The G20210A Prothrombin Mutation and the Physicians’ Health Study
To the Editor:
We would like to offer our interpretation of the recent article by Ridker et al,1 which was derived from the Physicians’ Health Study (PHS). To our knowledge, this is the only report that does not confirm the association between the G20210A prothrombin mutation and deep vein thrombosis. In this report, the mutation was also unrelated to myocardial infarction and stroke.1
The G20210A allele is associated with increased prothrombin levels in blood and, since 1996, it has been reported concordantly as a moderate and significant risk factor for deep vein thrombosis by ≥12 different case-control studies (published in extenso) involving 2657 patients and 4070 healthy controls (for partial review see Reference 2).
The PHS was conceived as a randomized trial to investigate the effects of aspirin versus placebo on cardiovascular mortality during a 5- to 10-year follow-up period.1 3 Enrolled subjects constituted a special group: American male physicians, 41 to 84 years of age, who were selected according to the absence of previous thrombotic events (ie, with a low cardiovascular risk profile, which was reduced even further by the intake of aspirin).1 3 The extremely low risk of the PHS population is supported by the participants’ initial healthy conditions, despite a mean age at enrollment of 60 years, and by the authors’ own statement at follow-up, that “cardiovascular death rates in this trial were exceptionally low… only 15% of that expected for a general population of white men with the same age distribution over a similar period.”3 Thus, the PHS participants, compared with the general population, might be more appropriate for the investigation of protective rather than predisposing factors for cardiovascular disease.
Approximately 6% of the PHS participants were not white; the rest represented the American heterogeneous racial “melting pot.”4 For studies aimed at identifying new genetic components of disease, it is considered pivotal to investigate homogeneous racial groups or families.
The outcome events in the PHS were self-reported by the participants through mailed questionnaires and were retrospectively confirmed through clinical records.3 Although it is fair to assume that a possible underestimation of events in the original randomized study was distributed equally among the aspirin and placebo groups and did not bias the end point, in this nested case-control study,1 silent events or missed diagnoses may have reduced the quality of controls by inappropriate assignment.
Therefore, in our opinion, such negative results from the PHS study should be considered with caution. The G20210A prothrombin mutation should not be judged irrelevant for venous thrombosis; indeed, in selected cases, it may also be relevant for arterial thrombosis, as shown in young patients with ischemic stroke and no other vascular risk factor.5
- Copyright © 2000 by American Heart Association
Ridker PM, Hennekens CH, Miletich JP. G20210A mutation in prothrombin gene and risk of myocardial infarction, stroke, and venous thrombosis in a large cohort of US men. Circulation. 1999;99:999–1004.
Alhenc-Gelas M, Arnaud E, Nicaud V, et al. Venous thromboembolic disease and the prothrombin, methylene tetrahydrofolate reductase and factor V genes. Thromb Haemost. 1999;81:506–510.
Steering Committee of the Physicians’ Health Study Research Group. Final report on the aspirin component of the ongoing Physicians’ Health Study. N Engl J Med. 1989;321:129–135.
Ridker PM, Miletich JP, Hennekens CH, et al. Ethnic distribution of factor V Leiden in 4047 men and women. Implications for venous thromboembolism screening. JAMA. 1997;277:1305–1307.
De Stefano V, Chiusolo P, Paciaroni K, et al. Prothrombin G20210A mutant genotype is a risk factor for cerebrovascular ischemic disease in young patients. Blood. 1998;91:3562–3565.
As a possible explanation for our null findings, Andreotti and colleagues suggest that the Physicians’ Health Study might be an inappropriate population in which to explore novel risk factors for cardiovascular disease. However, in this same cohort of apparently healthy middle-aged men, we have previously demonstrated that factor V LeidenR1 and homocysteineR2 are risk factors for future venous thrombosis and that C-reactive protein, fibrinogen, tissue-type plasminogen activator antigen, and D-dimer are all strong markers for future arterial thrombosis.R3 Andreotti and colleagues also suggest that the heterogeneity of the American population might explain our null results because ≈5% of our cohort is not white. However, the exclusion of these individuals had no impact on our null results. Finally, it is suggested that silent cardiovascular events or missed diagnoses might have resulted in an uninformative null result. We think this possibility is unlikely because all of our study participants were physicians with access to high-quality medical care. Further, our procedures for end point validation include full medical record review, and they have consistently provided complete and accurate data.
Our report is not the only null data presented concerning the G20210A mutation and arterial thrombosis. In a study of 560 men with a first myocardial infarction and 646 controls, Doggen et alR4 reported that the 95% confidence intervals associated with the presence of the G20210A mutation ranged from 0.6 to 3.8. Similar null data were reported in an American study of 106 young women with stroke.R5 With regard to venous thrombosis, we reiterate that our data are consistent with a modest, although nonsignificant, role for the prothrombin mutation (relative risk, 1.7; 95% confidence interval, 0.9 to 3.1; P=0.08). This risk, however, was substantially less than that associated with factor V Leiden (relative risk, 3.0; 95% confidence interval, 1.7 to 5.5; P<0.001).
Although we agree with Andreotti and colleagues that the generalization of any one study result must be done cautiously, we also think that an interpretation of conflicting genetic analyses must consider variations in mutation rates in control populations. As outlined in our article, reported background rates for the prothrombin mutation have ranged between 0.7% and 3.9%, depending on the sample size and country of origin. We evaluated 1774 controls, a sample substantially greater than that reported in any prior analysis. Gene sequencing was also performed on a subset of participants to confirm mutation status. Thus, we have confidence that our reported mutation rates are accurate.
Ridker PM, Hennekens CH, Lindpaintner K, et al. Mutation in the gene coding for coagulation factor V and the risk of myocardial infarction, stroke, and venous thrombosis in apparently healthy men. N Engl J Med.. 1995;332:912–917.
Ridker PM, Hennekens CH, Selhub J, et al. Interrelation of hyperhomocyst(e)inemia, factor V Leiden, and risks of future venous thromboembolism. Circulation.. 1997;95:1777–1782.
Ridker PM. Fibrinolytic and inflammatory markers for arterial occlusion: the evolving epidemiology of thrombosis and hemostasis. Thromb Haemost.. 1997;78:53–59.
Doggen JMC, Manger Cats V, Bertina RM, et al. Interaction of coagulation defects and cardiovascular risk factors: increased risk of myocardial infarction associated with factor V Leiden or prothrombin 20210A. Circulation.. 1998;97:1037–1041.
Longstreth WT, Rosendaal FR, Siscovick DS, et al. Risk of stroke in young women and two prothrombotic mutations: factor V Leiden and prothrombin gene variant (G20210A). Stroke.. 1998;29:577–580.