This post draws on some of the information provided in Dr. Tracy Minichiello's talk at the 17th annual UCSF conference on hospital medicine with particular focus on some less well appreciated findings form the clinical trials that led to approval of the three target specific oral anticoagulants (TSOACs) currently in use.
Compared to warfarin in RE-LY. Below in reference to the 150 mg bid dose.
Dabigatran favored, established superiority , RR 0.66 (0.53-0.82), for the primary outcome stroke or systemic embolism. NNT 172.
Dabigatran associated with borderline statistically significant reduction in all cause mortality, RR 0.88 (0.77 to 1.00) P=0.051.
Dabigatran associated with less ICH, RR 0.4 (0.27-0.6). NNT 227.
Dabigatran associated with increased GI bleed, RR 1.5 (1.19-1.89). NNH=204.
Dabigatran associated with trend toward higher risk of MI, not statistically significant, RR 1.27 (0.94-1.71). Absolute risk very low. NNH for the point estimates would be 588.
For the other safety outcomes the two drugs were comparable.
Pradaxa product labeling.
Comment: the trend toward a difference in MI risk suggests caution in the selection of anticoagulants in patients with coronary disease particularly taking into account that warfarin has an FDA approved indication for secondary prevention post MI. Consider avoidance in patient with history of GI bleed.
Compared to warfarin in ROCKET AF.
For the primary outcome of stroke/systemic embolism rivaroxaban, in the per protocol, as treated population, was non-inferior with HR 0.79 (0.66-0.96).
Overall bleeding rates comparable.
Critical (see the body of the paper for definition) and fatal bleeding rates were less with rivaroxaban, and the differences were statistically significant, HR 0.69 (0.53-0.91) and 0.50 (0.31-0.79) respectively.
ICH rates were statistically significantly less with rivaroxaban, HR 0.67 (0.47-0.93).
Rivaroxaban was associated with a statistically significant increase in GI bleeds, NNH 101.
There was no difference in MI.
At the end of the study treatment period there was a spike in the primary outcome in rivaroxaban patients transitioning to warfarin related to a long (13 day) time to reach target INR.
Compared to warfarin in ARISTOTLE.
Note the primary outcome: ischemic or hemorrhagic stroke or systemic embolism.
Apixaban was favored, HR 0.79 (0.66-0.95) p=.01 for superiority.
For ischemic or unknown mechanism stroke HR 0.97 (0.74-1.13).
For ICH HR 0.51 (0.35-0.75).
All cause mortality HR 0.89 (0.80-0.99).
Non-statistically significant trend toward reduced MI.
No significant difference in GI bleeds.
The labeling for all three agents carries warnings to consider bridging with discontinuation or transition to warfarin.
When selecting an anticoagulant for stroke prevention in a fib for patients with CAD consider 1) warfarin has an approved indication for secondary prevention in CAD patients; 2) a possible increased risk of MI with dabigatran; and 3) that we have a long experience using various combinations of warfarin and antiplatelet agents.
Consider avoidance of dabigatran and rivaroxaban in patients with GI bleed history or risk.
The true practice of EBM honors the preferences and values of an informed patient. Carve out time for a long discussion with your patient when you start them on an oral anticoagulant for stroke prevention in atrial fibrillation.