Performance of coagulation tests in patients on therapeutic doses of dabigatran: a cross-sectional pharmacodynamic study based on peak and trough plasma levels

Authors


  • Manuscript handled by: M. Cattaneo
  • Final decision: F. R. Rosendaal, 22 May 2013

Correspondence: Emily M. Hawes, Department of Family Medicine, University of North Carolina School of Medicine, 590 Manning Drive, Campus Box 7595, Chapel Hill, NC 27599, USA.

Tel.: +1 919 843 8408; fax: +1 919 966 6125.

E-mail: ehawes@unch.unc.edu

Summary

Background

Knowledge of anticoagulation status during dabigatran therapy may be desirable in certain clinical situations.

Objective

To determine the coagulation tests that are most useful for assessing dabigatran's anticoagulant effect.

Methods

Peak and trough blood samples from 35 patients taking dabigatran 150 mg twice daily, and one sample each from 30 non-anticoagulated individuals, were collected. Mass spectrometry and various coagulation assays were performed. ‘Therapeutic range’ was defined as the range of plasma dabigatran concentrations determined by mass spectrometry between the 2.5th and 97.5th percentiles of all values.

Results

The therapeutic range was 27–411 ng mL−1. The prothrombin time (PT) and activated partial thromboplastin time (APTT), determined with multiple reagents, and activated clotting time (ACT) were insensitive to therapeutic dabigatran: 29%, 18% and 40% of samples had a normal PT, APTT, and ACT, respectively. However, normal PT, ACT and APTT ruled out dabigatran levels above the 75th percentile. The thrombin clotting time (TCT) correlated well and linearly with dabigatran levels below the 50th percentile, but was unmeasurable above it. The dilute thrombin time, ecarin clotting time and ecarin chromogenic assay showed linear correlations with dabigatran levels over a broad range, and identified therapeutic and supratherapeutic levels.

Conclusions

The prothrombin time, APTT and ACT are often normal in spite of therapeutic dabigatran plasma levels. The TCT is useful for detecting minimal dabigatran levels. The dilute thrombin time and chromogenic and clotting ecarin assays accurately identify therapeutic and supratherapeutic dabigatran levels. This trial is registered at www.clinicaltrials.gov (#NCT01588327).

Introduction

Although anticoagulation monitoring of patients on therapeutic doses of the new oral anticoagulant dabigatran is not routinely necessary, understanding of a patient's anticoagulation status may be desirable in certain clinical situations, such as when assessing for subtherapeutic or supratherapeutic levels when thromboembolic or hemorrhagic events occur, determining minimal residual drug effects prior to major surgery or thrombolytic therapy, or assessing the efficacy of reversal strategies when managing a patient with major bleeding.

Limited information exists on the performance of coagulation tests in the determination of the level of anticoagulation in patients treated with therapeutic doses of dabigatran. Much of the existing information is derived from in vitro studies of dabigatran-spiked plasma, or ex vivo studies testing plasma samples from healthy volunteers who received various dabigatran doses [1-8]. Information on ex vivo data from patients treated with dabigatran is limited, owing to the use of doses other than 150 mg twice daily, lack of reporting of detailed data, testing of only one reagent for a particular coagulation assay, and evaluation of coagulation response in patients enrolled into pharmaceutical company-sponsored studies with strict enrollment exclusion criteria [9-13]. Very limited information exists on the performance of point-of-care (POC) coagulation assays in dabigatran-treated patients [1, 4, 14, 15].

Professional medical organizations have developed guidelines detailing how to appropriately evaluate patients on the new oral anticoagulants with coagulation tests. However, recommendations have been made on the basis of limited data on test performance [13, 16-18].

Given the potential need to assess anticoagulant status in certain patients, data on the performance of various coagulation assays in plasma and whole blood from actual patients treated with dabigatran are needed. The primary objective of this study was to determine the coagulation tests that are most useful for assessing dabigatran's anticoagulant effect, by correlating the results of various coagulation assays with plasma drug levels in patients taking therapeutic doses of dabigatran.

Patients and methods

Patients

The study was approved by the University of North Carolina (UNC) Institutional Review Board, and registered at www.clinicaltrials.gov as #NCT01588327. All 65 subjects provided written informed consent. Patients taking Food and Drug Administration-approved therapeutic doses of dabigatran at steady state receiving care at the UNC were eligible for inclusion. Volunteers who were not taking an anticoagulant were enrolled as controls.

Methods

For patients taking dabigatran, blood samples were obtained immediately prior to the next dabigatran dose (trough) and, on the same day, 2.5 h following the oral dosing (peak). A random blood sample was obtained for each control. For coagulation testing, blood samples were centrifuged at 1600 g for 15 min at 25 °C, spun again at 1600 g for 5 min to obtain platelet- poor plasma, aliquoted, and stored at − 70 °C until being analyzed. Whole blood samples obtained by venipuncture were added to disposable prothrombin time (PT), activated partial thromboplastin time (APTT) and activated clotting time-low range (ACT) cuvettes, and analyzed with the HEMOCHRON® (International Technidyne Corporation, Edison, NJ, USA) Signature Elite POC device. Table S1 contains a list of the reagents and devices used to perform the following tests: PT, APTT, ecarin chromogenic assay (ECA), ecarin clotting time (ECT), thrombin clotting time (TCT), ACT, and dilute thrombin time (dTT). Individual reagents and coagulation analyzers were used and operated according to the manufacturers’ instructions. Plasma dabigatran levels were determined by liquid chromatography–mass spectrometry by Boehringer-Ingelheim in Biberach, Germany.

Table 1. Summary of results for each coagulation test
TestReagent and device n Slope R 2 * Predicted rangePredicted trough rangePredicted peak range§Normal reference rangeMP** (%)
  1. ACT, activated clotting time; APTT, activated partial thromboplastin time; dTT, dilute thrombin time; ECA, ecarin chromogenic assay; ECT, ecarin clotting time; MP, misprediction; n, number of individual samples analyzed; PT, prothrombin time; TCT, thrombin clotting time; UCDMC, University of California Davis Health System; UNC, University of North Carolina. *The R-squared (R2) value is the square of the correlation coefficient. †The predicted range comprises the predicted test values for the therapeutic dabigatran range based on the fitted line. ‡The predicted trough range comprises the predicted test values for the therapeutic dabigatran range at trough. §The predicted peak range comprises the predicted test values for the therapeutic dabigatran range at peak. ¶The normal reference range represents the range of coagulation test results for all control participants. **The percentage of MP represents how often the coagulation test values were in the normal reference range while the plasma dabigatran concentration was in the therapeutic range. ††For APTT ratio, curves were fitted, and the slope can be considered to be scale parameter. The form of these curves are math formula. ‡‡Excludes unmeasurable TCT results. §§Below the assay's lower limit of detection.

PT ratioHEMOCHRON® Signature Elite [1]670.00290.861.09–2.21.06–1.601.14–2.420.86–1.2028
TriniCLOT™ PT HTF on MDA [2]690.00150.601.06–1.631.05–1.331.09–1.740.89–1.0613
HemosIL® RecombiPlasTin 2G on ACL TOP 700 (UNC) [3]690.00110.661.05–1.461.04–1.241.07–1.540.74–1.1029
HemosIL® RecombiPlasTin 2G on ACL TOP 700 (UCDMC) [4]690.00110.621.05–1.461.04–1.241.07–1.540.86–1.1443
Innovin® on Sysmex® CS-2000i System [5]690.00100.791.03–1.391.02–1.201.05–1.470.89–1.0724
Innovin® on BCS® XP System [6]690.00090.751.03–1.381.02–1.191.05–1.450.87–1.0839

APTT

ratio††

Dade® Actin® on Sysmex® CS-2000i System [1]690.0630.771.28–2.231.22–1.861.37–2.340.80–1.2619
HemosIL® SynthASil on ACL TOP 700 (UNC) [2]690.0590.811.27–2.161.21–1.811.36–2.270.82–1.3526
HemosIL® SynthASil on ACL TOP 700 (UCDMC) [3]690.0590.781.27–2.151.21–1.801.35–2.250.86–1.1911
HEMOCHRON® Signature Elite [4]670.0570.751.26–2.121.21–1.781.35–2.220.52–1.3219
Dade® Actin® FS on BCS® XP System [5]690.0560.821.24–2.091.19–1.761.33–2.190.90–1.2111
TriniCLOT™ automated aPTT on MDA [6]690.0520.741.24–2.021.19–1.711.31–2.120.86–1.2823
Dade® Actin® FSL on Sysmex® CS-2000i System [7]690.0440.791.20–1.861.16–1.601.27–1.940.89–1.1916
ACTHEMOCHRON® Signature Elite670.3110.71155–275152–211161–29899–18040

TCT

ratio

HemosIL® Thrombin Time on ACL TOP 700 [1]41‡‡0.1370.974.73–57.23.47–29.27.16–67.60.76–1.180
Pacific Hemostasis® Thrombin on Diagnostica Stago STart 4 Hemostasis Analyzer [2]40‡‡0.090.753.37–37.72.54–19.34.96–44.50.83–1.230
ECA (ng mL−1)ECA-T on BCS® XP System [1]690.9170.9924.0–37615.6–18840.3–445 §§ 0
ECA-T on Stago STA Compact [2]690.7640.9448.8–34241.8–18562.4–400 §§ 8.6
ECA-T on Diagnostica Stago STA-R Evolution® [3]690.6850.9534.4–29728.1–15746.6–349 §§ 0
ECTECARIN Prothrombin Activator on Diagnostica Stago STart 4 Hemostasis Analyzer690.3220.9834.7–15831.7–92.240.4–18322.0–26.80
dTTHemoclot® Thrombin Inhibitor on BCS® XP System670.7110.9235.7–30629.2–16148.2–359 §§ 13

Definition of ‘therapeutic range’

As there is no published ‘therapeutic range’ for dabigatran, owing to the absence of investigations correlating thrombotic and hemorrhagic complications with plasma drug levels, our study defined ‘therapeutic range’ as the range of plasma dabigatran concentrations determined by mass spectrometry that fell between the 2.5th and 97.5th percentiles of all dabigatran values (both trough and peak) of the study patients on dabigatran. ‘Subtherapeutic’ levels were defined as values below the lowest value of the ‘therapeutic range’, and ‘supratherapeutic’ levels as those above the highest value of the ‘therapeutic range’.

Statistical analysis

Differences in demographic and clinical characteristics between groups were evaluated with Fisher's exact tests for categorical variables and Wilcoxon rank sum tests for continuous variables. Mean and standard deviation (SD) are reported for continuous variables.

The normal range of each assay for each coagulation reagent–instrument combination was determined by using the results obtained in control subjects (n = 27–30). For assays in which ratios were used (PT, APTT, and TCT), the mean value of the normal range was determined, and the individual patient's test result was divided by this constant.

For each test and reagent, the same individual statistical analysis was performed. First, the relationship between the test and plasma dabigatran levels as determined by mass spectrometry was evaluated with regression models. For all tests except the APTT, a straight line best described the relationship, and lines of the y = slope(x) + intercept were fitted. A curve best described the relationship between dabigatran and the APTT, and lines of the math formula were fitted. In Table 1, the number in the slope column can be considered to be a scale parameter, and can still be compared with other APTT slopes. Once the line was fitted, the adjusted R-squared value and correlation coefficients (r) were determined. Coagulation assays with an R-squared value of 0.9–1 were considered to have a strong correlation with plasma dabigatran levels. The predicted range was determined by substituting 27 and 411 (the therapeutic range) for the dabigatran levels, and calculating the predicted coagulation assay values. The ‘misprediction percentage’ represents how often the coagulation test values were in the normal reference range while the plasma dabigatran concentration was in the therapeutic range. These misprediction percentages were calculated for all reagents in each assay category. All statistical analyses were performed with sas v 9.2 statistical software (SAS Institute Inc., Cary, NC, USA, www.sas.com).

Results

Study population

Thirty-five patients taking dabigatran 150 mg twice daily and 30 control subjects were enrolled between September 2011 and February 2012. Eighty-three per cent of patients were taking dabigatran for atrial fibrillation, and 17% were taking it for venous thromboembolism. The mean age was 38 years (SD 15) for control subjects and 61 years (SD 12) for dabigatran patients (P < 0.0001). Seventy-one per cent of dabigatran patients and 30% of controls were male (P = 0.001). There was no statistically significant difference in plasma dabigatran concentrations by gender, as determined by mass spectrometry (P = 0.3). The mean serum creatinine level of patients was 80.4 mm (SD 18.6; 0.91 mg dL−1 [SD 0.21]) and creatinine clearance was 113 mL min−1 (SD 32). The mean patient weight was 95 kg (SD 20), and the body mass index (BMI) was 30.4 kg m−2 (SD 5.1). Figures S1 and S2 show the effect of renal function, body weight, BMI and patient age on plasma dabigatran levels.

Therapeutic range

The on-therapy plasma drug level range of all values was 18–206 ng mL−1 at trough, and 45–487 ng mL−1 at peak. The mean time to peak blood draw after dabigatran dose administration was 151 min (95% confidence interval 146–156), ranging from 120 to 180 min. The ‘therapeutic range’ for plasma drug levels, representing 95% of all plasma levels, was 27–411 ng mL−1 (median of 103 ng mL−1); the 50th percentile for dabigatran concentrations was 103 ng mL−1, and the 75th percentile was 164 ng mL−1.

PT

The results for each coagulation assay were correlated with plasma drug levels (Table 1). The performance of the six different PT reagent–coagulometer combinations is shown in Fig. 1. Innovin® (Dade Behring Inc., Marburg, Germany) and HemosIL® (Instrumentation Laboratory, Bedford, MA, USA) RecombiPlasTin 2G are two of the PT reagents most commonly used in the USA [19]. All PT reagents performed similarly, except HEMOCHRON Signature Elite® (a POC instrument) and TriniCLOT™ (Tcoag Ireland Ltd, Wicklow, Ireland) PT HTF on MDA. The PT assays are insensitive to dabigatran, having a flat slope of the best-fit line (Fig. 1). In addition, they have low R-squared values and high misprediction percentages. Both the adjusted R-squared value and slope were highest for HEMOCHRON® Signature Elite, with a misprediction percentage of 28%. TriniCLOT™ PT HTF on MDA had the lowest percentage of misprediction and the second highest slope, but also the lowest adjusted R-squared value. In the overall analysis of results from all six PT reagents, 30% of PT values were normal despite therapeutic dabigatran drug levels. For all reagents except HemosIL® RecombiPlasTin 2G on ACL TOP 700 (University of California Davis Health System, UCDMC), there was an absence of normal PT values with plasma dabigatran concentrations above the 75th percentile. For HemosIL® RecombiPlasTin 2G on ACL TOP 700 (UCDMC), four PT results were normal despite dabigatran levels above the 75th percentile. No normal PT results for TriniCLOT™ PT HTF and only one normal PT value for HEMOCHRON® Signature Elite were present above the 50th percentile of the plasma dabigatran levels.

Figure 1.

The effect of dabigatran plasma levels on prothrombin time (PT) and activated partial thromboplastin time (APTT) assays. The vertical lines in the figures represent the ‘therapeutic range’ of dabigatran plasma concentrations. The horizontal lines represent the normal range of all of the combined PT ratio (A) or APTT ratio (B) values in control subjects. The best-fit lines in (A) appear in the following descending order: HEMOCHRON® Signature Elite (x, line 1), TriniCLOT™ PT HTF on MDA (○, line 2), HemosIL® RecombiPlasTin 2G on ACL TOP 700 (University of North Carolina [UNC]) (●, line 3), HemosIL® RecombiPlasTin 2G on ACL TOP 700 (UCDMC) (∆, line 4), Innovin® on Sysmex® CS-2000i System (□, line 5), and Innovin® on BCS® XP System (+, line 6). Lines 3 and 4 almost overlap. The best-fit lines in (B) appear in the following descending order: Dade® Actin® on Sysmex® CS-2000i System (x, line 1), HemosIL® SynthASil on ACL TOP 700 (UNC) (○, line 2), HemosIL® SynthASil on ACL TOP 700 (UCDMC) (●, line 3), HEMOCHRON® Signature Elite (∆, line 4), Dade® Actin® FS on BCS® XP System (□, line 5), TriniCLOT™ automated aPTT on MDA (+, line 6), and Dade® Actin® FSL on Sysmex® CS-2000i System (*, line 7).

APTT

Figure 1 also shows the effect of dabigatran on the different APTT reagent–coagulometer combinations. Dade® Actin® FSL and HemosIL® Synthasil are two of the APTT reagents most commonly used in the USA [19]. All APTT reagents performed similarly. Although the best-fit curve is initially steep at low plasma dabigatran levels, the curve flattens with increasing dabigatran levels. The R-squared values for the APTT tests were higher than those for PT, ranging from 0.74 to 0.82, and the percentages of misprediction were lower, ranging from 11% to 26%. Seventy-nine per cent of all APTT values were elevated above the normal range in the presence of therapeutic dabigatran, and 18% of values were normal despite therapeutic plasma drug levels. For all reagents and devices, three or fewer normal APTT values were present in patients with a plasma dabigatran concentration above the 50th percentile. Figure 2 shows the percentages of PT and APTT results within the normal range of a given test at different dabigatran plasma concentrations.

Figure 2.

The percentage of results within the normal range of a given test at different dabigatran plasma concentrations. The graphs show, for each prothrombin time (PT) and activated partial thromboplastin time (APTT) reagent, the percentage of test results (on the y-axis) from the dabigatran-treated patients that are in the normal range at the plasma dabigatran concentrations depicted on the x-axis. Example: with APTT reagent no. 3, 15% of APTT ratios are in the normal range when the dabigatran plasma concentration is up to 100 ng mL−1; with reagent no. 2, 35% are in the normal range.

ACT

The normal values for the ACT in controls ranged from 99 to 180 s. Forty per cent of patients had normal ACT values despite having therapeutic dabigatran levels (Fig. 3). Fifty-six per cent of patients with therapeutic dabigatran had ACT values above the normal range. There were no normal ACT values with plasma dabigatran levels above the 75th percentile.

Figure 3.

The performance of activated clotting time – low range (ACT-LR) and thrombin clotting time (TCT) in relation to dabigatran plasma levels. The vertical lines represent the ‘therapeutic range’ of dabigatran plasma concentrations. The horizontal lines denote the range of ACT results (A) and both of the TCT ratio results (B) in control subjects. Unmeasurable TCT values are not included in the analysis. Line 1 is the best-fit line for HemosIL® Thrombin Time on ACL TOP 700, and line 2 is the best-fit line for Pacific Hemostasis® Thrombin on Diagnostica Stago STart 4 Hemostasis Analyzer.

TCT

The two TCT tests performed similarly, and had linear relationships and strong correlations with plasma dabigatran concentration (Table 1; Fig. 3). Ninety-eight per cent of TCT values in the dabigatran arm were elevated above the normal range for controls, and no samples had normal TCT values despite therapeutic dabigatran levels. With HemosIL® Thrombin Time on ACL TOP, TCTs were often unmeasurable at or above a plasma dabigatran concentration of 108 ng mL−1, and, at dabigatran levels > 138 ng mL−1, all TCTs were unmeasurable. With Pacific Hemostasis® (Fisher Diagnostics, Middletown, VA, USA) Thrombin on Stago ST4, TCTs were often unmeasurable at or above a plasma dabigatran concentration of 112 ng mL−1, and all dabigatran levels greater than 164 ng mL−1 were unmeasurable. Thus, for both assays, TCTs were often unmeasurable at or above the 50th percentile.

ECA and ECT

Both the ECA and the ECT showed a linear relationship and strong correlation with plasma dabigatran concentration (Fig. 4). The ECAs’ lower limit of detection, as determined by the testing laboratory individual calibration method, was as low as 0 ng mL−1 on BCS® XP System, 40 ng mL−1 on Stago STA Compact, and 36 ng mL−1 on Diagnostica Stago STA-R Evolution® (Diagnostica Stago SAS, Asnieres, France).

Figure 4.

The performance of the ecarin chromogenic assay (ECA) (A) and ecarin clotting time (ECT) assay (B) in the presence of dabigatran. Results for ECA are reported as ng mL−1 from a calibration curve prepared with plasma calibrators. The vertical lines in both figures represent the ‘therapeutic range’ of dabigatran plasma concentrations. The best-fit line is displayed for each of the ECA assays in descending order: ECA-T on BCS® XP System (line 1), ECA-T on Stago STA Compact (line 2), and ECA-T on Diagnostica Stago STA-R Evolution® (line 3). The horizontal lines in (B) denote the range of ECT results for controls.

dTT

The dTT showed a linear relationship and strong correlation with dabigatran concentration (Fig. 5). It is of note that the linearity of dTT is not defined below the lower limit of detection, 30 ng mL−1.

Figure 5.

The impact of dabigatran on the dilute thrombin time (dTT). The vertical lines represent the ‘therapeutic range’ of dabigatran plasma concentrations. The horizontal line denotes the lower limit of detection for this assay.

Discussion

This study determined the performance of various coagulation assays in patients treated with dabigatran 150 mg twice daily, and draws conclusions regarding their potential usefulness in determining low (‘subtherapeutic’), intermediate (‘therapeutic’) and high (‘supratherapeutic’) levels.

Definition of ‘therapeutic range’

In the absence of published data correlating clinical events – hemorrhage and thromboembolism – with plasma drug levels, no true ‘therapeutic range’ for dabigatran exists. For this study, ‘therapeutic range’ was defined as the spread of plasma trough and peak dabigatran levels as determined by mass spectrometry, occurring in 95% of the patients. It is of note that the therapeutic ranges in our study were 18–206 ng mL−1 at trough and 45–487 ng mL−1 at peak, which are similar to previously reported ranges (also 2.5th to 95th percentiles) of 31–225 ng mL−1 at trough and 64–443 ng mL−1 at peak in patients receiving dabigatran 150 mg twice daily [4].

PT

We have shown that the PT, across different reagents and coagulometers, lacks sensitivity in detecting subtherapeutic and therapeutic levels of dabigatran. All instrument–reagent combinations performed similarly. Thus, the PT cannot be recommended for measuring the anticoagulation effects of dabigatran. However, for most PT assays, a normal PT value rules out plasma dabigatran concentrations above the 75th percentile, and may be considered for ruling out high therapeutic and supratherapeutic levels.

APTT

We have shown that the APTT cannot reliably distinguish subtherapeutic from therapeutic levels of dabigatran, regardless of the reagent–instrument combination. For all APTT reagents studied, 18% of patients had therapeutic dabigatran levels despite having a normal APTT. This is an important finding, as it is at variance with and does not support the recommendation in the ‘Practice Guide’ from the American Society of Hematology, which states that, in a patient on dabigatran who is bleeding, ‘a normal aPTT is an indicator that dabigatran would be unlikely to contribute to bleeding’ [16]. Similarly, the dabigatran prescribing information states that ‘at therapeutic doses dabigatran prolongs the aPTT’ and that ‘the aPTT provides an approximation of dabigatran's anticoagulant effect’ [13]. Finally, a recent publication from the British Committee for Standards in Haematology states that ‘above 100 ng mL−1 the aPTT is invariably prolonged’ and that ‘a normal aPTT ratio is likely to exclude a therapeutic intensity of anticoagulation due to dabigatran’ [18]. In our study, 15–35% of patients who had a plasma dabigatran level above 100 ng mL−1 had a normal APTT, depending on which reagent was used. Thus, the findings from our study do not support these guideline statements. However, we showed that normal APTT values rule out dabigatran levels above the 50th percentile, and the APTT may therefore be used to rule out high therapeutic and supratherapeutic levels.

POC device testing

With the exception of a few case reports that have demonstrated elevated POC International Normalized Ratio results in patients on dabigatran, and in vitro drug-spiking in whole blood from healthy volunteers, no data have previously been published on the performance of whole blood POC coagulation analyzers [14, 15, 20]. Our study investigated POC PT, APTT and ACT using the HEMOCHRON® Signature Elite whole blood coagulation device. The APTT assay with the POC device performed similarly to the plasma-based APTT assays. However, the whole blood PT assay performed better than most plasma-based PT assays. Like the plasma-based APTT assays, this specific whole blood PT assay may be considered for ruling out dabigatran levels above the 50th percentile. The large number of normal ACT values in the presence of therapeutic dabigatran indicates that the ACT lacks sensitivity in distinguishing subtherapeutic from therapeutic dabigatran. However, a normal ACT rules out dabigatran levels above the 75th percentile. The ACT cannot be used to distinguish between subtherapeutic and therapeutic levels, but it can be utilized as an alternative to the plasma-based PT assay to exclude excess dabigatran.

TCT

Our study expands previous knowledge by showing that, in patients treated with dabigatran, the TCT, also referred to as the thrombin time, is exquisitely sensitive to dabigatran, and therefore often has unmeasurable responses when plasma dabigatran levels are above the 50th percentile [1, 4, 7, 13]. A normal TCT indicates that no or only minimal dabigatran is present.

ECA, ECT, and dTT

The ECA, ECT and dTT have shown linear relationships with plasma dabigatran levels in in vitro and ex vivo studies [1, 3, 5-8, 10-13]. Our study expands these data by demonstrating that these assays performed on plasma from patients are sensitive to the presence of therapeutic and supratherapeutic dabigatran levels. The sensitivity of the ECA to low levels of dabigatran was dependent on the assays’ lower limits of detection, as determined by the levels of calibration used.

Clinical usefulness of coagulation tests

Figure 6 summarizes the conclusions from our study, depicting how various coagulation assays can potentially be utilized in clinical practice to identify supratherapeutic and subtherapeutic responses to dabigatran therapy. Three clinical scenarios in which measurement of the dabigatran anticoagulant effect may be clinically useful can be identified. (i) The most useful assay for determining subtherapeutic dabigatran is the TCT; to a lesser extent, the ECA, ECT and dTT can also be useful. However, the sensitivity of the ECA and dTT in the detection of dabigatran plasma concentrations of < 30–40 ng mL−1 depends on the assay's lower limit of detection as determined by the calibration. The assays of limited value in these situations are the PT and APTT. (ii) The coagulation tests most useful for identifying therapeutic dabigatran are the ECA, ECT, and dTT, which provide accurate information about dabigatran plasma levels throughout the majority of the therapeutic range. Given the lack of sensitivity, the ACT, APTT and PT are not useful, and the TCT is too sensitive. (iii) The best tests for identifying supratherapeutic dabigatran are the ECA, ECT, and dTT. However, the APTT, ACT and PT may also provide guidance, owing to the strong correlation with dabigatran concentrations. A normal APTT excludes excess dabigatran above the 50th percentile, and a normal ACT or PT rules out dabigatran above the 75th percentile. The TCT is too sensitive to identify supratherapeutic levels.

Figure 6.

Potential clinical usefulness of coagulation assays in relation to ranges of dabigatran plasma concentrations. The shaded parts of the rectangles for the dilute thrombin time (dTT), ecarin clotting time (ECT) and ecarin chromogenic assay (ECA) depict the area of dabigatran plasma concentrations below which accuracy of the assay is not established; < 30 ng mL−1 for the dTT, < 15 s for the ECT, and < 40 ng mL−1 for the ECA. ACT, activated clotting time; APTT, activated partial thromboplastin time; PT, prothrombin time; TCT, thrombin clotting time.

Study strengths and limitations

The strengths of our study are that therapeutic dabigatran levels were determined in patients on dabigatran, not in in vitro drug-spiked plasma samples or human volunteers, and that these individuals were ‘real-world’ patients, not patients enrolled in industry-sponsored efficacy and safety trials. In addition, a variety of different coagulation assays and reagent–analyzer combinations were tested, including the PT and APTT reagents most commonly used in clinical US laboratories [19]. Furthermore, in addition to plasma-based coagulation assays, several whole blood POC device tests were investigated. Finally, the ECA and the ECT assay were evaluated as established in several clinical laboratories, and the dTT assay was also incorporated.

A limitation of our study is the difference in age and gender distribution between dabigatran patients and controls. This may have influenced the normal ranges for the various coagulation tests. However, we would not expect this to influence the correlation of coagulation test results with dabigatran levels as presented in this study, as these normal ranges were similar to the reference ranges established for routine clinical use in the collaborating clinical laboratories. Although patients were selected without regard to weight, subjects in this study (95 kg) were heavier than patients included in the Randomized Evaluation of Long-Term Anticoagulation Therapy (RELY) study (83 kg) [21]. However, although there was a trend towards lower dabigatran levels with increasing weight, this difference was not statistically significant. Given the nature of our definition of ‘therapeutic range,’ a limited number of subtherapeutic and supratherapeutic values were included in our study. Although we were able to make inferences, this prevented definite conclusions to be drawn about the performance of these tests in subtherapeutic and supratherapeutic situations. In addition, the performance of the ECA and dTT in the subtherapeutic range below plasma dabigatran concentrations of 30–40 ng mL−1 requires further laboratory study, with investigatation of the use of standard curves calibrated to low drug concentrations.

In conclusion, this study evaluating the performance of a large variety of coagulation tests in patients on therapeutic doses of dabigatran shows that: the PT, APTT and ACT are often normal in spite of therapeutic dabigatran plasma levels; the TCT is useful for detecting minimal dabigatran levels; and the dTT, ECA and ECT accurately identify therapeutic and supratherapeutic dabigatran levels. This information should enable clinical laboratories to establish appropriate assays and the clinician to utilize the appropriate test when needing to assess a patient's anticoagulation status.

Addendum

E.M. Hawes, A.M. Deal, D. Funk-Adcock, R. Gosselin, A.M. Cook, and S. Moll: designed the study, acquired, analyzed and interpreted the data, and wrote the manuscript; E.M. Hawes: coordinated the work; S. Moll: provided study supervision; A.M. Deal: performed statistical analysis; C. Jeanneret: designed the study and acquired data; J.M. Taylor, H.C. Whinna, and A.M. Winkler: acquired, analyzed and interpreted the data, and performed critical manuscript revision.

Acknowledgements

We thank J. Looney and the staff from UNC McLendon Laboratories at Carolina Pointe II for providing laboratory support, P. Parker and the staff from the Specialty Coagulation Laboratory at UNC Hospitals for coagulation testing, and H. Yang from the UNC School of Public Health for statistical assistance. The HEMOCHRON® Signature Elite devices and cuvettes were provided free of charge by International Technidyne Corporation, Edison, NJ, USA. Plasma dabigatran concentration testing was provided free of charge by Boehringer-Ingelheim. dTT Hemoclot® (Thrombin Inhibitor, Hyphen Biomed, Neurillesur, Oise, France) Thrombin Inhibitor and calibrators were provided free of charge by Aniara, West Chester, OH, USA. The companies did not play a role in study design, data collection, evaluation, presentation, or manuscript preparation. The following individuals did not play a role in study design, data collection, evaluation, presentation, and manuscript preparation. We gratefully acknowledge J. van Ryn from Boehringer-Ingelheim for providing plasma dabigatran level testing, and S. El Rouby from International Technidyne Corporation for orchestrating access to HEMOCHRON® Signature Elite devices.

Disclosure of Conflict of Interests

E. M. Hawes, A. M. Deal, C. Jeanneret, A. M. Cook, J. M. Taylor, H. C. Whinna and A. M. Winkler declare no competing financial interest. R. Gosselin received speakers’ fees, funding for studies, and participated in focus group meetings for Siemens Healthcare Diagnostics. R. Gosselin and D. Funk-Adcock received honoraria for participating in Instrumentation Laboratory's North American Advisory Committee. S. Moll received an honorarium for consulting for Boehringer-Ingelheim.

Ancillary