Alteplase Not Harmful in NOAC Users, Even in a More Bleeding-Prone Cohort

Stroke patients treated with intravenous alteplase (Activase) did not have more intracranial hemorrhages if they had recently taken non-vitamin K antagonist oral anticoagulants (NOACs), according to an Asian cohort study that adds to the growing evidence for this off-label practice.

Between groups with recent NOAC therapy versus no treatment with anticoagulants, small differences in bleeding and mortality events after alteplase administration did not reach statistical significance in a propensity score-matched analysis:

• Intracranial hemorrhage during hospitalization after alteplase administration: 9.9% with NOACs vs 7.4% with no anticoagulants (OR 1.37, 95% CI 0.62-3.03)
• Major bleeding during the index hospitalization: 13.2% vs 8.2% (OR 1.69, 95% CI 0.83-3.45)
• All-cause 30-day mortality: 8.8% vs 11.0% (OR 0.78, 95% CI 0.35-1.73)
• In-hospital mortality: 4.4% vs 9.3% (OR 0.45, 95% CI 0.15-1.29)

Similarly, outcomes were no different between NOAC and warfarin groups. Results remained consistent in a meta-analysis pooling this Taiwanese analysis with prior studies, reported Huei-Kai Huang, MD, of Hualien Tzu Chi Hospital in Taiwan, and colleagues in JAMA Internal Medicine.

“Our meta-analysis synthesized the latest evidence, consistently demonstrating that pretreatment with NOACs was not associated with excess harm in patients receiving intravenous thrombolysis for acute ischemic stroke,” they wrote. “These findings provide robust evidence with potential implications for future research and guideline updates.”

Current stroke guidelines do not strongly endorse alteplase for these patients, giving thrombolysis a class III recommendation within 48 hours of last direct oral anticoagulant (DOAC) intake for fear of bleeding. This exclusion of DOAC users from stroke thrombolytics has been contested, however, as reports have suggested that carefully selected DOAC users do not have excess bleeding risks.

One large U.S. study found that intravenous thrombolysis could be safely given to ischemic stroke patients who had taken these medications in the preceding 7 days, while a Japanese analysis showed the same for thrombolysis within a day of DOAC administration. Furthermore, another group reported that off-label IV thrombolysis in DOAC users was in fact associated with a lower incidence of symptomatic intracranial hemorrhage compared with controls not on anticoagulant therapy.

In the present study, Huang and colleagues considered those who had prescriptions for dabigatran (Pradaxa), rivaroxaban (Xarelto), apixaban (Eliquis), or edoxaban (Savaysa) within 2 days before the index hospitalization as being pretreated with DOACs before receiving alteplase.

The setting of this cohort is notable, as Asian individuals are thought to have inherently higher bleeding risks during thrombolytic therapy.

“Furthermore, our study specifically involved patients who neither had their NOAC plasma levels measured nor received any antidote. While European guidelines recommend the measurement of NOAC plasma levels before administering intravenous alteplase, practical limitations in performing these measurements are a reality faced by many hospitals worldwide, including those in Taiwan,” the authors noted.

Their nationwide, population-based cohort study relied on Taiwan’s National Health Insurance Research Database that covers claims data for the entire country. They included 7,483 adults (mean age 67.4 years, 38.9% women) who had been treated with alteplase for acute ischemic stroke from 2011 to 2020.

The vast majority of participants were not on anticoagulants prior to their stroke, with only 1.2% on NOACs and another 2.4% on warfarin. The primary indication for oral anticoagulant therapy was atrial fibrillation.

The NOAC group was of disproportionately older age and comprised more women. This group also had a higher prevalence of prior stroke and coronary artery disease.

For the comparison of NOACs versus no anticoagulation, propensity score matching was performed 1:4 and yielded 91 and 364 patients, respectively, with balanced baseline differences. For the analysis comparing NOACs versus warfarin, 1:1 matching yielded comparator groups including 77 patients each in the NOAC and warfarin groups.

Huang’s group cautioned that each patient’s medication history was derived from drug dispensing records in the database, based on expiration dates of prescriptions. They were therefore unable to estimate exactly if or when patients were taking their prescribed oral anticoagulants.

Additionally, despite propensity score matching to balance baseline differences, the retrospective study left room for residual confounding.