SCAI Guidelines for the Management of Patent Foramen Ovale
Authors, Journal, Affiliations, Type, DOI
- Authors: Clifford J. Kavinsky (Rush University, chair), Molly Szerlip (Baylor Scott & White, vice-chair), Andrew M. Goldsweig (Nebraska), Zahid Amin (AdventHealth), Konstantinos Dean Boudoulas (Ohio State), John D. Carroll (Colorado), Megan Coylewright (Erlanger), Sammy Elmariah (Massachusetts General), Lee A. MacDonald (South Denver Cardiology), Atman P. Shah (Chicago), Christian Spies (Sutter Health), Jonathan M. Tobis (UCLA), Steven R. Messé (Penn, American Academy of Neurology representative), with Emily Senerth, Yngve Falck-Ytter, Ifeoluwa Babatunde and Rebecca L. Morgan (McMaster) providing GRADE methodological support through the Evidence Foundation
- Journal: Journal of the Society for Cardiovascular Angiography & Interventions. 2022;1:100039; published online 19 May 2022; open access (CC BY-NC-ND)
- Type: Clinical practice guideline using GRADE, built on original systematic reviews; IOM and Guidelines International Network standards; three patient representatives on the panel; 13 recommendations across five PICO questions
- DOI: https://doi.org/10.1016/j.jscai.2022.100039
- Companion documents: SCAI technical review (JSCAI 2022;1:100040) and the 2019 SCAI expert consensus on operator and institutional requirements for PFO closure
Overview
The interventional cardiology answer to the same question the neurologists answered, and it differs in one important respect: SCAI goes where the trials did not. Five of its thirteen recommendations concern patients with no prior PFO-associated stroke — migraine, decompression illness in divers, platypnea-orthodeoxia, thrombophilia, atrial septal aneurysm, systemic embolism, TIA, DVT — scenarios that were excluded from every RCT and that therefore had no guidance at all before this document.
The panel's one strong recommendation is the same as everyone else's: PFO closure rather than antiplatelet therapy alone in patients aged 18–60 with prior PFO-associated stroke (moderate certainty), and — unlike ESO 2024 and unlike the trial subgroups — independent of anatomy, because the data on subgroups are too thin to withhold a recommendation from anyone in that age band.
The document's most useful function may be its safety arithmetic. Its own summary of the evidence reports AF risk ratios of 7.33 versus antiplatelet therapy and 18.1 versus anticoagulation, both driven by events inside 45 days, against a stroke relative risk of 0.26 — and it then argues, reasonably, that the balance still favours closure in the population the trials enrolled.
Keywords
Patent foramen ovale · PFO closure · cryptogenic stroke · migraine · decompression illness · platypnea-orthodeoxia · thrombophilia
Key Takeaways
The recommendation set, in one table
| # | Population | Recommendation | Strength | Certainty |
|---|---|---|---|---|
| 1.1 | Migraine, no prior PFO stroke | Against routine closure | Conditional | Moderate |
| 1.2 | SCUBA divers with prior DCI, no prior PFO stroke | Against routine closure | Conditional | Very low |
| 1.3 | Platypnea-orthodeoxia, no prior PFO stroke | For closure | Conditional | Very low |
| 1.4 | Thrombophilia, no prior PFO stroke | Against closure in addition to antithrombotic therapy | Conditional | Very low |
| 1.5 | Atrial septal aneurysm, no prior PFO stroke | Against closure | Conditional | Very low |
| 1.6 | Systemic embolism, no prior PFO stroke, other causes excluded | For closure rather than medical therapy | Conditional | Very low |
| 1.7 | TIA, no prior PFO stroke | Against closure | Conditional | Very low |
| 1.8 | DVT, no prior PFO stroke | Against closure | Conditional | Very low |
| 2.1 | Prior PFO-associated stroke, age 18–60 | For closure rather than antiplatelet therapy alone | Strong | Moderate |
| 2.2 | Prior PFO-associated stroke, age ≥60 | For closure rather than long-term antiplatelet therapy | Conditional | Very low |
| 2.3 | History of AF with ischaemic stroke | Against routine closure | Conditional | Very low |
| 2.4 | Thrombophilia on antiplatelet therapy, prior PFO stroke | For closure rather than antiplatelet alone | Conditional | Very low |
| 3.1 | Prior PFO-associated stroke, age 18–60, no other anticoagulation indication | For closure plus antiplatelet therapy rather than anticoagulation alone | Conditional | Low |
| 3.2 | Same, age ≥60 | Same | Conditional | Very low |
| 4.1–4.3 | Thrombophilia / DVT / PE requiring lifelong anticoagulation, plus PFO stroke | For closure in addition to lifelong anticoagulation | Conditional | Very low |
| 5 | Post-closure DAPT beyond 1 month | No recommendation — knowledge gap | — | — |
Why 2.1 is strong and independent of anatomy
- Four RCTs compared closure with antiplatelet therapy: recurrent stroke RR 0.26 (0.12–0.58); DVT also lower.
- AF was higher after closure — RR 7.33 (2.41–22.25) — but 86% of AF events occurred within 45 days. Device- or procedure-related serious adverse events occurred in 4.5%.
- The panel's judgement: a transient AF risk is offset by a durable reduction in recurrent stroke, so the recommendation does not depend on shunt size, tunnel length or the presence of an aneurysm — the subgroup data are too sparse to justify withholding closure from the lower-risk anatomy.
- RoPE ≥7 may identify greater benefit (a pooled analysis of CLOSURE I, RESPECT and PC: HR 0.31 for RoPE ≥7 versus 0.82 for RoPE <7), but this is offered as a remark, not an entry criterion.
- Disaggregated subgroup data: atrial septal aneurysm, 0 events in 81 closed versus 9 in 74 on antiplatelet therapy (RR 0.05, 0.00–0.87); RoPE ≥7 RR 0.04 (0.01–0.71) versus RoPE <7 RR 0.15 (0.01–2.66).
Closure versus anticoagulation (3.1)
- Three RCTs: no reduction in recurrent stroke with closure versus anticoagulation (RR 0.94, 0.37–2.42).
- But closure had less major bleeding (RR 0.24, 0.06–0.91) and more AF (RR 18.1, 1.07–305.04; 86% within 45 days).
- Hence a conditional recommendation for closure plus antiplatelet therapy over anticoagulation — a mixed benefit profile rather than a clear efficacy win.
What the non-stroke recommendations actually rest on
| Scenario | Evidence | Effect |
|---|---|---|
| Migraine | MIST, PRIMA, PREMIUM | Attacks/month MD 0.59 (0.15–1.03); days/month MD 1.33 (0.33–2.32); none met its primary endpoint. Migraine cessation RR 3.46 (0.65–18.40). Post-hoc pooled PRIMA + PREMIUM (Amplatzer): −3.1 vs −1.9 days, P=0.02; cessation 9% vs 0.7%, P=0.001 |
| Diving / DCI | 3 observational studies | Recurrent DCI RR 0.31 (0.08–1.13) — non-significant. DCI risk ~5× higher in divers with a PFO |
| Platypnea-orthodeoxia | 3 observational studies | Oxygen saturation +14.21% (12.18–16.25); 76% improved quality of life; but AF 9.6%, mortality 8.7%, other complications 8.7% in the closure cohort |
| Thrombophilia | 2 observational studies | Recurrent stroke RR 0.17 (0.07–0.44) |
| Systemic embolism | No direct study — 3 patients in a subgroup analysis | Mechanistic extrapolation from stroke |
| TIA without prior stroke | No direct study — 4/65 recurred | Cannot be distinguished reliably from complex migraine |
| DVT without prior stroke | No direct study — 0/10 recurred | — |
- Migraine with aura is associated with PFO in 48% (Anzola), and aura patients may respond better than those without — but no trial has demonstrated that convincingly.
Safety
- A survey of 13,736 implants found surgery was required in only 0.2% — the panel's counter to the charge that closure is an invasive procedure in a young population.
- Post-procedural AF incidence 3.7 per 100 patient-years, concentrated in the first 45 days, and increasing with age. The proposed mechanism is local irritation and inflammation of the atrial septum by the newly placed occluder, resolving with endothelialisation.
- The panel comments that "oral anti-platelet and anti-coagulant therapies are associated with significant bleeding events" — a framing, not a trial result.
Knowledge gaps the panel names
- Duration of DAPT beyond one month after closure — no recommendation possible.
- Time from stroke to closure — no recommendation; trial enrolment occurred within a few months of the index event.
- Patients with AF and a PFO who have had a stroke — the panel recommends against routine closure despite acknowledging that some may have had a PFO-associated event, on the grounds that occult AF is the likelier mechanism and closure adds AF risk.
- PFO closure in the paediatric population is explicitly outside scope and discouraged.
Limitations of the document
- Eleven of thirteen recommendations are conditional, and eight rest on very low certainty evidence — mostly single observational studies, several with fewer than 15 exposed patients (three patients with systemic embolism, ten with DVT, eight with PE).
- The panel extrapolates where it says it should not. Recommendation 1.6 (systemic embolism) is supported by a subgroup of three patients, and the document itself concedes "randomized studies to demonstrate the prevention of paradoxical systemic embolism with PFO closure are not feasible because of the rarity of these events."
- The strong recommendation is anatomy-independent on the grounds that subgroup data are limited — but the same document elsewhere reports that PFO size and septal aneurysm morphology predict recurrence, and DEFENSE-PFO was built entirely on selecting by anatomy. Both cannot be right; the panel chooses the trial population over the mechanism.
- Recruitment of the evidence base is heavily interventional. The panel is drawn from interventional cardiology societies with one neurology representative; the guideline is funded by SCAI, and its Discussion argues against critics of the procedure by name. The GRADE process and three patient representatives mitigate but do not eliminate this.
- No long-term data beyond 5 years on devices, and no head-to-head comparison of FDA-approved devices — the panel states outright that observational data and meta-analyses "do not provide consistent conclusions favouring one over another."
- The migraine evidence is internally contradictory. Three RCTs failed their primary endpoints, and the panel nonetheless reports a post-hoc pooled analysis showing significant benefit on secondary endpoints — a signal that could equally reflect chance in a small pooled set.
- Diving advice is offered without a trial, and the recommended alternative (limit high-risk diving) is a behavioural change with no comparative outcome data either.
- Cost, access and equity are not addressed in a guideline recommending a device procedure for a population with a 2% two-year absolute risk reduction.
Key Concepts Mentioned
concepts/Patent-Foramen-Ovale · concepts/PFO-Closure-Selection · concepts/PFO-Detection-and-Anatomy · concepts/PFO-Non-Stroke-Indications · concepts/Sports-Cardiology-SDM
Key Entities Mentioned
entities/Ischemic-Stroke · entities/Atrial-Fibrillation · entities/Pulmonary-Embolism · concepts/Atrial-Septal-Defect
Named but unlinked: MIST · PRIMA · PREMIUM · NAVIGATE ESUS · RoPE · Amplatzer PFO Occluder · Gore Helex · Gore Cardioform · NobleStitch · platypnea-orthodeoxia syndrome · decompression illness
Wiki Pages Updated
- sources/pfo-jscai-2022 — created
- concepts/Patent-Foramen-Ovale — created
- concepts/PFO-Non-Stroke-Indications — created
- concepts/PFO-Closure-Selection — created
- concepts/PFO-Detection-and-Anatomy — created
- concepts/Sports-Cardiology-SDM — diving and PFO
- concepts/Atrial-Septal-Defect — PFO versus ASD
- wiki/wikiindex · wiki/sourceindex