PDF: DEFENSE-PFO
(2018)Objective
To evaluate whether PFO closure is superior to medical therapy alone in patients with cryptogenic stroke and high-risk PFO characteristics (large size ≥2mm, atrial septal aneurysm, or hypermobility ≥10mm)
Study Summary
• Primary composite endpoint (stroke, vascular death, or major bleeding) occurred in 0% vs 12.9% (p=0.013)
• All 53 PFO closures were successful with acceptable safety profile
Intervention
Transcatheter PFO closure using Amplatzer PFO Occluder plus medical therapy versus medical therapy alone (antiplatelet or anticoagulation at investigator discretion)
Inclusion Criteria
Cryptogenic ischemic stroke within 6 months; High-risk PFO on TEE defined as: PFO size ≥2mm OR atrial septal aneurysm (≥15mm protrusion) OR hypermobility (≥10mm septal excursion); Right-to-left shunting with agitated saline
Study Design
Arms: PFO Closure + Medical Therapy vs Medical Therapy Only
Patients per Arm: 60 per arm (120 total)
Outcome
• Ischemic stroke: 0% vs 10.5%
• Number needed to treat: 10 patients to prevent 1 stroke
Bottom Line
In patients with cryptogenic stroke and high-risk PFO characteristics (large size ≥2mm, atrial septal aneurysm, or hypermobility ≥10mm), PFO closure resulted in significantly lower rates of stroke recurrence (0% vs 10.5%, p=0.023) and the composite primary endpoint (0% vs 12.9%, p=0.013) compared to medical therapy alone over 2 years of follow-up.
Major Points
- Multicenter, randomized, open-label trial comparing PFO closure + medical therapy vs medical therapy alone
- Enrolled 120 patients with cryptogenic stroke and high-risk PFO features defined by TEE
- High-risk features: PFO size ≥2mm, atrial septal aneurysm (≥15mm protrusion), or hypermobility (≥10mm septal excursion)
- All 53 PFO closures were successful using Amplatzer PFO Occluder
- Primary endpoint (stroke, vascular death, or major bleeding) occurred in 0% of PFO closure group vs 12.9% of medication-only group (p=0.013)
- Ischemic stroke recurrence: 0% vs 10.5% (p=0.023)
- Number needed to treat: 10 patients to prevent 1 stroke at 2 years
- Trial terminated early due to safety concerns after publication of CLOSE trial with similar design
- Procedural complications included atrial fibrillation (n=2), pericardial effusion (n=1), and pseudoaneurysm (n=1)
- Results support morphologic risk stratification to select optimal candidates for PFO closure
Study Design
- Study Type
- Multicenter, randomized, open-label, superiority trial
- Randomization
- Yes
- Blinding
- Open-label with masked outcome assessment
- Sample Size
- 120
- Follow-up
- Median 2.8 years (IQR: 0.9-4.1 years)
- Centers
- 2
- Countries
- South Korea
Primary Outcome
Definition: Composite of stroke, vascular death, or TIMI-defined major bleeding during 2 years of follow-up
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| 6 events; 2-year Kaplan-Meier rate 12.9% (95% CI: 3.2-22.6%) | 0 events; 2-year Kaplan-Meier rate 0% | - | 0.013 |
Limitations & Criticisms
- Early termination due to safety concerns after publication of CLOSE trial, resulting in underpowered study (120 enrolled vs 210 planned)
- Open-label design, though outcome assessment was masked
- Only 2 centers in South Korea, limiting generalizability
- Seven patients randomized to PFO closure declined the intervention
- Four patients in medication-only group crossed over to receive PFO closure during follow-up
- Follow-up MRI was performed in only 60% of patients (72/120)
- Potential selection bias due to single-country enrollment
- Medication choice was at investigator discretion, introducing treatment heterogeneity
- Unable to calculate hazard ratio due to zero events in intervention arm
- Relatively short median follow-up (2.8 years) compared to some other PFO trials
Citation
J Am Coll Cardiol 2018;71(20):2335-42