IntroductionNearly 150 years ago, Virchow postulated that thrombosis was caused by changes in the flow of blood, the vessel wall, or the composition of blood. This concept created the foundation for subsequent investigation of hereditary and acquired hypercoagulable states. This review will focus on an example of the use of modern genetic epidemiologic analysis to evaluate the multigenic pathogenesis of the syndrome of juvenile thrombophilia.Juvenile thrombophilia has been observed clinically since the time of Virchow and is characterized by venous thrombosis onset at a young age, recurrent thrombosis, and a positive family history for thrombosis. The pathogenesis of juvenile thrombophilia remained obscure until the Egeberg observation, in 1965, of a four generation family with juvenile thrombophilia associated with a heterozygous antithrombin deficiency subsequently identified as antithrombin Oslo (G to A in the triplet coding for Ala 404).1,2 The association of a hereditary deficiency of antithrombin III with thrombosis appeared to support the hypothesis, first put forward by Astrup in 1958, of a thrombohemorrhagic balance.3 He postulated that there is a carefully controlled balance between clot formation and dissolution and that changes in conditions, such as Virchow’s widely encompassing triad, could tip the balance toward thrombus formation.The importance of the thrombohemorrhagic balance in hypercoagulable states has been born out of two lines of investigation: evidence supporting the tonic activation of the hemostatic mechanism and the subsequent description of additional families with antithrombin deficiency and other genetically abnormal hemostatic proteins associated with inherited thrombophilia. Assessing the activation of the hemostatic mechanism in vivo is achieved by a variety of measures, including assays for activation peptides generated by coagulation enzyme activity. Activation peptides, such as prothrombin fragment1+2, are measurable in normal individuals, due to tonic hemostatic activity and appear elevated in certain families with juvenile thrombophilia.4
In the past 25 years since Egeberg’s description of antithrombin deficiency, a number of seemingly monogenic, autosomal dominant, variably penetrant hereditary disorders have been well established as risk factors for venous thromboembolic disease. These disorders include protein C deficiency, protein S deficiency, antithrombin III deficiency, the presence of the factor V Leiden mutation, and the recently reported G20210A prothrombin polymorphism.5,6 These hereditary thrombophilic syndromes exhibit considerable variability in the severity of their clinical manifestations. A severe, life-threatening risk for thrombosis is conferred by homozygous protein C or protein S deficiency, which if left untreated, leads to death.7,8 Homozygous antithrombin III deficiency has not been reported but is also likely to be a lethal condition. Only a moderate risk for thrombosis is conferred by the homozygous state for factor V Leiden or the G20210A polymorphism.9,10 In contrast to homozygotes, the assessment of risk in heterozygotes, with these single gene disorders, has been complicated by variable clinical expression in family members with identical genotypes.11 Consideration of environmental interactions has not elucidated the variability of clinical expression. Consequently, it has been postulated that more than one genetic risk factor may co-segregate with a consequent cumulative or synergistic effect on thrombotic risk.12
A number of co-segregating risk factors have been described in the past few years. Probably the best characterized interactions are between the common factor V Leiden mutation, present in 3% to 6% of the Caucasian population,13,14 and the less common deficiencies of protein C, protein S, and antithrombin III. The factor V Leiden mutation does not, by itself, confer increased risk of thrombosis. The high prevalence of the mutation, however, creates ample opportunity for interaction with other risk factors when present.The G20210A prothrombin polymorphism has a prevalence of 1% to 2% in the Caucasian population and, thus, may play a similar role to factor V Leiden. A number of small studies have documented an interaction of G20210A with other risk factors.15-17 A limited evaluation of individuals with antithrombin III, protein C, or protein S deficiency revealed a frequency of 7.9% for the G20210A polymorphism, as compared to a frequency of 0.7% for controls.18 The G20210A polymorphism was observed in only 1 of the 6 protein C-deficient patients.18
In the present state, the elucidation of risk factors for venous thromboembolic disease attests to the effectiveness of the analytical framework constructed from the molecular components of Virchow’s triad, analyzed in the context of the thrombohemorrhagic balance hypothesis. Two investigative strategies have been used to study thromobophilia: clinical case-control studies and genetic epidemiologic studies. The latter strategy has gained considerable utility, based on the remarkable advances in molecular biology over the past two decades. Modern techniques of genetic analysis of families offer important opportunities to identify cosegregation of risk factors with disease.19 The essence of the genetic epidemiologic strategy is the association of clinical disease with alleles of specific genes. It is achieved either by the direct sequencing of candidate genes or by demonstration of linkage to genetic markers.