Dabigatran Monitoring: Observations From the Initial 127 Patients Tested

Dabigatran is the first oral anticoagulant since warfarin to be approved by the FDA for prevention of stroke in patients with non-valvular atrial fibrillation. Unlike warfarin, dabigatran monitoring has not yet been recommended. However, dabigatran accounted for 3781 serious adverse events (including 542 deaths) as reported in QuarterWatch by the Institute for Safe Medicine Practices in 2011. This has prompted clinicians to consider situations in which drug level measurements may provide clinically useful information to guide dosage. Aside from managing the acutely bleeding patient, drug level monitoring may be useful in the outpatient setting. For example, establishing a baseline drug level may be relevant given the direct correlation with bleeding risk, especially in patients with decreased creatinine clearance. It can also assess patient compliance given the short half-life of the drug. We performed a retrospective analysis of patient samples that were tested for dabigatranlevels at our facility to determine whether drug levels were within the therapeutic range.

Dabigatran levels were assayed for 172 frozen, citrated plasma samples from 127 patients (25 patients had multiple samples). The assay is a laboratory-developed, dilute thrombin clotting time test (CV ≤7.3%) that was validated by LC-MS/MS and is performed using drug-specific calibrators (Clin Chem 2012 abstract, in press). Specimens should be collected 90 to 120 minutes after the last dose and after ≥3 doses. The therapeutic range (64–443 ng/mL) was derived from the PETRO trial (van Ryn et al Throm Haemost2010). Patient demographics were correlated with drug levels, but clinical history was not available.

Dabigatran levels were within the therapeutic range for 76 samples (44.2%), in the subtherapeutic range for 82 samples (47.7%), and in the supratherapeutic range for 14 (8.1%) samples. The median dabigatran level for the samples within the therapeutic range was 152 ng/mL. These samples also had a skewed distribution such that 44 (57.9%) were within the lowest quartile and 62 (81.6%) were within the lower half.

Table 1 shows the distribution of dabigatranlevels by age and gender. Women ≥65 years of age were the most likely to have levels in the therapeutic range. Independent of gender, people aged <65 years were less likely to have levels in the therapeutic range. Four patients were not classified by age or gender.

Of the 25 patients with multiple samples (2–6 samples/patient), 17 patients were consistently in or out of range: 8 therapeutic, 7 subtherapeutic, and 2 supratherapeutic. The remaining 8 patients were variable (Table 2).

(1) A surprisingly high percentage of patients taking dabigatran were subtherapeutic. (2) The cause of the subtherapeutic levels is not readily discernable given the lack of clinical history, but may be explained by poor compliance, improper collection time, drug interactions, or decreased absorption (anecdotal evidence).

No relevant conflicts of interest to declare.