Log in to MyACC
Higher Incidence of Ischemic Stroke With NOACs
Study Questions:
Which drug class, novel oral anticoagulants (NOACs) or warfarin, is more effective in the real world for preventing acute ischemic stroke in patients with atrial fibrillation (AF) on home anticoagulation?
Methods:
This study retrospectively analyzed data for all patients in 2015-2016 with registered encounters in the US hospital network of Hospital Corporation of America Healthcare who were taking an oral anticoagulant as a home medication as reported by the patient or a family member at the time of hospitalization and who also carried a diagnosis of AF. The incidence of acute ischemic stroke, intracranial hemorrhage, or nontraumatic hemorrhage in these patients with AF taking home NOACs or warfarin was compared utilizing odds ratios and Fisher exact tests. For further analysis, a risk score analogous to CHADS2 was utilized to distinguish the effect of anticoagulant type versus underlying stroke risk factors.
Results:
The study identified >125,000 patients with AF on anticoagulation including 67,262 on NOACs, 59,066 on warfarin, and 7,033 who had reportedly taken both (possibly representing patients who had responded poorly to the initially prescribed anticoagulant and then switched to another). Acute ischemic stroke incidence was higher overall in patients on NOACs than in those on warfarin (odds ratio [OR] 1.291, p < 0.001). The incidence was higher for all 4 NOACs compared with warfarin, specifically edoxaban (OR 1.399, p = 0.117), apixaban (OR 1.447, p < 0.001), rivaroxaban (OR 1.069, p = 0.055), and dabigatran (OR 1.289, p < 0.001). The findings were statistically significant for apixaban and dabigatran but not for edoxaban or rivaroxaban. The CHADS2-analogous risk score was lower in the NOAC group (mean = 2.66) than in the warfarin group (mean = 2.85). Preexisting cerebrovascular disease was rare and was present in 0.2% of NOAC patients and 0.3% of warfarin patients. At presentation, NOAC patients had a slightly lower but statistically significant National Institutes of Health Stroke Scale (NIHSS) score compared with those on warfarin (4.78 vs. 5.23, p = 0.011). Hemorrhagic stroke incidence was lower in patients taking NOACs than in those taking warfarin (OR 0.7064, p < 0.001), and the incidence of general nontraumatic hemorrhage was also lower (OR 0.691, p < 0.001). Multianticoagulant patients (those identified as taking both an NOAC and warfarin who may have been switched from one to the other) had an even higher incidence of acute ischemic stroke than patients on only NOAC (OR 1.878, p < 0.001) or only warfarin (OR 2.425, p < 0.001). There was no statistically significant difference in incidence of intracranial hemorrhage, but these patients had higher rates of nontraumatic hemorrhage compared with those on only NOAC (OR 2.059, p < 0.001) or warfarin (OR 1.423, p < 0.001). Multianticoagulant patients differed significantly from those taking only NOAC or warfarin; they had higher overall CHADS2-analogous risk scores (mean 3.30 vs. 2.66 NOAC only and 2.85 warfarin only). They also had a much higher prevalence of congestive heart failure at 99.5% (compared with 50.3% and 64.6% for NOAC or warfarin only patients, respectively). NIHSS stroke severity at presentation was not statistically different in multianticoagulant patients than in those on only NOAC or warfarin.
Conclusions:
This study finds that patients with AF taking NOACs at home have a higher incidence of acute ischemic stroke than those taking warfarin. Increased stroke incidence was higher with each of the four NOACs and was statistically significant for apixaban and dabigatran. The authors suggest that real-world NOAC patients, who may have compliance gaps, might experience decreased anticoagulation effectiveness related to the shorter half-life of NOACs relative to warfarin and that transition to a NOAC from warfarin may be inadvisable in some patients.
Perspective:
These results seem to contradict recent studies based predominantly upon clinical trials or meta-analyses of clinical trials. Real-world patients may differ from those in clinical trials, and patients with hospital encounters, like those in this study, may differ from those in the community. The authors suggest that close attention to background covariate risk factors may be merited in future studies and that monitoring levels of anticoagulation in NOAC patients may be a significant unmet need in stroke prevention; comparative effectiveness studies with head-to-head trials of NOACs may clarify current discrepancies in research.
Keywords: Atrial Fibrillation, Anticoagulants, Stroke, Incidence, Brain Ischemia, Intracranial Hemorrhages, Warfarin, Antithrombins, Pyridones, Pyrazoles, Pyridines, Thiazoles, Hospitalization
< Back to Listings
