Lipoprotein(a): another risk factor for venous thrombosis?

Abstract

In this issue of the Journal, Marcucci and colleagues present additional evidence supporting the link between elevated levels of lipoprotein(a) [Lp(a)] and venous thromboembolism (1). Elevated levels of Lp(a) have been shown with a reasonable degree of certainty to be associated with arterial thrombosis. For example, Seed et al (2) found mean Lp(a) levels to be approximately threefold higher in patients with coronary heart disease than in healthy controls. However, the association between elevated levels of Lp(a) and venous thromboembolism has not been confirmed in adults, to date. The study by Marcucci et al is methodologically strong. In 603 patients with venous thrombosis and 430 matched controls, the authors demonstrated an association between elevated levels of Lp(a) and venous thromboembolism (odds ratio 2.1; 95% confidence interval: 1.4 to 3.2). This association remained after adjusting for covariates, and was present in both the overall sample and when the analysis was limited to patients with unprecipitated venous thromboembolism. Strengths of the study include the objective confirmation of venous thromboembolism and the use of a well-defined (and well-matched) control group. In univariate analyses, the authors found the expected association between venous thromboembolism and the presence of the factor V Leiden mutation, the prothrombin gene 20210A polymorphism, hyperhomocysteinemia, the presence of an anticardiolipin antibody, and known transient risk factors such as the use of oral contraceptives or hormone replacement therapy. In a multivariate analysis confined to the 341 patients with idiopathic venous thromboembolism, the factor V Leiden mutation, the prothrombin gene polymorphism, hyperhomocysteinemia, the presence of an anticardiolipin antibody, and Lp(a) levels of more than 300 mg/L remained independent predictors of venous thrombosis. Finally, patients were tested for their Lp(a) levels well after their episode of venous thrombosis to ensure that the presence of acute thrombosis did not influence the Lp(a) levels. The finding that elevated levels of Lp(a) are associated with venous thrombosis supports the work of von Depka and colleagues who found a similar association (3). Venous thromboembolism remains an important cause of morbidity and mortality, and identification of the cause of venous thromboembolism is important since the duration of anticoagulation and the need for supplementary therapy may be influenced by the identified precipitant. For example, patients with venous thromboembolism in the setting of cancer are treated most effectively with extended duration anticoagulation, whereas those with venous thromboembolism attributed to a transient, resolved risk factor are safely and effectively treated with 3 months of oral anticoagulant therapy. Confirmation that elevated levels of Lp(a) are associated with venous thrombosis is particularly important since preliminary evidence suggests that treatment with niacin can effectively lower Lp(a) levels. However, whether such treatment reduces the risk of thromboembolic complications remains to be proven. Lp(a) is a complex macromolecule consisting of lowdensity lipoprotein linked to apolipoprotein(a), a molecule that is highly homologous with plasminogen. Measurement of Lp(a) is problematic since the measured concentration of Lp(a) is affected by the temperature at which the samples are stored. In addition, freeze-thaw artifact can affect measured levels. There are also 34 described isoforms of apolipoprotein(a) with differing molecular masses; validated assays for Lp(a) should therefore be sensitive to all of these isoforms. Marcucci and colleagues used a commercially available enzyme-linked immunosorbent assay with a detection limit of 1 mg/L. Given the difficulties with measurement, additional confirmatory studies using other assays and finding similar associations will be required to confirm that elevated Lp(a) levels are associated with venous thromboembolism. Confidence in the association between Lp(a) levels and venous thromboembolism would be further enhanced if prospective studies with a standardized measurement procedure supported this association. Now that data are available from a large cross-sectional study, future investigations should employ a prospective design to provide a greater degree of confidence in this association. If Lp(a) is a risk factor for venous thromboembolism, it will join a legion of other recently identified risk factors, Am J Med. 2003;115:667– 668. From St. Joseph’s Hospital, Hamilton, Ontario, Canada. Requests for reprints should be addressed to Mark Crowther, MD, St. Joseph’s Hospital, 50 Charlton Avenue East, Room L-208, Hamilton, ON L8N 4A6, Canada, or crowthrm@mcmaster.ca.