509074

research-article2013

AOPXXX10.1177/1060028013509074Annals of PharmacotherapyGosselin et al

Research Report

Measuring Dabigatran Concentrations Using a Chromogenic Ecarin Clotting Time Assay

Annals of Pharmacotherapy 47(12) 1635­–1640 © The Author(s) 2013 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1060028013509074 aop.sagepub.com

Robert C. Gosselin, CLS1, Denis M. Dwyre, MD1, and William E. Dager, PharmD, BCPS, FCSHP, FCCP, FCCM, FASHP1,2,3,4

Abstract Background: Clinicians managing patients receiving the direct thrombin inhibitor dabigatran may benefit in being able to determine the amount of drug present in selected situations. This may include assessment of accumulation, concurrent drug interactions, or adequate removal from circulation. The ability to estimate the amount of dabigatran present using the chromogenic ecarin assay (ECA) requires further clarification. Objective: To describe the reliability of dabigatran measurements using a chromogenic ECA. Methods: This was an evaluation of the ECA method that incorporated assessment of imprecision, linearity, accuracy, carryover, and lower limits of detection or blank. Pooled normal plasma enriched with dabigatran at concentrations of 0, 25, 50, 75, 100, 125, 150, 200, 300, 400, and 500 ng/mL were sent blinded to 3 laboratories in the United States to compare our ECA results with those of laboratories reporting dilute thrombin time methods (HEMOCLOT thrombin inhibitor assay) for measuring dabigatran. Trough and peak levels from 35 patients were also compared with mass spectrophotometry for assessing ECA accuracy. Results: The within-run or day-to-day imprecision was less than 10%, with high linearity (R2 = 0.989) and high degree of accuracy (R2 = 0.985; slope = 0.908) for levels ranging between 18 and 470 ng/mL and no carryover at 0 ng/mL noted. The ECA approach appeared to be more reliable at lower dabigatran concentrations. Conclusions: The chromogenic ECA appears to be an effective approach to determine the amount of dabigatran present. Further insights are necessary to determine how it can be used to reduce thromboembolic or bleeding complications in patients receiving dabigatran Keywords dabigatran, chromogenic ecarin assay, laboratory

Introduction In late 2010, the FDA approved dabigatran etexilate (Pradaxa, Boehringer Ingelheim) for prevention of stroke in the setting of atrial fibrillation.1,2 Dabigatran is an oral direct thrombin inhibitor that does not require routine monitoring, but if a level is desired, the FDA recommends the “ecarin test.”3 At our institution, we have encountered several patients (eg, acute trauma and renal insufficiency) receiving dabigatran where there would have been a desire to quantitate the amount of dabigatran present. Complicating this issue is the lack of FDA-approved kits for ecarin testing in the United States. In addition, dabigatran etexilate is a pro-drug, so the active form of the drug is required for in vitro assessment of its ability to quantitate the amount of dabigatran present. Current recommendations from the subcommittee on control of anticoagulation of the Scientific and Standardisation Committee of the International Society on Thrombosis and Haemostasis indicate that the dilute thrombin time (dTT) in combination with dabigatran calibrant

plasmas can be used to determine the drug concentration4 but also note that ecarin clotting times are linear. However, there is no mention of chromogenic ecarin methods to determine dabigatran levels. We opted to evaluate a chromogenic ecarin kit (HaemoSys-ECA Ecarin Chromogenic Assay, Stago International, Parsippany, NJ) on the BCSXP analyzer (Siemens Healthcare Diagnostics, Glasgow, DE). Chromogenic ecarin testing has been previously described for use in patients on hirudin and argatroban anticoagulation5 and in one study in patients receiving dabigatran after major orthopedic surgery.6 At our site, the ecarin assay 1

University of California-Davis Health System, Sacramento, CA, USA University of California-San Francisco School of Pharmacy, CA, USA 3 University of California-Davis School of Medicine, CA, USA 4 Touro School of Pharmacy, Vallejo, CA, USA 2

Corresponding Author: William E. Dager, PharmD, BCPS, FCSHP, FCCP, FCCM, FASHP, Department of Pharmacy University of California Davis Medical Center, 2315 Stockton Blvd, Sacramento, CA 95817-2201, USA. Email: [email protected]

Downloaded from aop.sagepub.com at University of British Columbia Library on June 21, 2015

1636

Annals of Pharmacotherapy 47(12)

(ECA) method was adapted to the BCS instrument using a laboratory-developed protocol. Additional modifications to instrument programming were additionally explored to improve sensitivity of the test at lower concentrations (0.95 and slope between 0.85 and 1.15. For accuracy, we used Student’s paired t test to determine differences between methods, with a P > .05 indicating no significant differences between methods.

Results Imprecision For within-run imprecision, 4 different plasma samples covering the absence of drug and subtherapeutic, therapeutic, and supratherapeutic range with 0, ~25, ~150, and ~500 ng/mL of drug were tested 5 times concurrently, with the mean (CV) being 0 (0%), 23 (2.0%), 166 (9.6%), and 449 (1.0%) ng/mL, respectively. For day-to-day imprecision, 2 commercial dabigatran controls were evaluated over 12 days, with mean (CV) of 121 (7.4%) and 326 (8.2%) ng/ mL, respectively.

Linearity A MS-measured dabigatran sample (510 ng/mL) was diluted with normal pooled plasma and ECA testing performed. The regression statistics indicate favorable linearity, with an R2 = 0.998 (Figure 1A). At lower concentrations, higher-thanexpected recovery of drug was noted, with improvement of drug recovery using a test modification to increase sensitivity at lower drug concentrations (Figure 1B).

Accuracy We used controls, calibrators, and spiked and patient samples for assessing accuracy (n = 89). There was good correlation between BCSXP ECA and MS dabigatran measurements

Downloaded from aop.sagepub.com at University of British Columbia Library on June 21, 2015

1637

Gosselin et al A

A

400 300

y = 1.0507x - 5.983 R2 = 0.9988

200 100 0

0

100

200

300

400

500

500 450 400 350 300 250 200 150 100 50 0

600

Expected dabigatran concentration, ng/ml

B

Recovered dabigatran concentation, ng/ml

BCSXP ECA Dabigatran concentration, ng/ml

500

140 130 130 119 120 106 110 100 89 90 73 80 70 59 60 60 44 50 42 40 31 24 25 30 19 20 15 17 1011 10 0 0 130 117 104 91 78 65 52 39 26 13 6.5 0 Expected dabigatran concentration, ng/nl

0

50

100 150 200 250 300 350 400 450 500 MS Dabigatran concentration, ng/ml

B 20.00 BCS XP - Mean Dabigatran concentration, ng/ml

Recovered dabigratran concentration, ng/ml

600

10.00 0.00

0

50 100 150 200 250 300 350 400 450 500

-10.00 -20.00 -30.00 -40.00

Mean Dabigatran concentration, ng/ml

Figure 1.  Assessment of linearity for dabigatran assay using the chromogenic ecarin method. There is good linearity (A) between ECA and dabigatran measured using mass spectrophotometry, with improved recovery at lower concentrations of the drug when using a modified application (B).

Figure 2.  A. Regression analysis comparing the dabigatran level obtained using chromogenic ecarin assay (ECA) with that from mass spectrophotometry (MS). B. The Bland-Altman bias plots, depicting the areas of bias between the 2 laboratory methods (ECA and MS), indicate that most ECA measurements are less than levels obtained by MS.

(R2 = 0.985, slope = 0.908) for levels ranging between 18 and 470 ng/mL (Figure 2A). There were significant differences between BCSXP ECA– and MS-derived dabigatran levels (P < .001), with the majority of samples demonstrating a lower dabigatran level with BCSXP ECA methods (Figure 2B). When comparing with the dTT method for reporting dabigatran, there was good correlation between ECA and dTT methods (R2 > 0.95), with no significant differences (P > .05) between 2 of the reporting laboratories and ECA results in the dabigatran-enriched samples (Table 1). Of note, the dTT calibration used by reporting laboratories did not discriminate between 0 ng/mL and the lowest calibration point, which was 40 ng/mL, and the results from those sites were reported as 0.95) with the methods used at all 3 laboratories, and no statistical differences were noted between ECA and laboratories A and C.

Discussion Dabigatran anticoagulation is not routinely monitored. However, in selected situations such as overdose, traumatic bleeding, renal insufficiency, concurrent administration of a potential interacting drug (eg, one altering P glycoprotein), need for emergent surgery, or acute stroke, there may be a desire to know the amount of drug present.9 Previous studies for assessing dabigatran have described routine coagulation assays,10,11 such as prothrombin times or thrombin times, each with their limitations to drug sensitivity. Other methods for assessing direct-thrombin inhibitors, such as ecarin clotting time or dTT,12,13 may be useful in measuring dabigatran, but neither of these tests is approved for use in the United States. The additional limitation to assessing and evaluating a method for measuring dabigatran is the available drug; dabigatran etexilate is a pro-drug requiring in vivo metabolism to its active form, dabigatran. There are limited sources of dabigatran available to laboratory personnel for use in assessing and validating methods for measuring this drug. One company, Aniara, does provide calibrators and controls specifically for dabigatran measurement but optimized for using dTT principles. However, because the assigned values for these materials are determined by MS, these products can be used for other methods. In our evaluation, we found that the ECA correlated to drug level (as measured by MS) better than the dTT (HEMOCLOT).8 Other studies have indicated that dTT

correlated with drug concentrations and was superior to activated partial thromboplastin times and thrombin times for measuring drug effect.10,14,15 The HEMOCLOT method for measuring direct thrombin inhibitors must be adapted to automated instrumentation because the package insert for this method describes generic manual preparations of reagent/sample volume and incubation/ clotting times. Therefore, it is vital that automated instruments are appropriately validated when using kits describing generic laboratory preparations. It is possible that Laboratory B has not optimized the HEMOCLOT method for measuring dabigatran as suggested by the significant differences noted between their data and other laboratories performing the dTT method. We opted to evaluate chromogenic methods for measuring dabigatran because, typically, these methods are more robust methods than clot-based testing. The challenge was being able to adopt the test method onto an instrument platform not described by the reagent manufacturer. This process includes customizing and adapting the reagent to the instrument, calibration determination, detailing of the sample and reagent quantities, incubation times, detection times and methods, and finally result reporting. These processes may be relatively easy to determine in simple analyzers but may be more challenging with more complex systems that have multiple reagent probes, incubation wells, and reading or detection formats. In our initial assessment, we determined that our programming process was suitable when (1) we were able to create a linear calibration curve (using MS-determined calibrators) that had a relatively large span in absorbance readings between the highest and lowest calibrator, (2) the imprecision of within-run testing was satisfactory to our criteria (