A prospective cohort study on the absolute incidence of venous thromboembolism and arterial cardiovascular disease in asymptomatic carriers of the prothrombin 20210A mutation

After its discovery in 1996, the prothrombin 20210A mutation has become a well-established risk factor for venous thromboembolism (VTE).^1-5  Its role in arterial cardiovascular disease (CVD), however, has been a matter of debate. Meta-analyses have shown increased risks for myocardial infarction, ischemic stroke, and peripheral arterial disease at young ages and smaller, nonsignificant increases for overall arterial events.^6-8 

The majority of the evidence on the risk increases in carriers of the prothrombin 20210A mutation is based on case-control studies.^1-5  Although this type of study is an effective means of detecting associations between risk factors and disease, it may be less reliable to multiply these odds ratios by baseline risks in the general population in order to estimate absolute incidences in individuals with the risk factor of interest. Yet these absolute incidences are much needed. First, they are necessary for counseling individual patients about their risk. Second, these incidences will add evidence to the debate about whether there is any indication to test for this mutation in asymptomatic individuals in general or in specific risk groups, such as women with upcoming pregnancies or oral contraceptive use. We studied first-degree relatives of consecutive patients with the prothrombin 20210A mutation and either VTE or premature atherosclerosis. In our previously reported retrospective analysis of this cohort, we found an annual incidence of first VTE and arterial CVD of 0.35% (95% CI, 0.21-0.53) and 0.22% (95% CI, 0.11-0.34), respectively.⁹  Here we present the results of the prospective cohort study on the absolute incidence of first VTE and arterial CVD for carriers of the prothrombin 20210A mutation.

First-degree relatives, older than 14 years, of consecutive patients with objectively documented VTE or premature atherosclerosis (any arterial cardiovascular event before the age of 50 years) and the prothrombin 20210A mutation were enrolled in a multicenter study.⁹  Of this cohort, all asymptomatic participants were included in the prospective study. Participants with previous VTE were considered asymptomatic with regard to a first arterial cardiovascular event and vice versa. The last follow-up contact for all participants was scheduled in October 2004; the minimum duration of follow-up was 6 months. The study was approved by the local medical ethics committees, including that of the Academic Medical Center (Amsterdam, The Netherlands), and participants provided written informed consent in accordance with the Declaration of Helsinki. All participating centers are university hospitals serving their regions as well as functioning as tertiary referral centers for patients with VTE and/or premature atherosclerosis.

At time of inclusion, DNA analysis was performed for presence of the prothrombin 20210A mutation by polymerase chain reaction of genomic DNA extracted from peripheral leukocytes by standard methods, followed by digestion with the restriction enzyme HindIII to identify the variant allele.¹ 

Participants were instructed to seek immediate medical attention in case of symptoms of venous or arterial thromboembolism in order to undergo appropriate objective testing. Additionally, participants were contacted for follow-up every 6 to 12 months by means of structured questionnaires to identify occurrence of VTE or arterial cardiovascular events, exposure to risk factors, and use of medication including anticoagulants. All reported events were cross-checked with medical records and/or treating physicians.

Occurrence of VTE had to be confirmed by compression ultrasonography or venography in case of deep venous thrombosis and by ventilation-perfusion lung scanning, angiography, or computed tomography in case of pulmonary embolism.

Coronary and peripheral arterial disease had to be symptomatic and angiographically proven, while myocardial infarction had to be diagnosed according to clinical, enzymatic, and electrocardiographic criteria. Ischemic stroke was defined as the sudden onset of symptoms and signs of loss of cerebral function lasting at least 24 hours without any apparent nonvascular cause. Furthermore, it had to be confirmed by means of computed tomography or magnetic resonance imaging. If a cerebral event resolved completely within 24 hours without signs of cerebral lesions on scanning results, it was classified as a transient ischemic attack if considered as such by a neurologist.

Episodes of increased VTE risk included recent surgery, trauma, immobilization for more than 7 days, active malignancy, pregnancy, postpartum period, and use of oral contraceptives or hormone replacement therapy. Venous thromboembolism was classified as related to such an episode if it occurred within 3 months. A relationship with hormone-containing preparations was defined as occurring during exposure.

Incidence rates and 95% confidence intervals of first episodes of venous and arterial thromboembolic events for carriers (hetero- and homozygous) and noncarriers were calculated by dividing the number of events by the number of observation years. Observation years were defined as years since inclusion date until the date of the end of follow-up, or until the date of the first thromboembolic event or death, in case it had taken place. If a participant experienced both an arterial and venous event, separate observation years were calculated for event analysis. Comparison between carriers and noncarriers was quantified by a hazard ratio with 95% confidence interval using Cox regression. The hazard ratio for arterial cardiovascular disease was corrected for age, sex, smoking, hypertension, dyslipidemia, and diabetes mellitus by multivariate Cox regression.

For the sample size calculation, we concluded that a minimum of 600 prospective follow-up years was needed for both carriers and noncarriers to detect an incidence of approximately 0.5% per year with an upper limit of the 95% confidence interval of 1.5%. All analyses were performed using SPSS statistical software (version 11.5.2; SPSS, Chicago, IL).

Follow-up was complete for all 464 relatives who were included in the study (Table 1). DNA analysis revealed 230 heterozygous and 6 homozygous carriers, and 228 had the wild-type gene. The participants were first-degree family members of 134 probands.⁹  Fifty relatives had a history of VTE and were therefore excluded from the prospective analysis on VTE incidence but were still at risk for a first arterial cardiovascular event. Twenty-six relatives had a history of an arterial cardiovascular disease and were included only in the prospective analysis for first VTE. The overall number of follow-up years was 1817: 943 years for the carriers and 874 years for the noncarriers.

A total of 4 venous thromboembolic and 11 arterial cardiovascular events was recorded (Table 2). No events occurred in the 6 homozygous carriers. The absolute annual incidence of first VTE was 0.37% (95% CI, 0.08-1.08) for carriers and 0.12% (95% CI, 0.00-0.69) for noncarriers. The hazard ratio for carriers versus noncarriers was 3.1 (95% CI, 0.3-29.6).

For a first arterial cardiovascular event, the annual incidence was 0.56% (95% CI, 0.18-1.31) in carriers and 0.73% (95% CI, 0.27-1.58) in noncarriers (hazard ratio: 0.8; 95% CI, 0.2-2.6). Detailed incidences are shown in Table 3. Multivariate correction for age, male sex, smoking, hypertension, dyslipidemia, and diabetes mellitus did not materially alter the hazard ratio (0.7; 95% CI, 0.2-2.5). Thirteen carriers and 7 noncarriers had used aspirin during the follow-up period, all for primary prevention of arterial cardiovascular disease. None of these relatives experienced a cardiovascular event. Excluding the users of aspirin from the analysis only marginally affected the annual incidences (0.60% for carriers and 0.76% for noncarriers), and the hazard ratio between carriers and noncarriers remained the same.

Nine participants died during the follow-up period. In 1 patient, death was related to an arterial event: postoperative sepsis after surgery for aortic dissection. Three participants died at old age at home without a known cause, 3 other relatives died because of end-stage malignant diseases, and another 2 because of fatal infectious diseases. Of the deceased, 4 were heterozygous carriers and 5 were noncarriers; the mean age of death was 80 years (range, 69-93 years) in the carriers and 71 years (range, 54-88 years) in the noncarriers.

We recorded a total of 159 high-risk episodes for VTE. Sixty-three heterozygous carriers had 92 episodes (43 surgical procedures, 19 traumas, 13 periods of immobilization > 7 days, 4 malignancies, and 13 pregnancies), and 46% had received thrombosis prophylaxis. Sixty-seven episodes were recorded in 54 noncarriers (29 surgical procedures, 11 traumas, 10 periods of immobilization > 7 days, 4 malignancies, and 13 pregnancies), and standard thrombosis prophylaxis was given in 32%. None of these high-risk episodes resulted in a VTE.

Sixty-five women had a total of 171 years of oral contraceptive use or hormone replacement therapy: 84 years in the carriers and 87 years in the noncarriers. None of these women had a VTE, resulting in incidence rates of 0.0% (0.0-4.4%) in carriers and 0.0% (0.0-4.2%) in noncarriers.

In this prospective family study of asymptomatic carriers of the prothrombin 20210A mutation, we were able to confirm the findings of our retrospective analysis.⁹  The incidence of VTE in asymptomatic carriers of the prothrombin 20210A mutation seems to be in the same range as those found in prospective studies in carriers of the factor V Leiden mutation, but somewhat lower than the risks observed in asymptomatic individuals with deficiencies of the natural anticoagulants (antithrombin, protein C, protein S)^9-14  or elevated levels of clotting factor VIII:c.¹⁵ 

The prospectively observed annual incidences of a first arterial cardiovascular event in both carriers and noncarriers were somewhat higher than in our retrospective analysis. This can probably be explained by the fact that in this prospective follow-up study, only current years contributed to the analysis, which by definition was different in the retrospective study, in which all patient years from the age of 15 contributed. Since increasing age is a major risk factor for atherosclerotic disease, this might well explain the difference in the observed incidences.

The strengths of this study of asymptomatic carriers of the prothrombin 20210A mutation are its size and the fact that it is the first cohort that was followed prospectively. The advantage is thus that all VTE episodes and arterial cardiovascular events were confirmed by objective testing.

Some potential limitations warrant comment. First, the relatively short duration of follow-up with a fairly low number of events hampers comparison of carriers and noncarriers by means of hazard ratios, the estimates of risk episode-related VTE, as well as correction for unequally distributed characteristics of the participants. However, the purpose of this study was to obtain reliable estimates of absolute incidences for carriers, and the upper limit of the 95% CI (1.08%) well excludes the a priori defined 1.5% per year. Given the confirmative findings of this prospective study, we are confident that estimates of hazard ratios as well as the relationship between VTE, arterial cardiovascular events, and other risk factors can also be taken into account from retrospective studies.^9,16 

Second, some characteristics of our cohort could potentially lead to an overestimation of the observed incidences. We studied first-degree relatives of patients with either VTE or premature atherosclerosis, and therefore other inherited risk factors may be present. Also, the centers that included the probands were all university hospitals, implying a potential for referral bias. The majority of probands, however, were consecutive patients diagnosed with either VTE or premature atherosclerosis in these university hospitals. It is noteworthy that the incidence of VTE in noncarriers does not exceed the incidence in the general population, making such a bias unlikely.¹⁷  Furthermore, the participants in this study reflect the clinical reality in which, if testing is performed at all, this is most often done in the setting of a positive family history.

Finally, excluding symptomatic relatives from this analysis might bias the results toward a lower incidence of events. However, in our retrospective analysis, the symptomatic relatives were included and contributed to the outcome, and the observed incidences in that analysis were similar for VTE and lower for arterial cardiovascular events.⁹  Thus, we do not assume that this limits our findings.

We conclude that even though our and previous studies indicate an association between the prothrombin 20210A mutation and VTE^1-4  and arterial cardiovascular disease in the young,^6-8  the absolute incidence of these events in asymptomatic carriers is low. Therefore, the clinical implications of carriership of the prothrombin 20210A mutation are limited, and therefore routinely testing all first-degree relatives of probands with this mutation does not appear to be justified.

We thank research nurses Marja Voskuilen, Inge Paas, Jeannine van Suijlekom, and Mia Muller, as well as Joost C. M. Meijers, PhD, head of the Laboratory of Vascular Medicine, for their excellent help.