Medication adherence and non-vitamin K antagonist oral anticoagulants: what do we really know?

In their publication in EP-Europace, Vrijens and Heidbuchel1 conclude twice-daily (BID) dosing of non-vitamin K antagonist (VKA) oral anticoagulants ‘may be more forgiving in patients with suboptimal adherence’ than once-daily (OD) dosing. The authors submit the above conclusion is ‘exemplified’ by the PLATO trial, which compared OD clopidogrel to BID ticagrelor in acute coronary syndrome patients; suggesting the reduced rate of cardiovascular events with BID ticagrelor was the result of a ‘greater degree of continuity of drug action’ in the presence of suboptimal adherence.1,2 This theory ignores potential differences in antiplatelet potency between agents, in addition to, likely inferior platelet inhibition in some clopidogrel patients with CYP2C19 genetic polymorphisms.3 These points are more likely to explain differences in PLATO event rates than variances in blood concentrations/activity subsequent to suboptimal adherence. We agree with the authors that timing adherence (considering both the correct number of doses taken and their timing) is an important adherence metric to consider. If one embraces timing adherence, the authors statement that the pharmacological equivalent of missing a single dose of a QD regimen is missing three consecutive doses of a BID regimen (Figure 2, panel C) becomes representative of only an extreme instance of suboptimal adherence.1 The European Heart Rhythm Association guidance4 on new oral anticoagulants advises a forgotten dose can be taken as long as no more than half the dosing interval has passed. This means a patient taking an OD regimen need only remember to take their missed dose within 12-h of when it was scheduled to catch up; and consequently, need not go a full 48-h without a dose as depicted in panel C. It is also noteworthy that Figure 2 is stated to represent a single hypothetical drug given OD vs. BID, and not a comparison of different non-VKA oral anticoagulants. Perhaps most importantly, it is unclear what effect fluctuations in non-VKA oral anticoagulation (measured by drug concentration in the blood or degrees of factor Xa and/or thrombin inhibition) will have on efficacy and safety. Unlike antibiotics, for example, there is a paucity of data regarding what pharmacokinetic/pharmacodynamic parameters are most important in preventing thrombosis or bleeding. Perhaps the peak-to-trough ratio of the anticoagulant is most important (a kin to aminoglycoside antibiotics), or maybe only a minimal trough level of anticoagulation activity above a certain threshold is required to prevent or treat thrombosis? In our opinion, the only way to assure patients prescribed a novel target anticoagulant in the real-world will obtain similar efficacy and safety as reported in its designated randomized trial is for their medication adherence to rival that observed in the randomized trial. Adherence studies of cardiovascular medications using electronic monitoring devices to evaluate timing adherence have shown OD regimens to be adhered to more frequently than BID regimens (timing adherence: 76.3 vs. 50.4%, adjusted difference: −22.9% (95% confidence interval: −33.1 to −12.7) suggesting OD regimens may yield a better chance of obtaining randomized trial level anticoagulation adherence.5