- Top of page
- Patients and methods
- Group I: combined FVL and FII 20210A
- Group II: FII 20210A mutation alone
- Comparison of groups I and II
Summary. The risk of venous thromboembolism (VTE) in the absence of prophylaxis was evaluated in a retrospective study of 47 women (84 pregnancies) with combined thrombophilia [heterozygous factor V Leiden (FVL) plus prothrombin (FII) 20210A mutation (group I)] and in 82 women (193 pregnancies) with the FII alone (group II). VTE was more frequent in group I than in group II [17·8% versus 6·2%, P = 0·003, relative risk (RR) 2·9, 95% confidence interval (CI) 1·4–5·9], ante partum (7·1% and 2·1%) and post partum (11·5% and 4·2%). The risk was higher in index cases than in family members (RR 2·5, 95% CI 1·2–5·2 and RR 2·1, 95% CI 0·2–22·3 respectively) Even women who had no history of VTE before pregnancy had an increased risk (RR 2·2, 95% CI 1·0–4·8). Our results suggest that, during ante partum, prophylaxis is indicated in women with combined thrombophilia and with a VTE before pregnancy. In those without VTE before pregnancy, prophylaxis might be decided for each individual case, taking into consideration all risk factors. In women with the FII mutation alone, the low risk may not justify prophylaxis in the absence of previous VTE. In post partum, prophylaxis is indicated in all cases.
Factor V Leiden (FVL) and factor II (FII) 20210A heterozygous gene mutations are common and associated with a moderate risk of thrombosis. In contrast, antithrombin (AT), protein C (PC) and protein S (PS) deficiencies are at least 10 times less common but are considered to be associated with a higher risk of VTE (Bertina et al, 1994; De Stefano et al, 1996). Inheritable thrombophilia is either a single-gene or a multiple-gene disorder (Bertina, 2001). It is generally accepted that the combination of two genetic alterations increases the risk of VTE compared with that associated with a single mutation (Koeleman et al, 1994; Zoller et al, 1995; van Boven et al, 1996; Emmerich et al, 2001; Martinelli et al, 2001).
Pregnant women with thrombophilia are at increased risk of thrombosis in the ante-partum and post-partum period. The risk has been relatively well documented in AT, PC and PS deficiencies and, more recently, in the FVL mutation (Conard et al, 1990; Pabinger et al, 1992; Bokarewa et al, 1996; Pabinger & Schneider, 1996; Simioni et al, 1999; Martinelli et al, 2002). Few data are available for the FII 20210A gene mutation with pregnancy-associated thrombosis, which was first reported in 1998 (McColl et al, 1998a). Combined heterozygous FVL and FII 20210A mutation is rare even though it is the most frequent hereditary combined thrombophilia. Its estimated prevalence is about 1 per 1000 of a Caucasian control population and about 12 per 1000 in patients with a history of thrombosis. The thrombotic risk associated with pregnancy has been assessed in small groups of women with this combined genetic alteration (Grandone et al, 1998; Martinelli et al, 2001).
We conducted a single-centre retrospective study in 47 consecutive women with the double heterozygous FVL and FII 20210A mutation and in 83 women with the FII gene mutation only, who have been pregnant at least once and most of whom did not receive anticoagulant prophylaxis.
The aim of the study was to evaluate and compare the relative risk of pregnancy-related VTE in these two groups of women. It also aimed to determine the appropriate strategy regarding the prevention of VTE in pregnancy including prophylactic treatment with subcutaneous heparins during pregnancy (Brill-Edwards et al, 2000; Ginsberg et al, 2001).
Patients. A retrospective review of our files identified 69 consecutive women carriers of both heterozygous FVL and FII gene mutations (group I) and 116 women with a single-gene heterozygous FII mutation (group II) who were referred to our institution during the period 1997–99.
They were mostly referred by angiologists and gynaecologists because of a personal and/or family history of VTE and/or a family thrombophilia. In the 69 women included in group I (45 index cases, 24 family members), the mean age was 44·2 ± 15·8 years (median 46 years, range 19–78 years) and, in the 116 women in group II (70 index cases, 46 family members), the mean age was 46·4 ± 16·4 years (median 48 years, range 16–83 years). Only women older than 16 years, who had been pregnant at least once, were selected for the study.
All these thrombophilic women had been carefully followed in our outpatient clinic since their first referral to our institution. Full detailed medical histories were obtained using a standard file. The relevant information included the presence, type and site of objectively documented thrombosis before pregnancy or related to pregnancy, ante partum or post partum (in the 6 weeks after delivery).
Coagulation studies. A blood sample was collected from every patient to assess AT heparin cofactor activity, PC amidolytic activity, PS activity and free antigen, activated protein C resistance (APCR), plasminogen amidolytic activity and to screen for antiphospholipid syndrome. The homocysteine level was also measured in a subgroup of 102 patients. Assays for plasma FII and fibrinogen were included in the routine coagulation profile. Cell blood counts, including platelet counts, were also obtained.
Blood samples were drawn into evacuated tubes containing 0·129 mol/l sodium citrate for coagulation studies or EDTA for blood cell counts. Citrated blood was centrifuged twice at 4000 g for 15 min and tested immediately or frozen in aliquots at −30°C. All measurements on frozen plasma samples were performed within 2 weeks of blood collection, and they were all performed in our laboratory using previously described methods (Samama et al, 1999).
The presence of the FVL gene mutation was determined after DNA extraction, polymerase chain reaction (PCR) and MnlI restriction analyses (Bertina et al, 1994).
The presence of the FII 20210A allele was established after DNA extraction and HindIII cleavage of a 345-bp fragment amplified by PCR, as described previously (Poort et al, 1996). Each patient was informed about the genetic tests and gave written consent.
The diagnosis of thrombophilia was always confirmed in a second blood sample, collected after an interval of 1–3 months.
In the absence of prophylaxis. Among the 84 pregnancies, 15 objectively documented VTE episodes (17·8%) occurred in 11 women (four women had two episodes) (Table I). The mean age of women at the time of the pregnancy-associated thrombosis was 25·5 years (range 18–35 years). It was similar in women with an uneventful pregnancy: 27·6 years (range 20–39 years). The thrombosis occurred ante partum in six women (7·1%): three during the first trimester (one had an additional risk factor; treatment with oestradiol), three during the second trimester, and nine (11·5%) post partum (two after caesarean section, one aged 35 years) (Table II). It must be emphasized that the number of pregnancies with VTE in the ante-partum period was excluded for the calculations of the rate of VTE in the post-partum period, as these women subsequently received treatment for the rest of the pregnancy and post partum.
Table II. Characteristics of women who experienced pregnancy-related VTE.
|Patient||Pregnancies (n)||DVT/PE before pregnancy||Age (years)||Pregnancy-related VTE||VTE Pregnancy no.|| Ante/post partum|
|FVL + FII 20210A|
| 1||1||0||28||DVT||1st||Ante partum 1st trim.*|
| 2||1||Yes||27||DVT||1st||Ante partum 2nd trim.|
| 3||1||Yes||30||DVT DVT ||1st||Ante partum 2nd trim. Post partum |
| 4||4||0||23||DVT||1st||Ante partum 1st trim.|
|Yes||29||DVT||4th||Ante partum 2nd trim.|
| 5||2||0||28||DVT||1st||Ante partum (+ oestradiol)|
| 6||2||0||22||DVT + PE||1st||Post partum|
| 7||2||0||23||DVT||1st||Post partum|
| 8||2||0||26||DVT||1st||Post partum|
| 9||3||0||18||DVT||1st||Post partum|
|Yes||35||DVT||3rd||Post partum (caesarean)|
| 10||3||0||35||DVT||3rd||Post partum (caesarean)|
| 11||4||0||30||DVT||4th||Post partum|
| 1||2||0||29||DVT + PE||2nd||Ante partum 2nd trim.|
| 2||3||0||34||DVT||3rd (twin)||Ante partum 3rd trim.|
| 3||5||0||24||DVT||4th||Ante partum 3rd trim.|
| 4||3||0||26||DVT||2nd (twin)||Ante partum 2nd trim.|
| 5||1||0||40||DVT||1st||Post partum|
| 6||2||0||22||DVT||1st||Post partum caesarean|
| 7||2||0||22||DVT||2nd||Post partum|
| 8||2||0||25||DVT||2nd||Post partum|
| 9||2||0||26||DVT||1st||Post partum|
| 10||1||0||30||DVT + PE||1st||Post partum|
| 11||3||0||28||DVT||2nd||Post partum|
| 12||3||0||30||DVT||3rd||Post partum|
In previously asymptomatic women, 10 out of 74 pregnancies (13·5%) were associated with a thrombotic episode. In contrast, among 10 pregnancies in women who had a previous thrombotic episode, five (50%) were associated with thrombosis (P < 0·05) and one woman experienced deep vein thrombosis (DVT) during pregnancy and a recurrent episode in the early post-partum period (patient 3) (Tables I and II).
Interestingly, nine of the 47 women had a VTE event during their first pregnancy; seven of them were asymptomatic before the pregnancy, whereas two had a history of thrombotic episode before pregnancy. The other VTE episodes occurred during a second, third or fourth pregnancy (Table II).
The results in the index cases and family members were analysed separately (Table I). Of the 34 index women who had 66 pregnancies without prophylaxis, VTE occurred in 14 (21·2%): six (9·1%) during pregnancy and eight (13·3%) during the post-partum period. In the 13 family members who had 18 full-term pregnancies without prophylaxis, only one of them (5·5%) was associated with VTE, during the post-partum period. The difference between the results obtained in index cases and in family members did not reach statistical significance (P = 0·12).
A superficial venous thrombosis occurred in six pregnancies.
Among the 10 women in whom the first thrombotic episode occurred during pregnancy, five had a recurrence of thrombosis 6–31 years later, associated with a precipitating factor in three women (oestro–progestogen intake in one, spontaneous abortion in one and a fourth pregnancy in the third case). In another woman who was previously asymptomatic, a superficial vein thrombosis during her fourth pregnancy was followed by a DVT during the post-partum period and a severe vena cava thrombosis 15 years later.
In the cases that received prophylactic treatment. No thrombosis was observed in the six women (four asymptomatic and two with previous thrombosis) who had graduated compression stockings from the beginning of pregnancy and received low-molecular-weight heparin (LMWH) during the third trimester (n = 4) and systematically in the post-partum period (n = 6).
- Top of page
- Patients and methods
- Group I: combined FVL and FII 20210A
- Group II: FII 20210A mutation alone
- Comparison of groups I and II
The evaluation of the risk of pregnancy-associated VTE in women with thrombophilia is difficult to assess because the available data have often been obtained in small studies and the combined thrombophilias (such as FVL associated with FII G20210A or deficiencies in coagulation inhibitors) have not been identified in all women. Two main methods are commonly used to evaluate the magnitude of the risk of thrombosis: the frequency of pregnancy-associated VTE in known thrombophilic women (method used in the present study) or the frequency of thrombophilias in women who experienced a pregnancy-associated VTE. The frequency of DVT and PE during pregnancy and post partum was reported to be high in women with AT deficiency (36–46 per 100 pregnancies) and lower in protein C (20% and 15%) and protein S (16% and 27%) deficiency (Conard et al, 1990; Pabinger & Schneider, 1996). Index cases and family members, with or without previous thromboses, were included in these studies. When the study is restricted to family members with no personal history of VTE, the risks in the three coagulation inhibitor deficiencies are only 3%, 1·7% and 7% respectively (Friederich et al, 1996). In FVL or FII 20210A, the risk seems to be low. In family members with FVL who have no history of VTE, a frequency of two VTE per 100 pregnancies has been reported (Middeldorp et al, 1998; Simioni et al, 1999).
Our retrospective study included index cases and family members carrying both FVL and FII 20210A mutations or the FII 20210A mutation alone but not FVL alone. We found that, without prophylaxis, the risk of pregnancy-associated VTE was higher in women with both FVL and FII 20210A mutations than in those with FII mutation alone (17·8% versus 6·2%, RR 2·9, 95%CI 1·4–5·9). For the FII mutation alone, we found a frequency of 2·6% in family members, close to the 2% reported in the literature for FVL alone (Middeldorp et al, 1998; Simioni et al, 1999). We have not compared our results in FII 20210A carriers with other studies. To our knowledge, the frequency of pregnancy-associated VTE in this single thrombophilia has been studied in family members only (Martinelli et al, 2000). However, the frequency of FII 20210A has been reported to be 9·2–31% in women with pregnancy-associated VTE (Grandone et al, 1998; Gerhardt et al, 2000; McColl et al, 2000).
We found a VTE frequency of 5·5% in family members with both mutations, which is very similar to the 4% reported by Martinelli et al (2002). Our results in family members suggest that the risk of pregnancy-associated VTE in carriers of both mutations is additive rather than supra-additive: 5·5% compared with 2·6% in FII 20210A alone (present study) and 2% in FVL alone (Middeldorp et al, 1998; Simioni et al, 1999).
Most thromboses were observed in index cases, and one can argue that the status of index cases might introduce a bias. However, women who were considered as index cases were those in whom the thrombophilia was first discovered in each family, they were usually symptomatic at the time of the detection but may have experienced thrombosis before or long after pregnancies. Thus, it is important to compare the pregnancy-associated risk in women who were asymptomatic or not when they became pregnant (index cases or family members). Women with both mutations who had no history of VTE had a lower risk of pregnancy-associated VTE than those who had a VTE before pregnancy (13·5% versus 50%, P < 0·05). The same could not be done for FII 20210A mutation carriers because none of them had VTE before being pregnant, and a VTE occurred in 6·2%.
VTE was more frequent in the post-partum period than during pregnancy. In carriers of both mutations, VTE occurred during pregnancy at any trimester without another risk factor, except oral oestradiol intake in one, whereas in carriers of FII alone, two of the four women had a twin pregnancy. It is our belief that the magnitude of the risk during the ante-partum period needs more attention than the risk post partum because it is now clearly recommended to administer post-partum prophylaxis to women with thrombophilia. Ante partum, a VTE was observed in 7·1% of women with both mutations and 2·1% of women with the FII 20210A mutation alone.
One strength of our study is that all the patients were treated at a single centre, and all the tests were performed in the same laboratory. The main causes of hereditary thrombophilia have been screened in every patient. All patients were tested for antithrombin, protein C, protein S, anticardiolipin antibodies, FVL and FII 20210A mutations. Only patients with combined heterozygous mutations or FII 20210A mutation alone were included in the study, which represents the largest series reported. Homocysteinaemia was measured in a subgroup of patients (n = 102), and eight of them had a borderline level (results not shown).
Bias may exist in our study as in other retrospective studies. However, the demographic medical background and causes of referral were comparable in both groups of patients. We used a standardized record for the history of all the study participants to limit potential bias. The thrombotic events were not significantly more frequent in the index cases than in family members. Pregnancy is a characteristic situation, and symptomatic VTE was well documented in all patients. We took into consideration PE and DVT only. Distal vein thrombosis was more frequent than proximal DVT in both groups.
Superficial vein thrombosis (SVT) was excluded, although it was observed in a few women. Some authors consider that SVT plays a minor role (McColl et al, 1998b). For others, thrombophilia was associated with an increased risk of SVT during pregnancy (Martinelli et al, 1999).
The aetiology of retinal vein occlusion is uncertain, and the role of thrombophilic factors seems to be limited in this condition where thrombosis, compression and vasculitis may be involved (Vine & Samama, 2002). Thus, we did not include retinal vein occlusion in the pregnancy-associated thrombotic events.
It has been suggested that the rate of adverse pregnancy outcome is increased in women with thrombophilia (Preston et al, 1996; Kupferminc et al, 1999; Many et al, 2001; Alfirevic et al, 2002). In this study, the rate of fetal loss was higher in women with a combined heterozygous FV and FII gene mutation than in the group with FII 20210A alone (13·4% and 6·3%, respectively, P < 0·04). The small number of events does not allow a reliable conclusion to be formed, and it was not a primary objective of the study.
The results of our study may be useful when planning prophylaxis in a pregnant woman with combined heterozygous FVL and FII 20210A gene mutations or with the FII 20210A gene mutation alone. The American College of Chest Physicians' Consensus Conference recommendations (grade 1C) for thrombophilias other than AT deficiency were: clinical surveillance and/or prophylactic treatment during pregnancy and prophylaxis in post partum in all women (Ginsberg et al, 2001). Personal history of VTE may be taken into account (Conard & Horellou, 2002; Lockwood, 2002). Low-molecular-weight heparins (LMWH), such as enoxaparin and dalteparin, appear to be safe and efficient in pregnant women (Sanson et al, 1999; Lepercq et al, 2001, Anonymous, 2002). According to our results, in asymptomatic women with FII 20210A mutation alone, the magnitude of the risk is low during pregnancy. It may not justify prophylaxis ante partum but post partum only, as recommended for FVL alone, where the risk is similar (McColl et al, 1999; Lockwood, 2002). Our results did not provide information about the risk in women with FII mutation alone when a history of VTE exists.
There is agreement that women with the combined FVL and FII 20210A mutations should receive prophylaxis during the post-partum period. In women with a history of VTE, our study shows that the risk is high in the ante-partum period (three out of 10), and prophylaxis seems to be justified throughout pregnancy. In asymptomatic women (no history of VTE before pregnancy), the decision regarding prophylaxis during pregnancy is more difficult, and our results suggest that the risk in the ante-partum period is lower than anticipated (three out of 74). Prophylaxis may be decided case-by-case, taking into consideration other risk factors of thrombosis (age of the woman, multiparity, bed rest, family history) and the risk of long-term administration of LMWH, even if it is lower than with unfractioned heparin (heparin-induced thrombocytopenia, osteoporosis). Graduated compression stockings are often recommended. In contrast, carriers of these combined mutations who have had a previous thrombotic episode have a high probability of pregnancy-associated thrombosis, suggesting that prophylaxis is required throughout pregnancy.
In our series, 12 women with combined mutations had a recurrent thrombotic episode after pregnancy. The interval between the first and second episode was variable, and it was longer than 4 years in 10 women. This observation does not support systematic long-term anticoagulation as recommended by some authors (De Stefano et al, 1999; Seligsohn & Lubetsky, 2001).
The results of our study can be considered as relatively reassuring rather than alarming. The evaluation of the thrombotic risk in a larger group of patients, especially with combined heterozygous FVL and FII 20210A mutations, is desirable in order to enable more firm recommendations to be made on the use of ante-partum prophylactic treatment.