Prothrombin 20210 G→A, MTHFR C677T mutations in women with venous thromboembolism associated with pregnancy
Correspondence: Dr M. D. McColl, Department of Haematology, 3rd Floor MacEwen Building, Glasgow Royal Infirmary University NHS Trust, Castle Street, Glasgow G4 OSF, UK.
Over 50 unselected women with maternal venous thromboembolism were screened for the prothrombin 20210 G→A and MTHFR C677T mutations, in addition to screening for other thrombophilias. The prevalence of thrombophilia in these women was compared with its prevalence in the general population in our area. The prothrombin (OR 4.4; 95% CI 1.2-16) and factor V Leiden (OR 4.5; 95% CI 2.1-14.5) mutations were more common in our patients, compared with the general population, whereas women homozygous for the C677T mutation in the methylene tetrahydrofolate reductase gene (OR 0.45; 95% CI 0.13-1.58) were not. It is recommended that women with a personal or strong family history of venous thromboembolism should be screened for the prothrombin mutation either before or early in pregnancy, in addition to screening for other thrombophilias. Screening for the MTHFR mutation does not appear to identify women at increased risk of maternal venous thrombosis.
Venous thromboembolism associated with pregnancy has an incidence of approximately 1 per 1000 deliveries1–3, and remains the leading cause of maternal death in the UK4. The incidence may be increasing, perhaps as a consequence of an increased number of women of advanced maternal age, and an increased number who are managed by operative delivery. Since the description of the factor V Leiden mutation, the most common inherited risk factor for venous thromboembolism5, it is clear that genotype plays an important role in determining risk of venous thromboembolism. Poort et al. recently described a common mutation in the prothrombin gene (a G→A transition at position 20210 of the 3'untranslated region), associated with elevated plasma prothrombin levels and a significantly increased risk of venous thromboembolism6. These findings have been validated subsequently in a number of case-control studies7–11. One case report has highlighted this mutation as a possible risk factor for venous thromboembolism associated with pregnancy12, although thus far there are no data to determine whether the prothrombin mutation is associated with a significant number of venous thrombotic events related to pregnancy. In addition, there is current interest in a common mutation (C→T transition at position 677) in the methylene tetrahydrofolate reductase (MTHFR) gene, which is associated with elevated plasma homocysteine levels in the presence of concomitant folate deficiency13–15. Elevated plasma homocysteine is a risk factor for venous thromboembolism16.
The aim of our study was to determine, in addition to other established thrombophilias, the prevalence of the prothrombin and MTHFR gene mutations in a large cohort of unselected women who had suffered an episode of objectively confirmed venous thromboembolism either during pregnancy or the puerperium, and to compare this prevalence with that obtained in the general population.
Eighty-seven unselected consecutive women who had suffered an objectively confirmed episode of venous thromboembolism either during pregnancy or up to six weeks postpartum between 1985–1998 were identified at two maternity units in Glasgow. Patients were identified from discharge registry data using both ICD 9 and ICD 10 codes for deep vein thrombosis or pulmonary embolism. All events were objectively confirmed using either duplex Doppler scanning or venography (in the case of deep vein thrombosis) or ventilation perfusion scanning (in the case of pulmonary embolism). Seventy-five consented to participate in a follow up study which was commenced in 1996. After a brief interview, 30 mL blood was obtained for thrombophilia screening. This included full blood count, coagulation screen, antithrombin and protein C activities, total and free protein S antigens, IgG and IgM anticardiolipin antibodies and testing for lupus anticoagulant. Polymerase chain reaction was used to determine the presence or absence of the factor V Leiden, prothrombin 20210 G→A mutations, and C677T mutations in the methylene tetrahydrofolate reductase (MTHFR) gene. All 75 were screened for deficiency of antithrombin, protein C, protein S, for evidence of a lupus anticoagulant or elevated anticardiolipin antibodies, and for the factor V Leiden mutation. Screening for the prothrombin and MTHFR mutations was introduced in our centre in 1997. Those who had not been screened for these mutations were invited to re-attend. In total, 55 were screened for the prothrombin mutation, and 52 for the MTHFR mutation. The methodology used for coagulation assays and for polymerase chain reaction techniques has been described in detail elsewhere1,17. All were nonpregnant at the time that blood was withdrawn. The prevalence of antithrombin deficiency, and of the factor V Leiden, prothrombin 20210A and homozygous MTHFR mutations in our area (West of Scotland) has previously been described and was used for control purposes17,18.
Seventy-eight of 87 women presented with their first thrombotic event related to pregnancy; nine presented with a recurrent thrombosis. Sixty-four women presented with deep vein thrombosis, of whom 44 (69%) occurred in the antenatal period and 20 (31%) in the puerperium. Twenty-three women presented with pulmonary embolism, nine (39%) of these appeared in the antenatal period and 14 (61%) in the puerperium. Seventy-five women attended for interview and blood sampling (follow up rate of 86%). Of this number, 55 had deep vein thrombosis and 20 had pulmonary embolism. Nine of 75 women had a history of venous thromboembolism before the episode of pregnancy associated venous thromboembolism (eight with previous deep vein thrombosis and one with a previous episode of superficial thrombophlebitis). Fourteen individuals were on oral anticoagulants at the time of blood sampling.
Thrombophilic defects, identified in 75 women who attended for follow up, were as follows: seven (9.3%) with antithrombin deficiency (four with type I deficiency, three with type II deficiency); seven (9.3%) with the factor V Leiden mutation; one (1.3%) with protein C deficiency; and one (1.3%) with the antiphospholipid syndrome (consistently elevated IgG anticardiolipin antibody levels and intermittently positive testing in the dilute Russell viper venom time). In addition, of 55 women who were screened for the prothrombin 20210A mutation, five were found to be heterozygous. One woman was found to possess both the factor V Leiden and prothrombin mutations, both in heterozygous form. Of 52 women who were screened for the C677T mutation in the MTHFR gene, three (6%) were found to be homozygous for the C→T transition, 24 (46%) heterozygous and 25 (48%) homozygous for the normal allele. The total prevalence of established thrombophilia (excluding the MTHFR mutation) found in the cohort was 26.7% (20/75), although not all women were screened for the prothrombin mutation. Details of women found to possess the factor V Leiden or prothrombin mutations are shown in Table 1.
Table 1. Women with maternal venous thromboembolism (VTE) found to have either factor V Leiden (FVL [+/−= heterozygous; +/+= homozygous]) or the Prothrombin (FT) 20210A mutation. DVT = deep vein thrombosis; PE = pulmonary embolism; CS = caesarean section; PT20210A = heterozygous for prothrombin 20210 G→A mutation.
|1||DVT||1st trimester||FVL+/−||Previous VTE, parity|
|2||DVT||Postpartum||FVL+/−||Weight, family history of VTE|
|3||DVT||3rd trimester||FvL+/−||Ulcerative colitis, family history of VTE|
|4||DVT||3rd trimester||FVL+/−||Age > 35 years, parity, family history of VTE|
|5*||DVT||3rd trimester||FVL+/−, PT20210A||Previous VTE, weight > 80 kg|
|7||DVT||1st trimester||FVL+/+||Hyperemesis, immobility|
|8||PE||Postpartum||PT20210A||Emergency CS, pre-eclampsia|
|10||DVT||1st trimester||PT20210A||Previous VTE, weight > 80 kg|
|11||PE||Postpartum||PT20210A||Emergency CS, pre-eclampsia, weight > 80 kg|
The prevalence of antithrombin deficiency, factor V Leiden, prothrombin 20210A and the C677T mutation in the MTHFR gene was compared in cases to that obtained in the general population (Table 2). The prevalence of antithrombin deficiency (1/5000 for type 1 deficiency, 3/2000 for type 2 deficiency) and of factor V Leiden (2.2%), prothrombin 20210A (2.2%) and the MTHFR (12% for homozygous C677T) mutations in our area has been published previously17,18. The factor V Leiden (OR 4.5; 95% CI 2.1-14.5) and prothrombin mutations (OR 4.4; 95% CI 1.2-16) were significantly more common in cases compared with controls (Table 2). Women with type I (OR 282; 95% CI 31–2532) and type II (OR 28; 95% CI 5.5-142) antithrombin deficiencies were markedly more prevalent in cases, compared with controls. However, the homozygous C→T mutation in the MTHFR gene was not identified with increased frequency in cases (OR 0.45; 95% CI 0.13-1.58).
Table 2. Prevlence of factor V Leiden (FVL), prothrombin 20210A, C677T mutations nd ntithrombin deficiency in cases and in the general population.
|Prothrombin 20210A|| || || || |
| Absent||50||219|| || |
|FVL|| || || || |
| Absent||68||219|| || |
|MTHFR C677T|| || || || |
|+/−§and−/−¶||49||139|| || |
|Antithrombin|| || || || |
|Type I deficiency*|| || || || |
| Absent||71||4999|| || |
|Type II deficiency†|| || || || |
| Absent||72||1997|| || |
Risk factors for maternal venous thromboembolism include advanced age, obesity, high parity (≥ 4), operative delivery, hyperemesis, varicose veins, pre-eclampsia, infection or medical illness, and thrombophilia1,4,19. Natural anticoagulant deficiencies are relatively uncommon although significant risk factors for venous thromboembolism, with pregnancy a relatively frequent precipitating factor for venous thromboembolism in those affected20,23. Factor V Leiden and the prothrombin mutations are common in caucasian populations24,25. Data indicate that the factor V Leiden mutation is a risk factor for pregnancy associated venous thromboembolism1,26,27, although the magnitude of risk does not warrant large scale screening programmes.
Since the description of the prothrombin gene mutation6, there have been a number of case-control studies which have confirmed that carriers are at a significantly increased risk of venous thromboembolism7,11. The increased risk of venous thromboembolism is estimated at around two-to fourfold in heterozygotes. Interestingly, recent studies28,29 have also demonstrated an increased risk of myocardial infarction in selected patient groups, and also an increased risk of premature stroke. To our knowledge, our data are the first to demonstrate that the prothrombin mutation is associated with maternal venous thromboembolism. We found that the odds ratio for venous thromboembolism in carriers of the prothrombin mutation was increased (OR 4.4; 95% CI 1.2-16), compared with the general population. However, antithrombin deficiency was markedly more prevalent in cases, compared with controls (OR for antithrombin deficiency type I 282; 95% CI 31–2532; for antithrombin deficiency type II the OR was 28; 95% CI 5.5-142). However, not all women who suffered venous thromboembolism during the study period were screened for the prothrombin mutation, and therefore we must exercise some caution in interpretation of our findings. Additional risks for maternal venous thromboembolism were identified in 4/5 women with the prothrombin mutation who developed venous thromboembolism, and in 6/7 women with the factor V Leiden mutation, suggesting that other factors may be important in the development of thrombotic events in these individuals. It was also noteworthy that all seven women with factor V Leiden presented with deep vein thrombosis, and none had symptoms of pulmonary embolism. Data indicate a significantly lower prevalence of the factor V Leiden mutation in those presenting with isolated pulmonary embolism, compared with deep vein thrombosis or deep vein thrombosis and pulmonary embolism30–33. That is, the relative risk of pulmonary embolism in carriers of factor V Leiden appears lower than for DVT33. The reason for this observation is unclear, although it is hypothesised that higher thrombin levels seen in those with factor V Leiden may lead to the production of a more stable thrombus which is less likely to embolise33.
Women homozygous for the C677T mutation in the MTHFR gene were not identified with increased frequency in cases, compared with the general population (OR 045; 95% CI 0.13-1.58). This mutation may cause mild elevations of plasma homocysteine in the presence of folate deficiency15, which may in turn increase the risk of venous thromboembolism. However, Kluijtmans et al.34 recently concluded that this mutation was not an independent risk factor for venous thromboembolism. Plasma homocysteine appears to fall in pregnancy35, although there are no data to determine plasma homocysteine levels in pregnant women with the MTHFR mutation. Further data are required to determine changes in plasma homocysteine levels during pregnancy in women with this mutation.
In summary, our findings indicate that the recently described prothrombin 20210 G→A mutation is associated with maternal venous thromboembolism. The relative risk of venous thromboembolism can only be determined from a large prospective study, although our odds ratio gives some estimate of the relative risk. The odds ratio for the prothrombin mutation (4.4) was comparable to that identified for factor V Leiden (4.5), and as such would not warrant random screening for this mutation. Rather, women with a personal or family history of venous thromboembolism who conceive should be screened for the prothrombin mutation, in addition to screening for other thrombophilias. In our centre, we screen women with a personal history of venous thromboembolism, women with a first degree relative in whom there has been an episode of (preferably) objectively confirmed venous thromboembolism, or women with multiple (≥ 2) more distant relatives with venous thromboembolism. The management of women with such thrombophilic mutations during pregnancy is complex due to the absence of randomised controls trials of thromboprophylaxis, compared with observation. Suggested approaches to management have recently been published in a large review of the subject36.