Do Antiphospholipid Antibodies Interfere with Tissue Factor Pathway Inhibitor?

Thrombosis Research 94 (1999) 213–220

REGULAR ARTICLE

Do Antiphospholipid Antibodies Interfere with Tissue Factor Pathway Inhibitor? Eva Marie Jacobsen, Per Morten Sandset and Finn Wisløff Haematological Research Lab., Ulleva˚l University Hospital, Oslo, Norway. (Received 30 July 1998 by Editor H.C. Godal; revised/accepted 10 November 1998)

Abstract This study was conducted to investigate whether antiphospholipid antibodies (APA) can interfere with the phospholipid-dependent inhibition of coagulation exerted by tissue factor pathway inhibitor (TFPI). Eleven patients with APA and eleven healthy controls matched for age and gender were enrolled. Blood samples were drawn before and 5 minutes after an intravenous injection of unfractionated heparin 5000 IE, which is known to cause TFPI release in healthy individuals. The preheparin samples showed significantly higher TFPI free antigen levels in the APA positive patients than in the controls (21.7 vs. 14.2 ng/ml, p50.03). TFPI activity as measured in a chromogenic substrate assay also was higher in patients, but this difference was not statistically significant (1.13 vs. 1.01 U/ml, p50.2). The TFPI levels showed a considerable rise in both patients and controls after heparin injection. In both assays, the postheparin levels were significantly higher in patients than in controls (TFPI antigen: 179 vs. 153 ng/ml, p50.05; TFPI activAbbreviations: LA, lupus anticoagulant; APA, antiphospholipid antibodies; TF, tissue factor; TFPI, tissue factor pathway inhibitor; dPT, modified diluted prothrombin time assay; APS, antiphospholipid antibody syndrome; SLE, systemic lupus erythematosus; ACA, anticardiolipin antibodies; UFH, unfractionated heparin; NP, normal plasma; rTF, recombinant TF; HDB, hexadimethrine bromide; TBF, Tris-buffered saline; LR, Lupus Ratio; DIC, disseminated intravascular coagulation. Corresponding author: E.M. Jacobsen, Haematological Research Lab., Medical Clinic, Ulleva˚l University Hospital, Kirkevn. 166, 0407 Oslo, Norway. Tel: 147 (22) 119 240; Fax: 147 (22) 601 627; E-mail: ,[email protected]..

ity: 3.26 vs. 2.51 U/ml, p50.03). A modified diluted prothrombin time assay (dPT) was used to measure TFPI anticoagulant activity. In this assay, samples from the patients with the strongest effect of lupus anticoagulants (LAs) on preheparin coagulation times showed little or no increase after heparin injection. This result may reflect an inhibition of TFPI anticoagulant activity by strong LAs. In conclusion, we have found that patients with APA have higher TFPI amidolytic activity/antigen level both before and after heparin stimulation of TFPI release. These observations do not explain the higher thrombotic risk in these patients but may reflect an upregulated tissue factor activity, which has been demonstrated in these patients. TFPI anticoagulant activity, however, as measured in a dPT assay, may be inhibited by strong LAs.  1999 Elsevier Science Ltd. All rights reserved. Key Words: Antiphospholipid antibodies; Lupus anticoagulant; Tissue factor pathway inhibitor

A

ntiphospholipid antibodies (APA) are a heterogeneous group of autoantibodies with specificity towards complexes of phospholipids and proteins [1]. APA have been shown to interfere with procoagulant reactions in the coagulation cascade and with at least one anticoagulant system, the protein C pathway [2]. The possible influence of APA on phospholipid-dependent reactions involving other coagulation inhibitors has not been explored fully yet, although an antiheparin effect that attenuates the anticoagulant effect of antithrombin has been reported [3].

0049-3848/99 $–see front matter  1999 Elsevier Science Ltd. All rights reserved. PII S0049-3848(98)00195-9

hypertension HELLP habitual abortions 1 intrauterine death preeclampsia habitual abortions 1 intrauterine death

SLE

Wegener’s granulomatosis Sjøgren’s syndrome SLE? SLE preeclampsia, hemorrhage

SLE

chorea minor, endocarditis

Complications to pregnancy

70

40

31 34

31 53

43

M

F

F F

F F

F

cerebral

20 28 F F

cerebral

coronary

63 24 F F

cerebral

proximal DVT

Venous thrombosis Arterial thrombosis

Table 1. 11 patients with antiphospholipid antibodies

1.1. Patients and Control Subjects

Age (years)

1. Materials and Methods

Eleven patients with antiphospholipid antibodies and an equal number of control individuals were enrolled. Each control was matched with respect to age and gender. Individuals with trombocytopenia, peptic ulcers, or any other contraindication to heparin injection were excluded. Likewise; pregnancy, mental illness, alcohol or drug abuse, or regular medication with anticoagulant drugs were exclusion criteria. Informed consent was obtained from all participants. The protocol was approved by The Regional Committee for Medical Research Ethics. The clinical information on the patients are presented in Table 1. Six patients met the criteria for the antiphospholipid antibody syndrome (APS) as defined by Harris [13]. Four of these patients had primary APS and arterial thrombosis had been the main clinical event for all of them. The remaining two patients had APS in the setting of systemic lupus

SLE

Connective tissue disease

Other

The exposure of tissue factor (TF) to circulating blood and the formation of a complex between TF and activated factor (FVIIa) is the most important trigger of blood coagulation [4]. FVIIa/TF activity is modulated by its major inhibitor, tissue factor pathway inhibitor (TFPI). Most intravascular TFPI (50–80%) is bound to the vessel wall but may be released by heparin and other negatively charged ions [5–7]. We have earlier measured TFPI activity levels in patients with LA [8] and did not find any significant differences compared to a control group, even though there was a trend for higher TFPI levels in patients with LA. This is in accordance with earlier reports [5,9]. The inhibitory mechanism is complex and involves first the binding and inhibition of FXa [10,11]. The TFPI/FXa complex then builds a quaternary complex with TF and FVIIa [11]. The ability of TFPI/FXa to inhibit TF/FVIIa is phospholipid dependent as the binding of FXa to phospholipids is essential for the reaction [12]. In the present study, we have explored the possible interference of APA with this reaction and more specifically looked for a potentially modulating effect of APA on heparin-induced TFPI release and activity.

heart valve insuff., hypertension arteriosclerosis oblit.

E.M. Jacobsen et al./Thrombosis Research 94 (1999) 213–220

Sex

214

E.M. Jacobsen et al./Thrombosis Research 94 (1999) 213–220

erythematosus (SLE), and both had experienced habitual abortions/intrauterine death, whereas one also had suffered from a deep venous thrombosis. The other five patients had probable or verified connective tissue disease but did not meet the clinical criteria for APS. All patients included were LA positive, five had high positive LA, while the other six had moderate positive tests. Nine of the eleven patients had positive tests for anticardiolipin antibodies (ACA), and eight had positive tests for anticephalin antibodies (Acepha).

1.2. Sample Preparation Blood was drawn immediately before and exactly 5 minutes after intravenous injection of 5000 IU unfractionated heparin (UFH) (Løvens kemiske Fabrik, Ballerup, Denmark). Vacutainert tubes containing 1:10 volume of 0.129 M sodium citrate (Becton Dickinson, Meylan, France) were used throughout. The blood was centrifuged within 30 minutes at 20003g for 15 minutes. Plasma was stored in aliquots at 2708C. Plasma for LA testing was filtered (Millex-GV 0.22 mm Filter Unit; Millipore, Bedford, MA, USA) before freezing and storage.

1.2.1. Pooled normal plasma Pooled normal plasma (NP) was prepared with citrated plasma from 30 healthy blood donors. NP for LA testing was likewise collected from 20 blood donors. An APTT was performed on each plasma to ensure that all pooled plasmas had a normal clotting time. The reference plasma for LA testing also was filtered before storage. 1.3. Assays 1.3.1. Anti-FXa activity Heparin concentration, as anti-Xa activity, was measured using a commercial chromogenic assay (Coatestt LMW heparin/heparin; Chromogenix AB, Mo¨lndal, Sweden). Standard curves were prepared by diluting UFH 100 IE/ml (Løvens kemiske Fabrik) in NP to final concentrations of 0 to 1.0 IU/ml. 1.3.2. TFPI total activity TFPI activity was quantified with a two-stage amidolytic assay essentially as described earlier [14],

215

except that the TF used was recombinant TF (rTF) (Dadet Innovint; Baxter Diagnostics Inc., Deerfield, IL, USA).

1.3.3. TFPI free antigen TFPI free antigen was determined using a solidphase two-site enzyme immunoassay [15]. Microtiter wells were coated overnight with a monoclonal anti-TFPI antibody. The microtiter wells were then incubated for 1 hour at room temperature with diluted plasma samples or standards and bound TFPI was detected by using a peroxidase-labeled rabbit polyclonal anti-TFPI antibody. Standard curves were made by dilutions of a recombinant, full-length TFPI preparation with known concentration. All reagents were gifts from Novo Nordisk A/S (Gentofte, Denmark). 1.3.4. TFPI anticoagulant assay A modified diluted prothrombin time assay (dPT) was used to estimate the anticoagulant effect of the endothelial TFPI released by heparin [16]. The anticoagulant effect of TFPI in post-heparin plasma was estimated after neutralising the effect of heparin on antithrombin by the addition of hexadimethrine bromide (HDB) (Polybrenet; Sigma, St. Louis, MO, USA) to a final concentration of 5 mg/ml, which corresponds to 15 mg/ml plasma. Theoretically, this would neutralise heparin up to 2.7 U/ml plasma. When dPT was performed on NP with added heparin, the chosen concentration of HDB normalised coagulation times of plasma with heparin concentration up to, and including, 2.5 U/ml (data not shown). Plasma (40 ml) was mixed with 10 ml Tris-buffered saline (TBS-buffer) or, in the case of postheparin samples, with 10 ml HDB in TBS-buffer. After 3 minutes of incubation at 378C with 50 ml of rTF, coagulation was started with 50 ml, 35 mmol/L CaCl2, and the clotting time was recorded with an automated coagulometer (ACL Futura, Instrumentation Laboratories). rTF was diluted in TBS-buffer to yield clotting times of about 50 seconds with NP, which corresponded to a final dilution of 1:300. The results of the dPT assay were calculated as normalised values, i.e., the coagulation times were divided by the coagulation time of NP tested in the same run.

216

E.M. Jacobsen et al./Thrombosis Research 94 (1999) 213–220

1.3.5. Anticardiolipin antibodies (ACA) The enzyme-linked immunosorbent assay (ELISAtest) for ACA was done essentially as described by Gharavi et al. [17]. The standards from Louisville (Diagnostica, St. Louis, MO, USA) defined by Harris et al. [18] were used. 1.3.6. Anticephalin antibodies (Acepha) Antibodies with specificity for cephalin were determined as described earlier [19]. This assay was shown to have a high concordance with a dilute APTT screening test for LA. The results were divergent in some cases, though, and the ELISA test may have higher sensitivity [19]. 1.3.7. Lupus anticoagulant Two semi-automated, integrated tests were used; one was based on the activated partial thromboplastin time (APTT), and the other on the Russell viper venom time (RVVT) [20]. In both cases crude cephalin from bovine brain was the phosholipid source (a gift from Nycomed Pharma, Oslo, Norway). The tests were performed with an automated coagulation laboratory (ACL Futura, Instrumentation Laboratories). Each test was performed twice, with two different cephalin concentrations, and the ratio between these coagulation times divided with the corresponding ratio for pooled NP. This final calculation gives the Lupus Ratio (LR) of that plasma. The 97.5 percentile of the LR in a reference population has been calculated earlier for each test and has been defined as the upper reference limit. For the APTT-based test, a LR between 1.05 and 1.29 signifies a low positive LA, a ratio between 1.30 and 1.49 is defined as moderate positive, and a LR of 1.50 or more is a high positive result. In the RVVT-based test, the upper reference limit is 1.11, and results between 1.20 and 1.39 are considered moderate positive, while a LR of 1.4 or more signifies a high positive LA. 1.3.8. Statistics Results are given as means of duplicate analyses. Histograms of the data showed normal distribution in both the patient and control group except for the results of the TFPI anticoagulant assay where the data from the patient group were skewed. Comparisons between the patient group and control group were made with Student’s paired t-test except for the TFPI anticoagulant assay where we used the

Fig. 1. TFPI activity (a), TFPI antigen (b), and TFPI anticoagulant assay (c) before and after an intravenous injection of unfractionated heparin 5000 IE in 11 patients with antiphospholipid antibodies and 11 controls matched for age and gender.

Wilcoxon matched pairs signed rank sum test. Correlations were tested by Spearman’s rank correlation (rs). Statistical significance was set at p,0.05.

2. Results 2.1. Preheparin Values of TFPI Amidolytic Activity/Antigen The mean preheparin TFPI amidolytic activity was slightly higher in patients than in controls (1.13 vs.

E.M. Jacobsen et al./Thrombosis Research 94 (1999) 213–220

1.01 U/ml) (Figure 1A). The difference was not, however, statistically significant (p50.20). When TFPI was measured as free antigen, the mean values were significantly higher in patients than in controls (21.7 vs. 14.2 ng/ml; p50.03) (Figure 1B).

2.2. Preheparin Results of the TFPI Anticoagulant Assay The clotting time ratios of the dPT assay were significantly higher in the patient group than in the control group (1.99 vs. 1.04; p,0.01) (Figure 1C). As dPT may be used as a test for LA [21], this was not surprising. dPT ratios in the patient group had a high correlation with the results of the APTTbased LR test (rs50.75; p50.01). The correlation with the RVVT-based LR was lower and not statistically significant (rs50.57; 0.1.p.0.05).

2.3. The Effect of Heparin on TFPI Levels The mean heparin concentration measured as antiXa activity 5 minutes after the injection was 1.03 IU/ml for the patient group vs. 1.17 IU/ml for the controls (p50.24). Both TFPI activity and antigen levels were increased substantially in patients as well as controls after the heparin injection. The postheparin TFPI activity was significantly higher in patients with APA than in controls (3.26 vs. 2.51 U/ml; p50.03) (Figure 1A). However, the relative rise in TFPI activity expressed as the ratio between TFPI values after heparin and before heparin was not significantly higher in patients than in controls (3.00 vs. 2.57; p50.14). Again, as for the preheparin values, the TFPI antigen assay showed significantly higher postheparin values in patients than in controls (179 vs. 153 ng/ml, p50.05) (Figure 1B). The relative rise in antigen levels was much higher than the rise in activity levels (10–12-fold increase vs. approximately threefold increase). This is in accordance with earlier reports [5–7]. However, as for the ratio of TFPI activity, the ratio of antigen levels after heparin and before heparin was not different in the two groups (9.7 vs. 12.4; p50.18).

2.4. The Effect of Heparin on the TFPI Anticoagulant Assay Postheparin plasmas were analyzed with HDB added to the buffer. Even after neutralisation of

217

heparin by HDB, the coagulation times of postheparin plasmas were clearly longer than those of preheparin plasmas in both patients and controls (Figure 1C). The normalised clotting times were higher in patients than in controls, but in postheparin plasma the difference was not statistically significant (2.80 vs. 2.18, 0.1.p.0.05). The relative increase as compared to preheparin levels tended to be lower for patients than for controls (the ratio of postheparin to preheparin values: 1.63 vs. 2.12, 0.1.p.0.05). The absolute increase in clotting times (in seconds) was less pronounced in patients with long preheparin coagulation times than in patients with shorter preheparin times (Figure 2). Actually, the coagulation time of one of the patients with the longest preheparin time was shorter in the postheparin sample (after neutralizing with HDB) than in the preheparin sample. The correlation between preheparin clotting times and the difference (in seconds) between pre- and postheparin clotting times was 20.78 in patients (p,0.01). There was no obvious correlation in the control group (Figure 2).

3. Discussion Theoretically, there are several possible ways in which APA may influence a system that measures the inhibitory effect of TFPI on TF-induced coagulation and modify the effect of heparin on that system. TF is composed of a transmembrane apoprotein and membrane phospholipids. The procoagulant activity of TF is phospholipid dependent [22]. Anionic phospholipids stimulate TF/FVIIa activity by increasing the binding affinity between the complex TF/FVIIa and its substrate FX [23]. Antibodies to TF have been found in a high percentage of patients with SLE and were highly correlated to thrombosis as well as to anticardiolipin antibodies and LA [24]. Furthermore, both plasma and immunoglobulin from patients with APA have been reported to upregulate TF expression in cultured endothelial cells and in monocytes [25–27]. Soluble TF was significantly higher in patients with APA than in healthy controls when measuring TF by a commercial ELISA [28]. High TF expression could lead to a low-grade, ongoing coagulation. In accordance with this hypothesis, higher prothrombin fragment 112 and thrombin–antithrombin

218

E.M. Jacobsen et al./Thrombosis Research 94 (1999) 213–220

Fig. 2. (a,b) TFPI anticoagulant assay. Coagulation times in preheparin plasma plotted against the absolute rise (in seconds) 5 minutes after heparin injection. The heparin effect on the antithrombin system are neutralised by hexadimethrine bromide (see text).

complex levels have been found in APA positive patients with upregulated TF expression [27]. Higher plasma levels of TF is not caused necessarily, however, by APA per se but may—as reported by Koyama et al. [29]—reflect endothelial cell injury. These authors have reported elevated TF levels in patients with vasculitis syndrome, disseminated intravascular coagulation (DIC), thrombotic thrombocytopenic purpura, and diabetic patients with retinopathy and nephropathy [30]. The formation of the inhibitory complex of TFPI, FXa, TF, and FVIIa is dependent on the binding of FXa to a phospholipid surface [12]. Theoretically APA could interfere with TFPI activity in vivo by inhibiting this complex formation. In a previous study we did not find, however, any decrease in TFPI activity in patients with LA compared with age-matched controls when measured with a chromogenic substrate assay [8]. In fact, the patients showed a trend to higher TFPI levels than the controls. In accordance with our previous report, the present study showed higher preheparin TFPI levels in these patients. The difference in TFPI amidolytic activity was not statistically significant, while the TFPI antigen levels were significantly higher. Higher antigen levels in these patients could be a response to an upregulated TF expression and/or higher plasma TF levels. The lack of a statistically significant difference in the amidolytic activity may be due

to the relatively small number of patients but also could be a result of an interference of APA on TFPI activity. TFPI bound to the vascular wall is released by heparin [6]. In vitro, heparin-releasable TFPI contributes strongly to the postheparin anticoagulant effect as measured in dPT assays [16]. The TFPI anticoagulant assay, which is a modified dPT assay, is difficult to use and interpret in these patients because dPT assay is a well-known screening test for LA [21,31]. The difference in preheparin clotting times between patients and controls was therefore expected. Arnout et al. normalised clotting times by dividing with the clotting time of a reference plasma and found a normal range of 0.921.15 for this ratio when using rTF diluted 1:200 [21]. We used rTF diluted 1:300, and our patients had values between 1.00 and 3.50 (mean 1.99) while the controls had 0.96–1.26 (mean 1.04). As stated, the dPT clotting times were highly correlated with the APTT-based LR (rs50.75). Surprisingly, the postheparin coagulation times, after neutralisation of the effect of heparin on the antithrombin system, were hardly different from the preheparin times in the patients with the longest preheparin coagulation times, that is, in the patients with the strongest LA activity. Since there was a profound rise in TFPI antigen levels and TFPI chromogenic substrate activity, this can not be explained by an inhibition of TFPI release by

E.M. Jacobsen et al./Thrombosis Research 94 (1999) 213–220

APA. TFPI anticoagulant activity seems, however, to be inhibited by strong LAs. This may be due to an interference of APA on the binding of FXa to phospholipid surfaces. Interestingly, the three patients with the highest preheparin clotting times and the least difference between pre- and postheparin clotting times were all among the four patients that have had arterial thrombosis. In conclusion, we have found a trend for higher levels of unstimulated TFPI chromogenic substrate activity and significantly higher levels of free TFPI antigen in patients with APA. Both TFPI amidolytic activity and antigen levels were significantly higher in patients than in controls after heparin injection. The higher TFPI antigen level/amidolytic activity in these patients may be a response to an upregulated TF activity. In an assay for TFPI anticoagulant activity, it seems that some patients with strong LA activity have a weaker response to heparin injection than the controls.

References 1. Triplett DA. Antiphospholipid-protein antibodies: Laboratory detection and clinical relevance. Thromb Res 1995;78:1–31. 2. Oosting JD, Derksen RHWM, Bobbink IWG, Hackeng TM, Bouma BN, De Groot PG. Antiphospholipid antibodies directed against a combination of phospholipids with prothrombin, protein c, or proteins: An explanation for their pathogenic mechanism? Blood 1993;81: 2618–25. 3. Shibata S, Harpel PC, Gharavi A, Fillit H. Autoantibodies to heparin from patients with antiphospholipid antibody syndrome inhibit formation of antithrombin III-thrombin complexes. Blood 1994;83:2532–40. 4. Rapaport SI, Rao LVM. Initiation and regulation of tissue factor-dependent blood coagulation. Arterioscler Thromb 1992;12:1111–21. 5. Novotny WF, Brown SG, Miletich JP, Rader DJ, Broze GJ Jr. Plasma antigen levels of the lipoprotein-associated coagulation inhibitor in patient samples. Blood 1991;78:387–93. 6. Sandset PM, Abildgaard U, Larsen ML. Heparin induces release of extrinsic coagulation pathway inhibitor (EPI). Thromb Res 1988;50: 803–13.

219

7. Hubbard AR, Weller LJ, Gray E. Measurement of tissue factor pathway inhibitor in normal and post-heparin plasma. Blood Coag Fibrinol 1994;5:819–23. 8. Jacobsen EM, Sandset PM, Wisløff F. Lupus anticoagulant and a functional assay for tissue factor pathway inhibitor. Thromb Res 1996; 83:339–40. 9. Bajaj MS, Rana SV, Wysolmerski RB, Bajaj SP. Inhibitor of the factor VIIa-tissue factor complex is reduced in patients with disseminated intravascular coagulation but not in patients with severe hepatocellular disease. J Clin Invest 1987;79:1874–8. 10. Broze GJ Jr, Warren LA, Novotny WF, Higuchi DA, Girard JJ, Miletich JP. The lipoprotein-associated coagulation inhibitor that inhibits the factor VII-tissue factor complex also inhibits factor Xa: Insight into its possible mechanism of action. Blood 1988;71:335–43. 11. Warn-Cramer BJ, Rao LVM, Maki SL, Rapaport SI. Modifications of extrinsic pathway inhibitor (EPI) and factor Xa that affect their ability to interact and to inhibit factor VIIa/ tissue factor: Evidence for a two-step model of inhibition. Thromb Haemost 1988;60:453–6. 12. Kazama Y. The importance of the binding of factor Xa to phospholipids in the inhibitory mechanism of tissue factor pathway inhibitor: The transmembrane and cytoplasmic domains of tissue factor are not essential for the inhibitory action of tissue factor pathway inhibitor. Thromb Haemost 1997;77:492–7. 13. Harris EN. Syndrome of the black swan. Br J Rheumatol 1987;26:324–6. 14. Sandset PM, Larsen ML, Abildgaard U, Lindahl AK, Ødegaard OR. Chromogenic substrate assay of extrinsic pathway inhibitor (EPI): Levels in the normal population and relation to cholesterol. Blood Coag Fibrinol 1991;2: 425–33. 15. Østergaard PB, Beck TC, Ørsted H, Svendsen A, Nordfang O, Sandset PM, Hansen J-B. An enzyme linked immunosorption assay for tissue factor pathway inhibitor. Thromb Res 1997; 87:447–59. 16. Hansen J-B, Huseby KR, Huseby N-E, Sandset PM, Hanssen T-A, Nordøy A. Effect of cholesterol lowering on intravascular pools of TFPI and its anticoagulant potential in type II hy-

220

17.

18.

19.

20.

21.

22.

23.

24.

E.M. Jacobsen et al./Thrombosis Research 94 (1999) 213–220

perlipoproteinemia. Arterioscler Thromb Vasc Biol 1995;15:879–85. Gharavi AE, Harris EN, Asherson RA, Hughes GRV. Anticardiolipin antibodies: Isotype distribution and phospholipid specificity. Ann Rheum Dis 1987;46:1–6. Harris EN, Gharavi AE, Patel SP, Hughes GRV. Evaluation of the anti-cardiolipin antibody test: Report of an international workshop held 4 April 1986. Clin Exp Immunol 1987; 68:215–22. Sletnes KE, Keirung G, Wisløff F. Quantitation of anticephalin antibodies in a computerassisted enzyme-linked immunosorbent assay (ELISA): Relation to lupus anticoagulant. Thromb Res 1990;57:235–46. Schjetlein R, Sletnes KE, Wisløff F. A quantitative, semi-automated and computer-assisted test for lupus anticoagulant. Thromb Res 1993; 69:239–50. Arnout J, Vanrusselt M, Huybrechts E, Vermylen J. Optimization of the dilute prothrombin time for the detection of the lupus anticoagulant by use of a recombinant tissue thromboplastin. Br J Haematol 1994;87:94–9. Nemerson Y. The phospholipid requirement of tissue factor in blood coagulation. J Clin Invest 1968;47:72–80. Bach R, Gentry R, Nemerson Y. Factor VII binding to tissue factor in reconstituted phospholipid vesicles: Induction of cooperativity by phosphatidylserine. Biochemistry 1986;25:4007–20. Font J, Lo´pez-Soto A, Cervera R, Casals FJ, Reverter JC, Mun˜oz FJ, Miret C, Bove´ A, Ordinas A, Ingelmo M. Antibodies to thromboplastin in systemic lupus erythematosus: Iso-

25.

26.

27.

28.

29.

30.

31.

type distribution and clinical significance in a series of 92 patients. Thromb Res 1997;86: 37–48. Tannenbaum SH, Finko R, Cines DB. Antibody and immune complexes induce tissue factor production by human endothelial cells. J Immunol 1986;137:1532–7. ˚ -B, MiSletnes KE, Kierulf P, Westvik A chaelsen TE, Wisløff F. Increased fibrin generation by cryopreserved, lipopolysaccharidestimulated, normal human monocytes incubated with polyclonal antiphospholipid antibodies. Thromb Haemost 1993;69:542. Reverter J-C, Ta´ssies D, Font J, Monteagudo J, Escolar G, Ingelmo M, Ordinas A. Hypercoagulable state in patients with antiphospholipid syndrome is related to high induced tissue factor expression on monocytes and to low free protein S. Arterioscler Thromb Vasc Biol 1996;16:1319–26. Atsumi T, Khamashta MA, Amengual O, Hughes GRV. Up-regulated tissue factor expression in antiphospholipid syndrome. Thromb Haemost 1997;77:222–3. Koyama T, Ohdama S, Aoki N. Plasma tissue factor reflects endothelial cell injury rather than upregulation of tissue factor expression. Thromb Haemost 1997;78:972. Koyama T, Nishida K, Ohdama S, Sawada M, Murakami N, Hirosawa S, Kuriyama R, Matsuzawa K, Hasegawa R, Aoki N. Determination of plasma tissue factor antigen and its clinical significance. Brit J Haematol 1994;87:343–7. Schleider MA, Nachman RL, Jaffe EA, Coleman M. A clinical study of the lupus anticoagulant. Blood 1976;48:499–509.