Clot retraction of native whole blood in vitro was mentioned in the literature as early as 1772 (1). In contrast to most platelet function assays, such as aggregation, adhesion, and secretion studies, clot retraction measures a platelet-dependent variable that occurs only during clot formation. Clot retraction is typically assessed in vitro by measuring the volume of serum extruded from the clot or the decreases in the size of the clot mass (2). The characteristics of clot retraction have been used to monitor platelet function. The Hemodyne® hemostasis analyzer (Hemodyne Inc., Richmond, VA) assays platelet function by measuring the platelet contractile force generated during clot retraction (3,4). The Sonoclot® analyzer (Sienco, Morrison, CO) uses the clot retraction-induced change in the ultrasound signal to calculate platelet function (5). The Thrombelastograph® (TEG®; Haemoscope Corp., Niles, IL) is a coagulation monitor that provides qualitative information on clot formation and clot lysis (fibrinolysis).
The first aim of this study was to examine whether the decrease of TEG® amplitude reflects only fibrinolysis (6). We hypothesized that the TEG® amplitude would decrease when fibrin fibers connected to the cuvette walls were pulled away because of clot retraction and that the extent of the decrease would be proportional to platelet counts. Second, we examined the effects of small-dose tissue plasminogen activator (tPA) on TEG® tracings and d-dimer levels at different platelet counts to contrast retraction and fibrinolysis. The linkage force of fibrin-platelet bonding between a suspended pin and a cuvette wall can be disrupted by either plasmin-induced fibrinolysis or clot retraction mediated by the glycoprotein (GP)IIb/IIIa receptor. Accordingly, we hypothesized that the addition of abciximab, a GPIIb/IIIa inhibitor, would negate the platelet contribution to TEG® and help to distinguish clot retraction from lysis.
With approval of the IRB and written, informed consent, blood was obtained from 15 healthy volunteers with no history of aspirin ingestion over the preceding 2 wk or other medications that might interfere with platelet function. The whole-blood sample was collected into Vacutainer® (Becton Dickinson, Franklin Lakes, NJ) tubes containing 3.2% citrate (1:9 in volume). After incubation with prostaglandin E1 (final concentration, 500 ng/mL) for 10 min to minimize platelet activation by centrifugation, platelet-rich plasma (PRP) was prepared by centrifuging whole blood at 500g for 10 min at 22°C. PRP was then centrifuged at 3000g for 10 min to obtain platelet-poor plasma (PPP) and platelet pellet. The platelet pellet was diluted with PPP to make PRP with increasing concentrations of platelet at approximately 50, 100, 200, 600, and 1200 × 109/L by using Coulter® AC-T (Beckman Coulter, Miami, FL). Respective PRP samples were designated as P50, P100, P200, P600, and P1200, which represent thrombocytopenia (P50), near-normal count (P100–P200), and thrombocytosis (P600–P1200).
After 10 μL of 0.4 M CaCl2 and 5.5 μL of 1 mM adenosine-5′-diphosphate (ADP; final concentration, 15 μM) were added to TEG® cups, a six-channel TEG® was performed with 350 μL of PPP and PRP with different platelet counts. To investigate the contribution of GPIIb/IIIa receptors in clot retraction, an additional six-channel TEG® was performed simultaneously by using samples that were previously incubated with abciximab (Reopro®; Centocor, Malvern, PA) at a final concentration of 60 μg/mL for 5 min at room temperature.
TEG® analyses for both native and abciximab-treated samples were repeated in the presence of tPA (Sigma-Aldrich Inc., St. Louis, MO) at a final concentration of 0.1 μg/mL (40 IU/mL). The % decrease in amplitude to maximum amplitude (MA) was calculated at 30 and 60 min after MA was achieved (time points T30 and T60). The formula is as follows: % decrease = 100 × (MA − A)/MA, where A represents the amplitude at T30 or T60.
To investigate morphological (visual) sample changes over time after the addition of CaCl2 and ADP (final concentration 15 μM), native and abciximab-treated samples were also incubated in glass tubes at 37°C. Furthermore, to quantitatively measure the degree of fibrinolysis, if any, in the native samples with or without tPA, small aliquots were removed for d-dimer assay (Asserachrom D-Di; Diagnostica Stago, Asnieres, France), which is a specific marker of degradation of polymerized fibrin. d-Dimer level was measured at 60, 120, and 180 min after the start of incubation.
All data are expressed as mean ± sd. The Kruskal-Wallis H-test followed by the Mann-Whitney U-test with Bonferroni's correction was used to compare MA and the percentage decrease in amplitude among samples in each group. Wilcoxon's t-test was used to compare intergroup differences. P < 0.05 was considered significant.
Platelet counts for PPP and five PRPs were 2.4 ± 1.1, 50.9 ± 2.9, 103.1 ± 4.8, 212.6 ± 15.7, 632.5 ± 31.8, and 1272.3 ± 67.8 × 109/L, respectively. TEG® tracings of native and abciximab-treated samples are shown in Figure 1. In native samples, MA values increased with increasing platelet count, and there were statistical differences among PPP, thrombocytopenia (P50), near-normal count (P100–200), and thrombocytosis (P600–1200) samples (Fig. 2). In abciximab-treated samples, we found a statistically significant difference in MA only between PPP and P1200 (P < 0.05). There were significant differences in the percentage decrease in amplitude between PPP and all PRPs at 30 and 60 min after MA in the native group (P < 0.05 at 30 min after MA and P < 0.01 at 60 min after MA). In native samples of P200 or larger platelet count, the percentage decrease in amplitude at T30 and T60 nearly reached or exceeded the cutoff values of 7.5% and 15.0%, respectively, which in clinical practice would represent significant fibrinolysis (Fig. 3). However, in PRP treated with abciximab, percentage decreases in amplitude at T30 and T60 were smaller when compared with those in native PRP (P < 0.01), and there was a significant difference only between PPP and P1200 samples at T60. Abciximab did not change the percentage decrease in amplitude at T30 and T60 in native PPP.
TEG® tracings of native and abciximab groups with tPA are shown in Figure 4. Only native PPP and P50 showed a fibrinolysis (spindle-shaped) pattern, but samples with a larger platelet count were unaffected, and their TEG® tracings resembled those without tPA (Fig. 1A). The addition of tPA resulted in fibrinolysis patterns on TEG® in all abciximab-treated samples, regardless of the platelet count.
Morphological changes of PPP and PRP were seen after 15, 30, 60, and 120 min of incubation in glass tubes (Fig. 5A). Clots shrank, except in PPP, and the extent of shrinkage seemed more prominent with a larger platelet count. There were no morphological changes in abciximab-treated samples even after 120 min of incubation (Fig. 5B).
d-Dimer levels are shown in Table 1. d-Dimer levels of the native group did not exceed the normal range (<400 ng/mL) even after 180 min of incubation, and there was no significant difference in serial d-dimer levels at each platelet concentration. With the addition of tPA, d-dimer levels of the native group increased significantly, and the values exceeded the normal range. However, there was a trend toward a smaller increase of d-dimer concentration with increasing platelet count (Table 1).
We have demonstrated that activated platelets contribute to clot retraction, which resembles fibrinolytic patterns on TEG®. Contrary to the conventional interpretation of TEG® tracings (6), decreasing amplitudes after MA may not be simply interpreted as fibrinolysis. We simulated the effect of hyperactivity of platelets on TEG® by using PRP with increasing platelet concentrations. With a platelet count ≥200 × 109/L, clot retraction became more prominent, with an approximately 7.5% decrease in amplitude at T30 and 15% at T60. By abrogating the platelet-fibrin interaction with abciximab, the extent of decrease in amplitude lessened, and the changes were essentially comparable among different platelet concentrations (Fig. 3). Morphological changes in the glass tube demonstrated clot retraction and the effect of abciximab on clot retraction (Fig. 5). Fibrin gel was generated in the glass tube even in the presence of abciximab, but the size of the clot did not change (Fig. 5B).
Conversely, clot retraction was observed in native PRP, and this was more prominent with larger platelet count (Fig. 5A). Further, we measured d-dimer levels in native PRP to quantitate fibrinolysis. The result suggests that the progression of fibrinolysis does not occur for 180 minutes (normal range, <400 ng/mL; Table 1). Our data agree with the ex vivo perfusion circuit model, in which fibrinolysis was not observed because of a lack of endothelial surface, a major source of tPA (7). The lack of increases in d-dimer levels in native samples confirms the decrease in amplitude after MA observed on TEG® in the native PRP samples and reflects clot retraction rather than fibrinolysis. When d-dimer assays were repeated after a small dose of tPA (0.1 μg/mL) was added to native PRP, d-dimer levels were increased over the threshold range, 400 ng/mL, in all samples, but there was a trend toward a smaller increase of d-dimer concentration with increasing platelet count (Table 1). However, fibrinolytic TEG® patterns were observed only in PPP and PRP at 50 × 109/L (Fig. 4A).
d-Dimer is formed during fibrin degradation by plasmin after cross-linkage of fibrin polymers; therefore, our data indicate that tPA causes fibrinolysis regardless of platelet counts, although TEG® clot structures seem to be more stable against tPA treatment (Fig. 3A). We speculated that a clot formed by a complex of fibrin and activated platelets is less susceptible to fibrinolysis because 1) clot retraction makes the fibrin clot resistant to plasmin (8), 2) activated platelets release intracellular factor XIII and fibrinogen, thus adding strength to clot structure, and 3) more extensive thrombin generation with activated platelets leads to formation of factor XIIIa (9) and thrombin-activatable fibrinolysis inhibitor (10). Our experimental data with abciximab support these concepts. After treatment of samples with both tPA and abciximab, fibrinolysis patterns appeared on TEG® (Fig. 4B). Other investigators have also reported that abciximab augments the clot-inhibitory effects of tPA (11,12). Thus, our current data suggest that 1) platelets contribute to the stability of clot structure against fibrinolysis and that 2) underlying fibrinolysis can be unmasked by the in vitro addition of abciximab.
Technical errors should also be considered for interpretation of decreases in amplitude. With the use of disposable plastic cups, artifacts caused by contamination are unlikely (13). A strong impact or vibration to the TEG® may cause disruption of clot (13), but our TEG® devices were located on a steady surface away from any source of vibration.
Conventionally, tests for clot retraction have been performed to detect abnormal platelet numbers, abnormal platelet GP structure, or abnormal fibrinogen level/function. Although it remains controversial, clot retraction is thought to have physiological roles in promoting recanalization of vessels after appropriate hemostasis and in approximating the sides of a wound. Several investigators have shown the multiple roles of platelets in clot retraction, including enhancement of fibrin polymerization (14), increasing clot resistance to fibrinolysis (8), increasing the cross-link density of fibrin gels (15), and transmitting contractility force to fibrin fiber (16). The contractility force is transmitted via platelet GPIIb/IIIa receptors that bind the polymerized fibrin network to the platelets' actin cytoskeleton, which is the major contributor to clot strength. Thus, GPIIb/IIIa receptor antagonists (e.g., abciximab) prevent platelets from exerting contractile force to the fibrin network. However, as shown in Figure 2, decreasing amplitude after MA was still noted even in the presence of abciximab, particularly with the largest platelet count (P1200) at 60 minutes after MA. We used the supraclinical levels of abciximab (60 μg/mL), which were three times larger than those used in Khurana et al.'s (17) study (20 μg/mL), for a platelet count of 400 × 109/L. Incomplete inhibition might result from relative scarcity of abciximab to excess GPIIb/IIIa receptors or exocytosis of GPIIb/IIIa receptors by thrombin-induced platelet activation (18).
In summary, our data indicate that a decrease in TEG® amplitude does not always reflect fibrinolysis and may represent platelet-mediated clot retraction. Further, we have demonstrated that abciximab-modified TEG® may be suitable in distinguishing clot retraction from low-grade fibrinolysis on TEG®.
1. Owens CA Jr. Clot retraction. In: Nichols WL, Bowie EJW, eds. A history of blood coagulation. Minneapolis: Mayo Foundation for Medical Education and Research, 2001:201–2.
2. McCabe-White M, Jennings L. Laboratory evaluation of platelet function. In: Platelet protocols. San Diego: Academic Press, 1999:57–9.
3. Carr ME Jr. Measurement of platelet force: the Hemodyne hemostasis analyzer. Clin Lab Manage Rev 1995;9:312–4, 316–8, 320.
4. Carr ME Jr, Zekert SL. Measurement of platelet-mediated force development during plasma clot formation. Am J Med Sci 1991;302:13–8.
5. Hett DA, Walker D, Pilkington SN, Smith DC. Sonoclot analysis. Br J Anaesth 1995;75:771–6.
6. Kang Y, Lewis JH, Navalgund A, et al. Epsilon-aminocaproic acid for treatment of fibrinolysis during liver transplantation. Anesthesiology 1987;66:766–73.
7. Wachtfogel YT, Kucich U, Hack CE, et al. Aprotinin inhibits the contact, neutrophil, and platelet activation systems during simulated extracorporeal perfusion. J Thorac Cardiovasc Surg 1993;106:1–9.
8. Kunitada S, FitzGerald GA, Fitzgerald DJ. Inhibition of clot lysis and decreased binding of tissue-type plasminogen activator as a consequence of clot retraction. Blood 1992;79:1420–7.
9. Chandler WL, Patel MA, Gravelle L, et al. Factor XIIIA and clot strength after cardiopulmonary bypass. Blood Coagul Fibrinolysis 2001;12:101–8.
10. Nesheim M, Wang W, Boffa M, et al. Thrombin, thrombomodulin and TAFI in the molecular link between coagulation and fibrinolysis. Thromb Haemost 1997;78:386–91.
11. Huang TC, Jordan RE, Hantgan RR, Alevriadou BR. Differential effects of c7E3 Fab on thrombus formation and rt-PA-Mediated thrombolysis under flow conditions. Thromb Res 2001;102:411–25.
12. Collet JP, Montalescot G, Lesty C, Weisel JW. A structural and dynamic investigation of the facilitating effect of glycoprotein IIb/IIIa inhibitors in dissolving platelet-rich clots. Circ Res 2002;90:428–34.
13. Caprini JA, Traverso CI, Arcelus JI. Perspectives on thromboelastography. Semin Thromb Hemost 1995;21:91–3.
14. Cohen I, Gerrard JM, White JG. Ultrastructure of clots during isometric contraction. J Cell Biol 1982;93:775–96.
15. Pollard TD, Fujiwara K, Handin R, Weiss G. Contractile proteins in platelet activation and contraction. Ann N Y Acad Sci 1977;283:218–36.
16. Morgenstern E, Korell U, Richter J. Platelets and fibrin strands during clot retraction. Thromb Res 1984;33:617–23.
17. Khurana S, Mattson JC, Westley S, et al. Monitoring platelet glycoprotein IIb/IIIa-fibrin interaction with tissue factor-activated thromboelastography. J Lab Clin Med 1997;130:401–11.
18. Gawaz M, Ruf A, Pogatsa-Murray G, et al. Incomplete inhibition of platelet aggregation and glycoprotein IIb-IIIa receptor blockade by abciximab: importance of internal pool of glycoprotein IIb-IIIa receptors. Thromb Haemost 2000;83:915–22.