Risk factors for thrombosis in pregnancy

Best Practice & Research Clinical Haematology Vol. 16, No. 2, pp. 197 –210, 2003 doi:10.1053/ybeha.2003.257

5 Risk factors for thrombosis in pregnancy William M. Hague*

MD, FRCP, FRCOG

Senior Consultant Physician in Obstetric Medicine and Clinical Senior Lecturer Department of Obstetrics, University of Adelaide, Women’s and Children’s Hospital, North Adelaide, SA 5006, Australia

Gustaaf A. Dekker

MD, PhD, FRANZCOG , DCOG

Professor of Obstetrics and Gynaecology Department of Obstetrics, University of Adelaide, Lyell McEwin Health Service, Elizabeth Vale, SA 5118, Australia

Pregnancy is a hypercoagulable state affecting both the coagulation and the fibrinolytic systems. Any exacerbation of the pre-disposing factors for coagulation may well lead to a thrombotic event more often in pregnant women than in the general population. Arterial thrombosis is very rare in pregnancy. Pre-eclampsia may be a risk factor for the development of arterial disease in later life. Venous thromboembolism (VTE) in pregnancy, although still rare, is a major cause of maternal mortality. Risk factors, such as older age, increased weight and emergency Caesarean section, as well as acquired and genetic thrombophilia, often coexist and reinforce each other. Appropriate thromboprophylaxis needs to be considered and applied on an individual basis. Uteroplacental thrombosis provides a common pathophysiological link between various poor pregnancy outcomes, including recurrent miscarriage, stillbirth, placental abruption, fetal growth restriction and pre-eclampsia. Its significance depends on the gestational age. Acquired and genetic thrombophilia may be associated with such conditions, particularly in early-onset disease. More data are required to assess the significance of such thrombophilias in obstetric practice. Any treatment should be in the context of clinical trials. Key words: thrombosis; pre-eclampsia; maternal mortality; thrombophilia.

Pregnancy is well recognized as a hypercoagulable state. This follows a number of physiological changes in both the coagulation and the fibrinolytic systems, which may be viewed teleologically as mechanisms for minimizing blood loss resulting from placental separation at the time of birth. These changes include reduction in tissue plasminogen activator (tPA) concentrations and increased tPA inhibitor activity1, alteration in the concentration of various clotting factors with a marked increase in fibrinogen and a reduction in activity of protein S.2 Placental separation also removes a major source of tPA inhibitor.3 In addition, there is increased venous stasis during pregnancy and delivery, as well as the risk of damage to pelvic vessels at delivery. Given these natural tendencies, any exacerbation of the pre-disposing factors for coagulation may * Corresponding author. Tel.: þ61-8-8161-7000; Fax: þ61-8-8269-3200. E-mail address: [email protected] 1521-6926/03/$ - see front matter Q 2003 Elsevier Science Ltd. All rights reserved.

198 W. M. Hague and G. A. Dekker

well lead to a thrombotic event more often in pregnant women than in the general population. Arterial thrombosis is exceedingly rare in pregnancy. Data are hard to come by, and there are no large series recorded. In a series of women ðn ¼ 125Þ with myocardial infarction in pregnancy, of those found at autopsy or after angiographic examination ðn ¼ 68Þ to have a definite or probable intracoronary thrombosis in the absence of atherosclerosis ðn ¼ 14Þ; an equal proportion occurred antenatally and post-partum.4 Similarly, in a series of ischaemic carotid strokes in carefully studied young adults ðn ¼ 146Þ; 35% of the 65 women of child-bearing age were pregnant or puerperal, with similar numbers of events occurring in the puerperium as in the antenatal period.5 VTE, on the other hand, although still rare, is a major cause of maternal mortality in Australia, the U.K. and the U.S.A., occurring at an approximate rate of one death per 100 000 maternities.6 – 8 Placental thrombosis is commonly seen, and there is considerable variability as to its significance, which will depend on the gestation: a single thrombosis in a term placenta with an associated infarct is of no major concern, but a similar thrombosis in a 28-week placenta is likely to represent significant uteroplacental vascular disease.9 Major degrees of thrombosis leading to infarction are well recognized causes of fetal morbidity and mortality. The aetiology of arterial, venous and placental thrombosis includes environmental, acquired and genetic causes.

ARTERIAL THROMBOSIS In respect of early-onset arterial disease, well recognized risk factors include age, obesity, hypertension, family history of atherosclerosis, smoking, diabetes mellitus and dyslipidaemia. Antiphospholipid syndrome (both primary and secondary to other connective tissue disease, such as systemic lupus erythematosus) and hyperhomocysteinaemia have also been implicated in arterial as well as venous thrombosis. Hereditary deficiencies of the clotting proteins (protein C, protein S and antithrombin) and carriage of the factor V Leiden and prothrombin gene mutations do not in general predispose to arterial thrombosis. Variations, however, in the genes encoding fibrinogen, factor VII and plasminogen activator inhibitor (PAI-1) have all been identified in the complex interaction between genes and environment in the aetiology of arterial disease.10 In addition, common variants in the genes encoding enzymes involved in the metabolic pathways for homocysteine have been associated with an increased tendency to hyperhomocysteinaemia, especially in those with vitamin and folate deficiency, and may therefore be also associated with an increased risk of arteriopathy. Whether or not this association is causal is currently a matter for debate.11 In a case – control study of the obstetric histories of 46 parous women with earlyonset arterial disease and 53 parous controls, carefully selected to exclude those with risk factors for arteriopathy, the pregnancies in the arterial disease group were characterized by a higher incidence of pre-eclampsia (odds ratio (OR) 13.2: 95% confidence interval (CI) 3.94 –44.50) and non-hypertensive pregnancy complications (small for gestational age, pre-term delivery, loss after 20 weeks’ gestation, placental abruption) combined (OR 2.8: 95% CI 1.37 – 5.76).12 Women in the arterial disease group had higher homocysteine values than those in the control group ðP , 0:01Þ; and those women with homocysteine values . 11.1 mmol/l (the highest decile) had higher rates of pre-eclampsia (OR 7.0: 95% CI 1.17 –48.63). The incidence of thrombophilia was not otherwise increased in the arterial disease group. Further studies are necessary

Risk factors for thrombosis in pregnancy 199

to explore whether the hyperhomocysteinaemia characteristic of both vascular disease and disorders characterized by uteroplacental thrombosis may reflect a primary underlying endothelial dysfunction, first presenting during pregnancy. VENOUS THROMBOSIS Maternal mortality from VTE has been falling over the last 40 –50 years. The triennial Confidential Enquiries into Maternal Mortality in England, and subsequently in the UK13, and similar data from the U.S.A. and Australia, show that the rate has quartered. In the U.K., there was a doubling in the rate of deaths from VTE per million maternities from three in the triennium 1985 –1987 to six in the triennium 1994– 1996, perhaps due to better ascertainment or perhaps to a rise in the average age of pregnant women, but the rate has fallen again to less than two per million in the last reported triennium (Figure 1). The authors of the report suggest that this may have been due to the publication and introduction of guidelines on thromboprophylaxis for women undergoing Caesarean section from the Royal College of Obstetricians and Gynaecologists in 1995.14 The maternal mortality rate from VTE is much higher in older women: in the U.K., women over 39 years had a mortality rate of 1 per 3300 pregnancies, more than double that in younger women.13 VTE can occur at any time during pregnancy; its prevalence is approximately equally distributed between the three trimesters.15 Although two-thirds of events occur antenatally, the day-by-day risk is greatest in the first weeks after delivery.15 (Figure 2). Weight is also a significant risk factor for VTE: of 33 direct and late deaths attributed to VTE in the most recent triennium, 18 were in women who weighed over 80 kg or who were categorized as ‘obese’.13 Emergency caesarean section also tripled the risk of death from VTE, compared with vaginal delivery.13 (Figure 3). Other risk factors for VTE in pregnancy and postpartum include prolonged bed-rest and immobility, pre-eclampsia, nephrotic syndrome, current infection and other recent surgery, in addition to previous VTE and thrombophilia. These risk factors often coexist and reinforce each other. A risk assessment profile may be constructed, as suggested in the report from the RCOG.14 This consensus report recommended that:

Number of deaths

200

150

100

50

1997-99

1994-96

1991-93

1988-90

1985-87

1982-84

1979-81

1976-78

1973-75

1970-72

1967-69

1964-66

1961-63

1958-60

1955-57

0

Figure 1. Maternal mortality for VTE by triennia from U.K. Confidential Enquiries into Maternal Deaths 2001.13

200 W. M. Hague and G. A. Dekker

Rate per 1000 pregnant women

1.5

Antenatal

Postnatal

1

0.5

0 <35

>=35

<35

>=35

Maternal age (years) Figure 2. Maternal mortality rate by age and in relation to delivery from U.K. Confidential Enquiries into Maternal Deaths 2001.13

† all ‘at risk’ women should be monitored for symptoms and signs of VTE during the first week postpartum; † hydration should be maintained and early mobilization encouraged; † graduated compression stockings and/or calf stimulation should be used during and after Caesarean section in women at moderate risk (one or two risk factors); † in women at high risk (three or more risk factors), prophylaxis with low-molecularweight or unfractionated heparin should be utilized and continued for at least 5 days. Following the introduction of these guidelines, the rate of maternal mortality from VTE after Caesarean section fell from 2.1 to 1.4 per million.13 There are to date, however, no randomized trials to test the efficacy or benefit of such interventions, despite the demonstration in a pilot study of the feasibility of such.16 In women with previous VTE, the risk of recurrence will be influenced by a number of factors, including whether the index event was spontaneous or provoked (e.g. by Rate per 1000 pregnant women

1.5

1

Vaginal delivery Elective LSCS Emergency LSCS

0.5

0

<35

>=35 Maternal age (years)

Figure 3. Maternal mortality rate by age and mode of delivery from U.K. Confidential Enquiries into Maternal Deaths 2001.

Risk factors for thrombosis in pregnancy 201

trauma or surgery or pregnancy), the presence or absence of a family history of VTE, the presence of a known thrombophilia or whether there has been more than one episode of VTE. The decision as to whether VTE prophylaxis is required throughout pregnancy or only postpartum may be based on this information.17 (Table 1). The McMaster group observed prospectively a cohort of pregnant women with a single previous episode of VTE, withholding heparin thromboprophylaxis until after delivery, unless there were a documented recurrence of VTE, and blinding information as to underlying thrombophilia.18 Three of 125 women had such a recurrence (2.4%, 95% CI 0.2 – 6.9%). There were no recurrences in the 44 women who had no evidence of thrombophilia and whose previous episode of thrombosis was associated with a temporary risk factor (Table 2). Three of the 51 women with abnormal laboratory test results or a previous episode of thrombosis that was idiopathic had an antepartum recurrence of VTE (5.9%, 95% confidence interval 1.2 –16.2). Interestingly, of the 95 women from whom blood was taken on entry into the study, 48 (51%) had one or more abnormalities, including antithrombin deficiency in three and protein C deficiency in three. None of the women with these ‘more severe’ deficiencies developed recurrent thrombosis either during or after pregnancy (Table 3). No data were, however, provided as to the pregnancy outcomes of women in this study. VTE is a multifactorial disease, in many cases developing as a result of a thrombotic tendency (a thrombophilia) interacting with acquired factors, such as pregnancy.19,20 There are several ways in which such a thrombophilia may require consideration during pregnancy: † previous personal thromboembolic disease and known thrombophilia; † no previous VTE but strong family history (FH), i.e. one or more first-degree relatives, and known thrombophilia; † no previous VTE, strong FH, no previous investigations; † no previous VTE, weak FH, i.e. incidental finding of thrombophilia in a family member; † no personal or FH of VTE, but known thrombophilia detected following screening, for example, after obstetric complications or before commencing the combined oral contraceptive pill. The known causes of familial thromboembolism differ in their risk of associated thrombosis.21 The prevalence of such thrombophilic disorders varies between populations.22 A combination of any two or more inherited factors substantially increases the risk of thromboembolism. Table 1 also summarizes the risk profiles of the various genetic thrombophilias in respect of thromboembolism in pregnancy, and offers guidelines for therapy. The presence of a lupus anticoagulant or of moderate to strongly positive titres of anticardiolipin antibody (ACA) is a strong risk factor for recurrent VTE23, especially in pregnancy.24 On the other hand, it is contentious whether women with a positive lupus anticoagulant and/or ACA with no previous history of VTE, for example, women with systemic lupus erythematosus, are at significantly increased risk of VTE in pregnancy. UTEROPLACENTAL THROMBOSIS A common pathophysiological link between various poor pregnancy outcomes, including recurrent miscarriage, stillbirth, placental abruption, fetal growth restriction

Recurrent VTE Personal history of VTE independent of family history Family history of VTE in one or more 1st-degree relatives Family history of VTE in a distant relative No family history of VTE

.20%

10–20%

Single spontaneous VTE

10–20%

3–10%

Single VTE associated with surgery, combined oral contraceptive pill, etc.

3–10%

,3%

AT (very rare)

C (rare)

S (rare)

FVL or PGM homozygous (uncommon)

FVL or PGM heterozygous (common)

.20%

10– 20%

10–20%

10–20%

3–10%

10– 20%

10– 20%

10–20%

10–20%

3–10%

10– 20%

10– 20%

3–10%

3–10%

,3%

10– 30%

10– 20%

,3%

,3%

,3%

Women whose estimated pregnancy risk is . 20%, should be considered for therapeutic anticoagulation throughout pregnancy and post partum; those with an estimated risk of 10– 20%, should be considered for prophylactic anticoagulation throughout pregnancy and post partum; those with an estimated risk of 3–10%, negotiable on a case-by-case basis until further data available; those with an estimated risk of ,3%, should be considered for postpartum or no prophylaxis. AT, antithrombin deficiency; C, protein C deficiency; S, protein S deficiency; FVL, G1691A mutation in the factor V gene (factor V Leiden) causing activated protein C resistance; PGM, G20210A mutation in the prothrombin (factor II) gene. a Adapted from McLintock C et al (2001, Current Problems in Obstetrics, Gynecology and Fertility 24:109– 152) with permission.

202 W. M. Hague and G. A. Dekker

Table 1. Estimated pregnancy-related risk of VTE in pregnant women with either a previous VTE and no identified thrombophilia, adjusted for family history, or with an established genetic thrombophiliaa.

Risk factors for thrombosis in pregnancy 203

Table 2. Recurrence rate for VTE in pregnancya.

Recurrent events total Antepartum Postpartum None Recurrence rate % (95% confidence interval)

Idiopathic abnormal tests

Temporary risk abnormal tests

Idiopathic normal tests

Temporary risk normal tests

2

2

2

0

1 1 8 20 (2.5–55.6)

1 1 13 13 (1.7–40.5)

1 1 24 7.7 (0.1– 25.1)

44 0 (0.0–8.0)

a

Reproduced from Brill-Edwards P et al (2000, New England Journal of Medicine 343:1439–1444) with permission.

and pre-eclampsia, is thrombosis in the uteroplacental circulation. Inadequate maternal vascular response to placentation has been described in pre-eclampsia and in intrauterine growth restriction.25,26 Typically, there is absence of trophoblast-induced vascular changes in the spiral arteries of the placental bed. Acute atherosis and arterial thrombosis may further compromise the reduced blood supply, manifest in the placenta as accelerated villous maturity and infarction. The weakened arterial wall may predispose the placenta to abruption. A recent study of women with severe pre-eclampsia or eclampsia, placental abruption, intrauterine growth restriction or unexplained stillbirth reported placental infarction or other placental pathology in the majority, but could not demonstrate a difference in the rate of such events between those women with and those without a thrombophilia.27 This study, however, was flawed by poor placental histopathology techniques and inadequate clinical details28, and larger more robust studies are required. Table 3. Thrombophilia and recurrence of VTE in pregnancya. Abnormal test FVL C, S, AT, APL PGM Transient low protein S

Recurrent DVT n ¼ 6

No recurrent DVT n ¼ 89

3 2b 1 1

8 9c 6 28

Odds ratio (95% CI) 10.0 (1.0–80.0) 4.4 (0.2–35.4) 2.8 (0.7–32.9) 0.4 (0.1–4.0)

DVT, deep venous thrombosis; FVL, G1691A mutation in the factor V gene (factor V Leiden) causing activated protein C resistance; C, protein C deficiency; S, protein S deficiency; AT, antithrombin deficiency; APL, antiphospholipid antibody; PGM, G20210A mutation in the prothrombin (factor II) gene; CLAb, cardiolipin antibody positive, LAC, lupus anticoagulant positive. a Reproduced from Brill-Edwards P et al (2000, New England Journal of Medicine 343:1439–1444) with permission. b 1 C, 1 S. c 3 AT, 2 C, 2 CLAb, 2 LAC.

204 W. M. Hague and G. A. Dekker

A small, but more critical, study was undertaken of 14 placentas from a cohort of women diagnosed retrospectively as having hyperhomocysteinaemia, following a recent history of pre-eclampsia, intrauterine fetal growth restriction, abruption or of thromboembolic disease.29 Most of the placental findings indicated abnormal placentation but these were not specific to maternal hyperhomocysteinaemia nor found in every placenta. Features observed included absence of trophoblast-induced physiological vascular changes, acute atherosis, intraluminal endovascular trophoblast in the third trimester, infarction, retroplacental haematoma formation and accelerated villous maturity. Uteroplacental vascular thrombosis was seen also. Hyperhomocysteinaemia produces endothelial damage, smooth muscle proliferation, impairment of nitric oxide-mediated vasodilatation and thrombogenesis in vitro.30,31 The pathogenic mechanisms for thrombosis in hyperhomocysteinaemia include an interaction between inhibition of tPA by homocysteine.32 Any disruption of the normal interactions between trophoblast and maternal tissues by homocysteine could result in defective trophoblast migration and explain the placental findings of inadequate placentation. In a further instructive case report of a woman with thrombophilia and carrying dizygotic twins, one twin inherited thrombophilic genes from both the father and mother, resulting in placental fetal thrombotic vasculopathy and intrauterine growth restriction, whereas its co-twin inherited only one such gene from its mother and was unaffected.33 Amsterdam investigators were the first to evaluate the association of severe early-onset pre-eclampsia with thrombophilia.34 A total of 110 women were tested at least 10 weeks post-partum for the presence of hyperhomocysteinaemia, protein C, protein S and antithrombin deficiency, activated protein C resistance, lupus anticoagulant, and cardiolipin IgG and/or IgM antibodies. As in earlier studies, chronic hypertension was the most common (39%) underlying disorder. Of the 85 women tested for coagulation disturbances, 21 (24.7%) had protein S deficiency. Of the 50 women tested for activated protein C resistance, 8 (16.0%) were positive. Of the 79 women tested for hyperhomocysteinaemia, 14 (17.7%) had a positive methionine loading test. Finally, of 95 women tested for cardiolipin antibodies, 27 (29.4%) had detectable IgG and/or IgM antibodies. In a subsequent study in a larger group of more than 300 women with a history of severe pre-eclampsia or eclampsia, the same authors were able to include a control group of 65 women with only uncomplicated pregnancies in the obstetric history.35 The main findings of this study are presented in Table 4. Another landmark study was published from Israel.36 These authors studied 110 women with obstetric complications and 110 carefully matched controls. The factor V Leiden mutation, homozygosity for the thermolabile C677T polymorphism of the gene for methylenetetrahydrofolate reductase (MTHFR) and the prothrombin gene G20210A mutation were found to be significantly more common in the women with obstetric complications. Overall, 52% of women with an obstetric complication had a thrombophilic mutation versus 19% in women with normal pregnancies. In a subgroup analysis, the prevalence of mutations in the three genes was 53% among the women with severe pre-eclampsia. In a more recent study in 63 consecutive patients with severe pre-eclampsia and 126 controls, the same authors found a thrombophilia (one or more of factor V Leiden, prothrombin gene mutation, MTHFR C677T, protein C/protein S/antithrombin deficiency or cardiolipin antibody) in 67% of women who had severe pre-eclampsia versus 11% in the control group.37 In an Australian study investigating a second MTHFR A1298C polymorphism, the authors found that the 677 wild type (CC) was less frequent, while the 1298 wild type (AA) was increased in

Risk factors for thrombosis in pregnancy 205

Table 4. Prevalence of haemostatic abnormalities in women with a history of severe pre-eclampsiaa.

S APCR FVL HHC ACA

Controls

Pre-eclampsia # 28 weeks

Pre-eclampsia . 28 weeks

6/65 (9.0%) 1/67 (1.5%) 1/67 (1.5%) 3/67 (4.5%) 5/67 (7.5%)

11/59 (19.0%) 9/50 (18.0%)b 4/50 (8.0%) 11/58 (19.0)c 17/62b (27.4%)

26/251 (10.0%) 23/234 (9.8%)c 13/234 (5.6%) 35/289 (10.4%) 50/259c (19.3%)

S, protein S deficiency; FVL, G1691A mutation in the factor V gene (factor V Leiden) causing activated protein C resistance (APCR); HHC, hyperhomocysteinaemia; ACA, detectable cardiolipin IgG and/or IgM antibodies. a Reproduced from van Pampus MG et al (1999, American Journal of Obstetrics and Gynecology 180:1146– 1150) with permission. b P , 0:005: c P , 0:05 compared with controls.

women with pre-eclampsia.38 Because the C677T polymorphism is in linkage dysequilibrium with the 1298 wild-type allele, we suggested that the latter could possibly be protective. Other studies, however, have not found a significant association of pre-eclampsia with genetic thrombophilia. In a relatively large British case– control study ðn ¼ 483Þ; no evidence was found for the presence of an association between either the factor V Leiden mutation or the MTHFR C677T polymorphism and preeclampsia.39 Although about 50% of these patients had severe pre-eclampsia, no data were provided on gestational age and birth weight. Perhaps more importantly, most patients had pre-eclampsia relatively late in pregnancy (. 34 weeks’ gestation). In another prospective study of nearly 600 nulliparous Irish women, a similar frequency of the factor V Leiden and of the MTHFR C677T polymorphism was found in the 12 women with pre-eclampsia and the nine patients with IUGR as in those women with an uncomplicated pregnancy.40 The sample size, however, was much too small to evaluate properly the association of any genetic polymorphism and severe early-onset pre-eclampsia, given its very low incidence (0.5 – 1%). In a recent Australian case –control study including women with familial pre-eclampsia and eclampsia, no association could be found with the prothrombin gene G20210A mutation.41 A recent American prospective cross-sectional study compared the maternal and fetal genotype frequencies of factor V Leiden, MTHFR C677T polymorphism, and prothrombin gene G20210A mutation in 110 patients with severe pre-eclampsia matched for gestational age to 97 normotensive pregnant women.42 There were no significant differences between patients and controls regarding frequency of maternal genotype. In addition, there was no association between fetal carriage of a thrombophilia and the development of pre-eclampsia. Findings were similar in both white ðn ¼ 47Þ and African –American ðn ¼ 63Þ women. The mean gestation, however, in the pre-eclampsia group was 34.3 ^ 5.1 weeks and in the control group 35.6 ^ 4.8 weeks, with no separation of those women with early-onset disease. The patient cohorts studied in the two Amsterdam studies, on the other hand, represented very selected groups of very sick patients with early-onset pre-eclampsia referred to two large level 3 perinatal centres draining a densely populated area in The Netherlands with about 40 000 births per year.

206 W. M. Hague and G. A. Dekker

In a recent systematic review of inherited thrombophilia in relation to pre-eclampsia, overall factor V Leiden was increased twofold in women with pre-eclampsia, but there was significant heterogeneity between the studies43 (Table 5). There was a similar increase for hyperhomocysteinaemia, and, to a much lesser extent, the MTHFR C677T polymorphism. There was no association with the prothrombin gene mutation G20210A with pre-eclampsia, apart from those women with severe disease (OR 2.3: 95% CI 1.0 –5.1). The rarity of antithrombin and protein C deficiencies and the small number of patients reported with pre-eclampsia in relation to these made appropriate analysis impossible. A European group reported on the relationship between heritable thrombophilic defects and fetal loss in a cohort of women with factor V Leiden or deficiency of antithrombin, protein C, or protein S.20 The authors studied 1384 women enrolled in the European Prospective Cohort on Thrombophilia (EPCOT). Of 843 women with thrombophilia, 571 had 1524 pregnancies; of 541 control women, 395 had 1019 pregnancies. The controls were partners of male members of the EPCOT cohort or acquaintances of cases. The frequencies of miscarriage (fetal loss at or before 28 weeks of gestation) and stillbirth (fetal loss after 28 weeks of gestation) were analysed jointly and separately. The risk of fetal loss was increased in women with thrombophilia (168/571 versus 93/395; OR 1.35 [95% CI 1.01 – 1.82]). The OR was higher for stillbirth than for miscarriage (3.6 [95% CI 1.4 – 9.4] versus 1.27 [95% CI 0.94 –1.71]). The highest OR for stillbirth was in women with combined defects (14.3 [95% CI 2.4 –86.0]) compared with 5.2 (95% CI 1.5 – 18.1) in antithrombin deficiency, 2.3 (95% CI 0.6 – 8.3) in protein C deficiency, 3.3 (95% CI 1.0 – 11.3) in protein S deficiency, and 2.0 (95% CI 0.5 – 7.7) with factor V Leiden. The corresponding ORs for miscarriage in these subgroups were 0.8 (95% CI 0.2 – 3.6), 1.7 (95% CI 1.0 – 2.8), 1.4 (95% CI 0.9 –2.2), 1.2 (95% CI 0.7 – 1.9) and 0.9 (95% CI 0.5 –1.5) respectively. It was concluded that women with familial thrombophilia, especially those with combined defects or antithrombin deficiency, have an increased risk of fetal loss, with the relative risk of stillbirth greater than that for miscarriage. No data were provided on the incidence of pre-eclampsia in this study. The plethora of data in respect of thrombophilia and fetal loss that followed this initial study were very heterogeneous in their selection criteria. Several had no control data: these were excluded from the previously mentioned systematic review43, which found a significant increase in the prevalence of factor V Leiden, hyperhomocysteinaemia, the prothrombin gene G20210A mutation and protein S deficiency in women with fetal loss (recurrent first-trimester miscarriage, second-trimester fetal loss and stillbirth) across the board (Table 5), with no significant difference among those with antithrombin and protein C deficiency or who carried the C677T polymorphism of MTHFR. When the subgroups with recurrent early loss alone were analysed, hyperhomocysteinaemia remained unchanged (the same data), but factor V Leiden, prothrombin gene G20210A mutation and protein S deficiency were no longer significant factors, whereas among those with late fetal loss, these three factors retained their significance (Table 5). There are no controlled studies of hyperhomocysteinaemia in late fetal loss. SUMMARY The various genetic, acquired and environmental factors and their interaction in the promotion of thrombosis during pregnancy have been reviewed. Arterial disease in

Table 5. Prevalence of inherited thrombophilias in women with various pregnancy complicationsa.

Type of thrombophilia

Pregnancy complication

Factor V Leiden

Pre-eclampsia Fetal loss Late fetal loss

Hyperhomocysteinaemia

Pre-eclampsia Fetal loss Late fetal loss Pre-eclampsia

Prothrombin gene G20210A

Fetal loss Late fetal loss Pre-eclampsia Fetal loss Late fetal loss

Protein S deficiency

Pre-eclampsia Fetal loss Late fetal loss

a

Controls

10.0 (1569) 9.4 (0–26.5) 7.8 (2519) 8.0 (0–28) 9.5 (1220) 22 (6.5–31.3) 13.3 (309) 21.1 (12.1–30.0) 20.0 (137) 19.1 (17.1–21) No data 12.0 (2157) 12.5 (0.6–29.8) No data No data 4.2 (669) 5.3 (0–8.8) 4.5 (738) 3.6 (0–11.6) 3.3 (334) 5 (0–13) 2.2 (226) 4.0 (0–8.0) 2.4 (778) 4.7 (0–17.4) 4.7 (232)

5.0 (2197) 5.2 (1.5–11.1) 3.3 (3174) 3.9 (0 –10) 3.7 (1227) 4.2 (0 –8) 4.0 (100) 3.8 (3.0–4.5) 3.5 (108) 3.5 (2.4–4.5) No data 10.3 (3056) 10.8 (0 –18.6) No data No data 2.9 (809) 3.0 (1.1–3.7) 2.0 (1668) 3.0 (1 –4.5) 2.1 (961) 3 (1 –3.2) 0.4 (289) 0.4 (0 –0.8) 1.0 (681) 0.2 (0 –9) 0.2 (464)

Reproduced from McLintock C et al (2001, Current Problems in Obstetrics, Gynecology and Fertility 24:109– 152) with permission.

Pooled odds ratio (95% CI) 2.2 (1.7–3.0) 2.7 (2.0–3.6) 2.8 (1.8–4.3) 2.2 (1.5–3.4) 6.8 (2.1–22.4)

1.3 (1.1–1.6)

1.7 (0.9–3.0) 2.1 (1.3–3.6) 2.1 (1.0–4.6) – 5.0 (2.0–12.2) 22 (2.8 –170)

Risk factors for thrombosis in pregnancy 207

MTHFR C677T

Prevalence of thrombophilia weighted mean% ðnÞ median% (range%) Cases

208 W. M. Hague and G. A. Dekker

pregnancy is rare but there is evidence that pre-eclampsia, a disease characterized by endothelial activation, may be associated with its occurrence in later life. Venous thrombosis remains a major cause of maternal mortality. Those women who are older or obese are at increased risk during pregnancy, as are those undergoing Caesarean section, particularly as an emergency. Acquired or genetic thrombophilia also increases the risk of venous thrombosis during pregnancy, as well as the risk of poor pregnancy outcome associated with uteroplacental thrombosis. There are, however, no level 1 or 2 data to establish the benefit or safety of using anticoagulant therapy during pregnancy to reduce these risks. Each pregnant woman needs a careful assessment of risk, including enquiry as to a family history of thrombosis, and appropriate prophylaxis applied on a case-by-case basis, preferably within the context of a randomised controlled trial. More research is needed to answer the many questions as to the interplay of the individual components of the coagulation process in the development of clinical disease affecting both mother and fetus, and as to how such disease can best be avoided. Practice points † prophylactic doses of LMWH can be used to reduce the risk of recurrent thromboembolic events in pregnancy. The regimen employed will depend on the previous history, the family history and the presence of risk factors, including the genetic and acquired causes of thrombophilia † not all women with a genetic thrombophilia will develop a thrombosis during pregnancy † the presence of a thrombophilia does not necessarily increase the risk of a poor obstetric outcome

Research agenda † prevention of pregnancy-associated VTE: risks and benefits of thromboprophylaxis during pregnancy and postpartum for specific groups † prevention of adverse pregnancy complications related to placental insufficiency † randomized studies to determine efficacy of thromboprophylaxis and/or vitamin supplementation in improving subsequent pregnancy outcome in women with specific pregnancy complications and an underlying thrombophilia † a register to determine the clinical significance of thrombophilias in particular groups of patients

REFERENCES 1. Kruithof EK, Tran-Thang C, Gudinchet A et al. Fibrinolysis in pregnancy: a study of plasminogen activator inhibitors. Blood 1987; 69: 460– 466. 2. Letsky E. Coagulation Problems During Pregnancy. Edinburgh: Churchill Livingstone, 1985. 3. Astedt B, Lindoff C & Lecander I. Significance of the plasminogen activator inhibitor of placental type (PAI2) in pregnancy. Seminars in Thrombosis and Hemostasis 1998; 24: 431–435.

Risk factors for thrombosis in pregnancy 209 4. Roth A & Elkayam U. Acute myocardial infarction associated with pregnancy. Annals of Internal Medicine 1996; 125: 751–762. 5. Cross JN, Castro PO & Jennett WB. Cerebral strokes associated with pregnancy and the puerperium. British Medical Journal 1968; 3: 214 –218. * 6. Greer IA. Thrombosis in pregnancy: maternal and fetal issues. Lancet 1999; 353: 1258–1265. 7. Maternal mortality committee, Maternal Deaths in Australia 1991–1993. Canberra: NHMRC, 1998. * 8. Toglia M & Weg J. Current concepts: venous thromboembolism during pregnancy. New England Journal of Medicine 1996; 335: 108–114. 9. Porter H. The role of placental examination as a perinatal investigation. In Kingdom J, Jauniaux E & O’Brien P (eds) The placenta: Basic Science and Clinical Practice. London: RCOG Press, 2000, pp 109– 120. 10. Grant PJ & Humphries SE. Genetic determinants of arterial thrombosis. Baillie`re’s Clinical Haematology 1999; 12: 505– 532. * 11. Brattstrom L & Wilcken DE. Homocysteine and cardiovascular disease: cause or effect? American Journal of Clinical Nutrition 2000; 72: 315– 323. 12. Plantinga J, Hague W, Fitridge R et al. Obstetric history and thrombophilia in women with early onset atherosclerosis. Annual Scientific Meeting of the Australasian Society for the Study of Hypertension in Pregnancy; 2002; Akaroa, Christchurch, New Zealand 2002;. * 13. Department of Health, Why Mothers Die. Fifth Report on Confidential Enquiries into Maternal Deaths in the United Kingdom 1997– 1999, London: RCOG Press, 2001. 14. Royal College of Obstetricians and Gynaecologists, Report of the RCOG Working Party on Prophylaxis against Thromboembolism in Gynaecology and Obstetrics, London. ; 1995. 15. Ray JG & Chan WS. Deep vein thrombosis during pregnancy and the puerperium: a meta-analysis of the period of risk and the leg of presentation. Obstetric and Gynecological Survey 1999; 54: 265 –271. 16. Burrows RF, Gan ET, Gallus AS et al. A randomised double-blind placebo controlled trial of low molecular weight heparin as prophylaxis in preventing venous thrombolic events after caesarean section: a pilot study. BJOG: an International Journal of Obstetrics and Gynaecology 2001; 108: 835 –839. 17. McColl MD, Walker ID & Greer IA. The role of inherited thrombophilia in venous thromboembolism associated with pregnancy. British Journal of Obstetrics and Gynaecology 1999; 106: 756–766. * 18. Brill-Edwards P, Ginsberg J, Gent M et al. Safety of withholding antepartum heparin in women with a previous episode of venous thromboembolism. New England Journal of Medicine 2000; 343: 1439–1444. 19. Rosendaal FR. Venous thrombosis: a multicausal disease. Lancet 1999; 353: 1167–1173. * 20. Preston FE, Rosendaal FR, Walker ID et al. Increased fetal loss in women with heritable thrombophilia. Lancet 1996; 348: 913–916. 21. Gerhardt A, Scharf RE, Beckmann MW et al. Prothrombin and factor V mutations in women with a history of thrombosis during pregnancy and the puerperium. New England Journal of Medicine 2000; 342: 374–380. 22. Seligsohn U & Lubetsky A. Genetic susceptibility to venous thrombosis. New England Journal of Medicine 2001; 344: 1222–1231. 23. Levine JS, Branch DW & Rauch J. The antiphospholipid syndrome (see comments). New England Journal of Medicine 2002; 346: 752 –763. 24. Erkan D, Yazici Y, Peterson MG et al. A cross-sectional study of clinical thrombotic risk factors and preventive treatments in antiphospholipid syndrome. Rheumatology 2002; 41: 924–929. 25. Khong TY, De Wolf F, Robertson WB & Brosens I. Inadequate maternal vascular response to placentation in pregnancies complicated by pre-eclampsia and by small-for-gestational age infants. British Journal of Obstetrics and Gynaecology 1986; 93: 1049–1059. 26. Meekins JW, Pijnenborg R, Hanssens M et al. A study of placental bed spiral arteries and trophoblast invasion in normal and severe pre-eclamptic pregnancies. British Journal of Obstetrics and Gynaecology 1994; 101: 669 –674. 27. Mousa HA & Alfirevic Z. Do placental lesions reflect thrombophilia state in women with adverse pregnancy outcome? (see comments). Human Reproduction 2000; 15: 1830–1833. 28. Khong T, Moore L & Hague W. Thrombophilias and adverse pregnancy outcome (letter). Human Reproduction 2001; 16: 395–396. 29. Khong T & Hague W. The placenta in hyperhomocysteinaemia. British Journal of Obstetrics and Gynaecology 1999; 106: 273–278. 30. Chambers JC, McGregor A, Jean-Marie J & Kooner JS. Acute hyperhomocysteinaemia and endothelial dysfunction (see comments). Lancet 1998; 351: 36–37. 31. Tawakol A, Omland T, Gerhard M et al. Hyperhomocyst(e)inemia is associated with impaired endothelium-dependent vasodilation in humans [erratum appears in Circulation 2000;101:E116]. Circulation 1997; 95: 1119–1121. 32. Harpel PC, Zhang X & Borth W. Homocysteine and hemostasis: pathogenic mechanisms predisposing to thrombosis. Journal of Nutrition 1996; 126(supplement 4): 1285S–1289S.

210 W. M. Hague and G. A. Dekker * 33. Khong TY & Hague WM. Biparental contribution to fetal thrombophilia in discordant twin intrauterine growth restriction. American Journal of Obstetrics and Gynecology 2001; 185: 244 –245. * 34. Dekker G, de Vries J, Doelitzsch P et al. Underlying disorders associated with severe early-onset preeclampsia. American Journal of Obstetrics and Gynecology 1995; 173: 1042–1048. 35. van Pampus MG, Dekker GA, Wolf H et al. High prevalence of hemostatic abnormalities in women with a history of severe preeclampsia. American Journal of Obstetrics and Gynecology 1999; 180: 1146–1150. * 36. Kupferminc M, Eldor A, Steinman N et al. Increased frequency of genetic thrombophilia in women with complications of pregnancy (see comments) [erratum appears in New England Journal of Medicine 1999; 341:384). New England Journal of Medicine 1999; 340: 9–13. 37. Kupferminc MJ, Fait G, Many A et al. Severe preeclampsia and high frequency of genetic thrombophilic mutations. Obstetrics and Gynecology 2000; 96: 45–49. 38. Hague B, Wiltshire E, Nelson P et al. Polymorphisms of methylenetetrahydrofolate reductase in women with pre-eclampsia. Hypertension in Pregnancy 2000; 19(supplement 1): 23. 39. O’Shaughnessy KM, Fu B, Ferraro F et al. Factor V Leiden and thermolabile methylenetetrahydrofolate reductase gene variants in an East Anglian preeclampsia cohort. Hypertension 1999; 33: 1338–1341. 40. Murphy RP, Donoghue C, Nallen RJ et al. Prospective evaluation of the risk conferred by factor V Leiden and thermolabile methylenetetrahydrofolate reductase polymorphisms in pregnancy. Arteriosclerosis Thrombosis and Vascular Biology 2000; 20: 266 –270. 41. Higgins JR, Kaiser T, Moses EK et al. Prothrombin G20210A mutation: is it associated with pre-eclampsia? Gynecologic and Obstetric Investigation 2000; 50: 254 –257. 42. Livingston JC, Barton JR, Park V et al. Maternal and fetal inherited thrombophilias are not related to the development of severe preeclampsia. American Journal of Obstetrics and Gynecology 2001; 185: 153–157. * 43. McLintock C, North R & Dekker G. Inherited thrombophilias: implications for pregnancy-associated venous thromboembolism and obstetric complications. Current Problems in Obstetrics, Gynecology and Fertility 2001; 24: 109 –152.