INCREASED STROKE RISK DURING FIRST 30 DAYS OF WARFARIN TREATMENT

Ladies and gentlemen, I am Christoph Diener, a neurologist from the University of Essen in Germany. Today's topic is a possible prothrombotic state with the initiation of warfarin in patients with atrial fibrillation.

When the large clinical trials were conducted with novel anticoagulants, there was an interesting observation at the end of the studies. In the ROCKET AF study^[1] with rivaroxaban and the ARISTOTLE study^[2] with apixaban, patients were switched at the end of the study from the novel anticoagulant to warfarin. A very high risk for ischemic stroke was observed in patients who were treated de novo with warfarin, which was not observed in other conditions, so it is not a rebound phenomenon associated with rivaroxaban or apixaban.

To determine whether warfarin causes a prothrombotic state, data were derived from the United Kingdom general practice physicians' database, which has data from 60 million patients in the United Kingdom, including drugs taken and outcome events.^[3] In this study, 70,766 patients with permanent atrial fibrillation were identified and followed from 1993 to 2012. During this period, 5519 ischemic strokes occurred -- an annual stroke rate of about 2%.

The investigators performed a nested case-control study, and for each case with stroke, they identified 10 people who did not have a stroke. In the first 30 days after the initiation of warfarin, there was a 71% increased risk for ischemic stroke, followed by the accepted reduction of stroke risk of 50%-60% beyond 30 days. This is a very strong argument for a prothrombotic state with the initiation of warfarin.

We looked at this in the randomized trials, analyzing the first 30 days of initiation of anticoagulation in trials where novel anticoagulants were compared with warfarin.^[4] We observed that warfarin-naive patients who were treated de novo with warfarin had, on average, twice the risk for ischemic stroke than patients who continued on warfarin. This increased risk for ischemic stroke was not observed with the novel anticoagulants.

This is a very strong argument for using novel anticoagulants in patients who had a transient ischemic attack or stroke and have atrial fibrillation. One issue for which we have no data whatsoever is whether bridging (which is done in patients with deep vein thrombosis) would be of benefit when we initiate anticoagulation with warfarin. These trials, unfortunately, have not been done.

Clearly, the initiation of warfarin in people with atrial fibrillation might lead to an increased risk for ischemic stroke in the first 10 days, and most likely this is followed by long-term benefit.

I am Christoph Diener, a stroke neurologist from the University of Essen in Germany. Thank you for listening.