You are here
Warfarin Discontinuation Significantly Increases Stroke Risk
Warfarin prevents stroke in non-valvular atrial fibrillation (AF), but high early discontinuation rates place patients at risk of stroke. Freedman and colleagues presented the actual increase in stroke risk and risk duration after stopping vitamin K antagonists (VKA), primarily warfarin, at the European Society of Cardiology Congress 2015.
Stopping VKA in patients with AF more than doubles stroke risk and is highest in the first year after discontinuation and in the first year after AF diagnosis. VKA discontinuation leads to an excess of approximately 5 strokes per 100 persons in the 3 years following VKA discontinuation compared with current VKA use.
The United Kingdom Clinical Practice Research Datalink and other registry data were used to identify a cohort aged 45 to 89 years with VKA-naïve incident AF between 2001 and 2013. Within this cohort, patients with a new stroke were identified, and up to five age- and gender matched AF controls without stroke were randomly selected. The risk of VKA discontinuation was estimated using a conditional logistic regression model.
Among 68,288 AF patients, 5,116 with subsequent stroke were matched to 25,540 controls. The mean age of the sample was 79 years, with an average CHA2DS2VASc score of 4.4, and 52% were women.
During the first 180 days after VKA discontinuation, the odds ratio for developing a stroke was 1.89 compared with patients on continuous VKA. This jumped to 2.64 at 180 to 360 days after VKA discontinuation, and then decreased to 2.27 between 1 and 3 years.
The data showed that there was a steep risk increase during the first 6 months, already apparent within 3 months, to a peak at between 6 and 12 months, and a plateau between 2 and 3 years. In the first year after VKA cessation, this equated to an absolute increase in the stroke incidence rate of 2.2/100 patient years, and 2.7/100 patient years between 1 and 3 years. Restricting analysis to AF patients with CHA2DS2VASc scores of 2 or greater did not influence the results.
Although this study demonstrates the risk of stroke over time upon VKA discontinuation, it does not capture the reasons for discontinuation. For example, patients may have had a bleeding event or were lost to follow-up. This study calls for greater attention to issues surrounding anticoagulant discontinuation.
Source: European Society of Cardiology Congress; September 2, 2015