Author + information
- Paul K.L. Chin, MB ChB∗ (, )
- Daniel F.B. Wright, PhD,
- Matthew P. Doogue, MB ChB and
- Evan J. Begg, MD
- ↵∗University of Otago, 2 Riccarton Avenue, Christchurch, Canterbury 8011, New Zealand
We read with great interest the paper by Reilly et al. (1), which demonstrated that higher trough plasma dabigatran concentrations were associated with: 1) decreasing risk of stroke/systemic embolic event (SEE); and 2) increasing major bleeding risk in the RE-LY (Randomized Evaluation of Long-Term Anticoagulation Therapy) trial (2). The paper provides useful insights into the relationship between dabigatran exposure and clinical events but raises several questions.
1. Time lag from blood samples to events.
The dabigatran concentrations reported by Reilly et al. (1) were from blood samples that were mostly collected 1 month after enrollment. In contrast, the majority of the strokes/SEE in the RE-LY study occurred >12 months after enrollment (2). This difference limits the inferences that can be drawn from the observed relationship between dabigatran concentrations and adverse events, because of intraindividual changes during the time lag (e.g., altered renal function ) or addition/withdrawal of interacting medicines (4–7).
Could the temporal relationship between events and blood samples be described more specifically? Furthermore, was the correlation between trough concentrations and events stronger for those with a shorter time lag?
2. Variability in trough concentrations.
In the RE-LY study, a subgroup of those on dabigatran etexilate contributed additional blood samples at 3, 6, and 12 months after enrollment (1). These data could provide insights into intraindividual variability in trough plasma dabigatran concentrations over time.
3. Therapeutic index.
According to Reilly et al. (1), their data demonstrated that dabigatran has a wide therapeutic index. This is a strong claim, which we question. Data in their Table 1 and Figure 2 show that for a 72-year-old male with atrial fibrillation, diabetes mellitus, and previous stroke, there was a 3-fold increase in risk of major bleeding events between the 10th and 90th percentiles of trough concentrations. Furthermore, they found that the ischemic stroke/SEE risk declined steeply at low concentrations, with little additional benefit from higher concentrations. These data actually allow a more formal development of a possible target trough concentration range.
An example of a range for this individual may be derived by adjusting the lines of best fit of the event risk–concentration curves by using event weighting according to the hazard ratios of death described by Eikelboom et al. (8) and then combining the lines of best fit (Fig. 1). Note that for the patient described here, the therapeutic range is relatively narrow compared with the 10th to 90th percentiles of trough concentrations and would certainly not be considered a wide therapeutic range.
4. Trough concentrations outside the 10th and 90th percentiles.
By definition, 20% of individuals treated with dabigatran etexilate in the RE-LY study had concentrations outside the 10th to 90th percentiles of trough plasma dabigatran concentrations. Did the outcomes for these individuals differ from the rest of the cohort? Dose adjustment to “improve” the drug exposure for these 20% of patients may be beneficial. Could the authors describe the characteristics of the individuals with trough plasma dabigatran concentrations that were below the 10th percentile and above the 90th percentile?
- American College of Cardiology Foundation
- Reilly P.A.,
- Lehr T.,
- Haertter S.,
- et al.,
- RE-LY Investigators
- Eikelboom J.W.,
- Connolly S.J.,
- Hart R.G.,
- et al.