Cardiac Catheterization in Congenital Heart Disease: Pediatric and Adult - Part 5 ppt

CHAPTER 12 Foreign body removal
369
carrier catheter, the straight thread wire is advanced out
of the carrier catheter and through the eye of the hook.
The hook is then withdrawn into the carrier catheter and,
in turn, tightens against the thread wire on the opposite
side of the foreign body, thus encircling and grasping the
foreign body very securely (Figure 12.2).
The Needle’s Eye™ snare is the strongest of the re-
trieval systems and the most likely to be able to dislodge
securely embedded foreign bodies. This same character-
istic makes this device more dangerous to use. In addition
to being very large in diameter, the entire system is very
stiff. When the hook is extruded, the wire loop of the hook
is capable of digging into the tissues as well as excavating
around foreign bodies. As a consequence, perforations are
certainly possible both while extruding the hook and
when pulling the foreign body loose. The Needle’s Eye™
system is very strong, and the grip on a foreign body is
very tight and secure. Any tissues around a foreign body
that grip it securely will tear before the Needle’s Eye™
mechanism releases or gives way, so there is a real danger
of tearing vital tissues when extracting a foreign body
forcefully with the Needle’s Eye™ device. At the same
time, the Needle’s Eye™ can be released more readily
from a grasped foreign body. The hook of the device is
re-advanced to loosen it, the thread wire is withdrawn,
and then the hook can usually be withdrawn from around
the foreign body.
Adjunct devices for foreign body retrieval
There are several other catheter accessories or catheter

devices that are designed primarily for other uses which,
however, are very useful adjunct devices for the retrieval
of foreign bodies. These accessory devices are usually
used in conjunction with another specifically designed
retrieval device/system. The most commonly used of
these adjunct devices are the bioptome forceps and the
Amplatz™ controllable deflector wires (Cook Inc.,
Bloomington, IN).
Bioptome forceps
Relatively large bioptome forceps (6- or 7-French) are
used as an adjunct in the retrieval of many foreign bodies.
The bioptome is frequently used to loosen, to reorient or to
“expose” part of a trapped foreign body so that it can be
grasped with another, stronger retrieval device, although
occasionally the complete retrieval can be accomplished
with the bioptome. The bioptome has the one major
advantage of being able to grasp the side or a small por-
tion of a foreign body without the necessity of encircling it
completely. Similar to the other retrieval systems, the
bioptome is used through a large diameter, long sheath
with a back-bleed valve and a side arm flush port. The
sheath should be 3–4 French sizes larger than the biop-
tome catheter in order to accommodate the grasped for-
eign body.
The long sheath is delivered, as previously described, to
a position where the tip of the sheath is pressed against,
and abutting a portion of the foreign body. This usually
will be against a portion that extends off the surface of the
main mass of the foreign body. Keeping the sheath in this
position the bioptome is advanced in the sheath until the

tip of the bioptome is just within the tip of the sheath. The
bioptome jaws are opened as wide as possible within the
tip of the sheath and maintained in the maximum open
configuration as the bioptome catheter is advanced out of
the tip of the sheath. As the bioptome extends beyond the
end of the sheath, the jaws open further and completely.
The opened jaws are now forced against the loose or pro-
truding piece of foreign body, and as the bioptome is
advanced, this piece of foreign body becomes forced
within the open jaws. If the bioptome jaws are not posi-
tioned exactly over the piece of foreign body, the sheath
and the bioptome catheter are maneuvered together until
Figure 12.2 Needle and eye encircling a piece of catheter.
CHAPTER 12 Foreign body removal
370
the open jaws can be pushed against the desired portion of
the foreign body. By continuing to advance the bioptome
catheter with some forward force as the jaws are closed,
the force is applied against the foreign body and the jaws
close around a part of the foreign body. Once the jaws
have gripped the loose piece of foreign body tightly, an
attempt is made to withdraw the bioptome catheter with
the attached foreign body into the sheath and completely
out of the body through the sheath. The withdrawal is
observed on fluoroscopy to ensure that the foreign mater-
ial does not become dislodged within the sheath.
Because of the small size of the “mouth” and the smooth
but sharp jaws of the bioptome, it is only usable with small
diameter pieces of catheters or other foreign bodies of
fairly small diameter, such as errant guide wires. When

the foreign material is too large or rigid, the bioptome
jaws tend to slip off the foreign body. Although the biop-
tome may not be able to grasp the foreign body securely
enough to retrieve it completely, it can often dislodge an
edge or part of the foreign body enough for another
retrieval device to grasp it more securely. The bioptome is
most effective on pieces of foreign material that can be
crushed partially by the bioptome jaws such as a piece of
catheter, wire or small intravascular plastic tubing. At the
same time, the grasped material cannot be too soft.
Bioptome blades are designed for cutting rather than
grasping, and if the foreign material is too soft, the biop-
tome cuts through the material creating several pieces
instead of the original one piece of foreign material! The
bioptome is most useful when the foreign body is wedged
very far distally in a small vessel or when only the side of
an errant catheter, wire, or tubing is accessible with no
free ends protruding into the vessel lumen. The bioptome
is particularly useful for retrieving or unwinding coils
that have embolized to the distal pulmonary arteries.
The bioptome may not develop a strong enough grip to
pull the foreign material completely into the sheath or
completely free of the surrounding tissues without cutting
it, however, the bioptome is still very useful for dislodg-
ing or moving a foreign body from one area to another. It
is also useful for dislodging and then holding a piece of
foreign material in a fixed position, allowing the foreign
material to be grasped with another retrieval device.
For example, a partially opened snare can be intro-
duced over a bioptome catheter that is holding a piece or

the edge of a foreign body. The snare loop is advanced
into the large long sheath with the partially opened loop
of the snare positioned around the bioptome catheter,
while the snare catheter passes through the sheath
adjacent to the bioptome catheter. As the snare loop is
advanced out of the distal end of the sheath, the snare is
opened widely and advanced over the piece of foreign
material which is already grasped by the jaws of the biop-
tome. Once past the jaws and now over the foreign body,
the loop of the snare is constricted around the foreign
body. The foreign body is withdrawn into the sheath
while it is grasped by both the bioptome and the snare,
or by the snare alone after it has been released from the
bioptome.
A bioptome type apparatus with toothed or slightly ser-
rated jaws would be a great adjunct for foreign body
retrieval. This would allow a secure grasp directly on the
side of the foreign material and diversify the use of a biop-
tome as a retrieval tool. Even without serrated jaws, a
bioptome is a frequent adjunct device in the retrieval of
foreign bodies.
Amplatz™ controllable deflector wire
The common Amplatz™ controllable deflector wire
(Cook Inc., Bloomington, IN) with a distal curve radius of
10 mm is another very useful device that can be used as an
adjunct during foreign body retrieval. Although designed
specifically for “bending” or deflecting intravascular
catheters, the smaller 0.021″ or 0.025″ deflector wires form
a tight curve which creates a 360°, or greater, pigtail type
loop when deflected maximally outside of a catheter. The

deflector wire is most useful for dislodging or moving
embedded foreign bodies rather than for actually retriev-
ing them. A deflector wire that is curved around an errant
catheter or other foreign body usually does not hold the
catheter or foreign body tightly enough to withdraw it
completely. Often, however, the deflector wire can be
used either to loosen or to fix a free floating foreign
body/catheter enough to allow one of the other retrieval
devices to grasp the foreign body more securely. The
deflector wire is best suited in the situation where neither
end of the foreign body is free or only the middle or
side portion of an embolized tubing, wire or catheter is
accessible to the retrieval system.
The Amplatz™ deflector wire is used through an end-
hole carrier catheter that is long enough to pass com-
pletely through the long, large diameter retrieval sheath
that is being used. A catheter with good torque control
and a stiff shaft facilitates the maneuvering of the de-
flector wire into a position for grasping the foreign body.
The deflector wire is introduced through a wire back-
bleed valve/flush port on the carrier catheter in order
to prevent bleeding around the wire, clots within the
catheter, and binding of the wire within the catheter. The
distal end of the deflector wire must be able to extend far
enough beyond the tip of the carrier catheter (2–4 cm) for
the “deflected” portion of the wire to form a complete 360°
loop. The long carrier catheter is introduced through a
back-bleed valve of a valve/side flush port on the prox-
imal hub of the long retrieval sheath, which must be large
enough to accommodate the grasped foreign body and the

catheter for the deflector wire as well as another retrieval
device/catheter.
CHAPTER 12 Foreign body removal
371
The 0.021″ and 0.025″ wires usually form tight 360°
loops, however these loops are not very strong, and
straighten out if much force is applied to them. The larger
diameter deflector wires form stronger loops, but they
usually do not form as complete or tight a 360° curve
around the foreign body. Because of their minimal
strength, deflector wires are used primarily for freeing or
reorienting pieces of foreign bodies, wires, catheters or
tubing when there is no free end of the foreign body avail-
able. Once loosened, a freed end or tip of the foreign body
or catheter can be grasped more securely by one of the
previously discussed retrieval devices.
When using a deflector wire to grasp a foreign body, the
long retrieval sheath is positioned with its tip adjacent to,
and perpendicular to the piece of catheter, wire, or other for-
eign material. The catheter that carries the deflector wire is
introduced into the sheath and manipulated out of the
sheath to a position immediately adjacent to the catheter,
wire, or other foreign body that is to be removed. The
deflector wire is then introduced through the catheter and
advanced out of the tip of the carrier catheter, adjacent to
and eventually beyond the errant piece of foreign body.
The carrier catheter, deflector wire and deflector handle
are rotated so that the concave surface of the carrier
catheter (and hopefully the deflector wire) faces the for-
eign body. When oriented in this direction and the handle

of the deflector wire is activated, the concavity of the
curve of the deflector wire curves around the piece of
foreign body as the wire is advanced further.
When the loose piece of foreign body is definitely
within the curve of the deflecting wire, the deflector
handle is tightened maximally. With the end of the tip of
the deflector wire free outside of the catheter, this allows the
wire to create a tight, greater than 360° loop around the
loose piece, thus encircling it within the tightened loop.
While maintaining a tight grip on the deflector handle, the
carrier catheter is advanced against the tip of the deflector
wire, which is curved tightly back on itself and holding
the piece of foreign body. When the tip of the carrier
catheter is forced against the 360° loop of wire, it tends to
“lock” the curve on the wire. The combination deflector
wire, carrier catheter and foreign body is withdrawn
toward and, if possible, into the sheath. If the loop of the
deflector wire has a tight grip on the foreign body and the
tip of the curve on the deflector wire is entirely within
the sheath, the entire unit is withdrawn through and out
of the sheath. If, on the other hand, the grip on the foreign
body is loose or slipping, once the foreign body is within
the sheath, it is better to remove the entire retrieval sys-
tem, including the long sheath, together as a single unit.
As with all of the other foreign body retrievals, if the
foreign body or any portion of it that is grasped by the
loop of the deflector wire is still outside of the sheath, it
should not be withdrawn through a ventricle. If the entire
loop of the deflector wire or the loose piece of foreign
body cannot be withdrawn completely into the sheath, the

combined sheath, catheter, deflector wire and foreign
body as a unit is withdrawn very slightly within the distal
vessel until one end or a piece of the foreign body is free.
Once a part of the foreign body is loosened or the foreign
body is in a location where another retrieval device can be
used, a second, “true” retrieval device is introduced to
grasp the foreign body. The “true” retrieval device can be
passed through the same long sheath or through a separ-
ate, large diameter, long sheath. When the foreign body
is difficult to grasp with the curved deflector wire or is in
a precarious location, a separate sheath with the new
retrieval device is introduced while the foreign body is
held with the deflector wire. A tight enough grip then can
be created on the dislodged foreign material with the
specific retrieval device to allow its withdrawal into the
new sheath.
An Amplatz™ deflector wire can be used adjacent to a
true retrieval system or device as part of the planned
retrieval system. When a foreign body appears to be
tightly embedded, the true retrieval system is introduced
simultaneously with the deflector wire with the loop of
the snare pre-positioned over the deflector wire before the
two are introduced into the long sheath. The snare and the
deflector wire are advanced out of the tip of the sheath
together. Once the deflector wire is adjacent to, or passes
through, the foreign body, a loop is created on the
deflector wire and then tightened, as a result of which it
partially grabs and loosens the foreign body. Once an end
or a loose piece of the foreign body has been freed, the
snare is opened completely and advanced around or over

the foreign body and the curved deflector wire that is
grasping it. Once the foreign body has been grasped
securely with the snare, the curve on the deflector wire
can be relaxed to straighten and release the deflector from
the foreign body. The loosened and now straight deflector
wire can be withdrawn from the foreign body, into, and
then out of the sheath.
Occasionally, after the foreign body has been held
tightly for some time, the deflector becomes entangled
with the snare and/or the curve of the deflector wire does
not straighten, even after the tension on the deflector
handle has been released. When the deflector wire cannot be
withdrawn from the foreign body or becomes entrapped
within the second retrieval device, the deflector wire and
the true retrieval device, along with the foreign body,
which is held by both devices, are withdrawn together
into the sheath. This does require a slightly larger dia-
meter retrieval sheath.
A third alternative is to release the deflector wire com-
pletely once even a part of the foreign body has been freed
by the deflector wire and before the true retrieval device is
introduced. The loosened foreign body is released from
CHAPTER 12 Foreign body removal
372
the curve of the deflector wire and the deflector wire is
removed from the long sheath. The true retrieval device is
introduced through the same long sheath and used to
grab the now loosened piece of foreign body.
Combinations of retrieval devices
Each of the various retrieval devices has a unique applica-

tion for different types and orientations of foreign bodies,
as described previously in this chapter. However, the vari-
ous retrieval devices are frequently used together. This
was discussed briefly under “adjunct devices” but holds
true for any combination of the “true” retrieval devices.
One device may be used to loosen or reorient a foreign
body so that a different device can obtain a stronger grip
or better orientation on it. The second device may be used
in addition to, and simultaneously with, the first one to
obtain an even stronger grip on a tightly fixed foreign body.
Pacemaker lead extractioncspecialized “foreign
body” removal
Pacemaker leads are deliberately implanted very securely
in the myocardium. In this sense, they are not an errant or
embolized foreign body. However, an intravascular pace-
maker lead can be fractured and no longer function to
sense or pace the heart. As such, they serve no positive
function and at the same time act as a source of thrombi,
mechanical irritation to the heart and heart valves, and act
as a nidus for bacterial growth and endocarditis. In most
cases an attempt is made to remove intravascular pace-
maker leads when they are no longer functional. The
transvenous pacemaker lead is usually larger than an
errant piece of intravenous catheter, wire, or tubing, and it
has been fixed very securely into the myocardium of the
heart. Pacemaker leads have usually been in place in the
vascular system for very long periods of time, which
increases their adherence to the surrounding structures
throughout their course through the vasculature. All of
these peculiarities of pacemaker leads make their retrieval

unique and complex, and have resulted in the development
of some very specialized equipment for their retrieval.
The removal of intracardiac pacemaker leads is per-
formed only by individuals specifically trained in the use
of the specialized equipment and in the techniques of
pacemaker lead extraction. This often involves the com-
bined efforts of the electrophysiologist and the interven-
tionist. The tools for freeing up the leads are unique and
are designed especially for this procedure. The intravas-
cular leads become tightly encased in dense fibrous tis-
sues (old clot!). This fibrous adherence often extends from
the site of entrance of the lead into the more peripheral
vein, along the entire course of the lead through the vein,
to the attachment within the heart.
The special tools for the “excavation” and freeing-up of
the encased lead from the surrounding adherent fibrous
tissues include large diameter, thin walled metal sheaths
(Cook Inc., Bloomington, IN) and more recently, large
diameter, long “Laser™” cutting sheaths (Spectranetics,
Colorado Springs, CO)
4
. Laser™ sheaths emit a laser cut-
ting beam from around the entire circumference of the tip
of the sheath. They require a very large and expensive
Laser™ generator (Spectranetics, Colorado Springs, CO).
The Laser™ sheath has the advantage over thin-walled
metal sheaths of being more flexible, and it can follow the
lead through its curved course in the vein. Also, with the
Laser™’s cutting capabilities, somewhat less brute force
is required to push it over the encapsulated lead within

the vein.
The lead extraction procedure requires a cut-down at
the site of the previous, old, scarred cut-down where the
lead was originally introduced, and where it enters the
vein from the subcutaneous tissues. The entrance site of
the lead into the vein is where the lead extraction equip-
ment is introduced. This cut-down is usually in continuity
with the surgically formed “pacemaker pocket” and the
old pacemaker generator, which is usually removed or
replaced. The retrieval sheaths and dilators that are used
are larger than much of the comparable equipment used
for other transcatheter foreign body retrieval proced-
ures. One piece of large, specialized retrieval equipment
developed for lead extraction is the locking stylet and
sheath that is now commercially available as the Needle’s
Eye™ Snare (Cook Vascular Inc., Leechburg, PA), which
has already been discussed in this chapter
5,6
. The Needle’s
Eye™ is particularly useful for “digging” under a lead
or catheter that is embedded in the wall of a chamber or
large vein.
The first part of the lead extraction procedure is the
loosening of the lead from the scarred adhesions along its
course from its more peripheral entrance into the vein,
centrally through the vein, and to the fixation point in the
heart. The cut-down for the lead extraction is usually on
the anterior chest wall in the area of the subclavian vein.
The lead is freed-up from the subcutaneous tissues out-
side of the vein and back to the previously implanted

pacemaker by blunt and sharp dissection. The lead is
disconnected from the pacemaker and the connector is
excised from its proximal pacemaker end. The snare end
of a special wire snare is introduced from the proximal
end of the large extractor sheath and passed completely
through the extractor sheath. The extractor sheath can be
either a short metal extractor sheath or, preferably, a
Laser™ extractor sheath, depending upon the particular
circumstances and the preference of the operator.
The freed end of the pacemaker lead is grasped securely
with the snare wire and pulled into the distal end of the
extractor sheath while still outside of the body. Tension is
CHAPTER 12 Foreign body removal
373
applied to the snare and, in turn, the freed end of the lead,
as the extractor sheath is advanced over the lead and to
and into the vein, which is opened with a small incision.
While maintaining traction on the snare catheter (and
lead), the extractor sheath is forced into the vein and used
to cut around the lead, which is encased in scar or old
thrombus along its course in the vein. This process requires
considerable to-and-fro as well as drilling motion on the
extractor sheath, while continually holding tension on the
proximal end of the snare wire. As the extractor sheath
advances further over the encased lead in the vein and dis-
appears into the vein, eventually the original free end of
the lead appears at the proximal end of the extractor
sheath. Only the more flexible Laser™ sheath can be used
once the subclavian/innominate vein makes its turn caud-
ally into the superior vena cava toward the heart. With

the Laser™ sheath, intermittent laser energy is delivered
to the tip of the sheath as this pushing/drilling process
along the lead is carried out. This procedure is continued
along the entire course of the lead through the vein until
the lead has been freed all of the way to its attachment in
the myocardium.
Once the extractor sheath has reached the “tine” or
attachment mechanism of the lead into the myocardium
of the heart, the process is continued, but with consider-
ably less force and more cautiously. Ideally, the tip of the
extractor sheath will drill just to the tip of the lead within
the myocardium but no further. Specifically, the tip of the
extractor sheath must not drill through the myocardium!
While holding the long, stiff sheath in this fixed position
even more tension is applied to the proximal end of the
lead in an attempt to free the lead from the surrounding
myocardium. When the lead has been freed from the
myocardium, it is withdrawn through the long sheath.
If the lead cannot be withdrawn from its attachment in
the myocardium and, particularly, if the lead begins to
unravel or otherwise come apart, then a secondary tech-
nique for withdrawing the lead must be used.
Once the entire course of the lead has been dissected
free from the surrounding tissues with the extractor
sheath, most of the remaining extraction of the lead is per-
formed from the femoral venous approach. The 14-French
sheath of the Needle’s Eye™ snare set (Cook Vascular
Inc., Leechburg, PA) is introduced from a femoral vein
and advanced to a position adjacent to and perpendicular
to the side of the loosened lead and as close to the attach-

ment in the myocardium as possible. The tip of the
Needle’s Eye™ sheath will be in either the right atrium or
the right ventricle, depending upon the type of lead being
extracted. The Needle’s Eye™ snare in its carrier catheter
is introduced through this sheath and advanced to the tip
of the sheath. The carrier catheter and sheath are man-
euvered so that they are touching the side of the lead.
The hook of the snare is extruded while simultaneously
maneuvering it completely around a free portion of the
lead. Once the hook has encircled the lead completely, the
thread wire is extruded and advanced through the eye at
the tip of the hook (see Figure 12.2). When the eye has
been threaded, the hook is withdrawn back into the car-
rier catheter. This grips the lead very tightly and allows a
portion of the grasped lead to be folded on itself and with-
drawn into the large sheath. The lead is withdrawn as far
as possible into the sheath. By pulling the more proximal
lead into the sheath, the tip of the large, long retrieval
sheath is pulled against the attachment of the lead in the
myocardium. This occurs particularly when the proximal
end of the lead has been disrupted or completely with-
drawn away from the attachment in the myocardium by
the previous manipulations. The side of the lead can still
be withdrawn into the sheath, even when the proximal
end of the lead is intact and passing out through the super-
ior vena cava and the subclavian vein.
Considerable force with intermittent torsion is applied
to the Needle’s Eye™ retrieval device that is grasping the
embedded lead while counter force is applied by holding
or pushing the tip of the sheath over the lead and against

the myocardium. This is usually sufficient to pull the lead
attachment out of the myocardium. Occasionally, the lead
is disrupted further by these forces, leaving a loose seg-
ment attached to the myocardium with the free end dang-
ling in the cavities of the right heart. In that circumstance,
the Needle’s Eye™ retriever is replaced with a basket
retrieval device through the same large, long sheath. The
basket is manipulated to open around the free end of
the lead. The lead is grasped while closing and rotating
the basket and, again, as much of the grasped lead as pos-
sible is withdrawn into the sheath. Considerable traction
and some torque are applied to the basket while the tip
of the sheath is pushed against the myocardium to free
the lead.
Obviously, there are considerable and sometimes
poorly directed forces necessary during this procedure for
lead extraction. Also there is no direct control over where
the long extractor sheath containing the Needle’s Eye™
excavates between the lead and the vessel or chamber
walla only that the lead was originally within the vessel
or chamber and hopefully the extractor also remains close
to the lead and completely within the channel. All of these
factors make the extraction of pacemaker leads a fairly
dangerous procedure. In fact, because of the dense adher-
ence of the lead to the vascular wall and the tight, often
deep fixation of the tip of the lead in the myocardium, it is
not a question of whether there will be a complication
from a pacemaker lead extraction, but rather, when.
Cardiovascular surgical facilities must be available imme-
diately during the extraction of all transvenous pace-

maker leads because of the probability of a major vascular
or cardiac perforation.
CHAPTER 12 Foreign body removal
374
Special foreign body circumstances
The extensive use of coils and other devices in the PDA,
the larger intravascular stents, and the large occlusion
devices for both atrial and ventricular septal defects have
created a whole new generation of intravascular foreign
bodies. When any of these devices become free floating in
the circulation, they create an entirely different challenge
for foreign body removal. The handling and removal of
these devices are covered in the specific chapters dealing
with each device, and are not repeated in this section.
Intravascular thrombi and mechanical
thrombectomy
Although, strictly speaking, intravascular thrombi/emboli
do not represent a foreign body, they do constitute
intravascular masses that often need to be dissolved or
removed urgently. Intravascular thrombi can be the con-
sequence of vascular stenosis with obstruction, localized
vascular trauma, underlying low blood flow situations,
the presence of a foreign body in the vasculature, hyperco-
agulable states, or a combination of any or all of the above
causes. As a consequence of a high index of suspicion
combined with better imaging modalities, acute thrombo-
sis of large vessels can now be detected early, but unless
always considered, will go undetected. Better imaging,
the very complex surgical repairs now being performed,
particularly in the venous circulation, the more frequent

intravenous therapy through indwelling lines that are in
place over long periods of time, and the extensive manipu-
lations in the vascular system with large catheters/sheaths,
may account for what appears to be an increase in large
vessel thrombi/occlusions. When a large vessel thrombus
is detected, the earlier definitive management is started,
the better the chance of removing the thrombus and main-
taining or restoring patency of the vessel.
Thrombolytics are quite effective at “dissolving” thrombi,
however, they are less reliable and slower than mechanical
removal, which in the presence of “visible” thrombi, is
more rapid and more definitive than thrombolytic ther-
apy alone. Thrombolytic therapy can be used as a general
systemic infusion, infused directly into the thrombus, or it
can be used in conjunction with a mechanical thrombec-
tomy device. When used alone, thrombolytic medications
are most effective when infused directly into the throm-
bus. Thrombolytic therapy is covered in Chapter 2 on
medications.
There are a variety of thrombectomy devices which, in
one way or another, fragment large thrombi into minute
particles. Some of the devices fragment thrombi into par-
ticles that are similar in size to, or smaller than, blood cells,
which allows the debris to circulate, while some of the
devices also (attempt to) extract all (most of) the debris
created by the fragmentation. Only one of the current
devices that are available for the removal of thrombi
from large vessels has a “distal protection” capability of
capturing larger fragments that may be dislodged while
the thrombus is being fragmented into micro particles or

withdrawn.
The original thrombectomy device was the fairly
straightforward, Fogarty™ balloon (Edwards Lifesciences,
Irvine, CA), which is very effective for extracting thrombi
from peripheral vessels. The Fogarty™ device is a small
balloon mounted at the distal end of a 4-French catheter
similar to a very small Swan™ balloon (Edwards Life-
sciences, Irvine, CA). The catheter with the balloon
deflated is usually introduced into the thrombosed vessel
by means of a cut-down on the vessel. The tip of the
catheter with the balloon deflated is advanced past the
thrombus. The balloon is inflated and the catheter is with-
drawn from the vessel, which pulls the now trapped
thrombus, which is adjacent to the catheter and proximal
to (in front of) the balloon, back in the vessel. The trapped
thrombus is withdrawn through the incision in the vessel
or skin and out of the body. This is still an effective pro-
cedure for removing thrombi from peripheral vessels,
although it has been supplemented by the use of throm-
bolytics and other more sophisticated thrombectomy
devices.
The simplest and one of the earliest and most readily
available of the central thrombectomy devices was a
catheter with suction applied to it
7
. Using the same prin-
cipal, but using a long, large diameter (11–16-French)
sheath instead of a catheter, makes this technique even
more effective and very useful. When a small or fresh
thrombus is detected within the vasculature, it can often

be withdrawn from the intravascular site by positioning
the tip of the large diameter sheath immediately adjacent
to the thrombus and then applying a strong vacuum with
a large volume (60+ ml) syringe to the proximal end of the
sheath. This is frequently sufficient to suck a loose throm-
bus into the sheath. Large sheaths are readily available
and usually require no additional (or expensive) equip-
ment. Occasionally the thrombus must be fragmented or
dislodged with a separate catheter, wire, or basket device.
A basket retrieval device can be used to try to grasp the
thrombus, but usually it tends to slice through the throm-
bus. Distal embolization is always a potential problem
with this type of thrombus retrieval, but if performed
expeditiously, can prevent significant further problems
from very large thrombi. In addition to these improvised
thrombectomy devices there is a variety of commercially
manufactured devices.
The Helix Clot Buster™ (ev3, Plymouth, MN) is a pneum-
atic turbine driven impeller device that has been avail-
able for over a decade. The impeller has a rotating helical
screw which is contained within a capsule at the distal tip
CHAPTER 12 Foreign body removal
375
and rotates at 100,000 rpm. The rotating screw creates a
very strong vortex of fluid, which macerates the thrombus
into particles between 13 and 1000 microns in diameter.
The capsule of the Helix Clot Buster™ device is 7-French
and it is not introduced over a wire, which makes it more
difficult to maneuver into more remote or tortuous loca-
tions unless it is introduced through a long pre-positioned

sheath. There is no extraction system with the Helix Clot
Buster™ with the result that it depends entirely upon all
of the resultant fragments being small enough to pass
through any distal capillary bed! This makes this device
less desirable for use in the systemic arterial circulation.
The Thrombex™ device (Edwards Lifesciences, Irvine,
CA) is another rotating helical screw that macerates the
thrombus, but it then aspirates the thrombus into an evacu-
ation container. The Thrombex™ is mounted on a 6-
French catheter, which makes it usable in smaller patients.
Even with built-in aspiration of the debris, there is still a
potential for some distal embolization.
The AngioJet™ (Possis Medical, Inc., Minneapolis, MN)
macerates the thrombus with multiple, extremely high-
velocity jets of saline, which are directed backward from
openings adjacent to the tip of the catheter. The high-
velocity jets create a partial vacuum by a Bernoulli effect,
which sucks the fragments out of the circulation through a
separate lumen in the catheter. The extremely high velo-
city of the jets of saline macerates the thrombus very finely,
and the very localized distribution of the jets allows the
fragments to be withdrawn effectively from the circulation.
AngioJet™ catheters are available in the 4-French XMI™,
the 5-French XVG™, and the 6-French Xpeedior™, and
they all pass over a wire, which makes them more versa-
tile for entering precise locations and traversing thrombi
in smaller vessels. AngioJet™ catheters can be used with a
single-use pump set or with a fairly complex, multi-use
drive unit. Because of the small size and the over-the-wire
use of AngioJet™ catheters, these are the most suitable for

smaller pediatric and congenital patients
8
. However, the
drive unit, in addition to being fairly complex, is expens-
ive and is usually not available in a purely pediatric/
congenital cardiac catheterization laboratory, unless the
laboratory has some association with an adult cardio-
vascular service. The drive unit is mobile, so could be
moved between co-operating and near-by laboratories.
The AngioJet™ system and catheters can be used to infuse
thrombolytics locally into the thrombus as part of the total
thrombus removal procedure.
The Oasis™ Thrombectomy Device (Boston Scientific,
Natick, MA) has a single small nozzle at the tip of the
catheter, which directs a high-velocity jet of saline back
into the port of a separate lumen in the catheter. The high-
velocity jet creates a Venturi effect, which fragments the
thrombus while, at the same time, it sucks the fragments
back into the catheter. As with the other devices without
specific distal protection systems and although the par-
ticles are very small, particulate matter that is dislodged
can embolize distally.
The Trellis Infusion System™ (Bacchus Vascular Inc.,
Santa Clara, CA) utilized occlusion balloons in the lumen
of the vessel above and below the thrombus to “contain”
the debris. A catheter that rotated between the occlusion
balloons macerated the clot into fine particulate matter,
which, in turn, was evacuated from the area between
the occlusion balloons. The rotating catheter tended to
denude the endothelium of the vessel where it touched.

When the area of the vessel could be isolated with the
occlusion balloons, this device had one of the best chances
of preventing distal embolization; as of this writing, how-
ever, the Trellis™ device has been withdrawn from the
market.
The Trerotola Percutaneous Thrombectomy Device™
(PTD) (Arrow International Inc., Reading, PA) is a self-
expanding, 9 mm diameter, stainless steel, wire basket,
which rotates at 3,000 rpm. The rotating basket fragments
the clot into “macro” particles that are as large as 3 mm,
but there is no evacuation system for these large frag-
ments. The device can be delivered over a wire, which
makes it easier to deliver to distal, more circuitous loca-
tions such as branch pulmonary arteries. The rotating
basket denudes the wall of the adjacent vessel where it
touches. The major disadvantage is the potentially larger
resultant fragments. As the thrombus is being broken up,
the area distal to the thrombectomy must be able to toler-
ate the embolization of the larger fragments.
Complications of foreign body retrieval/removal
The removal of a foreign body is usually necessitated by
an adverse event that can become a complication, but the
retrieval procedure itself can create complications. As
with all other complications, the best treatment is prevent-
ing their occurrence. The retrieval of all foreign bodies
requires extensive manipulation of often large and stiff
catheters and multiple exchanges of wires, sheaths,
catheters and the retrieval devices themselves. All of these
manipulations and exchanges carry the same, or even
higher, risks of the complications that occur from the

routine manipulation of any cardiac catheter. Meticulous
attention to the details of normal catheterization pro-
cedures for the manipulation of catheters, wires and
sheaths, and the prevention of air and/or clot emboliza-
tion are mandatory during every retrieval procedure.
Each of the separate foreign body retrieval devices is asso-
ciated with specific complications, which have been dis-
cussed previously in the discussions of each of the devices.
Intravascular structures can be torn or perforated when
a foreign body is pulled loose or withdrawn through a
vessel. This occurs most frequently during intracardiac
CHAPTER 12 Foreign body removal
376
lead extractions and is a known, half-anticipated, part of
the procedure. It does occur, very rarely, with the forceful
extraction of any foreign body that has become embedded
in tissue and should always be anticipated. With the unex-
pected or unexplained deterioration of a patient during
any foreign body extraction a tear or perforation is always
considered. The treatment is intensive medical support of
the patient with the replacement of lost blood until sur-
gical intervention is available to repair the tear.
A foreign body becoming entangled in a ventricle or a
cardiac valve is another complication of foreign body
retrieval, which for the most part is iatrogenic and should
be avoidable. Most dislodged devices (e.g. coils, occlusion
devices, stents) in the systemic venous system embolize
directly to the pulmonary artery or tumble through the
ventricles into the pulmonary artery when they become
dislodged during or after an implant. They become lodged

in the ventricle when an attempt is made to withdraw a
partially or totally exposed device through the ventricle
during a retrieval procedure. Any foreign body that can-
not be withdrawn completely into a sheath, should not be
withdrawn through a ventricle. Avoiding this problem
requires a difficult judgment decision before the foreign
body is even grasped with a retrieval device. Once a large
foreign body is grasped securely with a retrieval device,
occasionally neither can it be withdrawn completely into a
sheath nor can it be released from the retrieval device!
The probability of this combined problem should
always be considered before grasping a foreign body
with a retrieval device when the foreign body must be
withdrawn through a ventricle before its final extraction.
When a large, catheter-delivered, intracardiac device
becomes trapped in the right ventricle, it usually requires
a sternotomy and cardiopulmonary bypass to remove the
device, while a direct surgical retrieval from the pul-
monary artery requires a thoracotomy, but usually does
not require cardiopulmonary bypass. This should be
considered before an attempt at grasping or withdraw-
ing an exposed foreign body through a ventricle is even
contemplated.
Occasionally, a foreign body is loosened with a retrieval
device, but before the foreign body can be withdrawn
from the body it becomes released inadvertently or is lost
back into the circulation. This results in the embolization
of the device to another location. Unless the device
embolizes to or through a ventricle, the inadvertent loss of
the device usually does not result in any permanent

sequelae, but does require a repeat retrieval procedure.
Summary
With the professional expertise which should be present
in a well equipped pediatric/congenital catheterization
laboratory, and using one or a combination of the retrieval
systems described, the removal of virtually all intravas-
cular foreign bodies is possible by a catheter technique.
However, even with all of the latest equipment and
the best techniques, there are a few circumstances when
a foreign body cannot, or should not be removed by a trans-
catheter technique. This applies particularly to large and
not easily compressible intracardiac therapeutic devices.
Very large devices (large ASD devices, fully expanded
intravascular stents) may not compress sufficiently or fold
enough to be withdrawn even partially into even the
largest, long sheath available. Persistent attempts at the
removal of such devices, particularly if the foreign body
must be withdrawn through a ventricle, have a high
likelihood of causing significant and often permanent
intravascular or intracardiac damage. In that circum-
stance, surgical assistance for the removal of the foreign
body is needed.
When a patient in whom a large or difficult-to-remove
device has embolized must undergo cardiac surgery to
correct the particular defect anyway, or there is some
other major defect requiring surgery, there is little wis-
dom or justification for pursuing an extensive trans-
catheter removal procedure unless the device is causing
an immediate life-threatening problem. Subjecting the
patient to the often long, complex, and potentially danger-

ous transcatheter removal procedure with its large dose
of radiation is in no way justified when cardiac surgery
is inevitable. This is true particularly when the original
defect is not treatable by a similar or larger catheter-
delivered device (e.g. an embolized ASD occlusion device
in an unusually positioned or large ASD).
It requires more skill, judgment and maturity on the
part of the interventional cardiologist to determine when
a foreign body is too difficult or dangerous to remove, and
then to make the decision not to pursue the catheter
retrieval any further.
References
1. Dotter CT, Rosch J, and Bilbao MK. Transluminal extraction of
catheter and guide fragments from the heart and great vessels;
29 collected cases. Am J Roentgenol Radium Ther Nucl Med 1971;
111(3): 467–472.
2. Lillehei CW, Bonnabeau RC Jr, and Grossling S. Removal of
iatrogenic foreign bodies within cardiac chambers and great
vessels. Circulation 1965; 32(5): 782–787.
3. Moncada R et al. Migratory traumatic cardiovascular foreign
bodies. Circulation 1978; 57(1): 186–189.
4. Bracke FA, Meijer A, and Van Gelder B. Learning curve char-
acteristics of pacing lead extraction with a laser sheath. Pacing
Clin Electrophysiol 1998; 21(11 Pt 2): 2309–2313.
CHAPTER 12 Foreign body removal
377
5. Fearnot NE et al. Intravascular lead extraction using locking
stylets, sheaths, and other techniques. Pacing Clin Electrophysiol
1990; 13(12 Pt 2): 1864–1870.
6. Byrd CL et al. Intravascular lead extraction using locking

stylets and sheaths. Pacing Clin Electrophysiol 1990; 13(12 Pt 2):
1871–1875.
7. Greenfield LJ, Kimmell GO, and McCurdy WC 3rd. Trans-
venous removal of pulmonary emboli by vacuum-cup catheter
technique. J Surg Res 1969; 9(6): 347–352.
8. Kirby WC, D’sa R, and Shapiro SR. Mechanical thrombectomy
for treatment of postoperative venous obstruction in pediatric
patients. J Invasive Cardiol 2004; 16(5/Suppl): S27–S29.
378
Rashkind™ Balloon Atrial Septostomy
The Rashkind™ Balloon Atrial Septostomy (BAS), which
was introduced in 1966, was the first intracardiac, non-
surgical interventional procedure to be developed and
used clinically
1
. The Rashkind™ BAS was not only the
first procedure, but it was exceptionally innovative and
daring for the timea and even for today. Dr William
Rashkind devised the non-surgical technique for the cre-
ation of an atrial septal defect as a palliation for newborns
with transposition of the great arteries at a time when
transposition was one of the most lethal congenital heart
defects. The Blalock–Hanlon surgical atrial septectomy
was an alternative for the palliative creation of an atrial
defect, however, in a critically ill infant, the surgical sep-
tectomy was associated with a high morbidity and mortal-
ity in most centers at that time. The Rashkind™ BAS
procedure, although “crude” by all standards, was dram-
atically and instantaneously successful, and has persisted
until today with little change from the original technique

as an essential procedure for many congenital heart lesions.
Indications for a balloon atrial
septostomy
The need for an atrial septostomy is determined by the
underlying cardiac lesion and from the associated clinical
findings. Infants or older patients whose clinical signs or
symptoms can be improved by better mixing or “venting”
of the systemic venous or pulmonary venous blood are
candidates for an atrial septostomy. Patients with trans-
position of the great arteries have parallel systemic and
pulmonary circuits and benefit dramatically from mixing
of their systemic and pulmonary venous blood at the atrial
level. Patients with pulmonary atresia, tricuspid atresia,
and other hypoplastic and poorly functioning right ven-
tricles require an adequate interatrial communication to
allow systemic venous blood to return back into the
systemic arterial circuit. Patients with severe mitral steno-
sis, mitral atresia or other varieties of hypoplastic left
ventricle represent an analogous situation, but for blood
to flow in the opposite direction where it is trapped on
the left or pulmonary venous side of the circulation. These
hypoplastic left heart patients require an atrial sept-
ostomy to vent the pulmonary venous blood back into
the functional systemic circulation. Patients with total
anomalous pulmonary venous connection require an
atrial opening to permit both the systemic and pulmonary
venous blood to re-enter the systemic arterial circulation.
The BAS is accomplished by forcefully pulling (“jerk-
ing”) a small spherical balloon through an essentially
intact atrial septum, thereby tearing an opening in it. The

adequacy of the opening that is created depends upon
the toughness of the septum, the size and compliance of
the balloon used for the septostomy and the force with
which the balloon is pulled through the septum.
Hemodynamics of restrictive atrial septal
communications
The necessity for septostomy is suggested by an echocar-
diogram with the demonstration of a patent foramen
ovale or tiny atrial septal defect with restricted flow in a
patient with any of the previously mentioned anatomic
defects. The restrictive nature of the defect is confirmed in
the catheterization laboratory by the pressure difference
between the two atria or by angiography demonstrating
restrictive flow, or minimal or no mixing at the atrial sep-
tal level. The difference in pressure will be very significant
when the predominant flow of blood is, or should be, from
left to right through the atrial defect. The left atrium is rel-
atively non-compliant, and develops very high pressures
when there is restriction to the outflow of the pulmonary
venous blood from the left atrium. The right atrium, on
the other hand, by itself and along with the hepatic veins
and the connected total systemic venous “pool”, is very
13
Balloon atrial septostomy
CHAPTER 13 Balloon atrial septostomy
379
compliant. The right atrium and the systemic venous
veins stretch or dilate almost infinitely and in doing so,
prevent the development of a gradient from right atrium
to left atrium across the atrial septum even with severe

restriction of flow and even when the atrial communica-
tion is the only outlet of blood from the right atrium!
Angiography will confirm the presence or absence of
obstruction at the atrial level.
Angiography of the atrial septum
For angiographic confirmation of the atrial obstruction, in
lesions with a potential left to right shunt, an angled, cra-
nial–LAO angiogram with an injection at the mouth of the
right upper pulmonary vein is used to profile the atrial
septum. This angiocardiogram usually demonstrates a
bulging of the septum toward the right atrium with a jet
of contrast passing through any existing opening in the
septum. In lesions with a potential right to left shunt, a
straight PA and lateral X-ray view with injection in the
right atrium just at the junction of the IVC with the right
atrium is used to demonstrate the restrictive flow through
the existing atrial defect. The lateral view demonstrates
the flow of contrast from right to left most clearly with a
vertical jet of contrast passing from the right atrium,
through the defect and toward the roof of the left atrium.
Sizing the atrial septal communications in the
catheterization laboratory
There are occasions when the existing atrial communica-
tion is sized. For larger patients there are static “sizing”
balloons that can be passed over a wire and through
the defect and used to demonstrate the exact size of
the existing atrial opening in questionable cases. These
NuMED™ “sausage”-shaped sizing balloons (NuMED
Inc. Hopkinton, NY), can be inflated at very lowaalmost
zeroapressure. At this extremely low pressure, the bal-

loon within the defect conforms to the size (and shape) of
the defect without stretching or distorting it. The true siz-
ing balloons, however, all are on large catheter shafts,
which make them unsatisfactory for use in newborn or
small infants. As an alternative, a very small, standard
angioplasty balloon, when inflated at very low pressures,
can be used in the defect in a similar fashion.
A Swan™ balloon (Edwards Lifesciences, Irvine, CA) or
the actual septostomy balloons were used in the past to
“calibrate” the atrial communications. The results of this
type of sizing can be very misleading, and the technique is
no longer recommended. When using a Swan™ or a bal-
loon septostomy catheter to size the defect, the balloon
catheter with the balloon deflated is passed through the
defect from the right atrium to the left atrium. The balloon
is inflated in the left atrium and gently withdrawn against
the septum until resistance is first encountered. As the
balloon is held gently against the septum, the balloon is
deflated in small decrements. The diameter of the balloon
when the balloon easily pulls through the septum pro-
vides a rough estimate of the size of the opening. This
technique is described in detail and illustrated in Chapter
28 (“ASD Occlusion”). There are several major flaws with
this type of sizing. The spherical, Swan™ or Rashkind™
type balloons create an acute angle between the edge of
the balloon attachment to the catheter and the catheter
shaft. As the inflated balloon is withdrawn against the
septum this acute, 90° anglea rather than the full diameter
of even a partially inflated balloonacatches on the edge of
the septal defect, providing resistance to the withdrawal,

and gives a false impression of a much smaller defect.
In addition, when the balloon is pulled from the left
atrium into the right atrium in the presence of a patent
foramen with a sizable “flap valve”, the balloon actually
pulls the flap of the foramen closed against the septal
opening and, in turn, give the impression of a very small
opening. When the balloon does pull through the defect
after several withdrawals of even the partially inflated
balloon through the defect, the defect actually can be
stretched and/or dilated significantly during this “sizing”
procedure.
The alternative technique for balloon sizing, when
using a Swan™ type balloon, is to pass a guide wire from
the right atrium through the defect into the left atrium.
The Swan™ balloon is advanced over the wire only to the
right atrial/inferior vena cava junction where the balloon
is inflated maximally. The inflated balloon catheter is
advanced over the wire from the right atrium until it
pushes against the defect in the atrial septum. The balloon
is deflated decrementally while pushing its catheter over
the wire against the defect. The size of the balloon as it
eventually passes across the defect into the left atrium
indicates the size of the defect. Because of the compliance
and compressibility of the partially inflated Swan™ type
balloons, the gas filled Swan™ balloons are easily com-
pressed, which allows them to “milk” through the defect
in an elongated shape. This can also give a false impres-
sion of the size of the defect. Of even more importance,
none of the sizing techniques provide any information
about the functional significance of the defect.

Equipment for a “classic” balloon atrial
septostomy
The Miller–Edwards™ (Miller™) balloon septostomy
catheter (Edwards Lifesciences, Irvine, CA) is the pre-
ferred septostomy balloon for all infants over 3 kg. The
Miller™ balloon septostomy catheter has a small latex
balloon of 4 ml capacity mounted at the distal end of a
CHAPTER 13 Balloon atrial septostomy
380
5-French catheter shaft. The balloon material is slightly
redundant over the shaft of the catheter and usually
requires a 7-French sheath for its introduction. Just at
the area where the balloon is attached, the Miller™ bal-
loon catheter has a fixed, slight angulation of the tip of
the catheter, which is very helpful for maneuvering the
catheter with the balloon deflated from the right atrium to
the left atrium across a “closed” patent foramen ovale
(PFO). The Miller™ septostomy catheter has only the one
lumen, which communicates with the balloon, and no
separate catheter lumen, so it cannot be advanced over
a wire nor can pressures be recorded or injections per-
formed through the balloon catheter. The Miller™ balloon
catheter comes with a very fine, smooth, metal wire stylet,
which extends the entire length of the catheter lumen
(including through the area of the balloon). The distal
end of the stylet can be bent and used like a rigid deflector
wire to form a greater or different curve on the tip of the
balloon catheter, or it can be used to clear the lumen of the
balloon catheter of any obstructions.
The left atrial cavity of an infant who is larger than 3 kg,

can safely accommodate the Miller™ balloon inflated
with 6 ml of fluid when the balloon is free within the left
atrium. Although rated and advertised at 4 ml capacity,
these balloons can actually accommodate 10–12 ml before
rupturing! When a large diameter septal opening is
desired and whenever the left atrium will accommodate
it, 6 ml of fluid is used in the Miller™ septostomy balloon.
At 6 ml, the balloon not only becomes larger, but the sur-
face of the balloon becomes tense and far less compliant.
These are very desirable characteristics for tearing and not
just stretching the septum. When the Miller™ balloon is
inflated to less than 6 ml, it will be proportionally softer
and more compliant. As a consequence Miller™ balloons
with lower volumes are smaller and less tense, and
“mold” through a smaller defect rather than tearing the
edges of it.
There are special circumstances where the Miller™ bal-
loon cannot be inflated to 6 ml. In order to purposefully
create a smaller (usually temporary) septostomy or to per-
form a balloon septostomy in a very small infant, there
are smaller balloons available that inflate with a more
rigid wall tension at the smaller diameters. The USCI
Rashkind™ balloon (United States Catheter, Inc. [USCI]
BARD, Glens Falls, NY) was on a 6-French catheter, while
the newer and currently available NuMED™ septostomy
balloons (NuMED Inc., Hopkinton, NY) are on a 4-French
catheter shaft and pass through slightly smaller intro-
ducer sheaths.
The more recent USCI Rashkind™ balloon (manufac-
turer as above) was a latex balloon mounted on and

recessed into a 6-French woven dacron shaft. If the bal-
loon deliberately was not fully inflated, when it had been
being prepared and cleared of air and bubbles, it passed
through a 6-French sheath (as a minimum). Like the
Miller™ balloon, this USCI Rashkind™ septostomy bal-
loon did not have a second lumen. The USCI Rashkind™
balloon had the additional major disadvantage of having
a much smaller balloon size, with only a 2.5 ml capacity at
its maximum inflation volume compared to the 6 ml of the
Miller–Edwards™ balloon. Even at its full inflation, the
USCI Rashkind™ septostomy balloon was fairly soft and
moderately compliant.
The NuMED™ Z-5™ septostomy balloons (NuMED
Inc., Hopkinton, NY) are available in two sizes. These
balloons are manufactured from a non-compliant thermo-
plastic elastomer, which results in an essentially fixed
diameter, rigid balloon when inflated with the recom-
mended volume. The smaller Z-5™ balloon reaches a
diameter of 9.5 mm when inflated with 1 ml of fluid while
the larger Z-5™ balloon has a 13.5 mm diameter when
inflated with 2 ml of fluid. As a consequence, although
they have a volume of only 1 or 2 ml when fully inflated,
they produce very rigid balloons with diameters of
exactly 9.5 and 13.5 mm. The deflated balloons pass
through 5- or 6-French sheaths, respectively.
To facilitate crossing the atrial septum, the distal shafts
of the Z-5™ septostomy balloons are angled to approx-
imately 35° just proximal to the location of the balloon. In
addition, both of the NuMED™ Z-5™ septostomy balloon
catheters have a separate catheter lumen from the balloon

lumen. The separate lumen allows the smaller and larger
catheters to be introduced over 0.014″ and 0.021″ wires,
respectively. The wire can be positioned securely in the
left atrium or a pulmonary vein separately through a more
maneuverable end-hole catheter, and then the balloon
septostomy catheter is passed over the pre-positioned
wire. This is very beneficial in directing the balloon into a
very small or malpositioned left atrium and to be abso-
lutely sure that the balloon is in the proper position. The
wire through the balloon catheter can be left in place dur-
ing the rapid balloon withdrawal during the septostomy.
The wire remaining in place is helpful in re-entering the
left atrium for a repeated balloon withdrawal, and pro-
vides a safety mechanism in the event of the balloon separ-
ating from the catheter. With the wire removed from the
catheter lumen, pressures can be recorded or small contrast
injections for angiograms can be performed through this
separate lumen to verify the exact location of the balloon.
In patients in whom the shunting is right to left at the
atrial level, particularly in those with total anomalous
pulmonary venous return, the left atrial chamber is very
small. In these very small left atria, the Miller™ balloon
conforms to the shape of the small left atrium and begins
to stretch and distort the atrium well before it reaches a
volume of 6 ml (and its maximum diameter). In that situ-
ation, the balloon will be fixed very firmly within the small
left atrium and will have no “bounce” or movement. In
CHAPTER 13 Balloon atrial septostomy
381
addition, when the balloon is inflated and either filling the

atrium or pulled against the atrial septum, all systemic
cardiac output will be stopped totally! Under these
circumstances, the balloon inflation must be carried out
very cautiously to avoid over distention of the atrium, but
at the same time, rapidly to avoid too long a period of
hypotension, bradycardia and even cardiac arrest. Once
the balloon has assumed the shape of the atrium and
stopped moving, inflation is stopped and the withdrawal
is performed rapidly with that degree of inflation (usually
less than 4 ml). This is the one situation where rapid over-
inflation to the full volume of the Miller™ balloon could
be disastrous and could cause rupture of the left atrium.
When a very small left atrium is found, particularly in
very small infants under 2 kg in weight, in infants where
the venous system does not tolerate the 7-French intro-
ductory sheath, or in larger patients with right to left
shunting, the NuMED™ 2 ml septostomy balloon is used
in preference. Neither of these balloons is satisfactory
when a large, permanent opening in the atrial septum
is desired. The technique and precautions for a balloon
septostomy with either of these alternative balloons are
similar to those for the Miller–Edwards™ catheter.
Balloon septostomy in the catheterization
laboratory
Whenever a balloon atrial septostomy is necessary in
neonatal patients, the patient’s hemodynamic stability is
often dependent upon the atrial communication. As a con-
sequence, the balloon atrial septostomy is performed at the
beginning of the procedure before any extensive catheter
diagnostic studies, whether the septostomy is performed

in the catheterization laboratory or the newborn intensive
care unit under echocardiography guidance.
The catheterization laboratory has many advantages
for the balloon septostomy. All of the ancillary equipment
necessary for catheter introduction and the septostomy
procedure are available and immediately accessible.
Overall visualization of the catheter in relation to the
remainder of the anatomy, and the septostomy procedure
itself are all visualized clearly using biplane fluoroscopy.
With the ancillary equipment in the laboratory and with
the use of fluoroscopy, the procedure can be altered to suit
any variation in the patient’s anatomy. The facilities for
cardiopulmonary resuscitation in the event of a problem
during the catheterization, and subsequent access to the
operating rooms, are usually better in the catheteriza-
tion laboratory. Performance of the septostomy in the
catheterization laboratory does not preclude the use of
echocardiography in the catheterization laboratory as an
adjunct means of visualizing the septum, the balloon and
the results.
Occasionally an umbilical venous line is already pres-
ent in these infants, and although it is a less desirable
approach, it is often used for balloon atrial septostomy out
of “expediency”
2
. If the umbilical vein is used, the existing
umbilical vein catheter is replaced over a wire with a
sheath/dilator set that will accommodate the septostomy
balloon. The sheath and dilator are advanced into the
umbilical vein until the tip of the sheath is well into the

right atrium. The positioning of the sheath with its tip
extending centrally past the entrance of the hepatic veins
into the inferior vena cava/right atrium assures repeated
access to the right atrium and septum with the septostomy
catheter. When attempts are made at replacing the umbil-
ical vein catheter directly with the septostomy catheter,
often the ductus venosus spasms and the septostomy
catheter becomes “detoured” and stuck in the hepatic/
portal vein system and cannot be maneuvered to the atrium
at all from the umbilical approach. This is a particular
problem when using a Miller™ septostomy balloon,
which has no separate lumen for a wire.
The femoral vein approach is preferred when a large
atrial septostomy is desired, and particularly when it is
planned for the atrial septal defect that is created, to last
any length of time after the procedure. Although the
umbilical vein is patent and even cannulated with an
umbilical catheter, the results of the septostomy are less
satisfactory from the umbilical vein approach. When the
balloon crosses the septum from the umbilical vein, there
is a very short distance between the balloon’s exit from the
small ductus venosus (end of the sheath) to the foramen
ovale as compared to the distance from further down in
the inferior vena cava to the foramen. This short distance
restricts the distance the balloon is able to travel during
and after the forceful pull through the septum and, in
turn, limits the force that can be applied to pull (“jerk”) the
balloon through the septum. If the balloon is pulled too far
and forcefully against the hepatic veins/ductus venosus,
the veins can be disrupted or pulled off the right atrium,

and if the balloon is pulled forcefully against the tip of the
sheath, it can be ruptured or even totally dislodged from
the catheter.
When a smaller opening in the septum is necessary,
and it only has to stay open for a few days before pro-
posed surgery, for example in a patient with a simple
transposition or a hypoplastic left heart, then the umbil-
ical approach is satisfactory. The smaller NuMED™ bal-
loons can be used and less force (a weaker “jerk”) applied
via the septostomy catheter. When using one of the
NuMED™ septostomy balloons, first an end-hole torque
catheter is introduced into the umbilical vein, maneu-
vered through the septum and into a pulmonary vein,
and replaced with a wire that the NuMED™ balloon sept-
ostomy catheter will accommodate. When there is a
pre-existing umbilical venous catheter, it is replaced with
CHAPTER 13 Balloon atrial septostomy
382
a small spring guide wire that the septostomy catheter
will accommodate, and a torque-controlled end-hole
catheter is introduced over the wire and maneuvered into
the left atriumaand preferably into a left pulmonary vein.
The wire is advanced as far as possible and fixed in the
pulmonary vein, the catheter is withdrawn over the wire,
and the septostomy balloon catheter advanced over the
wire and directly into the left atrium. When only a small or
temporary septostomy opening is necessary, this is the
most expeditious technique for performing a BAS. Without
the sheath through the umbilical vein, there is slightly
more room for the withdrawal (without hitting against the

tip of a sheath), but the withdrawal distance is still limited
in order not to pull the balloon into the hepatic veins/
ductus venosus.
During a balloon atrial septostomy procedure per-
formed in the neonatal intensive care unit or the catheter-
ization laboratory, an arterial monitoring line is desirable,
and in our laboratory is considered essential. In the new-
born infant with congenital heart disease, the arterial line
is often present as an umbilical artery line when the infant
arrives in the catheterization laboratory. If not, either the
umbilical artery is cannulated in the catheterization lab-
oratory with an umbilical artery catheter, or a femoral
artery is cannulated with a 20-gauge teflon Quick-cath™
while the sheath is being introduced into the femoral vein.
When a large or persistent opening is desired, the sept-
ostomy balloon is introduced from the femoral vein.
When a balloon atrial septostomy is the purpose of the
procedure or even if it is only a consideration, a sheath
that is large enough to accommodate the balloon sept-
ostomy catheter to be used is introduced into the femoral
vein at the onset of the procedure. A standard, short sheath
with a back-bleed valve and flush port is used. By starting
with the necessarily larger sheath, net trauma to the vein
is reduced by eliminating the subsequent exchange to a
larger sheath. Starting with the larger sheath also reduces
the possibility of not being able to introduce it at all later
during the catheterization because of venous spasm of the
small femoral vein. A 7-French sheath is required for
Miller™ balloons and a 6-French sheath for the larger Z-
5™ septostomy balloons. When a balloon septostomy was

not planned initially and a smaller sheath was used dur-
ing the initial part of the catheterization, the existing
smaller sheath used for the diagnostic catheterization
should be replaced with the larger sheath and dilator as
soon as the decision to perform the septostomy has been
made. The skin and subcutaneous tissues surrounding the
puncture site are re-infiltrated very liberally with local
anesthesia before exchanging the sheaths.
Once the appropriate sheath for the septostomy balloon
has been introduced into the vein, the dilator is removed
very slowly, and the indwelling sheath with its side arm
and proximal valve chamber is carefully cleared of all air
by allowing it to bleed-back as the valve of the side port is
opened very cautiously. If the infant is experiencing any
respiratory difficulty, particularly with excessive inspir-
atory effort, no attempt is made to clear the sheath by free
or passive back bleeding. With any strong inspiratory
effort on the part of the patient, air is easily sucked into an
open sheath or an open side port on the back-bleed valve.
When there is even minimal inspiratory obstruction, and
after the dilator and wire have been removed, the back-
bleed valve of the sheath is covered tightly with a gloved
finger, a syringe is attached to the side port, and the stop-
cock on the side port is opened while suction is applied
gently to the side port. This allows blood to flow freely
from the side port while preventing air from being sucked
into the sheath through the valve by the suction on the
syringe or the patient’s inspiratory effort. Once com-
pletely clear of any air, the side arm is attached to a separ-
ate flush line and placed on a slow and continuous flush.

The septostomy balloon is prepared outside of the
body. Normal saline (not dextrose with saline) is used for
flushing the balloon and to dilute the contrast to be used in
the balloon. Glucose solution, by itself, is sticky and can
clog the small lumen of the septostomy catheter. The sep-
tostomy balloon is inflated with one, to at most 1.5 ml,
of normal saline. This same volume is withdrawn along
with any air from the balloon or catheter lumen. The
filling with fluid and the withdrawal of the fluid and any
air are repeated several times in order to remove as much
trapped air from the balloon and catheter lumen as pos-
sible. Each time only a very small amount of saline flush is
used. The septostomy balloon should not be inflated to its
full capacity during this preparation. Inflation to the full
volume stretches the materials of the balloons, which
causes the balloon material to become more redundant
around the catheter shaft and, in turn, makes the introduc-
tion through the appropriate sized sheath more difficult
or even impossible.
The balloon is emptied completely. In a separate
syringe, one ml of contrast is diluted with 9 ml of saline to
provide exactly 10 ml of a 9:1 dilution of saline to contrast
solution. This syringe is attached to the previously flushed
and emptied balloon lumen. By starting with exactly 10 ml
in the syringe, it is easier and quicker to determine the
precise amount of diluted contrast solution that is filling
the balloon during the actual septostomy.
The Miller™ septostomy balloon catheter comes with a
fine steel stylet within the entire length of the balloon
lumen. In preparing the balloon for use, the stylet is

removed but is kept on the sterile field. The stylet has sev-
eral potential critical uses. If the balloon catheter cannot be
maneuvered readily from the right atrium into the left
atrium through a small or “tight” existing defect, a small,
short, smooth 45° curve is formed on the distal end of the
stylet. The syringe is removed from the hub of the balloon
CHAPTER 13 Balloon atrial septostomy
383
catheter and the stylet is inserted through the balloon
lumen to the tip of the balloon catheter. The stylet within
the lumen serves as a rigid deflector wire, creating an
additional curve on the distal end of the balloon catheter
and stiffening the shaft of the catheter, both of which facil-
itate advancing the balloon catheter into the left atrium.
The other very important use for the stylet is to assist in
the deflation of the balloon after the balloon septostomy
procedure has been completed. Even when using very
dilute solutions of contrast with saline to fill the balloon,
there are occasions when the balloon does not deflate
following the septostomy. In this situation, the stylet is
reintroduced into the lumen of the balloon septostomy
catheter to ream out any dried contrast or other obstruct-
ing material within the lumen. This clearing of the lumen
usually allows the subsequent deflation of the balloon
when the stylet has been withdrawn.
“Classic” balloon septostomy procedure
Regardless of the type of balloon septostomy catheter
being used, the surface of the balloon is moistened in the
saline flush solution and the rate of flush into the side port
of the hemostasis valve on the sheath is increased. Using

the syringe attached to the proximal end of the balloon
lumen, mild negative pressure is applied to the balloon
and the balloon is introduced through the valve of the
venous sheath. Very slight rotation on the shaft of the bal-
loon catheter often facilitates the passage of the deflated
balloon through the back-bleed valve and into the sheath
by wrapping the redundant balloon material around the
catheter slightly. The balloon is advanced slowly through
the sheath while continuing the flush into the side port
until the balloon passes beyond the distal tip of the sheath.
The redundant material of the deflated balloon advancing
within the lumen of the sheath acts as a “plunger” that can
create a partial vacuum behind it. The continual flush
through the side port fills the sheath behind the balloon as
the vacuum develops and, in turn, prevents air from being
sucked in through the back-bleed valve by this vacuum as
the balloon is advanced through the sheath. Once the
balloon has advanced past the end of the sheath, the nega-
tive pressure is released from the syringe attached to the
balloon lumen and the continual flush on the side port is
reduced to a slow drip. The balloon is maneuvered cepha-
lad in the IVC and manipulated through the pre-existing
interatrial septal defect and/or PFO into the left atrium.
In the presence of a very tense left atrium or any distor-
tion of the atrial anatomy, maneuvering across the atrial
septum can be difficult. With Miller™ septostomy bal-
loons, the solid stylet is introduced into the catheter lumen
to help direct the balloon through the atrial communica-
tion. When NuMED™ septostomy balloons are being
used there are several options to assist in crossing the

septum. The stiff end of a small spring guide wire or a
Mullins™ wire (Argon Medical Inc., Athens, TX) can be
used within the second lumen of the catheter to deflect the
tip and support the shaft of the catheter for better maneu-
verability. A soft-tipped wire can be advanced out of the
distal end of the NuMED™ septostomy catheter and the
wire maneuvered across the atrial septum and followed
by the septostomy catheter. Finally, a separate end-hole,
torque-controlled catheter can be maneuvered across the
tight atrial communication and into a pulmonary vein and
replaced with the fine wire that the NuMED™ septostomy
catheter will accommodate, and the septostomy catheter
advanced into the proper position over the wire.
The position of the catheter tip within the left atrium is
visualized on both the posterior–anterior (PA) and lateral
(LAT) fluoroscopy to ensure that the balloon tip is posi-
tioned properly in the left atrium, and when available the
balloon is visualized simultaneously on echocardiogram.
The balloon is initially inflated slowly while its position
and motion in the atrium are observed intermittently
but frequently on both the PA and LAT fluoroscopy (and
echo). The systemic blood pressure and the electrocardio-
gram (ECG) are monitored continuously and very care-
fully while the balloon is being inflated. As the balloon
begins to inflate, it should move freely within the left
atrium and, in fact, begin to “bounce” cephalad to cau-
dally toward the mitral valve or left ventricle with each
atrial contraction. The correct balloon bounce is perpen-
dicular to the shaft of the catheter. If this movement of the
balloon is not seen, its inflation is stopped immediately.

The exact position of the balloon is rechecked carefully on
the biplane fluoroscopy and echo. When balloon move-
ment or bounce does not occur or the balloon appears in
an abnormal location on echo, there are many dangers
and abnormal areas where the balloon may be located.
Abnormal positions of the balloon that must be excluded
include:
1 The left atrial appendage. A balloon in the left atrial
appendage appears slightly more anterior in the LAT
fluoroscopic view. With the initial balloon inflation, the
shape of the balloon begins to distort as the balloon con-
forms to the shape of the appendage. The balloon in the
appendage is seen clearly on echo. Further inflation in
the appendage or withdrawal of the balloon when it is
trapped in the appendage would tear or totally disrupt it.
2 A pulmonary vein. When the balloon is in a pulmonary
vein, the tip of the balloon catheter is positioned lateral to
(outside of ) the outline of the left atrium on the PA fluoro-
scopic view or posterior to the left atrium on the LAT
view. With the initial inflation, the shape of the balloon
elongates to conform to the size and shape of the pulmon-
ary vein. It is common that the balloon, or at least its tip,
begins to inflate in a pulmonary vein, but with the initial
part of the inflation the balloon should “milk” out of the
CHAPTER 13 Balloon atrial septostomy
384
vein and free into the left atrium. This is probably the
safest starting location for the initial inflation as long as
the balloon is observed very closely and does milk out of
the vein on both fluoroscopy and echo. If the balloon does

not milk out of the vein during inflation, the vein could be
split by the full inflation of the septostomy balloon.
3 In a juxtaposed right atrial appendage. A balloon posi-
tioned abnormally in a juxtaposed right atrial appendage
should always be considered a possibility. It is the most
difficult abnormal position to recognize and absolutely
rule out. The juxtaposed right atrial appendage passes
from the right atrium behind the great arteries to the left
heart boarder and is in very close approximation to the
left atrial cavity/left atrial appendage. A juxtaposed right
atrial appendage should have been identified by prior
echocardiographic or angiographic studies and when
present, there should be a high index of suspicion of this
area as the balloon is being positioned. A balloon posi-
tioned in a juxtaposed right atrial appendage appears in
a proper location (in the area of the left atrium) for a
septostomy on both PA and LAT fluoroscopy, but not on
echo when this is used simultaneously. However, the high
index of suspicion, the rigid fixation in the position of
the balloon during the initial inflation, the distortion of
the shape of the balloon with the early inflation, and the
absence of “bounce” toward the mitral valve, will differ-
entiate the position in the juxtaposed appendage from
the correct position even without echo. When there is a
known juxtaposed right atrial appendage, it is advisable
to visualize the balloon with echo as well as fluoroscopy
before withdrawing it. Withdrawal from a juxtaposed
atrial appendage can totally disrupt the appendage or tear
it loose from the right atrium.
4 In the left ventricle. As the balloon is inflated in the left

atrium and begins to “bounce” properly, it can easily be
sucked across the mitral valve and into the left ventricle
if there is significant slack on the shaft of the balloon
catheter as the balloon is inflated. If the balloon inflation
continues with the balloon in the ventricle, the balloon is
positioned caudally and laterally in the PA projection and
more anteriorly and caudally in the LAT projection. The
balloon in the left ventricle becomes relatively fixed in its
position on fluoroscopy compared to the bounce when it
is in the left atrium. The very caudal location in the ven-
tricle is very apparent unless there are marked distortion
and peculiarities in the overall position of the heart
and/or the chambers in relation to each other. The balloon
in the left ventricle moves and is distorted slightly with
each ventricular systole. In the left ventricle, the shaft of
the balloon catheter tends to pull away from the operator
along the long axis of the catheter in the direction of the
apex of the ventricle, rather than the usual bounce perpen-
dicular to the long axis when the catheter is correctly posi-
tioned in the left atrium. The balloon in the left ventricle
and the catheter passing through the mitral valve are very
obvious on echocardiography. A forceful withdrawal of
the balloon from the left ventricle would disrupt the left
atrioventricular valve.
5 The right ventricle. When the septostomy balloon is
grossly malpositioned in the right ventricle, the position
of the balloon is far anterior on the LAT fluoroscopy
and moves (bounces) perpendicular to the long axis of
the catheter, but cephalad toward the right ventricular
outflow tract during each systole. The abnormal location

on lateral fluoroscopy alone should identify this abnormal
balloon position and certainly, a simultaneous echo easily
confirms this abnormal location. A rapid, forceful with-
drawal of the septostomy balloon from the right ventricle
results in disruption of the tricuspid valve.
6 The coronary sinus. A balloon positioned in the coro-
nary sinus appears very posterior and caudal on LAT
fluoroscopy. The otherwise round septostomy balloon
assumes a sausage-like shape and becomes very fixed
with the initial inflation of the balloon. A septostomy
balloon inflated in the coronary sinus usually causes
discomfort for the patient, ST changes on the electro-
cardiogram, or bradycardia. Full inflation of the balloon
in, or a forceful withdrawal from, the coronary sinus
could tear or split it.
Correct positioning of the septostomy balloon
catheter
As the balloon begins to inflate when positioned properly
in the left atrium, it begins to “bounce” perpendicularly to
the long axis of the catheter shaft and toward the mitral
valve. As soon as this motion toward the mitral valve
is seen, the balloon is carefully and gently withdrawn
toward, and fixed against, the intact interatrial septum
during the remainder of the inflation to the desired vol-
ume. As the balloon is inflated further, continued care is
taken to prevent it from being sucked through the left A-V
valve and into the ventricle. This is avoided by maintain-
ing gentle tension on the balloon and careful observation
of its motion on fluoroscopy or echo during its initial posi-
tioning and entire inflation.

The final positioning, the inflation of the balloon in the
left atrium, and the withdrawal of the balloon are the
times when the adjunct use of the echocardiogram in
the catheterization laboratory is extremely useful, and
where echo guidance may have a slight advantage over
fluoroscopy guidance. The balloon itself and all of the
adjacent structures or areas mentioned are seen clearly on
the echo. Any abnormal position or motion of the balloon
is recognized on the echo image immediately. It is more
apparent on the echo image if the balloon is too large for
the particular atrium and the left atrium is being distorted
extensively during the inflation.
CHAPTER 13 Balloon atrial septostomy
385
Once confident that the balloon is correctly positioned
in the left atrium, and not in any of the danger areas, the
balloon is inflated more rapidly to its maximum volume,
or at least to the maximum volume and diameter tolerated
by the infant’s anatomy. For the first withdrawal (“pull
through”) and in order to achieve maximum effect from
the septostomy, the balloon is filled with the maximum
volume that it is planned to use or that can be tolerated by
the infant. The balloon withdrawals should not start with
a smaller volume and should not be performed with gradu-
ated and/or gradually increasing diameters of the bal-
loon. Performing balloon withdrawals with gradually
increasing volume is more likely to stretch the existing
opening temporarily, rather than actually tear a new opening.
As long as the balloon is being inflated, it is observed
very frequently or even continuously on biplane fluoro-

scopy and/or echocardiography. Fluoroscopic observation
is predominantly in the PA projection, but it is also inter-
mittently checked in the LAT view to watch for not only
the danger areas, but to be sure that the balloon is not
exceeding the volume/diameter of the left atrium. If the
balloon reaches the maximum atrial volume before it
reaches the desired maximum balloon volume, the balloon
becomes distorted as it assumes the shape of the atrium,
becomes unusually fixed in the atrium, and usually causes
the patient’s hemodynamics to “crater”. This is more
likely to occur in very small infants or in infants with right
to left shunt lesions with very small left atria.
The electrocardiogram and systemic blood pressure are
observed very carefully during the entire process. As the
balloon volume approaches the volume of the atrial cham-
ber, the heart rate slows and the systemic pressure drops.
This is a result either of the inflated balloon’s occluding all
systemic output or of the abnormal stretch on a small left
atrium. Such a reaction does not preclude continuation of
the septostomy, but is an indication for (1) reassessment of
the location of the balloon and its size-relationship to the
LA by both biplane fluoroscopy and echocardiogram; (2)
pre-treatment with atropine; and (3) a more rapid and
dextrous inflation and pull-through of the balloon.
When the balloon is fully inflated to the maximum
diameter for the septostomy (up to 6 ml for the Miller™
balloon), the catheter with the balloon is maintained fixed
gently against the septum. The shaft of the catheter is
grasped with a very firm grip (even to the point of wrap-
ping the shaft of the catheter around a finger several

times). A recording with biplane angiography or stored
fluoroscopy is started. While recording continuously, the
balloon catheter is withdrawn with a very rapid and force-
ful “jerk”. At the same time, the withdrawal or jerk from
the left atrium back into the right atrium is for a very short
distance and must be very controlled. Extreme care is
taken to assure that there is no “wind-up” on the catheter
just before this rapid withdrawalai.e. the operator must
avoid the almost instinctive tendency to push the shaft of
the balloon catheter (and the balloon) forward into the left
atrium just before withdrawing it through the septum.
Such a “wind-up” allows the balloon to drop away from
the septum and potentially through the left atrioventricu-
lar valve into the ventricle just before the pull on the
balloon starts!
The forceful pull or jerk on the balloon should be per-
formed with the fingers and wrist, not with the arm and
elbow. Withdrawal of the balloon should be so rapid that,
when filmed at sixty frames per second, the actual with-
drawal is visible on only one framea i.e. it lasts only
1/60th of a second! The balloon is yanked forcefully back,
and completely into, the right atrium during the with-
drawal, but it is not withdrawn any further down into the
inferior vena cava (or hepatic veins). In the presence of
a high-pressure left atrium, the atrial septum bulges to the
right, is displaced into the right atrium, and, as a con-
sequence, the foramen is often low and very close to the
inferior vena cava. In this situation (and during any sept-
ostomy withdrawal) it helps to position the hand that is
not pulling the catheter on the infant’s leg 5–6 cm caudal

to the site where the septostomy catheter is grasped with
the other hand or fingers. During the septostomy with-
drawal, the second hand serves as a stop to the with-
drawal and helps to prevent pulling the balloon too far
into the inferior vena cava.
The heart rate and blood pressure drop very signi-
ficantly immediately following a successful septostomy.
While observing the balloon on fluoroscopy, it is immedi-
ately pushed back toward (into) the mid right atrium
and deflated as rapidly as possible by suction applied to
the proximal end of the balloon lumen. When pushed
cephalad after a successful pull through the septum, the
balloon moves anteriorly on the LAT fluoroscopy while,
if the septum was only displaced and the balloon did
not pull through it, the balloon moves posteriorly back
into the left atrium when pushed cephalad. Following a
successful balloon atrial septostomy, the heart rate and
blood pressure return rapidly to normal without any extra
external effort.
Once the infant stabilizes, the deflated balloon is man-
euvered into the left atrium, and the procedure repeated
three to six times, or until absolutely no resistance is felt
as the balloon is withdrawn rapidly through the defect.
Each balloon withdrawal is performed with the balloon
inflated to its maximum volume and diameter as deter-
mined for that particular patient. Although no increase
in balloon size is used during the repeated balloon
withdrawals, the repeated withdrawals tend to extend
any tears in the septum that were initiated by the initial
withdrawal.

Once the operator feels secure that the septum is
opened to the maximum possible diameter with the
CHAPTER 13 Balloon atrial septostomy
386
particular balloon, the status of the opening is tested by
reintroducing the balloon into the left atrium, inflating
it with a slightly smaller volume of the dilute contrast,
and performing a slow, gentle withdrawal, allowing the
balloon to “float” through the newly created opening.
The amount of dilute contrast in the balloon is increased
in small amounts until slight resistance to the withdrawal
is felt, i.e. starting with 1 ml and working up in approx-
imately 1 ml increments. When an adequate opening has
been created, the balloon should float through the defect
even with the increasing increments of fluid.
When echocardiography is available in the catheteriza-
tion laboratory or when it is used for an echo-guided
septostomy, the atrial septum and the defect are viewed
on the echo. The actual defect is usually defined very
clearly and the flow through the defect can be seen on
color Doppler.
Once convinced that the opening is adequate, the
balloon catheter is withdrawn by applying suction to the
proximal end of the balloon lumen while it is being with-
drawn through the inferior vena cava, back into the sheath
and out of the body. Occasionally, the balloon cannot be
withdrawn into the venous sheath. In that case, the bal-
loon is re-advanced several centimeters into the IVC and a
0.025″, soft-tipped, spring guide wire is introduced into
the back-bleed valve adjacent to the balloon catheter at the

proximal end of the sheath. The guide wire is advanced
well into the cardiac silhouette, or even the superior vena
cava. The balloon is withdrawn back to the distal end
of the sheath and as far into the sheath as possible. The
sheath and the balloon catheter are withdrawn together
over the wire out of the skin. This allows the removal of
the septostomy balloon along with the sheath, but leaves
a wire in place for reintroduction of the sheath and a
subsequent diagnostic catheter.
An angiographic catheter is introduced and pressure
measurements are recorded during a recorded with-
drawal from the left to the right atrium. This documents
the elimination of the gradient between the atria. The best
visual documentation of the atrial communication is with
an angiocardiogram. In patients with left atrial to right
atrial shunting, an injection is performed in a right upper
pulmonary vein with recording in a 45° left anterior–
oblique X-ray projection with cephalad angulation (i.e. 4-
chambered view). In patients with predominately right to
left shunt lesions, the angiocardiogram is performed with
an injection into the right atrial–inferior vena cava junc-
tion, and recorded in a straight lateral projection.
Echo-guided balloon atrial septostomy
Considerable enthusiasm has been generated for “echo-
only-guided” balloon atrial septostomies, which are
performed in the neonatal intensive care unit (ICU)
3
.
Originally, all balloon atrial septostomies were performed
in the cardiac catheterization laboratory during the diag-

nostic catheterization, when the underlying defect was
being diagnosed and evaluated. The majority of infants
are now diagnosed definitively by an echocardiogram in
the neonatal or nursery ICU, and many balloon atrial sept-
ostomies in neonates are performed in the ICU using
echocardiographic guidance only. When the septostomy
is performed in the neonatal ICU, usually one or more,
nurses or technicians from the catheterization laboratory
who are familiar with the equipment and the procedure
assist with the procedure. The catheterization laboratory
personnel provide a special balloon septostomy tray or
“cart” containing most of the essential equipment for a
routine septostomy. The special tray is brought to the
bedside from the catheterization laboratory. The sep-
tostomy trays or carts are maintained and restocked in the
catheterization laboratory after each use.
The echocardiogram provides an excellent picture of
the atrial septum and of the immediately surrounding
structures, but not all of the structures can be seen at the
same time. Before the septostomy balloon has been
inflated and while it is being manipulated, the septostomy
catheter is often seen only fleetingly, if at all. Echo views
do not provide even two-dimensional perspectives to the
structures to aid in catheter manipulations. As a conse-
quence, the maneuvering of the septostomy catheter to the
atrium and across the existing defect into the left atrium is
semi-blind and more difficult (and occasionally impossible)
under echo-only guidance without adjunct fluoroscopy.
A very slight inflation of the balloon during these manipu-
lations will help the visualization of the balloon and the

tip of the catheter on echo.
Once the septostomy catheter is in the left atrium and
the septostomy balloon begins to inflate, the balloon, the
septum and the surrounding structures are all seen very
well with echocardiography. The exact relationship of the
balloon size to the size of the left atrium and the exact loca-
tion of the balloon against the septum are seen vividly,
and their relative sizes are appreciated even better by echo
images than by fluoroscopy. The same technique for the
actual “ballooning” that is used in the catheterization lab-
oratory is used in the echo-guided procedure in the ICU.
Once the balloon pulls through the septum, the hole in the
septum and the adequacy of the septostomy are apparent
immediately on the echo. At the same time, the balloon
disappears completely from the echo field until searched
for by readjustment of the position of the echo probe. Even
with the balloon out of the field of view, it is still re-
advanced from the position of maximum pull until it can
be seen in the echo field. After a successful septostomy,
the inflated balloon can very easily float from the right
atrium into the right ventricle as it is being re-advanced
from out of the field. This position must be watched for
CHAPTER 13 Balloon atrial septostomy
387
very carefully on the echo before further maneuvering of
the balloon.
Echo-guided septostomy that is performed in the
neonatal ICU has the one advantage of not having to
transport a sick baby to the catheterization laboratory.
Depending upon the availability of staffing, this may

be more expedient overall as long as everything goes
smoothly. However, in the neonatal ICU there is no option
for changes in technique or equipment without incurring
a significant delay and disruption of the patient’s environ-
ment. The size of the sterile field for the procedure in the
neonatal ICU is always compromised. Unless the infant
already has an indwelling arterial line (umbilical or radial
artery) or the catheterization team introduces a separate
arterial line, the infant undergoing a septostomy in the
neonatal ICU does not always have the benefit of arterial
monitoring during the septostomy. In the cyanotic infant
the determination of which femoral vessel has been
accessed is more difficult by echo where all of the blood is
very “blue” and the wire positions in the thorax cannot be
checked easily.
Because of the limitations with the totally echo-guided
balloon atrial septostomy procedure, it is used mostly
for those infants who need a very temporary opening in
their septumai.e. those patients who are expected to have
surgery within a few days or weeks and who will be
dependent upon their mixing or shunting only until the
time of the surgery. Under these circumstances, it is more
reasonable to use the umbilical vein for access and to use
smaller septostomy balloons for the procedure. In a com-
plex heart where the patient will be dependent on the sep-
tal communication for months or years, balloon atrial
septostomy is performed preferentially in the catheteriza-
tion laboratory, where there is in a much more controlled
environment and more optimal results can be obtained.
Dilation of the atrial septum (dilation

septostomy)
The creation or enlargement of a communication in the
atrial septum with a dilation balloon is a balloon dilation
procedure that is not exactly a vascular dilation nor is
it considered a usual septostomy. Dilation balloons are
used for the creation or enlargement of atrial septal
defects as an alternative to the fairly crude, and often
poorly controlled, rapid withdrawal of inflated, standard
atrial septostomy balloons through the septum
4
. A dila-
tion septostomy is used to enlarge a pre-existing opening
or, after a transseptal puncture of the septum, to create an
entirely new opening. Dilation septostomies are particu-
larly efficacious in the presence of a very thick or tough
atrial septum. The shunt created by the atrial septal defect
can be right to left or left to right depending upon the
underlying hemodynamics and anatomy. When a dilation
balloon is used, one or more low-profile angioplasty bal-
loons that are significantly larger than the existing atrial
opening, is (are) inflated exactly within the atrial septum.
As the balloon(s) expand(s) beyond the diameter of the
original opening, it (they) theoretically and hopefully
tear(s) a larger opening in the septum. Because of the high
degree of control over the rate of inflation and, in turn, the
exact diameter of the opening created with dilation bal-
loons, some operators prefer dilation septostomy for all
balloon septostomy procedures, particularly in very small
and/or very sick infants. At the other end of the spectrum,
atrial septostomy performed with a dilation balloon is

particularly useful in older, larger patients with very
thick or tough interatrial septae, especially when used in
conjunction with a blade atrial septostomy. Blade sept-
ostomy with dilation septostomy is covered in detail in
Chapter 14.
Balloon dilation atrial septostomies are used to enhance
both right to left and left to right shunting at the atrial
level. A right to left shunt at the atrial level is created in
order to vent blood away from the right ventricle while
enhancing systemic output by allowing the shunted sys-
temic venous blood to enter directly into the systemic arte-
rial flow. In patients with hypoplastic right ventricles this
is the only means of egress of blood from the systemic
venous system, and the atrial opening should be as large
as possible. The creation of a vent for the right ventricle
is also desirable in patients who have a high and fixed
pulmonary vascular resistance and have no other intra-
cardiac communication. When an atrial defect is created
or enlarged in these patients, the right to left shunting
causes the net systemic cardiac output to increase, and the
systemic pressure is expected to increase, while there is a
concomitant drop in the systemic arterial saturation, and
the arterial oxygen saturation is expected to decrease.
In older patients, the goal is to reduce the systemic oxy-
gen saturation by only 8–10%. A sudden decrease of more
than 10% in systemic oxygen saturation is usually not tol-
erated by these patients. The balloon dilation septostomy
that is performed in order to enhance right to left shunting
is initially performed very conservatively, starting with
smaller (6–8 mm) balloons and then gradually increasing

the diameter of the balloons as tolerated by the patient.
If the systemic arterial oxygen saturation does not drop by
at least 6%, the original balloon is replaced with a balloon
two millimeters larger in diameter, and the dilation and
observation procedure repeated until the desired drop in
systemic oxygen saturation and, hopefully, increase in
cardiac output, are accomplished. This slow, progressive
dilation is necessary to prevent a catastrophic drop in sys-
temic oxygen saturation, particularly in older patients
with Eisenmenger syndrome.
When atrial septal dilation is performed to vent a high left
atrial pressure, it is often performed to relieve inoperable,
CHAPTER 13 Balloon atrial septostomy
388
obstructive, left heart lesions in order to allow trapped
pulmonary venous return to rejoin the general circulation.
Infants and children with congenital mitral valve stenosis,
mitral atresia and aortic atresia, all with some degree of
hypoplastic left heart, are the usual patients undergoing
dilation of an interatrial communication for this purpose.
In these patients the atrial defect represents the only (or
dominant) means of egress of blood from the pulmonary
venous circulation, and the defect created should be as
large as possible to eliminate all resistance to flow across
the septum. The initial balloon (or balloons) used to dilate
the septum for this indication should be as large as can
be accommodated by the length of the atrial septum.
Often these openings are maintained with the implant of
an intravascular stent in the atrial septal opening. This is
discussed in Chapter 14.

Patients with transposed arterial (TGA) or venous
(TAPVC) circulations require an interatrial communica-
tion to allow mixing of blood in both directions in order
to maintain systemic circulation and oxygenation and to
lower the pulmonary venous pressure. The larger these
communications can be made, the better are the venting
and mixing. In these patients dilation of the atrial septum
is performed with the largest diameter balloon(s) that the
septum can accommodate.
Procedure for atrial septal dilation
When an atrial septal defect is being created using a dilation
balloon alone, the balloon that is used should be consider-
ably larger in diameter than the desired size of the defect.
An end-hole catheter is advanced across the septum through
either the small existing atrial septal opening or through a
transseptal puncture through an intact atrial septum. From
the left atrium the catheter is manipulated out into a left
upper pulmonary vein and is replaced with a Super Stiff™
(Medi-Tech, Boston Scientific, Natick, MA) exchange
length guide wire with a short floppy tip. The wire should
be of the largest diameter that the balloon dilation catheter
that is proposed for the dilation will accommodate.
A dilation balloon that is approximately twice the dia-
meter of the desired defect is advanced over the wire and
positioned precisely straddling the atrial septum. The bal-
loon should be long enough to easily straddle the septum
but at the same time must not be long enough to become
entrapped in adjacent, narrow vessels. Usually a 2 cm
long balloon is used in infants and a 3 or even 4 cm long
balloon is used in older patients. All balloons used

to dilate the atrial septum should have very short
“shoulders” in order for the tip of the balloon not to ex-
tend into unwanted, adjacent vessels.
The balloon is inflated slowly while recording each
inflation–deflation angiographically in both PA and LAT
views. The balloon is observed closely not only for the
“waist” near the center of the balloon, but also to ensure
that no “second waist” appears on the balloon as a result
of the balloon being positioned erroneously in, and/or
that it does not move forward into the pulmonary vein or
backward into the inferior vena cava during the inflation.
When the central waist on the balloon disappear during
the inflation or the balloon reaches its maximum pressure,
it is deflated. The inflation–deflation is repeated, moving
the balloon forward or backward very slightly with each
re-inflation.
Often two balloons, positioned side by side across the
atrial septum, are used to perform the dilation septostomy.
By using two, smaller diameter balloons with their
smaller deflated profiles instead of a single very large
diameter balloon with a large rough deflated profile, a
large diameter atrial septostomy can be accomplished
using two much smaller introductory venous sheaths in
the peripheral veins as opposed to one very large dia-
meter sheath for the introduction of the large balloon. For
example two 12 mm balloons can be introduced through
two 6-French sheaths as opposed to the necessity of a large
8- or 9-French sheath for a single 18–20 mm balloon.
When two small balloons are used for dilation of the
atrial septum, each balloon individually has an inflated

diameter of the desired atrial defect. Ideally, in order to
create one large atrial defect, both deflated balloons
should pass through the same atrial communication ini-
tially. With any pre-existing communication, the second
end-hole catheter is maneuvered through the same open-
ing and then into the same or even a separate left pul-
monary vein and replaced with a second, identical, Super
Stiff™ wire. The two balloons are placed side by side
within the defect and the dilation septostomy carried out
similarly to the single-balloon technique.
When there is no pre-existing opening in the septum,
there are two alternatives for introducing the two wires
and then the two balloons through the same initial open-
ing. The simplest procedure is to perform an initial, partial
dilation of the septum with the first smaller balloon,
which is introduced and positioned across the atrial
septum after an initial transseptal puncture. This creates
a septal opening through which the second end-hole
catheter and eventually the second balloon can be intro-
duced immediately adjacent to the first balloon and in the
exact same opening. The second, slightly more compli-
cated alternative is to perform a second transseptal punc-
ture immediately adjacent to (touching) the first transseptal
catheter before either balloon is introduced. When per-
formed precisely, using biplane visualization, the second
puncture will almost join the first opening, and when
dilated with separate balloons the openings will coalesce
into one larger hole.
Following the inflation–deflation with either one or two
balloons, the angiograms are reviewed for the appearance

CHAPTER 13 Balloon atrial septostomy
389
and then disappearance of the waist(s) with the initial
inflation(s) and the absence or reappearance of the waist(s)
at the lowest initial pressures during the subsequent
inflations. When the septostomy is successful, the waist(s)
on the balloon(s) does (do) not reappear at all, or only
reappear very late at full inflation during the subsequent
balloon inflations. Once satisfied with the inflations, the
balloon(s) is (are) withdrawn from the septum over the
wire(s). The patient is allowed to stabilize following
the septal dilation, the hemodynamics are remeasured, and
an angiogram is performed to visualize the atrial defect
which has been created. A straight lateral view with injec-
tion into the right atrium is usually optimal for visualiza-
tion of defects that shunt right to left, while a left anterior
oblique view with cranial angulation and an injection into
the right upper pulmonary vein is optimal for visualiza-
tion of lesions that shunt left to right.
When the septum is totally intact before the dilation or
it is very tough, a blade septostomy (covered in detail in
Chapter 14) is performed initially before any “ballooning”
of the septum is attempted. This includes all patients who
require an atrial septostomy who are more than three or
four weeks old. Although the septum can be ballooned at
this age, ballooning by itself usually only stretches rather
than tears the septum, and only a temporary defect is cre-
ated. By first making a small incision in the septum with
the blade, less wall tension is necessary on the balloon to
create a tear and there is more control over the opening

created with the balloon. Small incisions in the septum
become the beginning of more controlled, linear tears
in the specific directions of the cuts as they are split or
torn further with the balloon dilation. A controlled tear
with the use of less force is accomplished instead of the
whole circumference of the opening in the septum being
stretched until it forcefully “explodes” in any direction.
Once an opening is created in the septum by stretching the
hole with a balloon septostomy alone, then a subsequent
blade septostomy only stretches and elongates the defect
to accommodate the width of the blade as it slides through
the opening, and does not incise the septal tissues at all. A
dilation septostomy following an initial blade septostomy
is very effective when a controlled diameter of the open-
ing is desired.
The hole in the atrial septum formed by dilation, even
with preceding blade incisions, as with the other atrial
septostomies, often has the same problem of not remain-
ing open. There are now several techniques for stenting
the opening in the septum to keep it open. These techniques
are described in Chapter 14 on “Special Septostomies”.
There is also a new fenestrated Amplatzer™ device,
which was developed for this problem and for creating
atrial septal openings of very specific sizes; this is also dis-
cussed in Chapter 14. This device has been used success-
fully in experimental animals and in a few compassionate
cases and appears very promising, but is not available yet
for routine clinical use in the United States.
Bioptome septostomy
As an alternative to any type of balloon septostomy for the

creation of an atrial defect in a patient with a very small
left atrium that cannot accommodate even the smallest
septostomy balloon, a biopsy forceps catheter using echo
assistance in the positioning of the bioptome jaws has
been used successfully for creating a large septal open-
ing
5
. Small pieces of the septum are “chewed” away from
the edge of the atrial septal opening with the “jaws” of a 5-
or 6-French bioptome catheter. This has been particularly
useful in hypoplastic left heart patients, where the left
atrium is usually very small and the septum is very tough.
The biopsy forceps is advanced to the right atrium,
opened in the right atrium and, using both biplane fluoro-
scopic and echo guidance, purposefully engaged on the
edge of the existing atrial septal opening. A simultaneous
2-D echo image of the edge of the atrial septum and the
bioptome forceps is essential for the final precise position-
ing of the open bioptome jaws on the edge of the septal
opening before each “bite”. Once engaged on the edge of
the septal opening, small pieces of the edge of the original
septal opening are “nibbled away” by a series of bites
with the bioptome. Biopsy forceps septostomy is co-
vered in detail in Chapter 14, on blade and biopsy forceps
septostomy.
Complications of balloon atrial
septostomy
In spite of the crude and rough nature of the balloon atrial
septostomy procedure, there are surprisingly few compli-
cations from it when it is properly performed. Virtually

all patients have a significant bradycardia and drop in
systemic pressure immediately following the forceful
withdrawal of the balloon. Almost all of them recover
spontaneously and rapidly. For those few who have a sus-
tained, or even progressive bradycardia or hypotension, a
dose of atropine and epinephrine appropriate for the
patient’s weight is given intravenously. In order for this
to be administered expeditiously, the appropriate doses
of atropine and epinephrine are drawn up and ready to
administer before the actual septostomy procedure is
even begun for every patient undergoing an atrial sept-
ostomy. Unless there is an associated major catastrophic
injury during the septostomy, all of the patients recover
from the “vagal-like” response with atropine, epinephrine
and, rarely, a few external compressions of cardiac massage.
Catastrophic major tears due to gross malpositioning of
the balloon before the withdrawal have been discussed.
CHAPTER 13 Balloon atrial septostomy
390
Abnormal initial balloon positions and how to recognize
them and prevent the catastrophe are fairly straightfor-
ward. There seldom, if ever, are minor tears during an
atrial septostomy. Once a major tear of an appendage, an
atrioventricular valve or an atrial wall occurs, the result is
a massive blood loss and a rapid demise of the patient.
The only real treatment of these tears is prevention.
Prevention is possible in almost all cases by meticulous
attention to every detail of the procedure no matter how
mundane it seems, nor how many septostomies have
been performed.

The one major tear that might not be preventable, is a
tear at the junction of the right atrium and inferior vena
cava from the balloon’s being pulled into the inferior vena
cava at the end of the forceful jerk, particularly through a
bulging or prolapsing septum. Judging the precise dis-
tance and controlling the force and distance of the pull
on the balloon are often difficult and, unfortunately, are
learned only by experience. Fortunately, the inferior vena
cava is somewhat compliant or forgiving and “accepts”
the inflated balloon for a considerable distance. When a
tear occurs at this location, it is in a lower-pressure venous
area and often is not as extensiveaand therefore not as
catastrophicaas a tear within the heart itself. When recog-
nized, the infant is supported with volume replacement
and observation while preparing to move them to the
operating room. If deterioration continues rapidly, an
attempt is made at tamponading the area with a very-
low-pressure inflation of the septostomy balloon or an
equivalent diameter angioplasty balloon positioned just
caudal to or immediately adjacent to the suspected area
of tear.
Balloon septostomy, like the many other interventions
on the systemic side of the circulation, involves consider-
able manipulation in the blood that reaches the systemic
circulation directly. Most newborns do not receive sys-
temic heparin for a balloon atrial septostomy procedure,
so extra attention must be paid to keeping catheters and
sheaths flushed and not leaving them in place for long
periods of time. The atrial septostomy procedure involves
the introduction of balloon catheters into sheaths where

air and/or clot easily become trapped, with the ever-
present potential for systemic embolization if either the air
or thrombi are pushed out of the sheath. Again, the only
really effective treatment is recognition of the potential
and meticulous attention to the details for prevention.
Several complications are unique to the balloon atrial
septostomy procedure. Rupture of the septostomy bal-
loon with or without the loss of the balloon material has
occurred in the past, and probably will occur again in the
future
6
. This problem is less common with the current
universal single use of balloon septostomy catheters and
the newer Miller™ and NuMED™ septostomy balloons.
When a septostomy balloon that has been properly
emptied of air prior to inflation with fluid, ruptures but
does not lose part of the balloon, it creates an inconveni-
ence, but does not create a real problem or complication.
The ruptured balloon occasionally does not deflate com-
pletely and is more difficult to withdraw, but its with-
drawal can be accomplished by withdrawing the sheath
along with the septostomy catheter. A wire is introduced
through the valve in the sheath adjacent to the shaft of
the balloon catheter, through the sheath and advanced
into the central circulation before the original sheath is
withdrawn. This preserves the venous access for the intro-
duction of a new sheath.
When a septostomy balloon ruptures and all or part of
the balloon material embolizes, it potentially represents
a major catastrophe, but amazingly, usually does not

create even a small problem. The balloon material is not
radio-opaque. The only indicators of the location of the
embolized particle are the signs or symptoms of a vascular
occlusion somewhere in the patient, depending upon
which organ the piece of balloon lodges in. Any symp-
toms are treated supportively. Again, prevention by
inspection of the balloon and its specifications before
using it is the best treatment for balloon rupture. If the bal-
loon catheter is old, past its labeled expiry date, or has
been stored in an unfavorable environment, rupture with
disruption of the balloon is more likely and the balloon
should not be used.
A balloon septostomy catheter shaft can break off and a
piece of the catheter embolize with the entire balloon
7
. The
piece of catheter is radio-opaque and has a radio-opaque
marker that allows the location of the errant piece to be
ascertained. In the one reported case of such an incident,
the piece lodged in a systemic artery and was retrievable
with a catheter retrieval (snare) device. This apparently
was an isolated incident and hopefully was a quality
assurance problem that will not recur.
Another problem, which is a problem of balloon catheters
in general and not unique to septostomy balloons, is the
failure of the balloon to deflate after inflation for a sept-
ostomy or dilation. Contrast in the balloon catheter lumen
which is not diluted sufficiently, or during a prolonged
procedure when the contrast remains in the catheter
lumen for a long time, both contribute to, but are probably

not the only causes of, this problem. Assuming that the
balloon is in the right atrium and not lodged in a vital
structure or critical orifice, the failure of the septostomy
balloon to deflate does not compromise the hemodynam-
ics, but does create a problem of balloon removal. There
are several alternative means of overcoming a “non-
deflate” problem. The first means of deflating the balloon
is to use the stylet that comes with the balloon and try
cleaning out the lumen of the catheter by passing the
moistened stylet repeatedly in and out of the catheter
lumen and all of the way to the balloon. Usually the
CHAPTER 13 Balloon atrial septostomy
391
balloon does deflate very slowly after this reaming out
with the stylet.
On the rare occasions that the balloon does not deflate
using a stylet, it must be punctured externally. This is
accomplished using a second catheter introduced through
a second vein. A standard transseptal dilator (alone) as the
second catheter and used with a transseptal needle pro-
vides the best control as a puncturing device. A small end-
hole catheter along with a Mullins™ wire with the safety
tip bead cut off, also functions similarly and may allow the
use of a smaller second catheteraparticularly in small
infants. Once the second catheter has been introduced and
advanced almost to the right atrium, the non-deflated sept-
ostomy balloon is withdrawn into the right atrial–inferior
vena caval junction in order to stabilize the balloon in the
side-to-side direction. The septostomy balloon is pulled
tightly into the inferior vena cava, which fixes it in a fairly

stable, non-mobile, position in this location while the sec-
ond, puncturing catheter approaching from the inferior
vena cava is directed at the caudal surface of the balloon.
When the catheter tip is against the balloon in both planes
of the fluoroscopy, the needle or sharp wire is extruded
from the end of the catheter and poked into the balloon. This
maneuver is repeated until the balloon has been punctured.
If the fully inflated balloon that will not deflate becomes
trapped in a critical location in the course of the blood
flow (the tricuspid valve, right ventricular outflow tract
or mitral valve), there is usually no time to introduce the
second catheter, or the location of the balloon prohibits
access for a puncture. The only recourse is to over-inflate
the balloon and rupture it deliberately! For Miller™
balloons, this requires 12–13 ml of fluid. This, in itself, is a
potential hazard to the infant but when cardiac output is
totally blocked by the inflated balloon, there is no alterna-
tive. This emphasizes the importance of always knowing
the location of the inflated septostomy balloon and the
prevention of its moving into critical locations.
Meticulous attention to the details of the septostomy pro-
cedure prevents most, if not all, of the previously mentioned
complications and remains the best treatment for them.
References
1. Rashkind WJ and Miller WW. Creation of an atrial septal
defect without thoracotomy; a palliative approach to complete
transposition of the great arteries. JAMA 1966; 196: 991–992.
2. Abinader E, Zeltzer M, and Riss E. Transumbilical atrial sept-
ostomy in the newborn. Am J Dis Child 1970; 119(4): 354–355.
3. Perry LW et al. Echocardiographically assisted balloon atrial

septostomy. Pediatrics 1982; 70(3): 403–408.
4. Mitchell SE et al. Atrial septostomy: stationary angioplasty
balloon techniqueaexperimental work and preliminary clin-
ical applications. Pediatr Cardiol 1994; 15(1): 1–7.
5. Boucek MM et al. Management of the critically restrictive atrial
septal defect awaiting infant cardiac transplantation. Am J
Cardiol 1992; 70: 559.
6. Vogel JH. Balloon embolization during atrial septostomy.
Circulation 1970; 42(1): 155–156.
7. Akagi T et al. Torn-off balloon tip of Z-5 atrial septostomy
catheter. Catheter Cardiovasc Interv 2001; 52(4): 500–503.
392
Blade atrial septostomy
The blade atrial septostomy catheter and procedure, as the
name implies, utilize a blade to incise an opening in the
atrial septum. It was designed to create an incision in a
resistant or tough atrial septum using the blade to initiate
the creation of a larger opening in the septum. The proced-
ure and the unique blade catheter were developed by Dr
Sang Park in collaboration with Cook Inc. (Bloomington,
IN)
1
. For introduction into the heart, the blade is recessed
in a metal housing or “pod” at the distal end of the
catheter. Once across the septum, the blade is extended
out of the metal housing and withdrawn across the sep-
tum to create the incision. A standard Rashkind™ balloon
atrial septostomy or a dilation septostomy is used to
extend or tear the incision created by the blade.
Indications for a blade atrial

septostomy
The indications for a blade atrial septostomy are the same
as those for a balloon atrial septostomy, however the
blade procedure is used in older patients or in infants
with a particularly tough septum. Any patient who has
a restriction in the communication between the right and
left atria where access of the systemic and/or pulmonary
venous blood back into the functional circulation is essen-
tial, requires the creation or enlargement of an opening
between the two atria. Often such patients do not present
clinically until two to three months or even many years of
age. There are six general categories of these older congen-
ital heart patients who present from weeks to years after
the newborn period, and who are often in desperate need
of a new or enlarged atrial communication that usually
requires a blade incision to initiate the opening. There is
also a growing group of older patients with acquired car-
diac problems who benefit from the creation of an atrial
communication and require a blade incision to initiate
catheter atrial septostomy.
1 The first group of patients who require the creation
of an interatrial communication are those with transposi-
tion of the great arteries, who have insufficient mixing of
systemic and pulmonary venous blood with resultant
hypoxemia or pulmonary over-circulation/congestion.
These patients fall into two subgroups. The first comprises
those with an intact ventricular septum without an atrial
communication or with an atrial communication that is
too small. Patients with an intact ventricular septum or a
small or non-existent atrial septal defect have no mixing of

systemic and pulmonary blood and usually present very
early or are very ill. These infants need an atrial commun-
ication created as an emergency regardless of what surgery
is planned for them subsequently.
The second subgroup of transposition patients are those
with transposition of the great arteries and an associated
ventricular septal defect. These patients either have a pre-
existing atrial septal defect that is too small or, more often,
no atrial communication at all
2
. Patients with transposi-
tion of the great arteries, a ventricular septal defect, and an
inadequate or absent atrial defect, all need a larger atrial
communication for free atrial mixing, but usually not as
urgently, and often they do not present for this procedure
until several months of age. These patients with only a
large ventricular communication “trap” a large percent-
age of their blood in their pulmonary circulation, develop
pulmonary congestion and pulmonary hypertension and,
in short order, pulmonary vascular disease unless they
have an adequate interatrial communication created.
2 The second major group of patients who require a
blade septostomy are those with left atrioventricular valve
stenosis/atresia and an associated underdeveloped or
hypoplastic left ventricle
2
. Regardless of other intracardiac
communications, without an adequate atrial septal open-
ing, the pulmonary venous blood is trapped in the left
atrium, creating very high pulmonary venous pressures,

preventing adequate systemic oxygenation, and leading
14
Blade/balloon atrial septostomy,
special atrial septostomies, atrial
“stent septostomy”
CHAPTER 14 Blade and special atrial septostomies
393
to acidosis. Usually the existing atrial communication in
hypoplastic left heart patients is very small, making the
creation of an opening in the atrial septum an urgent pro-
cedure. Even though these patients are very young with
very small left atria, they usually have very tough inter-
atrial septae which require some type of blade or other
special septostomy.
3 The third group of patients is those with pulmonary
valve or tricuspid valve atresia/hypoplasia with hypo-
plastic right ventricles, in whom the systemic venous return
is trapped in the right atrium by the absence ofaor a small
aatrial communication
3
. In the presence of either a very
tiny right ventricular cavity or persistent very high right
ventricular end-diastolic and systolic pressures, the right
ventricle is not capable of accommodating an adequate
volume of systemic venous return with each diastole. In
turn, the very small or restrictive right ventricle does not
pump an adequate volume through the lungs to the left
ventricle to supply an effective systemic output. Without
an adequate vent at the atrial level, the systemic venous
return remains in the right atrium, the right atrium

becomes massively dilated with stasis of systemic venous
return, and the cardiac output remains low. Under these
extreme circumstances, balloon or blade and balloon atrial
septostomy is performed during the initial diagnostic
catheterization.
When there is a possibility that the right ventricle will
be able to sustain cardiac output once the pulmonary
stenosis/atresia has been relieved, the atrial septum is left
intact during the initial catheterization. Once the infant
has stabilized after the valve perforation/dilation with or
without stenting of the ductus or a systemic to pulmonary
shunt, then clinical assessment will determine the need for
a subsequent atrial septostomy or other intervention.
If the systemic output remains low or the right atrium
and liver become distended and enlarged, the atrial sept-
ostomy can be performed hours, days or weeks later.
If however, the atrial septostomy was performed at the
initial catheterization when, in fact, an elevation in right
atrial pressure would be essential to augment the right
ventricle filling, the potential growth of the right ventricle
could be compromised permanently by the adequate
atrial communication created by the septostomy. In each
individual case of hypoplasia of the right ventricleaand
until more data are available on which ventricles will
grow with adequate stimulus and with what type of
stimulusa the decision as to when to create the septost-
omy remains an-on-the-spot judgment decision for each
individual case in the catheterization laboratory.
4 The fourth group of patients who benefit dramatically
from an atrial septostomy are those with total anomalous

pulmonary venous connection (TAPVC). In TAPVC both
the systemic and pulmonary venous blood returns to the
systemic venous circulation and is restricted from return-
ing to the systemic circulation for systemic cardiac out-
put
4
. All of these patients need, or benefit markedly from,
a large interatrial communication. Unless these patients
have associated, separate obstruction of the pulmonary
veins, they present much later in life (weeks, months or
even years!) and a blade septostomy is required to initiate
the atrial septostomy. When the atrial septum is wide
open and the pulmonary veins are non-obstructed, the
clinical presentation of patients with TAPVC can be very
similar to that of patients with a large atrial septal defect!
5 The fifth group of older congenital heart patients who
would benefit from an atrial septostomy are those with
pulmonary vascular disease and severe pulmonary hyper-
tension
5
. An atrial septal defect in these patients allows a
finite amount of systemic venous blood to shunt from
right to left at the atrial level. The right to left shunt vents
or decompresses the right heart volume/pressure slightly
through the atrial communication and also supplements
the blood volume for their systemic cardiac output, as this
systemic venous blood is added to the pulmonary venous
blood in the left atrium. The admixture of systemic venous
blood in the systemic circulation creates some systemic
desaturation.

6 The sixth and final group or category of congenital heart
patients who benefit from or require the creation of a
special atrial communication are those who have under-
gone a single ventricle cavopulmonary repair and who for
one reason or another are “failing”, with resultant low
systemic cardiac output or refractory protean loosing
enteropathy. These patients frequently receive consider-
able symptomatic relief by the creation of a small restrict-
ive interatrial communication. The tissues are always
tough and always require some special procedures both to
create and to maintain a septostomy. The interatrial com-
munication usually increases systemic cardiac output, but
this is always at the expense of some systemic desatura-
tion. The size of this opening is critical in order to create
the optimal balance between increased cardiac output
and some desaturation, but without causing symptoms
of hypoxia or even resulting in the death of the patient.
As already mentioned, there is also a group of older
patients with acquired cardiac problems who would
also benefit from an atrial communication. These are the
patients with acquired severe progressive right or left
heart failure where blood from the failing side of the heart
needs to be vented acutely into the better functioning
or supported side of the circulation. The creation of an
atrial septal communication is required to vent the left
heart in patients on extracorporeal membrane oxygena-
tion (ECMO), where left-sided over-circulation and dis-
tention can be a significant problem
6,7
. Venting the right

heart with an emergent septostomy following acquired
acute right heart decompensation has also been done
8,9
.
Almost all patients requiring ECMO are older, and in