Three large trials (ie, the Randomized Evaluation of Long-Term Anticoagulation Therapy [RE-LY], the Rivaroxaban Once Daily Oral Direct Factor Xa Inhibition Compared With Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation [ROCKET-AF], and the Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation [ARISTOTLE] trials)1–3 were recently published to test the efficacy and safety of new oral anticoagulants in preventing stroke and systemic embolism in patients with nonvalvular atrial fibrillation. The three studies, despite their important differences in design and clinical profile of the participants, consistently demonstrated at least noninferiority for both efficacy and safety for the newer anticoagulants (ie, dabigatran, rivaroxaban, apixaban) vs warfarin. Beyond the already emerged debate relative to the clinical application of the newer agents,4,5 there are further obscured points seeking clear answers in order to better evaluate whether there is a clinically relevant link between implemented randomized treatments and investigated outcome.
First, we should underline that the comparator drug was used at least suboptimally in all of the three studies, with ROCKET-AF scoring the less acceptable rate of anticoagulant therapeutic efficacy for warfarin as compared with both RE-LY and ARISTOTLE. In these lines, the new drugs were substantially compared with real-world warfarin misusage, and not to its optimal therapeutic potential, which is a mandatory requirement in the setting of randomized clinical trials. A step further, it would be quite interesting to know the pre-events international normalized ratio (INR) of patients in the arm of warfarin treatment in order to confirm or not the time-dependent causative associations between the randomized treatment and outcome. In our view, the time in therapeutic range used in prespecified studies’ analyses represents a by far vague approach to assess warfarin’s efficacy. By contrast, INR values in every single patient just before the occurrence of any outcome would reflect the individualized clinical relevant efficacy of warfarin.
Second, in all studies it is reported that participants had their blood pressure (BP) controlled at baseline. However, no information is provided for both BP control rate and BP levels during the follow-up period, since stroke and bleedings are strongly correlated with BP-oriented measures. It might be that participants randomized in the two treatment arms had unequal control of their BP at least in the period around the occurrence of the outcome. The same discussion is also applicable for participants with diabetes mellitus. Did all of the participants have their plasma glucose well controlled or eventual hypoglycemic phenomena interfered with the outcome under investigation? Taken together, the above inadequate control of clinically relevant outcome confounders might have seriously biased the results.
Third, it is already proven that the new drugs demonstrated interactions with other drugs at the level of intestinal epithelium (ie, glycoprotein 1).6 Accordingly, it would be interesting to know the coadministration of drugs affecting the bioavailability of the new anticoagulants and possibly project correlations of their implementation with adverse outcomes.
Finally, all three studies recruited patients older than 80 years; however, in clinical practice, a huge number of patients with nonvalvular atrial fibrillation reside in the age range of the very elderly (>80 years). Extending the findings of these trials in very elderly patients with an enhanced burden of cardiovascular risk and comorbidities at least represents a foggy clinical suggestion.
We steadily believe that projecting studies that evaluate hard clinical end points with suboptimal evaluation and management of the underlying traditional cardiovascular risk factors seriously questions the merit of the results. A second look and analysis of the results is mandatory in order to assess the impact of parameters not taken into consideration in the original release of the “new anticoagulation era” studies.
Beyond all of these remarks, the new drugs for nonvalvular atrial fibrillation promise to overcome the limitations related to warfarin-related anticoagulation strategy. At present, whether these drugs would demonstrate adequate efficacy and safety in real-world clinical practice remains yet undefined.