Thrombosis Research 111 (2003) 311 – 315
Regular Article
Validation of a novel tissue factor assay in experimental $ human endotoxemia Claudia Marsik a,b, Peter Quehenberger b, Nigel Mackman c, Bjarne Osterud d, Thomas Luther e, Bernd Jilma a,* a
Department of Clinical Pharmacology, Vienna University School of Medicine Waehringer Guertel 18-20, A-1090 Vienna, Austria b Institute for Medical and Chemical Laboratory Diagnostics, Vienna University, Vienna, Austria c Department of Immunology and Vascular Biology, The Scripps Research Institute, CA, USA d Institute of Biochemistry, University of Tromso, Tromso, Norway e Institute of Pathology, University of Dresden, Dresden, Germany Received 5 December 2002; received in revised form 21 September 2003; accepted 24 September 2003
Abstract Background: Nuclear factor kappa B (NF-nB) activation and tissue factor (TF) expression may contribute to lethality in sepsis. Inappropriate in vivo expression of TF is likely responsible for fibrin deposition in sepsis-associated disseminated intravascular coagulation (DIC). Clinical assessment of TF expression has remained a major challenge. No point-of-care assays are currently available to measure the level of TF activity in whole blood. The current study examined the suitability of the TiFaCT assay as a point-of-care assay to detect TF expression. Methods: 30 healthy male volunteers received 2 ng/kg of LPS. Tissue factor-dependent coagulation was quantified with a novel assay called tissue factor clotting time (TiFaCT), and by measurement of activation markers of downstream coagulation. Results: Ex vivo addition of anti-TF antibodies to blood slightly increased clotting times at 0 – 24 h ( p < 0.01) indicating that some tissue factor activity was present in whole blood at any time. LPS bolus infusion decreased TiFaCT clotting time by 23% compared to baseline ( p < 0.01), when in vivo clotting increased, as demonstrated by a 10-fold increase in prothrombin fragment levels (F1 + 2). Ex vivo incubation with LPS considerably shortened TiFaCT (from 1000s to 400s as compared to control incubation; p < 0.01). This effect was blunted at 2 – 4 h after LPS infusion (i.e. the time of monocytopenia), but twofold enhanced 24 h after LPS challenge ( p < 0.01). Conclusions: In summary, the TiFaCT assay was validated in our in vivo model of LPS-induced coagulation. It detected minute quantities of circulating TF even at baseline. TiFaCT is shortened at times of in vivo thrombin generation. D 2003 Elsevier Ltd. All rights reserved. Keywords: Endotoxin; TiFaCT assay; Prothrombin fragment; D-dimer; Randomized controlled trial
1. Introduction Tissue factor, a membrane-bound procoagulant glycoprotein, is the initiator of the extrinsic clotting cascade, which is the predominant coagulation pathway in vivo. The expression of tissue factor (TF) by monocytes/macrophages leads to thrombin generation and contributes to their physiological and pathophysiological roles in wound repair, disseminated intravascular coagulation linked to sepsis, postoperative thrombosis, unstable angina, athero$ ¨ sterreichischen Supported in part by the Jubila¨umsfonds der O Nationalbank Project no. 8917. * Corresponding author. Tel.: +43-1-40400/2980; fax: +43-1-40400/ 2998. E-mail address:
[email protected] (B. Jilma).
0049-3848/$ - see front matter D 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2003.09.017
sclerosis, chronic inflammation and cancer. Regulation of TF expression in monocytes is controlled by the transcription factors nuclear factor kappa B (NF-nB) and AP-1 [1,3]. Lipopolysaccharide (LPS) or cytokines induce gene expression of TF by nuclear factor kappa B (NF-nB) [2]. Interestingly, NF-nB binding activity is higher in nonsurvivors than in survivors of sepsis. Along similar lines, high levels of TF expression are predictive of clinical outcome in meningococcal sepsis [4] and TF contributes to lethality in animal models of sepsis [5,6]. Tissue factor, the primary initiator of blood coagulation [1], forms a highly pro-coagulant complex with activated factor VII (FVIIa), which initiates the coagulation cascade during endotoxemia. Inappropriate in vivo expression of TF is likely responsible for fibrin deposition in
312
C. Marsik et al. / Thrombosis Research 111 (2003) 311–315
sepsis-associated disseminated intravascular coagulation (DIC). Despite increasing knowledge about the in vitro regulation of TF expression, clinical assessment of TF expression has remained a major challenge. Thus, no point-of-care assays are currently available to measure the level of TF activity in whole blood, which could help tailor effective anti-TF treatment strategies. Recently, a so-called tissue factor clotting time (TiFaCT) assay, which measures the effects of spontaneous and LPSinduced TF expression on fibrin formation, was introduced [7]. The LPS-induced TiFaCT clotting time is NF-nBdependent in vitro [7]. It is currently unknown whether this assay could be suitable to measure TF activity in humans. Hence, the current study aimed to characterize the functional properties of the newly developed TiFaCT assay [7] in a well-defined human endotoxin model [8], which induces TF-triggered coagulation [9,10].
2. Materials and methods 2.1. Study design The study was approved by the Institutional Ethics Committee. Written informed consent was obtained from all participants. 2.2. Study subjects Thirty healthy male volunteers were invited to participate in this trial. All subjects were 19– 35 years of age with a body mass index between the 15th and the 85th percentile. Determination of health status included medical history, physical examination, laboratory parameters, virological and standard drug screening. In addition, study subjects were tested for hereditary thrombophilia, i.e. factor V Leiden, protein C and S deficiency, to minimize potential risks imposed by endotoxin-induced coagulation activation. Exclusion criteria were regular or recent intake of medication including non-prescription medication, and relevant abnormal findings in medical history or laboratory parameters. 2.3. Study protocol The experimental procedures of the endotoxin infusion studies have been described in detail previously in other trials [11 – 13]. Briefly, volunteers were admitted to the study ward at 8:00 a.m. after an overnight fast, because the response to endotoxin varies with daytime. Throughout the entire study period, subjects were confined to bed rest and kept fasting for 8.5 h following LPS infusion. A 5% glucose infusion (Leopold Pharma, Vienna, Austria) was started at 8:30 a.m. and continued over 8.5 h at 3 ml/
kg/h to maintain adequate hydration. All subjects received a bolus of 2 ng/kg LPS i.v. (National Reference Endotoxin, E. coli; USP Convention, Rockville, MD 20852, USA). 2.4. Sampling and analysis Sampling times were selected based on the kinetics of coagulation effects seen in subjects challenged with LPS in previous trials and on the kinetics of TF up-regulation [9,13]. Blood samples were collected by repeated venipunctures into EDTA anticoagulated vacutainer tubes (Becton Dickinson, Vienna) before LPS infusion, and thereafter at times indicated in the figures (except leukocyte counts which were obtained from an indwelling venous line on the contralateral arm from where LPS had been administered). Plasma samples were processed immediately by centrifugation at 2000 g at 4 jC for 15 min and stored at 80 jC before analysis. TF expression was quantified by flow cytometry on a FACSCalibur flow cytometer (Becton Dickinson) as described previously [8]. The antiTF antibody (American Diagnostica, USA) was FITC-labeled. Results are presented as mean fluorescence intensity (MFI) and percentage of positive cells. Neutrophil counts were obtained with a cell counter (Sysmex, Milton Keynes, UK), and monocyte counts were estimated from FSS/SSC plot flow cytometry counts [13]. The following commercially available assays were used: prothrombin fragment F1 + 2 (Behring; normal value < 1.9 nmol/l); plasmin – antiplasmin (PAP) complexes (Enzygnost PAP micro, Behring; normal range: 120– 700 Ag/l), measuring plasmin activity; the fibrin split product D-dimer (Boehringer Mannheim; normal values < 400 ng/ml), which reflects fibrinolytic digestion of cross-linked fibrin. TF levels in human plasma were quantified as described previously [14]. 2.5. TiFaCT assay This assay has been described in detail previously [7]. Briefly it employs a Sonoclot Coagulation and Platelet Function Analyser (Sienco, Wheat Rige), which uses a disposable vibrating probe immersed in 300 Al of whole blood to measure the viscous drag of fibrin strands. Whole blood samples were collected in test tubes containing 0.5 ml of 3.2% sodium citrate and a corn trypsin inhibitor (CTI; 50 Ag/ml) to inhibit FXIIa (Haematologic Technology, Essex Junction, VT). Blood was immediately processed and incubated in plastic tubes supplied by Sienco for 10 min with and without two inhibitory antiTF antibodies (TF mAbs VD10 and VIC12 generated with native, factor VII affinity-purified TF as the immunogen [14,15]) and for 2 h with and without LPS. Previous in vitro assessment showed a shortened clotting time in LPS-stimulated blood. Inhibitory anti-TF antibodies prolonged the clotting time of LPS-stimulated
C. Marsik et al. / Thrombosis Research 111 (2003) 311–315
313
blood, indicating that the shortened clotting time was due to induction of TF expression [7]. 2.6. Data analysis Data are expressed as mean and the 95% confidence intervals for description in the text. Nonparametric statistics were applied. All statistical comparisons were done with the Friedman ANOVA and the Wilcoxon signed rank test for post hoc comparisons. A two-tailed p-value of < 0.05 was considered significant.
3. Results The purpose of the study was to evaluate a new whole blood clotting test for tissue factor in a human disease model. Previous publications [9] have demonstrated that TF is induced in human endotoxemia. 3.1. Regulation of tissue factor in humans 3.1.1. Plasma TF Baseline levels of plasma TF averaged 149 pg/ml (95% CI: 115 – 184) and did not significantly vary over time, although a trendwise increase was observed after 24 h (mean: 164; 95% CI: 126 –203, data not shown). 3.1.2. TiFaCT The TiFaCT assay measures tissue factor clotting time in whole blood and coagulation after re-stimulation with LPS in vitro [7]. 3.1.3. Spontaneous TiFaCT TiFaCT clotting time decreased in a time-dependent manner ( 23% compared to baseline, Fig. 1A), as in vivo thrombin formation increased (vide infra)—even though the number of circulating monocytes decreased (Fig. 1C). This effect occurred as early as 2 h, and peak activation, i.e. minimum clotting time, was measured after 6 h ( p < 0.001). Addition of anti-TF antibodies to blood ex vivo slightly but significantly increased clotting times at 0 –24 h ( p < 0.01). 3.1.4. TiFaCT after (re-)stimulation with LPS ex vivo When blood was incubated for 2 h with LPS ex vivo, clotting time substantially decreased when compared to unstimulated blood (Fig. 1B). Ex vivo responsiveness of whole blood to LPS decreased within 2 h, when circulating monocytes decreased. Interestingly 24 h after LPS, TiFACT clotting time was markedly shorter than at baseline ( 15% compared to baseline, p = 0.01). 3.1.5. Monocyte counts Monocyte counts fell to almost undetectable values 2 h after LPS infusion (Fig. 1C), and nearly recovered at 24 h in all study groups.
Fig. 1. Effects of endotoxin infusion on the tissue factor clotting time (TiFaCT) assay. Effects of i.v. LPS (2 ng/kg) in a modified TIFaCT assay. Tissue factor clotting time is measured after 10 min incubation with (up triangle)/without (circle) ex vivo added TF antibody (A, top) and after 2 h incubation with (circle, filled)/without (down triangle) ex vivo added LPS (B, middle). The monocytopenia (C, bottom) explains the reduced ex vivo inducible TF clotting time at 2 h after LPS challenge. Data are presented as mean ( F 95% CI).
3.2. Effect of LPS on systemic coagulation 3.2.1. TF expression As expected, neutrophils and monocytes became highly activated as measured by a threefold increase in CD11b expression; yet, mean TF fluorescence intensity of peripheral monocytes increased only by 30% (MFI 19.3 at baseline compared to 25.6 at 4 h CI: 2– 60%) at 4 h and returned to baseline levels at 24 h (data not shown). Twenty-four hours after LPS infusion, monocyte counts returned to baseline and TF expression was no longer elevated. 3.2.2. Coagulation To relate the changes in TiFaCT changes to in vivo thrombin generation, we also assessed downstream coagu-
314
C. Marsik et al. / Thrombosis Research 111 (2003) 311–315
Fig. 2. Effects of LPS on TF-induced coagulation. Effects of 2 ng/kg lipopolysaccharide (LPS) on LPS-induced coagulation (fibrin split product: D-dimer, F1 + 2. . . prothrombin fragment). All subjects received bolus infusions of 2 ng/kg LPS. Data are presented as mean F 95% CI.
lation factors in LPS-induced coagulation (Fig. 2). Plasma levels of F1 + 2 increased almost 10-fold at 4 h ( p < 0.001 vs. baseline, Fig. 2). D-dimer increased constantly and peaked 4-fold over baseline at 24 h ( p < 0.01 compared to baseline).
4. Discussion While advances in molecular biology have helped to elucidate the molecular mechanism of TF regulation in vitro, quantification of TF regulation in humans remains a major challenge due to limitations of the test systems. We used a well-established model of LPS-induced TF-dependent coagulation to test a new (TiFaCT) assay as to its suitability to quantify TF activity in systemic inflammation. In this model, plasma levels of TFPI do not change [12]. The purpose of the study was to evaluate a new whole blood clotting test for tissue factor in a human disease model. A previous publication [9] has demonstrated that TF is induced in human endotoxemia, and that inhibitors of tissue factor fully blunted LPS-induced coagulation [8,10]. Measurement of plasma TF and TF expression on circulating monocytes failed to demonstrate a clear-cut effect of
LPS infusion on TF levels. In contrast, the TiFaCT assay could have particular benefits. The TiFaCT assay measures the clotting time of whole blood samples. The current assay now includes corn trypsin inhibitor to antagonize FXIIa and thus contact activation of blood coagulation. This is one possible explanation for the prolongation in baseline TiFaCT (1025 s in our study, Fig. 1). However, it was more than what we expected from the preliminary results with CTI presented by Santucci et al. [7]. Thus, standardization of this novel assay should be addressed as early as possible before widespread use. After in vivo LPS challenge, TiFaCT started to decrease as early as 2 h with a maximum at 4 – 6 h ( 23% compared to baseline, Fig. 1A), which corresponds well with the kinetics of TF expression in vivo [9]. Hence, LPS infusion shortened TiFaCT concurrently with increased thrombin formation in vivo (Fig. 2), despite the substantial decrease in monocyte counts (Fig. 1C). Ex vivo addition of anti-TF antibodies to whole blood prolonged TiFaCT slightly at all times (Fig. 1A). This could indicate that at least part of the activation of coagulation in the TiFaCT system is induced by TF, potentially even by non-monocyte-associated TF such as TF-positive vesicles in plasma [16] at times of monocytopenia. However, this inhibition was rather incomplete, which can be explained by the fact that the TF antibody is added too late to effectively block the LPS-induced shortening of the clotting time. Downstream coagulation products have already been generated in vivo that are not inhibited. In any case, the whole blood clotting time measures a functional activation of the coagulation cascade but the anti-TF antibody cannot be used to show a TF dependence because of the delay in addition. Hence, the TiFaCT assay may best be described as a whole blood clotting assay and is different from a recalcification time due to the presence of corn trypsin inhibitor. TiFaCT was shortened relative to the 10-min incubation when blood was incubated at 37 jC for 2 h with saline, indicative of a small degree of in vitro activation of blood coagulation (Fig. 1). When blood was stimulated ex vivo by LPS during that 2 h, a pronounced shortening of TiFaCT was observed. Ex vivo LPS-stimulated TiFaCT was prolonged first when monocyte counts decreased at 2 h (Fig. 1C). Even though the estimated number of monocytes had decreased to < 1% of the initial monocyte counts, LPS-induced TiFaCT was still shorter as compared to 2 h incubation with saline. This suggests that the TiFaCT assay is highly sensitive to even minute quantities of TF-expressing monocytes. Interestingly, ex vivo LPS-stimulated TiFaCT was markedly shortened at 24 h as compared to baseline, potentially suggesting monocyte priming. It may be speculated that this could have clinical consequences and may help in part explain the increased incidence of thrombotic events following infectious diseases [17 – 19]. As demonstrated previously [7], TiFaCT clotting time is also prolonged by ex vivo addition of anticoagulants such as
C. Marsik et al. / Thrombosis Research 111 (2003) 311–315
heparin, low molecular weight heparin or hirudin. This suggests that the TiFaCT assay may have clinical utility in monitoring anticoagulant therapy, particularly when patients are switched from one anticoagulant to another. The TiFaCT assay has advantages over PT or aPTT because it can measure the net effect of anticoagulants present in the blood. The studies with endotoxin demonstrate that the TiFaCT assay can be used to detect TF-dependent shortening of the clotting time of whole blood by TF expressed in vivo. This might render the TiFaCT assay superior to evaluating levels of TF antigen in plasma. A limitation of our trial might be the missing control group. However, clotting times (without LPS incubation) returned to baseline after 24 h after LPS challenge, indicating that the observed changes were not merely due to chance. This study was an experimental approach to validate the TiFaCT assay, rather than a validation of the TiFaCT performance in the assessment of therapeutic anti-coagulation. Hence the lack of a control group is probably less important. In conclusion, the TiFaCT assay—validated in our in vivo model of LPS-induced coagulation—was highly sensitive to LPS and demonstrated increased responsiveness to ex vivo added LPS 24 h after in vivo endotoxin challenge. Acknowledgements We are indebted to Lena Carpenter, Coagulation Diagnostics, for the setup of the Sonoclot analyzers, and also for her invaluable administrative help, and to Christa Drucker for running the Sonoclot assays. References [1] Osterud B. Tissue factor: a complex biological role. Thromb Haemost 1997;78:755 – 8. [2] Mackman N. Lipopolysaccharide induction of gene expression in human monocytic cells. Immunol Res 2000;21:247 – 51. [3] Magdolen V, Albrecht S, Kotzsch M, Haller C, Burgle M, Jacob U, et al. Immunological and functional analyses of the extracellular domain of human tissue factor. Biol Chem 1998;379:157 – 65. [4] Osterud B, Flaegstad T. Increased tissue thromboplastin activity in monocytes of patients with meningococcal infection: related to an unfavourable prognosis. Thromb Haemost 1983;49:5 – 7.
315
[5] de Boer JP, Creasy AA, Chang A, Roem D, Brouwer MC, Eerenberg AJ, et al. Activation patterns of coagulation and fibrinolysis in baboons following infusion with lethal or sublethal dose of Escherichia coli. Circ Shock 1993;39:59 – 67. [6] Taylor FB, Chang A, Ruf W, Morrissey JH, Hinshaw L, Catlett R, et al. Lethal E. coli septic shock is prevented by blocking tissue factor with monoclonal antibody. Circ Shock 1991;33:127 – 34. [7] Santucci RA, Erlich J, Labriola J, Wilson M, Kao KJ, Kickler TS, et al. Measurement of tissue factor activity in whole blood. Thromb Haemost 2000;83:445 – 54. [8] Jilma B, Marsik C, Mayr F, Graninger MT, Taylor FB, Ribel MC, et al. Pharmacodynamics of active site-inhibited factor VIIa in endotoxin-induced coagulation in humans. Clin Pharmacol Ther 2002;72: 403 – 10. [9] Franco RF, de Jonge E, Dekkers PE, Timmerman JJ, Spek CA, van Deventer SJ, et al. The in vivo kinetics of tissue factor messenger RNA expression during human endotoxemia: relationship with activation of coagulation. Blood 2000;96:554 – 9. [10] de Jonge E, Dekkers PE, Creasey AA, Hack CE, Paulson SK, Karim A, et al. Tissue factor pathway inhibitor dose-dependently inhibits coagulation activation without influencing the fibrinolytic and cytokine response during human endotoxemia. Blood 2000;95: 1124 – 9. [11] Hollenstein U, Homoncik M, Stohlawetz PJ, Marsik C, Sieder A, Eichler HG, et al. Endotoxin down-modulates granulocyte colonystimulating factor receptor (CD114) on human neutrophils. J Infect Dis 2000;182:343 – 6. [12] Hollenstein U, Homoncik M, Knobl P, Pernerstorfer T, Graggaber J, Eichler HG, et al. Acenocoumarol decreases tissue factor-dependent coagulation during systemic inflammation in humans. Clin Pharmacol Ther 2002;71:368 – 74. [13] Pernerstorfer T, Hollenstein U, Hansen JB, Stohlawetz P, Eichler HG, Handler S, et al. Lepirudin blunts endotoxin-induced coagulation activation. Blood 2000;95:1729 – 34. [14] Albrecht S, Kotzsch M, Siegert G, Luther T, Grossmann H, Grosser M, et al. Detection of circulating tissue factor and factor VII in a normal population. Thromb Haemost 1996;75:772 – 7. [15] Albrecht S, Luther T, Grossmann H, Flossel C, Kotzsch M, Muller M. An ELISA for tissue factor using monoclonal antibodies. Blood Coagul Fibrinolysis 1992;3:263 – 70. [16] Casciola-Rosen L, Rosen A, Petri M, Schlissel M. Surface blebs on apoptotic cells are sites of enhanced procoagulant activity: implications for coagulation events and antigenic spread in systemic lupus erythematosus. Proc Natl Acad Sci U S A 1996;93:1624 – 9. [17] Blann AD, Lip GY. Relation between endothelial-cell activation and infection, inflammation, and infarction. Lancet 1997;350:293 – 4. [18] Meier CR, Jick SS, Derby LE, Vasilakis C, Jick H. Acute respiratorytract infections and risk of first-time acute myocardial infarction. Lancet 1998;351:1467 – 71. [19] Vallance P, Collier J, Bhagat K. Infection, inflammation, and infarction: does acute endothelial dysfunction provide a link? Lancet 1997;349:1391 – 2.
