Full text of DOI:10.1016/S1051-0443(07)61369-4

The Square Stent-based Large Vessel
Occluder: An Experimental
Pilot Study1
Dusan Pavcnik, MD, PhD
Barry T. Uchida, BS
Hans Timmermans, BFA
Christopher L. Corless, MD,
PhD
Marc Loriaux, MD, PhD
Frederick S. Keller, MD
Josef Ro¨sch, MD
PURPOSE: The purpose of this study is in vitro and in vivo experi-
mental evaluation of a square stent-based vascular occlusion de-
vice for large vessels.
MATERIALS AND METHODS: Square stent-based large vessel oc-
cluders (LVO) 5 mm–50 mm in size were constructed from stain-
less-steel square stents covered by porcine small intestine submu-
cosa (SIS). The LVOs with two back-side barbs were delivered
through a guiding catheter. The LVOs with two back-side barbs
and two frontal barbs were front-loaded and delivered coaxially. A
pusher with a retention mechanism at its end was used for deploy-
ment. In vitro testing for competency was performed with use of a
flow model with pressure increases. In an experimental pilot study
in seven pigs and five dogs, 16 LVOs were placed into the aorta
(n ؍ 4), common iliac artery (n ؍ 2), pulmonary artery (n ؍ 4), and
medial sacral artery (n ؍ 6). Four animals received two LVOs in
different locations. Angiography was performed before and after
placement of each LVO. Animals were followed for as long as 3
months with use of angiography and were then killed for gross and
histologic evaluation.
Index terms: Stents and prostheses
● Interventional procedures, experimental
● Embolization
JVIR 2000; 11:1227–1234
Abbreviations: LVO ϭ large vessel oc-
cluders, SIS ϭ small intestine submucosa
RESULTS: In vitro LVOs with two and four barbs were easily col-
lapsed and pushed through or front-loaded into guiding catheters
(6-F for a 5-mm occluder, 10-F for a 50-mm occluder). A 20-mm LVO
adapted to tubular structures 10–15 mm in diameter, forming poly-
gons 17–18.5 mm in length. In the flow model, LVOs endured pres-
sure increases to 300 mm Hg. In vivo, the LVOs self-expanded and
adapted to the vessel without migration in all cases. The locking
pusher allowed precise LVO placement and engagement of its
barbs into the vessel wall before complete deployment, preventing
dislodgment by blood flow. Complete arterial occlusion occurred
within 10–20 minutes and arteries remained occluded until the an-
imal was killed in all cases. After 2 months, histologic evaluation
revealed replacement of SIS by host tissue and its remodeling with
variable fibrocytes, fibroblasts, and some inflammatory cells. Com-
plete endothelialization was seen on both sides of the LVO.
CONCLUSION: The SIS LVO is effective and reliable for acute and
chronic occlusion in a high flow model in an experimental animal.
1 From the Dotter Interventional Insti-
tute (D.P., B.T.U., H.T., F.S.K., J.R.) and
Department of Pathology (C.L.C., M.L.),
Oregon Health Sciences University, L342,
3181 SW Sam Jackson Park Road, Port-
land, Oregon, 97201. Received February
7, 2000; revision requested February 8,
final revision received April 22, accepted
April 24. Address correspondence to
D.P.; E-mail: pavcnikd@ohsu.edu
EXPANDABLE stents have been
used commonly for more than 10
years for treatment of obstructions
in vascular and nonvascular sys-
tems. Expandable stents also have
taneously placed intravascular oc-
cluders. Self-expandable Wallstents
(Boston Scientific/Meditech, Water-
town, MA), balloon-expandable
Palmaz stents (Cordis, Warren, NJ),
© SCVIR, 2000
great potential as carriers for percu- and self-expandable Gianturco
1227

1228 ● Square Stent-based Large Vessel Occluder
October 2000 JVIR
stents were created as single stents
or connected to other stents with
elongated barbs (Fig 1). When in-
troduced to the tubular structure,
the square stent self-expanded with
complete wall contact to take the
shape of the tubular structure (Fig
2).
● Large Vessel Occluder
A square frame with two or four
barbs, when covered with low-poros-
ity material such as porcine SIS,
becomes an occluder (Fig 3a). A
double LVO is composed of two SIS-
covered square stents connected by
an elongated barb.
To determine an ideal fit in an
artery with a circumference of 2␲r,
the diagonal axis of the square stent
was constrained to the length of ␲r,
forming a diamond. An artery 7 mm
in diameter required a 12-mm square
stent, the diagonal axis of which was
constrained to 13–15 mm; a 14-mm
artery required a 20-mm square stent
with a diagonal axis constrained to
20–22 mm (Fig 3b). The covering
was attached to the corners with 7.0
Prolene (Ethicon, Somerville, NJ)
monofilament suture. The sides of
the covering were then folded back
around the wire and continuously
sutured to the stent frame. When
sutured to the square frame, the
layer of SIS had a flap approximately
2 mm wide, allowing a better seal
between the artery wall and the
LVO. Large vessel occluders were
made in various sizes from 5 mm to
50 mm and were connected with com-
binations of two or three occluders.
Figure 1. The square stents. (a) Single square stent 28 mm in length with four
barbs for self-attachment to the vessel wall (arrows). (b) Two square stents 18 mm
in length connected by an elongated barb into square stent combinations.
Z-stents (Cook, Bloomington, IN)
have been explored for this purpose
(1–5). We present a preliminary re-
port of a new self-expandable square
stent and its potential as a carrier for
a large vessel occluder (LVO). The
new cover material, small intestinal
submucosa (SIS), is derived from the
middle layer of the small intestine of
pigs. It provides an acellular frame-
work that is remodeled by host tissue
while degrading and reabsorbing over
the course of time (6). This report
describes the square stent and the
square stent-based LVO and summa-
rizes the results of our tests of them
in vitro and in pilot animal experi-
mental studies.
● Square Stent
Square stents were constructed
in our research laboratory of stain-
less-steel wire 0.006–0.018 inches
in diameter. Selection of the wire
diameter depends on the desired
size and degree of expansile force of
the square stent. The selected wire
is hand-formed on a wooden tem-
plate with fixed metal pegs, en-
abling the wire to be bent into an
exact square. The distance between
the pegs determines the size of the
square stent. Stent sizes ranged
from 5 mm to 50 mm. The corners
of the square stent were formed
into a coil to reduce stress and fa-
tigue of the wire. The wire ends
were joined by a metal cannula. For
better fixation of the stent, we
● Deployment
The square stent, LVO, or the
combination of the two was deliv-
ered through a 5–10-F guiding cath-
eter or sheath, depending on the
size of the stent, wire diameter, and
the amount of SIS. Square stents,
LVOs without barbs, and LVOs
with only proximal barbs were de-
livered through a single guiding
catheter. Square stents and LVOs
with distal barbs were delivered
coaxially. For deployment, a modi-
fied locking wire guide pusher was
MATERIALS AND METHODS
added 2 or 4 barbs. The square
stents with two barbs had both wire
ends extended 1–2 mm over the
stent frame, forming barbs on oppo-
site corners. On stents with four
barbs, two anchoring barbs were
attached to the other sides of the
stent frame with metal cannulas.
To ensure stability of the square
stent and the barbs, the cannulas
were crimped and soldered. Square
The protocol for the animal stud-
ies was approved by the Oregon
Health Sciences University Animal
Care and Use Committee. All ani-
mal facilities are accredited by the
American Association for the Ac-
creditation of Laboratory Animal
Care and meet all federal (Public
Health Service and National Society
for Medical Research) guidelines for
animal care.

Pavcnik et al ● 1229
Volume 11 Number 9
used for measurement of the metal-
to-surface ratio. For measurement
of the expansile force of the square
stent, a similar method to that de-
scribed by Fallone et al (7) for
Gianturco Z-stents was used. Test-
ing was done at room temperature.
● In Vitro Testing: Wet
A model with a flow rate of 2.5
L/min was used to test LVO compe-
tency and stability with pressure
measurement proximal and distal to
the occluder. The test system was
constructed from a plastic tube flow
loop with an internal diameter of 15
mm. A pump with a flow rate of 2.5
L/min was connected by side arms
to the lower end of the flow model.
At rest without flow, the occluders
were exposed to hydrostatic pres-
sure of 60 mm Hg provided by a
70-cm water column in a plastic
tube above the occluders. Two LVOs
with four barbs were repeatedly
tested with pressure increases to
300 mm Hg. Testing was done at
room temperature with the flow
model in a vertical position.
Figure 2. Deployment of a square stent 28 mm in length. (a) Square stent re-
tained by a wire pusher connected to one barb. (b) Square stent deployed into a
tube 20 mm in diameter.
used with a Bird’s Nest filter (Cook, and catheter with the retention
wire were removed.
Bloomington, IN). The stent or LVO
was folded by bringing its opposite
bends together. The collapsed
square stent or LVO without barbs
or with only back-side barbs was
loaded into a cartridge. From the
cartridge, a pusher wire then
● In Vitro Testing: Dry
To test the compressibility of
folded stent profiles and LVOs, we
used 5-mm and 50-mm square
stents made of stainless-steel wire,
0.0075 inches and 0.018 inches, re-
spectively. Square stents and LVOs
were folded gently without force,
avoiding deformation of the coiled
bends. The minimum inner diame-
ter of the guiding catheter or
● Animal Experiment
Sixteen LVOs 8–20 mm were
tested in vivo in seven pigs and five
dogs. Seven young pigs weighing
25–40 kg were used for acute (n ϭ
4) and chronic studies (n ϭ 3); five
adult dogs weighing 24–36 kg were
used for chronic studies. Each ani-
mal was sedated with Telazol 3–6
mg/kg intramuscularly (Fort Dodge
pushed the device into the delivery
catheter and to the site of its place-
ment. The square stent or LVO
with frontal barbs was front-loaded
into the guiding catheter with the
back barbs protected with cannulas.
The guiding catheter with the de-
vice and introduced pusher was
then coaxially advanced through a
sheath to the desired position. De-
livery in both instances was accom-
plished by withdrawing the guiding
catheter or guiding catheter and
sheath while holding the pusher.
The square stent or LVO self-ex-
panded and adapted to the tubular
structure of the artery when deliv-
ered. The locking pusher assured
placement of the square stent or
LVO in the proper position and pre-
vented its dislodgment by blood
flow before its barbs engaged into
the vessel wall. When the device
was anchored against the vessel
sheath was determined as the di-
ameter at which a device with back- Laboratories, Fort Dodge, IA) and
side barbs or without barbs could be Atropine sulfates 1 mL intramuscu-
delivered smoothly or a square
stent and LVO with four barbs
could be front-loaded without de-
forming the coiled corners.
For self-expansion, wall contact,
metal-to-surface ratio, and expan-
sile force, 20-mm stainless-steel
stents 0.0075, 0.010, and 0.012
inches in diameter were tested.
They were tested in rigid and elas-
tic transparent plastic tubes 10–20
mm in diameter.
larly (American Regent Laborato-
ries, Shirley, NY). The animals
were ventilated by mask with
isoflurane (Burns Veterinary Sup-
ply, Rockville Center, NY) and were
subsequently intubated and ventila-
tion was maintained at 2% isoflu-
rane with 2 L/min of O2.
After induction of general anes-
thesia, a cut-down was performed to
access the left or right femoral or
carotid artery and a 9-F Teflon
sheath (Cook) was inserted for LVO
delivery into the arterial system.
The equation SA ϭ L ϫ 2␲r,
where L is the length of deployed
square stent, yields the surface area For single or coaxial LVO place-
wall, it was released and the pusher of the cylinder. The equation was ment, a 7–9-F LuMax guiding cath-

1230 ● Square Stent-based Large Vessel Occluder
October 2000 JVIR
Figure 3. Large vessel occluder 20 mm in length covered with SIS. (a) Unrestricted previous introduction. (b) LVO after de-
ployment into tubular structure 14 mm in diameter.
eter (Cook) was advanced through
the sheath into the aorta. A percu-
taneous transjugular puncture was
used to access the pulmonary ar-
eter arteries. Two animals received
LVOs in systemic and pulmonary
arteries. Follow-up arteriography
was performed immediately, 15–45
RESULTS
● In Vitro Study: Square Stent
tery. A 9-F Teflon sheath was intro- minutes after deployment, and at
Compressibility-folded stent pro-
file.—The guiding catheter had to
have an inner diameter of at least
1.8 mm (6 F) for 5-mm square
stents or LVOs and at least 3 mm
(10 F) for 50-mm square stents or
LVOs.
duced and an 8-F LuMax guiding
catheter was advanced through it
into the pulmonary artery for LVO
placement.
the time the animals were killed.
Animals were followed by angiogra-
phy at 6 weeks (n ϭ 5), 8 weeks
(n ϭ 2) and 3 months (n ϭ 1). They
were killed with an overdose of pen-
tobarbital sodium (Euthasol; King
Pharmaceuticals, Bristol, TN) ad-
ministered intravenously, and nec-
ropsy with histologic examination
was performed.
Gross examination concentrated
mainly on the occluded aortas and
arteries and surrounding struc-
tures. After a minimum of 48 hours
of fixation in buffered 10% formalin,
the specimens were further sec-
tioned into tissue cassettes, pro-
cessed through alcohol and xylene,
Four animals received one LVO
in their systemic arterial system
and two animals received one addi-
tional pulmonary LVO to evaluate
short-term delivery, stability, and
efficiency of LVO placement. In an
acute experimental pilot study in
four pigs, six LVOs were studied in
the abdominal (n ϭ 1) and thoracic
aorta (n ϭ 1), common iliac artery
(n ϭ 1), pulmonary artery (n ϭ 2),
and medial sacral artery (n ϭ 1).
Two barbed LVOs were placed into
medial sacral and pulmonary arter-
Self-expansion and wall contact.
—A square stent, after being col-
lapsed, did not return to its initial
shape. Its shape changed to that of
a four-sided polygon. The square
stent’s diagonal axis actually short-
ened from 28 mm to 25 mm and the
four 90° angles changed to two 80°
and two 100° angles. This deforma-
tion was the same for both wire
sizes tested. When introduced to the
tubular structure with a diameter
half that of the diagonal axis of the
square stent, the square stent
shortened slightly, self-expanded,
centered, and adapted, with com-
plete wall contact, to the shape of
the tubular structure. A 20-mm
ies. The other LVOs had four barbs. and embedded in paraffin. Five-mi-
For longer-term testing in three
pigs and five dogs, 10 LVOs were
placed into the abdominal aorta
(n ϭ 2), common iliac artery (n ϭ
1), pulmonary artery (n ϭ 2), and
cron paraffin sections were cut and
stained with hematoxylin and eosin
or with Masson’s trichrome stain.
Digital photomicrographs were
made on a Zeiss Axiophot micro-
medial sacral artery (n ϭ 5) to eval- scope equipped with a Polaroid Dig- square stent with a diagonal axis of
uate long-term stability and efficacy ital Microscope camera (Carl Zeiss,
of the LVOs in different large-diam- Jena, Germany).
28 mm fit into tubular structures
10–15 mm in diameter, forming a

Pavcnik et al ● 1231
Volume 11 Number 9
was identical to the position on fol-
low-up arteriography at the time
the animal was killed. In all eight
animals followed for 3 months, fol-
low-up arteriography performed at
6 weeks (n ϭ 6), 8 weeks (n ϭ 3),
and 3 months (n ϭ 1) showed com-
plete permanent occlusion of two
aortas, eight systemic or pulmonary
arteries with multiple collateral
vessels around the excluded vessel
segments, and excellent distal filling
(Figs 5a, 6). No evidence of recanali-
zation of occluded arteries was found
on follow-up arteriography.
● Pathologic Examination
Postmortem examination of four
animals that had been treated
Figure 4. Expansile force of square stents 20 mm in length for an ideal tube di-
ameter of 14 mm made of different wire sizes. Graph shows reduction of the square
stents’ diagonal diameter as a function of increased radial pressure.
acutely showed that the LVO was
securely anchored against the aorta
or arterial wall by the barbs and no
damage to the surrounding struc-
tures was noted. The aorta or arter-
ies were occluded and the LVOs
were covered with thrombus and
fibrinous apposition, filling the
space around the device. Traces of
barbs penetrating into the arterial
wall were seen after removal of the
LVO. Necropsies of the eight ani-
mals undergoing longer-term fol-
low-up revealed incorporation of the
SIS LVO into the aortic or arterial
wall. Smooth, shiny vascular sur-
face was seen on both sides of the
LVO in all animals (Fig 5b). There
was no sign of stainless steel corro-
sion or fracture. Histologic evalua-
tion revealed SIS replaced by host
tissue and remodeled with variable
fibrocytes, fibroblasts, some inflam-
matory cells, and vascular endothe-
lial cells. The suture lines of the
LVO were inflamed with foreign-
body reaction. There was complete
endothelialization of the LVO cov-
ered by endothelialized neointima
on both sides (Fig 5c).
cylinder 17–18.5 mm in length. In
tubular structures 16–20 mm in
diameter, square stents with barbs
anchored well against the wall but
did not achieve complete wall con-
tact because their shortened diago-
nal axes exceeded the vessel diame-
ter. Stents without barbs had a ten-
dency to move in tubes larger than
15 mm in diameter. These charac-
teristics were the same for both
wire sizes.
Metal-to-surface ratio.—It was
calculated that, because of the square
stent’s low profile, only 5%–7% of the
vessel surface is covered by metal.
Expansile force.—Expansile force
was measured by attaching weights
to the coiled corners of the folded
stent. Values of these measure-
ments are plotted graphically in
Figure 4. Force of expansion de-
pended on wire diameter, con-
strained stent diameter, and length
of the square stent. Under increas-
ing external pressure, the square
stents changed shape, forming a
four-sided polygon. After this defor-
mation, stents were completely elas-
tic and compliant and they retained
their original diameter after re-
moval of the load.
LVOs endured pressure increases to
300 mm Hg and there was no frac-
ture of the material.
● In Vivo Study
All 16 LVOs were easily intro-
duced via either the femoral artery
(n ϭ 6), common carotid artery (n ϭ
6), or jugular vein (n ϭ 4) into the
abdominal aorta (n ϭ 3), thoracic
aorta (n ϭ 1), common iliac artery
(n ϭ 2), middle sacral artery (n ϭ
6), or pulmonary artery (n ϭ 4).
Large vessel occluders with proxi-
mal barbs were delivered through a
single guiding catheter (n ϭ 10).
The LVOs self-expanded, centered,
and adapted properly in large ves-
sels after placement in all animals.
The locking pusher allowed precise
LVO placement and enabled its
barbs to engage the vessel wall in
all cases. The introduced LVOs re-
mained stable and no migration
was observed. Arteriography imme-
diately after placement revealed
minimal leak around the LVOs. Fol-
low-up arteriography performed 20
DISCUSSION
To serve efficiently as a device
minutes postprocedure showed com- carrier, an expandable stent should
plete arterial and aortic occlusion
with no leaks around the LVOs.
In all cases, the position of the
LVO immediately after placement
have a low enough profile to be in-
troduced through a small catheter
but should have sufficient expansile
force to securely attach to the arte-
Flow model.—Results of testing
in the flow model showed that the

1232 ● Square Stent-based Large Vessel Occluder
October 2000 JVIR
rial wall in high-volume, high-pres-
sure flow. The self-expandable
square stent fulfills these conditions
and has potential to be a suitable
carrier for intravascular and intra-
cardiac devices.
To construct an LVO, we covered
the square stent with a biologic ma-
terial, porcine SIS, which is very
thin and does not significantly in-
crease the size of guiding catheters
used for introduction. Depending on
their size, LVOs can be introduced
through 6–10-F catheters. Even
though it is very thin, SIS cover is
durable, and when attached to a
square stent with two or four barbs,
it forms an LVO that is stable and
withstands high-volume, high-pres-
sure flow. However, it is essential
that the LVO is properly sized in
relation to the vessel that will be
occluded. In our experience, the for-
mula that half the diagonal axis of
the square stent represents the op-
timal diameter for the tubular
structure to be occluded is valuable
in estimation of the proper LVO
size. Complete contact between the
diamond-shaped stent frame and
the cylindrical vessel and a flap of
overhanging SIS material provided
a good seal between the LVO and
the vessel wall. The SIS overhang
also provided a basis for ingrowth of
native intimal tissue from the ves-
sel.
Figure 5. Placement of LVO into ab-
dominal aorta. (a) Baseline aortogram.
(b) Follow-up aortogram immediately
after LVO placement shows occlusion
of aorta with minimal leak (arrows).
(c) Aortogram 2 months after place-
ment shows complete aortic occlusion
with extensive collateral circulation
with filling of aorta and its branches
distal to occlusion. (d) Autopsy speci-
men of the occluded aorta shows neo-
intima completely covering the oc-
cluder. Large collateral lumbar arter-
ies (arrow). (e) Photomicrograph of a
longitudinal cross section of aortic SIS
occluder. The area of loose connective
tissue in the left half of the image cor-
responds to organized thrombus,
Large vessel occluders are ra-
diopaque and easy to place. The
tendency of a single-body LVO to
spring out of the catheter during
release is eliminated by a retention
mechanism attached to a wire
pusher. In all our 16 LVO place-
ment procedures in animal studies,
we did not observe any misplace-
ment during release. Also, all LVOs
stayed in place during the study
with no evidence of migration.
As a biomaterial capable of pro-
moting reconstructive tissue re-
sponse rather than nonspecific scar
tissue formation (11–17), SIS has
several advantages over other cov-
ers for LVOs, such as Dacron, poly-
tetrafluouroethylene, silicone, or
polyurethane. Present experience
with these materials show that they
do not remodel (1,3,5,8,9). There-
fore, no ingrowth of host cells is
which was lined by intact endothelium
when viewed at high power. This area
includes the region in which the SIS
material was originally present (ar-
row), but the SIS is no longer distin-
guishable from the host tissue. The
adjacent elastic wall of the vessel
shows slight compression, but there
are no significant changes within the
adventitia. ͓Trichome stain, original
magnification 25ϫ͔.

Pavcnik et al ● 1233
Volume 11 Number 9
Figure 6. Placement of LVO in the pulmonary artery. (a) Pulmonary arteriogram before occluder placement. Pulmonary arte-
riograms 6 weeks after occluder placement show complete pulmonary artery occlusion, with extensive collaterals and filling of
artery distal to occlusion. (b) Early phase. (c) Late phase.
possible, and this may be a reason
that these prosthetic materials oc-
casionally fail and recanalize in
long-term studies (8,10). In addi-
tion, they are thicker and require a
larger introducing catheter. As an
acellular biomaterial, SIS has been
shown to be a bioscaffold with ex-
cellent strength in animals and hu-
mans in intraarticular ligament
grafts (11,12), articular cartilage
repair (13), vascular grafts (14),
dermal grafts (15), and grafts for
urinary (16) and hernia repair (17).
With immunohistochemical analysis
of SIS-remodeled aortic tissue, Hiles
et al (6) demonstrated that SIS ma-
terial degrades and reabsorbs over
the course of time. Like previous
studies, our experimental study
demonstrated the endovascular re-
modeling ability of SIS with vascu-
lar tissue in two different species.
Although SIS comes from a biologic
source, studies to date have demon-
strated no problems with rejection
(6,14,16). The LVO remodeled as an
ingrowth of the host cells, incorpo-
rating it into the vascular wall and
making it less likely to fail.
and aortic valves, and branched
adapter stent-grafts.
spindle for transcatheter vascular
occlusion: in vivo experiments: work
in progress. Radiology 1993; 188:95–
100.
2. Nazarian GK, Qian Z, Vlodaver Z,
et al. Evaluation of a new vascular
occlusion device. Invest Radiol 1993;
28:1165-1169.
3. Moss JG, Laborde JC, Clem MC, et
al. Vascular occlusion with a bal-
loon-expandable stent occluder. Ra-
diology 1994; 191:483–486.
4. Pozza CH, Gomes MR, Qian Z, Am-
brozaitis R, Kim JK, Amplatz K.
Transcatheter occlusion of patent
ductus arteriosus using a newly de-
veloped self-expanding device: eval-
uation in a canine model Invest Ra-
diol 1995; 30:104–109.
5. Kato N, Semba CP, Dake MD. Use
of a self-expanding vascular oc-
cluder for embolization during endo-
vascular aortic aneurysm repair.
JVIR 1997; 8:27–33.
6. Hiles MC, Badylak SF, Lantz GC,
Kokini K, Geddes LA, Morf RJ.
Mechanical properties of xenogenic
small-intestinal submucosa when
used as an aortic graft in the dog.
J Biomed Mater Res 1995; 29:883–
895.
7. Fallone BG, Wallace S, Gianturco C.
Elastic characteristics of the self-
expanding metallic stents. Invest
Radiol 1988; 23:370–376.
8. Konya A, Wright KC. Preliminary
results with a new vascular basket
occluder in swine. JVIR 1999; 10:
1043–1049.
9. Pollak JS, Egglin TK, Rosenblatt
MM, Dickey KW, White RI. Clini-
SUMMARY
Our experience has shown that
the self-expandable square stent is
a good carrier for intravascular de-
vices, particularly the LVO. It has a
low profile, is easy to place, and its
barbs keep it in stable position.
LVOs covered with SIS are very ef-
ficient and durable occlusive devices
for medium-sized and large vessels
in a high-flow model and in experi-
mental animals. Possible clinical
applications include occlusion of il-
iac and other large-diameter arter-
ies during endoluminal aortic and
arterial aneurysm repair, occlusion
of the aorta, occlusion of large-di-
ameter patent ductus arteriosus,
and percutaneous treatment of
large arteriovenous fistulas. How-
ever, more experience in animals
other than pigs will be necessary to
document the long-term potential of
SIS-covered square stents as LVOs
before their clinical application.
Acknowledgment: The authors
thank Martina Basin Pavcnik and Sheri
Imai for their assistance.
The ability of SIS to stimulate
the body to endoluminally build its
own new tissue suggests great po-
tential for future research with SIS-
covered square stent-based devices
such as vascular occluders, venous
References
1. Schmitz-Rode T, Timmermans H,
Uchida B, et al. Self-expandable

1234 ● Square Stent-based Large Vessel Occluder
October 2000 JVIR
cal results of transvenous systemic
embolotherapy with neuroradiologic
detachable balloon. Radiology 1994;
191:477–482.
as an intra-articular ligamentous
cosa (SIS): utilization as a wound
dressing in full thickness rodent
wound. Ann Plast Surg 1995; 35:
381–388.
graft material: a pilot study in dogs.
Vet Comp Orthop Traumatol 1994;
7:124–128.
10. Sagara K, Miyazono N, Innue H,
Ueno K, Nishida H, Nakajo M. Re-
canalization after coil embolo-
13. Peel SF, Chen H, Renlund R, Bady-
lak SF, Kandel RA. Formation of a
SIS-cartilage composite graft in-
vitro and its use in the repair of ar-
ticular cartilage defect. Tissue Eng
1998; 4:143–153.
14. Lantz GC, Badylak SF, Hiles MC, et
al. Small-intestinal submucosa as
a vascular graft: a review. J Invest
Surg 1993; 6:297–310.
16. Knapp PM, Lingeman JE, Siegel Yi,
Badylak SF, Demeter RJ. Biocom-
patibility of small-intestinal submu-
cosa in urinary tract augmentation
cystoplasty graft and injectable sus-
pension. J Endourol 1994; 8:125–
130.
17. Prevel CD, Eppley BL, Summerlin
DJ, Jackson JR, McCarthy M, Bady-
lak SF. Small-intestinal submu-
cosa (SIS): utilization for repair of
rodent abdominal wall defect. Ann
Plast Surg 1995; 35:347–380.
therapy of pulmonary arteriovenous
malformations study of long term
outcome and mechanism for recana-
lization. AJR 1998; 170:727–730.
11. McGee P. Small intestinal submu-
cosa serves as scaffold for ligament
repair [abstract]. Orthopedics Today
1998; 18:43.
15. Prevel CD, Eppley BL, Summerlin
DJ, Jackson JR, McCarthy M, Bady-
lak SF. Small-intestinal submu-
12. Aiken SW, Badylak SF, Toombs JP,
et al. Small intestine submucosa