koagulasi pada sle

Lupus (2011) 20, 1474–1483

http://lup.sagepub.com

PAPER

Impaired control of the tissue factor pathway of blood

coagulation in systemic lupus erythematosus

MJ Adams1, AA Palatinus1, AM Harvey1 and AA Khalafallah1,21School of Human Life Sciences, University of Tasmania, Launceston, TAS, Australia; and 2Haematology Research Unit, Pathology Department,

Launceston General Hospital, Launceston, TAS, Australia

Thrombosis is a frequent manifestation in patients with systemic lupus erythematosus (SLE),although precise mechanisms remain unclear. This study investigated whether the major phys-iological trigger of blood coagulation, the tissue factor (TF) pathway, was altered in SLEpatients. Furthermore, we investigated potential associations between the TF pathway, thepresence of antiphospholipid (APL) antibodies and other abnormalities present in SLE.A total of 101 participants (40 SLE patients and 61 age- and sex-matched controls) wererecruited from Tasmania, Australia. Markers of the TF pathway, hypercoagulability, inflam-mation and endothelial cell damage were measured in plasma. Serum levels of APL antibodies(anti-cardiolipin antibodies [ACL], lupus anticoagulants [LAC], anti-beta2-glycoprotein-1[anti-b2GP1] and anti-prothrombin antibodies) were also determined. Despite similar TFand TF pathway inhibitor (TFPI) total antigen levels, SLE patients had significantly increasedlevels of TFPI free antigen (patients vs controls; mean� SD) (11.6� 0.9 ng/mL vs6.4� 0.4 ng/mL; p< 0.001) but significantly reduced TFPI activity (0.66� 0.07U/mL vs1.22� 0.03U/mL; p< 0.001), compared with healthy controls. Anti-TFPI activity, designatedas the ability of isolated IgG fractions to inhibit TFPI activity in normal plasma, was detectedin 19/40 (47.5%) of SLE patients and 3/40 (7.5%) of healthy controls. The significant reduc-tion in TFPI activity in SLE patients reflects impaired functional control of the TF pathway.Moreover, SLE patients with a history of thrombosis demonstrated higher levels of TFPIactivity compared with patients without a previous thrombotic event (0.97� 0.07U/mL vs0.53� 0.14U/mL; p¼ 0.0026). Changes to the TF pathway were not associated with manifes-tations of SLE such as inflammation or endothelial cell damage. The results from this studysuggest hypercoagulability in SLE may (in part) be due to reduced TFPI activity, a mechanismthat appears to be independent of other abnormalities in SLE. Lupus (2011) 20, 1474–1483.

Key words: anti-TFPI; systemic lupus erythematosus; thrombosis; tissue factor; tissue factorpathway inhibitor

Introduction

SLE is a chronic inflammatory disease that canaffect any part of the body, including the skin,joints, liver, kidneys and blood. Haematologicalcomplications are common in SLE, including anae-mia, the presence of circulating APL antibodies(e.g. ACL, LAC, and anti-b2GP1 antibodies) andthrombotic manifestations. Indeed, it has beenreported that patients with SLE who are positivefor APL antibodies are three times more likely to

have a thrombotic event compared with subjectswho are negative for APL antibodies,1 althoughthe precise mechanisms are yet to be fully eluci-dated.2 Other haemostatic factors, includingFactor V Leiden genotype,3 platelet activation4,5

and the action of APL antibodies on monocytesto induce tissue factor (TF) expression and procoa-gulant activity,6–8 have also been proposed to con-tribute to hypercoagulability in SLE patients.

TF is an integral membrane protein that is notnormally expressed on cells in contact with circu-lating blood. Following vessel injury TF complexeswith coagulation factor VII(a) to initiate a series ofprotein/enzyme interactions through the TF path-way, resulting in thrombin generation, coagulationamplification, platelet activation, and fibrin

Correspondence to: Murray Adams, BSc(Hons) PhD, FFSc (RCPA),

School of Human Life Sciences, University of Tasmania, Bag 1320

Launceston, Tasmania, 7250, Australia

Email: [email protected]

Received 2 December 2010; accepted 5 July 2011

! The Author(s), 2011. Reprints and permissions: http://www.sagepub.co.uk/journalsPermissions.nav 10.1177/0961203311418267

deposition. TF is regulated by tissue factor path-way inhibitor (TFPI), a Kunitz-type protease inhib-itor that is well described as the major regulator ofthe TF pathway.9–12 TFPI acts in a factor Xa-dependent manner to inhibit the TF-factor VIIacomplex, and thus regulate thrombin generationand fibrin formation. Although varying circulatinglevels of TFPI have been reported in many diseases,a ‘TFPI deficiency’ phenotype has yet to be clearlydefined.

Altered balance of the TF pathway, usuallythrough increased TF expression that is not com-pensated by a similar increase in TFPI, is thoughtto contribute to increased hypercoagulability in avariety of conditions including sepsis,13 acute respi-ratory distress,14 disseminated intravascular coagu-lation,15 and the antiphospholipid syndrome,10 buthas yet to be extensively investigated in patientswith SLE. The aim of the present study was totherefore investigate whether there were changesto the TF pathway in a well-defined cohort ofSLE patients, compared with healthy controls.Furthermore, we investigated potential associationsbetween markers of the TF pathway, and the pres-ence of different APL antibodies, and other abnor-malities present in SLE (i.e. inflammation andendothelial cell damage).

Materials and methods

Controls and patients

This study was approved by the Tasmanian Healthand Medical Human Research Ethics Committee(Reference number: H009415). A total of 101 sub-jects were recruited, comprising 40 SLE patientsand 61 age- and sex-matched healthy controls.SLE patients met the American College ofRheumatology Classification,16 and time(mean� SD) since diagnosis was 10.1� 7.3 years.All participants were recruited through theUniversity of Tasmania, the Lupus Association ofTasmania and the Launceston General Hospital(Tasmania), and provided informed consent.

Approximately one-third of SLE patients(n¼ 12) reported a previous thrombotic event.These included lower limb deep vein thrombosis(DVT) (n¼ 8), thrombosis in the groin (n¼ 2),ovary (n¼ 2) and knee (n¼ 1), pulmonary embo-lism (n¼ 3), stroke (n¼ 2), post-operative DVT(n¼ 3), post-partum DVT (n¼ 2), miscarriage(n¼ 1) and thrombophlebitis (n¼ 1). More thanone type of thrombotic episode was reported by

six (16%) SLE patients. No healthy controlreported a previous thrombotic event.

Blood sampling

Whole blood (approximately 6mL) was collectedfrom the cubital vein. For the detection of APLantibodies, blood was collected into tubes withoutanticoagulant to obtain serum. For the measure-ment of markers of the TF pathway, thrombin gen-eration, inflammation and endothelial cell damage,one part whole blood was collected into nine partssodium citrate (3.8%) and centrifuged at 2000 g for10 minutes. The plasma was separated and centri-fuged a second time at 2000 g for 10 minutes toobtain platelet poor plasma (PPP). Whole bloodwas collected from a further 20 healthy subjectsto obtain pooled normal plasma (PNP) and usedto generate standard curves for the TFPI and anti-TFPI activity assays. All samples were stored in1mL aliquots at �80�C until assayed.

Isolation of IgG fractions

IgG fractions were isolated from PPP by affinitypurification using protein G columns (GEHealthcare, Uppsala, Sweden). Samples werediluted 1:5 in 0.02mol/L phosphate bufferedsaline (PBS), pH 7.4. Prior to use, columns werewashed and equilibrated with 10mL of PBS. Onehundred mL 1M Tris-HCl (pH 9.0) was added tocollection tubes to neutralize the pH of the fractioncollected. A 5mL diluted sample was applied to thecolumn, which was then washed with 5mL of PBS.IgG was eluted from the column using 5mL gly-cine, pH 2.7. The purified fractions were bufferexchanged with PBS using a desalting column(GE Healthcare, Buckingham, UK). After elution,the absorbance of each 1mL fraction was measuredat 280 nm, and the aliquot with the highest absor-bance was added to 1.5mL PBS pH 7.4. Thisvolume was applied to the desalting column thatwas equilibrated using 25mL PBS, pH 7.4. Theflow through was discarded, then IgG fractionswere eluted using 3.5mL buffer. Equal volumes ofeach fraction were pooled and stored at -80�C untilrequired.

Antiphospholipid antibodies

Serum samples were analysed using commerciallyavailable assays to determine levels of APL antibo-dies. The assays were; Staclot� dRVV LADetection Kit for lupus anticoagulants(Diagnostica Stago, Asnieres, France), enzymelinked immunosorbent assay (ELISA) for IgG

Impaired control of the tissue factor pathway of blood coagulation in SLEMJ Adams et al.

1475

Lupus

and IgM isotypes of ACL (Enzyme LinkedImmunoConcepts�, Sacramento, USA), and ELISAs foranti-b2GP1 and anti-prothrombin IgG antibodies(Corgenix, Inc., Broomfield, USA). All assayswere performed according to manufacturer’sinstructions.

Laboratory methods

Markers of the TF pathway, thrombin generation,inflammation and endothelial cell damage weremeasured using commercially available assaysaccording to manufacturer’s instructions, unlessotherwise stated. TF, TFPI total antigen andTFPI free antigen were measured using ELISA(American Diagnostica Incorporated, Greenwich,USA). TF activity was measured using a chromo-genic activity assay (Abnova, Taipei City, Taiwan).Thrombin-antithrombin (TAT) complexes and pro-thrombin fragment 1þ2 (F1þ2 fragments) weremeasured using ELISA (Dade-Behring, Marburg,Germany). Interleukin 6 (IL-6) was measuredusing a high sensitivity quantitative enzyme immu-noassay (Quantikine1 HS; R&D Systems Inc.,Minneapolis, USA). Human soluble E-selectin(sE-selectin) was measured by ELISA (HycultBiotechnology b.v., Uden, The Netherlands).

TFPI activity was measured using an amidolyticassay as previously described.17–19 Anti-TFPI activ-ity was determined using a slight modification ofthe TFPI activity assay. Equal volumes of IgG frac-tions (25 mg/mL) isolated from SLE patients andhealthy controls were mixed with PNP (containing1U/mL TFPI activity) for 30 minutes at 37�C. Todetermine anti-TFPI activity, the TFPI activity ofPNP with IgG was subtracted from the TFPI activ-ity of PNP mixed with TFPI activity buffer

(standard) to demonstrate the effect of IgG onTFPI activity.

Statistical analysis

Statistical analysis was performed using StatisticalPackage for Social Sciences (SPSS v18, Chicago,USA). The data is presented as mean � standarddeviation (SD) for normally distributed data, or asmedian (interquartile range; IQR) for non-nor-mally distributed data. Differences betweengroups were determined using either an unpairedt-test or Mann–Whitney test, depending on the dis-tribution of data. Associations between the variouslaboratory markers were determined usingPearson’s correlation coefficient (r) or Spearman’srank correlation coefficient (rho), depending on thedistribution of data. P-values <0.05 were consid-ered statistically significant.

Results

Controls and patients

Demographic, APL antibody status and throm-botic history of healthy controls and SLE patientsare summarized in Table 1. Healthy controls andSLE patients were matched for age and gender.SLE patients had a significantly higher incidenceof LAC, anti-b2GP1 and anti-prothrombin antibo-dies, compared with healthy controls, but there wasno difference in the incidence of ACL (IgG andIgM) antibodies between the two groups. Asexpected, SLE patients had a higher overall inci-dence of both APL antibodies and thromboticevents.

Table 1 Demographic and APL antibody data of healthy controls and SLE patients

Controls Patients p value

n 61 40

Age (mean� 2SD) years 50.3� 8.3 55.4� 12.2 NS

Female % 82 92 NS

LAC 5/61 (8%) 11/40 (28%) 0.0102

Anti-b2GP1 0/61 (0%) 6/40 (15%) 0.0139

Anti-PT 0/61 (0%) 4/40 (25%) 0.012

ACL IgG 1/61 (2%) 5/40 (11%) NS

ACL IgM 5/61 (8%) 6/40 (14%) NS

Overall APL antibodya 10/61 (16%) 16/40 (40%) 0.0084

History of thrombosis 0/61 (0%) 12/40 (30%) <0.0001

aOverall APL antibody; subject was positive for one or more of LAC, ACL (IgG or IgM), anti-b2GP1 or anti-prothrombin. ACL,

anti-cardiolipin antibodies; anti-b2GP1, anti-beta2-glycoprotein-1; anti-PT, anti-prothrombin; APL, antiphospholipid; LAC, lupus

anticoagulants; NS, not significant.


1476

Lupus

Markers of the TF pathway

Results are summarized in Table 2. The medianlevel of TF antigen in SLE patients was increasedcompared with healthy controls, but this was notstatistically significant (77.5 vs 52.0 pg/mL;p¼ 0.224) (Figure 1A). Plasma levels of TFPI anti-gen (total) were not statistically different betweencontrols and patients (Figure 1B). SLE patients hadhigher levels of TFPI antigen (free) (mean� SD)(11.6� 0.9 vs 6.4� 0.4 ng/mL; p< 0.001)(Figure 1C), but lower mean TFPI activity(0.66� 0.07 vs 1.22� 0.03U/mL; p< 0.001)(Figure 2A), compared with healthy controls.Anti-TFPI activity was significantly higher inSLE patients compared with healthy controls(Figure 2B), and was detected in 19/40 (47.5%) ofpatients and 3/40 (7.5%) of controls tested.

Analysis of TF:TFPI ratio

To further evaluate changes to the balance of the TFpathway, the ratio of TF to various measures ofTFPI were determined in healthy controls andSLE patients (Figure 3). There was no significantdifference in the ratio of TF(antigen): TFPI(totalantigen), (controls vs patients; median[IQR]) (0.71[0.34–1.60] vs 1.00 [0.50–1.50]; p¼ 0.42) orTF(antigen):TFPI(free antigen) (9.1 [3.0–19.9]vs 6.6 [3.5–9.2]; p¼ 0.054). However, there wasa statistically significant difference in the ratioof TF(antigen): TFPI(activity) (42.1 [18.5–95.2]vs 99.0 [63.4–281.4]; p< 0.001) and TF(activity):TFPI(activity) (112.5 [77.1–135.9] vs 201.5[105.50–407.7]; p< 0.001), between the two cohorts(Figure 3).

Markers of thrombin generation, inflammationand endothelial cell damage

SLE patients had significantly higher levels of inter-feron-gamma (IFN-g) compared to healthy con-trols. SLE patients had higher levels of F1þ2fragments than controls, but this was not statisti-cally significant (p¼ 0.091). There was no signifi-cant difference in the level of TAT complexes inSLE patients and healthy controls. Similarly,there were no significant differences in laboratorymarkers of inflammation (IL-6, tumour necrosisfactor alpha [TNF-a]) or endothelial cell damage(sE-selectin) between patients and controls.Results are summarized in Table 2.

Correlation analysis

TFPI free antigen was significantly associated withTFPI total antigen in both healthy controls(r¼ 0.348; p¼ 0.006) and SLE patients (r¼ 0.734;p< 0.001) (Figure 4). However, TFPI activity wasnot associated with TFPI total or free antigen ineither controls or patients (both p> 0.05).Furthermore, there was no association betweenTFPI activity and anti-TFPI activity in either con-trols or patients (p> 0.05).

In healthy controls, there were significant associ-ations between TFPI activity and ACL IgG(r¼�0.355; p¼ 0.005), anti-b2GP1 and TFPI freeantigen (r¼ 0.395; p¼ 0.002), anti-prothrombinantibodies and TFPI free antigen (rho¼ 0.439;p< 0.001), F1þ 2 and TFPI free antigen(rho¼ 0.471; p< 0.001), and IL-6 and TFPI freeantigen (rho¼ 0.411; p¼ 0.001). In SLE patients,

Table 2 Markers of the TF pathway, thrombin generation, inflammation and endothelial cell damage in healthycontrols and SLE patients

Controls Patients p value

TF activity (pg/M)a 132.6� 46.7 128.1� 66.2 0.700

TF antigen (pg/mL)b 52.0 (24.0–103.5) 77.5 (35.8–109.3) 0.224

TFPI total (ng/mL)a 66.2� 2.5 74.4� 4.7 0.091

TFPI free (ng/mL)a 6.4� 0.4 11.6� 0.9 <0.001

TFPI activity (U/mL)a 1.22� 0.03 0.66� 0.07 <0.001

TAT complexes (mg/L)b 2.5 (0.9–5.8) 2.3 (1.0–6.8) 0.673

F1þ2 fragments (nmol/L)b 329 (230–402) 399 (236–677) 0.097

IL-6 (pg/mL)b 1.1(1.0–1.6) 1.5 (1.0–1.8) 0.169

IFN-g (pg/mL)b 0.0 (0.0–0.0) 1.3 (0.0–22.6) <0.001

TNF-a (pg/mL)b 1.0 (0.5–1.9) 1.4 (0.9–2.0) 0.066

sE-selectin (pg/mL)a 19.1� 1.2 21.0� 1.9 0.385

aData presented as mean� SD; differences between groups determined using unpaired t-test.bData presented as median (IQR); differences between groups determined using Mann–Whitney test. SLE patients had significantly

increased TFPI free antigen, but significantly decreased TFPI activity, compared with healthy controls. IFN, interferon;

IL, interleukin; TAT, thrombin-antithrombin; TF, tissue factor; TFPI, TF pathway inhibitor; TNF, tumour necrosis factor.


1477

Lupus

there were no significant associations between theparameters measured.

TFPI activity was significantly elevated in SLEpatients with a history of thrombosis, comparedwith patients who had not experienced a throm-botic event (n¼ 12 (0.97� 0.07U/mL) vs n¼ 28(0.53� 0.14U/mL); p¼ 0.0026) (Figure 5).Furthermore, there were no significant differencesbetween SLE patients according to APL status i.e.APL negative vs SLE APL positive, for any of themeasured parameters (data not shown).

Discussion

The present study has demonstrated impaired TFPIregulation of the TF pathway of blood coagulation

in patients with SLE, compared with healthy con-trols. Interestingly, this was not reflected bychanges in either TF (antigen or activity) or TFPItotal antigen levels, but was rather due to signifi-cantly reduced plasma levels of TFPI activity inSLE patients. Furthermore, we demonstrated thatIgG fractions isolated from SLE patients signifi-cantly inhibit TFPI activity of normal pooledplasma. Anti-TFPI activity was detected in approx-imately half of the IgG fractions isolated from SLEpatients in this study cohort, although it was notassociated with history of thrombosis. Moreover,changes to the regulation of the TF pathway werenot associated with inflammation or endothelial celldamage, suggesting that the development of hyper-coagulability is independent of these clinical mani-festations in SLE.

Controls Patients

0

100

200

300

500

550p = NS

TF

an

tig

en (

pg

/mL

)

Controls Patients

0

50

100

150

200p = NS

TF

PI a

nti

gen

(to

tal)

(n

g/m

L)

Controls Patients

0

10

20

30p < 0.001

TF

PI

anti

gen

(fr

ee)

(ng

/mL

)

(B)(A)

(D)(C)

Controls Patients

0

100

200

300

400p = NS

TF

act

ivit

y (p

M)

Figure 1 Plasma levels of TF and TFPI in healthy controls and SLE patients. Although plasma levels of both TF antigen (A),TF activity (B) and TFPI total antigen (D) were not significantly different, SLE patients (n¼ 40) had significantly increased levelsof TFPI free antigen (C) compared with healthy controls (n¼ 61). Solid horizontal lines indicate mean values for TF activity, TFPItotal antigen and TFPI free antigen, and median values for TF antigen. TF, tissue factor; TFPI, TF pathway inhibitor.


1478

Lupus

Controls Patients

0.0

2.5

5.07.0

12.0 p = NS

TF

(an

tig

en):

TF

PI (

tota

l an

tig

en)

rati

o

Controls Patients

0

25

5060

110 p = NS

TF

(an

tig

en):

TF

PI (

free

an

tig

en)

rati

o

Controls Patients

0

125

250300

1200 p < 0.001

TF

(an

tig

en):

TF

PI (

acti

vity

) ra

tio

Controls Patients

0

350

700

1400 p < 0.001

TF

(ac

tivi

ty):

TF

PI (

acti

vity

) ra

tio

(A)

(D)(C)

(B)

Figure 3 Ratio of TF:TFPI. There was no significant difference in the ratio of TF antigen to either TFPI total (A) or TFPI freeantigen (B) between SLE patients (n¼ 40) and healthy controls (n¼ 61). However, SLE patients had a significantly increased ratioof TF antigen to TFPI activity (C) and TF activity to TFPI activity (D), reflecting impaired control of the TF pathway of bloodcoagulation. Solid horizontal lines indicate median values. TF, tissue factor; TFPI, TF pathway inhibitor.

Controls Patients

0.00

0.50

1.00

1.50

2.00p < 0.001

TF

PI a

ctiv

ity

(U/m

L)

Controls Patients

0.0

0.2

0.4

0.6

0.8

1.0p < 0.001

An

ti-T

FP

I act

ivit

y (U

/mL

)

(B)(A)

Figure 2 TFPI and anti-TFPI activity in healthy controls and SLE patients. SLE patients (n¼ 40) had significantly reduced levelsof TFPI activity (A) i.e. function, but increased levels of anti-TFPI activity (B) compared with healthy controls (n¼ 61). Solidhorizontal lines indicate mean values. TFPI, TF pathway inhibitor.


1479

Lupus

SLE patients demonstrated higher levels of APLantibodies (LAC, anti-b2GP1 and anti-prothrom-bin) compared with healthy controls, consistentwith other studies.20,21 It has previously been dem-onstrated that approximately 50% of SLE patientsand 1–5% of the general population are positive forAPL antibodies.22–24 The heterogeneity and tran-sient nature of APL,25,26 as well as the lack of stan-dardized diagnostic methods, contribute to somedifficulty in detecting the presence of APL,27

which may also help explain the lack of associationbetween APL and markers of the TF pathway.Moreover, APL antibody levels increase in patientswith chronic disease and older individuals,23 andtogether with the relatively small sample size ofthe current study, may explain the higher incidenceof APL antibodies demonstrated in healthy con-trols of the present study (16%), where the meanage was 50 years.

There is minimal information available in theliterature about the role of the TF pathway inSLE. In this study, we have demonstrated no sig-nificant difference in TF antigen levels between SLEpatients and healthy controls. It was somewhatunexpected that increased levels of TF were not

demonstrated as it has previously been reportedthat monocytes from SLE patients expressincreased levels of cell surface TF, particularly inpatients who had a previous thrombotic event.28 Inthe current study we hypothesized that higher cir-culating TF would reflect increased expression dueto an ongoing inflammatory state and endothelialcell disruption. It may be that TF expressed bymonocytes has a more significant role in modulat-ing the ongoing thrombin generation in SLE, com-pared with that expressed by endothelial cells. Ourdata provides some support for this statement inthat sE-selectin levels in SLE patients were similarto healthy controls, suggesting minimal or no endo-thelial cell disruption, although this may be con-founded by immunosuppressive treatment.

Our data demonstrates no significant changes toTFPI total antigen levels between the patient andcontrol groups. This is in contrast to previous stud-ies that have reported both increased29 anddecreased30 TFPI total antigen levels. The reasonsfor the varying results are unknown, however,increased TFPI total antigen may reflect endothe-lial cell injury. Again, sE-selectin levels were notaltered in our study, which may in part providean explanation of the similar patient and controlTFPI total antigen levels. Decreased circulatingTFPI total antigen levels have also previouslybeen reported, however the vascular endothelialpool of TFPI was not affected following heparinadministration.30

Free TFPI is the 43 kDa full length form of theinhibitor that circulates in plasma, but is not asso-ciated with lipoproteins. It is thought to play amore biologically active role in regulating the TFpathway as it has greater anticoagulant activitythan the lipoprotein bound forms of TFPI.31,32

The increase in TFPI free antigen levels in SLEpatients reported in the current study is in agree-ment with earlier work,29 and may reflect eitherincreased release from intracellular/cell surface-associated stores, endothelial cell injury,29 releasefrom platelets following activation,33 or possiblyredistribution from (primarily) low density lipopro-tein stores.

A limitation of many studies that investigate therole of the TF pathway in normal haemostasis anddisease is that they do not assess the activity ofTFPI using functional assays. These techniquesmore appropriately reflect the in vivo situationcompared with antigen methods, as their principledepends on the ability of TFPI to inhibitTF-FVIIa-FXa (the complex formed followingthe initiation of coagulation). To our knowledge,the current study is the first to report TFPI activity

20 70 120 170

0

10

20

30

TFPI total (ng/mL)

TF

PI f

ree

(ng

/mL

)

Figure 4 Association between TFPI total and free antigenlevels in healthy controls (�) and SLE patients (x). Therewere significant, positive associations between TFPI totaland free antigen levels in both healthy controls (r¼ 0.348;p¼ 0.006) and SLE patients (r¼ 0.734; p< 0.001), howeverTFPI activity was not associated with TFPI total or free anti-gen levels, in either controls or patients. TFPI, TF pathwayinhibitor.


1480

Lupus

in a well-defined cohort of SLE patients, comparedwith healthy controls.

We demonstrated that TFPI activity wasdecreased in SLE patients despite normal TFPItotal antigen and increased TFPI free antigenlevels. These results suggest that although TFPI,particularly the more biologically active free form,is present at normal to increased levels in SLEpatients, it does not have the same inhibitorycapacity as TFPI in normal plasma. Furthermore,in SLE patients with a history of thrombosis, TFPIactivity levels were unexpectedly increased, com-pared with patients who had not previously had athrombotic event. Reduced circulating TFPI activ-ity levels in SLE patients may therefore reflect apre-thrombotic environment and contribute to the

development of thrombosis in conjunction withother risk factors.

It is interesting to speculate why the SLE cohortin this study demonstrated reduced TFPI activity.Possible reasons may include the action of serineproteases such as plasmin, thrombin and neutrophilelastase,34–36 that have previously been demon-strated to cleave TFPI, possibly causing reducedanticoagulant function. Other possibilities mayinclude the presence of autoantibodies and/orcross-reacting APL antibodies that partially orcompletely attenuate TFPI function. The currentstudy has also confirmed the presence of an inhib-itory component to TFPI in approximately half(48%) of the IgG fractions isolated from theplasma of SLE patients. This is in agreement with

No thrombosis Thrombosis

0.00

0.50

1.00

1.50

2.00p = 0.0026

TF

PI a

ctiv

ity

(U/m

L)


0

50

100

150

200p = NS

TF

PI a

nti

gen

(to

tal)

(n

g/m

L)


0

10

20

30

40p = NS

TF

PI

anti

gen

(fr

ee)

(ng

/mL

)


0.00

0.20

0.40

0.60

0.80

1.00 p = NSA

nti

-TF

PI (

U/m

L)

(B)(A)

(D)(C)

Figure 5 TFPI and anti-TFPI levels in SLE patients (grouped according to history of thrombosis). TFPI activity was significantlyelevated in SLE patients with a history of thrombosis. There was no significant difference in TFPI antigen (total or free), or anti-TFPI activity, in SLE patients with a history of thrombosis compared with those without history of thrombotic event. Solidhorizontal lines indicate mean values. TFPI, TF pathway inhibitor.


1481

Lupus

earlier studies using samples obtained from patientswith the antiphospholipid syndrome, which found65%17 and 48%.19 Our data suggests that the inhib-itory component in IgG fractions from SLEpatients is unlikely to be an APL antibody, asthere was no clear association between any of theAPL antibodies studied and anti-TFPI activity.Moreover, it is noted that the majority of theSLE cohort in this study was receiving immunosup-pressive agents which would, 1) suggest that themechanism of impaired control of the TF pathwayis independent of inflammation, 2) reduce the like-lihood of immune production of autoantibodies,and 3) provide a reason why IL-6 and IFN-glevels were not significantly increased, comparedwith healthy controls.

It is acknowledged that the current study has twoimportant limitations. Firstly, there was a relativelysmall sample size to detect significant differencesand/or potential associations between the markersinvestigated. Secondly, methods used to measureTFPI demonstrate variable correlation when com-paring results obtained using functional (activity)and protein (antigenic) techniques,37–39 which isdue to the different principles of detection used bythese methods. ‘Activity’ assays measure the func-tional capacity of TFPI to inhibit TF-FVIIa-FXacomplexes, whereas ‘antigenic’ methods use antibo-dies to capture different regions of the TFPI mole-cule, and thus detect the overall protein level ofTFPI. Moreover, the plasma pool of TFPI is a het-erogeneous mixture of different molecular weightforms of the inhibitor that contains variable func-tional capacity due to the association with lipopro-teins and the probable degradation of the moleculeby plasma proteases. Therefore, it was not surpris-ing that the data in the current study demonstratedgood correlation when comparing total and freeantigen levels of TFPI (Figure 5), but not whencomparing antigen with activity. This confirms theimportance of determining the functional ability,rather than just the presence of the protein, in stud-ies investigating TFPI.

In conclusion, this study has demonstrated thatdespite normal protein levels of TFPI, SLE patientshave both decreased TFPI activity and increasedanti-TFPI activity, compared with healthy controls.These changes reflect impaired regulation of the TFpathway that may contribute to hypercoagulabilityin SLE. Furthermore, these changes were not asso-ciated with other markers that reflect clinical man-ifestations of SLE, i.e. inflammation andendothelial cell damage. Further studies are clearlywarranted to determine how TFPI activity is

impaired, and whether there are other co-factorsinvolved in this interference to the TF pathway.

Acknowledgements

The authors would like to thank the LupusAssociation of Tasmania for their support of thisstudy. We are also grateful for the cooperation ofeveryone at Launceston Pathology, HobartPathology, North-West Pathology (Burnie andLatrobe) and the Pathology Department at theLaunceston General Hospital in helping withsample collection and processing. We would alsolike to thank Mrs Merrilyn Johnson and MrsNatasha Betts for their excellent technicalassistance.

Funding

This work was supported by a grant from theClifford Craig Medical Research Trust (Project66), and a Faculty of Health Science StrategicGrant, University of Tasmania.

References

1 Kaiser R, Cleveland CM, Criswell LA. Risk and protective factorsfor thrombosis in systemic lupus erythematosus: results from alarge, multi-ethnic cohort. Ann Rheum Dis 2009; 68: 238–241.

2 Palatinus A, Adams M. Thrombosis in systemic lupus erythema-tosus. Semin Thromb Hemost 2009; 35: 621–629.

3 Kaiser R, Barton JL, Chang M, et al. Factor V Leiden and throm-bosis in patients with systemic lupus erythematosus: a meta-analy-sis. Genes Immun 2009; 10: 495–502.

4 Dhar JP, Andersen J, Essenmacher L, Ager J, Bentley G, Sokol RJ.Thrombophilic patterns of coagulation factors in lupus. Lupus2009; 18: 400–406.

5 Pereira J, Alfaro G, Goycoolea M, et al. Circulating platelet-derived microparticles in systemic lupus erythematosus.Association with increased thrombin generation and procoagulantstate. Thromb Haemost 2006; 95: 94–99.

6 de Prost D, Ollivier V, Ternisien C, Chollet-Martin S. Increasedmonocyte procoagulant activity independent of the lupus antico-agulant in patients with systemic lupus erythematosus. ThrombHaemost 1990; 64: 216–221.

7 Martini F, Farsi A, Gori AM, et al. Antiphospholipid antibodies(aPL) increase the potential monocyte procoagulant activity inpatients with systemic lupus erythematosus. Lupus 1996; 5:206–211.

8 Ternisien C, Ollivier V, Chollet-Martin S, de Prost D. Increasedmonocyte procoagulant activity in patients with systemic lupuserythematosus. Nouv Rev Fr Hematol 1990; 32: 373–375.

9 Kato H. Regulation of functions of vascular wall cells by tissuefactor pathway inhibitor: basic and clinical aspects. ArteriosclerThromb Vasc Biol 2002; 22: 539–548.

10 Adams M. Novel considerations in the pathogenesis of the anti-phospholipid syndrome: involvement of the tissue factor pathwayof blood coagulation. Semin Thromb Hemost 2008; 34: 251–255.

11 Crawley JT, Lane DA. The haemostatic role of tissue factor path-way inhibitor. Arterioscler Thromb Vasc Biol 2008; 28: 233–242.


1482

Lupus

12 Lwaleed BA, Bass PS. Tissue factor pathway inhibitor: structure,biology and involvement in disease. J Pathol 2006; 208: 327–339.

13 Gando S, Kameue T, Morimoto Y, Matsuda N, Hayakawa M,Kemmotsu O. Tissue factor production not balanced by tissuefactor pathway inhibitor in sepsis promotes poor prognosis. CritCare Med 2002; 30: 1729–1734.

14 Gando S, Kameue T, Matsuda N, et al. Imbalances between thelevels of tissue factor and tissue factor pathway inhibitor in ARDSpatients. Thromb Res 2003; 109: 119–124.

15 Gando S, Nanzaki S, Morimoto Y, Ishitani T, Kemmotsu O.Tissue factor pathway inhibitor response does not correlate withtissue factor-induced disseminated intravascular coagulation andmultiple organ dysfunction syndrome in trauma patients. CritCare Med 2001; 29: 262–266.

16 Hochberg MC. Updating the American College of Rheumatologyrevised criteria for the classification of systemic lupus erythemato-sus. Arthritis Rheum 1997; 40: 1725.

17 Adams M, Breckler L, Stevens P, Thom J, Baker R, Oostryck R.Anti-tissue factor pathway inhibitor activity in subjects with anti-phospholipid syndrome is associated with increased thrombin gen-eration. Haematologica 2004; 89: 985–990.

18 Adams M, Donohoe S, Mackie I, Machin S. Anti-tissue factorpathway inhibitor activity in patients with primary antiphospholi-pid syndrome. Br J Haematol 2001; 114: 375–379.

19 Lean S, Ellery P, Ivey L, et al. The effects of tissue factor pathwayinhibitor and anti-b-2-glycoprotein-I IgG on thrombin generation.Haematologica 2006; 91: 1360–1366.

20 Hasselaar P, Derksen RH, Blokzijl L, et al. Risk factors for throm-bosis in lupus patients. Ann Rheum Dis 1989; 48: 933–940.

21 Rauch J. Lupus anticoagulant antibodies: recognition ofphospholipid-binding protein complexes. Lupus 1998; 7 Suppl 2:S29–S31.

22 Alarcon-Segovia D, Perez-Vazquez ME, Villa AR, Drenkard C,Cabiedes J. Preliminary classification criteria for the antiphospho-lipid syndrome within systemic lupus erythematosus. SeminArthritis Rheum 1992; 21: 275–286.

23 Petri M. Epidemiology of the antiphospholipid antibody syn-drome. J Autoimmun 2000; 15: 145–151.

24 Triplett DA. Laboratory diagnosis of lupus anticoagulants. SeminThromb Hemost 1990; 16: 182–192.

25 Martinuzzo M, Iglesias Varela ML, Adamczuk Y, Broze GJ,Forastiero R. Antiphospholipid antibodies and antibodies totissue factor pathway inhibitor in women with implantation fail-ures or early and late pregnancy losses. J Thromb Haemost 2005; 3:2587–2589.

26 Urbanus RT, Derksen RH, de Groot PG. Current insight intodiagnostics and pathophysiology of the antiphospolipid syndrome.Blood Rev 2008; 22: 93–105.

27 Greaves M. Antiphospholipid syndrome: state of the art withemphasis on laboratory evaluation. Haemostasis 2000; 30 Suppl2: 16–25.

28 Nojima J, Masuda Y, Iwatani Y, et al. Tissue factor expression onmonocytes induced by anti-phospholipid antibodies as a strongrisk factor for thromboembolic complications in SLE patients.Biochem Biophys Res Commun 2008; 365: 195–200.

29 Roldan V, Marco P, Fernandez C, Pascual E. Levels of tissuefactor pathway inhibitor in lupus patients correlate with lupusactivity and endothelial damage markers. Haematologica 2002;87: 1231–1232.

30 Ertenli I, Kiraz S, Celik IC, et al. Changes in the concentration anddistribution of tissue factor pathway inhibitor in Behcet’s diseaseand systemic lupus erythematosus: effect on the prethromboticstate. Ann Rheum Dis 2001; 60: 1149–1151.

31 Hansen JB, Huseby KR, Huseby NE, Ezban M, Nordoy A. Tissuefactor pathway inhibitor in complex with low density lipoproteinisolated from human plasma does not possess anticoagulant func-tion in tissue factor-induced coagulation in vitro. Thromb Res1997; 85: 413–425.

32 Hansen JB, Huseby KR, Huseby NE, Sandset PM, Hanssen TA,Nordoy A. Effect of cholesterol lowering on intravascular pools ofTFPI and its anticoagulant potential in type II hyperlipoproteine-mia. Arterioscler Thromb Vasc Biol 1995; 15: 879–885.

33 Maroney SA, Haberichter SL, Friese P, et al. Active tissue factorpathway inhibitor is expressed on the surface of coated platelets.Blood 2007; 109: 1931–1937.

34 Li A, Wun TC. Proteolysis of tissue factor pathway inhibitor(TFPI) by plasmin: effect on TFPI activity. Thromb Haemost1998; 80: 423–427.

35 Salemink I, Franssen J, Willems GM, et al. Factor Xa cleavage oftissue factor pathway inhibitor is associated with loss of anticoag-ulant activity. Thromb Haemost 1998; 80: 273–280.

36 Ohkura N, Enjyoji K, Kamikubo Y, Kato H. A novel degradationpathway of tissue factor pathway inhibitor: incorporation intofibrin clot and degradation by thrombin. Blood 1997; 90:1883–1892.

37 Adams MJ, Oostryck R. A comparative study of functional assaysfor tissue factor pathway inhibitor using normal plasma and clin-ical samples. Blood Coagul Fibrinolysis 2000; 11: 327–333.

38 Bognacki J, Hammelburger J. Functional and immunologic meth-ods for the measurement of human tissue factor pathway inhibitor.Blood Coagul Fibrinolysis 1995; 6(Suppl 1): S65–72.

39 Jeske W, Hoppensteadt D, Fareed J, Bermes E. Measurement offunctional and immunologic levels of tissue factor pathway inhib-itor. Some methodologic considerations. Blood Coagul Fibrinolysis1995; 6 Suppl 1: S73–S80.


1483

Lupus

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

koagulasi pada sle

Documents