The measurement results for the procoagulant and anticoagulant proteins are shown in Figures 2 and 3, respectively. As expected, 3-fold dilution of the plasma samples resulted in a decrease in the measured coagulation factor levels. Likewise, the postsupplementation protein levels were close to their predilution levels. The measured levels of anticoagulants were near the expected values for the considered scenarios (Figure 3). However, for all the procoagulant proteins, rFVIIa supplementation resulted in increased protein levels (Figure 2); this effect was particularly extreme in the case of FVII (3686% FVII activity; Figure 2C). Such bias was not entirely unexpected and is consistent with the 1-stage assay design,40 whose results may be skewed when the test plasma contains a supraphysiologic concentrations of a clotting factor.
Thrombin Generation Trajectories
The TG trajectory data for the 5 considered dilution/supplementation scenarios are summarized in Figure 4, and the individual subject trajectories are shown in Figures S1 and S2 (Supplemental Digital Content, http://links.lww.com/AA/B432). All the trajectories had the expected 1-peak shape that allowed the automatic calculation of the TG parameters defined in Figure 1. In the case of CCF-FVII supplementation, the right-hand “tails” of the thrombin curves were characterized by high variability for certain subjects (eg, Figure S2, A and B), which occurred apparently because of an increased noise-to-signal ratio resulting from fluorogenic substrate depletion caused by the strong CCF-FVII-triggered TG in the samples.
Direct visual inspection of the summarized thrombin trajectories (Figure 4) and the thrombin trajectories for individual subjects (Figures S1 and S2) allowed us to identify patterns that held for each subject despite the intersubject variability in the data set. Specifically, dilution always reduced the height of the TG peak but did not increase the ttP. Moreover, rFVIIa supplementation visibly accelerated the onset of TG in diluted plasma but it did not considerably change the PH. Supplementation with rFVIIa resulted in the earliest onset of TG (in some subjects, closely matched by that in CCF-FVII-supplemented plasma). Finally, among the 5 considered dilution/supplementation scenarios, the largest TG peak resulted from CCF-FVII supplementation and the lowest peak was either the one in the diluted plasma or the rFVIIa-supplemented plasma. In several subjects, CCF-AT supplementation resulted in a nearly normal TG trajectory (Figures S1, D and F, and S2, B and D). In contrast, the rFVIIa and CCF-FVII supplementations did not result in accurate thrombin trajectory normalization in any of the 10 subjects (Figures S1 and S2). The ability of CCF-AT, and the inability of both rFVIIa and CCF-FVII, to restore normal TG in diluted plasma is clearly demonstrated by the summarized TG data (Figure 4).
Standard Coagulation Tests and Thrombin Generation Parameters
In our experiments, PT was prolonged by dilution (Figure 5A). Supplementation with rFVIIa shortened PT to a near-normal level, whereas CCF-AT and CCF-FVII supplementation resulted in prolonged PT (compared with undiluted plasma). The PT after CCF-AT supplementation was somewhat prolonged compared with those after CCF-FVII supplementation, which is consistent with the anticoagulant action of AT present in CCF-AT. aPTT modulation was qualitatively similar to that of PT (Figure 5A).
Both PT and LT reflect the duration of the initiation phase of TF-triggered TG. However, in contrast to the dilution-induced prolongation of PT and aPTT (Figure 5A), no significant difference in LT was detected between undiluted and diluted plasma (P = 0.19, Figure 5B), which is consistent with our earlier in vitro dilution experiments.12,13 Both the VI and the thrombin PH were decreased in diluted plasma (Figure 5, D and E).
Supplementation with rFVIIa noticeably shortened both LT and ttP in diluted plasma (Figure 5, B and C). Thrombin PH in the diluted plasma supplemented with rFVIIa, although, was similar to that in diluted plasma and significantly smaller than that in undiluted plasma (Figure 5E). These results were consistent with previous computational predictions.21,28 As predicted,38 dilution reduced the VI, but contrary to the prediction,21 rFVIIa supplementation did not return this parameter to baseline (ie, to its value in undiluted plasma; Figure 5D). CCF-AT and CCF-FVII supplementation resulted in VI values considerably exceeding the baseline.
Previous simulations have suggested that CCF-FVII supplementation, in contrast to rFVIIa, should primarily affect those parameters reflecting peak TG, such as the thrombin PH and endogenous thrombin potential.21 In our experiments, CCF-FVII supplementation not only resulted in higher-than-baseline values for both peak TG and endogenous thrombin potential, but also both these values were the largest among the 5 considered dilution/supplementation scenarios (Figure 5, E and F). The respective P values were P < 0.001 and P < 0.055 (the Wilcoxon test for the latter). These P values were Bonferroni corrected with denominator 4 because, in this analysis, it was sufficient to compare the samples for the CCF-FVII supplementation with each of the corresponding samples for the remaining 4 scenarios). Moreover, as predicted,21 CCF-FVII supplementation was the only dilution/supplementation scenario that caused a significant difference in the endogenous thrombin potential compared with undiluted plasma (Figure 5F).
CCF-AT supplementation restored ttP to baseline level (Figure 5C) and yielded a thrombin PH value that was, on average, the closest to its predilution level (Figure 5E). Unexpectedly, however, CCF-AT supplementation resulted in a somewhat prolonged LT (Figure 5B). Supplementation with CCF-AT resulted in a VI value approximately 2-fold higher than the baseline value, whereas the value of this parameter in diluted plasma was approximately 2-fold lower than the baseline (Figure 5D). CCF-FVII supplementation resulted in the largest VI value across the 5 considered dilution/supplementation scenarios (Figure 5D; P ≤ 0.039 [the Wilcoxon test], Bonferroni-corrected with denominator 4, similarly to the analysis described earlier). The above-normal levels of the VI (Figure 5D) and thrombin PH (Figure 5E) after both CCF-AT and CCF-FVII supplementations can be explained by noticing that dilution reduces the concentrations of all natural procoagulants and anticoagulants, whereas CCF-FVII replenishes none of the anticoagulants and CCF-AT replenishes only 1 of the anticoagulants (ie, AT). Indeed, the levels of TFPI, as well as those of other coagulation inhibitors,35 were not restored in our clotting factor supplementation scenarios.
To test whether the effects of CCF-AT were different from those of rFVIIa and CCF-FVII, we compared the TG parameters for the CCF-AT supplementation with the corresponding samples for the rFVIIa and CCF-FVII supplementations (Figure 5, B–F). The differences between the effects of CCF-AT versus those of rFVIIa and CCF-FVII were significant for LT (P < 0.001 and P = 0.005, respectively), ttP (P < 0.001 and P = 0.004, respectively), VI (P < 0.001 and P = 0.019 [the Wilcoxon test for the latter], respectively), thrombin PH(P< 0.001 for both comparisons), and endogenous thrombin potential (P =0.034 and P = 0.019 [the Wilcoxon test for the latter], respectively). These P values were Bonferroni corrected with denominator 2 for each TG parameter analyzed independently. Similar comparisons for the corresponding PT and aPTT also indicated significant differences (all corresponding P values did not exceed 0.05, except the CCF-AT versus CCF-FVII comparison for aPTT).
Predictive Analysis of the Thrombin Trajectories Using the Extended Computational Model
As a result of 10 cross-validation rounds (1 round for each subject), we obtained 10 model variants with subject-specific initial conditions, which were somewhat different in their kinetic parameter values but were similar in their subject-specific predictive capacity. In each of the 10 model variants, the kinetic parameters reflected the training data from the 9 subjects who were used in model training (but not in model validation). To obtain 1 model that would contain information about all 10 subjects in our subject group, each kinetic parameter value was averaged across the 10 model variants, which resulted in an averaged-parameter model. This model was used with subject-specific initial conditions to generate thrombin trajectories for distinct subjects.
Model cross-validation results for the subject group are summarized in Figure 6; the modeling results for individual subjects are shown in Figures S3 to S12. The trained model accurately predicted TG kinetics in undiluted plasma (Figure 6A), and the predictions for rFVIIa and CCF-AT supplementation were also rather accurate (Figure 6, C and D, respectively). Less accurate were the predictions of TG in unsupplemented diluted plasma (Figure 6B). Yet, the model qualitatively predicted the thrombin PH decrease in diluted plasma (Figure 6, A and B), and the exaggerated PH and endogenous thrombin potential in diluted plasma supplemented with CCF-FVII (Figure 6, A and E). In the case of CCF-FVII supplementation, the model prediction for thrombin PH was quantitatively accurate, and ttP was similar to that detected experimentally (Figure 6E). However, the decay of the thrombin trajectory after the peak was noticeably slower in the model-predicted thrombin trajectories than in the experimental data. For most scenarios, using the averaged-parameter model led to increased modeling accuracy (cf. green lines and black lines in Figure 6, C–E) but somewhat reduced the modeling accuracy for undiluted plasma (Figure 6A).
Trauma-induced coagulopathy is a significant cause of morbidity and mortality worldwide,1,2 with limited therapeutic approaches. In our experiments, the constituent coagulation factor levels for the CCF-AT and CCF-FVII supplementations were restored to their predilution values, which is consistent with the factor levels detected after PCC administration in a porcine model of dilutional coagulopathy.29,30 The CCF-AT supplementation scenario achieved a more balanced enhancement of TG compared with rFVIIa and CCF-FVII supplementation. This is because the latter 2 scenarios strongly accelerated the onset (especially, rFVIIa) and increased the abundance (CCF-FVII) of TG to suprabaseline levels.
Plasma dilution in vitro is only a simplified representation of some of the changes to the hemostatic proteome occurring in vivo during trauma-induced coagulopathy. Yet, blood coagulation analysis in diluted/supplemented plasma may be regarded as a first step toward the mechanistic understanding of coagulopathic conditions. Although different resuscitation fluids have distinct effects on TG,41 we chose normal saline as an example of a resuscitation fluid in trauma42 and priming solution in cardiac surgery.43 Clinical studies have shown that, in human trauma, the degrees of coagulation factor depletion are both factor- and study-dependent.3–5 Given this uncertainty, we chose an equal-degree dilution scenario, in which every coagulation factor was diluted to the same degree (ie, 3-fold).
Our results suggest that our dilution/supplementation procedure achieved the target coagulation factor levels in the subjects’ plasma samples. However, measurement assays using PT-based and chromogenic methods depend on FXa generation, which is increased in rFVIIa-supplemented plasma.44 As another example, a commercially available (from Abcam®, Cambridge, UK) immunofunctional/chromogenic assay for FVII can also be biased by the presence of FVIIa (Abcam scientific support, email communication). Thus, accurate clotting factor measurements in the presence of high rFVIIa levels appear to be a common problem for various methods, and the 1-stage clotting assays should still be preferred in general situations.
In this study, we specifically focused on the effects of normalizing the levels of the major coagulation components found in PCCs rather than using the preformulated, commercially available products. This unique approach allowed us to independently control the concentration of supplemented AT. AT is the most abundant natural inhibitor of the blood coagulation proteases and may be regarded as the strongest natural anticoagulant.10,45,46 Therefore, the design of the CCF-AT supplementation strategy indicates its ability to considerably strengthen both coagulation and anticoagulation processes in plasma, which may be more difficult to achieve with other natural TG inhibitors, such as TFPI and protein C. Most commercially available PCCs only include small amounts of AT (1%–5%) compared with higher concentrations of procoagulant factors.47,48
The possibility of thromboembolic complications is a major concern when procoagulants are used in traumatic and surgical coagulopathy. Indeed, rFVIIa is known to cause thromboembolic complications49 (which may depend on the dose, patient age, and other factors50). Similar concerns are valid for PCCs, even though the information from clinical studies may not be sufficient for a definitive conclusion.15,51 In our opinion, administration of AT together with commercially available PCCs is a promising approach, but the relative dosage and administration time for the 2 distinct components will require careful investigation. With properly balanced procoagulants and anticoagulants, both 3- and 4-factor PCCs could likely achieve the desired improved level of safety without sacrificing efficacy. However, 3-factor PCCs should probably be preferred because their activity lacks the uncertainty associated with the function of FVII. Indeed, although FVII is generally considered a procoagulant, it has been shown to inhibit TG in vitro,52 which suggests that its overall contribution to blood coagulation may depend on the specific situation and be difficult to predict or control.
Our experimental approach has a number of limitations. First, our study relied on a widely used experimental model that measures TG in a static, cell-free in vitro system. Therefore, the effects of platelets, blood vessel endothelial cells, and blood flow on TG could not be captured in this study. Although the use of PPP (rather than platelet-rich plasma or whole blood) may be regarded as a limitation of this study, the CAT methodology for TG measurement is particularly well established for PPP, which allows for reliable data generation and direct comparisons with available literature. Second, the effects of thrombomodulin and protein C were not investigated. Third, because we focused solely on TG, other aspects of blood coagulation, such as fibrin generation and fibrinolysis, were not addressed. Indeed fibrinogen deficiency, which may affect fibrin accumulation, is likely to be the primary manifestation of trauma-induced coagulopathy. The final limitation stems from our use of an in vitro model of clotting factor dilution/depletion and supplementation scenarios that may not fully capture the characteristics of these processes in vivo. Nevertheless, the design and results of this study form a foundation for future investigations into the biological reactions taking place in vivo.
The results of this study indicate that the possible detrimental effects of rFVIIa and current PCC products may be caused by a distortion of TG kinetics caused by these treatment strategies; this distortion may be avoided if procoagulant components are combined with AT. Overall, our results suggest that if hemostatic balance is to be achieved, the therapeutic intervention strategy itself should be balanced. The notion of balance is a key concept in our investigation and can be defined as the presence of both procoagulant and anticoagulant components at comparable levels of abundance or activity in the therapeutic strategy. We believe that this concept should guide the search for optimized clotting factor compositions, and one approach to perform such searches is through computational modeling.
We devised a computational modeling strategy to predict TG in the diluted/supplemented subject plasma samples. The strategy involved subject-specific parameter tuning during model training. This was based on the use of TG data obtained from the subjects’ plasma, an approach advocated in a critique of mathematical modeling in blood coagulation research.53 It appears that to capture the specifics of the experimental protocol and to better reflect intersubject variability, such parameter tuning is necessary. Without it, the modeling accuracy can typically be only qualitative or semiquantitative at best.23 The cross-validation predictions of TG for the undiluted plasma scenario were quantitatively accurate for most subjects (Figure 6 and Figures S3 to S12). The TG model prediction accuracy for the diluted and supplemented plasma was somewhat lower for certain subject/scenario combinations (particularly, for dilution and CCF-FVII supplementation), which warrants future efforts to improve both the mathematical model and training/validation algorithm.
One of our model’s main limitations is its limited accuracy in certain situations. It is plausible that including TG data for diluted/supplemented plasma in the training data sets would further improve the model’s accuracy in predicting TG under these conditions. Moreover, the limitations indicated for our experimental phase also apply to the computational modeling phase and suggest directions for model improvement. For example, our recently developed computational model23 of fibrin formation and fibrinolysis may be combined with the subject-specific model training algorithm presented here and may be used to predict fibrin accumulation in blood samples from individual subjects.
The results of this study support the notion that mathematical modeling can facilitate hypothesis generation and thereby successfully guide experimental and clinical research. The presented modeling strategy can be applied as a “virtual test bed” to investigate subject-specific effects of coagulopathic conditions (including hypothermia36 and acidosis37) and clotting factor supplementation on TG. Research objectives for future simulation-driven predictive analyses include understanding of the effects of current commercial PCC products under different coagulopathic conditions, the effects of plausible therapeutics, such as combinations of rFVIIa with natural anticoagulants, and the effects of intervention dosage and timing variations on the therapeutic outcomes.
The computer code implementing our mathematical model and the computational analyses is available from the authors on request.
The opinions and assertions contained herein are private views of the authors and are not to be construed as official or as reflecting the views of the US Army or the US Department of Defense. This article has been approved for public release with unlimited distribution.
Name: Alexander Y. Mitrophanov, PhD.
Contribution: This author helped design the research, perform the computational modeling, analyze the data, and write the manuscript.
Conflicts of Interest: Alexander Y. Mitrophanov declares no conflicts of interest.
Name: Fania Szlam, MMSc.
Contribution: This author helped design the study, carry out in vitro experiments, analyze the data, and prepare the manuscript.
Conflicts of Interest: Fania Szlam declares no conflicts of interest.
Name: Roman M. Sniecinski, MD.
Contribution: This author helped collect the data, analyze the data, and prepare the manuscript.
Conflicts of Interest: Roman M. Sniecinski reports grants and personal fees from Grifols and grants from Shire ViroPharma outside the submitted work.
Name: Jerrold H. Levy, MD.
Contribution: This author helped collect the data, analyze the data, and prepare the manuscript.
Conflicts of Interest: Jerrold H. Levy reports participation in the Steering Committees for CSL Behring, Boehringer-Ingelheim, Jansen, and Grifols during the conduct of the study.
Name: Jaques Reifman, PhD.
Contribution: This author helped design the study, analyze the data, and edit the manuscript.
Conflicts of Interest: Jaques Reifman declares no conflicts of interest.
This manuscript was handled by: Charles W. Hogue, MD.
Dr. Roman M. Sniecinski is the Section Editor for Hemostasis for Anesthesia & Analgesia. This manuscript was handled by Dr. Charles W. Hogue, then Associate Editor-in-Chief and Section Editor for Cardiovascular Anesthesia, and Dr. Sniecinski was not involved in any way with the editorial process or decision.
The authors are grateful to the editor and 2 anonymous reviewers whose comments have helped to improve the article, and to Dr. Franklin Dexter for his comments and suggestions regarding the statistical methods used in this work. The authors are grateful to Ms. Romy Kremers and Drs. Srinivas Laxminarayan, Maurizio Tomaiuolo, Paul Riley, and Peter Giesen for valuable discussions.
1. Cap A, Hunt BJ. The pathogenesis of traumatic coagulopathy. Anaesthesia 2015;70(suppl 1):96–101e32–4.
2. Schochl H, Voelckel W, Schlimp CJ. Management of traumatic haemorrhage—the European perspective. Anaesthesia 2015;70(suppl 1):102–7e35–7.
3. Cohen MJ, Kutcher M, Redick B, Nelson M, Call M, Knudson MM, Schreiber MA, Bulger EM, Muskat P, Alarcon LH, Myers JG, Rahbar MH, Brasel KJ, Phelan HA, del Junco DJ, Fox EE, Wade CE, Holcomb JB, Cotton BA, Matijevic N; PROMMTT Study Group. Clinical and mechanistic drivers of acute traumatic coagulopathy. J Trauma Acute Care Surg 2013;75:S40–7.
4. Rizoli SB, Scarpelini S, Callum J, Nascimento B, Mann KG, Pinto R, Jansen J, Tien HC. Clotting factor deficiency in early trauma-associated coagulopathy. J Trauma 2011;71:S427–34.
5. Shaz BH, Winkler AM, James AB, Hillyer CD, MacLeod JB. Pathophysiology of early trauma-induced coagulopathy: emerging evidence for hemodilution and coagulation factor depletion. J Trauma 2011;70:1401–7.
6. Cardenas JC, Rahbar E, Pommerening MJ, Baer LA, Matijevic N, Cotton BA, Holcomb JB, Wade CE. Measuring thrombin generation as a tool for predicting hemostatic potential and transfusion requirements following trauma. J Trauma Acute Care Surg 2014;77:839–45.
7. Dunbar NM, Chandler WL. Thrombin generation in trauma patients. Transfusion 2009;49:2652–60.
8. Guzzetta NA, Szlam F, Kiser AS, Fernandez JD, Szlam AD, Leong T, Tanaka KA. Augmentation of thrombin generation in neonates undergoing cardiopulmonary bypass. Br J Anaesth 2014;112:319–27.
9. Sniecinski R, Szlam F, Chen EP, Bader SO, Levy JH, Tanaka KA. Antithrombin deficiency increases thrombin activity after prolonged cardiopulmonary bypass. Anesth Analg 2008;106:713–8.
10. Hockin MF, Jones KC, Everse SJ, Mann KG. A model for the stoichiometric regulation of blood coagulation. J Biol Chem 2002;277:18322–33.
11. Mann KG, Butenas S, Brummel K. The dynamics of thrombin formation. Arterioscler Thromb Vasc Biol 2003;23:17–25.
12. Bolliger D, Szlam F, Levy JH, Molinaro RJ, Tanaka KA. Haemodilution-induced profibrinolytic state is mitigated by fresh-frozen plasma: implications for early haemostatic intervention in massive haemorrhage. Br J Anaesth 2010;104:318–25.
13. Bolliger D, Szlam F, Molinaro RJ, Rahe-Meyer N, Levy JH, Tanaka KA. Finding the optimal concentration range for fibrinogen replacement after severe haemodilution: an in vitro model. Br J Anaesth 2009;102:793–9.
14. De Smedt E, Wagenvoord R, Coen Hemker H. The technique of measuring thrombin generation with fluorogenic substrates: 3. The effects of sample dilution. Thromb Haemost 2009;101:165–70.
15. Grottke O, Rossaint R, Henskens Y, van Oerle R, Ten Cate H, Spronk HM. Thrombin generation capacity of prothrombin complex concentrate in an in vitro dilutional model. PLoS One 2013;8:e64100.
16. Schols SE, Feijge MA, Lancé MD, Hamulyák K, ten Cate H, Heemskerk JW, van Pampus EC. Effects of plasma dilution on tissue-factor-induced thrombin generation and thromboelastography: partly compensating role of platelets. Transfusion 2008;48:2384–94.
17. Schols SE, Lancé MD, Feijge MA, Damoiseaux J, Marcus MA, Hamulyák K, Ten Cate H, Heemskerk JW, van Pampus EC. Impaired thrombin generation and fibrin clot formation in patients with dilutional coagulopathy during major surgery. Thromb Haemost 2010;103:318–28.
18. Schols SE, van der Meijden PE, van Oerle R, Curvers J, Heemskerk JW, van Pampus EC. Increased thrombin generation and fibrinogen level after therapeutic plasma transfusion: relation to bleeding. Thromb Haemost 2008;99:64–70.
19. Schöchl H, Maegele M, Solomon C, Görlinger K, Voelckel W. Early and individualized goal-directed therapy for trauma-induced coagulopathy. Scand J Trauma Resusc Emerg Med 2012;20:15.
20. Tanaka KA, Mazzeffi M, Durila M. Role of prothrombin complex concentrate in perioperative coagulation therapy. J Intensive Care 2014;2:60.
21. Mitrophanov AY, Rosendaal FR, Reifman J. Therapeutic correction of thrombin generation in dilution-induced coagulopathy: computational analysis based on a data set of healthy subjects. J Trauma Acute Care Surg 2012;73:S95–102.
22. Luan D, Szlam F, Tanaka KA, Barie PS, Varner JD. Ensembles of uncertain mathematical models can identify network response to therapeutic interventions. Mol Biosyst 2010;6:2272–86.
23. Mitrophanov AY, Wolberg AS, Reifman J. Kinetic model facilitates analysis of fibrin generation and its modulation by clotting factors: implications for hemostasis-enhancing therapies. Mol Biosyst 2014;10:2347–57.
24. Panteleev MA, Sveshnikova AN, Belyaev AV, Nechipurenko DY, Gudich I, Obydenny SI, Dovlatova N, Fox SC, Holmuhamedov EL. Systems biology and systems pharmacology of thrombosis. Math Model Nat Phenom 2014;9:4–16.
25. Shibeko AM, Woodle SA, Mahmood I, Jain N, Ovanesov MV. Predicting dosing advantages of factor VIIa variants with altered tissue factor-dependent and lipid-dependent activities. J Thromb Haemost 2014;12:1302–12.
26. Tanaka KA, Mazzeffi MA, Strauss ER, Szlam F, Guzzetta NA. Computational simulation and comparison of prothrombin complex concentrate dosing schemes for warfarin reversal in cardiac surgery. J Anesth 2016;30:369–76.
27. Dickneite G, Doerr B, Kaspereit F. Characterization of the coagulation deficit in porcine dilutional coagulopathy and substitution with a prothrombin complex concentrate. Anesth Analg 2008;106:1070–7.
28. Mitrophanov AY, Reifman J. Kinetic modeling sheds light on the mode of action of recombinant factor VIIa on thrombin generation. Thromb Res 2011;128:381–90.
29. Dickneite G, Dörr B, Kaspereit F, Tanaka KA. Prothrombin complex concentrate versus recombinant factor VIIa for reversal of hemodilutional coagulopathy in a porcine trauma model. J Trauma 2010;68:1151–7.
30. Dickneite G, Pragst I. Prothrombin complex concentrate vs fresh frozen plasma for reversal of dilutional coagulopathy in a porcine trauma model. Br J Anaesth 2009;102:345–54.
31. Brummel-Ziedins K, Vossen CY, Rosendaal FR, Umezaki K, Mann KG. The plasma hemostatic proteome: thrombin generation in healthy individuals. J Thromb Haemost 2005;3:1472–81.
32. Hemker HC, Giesen P, Al Dieri R, Regnault V, de Smedt E, Wagenvoord R, Lecompte T, Béguin S. Calibrated automated thrombin generation measurement in clotting plasma. Pathophysiol Haemost Thromb 2003;33:4–15.
33. Jarque CM, Bera AK. A test for normality of observations and regression residuals. Internat Statist Rev 1987;55:163–72.
34. Danforth CM, Orfeo T, Mann KG, Brummel-Ziedins KE, Everse SJ. The impact of uncertainty in a blood coagulation model. Math Med Biol 2009;26:323–36.
35. Kremers RM, Peters TC, Wagenvoord RJ, Hemker HC. The balance of pro- and anticoagulant processes underlying thrombin generation. J Thromb Haemost 2015;13:437–47.
36. Mitrophanov AY, Rosendaal FR, Reifman J. Computational analysis of the effects of reduced temperature on thrombin generation: the contributions of hypothermia to coagulopathy. Anesth Analg 2013;117:565–74.
37. Mitrophanov AY, Rosendaal FR, Reifman J. Mechanistic modeling of the effects of acidosis on thrombin generation. Anesth Analg 2015;121:278–88.
38. Mitrophanov AY, Rosendaal FR, Reifman J. Computational analysis of intersubject variability and thrombin generation in dilutional coagulopathy. Transfusion 2012;52:2475–86.
39. Prechelt L. Automatic early stopping using cross validation: quantifying the criteria. Neural Netw 1998;11:761–7.
40. Kitchen S, McCraw A, Echenagucia M.Diagnosis of hemophilia and other bleeding disorders. A Laboratory Manual. 20102nd ed. Montreal, Canada, World Federation of Hemophilia, .
41. Brummel-Ziedins K, Whelihan MF, Ziedins EG, Mann KG. The resuscitative fluid you choose may potentiate bleeding. J Trauma 2006;61:1350–8.
42. Schreiber MA. The use of normal saline for resuscitation in trauma. J Trauma 2011;70:S13–4.
43. Holley FO, Ponganis KV, Stanski DR. Effect of cardiopulmonary bypass on the pharmacokinetics of drugs. Clin Pharmacokinet 1982;7:234–51.
44. Bolliger D, Szlam F, Molinaro RJ, Escobar MA, Levy JH, Tanaka KA. Thrombin generation and fibrinolysis in anti-factor IX treated blood and plasma spiked with factor VIII inhibitor bypassing activity or recombinant factor VIIa. Haemophilia 2010;16:510–7.
45. Butenas S, van’t Veer C, Mann KG. “Normal” thrombin generation. Blood 1999;94:2169–78.
46. van ‘t Veer C, Golden NJ, Kalafatis M, Mann KG. Inhibitory mechanism of the protein C pathway on tissue factor-induced thrombin generation. Synergistic effect in combination with tissue factor pathway inhibitor. J Biol Chem 1997;272:7983–94.
47. Grottke O, Levy JH. Prothrombin complex concentrates in trauma and perioperative bleeding. Anesthesiology 2015;122:923–31.
48. Levy JH, Tanaka KA, Dietrich W. Perioperative hemostatic management of patients treated with vitamin K antagonists. Anesthesiology 2008;109:918–26.
49. Thomas GO, Dutton RP, Hemlock B, Stein DM, Hyder M, Shere-Wolfe R, Hess JR, Scalea TM. Thromboembolic complications associated with factor VIIa administration. J Trauma 2007;62:564–9.
50. Levi M, Levy JH, Andersen HF, Truloff D. Safety of recombinant activated factor VII in randomized clinical trials. N Engl J Med 2010;363:1791–800.
51. Schöchl H, Voelckel W, Maegele M, Kirchmair L, Schlimp CJ. Endogenous thrombin potential following hemostatic therapy with 4-factor prothrombin complex concentrate: a 7-day observational study of trauma patients. Crit Care 2014;18:R147.
52. van ‘t Veer C, Golden NJ, Mann KG. Inhibition of thrombin generation by the zymogen factor VII: implications for the treatment of hemophilia A by factor VIIa. Blood 2000;95:1330–5.
53. Hemker HC, Kerdelo S, Kremers RM. Is there value in kinetic modeling of thrombin generation? No (unless…). J Thromb Haemost 2012;10:1470–7.
Supplemental Digital Content
© 2016 International Anesthesia Research Society