Genetic hypercoagulability: prevention suggests testing family members.

Dahlback in 1993 and Bertina in 1994 made the landmark discovery that resistance to activated protein C due to a point mutation in the factor V gene (G1691A in exon 10, leading to Arg506Gln) is associated with a 6to 8-fold increased risk of venous thrombosis.1,2 The importance of these findings, which have since been confirmed and cited more than 2000 times in the medical literature, is due to the fact that heterozygosity for factor V Arg506Gln (generally known as factor V Leiden) is present in approximately 20% of unselected patients who present with a first episode of venous thrombosis3 and, most important, in approximately 4% of general white populations.4 Another piece of evidence supporting the importance of genetic risk factors in causing venous thrombosis was added in 1996, when again Bertina and his colleagues found that the prothrombin G20210A polymorphism has a frequency of 6% in patients with a first episode of venous thrombosis, contrasting with approximately 2% in the general white population.5 Adding to factor V Leiden and prothrombin G20210A the rarer deficiencies of the naturally occurring anticoagulant proteins antithrombin, protein C, and protein S, the proportion of venous thromboses attributable to genetic factors is 25% to 30%, at least as large as that attributable to well-established acquired risk factors such as surgery, trauma, prolonged immobilization in bed (more than 7 days), pregnancy and puerperium, estrogen use, and cancer. With these genetic risk factors of venous thrombosis having such a weight and prevalence in the general population, the first obvious question is whether they should be looked for in healthy individuals with no personal or family history of thrombosis, particularly when they are exposed to acquired factors that may interact with genetic factors to increase the risk of venous thrombosis. The magnitude of the interaction between genetic and acquired factors can be exemplified by one of the most frequently occurring interactions, that is, that between heterozygosity for factor V Leiden and oral contraceptive intake. Given that oral contraceptives alone increase from 4to 6-fold the relative risk of venous thrombosis, the concomitant presence of factor V Leiden brings the relative risk up to 30-fold or more.6 Despite this impressive increase in the relative risk, one must consider that the annual incidence of venous thrombosis in women of childbearing age is approximately 1 in 10 000. Hence, more than 20 000 women should be tested and counseled to prevent one episode of venous thrombosis, and more than 2 million tested and counseled to prevent one death from pulmonary embolism.7 These figures clearly indicate that it is hardly justified to screen for thrombophilia all women who are considering oral contraceptive intake. By analogy, the same negative consideration applies to pregnant women and to individuals of both sexes exposed to major surgery, immobilization, cancer, and hormone replacement therapy, because there is no evidence that the interaction among these risk factors and heterozygosity for factor V Leiden is more thrombogenic than that between the latter and oral contraceptive intake. A special situation is that of asymptomatic relatives of index patients with a thrombophilic defect diagnosed after an episode of venous thrombosis. In these individuals, screening is obviously more focused and potentially more fruitful than in the general population because all the genetic risk factors are transmitted as autosomal dominant traits. The first issue is whether the results of screening would be advantageous for family members. In principle, screening entails the possibility of giving advice for antithrombotic prophylaxis to these individuals. There is no evidence that their lifelong risk of thrombosis justifies exposure to the risks associated with primary prophylaxis with oral anticoagulants, because many individuals with a genetic thrombophilic abnormality never develop thrombotic symptoms, nor do these defects increase mortality.8,9 However, when asymptomatic thrombophilic individuals are exposed to acquired risk factors, the increased risk of thrombosis might in principle be reduced with relatively little risk and cost by adopting short-term prophylaxis with low-dose unfractionated or low-molecular-weight heparin. What is the evidence in favor of this diagnostic strategy and of the ensuing prophylaxis strategy? In a retrospective study conducted in Italy in 1994, 238 patients with antithrombin, protein C deficiency, or protein S deficiency were analyzed for the incidence of thrombotic events prediagnosis and postdiagnosis of the thrombophilic abnormality.10 In the prediagnosis period, antithrombotic prophylaxis was not implemented at the time of exposure to acquired thrombosis risk factors. However, after diagnosis, shortterm prophylaxis with heparin was usually implemented. This approach reduced the incidence of both first (0.7/100 patient-years postdiagnosis vs 1.7 prediagnosis) and recurrent (1.3/100 vs 4.8/100 patient-years) thrombotic episodes.10 Among the limitations of this study are its retrospective design and the fact that factor V Leiden and prothrombin G20210A were not known as risk factors at the time of the study and so they were not considered in the analysis.10 Another piece of evidence in favor of the usefulness of screening and of antithrombotic prophylaxis during risk periods stems from a prospective cohort study carried out in Europe and Canada in 208 affected but asymptomatic family members of 94 symptomatic antithrombin, protein C–, and protein S–deficient patients who had a venous thrombotic event.11 During 40 periods of exposure of these individuals to acquired risk factors, the decision to use antithrombotic prophylaxis and the regimens used were left to the choice of the treating physicians. The incidence of risk period–related, objectively documented, first venous thromboembolic episodes was 4.5% in the individuals who received prophylaxis, as compared with 16.7% in those who did not receive prophylaxis. Limitations of this study are the very small number of