Thromboembolic events secondary to hypercoagulability are a major cause of morbidity and death in dogs.1–4 Respiratory distress caused by pulmonary thromboemboli as well as acute development of arterial thromboemboli are being reported with increasing frequency with availability of advanced imaging for veterinary patients.1,2,4 Hypercoagulability may occur because of increased coagulation factor activity, decreased endogenous anticoagulants, decreased fibrinolysis,5 or, as in most cases, a combination of these factors. Arterial events such as ischemic stroke or aortic thromboembolism as well as venous thrombosis have potentially life-threatening consequences; however, testing to detect this risk remains limited in veterinary medicine.1,2,4,6
Standard coagulation tests such as PT and APTT quantify time to clot formation, but lack sensitivity for decreased coagulation times in hypercoagulable states.7 Quantification assays of individual components of the coagulation system such as antithrombin, thrombinantithrombin complexes, or D-dimer may help to characterize hypercoagulability, but it is unclear whether these assays are correlated with thrombotic events in veterinary medicine.7–10 This lack of correlation may in part be attributable to the vast capacity of the coagulation system to compensate for deficiencies or excesses of its components. To overcome this, global coagulation assays such as TEG and thrombin generation assays have been developed, giving an overall impression of the state of the coagulation system, with particular sensitivity for the detection of hypercoagulable states.11–13,a However, the routine clinical use of these tests is currently limited by their high cost and limited availability.
The OHP evaluates fibrin generation and lysis over time.14–16 Originally developed in 1990 for use with human plasma,18 this assay has never been used with canine plasma, to the authors' knowledge. The OHP can be used to detect hypercoagulability in human plasma by calculating the area under a fibrin generation curve, or OCP curve, as well as assessing fibrinolysis.5,14,16 In humans, the OHP is used to identify hypercoagulable states in situations associated with increased thrombotic risk including pregnancy, preeclampsia, diabetes, antiphospholipid antibody syndrome, and stroke and in vascular surgery patients.5,14,16 The objectives of the study reported here were to determine whether the OHP assay is suitable for use in canine plasma, to optimize the assay, and to formulate reference intervals from samples obtained from healthy dogs.
Materials and Methods
The trial was approved by the Director General's Animal Ethics Committee. Signed client consent was obtained.
Dogs—Healthy dogs evaluated at North Shore Veterinary Specialist Centre for general health status or those in the boarding facility that had no history of illness, no abnormal findings with clinical examination, and APTT and PT in the reference range were eligible for inclusion in the study. Dogs were excluded if they were returned to the hospital for illness within 4 weeks after sampling, to prevent the effect of occult disease on reference intervals.
Of the 53 dogs recruited, 5 were excluded because of sample lipemia or hemolysis, and 8 were used as donors for pooled plasma. The remaining 40 dogs were used to establish reference intervals. This population comprised 16 mixed-breed dogs, 5 Bichon Frise, 3 Cavalier King Charles Spaniels, 2 Labrador Retrievers, 2 American Staffordshire Terriers, 2 Jack Russell Terriers, 1 Border Collie, 1 Boxer, 1 Cocker Spaniel, 1 Golden Retriever, 1 German Shepherd, 1 German Short-haired Pointer, 1 Italian Greyhound, 1 Australian Kelpie, 1 Maltese, and 1 Australian Silky Terrier, with a median age of 5.5 years (range, 0.3 to 13 years). Sex was evenly distributed, with 18 spayed females, 1 sexually intact male, and 21 neutered males.
Blood samples—Blood samples were collected by gentle jugular venipuncture, into a 5-mL syringe. Samples were immediately transferred into a tube containing sodium citrateb and gently mixed. Each sample was centrifuged (2,500 × g for 10 minutes) at room temperature (22°C) within 30 minutes of collection. A second centrifugation of the supernatant was performed under the same conditions. The resulting PPP was removed and stored in 500-μL aliquots at −80°C until analysis was performed within 3 months after preparation. Sample stability for < 3 months has been established.c Plasma from 8 healthy adult dogs (5 neutered males and 3 spayed females) with a median age of 4 years (range, 1.75 to 14 years) was pooled. Breeds included Labrador Retriever (n = 2), American Staffordshire Terrier (2), and Cocker Spaniel, Bichon Frise, Cavalier King Charles Spaniel, and mix (1 each). This pooled plasma was used as a control sample during all assay runs.
Standard coagulation assays—Double-spun citrated PPP was used for all tests. Fibrinogen concentration was measured with the Clauss method.18 Coagulation tests (PTc and APTTd) were performed on an automated analyzer.e Results of these tests were within reference ranges for all dogs included in the calculation of OHP reference intervals.18
OHP assay—Initial optimization studies were performed with plasma from 2 staff-owned dogs. Assays were performed in flat-bottom microplates as described.a Buffer (Tris 66mM, NaCl 130mM, and CaCl2 35mM) was adjusted to pH 7.0, filtered, and stored at 4°C. Stock solutions of bovine thrombin (100 U/mL)f and human r-tPAg (1 mg/mL) were diluted 1:100 and 1:10, respectively. Thrombin (0.006 U/mL) was added to the buffer used for the OCP curve acquisition. Optimal thrombin concentration was determined by serial dilutions including 0.03 U/mL, 0.01 U/mL, and 0.003 U/mL, analyzed in duplicate. Thrombin (0.006 U/mL) and r-tPA (700 ng/mL) were both added to the buffer for the OHP curve. Seventy-five microliters of PPP and 75 μL of the respective buffer were added to each well. Optimal absorbance was obtained at 390 nm by testing over a spectrum of wavelengths from 300 to 900 nm. Absorbance data were recorded every minute for 60 minutes with an automated plate readerh operating at 37°C. Samples were run in duplicate, and fibrin time curves were generated. Pooled human plasma curves were obtained from a published study.5
Variables measured included OHP and OCP (area under respective curves), overall fibrinolysis potential (calculated from formula [OCP – OHP]/[OCP × 100%]), Max OD, Max slope, and lag time in onset of fibrin generation (Figure 1). All outputs were expressed as the mean of duplicate samples run simultaneously. Acceptance criteria for the assay were absence of hemolysis and lipemia and generation of each of the outputs during the 1-hour time frame.
Graphic representation of outputs generated by an OHP assay evaluated for use in dogs. The OFP is calculated as (OCP – OHP)/(OCP × 100).
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Graphic representation of outputs generated by an OHP assay evaluated for use in dogs. The OFP is calculated as (OCP – OHP)/(OCP × 100).
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Graphic representation of outputs generated by an OHP assay evaluated for use in dogs. The OFP is calculated as (OCP – OHP)/(OCP × 100).
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Statistical analysis—Prothrombin time, APTT, fibrinogen concentration, and age were log transformed, when necessary, to attain an acceptably normal distribution. Log transformations were applied to OHP, OCP, Max OD, and Max slope, and OFP was raised to the power 3. The Pearson product moment correlation coefficients (Pearson r) and associated P values were calculated between these variables and the OHP outputs, also transformed as necessary. The association between sex and OHP outputs was examined by use of a Welch 2-sample t test.i
Pooled canine plasma was included in each of 16 plates for calculation of interassay CV. Sixteen sample wells of pooled normal plasma were tested in a single assay to calculate the intra-assay CV. A medium-level CV was calculated; high- and low-level CVs were not calculated because of lack of plasma availability. The reference intervals were calculated with statistical software by use of the robust method following Box-Cox transformation of data after symmetry of distribution was established, with 90% confidence intervals of the upper and lower reference limits.19,j A value of P < 0.05 was considered significant.
Results
Validation of OHP assay and modifications of test methods—Marked differences were detected between normal human and canine plasma OHP curves, with shortened delay and decreased Max OD observed in canine plasma (Figure 2). A thrombin concentration of 0.003 U/mL in each well was identified as optimal to capture the delay phase in canine plasma (Figure 3). At this concentration, delay was prolonged, although OD was not significantly decreased.
Graph of OCP and OHP curves for pooled normal human (square) and pooled canine (diamond) plasma samples generated by use of 0.003 U of thrombin/mL and 350 ng of r-tPA/mL in an OHP assay evaluated for use in dogs. Notice the reduced delay and Max OD of the canine sample, compared with the human sample.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Graph of OCP and OHP curves for pooled normal human (square) and pooled canine (diamond) plasma samples generated by use of 0.003 U of thrombin/mL and 350 ng of r-tPA/mL in an OHP assay evaluated for use in dogs. Notice the reduced delay and Max OD of the canine sample, compared with the human sample.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Graph of OCP and OHP curves for pooled normal human (square) and pooled canine (diamond) plasma samples generated by use of 0.003 U of thrombin/mL and 350 ng of r-tPA/mL in an OHP assay evaluated for use in dogs. Notice the reduced delay and Max OD of the canine sample, compared with the human sample.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Results of optimization of use of thrombin (Thr) in an OHP assay evaluated for use in dogs. Notice the subtle prolongation of delay with lower thrombin concentration and an associated decrease in OD.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Results of optimization of use of thrombin (Thr) in an OHP assay evaluated for use in dogs. Notice the subtle prolongation of delay with lower thrombin concentration and an associated decrease in OD.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Results of optimization of use of thrombin (Thr) in an OHP assay evaluated for use in dogs. Notice the subtle prolongation of delay with lower thrombin concentration and an associated decrease in OD.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Both lipemia and marked hemolysis interfered with results, disrupting OD measurements and giving nonsensical OD readings (Figure 4). Samples that were lipemic or hemolytic (5/53) were subsequently excluded from all further analyses.
Example of curve generated from lipemic plasma used in an OHP assay evaluated for use in dogs.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Example of curve generated from lipemic plasma used in an OHP assay evaluated for use in dogs.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Example of curve generated from lipemic plasma used in an OHP assay evaluated for use in dogs.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
To investigate the difference in fibrin generation between canine and human plasma, the Max OD of human and canine plasma samples was plotted against fibrinogen concentration (Figure 5). The OD of the canine fibrin clot was directly correlated with the fibrinogen concentration, as in human samples. However, the turbidity of the clot was consistently less in canine plasma.
Comparison of OD of canine (diamond) and human (square) clots at known fibrinogen concentrations in an OHP assay evaluated for use in dogs.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Comparison of OD of canine (diamond) and human (square) clots at known fibrinogen concentrations in an OHP assay evaluated for use in dogs.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Comparison of OD of canine (diamond) and human (square) clots at known fibrinogen concentrations in an OHP assay evaluated for use in dogs.
Citation: American Journal of Veterinary Research 74, 12; 10.2460/ajvr.74.12.1493
Coefficients of variation were calculated in repeated samples between and within assay runs. For OCP, OHP, OFP and Max OD, variance was < 10% within and between assays. There was unacceptable variance both within (24%) and between (42%) assays for the delay. The Max slope had 12% variance between assays and 10% variance within assays (Table 1).
Summary statistics and reference intervals for an OHP assay evaluated for use in dogs.
| Variable | Mean ± SD | Reference interval | 90% CI | Intra-assay CV (%) | Interassay CV (%) |
|---|---|---|---|---|---|
| Delay (min) | 2:39 ± 2:00 | 0–8:14 | 0:14–6:25 | 42 | 24 |
| Max slope (Max OD/min) | 131.90 ± 80.30 | 37.84–298.91 | 48.90–246.24 | 12 | 10 |
| OCP | 18.22 ± 9.90 | 5.12–44.26 | 6.82–34.89 | 5.8 | 8 |
| Max OD | 0.33 ± 0.19 | 0.06–0.79 | 0.10–0.63 | 7 | 0.02 |
| OHP | 2.46 ± 1.71 | 0.48–5.85 | 0.68–4.58 | 10 | 0.5 |
| OFP (%) | 84.58 ± 12.94 | 68.49–98.68 | 77.37–96.47 | 0.5 | 3 |
CI = Confidence interval.
Reference intervals—Summary statistics for all OHP output variables were summarized (Table 1). The reference range was considered the middle 95th percentile of the study population. Most variables had a right-skewed distribution and were normally distributed after logarithmic transformation. Males generally had an increased OHP, compared with females (P = 0.04). Max OD was often increased with age but did not reach significance (P = 0.079).
Correlation with routine coagulation testing—There was no significant correlation between the APTT findings and any of the OHP output variables. The PT was strongly negatively correlated with the OCP, OHP, Max OD, and Max slope outputs (Pearson r, −0.69 to −0.75 for each variable [P ≤ 0.01 for all comparisons]). No significant association with delay or fibrinolytic potential was found. The fibrinogen reference range was 1.3 to 4.5 g/dL. The fibrinogen concentration was strongly positively correlated with OCP (P ≤ 0.001) and Max OD (P = 0.001) and more moderately correlated with Max slope outputs (Pearson r = 0.69; P = 0.001) and OHP (Pearson r = 0.49; P = 0.003). A weak positive correlation between delay and fibrinogen concentration did not reach significance (Pearson r = 0.33; P = 0.06).
Discussion
Marked differences between human and canine OHP curves were evident. Inherent differences between coagulation factor activities may contribute to this interspecies variation. In a previous studya with human plasma, an association between increased factor VIII activity and increased OCP was found. Factor VIII activity in canine plasma may be increased up to 7.9-fold, compared with human factor VIII activity20; therefore, the expectation was that the canine OHP curve would be at least equivalent to the human curve. It is likely that the shortened delay in dogs, compared with humans, is linked to the inherently higher factor V and factor VIII activities,20 and further investigation is warranted to examine a correlation between OHP and coagulation factor activity and antigen concentrations in the canine population.
The range of fibrinogen concentrations was similar between the human and canine populations studied5; however, the Max OD (which correlated strongly with fibrinogen concentrations in both studies) differed significantly. This reflects variance in clot opacity between the species, which is likely a reflection of both the fibrin tendril width and density in the clot.22 Different agonists may result in variation in fibrin tendril width; therefore, further research using different coagulation initiators, including species-specific thrombin, factor Xa, and TF, would help to clarify this.21 Despite the differences between canine and human curves, the assay was successfully adapted to detect increases in fibrin generation and decreases in fibrinolysis in dogs, both of which contribute to a hypercoagulable state.
The traditional coagulation assays PT and APTT measure time to fibrin generation under the influence of TF or in the absence of exogenous TF, respectively.22 The OHP also measures time to fibrin generation in the absence of exogenous TF and the time to clot breakdown and therefore has the theoretical potential to detect more subtle hemostatic disorders, compared with PT and APTT. The correlation between the OHP and PT assay in this study was not surprising given that both measure time to clot formation, although the lack of significant correlation with the APTT assay was interesting. This may have been related to variability of the OHP assay in the delay time, to insufficient study size to reach significance, or to the fact that the assays examine different variables. Despite this, the additional information offered by the OHP assay may complement the information offered by the APTT assay in a clinical setting.
The OHP assay is a plasma-based assay; therefore, the platelet, RBC, and endothelial cell contribution to hypercoagulability may be underestimated. However, the presence of microparticles in the plasma, released when platelets are activated, likely provides the phospholipid medium required for effective coagulation to occur.23 In addition, the quantity of microparticles present in plasma is proportional to the degree of thrombin generation.23,24 This likely relates also to fibrin generation as measured by the OHP, although more research is required to state this conclusively.
Thromboelastography remains the only whole blood global coagulation assay available, aside from whole-blood clotting time.11 Although this has the advantage of measuring the cell surface contribution to coagulation in addition to the soluble components, it is limited by its expense, availability, and preanalytic conditions. Whole blood assays such as TEG must be performed within 30 minutes to 2 hours (point-of-care) to avoid artifactual coagulation activation13 and are influenced by Hct and platelet number and function. Plasma-based assays such as the OHP can be performed on stored samples as a batched assay, simplifying comparison with a reference standard, and are not affected by Hct. An additional advantage of the OHP over TEG is its sensitivity for reduced fibrinolytic capacity. Although TEG may be performed with tissue plasminogen activator added to detect fibrinolytic potential, this aspect of the assay is not optimized for canine blood.12,13 Decreased fibrinolysis is a potential contributor to the hypercoagulable state, which has proven difficult to detect in dogs.
Substantial variation was detected in the delay both between and within runs by use of the OHP assay. The delay time was short, compared with human plasma, making it difficult to identify clinically relevant decreases in this variable in dogs. In addition, this variable will be more susceptible to preanalytic conditions (eg, the time difference between addition of thrombin to the first well and the last). Although a substantially prolonged delay may be detected, a short delay is unlikely to be diagnostic for hypercoagulability. The variability among assay runs of the Max slope was higher than expected, particularly given the gross visual similarity of the curves generated. However, the potential of the OHP assay as a tool for detection and treatment of hypercoagulability should not be discounted on the basis of its variability alone, particularly given the low variability in OCP, OHP, and fibrinolysis.
The present study identified greater OHP in males, compared with females, whereas in human populations, female patients typically have greater OHP than do males.25,a This has been associated with use of oral contraceptives and pregnancy25,a and is therefore less relevant in the veterinary population. In the canine population, the difference between sexes is likely influenced by the attenuated sex hormone concentrations associated with neutering. Further research is necessary to determine whether this difference exists in sexually intact males and females as well as pregnant females. The lack of sexually intact dogs in the study population was a limitation of this study.
Another limitation of this study was the lack of standardization of preanalytic procedures. Ideally, blood would have been drawn directly into citrated tubes to prevent time-dependent coagulation system activation and turbulence during transfer from syringe to tube causing platelet activation and microparticle release. The optimization of this assay was aimed at developing a methodology that would be easily used by clinicians, and the methodology used for citrated blood sample collection is standard procedure at the authors' center.
The methodology and reference intervals established in this study may provide a standardized approach for future studies to investigate the application of the OHP assay to dogs with thrombosis or thromboembolism. The potential use of this assay as an affordable substitute for other global assays is dependent on further research in dogs with thromboembolic disease.
ABBREVIATIONS
| APTT | Activated partial thromboplastin time |
| CV | Coefficient of variation |
| Max OD | Maximum optical density |
| Max | slope Maximum slope |
| OCP | Overall coagulation potential |
| OFP | Overall fibrinolytic potential |
| OHP | Overall hemostasis potential |
| PPP | Platelet-poor plasma |
| PT | Prothrombin time |
| r-tPA | Recombinant tissue plasminogen activator |
| TEG | Thromboelastography |
| TF | Tissue factor |
Curnow J. Evaluation of the overall haemostatic potential assay for the diagnosis and management of hypercoagulable states. Sydney: University of Sydney, 2010.
BD Vacutainer 0.109M, Becton-Dickinson, Franklin Lakes, NJ.
STA Neoplastin Cl plus 10, Stago Diagnostica, Doncaster, VIC, Australia.
STA Cephascreen 4, Stago Diagnostica, Doncaster, VIC, Australia.
Stago Compact analyzer, Stago Diagnostica, Doncaster, VIC, Australia.
BC thrombin reagent, Dade Behring, Newark, Del.
Actilyse, Boehringer Ingelheim, Ingelheim, Germany.
Powerwave XS Biotek microplate reader, Biotek, Winooski, Vt.
R, version 2.13.0, R Foundation for Statistical Computing, Vienna, Austria. Available at: http://www.r-project.org/index.html. Accessed May 13, 2012.
Reference Value Advisor, version 2.1, Biostatistiques Toulouse, Toulouse, France.
References
1. Boswood A, Lamb CR, White RN. Aortic and iliac thrombosis in six dogs. J Small Anim Pract 2000; 41:109–114.
2. Garosi L, McConnell JF, Platt SR, et al. Results of diagnostic investigations and long-term outcome of 33 dogs with brain infarction (2000–2004). J Vet Intern Med 2005; 19:725–731.
3. Klose TC, Creevy KE, Brainard BM. Evaluation of coagulation status in dogs with naturally occurring canine hyperadrenocorticism. J Vet Emerg Crit Care (San Antonio) 2011; 21:625–632.
4. Laurenson MP, Hopper K, Herrera MA, et al. Concurrent diseases and conditions in dogs with splenic vein thrombosis. J Vet Intern Med 2010; 24:1298–1304.
5. Curnow JL, Morel-Kopp MC, Roddie C, et al. Reduced fibrinolysis and increased fibrin generation can be detected in hypercoagulable patients using the overall hemostatic potential assay. J Thromb Haemost 2007; 5:528–534.
6. Burns MG, Kelly AB, Hornof WJ, et al. Pulmonary artery thrombosis in three dogs with hyperadrenocorticism. J Am Vet Med Assoc 1981; 178:388–393.
7. Schaer H. Thrombotic disorders. In: Weiss DJ, Wardrop KJ, ed. Schalm's veterinary hematology. 6th ed. Ames, Iowa: Wiley-Blackwell, 2010;668–678.
8. Nelson OL. Use of the D-dimer assay for diagnosing thromboembolic disease in the dog. J Am Anim Hosp Assoc 2005; 41:145–149.
9. Stokol T, Brooks M, Rush JE, et al. Hypercoagulability in cats with cardiomyopathy (Erratum published in J Vet Intern Med 2009; 23:224). J Vet Intern Med 2008; 22:546–552.
10. Tarnow I, Falk T, Tidholm A, et al. Hemostatic biomarkers in dogs with chronic congestive heart failure. J Vet Intern Med 2007; 21:451–457.
11. Donahue SM, Otto CM. Thromboelastography: a tool for measuring hypercoagulability, hypocoagulability, and fibrinolysis. J Vet Emerg Crit Care (San Antonio) 2005; 15:9–16.
12. Wiinberg B, Kristensen AT. Thromboelastography in veterinary medicine. Semin Thromb Hemost 2010; 36:747–756.
13. Kol A, Borjesson DL. Application of thrombelastography/thromboelastometry to veterinary medicine. Vet Clin Pathol 2010; 39:405–416.
14. Antovic A. The overall hemostasis potential: a laboratory tool for the investigation of global hemostasis. Semin Thromb Hemost 2010; 36:772–779.
15. Antovic JP, Antovic A, He S, et al. Overall haemostatic potential can be used for estimation of thrombin-activatable fibrinolysis inhibitor-dependent fibrinolysis in vivo and for possible follow-up of recombinant factor VIIa treatment in patients with inhibitors to factor VIII. Haemophilia 2002; 8:781–786.
16. He S, Antovic A, Blomback M. A simple and rapid laboratory method for determination of haemostasis potential in plasma II. Modifications for use in routine laboratories and research work. Thromb Res 2001; 103:355–361.
17. Jones AJ, Meunier AM. A precise and rapid microtitre plate clot lysis assay: methodology, kinetic modeling and measurement of catalytic constants for plasminogen activation during fibrinolysis. Thromb Haemost 1990; 64:455–463.
18. Bauer N, Eralp O, Moritz A. Reference intervals and method optimization for variables reflecting hypocoagulatory and hypercoagulatory states in dogs using the sta compact® automated analyzer. J Vet Diagn Invest 2009; 21:803–814.
19. Geffré A, Concordet D, Braun JP, et al. Reference Value Advisor: a new freeware set of macroinstructions to calculate reference intervals with Microsoft Excel. Vet Clin Pathol 2011; 40:107–112.
20. Mischke R. Optimization of coagulometric tests that incorporate human plasma for determination of coagulation factor activities in canine plasma. Am J Vet Res 2001; 62:625–629.
21. Smith SA, Morrissey JA. Polyphosphate enhances fibrin clot structure. Blood 2008; 112:2810–2816.
22. Laffan MA, Manning R. Investigation of haemostasis. In: Bain JB, Bates I, Laffan MA, et al, eds. Dacie and Lewis practical haematology. 11th ed. Philadelphia: Churchill Livingstone Elsevier, 2012;393–447.
23. Bidot L, Jy W, Bidot C, et al. Microparticle-mediated thrombin generation assay: increased activity in patients with recurrent thrombosis. J Thromb Haemost 2008; 6:913–919.
24. Wielders SJH, Ungethum L, Reutelingsperger CPM, et al. Factor Xa-driven thrombin generation in plasma: dependency on the aminophospholipid density of membranes and inhibition by phospholipid-binding proteins. Thromb Haemost 2007; 98:1056–1062.
25. Rosing J, Tans G. Effects of oral contraceptives on hemostasis and thrombosis. Am J Obstet Gynecol 1999; 180:S375–S382.
