Thromboelastography has emerged as an important method for evaluating hemostasis in several clinical settings (1–7) and has been used to guide transfusion therapy (3). To mechanistically determine the contribution of platelets and soluble components of the coagulation pathway to clot strength, various modifications of thromboelastographic samples have been characterized (4–7). Rather than separate blood into platelet-rich plasma and platelet-poor plasma, the inhibition or enhancement of platelet-fibrin interaction via the platelet integrin GPIIb/IIIa in whole blood has been used to determine the contribution of platelets to clot strength (4–7). In particular, the contribution of platelets to clot strength has been elucidated by the inhibition of GPIIb/IIIa function with a monoclonal antibody fragment (c7E3 Fab, ReoPro) (4–7) or by inhibition of platelet actin polymerization with cytochalasin D (7). The clot strength observed after such platelet inhibition has been attributed to the soluble components of the coagulation pathway (4–7). The addition of tissue factor (TF) to thromboelastographic samples has also been recently advocated (6,7), as there is considerable patient variability in the number and function of GPIIb/IIIa receptors (8,9), and TF-mediated increases in thrombin result in highly reproducible maximum stimulation of platelet function (7). Although in its infancy, the paradigm of evaluating the components of clot strength by modulation of GPIIb/IIIa receptor function will likely be applied to several clinical settings in the near future.
Although the modulation of GPIIb/IIIa receptor function in the clinical arena may provide additional mechanistic insights into hemostatic disorders, there is a need for an animal model wherein hypotheses concerning the etiology and treatment of coagulopathic states may be rigorously tested. The hemostatic responses of this animal model should closely mimic those of the human model (1–7). Consequently, the purpose of our study was to develop and characterize a rabbit model of hemostasis wherein the contribution of platelet function to clot strength would be determined.
The study was approved by our animal review committee. Conscious New Zealand White rabbits (2–3 kg) were briefly restrained (<2 min) once a week and had 4 mL of blood aseptically drawn from central ear arteries. Within 1 min, 350 μL of blood and 10 μL of various solutions characterized below were placed into a disposable cup and inserted into a Thrombelastograph® Coagulation Analyzer (Haemoscope, Skokie, IL) that was computer controlled. The proper functioning of the Thrombelastograph® was confirmed daily with quality control standards purchased from Haemoscope. The following thromboelastographic variables were measured for each sample for a 1-h period at 39°C (the normal temperature of the rabbit): reaction time (R, min), coagulation time (K, min), angle (α, degrees), maximum amplitude (MA, mm), and G (dyne/cm2). A detailed description of the methodology of thromboelastography has been presented in great detail elsewhere (3,4,7,10). In brief, R is defined as the time from when the blood sample is placed into the thromboelastograph cuvette until initial fibrin formation occurs as noted by a signal of 2-mm amplitude. K, a measure of the speed at which a clot forms with a certain viscoelastic strength, is defined as the time from R until the amplitude of the thromboelastographic signal is 20 mm in amplitude. α is the angle formed from R to the inflection point of the thromboelastographic signal as clot strength stabilizes; it is a measure of the speed of clot formation. MA is the largest amplitude of the thromboelastographic signal and is a measure of clot strength. A diagram of a prototypical thromboelastogram illustrating R, K, α, and MA is depicted in Figure 1. Finally, G (dyne/cm2) is a measure of clot strength (10) calculated from MA as follows: G = (5000 × MA)/(100 − MA). The relationship between MA and G is curvilinear. As MA varies from 0 to 100, G concordantly varies from 0 to infinity. Given this relationship, it is conceptually and statistically important to express clot strength as G (7).
To determine the concentration of TF required to maximally stimulate rabbit platelet function, four concentrations of TF were added to of blood samples placed in the Thrombelastograph® (n = 6). The concentrations of TF used were designed to determine a dose that would decrease R to 2–3 min, as this has been associated with maximal G values in humans (7). Recombinant human TF (Ortho Diagnostic Systems, Inc., Raritan, NJ) with an international sensitivity index of 1.0 was diluted in 0.9% NaCl so that the addition of 10 μL of the diluted solution to 350 μL of blood resulted in final concentrations of 0%, 0.025%, 0.0125%, or 0.00625% TF. Similarly, to determine the concentration of ReoPro (2 mg/mL) required to maximally inhibit rabbit platelet function, 10 μL (n = 1) and 30 μL (n = 1) of ReoPro was added to 350 μL and 330 μL of blood, respectively. Finally, the concentration of cytochalasin D required to maximally inhibit rabbit platelet function was determined by adding 10 μL of diluted cytochalasin D in 0.9% NaCl to 350 μL of blood that resulted in a final concentration of 0, 1, 5, or 10 μM cytochalasin D (n = 6). To confirm that 10 μM was the concentration required to maximally inhibit platelet function, blood was removed from another group of rabbits (n = 6) and exposed to either 10 μM or 15 μM cytochalasin D. The final concentration of dimethylsulfoxide in each sample exposed to cytochalasin D was 0.28% (vol/vol).
Based on the results of the preceding dose-response studies, a series of rabbits (n = 12) had blood sampled for thromboelastographic analysis to which was added 10 μL of 0.9% NaCl, TF (final concentration 0.00625%), or cytochalasin D (final concentration 10 μM). In addition to measuring R, K, α, MA, and G, the separate contribution of platelets and the soluble components of the coagulation pathway to G were calculated as follows in this series of animals. Total G (GT) = G caused by platelet function (GP) + G caused by the soluble components of coagulation (GSC). GSC was defined as the G remaining after maximal platelet inhibition with cytochalasin D. The method for defining the relative % of each component of clot strength is as follows: (GSC/GT) × 100 = % of GT attributable to GSC; (GT − GSC/GT) × 100 = % of GT attributable to GP. Finally, blood samples had concurrent platelet concentrations and hematocrit determined with a Sysmex K-800 (TOA Medical Electronics Co., LTD, Kobe, Japan).
All variables were expressed as mean ± sd. With regard to thromboelastography, it was decided a priori that blood samples that did not clot were to be assigned an R value of 60 min, a K value of 60 min, an α value of 0°, an MA value of 0 mm, and a G value of 0 dyne/cm2. Analyses of the effects of differing doses of TF and cytochalasin D on all thromboelastographic variables were conducted with one-way analysis of variance. Comparisons between the effects of TF and cytochalasin D on whole blood thromboelastographic variables were conducted with one-way analysis of variance. Post hoc analysis was conducted with the Tukey test. Comparison of GP in the presence or absence of TF was performed with paired Student’s t-tests. Linear regression was performed to determine associations between platelet concentrations and GP. An α error of ≤ 0.05 was considered significant.
Both TF and cytochalasin D significantly modified thromboelastographic variables in the rabbit (Tables 1 and 2). All doses of TF significantly decreased R and K while increasing α, MA, and G. There were no significant differences in thromboelastographic variables among the 0.00625%, 0.0125%, or 0.025% doses of TF. However, as the 0.00625% dose was observed to have an R of 2–3 min, this dose was used in subsequent experimentation. Cytochalasin D dose significantly increased R and K while decreasing α, MA, and G in a dose-dependent fashion. In the second experimental series wherein rabbit blood was exposed to either 10 μM or 15 μM cytochalasin D, the G values were 1040 ± 272 and 962 ± 261, respectively (P > 0.05). As 10 μM cytochalasin D was associated with maximal inhibition of clotting activity, this concentration was used in subsequent experimentation. Finally, neither dose of ReoPro significantly affected any thromboelastographic variable (data not shown), so no further experimentation was performed using the antibody fragment.
The thromboelastographic variables R, K, α, MA, and G of the second series of animals are depicted in Table 3. TF and cytochalasin D affected hemostasis in a fashion similar to that observed in the first series of experiments. The only exception to this trend was that there was no significant difference in R values between blood samples exposed to 0.9% NaCl or cytochalasin D. In the absence of TF, GP and GSC accounted for 87% and 13% of GT, respectively. After exposure to TF, GP more than doubled, accounting for 94% of GT. Although all rabbits had larger GP values after TF stimulation, there was up to a four-fold difference in the individual rabbit response to TF. The platelet count and hematocrit of this series of experiments were 403 ± 70 × 103/mm3 and 37% ± 2%, respectively. Platelet count was not significantly associated with GP in the presence (R2=0.03, P > 0.05) or absence (R2=0.17, P > 0.05) of TF.
The hemostatic response of rabbit whole blood to the inhibition or stimulation of platelet function is remarkably similar to that of humans (7). Cytochalasin D (10 μM) decreased GT to the same extent (approximately 90%) in rabbit blood that is observed in human blood (7). It is notable that ReoPro did not affect rabbit platelet function, probably secondary to antigenic differences of the GPIIb/IIIa receptor between the two species. Stimulation of rabbit platelet function with TF in whole blood resulted in a two-fold increase in GT, similar to that observed in human samples (7). It is of interest that GP was not significantly associated with whole blood platelet concentration in the presence or absence of TF in the rabbit. In individual humans, GT remains relatively constant as platelet count is decreased in the presence of equivalent concentrations of the soluble components of the coagulation pathway (7) until a critical concentration of platelets is reached (approximately 100,000/mm3). Thereafter, GT decreases precipitously, as platelet count is decreased (7). As in humans, the primary determinant of GP in the rabbit in a normal homeostatic state is the particular subject’s platelet responsiveness to stimulation (e.g., TF).
The rabbit paradigm of hemostasis presented may aid in gaining mechanistic insight into pathoetiologies of coagulation. For example, if GP is remarkably different between two experimental groups in the presence of similar platelet concentrations, one may be able to make inferences concerning the functional state of the platelets. Similarly, if platelet concentrations between groups are different but GP values are equivalent, one may ascertain differences in platelet function. Relative differences in the contribution of platelets and coagulation cascade proteins to GT have been discerned as demonstrated in pregnant patients (4) whose GSC accounted for 18% of clot strength. It is important to note that, unlike G values, one cannot simply subtract MA values from one another to determine the contributions of platelets and coagulation proteins to overall clot strength. G values are defined with absolute viscoelastic units (dyne/cm2), whereas MA values are not. A clot with an MA of 70 mm (GT = 11,667 dyne/cm2) is more than 4 times as strong as one with an MA of 35 mm (GT = 2692 dyne/cm2). Consequently, whereas inhibition of platelet function may aid in discerning which factors may contribute to clot strength, the reader must carefully note what units are used to express clot strength.
Although one may infer changes in platelet function by measuring GP and platelet concentration, which soluble component of the coagulation pathway should to be used in interpretation of changes in GSC remains an area of controversy. Fibrinogen concentration has been demonstrated to correlate to MA in pregnant patients after inhibition of platelet function with ReoPro, but whether this relationship persists in other patient populations or pathologic states remains to be elucidated. Another important issue is determining how decreases in the concentration of coagulation proteins affect clot strength in the presence of equivalent concentrations of platelets. Further, the contribution of coagulation proteins other than fibrinogen could profoundly affect GSC. These important issues, although beyond the scope of this study, are worthy of further investigation.
It is of interest that, in the first series of rabbits (n = 6) used to determine the dose response of G to cytochalasin D, it was noted that R increased in a dose-dependent fashion, whereas in the group of animals comparing the effects of TF and cytochalasin D (n = 12), there was no difference in R noted between the blood samples exposed to cytochalasin D or saline with DMSO. Further, in the series of experiments wherein rabbit blood was exposed to either 10 or 15 μM cytochalasin D (n = 6), the R values were not different; further, they were similar to that observed in the experimental series with n = 12. Consequently, the differences in R between the first study with cytochalasin D and the subsequent two series of experiments may reflect individual rabbit variability in the first series and the greater statistical power (higher n) of the subsequent studies. Most importantly, the primary endpoint of cytochalasin D-mediated platelet inhibition is a reliable, maximal decrease in G. To that end, the present study was successful.
In conclusion, our study presents a rabbit model that clarifies the contribution of platelet function to clot strength. The precise combination of inhibition/stimulation required for future studies of hemostasis in pathologic states (e.g., ischemia-reperfusion injury, hemorrhagic shock) remains to be determined. It may be of benefit not to stimulate platelet function with TF if an increase in platelet function is the expected result (e.g., a thrombotic state). On the contrary, if a decrease in platelet function is likely (e.g., presence of increased nitric oxide), it may be important in assessing changes in platelet function by comparing the relative increases in GP between groups after TF-mediated stimulation. Given the hemostatic similarities between rabbits and humans, the rabbit paradigm of hemostasis may serve as a clinical surrogate when investigating the effects of pathologic states on the contribution of platelet function to clot strength.
1. Miller BE, Bailey JY, Mancuso TJ, et al. Functional maturity of the coagulation system in children: an evaluation using thrombelastography. Anesth Analg 1997; 84:745–8.
2. Gan TJ, Bennett-Guerrero E, Phillips-Bute B, et al. Hextend, a physiologically balanced plasma expander for large volume use in major surgery: a randomized phase III clinical trial. Anesth Analg 1999; 88:992–8.
3. Lesserson LD, Manspeizer HE, DePerio M, et al. Thromboelastography guided transfusion algorithm reduces transfusion in complex cardiac surgery. Anesth Analg 1999; 88:312–9.
4. Gottumukkala VNR, Sharma SK, Philip J. Assessing platelet and fibrinogen contribution to clot strength using modified thromboelastography in pregnant women. Anesth Analg 1999; 89:1453–5.
5. Gottumukkala VNR, Sharma SK, Philip J, et al. Assessment of platelet function using modified thromboelastography in women with severe preeclampsia and thrombocytopenia [absract]. Anesthesiology 1999;SOAP suppl:A2.
6. Morewood GH, Jobes DR, Siegel MS. Changes in platelet function and fibrinogen during cardiopulmonary bypass as detected by modified thromboelastography (MTEG) [abstract]. Anesthesiology 1999; 91:A507.
7. 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.
8. Niiya K Hodson E, Bader R, et al. Increased surface expression of the membrane glycoprotein IIb/IIIa complex induced by platelet activation: relationship to the binding of fibrinogen and platelet aggregation. Blood 1987; 72:475–83.
9. Jackson CW, Jennings LK. Heterogeneity of fibrinogen receptor expression on platelets activated in normal plasma with ADP: analysis with flow cytometry. Br J Haematol 1989; 72:407–14.
© 2000 International Anesthesia Research Society
10. Hartert H, Schaeder JA. The physical and biologic constants of thromboelastography. Biorheology 1962; 1:31–9.