What you need to know
The recommendations apply to patients under 60 years old with patent foramen ovale
(PFO) who have had a cryptogenic ischaemic stroke, when extensive workup for other
aetiologies of stroke is negative
For patients who are open to all options, we make a weak recommendation for PFO closure
plus antiplatelet therapy rather than anticoagulant therapy
For patients in whom anticoagulation is contraindicated or declined, we make a strong
recommendation for PFO closure plus antiplatelet therapy versus antiplatelet therapy
alone
For patients in whom closure is contraindicated or declined, we make a weak recommendation
for anticoagulant therapy rather than antiplatelet therapy.
Further research may alter the recommendations that involve anticoagulant therapy
Options for the secondary prevention of stroke in patients younger than 60 years who
have had a cryptogenic ischaemic stroke thought to be secondary to patent foramen
ovale (PFO) include PFO closure (with antiplatelet therapy), antiplatelet therapy
alone, or anticoagulants. International guidance and practice differ on which option
is preferable.
The BMJ Rapid Recommendations panel used a linked systematic review1 triggered by
three large randomised trials published in September 2017 that suggested PFO closure
might reduce the risk of ischaemic stroke more than alternatives.2
3
4 The panel felt that the studies, when considered in the context of the full body
of evidence, might change current clinical practice.5 The linked systematic review
finds that PFO closure prevents recurrent stroke relative to antiplatelet therapy,
but possibly not relative to anticoagulants, and is associated with procedural complications
and persistent atrial fibrillation.1 The review also presents evidence regarding the
role of anticoagulants or antiplatelet therapy when PFO closure is not acceptable
or is contraindicated.
This expert panel make a
Strong recommendation in favour of PFO closure plus antiplatelet therapy compared
with antiplatelet therapy alone
Weak recommendation in favour of PFO closure plus antiplatelet therapy compared with
anticoagulants
Weak recommendation in favour of anticoagulants compared with antiplatelet therapy.
The largest challenge in making our recommendation was the low quality evidence for
the comparisons that included anticoagulants. We summarised all the highest quality
available evidence separately for antiplatelet therapy and anticoagulants because
the evidence suggests it is likely their effectiveness and adverse effects differ,
and clinicians and patients should be aware of these likely differences. Our panel
believes that the mechanism of benefit with PFO closure is prevention of venous clots
crossing the PFO. Anticoagulants are likely to be substantially more effective in
preventing such clots from initially arising than antiplatelet agents.
Box 1 shows the articles and linked evidence in this Rapid Recommendation package.
The main infographic presents the recommendations as three paired comparisons, together
with an overview of the absolute benefits and harms informing each recommendation,
according to the GRADE methodology.
Box 1
Linked resources for this BMJ Rapid Recommendations cluster
Kuijpers T, Spencer FA, Siemieniuk RAC, et al. Patent foramen ovale closure, antiplatelet
therapy or anticoagulation therapy alone for management of cryptogenic stroke? A clinical
practice guideline. BMJ 2018;362:k2515
Summary of the results from the Rapid Recommendation process
Mir H, Siemieniuk R, Ge L, et al. Percutaneous closure plus antiplatelet therapy versus
antiplatelet or anticoagulation therapy alone in patients with patent foramen ovale
and cryptogenic stroke: a systematic review and network meta-analysis incorporating
complementary external evidence. BMJ Open 2018;0:e023761. doi:10.1136/bmjopen-2018-023761
Review and network meta-analysis of all available randomised trials that assessed
PFO closure as adjunct treatment to antiplatelet versus antiplatelet therapy or anticoagulation,
and comparing anticoagulants to antiplatelet therapy
MAGICapp (https://app.magicapp.org/app#/guideline/2191)
Expanded version of the results with multilayered recommendations, evidence summaries,
and decision aids for use on all devices
Current practice
Management options for those with patent foramen ovale (PFO) and cryptogenic stroke
Typically, patients with cryptogenic stroke and PFO have three treatment options to
reduce the risk of future stroke:
(a) Closure of the PFO with subsequent antiplatelet therapy that may be continued
indefinitely or discontinued some months after PFO closure
(b) Antiplatelet therapy alone
(c) Anticoagulant therapy alone.
Most current guidelines recommend against routine closure of the PFO in patients with
cryptogenic stroke and instead recommend antiplatelets or anticoagulation (the latter
if indicated for another reason) (box 2).6
7
8
9
Box 2
Current guidance for closure of patent foramen ovale (PFO) in patients with PFO and
cryptogenic stroke
American Academy of Neurology 20176
PFO v medical therapy alone—Clinicians must counsel patients considering percutaneous
PFO closure that having a PFO is common in the general population; it is impossible
to determine with certainty whether their PFO caused their stroke or transient ischaemic
attack; the effectiveness of the procedure for reducing stroke risk remains uncertain;
and the procedure is associated with relatively uncommon, yet potentially serious,
complications
Anticoagulation v antiplatelet—In the absence of another indication for anticoagulation,
clinicians may routinely offer antiplatelet drugs instead of anticoagulation to patients
with cryptogenic stroke and PFO
American Heart Association/American Stroke Association7
For patients with an ischaemic stroke or transient ischaemic attack and a PFO who
are not undergoing anticoagulation therapy, antiplatelet therapy is recommended
For patients with an ischaemic stroke or transient ischaemic attack and both a PFO
and a venous source of embolism, anticoagulation is indicated depending on stroke
characteristics. When anticoagulation is contraindicated, an inferior vena cava filter
is reasonable
For patients with a cryptogenic ischaemic stroke or transient ischaemic attack and
a PFO without evidence for deep vein thrombosis, available data do not support a benefit
for PFO closure
In the setting of PFO and deep vein thrombosis, PFO closure by a transcatheter device
might be considered depending on the risk of recurrent deep vein thrombosis
NICE 20138
Evidence on the safety of percutaneous closure of PFO to prevent recurrent cerebral
embolic events shows serious but infrequent complications. Evidence on its efficacy
is adequate. Therefore, this procedure may be used with normal arrangements for clinical
governance, consent, and audit
Netherlands Society of Cardiology 20169
Closure of a PFO is not beneficial in unselected patients with transient ischaemic
attack or cryptogenic stroke
Closure of a PFO should be considered in patients with transient ischaemic attack
or cryptogenic stroke and a Risk of Paradoxical Embolism (RoPE) score >8 and at least
one clinical risk factor
Identification of cryptogenic stroke
In about a third of patients in the general population who are diagnosed with an acute
ischaemic stroke, investigation finds no clear cause; it is cryptogenic.10 Clinicians
reach the diagnosis by ruling out alternative reasons for stroke through prolonged
rhythm monitoring to exclude atrial fibrillation; transoesophageal echocardiography
or alternative imaging of the aorta and left atrial appendage to rule out aortic atherothrombosis
or left atrial clot; and carotid ultrasonography, computed tomography, or magnetic
resonance imaging to rule out cerebrovascular disease.
Patients diagnosed with cryptogenic stroke are less likely to have classic risk factors
for atheroembolic stroke such as older age, hypertension, hyperlipidaemia, and diabetes.11
They are more likely to have a PFO than patients in the general population.12
Implications of a patent foramen ovale (PFO)
The presence of a PFO does not result in an identifiable increased risk of stroke
in the general population.13
14
15 Many meta-analyses have addressed whether closure of a PFO reduces the long term
risk of subsequent stroke,12
16
17
18 but most have concluded that there is insufficient evidence.6
PFO is a communication between the right and left atrium, typically diagnosed by transthoracic
echocardiography with observed flow between the left to right atrium by colour Doppler
ultrasonography.19 If the shunt direction reverses, this communication may allow a
venous thrombus or right atrial thrombus to travel directly into the arterial circulation
and cause a stroke—a phenomenon known as a paradoxical embolism.20
21 This can be characterised with echocardiography (box 3).
Box 3
Details of echocardiographic diagnosis, risk profile, and patent foramen ovale (PFO)
procedure planning
Which route—Transesophageal echocardiography has a higher sensitivity for detection
of a PFO compared with transthoracic imaging and is recommended in younger adults
with unexplained cerebrovascular events
Work-up of cryptogenic stroke—In addition to detection of PFO, rarer causes of embolic
events include an atrial septal defect, cardiac tumours (such as myxoma or papillary
fibroelastoma), bacterial or non-bacterial valve vegetations, and atrial thrombi
Detection of PFO—Microbubbles enter the right atrium, and, if a PFO is present, they
pass into the left atrium within a few beats of appearance in the right atrium. Although
shunting usually is predominantly left to right, there is some right to left shunting
as the relative pressures in the two chambers change during the cardiac cycle and
with respiration
Sensitivity of saline contrast for detection of a PFO is increased by asking the patient
to perform a Valsalva manoeuvre, which transiently increases right atrial pressure
Estimating the size of a PFO based on the amount of contrast seen in the left atrium
may be unreliable22
Those with PFO at greater risk—An atrial septal aneurysm, defined as excessive bulging
of atrial septal fossa ovalis, is often associated with septal fenestrations and may
be a marker of increased embolic risk
Ahead of planned PFO closure—Transeophageal echocardiography is recommended for more
detailed visualisation of the atrial septal anatomy when PFO closure is planned22
How the recommendation was created
Our international panel included general internists, interventional and non-interventional
cardiologists, stroke physicians, epidemiologists, methodologists, statisticians,
and people with personal experience of cryptogenic stroke and patent foramen ovale
(PFO). They decided on the scope of the recommendation and the outcomes that are most
important to patients. The panel identified eight patient-important outcomes needed
to inform the recommendation: non-fatal ischaemic stroke, death, major bleeding, pulmonary
embolism, serious procedure related or device related adverse events, atrial fibrillation,
transient ischaemic attack, and systemic embolism.
A parallel team conducted a systematic review addressing the benefits and harms of
three patient-relevant clinical questions framed by the panel: (a) PFO closure with
subsequent antiplatelet therapy versus antiplatelet therapy alone, (b) PFO closure
with subsequent antiplatelet therapy versus anticoagulation, and (c) anticoagulation
versus antiplatelet therapy.1
Because of a lack of evidence in those with PFO, particularly for the anticoagulation
option, the panel asked for a summary of the indirect evidence regarding prevention
of thrombosis from trials of venous thromboembolism and atrial fibrillation.
We also performed a systematic search for evidence regarding patients’ values and
preferences (see appendix 1 on bmj.com).
No panel member had financial conflicts of interest; intellectual and professional
conflicts were minimised and managed (for full summary see appendix 2 on bmj.com).
The panel followed the BMJ Rapid Recommendations procedures for creating a trustworthy
recommendation,5
28 including using the GRADE approach to critically appraise the evidence and create
recommendations (see appendix 3 on bmj.com).29 The panel considered the balance of
benefits, harms, and burdens of the procedure, the quality of the evidence for each
outcome, typical and expected variations in patient values and preferences, and acceptability.30
Recommendations can be strong or weak, for or against a course of action.
The evidence
The linked systematic review reports the relative and the absolute effects of PFO
closure followed by antiplatelet therapy versus antiplatelet therapy alone or versus
anticoagulation and the effect of anticoagulation versus antiplatelet therapy in patients
with cryptogenic stroke and PFO.1
Figure 2 provides an overview of the number and types of patients included, the study
funding, and patient involvement.
Fig 2
Characteristics of patients and trials included in systematic review of the effects
of percutaneous closure followed by antiplatelet therapy versus antiplatelet or anticoagulation
therapy alone in patients with patent foramen ovale (PFO) and cryptogenic stroke.
Evidence used from 6 randomised clinical trials2
3
4
31
32
33 (plus 2 further trials for comparison of antiplatelets and anticoagulants34
35)
We conducted a network meta-analysis combining direct evidence (from studies of management
in people with cryptogenic stroke comparing at least two of the three options) with
indirect evidence (inferring benefits and harms of two alternatives through relative
effects on a third option) to obtain more informative estimates of effect. The paucity
of data regarding anticoagulation for this intervention resulted in a sparsely populated
network with low certainty evidence. The estimates of relative effect of PFO closure
versus anticoagulation were extremely imprecise. Only 353 patients were randomised
to PFO closure versus anticoagulation, and 405 patients to anticoagulation versus
antiplatelet agents, and events were infrequent. Therefore, to obtain more precise
estimates, we performed additional analyses based on indirect evidence.
The systematic review also reports indirect evidence, from participants who did not
have PFO and cryptogenic stroke, but venous thromboembolism.23 This evidence was used
to inform the effects of anticoagulation versus on stroke. Similarly, for the outcome
of major bleeding, we performed additional analyses based on indirect evidence comparing
anticoagulation with antiplatelet therapy for several non-PFO associated indications.1
Specific groups of PFO patients with cryptogenic stroke
We hypothesised that studies including more patients with larger shunt sizes, and
those that included more patients treated with anticoagulants, would demonstrate larger
effects. A separate systematic review24 reported that PFO closure, compared with any
medical therapy, was more effective in patients with moderate or large size shunts.
However, the same clinical trials that included more patients with larger shunts also
included fewer patients who were prescribed anticoagulants in the medical therapy
arm; this confounding makes it impossible to sort out which association (if either)
was responsible for the larger effect. Therefore, the shunt size subgroup effect has
low credibility (for more details see the linked systematic review).1
We were unable to stratify our analyses and recommendations by type or generation
of PFO closure device because of the limitations in published data and small subset
sample sizes.
Procedure or device related adverse events
Procedure or device related adverse events included vascular complications (1%), conduction
abnormalities (1%), device dislocation (0.7%), and device thrombosis (0.5%). Less
serious adverse events such as minor bleeding and supraventricular tachycardia were
inconsistently reported; the panel judged them as important, however, and took them
into account in making recommendations.
Values and preferences
No studies had relevant information on values and preferences. We screened 455 titles
and abstracts, and six full text articles. Appendix 1 on bmj.com presents our systematic
review of the limited evidence. Three people with experience of living with cryptogenic
stroke and PFO provided input regarding the choice of outcomes.
Understanding the recommendations
Absolute benefits and harms
The panel considered PFO closure plus antiplatelets better than antiplatelet agents
alone. This is a strong recommendation because the absolute differences and patient
preferences were aligned to place a high value on stroke prevention. Patients are
likely to find an absolute reduction of stroke with PFO closure of 8.7% at five years
very important. Although 3.6% will experience an adverse event, such events, including
1.8% increase in atrial fibrillation, do not usually result in long term disability
and so were considered less important.
The possible small reduction in stroke and decreased bleeding risk with PFO closure
versus anticoagulants alone mandated a weak recommendation for PFO closure.
For those patients who need or want to avoid PFO, the panel judged anticoagulation
the best alternative, although the evidence regarding stroke reduction was of low
certainty. The risk of major bleeding probably increased with anticoagulation. Although
direct anticoagulants have not been evaluated in PFO, their advantages in terms of
convenience may render them, rather than warfarin, the best option for those who choose
anticoagulants.
The main infographic explains the recommendations and provides an overview (GRADE
summary of findings) of the absolute benefits (reduction in recurrent ischaemic stroke)
and harms of:
PFO closure followed by antiplatelet therapy versus antiplatelet therapy alone
PFO closure followed by antiplatelet therapy versus anticoagulants alone
Anticoagulants versus antiplatelet therapy.
Estimates of baseline risk for effects come from the control arm of the trials, using
the median estimate of risk where available.1
The panel agreed that, compared with antiplatelet therapy alone, PFO closure followed
by antiplatelet therapy:
Probably has a large decrease in ischaemic stroke (8.7% absolute risk reduction, moderate
quality evidence) over five years
Has a risk of device or procedure related adverse events (3.6% absolute risk, high
quality evidence) at one year
Probably has an increase in persistent atrial fibrillation or flutter (1.8% absolute
risk increase, moderate quality evidence) and transient atrial fibrillation or flutter
(1.2% absolute risk increase, moderate quality evidence) at one year
Probably has little or no difference in death, major bleeding, pulmonary embolism,
transient ischaemic attack, or systemic embolism (moderate to high quality evidence)
at five years.
The panel agreed that, compared with anticoagulation, PFO closure followed by antiplatelet
therapy:
May result in little or no difference in ischaemic stroke (1.6% absolute risk reduction,
low quality evidence) at five years
Probably decreases major bleeding (2.0% absolute risk reduction, moderate quality
evidence) at five years
Has a risk of device or procedure related adverse events (3.6% absolute risk, high
quality evidence) at one year
Probably has an increase in persistent atrial fibrillation or flutter (1.8% absolute
risk increase, moderate quality evidence) and transient atrial fibrillation or flutter
(1.2% absolute risk increase, moderate quality evidence) at one year
Probably has little or no difference in death, pulmonary embolism, transient ischaemic
attack, or systemic embolism (moderate quality evidence) at five years.
The panel agreed that anticoagulation versus antiplatelet therapy at five years’ duration:
May decrease ischaemic stroke (7.1% absolute risk reduction over 5 years, low quality
evidence)
Probably increases major bleeding (1.2% absolute risk increase over 5 years, moderate
quality evidence)
Probably has little or no difference in death, pulmonary embolism, transient ischaemic
attack, or systemic embolism (moderate quality evidence).
Values and preferences
PFO closure followed by antiplatelet therapy versus antiplatelet therapy alone
Patients for whom anticoagulation is unacceptable or contraindicated should consider
PFO closure. Our strong recommendation for PFO closure for such patients reflects
the high value they place on avoiding recurrent ischaemic stroke. Patients are likely
to find absolute reduction of stroke with PFO closure of 8.7% in five years important.
Although 3.6% experience serious device or procedure related adverse events, these
do not usually result in long term disability, and so we considered them less important.
Persistent atrial fibrillation after PFO closure procedure might be a concern; however,
the main adverse consequence of atrial fibrillation is increased risk of stroke, which
was already shown to be substantially lower in patients randomised to PFO closure.
PFO closure followed by antiplatelet therapy versus anticoagulation
The major downsides of PFO closure are the 3.6% incidence of complications from the
procedure and the probable 1.8% absolute increase in persistent atrial fibrillation.
The major downside of anticoagulation is the probable 2.0% absolute increase in bleeding
risk over five years. Other issues to consider are the burden and costs of long term
anticoagulation. Our weak recommendation for PFO closure reflects (in addition to
the low certainty in the estimates of effect) that most serious complications of PFO
closure are usually short term, whereas anticoagulation imposes a long term burden
and increased risk of major bleeding. Most fully informed patients would probably
accept the transient risk of major adverse events rather than the long term bleeding
risk, but a substantial minority would probably choose anticoagulation.
Anticoagulation versus antiplatelet therapy
Patients to whom PFO closure is unacceptable or contraindicated have to choose between
anticoagulant or antiplatelet therapy. A typical patient places a high value in a
possible absolute reduction of stroke with anticoagulation of 7.1% over five years
and would therefore place higher value on the possible benefit of stroke reduction
than the probable increased risk of major bleeding. A systematic review25 and a primary
study26 of values and preferences on thromboprophylaxis treatment of patients with
atrial fibrillation showed that, though preferences were highly variable, most patients
value preventing strokes considerably more than they are concerned about increased
risk of bleeding. However, there is substantial uncertainty in our estimates for stroke
reduction—how this uncertainty would influence decisions is likely to vary substantially.
Therefore, we issue a weak recommendation for anticoagulation. Both options need to
be discussed with the patient, ideally in a process of shared decision making.
Practical issues and other considerations
Figure 3 outlines the key practical issues for patients and clinicians discussing
PFO closure and is based on the content expertise of the panel members; practical
issues are also accessible, along with the evidence, as decision aids to support shared
decision making in MAGICapp. Antiplatelet therapy or anticoagulation are typically
given as an oral medication once or twice a day.
Fig 3
Practical issues about use of percutaneous closure followed by antiplatelet therapy
versus antiplatelet or anticoagulation therapy alone in patients with patent foramen
ovale (PFO) and cryptogenic stroke
Costs and resources
The panel focused on the patient’s perspective rather than that of society when formulating
the recommendation. Because PFO closure is associated with higher costs related to
the procedure, implementation of this recommendation is likely to have an important
impact on the costs for health funders in the short term. Over the long term, however,
PFO closure may reduce costs as a result of reduced stroke rates and reduction in
associated costs.27 Addressing this issue formally would require a cost effectiveness
analysis.
Uncertainties to be addressed in future research
The key remaining research question is the relative merit of PFO closure versus anticoagulation
alone. It may also be appropriate to conduct further trials of PFO closure versus
antiplatelet agents alone in those with small PFOs. Longer trials are also needed
to address the longevity of the PFO closure device and ongoing need for monitoring
of device performance.
Key research questions to inform decision makers and future guidelines include:
What are the benefits and harms of PFO closure versus anticoagulants (including direct
oral anticoagulants) in patients with PFO and cryptogenic stroke?
What patient groups are more likely to benefit from PFO closure versus medical therapy?
(That is, explore whether the effect of PFO closure versus medical therapy varies
with shunt size, presence of atrial septal aneurysm, and age.)
Which device for PFO closure is best?
What is the longevity of the PFO closure device and ongoing need for monitoring of
device performance?
Updates to this article
Table 1 shows evidence which has emerged since the publication of this article. As
new evidence is published, a group will assess the new evidence and make a judgment
on to what extent it is expected to alter the recommendation.
Table 1
New evidence which has emerged after initial publication
Date
New evidence
Citation
Findings
Implications for recommendation(s)
There are currently no updates to the article.
Education into practice
Does this article offer you new ways to approach advising patients with cryptogenic
ischaemic stroke presumed to be related to a patent foramen ovale (PFO)?
How might you better respect differences in patients’ preferences, particularly their
perspective regarding the bleeding risk associated with long term anticoagulation
or their feelings about undergoing an invasive procedure?
What information could you share with your patients to help them reach a decision?
How might you share this information with colleagues to learn together?
How patients were involved in the creation of this article.
The panel included three people with personal experience of cryptogenic stroke and
patent foramen ovale (PFO). These panel members identified important outcomes, and
led the discussion on values and preferences. The patients agreed that, in general,
small reductions in risk of ischaemic stroke are more important to them than small
increases in risk of atrial fibrillation or of device or procedure related adverse
events. We expect these values to be shared by most patients for ischaemic stroke.
The patients participated as full panel members in the teleconferences and email discussions
and met all authorship criteria. They had equal input as any other author on the recommendation.