Skip to main content
NeuroWiki
Antiplatelets

RESPECT (Long-Term): Extended Follow-up of PFO Closure with the Amplatzer Occluder for Cryptogenic Stroke (Saver et al., 2017)

In patients 18–60 with cryptogenic ischemic stroke and a PFO, does transcatheter closure with the Amplatzer Occluder reduce recurrent ischemic stroke compared with medical therapy over extended follow-up? Note: original 2013 primary analysis at 2.1 years missed ITT significance (P=0.08); this is the pre-specified extended-follow-up analysis at median 5.9 years.

Saver JL et al. (NEJM 2017;377:1022-1032) · doi:10.1056/NEJMoa1610057 · 980 patients

Population

Included

  • Age 18–60 years
  • Cryptogenic ischemic stroke within 270 days before randomization
  • TEE-confirmed PFO
  • Mechanism workup excluded large-vessel arteriopathy, established cardioembolic source, intrinsic small-vessel disease, arterial hypercoagulable state (antiphospholipid antibodies, hyperhomocysteinemia)

Excluded

  • Identifiable competing stroke mechanism
  • Age outside 18–60
  • Inability to undergo randomization within 270 days of index stroke

Source: Saver et al., N Engl J Med 2017;377:1022-1032· Retrieved 2026-06-09

Primary Outcome

Recurrent ischemic stroke at median 5.9-year follow-up

PFO Closure (Amplatzer)Better outcome
3 / 100
Medical Therapy
5 / 100

Negligible absolute difference

Recurrent ischemic stroke at 5.9y

Risk ratio HR 0.5595% CI 0.31–0.999p = 0.046
NNT42to prevent one recurrent stroke over 5 years

Study Arms

Agent
Amplatzer PFO Occluder (St. Jude Medical)
Dose
Single device implantation
Route
Percutaneous transcatheter implantation under fluoroscopic and echocardiographic guidance
Frequency
Single closure procedure within 21 days of randomisation
Duration
Permanent implant; antithrombotic therapy at site investigator discretion after the first 6 months
Co-interventions
Patients continued their prerandomisation antithrombotic regimen until placement; after implantation, aspirin 81 to 325 mg plus clopidogrel 75 mg daily for 1 month, then aspirin monotherapy for 5 months

Extended follow-up (median 5.9 years) of the intention-to-treat population from the original RESPECT trial.

Safety

Procedure-related atrial fibrillation/flutter

PFO Closure (Amplatzer)

1.4%

Medical Therapy

0%

Procedure-related atrial fibrillation/flutter: 7 events in 499 PFO closure patients (1.4%), all resolved before discharge from the hospital. Source: Saver et al., NEJM 2017, Table 3.

Trial Design

Type

  • Multicenter, randomized, open-label, controlled trial with blinded endpoint adjudication
  • 1:1 PFO closure vs medical therapy
  • Stratified randomization by site, planned antithrombotic regimen, and presence/absence of atrial septal aneurysm
  • 69 sites in US and Canada
  • Original primary analysis (Carroll JD et al., NEJM 2013;368:1092–1100) at median 2.1-year follow-up: HR 0.49 (95% CI 0.22–1.11, P=0.08). Did not meet ITT primary
  • Extended follow-up (this paper) at median 5.9 years. Pre-specified as exploratory
  • No alpha-spending adjustment applied to long-term P-value; group-sequential Lan–DeMets threshold for final analysis would have been 0.043

Timeline

Enrolled August 2003 to December 2011; original primary analysis reported 2013 at median 2.1y; extended follow-up database lock May 31, 2016, median 5.9y

N

980

Enrollment

980 patients (499 PFO closure / 481 medical therapy) at 69 sites in US + Canada. Enrolled August 2003 to December 2011. Original primary analysis 2013 at median 2.1y was borderline (P=0.08); extended follow-up reported here at median 5.9y. Published NEJM 2017.

ClinicalTrials.gov

NCT00465270

Bedside Pearl

RESPECT (long-term): in patients 18–60 with cryptogenic ischemic stroke and PFO, Amplatzer Occluder vs medical therapy reduced recurrent ischemic stroke over median 5.9y (3.6% vs 5.8%, HR 0.55, P=0.046, NNT 42). The original 2013 primary at 2.1y was negative; benefit emerges with time. Read with CLOSE and REDUCE (same NEJM 2017 issue).

NeuroWiki is a clinical reference. It does not substitute for your clinical judgment, current guidelines, or your institution's protocol. Verify before acting. Do not enter patient names, MRNs, or dates of birth. Privacy Policy