The importance of dosing frequency on the safety of DOACs

Abstract

Direct oral anticoagulants (DOACs) represent a significant advance over vitamin K antagonists, such as warfarin in patients with atrial fibrillation (AF), due to their favorable risk-benefit profile, with significant reductions in stroke and serious bleeding, particularly intracranial hemorrhage.1 Although as class, DOACs tend to reduce major bleeding in general compared to warfarin, they are associated with a significant excess of gastrointestinal (GI) bleeding. A rationale for this observation is that unlike warfarin, the DOACs are active anticoagulants in the gastrointestinal lumen which contributes to GI bleeding above and beyond the systemic exposure to anticoagulant activity. Ido et al. present an interesting study in this issue investigating whether dosing regimen, specifically once vs. twice daily, is the major contributor to a differential GI bleeding profile across the DOACs.

Direct oral anticoagulants (DOACs) represent a significant advance over vitamin K antagonists, such as warfarin in patients with atrial fibrillation (AF), due to their favorable risk-benefit profile, with significant reductions in stroke and serious bleeding, particularly intracranial hemorrhage [1]. Although as class, DOACs tend to reduce major bleeding in general compared to warfarin, they are associated with a significant excess of gastrointestinal (GI) bleeding. A rationale for this observation is that unlike warfarin, the DOACs are active anticoagulants in the gastrointestinal lumen which contributes to GI bleeding above and beyond the systemic exposure to anticoagulant activity. Ido et al. present an interesting study in this issue investigating whether dosing regimen, specifically once vs. twice daily, is the major contributor to a differential GI bleeding profile across the DOACs.

The authors report their finding from the DIRECT registry, which was a single-center prospective observational registry of 2216 AF patients followed for 650 days [2]. Patients were divided into twice daily (dabigatran and apixaban) and once daily (rivaroxaban and edoxaban) DOAC groups [2]. The analysis consisted of the 904 patients in each group that were successfully matched using propensity score adjustment. The primary endpoint was GI bleeding identified through the medical records. The twice daily group demonstrated a significantly lower rate of GI bleeding than the once daily group (3.5/100 person-year vs. 6.2/100 person-year, HR 0.556, 95 % CI 0.416–0.743, log-rank P < 0.0001). They also observed a significantly lower rate of major bleeding in general with the twice daily group compared to once daily group (log-rank P = 0.040). There was no difference between the regimens with respect to stroke or mortality.

To put these results in context, there are several important considerations. The first is to address the pharmacokinetic and pharmacodynamic rationale of once vs. twice daily dosing. The authors comment that their findings of the superior safety with twice daily regimens are supported by the premise that the peak level of drug concentration is thought to be associated with bleeding events. Twice daily dosing results in lower peak concentration but a higher trough concentration while once daily dosing results in a higher peak concentration but a lower trough concentration. The best data we have on this topic, however, suggests the exact opposite. The ENGAGE AF-TIMI 48 study, the largest and longest of the seminal phase 3 DOAC vs. warfarin trials, found that peak drug concentration and the endogenous factor Xa activity (the true pharmacodynamic effect of the drug, i.e., the anticoagulant action) was found not be a significant predictor of the risk of bleeding but both average and minimum concentration/inhibition were [3]. These results argue that drug concentration and anticoagulant activity need to be below a certain threshold for optimal hemostasis and vessel repair to occur. For many different drugs it is the minimum concentration (trough) that is correlated with toxicity (e.g., aminoglycosides).

The second important consideration is how do the results of the DIRECT registry compare to the extensive DOAC clinical trial data we have in over 70,000 patients, the gold standard when assessing the efficacy and safety of drugs. The authors correctly note that the once daily regimens of rivaroxaban and edoxaban were associated with a higher risk of GI bleeding but dismiss the fact that dabigatran, given twice daily, was also associated with a significant excess of GI bleeding by stating that the tartaric acid core that coats dabigatran causes direct caustic GI injury which obviates the superiority of the twice daily regimen. The omit the important fact that the only DOAC regimen studied associated with a significantly lower risk of GI bleeding was the lower dose twice daily edoxaban regimen (30 mg with dose reduction to 15 mg in selected patients) [4]. Apixaban and dabigatran 110 mg had only similar GI bleeding compared with warfarin. These results suggest that dose and ultimately the level of anticoagulation is potentially more important than dosing frequency in determining the safety (and efficacy) of DOACs. This is further supported by the findings from the clinical trials that two safest DOACs with respect to major bleeding overall (not just GI bleeding) were apixaban (twice daily) and edoxaban (once daily).

The final consideration as to why these results of the DIRECT registry are discordant with prior studies has to do with the nature of the study. An important limitation, as acknowledged by the authors, is that the registry is a relatively small, single-center observational study. The assignment of once vs. twice daily DOAC was not randomized. To deal with factors that influence why a physician may choose one drug over another, the authors appropriately use propensity score-matching technique, but it is important to note that observational studies can't account for unmeasured confounders and the propensity score model in their study had an area under the receiver operating characteristic (ROC) curve (AUC) of 0.061, which is modest at best, suggesting likely residual confounding.

The authors should be congratulated for providing important new data regarding the relative safety of DOACs with respect to GI bleeding, which is the most common type of anticoagulant associated bleeding in AF patients. While dosing frequency likely plays a role, the safety profile of DOACs is more complex and requires integration of dose, dose frequency, ultimate anticoagulant effect, and patient characteristics.

Declaration of competing interest

Honoria from Anthos, Bayer, Bristol Myers Squibb, Boehringer Ingelheim, Daiichi Sankyo, Janssen, Pfizer. Member of the TIMI study Group which has received institutional research grant support through Brigham and Women's Hospital from Abbott, Amgen, Anthos Therapeutics, ARCA Biopharma, Inc., AstraZeneca, Bayer HealthCare Pharmaceuticals, Inc., Daiichi-Sankyo, Eisai, Intarcia, Ionis Pharmaceuticals, Inc., Janssen Research and Development, LLC, MedImmune, Merck, Novartis, Pfizer, Quark Pharmaceuticals, Regeneron Pharmaceuticals, Inc., Roche, Siemens Healthcare Diagnostics, Inc., Softcell Medical Limited, The Medicines Company, Zora Biosciences.