- Top of page
- PATIENTS and METHODS
Factor V Leiden (FVL) leads to a sevenfold increased risk of venous thrombosis and is present in 50% of individuals from families referred because of unexplained familial thrombophilia. We assessed the association of FVL with venous thromboembolism (VTE) in 12 thrombophilic families of symptomatic probands with FVL in a retrospective follow-up study. We screened 182 first- and second-degree relatives of the 12 unrelated propositi for the FVL mutation and the occurrence of VTE. The incidence rate of VTE in carriers of FVL (0·56%/year) was about six times the incidence for the Dutch population (0·1%/year). The incidence rate in non-carriers also appeared to be higher (0·15% per year). At the age of 50 years, the probability of not being affected by VTE was reduced to 75% for carriers and to 93% for non-carriers (P = 0·009). Identification of carriers of FV Leiden may be worthwhile in young symptomatic individuals and their relatives with a strong positive family history of venous thromboembolism or a history of recurrent venous thrombosis who may be at risk (e.g. pregnancy, use of oral contraceptives). After adjustment for prothrombin G20210A (present in two families), even higher thrombotic incidence rates were found in carriers and non-carriers of FVL. This makes the presence of other unknown prothrombotic risk factors more probable in these families.
Venous thromboembolism is a major cause of morbidity with an incidence of about one per 1000 per year ( Nordström et al, 1992 ; National Medical Registration of the Foundation Information Centre for Health, 1996). Predisposing factors can be either genetic or environmental. Environmental risk factors include recent surgery, immobilization, oral contraceptives, pregnancy and puerperium ( Nachman & Silverstein, 1993). Until 1993, a specific genetic defect was identified in a maximum of 10–15% of affected subjects ( Allaart & Briët, 1994). These defects included deficiencies of the main inhibitors of the clotting system: protein C, protein S and anti-thrombin ( Hirsh et al, 1986 ). Familial resistance to activated protein C (APC-resistance), first described in 1990 ( Dahlbäck et al, 1990 ), is the most frequent genetic risk factor for thrombosis ( Griffin et al, 1993 ; Koster et al, 1993 ; Svensson & Dahlbäck, 1994 ; Dahlbäck, 1995; Lane et al, 1996 ). In the large majority of cases, APC resistance is associated with a single point mutation (1691 G to A substitution) in the factor V gene that predicts the synthesis of a factor V molecule (Factor V Leiden) that is not properly inactivated by activated protein C (APC) ( Bertina et al, 1994 ; Greengard et al, 1994 ; Voorberg et al, 1994 ). In Caucasians, this mutation is present in about 5% of healthy individuals ( Rees et al, 1995 ; Zivelin et al, 1997 ). Factor V Leiden leads to a sevenfold increased risk of venous thrombosis ( Rosendaal et al, 1995 ). It is present in 20% of unselected, consecutive patients with deep vein thrombosis and in 50% of individuals from families referred because of unexplained familial thrombophilia ( Bertina et al, 1994 ; Zöller et al, 1994 ; Lane et al, 1996 ).
We demonstrated an earlier age of onset in a series of selected patients from thrombophilic families with factor V Leiden than in a panel of unselected patients with a first venous thrombosis who turned out to be carriers of the factor V Leiden mutation ( Lensen et al, 1996 ). This suggests a higher thrombotic tendency in members from selected families than in consecutively diagnosed patients, even if both carry the same or a similar molecular defect. Important to clinicians is the question of what prophylactic measures are advisable for patients and their relatives in these selected thrombophilic families with factor V Leiden. Before this question can be answered, the risk of thrombosis in these individuals needs to be assessed. Therefore, we studied 12 thrombophilic families with the factor V Leiden mutation in which, next to the proband, at least two persons had experienced thrombosis.
- Top of page
- PATIENTS and METHODS
We performed a study on 182 members of 12 referred families with thrombophilia. At the age of 50 years, 25% of carriers had experienced at least one venous thrombotic event (vs. 7% in non-carriers). The incidence rate of deep vein thrombosis (superficial thrombophlebitis excluded) in carriers is six times higher than the reported incidence rate for the general Dutch population (about 0·11%/year) ( Nordström et al, 1992 ; Briët et al, 1994 ; National Medical Registration of the Foundation Information Centre for Health Care, 1996). This incidence rate of 0·11%/year for the general population is probably an overestimation as recurrences are included. Interestingly, in the families we studied the incidence among non-carriers (0·17%/year, adjusted for the prothrombin G20210A allele) also exceeded this population figure. This increased incidence of venous thrombosis in non-carriers is probably caused by the selection of families that are very prone to thrombosis, in which the presence of more than one defect may be suspected ( Lane et al, 1996 ; Lensen et al, 1996 ; Rosendaal, 1997). We also think that the significant higher thrombotic incidence found in first-degree relatives as compared with second-degree relatives (which is characteristic in polygenic inheritance) gives support to the hypothesized presence of more co-existing genetic defects and justifies further research work into this matter.
Previously, in a population-based case-control study on venous thrombosis (Leiden Thrombophilia Study, LETS; Rosendaal et al, 1995 ) among 471 unselected consecutive patients with a first, objectively confirmed deep vein thrombosis (all younger than 70 years) and 471 unrelated age- and sex-matched controls, we found an odds ratio for heterozygosity of 7·3. The lower risk ratio of 3·7 for heterozygosity we found in thrombophilic families with factor V Leiden may be the result of the high incidence of venous thrombosis in non-carriers and the difference in study design.
Considering Tables I and II, we think that the minimally higher thrombotic risk found in carriers older than 45 years as compared with the 25–45 year group is caused by the occurrence of many thrombotic events during pregnancy and puerperium or the use of oral contraceptives in the 25–45 years group, demonstrating their important role as a thrombogenic risk factors in this age group.
Recently, Middeldorp et al (1998) presented their results of a family study on factor V Leiden. They investigated first-degree family members of consecutive patients with venous thromboembolism, as opposed to referred high-risk families in our study. They and Zöller et al (1994) found that 97% of carriers were still free of thrombosis at age 30 years. We found that 28% of carriers in thrombophilic families had experienced thrombosis at this age ( Fig 2). The incidence rate for first-degree heterozygous family members younger than 61 years of age was 0·9%/year, more than two times the incidence rate found by Middeldorp et al (1998) (0·4%/year; in both calculations, follow-up started at the age of 15 years). These comparisons emphasize that the estimated risk of thrombosis for carriers of factor V Leiden depends on the inclusion criteria and is much higher in selected and referred families than in relatives of consecutive unselected patients. This is also illustrated by comparing our data with data from a study recently performed by Simioni et al (1999) . They included relatives of unselected patients and found an annual thrombotic incidence of 0·28% in carriers and 0·09% in non-carriers (rate ratio 2·8, 95% CI 1·1–8·6) (in first-degree relatives we found 1·7% in carriers and 0·3% in non-carriers, resulting in a doubled rate ratio of 5·7, 95% CI 2·0–26·8). We recently found thrombotic incidences in first-degree relatives of 47 unselected consecutive patients with a first venous thrombosis that were five times lower in carriers and three times lower in non-carriers than the thrombotic incidences we found in the thrombophilic families. Only one of these 47 families (2%) met the inclusion criteria used in the current study (more than one symptomatic relative besides the proband) and 8% of the first-degree relatives had experienced a deep vein thrombosis (as Middeldorp et al, 1998 found in their study population; Lensen et al 2000 ) vs. 47% in the current study. This illustrates the role of a strong family history as a probable risk factor for venous thrombosis.
The inclusion of thrombotic events that were not objectively diagnosed (20%) could have led to an overestimation of the thrombotic incidence. As the possibility of misclassification concerns only this fraction of all events, and as the medical history in these patients strongly indicated a deep venous thrombosis, this cannot have affected our figures materially (after exclusion of the 20% who were not objectively diagnosed, we found similar risk estimates with wider confidence intervals).
Two years and 6 years after the first event, 10% (two out of 20) and 25% (five out of 20) of the symptomatic carriers had experienced a second event (superficial thrombophlebitis excluded). These high recurrence risks in factor V Leiden carriers are similar to the results of previous studies from the United States and Italy ( Ridker et al, 1995 ; Simioni et al, 1997 ; Prandoni et al, 1998 ). The protective role of oral anticoagulants is well illustrated in these 20 symptomatic carriers. None of the eight carriers who received oral anticoagulant treatment experienced a recurrence, while ten of the 12 carriers who received no oral anticoagulant treatment experienced a recurrence.
Remarkably, two (16·7%) of the propositi had a combined defect, i.e. factor V Leiden and the prothrombin G20210A variant. This mutation has a population prevalence of about 2% and, therefore, it seems probable that the combination of defects led to these patients becoming index patients ( Poort et al, 1996 ). The elevated incidence in non-carriers could not be explained by the presence of the prothrombin G20210A variant: none of the relatives carrying only the prothrombin G20210A variant had experienced a venous thrombosis and the thrombotic incidence rate in non-carriers was even higher after exclusion of the two families with the prothrombin G20210A variant.
We compared our data with data from a study performed by Allaart et al (1993) on similar selected thrombophilic families with heterozygous protein C deficiency (who were also known to our centre) for thrombophilia work-up. Using these data, we found that the median of thrombotic incidence and recurrence rates were very similar for both defects, which does not lend support to different thrombotic risks for protein C deficiency and factor V Leiden.
In conclusion, in clinical practice special attention should be paid to young symptomatic individuals and their relatives with a strong positive family history of venous thromboembolism or a history of recurrent venous thrombosis who are at risk, especially women who would like to use oral contraceptives or who intend to become pregnant. Identification of carriers of factor V Leiden may be worthwhile in these persons in order to discourage contraceptive pill use among carriers and to protect carriers during pregnancies against venous thrombosis. However, data that such a policy would be beneficial are lacking. In addition, our data are based on a study among selected families with thrombophilia and should not be applied to screening of other asymptomatic individuals (i.e. prior to prescribing oral contraceptives). Finally, our data provide no grounds to treat patients with factor V Leiden differently from patients with heterozygous protein C deficiency. However, considering the fact that the prevalence of factor V Leiden is at least tenfold higher than the prevalence of all other known genetic deficiencies, further prospective studies will be needed to evaluate clinical policy.