Diagnosis and management of heritable thrombophilias

Factor V Leiden

The most common heritable thrombophilia within the UK population and identified in clinical practice is factor V Leiden. As a result, knowledge about the epidemiology of factor V Leiden is greater than for the other heritable thrombophilias. The prevalence varies considerably according to ethnic origin. Within white populations of European origin, the prevalence is approximately 3-7% and even higher in some areas, such as southern Sweden.7 8 To explain the relatively high prevalence within these populations it has been suggested that the heterozygous state may confer a survival advantage, possibly because of reduced bleeding with childbirth or trauma; however, this is unconfirmed. A similar prevalence is also seen in populations of the Mediterranean and Middle East. In contrast, factor V Leiden is essentially not found in South East Asia, Japan, and sub-Saharan Africa, or in indigenous Australian populations. The predicted prevalence of people homozygous for factor V Leiden in European populations would be around 1 in 1600.

The factor V Leiden polymorphism is a mis-sense mutation. The result of this mutation is that activated factor V, a clotting factor, is inactivated around 10-fold more slowly than normal by activated protein C (so called APC resistance) and this leads to increased thrombin generation. The molecular basis of this activated protein C resistance laboratory phenotype was initially identified by investigators from Leiden University in the Netherlands.9

Compared with non-carriers, heterozygous carriers of factor V Leiden have a threefold to fivefold increased risk of VTE throughout life.10 11 12 The absolute risk, however, is low, with only 5% of such people developing an episode of VTE by age 65 years.13 14 This is a reflection of the background risk of VTE within the population as a whole, which is age dependent but on average affects 1 per 1000 people each year. Approximately 20% of patients with a first unprovoked episode of VTE are heterozygous carriers of factor V Leiden. For reasons that remain unexplained, factor V Leiden seems to be a stronger risk factor for deep vein thrombosis than for pulmonary embolism.15 Those who are homozygous for factor V Leiden are at a 10-fold or greater risk of VTE than non-carriers.10 11 12 However, because such people are uncommon within the population, they do not contribute substantially to the overall burden of VTE.

Interactions have been found with other risk factors for VTE, in particular combined hormonal contraceptives, hormone replacement therapy, and pregnancy. For example, data from a case-control study suggested that the risk of deep vein thrombosis in users of combined hormonal contraceptives was increased fourfold compared with non-users and that the risk associated with combined hormonal contraceptive use in heterozygous carriers of factor V Leiden was multiplicative, resulting in an approximately 30-fold increased risk compared with non-users who did not carry the factor V Leiden gene.16 Increased risks have also been observed in studies of hormone replacement therapy.17 However, the absolute (as opposed to relative) risk of VTE while taking combined hormonal contraceptives in the 5% of the population who carry the factor V Leiden gene remains too small (about 15-30 per 10 000 women each year) to justify widespread screening.18 19

Prothrombin gene mutation

The prothrombin gene mutation is the second most commonly identified heritable thrombophilia, with around 1-2% of white populations of European origin being heterozygous carriers. The mutation is associated with a twofold to fourfold increased risk of VTE throughout life, slightly less overall than the risk associated with factor V Leiden.11 12 20 21 22 It is present in around 5% of people with a first unprovoked episode of VTE. Prothrombin levels are increased by about 30% in heterozygotes and 70% in homozygotes.

Antithrombin deficiency

Heterozygous mutations of the gene for antithrombin, previously called antithrombin III, result in antithrombin deficiency leading to reduced inhibition of factor Xa and thrombin and therefore increased thrombin generation and activity. Two major types of antithrombin deficiency occur: type I (a quantitative defect with reduced production of a normal molecule) and type II (a qualitative defect in which the molecule itself is abnormal). Type II antithrombin deficiency can be subdivided according to the site of the molecular defect: heparin binding site (which carries a lower risk of thrombosis), reactive site, or both. In clinical practice it is difficult to distinguish between the subtypes of type II antithrombin deficiency, and specialist advice should be considered.

Type I antithrombin deficiency is rare (0.02% of the population) and homozygous mutations are incompatible with life.23 The degree of increased risk of VTE associated with heterozygous antithrombin deficiency remains the subject of debate. This reflects in part on the low prevalence of antithrombin deficiency within the population and therefore the uncertainties around any estimates. Different study designs (for example, case-control versus family cohort) have reached different conclusions, with estimated relative risks varying from similar to those of factor V Leiden to risks that are around fivefold higher.24 25 26 The absolute risk of a first episode of VTE in previously asymptomatic people with antithrombin deficiency in families in whom a proband with VTE is antithrombin deficient has been estimated at around 2% yearly.26 27 An antithrombin concentrate is available and some advocate its use in certain high risk patients with antithrombin deficiency and previous VTE who are undergoing procedures where anticoagulation has to be withheld; such situations are rare and expert advice should be sought.

Protein C and protein S deficiency

Protein C and protein S are vitamin K dependent glycoproteins. Heritable deficiencies of either lead to impaired inactivation of factor Va and factor VIIIa and thereby increased thrombin generation. Protein C deficiency is classified into type I (quantitative) and type II (qualitative) defects, although unlike with antithrombin deficiency the thrombotic risk between the different types does not differ. Similarly, type I and type II abnormalities of protein S exist, although type II is rare. A type III deficiency characterised by a normal level of total protein S but a reduced level of the free active protein has been described and probably represents defective production (type I deficiency) not recognised with the less sensitive assay for total protein S (which therefore incorrectly suggests a normal result).4

About 0.3% of the population has protein C deficiency.28 Estimates for risk of VTE associated with protein C deficiency vary from those that are similar to factor V Leiden to those that are closer to antithrombin deficiency. Unlike antithrombin deficiency, homozygous protein C deficiency, although rare, is compatible with life, typically presenting with neonatal purpura fulminans or cerebral vein thrombosis.23

The prevalence of protein S deficiency is uncertain but is probably around 0.1%.28 The risk of VTE associated with protein S deficiency is uncertain: family studies have suggested risks similar in size to those observed in protein C deficiency,26 27 but population studies suggest a lower risk.29

Risk factors for recurrence

For the index case an increased risk of recurrence is predicted by a persisting risk factor (for example, active cancer) or an unprovoked event.36 37 38 The risk of recurrence is increased after a first unprovoked episode of VTE if the patient is male39 40 41 and/or the D-dimer level is increased after completing anticoagulation,40 41 42 and the identification of such risk factors makes a stronger case for long term anticoagulation for secondary prevention. A family history of thrombosis does not increase the risk of recurrence,43 and prospective cohort studies have shown that heritable thrombophilia does not usefully predict recurrence.44 45 A systematic review looking at the two most common heritable thrombophilias found that the odds ratio for recurrence in association with heterozygosity for factor V Leiden was 1.56 (95% confidence interval 1.14 to 2.12) and with heterozygosity for the prothrombin gene mutation was 1.45 (0.96 to 2.21).46 These increases in risk are not thought of sufficient size to justify long term anticoagulation if such treatment would not otherwise have been given. It has been suggested that combined defects carry a higher risk of recurrence, but patients who are homozygous for factor V Leiden or the prothrombin gene mutation or double heterozygous carriers of factor V Leiden and the prothrombin gene mutation do not seem to have a high risk of recurrence.46 47 Deficiencies of antithrombin, protein C, or protein S are often described as high risk thrombophilias.26 The risk of recurrence had been claimed to be high in studies of families prone to thrombosis,48 but in patients who are not selected from such families the relative risk of recurrence seems to be <2.44 45 49 One recent study in patients with unprovoked VTE suggested that antithrombin deficiency was associated with a twofold increase in the risk of recurrence after stopping anticoagulation50; this requires confirmation. No randomised trials have assessed the benefit of testing for heritable thrombophilias on the risk of recurrent VTE.51 Additionally, heritable thrombophilias do not predict the risk of either the post-thrombotic syndrome after a deep vein thrombosis or mortality.52 53

The observation that heritable thrombophilias are weaker risk factors for recurrent rather than first VTE events to some seems counterintuitive. The simplest explanation is that in the setting of first episodes the comparison is between those who have had events and those who have not and, as expected, identifiable thrombophilias are found more commonly in the former than in the latter. In contrast, in the setting of recurrences, the comparison is between those who have a detectable heritable thrombophilia and those who are likely to have a thrombophilic tendency (to cause the VTE in the first place) that cannot be easily identified with current laboratory tests.

Duration of anticoagulation

Heritable thrombophilias would seem to have only a minor role in the duration of anticoagulant treatment.54 Recent British guidelines5 6 are against indiscriminate testing of patients presenting with VTE and suggest testing should be considered only in selected patients such as those with unprovoked events and a family history of thrombosis who are planning to stop anticoagulation.

A separate situation is seen in venous thrombosis in unusual sites. Here there are even fewer data and although some suggest that the presence of heritable thrombophilia should result in a longer course of anticoagulation,55 there is no good evidence base for such recommendations.56

Pregnancy and contraception

Combined hormonal contraceptives are contraindicated in women with a previous episode of VTE,57 and this is unaffected by the presence or absence of a heritable thrombophilia. However, the levonorgestrel releasing intrauterine device and low dose oral progestogen preparations seem to be safe.58 In the context of pregnancy, all women with a previous episode of VTE will be offered thromboprophylaxis for six weeks post partum.59 60 Women with a history of unprovoked or oestrogen provoked VTE will also be offered antenatal thromboprophylaxis.59 60 Those with a previous event provoked by a minor transient risk factor would not routinely have antenatal thromboprophylaxis, but they should then be tested for a thrombophilic tendency as a positive test result may change this decision.60 The presence or absence of a heritable thrombophilia does not affect the dose of low molecular weight heparin used for thromboprophylaxis in pregnancy, with the exception of women with antithrombin deficiency, where guidelines from the Royal College of Obstetricians and Gynaecologists recommend higher (intermediate or treatment) doses, although the equivalent North American guidelines do not draw this distinction.59 60

Pregnancy, contraception, and hormone replacement therapy

Specific problems are associated with contraception and pregnancy in female relatives of childbearing age, and hormone replacement therapy in older women. According to a recent European review, combined hormonal contraceptives are estimated to increase the annual risk of VTE from 2 in 10 000 women to approximately 5-10 in 10 000 women (depending on the preparation).61 The risks seem to be higher in women with heritable thrombophilias from thrombosis prone families and have been estimated at 0.2% per year for the prothrombin gene mutation and 0.5% per year for factor V Leiden, and up to 4% per year for women with deficiencies of antithrombin, protein C, or protein S.58 The recent checklist for prescribers issued by the Medicines and Healthcare products Regulatory Agency after review of combined hormonal contraceptives by the European Medicines Agency states that combined hormonal contraceptives should not be prescribed if women have a predisposition for a blood clotting disorder, and the suitability of a combined hormonal contraceptive should be discussed with women if a close relative (parent or sibling) has had a thromboembolic event at a young age (<50 years).57 Since women with a history of VTE in a first degree relative before 50 years of age should consider an alternative form of contraception to combined hormonal contraception, thrombophilia testing generally has little role. The exception might be where a high risk thrombophilia (deficiency of a naturally occurring anticoagulant or homozygosity or double heterozygosity for defects) is identified in the symptomatic relative, when its absence in women might allow combined oral contraceptives to be considered. For the most part, however, while the finding of a heritable thrombophilia would mean that combined hormonal contraceptives should not be prescribed, the absence of a detectable tendency for thrombophilia in the setting of a family history of VTE would not provide sufficient reassurance to enable combined hormonal contraceptives to be prescribed without discussion of the risks and alternatives.

Women with a history of unprovoked VTE in a first degree relative should not use oral hormone replacement therapy, but if it is considered essential they can be offered transdermal therapy, which seems not to carry a clinically significant risk of VTE.5 62 63 64

Pregnant women, or women planning a pregnancy, who have a first degree relative with unprovoked or oestrogen associated VTE should be assessed for their risk of pregnancy associated venous thrombosis. This should include a thrombophilia screen, as the Royal College of Obstetricians and Gynaecologists guidelines suggest antenatal thromboprophylaxis if such women have antithrombin deficiency or homozygosity or double heterozygosity for heritable thrombophilia,60 although North American guidelines59 would only recommend antenatal thromboprophylaxis in the presence of homozygosity for factor V Leiden or the prothrombin gene mutation. Postpartum thromboprophylaxis should be considered in all women with a family history of VTE and a known thrombophilia,59 60 particularly if there are additional risk factors.