- Top of page
- Author contributions
Thrombophilia is a well-established risk factor for a venous thromboembolic event (VTE), and it has been proposed that hereditary thrombophilia may substantially contribute to the development of VTE in young patients. We aimed to analyse the prevalence of thrombophilia with special regard to the age of VTE manifestation. The study cohort consisted of 1490 patients (58% females) with a median age 43 years at the time of their first VTE. At least one thrombophilic disorder was identified in 50·1% of patients. The probability of detecting a hereditary thrombophilia declined significantly with advancing age (from 49·3% in patients aged 20 years and younger to 21·9% in patients over the age of 70 years; P < 0·001). This may be primarily attributed to the decreasing frequencies of the F5 R506Q (factor V Leiden) mutation and deficiencies of protein C or protein S with older age at the time of the initial VTE event. Moreover, thrombophilia was more prevalent in unprovoked compared with risk-associated VTE (57·7% vs. 47·7%; P = 0·001). The decline in the prevalence of hereditary thrombophilia with older ages supports the use of a selected thrombophilia screening strategy dependent on age and the presence or absence of additional VTE risk factors.
- Top of page
- Author contributions
Using a cross-sectional study, we revealed a 43–61% prevalence of thrombophilia for each decade of VTE manifestation age. The prevalence of any thrombophilia decreased slightly with age. There was a significant decline in the prevalence of hereditary thrombophilia with age, which could be primarily attributed to the decreasing frequencies of the F5 R506Q mutation and deficiencies of PC or PS with older age at the time of the initial VTE event. Moreover, the prevalence of hereditary thrombophilia was higher among patients with unprovoked or risk-associated VTE who were younger than 40 years old when compared with older patients. Interestingly, we discovered the F5 R506Q mutation to be more prevalent in males than in females without hormonal treatment, whereas the prevalence was approximately the same between females using OC and males of the same age group.
Only a few studies have analysed the prevalence of thrombophilia with special regard to the age of VTE manifestation. In one investigation (Paschoa & Guillaumon, 2006), 84 patients with previous VTE underwent screening for thrombophilia (i.e., APL; deficiencies of AT, PC and PS; F5 R506Q and F2 G20210A mutations, hyperhomocysteinaemia). The authors discovered no difference in the prevalence of thrombophilia when comparing patients older and younger than 50 years of age. Pottier et al (2005) investigated the presence of thrombophilia (i.e., APL; deficiencies of AT, PC or PS; F5 R506Q or F2 G20210A mutations) and common VTE risk factors in a cohort of 66 patients with unprovoked VTE. These authors discovered a higher frequency of thrombophilia among patients younger than 40 years old when compared with older patients (64% vs. 32%; P = 0·03). However, after exclusion of patients with lupus anticoagulant (N = 5), the frequency of thrombophilia did not differ between the two age groups (Pottier et al, 2005). In the Spanish Multicentric Study on Thrombophilia (EMET), which included 2132 VTE patients, a higher frequency of inhibitor deficiencies (AT, PC or PS) in VTE patients younger than 45 years was observed when compared with patients older than 45 years [17·9% vs. 10·7%; OR 1·8 (1·4–2·4)] (Mateo et al, 1997). In summary, the literature data are limited and contradictory. However, recently published findings of the large RIETE (Registro Informatizado de Enfermedad TromboEmbólica) registry support our results insofar as hereditary thrombophilia is more prevalent in younger VTE patients (Monreal et al, 2012). However, in the RIETE study, thrombophilia screening was not routinely performed, and only 21% of patients underwent testing (Monreal et al, 2012). Another study reported that hereditary thrombophilia (F5 R506Q or F2 G20210A mutation, deficiencies of AT, PC or PS) is especially more prevalent among younger patients with unprovoked DVT when compared with patients older than 45 years or with those with risk-associated DVT (De Stefano et al, 2002). This observation is consistent with our findings, and our study results indicate that a family history of VTE further enhances the chance of detecting a prothrombotic disorder.
The most common hereditary thrombophilias are the F5 R506Q and F2 G20210A mutations. The LITE study (Longitudinal Investigation of Thromboembolism Etiology) combined the data from the Atherosclerosis Risk in Communities (ARIC) study and the Cardiovascular Health Study (CHS). The combined study sample comprised 21 680 subjects. In a nested case-control analysis, the presence of an F5 R506Q mutation increased the risk of VTE by a factor of 4·6 in individuals aged 45–64 years in the ARIC study and by a factor of 2·2 in individuals aged 65 years and older from the CHS (Folsom, 2007). In contrast to these data, the results of earlier studies suggested that the relative risk of VTE with the F5 R506Q and F2 G20210A mutations is age-dependent and increases with age (Folsom et al, 2002; Heit et al, 2005). In the EPCOT (European Prospective Cohort on Thrombophilia) study, a large multicentre cohort study of familial thrombophilia, 575 asymptomatic carriers of a hereditary thrombophilia (F5 R506Q mutation or AT, PC or PS deficiency) were followed for an average of 5·7 years. The annual incidence rate for a first VTE was 0·8% compared with 0·1% per year among 1118 non-carriers who served as controls (Vossen et al, 2005). In this study, the initial VTE event occurred approximately 20 years earlier in patients with thrombophilia when compared with individuals without thrombophilia.
The overall incidence of VTE in the general population has been estimated to be 1–2 per 1000 per year (Rosendaal et al, 2007) and has been described to be similar between males and females. The risk is elevated among young females with risk factors associated with reproduction (i.e., OC use, pregnancy and lactation period), whereas a higher VTE risk has been described for elderly men when compared with women of the same age (Tsai et al, 2002; Naess et al, 2007). Furthermore, males have a higher risk of VTE recurrence than females (Kyrle et al, 2004; Cushman et al, 2006; Linnemann et al, 2008). The lower recurrence risk in females has been attributed to the lower risk after the removal of hormonal risk factors (Cushman et al, 2006). However, there is evidence from several studies that males exhibit a higher recurrence risk, irrespective of whether the initial VTE event was risk-associated or unprovoked (Christiansen et al, 2005; McRae et al, 2006; Linnemann et al, 2008). The higher recurrence risk in males may be explained by a higher prevalence of hereditary thrombophilia among males. Our own results indicate that males are more likely to have a F5 R506Q mutation or a deficiency of natural inhibitors than females with a VTE that was not related to OC use. However, other study groups did not find any difference in the prevalence of hereditary thrombophilia among males and females (Kyrle et al, 2004). Moreover, there is still controversy regarding whether hereditary thrombophilia significantly increases the risk of VTE recurrence.
In contrast to the hereditary thrombophilias, we observed higher frequencies of persistently elevated FVIII activity that was not related to malignancy or inflammatory disease with VTE manifestation at higher ages. It is known that FVIII activities are not only elevated in patients with cancer, inflammation or pregnancy but also increase with age (Conlan et al, 1993; Luxembourg et al, 2009). Therefore, it is not surprising that, even after the exclusion of patients with cancer, inflammatory disease or pregnancy, we observed higher frequencies of elevated FVIII activity among older VTE patients. It is debatable whether an age-dependent cut-off definition will offer any advantage. Plasma FVIII levels indicate a wide inter-individual variation that is due to environmental and genetic determinants. However, as shown in a previous study completed by our group, only 3% of the variance of FVIII activity was predicted by age, sex, body mass index, and OC use (Luxembourg et al, 2009). Although no specific F8 gene polymorphisms have been associated with elevated plasma FVIII levels, the variation due to heritability was estimated to be approximately 40% (Souto et al, 2000). In addition, FVIII levels have been demonstrated to be mainly influenced by the amount of von Willebrand factor (VWF), polymorphisms of the low-density lipoprotein receptor-related protein (LRP) and the ABO blood group (Jenkens et al, 2012). It has been demonstrated that elevated FVIII levels are associated with an approximately 5- to 7-fold increased relative risk of VTE and VTE recurrence. It has also been suggested that increasing plasma levels of coagulation factors (e.g., FVIII, FIX, FXI, fibrinogen) with advancing age are likely to contribute to the higher incidence of VTE in the elderly (Rosendaal et al, 2007; Jenkens et al, 2012).
There are only a few studies analysing the familial component of VTE. In a recently published large Danish nationwide study including 30 179 siblings of 19 599 VTE cases, siblings of the index patient had a 2- to 3-fold higher risk of VTE compared with the general population (Christiansen et al, 2005). The calculated standardized incidence ratios decreased with the age of the index case, from 11·4 (95% CI 6·8–19·3) for index cases below the age of 19 years to 1·9 (95% CI 1·4–2·5) for index cases aged more than 50 years at the time of their first VTE (P < 0·001). The family history was also evaluated among VTE patients and control subjects from the MEGA (Multiple Environmental and Genetic Assessment of Risk Factors for Venous Thrombosis) study (Hansson et al, 2000). In this study, having a relative affected at a younger age and a higher number of affected relatives increased the chance of detecting a genetic risk factor up to 36%.
The identification of a thrombophilic defect may help clarify the cause of the first VTE. However, in the majority of cases, VTE is multicausal, and the laboratory diagnosis of thrombophilia does not exclude the presence of other diseases associated with increased thrombotic risk (e.g., malignant disease). It remains a matter of debate as to whether screening for thrombophilia is justified in clinical practice in symptomatic VTE patients. General testing for thrombophilia would be logical only if the knowledge of a thrombophilic defect would influence the future management of VTE patients in terms of the intensity and duration of anticoagulant therapy, taking preventive measures or avoiding exposure to avoidable risk factors. To date, there have been no prospective studies answering this question (Cohn & Middeldorp, 2008; Cohn et al, 2009). The only prospective randomized trial that aimed to assess the potential benefits and disadvantages of testing for thrombophilia, the Dutch NOSTRADAMUS trial, was terminated early due to insufficient patient enrolment (Cohn & Middeldorp, 2008).
There is agreement that the male sex, unprovoked VTE, proximal DVT, the presence of malignant disease, a positive D-dimer status after discontinuation of vitamin K-antagonists and residual venous thrombosis are risk factors for VTE recurrence (Zhu et al, 2009; Kearon et al, 2012). The relative risk of VTE recurrence with the presence of the common F5 R506Q and F2 G20210A gene mutations appears to be only mildly elevated (Ho et al, 2006; Segal et al, 2009). However, data from recently published studies provide evidence that inherited deficiencies of AT, PC or PS (Brouwer et al, 2009) and triple positivity of antiphospholipid antibodies are associated with a substantial increase to VTE recurrence risk (Pengo et al, 2010). Current guidelines do not recommend general screening for thrombophilia after VTE because the intensity or duration of anticoagulant therapy is not influenced by the test results in the majority of cases (Pernod et al, 2009; Baglin et al, 2010; Blaettler et al, 2010; Chong et al, 2012; Kearon et al, 2012). Moreover, one has to take into account the psychological and social consequences for the individual patient, such as persistent anxiety due to the knowledge of genetic test results or problems with acquiring life or disability insurance (Cohn et al, 2008). Another disadvantage is the cost of testing, which is approximately €500 for a complete thrombophilia screen (Machin, 2003). Although some cost-effectiveness studies have been published regarding testing for thrombophilia, which concluded that testing could indeed be cost-effective in some scenarios, the number of assumptions from inconsistent observational studies complicates the interpretation of the results of these studies (Wu et al, 2006; Simpson et al, 2009).
The present study has several potential limitations. We obtained data from a single-centre cross-sectional registry trial. This trial enrolled consecutive patients with acute or documented histories of VTE observed in our outpatients department. Therefore, there is a possibility of a patient selection bias. First, the proportion of patients with a first VTE at younger ages was high (median age of the entire cohort 43 years). Second, our university hospital is located near an international airport, which may heighten the proportion of patients with VTE related to long-term travel. Third, our university hospital operates several specialized tumour centres, which may increase the number of cancer-related VTEs at our institution.
Despite these limitations, our findings have several important implications. Because of the decline in the frequency of hereditary thrombophilia, selected screening strategies appear to be superior to the concept of overall general screening for thrombophilia. Screening for hereditary thrombophilia should be considered for patients younger than 40 years old after a first VTE episode, especially in cases of unprovoked VTE. The knowledge of presence or absence of a thrombophilic disorder may facilitate the decision as to whether to prolonged anticoagulant therapy should be continued in cases where there is uncertainty.