Q J Med 2000; 93: 523-530
© 2000 Association of Physicians
Prognostic factors and clustering of serious clinical outcomes in antiphospholipid syndrome
From the Department of Pathophysiology, Medical School, National University of Athens, Athens, Greece 1 Clinical and Molecular Epidemiology Unit, Department of Hygiene and Epidemiology, University of Ioannina School of Medicine, Ioannina, Greece and Department of Medicine, Tufts University School of Medicine, Boston, USA
Received 7 April 2000 and in revised form 7 June 2000
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
We assessed whether initial clinical presentations suggestive of antiphospholipid syndrome (APS) predicted the subsequent rate and type of serious clinical outcomes. Eighty-two consecutive patients with anticardiolipin antibodies or lupus anticoagulant were followed for 814 person-years after a first event suggestive of APS (livedo reticularis, thrombocytopenia, autoimmune haemolysis, thrombosis, central nervous system manifestations, recurrent abortions). The hazard of developing a second event was largest in patients with antibodies recognizing ß2 glycoprotein I who had autoimmune haemolysis as the first event (hazard ratio HR 2.70, p=0.018) and smallest in patients without such antibodies who had recurrent abortions as their first event (HR 0.37, p=0.028). Subsequent serious events in patients with venous and arterial thromboses, recurrent abortions, central nervous system manifestations and autoimmune haemolytic anaemia were likely to be of the same type as the presenting event (odds ratio (OR) 3.76, 5.90, 77.7, 6.92, and 7.13, respectively. Adjusting for therapy, the rate of subsequent serious events was 6.86-fold higher (p=0.0001) in patients presenting with two events, 1.56-fold higher (p=0.038) in autoimmune haemolysis presentations, 1.69-fold higher (p=0.004) in patients with anti-ß2-glycoprotein-I antibodies, and 46% (p=0.063) lower in thrombocytopenia presentations. Initial clinical features determine the long-term evolution of APS, and specific types of clinical manifestations cluster during the course of the disease.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
The antiphospholipid syndrome (APS) is a multisystem disorder characterized by various combinations of vascular thromboses, pregnancy morbidity, and potentially an array of other clinical manifestations1 in the setting of antibodies against negatively charged phospholipids.2,3 The strength of the associations with specific clinical presentations has varied, and the complexity of the clinical presentations defies a straightforward classification. Laboratory diagnosis is also not straightforward. The most popular antiphospholipid assay uses cardiolipin4 as the antigen, but anticardiolipin antibodies do not recognize cardiolipin in the absence of ß2 glycoprotein I (ß2GPI). ß2GPI is present in human as well as bovine serum used in the conventional anticardiolipin assays.5,6 The introduction of phospholipid-free anti-ß2-glycoprotein detection assays7–9 has revealed that anti-ß2-glycoprotein-I antibodies may be more strongly associated with APS than anticardiolipin antibodies per se.10,11
The outcome of patients with APS can be markedly variable. Some patients have recurrent or even catastrophic events, while others never have a second event after their initial clinical presentation. There are only sparse data on overall prognosis and clinical evolution of this heterogeneous syndrome. Prior studies have predominantly focused on the recurrence of thrombotic events12,13 and their therapeutic management.14,15 According to a recent international consensus panel, thromboses and pregnancy morbidity should qualify as clinical criteria for definite APS when accompanied by adequate laboratory markers.16 The consensus panel has nevertheless recognized that we still need more evidence on the importance and associations of other clinical features such as autoimmune haemolytic anaemia, thrombocytopenia and livedo reticularis.16 It is unknown whether we can predict the outcome of patients with symptoms suggestive of APS on the basis of their initial disease presentation. Therefore, we undertook a study to determine whether various initial clinical presentations suggestive, but not necessarily definitive, of APS in patients with antiphospholipid antibodies or lupus anticoagulant could predict the rate and types of serious subsequent clinical outcomes in a cohort with long-term follow-up. In this setting, we also evaluated the potential effect of the auto-antibody immune response for antiphospholipid and anti-ß2GPI specificities in terms of their clinical significance and interactions with specific clinical manifestations.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Study cohort and definitions
This is the complete cohort of 82 consecutive patients with clinical manifestations suggestive, but not necessarily definitive, of APS and laboratory support for the syndrome who were seen at the University of Athens Department of Pathophysiology, the largest referral centre for autoimmune diseases in Greece. Patients were followed on a regular basis as out-patients every 2 months, with additional unscheduled visits and hospitalizations for serious events. Clinical information included also retrospective details on major clinical diagnoses and laboratory evaluations for patients already diagnosed with APS before referral to our clinic. Patients were prospectively followed from January 1994 until November 1998, and additional retrospective information was collected on events preceding this time period. Retrospective data were cross-validated on hospitalization reports.
Patients were included in the study if they had high anticardiolipin IgG or IgM titres (at least 5 SD above the mean of 100 normal individuals, typically at two measurements at least 6 weeks apart) or positive lupus anticoagulant test, along with a first clinical manifestation suggestive, but not necessarily definitive, of APS. The following clinical manifestations were considered as first suggestive events: livedo reticularis, thrombocytopenia (<100x109/l platelets), leg ulcers, venous thrombosis (deep vein thrombosis, pulmonary embolism, Budd-Chiari syndrome, renal vein thrombosis, diffuse alveolar haemorrhage), arterial thrombosis (peripheral arterial thrombosis, myocardial infarction, retinal artery thrombosis, or renal artery thrombosis), recurrent abortion (second or third trimester), autoimmune haemolytic anaemia (drop in haemoglobin >30 g/l within 
2 months, positive Coombs test, reticulocytes >5% and/or increased unconjugated bilirubin), and central nervous system involvement (stroke, multi-infarct dementia, anterior ischaemic optic neuropathy, chorea, epilepsy and organic psychosis, associated with infarcts on magnetic resonance imaging).17,18 For patients with thrombotic events we excluded protein C, protein S, and antithrombin III deficiency.
Laboratory data
IgG and IgM anticardiolipin antibodies and anti-ß2GPI antibodies were determined by enzyme-linked immunosorbent assay (ELISA) as previously described.19,10 A titre of 100 units corresponds to 3 SD above the mean of controls. For antibodies to cardiolipin (aCL), the ELISA used 50 µg/ml cardiolipin in ethanol (cardiolipin 1649, Sigma) as antigen on flat-bottom plates (Nunc) which were left to dry overnight at 4 °C. After washing with phosphate-buffered saline, non-specific binding was blocked by 10% bovine serum in phosphate-buffered saline.19 For determination of antibodies to ß2GPI, ELISA was done on polystyrene plates (Lindro/Titertek; ICN Biomedical; Australian Nuclear Science and Technology Organization) incubated overnight with 10 µg/ml ß2GPI (Behring) in phosphate-buffered saline. Blocking was performed with phosphate-buffered saline-gelatin 0.5%.10 Anti-ß2GPI antibodies detected in a cardiolipin-free system have been reported to be better predictors of APS.20,21 Lupus anticoagulant was assayed using activated partial thromboplastin time, kaolin clotting time and dilute viper venom time.22
Predictive models
Two separate analyses were performed. First, we evaluated predictors of developing a second clinical event from those among the list of first suggestive events (livedo reticularis, thrombocytopenia, Coombs-positive haemolytic anaemia, thrombosis, central nervous system manifestations, leg ulcers, or recurrent abortion).17,18 The second event had to be of different type than the first event.
Second, we evaluated predictors of the rate of serious clinical outcomes during follow-up. Recurrent events of the same type counted as separate outcomes. We included all serious clinical events which may be related directly or indirectly to APS. We specified a priori that these should include venous or arterial thrombosis (as defined above), recurrent abortions (counting all abortions separately after the second one), haemolytic anaemia (as defined above), symptomatic pulmonary hypertension (>35 mmHg), chronic renal failure (progressive and irreversible decline in glomerular filtration rate <60 ml/min), and central nervous system involvement (as defined above). We did not consider livedo reticularis, leg ulcers and thrombocytopenia as serious events, because they are not as severe as the events listed above (no patient had <50;x109/l platelets during follow-up and no patient had severe leg ulcers) and their documentation during follow-up might have been erratic.
For both analyses, the following variables at the time of first manifestation were considered as predictors: type of first clinical manifestation(s), age, sex, diagnosis of systemic lupus erythematosus (as defined by standard criteria),23 IgG and IgM anticardiolipin titres, anti-ß2GPI titres, and calendar time. Therapy was divided into three categories on the basis of prior evidence of its efficacy in preventing thrombotic manifestations: (a) warfarin (typically aiming for high INR),13,14 (b) aspirin, and (c) neither of the two. Therapy was considered as a predictor for the rate of serious events, and the incidence of events per time period was analysed separately for the three categories of treatment. No patients had received warfarin or aspirin before having a second event.
Statistical analysis
Proportional hazards models evaluated predictors of the time to a second event. We also considered models based on the interaction of clinical presentations with anti-ß2GPI in order to assess the interaction of the initial clinical and immunological features of the syndrome. Multivariate models were fit using backward elimination of variables based on a likelihood ratio test, with p>0.10 for removal and p<0.05 for entry of variables. We also performed Kaplan-Meier analyses, and comparisons were made using the log-rank test.
Poisson regression analyses were used to evaluate the effect of candidate predictors on the rate of subsequent serious clinical events. The regression form is ln(I) = b0+b1X1+ ... +bnXn where ln(I) is the natural logarithm of the incidence of events.24 For each patient, periods with different types of therapy were considered as separate observations.
Further, we assessed whether subsequent events were likely to be the same as the initial presenting clinical event. When there were zero counts in any cells of the 2x2 table, 0.5 was added, to calculate the odds ratio (OR).
Analyses were done using Advanced SPSS.25All p values are two-tailed.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Description of the cohort
The 82 patients (65 women) had a total follow-up of 814 patient-years after the first suggestive clinical event (mean 119.1 months, median 108 months, interquartile range IQR 57–156 months). Five patients presented immediately with two different types of clinical events, and another 68 developed a second different event during follow-up. Three patients died during follow-up (catastrophic syndrome n=2, recurrent ischaemic strokes n=1). Mean age at the first clinical event was 27.8 (IQR 21–32.25) years. At initial presentation, 34 patients had systemic lupus erythematosus. Mean IgG anticardiolipin antibodies titre was 376 (IQR, 200 to 500) units and mean IgM anticardiolipin antibodies titre was 229 (IQR 100–300) units. Fifty-four patients had anti-ß2GPI antibodies.
On average, patients spent 17.3% of their time on aspirin (IQR 0–21.3%) and 29.8% on warfarin (IQR 0–50%). Three patients had bleeding that was possibly related to warfarin therapy (recurrent nosebleeds, bleeding mouth ulcers and genital haemorrhages), but none was severe. No patient had major side-effects from aspirin use.
Time to second event
The median time from the first clinical manifestation to a second different event was 3 years (mean 5.54 years). The type of presenting clinical event was a strong predictor of this disease progression. Specifically, in univariate Cox regressions, the hazard ratio (HR) for progression was 1.95 (p=0.04) when the presenting clinical event was haemolytic anaemia and 0.54 (p=0.05) when it was recurrent abortions. There were also suggestions that a second event occurred earlier in patients with anti-ß2GPI antibodies (HR 1.48, p=0.14). Under multivariate analysis, presentation with haemolytic anaemia (HR 2.24, p=0.017) and anti-ß2GPI antibodies (HR 1.64, p=0.053) were the two selected independent predictors. A multivariate model considering interactions between type of presenting event and anti-ß2GPI response had the best fit to the data. The two independent predictors in this model were presentation with both haemolytic anaemia and anti-ß2GPI antibodies (HR 2.70, p=0.018) and presentation with abortions, with absence of anti-ß2GPI (HR 0.37, p=0.028).
The figure shows Kaplan-Meier analyses, excluding patients who presented with two events. The mean (median) time to a second event was 0.93 (1.00) year for patients presenting with haemolytic anaemia and anti-ß2GPI antibodies, while it was 13.6 (17.0) years for patients presenting with recurrent abortions and no anti-ß2GPI response (Figure 1
).
|
Distribution and types of clinical events
Table 1 shows the types of the 87 presenting events and of the 153 additional serious events which were recorded during follow-up. Of the 153 serious events during follow-up, 132 occurred on no treatment, 12 on warfarin and nine on aspirin. The mean incidence per month was 0.021 (SD 0.03). The respective mean incidence rates on no treatment, aspirin, and warfarin were 0.064, 0.010 and 0.004. Thrombosis was more common than haemolytic anaemia in periods of no therapy (including the initial presentations) accounting for 34.5% (68/197) vs. 12.7% (25/197) of all serious events, respectively. The proportions were almost reversed during therapy with aspirin [22.2% (2/9) and 33.3% (3/9), respectively] or warfarin [16.7% (2/12) and 25% (3/12), respectively]. However, the odds ratios for the comparison between no therapy and therapy did not reach formal statistical significance for either thrombosis (OR 2.24, 95% 0.73–6.92, p=0.16) or haemolytic anaemia (OR 0.36, 95% 0.13–1.02, p=0.055).
|
Predictive modelling for the rate of subsequent serious clinical events
Warfarin and aspirin therapy were strong predictors of the rate of serious clinical events, with 91% and 78% reductions, respectively, in a multivariate model considering both treatment modalities and adjusting for individual patients. Results were similar for unadjusted analyses (Table 2). After adjusting for the effect of therapy and initial clinical presentation, the presence of anti-ß2GPI antibodies independently increased the rate of events by 69% (Table 2).
|
The initial clinical presentation also offered strong independent predictive information on the subsequent rate of events. Patients who presented with two clinical events had substantially higher event rates after adjusting for therapy and anti-ß2GPI specificity. Prognosis was also independently worse in patients presenting with autoimmune haemolytic anaemia and marginally better in patients with thrombocytopenia (Table 2
). Considering interactions of anti-ß2GPI antibodies and initial clinical manifestations did not improve the model fit substantially.
Clustering of similar clinical events
For all types of presenting events, the odds were high that subsequent events during follow-up would be of the same type as the presenting event(s). We excluded kidney involvement and pulmonary hypertension from this analysis, because they were not considered as potential initial events. Fifteen of 22 subsequent events in patients initially presenting with venous thromboses were also venous thromboses, as compared with only 41/113 subsequent events (excluding kidney involvement and pulmonary hypertension) in patients initially presenting with other symptoms (OR 3.76, 95% CI 1.42–9.98). The respective ORs for abortions (15/49 vs. 0/86), central nervous system manifestations (14/20 vs. 29/115), haemolytic anaemia (9/22 vs. 10/113), and arterial thromboses (1/4 vs. 7/131) were 77.7 (95% CI 4.52–1335), 6.92 (95% CI 2.43–19.7), 7.13 (95% CI 2.45–20.8), and 5.90 (95% CI, 0.54–64.3).
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
We have identified predictive factors affecting the prognosis of patients who present with clinical events suggestive of APS and laboratory support for this diagnosis. The type of initial clinical presentation affected both the rate and type of subsequent clinical events. In the case of abortions, haemolytic anaemia, thromboses and central nervous system manifestations, the odds were high for subsequent events to be the same as the initial manifestation. A first presentation with two clinical events or haemolytic anaemia carried a worse overall prognosis. The prognosis was also worse in the presence of anti-ß2GPI antibodies. The type of initial clinical presentation and anti-ß2GPI response affected also the time to the occurrence of a second clinical event. Overall prognosis was strongly affected by therapy and, interestingly, not only warfarin, but also aspirin seemed to have beneficial effects.
Our cohort is one of the largest in sample size and length of follow-up in the literature. The retrospective nature of the study poses some limitations. However, we used stringent definitions so as to avoid typical biases of retrospective studies such as recall bias and information bias. We selected as serious outcomes only major clinical events that would certainly have drawn clinical attention. Moreover, we posed stringent laboratory criteria for titres of anticardiolipin antibodies (at least 5 SD above the mean of normals) required to enter the study. In this way, we avoided patients with low or even negative titres on repeated measurement and questionable diagnoses.18,26
Our models allow the separation of patient groups with markedly diverse clinical outcomes. Patients presenting with haemolytic anaemia and who have anti-ß2GPI antibodies are expected to develop a second clinical manifestation in about 1 year, on average. On the other hand patients presenting simply with recurrent abortions and without anti-ß2GPI antibodies have a mean of over 13 years before developing another type of clinical manifestation. It should be noted that with the exception of one patient, all patients who were initially diagnosed with a first event suggestive of APS on the basis of a recurrent abortion did have another event during the prolonged follow-up of the study cohort. Therefore, it is unlikely that this group of patients was given a spurious probable APS diagnosis based on our diagnostic criteria. It is far more likely that patients who present simply with recurrent abortions (and especially if they lack an anti-ß2GPI response) represent a subgroup of patients who take very long to develop other types of APS-related manifestations. Given the relatively small sample size of the subgroup, the finding should be verified in other cohorts. The use of therapy further diversifies the risk of different patients and may also change the profile of occurring clinical events. Such information may be used in the future for assessing the risk of patients with a presentation which is suggestive, but not necessarily definitive of APS diagnosis and for making decisions on the urgency of intervening therapeutically at the time of initial presentation.
Our study further documents the importance of the anti-ß2GPI response in determining patient outcome. A limitation of our study is the fact that auto-antibodies were often measured many years after the first clinical event. There are insufficient data on the time of appearance and longitudinal persistence of these auto-antibodies. Nevertheless, the strength of the evidence that anti-ß2GPI antibodies may define a group of patients with worse clinical outcomes, especially in conjunction with certain initial first clinical manifestations. The anti-ß2GPI response may be more strongly related to APS than anticardiolipin antibodies10,11,20,21,27 and has also shown stronger relationships with the genetic HLA-related background of the syndrome.28
We also observed a clustering of serious event types in individual patients. Previous investigators have noticed that venous thromboses tend to be followed by more venous thromboses, and arterial thromboses tend to be followed by more arterial thromboses.29 A similar clustering has been described for abortions.30,31 In our cohort, we observed clustering of subsequent events similar with the initial event type not only for abortions and thromboses, but for all serious events considered, including autoimmune haemolytic anaemia and central nervous system manifestations.
The recent preliminary classification criteria put forth by an international panel suggested that thromboses and pregnancy morbidity are the sole clinical criteria that can carry the weight of a clinical diagnosis, and it was agreed that further evidence is required on the importance and associations of other clinical manifestations of APS.16 Our study has examined the predictive importance of some of these manifestations. On the basis of our data, APS may be conceptualized as a cluster of sub-syndromes, each having a specific type of pathophysiology which tends to predominate among its clinical manifestations. Strong clustering of events and diversified, yet predictable, evolution, are important characteristics of these syndromes.
|
Acknowledgments |
|---|
The authors thank the Hellenic Rheumatology Society and the Greek Secretariat for Research and Technology (Grant EPET 2728) for financial support and Mrs P. Papadopoulou for excellent secretarial assistance.
|
Notes |
|---|
Address correspondence to Dr J.P.A. Ioannidis, Chairman, Department of Hygiene and Epidemiology, University of Ioannina School of Medicine, Ioannina 45110, Greece. e-mail: jioannid{at}cc.uoi.gr
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Hughes GRV. The antiphospholipid syndrome: ten years on. Lancet1993; 342:341–4.[Web of Science][Medline]
2. Asherson RA. The catastrophic antiphospholipid syndrome. J Rheumatol1992; 19:508–12.[Web of Science][Medline]
3. McNeil HP, Chesterman CN, Krilis SA. Immunology and clinical importance of antiphospholipid antibodies. Adv Immunol1991; 49:193–280.[Web of Science][Medline]
4. Loizou S, McCrea JD, Rudge AC, Reynolds R, Boyle CC, Harris EN. Measurement of anti-cardiolipin antibodies by an enzyme-linked immunosorbent assay (ELISA): Standardization and quantitation of results. Clin Exp Immunol1985; 62:738–45.[Web of Science][Medline]
5.
McNeil HP, Simpson RJ, Chesterman CN, Krilis SA. Antiphospholipid antibodies are directed against a complex antigen that includes a lipid binding inhibitor of coagulation: ß2-glycoprotein I (apolipoprotein H). Proc Natl Acad Sci USA1990; 87:4120–4.
6. Matsuura E, Igarashi Y, Fujimoto M, Ichikawa K, Suzuki T, Sumida T, Yasuda T, Koike T. Heterogeneity of anticardiolipin antibodies defined by anticardiolipin antibodies defined by anticardiolipin cofactor. J Immunol1992; 148:3885–91.[Abstract]
7.
Matsuura E, Igarashi Y, Yasuda T, Triplett DA, Koike T. Anticardiolipin antibodies recognize ß2-glycoprotein I structure altered by interacting with an oxygen modified solid phase surface. J Exp Med1994; 179:457–62.
8. Roubey RA, Eisenberg RA, Harper MF, Winfield JB. Anticardiolipin autoantibodies recognize ß2-glycoprotein I in the absence of phospholipid. Importance of Ag density and bivalent binding. J Immunol1995; 154:954–60.[Abstract]
9. Viard JP, Amoura Z, Bach JE. Association of anti ß2-glycoprotein I antibodies with lupus-type circulating anticoagulant and thrombosis in systemic lupus erythematosus. Am J Med1992; 93:181–6.[Web of Science][Medline]
10. Vlachoyiannopoulos PG, Petrovas C, Tektonidou M, Krilis S, Moutsopoulos HM. Antibodies to ß2-Glycoprotein-I: urea resistance, binding specificity and association with thrombosis. J Clin Immun1998; 18:380–91.[Web of Science][Medline]
11. Tsutsumi A, Matsuura E, Ichikawa K, Fujisaku A, Mukai M, Kobayashi S, Koike T. Antibodies to ß2-glycoprotein I and clinical manifestations in patients with systemic lupus erythematosus. Arth Rheum1996; 39:1466–74.[Web of Science][Medline]
12. Rosove MH, Brewer MC. Antiphospholipid thrombosis: Clinical course after the first thrombotic event in 70 patients. Ann Intern Med1992; 117:303–8.
13.
Asherson RA, Baguley E, Pal C, Hughes GRV. Antiphospholipid syndrome: five year follow-up. Ann Rheum Dis1991; 50:805–10
14.
Khamashta MA, Guadrado MJ, Mujic F, Taub NA, Hunt BJ, Hughes GRV.The management of thrombosis in the antiphospholipid-antibody syndrome. N Engl J Med1995; 332:993–7.
15. Derksen RH, de Groot PG, Kater L, Nieuwenhuis HK.Patients with antiphospholipid and venous thrombosis should receive long term anticoagulant treatment. Ann Rheum Dis1993; 32:689–92.
16. Wilson WA, Gharavi AE, Koike T, Locksin MD, Branch DW, Piette JC, Brey R, Derksen R, Harris EN, Hughes GR, Triplett DA, Khamashta MA. International consensus statement on preliminary classification criteria for definite antiphospholipid syndrome: report of an international workshop. Arthr Rheum1999; 42:1309–11.[Web of Science][Medline]
17. Alarcon-Segovia D, Sanchez-Guerrero J.Primary antiphospholipid syndrome. J Rheumatol1989; 16:482–8.[Web of Science][Medline]
18. Alarcon-Segovia D, Perez-Vazquez ME, Villa AR, Drenkard C, Cabiedes J. Preliminary classification criteria for the antiphospholipid syndrome within systemic lupus erythematosus. Semin Arthritis Rheum1992; 21:275–86.[Web of Science][Medline]
19. Manoussakis MN, Gharavi AE, Drosos AA, Kitridou RC, Moutsopoulos HM. Anticardiolipin antibodies in unselected autoimmune rheumatic disease patients. Clin Immunol Immunopathol1987; 44:297–307.[Web of Science][Medline]
20. Cabiedes J, Cabral AR, Alarcon- Segovia D. Clinical manifestations of the antiphospholipid syndrome in patients with systemic lupus erythematosus associate more strongly with anti-ß2Glycoprotein-I than with antiphospholipid antibodies. J Rheumatol1995; 22:1899–906.[Web of Science][Medline]
21. Martinuzzo MF, Forastiero RR, Carreras LO. Antiß2-glycoprotein I antibodies: determination and association with thrombosis. Br J Haematol1995; 89:397–402.[Web of Science][Medline]
22. Exner T, Triplett DA, Taberner D, Machin SJ. Guidelines for testing and revised criteria for lupus anticoagulants. SSC Subcommitee for the standardization of lupus anticoagulants. Thromb Haemost1991; 65:320–2.[Web of Science][Medline]
23. Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, Rothfield NF, Schaller JG, Talal N, Winchester RJ. The 1982 revised criteria for the classification of lupus erythematosus. Arthritis Rheum1982; 25:1271–7.[Web of Science][Medline]
24. Koch GG, Atkinson SS, Stokes ME. Poisson regression. In: S Kotz, NL Johnson and CB Read, eds. Encyclopedia of Statistical Sciences, vol. 7. New York, John Wiley and Sons, 1986.
25. SPSS, Inc. SPSS. Chicago, 1994.
26. Piette JC. 1996 diagnostic and classification criteria for the antiphospholipid/cofactors syndrome: a ‘mission impossible’? Lupus1996; 5:354–63.[Web of Science][Medline]
27.
Wahl DG, De Maistre E, Guillemin F, Regnault V, Perret-Guillaume C, Lecompte T. Antibodies against phospholipids and beta2-glycoprotein I increase the risk of recurrent venous thromboembolism in patients without systemic lupus erythematosus. Q J Med1998; 91:125–30.
28. Ioannidis JPA, Tektonidou MG, Vlachoyiannopoulos PG, Stavropoulos-Giokas C, Spyropoulou-Vlachou M, Reveille JD, Arnett FC, Moutsopoulos HM. HLA associations of anti-beta2 glycoprotein I response in a Greek cohort with antiphospholipid syndrome and meta-analysis of four ethnic groups. Hum Immunol1999; 60:1274–80.[Web of Science][Medline]
29. Finazzi G, Brancaccio V, Moia M, Ciaverella N, Mazzucconi MG, Schinco PC, Ruggeri M, Pogliani EM, Gamba E, Rossi E, Baudo F, Manotti C, D'Angelo A, Polareti G, De Stefano V, Berrettini M, Barbui T. Natural history and risk factors for thrombosis in 360 patients with antiphospholipid antibodies: a four year prospective study from the Italian registry. Am J Med1996; 100:530–6.[Web of Science][Medline]
30.
Ramsey-Goldman R, Kutzer LH, Kuller LH, Guzick D, Carpenter AB, Medsger TA Jr. Pregnancy outcome and anticardiolipin antibody in women with systemic lupus erythematosus. Am J Epidemiol1993; 138:1057–69.
31.
Shah NM, Khamashta MA, Atsumi T, Hughes GRV. Outcome of patients with anticardiolipin antibodies: a 10 year follow-up of 52 patients. Lupus1998; 7:3–6.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  I Krause, L Leibovici, M Blank, and Y Shoenfeld Clusters of disease manifestations in patients with antiphospholipid syndrome demonstrated by factor analysis Lupus, March 1, 2007; 16(3): 176 - 180. [Abstract] [PDF]
|
|
|
|
 |
|
 G. Ruiz-Irastorza, M. A. Khamashta, B. J. Hunt, A. Escudero, M. J. Cuadrado, and G. R. V. Hughes Bleeding and Recurrent Thrombosis in Definite Antiphospholipid Syndrome: Analysis of a Series of 66 Patients Treated With Oral Anticoagulation to a Target International Normalized Ratio of 3.5 Arch Intern Med, May 27, 2002; 162(10): 1164 - 1169. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||