High-Risk of Cerebral Vein Thrombosis in Prothrombin Gene Mutation Carriers and in Oral Contraceptive Users

Abstract & Commentary

The two most common causes of familial thrombophilia are mutations in the genes for factor V and prothrombin. These mutations are known to predispose to deep vein thrombosis of the lower extremities. In this study, Martinelli and colleagues sought to determine if these familial thrombophilias predisposed to cerebral vein thrombosis. The mortality of cerebral vein thrombosis is as high as 30%. Martinelli et al compared the prevalence of genetic and nongenetic risk factors in 40 patients with cerebral vein thrombosis, 80 patients with deep vein thrombosis of the lower extremities, and 120 healthy controls. The controls were matched to the patients for sex, age, geographic region, and level of education. The prevalence of the prothrombin gene mutation was higher in patients with cerebral vein thrombosis (20%) than in controls (3%), odds ratio 10.2 (95 confidence interval 2.3-31.0) and comparable to that of patients with deep vein thrombosis (18 %). Similar results were obtained for factor V gene mutation. Oral contraceptive use was more frequent among women with cerebral vein thrombosis (96%) than among control women (32%), odds ratio 22.1 (5.9-84.2). Women taking oral contraceptives who had a prothrombin gene mutation (7 patients and 1 control) had an odds ratio of 149.3 (31-711). The mean age of the women with cerebral vein thrombosis was 30 years and, for lower extremity DVT, also 30 years. (Martinelli I, et al. N Engl J Med 1998;338:1793-1797.)

COMMENT BY SARAH L. BERGA, MD

My companion article in the October 1998 issue of OB/GYN Clinical Alert reviews the recent study demonstrating that oral contraceptive use was found to protect women with the BRCA1 or BRCA2 mutation from ovarian cancer. It provides good news about oral contraceptives. On the other hand, these findings about oral contraceptive use and the risk of thrombosis sound a cautionary note. More importantly, the articles demonstrate the emergence of molecular medicine as a tool for individualizing therapy. It may not yet be worthwhile or appropriate to genotype our patients as part of medical decision-making. However, as molecular epidemiology clarifies individual risk and as genotyping technologies advance, sooner or later there will come a time when it will be both feasible and worthwhile to genotype. When that intersection will occur depends on a number of factors, including the acceptance by society in general (and insurers in particular) that having an abnormal gene is not a "flaw" for which the individual must "pay," but a vulnerability or predisposition that must be taken into account when making a rational medical decision. Further, at the present time, physicians and pharmaceutical companies are often held liable for "bad" outcomes that are not knowable and, therefore, not preventable. Genotyping could help to lessen this liability.

The accompanying editorial (Bertina RM, Rosendaal FR. N Engl J Med 1998;338:1840-1841) also makes some important points. The proportion of carriers of factor V Leiden in the white population ranges from 2-15%. Heterozygotes have a seven-fold risk of DVT and homozygotes an 80-fold increase. All alleles are thought to derive from a common ancestor. The prevalence of prothrombin mutation is 0.7-4% of the white population and is rare in nonwhites. Oral contraceptives also change the balance between hemostatic and fibrinolytic components of the blood and are an important cause of nongenetic thrombophilia. This study suggests that oral contraceptive use in women with familial thrombophilia is hazardous, but neither group of authors called for routine screening of women contemplating oral contraceptive use. Since large-scale screening is not currently feasible, the next best step would be to design a less thrombogenic oral contraceptive. This makes more sense than screening, anyway, because oral contraceptives increase the risk of thrombosis even in women without familial thrombophilia. One strategy might be to avoid the use of ethinyl estradiol. Another strategy might be to nonorally deliver ethinyl estradiol in combination with a progestin.