Coexistence of Anti-Phospholipid Antibodies and Endothelial Perturbation in Systemic Lupus Erythematosus Patients With Ongoing Prothrombotic State
Background Anti-phospholipid antibodies (aPLs) were associated with an ongoing prothrombotic state in patients with systemic lupus erythematosus (SLE). Because aPLs are able to shift endothelial function toward procoagulant activity in vitro, we investigated the relationship among aPLs, ongoing prothrombotic state, and endothelial perturbation in SLE patients.
Methods and Results We measured aPLs, anti-EC antibodies, circulating levels of prothrombin fragment F1+2 (F1+2), tumor necrosis factor-α (TNF-α), tissue-type plasminogen activator (TPA), and von Willebrand factor (vWF) in 43 SLE patients and 25 healthy subjects. Patients positive for aPLs (n=23) had a higher prevalence of anti-EC antibodies (P=.02) and higher levels of F1+2 (P=.003) than aPL(−) patients. Endothelial perturbation, defined by elevated plasma levels of both TPA and vWF, was significantly associated with aPL positivity (P=.001). F1+2 >1 nmol/L (mean+2 SD of controls) was detected in all but one patient in whom aPL positivity and endothelial perturbation coexisted and in no aPL(+) patient without endothelial perturbation (P=.0039). F1+2 was significantly correlated with vWF (ρ=.6, P=.004) and TPA (114 =.70, P=.0006) only in aPL(+) patients. Endothelial perturbation was closely associated with high values of TNF-α (P=.0001), anti-phospholipid (P=.001), and anti-EC antibodies (P=.012). In 31 patients without a clinical history of thrombosis followed up for 3 years, aPL(+) patients with endothelial perturbation showed higher F1+2 and TNF-α values than aPL(+) patients without endothelial dysfunction.
Conclusions This study shows that in SLE patients, aPL positivity is associated with an ongoing prothrombotic state only in the presence of endothelial perturbation. Our findings also suggest that aPLs and TNF-α might cooperate in inducing endothelial perturbation.
Anti-phospholipid antibodies are frequently associated with venous and arterial thrombosis in patients with and without autoimmune diseases,1 2 3 suggesting that aPLs might predispose to thrombosis. Two in vivo studies4 5 reinforced this hypothesis, showing that patients with SLE and aPLs have an ongoing prothrombotic state, as demonstrated by high circulating levels of F1+2, a marker of thrombin generation.
Even if many studies sought to elucidate the mechanism accounting for this association, it is still unclear whether and how aPLs predispose to thrombosis.
ECs were indicated to be an important target of aPLs,6 7 but it has not been clearly established whether the aPLs per se bind to endothelial phospholipids or phospholipid-binding plasma proteins, such as β2-glycoprotein I.8 9 10 Anti-EC antibodies were demonstrated in 65% of SLE patients and correlated to the presence of LA,11 whereas in non-SLE patients, 62% of aPL(+) sera showed an increased binding of IgG to endothelium.9 Even if immunoglobulin fractions from SLE sera cross-reacted with ECs and cardiolipin,12 the anti-EC activity was not diminished by adsorption of aCLs by liposomes,9 thus suggesting that anti-EC antibodies and aPLs represent two distinct families of antibodies, which can frequently be associated. Furthermore, no association between anti-EC antibodies and thrombosis was observed in SLE patients.8 11
Conversely, a rich and somewhat conflicting literature suggests that aPLs interact with endothelium, leading to a complex of functional changes, including interference with the regulatory protein C/thrombomodulin and protein S pathways,13 14 15 prostacyclin synthesis,16 fibrinolysis activation,17 18 and inhibited formation of antithrombin-III–thrombin complexes.19 On the other hand, Oosting et al20 showed that aPLs were able to enhance EC procoagulant activity in the extracellular matrix only in the presence of a suboptimal dose of TNF. Indirect support for the suggestion of an interaction between aPLs and endothelium also comes from an in vivo study that demonstrated a significant association between aPLs and circulating levels of thrombomodulin, a marker of endothelial perturbation.21
Taken together, these findings may lead to the hypothesis that in aPL patients, clotting activation could occur through an interaction between aPLs and ECs. This hypothesis has been tested in vivo in SLE patients followed up for 3 years.
Between May 1991 and September 1993, we studied 43 patients (38 women and 5 men 21 to 59 years old) diagnosed as having SLE in accordance with the criteria of the American College of Rheumatology (formerly the American Rheumatism Association)22 and 25 healthy subjects (20 women and 5 men 23 to 60 years old) as control subjects. The SLE patients, who were referred to our clinic as outpatients, were included in the study if they were not being treated with anticoagulants. In SLE patients, the duration of disease averaged 7±4 years (range, 1 to 15 years). Thirty-three patients were being treated with corticosteroids (prednisone 5 to 25 mg/d or methylprednisolone 4 to 24 mg/d) and/or methotrexate (0.25 to 0.30 mg/kg IV once a week). Fourteen patients were considered hypertensive, having blood pressure values >140/90 mm Hg on at least two different occasions; 12 of these were being treated with diuretics, ACE inhibitors, or calcium antagonists. Five patients had diabetes mellitus, 3 of whom were being treated with insulin. Five patients showed clinical signs of vasculitis.23
Twelve patients (28%) (9 women and 3 men 21 to 52 years old) had a history of previous arterial and/or venous thrombosis and/or fetal loss in the previous 6 to 22 months: 6 DVT, 1 thromboembolic stroke, 2 DVT and thromboembolic stroke, 2 recurrent fetal loss (two or more prior unexplained first trimester miscarriages; two or more unexplained fetal deaths occurring after 12 weeks of gestation, with prior documentation of viability), and 1 recurrent fetal loss and DVT. DVT was confirmed by venous Doppler ultrasound and thromboembolic stroke by CT scan.
Eleven (92%) of 12 patients with thrombosis were aPL(+). At the time of the study, all patients with previous thrombotic events were treated with antiplatelet agents (aspirin 325 mg/d).
The patients were considered to be in a severely active, moderately active, or inactive phase of the disease on the basis of the occurrence of three, one or two, or none of the following in the 3 months preceding the study: arthralgia, pleuritis, pericarditis, vasculitis, myalgia, or renal or nervous system involvement.24 Thirty-one patients were considered to be in the active phase (11 severe, 20 moderate) of the disease.
Among laboratory indexes, we measured the serum levels of some proteins that are known to change during the acute phase of the disease, namely C3 and C4, measured by radial immunodiffusion25 ; C-reactive protein, measured by an agglutination test26 ; and clottable fibrinogen, evaluated by the Clauss method (Ortho Thrombin, Ortho Diagnostics; reference value, 140 to 400 mg/dL).27 No patient had had active infections, surgery, or trauma in the previous 3 months.
Baseline laboratory measurements. A cross-sectional comparison of vWF:Ag, TPA:Ag, F1+2, TNF-α, and aPLs and anti-EC antibodies was performed in both SLE patients and control subjects. We considered patients with signs of endothelial perturbation to be those with plasma values of both vWF:Ag and TPA:Ag greater than the mean+2 SD of control values.
Prospective study. The prospective study included 31 SLE patients (29 women and 2 men 21 to 59 years old) without a history of thrombosis, whereas the 12 patients with previous thrombotic events were given oral anticoagulants (warfarin; international normalized ratio, 2.5 to 3.5) and excluded from the follow-up.
Patients were examined every 12 months for a follow-up of 3 years (median [range], 36 [18 to 36] months; mean±SD, 33±6 months). Doppler ultrasound was performed at each follow-up contact.
Any thrombotic episode during follow-up was checked by clinical and laboratory methods. In case of occurrence of DVT, the diagnosis had to be confirmed by venography or ultrasonography, and the diagnosis of thrombosis of intracerebral vessels by CT scanning or angiography.
On each visit, vWF:Ag, TPA:Ag, F1+2, TNF-α, anti-EC antibodies, aPLs, and disease activity were investigated. Patients showing high circulating levels (mean+2 SD of control values) of both vWF:Ag and TPA:Ag at each follow-up were considered to have persistent circulating signs of endothelial perturbation.
Concomitant treatment throughout the follow-up. Twenty-four patients were treated with corticosteroids (prednisone 5 to 50 mg/d or methylprednisolone 4 to 24 mg/d). Among these subjects, 11 were also treated with methotrexate (0.25 to 0.30 mg/kg IV once a week) for 12 to 36 weeks.
After an overnight fast and rest for at least 10 minutes, a blood sample was taken from each patient and mixed in tubes containing 3.8% trisodium citrate (Becton Dickinson Vacutainer; ratio, 9:1). The samples were centrifuged for 10 minutes at 2000g at room temperature. The supernatant was used immediately for the evaluation of fibrinogen and LA and in part stored at −80°C for measurement of TPA:Ag and fragment F1+2. Another blood sample was taken to measure aCLs, anti-EC antibodies, C reactive protein, complement components C3 and C4, and TNF-α. Sera were immediately frozen at −80°C until assay.
vWF:Ag was evaluated by a “sandwich ELISA” system (Imubind vWF ELISA, American Diagnostica) as previously described.28 Intra-assay and interassay coefficients of variation were 9% and 13%, respectively. In 25 healthy control subjects, plasma values of vWF:Ag were 991±291 U/L.
TPA:Ag was measured with an ELISA (Imubind-5-TPA, American Diagnostica).17 Intra-assay and interassay coefficients of variation were 7% and 9%, respectively. In 25 healthy subjects, plasma values of TPA:Ag were 5.2±1.8 ng/mL.
F1+2 was assayed by an ELISA based on the sandwich principle (Enzygnost F1+2, Behringwerke).29 Intra-assay and interassay coefficients of variation were 8% and 9%, respectively. In 25 healthy subjects, plasma values of F1+2 were 0.6±0.2 nmol/L.
LA was performed in platelet-poor plasma centrifuged twice at 5000g with four different coagulation tests, as previously described.30 Patients were considered LA(+) if they had at least two abnormal (prolonged) clotting tests that returned to normal values by the addition of 0.05 mmol/L phosphatidylcholine-phosphatidylserine liposomes (confirmatory test).31
For aCL evaluation, an ELISA validated in an international workshop32 was used. IgG or IgM aCLs were considered positive when the activity was >10 GPL or 10 MPL units, respectively.33 Patients were considered aPL(+) if LA and/or aCLs were detected on two separate occasions at least 2 months apart.4 5
During the follow-up, patients were considered to be persistently aPL(+) (LA and/or aCL) if positivity persisted at each follow-up control.
EC Cultures and Anti-EC Antibodies
Human umbilical vein ECs were isolated from normal-term umbilical cord veins by collagenase perfusion as previously described.8 First-passage cultures were used.
Anti-EC antibodies were detected by use of a cell-surface ELISA assay on living cells as previously reported by Del Papa et al.34 The results of the tested samples were expressed as a percentage of a positive reference serum. Values >2 SD of the mean of 40 healthy subjects were considered positive, namely, 50.1% for IgG anti-EC antibodies.
During the follow-up, patients were considered to be persistently positive for anti-EC antibodies if positivity persisted at each follow-up check.
Serum TNF-α was assayed in duplicate by an enzyme immunoassay (Biokine Tumor Necrosis Factor Alpha test kit, T Cell Diagnostics Inc).35 The detection limit was calculated to be 10 pg/mL. Among 20 healthy subjects (17 women and 3 men 23 to 58 years old), 3 showed detectable TNF-α serum levels (median, <10 pg/mL; range, <10 to 34 pg/mL). Intra-assay and interassay coefficients of variation were 8% and 9%, respectively.
Statistical analysis was performed by χ2 statistics or Fisher’s exact test (if n≤5) for independence and by unpaired t test. The linear regression test was used to study the different correlations. When necessary, log transformation was used to normalize the data, or appropriate nonparametric tests were used.
Data are presented as mean±SD and 95% confidence limits. Median and range are given for TNF-α, because it shows appreciably skewed distribution. Only values of P<.05 were regarded as statistically significant. All calculations were made with the computer program STAT-View II (Abacus Concepts).36
Clinical and laboratory features of SLE patients are reported in Table 1⇓. Compared with control subjects, SLE patients had significantly higher circulating levels of vWF:Ag (2437±1411 versus 991±291 U/L, P=.0001) and TPA:Ag (12.2±8.3 versus 5.2±1.8 ng/mL, P=.0001). A significant direct correlation (r=.83, P=.0001) was observed between these two parameters. Moreover, SLE patients had significantly higher F1+2 plasma levels than control subjects (1.03±0.66 versus 0.63±0.23 nmol/L, P=.0102).
Relationship Among aPLs, Thrombin Generation, and Endothelial Perturbation
Among SLE patients, 23 (53%) showed aPL positivity: they had a higher prevalence of anti-EC antibody positivity (P=.02) and higher vWF:Ag (P=.004), TPA:Ag (P=.001), and F1+2 (P=.003) plasma levels (Table 2⇓). These differences persisted when patients with previous thrombotic events were excluded (data not shown).
Seventy percent of aPL(+) subjects had signs of endothelial perturbation (see “Methods”), compared with 15% of aPL(−) patients (P=.001) (Table 2⇑). Sixty-five percent of aPL(+) subjects had F1+2 plasma levels >1 nmol/L (mean+2 SD of control values), compared with 5% of aPL(−) patients (P=.0002) (Table 2⇑).
aPL(+) patients with endothelial dysfunction (n=16) had significantly higher F1+2 plasma levels than aPL(+) subjects without endothelial perturbation (n=7) (1.7±0.5 versus 0.5±3 nmol/L; P=.0003) (Fig 1⇓). In particular, all but one aPL(+) patient with endothelial perturbation showed elevated F1+2 plasma levels, whereas all aPL(+) patients without endothelial perturbation showed normal F1+2 values (94% versus 0%, P=.0039) (Fig 1⇓). Finally, F1+2 circulating levels were significantly correlated with vWF:Ag (ρ=.6, P=.004) and TPA:Ag (ρ=.70, P=.0006) in aPL(+) (Fig 2⇓) but not in aPL(−) SLE patients (ρ=.20, P=.51 and ρ=.40, P=.063 for vWF:Ag and TPA:Ag, respectively).
Clinical and Laboratory Characteristics of Patients With Endothelial Perturbation
In the whole SLE series, 19 patients (44%) were considered to have signs of endothelial perturbation (Table 3⇓); they did not show differences in acute-phase reactant proteins compared with SLE patients without endothelial perturbation (not shown). All patients with vasculitis had signs of endothelial perturbation.
Disease duration, hypertension, diabetes mellitus, and serum cholesterol did not distinguish patients with and without endothelial dysfunction. Conversely, SLE patients with endothelial perturbation had more severe disease activity (P=.019), higher prevalence of aPLs (P=.001), and anti-EC antibody positivity (P=.012) (Table 3⇑). Corticosteroid daily dosage was significantly higher in SLE patients with endothelial perturbation (P=.0093) (Table 3⇑).
TNF-α was significantly higher in SLE patients than in control subjects (median [range], 91.4 [<10 to 311] versus <10 [<10 to 34] pg/mL; P=.0001). Forty SLE patients (93%) and 3 healthy subjects (16%) showed TNF-α serum levels >10 pg/mL.
Patients with endothelial perturbation had significantly higher TNF-α values than patients without endothelial perturbation (P=.0001) (Table 3⇑). aPL(+) patients with endothelial perturbation (n=16) had significantly higher TNF-α values than aPL(+) patients without endothelial perturbation (n=7) (median [range], 140.8 [70.1 to 310.7] versus 14.9 [<10 to 81.6] pg/mL, P=.0003) (Fig 3⇓).
Serum values of TNF-α were closely related to disease activity. Indeed, patients with severe disease had TNF-α values (median [range], 199.2 [91 to 301] pg/mL) higher than patients with moderate (median [range], 93.1 [40 to 311] pg/mL, P=.0005) or inactive (median [range], 13 [<10 to 140] pg/mL, P=.0001) disease. TNF-α was significantly correlated with vWF:Ag (ρ=.65, P=.0001) and TPA:Ag (ρ=.64, P=.0001).
During a 3-year follow-up (mean±SD [range], 33±6 [18 to 36] months), all patients aPL(+) at baseline (n=12) showed persistent aPL positivity; conversely, 3 of 19 SLE patients aPL(−) showed transient aPL positivity. Patients with persistent aPL positivity had higher prevalence of persistent anti-EC antibody positivity than transiently aPL(+) and aPL(−) patients (42% versus 0% versus 5%, respectively, P=.043). Among 12 patients with persistent aPL positivity, 7 (58%) showed persistent signs of endothelial perturbation; at baseline and throughout the follow-up, they had higher values of F1+2 (P=.002) and TNF-α (P=.04) than aPL(+) and aPL(−) patients without endothelial perturbation (Fig 4A⇓ and 4B⇓).
Among 7 patients with persistent aPL positivity and endothelial dysfunction, 4 (57%) had a thrombotic event (2 DVT, 2 stroke) after 18 to 29 (mean±SD, 24±5) months of follow-up, whereas no thrombotic event was observed in aPL(+) and aPL(−) patients without endothelial perturbation (odds ratio, 3.43; 95% confidence limits, 1.6 to 7.5; P=.026).
An ongoing prothrombotic state has been detected in aPL(+) SLE patients,4 5 but the underlying mechanism is still unclear. Several lines of experimental evidence point to a functional interaction between aPLs and ECs in which, at least in vitro, aPLs induce a shift toward procoagulant activity.13 14 15 16 17 18 19 20 37 38 39 We tested this hypothesis in the SLE population in whom endothelial perturbation has been assessed by measurement of the plasma levels of two parameters, TPA and vWF, both synthesized and secreted by endothelium via constitutive and induced pathways.40 41 The definition of endothelial perturbation we used could have underestimated the rate of endothelial dysfunction, but it should be more specific, taking into account, for instance, that vWF circulating levels may also result from platelet activation.42 43
Our data show that 70% of aPL(+) SLE patients have signs of endothelial dysfunction. However, the most important finding of this study is a close association among aPL positivity, endothelial perturbation, and increased rate of thrombin generation. In fact, 94% of aPL(+) subjects with endothelial perturbation showed elevated F1+2 plasma levels, whereas among aPL(+) patients without endothelial perturbation, none had elevated F1+2 values, indicating that aPL positivity per se would not play a role in inducing clotting activation. This finding was corroborated by the significant correlation between F1+2, vWF, and TPA observed in aPL(+) patients and seems to suggest that the coexistence of endothelial dysfunction is crucial to observe clotting activation in aPL(+) patients.
The aPL positivity does not seem to be an essential condition to observe endothelial perturbation. Thus, among 23 aPL(+) patients, 7 had no endothelial perturbation. Also, despite the significant association between anti-EC antibodies and endothelial perturbation, nearly 50% of patients with endothelial dysfunction had no anti-EC antibodies, suggesting that other mechanism(s) would contribute to endothelial perturbation.
Interestingly, 94% of patients with high values of vWF:Ag and TPA:Ag had active disease, suggesting a possible relationship between inflammation and endothelial activation.
Among the mechanisms linking immune-mediated inflammatory response and endothelial perturbation in SLE, in vitro experiments focused on the possible role of cytokines.44 In particular, TNF-α may play an important role, in that it is cytotoxic for endothelium in vitro and induces in vitro and in vivo vWF release from ECs.45 46 In vitro studies also showed that TNF-α is able to induce procoagulant activity in cultured ECs,47 48 suggesting that TNF-α could mediate activation of the clotting system by enhancing endothelial tissue factor expression.
A previous study showed that TNF-α was elevated in SLE patients with severe disease,49 but the relationship between TNF-α and endothelial perturbation was not investigated. Our study demonstrates that endothelial perturbation is prevalent in patients with active disease and that TNF-α is likely to play an important role. The very close correlation of TNF-α with vWF:Ag and TPA:Ag supports this hypothesis.
It is noteworthy that all but two aPL(+) patients with endothelial perturbation also showed high TNF-α serum levels, whereas all aPL(+) patients without endothelial perturbation had normal TNF-α values. These data lead us to hypothesize that the ongoing prothrombotic state in the SLE population might result from the coexistence of aPL positivity and elevated TNF-α, which would concur to induce endothelial perturbation and eventually clotting activation. This hypothesis is supported by in vitro studies in which the coincubation of aPL(+) IgG and TNF-α was able to enhance the expression of endothelial tissue factor.20 50
The prospective study reinforced the suggestion that aPL positivity and endothelial perturbation coexist in SLE patients with an ongoing prothrombotic state. Thus, all patients with persistent aPL positivity and endothelial perturbation showed elevated F1+2 plasma levels during the follow-up. Patients with these laboratory characteristics also had elevated TNF-α values, further indicating that inflammation may play an important role in favoring endothelial dysfunction and clotting activation.
In conclusion, this study demonstrates that aPL positivity per se is not associated with an ongoing prothrombotic state, which, on the contrary, is present only in case of coexistence with endothelial dysfunction. The presence of high TNF-α serum levels in aPL(+) patients with endothelial perturbation also suggests that inflammation may be crucial for the development of endothelial dysfunction and eventually clotting activation. However, because this study is observational, a cause-and-effect relationship between endothelial dysfunction and an ongoing prothrombotic state cannot be firmly established, and further study is necessary to support this hypothesis. The higher incidence of thrombosis in aPL(+) patients with endothelial dysfunction compared with patients without endothelial perturbation is potentially of clinical relevance, because it suggests that within aPL(+) patients, there are subgroups with different risks of thrombosis. These data need to be confirmed in a larger cohort of the SLE population.
Selected Abbreviations and Acronyms
|DVT||=||deep venous thrombosis|
|F1+2||=||human prothrombin fragment F1+2|
|SLE||=||systemic lupus erythematosus|
|TNF-α||=||tumor necrosis factor-α|
|TPA||=||tissue plasminogen activator|
|vWF||=||von Willebrand factor|
We wish to thank Drs Maria Antonietta Grandilli, Mario Vieri, and Doloretta Cara for their fruitful collaboration.
- Received August 5, 1996.
- Revision received October 18, 1996.
- Accepted November 14, 1996.
- Copyright © 1997 by American Heart Association
Love PE, Santoro SA. Antiphospholipid antibodies: anticardiolipin and the lupus anticoagulant in systemic lupus erythematosus and in non-SLE disorders: prevalence and clinical significance. Ann Intern Med. 1990;112:682-698.
Ginsberg JS, Demers C, Brill-Edwards P, Johnston M, Bona R, Burrows RF, Weitzx J, Denburg JA. Increased thrombin generation and activity in patients with systemic lupus erythematosus and anticardiolipin antibodies: evidence for a prothrombotic state. Blood. 1993;81:2958-2963.
Ferro D, Quintarelli C, Valesini G, Violi F. Lupus anticoagulant and increased thrombin generation in patients with systemic lupus erythematosus. Blood. 1994;83:304-305. Letter.
De Groot PG, Oosting JD, Derksen RHWM. Antiphospholipid antibodies: specificity and pathophysiology. Baillieres Clin Haematol. 1994;6:691-709.
Roubey RAS. Autoantibodies to phospholipid-binding plasma proteins: a new view of lupus anticoagulants and other ‘antiphospholipid’ antibodies. Blood. 1994;84:2854-2867.
Del Papa N, Guidali L, Spatola L, Bonara P, Borghi MO, Tincani A, Balestrieri G, Meroni PL. Relationship between anti-phospholipid and antiendothelial cell antibodies, III: beta-2-glycoprotein I mediates the antibody binding to endothelial membranes and induces the expression of adhesion molecules. Clin Exp Rheumatol. 1995;13:179-185.
Oosting JD, Derksen RHWM, Bobbink IWG, Hackeng TM, Bouma BN, De Groot PG. Antiphospholipid antibodies directed against a combination of phospholipid with prothrombin, protein C or protein S: an explanation for their pathologic mechanism? Blood. 1993;81:2618-2625.
Violi F, Ferro D, Quintarelli C, Saliola M, Grandilli MA, Balsano F. Tissue plasminogen activator inhibitor in patients with systemic lupus erythematosus. Br Med J. 1990;300:1099-1102.
Shibata S, Harpel PC, Gharavi A, Rand J, Fillit H. Autoantibodies to heparin from patients with antiphospholipid antibody syndrome inhibit formation of antithrombin III-thrombin complexes. Blood. 1994;83:2532-2540.
Karmochkine M, Boffa MC, Piette JC, Cacoub P, Wechsler B, Godeau P, Juhan L, Weiller PJ. Increase in plasma thrombomodulin in lupus erythematosus with antiphospholipid antibodies. Blood. 1992;79:837-838. Letter.
Gladman DD, Urowitz MB. Systemic lupus erythematosus: clinical features. In: Klippel JM, Dieppe PA, eds. Rheumatology. London, UK: Mosby; 1994:2.4.
Morrow WJW, Isenberg DA, Todd-Pokropek A, Parry HF, Snarth ML. Useful laboratory measurement in the management of systemic lupus erythematosus. Q J Med. 1982;202:125-138.
Singer JM, Pratz CM, Pader E, Elster SK. The latex fixation test: agglutination test for the C-reactive protein and comparison with the capillary precipitin method. Am J Clin Pathol. 1957;28:614.
Ferro D, Saliola M, Quintarelli C, Valesini G, Basili S, Grandilli MA, Bonavita MS, Violi F. Methods for detecting lupus anticoagulants and their relation to thrombosis and miscarriage in patients with systemic lupus erythematosus. J Clin Pathol. 1992;45:332-338.
Brandt JT, Triplett DA, Alving B, Scharrer I. Criteria for the diagnosis of lupus anticoagulants: an update on behalf of the subcommittee on lupus anticoagulant/antiphospholipid antibody of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis (ISTH). Thromb Haemost. 1995;74:1185-1190.
Harris EN, Gharavi AE, Patel SP, Hughes GRV. Evaluation of the anticardiolipin antibody test: report of an international workshop held 4 April 1986. Clin Exp Immunol. 1987;68:214-222.
Del Papa N, Meroni PL, Tincani A, Harris EN, Pierangeli SS, Barcellini W, Borghi MO, Balestrieri G, Zanussi C. Relationship between anti-phospholipid and anti-endothelial cell antibodies: further characterization of the reactivity on resting and cytokine-activated endothelial cells. Clin Exp Rheumatol. 1992;10:37-42.
Armitage P, Berry G. Statistical Methods in Medical Research. 2nd ed. Oxford, UK: Blackwell Scientific Publications; 1990:200-215.
Nilsson TK, Lofvenberg E. Decreased fibrinolytic capacity and increased von Willebrand factor levels as indicators of endothelial dysfunction in patients with the lupus anticoagulant. Clin Rheumatol. 1989;1:58-63.
Paleolog EM, Crossman DC, McVey J, Pearson JD. Differential regulation of constitutive and stimulated secretion of von Willebrand factor from endothelial cells. Blood. 1990;75:688-695.
Emeis JJ. Mechanisms involved in short-term changes in blood levels of t-PA. In: Kluft C, ed. Tissue-Type Plasminogen Activator: Physiological and Clinical Aspects. Boca Raton, Fla: CRC Press; 1988;2:21-35.
Sporn LA, Chavin SI, Marder VJ, Wagner DD. Biosynthesis of von Willebrand protein by human megakaryocytes. J Clin Invest. 1985;76:1102-1106.
Bevilacqua MP, Pober JS, Majeau GR, Fiers W, Cotran RS, Gimbrone MA Jr. Recombinant tumor necrosis factor induces procoagulant activity in cultured human vascular endothelium: characterization and comparison with the actions of interleukin-1. Proc Natl Acad Sci U S A. 1986;83:4533-4537.
Mulder AB, Hegge-Paping KSM, Magielse CPE, Blom NR, Smit JW, Van der Meer J, Halie MR, Bom VJJ. Tumor necrosis factor alpha-induced endothelial tissue factor is located on the cell surface rather than in the subendothelial matrix. Blood. 1994;84:1559-1566.