CLOSURE I Trial: STARFlex Device Closure vs Medical Therapy for Cryptogenic Stroke or TIA with PFO (Furlan et al., 2012)
In patients 18–60 with cryptogenic ischemic stroke or TIA and a PFO, does transcatheter closure with the STARFlex Septal Closure System plus 6 months clopidogrel and 2 years aspirin reduce stroke / TIA / death vs medical therapy alone? First major PFO closure RCT; the STARFlex device has since been withdrawn from the market.
Furlan AJ et al. (NEJM 2012;366:991-999) · doi:10.1056/NEJMoa1009639 · 909 patients
Population
Included
- Age 18–60 years
- Cryptogenic ischemic stroke or TIA within 6 months of randomization
- PFO confirmed by transesophageal echocardiography with agitated saline
- Workup excluding alternative stroke etiologies (large-artery atherosclerosis, established cardioembolic source, small-vessel disease, hypercoagulable state)
Excluded
- Identifiable alternative stroke mechanism
- Age outside 18–60
- Other indication for closure or for anticoagulation
- Pregnancy
Primary Outcome
2-year composite: stroke or TIA, 30-day all-cause death, or days 31–730 death from neurologic causes
Negligible absolute difference
Primary composite at 2 years
Safety
Atrial fibrillation
5.7%
0.7%
New-onset atrial fibrillation: 5.7% (PFO closure) vs 0.7% (medical therapy), P<0.001. Major vascular procedural complications occurred in 3.2% of the closure arm. Source: Furlan AJ et al., NEJM 2012;366:991-999.
Trial Design
Type
- Multicenter, randomized, open-label superiority trial with blinded endpoint adjudication
- 1:1 randomization (PFO closure + antithrombotic : medical therapy)
- 87 sites in US and Canada
- First major randomized trial of PFO closure for cryptogenic stroke/TIA
- STARFlex device exclusively (NMT Medical; device subsequently withdrawn from market)
Timeline
Enrolled June 2003 to October 2008; published NEJM 2012
N
909
Enrollment
909 patients (447 closure / 462 medical therapy) at 87 US + Canadian sites. Enrolled June 2003 to October 2008. 2-year follow-up. STARFlex Septal Closure System (NMT Medical; subsequently withdrawn from market). Published NEJM 2012.
ClinicalTrials.gov
NCT00201461Bedside Pearl
CLOSURE-I (NEJM 2012) was the first major PFO closure RCT and was NEGATIVE (HR 0.78, P=0.37). STARFlex device (now withdrawn) and broad inclusion likely contributed. Established the AF excess pattern (5.7% vs 0.7%) seen in every subsequent PFO trial. Historical context only. Do not cite against modern PFO closure, which is supported by the 2017 NEJM cluster (CLOSE, REDUCE, RESPECT long-term).