oxygen delivery is thereby reduced. The pathophysiology
associated with each congenital shunt is reviewed in more
With normal cardiac anatomy, lung function, and hemoglobin
levels, arterial blood oxygen contents vary only to the extent
that pulmonary alveolar oxygenation changes. Under most
physiological conditions, the blood oxygen content changes
little and is more than adequate to supply the needs of the
tissues. With a right-to-left shunt, however, deoxygenated
systemic venous blood returns directly to the systemic arterial
circulation. The oxygen content of the systemic arterial blood
falls in proportion to the volume of systemic venous blood
mixing with the normal pulmonary venous return. With
reduced oxygen content, even with normal cardiac output,
tissue oxygen delivery falls and the work capacity of the
muscles is limited.6
Quantifying Shunt Volumes
The ratio of total pulmonary blood flow to total systemic blood
flow, the Qp/Qs ratio, is a useful tool for quantifying the net
shunt. A Qp/Qs ratio of 1:1 is normal and usually indicates that
there is no shunting. A Qp/Qs ratio of Ͼ1:1 indicates that
pulmonary flow exceeds systemic flow and defines a net
left-to-right shunt. Similarly, a Qp/Qs ratio of Ͻ1:1 indicates
a net right-to-left shunt. Both left-to-right and right-to-left
(bidirectional) shunting may be present in the same patient. If
the left-to-right shunt equals the right-to-left shunt in magni-
tude, it is possible to have a Qp/Qs of exactly 1:1.
Atrial Septal Defect
The formation of the atrial septum is a complex process,
consisting of the growth and partial reabsorption of 2 tissue
membranes, septum primum and septum secundum; the
fusion of these membranes to the forming endocardial cush-
ions; and the reabsorption of the fetal sinus venosus into the
structure that will ultimately become the RA. In Ϸ4 of
100 000 newborns,7 an error in this developmental process
will result in a defect in the wall separating the 2 atria, an
atrial septal defect (ASD). There are a number of types of
ASD (Figure 1), including the ostium primum defect (a result
of the deficiency of endocardial cushion tissue), the ostium
secundum defect (a result of excessive reabsorption of sep-
tum primum), and the sinus venosus defect (resulting from an
error in the incorporation of the sinus venosus chamber into
the RA).8 Although the following discussion of ASD patho-
physiology is true for all types of ASD, the sinus venosus
ASD also may be associated with anomalous pulmonary
venous return and the ostium primum ASD with significant
atrioventricular (AV) valve abnormalities. These additional
features may complicate the physiology further and are
beyond the scope of this review.
The pathophysiology of an ASD is complex and multifacto-
rial.9 Flow across the defect occurs in both systole and
diastole. In most patients, flow is predominantly left to right,
but transient right-to-left shunts are common, particularly
with isometric strain. The bulk of the shunt flow occurs
during diastole. In this phase, blood in each atrium has 2
alternative pathways: following the normal route through the
AV valve to the ventricle on that side or passing through the
ASD to fill the opposite ventricle. The direction of flow
across the ASD during diastole is determined by the instan-
taneous differences in the compliance and the capacity of the
Ventricular chamber compliance is determined to a large
extent by afterload. (Other factors such as intravascular
volume status, myocardial muscle mass, chamber geometry,
coronary perfusion, and pericardial and intrathoracic pres-
sures also contribute to the intrinsic distensibility of the
chamber.) In an otherwise normal patient, the left ventricle
(LV), pumping to the systemic circulation, faces a substan-
tially larger workload than the right ventricle (RV), pumping
to the lungs. The LV becomes physiologically hypertrophied,
reflecting its level of work, whereas the RV myocardium
remains thin. The thick-walled LV will stretch/distend to
accept additional volume less readily than the thinner RV. As
a result, in the usual ASD patient, the difference in chamber
compliance favors a left-to-right shunt because the blood in
the LA finds it easier to fill the more compliant RV. In
patients with increased RV afterload resulting from congen-
ital obstructions in the pulmonary arteries or veins or with
high pulmonary vascular resistance resulting from pulmonary
parenchymal disease or primary pulmonary hypertension, the
RV will be hypertrophied and less compliant. Left-to-right
shunting at the ASD may be minimal, reflecting little overall
difference between the 2 ventricles. With more severe RV
noncompliance or distensibility, flow across the ASD may be
predominantly from right to left. In some patients with
reduced RV compliance, the RV may be able to handle a
Figure 1. Types and typical locations of ASDs. In this view, the
free walls of the RA and RV have been removed, looking left-
ward toward the septal surface. MPA indicates main pulmonary
artery; 1, sinus venosus ASD at the junction of the superior vena
cava (SVC) and RA; 2, ostium secundum ASD; and 3, ostium
primum/AV canal type ASD.
Sommer et al Shunts in Congenital Heart Disease 1091
normal cardiac output at rest but will not readily accept
additional flow (when cardiac output increases), and a right-
to-left shunt may occur only with exertion. Other factors such
as a prior RV myocardial infarction, extrinsic compression of
the RV, or any associated congenital abnormality that results
in RV or tricuspid valve hypoplasia may reduce the effective
RV capacity and impede its filling.
When the AV valves are closed, ventricular compliance no
longer affects blood flow across the defect at the atrial level.
Several factors determine flow volume and direction in
systole. As in diastole, the size of the defect is a critical
determinant in the volume but not the direction of flow. The
atria may have differential capacities or compliances them-
selves, which may affect flow direction during systole. In
patients with a large left-to-right shunt in diastole, pulmonary
venous return will exceed systemic venous return (because
QpϾQs). Volumes in the 2 atria tend to equilibrate during
systole, resulting in a left-to-right flow during that phase of
the cardiac cycle. AV valve regurgitation also may affect the
direction of ASD flow during systole. Significant mitral or
tricuspid regurgitation may impede flow across an ASD,
increasing or decreasing shunting during systole.
Finally, the size of the ASD itself helps to determine the
volume of shunting. If the ASD is large, the defect creates
little or no resistance to flow. Blood flow across the defect in
diastole is determined entirely by the relative properties of the
ventricles as above. With a smaller, restrictive defect, blood
flow is limited by the resistance of the ASD itself, no matter
how large the difference in ventricular compliance.
In a patient with an ASD, shunt direction and magnitude are
variable and age dependent. In fetal life, RV noncompliance,
a result of high pulmonary vascular resistance, allows nearly
unidirectional right-to-left flow at the atrial level. Immedi-
ately after birth, with RV compliance comparable to that of
the LV, there may be little net shunting through an ASD.
Over several months, with the physiological fall in pulmonary
vascular resistance, the RV thins, compliance falls, and the
typical left-to-right shunt develops in children and young
As a result of normal physiological changes associated
with aging, the LV myocardium tends to become more
hypertrophied and less compliant.10 With similarly sized
ASDs, therefore, adults tend to have larger shunts as they age.
It is part of the reason why children are rarely symptomatic
but patients in their fourth or fifth decade may begin to
develop the symptoms frequently associated with ASD. There
are 4 common clinical presentations of ASD in the adult
Most frequently, adult patients complain of progressive
shortness of breath with exertion. Studies have shown a
reduction in maximum oxygen consumption in the unrepaired
ASD population because of the inherent inefficiency of a
continuously preload-reduced LV in combination with a
volume overload in the pulmonary circulation. After repair of
the ASD, exercise capacity improves within days to
Atrial arrhythmia, resulting from stretching of the conduc-
tion system, may be the first presenting sign of an ASD. An
adult who presents with atrial arrhythmia at a young age
should be evaluated for dilatation of the right-side cardiac
chambers and evidence of an atrial level shunt. Prevention of
long-term atrial fibrillation is one of the reasons for repairing
ASD in young asymptomatic patients,13 although the subse-
quent development of atrial fibrillation may depend more on
the patient’s age at intervention and may occur despite
surgery in patients Ͼ25 years of age.14 Although both RA and
RV volumes are reduced acutely with ASD repair and both
chambers return to normal dimensions in children, there
appears to be persistent RA enlargement when the ASD is
closed in adult patients.15 This may explain an ongoing,
increased risk of atrial fibrillation after adult ASD repairs
compared with patients who underwent closure at a younger
Rarely, a patient with ASD will present with stroke or other
systemic ischemic event caused by paradoxical embolization
of thrombus through the defect17 similar to the patent foramen
ovale (PFO; see below). Although most patients have a
significant net left-to-right shunt through the defect, virtually
all have transient flow reversal with the Valsalva maneuver or
other isometric strain.
A small number of adult patients also may be identified
echocardiographically when a heart murmur or unrelated
cardiac symptoms in the absence of exercise, rhythm, or
embolic symptoms bring them to a physician’s attention.
Pulmonary hypertension is uncommon with ASD, even in
patients with large defects, in part because of the large
capacitance of the pulmonary bed. Natural history studies
dating to the presurgical and pre-echocardiography eras
suggested an incidence of Ϸ15% in the ASD population.18
The observations that pulmonary vascular disease may de-
velop in patients with a tiny ASD and that it is absent in the
vast majority of patients with large ASDs suggest that the
ASD may be an associated marker of pulmonary hyperten-
sion but not necessarily causative. More recent reviews
suggest a rate of 6% to 9%.19,20 Once a patient has reached
adulthood with normal PA pressures, the natural history is
established: They no longer develop significant pulmonary
hypertension related to the shunt, but they may have pressure
elevation, like any other patient, as a result of the develop-
ment of pulmonary parenchymal disease, left-sided heart
dysfunction, or obstructive sleep apnea. It would be fair to
say that the overall risk of and specific risk factors for
developing pulmonary vascular disease with an ASD remain
When severe pulmonary hypertension from any cause
results in RV systolic failure, high RV end-systolic volumes
impede filling from the RA. With an intact atrial septum,
there is systemic venous stasis and symptoms of “classic”
right-heart congestive failure (anasarca and low cardiac
output) because the LV can pump out only what it receives
back from the lungs. In a patient with pulmonary hyperten-
sion and an ASD, however, the defect allows decompression
of the right heart via a right-to-left shunt. The systemic
venous blood does not need to traverse the lungs to reach the
LV. It may cross through the ASD, mixing with the pulmo-
1092 Circulation February 26, 2008
nary venous return in the LA, to augment LV preload. These
patients are cyanosed from the right-to-left shunt and show
minimal response to supplemental oxygen.21 However, tissue
oxygen delivery is often better than in the patient without an
ASD because the detrimental reduction in blood oxygen
content is far outweighed by the maintenance of a normal or
nearly normal cardiac output. For this reason, it is likely that
repair of adult patients with ASD and moderate or severe
pulmonary hypertension may not improve survival.16 Simi-
larly, in patients with end-stage primary pulmonary hyperten-
sion, the creation of an ASD has been demonstrated to be of
benefit in prolonging life and as a bridge to lung
Indications for Intervention
Significant questions remain about which ASDs should be
closed because the natural history of the disease remains
elusive. Generally accepted indications for closure of ASD
include patients with ASD and echocardiographic evidence of
right-sided cardiac volume loading, patients with ASD who
are symptomatic (principally exercise related), patients with
ASD with exercise-related cyanosis (without pulmonary hy-
pertension), patients with ASD who have suffered an episode
of paradoxical thromboembolization, and potentially as pro-
phylaxis in patients who require a noncardiac procedure with
high risk of paradoxical embolization (ie, joint replacement
therapy). A traditional open heart approach, dating back to
1954,23 and newer minimally invasive and robotic tech-
niques24 are all available for surgical repair of the defects.
The success and safety of transcatheter techniques have
significantly increased the number of percutaneous closure
As above, pulmonary hypertension, with right-to-left
shunting at rest (or a pulmonary vascular resistance of Ͼ14
Woods units), is the principal contraindication to ASD
closure.20 However, patients may present with intermediate
degrees of pulmonary hypertension with bidirectional or
predominantly left-to-right shunts. It may be difficult with the
limited data available in the literature to decide which of
these patients should be offered closure. We have used
temporary balloon occlusion of the ASD in the catheteriza-
tion laboratory as our primary determinant. With occlusion of
the defect and elimination of the right-to-left shunt, one can
directly assess the acute impact on RA and pulmonary artery
pressures and on cardiac output.
Similarly, patients may present with signs of left-sided
congestive heart failure with a left-to-right shunt at the atrial
level. Typically occurring in the elderly, this is related to
changes in LV compliance (ie, status after myocardial infarc-
tion). In this scenario, the stiffer the LV chamber becomes,
the larger the left-to-right shunt is. The shunt acts as a
“popoff” for the LA, allowing the patient to maintain man-
ageable pulmonary venous pressures, but also reduces LV
preload and cardiac output. With closure of such a defect, the
full impact of the LV diastolic dysfunction is felt (because all
pulmonary venous return is forced into the LV), resulting in
an acute rise in LA pressure and the potential development of
pulmonary edema. Temporary balloon occlusion in the cath-
eterization laboratory, with simultaneous LA pressure mea-
surement, can be used to decide which patient will tolerate
Patent Foramen Ovale
PFO, another communication in the atrial septum, is a
remnant of the normal fetal circulation. Anatomically, the
foramen ovale comprises overlapping portions of septum
primum and septum secundum, acting as a 1-way flap valve
allowing continuous right-to-left flow during fetal life. Be-
cause all venous blood (both deoxygenated systemic and the
high-oxygen umbilical venous return) drains to the RA
throughout fetal life, right-to-left flow is critical for perfusion
and growth of the left heart. Postnatally, septum primum
fuses to septum secundum, completing septation of the atria.8
However, in 20% to 25%, incomplete fusion leads to the
persistence of the flap valve, leaving a PFO. Because it is
present in all newborns, a PFO technically is not a “congen-
ital” defect. But as the most common “hole in the heart” and
currently the most common catheter intervention for struc-
tural heart defects,27 the PFO is deserving of mention in this
Immediately after birth, with the acute increase in pulmonary
blood flow, LA pressure rises to exceed RA pressure, pushing
septum primum rightward, against septum secundum, shut-
ting the flap of the PFO. When RA pressure rises intermit-
tently with Valsalva or other isometric strain, the leaflets of
the PFO may separate, with leftward excursion of septum
primum, allowing flow from RA to LA, as was the norm
throughout fetal life.
The degree to which the leaflets separate, the frequency
with which the flap valve opens, and the amount of blood that
crosses from right to left depend on a number of variables,
including RV compliance, the difference in RA and LA
pressure, distortion of the anatomy (ie, dilated aortic root,
elevation of right hemidiaphragm), and most important, the
tissue characteristics of septum primum. With a relatively
rigid septum primum, right-to-left flow may occur only with
strain. At the other extreme, the septum primum may be
extremely thin and mobile, the so-called atrial septal aneu-
rysm, and may open spontaneously with the changing phases
of the cardiac or respiratory cycle. If septum secundum is less
sturdy or in the case of LA hypertension, septum secundum
may be pushed rightward and left-to-right shunting can be
seen, the functional equivalent of a prolapsing or regurgitant
valve. It is common for such a PFO to be labeled an ASD,
although with clear overlap of septum primum and septum
secundum, it is preferentially called a PFO.
In general, patients with PFO are never identified because
they have no symptoms. Paradoxical embolization of throm-
botic material is the most frequent clinical presentation.28
Although thromboembolic events to noncritical systemic
structures may go unnoticed, cerebral embolization may
produce stroke or transient ischemic attack,29 and coronary
embolization may result in myocardial infarction in the
absence of atherosclerotic disease.30 Splenic, hepatic, renal,
Sommer et al Shunts in Congenital Heart Disease 1093
and retinal infarcts have all been seen in our practice in
patients with no risk factors other than PFO. Therefore, when
no other source of thromboembolism can be detected, a PFO
may be suspected.31
A number of other clinical syndromes are being studied for
their apparent relationship to PFO, including migraine head-
ache,32,33 hypoxemia,34 and decompression illness in divers
and in people who work at altitude,35 as well as high-altitude
pulmonary edema.36 The mechanism of the migraine relation-
ship to PFO remains entirely unknown. Hypoxemia may
occur in a patient with PFO, like in ASD, when RV
compliance or capacitance is diminished and hemodynamics
favor a right-to-left flow across the PFO. This physiology has
been described for patients with obstructive sleep apnea,37
after pulmonary embolic events,38 and with RV infarction.39
In addition, significant hypoxemia can occur in the presence
of normal right-heart pressures when anatomic distortion of
the atrial septum accentuates right-to-left shunt at the atrial
level, the platypnea orthodeoxia syndrome.40 These patients
experience an acute fall in arterial oxygen saturation on
changing from a supine to an upright position. Decompres-
sion illness is presumed secondary to right-to-left air embo-
lization through the PFO in divers or people who work at
Indication for Intervention
Despite the rapid growth, relative ease, safety, and efficacy of
transcatheter devices,41–44 particularly compared with surgi-
cal closure, there are currently no approved indications for
transcatheter PFO closure in the United States. Each patient
with PFO should be considered individually. Treatment
depends on the presenting symptom. Because the vast major-
ity of patients with isolated PFO never have symptoms, there
is no rationale for prophylactic closure in asymptomatic
patients at this time.
For patients who have had paradoxical embolization,
treatment options include antithrombotic agents, entrance in a
randomized trial comparing blood thinners and closure,
“off-label” transcatheter closure with a non-PFO device, or
surgical repair. Because prospective, randomized, controlled
trials are not yet completed, there is no definitive answer as to
which therapy provides the best long-term prophylaxis from
recurrent stroke. Similarly, there are few prospective data to
date to support the closure of PFO in migraine sufferers.
Multiple studies addressing this question are underway in the
United States and Europe.
In contrast, patients with the platypnea orthodeoxia syn-
drome should have their PFO closed in all cases. Closure
results in immediate elimination of right-to-left shunting and
normalization of oxygen saturation. Recurrent decompression
illness in divers may be preventable with PFO closure,
although few data exist at this point.
Ventricular Septal Defect
Ventricular septal defect (VSD) is the most common form of
congenital heart defect in children, accounting for Ϸ20% of
human cardiac malformations. This is undoubtedly based on
the complexity of the embryological development of the
ventricular septum, which involves the fusion of multiple
distinct septal components.45 The membranous septum, the
site at which all of these components fuse, sits behind the
septal leaflet of the tricuspid valve and immediately below
the aortic valve in the LV outflow tract. The perimembranous
type of VSD, which accounts for Ϸ80% of all VSDs, occurs
in this location. Defects of the muscular septum are the next
most frequent, are the result of excessive fetal muscular
resorption, can be single or multiple, and can be located
anywhere in the muscular septum. Endocardial-cushion–type
VSDs (associated frequently with primum ASD as part of a
complete AV canal defect), malalignment defects, and other
defects of the outlet septum are less common (Figure 2).
Endocardial-cushion VSDs often are associated with AV
valve insufficiency, and outflow defects often are associated
with aortic insufficiency, which further complicate the phys-
iological consequences of the lesion.
The hemodynamic effects of a VSD are very different than
those of an ASD. With a VSD, blood in each ventricle has 2
possible systolic pathways: through the usual outflow tract of
that ventricle or through the VSD to the outflow tract of the
other ventricle. It is the comparative resistance of each
pathway that determines both the direction and volume of
systolic flow across the VSD according to Ohm’s law.45 For
example, with normal pulmonary vascular resistance and a
large, nonrestrictive VSD, the sum of resistors from the LV to
the pulmonary artery is very low compared with the resis-
tance of flow to the systemic circulation, resulting in a large
left-to-right systolic flow across the defect. If the VSD is very
small, there is high resistance at the defect itself, limiting the
left-to-right shunt, even with low pulmonary resistance. With
elevation of pulmonary vascular resistance and a large defect,
the sum of the resistors may approximate the aortic resis-
Figure 2. Types and typical locations of VSDs. In this view, the
free walls of the RA and RV have been removed, looking left-
ward toward the septal surface. SVC indicates superior vena
cava; 1, perimembranous VSD; 2, muscular VSD; 3, inlet/AV
canal type VSD; and 4, subpulmonary VSD.
1094 Circulation February 26, 2008
tance, and net shunting will be minimal. With pulmonary
resistance higher than systemic resistance, the shunt will be
right to left, no matter what size the defect is.
A left-to-right shunt at the ventricular level reduces LV
output by the amount of the shunt. However, the body’s
compensatory mechanisms will increase intravascular vol-
ume in response to a lower cardiac output until LV end-dia-
stolic volume is sufficient to pump both a normal cardiac
output and the proportionate left-to-right shunt. This results in
significant LV volume overload. If, for example, 50% of the
blood is shunted across the VSD to the lungs, the LV volume
at end diastole must be 200% of normal to maintain a normal
cardiac output. High LV volumes elevate LA filling pressures
and may cause pulmonary venous congestion at rest or during
periods of exertion.
A secondary effect of the large VSD is the transmission of
LV pressure to the pulmonary vascular bed. With a large
enough defect, the LV and RV become a functionally
common chamber. With no RV outflow obstruction, the
pressure in the pulmonary artery in systole equals the pres-
sure in the aorta in systole. This pulmonary artery pressure
elevation distinguishes a shunt at the ventricular level from a
shunt at the atrial level. Although the volumes may be the
same, the ASD does not generate the same pressure head/
shear forces in the pulmonary bed. The combination of the
volume and pressure contributes to the development of
pulmonary vascular disease or Eisenmenger’s syndrome.46
Compared with the volume that crosses a VSD in systole,
there is little flow in diastole. Diastolic flow across the VSD
is much like that across the ASD in systole. Relative chamber
volumes and compliance, as well as differences in ventricular
contractile and relaxation properties,47,48 determine instanta-
neous flow direction. Flow outside of systole does not
contribute significantly to the net shunt volume.
Natural History: Small VSDs
Patients with VSD present initially in infancy. With a small
VSD and normally falling pulmonary vascular resistance, the
patient develops a heart murmur in the first week as left-to-
right flow develops. But because the shunt is small, these
babies remain asymptomatic, and they grow and develop
normally. With time, the septal leaflet of the tricuspid valve
may become adherent to the ventricular septal surface,
spontaneously closing the perimembranous VSD. Similarly, a
muscular VSD may close spontaneously with progressive
compression of the defect by the growing myocardial tissue.
Once closed, these defects usually present no further medical
issue for the patient. Endocardial-cushion defects and outlet
defects do not close without intervention.
For adults with small persistent VSDs that have not closed
who have nondilated LV and normal pulmonary artery
pressure, there are only 2 ongoing risks. Infective endocardi-
tis (any type of VSD) may be prevented with antibiotic
prophylaxis for dental procedures.49 Progressive aortic insuf-
ficiency also may develop with either perimembranous or
subpulmonary VSD as a result of aortic cusp prolapse into the
VSD.50 Lifelong follow-up with echocardiographic evalua-
tion is recommended.
Natural History: Large VSDs
In patients with moderate to large defects, symptoms of
congestive heart failure become evident in the first few weeks
of life. Untreated, these infants with high pulmonary blood
flow and high pulmonary artery pressures will present with
congestive heart failure: poor growth, rapid/labored breath-
ing, tachycardia, and diaphoresis. They will be at risk for
recurrent pulmonary infection. If they survive without ther-
apy, they will develop pulmonary vascular disease in the first
few years of life and progress to Eisenmenger’s physiology.
In this process, the medial layer of the pulmonary arterioles
becomes hypertrophied in response to the volume and the
pressure transmitted from the LV.46 As pulmonary arteriolar
vessel lumens shrink, pulmonary vascular resistance begins to
rise. The pathway from the LV through the VSD to the lungs
is no longer a low-resistance alternative, and the left-to-right
flow diminishes. By history, children who have been in
severe heart failure for the first few years of life “get better”
as Qp/Qs returns to 1:1. The medial muscle layer of the
pulmonary arterioles continues to hypertrophy, however,
until small vessels are obliterated. Pulmonary resistance will
eventually rise to exceed systemic resistance. At this point,
patients will shunt from the RV through the VSD to the aorta,
and the patient will become cyanotic. Intervention to close the
defect at this time would lead to acute RV failure, venous
stasis, low cardiac output, and a shorter life expectancy than
if the defect were left open.51
Rarely, adults may present with new symptoms of exercise
intolerance as a result of a moderate VSD associated with a
left-to-right shunt sufficient to cause LV dilatation. These
patients have had no symptoms through childhood but may
become ill as the LV becomes less compliant with normal
aging. In contrast to the ASD in which the volume is shunted
to the right heart in response to LV chamber noncompliance,
the LV must receive the shunted blood back from the lungs.
When diastolic filling properties begin to change, the LV can
no longer accommodate the additional volume at diastolic
pressures as low as it had in childhood. LA pressures rise, and
pulmonary venous congestion leads to symptoms of dyspnea
In Ϸ0.2% of patients after myocardial infarction, rupture
of the ventricular septum can produce a significant left-to-
right shunt, with physiological consequences identical to
those of the congenital VSD.52 Unfortunately for these
patients, the acute reduction in LV stroke volume (reduced by
the amount of the shunt) and the resulting volume load
imposed on the LV compound the global or segmental
myocardial dysfunction from the ischemic insult. Over the
first 10 to 14 days after the infarct, shunting across the septal
defect may increase, potentially dramatically, as surrounding
tissue continues to necrose, enlarging the diameter of the
defect.53 Supportive care alone results in Ϸ90% mortali-
ty.54,55 Iatrogenic VSDs also may occur after aortic valve
replacement, septal ablations, and the Konno procedure to
enlarge the LV outflow tract.
Indication for Intervention
Because the natural history of the congenital VSD is well
established by the time patients reach adulthood, intervention
Sommer et al Shunts in Congenital Heart Disease 1095
is rarely required in the adult population. Adult patients with
large VSDs will have developed irreversible pulmonary
vascular disease decades earlier and should never undergo
intervention to close the defect because RV failure and
sudden death may ensue. Small VSDs from childhood have
usually closed spontaneously or, if they remain open, do not
create a clinically important shunt. Intermediate-sized de-
fects, large enough to create symptoms but not large enough
to have damaged the pulmonary vasculature, are extremely
rare. A patient with exercise intolerance, orthopnea, or other
signs of heart failure with a dilated heart on echocardiography
and normal pulmonary vascular resistance should be consid-
ered for VSD closure.
More often repair of the VSD is required for non–shunt-
related issues such as endocarditis or the development of
clinically important aortic insufficiency related to inadequate
support of the coronary cusps (particularly with subpulmo-
nary VSD). Surgical repair remains the gold standard for
treatment of VSD.51 Transcatheter closure of VSD remains
largely an investigational technique despite the development
of defect-specific devices.56–60 Two devices designed for
congenital muscular VSD have now been approved by the
Food and Drug Administration for use in the United States.
In patients with ventricular septal rupture, congestive heart
failure symptoms may be so severe that even maximal
medical support may not be sufficient to maintain cardiac
output. The need for and the timing of intervention depend on
the severity of symptoms. Patient age, baseline clinical status,
and the family’s wishes may significantly affect the decision-
making process. To date, insufficient data are available to
make definitive comparisons between surgical and transcath-
eter closure outcomes.53,61
Patent Ductus Arteriosus
The ductus arteriosus, like the PFO, is a critical component in
the fetal circulation. It is a tubular arterial structure connect-
ing the aorta and main pulmonary artery. In utero, the ductus
allows blood flow from the RV to bypass the nonfunctioning
lungs to return to the placenta via the descending aorta.
Within 72 hours of birth, the ductus closes in most newborns
through the contraction of an arteriolar smooth muscle layer,
a mechanism signaled by the rise in postnatal systemic
oxygen levels.62 If the lumen of the ductus is not fully
obliterated, an arterial connection remains between the sys-
temic and pulmonary circulations, a “patent” ductus arterio-
As in the patient with VSD, shunt direction and volume
depend on the relative resistances to flow in each pathway. In
most patients, systemic resistance is significantly higher than
pulmonary resistance, resulting in left-to-right flow (aorta to
main pulmonary artery). As in VSD physiology, the size of
the PDA is the critical resistor in the circuit and is the
principal determinant of the volume of flow. As with a VSD,
in a patient with a large PDA, the LV end-diastolic volumes
(preload) must increase to allow the stroke volume to supply
both the normal cardiac output and the left-to-right shunt at
the PDA. LA filling pressures rise, and pulmonary venous
congestion may limit exertion. In contrast to the VSD patient,
aortic blood has a route to the pulmonary circuit in both
systole and diastole, so flow occurs throughout the cardiac
cycle. In large enough defects, this diastolic “runoff” (similar
to a patient with aortic valve regurgitation) may result in
impaired coronary and splanchnic perfusion.
The clinical course of PDA is similar to that of patients with
VSD. If the PDA is large, the patient presents with symptoms
of congestive heart failure in infancy and usually goes on to
develop Eisenmenger’s physiology if the defect is not re-
paired.63 As pulmonary vascular resistance rises, a right-to-
left shunt develops at the PDA. This creates a unique clinical
picture. The aorta proximal to the PDA receives fully satu-
rated flow from the LV, and the postductal aorta receives a
mixture of LV and desaturated RV blood. “Differential
cyanosis” is the result, with normal or nearly normal oxygen
saturation in the arms and head and hypoxemia/desaturation
in the lower body. Patients may be identified by this differ-
ential coloration and by selective clubbing in the feet.63 These
PDAs cannot be closed for the same physiological reasons
that the Eisenmenger patients do not tolerate VSD closure.
If the PDA is small in the child or young adult, there are no
clinical symptoms because of the relatively small volume of
left-to-right flow. The risk of endarteritis remains, and pro-
phylaxis is required for dental work.
More frequently than in the VSD population, we have seen
small to moderate PDAs present with new symptoms in
adulthood. The development of systemic hypertension, with
its rise in systemic resistance, increases shunting at the PDA.
This, combined with diminishing LV compliance (as the
ventricle hypertrophies), may substantially increase LV fill-
ing pressures and lead to pulmonary venous congestion. In
patients with moderate shunts but normal PA pressure,
chronic LA volume loading and dilatation may lead to the
development of atrial arrhythmia.
Finally, the diastolic runoff from the aorta may lead to a
coronary “steal” phenomenon as the ductus competes for
aortic diastolic flow. Patients may present with exertional
angina in the absence of obstructive coronary artery disease.
Indications for Intervention
There are several clinical indications for intervention in adult
patients with PDA, including a significant left-to-right shunt
with LV and LA volume overload (without pulmonary
hypertension), clinical symptoms as above, or a history of
endarteritis at the PDA or adjacent structures. There are no
data to suggest that closure of a small, non–volume-loading
PDA in an asymptomatic adult is superior to dental prophy-
laxis alone. Surgical closure of PDA, first performed in
1938,64 has been replaced over the last decade by transcath-
eter closure devices in both children and adults.65–70
Pulmonary Arteriovenous Malformations
Pulmonary arteriovenous malformations (AVMs) are abnor-
mal connections between branches of the pulmonary arterial
and pulmonary venous systems that bypass the small arte-
rioles and the air-containing spaces of the lung. They occur
1096 Circulation February 26, 2008
most frequently as part of the hereditary hemorrhagic telan-
giectasia syndrome, an autosomal-dominant disorder,71 and
may occur as isolated or multiple defects that develop later in
life. Isolated pulmonary AVM also may be congenital in
nature or secondary to trauma or infection. Small pulmonary
AVMs distributed diffusely throughout the lung fields may be
a product of pulmonary parenchymal disease or the result of
end-stage liver disease.72 In complex congenital heart disease
patients, diffuse small pulmonary AVMs also may occur with
a long-standing Glenn shunt (superior vena cava to pulmo-
In the presence of a pulmonary AVM, flow through the
pulmonary artery has 2 potential pathways: the high-
resistance blood vessels of the pulmonary arteriolar and
capillary bed and the low-resistance, fistulous pathway back
to the pulmonary vein and LA. This creates an intrapulmo-
nary right-to-left shunt because the desaturated blood by-
passes the filtering and oxygenation mechanisms at the
The magnitude of the shunt depends on the resistance of
the fistulous pathway and on the number of fistulas present.
With smaller shunts, patients are usually asymptomatic. In
such a patient, the first clinical presentation of the pulmonary
AVM may be occlusive stroke or another thromboembolic
event secondary to paradoxical embolization. With much
larger single lesions or more numerous pulmonary AVMs,
patients may be hypoxemic because of the flow of desatu-
rated blood to the LA. Rarely, postural changes may
redistribute pulmonary flow to the lung segment(s) that
contain the pulmonary AVM, increasing shunting and
creating postural desaturation (orthodeoxia). Recent studies
also have linked migraine headache with aura to right-to-left
shunting in the hereditary hemorrhagic telangiectasia syn-
Indications for Intervention
In the presence of clinical cyanosis, transcatheter emboliza-
tion is the treatment of choice for individual pulmonary
AVMs. Surgical lung resections are rarely required. The
choice of transcatheter closure device depends entirely on the
size and location of the fistula to be closed.76–78 With multiple
pulmonary AVMs, some or all of the defects may be closed,
provided that pulmonary blood flow and pulmonary artery
pressures are not adversely affected. Diffuse, microscopic
pulmonary AVMs such as those seen in end-stage liver
disease cannot be treated with embolic therapy. Closure of
pulmonary AVMs to prevent recurrent paradoxical emboli-
zation and migraine headache79 remains unproven, similar to
Communications between the coronary arteries and the car-
diac chambers (coronary-cameral fistulas) or low-pressure
veins (coronary arteriovenous malformations) are most often
congenital in nature. They also may be acquired secondary to
trauma or from invasive cardiac procedures such as pace-
maker implantation, endomyocardial biopsy, coronary artery
bypass grafting, or coronary angiography. Fistulas may arise
from any branch of the coronary artery system. In congenital
fistulas, drainage is most often to the RV, RA, or the
pulmonary arteries and less frequently to the superior vena
cava, coronary sinus, pulmonary veins, or LA.80
The physiological derangement depends principally on the
resistance of the fistulous connection and on the site of fistula
termination. The resistance is determined by the size, tortu-
osity, and length of the pathway. As in a PDA, flow from the
coronary artery to a venous structure, or right-sided cardiac
chamber, occurs throughout the cardiac cycle. Blood follows
the lower-resistance pathway through the fistula rather than
traversing the smaller arterioles and capillaries of the myo-
cardium. With larger fistulas, a “diastolic runoff” may occur,
drawing blood away from the normal coronary pathway with
a widened pulse pressure and a coronary “steal.” There is a
left-to-right shunt if the fistula drains to the systemic venous
side of the circulation. Unlike the left-to-right shunt of either
ASD (RV volume load) or VSD/PDA (LV volume load),
such a shunt volume loads both ventricles. When the drainage
site is the LA or pulmonary vein, there is an effective
“left-to-left” shunt, volume overloading the left heart only.
Most coronary artery fistulas are small, and patients are
asymptomatic because myocardial blood flow is not compro-
mised. A continuous murmur may be audible at the left lower
sternal border. With increased flow, the coronary artery
branches proximal to the shunt site become significantly
enlarged. Small coronary fistulas in children tend to grow
with age. If untreated, fistulas cause clinical symptoms in
19% of patients Ͻ20 years of age and in 63% of older
patients.81 Symptoms and sequelae include chronic myocar-
dial ischemia and angina, congestive heart failure, cardiomy-
opathy, myocardial infarction, pulmonary hypertension, en-
docarditis, and rarely fistula rupture.
Indications for Intervention
Small fistulas in asymptomatic patients should be followed
up clinically for signs of growth and increasing flow. Large,
hemodynamically significant fistulas should be closed elec-
tively at the time of diagnosis. Surgical closure usually is
performed on a beating heart from the epicardial surface, is
associated with low mortality and morbidity, and has excel-
lent immediate and long-term outcomes.81 Various transcath-
eter occlusion techniques have been used82–84 with excellent
outcomes for fistulas with shorter, less tortuous courses.
Intracardiac and intravascular shunting is a common clinical
condition for patients with adult congenital heart disease.
These patients can present de novo with a previously unde-
tected lesion when shunt-related symptoms first occur in
adulthood. They may suffer from paradoxical embolization,
endocarditis, or valve deterioration indirectly related to the
defect, which draws attention to the defect for the first time.
Sommer et al Shunts in Congenital Heart Disease 1097
Familiarity with the pathophysiology of each lesion will help
the cardiologist care for these unusual patients.
We would like to thank Derron Cooke for his contribution to the
1. Hoffman JL. Incidence of congenital heart disease, I: postnatal incidence.
Pediatr Cardiol. 1995;16:103–113.
2. Garson A Jr, Allen HD, Gersony WM, Gillette PC, Hohn AR, Pinsky
WW, Mikhail O. The cost of congenital heart disease in children and
adults: a model for multicenter assessment of price and practice variation.
Arch Pediatr Adolesc Med. 1994;148:1039–1045.
3. Warnes CA, Liberthson R, Danielson GK, Dore A, Harris L, Hoffman JI,
Somerville J, Willians RG, Webb GD. Task force 1: the changing profile
of congenital heart disease in adult life: care of the adult with congenital
heart disease. Presented at the 32nd Bethesda Conference; Bethesda, Md;
October 2–3, 2000. J Am Coll Cardiol. 2001;37:1170–1175.
4. Hoffman JI, Kaplan S, Liberthson RR. Prevalence of congenital heart
disease. Am Heart J. 2004;147:425–439.
5. Williams RG, Pearson GD, Barst RJ, Child JS, del Nido P, Gersony WM,
Kuehl KS, Landzberg MJ, Myerson M, Neish SR, Sahn DJ, Verstappen
A, Warnes CA, Webb CL. Report of the National Heart, Lung, and Blood
Institute Working Group on Research in Adult Congenital Heart Disease.
J Am Coll Cardiol. 2006;47:701–707.
6. Diller GP, Uebing A, Willson K, Davies LC, Dimopoulos K, Thorne SA,
Gatzoulis MA, Francis DP. Analytical identification of ideal pulmonary-
systemic flow balance in patients with bidirectional cavopulmonary shunt
and univentricular circulation: oxygen delivery or tissue oxygenation?
7. Lindsey JB, Hillis LD. Clinical update: atrial septal defect in adults.
8. Kiraly L, Hubay M, Cook AC, Ho SY, Anderson RH. Morphologic
features of the uniatrial but biventricular atrioventricular connection.
J Thorac Cardiovasc Surg. 2007;133:229–234.
9. Webb G, Gatzoulis MA. Atrial septal defects in the adult: recent progress
and overview. Circulation. 2006;114:1645–1653.
10. Booth DC, Wisenbaugh T, Smith M, DeMaria AN. Left ventricular
distensibility and passive elastic stiffness in atrial septal defect. J Am Coll
11. Brochu MC, Baril JF, Dore A, Juneau M, De Guise P, Mercier LA.
Improvement in exercise capacity in asymptomatic and mildly symptom-
atic adults after atrial septal defect percutaneous closure. Circulation.
12. Giardini A, Donti A, Specchia S, Coutsoumbas G, Formigari R, Prandstraller
D, Bronzetti G, Bonvicini M, Picchio FM. Recovery kinetics of oxygen
uptake is prolonged in adults with an atrial septal defect and improves after
transcatheter closure. Am Heart J. 2004;147:910–914.
13. Silversides CK, Siu SC, McLaughlin PR, Haberer KL, Webb GD, Benson
L, Harris L. Symptomatic atrial arrhythmias and transcatheter closure of
atrial septal defects in adult patients. Heart. 2004;90:1194–1198.
14. Murphy JG, Gersh BJ, McGoon MD, Mair DD, Porter CJ, Ilstrup DM,
McGoon DC, Puga FJ, Kirklin JW, Danielson GK. Long-term outcome
after surgical repair of isolated atrial septal defect: follow-up at 27–32
years. N Engl J Med. 1990;323:1645–1650.
15. Kort HW, Balzer DT, Johnson MC. Resolution of right heart enlargement
after closure of secundum atrial septal defect with transcatheter technique.
J Am Coll Cardiol. 2001;38:1528–1532.
16. Gatzoulis MA, Freeman MA, Siu SC, Webb GD, Harris L. Atrial ar-
rhythmia after surgical closure of atrial septal defects in adults. N Engl
J Med. 1999;340:839–846.
17. Berthet K, Lavergne T, Cohen A, Guize L, Bousser MG, Le Heuzey JY,
Amarenco P. Significant association of atrial vulnerability with atrial
septal abnormalities in young patients with ischemic stroke of unknown
cause. Stroke. 2000;31:398–403.
18. Cherian G, Uthaman CB, Durairaj M, Sukumar IP, Krishnaswami S,
Jairaj PS, John S, Krishnaswami H, Bhaktaviziam A. Pulmonary hyper-
tension in isolated secundum atrial septal defect: high frequency in young
patients. Am Heart J. 1983;105:952–957.
19. Vogel M, Berger F, Kramer A, Alexi-Meshkishvili V, Lange PE.
Incidence of secondary pulmonary hypertension in adults with atrial
septal or sinus venosus defects. Heart. 1999;82:30–33.
20. Steele PM, Fuster V, Cohen M, Ritter DG, McGoon DC. Isolated atrial septal
defect with pulmonary vascular obstructive disease: long-term follow-up and
prediction of outcome after surgical correction. Circulation. 1987;76:
21. Frost AE, Quinones MA, Zoghbi WA, Noon GP. Reversal of pulmonary
hypertension and subsequent repair of atrial septal defect after treatment
with continuous intravenous epoprostenol. J Heart Lung Transplant.
22. Kerstein D, Levy PS, Hsu DT, Hordof AJ, Gersony WM, Barst RJ. Blade
balloon atrial septostomy in patients with severe primary pulmonary
hypertension. Circulation. 1995;91:2028–2035.
23. Blount SG, Swan H, Gensini G, McCord MC. Atrial septal defect: clinical
and physiologic response to complete closure in five patients. Circulation.
24. Argenziano M, Ox MC, Kohmoto T, Morgan J, Dimitui J, Mongero L,
Beck J, Smith CR. Totally endoscopic atrial septal defect repair with
robotic assistance. Circulation. 2003;108(suppl I):II-191–II-194.
25. Du ZD, Koenig P, Cao QL, Waight D, Heitschmidt M, Hijazi ZM.
Comparison of transcatheter closure of secundum atrial septal defect
using the Amplatzer Septal Occluder associated with deficient-vs-
sufficient rims. Am J Cardiol. 2002;90:865–869.
26. Du ZD, Hijazi ZM, Kleinman CS, Silverman NH, Larntz K. Comparison
between transcatheter and surgical closure of secundum atrial septal
defect in children and adults: results of a multicenter non-randomized
trial. J Am Coll Cardiol. 2002;39:1836–1844.
27. Valente AM, Rhodes JF. Current indications and contraindications for
transcatheter atrial septal defect and patent foramen ovale device closure.
Am Heart J. 2007;153:81–84.
28. Lechat P, Mas JL, Lascault G, Loron P, Theard M, Klimczac M,
Drobinski G, Thomas D, Grosgogeat Y. Prevalence of patent foramen
ovale in patients with stroke. N Engl J Med. 1988;318:1148–1152.
29. Webster MW, Chancellor AM, Smith HJ, Swift DL, Sharpe DN, Bass
NM, Glasgow GL. Patent foramen ovale in young stroke patients. Lancet.
30. Gersony DR, Kim SH, Di Tulio M, Fard A, Rabbani L, Homma S. Acute
myocardial infarction caused by paradoxical coronary embolization in a
patient with a patent foramen ovale. J Am Soc Echocardiogr. 2001;14:
31. Lock JE. Patent foramen ovale is indicted, but the case hasn’t gone to
trial. Circulation. 2000;101:838.
32. Reisman M, Christofferson RD, Jesurum J, Olsen JV, Spencer MP,
Krabill KA, Diehl L, Aurora S, Gray WA. Migraine headache relief after
transcatheter closure of patent foramen ovale. J Am Coll Cardiol. 2005;
33. Azarbal B, Tobis J, Suh W, Chan V, Dao C, Gaster R. Association of
interatrial shunts and migraine headaches: impact of transcatheter closure.
J Am Coll Cardiol. 2005;45:489–492.
34. Ilkhanoff L, Naidu SS, Rohatgi S, Ross MJ, Silvestry FE, Herrmann HC.
Transcatheter device closure of interatrial septal defects in patients with
hypoxia. J Interv Cardiol. 2005;18:227–232.
35. Germonpre P, Balestra C. Risk of decompression illness among 230
divers in relation to the presence and size of patent foramen ovale. Eur
Heart J. 2004;25:2173–2174.
36. Allemann Y, Hutter D, Lipp E, Sartori C, Duplain H, Egli M, Cook S,
Scherrer U, Seiler C. Patent foramen ovale and high altitude pulmonary
edema. JAMA. 2006;296:2954–2958.
37. Agnoletti G, Iserin L, Lafont A, Sidi D, Desnos M. Obstructive sleep
apnoea and patent foramen ovale: successful treatment of symptoms by
percutaneous foramen ovale closure. J Interv Cardiol. 2005;18:393–395.
38. Miller RL, Das S, Anandarangam T, Leibowitz DW, Alderson PO,
Thomashow B, Homma S. Relation between patent foramen ovale and
perfusion abnormalities in acute pulmonary embolism. Am J Cardiol.
39. Bansal RC, Marsa RJ, Holland D, Beehler C, Gold PM. Severe
hypoxemia due to shunting through a patent foramen ovale: a correctable
complication of right ventricular infarction. J Am Coll Cardiol. 1985;5:
40. Landzberg MJ, Sloss LJ, Faherty CE, Morrison BJ, Bittl JA, Bridges ND,
Casale PN, Keane JF, Lock JE. Orthodeoxia-platypnea due to intra-
cardiac shunting: relief with transcatheter double umbrella closure.
Cathet Cardiovasc Diagn. 1995;36:247–250.
1098 Circulation February 26, 2008
41. Windecker S, Wahl A, Nedeltchev K, Arnold M, Schwerzmann M, Seiler
C, Mattle HP, Meier B. Comparison of medical treatment with percuta-
neous closure of patent foramen ovale in patients with cryptogenic stroke.
J Am Coll Cardiol. 2004;44:750–758.
42. Schuchlenz HW, Weihs W, Berghold A, Lechner A, Schmidt R. Sec-
ondary prevention after cryptogenic cerebrovascular events in patients
with patent foramen ovale. Int J Cardiol. 2005;101:77–82.
43. Hung J, Landzberg MJ, Jenkins KJ, King ME, Lock JE, Palacios IF, Lang
P. Closure of patent foramen ovale for paradoxical emboli:
intermediate-term risk of recurrent neurological events following trans-
catheter device placement. J Am Coll Cardiol. 2000;35:1311–1316.
44. Braun M, Gliech V, Boscheri A, Schoen S, Gahn G, Reichmann H, Haass
M, Schraeder R, Strasser RH. Transcatheter closure of patent foramen
ovale (PFO) in patients with paradoxical embolism: periprocedural safety
and mid-term follow-up results of three different device occluder systems.
Eur Heart J. 2004;25:361–362.
45. Colvin EV. Cardiac embryology. In: Garson A, Bricker TJ, Fisher DJ,
Neish SR, eds. The Science and Practice of Pediatric Cardiology. Bal-
timore, Md: Williams & Wilkins; 1998:105–109.
46. Fried R, Falkovsky G, Newburger J, Gorchakova AI, Rabinovitch M,
Gordonova MI, Fyler D, Reid L, Burakovsky V. Pulmonary arterial
changes in patients with ventricular septal defects and severe pulmonary
hypertension. Pediatr Cardiol. 1986;7:147–154.
47. Levin AR, Jarmakani MM, Spach MS, Canent RVJ R, Capp MP, Boineau
JP, Barr RC. Ventricular intracardiac shunting mechanisms in congenital
heart disease. UCLA Forum Med Sci. 1970;10:247–263.
48. Sommer RJ, Golinko RJ, Ritter SB. Intracardiac shunting in children with
ventricular septal defect: evaluation with Doppler color flow mapping.
J Am Coll Cardiol. 1990;16:1437–1444.
49. Danchin N, Duval X, Leport C. Prophylaxis of infective endocarditis:
French recommendations 2002. Heart. 2005;91:715–718.
50. Chiu Sn, Wang JK, Lin MT, Chen CA, Chen HC, Chang CI, Chen YS,
Chiu IS, Lue HC, Wu MH. Progression of aortic regurgitation after
surgical repair of outlet-type ventricular septal defects. Am Heart J.
51. Fuster V, Brandenberg RO, McGoon DC, Giulianni ER. Clinical
approach and management of congenital heart disease in the adolescent
and adult. Cardiovasc Clin. 1980;10:161–197.
52. Crenshaw BS, Granger CB, Birnbaum Y, Pieper KS, Morris DC, Kleiman
NS, Vahanian A, Califf RM, Topol EJ. Risk factors, angiographic
patterns, and outcomes in patients with ventricular septal defect compli-
cating acute myocardial infarction: GUSTO-I (Global Utilization of
Streptokinase and TPA for Occluded Coronary Arteries) Trial Investi-
gators. Circulation. 2000;101:27–32.
53. Birnbaum Y, Fishbein MC, Blanche C, Siegel RJ. Ventricular septal
rupture after acute myocardial infarction. N Engl J Med. 2002;347:
54. Backer CL, Winters RC, Zales VR, Takami H, Muster AJ, Benson DW
Jr, Mavroudis C. The restrictive ventricular septal defect: how small is too
small to close? Ann Thorac Surg. 1993;56:101–104.
55. Seddio F, Reddy VM, McElhinney DB, Tworetsky W, Silverman NH,
Hanley FL. Multiple ventricular septal defects: how and when should they
be repaired? J Thorac Cardiovasc Surg. 1999;117:134–139.
56. Kitagawa T, Durham LA 3rd, Mosca RS, Bove EL. Techniques and
results in the management of multiple ventricular septal defects. J Thorac
Cardiovasc Surg. 1998;115:848–856.
57. Lock JE, Block PC, McKay RG, Baim DS, Keane JF. Transcatheter
closure of ventricular septal defects. Circulation. 1988;78:361–368.
58. Hijazi ZM, Hakim F, Haweleh A, Madani A, Tarawna W, Hiari A, Cao
QL. Catheter closure of perimembranous ventricular septal defects using
the new Amplatzer membranous VSD occluder: initial clinical expe-
rience. Cathet Cardiovasc Intervent. 2002;56:508–515.
59. Hijazi ZM. Device closure of ventricular septal defects. Cathet Cardiovasc
60. Holzer R, Cao QL, Hijazi ZM. An overview of device closure of ven-
tricular septal defects using the Amplatzer devices. In: Herrmann, HC, ed.
Interventional Cardiology: Percutaneous Non-Coronary Intervention.
Totowa, NJ: Humana Press; 2005:179–195.
61. Landzberg MJ, Lock JE. Transcatheter management of ventricular septal
rupture after myocardial infarction. Semin Thorac Cardiovasc Surg.
62. Hermes-DeSantis ER, Clyman RI. Patent ductus arteriosus: pathophys-
iology and management. J Perinatol. 2006;26:S14–S18.
63. Coggin CJ, Parker KR, Keith JD. Natural history of isolated patent ductus
arteriosus and the effect of surgical correction: twenty years’ experience
at the Hospital for Sick Children, Toronto. Can Med Assoc J. 1970;102:
64. Gross RE, Hubbard JP. Landmark article Feb 25, 1939: surgical ligation
of a patent ductus arteriosus: report of first successful case: by Robert E
Gross and John P Hubbard. JAMA. 1984;251:1201–1202.
65. Cambier PA, Kirby WC, Wortham DC, Moore JW. Percutaneous closure
of small (Ͻ2.5 mm) patent ductus arteriosus using coil embolization.
Am J Cardiol. 1992;69:815–818.
66. Hijazi ZM, Geggel RL. Results of antegrade transcatheter closure of
patent ductus arteriosus using single or multiple Gianturco coils.
Am J Cardiol. 1994;74:925–929.
67. Masura J, Walsh KP, Thanopoulos B, Chan C, Bass J, Goussous Y,
Gavora P, Hijazi ZM. Catheter closure of moderate- to large-sized patent
ductus arteriosus using the new Amplatzer duct occluder: immediate and
short-term results. J Am Coll Cardiol. 1998;31:878–882.
68. Faella HJ, Hijazi ZM. Closure of the patent ductus arteriosus with the
Amplatzer PDA device: immediate results of the international clinical
trial. Cathet Cardiovasc Intervent. 2000;51:50–54.
69. Hong TE, Hellenbrand WE, Hijazi ZM. Transcatheter closure of patent
ductus arteriosus in adults using the Amplatzer duct occluder: initial
results and follow up. Indian Heart J. 2002;54:384–389.
70. Pass RH, Hijazi ZM, Hsu DT, Lewis V, Hellenbrand WE. Multicenter
USA Amplatzer Patent Ductus Arteriosus Occlusion Device Trial: initial
and one year results. J Am Coll Cardiol. 2004;44:513–519.
71. Sadick H, Sadick M, Gotte K, Naim R, Riedel F, Bran G, Hormann K.
Hereditary hemorrhagic telangiectasia: an update on clinical manifes-
tations and diagnostic measures. Wien Klin Wochenschr. 2006;118:
72. Muller C, Schenk P. Hepatopulmonary syndrome. Wien Klin
73. Jacobs ML. Complications associated with heterotaxy syndrome in
Fontan patients. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu.
74. Post MC, Letteboer TG, Mager JJ, Plokker TH, Kelder JC, Westermann
CJ. A pulmonary right-to-left shunt in patients with hereditary hemor-
rhagic telangiectasia is associated with an increased prevalence of
migraine. Chest. 2005;128:2485–2489.
75. Post MC, Thijs V, Schonewille WJ, Budts W, Snijder RJ, Plokker HW,
Westermann CJ. Embolization of pulmonary arteriovenous malfor-
mations and decrease in prevalence of migraine. Neurology. 2006;66:
76. White RI Jr, Lynch-Nylan A, Terry P. pulmonary arteriovenous malfor-
mations: techniques and long-term outcome of embolotherapy.
77. Swanson KL, Prakash UBS, Stanson AW. Pulmonary arteriovenous fis-
tulas: Mayo Clinic experience, 1982–1997. Mayo Clin Proc. 1999;74:
78. Baldi S, Rostagno RD, Zander T, Rabellino M, Maynar M. Occlusion of
a pulmonary arteriovenous fistula with an Amplatzer vascular plug. Arch
79. Post MC, Thijs V, Schonewille WJ, Budts W, Snijder RJ, Plokker HWM,
Westermann CJJ. Embolization of pulmonary arteriovenous malfor-
mations and decrease in prevalence of migraine. Neurology. 2006;66:
80. Liberthson RR, Sagar K, Berkoben JP, Weintraub RM, Levine FH.
Congenital coronary arteriovenous fistula: report of 13 patients, review of
the literature and delineation of management. Circulation. 1979;59:
81. Mavroudis C, Backer CL, Rocchini AP, Muster AJ, Gevitz M. Coronary
artery fistulas in infants and children: a surgical review and discussion of
coil embolization. Ann Thorac Surg. 1997;63:1235–1242.
82. Armsby LR, Keane JF, Sherwood MC, Forbess JM, Perry SB, Lock JE.
Management of coronary artery fistulae: patient selection and results of
transcatheter closure. J Am Coll Cardiol. 2002;39:1026–1032.
83. Reidy JF, Anjos RT, Qureshi SA, Baker EJ, Tynan MJ. Transcatheter
embolization in the treatment of coronary artery fistulas. J Am Coll
84. Hakim F, Madani A, Goussous Y, Cao QL, Hijazi ZM. Transcatheter
closure of a large coronary arteriovenous fistula using the new Amplatzer
duct occluder. Cathet Cardiovasc Diagn. 1998;45:155–157.
KEY WORDS: heart defects, congenital Ⅲ physiology Ⅲ shunts
Sommer et al Shunts in Congenital Heart Disease 1099