OUP user menu

Influence of Coagulation and Anticoagulant Factors on Global Coagulation Assays in Healthy Adults

Seon Young Kim MD, Ji-Eun Kim MS, Hyun Kyung Kim MD, PhD, Inho Kim MD, PhD, Sung-Soo Yoon MD, PhD, Seonyang Park MD, PhD
DOI: http://dx.doi.org/10.1309/AJCPC5C4AGFRDKMX 370-379 First published online: 1 March 2013

Abstract

It remains unclear how coagulation and anticoagulant factors influence global coagulation assays such as prothrombin time (PT), activated partial thromboplastin time (aPTT), and thrombin generation assay (TGA). We measured PT, aPTT, coagulation factor and protein levels, and TGA parameters (lag time, endogenous thrombin potential [ETP], and peak thrombin) in 252 apparently healthy adults. Vitamin K–dependent coagulation and anticoagulant factors were significantly correlated with blood lipids. PT was determined by factor (F) V and FVII; aPTT was dependent on antithrombin, protein C, FVIII, and FXII. Lag time was mainly determined by FVII, FXII, and protein S and peak thrombin by FVIII and FIX. Antithrombin (for ETP and lag time) and protein S (for lag time) contributed significantly to TGA inhibition. This knowledge about determinants of global coagulation assays may help interpret the results of coagulation assays and contribute to the future development of diagnostic tools. The synchronized plasma levels of vitamin K–dependent proteins with opposite functionalities may compensate a propensity to hyper- or hypocoagulability in a normal population.

Key Words
  • Global coagulation assay
  • Prothrombin time
  • Activated partial thromboplastin time
  • Thrombin generation assay
  • Coagulation factor
  • Anticoagulant factor
  • Vitamin K

Normal hemostasis is a state of equilibrium between procoagulant and anticoagulant factors in circulating blood. An imbalance in these hemostatic systems produces either hemorrhagic or thrombotic conditions. For the early detection and management of coagulopathy, global coagulation tests that can detect delicate hemostatic changes are essential. Prothrombin time (PT) and activated partial thromboplastin time (aPTT) are widely used in clinical laboratories as routine screening tests of the coagulation system. However, these tests are not strongly predictive of hemorrhagic risk and vary among individuals.1 There is increasing evidence that the evaluation of thrombin generation gives useful information regarding coagulation status.2-5 In the thrombin generation assay (TGA) using an automated calibrated thrombogram, thrombin generation curves can be operationally characterized as displaying initiation, propagation, and termination phases. After stimulation with tissue factor (TF), the consequent formation of endogenous thrombin potential (ETP) is measured in plasma. The ETP has been shown to be a good indicator of prothrombotic4,6-8 and hemorrhagic tendencies.3,9,10 Several studies have shown that elevated ETP is associated with a risk for venous thromboembolism 2,4,8,11; elevated levels of factor (F) XI, FIX, FVIII, FX, prothrombin, or fibrinogen are seen to be correlated with a risk for venous thromboembolism risk.12-16 Although elevated procoagulant levels were shown to correlate with abnormal in vitro thrombin generation,17-20 their individual relationships with in vivo thrombin generation remain unclear. The subclassification of individuals with elevated thrombin generation based on the underlying changes in different coagulation factors and anticoagulant factors may help in understanding the clinical risk of elevated thrombin generation.

In previous studies of normal coagulation, the normal ranges of procoagulant and anticoagulant factors have been generally found to be between 50% and 150%.21-23 However, a number of studies have shown that variations of coagulation proteins in this range are associated with thrombotic risk.23-26 An individual without apparent coagulopathy whose coagulation factors fall within the reference range may have a potential risk for thrombosis or hemorrhage because of an imbalance between coagulation and anticoagulation factors. Given that global coagulation assays such as PT, aPTT, and TGA are widely used for the detection of hemostatic dysfunction, it is necessary to know how individual levels of coagulation and anticoagulant factors influence global coagulation test results.

Sporadic studies have reported the coagulatory determinants of aPTT27 and TGA18 in normal populations. However, to our knowledge, no report has compared the influence of coagulation and anticoagulant factors on 3 global coagulation tests (PT, aPTT, TGA) through their simultaneous measurement in the same healthy population. In this study, we investigated the influence of coagulation and anticoagulant factors on 3 global coagulation tests, including PT, aPTT, and TGA, in a population of normal healthy adults.

Materials and Methods

Study Population and Specimen Collection

A total of 252 consecutive apparently healthy adults for whom coagulation screening tests were requested in routine health visits were included in the present study. The mean age of the subjects was 54 years (range, 24-78 years). All of the subjects were Korean; 151 were male and 101 were female subjects. Mean ages of the men and women were similar (54.7 and 52.9 years, respectively; P = .245). Peripheral venous blood samples were collected in commercially available tubes containing 0.109 mol/L sodium citrate (Becton Dickinson, San Jose, CA). Plasma was separated by centrifugation of whole blood at 1,550g for 15 minutes within 2 hours after blood collection. The aliquots of plasma were stored at –80°C. This study was reviewed and approved by the institutional review board of Seoul National University College of Medicine (Seoul, Korea).

Thrombin Generation Assay and Other Coagulation Tests

TF-triggered thrombin generation in platelet-poor plasma was measured using the calibrated automated thrombogram method (Thrombinoscope, Maastricht, The Netherlands) as described previously.28 Briefly, 20 μL of reagent containing TF at a final concentration of 5 pmol/L or 1 pmol/L (PPP Reagent 5 pmol/L or PPP Reagent Low, respectively; Thrombinoscope BV), along with phospholipids or thrombin calibrators, was dispensed into each well of round-bottomed 96-well plates, and then 80 μL of test plasma specimen was added. After the addition of 20 μL of fluorogenic substrate in HEPES buffer containing calcium chloride, the fluorescent signal was read in a Fluoroskan Ascent fluorometer (Thermo Labsystems, Helsinki, Finland) and thrombin generation curves were calculated using the Thrombinoscope software (Thrombinoscope). Thrombin generation curves were evaluated using parameters that describe the initiation, propagation, and termination phases of thrombin generation, namely, lag time, ETP, and peak thrombin height (peak thrombin). The lag time is defined as the time to reach one sixth of the peak height and is a measure of the initiation phase. It is equivalent to the clotting time. The peak height is defined as the maximum thrombin concentration produced. ETP is the area under the thrombin generation curve and represents the total amount of generated thrombin.

Coagulation tests, including PT, aPTT, and factor assays, were performed on an automated coagulation analyzer, ACL TOP (Beckman Coulter, Fullerton, CA). PT was measured using the HemosIL RecombiPlasTin reagent (Instrumentation Laboratory, Milan, Italy), and aPTT was measured using the SynthASil reagent (Instrumentation Laboratory). Fibrinogen was measured using the HemosIL Fibrinogen-C XL reagent (Instrumentation Laboratory) based on the Clauss method. Coagulation factors were tested using a PT-based clotting assay with the HemosIL RecombiPlasTin reagent (for FII, FV, FVII, and FX) and an aPTT-based clotting assay using the SynthASil reagent (for FVIII, FXI, FXI, and FXII). Antithrombin, protein C, and protein S activity were determined with chromogenic assays (HemosIL liquid antithrombin and HemosIL Protein C; Instrumentation Laboratory) and an immunoassay (HemosIL Free Protein S, Instrumentation Laboratory).

Statistical Analysis

Data were compared using the Mann-Whitney U test and Kruskal-Wallis analysis for continuous variables and the χ2 test for categorical variables. The coefficient of variation was calculated as the ratio of the standard deviation to the mean so as to assess interindividual variability. Correlations are expressed as Pearson coefficients. The relative effects of coagulation and anticoagulant factors on thrombin generation, PT, and aPTT were assessed using multiple linear regression analysis. For each model, the adjusted R2 and the standardized regression coefficients (β) of the independent variables were calculated. All analyses were carried out using SPSS 12.0 software (SPSS, Chicago, IL). A probability value (P) of less than .05 was considered significant.

Results

Coagulation Protein Levels and Thrombin Generation Parameters in a Healthy Population

The mean and 2.5 to 97.5 percentile values of each coagulation and anticoagulant factor and the 3 global coagulation tests (PT, aPTT, and TGA) are summarized in Table 1. Women had higher levels of FII and antithrombin and lower levels of FIX and protein S. Women had lower peak thrombin and ETP values than men on stimulation with 1 pmol/L TF. The reference ranges of coagulation and anticoagulant factors from our study were generally similar to those of previous studies.29,30 Significantly higher lower limits of FII were found in our study compared with reference intervals from the previous study (77%-121%), but the median value (114%) was similar to those reported previously (95%-110%).30 Compared with 1 pmol/L TF, TGA results with 5 pmol/L TF demonstrated shortened lag time and time to peak as well as higher peak thrombin and ETP, consistent with previous studies.18,20 With 1 pmol/L TF, the lag time varied 2-fold (range, 6.3-15.2 min) and peak thrombin varied over a 4-fold range in this healthy population. The interindividual variabilities in lag time, peak thrombin, and ETP with 1 pmol/L TF were 23.7%, 34.1%, and 23.1%, respectively, consistent with previously reported findings.20,31-33

View this table:
Table 1

Correlation Among Levels of Coagulation Factors, Anticoagulant Factors, and Other Laboratory Results

No significant correlation was seen between PT and any of the TGA parameters. With 1 pmol/L TF, lag time showed a positive correlation with aPTT (r = 0.161), whereas peak thrombin and ETP showed inverse correlations with aPTT (r = –0.449 and –0.277, respectively). With 5 pmol/L TF, peak thrombin and ETP showed inverse correlations with aPTT (r = –0.270 and –0.129, respectively).

Weak to moderate correlations were observed between various coagulation factors and anticoagulant factors Table 2. Interestingly, the levels of vitamin K–dependent anticoagulant factors such as protein C and protein S showed strong correlations with those of vitamin K–dependent coagulation factors including FII, FVII, and FX. Correlations among vitamin K–dependent coagulation factors (FII, FVII, FIX, and FX) were higher than among other non–vitamin K–dependent coagulation factors.

View this table:
Table 2

FV, FVIII, FIX, and FXII were positively correlated to age (r = 0.139 to approximately 0.263) Table 3. FII and antithrombin levels tended to decrease with age (r = –0.119 and r = –0.131, respectively). The various laboratory test results revealed weak correlations with several coagulation proteins. Of note, cholesterol and triglyceride levels showed significant correlations with vitamin K–dependent coagulation proteins including FII, FVII, FIX, FX, protein C, and protein S (r = 0.159-0.370).

View this table:
Table 3

Effects of Coagulation and Anticoagulant Factors on PT and aPTT

The concentrations of FVII and FV decreased significantly with prolonged PT Figure 1A. On the other hand, FV, FVIII, FIX, FX, FXI, and FXII decreased with prolonged aPTT Figure 1B. In a multiple linear regression analysis Table 4, the β value of FV for the dependent variable PT was –0.420. This value means that when FV increases by 1 standard deviation (18.3), PT shortens by 0.42. The strongest determinants of PT results were FV and FVII. FII was also a determinant of PT. The aPTT value was mainly dependent on FVIII and FXII levels and, to a lesser extent, negatively determined by antithrombin and protein C.

Figure 1

Changes in the mean levels of coagulation factors and anticoagulant factors based on prothrombin time (PT), activated partial thromboplastin time (aPTT), and lag time. On the x-axis, each point represents the mean value of coagulation factor levels in the intervening range indicated (–2 SD indicates the range of mean – 2.5 SD to mean – 1.5 SD; –1 SD, mean – 1.5 SD to mean – 0.5 SD; mean, mean – 0.5 SD to mean + 0.5 SD; 1 SD, mean + 0.5 SD to mean + 1.5 SD; 2 SD, mean + 1.5 SD to mean + 2.5 SD) of PT (A), aPTT (B), lag time at 5 pmol/L tissue factor (TF) (C), and lag time at 1 pmol/L TF (D). Solid lines indicate P < .05 by test for trend, with meaningful directional changes compatible with test results. AT, antithrombin; F, factor; PC, protein C; PS, protein S.

View this table:
Table 4

Effects of Coagulation and Anticoagulant Factors on the Thrombin Generation Assay

Lag times were progressively prolonged with increasing concentrations of anticoagulant factors such as antithrombin and protein S Figure 1C and Figure 1D. In a multiple linear regression analysis, the significant negative determinants of lag time with 5 pmol/L TF were FVII and FXII, and the significant positive determinants of the same were antithrombin and protein S (Table 4). With 1 pmol/L TF, the only anticoagulant that emerged as a significant positive determinant of lag time was protein S. FX was a positive determinant, revealing an inverse influence compared with other coagulation factors.

As expected, apparent increases in FVIII, FIX, and FXI were seen as peak thrombin increased with both 1 pmol/L and 5 pmol/L TF Figure 2A and Figure 2B. Antithrombin levels decreased as peak thrombin and ETP increased. In a multiple linear regression analysis, the significant positive determinants of peak thrombin were FVIII and FIX with both 5 pmol/L and 1 pmol/L TF (Table 4). FII was a positive determinant of peak thrombin with 5 pmol/L TF. FV and FX had a negative influence on peak thrombin. Antithrombin was a significant negative determinant of peak thrombin with 5 pmol/L TF. The significant determinants of ETP were FII, FIX, and antithrombin with 5 pmol/L TF and FIX and antithrombin with 1 pmol/L TF.

Figure 2

Changes in the mean concentrations of coagulation factors and anticoagulant factors based on peak height and endogenous thrombin potential (ETP) of thrombin generation assay. Each dot indicates the mean of coagulation protein levels in the intervening range indicated (–2 SD indicates the range of mean – 2.5 SD to mean – 1.5 SD; –1 SD, mean – 1.5 SD to mean – 0.5 SD; mean, mean – 0.5 SD to mean + 0.5 SD; 1 SD, mean + 0.5 SD to mean + 1.5 SD; 2 SD, mean + 1.5 SD to mean + 2.5 SD) of peak height at 5 pmol/L tissue factor (TF) (A), peak height at 1 pmol/L TF (B), ETP at 5 pmol/L TF (C), and ETP at 1 pmol/L TF (D). Solid lines indicate P < .05 by test for trend, with meaningful directional changes compatible with test results. AT, antithrombin; F, factor; PC, protein C; PS, protein S.

Discussion

PT and aPTT are routine coagulation screening tests that estimate deficiencies in coagulation factors. Because the normal values of these global tests showed wide interindividual variations in our study, it is assumed that wide variations in coagulation protein levels affect the PT and aPTT results. Although several reports have described the determinants of PT and aPTT in pathologic conditions, such as oral anticoagulation34 or hemophilia,1 the dependence of these tests on the levels of coagulation and anticoagulant factors in a healthy population has not been studied in detail. By simultaneously measuring coagulation and anticoagulant factors, as well as global coagulation assays in plasma specimens from normal healthy persons, our study demonstrated that the PT value was significantly determined by FV and FVII and that aPTT was mainly dependent on FVIII and FXII. Moreover, the levels of anticoagulant factors such as antithrombin and protein C were shown to be significant determinants of the aPTT value.

Given that the clotting process is initiated by the binding of TF to FVII(a), the FVII level may be important in determining the PT value. Interestingly, we observed that FV levels also significantly influenced PT values. The endpoint of PT is defined as the time at which only around 5% of all physiologically relevant thrombin is formed.35 To generate an initial small amount of thrombin, the level of FV appears to be important in the PT system. Although the principle is not known in detail, the knowledge that FV and FVII have significant effects on PT is of note for the interpretation of PT values in a normal population.

As expected, FVIII and FXII levels were identified as major determinants of aPTT in our results. This is in agreement with observations from a previous study that FVIII and FXII were the 2 strongest contributors determining the aPTT level.27 Because FVIII is known to be a risk factor for thrombosis,36 aPTT may be used as a supplementary indicator of thrombosis risk. In our study, FXII was also a major determinant of aPTT. If the reagent used in the aPTT test first activates a large amount of FXII, this activated FXII can drive the subsequent coagulation cascade until a small amount of thrombin is formed. Because FXII is not the main contributor to the physiologic coagulation process,1 the dependency of aPTT on the FXII value is not a desirable finding because it indicates poor correlation of aPTT with the hemorrhagic potential. Until now, to our knowledge, the effect of anticoagulant protein levels on the aPTT value has not been reported. The current study demonstrated that the aPTT could be decreased by low levels of antithrombin and protein C. The aPTT might be decreased in individuals who are heterozygous carriers of thrombophilia and contain mildly low levels of antithrombin or protein C. Future studies in a heterozygous carrier are required to confirm this.

In our TGA experiments, the TGA lag time with 5 pmol/L TF was negatively determined by FVII and FXII. Peak thrombin was significantly dependent on FVIII and FIX levels with both 1 pmol/L and 5 pmol/L TF. In a previous study using an artificial plasma mixture spiked with coagulation factors, peak thrombin was found to be a sensitive parameter for the detection of FVIII and FIX deficiency with 1 pmol/L TF but not with 5 pmol/L TF.19 However, in our study, using normal plasma, both the 1 pmol/L and 5 pmol/L TF-stimulated peak thrombin levels were good representatives of FVIII and FIX levels. This finding suggests the potential usefulness of peak thrombin as a screening test for hemophilia carriers; however, future studies are required to confirm this. Prothrombin was also a determinant of peak thrombin with 5 pmol/L TF but not 1 pmol/L TF. This is in agreement with a previous study that used 13.6 pmol/L TF.18 Because a very small fraction of the prothrombin present in plasma is activated at low TF, the effect of prothrombin on peak thrombin seems to be significant only at high TF. In our study, the ETP values with both 1 pmol/L and 5 pmol/L TF were positively dependent only on FIX but not on FVIII. Given that peak thrombin was determined by both FVIII and FIX, ETP is considered a less sensitive parameter than peak thrombin. The high sensitivity of peak thrombin to FVIII and FIX could be used for diagnostic purposes in the future.

Antithrombin was an important determinant of ETP. Antithrombin is a serine protease inhibitor that regulates thrombin and FXa and, to a lesser extent, FIX, FXI, FXII, and FVII. It regulates the activities of coagulation factors that have already formed and therefore has a larger contribution to the ablation of the propagation phase37; this can explain the importance of antithrombin as a determinant of ETP. The TGA lag time with 5 pmol/L TF was determined by the anticoagulant proteins antithrombin and protein S. The activated protein C/protein S system, which inactivates activated FV and FVIII, may play a role in regulating the initiation phase of coagulation.37 Although the effect of protein S on lag time can be partially explained through this knowledge of the coagulation cascade, the detailed mechanism requires future investigation.

Our study found significant positive correlations between several coagulation factors and anticoagulant factors, which are likely based on the specific conditions of the individual. Of note, vitamin K–dependent coagulation factors (FII, FVII, FIX, FX) significantly correlated with vitamin K–dependent anticoagulant factors (protein C, protein S), suggesting that the vitamin K status determined the plasma concentrations of coagulation and anticoagulant factors in each individual. Moreover, the levels of vitamin K–dependent proteins correlated with levels of blood lipids such as cholesterol and triglycerides. Consistent with our findings, one study reported significant correlations of vitamin K–dependent coagulation factors (FII, FVII, FX) with cholesterol and triglycerides in healthy adults.38 Vitamin K–dependent coagulation proteins can bind to triglyceride-rich lipoproteins via hydrophobic interactions. Although we could not measure the concentration of circulating vitamin K in our healthy population, it is assumed that high levels of circulating lipids can increase the in vivo reservoir of lipid-soluble vitamin K and can eventually drive the synthetic cycle of vitamin K–dependent coagulation proteins forward in the liver. This might be a part of a mechanism by which circulating lipids influence the levels of vitamin K–dependent coagulation proteins. In addition, it is also possible that subclinical liver disease undetectable with routine liver function tests could be the common factors causing the correlation between the coagulation proteins and lipid levels. Of note, the levels of vitamin K– dependent anticoagulation proteins were positively correlated with those of vitamin K–dependent coagulation factors in our results. These synchronized plasma levels of proteins with opposite functionalities is likely to offset a propensity toward hypercoagulability or hypocoagulability in global coagulation assays. Future studies are required to demonstrate in detail the mechanisms of this potentially interesting finding.

The current study has 3 potential limitations. First, the level of TF pathway inhibitor (TFPI) was not investigated despite the fact that TFPI plays an inhibitory role in the coagulation cascade through the inactivation of activated FX and FVII. Second, to prevent contact activation, contact factor inhibitor can be used for the measurement of TGA at low concentrations of TF.39 In our experiments, we did not use a contact factor inhibitor in TGA; therefore, the level of FXII as a determinant of lag time with 1 pmol/L TF can be attributed to in vitro contact activation. Third, we could not investigate the effects of factors known to be correlated with hypercoagulable status such as menstrual status, hormonal treatment, or obesity.

In summary, the current study demonstrated significant correlations among vitamin K–dependent coagulation protein and blood lipids, suggesting that the plasma levels of vitamin K–dependent coagulation proteins are controlled by the individual's lipid metabolism status in normal physiology. The global coagulation assays (PT, APTT, and TGA) were dependent on the levels of anticoagulant proteins as well as coagulation factors. In particular, the PT value was significantly determined by FV and FVII, and aPTT was dependent on FVIII and FXII. Moreover, the levels of anticoagulant factors such as antithrombin and protein C were shown to be significant determinants of aPTT. The TGA lag time was mainly determined by FVII and FXII. Peak thrombin was significantly dependent on FVIII and FIX levels. ETP was dependent only on FIX, suggesting that peak thrombin is a more sensitive parameter than ETP. Antithrombin (for ETP and lag time) and protein S (for lag time) were important contributors to TGA inhibition. It is expected that these findings will help in interpreting the results of global coagulation assays in a normal population and contribute to the development of alternative diagnostic tools in the future. To clearly determine the effects of specific factor deficiencies and excesses and to elucidate the detailed underlying mechanisms, further studies using factor-deficient plasma specimens would be required.

Acknowledgments

This work was supported by a National Research Foundation of Korea grant from the Korean Government (2009-0075731).

References

View Abstract