UK
Europe
USA&Canada
Aust.&NZ
Asia
Editorials Online™
Medical
Law News
E-texts
Links
Classifieds
Home
|
|
|
NEJM
Michael D. Lockshin, M.D., and Doruk Erkan, M.D. Two decades after the presence of anticardiolipin antibodies and a syndrome of spontaneous thrombosis and fetal death were linked,1 and after the publication of approximately 3500 reports in which the term "antiphospholipid syndrome" has been used, we still lack scientifically robust, evidence-based rules for the treatment of the antiphospholipid syndrome. This failure is due in part to a lack of well-designed prospective studies and in part to the clinical complexity of the syndrome. The complexity stems in part from the fact that the antiphospholipid syndrome actually comprises two syndromes, one that is not associated with another illness (primary antiphospholipid syndrome), and another that entails the additional burden of systemic lupus erythematosus or other rheumatic disease (secondary antiphospholipid syndrome; in up to one third of patients with lupus, the antiphospholipid antibody is detected). The term "antiphospholipid antibody syndrome" is also a point of
confusion. Originally, the most important autoantigens in the
disorder were thought to be negatively charged phospholipids, such as
cardiolipin. More recently, however, the principal autoantigen has
been shown to be A second point of confusion is that the criteria for the classification of the antiphospholipid syndrome2 include a range of illnesses: either thromboembolism or fetal loss and the persistence of moderate-to-high levels of IgG or IgM anticardiolipin antibodies or lupus anticoagulant. In the diagnosis of the antiphospholipid syndrome, the persistence, isotype, and titer of the autoantibodies are important considerations, because transient low titers of anticardiolipin antibodies occur in up to 5 percent of otherwise healthy people and the association of only IgA antibody with the antiphospholipid syndrome is not firmly established. The spectrum of the antiphospholipid syndrome allows for the stratification of risks and therapeutic options. Among patients with antiphospholipid antibodies, the risk of thrombosis at some future time is moderate in women with pregnancy loss alone, higher among those in whom thrombosis occurs, and extremely high in patients with the rare but devastating syndrome of rapidly progressive, multisystemic vascular occlusion, known as catastrophic antiphospholipid syndrome.3 By contrast, asymptomatic persons who are incidentally found to have positive antibody tests are at the lowest risk for any of these events. A high titer of autoantibodies, especially lupus anticoagulant, increases the likelihood of thrombosis, as do smoking, the use of oral contraceptives, surgery, immobilization, pregnancy, and heritable procoagulant states. Another measure of complexity is that patients with antiphospholipid antibodies who do not have recurrent thrombosis but who do have livedo reticularis, thrombocytopenia, cardiac-valve vegetations, a multiple sclerosis–like condition, or progressive cognitive dysfunction do not fulfill the usual criteria for the diagnosis of the antiphospholipid syndrome. In these patients, the risk of thrombosis, and hence the urgency of treatment with an anticoagulant agent, is unknown. Thrombosis and fetal loss can also be caused by mutations involving factor V (factor V Leiden), prothrombin, or methylenetetrahydrofolate reductase, or by deficiencies of protein C, protein S, or antithrombin III. Opinions differ, but the concomitant presence of two or more prothrombotic risk factors may increase the risk. Thus, a precise diagnosis requires an expensive and inconvenient panel of tests. Very few clinical clues differentiate the various types of thrombophilia. Arthralgia, leukopenia, thrombocytopenia, livedo reticularis, or a rheumatic illness (or a family history of rheumatic illness) increases the likelihood that the antiphospholipid syndrome is the cause of a newly diagnosed thrombosis. Two tests, a standardized enzyme-linked immunosorbent assay for anticardiolipin antibodies and a clotting test for lupus anticoagulant, readily and inexpensively confirm or rule out the antiphospholipid syndrome. Population studies usually identify the antiphospholipid syndrome and factor V Leiden as the most common causes of thrombophilia, each accounting for about 20 percent of cases of recurrent thrombosis among young patients. Approximately 15 percent of women with recurrent pregnancy loss have the antiphospholipid syndrome. The antiphospholipid syndrome and the genetic thrombophilias are proportionately less frequent among persons 65 years of age or older, because with increased age, atherosclerosis and cancer are more frequent causes of thrombosis. A heterogeneous clinical picture has led to discrepant recommendations for treatment of the antiphospholipid syndrome. Corticosteroids, prescribed in the 1980s because of the association of the syndrome with lupus, proved ineffective in preventing pregnancy loss.4 In pregnant women with a history of recurrent fetal loss, low doses of unfractionated heparin (5000 U administered subcutaneously twice daily) are as effective, with respect to the outcomes of pregnancy, as high doses (10,000 U or more twice daily).5 Patients with previous thrombosis require high doses. Prospective dose-ranging studies of warfarin for the treatment of thrombosis in patients who are not pregnant have not been done. A 1995 report of a retrospective study concluded that anticoagulant therapy that results in an international normalized ratio (INR) of 3.0 or higher affords better protection against recurrence than does less intense anticoagulant therapy.6 In this issue of the Journal, a Canadian consortium reports the results of a prospective, randomized, controlled study of the use of two doses of warfarin for prevention of thrombosis in patients with the antiphospholipid syndrome who had previous thrombosis.7 A total of 114 patients were treated with either moderate-intensity warfarin (target INR, 2.0 to 3.0) or high-intensity warfarin (target INR, 3.1 to 4.0). The follow-up period was 2.7 years. Only patients with moderate or high titers of anticardiolipin antibodies or lupus anticoagulant were included in the study. The investigators did not analyze thrombosis-triggering events or additional risk factors. Two surprising results emerged: in both treatment groups, the rate of recurrent thrombosis was low, and the lower dose of warfarin was as effective as the higher dose. Adverse effects were similar among patients in the two groups. These results refute recent calls for treatment with even higher doses of anticoagulant agents.8 The Canadian study warrants several caveats, however. Patients with arterial thrombosis, which is generally harder to treat than venous thrombosis, accounted for only approximately one fifth of the patients included in the study, although outcomes among patients with arterial thrombosis were similar to those among patients with venous thrombosis. Patients with a previous recurrence despite an INR of 2.0 or higher, which is relatively common among patients with the antiphospholipid syndrome who are followed in university clinics, were excluded. The dose and duration of anticoagulant treatment without a recurrence before enrollment are not discussed. Adherence to the protocol was imperfect: in five of eight patients who had recurrent thrombosis during treatment with the study drug, the INR was lower than 2.0. Nonetheless, this real-world demonstration now definitively permits the use of less intense anticoagulant therapy for most patients with the antiphospholipid syndrome and thrombosis. Several questions are not addressed by the Canadian study. Should asymptomatic persons in whom antiphospholipid antibodies are incidentally detected be treated? Many, if not most, such persons remain clinically well for decades, but a recent retrospective study of women in whom the antiphospholipid syndrome was diagnosed because of pregnancy loss (none of whom had a previous vascular thrombosis) suggests that treatment with low-dose aspirin provides protection against thrombosis.9 A prospective study of aspirin as compared with placebo, now under way in the United States, and a trial of aspirin as compared with low-dose warfarin plus aspirin, under way in the United Kingdom, should provide definitive answers.10,11 If thrombosis is triggered by surgery, by the use of oral contraceptives, or by other nonrecurrent triggers, is it necessary to treat patients with antiphospholipid antibodies indefinitely? The answer is unknown. Should all pregnant women be tested for the antiphospholipid antibody, and, if the results of the test are positive, should they be treated? The high frequency of antiphospholipid antibodies in normal persons suggests that the answer to this question, certainly for patients with a low titer of anticardiolipin antibodies, is no. Can anticoagulant therapy end after the first trimester of pregnancy? No studies definitively support this option, and treatment throughout the pregnancy may prevent pregnancy-associated intravascular thrombosis. Can anticoagulant therapy prevent or treat livedo reticularis, valvular heart disease, or cognitive dysfunction? The answer to this question, too, is unknown. What is the best treatment for patients in whom new thromboses develop despite high-intensity anticoagulant therapy? Most experts favor the use of plasmapheresis, intravenous immune globulin, or both, particularly in the case of patients with catastrophic antiphospholipid syndrome; support for this approach is anecdotal but justifiable, because the mortality rate among patients with catastrophic antiphospholipid syndrome is very high. At the 10th International Congress on Antiphospholipid Antibodies, consensus groups summarized the available data, which are meager, and offered expert opinions on most of these questions.12 In the future, the treatment of patients with the antiphospholipid syndrome
may not be anticoagulant therapy. Mechanisms by which antiphospholipid
antibodies induce thrombosis and pregnancy loss include endothelial
activation, interference with placental anticoagulant protein I (annexin
V) by anti– When new treatments are introduced, prospective, risk-stratified, blinded, controlled studies like that reported by Crowther et al.7 must be conducted. An American antiphospholipid syndrome registry funded by the National Institutes of Health, the European Euro-Phospholipid Project,13 and several recently established Canadian and American clinical trial consortia provide opportunities to conduct the correct type of study from the beginning. If these opportunities are taken, we will not have to wait another 20 years to know the best treatment for patients with the antiphospholipid syndrome.
From the Hospital for Special Surgery, Joan and Sanford Weill College of Medicine of Cornell University, New York. References
|
2-glycoprotein
I (apolipoprotein H), a protein that binds cardiolipin and other
anionic phospholipids and functions in a physiologic disposal system
for apoptotic cells. The phospholipid that is bound by