STROKE RISK INCREASES THE FIRST MONTH OF WARFARIN TREATMENT

Montreal, QC — Ischemic stroke risk actually goes up during the first month after patients with chronic atrial fibrillation (AF) start warfarin, especially in the first week, and it plunges below baseline only after the first month of oral anticoagulation, suggests a retrospective case-control study ^[1].

The findings, which support post-hoc observations from some large randomized AF trials, could potentially encourage the field to reconsider how such patients start on warfarin, according to the study's authors, led by Dr Laurent Azoulay (Jewish General Hospital, Montreal, QC).

"If our results are confirmed, one possible way of minimizing the risk would be to bridge patients with therapies such as [standard] heparin and low-molecular–weight heparins. However, the appropriateness of such bridging strategies should be tailored to the individual patient, ensuring that the benefits always outweigh the risks," Azoulay toldheartwire by email. "In the meantime, physicians should exercise more vigilance during the initial days of [warfarin] treatment."

In the analysis, published December 19, 2013 in the European Heart Journal, a cohort of 70 766 chronic-AF patients followed a mean of four years showed an ischemic stroke rate of 2% per year overall. After matching stroke cases with up to 10 nonstroke patients based on demographics, CHADS 2 score, AF duration, comorbidities, medications used, and other factors, warfarin use—compared with no antithrombotic therapy—was associated with a 71% jump in ischemic stroke risk in the first 30 days of warfarin exposure. The risk fell to about one-half of baseline after the first 30 days.

In a separate analysis, the elevated ischemic-stroke risk was highest in the first week after starting warfarin, and peaked at three days (relative risk [RR], 2.33; 95% CI, 1.50–3.61).

The authors note that at the conclusions of the ROCKET AF and ARISTOTLE trials, patients who had been randomized to the new oral anticoagulants (NOACs) rivaroxaban (Xarelto, Bayer Pharma/Janssen Pharmaceuticals) or apixaban (Eliquis, Pfizer/Bristol-Myers Squibb), respectively, showed an increased stroke risk after transitioning to open-label warfarin. The current analysis corroborates those post-hoc findings, they write.

Relative Risk (95% CI) for Ischemic Stroke by Time After Starting Warfarin Vs No Antithrombotic Therapy

Time after starting warfarin	RR (95% CI)*
First 30 days	1.71 (1.39–2.12)
31–90 days	0.50 (0.34–0.75)
>90 days	0.55 (0.49–0.61)

*Reference: no antithrombotic use for at least one year. RR adjusted for birth year, sex, date of and time since AF diagnosis, alcohol use, smoking status, body-mass index, CHADS2 score, comorbidities, and current medications.

As to whether the NOACs could be used to lessen stroke risk in the first month after AF patients start warfarin, "it is unclear yet if they can be used safely as bridging therapies," Azoulay said. That the "safety signal" in ROCKET-AF and ARISTOTLE occurred after actively treated patients transitioned to warfarin casts doubt on a bridging role, "although it is important to mention that the short two-day bridging period in those trials was likely insufficient."

Importantly, according to Azoulay, "warfarin has been around for a long time, and has been extensively studied. This is a treatment that works extremely well, and has set a very high standard for the novel anticoagulants. As such, it would be premature at this point to advise against using it in favor of the novel anticoagulants on the basis of this possible initial stroke risk."

He and his colleagues offer several possible explanations for the early stroke hazard. Warfarin may induce an initial, transient hypercoagulable state, as it blocks various clotting factors but also some endogenous anticoagulant proteins, they write. Or it may simply be that, at least in the current study and the two randomized trials, achieved anticoagulation levels tended to be inadequate early after starting warfarin but better later on.

Commenting on the analysis by email, Dr Ralph L Sacco (University of Miami Miller School of Medicine, FL), who was not a coauthor, agreed with the two potential mechanisms proposed by Azoulay et al.

"The newer oral anticoagulants have some advantages over warfarin, since they have more specific anticoagulant effects, do not affect [the endogenous anticoagulant proteins to which Azoulay et al referred], and have a quicker onset of action," Sacco observed. "When we use warfarin we need to monitor it carefully, sometimes start with lower doses, and in high-risk cases, may need to use concomitant medications to anticoagulate patients before warfarin reaches the therapeutic range."

This study was funded by Bristol-Myers Squibb and Pfizer. Azoulay had no disclosures.