2 The procedure

2.1 Indications and current treatments

2.1.1 A PFO is the persistence of an opening (the foramen ovale) in the septum between the right atrium and left atrium of the heart. In the fetus, the foramen ovale allows blood to bypass the lungs, directly from the venous to the arterial side of the circulation. After birth the foramen ovale normally closes but in approximately 25% of people it remains either fully or partially patent throughout life. Studies evaluating PFO closure to prevent paradoxical thromboembolism noted a change in the incidence of migraine amongst patients. Any physiological effect of PFO closure in migraine treatment is not understood.

2.1.2 Current treatment for patients with recurrent migraine is aimed at either preventing or aborting episodes through medical management. Invasive treatments such as nerve blocks or physical therapies such as acupuncture are sometimes used if medical therapy has failed.

2.2 Outline of the procedure

2.2.1 Percutaneous closure of PFO for recurrent migraine is carried out with the patient under local anaesthesia and intravenous sedation, or general anaesthesia. A guidewire and delivery sheath are introduced via a small incision in the femoral vein into the heart and across the PFO. A closure device is then inserted through the opening via the delivery sheath and released, closing the PFO.

2.2.2 A range of different devices are available for this procedure.

Sections 2.3 and 2.4 describe efficacy and safety outcomes from the published literature that the Committee considered as part of the evidence about this procedure. For more detailed information on the evidence, see the migraine- and embolism-related overviews.

2.3 Efficacy

2.3.1 Immediate closure of the PFO (confirmed with echocardiography) was reported in 99% (148/150), 89% (42/47), 97% (179/185), 100% (76/76) and 99.8% (823/825) of patients in studies across a range of indications.

2.3.2 A randomised controlled trial (RCT) of 147 patients treated either by the procedure (n = 74) or a sham procedure (n = 73) reported that 3 patients in each group had experienced no further migraines at 6-month follow-up. The study reported no significant difference in the reduction in median MIDAS score (a measure of migraine-related disability on a scale of 0–21+; higher score indicates more severe disability) between the procedure and sham groups (from 36 to 17 vs 34 to 18 respectively), or mean migraine headache days (from 26 to 19 vs 30 to 21 days respectively) over 6 months.

2.3.3 A non-randomised comparative study of 86 patients reported a significant reduction in mean MIDAS score in the 40 patients treated by the procedure and the 46 patients treated by medical therapy at a mean follow-up of 29.2 months (from 35.8 to 8.3, p < 0.003 vs 22.6 to 19.1, p = 0.059 respectively; significance of between-group difference not stated).

2.3.4 The Specialist Advisers listed key efficacy outcomes as evidence of complete closure, and frequency and severity of migraine.

2.4 Safety

2.4.1 The following safety data were obtained from studies of PFO closure for a range of indications because:

  • safety data are likely to be similar for the various indications

  • the larger numbers of patients provide more robust evidence on safety than those from studies specifically relating to migraine.

2.4.2 Cardiac tamponade requiring surgery was reported in 2 patients in a non randomised comparative study of 280 patients: 1 occurred 5 weeks after the procedure because of left atrial laceration.

2.4.3 Late perforation of the aortic root by the device requiring pericardiocentesis and emergency cardiothoracic surgery occurred in 1 patient in a case report.

2.4.4 Device embolisation was reported in 0.6% (5/825) and 1% (2/167) of patients treated by the procedure in a case series of 825 patients and a non randomised comparative study of 280 patients respectively (device removed percutaneously in the first study but no further details given for the second).

2.4.5 Post- or peri-procedural arrhythmia was reported in 17% (8/47) and 10% (5/48) of patients in non-randomised comparative studies of 121 and 92 patients respectively.

2.4.6 The Specialist Advisers considered an additional theoretical adverse event to be valve dysfunction.

  • National Institute for Health and Care Excellence (NICE)