrillation N=2614


N=828 Thromboembolism (N)

deep vein thrombosis pulmonary embolism ischemic strokes retinal artery thrombosis peripheral artery thrombosis non-fatal myocardial infarction fatal myocardial infarction prosthetic heart valve thrombosis incidence rate (95% confidence interval), per 100 patient-years

41 0 0 3 0 0 34 4† 0

4.0 (2.9-5.4)

45 2 1 23 1 1 13 4 0

1.7 (1.2-2.3)

21 2 0 10 0 0 5 3 1

1.4 (0.9-2.1)

Major bleed (N) non-fatal intracranial fatal intracranial non-fatal gastrointestinal fatal gastrointestinal other major bleed incidence rate (95% confidence interval), per 100 patient-years

10 1 1 4 - 4

0.9 (0.5-1.7)

43 5 5 7

5 21

1.6 (1.2-2.2)

26 1 -

13 -

12

1.7 (1.1-2.5)

* For thromboembolism and major bleeding, the number of patients with at least one event

is presented. Within each main event category, the first event was counted. In case of multiple events of both origins, e.g. a patient suffering a recurrent myocardial infarction and a major bleed, events were counted once in each category.

† One patient had a fatal myocardial infarction after an earlier non-fatal myocardial infarction. As only one thromboembolic event is counted, the fatal myocardial infarction was counted and the non-fatal myocardial infarction is not included in the number of non-fatal myocardial infarctions.

When both major and minor bleeding were taken into account, the incidence rate of any bleed was 18.3 (95% CI, 15.7-21.3) in the myocardial infarction cohort,


- 50 -

26.9 (95% CI, 24.7-29.2) in the atrial fibrillation cohort, and 25.9 per 100 person-years (95% CI, 23.0-29.1) in the prosthetic heart valve cohort.

Table 3 Individual time within, below and above the target range and incidence rates of

thromboembolism and major bleeding.

Mean time (%)

Within Below Abovetarget range

Incidence rates (95% CI) (per 100 person-years)

Thrombo- Major bleeding embolism

Myocardial infarction (target INR 3.0 to 4.0)*

ITTR < 25% ITTR 25 to 39% ITTR 39 to 53% ITTR > 53% Total

10 32 45 67

39

65 44 35 19

40

25 24 20 14

21

19.2 (11.7-29.6) 2.3 (0.9-4.8) 2.0 (0.8-4.0) 2.6 (1.0-5.4)

4.0 (2.9-5.4)

1.9 (0.2-6.8) 1.0 (0.2-2.9) 0.8 (0.2-2.3) 0.7 (0.1-2.7)

0.9 (0.5-1.7)

Atrial fibrillation (target INR 2.5 to 3.5)*

ITTR < 30% ITTR 30 to 43% ITTR 43 to 55% ITTR > 55% Total

16 37 49 68

42

43 24 18 12

25

41 39 33 20

33

4.1 (2.3-7.0) 1.6 (0.8-2.7) 1.1 (0.5-2.1) 1.4 (0.6-2.6)

1.7 (1.2-2.3)

4.1 (2.2-6.9) 1.3 (0.7-2.4) 1.3 (0.7-2.4) 1.1 (0.4-2.6)

1.6 (1.2-2.2)

Prosthetic heart valve (target INR 3.5 to 4.8)*

ITTR < 34% ITTR 34 to 45 % ITTR 45 to 54% ITTR > 54% Total

20 41 50 64

44

59 42 35 26

41

21 17 15 10

15

3.4 (1.4-7.1) 0.8 (0.2-2.2) 1.3 (0.5-2.9) 1.1 (0.4-2.6)

1.4 (0.9-2.1)

3.9 (1.7-7.7) 0.8 (0.2-2.2) 1.4 (0.5-3.0) 2.0 (0.9-3.7)

1.7 (1.1-2.5)

* For prosthetic heart valve patients, a lower target range of INR 3.0 to 4.0 was used from January

1st, 1997 onwards.

ITTR The mean individual time within the target range (ITTR) was 39%, 42% and 44% in patients with myocardial infarction, atrial fibrillation and a prosthetic heart valve respectively.

As presented in Table 3, within each cohort the patients were categorized into

Chapter 3

- 51 -

quartiles based on the ITTR. Patients with atrial fibrillation spend less time below than above the target range (25% vs. 33%). This was consistent over all ITTR quartiles, except for patients in the lowest quartile in whom time below and above the target range were equally high (43% and 41% respectively). In the myocardial infarction and prosthetic heart valve cohorts, patients spend more time below than above the target range, 40% vs. 21% and 41% vs. 15% respectively. This was consistent over all quartiles of both cohorts.

Figure 2 Thromboembolism and major bleeding in patients with different individual times within the therapeutic range (ITTR).

0 6 12 18 24 30Tim

e-re

lated

occ

urre

nce

(in %

)

0

5

10

15

20 ITTR below 30%ITTR above 30%

0 6 12 18 24 30

Tim

e-re

lated

occ

urre

nce

(in %

)

0

5

10

15


Time in months0 6 12 18 24 30

Tim

e-re

lated

occ

urre

nce

(in %

)

0

5

10

15


0 6 12 18 24 300

5

10

15


Thromboembolism Major bleeding

0 6 12 18 24 300

5

10

15


Time in months0 6 12 18 24 30

0

5

10

15


Atri

al fib

rillat

ion

Myo

card

ial in

farc

tion

Pros

thet

ic he

art v

alve


- 52 -

ITTR related clinical outcome The incidence rates of thromboembolism and major bleeding were related to ITTR, as shown in Table 3. In all three cohorts, the highest incidence rates of thromboembolism and major bleeds were observed in the lowest ITTR quartile, whereas in the other three quartiles incidence rates of thromboembolism as well as major bleeding were similar. In the myocardial infarction cohort, the hazard ratio (HR) of thromboembolism in patients with an ITTR less than 25%, i.e. the lowest quartile was 6.4 (95% CI, 3.4-12.0; P < 0.001), as compared to patients with higher ITTR values, adjusted for age, sex, diabetes mellitus, malignancy and concomitant drugs.

For the lowest ITTR quartile of patients with atrial fibrillation (ITTR < 30%) the HR of thromboembolism was 2.7 (95% CI, 1.4-5.2; P = 0.002) and for the lowest quartile of patients from the heart valve cohort (ITTR < 34%) this was 3.1 (95% CI, 1.2-8.8; P = 0.02).

The risk of major bleeding in patients with the lowest ITTRs (lowest quartile) vs. higher ITTR values were 2.3 (95% CI, 0.5-11.3; P = 0.32), 3.2 (95% CI, 1.7-6.0; P < 0.001) and 2.6 (95% CI, 1.1-6.1; P = 0.03) in patients with myocardial infarction, atrial fibrillation and a prosthetic heart valve respectively. Figure 2 shows the cumulative incidence of thromboembolism and major bleeding over time for the lowest quartile vs. the combined other quartiles of each cohort.

Next to ITTR, age above 69 years (HR = 2.3; 95% CI, 1.2-4.5; P = 0.01) and male gender (HR = 2.2; 95% CI, 1.0-4.9; P = 0.045) were independent risk factors for thromboembolic events in patients with myocardial infarction. In patients with atrial fibrillation, age above 74 years (i.e. the median age) and non-cardiovascular concomitant drugs were univariately associated with clinical outcome. In multivariable analysis these were not identified as independent risk factors, neither for thromboembolism nor for major bleeding. In the prosthetic heart valve cohort, age above 69 years (HR = 0.4; 95% CI, 0.1-1.0; P = 0.045) and concomitant cardiovascular drugs (HR = 0.3; 95% CI, 0.1-0.9; P = 0.02) were independent risk factors for thromboembolic events, and age above 69 years (HR = 3.1; 95% CI, 1.3-7.5; P = 0.01), male gender (HR = 3.3; 95% CI, 1.3-8.1; P = 0.01) and diabetes mellitus (HR = 3.2; 95% CI, 1.2-8.1; P = 0.02) for major bleeding.

When comparing the risk of thromboembolism with major bleeding, in the myocardial infarction cohort the risk of thromboembolism was 4-fold higher, and in

Chapter 3

- 53 -

patients with the lowest ITTRs even 10-fold higher than the risk of major bleeding (Table 3). In the atrial fibrillation cohort and the prosthetic heart valve cohort, incidence rates of thromboembolism and major bleeding were similar. This difference in the balance between thromboembolism and major bleeding was related to differences in the time below and time above INR target range, while the ITTR was similar (Figure 3).

Figure 3 Thromboembolism and major bleeding and percentage of time spent below and above the target range. Presented are the incidence rates with 95% CI, per quartile, indicated by their mean values.

Early ITTR as an indicator of overall ITTR As is shown in Figure 4, low ITTR values during the first 30 days of treatment were associated with a poor overall ITTR. In all three cohorts, the chance of a low overall ITTR was most pronounced in patients from the lowest quartile of 30-days ITTR. Comparing the patients from the lowest with those from the highest 30-days ITTR quartile in patients with myocardial infarction, the chance of a poor overall ITTR was 8.2-fold higher (95% CI, 4.6-14.5; P < 0.001). In patients with atrial fibrillation, it was 7.7-fold higher (95% CI, 5.6-10.6; P < 0.001) and in patients with a prosthetic heart valve 15.2-fold higher (95% CI, 7.9-29.3; P < 0.001).

Percentage of time above target range

10 20 30 40 50 60 70 80

1

3

5

7

9

11

13

15

24

Atrial fibrillationMyocardial infarctionProstethic heart valve

Major bleeding

Percentage of time below target range

10 20 30 40 50 60 70 80

Incid

ence

rate

(per

100

per

son

year

s)

1

3

5

7

9

11

13

15

24Thromboembolism


- 54 -

Individual time within target range during first 30 days (%)

0 20 40 60 80 100

Relat

ive

risk

for p

oor o

vera

ll IT

TR

1

3

5

7

9

11

13

15

30Atrial fibrillationMyocardial infarctionProsthetic heart valve

DISCUSSION In this study we evaluated the ITTR as a measure of quality of anticoagulation with VKA in individual patients. It was demonstrated that ITTR is related to clinical outcome in patients with a variety of clinical cardiac conditions. Moreover, ITTR for the first 30 days of VKA treatment is strongly associated with ITTR during total treatment.

Figure 4 Risk of poor overall ITTR in patients with different first 30-day ITTR.

Presented are the odds ratios with 95% confidence intervals, adjusted for age, sex, diabetes mellitus, malignancy and concomitant drugs, per quartile of 30-day ITTR, indicated by their median values.

On average, ITTRs were 39%, 42% and 44% in patients with myocardial infarction, atrial fibrillation and prosthetic heart valve respectively. These ITTR values are in agreement with TTR reported in previous studies [4,18]. Whereas TTR represents the distribution of INR values in a group of patients, ITTR enabled us to classify individual patients in accordance with INR values over time. Our study showed that variations in the level of anticoagulation were not randomly distributed among all patients, but that patients persistently revealed either adequate or poor anticoagulation. Although mean ITTRs in the three cohorts were comparable, there were differences in time below and time above INR target ranges. We have no simple explanation for this difference between cohorts. In all three cohorts, more intensive

Chapter 3

- 55 -

monitoring with dose adjustments of VKA was observed in patients with the lowest percentage of time spend within the target range and differences in ITTR were not attributable to a lack of VKA monitoring (data not shown). Although there was a difference between the cohorts in the use of concomitant drugs interacting with VKA, this could not explain the observed differences in time below vs. above target range. A difference in patient compliance between the cohorts also is an unlikely explanation. Changing the INR target range during the study period in patients with a prosthetic heart valve may have influenced the results in this cohort, although the ratio of time between below and above target range remained the same.

The incidence rates of thromboembolism and major bleeding in our study were similar to those reported in previous studies in patients with atrial fibrillation, myocardial infarction and a prosthetic heart valve [6,7,19-26]. In all three cohorts we demonstrated a similar strong relationship of thromboembolism and major bleeding with ITTR, with the highest incidence rates observed in the lowest ITTR quartile and no differences in the three higher ITTR quartiles.

Observed differences in clinical outcome between the cohorts may be explained by differences in pathophysiology of thrombosis in the heart cavities vs. coronary artery thrombosis. The inability to maintain INR values within its target range will have a much greater clinical impact when VKA treatment is applied to prevent the development of a small thrombus in a thrombogenic atherosclerotic coronary artery than to prevent a first thrombus in the much wider heart cavities. On the other hand, the cohorts differed with respect to the time below and the time above INR target range, while ITTR was similar. As a consequence of more frequent under-anticoagulation, the risk of major bleeding was lower in patients with myocardial infarction, especially in the lowest ITTR quartile, than in patients from the two other cohorts. The excess of over-anticoagulation in patients with atrial fibrillation was associated with an excess of fatal bleeds, while most fatal events were of thromboembolic origin in the other cohorts, reflecting excessive under-anticoagulation.

Our data emphasizes the great importance to maintain INR values within the optimal range, as recently argued by O’Donnell and Hirsh [27]. A limitation of our study is that the target ranges used are different from internationally recommended target ranges. However, our findings were consistent over all three cohorts indicating


- 56 -

that the ability to identify a poor response to VKA was independent of the target ranges used.

Considering the imbalance of under- and over-anticoagulation and the associated types of events, there seems to be room for improvement only in patients with myocardial infarction. As Schafer elegantly described as “walking the tightrope of dosing VKA therapy”, a shift of INR target value towards the upper part of the current target range, could improve clinical outcome in these patients [28]. For this reason higher target ranges are used in the Netherlands in an attempt to avoid underanticoagulation and thereby reducing the risk of thromboembolism, being the primary aim of VKA treatment. Improved algorithms for the dosing of VKA are currently under development [29].

Furthermore, our findings confirm the notion that in patients with myocardial infarction a more moderate-intensity of anticoagulation might not be suitable [4]. However, the more frequent use of concomitant antiaggregant therapy nowadays must also be taken into account. For patients on acenocoumarol, an early switch to a coumarin derivative with a longer half-life might be considered [18]. Moreover, coaching of patients identified as poor responders is indicated. As stated by Arnesen in a review on oral anticoagulation after myocardial infarction, patients with a poor response to VKA (in terms of stability of INR) should possibly be taken off the VKA treatment [25]. However, an acceptable alternative must then be available for these patients. Aspirin may be considered, awaiting for the results of clinical trials on newly developed anticoagulants.

In conclusion, ITTR can be used to assess quality of anticoagulation with VKA in individual patients. We demonstrated that differences in risk of thromboembolism and major bleeding between patients with atrial fibrillation, myocardial infarction and a prosthetic heart valve were strongly related to ITTR. Despite intensive INR monitoring and VKA dose adjustments accordingly, clinical outcome was compromised in 25% of patients with an ITTR of 25-34%. Depending on the proportion of time spent below or above the target range, either thromboembolism or major bleeding was predominantly seen. These patients could be identified early during treatment, as ITTR for the first 30 days was strongly associated with the ITTR for the total treatment period.

Chapter 3

- 57 -

References 1. Hirsh J, Fuster V, Ansell J, Halperin JL. American Heart Association/American College of

Cardiology Foundation guide to warfarin therapy. Circulation 2003; 107: 1692-1711. 2. Schulman S. Clinical practice. Care of patients receiving long-term anticoagulant therapy. N Engl J

Med 2003; 349: 675-683. 3. Hirsh J. Oral anticoagulant drugs. N Engl J Med 1991; 324: 1865-1875. 4. Ansell J, Hirsh J, Poller L, Bussey H, Jacobson A, Hylek E. The pharmacology and management

of the vitamin K antagonists: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 204S-233S.

5. Hirsh J. Current anticoagulant therapy--unmet clinical needs. Thromb Res 2003; 109 Suppl 1: S1-S8.

6. Levine MN, Raskob G, Beyth RJ, Kearon C, Schulman S. Hemorrhagic complications of anticoagulant treatment: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 287S-310S.






12. Petersen P, Grind M, Adler J. Ximelagatran versus warfarin for stroke prevention in patients with nonvalvular atrial fibrillation. SPORTIF II: a dose-guiding, tolerability, and safety study. J Am Coll Cardiol 2003; 41: 1445-1451.

13. Veeger NJ, Piersma-Wichers M, Tijssen JG, Hillege HL, van der Meer J. Individual time within target range in patients treated with vitamin K antagonists: main determinant of quality of anticoagulation and predictor of clinical outcome. A retrospective study of 2300 consecutive patients with venous thromboembolism. Br J Haematol 2005; 128: 513-519.

14. Sackett DL, Hayes RB, Guyatt GH, Tugwell P. Clinical Epidemiology. Boston: Little, Brown and Company, 1991; 173-185.

15. Rosner B. Fundamentals of biostatistics. Belmont: Wadsworth Publishing Company, 1995. 16. Kleinbaum DG, Kupper LL, Muller KE, Nizam A. Applied regression analysis and other multivariable

methods. Pacific Grove: Brooks/Cole Publishing Company, 1998. 17. Rothman KJ, Greenland S. Modern epidemiology. Philadelphia: Lippincott-Raven Publishers, 1998. 18. Fihn SD, Gadisseur AA, Pasterkamp E, van der Meer FJ, Breukink-Engbers WG, Geven-Boere

LM, van Meegen E, de Vries-Goldschmeding H, Antheunissen-Anneveld I, van't Hoff AR,


- 58 -

Harderman D, Smink M, Rosendaal FR. Comparison of control and stability of oral anticoagulant therapy using acenocoumarol versus phenprocoumon. Thromb Haemost 2003; 90: 260-266.

19. Singer DE, Albers GW, Dalen JE, Go AS, Halperin JL, Manning WJ. Antithrombotic therapy in atrial fibrillation: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 429S-456S.

20. Hylek EM, Go AS, Chang Y, Jensvold NG, Henault LE, Selby JV, Singer DE. Effect of intensity of oral anticoagulation on stroke severity and mortality in atrial fibrillation. N Engl J Med 2003; 349: 1019-1026.

21. Cannegieter SC, van der Meer FJ, Briet E, Rosendaal FR. Warfarin and aspirin after heart-valve replacement. N Engl J Med 1994; 330: 507-508.

22. Cannegieter SC, Rosendaal FR, Briet E. Thromboembolic and bleeding complications in patients with mechanical heart valve prostheses. Circulation 1994; 89: 635-641.

23. Salem DN, Stein PD, Al-Ahmad A, Bussey HI, Horstkotte D, Miller N, Pauker SG. Antithrombotic therapy in valvular heart disease--native and prosthetic: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 457S-482S.

24. Cannegieter SC, Torn M, Rosendaal FR. Oral anticoagulant treatment in patients with mechanical heart valves: how to reduce the risk of thromboembolic and bleeding complications. J Intern Med 1999; 245: 369-374.

25. Arnesen H. Oral anticoagulation after myocardial infarction. Thromb Res 2003; 109: 163-170. 26. Brouwer MA, Verheugt FW. Oral anticoagulation for acute coronary syndromes. Circulation 2002;

105: 1270-1274. 27. O'Donnell M, Hirsh J. Establishing an optimal therapeutic range for coumarins: filling in the

gaps. Arch Intern Med 2004; 164: 588-590. 28. Schafer AI. Warfarin for venous thromboembolism - walking the dosing tightrope. N Engl J Med

2003; 348: 1478-1480. 29. Pasterkamp E, Kruithof CJ, van der Meer FJ, Rosendaal FR, Vanderschoot JP. A model-based

algorithm for the monitoring of long-term anticoagulation therapy. J Thromb Haemost 2005; 3: 915-921.

Chapter 4

- 59 -

Chapter 4

Minor Bleeds Alert

for Subsequent Major Bleeding in Patients Using


Nic J.G.M. Veeger


Hans L. Hillege

Karina Meijer

Jan van der Meer

Submitted

Minor Bleeds Alert for Major Bleeding

- 60 -

SUMMARY Background: The majority of bleeding complications in patients on vitamin K antagonists (VKA) are clinically mild. These minor bleeds might be due to over-anticoagulation or a pre-existing mild bleeding disorder, which is enhanced by VKA. Objective: We investigated whether minor bleeds were associated with an increased risk of major bleeding, whereas possible intensified monitoring of VKA therapy after a minor bleed would improve quality of anticoagulation, thereby modifying the risk of major bleeding.

Methods: In an inception cohort of 6758 consecutive patients followed at a specialist anticoagulation clinic, this dual hypothesis was retrospectively studied. The risk of major bleeding was estimated using a multivariable Cox proportional hazards regression model with including a time-varying exposure for occurring minor bleeds. The achieved individual time within INR target range (ITTR) was specifically addressed.

Results: In patients without a minor bleed (N=5410) the incidence rate of major bleeding was 1.6 per 100 person-years (95%CI 1.3-2.1) and in patients with a minor bleed 2.1 per 100 person-years (95% CI 1.5-2.9) (N=1348). Patients with a minor bleed had a 3.6-fold higher risk of subsequent major bleeding (95%CI 2.0-6.6, P<0.001). ITTR was also independently associated with major bleeding, with a 2.8-fold increased risk in patients with an ITTR<25% of the total treatment time (95% confidence interval (CI) 1.7-4.5, P<0.001). Comparing patients with and without a minor bleed, there was no difference in ITTR (43% vs. 42% respectively).

Conclusion: Minor bleeds alert for subsequent major bleeding, independent of the achieved quality of anticoagulation.

Chapter 4

- 61 -

INTRODUCTION Vitamin K antagonists (VKA) are widely used in primary and secondary prevention of thromboembolism. Although these drugs have shown to be effective, the associated risk of bleeding is an important limitation [1]. Both effectiveness and risk of bleeding are related to the actual level of anticoagulation, expressed by the international normalized ratio (INR) [1,2]. The intensity of anticoagulation widely varies within and between patients, due to environmental and genetic factors [3-9], as well as patient’s compliance [10,11]. In this respect, the individual time within INR target range (ITTR) can be used to identify patients on VKA who are at high risk of (recurrent) thromboembolism and bleeding [2,12].

In most of the research on the risk of bleeding while using VKA, the primary focus was on major bleeding [1]. The risk of major bleeding is strongly associated with the level of anticoagulation. Furthermore, a history of major bleeding is also identified as a risk factor for subsequent major bleeding, although not consistently in all studies [13-17]. However, the majority of bleeding complications in patients on VKA are clinically mild. Incidence rates of major bleeding range from 1.4 to 3.3 per 100 person-years, whereas for minor bleeding incidence rates up to 30 per 100 person-years are reported [1,4,18-20]. Minor bleeding might be due to over-anticoagulation or a pre-existing mild bleeding disorder, which is enhanced by VKA. In either case, as with a history of major bleeding, the occurrence of a minor bleed during VKA therapy could be associated with an increased risk of major bleeding.

In this study, we hypothesized that patients with a minor bleeding are at increased risk for subsequent major bleeding, similar to a history of major bleeding. In this, minor bleeds alert for subsequent major bleeding. On the other hand, minor bleeds occurring during VKA therapy could lead to intensified monitoring of VKA therapy to maximize the quality of anticoagulation and thereby reducing the risk of major bleeding. METHODS Patients and study design A total of 6758 patients were retrospectively studied over a 3-year period. Patients already on VKA at the start of the study period were excluded (inception cohort). The


- 62 -

patients originated from two cohorts of consecutive patients referred to a large Dutch community based anticoagulation clinic for management of their VKA therapy from 1 January 1995 to 1 February 1998. Both cohorts were described in greater detail previously [2,12]. One cohort consisted of patients with deep vein thrombosis (n=1599) and patients with pulmonary embolism (n=705), and the other cohort of patients with atrial fibrillation (n=2614), patients with a myocardial infarction (n=1012) and patients with a prosthetic heart valve (n=828).

All patients were followed from the first INR measurement by the anticoagulation clinic until the end of treatment or the end of the study.

In the Netherlands, a unique nationwide network of 61 specialized anticoagulation clinics is responsible for the VKA management of all outpatients. In these specialized clinics, experienced physicians and dosing assistants are involved in optimizing VKA therapy, using computerized dosing programs and protocols for dose adjustments in case of under- and over-anticoagulation. In addition, patient education and awareness regarding factors influencing the quality of VKA therapy and signs of over-anticoagulation, i.e. minor bleeding, are specifically addressed. Data collection Patients’ characteristics Patients’ characteristics were collected from the records on file at the anticoagulation clinic. Baseline data included age, sex, co-morbidity, concomitant drugs and indication for VKA therapy. Anticoagulation data INR target range, as recommended by the Dutch Federation of Anticoagulation Clinics, was 2.5 to 3.5 in patients with venous thromboembolism or atrial fibrillation, and 3.0 to 4.0 in patients with myocardial infarction. In patients with a prosthetic heart valve, up to January 1st 1997 the target range was INR 3.5 to 4.8. From January 1st 1997 onwards, it was lowered to INR 3.0 to 4.0. In all patients, VKA dosing was performed using a nomogram based automated system, with control by experienced physicians. INR was measured once every 3 to 4 weeks or more frequently when appropriate. In the initial phase, INR was checked more frequently, i.e. every 3 to 5 days. The dose was increased with predefined dose steps at INR values below the target range, and tapered at INR values exceeding the target range but less than 6.4.

Chapter 4

- 63 -

Treatment was interrupted for one day at INR values above 6.4 and vitamin K was additionally given in case of INR > 10.0. The next measurement of INR was then performed within 3 days. The VKA of first choice was acenocoumarol. All data regarding anticoagulant therapy were stored in a computerized registration system. INR values, dates of measurements and dosage schedules were extracted from this system. Adverse events The physicians of the anticoagulation clinic registered all major bleeds, as a mandatory part of their quality system. Within the unique setting of VKA management in the Netherlands, patient education and awareness regarding the risks of VKA therapy also included reporting the occurrence of minor bleeds to the treating physicians. If treatment was interrupted due to a minor bleed, this was also included. For this study, all information about major and minor bleeds, hospital admissions, treatment interruptions or cessation, and death during the treatment period was collected. To minimize misclassification, all events were adjudicated by an experienced haematologist, who was not informed about the level of anticoagulation to assure a blinded classification. Clinical outcome Endpoints of clinical outcome were major and minor bleeds. Major bleeding was defined as a clinically overt bleed leading to transfusion, hospitalization and/or death, as well as a retroperitoneal, intracranial or intra-ocular bleed [21]. An overt bleed not classified as major was considered minor. When a minor bleed required any medical intervention, it was classified as clinically relevant minor bleed. Statistical analysis Based on actual INR values, the day-to-day INR values were calculated by linear estimation, as proposed by Rosendaal et al. with an adjustment of estimated values towards the next actual INR value [2,22]. Linear estimation was not performed when time between two assessments was more than 8 weeks, since the assumption of linearity was no longer judged as valid [23]. From the estimated day-to-day INR values, the percentage of time within the predefined target range was calculated for each individual patient (ITTR). Post-event data were not included in the calculation of


- 64 -

ITTR. Stability of anticoagulation was expressed as mean absolute difference (SD) in daily INR values.

Absolute risks (incidence rates per 100 person years) of minor and major bleeding were assessed. Ninety-five percent confidence intervals (95%CI) around the incidence rates were calculated under the Poisson distribution assumption. In these calculations, the post-event follow-up time was not included.

Table 1 Characteristics of patients at baseline and anticoagulation data.

Total N=6758

Patients’ characteristics Age, years *Male, % Diabetes mellitus, %Malignancy, % Concomitant drugs, % low dose aspirin beta blockers calcium antagonists nitrates antiarrhythmic drugs diuretics lipid lowering drugs oral contraceptives/HRT† NSAIDs

67 ± 15 51 9 5 4 7 9 1 10 13 8 1 14

Anticoagulation data * Follow-up time (months) Number of INRs Time between INRs (days) INR‡ ITTR (%) § Time with INR > 5.0 (%) INR stability║

11.9 ± 11.4 28 ± 25 13 ± 8 3.3± 0.7 42 ± 22 7 ± 12 0.1 ± 0.1

* mean ± SD; † hormone replacement therapy; ‡ International Normalized Ratio; § Individual Time within Target Range; ║ INR stability is expressed as absolute daily INR change.

Chapter 4

- 65 -

The effect of ITTR and minor bleeding on the risk of major bleeding was assessed by multivariable Cox proportional hazard regression analysis. Minor bleeding was incorporated in the analysis using a time-varying exposure approach, with a change in the exposure pattern at the time of the minor bleed up to 30 days thereafter, a priori defined. In the multivariable analyses, sex, age, co-morbidity, concomitant drugs, INR target range and indication for VKA therapy were included as potential confounders. As a sensitivity analysis on the a priori chosen exposure pattern, we repeated our time-varying exposure analysis using an exposure duration of 60 days.

Reported P-values are two-sided and a P-value < 0.05 was considered statistically significant. For all analyses, commercially available computer software (Statistical Analysis System version 9.1, SAS Institute, Cary, NC) was used. RESULTS A total of 6758 consecutive patients were included in our study. Their characteristics and anticoagulation data are summarized in Table 1. Cumulative treatment time was 6681 patients-years, during which 189762 INR measurements were performed. Acenocoumarol was the predominantly used VKA (96.7 %). The remaining patients used phenprocoumon (2.3%) or switched during their treatment period (1%). VKA was combined with low dose aspirin in 3.5% of patients. Mean individual treatment time was 11.9 months (SD, 11.4), during which 28 (SD, 25) INR measurements were performed. Mean INR was 3.3 (SD, 0.7), mean ITTR was 42% (SD, 22) and stability of anticoagulation, i.e. mean absolute change in daily INR values was 0.1 (SD, 0.1). Clinical outcome Minor bleeding A total of 2349 minor bleeds had occurred in 1348 patients (20%), 857 patients had a single minor bleed (64%), 272 patients two minor bleeds (20%), and 219 patients multiple recurrent minor bleeds (19%). INR at the time of the minor bleed was on average 3.7 (95% CI 3.6-3.8). The absolute risk, expressed as the incidence rate per 100 person-years was 25.1 (95%CI, 23.8-26.5). Haematoma (39%), nose bleeds (28%), conjunctiva bleeds (11%) and haematuria (9%) were the most frequently observed minor bleeds. In women, menorrhagia accounted for 3% of all reported minor bleeds. The most apparent difference between women and men was the occurrence of


- 66 -

haematoma, with 46% in women vs. 30% in men, leading to a significantly higher risk of minor bleeding in women than men (incidence rates 31.7, 95%CI, 29.5-34.0 in women vs. 19.8, 95%CI, 18.2-21.4 in men).

A medical intervention, including temporary interruption, was required for 111 minor bleeds in 99 patients. These minor bleeds were classified as clinically relevant. Major bleeding Major bleeding had occurred in 119 patients (1.8%), with an incidence rate of 1.8 per 100 person-years (95% CI, 1.5-2.2), of which 18 were fatal (0.3%). In 22 patients (18%) major bleeding had occurred within 30 days, and in 31 patients (26%) within 60 days. At the time of the major bleed, the INR was on average 4.4 (95% CI 3.8-5.1). The type of major bleeding is presented in Table 2.

When stratifying patients by minor bleeding, in patients without a minor bleed the incidence rate of major bleeding was 1.6 per 100 person-years (95% CI, 1.3-2.1), whereas in patient with a minor bleed the incidence rate was 2.1 per 100 person-years (95%CI, 1.5-2.9). The relative risk associated with minor bleeding was estimated using a multivariable time-varying exposure Cox regression model in which the exposure duration of the minor bleeds was set to 30 days. Using this exposure pattern, after a minor bleed patients were at a 3.6-fold increased risk of major bleeding up to 30 days after a minor bleed (HR=3.6, 95% CI, 2.0-6.6, P < 0.001).

Changing the exposure duration to 60 days, as a sensitivity analysis on the exposure duration, results were comparable (HR=3.6, 95% CI 2.2-5.9, P < 0.001), indicated a robustness of our finding of an increased risk after a minor bleed also over a substantial amount of time.

The composite of major bleeding and clinically relevant minor bleeding had occurred in 3.2% of patients (n=216) with an incidence rate of 3.3 (95%CI, 2.9-3.8). Also for the composite endpoint, patients with minor bleeding were at higher risk. The incidence rate in these patients was 7.0 per 100 person-years (95%CI, 5.9-8.3), whereas in patients without minor bleeding the incidence rate was 1.6 (95%CI, 1.3-2.1). Using the time-varying exposure model for this composite endpoint, the hazard ratio of minor bleeding (with the 30-day exposure duration pattern) was 4.7, 95% CI, 3.1-7.1, P < 0.001). Expanding the exposure duration to 60 days resulted in a HR of 3.9 (95% CI 2.7-5.7, P < 0.001).

Chapter 4

- 67 -

Table 2 Major bleeding and clinically relevant minor bleeding. TYPE OF EVENT * Minor bleeding Total

Yes n=1348

No n=5410

n=6758

Major bleeding n (%) non-fatal gastrointestinal fatal gastrointestinal non-fatal intracranial fatal intracranial intra-ocular other major bleeding incidence rate (95% CI) per 100 patient-years

Composite of major or clinically relevant minor bleeding n (%)†

non-fatal gastrointestinal fatal gastrointestinal non-fatal intracranial fatal intracranial intra-ocular other major or clinically relevant bleeding incidence rate (95% CI) per 100 patient-years

44 (3.3) 11 3 4 2 1 23 2.1 (1.5-2.9) 141 (10.5) 23 3 4 2 1 108 7.0 (5.9-8.3)

75 (1.4) 19 6 4 7 4 35 1.6 (1.3-2.1) 75 (1.4) 19 6 4 7 4 35 1.6 (1.3-2.1)

119 (1.8) 30 9 8 9 5 58 1.8 (1.5-2.2) 216 (3.2) 42 9 8 9 5 143 3.3 (2.9-3.8)

* The number of patients with at least one event is presented. In case of multiple bleeding, only the first event was counted; † Minor bleeding requiring any medical intervention was classified as clinically relevant minor bleeding.

Achieved level of anticoagulation Overall, the mean individual time within INR target range (ITTR) was 42%. In one-fifth of all patients, ITTR was < 25%. Of all time spend outside the target range on average 27% was above the target range. Based on the 0.5 INR wider therapeutic range, as this was applied by the Dutch anticoagulation clinics in the day-to-day monitoring, the percentage of time spend within the therapeutic range was 60%. The percentage of time spend above INR 5.0 was 7%.

There was no difference between the patients with a minor bleed (ITTR = 43%) and the patients without a minor bleed (ITTR = 42%). Also the percentage of


- 68 -

time spend above the target range was similar, 28% and 26% respectively. In addition, the percentage of time spend above INR 5.0 was 7% in patients with as well as in patients without a minor bleed. Figure 1 shows the highly comparable day-to-day level of anticoagulation during the course of VKA therapy in patients without a minor bleed or before the occurrence of a minor bleed, and in the patients after the occurrence of a minor bleed.

Figure 1 Achieved day-to-day level of anticoagulation, expressed as the daily percentage of patients below (dark grey), within (light grey) and above (black) the therapeutic range during the course of VKA therapy. The number of patients on treatment are shown between brackets.

From the multivariable Cox regression analysis, ITTR was identified as an independent risk indicator for major bleeding, irrespective of the presence of a minor bleed. Patients with an ITTR less than 25% (lowest quintile) were at 2.8-fold increased risk of major bleeding (HR=2.8, 95%CI, 1.7-4.5, P<0.001). For the composite of major and clinically relevant minor bleeding, this was 2.9-fold increased. (HR=2.9, 95%CI, 2.1-4.2, P<0.001).

Figure 2 shows all risk independent risk indicators for major bleeding, as these were identified by multivariable Cox regression analysis. In addition to minor bleeding and ITTR < 25%, these were age (55-70 years and especially above 70 years), sex (male), malignancy, indication for VKA therapy and use of NSAIDs.

Time since start vitamin K antagonists therapy (months)

0 3 6 9 12 15 18 21 24 27 30Patie

nts b

elow

, with

in a

nd a

bove

targ

et ra

nge

(%)

10

20

30

40

50

60

70

80

90

100

(6758) (4351) (2421)(3170) (1966) (1630) (1358) (1110) (878) (707) (520)

No minor bleeding or prior to first occurrence

Time since occurrence minor bleed (months)

0 3 6 9 12 15 18 21 24 27 30

10

20

30

40

50

60

70

80

90

100

(1348) (987) (638)(783) (524) (429) (353) (263) (199) (137) (79)

After first minor bleeding

Chapter 4

- 69 -

Figure 2 Adjusted hazard ratios of major bleeding and the composite of major and clinically

relevant minor bleeding, from multivariable Cox proportional hazards regression analysis.

DISCUSSION To our knowledge, this study is the first to report on minor bleeding as potential risk indicator for major bleeding. We showed that the occurrence of a minor bleeding was associated with a 3.6-fold increased risk of major bleeding during the subsequent month. The potential of minor bleeding as risk indicator was further corroborated by the similar risk when the duration of the exposure pattern was extended to 2 months. When including clinically relevant minor bleeding as a composite endpoint with major bleeding, minor bleeding remained a strong risk indicator for subsequent adverse clinical outcome.

The absolute risk of major bleeding in our study was 1.8 per 100 person-years. Taking into consideration that our cohort consisted of unselected patients in whom VKA therapy was managed by a specialized anticoagulation clinic, this absolute risk is in line with those reported previously [24,25]. Our rate of minor bleeding of 25.1 per 100 person-years is more difficult to compare, however in a recent large clinical trial in AF patients the rate of minor bleeding was 18.2 [26]. Two other large trials reported

0.2 0.5 2 50.1 1.0 10.0

Major bleedingMajor and clinically relevant minor bleeding

Variable

Minor bleeding (exposure duration 30 days)

ITTR < 25%

SexAge

Malignancy

Indication VKA

Use of NSAIDS

Male

55 to 70 years

70 years or older

References below 55 years

DVT/PE

Myocardial infarction


Reference atrial fibrillation

3.6 (2.0-6.6) P<0.001 4.7 (3.1-7.1) P<0.001

2.9 (2.1-4.2) P<0.001

1.2 (0.9-1.6) P=0.123

1.8 (1.0-3.2) P=0.040

2.7 (1.6-4.2) P<0.001

2.4(1.5-3.8) P<0.001

1.3 (0.9-1.8) P=0.104

0.5 (0.3-0.8) P=0.003

0.7 (0.5-1.0) P=0.067

1.4 (1.0-1.9) P=0.052

2.8 (1.7-4.5) P<0.001

1.4 (1.0-2.1) P=0.050

2.0 (0.9-4.3) P=0.069

2.8 (1.4-5.7) P=0.004

2.8 (1.5-5.0) P<0.001

1.7 (1.1-2.7) P=0.021

0.6 (0.3-1.2) P=0.144

1.3 (0.8-2.1) P=0.33

1.5 (1.0-2.2) P=0.058

1

1

Major bleeding

AdjustedHazard Ratio

(95%CI)

Major and clinicallyrelevant minor

bleedingAdjusted

Hazard Ratio(95%CI)

1

1


- 70 -

higher incidence rates for combined major and minor bleeding, ranging from 30 to 47 per 100 person-years [27,28].

Our finding regarding the quality of anticoagulation as risk indicator is consistent with previous studies, where an association between ITTR and major bleeding also was observed [29,30]. Although quality of anticoagulation, i.e. the ITTR was an important risk indicator for major bleeding, the ITTR was comparable in patients with a minor bleed vs. patients without a minor bleed (43% vs. 42%). Time spend above the target range was also similar, 28% and 26% respectively, and no apparent differences were observed in the achieved quality of anticoagulation before and after the occurrences of minor bleeds. In this respect, our hypothesis of risk modification after the occurrence of a minor bleed by increasing quality of anticoagulation was not confirmed.

Our study has some limitations to consider. Unfortunately, due to the retrospective design of our study, we did not have data on genetic factors related to the individual sensitivity to VKA [31,32]. Another limitation is the potential underreporting of bleeding complications in a study with a retrospective design. However, we included new patients only (inception cohort, excluding patients already on VKA) and had access to detailed information regarding all bleeding complications, including minor bleeds. These were assessed by the experienced staff of the anticoagulation clinic responsible for monitoring the VKA therapy and/or reported spontaneously by the patients themselves. Therefore, in our opinion underreporting was very limited, which is also indicated by the observed bleeding risk that is consistent with previous reports [1,18,19,24-28,33-35].

The intensity of anticoagulation used in the Netherlands might limit the generalizability of our findings. To avoid, particularly, under-anticoagulation, the Dutch anticoagulation clinic used a therapeutic range, next to the target range, which was 0.5 INR wider towards a higher intensity, than the conventionally employed target ranges in other countries (e.g. INR 2.0-3.5 vs. INR 2.0-3.0). The target ranges that were used were 0.5 INR higher (e.g. INR 2.5-3.5 vs. INR 2.0-3.0). Due to this strategy, the ITTRs as observed in our study were at the lower end of those reported in previous studies [4,36]. When considering the therapeutic range, the percentage within range was 60%, with only 13% of time spend below the range, confirming the strategy of avoiding under-anticoagulation. With an absolute risk of major bleeding of

Chapter 4

- 71 -

1.8 per 100 person-years, in the setting of a specialized anticoagulation clinic this strategy did not lead to excessive major bleeding.

Regarding the type of VKA, acenocoumarol was predominantly used. This short-acting VKA is associated with less stable anticoagulation [36]. Although acenocoumarol may have reduced the ITTR, minor bleeding was identified as a risk indicator of major bleeding, also independent of the ITTR.

In conclusion, minor bleeding is associated with subsequent major bleeding. The occurrence of a minor bleed should increase the awareness towards a potentially high-risk situation. In this, the achieved level of anticoagulation should also be considered, not just the INR at the time of the minor bleed, but especially the ITTR of the preceding treatment period.

In line with this awareness, a renewed risk-benefit assessment should be performed. Whether a lower intensity of anticoagulation might be more suitable for these high-risk patients, without compromising the efficacy of VKA therapy to prevent thromboembolic complications is speculative. In this, also VKA as the anticoagulant drug of choice could be reconsidered. For some patients, no anticoagulant agent, aspirin, or – in the near future – an oral II or Xa inhibitor might be a more suitable option.


- 72 -

References 1. Schulman S, Beyth RJ, Kearon C, Levine MN. Hemorrhagic Complications of Anticoagulant and

Thrombolytic Treatment: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008; 133: 257S-298S.

2. Veeger NJ, Piersma-Wichers M, Tijssen JG, Hillege HL, van der Meer J. Individual time within target range in patients treated with vitamin K antagonists: main determinant of quality of anticoagulation and predictor of clinical outcome. a retrospective study of 2300 consecutive patients with venous thromboembolism. Br J Haematol 2005; 128: 513-519.

3. Wilkinson GR. Drug therapy - Drug metabolism and variability among patients in drug response. New England Journal of Medicine 2005; 352: 2211-2221.

4. Ansell J, Hirsh J, Hylek E, Jacobson A, Crowther M, Palareti G. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008; 133: 160S-198S.

5. Kucher N, Connolly S, Beckman JA, Cheng LH, Tsilimingras KV, Fanikos J, Goldhaber SZ. International normalized ratio increase before warfarin-associated hemorrhage - Brief and subtle. Archives of Internal Medicine 2004; 164: 2176-2179.

6. Wittkowsky AK, Devine EB. Frequency and causes of overanticoagulation and underanticoagulation in patients treated with warfarin. Pharmacotherapy 2004; 24: 1311-1316.

7. Wittkowsky AK. Factors associated with INR elevation and bleeding complications during warfarin therapy. Archives of Internal Medicine 2005; 165: 703.

8. Palareti G, Legnani C, Guazzaloca G, Lelia V, Cosmi B, Lunghi B, Marchetti G, Poli D, Pengo V. Risks factors for highly unstable response to oral anticoagulation: a case-control study. British Journal of Haematology 2005; 129: 72-78.

9. Penning-van Beest FJA, van Meegen E, Rosendaal FR, Stricker BHC. Characteristics of anticoagulant therapy and comorbidity related to overanticoagulation. Thrombosis and Haemostasis 2001; 86: 569-574.

10. van der Meer FJM, Briet E, Vandenbroucke JP, Sramek DI, Versluijs MHPM, Rosendaal FR. The role of compliance as a cause of instability in oral anticoagulant therapy. British Journal of Haematology 1997; 98: 893-900.

11. McCormick D, Gurwitz JH, Goldberg RJ, Becker R, Tate JP, Elwell A, Radford MJ. Prevalence and quality of warfarin use for patients with atrial fibrillation in the long-term care setting. Archives of Internal Medicine 2001; 161: 2458-2463.

12. Veeger NJ, Piersma-Wichers M, Hillege HL, Crijns HJ, van der Meer J. Early detection of patients with a poor response to vitamin K antagonists: the clinical impact of individual time within target range in patients with heart disease. J Thromb Haemost 2006; 4: 1625-1627.

13. The Stroke Prevention in Atrial Fibrillation Investigators. Bleeding during antithrombotic therapy in patients with atrial fibrillation. Arch Intern Med 1996; 156: 409-416.

14. Beyth RJ, Quinn LM, Landefeld CS. Prospective evaluation of an index for predicting the risk of major bleeding in outpatients treated with warfarin. Am J Med 1998; 105: 91-99.

15. Kuijer PM, Hutten BA, Prins MH, Buller HR. Prediction of the risk of bleeding during anticoagulant treatment for venous thromboembolism. Arch Intern Med 1999; 159: 457-460.

16. White RH, Beyth RJ, Zhou H, Romano PS. Major bleeding after hospitalization for deep-venous thrombosis. Am J Med 1999; 107: 414-424.

Chapter 4

- 73 -

17. Shireman TI, Mahnken JD, Howard PA, Kresowik TF, Hou Q, Ellerbeck EF. Development of a contemporary bleeding risk model for elderly warfarin recipients. Chest 2006; 130: 1390-1396.

18. van der Meer FJ, Rosendaal FR, Vandenbroucke JP, Briet E. Bleeding complications in oral anticoagulant therapy. An analysis of risk factors. Arch Intern Med 1993; 153: 1557-1562.

19. Palareti G, Leali N, Coccheri S, Poggi M, Manotti C, D'Angelo A, Pengo V, Erba N, Moia M, Ciavarella N, Devoto G, Berrettini M, Musolesi S. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Italian Study on Complications of Oral Anticoagulant Therapy. Lancet 1996; 348: 423-428.

20. van Es RF, Jonker JJ, Verheugt FW, Deckers JW, Grobbee DE. Aspirin and coumadin after acute coronary syndromes (the ASPECT-2 study): a randomised controlled trial. Lancet 2002; 360: 109-113.

21. Hirsh J. Current anticoagulant therapy--unmet clinical needs. Thromb Res 2003; 109 Suppl 1: S1-S8.



24. Linkins LA, Choi PT, Douketis JD. Clinical impact of bleeding in patients taking oral anticoagulant therapy for venous thromboembolism: a meta-analysis. Ann Intern Med 2003; 139: 893-900.

25. Palareti G, Cosmi B. Bleeding with anticoagulation therapy - who is at risk, and how best to identify such patients. Thromb Haemost 2009; 102: 268-278.

26. Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Oldgren J, Parekh A, Pogue J, Reilly PA, Themeles E, Varrone J, Wang S, Alings M, Xavier D, Zhu J, Diaz R, Lewis BS, Darius H, Diener HC, Joyner CD, Wallentin L. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med 2009; 361: 1139-1151.

27. Albers GW, Diener HC, Frison L, Grind M, Nevinson M, Partridge S, Halperin JL, Horrow J, Olsson SB, Petersen P, Vahanian A. Ximelagatran vs warfarin for stroke prevention in patients with nonvalvular atrial fibrillation: a randomized trial. JAMA 2005; 293: 690-698.

28. Olsson SB. Stroke prevention with the oral direct thrombin inhibitor ximelagatran compared with warfarin in patients with non-valvular atrial fibrillation (SPORTIF III): randomised controlled trial. Lancet 2003; 362: 1691-1698.

29. Reynolds MW, Fahrbach K, Hauch O, Wygant G, Estok R, Cella C, Nalysnyk L. Warfarin anticoagulation and outcomes in patients with atrial fibrillation: a systematic review and metaanalysis. Chest 2004; 126: 1938-1945.

30. Samsa GP, Matchar DB, Phillips DL, McGrann J. Which approach to anticoagulation management is best? Illustration of an interactive mathematical model to support informed decision making. J Thromb Thrombolysis 2002; 14: 103-111.

31. Gage BF, Lesko LJ. Pharmacogenetics of warfarin: regulatory, scientific, and clinical issues. J Thromb Thrombolysis 2008; 25: 45-51.

32. Hermida J, Zarza J, Alberca I, Montes R, Lopez ML, Molina E, Rocha E. Differential effects of 2C9*3 and 2C9*2 variants of cytochrome P-450 CYP2C9 on sensitivity to acenocoumarol. Blood 2002; 99: 4237-4239.


- 74 -

33. DiMarco JP, Flaker G, Waldo AL, Corley SD, Greene HL, Safford RE, Rosenfeld LE, Mitrani G, Nemeth M. Factors affecting bleeding risk during anticoagulant therapy in patients with atrial fibrillation: observations from the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study. Am Heart J 2005; 149: 650-656.

34. Hylek EM, Evans-Molina C, Shea C, Henault LE, Regan S. Major hemorrhage and tolerability of warfarin in the first year of therapy among elderly patients with atrial fibrillation. Circulation 2007; 115: 2689-2696.

35. Go AS, Hylek EM, Chang Y, Phillips KA, Henault LE, Capra AM, Jensvold NG, Selby JV, Singer DE. Anticoagulation therapy for stroke prevention in atrial fibrillation: how well do randomized trials translate into clinical practice? JAMA 2003; 290: 2685-2692.

36. Fihn SD, Gadisseur AA, Pasterkamp E, van der Meer FJ, Breukink-Engbers WG, Geven-Boere LM, van Meegen E, Vries-Goldschmeding H, Antheunissen-Anneveld I, van't Hoff AR, Harderman D, Smink M, Rosendaal FR. Comparison of control and stability of oral anticoagulant therapy using acenocoumarol versus phenprocoumon. Thromb Haemost 2003; 90: 260-266.

Chapter 5

- 75 -

Chapter 5

Quality of Anticoagulation Associated with Late

Occurrence of Cardiac Events after Coronary-Artery-Bypass-Graft Surgery in Patients Treated

with Vitamin K Antagonists; Fourteen Year Follow-up Results of CABADAS

Nic J.G.M. Veeger

Felix Zijlstra

Hans L. Hillege

Jan van der Meer

Submitted

Post-CABG quality of VKA and late cardiac events

- 76 -

SUMMARY Introduction: Despite prophylactic therapy initiated after coronary-artery-bypass-graft surgery, graft occlusion as well as extension of the native coronary artery disease will gradually occur. We assessed the impact of the individually achieved level of anticoagulation during the first year on the 14-year clinical outcome of 235 patients treated with vitamin K antagonists (VKA), originally included in the CABADAS study.

Materials and methods: Based on the individual time spend above INR 3.0 and the variability of consecutive INRs measured during the first year after the coronary-artery-bypass-graft surgery, all patients were categorized as having non-optimal VKA (N=142) or optimal VKA (N=93) therapy. Primary endpoint was long-term occurrence of the composite of cardiac death and myocardial infarction.

Results: Median follow-up was 12.7 years (range 1.0-14.8). Cumulative rates of the composite of cardiac death and myocardial infarction were 44% in patients with non-optimal and 26% in patients with optimal VKA. However, up to 8 years, outcome was comparable (Hazard ratio (HR)=0.90; 95%CI 0.47 to 1.7; P=0.75). Thereafter, patients with non-optimal VKA had a 3.4-fold increased risk (95%CI 1.4 to 8.1; P=0.007), compared to patients with optimal VKA. The need for revascularisation was not influenced by non-optimal VKA (HR=0.87, 95%CI 0.50-1.5; P=0.63), which might be attributed to the frequent patient-doctor contacts in all VKA using patients.

Conclusion: Late clinical outcome is strongly affected by the achieved quality of anticoagulation. Considering the limited number of patients with an optimal response to VKA, in patients undergoing coronary-artery-bypass-graft surgery, VKA should be not considered as first choice.

Chapter 5

- 77 -

INTRODUCTION Nowadays, secondary prophylaxis using aspirin is standard of care in patients undergoing coronary-artery-bypass-graft surgery (CABG) to prevent vein graft closure [1,2]. This standard is based on a high level of evidence coming from several prospective controlled trials [3]. One of these clinical trials was CABADAS (CABADAS = prevention of Coronary Artery Bypass graft occlusion by Aspirin, Dipyridamole and Acenocoumarol/phenprocoumon Study) [4]. CABADAS evaluated the effect of aspirin alone, aspirin with dipyridamole and vitamin K antagonists (VKA) on one-year graft patency and early clinical outcome, and concluded that there was no convincing evidence that addition of dipyridamole to a low dose of aspirin was beneficial. Oral anticoagulants, compared with aspirin, provided no benefit [4].

Despite prophylactic therapy, occlusion of vein grafts will gradually occur. Approximately 50% of vein grafts are occluded 10 years after surgery. This process is reflected by an increasing risk of cardiovascular events starting from 5 to 10 years after surgery [2]. Optimal anticoagulant therapy already early after coronary-artery-bypass-graft surgery might proof crucial for long-term clinical outcome. Considering the large intra- and inter-variability in dose-response of VKA, resulting in a limited time in which optimal treatment is achieved, these drugs could be less beneficial. In this respect, the one-year results of the CABADAS study indicated a the lack of superiority of VKA over aspirin alone and aspirin combined with dipyridamole in part due to a non-optimal quality of VKA therapy [4].

In the present study, we specifically address the impact of the quality of VKA therapy on long-term clinical outcome, comparing patients with a optimal level of anticoagulation during the first year of treatment with patients with a non-optimal level of anticoagulation. METHODS The CABADAS study was an international multicenter randomized controlled clinical trial evaluating the effects of low dose aspirin alone, aspirin combined with dipyridamole and vitamin K antagonists on graft patency and clinical outcome at one year. The design of the study was described in great detail elsewere [4]. For the extended long-term follow-up, only the 726 Dutch participants of the CABADAS


- 78 -

study were considered. Of these, 244 patients were randomly allocated to receive VKA. VKA was started one day prior to initial CABG and continued on the first post-operative day. Initial study protocol was approved by the Ethics Committee of each participating center and all patients gave written informed consent. Patients All patients underwent elective CABG for disabling angina between July 1987 and August 1990. They received saphenous vein grafts and additional internal mammary artery (IMA) grafts were used at the discretion of the surgeon. Surgery was performed using routine techniques in each participating center. Heparin, administrated during surgery was antagonized by protamine sulphate at the end of the procedure, unless it was continued because of the introduction of an intra-aortic balloon pump. Main exclusion criteria were age over 70 years, previous or concurrent cardiac surgery, unstable angina at the time of surgery (< 2 days) or myocardial infarction within 7 days prior to surgery and severe concomitant disease. Follow up After discharge from the hospital, patients were seen for the duration of one year at a 3-month interval. The extended long-term follow-up was retrospectively performed for all patients up to June 2002, through telephonic interviews of the patients using a standard questionnaire and by reviewing the medical files of their cardiologists and/or general practitioners. This evaluation was performed for the first time 2 to 3 years after surgery and repeated 12 years thereafter. Clinical endpoints The primary endpoint of the extended follow-up study was a composite of the hard clinical endpoints cardiac mortality and myocardial infarction. Cardiac mortality was defined as mortality due to acute myocardial infarction, acute cardiac arrest, sudden death and mortality caused by progressive congestive heart failure. In case of an unknown cause of death, it was classified as cardiac death. For myocardial infarction, pathological Q-waves had to be present on the ECG in combination with CPK and CPK-MB serum levels that exceeded the upper limits of normal ranges.

Repeat revascularisation was evaluated as a secondary endpoint. Repeat revascularisation was defined as percutaneous coronary intervention (PCI) or redo-

Chapter 5

- 79 -

CABG during follow-up. From the initial CABADAS study, patients’ characteristics, general and disease

specific medical history and clinical outcome events were available up to one year after discharge. Adjudication of these events was performed by an independent classification committee, without knowledge of the patient’s treatment assignment. Anticoagulation data The patients allocated to VKA therapy were all referred to specialized Thrombosis Services for monitoring of the anticoagulant therapy. Based on actual INR values, day-to-day INR values were calculated using linear estimation [5]. From the estimated day-to-day INR values, the percentage of time within the predefined target range was calculated for each individual patient (ITTR) [6,7].

At the time of the original study, the target range used was INR 2.8 to 4.8. However, to asses quality of anticoagulation in reference to the currently recommended target range, we used INR 3.0 to 4.0 for our calculations of ITTR. The individual time below and above the target range was estimated using the same methodology. As a measure of stability of INRs over time, the variance growth rate was used [8,9].

Quality of anticoagulation in terms of “optimal VKA” was determined combining the individual time spend above INR 3.0 and the variance growth rate. INR 3.0 is the lower boundary of the currently recommended target range, indicating the minimum required intensity to achieve adequate prophylaxis against thromboembolic events. “Optimal VKA” was defined as a high percentage of time above INR 3.0 (middle and highest tertile), together with a moderate to low variance growth rate (middle and lowest tertile). In case of either a low percentage of time above INR 3.0 (lowest tertile) and/or a high variance growth rate (highest tertile), VKA treatment was considered non-optimal. Statistical analysis The primary objective of this study was to evaluate the effect of optimal anticoagulation in patients using VKA during the first year after surgery, on long-term clinical outcome in patients who underwent CABG for disabling angina. For this, in all patients the individually achieved quality of anticoagulation was classified as “optimal VKA” or “non-optimal VKA”. Differences in long-term clinical outcome


- 80 -

was evaluated using survival analysis. The event-free survival was graphically depicted using the method of Kaplan-Meier. A multivariable Cox proportional hazards regression analysis was performed to estimate adjusted hazard ratios (HR).

Predefined risk factors for adverse clinical outcome included age, sex, body mass index, smoking, diabetes mellitus, hypertension, hypercholesterolemia, history of arterial thrombosis, NYHA classification, concomitant cardiovascular medication, severity of vessel disease, type of grafts used (vein or arterial) and number of grafts used.

Secondary survival analyses were performed on separate components of the composite, as well as on repeat revascularisation, using the same strategy as in the primary analysis.

All tests performed to test the (null-) hypothesis of no differences between groups, were two-sided. A p-value < 0.05 was considered statistically significant. For all analyses, commercially available computer software (Statistical Analysis System version 9.1, SAS Institute, Cary, North Carolina) was used. RESULTS Patients’ characteristics In the original CABADAS study, 307 patients were randomized to VKA. Of these, 244 patients were from Dutch centers. In 9 patients, VKA was not continued post-operatively, leaving 235 patients treated with VKA during the one-year follow-up period of the original study eligible for the extended long-term follow-up study.

Of the 9 patients not included in our study, VKA was stopped for surgical reasons in two patients. Furthermore, one patient was diagnosed with malignancy requiring chemotherapy, one patient suffered from a TIA and another from a major bleeding, all during initial hospitalisation. In four patients VKA was not continued post-operatively for unknown reasons.

The baseline characteristics are summarized in Table 1. Eighty-eight % of the 235 patients were males and the mean age at the time of surgery was 58 years. All patients had disabling angina with 57 % NYHA class III and 17% NYHA class IV. Fifty-one % had a previous myocardial infarction, hypertension was diagnosed in 39% and hyperlipidaemia in 42%. Bypass grafting was performed on two coronary arteries

Chapter 5

- 81 -

in 36% and on all three coronary arteries in 64% of patients. In 58% of the patients, venous grafts were combined with an arterial graft.

Table 1 Baseline characteristics.

Optimal VKA †

N=93

Non-optimal VKA † N=142

Total

N=235

P-value

Patients’ characteristics Age, years *Male gender, % Weight , kg * Height, cm * Body mass index (kg/m2)Angina NYHA class, % II III IV Myocardial infarction, % CVA or TIA, % Hypertension, % Diabetes mellitus, % Hyperlipidaemia, % Smoking (current/past), % Concomitant drugs, % beta blockers calcium antagonists nitrates triple therapy none single double triple diuretics lipid lowering drugs ACE inhibitors

59 ± 7.0

84 78 ± 9.6 173 ± 8.3 26.1 ± 2.7

23 56 21 52 2 46 10 40

33/54

33 14 1 53 46 1 0 16 6 6

58 ± 7.8

91 77 ± 10.2 173 ± 7.6 25.9 ± 2.7

28 57 15 51 1 34 9 44

43/45

37 13 1 53 44 3 0 11 8 4

58 ± 7.5

88 78 ± 9.9 173 ± 7.9 26.0 ± 2.7

26 57 17 51 1 39 9 42

39/49

36 13 1 53 45 2 0 13 7 5

0.58 0.15 0.68 0.77 0.99 0.32

0.89 0.56 0.08 1.00 0.59 0.28

0.58 0.84 1.00 0.50

0.33 0.80 0.35

Surgical data Vessel disease, % single double tripleNumber of grafts * Vein plus arterial grafts, % Heparin, 103 U *

0 38 62

2.2 ± 0.7 51

28.9 ± 6.9

0 35 65

2.3 ± 0.7 63

27.2 ± 7.0

0 36 64

2.2 ± 0.7 58

27.8 ± 7.0

0.68

0.26 0.06 0.06

* mean ± SD; † Vitamin K antagonists.


- 82 -

VKA therapy was initiated one day prior to surgery and continued on the first postoperative day. In all patients acenocoumarol or phenprocoumon was used. After discharge, patients were referred to the specialized Thrombosis Services for monitoring the VKA therapy. On average, 19 (SD, 5.8) INR assessment were performed during the first year of VKA therapy. As shown in Table 2, the mean INR achieved was 3.1 (SD, 0.43), which was at the lower end of the target range INR 2.8 to 4.8 that was used during VKA monitoring. In the first year the individual time spend within the target range was 56%, below target range 39% and above target range 5%. Using the range of INR 3.0 to 4.0, the individual time within this range was 36%. The percentage of time spend below INR 3.0 was 49% and time spend above INR 4.0 was 15%. Figure 1 shows the percentage of patients below, within and above INR 3.0 to 4.0 during the first year of treatment. Table 2 Anticoagulation data.

Optimal VKA † N=93

Non-optimal VKA † N=142

Total

N=235

P-value

Anticoagulant data

INR assessments 17 ± 4.2 20 ± 6.1 19 ± 5.8 < 0.001

Mean INR 3.2 ± 0.28 3.0 ± 0.48 3.1 ± 0.43 < 0.001

ITTR, % 49 ± 13.3 28 ± 12.9 36 ± 16.7 < 0.001

Time above INR 4.0, % 16 ± 12.1 14 ± 13.3 15 ± 12.8 0.16

Time below INR 3.0, % 35 ± 15.4 58 ± 21.0 49 ± 21.9 < 0.001

Variance growth rate 0.08 ± 0.04 0.31 ± 0.79 0.22 ± 0.63 < 0.001

30 days ITTR, % 40 ± 31.4 21 ± 23.6 28 ± 28.5 < 0.001

30 days above INR 4.0, % 13 ± 20.7 17 ± 24.8 16 ± 23.3 0.29

30 days below INR 3.0, % 47 ± 36.7 62 ± 35.4 26 ± 36.6 0.004 * † Vitamin K antagonists.

Chapter 5

- 83 -

Achieved level of anticoagulation To assess the impact of the achieved level of anticoagulation on the occurrence of cardiac death and myocardial infarction, patients were categorized as having “optimal” or “non-optimal” VKA, using both the percentage of time below INR 3.0 and the variability of INRs during the first year of VKA therapy.

As shown in Table 2, 93 patients (40%) had optimal VKA therapy, whereas 142 patients (60%) were classified as non-optimal. Of these, 17 patients (7%) were both unstable and frequently below INR 3.0, 63 patients (27%) were stable but frequently below INR 3.0, and 62 patients (26%) were frequently above INR 3.0 but unstable.

Figure 1 Quality of VKA therapy during the first year, percentage of time below

(dark grey), within (light grey) and above (black) INR 3.0 to 4.0.

In the patients with optimal VKA, the mean individual time spend within INR 3.0 to 4.0 during the first year after the CABG was 49%, with 35% below INR 3.0 and 16% above INR 4.0, whereas in patients with non-optimal VKA, this was 28%, 58% and 14% respectively. This difference between patient with optimal and non-optimal VKA was already present early after the initiation of VKA therapy, as illustrated by 47% vs. 62% of time spend below INR 3.0 and 40% vs. 21% of time spend within INR 3.0-4.0 during the first month respectively.

VKA treatment duration (months)

0 1 2 3 4 5 6 7 8 9 10 11 12Patie

nts b

elow

, with

in a

nd a

bove

INR

3.0

to 4

.0 (%

)

10

20

30

40

50

60

70

80

90

100


- 84 -

In this, the risk of non-optimal VKA was 2.5 higher (OR=2.5; 95%CI 1.3 to 4.8; P=0.012) in patients who spend the most time below INR 3.0 during the first month (80 to 100%, highest tertile), compared to patients who were the least frequent below INR 3.0 (0% to 33%, lowest tertile). For the patients in the middle tertile (33 to 80%), with an odds ratio of 1.5 (0.75 to 2.9; P=0.82), the risk of non-optimal VKA was not statistically significantly different from the patients with the least time below INR 3.0 during the first month.

As also shown in Table 1, at baseline there were only a view differences between the patients in the optimal and non-optimal VKA group. In the non-optimal group there where more males (91% vs. 84%; P=0.15), more patients had hypertension (46% vs. 34%; P=0.08) and Ca antagonists were more frequently used (73% vs. 59%; P=0.04). At discharge, the use of Ca antagonists was not significantly different (13% vs. 14%; P=0.84). There were no differences in the number of coronary arteries that received grafts (65% vs. 62% triple vessel; P=0.68), however, arterial grafts were more frequently used in the non-optimal group (63% vs. 51%; P=0.06). In patients with non-optimal VKA, the amount of heparin used during surgery was on average 1675 U less than in the optimal group (27193 vs. 28868 U; P=0.08). Continuation of treatment VKA was prematurely discontinued in 17 patients, in 5 patients due to a major bleeding, in 4 patients after the occurrence of a thrombotic event, in 1 patients after developing a contra-indication for VKA and 7 patients stopped VKA without medical reason. After one year, when the CABADAS study had ended, continuation the allocated VKA treatment was left at the discretion of the treating physician. At the time of first extended follow-up after on average 2.5 years, 48% of the patients were still using VKA in both patients with optimal and non-optimal VKA therapy during the first year (P=1.0). Of the patients no longer using VKA, 31% vs. 30% switched to aspirin (3% vs. 3% combined with dipyridamol) and 20% vs. 21% stopped taking anticoaguant/antiplatelets drugs, in patients with optimal vs. non-optimal VKA respectively.

Regarding concomitant cardiovascular medication, particularly the use of lipid lowering drugs and ACE inhibitors had increased from 7% to 28% and 5% to 15% respectively. Furthermore, the use of nitrates had increased from 1% to 10%,

Chapter 5

- 85 -

consistent with the presence of anginal complaints at the first extended follow-up evaluation, i.e. 89% NYHA I and 11% NYHA II. Triple drug therapy (nitrates, betablockers and calcium antagonists) had increased accordingly, without a significant difference between patients with optimal and non-optmal VKA. Cardiac death and myocardial infarction The median duration of the extended follow-up was 12.7 years (range 1.0 to 14.8) for the “optimal VKA” group and 12.6 years (range 0.25 to 15.1) for the “non-optimal VKA” group. Cardiac death or myocardial infarction had occurred in 72 patients (31%). Figure 2 Kaplan Meier of the composite of cardiac death and myocardial infarction after

coronary-artery-bypass-graft surgery.

As shown in Figure 2, over a 14-year period, cumulative rate of the composite of cardiac death and myocardial infarction was 44% in patients with non-optimal VKA and 26% in patients with optimal VKA. Up to 8 years, the risk of cardiac death and myocardial infarction was comparable. Thereafter, the risk started to differ between optimal vs. non-optimal VKA, indicating a time-dependence in the association between quality of VKA and the occurrence of cardiac death and myocardial infarction. Taking this time-dependence into account in the survival analysis (significant treatment*time interaction), there was no significant difference

Time (years)0 2 4 6 8 10 12 14

Com

posit

e of

car

diac

dea

th a

nd m

yoca

rdial

infa

rctio

n

0%

10%

20%

30%

40%

50%

60% Optimal VKA Non-optimal VKA

HR = 0.90

(95%CI 0.47 to 1.7; P = 0.75)

Non-optimal VKA: 142 Optimal VKA: 93

HR = 3.4

(95%CI 1.4 to 8.; P = 0.007)

81 77 636769 59 13122 111 88100102 78 16


- 86 -

during the first 8 years (HR=0.90; 95%CI 0.47 to 1.7; P=0.75), but after 8 years patients with non-optimal VKA had a 3.4-fold increased risk (95%CI 1.4 to 8.1; P=0.007), compared to patients with optimal VKA.

Additional risk indicators were “history of myocardial infarction” (HR=1.7 , 95%CI, 1.03 to 2.7; P=0.039), “age above 60 years” (HR=1.7, 95%CI, 1.04 to 2.7; p=0.034) and “use of beta blockers” (HR=0.59 , 95%CI, 0.36 to 0.97; P=0.039). The “use of lipid lowering drugs” was borderline significant (HR=0.39 , 95%CI, 0.14 to 1.1; P=0.069). Figure 3 Kaplan Meier of repeat revascularisation, cardiac death and myocardial infarction,

separately.

Repeat revascularisation Regarding repeat revascularisation, 51 patients underwent repeat revascularisation, without a difference between patient with non-optimal and optimal VKA. After 14 years, the cumulative rates of repeat revascularisation for patients with non-optimal versus optimal VKA were 34% and 35%, respectively, indicating that there was no association between the quality of anticoagulant therapy and repeat revascularisation (HR=0.87, 95%CI 0.50 to 1.5; P=0.63) (see Figure 3). In this, age below 60 years at the time of surgery (HR=2.2, 95%CI 1.2 to 4.1, P=0.009) was the dominant determinant for repeat revascularisation, with more frequent repeat revascularisation in the younger patients.

In addition to repeat revascularisation, Figure 3 also show the cumulative rates of cardiac death and myocardial infarction separately.

0 2 4 6 8 10 12 14

Card

iac d

eath

0%

10%

20%

30%

40%Optimal VKA Non-optimal VKA

Time (years)0 2 4 6 8 10 12 14

Myo

card

ial in

farc

tion

0%

10%

20%

30%

40%Optimal VKANon-optimal VKA

0 2 4 6 8 10 12 14

Repe

at re

vasc

ular

isatio

n

0%

10%

20%

30%


Time (years)Time (years)

Chapter 5

- 87 -

DISCUSSION We assessed the impact of the quality of anticoagulation during the first year of VKA therapy on long-term clinical outcome of patients undergoing coronary-artery-bypass-graft surgery. Considering VKA therapy, initially, the majority of patients were under-anticoagulated, with a rapid improvement during the first 4 weeks. Thereafter, the percentage of patients above, within and below the target range stabilised. This difficulty to achieve adequate anticoagulation in the initial phase is also reported elsewhere [10,11]. To assess quality of anticoagulation, patients were classified as optimal or non-optimal VKA, based on the individual percentage of time spend below INR 3.0 and the stability of consecutive INRs. This is the lower limit of the intensity level for primary and secondary prophylaxis of arterial thromboembolic complications currently advocated by the Dutch Federation of Thrombosis Services. This is slightly narrower than the target range of INR 2.8 to 4.8 that was actually used for monitoring and subsequent adjustments of VKA therapy at the time of the CABADAS study. Therefore, the percentage patients with non-optimal VKA does not reflect the performance of the monitoring per se.

Using our definition, we allocated 40% of the patients to optimal VKA therapy. These patients were most frequently above the therapeutic threshold of INR 3.0 with the least frequent fluctuations on consecutive days, theoretically benefiting the most for secondary prophylaxis with VKA therapy. Non-optimal VKA was not associated with a lack of INR monitoring. In fact, with 20 versus 17 INR assessments, patients with non-optimal VKA were significantly more frequently contacted during the first year.

Overall, the cumulative rate of the composite of cardiac death and myocardial infarction was 37% and for repeat revascularisation 35%. Considering that studies with follow-up up to 14 years are limited and taking the difference in follow-up into account, the cumulative rates observed in our study were comparable with those from previous reports [12-16].

Addressing the risk of cardiac death and myocardial infarction, optimal VKA was associated with a risk reduction. This beneficial effect of optimal VKA became apparent 8 years after CABG surgery (hazard ratio 3.4), whereas during the first 8 years no differences were observed. As suggested by others, clinical outcome after initial CABG might be influenced more strongly by the extension of the native


- 88 -

coronary artery disease (CAD), rather than disease progression at the site of grafts. This extension of CAD might be reduced by optimal VKA [16,17]. For both cardiac death and myocardial infarction, patients with non-optimal VKA showed a more than gradual increase of adverse clinical outcome after 8 years, significantly different from the patients with optimal VKA.

The more than gradual increase was also observed for repeat revascularisation, however in both patients with non-optimal and optimal VKA. As one would rather expect non-inferiority regarding cardiac death and myocardial infraction, at the cost of an increased need of repeat revascularisation, this finding is somewhat puzzling [14,18-20].

Due to the retrospective setting of the extended follow-up of the CABADAS patients, which is a important limitation of our study, we were unable to further assess the apparent lack of association between achieved quality of anticoagulation and repeat revascularisation. Both treatment and clinical outcome were evaluated at the first extended follow-up at 2.5 years. But after this first follow-up, only clinical outcome was thoroughly assessed but detailed information on continuation of medical treatments as well as co-morbidity were no longer available. Therefore, we can only speculate regarding the lack of effect of optimal VKA on repeat revascularisation. As the overall high cumulative rate of repeat revascularisation at 14 years was high (35%), and doubled over the last 4 years of follow-up, patients might have been referred to cardiac specialists more quickly after first presentation of extension of CAD, due to their frequent contacts with treating physicians. This lower threshold for referral could have limited the potential impact of optimal VKA on the “need” for repeat revascularisation. Furthermore, especially patients below 60 years at the time of initial CABG (and 68 years at time of increase in rate repeat revascularisation) were at increased risk. One could hypothesize that in these younger patients, the threshold for referral and consequent repeat revascularisation was lower than for the elderly patients.

Of note, a comparison of all VKA patients with the patients randomised to aspirin alone or aspirin combined with dipyridamole (from original CABADAS study) showed a significantly lower cumulative rate of 18% and 17% in patients treated with aspirin alone or aspirin combined with dipyridamole respectively, without a differences in cardiac death and myocardial infarction (data not shown). In these

Chapter 5

- 89 -

patients, frequent follow-up visits related to the control of their anticoagulant treatment were not required.

Remarkably, also in 50% of the patients with non-optimal VKA, the treatment was continued, clearly indicating that the decision to continue VKA was not based on information regarding the previously achieved (poor) quality of anticoagulation. At the time of the decision, such information was not provided to the treating physician. Conclusions Late clinical outcome is strongly affected by the achieved quality of anticoagulation. Considering the limited number of patients with an optimal response to VKA, in patients undergoing coronary-artery-bypass-graft surgery, VKA should not be considered as first choice. If indicated because of concomitant condition, e.g. heart valve replacement, VKA combined with aspirin is suggested, which could especially be important in patients with non-optimal VKA therapy [1]. In this, early recognition of non-optimal response is important in assessing the appropriateness of VKA therapy.


- 90 -

References 1. Becker RC, Meade TW, Berger PB, Ezekowitz M, O'Connor CM, Vorchheimer DA, Guyatt GH,

Mark DB, Harrington RA. The primary and secondary prevention of coronary artery disease: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008; 133: 776S-814S.

2. Eagle KA, Guyton RA, Davidoff R, Edwards FH, Ewy GA, Gardner TJ, Hart JC, Herrmann HC, Hillis LD, Hutter AMJr, Lytle BW, Marlow RA, Nugent WC, Orszulak TA, Antman EM, Smith SCJr, Alpert JS, Anderson JL, Faxon DP, Fuster V, Gibbons RJ, Gregoratos G, Halperin JL, Hiratzka LF, Hunt SA, Jacobs AK, Ornato JP. ACC/AHA 2004 guideline update for coronary artery bypass graft surgery: summary article. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1999 Guidelines for Coronary Artery Bypass Graft Surgery). J Am Coll Cardiol 2004; 44: e213-e310.

3. Antiplatelet Trialists' Collaboration. Collaborative overview of randomised trials of antiplatelet therapy--II: Maintenance of vascular graft or arterial patency by antiplatelet therapy. BMJ 1994; 308: 159-168.



6. Veeger NJ, Piersma-Wichers M, Tijssen JG, Hillege HL, van der Meer J. Individual time within target range in patients treated with vitamin K antagonists: main determinant of quality of anticoagulation and predictor of clinical outcome. a retrospective study of 2300 consecutive patients with venous thromboembolism. Br J Haematol 2005; 128: 513-519.


8. Fihn SD, Callahan CM, Martin DC, McDonell MB, Henikoff JG, White RH. The risk for and severity of bleeding complications in elderly patients treated with warfarin. The National Consortium of Anticoagulation Clinics. Ann Intern Med 1996; 124: 970-979.

9. van Leeuwen Y, Rosendaal FR, Cannegieter SC. Prediction of hemorrhagic and thrombotic events in patients with mechanical heart valve prostheses treated with oral anticoagulants. J Thromb Haemost 2008; 6: 451-456.

10. Ansell J, Hirsh J, Hylek E, Jacobson A, Crowther M, Palareti G. Pharmacology and management of the vitamin K antagonists: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008; 133: 160S-198S.

11. Veeger NJ, Piersma-Wichers M, Hillege HL, Crijns HJ, van der Meer J. Early detection of patients with a poor response to vitamin K antagonists: the clinical impact of individual time within target range in patients with heart disease. J Thromb Haemost 2006; 4: 1625-1627.

12. Cameron AA, Green GE, Brogno DA, Thornton J. Internal thoracic artery grafts: 20-year clinical follow-up. J Am Coll Cardiol 1995; 25: 188-192.

Chapter 5

- 91 -

13. Daemen J, Boersma E, Flather M, Booth J, Stables R, Rodriguez A, Rodriguez-Granillo G, Hueb WA, Lemos PA, Serruys PW. Long-term safety and efficacy of percutaneous coronary intervention with stenting and coronary artery bypass surgery for multivessel coronary artery disease: a meta-analysis with 5-year patient-level data from the ARTS, ERACI-II, MASS-II, and SoS trials. Circulation 2008; 118: 1146-1154.

14. The BARI Investigators. The final 10-year follow-up results from the BARI randomized trial. J Am Coll Cardiol 2007; 49: 1600-1606.

15. van Brussel BL, Voors AA, Ernst JM, Knaepen PJ, Plokker HW. Venous coronary artery bypass surgery: a more than 20-year follow-up study. Eur Heart J 2003; 24: 927-936.

16. Gupta A, Burke J, Bove A. Coronary arterial revascularization: past, present, future: part I--historical trials. Clin Cardiol 2006; 29: 290-294.

17. Alderman EL, Kip KE, Whitlow PL, Bashore T, Fortin D, Bourassa MG, Lesperance J, Schwartz L, Stadius M. Native coronary disease progression exceeds failed revascularization as cause of angina after five years in the Bypass Angioplasty Revascularization Investigation (BARI). J Am Coll Cardiol 2004; 44: 766-774.

18. The Bypass Angioplasty Revascularization Investigation (BARI) Investigators. Comparison of coronary bypass surgery with angioplasty in patients with multivessel disease. N Engl J Med 1996; 335: 217-225.

19. Serruys PW, Unger F, Sousa JE, Jatene A, Bonnier HJ, Schonberger JP, Buller N, Bonser R, van den Brand MJ, van Herwerden LA, Morel MA, van Hout BA. Comparison of coronary-artery bypass surgery and stenting for the treatment of multivessel disease. N Engl J Med 2001; 344: 1117-1124.

20. Serruys PW, Ong AT, van Herwerden LA, Sousa JE, Jatene A, Bonnier JJ, Schonberger JP, Buller N, Bonser R, Disco C, Backx B, Hugenholtz PG, Firth BG, Unger F. Five-year outcomes after coronary stenting versus bypass surgery for the treatment of multivessel disease: the final analysis of the Arterial Revascularization Therapies Study (ARTS) randomized trial. J Am Coll Cardiol 2005; 46: 575-581.

Chapter 6

- 93 -

Chapter 6


or Aspirin Combined with Dipyridamole Not Superior to Aspirin Alone;

Fourteen Year Follow-up Results of CABADAS

Nic J.G.M. Veeger

Felix Zijlstra

Hans L. Hillege

Jan van der Meer

Submitted

CABADAS 14-year follow-up study

- 94 -

SUMMARY Background. Secondary prophylaxis using aspirin is standard of care after coronary-artery-bypass-graft surgery. Limited data are available for long-term results. We evaluated the effect of aspirin, aspirin with dipyridamole and vitamin K antagonists (VKA) on 14-year clinical outcome of patients included in the CABADAS study.

Methods. All 726 Dutch patients in whom antithrombotic therapy with aspirin (N=248), aspirin with dipyridamole (N=234) or VKA (N=244) was randomly allocated, were included. Primary endpoint was occurrence of Major Adverse Cardiac Events (MACE). Outcome was retrospectively evaluated during 14 year follow-up.

Results. Cumulative incidences for MACE over 14 years were 49%, 50% and 59% in patients treated with aspirin, aspirin with dipyridamole and VKA, respectively. Although overall occurrence of MACE did not significantly differ between the three treatment groups (P=0.12), patients treated with VKA were at higher risk of MACE than patients treated with aspirin with dipyridamole (HR=1.3, 95%CI, 1.0-1.8; P=0.041) and patients treated with aspirin alone (HR=1.1, 95%CI, 0.86-1.5; P=0.37). This difference was attributed to an increased risk of repeat revascularization in patients treated with VKA, without any differences in cardiac death and myocardial infarction between the three treatment groups. However, the observed high rate of repeat revascularization in patients treated with VKA could reflect an a priori increased probability for repeat revascularization due to the specific conditions surrounding VKA therapy (i.e. more intense patient-doctor contacts).

Conclusions. This study with 14-year clinical outcome provide further evidence for the use of aspirin as secondary prophylaxis after coronary-artery-bypass-graft surgery.

Chapter 6

- 95 -

INTRODUCTION Several prospective controlled clinical trials have evaluated efficacy and safety of antithrombotic drug therapies for secondary prophylaxis after coronary-artery-bypass-graft surgery to prevent vein graft closure [1]. This resulted in a high level of evidence, identifying aspirin as standard of care [2,3]. To maximize the effect of anticoagulant treatment it must be initiated already early after surgery [4-6]. In this respect, current guideline stipulates the initiation of aspirin within 48 hours after surgery and to be continued indefinitely [4,5]. When started more than 48 hours after surgery, efficacy of aspirin is hampered and early graft occlusion more frequently observed [7].

One of the clinical trials was CABADAS (CABADAS = prevention of Coronary Artery Bypass graft occlusion by Aspirin, Dipyridamole and Acenocoumarol/phenprocoumon Study). CABADAS evaluated the effect of aspirin, aspirin with dipyridamole and vitamin K antagonists (VKA) on one-year graft patency and early clinical outcome. The CABADAS study concluded that there was no convincing evidence that the addition of dipyridamole to a low dose of aspirin improves one-year vein-graft patency after coronary artery bypass grafting. Oral anticoagulants, compared with aspirin, provided no benefit [8].

In many of the clinical trials that contributed to the evidence of aspirin as primary choice for secondary prophylaxis after coronary-artery-bypass-graft surgery, data on long-term clinical outcome are limited. As in CABADAS, mainly early effects on graft patency and short- and mid-term clinical outcome are available. Therefore, in the present study we evaluated the clinical outcome of patients included in the CABADAS study after an extended follow-up up to 14 years. These patients were originally treated with aspirin, aspirin with dipyridamole or VKA. PATIENTS AND METHODS The Extended CABADAS study is a retrospective follow-up study in 726 Dutch participants of the CABADAS study (CABADAS = prevention of Coronary Artery Bypass graft occlusion by Aspirin, Dipyridamole and Acenocoumarol/ phenprocoumon Study). A priori, it was decided to only include the Dutch participants for the extension of the CABADAS study, for logistical reasons. The original CABADAS study was an international multicenter randomized controlled


- 96 -

clinical trial evaluating the effects of low dose aspirin, aspirin combined with dipyridamole and vitamin K antagonists on graft patency and clinical outcome at one year. In this study, 912 patients were evaluated after one year of study treatment. The design of the study is described in great detail elsewhere [8]. Initial study protocol was approved by the Ethics Committee of each participating center and all patients gave written informed consent. Patients All patients underwent elective CABG for disabling angina between July 1987 and August 1990. Surgery was performed using routine techniques in each participating center. They all received saphenous vein grafts, and in addition, internal mammary artery (IMA) grafts could be used at the discretion of the surgeon. Heparin, administrated during surgery was antagonized by protamine sulphate at the end of the procedure, unless it was continued because of the introduction of an intra-aortic balloon pump. Main exclusion criteria were age over 70 years, previous or concurrent cardiac surgery, unstable angina at the time of surgery (< 2 days) or myocardial infarction within 7 days prior to surgery and severe concomitant disease. Randomization Study treatment was initiated after random allocation, i.e. aspirin, aspirin with dipyridamole or vitamin K antagonists (acenocoumarol or phenprocoumon) to all eligible patients, prior to surgery. To ensure a balanced treatment in all participating hospitals, allocation was stratified by study site. Follow up After discharge from the hospital, patients were seen for the duration of one year at a 3-month interval. The extended long-term follow-up was retrospectively performed for all patients up to June 2002, through telephonic interviews of the patients using a standard questionnaire and by reviewing the medical files of their cardiologists and/or general practitioners. This evaluation was performed for the first time 2 to 3 years after surgery and repeated 12 years thereafter.

Chapter 6

- 97 -

Clinical endpoints The primary endpoint of the extended follow-up study was the composite of Major Adverse Cardiac Events (MACE), i.e. cardiac mortality, myocardial infarction and coronary repeat revascularization. Cardiac mortality was defined as mortality due to acute myocardial infarction, acute cardiac arrest, sudden death and mortality caused by progressive congestive heart failure. In case of an unknown cause of death, it was classified as cardiac death. For myocardial infarction, pathological Q-waves had to be present on the ECG in combination with CPK and CPK-MB serum levels that exceeded the upper limits of normal ranges. Repeat revascularization was defined as percutaneous coronary intervention (PCI) or redo-CABG during follow-up.

From the original CABADAS study, patients’ characteristics, general and disease specific medical history and clinical outcome events were available up to one year after discharge. Adjudication of these events was performed by an blinded classification committee, without knowledge of the patient’s treatment assignment. Anticoagulation data Patients allocated to VKA therapy were referred to specialized Thrombosis Services for frequent control of the level of anticoagulation. The VKA of choice was acenocoumarol or phenprocoumon. The target range used at the time of the original study was INR 2.8 to 4.8. Statistical analysis The primary objective of the study was to evaluate the effect of three anti-thrombotic drugs on long-term clinical outcome in patients who underwent CABG for disabling angina. Differences between groups were evaluated by survival analysis using an Intention-To-Treat approach. In this, patients were evaluated by their originally allocated treatment. The event-free survival was graphically depicted using the method of Kaplan-Meier. A multivariable Cox proportional hazards regression analysis was performed to estimate adjusted hazard ratios (HR). Predefined risk factors for MACE were age, sex, Body Mass Index, smoking, diabetes mellitus, hypertension, hypercholesterolemia, history of arterial thrombosis, NYHA classification, concomitant cardiovascular medication, severity of vessel disease, type of grafts used (vein or arterial) and number of grafts used.


- 98 -

Secondary survival analyses were performed on separate components of MACE using the same strategy as in the primary analysis, as well as a sensitivity analysis based on an “As Treated” approach. Table 1 Baseline characteristics of patients and anticoagulation data. Aspirin

N=248

Aspirin + Dipyridamole

N=234

VKA†

N=244

Total

N=726

P-value

Patients’ characteristics Age, years * Male gender, % Weight, kg * Height, cm * Angina NYHA class, % II III IV Myocardial infarction, % CVA or TIA, % Hypertension, % Diabetes mellitus, % Hyperlipidemia, % Smoking (current/past), % Concomitant drugs at admission, % beta blockers calcium antagonists nitrates diuretics lipid lowering drugs ACE inhibitors digitalis

59 ± 7.7

86 77 ± 10.4 173 ± 7.6

29 55 16 52 4 29 8 35

36/50

80 66 81 15 10 4 2

59 ± 7.7

82 77 ± 10.2 173 ± 8.1

26 56 18 48 3 30 9 34

34/46

76 62 80 13 10 4 2

58 ± 7.5

88 78 ± 9.9 173 ± 7.8

25 57 18 51 1 37 9 43

40/48

77 64 83 17 11 5 2

58 ± 7.6

86 78 ± 10.1 173 ± 7.9

27 56 17 51 3 32 9 37

37/48

78 64 81 15 10 4 2

0.46 0.21 0.93 0.99 0.92

0.62 0.16 0.11 0.93 0.12 0.15

0.46 0.75 0.61 0.53 0.95 0.99 0.99

Surgical data Vessel disease, % single double tripleNumber of grafts * Graft type, % vein grafts only vein plus arterial grafts Concomitant drugs at discharge, % beta blockers calcium antagonists nitrates diuretics lipid lowering drugs ACE inhibitor digitalis

3 39 58

2.2 ± 0.9 29 61

36 8 1 15 5 2 13

1 39 60

2.5 ± 1.2 29 61

36 10 1 16 3 3 15

2 33 65

2.3± 0.7 42 58

35 14 1 14 7 5 12

2 37 61

2.3 ± 0.9 40 60

36 11 1 15 5 3 13

0.23

0.003 0.70

0.97 0.11 0.83 0.76 0.091 0.28 0.54

* mean ± SD; † Vitamin K antagonists.

Chapter 6

- 99 -

All analyses performed to test the (null-) hypothesis of no differences between groups, were two-sided. A p-value < 0.05 was considered statistically significant. For all analyses, commercially available computer software (Statistical Analysis System version 9.1, SAS Institute, Cary, North Carolina) was used.

RESULTS Patients’ characteristics The study cohort comprised of 726 patients from the 912 eligible patients in the original CABADAS study. These were the Dutch patients, of whom 248 were allocated to receive aspirin plus placebo, 234 aspirin with dipyridamole and 244 vitamin K antagonists. Baseline characteristics of the patients are summarized in Table 1. Eighty-eight % of the patients were males and the mean age at the time of surgery was 59 years. All patients had disabling angina, with 56 % NYHA class III and 17% NYHA class IV. Fifty-one % had a previous myocardial infarction, hypertension was diagnosed in 32% and hyperlipidemia in 37%. Bypass grafting was performed on all three coronary arteries in 61%, in 37% on two of the three arteries and a single artery in 2% of patients. Arterial grafts, together with vein grafts, were used in 60% of patients. Due to the randomized study design with a balanced treatment allocation over participating study sites, the three treatment groups were highly comparable at baseline. Clinical outcome MACE Median duration of the extended follow-up was 12.6 years (range 0.08 -14.8) for the aspirin group, 12.7 years (range 0.02 - 14.8) for the aspirin with dipyridamole group and 12.4 years (range 0.04 - 15.1) for the VKA group.

The cumulative incidence of MACE over the 14-year period was 49% in patients treated with aspirin, 50% in patients treated with aspirin with dipyridamole and 59% in patients treated with VKA. These observed differences between aspirin, aspirin with dipyridamol and VKA were not significant (P=0.12; from multivariable Cox regression), although the patients treated with VKA were at higher risk for MACE than the patients treated with aspirin with dipyridamole and aspirin (HR=1.3, 95%CI, 1.01 to 1.8; P=0.041 and HR=1.1, 95%CI, 0.86 to 1.5; P=0.37, respectively;


- 100 -

aspirin compared to aspirin with dipyridamole HR=1.2, 95%CI, 0.89 to 1.6; P=0.24; see Figure 1). Figure 1 Kaplan Meier of the composite of cardiac death, myocardial infarction or repeat

revascularization (MACE) after coronary-artery-bypass-graft surgery.

As presented in Figure 2, significantly associated with MACE were Age at time of surgery (HR 56 to 62 years = 1.2, 95%CI, 0.90 to 1.6, P=0.22 and HR > 62 years = 1.4, 95%CI, 1.1 to 1.9; P=0.009, compared to patients up to the age of 55 years), NYHA class III or IV (HR=1.4, 95%CI, 1.1 to 1.8; P=0.018), Use of digitalis (HR=2.3, 95%CI, 1.3 to 4.3; P=0.006), Beta-blocking agents (HR=0.73, 95%CI, 0.56 to 0.94; P=0.014), and Lipid lowering drugs (HR=0.55, 95%CI, 0.35 to 0.87; P=0.011), all prior to surgery and Type of graft used during surgery (HR vein graft only = 1.3, 95%CI, 1.1 to 1.6; P=0.022). The number of grafts used during surgery were not associated with MACE (HR 1 or 2 grafts vs. 3 or more = 1.0, 95%CI, 0.8 to1.3; p=0.75). Components of MACE Regarding the components of MACE separately, cardiac death had occurred in 135 patients with a cumulative incidence over the 14-year period of 23%. In the patients

Time (years)0 2 4 6 8 10 12 14

MA

CE

0%

10%

20%

30%

40%

50%

60%AspirinAspirin + dipyridamoleVKA

P = 0.12

248234244

208205200

192188183

181171162

158157156

139142134

115125116

151916

Patients at risk:

Chapter 6

- 101 -

treated with aspirin, aspirin with dipyridamole and VKA, this was 21%, 24% and 25%, respectively (HR=0.96 (95%CI, 0.64 to 1.5; P=0.85) and HR= 1.0 (95%CI, 0.67 to 1.5; P=0.94) of aspirin and aspirin with dipyridamole, compared to VKA, P=0.96 (aspirin compared to aspirin with dipyridamole HR=0.95, 95%CI, 0.62 to 1.4; P=0.79); see Figure 3). Figure 2 Univariate and multivariable results of Cox proportional hazards regression analysis

of MACE.

For myocardial infarction, which had occurred in 177 patients (82 fatal), the cumulative incidence over the 14-year period was 31%. In the patients treated with aspirin, aspirin with dipyridamole and VKA, this was 33%, 29% and 30%, again not significantly different (HR= 1.1 (95%CI, 0.80 to 1.6 P=0.46) and HR=0.96 (95%CI, 0.66 to 1.4; P=0.82) for aspirin and aspirin with dipyridamole, compared to VKA, P=0.60 (aspirin compared to aspirin with dipyridamole HR=1.2, 95%CI, 0.83 to 1.7; P=0.34); see Figure 3).

0.2 0.5 2 50.1 1.0 10.0

Variable

Allocated treatment AspirinAspirin + dipyridamoleVitamin K antagonists

0.87 (0.67-1.14)0.78 (0.59-1.03)

Reference

UnivariateHazard Ratio

(95%CI)

AdjustedHazard Ratio

(95%CI)

Univariate Hazard Ratio

0.300.078

P P

Age 56-62 years> 62 years

up to 55 years

1.2 (0.90-1.6)1.6 (1.2-2.1)

Reference

0.21<0.001

Sex 1.2 (0.86-1.6) 0.31Smoking 1.3 (0.97-1.9) 0.079NYHA 1.4 (1.1-1.8) 0.018

Female

Class III or IVDiabetes Mellitus 1.2 (0.82-1.8) 0.35Body Mass Index (kg/m2) 0.99 (0.95-1.03) 0.70Previous ATE (AMI or stroke) 1.2 (0.93-1.5) 0.187Hyperlipidaemia 0.83 (0.66-1.05) 0.121Hypertension 0.99 (0.78-1.3) 0.92Type of graft used 1.4 (1.1-1.7) 0.008Vein graft only

IMA + vein grafts ReferenceVessel disease 1.03 (0.46-2.3) 0.94Triple vessel

Two vesselSingle vessel

1.1 (0.49-2.5) 0.81Reference

Digitalis 3.1 (1.8-5.6) <0.001Lipid lowering drugs 0.56 (0.35-0.87) 0.011Beta blocking agents 0.70 (0.54-0.90) 0.005Diuretics 1.3 (0.96-1.7) 0.087

0.89 (0.68-1.16)0.75 (0.57-0.99Reference

0.370.041

1.2 (0.90-1.6)1.4 (1.1-1.9)Reference

0.220.009

--

1.4 (1.1-1.8) 0.018-----

1.3 (1.1-1.8) 0.018Reference

--

2.3 (1.3-4.3) 0.0060.55 (0.35-0.87) 0.011

--


- 102 -

Repeat revascularization was performed in 106 patients (in 76 patients as the only MACE event), resulting in a cumulative incidence at 14 years of 23%. Of these, 72% was a percutaneous coronary intervention and 28% repeat coronary artery bypass surgery (comparable between aspirin, aspirin with dipyridamole and VKA; p=0.30). Figure 3 Kaplan Meier of cardiac death, myocardial infarction and repeat revascularization

after coronary-artery-bypass-graft surgery, separately.

With 35% at 14 years, the need for repeat revascularization was most pronounced in patients treated with VKA, compared to 17% in patients treated with aspirin and 18% in patients treated with aspirin with dipyridamole (HR= 2.2 (95%CI, 1.4 to 3.5; P=0.001) and HR=2.2 (95%CI, 1.3 to 3.5; P=0.002), respectively, p=0.006 (aspirin compared to aspirin with dipyridamole HR=0.99, 95%CI, 0.57 to 1.7; P=0.97); see Figure 3). Furthermore, repeat revascularization was most frequently performed in younger patients below 56 years of age at the time of surgery, with a 2.8-fold increased risk of repeat revascularization compared to patients aged above 62 years (95%CI, 1.6 to 4.8, P=0.001). When comparing the patients aged 56-62 years with patients above 62 years, a 1.8-fold increased risk was observed (95%CI, 1.03 to 3.3, P=0.039). Additional risk-indicators were NYHA class III or IV (HR=2.0, 95%CI, 1.2 to 3.2; p=0.007), Use of digitalis (HR=5.3, 95%CI, 1.6 to 17.8;p=0.006), both prior to surgery and Number of grafts used during surgery (HR 1 or 2 grafts vs. 3 or more = 1.8, 95%CI, 1.1 to 2.8; p=0.018).

0 2 4 6 8 10 12 140%

10%

20%

30%


0 2 4 6 8 10 12 140%

10%

20%

30%


Time (years)

0 2 4 6 8 10 12 140%

10%

20%

30%


P=0.96 P=0.60 P=0.006

Time (years)Time (years)

Cardiac death Myocardial infarction Repeat revascularization

Chapter 6

- 103 -

Continuation of allocated treatment After one year, when the CABADAS study had ended, continuation of the allocated treatment was left at the discretion of the treating cardiologist. As presented in Table 2, at the first extended follow-up evaluation, after on average 2.5 years (SD, 0.8), the majority of the patients originally randomized to aspirin continued with aspirin (78%, 13% in combination with dipyridamole). Dipyridamole alone was used by 1%, 4% had switched to VKA and 17 % had stopped without switching. In the patients originally randomized to aspirin with dipyridamole, aspirin was still used by 77% of the patients (31% still in combination with dipyridamole), dipyridamole alone by 2%, 4% had switched to VKA and 17% had stopped without switching. Almost half of the patients originally randomized to VKA were still using VKA (49%), 30% had switched to aspirin (3% combined with dipyridamole) and 21% of patients had stopped without switching to another drug. Table 2 Continuation of allocated treatment. Aspirin

N=248

Aspirin + Dipyridamole

N=234

VKA†

N=244

Total

N=726

P-value

Medication at 2.5 years Anticoagulant therapy, % Aspirin Aspirin with dipyridamole Dipyridamole Vitamin K antagonists None Concomitant drugs, % beta blockers calcium antagonists nitrates diuretics lipid lowering drugs ACE inhibitor digitalis

65 13 1 4 17

31 10 9 14 25 12 4

46 31 2 4 17

29 11 7 12 30 14 4

27 3 0 49 21

36 15 10 15 28 15 3

46 16 1 19 18

32 12 9 14 27 14 4

<0.01

0.25 0.23 0.60 0.58 0.46 0.60 0.97


- 104 -

Regarding concomitant cardiovascular medication, in all three treatment groups, particularly the use of lipid lowering drugs and ACE inhibitors had increased from 5% to 27% and 3% to 14% respectively. Furthermore, the use of nitrates had increased from 1% to 9%, consistent with the presence of anginal complaints at the first extended follow-up evaluation, i.e. 89% NYHA I, 10% NYHA II and 1% NYHA III. When comparing the use of concomitant medication in the patients originally treated with aspirin, aspirin with dipyridamole and VKA, no apparent differences were observed. Sensitivity analysis In addition to the primary “Intention-To-Treat” analysis, a sensitivity analysis was performed using an “As Treated” approach. In this analysis, the continuation of allocated treatment was taken into account by using a varying exposure model in which treatment was allowed to change according to actual treatment after the time of the first extended follow-up. Overall, for MACE, differences between treatments were non-significant (P=0.20). Furthermore, the hazard ratio’s from this analysis were comparable to our “Intention-To-Treat” findings. Again, VKA was associated with a higher risk for MACE than aspirin with dipyridamole and aspirin (HR=1.4, 95%CI, 1.04 to 2.0; P=0.030 and HR=1.2, 95%CI, 0.94 to 1.6; P=0.136, respectively. When comparing VKA with no treatment, VKA patients were at 1.2 higher risk for MACE (95%CI, 0.76 to 1.7; P=0.51). Aspirin with dipyridamole and aspirin did not differ from those that stopped. In reference to repeat revascularization, also in the varying exposure approach, the risk was higher in VKA, compared to aspirin with dipyridamole and aspirin, as well as to stopping treatment (HR=3.1, 95%CI, 1.7 to 5.7; P=0.002 and HR=2.2, 95%CI, 1.4 to 3.4; P=0.001 and HR=2.4, 95%CI, 1.3 to 4.7; P=0.007, respectively). In this, aspirin with dipyridamole, aspirin and no treatment were not significantly different. COMMENT We evaluated the 14-year clinical outcome of patients who underwent coronary artery bypass grafting and subsequent treatment with aspirin, aspirin with dipyridamol or VKA. Studies with such a long-term follow-up are limited. In our study, cumulative rates of major adverse cardiac events reached 52%, with 17% at 5 year, 34% at 10 year

Chapter 6

- 105 -

and 41% at 12 year. When taking differences in follow-up into account, these rates are comparable with those from previous reports [9-12].

Overall, with 49%, 50% and 59% at 14-year, occurrence of MACE did not differ between the three treatment groups.

When evaluating the three components of MACE separately, the risk of both cardiac mortality and myocardial infarction was comparable in patients treated with aspirin, aspirin with dipyridamol and VKA. For repeat revascularization, however, a significant difference in favor of aspirin and aspirin plus dypiridamole was observed. Patients treated with VKA had a 2-fold increased risk for repeat revascularization. In this, patients treated with VKA already started to differ from aspirin and aspirin with dipyridamol early after completion of the original CABADAS one-year follow-up.

Figure 4 Kaplan Meier of repeat revascularization after coronary-artery-bypass-graft surgery, in patients treated with VKA.

Although differences in the separate components of MACE are well recognized [11,13-15], in our study this finding is intriguing. Firstly, one-year angiographic results from the CABADAS study showed comparable graft occlusion in the three treatment groups, which is not consistent with the observed difference already present within the second year of follow-up [8,16]. Secondly, we confirmed this finding in a secondary analysis using a varying exposure approach, taking into account the changes made to the randomized study treatment. Thirdly, when evaluating the quality of VKA therapy (i.e. percentage of time spend below the target range and the stability of INR)

Time (years)

0 2 4 6 8 10 12 14

Repe

at re

vasc

ular

izat

ion

0%

10%

20%

30%


P = 0.63


- 106 -

during the first year, no difference in repeat revascularization was observed in patients with optimal VKA and patients with a non-optimal VKA (additional evaluation, see Figure 4)

These three observations induced us to speculate that the high rate of repeat revascularization in patients treated with VKA is in part due to a referral bias. VKA patients were seen every two to three weeks by the specialized thrombosis service for the monitoring of the level of anticoagulation, whereas patients treated with aspirin or aspirin with dipyridamole were seen at a 3-montly interval only during the first year, and thereafter only when clinically indicated. The frequent contacts in VKA patients and continued care provided by the specialized thrombosis service might have increased sensitivity towards referral to cardiac specialists, resulting in a lower threshold towards coronary angiography and an increased probability for repeat revascularization.

This could explain the difference between repeat revascularisation and the “hard” clinical outcome events of cardiac death and myocardial infarction. Repeat revascularisation might be less suitable as a study endpoint in the setting of an observational follow-up study particularly when there are inherent differences in the structure of care associated with the treatments that are evaluated. In this setting, the “hard” clinical outcome events cardiac death and myocardial infarction should be used for valid risk estimates associated with the different treatments. In our study, using these endpoints no difference between aspirin, aspirin with dipyridamole and VKA was observed.

Although CABADAS was a randomized controlled clinical trial, a major limitation of our study is the retrospective setting of the extended follow-up. As already mentioned, referral bias might have lead to the observed difference in repeat revascularization. A prospective design with visits at regular intervals in all patients would have reduced this bias.

Furthermore, treatment and clinical outcome were evaluated at the first extended follow-up at 2.5 years. After this first follow-up, occurrence of MACE was thoroughly assessed but detailed information on continuation of medical treatments after 2.5 years is no longer available, limiting the ability to further assess the differences in repeat revascularization rates.

As the risk of both cardiac mortality and myocardial infarction did not differ between the three treatments groups, and irrespective of the extend to which referral

Chapter 6

- 107 -

bias might have led to the increased rate of repeat revascularization in patients treated with VKA, our long-term results add further evidence to the recommendations for antithrombotic therapy in CABG patients, as recently re-established in the 8th edition of the American College of Chest Physicians evidence-Based Clinical Practice Guidelines [2]. For all patients with coronary artery disease undergoing CABG, aspirin indefinitely is recommended and as there is no evidence that adding dipyridamole to aspirin improves outcome, this is not recommended. VKA should only be considered if indicated because of concomitant condition, e.g. heart valve replacement. In those instances, VKA combined with aspirin is suggested [2].

In conclusion, this study with 14-year clinical outcome provide further evidence for the use of aspirin as secondary prophylaxis after coronary-artery-bypass-graft surgery.


- 108 -

References 1. Antiplatelet Trialists' Collaboration. Collaborative overview of randomised trials of antiplatelet

therapy--II: Maintenance of vascular graft or arterial patency by antiplatelet therapy. BMJ 1994; 308: 159-168.

2. Becker RC, Meade TW, Berger PB, Ezekowitz M, O'Connor CM, Vorchheimer DA, Guyatt GH, Mark DB, Harrington RA. The primary and secondary prevention of coronary artery disease: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008; 133: 776S-814S.

3. Eagle KA, Guyton RA, Davidoff R, Edwards FH, Ewy GA, Gardner TJ, Hart JC, Herrmann HC, Hillis LD, Hutter AM, Jr., Lytle BW, Marlow RA, Nugent WC, Orszulak TA, Antman EM, Smith SC, Jr., Alpert JS, Anderson JL, Faxon DP, Fuster V, Gibbons RJ, Gregoratos G, Halperin JL, Hiratzka LF, Hunt SA, Jacobs AK, Ornato JP. ACC/AHA 2004 guideline update for coronary artery bypass graft surgery: summary article. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1999 Guidelines for Coronary Artery Bypass Graft Surgery). J Am Coll Cardiol 2004; 44: e213-e310.

4. Chesebro JH, Fuster V, Elveback LR, Clements IP, Smith HC, Holmes DR, Jr., Bardsley WT, Pluth JR, Wallace RB, Puga FJ, et al. Effect of dipyridamole and aspirin on late vein-graft patency after coronary bypass operations. N Engl J Med 1984; 310: 209-214.

5. Lorenz RL, Schacky CV, Weber M, Meister W, Kotzur J, Reichardt B, Theisen K, Weber PC. Improved aortocoronary bypass patency by low-dose aspirin (100 mg daily). Effects on platelet aggregation and thromboxane formation. Lancet 1984; 1: 1261-1264.

6. Mangano DT. Aspirin and mortality from coronary bypass surgery. N Engl J Med 2002; 347: 1309-1317.

7. Sharma GV, Khuri SF, Josa M, Folland ED, Parisi AF. The effect of antiplatelet therapy on saphenous vein coronary artery bypass graft patency. Circulation 1983; 68: II218-II221.


9. Cameron AA, Green GE, Brogno DA, Thornton J. Internal thoracic artery grafts: 20-year clinical follow-up. J Am Coll Cardiol 1995; 25: 188-192.

10. Daemen J, Boersma E, Flather M, Booth J, Stables R, Rodriguez A, Rodriguez-Granillo G, Hueb WA, Lemos PA, Serruys PW. Long-term safety and efficacy of percutaneous coronary intervention with stenting and coronary artery bypass surgery for multivessel coronary artery disease: a meta-analysis with 5-year patient-level data from the ARTS, ERACI-II, MASS-II, and SoS trials. Circulation 2008; 118: 1146-1154.

11. The BARI Investigators. The final 10-year follow-up results from the BARI randomized trial. J Am Coll Cardiol 2007; 49: 1600-1606.


Chapter 6

- 109 -

13. The Bypass Angioplasty Revascularization Investigation (BARI) Investigators. Comparison of coronary bypass surgery with angioplasty in patients with multivessel disease. N Engl J Med 1996; 335: 217-225.

14. Serruys PW, Unger F, Sousa JE, Jatene A, Bonnier HJ, Schonberger JP, Buller N, Bonser R, van den Brand MJ, van Herwerden LA, Morel MA, van Hout BA. Comparison of coronary-artery bypass surgery and stenting for the treatment of multivessel disease. N Engl J Med 2001; 344: 1117-1124.

15. Serruys PW, Ong AT, van Herwerden LA, Sousa JE, Jatene A, Bonnier JJ, Schonberger JP, Buller N, Bonser R, Disco C, Backx B, Hugenholtz PG, Firth BG, Unger F. Five-year outcomes after coronary stenting versus bypass surgery for the treatment of multivessel disease: the final analysis of the Arterial Revascularization Therapies Study (ARTS) randomized trial. J Am Coll Cardiol 2005; 46: 575-581.

16. van der Meer J, Brutel de la Riviere A, van Gilst WH, Hillege HL, Pfisterer M, Kootstra GJ, Dunselman PH, Mulder BJ, Lie KI. Effects of low dose aspirin (50 mg/day), low dose aspirin plus dipyridamole, and oral anticoagulant agents after internal mammary artery bypass grafting: patency and clinical outcome at 1 year. CABADAS Research Group of the Interuniversity Cardiology Institute of The Netherlands. Prevention of Coronary Artery Bypass Graft Occlusion by Aspirin, Dipyridamole and Acenocoumarol/Phenprocoumon Study. J Am Coll Cardiol 1994; 24: 1181-1188.

Chapter 7

- 111 -

Chapter 7

Excellent Long-Term Clinical Outcome

after Coronary Artery Bypass Surgery Using Three Pedicled Arterial Grafts

in Patients with Three-Vessel Disease

Nic J.G.M. Veeger

Gerald F.V. Panday

Adriaan A. Voors

Jan G. Grandjean

Jan van der Meer

Piet W. Boonstra

Ann Thorac Surg 2008;85:508-12

Complete arterial grafting in three-vessel disease

- 112 -

SUMMARY Background. Long-term clinical outcome of complete arterial grafting in three-vessel disease is unknown.

Methods. We studied 344 patients who underwent complete arterial revascularization using the internal thoracic arteries and the right gastroepiploic artery. Freedom from major adverse cardiac events (MACE) was evaluated by the Kaplan-Meier method, and homogeneity of outcome in strata of patients was assessed using Cox proportional hazards modeling.

Results. Median follow-up of survivors was 9.3 years (range, 0.01-12.8 years). The 12-year freedom from MACE was 75.5%. For the composite of MACE, this was 86.9% for cardiovascular death, 93.3% for myocardial infarction, and 89.4% for reintervention. In patients aged older than 65 years, MACE occurred significantly more frequent, with a freedom from MACE of 65.8% compared with 82.6% in younger patients (hazard ratio, 3.4; 95% confidence interval, 2.1-5.6, P < 0.001).

Conclusions. Complete arterial revascularization using both pedicled internal thoracic arteries and the gastroepiploic artery in patients with three-vessel disease resulted in an excellent long-term clinical outcome, especially in patients aged younger than 65 years.

Chapter 7

- 113 -

INTRODUCTION Venous grafts are used in most patients who have coronary artery bypass grafting (CABG), although it is well known that the use of at least one arterial graft to the left anterior descending coronary artery significantly reduces mortality and morbidity [1-3]. In addition, using both internal mammary arteries resulted in an even better clinical outcome [4-6]. However, two arteries often do not provide enough grafting material to obtain total arterial revascularization, especially in patients with three vessel disease.

There are two possibilities for additional arterial grafting material. First, the radial artery provides enough material, it can be used safely, and its reported short- and mid-term clinical results are satisfactory [7]. Second, we and others demonstrated that the right gastroepiploic artery also can be used for complete arterial revascularization in three-vessel disease [8]. Our mid-term follow-up results of CABG using both internal mammary arteries combined with the right gastroepiploic artery appeared to be promising [9,10]; however, the follow-up was too short to draw definite conclusions. In the present study, we therefore evaluated the long-term clinical outcome of patients with three-vessel disease who underwent CABG with both internal thoracic arteries and the right gastroepiploic artery. PATIENTS AND METHODS Patients Between 1989 and 1994, 344 patients underwent complete arterial revascularization using both internal thoracic arteries in combination with the right gastroepiploic artery for three-vessel coronary artery disease. All patients were operated on in the University Medical Center of Groningen. Long-term clinical outcome was retrospectively assessed up to June 2002 through telephone interviews of the patients using a standard questionnaire and by reviewing the medical files of their cardiologists or general practitioners, or both, when appropriate.

For this retrospective study approval from the local Ethics Committee was obtained. Individual patient data were collected without patient identification information. Owing to the retrospective nature of the study, the need for written informed consent was waived.


- 114 -

Surgical Techniques The surgical techniques are extensively described elsewhere [9,10]. In brief, all patients were operated on with the use of extracorporeal circulation, using mild hypothermia (28-32°C nasopharyngeal temperature) and crystalloid-induced cardioplegic arrest. The right internal thoracic artery (ITA) was routed through the transverse sinus in 75% of the patients to reach the circumflex coronary artery; in 25%, it was routed directly to the left anterior descending coronary artery. The left ITA was routed under the mediastinal fat deep along the parietal left pleura to the left anterior descending coronary artery in 75% of patients and to the circumflex area in 25%. The right gastroepiploic artery was always routed anterior to the liver and through a hole in the diaphragm.

All patients received anticoagulant or antiplatelet drug therapy (mostly aspirin) for at least 1 year after CABG. In most patients, this was continued afterwards to maintain graft patency. Clinical End Points Major adverse cardiac events (MACE) were defined as cardiac mortality, myocardial infarction, and coronary revascularization. Cardiac mortality was defined as death due to acute myocardial infarction, acute cardiac arrest, sudden death, or progressive congestive heart failure. Unknown causes of death were classified as cardiac death. For myocardial infarction, pathologic Q-waves had to be present on the electrocardiogram (ECG) in combination with creatine phosphokinase (CPK) and CPK-MB serum levels that exceeded the upper limits of normal ranges. Revascularization was defined as percutaneous coronary intervention (PCI) or redo-CABG during follow-up. Statistical Analysis The objective of this study was to evaluate long-term clinical outcome in patients who underwent complete arterial revascularization using three arterial grafts. The primary composite end point was MACE. Event-free survival was graphically depicted using the Kaplan-Meier method. To evaluate the homogeneity of outcome in different categories of patients, Cox proportional hazards regression analysis was performed. Predefined risk factors for MACE included in univariate analyses were baseline data for age, sex, body mass index, left ventricle ejection fraction (LVEF) of 0.30 or less,

Chapter 7

- 115 -

hypertension, diabetes mellitus, smoking, hypercholesterolemia, history of myocardial infarction, PCI, and concomitant cardiovascular medication. Other factors included the number of distal anastomoses, perfusion duration, and year of surgery.

All tests performed to test the (null) hypothesis of no difference were two-sided. A value of P < 0.05 was considered statistically significant. For all analyses, SAS 9.1 software (SAS Institute, Cary, NC) was used. RESULTS Patient Characteristics The study included 344 consecutive patients (89% men) who underwent CABG from September 1989 to December 1994. The total follow-up time was 3055 patient-years. Mean age at time of operation was 59 years. Anginal complaints at inclusion were 62% at New York Heart Association (NYHA) functional class III and 33% at class IV. Left ventricular function was normal in 76%, moderate in 21%, and poor (LVEF < 0.30) in 3% of the patients. All patients used cardiovascular drugs, as summarized in Table 1. Death and Additional Procedures Post-operative mortality (≤ 30 days) was 1.5%. During follow-up, 3 patients had abdominal surgery: twice for aneurysm of the abdominal aorta and once for a tumor of the stomach. These operations were completed without damage to the gastroepiploic artery. Clinical outcome Median follow-up was 9.3 years (range, 0.01–12.8 years). During follow-up, MACE events occurred in 69 patients (20.1%). MACE-free survival is presented in Figure 1, indicating a 12-year freedom from MACE of 75.5%. The12-year event-free survival for the components of MACE was 86.9% for cardiovascular death, 93.3% for myocardial infarction, and 89.4% for revascularization (also see Figure 1).


- 116 -

Table 1 Baseline characteristics of patients with complete arterial revascularization

Characteristicsa Value (N=344)

Age, years Male, % Body mass index, kg/m2 Blood pressure, mmHg Systolic Diastolic LVEF, % Normal Moderate Poor (≤ 0.30) NYHA class, % II III IV Medical history, % Diabetes Mellitus Smoking Hypertension Myocardial infarction PCI Hypercholesterolemia Cholesterol, mmol/L HDL, mmol/L Triglycerides, mmol/L Medication, % ACE inhibitors Calcium antagonists Diuretics Oral anticoagulants Platelet aggregation inhibitors β-blockers Lipid-lowering drugs Nitrates Triple therapyb

Perfusion time, min ≥4 distal anastomosis, %

59 ± 10 89 26 ± 3 138 ± 20 82 ± 10 76 21 3 5 62 33 9 33 23 56 16 56 6.6 ± 1.3 1.0 ± 0.3 2.4 ± 1.3 8 58 9 13 33 71 15 61 29 92 ± 36 39

a Continious variables are means ± SD. b Nitrate, β-blocker, calcium antagonist. ACE = Angiotensin Converting Enzyme; HDL = High Density Lipoprotein; LVEF = Left Ventricular Ejection Fraction; NYHA= New York Heart Association; PCI = Percutaneous Coronary Intervention.

Chapter 7

- 117 -

Angina pectoris status was assessed in 279 patients (81%). Of these, 88% were in NYHA class I, 10% in class II, 1% in class III and 1% in class IV. Anginal status was not assessed in 37 patients due to a fatal cardiovascular event. Figure 1 Freedom from major adverse cardiac events (MACE, black line) during 12 years of

follow-up as well as from the separate components of MACE, consisting of cardiac death (solid grey), myocardial infarction (dash-dot line), and revascularization either by percutaneous coronary intervention or redo coronary artery bypass grafting (dash line).

Predictors of Major Adverse Cardiac Events Cox proportional hazards regression analysis showed differences in the risk for MACE between various strata of known risk factors. Especially in elderly patients (≥ 65 years), the 12-year freedom from MACE was 65.8% compared with 82.6% for younger patients, resulting in a 3.4-fold (95% confidence interval (CI), 2.1-5.6, P < 0.001) higher risk (see Figure 2) in the elderly patients. In patients with a moderate to poor left ventricular function before operation, the risk was 1.9-fold higher (95% CI, 1.1-3.1, P = 0.02) compared with patients with a normal left ventricular function. A history of hypertension, myocardial infarction before CABG, the number of distal anastomoses, and diabetes mellitus was also associated with the occurrence of MACE. Multivariable analysis showed that myocardial infarction before CABG, the number of distal anastomoses, and diabetes mellitus were no longer independently associated

Patients at risk:344 331 321 306 250 122 21

Time in months0 12 24 36 48 60 72 84 96 108 120 132 144

Eve

nt fr

ee su

rviv

al (%

)

075

80

85

90

95

100

MACECardiac deathRevascularizationMyocardial infarction


- 118 -

with MACE, whereas age, a history of hypertension, and left ventricular function remained as important predictors for MACE (Figure 3).

Figure 2 Freedom from major adverse cardiac events (MACE) during 12 years of follow-up in

patients younger than 65 years (black line) vs patients 65 years or older (gray line) at the time of surgery.

When the components of MACE were considered separately, the hazard ratios for the respective risk factors of cardiovascular death, myocardial infarction, and reintervention were comparable with those observed in the evaluation of the composite of MACE (data not shown). COMMENT We evaluated a cohort of 344 patients who underwent complete arterial revascularization for triple-vessel disease using the internal thoracic arteries and right gastroepiploic artery. At 12 years, the freedom from MACE was 75.5%, and freedom from cardiovascular death was 86.9%, myocardial infarction was 93.3%, and revascularization was 89.4%.

The beneficial effects of arterial grafts were reported in 1986 [2,3]. In a 15-year follow-up study of 748 CABG patients, the use of at least one ITA was associated

Patients at risk:233111

228103

22497

22086

18367

9329

174

Time in months

0 12 24 36 48 60 72 84 96 108 120 132 144

MA

CE fr

ee su

rviv

al (%

)

050556065707580859095

100

up to 65 years65 years or older

P < 0.001

65 years or olderup to 65 years

Chapter 7

- 119 -

with decreased death, risk of myocardial infarction, reoperations, and less recurrence of angina pectoris [3]. These findings were supported by those from a larger series of 5931 patients in whom the use of only vein grafts revealed a 1.6-fold increased death, compared with the use of one ITA combined with vein grafts [2]. Using both left and right ITA resulted in an even better clinical outcome [5,6]. A large retrospective study of 8123 patients who underwent CABG using both ITAs showed a small but significant reduction in the death rate compared with the use of one ITA [6]. In a 13-year follow-up study comparing bilateral with single ITA grafting, Berreklouw and colleagues [5] demonstrated a reduction of the combined end point of cardiac death, myocardial infarction, reintervention, or recurrent angina.

Despite these data, one arterial graft and a vein graft are still used in most patients. This can be explained by a lack of arterial graft material, especially in patients with three-vessel coronary artery disease. Another explanation might be that there is no evidence of superiority for three pedicled arterial grafts. Figure 3 Multivariable analysis of the predictors of major adverse cardiac events (MACE)

during long-term follow-up in patients with complete arterial revascularization presented as hazard ratios (diamonds) and 95% confidences intervals (whiskers). (LVEF = left ventricular ejection fraction)

In combination with both left and right ITAs, there are several options to obtain additional arterial grafts. First, the radial artery is being used in an increasing number of CABG patients. Several authors reported beneficial angiographic and clinical short-term effects when this artery was used. Possati and colleagues [11] demonstrated a patency rate of 92% after a mean follow-up of 9 years, compared with 98% for ITAs [11]; however, long-term angiographic graft patency varied widely.

Hazard ratios with 95% confidence intervals0 1 2 3 4 5 6

Age: > 65 versus < 65

History of hypertension

Distal anastomosis: less than 4 versus 4 or more

History of myocardial infarction

Moderate to poor LVEF

P < 0.001

P = 0.013

P = 0.094

P = 0.19

P = 0.019

Diabetes mellitus P = 0.69


- 120 -

Like other arterial grafts, the radial artery is very vulnerable to surgical trauma, hypotension, and early graft malfunction whenever α-adrenergic drugs are used. Recent improvements in surgical harvesting technique of the radial artery and in perioperative drug therapy have reduced these problems and consequently have increased the use of the radial artery [12-14]. Table 2. Literature Reviewa

Freedom from Published Reports, First Author Survival (%) MI (%) Reintervention (%)

Vein grafts only - van Brussel (20) - Cameron (3)

83 73

91 75

86 82

Single ITA grafts - Sergeant (1) - Fiore (21) - Cameron (3) - Pick (22)

- 82 82 75

85 81 83 -

- 95 96b

-

Double ITA grafts - Fiore (21) - Cameron (3) - Pick (22)

84 87 85

81 - -

95 - -

Double ITA and GEA graft - Tavilla (23) - Veeger, present study

87 87

- 93

- 89

a Overview of published reports on actuarial survival and actuarial probability of the patient remaining free from myocardial infarction or reintervention (repeat coronary artery bypass graft surgery or percutaneous coronary intervention) or both, 12 years after primary operation. b Redo coronary artery bypass grafting only.

GEA = gastroepiploic artery; ITA = internal thoracic artery; MI = myocardial infarction.

Another option is the gastroepiploic artery. In contrast with the radial artery, this is a pedicled graft. Beneficial short-term effects, minimal operative risks, and good patency rates have been reported using the gastroepiploic artery [10,15], and mid-term clinical outcome has also been promising [7,16]. We previously reported the results of 7 years of follow-up in 256 patients with three-vessel disease who were grafted with

Chapter 7

- 121 -

the gastroepiploic artery together with both ITAs [7]. After 7 years, 85% were free from angina pectoris, a considerably better result than that reported from studies in patients in whom vein grafts, single ITA grafts, or double ITA grafts were used (59-77%) [7].

Other options for additional arterial grafts are the inferior epigastric artery [17] and, rarely used, bovine grafts [18]. However, their short-term patency rates do not warrant their use in routine CABG surgery [17,18].

The benefit of complete arterial revascularization was less pronounced in elderly patients (≥ 65 years, see Figure 2). Moreover, the gastroepiploic artery should be considered in case of a calcified ascending aorta to avoid side clamping of the aorta for performing the proximal anastomosis. Previous studies suggested that the use of arterial grafts should particularly be preferred in patients with diabetes mellitus [12,19]. We can neither confirm nor reject this hypothesis. Although hazard ratios indicated a higher risk in our patients with diabetes mellitus, their small number led to wide confidence limits.

Data published on results of patients with coronary revascularization are difficult to compare owing to the wide differences in used types of grafts and inclusion and exclusion criteria. We did, nevertheless, compare our results with those obtained from recently published studies [1,3,20-23] (see Table 2) and concluded that actuarial survival and actuarial probability of remaining free from myocardial infarction or reintervention, or both, 10 years after the primary operation, is excellent. In addition to the literature review, we also compared the clinical outcome in our study cohort with that in a cohort of 281 patients with three-vessel disease who underwent revascularization and received ITAs in addition to vein grafts. The patients in this cohort originated from a multicenter clinical trial (CABADAS) [24] executed between 1987 and 1990. Patient outcome was assessed using the same methodology as for the patients in our study cohort. This comparison was done using a stepwise approach. First, considering the lack of a random treatment allocation, a propensity score analysis was performed to estimate the probability of being assigned to either one of the three grafting strategies [25-27]. The individual propensity scores were than included in a multivariable Cox proportional hazards regression analysis.

At baseline the patients from the mixed arterial-vein grafts cohort were comparable with our study cohort for most of the important risk factors of MACE.


- 122 -

Differences were observed in hypertension (23% in complete arterial vs 37% in mixed arterial-vein grafts), hypercholesterolemia (56% in complete arterial vs 68% in mixed arterial-vein grafts), previous revascularization (15% in complete arterial vs 3% in mixed arterial-vein grafts), preoperative NYHA classification (62% class III and 33% class IV in complete arterial vs 57% class III and 18% class IV in mixed arterial-vein grafts), and triple drug therapy (29% in complete arterial vs 37% in mixed arterial-vein grafts) (further data not shown).

MACE was significantly reduced in patients who underwent complete arterial revascularization, with a hazard ratio of 0.43 (95% CI, 0.29-0.63, P < 0.001) (derived from multivariable Cox proportional hazards regression analysis, including propensity scores; full analysis results not shown).

Despite the use of propensity scores, this comparison is limited by the lack of random treatment allocation with potential selection bias. We therefore cannot consider this comparison as conclusive, but it does provide us with an estimate of the benefit of complete arterial revascularization. In conclusion, we demonstrated that the use of complete arterial revascularization using both pedicled ITAs and the gastroepiploic artery in patients with three-vessel disease resulted in excellent long-term clinical outcome. These long-term results confirm our earlier mid-term findings of excellent results with complete arterial grafting in these patients.

Chapter 7

- 123 -

References 1. Sergeant PT, Blackstone EH, Meyns BP. Does arterial revascularization decrease the risk of

infarction after coronary artery bypass grafting? Ann Thorac Surg 1998; 66: 1-10. 2. Loop FD, Lytle BW, Cosgrove DM, Stewart RW, Goormastic M, Williams GW, Golding LA, Gill

CC, Taylor PC, Sheldon WC, . Influence of the internal-mammary-artery graft on 10-year survival and other cardiac events. N Engl J Med 1986; 314: 1-6.

3. Cameron A, Kemp HG, Jr., Green GE. Bypass surgery with the internal mammary artery graft: 15 year follow-up. Circulation 1986; 74: III30-III36.

4. Endo M, Nishida H, Tomizawa Y, Kasanuki H. Benefit of bilateral over single internal mammary artery grafts for multiple coronary artery bypass grafting. Circulation 2001; 104: 2164-2170.

5. Berreklouw E, Rademakers PP, Koster JM, van Leur L, van der Wielen BJ, Westers P. Better ischemic event-free survival after two internal thoracic artery grafts: 13 years of follow-up. Ann Thorac Surg 2001; 72: 1535-1541.

6. Lytle BW, Blackstone EH, Sabik JF, Houghtaling P, Loop FD, Cosgrove DM. The effect of bilateral internal thoracic artery grafting on survival during 20 postoperative years. Ann Thorac Surg 2004; 78: 2005-2012.

7. Bergsma TM, Grandjean JG, Voors AA, Boonstra PW, den Heyer P, Ebels T. Low recurrence of angina pectoris after coronary artery bypass graft surgery with bilateral internal thoracic and right gastroepiploic arteries. Circulation 1998; 97: 2402-2405.

8. Hirose H, Amano A, Takanashi S, Takahashi A. Coronary artery bypass grafting using the gastroepiploic artery in 1,000 patients. Ann Thorac Surg 2002; 73: 1371-1379.

9. Grandjean JG, Boonstra PW, den Heyer P, Ebels T. Arterial revascularization with the right gastroepiploic artery and internal mammary arteries in 300 patients. J Thorac Cardiovasc Surg 1994; 107: 1309-1315.

10. Grandjean JG, Voors AA, Boonstra PW, den Heyer P, Ebels T. Exclusive use of arterial grafts in coronary artery bypass operations for three-vessel disease: use of both thoracic arteries and the gastroepiploic artery in 256 consecutive patients. J Thorac Cardiovasc Surg 1996; 112: 935-942.

11. Possati G, Gaudino M, Prati F, Alessandrini F, Trani C, Glieca F, Mazzari MA, Luciani N, Schiavoni G. Long-term results of the radial artery used for myocardial revascularization. Circulation 2003; 108: 1350-1354.

12. Tatoulis J, Buxton BF, Fuller JA, Royse AG. Total arterial coronary revascularization: techniques and results in 3,220 patients. Ann Thorac Surg 1999; 68: 2093-2099.

13. Tatoulis J, Buxton BF, Fuller JA. Patencies of 2127 arterial to coronary conduits over 15 years. Ann Thorac Surg 2004; 77: 93-101.

14. He GW. Nitric oxide and endothelium-derived hyperpolarizing factor in human arteries and veins. J Card Surg 2002; 17: 317-323.

15. Suma H, Isomura T, Horii T, Sato T. Late angiographic result of using the right gastroepiploic artery as a graft. J Thorac Cardiovasc Surg 2000; 120: 496-498.

16. Nishida H, Tomizawa Y, Endo M, Koyanagi H, Kasanuki H. Coronary artery bypass with only in situ bilateral internal thoracic arteries and right gastroepiploic artery. Circulation 2001; 104: I76-I80.

17. Cremer J, Mugge A, Schulze M, Trappe HJ, Schneider M, Heublein B, Haverich A. The inferior epigastric artery for coronary bypass grafting. Functional assessment and clinical results. Eur J Cardiothorac Surg 1993; 7: 423-427.


- 124 -

18. Suma H, Wanibuchi Y, Takeuchi A. Bovine internal thoracic artery graft for myocardial revascularization: late results. Ann Thorac Surg 1994; 57: 704-707.

19. Thourani VH, Weintraub WS, Stein B, Gebhart SS, Craver JM, Jones EL, Guyton RA. Influence of diabetes mellitus on early and late outcome after coronary artery bypass grafting. Ann Thorac Surg 1999; 67: 1045-1052.


21. Fiore AC, Naunheim KS, Dean P, Kaiser GC, Pennington G, Willman VL, McBride LR, Barner HB. Results of internal thoracic artery grafting over 15 years: single versus double grafts. Ann Thorac Surg 1990; 49: 202-208.

22. Pick AW, Orszulak TA, Anderson BJ, Schaff HV. Single versus bilateral internal mammary artery grafts: 10-year outcome analysis. Ann Thorac Surg 1997; 64: 599-605.

23. Tavilla G, Kappetein AP, Braun J, Gopie J, Tjien AT, Dion RA. Long-term follow-up of coronary artery bypass grafting in three-vessel disease using exclusively pedicled bilateral internal thoracic and right gastroepiploic arteries. Ann Thorac Surg 2004; 77: 794-799.


25. D'Agostino Jr RB. Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med 1998; 17: 2265-2281.

26. Winkelmayer WC, Glynn RJ, Mittleman MA, Levin R, Pliskin JS, Avorn J. Comparing mortality of elderly patients on hemodialysis versus peritoneal dialysis: a propensity score approach. J Am Soc Nephrol 2002; 13: 2353-2362.

27. Moss RR, Humphries KH, Gao M, Thompson CR, Abel JG, Fradet G, Munt BI. Outcome of mitral valve repair or replacement: a comparison by propensity score analysis. Circulation 2003; 108 Suppl 1: II90-II97.

Chapter 8

- 125 -

Chapter 8

Summary, Discussion

and Future Perspectives

Summary and discussion

- 126 -

SUMMARY, DISCUSSION AND FUTURE PERSPECTIVES BACKGROUND Vitamin K antagonists (VKA) are still the only oral anticoagulants for long-term therapy, despite the need and ongoing search for new and better drugs. Since its approval for medical use in humans almost 60 years ago their use is still increasing. This is remarkable, especially since VKA therapy is problematic and has several limitations, of which its large inter- and intra-individual variability in dose-response is the most important one. To indicate the magnitude of both use and limitation of VKAs the following numbers are informative: in the Netherlands, in 2007 in more than 300000 outpatients VKA therapy was managed by the anticoagulation clinics, in a nationwide network. From 2002 to 2007, this number had increased by 10%. Furthermore, in the USA in 2004 the number of outpatient prescriptions written for VKA amounted to nearly 31 million, and VKA was among the top 10 drugs with the largest number of serious adverse event reports submitted during the 1990 and 2000 decades. All this adds up to a growing need for new and better drugs, of which promising drugs are currently investigated, some already in large phase III studies, using VKA as the reference therapy. In this respect, it is intriguing that in most of the studies the aim is non-inferiority. The concept of non-inferiority of treatments indicates a choice. When choosing, other factors than efficacy and safety could be relevant, but also the between-patient differences in the efficacy and safety must be taken into account. This is especially the case with a drug like VKA, where this problem is the most important limitation. The question that must be answered is “which patient is best treated with which drug?” In this respect, the ability to predict a patient response to VKA prior to the initiation of the treatment is still very limited.

In this thesis, both the efficacy and safety of VKA therapy in different groups of patients was addressed. Its focus lies on the early identification of those patients who will, or will not do well on VKA. The cohorts that were extensively studied were patients with deep vein thrombosis, pulmonary embolism, atrial fibrillation, heart valve replacement and myocardial infarction, as well as patients who underwent coronary artery bypass grafting surgery, in whom VKA was used in a comparative setting, next to dipyridamole, aspirin, or both.

Chapter 8

- 127 -

SUMMARY In chapter one, an overview of the current state of affairs of VKA therapy was presented, and several issues discussed. These included the optimal level of anticoagulation, but also the difficulty to achieve this optimum. Although many factors are known to influence this relationship, including genetic factors, age, co-morbidity, concomitant drugs, changes in activity level and dietary intake of vitamin K, still a large percentage of the instability that is observed cannot be attributed to these known risk factors. For this reason, early signs of under- and overanticoagulation could be valuable in identifying those patients at higher risk of recurrent thrombosis or major bleeding. Due to this highly variable dose-response relationship, nonstop monitoring of the response is mandatory.

To assess the extent of the variability in individual patients, a valid methodology should be used. The Time within the Target Range (TTR) has proven its value in assessing this variability in relation to clinical outcome. Although TTR can be calculated differently, the method proposed by Rosendaal et al. was our method of choice. Originally advocated for its applicability in assessing INR-related incidence rates at a group level, it also makes it possible to assess the level of anticoagulation during the course of VKA therapy in individual patients (ITTR). Considering the huge amount of variation, with only a limited predictability beforehand, especially the evaluation of the individual course of INR through time is of importance. In chapter two, efficacy and safety of vitamin K antagonists (VKA) were evaluated in relation to the actually achieved level of anticoagulation (INR) in a cohort of 2304 consecutive patients with venous thromboembolism. The patients were all referred to a specialized anticoagulant clinic in The Netherlands for control of their VKA therapy. The target range used in these patients was INR 2.0 to 3.5. Using the individual percentage of time within the target range (ITTR) during VKA therapy, we confirmed that it is difficult to maintain an optimal INR over time in all patients. Although the mean ITTR was 63% (11% below and 26% above the range), in one quarter of the patients the ITTR was below 45%. Especially these patients were at higher risk for both recurrent thromboembolism and major bleeding. The absolute risks of recurrent thromboembolism and major bleeding, expressed as incidence rates were 6.2 and 2.8 per 100 person-years, respectively. In the patients with an ITTR <


- 128 -

45%, incidence rates amounted to 16.6 and 8.7 per 100 person-years, whereas in the patients with an ITTR between 45-65% and above 65% incidence rates were comparable (4.9 and 2.1 versus 4.6 and 1.9 per 100-person-years respectively).

This finding stressed the importance of being able to identify the poor responders, prior to or soon after initiation of VKA therapy. When using the achieved level of anticoagulation during the first 30 days of treatment, i.e. 30-days ITTR, a low ITTR was highly associated with an overall poor response to VKA. In this cohort, the lowest quarter of patients with a 30-days ITTR < 37% (median 20%) had an approximately 25-fold increased risk of having an overall poor ITTR < 45%, indicating that 30-days ITTR is highly predictive for the total treatment ITTR. In chapter three, the association between quality of anticoagulation and occurrence of thromboembolism and major bleeding was also assessed in three other cohorts. In total, 4454 consecutive patients with myocardial infarction (N = 1012), atrial fibrillation (N = 2614) and a prosthetic heart valve (N = 828) were studied. Again, all patients were referred to a specialized anticoagulant clinic in The Netherlands for control of their VKA therapy. The quality of anticoagulation was measured using the ITTR, now in patients from three cohorts with different target ranges. In addition, differences in percentage of time below and above target was also considered. Furthermore, the value of 30-days ITTR on the ability to early identify patients with a poor response to VKA was addressed.

On average, ITTRs were 39%, 42% and 44% in patients with myocardial infarction, atrial fibrillation and prosthetic heart valve respectively. This lack in differences between ITTRs can be explained by the comparable width of the target range. However, there were differences in clinical outcome. Thromboembolic events had occurred more frequently in patients with myocardial infarction (incidence rate 4.0 per 100 person-years) than in patients with atrial fibrillation (incidence rate 1.7 per 100 person-years) and patients with a prosthetic heart valve (incidence rate 1.4 per 100 person-years). For major bleeding, with an incidence rate of 0.9, 1.6 and 1.7 per 100 person-years, this was the opposite. As was observed in the cohort of deep vein thrombosis and pulmonary embolism (chapter 2), the highest risk of both thromboembolic events and major bleeding was observed in the quartile of patients with the lowest ITTR (less than 25% to 34%). Remarkably, incidence rates did not differ strongly between patients with higher ITTRs in all three cohorts, indicating a

Chapter 8

- 129 -

threshold in quality of anticoagulation beyond which clinical outcome is compromised.

In the atrial fibrillation and the prosthetic heart valve cohort, thromboembolism and major bleeding were balanced. This was not the case in the myocardial infarction cohort, were thromboembolism had occurred more frequently than major bleeding, especially in the patients with the lowest ITTR. This difference can be explained by differences in pathophysiology of thrombosis in the heart cavities versus coronary artery thrombosis. The inability to maintain the level of anticoagulation within its target range will have a much greater clinical impact when VKA treatment is applied to prevent the development of a small thrombus in a thrombogenic atherosclerotic coronary artery than to prevent a first thrombus in the much wider heart cavities. On the other hand, the cohorts differed with respect to the time below and the time above INR target range, while ITTR was similar. In the myocardial infarction cohort more frequent under-anticoagulation was observed, with an increased risk of thromboembolism - mainly recurrent myocardial infarction - and a reduced risk of major bleeding, as compared with patients from the two other cohorts. This was especially observed in the patients in the lowest ITTR quartiles.

Considering the imbalance of under- and over-anticoagulation and the associated types of events, there seems to be room for improvement in managing the patients with a myocardial infarction, especially those with a poor ITTR.

Managing anticoagulant therapy using VKA has been compared with walking a tightrope, swinging from left to right and always at risk of falling. The majority of patients are sufficiently equipped to do the walk, but those who are not, despite intensive INR monitoring and adjustments of the dose of VKA, need to be identified. Such an identification is possible using the 30-days ITTR, as this was strongly associated with the overall ITTR. Especially these patients may benefit from treatment with new anticoagulant drugs currently being developed. Chapter four addresses the risk of bleeding complications, and the possibility to identify patients at high risk of major bleeding. The majority of bleeding complications in patients on VKA are clinically mild. Incidence rates of major bleeding range from 1.4 to 3.3 per 100 person-years, whereas for minor bleeding incidence rates up to 30 per 100 person-years are reported. Minor bleeding might be due to over-anticoagulation or a pre-existing mild bleeding disorder, which is enhanced by VKA.


- 130 -

In either case, the occurrence of a minor bleed during VKA therapy could be associated with an increased risk of major bleeding. We hypothesized that minor bleeding occurring during VKA therapy could lead to intensified monitoring of VKA therapy to maximize the quality of anticoagulation and thereby reducing the risk of major bleeding. On the other hand, minor bleeding could be an indicator for an increased risk of major bleeding, similar to a history of major bleeding.

To assess the association between minor and subsequent major bleeding, a cohort of 6758 consecutive patients was retrospectively studied. In 20.0% of the patients (N = 1348), a minor bleeding had occurred (incidence rate 25.1 per 100 person-years (95%CI, 23.8-26.5)). These patients were at 3.6-fold increased risk of a subsequent major bleeding within one month (risk estimated using a time-varying exposure model). A change in managing VKA therapy towards more intense monitoring with more frequent dose adjustments was not observed, and the percentage of time within, below and above the target range did not differ from the patients without or after the occurrence of a minor bleed.

The quality of anticoagulation itself was also independently associated with major bleeding, with a 2.8-fold increased risk in patients with a poor response to VKA, i.e. individual time within target range of less than 25% of the total treatment time. Other risk factors of major bleeding were malignancy, increasing age (between 55-70 years and especially above 70 years), sex (male), and the use of NSAIDs, adjusted for the indication for VKA therapy (atrial fibrillation, deep vein thrombosis, pulmonary embolism, prosthetic heart valve and myocardial infarction). We concluded that minor bleeding was associated with subsequent major bleeding. The occurrence of a minor bleed should increase the awareness towards a potentially high-risk situation. In this, the achieved level of anticoagulation should also be considered, not just the INR at the time of the minor bleed, but especially the ITTR of the preceding treatment period.

Whether a lower intensity of anticoagulation might be suitable for these high-risk patients, without compromising the efficacy of VKA therapy to prevent thromboembolic complications is speculative. But still, the indication for VKA therapy should be reconsidered, with a new risk-benefit assessment. Given the very promising safety profile of one of the new direct thrombin and factor Xa inhibitors, i.e. dabigatran, as recently established in the RE-LY study (N Engl J Med

Chapter 8

- 131 -

2009;361:1139-51), a switch to this new oral thrombin inhibitor, when it is approved, could be indicated. The presence of non-optimal VKA therapy, not momentary but structural, and its consequences for clinical outcome was further assessed in a cohort of coronary-bypass-graft-surgery patients who received VKA for secondary prophylaxis. As described in chapter five, these patients were included in a large clinical trial, the CABADAS, evaluating efficacy and safety of aspirin, aspirin plus dipyridamole and VKA over a one-year period. Two-hundred-thirty-three patients randomized to VKA were selected and their individual response to VKA during the one-year of therapy assessed. Furthermore, the occurrence of cardiac events during a 14-year follow-up was evaluated.

Although a target range of INR 2.8-4.8 was used for managing VKA therapy at the time of treatment, we used the target range of INR 3.0-4.0 to assess the association between non-optimal VKA therapy and cardiac events. This is the intensity level for primary and secondary prophylaxis of arterial thromboembolic complications currently advocated by the Dutch Federation of Thrombosis Services. Patients had non-optimal VKA therapy when they spent a large amount of time below the target range INR 3.0-4.0 (lowest tertile of patients to indicate non-optimal VKA therapy), or their response over time was highly variable (highest tertile of patients to indicate non-optimal VKA therapy). For the latter, the variance growth rate was used.

The mean INR achieved in this cohort was 3.1, which was at the lower end of the target range of INR 2.8-4.8. On average, the individual time spent within the target range (ITTR) was 56%, with 39% of time below INR 2.8. When considering the achieved quality of anticoagulation as defined in terms of both the time spent below INR 3.0 and stability in INR over time, 93 patients (40%) had optimal VKA therapy, i.e. least frequently below INR 3.0 with most stable INRs. In these patients, time spent below INR 3.0 was 35%, compared to 58% in the 142 patients (60%) classified as non-optimal.

After one year, at the end of the CABADAS study, the continuation of the VKA therapy was left at the discretion of the treating physician. Overall, in half of the patients, VKA was continued. Remarkably, also in half of the patients with non-optimal VKA, the treatment was continued, clearly indicating that the decision to continue VKA was not based on information regarding the previously achieved (poor) quality of anticoagulation. At the time of the decision, such information was not


- 132 -

provided to the treating physician. When considering the clinical impact of optimal versus non-optimal VKA during the first year, we analyzed the events during follow-up. No difference in cardiac death and myocardial infarction were observed up to 8 years, but after 8 years, non-optimal VKA was associated with a more than 3-fold increased risk of cardiac death and myocardial infarction. In this, optimal VKA therapy could be more effective against the extension of the native coronary artery disease, rather than disease progression at the site of grafts.

In addition to the “hard” clinical endpoints of cardiac mortality and myocardial infarction, the need for repeat-revascularization was evaluated as a secondary endpoint. Repeat-revascularization was not associated with the first year quality of anticoagulation, with a cumulative rate of 35% at 14 years. In this, with a more than 2-fold increase, an age below 60 years at the time of the initial CABG was the most important determinant for a repeat-revascularization. Possibly, a lower threshold for referral and consequent revascularization in younger patients, irrespective of the achieved quality of anticoagulation, might explain this difference. The possibility of a referral bias for revascularization was further illustrated by comparing the patients treated with VKA with those patients from the original CABADAS study randomized to receive aspirin or aspirin plus dipyridamole. In these patients, without the frequent follow-up visits to manage their treatment, the cumulative rates of repeat-revascularization were significantly lower (18% and 17%, respectively). Further comparisons of VKA, aspirin and aspirin plus dipyridamole are described in detail in chapter six. In this chapter, the long-term clinical outcome of all Dutch patients included in the CABADAS study was evaluated (N = 726). In these patients, cumulative incidence of the composite of cardiac death, myocardial infarction and repeat-revascularization (MACE) reached 52% at 14 years. Although there was a large difference in repeat-revascularization (2-fold increased risk in patients treated with VKA), regarding MACE no difference between the three treatment groups was observed. For the “hard” clinical endpoints cardiac death and myocardial infarction separately, also no differences were observed. We speculated that the increased risk of repeat-revascularization in patients treated with VKA was in part due to a referral bias. As already discussed in chapter 5, due to the monitoring requirements associated with VKA therapy, with frequent patient-physician contacts, the threshold towards referral

Chapter 8

- 133 -

to cardiac specialists might have been lower, resulting in earlier and more frequent repeat-revascularizations.

Based on our findings, we concluded that in patients undergoing coronary-artery-bypass-graft surgery, aspirin was the drug of choice. In this, we added further evidence to the recommendations for antithrombotic therapy in CABG patients, as recently re-established in the 8th edition of the American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. VKA should only be considered if indicated for a concomitant condition, and preferably combined with aspirin. The combination should especially be considered in patients with non-optimal VKA therapy. In CABADAS, patients who underwent elective aortocoronary bypass surgery with saphenous vein grafts were included. More than half of the patients also received an internal mammary artery graft in addition to the vein grafts. Reported long-term patency rates of arterial grafts are superior to vein grafts, and associated with improved survival. Especially in patients with three-vessel disease, total arterial grafting could further reduce the recurrence of ischemic cardiac events. Although mid-term results were promising, long-term results were still unknown. In chapter seven, we evaluated clinical outcome in patients with three-vessel disease who underwent complete arterial revascularization using both pedicled ITAs and the gastroepiploic artery. The 12-year freedom from MACE was 75.5%. This was 86.9% for cardiovascular death, 93.3% for myocardial infarction, and 89.4% for re-intervention. When comparing these results with those obtained from recently published studies, our study indicated that the actuarial survival and actuarial probability of remaining free from myocardial infarction or re-intervention, or both, 10 years after the primary operation, was excellent. We also demonstrated that the beneficial effect of complete arterial revascularization was rather consistent through different strata of patients. Only in elderly patients (≥ 65 years) the benefit of complete arterial revascularization was less pronounced.

In addition to the literature review, we also compared the clinical outcome in our study cohort with that in a cohort of 281 patients with three-vessel disease who underwent revascularization and received ITAs in addition to vein grafts. These patients originated from the CABADAS study. MACE was significantly reduced in


- 134 -

patients who underwent complete arterial revascularization, with a hazard ratio of 0.43 (95% CI, 0.29-0.63, P < 0.001).

As with the literature review, this comparison is limited by potential bias, particularly due to the lack of random treatment allocation. We therefore cannot consider the analysis as conclusive, despite the use of propensity scores to address the non-randomized setting. It did, however, provide us with an estimate of the benefit of complete arterial revascularization, and added further evidence that the use of complete arterial revascularization using both pedicled ITAs and the gastroepiploic artery in patients with three-vessel disease resulted in excellent long-term clinical outcome. DISCUSSION AND FUTURE PERSPECTIVES In this thesis, the use of vitamin K antagonists was evaluated in several large patient populations. Further, a comparison between VKA, aspirin and dipyridamole was made in patients who underwent coronary-artery-bypass-grafting. The achieved level of anticoagulation is the most important determinant of effective and safe VKA therapy. We assessed the quality of anticoagulation that was achieved and the associated risk of thromboembolism and major bleeding in a number of cohorts with different indications for VKA therapy. We also evaluated differences in response to VKA and clinical outcome between different types of patients.

Considering the large amount of within and between patient variability, with only a small part of the variability explained by known risk factors, we specifically addressed the possibility to early identify patients who do not respond well to VKA after therapy is started. We evaluated the initial response after 30 days. The period of 30 days is long enough to generate a valid estimate of the initial response, and short enough to still have sufficient time to adjust treatment to prevent adverse clinical outcome. In this, we used the percentage of time within (ITTR), above and below the target range for each individual patient.

In the same line of reasoning, the occurrence of a minor bleed was evaluated. Next to the achieved quality of anticoagulation, a minor bleed proved to be an important predictor of subsequent major bleeding. In this respect, also clinically relevant minor bleeding must be mentioned. These were bleeds that were not serious by definition, but did require intervention and/or dose adjustment. A nose bleed, not

Chapter 8

- 135 -

yet self-limited after several hours is an example of such a bleed, which does have an impact on the subjective wellbeing of patients. It is good to notice that recent trials with new oral anticoagulants have included this type of severe minor bleeding as endpoint in their safety evaluations. The potential benefit of identifying poor response was demonstrated in our cohort of patients who underwent coronary-artery-bypass-grafting and were treated with VKA as part of the CABADAS study investigating the effects of aspirin, aspirin plus dipyridamole and VKA. Long-term beneficial effects of VKA over aspirin was achieved, but in a limited number of the patients with an optimal response of VKA. In the CABADAS study, continuation of VKA therapy after one year was left at the discretion of the treating physician. Remarkably, in the patients who continued VKA, the percentage non-optimal VKA was equal to that in the group of patients who stopped VKA. Using “response to VKA” data at the time of the decision to continue VKA therapy would have resulted in a higher percentage of patients with an optimal response in those who continued VKA, whereas more patients with non-optimal response would have switched to aspirin. Probably, patients would have benefited from such a different decision. Feedback to the treating physician on the achieved quality of anticoagulation is still not done regularly, but as illustrated above, could benefit the patients treated with VKA.

Promising alternatives for long-term VKA therapy are currently under development. These new oral anticoagulant drugs target on direct thrombin or factor Xa inhibition. The concept of factor Xa inhibition was confirmed in clinical trials using fondaparinux, and several factor Xa inhibitors are currently investigated. Ximelagatran was the first direct thrombin inhibitor tested in phase III trials and although development of this drug was stopped due to unexpected serious hepatic toxicity, the results confirmed that thrombin is a suitable target for new oral anticoagulants. It also showed that fixed dosing without monitoring the level of anticoagulation is possible, with efficacy and safety comparable to VKA.

Other direct thrombin inhibitors are now under investigation for various indications. Taken into consideration both the established efficacy of VKA and the requirements for a safe use, an equally effective and safe new oral anticoagulant could replace VKA as the primary choice of oral anticoagulant. Therefore, most studies target on non-inferiority. Recently, the RE-LY study was presented, in which


- 136 -

dabigatran was compared with warfarin in patients with atrial fibrillation. In more than one aspect, the RE-LY study generated very important results. Beyond expectation, the higher dose dabigatran (150 mg) was found to be superior to warfarin, without an increased risk of bleeding. The lower dose (110 mg) was declared non-inferior. In this study, the quality of warfarin therapy was considered adequate, with a mean TTR of 64%. After the RE-LY study, the approval of dabigatran is probably only a matter of time. Once the new oral anticoagulant drugs become available, also in patients already on VKA a switch must be considered. Response to VKA could then be used. In case of a poor response, new oral thrombin inhibitors should be considered.

A note of caution is related to the ability to counteract anticoagulation. When a rapid reversal of the action of VKA is indicated, this can be achieved by withholding VKA, in combination with the administration vitamin K and when further indicated the replacement of the depleted coagulation factors by administration of Prothrombin Complex Concentrate or Fresh Frozen Plasma. The new oral anticoagulants lack specific antidotes, which could especially be problematic in those drug with a longer half-life.

In patients who do well on VKA, a switch might still be less beneficial, especially when compliance is also taken into account. In RE-LY, where treatment was administered under clinical trial conditions, 15%-16% of the patients discontinued dabigatran, compared with 10% of the patients treated with warfarin. This was in part related to a higher rate of gastrointestinal complaints (11%-12% in dabigatran vs. 6% in warfarin). The higher rate, combined with the notion that compliance will be compromised in the “real life” setting, could influence the established benefit of dabigatran (also requiring daily dosing) over warfarin.

VKAs differ from many other drugs. Their big drawback, i.e. the highly variable dose-response relationship requiring continuous monitoring of the level of anticoagulation, also provides an advantage. The monitoring, with regular patient-physician contacts, also provides feedback on compliance. This feedback, but also the intensity and continuity of care itself are important factors that influence compliance/non-adherence. Furthermore, also patient knowledge and awareness towards the risks and benefits of their anticoagulant therapy are of importance. In this, the Dutch situation with a nationwide network of thrombosis services provides a unique setting for optimal compliance to VKA therapy. For the new oral anticoagulants, such a compliance might not be achieved, and especially in those drugs

Chapter 8

- 137 -

requiring a daily dosing, as dabigatran, the benefit that is observed under the clinical trial conditions might be reduced in “real life”. Therefore, after market approval a focus of further research should be on compliance as part of the postmarketing surveillance, especially where conditions under which VKA therapy is provided are optimal, as is the case in the Netherlands. The current presence of the nationwide network of Thrombosis Services in the Netherlands gives the unique opportunity to do this research, using the extensive knowledge of these highly experienced thrombosis services, before VKAs are abandoned to the past.

Appendix I

- 139 -

Nederlandse Samenvatting

Popular Summary in Dutch

- 140 -

NEDERLANDSE SAMENVATTING Bij langdurige antistollingsbehandeling wordt gebruik gemaakt van vitamine K antagonisten. Vitamine K is een belangrijke stof die verantwoordelijk is voor de aanmaak van stollingsfactoren. Een stof die vitamine K remt (een antagonist) kan daarom een goed antistollingsmiddel zijn. In Nederland zijn dergelijke antagonisten fenprocoumon (marcoumar ®) en acenocoumarol (sintrom ®). Sinds de registratie van vitamine K antagonisten als antistollingsmiddel nu 60 jaar geleden neemt het gebruik ervan nog steeds toe. In Nederland werden in 2007 meer dan 300.000 patiënten behandeld met vitamine K antagonisten (ongeveer 2% van de bevolking), met een toename van 10% over de laatste 5 jaren.

Ondanks de uitgebreide ervaring met vitamine K antagonisten behoren ze tot de top-10 van geneesmiddelen met het grootste aantal ernstige bijwerkingen. Dit heeft te maken met het feit dat het erg moeilijk is deze middelen goed toe te passen. Zij hebben een zeer belangrijke tekortkoming, namelijk een grote spreiding in de mate van ontstolling bij een gegeven dosering, zowel tussen verschillende patiënten als ook binnen dezelfde patiënt. Het is dan ook noodzakelijk dat het niveau van antistolling regelmatig (1 tot 2x per maand) wordt gecontroleerd en zo nodig bijgesteld. Dit gebeurt in Nederland door speciaal daarvoor toegeruste trombosediensten. Desondanks zijn er nog steeds momenten van onderbehandeling en overbehandeling, welke elkaar kunnen afwisselen. Bij onderbehandeling (te laag niveau van antistolling) is er sprake van een verhoogde kans op (opnieuw) optreden van een trombo-embolie (d.w.z. trombosebeen, longembolie of een beroerte), terwijl bij overbehandeling de veiligheid van de behandeling in het geding komt. Dit uit zich in een verhoogde kans op een ernstige bloeding.

Dit proefschrift beschrijft het resultaat van een aantal onderzoeken naar de werkzaamheid en veiligheid van vitamine K antagonisten in verschillende patiëntgroepen waarbij een langdurige antistollingsbehandeling noodzakelijk was. De rode draad daarbij was de relatie tussen het bereikte niveau van antistolling en het optreden van trombo-embolieën en bloedingen op de korte en middellange termijn. De focus lag op het bereikte niveau van antistolling in individuele patiënten, met als belangrijk extra aandachtspunt het in een vroeg stadium van de behandeling herkennen van de patiënten die niet goed instelbaar zijn (veel momenten van onder- en of overbehandeling).

Appendix I

- 141 -

Het eerste onderzoek dat in het kader van dit proefschrift is uitgevoerd (hoofdstuk 2) beschrijft het beloop van de behandeling met vitamine K antagonisten in een groep van 2300 patiënten die na het optreden van een trombosebeen of een longembolie langdurige antistollingsbehandeling nodig hadden. Alle patiënten werden daarbij regelmatig gecontroleerd door de trombosedienst. Hierbij werd het niveau van antistolling gemeten en de dosering zo nodig aangepast. Bij het meten van het niveau van antistolling wordt gekeken naar de snelheid waarmee het bloed van de patiënt stolt, welke uitgedrukt wordt in de stollingswaarde INR (International Normalized Ratio). De INR moest in deze patiëntengroep liggen binnen een smal bereik van INR 2.0 tot 3.0. Dit wordt het streefgebied genoemd, waarbinnen het niveau van antistolling optimaal is. Hoe hoger de INR, hoe sterker de antistolling en hoe “dunner” het bloed. Om de kans op met name onderbehandeling te verlagen is bij het controleren en bijstellen een iets ruimer gebied gehanteerd, te weten INR 2.0 tot 3.5.

Uit de controles bleek dat, ondanks veelvuldige aanpassingen, het moeilijk is om alle patiënten langdurig op het optimale niveau van antistolling te houden. Om dit te evalueren is gebruik gemaakt van het individuele percentage van de tijd dat een patiënt binnen het streefgebied verbleef (Individual Time within Target Range, ITTR). In de onderzoekgroep was de gemiddelde ITTR 63%, waarbij men gemiddeld 11% onder het streefgebied zat en 26% erboven. Hierbij is uitgegaan van het ruimere gebied van INR 2.0 tot 3.5. In een kwart van de patiënten was het zeer moeilijk om het optimale niveau te bereiken en langdurig te behouden. Deze patiënten waren maximaal in slechts 45% van hun behandeltijd binnen het streefgebied. Met name deze patiënten lieten een hoog risico op zowel een hernieuwde trombo-embolie als ook een ernstige bloeding zien (166 trombo-embolieën en 87 ernstige bloedingen per 1000 behandeljaren). In de patiënten waarbij een stabieler niveau van antistolling was bereikt was dit significant lager (49 om 21 bij 45-65% en 46 om 19 trombo-embolieën en ernstige bloedingen per 1000 behandeljaren bij de patiënten met een percentage boven de 65%).

Juist vanwege het hoge risico op een trombo-embolie en een ernstige bloeding bij de patiënten met zeer lage ITTRs is het belangrijk om deze patiënten zo vroeg mogelijk te herkennen. Daarom hebben we gekeken naar de ITTR gedurende de eerste 30 dagen. Daaruit bleek dat de patiënten met de laagste ITTRs gedurende de eerste 30 dagen de grootste kans hadden om uiteindelijk gedurende de totale behandelduur veelvuldig buiten het streefgebied te verblijven. Het ging daarbij om een


- 142 -

kwart van de patiënten waarbij de 30-dagen ITTR maximaal 37% bedroeg, met een mediaan van slechts 20%. Hiermee kan een vroege evaluatie van het bereikte niveau van antistolling een bijdrage leveren aan het vroegtijdig herkennen van patiënten met een verhoogde kans op een onvoldoende mate van antistolling.

In het tweede onderzoek is gekeken naar de kwaliteit van antistolling en het optreden van trombo-embolieën en ernstige bloedingen bij drie andere patiëntgroepen, nu met cardiovasculaire problemen. Dit waren 4454 patiënten met respectievelijk een hartinfarct (N=1012), boezemfibrilleren (N=2614) of een kunsthartklep (N=828), allen langdurig behandeld met vitamine K antagonisten (hoofdstuk 3).

De resultaten van dit onderzoek bevestigen het beeld van het eerste onderzoek. Ook in het tweede onderzoek was er regelmatig sprake van een te lage of te hoge mate van antistolling, waarbij een hoog risico op het optreden van een trombo-embolie en een ernstige bloeding werd gezien in de patiënten met de laagste ITTRs (het “slechtste” kwart van de patiënten). Bij patiënten met hogere ITTRs was hiervan geen sprake. Dit beeld duidt op het bestaan van een kritische grens voor de ITTR, waaronder zowel de effectiviteit alsook de veiligheid van de antistollingsbehandeling in het geding komt.

Men kan de antistollingsbehandeling met vitamine K antagonisten vergelijken met het lopen over een evenwichtsbalk. Er zijn voortdurend correcties nodig om het evenwicht te bewaren, waarbij de kans dat het mis gaat altijd aanwezig is. De meeste patiënten zijn voldoende in staat om veilig de andere kant van de balk te bereiken, ondanks toch ook veelvuldige momenten van disbalans. Echter, ondanks alle “steun” van de trombosedienst (controles en bijstellingen) is er een groep patiënten die teveel “wankelen”. Het zijn deze patiënten die vroegtijdig herkend moeten worden. Ook in dit onderzoek was dit mogelijk met behulp van de response gedurende de eerste 30 dagen. Patiënten met een hoge mate van instabiliteit gedurende de eerste 30 dagen hadden de grootste kans op een onvoldoende niveau van antistolling over de totale behandelingsduur, met uiteindelijk een grote kans om van de balk af te vallen. Tijdens het gebruik van vitamine K antagonisten treden per 1000 behandelingsjaren 14 tot 33 ernstige bloedingen op, welke levensbedreigend zijn. Daarnaast gaat een langdurige antistollingsbehandeling ook gepaard met soms meerdere niet-ernstige bloedingen. Dit zijn dan bijvoorbeeld bloed bij de urine, een bloedneus of een

Appendix I

- 143 -

bloeding van het oogslijmvlies (waarbij het oogwit rood geworden is), maar ook grote blauwe plekken zonder duidelijk voorafgaand stoten. Het aantal bloedingen kan oplopen tot 200 tot 300 bloedingen per 1000 behandelingsjaren. Deze bloedingen zijn veelal mild maar kunnen ook aanleiding geven tot het tijdelijk onderbreken van de behandeling.

Het optreden van zo’n bloeding kan leiden tot een “verscherpt toezicht” door de trombosedienst, hetgeen de kwaliteit van de behandeling op de lange termijn positief kan beïnvloeden, met een lager risico op een ernstige bloeding als gevolg. Het kan ook een eerste teken zijn van een verhoogd bloedingsrisico, waarbij ondanks extra inspanningen om de behandeling te optimaliseren een ernstige bloeding toch optreedt. Dit is onderzocht in alle 6758 patiënten uit de eerdere onderzoeken (hoofdstuk 4).

Bij 20% van alle patiënten was een niet-ernstige bloeding opgetreden (251 bloedingen per 1000 behandelingsjaren). Na zo’n bloeding was het risico op een ernstige bloeding significant hoger; binnen 3 maanden na het optreden ervan zelfs 3.6-maal hoger. Daarnaast hadden de patiënten met een lagere kwaliteit van antistolling een verhoogd risico op een ernstige bloeding, namelijk 2.8-maal hoger. Overige risicofactoren voor een ernstige bloeding waren maligniteit, leeftijd (tussen 55-70 jaar en met name boven 70 jaar, geslacht (man), en het gebruik van NSAIDs tijdens de behandeling (NSAIDs zijn ontstekingsremmers zonder steroïden, zoals bijvoorbeeld ibuprofen of aspirine, welke ook de mate van antistolling kunnen beïnvloeden).

De conclusie van dit onderzoek was dat een niet-ernstige bloeding geassocieerd is met een verhoogde kans op een daaropvolgende ernstige bloeding. Ook de bereikte kwaliteit van antistolling is daarbij van groot belang. Het gaat dan niet alleen om de mate van ontstolling op het moment van de bloeding, maar met name om het bereikte niveau over de gehele voorafgaande behandelingsduur. In aansluiting op het onderzoek naar het effect van de kwaliteit van antistolling op de korte en middellange termijn is er vervolgens onderzoek gedaan naar de impact van een niet-optimale antistollingsbehandeling met vitamine K antagonisten op de lange termijn. Hierbij is gebruik gemaakt van een groep patiënten die eind jaren ‘80 begin jaren ‘90 hadden deelgenomen aan de CABADAS studie, een groot onderzoek naar het effect van respectievelijk aspirine, dipyridamol en vitamine K antagonisten (zie hoofdstuk 5). Het ging hierbij om patiënten die een coronaire bypassoperatie hadden ondergaan. Bij zo’n bypassoperatie wordt met een stuk bloedvat een omleiding


- 144 -

gemaakt om een afsluiting van één van de kransslagaders heen. Vervolgens worden de patiënten nabehandeld om nieuwe afsluitingen te voorkomen.

Er werden 233 patiënten geselecteerd die allen behandeld waren met vitamine K antagonisten. Van deze patiënten is de bereikte kwaliteit van antistolling gedurende het eerste jaar in kaart gebracht. Vervolgens is over een periode van 14 jaar gekeken naar eventuele verschillen in cardiale sterfte en het optreden van een hartinfarct.

De bereikte kwaliteit van antistolling was bij 40% (N=93) van de patiënten als optimaal beoordeeld. Gedurende de eerste 8 jaar na de bypassoperatie was er geen verschil in cardiale sterfte of het optreden van een hartinfarct tussen de patiënten met een optimale versus niet-optimale kwaliteit van antistolling. Na 8 jaar was er echter sprake van een 3-maal hoger risico in de patiënten met een niet-optimale kwaliteit van antistolling gedurende het eerste jaar.

Naast cardiale sterfte of het optreden van een hartinfarct is ook gekeken naar de een hernieuwde bypassoperatie of dotterprocedure. Er was geen relatie tussen de kwaliteit van antistolling en het uitvoeren van een nieuwe ingreep. In beide groepen was het percentage patiënten die een nieuwe ingreep moest ondergaan 35%. Een mogelijk verklaring hiervoor zijn de frequente controles die bij deze patiënten werden gedaan vanwege hun vitamine K antagonisten behandeling. Door deze frequente controles zijn patiënten bij terugkeren van hartklachten mogelijk sneller terugverwezen naar de cardioloog voor verder onderzoek. Dit kan leiden tot eerder en vaker hernieuwd opereren. Deze mogelijke beïnvloeding door frequente controles werd bevestigd door een vergelijking met de overige patiënten uit de CABADAS studie die met aspirine of dipyridamole waren behandeld. In deze patiënten, zonder frequente controles, was het percentage re-operaties slechts 18% en 17%. Naast de bovenstaande vergelijking op alleen re-operaties is bij alle Nederlandse patiënten uit de CABADAS studie (N = 726) ook gekeken naar het optreden van cardiale sterfte, het optreden van een hartinfarct en het uitvoeren van een re-operatie (inclusief dotterprocedure) als een gecombineerde uitkomstmaat (hoofdstuk 6). De afwezigheid hiervan is een maat voor succes van de eerste operatie. Na 14 jaar bleek bij 52% van de patiënten een van deze drie ernstige gebeurtenissen te zijn opgetreden. Zoals al eerder opgemerkt was er een groot verschil in de noodzaak tot re-operaties. Dit was mogelijk toe te schrijven aan het grote verschil in patiënt-dokter contacten als gevolg van behandelingen, en dus niet zozeer vanwege een verschil in de effectiviteit

Appendix I

- 145 -

van de behandelingen op zich. Met betrekking tot cardiale sterfte en een hartinfarct was er geen significant verschil.

Onze resultaten bevestigden daarmee het huidige beleid bij patiënten na een bypassoperatie, waarbij aspirine het eerste keuze geneesmiddel is ter voorkoming van verdergaande hartziekte. Deze conclusie is gebaseerd op gegevens verkregen uit een zeer lange follow-up duur. Het gebruik van vitamine K antagonisten moet daarbij beperkt worden tot die patiënten die vanwege een andere ziekte langdurige antistolling met vitamine K antagonisten nodig hebben. Het laatste onderwerp van onderzoek was het gebruik van arteriële omleidingen bij de eerder genoemde bypassoperatie (hoofdstuk 7). Bij deze omleidingen, die bij de kransslagaders van het hart worden “aangelegd”, kan gebruik gemaakt worden van een stuk ader, maar ook van een stuk slagader. Het gebruik van een slagaderlijke omleiding is complexer, maar het voordeel is dat op de korte en middellange termijn de kans op een afsluiting van de omleiding kleiner is dan bij een omleiding met een ader. Zo’n voordeel is met name van belang voor patiënten met zogenaamd drievaatslijden, waarbij in alle drie de kransslagaders van het hart de bloeddoorstroming ernstig belemmerd wordt. Echter, over het voordeel op de lange termijn waren minder gegevens beschikbaar.

In het UMC Groningen is begin jaren ‘90 bij een serie patiënten met drievaatslijden gebruik gemaakt van alleen slagaderlijk materiaal om de omleidingen te maken. Vervolgens is bij deze groep de lange termijn uitkomst gemeten, waarbij over een periode van 12 jaar in alle patiënten de cardiale sterfte, het optreden van een hartinfarct en het uitvoeren van een re-operatie werd geëvalueerd. Na 12 jaar was bij 25% van de patiënten zo’n ernstige gebeurtenis opgetreden. Voor alleen cardiale sterfte was dit 13%, voor een hartinfarct 7% en voor een re-operatie 11%. In vergelijking met eerdere gepubliceerde gegevens, alhoewel moeilijk te vergelijken, waren deze lange termijn resultaten zeer goed. Een vergelijking met 281 patiënten met drievaatslijden uit de CABADAS studie, waarbij naast een slagader ook aders waren gebruikt om omleidingen te maken, bevestigde dit. Uit deze vergelijking kwam naar voren dat bij het gebruik van alleen slagaderlijke omleidingen de kans op een ernstige gebeurtenis (cardiale sterfte, hartinfarct of re-operatie) meer dan 2-maal lager was. Dit verschil was sterk significant.


- 146 -

Samenvattend is in dit proefschrift verslag gedaan van een aantal onderzoeken naar het gebruik van vitamine K antagonisten in verschillende patiëntgroepen. Hierbij is gekeken naar effectiviteit en veiligheid, met het accent op het bereikte niveau van antistolling in individuele patiënten. Daarnaast zijn de lange termijn resultaten van de behandeling met vitamine K antagonisten geëvalueerd en vergeleken met twee andere middelen.

De behandeling met vitamine K antagonisten kenmerkte zich door een grote variabiliteit in de mate van ontstolling. In dat kader was een belangrijk aspect van onderzoek de mogelijkheid om de patiënten die een onvoldoende niveau van antistolling hadden vroegtijdig te herkennen. Dit onderzoek laat zien dat met een evaluatie van het bereikte niveau van antistolling gedurende de eerste 30 dagen de patiënten met een sterk verhoogde kans op een onvoldoende niveau van antistolling daadwerkelijk vroegtijdig herkenbaar zijn. Op dit moment worden er nieuwe alternatieven ontwikkeld voor de langdurige antistollingsbehandeling met vitamine K antagonisten, waarbij de regelmatige controles van het bereikte niveau van antistolling niet langer nodig zijn. De resultaten van deze onderzoeken zijn zeer veelbelovend en het is dan ook slechts een kwestie van tijd totdat deze alternatieven beschikbaar zijn. Een belangrijke vraag wordt dan of alle patiënten baat zullen hebben bij een keuze voor deze nieuwe middelen. Een groot deel van de patiënten “doen het goed” op vitamine K antagonisten. Wel zijn hier veel controles voor nodig, waarmee het een zeer intensieve behandeling is. Dit is een groot nadeel, maar geeft ook een groot voordeel. Door de regelmatige controles wordt de therapietrouw positief beïnvloed. Bij de nieuwe middelen zijn deze controles niet langer noodzakelijk, en daarmee kan in de dagelijkse praktijk de bewezen effectiviteit en veiligheid van de nieuwe middelen door een lagere therapietrouw nadelig worden beïnvloed. Juist in de Nederlandse situatie, waar alle patiënten die behandeld worden met vitamine K antagonisten door gespecialiseerde trombosediensten intensief gecontroleerd en begeleid worden, kan dit een factor van betekenis zijn.

Ook moet de mogelijkheid tot het snel corrigeren van de antistolling worden genoemd. Ruime ervaring heeft geleerd dat indien noodzakelijk, zoals bijvoorbeeld bij een bloeding of een spoedoperatie, de mate van ontstolling bij het gebruik van vitamine K antagonisten snel en adequate teniet is te doen. Vooralsnog is hiervan bij de nieuwe middelen nog geen sprake.

Appendix I

- 147 -

De vraag of alle patiënten beter af zijn met de nieuwe middelen is daarmee dus nog niet beantwoord. Het huidige netwerk van zeer ervaren trombosediensten, zoals we dat in Nederland kennen, biedt unieke mogelijkheden om naar deze belangrijke vraag verder onderzoek te kunnen doen.

Appendix II

- 149 -

Dankwoord

Dankwoord

- 150 -

DANKWOORD Eigenlijk had ik mijn dankwoord al een paar jaar geleden geschreven. Met weinig woorden, maar wel een krachtige boodschap. Prof. dr. Jan van der Meer, Jan, heel veel dank! Zo kort, dat moet je aangesproken hebben. Natuurlijk, om recht te doen aan de andere mensen die hebben bijgedragen aan deze promotie wil ik het hier natuurlijk niet bij laten. Maar toch, vanaf het begin was dit onderzoek iets wat ik samen met jou mocht doen, en daar ben ik erg dankbaar voor. In 2009 moest het gaan gebeuren, de omslag moest er maar eens omheen, het was tijd voor een feest. Hoe anders liep het. De basis van dit proefschrift werd gelegd in 2000, in de vorm van een afstudeerproject van de mastersopleiding Klinische Epidemiologie aan de Erasmus Universiteit Rotterdam. Onder de stimulerende begeleiding van prof. dr. Jan Tijssen werd het al snel duidelijk dat het niet bij dat ene project zou blijven. Ook prof. dr. Harrie Crijns wil ik hierbij noemen. Als toenmalig hoofd van de afdeling Cardiologie vond ook jij het een heel goed idee om met eigen onderzoek verder te gaan. Iets wat binnen het Trial Coordination Center, een afdeling gericht op het ondersteunen van onderzoek van anderen, toen niet gebruikelijk was. Harrie en Jan, dank voor de steun en het uitspreken van het vertrouwen in die vroege fase, ook al moeten jullie gedacht hebben dat het wel eens moeilijk kon gaan worden. Harrie, ik waardeer het zeer dat je zitting hebt willen nemen in de leescommissie. De andere leden van de leescommissie, prof. dr. Harrie Büller en prof. dr. Martin Prins, ik ben blij dat ook jullie bij mijn promotie deze belangrijke rol hebben willen vervullen. Wat is een onderzoek zonder betrouwbare gegevens. Dit onderzoek is grotendeels gebaseerd op gegevens die belangeloos beschikbaar zijn gesteld door de Trombosedienst Groningen. Daarbij kon ik ook gebruik maken van hun grote expertise op het terrein van de behandeling met vitamine K antagonisten. Hiervoor ben ik dr. Peter Bijster van LabNoord dankbaar. Indrukwekkend was het om te zien hoe het bij de Trombosedienst niet alleen gaat om meten en bijstellen, maar zeker ook om het begeleiden van de patiënten. In dit verband wil ik zeker ook Margriet Piersma-

Appendix II

- 151 -

Wichers noemen. Margriet, het begon met een sprong in het diepe, en om een duikterm te gebruiken, je was een buddy waar ik op kon vertrouwen. Met name door jouw medewerking ligt dit boekje er nu. Vaak op de achtergrond, maar altijd beschikbaar. Dank je, en laat ik hier de wens uitspreken dat we het gezamenlijke onderzoek naar de optimale zorg voor patiënten met langdurige antistollingsbehandeling kunnen voortzetten. Ook dr. Jan Grandjean, dr. Adriaan Voors, drs. Gerald Panday en prof. dr. Piet Boonstra wil ik bedanken. Het onderzoek naar de lange termijn resultaten na totale arteriële revascularisatie was een lang en leerzaam traject. Beste Piet, mede door jouw volharding is het een mooi hoofdstuk geworden. De laatste loodjes waren zwaar. Zowel prof. dr. Felix Zijlstra, dr. Karina Meijer, prof. dr. Hanneke Kluin en prof. dr. Hans Hillege waren er om mij te helpen. Felix en Karina, jullie inbreng bij de laatste stukken was zeer waardevol en heb ik zeer gewaardeerd. Hanneke, ik ben je dankbaar dat je de rol van Jan als promotor hebt willen overnemen, en zeker ook voor de manier waarop je dat hebt gedaan. Het voelt goed. Hans, vanaf je terugkomst bij het TCC heb je het belang van deze promotie benadrukt. Jij was degene die het tempo erin wilde houden. Dit was gezien de setting waarin het onderzoek plaatsvond, naast het andere werk, vaak moeilijk. De extra ruimte en steun die ik van je kreeg om toch meer tempo te kunnen maken was voor mij erg belangrijk. Mede daardoor is de eindstreep gehaald, veel dank daarvoor. Ten slotte wil ik de anonieme verpleegkundige noemen, die mij een lift gaf naar het (toen nog) AZG waardoor ik nog net op tijd de toelatingstest voor de Academie voor Fysiotherapie kon doen. De deur tot Groningen werd daarmee geopend. Verder natuurlijk al die andere toevallige ontmoetingen met mensen die mij uiteindelijk hebben gebracht waar ik nu ben. De toevoeging “toevallige” is daarbij zeker geen (dis)kwalificatie van de persoon die het betreft, maar een benadrukken van de kansen die ik gekregen heb om al die voor mij belangrijke mensen te ontmoeten. In de loop van de jaren zijn het er veel geweest, met één ben ik heel erg gelukkig. Astrid, dit boek is geschreven, ons boek is nog lang niet af.

university of groningen individual approach towards ... · university of groningen individual...

Documents