Cardiac Catheterization in Congenital Heart Disease: Pediatric and Adult - Part 8 doc

Once the catheter is in the distal pulmonary artery, the original floppy, torque-controlled, guide wire that was used to manipulate across the ductus is replaced with the stiffest possibl

over the wire If an adequate angiogram cannot be

obtained by either of these techniques, a second, small

ret-rograde catheter should be introduced A freeze-frame

“road map” of this angiogram of the ductus is stored for

reference use

After the angiogram has been obtained, an end-hole

catheter with a smooth tip is advanced over the wire

very gently, through the ductus and into the distal branch

pulmonary artery Once the catheter is in the distal

pulmonary artery, the original floppy, torque-controlled,

guide wire that was used to manipulate across the ductus

is replaced with the stiffest possible exchange length

guide wire that the delivery balloon/stent that will be

used to deliver the stent to the ductus, will accommodate

The exact length of the stent used in the PDA is

deter-mined from the angiogram of each particular ductus

Ideally the stent must cover, and extend slightly past both

ends of the ductus but, at the same time, should not extend

too far into the lumen of either the pulmonary artery or

the aorta Once the delivery wire is secure through the

ductus and well into a distal pulmonary branch, the

sec-ond angiographic catheter is positioned adjacent to the

aortic end of the ductus The correct stent/balloon is

chosen and prepared on the catheterization table and the

prostaglandin infusion is stopped The patient is observed

closely for at least 30 minutes, and assuming no acute or

severe deterioration of the patient during that time, the

aortogram is repeated Usually the ductus constricts and

changes in configuration rapidly once the infusion of

prostaglandin is stopped Occasionally the rebound

con-striction of the ductus is very severe and very rapid which,

in turn, necessitates either the reinstitution of the

pro-staglandin infusion or the rapid introduction of, and the

implant of, the stent The ductus tissues are friable, so

the balloon/stent introduction must be very gentle and

very precise, and the stent must pass easily through the

ductus Because of these factors, some operators prefer to

prepare the balloon stent and position the stent/balloon

across the ductus before stopping the prostaglandin

infu-sion and then wait for the prostaglandin to wear off with

the deflated balloon/stent already in place across the

duc-tus The second catheter adjacent to the aortic end of the

ductus is essential at this stage of the procedure in order to

allow very rapid and precise final positioning of the stent

within the ductus just before deployment

With any deterioration of the patient, or after 30

min-utes, the angiogram is repeated, the stent/balloon

posi-tion adjusted, and the balloon/stent is inflated to its

advertised pressure or until all indentations in the

bal-loon/stent have disappearedawhichever comes first The

balloon is deflated immediately and rapidly With a

pro-perly implanted stent, the patient immediately will

oxy-genate better and stabilize as soon as the balloon is

deflated The balloon is withdrawn carefully out of the

stent/ductus and with the wire still through the stent andductus, the aortogram is repeated to visualize the ad-equacy of the coverage of the stent in the ductus and the

new flow to the pulmonary arteries If there are any areas

of the ductus, particularly at either end of the ductus,which are not covered by the stent, a second (or more)stent(s) is/are implanted during the same catheterization.The additional stent(s) is/are placed overlapping the first

stent and covering all of the ductal tissues entirely Any

“exposed” ductus tissue is notorious for constricting andclosing completely, even with a stent in the remainder ofthe ductus For the implant of an additional stent immedi-ately after the implant of the first stent, all of the cathetersand wires are already in place and minimal furthermanipulation or time in the catheterization laboratory isnecessary for the implant of an additional stent Once theductus is covered adequately and the patient is stable, thedelivery wire is withdrawn from the stent and the catheterwithdrawn The infants are maintained on 21 mg aspirinper day

Stenting of the ductus in patients with pulmonary atresia with intact ventricular septum

In infants with pulmonary atresia and intact ventricularseptum, along with the opening of the pulmonary valve, asystemic to pulmonary “shunt” can be performed in thecatheterization laboratory by implanting a stent in theductus arteriosus The stenting of the ductus is addressedafter the pulmonary valve has been perforated and dilatedsuccessfully This allows the delivery and implant of thestent into the ductus arteriosus through a venous routeand the use of a smaller catheter system in the artery.When a stent is to be implanted in the ductus during thecatheterization, before any intervention on the ductus and

if not administered earlier when the lines were lished, the patient is administered heparin systemically.After the pulmonary valve has been perforated anddilated, the final balloon dilation catheter for the pul-monary valvotomy is removed over the guide wire.Following the perforation and balloon dilation of anatretic pulmonary valve, a long guide wire usually hasalready been advanced through the ductus creating

estab-a “through-estab-and-through” route from estab-a femorestab-al vein,through the right heart and ductus, down the descendingaorta and out through a femoral artery sheath/catheter

If the through-and-through “rail” wire was not lished during the valve perforation/dilation, the “rail” isestablished at this time with an exchange length wirewhich the proposed balloon catheter for stent deliverywill accommodate The balloon dilation catheter used forthe pulmonary valve is replaced with a long 4- or 5-Frenchend-hole catheter, which is advanced through the venoussheath, over the original guide wire and into the pulmonary

estab-artery or ductus (wherever the end of the wire is

posi-tioned) If the tip of the prograde venous catheter is in the

pulmonary artery, it is manipulated through the ductus

into the descending aorta and maneuvered through and

out of the femoral artery sheath with the help of a

torque-controlled wire Occasionally the prograde catheter in the

pulmonary artery cannot cross the ductus easily into the

descending aorta In that circumstance a floppy tipped

wire and then a snare catheter is maneuvered from the

aorta, through the ductus and into the pulmonary artery.

A soft tipped wire is advanced through the venous

catheter prograde into the main pulmonary artery and is

snared with a Microvena™ snare (ev3, Plymouth, MN),

which has been passed through the retrograde catheter

into the main pulmonary artery The snared prograde

wire and the prograde venous catheter are withdrawn

through the ductus and out through the femoral arterial

sheath with the retrograde snare Once both ends of the

catheter are available outside of the body, the original

floppy tipped exchange wire is removed and replaced

with a stiff exchange wire of the maximum wire diameter

that the particular coronary stent/balloon catheter or

other pre-mounted stent/balloon that will be used for

implanting the stent in the ductus, requires This wire is

secured outside of the body at both the arterial and

venous ends

An equally effective alternative is to advance the floppy

tipped wire from the retrograde catheter through the

duc-tus and into the main pulmonary artery and to introduce

the snare catheter into the pulmonary artery from the

venous route The retrograde wire then is withdrawn

through the right heart and out through the venous sheath

to complete the through-and-through wire

Even when previously visualized very adequately, a

repeat aortogram with the injection adjacent to the ductus

is performed either through a Tuohy™ adapter attached

to the hub of the catheter over the through-and-through

wire or through a separate prograde or retrograde

catheter positioned in the descending aorta adjacent to the

ductus This angiogram of the ductus is imperative

because of spasm or distortion of the ductus caused by the

various wire/catheter manipulations through it The

diameter and length of the ductus are remeasured very

accurately As with the implant of a stent in any other

duc-tus, the goal is to “line” the entire lumen of the ducduc-tus,

including covering both ends of the ductus with the stent

Similarly to the patients with pure ductus-dependent

pul-monary circulations, a 4 mm diameter stent is used in

these patients Once the through-and-through wire is

established and the appropriate balloon/stent for the

particular ductus is chosen and prepared for delivery, the

prostaglandin infusion is discontinued Assuming no

acute or sudden deterioration in the infant’s saturation/

hemodynamics, the infant is observed for at least

30 minutes while off the prostaglandin infusion and theaortogram repeated Often the combination of the irrita-tion of the through-and-through wire/catheter, the pre-ceding manipulations during the balloon dilations of the valve, and the discontinued prostaglandin infusiondistorts the ductal anatomy significantly The ductus anatomy is re-examined carefully and the measurementsrepeated

If the ductus acutely goes into spasm or the infant riorates significantly when the prostaglandin infusion isstopped, the prostaglandin infusion is restarted and thestent is delivered to the ductus before re-stopping theprostaglandin as described previously

dete-The pre-mounted stent on a 4 mm balloon dilation

catheter is introduced over the venous end of the

through-and-through wire and advanced over the wire, throughthe right ventricle, pulmonary valve and into the ductusarteriosus Although small stents notoriously are poorlyvisible, or even invisible in large adult patients, in smallinfants they are seen clearly in both the collapsed and theexpanded states Once the balloon/stent combination is inposition in the ductus, and before the stent is expanded inthe ductus, the anatomy of the whole ductus and, in par-ticular, the exact location and diameter of the areas of min-imal ductal diameter are re-imaged angiographically Ifthere is not a second catheter in the aorta, the aortogram isaccomplished by advancing a 4- or, preferably, a 5-French,end and side-hole catheter retrograde over the arterial end of the through-and-through wire The catheter isadvanced just to the aortic end of the ductus and a pres-sure, Tuohy™ “Y” adaptor is attached over the wire at theproximal end of the retrograde catheter An angiogram isperformed over the wire through this retrograde catheter,the tip of which should be adjacent to the ductus and thestent/balloon catheter coming from the other end of thewire With the anatomy precisely identified and recorded,the retrograde catheter is withdrawn back into the des-cending aorta as the stent/balloon catheter is positionedexactly in the ductus

The stent should cover both ends of the ductus

com-pletely including, in particular, the area of the narrowestportion of the ductus, which usually is at the pulmonaryend of a long tortuous ductus When satisfied with thestent length and location, the stent is expanded byinflation of the balloon to its full diameter or to the recom-

mended maximum pressure of the balloonawhichever comes firstafollowed by a rapid deflation A repeat

angiogram is recorded with injection through the aortic

catheter before the deflated balloon is withdrawn from the

stent positioned in the ductus When satisfied that thestent is fully inflated and fixed in the ductus, the deflatedballoon is withdrawn cautiously out of the stent, over thewire and out of the body The retrograde catheter is re-advanced to the aortic end of the stent and a final repeat

angiocardiogram is recorded through this catheter before

the wire is withdrawn Similarly to the other stents in the

neonatal ductus, if there is an area of the ductus which

is not covered completely, a second, overlapping stent

should be implanted at that time

When satisfied with the position, the adequacy of

expansion of the stent and the “lining” of the entire ductus

by the stent(s), an end-hole catheter is introduced over the

venous end of the wire and advanced into the pulmonary

artery The wire is carefully withdrawn out of the stent

through the venous catheter

The hemodynamics and anatomy are carefully

re-assessed and a decision is made whether an atrial

sept-ostomy is to be performed Hemodynamically, an atrial

communication in the presence of some elevation of the

right atrial pressure enhances left ventricular filling and

systemic output, albeit with some systemic desaturation

and at the expense of some forward flow through the right

ventricle/pulmonary arteries On the other hand, with at

least a moderate sized right ventricle, the right ventricular

pressure near normal, and even in the presence of

signi-ficant tricuspid valve regurgitation, a restrictive atrial

communication and the associated elevated right atrial

pressure should enhance right ventricular filling and, in

turn, encourage forward blood flow through the right

ventricle, pulmonary artery, and lungs Until more

definitive data are available on which ventricles will grow

with adequate flow, and with what type of stimulus, this

remains an on the spot judgment decision, which must be

made in the catheterization laboratory, during each

indi-vidual case

Stenting of the ductus in the hypoplastic left

heart syndrome

The patent ductus is essential for the systemic output in

the infant with severe left heart obstructive lesions and,

particularly, an associated “hypoplastic left heart”

syn-drome In most cases, the open ductus is maintained with

prostaglandins until the infant undergoes the first stage

“Norwood” surgical palliation In that surgery, the entire

ductal tissue is excised purposefully and widely when

the distal pulmonary artery is anastomosed to the aorta

and no catheter intervention on the patent ductus should

be considered when a “Norwood” surgical palliation is

considered

An alternative approach to the “Norwood” and “single

ventricle” approach to the patient with a hypoplastic

left heart is an orthotopic cardiac transplant However,

when the infant is “listed” and awaiting the transplant,

the ductal patentcy must be maintained until the

trans-plant is performedaoften for weeks or months To

main-tain the patency of the ductus medically requires a

continuous, precisely controlled, intravenous (IV) infusion

of prostaglandin The maintenance of the IV and the precise control of the rate of the prostaglandin in aneonate require a 24/7 neonatal intensive care environ-ment Even in this environment the infant is in a very pre-carious situation An alternative technique is to maintainthe ductal patency with an intravascular stent20,21.Once the decision is made to “list” the patient for atransplant, then the implant of a stent in the ductus should

be considered immediately The longer the patient waits,the lower the pulmonary resistance becomes and thesicker the infant becomes The longer the patient remains

on prostaglandin, the greater the likelihood of systemicinfection and the more friable the ductal tissues become.The infant with a “hypoplastic left heart” who is toundergo the implant of a stent in the ductus is taken to thecatheterization laboratory with the ductus patency main-tained with the prostaglandin infusion The infant is intub-ated and ventilated on 17–18% oxygen If the patient doesnot have an indwelling arterial line, a femoral artery iscannulated with a 20-gauge Quick-Cath™, and a multi-purpose, end and side-hole catheter is introduced through

a sheath in the femoral vein This catheter is manipulatedthrough the right ventricle to the pulmonary artery and tothe ductus arteriosus A biplane angiogram is recorded

in the PA and lateral views with the injection directly intothe ductus This angiogram is to visualize the exact dia-meter, length and configuration of the ductus If necessary,the X-ray tubes are re-angled to “cut the ductus on edge”more precisely and the angiogram repeated Precise meas-urements are made of the length and diameter of the duc-tus from the views that elongate the ductus optimally.Freeze-frame images of the desired views are stored for use as “road maps” during the implant of a stent intothe ductus

The end and side-hole catheter is advanced through theductus into the distal descending aorta A 0.018″ or 0.035″exchange length guide wire (depending upon which stentand which balloon catheter is to be used), is advanced farinto the distal aorta, the tip of the wire is fixed in the ilio-femoral artery, and the catheter is removed over the wire

If there appears to be any distortion of the pulmonaryartery–ductus–descending aorta anatomy by the wire, arepeat angiogram is recorded over the wire and throughthe catheter in the ductus before the catheter is removed.This angiogram performed with the catheter over the wire is accomplished by injecting the contrast through

a Tuohy™ high-pressure “Y” adaptor attached to the hub

of the catheter while the catheter is still positioned overthe wire

A stent is chosen that is as large in diameter as the tus/descending aorta can accommodate and long enough

duc-to extend completely through the ductus The expandedstent must extend from well within the pulmonary artery,through the ductus, to well into the descending aorta

Unlike the stents in the ductus for pulmonary atresia

patients, where a very controlled flow using a small stent

(3– 4 mm) is desired, in patients with hypoplastic left

heart, the stent must be large enough to accommodate all

of the cardiac output and not create any resistance to this

flow because of restriction from a limited diameter of the

stent Depending upon the patient’s size, this usually

requires an 8–12 mm diameter stent The pre-mounted

standard “Large” Genesis™ stents (Johnson & Johnson–

Cordis Corp., Miami Lakes, FL) are satisfactory for this

use up to 10 mm in diameter The Genesis™ stents are

available pre-mounted in diameters up to 9 mm and in

lengths of 19, 29 and 39 mm These stents can be

intro-duced through a short, 6-French sheath and advanced

over a 0.035″ guide wire without the necessity of

pre-positioning a long sheath across the lesion “Large”

Genesis™ stents are very flexible and conform to the

cur-vature of the ductus None of the coronary artery stents

are suitable for this use because of the maximum diameter

of 4 to 5 mm The Express™ Biliary LD stents (Medi-Tech,

Boston Scientific, Natick, MA) are available pre-mounted

in similar sizes, but have larger side “cells” as a

conse-quence of their open-cell design, which may be a problem

for any use in the ductus because of the propensity for

rapid ingrowth of the ductal tissues through any space

Once the wire is in place and the stent is ready for

deliv-ery, the prostaglandin infusion is stopped The stent is

advanced over the wire and positioned across the ductus

The balloon/stent is maintained in this position while

monitoring the patient’s distal arterial pressure and

sys-temic saturation for 30 minutes or until the patient’s

hemodynamics begin to deteriorate The idea is to allow

the large diameter, often patulous, ductus to constrict

enough after the prostaglandin is stopped to hold the

stent in place The blood pressure and systemic saturation

decrease as the prostaglandin wears off and the ductus

begins to close Unless a second venous catheter has been

introduced into the pulmonary artery or the stent/balloon

was delivered to the ductus through a long sheath, an

angiogram in the pulmonary artery to verify the status of

the ductus and the stent position is not possible at this time

After the 30 minutes or with any deterioration of the

patient, the stent’s position is compared to the

freeze-frame “road map” images of the ductus, and when in the

appropriate position, the balloon is inflated slowly to

deploy the stent precisely in the ductus Once fully

inflated, the balloon is deflated rapidly The large

dia-meter stent should fix very securely in place in the ductus

The inflation/deflation is repeated one, or more, times

and then the balloon is removed from the body over

the wire

An end- and side-hole catheter is advanced over the

wire and into the stent An angiogram is recorded within

the ductus (stent), injecting over the wire with the use of a

Tuohy™ “Y” connector on the catheter If the stent is notfully expanded, if it is at all unstable, or if there are areas ofductal tissue that are “exposed” or “uncovered” by thestent, a further dilation of the stent or the implant of anadditional stent will be necessary to cover the ductal tis-sue completely If a second stent is necessary and the wire still is in place, the additional stent can be implantedduring this same procedure without any significant further or repeated manipulations through the freshlyimplanted stent

Once satisfied with the location and expansion of the stent(s), the wire is removed from the catheter and

the catheter withdrawn The infant is maintained off the

prostaglandin infusion Depending upon the status of thepulmonary vascular resistance, the infant usually can

be discharged and observed as an outpatient until a donor heart becomes available Even with a very satisfac-tory stent in the ductus, the infant still needs very close follow-up As the pulmonary vascular resistance fallssignificantly, the infant’s lungs can become flooded, with

a proportionate decrease in the effective systemic cardiacoutput through the ductus

The presence of the stent in the ductus does not interferewith a cardiac transplant procedure since the stented duc-tus and the adjacent tissues are removed at the time of the

transplant The main disadvantage to stenting the ductus in

an infant with hypoplastic left heart syndrome is when a

donor heart does not become available of for some other

reason, the decision for the course of treatment is reversedand a “Norwood” procedure is required A large stent inthe ductus arteriosus complicates, or even possibly ren-ders impossible, the usual “Norwood” surgery

Complications of arterial stents

Complications of systemic arterial stents include ation or extension of any of the complications of balloondilation of any vessel including, particularly, coarctation

exagger-of the aorta or dilation exagger-of other arterial lesions At the

same time, since the implant of a stent requires no dilation of the artery, complications related to dilation with

over-the balloons are very rare

Iatrogenic obstructions caused by the implant of stentsthat have a maximum diameter that will be too small for

the eventual size of the adult aorta, should be a totally

avoidable “complication” of stents used in the treatment

of coarctation of the aorta and other arteries There noware stents available that can be dilated to 25+ mm, whichmakes these stents large enough in diameter for any adultproximal descending aorta These potentially larger dia-meter stents should be used in any growing patient witheven the potential of having the aorta in the area of coarc-tation grow to larger than 16–18 mm in diameter A stent

with a limited diameter, which creates an obstruction

because it is too small for the aorta, creates an obstruction

that is much more difficult to repair than the usual native

or “residual” post-operative obstruction When the patient

is too small to implant a stent that can be dilated to the

anticipated diameter of the adult aorta, it is preferable to

treat the patient with balloon dilation without a stent or

with surgical repair initially A stent can be implanted

later to “finalize” the correction of any residual lesion

fol-lowing prior dilation or surgery

A very serious complication of the catheter treatment of

coarctation of the aorta using balloon dilation with or

without stent implant is injury to the central nervous

sys-tem (CNS) during the catheterization procedure CNS

injury most likely occurs from air and/or clot

emboliza-tion coming from catheters, sheaths or wires that are

pre-sent in the systemic circulation proximal to the carotid

and vertebral vessels Meticulous attention to clearing

fluid lines, catheters and sheaths of any air and/or clot,

keeping guide wires in the circulation “covered” with a

catheter that is maintained on a continual flush for as long

as possible, the continual flushing of all catheters and

sheaths that are positioned in the heart, and the routine

use of systemic heparin, particularly during “left heart”

procedures, should reduce or eliminate CNS problems

Direct injury from the tip of a wire positioned in a cranial

artery has been implicated in central nervous system

injury Never positioning a wire tip in either a carotid or a

vertebral artery eliminates this particular possibility

Probably the most common complication of the use of

stents in the arteries is injury to the local artery at the site

of catheter/stent introduction with subsequent

compro-mise of arterial blood flow in the involved extremity

These injuries occasionally are unavoidable because of the

necessarily very large sized balloon catheters/sheaths

that are used to deliver stents to the aorta Meticulous care

of the arteries, which was discussed earlier in this chapter

and in Chapter 4, is the best prevention and, in turn, best

treatment of this problem Local anesthesia is used

liber-ally and repeatedly around the artery and surrounding

tissues before, during, and at the end of the procedure

Local anesthesia is administered even if the patient is

receiving general anesthesia A precise, single-wall

punc-ture is utilized for the entrance into the artery Although

an indwelling sheath often is as much as 2–3 French sizes

larger than the dilation balloon or balloon catheter that

is used for the stent delivery, an indwelling sheath, which

is relatively fixed in the artery, is always less traumatic to

the artery than a constantly moving catheter or a bare,

rough, “folded” balloon/stent being introduced into

and withdrawn out of the artery When the sheath is

removed from the artery, the puncture site is compressed

manually and personally while continually monitoring

the puncture site for bleeding and, at the same time, the

artery distal to the puncture for a pulse This hemostasiscan take 30, 60, or more minutes, but should be accom-

plished before the patient leaves the observation and care

of the catheterizing physician!

Stent displacement is relatively common during theimplant of stents in coarctation of the aorta There often is

a large discrepancy in diameter between the aorta imal to the coarctation and the aorta distal to the coarcta-tion A stent that is expanded to a diameter that is calculated to fix the stent in the aorta proximal to thecoarctation, can easily become free-floating in the muchlarger distal aorta At the same time, expanding the stent

prox-to a diameter satisfacprox-tory for the diameter of the aorta tal to the coarctation would split the smaller-diameteraorta that is proximal to the obstruction In order to avoiddisplacement, the stent is implanted with the majority ofthe stent positioned in the aorta proximal to the narrow-

dis-ing and no attempt is made at “approximatdis-ing” the distal

end of the stent to the larger distal aorta during the initialimplant procedure If a stent dislodges and moves furtherdistally in the aorta, the stent is purposefully repositioned

in the distal aorta so that it does not compromise vital sidebranches, and then it is expanded and fixed in the moredistal location

Tears in the arterial wall or flaps off the intima/mediaare less common, or at least, less recognized, during stentimplants in systemic arteries, including congenital coarc-tation of the aorta, than with balloon dilation alone ofthese same lesions When the proper sized stents are usedafter the artery is measured properly and very accurately,the adjacent aorta should not be “over-dilated” at all and,

at the same time, any slight disruption of the intima/medial tissue within the area of the stent is compressedback against the wall of the aorta by the stent As theendothelium and neointima develop over and around the stent, the combination of the wall “thickening” by the

“new endothelial tissues”, the scarring as a consequence

of any tears that did occur, and the “metal scaffolding” ofthe stent within the area, all together create a very solidarterial wall The artery in the precise area of the stent isvery non-compliant, but no more so than a surgical scarinvolving the same area!

Aneurysms have occurred acutely during the implant

of stents in coarctations of the aorta These occurred morecommonly with the use of the larger Palmaz™ stents( Johnson & Johnson, Warren, NJ) and occurred predomin-antly (only?) when the stents were dilated acutely to their final large diameters with a single inflation of a large diameter balloon Acute aneurysms are not reportedwith the sequential dilation of stents to their precise, eventual large diameters, or with the use of stents that donot develop sharp tips at their ends as they expand.Aneurysms following the implant of stents in coarctation

of the aorta are still being studied, and should be looked

for in every patient who undergoes a stent implant in

the aorta

Tears or ruptures of the aorta occur with dilation of

native coarctation, re-coarctation of the aorta, and middle

aortic syndrome, but should not occur with the

conserva-tive implant of the correct size stent in coarctations.

Meticulous, accurate measurements of the lesion itself

and the adjacent vessels, and avoiding oversized

dila-tions/stents compared to the size of the lesion itself and

the adjacent vessel, presumably should prevent this

com-plication during stent implant In cases of very severe

stenosis of classic coarctation or middle aortic stenosis,

a staged dilation/implant during several sequential

catheterizations is utilized to avoid splitting very

nar-rowed vessels by a single dilation to a very large diameter

A stent that is still narrowed within a vessel can always be

dilated further at a later date Once the artery/aorta is

split or torn, there is little or no “turning back”, although

several operators have reported on the successful use of a

covered stent as an emergency “bail-out” therapy in the

catheterization laboratory22,23

Stents implanted in coarctation of the aorta have been

reported to fracture or kink24 This probably is a result of

the type of stent used in the area There have been no

adverse events from these findings and a recurrent

nar-rowing as a result of a fracture or kink can be treated with

the implant of an additional stent within the original stent

The implant of stents into the patent ductus of newborn

infants has its own specific complications These

compli-cations are in addition to the inherent complicompli-cations of

extensive catheter manipulations in very sick newborn or

small infants As with all complications, the best

treat-ment is prevention by paying meticulous attention to the

details of the procedure and the use of known, established

techniques until newer/better techniques are proven

Irritation and spasm of the ductus is a potential problem

with any manipulation around or through the ductus

This spasm is not always responsive to prostaglandin

infusion or re-infusion Should intractable ductal spasm

occur, having the equipment ready for immediate

deploy-ment of a stent is the best treatdeploy-ment This, however, is not

a guarantee of successful recovery since the “mass” of

even the “tiny” stent/balloon occasionally cannot be

advanced through the ductus once it begins to spasm

Disruption/tears of the ductal tissues is another potential

with stent implant into the patent ductus, particularly

when the stent delivery is rushed The tissue is inherently

very friable and cannot tolerate rough handling

When disruption of the ductus does occur, it usually is

catastrophic

Stent displacement during implant into the “patulous”

ductal tissue in an infant is a real problem Prevention, by

the use of a slow meticulous positioning and by waiting

for the prostaglandin to wear off before deploying the

stent, is the best treatment When a stent displaces fromthe ductus, an attempt is made to capture the stent on aballoon and reposition it back into the ductus A new bal-loon which is at least 1 mm larger than the maximumdiameter of the stent is more effective for “capturing” an

errant coronary stent When a coronary stent becomes placed from the ductus and is positioned or implanted in any other artery (even very peripherally), the coronary

dis-stent, unequivocally, eventually will produce a verysignificant stenosis in that vessel because of its very smallmaximum diameter If a displaced stent cannot be cap-tured and reimplanted successfully in the ductus orremoved from the patient with a catheter, the stent should

be removed surgically from the errant vessel within a fewdays after the attempted implant procedure, unless theerrant vessel is considered “expendable”

The majority of the complications of stent implants inarterial locations are eliminated by the use of extremelyaccurate measurements, a conservative diameter at theinitial implant, and by paying meticulous attention to the

details of every step of the procedure The morbidity and

complications of dilation with intravascular stent implantfor systemic arteries appear to be comparable to or evenless than surgical therapy of these same lesions Dilationwith stent implants in coarctation and other congenitalsystemic arterial lesions still represents a “new” treat-ment, which requires decades of follow-up to determineits real efficacy and safety

References

1 Dotter CT et al Transluminal expandable nitinol coil

stent grafting: preliminary report Radiology 1983; 147(1):

259–260.

2 Palmaz JC et al Atherosclerotic rabbit aortas: expandable

intraluminal grafting Radiology 1986; 160: 723–726.

3 Shaffer KM et al Intravascular stents in congenital heart

dis-ease: short- and long-term results from a large single-center

experience J Am Coll Cardiol 1998; 31(3): 661–667.

4 Morrow WR et al Balloon angioplasty with stent tion in experimental coarctation of the aorta Circulation 1994;

implanta-89(6): 2677–2683.

5 Grifka RG et al Balloon expandable intravascular stents:

aortic implantation and late further dilation in growing

minipigs Am Heart J 1993; 126(4): 979–984.

6 Suarez de Lezo J et al Balloon-expandable stent repair of

severe coarctation of the aorta Am Heart J 1995; 129(5):

1002–1008.

7 Rosenthal E, Qureshi SA, and Tynan M Stent implantation

for aortic recoarctation Am Heart J 1995; 129(6): 1220–1221.

8 Bulbul ZR et al Implantation of balloon-expandable stents

for coarctation of the aorta: implantation data and short-term

results Cathet Cardiovasc Diagn 1996; 39(1): 36–42.

9 Cheatham JP Stenting of coarctation of the aorta Catheter

Cardiovasc Interv 2001; 54(1): 112–125.

10 Mendelsohn AM et al Stent redilation in canine models of

congenital heart disease: pulmonary artery stenosis and

coarctation of the aorta Cathet Cardiovasc Diagn 1996; 38(4):

430– 440.

11 Cheatham J et al Early experience using endovascular stents

in children with coarctation of the aorta: promising results

but proceed with caution (abstr) Cardiol Young 1998;

9(Suppl 1:11): (abstr).

12 Suarez de Lezo J et al Immediate and follow-up findings

after stent treatment for severe coarctation of the aorta Am J

Cardiol 1999; 83(3): 400–406.

13 Sapin SO, Rosengart RM, and Salem MM Chest pain

dur-ing stentdur-ing of a native aortic coarctation: a case for acute

intercostal muscle ischemia and rhabdomyolysis Catheter

Cardiovasc Interv 2002; 57(2): 217–220.

14 Thanopoulos BV et al Long segment coarctation of the

thor-acic aorta: treatment with multiple balloon-expandable stent

implantation Am Heart J 1997; 133(4): 470–473.

15 Tyagi S et al Percutaneous transluminal angioplasty for

stenosis of the aorta due to aortic arteritis in children Pediatr

Cardiol 1999; 20(6): 404–410.

16 Redington AN and Somerville J Stenting of aortopulmonary

collaterals in complex pulmonary atresia Circulation 1996;

94(10): 2479–2484.

17 El-Said HG et al Stenting of stenosed aortopulmonary

collaterals and shunts for palliation of pulmonary atresia/

ventricular septal defect Catheter Cardiovasc Interv 2000;

49(4): 430–436.

18 Gibbs JL et al Stenting of the arterial duct: a new approach to

palliation for pulmonary atresia Br Heart J 1992; 67(3):

240–245.

19 Michel-Behnke I et al Stent implantation in the ductus

arte-riosus for pulmonary blood supply in congenital heart

dis-ease Catheter Cardiovasc Interv 2004; 61(2): 242–252.

20 Ruiz CE et al Brief report: stenting of the ductus arteriosus as

a bridge to cardiac transplantation in infants with the

hypoplastic left-heart syndrome N Engl J Med 1993; 328(22):

1605–1608.

21 Slack MC et al Stenting of the ductus arteriosus in

hypoplas-tic left heart syndrome as an ambulatory bridge to cardiac

transplantation Am J Cardiol 1994; 74(6): 636–637.

22 Khan MS and Moore JW Treatment of abdominal aortic pseudoaneurysm with covered stents in a pediatric patient.

Catheter Cardiovasc Interv 2000; 50(4): 445–448.

23 Tyagi S, Rangesetty UC, and Kaul UA Endovascular ment of aortic rupture during angioplasty for aortic in-stent

treat-restenosis in aortoarteritis Catheter Cardiovasc Interv 2003;

58(1): 103–106.

24 Ledesma M et al Stent fracture after stent therapy for aortic

coarctation J Invasive Cardiol 2003; 15(12): 719–721.

Occlusion of abnormal or persistent arterial or

arterioven-ous structures or vessels feeding vascular leaks or tumors

by catheter embolization techniques has been utilized for

over thirty years1 The embolization techniques were

developed and perfected primarily by the vascular

radi-ologists working in the abdominal viscera, gastrointestinal

areas and the central nervous system, particularly in “end

artery” vessels Many materials and devices, including the

patient’s own clotted blood, Gelfoam™, colloidal plugs,

“glues”, detachable balloons and coil occlusion devices

have been used for these peripheral occlusions1– 6

In the pediatric and congenital heart population there

are numerous abnormal congenital and acquired vascular

communications and intravascular “leaks” which require

or, at least, can be benefited by transcatheter occlusion

The occlusion of these vascular lesions in pediatric and

congenital heart patients has been performed in the

cath-eterization laboratory for over two decades The

abnor-mal flow through these communications usually results

in significant abnormalities of the underlying

hemody-namics and compromises the patient’s symptomatic and

hemodynamic status The abnormal vascular

communica-tions which are encountered in pediatric and congenital

heart patients include traumatic fistulae, systemic to

pul-monary artery collaterals, systemic arteriovenous fistulae,

pulmonary arteriovenous fistulae, coronary arterial-cameral

fistulae, perivalvular leaks and a variety of residual,

surgi-cally created systemic to pulmonary artery

communica-tions including Blalock–Taussig, modified Blalock–Taussig,

Potts, and Waterston/Cooley shunts

There are numerous different catheter-delivered devices

and techniques available for the occlusion of these

abnor-mal vascular structures There is no single device

applic-able for every lesion and multiple devices may be suitapplic-able,

and used, for any one lesion These devices/materials are

used either by themselves or (frequently) in combination

with one or more of the other devices The specific sion device used depends upon the type, size and location

occlu-of the communication/leak as well as the availability occlu-of

a particular device either locally or as approved, in theparticular country Some of these devices are designedspecifically for a particular lesion and are discussed indetail in other chapters in the description of the occlusion

of the specific intracardiac defect These same descriptionsare not repeated in this chapter

Since most of the devices can be utilized for the sion of multiple different structures and many of theabnormal vascular communications can be occluded withseveral different devices, each of these miscellaneous vas-cular lesions and the separate vascular occlusion devicesthat are used for that lesion are included in the discussion

occlu-of the particular lesions in this chapter The multipledevices themselves that are available for these occlusionsare discussed initially in this chapter, before the details oftheir use in the various lesions for which they can be used

Devices/equipment for vascular occlusions

Occlusion coils

Stainless steel occlusion coils are the most widely used

of the catheter-delivered occlusion devices and have hadthe longest continued use in pediatric and congenitalheart lesions They are particularly useful for small or tor-tuous vessels and have gained enormous popularity anduse for the catheter occlusion of the patent ductus arterio-sus (PDA) The specific coils used for PDA occlusion and the modifications of the delivery system/techniquesspecifically for the PDA are discussed separately and indetail in Chapter 27 (“PDA Occlusion”) Many of thesemodifications, which were developed specifically forPDA occlusions, are useful for the occlusion of generalvascular structures

including perivalvular leaks

Occlusion coils are available as specific lengths of

vari-ous sizes (gauges) of stainless steel, spring guide wire The

“guide” wires are pre-formed during manufacturing so

that they will coil into a cylindrical tube of a specific

dia-meter of the wire in their “resting” state The number of

loops of coil and the length of the “tube” of coils depends

upon the original length of the straightened wire Most

of the “occlusion coils” have multiple tiny threads or

fila-ments of nylon fabric intertwined within the windings

of the spring wires to promote better thrombosis within

a vessel

Occlusion coils are best suited for the occlusion of long

and/or tortuous vessels with irregular internal diameters,

and especially those which have a significant narrowing

somewhere along the course of the vessel Since a

success-ful occlusion is expected to cut off all blood flow through

the vessel, the vessels being occluded should not be the

sole blood supply to a particular area of tissue unless

necrosis of the tissues that are supplied by the vessel is

desired

The coil occluder with the most extensive use in pediatric

and congenital heart patients is the standard Gianturco™

coil The Gianturco™ coil is a length of special stainless

steel spring guide wire in which the stiffening “core” wire

is pre-formed to curl into a “coil” or “wire cylinder” of a

specific diameter in its “free” or “resting” state These

coils are available in multiple sizes of the spring wire,

multiple diameters of the coil (“cylinder”) and multiple

lengths of the coil wire The Gianturco™ coil has multiple

fine nylon fiber segments embedded within the windings

of the spring wire in order to increase the thrombogenicity

of the implanted coil Gianturco™ coils are available in

spring wire diameters of 0.025″, 0.035″, 0.038″ and now, an

additional coil of 0.052″ wire diameter, in lengths between

1.2 and 15 cm and in coil diameters from as small as 2 mm to

as large as 20 mm in diameter The very smallest diameter,

short coils are available only in the 0.025″ diameter wires

while very large diameter coils are available only in the

more recently available 0.052″ wires The total length of

the straightened segment of coil wire in conjunction with

the particular diameter of the preformed loops of the coils,

determine the number of loops which are formed by any

particular length of coil Each Gianturco™ coil comes

from the manufacturer in a straight metal introducer tube

The internal lumen and the length of the introducer tube

are specific for the diameter of each wire and the

straight-ened length of the wire

The mass of the wire of the coil itself creates a

mechan-ical occlusion and the embedded nylon fibers add to the

thromboses in the area where the coil is deposited and, in

turn, occlude the vessel or communication The coils are

best suited for deposit into tubular vessels that have some length and vessels that have an area of narrowingsomewhere within the channel of the vessel or the abnor-mal communication A stenosis distally in the channel ofthe vessel prevents even coils that are undersized frommigrating out of the target vessel and embolizing to anarea or vital organ distally beyond the vessel

The coils are delivered through polyethylene, end-hole

only, catheters, which have an internal diameter which

is just slightly greater than the diameter of the wire of the spring coil Other end-hole only catheters manufactured

from materials that impart a smooth or slick inner surfaceand that are slightly larger in their internal diameters than the wire of the coil can be used for coil delivery The

catheter for coil delivery must not have side holes Side

holes allow the potentially curved tip of the coil to catch

in, or pass into, a side hole of the catheter as the tip of thecoil crosses the side hole The tip of a coil catching in a sidehole of the catheter will prevent the coil from being deliv-ered through the tip of the catheter Since the standardGianturco™ coils have no attachment to the delivery wire,the coil catching in a side hole also prevents any retrieval

of the coil without totally removing the delivery catheter.Both the material of the catheter and the internal diameter

of the catheter are critically important in order to preventthe coil from “binding” within the lumen of the catheterduring the delivery through the catheter

A catheter that is smaller in internal diameter than thecoil wire obviously does not allow the coil with its imbed-ded fibers to be introduced into, or advanced through, the catheter A catheter with an internal diameter signific-antly larger than the diameter of the wire of the coil allowsthe coil to bend and partially “coil” within the catheter orallows the pusher wire to push past the coil instead ofactually “pushing” the coil through the catheter Eitheroccurrence will cause the coil to bind within the catheter.The delivery catheters are available with many pre-formed tip configurations in order to facilitate entry intospecific areas Straight delivery catheters, which the operator can pre-shape to suit his particular needs, arealso used to deliver coils End-hole, only, floating ballooncatheters can be used to deliver the coils to certain locations or in particular circumstances The inflated balloon helps to fix the tip of the catheter in place and/or

prevents portions of the coil from extending back into

a more proximal main vascular channel The catheterlumen of the floating balloon catheter obviously must be

of a slightly larger internal diameter than the diameter of

the coil wire that is being delivered through the balloon

catheter

The coil is introduced into the proximal hub of thedelivery catheter through the straight metal “loader” as astraight length of wire The straightened coil is pushed out

of the loader, into the delivery catheter and through the

delivery catheter with a teflon-coated, spring guide wire

of the same or similar wire size as the coil wire The coil is

delivered by pushing it completely through and out of the

distal end of the delivery catheter As the coil is extruded

out of the delivery catheter, it immediately begins to form

the small loop of its predetermined “coil” diameter as it

opens into its coiled configuration Once the extrusion

from the catheter starts with the standard Gianturco™ or

the 0.052″ coils, there is no way of withdrawing the coil

back into the catheter or stopping or reversing the

deliv-ery Even if the coil is noted to be in an unsatisfactory

posi-tion as it starts to extrude from the catheter, it can only be

extruded completely and then retrieved with a separate

retrieval catheter and system

When choosing the appropriate occlusion coil, the

diameter, the length and the general configuration of

the vessel to be occluded are imaged angiographically

The length and diameters of the vessel are measured

very accurately on the angiograms The Gianturco™ coil

occludes the vessel by the creation of an irregular mass of

the coil wire and the nylon fiber strands that are

incorpor-ated within the wire in which a thrombus forms The coil

used should be 1–2 mm larger in diameter than the vessel

that is to be occluded The slightly larger diameter results

in the coil unraveling in an irregular configuration within

and across the vessel lumen rather than into a neat

“donut” like cylinder or smoothly coiled configuration

If the diameter of the coil is far larger than the diameter of

the vessel, the coil does not “coil” at all, but rather tends to

align straight within the vessel lumen and, in turn, does

not form an effective occlusive mass in the vessel When

the coil is extruded from the catheter, it not only must

have the appropriate diameter and length to fix to the

walls of the vessel, but also must not be excessively long

When the coil is too long, it can extend proximally out of

the target vessel and into the more central feeder vessel,

which potentially can be back into the normal circulation

and interfere with vital structures If, at the other extreme,

the diameter of the formed coil is too small for the vessel,

the coil rolls up into a tight “donut”, does not occlude the

entire vessel, and is likely to tumble distally or even out of

the desired vessel Once one coil is secured within a

ves-sel, additional coils of different sizes and/or diameters

can, and frequently are, intertwined within or deposited

proximal to the original coil to complete the occlusion

Even when used in tandem, but without a distal

narrow-ing or some other type of device for fixation, the standard

Gianturco™ coil generally is only usable in tubular

struc-tures of no more than 7–8 mm in their distended diameter

For larger vessels and vessels without an area of discrete

stenosis, either the standard Gianturco™ coils are used in

conjunction with other intravascular occlusion devices or

the 0.052″ coils are used initially to begin the occlusion

of the vessel

Occlusion coils can be deposited into long vessels thathave a discrete distal narrowing, where the coil thenlodges in place at the narrowing, although it is not “fixed”against the wall In that circumstance, coils can be

“floated” into the vessel, one after another to create a mass

of coils, proximal to the stenosis in the vessel When thistechnique is used, the final coil in the vessel should be of a

slightly larger diameter then the diameter of the vessel in

order to wedge the last coil against the walls of the vessel.The one “fixed” coil assures that the “loose” coils packedwithin the vessel do not “float” back out of the vessel intothe vital circulation

Currently, by far the most common use of theGianturco™ coil in congenital heart lesions is for the clo-sure of the patent ductus arteriosus This is an entirelyseparate subject and is discussed in detail in Chapter 27and is not covered in this chapter at all There are manyabnormal vessels, collaterals and persistent surgically cre-ated systemic to pulmonary artery shunts, which fre-quently are associated with more complex lesions Thesevessels require occlusion when the additional systemicflow competes with normal pulmonary flow, particularlywhen the abnormal communication persists after themajor intracardiac defect has been corrected These communications traditionally required surgical divisionduring the corrective procedure or as a separate, later, surgical procedure When the occlusion of these defects

is performed surgically during the intracardiac repair, itsignificantly prolongs or complicates the surgery Most ofthese abnormal communications now are occluded withGianturco™ coils either before or shortly after the majorsurgery7 With the use of coils, further extensive extrasurgery or repeat surgery is unnecessary for the elimina-tion of persistent systemic to pulmonary artery collaterals

or for any surgically created systemic to pulmonary arteryshunts that are present at the time of, or following, the

“total” correction

Other lesions in which the coils are useful are venous fistulae, including systemic coronary-cameral,peripheral arteriovenous fistulae as well as pulmonaryarteriovenous fistulae These lesions can produce eitherleft to right or right to left shunts In these lesions it is critical to identify a stenotic or “end” vessel into which thedevice can be fixed very securely in order to reduce thedangers of embolization to an essential more distal vessel

arterio-or vital structure in the systemic circulation

The 0.052 inch stainless steel coil

In order to provide a more robust coil, a more occlusivecoil and a coil particularly for use in the patent ductusarteriosus, the 0.052″ stainless steel coil was developed.The 0.052″ coil is a larger, stiffer version of the standardGianturco™ coil with the wire of the coil being the heavier

gauge 0.052″ diameter The 0.052″ coil provides a much

sturdier occluding device for the miscellaneous vessels

and is particularly useful for the occlusion of much larger

vessels with higher pressures and higher flows8 The use

of the 0.052″ coils for the occlusion of the patent ductus

arteriosus for which they were developed, is discussed in

detail in Chapter 27

In an end vessel or in a vessel with definite distal

stenosis, the 0.052″ coils are delivered with a “free release”

technique exactly like the standard, smaller sized,

Gianturco™ coils However, when choosing the diameter

of the 0.052″ coil for a particular constrained vessel, the

diameter of the coil used should be no more than 1–1.5 mm

larger than the stretched diameter of the vessel being

occluded The 0.052″ coils are very rigid and have a much

greater tendency to form into a symmetrical “donut”

shape after they are extruded, and they have very little

tendency to form into irregular or elongated shapes In a

vessel that is significantly small in diameter for the coil

and non-elastic, the 0.052″ coil is likely to elongate into

an almost straight wire rather than to bunch up into an

irregular coil

Because of the larger wire size alone, the 0.052″ coil

requires a delivery “catheter” of a larger internal diameter

for implant than the standard Gianturco™ coils The

necessity of using the larger delivery catheters can

pro-hibit the use of the 0.052″ coils in infants and very small

children This becomes an even greater problem when the

0.052″ coil embolizes away from the implant location and

must be retrieved To overcome the necessity of using

the larger French sized delivery catheters, the 0.052″ coil is

usually delivered through a 4- or 5-French, long,

transsep-tal type sheath with a radio-opaque band at the tip rather

than through a separate delivery catheter This allows the

delivery of the larger diameter coil wire without an

over-all increase in outside diameter of the delivery system The

long sheaths have the disadvantages of having less

flex-ibility, less ability to bend at acute angles and, as a

con-sequence, a greater tendency to kink than most “delivery

catheters” This tendency to kink compromises the access

to vessels or lesions arising at acute angles off the major

vessel (aorta) When the long 4- or 5-French sheath is used

to deliver the 0.052″ coil, it is advisable to introduce the

long sheath into the peripheral entry vessel through a

short, 6- or 7-F sheath Then, if the totally extruded coil

needs to be withdrawn, it can be withdrawn into the

larger, short 6- or 7-French sheath, which is already in the

vessel In the larger patient, the 0.052″ coil, of course, can

be delivered through a 6- or 7-French guiding catheter

The “bioptome controlled” delivery technique is

prefer-able to the “free release” technique for the delivery of the

0.052″ coils to all locations, but particularly when there are

any concerns about the proper “seating” or possible distal

embolization of this coil Bioptome-controlled delivery is

described in detail, later in this chapter and in Chapter 27

on “PDA Occlusion”

Delivery techniques for the Gianturco™

“Free-release” technique for coil delivery

An end-hole only delivery catheter is chosen with a shape

of the tip of the catheter that will facilitate entry into thespecific area to be occluded and of an internal diameter to

match the diameter of the coil wire being used The

deliv-ery catheter/sheath is manipulated and advanced as far

as possible into the vessel to be occluded The tip of thedelivery catheter/sheath is fixed securely in the vessel at,

or distal to the site for, implant of the occlusion device.Extra effort should always be made to ensure that thedelivery catheter/sheath is positioned very deep into thevessel that is to be occluded Often during the process ofadvancing the coil through the catheter or extruding thecoil wire out of the tip of the delivery catheter/sheath, the tip of the delivery catheter/sheath can be pushedbackward in the vessel The catheter/sheath must be farenough into the vessel initially to allow for this

Once the catheter is positioned properly, the thin lar end of the coil introducer is introduced into the prox-imal hub of the pre-positioned delivery catheter or sheath.The coil introducer should be introduced through a wireback-bleed/flush device that previously was attached tothe proximal hub of the coil delivery catheter/sheath Theback-bleed/flush device allows a continual flush of thedelivery catheter or sheath during the introduction of the coil and during any exchange of pusher wires, which, inturn, “lubricates” the lumen of the catheter The stiff end

tubu-of a straight teflon-coated spring guide wire (tubu-of the samesize as the internal diameter of the delivery catheter/sheath and the same size or slightly larger [if possible]

than the wire of the coil itself) is introduced into the

prox-imal end of the straight tubular coil introducer As thestraight spring guide wire is introduced into the proximalend of the coil introducer, the coil is pushed through andout of the distal end of the introducer tube and into thedelivery catheter/sheath by the “pusher” spring guide

wire The coil wire has no attachment or connection to the

pusher wire so that once the introduction of the coil into

the catheter has begun, the coil can be advanced only within

the catheter Standard Gianturco™ or 0.052″ coils cannot

be withdrawn at all! Once the coil is completely within the proximal end of the delivery catheter/sheath, the coil introducer is withdrawn from the catheter over the

“pusher” wire and pulled back to the proximal end of, oroff, the pusher wire The teflon-coated spring pusher wire

is reversed and the soft end of the wire is advanced within

the catheter, which, in turn, pushes the coil (still as

a straight segment of wire) through the length of the

delivery catheter/sheath Once the wire and coil are

well within the catheter, the catheter, pusher wire and

enclosed coil wire are observed carefully on fluoroscopy

as the coil is advanced within the catheter or long sheath

The coil is distinguishable from the pusher wire by its

slightly “undulating” configuration and the different

dens-ity of the wire of the coil compared to the pusher wire As

a check of the two wires, a small radiolucent “interspace”

can be created between the proximal end of the coil and

the distal end of the pusher wire by withdrawing the

pusher wire several millimeters while observing under

fluoroscopy

Once the distal end of the coil has reached the tip of the

catheter/long sheath, the coil begins to extrude from the

tip of the catheter/sheath by continuing to advance

the “pusher” wire As the coil is extruded from the tip of

the catheter or sheath, the “coil wire” immediately begins

to curl into its pre-formed coiled configuration and/or

push the tip of the delivery catheter/sheath away from

the site Exactly how the coil positions itself within the

vessel depends upon the size relationship of the diameter

of the pre-formed coil and the internal diameter of the

ves-sel at that particular location

It is important to choose a coil diameter of the standard

Gianturco™ coil that is 15–20% larger than the stretched

or expanded internal diameter of the vessel at the area

where the coil is to be implanted If the coil diameter is

smaller than the vessel diameter, the coil wire “coils” into

a tight, smooth circular coil with the appearance of a small

“donut” and is likely to embolize further along in the

ves-sel When it tumbles and if it does become lodged in the

more distal vessel, the lumen within the “donut” of the

coiled wire can line up with the vessel lumen and prevent

effective occlusion On the other hand, when the diameter

of the coil is much larger than the diameter of the vessel

lumen, then the coil cannot “coil” on itself at all and

stretches out longitudinally in the lumen of the vessel and

again, probably will not occlude the vessel When the coil

diameter is so large that the coil cannot “coil” at all, as the

pusher wire and coil wire are advanced through the

catheter or sheath and as the coil wire is pushed out of

the tip of the delivery catheter or sheath, the coil wire

re-mains straight and actually pushes the tip of the delivery

catheter/sheath back in the vessel The catheter or sheath

tip can be pushed completely back out of the vessel into

which the coil is being delivered This leaves some, or all,

of the coil extending proximally out of the target vessel

The ideal coil/vessel size relationship allows the coil to

partially coil on itself yet partially stretch out into a very

irregular shape or mass When the coil is to be delivered

into a very critical and short segment of vessel, the

“dis-tensibility” of the particular vessel can be tested precisely

at that location by a low-pressure inflation of a smallangioplasty balloon that is slightly larger in diameter thanthe vessel at that area

After the successful implant of a coil, a small injection ofcontrast is performed through the delivery catheter orsheath to verify the degree of occlusion Often, it is neces-sary to deposit more than one coil, and often even mul-tiple coils of different sizes, in any single vessel to completethe occlusion As long as there is room in the vessel, addi-tional coils are deposited at the same location, through thesame delivery catheter or sheath, and during the sameprocedure until the vessel is occluded completely

In most cases of the occlusions of discrete vessels, ard coils are extruded into a confined segment of the ves-sel and are released automatically by the “free-release”technique as they exit the delivery catheter The majority

stand-of vessels that are occluded (except the PDA), have somedistal tortuosity or narrowing on which the extruded coilbecomes trapped The very accurate “teetering” across thenarrowest segment of a vessel that is required for coilocclusion of the PDA, seldom is encountered in the major-ity of other vascular/small vessel occlusions With accur-ate information about the size of the vessel to be occludedand when the proper size and type of coil is used for theocclusion of discrete vessels, there is full expectation thatthe coil will “fold up” into a compacted, irregular shape,reorient and move (usually distally) after its release andthen lodge securely in the vessel When there is a highlikelihood of this type of seating in an appropriate vesselthere is little need for the more complicated and expensivedetachable/retractable coils Although unnecessary forthe majority of coil deliveries for vascular occlusions, con-trol of the release and retrievability of the coil, however,does become essential in some circumstances

Special techniques for the delivery or

to improve the safety of their delivery

When the coil must be delivered in a very specific site in

order not to occlude adjacent or branching vessels that arecritical for the supply of essential viable tissues (e.g coron-ary branches adjacent to a coronary-cameral fistula), then a controlled release/retrievable system must beused When the vessel more proximal to the area beingoccluded is significantly wider than the area of the vesselwhere the coils are being implanted or when the coils aredelivered near to the aortic entrance of the vessel and asthe vessel becomes full of coils, then a retractable or con-trolled release coil is desirable or even essential Similarly,

if the vessel has only minimal length or the coil is at alllarge in diameter relative to the vessel diameter, there is

a high probability that the coil will elongate too much ing delivery and will extend back into the central vessel

dur-(aorta!) as it is extruded from the delivery catheter, and a

control of the release of the coil is essential Even when

multiple coils are being “wadded” into a long vessel, easy

and immediate retrievability of the last and most

prox-imal coil always is desirable

When the free-release coil is too small in diameter for

the vessel, it “balls-up” into its “donut” configuration,

“bounces around” in the vessel, which is too large, and

easily can embolize backward and out of the vessel into

the central, critical vessel With standard free-release

Gianturco™ coils, there is absolutely no control over this

tumbling, once the extrusion of the coil has started

“Retrieval” of a free-release Gianturco™ coil after the

extrusion of the coil has started, thereafter, represents a

“foreign body” retrieval of a coil which has embolized to a

distal location from the site of implant!

Special techniques to achieve control over

steel coils

In order to overcome the non-retrievability, “all or

noth-ing” delivery of standard Gianturco™ and 0.052″ coils,

multiple modifications of the delivery system/technique

have been developed With a controlled attach/release coil,

the fit and fixation of the completely extruded coil can be

tested once the coil is extruded completely into the vessel

before the coil is released from the attached delivery

wire/cable If the position/fixation is not satisfactory, the

coil can be withdrawn back into the delivery catheter and

repositioned before it is released, or the coil can be

with-drawn totally out of the body and the procedure restarted

with a more appropriately sized coil When there is a

residual leak through the coils and even when the vessel

proximal to the coils is short or otherwise not ideal, by

using a controlled-release coil system, an additional coil

can be extruded safely into earlier coils in the vessel and

tested for both occlusion and fixation before release If

either fixation or occlusion is not satisfactory, the

addi-tional coil can still be withdrawn from the original coil(s),

although care must be taken because each successive coil

tends to become entangled with any previous coil(s)

The various commercially available and the self-made

techniques for controlled release of the coil, all are

effect-ive to some degree at accomplishing retrievability of the

coil during delivery The minute details of all of these

modifications are described in detail in Chapter 27 on

“PDA Occlusion”, where these coil “control” systems

are more of a necessity The specific uses of the various

“controlled-release” systems for general vascular

occlu-sions are discussed here

Because of the problem of a “no-return” delivery once

the extrusion of standard Gianturco™ and 0.052″ coils

is started and, in the long absence of a commercially

available and viable “detachable coil” in the United States,there have been several very innovative techniques devel-oped in order to overcome this shortcoming and make theGianturco™ and 0.052″ coils safer and more reliable

Latson catheter modification for Gianturco™

coil delivery

The Latson™ modification of the delivery catheter provides

some degree of control and retrievability for the standardGianturco™ coil9 The tip of the delivery catheter is heatedand then pulled into a tapered tip over a short, solid,

smooth wire or “mandril” of exactly the diameter of the bare coil spring wire of the particular coil The mandril is

removed and the pulled tip of the catheter is cut off at thenarrowest area of the “pulled taper” on the catheter Thiscreates an opening in the tapered tip of the catheter, which

is tight around the bare spring wire of the coil and very

tight around the coil wire when the nylon fabric is ded in the wire of the coil This modified tip configura-tion is now available commercially as the Latson™ multipurpose catheter: #248498 (Cook Inc., Bloomington,IN) and as the modified Vertebral catheter: #WN27750(Mallinckrodt Inc., St Louis, MO)

embed-As a result of the narrowed orifice of the tip of thecatheter, the coil now is gripped very tightly as it passes

through the narrowed orifice of the tightened, modified

tip of the catheter As a consequence of the tight grip

on the wire, it is necessary to apply considerable force

to the delivery/pusher wire to extrude the standardGianturco™ coil through the tip of the Latson™ catheter.Because of this tight grip on the coil created by themodified tip of the catheter, and unless the coil hasbecome entrapped on something in the vessel (a previouscoil!), the now “dangling” coil, which has been extruded

almost entirely but is now gripped tightly by the tip of the

modified catheter, can be withdrawn from a vessel andout of the body along with the catheter when the deliverycatheter is withdrawn through a peripheral introductorysheath

During the withdrawal of the Latson™ catheter, the coilcannot be withdrawn back into the Latson™ catheter butrather the coil, which is now mostly extruded, will be

“dangling” at the tip of the catheter, exposed in the

circu-lation and not “protected” as it is being withdrawn within

the circulation The coil cannot be repositioned or reusedwith this catheter while it still is within the circulation Tostart over, the coil and delivery catheter are withdrawncompletely out of the body through the peripheral intro-ductory sheath, and a new coil is introduced through thesame delivery catheter

The technique for the delivery of a Gianturco™ coil to avessel using the Latson™ modification is almost identical

to the delivery of the “free-release” coils Usually, but not

necessarily, the retrograde approach is used with the

Latson™ catheter Once in position the coil is pushed out

of the catheter with the pusher wire, but now with the

nec-essary considerable additional force on the pusher wire

Because of the force required, more attention must be paid

to the extrusion process in order to extrude the coil in

small, controlled increments in order to deliver the coil to

the precise location The grip on the coil by the tip of

the catheter is dependent completely on the very accurate

“tolerances” of the internal diameter of the tip of the

catheter whether the catheter is built by hand or by

com-mercial manufacturers The Latson™ catheter tip is not as

strong or dependable as any of the other attach/release

mechanisms and does not allow the reuse of the coil

during the same procedure without first withdrawing the

coil out of the body completely Although the Latson™

catheter does provide some extra safety in the delivery of

a standard Gianturco™ coil, the necessary tolerances and

the inability to withdraw the coil into the catheter are

dis-advantages to this catheter modification for coil delivery

Balloon-assisted coil delivery to branch vessels

Floating end-hole balloon catheters occasionally are used

to deliver any 0.035″ or smaller Gianturco™ coils to

abnor-mal branch vessels Balloon catheters are used with

several different techniques for the delivery of the coils

The end-hole, floating balloon catheter may be the only

catheter that can be manipulated into the vessel by the

particular operator In that circumstance, usually, once in

the vessel, the balloon is deflated within the vessel and the

end-hole balloon catheter is used like any other end-hole

catheter for the delivery of the coil Because of the “ragged”

surface of the deflated balloon over the tip of the catheter,

extra attention is necessary during the withdrawal of the

catheter after the coil has been delivered to prevent the

balloon from snagging on the fabric strands that are

dang-ling from the freshly implanted Gianturco™ coil

The end-hole floating balloon catheter is also used with

the balloon inflated securely in the vessel in order to wedge

coils forcibly into distal locations or to keep the coils from

extruding back out of the target vessel For these

pur-poses, the balloon catheter is manipulated into the vessel

to be occluded and then the balloon is inflated tightly

against the walls within the particular vessel With the

balloon catheter fixed tightly within the vessel, the

“free-release” Gianturco™ coil is extruded into the vessel distal

to the balloon When the inflated balloon is fixed tightly

enough in the vessel, the coils can be pushed into the

vessel with some force to “pack” them more tightly in the

vessel This usually represents a delicate balance between

the fixation of the balloon in the vessel and the force used

against the pusher wire If too much force is applied, the

inflated balloon can be pushed out of the vessel along

with the coils that are being packed into the vessel If the entrance of the vessel being occluded is close to theentrance of a vital branch vessel, some of the coils canextend into the adjacent vessel or the entire coil canembolize distally into the more central circulation!

Bioptome-assisted coil delivery

The bioptome-assisted delivery of coils for PDA sion, and in particular the 0.052″ coils, is discussed indetail in Chapter 27 (“PDA Occlusion”)8 The bioptome-assisted technique is effective for the precise control of the delivery of any Gianturco™ coil to locations otherthan the PDA where the exact localization of the coil isextremely critical Bioptome-assisted delivery is equally

occlu-as effective when using the more frequently used 0.038″coils as it is with the 0.052″ coils Bioptome-controlleddelivery is offered an alternative to the commercially

manufactured, detachable coils, which in their more robust form are only available outside of the USA The bioptome

technique allows complete retrievability of the coil at anytime until the purposeful release of the grasp with thebioptome The bioptome technique is used as an adjuncttechnique to the “free-release” technique in the placement

of the coils into very specific locations or the “final” coils

in more critical or precarious locations

The bioptome attachment to the coil requires the lumen

or internal diameter (ID) of the delivery system to be at

least 1.3 mm (4-French ID) A 4-French long sheath with

a back-bleed valve and distal radio-opaque marker is

used as the minimum sized delivery system for

bioptome-controlled delivery However, a 5-French sheath allows

a “more comfortable” delivery and more reliable drawal of the coil/bioptome back into the delivery systemwhen necessary particularly with the 0.052″ coils As aconsequence, a 5-French sheath with distal radio-opaquemarker is usually used with bioptome-controlled coilsexcept in the very smallest infants

with-An end-hole catheter is introduced through a 6- or 7-French short sheath and advanced well into the vessel

to be occluded The end-hole catheter is replaced with

a stiff, exchange length wire The 4- or 5-French longsheath/dilator that is to be used to deliver the coil/biop-tome is introduced over the wire and through the short

introductory sheath and advanced until the tip of the sheath is significantly past the “target” area within the ves-

sel to be occluded Because of the use of a sheath for thedelivery and the stiffness of the bioptome itself, bioptome-controlled delivery does have some limitations as towhich vessels it can be used in Often, if there is significanttortuosity of the vessel proximal to the implant site, thebioptome-controlled technique cannot be used

The bioptome technique requires a special tion of the occlusion coil, which is used in order for the

prepara-coil to be grasped with the bioptome All Gianturco™

and 0.052″ coils consist of a very fine stainless steel wire

tightly wound to form the larger “spring guide” type

wire of the advertised coil “wire” diameter One end of

the “spring” wire is sealed and polished with a small

“weld”, while the opposite end is open with a very

tiny hollow tube created by the fine wires The coils,

including the 0.052″ coils, come loaded in a metal “loading

tube” with the welded end of the coil at the distal end of

the loader

The stiff end of a separate 0.038″ spring guide wire

is introduced into the proximal end of the coil holder/

loader and advanced until just 1–2 mm of the distal, closed

(welded) end of the coil is exposed exiting the distal end of

the metal loading tube This exposes the tiny welded tip,

which seals the end of the fine coiled “wire tube” of which

the “spring coil” is made While holding the short more

proximal and exposed portion of the coil tightly between

the fingers, the small welded tip (only) is grasped tightly

with a forceps and pulled 0.25 to 0.5 mm away from the

remainder of the spring coil which is held by the fingers

of the other hand This maneuver separates the small

welded ball approximately 0.5 mm away from the

wind-ings of the coiled portion of the “spring wire”, but still

attached by a single, stretched-out, strand of the fine, very

stiff wire This separation of the welded “ball” allows the

bioptome to grasp the “ball” firmly while at the same time

allowing the bioptome jaws to close almost completely

over the stretched out strand of fine wire This allows the

outside diameter of the combination of the closed,

3-French bioptome jaws, over and holding the welded

“ball” of even the 0.052″ coil, still to be less than 4-French

in outside diameter

The closed bioptome jaws over the tip of the coil,

however, do not fit back within the original metal loader

The combination must be withdrawn into a separate

slightly larger “loader” in order to be front-loaded into the

delivery sheath Before the coil is grasped with the

biop-tome, the bioptome is passed through the separate slightly

larger metal loading tube (which comes with the 0.052″

coils) or through a segment of a 4-French short sheath

which can be used as a loader equally as well as the new,

larger metal loading tube which is provided only with the

0.052″ coils The 4-French sheath must be long enough

to contain the entire straightened length of whichever coil

is being used After the bioptome is passed through the

separate metal loading tube or the 4-French sheath and

during the attachment to the “prepared coil”, the metal

loading tube or length of sheath is withdrawn on the

biop-tome shaft back to its proximal control handle Once the

new “loading tube” is back on the shaft of the bioptome

catheter, the bioptome jaws are opened and then closed

tightly around the “prepared” separated, welded ball at

the end of the coil

With one operator/assistant holding the bioptome jaws closed tightly over the “ball” at the tip of the coil, the “new”, larger diameter “loading tube” through which the bioptome was previously passed, is advancedover the bioptome jaws which now have the tip of the coil grasped within them The entire coil is drawn out

of the original metal loading tube directly into the newlarger “loading tube” or segment of short sheath The coil is withdrawn completely into this new tube until

the now distal (originally proximal) end of the coil is just within the tip of the new loading tube The bioptome-

controlled coil is now ready to be delivered from the newloading tube into the pre-positioned delivery sheath Thedistal end of the loading tube is introduced into the back-bleed valve at the hub of the pre-positioned deliverysheath and the bioptome catheter (with the attached coil)

is advanced into the loader/sheath and the shaft of thedelivery sheath

The bioptome jaws are held closed very tightly whilethe coil is advanced to the end of the delivery sheath Theentire delivery is observed closely on fluoroscopy Thecoil and the bioptome jaws can be visualized very clearly

on fluoroscopy The sheath is withdrawn slightly until thetip of the sheath is exactly in position for implanting the coil The bioptome catheter and coil are advancedtogether, extruding the coil out of the tip of the sheath Thecombination bioptome/coil is advanced together with the sheath held in position until the bioptome jaws with

the grasped coil are just 1–2 mm within the tip of the sheath

At this point the security of the extruded coil is tested byvery slight, gentle, to-and-fro movement of the sheath and bioptome catheter together The degree of occlusioncan be tested by a contrast injection into the involved ves-sel through a second catheter This obviously can only beaccomplished if such a catheter is in place If not satisfiedwith the fixation of the coil in the vessel or the degree ofocclusion, the coil can easily be withdrawn into (and, ifdesired, out of) the sheath, and the procedure restarted in

a new location of the delivery sheath or with a differentcoil When satisfied with the fixation in the vessel and thedegree of occlusion, the bioptome catheter is advanceduntil the jaws have advanced completely out of the tip

of the sheath The bioptome jaws are opened, releasing the coil

The advantages of the bioptome-controlled delivery ofcoils are the obvious complete control over the actualrelease of the coil and the complete retrievability of the

coil until the moment of its purposeful release The major

disadvantages to this technique are the requirement for

a slightly larger delivery system, the lower flexibility ofthe long sheath delivery system for entering more tortu-ous locations or vessels arising at acute angles off themajor feeding vessel (aorta), and the added expense of thebioptome

Special commercial modifications of the delivery

system or the coil to control the delivery of the

Gianturco™ type coil

Detachable™ coils: Jackson™ coils, Cook detachable

and Flipper™ coils

Outside the United Statesai.e in countries not under US

FDA jurisdictionathe Jackson™ (or Detachable™) coil is

available commercially from Cook™ Inc., Europe This is

a safe, detachable or controlled-release variation of the

standard Gianturco™ coil The Detachable™ coil is in

widespread, standard, use for PDA occlusion as well as all

other types of vascular occlusions, particularly in Europe10

The European Detachable™ coil is not available in the

United States The Detachable™ coil is slightly less robust

than the standard Gianturco™ coil, however, for the

occlusion of most vascular lesions other than the PDA,

this relative flimsiness is of little consequence The

relat-ively simple attach/release mechanism of the Detachable™

coil gives complete control over the release and

retract-ability of the coil, making vascular occlusions in even the

most precarious vessels a safer and more effective

pro-cedure The Detachable™ coil, with its specific pusher/

delivery wire is, however, slightly more expensive than

the standard “free-release” Gianturco™ coil

The Detachable™ coil essentially is a standard

Gianturco™ coil that has been fitted with a screw

mechan-ism for the purposeful attachment and release from the

delivery or “pusher” wire The European Detachable™

coil comes from the manufacturer as a “set” containing the

coil, a special clear loader, a special delivery/pusher wire

and a fine movable mandril, which passes through the

delivery wire and the coil The clear loader has a slight

funnel at the proximal end where the “female” screw

mechanism in the coil is located The Detachable™ coil

itself outwardly has the same appearance as a standard

Gianturco™ coil, with the distal end of the occlusion coil

sealed with a rounded, smooth and polished “weld”

while the proximal end of the coil is hollow The open,

proximal end of the straightened occluder coil is “squared

off”, hollow and appears slightly irregular on very close

inspection The fine wire windings within the proximal,

hollow end of the coil form a “female screw thread” for

the attachment of the “pusher” wire The special attach/

release delivery or “pusher” wire consists of a spring

guide wire of the same diameter as the coil to be used The

pusher wire has a long, tapered “male” screw mechanism

as an integral part of its distal end and comes with a

removable “torquing” device for purposeful rotation

of the wire/screw tip The coil attaches to the delivery/

pusher wire with this very simple screw mechanism

In addition, the special delivery/pusher wire has no

fixed “core” wire, but is hollow throughout its entire length,

including through the fine screw at the distal end The

small lumen throughout the wire allows a very fine,smooth, steel “mandril” or stiffening wire to pass com-pletely through, out of the end of the pusher wire andthrough the coil The fine, totally removable “mandril”wire, which is approximately 10 cm longer than the com-bined pusher wire and the straight coil wire, comes pack-aged within each pusher wire The mandril passingwithin the delivery/pusher wire and the coil acts as a stiff-ening “core” wire The mandril has a short segment of

“spring” wire attached at its proximal end to serve toidentify the proximal end and as a “hub” for moving andtorquing the mandril Detachable™ coils commerciallycome stretched out as a straight spring wire positioned

within a thin, clear, straight loading tube of approximately

the same ID as the OD of the coil

Once the delivery catheter is positioned properly andsecured in place, the coil is attached to the pusher wire.The clear loader containing the occluder coil is inspectedvery carefully Each end of the occluder coil is positionedslightly more than one centimeter within the ends of theloader The proximal and distal ends of the occludercoil/loader are identified The screw end of the deliv-ery/pusher wire is introduced into the proximal (fun-neled) end of the loader The mandril wire is advancedthrough the delivery/pusher wire and 8–10 mm beyondthe distal tip of the “screw” mechanism of the delivery/pusher Very carefully and without pushing the coil for-ward in the loader, the mandril is introduced into theproximal end of the coil by gentle trial and error probing

and then advanced 8–10 mm into the hollow coil The

delivery/pusher wire is advanced over the mandril untilthe screw at the tip of the delivery/pusher wire enters into the proximal end of the coil When the screw tip hasengaged in the proximal end of the coil, the coil andpusher wire are attached by two and a half clockwiseturns on the pusher wire or on the coil loader when it isgripped facing the pusher wire The screw at the tip of thepusher wire enters and engages with the proximal end

of the coil Once the screw has tightened within the coil,

the delivery/pusher wire is turned counter-clockwise one half turn to slightly loosen the attachment Gentle, and very slight to-and-fro motion of the pusher wire within

the loader is used to test the attachment of the coil to thepusher wire Holding the pusher wire, the core/mandrilwire is advanced the remainder of the way into the coiluntil the mandril reaches the distal, closed tip of thestraightened coil within the loader

The distal end of the clear coil loader is introducedthrough a wire back-bleed/flush device into the hub ofthe previously positioned delivery catheter The deliv-ery/pusher wire with the enclosed mandril is advancedinto the loader, which, in turn, pushes the attached,straightened Detachable™ coil out of the loader and into

the delivery catheter The mandril remains in the

deliv-ery/pusher wire as the pusher wire is advanced The

distal end of the catheter is observed continually on

fluoroscopy as the coil advances through the catheter to

the tip of the catheter and as the coil is advanced out of the

tip of the catheter into the desired location for implant by

advancing the delivery/pusher wire further The mandril

is maintained within the delivery wire and coil until the

coil is at least partially out of the delivery catheter and the

tip of the coil is in or slightly distal to the correct position

for implant The mandril still within the coil keeps the coil

straight and allows easy adjustments in the position of the

coil Once the coil is partially out of the catheter and in

proper position for implant, the mandril is withdrawn

slowly, allowing the coil to curl into its “coiled”

configura-tion in the vessel/lesion The pusher catheter and coil are

advanced further while the mandril is withdrawn to just

within the proximal end of the coil and until the coil is totally

extruded from the catheter and formed into its full coil

configuration At this point the coil is tested for its proper

location and its fixation in the vessel by slight to-and-fro

motion of the pusher wire or with an angiogram When

satisfied with the fixation of the coil and the occlusion of

the vessel, the small “vise” is attached to the proximal end

of the delivery/pusher wire and the vise with the pusher

wire/mandril is rotated counter-clockwise until the coil

detaches from the pusher wire The release of the coil as

the coil is being detached is observed very closely under

fluoroscopy to ensure a smooth “unscrewing”, and that

there is no binding or twisting of the still-attached coil

If the Detachable™ coil is not in the exact, desired

posi-tion, at any time before it is purposefully “unscrewed”, it

can easily be withdrawn back into the delivery catheter

As the coil is withdrawn into the catheter, the mandril

remains positioned proximal to and outside of the coil

If the same coil is to be repositioned using the same

catheter, the mandril is re-advanced into the coil (which

now is straightened within the catheter) before the coil

is re-extruded into the vessel This retrievability during

delivery adds total control and a significant degree of

safety to the occlusion of any vascular structure The total

European Detachable™ coil systems are slightly more

complicated and significantly more expensive than the

standard Gianturco™ coils by themselves

Flipper™ coils

The Flipper™ coil (Cook Inc., Bloomington, IN) is

an attempt in the US at a version of the European

Detachable™ coil The earlier US version of a detachable

stainless steel coil of a size considered reasonable to use in

PDA occlusions was a catastrophe The original “larger”

detachable coil which became available in the US was

made from a smaller, less robust spring wire, had far fewer

incorporated “thrombotic” fibers, and had a bizarre,

unnecessarily complicated attach/release mechanism As

a consequence this coil had very little use The Flipper™coil is a revised version of the Cook, Inc., US detachablecoil, which became available in 2001 with an attach/releasemechanism similar to the European Detachable™ coils.The Flipper™ coil wires are 0.035″ wires and still do nothave comparable robustness nor the occlusive capabilities

of the 0.038″ standard Gianturco™ coils Flipper™ coilsare available in 3–12 cm lengths and with coil diametersbetween 3 and 8 mm Current Flipper™ coils have morefiber strands embedded in the coil wire than the original

US version of the detachable coil, and for a 0.035″coil doseem to have comparable occlusive properties to thesmaller 0.035″ standard Gianturco™ coils Flipper™ coilsrequire a delivery catheter with a 0.041″ inner diameter,which generally is a 5-French catheter

The lack of robustness of the 0.035″ coil compared to the0.038″ coil is a limitation of this coil for its use for primarilyclosing a patent ductus arteriosus, but the controllabil-ity and retrievability of the Flipper™ coil make it veryappealing for most other vascular occlusions or for theimplant of additional coils for the “final and total” occlu-sion of a patent ductus which already has a larger, securecoil in place

The packaging, attaching, loading, delivery and release

of the Flipper™ coils are all similar to the EuropeanDetachable™ coils described above

Alternative occlusion coils including “Micro” coils

In addition to the standard and large sized Gianturco™

type coils, there is a large variety of other small coils able for small vessel occlusion The alternative coils are

avail-delivered through significantly smaller and more flexiblecatheters, and as a consequence can be placed in more circuitous and distal locations None of the smaller alter-native coils are as robust as the stainless steel Gianturco™coils

Target™ platinum coils and their delivery technique

Target™ Coils (Target Therapeutics, Fremont, CA) are similar in concept to the Gianturco™ coils They aresegments of very fine spring guidewire-like wires withthrombogenic fibers intertwined in the spaces betweenthe wire coils of the spring wire There, the similarityends Target™ coil wires are made of very small diameter0.014″ and 0.018″ platinum wires In spite of their verysmall diameters, and because of the platinum material,these coils are very radio-opaque and easily visible underfluoroscopy in the catheterization laboratory Because ofthe material and tiny size of these coils, they are moreflexible and pass through tortuous catheters/vessels eas-ier than Gianturco™ coils Target™ coils do not open into

a cylindrical “coil” in their relaxed state but rather, into

complex “helix” configurations which vary between 4

and 7 mm in diameter Because of their size and less

“fiber material”, these coils are more compressible in a

particular vessel and, in turn, not as occlusive as the larger

stainless steel Gianturco™ coils

Target™ coils are delivered through specific, very

small (3-French), very flexible Tracker™ delivery catheters

(Target Therapeutics, Fremont, CA) The Tracker™

catheters are small and “trackable” but are

non-radio-opaque throughout their entire length except for a tiny

radio-opaque marker at the very tip of the catheter This

tip marker identifies the most distal location of the tip

but, once the delivery/pusher wire is removed from the

Tracker™ catheter, the course of the catheter proximal to

the tip is invisible! Tracker™ catheters are advanced to the

orifice of the vessel to be occluded through a larger

torque-controlled 5-French, guiding catheter From the guiding

catheter, the Tracker™ catheter is advanced through

the vessel over a special, very fine, Dasher™ guide wire

(Target Therapeutics, Fremont, CA) The Dasher™ guide

wire is pre-positioned in the target vessel Because of its

torque control, its very floppy tip and its very small size,

this wire can be manipulated very readily to very distal

locations and through very tortuous and long vessels

To deliver the Target™ coils, a pre-formed,

torque-controlled, 5-French guiding catheter which accommodates

the Tracker™ catheter is positioned in the orifice of the

vessel to be occluded or, alternatively, into a trunk vessel

off of which the target vessel arises The special Dasher™

tracking wire, which is 0.014″ in diameter and 175 cm long

with a very flexible tip, is passed through the guiding

catheter and into the target vessel The specifically curved

floppy tip of the Dasher™ wires can be manipulated

select-ively through very tortuous and very circuitous, distal

vessels using a “torquer vice” on the stiff shaft of the

wire to help to direct the tip of the wire to the specific

loca-tion An appropriately sized Tracker™ (“Tracker™-18”,

150 cm) catheter (Target Therapeutics, Fremont, CA) is

then advanced over the Dasher™ wire and through the

guiding catheter to the orifice of the target vessel As the

Tracker™ catheter is being advanced over the wire a

continual flush is maintained through the guide catheter

and the Tracker™ catheter through the special double

“Y” adaptors attached to the proximal ends of both the

Tracker™ and the guiding catheters

When the Tracker™ catheter is being advanced over the

wire beyond the tip of the guiding catheter and through

the more tortuous areas of the vessel, the Tracker™

catheter is advanced several centimeters at a time while

alternately withdrawing or tightening the Dasher™

wire very slightly In this way the Tracker™ catheter is

“inched” along the Dasher™ wire without displacing the

wire All of this time, only the tip of the Tracker™ catheter

is visible over the wire Occasionally the Tracker™catheter does not follow over the Dasher™ wire and thewire is pulled out of position In that circumstance, oftenthe very flimsy Tracker™ catheter with just the tip of the exposed, very floppy Dasher™ wire maintained justbeyond the tip of the catheter can be advanced (manipu-lated) together as a unit back to the appropriate location.Once the Tracker™ catheter has reached the desired loca-tion, the Dasher™ wire is withdrawn very carefully andvery slowly from the Tracker™ catheter The Tracker™catheter is maintained on a slow constant flush while theopaque tip of the catheter is observed continuously onfluoroscopy to be sure that the tip is not being displacedduring the removal of the wire Once the wire is removed,

again, only the opaque marker at the tip of the Tracker™

catheter will be visible A small, slow hand injection of

contrast is performed through the Tracker™ catheter andrecorded on biplane angiography or stored fluoroscopy inorder to “road map” the actual “course” of the Tracker™catheter for future reference After the “road map” is

recorded, the Tracker™ catheter is flushed very slowly but

very thoroughly to clear it of all contrast material A rapid

or forceful flush of the Tracker™ catheter can easily “blow”the tip of the Tracker™ catheter back out of the vessel.Each Target™ coil comes straightened in a separate,tiny, metal tubular holder/introducer The introducertube is flushed gently and the end, which does not have ahub, is fitted into the straight segment of the “Y” connec-tor at the proximal end of the Tracker™ catheter The coil

is pushed out of the introducer and into the proximal end

of the Tracker catheter with the short “plunger tool”which comes with the coil Once the plunger tool has beenadvanced to the “hilt”, the coil is completely within the

Tracker™ catheter as a short, straight, free segment of

wire Target™ coils fortunately are very radio-opaque.The plunger tool and the introducer tube are removed

The stiff end of the special “coil pusher wire” is introduced

into the proximal end of the Tracker™ catheter and

advanced approximately 30 cms This “pre-advances” the

coil that distance within the Tracker™ catheter The stiffend of the coil pusher is withdrawn, the pusher wirereversed, and the soft nylon distal end of the pusher

“wire” introduced into the Tracker™ catheter All of this

time the Tracker™ catheter and guiding catheter are tained on a slow continuous flush As the coil pusher (soft

main-nylon end first) is advanced in the Tracker™ catheter, thecoil is advanced proportionately through the catheter.This part of the delivery is observed particularly carefully

on the fluoroscopy Extreme care is taken as the coil isadvanced within the Tracker™ catheter to ensure that the tip of the Tracker™ catheter is not withdrawn evenslightly by inadvertent traction applied to it When theadvancing coil and coil pusher wire begin to enter curvesand bends in the course of the Tracker™ catheter, the

advancing coil and coil pusher wire tend to “straighten”

the Tracker™ which, if not compensated for by advancing

the proximal end of the Tracker™ catheter, can pull the tip

of the Tracker™ catheter back out of the desired position

The course of the coil, which now is visible through

the “invisible” catheter, is compared to the previously

obtained angiographic “road map” of the catheter course

This is to help ensure that the coil and pusher are not

dis-placing the “invisible” catheter during the introduction

of the coil

When the coil reaches the tip of the Tracker™ catheter,

by further advancing the coil pusher, the coil is extruded

from the distal end of the catheter and is free in the vessel

to be occluded Like the standard Gianturco™ coils, the

Target™ coils have no attachment to the pusher wire

and once extrusion begins, the coil cannot be withdrawn

Also like the Gianturco™ coils, the way the coils unwind

and coil in the vessel depends upon the relative size of the

coil compared to the vessel size In order to accomplish

complete occlusion of a vessel, usually many coils or

mul-tiple sizes of the coils are used As with other occlusions

and whenever possible, the total occlusion of the vessel

should be accomplished completely before the procedure

is abandoned

Controlled release micro coils

There are two varieties of controlled release systems for

the “micro” coils, which do provide retrievability of the

coils even after they have started to be extruded The two

mechanisms are entirely different, but both are

manu-factured by Target™ Therapeutics (Target Therapeutics,

Fremont, CA) Although the primary use of these coils is

by neuroradiologists, they are equally suitable for small

vessel occlusions in the pediatric/congenital population

The first of these controlled release micro coils is

the Guglielmi™ electrolytically detachable coil (GDC)

(Target Therapeutics, Fremont, CA) These are, as the

name implies, released from the delivery wire by a micro

current delivered through the wire, which, in effect, melts

a connection between the delivery/pusher wire and the

coil The second type of controlled release coil is the

Interlocking Target™ coil (Target Therapeutics, Fremont,

CA) These coils have a unique system of micro machined,

overlapping pins or “couplers” which are compressed

within a tiny delivery catheter until the final millimeter of

the coil is extruded from the tip of the catheter Both of

these controlled release micro coils otherwise are used in

similar circumstances to the other micro coils

Tornado™ coils

The Cook Tornado™ coils (Cook Inc., Bloomington, IN)

functionally are almost a cross between the standard

Gianturco™ coils and the Target™ coils As their nameimplies, these coils are shaped like a tiny tornado “funnel”which along with their size and material gives them morecompressibility and, supposedly, more coil “exposure” tothe lumen of the vessel These coils are constructed of plat-inum wire, which is more radio-opaque than stainlesssteel wire of comparable size The coil wires are available

in both 0.018″ and 0.025″ spring wire sizes, which, in turn,are less robust than the standard Gianturco™ coils Eachcoil has multiple strands of tiny synthetic fibers embed-ded along the coil wire to facilitate thrombosis TheTornado™ coil sizes are labeled corresponding to thelargest and the smallest diameters at the ends of the “fun-nel” For example, a 6/2 coil has a 6 mm diameter largeend, which tapers down to a 2 mm diameter small end.The 0.018″ Tornado™ coils are available in sizes from 3/2

to 10/4 with lengths of the straightened coil varying from

2 to 14 cm, while the 0.025″ Tornado™ coils come in sizesfrom 5/3 to 10/5, with lengths of the straightened coilfrom 4 to 12 cm From the standard packaging of theTornado™ coils, the small end of the “funnel” is deliveredfirst from the “coil” holder and loads first By special orderthe Tornado™ coils can be packaged so the large end isdelivered first Tornado™ coils are delivered throughcatheters with internal diameters of 0.025″ or 0.032″,which allow the use of very small, trackable catheters.The 3-French Slip-Cath™ (Cook Inc., Bloomington, IN)

is ideal for the delivery of these coils The Slip-Cath™ isvery flexible and has a very slick hydrophilic coating,which allows it to track through very circuitous and smallvessels Its small outer diameter allows it to be advanced

to the orifice of the vessel that is to be occluded through atorque-controlled, 5- or 6-French guiding catheter Thenylon 3-French infusion catheter (Cook Inc., Bloomington,IN) with a radio-opaque tip is less flexible than the Slip-Cath™ but also makes a good delivery catheter for delivery

to more proximal areas in small, but less tortuous vessels.The technique for loading and delivering theTornado™ coil is similar to the delivery of a “free-release”Gianturco™ coil There is no attach/release mechanism.The coils are introduced into the pre-positioned deliverycatheter from their loader and advanced through thecatheter with a 0.025″ pusher wire When there is a verytortuous vessel proximal to the site of delivery, a very softtipped pusher wire is used A concomitant slow, continu-ous flush through the delivery catheter assists in advanc-ing the coil through the catheter The smaller Tornado™coils occasionally can be advanced through the deliverycatheter by a strong flush on the catheter alone The strongflush, however, may cause the tip of the catheter to recoilout of position, making the “flush” delivery much lessreliable and less precise As with other micro coil occlu-sions, usually more than one coil essentially is alwaysrequired to complete the occlusion After one or more

coils are delivered, a small hand angiogram is performed

through the delivery catheter to check the degree of

occlu-sion If there still is any residual leak, additional coils are

delivered until the vessel is occluded or there is no more

space in the vessel for additional coils

Nester™ coils

The Nester™ coils (Cook Inc., Bloomington, IN) are still

another variety of coil for vascular occlusions These are

0.035″ platinum coils with synthetic fibers entwined in the

coil wire In their relaxed, free shape, the coils have a

cylindrical configuration Nester™ coils are all 14 cm in

length and available in 4–12 mm diameter “cylinders”

These coils are softer than even the comparable diameter

Tornado™ coils The Nester™ coils are designed to

occlude by bunching up and packing into a relatively

large (1–2 cm) mass within the vessel

The Nester™ coils are delivered through a catheter with

a 0.035″ or 0.038″ inner diameter and advanced through

the delivery catheter with a 0.035″ teflon-coated spring

guide “pusher” wire identically to the delivery of the

Gianturco™ free-release coils The standard Nestor™

coils have no attach/release mechanism so, like the

Gianturco™ coils, once extrusion of the coil begins, there

is no turning back Because they are so soft and flimsy,

these coils are used mostly as additional or supplemental

“stuffing” coils rather than primary occlusion devices

They are added as “packing” to complete the occlusion

within a previous, stiffer coil or other device already

placed in a vessel

Particulate materials and non-coil intravascular

devices available for vascular occlusions

There are several materials and intravascular devices in

addition to the intravascular coils that are available and

occasionally used for the occlusion of abnormal vascular

communications

Particulate materials

Vascular radiologists use particulate materials

exten-sively for the acute occlusion of vessels and abnormal

communications These materials usually are used to

control bleeding from a specific vessel and/or to totally

occlude vessels with the intent of necrosing the tissues

“downstream” from the occlusion (tumors, neoplasms)

The particulate materials are useful mostly for very

dif-fuse, but, at the same time, “end artery” lesions The

mater-ials used include autologous blood clots, pieces of

Gelfoam™ and pieces of Polyvinyl Alcohol (Ivalon™)

The autologous blood clots and Gelfoam™ usually only

provide a temporary occlusion and for that reason seldom

are used in congenital vascular lesions where a more manent” occlusion is desired

“per-The autologous clots are just thatasmall solid particles

of the patient’s recently clotted blood, which are injectedfrom a syringe through a catheter that is pre-positioned

in the culprit vessel1 The “clot particles” that are usedshould be slightly larger than the vessel that is to beoccluded The already clotted blood acts as a temporaryocclusive mass to occlude the small, preferably “end” ves-sel The clot usually thrombolyses over a short period oftime, however, it usually remains in place long enough toallow acute bleeding from the vessel to stop and occasion-ally long enough for the vessel to thrombose

Gelfoam™ is an insoluble material made from driedpork skin gelatin and formed into porous foam sheets Thefoam material can absorb many times its own weight

in blood and/or other fluids When left in the tissues

it resorbs into the body within 4–6 weeks Gelfoam™ is

approved to enhance coagulation on the surface of

bleed-ing tissues but is not “approved” for intravascular use Itcomes in sterile sheets, which can be cut to any desiredsize or shape, and the pieces can be compressed into verysmall particles

Very small, cut and compressed pieces of Gelfoam™are soaked in a dilute solution of contrast and then areforced into the end vessel that is to be occluded through acatheter by a strong flush with a syringe2 In the vessel, the

“wad” of Gelfoam™ expands and absorbs blood to create

an occlusive mass Gelfoam™ is useful to “complete” or

“finalize” the occlusion of a vessel/fistula which has been started with another occlusion device Obviously,this soft gelatin like material cannot be used to occludehigh-flow and/or high-pressure vessels, and like autolog-ous clot, the Gelfoam™ itself may not create a permanentocclusion The Gelfoam™ particles themselves are notradio-opaque, so embolization to distal locations is onlyapparent from any contrast solution that might beretained in the particles and/or signs or symptoms thatare produced by the embolized particles

Ivalon™ (Polyvinyl-alcohol or PVA) foam particles are available for a similar use to the autologous clots orGelfoam™ pieces The Ivalon™ does provide a more per-manent occlusion but, as small soft particles, still is usefulonly for end vessel and/or “completion” of other vesselocclusions Ivalon™ particles are available commerciallyfrom Cook Inc (Bloomington, IN) as dried particles inmultiple different particle sizes between the smallest50–100 micron and the largest 2000–2800 micron sizes toaccommodate multiple different vessel sizes The driedIvalon™ particles are mixed with contrast material tosoften them and give them some radio-opacity for injec-tion into the circulation The particles mixed with dilutecontrast are drawn out of their sterile container into

a syringe and injected with the same syringe into the

desired site through a pre-positioned catheter Similar to

the autologous clots and Gelfoam™, there is little control

over the location where the particles are delivered and

essentially, no retrievability

“Intravascular” glues

There was a transient interest by several different

investigators for the use of the tissue adhesive isobutyl

2-cyanoacrylate (Bucrylate™) as an occlusive/embolic

material to occlude large vessels/vascular

communica-tions4 The Bucrylate™ polymerizes into a solid mass

instantaneously on contact with blood The area for

deposit of the Bucrylate™ must be “isolated” from the

surrounding circulating blood and/or the Bucrylate™

must be injected very specifically into the area

Bucrylate™ was very difficult to handle and to control

during delivery both during animal testing and during

several clinical uses If injected too fast it embolized

dis-tally and/or backward into proximal branches and, in

doing so, occluded all areas that it entered If injected

too slowly, it occluded the injecting catheter when only

partially extruded The difficulties with delivery of

Bucrylate™ compared to other occlusion materials/

devices, and the unknowns about long-term carcinogenic

effects of Bucrylate™ in humans, led to it being

aban-doned for human trials in the United States

Occlusion devices for large vascular

communications

Gianturco-Grifka Vascular Occlusion Device™

(GGVOD™)

Although it does contain spring guide wire and is

ap-proved for use in the US as a variation of the Gianturco™

coil, the Gianturco-Grifka Vascular Occlusion Device™

(GGVOD™) (Cook Inc., Bloomington, IN) has little

resem-blance in either appearance or use to the Gianturco™

coils The GGVOD™ is a nylon bag or sack of a

predeter-mined fixed diameter into which a specific length of

spring wire is wadded to achieve a tense, fixed diameter,

mass of spring wire11 The wire for the packing of the bag

of the GGVOD™ is a spring guide wire with the stiffening

core and safety wires removed The bag only serves to

contain the mass of wire within the fixed diameter of the

bag The bags are available in diameters of 3, 5, 7 and

9 mm, each coming with a specific length of “packing”

spring guide wire The bags are somewhat flattened and

elongated so that they do not form a circular or spherical

configuration in their length or cross section The stated

diameter is the largest measured cross-section of the

par-ticular bag and not the actual circumferential diameter

of the bag

The GGVOD™ is usable only in tubular vascular tures that are at least 1.5 times longer than their respectivediameters The GGVOD™ is held in the vessel by theradial force created by the wad of spring wire within the bag and exerted against the surrounding vessel wall.The bag diameter should be at least one millimeter largerthan the diameter of the vessel to be occluded If the par-ticular vessel is very distensible, then a larger diameterbag to vessel ratio is used

struc-The technique for delivery of the GGVOD™ to mostabnormal vascular communication is similar to the deliv-ery of the GGVOD™ to the PDA (as described subse-quently in Chapter 27) with one major exception Whendelivering the GGVOD™ to a collateral or branch vesselcompared to the delivery to the PDA, the delivery sheathinitially and, in turn, the bag is delivered exactly to the

implant site and all of the wire extrusion into the bag is

directly into the site where the bag is to be implanted This

is in comparison to the delivery to the usual patent ductus,

where the sheath and the bag are advanced beyond the

implant location and some of the filler wire is extrudedinto the bag in this distal location before the bag/deliverycatheter is withdrawn back into the specific site within theductus for the implant The fixation of the bag and/or theability to push all of the filler wire into the bag depend(s)upon a relatively precise and, at the same time, tight fit ofthe filled bag into the vessel/channel As with the delivery

of coils to critical locations, the vessel/channel can be

“sized” with a small angioplasty balloon that is slightlylarger than the vessel/channel and inflated at a low pres-sure in the vessel/channel

Because of the complexity of the delivery and release

of the GGVOD™ and the critical importance of each vidual step, the details of the GGVOD™ delivery to vas-cular communications other than the PDA are listed here

indi-in a tabular, “cook book” form

Components of the GGVOD™ system

• A long, 8-French, valved outer delivery sheath with adistal marker band

• The nylon “sack” or “bag”: in 3, 5, 7 and 9 mm sizes(diameters)

• An inner, stiff walled, pusher catheter to which the bag

• A second, stiff, middle, “bag release” catheter which ispre-positioned over the pusher catheter

• A complex attach/release system which joins all ofthese components

• A short length of 8-French “peal away” sheath which

comes over the distal ends of the two catheters and the

attached “bag”

Six major steps in the implant of a GGVOD to an

abnormal vascular communication (except a PDA)

• The long delivery sheath is advanced into the abnormal

vascular communication until the sheath tip is just within

the area where the “sack” is to be implanted

• The “sack” is introduced into the long sheath through

the “peal away” sheath

• The “sack” is positioned precisely in the area to be

occluded and the sheath is withdrawn off the sack

• The filler wire is pushed into the sack

• The pusher wire is separated from the filler wire

• The filled “sack” is separated from the pusher catheter

Details of the six major steps for the GGVOD

delivery to an abnormal vascular communication

Step 1: The long delivery sheath is positioned in

vascular communication with the tip of the sheath

1–2 cm beyond the area where the sack is to be

implanted

• The abnormal vascular channel is entered with an

end-hole catheter which is advanced as far beyond the area to

be occluded as possible

• A second catheter from a second vascular entrance site

is placed adjacent to the proximal (in the direction of flow)

end of the communication

• The end-hole catheter is replaced with a 0.035″ stiff

exchange wire

• The special 8-French, valved delivery sheath/dilator

set with a marker band at the distal tip of the sheath is

advanced over the wire into the abnormal channel until the

tip of the sheath is just beyond the area of the channel where

the GGVOD™ is to be implanted Unless the delivery

sheath tip can be advanced just distal to the location for

occlusion, the GGVOD™ cannot be used

• The dilator and wire are removed, the system is

cleared of air and clot and then the sheath is flushed

thoroughly

Step 2: The “sack” and delivery system introduction

into the long sheath

• All of the components of the GGVOD system are

inspected carefully

• The “sack”, which is attached to the pusher catheter

and the wire/delivery/release system, is introduced into

the pre-positioned long sheath as a single unit through the

short “peal away” sheath

• The short “peal-away” sheath is removed after the

“sack” is completely within the sheath

• The “sack” is advanced with the attached delivery system

to the tip the sheath (but still within the sheath tip) to theexact location where the GGVOD™ is to be implanted

Step 3: The “sack” is positioned in the vessel/

channel

• With the pusher catheter held in place, the sheath iswithdrawn off the “sack” which now is in the precise position for implant The “sack” itself is invisible, but the attaching band on the proximal neck of the “sack” isvisible and defines the proximal limit of the “sack”

• While observing on fluoroscopy, several loops of the

“filler” wire are advanced very loosely into the “sack” by

advancing the stiff pusher wire 5–10 cm The “sack” is not

packed tightly with the filler wire at this point

• The position of the “sack” is checked with a smallangiogram through the second catheter positioned in oradjacent to the proximal end of the channel

Step 4: The filler wire is pushed into the sack

• With the “sack” in the proper position, the remainder

of the filler wire is fed completely into the “sack” This is

checked on fluoroscopy being sure that the connection

point of the filler wire with the pusher wire is within the sack This is identified by a difference in X-ray densities at

the connection point

• The exact “sack” position in the vessel and the degree ofocclusion are checked by another angiogram through thesecond catheter

Step 5: The pusher wire is separated from the filler wire

• When sure the filler wire is completely within the

“sack” and while observing on fluoroscopy, the wire

release mechanism is activated by pushing forward

on the two side “loops” of the special attach/release handle This detaches the pusher wire from the filler wire

• The pusher wire is withdrawn into the pusher catheter athis releases and separates the pusher wire from the

“sack” This withdrawal of the pusher wire should be

very slowauntil absolutely sure that the pusher wire is

separated from the filler wire

Step 6: The pusher catheter is separated from the sack

• The stiffer, middle, release catheter, which is over the

pusher catheter but within the sheath, is advanced snuglyagainst the neck of the “sack”

• With the tip of the release catheter held exactly in place

against the sack, the inner “pusher” catheter is withdrawn

forcibly from within the neck of the “sack” (If the release

catheter moves at all, the position of the “sack” very likely

will move with it!!)

• The forceful withdrawal of the “pusher catheter”

releases the “sack” in the channel entirely free from the

delivery system

The advantages of the GGVOD™ for vascular occlusions

are the same as its advantages for occlusion of the PDA

The GGVOD™ is excellent for the occlusion of relatively

large and/or high flow vascular channels or structures

that have any associated length In this type of lesion, the

GGVOD™ usually produces immediate, complete

occlu-sion Even the largest “sack” is delivered through only an

8-French sheath, which is reasonable in all patients past

infancy Once delivered and before it is released

purpose-fully, if the “filled bag” does not fit in the vessel, is loose

in the vessel or otherwise does not occlude the vessel, the

wire can be withdrawn and the bag repositioned or even

replaced with a bag of a different size The larger diameter

GGVOD™ devices also are applicable to some larger,

high flow vessels and channels where coils alone are not

sufficientafor example in large abnormal venous

chan-nels or large pulmonary arteriovenous malformations

One of the greatest advantages of the GGVOD™ is that

it is approved for this use and is available in the United

States

The GGVOD certainly is not a “universal” device,

only being applicable to tubular vascular lesions that are

somewhat longer than they are in diameter Another

dis-advantage of the GGVOD for vascular occlusions relates

to the size and stiffness of the delivery sheath Since most

small vessels requiring occlusion arise from the aorta and

require arterial access, the 8-French delivery sheath

repres-ents a significant disadvantage for use in infants or small

children In addition to its diameter, the GGVOD™ is

delivered through a long sheath which is not as flexible as

most delivery catheters, so delivery of the device into

acutely angled and/or tortuous vessels is difficult, if not

impossible Replacing the delivery sheath that comes with

the GGVOD™ with a Flexor™ type sheath overcomes

some of this problem A final disadvantage is that the

GGVOD™ is quite complicated to use and when used

infrequently, the delivery/release technique must be

“relearned” with each use

There have been rare complications encountered with

the use of the GGVOD™ even in the occlusion of vascular

structures other than a PDA On several occasions the

filler wire either could not be or was not pushed entirely

into the “sack” and/or was pulled partially out of the

“sack” as the pusher wire was withdrawn (incompletely

released?), resulting in a segment of the filler wire

extend-ing into the lumen of the vascular channel after the release

of the “sack” In most occlusions other than the PDA, thiscreates little or no problem since the goal is to occlude theentire channel If the filler wire extends back into a neces-sary or vital feeding channel and cannot be “wadded”back into the channel that is being occluded, the wire can

be captured with a vascular snare The wire can be bentback and forth repeatedly against the orifice of the targetvessel and, in doing so, broken off from the more distalwire, which is within the target vessel If the wire cannot

be broken, the entire wire can be withdrawn from the sackand out of the body This, unfortunately, leaves the “sack”empty and often destined to embolize to a more distallocation Under these circumstances and before with-drawing the wire completely out of the “sack”, the “sack”should be grasped with a separate retrieval device and/orthe vessel distal to the “sack” selectively occluded with atemporary occlusion balloon unless it is determined thatthe more distal vessel is expendable if permanentlyoccluded with the embolized empty sack

When undersized and/or in a very compliant vessel,the entire, full “sack” can embolize distally from itsdesired site in the abnormal channel Occasionally thismerely occludes the desired channel in a different, but still effective location If, however, the embolized “sack”migrates into the central or other vital areas of the circula-tion, the embolized sack must be removed, which cannot

be accomplished as a full sack First, the full bag is snared

to hold it in position If the neck of the sack can be snared,

it should not be pulled with any force Unusual tension onthe neck can detach the only radio-opaque part of the bagitself!! Once the bag is grasped, a hole is “chewed” in thebag with a bioptome forceps introduced from a separatevessel and through a slightly larger sheath Once a hole

is created in the full “sack”, loops of the packed filler wire usually will extrude immediately through the hole,and/or while “chewing” the hole, a loose portion of thefiller wire becomes exposed and is grabbed The graspedwire then is carefully withdrawn with the forceps whilethe “sack” is held with the snare or other retrieval device.The emptied bag is then withdrawn utilizing the devicethat is holding the sack

In spite of the disadvantages and rare complications,the GGVOD™ has some very specific uses and it should

be available to any laboratory heavily engaged in peutic catheterizations

thera-Spring wire alone

The steel wire of a spring guide wire itself is genic This property of the wire causes unwanted throm-bosis on the wires within the blood stream and withinsheaths and catheters Thrombosis on spring guide wiresrepresents a continual potential for embolic complications.This same property of the wire can be used purposefully

thrombo-to promote thrombosis and/or occlusion in very large

abnormal chambers (aneurysms) or channels, either alone

or packed on top of other occlusion devices In order for a

spring guide wire to be used as a stimulant to thrombosis

and vascular occlusion, the wire should be capable of being

wadded into a fairly compact mass Flexibility of the

wire is accomplished by removing the straight, stiffening

safety core wire(s) from inside of a standard stainless steel

spring guide wire Once the safety core is removed from a

spring guide wire, the remaining spring wire becomes

very soft and can be compacted easily into a tight mass

Once a guide wire has the core wire removed, several

meters of 0.025–0.038″ spring wire can be delivered into a

2–3 cm diameter space Once packed within any confined

area that has little or no flow within it, such as a

peduncu-lated aneurysm, thrombosis and obliteration of the area

occur very effectively

Like a “free-release” coil, long spring wires used in this

fashion cannot be retracted back into the delivery catheter

once the extrusion of the wire has started If part of the

wire inadvertently protrudes out of the specific area being

“packed”, usually the loose end of the wire can be

recap-tured with a snare catheter and can be withdrawn

com-pletely if it cannot be positioned ideally

Amplatzer™ Vascular Plug

The Amplatzer™ Vascular Plug (AGA Medical Corp.,

Golden Valley, MN) is the latest addition to the

armamen-tarium for the occlusion of abnormal vascular

commun-ications and is available even in the United States This

unique device is still another modification of the Nitinol™

wire weave or “basket-like” occlusion devices already in

common use for the occlusion of the patent ductus, atrial

septal defects, patent foramen ovale and ventricular

sep-tal defects The Amplatzer™ Vascular Plugs are small,

cylindrical, Nitinol™ wire mesh “plugs”, which are

avail-able in various diameters, but have no flanges or retention

disks at either end of the “plug” The plugs are

manufac-tured of 144 strands of a finer, 0.004″ Nitinol™ wire

Unlike most of the other Amplatzer™ occlusion devices,

the vascular plugs have no polyester disks within them

and, in turn, rely on the fine mesh of the Nitinol™ metal

mesh for occlusion of the vascular structure Similar to

the other Amplatzer™ devices, the plugs do have similar

metal posts or markers at the center of each end of the

plug, with a female micro screw within the post at

the proximal end of the device The female micro screw

allows attachment with an identical screw mechanism to a

standard Amplatzer™ delivery cable The plugs achieve

their fixation in the particular vessel by the tension against

the wall of the vessel by the expansion of the device

against the wall of the vessel/structure similarly to the

other Amplatzer™ devices

The plugs are available in diameters from the smallest

of 4 mm increasing in 2 mm increments up to a maximum

of 16 mm in diameter The 4–10 mm plugs are 7 mm longwhile the 12–16 mm plugs are 8 mm in length The finermetal wires and the lack of polyester disks allow the plugs

to be implanted through smaller delivery systems thanthe comparably sized Amplatzer™ PDA and VSD occlu-sion devices The 4–8 mm diameter devices are deliveredthrough a 5-French, 0.056″ internal diameter sheath, the

10 & 12 mm devices through 6-French, 0.067″ internaldiameter sheaths and the 14 & 16 mm devices are deliv-ered through 8-French, 0.088″ internal diameter sheaths.The Amplatzer™ Vascular Plugs are usable for theocclusion of multiple different venous and arterial struc-tures, all of which, however, must have some “tubular”configuration This includes residual Blalock–Taussigtype shunts, the tubular PDA, systemic to pulmonary collaterals, systemic to pulmonary vein communications,pulmonary arteriovenous fistulae, and even some peri-valvular leaks The diameter of the plug that is usedshould be 2–3 mm larger than the diameter of the lumenthat is to be occluded The Amplatzer™ Plugs, like theother Amplatzer™ devices, have the advantage of beingfully retrievable until they are purposefully released,which makes a test occlusion with the device possiblebefore the operator is committed to releasing the plug

In addition to “the tubular” characteristics of the lesion,the use of the Amplatzer™ Vascular Plug depends uponthe ability to maneuver the delivery catheter into (andpast) the lesion that is to be occluded The current deliv-ery sheath along with the relatively stiff delivery cable precludes the use of the plugs distally in very tortuouslocations, while, at the same time, the retrievability of theplug allows its placement more proximally in tortuousvessels where other non-controlled release devices would

abnor-is dabnor-iscussed in detail for its use for PDA occlusion inChapter 27 Unfortunately, this device is only available

outside of the United States and in a limited clinical trial in

the US Nit-Occlud™ devices are tightly wound coils ofNitinol™ wire, which are pre-shaped to conform to sev-eral different shapes of “typical” conical PDAs Like itspredecessor, the Duct-Occlud™ device, the Nit-Occlud™devices have no fibers intertwined in their coil windings

to help promote thrombosis/occlusion Nit-Occlud™

devices depend entirely on the tightness of the coil

con-figurations, the mass of Nitinol™ wire, and the

thrombo-genicity of the metal itself to occlude the vessels The

precise control/release mechanism of the Nit-Occlud™

device makes it useful in situations in vascular channels

where a precise and very controlled positioning is

essen-tial Like the bioptome control for the delivery of standard

stainless steel coils, the attach/release system for the

Nit-Occlud™ makes the delivery system somewhat stiffer

than the usual Gianturco™ type coil delivery catheters

which, in turn, makes these devices less useful for lesions

in more tortuous vessels and/or circuitous locations Where

it is available outside of the US (and hopefully sometime

eventually in the US) the Nit-Occlud™ device should be

considered for the occlusion of larger and unusual and/or

abnormal vessels/vascular communications/leaks where

the placement of the occluder is very precarious

Catheter delivered detachable vascular occlusion

balloons

Several types of detachable occlusion balloon were

avail-able around the world and even in the United States in the

past Although still available in some locations around the

world, none of the detachable occlusion balloons are

available any longer in the US market

B-D Mini-Balloon™

The B-D Mini-Balloon™ occlusion device (Becton Dickson

Co.) probably had the widest use in congenital heart

defects when it was available12 These were very tiny

occlusion devices The balloons were 1 mm in diameter

when deflated and up to 5.3 mm in diameter when

inflated They were delivered “hydraulically” by a unique

(and complex) delivery system The deflated balloon came

attached to the tip of a very soft, flexible, 1 mm diameter

delivery catheter This catheter, with the attached deflated

balloon, was coiled into a spherical “delivery chamber”

These delivery chambers were flattened at two opposite

ends, giving the chamber the appearance of a toy

“rotat-ing top” The delivery chamber had an inlet and outlet

port at the opposite flatter ends of the chamber

A larger, more maneuverable, guiding catheter, which

accommodates the 1 mm delivery catheter, is maneuvered

into the origin of the vessel to be occluded The outlet port

of the delivery chamber which contained the delivery

catheter and attached balloon was attached to the

prox-imal end of the guiding catheter with a Lure-lock

connec-tion By a rapid, forceful, hand flush into an inlet port

in the chamber, the delivery catheter with the deflated

balloon attached at the tip was hydraulically forced into

and through the guiding catheter, and from there into the

vessel The catheter with the deflated balloon actually

floated along with the flow of the fluid and blood in thevessel With this delivery technique, the tiny deflated bal-loon and catheter would traverse almost any bend, curve

or loop throughout the course of the vessel Usually thecatheter with the deflated balloon traveled completelythrough the vessel to, or even past, the end of the vessel.There was, however, no way to direct the course of the bal-loon/catheter purposefully if there was a branch or bifur-cation in the vessel Once delivered, the fine soft, deliverycatheter with the attached deflated balloon was with-drawn slowly until the deflated balloon reached a positionprecisely at the site within the vessel where the occlusionwas to occur

The balloon then was inflated with a predeterminedamount of contrast solution, which was diluted to beexactly isotonic With this inflation and with the balloonstill attached, the degree of occlusion and, of more import-ance, the tightness of the fixation of the balloon within thevessel were tested If the balloon migrated distally and/ordid not fix securely in a suitable position in the vessel, itwas deflated and withdrawn to a different more proximalarea or withdrawn completely If the fixation and occlu-sion seemed satisfactory, then the balloon was furtherinflated with several tenths of a ml more of the dilute contrast, which, in turn, fixed the balloon in place moresecurely and occluded the vessel The delivery catheterwas pulled away forcefully from and out of the “neck” ofthe “fixed” balloon, hopefully leaving the balloon in place.The B-D Mini-Balloon™ had a self-sealing valve in theneck of the balloon, which kept the balloon inflated afterthe delivery catheter was pulled out of the valve The iso-tonic contrast within the balloon would neither leach innor out of the balloon, once the balloon was inflated inposition Occasionally, even with this technique, the tinyballoons would work loose and migrate (embolize) more

distallyausually with no consequences as long as they

remained in the same vessel A 10 mm occlusion balloon

on a 2 mm diameter catheter, which was delivered withthe same system, was developed at about the time theentire system/concept was abandoned

The B-D Mini-Balloon™ had the capability of being able

to be delivered through very tortuous channels to wise inaccessible locations At the same time, the deliverytechnique was quite complex and initially was difficult for many operators to master, so most operators wouldchoose an alternative device (coil) unless the locationmandated the use of the occlusion balloon In addition,because of the balloon material, these occlusion balloonshad a definite and relatively short (one year) shelf lifebefore the materials of the balloon deteriorated andbecame unusable There have been other detachable

other-“mini-balloons” used sporadically and in isolated casesfor the occlusion of congenital heart lesions, but like the B-D Mini-Balloon™ they have been withdrawn from the

market, not because of complications, but because of the

very small overall market, the small demand for the more

complicated balloon occlusion techniques, and the

avail-ability of alternative devices which are simpler to use

A second larger Goldvalve™ Detachable Balloon

Embolization Device™ (DB™) was, and may still be,

available in Japan The Goldvalve™ balloons are

manu-factured of latex and are available in multiple sizes

varying from a small 6 mm cylindrical to a large 15 mm

diameter, 30 mm long balloon The Goldvalve™ balloons

attached to a 2–3-French delivery catheter are delivered

through a 9-French catheter The 9-French delivery catheter

is advanced over a long, stiff wire all of the way to the site

for the implant/occlusion and the balloon advanced

through the delivery catheter to the site Once in its proper

position the balloon is filled with a 2-hydroxy-ethyl

methacrylate (HEMA) solidifying solution Two different

liquid components of the HEMA compound are

intro-duced into the balloon, and because of a glucose oxidase

in one of the liquids, the liquids within the balloon

spon-taneously solidified over 40–60 minutes to form a solid

occlusive device13 Once the material solidifies, the

9-French outer catheter is pushed against the inflated

bal-loon, the attached small delivery catheter is pulled against

the 9-French outer catheter to free it from the balloon, and

both catheters are removed

Because of the necessity of delivering the large 9-French

delivery catheter to the lesion, these balloons had no

advantages in either delivery or occlusion capabilities

compared to standard occlusion coils and other occlusion

devices Because of the relative unknowns of the HEMA

materials in the human body over long periods of

time, these DB™ devices probably will not make it to the

United States

At this time there are no balloon occlusion devices

commercially available on the US market for permanent

occlusion of vessels and abnormal vascular channels

Amplatzer™ PDA & VSD devices for the occlusion

of abnormal vascular channels

The Amplatzer™ PDA and VSD occlusion devices

(AGA Corp., Golden Valley, MN) are better suited for the

occlusion of larger vascular channels than many of the

previously discussed devices, and are relatively recent

additions to the devices available for the occlusion of

abnormal and/or unwanted vascular channelsaeven in

the United States The Amplatzer PDA and VSD devices

are described in detail in Chapters 27 and 30, respectively

There also is the Amplatzer™ Plug (AGA Corp., Golden

Valley, MN), which is discussed earlier in this chapter and

is somewhat similar to the VSD and PDA devices

The three devices vary only a little from each other

in their design All three are roughly cylindrical, short,

closed tubes of a fine weave of Nitinol™ memory metal

The VSD devices have small 2 mm lips or rims at each end

of a “cylinder” and the cylinder or hub is not tapered Thecentral “tube” of the PDA device is tapered slightly andhas a rim, but only at the more distal end, which has theslightly larger diameter The “Plug” is a short cylindricaltube with no “rim” at either end and is manufactured of afiner Nitinol™ wire Which one of the devices is prefer-able for any particular vascular channel, depends uponthe particular anatomy of the lesion and the flow withinthe particular channel

The Amplatzer™ PDA and VSD devices as well as the Plug, all occlude by the same mechanism TheAmplatzer™ devices all self expand into their “resting”configurations in a vascular channel when extruded fromtheir delivery sheath When the channel is smaller indiameter than the particular device that is used, the deviceapplies pressure against the vessel walls outward and allaround the lumen to hold the device in place and fill thelumen of the channel Although none of these devices wasdesigned specifically for other types of vascular lesions orchannels, this mechanism of occlusion does make themideal for the occlusion of larger abnormal vascular chan-nels/leaks, which are broad, not necessarily very long,and which have high flow through them

As opposed to the delivery to the PDA or VSD, wherethe Amplatzer™ devices are positioned more or lessstraddling the lesion, their delivery to vascular channelsand/or fistulae usually involves implanting the deviceswithin the body of the channel or vessel or even in a channel which tapers further distally As a consequence,these devices, when expanded, potentially fix even more securely into the walls of the channel rather thanwhen “teetering” on a specific area of narrowing Sincethese devices are designed to fit into a specific sizeddefect, when used in vascular channels they often cannotopen immediately to their designed, fully expanded size and/or shape In those circumstances, they initiallyassume unusual configurations within the vessels, butgenerally occlude the abnormal channel very completely.With the “memory” of the Nitinol™ metal in these devices,the devices eventually will achieve their designed con-figuration even within a constrained vessel!

Both the Amplatzer™ VSD and ASD devices have been used in a variety of abnormal vascular channels inpatients outside of the United States and in a few extenu-ating, compassionate circumstances in the US Now theAmplatzer™ muscular VSD device, the Amplatzer™PDA devices and the Amplatzer™ Plugs are available foruse in the United States This does give physicians aroundthe world including in the US, a better choice of occlusiondevices in order to be able to use the most appropriatedevice for the occlusion of unusual and/or large vascularchannels

Atrial septal occluders for miscellaneous vascular

occlusions

Occasionally there are vascular channels that require

occlusion that are very large in diameter and relatively

short All of these lesions are large enough in diameter to

be hemodynamically significant, yet so short in length

that they will not accommodate any of the more common

and accepted vascular occlusion devices without

com-promising the adjacent vital blood channels that are

connected by the abnormal communication These very

large diameter channels include but are not limited to

the afferent vessels of pulmonary arteriovenous fistulae,

anomalous left superior vena cava draining into the left

atrium, and persistent azygos/hemiazygos veins

follow-ing “Glenn” type shunts In these cases, an Amplatzer™

ASD device (AGA Corp., Golden Valley, MN), an

Amplatzer™ PFO device (AGA Corp., Golden Valley,

MN), a Rashkind™ PDA device (USCI, Glens Falls, NY) a

CardioSEAL™ ASD device (NMT Medical Inc., Boston,

MA) and a Sideris Button™ device (Pediatric Cardiology

Custom Medical Devices, Athens, Greece) all can be of an

appropriate size and configuration to fit into, and all of

these devices have been used to occlude, these unusual

communications Some persistent aortic pulmonary

com-munications that were created surgically

(Waterston-Cooley or Potts shunts) are very short, they have high

flow, and they are present between very large,

high-pres-sure and very essential vessels As a consequence, they

require a very low-profile and sturdy device for occlusion

These devices and their delivery to their designed

lesions are described in detail in Chapters 28 (“ASD

Occlusion”) and Chapter 29 (“PFO, Fenestration, and

Baffle Occlusions”) In the large diameter, short and more

or less, “flat” lesions between two large, adjacent “tubes”

(great arteries), the positioning and fixation of the device

in the vessel are critical in order to prevent “movement” of

the device during or after implant The occlusion of the

short aortopulmonary vascular communications often

involves a more circuitous route to the lesion in order to

deliver a device, and the implant procedure usually

requires a very unusual and/or acute angle for the

deliv-ery of the device This makes the delivdeliv-ery to these unusual

locations more difficult and the release/implant of the

device more precarious Once the delivery sheath is in

position through the lesion while at the same time, not

kinking the sheath, the particular device extrusion and

implant is similar to the device delivery into an atrial

sep-tal defect/PFO

Caval filters or Cook “Spider”

Rarely there are very large diameter, abnormal, venous

channels and/or communications with very compliant

walls that require occlusion Because these veins distendvery easily and extensively, none of the previously de-scribed devices can be fixed securely in the vessels In thesevessels, the Spider™ device (Cook Inc., Bloomington, IN)

is used as a “scaffold” or framework for the implant ofother devices, particularly coils The Spider™ “springs”into a broad, spread-out configuration as it is pushed out

of its delivery catheter and attaches to the walls of the vessel by tiny, very sharp hooks in addition to the springtension of the arms of the device against the vessel wall.The Spider™ itself is not occlusive, but other devices such as coils are intertwined in the Spider™ to form anocclusive mass

Stents as a scafold for occlusion devices

Although designed for, and used primarily for, ing the patency of vessels, specially prepared intravascu-lar stents are used to assist in the occlusion of very largediameter channels/vessels14 The stents are used in amanner similar to the previously described Spider™ as a

maintain-“scaffold” for coils or other occlusions devices By fixingthe center of the stent at a restricted diameter by means

of a circumferential suture tied around the stent when

it is mounted on the balloon, a large diameter stent can

be expanded purposefully into an “hour-glass” shape.The suture is “woven” into the spaces between the struts

of the collapsed stent and tied around the stent in a cumference with a small fixed diameter before the stent

cir-is mounted on the delivery balloon The diameter of thesuture around the stent maintains the center of the stentexpanded only to the desired restricted diameter of thesuture When the stent is expanded on a large diameterballoon (15+ mm), the ends of the stent flare widely withthis “narrowing” at the center of the stent creating an

“hour-glass” configuration of the stent With the center

of the stent fixed at the narrow diameter by the suture, thetendency of the ends of the stent to extend and alignalmost perpendicular to the walls of the particular vessel/channel is exaggerated even more than usual The verysharp ends of each strut of particularly the J & J P _ _ 8 or

P _ _ 10 stents create sharp fixation points into the wall ofthe vessel This fixes the stent very securely in the vessel

A stent can be implanted using this technique and using

a P _ _ 10 stent, into vessels/channels up to 25 mm indiameter Once the stent is in place, then coils and/orocclusion devices can be intertwined into or packed intothe “afferent end” of the stent/vessel to achieve the occlu-sion of the vessel

Covered stents

Another mechanism with which intravascular stents areused to “occlude” vessels is by the use of “covered stents”

to purposefully occlude side or branching vessels and/or

tears in the wall off a main channel/vessel The covered

stent will cover all areas that are adjacent to the stent

and prevent flow into those vessels and other abnormal

openings/tears This is a very selected use and requires

the availability of a covered stent of the exact diameter of

the torn central or main channel or vessel The covered

stents are available commercially only for adult

peri-pheral vascular use and/or in investigational studies For

use in the isolated congenital patient, commercially

manufactured covered stents have been available only for

compassionate use in very extenuating circumstances

More frequently the stents are modified extensively from

the commercially available covered stents and/or hand

fashioned using ultrathin expanded

polytetrafluoroethy-lene (ePTFE, Gore-Tex™) membranes (W L Gore &

Associates, Flagstaff, AZ) over existing intravascular

stents of appropriate size for the involved vessel15 The

ultrathin ePTFE material has a considerable capability of

stretching acutely so that each diameter of ePTFE tubing

can be used with several sizes of stents However, none of

the means of acquiring and producing covered stents for

congenital heart patients in the United States are rapid

enough or satisfactory for emergency use For such use,

the stent should be available immediately from the

inven-tory of the catheterization laborainven-tory for emergency use

Except in emergencies, the covered stents probably only

should be used in fully-grown patients since there is no

suggestion or information about the ability to re-dilate the

covering materials on the stents even months, much less

years, after implant and/or in the presence of significant

growth of the vessels At the same time, in a life-threatening

emergency, the use of a covered stent that was rapidly

available and of a size to fit even a small vessel would be

preferable to massive blood loss and exsanguination from

an acute vascular tear

Occlusion procedures for specific lesions

Systemic to pulmonary collaterals

The systemic to pulmonary collateral vessels originate

from the aorta or from major branches off the aorta They

usually are relatively long, tortuous, tubular lesions of

varying diameters throughout their course and usually

have one or more discrete narrowing as they course

distally into the pulmonary parenchyma Because of this

anatomy the occluding device usually does not have to

“fix” tightly against the vessel wall and, in fact, almost is

expected to move distally a small distance after release

The anatomy of these lesions makes them the ideal lesions

for occlusion with “free-release” Gianturco™ coils The

most difficult part of the occlusion of these lesions

frequently is the initial access into, and then obtaining a

secure position well within, the vessel The cannulation of

some of the collaterals often requires a great deal of gination as well as a large inventory of specific cathetersand/or deflector wires Very often, a more secure accessinto a collateral off the descending aorta is achieved whenthe catheter is approaching from the ascending aorta tothe descending aorta This approach to the collateral ves-sel can be accomplished through a co-existing ventricularseptal defect, through a transseptal atrial septal puncture

ima-in conjunction with the use of a floatima-ing balloon catheter,

or by introduction of the catheter into a brachial artery.The collateral vessels that come directly off the aortausually are entered directly with the catheter for the deliv-ery of the coil Those vessels arising more remotely off

a brachiocephalic branch of the aorta, those with a verytortuous course, and/or those where the occlusion must

be accomplished in a more circuitous and distal location

in the vessel, often require the use of a separate “guidecatheter” to engage the orifice of the main collateral ves-sel Then, a small, very flexible, or even “floating” smallerdelivery catheter such as the Target Tracker™ catheter(Target Therapeutics, Fremont, CA) or the Cook Slip-Cath™ (Cook Inc., Bloomington, IN) is passed throughthe guide catheter to deliver the smaller coils and maneu-vered to a more secure distal location into the vessel that is

to be occluded

The exact type and size of coil that is used depend upon the access to the vessel as well as the length and thediameter of the vessel The “free-release” Gianturco™ andTornado™ coils usually are ideal for the smaller vesselswith sufficient length to accommodate several “com-pacted” coils placed in series In the same size but shorterlength vessels or in vessels where multiple coils alreadyhave been implanted and are filling the vessel back to

or near the orifice of the entrance from the main vessel,

a controlled release coil or device is desirable Using a controlled release coil during the extrusion of the finalcoil, if the entire coil does not fit completely and securelywithin the vessel to be occluded, the detachable coil can bewithdrawn, repositioned, “packed” into the vessel, oreven completely and simply withdrawn

The 0.052″ coils (Cook Inc., Bloomington, IN), larly with a bioptome-controlled release system, are verygood for the occlusion of larger diameter collaterals

particu-in smaller patients As long as the vessel that is to beoccluded does not arise too acutely off the aorta, thesecoils can be delivered through as small as a 4-Frenchsheath The controlled release allows repositioning, oreven complete withdrawal of the coil if either the posi-tion and/or occlusion are not satisfactory In the larger patient with a large, longer collateral, the Gianturco-Grifka Vascular Occlusion Device™ (GGVOD™) (CookInc., Bloomington, IN) and the Amplatzer™ plug (AGA

Medical Corp., Golden Valley, MN) are excellent choices

for these occlusions The use of the GGVOD™ and the

larger diameter plugs is dependent upon being able to use

an 8-French delivery sheath safely through an arterial

approach and to have the stiff sheath make the acute curve

from the aorta into the collateral

Surgical shunts

Occasionally, systemic to pulmonary shunts persist after

“corrective” surgery and require closure following a

pre-vious repair of the underlying congenital lesion Although

the “Blalock” and “modified Blalock” shunts are similar to

the collateral vessels in that they are tubular “vessels”, the

remaining characteristics of the shunts are very different

The shunt, which is created surgically, usually arises at

a very acute angle off the base of a brachiocephalic

vessel, which, in turn, makes access into the shunt from

the arterial approach very difficult The “modern” shunts

usually are made of synthetic materials that are very

non-compliant, they usually are uniform in diameter with no

tortuosity, no narrowing and no distal tapering, which

makes the fixation of an occlusion device within a Blalock

type shunt depend entirely upon the firm pressure of the

occluder against the walls of the shunt

The “free-release” Gianturco™ coil is still the most

commonly used device for transcatheter occlusion of

these lesions The larger, 0.038″ or preferably, 0.052″ coils

are used to ensure firm pressure against the walls of

the shunt Ancillary maneuvers frequently are employed

to assure the fixation of the coils in these shunts

Occasionally, there is a narrowing of the pulmonary artery

at the entrance/anastomosis site of the shunt into the

pul-monary artery In this circumstance and when the patient

is large enough to accommodate it, an intravascular stent

can be implanted in the pulmonary vessel in the area of

the stenosis, which will cross the exit orifice of the shunt.

This effectively “jails” the exit of the shunt and allows

coils to be packed into the shunt against the struts of the

stent, which are crossing the orifice of the shunt into the

pulmonary artery

When there is no stenosis of the shunt and/or no

steno-sis of the pulmonary artery (or the patient is not a

candid-ate for a pulmonary artery stent), an angioplasty type

balloon can be inflated in the pulmonary artery across the

distal end of the shunt for the purpose of temporarily

occluding flow through the shunt while the coils are being

implanted The balloon inflation is maintained as coils or

other devices are placed in the shunt The inflated balloon

occludes the distal end of the shunt and stops blood flow

through the shunt while coils are packed and intertwined

with each other within the shunt Unfortunately, this

tech-nique is not foolproof Unless the coils are packed very

tightly against the walls of the shunt and/or when the

coils are placed in a shunt with a very uniform diameterand/or no distal narrowing, the implanted coils still canembolize into the pulmonary artery when the balloon

is deflated

A sturdy, 0.038″ or even 0.052″ controlled release coil is

the preferred device for the residual surgical shunt whichhas no narrowing within its lumen The controlled releasecoil is implanted in the shunt and, before the coil is

released, tension on the attached delivery wire is relaxed.

In this way the coil is allowed to reposition itself withinthe shunt and usually, if it is not going to hold, the coil willpass through the shunt and to the pulmonary artery whilestill attached to the delivery, attach/release wire Whenthis occurs with the coil still attached, the coil can be with-drawn into the delivery catheter

Another alternative for occluding residual surgicalshunts is to deliver a much sturdier device into the shunt from the venous/pulmonary artery approach

A 0.052″ coil, a Gianturco-Grifka Bag™, or even a smallAmplatzer™ Vascular Plug can be delivered from thevenous approach The venous and pulmonary arteryapproach may have a straighter course to the shunt Thevenous introduction and straighter course allow the use of

a larger and/or stiffer delivery system

An end-hole catheter introduced into the venous system is manipulated into the pulmonary artery wherethe shunt enters, into the shunt from the pulmonary artery end of the shunt, through the shunt and into theaorta The catheter is replaced with an exchange length,long floppy tipped, Super Stiff™ wire The specific deliv-ery catheter/sheath for the device to be used is advancedover the wire through the venous system and into theshunt The dilator and wire are removed, the sheath iscleared of all air and/or clot and placed on a slow flush

Once the delivery catheter/sheath is in place through the shunt, the device is delivered to the site within the shunt

for implant/occlusion With the device maintained in theprecise position, the sheath is withdrawn, which, in turn,deploys the device in the shunt The degree of occlusion isverified with an angiogram, injecting into the arterial end

of the shunt through a separate catheter

Occasionally, entry with any catheter or wire into theshunt from the pulmonary artery end is very difficult

or even impossible In that circumstance, an exchangelength, floppy tipped, 0.035″ wire is advanced through anend-hole catheter, which is passed retrograde from theartery, through the shunt and into the pulmonary artery.The end of the wire is snared in the pulmonary artery with

a snare introduced from a vein and advanced through theright heart to the pulmonary artery Once snared, the wire

is withdrawn through the right ventricle, right atrium andout of the venous system, creating a through-and-throughwire, which passes through the shunt The delivery system for the occlusion device can then be advanced over

the wire from the venous approach and into the shunt for

venous delivery of the desired device

Short, large-diameter aorto-pulmonary

communications

Congenital aorto-pulmonary (A-P) windows and

resid-ual, direct aortic to pulmonary artery (Cooley–Waterston

or Potts) shunts can be occluded with devices in the

catheterization laboratory Coils are unsatisfactory for

these lesions because of the very short, broad, “window”

nature of these communications The most satisfactory

device for these “window-like” shunts is one of the

“double umbrellas” The double umbrella (Rashkind™

PDA, Clamshell™ and CardioSEAL™ ASD) devices, the

Sideris double umbrella modification of the Inverted

Button™, and a modified Amplatzer™ ASD device all

have been used to occlude this type of lesion16 –18

A-P window type communications are occluded from

either the arterial or venous approach The choice of

approach depends upon the vascular access for the

particu-lar approach and which approach gives the best “angle”

from the vessel into the lesion for the implant of the device

and, in turn, the best control over the particular delivery

system Because these communications essentially are

part of, and parallel to, the walls of the adjacent vessels

and the vessels have relatively small diameters compared

to a cardiac chamber, there usually is no good angle for the

delivery of any one of these devices! The devices initially

used for these lesions required an 11-French delivery

sheath, however, the Amplatzer™ PFO device potentially

can be delivered through a smaller sheath Because of the

sheath size, the venous approach is preferred when all of

the other factors in the choice in the approach are

other-wise equal

The congenital or residual surgical A-P communication

is identified and measured accurately by echocardiogram

and by selective angiograms with injection immediately

adjacent to the systemic arterial side of the lesion and/or

actually within the communication The X-ray tubes are

angled so that the communication is “cut on edge” at least

in one plane Once identified and measured, the defect is

crossed with an end-hole catheter Preferably, a

through-and-through, artery to vein wire “rail” is created with

an exchange length wire The catheter/wire can be

intro-duced from either the arterial or venous approach and

then captured with a snare catheter, which is introduced

from the opposite vascular access When there is any

question about the size and/or configuration of the defect,

the communication can be sized and “shaped” using a

NuMED™ sizing balloon (NuMED Inc., Hopkinton, NY)

advanced over the wire It is advantageous, if not

essent-ial, to have a second systemic arterial catheter to be used

for “positioning” angiograms during the procedure

For occlusion of the “window type” communicationwith any of these devices, a long delivery sheath first is

positioned through the defect with great care taken not

to kink the sheath because of the unfavorable anglesinvolved The through-and-through wire facilitates thedelivery considerably of the delivery sheath/dilatorthrough the communication Once the delivery sheath isthrough the defect and the wire and dilator are removed,the “umbrella” type devices are delivered exactly as for the occlusion of a PDA with the old Rashkind™ PDA device or the occlusion of a fenestration with aCardioSEAL™ device The distal umbrella/disk isopened on the “distal” side of the defect The sheath, thedelivery system and the device are withdrawn as a unit,pulling the open umbrella/disk tightly against the defect.The device may have to be pulled very firmly against thedefect, in which case, the device becomes markedly dis-torted, very similar to the delivery of an umbrella/diskduring the closure of a fenestration, particularly wherethere is a long communication Usually, the vessel wallssurrounding and adjacent to A-P communications arefairly rigid and the defects are small, so that a marked dis-tortion of the device from the strong “pull” on the deviceagainst the vessel wall is possible during the deliverywithout pulling the device through the defect When thecenter “hinge” or “hub” of the device is within and/orthrough the defect and the proximal legs/disk still arewithin the delivery sheath but completely through thedefect, the sheath alone is withdrawn, opening the proximalumbrella/disk on the more “proximal” side of the defect.After angiographic confirmation with an injection on thesystemic arterial side of the lesion, the device is released.The device often reorients itself markedly after releasefrom the torsion of an acutely angled delivery system

Coronary cameral fistulae

Closure of most coronary-cameral fistulae with a delivered device in the cardiac catheterization laboratorynow is the standard approach for these lesions The fistu-lae arise off either the right, left and/or any branch orcombination of branches off the coronary arterial system.They usually have a fairly long and often tortuous courseand empty into a right-sided cardiac chamber or the pul-monary artery They rarely empty into the left atrium oreven the left ventricle The choice of device and deliverytechnique depends upon the specific anatomy of the fistu-lae, the length of the fistulous tract and the distance awayfrom the true and essential coronary arteries Test occlu-sion of the fistula with a balloon is recommended in allcases to help to define the anatomy and distensibility ofthe fistulous tract, and to determine if there is any myocar-dial tissue dependent on the blood supply off the fistula.This is tested by the inflation of a small Swan™ balloon in

catheter-the fistula tract while observing catheter-the electrocardiogram

very closely for any ST-T wave changes

A retrograde, end-hole arterial catheter is introduced

into the “feeding” coronary artery and maneuvered

into the fistula as far as possible using torque-controlled

and/or Glide™ wires as necessary The fistula is defined

angiographically with selective coronary angiograms into

the feeding coronary artery or, preferably, into the orifice

of the fistula itself Multiple views are obtained to define

the course, branchings, narrowing and the site of

empty-ing of the fistula Often there is a narrowempty-ing and/or an

acute tortuosity in the fistula tract, which will help to fix

an occlusion device

An exchange guide wire is passed far into, and if

pos-sible, completely through the fistula The original catheter

is replaced over the wire with a small Swan™ end-hole,

balloon catheter or a very short, small balloon angioplasty

catheter, which is to be used for sizing and test occlusion

Several (multiple) low-pressure test inflations at different

locations along the fistulous tract are performed with the

balloon It is useful to have a second arterial catheter

already in position in the aortic root or even in the

select-ive coronary artery in order to perform small selectselect-ive

angiograms in the aortic root and/or in the coronary

artery proximal to the fistula and the site of test occlusion

In addition to the degree of occlusion of the fistula, this

injection demonstrates additional tracts off the fistula

and any significantly under-perfused areas of the

myo-cardium distal to the balloon occlusion

When the fistula is to be closed from the retrograde

arter-ial approach, the balloon catheter in the fistula is replaced

over the exchange wire with an end-hole only,

torque-controlled catheter The end-hole catheter is passed as far

distally in the fistula as is possible A “free-release” type

coil is satisfactory for coronary-cameral fistulae only when

(1) there is no myocardial tissue dependent on the flow

from the proximal fistula, (2) there is a long tract of the

fistula that is distal to the origin of the fistula off the true

coronary system and (3) the delivery catheter can be

manipulated reasonably easily and far distally into the

fistula With lesions that fulfill these criteria, the coils

usu-ally are delivered from the retrograde approach through a

catheter maneuvered selectively into the fistula from the

feeding coronary artery

When delivered retrograde, the occlusion device is

implanted as far distally in the fistulous tract and as far

away from the orifice of the fistula off the true coronary

artery as possible The distal location keeps the coil away

from the true coronary vessel and allows the deposit of

additional coils more proximally in the fistula Once the

delivery catheter is in position for the coil delivery, a

select-ive angiogram is recorded in the fistula through this

catheter to verify the precise location A coil for occlusion

of the fistula is chosen which is 1–2 mm larger in diameter

than the “balloon sized” diameter of the lumen of thefistula The coil must expand against the walls of thefistula in order to fix the coil in place in a non-stenotic loca-tion In a fistula with a significant narrowing and/or anarea of acute tortuosity more distally in the fistula, a coil is

chosen just larger than the diameter proximal to the

nar-rowing or to straddle a specific narnar-rowing similar to theretrograde implant of a coil into the “narrowing” of thePDA How the coil is delivered into the fistula dependsupon the precise anatomy, and is different in each case

At least 2–3 minutes after the first coil is implanted, thedegree of occlusion is tested with a repeat selective injec-tion into the fistula through the coil delivery catheter.Frequently there are residual leaks through one, or evenseveral, coils in a coronary-cameral fistula Leaving asmall residual leak through the fistulae, particularly withthe coils in place, increases the patient’s likelihood ofdeveloping endarteritis in the future, and when there is a

“jet” of residual leak through the coils, there also is a highchance of creating hemolysis

Several (or multiple) more coils are deposited in thefistula sequentially until the flow through the fistula isstopped The choice of the additional coils depends uponthe residual anatomy and the remaining proximal length

of the fistula Additional coils are intertwined with the initial (earlier) coils to prevent their movement backwardtoward the coronary artery When the most proximal coilswill be close to the entrance of the fistula from the cor-

onary artery, a controlled release coil and technique are used

for the final coil(s)

The prograde venous (retrograde into the fistula)approach for the transcatheter closure of some coronary-cameral fistulae has several distinct advantages over theretrograde arterial approach Fewer catheter manipula-tions are required in and/or through the somewhat pre-carious true coronary arteries and the central arterialsystem in general Larger and stiffer (less bendable) deliv-ery systems can be introduced from the venous approachthan from the retrograde arterial approach This extendsthe options to a larger variety of occlusion devices that can be used to occlude coronary-cameral fistulae Fromthe venous approach, the delivery sheaths for the large 0.052″ coils and Duct-Occlud™ device as well as the largersheaths for the Gianturco-Grifka™ Bag, the Amplatzer™Vascular Plug or the Amplatzer™ muscular VSD occlud-ers can be positioned in the fistula without kinking the sheaths and without compromising the systemic arter-ial entrance site and/or the proximal coronary arteriesthemselves

In order to establish the venous access to the fistula, italmost always is necessary first to create an arterial tovenous, through-and-through wire The catheter/wire isintroduced from the arterial approach, into the coronaryartery, through the fistula and into the venous circulation,

where the wire is snared with a catheter introduced from a

vein The introduction site for the venous catheter should

be from the peripheral vein (femoral or jugular) that will

give the best angle of approach into the distal end of the

fistula An end-hole retrograde catheter is introduced into

the feeding coronary artery and from there, manipulated

into the origin of the fistula A torque-controlled, floppy

tipped or curved tip Glide™ wire is maneuvered through

the fistula and into the “exit chamber” of the fistula within

the heart The distal end of the wire is snared at the

“venous” end, where it exits the fistula and is exteriorized

through the venous entry site

The retrograde catheter should remain over the arterial

end of the wire while the delivery catheter or sheath/

dilator is advanced from the venous end of the

through-and-through wire and into the fistula The arterial catheter

over the wire protects the tissues within the coronary

artery from damage by the bare wire and allows a

contin-uous flush around the wire to prevent clotting in the

cor-onary system The tip of the delivery catheter/sheath is

positioned according to both the anatomy of the fistula

and what type of device is being used In general, the tip

of the delivery catheter/sheath is positioned just

prox-imal (in the direction of blood flow in the fistula) to the

location where the device(s) is(are) to be implanted Once

the delivery catheter/sheath is securely in position, the

through-and-through wire is withdrawn through the

arter-ial catheter and the catheter is cleared carefully of any air

and/or clot The retrograde arterial catheter remains

posi-tioned in the coronary artery in the end of the fistula or at

least in the coronary artery just proximal to the fistula The

retrograde catheter is used to perform sequential selective

angiograms during the positioning and delivery of the

device(s) and/or to add additional occlusion devices after

the implant of the original device

An occlusion device/coil is chosen which is appropriate

for the particular anatomy of the coronary fistula When

delivered from the venous approach the choice of type

and size of device is much greater, allowing the delivery

of very retrievable devices such as the Amplatzer™

vas-cular plug and the GGVOD™ devices When delivered

from the venous approach, the device is implanted in the

fistula straddling an area of narrowing or proximal (in the

direction of blood flow) to any areas of narrowing and/or

sharp bends in the fistulous tract, but at the same time

distal to any true coronary artery branches off the fistula

The one disadvantage to the venous approach is that

sub-sequent devices that are delivered through the same

delivery system, must be implanted “downstream” or

dis-tal in the flow from the original device and often into a

larger diameter area of the fistula than the site of the

original device The “downstream” devices, which are

predominately coils, may not fix in the vessel as securely,

and complete occlusion is less likely Any subsequent

device and, in particular, coils delivered in a stream” location must be significantly larger than thediameter of the fistula in that location and must be “inter-twined” and attached to the original device

“down-It always is preferable to implant additional coils/devices proximal to the original device If additional coils

are to be delivered from the venous side and are to be

placed proximal to the first device, a delivery catheter

must be advanced through and/or past the original device and

into the more proximal fistula without dislodging theoriginal device/coil This is accomplished using a TerumoGlide™ (Boston Scientific, Natick, MA) wire and a tiny,end-hole-only catheter similar to crossing a freshlyimplanted coil in a PDA As the coil is being extrudedfrom the catheter, which has passed through/past theoriginal device/coil, the catheter is withdrawn into, oradjacent to, the original device/coil in order to catch orentwine the additional coil(s) on the original device/coil

A preferable technique for the implant of additionalcoils proximal to the original device, which was deliveredfrom the venous approach, and particularly when the firstdevice is implanted distally in the fistula away from theentrance of the fistula off the coronary artery, is to deliver

additional coils “upstream” through the retrograde catheter

that is already positioned in the proximal part of the

fistula Using a controlled-release system additional coils

are “stacked on top of ” the original device to plete the occlusion Often, with a very large coronary-cameral fistula, this combination of venous and arterialapproaches is the planned procedure

com-In certain circumstances during the catheter occlusion

of coronary-cameral fistulae it is imperative to use a controlled-release device for the delivery of the occlusiondevice The presence of a particularly short fistulous tractand particularly one with a wide orifice off the true coronary system necessitates a controlled-release device

A fistula of any length into which multiple coils areimplanted and when the next, and most proximal, coilmust be implanted near the true coronary artery, is theother definite indication for a controlled-release system

In these circumstances, the coil and/or any other type ofdevice must be retrievable immediately in the event thatthe device or part of the device extends back into thelumen of the true coronary artery during the delivery ofthe device The tip of a coil with its attached filamentstrands dangling into the lumen of a true coronary artery

is very likely to shed thrombotic emboli into the high flow

in the coronary artery and can result in infarction ofmyocardium distal to that area

Pulmonary arteriovenous fistulae

Pulmonary arteriovenous (AV) fistulae can be congenital

or “acquired”, they can be isolated or multiple, they can

occur in thousands and they can be large or small These

lesions can be occluded with catheter-delivered devices.

How they are occluded and with what particular device

are determined by the exact anatomy of the lesion(s) Each

pulmonary AV “fistula” may have several afferent and

efferent vessels The central communicating vessels usually

are small, tortuous and restrictive, but can be relatively

large and even form an almost straight communication

from the pulmonary artery to the pulmonary vein

The rare, isolated, congenital pulmonary AV fistula

usually is large, with multiple and usually tortuous

affer-ent pulmonary arterial and efferaffer-ent pulmonary venous

channels The size(s) of the major central AV

communica-tion(s) through the fistula is(are) determined by detailed

angiography An occluding device significantly larger

than the major central communicating AV channel must

be used initially to prevent the occlusion device from

passing through the fistula and into the pulmonary

vein/systemic circulation The goal is to occlude the major

afferent channel(s) while, at the same time, occluding no,

or as few as possible, normal pulmonary arterial vessels,

which supply adjacent normal lung tissue In the presence

of multiple channels, test occlusions of the various

affer-ent channels using a balloon angiographic catheter or a

Swan™ balloon catheter is performed to determine the

largest and/or most significant feeding vessel(s) In most

cases, the actual communication between the artery and

vein is relatively small and one or many standard

“free-release” Gianturco™ coils can be used for the occlusion

When the communication is large, the large 0.052″ coils,

the Gianturco-Grifka™ bag, the Amplatzer™ PDA, VSD

devices or vascular Plug or even an “umbrella” occluding

device all can be used to at least begin the occlusion in

these lesions Once a larger device is in place, standard

Gianturco™ coils can be packed proximally to the larger

device in the afferent channel to complete the occlusion of

that particular vessel/segment

Biplane fluoroscopic and angiographic systems are

absolutely essential for the treatment of these lesions The

branching pulmonary arteries are very “three-dimensional”

and arise at multiple angles to, and at different planes

from, each other The branches of the arteries to be

occluded must be cannulated selectively, and then

sub-selectively A single-plane system does not provide the

depth relationships for this precise cannulation Probing

with a single-plane system results in exponential increases

in the radiation use/exposure and of the amount of

contrast used for localizing and cannulating any specific

lesion

It is desirable to have a second venous angiographic

catheter, which is positioned selectively in the pulmonary

artery and immediately adjacent to the delivery catheter

for the occlusion device Once the precise afferent vessel

has been identified, it is cannulated selectively with an

end-hole catheter The catheter is replaced with anexchange length, short tipped, 0.035″ Super Stiff™ wire.The delivery dilator/sheath set with a sheath appropriatefor the device to be used is advanced over the wire into

the major afferent vessel and as close to the fistulous munication as possible Multiple normal pulmonary artery

com-branches, which supply vital normal lung tissue, usuallyarise off the major afferent pulmonary artery proximal tothe fistula The more proximally the occlusion is per-formed in the afferent pulmonary artery that feeds thefistula, the greater the number of vital vessels that supplythe normal adjacent functional lung parenchyma will

be occluded

The occlusion device is introduced into the preciselypositioned sheath and extruded as accurately as is pos-sible into the exact location and released Before release ofthe device, an angiogram is performed through the sec-ond venous catheter to determine the effectiveness of thefirst occlusion device and to determine the location(s) ofthe next most important afferent vessel(s) With residualflow through the fistula from this artery, either a secondlarge device is implanted through the same deliverycatheter or a coil delivery catheter is introduced throughthe delivery sheath and the occlusion is completed withsupplemental coils packed proximal to the original device

in the afferent vessel of the fistula

Once the first feeding artery is occluded, the deliverysheath/catheter is repositioned into the next most import-ant feeding vessel and the procedure repeated Once theinitial, larger vessels are occluded, the remaining vesselsmay be occluded more satisfactorily with a differentand/or smaller device In that case the original deliverysheath/catheter is exchanged for the delivery sheath/catheter for the different device or the new deliverycatheter is advanced to the lesion through the originaldelivery sheath/catheter The procedure is repeated untilthe individual fistula is occluded Each time an occlusiondevice is implanted, it is important to perform the implant

as close to the lesion as possible

In the presence of multiple pulmonary arteriovenousfistulae, patients often are very cyanotic and very symp-tomatic At the same time, there usually are too many sep-arate pulmonary AV fistulae to consider occluding all ofthem In the very symptomatic patient, the largest andmost important AV lesions are identified with detailedangiography and an attempt is made to occlude as many

of the major fistulae as possible without further mising the remaining normal lung tissue This, of course,does not cure the patient, but often, enough of the lesionscan be occluded to provide a significant, albeit temporary,symptomatic palliation for the patient The technique isthe same for multiple fistulae as for an isolated largefistula The exact device and the technique used are deter-mined from the selective angiographic anatomy of the

compro-individual lesions With each successful occlusion of one

or more significant lesions, the patient’s systemic arterial

saturation rises slightly If this does not occur after the

occlusion of three or four more significant fistulae, the

patient probably will not receive benefit from further

occlusions

The multiple diffuse (thousands) of tiny pulmonary

arteriovenous fistulae seen frequently after isolated

sup-erior caval to pulmonary artery anastomoses are not

directly amenable to occlusion therapy The exception is

when there are a few separate large fistulae associated

with the other multiple diffuse tiny fistulae These

larger systemic to pulmonary fistulae can, and should, be

occluded Even with the multiple small fistulae, occlusion

of associated larger pulmonary fistulae provides some

symptomatic relief for the patient

The approach to the multiple tiny pulmonary

arterio-venous fistulae is to anticipate their development, and in

turn, prevent their occurrence by proceeding with the

completion of the connection of the inferior vena cava so

that hepatic blood flow reaches both lungs Even after these

multiple tiny pulmonary A-V fistulae have appeared, the

patient’s systemic saturations seem to improve with the

completion of the inferior caval connection to both lungs,

even though the multiple tiny fistulae do not, objectively,

seem to regress

Systemic artery to pulmonary artery and

veno-venous collaterals post “Fontan corrections”

The complete caval-pulmonary repair of single ventricle

lesions has created a whole new spectrum of systemic

artery to pulmonary artery and systemic vein to

pul-monary vein collaterals/fistulae These are usually small

but can be multiple and/or very large, especially some of

the hepatic vein to pulmonary venous fistulae Regardless

of the size of an individual fistula, each fistula provides

“competition” to the flow to the already precarious,

“for-ward” pulmonary blood flow, produces extra volume on

the “single ventricle” and/or creates a “right to left” shunt

with systemic desaturation of the patient These abnormal

vascular communications usually are not obvious from

clinical or noninvasive studies They are detected only

during a cardiac catheterization and then, only when they

are looked for specifically with selective angiography

These abnormal fistulae arise from any systemic artery or

vein within the thorax, neck or even from the abdomen

The arteries communicate with the pulmonary arteries,

the pulmonary veins and/or even the systemic veins The

systemic venous channels communicate with the

pul-monary veins and/or with other systemic veins draining

to an entirely different systemic venous bed The arterial

communications usually are small tortuous channels

but on aortography or systemic venous angiography, the

pulmonary arteries and/or veins light up with contrastlong before the normal “recirculation” from the pul-monary arteries has time to reach the pulmonary veins.When a catheterization is performed for any reason onany of these single ventricle, “Fontan” type patients, theselesions should be looked for specifically while the patient

is in the catheterization laboratory and when found,occluded

The major challenge in occluding these lesions is ing and then cannulating their origins off the systemicarterial or systemic venous systems Once located andcannulated, they usually are occluded quite simply withfree-release coils or other available occlusion devices Thesmaller of these lesions are ideal for the small Tornado™(Cook Inc., Bloomington, IN) or Target™ (Target Thera-peutics, Fremont, CA) coils and delivery systems Thehepatic vein to pulmonary venous fistulae usually aremuch larger and require one or more of the larger occlud-ing systems Fortunately, access to these hepatic to pul-monary venous channels is usually directly from thecentral hepatic veins, allowing introduction of the larger,stiffer delivery systems for the large coils, Gianturco-Grifka™ bags (Cook Inc., Bloomington, IN) or one of theAmplatzer™ occlusion devices (AGA Medical Corp.,Golden Valley, MN)

locat-In 2004, most of the abnormal vascular communicationswere treated in the catheterization laboratory with theavailable devices even in the United States With the addition of the devices that are on the horizon in protocolstudies in the US and/or routinely available in the rest ofthe world, even more of these lesions will be amenable tocatheter therapy The catheter management of all abnor-mal vascular communications will be safer and morewidely distributed with the introduction of a wider choice

of devices that are more appropriate for specific lesions

Occlusion of aortic root/coronary sinus to left ventricular tunnels

Congenital tunnels from the aorta to the left ventricle arerare fistulous tracts, which usually present in infancy Thecombination of the clinical findings of a nondescript systolic murmur, which is followed immediately by a longdiastolic, decrescendo murmur over the base of the heartand in association with a widened pulse pressure areindistinguishable from the findings of aortic valve regur-gitation The diagnosis of an aortic to left ventricular tun-nel is made by echocardiogram and confirmed with anaortic root angiocardiogram A biplane angiogram is nec-essary to determine the precise origin, the course and theexit of the tunnel and to obtain accurate measurements

of the diameter of various areas of the tunnel Generally,these lesions have a fairly straight origin off the aorticroot, are some distance from the aortic valve and coronary

arteries and, as a consequence, are very amenable to

occlusion with a catheter-delivered device

Once the tunnel has been identified angiographically,

it is cannulated selectively with a floppy tipped spring

guide wire and then a catheter over the wire Selective

angiograms are recorded with injections directly into the

tunnel in order to define the anatomy of the tunnel more

precisely The X-ray tubes are angled in order to align the

tunnel on edge and in its longest dimension If there is any

question about the anatomy and/or diameters of the

tun-nel, the guide wire is passed through the tunnel into the

left ventricle and a Swan™ floating balloon catheter is

advanced over the wire into the tunnel The balloon is

inflated very gently and at a low pressure within the

tun-nel and, if the tuntun-nel has any significant length, in several

different locations along the course of the tunnel The

bal-loon inflation in the tunnel provides information about

the distensibility of the fistulous tract and how the patient

will tolerate an occlusive device in the tunnel The

com-bined angiographic and “sizing” information determines

the most appropriate device for occlusion of the tunnel A

static sizing balloon would give the same or better

infor-mation, but would require the use of a 9-French sheath in

the artery!

The tunnels usually are large relative to the patient’s

size and all of the tunnels from aorta to left ventricle have

a high-velocity flow through them This requires the use

of a sturdy occlusion device, which will fix securely in

the high-pressure, high-flow tract Because of the usual

small size of the patient, usually either a

controlled-release, large coil and/or a small Amplatzer™ PDA or

Vascular Plug is/are used A coil with a

bioptome-controlled release (in the US) requires a 5-French long

sheath for true controllability and the smallest sizes of

both the Amplatzer™ PDA and Vascular Plug devices

(4 mm) pass through 5-French sheaths The Detachable™

coils, available in the rest of the world, utilize a 4-French

delivery catheter, but these coils often are not robust

enough for these lesions In the rare larger patients, larger

devices including the GGVOD™ can be considered for the

occlusion of these tunnels

A wire is maneuvered from the retrograde catheter,

through the tunnel and into the left ventricle The

retro-grade catheter is replaced over this wire with the delivery

sheath/dilator or catheter that is to be used for the device

chosen for the particular tunnel The delivery sheath is

advanced through the tunnel, to the left ventricle or, at

least, to the distal end of the tunnel Once the delivery

sheath is positioned securely in the tunnel, the wire and

the dilator are removed from the delivery sheath The

sheath is cleared passively and meticulously of all air

and/or clot as with all other delivery systems particularly

in the systemic circulation The appropriate device is

loaded into, and advanced to the end of the sheath within

the tunnel The sheath and occlusion device are drawn together until the device, which is still within thesheath, is positioned in the desired location for implant

with-in the tunnel If the device bewith-ing used is a coil, the coil

is extruded from the sheath into the tunnel until theattach/release mechanism of the coil is just within the tip of the sheath With the other devices, the deliverycatheter/cable and device are fixed in position and thesheath withdrawn off the device until the device is fullydeployed in the tunnel

The fixation of the occlusion device is tested by fairlyvigorous to-and-fro motion of the delivery system, and

a small, hand injected angiogram is performed with aninjection of contrast through the delivery sheath Thedelivery sheath still is over the delivery catheter/cablewith the tip of the sheath just proximal to the devicewithin the tunnel and even a slow, small injection demon-strates the position of the occlusion device and the degree

of occlusion of the tunnelaparticularly if the tunnel is

occluded satisfactorily When satisfied with the fixationand the degree of occlusion, the release mechanism of theparticular device is activated By using a device with acontrolled-release system, if the fixation of the device andocclusion are not satisfactory, the device readily can bewithdrawn into the delivery system and the procedurerestarted with a more appropriate size or type of device

mul-of the “sewing ring” mul-of the prosthetic valve, which usuallyplaces the fistulous communication away from the func-tioning “leaflets” of the prosthetic valve

The aortic prosthetic perivalvular leak cally is similar to a coronary sinus to left ventricular fistula(previously in this chapter), although perivalvular leaksfrequently are multiple and usually occur in significantlylarger and older patients The diagnosis of, and the number of aortic perivalvular leak(s) are suggested bytransthoracic echocardiography and/or from biplane,aortic root aortograms The treatment of perivalvularleaks is monitored in the catheterization laboratory with

hemodynami-both selective biplane angiograms, which are performed

actually within the perivalvular leak, and by esophageal echo (TEE) The aortic perivalvular leak isaccessed with a retrograde catheter from the aorta above the prosthetic valve The orifice of the perivalvu-lar leak is usually located directly off the ascending aorta adjacent to the “ring” of the prosthetic aortic valve

trans-The location(s) of the aortic perivalvular leak(s) is(are)

identified with TEE

Once an aortic perivalvular leak is suspected and then

identified, the leak is cannulated selectively with a

retro-grade, end-and-side hole catheter The exact location of the

aortic opening of the perivalvular leak in relation to the

aortic valve prosthesis, its shape, its maximum and minimal

diameters, its length and its exit site into the left ventricle

all are defined very accurately with selective biplane

angiograms with injection directly into the fistulous tract,

with the X-ray tubes angled to elongate the tract

max-imally The perivalvular leaks often are short and have

an elliptical shape adjacent to the rigid, prosthetic valve

“sewing ring” Once the precise anatomy and size are

defined angiographically and correlated with the TEE

findings, an occlusion device is chosen which will occlude

the perivalvular leak, but, at the same time will not extend

out of the orifice of the fistula and interfere with the

func-tion of the adjacent prosthetic valve

An exchange wire is placed through the retrograde

catheter, into and through the tract of the perivalvular

leak and into the left ventricle The original catheter is

replaced over the exchange wire with the specific delivery

system for the device to be used and the delivery system is

advanced deep into the tract of the perivalvular leak The

tract of the perivalvular leak and the delivery system

are interrogated continuously with the TEE It is useful, if

not absolutely necessary, to introduce a second,

retro-grade catheter into the aortic root in order to perform

small, repeated, selective angiograms during the actual

delivery of the device, which is being performed through

the other retrograde catheter/delivery system

When the fistulous tract has any length and has any

constriction within its course, a controlled-release, large

coil often is the ideal device for these lesions, but only

when the coil can be implanted completely within the tract

so that none of the attached coil fibers extend out of the

tract and into the area of the moving leaflets of the

prosthesis

For the short, broader perivalvular leaks, the smaller

diameter Amplatzer™ PDA, the Amplatzer™ muscular

VSD occluders or the newer Amplatzer™ Vascular Plug

are used The choice of which of these devices to use

depends upon the size of the proximal orifice and the

presence of any narrowing along the course of the leak In

spite of the usual elliptical shape of the orifice of these

perivalvular leaks, the Amplatzer™ devices, when

ini-tially implanted, often conform to the shape of the tract

The Amplatzer™ devices are placed either at the aortic

orifice of the leak or within the tract of the leak utilizing

the “ridge” created by the valve ring as the “central

nar-rowing” to straddle with the “waist” of the device The

exact location where the occlusion device is implanted

within a perivalvular leak depends upon the length,

diameter and configuration of the perivalvular leak, theproximity to the prosthetic leaflets and, often, a trial anderror deployment of the device The “lips” or “rims” of theAmplatzer™ PDA and muscular VSD devices usually areshort enough that they do not interfere with the leaflets ofthe prosthesis, even when the “rim” of the device is out-side of the orifice of the fistulous tract

Once a coil or one of the Amplatzer™ occluders hasbeen delivered into the tract, but before its release, arepeat biplane angiogram in, or very close to, the aorticorifice of the leak is performed to visualize the tract, tovisualize the degree of occlusion, and to check the position

of the coil/device in relation to the movement of the thetic valve leaflets The area is interrogated with theTEE/Doppler™ to identify the distance of the particularcoil/device away from the leaflets of the prosthetic valveand to check for residual leak through the tract When theocclusion is satisfactory and there is no interference withthe function of the prosthetic valve as seen on the TEE, thecoil/device is released The angiogram and the TEE arerepeated after the release of the coil/device to confirm the degree of occlusion, changes in the position of thecoil/device, the freedom from interference with the pros-thetic valve apparatus, and/or the presence of any addi-tional perivalvular leaks

pros-Any residual and/or additional perivalvular leak isaddressed during the same catheterization procedure.Residual leaks through devices can result in hemolysisand, if not closed during the initial implant procedure,may have to be closed later as an emergency In addition,when performing additional occlusions during the sameprocedure, all of the anatomy is fresh in the operator’smind and the wires and catheters already are in place forthe implant of an additional occluder, even though it may

be a different type from the original occluder

Small PDA “double umbrella” occlusion devices, which could be implanted completely within the tract ofperivalvular leaks, were used in the past for the occlusion

of perivalvular aortic leaks, however, all of the doubleumbrella devices require a much larger and stiffer deliv-ery system As a consequence, the double umbrelladevices on a rigid frame no longer are used for the occlu-sion of aortic perivalvular leaks

Perivalvular leaks around a prosthetic mitral valve resent significantly greater challenges for transcatheterocclusion Mitral perivalvular leaks frequently are multi-ple and they are more difficult to image precisely by eitherecho or angiography Both prograde access through atransseptal atrial septal puncture to the left atrium andretrograde access from the left ventricle are used for thecatheter occlusions of mitral perivalvular leaks Devices

rep-in perivalvular mitral valve leaks are delivered andimplanted from the left atrial (outlet) end of the leak,which is against the direction of the flow of the blood

through the mitral perivalvular leak As with the

occlu-sion of the aortic perivalvular leaks, the delivery and the

implant of occlusion devices for mitral perivalvular leaks

are visualized and controlled by both selective biplane

angiography and TEE/Doppler

Mitral perivalvular leaks initially are identified by

transthoracic echocardiography and biplane, left

ventric-ular angiography Once the perivalvventric-ular mitral leak is

defined angiographically and a decision is made to

perform a transcatheter occlusion, a pre-curved,

end-and-side hole, retrograde catheter is positioned in the left

ventricle, a prograde, long, transseptal sheath is

posi-tioned in the left atrium (usually by transseptal atrial

puncture), and a TEE probe is introduced to visualize the

mitral annulus The specific leak(s) is(are) identified, then

they are defined more definitively by TEE/Doppler™

interrogation and more selective biplane angiograms

with the injections preferably performed directly into the

perivalvular tract from either a prograde or retrograde

approach Selective views of the angiograms in the left

ventricle and/or selective views in the tract of the leak

are displayed as “road maps” of the overall anatomy of

the area

A Judkins™ left coronary catheter or another “J”

shaped catheter is introduced retrograde into the left

vent-ricle The 180° curve on the pre-curved, retrograde, left

ventricular catheter helps to direct the tip of the catheter

toward the mitral annulus or, more specifically, toward

the left ventricular (entrance) end or into the left

ventricu-lar “entrance” of the perivalvuventricu-lar leak(s) in order to obtain

more “selective” injections

Using the angiographic “road maps” and TEE, from the

venous approach, a separate 6- or 7-French, end-and-side

hole torque-controlled catheter with a fixed slight angle

at the tip is maneuvered, usually along with a

torque-controlled guide wire, into the left atrial (exit) end of the

perivalvular leak In order to prevent extensive,

unneces-sary and “blind” maneuvering, biplane fluoroscopy is

absolutely essential for this maneuvering in order to

direct the catheter purposefully both from side to side and

from front to back The catheter is advanced well into, or

through, the fistulous tract The X-ray tubes are angled as

perpendicular to the long axis of the catheter (and tract)

as possible and to place the catheter, which is positioned

in the tract, at one “edge” of the valve ring as seen in at

least one view A high-pressure, Tuohy™ “Y” adaptor is

attached over the wire to the hub of the catheter and a

biplane angiogram recorded while injecting through this

catheter, over the wire and directly into the tract of the

perivalvular leak The biplane angiograms define the

diameter, shape and length of the tract and demonstrate

the distance of the proximal and distal openings of the

tract from the prosthetic valve leaflets The diameters,

lengths and distances are measured very accurately on

both the angiogram and on the TEE images The properdevice to occlude the perivalvular leak is chosen on thebasis of these angiographic measurements and the cor-responding TEE images The original torque wire, whichstill is through the catheter, is replaced with a long, pre-formed, floppy tipped Super Stiff™ (Boston Scientific,Natick, MA) wire, which is looped in the ventricle afterpassing through the perivalvular tract

The choice of the device used to occlude a perivalvularmitral valve leak is similar to the choice of device used

to occlude perivalvular aortic valve leaks The

avail-able devices include large, controlled-release coils, the

Amplatzer™ PDA, muscular VSD and Vascular Plugdevices, and the Gianturco-Grifka Vascular OcclusionDevices (GGVOD™) In addition to these devices, thelarger and stiffer delivery systems that are necessary toaccommodate the “double umbrella” devices, can be tol-erated from the venous/transseptal approach, allowingthe use of virtually any occlusion device that is suited forthe particular anatomy Once the exact device is chosen, theappropriate delivery system/catheter is delivered to theperivalvular leak from the prograde/transseptal/leftatrial approach The specific delivery system/catheter forthe occlusion device can be delivered to the perivalvular

leak over the Super Stiff™ wire through the original

trans-septal sheath, or the new long delivery sheath/dilator isintroduced over the exchange wire after removing theoriginal transseptal sheath If the original transseptalsheath does not accommodate the delivery system/catheter for the device that is chosen, the originaltransseptal sheath is replaced with a larger diameter longtransseptal sheath which is one or two French sizes largerthan the particular delivery system that is to be used forthe device delivery The larger diameter transseptalsheath is pre-curved before it is introduced to conform tothe course from the inferior vena cava, through the rightatrium and atrial septum and to a position just over themitral annulus The large long sheath positioned in the leftatrium provides extra support for the delivery catheter,and when the larger sheath is advanced to the left atrialend of the perivalvular leak over the delivery catheter, the large sheath allows “postioning” angiograms to beobtained just at the distal end of the leak during the deliv-ery of the device without the necessity of another catheter

in the left atrium When the device is deployed and flowthrough the tract is decreased/stopped, the contrast willreflux “backward” into the tract slightly

The device is delivered into the tract of the perivalvularmitral leak and positioned in the tract similarly to thepositioning in perivalvular aortic leaks and according tothe type of device being used Once opened in the tract,the positioning, degree of occlusion and distance awayfrom the leaflets of the prosthetic valve are examined

by repeat biplane angiography, injecting through the

retrograde catheter or through the long sheath over the

delivery catheter, and by TEE/Doppler™ interrogation

When satisfied with the occlusion and the position, the

device is released A large volume left ventricular

angio-cardiogram is performed through the retrograde catheter

to visualize the final occlusion of the perivalvular leak and

the function of the prosthetic mitral valve, and to exclude

any additional perivalvular leaks If there is persistent

flow through the original perivalvular communication

and/or if there are any additional perivalvular leaks

detected, these are addressed during this same

catheter-ization for the same reasons as for residual perivalvular

aortic leaks

Complications of vascular

occlusions

As with all other therapeutic catheterization procedures,

there are potential complications related to the

catheter-ization itself and those more specifically related to the

occlusion procedure Since a majority of the abnormal

ves-sels and/or vascular channels requiring occlusion arise

off a systemic artery, most transcatheter occlusion

proced-ures are performed entirely or, at least, partially from an

arterial approach This, alone, increases the likelihood

of local arterial injury and of systemic embolic

complica-tions These complications and the steps taken to prevent

them are discussed in detail in Chapter 4 (“Vascular

Access”) and Chapter 9 (“Retrograde Technique”)

Embolization of the occlusion device and/or

inadvert-ent occlusion of the blood supply to a vital structure or

organ are the major complications unique to intravascular

occlusion of systemic vessels, fistulae and perivalvular

leaks Occlusion of unwanted vessels/structures is

pre-vented by test occlusion with a balloon catheter before

depositing an occlusion device and by the use of

controlled-release/retrievable devices whenever there is

any question about compromise of a vital adjacent vessel/

structure This is particularly important in the occlusion of

coronary-cameral fistulae, where the extension of even a

small part of the occlusion device into a vital coronary

artery can obstruct and/or thrombose an essential

cor-onary blood supply

Remote embolization of a vascular occlusion device

is always possible, but almost always is preventable

Certainly, the use of the controllable and retrievable

devices adds to the safety and secure positioning of

occlu-sion devices for peripheral vascular occluocclu-sions When

“controllable” release devices are not available or cannot

be used, extra care is taken in the sizing of the device for

the lesion and in choosing the safest location for the

implant of the occlusion device When a vascular

occlu-sion device does embolize to a different unwanted and/or

dangerous area or structure, it becomes a foreign bodythat requires removal It is imperative that the catheteriza-tion laboratory performing vascular occlusion procedureshas the equipment and expertise for the removal of alltypes of foreign bodies

The migration of a coil or other small occlusion devicecompletely through a large pulmonary arteriovenousfistulous communication into the pulmonary vein and, inturn, embolization into the systemic arterial circulationalways is a possibility and potentially, results in very seri-ous consequences This type of embolization is prevented

by the use of a much larger device and/or an “irregular”device, like an umbrella, which becomes trapped securely

in the afferent vessel of the fistula

Whenever there is incomplete occlusion of a vessel orother lesion with a resultant high-velocity residual “jet”leak through or adjacent to an implanted device, any ofthe vascular occlusion devices can produce hemolysis.Again, treatment is prevention by occluding the lesioncompletely during the initial catheterization for the occlu-sion When a high-velocity lesion does remain and doesresult in hemolysis, the patient initially is treated medic-ally with extra volume infusions and blood replacement

as necessary for symptoms Often the hemolysis is limited after 2–7 days, but when it is unrelenting and/orprogressive, re-intervention to eliminate the high-velocitylesion is necessary Usually additional transcatheter deliv-ered occlusion devices to stop the flow will solve the prob-lem, but rarely surgical removal of the device and closure

self-of the vessel/leak will be necessary

Complications of catheter occlusions of aorta to left ventricular tunnels and peri-prosthetic valvular leaks

Like all other therapeutic catheterization procedures, andparticularly those within the systemic arterial system, all

of the basic hazards of the cardiac catheterization ure are present with these procedures The occlusion ofaortic to left ventricular tunnels and perivalvular leaks ofboth the aortic and mitral valves with catheter-delivereddevices involves very extensive manipulations proximal

proced-to the head vessels in the systemic arterial system withwires, catheters and delivery systems, which, in turn, cre-ates a very high potential for cerebrovascular embolic

problems There must be constant vigilance to prevent the

introduction of air into the system and to keep all wirespassing through catheters, catheters and delivery systems

on a continuous flush to prevent thrombi from forming.The potential for embolization of the occlusion deviceitself from the defect to the systemic circulation is alwayspresent The equipment must be available for the immedi-ate capture and retrieval of any embolized device before itresults in permanent damage to a vital organ

The occlusion of perivalvular leaks around mechanical

prosthetic valves involves extensive

catheter/wire/deliv-ery system manipulations adjacent to the valve orifice

and the mechanical leaflets of the involved prosthetic

valves with the potential for entrapment in the valve

and significant interference with the valve’s function The

manipulations in the area of these valves must be directed

very precisely and cautiously, and observed continually

on biplane fluoroscopy to avoid the valves.

Any of the occlusion devices placed in perivalvular

leaks can interfere with the function of a mechanical

pros-thetic valve Prevention is the optimal management by the

use of retrievable devices and the very careful positioning

and testing of each device as it is delivered If the function

of a prosthetic valve is compromised significantly, the

device must be removed either by retrieval with a catheter

technique or surgically

Summary

Unwanted shunts and/or vascular leaks in virtually all

cases can be occluded with a catheter-delivered device

Refinements in the imaging systems, the variety of

avail-able devices, and improvements in delivery have made

these procedures more effective and far safer In 2005,

trans catheter occlusion of residual and/or unwanted

vascular communications and/or leaks should be the

primary approach for their management

References

1 Bookstein JJ et al Transcatheter hemostasis of

gastrointest-inal bleeding using modified autogenous clot Radiology 1974;

113(2): 277–285.

2 Gold RE and Grace DM Gelfoam embolization of the left

gastric artery for bleeding ulcer: experimental

considera-tions Radiology 1975; 116(3): 575–580.

3 Kaufman SL et al Transcatheter embolization with

micro-fibrillar collagen in swine Invest Radiol 1978; 13(3): 200–

204.

4 Zuberbuhler JR et al Tissue adhesive closure of

aortic-pulmonary communications Am Heart J 1974; 88(1): 41–46.

5 Serbinenko FA [Balloon occlusion of saccular aneurysms of

the cerebral arteries] Vopr Neirokhir 1974; 4: 8–15.

6 Gianturco C, Anderson JH, and Wallace S Mechanical

devices for arterial occlusion Am J Roentgenol Radium Ther

Nucl Med 1975; 124(3): 428–435.

7 Yamamoto S et al Transcatheter embolization of bronchial

collateral arteries prior to intracardiac operation for tetralogy

of Fallot J Thorac Cardiovasc Surg 1979; 78(5): 739–743.

8 Grifka MR and Jones TK Transcatheter closure of large PDA using 0.052″ gianturco coils: controlled delivery using a biop-

tome catheter through a 4 French sheath Catheter Cardiovasc

Interv 2000; 49(3): 301–306.

9 Kuhn MA and Latson LA Transcatheter embolization coil

closure of patent ductus arteriosusamodified delivery for enhanced control during coil positioning Cathet Cardiovasc

Diagn 1995; 36(3): 288–290.

10 Uzun O et al Transcatheter occlusion of the arterial duct with

Cook detachable coils: early experience Heart 1996; 76(3):

269–273.

11 Grifka RG et al New Gianturco-Grifka vascular occlusion device Initial studies in a canine model Circulation 1995;

91(6): 1840–1846.

12 White RI Jr et al Therapeutic embolization with detachable

balloons Physical factors influencing permanent occlusion.

Radiology 1978; 126(2): 521–523.

13 Goto K et al Permanent inflation of detachable balloons with

a low-viscosity, hydrophilic polymerizing system Radiology

1988; 169(3): 787–790.

14 Moore JW and Murphy JD Use of a bow tie stent occluder for

transcatheter closure of a large anomalous vein Catheter

Cardiovasc Interv 2000; 49(4): 437–440.

15 Khan MS and Moore JW Treatment of abdominal aortic pseudoaneurysm with covered stents in a pediatric patient.

Catheter Cardiovasc Interv 2000; 50(4): 445–448.

16 Tulloh RM and Rigby ML Transcatheter umbrella closure of

aorto-pulmonary window Heart 1997; 77(5): 479–480.

17 Jureidini SB, Spadaro JJ, and Rao PS Successful transcatheter closure with the buttoned device of aortopulmonary win-

dow in an adult Am J Cardiol 1998; 81(3): 371–372.

18 Richens T and Wilson N Amplatzer device closure of a

resid-ual aortopulmonary window Catheter Cardiovasc Interv 2000;

50(4): 431–433.

Transcatheter closure of patent ductus arteriosus (PDA)

with one or more of several (many) devices/techniques

now represents the established, standard approach for the

correction of a PDA throughout the entire world

includ-ing even the United States (US) There now is one device

approved even by the US Food and Drug Administration

(FDA) specifically for closure of the PDA

The patent ductus arteriosus presents in an infinite

vari-ety of shapes and each shape appears in an equally wide

range of sizes1 The most common shape is conical with a

large aortic end (ampulla) tapering to a narrower

pul-monary end The persistent ductus however, can be long

and “finger-like” with or without an additional tapering

at the pulmonary end; it can have several areas of

narrow-ing or even be shaped like a long tube with no tapernarrow-ing at

any area The long tubular ductus usually is relatively

straight but can be tortuous At the other extreme the PDA

can be a very short and flat communication or “window”,

between the aorta and the pulmonary artery Each shape

of the PDA can vary in diameter at its narrowest portion

from less than 1 mm to greater than 10 mm The flow

through the ductus and the differential in pressure

be-tween the two ends of the ductus depend upon the size of

the ductus and the relative resistances of the pulmonary

arterial and systemic arterial circulations Obviously, no

single device can be optimal or even applicable for all

of the varieties in the size and shape of the PDA, and

fortunately over the past three plus decades multiple

devices and the procedures for delivering them have

been developed for transcatheter closure of the patent

ductus arteriosus

Transcatheter occlusion of the PDA was first introduced

by Porstmann in 1967 et al.2, however, PDA occlusion in

the catheterization laboratory was not popularized until

the early 1980s, when the more practical Rashkind™ PDA

occluder was introduced into clinical trials in the US and

the rest of the world Since the introduction of theRashkind™ PDA device, many different devices and tech-niques for transcatheter occlusion of the PDA have beenintroduced and used clinically for almost 25 years with

minimal morbidity and no mortality from the catheter cedures The success and complications of the procedures

pro-for transcatheter occlusion of the PDA have been variableand are related both to the type and size of the ductus, theparticular device used, the delivery procedure utilized,and to the experience and skill of the operators perform-ing the occlusion

Depending upon where one resides in the world, in

2004 there were between three and ten devices and/or

techniques available for PDA occlusion in the

catheteriza-tion laboratory The availability of devices in any lar area or country depends upon the developmentalstage of new devices and the approval for human use

particu-by the regulatory agencies of the various countries.Ironically, most of the viable devices and techniques forocclusion of the PDA in the catheterization laboratorywere designed, developed and clinically validated in the

US Unfortunately for the pediatric and adult patientswith congenital heart lesions in the US, the FDA has beenultra conservative toward children in their review andapproval of specific devices As a consequence, the UnitedStates has the fewest devices available and the mostrestricted “approval” for the use of those devices that areavailable for specific congenital heart lesions throughoutthe rest of the world

There are several devices that are approved by the FDA

in the US for the occlusion of “vascular lesions” that havebeen adapted for transcatheter PDA occlusion and areused for PDA occlusions in the catheterization laboratory

in the US These devices and their special delivery niques have been demonstrated to be extremely safe and

tech-effective and are accepted as the standard therapy for PDA occlusion by all knowledgeable medical professionals

throughout the world, including in the US However, in

2004, the Amplatzer™ PDA device was the only device