JACC: CARDIOVASCULAR INTERVENTIONS VOL. 11, NO.

11, 2018

ª 2018 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION

PUBLISHED BY ELSEVIER

Incidence and Outcomes of

Positive Bubble Contrast Study Results
After Transcatheter Closure of a
Patent Foramen Ovale
Ashish H. Shah, MD, MD-RESEARCH,a Mark Osten, MD,a Lee Benson, MD,b Sami Alnasser, MD,a Yvonne Bach, BSCH,a
Rohan Vishwanath, BSC,a Alex Van De Bruaene, MD, PHD,a Healey Shulman, BSC,a Jeneka Navaranjan, BSC,a
Rafique Khan, BHSC,a Eric Horlick, MDCMa

ABSTRACT

OBJECTIVES The aim of this study was to assess the incidence of persistently positive results on agitated saline
contrast injection after patent foramen ovale (PFO) closure, the underlying mechanism, and management.

BACKGROUND Transcatheter intervention to close a PFO is reasonable in highly selected patients younger than 60
years, after a thorough cardioneurological investigation following a cryptogenic stroke, particularly in the presence of
thromboembolic disease or in patients at high risk for venous thrombosis. The U.S. Food and Drug Administration
approved the Amplatzer PFO Occluder in October 2016 for such an indication. Confirmation of PFO closure is usually
verified by an agitated saline contrast injection during an echocardiographic examination. The appearance of bubbles in
the left atrium raises the concern of incomplete closure or other sources of shunting.

METHODS The medical records and echocardiograms of patients who were treated with transcatheter closure of a PFO
for cryptogenic stroke were reviewed.

RESULTS From January 1998 through December 2015, 880 patients were taken to the catheter laboratory for PFO
closure, of whom 568 patients, 320 men (56.3%), underwent transcatheter closure of a PFO using an Amplatzer PFO
Occluder, at a mean age of 48.1
12.9 years. The incidence of right-to-left shunting (RLS) was 19.5% at a mean of
4 months’ follow-up, which reduced to 8.4% at 11
2 months. Sources of RLS were identified in 10 (1.8%); pulmonary
arteriovenous malformation (n ¼ 4) was the most common etiology, followed by leak through the device (n ¼ 3). All
patients with additional sources of RLS were treated percutaneously. At 2-year follow-up, 16 patients (2.8%) persisted
with only mildly positive results on agitated saline contrast injection, without an apparent additional source of shunting.

CONCLUSIONS Coexistence of a PFO and an additional lesion responsible for RLS is uncommon, but not rare;
the majority are amenable to transcatheter or surgical intervention. (J Am Coll Cardiol Intv 2018;11:1095–104)
© 2018 by the American College of Cardiology Foundation.

P atent foramen ovale (PFO) is the most common

interatrial communication and is presumed
responsible for a proportion of cryptogenic
strokes through paradoxical embolization, especially
in young patients (1). Meta-analysis of randomized tri-
als has provided different conclusions with respect to
the value of PFO closure in this population (2). Recently
published results from the intention-to-treat analysis

From the aPeter Munk Cardiac Centre and Toronto Congenital Cardiac Centre for Adults, Toronto General Hospital, University
Health Network, Toronto, Ontario, Canada; and bThe Labatt Family Heart Centre, The Hospital for Sick Children, Division of
Cardiology, The University of Toronto School of Medicine, Toronto, Ontario, Canada. The structural heart disease program at the
Toronto General Hospital receives support for its educational and research missions from Abbott Vascular. Dr. Horlick is supported
by the Peter Munk Chair in Structural and Congenital Heart Disease Intervention. Dr. Horlick is a proctor and consultant for Abbott
Vascular. Dr. Osten is a consultant for Abbott Vascular. All other authors have reported that they have no relationships relevant to
the contents of this paper to disclose.

Manuscript received October 26, 2017; revised manuscript received February 5, 2018, accepted March 6, 2018.

ISSN 1936-8798/$36.00 https://doi.org/10.1016/j.jcin.2018.03.008

1096 Shah et al. JACC: CARDIOVASCULAR INTERVENTIONS VOL. 11, NO. 11, 2018

Bubble Study Post-Transcatheter PFO Closure JUNE 11, 2018:1095–104

ABBREVIATIONS of the RESPECT (Randomized Evaluation of the results of the cohort treated with the Amplatzer
AND ACRONYMS Recurrent Stroke Comparing PFO Closure to PFO Occluder. We sought to assess the incidence of
Established Current Standard of Care Treat- persistently positive ASCI after PFO closure, the un-
ASCI = agitated saline
contrast injection
ment) trial have shown that long-term derlying mechanism, and subsequent management.
follow-up in a highly selected group of pa-
CS = coronary sinus
tients with cryptogenic stroke, transcatheter
CT = computed tomography METHODS
closure (TC) of PFO reduced the rate of recur-
HHT = hereditary hemorrhagic
telangiectasia
rent cryptogenic but not all-cause stroke
PATIENTS. We reviewed our institutional database of
compared with medical therapy (3). The re-
PAVM = pulmonary patients who underwent transcatheter PFO closure,
arteriovenous malformation sults of the REDUCE trial, a controlled, open-
primarily for cryptogenic stroke, between January
PFO = patent foramen ovale label, randomized trial of 664 patients 18 to
1998 and December 2015. PFO closure to treat the
PSLVC = persistent left-sided
59 years of age with cryptogenic stroke
platypnea-orthodeoxia syndrome was not included.
superior vena cava randomly assigned to PFO closure with the
After the procedure, patients were reviewed in the
RLS = right-to-left shunting Gore septal occluder and antiplatelet therapy
outpatient clinic at 2 to 3 months and yearly and
TC = transcatheter closure versus antiplatelet therapy alone, met its pri-
investigated using transthoracic echocardiography
TEE = transesophageal
mary endpoint with a 76.6% reduction in
(TTE) with ASCI, both at rest and after a Valsalva
echocardiography stroke over an average of 3.4 years of follow-
maneuver. Results of the bubble study were defined
TTE = transthoracic up. That study also showed a reduction in
as mild, moderate, or strongly positive if 3 to 9, 10 to
echocardiography new brain infarcts by core laboratory interpre-
30, or >30 bubbles appeared in the left atrium, as
tation of baseline versus 2-year magnetic resonance
described previously (3). The shunt was thought to be
imaging from 11.3% in the control group to 5.7% in the
at the atrial level if bubbles appeared in the left
device arm (4). The CLOSE (Closure of Patent Foramen
atrium within 3 heartbeats after opacifying the right
Ovale, Oral Anticoagulants or Antiplatelet Therapy to
atrium, whereas bubbles appearing late were sug-
Prevent Stroke Recurrence) trial randomized 524 pa-
gestive of an extracardiac communication.
tients to oral anticoagulation, PFO closure, or anti-
platelet therapy. All patients had either atrial septal INVESTIGATIONS. A diagnosis of cryptogenic stroke
aneurysms or significantly positive results on bubble was made only after excluding known secondary
study. The study met its primary endpoint, demon- causes. Patients with cryptogenic stroke were
strating that PFO closure plus antiplatelet therapy reviewed by a neurologist and underwent magnetic
reduced the risk for stroke recurrence compared with resonance imaging, magnetic resonance angiography,
antiplatelet therapy alone. Anticoagulants did not computed tomography (CT), or computed tomo-
significantly reduce the risk for recurrent stroke graphic angiography of the brain if necessary; TTE or
compared with antiplatelet therapy (5). The U.S. Food transesophageal echocardiography (TEE); thrombo-
and Drug Administration approved the Amplatzer philia work-up; and ambulatory cardiac rhythm
PFO Occluder (St. Jude Medical, St. Paul, Minnesota) monitoring. Since evidence supporting up to 4 weeks
in October 2016. of cardiac monitoring has emerged, we have changed
our practice from 48 h to 2 to 4 weeks of Holter
SEE PAGE 1105
monitoring (4,6).
Atrial-level defects are not the sole source of par-
PROCEDURE. TC of PFOs was performed under fluo-
adoxical emboli. Pulmonary arteriovenous malfor-
roscopy, local anesthesia, and appropriate use of
mations (PAVMs), venous abnormalities including
intravenous sedation, as previously described (5).
persistent left-sided superior vena cava (PLSVC) to
Intracardiac echocardiography was available as a
the left atrium, or an unroofed coronary sinus (CS)
standby if issues arose during the procedure or if TEE
may also cause paradoxical embolization. Coexis-
suggested potential challenges (e.g., coexistent atrial
tence of a PFO and another anomaly, both of which
septal defect). One hundred units per kilogram of
may result in right-to-left shunting (RLS), is uncom-
intravenous unfractionated heparin was adminis-
mon, and prevalence is unknown. Successful closure
tered, adjusted to an activated coagulation time of
of a PFO is confirmed by agitated saline contrast in-
>250 s, and prophylactic antibiotics were adminis-
jection (ASCI) during an echocardiographic examina-
tered to each patient. Device positioning and ade-
tion in follow-up. The appearance of bubbles in the
quacy of closure was assessed using right atrial
left atrium raises the concern of incomplete closure or
angiography.
other sources of shunting. Our PFO closure experi-
ence involved multiple devices, some of which are of FOLLOW-UP. Patients were investigated using TTE
historical interest at this time; this report focuses on with ASCI at the time of follow-up, 2 to 3 months after
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 11, NO. 11, 2018
JUNE 11, 2018:1095–104
F I G U R E 1 Follow-Up After Transcatheter Closure of a Patent Foramen Ovale
LA ¼ left atrium; LSVC ¼ left superior vena cava; PAVM ¼ pulmonary arteriovenous malformation; PFO ¼ patent foramen ovale.
the procedure. Patients with persistent RLS were
further reviewed in the clinic at 6 to 12 months
post-procedure, and those who persisted with mod-
erate or strongly positive results on ASCI were
investigated in stepwise fashion using TEE and/or
noncontrast chest CT and/or were brought to the
catheterization laboratory and assessed using intra-
cardiac echocardiography and angiography to iden-
tify possible additional sources of RLS depending on
the era and individualized to patient factors. Those
identified with PAVMs or other congenital anomalies
were treated with device closure as appropriate.
RESULTS
In total, 880 patients were taken to the catheteriza-
tion laboratory for PFO closure. Of 570 patients
referred for PFO closure in the Amplatzer PFO
Occluder era, 568 underwent TC of PFOs using the
Amplatzer PFO Occluder, whereas 2 patients were
identified to have a large PAVM and a PLSVC draining
into the left atrium through an unroofed CS each, but
not a PFO. Both of these patients were treated using
Shah et al. 1097
Bubble Study Post-Transcatheter PFO Closure
transcatheter interventions. The mean age was
48.1
12.9 years, and 320 were men (56.3%). A 35-mm
device was used in those with aneurysmal septa,
whereas a 25-mm device was used in the remainder.
In 3 patients, an Amplatzer cribriform device was
used.
Patients’ clinical follow-up, after PFO closure, is
described in Figure 1. During the first visit at 4
2
months, 29 patients (5.1%) were lost to follow-up, and
residual RLS was noted in 105 of 539 patients (19.5%).
During subsequent follow-up at 11
2 months, 6
further patients were lost to follow-up, and 45 of 99
patients had residual RLS. Twenty-four patients with
at least moderately positive results on ASCI were
investigated (Figure 2), of whom 8 were identified to
have sources of RLS, whereas 2 other patients were
diagnosed with a PAVM and a PLSVC draining into the
left atrium through an unroofed CS but no PFO, as
described in Table 1. All patients with additional
sources of RLS were treated percutaneously. During
subsequent follow-up, further reduction in bubble
transmission was observed, or the results of ASCI were
reported as negative. At 21
3 months follow-up,
1098 Shah et al. JACC: CARDIOVASCULAR INTERVENTIONS VOL. 11, NO. 11, 2018

Bubble Study Post-Transcatheter PFO Closure JUNE 11, 2018:1095–104

diagnosed to have HHT after PFO closure. These

T A B L E 1 Follow-Up After Transcatheter Closure of a
Patent Foramen Ovale
Total patients considered for PFO closure
Total patients treated with PFO closure
Male
Age, yrs
Patients investigated for additional source of RLS
Identified sources of RLS
PAVM
Leak through the device
PLSVC with unroofed coronary sinus
Septal fenestrations (2–4 mm); normal RV size
Unidentified sources of RLS
Values are n, n (%), or mean
SD.
PAVM ¼ pulmonary arteriovenous malformation; PFO ¼ patent foramen ovale;
PLSVC ¼ persistent left-sided superior vena cava; RLS ¼ right-to-left shunting;
RV ¼ right ventricular.
residual RLS was observed in 16 of 45 patients, all of
which was only mild in severity.
Four patients were diagnosed with PAVMs, of whom
1 patient was known to have hereditary hemorrhagic
telangiectasia (HHT), whereas 1 additional patient was
PAVMs were treated using TC as described in Figure 3.
In 3 patients who were treated with the Amplatzer PFO
570
568
Occluder, residual leaks at the edge or through the
320 (56.3) device were noted; however, either the defect was not
48.1
12.9 considered large enough or it was not believed to be
24 (4.2) safe enough to warrant additional device deployment.
10 (1.8) Interestingly, during their latest clinic visits, ASCI re-
4
sults were negative in all patients. A patient with
3
PLSVC with an unroofed CS draining into the left
1
atrium was treated with transcatheter intervention
2
16 (2.8) using a 20-mm Amplatzer post-infarct ventricular
septal defect occluder (Figure 4). Two patients were
diagnosed with very small septal fenestrations (2 to 4
mm) that were not thought to pose an additional risk
for thromboembolism. Additionally, as there was no
evidence of right ventricular dilatation, no in-
terventions were offered. Individual patient details of
TC interventions to treat additional sources of RLS are
described in Table 2. The remaining 16 patients (2.8%)
had only mildly positive results on ASCI after PFO
closure; CT was used to investigate 11 of these, and
PAVM and other venous abnormalities were ruled out.

F I G U R E 2 Identifying a Source of Right-to-Left Shunting in the Catheterization Laboratory

Investigating a patient presenting with residual right-to-left shunting after patent foramen ovale closure in a catheterization laboratory
includes interrogating an interatrial septum (A), pulmonary angiography to identify micro or larger pulmonary arteriovenous malformation
(B,C), and innominate vein (D).
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 11, NO. 11, 2018 Shah et al. 1099
JUNE 11, 2018:1095–104 Bubble Study Post-Transcatheter PFO Closure

F I G U R E 3 Transcatheter Intervention to Treat Pulmonary Arteriovenous Malformation

Pulmonary angiography demonstrates arteriovenous malformation in the right middle lobe (A) that was selectively cannulated (B), and a
vascular plug was deployed (C) to successfully close the defect (D). Computed tomography of the thorax demonstrates the pulmonary
arteriovenous malformation (E) that is better delineated on 3-dimensional reconstruction. The red vessels are pulmonary arterial branches,
and the white vessel is a pulmonary vein draining to the left atrium.

DISCUSSION number and size of malformations (10). Nearly half of

TC of a PFO in a patient presenting with paradoxical
embolization is intended to achieve complete elim-
ination of RLS, as persistent residual RLS can
potentially result in recurrent stroke after PFO
closure (7). A common practice to assess the effec-
tiveness of such interventions is by TTE or trans-
cranial Doppler and ASCI in follow-up. As noted in
this cohort, observing the presence of a few bubbles
in the left atrium within a few months after TC of a
PFO is not uncommon (3–5,8), likely because of
incomplete endothelization of the device. Although
supporting evidence describing this endothelization
process is limited, we suggest this as a plausible
mechanism (9).
In patients with persistently positive moderate to
strongly positive results on ASCI, defining the mech-
anism for RLS, either a residual defect or an addi-
tional RLS lesion not identified earlier, is important
(7). PAVM is one such anomaly that has been re-
ported to be responsible for cryptogenic stroke, as the
prevalence of neurological events increases with the
the patients with PAVMs were also known to have
HHT, an autosomal-dominant condition (11), whereas
only 15% to 20% of patients with HHT were reported
to have PAVMs. There are also isolated cases of PAVM
unrelated to HHT (12). The size of these vascular
anomalies ranges from small connections to large
malformations, which may involve entire lobes, with
multiple feeders and draining vessels. As such, pa-
tients thought to have paradoxical embolization
should also be examined for the presence of muco-
cutaneous telangiectasia, a marker of HHT. PAVM can
be diagnosed with noncontrast CT of the chest or
pulmonary angiography (13).
The incidence of PAVM in the general population is
reported as 1:5,000 (14,15), whereas various other
studies in patients after PFO closure have reported
the rate to be 20% to 34%, on the basis of the finding
of persistent RLS, although these patients were not
formally investigated to confirm the diagnosis of
PAVM (8,16,17). As noted in 2 patients from our
cohort, 5 of 74 patients (6.8%) randomized to PFO
closure in the MIST (Migraine Intervention With
1100 Shah et al. JACC: CARDIOVASCULAR INTERVENTIONS VOL. 11, NO. 11, 2018

Bubble Study Post-Transcatheter PFO Closure JUNE 11, 2018:1095–104

F I G U R E 4 Transcatheter Intervention to Treat Persistent Left Superior Vena Cava

Angiography in persistent left superior vena cava (SVC) with unroofed coronary sinus demonstrates left SVC as well as ascending aorta (Ao) at
the same time (A). Balloon sizing to assess the compliance and sizing of the SVC, as well as measurement of pressure change subsequent to
the SVC occlusion (B). Check angiogram after deploying a 20-mm post-infarction muscular ventricular septal defect closure device in the left
SVC (C), and after the device release (D). Presence of right and left SVC is marked R and L, and bridging veins are denoted by the white
arrow (D).

STARFlex Technology) trial (18), and 1 patient in the
CLOSE study (5), had no PFO identified at the time of
the procedure, and PAVM was suspected to be
responsible for early RLS; however, no investigations
were performed to confirm the presence of PAVM. It
stands to reason that the incidence of PAVM may be
higher in a population with stroke and positive re-
sults on saline contrast study. However, the most
likely explanation for the discrepancy between sus-
pected and reported incidence of PAVM lies in the
ability of ASCI to distinguish between PAVM and
intracardiac shunting.
Bubbles appearing in the left atrium 3 or more
heartbeats after opacifying the right atrium are sug-
gestive of PAVM (extracardiac), whereas an earlier
appearance suggests the presence of intracardiac
shunting; however, there is an overlap in the time
delay before bubble appearance in the left atrium be-
tween these conditions (19). Physiological character-
istics of pulmonary capillaries, such as recruitment
and vasodilation, offer an explanation for the phe-
nomenon of RLS in the absence of PAVM. Pulsatile
pulmonary flow can recruit capillaries, and such an
effect can be observed within a single cardiac cycle that
can be maintained for few seconds afterward (20). Such
pulsatility can also be observed after a Valsalva ma-
neuver (21). One can conceptualize that it is feasible to
recruit and dilate pulmonary capillaries with repeated
Valsalva maneuvers, resulting in false-positive results
on ACSI by the mechanism described later.
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 11, NO. 11, 2018 Shah et al. 1101
JUNE 11, 2018:1095–104 Bubble Study Post-Transcatheter PFO Closure

T A B L E 2 Interventions to Treat Persistent Right-to-Left Shunting

Follow-Up
PFO PFO Mechanism Duration
Age, yrs/Sex Presentation Present Closed of RLS Intervention (months) Outcome

50/M Stroke Yes Vestigial PFO, PAVM PAVM coiled 6 Asymptomatic

not closed
55/F Stroke Yes 35-mm PAVM PAVM closed with 12 Asymptomatic; negative
Amplatzer 2 AVP-4 (8 mm) bubble results
PFO Occluder
33/F Stroke Yes 25-mm PAVM PAVM closed with AVP 72 Few bubbles appear in LA
Amplatzer after 5 beats, but
PFO Occluder no cause identified for RLS
42/M Stroke, HHT, Yes 25-mm PAVM PAVM coiled; PAPVC, 16 Asymptomatic; negative
PAPVC (RUPV Amplatzer left untreated bubble results
to SVC) PFO Occluder
38/M Stroke No NA PLSVC PLSVC closed with a 7 Asymptomatic, negative
to LA 20-mm Amplatzer bubble test
post- infarct VSD
device

AVP ¼ Amplatzer Vascular Plug; HHT ¼ hereditary hemorrhagic telangiectasia; LA ¼ left atrium; NA ¼ not applicable; PAPVC ¼ partial anomalous pulmonary venous
connection; RUPA ¼ right upper pulmonary artery; SVC ¼ superior vena cava; VSD ¼ ventricular septal defect; other abbreviations as in Table 1.

Although the actual size of saline contrast micro- supine position, by breathing a hypoxic air mixture or
bubbles is not known, they are suspected to be in the during exercise-induced hyperdynamic state
range of 60 to 90 m m in diameter (22), whereas the (24,26,27). Passage of agitated saline bubbles though
pulmonary capillaries measure 8 to 9 m m at rest (23). the pulmonary circulation during exercise but not at
Experiments using 25- to 50-m m microspheres injec- rest in the lungs of healthy volunteers suggests the
ted in healthy human donor lungs without known existence of recruitable arteriovenous shunts or
PAVM demonstrated the availability of arteriovenous anastomosis (27). In the presence of a PFO, ASCI is not
pathways >50 mm in two-thirds of the lungs (22), uniformly helpful in identifying the presence of
adequate enough to result in mildly positive results PAVM, unless agitated saline is injected in pulmonary
on ASCI (<10 bubbles). Normal human lungs are also arteries (downstream from a PFO) or with balloon
reported to have larger caliber pulmonary arteriove- occlusion of the PFO, and an operator actively as-
nous anastomoses that bypass pulmonary capillaries sesses individual pulmonary veins. Such a vigilant
(24). Different from PAVM, such anastomoses are approach also helps in identifying the potential
located in the lung apices (25), closed at rest and in location of a PAVM. Interestingly, in a study
the upright position, but can be recruited in the comparing diagnostic efficacy of ASCI in patients with
HHT versus control subjects, 7% of healthy subjects
without PFO or PAVM were noted to have mildly
T A B L E 3 Possible Etiologies That Can Result in positive results on bubble testing (28). It is also
Right-to-Left Shunting possible that such RLS can occur through a PAVM that
Interatrial-level Residual flow through the device may be difficult to detect by CT of the thorax (29). The
communication (incomplete endothelization)
Residual leak at the edge of the
presence of bubbles in the left atrium is not enough to
device Atrial septal defect confirm the diagnosis of PAVM, and special care
Fenestrated septum
should be taken with interpretation depending on
Pulmonary vasculature Large PAVM (evaluate for HHT)
Micro PAVM (consider congenital body positioning and the disease state of the lungs.
heart disease) Finally, transcranial Doppler has been reported to be
Other congenital conditions Persistent left superior vena cava
more sensitive at diagnosing RLS in comparison with
with unroofed coronary sinus
(bubble injection through the TEE with saline contrast study; however, interpreting
left hand)
the findings can be challenging, with similar limita-
Pulmonary artery–to–left
atrium connection tions described with TTE- or TEE-based ACSI.
Acquired conditions Hepatic vein–pulmonary vein collateral The location of the ACSI site may affect the de-
vessels (observed in patients with
congenital heart disease)
gree of shunting, as blood flow from the inferior
Esophageal vein–pulmonary vena cava is directed toward the fossa ovalis,
vein collateral vessels
whereas flow from the superior vena cava is
Abbreviations as in Tables 1 and 2. directed to the tricuspid valve. Agitated saline
injected through the cubital vein is less sensitive in
1102 Shah et al. JACC: CARDIOVASCULAR INTERVENTIONS VOL. 11, NO. 11, 2018

Bubble Study Post-Transcatheter PFO Closure JUNE 11, 2018:1095–104

F I G U R E 5 Algorithm to Manage Patients With Persistently Positive Results on Bubble Study After Patent Foramen Ovale Closure

ASD ¼ atrial septal defect; AVM ¼ arteriovenous malformation; CS ¼ coronary sinus; CT ¼ computed tomography; IAS ¼ interatrial septum;
IVC ¼ inferior vena cava; LA ¼ left atrium; PA ¼ pulmonary artery; PAVM ¼ pulmonary arteriovenous malformation; PFO ¼ patent foramen
ovale; PLSVC ¼ persistent left-sided superior vena cava; PV ¼ pulmonary vein; SVC ¼ superior vena cava; TEE ¼ transesophageal echocar-
diography; TTE ¼ transthoracic echocardiography.

comparison with an injection from a lower limb
vein to detect RLS (30). Additionally, an ACSI study
performed from the right antecubital vein is likely
to miss a PLSVC with an unroofed CS or direct
connection of the left superior vena cava to the left
atrium.
A PLSVC is the most common anomalous systemic
venous return anomaly (31). In the majority of pa-
tients, it drains into the right atrium through the CS,
but in the absence of an intact CS roof, it can
communicate with the left atrium (32). Its presen-
tation may range from asymptomatic to right heart
dilatation or even resting cyanosis; however, one
should also consider it as an alternative diagnosis in
patients presenting with paradoxical embolization.
TEE, cardiac CT, or magnetic resonance imaging can
confirm the diagnosis and help define the extent of
the defect and associated other congenital anoma-
lies. In the presence of a bridging vein, such an
anomaly can be treated using a TC procedure (33,34),
whereas in the absence of a bridging vein, balloon
occlusion of a left superior vena cava to the left
atrium may unmask adequate venovenous collateral
vessels to prevent a pressure rise in the venous
system upstream from a potential closure device
(Figure 4).
RLS due to a right pulmonary artery–to–left atrium
fistula is a rare anomaly (35). It may be associated
with an atrial septal defect, anomalies of the right
lung, and diverticulum of the right main bronchus
(36). A large shunt can result in heart failure in
childhood, whereas adult patients present with
cyanosis. Conventional angiography or CT is neces-
sary to confirm the diagnosis. In the majority of pa-
tients, a communication is located between the
posterior wall of the right pulmonary artery to the left
atrium and traverses between the right and left pul-
monary veins. Early diagnosis in infancy requires
surgical or TC repair (37).
Coexistence of a PFO and an additional condition
that can result in RLS is uncommon but not rare.
None of the published large prospective studies
evaluating efficacy of transcatheter intervention to
close PFO has reported these findings (18,38–40),
despite having residual shunt evident on ASCI after
PFO closure. It is imperative that treating physicians
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 11, NO. 11, 2018
JUNE 11, 2018:1095–104
be aware of these diagnoses (Table 3), as they are
associated with a potential risk for paradoxical
embolization, and the majority of such anomalies are
amenable to TC or surgical intervention. The pres-
ence of bubbles in the left atrium does not equate to
the presence of PAVM, and one should be thorough
and meticulous with subsequent investigations and
management plans. An algorithm to deal with such a
clinical situation is presented in Figure 5. Despite a
thorough search, it is possible to fail to identify the
mechanism leading to RLS in a small number of
patients.
STUDY LIMITATIONS. This was a nonrandomized,
retrospective, single-institutional
study with its inherent limitations. The reported
incidence of PFO and PAVM in our cohort is likely to
underestimate the true prevalence, as not all patients
with persistent RLS were investigated, and as our
institution is a supraregional referral center, many
patients we treated were referred from outlying
areas, and some did not return for follow-up to our
institution.
CONCLUSIONS
Each patient presenting with paradoxical emboliza-
tion or RLS, thought to be through a PFO, should also
be assessed for clinical features of HHT. Persistently
positive results on bubble testing after PFO closure
are not uncommon, but stratification of these patients
by the degree of positivity and allowing time for the
REFERENCES
1. Handke M, Harloff A, Olschewski M, Hetzel A,
Geibel A. Patent foramen ovale and cryptogenic
stroke in older patients. N Engl J Med 2007;357:
2262–8.
2. Udell JA, Opotowsky AR, Khairy P, et al. Patent
foramen ovale closure vs medical therapy for
stroke prevention: meta-analysis of randomized
trials and review of heterogeneity in meta-ana-
lyses. Can J Cardiol 2014;30:1216–24.
3. Saver JL, Carroll JD, Thaler DE, et al. Long-term
outcomes of patent foramen ovale closure or
medical therapy after stroke. N Engl J Med 2017;
377:1022–32.
4. Sondergaard L, Kasner SE, Rhodes JF, et al.
Patent foramen ovale closure or antiplatelet
therapy for cryptogenic stroke. N Engl J Med 2017;
377:1033–42.
5. Mas JL, Derumeaux G, Guillon B, et al. Patent
foramen ovale closure or anticoagulation vs.
antiplatelets after stroke. N Engl J Med 2017;377:
1011–21.
6. Albers GW, Bernstein RA, Brachmann J, et al.
Heart rhythm monitoring strategies for
Shah et al. 1103
Bubble Study Post-Transcatheter PFO Closure
device to endothelialize are reasonable consider-
ations. It is likely that among those patients with
persistently positive results on bubble studies,
important anatomic lesions may coexist and remain a
potential risk factor for recurrent paradoxical
embolization.
ACKNOWLEDGMENT The authors thank Dr. Madhu-
sudan Paravasthu, MBBS, clinical fellow in cardiotho-
racic imaging, Toronto General Hospital, for helping
with radiological image preparation.
ADDRESS FOR CORRESPONDENCE: Dr. Eric Horlick,
Toronto General Hospital, 200 Elizabeth Street, Room
observational 6E-249, Toronto, Ontario M5G 2C4, Canada. E-mail:
eric.horlick@uhn.ca.
PERSPECTIVES
WHAT IS KNOWN? PFO and PAVM are the most commonly
observed anomalies thought to be responsible for RLS that can
result in cryptogenic stroke due to paradoxical embolization.
WHAT IS NEW? For the first time, we have reported the inci-
dence rate of a coexisting second etiology that can also result in
RLS in patients who have undergone TC of a PFO. Majority of
these conditions can be treated with transcatheter interventions.
WHAT IS NEXT? Impact of residual leak after TC of a PFO on
recurrent neurological events should be evaluated in a prospec-
tive multicenter registry.
cryptogenic stroke: 2015 Diagnostics and Moni- and pulmonary arteriovenous fistula: survey of a
toring Stroke Focus Group report. J Am Heart large family. N Engl J Med 1959;261:625–36.
Assoc 2016;5:e002944.
12. Lucas RV Jr., Lund GW, Edwards JE. Direct
7. Windecker S, Wahl A, Chatterjee T, et al. communication of a pulmonary artery with the left
Percutaneous closure of patent foramen ovale in atrium. An unusual variant of pulmonary arterio-
patients with paradoxical embolism: long-term venous fistula. Circulation 1961;24:1409–14.
risk of recurrent thromboembolic events. Circula-
13. Faughnan ME, Palda VA, Garcia-Tsao G, et al.
tion 2000;101:893–8.
International guidelines for the diagnosis and
8. Harms V, Reisman M, Fuller CJ, et al. Outcomes management of hereditary haemorrhagic telangi-
after transcatheter closure of patent foramen ectasia. J Med Genet 2011;48:73–87.
ovale in patients with paradoxical embolism. Am J
14. Sloan RD, Cooley RN. Congenital pulmonary
Cardiol 2007;99:1312–5.
arteriovenous aneurysm. Am J Roentgenol Radium
9. Chen F, Zhao X, Zheng X, Chen S, Xu R, Ther Nucl Med 1953;70:183–210.
Qin Y. Incomplete endothelialization and late
15. Swanson KL, Prakash UB, Stanson AW. Pul-
dislocation after implantation of an Amplatzer
monary arteriovenous fistulas: Mayo Clinic expe-
septal occluder device. Circulation 2011;124:
rience 1982–1997. Mayo Clin Proc 1999;74:
e188–9.
671–80.
10. Moussouttas M, Fayad P, Rosenblatt M, et al.
16. Jesurum JT, Fuller CJ, Renz J, Krabill KA,
Pulmonary arteriovenous malformations: cerebral
Spencer MP, Reisman M. Diagnosis of secondary
ischemia and neurologic manifestations.
source of right-to-left shunt with balloon occlu-
Neurology 2000;55:959–64.
sion of patent foramen ovale and power M-mode
11. Hodgson CH, Burchell HB, Good CA, transcranial Doppler. J Am Coll Cardiol Intv 2009;
Clagett OT. Hereditary hemorrhagic telangiectasia 2:561–7.
1104 Shah et al.
Bubble Study Post-Transcatheter PFO Closure
17. Anzola GP, Morandi E, Casilli F, Onorato E.
Does transcatheter closure of patent foramen
ovale really “shut the door?” A prospective study
with transcranial Doppler. Stroke 2004;35:
2140–4.
18. Dowson A, Mullen MJ, Peatfield R, et al.
Migraine Intervention With STARFlex Technology
(MIST) trial: a prospective, multicenter, double-
blind, sham-controlled trial to evaluate the
effectiveness of patent foramen ovale closure
with STARFlex septal repair implant to resolve
refractory migraine headache. Circulation 2008;
117:1397–404.
19. Zukotynski K, Chan RP, Chow CM, Cohen JH,
Faughnan ME. Contrast echocardiography grading
predicts pulmonary arteriovenous malformations
on CT. Chest 2007;132:18–23.
20. Presson RG Jr., Baumgartner WA Jr.,
Peterson AJ, Glenny RW, Wagner WW Jr. Pulmo-
nary capillaries are recruited during pulsatile flow.
J Appl Physiol 2002;92:1183–90.
21. Mc IH, Burnum JF, Hickam JB, Warren JV. Cir-
culatory changes produced by the Valsalva ma-
neuver in normal subjects, patients with mitral
stenosis, and autonomic nervous system alter-
ations. Circulation 1954;9:511–20.
22. Lovering AT, Stickland MK, Kelso AJ,
Eldridge MW. Direct demonstration of 25- and 50-
microm arteriovenous pathways in healthy human
and baboon lungs. Am J Physiol Heart Circ Physiol
2007;292:H1777–81.
23. La Gerche A, MacIsaac AI, Burns AT, et al.
Pulmonary transit of agitated contrast is associ-
ated with enhanced pulmonary vascular reserve
and right ventricular function during exercise.
J Appl Physiol 2010;109:1307–17.
24. Lovering AT, Elliott JE, Beasley KM, Laurie SS.
Pulmonary pathways and mechanisms regulating
transpulmonary shunting into the general circu-
lation: an update. Injury 2010;41 suppl 2:S16–23.
JACC: CARDIOVASCULAR INTERVENTIONS VOL. 11, NO. 11, 2018
25. Tobin CE, Zariquiey MO. Arteriovenous shunts
in the human lung. Proc Soc Exp Biol Med 1950;75:
827–9.
26. Stickland MK, Welsh RC, Haykowsky MJ, et al.
Intra-pulmonary shunt and pulmonary gas ex-
change during exercise in humans. J Physiol 2004;
561:321–9.
27. Eldridge MW, Dempsey JA, Haverkamp HC,
Lovering AT, Hokanson JS. Exercise-induced
intrapulmonary arteriovenous shunting in healthy
humans. J Appl Physiol 2004;97:797–805.
28. Gazzaniga P, Buscarini E, Leandro G, et al.
Contrast echocardiography for pulmonary arte-
riovenous malformations screening: does any
bubble matter? Eur J Echocardiogr 2009;10:
513–8.
29. Nawaz A, Litt HI, Stavropoulos SW, et al.
Digital subtraction pulmonary arteriography
versus multidetector CT in the detection of pul-
monary arteriovenous malformations. J Vasc
Interv Radiol 2008;19:1582–8.
30. Gin KG, Huckell VF, Pollick C. Femoral vein
delivery of contrast medium enhances trans-
thoracic echocardiographic detection of patent
foramen ovale. J Am Coll Cardiol 1993;22:
1994–2000.
31. Campbell M, Deuchar DC. The left-sided su-
perior vena cava. Br Heart J 1954;16:423–39.
32. Attenhofer Jost CH, Connolly HM,
Danielson GK, Dearani JA, Warnes CA, Jamil
Tajik A. Clinical features and surgical outcome in
25 patients with fenestrations of the coronary si-
nus. Cardiol Young 2007;17:592–600.
33. Hutyra M, Skala T, Sanak D, Novotny J,
Kocher M, Taborsky M. Persistent left superior
vena cava connected through the left upper pul-
monary vein to the left atrium: an unusual
pathway for paradoxical embolization and a rare
cause of recurrent transient ischaemic attack. Eur J
Echocardiogr 2010;11:E35.
JUNE 11, 2018:1095–104
34. Leventhal AR, Shah AH, Crean AM, Osten M,
Horlick E, Benson L. Percutaneous correction of
right superior vena cava to left atrium. J Am Coll
Cardiol Intv 2015;8:e221–2.
35. Friedlich A, Bing RJ, Blount SG Jr. Physiolog-
ical studies in congenital heart disease; circulatory
dynamics in the anomalies of venous return to the
heart including pulmonary arteriovenous fistula.
Bull Johns Hopkins Hosp 1950;86:20–57.
36. Kroeker EJ, Adams HD, Leon AS, Pouget JM.
Congenital communication between a pulmonary
artery and the left atrium. Physiologic observa-
tions and review of the literature. Am J Med 1963;
34:721–5.
37. Zeebregts CJ, Nijveld A, Lam J, van Oort AM,
Lacquet LK. Surgical treatment of a fistula between
the right pulmonary artery and the left atrium:
presentation of two cases and review of literature.
Eur J Cardiothorac Surg 1997;11:1056–61.
38. Agarwal S, Bajaj NS, Kumbhani DJ, Tuzcu EM,
Kapadia SR. Meta-analysis of transcatheter closure
versus medical therapy for patent foramen ovale
in prevention of recurrent neurological events af-
ter presumed paradoxical embolism. J Am Coll
Cardiol Intv 2012;5:777–89.
39. Carroll JD, Saver JL, Thaler DE, et al. Closure
of patent foramen ovale versus medical therapy
after cryptogenic stroke. N Engl J Med 2013;368:
1092–100.
40. Pineda AM, Nascimento FO, Yang SC,
Kirtane AJ, Sommer RJ, Beohar N. A meta-analysis
of transcatheter closure of patent foramen ovale
versus medical therapy for prevention of recurrent
thromboembolic events in patients with crypto-
genic cerebrovascular events. Catheter Cardiovasc
Interv 2013;82:968–75.
KEY WORDS cryptogenic stroke, patent
foramen ovale, pulmonary arteriovenous
malformation, right-to-left shunt