Early Apixaban use Following Stroke in Patients With Atrial Fibrillation

Results of the AREST Trial

Arthur J. Labovitz, MD; David Z. Rose, MD; Michael G. Fradley, MD; John N. Meriwether, MD; Swetha Renati, MD; Ryan Martin, MD; Thomas Kasprowicz, MD; Ryan Murtagh, MD; Kevin Kip, PhD; Theresa M. Beckie, PhD, MN, RN; Marcus Stoddard, MD; Andrea C. Bozeman, APRN; Tara McTigue, RN; Bonnie Kirby, RN, MSN; Nhi Tran, MS; W. Scott Burgin, MD


Stroke. 2021;52(4):1164-1171. 

In This Article

Abstract and Introduction


Background and Purpose: It is unknown when to start anticoagulation after acute ischemic stroke (AIS) from atrial fibrillation (AF). Early anticoagulation may prevent recurrent infarctions but may provoke hemorrhagic transformation as AF strokes are typically larger and hemorrhagic transformation-prone. Later anticoagulation may prevent hemorrhagic transformation but increases risk of secondary stroke in this time frame. Our aim was to compare early anticoagulation with apixaban in AF patients with stroke or transient ischemic attack (TIA) versus warfarin administration at later intervals.

Methods: AREST (Apixaban for Early Prevention of Recurrent Embolic Stroke and Hemorrhagic Transformation) was an open-label, randomized controlled trial comparing the safety of early use of apixaban at day 0 to 3 for TIA, day 3 to 5 for small-sized AIS (<1.5 cm), and day 7 to 9 for medium-sized AIS (≥1.5 cm, excluding full cortical territory), to warfarin, in a 1:1 ratio at 1 week post-TIA, or 2 weeks post-AIS.

Results: Although AREST ended prematurely after a national guideline focused update recommended direct oral anticoagulants over warfarin for AF, it revealed that apixaban had statistically similar yet generally numerically lower rates of recurrent strokes/TIA (14.6% versus 19.2%, P=0.78), death (4.9% versus 8.5%, P=0.68), fatal strokes (2.4% versus 8.5%, P=0.37), symptomatic hemorrhages (0% versus 2.1%), and the primary composite outcome of fatal stroke, recurrent ischemic stroke, or TIA (17.1% versus 25.5%, P=0.44). One symptomatic intracerebral hemorrhage occurred on warfarin, none on apixaban. Five asymptomatic hemorrhagic transformation occurred in each arm.

Conclusions: Early initiation of anticoagulation after TIA, small-, or medium-sized AIS from AF does not appear to compromise patient safety. Potential efficacy of early initiation of anticoagulation remains to be determined from larger pivotal trials.

Registration: URL: https://www.clinicaltrials.gov/; Unique identifier: NCT02283294.


After an acute ischemic stroke (AIS) or transient ischemic attack (TIA) due to atrial fibrillation (AF), the most significant complications are hemorrhagic transformation (HT), recurrent stroke, and death.[1,2] Minimal data exist to direct optimal timing for initiation of anticoagulation: heparin and the vitamin K antagonist (VKA) warfarin reduced recurrent ischemic stroke rates, but this benefit was offset by a higher incidence of intracerebral hemorrhage (ICH) and HT.[3–5] Four major randomized controlled trials between 2009 and 2013 evaluating direct oral anticoagulants (DOACs) in AF reported noninferiority (or in some cases superiority) to VKA due to decreased rates of stroke and embolic events, with fewer ICH.[6–9] However, these trials excluded strokes within 7 to 30 days, presumably due to an assumption that HT risk on VKA or DOAC is higher than ischemic stroke reduction.[6–9] This threshold delay to start anticoagulation poststroke represents a patient-specific timepoint, when the value of anticoagulation switches from neutral (or even harmful) to beneficial.[8] About 10% of AF patients with stroke have recurrence within 1 month (1.3% daily risk in the first 2 weeks).[7–11]

Early DOAC use may result in a favorable outcome for secondary prevention,[12–14] if started between 4 and 14 days poststroke, potentially even within 2 to 3 days.[9] Currently, the decision on anticoagulation timing is patient-centered, without randomized trials,[3] and delay is driven by perception of elevated HT risk influenced by historical data, provider recall bias, knowledge gaps,[15,16] and lenient guidelines on the subject (class IIa recommendations for 3–14 days, level of evidence B [United States] and C [Europe]).[17,18] Thus, we conducted the investigator-initiated, randomized, open-label AREST study (Apixaban for Early Prevention of Recurrent Embolic Stroke and Hemorrhagic Transformation), to examine early DOAC use at prespecified timepoints based on stroke size, versus conventional VKA at 14 days, or 7 days for TIA.[19]