PDF: DOAC-IVT
(2023)Objective
To determine the risk of symptomatic intracranial hemorrhage (sICH) associated with intravenous thrombolysis in patients with ischemic stroke who had recent ingestion of DOAC within 48 hours
Study Summary
• Primary safety outcome (sICH): 2.5% (95% CI 1.6-3.8) DOAC vs 4.1% (95% CI 3.9-4.4) controls
• 42.7% received IVT without levels/reversal
• Results consistent across all selection strategies
Intervention
International, multicenter, retrospective cohort study from 64 centers (2008-2021) • 832 patients with recent DOAC use (confirmed within 48h) vs 32,375 controls without anticoagulation • IV alteplase 0.9 mg/kg (0.6 mg/kg in Japan) or tenecteplase 0.25 mg/kg, administered despite confirmed DOAC ingestion within 48 hours. Three selection strategies: (1) idarucizumab reversal before IVT (n=252, all dabigatran), (2) DOAC plasma level measurement (n=225), (3) IVT without levels or reversal (n=355).
Inclusion Criteria
Adult patients ≥18 years with acute ischemic stroke, confirmed DOAC ingestion within 48 hours, received IVT (with or without thrombectomy). Control group: patients with ischemic stroke receiving IVT without prior anticoagulation (no VKA with INR >1.7 or DOAC).
Study Design
Arms: DOAC group (n=832, confirmed intake <48h) vs Control group (n=32,375, no anticoagulation)
Patients per Arm: 832 DOAC, 32,375 controls (33,207 total)
Outcome
• Any ICH: 18.0% DOAC vs 17.4% controls, adjusted OR 1.18 (95% CI 0.95-1.45, p=0.14)
• Functional independence (mRS 0-2) at 90d: 45% DOAC vs 57% controls (unadjusted), adjusted OR 1.13 (95% CI 0.94-1.36, p=0.20)
• 90-day mortality: 17.9% DOAC vs 13.2% controls (unadjusted), adjusted OR 0.97 (95% CI 0.77-1.23, p=0.82)
• Results consistent across selection strategies: idarucizumab (sICH 1.2%), DOAC levels measured (sICH 3.1%), neither (sICH 3.1%)
Bottom Line
In this international observational cohort study, there was insufficient evidence of excess harm associated with off-label IVT in selected patients with recent DOAC ingestion. DOAC patients actually had lower odds of sICH compared to controls (2.5% vs 4.1%, adjusted OR 0.57). Results were consistent across different selection strategies (reversal, DOAC levels, or neither). However, selection bias is likely, as clinicians may have only offered IVT to lower-risk DOAC patients.
Major Points
- International multicenter retrospective cohort from 64 centers across Europe, Asia, Australia, New Zealand (2008-2021)
- 832 patients with confirmed DOAC intake <48h receiving IVT vs 32,375 controls without anticoagulation
- Median age 73 years, median NIHSS 9, 43.5% female
- Primary outcome: sICH lower in DOAC group (2.5% vs 4.1%, adjusted OR 0.57, p=0.02)
- Selection strategies: 30.3% idarucizumab reversal, 27% DOAC levels measured, 42.7% neither
- All selection strategies showed similar safety (sICH 1.2-3.1% across groups)
- DOAC patients were older, had more severe strokes, more LVOs, longer time to treatment
- DOAC types: dabigatran 41%, rivaroxaban 31%, apixaban 20%, edoxaban 8%
- Time from last DOAC intake: <12h (51.6% idarucizumab, 20.6% others), 12-24h (22%), >24h (17%)
- Median DOAC plasma level: 21 ng/mL (when measured); 25% had >50 ng/mL, 10% had >100 ng/mL
- Any ICH similar between groups (18.0% vs 17.4%, adjusted OR 1.18, p=0.14)
- Functional independence numerically lower in DOAC group but not significant after adjustment
- Results consistent across geographic regions and in sensitivity analyses
- No association between DOAC plasma levels or time from intake with sICH risk
- Likely selection bias: clinicians may have selected lower-risk patients for IVT
- 19% of IVT-eligible DOAC patients received IVT in selected centers providing data
- First large study specifically addressing DOAC patients with confirmed recent intake
Study Design
- Study Type
- International, multicenter, retrospective observational cohort study
- Randomization
- No
- Blinding
- Not applicable (observational)
- Sample Size
- 33207
- Follow-up
- 90 days
- Centers
- 64
- Countries
- Multiple countries across Europe, Asia, Australia, New Zealand
Primary Outcome
Definition: Symptomatic intracranial hemorrhage (sICH) within 36 hours after IVT, defined as any intracranial hemorrhage with associated neurological worsening of ≥4 points on NIHSS compared to immediately before deterioration, attributed to the radiologically evident ICH. Assessed by site investigators.
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 1,345/32,375 (4.1%, 95% CI 3.9-4.4%) | 21/832 (2.5%, 95% CI 1.6-3.8%) | - (Unadjusted OR 0.62 (95% CI 0.40-0.96, p=0.03); Adjusted OR 0.57 (95% CI 0.36-0.92, p=0.02)) | 0.02 (adjusted) |
Limitations & Criticisms
- Observational cohort design - inherent risk of selection bias
- Clinicians likely selected lower-risk patients for IVT (selection bias evident in baseline differences)
- DOAC patients were older, had more severe strokes, more LVOs, but still had lower sICH - counterintuitive
- Control population partially from different time frame and additional centers
- Only 19% of IVT-eligible DOAC patients received IVT - highly selected population
- Exact time of last DOAC ingestion unknown for 39.5% of patients
- Limited number of patients with very high DOAC levels (>100 ng/mL)
- DOAC plasma levels only measured in 27% of IVT patients, 40% of non-IVT patients
- No central adjudication of outcomes - site investigators assessed sICH
- IVT dosing heterogeneous (0.6 mg/kg in Japan vs 0.9 mg/kg elsewhere)
- Only 51 patients received tenecteplase - limited data on this agent
- Low data quality for DOAC patients not receiving IVT
- Functional outcome missing in 20% of DOAC patients, 12.6% of controls
- No information on systemic bleeding complications
- Baseline and follow-up imaging modality not standardized
- Residual and unmeasured confounding despite adjustment
- Time from onset to treatment longer in DOAC group - potential confounding
- Cannot exclude that lower sICH is due to protective effect of DOACs rather than safe selection
Citation
JAMA Neurol. 2023;80(3):233-243