GORE-REDUCE
Patent Foramen Ovale Closure or Antiplatelet Therapy for Cryptogenic Stroke
Year of Publication: 2017
Authors: Lars Søndergaard, Scott E. Kasner, John F. Rhodes, ..., for the Gore REDUCE Clinical Study Investigators
Journal: New England Journal of Medicine
Citation: N Engl J Med 2017;377:1033-42
PDF: Clinical Question
Does PFO closure combined with antiplatelet therapy reduce recurrent stroke and new brain infarctions compared to antiplatelet therapy alone in patients with cryptogenic stroke and PFO?
Study Overview
Objective
PFO closure plus antiplatelet therapy versus antiplatelet therapy alone in patients with cryptogenic stroke and right-to-left shunt.
Study Summary
PFO closure combined with antiplatelet therapy was associated with a lower risk of recurrent ischemic stroke compared to antiplatelet therapy alone.
Intervention
PFO closure with HELEX Septal Occluder or Cardioform Septal Occluder plus antiplatelet therapy vs. antiplatelet therapy alone.
Patients per Arm
PFO Closure: 441, Antiplatelet Therapy: 223
Bottom Line
Among patients with PFO and cryptogenic stroke, PFO closure plus antiplatelet therapy significantly reduced recurrent ischemic stroke (1.4% vs 5.4%, HR 0.23, p=0.002) and new brain infarctions (4.7% vs 10.7%, p=0.02) compared to antiplatelet therapy alone, though closure was associated with increased atrial fibrillation (6.6% vs 0.4%).
Major Points
- First PFO closure trial mandating antiplatelet-only therapy in both arms (no anticoagulation allowed)
- 2:1 randomization: 441 patients to PFO closure + antiplatelet vs 223 to antiplatelet alone
- Two coprimary endpoints: clinical stroke and composite of clinical stroke or silent brain infarction
- 81% of patients had moderate or large interatrial shunts at baseline
- Median follow-up 3.2 years (range 2-5 years)
- Clinical stroke significantly reduced: 1.4% vs 5.4% (HR 0.23, 95% CI 0.09-0.62, p=0.002)
- New brain infarction reduced: 4.7% vs 10.7% (RR 0.44, 95% CI 0.24-0.81, p=0.02)
- Silent brain infarction alone showed no difference: 3.4% vs 4.0% (p=0.75)
- NNT = 28 patients to prevent 1 stroke at 24 months
- Complete PFO closure achieved in 75.6% at 12 months; effective closure in 94.5%
- Atrial fibrillation significantly higher with closure (6.6% vs 0.4%, p<0.001), but 83% occurred within 45 days and 59% resolved within 2 weeks
- Used GORE HELEX (through 2012) and CARDIOFORM (from 2012 onward) devices
- Statistical plan modified mid-trial: neuroimaging endpoint elevated from secondary to coprimary after enrollment of 467 patients
Design
Study Type: Multinational, prospective, randomized, controlled, open-label trial with blinded adjudication
Randomization: 1
Blinding: Open-label treatment assignment; blinded clinical events committee and MRI core laboratory for outcome adjudication
Enrollment Period: December 2008 to February 2015
Follow-up Duration: Median 3.2 years (IQR 2.2-4.8); minimum 2 years, maximum 5 years
Centers: 63
Countries: Canada, Denmark, Finland, Norway, Sweden, United Kingdom, United States
Sample Size: 664
Analysis: Intention-to-treat for primary analysis; Log-rank test for time-to-event (clinical stroke); Two-sample binomial test for brain infarction composite; Kaplan-Meier survival curves; P-values adjusted for multiplicity using Dubey and Armitage-Parmar procedure; Per-protocol and as-treated sensitivity analyses performed
Inclusion Criteria
- Age 18-59 years
- Cryptogenic ischemic stroke within 180 days before randomization
- Focal neurologic deficit lasting ≥24 hours OR <24 hours with MRI/CT evidence of infarction
- PFO with right-to-left shunt on transesophageal echocardiography with agitated saline (at rest or during Valsalva)
- Comprehensive workup ruling out other stroke mechanisms: no large-artery disease (≥50% stenosis or occlusion), no lacunar stroke, no cardioembolic source, no hypercoagulable disorder requiring anticoagulation, no arterial dissection
Exclusion Criteria
- Age ≥60 years
- Large-artery atherosclerotic disease (≥50% stenosis or vessel occlusion on CTA/MRA/ultrasound/catheter angiography)
- Small-vessel occlusive disease (lacunar stroke: <1.5cm deep infarction or typical lacunar syndrome)
- Established cardioembolic source
- Specific indication for anticoagulation
- Hypercoagulable disorder requiring anticoagulation
- Arterial dissection
- Uncontrolled diabetes mellitus
- Uncontrolled hypertension
- Autoimmune disease
- Recent history of alcohol or drug abuse
Baseline Characteristics
| Characteristic | Control | Active |
|---|---|---|
| N | 223 | 441 |
| Age (years) | 44.8 ± 9.6 | 45.4 ± 9.3 |
| Days from stroke to randomization | 101 ± 53 | 100 ± 52 |
| Male sex | 61.9% | 59.2% |
| Current smoking | 11.2% | 14.3% |
| Hypertension | 26.0% | 25.4% |
| Diabetes | 4.5% | 4.1% |
| Previous stroke or TIA | 10.3% | 14.1% |
| Previous stroke | 5.8% | 9.5% |
| Previous TIA | 4.9% | 5.9% |
| Stroke symptoms ≥24 hours | 89.2% | 91.2% |
| Stroke symptoms <24 hours with imaging | 10.8% | 8.8% |
| Small shunt | 19.9% | 18.1% |
| Moderate shunt | 43.5% | 39.1% |
| Large shunt | 36.6% | 42.8% |
| Atrial septal aneurysm | Not assessed (determined only at time of closure procedure) | 20.4% |
Arms
| Field | Control | PFO Closure + Antiplatelet Therapy |
|---|---|---|
| Intervention | Antiplatelet therapy chosen by investigator (uniform within each site): aspirin 75-325mg daily alone, OR aspirin 50-100mg + dipyridamole 225-400mg daily, OR clopidogrel 75mg daily. Continued for duration of trial. | Transcatheter PFO closure within 90 days using GORE HELEX Septal Occluder (through late 2012) or CARDIOFORM Septal Occluder (from late 2012 onward). Loading dose clopidogrel 300mg before/after procedure if not already on it, then 75mg daily for 3 days, then resume chosen antiplatelet regimen (same options as control). Median 28 days from randomization to procedure. |
| Duration | Minimum 2 years, median 3.2 years | Minimum 2 years, median 3.2 years |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Coprimary endpoint 1: Freedom from clinical ischemic stroke through ≥24 months (reported as recurrent stroke rate) | Primary | 12/223 (5.4%); 1.71 per 100 patient-years | 6/441 (1.4%); 0.39 per 100 patient-years | 0.23 | 0.002 |
| Coprimary endpoint 2: New brain infarction at 24 months (composite of clinical stroke or silent infarct ≥3mm on T2-weighted MRI) | Secondary | 19/177 (10.7%) with available MRI | 18/383 (4.7%) with available MRI | RR 0.44 (95% CI 0.24-0.81); Absolute difference 6.0% (95% CI 1.0-11.1) | 0.02 |
| Recurrent clinical ischemic stroke (component of coprimary 2) | Secondary | 12/177 (6.8%) | 5/383 (1.3%) | RR 0.19 (95% CI 0.07-0.54) | 0.005 |
| Silent brain infarction only (component of coprimary 2) | Secondary | 7/177 (4.0%) | 13/383 (3.4%) | RR 0.86 (95% CI 0.35-2.11) | 0.75 |
| Successful device implantation and retention | Secondary | N/A | 408/413 attempted (98.8%); 158 HELEX, 250 CARDIOFORM | ||
| Complete PFO closure immediately post-procedure | Secondary | N/A | 73.2% | ||
| Complete PFO closure at 12 months | Secondary | N/A | 75.6% | ||
| Effective closure (freedom from large shunt) at 12 months | Secondary | N/A | 94.5% | ||
| Any serious adverse event | Adverse | 62/223 (27.8%) | 102/441 (23.1%) | 0.22 | |
| Death | Adverse | 0 | 2/441 (0.5%): 1 suicide at 131 days, 1 MI at 1045 days | 0.55 | |
| Any atrial fibrillation or flutter | Adverse | 1/223 (0.4%) | 29/441 (6.6%); 83% within 45 days, 59% resolved within 2 weeks | <0.001 | |
| Serious atrial fibrillation or flutter | Adverse | 1/223 (0.4%) | 10/441 (2.3%) | 0.11 | |
| Device-related serious adverse events | Adverse | N/A | 6/441 (1.4%): 3 device dislocation, 2 device thrombosis, 1 aortic dissection | ||
| Procedure-related serious adverse events | Adverse | N/A | 11/441 (2.5%) | ||
| Serious bleeding adverse event | Adverse | 6/223 (2.7%) | 8/441 (1.8%); 4 procedure-associated (3 vascular access, 1 tamponade) | 0.57 | |
| Deep-vein thrombosis or pulmonary embolism | Adverse | 2/223 (0.9%) | 3/441 (0.7%) | 1.00 |
Subgroup Analysis
Subgroup analyses by age (18-45 vs 46-59), sex, region (Europe/Canada vs US), and shunt size (small vs moderate-to-large) showed no significant interactions. Treatment effect favoring PFO closure was consistent across all subgroups: Age 18-45: HR 0.26 (p=0.04); Age 46-59: HR 0.21 (p=0.02); Male: HR 0.19 (p=0.01); Female: HR 0.31 (p=0.11); Europe/Canada: HR 0.23 (p=0.03); US: HR 0.24 (p=0.03); Small shunt: HR 0.27 (p=0.26); Moderate-to-large shunt: HR 0.18 (p=0.001). No interaction P-values were significant.
Criticisms
- Open-label design increases risk of ascertainment bias despite blinded outcome adjudication
- Differential dropout: 8.8% in closure group vs 14.8% in antiplatelet group discontinued trial
- Crossover issues: 14 patients (6.3%) in antiplatelet group underwent closure outside trial; 28 patients (6.3%) assigned to closure never attempted device implantation
- Statistical plan changed mid-trial: neuroimaging endpoint elevated from secondary to coprimary after 467 patients enrolled due to lower-than-expected event rates in other PFO trials
- Silent infarcts not reduced despite reduction in clinical strokes - raises possibility of referral bias in open-label setting
- No requirement for Holter or prolonged cardiac monitoring to exclude paroxysmal atrial fibrillation
- MRI detection threshold (≥3mm on T2-weighted MRI) may be insensitive for smaller silent infarcts
- Higher than expected stroke rate in antiplatelet arm compared to previous trials
- Relatively low total event rate limits subgroup analysis power
- Atrial fibrillation significantly increased with closure - long-term stroke implications unclear
- Younger, carefully selected population (age <60, strict cryptogenic criteria, controlled risk factors) may limit generalizability
- Missing follow-up MRI data (final neuroimaging available in 383/441 [86.8%] closure vs 177/223 [79.4%] antiplatelet, i.e., missing 13.2% vs 20.6%)
- Two different devices used (HELEX vs CARDIOFORM) though outcomes were similar
- Median time from stroke to randomization was 102 days - relatively long interval
Funding
W.L. Gore and Associates (device manufacturer and sponsor)
Based on: GORE-REDUCE (New England Journal of Medicine, 2017)
Authors: Lars Søndergaard, Scott E. Kasner, John F. Rhodes, ..., for the Gore REDUCE Clinical Study Investigators
Citation: N Engl J Med 2017;377:1033-42
Content summarized and formatted by NeuroTrials.ai.