Background: Optimal secondary stroke prevention for patients with cryptogenic stroke (CS) or transient ischemic attack (TIA) and patent foramen ovale (PFO) is not established. Observational studies imply a benefit for PFO closure over conventional medical management. Data from randomized trials are more equivocal or pending. Most studies report short term follow-up that will miss late adverse events (AEs). Methods: A systematic review was conducted of the Manufacturer and User Facility Device Experience (MAUDE) database for all septal closure device reports of late (>30 days) AEs between 1/1/00-9/10/11. Search terms were brand names of closure devices used in PFOs. To exclude patients not treated for a PFO, cases were those for whom 1) the narrative clearly described PFO as the target lesion, or 2) the device was designed only for PFO closure. Data extracted were device type, indication for closure, time from implantation to AE, and explantation. We collected clinical (e.g. stroke) and device-related (e.g. thrombus on device) AEs. Results: We reviewed 1250 reports - 50 (4%) met inclusion criteria. Devices were CardioSEAL (n=28), Gore Helex (n=11) and Amplatzer PFO or ASD Occluders (n=11). Clinical AEs were stroke/TIA or systemic embolism (n=20), arrhythmia/palpitations (n=6), pericardial effusion (n=3), visual symptoms (n=3), chest pain (n=2), asymptomatic (n=7), and unknown/other (n=18). Device AEs were thrombus (n=24), residual shunt (n=8), embolization (n=7), erosion (n=3), infection (n=2), fracture (n=1), and unknown/other (n=8). AEs led to device explantation in 28 patients. The range of time between device placement and AE was 30 days to 6 years; 23 reports (46%) were >1 year. Conclusion: The most common late clinical AE for device PFO closure was embolization and most common device event was thrombus. Nearly half of reported devices were explanted. These data cannot produce AE incidence rates because the number of implanted devices is unknown. We are also unable to determine the comparative safety of devices. However, the data emphasize that late AEs should be included when calculating the overall risk with the possible benefits of PFO closure.
Read full abstract