University of Groningen the Pill and Thrombosis Van Vlijmen, Elizabeth

University of Groningen

The pill and thrombosis van Vlijmen, Elizabeth Femma Willemien

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record

Publication date: 2016

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA): van Vlijmen, E. F. W. (2016). The pill and thrombosis. Rijksuniversiteit Groningen.

Copyright Other 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 the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne- amendment.

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

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

Download date: 08-10-2021

Financial support for the printing of this thesis was kindly provided by the Dutch Medicines Evaluation Board and UMCG/GUIDE.

ISBN: 978-90-367-8569-3

ISBN: 978-90-367-8566-2 (printed version) ISBN: 978-90-367-8569-3 (digital version)

Cover design: Harry Zijderveld Cover painting: Lita Cabellut 2013© ‘Billy Holiday’ from the series ‘Strong women’ Layout: Harry Zijderveld - Graphic Design Printed by: drukkerij van Ark, Haren The Pill and thrombosis

Proefschrift

ter verkrijging van de graad van de doctor aan de Rijksuniversiteit Groningen op gezag van de rector magnificus prof. dr. E. Sterken en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op woensdag 17 februari 2016 om 14.30 uur

door

Elizabeth Femma Willemien van Vlijmen

geboren op 5 februari 1958 te Naarden Promotores Prof. Dr. Karina Meijer Prof. Dr. J. C. Kluin-Nelemans

Co-promotor Dr. N.J.G.M. Veeger

Beoordelingscommissie Prof. Dr. M. V. Huisman Prof. Dr. F. R. Rosendaal Prof. Dr. S. A. Scherjon Paranimfen: Clarissa van Vlijmen-van Keulen Annemarie van Leijsen-de Vries

Voor mijn ouders

TABLE OF CONTENTS

Chapter 1 9 General introduction and outline of thesis

Chapter 2 41 Oral contraceptives and the absolute risk of venous thromboembolism in women with single or multiple thrombophilic defects Archives of Internal Medicine 2007;167:282-289

Chapter 3 59 Thrombotic risk during oral contraceptive use and pregnancy in women with factor V Leiden or Prothrombin mutation: a rational approach to contraception Blood 2011;118:2055-2061

Chapter 4 79 Combined hormonal contraceptives, thrombophilia and the risk of venous thromboembolism: a systematic review and meta-analysis Submitted

Chapter 5 101 A male or female family history of VTE: does it influence hormone-related VTE risk? Submitted

Chapter 6 117 Clinical profile and recurrence rate in women with venous thromboembolism during combined hormonal contraceptive use: results of a prospective cohort study British Journal of Hematology 2015;DOI 10.1111/bjh.13534

Chapter 7 127 Summary, discussion and future perspectives

Chapter 8 157 Nederlandse samenvatting

Chapter 9 173 Dankwoord

Chapter1 General introduction and outline of thesis

Introduction

Chapter 1

Introduction

Combined oral contraceptives and venous thromboembolism

The combined oral contraceptive, i.e. the ‘Pill’, was the first medicinal product developed that had to be taken every day on a regular basis by persons not suffering from any disease. Today, world-wide more than 100 million women start the day with taking the ‘Pill’.1 Although the Pill is a safe and highly effective contraceptive method in the vast majority of women, in a small group of women its use is associated with a rare but potentially serious adverse event, i.e. venous thromboembolism (VTE). Up to now, this association has been the topic of many studies, including the studies presented in this thesis, and probably there are many more to follow, as even after the recent 50th anniversary of the Pill this risk still has not been solved.

Combined oral contraceptives

Composition

Combined oral contraceptives (COCs) contain a combination of a synthetic estrogen and a synthetic progestogen. The estrogen component is ethinylestradiol and recently also estradiol, but there are many different types of synthetic progestogens, e.g. levonorgestrel, norethisterone, desogestrel, gestodene, dienogest, drospirenone, norgestimate, etonogestrel, norelgestromin, nomegestrol, chlormadinone, and cyproterone. Initially, the dose of ethinylestradiol was up to 150 micrograms, but since the early seventies the dose is steadily lowered to 35-15 micrograms (so-called sub-50 Pills) that are used today. Also the progestogen doses have decreased substantially. Further, currently several COCs have standard and low-dose versions. Additionally, there are not only oral combined contraceptives, but also non-oral preparations such as a vaginal ring to be applied once a month and a patch to be applied once weekly.

11 Introduction

The reason why there are so many synthetic progestogens is that the initially developed progestagens, besides the desired progestogenic effects, also had androgenic effects. Androgenic properties could result in androgenic side effects like acne, hirsutism, and negative effects on the lipid metabolism. Development therefore focused on progestagens having less androgenic properties.

The consecutive development of new COCs, i.e. with a new synthetic progestogen, is divided into ‘generations‘, and is related to the time point of their market launch. ‘First generation’ is related to the very first COCs, which were approved around 1960. An overview of several generations of COCs with different synthetic progestogens and their first date of introduction is given in Table 1.

Table 1. Combined hormonal contraceptives divided into generations

Type of progestogen Generation Market introduction Ethinylestradiol/norethisterone-containing COC (Modicon©) 1st generation 1960

Ethinylestradiol/levonorgestrel-containing COC (Microgynon 30©) 2nd generation 1974 Ethinylestradiol/norgestimate-containing COC (Cilest©) 2nd generation 1990

Ethinylestradiol/desogestrel-containing COC (Marvelon©) 3rd generation 1984 Ethinylestradiol/gestodene-containing COC (Femodeen©) 3rd generation 1984

Ethinylestradiol/cyproterone-containing COC (Diane 35©) 4th generation 1987 Ethinylestradiol/chlormadinone-containing COC (Madinelle©)a 4th generation 1995 Ethinylestradiol/drospirenone-containing COC (Yasmin©) 4th generation 2000 Ethinylestradiol/etonogestrel-containing vaginal ring (NuvaRing©) 4th generation 2001 Ethinylestradiol/norelgestromin-containing patch (Evra©) 4th generation 2002

Estradiol/dienogest-containing COC (Qlaira©) ‘5th’ generation 2008 Estradiol/nomegestrol-containing COC (Zoely©) ‘5th’ generation 2011

a: ethinylestradiol/chlormadinone-containing COCs are not available in NL

However, as there are now many different preparations, it is suggested not to use these terms any longer, as they lead to confusion. In practice this has already resulted in studies on COCs and thromboembolism with e.g. COCs being catego-

12 Introduction

rized as a ‘3rd generation’, ‘4th generation’ or as a ‘2nd generation’, which makes- the interpretation of clinical results for the individual progestogens such as risk of VTE, challenging.2

History of combined hormonal contraceptives

Around 1920, several Austrian scientists demonstrated the occurrence of temporary contraception by transplantation of ovaries of a pregnant animal into a non- pregnant animal. These experiments were followed by studies in mice showing that extracts of ovaries could inhibit ovulation. These first studies were performed by Ludwig Haberlandt, an Austrian professor of physiology. He is seen as the Grandfather of the ‘Pill’, as he was the first to perform experiments with the goal to obtain contraception by using female sex hormones. His aim was to develop hormonal contraception by mimicking a pregnant state, which would create a ‘temporary hormonal sterilization’. In 1931, he started the production of a hormonal extract ‘Infecundin’ in co-operation with the Hungarian pharmaceutical company Gedeon-Richter. However, he was severely criticized by his colleagues and in the lay public press, who accused him of hindering the unborn life. Despite all opposition, he started the first clinical trials. Unfortunately, he died at an early age and his co-worker Otto Fellner lost his life during World War II, which brought an end to this first initiative.3,4

Although Gabriel Fallopinus already in 1561 mentioned the existence of a ‘corpus luteum’, of which the structure is described by the Dutch anatomist Volcher Coiter in 1573,5 its function in regulating the menstrual cycle remained unknown for long. Only at the end of the nineteenth century scientists in Europe and USA discovered the endocrine function of this temporary existing ‘gland’ after ovulation has taken place. Between 1930-1937, scientists unraveled the action of the corpus luteum hormone as well as its chemical structure, and named the hormone ‘progesterone’.6,7

But at the time, it was still difficult and very expensive to produce sex hormones, as the only way was to process thousands of animal ovaries to retrieve a small amount of sex hormones. In early 1941, the American chemist Russell Marker solved this problem by the discovery of the Mexican wild Yam plant, which contained the plant steroid diosgenin. This plant was said used by Mexican

13 Introduction

women for treatment of menstrual cramps. This phytoestrogen could be used as a source to produce large amounts of progesterone. Carl Djerassi in Mexico and other chemists in the US improved the potency of the ‘Yam’ progesterone by removing the carbon at the 19-position, which led to the first synthetic progestogen ‘norethindrone’. Meantime in Europe, chemists at Schering AG in Germany developed the potent synthetic estrogen ‘ethinylestradiol’. The latter became the estrogen compound in future developed combined oral contraceptives.8

In 1944, disappointed by the negative attitude of politics of academic institutions at the time, the physiologists Gregory Pincus, Hudson Hoagland, Min-Chueh Chang, and John Rock, gynecologist, started the ‘Worchester foundation’, a private research center for experimental biology. With financial support and encouragement of biologist Katharine McCormick and Margaret Sanger of the Planned Parenthood association, they focused on development of oral hormonal contraception.9.10 This collaboration led to a first study in women who received oral natural progesterone. In this first study, a 20-day dose regimen starting at day 5 of the menstrual cycle was chosen as it covers the period in the cycle during which ovulation occurs. Discontinuation after these 20 days would result in a withdrawal bleeding thereby mimicking the normal menstrual bleeding, which could reassure women that they are not pregnant.

In 1956, a first study was performed in women who received a synthetic progestogen and results were published in Science.11 The results indicated that ovulation inhibition could be achieved and that after discontinuation a rapid return to fertility was observed. However, the synthetic progestogen norethindrone used appeared to be contaminated with the estrogen ‘mestranol’. When repeating this study using non-contaminated progestogen, the women reported unwanted irregular (breakthrough) bleeding. Based on these results, it was concluded that the progestogen should be combined with a certain amount of estrogen to keep the endometrium stabilized, which avoids occurrence of breakthrough bleeding. The results of these studies showed the usefulness of combining a progestogen with an estrogen, i.e. the concept of the combined contraceptive pill was born. An additional study in which women used norethindrone 10 mg combined with ethinylestradiol 0.15 mg showed a decrease in risk of pregnancy to 2.4%.12 When in 1957 the pharmaceutical company Searle had submitted an application for approval of this combined oral contraceptive pill, Pincus and Rock strongly contributed to convince the US food and drug administration to approve Enovid©.

14 Introduction

This first ‘Pill’, with a huge dose of 150 micrograms mestranol and 10 mg norethy- nodrel, was initially approved for ‘treatment of miscarriages and severe menstrual disorders’, with a patient leaflet mentioning ‘temporary infertility’ as an ‘adverse effect’. But in 1960, the indication of contraception was accepted.13

Mechanism of contraceptive action

Combined oral contraceptives contain an estrogen and a progestogen. The progestogen inhibits ovulation, as it primarily suppresses the production of lutei- nizing hormone (LH), thereby preventing the LH-surge needed to start ovulation. The estrogen component suppresses the production of follicle stimulating hormone (FSH), which will prevent follicular growth and development of a domi- nant follicle. The estrogen component also provides stabilization of the endometrium, which will reduce risk of breakthrough bleeding and it also potentiates the progestogenic activity of the progestogens.10 This potentiating has the advantage that the progestogen dose can be reduced. Initially, a dose regimen was chosen to mimic the menstrual cycle, which included a period of 21 days of tablet intake, followed by a 7-day break in which a withdrawal bleeding occurred due to the drop in hormone levels. Nowadays, several variants are introduced in which the 7-day break is reduced to 4 or even 2 days per month, and recently there are dosage schemes approved in which the ‘Pill’ can be taken continuously for up to 120 days.

Contraceptive efficacy in comparison to other contraceptive methods

Combined oral contraceptives (COCs) are one of the most efficacious contraceptives available. In Europe, combined oral contraceptives are the most frequently used type of birth control; about 85% of women in Western Europe have taken the pill during their fertile life. When taken according to the recommended dose regimen (perfect use), the risk of pregnancy, expressed as the Pearl Index, is only about 0.3% (Pearl Index: the number of failures of a contraceptive method [pregnancies] per 100 woman who use the contraceptive for one year). Only those hormonal contraceptive methods which are not dependent of compliance to regular intake, i.e. progestogen-containing IUDs and implants have an even lower pregnancy risk of around 0.1%, as presented in the table of failure rates based on US figures. However, Pearl Index for typical use (actual use which also includes pregnancies due to non-compliance to the dose regimen) is far lower in Europe than in the US, of which the reason is still not elucidated, although it is suggested that less compliance and the higher BMI of US women may play a role.14,15

15 Introduction

Table 2. Percentage of US women experiencing an unintended pregnancy during the first year of use per contraceptive method

% Women experiencing an unintended pregnancy within the first year of use

Method Typical use1 Perfect use2 No method 85 85 Spermicides 29 18 Diaphragm 16 6 Condom: - female 21 5 - male Combined oral contraceptive 8 0.3 Desogestrel-only pill 8 0.3 Combined hormonal patch (Evra©) 8 0.3 Combined hormonal vaginal ring (NuvaRing©) 8 0.3 Medroxyprogesterone injection (Depo-Provera©)* 3 0.3 Levonorgestrel-containing IUD (Mirena©)* 0.2 0.2 Copper-containing IUD* 0.8 0.6 Etonogestrel containing implant (Implanon©)* 0.05 0.05 Adapted from Trussel J. Contraception 2011;83:397-40414 1: typical use shows how effective the different methods are during actual use (including inconsistent or incorrect use). 2: Pregnancy rates during perfect use show how effective methods can be, where perfect use is defined as following the directions for use. * : LARC’s, long-acting reversible contraception

The social, cultural and economic impact of the approval of the Pill

The introduction of the pill caused an earthquake in the Western world, as it rearranged the relationships between men and women completely. For the first time, women were able to plan their own life, by having the opportunity to decide when to get pregnant and how many children she would have. This meant that she could finish an education, have a job and have a life of her own. Up till then, the fear of unexpected pregnancy before marriage and in marriage, and the frequent state of pregnancy were women’s major burden. So there is a straight line between the introduction of the Pill and the changing in family structure. Initially only married women could have a receipt, but when the Pill became also available to unmarried women, this coincided with the increase in the age at first marriage and the increase in women in professional degree programs.16 Its introduction was the basis of the large positive change up to the economic and social status of the women in the world today.

16 Introduction

At its recent 50th anniversary, the Pill was voted by many as the greatest medical advance of the 20th century, as it enabled women to change their domestic position and plan families around their own careers.

As to the situation in the Netherlands, until the introduction of the ‘Pill’, the Netherlands was prudish society with high birth rates. Contraception was an inde- cent subject, and physicians were ignorant to the subject or didn’t want to have anything to do with it. To sell contraceptives in public was forbidden by law. In 1950, the editors of the ‘Nederlands Tijdschrift voor Geneeskunde’ apologized for having published two articles dealing with different methods for birth control.17,18 With the introduction of the ‘Pill’, a rigorous change in attitude started which was not to be stopped, as in no time a large number of Dutch women started using it. Initially, there was lot of opposition of the church, politicians and doctors, the latter as they were not convinced it would actually work. But in 1963, by wording of the Dutch Bishop Bekkers, the Dutch Catholic church agreed that married couples could decide themselves if they would use the Pill. Within 10 years after its introduction, the Netherlands had highest percentage of Pill users in the world.

Venous thromboembolism

Venous thromboembolism (VTE) is defined as a non-physiologic formation of a thrombus in a vein. In 1846, the German pathologist Virchow hypothesized 3 main mechanisms to cause thrombosis: damage to the vessel wall, stasis of the blood stream, and alterations in the blood composition. The most common location is deep vein thrombosis of the leg and when this thrombus partly deta- ches (embolizes) it can be transported upwards through the right heart into one of the arteries of the lung and cause pulmonary embolism.19,20

VTE is considered a multifactorial disease, based on interactions between genetic and acquired risk factors. In the general population, many risk factors for VTE are present, and often within the same individual, but an interaction of hereditary and acquired risk factors may eventually lead to the development of VTE. Actually, the current view is that the presence of several e.g. more than one risk factors appears to be a requirement for the development of VTE.21-23 There are many factors that contribute to the individual baseline risk of VTE. In this, hereditary thrombophilia and first-degree family history both are established risk factors. The baseline risk

17 Introduction

can be further increased by underlying conditions like malignant disease and obesity, advanced age, and by exogenous risk factors such as surgery, trauma, immobilization, and specifically in women, hormonal exposure due to use of combined oral contraceptives (COCs) hormone replacement therapy, hormonal treatments in assisted reproduction techniques (in vitro fertilization), and the pregnancy-postpartum period.

Risk of venous thromboembolism during combined oral contraceptive use

Since their introduction in 1960, the first case reports appeared in public literature suggesting that COCs are associated with an increased risk of VTE.24,25 Despite an up to 10-fold reduction in dose of both components and the development of new progestogens, this risk, albeit lower, remained.26

Up to 1995, the risk of VTE was assumed attributable to ethinylestradiol (EE), because this component is identical in all COCs, whereas the type of synthetic progestogen included could differ.

Currently, there is consensus that the type of progestogen can influence this estrogen-mediated VTE risk. Such difference in VTE risk became apparent in 1995 with the simultaneous publication of three epidemiological studies in the Lancet, which observed that the use of 3rd-generation COCs (containing the progestogens desogestrel or gestodene) in healthy women and in women with factor V Leiden mutation resulted in a higher VTE risk than the use of 2nd-generation COCs (containing levonorgestrel).26-28 Many other studies followed that further evaluated differences in VTE risk between COCs containing 2nd- and 3rd-generation progestogens.29 In the Netherlands, the results of these studies led to a change in the patient leaflets of all COCs in 1997, by informing that the risk of VTE during use of COCs containing desogestrel- or gestodene-containing COCs (3rd-generation) is 1.5-2 times higher than with levonorgestrel-containing COCs (2nd-generation). Additionally, the Dutch General Practitioners Standard on combined oral contraceptives was changed in 1998 to include a statement that second-generation COCs should be considered first choice.30

Based on the results of the most recent epidemiological studies, lowest risk of VTE is observed during use of COCs containing levonorgestrel and norgestimate

18 Introduction

(both 2nd-generation COCs), or norethisterone (a 1st-generation COC), whereas a higher risk is observed during use of COCs containing desogestrel or gestodene, drospirenone, and cyproterone.32-37 The COC containing cyproterone is approved for hormonal treatment in women with androgen-sensitive acne and hirsutism, but has similar contraceptive properties. Recently, also the transdermal patch containing norelgestromin and ethinylestradiol and the vaginal ring containing etonogestrel and ethinylestradiol were shown to have similarly higher VTE risk.34 The relative risk of VTE of recently introduced COCs containing estradiol instead of ethinylestradiol compared to ethinylestradiol-containing COCs is yet unknown. Nevertheless, the absolute risk of VTE is low: varying between 6-9 cases of VTE per 10.000 years of COC-use in comparison to 2-3 cases of VTE per 10.000 women per year who do not use COCs.33,34 Based on these risk profiles, it is now generally recommended in European guidelines,40 including Dutch guidelines (Nederlands Huisartsen Genoodschap already in 1998)30 to preferentially prescribe 2nd-generation COCs to women who start using a COC for the first time.

Recently, a re-evaluation of the risk of VTE during COC-use and during use of non-oral combined hormonal contraceptives has been performed by the Euro- pean Medicines Agency based on all published and unpublished data currently available, including those studies discussed above.2 On the basis of this review, also the COCs containing ethinylestradiol/ norgestimate or ethinylestradiol/ norethisterone have similar VTE risk as noted in women using ethinylestradiol/ levonorgestrel-containing COCs. This re-evaluation subsequently resulted in adap- ting the information in the patient leaflets regarding the risk of VTE of all COCs available in the EU to include the latest VTE risk estimations. The results of this review of VTE risk in different COCs and non-oral combined hormonal contraceptives are shown in Table 3.

19 Introduction

Table 3. Estimated incidence of VTE risk with different combined hormonal contraceptives

RR against Incidence per Generation levonorge- 10.000 women/ strel-COCs per year of use

Non-pregnant non-user 2-3 Ethinylestradiol/levonorgestrel-containing COC (Microgynon 30©) Referentie 5-7 2nd

Ethinylestradiol/norethisterone-containing COC (Modicon©) 1 5-7 1st

Ethinylestradiol/norgestimate-containing COC (Cilest©) 1 5-7 2nd

Ethinylestradiol/desogestrel-containing COC (Marvelon©) 1.5-2 9-12 3rd Ethinylestradiol/gestodene-containing COC (Femodeen©) 1.5-2 9-12 3rd Ethinylestradiol/drospirenone-containing COC (Yasmin©) 1.5-2 9-12 4th Ethinylestradiol/cyproterone-containing COCs (Diane 35©) 1.5-2 9-12 4th Ethinylestradiol/etonogestrel-containing vaginal ring (NuvaRing©) 1-2 6-12 4th Ethinylestradiol/norelgestromin-containing patch (Evra©) 1-2 6-12 4th Ethinylestradiol/chlormadinone-containing COC (Madinelle©)a Unknown Unknown 4th Estradiol/dienogest-containing-COC (Qlaira©) Unknown Unknown 4th Estradiol/nomegestrol-containing COCs (Zoely©) Unknown Unknown ‘5th’

RR: relative risk. a: not available in the Netherlands. Adapted from the ‘Assessment report for combined hormonal contraceptives containing medicinal products’ 16 January 2014, EMA/739865/20132

Risk of VTE also changes with duration of COC-use: VTE risk is highest in the first year following initiation of COC-use and when restarting after a break of at least one month. The risk then reduces over the next year and remains stable thereafter.33,34,38-40 Although the risk is high in the first year of COC-use and then falls, it remains higher than in non-users.

Further, risk of VTE is estrogen-dose dependent; Current low dose (30 microgram ethinylestradiol) has lower risk than COCs containing 50 micrograms or more. Limited data suggest risk of VTE to be slightly lower with COCs containing 20 microgram ethinylestradiol than with COCs containing 30 microgram ethinylestradiol.33

20 Introduction

“Estrogenicity” of combined oral contraceptives

Based on the differences in VTE risk with COCs that have the same estrogen dose but different progestogen type, the prothrombotic effect of COCs is considered not only influenced by the estrogen compound, but is based on the “total estrogenicity” of the specific COC.41 A higher estrogen dose will increase the “total estrogenicity”, but the net estrogenic effect of a COC is also dependent on the anti- estrogenic properties of the synthetic progestogen type. It is assumed that desogestrel- and gestodene-containg COCs (3rd-generation progestogens), but also the more recent COCs containing drospirenone, cyproterone, etonogestrel, norelgestromin (4th-generation COCs), have no or weaker anti-estrogenic activity than levonorgestrel-containing COCs. It is therefore postulated that because these COCs have less or no antagonistic effects on the prothrombotic effect of the estrogen component, the net thrombotic effects will be higher and therefore could result in a higher risk of VTE.

Sex Hormone Binding Globulin (SHBG), produced in the liver, dose-dependently increases after oral intake of EE alone, whereas progestogen intake results in a SHBG decrease, which is dependent on dose and type of progestogen. In view of this, it is postulated that the SHBG increases with increasing estrogenicity, and it was suggested to use SHBG as a surrogate for predicting VTE risk.42

In addition, an association between SHBG levels with a thrombin generation- based APC-resistance test was reported.43 However, others did not share this opinion. Although estrogen intake and more particularly a pregnancy increases SHBG, SHBG is not involved in the clotting cascade, and various metabolic systems also affect the SHBG level.43 However, SHBG, thrombin-based APC resistance tests (APCsr), or any other postulated tests or coagulation factors are as yet not validated as a surrogate endpoint for the clinical outcome of VTE. Some studies have evaluated whether the level of APCsr could be used to explain risk differences observed between users of 2nd or 3rd generation COCs. One study found no association.44 In another study, the large range in APCsr values observed in the plasma samples of women and the small sample sizes resulted in large confidence intervals for the estimated risk. This precluded a meaningful analysis in the female population with respect to the question of whether venous thrombotic risk gradually increased with increasing APCsr.45 This means that it has yet to be demonstrated that such alterations are related to an increase in clinical risk of VTE among COC users, and in particular differences in risks between COCs.

21 Introduction

Biological background of risk of venous thromboembolism during combined oral contraceptive use

The biological background of this association between the use of COCs and the rare adverse event of VTE is complex. The ‘estrogenic’ effect of COCs falls into the category of modifying the blood composition, as its use induces changes in hemostatic balance into a prothrombotic direction by influencing coagulant- and anticoagulant factors and fibrinolysis in Table 4. Results are included which were observed during two months treatment of a 2nd-generation and a 3rd generation COC, with a washout of 2 months between treatments.46

Table 4. Changes in coagulation-, anticoagulation-, and fibrinolytic factors

nd rd 2 generation COCs 3 generation COCs Coagulation factorsa Prothrombin ↑ ↑* Fibrinogen ↑ ↑ Factor V ↓ ↓* Factor VII ↑ ↑* Factor VIII ↑↓ ↑ Factor X ↑ ↑

Anticoagulant systemb Antithrombin ↑↓ ↓

α2-macroglobulin ↑ ↑

α1-antitrypsin ↑ ↑ Protein C inhibitor ↑ ↑ Protein C ↑ ↑*

Protein Stotal and free ↓ ↑* APCsr (aPTT-based assay) ↑↓ ↓* APCsr (EPT-based assay) ↑

Fibrinolytic systemc tPA, antigen, ng/ml ↓ ↓ tPA activity U/ml ↑ ↑ PAI-1 Ag, ng/ml ↓ ↓ PAI-1, act U/ml ↓ ↓ Plasminogen, activity, % ↑ ↑ TAFI Ag, % ↑ ↑*

Significant differences between users of second (2nd gen) and third (3rd gen) generation oral contraceptives (p<0.05). APCsr: activated protein C sensitivity ratio; aPTT: activated partial prothrombin time; ETP: endogenous thrombin potential; PAI: plasminogen activator-inhibitor; TAFI: thrombin-activable fibrinolysis inhibitor; tPA: tissue-type plasminogen activator; Ag: antigen; act: activity; SD: standard deviation. a,47b,48 c,49

22 Introduction

The observed changes are more pronounced during use of 3rd generation COCs than noted during use of 2nd generation COCs, see table above.46-49 It is expected that changes induced by 4th generation COCs will follow the same pattern as noted of 3rd generation COCs. However, in healthy women, these changes in hemostatic variables remain within the normal ranges.21 As stated previously, up to now there are no hemostatic tests that can be used as a validated surrogate for the clinical endpoint of VTE.

Risk of venous thromboembolism during pregnancy and post-partum period

Risk of VTE is substantially higher during pregnancy, especially in the post-partum period, than observed during the use of COCs. In comparison to non-pregnant women, the VTE risk during pregnancy is increased about 9-fold, and up to 50-fold during the postpartum period.50 The main reason of the increased risk of VTE in pregnancy is the hypercoagulable state. This increased coagulation tendency is a physiological adaptation in order to protect the woman at time of delivery from bleeding complications at the placental site of the uterus after placental separation. Like COC-use, a pregnancy is accompanied by changes in sex hormone levels. In comparison to the non-pregnant state, pronounced hemostatic changes are noted, which are presumed due to the huge increase in endogenous estrogen levels already starting from the first trimester in pregnancy; the production of estradiol and estrone is increased 100-fold, and that of estriol a 1000-fold. An accelerated increase of both estradiol and estriol is noted at week 36 of gestation, while at the same time the progesterone level decreases. An increase of coagulation factors, a slight decrease in natural anticoagulants and a decreased fibrinolytic activity are observed. The most pronounced hemostatic changes towards hypercoagulability are noted in the late third trimester, presumably preparing for the upcoming delivery. Initiated by delivery and uterine contractions, activation of blood coagulation occurs, as after placental separation about more than 100 spiral arteries torn from the uterus wall must thrombose. After delivery, fibrinolytic activity rapidly returns to baseline values, while blood coagulation factors mainly normalize during the first 4 to 6 weeks postpartum, although the level of the natural anticoagulant protein S (free fraction) can be decreased for as long as 8 weeks postpartum. Although the increased risk of VTE especially during the postpartum period is firmly established, the exact mechanism of raising sex hormone levels, accompanied with prothrombotic alterations in hemostasis, and the time to return to normal hemostasis is as yet not fully understood.51,52

23 Introduction

Thrombophilia

Inherited thrombophilias are coagulation disorders, leading to a genetic tendency to develop VTE. There are several relevant hereditary thrombophilias identified, which can be divided into rare but severe thrombophilias consisting of a deficiency of the natural anticoagulants protein C, protein S and antithrombin,53-55 of which the prevalence in the general population is about 0.1% each.56-58 These deficiencies increase the risk of VTE more than 10-fold. The absolute risk of VTE in antithrombin-, protein C–, or protein S–deficient persons ranges from 0.7 to 1.7 per 100 person-years. However, these absolute risks are estimated in thrombophilic cohort studies only,59-62 which risks cannot be extrapolated directly to the general population. However, though based on a very low prevalence, some population based case-control studies showed that antithrombin and protein C levels in the deficient range clearly increased risk of VTE 5- and 7-fold.63 However, protein S levels in the deficient range were not associated with an increased VTE risk, except at extremely low level.64

Additionally, mild but more frequent hereditary thrombophilic defects are a mutation in the coagulant factor V Leiden, preventing its inactivation by activated anticoagulant protein C65 and a mutation in prothrombin G20210A, which leads to increased prothrombin, the precursor of thrombin.66 These defects have a prevalence of 5%67 and 2%,68 respectively, in the general Dutch population. Factor V Leiden mutation is the most common thrombophilic defect in the Caucasian population, but far less common or even rare in populations of other ethnic origin. Prothrombin mutation is only prevalent in the Caucasian population.22 In carriers of these mutations, the risk reported in case-control studies is increased about 2-5 fold,66,70 whereas homozygote carriers and persons with double heterozygosity for factor V Leiden and prothrombin mutation have further increased VTE risk.71,72 The absolute risk of VTE in factor V Leiden carriers is estimated being 0.15 per 100 person-years.61,69 Nevertheless, in about 50% of families suffering from VTE, no hereditary thrombophilic defects could be identified. It is possible that some further genetic defects will be identified through genome wide association studies, but if so, such defects are expected to have low risk.

24 Introduction

Hereditary thrombophilia and risk of venous thromboembolism during combined oral contraceptive use

The risk of VTE associated with the use of COCs will increase when the woman’s baseline risk is altered. One of the strongest factors that increase baseline risk is the presence of hereditary thrombophilia. Information on the risk of VTE in COC- users who have severe thrombophilic defects (antithrombin-, protein C, or protein S-deficiency) is very limited,60,73 and only evaluated in thrombophilic cohort studies due to the rarity of these thrombophilic defects. In these families, the risk of VTE in COC-users with severe thrombophilias risk increased risk about 10-fold versus non-thrombophilic COC-users; the observed increased absolute risk varied between 4.3 to 27.5 per 100 pill-years of use.

As to the risk of VTE in COC-users who have mild thrombophilic defects, i.e. factor V Leiden or prothrombin 20210A mutation, a large number of case-control studies have evaluated the risk of VTE in COC-users with factor V Leiden and prothrombin-G20210A, which reported odds ratios varying between a 2 and 26-fold increased risk in comparison to matched control groups.29,74-90 Few subgroup-analyses in family cohort studies presented the absolute risk of VTE in COC-users with factor V Leiden or prothrombin mutation, in which the incidence rate varied between 0.48 and 2.0 per 100 pill-years of use.60,73,91,92

Family history and risk of venous thromboembolism during combined hormonal contraceptive use

Several studies have explored the impact of positive family history and currently it is considered an independent risk factor of VTE, regardless of the presence of other risk factors, including hereditary thrombophilia.93 Several case-control studies reported family history to increase the risk of VTE two to three-fold.93-95 Additionally, two studies reported the VTE risk of a positive family history as higher in first-degree female relatives during fertile age than in male relatives of that age.96,97

However, only one study has considered whether also specific characteristics of patients (probands) with VTE could influence the risk of VTE in their relatives.98 Additionally, it was taken into account whether the proband had experienced a

25 Introduction

provoked or unprovoked VTE. Results of this study indicated that relatives of younger patients (<45 years) with either provoked or unprovoked VTE, have much higher risk for VTE than relatives of older patients. An interesting follow-up question would be whether these family-conferred risks could also be hormone- or gender specific. In the study described earlier, gender of the patient with VTE was reported as having no influence on the risk of VTE in their first-degree relatives (OR: 0.96 (95% confidence interval: 0.67-1.38).98 However, no study has yet evaluated whether a female patient with a hormonally-related VTE could be an additional predictor of (hormonally-related) VTE risk in her first-degree female relatives.

Recurrence rate in women with venous thromboembolism associated with combined oral contraceptive use

Several comparative studies evaluated the risk of VTE recurrence in women with COC-associated VTE, in which the recurrence risk varied between 1.8% to 5.6%.99- 103 However, often these women were included as a subgroup, or COC-exposure was combined with exposure to pregnancy and/or hormone replacement, or studies were restricted to women having COC-use as single risk factor.101,103 An additional difficulty is that in several studies not all women discontinued COC-use or categorized COC-associated VTE as unprovoked. The recurrence rate in women with COC-associated VTE who post-VTE have discontinued further COC-use is therefore not firmly established.

Clinical profile of a woman with COC-associated VTE

Up to now, it is not yet fully understood who will develop a COC-associated VTE. Hereditary thrombophilia and first-degree family history, even in the absence of thrombophilia increases the risk of VTE, but studies evaluating the interaction between family history and COC-use are contradictory.104,105 Concomitant presence of other risk factors besides COC-use additionally increases the risk of VTE, like surgery, trauma, immobilization, postpartum period, malignant disease, and obesity. Many of these risk factors are frequently present in COC-users, but only few women actually develop a COC-associated VTE.

26 Introduction

Outline of this thesis

Central theme of this thesis is the association between COC-use and VTE. In this context, the contribution of other risk factors of VTE, including thrombophilia and family history, have been evaluated. Further, comparisons were made against the risk of VTE in women during pregnancy; the condition that is to be prevented by COC-use.

Information on the absolute risk of VTE in COC-users who have thrombophilic defects is limited, whereas this risk in women in combination with other defects was hardly considered in any study. In Chapter 2, we present a thrombophilic family cohort study, in which we assessed the absolute and relative risk of VTE during use of COCs in women with or without protein S-, protein C-, or antithrombin-deficiency and additionally assessed the contribution of other known thrombophilic defects. In Chapter 3 we present a thrombophilic cohort study in which we assessed the absolute and relative risk of first VTE during COC- use and pregnancy-postpartum in women with or without heterozygous, double heterozygous, or homozygous factor V Leiden or prothrombin-G20210A mutation. Furthermore, the absolute risk of VTE during COC use in women with or without mild hereditary thrombophilia was put into the perspective of contraceptive failure of COC and alternative contraceptives.

In Chapter 4, we present a systemic review and meta-analysis of studies, which compared the risk of VTE in COC-users with or without hereditary thrombophilia, in which a case presentation is taken as a starting point.

As shown in cohort studies described in chapter 3 and 4, women from thrombophilic families have an increased risk of venous thromboembolism (VTE), which increases further during combined oral contraceptive use (COCs) and pregnancy. It is unknown whether this additional risk differs between relatives of male and female patients with VTE, and also whether it matters if the female patient had a hormonally-related VTE (during COC-use or pregnancy). In order to explore this question, we performed a large retrospective family cohort study, which is presented in Chapter 5. In a retrospective thrombophilic family cohort of 1005 first-degree female relatives of reproductive age, we compared VTE risk in relatives of female versus male patients, and between relatives of female patients with and without hormonally-related VTE.

27 Introduction

Although COC-use increases the risk of VTE, it is yet not fully understood which woman will develop COC-associated VTE and evidence on recurrence risk varies. In Chapter 6 we present a prospective cohort of 125 consecutive Dutch women with COC-associated VTE who are, according to Dutch General Practitioner (GP) guidelines, was preferentially prescribed levonorgestrel-containing COCs since 1998.29 In this study, we describe clinical characteristics and post-VTE contraception choices, and have prospectively assessed VTE recurrence after discontinuation of further COC-use and thromboprophylaxis in subsequent pregnancies.

References

1. WHO scientific group meeting on cardiovascular disease and steroid hormone contraceptives. Weekly Epidemiol Rec 1997;72:361-363.

2. Assessment report for combined hormonal contraceptives containing medicinal products. Procedure under Article 31 of Directive 2001/83/EC. Available at: http://www.ema.europa.eu/docs/en_GB/document_library/Referrals_document/ Combined_hormonal_contraceptives/Recommendation_provided_by_ Pharmacovigilance_Risk_Assessment_Committee/WC500160272.pdf

3. Haberlandt E. Ludwig Haberlandt - A pioneer in hormonal contraception. Wien Klin Wochenschr 2009;121:746-749.

4. Djerassi C. Ludwig Haberlandt - Grandfather of the ’Pill’. Wien Klin Wochen- schr. 2009;121:727-728.

5. The Edinburgh Medical and Surgical Journal. A concise view of the latest important discoveries in medicine, surgery, and pharmacy, volume 55, 1841.

6. Frobemius F. Ludwig Fraenkel: ‘spiritus rector’ of the early progesterone research. Eur J Obstet Gyn Reprod Biol 1999;83:115-119.

7. Morgan TS and Davidson SG. The action of the corpus luteum hormone on the human menstrual cycle. Lancet 1937;April 10:861-64.

8. Innhoffen HH, Hohlweg W. Naturwissenschaften 1938;36(6):96.

9. Bates RW. Pincus G. Recent Prog Horm Res 1968;4:V-VI.

28 Introduction

10. Fritz MA and Speroff L. Clinical gynecologic endocrinology and infertility. 2011; 8th Edition Lippincott, Williams & Wilkins, Philadelphia.

11. Rock J, Pincus G. Garcia CR. Effects of certain 19-nor- steroids on the normal human menstrual cycle. Science 1956;124:891-893.

12. Pincus G, Garcia CR, Rock J, Paniagua M, Pendleton A, Laraque F, Nicolas R, Borno R, Pean V. Effectiveness of a contraceptive; effects of a progestin- estrogen combination upon fertility, menstrual phenomena, and health. Science 1959;130:81-83.

13. White Junod S. FDA's Approval of the First Oral Contraceptive, Enovid. Avai- lable at: http://www.fda.gov/AboutFDA/WhatWeDo/History/ProductRegulation/Selecti- onsFromFDLIUpdateSeriesonFDAHistory/ucm092009.htm

14. Trussel J. Contraceptive failure in the United States. Contraception 2011;83:397-404.

15. Trussel J. and Portman D. The creeping pearl: why has the rate of contraceptive failure increased in clinical trials of combined hormonal contraceptive pills? Contraception 2013;88:604-610.

16. Goldin, C, and Katz LF. The Power of the Pill: Oral Contraceptives and Women's Career and Marriage Decisions". J Political Economy 2002;110: 730-770.

17. Wijnberg R. Voorbehoedmiddelen tegen zwangerschap. Ned Tijdschr Geneeskd 1949;93:4242-4245.

18. Kouwenaar W. De taak van het Tijdschrift. Ned Tijdschr Geneeskd 1950;94:290.

19. Virchow R. Die verstopfung der Lungenarterie und ihre Folgen. Beitr Exper Path Physiol 1846;2(1).

20. Virchow R. Phlogose und Thrombose im GeFäβsystem. Gesammelte Abhandlungen Zur Wissenschaftlichen Medizin Frankfurt-am-Main: Von Meidinger Sohn & Comp. 1856. P. 458-636.

29 Introduction

21. Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet 1993;353:1167-1173.

22. Rosendaal FR, Reitsma PH. Genetics of venous thrombosis. J Thromb Haemost 2009; 7(S1): 301-304.

23. Rosendaal FR. Risk factors for venous thrombosis: prevalence, risk and interaction. Semin Hematol 1997;34:171-187.

24. Jordan W. Pulmonary embolism. Lancet 1961;18:1146-1147.

25. Tyler ET. Oral contraception and venous thrombosis. JAMA 1963;185:131-132.

26. Gerstman BB, Piper JM, Freiman JP, Tomita DK, Kennedy DL, Ferguson WJ, Bennett RC. Oral contraceptive oestrogen and progestin potencies and the incidence of deep venous thromboembolism. Int J Epidemiol 1990;19:931- 936.

27. Effect of different progestagens in low oestrogen oral contraceptives on venous thromboembolic disease: World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Lancet 1995;346:1582-1588.

28. Jick H, Jick S, Gurewich V, Wald Myers M, Vasilakis C. Risk of idiopathic venous thromboembolism in women using oral contraceptives with differing progestagen components. Lancet. 1995;346:1589-1593.

29. Bloemenkamp KW, Rosendaal FR, Helmerhorst FM, Büller HR, Vanden- broucke JP. Enhancement by factor V Leiden mutation of risk of deep-vein thrombosis associated with oral contraceptives containing a third-generation progestagen. Lancet 1995;346:1593-1596.

30. Kemmeren JM, Algra A, Meijers JC, Bouma BN, Grobbee DE. Effect of second- and third-generation oral contraceptives on the protein C system in the absence or presence of the factorVLeiden mutation: a randomized trial. Blood 2004;103:927–933.

30 Introduction

31. Sips AJBI, Beijderwellen L, Kardolus GJ, Burgers JS. Dutch General Practitioner’s Standard on hormonal contraception, first revision (Nederlands Huisartsen Genootschap-Standaard Hormonale Anticonceptie (eerste revisie). Huisarts Wet 1998;41:385-394.

32. van Hylckama V, Helmerhorst FM, Vandenbroucke JP, Doggen CJ, Rosen- daal FR. The venous thrombotic risk of oral contraceptives, effects of oestrogen dose and progestogen type: results of the MEGA case-control study. BMJ 2009;339:b2921.

33. Lidegaard Ø, Løkkegaard E, Svendsen AL, Agger C. Hormonal contraception and risk of venous thromboembolism: national follow-up study. BMJ 2009;33:b2890.

34. Lidegaard Ø, Nielsen LH, Skovlund CW, Løkkegaard E. Venous thrombosis in users of non-oral hormonal contraception: follow-up study, Denmark 2001-10. BMJ 2012;344:e2990.

35. Stegeman BH, de Bastos M, Rosendaal FR, van Hylckama Vlieg A, Helmer- horst FM, Stijnen T, Dekkers OM. Different combined oral contraceptives and the risk of venous thrombosis: systematic review and network meta- analysis. BMJ 2013;347:f5298.

36. de Bastos M, Stegeman BH, Rosendaal FR, Van Hylckama Vlieg A, Helmer- horst FM, Stijnen T, Dekkers OM. Combined oral contraceptives: venous thrombosis. Cochrane Database of Systematic Reviews 2014, Issue 3. Art. No.: CD010813. DOI:10.1002/14651858.CD010813.pub2.

37. Vinogradova Y, Coupland C, Hippisley-Cox J. Use of combined oral contraceptives and risk of venous thromboembolism: nested case-control studies using the QResearch and CPRD databases. BMJ 2015;350:h2135.

38. Dinger JC, Heinemann LAJ, and Kuhl-Habichl, D. The safety of drospirenone- containing oral contraceptive: final results from the European Active Surveil lance study on Oral Contraceptives based on 142,475 women-years of obser vation. Contraception 2007;75:344-354.

31 Introduction

39. Sidney S, Cheethamb TC, Connell FA, Ouellet-Hellstrom R, Ouellet- Hellstrom R, Graham DJ, Davis D, Sorela M, Quesenberry CP, Cooper WO. Recent combined hormonal contraceptives (CHCs) and the risk of thromboembolism and other cardiovascular events in new users. Contraception 2013;87:93-100.

40. RCOGs’ Faculty of sexual & reproductive health care (2014) statement on Venous thromboembolism and hormonal contraception, November 2014. Available at: http://www.fsrh.org/pdfs/FSRHStatementVTEandHormonalContraception.pdf

41. Odlind V, Milsom I, Persson I, Victor A. Can changes in sex hormone binding globulin predict the risk of venous thromboembolism with combined oral contraceptive pills? Acta Obstet Gynecol Scand 2002;81:482-490.

42. van Vliet HAAM, Frolich M, Thomassen MCLGD, Doggen CJ, Rosendaal FR, Rosing J, Helmerhorst FM. Association between sex hormone-binding globulin levels and activated protein C resistance in explaining the risk of thrombosis in users of oral contraceptives containing different progestogens. Hum Reprod 2005;20:563–568.

43. Stanczyk FZ, Grimes DA. Sex hormone-binding globulin: not a surrogate marker for venous thromboembolism in women using oral contraceptives. Contraception 2008;78:201-203.

44. Heinemann LAJ, Assmann A, Spannagl M, Schramm W, Dick A, Kluft C, de Maat MPM. Normalized Activated Protein C Ratio itself not associated with increased risk of venous thromboembolism. Contraception 1998;58:321-322.

45. Tans G, van Hylckama Vlieg A, Christella M, Thomassen MLGD, Curvers J, Bertina RM, Rosing J, Rosendaal FR. Activated protein C resistance determined with a thrombin generation-based test predicts for venous thrombosis in men and women. Br J Hematol 2003;122:465-470.

46. Tchaikovski S, Rosing J. Mechanisms of Estrogen-Induced Venous Throm- boembolism. Thromb Res 2010;126:5-11.

32 Introduction

47. Middeldorp S, Meijers JC, van den Ende AE, van Enk A, Bouma BN, Tans G, Rosing J, Prins MH, Büller HR. Effects on coagulation of levonorgestrel- and desogestrel-containing low dose oral contraceptives: a cross-over study. Thromb Haemost 2000;84:4-8.

48. Tans G, Curvers J, Middeldorp S, Thomassen MC, Meijers JC, Prins MH, Bouma BN, Büller HR, Rosing J. A randomized cross-over study on the effects of levonorgestrel- and desogestrel-containing oral contraceptives on the anticoagulant pathways. Thromb Haemost 2000;84:15-21.

49. Meijers JC, Middeldorp S, Tekelenburg W, van den Ende AE, Tans G, Prins MH, Rosing J, Büller HR, Bouma BN. Increased fibrinolytic activity during use of oral contraceptives is counteracted by an enhanced factor XI independent down regulation of fibrinolysis: a randomized cross-over study of two low-dose oral contraceptives. Thromb Haemost 2000;84:9-14.

50. Heit JA, Kobbervig CE, James AH, Petterson TM, MS; Bailey KR, Melton III LJ. Trends in the incidence of venous thromboembolism during pregnancy or postpartum: a 30-year population-based study. Ann Intern Med. 2005;143:697-706.

51. Hellgren M. Haemostasis during normal pregnancy and puerperium. Semi- nars in thrombosis and haemostasis 2003;26:126-130.

52. James AH. Venous Thromboembolism in Pregnancy. Arterioscler Thromb Vasc Biol 2009;29:326-331.

53. Egeberg O. Inherited antithrombin deficiency causing thrombophilia. Thromb Diath Haemorrh 1965;13:516.

54. Griffin JH, Evatt B, Zimmerman TS, Kleiss AJ, Wideman C. Deficiency of protein C in congenital thrombotic disease. J Clin Invest 1981;68:1370-1373.

55. Schwarz HP, Fischer M, Hopmeier P, Batard MA, Griffin JH. Plasma protein S deficiency in familial thrombotic disease. Blood 1984;64:1297-1300.

56. Tait RC, Walker ID, Perry DJ, Islam SI, Daly ME, McCall F, Conkie JA, Carrell RW. Prevalence of antithrombin deficiency in the healthy population. Br J Haematol 1994;87:106-112.

33 Introduction

57. Tait RC, Walker ID, Reitsma PH, Islam SI, McCall F, Poort SR, Conkie JA, Bertina RM. Prevalence of protein C deficiency in the healthy population. Thromb Haemost 1995;73:87-93.

58. Dykes AC, Walker ID, McMahon AD, Islam SI, Tait RC. A study of Protein S antigen levels in 3788 healthy volunteers: influence of age, sex and hormone use, and estimate for prevalence of deficiency state. Br J Haematol 2001;113:636.

59. Martinelli I, Mannucci PM, De Stefano V, Taioli E, Rossi V, Crosti F, Pacia- roni K, Leone G, Faioni EM,. Different risks of thrombosis in four coagulation defects associated with inherited thrombophilia: a study of 150 families. Blood 1998;92:2353-2358.

60. Simioni P, Sanson BJ, Prandoni P, Tormene D, Friederich PW, Girolami B, Prins MH. Incidence of venous thromboembolism in families with inherited thrombophilia. Thromb Haemost 1999;81:198-202.

61. Vossen CY, Conard J, Fontcuberta J, Makris M, van der Meer FJM, Pabinger I, Palareti G, Preston FE, Scharrer I, Souto JC, Svensson P, Walker ID, Rosen- daal FR. Familial thrombophilia and lifetime risk of venous thrombosis. J Thromb Haemost 2004; 2:1526-1532.

62. Brouwer JLP, Veeger NJGM, Kluin-Nelemans HC, van der Meer J. Multicau- sal pathogenesis of venous thromboembolism: evidence from a study in families with hereditary deficiencies of protein C, protein S, or antithrombin. Ann Intern Med. 2006;145:807-815.

63. Koster T, Rosendaal FR, Briet E, van der Meer FJM, Colly LP, Trienekens PH, Poort SR. Pieter H. Reitsma PH, Vandenbroucke JP. Protein C Deficiency in a Controlled Series of Unselected Outpatients: An Infrequent But Clear Risk Factor for Venous Thrombosis (Leiden Thrombophilia Study). Blood 1995;85:2756-2761.

64. Pintao MC, Ribeiro DD, Bezemer ID, Garcia AA, de Visser MCH, Doggen CJM, Lijfering WM, Reitsma PH, Rosendaal FR. Protein S levels and the risk of venous thrombosis: results from the MEGA case-control study. Blood 2013;122:3210-3219.

34 Introduction

65. Bertina RM, Koeleman BP, Koster T, Rosendaal FR, Dirven RJ, de Ronde H, van der Velden PA, Reitsma PH. Mutation in blood coagulation factor V associated with resistance to activated protein C. Nature 1994;369:64-67.

66. Poort SR, Rosendaal FR, Reitsma PH, Bertina RM. A common genetic varia- tion in the 3’-untranslated region of the prothrombin gene is associated with elevated plasma prothrombin levels and an increase in venous thrombosis. Blood 1996;88:3698-3703.

67. Rees DC, Cox M, Clegg JB. World distribution of factor V Leiden. Lancet 1995;346:1133-1134.

68. Rosendaal FR, Doggen CJM, Zivelin A, Arruda VR, Aiach M, Siskovick DS, Hillarp H, Watzke HH, Bernardi F, Cumming AM, Reitsma PH. Geographic distribution of the 20210 G to A prothrombin variant. Thromb Haemost 1998;79:706-708.

69. Koster T, Rosendaal FR, De Ronde H, Briët E, Vandenbroucke JP, Bertina RM. Venous thrombosis due to a poor anticoagulant response to activated protein C: Leiden Thrombophilia Study. Lancet 1993;342:1503-1506.

70. Middeldorp S, CM, Koopman MM, van Pampus EC, Hamulyák K, van der Meer, Büller HR. The incidence of venous thromboembolism in family members of patients with factor V Leiden mutation and venous thrombosis. Ann Intern Med 1998;128:15-20.

71. Rosendaal FR, Koster T, Vandenbroucke JP, Reitsma PH. High risk of thrombosis in patients homozygous for factor V Leiden (activated protein C resistance). Blood 1995;85:1504-1508.

72. Emmerich J1, Rosendaal FR, Cattaneo M, Margaglione M, De Stefano V, Cumming T, Arruda V, Hillarp A, Reny JL. Combined effect of factor V Leiden and prothrombin 20210A on the risk of venous thromboembolism--pooled analysis of 8 case-control studies including 2310 cases and 3204 controls. Study Group for Pooled-Analysis in Venous Thromboembolism. Thromb Haemost 2001;86:809-816.

73. Pabinger I, Schneider B, and the study GTH group on natural inhibitors. Thrombotic risk of women with hereditary antithrombin III, protein C- and protein S deficiency taking oral contraceptive medication. Thromb Haemost 1994;71:548-552.

35 Introduction

74. de Bruijn SFTM, Stam J, Koopman MMW, Vandenbroucke JP for the Cerebral Venous Sinus Thrombosis Study Group. Case-control study of cerebral sinus thrombosis in oral contraceptive users who are carriers of hereditary prothrombotic conditions. BMJ 1998;316:589-592.

75. Andersen BS, Olsen J, Nielsen GL, Steffensen FH, Sorensen HT, Baech J, Gregersen H. Third generation oral contraceptives and heritable thrombophilia as risk factors of non-fatal venous thromboembolism. Thromb Haemost 1998;79:28-31.

76. Bennet L, Odeberg H. Resistance to activated protein C, highly prevalent amongst users of oral contraceptives with venous thromboembolism. J Intern Med 1998;244:27-32.

77. Martinelli I, Sacchi E, landi G, Taioli E, Duca F, Manucci PM. High risk of cerebral vein thrombosis in carriers of a prothrombin-gene mutation and in users of oral contraceptives. N Engl J Med 1998;338:1793-1797.

78. Martinelli I, Taioli E, Bucciarelli P, Akhavan S, Manucci PM. Interaction between the G20210A mutation of the prothrombin gene and oral contraceptives use in deep vein thrombosis. Arterioscler Thromb Vasc Biol 1999;19:700-703.

79. Spannagl M, Heinemann LA, Schramm W. Are factor V Leiden carriers who use oral contraceptives at extreme risk for venous thromboembolism? Eur J contracept reprod health care. 2000; 5:105-112.

80. Aznar J, Vaya A, Estelles A, MiraY, Segui R, Villa P, Ferrando F, Falco C, Corella D, Espana F. Risk of venous thrombosis in carriers of the prothrombin G20210A variant and factor V Leiden and their interaction with oral contraceptives. Haematologica 2000;85:1271-1276.

81. Bloemenkamp, K.W., Rosendaal, F.R., Helmerhorst, F.M., Vanden broucke, J.P. Higher risk of venous thrombosis during early use of oral contraceptives in women with inherited clotting defects. Arch Intern Med 2000;160:49-52.

82. Legnani C, Palareti G, Guazzaloca G, Cosmi B, Lunghi B, Bernardi F, Coccheri S. Venous thromboembolism in young women: role of throm bophilic mutations and oral contraceptive use. Eur Heart J 2002;23;984-990.

36 Introduction

83. Vaya A, Mira Y, Mateo J, Falco C, Villa P, Estelles A, Aznar J. Prothrombin G20210A mutation and oral contraceptive use increase upper extremity deep vein thrombotic risk. Thromb Haemost. 2003;89:452-457.

84. Martinelli I, Battaglioli T, Buccarelli P, Passamonti SM, Manucci PM. Risk factors and recurrence rate of primary deep vein thrombosis of the upper extremities. Circulation 2004;110:566-570.

85. Ehrenforth S, Nemes L, Mannhalter C, Rosendaal FR, Koder S, Zoghlami-Rintelen C, Pabinger I. Impact of environmental and hereditary risk factors on the clinical manifestation of thrombophilia in homozygous carriers of factor V:1691A. J Thromb Haemost 2004;2:430-436.

86. Sidney S, Petitti DB, Soff GA, Cundiff DL, Tolan KK, Quesenberry CP. Venous thromboembolic disease in users of low-estrogen combined estrogen-progestin oral contraceptives. Contraception 2004;70;3-10.

87. Gadelha T, André C, Jucá AAV, Nucci M. Prothrombin 20210a and oral contraceptive use as risk factors for cerebral venous thrombosis. Cere brovasc Dis 2005;19:49-52.

88. Roach REJ, Lijfering WM, Helmerhorst FM, Cannegieter SC, RosendaalFR, van Hylckama Vlieg. The risk of venous thrombosis in women over 50 years old using oral contraception or postmenopausal hormone therapy. J Thromb Haemost 2013;11:124-131.

89. Tufano A, Guida A, Coppola A, Nardo A, Capua MD, Quintavalle G, Di Minno MND, Cerbone, AM, Di Minno G. Risk factors and recurrent thrombotic episodes in patients with cerebral venous thrombosis. Blood Transfus 2014;12:s337-342.

90. Bergendal A, Persson I, Odeberg J, Sundström A, Holmström M, Schulman S, Björgell O, Kieler H. Association of venous thromboembolism with hormonal contraception and thrombophilic genotypes. Obstet Gynecol 2014;124:600-609.

91. Van Boven HH, Vandenbroucke JP, Briët E, Rosendaal FR. Gene-gene and gene-environment interactions determine the risk of thrombosis in families with inherited antithrombin deficiency. Blood 1999;94:2590-2594.

37 Introduction

92. Santamaria A, Mateo J, Oliver A, Menendez B, Souto J, Borrel M, Fontcu berta J. Risk of thrombosis associated with oral contraceptives of women from 97 families with inherited thrombophilia: high risk of thrombosis in carriers of the G20210A mutation of the thrombin gene. Haematologica 2001;86:965-971.

93. Bezemer ID, van der Meer FJM, Eikenboom JCJ, Rosendaal FR, Doggen CJM. The value of family history as a risk indicator for venous thrombosis. Arch Intern Med 2009;169:610-615.

94. Noboa S, Le Gal, Lacut K, Mercier B, Leroyer C, Nowak E, Mottier D, Oger E: for the EDITH Collaborative Study Group. Family history as a risk factor for venous thromboembolism. Thromb Res 2008;122:624-629.

95. Dowling NF, Austin H, Diley A, Whitsett C, Evatt BL, Hooper WC. The epidemiology of venous thromboembolism in Caucasians and African- Americans: the GATE Study. J Thromb Haemost 2003;1:80-87.

96. Silverstein MD, Heit JA, Mohr DN, Petterson TM, O'Fallon WM, Melton III LJ. Trends in incidence of deep vein thrombosis and pulmonary embolism. Arch Intern Med 1998;158:585-593.

97. Zöller B, Li X, Sundquist J, Sundquist K. Age- and Gender-Specific Familial Risks for Venous Thromboembolism. A Nationwide Epidemiological Study Based on Hospitalizations in Sweden. Circulation 2011;124:1012-1020.

98. Couturaud, F, Leroyer C, Tromeur C, Julian JA, Kahn SR, Ginsberg JS, Wells PS, Douketis JD, Mottier D, Kearon C. Factors that predict thrombosis in relatives of patients with venous thrombosis. Blood 2014;124:2124-2130.

99. Cushman M, Glynn RJ, Goldhaber SZ, Moll S, Bauer KA, Deitcher S, Shrivastava S, Ridker PM. Hormonal factors and risk of recurrent venous thrombosis: the Prevention of Recurrent Venous Thromboembolism trial. J Thromb Haemost 2006;4:2199-2203.

100. Le Gal G, Kovacs MJ, Carrier M, Do K, Kahn SR, Wells PS, Anderson DA, Chagnon I, Solymoss S, Crowther M, Righini M, Lacut K, White RH, Vickars L, Rodger M. Risk of recurrent venous thromboembolism after a first oestrogen-associated episode. Data from the REVERSE cohort study. Thromb Haemost 2010;104:498-503.

38 Introduction

101. Vaillant-Roussel H, Ouchchane L, Dauphin C, Philippe P, Ruivard M. Risk factors for recurrence of venous thromboembolism associated with the use of oral contraceptives. Contraception 2011;84:e23-30. 102. Le Moigne E, Delluc A, Tromeur C, Nowak E, Mottier D, Lacut K, Le Gal G. Risk of recurrent venous thromboembolism among young women after a first event while exposed to combined oral contraception versus not exposed to: A cohort study. Thromb Res 2013;132:51-55. 103. Eischer L, Eichinger S, Kyrle PA. The risk of recurrence in women with venous thromboembolism while using estrogens: a prospective cohort study. J Thromb Haemost 2014;12:635-640. 104. Cosmi B, Legnani C, Bernardi F, Coccheri S, Palareti G. Value of family history in identifying women at risk of venous thromboembolism during oral contraception: observational study. BMJ 2001;322:1024–1025. 105. Vandenbroucke JP, van der Meer FJM, Helmerhorst FM, Rosendaal FR. Family history and risk of venous thromboembolism with oral contraception. BMJ 2001;323:752.

39 40 Oral contraceptives and the absolute risk of venous thromboembolism in women with single or multiple thrombophilic defects

Results From a Retrospective Family Cohort Study

Elizabeth F.W. van Vlijmen Jan-Leendert P. Brouwer Nic J.G.M. Veeger Tom K.A.B. Eskes† Pieter A. de Graaff Jan van der Meer†

Archives of Internal Medicine 2007;167:282-289 Chapter 2

ABSTRACT

Background: The risk of venous thromboembolism (VTE) in women taking combined oral contraceptives (COCs) is attributed to changes in coagulation and fibrinolysis. Their impact may be greater in women with preexistent thrombophilic defects.

Methods: We assessed the effects of COCs on absolute VTE risk in women with single or multiple thrombophilic defects in a retrospective family cohort study. Female relatives of probands with VTE and hereditary deficiencies of protein S, protein C, or antithrombin were tested for known thrombophilic defects, including the index deficiency. Absolute incidences of VTE were compared in deficient versus non-deficient women, in deficient and non-deficient women who ever or never used COCs, and in deficient and non-deficient women with 0, 1, or more than 1 other thrombophilic defect during exposure to COCs.

Results: Of 222 women, 135 (61%) ever used COCs. Overall, annual incidences of VTE were 1.64% and 0.18% in deficient and non-deficient women, respectively; the adjusted relative risk was 11.9 (95% confidence interval, 3.9-36.2). The risk was comparable in deficient ever and never users (1.73% versus 1.54%). Annual incidences during actual COC use were 4.62% in deficient women and 0.48% in non-deficient women; the relative risk was 9.7(95% confidence interval, 3.0-42.4). The incidence increased by concomitant thrombophilic defects, from 3.49% to 12.00% in deficient women and from 0% to 3.13% in non-deficient women.

Conclusions: Women with hereditary deficiencies of protein S, protein C, or antithrombin are at high risk of VTE during use of COCs, particularly when other thrombophilic defects are present. They have VTE at a younger age, but the overall risk is not increased by COCs.

42 Oral contraceptives, antithrombin-, protein C- or protein S deficiency and risk of VTE

INTRODUCTION

Since their introduction in 1960, combined oral contraceptives (COCs), containing a combination of ethinylestradiol and a progestogen, are associated with an increased risk of venous thromboembolism (VTE).1,2 Despite an up to 10-fold reduction in the dose of both components, this risk, albeit lower, has persisted.3 Up to 1995, the risk of VTE was attributed to ethinylestradiol, because this component is identical in all COCs, while the type of progestogen included may differ. It is the opinion that the type of progestogen may also influence the VTE risk, because the use of third-generation COCs (containing desogestrel or gestodene) in healthy women was reported to result in a higher risk than the use of second generation COCs (containing levonorgestrel).4-6 Nevertheless, because of the low baseline risk of VTE in healthy women of fertile age, ranging from 1 per 100 000 years at the age of 15 years to 10 per 100 000 years at the age of 44 years,7,8 the absolute risk of VTE in COC-using women is low (20-40 per 100 000 pill-years).4-6

The biological background of this association is complex. Combined oral contraceptives induce changes in coagulation and fibrinolysis but, in healthy women, hemostatic variables remain within their normal ranges.9,10 It is likely that these changes may have more impact in women who are already at increased risk of VTE because of preexisting thrombophilic defects. Although several observational studies11-17 reported on the interaction of COCs and various single thrombophilic defects, little is known about the absolute risk of VTE attributable to COCs in women with single or combined thrombophilic defects.16,17 Recently, aggregation of multiple thrombophilic defects in families with hereditary deficiencies of protein S, protein C, or antithrombin was demonstrated, and it was shown that the risk of VTE strongly depended on the presence of 1 or more other thrombophilic defects.18 In the present study, we analyzed the data from this large family cohort to assess the effects of COCs on the absolute risk of VTE in protein S–, protein C–, or antithrombin-deficient women and the contribution of other known thrombophilic defects.

METHODS

Subjects

The original retrospective study18 was performed in a cohort of protein S–, protein

43 Chapter 2

C-, and antithrombin-deficient families. Briefly, probands were consecutive patients with documented VTE, in whom 1 of these deficiencies was demonstrated. First-degree relatives older than 15 years were identified by pedigree analysis. Because the number of antithrombin deficient probands was small, second- degree relatives with a parent who was deficient were also identified. Relatives were enrolled after informed consent was obtained. Information on previous episodes of VTE, exposure to exogenous risk factors for VTE, and anticoagulant treatment was collected by physicians at the thrombosis outpatient clinic of our hospital, using a validated questionnaire,19 and by reviewing medical records. In women, COC use and obstetric history were documented in detail. Blood samples were taken after collection of clinical data. All relatives were tested for additional thrombophilic defects, including a second deficiency of protein S, protein C, or antithrombin; factor V Leiden, or prothrombin G20210A; increased levels of factor VIII, IX, and XI; lupus anticoagulant; and hyperhomocysteinemia. In the present study, we analyzed female relatives of fertile age from the family cohort. Male relatives of comparable age were used as a reference group. The study was approved by the institutional review board of University Medical Center Groningen.

Diagnosis of VTE

Venous thromboembolism was considered established when diagnosed by compression ultrasonography or venography (deep vein thrombosis) or by ventilation and perfusion lung scanning, spiral computed tomographic scanning, or pulmonary angiography (pulmonary embolism); or when the patient had received full-dose heparin and vitamin K antagonists for at least 3 months without objective testing when these techniques were not yet available. Venous thromboembolism was classified as secondary when occurring within 3 months after exposure to exogenous risk factors, including surgery, trauma, immobilization for 7 days or more, COC use, hormone therapy, pregnancy and puerperium, and malignancy. In the absence of these risk factors, VTE was defined as primary. Superficial phlebitis was not counted as a thrombotic event, because it was usually not confirmed by an objective technique and consequently might not be accurately classified in the setting of this retrospective study.

44 Oral contraceptives, antithrombin-, protein C- or protein S deficiency and risk of VTE

Laboratory studies

Protein S and protein C antigen levels were measured by enzyme linked immuno- sorbent assay using reagents from DAKO, Glostrup, Denmark. The activity of protein C (Berichrom Protein C; Dade Behring, Marburg, Germany) and antithrombin (Coatest; Chromogenix, Mölndal, Sweden) was measured by chromogenic substrate assays. Levels of protein S, protein C, and antithrombin were expressed as percentages of the levels measured in pooled plasma set at 100%. Normal ranges were determined in 393 healthy blood donors without a (family) history of VTE, who were neither pregnant nor used COCs in the 3 months before blood sampling. Protein S deficiency type I was defined by total protein S levels below the lower limit of its normal range (67%), and protein S deficiency type III by lowered free protein S levels (65%), but normal total protein S levels. After we had demonstrated that protein S deficiency type III was not a risk factor for thrombosis in these families, we excluded families with this deficiency from the analysis.20 Protein C deficiency types I and II were defined by lowered levels of either protein C antigen (63%) or activity (64%), and antithrombin deficiency by lowered levels of antithrombin activity (74%). Deficiencies were considered inherited when confirmed by measurement of a second sample collected after a 3-month interval and demonstrated in at least 2 family members. Factor V Leiden and prothrombin G20210A were demonstrated by polymerase chain reactions.21,22 Factors VIII:C, IX:C, and XI:C were measured by 1-stage clotting assays (Amelung, Lemgo, Germany) and considered increased at levels greater than 150%. Lupus anticoagulant was demonstrated by abnormal dilute Russell viper venom time and activated partial thromboplastin time or tissue thromboplastin inhibition, normalized by adding phospholipids.23 Fasting and post–methionine-loading levels of homocysteine were measured by high-performance liquid chromatography.24 Hyperhomocysteinemia was defined as a fasting homocysteine level greater than 2.50 mg/L (18.5 µmol/L) and/or a post-loading level greater than 7.95 mg/L (58.8 µmol/L), as described in the Dutch population.25 In probands and symptomatic relatives, blood samples were collected at least 3 months after VTE had occurred. If they were then still treated with a short-acting vitamin K antagonist (acenocoumarol), samples were taken after temporary change of this therapy into subcutaneous nadroparin treatment for at least 2 weeks.

45 Chapter 2

Statistical analysis

We compared the absolute risk of a first episode of VTE in deficient and non- deficient female relatives, who were divided into women who ever or never used COCs during their fertile lifetime. Probands were excluded from analysis to avoid bias. Annual incidences were calculated by dividing the number of first episodes of VTE by the number of person-years. Person-years were counted from the age of 15 years until the age of 50 years, the first episode of VTE, or the end of the study period. A minimum age of 15 years was chosen because VTE is rare at a younger age, and a maximum age of 50 years was chosen as the end of a fertile lifetime. Event-free survival was analyzed by the Kaplan-Meier method. In this analysis, male-deficient relatives were used as a reference group. In addition to the overall risk of VTE during fertile lifetime in ever and never users, we calculated the risk of VTE related to actual COC exposure in deficient and non- deficient women. In this calculation, person-years was defined as the number of pill-years, including a 3-month exposure window after COC use was discontinued. Moreover, we estimated the contribution of concomitant thrombophilic defects in deficient and non-deficient actual COC users. Concomitance was classified as no other thrombophilic defect, 1 other thrombophilic defect, and more than 1 other thrombophilic defect. Crude relative risks (RRs) were based on annual incidences, and 95% confidence intervals (CIs) were calculated using the binomial probability model (conditional small sample approach).26 To account for clustering of women within families, outcome rates were analyzed by random-effects logistic regression with Gaussian distribution, resulting in adjusted RRs and 95% CIs. Continuous variables were expressed as mean values and standard deviation or median values and range, and categorical data as counts and percentages. Diffe- rences between groups were evaluated by the t test or the Mann-Whitney test, depending on the normality of data, for continuous data and by the Fisher exact test for categorical data. A 2-sided P.05 indicated statistical significance. Analyses were performed using SAS statistical software, version 8.2 (SAS Institute Inc, Cary, NC), and Stata software, version 9.1 (Stata Corp, College Station, Tex).

46 Oral contraceptives, antithrombin-, protein C- or protein S deficiency and risk of VTE

39 With Protein S 40 With Protein C 12 With Antithrombin Deficient Probands 91 Total Deficiency Deficiency Deficiency

All Relatives 725 263 277 185

Aged 15 y 92 36 33 23

3 Deceased 136 50 43 43

1 No Consent 30 15 12

Geographic Reasons 11 4 6

Tested Relatives 456 158 183 115 0

Inheritance Not Established 19 15 4

Analyzed Relatives 437 143 179 115

52 Women 222 71 88 63

Men 215 72 91

Figure 1. Recruitment of the cohort of families with hereditary deficiencies of protein S, protein C, or antithrombin.

RESULTS

The original family cohort contained 91 unrelated families from 39 protein S-deficient probands, 40 protein C–deficient probands, and 12 antithrombin- deficient probands, including 263, 277, and 185 relatives, respectively (Figure 1). Seven families (19 relatives) were excluded because inheritance of the deficiency in the probands could not be established by testing their relatives. The mean (SD) age at the time of VTE was 29 (11) years in the 49 female probands and 38 (14) years in the 35 male probands. Of the female probands, 43 (88%) had secondary VTE (14 because of COC use, 6 because of pregnancy, 16 because of puerperium, and 7 because of trauma, surgery, or immobilization), compared with 8 (23%) of male probands (because of trauma, surgery, or immobilization). Overall, 343 relatives were female, of whom 121 could not be enrolled because of prior death (n=50), being 15 years or younger (n=59), no consent (n=9), or geographic distance (n=3). Death was possibly related to VTE in 5 women, of whom 2 died at a fertile age.

47 Chapter 2

The response rate of eligible women was 95%. The remaining 222 women were analyzed. Their characteristics are summarized in Table 1. One hundred one (45%) of these women were deficient. Fifty-six deficient and 79 non-deficient women had ever used COCs. The mean age at the start of COC use was 22 years. Most women (78%) had only 1 period of COC use. The median duration of COC use was 5 years (range, 0.03-27 years). Thirty-four events were reported, of which 31 (91%) were classified as secondary; 16 events were related to COC use, and 11 to pregnancy or puerperium.

Table 1. Characteristics of 222 female and 215 male relatives of 91 probands with protein S, protein C, or antithrombin deficiency*

Data for Women Data for Men Deficient Nondeficient P Deficient Nondeficient P Characteristic (n=101) (n=121) Value (n=106) (n=109) Value Age at enrollment, y 39 (18) 44 (18) .08 40 (17) 41 (16) .09 Follow-up, y† 18 (12) 24 (12) .001 19 (12) 24 (11) .001 Ever users of COCs‡ 56 (55) 79 (65) .17 NA NA NA Age at start of COC use, y 22 (7) 23 (7) .22 NA NA NA Duration of COC use, y 5 (4) 8 (7) .008 NA NA NA VTE†‡ 29 (29) 5 (4) .001 27 (25) 1 (1) .001 Age at time of event, y 27 (6) 27 (9) .73 30 (11) 48 (NA) .22 Primary§ 3 (10) 0 18 (67) 1 (100) Secondary to§ Oral contraception 13 (45) 3 (60) .87 NA NA .99 Pregnancy or puerperium 10 (34) 1 (20) NA NA Major trauma, surgery, or 3 (10) 1 (20) 9 (33) 0 immobilization Other thrombophilic defects Factor V Leiden 13 (90) 15 (110) .84 12 (104) 20 (97) .48 Prothrombin G20210A 6 (89) 9 (109) .42 8 (93) 9 (95) .79 Increased factor VIII 39 (84) 42 (102) .45 44 (90) 27 (89) .02 Increased factor IX 9 (68) 13 (104) .47 9 (67) 18 (93) .11 Increased factor XI 9 (88) 16 (105) .20 14 (93) 5 (94) .05 Hyperhomocysteinemia 16 (83) 19 (96) .69 22 (88) 16 (83) .34 Any other thrombophilic defects 57 (101) 70 (121) .18 60 (106) 54 (109) .41

Abbreviations: COC, combined oral contraceptive; NA, data not applicable; VTE, venous thromboembolism. * Data are given as mean (SD) unless otherwise indicated. †Restricted to those aged 15 to 50 years. ‡Data are given as number (percentage) of the total for each group. §Data are given as number (percentage) of VTE total for each group. Percentages may not total 100 because of rounding. Data are given as percentage (number tested). The prevalences in the healthy population are as follows: factor V Leiden, 5%; prothrombin G20210A, 2%; increased levels of factor VIII (25%), factor IX (10%), and factor XI (10%); and hyperhomocysteinemia, 10%.

Overall absolute risk of VTE during reproductive age

Overall, the annual incidence of VTE was 1.64% versus 0.18% in deficient versus non-deficient women; the adjusted RR was 11.9 (Table 2). The highest risk of VTE was observed in antithrombin-deficient women (2.06%), compared with 1.89% in protein C–deficient women and 1.01% in protein S–deficient women. Antithrombin non-deficient women had the lowest risk (0%); the risk was 0.17% and 0.31% in protein C and protein S non-deficient women, respectively.

48 Oral contraceptives, antithrombin-, protein C- or protein S deficiency and risk of VTE

The annual incidences were 1.73% versus 0.16% in deficient versus non-deficient ever users of COCs, and 1.54% versus 0.21% in deficient versus non-deficient never users. There was no significant difference between deficient ever and never users (adjusted RR, 1.3; 95% CI, 0.5-3.8). Fifty-five percent of ever users and 58% of never users experienced 1 or more pregnancies. In deficient ever users, 80% of VTE episodes were related to use of COCs, whereas in deficient never users, 67% were related to pregnancy.

Table 2. Risk of VTE in female relatives of probands with deficiencies of protein S, protein C, or antithrombin, who ever or never used COCs, and during actual exposure to COCs

Event free survival analysis showed that there were no differences between deficient ever users, deficient never users, and deficient men at the age of 50 years. First events occurred earlier in deficient ever users and, although less pronounced, in deficient never users than in deficient men (Figure 2).

49 Chapter 2

100 95 90 85 80 vival, % r 75 70

65 Group 1 Event-Free Su 60 Group 2 Group 3 55 Group 4 Group 5 50

0 15 20 25 30 35 40 45 50 Age, y Sample Size Group 1 114 83 66 38 19 Group 2 56 51 28 18 11 Group 3 45 36 24 14 11 Group 4 42 31 28 24 16 Group 5 79 75 61 47 25

Figure 2. Event-free survival in protein S–, protein C–, and antithrombin-deficient and nondeficient female relatives who ever or never used combined oral contraceptives and in deficient male relatives. The sample sizes represent women at risk. Group 1 indicates deficient men; 2, deficient ever users; 3, deficient never users; 4, nondeficient never users; and 5, nondeficient ever users.

Absolute risk of VTE during exposure to COCs

The annual incidence of VTE during actual COC use was 4.62% in deficient COC users, compared with 0.48% in non-deficient COC users (Table 2). The annual incidences were 2.42%, 7.06%, and 5.14% in protein S–, protein C–, and antithrombin-deficient women, respectively, compared with 1.01%, 0.29%, and 0% in non-deficient COC users, respectively (Table 2). Details of the 16 women who experienced a first episode of VTE while using COCs are presented in Table 3. Thirteen women (81%) were protein S, protein C, or antithrombin deficient. Concomitance of 1 or more other thrombophilic defects was demonstrated in 10 (77%) of the 13 deficient women, and in all 3 non-deficient women. Thirteen women (81%) were first-ever users of COCs. The mean age at onset of VTE was 24 years. Of the women, 69% experienced their first event within 6 months of the start of COC use.

50 Chapter 2 Oral contraceptives, antithrombin-, protein C- or protein S deficiency and risk of VTE

Table 3. Characteristics of female relatives with or without protein S, protein C, and antithrombin deficiency, who developed VTE during actual exposure to COCs

Age When Starting to Previous COC COC Type Exposure Age at Receive Starting (Type/Exposure, Used at Until VTE, First VTE, Type of COC, y Year mo)* Onset of VTE * mo y VTE Deficiency Concomitant Thrombophilic Defects 24 1984 No Unknow 1 24 PE Antithrombin None 27 1986 No First 6 27 DVT Protein C Factor V Leiden 16 1974 No Second 3 16 DVT plus PE Antithrombin None 31 1975 No Second 3 32 DVT Protein S Increased factor VIII 17 1978 No Second 6 17 DVT plus PE Nondeficient Increased factor VIII and factor XI; factor V Leiden; prothrombin G20210A 17 1980 No Second 120 26 DVT Protein S Increased factor VIII 17 1986 No Third 1 17 DVT Protein C Increased factor VIII 24 1982 No Third 3 25 DVT plus PE Protein C Increased factor VIII and factor XI 20 1990 No Third 3 20 DVT Protein C Increased factor VIII; factor V Leiden 25 1993 No Third 6 25 CST Nondeficient Factor V Leiden 16 1991 No Third 44 20 DVT Nondeficient Factor V Leiden; prothrombin G20210A 15 1989 No Third 56 23 PE Protein S Factor V Leiden 15 1983 No Third 60 20 DVT Protein C Factor V Leiden; prothrombin G20210A 15 1982 Yes (third/132) Third 1 31 DVT plus PE Antithrombin Increased factor VIII 18 1987 Yes (second/48; Second 1 28 PE Protein C None third/24; third/48) 15 1980 Yes (third/48) Third 62 27 DVT Protein C High factor VIII

Abbreviations: COC, combined oral contraceptive; CST, cerebral sinus thrombosis; DVT, deep venous thrombosis; PE, pulmonary embolism; VTE, venous thromboembolism. * The COCs were categorized into first-generation agents (containing norethisterone or lynestrenol), second-generation agents (containing levonorgestrel), and third-generation agents (containing desogestrel, gestodene, norgestimate, cyproterone acetate, or drospirenone).

Exposure to third-generation COCs was associated with a higher risk of VTE than exposure to second generation COCs. Overall, annual incidences were 5.43% versus 1.19%; the RR was 4.5 (95% CI, 1.4-15.7). In deficient women, these were 6.85% and 4.35%, respectively (RR, 1.6; 95% CI, 0.4-6.6); in non-deficient women, 3.57% and 0.31%, respectively (RR, 11.7; 95% CI, 0.9-344.4). However, risk estimates might not be accurate because of the inability to retrieve the brands used in 19 women.

Table 4. Risk of VTE during actual COC exposure in female relatives with or without protein S, protein C, or antithrombin deficiency, relative to the number of concomitant thrombophilic defects*

No. of Concomitant Thrombophilic Defects

Variable 0 1 >1

Deficient Group Total No. 21 21 7 No. of women with event 3 7 3 Sum of pill-years 86 126 25 Annual incidents, % 3.94 (0.7-10.2) 5.56 (2.2-11.5) 12.00 (2.5-35.1)

Nondeficient Group Total No. 21 39 12 No. of women with event 0 1 2 Sum of pill-years 161 319 64 Annual incidents, % 0 (0-2.3) 0.31 (0.01-1.8) 3.13 (0.4-11.3)

Abbreviations: See Table 2. *The relative risk (95% CI) for deficient vs nondeficient women with 1 concomitant thrombophilic defect was 17.7 (2.7-401.9); and for women with more than 1 concomitant thrombophilic defect, 3.8 (0.6-32.3).

51 Chapter 2

Concomitant thrombophilic defects and the absolute risk of VTE

Concomitance of other thrombophilic defects was demonstrated in 57% of deficient women and in 70% of non-deficient women (Table 1). Of these women, a second deficiency of protein S, protein C, or antithrombin, and lupus anticoagulant was not found. The remaining concomitant defects were equally distributed among deficient and non-deficient women. The annual incidences of VTE in deficient actual COC users with 0, 1, or more than 1 concomitant defects were 3.49%, 5.56%, and 12.00%, respectively, compared with 0%, 0.31%, and 3.13% in non-deficient actual COC users, respectively (Table 4).

DISCUSSION

This study shows a high absolute risk of VTE in protein S–, protein C–, and antithrombin-deficient women, compared with their non-deficient female relatives. Combined oral contraceptives hardly contributed to the overall risk during fertile lifetime. Cumulative event rates at the age of 50 years were similar in women, either ever or never users of COCs, and in men. However, event free survival curves showed VTE to occur earlier in deficient ever users and, although less pronounced, in deficient never users than in deficient men. These differences between women and men are most likely because of COC use in deficient ever users and because of pregnancy in deficient never users. Therefore, COCs and pregnancy resulted in the occurrence of VTE in deficient women at a younger age, but did not increase the overall risk of VTE during fertile lifetime. The absolute risk of VTE in deficient women during actual COC use was almost 10-fold higher than in non-deficient COC users (4.62% versus 0.48%). This risk is 115- to 230-fold higher than the published absolute risk in healthy women of fertile age during actual exposure to COCs (0.02%-0.04%).4-6 The 12- to 24-fold higher risk in non-deficient COC users is explained by the presence of other thrombophilic defects (described later).

Several studies have evaluated the contribution of COC use to the risk of VTE in women with thrombophilic defects. Most studies were addressed to factor V Leiden or prothrombin G20210A, both more prevalent but milder risk factors for VTE than the previously mentioned deficiencies.11-15 Only 2 studies16,17 reported on the absolute risk of VTE in protein S–, protein C–, and antithrombin-deficient women.

52 Chapter 2 Oral contraceptives, antithrombin-, protein C- or protein S deficiency and risk of VTE

The first, a retrospective case-control study,16 compared deficient females who used COCs with those who did not use COCs. The incidence of VTE was increased only in antithrombin-deficient COC users. A comparison with our results is hampered by differences in design because this study also included probands, who by definition had experienced VTE, and patients with prior VTE, whereas superficial phlebitis was counted as an event. The second, a family cohort study,17 reported incidences of 4.3% and 0.7% per year in deficient and non-deficient female relatives who used COCs, respectively, in line with our results. We demonstrated 1 or more other thrombophilic defects in 57% of deficient and in 70% of non- deficient relatives. These prevalences were higher than is expected in the general population, particularly for factor V Leiden, prothrombin G20210A, increased factor VIII plasma levels, and hyperhomocysteinemia. Apparently, these defects aggregated in these families.

The risk of VTE in deficient women during COC use strongly increased in the presence of 1 or more other thrombophilic defects. The annual incidences of VTE increased from 3.49% (no defects) to 12.00% (1 concomitant defect). In non- deficient women who used COCs, other thrombophilic disorders increased the risk from 0% (no defects) to 3.13% (1 defect). The combination of a deficiency with more than 1 other thrombophilic defect exceeded the reported risk of VTE during COC use in healthy women up to 300- to 600-fold. This finding indicates a synergistic interaction of various thrombophilic defects and COC use. Two previous studies27,28 that evaluated the risk of multiple thrombophilic factors in carriers of factor V Leiden or prothrombin G20210A showed a similar effect.

Based on our results, we recommend that the use of COCs be strongly discouraged in women with inherited deficiencies of protein S, protein C, or antithrombin. Even in their non-deficient female relatives, this might be considered, unless they have negative test results for all known thrombophilic defects. Because 45% of female relatives were deficient and 70% of non-deficient women had 1 or more other thrombophilic defects, one might suggest not to use COCs to all female relatives of patients with deficiencies, without further testing. Despite an excessive risk of VTE during the use of COCs in women with deficiencies, the impact of this finding is limited because the prevalence of these deficiencies is low. However, it is likely that COCs will also increase the risk of VTE in women with other single or combined thrombophilic defects.27,28

53 Chapter 2

Although the latter are milder risk factors, their contribution to the risk of VTE in the female population may be greater because they are more prevalent. We specu- late that COC-related VTE observed in a small proportion of the female population might be attributed to a preexistent higher risk of VTE as a result of single or combined thrombophilic defects. If so, COCs can safely be prescribed to the remaining majority of women without these defects. We agree that it is neither feasible nor cost-effective to test all women before the use of COCs. In clinical practice, however, testing for hereditary deficiencies of protein S, protein C, and antithrombin should be considered in women with a family history of VTE, especially when VTE had occurred at a young age and was related to the use of COCs.29

This retrospective study has several limitations. The small sample size resulted in wide CIs and unstable estimates. Several events were not established by objective techniques, because these were not available yet. Several relatives were not tested for all concomitant thrombophilic defects, mainly because of insufficient amounts of plasma collected at enrollment. The concomitance of thrombophilic defects was classified by the number of defects, although the thrombotic potency of different defects may vary. However, subgroups of relatives with specific combinations were numerous and consequently too small to allow accurate estimates. We tested 1600 patients to obtain 91 probands, corresponding with a prevalence of hereditary deficiencies of 5.7%, in line with previous series30 of unselected patients with VTE. Therefore, referral bias seems unlikely. Selection bias was avoided by testing consecutive patients with VTE and an extraordinarily high response rate of relatives. Because only a few women died of VTE at a fertile age, the results of this study will not be influenced by an excess of fatal events in deficient relatives. Finally, the risk of VTE might have been overestimated by selecting symptomatic deficient patients, but asymptomatic deficient subjects and their relatives are not identified in clinical practice. Despite these limitations, to our knowledge, this is the first study to address the multicausality of VTE in COC-using women, including all known thrombophilic defects and deficiencies. Further studies are warranted to establish our findings.

In conclusion, the high risk of COC-related VTE in women with hereditary deficiencies of protein S, protein C, or antithrombin strongly depends on frequent concomitance of other thrombophilic defects. These women have VTE at a younger age, but COCs do not increase the overall absolute risk of VTE over fertile lifetime.

54 Chapter 2 Oral contraceptives, antithrombin-, protein C- or protein S deficiency and risk of VTE

REFERENCES

1. Jordan WM. Pulmonary embolism. Lancet. 1961;2:1146-1147.

2. Records Unit and Research Advisory Service, Royal College of General Practitioners. Oral contraceptives and thromboembolic disease. J R Coll Gen Pract 1967;13:267-279.

3. Gerstman BB, Piper JM, Freiman JP, Kennedy DL, Ferguson WJ, Bennett RC. Oral contraceptive oestrogen and progestin potencies and the incidence of deep venous thromboembolism. Int J Epidemiol 1990;19:931-936.

4. Effect of different progestagens in low oestrogen oral contraceptives on venous thromboembolic disease: World Health Organization Collabora- tive Study of Cardiovascular Disease and Steroid Hormone Contracep- tion. Lancet 1995;346:1582-1588.

5. Jick H, Jick S, Gurewich V, Wald Myers M, Vasilakis C. Risk of idiopathic venous thromboembolism in women using oral contraceptives with differing progestagen components. Lancet 1995;346:1589-1593.

6. Kemmeren JM, Algra A, Grobbee DE. Third generation oral contraceptives and risk of venous thrombosis: meta-analysis. BMJ 2001;323:131-134.

7. Anderson FA Jr, Brownell Wheeler HB, Goldberg RJ, Hosmer DW, Patwardhan NA, Jovanovic B, Forcier A, Dalen JE. A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism: the Worcester DVT Study. Arch Intern Med 1991;151:933-938.

8. Nordström M, Lindblad B, Bergqvist D, Kjellström T. A prospective study of the incidence of deep-vein thrombosis within a defined urban population. J Intern Med 1992;232:155-160.

9. Middeldorp S, Meijers JCM, van den Ende AE, van Enk A, Bouma BN, Tans G, Rosing J, Prins MH, Büller HR. Effects on coagulation of levonorgestrel- and desogestrel-containing low dose oral contraceptives: a cross- over study. Thromb Haemost 2000;84:4-8.

55 Chapter 2

10. Tans G, Curvers J, Middeldorp S, Thomassen MC, Meijers JC, Prins MH, Bouma BN, Büller HR, Rosing J. A randomized cross-over study on the effects of levonorgestrel- and desogestrel-containing oral contraceptives on the anticoagulant pathways. Thromb Haemost 2000;84:15-21.

11. Middeldorp S, Henkens CMA, Koopman MW, Hamulyák K, van der Meer, Büller HR. The incidence of venous thromboembolism in family members of patients with factor V Leiden mutation and venous thrombosis. Ann Intern Med 1998;128:15-20.

12. Martinelli I, Taioli E, Buccarelli P, Akhavan S, Manucci M. Interaction between the G20210A mutation of the prothrombin gene and oral contraceptive use in deep vein thrombosis. Arterioscler Thromb Vasc Biol 1999;19:700-703.

13. Aznar J, Vaya A, Estelles A, Segui R, Villa P, Ferrando F, Falco C, Corella D, Espana F. Risk of venous thrombosis in carriers of the thrombin G20210A variant and factor V Leiden and their interaction with oral contraceptives. Haematologica 2000;85:1271-1276.

14. Legnani C, Palareti G, Guazzaloca G, Cosmi B, Lunghi B, Bernardi F, Coccheri S. Venous thromboembolism in young women: role of thrombophilic mutations and oral contraceptive use. Eur Heart J 2002;23:984- 990.

15. Santamaria A, Mateo J, Oliver A, Menendez B, Souto J, Borrel M, Fontcu- berta J. Risk of thrombosis associated with oral contraceptives in women from 97 families with inherited thrombophilia: high risk of thrombosis in carriers of the G2210A mutation of the prothrombin gene. Haematologica 2001;86:965-971.

16. Pabinger I, Schneider B; GTH Study Group on Natural Inhibitors. Throm- botic risk of women with hereditary antithrombin III-, protein C- and protein S-deficiency taking oral contraceptive medication. Thromb Haemost 1994;71:548-552.

17. Simioni P, Sanson B-J, Prandoni P, Tormene D, Friederich PW, Girolami B, Gavasso S, Huisman MV, Büller HR, ten Cate JW, Girolami A, Prins MH. Incidence of venous thromboembolism in families with inherited thrombophilia. Thromb Haemost 1999;81:198-202.

56 Chapter 2 Oral contraceptives, antithrombin-, protein C- or protein S deficiency and risk of VTE

18. Brouwer JLP, Veeger NJGM, Kluin-Nelemans HC, van der Meer J. Multi- causal pathogenesis of venous thromboembolism: evidence from a study in families with hereditary deficiencies of protein C, protein S, or antithrombin. Ann Intern Med 2006;145:807-815.

19. Frezzato M, Tosetto A, Rodeghiero F. Validated questionnaire for the identification of previous personal or familial venous thromboembolism. Am J Epidemiol 1996;143:1257-1265.

20. Brouwer JL, Veeger NJ, Schaaf W, Kluin-Nelemans HC, van der Meer J. Difference in absolute risk of venous and arterial thrombosis between familial protein S deficiency type I and type III: results from a family cohort study to assess the clinical impact of a laboratory test-based classifica- tion. Br J Haematol 2005;128:703-710.

21. Bertina RM, Koeleman BP, Koster T, Rosendaal FR, Dirven RJ, de Ronde H, van der Velden PA, Reitsma PH. Mutation in blood coagulation factor V associated with resistance to activated protein C. Nature 1994;369:64-67.

22. Danneberg J, Abbes AP, Bruggeman BJ, Engel H, Gerrits J, Martens A. Reliable genotyping of the G-20210-A mutation of coagulation factor II (prothrombin). Clin Chem1998;44:349-351.

23. Exner T, Triplett DA, Taberner D, Machin SJ; SSC Subcommittee for the Standardization of Lupus Anticoagulants. Guidelines for testing and revised criteria for lupus anticoagulants. Thromb Haemost 1991;65:320- 322.

24. Araki A, Sako Y. Determination of free and total homocysteine in human plasma by high- performance liquid chromatography with fluorescence detection. J Chromatogr 1987;422:43-52.

25. den Heijer M, Koster T, Blom HJ, Bos GM, Briët E, Reitsma PH, Vanden- broucke JP, Rosendaal FR. Hyperhomocysteinemia as a risk factor for deep-vein thrombosis. N Engl J Med 1996;334:759-762.

26. Rothman KJ, Greenland S. Modern Epidemiology. 2nd ed. Philadelphia, Pa: Lippincott-Raven Publishers; 1998:247-252.

57 Chapter 2

27. Libourel EJ, Bank I, Meinardi JR, Baljé-Volkers CP, Hamulyák K, Middel- dorp S, Koopman MM, van Pampus EC, Prins MH, Büller HR, van der Meer J. Co-segregation of thrombophilic disorders in factor V Leiden carriers: the contributions of factor VIII, factor XI, thrombin activatable fibrinolysis inhibitor and lipoprotein(a) to the absolute risk of venous thromboembolism. Haematologica 2002;87:1068-1073.

28. Vossen CY, Conard J, Fontcuberta J, Fontcuberta J, Makris M, van der Meer FJM, Pabinger I, Palareti G, Preston FE, Scharrer I, Souto JC, Svens- son P, Walker ID, Rosendaal FR; European Prospective Cohort on Throm- bophilia (EPCOT). Risk of a first thrombotic event in carriers of a familial thrombophilic defect. J Thromb Haemost 2005;3:459-464.

29. Vandenbroucke JP, van der Meer FJM, Helmerhorst FM, Rosendaal FR. Family history and risk of venous thromboembolism with contraception: family history is important tool. BMJ 2001;323:752.

30. Heijboer H, Brandjes DPM, Büller HR, Sturk A, ten Cate JW. Deficiencies of coagulation-inhibiting and fibrinolytic proteins in outpatients with deep-vein thrombosis. N Engl J Med 1990;323:1512-1516.

58 Thrombotic risk during oral contraceptive use and pregnancy in women with factor V Leiden or prothrombin mutation: a rational approach to contraception

Elizabeth F.W. van Vlijmen Nic J.G.M. Veeger Saskia Middeldorp Karly Hamulyák Martin H. Prins Harry R. Büller Karina Meijer

Blood 2011;118(8): 2055-2061 Chapter 3

ABSTRACT

Background: Current guidelines discourage combined oral contraceptive (COC) use in women with hereditary thrombophilic defects. However, qualifying all hereditary thrombophilic defects as similarly strong risk factors might be questioned. Recent studies indicate the risk of venous thromboembolism (VTE) of a factor V Leiden mutation as considerably lower than a deficiency of protein C, protein S, or antithrombin.

Methods: In a retrospective family cohort, the VTE risk during COC use and pregnancy (including postpartum) was assessed in 798 female relatives with or without a heterozygous, double heterozygous, or homozygous factor V Leiden or prothrombin G20210A mutation.

Results: Overall, absolute VTE risk in women with no, single, or combined defects was 0.13 (95% confidence interval 0.08-0.21), 0.35 (0.22-0.53), and 0.94 (0.47- 1.67) per 100 person-years, while these were 0.19 (0.07-0.41), 0.49 (0.18-1.07), and 0.86 (0.10-3.11) during COC use, and 0.73 (0.30-1.51), 1.97 (0.94-3.63), and 7.65 (3.08-15.76) during pregnancy. COC use and pregnancy were independent risk factors for VTE, with highest risk during pregnancy postpartum, as demonstrated by adjusted hazard ratios of 16.0 (8.0-32.2) versus 2.2 (1.1-4.0) during COC use.

Conclusions: Rather than strictly contraindicating COC use, we advocate detailed counseling on all contraceptive options, including COCs, addressing the associated risks of both VTE and unintended pregnancy, enabling these women to make an informed choice.

60 Oral contraceptives, pregnancy, factor V Leiden or Prothrombin mutation, and risk of VTE

INTRODUCTION

The risk of venous thromboembolism (VTE) in women using combined oral contraceptives (COCs) is attributed to changes in hemostasis.1-3 These changes may have greater impact in women with thrombophilic defects. Therefore, WHO recommendations state COC use in women with thrombophilic mutations, that is, antithrombin deficiency, protein C deficiency, or protein S deficiency, factor V Leiden, and prothrombin-20210A, as associated with an unacceptable health risk.4 These recommendations are mainly based on case-control studies reporting increased relative risks of VTE during COC use in women with hereditary thrombophilic defects.5-11 However, to qualify all hereditary thrombophilic defects as similarly strong risk factors might be questioned. The absolute risk of VTE in factor V Leiden carriers is estimated being 0.15 per 100 person-years,12 whereas in antithrombin-, protein C–, or protein S–deficient persons these estimates range from 0.7 to 1.7 per 100 person-years, indicating a considerably higher degree of risk.12,13

We previously demonstrated that women with severe hereditary thrombophilic defects, that is, a deficiency of antithrombin, protein C, or protein S, are at very high risk during actual COC use, particularly when concomitant thrombophilic defects are present (4.6 per 100 pill-years), and the use of COCs should be strongly discouraged in these women.14,15

As the absolute risk in women with mild thrombophilic defects is substantially lower than in women with severe thrombophilic defects, withholding COCs in women with mild hereditary thrombophilic defects might be less favorable. When discouraging COC use, an increased risk of unintended pregnancy must be taken into account as alternative non-hormonal contraception is less reliable.16,17 Pregnancy and especially the postpartum period is a strong risk factor for VTE, with an absolute risk of VTE in the general population that is higher than noted for COCs, that is, 0.2 per 100 pregnancy-years18 versus 0.06 per 100 pill-years.19 To balance the risk and benefits of COCs, reliable estimates of the VTE risk associated with both COC use and pregnancy are needed.

To evaluate whether mild thrombophilic defects indeed have lower risk of VTE during COC use, which would be the basis for counseling, the risk of first VTE

61 Chapter 3

during COC use and pregnancy postpartum was assessed in a large retrospective family cohort of women with heterozygous, double heterozygous, or homozygous factor V Leiden or prothrombin-G20210A mutation. Women were divided into those with no, single, or combined mild thrombophilic defects, as we recently demonstrated that aggregation of defects is often noted in thrombophilic families, and that especially the presence of multiple defects may significantly increase the absolute risk of VTE.13-15 To put results into perspective, obtained absolute risks are compared with those during use of alternative contraception.

METHODS

Subjects

In the present retrospective family cohort study, all female relatives were included from 4 family cohorts, which were described in detail elsewhere.20-23 These were first-degree relatives of consecutive patients (probands) with VTE or premature atherosclerosis (≤ 50 years) and factor V Leiden, prothrombin-G20210A, high factor VIII levels (≥ 150 IU/dL), or hyperhomocysteinemia, respectively (Figure 1). Probands and their first-degree relatives were enrolled between 1995 and 1998 (factor V Leiden study) and 1998 and 2004 (prothrombin-G20210A, hyperhomocysteinemia, and factor VIII studies) in 3 university hospitals in The Netherlands (University Medical Center Groningen, University Medical Center Amsterdam, and University Hospital Maastricht). Probands were excluded to avoid bias, as they have experienced VTE by definition. Relatives of probands with premature atherosclerosis were not included in the present study. First-degree relatives aged 15 years or older were identified by pedigree analysis. Relatives were contacted through the probands and were all seen in person at our clinics.

Information on VTE, exposure to exogenous risk factors for VTE, and anticoagulant treatment was collected by physicians and research nurses through medical interviews using a validated questionnaire,24 and by reviewing medical records. All relatives were tested for the presence of factor V Leiden, prothrombin- G20210A, high factor VIII levels, hyperhomocysteinemia (original index defects), and for a deficiency of antithrombin, protein C, and protein S. Because of the retrospective nature of the study, collection of medical history took place at the end of the observation period. As thrombophilia testing was performed after collection of

62 Oral contraceptives, pregnancy, factor V Leiden or Prothrombin mutation, and risk of VTE

these data, the taking of medical histories of relatives was not influenced by the results of thrombophilia testing. For the same reason, any decisions on diagnos- tic outcome and treatments during the observation periods were made without knowledge on presence of thrombophilia. Female relatives with a deficiency of antithrombin, protein C, or protein S were excluded, as these deficiencies are strong risk factors for VTE.12-15 As hyperhomocysteinemia is no longer considered an independent thrombophilic risk factor,23,25 this defect was not taken into account, but these women were not excluded from our cohort. For the purpose of our study, information on contraceptive use and pregnancies (including pregnancy losses) was reconfirmed by a written questionnaire sent by mail. In addition, also general practitioners were contacted for further information. All relatives gave informed consent in accordance with the Declaration of Helsinki and the studies were approved by the institutional review boards of the 3 participating Dutch hospitals.

Diagnosis of VTE

VTE was considered established when diagnosed by compression ultra- sound or venography (deep vein thrombosis), by ventilation/perfusion lung-scan, spiral computed tomography scan or pulmonary angiography (pulmonary embolism), or when the patient had received full-dose heparin and vitamin K antagonists for at least 3 months without objective testing at a time when these techniques were not available. VTE was classified as ‘provoked’ when occurring up to 3 months after exposure to exogenous risk factors, which included surgery, trauma, immobilization for at least 7 days, COC use, pregnancy-postpartum up to 3 months, and malignancy. In the absence of these risk factors, VTE was defined as “unprovoked.” Superficial phlebitis was not considered a thrombotic event.

Laboratory studies

Factor V Leiden and prothrombin-G20210A were demonstrated by PCRs.26,27 Factor VIII:C was measured by 1-stage clotting assay and considered increased at levels above 150 IU/dL.28 Protein S– and protein C–Ag levels were measured by ELISA (DAKO); protein C activity and antithrombin levels (Chromogenix) were measured by chromogenic substrate assays. Normal ranges were determined in

63 Chapter 3

healthy volunteers without a (family) history of VTE, who were neither pregnant nor used COCs within 3 months before blood sampling. Deficiency of antithrombin, protein S, and protein C was defined by levels below the lower limit of their normal ranges. In probands and symptomatic relatives, blood samples were collected at least 3 months after VTE had occurred. If they were still treated with vitamin K antagonists, samples were taken after temporary change of this therapy to low-molecular-weight heparin for at least 2 weeks.

Statistical analysis

We estimated the overall absolute risk of first VTE in female relatives with no defects, a single defect, or a combination of factor V Leiden and prothrombin- G20210A. Homozygosity for factor V Leiden or prothrombin-G20210A was classified as a combined defect, as their prevalence is too low to calculate reliable estimates separately. Furthermore, the absolute risk during actual COC use and during the pregnancy/postpartum period was estimated. The absolute risk was expressed as the incidence rate per 100 person- years. Corresponding 95% confidence intervals (CI) were calculated by using the binomial probability model (conditional small-sample approach).29 Person-years were counted from age 15 until age 50 years, first VTE, or end of study. A minimum age of 15 years was chosen because VTE is rare at younger age, and a maximum age of 50 years as end of fertile lifetime. The duration of exposure to COCs (pill-years) included actual use including a 3-month exposure window after COC use was discontinued. For pregnancy, including pregnancy losses, exposure was defined as the gestation time plus 3-month postpartum.

We used a time-varying exposure Cox proportional-hazard model to estimate adjusted hazard ratios (HRs) of actual COC use and the pregnancy- postpartum period next to the presence of single or combined hereditary mild thrombophilic defects. With this model, we specifically took into account that both COC use and the pregnancy-postpartum period are temporary risk periods during fertile lifetime. Effect modification (interactions) of mild thrombophilic defects on both COC use and the pregnancy- postpartum period were also considered. In addition, the influence of an increased factor VIII level (≥ 150 IU/mL) as an acquired independent risk factor for VTE was estimated.

64 Oral contraceptives, pregnancy, factor V Leiden or Prothrombin mutation, and risk of VTE

Furthermore, the absolute risk of VTE during COC use in women with or without mild hereditary thrombophilia was put into the perspective of contraceptive failure of COC and alternative contraceptives. Alternative contraceptives include condom, the copper-intrauterine devices (IUD; 380 mm2), and the levonorgestrel IUD (LNG-IUD). The LNG-IUD is presented as having no increased risk of VTE, as recently reported in a large study.19 Continuous variables were expressed as mean values and SD or median values and range, and categorical data as counts and percentages. A 2-sided P value < .05 indicated statistical significance. Analyses were performed using SAS Version 9.1 software.

Family cohorts of affected probands with VTE

Affected probands Factor V Leiden Prothrombin High factor VIII Hyperhomocysteinaemia 271 109 156 103

All relatives 15 years 4315

Deceased 921

Non-responders 1371

Relatives enrolled 2023

Male relatives 1034

Female relatives with antithrombin, 54 protein C, or protein S deficiency

Female relatives eligible 935

Female relatives with incomplete thrombophilic tests 137

Female relatives evaluable for analysis 798

Figure 1. Recruitment of the study population from families with factor V Leiden, prothrombin G20120A mutation, high factor VIII levels, and hyperhomocysteinemia, respectively.

65 Chapter 3

RESULTS

Clinical characteristics

The study consisted of 639 unrelated families including 271 probands with factor V Leiden, 109 with prothrombin-G20210A, 156 with high factor VIII levels, and 103 with hyperhomocysteinemia (Figure 1). Of their 4315 relatives aged 15 years or older, 2292 relatives could not be enrolled; 1371 were non-responders (no consent, geographic distance); and 921 had deceased. Of the 2023 relatives enrolled, 989 were female. Of these women, 54 were not eligible as they had a deficiency of antithrombin, protein C, or protein S, leaving 935 eligible female relatives. A total of 137 eligible women were excluded for incomplete thrombophilic tests. The remaining 798 women were available for analysis (Figure 1).

Table 1 lists the characteristics of the 798 female relatives. Of those, 301 had one or more defects. Among these women, 14 were homozygous for factor V Leiden, of whom 3 were also heterozygous for prothrombin-G20210A mutation, and 4 women were homozygous for the prothrombin-G20210A mutation. A total of 205 (68%) of the 301 women with single or combined defects, and 366 (74%) of the 497 women without a defect reported COC use during their lifetime. The majority of women (76%) had only one period of COC use. Seventy-two percent of women had 1 or more pregnancies, with a median number of 3 (range 1-10) pregnancies. Sixty-nine percent of the women who had used COCs, and 75% of women who never used COCs had 1 or more pregnancies. A total of 50 first episodes of VTE were reported, of which 44 (88%) were provoked by exogenous risk factors. The majority, that is, 35 provoked episodes of VTE, occurred during COC use (11), of which 4 in the presence of another risk factor, or during the pregnancy- postpartum period (21), and 3 VTEs occurred after starting COC use during the 12 weeks postpartum (Table 1).

66 Oral contraceptives, pregnancy, factor V Leiden or Prothrombin mutation, and risk of VTE

Table 1. Characteristics of 798 female ﬁrst-degree relatives with either factor V Leiden, prothrombin G20210A, or a combination of these defects, including homozygosity

All female relatives No defects Single defects Combined defects*

Female relatives, n 497 251 50 Follow-up, y† 32 (0.1-35) 28 (0.1-35) 24 (2-35) Factor V Leiden (%) NA 160 (64) 46 (92) Prothrombin G20210A (%) NA 91 (36) 39 (81) Ever COC users, n (%) 366 (74) 171 (68) 34 (68) Age at start of COC, 19 (15-47) 20 (15-49) 19 (15-49) Duration of COC use, 8 (0.1-30) 6 (0.1-27) 6 (0.5-16) Ever pregnant, n (%) 364 (73) 175 (70) 36 (72) Age at ﬁrst pregnancy 24 (15-41) 24 (15-36) 23 (17-38) Number of pregnancies, 2 (1-10) 3 (1-10) 2 (1-7) Total pregnancy time, 2 (0.3-9) 3 (0.3-9) 2 (0.8-6) Age at time of VTE, 31(18-46) 34 (17-48) 26 (17-39) VTE, n† 17 22 11 Unprovoked, 1 4 1 Provoked, n COC use 4 1 2 Pregnancy/postpartum 5 9 7 COC use postpartum 2 1 0 COC use other risk factor 0 4 0 Major trauma, surgery, immobilization 5 3 1

High factor VIII level was present in 8 (22%), 99 (39%), and 197 (40%) women with combined defects, single defects, and no defects, respectively (in 14 women factor VIII levels were not available). Data are given as median (min max) unless otherwise indicated. COC indicates combined oral contraceptive; NA, data not applicable; and VTE, venous thromboembolism. *Including 14 homozygote carriers of factor V Leiden, of whom 3 were also heterozygous for prothrombin G20210A, and 4 homozygote carriers of the prothrombin G20210A mutation, respectively. †Restricted to those aged 15-50 years.

Absolute risk of first VTE

Table 2 lists the absolute risk of VTE associated with COC use and the pregnancy- postpartum period. The crude overall absolute risk of first VTE in our cohort was 0.25 per 100 person-years (95% CI, 0.18-0.32), based on 50 VTEs in 798 women with 20 317 person- years. In women with no, a single, or combined defects, these were 0.13 (95% CI, 0.08-0.21), 0.35 (95% CI, 0.22-0.53), and 0.94 (95% CI, 0.47- 1.67) per 100 person-years, respectively.

Restricting the observation time to actual COC use, the crude incidence of VTE was 0.30 (95% CI, 0.16-0.50) per 100 pill-years, based on 14 VTEs during 4661 pill-years. In women with no, a single, or combined defects, the incidences were 0.19 (95% CI, 0.07-0.41), 0.49 (95% CI, 0.18-1.07), and 0.86 (95% CI, 0.10-3.11) per 100 pill-years, respectively (Table 2).

67 Chapter 3

When considering only the pregnancy-postpartum periods, the crude incidence of first VTE was 1.55 (95% CI, 0.99-2.30) per 100 pregnancy-years, based on 24 VTEs during 1553 pregnancy- years. In women with no, a single, or combined defects, the incidences rose from 0.73 (95% CI, 0.30-1.51) and 1.97 (95% CI, 0.94-3.63) to 7.65 (95% CI, 3.08-15.76) per 100 pregnancy-years, respectively.

Table 2. Absolute risk of VTE in all 798 female relatives

All female relatives

Defects None Single Combined*

All women Total no. 497 251 50 No. with event 17 22 11 Observation period, y 12 908 6234 1175 Incidence rate per 100 person-years (95% CI) 0.13 (0.08-0.21) 0.35 (0.22-0.53) 0.94 (0.47-1.67) Actual pill use Total no. 366 171 34 No. with event 6 6 2 Observation period, pill-years 3211 1218 232 Incidence rate per 100 pill-years (95% CI) 0.19 (0.07-0.41) 0.49 (0.18-1.07) 0.86 (0.10-3.11) Actual pregnancy Total no. 364 175 36 No. with event 7 10 7 Observation period, pregnancy-years 955 507 92 Incidence rate per 100 pregnancy-years (95% CI) 0.73 (0.30-1.51) 1.97 (0.94-3.63) 7.65 (3.08-15.76)

Absolute risk of VTE in all 798 female relatives with no defects, a single defect (factor V Leiden or prothrombin G20210A), or a combination of these defects (including homozygosity) and during actual use of COCs and during actual pregnancy. VTE indicates venous thromboembolism; and CI, conﬁdence interval. *Including 14 homozygote carriers of factor V Leiden, of whom 3 were also heterozygous for prothrombin G20210A, and 4 homozygote carriers of the prothrombin G20210A mutation, respectively.

Relative risk of first VTE

The increased risk of VTE in women with single and combined defects in comparison to women without defects was confirmed in our time-dependent multivariable Cox regression analysis (also including pregnancy and COC use). The HRs were 2.7 (95% CI, 1.4-5.1) for a single defect and 8.5 (95% CI, 3.8-19.2) for combined defects. The substantially higher risk of VTE during the pregnancy-partum period was confirmed by a HR of 16.4 (95% CI, 8.2-32.8). For actual COC use, the HR was 2.1 (95% CI, 1.1-4.1).

When adjusted for factor VIII, HRs were 2.7 (95% CI, 1.4-5.0) and 10.1 (95% CI, 4.4-23.0) for single and combined defects, 2.2 (95% CI, 1.1-4.0) for COC use and 16.0 (95% CI, 8.0-32.2) for pregnancy.

68 Oral contraceptives, pregnancy, factor V Leiden or Prothrombin mutation, and risk of VTE

A high factor VIII level was present in 197 (40%), 99 (39%), and 8 (22%) of women with no, single, and combined defects, respectively. Independent of the presence of factor V Leiden or prothrombin-G20210A mutation, the presence of high factor VIII level was associated with an increased risk of VTE (adjusted HR 2.3; 95% CI, 1.3-4.2).

Absolute risk of VTE in COC users versus users of alternative contraception

The absolute VTE incidence in COC users and in users of alternative contraceptive methods, that is, LNG-IUD users, copper-IUD users, and users of male condom, for a hypothetical group of 100 000 women over 1 year is presented in Table 3. In addition to the absolute risk related to COC use, the hypothetical risks of VTE associated with contraception failure were estimated. In this analysis, both the direct risk of VTE because of the contraceptive method (only present in COC users) and the additional risk of VTE because of contraceptive failure resulting in unintended pregnancies are taken into account.

In Table 3, the absolute risk of VTE of 0.55 per 100 pill-years in COC users with thrombophilic defect(s) is derived from combining both the number of VTEs and observation years in women with single and combined defects as presented in Table 2. The absolute risk of VTE of in users of alternative contraceptives is 2.2-fold lower, 0.25 per 100 years in users of alternative contraceptive methods, that is, based on the adjusted HR of 2.2 for COC use. The absolute risk of VTE of 2.8 per 100 pregnancy years is derived from combining both the number of VTEs and observation years noted in pregnant women with single and combined defects, see Table 2. A similar approach is used in the calculations in women without thrombophilic defects.

In women with a thrombophilic defect, the total number of VTEs is 556 in COC users, 270 in LNG-IUD users, 290 in copper-IUD users, and 586 in condom users. In women without thrombophilic defects, the total number of VTEs is 192, 174, 95, and 100, respectively. In these estimations, the possibility that unintended pregnancies are interrupted is not taken into consideration; in that situation, it is expected that the risk of VTE is lower than presented here.

69 Chapter 3

T able 3. Comparison of thrombosis outcome in women with factor V Leiden or prothrombin G20210A, or a combination of these defects (including homozygosity)

Defects No defects Copper IUD Copper IUD 2 COC LNG-IUD (380 mm ) Condom* COC LNG-IUD (380 mm2 ) Condom* Incidence of ﬁrst VTE per 100 pregnancy-years 0.55† 0.25‡ 0.25‡ 0.25‡ 0.19 0.09 0.09 0.09 Cases of VTE per 100 000 pregnancy-years 550 250 250 250 190 90 90 90 Contraceptive failure rate, per 100 women-years§ 0.2 0.7 1.4 12 0.2 0.7 1.4 12 Unintended pregnancies per 100 000 pregnancy-years 200 700 1400 12 000 200 700 1400 12 000 Incidence of VTE per 100 pregnancy-years¶ 2.8 2.8 2.8 2.8 0.7 0.7 0.7 0.7 Additional cases of VTE 6 20 40 336 2 5 10 84 Total number of VTE 556 270 290 586 192 95 100 174

According to the method described by Koster et al.39 LNG-IUD indicates levonorgestrel intrauterine device; HR, hazard ratio; and VTE, venous thromboembolism. *US data. †(6 2) events during (1218 232) pill-years (see Table 2). ‡Based on an adjusted HR of 2.2 for COC use when compared to no COC use. §Pearl Index for correct use (method failure). ¶(10 7) events during (507 92) pregnancy-years (see Table 2).

DISCUSSION

In a cohort of women with a positive family history of VTE, the presence of single or combined factor V Leiden and prothrombin- G20210A mutation, including homozygotes, resulted in a modest increase in absolute risk of VTE. COC use and the pregnancy-postpartum period were confirmed as risk factors for VTE. Although the absolute risk of VTE significantly increased during COC use (up to 0.86 [95% CI, 0.10-3.11] per 100 pill-years) in women with combined defects, the absolute risk during the pregnancy-postpartum period was by far the most important (up to 7.65 [95% CI, 3.08-15.76] per 100 pregnancy years). The substantially higher risk of VTE during the pregnancy-postpartum period was confirmed by an adjusted HR of 16.0 (95% CI, 8.0-32.2), compared with an adjusted HR of 2.2 (95% CI, 1.1-4.0) for COC use.

Our results further show that, in line with recent publications, the a priori absolute risk of VTE during the pregnancy-postpartum period noted in women without any thrombophilic defect is higher than noted in COC users, that is, 0.73 versus 0.19 per 100 person- years. However, as these risks were observed in thrombophilic families, the absolute risk during the pregnancy-postpartum period and during COC use in our study was approximately 3.5 to 5 times higher than reported in the general community, incidences of 0.2 per 100 pregnancy-years18 and 0.06 per 100 pill-years.19

70 Oral contraceptives, pregnancy, factor V Leiden or Prothrombin mutation, and risk of VTE

As our cohort also included women from one of the original family studies with familial high factor VIII level, which is an acquired risk factor for VTE, the effect of high factor VIII level was also estimated.21 Independent of the presence or absence of factor V Leiden or prothrombin-G20210A, the presence of high factor VIII level was associated with an increased risk of VTE (adjusted HR 2.3; 95% CI, 1.3-4.2).

Several studies have reported COC use5-11,30-33 and pregnancy- postpartum33-36 as contributing factors to the risk of VTE in women with factor V Leiden and/or prothrombin-G20210A. However, adequate interpretation of risks is hampered as most studies only reported relative risks and these estimates differ considerably. In carriers of factor V Leiden or prothrombin-20210A, odds ratios for COC use ranged from 1.3 to 30, while for the pregnancy-postpartum period these ranged from 2 to 53. In addition, it is often unclear whether the presence of other thrombophilic defects was excluded. Furthermore, although COCs prevent from pregnancy, the clinical consequences of both hormonal risks of VTE are seldom considered simultaneously within 1 study.

The major finding in the present study is the high pregnancy- related risk of VTE, relative to the observed risk during COC use. Therefore, if it is decided against COC use, adequate alternative contraception is needed to keep the risk of unintended pregnancy in these women as low as possible. The actual choice in alternative contraception is very limited; only copper-intrauterine devices (IUDs) and condoms. All hormonal contraceptives are considered to increase the risk of VTE, including progestagen-only contraceptives (desogestrel-only pill: Cerazette; etonogestrel implant: Implanon; medroxyprogesterone-acetate injections: Depo- Provera; and LNG-IUD: Mirena©), though this is merely on theoretical grounds by extrapolation of data from combined hormonal contraceptives. However, a recent large cohort study reported the LNG-IUD to not increase the risk of VTE, based on >100 000 women-years of use.19

To put our results into a risk-benefit perspective, we estimated absolute VTE risk of COC use and alternative contraceptive methods, that is, LNG-IUD, copper- IUD, and condoms for a hypothetical group of 100 000 women over 1 year of use, together with VTE risk associated with contraception failure (Table 3). Although COC use results in COC-related VTEs, because of its excellent contraceptive

71 Chapter 3

efficacy,37 the number of unintended pregnancies and subsequent number of pregnancy-related VTEs (6) is very low. The pregnancy-related VTE risk associated with the LNG-IUD and Cu-IUD (380 mm2 Cu-IUD) is estimated to be higher because of their slightly lesser contraceptive efficacy.16,38 Condom use has the lowest contraceptive efficacy, with an up to 60-fold increased risk of unintended pregnancy,17 which makes this option the least favorable alternative. Summarizing these extrapolations, the LNG-IUD and copper-IUD both carry a lower overall risk of VTE and are therefore good alternatives to COCs. The reported higher rate of unintended pregnancies versus COC use is expected not clinically relevant in daily practice, as these contraceptives are not dependent on compliance. However, the specific side effects of LNG-IUD and copper-IUD related to changes in bleeding pattern, that often lead to discontinuation of use, and women’s preferences need be taken into account.

Our study has its limitations. With the retrospective design, not all events were established by objective techniques because these were not yet available at the time. Consequently, the reported absolute risk of VTE may have been overestimated. On the other hand, because of the retrospective design, treating physicians were unaware of the presence of any thrombophilic defects. Furthermore, there was a potential for recall bias on COC exposure. Therefore, extensive efforts were made to minimize the recall bias on COC exposure by verification of patient information through medical files and treating physicians. Not all relatives were tested for all thrombophilic defects. Furthermore, the overall risk of VTE might be overestimated because of the family cohort design by selecting symptomatic affected probands. The risk of VTE associated with these mild hereditary thrombophilic defects will probably be even lower when unselected women are tested for those defects as part of counseling before COC use.

Strong points of our study are its size and the testing for all known thrombophilic defects. Furthermore, we estimated the absolute and relative risk of VTE and took both COC use and pregnancy into account. Moreover, these risks were put into perspective in a modeling exercise, in which both the risks of contraceptive- related VTE and the risk of pregnancy-related VTE (resulting from contraceptive failure) are considered.

In conclusion, factor V Leiden and prothrombin-G20210 are mild risk factors for

72 Oral contraceptives, pregnancy, factor V Leiden or Prothrombin mutation, and risk of VTE

VTE in fertile women. Although in the women with a factor Leiden and prothrombin-G20210 mutation, the absolute risk of VTE increased during COC use, this risk was importantly lower than the absolute risk observed during the pregnancy-postpartum period. These data provide evidence that the policy to contraindicate COC use in these women needs reconsideration. The results of the study do not allow to “promote” a COC use in asymptomatic family-carriers of FVL or PT G20210A, but indicate that when COC use is discontinued the need for adequate alternative contraception has high priority. Rather than strictly contraindicating COC use, we advocate that detailed counseling on all contraceptive options, including COCs, should be performed, addressing the associated risks of both VTE and unintended pregnancy, to enable these women to make an informed choice.

73 Chapter 3

REFERENCES

1. Middeldorp S, Meijers JCM, van den Ende AE, van Enk A, Bouma BN, Tans G, Rosing J, Prins MH, Büller HR. Effects on coagulation of levonorgestrel- and desogestrel-containing low dose oral contraceptives: a cross- over study. Thromb Haemost 2000:84:4-8.

2. Tans G, Curvers J, Middeldorp S, Thomassen MC, Meijers JC, Prins MH, Bouma BN, Büller HR, Rosing J. A randomized cross-over study on the effects of levonorgestrel- and desogestrel-containing oral contraceptives on the anticoagulant pathways. Thromb Haemost 2000;84:15-21.

3. Meijers JC, Middeldorp S, Tekelenburg W, van den Ende AE, Tans G, Prins MH, Rosing J, Büller HR, Bouma BN. Increased fibrinolytic activity during use of oral contraceptives is counteracted by an enhanced factor XI-independent down regulation of fibrinolysis: a randomized cross-over study of two low- dose oral contraceptives. Thromb Haemost 2000;84:9-14.

4. WHO Expert Group Medical eligibility criteria for contraceptive use. Reproductive Health and Research, 4th ed. Geneva, Switzerland: World Health Organization; 2009.

5. Vandenbroucke JP, Koster T, Briët E, Reitsma PH, Bertina RM, Rosendaal FR. Increased risk of venous thrombosis in oral contraceptive users who are carriers of factor V Leiden mutation. Lancet 1994;344:1453-1457.

6. Martinelli I, Taioli E, Buccarelli P, Akhavan S, Manucci PM. Interaction between the G20210A mutation of the prothrombin gene and oral contraceptive use in deep vein thrombosis. Arterioscler Thromb Vasc Biol 1999;19:700-703.

7. Bloemenkamp KW, Rosendaal FR, Helmerhorst FM, Vandenbroucke, JP. Higher risk of venous thrombosis during early use of oral contraceptives in women with hereditary clotting defects. Arch Int Med 2000;160:49-52.

8. Spannagl M, Heinemann LA, Schramm W. Are factor V Leiden carriers who use oral contraceptives at extreme risk of venous thromboembolism? Eur J Contracept Reprod Health Care 2000;5:105-112.

74 Oral contraceptives, pregnancy, factor V Leiden or Prothrombin mutation, and risk of VTE

9. Santamaria A, Mateo J, Oliver A, Menendez B, Souto J, Borrel M, Fontcu- berta J. Risk of thrombosis associated with oral contraceptives in women from 97 families with inherited thrombophilia: high risk of thrombosis in carriers of the G2210A mutation of the prothrombin gene. Haematologica 2001;86:965-971.

10. Legnani C, Palareti G, Guazzaloca G, Cosmi B, Lunghi B, Bernardi F, Coccheri S. Venous thromboembolism in young women: role of thrombophilic mutations and oral contraceptive use. Eur Heart J 2002;23:984- 990.

11. Sidney S, Petitti DB, Soff GA, Cundiff DL, Tolan KK, Quesenberry CP. Venous thromboembolic disease in users of low-estrogen combined estrogen-progestin oral contraceptives. Contraception 2004;70:3-10.

12. Vossen CY, Conard J, Fontcuberta J, Makris JM, van der Meer FJM, Pabin- ger I, Palareti G, Preston FE, Scharrer I, Souto JC, Svensson P, Walker ID, Rosendaal FR. Familial thrombophilia and lifetime risk of venous thrombosis. J Thromb Haemost 2004; 2:1526-1532.

13. Brouwer JL, Veeger NJGM, Kluin-Nelemans HC, van der Meer J. Multicau- sal pathogenesis of venous thromboembolism. Evidence from a study in families with hereditary deficiencies of protein C, protein S, or antithrombin. Ann Int Med 2006;145:807-815.

14. van Vlijmen EFW, Brouwer JLP, Veeger NJGM, Eskes TKAB, de Graeff PA, van der Meer J. Oral contraceptives and the absolute risk of venous thromboembolism in women with single or multiple thrombophilic defects. Arch Int Med 2007;167:282-289.

15. Folkeringa N, Brouwer JLP, Korteweg F, Veeger NJGM, Erwich JJHM, van der Meer J. High risk of pregnancy-related venous thromboembolism in women with multiple thrombophilic defects. Br J Haematol 2007;138:110- 116.

16. Cox M, Tripp J, Blacksell S. Clinical performance of the Nova T380 intrauterine device in routine use by the UK Family Planning and Reproductive Health Research Network: 5-year report. J Fam Plann Reprod Health Care 2002;28:69-72.

75 Chapter 3

17. Bátár I, Sivin I. State-of-the-art of non-hormonal methods of contraception: I. Mechanical barrier contraception. Eur J Contracept Reprod Health Care 2010;15:67-88.

18. Heit JA, Kobbervig CE, James AH, Heit JA, Kobbervig CE, James AH, Petterson TM, MS; Bailey KR, Melton III LJ. Trends in the incidence of venous thromboembolism during pregnancy or postpartum: a 30-year population-based study. Ann Intern Med 2005;143: 697-706.

19. Lidegaard Ø, Løkkegaard E, Svendsen AL, Agger C. Hormonal contraception and risk of venous thromboembolism: national follow-up study. BMJ 2009;339:b2890.

20. Middeldorp S, Henkens CMA, Koopman MW, Hamulyák K, van der Meer, Büller HR. The incidence of venous thromboembolism in family members of patients with factor V Leiden mutation and venous thrombosis. Ann Intern Med 1998;128:15-20.

21. Bank I, Libourel EJ, Middeldorp S, Pampus ECM, Koopman MMW, Hamulyák K, Prins MH, van der Meer J, Büller HR. Elevated levels of FVIII:C within families are associated with an increased risk for venous and arterial thrombosis. J Thromb Haemost 2005;3:79-84.

22. Bank I, Libourel EJ, Middeldorp S, Pampus ECM, Koopman MMW, Hamulyák K, Prins MH, van der Meer J, Büller HR. Prothrombin 20210A mutation: a mild risk factor for venous thromboembolism but not for arterial thrombotic disease and pregnancy-related complications in a family study. Arch Int Med 2004;164:1932-1937.

23. Lijfering WM, Coppens M, van de Poel MHW, Middeldorp S, Hamulyák K, Bank I, Veeger NJGM, Prins MH, Büller HR, van der Meer J. The risk of venous and arterial thrombosis in hyperhomocysteinaemia is low and mainly depends on concomitant thrombophilic defects. Thromb Haemost 2007;98:457-463.

24. Frezzato M, Tosetto A, Rodeghiero F. Validated questionnaire for the identification of previous personal or familial venous thromboembolism. Am J Epidemiol 1996;143:1257-1265.

76 Oral contraceptives, pregnancy, factor V Leiden or Prothrombin mutation, and risk of VTE

25. Lijfering WM, Coppens M, Veeger NJ, Middeldorp S, Hamulyák K, Prins MH, Büller HR, van der Meer J. Hyperhomocysteinemia is not a risk factor for venous and arterial thrombosis, and is associated with elevated factor VIII levels. Thromb Res 2008;123:244-250.

26. Bertina RM, Koeleman BP, Koster T, Rosendaal FR, Dirven RJ, de Ronde H, van der Velden PA, Reitsma PH. Mutation in blood coagulation factor V associated with resistance to activated protein C. Nature 1994;369:64- 67.

27. Danneberg J, Abbes AP, Bruggeman BJ, Engel H, Gerrits J, Martens A. Reliable genotyping of the G-20210-A mutation of coagulation factor II (prothrombin). Clin Chem 1998;44:349-351.

28. Koster T, Blann AD, Briët E, Vandenbroucke JP, Rosendaal FR. Role of the clotting factor VIII in effect of von Willebrand factor on occurrence of deep vein thrombosis. Lancet 1995;345:152-155.

29. Rothman KJ, Greenland S. Modern Epidemiology, 2nd ed. Philadelphia, PA: Lippincott-Raven Publishers;1998:247-252.

30. Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet 1999;353:1167-1173.

31. Couturaud F, Kearon C, Leroyer C, Mercier B, Abgrall JF, Le Gal G, Lacut K, Oger E, Bressollette L, Ferec C, Lamure M, Mottier D; Groupe d’Etude de la Thrombose de Bretagne Occidentale). Incidence of venous thromboembolism in first-degree relatives of patients with venous thromboembolism who have factor V Leiden. Thromb Haemost 2006;96:744-749.

32. Martinelli I, Bucciarelli P, Margaglione M, DeStefano V, Castaman G, Pier Mannuccio M. The risk of venous thromboembolism in family members with mutations in the genes of factor V or prothrombin or both. Br J Haematol 2000;111:1223-1229.

33. Simioni P, Sanson B-J, Prandoni P, Tormene D, Friederich PW, Girolami B, Gavasso S, Huisman MV, Bu ller HR, ten Cate JW, Girolami A, Prins MH. Incidence of venous thromboembolism in families with inherited thrombophilia. Thromb Haemost 1999;81:198-202.

77 Chapter 3

34. Pomp ER, Lenselink AM, Rosendaal FR, Doggen CJ. Pregnancy, the postpartum period and prothrombotic effects: risk of venous thrombosis in the MEGA study. J Thromb Haemost 2008;6:632-637.

35. Martinelli I, Battaglioli T, De Stefano V, Tormene D, Valdrè L, Grandone E, Tosetto A, Mannucci PM; GIT (Gruppo Italiano Trombofilia). The risk of first venous thromboembolism during pregnancy and puerperium in double heterozygotes for factor V Leiden and prothrombin G20210A. J Thromb Haemost 2008;6:494-498.

36. Marik PE, Plante LA. Venous thromboembolic disease and pregnancy. N Engl J Med 2008; 359:2025-2033.

37. Mansour D, Inki P, Gemzel-Danielsson K. Efficacy of contraceptive methods: a review of the literature. Eur J Contr Reprod Health Care 2010;15:4-16.

38. Cox M, Tripp J, Blacksell S. Clinical performance of the levonorgestrel intrauterine system in routine use by the UK Family Planning and Repro- ductive Health Research Network: 5-year report. J Fam Plann Reprod Health Care 2002;28:73-75.

39. Koster T, Small A, Rosendaal FR, Helmerhorst FM. Oral contraceptives and venous thromboembolism: a quantitative discussion of the uncertainties. J Int Med 1995;238:31-37.

78 Combined oral contraceptives, thrombophilia and the risk of venous thromboembolism: a systematic review and meta- analysis

Elizabeth F.W. van Vlijmen Sophie Wiewel-Verschueren Taco B. M. Monster Karina Meijer

Submitted Chapter 4

ABSTRACT

Case presentation A 29-year-old woman previously tested heterozygote for Factor V Leiden mutation (in a research setting 6 years ago) asks: what is my risk of thrombosis if I would start using combined oral contraceptives (COCs)?

Methods We performed a meta-analysis to evaluate the risk of VTE in COC-users with thrombophilia. A distinction was made in ‘mild’ (factor V Leiden, prothrombin- G20210A mutation) and ‘severe’ thrombophilia (antithrombin-, protein C-, protein S-deficiency, double heterozygosity or homozygosity of factor V Leiden and prothrombin-G20210A mutation). We identified 12 case-control- and three cohort studies.

Results In COC-users, mild thrombophilia increased VTE risk six-fold (relative risk (RR) 5.89, 95% Confidence Interval [CI]: 4.21-8.23), while severe thrombophilia increased VTE risk seven-fold (RR 7.15, 95% CI: 2.93-17.45). The cohort studies showed that absolute VTE risk was far higher in COC-users with severe than with mild thrombophilia (4.3-4.6 versus 0.49-2.0 per 100 pill-years, respectively), but with the caveat that absolute risk was estimated in relatives of thrombophilic patients with VTE, i.e. with positive family history.

Conclusion We recommend discouraging COC-use in all women with severe thrombophilia. Contrary, additive VTE risk of mild thrombophilia is modest and with no other risk factors present, e.g. family history, COC-use should not be denied in women who consider alternative adequate contraception not acceptable.

80 Oral contraceptives, thrombophilia and risk of VTE: a meta-analysis

INTRODUCTION

Since their introduction in 1960, combined oral contraceptives (COCs), containing ethinylestradiol and a progestogen, are associated with an increased risk of venous thromboembolism (VTE).1-4 This association is considered related to COC-induced changes in coagulation, anticoagulation, and fibrinolysis in a prothrombotic direction, which alters hemostatic balance. These changes have more impact in women who are already at increased risk of VTE, for instance because of pre-existing hereditary thrombophilia, i.e. protein C, protein S and antithrombin deficiencies and factor V Leiden and prothrombin G20210A mutation. In 1994, a first publication reported increased VTE risk in COC-users who are factor V Leiden mutation carriers.5 Many studies followed of which the majority evaluated the risk of factor V Leiden and prothrombin G20210A mutation as these are more prevalent in the general population, 5%6,7 and 2%,8 respectively, than protein C-, protein S- and antithrombin deficiencies, which have a prevalence of about 0.1% each.9-11 Further, some studies evaluated the risk in COC-users with additional hereditary thrombophilias.12,26,27 The number of cohort studies reporting the absolute VTE risk in COC-users with thrombophilia is limited and restricted to thrombophilic family cohort studies, which reported higher risk with natural anticoagulant deficiencies than with factor V Leiden or prothrombin G20210A mutation. Analyses are mostly based on small subgroups. Currently, WHO Medi- cal Eligibility Criteria for contraceptive use state that COC-use in women with hereditary thrombophilias (antithrombin-, protein C-, protein C-deficiency, factor V Leiden, prothrombin G20210A mutation)13 is associated with an unacceptable health risk.

The aim of this systematic review and meta-analysis is to present summary statis- tics of the risk of 1st VTE in COC-users with mild and severe hereditary thrombophilia.

81 Chapter 4

METHODS

Identification of studies

MEDLINE and EMBASE databases were searched for potential studies published from inception to February 10, 2015 (date of search performed) presenting relevant evidence on VTE risk in COC-users with hereditary thrombophilia, with no restriction in language.

The PubMed database was searched by applying the following search terms: ("Thrombosis"[Mesh] OR "thrombosis" OR "thrombotic" OR thromboembolism*) AND ("Contraceptives, Oral"[Mesh] OR contracept*) AND ("Thrombo- philia"[Mesh] OR "thrombophilia" OR "protein C" OR "factor V Leiden" OR "protein S" OR antithrombin* OR "prothrombin" OR "hereditary" OR "inherited" OR "family history" OR "genetics" OR "genetic").

The EMBASE database was searched by using the search terms: 'thromboembolism'/exp OR thrombosis:ab,ti OR thrombotic:ab,ti OR thromboembolism*:ab,ti AND ('oral contraceptive agent'/exp OR contracept*:ab,ti) AND ('throm-bophilia'/exp OR 'protein c deficiency'/exp OR 'protein s deficiency'/exp OR 'blood clotting factor 5 leiden'/exp OR 'antithrombin iii'/exp OR 'antithrombin'/exp OR 'prothrombin'/exp OR 'family history'/exp OR 'genetic predisposition'/exp OR thrombophil*:ab,ti OR 'protein c':ab,ti OR 'protein s':ab,ti OR 'factor v leiden':ab,ti OR 'factor 5 leiden':ab,ti OR antithrombin*:ab,ti OR prothrombin*:ab,ti OR 'family history':ab,ti OR familial:ab,ti OR genetic:ab,ti OR hereditary:ab,ti OR inherited:ab,ti) NOT [medline]/lim.

The search was developed with the expertise of a professional librarian of the Central Medical Library of the University Medical Center Groningen. The search was extended by manual review of the retrieved papers’ reference lists.

Titles and abstracts of potentially relevant papers retrieved were checked independently by two investigators (E.F.W.v.V. and S.W.-V.) for eligibility of full paper evaluation. Discrepancies in opinion between the 2 principal reviewers regarding eligibility were resolved by discussion with a third investigator (K.M.). Likewise, the same two investigators performed full paper evaluation, and discrepancies were resolved by the third investigator.

82 Oral contraceptives, thrombophilia and risk of VTE: a meta-analysis

Study selection criteria

Regarding selection criteria, we anticipated to include both case-control and cohort studies. Studies were considered eligible if the following criteria were met: original data presented, hereditary thrombophilia was considered (antithrombin-, protein C-, or protein S deficiency, factor V Leiden or prothrombin G20210A mutation, double heterozygosity or homozygosity of factor V Leiden or prothrombin G20210A mutation), restriction to 1st VTE of any type, in case-control studies an analysis was performed comparing the prevalence of thrombophilia in COC- associated VTE cases versus the prevalence in COC-using control persons, in cohort studies the incidence rate of VTE in thrombophilic COC-users was compared to non-thrombophilic COC-users, odds ratio (OR) or rate ratio (RR) provided with underlying data or retrievable based on available data. An additional inclusion criterion for cohort studies was that probands were not included in the analyses. Case series of patients were excluded and in case of studies with multiple publications, the publication with the most inclusive dataset was included.

Quality assessment

The two investigators independently performed a quality assessment of the selected papers, in which components of the Newcastle-Ottawa tool for epidemiological studies were taken into account.14 The following quality issues were considered as relevant: the characteristics of the participants, in- and exclusion criteria for VTE cases, diagnostics of VTE, methods collecting information on COC-use, source of control group, methods of matching cases to controls, and adjustment for confounding.

Statistical analysis

Data synthesis was conducted with Review manager (RevMan, version 5.3. Copen- hagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014), which was used to pool the data for each risk factor, using the Mantel-Haenszel method with a random effects model. When risks of a certain disease are small, Odds ratios (ORs) are considered to reliably estimate the Relative risk (RR); therefore the same method was used for case-control and cohort studies. A distinction is made between studies evaluating COC-users with severe (antithrombin-, protein C-, or protein S deficiency, and double heterozygosity or homozygosity of factor V Leiden or prothrombin G20210A mutation) and mild hereditary thrombophilia

83 Chapter 4

(factor V Leiden or prothrombin G20210A mutation). Pooled results are presented as RR with corresponding 95% confidence intervals. Heterogeneity across studies was tested using the I2 statistic, homogeneity was considered unlikely when p<0.10. Funnel plots were performed to examine publication bias.

RESULTS

The search resulted in 2087 hits, of which 2027 remained after deletion of 60 duplicates. Based on title and abstract screening, 1929 papers were excluded, leaving 98 articles of possible relevance, which were retrieved for full paper evaluation. Additionally, 22 references were identified from these papers.

Based on the predefined in- and exclusion criteria, initially 25 studies (18 case- control studies and seven cohort studies)5,15-38 were selected for detailed evaluation. Of these, six case-control studies15,17,22,26,28,30 were excluded for the following reasons: no subdivision between men and women was provided for the prevalence of thrombophilia;30 prevalence of COC-use and/or thrombophilia in controls was based on estimations only;15,17 cases of VTE had occurred in patients who were all affected with thrombophilia;26 information on the number of COC-users with thrombophilia in the patient- and/or control-groups was not provided or incomplete,28,30 and a re-analysis based on the same dataset was used to assess influence of duration of COC-use.22 Additionally, a subgroup analysis in COC- users with or without double heterozygosity of FVL and prothrombin G20210A mutation was excluded, as the study did not take into account the cases that were identified with homozygosity.23

Further, four family cohort studies32-34,36 were excluded for the following reasons: inclusion of probands with symptomatic VTE in risk estimations performed in thrombophilic family cohorts;32,36 inclusion of women with personal history of VTE32,34 and a larger study dataset available than present in the initial study.33 As a result, 15 studies, consisting of 12 case-control5,16,18-21,23-25,27,29,31 and three cohort studies,35,37,38 were selected. There were no disagreements between the independent reviewers with respect to study eligibility that needed to be resolved by the third investigator. A flow chart is presented in Figure 1.

84 Oral contraceptives, thrombophilia and risk of VTE: a meta-analysis

Studies identified and screened for potential eligibility (n=2027)

Studies excluded after title and abstract screening (n=1929)

Studies retrieved for full evaluation (n=98)

References extracted from studies retrieved for full-text evaluation (n=22)

Studies excluded after full-text evaluation based on selection critera (n=105)

Studies included in review (n=15)

Figure 1. Study flow diagram

All 15 studies were written in the English language. Fourteen studies5,18-21,23- 25,27,29,31,35,38 evaluated additional risk of mild thrombophilia (factor V Leiden, prothrombin G20210A mutation or both). Three studies evaluated additional risk of severe thrombophilias (antithrombin-, protein C-, or protein S deficiency,35,37 and double heterozygous, or homozygous factor V Leiden or prothrombin G20210A mutation38). Characteristics of the studies are presented in Table 1. The number of COC-users with or without thrombophilia in cases- and control-groups contributing to the analysis is presented in Table 2; incidence rates of the cohort studies are presented in Table 3.

Study characteristics

Characteristics of the studies are presented in Table 1. Except for two,5,38 in none of the studies selected COC-users with or without hereditary thrombophilia were the main population to be studied. All case-control studies collected cases from hospitals, except one which used a medical record database.27 All cases occurred in women aged between 15 or 18 to 49 years, except for one study, which included

85 Chapter 4

women above 50 years.29 All case-control studies matched controls for age, or an adjustment for age was performed, and five additionally matched according to region. All studies adjusted for various confounders, of which the majority adjusted for BMI and family history. Seven case-control studies tested also for antithrombin-, protein S-, and protein C deficiency,16,18,19,21,23-25 of which three studies adjusted for other thrombophilias.19,23,25 One cohort study adjusted for clustering of women within families.37 The majority of studies included any VTE or cases of deep vein thrombosis (DVT) and/or pulmonary embolism (PE). Howe- ver, some studies selected cases based on specific VTE type; two included DVT of upper extremity,25 and one study focused on cerebral vein thrombosis (CVT). The quality of the studies was high: degree of information provided on methods applied to collect information on COC-use, the objective methods used in diagno- sing VTE, source of controls, degree of matching to controls, adjustment, and description of in- and exclusion criteria were generally considered adequate.

The number of COC-users with or without thrombophilia in cases- and control- groups contributing to the analysis is presented in Table 2.

86 Oral contraceptives, thrombophilia and risk of VTE: a meta-analysis

Table 2 Individual results of case-controls studies Studies Thrombophilia VTE Cases Controls Odds ratio (95% CI) examined COC*-users COC-users with/without with/without thrombophilia thrombophilia Bloemenkamp5 FVL** 15/65 0/55 26.3 (1.5-449.0) 1995 Andersen16 FVL 14/26 2/26 7.0 (1.4-33.9) 1998 Martinelli18 FVL 3/14 0/27 13.4 (0.6-275.0) 1998 FII*** 7/14 1/27 13.5 (1.5-120.9)

Martinelli19 FVL 11/52 2/41 4.3 (0.9-20.7) 1999 FII 9/52 2/41 3.6 (0.7-17.3) Aznar21 FVL or FII 9/10 0/10 19.0 (0.9-8.2) 2000 Spannagl20 FVL 12/34 10/109 3.9 (1.5-9.7) 2000 Legnani23 FVL 26/86 4/166 12.6 (4.2-37.1) 2002 FII 18/86 2/166 17.4 (3.9-76.6) Vaya24 FII 4/9 0/12 11.8 (0.6-247.8) 2003 Martinell25 FVL or FII 5/18 3/82 7.6 (1.7-34.7) 2004 Sidney27 FVL 10/56 3/59 3.5 (0.9-13.4) 2004 FII 3/56 2/59 1.6 (0.3-9.8)

Roach29 FVL 25/10 1/36 11.5 (1.5-87.1) 2013 FII 11/10 2/36 2.5 (0.5-11.8) Bergendal31 FVL 89/221 6/91 6.1 (2.6-14.5) 2014 FII 29/273 2/97 5.2 (1.2-22.0)

*COC: Combined oral contraceptive; **FVL: factor V Leiden mutation; ***FII: prothrombin G20210A mutation

Risk of VTE

Mild thrombophilia

Based on combined results of 14 studies,6,15-26,28 presence of mild thrombophilia increased the risk of VTE in COC-users almost six-fold (RR 5.89, 95%CI:4.21-8.23) (Figure 2). Heterogeneity between studies was low (I2 = 0%;P=0.47). In separate analyses, presence of factor V Leiden mutation increased risk slightly more than six-fold (RR 6.14 [95%CI: 2.58-14.46]). Between-study heterogeneity was low (I2 = 0%; P = 0.81); presence of prothrombin G20210A mutation increased risk was slightly more than five-fold (RR 5.24 [95%CI: 2.69-10.20]). Heterogeneity between studies was very low (I2 = 4%; P = 0.40) (data not shown).

87 Chapter 4

Figure 2. Overall risk ratio for VTE in COC-users with mild thrombophilia (factor V Leiden or Prothrombin G20210A mutation)

Severe thrombophilia

Based on combined results of three studies,35,37,38 presence of severe thrombophilia increased the risk in COC-users more than seven-fold (RR 7.15, 95%CI:2.93- 17.45) (Figure 3). Heterogeneity between studies was low (I2 = 0%;P = 0.77).

Figure 3. Overall risk ratio for VTE in COC-users with severe thrombophilia (antithrombin-, protein S-, or protein C deficiency, and homozygosity or heterozygosity of factor V Leiden or prothrombin G20210A mutation).

88 Oral contraceptives, thrombophilia and risk of VTE: a meta-analysis

Absolute risk of VTE

Mild thrombophilia

Incidence of VTE in COC-users with mild thrombophilia (factor V Leiden, prothrombin G20210A-mutation) was 0.49 (95%CI:0.18-1.07) and 2.0 (95%CI:0.3-7.2) versus 0.19 (95%CI: 0.07-0.41) and 0.0 (95%CI:0-5.5) per 100 pill-years in COC-users without these mutations28 (Table 3).

Severe thrombophilia

Incidence of VTE in COC-users with double heterozygosity or homozygosity of factor V Leiden or prothrombin G20210A mutation was 0.86 (95%CI:0.10-3.11) versus 0.19 (95%CI:0.07-0.41) per 100 pill-years in COC-users without these mutations.38

Incidence of VTE in COC-users with antithrombin-, protein C-, or protein S deficiency was 4.3 (95%CI:1.4-9.7) and 4.62 (95%CI:2.5-7.9) versus 0.48 (95%CI:0.1-1.4) and 0.7 (95%CI:0.0-3.7) per 100 pill-years in non-deficient COC- users35,38 (Table 3).

Table 3: Individual results of thrombophilic family cohort studies

Study Thrombophilas VTE cases/100 pill-years of use Risk ratio 95% CI examined Incidence rate (95% CI)

With Without thrombophilia thrombophilia

Simioni FVL 2/98 0/65 199935 2.0 (0.3-7.2) 0.0 (0.0-5.5) 3.3 (0.4-120) ACS 3/117 1/150 4.3 (1.4-9.7) 0.7 (0.03-3.7) 6.4 (1.0-41.1) Vlijmen ACS 13/281 3/629 200737 4.62 (2.5-7.9) 0.48 (0.1-1.4) 9.7 (3.0-42.4) Vlijmen FVL or FII 6/1218 6/3211 201138 0.49 (0.18-1.07) 0.19 (0.07-0.41) 2.6 (0.85-8.17) Homozygosity or 2/232 6/3211 heterozygosity of 0.86 (0.10-3.11) 0.19 (0.07-0.41) 4.6 (0.93-22.86) FVL + FII ACS: antithrombin, protein C, protein S deficiency, FVL: factor V Leiden mutation, FII: prothrombin G20210A mutation, CI: confidence interval

89 Chapter 4

Potential sources of bias

As to evaluation of possible sources of bias, in the case-control studies the inclusion criteria applied in the analyses in COC-users were generally comparable, i.e. 1st VTE, although inclusion of cases in some studies depended on VTE type (CVT,18 DVT upper extremity,24,25 DVT lower extremity).19 One case-control study29 had included older women (50-63 years) in comparison to the age range in the other studies (15-49), therefore the meta-analysis for mild thrombophilia was re-performed without this study, but the outcome hardly changed (RR: 5.91 [95%CI:4.10-8.51]). Upper extremity DVT is viewed as somewhat different than lower extremity DVT (especially with respect to possible influence of thrombophilia on the development of a first VTE). However, the results did not change if those studies24,25 are excluded (5.73 [95% CI: 3.92-8.36]). The majority excluded patients with recent risk factors (pregnancy, postpartum, surgery, trauma, immobilization),5,16,27 or co-morbidity (malignancy, systemic disease),21,24,25 or both.16,18,19,31 Information on COC-use was collected by interview and also discharge letters in one study,5 except for two, which used written questionnaires, and similar for both cases and controls.20,29 In 3 studies, the source was not stated.18,19,25

In all three family cohort studies, only relatives were included, as including also probands would have introduced bias since they all had VTE; information on COC-use was collected by interviews and medical record review prior to the thrombophilia testing. VTE was objectively established in all studies, but 4 studies,20,35,37,38 also had included patients based on clinical diagnosis (full-dose anticoagulants for ≥ 3 months). Several studies discussed missing data; one adjusted for missing data.15 Finally, the funnel plots appeared symmetric, suggesting no publication bias.

DISCUSSION

We performed a systematic review and meta-analysis based on 15 studies. In COC-users, mild thrombophilia (factor V Leiden, prothrombin G20210A mutation) increased the risk of VTE six-fold, whereas severe thrombophilia (antithrombin-, protein C- or protein S deficiency, double heterozygosity and homozygosity of factor V Leiden or prothrombin G20210A mutation) increased risk seven-fold.

90 Oral contraceptives, thrombophilia and risk of VTE: a meta-analysis

However, to adequately assess the impact of the relative increase in risk, information on the absolute risk is needed. Two cohort study reported a VTE incidence of 0.49 and 2.0 per 100 pill-years of use in COC-users with mild thrombophilia,35,38 indicating inconsistent outcomes. The noted high incidence of 2.0 per 100 pill-years could be due to the very small subgroup of COC-users with or without factor V Leiden mutation providing pill-years and zero cases in non-thrombophilic COC-users. Two cohort studies35,37 reported incidences of 4.3 and 4.6 per 100 pill-years in COC-users with severe thrombophilia, indicating a far higher risk (Table 3). These differences in absolute risks are also noted in non-affected women from families with severe versus mild thrombophilias. Co-inheritance of other thrombophilic defects could explain the more heightened risk in non- affected women from families with severe thrombophilia; one of the family cohort studies indicated frequent co-existence of other thrombophilic defects.37 The heightened risk in non-deficient users may also explain that the increase in relative risk is not clearly different from that seen in COC-users with mild thrombophilia. The incidence of VTE in COC-users with double heterozygosity or homozygosity of factor V Leiden or prothrombin G20210A mutation was 0.86 per 100 pill-years, suggesting that the absolute risk of this double defect is less serious than an antithrombin-, protein C-, or protein S deficiency.38 All absolute risks are estimated in family members of thrombophilic families, i.e. relatives of thrombophilic patients with VTE, and therefore also have a positive family history, which increases their baseline risk of VTE two- to three-fold.39-42 To put observed risks into perspective, in COC-users with mild thrombophilia and positive family history the absolute VTE risk is increased eight- to 33-fold, and 70-fold in COC-users with severe thrombophilia, when compared to the VTE risk of about 0.06 per 100 person-years,1 estimated in the general population of COC- users. This meta-analysis has some limitations; the risk of mild thrombophilia was largely estimated in a community setting, while risk of severe thrombophilia was exclusively evaluated in a limited number of thrombophilic family cohorts. This is inherently due to the very low prevalence of severe thrombophilia. Further, absolute risks were all estimated in members of thrombophilic family cohort studies, risks will therefore be more pronounced because of the co-existing family history, than in the general population of COC-users tested positive.

91 Chapter 4

In conclusion, the presence of mild and severe thrombophilia increases the risk of VTE in COC-users six, and seven-fold, respectively. However, absolute VTE risk estimates indicate the contribution of severe thrombophilia to the VTE risk in COC-users as considerably higher (4.3 and 4.6 per 100 pill-years) than the additional risk of mild thrombophilia (0.49 and 2.0 per 100 pill-years), but with the caveat that these risks were estimated in thrombophilic COC-users who also had a positive family history. As a co-existing family history increases VTE risk two- to threefold, estimated risks are more pronounced than in a general population of COC-users tested positive for thrombophilia.

Recommendations According to GRADE quality of evidence grading, sound epidemiological studies as included in this meta-analysis would be graded as low evidence, but when observed effects are large and there is no obvious bias explaining those effects, the evidence may be rated as moderate or even high, as further research is very unlikely to change our confidence in the estimate of effect (http://gradeworkingroup.org). From this standpoint the strong recommendation is given based on the high additive risk, to avoid COC-use in women with severe hereditary thrombophilia in all cases. The additive VTE risk of mild thrombophilia is modest and when no other risk factors are present, e.g. family history, COC-use should not be denied in women who consider alternative adequate contraception not acceptable, as in this situation the increased risk of pregnancy-related VTE and increased risk of pregnancy outweigh the COC-associated risk.

92 Oral contraceptives, thrombophilia and risk of VTE: a meta-analysis

References

1. Lidegaard Ø, Løkkegaard E, Svendsen AL, Agger C. Hormonal contraception and risk of venous thromboembolism: national follow-up study. BMJ 2009;339:b2890.

2. van Hylckama V, Helmerhost FM, Vandenbroucke JP, Doggen CJ, & Rosendaal FR. The venous thrombotic risk of oral contraceptives, effects of oestrogen dose and progestogen type: results of the MEGA case- control study. BMJ 2009:339:b2921.

3. de Bastos M, Stegeman BH, Rosendaal FR, Van Hylckama Vlieg A, Helmerhorst FM, Stijnen T, Dekkers OM. Combined oral contraceptives: venous thrombosis. Cochrane Database of Systematic Reviews 2014, Issue 3. Art. No.: CD010813.

4. Vinogradova Y, Coupland C, Hippisley-Cox J. Use of combined oral contraceptives and risk of venous thromboembolism: nested case-control studies using the QResearch and CPRD databases. BMJ 2015;350:h2135.

5. Bloemenkamp KW, Rosendaal FR, Helmerhorst FM, Büller HR, Vanden- broucke JP. Enhancement by factor V Leiden mutation of risk of deep-vein thrombosis associated with oral contraceptives containing a third- generation progestagen. Lancet 1995;346:1593-1596.

6. Rosendaal FR. Risk factors for venous thrombosis: prevalence, risk and interaction. Semin Hematol 1997;34:171-87.

7. Rees DC, Cox M, Clegg JB. World distribution of factor V Leiden. Lancet. 1995;346:1133-34.

8. Rosendaal FR, Doggen CJM, Zivelin A, Arruda VR, Aiach M, Siskovick DS, Hillarp H, Watzke HH, Bernardi F, Cumming AM, Reitsma PH. Geograp- hic distribution of the 20210 G to A prothrombin variant. Thromb Haemost 1998;79:706-08.

9. Tait RC, Walker ID, Perry DJ, Islam SI, Daly ME, McCall F, Conkie JA, Carrell RW. Prevalence of antithrombin deficiency in the healthy population. Br J Haematol 1994;87:106-12.

10. Tait RC, Walker ID, Reitsma PH, Islam SI, McCall F, Poort SR, Conkie JA, Bertina RM. Prevalence of protein C deficiency in the healthy population. Thromb Haemost 1995;73:87-93.

93 Chapter 4

11. Dykes AC, Walker ID, McMahon AD, Islam SI, Tait RC. A study of Protein S antigen levels in 3788 healthy volunteers: influence of age, sex and hormone use, and estimate for prevalence of deficiency state. Br J Haema- tol 2001;113:636.

12. Emmerich J1, Rosendaal FR, Cattaneo M, Margaglione M, De Stefano V, Cumming T, Arruda V, Hillarp A, Reny JL. Combined effect of factor V Leiden and prothrombin 20210A on the risk of venous thromboembolism - pooled analysis of 8 case-control studies including 2310 cases and 3204 controls. Study Group for Pooled-Analysis in Venous Thromboembolism. Thromb Haemost 2001;86:809-16.

13. WHO Expert Group Medical eligibility criteria for contraceptive use. Reproductive Health and Research, 5th ed. Geneva, Switzerland: World Health Organization; 2015.

14. Wells G, Shea B, OConnell D, Peterson J, Welch V, Losos M, Tugwell P. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonran- domised studies in meta-analyses. Available at: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp

15. de Bruijn SFTM, Stam J, Koopman MMW, Vandenbroucke JP for the Cere- bral Venous Sinus Thrombosis Study Group. Case-control study of cerebral sinus thrombosis in oral contraceptive users who are carriers of hereditary prothrombotic conditions. BMJ 1998;316:589-592.

16. Andersen BS, Olsen J, Nielsen GL, Steffensen FH, Sorensen HT, Baech J, Gregersen H. Third generation oral contraceptives and heritable thrombophilia as risk factors of non-fatal venous thromboembolism. Thromb Haemost 1998;79:28-31.

17. Bennet L, Odeberg H. Resistance to activated protein C, highly prevalent amongst users of oral contraceptives with venous thromboembolism. J Intern Med 1998;244:27-32.

18. Martinelli I, Sacchi E, Landi G, Taioli E, Duca F, Manucci PM. High risk of cerebral vein thrombosis in carriers of a prothrombin-gene mutation and in users of oral contraceptives. N Engl J Med 1998;338:1793-1797.

94 Oral contraceptives, thrombophilia and risk of VTE: a meta-analysis

19. Martinelli I, Tailoi E, Bucciarelli P, Akhavan S, Manucci PM. Interaction between the G20210A mutation of the prothrombin gene and oral contraceptives use in deep vein thrombosis. Arterioscler Thromb Vasc Biol 1999;19:700-703.

20. Aznar J, Vaya A, Estelles A, Mira Y, Segui R, Villa P, Ferrando F, Falco C, Corella D, Espana F. Risk of venous thrombosis in carriers of the prothrombin G20210A variant and factor V Leiden and their interaction with oral contraceptives. Haematologica 2000;85;1271-1276.

21. Spannagl M, Heinemann LA, Schramm W. Are factor V Leiden carriers who use oral contraceptives at extreme risk for venous thromboembolism? Eur J Contracept Reprod Health Care 2000;5:105-112.

22. Bloemenkamp KW, Rosendaal FR, Helmerhorst F.M, Vandenbroucke, JP. Higher risk of venous thrombosis during early use of oral contraceptives in women with inherited clotting defects. Arch Intern Med 2000;160:49-52.

23. Legnani C, Palareti G, Guazzaloca G, Cosmi B, Lunghi B, Bernardi F, Coccheri S. Venous thromboembolism in young women: role of thrombophilic mutations and oral contraceptive use. Eur Heart J 2002;23:984-990.

24. Vaya A, Mira Y, Mateo J, Falco C, Villa P, Estelles A, Aznar J. Prothrombin G20210A mutation and oral contraceptive use increase upper extremity deep vein thrombotic risk. Thromb Haemost 2003;89:452-457.

25. Martinelli I, Battaglioli T, Buccarelli P, Passamonti SM, Manucci PM. Risk factors and recurrence rate of primary deep vein thrombosis of the upper extremities. Circulation 2004;110:566-570.

26. Ehrenforth S, Nemes L, Mannhalter C, Rosendaal FR, Koder S, Zoghlami- Rintelen C, Pabinger I. Impact of environmental and hereditary risk factors on the clinical manifestation of thrombophilia in homozygous carriers of factor V:1691A. J Thromb Haemost 2004;2:430-436.

27. Sidney S, Petitti DB, Soff GA, Cundiff DL, Tolan KK, Quesenberry CP. Venous thromboembolic disease in users of low-estrogen combined estrogen-progestin oral contraceptives. Contraception 2004;70:3-10.

95 Chapter 4

28. Gadelha T, André C, Jucá AAV, Nucci M. Prothrombin 20210ª and oral contraceptive use as risk factors for cerebral venous thrombosis. Cerebro- vasc Dis 2005;19:49-52.

29. Roach REJ, Lijfering WM, Helmerhorst FM, Cannegieter SC, Rosendaal FR, van Hylckama Vlieg. The risk of venous thrombosis in women over 50 years old using oral contraception or postmenopausal hormone therapy. J Thromb Haemost 2013;11:124-131.

30. Tufano A, Guida A, Coppola A, Nardo A, Capua MD, Quintavalle G, Di Minno MND, Cerbone, AM, Di Minno G. Risk factors and recurrent thrombotic episodes in patients with cerebral venous thrombosis. Blood Transfus 2014;12:s337-342.

31. Bergendal A, Persson I, Odeberg J, Sundström A, Holmström M, Schul- man S, Björgell O, Kieler H. Association of venous thromboembolism with hormonal contraception and thrombophilic genotypes. Obstet Gyne- col 2014;124:600-609.

32. Pabinger I, Schneider B, and the study GTH group on natural inhibitors. Thrombotic risk of women with hereditary antithrombin III, protein C- and protein S deficiency taking oral contraceptive medication. Thromb Haemost 1994;71:548-552.

33. Middeldorp S, Henkens CM, Koopman MM, van Pampus EC, Hamulyák K, van der Meer J, Büller HR. The incidence of venous thromboembolism in family members of patients with factor V Leiden mutation and venous thrombosis. Ann Intern Med 1998;128:15-20.

34. Van Boven HH, Vandenbroucke JP, Briët E, Rosendaal FR. Gene-gene and gene-environment interactions determine the risk of thrombosis in families with inherited antithrombin deficiency. Blood 1999;94:2590-2594.

35. Simioni P, Sanson BJ, Prandoni P, Tormene D, Friederich PW, Girolami B, Prins MH. Incidence of venous thromboembolism in families with inherited thrombophilia. Thromb Haemost 1999;81:198-202.

96 Oral contraceptives, thrombophilia and risk of VTE: a meta-analysis

36. Santamaria A, Mateo J, Oliver A, Menendez B, Souto J, Borrel M, Fontcu- berta J. Risk of thrombosis associated with oral contraceptives of women from 97 families with inherited thrombophilia: high risk of thrombosis in carriers of the G20210A mutation of the thrombin gene. Haematologica 2001;86:965-971.

37. Van Vlijmen EFW, Brouwer JL, Veeger N, Eskes TK, Graeff PA, van der Meer J. Oral Contraceptives and the Absolute Risk of Venous Throm- boembolism in Women With Single or Multiple Thrombophilic Defects. Arch Intern Med 2007;167:282-289.

38. Van Vlijmen EF, Veeger N, Middeldorp S, Hamulyák K, Prins MH, Büller, Meijer K. Thrombotic risk during oral contraceptive use and pregnancy in women with factor V Leiden or prothrombin mutation: a rational approach to contraception. Blood 2011;118:2055-2061.

39. Bezemer ID, van der Meer FJM, Eikenboom JCJ, Rosendaal FR, Doggen CJM. The value of family history as a risk indicator for venous thrombosis. Arch Intern Med 2009;169:610-615.

40. Noboa S, Le Gal, Lacut K, Mercier B, Leroyer C, Nowak E, Mattier D, Oger E for the EDITH Collaborative Study Group. Family history as a risk factor for venous thromboembolism. Thromb Res 2008;122:624-629.

41. Dowling NF, Austin H, Diley A, Whitsett C, Evatt BL, Hooper WC. The epidemiology of venous thromboembolism in Caucasians and African- Americans: the GATE Study. J Thromb Haemost 2003;1:80-87.

42. Silverstein MD, Heit JA, Mohr DN, Petterson TM, O'Fallon WM, Melton III LJ. Trends in incidence of deep vein thrombosis and pulmonary embolism. Arch Intern Med 1998;158:585-593.

97 Chapter 4 Age, Other thrombophilias Age, Family history, FVL jointly Adjustment Matching factors BMI, Smoking, Parity, Missing data 1977-? Study period 1988-1992 1991-1997 1995-1998 Age, Region, Education Age, Region No Matching factors No 3 2 2 ≤ 2 Cases: COC-use at time of VTE Controls: COC-use at inclusion Cases: COC-use ≤ 3 months prior to VTE Controls: COC-use ≤ months prior to inclusion Cases: COC use weeks prior to VTE Controls: COC-use at time of blood sampling Definition COC-use Cases: COC use ≤ weeks prior to VTE Controls: COC-use ≤ weeks prior to inclusion Cases: interview Controls: Questionnaire Source COC-use Cases: interview/ hospital discharge letter Controls: interview Source COC-use not stated Source COC-use not stated Diagnosis VTE* Objectively confirmed Objectively confirmed Not stated Objectively confirmed Source cases/ controls Cases: hospital discharge registries Controls: blood donors Cases: Hospital Controls; friends/ partners Cases; Hospital Controls: friends/ partners/other patients Cases: hospitals Controls: Friends/ partner/other patients 33/28 30/46 30/32 Mean/median age case/ controls 35/Not stated FVL FII ACS FVL FVL ACS Thron bophilia testing FVL FII ACS 3 months: ≤ Surgery Trauma Pregnancy Postpartum Malignancy Immobility SVT Crohn/colitis Heart failure Malignancy Autoimmune disease Pregnancy Postpartum Postmenopause Malignancy Autoimmune disease, Pregnancy Postpartum Malignancy Pregnancy Postpartum, Recent miscarriage Exclusion criteria Cases DVT, aged st Women with 1st VTE, aged 18-49 years Inclusion criteria Cases 15-49 years Patients with 1st CVT aged 15-64 years Women with 1 Consecutive women with 1st DVT lower extremities aged 15-48 years

p m a k n 5 18 19 16 me e o l Study 1995 Case-control studies B Netherlands FVL Italy FVL, FII Martinelli 1998 Martinelli 1999 Italy FVL, FII Andersen 1998 Denmark FVL, ACS Table 1. Study characteristics of case-control- and cohort studies included in the analysis

98 Study Inclusion Exclusion Thron Mean/median Source Diagnosis Source Definition Matching Study Adjustment criteria criteria bophilia age case/ cases/ VTE* COC-use COC-use factors period Cases Cases testing controls controls

Aznar Consecutive Malignancy FII Not stated Cases: Objectively Cases: Cases: Age, Not stated Age 200020 patients with FVL hospital confirmed Interview COC-use at Region Spain 1st DVT/PE ACS Controls: Controls: time of the FV, FI No age blood donors Interview study restriction Controls: COC-use at time of the study

Spannagl Consecutive Malignancy FVL 34/34 Cases: Objectively Cases: Cases: Age 1995-1997 Age, 200021 women with Infectious, hospitals confirmed or Questionnaire COC-use Family history, Germany DVT/PE aged autoimmune/ Controls: anticoagulant Controls: at time of VTE BMI, FVL 15-49 years liver/renal random treatment Questionnaire Controls: Varicose veins disease sample COC-use at History drug population - inclusion abuse database 99 Oral contraceptives,thrombophiliaandriskofVTE:ameta-analysis Legnani Consecutive Isolated PE FVL 30/33 Cases: Objectively Cases: Cases: COC use Region 1994-2000 Age, 200223 women with SVT FII hospital confirmed Interview at time of VTE Other Italy DVT/PE aged DVT upper ACS Controls: Controls: Controls: COC thrombophilias FVL, FII 15-49 years limb or unusual from general Interview at inclusion FVL+FII site population

Vaya Consecutive Malignancy FVL Not stated Cases: Objectively Source COC- Cases: COC-use Age 1997-2001 Confounding 200324 patients with Infection FII hospital confirmed use not stated ≤ 2 weeks prior factors not Spain 1st DVT upper Autoimmune/ AT Controls: to VTE specified FVL, FII extremity Liver/renal healthy Controls: COC- aged 15-75 disease volunteers use ≤ 2 weeks years History drug from same prior to sampling abuse hospitals

Martinelli Patients with DVT due to FVL 32/30 Cases: Objectively Source COC- Cases: Age 1994-2003 Age, 200425 1st DVT of central vein FII hospital confirmed use not stated COC use at time Other Italy upper- catheter ACS Controls: of VTE thrombophilias FVL, FII extremity Malignancy partners or Controls: COC No age Related to friends use at inclusion restriction thrombophilic family Chapter 4 Smoking, BMI, immobilization Cohort studies Race/ethnicity, Income, BMI Adjustment No No Clustering within families Age, BMI, Smoking, Positive family history 2000-2004 1995-1998 and 1998-2004 Study period Not stated 2003-2009 1998-2000 1999-2004 Age NA Age NA Matching factors NA Region 3 3 months Idem COC-use and duration collected Cases: COC-use ≤ months prior to VTE Controls: ≤ prior to inclusion Idem Cases: COC-use at time of VTE Controls: COC use at inclusion Definition COC-use Cases: COC-use ≤ 12 months prior to VTE Controls: ≤ 12 months prior to inclusion Interview and medical record and/or GP Interview Source COC-use Cases: interview Controls: interview Interview and medical record and/or GP Cases: interview Controls: Interview Cases: Questionnaire Control: Questionnaire or V Leiden mutation, FII: prothrombin-G20210A MRI, VP-lung scan, pulmonary angiogram. Objectively confirmed or anticoagulant treatment Diagnosis VTE* Objectively confirmed Objectively confirmed or clinical suspicion PE Idem Idem Objectively confirmed e Cases: Hospitals Controls: Swedish population register NA NA Source cases/ controls Cases: medical records Database Controls: databas same NA Cases: hospital Controls: Partners/ Random digit dialling 150%), BMI: body mass index, NA: not applicable ≥ NA NA Mean/median age case/ controls 41/42 NA 53/53 35/33 ACS FVL ACS FVL FII Thron bophilia testing FVL FII FVL FII ACS FVL FII FVL FII Double defect Malignancy No consent Probands Dead Geographic distance No consent Laboratory incomplete Pregnancy Hysterectomy Ovariectomy HRT-use Missing COC data, estrogen dose >50 mcg Exclusion criteria Cases Pregnancy < 3 months Malignancy Probands Dead Geographic distance No consent Laboratory incomplete Severe psychiatric problems Inability to speak Dutch DVT/PE

st Inclusion criteria Cases Women with 1st DVT/PE aged 18-54 years Relatives, aged >15 years, of patients with VTE and ACS or FVL 1st degree female relatives, aged 15-50 years, of patients with VTE and FVL/F2 1st degree female relatives, aged 15-50 years, of patients with VTE and ACS Women with 1 Aged 15-44 years Women with 1st DVT/PE aged 50-63 years

2 n s s s ote 35 nd nd nd g me men a a a l l l lij 38 lij 37 27 31 29 er er V er V

h h h n n et et et Cohort studies *Venography, pletysmography, Doppler US, duplex color Intra-arterial angiography, CT-scan, IV digital angiographic imaging, ACS: antithrombin, protein S or C deficiency, FVIII: increased factor VIII levels ( VTE: venous thromboembolism, PE: pulmonary embolism, CVT: cerebral vein thrombosis, SVT: superficial FVL: fact Study N FVL or FII FVL+FII Homozy 2011 va N ACS va 2007 USA FVL, FII Sidney 2004 1999 Italy N FVL, ACS Simioni Bergendal 2014 Sweden FVL, FII Netherlands FVL, FII Roach 2013

100 A male or female family history of VTE: does it influence VTE risk in women?

Elizabeth F. W. van Vlijmen Nic J. G. M. Veeger Saskia Middeldorp Karly Hamulyák Martin H. Prins Hanneke Kluin-Nelemans Karina Meijer

Submitted

Chapter 5

ABSTRACT

Background Women from thrombophilic families have an increased risk of venous thromboembolism (VTE), which increases further during combined oral contraceptive use (COCs) and pregnancy. It is unknown whether this additional risk differs between relatives of male and female patients with VTE, and also whether it matters if the female patient had a hormonally-related VTE (during COC-use or pregnancy).

Methods In a retrospective thrombophilic family cohort of 1005 first-degree female relatives of reproductive age, we compared VTE risk in relatives of female versus male patients (probands), and between relatives of female patients with and without hormonally-related VTE.

Results Follow-up covered 24.195 person-years with 84 VTEs, of which 82% was related to COC-use or pregnancy. Crude absolute VTE risk was 0.35 (95% confidence interval [CI]:0.28-0.43) per 100 person-years. Risk in relatives of female (0.32, 95%CI: 0.23-0.43) versus male patients (0.39, 95%CI:0.28-0.53) was comparable, but young age (<45 years at time of 1st VTE) of the patient, and presence of thrombophilia, COC-use and pregnancy of the relative were factors which significantly increased risk of VTE. In the multivariable analysis, thrombophilia, COC-use and pregnancy of relative, but not patients’ age, remained independent risk factors. Additionally, the heterogeneity analysis of risk estimates suggested that VTE risk in relatives of female versus male patients differs during pregnancy-postpartum (Hazard Ratio [HR], 11.6 [95%CI:5.9-22.7] versus 6.6 [95%CI:3.2-13.6]) and to a lesser extent during COC-use (HR, 3.6 [95%CI:1.9-6.7] versus 2.7 [95%CI:1.4-5.3]), although not statistically significant.

Taking hormonal exposure in the patient at the time of VTE into account, incidence of VTE was significantly higher, i.e. 0.43 (95%CI:0.31-0.59) per 100 person-years in relatives of patients with hormonally-related VTE, than in relatives of female patients without hormonally-related VTE (0.13 [95%CI:0.05-0.27], HR, 3.1 [95%CI:1.3-7.4]).

102 A male or female family history and risk of VTE

When considering only relatives of female patients, the higher HRs of pregnancy and COC-use are mainly observed in relatives of patients with hormonally-related VTE. The risk estimates in relatives of female patients without hormonally-related VTE were substantially lower.

Conclusion Summarized, these findings suggest that a family history originating from a female patient, especially when that patient experienced a hormonally-related VTE, may further increase VTE risk in her female relatives. This information could be of importance when counseling women on contraceptive choices.

103 Chapter 5

INTRODUCTION

There are many factors that contribute to the individual baseline risk of venous thromboembolism (VTE). In this, hereditary thrombophilia and first degree family history are both established independent risk factors.1-4 The baseline risk can be further increased by underlying conditions like cancer and obesity, by exogenous risk factors such as surgery, trauma, and in women of reproductive age the use of combined oral contraceptives (COCs) and the pregnancy-postpartum period.5 Recent studies indicate that exposure to hormonal factors plays a decisive role in the development of VTEs in women during the period of reproductive age. Adjust- ment for reproductive risk factors was reported to reduce risk of VTE in women of reproductive age to half of that in men of the same age.6

Previously we have studied the contribution of hormonal exposure, i.e. COC-use and pregnancy-postpartum, to the baseline risk of VTE in female relatives of patients with VTE from families with mild (factor V Leiden and prothrombin 2020A mutation) and severe hereditary thrombophilia (antithrombin, protein C, and protein S deficiency).7-9 It was shown that in female relatives of reproductive age with thrombophilia the risk of VTE was further increased during hormonal exposure, of which the impact was highest in women with severe thrombophilic defects. Additionally, in these cohorts of female relatives of patients with VTE, who per definition all had a positive family history, those without thrombophilia also had a higher risk of VTE during hormonal exposure than reported in the general population.7,8 This is in line with publications indicating that family history is an independent risk factor, regardless of other risk factors present.4 It is unknown whether this family-conferred risk is hormone- or gender specific: is a young woman at higher hormone-related risk if her mother or sister had COC- or pregnancy-related VTE than if her father or brother had thrombosis? This would be relevant for counseling purposes.

In order to investigate this question, we have included in the present study the first-degree female relatives of reproductive age of consecutive patients with VTE and hereditary thrombophilia from our thrombophilic family cohorts. As all patients had experienced VTE, all female relatives had an established positive family history of VTE. This provided us with the unique opportunity to evaluate in these female relatives the impact of a family history originating from a male or female patient, and also to investigate whether a positive family history originating

104 A male or female family history and risk of VTE

from a female patient with a VTE during hormonal exposure (during COC-use or pregnancy) has an additional impact on their risk profile. For this purpose, we estimated absolute and relative risk of VTE in female relatives of male patients, of female patients with hormonally-related VTE, and of female patients without hormonally-related VTE.

METHODS

Subjects

In the present retrospective family cohort study, female relatives were included from five cohorts of thrombophilic families from three university hospitals in the Netherlands, which were described in detail elsewhere.1-3 The first cohort was single center study of first-degree relatives of consecutive patients (probands) with VTE and antithrombin-, protein C-, or protein S-deficiency. As the number of antithrombin deficient probands was small, second-degree relatives with a deficient parent were also identified. Enrolment took place from 1999 to 2004. Three multicenter studies included first-degree relatives of consecutive probands with VTE or premature atherosclerosis (<50 years) and prothrombin-G20210A, high factor VIII levels (150 IU/dL), or hyperhomocysteinemia, respectively. The fifth study was a multicenter study, which included first-degree relatives of patients with VTE and a factor V Leiden mutation. These relatives were enrolled between 1995 and 1998 (factor V Leiden study) and 1998 and 2004 (prothrombin- G20210A, hyperhomocysteinemia, and factor VIII studies) in 3 university hospitals in The Netherlands. Probands were excluded to avoid bias, as they have experienced VTE by definition. Relatives of patients with premature atherosclerosis were not included in the present study. In these cohorts, only hereditary thrombophilic defects were taken into account; a high factor VIII level and hyperhomocysteinemia are not considered hereditary thrombophilic defects (Figure 1). Strategies applied for collection of information on exposure to exogenous risk factors for VTE, including contraceptive use and pregnancies (including pregnancy losses), are described in detail in our previous studies based on these cohorts.7-9 For the present cohort study, all data collected on hormonal exposure and risk factors present at time of VTE of the patients was confirmed by review of the patient files. All relatives had given informed consent and the original study cohorts were approved by the institutional review boards of the three participating

105 Chapter 5

Dutch hospitals (University Medical Centers of Groningen, Amsterdam, and Maastricht). Person-years of exposure were counted from age 15 until 50 years, first VTE, or end of study. A minimum age of 15 years was chosen because VTE occurrence below this age is rare, and a maximum age of 50 years as end of fertile lifetime. The duration of exposure to COCs (years of COC-use) included actual use including a 3-month exposure window after COC-use was discontinued. For pregnancy, including pregnancy losses, exposure was defined as the gestation time plus 3-month postpartum.

Diagnosis of venous thromboembolism Details on the methods applied for diagnosis of VTE can be found in the original studies.1-3

Laboratory studies Methods applied to demonstrate presence of Factor V Leiden and prothrombin- G20210A mutation and antithrombin-, protein S- and protein C-deficiency are described in detail elsewhere.1-3

Statistical analysis VTE risk estimation: Next to estimation of the overall absolute risk of VTE in this cohort, absolute VTE risk was estimated in female relatives originating from a female versus a male proband (patient). The absolute risk was expressed as the incidence rate (IR) per 100 person-years, with 95% confidence intervals (CI). Relative risks were estimated using Cox regression analysis. To account for time-varying exposures of COC- use and pregnancies, an extended Cox model (with time-varying exposures) was used. In this multivariable model, hazard ratios of female versus male proband, age of the patient at time of 1st VTE (<45 years versus ≥45 years), COC-use versus no use and of pregnancy-postpartum versus no pregnancy and presence of thrombophilia versus no thrombophilia were estimated. Possible effect modification (interactions) of mild or severe thrombophilic defects on risk of COC-use and the pregnancy-postpartum period were also taken into account.

Furthermore, within the group of female relatives originating from a female proband, absolute and relative risk estimations as described above were calculated

106 A male or female family history and risk of VTE

for the female relatives with a female proband with hormonally-related VTE (VTE during COC-use or pregnancy) versus female relatives with a female proband without hormonally-related VTE. Continuous variables were expressed as mean values and standard deviation (SD) or median values and range, and categorical data as counts and percentages. A two-sided p-value of less than 0.05 indicated statistical significance. Analyses were performed using SAS software, version 9.4 (SAS-Institute-Inc., USA).

Family cohorts of affected probands with VTE

Affected probands Protein C-S antithrombin Factor V Leiden Prothrombin Hyperhomocysteinaemia High factor VIII 91 271 109 103 156

All relatives 15 years 4948

Deceased 1057

Non-responders 1412

Relatives enrolled 2479

Male relatives 1249

Female relatives with antithrombin, 38 protein C, or protein S deficiency

Female relatives eligible 1192

Female relatives with incomplete thrombophilic tests 187

Female relatives evaluable for analysis 1005

Figure 1. Recruitment of probands with VTE and first-degree female relatives

RESULTS Clinical characteristics of probands and their female relatives The original cohorts consisted of 730 thrombophilic patients (probands) with VTE. The flow chart, presented in Figure 1, displays the number of relatives of

107 Chapter 5

these probands who could not be enrolled, which included non-responders (no consent, geographical distance) and relatives who were deceased. In total, 1230 relatives were female. After excluding female relatives in whom inheritance of the index defect could not be established, and those who had incomplete thrombophilia testing, 1005 female relatives were available for analysis, originating from 465 probands.

The upper part of Table 1 presents the characteristics of the probands, divided into male and female probands of whom the latter group is divided further into female probands with or without a hormonally-related VTE. Characteristics of the probands in these three groups were comparable with regard to hormonal exposure (COC-use and pregnancies), but probands with hormonally-related VTE had their first VTE at considerably younger median age, i.e. 26.5 years, than female probands without hormonally-related VTE (44.0 years) and male probands (45.0 years).

The lower part of Table 1 presents the characteristics of the 1005 female relatives of these probands, of whom 604 originated from a female proband, and 401 from a male proband. Of the 604 relatives of female probands, 394 originated from female probands with hormonally-related VTE and 206 from female probands without hormonally-related VTE. For 4 relatives (who experienced no VTE during follow-up), the female probands’ hormonal exposure at the time of VTE was unknown. The characteristics did not differ strongly between relatives of male proband and the two female proband groups. However, in the relatives suffering from a VTE, a median duration of COC-use of 0.46 years up to time of VTE was considerably shorter in relatives of female probands with hormonally-related VTE than in female relatives of female probands without hormonally related VTE (5.3 years) and female relatives of male probands (4.0 years).

108 A male or female family history and risk of VTE

Table 1. Clinical characteristics of male and female probands with or without hormonally-related VTE and their 1005 1st degree female relatives

Male All Female‡ Female‡ proband Female proband proband probands with without hormonally- hormonally- related related VTE VTE Probands, n 179 286 195 91

Ever COC-users, % NA 66.0 85.6 68.1 Ever pregnant, % NA 70.2 68.0 76.9 Ever pregnant + ever COC-use % NA 56.1 56.0 56.2 Median age at time of VTE 46.1 31.9 26.5 44.0

Female relatives, n 401 604 394 206

Thrombophilia 45% (45) 306 (52) 198 (50) 70 (34) Mild thrombophilic defects, n (%) 111 (28) 209 (35) 113 (29) 48 (23) Severe thrombophilic defects†, n (%) 68 (17) 107 (18) 85 (22) 22 (11)

Total follow-up, y* 9.757 14.438 9.014 5.299

Ever COC users, n (%) 285 (71) 428 (71) 280 (71) 145 (70) Cumulative exposure of COC use, y 2173 3340 1941 1360 Duration of use, y 3.56 3.26 2.76 4.78

Ever pregnant, n (%) 283 (71) 407 (67) 258 (66) 145 (70) Cumulative pregnancy time, y 759 1073 687 37 Pregnancy time, y 2.0 2.0 1.9 2.0

VTE, n* 38 46 39 7 VTE without hormonal exposure 9 6 5 1 VTE during COC-use 15 (40) 19 (41) 15 4 • Duration COC-use, y 4.0 - 0.46 5.3 VTE during pregnancy/post-partum 14 (37) 21 (46) 19 2

Abbreviations: COC, combined oral contraceptive; NA, data not applicable; VTE, venous thromboembolism. Data are given as median unless otherwise indicated. ‡, In one female patient information on exposure hormonal exposure at time of VTE was unknown, therefore her 4 relatives (who experienced no VTE during follow-up) were excluded. * Restricted to time between age 15 and 50 years. † Including 47 (4.7%) women with homozygosity for FVL or prothrombin G20210A or heterozygosity)

109 Chapter 5

VTE risk in female relatives of female versus male probands

Crude absolute VTE risk was 0.35 (95% confidence interval [CI]:0.28-0.43) per 100 person-years. As shown in Table 2, the risk in relatives of female (0.32, 95% CI:0.23-0.43) versus male probands (0.39, 95% CI:0.28-0.53) was comparable, but young age (<45 years at time of 1st VTE) of the proband, and presence of thrombophilia, COC-use and pregnancy of the relative were factors which significantly increased risk of VTE. In the multivariable analysis, thrombophilia, COC-use and pregnancy of relative, but not proband’s age, remained independent risk factors.

Table 2. Crude and Adjusted Hazard ratios in female relatives of all probands

Parameter Crude HR Adjusted HR Proband factors Female proband vs. male proband 0.8 (0.5-1.3, p=0.37) - Proband aged < 45 years vs. ≥ 45 years 1.8 (1.1-3.0, p=0.023) - Relative factors Mild thrombophilia vs. no thrombophilia 3.2 (1.8 - 5.6) 3.4 (1.9 - 5.9) Severe thrombophilia vs. no thrombophilia 10.4 (5.9 - 18.4) 11.2 (6.3 - 19.9) COC-use vs. no COC-use 1.9 (1.2 - 3.0) 3.6 (2.0 - 5.1) Pregnancy vs. no pregnancy 7.4 (3.1 - 13.6) 9.0 (5.5 - 14.7)

HR, hazard ratio; COC, combined oral conceptive

However, the heterogeneity analysis of risk estimates for relatives with a male versus female probands showed a substantially higher pregnancy-related risk for VTE (HR 11.6 [95% CI: 5.9-22.7]) in relatives of female probands than in relatives of male probands (HR 6.6 [95% CI: 3.1-13.6]), although the formal test for interaction between pregnancy and sex of proband was not significant (p=0.34). For COC-use, the difference in risk for VTE was less pronounced, i.e. HR of 3.6 (95% CI: 1.9-6.7) in relatives of female probands versus HR 2.7 (95% CI: 1.4-5.3) in relatives of male probands (Table 3a).

110 A male or female family history and risk of VTE

Table 3a. Adjusted Hazard Ratio of mild or severe thrombophilia, COC-use and pregnancy in female relatives of male or female probands Parameter Relatives of male probands Relatives of female probands Mild thrombophilia vs no thrombophilia 3.2 (1.3 - 7.7) 3.4 (1.6 - 7.2) Severe thrombophilia vs no thrombophilia 13.0 (5.6 - 30.6) 10.2 (4.7 - 22.2) COC-use vs no COC-use 2.7 (1.4 - 5.3) 3.6 (1.9 - 6.7) Pregnancy vs no pregnancy 6.6 (3.1 - 13.6) 11.6 (5.9 - 22.7) COC, combined oral contraceptive

VTE risk in female relatives of female probands with or without hormonally-related VTE

When taking hormonal exposure of the proband at the time of VTE into account, the incidence rate of VTE was significantly higher, i.e. 0.43 (95% CI:0.31-0.59) per 100 person-years in relatives of probands with hormonally-related VTE, than noted in relatives of female probands without hormonally-related VTE (0.13 (0.05-0.27) (HR= 3.1, 95% CI:1.3-7.4; p = 0.011).

When considering only these relatives with a female proband, the noted heterogeneity with the higher hazard ratios of pregnancy and to a lesser extent COC-use (Table 3a) is mainly observed in the relatives with a proband with a hormonally- related VTE (see Table 3b). The risk estimates in the relatives with a female proband without hormonally-related VTE risk estimates were substantially lower.

Table 3b. Adjusted Hazard Ratio of mild or severe thrombophilia, COC-use and pregnancy in female relatives of female probands with or without hormonal-related VTE

Parameter Relatives of female Relatives of female probands with probands without hormonally-related VTE hormonally-related VTE Mild thrombophilia vs no thrombophilia 3.7 (1.5 – 9.0) 1.5 (0.3-8.2) Severe thrombophilia vs no thrombophilia 10.5 (4.2-26.0) 4.0 (0.4-38.4) COC-use vs no COC-use 3.3 (1.7-6.7) 4.6 (0.92-22.6) Pregnancy vs no pregnancy 12.2 (5.8-25.6) 7.3 (1.2-44.8) COC, combined oral contraceptive

111 Chapter 5

DISCUSSION

In this large cohort study of first-degree female relatives of thrombophilic patients (probands) with VTE, which evaluated possible gender- and hormone specific aspects in family history, the overall absolute risk of VTE was not significantly different between female relatives of male or female patients. However, extended Cox regression analyses showed that the risk associated with pregnancy- postpartum was almost double in female relatives of female patients (HR, 11.6) compared to female relatives of male patients (HR, 6.6), although not significant (p=0.34). When considering only these relatives with a female proband, the observed heterogeneity with the higher hazard ratios of pregnancy and to a lesser extent COC-use in relatives of female patients is mainly observed in the relatives of patients with a hormonally-related VTE. This difference is also reflected by the 3-fold higher absolute risk in female relatives of female patients with hormonally- related VTE (HR, 3.1, 95% CI:1.3-7.4; p = 0.011). The observed heightened risk suggests that female relatives of patients with hormonally-related VTE are more susceptible to hormonal exposure.

Several studies have explored the impact of positive family history and it is considered an independent risk factor of VTE with reported odds ratios varying between 2.2-2.7.4,10,11 Additionally, two studies have reported the VTE risk of a positive family history as higher in female relatives during fertile age than in male relatives of that age.12,13

To our knowledge, only one study has evaluated sex of patient as one of the potential predictors of VTE risk in their first-degree relatives. The authors reported highest VTE risk in relatives of patients with unprovoked VTE and in relatives of patients who had VTE at younger age (below 45 years), whereas the sex of patient was no significant risk indicator for risk of VTE in their relatives (OR: 0.96 [95% CI: 0.67-1.38]).14 However, in this study those patients with estrogen-associated VTE’s were classified as having unprovoked VTE. It could well be that the outcome in this study that the highest risk in relatives of young patients with VTE (age below 45 years) is partly explained by a heightened risk in female relatives of young patients with hormonally (estrogen)-related VTE’s, as observed in our study. In our study, we also evaluated the influence of patient's age at time of first VTE and observed an overall increased risk of VTE in female relatives of younger

112 A male or female family history and risk of VTE

patients (first VTE at age below 45 years), but in multivariable analyses, influence of age was no longer an independent risk indicator. Although the authors of this study indicate that it is possible that inclusion of patients with estrogen-associated VTE among the patients with unprovoked VTE could have diluted the comparison of VTE risk between first-degree relatives of patients with unprovoked versus provoked VTE, they were unable to assess this directly. Thus, no previous studies with a similar objective are available for comparison. Our study has its limitations as discussed previously in the separate publications based on these family cohorts.1-3,7,8 With the retrospective design, not all events were established by objective techniques, because these were not yet available at the time. Consequently, the reported absolute risk of VTE may have been overestimated. Although we a priori defined subgroup analyses, the main limitation is that results of our analyses are based on small numbers of VTE. Strong points of our pooled family cohort are the inclusion of consecutive patients with symptomatic VTE, and the inclusion of their female relatives who all have confirmed positive family history. Further, extensive efforts made to minimize the recall bias on hormonal exposure resulted in an almost complete data collection on exposure to hormonal factors during reproductive age, both in female relatives and patients. Moreover, due to the cohort design we were able to estimate the absolute risk of VTE.

In conclusion, this is the first analysis taking into account a positive family history, based on a hormonally-related VTE. Although further research is needed, our data suggest that a positive family history originating from a female patient, especially when that patient experienced a COC or pregnancy-related VTE may further increase VTE risk in her female relatives. This information could be important in the counseling women on contraceptive choices.

113 Chapter 5

REFERENCES

1. Middeldorp S, Henkens CMA, Koopman MW, Hamulyák K, van der Meer, Büller HR. The incidence of venous thromboembolism in family members of patients with factor V Leiden mutation and venous thrombosis. Ann Intern Med 1998;128:15-20.

2. Bank I, Libourel EJ, Middeldorp S, Pampus ECM, Koopman MMW, Hamulyák K, Prins MH, van der Meer J, Büller HR. Prothrombin 20210A mutation: a mild risk factor for venous thromboembolism but not for arterial thrombotic disease and pregnancy-related complications in a family study. Arch Intern Med 2004;164:1932-1937.

3. Brouwer JLP, Veeger NJGM, Kluin-Nelemans HC, van der Meer J. Multicausal pathogenesis of venous thromboembolism. Evidence from a study in families with hereditary deficiencies of protein C, protein S, or antithrombin. Ann Intern Med 2006;145:807-815.

4. Bezemer ID, van der Meer FJM, Eikenboom JCJ, Rosendaal FR, Doggen CJM. The value of family history as a risk indicator for venous thrombosis. Arch Intern Med 2009;169:610-615.

5. Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet 1999;353:1167-1173.

6. Roach REJ, Lijfering WM, Rosendaal FR, Cannegieter SC, Cessie S. Sex difference in risk of second but not of first venous thrombosis. Paradox explained. Circulation 2014;129:51-56.

7. van Vlijmen EFW, Brouwer JLP, Veeger NJGM, Eskes TKAB, de Graeff PA, van der Meer J. Oral contraceptives and the absolute risk of venous thromboembolism in women with single or multiple thrombophilic defects. Arch Intern Med 2007;167:282-289.

114 A male or female family history and risk of VTE

8. van Vlijmen EFW, Veeger NJGM, Middeldorp S, Hamulyák K, Prins MH, Büller HR, Meijer K. Thrombotic risk during oral contraceptive use and pregnancy in women with factor V Leiden or prothrombin mutation: a rational approach to contraception. Blood 2011;118:2055-2061.

9. Folkeringa N, Brouwer JLP, Korteweg F, Veeger NJGM, Erwich JJHM, van der Meer J. High risk of pregnancy-related venous thromboembolism in women with multiple thrombophilic defects. Br J Haematol 2007;138:110-116.

10. Noboa S, Le Gal, Lacut K, Mercier B, Leroyer C, Nowak E, Mottier D, Oger E: for the EDITH Collaborative Study Group. Family history as a risk factor for venous thromboembolism. Thromb Res 2008;122:624-629.

11. Dowling NF, Austin H, Diley A, Whitsett C, Evatt BL, Hooper WC. The epide miology of venous thromboembolism in Caucasians and African-Americans: the GATE Study. J Thromb Haemost 2003;1:80-87.

12. Silverstein MD, Heit JA, Mohr DN, Petterson TM, O'Fallon WM, Melton III LJ. Trends in incidence of deep vein thrombosis and pulmonary embolism. Arch Intern Med 1998;158:585-593.

13. Zöller B, Li X, Sundquist J, Sundquist K. Age- and Gender-Specific Familial Risks for Venous Thromboembolism. A Nationwide Epidemiological Study Based on Hospitalizations in Sweden. Circulation 2011;124:1012-1020.

14. Couturaud, F, Leroyer C, Tromeur C, Julian JA, Kahn SR, Ginsberg JS, Wells PS, Douketis JD, Mottier D, Kearon C. Factors that predict thrombosis in relatives of patients with venous thrombosis. Blood 2014;124:2124-2130.

115 116 Clinical profile and recurrence rate in women with venous thromboembolism during combined hormonal contraceptive use: results from a prospective cohort study

Elizabeth F. W. van Vlijmen Anja B. U. Mäkelburg H. Marieke Knol Vladimir I. G. V. Tichelaar Hanneke C. Kluin-Nelemans Nic J. G. M. Veeger Karina Meijer

British Journal of Hematology 2015 DOI: 10.1111/bjh.13534 (adapted version) Chapter 6

ABSTRACT

Background Combined hormonal contraceptives (CHCs) increase risk of venous thromboembolism (VTE). However, it is yet not fully understood who will develop CHC- associated VTE and evidence on recurrence risk varies.

Methods We set up a prospective cohort of consecutive Dutch women with CHC- associated VTE that, according to Dutch General Practitioner (GP) guidelines, was preferentially prescribed levonorgestrel-containing CHCs since 1998, to describe clinical characteristics and post-VTE contraception choices, and to prospectively assess VTE recurrence.

Results The cohort included 125 women. The majority of VTE were deep vein thrombosis and pulmonary embolism. Median age was 29 years and 69/125 (55%) used levonorgestrel-containing CHCs at time of VTE. In 63%, CHCs were already used for more than 1 year. Prior to the CHC-associated VTE, 68% reported previous periods of uneventful CHC-use and/or pregnancies. Many had additional risk factors: thrombophilia (34%), first-degree family history (31%), obesity (24%), and temporary risks (14%, postpartum, trauma, surgery, and immobilization). Post-VTE, all women had discontinued CHC-use and received thromboprophylaxis in subsequent pregnancies (20%). VTE recurred in 5 women, with an annual recurrence rate of 1.2% (95% CI: 0.44-2.63), and 1.8% (95% CI: 0.30-5.90) in the first year.

Conclusions Our findings did not show any clear pattern in risk factors in this random sample of women with CHC-associated VTE. The only avoidable risk factor seemed to be a positive family history. A surprising 31% of women had a first-degree family member with VTE, which currently is not considered a contra-indication to CHC use. Recurrence was low when CHC use was discontinued and thromboprophylaxis was offered in pregnancies, but larger studies are needed to confirm this.

118 VTE during oral contraceptive use: clinical profile and VTE recurrence rate

INTRODUCTION

Combined hormonal contraceptives (CHCs) increase the risk of venous thromboembolism (VTE). The excess risk of VTE is reported to be highest during the first year on initial use of CHCs.1,2 Recent data suggest that restarting CHC after a break also introduces a temporarily higher risk.3 The lowest VTE risk is observed during use of CHCs containing ethinylestradiol combined with levonorgestrel, norgestimate (2nd-generation CHCs), or norethisterone (1st-generation CHCs), whereas highest risk is observed with CHCs containing desogestrel, gestodene, drospirenone, cyproterone, etonogestrel (vaginal ring) and norelgestromin (patch) (3rd- or 4th-generation CHCs).1-3 Based on these differences, in Europe it is generally recommended to prescribe second-generation CHCs to women initially starting CHC-use. It is not yet fully understood who will develop a CHC associated VTE. Hereditary thrombophilia increases risk of CHC-associated VTE.4,5 Additionally, first-degree family history, even in the absence of thrombophilia6 increases risk of VTE, but studies evaluating interaction between family history and CHC use are contradictory. The presence of other risk factors additionally increases risk of VTE, such as surgery, immobilization, postpartum period and obesity. Many of these risk factors are frequently present in CHC users, but few women actually develop a CHC-associated VTE.

METHODS We set up a prospective cohort of consecutive Dutch women with CHC- associated VTE that, according to Dutch General Practitioner (GP) guidelines, was preferentially prescribed levonorgestrel-containing CHCs since 1998,7 to describe clinical characteristics and post-VTE contraception choices, and to prospectively assess VTE recurrence. We included all consecutive women with CHC- associated first VTE, referred to our service between 2005 and 2010. Information on additional risk factors (surgery, trauma, immobilization, postpartum period, comorbidity, first-degree family history, obstetric history) was collected at time of VTE presentation. VTE was objectively diagnosed and thrombophilia testing was performed according to local protocol. Counselling on alternative contraception was given, which discouraged continued CHC use and thromboprophylaxis was offered in subsequent pregnancies. All patients gave informed consent for medical record access and participation in telephone interviews to follow-up on contraception changes and verifying that no recurrence was missed.

119 Chapter 6

Information was completed by medical record review and GP contact. We prospectively estimated absolute risk of recurrent VTE, expressed as recurrence rate per 100 person-years with 95% confidence intervals (CI). Observation time was defined as time since anticoagulant treatment completion until VTE recurrence or date of final telephone interview conducted in the second half of 2012.

Table I. Clinical Characteristics of women with CHC-associated VTE at time of inclusion

Variable* Median age (range) at time of VTE, years 29 (15-53) Type of VTE DVT (67% left leg) 54 (43) PE (including PE + DVT) 53 (42) Cerebral sinus thrombosis 11 (9) Calf vein thrombosis 4 (3) Other location (right arm, mesenteric vein, portal vein) 3 (2) Family history First-degree family history 39 (31) Hereditary thrombophilia F5 R506Q (Factor V Leiden) mutation 36 (26) † F2 (Prothrombin) G20210A mutation 8 (6) Protein C deficiency 2 (2) Protein S deficiency 1 (1) Antithrombin deficiency 1 (1) CHC-use Duration (range) of current CHC-use at time of VTE, months 36 (1-300) Women with previous use of CHCs 73 (58) Duration (range) of previous CHC-use, months 36 (1-312) Type of CHC at time of VTE Levonorgestrel-containing CHC (2nd generation) 69 (55.2) Desogestrel/gestodene-containing CHC (3rd generation) 35 (28.0) Drospirenone-containing CHC (4th generation) 8 (2.4) Cyproterone-containing CHC (4th generation) 3 (2.4) Norethisterone-containing CHC (1st generation) 2 (1.6) Combined hormonal contraceptive ring/patch 3 (2.4) Unknown 5 (4.0) Pregnancy Women with pregnancies prior to VTE 52 (42) Number of pregnancies prior to VTE, n (range) 112 (1-6)

CHC, combined hormonal contraceptive; VTE, venous thromboembolism; PE, pulmonary embolism; mo, month. *Unless otherwise indicated, data are expressed as number (percentage) of patients. †Including 3 homozygotes.

120 VTE during oral contraceptive use: clinical profile and VTE recurrence rate

RESULTS

This cohort included 125 women (Table 1). The majority of VTE were deep venous thrombosis (43%) and pulmonary embolism (42%). Median age was 29 (15–53) years and 55% (69/125) were using levonorgestrel-containing CHCs at time of VTE; 63% of women had used this CHC for more than 1 year. Prior to the CHC- associated VTE, 68% reported previous periods of uneventful CHC use and/or pregnancies. Many had additional risk factors: thrombophilia (34%), first-degree family history (31%), obesity (24%) and temporary risks (14%, postpartum/ trauma/surgery/immobilization). Post-VTE, all women discontinued CHC use, except one woman with endometrio- sis who also continued anticoagulant treatment. However, 53 (42%) women used no further contraception, including 20% because of a pregnancy wish. In women who restarted contraception, levonorgestrel-containing intra-uterine device was the most popular (28%), followed by injectable medroxyprogesterone (10%) (Table 2).

Table 2. Alternative contraception choices after CHC-associated VTE

Contraception method n (%) No contraception 53 (42)* Levonorgestrel-containing IUD 35 (28) Medroxyprogesterone 3-month injections 13 (10) Condom 10 (8) Oral desogestrel 5 (4) Copper-IUD 4 (3) Etonogestrel-containing implant 2 (2) Tubal ligation 1 (1) Unknown 2 (2)

CHC, combined hormonal contraceptive; VTE, venous thromboembolism; IUD, intra-uterine device *Including 25 women who wished to become pregnant

Follow-up was available for 123 women; two were lost to follow-up after anticoagulant treatment completion. Most women used vitamin-K-antagonists for 6 months, but 16% received rivaroxaban in an ongoing study setting. Post-VTE, eight (6%) women received long-term anticoagulant therapy, among others due

121 Chapter 6

to underlying conditions, and therefore did not contribute data to follow-up. During follow-up, 25 women (20%) became pregnant; all received prophylactic or therapeutic low molecular weight heparin up to delivery and 6 weeks postpartum. Median follow-up after anticoagulant therapy completion was 37 months (range 7–98 months). VTE recurred in five (4.1%) women, with an annual recurrence rate of 1.2% (95% CI: 0.45–2.70). In the first year this was 1.8% (95% CI: 0.30–5.90). All recurrences were idiopathic. Concomitant morbidity was present in two women (rheumatoid arthritis, hypothyroidism).

DISCUSSION

We described the clinical characteristics in a cohort of women with CHC- associated VTE, who preferentially used 2nd-generation CHCs. We observed that 68% of women had one or more additional risk factors besides CHC-use, in particular a high prevalence of thrombophilia, positive family history, and obesity. Regardless of the high prevalence of these risk factors, only 37% of women had the VTE event within the first year of starting CHC-use. Additionally, many women had previous periods of CHC-use and/or pregnancies without any thrombotic complications prior to the occurrence of the CHC-associated VTE. This suggests that previous uncomplicated hormonal exposure may not preclude the occurrence of a CHC-associated VTE, neither during long-term ongoing CHC-use nor when starting a new period of CHC-use at a later time point in reproductive life. Prospective assessment of their risk of VTE recurrence after a CHC-associated VTE showed low risk with subsequent risk management, i.e. discontinuation of CHC and prophylaxis in subsequent pregnancies.

Several comparative studies evaluated the risk of recurrence in women with CHC- associated VTE.9-11 However, these studies often included these women as a subgroup, or CHC exposure was combined with exposure to pregnancy and/or hormone replacement, or studies were restricted to women with CHC use as single risk factor. Our cohort study presents an annual recurrence rate of 1.2% and 1.8% in the first year post index-VTE in women without further CHC use. Indirect comparison suggests that this rate compares favorably to that observed in previous studies (1.8–5.6%).

122 VTE during oral contraceptive use: clinical profile and VTE recurrence rate

The strong points of our study include the capture of all consecutive women with CHC-associated first VTE referred to our service between 2005 and 2010, its high degree of detail, almost complete thrombophilia testing (94%) and minimal loss to follow-up.

Although a comparator group of women of fertile age with idiopathic VTE was considered, this did not appear feasible due to the very low number of women of fertile age presenting with idiopathic VTE, which is also related with the high prevalence of CHC-use in the Netherlands (61% between 18 and 30 years of age). However, studies reporting a recurrence rate after idiopathic VTE indicated a far higher recurrence risk, i.e. 15–17%,9,11 than noted after a CHC-associated VTE with subsequent risk management, i.e. discontinuation of CHC. The design of our study does not enable any prediction on who will develop CHC- associated VTE, and we did not see any clear pattern. In this random sample of women with CHC-associated VTE, the majority had already used their present CHC for more than 1 year, had previous uneventful CHC use and/or pregnancies even in the presence of other risk factors and was on the preferred second- generation CHC. The only avoidable risk factor seemed to be a positive family history. A surprising 31% of women had a first-degree family member with VTE, which currently is not considered a contra-indication to CHC use. Recurrence was low when CHC use was discontinued and thromboprophylaxis was offered in pregnancies, but larger studies are needed to confirm this.

123 Chapter 6

REFERENCES

1. van Hylckama V, Helmerhost FM, Vandenbroucke JP, Doggen CJ, Rosendaal FR. The venous thrombotic risk of oral contraceptives, effects of oestrogen dose and progestogen type: results of the MEGA case-control study. BMJ 2009;339:b2921.

2. Lidegaard Ø, Løkkegaard E, Svendsen AL, Agger C. Hormonal contraception and risk of venous thromboembolism: national follow-up study. BMJ 2009;339:b2890.

3. Lidegaard Ø, Nielsen LH, Skovlund CW, Skjeldestad FE, Løkkegaard E. Risk of venous thromboembolism from use of oral contraceptives containing different progestogens and oestrogen doses: Danish cohort study, 2001-9. BMJ 2011;343:d6423.

4. Faculty of Sexual & Reproductive Healthcare. Statement on Venous throm boembolism and hormonal contraception, November 2014. Available at: http://www.fsrh.org/pdfs/FSRHStatementVTEandHormonalContraception.pdf

5. Van Vlijmen EFW, Brouwer JLP, Veeger NJGM, Eskes TKAB, de Graeff PA, van der Meer J. Oral contraceptives and the absolute risk of venous thromboembolism in women with single or multiple thrombophilic defects. Arch Intern Med 2007;167:282-289.

6. Van Vlijmen EFW, Veeger NJGM, Middeldorp S, Hamulyák K, Prins MH, Büller HR, Meijer K. Thrombotic risk during oral contraceptive use and pregnancy in women with factor V Leiden or prothrombin mutation: a rational approach to contraception. Blood 2011;118,2055-2061.

7. Bezemer ID, van der Meer FJ, Eikenboom JC, Rosendaal FR, Doggen JCM. The value of family history as a risk indicator for venous thrombosis. Arch Intern Med 2009;169:610-615.

124 VTE during oral contraceptive use: clinical profile and VTE recurrence rate

8. Sips AJBI, Beijderwellen L, Kardolus GJ, Burgers JS. Dutch General Practitioner’s Standard on hormonal contraception, first revision (Nederlands Huisartsen Genootschap-Standaard Hormonale Anticonceptie (eersterevisie). Huisarts Wet 1998;41:385-394.

9. Cushman M, Glynn RJ, Goldhaber SZ, Moll S, Bauer KA, Deitcher S, Shrivas tava S, Ridker PM. Hormonal factors and risk of recurrent venous thrombosis: the Prevention of Recurrent Venous Thromboembolism trial. J Thromb Haemost 2006;4:2199-2203.

10. Le Gal G, Kovacs MJ, Carrier M, Do K, Kahn SR, Wells PS, Anderson DA, Chag non I, Solymoss S, Crowther M, Righini M, Lacut K, White RH, Vickars L, Rodger M. Risk of recurrent venous thromboembolism after a first oestrogen- associated episode. Data from the REVERSE cohort study. Thromb Haemost 2010;104:498-503.

11. Eischer L, Eichinger S, Kyrle PA. The risk of recurrence in women with venous thromboembolism while using estrogens: a prospective cohort study. J Thromb Haemost 2014;12:635-640.

125 126 Summary, discussion and future perspectives Chapter 7

Summary

Chapter 1 presents the history, mechanism of action, and efficacy of combined oral contraceptives (COCs) and the association of its use to the rare but serious adverse event of venous thromboembolism (VTE). As VTE is seen as a multicausal disease, also the contribution of other risk factors to the risk of VTE in COC- users, including thrombophilia and family history, is discussed. Further, a comparison is made with the risk of VTE in women during pregnancy, a condition, which is hormonally comparable. This chapter ends with the outline of this thesis and the rationale for initiating the different studies that are part of this thesis.

Hereditary thrombophilia and the risk of venous thromboembolism during combined hormonal contraceptive use

Severe hereditary thrombophilia (protein S-, protein C-, or antithrombin deficiency)

Information on the risk of VTE in COC-users who have severe thrombophilic defects, i.e. protein S-, protein C- or prothrombindeficiency, is very limited, whereas this risk in women in combination with other thrombophilic defects is hardly considered in any study. In Chapter 2 a retrospective cohort study was described, in which we assessed the effects of COC-use on the absolute risk of VTE in protein S, protein C, or antithrombin-deficient women and the contribution of other known thrombophilic defects. From a large family cohort of protein S-, protein C-, and antithrombin-deficient families, we enrolled female relatives of fertile age and male relatives in the same group who were used as a reference group. In total, 222 women were analyzed, of whom 101 (45%) were deficient. Fifty-six deficient and 79 non-deficient women had ever used COCs.

Absolute risk of VTE in protein S–, protein C–, and antithrombin-deficient women was high in comparison to non-deficient female relatives, as shown by the incidence of VTE of 1.64 versus 0.18 per 100 person years in deficient versus non- deficient women and the adjusted RR was 11.9 (95% confidence interval (CI), 3.9-36.2). COC-use hardly contributed to the overall risk of VTE during fertile lifetime: cumulative event rates at from the age of 15 up to 50 years were similar in women, who ever- or never used COCs, and in men. However, event-free survival

128 Summary, discussion, and future perspectives

curves showed VTE to occur earlier in deficient ever-users and, although less pronounced, in deficient never-users than in deficient men. We hypothesized that these differences between women and men are related to COC-use in deficient ever-users and pregnancy in deficient never-users. Therefore, both COC-use and pregnancy resulted in the occurrence of VTE in deficient women at a younger age, but did not increase the overall risk of VTE during fertile lifetime.

The absolute risk of VTE in deficient women during COC-use was almost 10-fold higher than in non-deficient COC-users, as demonstrated by an incidence rate of 4.62 versus 0.48 per 100 pill-years of use. This risk is almost 100-fold higher than the most recently published absolute risk noted in the general population of COC-users, in whom the incidence rate according to the latest evidence varies between 0.05-0.09 per 100 pill-years of use.1-3 The 10-fold higher risk observed in non-deficient COC-users in this family cohort is explained by the presence of other thrombophilic defects, such as presence of factor V Leiden or prothrombin 20210A mutation or an increased factor VIII: the incidence rate during actual COC-use increased by concomitant thrombophilic defects, from 3.49 to 12.00 per 100 pill-years in deficient women and from 0 to 3.13 per 100 pill-years in non-deficient women. Based on these results, it is concluded that women with hereditary deficiencies of protein S, protein C, or antithrombin are at high risk of VTE during use of COCs, particularly when other thrombophilic defects are present. Therefore, we recommend that the use of COCs in these women is strongly discouraged.

Mild hereditary thrombophilia (factor V Leiden or prothrombin 20210A mutation)

Current contraception guidelines discourage COC-use in women with hereditary thrombophilic defects.4 However, qualifying all hereditary thrombophilic defects as similarly strong risk factors might be questioned. Several studies indicated the risk of VTE associated with a factor V Leiden and prothrombin 20210A mutation as considerably lower than a deficiency of protein C, protein S, or antithrombin.5,6 In view of these differences, withholding COCs in women with these mild hereditary thrombophilic defects might be less favorable, as pregnancy and especially the postpartum period is a stronger risk factor for VTE than is COC-use. There- fore, when discouraging COC use, an increased risk of unintended pregnancy must be taken into account, as alternative contraception may be less reliable.

129 Summary, discussion, and future perspectives

Hence, to adequately balance the risk and benefits of COCs, reliable estimates of the VTE risk associated with both COC-use and pregnancy are needed for women with mild thrombophilic disorders. In Chapter 3 a retrospective family cohort was presented, in which the VTE risk during COC-use and pregnancy (including postpartum) was assessed in 798 female relatives with or without a heterozygous, double heterozygous, or homozygous factor V Leiden or prothrombin G20210A mutation.

Overall, absolute VTE risk in women with no, single, or combined defect was 0.13 (95% CI: 0.08-0.21), 0.35 (95% CI: 0.22-0.53), and 0.94 (95% CI: 0.47-1.67) per 100 person-years, while these were 0.19 (95% CI: 0.07-0.41), 0.49 (95% CI: 0.18- 1.07), and 0.86 (95% CI: 0.10-3.11) during COC-use, and 0.73 (95% CI: 0.30-1.51), 1.97 (95% CI: 0.94-3.63), and 7.65 (95% CI: 3.08- 15.76) during pregnancy. COC- use and pregnancy were independent risk factors for VTE, with the highest risk during the postpartum period, as demonstrated by adjusted Hazard Ratios (HR) of 16.0 (95% CI: 8.0-32.2) versus HR of 2.2 (95% CI: 1.1-4.0) during COC use.

Obtained absolute risks were put into perspective in a modelling exercise, in which both the risks of alternative contraceptive-related VTE and the risk of pregnancy-related VTE (resulting from contraceptive failure) were considered. Summarizing these extrapolations, both the levonorgestrel-containing (Mirena) and copper-containing intrauterine device carry a lower overall risk of VTE and have adequate contraceptive efficacy; these contraceptives were therefore considered good alternatives to COCs. Condom use has the lowest contraceptive efficacy, with an up to 60-fold increased risk of unintended pregnancy, which makes this option the least favorable alternative.

Based on these results, rather than strictly contraindicating COC-use, we recommended that detailed counseling is given on all contraceptive options, including COCs, with addressing the associated risks of both VTE and unintended pregnancy, enabling these women to make an informed choice.

Meta-analysis of studies evaluating risk of COC-use in women with thrombophilia

In chapter 4, a systematic review and meta-analysis was presented which took a clinical case presentation “A 29-year-old woman previously tested heterozygote for

130 Summary, discussion, and future perspectives

Factor V Leiden mutation (in a research setting 6 years ago) asks: what is my risk of thrombosis if I would start using combined oral contraceptives (COCs)?” as a starting point. A meta-analysis was performed to evaluate the risk of VTE in COC-users with thrombophilia. A distinction was made in ‘mild’ (factor V Leiden, prothrombin-G20210A mutation) and ‘severe’ thrombophilia (antithrombin-, protein C-, protein S-deficiency, double heterozygosity or homozygosity of factor V Leiden and prothrombin-G20210A mutation). Twelve case-control- and three cohort studies were identified.7-20

In COC-users, mild thrombophilia increased VTE risk six-fold (relative risk (RR) 5.89, 95% Confidence Interval [CI]: 4.21-8.23), while severe thrombophilia increased VTE risk seven-fold (RR 7.15, 95% CI: 2.93-17.45).

However, to adequately assess the impact of the observed relative increase in risk, information on the absolute risk is needed. One cohort study reported a VTE incidence of 0.49 and 2.0 per 100 pill-years of use in COC-users with mild thrombophilia,20 while two studies18,19 reported incidences of 4.3 and 4.6 per 100 pill-years in COC-users with severe thrombophilia, indicating a far higher risk (Table 3). The incidence of VTE in COC-users with double heterozygosity or homozygosity of factor V Leiden or prothrombin-G20210A mutation was 0.86 per 100 pill-years, suggesting that the absolute risk of this double defect is less serious than a antithrombin-, protein C-, or protein S deficiency.20 All absolute risks are estimated in family members of thrombophilic families, i.e. relatives of thrombophilic patients with VTE, and therefore also have a positive family history, which increases their baseline risk of VTE two- to three-fold.29-32 To put observed risks into perspective, in COC-users with mild thrombophilia and positive family history the absolute VTE risk is increased eight- to 33-fold, and 70-fold in COC- users with severe thrombophilia, when compared to the VTE risk of 0.06 per 100 person-years,1 estimated in the general population of COC-users.

However, it was noted that the risk of mild thrombophilia was largely estimated in a community setting, while risk of severe thrombophilia was exclusively evaluated in thrombophilic family cohorts. This is inherently due to the very low prevalence of severe thrombophilia. Further, absolute risks were all estimated in members of thrombophilic family cohort studies, risks will therefore be more pronounced because of the co-existing family history, than in the general population of COC- users tested positive.

131 Chapter 7

In conclusion, the presence of mild and severe thrombophilia increases the risk of VTE in COC-users six- to seven-fold. However, absolute VTE risk estimates indicate the contribution of severe thrombophilia to the VTE risk in COC-users as considerably higher (4.3 and 4.6 per 100 pill-years) than the additional risk of mild thrombophilia (0.49 and 2.0 per 100 pill-years), but with the caveat that these risks were estimated in thrombophilic COC-users who also had a positive family history. As a co-existing family history increases VTE risk two- to threefold,21-23 estimated risks are more pronounced than in a general population of COC-users tested positive for thrombophilia.

In the weighing of above results, we took into account the GRADE quality of evidence grading, that sound epidemiological studies as included in this meta- analysis would be graded as low evidence, but when observed effects are large and there is no obvious bias explaining those effects, the evidence may be rated as moderate or even high, as further research is very unlikely to change our confidence in the estimate of effect (http://gradeworkingroup.org). From this standpoint we propose the strong recommendation is given that based on the high additive risk, to avoid COC-use in women with severe thrombophilia in all cases. The additive VTE risk of mild thrombophilia is modest and when no other risk factors are present, e.g. family history, we recommend that COC-use should not be denied in women who consider alternative adequate contraception not acceptable, as in this situation the increased risk of pregnancy-related VTE and increased risk of pregnancy outweigh the COC-associated risk.

Family history and risk of venous thromboembolism during combined hormonal contraceptive use

As demonstrated in the studies described above (Chapter 2, 3 and 4), women from thrombophilic families have an increased risk of VTE, which increases further during COC-use and pregnancy. It is unknown whether this additional risk differs between relatives of male and female patients with VTE, and also whether it matters if the female patient had a hormonally-related VTE (during COC-use or pregnancy). To explore this question, we performed a retrospective family cohort study, which was presented in Chapter 5. In this family cohort of 1005 first-degree female relatives of reproductive age, we compared VTE risk in relatives of female versus male patients, and between relatives of female patients with and without hormonally-related VTE.

132 Summary, discussion, and future perspectives

In this cohort, VTE risk in relatives of female (0.32, 95%CI:0.23-0.43) versus male patients (0.39, 95%CI:0.28-0.53) was comparable, but young age (<45 years at time of 1st VTE) of the patient, and presence of thrombophilia, COC-use and pregnancy of the relative were factors which significantly increased risk of VTE. In the multivariable analysis, thrombophilia, COC-use and pregnancy of relative, but not patients’ age, remained independent risk factors.

Additionally, the heterogeneity analysis of risk estimates suggested that VTE risk in relatives of female versus male patients differs during pregnancy-postpartum (Hazard Ratio [HR], 11.6 [95%CI:5.9-22.7] versus 6.6 [95%CI:3.2-13.6]) and to a lesser extent during COC-use (HR, 3.6 [95%CI:1.9-6.7] versus 2.7 [95%CI:1.4-5.3]), although not statistically significant.

When considering only relatives of female patients, the noted higher HRs of pregnancy and COC-use are mainly observed in relatives of patients with hormonally- related VTE. The risk estimates in relatives of female patients without hormonally-related VTE were substantially lower.

The observed heightened risks suggest that female relatives of patients with hormonally-related VTE are more susceptible to hormonal exposure.

Several studies have explored the impact of positive family history and it is considered an independent risk factor of VTE with reported odds ratios varying between 2.2 to 2.7.21-23 Of these, two studies have reported the VTE risk of a positive family history as higher in female relatives during fertile age than in male relatives of that age.24,25 To our knowledge, only one study has evaluated sex of patient as one of the potential indicators of VTE risk in their first-degree relatives.26 They reported highest VTE risk in relatives of patients with unprovoked VTE and in relatives of patients who had VTE at younger age (below 45 years), whereas the sex of patient was no significant risk indicator for VTE (OR: 0.96 [95% CI: 0.67-1.38]).

133 Chapter 7

However, in this study those patients with estrogen-associated VTE’s were classified as having unprovoked VTE. It is well possible that the highest risk in relatives of young patients with VTE (age below 45 years) as found in this study, is partly explained by an increased risk in female relatives of patients with hormonally (estrogen)-related VTEs (who are inherently young), as observed in our study.

In our study, we also evaluated the influence of patient's age at time of first VTE and observed an overall increased risk of VTE in female relatives of younger patients (first VTE below 45 years), but in multivariable analyses, influence of patient’s age was no longer an independent risk indicator. Although the authors of this study indicate that it is possible that inclusion of patients with estrogen- associated VTE among the patients with unprovoked VTE could have diluted the comparison of VTE risk between first-degree relatives of patients with unprovoked versus provoked VTE, they were unable to assess this directly. Thus, no previous studies with a similar objective are available for comparison.

In conclusion, this is the first analysis taking into account a positive family history, based on a hormonally-related VTE. Although further research is needed, our data suggest that a positive family history originating from a female patient, especially when that patient experienced a COC or pregnancy-related VTE may further increase VTE risk in her female relatives. We recommend to consider this information in the counseling women on contraceptive choices.

Clinical profile and recurrence rate in women with venous thromboembolism during combined oral contraceptive use

COCs increase risk of venous thromboembolism (VTE). However, it is yet not fully understood who will develop COC-associated VTE and evidence on recurrence risk varies.

In order to gather more information, we set up a prospective cohort, presented in Chapter 6, of consecutive Dutch women with COC-associated VTE that, according to Dutch General Practitioner (GP) guidelines, was preferentially prescribed levonorgestrel-containing CHCs since 1998. The aim of this study was to describe clinical characteristics and post-VTE contraception choices, and to prospectively assess VTE recurrence.

134 Summary, discussion, and future perspectives

The cohort included 125 women with COC-associated VTE, of whom the majority had experienced deep vein thrombosis and pulmonary embolism. Their median age was 29 years and 69/125 (55%) used levonorgestrel-containing COCs at time of VTE. In 63%, the women had already used the COC for more than 1 year. Prior to the COC-associated VTE, 68% reported previous periods of uneventful COC- use and/or pregnancies. Many had additional risk factors: thrombophilia (34%), first-degree family history (31%), obesity (24%), and temporary risks (14%, postpartum, trauma, surgery, and immobilization). Post-VTE, all women had discontinued COC-use and received thromboprophylaxis in subsequent pregnancies (20%). Median follow-up after anticoagulant therapy completion was 37 months (range 7–98 months). VTE recurred in 5 women, with an annual recurrence rate of 1.2% (95% CI: 0.44-2.63), and 1.8% (95% CI: 0.30-5.90) in the first year.

Several comparative studies evaluated the risk of recurrence in women with CHC- associated VTE.27-29 However, these studies often included these women as a subgroup, or CHC exposure was combined with exposure to pregnancy and/or hormone replacement, or studies were restricted to women with CHC use as single risk factor. Our cohort study presents an annual recurrence rate of 1.2% and 1.8% in the first year post index-VTE in women without further CHC use. Indirect comparison suggests that this rate compares favorably to that observed in previous studies (1.8–5.6%). Although a comparator group of women of fertile age with idiopathic VTE was considered, this did not appear feasible due to the very low number of women of fertile age presenting with idiopathic VTE, which is also related with the high prevalence of CHC-use in the Netherlands (61% between 18 and 30 years of age). However, studies reporting a recurrence rate after idiopathic VTE indicated a far higher recurrence risk, i.e. 15–17%,27,29 than noted after a CHC-associated VTE with subsequent risk management, i.e. discontinuation of CHC.

The design of our study did not enable any prediction up front on who will develop CHC-associated VTE, and we did not see any clear pattern. In this random sample of women with CHC-associated VTE, the majority had already used their present CHC for more than 1 year, had previous uneventful CHC use and/or pregnancies even in the presence of other risk factors and was on the preferred second- generation CHC. The only avoidable risk factor seemed to be a positive family history.

135 Chapter 7

A surprising 31% of women had a first-degree family member with VTE, which currently is not considered a contra-indication to CHC use. Recurrence was low when CHC use was discontinued and thromboprophylaxis was offered in pregnancies, but larger studies are needed to confirm this.

Summarized, our findings did not show any clear pattern in risk factors in this random sample of women with CHC-associated VTE. The only avoidable risk factor seemed to be a positive family history. A surprising 31% of women had a first-degree family member with VTE, which currently is not considered a contra- indication to CHC use. The recurrence was low with subsequent risk management of COC-discontinuation and thromboprophylaxis in subsequent pregnancies.

Discussion and implications for future research

The studies described in this thesis show that the risk of VTE in COC-users is increased substantially in women with severe thrombophilia (protein S-, C- or antithrombin-deficiency), leading us to the recommendation that COC-use should be strongly discouraged (Chapter 2). However, in women with a factor V Leiden or a prothrombin-G20210A mutation, the absolute risk of VTE during COC use was only modestly increased, and importantly lower than the absolute risk observed during the pregnancy-postpartum period (Chapter 3). These data provide evidence that the current WHO guideline (2015)4 on medical eligibility criteria for contraceptive use, which strictly contraindicates COC use in these women, needs reconsideration. The results of the study do not allow to ‘promote’ COC use in asymptomatic family-carriers of FVL or PT G20210A mutation, but indicate that when COC use is discontinued or avoided, the need for adequate alternative contraception has high priority. Rather than strictly contraindicating COC-use, we recommend that these women should be given detailed counseling on all contraceptive options, including COCs, in which the associated risks of both VTE and unintended pregnancy and additional VTE risk factors present need to be addressed. Based the results of our meta-analysis (Chapter 4), which evaluated the VTE risk in COC-users with or without thrombophilia, we proposed recommendations regarding COC-use in women with mild and severe thrombophilia. In this, a woman tested positive for factor V Leiden in the past, who wishes to start combined oral contraception was taken as a starting point, as such situation is often a reason for consulting a hematologist.

136 Summary, discussion, and future perspectives

The results of the meta-analysis showed that in COC-users mild thrombophilia (factor V Leiden, prothrombin-G20210A mutation) and severe thrombophilia (antithrombin-, protein C- or protein S-deficiency, double heterozygosity and homozygosity of factor V Leiden or prothrombin-G20210A mutation) increased VTE risk six to seven-fold. However, to value the impact of the relative increase in risk, the increase in absolute risk needs to be taken into account. The cohort studies included in this meta-analysis reported a VTE incidence of 0.49 and 0.2 per 100 pill-years of use in COC-users with mild thrombophilia, but incidences of 4.3 and 4.6 per 100 pill-years in COC-users with severe thrombophilia, indicating a considerably higher risk. However, all absolute risks are estimated in family members of thrombophilic families, i.e. relatives of thrombophilic patients with VTE, and therefore also have a positive family history. As a co-existing family history increases VTE risk two- to threefold, the observed absolute risk estimates are more pronounced than in a general population of COC-users tested positive for thrombophilia. Taking into account the GRADE quality of evidence grading, we proposed a strong recommendation that given the high additive risk, COC-use should be avoided in women with severe thrombophilia in all cases. The additive VTE risk of mild thrombophilia is modest and when no other risk factors are present, e.g. no family history, COC-use should not be denied in women who consider alternative adequate contraception not acceptable, as in this situation the increased risk of pregnancy-related VTE and increased risk of pregnancy outweigh the COC-associated risk. The VTE risk during pregnancy in women with mild thrombophilia (heterozygote for factor V Leiden or prothrombin G20210A mutation) was estimated (Chapter 3) to be as high as 1.97 per 100 pill-years. In this study, also absolute risk of VTE in COC-users versus users of alternative contraceptive methods, i.e. LNG-IUD users, copper-IUD users, and users of male condom, were compared. In this analysis both the risks of contraceptive-related VTE and the risk of pregnancy-related VTE (resulting from contraceptive failure) were considered. In these extrapolations, the LNG-IUD and copper-IUD both carried a lower overall risk of VTE and therefore good alternatives to COCs. Condom use had the lowest contraceptive efficacy, with an up to 60-fold increased risk of unintended pregnancy, which makes this option the least favorable alternative contraceptive method. As shown in these two studies (Chapter 2, 3) and meta-analysis (Chapter 4) discussed above, women from thrombophilic families have increased risk of venous thromboembolism (VTE), which increases further during hormonal exposure.

137 Chapter 7

Our cohort study, which evaluated whether this family-conferred risk is gender- or hormone-specific (Chapter 5), suggested that it does matter when a positive family history is originating from a female patient, especially whether the patient experienced a VTE during COC-use or during pregnancy or the post-partum period. Regarding possible clinical implications, we recommend to consider that women could be at higher VTE risk during COC-use and during pregnancy if they have a mother or sister with VTE, especially if that VTE also occurred during COC- use or in the pregnancy or postpartum period. Therefore, reviewing the family history prior to prescribing COCs should include specific data on who experienced VTE under what circumstances.

Finally, detailed information on clinical characteristics is presented, collected in a cohort of 125 consecutive women who had experienced a VTE while using a COC that we followed to estimate recurrence rate post COC-associated VTE. Detailed information on the clinical characteristics of these women was collected (Chapter 6). In 68% of women additional VTE risk factors besides COC-use were present, of which thrombophilia, family history and obesity reported in a high frequency. The design of this study did not permit any conclusion to predict up front on who will develop COC-associated VTE, and we did not see any clear pattern. However, this is in line with the concept of VTE being a multicausal disease: besides COC- use a variety of additional risk factors in a high frequency were concomitantly present in these women and that it can take years before a combination of these risks at one point in time results in a symptomatic VTE. This is supported by the observation that in this sample of consecutive women with COC-associated VTE, two-third of women already had used their COC for more than 1 year, had used COCs before and also had pregnancies without any complication. As in clinical practice routine screening for thrombophilia is considered not useful prior to start of COC-use,30 the only ‘avoidable’ risk factor seems to be a positive family history. A surprisingly high percentage of women (31%) had a first-degree family member with VTE. In two-third of cases, the family history originated from the mother or sister and in most cases was related to VTE during COC-use or the pregnancy- postpartum period. In current guidelines a family history of VTE is not a strict contra-indication to COC-use. It is therefore recommended, also in view of the outcome of our study suggesting that a positive family history could also be hormone-specific, to put a detailed first-degree family history in at a prominent place in the counseling on potential risk factors to COC-use.

138 Summary, discussion, and future perspectives

Based on median follow-up after anticoagulant therapy completion of 37 months (range 7–98 months) a low annual VTE recurrence (1.2%) was observed with post-VTE risk management, which consisted of discontinuation of COC-use and thromboprophylaxis in pregnancies. Therefore, detailed counseling in the choice of adequate alternative contraception post-COC-related VTE is important.

In conclusion, there is “no other medication in the history of medical science that has been studied more extensively than the combined oral contraceptive pill’ (US Food and Drug Administration), which recently had its 50-year anniversary. Although the studies in this thesis and many other studies available gave us further detailed insight in the contribution of VTE risk factors to the VTE risk associated with the use of COCs and pregnancy, as well as new starting points for further research, the risk of COC-associated VTE is still not resolved.

Current developments in contraceptive methods Combined hormonal contraceptives

Up to today, a combined hormonal contraceptive that has no increased risk of VTE has not been developed, although science does not stand still. As for the currently approved COCs, despite a more than 10-fold decrease in dose, and development of many new progestogens, the risk of VTE, although considerably lower, persisted.

Several options have been developed or are ongoing based on the theory that the risk of VTE may be modified by the type of estrogen, the route of administration, or lowering the estrogen dose. Some new estrogens are being developed such as the fetal estrogen estretol,31 but it is as yet unclear whether these will have different effects on the risk of VTE. Further, recently two new COCs (Qlaira©, Zoely©) became available on the market, which contain estradiol instead of ethinylestradiol. However, there is yet no sufficient clinical trial information to conclude that changing ethinylestradiol into estradiol will change the risk of VTE. Published data showed that alterations regarding hemostatic variables, SHBG-levels and thrombin-generation activated protein-C resistance (APC-resistance) were in the range of that noted with levonorgestrel containing COCs, but unfortunately none of these are validated surrogates for the clinical endpoint of VTE.32-34

139 Chapter 7

As to a different route of administration, it is suggested that orally administered estrogen has higher prothrombotic effect that is possibly related to high concen- trations of estrogen in the liver due to the “first-pass” effect. However, the combined hormonal contraceptive vaginal ring (ethinylestradiol-desogestrel) and transdermal patch (ethinylestradiol-norelgestromin) were recently shown to have an increased risk of VTE that was comparable to that noted for 3rd and 4th generation COCs.36 On the other hand, a case-control study in postmenopausal women who used either oral or estradiol patches observed no increased risk of VTE in women using patches,37 and hemostatic variables might be less influenced with transdermal administration of estradiol,38 but more robust clinical trial data are lacking. There- fore, the patient leaflet of estradiol-containing patches has not been adapted regarding the risk of VTE. However, gynecological organizations recommend considering the possibly lower risk of VTE with transdermal administration of estradiol.39 Finally, there are some data in postmenopausal women suggesting that statins could potentially attenuate the estrogen-related risk of VTE, but adequate controlled studies are needed to further explore this observation.40

Non-hormonal contraceptives

The development of non-hormonal contraceptives is ongoing,41 but the number of potential candidates in the clinical phase is still limited: several clinical contraception studies with selective progestogen-receptor modulators (SPRMs) like mife- pristone, ulipristal,42 have been published, but as yet no product has been put forward as an alternative to a COC. In India, since the early 1990s, a selective estrogen-receptor modulator (SERM) - ormeloxifene – is on the market for contraception, which has to be taken once a week.43 This product is not approved for contraception in the Western world. The documentation on the effectiveness and safety of this product in the public domain is too sparse to draw any meaningful conclusion, but those SERMs that are approved in indications for postmenopausal women indicate that these products induce a 2-fold increased risk of VTE.44 Other potential candidates could be prostaglandin E2 receptor antagonists but research is still in the preclinical development phase.45

140 Summary, discussion, and future perspectives

Contraception and pregnancy in women at increased risk of VTE

As to the possibilities of reliable alternative contraception for women post-VTE, the progestogen-only preparations are considered a very good alternative. Although traditionally they are seen as contraceptives especially for women who have completed their family, these products can compete with COCs as these have excellent contraceptive efficacy and rapid return of fertility when discontinued.46,47 The amount of data supporting the progestogen-only preparations to be non-thrombotic is increasing; the largest amount of data up to now is collected on the levonorgestrel-containing IUD (Mirena®) showing that this levonorgestrel-containing intrauterine device does not increase the risk of VTE.36,48 Recently, also a slightly smaller and lowed-dosed levonorgestrel-containing IUD (Jaydess®) has been registered, but this product is not yet available in all EU countries. Also the data now available for an etonogestrel-containing implant (Implanon®), and the oral progestogen-only pill (Cerazette®) containing desogestrel do not show an increased risk of VTE.1,36,48 However, limited data indicate that the medroxyprogesterone injectable (Depo-Provera®) increases the risk of VTE.17,48 It is difficult to explain why one of the progestogen-only preparations would have an opposite safety pattern regarding risk of VTE. A possible explana- tion might be that the characteristics of the women who use this contraceptive are different, as this contraceptive may be used only when other contraceptives are unsuitable. The reasons for its restricted indication are the considerable delay in return to fertility after discontinuation of this product and the observed decrease in bone mineral density noted in adolescent women. As to recommending non- hormonal contraception, the copper-IUD is a very good alternative with adequate contraceptive properties. However, above mentioned alternative contraception methods have effects on the menstrual bleeding pattern: progestogen-only contraception, including the levonorgestrel-containing intrauterine device induce unpredictable bleeding pattern and the copper IUD will increase the duration and amount of menstrual blood loss. These adverse events could lead to discontinuation and it is recommended to discuss beforehand that such effects on the menstrual bleeding pattern can be expected and whether such adverse effects would be acceptable to the woman. Guidance on the expected bleeding pattern, which could improve with longer duration, is shown to increase continuation rates.49,50 The current options of adequate progestogen-only contraception, especially oral progestagen-only, are still limited, but several new products are currently in the clinical development phase.51

141 Chapter 7

While the hormonal (estrogenic) exposure of a COC is avoidable, this is not the case for a pregnancy. This introduces the difficult dilemma what should be done if the VTE risk pattern of the woman is not straightforward. In case of a previous history of thrombosis, the decision is clear: several guidelines recommend thromboprophylaxis during pregnancy and in the postpartum period. But what if the risk is less clear, i.e. there are known risk factors but these have not resulted in symptomatic VTE? International clinical guidelines still rely heavily on expert opinion, and there are contradictory views in these on initiation, dosing and duration of thromboprophylaxis in pregnancy and postpartum period due to the absence of adequate controlled trials. Thromboprophylaxis in pregnancy in high-risk patients is also poorly discussed in the product information of the low molecular weight heparins (LMWHs), and dose recommendations specifically for use in pregnancy are not given. More pharmacodynamic and pharmacokinetic information is needed to support an adequate dose/per day, taking into account the differences in hemostatic balance that are induced by the pregnant state. Additionally, it might be worthwhile to gather more information on the increasing estrogen levels and hemostatic changes in pregnancy and the post-partum period. More insight in this pattern could be helpful on what would be the most appropriate time point to start and stop thromboprophylaxis. A better docu-mented timing will also improve the balance between coagulation and bleeding in the critical period after the partus.

Improving safe use of combined oral contraceptives

Lastly, in view of the current situation that there is no thrombosis risk-free combined hormonal contraceptive to be expected soon, the focus should be on who can safely use COCs and who should not be prescribed a COC. Recently, the European Medicines Agency has performed an extensive review of available published and unpublished data available on the risk of VTE during use of all combined oral contraceptives and non-oral combined hormonal contraceptives (CHCs) available in the EU.52 This committee has concluded that “the benefits of CHCs in preventing unwanted pregnancies continue to outweigh their risks, and that the well-known risk of VTE with all CHCs is small. But the review has reinforced the importance of ensuring that clear and up-to-date information is provided to women who use these medicines and to the healthcare professionals giving advice and clinical care.”

142 Summary, discussion, and future perspectives

EU-wide, this review resulted in the implementation of updated texts in the physicians’ product information and patient leaflet reflecting the VTE risk differences between COCs and contributing risk factors. Additionally, several tools (a VTE physicians’ and patient card) have recently been made available. In this material, also special focus is made on recognition of VTE in young women, as VTE can present itself as atypical, and on the impact of other risk factors for VTE present. An overview of the guidance on VTE risk factors is given in the tables below.

Table 1. Information on VTE risk factors included in the physicians’ product information*

Risk factor Comment

Obesity (body mass index over 30 kg/m2) Risk increases substantially as BMI rises.

Particularly important to consider if other risk factors also present.

Prolonged immobilization, major surgery, In these situations it is advisable to discontinue any surgery to the legs or pelvis, neurosurgery, use of the pill (in the case of elective surgery at or major trauma least four weeks in advance) and not resume until two weeks after complete remobilization. Another method of contraception should be used to avoid unintentional pregnancy.

Antithrombotic treatment should be considered if COC has not been discontinued in advance.

Positive family history (venous thromboembolism If a hereditary predisposition is suspected, the ever in a sibling or parent especially at a relatively woman should be referred to a specialist for early age e.g. before 50). advice before deciding about any CHC use

Other medical conditions associated with Cancer, systemic lupus erythematosus, VTE haemolytic uraemic syndrome, chronic inflammatory bowel disease and sickle cell disease

Increasing age Particularly above 35 years

*Adapted from ‘Assessment report on combined hormonal contraceptives’ of the European Medicines Agency 201452

143 Chapter 7

Table 2. Text on VTE risk factors included in the patient leaflet of a COC*

Factors that increase your risk of a blood clot in a vein

The risk of a blood clot with X is small but some conditions will increase the risk. Your risk is higher:

• if you are very overweight (body mass index or BMI over 30kg/m2); • if one of your immediate family has had a blood clot in the leg, lung or other organ at a young age (eg. below the age of about 50). In this case you could have a hereditary blood clotting disorder; • if you need to have an operation, or if you are off your feet for a long time because of an injury or illness, or you have your leg in a cast. The use of X may need to be stopped several weeks before surgery or while you are less mobile. If you need to stop X ask your doctor when you can start using it again. • as you get older (particularly above about 35 years); • if you gave birth less than a few weeks ago.

The risk of developing a blood clot increases the more conditions you have.

*Adapted from ‘Assessment report on combined hormonal contraceptives’ of the European Medicines Agency 201452

Future perspectives

Up to now, the association between COC-use and risk of VTE has been the topic of many studies, including those presented in this thesis. It is expected that many more will follow, as this risk is still unsolved. Yet, the situation today is that the majority of women in the Western world uses a COC for contraception, and will do so for the upcoming years. Therefore, three lines are proposed: First, continuing further research to increase knowledge of risk factors influencing VTE risk in COC-users that can differentiate between women who can and who cannot safely use a COC. Second, further research may not result in absolute certainty in predicting individual VTE risk in COC-users, therefore uncertainties in VTE risk needs to be taken into account in the counseling in contraceptive choices. Three: alongside, there is a necessity to increase awareness that VTE, although rarely, can occur in healthy (young) COC-users. Indeed, awareness and education of thrombosis is the goal of the initiated World Thrombosis Day, which was first held on October 13th, 2014.53

144 Summary, discussion, and future perspectives

Increasing knowledge of risk factors predicting VTE risk in COC-users

As to predicting individual VTE risk, it was already put forward years ago that, despite increasing knowledge and better prophylaxis in risk situations, the prevalence of VTE was constant since 1979. This was then believed largely due to the inability to recognize those persons at greatest risk.54 However, the prevalence in 2015 is actually not much lower, as it is still referred to as about 1 in 1000 persons.55 This indicates that identifying persons at greatest risk are still insufficient.

The common theme in this thesis is the contribution of other VTE risk factors to the risk of VTE during COC-use. Of these, the increased risk of VTE in COC-users with a known hereditary thrombophilia is reasonably well established, of which 2 studies in this thesis have further contributed that there are large differences in absolute risk between mild and severe thrombophilias (Chapters 2 and 3). Howe- ver, in clinical practice, testing for any thrombophilia is not considered contributing in predicting VTE risk prior to start of COC-use, as due to the low prevalence, a very large number of women needs to be tested to prevent one VTE.56 Further, withholding COCs in non-symptomatic women tested positive for mild thrombophilia (factor V Leiden or prothrombin G20210A mutation), might be reconsi- dered when the woman has no adequate alternative contraception, as in this situation the increased risk of pregnancy and pregnancy-related VTE outweigh the COC-associated risk.

Further, there are a number of recognized exogenous or temporary risk factors that add up to the VTE risk of COC-use, e.g. surgery, trauma, immobilization, or obesity. These risk factors are indeed frequently present in COC-users who experienced VTE, as is also reported in one of the studies in this thesis that reported on the clinical profile of these women. However, these daily life risks are also very common in women who do not experience COC-associated VTE. Therefore, the predictive value of these risks for the individual woman is uncertain.

Also a positive family history is an established predictor of VTE risk, and recently the number of studies exploring the association between family history and VTE is increasing. In 2004, a large family-based study, which addressed the heritability of VTE and the potential mode of inheritance, showed that VTE is not solely due to environmental exposures.

145 Chapter 7

The authors found that a multifactorial non-Mendelian inheritance model fitted best as the cause for VTE, and concluded that VTE is probably a result of multige- nic action as well as environmental exposures.54 This outcome is consistent with the commonly accepted hypothesis that VTE is a multifactorial disease.47 Indeed, up to now several family cohort studies have shown VTE to occur more often in first-degree relatives of patients with VTE than in the general population. These studies showed that the increased VTE risk observed in relatives is not fully explained by the presence of known hereditary thrombophilias, as risk of VTE is also increased in relatives without thrombophilia. An increased risk of VTE in non-thrombophilic women was also observed in our family cohort studies. Further to this, emerging data indicate that a family-conferred risk might be of particular importance in young relatives, as pointed out in comments accompan- ying recent published studies evaluating details of VTE risk of a family history.57,58 These studies, one focusing on characteristics of relatives of symptomatic probands26 and the other focusing on characteristics of symptomatic probands with or without unprovoked VTE,27 showed that VTE is aggregated in families and that “uncovering the sources of this familial aggregation (genetic and non- genetic) may be worthwhile”. It was suggested that for younger relatives, a positive family history is a major risk factor for VTE that trumps known thrombophilic disorders.57 This more increased risk was especially observed in younger female relatives. With respect to above studies exploring predictors of risk of VTE, the recommendation to further explore the role of family history is supported, as information on family history is ready to be used within the counseling on VTE risks and contraception. In these studies, no specific attention was given yet to a possible familial susceptibility to hormonal exposure (COC-use and pregnancy). Our study, discussed in Chapter 5, focused on this question and although this is the first study, the results suggested that a hormonally (COC-use or pregnancy)- associated VTE in a female proband could be of additional relevance in predicting the VTE risk in her female relatives.

Options to further expand research on VTE risk factors influencing VTE risk during COC-use (and pregnancy) within families are using currently available data in order to increase the population. Options are to link existing family cohorts. Another approach could be to pool existing studies in women who experienced a COC- or pregnancy-associated VTE and focus on data collected on first-degree family history, although these studies have the drawback that most often a family history is self-reported.

146 Summary, discussion, and future perspectives

Further, existing registries could be used to evaluate VTE risk patterns in women between generations of families. In this, the multigenetical data collected in the LifeLines cohort study and biobank59 would be an excellent source.

Improvement in counseling on contraceptive options

Although above discussed focus on further research will bring more differentia- tion in VTE risks and may find additional risk factors, it needs to be considered that absolute certainty in predicting individual VTE risk may not be reached. We didn’t achieve this in the past 50 years: although there are few definite situations in which COCs should not be prescribed, the majority of the defined risk factors result in a mild increase in VTE for which it remains difficult to draw definite conclusions. Also the results of the studies in this thesis underline that with few exceptions, single VTE risk factors result in a mild increase in risk. From these uncertainties in predicting VTE risk it follows that from the first contraception consult on detailed counseling is recommended on all contraceptive options, including COCs, addressing both existing risk factors of VTE and risk of unintended pregnancy. In the counseling, existing adequate alternatives (levonorgestrel- containing IUD, progestogen-only implant, oral progestogen-only tablet, copper IUD) could be far better promoted as equally good alternatives, as contraceptive adequacy and tolerability pattern can compete with that of a COC. Although, no direct comparison versus COCs is available, as such studies are difficult to perform, several studies compared different progestogen-only products which showed good efficacy, tolerability and acceptance47,48 In comparison to COCs, the return of fertility after discontinuation or removal of these products is equally rapidly restored. Prior counseling on the unpredictable bleeding pattern to be expected is shown to reduce discontinuation rates.49,50 The one exception is the depot medroxyprogesterone (Depo-Provera©) of which data are disadvantage- ous with regard to return of fertility and its risk of VTE is yet uncertain.

To improve counseling on contraceptive options, the following lines to take are proposed: • Training of health care professionals in risks and benefits of all contraceptive options, taking into account relevant risk factors of VTE;

147 Chapter 7

• Developing decision aids, such as information cards, which present all adequate contraceptive options, stating their pro’s and con’s.

• Evaluation of the effectiveness of such training in prescribing practices on contraception in implementation studies; in this evaluation also the patient view can be taken into account.

• Alongside, a well-informed public, i.e. the women in need for contraception, will also a good safeguard to improve safe use of contraceptives. Therefore, ideally all sources of information accessible for the public related to contraceptive choices, need to be updated with adequate objective information that will help women to select the most appropriate method of contraception.

Increasing awareness of VTE in COC-users and health care providers

Apart from continuing research, a further focus on measures to improve physicians and patient’s awareness of VTE during use of COCs is needed. Although this seems obvious, we learned from the recent discussions on Diane 35 in public literature and the media, that a major issue appeared the unawareness that healthy young women could suffer from a VTE during COC-use, as the doctor and patient had not in all cases recognized the presence of a VTE. Therefore, diagnosis and treatment was delayed. Additionally, temporary risk factors that add up to the VTE risk during COC-use, such operation, or immobilization, are not common knowledge.

Recently, the European Medicines Agency has requested to implement a EU-wide update of the official information in physicians’ product information and patient leaflet on differences in risk of VTE between COCs. Additionally, guidance is included regarding the complexity of other risk factors that can increase risk. Although these updated texts present the latest information, these are difficult to read due to the large amount of data presented and the complexity of the VTE risk as many additional risk factors can additionally contribute. Therefore, there is a need for clear and concise information that presents contraceptive effectiveness and risks of all contraceptives currently available. Such an information card could be used as a tool in the counseling on the most suitable contraceptive for the individual woman.

148 Summary, discussion, and future perspectives

Within this effort, it is also recommended to provide more consistency in the information on risk of VTE during contraceptive use that is available in all information areas, including the many (commercial) websites that are focusing on providing information on contraceptive options for women.

In conclusion, increasing awareness of risk factors for VTE that can be used to better predict who can safely use COCs will improve the balance between preven- tive interventions to avoid this rare but serious complication of COCs and withholding an adequate contraceptive; increasing knowledge of VTE as a rare adverse event will improve recognition of VTE and will result in a more timely diagnosis and treatment.

149 Chapter 7

References

1. Lidegaard Ø, Løkkegaard E, Svendsen AL, Agger C. Hormonal contraception and risk of venous thromboembolism: national follow-up study. BMJ 2009;33:b2890.

2. van Hylckama V, Helmerhost FM, Vandenbroucke JP, Doggen CJ, & Rosendaal FR. (2009) The venous thrombotic risk of oral contraceptives, effects of oestrogen dose and progestogen type: results of the MEGA case-control study. BMJ:339:b2921.

4. WHO Expert Group Medical eligibility criteria for contraceptive use. Reproductive Health and Research, 4th ed. Geneva, Switzerland: World Health Organization; 2009/2015.

5. Middeldorp S, Henkens CM, Koopman MM, van Pampus EC, Hamulyák K, van der Meer J, Büller HR. The incidence of venous thromboembolism in family members of patients with factor V Leiden mutation and venous thrombosis. Ann Intern Med 1998;128:15-20.

6. Vossen CY, Conard J, Fontcuberta J, Makris JM, van der Meer FJM, Pabinger I, Palareti G, Preston FE, Scharrer I, Souto JC, Svensson P, Walker ID, Rosendaal FR. Familial thrombophilia and lifetime risk of venous thrombosis. J Thromb Haemost 2004;2:1526-1532.

7. Andersen BS, Olsen J, Nielsen GL, Steffensen FH, Sorensen HT, Baech J, Greger- sen H. Third generation oral contraceptives and heritable thrombophilia as risk factors of non-fatal venous thromboembolism. Thromb Haemost 1998;79:28-31.

8. Martinelli I, Sacchi E, landi G, Taioli E, Duca F, Manucci PM. High risk of cerebral vein thrombosis in carriers of a prothrombin-gene mutation and in users of oral contraceptives. N Engl J Med 1998;338:1793-1797.

150 Summary, discussion, and future perspectives

9. Martinelli I, Tailoi E, Bucciarelli P, Akhavan S, Manucci PM. Interaction between the G20210A mutation of the prothrombin gene and oral contraceptives use in deep vein thrombosis. Arterioscler Thromb Vasc Biol 1999;19:700-703.

10. Spannagl M, Heinemann LA, Schramm W. Are factor V Leiden carriers who use oral contraceptives at extreme risk for venous thromboembolism? Eur J contracept reprod health care 2000; 5:105-112.

11. Aznar J, Vaya A, Estelles A, MiraY, Segui R, Villa P, Ferrando F, Falco C, Corella D, Espana F. Risk of venous thrombosis in carriers of the prothrombin G20210A variant and factor V Leiden and their interaction with oral contraceptives. Haema- tologica 2000;85;1271-1276.

12. Legnani C, Palareti G, Guazzaloca G, Cosmi B, Lunghi B, Bernardi F, Coccheri S. Venous thromboembolism in young women: role of thrombophilic mutations and oral contraceptive use. Eur Heart J 2002;239:984-990.

13. Vaya A, Mira Y, Mateo J, Falco C, Villa P, Estelles A, Aznar J. Prothrombin G20210A mutation and oral contraceptive use increase upper extremity deep vein thrombotic risk. Thromb Haemost 2003; 89:452-457.

14. Martinelli I, Battaglioli T, Buccarelli P, Passamonti SM, Manucci PM. Risk factors and recurrence rate of primary deep vein thrombosis of the upper extremities. Circulation 2004;110:566-570.

15. Sidney S, Petitti DB, Soff GA, Cundiff DL, Tolan KK, Quesenberry CP. Venous thromboembolic disease in users of low-estrogen combined estrogen-progestin oral contraceptives. Contraception 2004;70;3-10.

16. Roach REJ, Lijfering WM, Helmerhorst FM, Cannegieter SC, Rosendaal FR, van Hylckama Vlieg A. The risk of venous thrombosis in women over 50 years old using oral contraception or postmenopausal hormone therapy. J Thromb Haemost 2013;11:124-131.

17. Bergendal A, Persson I, Odeberg J, Sundström A, Holmström M, Schulman s, Björgell O, Kieler H. Association of venous thromboembolism with hormonal contraception and thrombophilic genotypes. Obstet Gynecol 2014;124:600-609.

151 Chapter 7

18. Simioni P, Sanson BJ, Prandoni P, Tormene D, Friederich PW, Girolami B, Prins MH. Incidence of venous thromboembolism in families with inherited thrombophilia. Thromb Haemost 1999;81:198-202.

19. Van Vlijmen EFW, Brouwer JL, Veeger N, Eskes TK, Graeff PA, van der Meer J. Oral Contraceptives and the Absolute Risk of Venous Thromboembolism in Women With Single or Multiple Thrombophilic Defects. Arch Intern Med 2007;167:282- 289.

20. Van Vlijmen EF, Veeger N, Middeldorp S, Hamulyák K, Prins MH, Büller HR, Meijer K. Thrombotic risk during oral contraceptive use an pregnancy in women with factor V Leiden or prothrombin mutation: a rational approach to contraception. Blood 2011;118:2055-2061.

21. Bezemer ID, van der Meer FJM, Eikenboom JCJ, Rosendaal FR, Doggen CJM. The value of family history as a risk indicator for venous thrombosis. Arch Intern Med 2009;169:610-615.

22. Noboa S, Le Gal, Lacut K, Mercier B, Leroyer C, Nowak E, Mottier D, Oger E: for the EDITH Collaborative Study Group. Family history as a risk factor for venous thromboembolism. Thromb Res 2008;122:624-629.

23. Dowling NF, Austin H, Diley A, Whitsett C, Evatt BL, Hooper WC. The epidemiology of venous thromboembolism in Caucasians and African-Americans: the GATE Study. J Thromb Haemost 2003;1:80-87.

24. Silverstein MD, Heit JA, Mohr DN, Petterson TM, O'Fallon WM, Melton III LJ. Trends in incidence of deep vein thrombosis and pulmonary embolism. Arch Intern Med 1998;158:585-593.

25. Zöller B, Li X, Sundquist J, Sundquist K. Age- and Gender-Specific Familial Risks for Venous Thromboembolism. A Nationwide Epidemiological Study Based on Hospitalizations in Sweden. Circulation 2011;124:1012-1020.

26. Couturaud, F, Leroyer C, Tromeur C, Julian JA, Kahn SR, Ginsberg JS, Wells PS, Douketis JD, Mottier D, Kearon C. Factors that predict thrombosis in relatives of patients with venous thrombosis. Blood 2014;124:2124-2130.

152 Summary, discussion, and future perspectives

27. Cushman M, Glynn RJ, Goldhaber SZ, Moll S, Bauer KA, Deitcher S, Shrivastava S, Ridker PM. Hormonal factors and risk of recurrent venous thrombosis: the Prevention of Recurrent Venous Thromboembolism trial. J Thromb Haemost 2006;4:2199-2203.

28. Le Gal G, Kovacs MJ, Carrier M, Do K, Kahn SR, Wells PS, Anderson DA, Chagnon I, Solymoss S, Crowther M, Righini M, Lacut K, White RH, Vickars L, Rodger M. Risk of recurrent venous thromboembolism after a first oestrogen-associated episode. Data from the REVERSE cohort study. Thromb Haemost 2010;104:498- 503.

29. Eischer L, Eichinger S, Kyrle PA. The risk of recurrence in women with venous thromboembolism while using estrogens: a prospective cohort study. J Thromb Haemost 2014;12:635-640.

30. Middeldorp S. Is Thrombophilia Testing Useful? Hematology Am Soc Hematol Educ Program 2011;2011:150-155.

31. Visser M, Coelingh Bennink HJ. Clinical applications for estetrol. J Steroid Biochem Mol Biol 2009;114:85-89.

32. Odlind V, Milsom I, Persson I, Victor A. Can changes in sex hormone binding globulin predict the risk of venous thromboembolism with combined oral contraceptive pills? Acta Obstet Gynecol Scand 2002;81:482-490.

33. van Vliet HAAM, Frolich M, Thomassen MCLGD, Doggen CJ, Rosendaal FR, Rosing J, Helmerhorst FM. Association between sex hormone-binding globulin levels and activated protein C resistance in explaining the risk of thrombosis in users of oral contraceptives containing different progestogens. Hum Reprod 2005;20:563-568.

34. Stanczyk FZ, Grimes DA. Sex hormone-binding globulin: not a surrogate marker for venous thromboembolism in women using oral contraceptives. Contraception 2008;78:201-203.

35. Tchaikovski S, Rosing J. Mechanisms of Estrogen-Induced Venous Thromboem- bolism. Thromb Res 2010;126:5-11.

153 Chapter 7

36. Lidegaard Ø, Nielsen LH, Skovlund CW, Løkkegaard E. Venous thrombosis in users of non-oral hormonal contraception: follow-up study, Denmark 2001-10. BMJ 2012;344:e2990.

37. Scarabin PY, Oger E, Plu-Bureau G. Differential association of oral and transdermal oestrogen-replacement therapy with venous thromboembolism risk. EStro- gen and THromboEmbolism Risk Study Group. Lancet 2003;362:428-32.

38. Post MS, Christella M, Thomassen LG, van der Mooren MJ, van Baal WM, Rosing J, Rosing J, Kenemans P, Stehouwer CD. Effect of oral and transdermal estrogen replacement therapy on hemostatic variables associated with venous thrombosis: a randomized, placebo-controlled study in postmenopausal women. Arterioscler Thromb Vasc Biol 2003;23:1116-1121.

39. AGOG Committee opinion 556 of the Committee of gynecologic practice 2013: Postmenopausal estrogen therapy: rout of administration and risk of venous thromboembolism. Available at: http://www.acog.org/-/media/Committee-Opinions/Committee-on-Gynecologic- Practice/co556.pdf?dmc=1&ts=20150718T1114397055

40. Herrington DM, Vittinghoff E, Lin F, Fong J, Harris F, Hunninghake D, Bittner V, Schrott HG, Blumenthal RS, Levy R, HERS Study Group. Statin therapy, cardiovascular events, and total mortality in the Heart and Estrogen/Progestin Replace- ment Study (HERS). Circulation 2002;105:2962-2967.

41. The contraception pipeline database. Available at: http://pipeline.ctiexchange.org/

42. Brache V, Sitruk-Ware R, Williams A, Blithe D, Croxatto H, Kumar N, Kumar S, Tsong Y, Sivin I, Nath A, Sussman H, Cochon L, Miranda MJ, Reyes V, Faundes A, Mishell D Jr. Effects of a novel estrogen-free, progesterone receptor modulator contraceptive vaginal ring on inhibition of ovulation, bleeding patterns and endometrium in normal women. Contraception 2012;85:480-488.

43. Lal J, Nitynand S, Asthana OP, Nagaraja NV, Gupta RC. Optimization of contraceptive dosage regimen of Centchroman. Contraception 2001;3:47-51.

154 Summary, discussion, and future perspectives

44. Adomaityte J, Farooq M, Qayyum R. Effect of raloxifene therapy on venous thromboembolism in postmenopausal women. A meta-analysis. Thromb Haemost 2008;99:338-342.

45. Peluffo MC, Braeuer SN, Rotgeri A, Fritzemeier KH, Fuhrmann U, Buchmann B, Adevai T, Murphy MJ, Zelinski MB, Lindenthal B, Hennebold JD, Stouffer RL. A prostaglandin E2 receptor antagonist prevents pregnancies during a preclinical contraceptive trial with female macaques. Hum Reprod 2014;29:1400-1412.

46. Glasier A. Implantable contraceptives for women: effectiveness, discontinuation rates, return of fertility, and outcome of pregnancies. Contraception 2002;65:29-37.

47. Weisberg E, Bateson D, McGeechan K, Mohapatra L. A three year comparative study of continuation rates, bleeding patterns, and satisfaction in Australian using a subdermal contraceptive implant or progestogen-releasing intrauterine system. Eur J Contracept Reprod Health Care 2014;19:5-14.

48. Mantha S, Karp R, Raghavan V, Terrin V, Bauer KA, Zwicker JI. Assessing the risk of venous thromboembolic events in women taking progestin-only contraception: a meta-analysis. BMJ 2012;345:e4944.

49. Mansour D, Bahamondes L, Critchley L, Darney P, Frazer IS. The management of unacceptable bleeding patterns in etonogestrel-releasing contraceptive implant users. Contraception 2011;83:202-210.

50. Modesto W, Bahamondes M, Bahamondes L. A randomized clinical trial of the effect of intensive versus non-intensive counseling on discontinuation rates due to bleeding disturbances of three long-acting reversible contraceptives. Hum Reprod 2014;29:1393-1399.

51. A randomized study comparing the effect on ovarian activity on a progestogen- only pill (POP) containing desogestrel and a new POP containing drospirenone in a 24/4 regimen. J Contracept Reprod Health Care 2015 early on line June 15th DOI103109/13625187.2015.1044082

155 Chapter 7

52. Assessment report for combined hormonal contraceptives containing medicinal products. Procedure under Article 31 of Directive 2001/83/EC. Available at: http://www.ema.europa.eu/docs/en_GB/document_library/Referrals_document/Co mbined_hormonal_contraceptives/Recommendation_provided_by_Pharmacovigilan ce_Risk_Assessment_Committee/WC500160272.pdf

53. Rosendaal FR, Raskob GE. On World Thrombosis Day. Lancet 2014;384:1653- 1654.

54. Heit JA, Phelps MA, Ward SA, Slusser JP, Petterson TM, De Andrade TM. Familial segregation of venous thromboembolism. J Thromb Haemost 2004;2:731-736.

55. Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet 1999;353:1167-73.

56. Vandenbroucke J, van der Meer F, Helmerhorst F, Rosendaal FR. Factor V Leiden: should we screen oral contraceptive users and pregnant women? BMJ 1996;313:1127-1130.

57. Eikelboom JW, Weitz JI. Importance of Family History as a Risk Factor for Venous Thromboembolism. Circulation. 2011;124:996-997.

58. Garcia DA. VTE risk and family history: provocative findings. Blood 2014;124:2011.

59. www.lifelines.nl

156 Nederlandse samenvatting Hoofdstuk 8

INLEIDING

Dit proefschrift heeft als centraal thema de orale gecombineerde anticonceptiva (OACs), beter bekend als de ‘Pil’ en het risico op veneuze trombose, een zeldzame bijwerking die kan optreden tijdens pilgebruik.

In deze inleiding wordt een kort overzicht gegeven van de werking van OACs, de verschillende OACs die in Nederland verkrijgbaar zijn en de anticonceptieve werking ten opzichte van andere anticonceptiemethoden. Daarnaast wordt inge- gaan op het ontstaan van veneuze trombose en de verschillende factoren die het risico op trombose kunnen verhogen.

Gecombineerde orale anticonceptiva Samenstelling OACs bestaan uit een combinatie van een oestrogeen en progestageen, beide afgeleid van vrouwelijke geslachtshormonen. De oestrogene component bestaat uit ethinylestradiol en recent ook estradiol, maar in de loop der jaren zijn er veel nieuwe progestagenen ontwikkeld, o.a. norethisteron, levonorgestrel, desogestrel, gestodeen, norgestimaat, cyproteron, chlormadinon, dienogest, etonogestrel, drospirenon, norelgestromin en nomegestrol.

De allereerste OAC bevatte een hoge dosis ethinylestradiol van meer dan 100 microgram, maar vanaf de zeventiger jaren is de dosering gestaag gedaald tot 35-15 microgram. Daarnaast zijn er nu ook niet-orale gecombineerde anticonceptiva beschikbaar, zoals een pleister en een vaginale ring.

De eerste progestagenen gaven naast het gewenste progestagene effect ook androgene bijwerkingen, zoals acne, overmatige haargroei (hirsutisme) en een mogelijk ongunstig effect op het vetmetabolisme. Daarom richtte men zich op de ontwikkeling van nieuwe progestagenen met minder androgene bijwerkingen. Het tijdstip waarop een OAC met een nieuw progestageen beschikbaar kwam is onderverdeeld in opeenvolgende ‘generaties’. De allereerste OACs vallen onder de 1e generatie. Een overzicht is gegeven in Tabel 1.

158 Nederlandse samenvatting

Tabel 1: Overzicht van de in Nederland verkrijgbare OACs ingedeeld in generatie en jaar van introductie

Type progestageen Generatie Introductie

Ethinylestradiol/levonorgestrel-bevattende OAC (Microgynon©) 2nd generatie 1974 Ethinylestradiol/norgestimaat-bevattende OAC (Cilest©) 2nd generatie 1990

Ethinylestradiol/desogestrel-bevattende OAC (Marvelon©) 3rd generatie 1984 Ethinylestradiol/gestodeen-bevattende OAC (Femodeen©) 3rd generatie 1984

Ethinylestradiol/cyproteron-bevattende OAC (Diane 35©) 4th generatie 1987 Ethinylestradiol/chlormadinona-bevattende OAC (Madinelle©) 4th generatie 1995 Ethinylestradiol/drospirenon-bevattende OAC (Yasmin©) 4th generatie 2000 Ethinylestradiol/etonogestrel-bevattende vaginale ring (NuvaRing©) 4th generatie 2001 Ethinylestradiol/norelgestromin-bevattende pleister (Evra©) 4th generatie 2002

Estradiol/dienogest-bevattende OAC (Qlaira©) ‘5th’ generatie 2008 Estradiol/nomegestrol-bevattende OAC (Zoely©) ‘5th’ generatie 2011 OAC: oraal anticonceptivum, a: niet geregistreerd in Nederland

Werkingsmechanisme Het progestageen remt de aanmaak van luteïniserend hormoon (LH), waardoor de toename in de LH-spiegel, die nodig is om het ovulatieproces te starten, wordt voorkomen. Het oestrogeen remt de aanmaak van follikelstimulerend hormoon (FSH), waardoor de groei en ontwikkeling van een dominante follikel wordt geremd. Door deze processen wordt de ovulatie voorkomen. Daarnaast zorgt het oestrogeen er voor dat het effect van het progestageen op het endometrium (baarmoederslijmvlies) wordt gestabiliseerd, zodat ongewenst tussentijds bloedverlies wordt voorkomen.

Anticonceptieve werking van OACs ten opzichte van andere anticonceptiemethoden Een OAC is een van de meest betrouwbare anticonceptiemethoden. Als de OAC correct wordt ingenomen dan is het risico op zwangerschap gedaald tot 0.3% (Pearl Index = het aantal zwangerschappen per 100 vrouwen die de OAC gedurende 1 jaar hebben gebruikt). Alleen anticonceptiemethoden waarbij de effectivi- teit niet afhankelijk is van een correcte inname, zoals een progestageen (Mirena©)- of koper-bevattende spiraal of een progestageen-bevattend implantatiestaafje (Implanon©), hebben een nog betere anticonceptieve werking, zie ook

159 Hoofdstuk 8

onderstaande Tabel 2. Daarin wordt ook de Pearl Index genoemd, die is gebaseerd op alle zwangerschappen die zijn opgetreden tijdens gebruik, inclusief de zwangerschappen opgetreden bij incorrect gebruik zoals het vergeten van de pil.

Tabel 2. Percentage vrouwen met onbedoelde zwangerschap tijdens 1e jaar van gebruik per methode

% Vrouwen met onbedoelde zwangerschap tijdens 1e jaar van gebruik

Methode Typisch gebruik1 Correct gebruik2 Geen anticonceptie 85 85 Spermicide 29 18 Pessarium 16 6 Condoom: - vrouwen 21 5 - mannen 15 2 Gecombineerd oraal anticonceptivum 8 0.3 Desogestrel-bevattende pil (Cerazette©) 8 0.3 Gecombineerde hormonale pleister (Evra©) 8 0.3 Gecombineerde hormonale vaginale ring (NuvaRing©) 8 0.3 Medroxyprogesteron-bevattende injectie (Depo-Provera©)* 3 0.3 Levonorgestrel-bevattende spiraal (Mirena©)* 0.2 0.2 Koper-bevattende spiraal* 0.8 0.6 Etonogestrel-bevattend implantatiestaafje (Implanon©)* 0.05 0.05

Afgeleid van Trussel J. Contraception 2011;83:397-404 1: Typisch gebruik laat zien hoe betrouwbaar de verschillende methoden zijn tijdens gebruik in de praktijk (op basis van zwangerschap tgv falen van de methode en als gevolg van incorrect gebruik). 2: Zwangerschapscijfer bij correct gebruik laat zien hoe betrouwbaar de methode is bij correcte inname *: LARCs, long acting reversable contraceptives

Veneuze trombose Veneuze trombose is het ontstaan van een bloedstolsel in een ader zonder dat daar een fysiologische aanleiding voor is. De meest voorkomende locatie is het been. Als complicatie kan de in een beenader ontstane trombus gedeeltelijk loslaten en met de bloedstroom opwaarts verder getransporteerd worden naar de longslagader en daar een longembolie veroorzaken.Veneuze trombose wordt gezien als een multifactoriële aandoening, waarbij zowel genetische als verworven risicofactoren voor trombose een rol kunnen spelen. De interactie tussen deze risicofactoren bepaalt uiteindelijk of er daadwerkelijk een trombose zal optreden. In de algemene populatie komen veel risicofactoren voor die de kans op trombose verhogen en vaak ook meerdere tegelijkertijd, zoals een operatie, langdurige immobilisatie, trauma, zwangerschap en postpartum periode, pilgebruik, hormoongebruik in de postmenopauze, hormoongebruik bij in-vitro fertilisatie, obesitas,

160 Nederlandse samenvatting

een maligniteit gebruik bij in-vitro fertilisatie, obesitas, een maligniteit en andere onderliggende aandoeningen. De huidige opvatting is dat de aanwezigheid van meer dan één risicofactor noodzakelijk is voor het optreden van een trombose.

Risico op trombose tijdens pilgebruik Het verhoogde risico op trombose bij pilgebruik wordt veroorzaakt door de oestrogene component. Daarnaast kan het type progestageen het oestrogeen- gerelateerde risico op trombose beïnvloeden. Recent is er een re-evaluatie van het tromboserisico voor alle OACs, de vaginale ring en de pleister uitgevoerd door de Europese registratie-autoriteiten. Deze re-evaluatie had tot gevolg dat de bijslui- terteksten t.a.v. het tromboserisico hierop zijn aangepast, zie Tabel 3.

Tabel 3. De geschatte aantal trombose gevallen per 10.000 vrouwen per jaar van gebruik

RR t.o.v. Incidentie Generatie levonorgestrel per 10.000 OAC vrouwen/ jaar van gebruik

Niet-gebruikster 2-3

Ethinylestradiol/levonorgestrel-bevattende OAC (Microgynon 30©) Referentie 5-7 2e Ethinylestradiol/norethisteron-bevattende OAC (Modicon©) 1 5-7 1e Ethinylestradiol/norgestimaat-bevattende OAC (Cilest©) 1 5-7 2e Ethinylestradiol/desogestrel-bevattende OAC (Marvelon©) 1.5-2 9-12 3e Ethinylestradiol/gestodeen-bevattende OAC (Femodeen©) 1.5-2 9-12 3e Ethinylestradiol/drospirenon-bevattende OAC (Yasmin©) 1.5-2 9-12 4e Ethinylestradiol/cyproterone-bevattende OAC (Diane 35©) 1.5-2 9-12 4e Ethinylestradiol/etonogestrel-bevattende vaginale ring (NuvaRing©) 1.2 6-12 4e Ethinylestradiol/norelgestromin-bevattende pleister (Evra©) 1.2 6-12 4e Ethinylestradiol/chlormadinon-bevattende OAC (Madinelle©)a Onbekend Onbekend 4e Estradiol/dienogest-bevattende-OAC (Qlaira©) Onbekend Onbekend ‘5e’ Estradiol/nomegestrol-bevattende OAC (Zoely©) Onbekend Onbekend ‘5e’

RR: relatieve risico. a: niet beschikbaar in Nederland. Gebaseerd op ‘Assessment report for combined hormonal contraceptives containing medicinal products’ 16 Januari 2014, EMA/739865/2013

161 Hoofdstuk 8

Uit deze re-evaluatie blijkt dat het tromboserisico van norgestimaat- en norethisteron-bevattende OACs vergelijkbaar is met dat van levonorgestrel- bevattende OACs. Een hoger tromboserisico wordt gezien bij gebruik van een desogestrel-, gestodeen-, drospirenon- en cyproteron-bevattende OACs. De cyproteron-bevattende ‘OAC’ is alleen geïndiceerd voor hormonale behandeling van vrouwen met androgeen-sensitieve acne en hirsutisme, maar heeft dezelfde anticonceptieve werking als de andere OACs. Ook de vaginale ring en de pleister hebben een hoger tromboserisico. Het absolute risico op trombose is echter voor alle OACs heel laag; 5-12 gevallen van trombose per 10.000 piljaren van gebruik in vergelijking met 2-3 gevallen van trombose per 10.000 vrouwen die geen OAC gebruiken.

Het tromboserisico is ook afhankelijk van de duur van het gebruik: het risico is het hoogst in het eerste jaar na starten van de OAC en mogelijk ook na het opnieuw starten van pilgebruik na een onderbreking van tenminste 4 weken. Het risico neemt bij langer gebruik geleidelijk af, maar blijft altijd hoger dan het risico bij vrouwen die geen OAC gebruiken.

Daarnaast is het tromboserisico ook afhankelijk van de dosering. De oudere pillen die 50 microgram ethinylestradiol bevatten hebben een hoger risico dan de huidige OACs met een lagere dosis (35-15 microgram), de zogenaamde sub-50 OACs. Mogelijk geeft een pil met 20 of 15 microgram ethinylestradiol een nog lager risico, maar de gegevens zijn nog beperkt.

Biologische achtergrond van het risico op trombose bij pilgebruik De biologische achtergrond van de associatie tussen trombose en pilgebruik is complex. Pilgebruik veroorzaakt veranderingen in de haemostatische balans naar een protrombotische richting. Deze effecten zijn meer uitgesproken bij 3e en 4e generatie OACs dan bij 2e generatie OACs. Bij gezonde vrouwen vallen deze veranderingen echter binnen de normale variatie.

Risico op trombose tijdens de zwangerschap en post-partum periode Het risico op trombose is veel hoger tijdens zwangerschap en met name tijdens de postpartum periode dan bij pilgebruik. In vergelijking met niet zwangere vrouwen is het risico tijdens de zwangerschap ongeveer 9-voudig verhoogd en tot 50-keer verhoogd tijdens de postpartum periode.

162 Nederlandse samenvatting

De belangrijkste oorzaak is de hypercoagulante staat, een fysiologisch mecha- nisme dat de aanstaande moeder moet behoeden voor te veel bloedverlies nadat de placentageboorte heeft plaats gevonden. Na de bevalling keren de afwijkende waarden weer terug naar de uitgangswaarden, maar dat kan wel tot 8 weken na de bevalling duren.

Trombofilie Erfelijke trombofiele afwijkingen geven een verhoogd tromboserisico. Er zijn op dit moment vijf relevante erfelijke trombofiele afwijkingen aangetoond. De zeldzame, maar ernstige trombofilie afwijkingen (antitrombine-, proteïne C- of proteïne S-deficiëntie) hebben een prevalentie van elk 0.1%. De meer frequenter voorkomende mildere trombofiele afwijkingen (factor V Leiden of protrombine G20210A mutatie), hebben een prevalentie van respectievelijk 5% en 2% in de Nederlandse bevolking. Echter, in 50% van de families, waarin veel trombose wordt gezien kan geen erfelijk trombofiel defect worden aangetoond.

HOOFDSTUK 1 Doelstelling van dit proefschrift Het centrale thema in dit proefschrift is de associatie tussen pilgebruik en het risico op trombose. In deze context is de bijdrage van additionele risicofactoren voor het optreden van trombose, inclusief trombofilie en een positieve familieanamnese onderzocht. Daarnaast zijn de risico’s vergeleken met het risico op trombose tijdens de zwangerschap, omdat een OAC wordt gebruikt om zwangerschap te voorkomen.

HOOFDSTUK 2 Ernstige erfelijke trombofilie en het risico op trombose tijdens pilgebruik Het risico op een veneuze trombose tijdens pilgebruik wordt toegeschreven aan veranderingen in de hemostatische balans. De impact van deze effecten is mogelijk groter in vrouwen met een erfelijke trombofiele afwijking. In dit hoofdstuk wordt een retrospectieve familiestudie beschreven, waarin we het tromboserisico in pilgebruiksters met een ernstige trombofiele afwijking en de bijdrage van even- tuele andere aanwezige risicofactoren aan dit tromboserisico hebben geëvalu- eerd. In dit cohort, werden de vrouwelijke en mannelijke familieleden van patiën- ten met een antitrombine-, proteïne C- en proteïne S deficiëntie en een doorgemaakte trombose geïncludeerd, waarbij de mannen fungeerden als controle- groep.

163 Hoofdstuk 8

Van de 222 geïncludeerde vrouwen hadden 135 (62%) ooit een OAC gebruikt. Het totale absolute risico op trombose per 100 persoonsjaren in deficiënte vrouwen was respectievelijk 1.64 versus 0.18 in niet-deficiënte vrouwen; het geadjusteerde relatieve risico was 11.9 (95% betrouwbaarheidsinterval [CI] 3.9-36.2). Het tromboserisico was vergelijkbaar in mannen en vrouwen die ooit of nooit een OAC hadden gebruikt. Wel trad een trombose in deficiënte vrouwen op jongere leeftijd op dan in deficiënte mannen. Het absolute tromboserisico was 4.62 in deficiënte- versus 0.48 in niet-deficiënte pilgebruiksters; het relatieve risico was 9.7 (95% CI 3.01-42.4). Het absolute risico nam verder toe bij aanwezigheid van andere trombofiele afwijkingen; respectievelijk van 3.49 (alleen deficiëntie) naar 12.00 (meer dan 1 extra defect) per 100 piljaren van gebruik in deficiënte pilgebruiksters en van 0 (geen defect) naar 3.13 per 100 piljaren van gebruik in niet-deficiënte pilgebruiksters.

Op basis van deze resultaten kan men concluderen dat vrouwen met een erfelijke antitrombine-, proteïne C-, of proteïne S deficiëntie een hoog tromboserisico hebben tijdens pilgebruik, met name als er ook andere trombofiele afwijkingen aanwezig zijn.

HOOFDSTUK 3 Milde erfelijke trombofilie en het risico op trombose tijdens pilgebruik Internationale anticonceptierichtlijnen, o.a. die van de WHO, adviseren geen OAC voor te schrijven aan vrouwen met een erfelijke trombofiele afwijking. Maar de veronderstelling dat alle trombofiele afwijkingen hetzelfde risico op trombose geven zou ter discussie kunnen worden gesteld. Verschillende studies hebben aangetoond dat het absolute risico op trombose geassocieerd met een factor V leiden of protrombine G20210A mutatie aanmerkelijk lager is dan wordt gezien bij een antitrombine-, proteïne C- of proteïne S-deficiëntie. In dit hoofdstuk wordt een retrospectief familiecohort gepresenteerd waarin we het absolute tromboserisico tijdens pilgebruik en zwangerschap in vrouwen met een milde trombofiele afwijking hebben geëvalueerd. In dit cohort werden 798 vrouwelijke familieleden van patiënten met factor V Leiden, protrombin-G20210A mutatie, een dubbele heterozygotie of homozygotie voor factor V Leiden en protrombine-G20210A mutatie en een doorgemaakte trombose, geïncludeerd.

164 Nederlandse samenvatting

Het totale absolute risico op trombose in vrouwen zonder defect, een enkel defect of een dubbel defect was 0.13, 0.35 en 0.94 per 100 persoonsjaren, terwijl deze risico’s tijdens pilgebruik stegen naar 0.19, 0.49, en 0.86, en tijdens de zwangerschap-postpartum periode stegen naar 0.73, 1.97, en 7.65. pilgebruik en zwangerschap waren onafhankelijke risicofactoren voor trombose, waarbij het risico tijdens de zwangerschap veel hoger was dan tijdens pilgebruik, zoals wordt aangetoond door de gecorrigeerde Hazard ratio (HR) van 16.0 tijdens zwangerschap versus een HR van 2.2 tijdens pilgebruik. Deze absolute risico’s zijn vervol- gens gebruikt voor een extrapolatie naar het risico op trombose bij gebruik van verschillende alternatieve anticonceptiemethoden. Hierbij zijn zowel het tromboserisico van de alternatieve anticonceptiemethode als het tromboserisico tijdens zwangerschap (als gevolg van het falen van de anticonceptiemethode) berekend. Op basis van deze extrapolaties kan men concluderen dat zowel het levonorgestrel-bevattende spiraaltje als het koper-bevattende spiraaltje een geen verhoogd risico op trombose geven en een adequate anticonceptieve werking hebben; daarom zijn beide goede alternatieven voor pilgebruik. Condoomgebruik heeft een zeer onbetrouwbare anticonceptieve werking; met een tot 60-keer hoger risico op zwangerschap ten opzichte van een OAC kan het condoom kan worden beschouwd als de minst geschikte alternatieve anticonceptiemethode voor pilgebruik bij deze vrouwen. Op basis van deze resultaten doen wij de aanbeveling om, in plaats van het strikt contra-indiceren van pilgebruik bij deze vrouwen, een uitge- breide counseling te geven over alle beschikbare anticonceptiemethoden, inclusief een OAC, waarbij zowel het tromboserisico als het risico op onbedoelde zwangerschap wordt besproken.

HOOFDSTUK 4 Systematische review en meet-analyse van het tromboserisico in pilgebruiksters met trombofilie Dit hoofdstuk wordt begint met een klinische vraagstelling: “Een 29-jarige vrouw bij wie in het verleden in een research setting een heterozygotie voor factor V Leiden mutatie was vastgesteld) vraagt: wat is mijn risico op trombose als ik start met de Pil” Deze casus wordt gevolgd door de presentatie van een meet-analyse waarin het risico op trombose in pilgebruiksters met of zonder trombofilie is geëvalueerd.

165 Hoofdstuk 8

Een onderscheid werd gemaakt tussen milde (factor V Leiden, protrombin- G20210A mutatie) en ernstige trombofilie (antitrombine-, proteïne C-, proteïne S deficiëntie, een dubbele heterozygotie of homozygotie voor factor V Leiden en protrombine-G20210A mutatie). Twaalf case-control- en drie cohort studies werden geïdentificeerd. In pilgebruiksters met milde trombofilie werd een 6-voudige (relatieve risico (RR) 5.89) en in pilgebruiksters met ernstige trombofilie een 7-voudige (RR 7.15) toename in tromboserisico gezien. De familie cohortstudies lieten een veel hoger absoluut risico zien in pilgebruiksters met ernstige trombofilie van respectievelijk 4.3 en 4.6 per 100 piljaren van gebruik dan in pilgebruiksters met milde trombofilie (respectievelijk 0.49 en 2.0 per 100 piljaren van gebruik). Alle absolute risico’s zijn echter berekend in familieleden van patiënten met erfelijke trombofilie die een trombose hebben doorgemaakt en hebben daarom ook een positieve familieanamnese. Op basis van bovengenoemde uitkomsten adviseren wij een OAC niet voor te schrijven aan vrouwen met ernstige trombofilie. Het extra risico van een milde trombofilie afwijking is beperkt. Als er geen andere risicofactoren voor trombose aanwezig zijn dan zou een OAC kunnen worden gebruikt als de vrouw andere adequate anticonceptiemethoden niet accep- tabel vindt, omdat in deze situatie het toegenomen risico op zwangerschap- gerelateerde trombose en toegenomen risico op een onbedoelde zwangerschap hoger is dan het pilgerelateerde risico op trombose.

HOOFDSTUK 5 Een positieve familieanamnese afkomstig van een mannelijk of vrouwelijk familielid: is dit van invloed op het tromboserisico in vrouwen? Zoals eerder aangetoond in de hierboven beschreven studies, hebben vrouwen afkomstig uit trombofiele families een verhoogd tromboserisico dat verder stijgt tijdens pilgebruik en tijdens de zwangerschap en postpartum periode. Het is onbekend of dit additionele risico verschilt tussen vrouwelijke familieleden van een mannelijke of vrouwelijke patiënt en of het uitmaakt dat een vrouwelijke patiënt een trombose heeft doorgemaakt tijdens pilgebruik of de zwangerschap (hormonaal-gerelateerde trombose). Om deze vragen te onderzoeken is een grote cohortstudie uitgevoerd, beschreven in hoofdstuk 5. In dit familiecohort van 1005 vrouwelijke familieleden werd het tromboserisico vergeleken tussen familieleden van een mannelijke en een vrouwelijke patiënt en tussen familieleden van vrouwelijke patiënten met of zonder een hormonaal-gerelateerde trombose.

166 Nederlandse samenvatting

Het absolute tromboserisico in vrouwelijke familieleden van vrouwelijke (0.32) en mannelijke patiënten (0.39) was vergelijkbaar, maar een jonge leeftijd (onder de 45 jaar) van de patiënt op moment van optreden trombose en bij het familielid het hebben van een trombofiele afwijking, pilgebruik en zwangerschap waren factoren die het risico op trombose significant verhoogden. In de multivariabele analyses was een jonge leeftijd van de patiënt echter niet langer significant.De heterogeni- teitsanalyse van de verkregen risicoschattingen suggereerde echter dat het risico in familieleden van vrouwelijke versus mannelijke patiënten verschilt tijdens de zwangerschap en postpartum periode (Hazard Ratio [HR], 11.6 versus 6.6) en in mindere mate tijdens pilgebruik (HR, 3.6 versus 2.7), hoewel niet significant. Daarnaast was het absolute risico statistisch significant hoger (0.43 per 100 persoonsjaren) in familieleden van vrouwelijke patiënten met een hormoon- gerelateerde trombose dan in vrouwelijke familieleden van vrouwelijke patiënten met een niet hormonaal gerelateerde trombose (0.13 per 100 persoonsjaren). Bovendien werden de hogere HRs tijdens zwangerschap en pilgebruik in familieleden van vrouwelijke patiënten t.o.v. familieleden met mannelijke patiënten werden vooral gezien in familieleden van vrouwelijke patiënten met een hormonaal- gerelateerde trombose. Dit toegenomen risico op trombose suggereert dat vrouwelijke familieleden van vrouwelijke patiënten met een hormonaal gerelateerde trombose gevoeliger zijn voor hormonale blootstelling. Deze informatie kan van belang zijn bij de keuze van meest geschikte anticonceptiemethode.

HOOFDSTUK 6 Klinische kenmerken en het recidief risico in vrouwen met een pilgerelateerde trombose Pilgebruik verhoogt het risico op trombose, maar het is op voorhand niet goed duidelijk welke vrouw een trombose zal ontwikkelen tijdens pilgebruik. Daarnaast is er veel variatie in de uitkomsten ten aanzien van het risico op een recidieftrom- bose. Om deze beide vragen verder te onderzoeken is een prospectieve cohortstudie uitgevoerd, beschreven in hoofdstuk 6. Het doel van deze studie was tweele- dig: het beschrijven van de klinische kenmerken van vrouwen met een pilgerelateerde trombose en hun anticonceptiekeuze na het optreden van de trombose en het prospectief analyseren van hun risico op een recidief trombose. In dit cohort werden 125 vrouwen met pilgerelateerde trombose geïncludeerd. Hun mediane leeftijd ten tijde van de trombose was 29 jaar.

167 Hoofdstuk 8

Op het moment van optreden van de trombose gebruikte 69/125 (55%) van de vrouwen een levonorgestrel-bevattende OAC. In 69% van de gevallen gebruikten de vrouwen de OAC al langer dan 1 jaar. Voorafgaande aan het pilgebruik waarbij de trombose is opgetreden, had 68% van de vrouwen al eerdere periodes van pilgebruik en/of zwangerschap doorgemaakt. De meerderheid van de vrouwen had naast pilgebruik ook andere risicofactoren voor trombose: trombofilie (34%), 1e graad positieve familieanamnese (31%), obesitas (BMI ≥ 30kg/m2, 24%) en tijdelijke risico’s (postpartum periode, trauma, operatie en langdurige immobilisatie, 14%). Na het optreden van de trombose zijn alle vrouwen gestopt met pilgebruik en hebben zij tromboprofylaxe gekregen bij een volgende zwangerschap (25 [20%]). Wat betreft de keuze van alternatieve anticonceptie was de levonorgestrel-bevattende spiraal favoriet. De meeste vrouwen werden gedurende 6 maanden behandeld met orale anticoagulantia. De mediane follow-up na anticoa- gulante therapie was 37 maanden (7-98 maanden). Tijdens de follow-up traden 5 recidief tromboses op. Op basis hiervan werd een jaarlijkse kans op recidief berekend van 1.2% (95% CI: 0.44-2.63), met een risico van 1.8% (95% CI: 0.30-5.90) in het eerste jaar na de eerste trombose. De resultaten van deze groep van opeenvol- gend geïncludeerde vrouwen met pilgerelateerde trombose lieten geen duidelijk risicopatroon zien. De enige vermijdbare risicofactor betreft een positieve familieanamnese, die in een verrassend hoog percentage (31%) werd gemeld. Op dit moment wordt een positieve familieanamnese niet gezien als een strikte contra- indicatie voor pilgebruik. Het risico op een recidief trombose was laag wanneer verder oestrogeen-bevattende anticonceptie werd gestopt en tromboprofylaxe werd gegeven in de zwangerschap.

DISCUSSIE De studies beschreven in dit proefschrift hebben laten zien dat het tromboserisico in pilgebruiksters met een ernstige erfelijke trombofilie (antitrombine-, proteïne C-, of proteïne S deficiëntie) aanzienlijk is toegenomen. Op basis hiervan we adviseren dat pilgebruik in deze vrouwen sterk moet worden afgeraden. Echter, in pilgebruiksters met een milde erfelijke trombofilie (factor V Leiden of protrombin-G20210A mutatie), is het absolute tromboserisico substantieel lager dan hun tromboserisico tijdens de zwangerschap en postpartumperiode. Deze data indiceren dat de huidige WHO Medical Eligibility criteria (2015), waarin een factor V Leiden of protrombine G20210A mutatie als een strikte contra-indicatie

168 Hoofdstuk 8

voor pilgebruik wordt gezien, zou kunnen worden heroverwogen. Hoewel deze studieresultaten het pilgebruik in deze vrouwen zeker niet promoten, indiceren zij dat, in het geval pilgebruik is gestopt of wordt afgeraden, adequate alternatieve anticonceptie noodzakelijk is vanwege het hogere tromboserisico tijdens de zwangerschap. Op basis hiervan adviseren wij om in plaats van het contra-indiceren van pilgebruik, deze vrouwen uitleg te geven over alle beschikbare anticonceptiemethoden, inclusief een OAC, waarbij zowel het tromboserisico als het risico op onbedoelde zwangerschap wordt besproken, zodat deze vrouwen een geïnfor- meerde keuze kunnen maken wat betreft de meest geschikte anticonceptiemethode. De resultaten van de door ons uitgevoerde meta-analyse, waarin het tromboserisico in pilgebruiksters met en zonder trombofilie is geëvalueerd, sluiten aan op bovengenoemde studies. De resultaten laten zien dat in pilgebruiksters met milde en ernstige trombofilie het tromboserisico 6- tot 7-voudig is verhoogd ten opzichte van pilgebruiksters zonder trombofilie. Het absolute risico in pilgebruiksters met ernstige trombofilie is echter veel hoger dan in pilgebruiksters met milde trombofilie. Deze verschillen in absoluut risico werden ook gezien in niet- aangedane vrouwen uit families met ernstige versus milde trombofilie. Een mogelijke verklaring hiervoor is dat in deze families met ernstige trombofilie vaak ook andere stollingsafwijkingen voorkomen. Daarnaast hebben familieleden van trombofiele patiënten die een trombose hebben doorgemaakt, ook een positieve familieanamnese dat het basisrisico van deze familieleden 2 tot 3-voudig verhoogt. Daarom zijn absolute risico’s berekend in familieleden binnen trombofiele families extra toegenomen dan in de algemene populatie van pilgebruiksters waarbij een trombofiele afwijking wordt aangetoond. Zoals aangetoond in bovengenoemde studies en de meetanalyse, hebben vrouwen uit trombofiele families een verhoogd risico op trombose dat verder stijgt tijdens pilgebruik en tijdens de zwangerschap en postpartum periode. In een door ons uitgevoerde familie cohortstudie, waarin we hebben geëvalueerd of een positieve familieanamnese geslachtsspecifiek of hormoonspecifiek zou kunnen zijn, suggereren de resultaten dat het uitmaakt dat positieve familieanamnese afkomstig is van een vrouwelijk familielid, vooral als dit familielid een trombose heeft doorgemaakt tijdens pilgebruik of zwangerschap. Hoewel onze studie de eerste is die dit heeft onderzocht en meer studies nodig zijn om dit te bevestigen, raden we aan dit mee te nemen in de counseling dat vrouwen mogelijk een hoger risico hebben wanneer zij een moeder of zuster hebben die een hormoon-gerelateerde trombose heeft meegemaakt.

169 Hoofdstuk 8

Onze laatste studie geeft gedetailleerde informatie over de klinische karakteristie- ken van vrouwen met een pilgerelateerde trombose. Het merendeel van deze vrouwen (68%) had naast pilgebruik ook andere risicofactoren, met name trombofilie, een positieve familieanamnese en obesitas. De studieopzet is echter niet geschikt is om te voorspellen welke vrouw een pilgerelateerde trombose zal ontwikkelen en we zagen ook geen duidelijk patroon in de verschillende risico’s. Dit is ook in lijn met de aanname dat trombose als een multifactoriële aandoening moet worden gezien. Dit is wordt verder ondersteund door het feit dat twee derde van deze vrouwen de OAC al meer dan een jaar gebruikte voordat de trombose optrad en dat zij eerder zonder problemen een OAC hadden gebruikt en zwangerschappen hadden doorgemaakt. Het recidiefrisico in deze vrouwen, die allen waren gestopt met pilgebruik en tromboprofylaxe gebruikten bij een volgende zwangerschap, was zeer laag.

TOEKOMSTPERSPECTIEF De associatie tussen pilgebruik en trombose is al 50 jaar onderwerp van onderzoek, maar deze bijwerking is nog steeds aanwezig. De huidige situatie is echter dat miljoenen vrouwen een OAC gebruiken en ook in de komende jaren daar mee door zullen gaan. Daarom wordt voorgesteld 3 mogelijke lijnen voor de toekomst uit te zetten.

Ten eerste: verder onderzoek naar risicofactoren die beter kunnen bijdragen in het voorspellen welke vrouwen wel en niet een OAC veilig kunnen gebruiken. Er zijn verschillende bekende risicofactoren. Hierbij is een trombofilie een belangrijke factor, maar het vooraf testen van vrouwen is niet zinvol gezien de lage prevalentie van deze afwijkingen. Daarnaast zijn er een aantal veel voorkomende risicofactoren, zoals operatie, trauma, immobilisatie en obesitas. Maar deze komen ook vaak voor in vrouwen die geen pilgerelateerde trombose ontwikkelen. Daarom is de voorspellende waarde van deze risico’s lastig te beoordelen. Maar het spreekt voor zich dat hoe meer risicofactoren aanwezig zijn, des te meer het risico zal toenemen. Recente studies hebben laten zien dat verder onderzoek naar de oorzaken (genetisch en non-genetisch) van familiair voorkomen van trombose zou kunnen bijdragen, ook omdat voor jonge familieleden deze risicofactor zelfs belangrijker zou kunnen zijn dan het hebben van een trombofiele afwijking.

170 Nederlandse samenvatting

Ten tweede: omdat men rekening moet houden met het feit dat er geen absolute zekerheid kan worden gegarandeerd in het voorspellen van het individuele risico op trombose tijdens pilgebruik, moet deze onzekerheid ingebouwd worden in de afweging om wel of niet een OAC voor te schrijven. Er zijn maar een paar situaties waarbij een OAC moet worden vermeden, maar de meeste risicofactoren geven slechts een geringe verhoging van het risico. Vanuit dit gegeven is het daarom belangrijk bij de anticonceptie counseling alle risicofactoren in kaart te brengen en een uitgebreid advies te geven over de beschikbare anticonceptiemethoden, waarbij zowel het tromboserisico als het risico op onbedoelde zwangerschap wordt besproken. In deze counseling zouden de progestageen-alleen preparaten een veel prominentere plaats moeten krijgen.

Ten derde: tegelijkertijd is het noodzakelijk om voorschrijvers en vrouwen meer bewust te maken van de mogelijkheid dat een trombose kan optreden bij pilgebruik, ook bij jonge vrouwen. Trombose wordt niet altijd herkend, waardoor het tijdstip van diagnose en het starten van adequate behandeling wordt vertraagd. Daarnaast zijn additionele risicofactoren die het tromboserisico tijdens pilgebruik kunnen verhogen, zoals een operatie, postpartum periode en immobilisatie, niet bij iedereen bekend. Recent zijn de bijsluiters van alle OACs nu voorzien van een tabel met risicofactoren voor trombose en de symptomen (zie Tabel 4). Daarnaast is het van belang dat alle informatiebronnen, inclusief de commerciële websites, eensluidende informatieve op dit punt verstrekken.

Concluderend, met het verbeteren van de kennis van omtrent de verschillende risicofactoren voor trombose naast pilgebruik kan men het individuele risico op trombose beter inschatten, hoewel een VTE nooit is uit te sluiten; een toename in kennis over trombose als zeldzaam maar belangrijke bijwerking zal herkenning van trombose verbeteren en leiden tot tijdige diagnose en behandeling.

171 Hoofdstuk 8

Tabel 4. Samenvatting risicofactoren voor trombose en symptomen trombose opgenomen in de bijsluiter van alle OACs

BELANGRIJKE INFORMATIE OVER GECOMBINEERDE HORMOON-BEVATTENDE ANTICONCEPTIE- MIDDELEN (DE COMBINATIE-PIL) EN HET RISICO OP BLOEDSTOLSELS (TROMBOSE)

 Gecombineerde hormoon-bevattende anticonceptiemiddelen geven een licht verhoogde kans op een bloedstolsel (trombose).  Het risico is het grootst in het eerste jaar van gebruik (of als u weer begint nadat u meer dan vier weken bent gestopt).  Het risico op bloedstolsels is echter lager dan bij vrouwen tijdens de zwangerschap of in de periode na de bevalling.  Het is belangrijk dat u, als gebruikster van een gecombineerd hormoon-bevattend anticonceptiemiddel, goed op de hoogte bent van de verschijnselen die kunnen optreden bij een bloedstolsel. Lees daarom de bijsluiter zorgvuldig door en lees deze met enige regelmaat opnieuw, bij voorkeur eens per jaar, omdat de bijsluiter soms aangepast wordt.

De volgende klachten zouden kunnen wijzen op een bloedstolsel. Neem DIRECT contact met uw arts op bij:  pijn, roodheid en zwelling van het been  plotseling optreden van kortademigheid en hoesten, soms met wat bloed, pijn op de borst tijdens het ademhalen, een versnelde ademhaling en verhoogde hartslag  onverklaarde zware hoofdpijn

De kans op een bloedstolsel is verhoogd in de volgende situaties:  als u al eerder trombose (bloedstolsels) heeft gehad (dan mag u géén gecombineerd hormoon- bevattend anticonceptiemiddel gebruiken)  als er in de naaste familie (ouders, broer of zuster) trombose voorkomt op een leeftijd onder de 50 jaar  in de periode na de bevalling  als u te zwaar bent  als u net geopereerd bent of een ernstige verwonding hebt gehad  als u een tijd weinig bewogen heeft, bijvoorbeeld als u bedlegerig bent door ziekte

Vertel uw arts of andere zorgverlener dat u een gecombineerd hormoon-bevattend anticonceptiemiddel gebruikt:  als u gevraagd wordt of u medicijnen gebruikt (de anticonceptiepil, of vaginale ring of pleister, is óók medicatie)  als u ingepland wordt voor een operatie, of als u net bent geopereerd

Als u vragen heeft over dit geneesmiddel, neem dan contact op met uw arts. Voor meer informatie leest u de bijsluiter uit de verpakking of de informatie op de website www.cbg-meb.nl. Wilt u een bijwerking melden, dan kunt u dat doen bij uw arts of bij het Nederlands Bijwerkingen Centrum Lareb: www.lareb.nl.

172 Dankwoord Hoofdstuk 9

DANKWOORD

Dit proefschrift was niet tot stand gekomen zonder de inspirerende invloed van Jan van der Meer. Hij was degene die ja zei op de vraag of ik niet in mijn vrije tijd kon aanhaken in het lopende onderzoek van zijn afdeling, terwijl wij een kopje heel vieze koffie dronken bij de koffieautomaat op de gang. Vooral trombofilie en de interactie met pilgebruik vond ik vanwege mijn werk een interessant onderwerp. Jan zei dat hierover al veel was geschreven en de interesse in deze richting behoorlijk was afgenomen. Maar, doe toch maar en we kijken wel hoe het loopt. Ik begon onderaan met het invoeren van gegevens en het interviewen van familieleden van patiënten met trombose en bewezen trombofilie. Met de hulp van Jan-Leendert Brouwer vond ik mijn weg in de enorme databases van de DESCARTES en BEETHOVEN studies. Met de grote inzet van Nic Veeger, die mij al die jaren heeft gesteund in de ene na de andere analyse, groeide in de vrije uren uiteindelijk de eerste mooie studie in vrouwen met ernstige trombofilie. Ook de vervolgstudie in vrouwen met milde trombofilie kwam goed van de grond, maar op de dag dat we deze studie zouden insturen kwam het plotselinge bericht dat Jan ernstig ziek was en na een kort ziekbed is hij overleden op een veel te jonge leeftijd van 58 jaar. Na anderhalf jaar het promotietraject de rug toegekeerd te hebben, heb ik toch de draad weer opgepakt met de geweldige hulp van prof. Karina Meijer en prof. Han- neke Kluin-Nelemans, die de grote groep promovendi van Jan naar een goed einde wisten te leiden, met mij als laatste kandidaat. Ik heb uiteindelijk alle studies kunnen uitvoeren die we hadden besproken en ik hoop Jan dat je tevreden bent!

Zoals al hierboven genoemd, heb jij Karina, inmiddels professor Karina, aangebo- den om samen met professor Hanneke Kluin-Nelemans de begeleiding over te nemen en ik heb het niet beter kunnen treffen! Je hebt mij eigenlijk geheel vrijgela- ten in de keuzes van het verdere onderzoek, maar je was er als ik vastliep en hebt me vaak geholpen met het vinden van de juiste woorden en verder uitwerken van mijn soms in het begin wat warrige ideeën hoe iets aan te pakken.

Professor Kluin-Nelemans, beste Hanneke, u was er altijd als vast baken op de achtergrond met veel geduld. Met een nuchtere kijk en concreet commentaar als de prefinale versie van een artikel aan u werd voorgelegd. En toen het boekje bijna af was, heeft u in een razend tempo, gauw nog voor uw boot zou uitvaren, mijn in- en uitleiding geredigeerd en gecorrigeerd, zodat ik met een gerust hart de rest van de laatste loodjes kon gaan afronden. Het spreekt voor zich dat ik dit enorm heb gewaardeerd.

174 Dankwoord

Co-promotor dr. Veeger, beste Nic, zoals al gezegd, jij hebt mij tijdens al die jaren steeds maar weer geholpen met weer een nieuwe analyse, nog weer een aan- passing, toch nog weer iets extra’s er bij tot het uiteindelijk prachtige studies zijn geworden met mooie resultaten. Ik ben je daar heel dankbaar voor, want het heeft heel veel tijd gekost en tijd, vooral nadat je een nieuwe baan had aangenomen in Leeuwarden, kwam je vaak te kort!

Geachte leden van de leescommissie, prof. Huisman, prof. Rosendaal en prof. Scherjon, dank voor het beoordelen en goedkeuren van mijn proefschrift!

Prof. Middeldorp, beste Saskia, we spraken elkaar niet vaak, maar het had wel als resultaat dat jouw suggesties een heel gunstig effect hadden op de uiteindelijke versie van de stukken!

Prof. de Graeff, beste Pieter, je was er in het begin aan de zijlijn ook bij betrokken, maar liet de regie volledig aan Jan van der Meer. Af en toe gaf je wel aan dat het wel lang duurde en was dit de standaard grap op elk CBG-etentje, je bleef geïnteres- seerd en gaf me alle ruimte om deze ‘vrijetijdsbesteding’ voort te zetten. Maar het is nu eindelijk zover, het ei is gelegd!

Ook de andere medeauteurs, dr. Hamulyák, prof. Prins en prof. Büller wil ik graag bedanken voor hun bijdrage aan mijn artikelen. Ook helaas overleden professor Eskes wil ik speciaal hier noemen.

Jan-Leendert, dank voor jouw hulp en steun bij het opzetten van mijn eerste studie en het invoeren en gebruiken van gegevens in de DESCARTES- en BEETHOVEN database.

Sophie en Taco, fantastisch dat jullie met mij in zee gingen om een meet-analyse uit te voeren. Het is een prachtig stuk geworden!

Ina, naast jouw drukke baan heb je tijd vrijgemaakt om alle gegevens van de ‘pil- vrouwen’ in te voeren in een mooie database en hoewel het lang heeft geduurd, deze studie is nu ook gepubliceerd!

175 Hoofdstuk 9

En vanzelfsprekend dank aan alle patiëntes die belangeloos hebben meegewerkt; zonder hen geen studies!

Natuurlijk wil ik ook alle promovendi waar ik in de afgelopen jaren mee heb samen- gewerkt, gepubliceerd en congressen op prachtige locaties (Sidney, Geneve, Boston, Kyoto, Berlijn, Wenen) heb bezocht, bedanken voor de gezelligheid en het luisterend oor: Jan-Leendert, Ward, Liffert, Min Ki, Joop, Nienke, Willem, Marjan, Khan, Marieke, Inge, Vladimir, Renée, Anja, Nakisa, Hilde, Jaime en Sophie. En Margriet als vaste factor in deze lange rij van tijdelijke bewoners van de assistenten- kamer van de stolling. Dank ook voor het ‘aanleveren’ van patiëntes met pil- gere- lateerde trombose!

De secretaresses van het stollingssecretariaat bedankt voor de hulp bij de vele administratieve zaken en de researchverpleegkundigen voor het trouw aanmelden van nieuwe pildames!

Mijn eigen collega’s van de afdeling van het College ter beoordeling van Genees- middelen (CBG) in het UMCG en RadboudUMC: Marjon, Margje, Peter, Patrick, Frank, Amany, Else, Huub, jullie hebben mij er steeds weer van overtuigd dat het de moeite waard is om dit langdurig ‘project’ door te blijven zetten. Hierbij wil speciaal Marjon en Anouk bedanken voor hun peptalk omdat het niet allemaal liep zoals ik gehoopt had. En ook dank voor de steun van collega’s vanuit ons nieuwe hoofdkantoor in Utrecht.

Lieve vrienden en familie, na dit laatste jaar van drukte en laatste loodjes gaat de tijd weer aanbreken voor leuke dingen, zonder te hoeven nadenken of er wel tijd voor is! Dank voor jullie steun en interesse al die jaren!

Lieve mama en papa, dank voor jullie lieve zorgen, goede raad en jullie niet aflatend vertrouwen dat ik dit project met succes zal gaan afronden! Ik draag dit boekje aan jullie beiden op, fantastisch dat jullie er bij zijn; het heeft lang geduurd maar het is eindelijk zo ver!

176 Dankwoord

Lieve Rutger en Allard, we hebben al heel wat samen meegemaakt en ik heb jullie zien opgroeien als dwarse pubers, aartslui en niet vooruit te branden! Maar nu hebben jullie beiden sinds kort het nest verlaten en ik ben benieuwd wat de toekomst gaat brengen! Het zal wel wennen zijn dat jullie moeder niet meer ook nog ’s avonds aan het werk is!

Mijn paranimfen, Clarissa en Annemarie, geweldig dat jullie naast mij zullen staan! Lieve Annemarie, mijn beste vriendin, we hebben samen de studie geneeskunde doorgeworsteld! Wat een leuke tijd met veel drama, feesten en gezamen- lijke vakanties! Daarna kwam het echte leven: werken, man, kinderen, verhuizen… We hebben veel meegemaakt, maar onze vriendschap is nog altijd even hecht. Lieve Clarissa, mijn superschoonzus! Samen zijn we de enige dokters in een familie van economen en andere liberale denkers, die zich voortdurend afvragen waarom die dokters toch al maar moeten bijleren en publiceren! Maar jij hebt laten zien, dat wat er ook gebeurt, een promotie een prima doel kan zijn om je tijd in te steken ook al moet je er wat voor over hebben!

Als laatste Harry, mijn allerliefste. Dank voor je hulp op alle fronten, inclusief dit mooie boekje, en jouw niet aflatende pogingen om mij te laten zien dat er vele mooie dingen zijn op deze wereld die het waard zijn om te zien en tijd aan te besteden. En dat het leven kort is en dat je daarom veel moet genieten!

177 Hoofdstuk 9

About the author Liesbeth van Vlijmen is born on February 5th 1958 in Naarden. She attended the Willem de Zwijger lyceum in Bussum. She did not finish, but started to work as a dentist assistant. Few years later she passed State exams VWO. She started the Medicines study at the Vrije Universiteit at Amsterdam, but switched to the Rijks- universiteit Groningen at Groningen. After she passed her medical exam with specialty in clinical genetics, she started with a temporary project at the WHO department at Copenhagen and Groningen under supervision of professor Graham Dukes, who was at the time also an appointed professor of pharmacy in Groningen. Several years later she started a temporary job at the Dutch Medicine evaluation Board (MEB) to cover the time span until she could start training as a general practitioner. The MEB clinical assessment division is located at the University Medical Centre in Groningen (UMCG) and headed by Professor Pieter de Graeff. However, this job she found so interesting that she decided to change her carrier from patient care into regulatory medicine. She now is a senior clinical expert in gynaecology at the Dutch Medicine Evaluation Board, and expert in gynaecology at the European Medicines Agency. She is involved in the European registration of numerous hormonal contraceptives and other medicines in the area of women’s health. Next to this job, she started this PhD traject at the Department of Hematology, Division Thrombosis and Haemostasis under the supervision of professor Jan van der Meer. After his death, she restarted a year later under the supervision of professor Karina Meijer and professor Hanneke Kluin-Nelemans. Liesbeth has two sons, aged 19 and 22.

178