Disturbances in the tissue plasminogen activator/plasminogen activator inhibitor (TPA/PAI) system in systemic lupus erythematosus

American Journal of Hematology 379-1 3 (1991)

Disturbances in the Tissue Plasminogen Activator/ Plasminogen Activator Inhibitor (TPAIPAI) System in Systemic Lupus Erythematosus Guillermo J. Ruiz-Arguelles, Alejandro Ruiz-Arguelles, Eduardo Lobato-Mendizabal, Felipe Diaz-Gomez, Elda Pacheco-Pantoja, Cristina Drenkard, and Donato Alarcon-Segovia Laboratorios Clinicos de Puebla (G.R.-A., A.R.-A., F.D.-G., E.P.-P.), Centro de Hematologia y Medicina lnterna de Puebla (G.R.-A., E.L.-M.), Puebla, lnstituto Nacional de la Nutricion Salvador Zubiran (C.D., D.A.-S.), Mexico City, Mexico

Increased thrombogenesis observed in systemic lupus erythematosus (SLE) is derived from multiple mechanisms, including: Enhanced coagulation factor VIII:VWf activity, lupus anticoagulants, anti-phospholipid antibodies, acquired deficiencies of natural anti-thrombotic mechanisms (protein C, protein S, anti-thrombin Ill), and impaired fibrinolytic mechanisms. We studied the fibrinolytic mechanisms of 18 patients with systemic lupus erythematosus, selected carefully to avoid other possible causes of abnormalities in the fibrinolytic activity. Despite the fact that the euglobulin lysis time in steady state was normal in all instances, disturbances in the tissue plasminogen activator/plasminogenactivator inhibitor (TPAIPAI) system were found in all SLE patients: TPA activity was undetectable in all cases, whereas it was above 0.4 IU/ml in a control group. In 72 percent of patients, the undetectable TPA activity was correlated with abnormally high PA1 activity; PA1 levels were normal in all members of the control group, their mean value being 0.74 versus 8.63 IU/ml for SLE patients (P < .01). Coagulation protein C deficiency was found in 3 patients (17%). Even though within normal range, fibrinogen levels were significantly higher in SLE than in normal controls (219 versus 192 mg/dl, P < .01) and plasminogen levels were significantly higher in SLE than in controls (117 versus 78.2%, P < .01). Cross-linked fibrin derivatives (D-D dimers) were negative in all patients with SLE. Sixty-eight percent of SLE patients had high levels of antiphospholipid antibodies, but no correlation with the disturbances of the TPAiPAl system was found. It is concluded that most patients with SLE display severe abnormalities in the TPNPAI anti-thrombotic system and that these abnormalities may be related to the lupus thrombophilia, apparently multifactorial in its origin. Key words: lupus, thrombosis, TPA, PAI, fibrinolysis

Acquired coagulation protein C, protein S, and antithrombin I11 deficiencies occur in some patients and are Patients with systemic lupus erythematosus (SLE) related to vaso-occlusive episodes [ 5 , 6 ] ;acquired prossuffer a thrombogenic state apparently derived from tacyclin deficiency has also been recorded in SLE [7], as multiple mechanisms, Imbalances of the natural pro- well as have other abnormalities in anticoagulant mechcoagulant and anti-coagulant activities have been re- anisms [8]. In addition to abnormalities in natural antiported in such patients in association to a hypercoagula- coagulants, changes in other natural anti-thrombotic ble state: Modifications in the coagulation factor VIII systems may be responsible for the thrombogenesis complex with elevated von Willebrand factor activity observed in these patients. Defective fibrinolysis in SLE have been described in the disease [ 11, the presence of the lupus anticoagulant (LA), a form of anti-phospholipid antibody (APLA) is actually related to thrombogenesis in Received for publication June 6, 1990; accepted December 20, 1990. vivo despite its behavior as anticoagulant in vitro [2-4]. Address reprint requests to Dr. Guillermo J. Ruiz-Arguelles, ConsultIn addition, acquired deficiencies of some natural anti- ant in Hematology, Laboratorios Clinicos de Puebla, Diaz Ordaz 808, coagulant mechanisms have also been found in SLE: 72530 Puebla, Pue. MCxico. 0 1991 Wiley-Liss, Inc. INTRODUCTION

10

Ruiz-Argiielles et al.

has been previously recorded and identified as another cause of thrombogenesis [ 1,9, lo]. Herein we report on the study of the fibrinolytic mechanisms of 18 patients with SLE, stringently selected to avoid other causes of abnormalities in them. Impaired fibrinolysis was found in every patient as follows: Tissue-type plasminogen activator (TPA) activity was undetectable in all patients whereas plasminogen activator inhibitors (PAI) were abnormally increased in 88%, and coagulation protein C was abnormally low in 17% of them. Fibrinogen, plasminogen, and cross-linked fibrin derivatives were also measured and found within normal limits. No relationship between the presence of APLA and the defective fibrinolysis was observed in these patients. Our results are consonant with the idea that the thrombophilia observed in patients with SLE (lupus thrombophilia) is multifactorial and that impaired fibrinolysis, dependent on disturbances of the TPAiPAI system, may account for at least part of it. MATERIAL AND METHODS Subjects Parents fullfilling criteria for the diagnosis of SLE [ 1 I] were prospectively entered in the study. All patients had inactive forms of SLE, that is, no new clinical signs of the disease had appeared in the last 3 months prior to the study, despite a decreased or unchanged corticosteroid regimen. The history of clinically apparent past vasoocclusive episodes was also recorded. A normal control group, matched in age and sex, was also included. Informed, written consent was obtained from all individuals. Subjects in both groups were eliminated if they were 1. 2. 3. 4.

5. 6. 7. 8. 9.

smokers. regular alcohol drinkers. oral contraceptive users. receiving salicylates, cloroquine, or cyclophosphamide. receiving oral anticoagulants or heparin. displaying an obvious infection. recently operated on (within 3 months) suffering diabetes, liver failure, malignant disease, or myocardial infarction. pregnant.

Studies The following studies of the fibrinolytic mechanisms were done to all subjects: Euglobulin lysis time, fibrinogen, levels coagulation protein C activity, cross-linked fibrin derivatives (D-D dimers) assessment, plasminogen activity, tissue-type plasminogen activator activity, rapid plasminogen activator inhibitor activity, and endothelial cell (type I) plasminogen activator inhibitor antigen.

Anti-phospholipid (cardiolipin) antibodies were also assessed. Euglobulin lysis time (ELT). It was evaluated according to the method of Milstone [12]. Normal values were considered above 180 min. Fibrinogen (FG). It was measured by 2 methods: Thermoprecipitation as described by Ruiz-Reyes and Jimenez-Vazquez [ 131 and coagulation by thrombin, i.e., the Clauss method [14]. Normal values have been reported in the range of 200-400 mgidl and 200-500 mgi dl, respectively. Coagulation protein c (PC). Functional anticoagulant PC levels were measured by the method of Francis and Seyfert [ 151, using the venom of the copperhead snake Agkistrodon contortrix to activate PC in the plasma sample and using an end-point chromogenic substrate (H-D-Lys (Cho)-Pro-Arg-pNA.2AcOH) system as indicator (Spectrozyme PCA, American Diagnostica, Greenwich, Connecticut). Deficiency was defined at levels below 60%, with normal range of 60-140%. D-D Dimers. Cross-linked fibrin derivatives (D-D dimers) were sought according to Rylatt [16], using a specific monoclonal antibody linked to latex particles in an agglutination test (Dimertest, American Diagnostica, Greenwich, Connecticut). Plasminogen (PLG). It was assessed by a chromogenic substrate (H-D-N 1e-HHT-Lys-pNA.2AcOH) assay, employing streptokinase as activator, according to Soria, Soria and Samama [ 171; (Actichrome PLG, American Diagnostica, Greenwich, Connecticut). Normal ranges are 60-140%. Tissue-type plasminogen activator activity (TPA). It was evaluated in unstimulated plasma, by the method of Chmielewska et al. [18], using a chromogenic substrate (CH3S02-D-CHT-Gly-Arg-pNA.AcOH) and physiologic fibrin stimulation. (SpectrolyseiDesafib, American Diagnostica, Greenwich, Connecticut). Normal ranges are 0.01-0.8 IUiml. Due to the great variability even in normal subjects, TPA release after venous occlusion was not assessed. Rapid plasminogen activator inhibitor activity (PAI). Using the same chromogenic substrate for TPA as end-point, the assay with physiologic fibrin stimulation was done simultaneously with the TPA assessment [ 181 (Spectrolyse/Desafib, American Diagnostica, Greenwich, Connecticut); PA1 normal ranges have been determined between M IU/ml. Plasminogen activator inhibitor type 1 (PAI-1). Its antigenic levels were assessed using a double antibody enzyme immunoassay (ELISA), using 2 different monoclonal antibodies, to detect both active (free) and inactive forms of PAI- 1, according to Urden and Blomback [ 191; (Imubind- 1 PAI-1, American Diagnostica, Greenwich, Connecticut). Normal values were 1 .53-3.87 ng/ml. Anti-phospholipid antibodies (APLA). Both IgG

Fibrinolysis Abnormalities in Lupus

11

1979, by Angles-Can0 et al. [l]. Deficiencies in plasminogen activators (PA) have been recorded as the most frequent fibrinolytic system abnormality in SLE patients, since the description of Angles-Can0 et al. [ 1,9,10,23271. Abnormalities in the TPNPAI system have been repeatedly described in SLE: Byron et al. reported that about one third of patients with SLE exhibit a decreased fibrinolytic potential [25]. Glas-Grenwalt et al. found that two-thirds of patients with SLE had abnormally low RESULTS levels of plasminogen activators and that the level of PA1 was significantly increased in 87% of them [23]. Moroz Table I summarizes the results of the 18 SLE patients et al. reported that patients with SLE have impaired and 11 normal controls studied. ELT in steady state was whole blood, but not serum-dependent, fibrinolytic acnormal in all patients with SLE and controls. It is tivity [26]. Awada et al. found decreased TPA and noteworthy that TPA activity was undetectable in all increased rapid PA1 activities in 16 patients with SLE; the patients with SLE, whereas it was above 0.4 IUiml in all extent of the disturbances in the TPA-PA1 system was normal controls, the mean levels of TPA were significorrelated with disease activity [lo]. Despite severe cantly lower in SLE than in the group of normal controls abnormalities in the fibrinolytic system, all patients (CO.01 versus 0.92 IU/ml, P < 0.01). In thirteen paincluded in our group had inactive forms of the disease as tients with SLE (72 percent) the undetectable TPA described above. activity was correlated with abnormally high levels of The mechanisms underlying the disturbances in the PA1 activity and in 8 of them (44 percent) the antigenic TPA/PAI system in SLE have not been recognized; in an levels of PAI- 1 were also increased. In 3 cases high levels effort to explain them, it has been speculated that they of PAI-1 antigen were found, despite normal values for could be related to the presence of the lupus anticoagulant PA1 activity, thus making a total of 16 cases (88 percent) (LA), to monocytelmacrophage activation with concomin which PA1 was abnormally high, either in its function itant elevation in procoagulant expression and monokine or its antigenic levels. PA1 levels were normal in all the secretion (interleukin 1 for example), or to the presence members of the control group: The mean levels of PA1 of anti-endothelial cell antibodies [24]. We have previactivity were significantly higher in the group of patients ously shown that APLA may interfere with coagulation with SLE than in the normal controls (8.63 vs. 0.74 IU/ ml, P < 0.01). Coagulation PC subnormal activity was PC activation at the endothelial cell level by blocking the found in 3 cases of SLE (17 percent) and it was normal action of thrombomodulin [6]; however, Cariou et al. in all the controls. Clottable FG levels were significantly have recently shown that APLA do not inhibit the higher in the group of SLE than in the normal controls endothelial release of TPA or PA1 [29]. Tsakiris et al. (219.0 vs. 192.0 mg/dl, P < 0.01). Plasminogen levels reported preliminary findings of a high correlation bewere significantly higher in the SLE group (1 17.0 versus tween the presence of LA and abnormally high levels of 78.2%, P < 0.01), however, the levels of FG, PLG, and PAI-1 [30], but Awada et al. found no correlation D dimers were either normal or negative in all SLE between APLA and abnormalities in the TPA/PAI fibrinpatients. Two patients (numbers 5 and 11) had clinically olytic system [lo]. We have not found any correlation apparent vaso-occlusive episodes more than 6 months between presence of APLA and impaired fibrinolysis due before the study; 13 patients had high levels of APLA. to changes in the TPA/PAI system. Our results confirm previous observations regarding the abnormalities of the TPAiPAI system in SLE. We have found undetectable activity of TPA in all patients DISCUSSION with SLE, using a sensitive assay, specific for TPA. Thromboembolic events in SLE, recognized since Glas-Grenwalt, with a less sensitive assay [23] found 1935 [21], may precede the full clinical picture of the decreased levels of TPA in 63% of patients with SLE. disease by months or even years. Vaso-occlusive epi- When more specific methods to assess TPA were used, sodes occur more frequently in patients with SLE than in persistently undetectable TPA activity has been recorded the general population, its overall incidence in large [lo]. Some papers describing low levels of activity of series being 4 to 5% [I]. Current knowledge of the TPA have also found increased activities of PAI: Glasso-called “lupus thrombophilia” points to the possibility Grenwalt et al. [23] found increased levels of an “inhibof its multifactorial origin, most likely derived from an itor of plasminogen activation” in 85% of SLE patients. acquired imbalance of the procoagulant and anti-coagu- Awada et al. [lo], using a similar assay to ours found PA1 lant systems. Impaired fibrinolysis has been identified in activities above 5 IU in 31% of patients with SLE; we autoimmune vasculitis since 1971 [22] and in SLE since have found such high levels in 72% of patients. Unlike and IgM isotypes were determined in serum by a solid phase immunoenzymatic assay (ELISA) using cardiolipin as antigen as previously described [20]. Known positive and negative sera were kindly provided by Dr. G.R.V. Hughes, London, for standardization. Normal values were determined in our laboratory as follows: IgG below 1.9 units and IgM below 2.4 units.

12

Ruiz-Arguelles et al.

TABLE 1. Results of the Assessment of the Fibrinolytic System of Both Patients With Systemic Lupus Erythematosus (SLE) and Normal Controls

Units Normal range SLE Controls P value

ELT

TPA

PA1

PAI- 1

PC

PLG

Fibrinogen

APLA IgG

APLA IgM

min

IUlml

IUiml

nglnil

%

%

nigldl

units

units

>I80

.o 1 - 3