FUTURE TRENDS
PERCUTANEOUS VALVE IMPLANTATION: PAST, PRESENT AND FUTURE.
Younes Boudjemline*, MD; Philipp Bonhoeffer**,
MD. * Service de Cardiologie Pédiatrique, Hôpital
Necker Enfants Malades, Paris, France. ** Cardiothoracic
Unit, Great Ormond Street Hospital, London, UK,
and The Institute of Child Health, London, UK.
The percutaneous approach for valve replacement
has recently been introduced. This procedure is
presently available for patients with artificial
right ventricular outflow tract conduits only.
The technique is safe, but is limited to rare
stereotyped clinical situations. Further technical
improvements will broaden the spectrum of indications
to pulmonary valve replacement regardless of the
anatomy of the outflow tract to the aortic valve
and possibly to atrio-ventricular valves. In this
report, we review our experience in valve replacement
through a percutaneous technique in humans, describing
the ongoing experimental work.
(Heart Views. 2002;3(2):55-60) © 2002 Gulf Heart
Association.
Key Words:
Catheterization
valve diseases
stent
he first surgical attempt of valve replacement
was by Hufnagel in the early 1950’s. He surgically
implanted a mechanical ball valve in the descending
aorta to palliate chronic aortic insufficiency[1,2].
With the development of extracorporeal circulation,
native valve replacement became the conventional
treatment for valve diseases making implantation
in heterotopic position a forgotten entity. In
parallel, over the last years, cardiac catheterization
has developed progressively, shifting from diagnostic
to interventional procedures. Various diseases
namely valvular or vascular stenosis, and septal
defect or shunt closures, are now treated primarily
by a transcatheter technique [3-6]. Until recently,
valve replacement has, however, remained entirely
in the surgical domain. In this manuscript, we
give an insight into our experience in transcatheter
implantation of cardiac valves, describing the
device and discussing present and future indications.
The valve of choice for percutaneous implantation
is a valve which 1) is easily available at variable
sizes; 2) is biocompatible; 3) has excellent intrinsic
properties; 4) has a low profile; 5) can be sutured
into an expandable stent; and 6) does not lose
its property after crimping and re-expansion.
After testing different types of valves (Figure
1), we finally opted for the bovine jugular venous
valve. Bovines have native valves in their
Figure 1: Various valves have been
tested. This picture is showing one of this attempt
using an aortic homograft after its suture into
a platinum stent.
jugular veins which allow the filling
of the right heart, and avoid stasis of the blood
while these animals hold their head at a low level
such as during feeding. The bovine jugular venous
valve was introduced in the late 1990s in surgical
practice as a right ventricular to pulmonary artery
substitute (Contegra©, Medtronic) and preliminary
results were excellent [7]. Various sizes are
available from 8mm to 22mm. The leaflets of the
valve are highly mobile, thin and redundant. Despite
its excellent properties, the size of the venous
wall is too thick to allow it to be used as a
substitute for percutaneous insertion. Fortunately,
the wall can be reduced in profile significantly
without interfering with the valve function.
Figure 2: Preparation of the device: the venous
wall is too thick to allow for percutaneous implantation.
Therefore, the venous wall is firstly reduced
in profile (left) and then sutured in a platinum
stent (right).
Figure 3: After preparation and
sterilization, the device is reduced onto the
balloon of the delivery system (up) and then covered
previous to its skin insertion (down).
After removal of unnecessary tissue
from the external venous wall, a section of the
prepared valved vein is sutured into a platinum-iridium
stent, which we developed in cooperation with
Numed Inc (Figure 2). This is a highly malleable
stent that can be crimped and re-expanded several
times without damage. After preparation, the device
is sterilized, cross-linked with a buffered glutaraldehyde
solution and stored in an alcoholic solution according
to industrial protocols.
A custom-made delivery system was developed in
parallel with the valved stent. It is based on
a frontloading technique using the BiB (balloon
in balloon) technology (Numed Inc) for stent delivery.
At the time of implantation, the device is delicately
hand-crimped on the outer balloon of the delivery
system and then covered (Figure 3). The outer
diameter of the device when fitted on the balloons
is approximately eighteen French.
Eleven lambs underwent catheterization for percutaneous
pulmonary valve implantation under general anesthesia[8].
The valve device was inserted percutaneously in
pulmonary position on to a previously positioned
guide wire according to standard stent placement
technique. Seven of the lambs had a pulmonary
valve insertion through a right jugular venous
access. Technical failure occurred in the remaining
four lambs because of the narrow angle between
the tricuspid valve and the right ventricular
outflow in our model. In the other seven lambs,
five valved stents impinged on the function of
the native valve and two were unsatisfactorily
positioned just adjacent to the native valve.
At 2 months evaluation, one stent was slightly
stenotic, with a systolic pressure gradient between
the right ventricle and the pulmonary artery of
15-mm Hg. Six of the seven successfully implanted
valves were angiographically competent. One mild
regurgitation was noted on the last animal, probably
aggravated by the position of the catheter through
the pulmonary valve during the contrast injections.
Two months after the insertion, valves were electively
explanted. Macroscopic examinations showed that
four of the five precisely inserted valves had
no sign of valvar calcification. The remaining
valve showed early signs of degeneration including
macroscopic calcification, cuspal retraction and
partial fusion of the commissures. This was attributed
to the sub optimal sterilization process used
in this initial setting. The two stents incorrectly
positioned were malfunctioning. Fibrous tissue
covered the leaflets of these valves. Their functions
were restored in vitro after the removal of fibrous
tissue. This animal experimentation confirmed
the feasibility of the technique. Technical implantation
difficulties were mainly related to the animal
model.
The main concern was about the durability of the
valves. However, as far as the durability is concerned,
the surgery and the percutaneous approach shared
the same limitations. Therefore, the excellent
results of the initial experience of the bovine
jugular vein in the surgical setting encouraged
us to introduce percutaneous approach in humans.
We initially limited the indication
to pulmonary valve insertion to patients with
surgically created communications between the
right ventricle and the pulmonary artery namely
prosthetic conduits, valves, and patch reconstructions[9].
This stereotyped situation allowed us the precise
knowledge in advance of the anatomy of the right
ventricular outflow tract. Nine patients with
significant pulmonary regurgitation and/or right
ventricular outflow tract obstruction were selected
for a preliminary study. Approval for percutaneous
pulmonary valve replacement was given by a certified
ethical committee (CCPPRB, Paris Cochin, Paris,
France). Fully informed consent was obtained from
the parents, where the patient was a child, and
from the patient themselves, if an adult. The
seven children ranged in age between 10-17 years,
with a mean age of 12.1± 2.3 years. Three had
a tetralogy of Fallot without pulmonary atresia,
and 3 with pulmonary atresia. One patient had
absent pulmonary valve syndrome. Six of the seven
patients had previous palliative surgery with
one or more modified Blalock-Taussig shunts followed
by a total repair with conduit placement. Most
of the children underwent reoperation for replacement
of the initial conduit with a larger one during
infancy. One adult patient (aged 38 years) had
a tetralogy of Fallot repaired initially at 3
years of age, followed by two reoperations to
replace the pulmonary valve 15 and 25 years after
the original repair. The last patient (aged 18
years) had a congenital aortic stenosis with a
sub-valvular membrane. His membrane was resected
at 5 years of age. In July 2001, a Ross procedure
was performed to replace his aortic valve, and
a 26mm homograft was inserted in the pulmonary
position. All patients were symptomatic with effort
intolerance and breathlessness, and needed surgery
for a conduit replacement or pulmonary valvulation
due to significant stenosis and insufficiency
of the conduit. Before the procedure, six patients
were in New
Figure 4
 |
Figure 5: Doppler images showing pulmonary
regurgitation (left) prior to valve insertion
(left) and competence of the newly implanted
valve (right). Note the click at the closure
of the valve.
York Heart Association (NYHA) class II
and three were in NYHA class III with cardiomegaly,
moderate to severe right ventricular dilatation
and dysfunction on echocardiography. The
valved/stent was successfully implanted
in all nine patients. Immediately after
implantation, the hemodynamic and angiographic
evaluation confirmed competence of the newly
implanted valve in seven patients (Figure
4). In two patients, the valve was deployed
slightly lower than intended, lying slightly
in the infundibulum leading to insignificant
paraprosthetic regurgitation. The relief
of the conduit obstruction was partial in
three patients. The fluoroscopy time ranged
from 17-129 minutes with a mean of 48 minutes.
The time for the procedure of valved/stent
implantation improved significantly after
the first cases. All the patients were discharged
between 1 and 5 days after the procedure
with aspirin at 2mg/kilo/day. Echocardiography
immediately before discharge confirmed the
perfect competence of the implanted valve
in all patients (Figure 5). The slight paraprosthetic
leak present in two patients after the procedure
disappeared on color Doppler echocardiography.
At the latest follow-up ranging from 1-17
months (mean 11 months), all the patients
had improvement of their symptoms especially
in adults patients who were the most symptomatic.
Color Doppler echocardiography showed a
fully competent pulmonary valve in six,
and trivial regurgitation in the remaining
three.
A reduction of the right ventricular size
and an improvement of systolic function
of the right ventricle was suspected on
transthoracic echocardiography evaluation,
and confirmed by MRI scan. Echocardiography
also showed the persistence of a moderate
elevation of the right ventricular systolic
pressure in four patients.
One of the major challenges of the future
will be to implant such a valve in more
variable anatomy of the right heart. Indeed,
conduit valvulation is a limited indication.
Most pulmonary insufficiency occurs after
surgical repair of tetralogy of Fallot.
These patients have extremely dilated pulmonary
trunks that make percutaneous implantation
of valved stents as presently designed impossible.
One alternative would be to implant a valve
in the two proximal pulmonary branches.
However, this stenting would leave a regurgitant
fraction originating from the pulmonary
trunk. The clinical benefit of this technique
would need to be demonstrated. The valvulation
of a Fontan circulation could also be an
interesting indication to evaluate in the
near future. Indeed, right atrial dilatation
is a frequent complication occurring in
long-term follow-up [10]. It generates arrhythmia
that has major repercussions on cardiac
function [11]. The valvulation of such patient
might prevent the occurrence of arrhythmia.
Percutaneous aortic valve replacement
is a major challenge. This has been considered
not possible because of 1) the proximity
of coronary arteries, and 2) the anatomic
continuity between the aortic and the mitral
valve. The design of the valved stent in
the present study does not allow its implantation
in sub-coronary position. But, as a first
step, we evaluated the function of the venous
valve at high pressure in vitro and more
recently in an animal setting. To achieve
this goal, we implanted valved stent in
the descending aorta of lambs after the
creation of a moderate or severe aortic
regurgitation. All animals with severe aortic
insufficiency died suddenly despite perfect
competence of the implanted valve. The deaths
were attributed to coronary flow impairment
secondary to aortic regurgitation. All animals
with moderate aortic regurgitation survived.
All implanted valves were perfectly functioning
during the first 2 months of implantation.
At the 3-month evaluation, the aortic regurgitation
has disappeared and none of the implanted
valves were competent. At macroscopic evaluation,
all implanted valves were covered with a
fibrinous tissue and no tear was found on
the aortic valve. According to our previous
study, which showed that unused valves are
rapidly covered with a fibrous tissue, we
speculated that the healing of the aortic
valve led to the dysfunction of the implanted
valves. In this initial experiment [12,13],
we demonstrated the feasibility of implantation
in the descending aorta, reproducing through
a percutaneous approach the pioneering work
of Hufnagel et al [1]. As a second step
to approach the native aortic valve, we
redesigned the device [14,15]. Initially,
we wrongly thought that the venous wall
was necessary for the valve to function.
In experimental studies, we first verified
that the removal along the commissures of
the venous wall did not alter the function
of the valve in vitro and secondarily that
these dissected venous valves could function
at high pressures (in the descending aorta)
in an animal setting very similar to the
previous one. The next step was implantation
in the native position. We first tested
second-generation valved stents. This type
of stent liberates space for coronary orifices
but its orientation is impossible. Therefore,
the stent was deployed 1 cm below the aortic
annulus. All lambs died suddenly during
the procedure despite “successful” (but
inappropriate) delivery of the device: one
from a severe mitral valve insufficiency;
the second from an acute obstruction of
the coronary artery orifices; the third
migrated prematurely in the ascending aorta.
The last device was inappropriately placed
in the left ventricle leading to a paraprosthetic
leak. This study highlighted the need for
perfect orientation and anchoring of the
device. To achieve these goals, we redesigned
the valved stent (“third generation”) with
a deployment strategy in two steps. The
first step assured the orientation and the
locking of the device in the aortic orifice.
The second acted as a supporting structure
for the graft. To guaranty the orientation
of the device, we fixed an autoexpandable
nitinol stent on a second generation valved
stent (Figure 6). The branches congruous
with the commissures of the valve were interdependent
with the platinum stent wires and could
not be deployed separately. Contrarily,
the branches congruous with the leaflets
were not sutured to the platinum stent
Figure 6
|
|
|
Figure 6: On the left (a), the nitinol
stent is shown. On the center (b), the branches
of the stents congruous with the commissures
of the valve are sutured so that they can
not be deployed separately. On the right,
the non-sutured branches of the nitinol
stent are deployed defining 3 free spaces
between the two stents for the native valve
leaflets (c and d).
|
Figure 7: Aortogram showing the
competence of the implanted valve and the non-obstruction
of the coronary arteries.
wires so that when the platinum
stent was reduced, the non-sutured branches of
the nitinol stent were deployed at a diameter
of 23mm defining a free space between the inner
valved stent and the nitinol stent. In an acute
setting, we successfully implanted this third
generation-valved stent in five animals. No coronary
orifices were obstructed and no mitral valve impairment
was noted. There was no stent migration. Two valves
were dilated to the proper diameter and were perfectly
competent (Figure 7). Three implanted valves were
overdilated and were incompetent from mild to
severe because of a non-coaptation of the valve
leaflets. Further experiments are obviously needed
to confirm these early results. In particular,
questions regarding the durability of the valve
in systemic pressure have to be answered before
considering human application.
We report in this study the development of a non-surgical technique
to implant valves. This technique has already
been applied in humans in patients with artificial
pulmonary artery trunk. The technique is safe
but is presently limited to a rare and stereotyped
clinical situation. Further technical improvements
will allow expanding the spectrum of indications
to patients with large pulmonary trunk or with
Fontan circulation. More recently, we approached
the aortic valve and opened the field of percutaneous
valve replacement in heterotopic position as well
as in native position using newly designed valved
stents. Studies with longer follow-up are presently
missing but will permit evaluating the function
and durability of bovine venous valve in systemic
pressures. However, even if the results of these
studies are good, large diameters of valves are
difficult to find making its use as a common substitute
in adults improbable. Therefore, other valvular
substitutes (artificial or biological) have to
be found to overlap all the sizes used in clinical
practice and to be suitable for right and left
sided valves. The technology we developed for
bovine valve placement can be used for any valvular
substitutes. Non semi-lunar valves were not approached
yet through a percutaneous technique but would
probably be in the next decade.
The "Fondation de l'Avenir", Paris, France and
the "Fédération Française de Cardiologie", Paris,
France. Research at the Institute of Child Health
and Great Ormond Street Hospital for Children
NHS Trust benefits from R&D funding received from
the NHS Executive.
1. Hufnagel CA, Harvey WP, Rabil P, McDermott
TF.
Surgical correction of aortic
valve insufficiency. Surgery 1954;35 :673-77.
2. Hufnagel CA, Gomes MN. Late follow-up of ball-valve
prostheses in the
descending thoracic aorta.
J Thorac Cardiovasc Surg 1976;72:900-9.
3. Letac B, Cribier A, Koning R, Lefebvre
E. Aortic stenosis in elderly patients aged 80
or older. Treatment by percutaneous
balloon valvuloplasty in a series of 92 cases.
Circulation
1989;80:1514-20.
4. Hockings B, Wilkinson JL, Blanton C. Amplatzer
atrial septal defect closure devic
used in adult patients.
Heart Views 2000;1:248-50.
5. Powell AJ, Lock JE, Keane JF, Perry SB. Prolongation
of right ventricular to pulmonary
artery conduit life span by percutaneous
stent implantation intermediate-term results.
Circulation 1995;92:3282-88.
6. Bonhoeffer P, Esteves C, Casal U, Tortoledo
F, Yonga G, Patel T, Chisholm R,
Luxereau P, Ruiz C. Percutaneous mitral valve
dilatation with the Multi-Track System.
Catheter Cardiovasc Interv 1999;48:178-83.
7. Corno AF, Hurni M, Griffin H, Jeanrenaud
X, von Segesser LK. Glutaraldehyde-fixed
bovine jugular vein as a substitute
for the pulmonary valve in the Ross operation.
J Thorac Cardiovasc Surg 2001;122:493-4.
8. Bonhoeffer P, Boudjemline Y, Saliba Z, Hausse
AO, Aggoun Y, Bonnet D, Sidi D,
Kachaner J. Transcatheter implantation
of a bovine valve in pulmonary position:
a lamb study. Circulation 2000;102:813-6.
9. Bonhoeffer P, Boudjemline Y, Saliba Z, Merckx
J, Aggoun Y, Bonnet D, Acar P, Le
Bidois J,
Sidi D, Kachaner J. Percutaneous replacement of
pulmonary valve in a right- ventricle
to pulmonary-artery prosthetic conduit with valve
dysfunction.
Lancet 2000;356:1403-5.
10. Conte S, Gewillig M, Eyskens B, Dumoulin M,
Daenen W.
Management of late
complications after classic Fontan procedure by
conversion to
total cavopulmonary connection. Cardiovasc
Surg 1999;7:651-5.
11. Agarwal KC, Edwards WD, Mair DD, Julsrud PR,
Seward JB, Danielson GK,
Feldt RH. Severe
fibrotic obstruction of Hancock conduit after
Fontan operation.
J Thorac Cardiovasc
Surg 1982;83:791-4.
12. Boudjemline Y, Abdel Massih T, Bonnet D, Sidi
D, Bonhoeffer P. Percutaneous
implantation
of a biological valve in the descending aorta
to treat aortic valve
insufficiency : a sheep study. Med
Science Monitor 2002;8(4):BR20-25.
13. Boudjemline Y, Bonhoeffer P. Percutaneous
implantation of a valve in the descending
aorta in lambs. Eur Heart J
2002. In press.
14. Boudjemline Y, Bonhoeffer P. Percutaneous
aortic valve replacement: will we get
there?
Heart 2001:86;705-6. 15. Boudjemline Y, Bonhoeffer
P.
Steps towards percutaneous
aortic valve replacement. Circulation 2002:105;775-8.
Luigi Nono (oil on canvas, 1889,
Milan, Italy)
|
