APACHE-AF
Apixaban versus no anticoagulation after anticoagulation-associated intracerebral haemorrhage in patients with atrial fibrillation in the Netherlands (APACHE-AF): a randomised, open-label, phase 2 trial
Year of Publication: 2021
Authors: Schreuder FHBM, van Nieuwenhuizen KM, Hofmeijer J, ..., Klijn CJM; APACHE-AF Trial Investigators
Journal: The Lancet Neurology
Citation: Lancet Neurol 2021;20(11):907-916
PDF: Clinical Question
In patients with atrial fibrillation who survive anticoagulation-associated intracerebral hemorrhage, what are the rates of non-fatal stroke or vascular death when treated with apixaban compared to avoiding anticoagulation?
Study Overview
Objective
To estimate the rates of non-fatal stroke or vascular death in AF patients who survived anticoagulation-associated ICH when treated with apixaban versus avoiding anticoagulation, to inform design of a larger phase 3 trial
Study Summary
- No difference in primary outcome between apixaban and avoiding anticoagulation (aHR 1.05, p=0.90)
- Very high annual event rate (~12%) in both groups regardless of treatment strategy
- More recurrent ICH with apixaban (8% vs 2%) but similar ischemic stroke rates (12% vs 12%)
Intervention
Apixaban 5mg BID (or 2.5mg BID if dose reduction criteria met) vs avoid anticoagulation (antiplatelet therapy permitted at physician discretion)
Patients per Arm
50 vs 51
Bottom Line
Patients with AF who had anticoagulation-associated ICH have a very high annual risk of non-fatal stroke or vascular death (~12%) regardless of whether they receive apixaban or avoid anticoagulation. This phase 2 trial was underpowered to detect a treatment difference and underscores the need for larger trials.
Major Points
- Phase 2 trial designed to inform sample size for a larger phase 3 trial
- Specifically enrolled patients who had ICH WHILE ON anticoagulation (anticoagulation-associated ICH)
- No significant difference in primary composite outcome between groups (adjusted HR 1.05)
- Very high annual event rates in both groups (~12%) highlight the severity of this clinical situation
- Numerically more recurrent ICH with apixaban (8% vs 2%) but not statistically significant
- Similar rates of ischemic stroke in both groups (12% vs 12%)
- 22% crossover rate in avoid group (11/51 crossed to anticoagulation)
- Trial underpowered; authors estimate 5744 participants per group needed for definitive trial
Design
Study Type: Randomized, open-label, phase 2 trial with masked endpoint assessment (PROBE design)
Randomization: 1
Blinding: Open-label for participants and treating physicians; outcome adjudicators masked to treatment allocation and antithrombotic drug use
Enrollment Period: January 2015 to July 2020
Follow-up Duration: Median 1.9 years (IQR 1.0-3.1); minimum 6 months
Centers: 16
Countries: Netherlands
Sample Size: 101
Analysis: Intention-to-treat; Cox proportional hazards regression adjusted for age and ICH location (minimization variables); on-treatment analysis as secondary
Inclusion Criteria
- Adults ≥18 years
- Spontaneous ICH (including isolated intraventricular hemorrhage) in previous 7-90 days
- ICH occurred during treatment with anticoagulation (VKA, DOAC, or therapeutic-dose heparin/LMWH)
- Documented paroxysmal or non-paroxysmal non-valvular atrial fibrillation
- CHA₂DS₂-VASc score ≥2
- Modified Rankin Scale ≤4
- Clinical equipoise regarding optimal treatment
Exclusion Criteria
- Conditions other than AF requiring long-term anticoagulation (e.g., mechanical heart valve)
- Other serious bleeding events besides ICH in previous 6 months
- High risk of bleeding
- Ischemic stroke in previous 7 days
- Active alcohol or drug misuse
- Life expectancy <1 year
- Severe renal insufficiency
- Liver test abnormalities
- Women with childbearing potential, pregnant, or breastfeeding
Baseline Characteristics
| Characteristic | Apixaban | Avoid Anticoagulation |
|---|---|---|
| N | 50 | 51 |
| Median Age (IQR) | 77 (74-83) | 79 (72-83) |
| Age ≥75 years | 72% | 59% |
| Male Sex | 54% | 55% |
| White Ethnicity | 98% | 100% |
| CHA₂DS₂-VASc score median (IQR) | 4 (3-5) | 4 (3-5) |
| Previous Ischemic Stroke | 20% | 27% |
| Previous ICH | 8% | 6% |
| Previous MI | 16% | 8% |
| Hypertension | 90% | 98% |
| Diabetes | 18% | 14% |
| OAC at time of ICH - VKA | 66% | 73% |
| OAC at time of ICH - DOAC | 34% | 27% |
| ICH Volume mL median (IQR) | 5.7 (1.8-11.9) | 6.6 (2.3-14.3) |
| ICH Location - Lobar | 28% | 27% |
| ICH Location - Deep | 54% | 49% |
| ICH Location - Brainstem | 4% | 2% |
| ICH Location - Cerebellum | 14% | 20% |
| Small Vessel Disease CT score >0 | 42% | 35% |
| Edinburgh CT criteria for CAA | 6% | 12% |
| Time from ICH to randomization days (IQR) | 45 (22-70) | 46 (20-76) |
| Systolic BP at randomization mmHg (IQR) | 142 (130-159) | 140 (129-157) |
| mRS at randomization median (IQR) | 2 (1-3) | 3 (1-4) |
| ICH Location - IVH only | 2% |
Arms
| Field | Apixaban | Control |
|---|---|---|
| Intervention | Apixaban 5 mg twice daily orally; reduced dose 2.5 mg twice daily if CrCl ≤30 mL/min OR if 2 of 3 criteria present: age ≥80, weight ≤60 kg, serum creatinine ≥133 µmol/L | No anticoagulation; oral antiplatelet therapy permitted at treating physician discretion (51% received antiplatelet: ASA 80mg, carbasalate calcium 100mg, clopidogrel 75mg, or dipyridamole combinations) |
| Duration | Ongoing treatment with follow-up minimum 6 months | Ongoing with follow-up minimum 6 months |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Composite of non-fatal stroke (ischemic stroke, ICH, or SAH) or vascular death, whichever came first | Primary | 12 (24%); Annual event rate 11.9% (95% CI 6.2-20.8) | 13 (26%); Annual event rate 12.6% (95% CI 6.7-21.5) | 0.90 | |
| Recurrent Intracerebral Hemorrhage | Secondary | 1 (2%) | 4 (8%) | 0.21 | |
| All Major Haemorrhagic Events | Secondary | 3 (6%) | 6 (12%) | 0.29 | |
| Subarachnoid Haemorrhage | Secondary | 0 (0%) | 0 (0%) | ||
| Traumatic Intracranial Haemorrhage | Secondary | 0 (0%) | 0 (0%) | ||
| Ischemic Stroke | Secondary | 6 (12%) | 6 (12%) | 0.94 | |
| Any Stroke | Secondary | 7 (14%) | 10 (20%) | ||
| Unclassified Stroke | Secondary | 0 (0%) | 0 (0%) | ||
| All Major Occlusive Events (ischemic stroke, MI, PE, systemic embolism) | Secondary | 11 (22%) | 6 (12%) | 0.13 | |
| Myocardial Infarction | Secondary | 2 (4%) | 0 (0%) | ||
| Pulmonary Embolism | Secondary | 4 (8%) | 0 (0%) | ||
| Systemic Embolism | Secondary | 0 (0%) | 0 (0%) | ||
| Vascular Death | Secondary | 7 (14%) | 5 (10%) | ||
| All-cause Death | Secondary | 11 (22%) | 9 (18%) | ||
| All Major Vascular Events according to the protocol (major haemorrhagic, major occlusive, or vascular death) | Secondary | 16 (31%) | 14 (28%) | 0.54 | |
| All Major Vascular Events (MI, stroke, or vascular death - ATT definition) | Secondary | 13 (25%) | 13 (26%) | 0.85 | |
| Serious Adverse Events (non-outcome) | Adverse | 29 (57%) | 29 (58%) | ||
| Major Extracranial Hemorrhage | Adverse | 2 (4%) | 2 (4%) | ||
| Clinically Relevant Non-major Bleeding | Adverse | 0 (0%) | 1 (2%) |
Subgroup Analysis
Prespecified exploratory subgroup analyses showed no evidence of heterogeneity for the primary outcome based on: ICH location (lobar vs non-lobar), age (<75 vs ≥75), sex, time since ICH (<7 weeks vs ≥7 weeks), CHA₂DS₂-VASc score, intention for antiplatelet in comparator group, or CT small vessel disease score.
Criticisms
- Small sample size (n=101) - phase 2 trial underpowered to detect treatment differences
- Wide 95% confidence intervals preclude definitive conclusions
- Open-label design may introduce reporting, observation, and detection bias
- High crossover rate: 22% (11/51) in avoid group crossed over to anticoagulation
- 9 participants in apixaban group did not start or discontinued treatment
- Blood pressure control during follow-up was suboptimal
- Nearly all participants (99%) were White - limited generalizability to other ethnicities
- Small median ICH volume (5-7 mL) suggests selection of patients who recovered well
- Single country (Netherlands) enrollment only
Funding
Dutch Heart Foundation (grant 2012T077)
Based on: APACHE-AF (The Lancet Neurology, 2021)
Authors: Schreuder FHBM, van Nieuwenhuizen KM, Hofmeijer J, ..., Klijn CJM; APACHE-AF Trial Investigators
Citation: Lancet Neurol 2021;20(11):907-916
Content summarized and formatted by NeuroTrials.ai.