Advertisement
Advertisement
November 2009
Management of Paravalvular Regurgitation
An overview of the current imaging technology, devices, and outcomes associated with closure.
The management of periprosthetic valvular regurgitation (PVR) is complex and challenging. These challenges include the identification and quantification of periprosthetic regurgitant defects, choosing the most appropriate treatment option (be it medical, surgical, or percutaneous), and the technical aspects involved in percutaneous closure of these defects.
SCOPE OF THE PROBLEM
PVR is a recognized complication of surgical valve
replacement and is the most common cause of nonstructural
prosthetic valve failure—comprising approximately
one-third of prosthetic valve failures.1 The reported incidence of significant PVR ranges from 1% to 7% in longterm
studies, although the exact incidence is difficult to
quantify and likely remains unknown for a variety of reasons
(including lack of systematic follow-up and technically
difficult imaging).2-6 These defects are difficult to visualize on transthoracic echocardiography and often require transesophageal echocardiographic imaging to provide an accurate
diagnosis. Most leaks are small and asymptomatic, but
when they are moderate or severe, they can lead to significant
morbidity, including heart failure, hemolysis, and
endocarditis. Serial echocardiographic studies suggest that
these defects are not due to a slowly progressive lesion, but
rather, they may occur acutely due to dehiscence of suture
material or chronic low-grade inflammation. The median
time to diagnosis is 1.5 years postoperatively, although the
range is wide and significant leaks can develop years after
an operation.1
PATHOLOGY AND CLINiCAL PRESENTATION
In general, the pathology underlying these defects
involves incomplete apposition of the sewing ring to the
native annular tissue. This may be due to tissue friability, as
seen in degenerative disease, previous infective endocarditis,
chronic low-grade inflammation, or inherent tissue
defects as found in Marfan syndrome.1 Other potential
contributing pathologic factors include extensive tissue calcification
and suture dehiscence or tension, which is frequent
with advanced mitral valve disease. In our experience,
defects are more commonly seen with mechanical
prostheses (approximately two-thirds), in the mitral position
(approximately 80%), and occur most often at the
posteromedial or anterolateral annulus.
Patients with significant PVR can present with a wide variety of clinical symptoms and signs ranging from asymptomatic murmurs to decompensated heart failure and hemolysis. Cardiac cachexia is a frequent finding in advanced cases and may be profound. Heart failure is the most common indication for treatment (over 90% of cases in our experience), while some degree of hemolysis can be demonstrated in more than 40% of cases, if screened carefully.
TREATMENT OPTIONS
Once identified, the choice of treatment options for
patients with significant PVR can be difficult. Medical
therapy is usually of limited success in those with severe
symptoms of heart failure and/or hemolysis. Reoperation
for surgical repair needs to be considered, and a surgical
opinion should be obtained in all cases. Surgical repair is often difficult and complicated, especially in patients with
multiple prior sternotomies. Perioperative mortality is
high and in the range of 6% to 15%.2-4 Increased risks of
perioperative complications, residual paravalvular regurgitation,
and poor outcomes among hospitalized survivors
of operation exist.2
Percutaneous device closure of these defects would be ideal if suitable devices existed and procedural techniques were developed to deliver these devices in a safe and effective manner. Percutaneous closure would avoid surgical morbidity and mortality in this complex and high-risk patient subset but is associated with its own issues. Reducing the degree of PVR may provide patients symptomatic relief, even if some residual leak persists. Interest in developing these techniques has grown rapidly, although the procedures are far from being perfected.
Appropriate patient selection for percutaneous device closure can be challenging (Figure 1). Surgical repair of severe, symptomatic paravalvular leaks should be considered and a surgical opinion obtained. Bacterial endocarditis must be excluded as an etiology before consideration of device implantation, although there is concern that standard peripheral blood cultures may not detect localized infections or inflammation. A disturbing finding in a singlesurgeon series of 24 cases of periprosthetic regurgitation was macroscopic evidence of infection (broadly defined as bland abscess, vegetation, or friable annulus) at reoperation in 67% of aortic valves and 79% of mitral valves.7 This was found despite negative blood cultures. Clearly, vigilance and a high index of suspicion is required in excluding infection before considering device implantation.
Technical challenges involved in percutaneous closure of paravalvular defects center around three main issues: (1) adequate visualization of the 3D geometry of these defects, (2) procedural planning and execution of the varying approaches required to deploy closure devices, and (3) the limitations of the currently available devices.
IMAGING
Imaging of periprosthetic defects is crucial not only in
diagnosis, but in planning successful device closure.
Knowledge of the number of defects and their size and
location in 3D, impacts procedural planning and execution.
Choice of access route and catheters, transeptal puncture
location, and device type are all influenced by periprocedural
imaging. A variety of imaging techniques are currently
available (Table 1). Transthoracic echocardiography is frequently
sufficient for periaortic leaks (as most tend to be
anterior). An important issue is flaring of the regurgitant
color Doppler jet in the left ventricular outflow tract,
which can lead to overestimation of the degree of regurgitation.
In some instances, aortography may be necessary to
assess the degree of regurgitation. Transthoracic imaging
for defects involving the mitral valve is problematic
because of extensive acoustic shadowing from valve rings,
valve leaflets, and heavily calcified mitral annulae. For
perimitral leaks, the most important modality is transesophageal
echocardiography (TEE), and we advocate a
low threshold for performing TEE in patients with prosthetic
mitral valves with any systolic murmur. Real-time (RT) 3D
TEE guidance is now available and is our preferred modality
for diagnosis, as well as for identifying and successfully
crossing defects in the mitral position. RT 3D TEE allows for
precise anatomic localization of defect(s) with attention to
the following characteristics: shape (crescentic vs round),
valve dehiscence, distance from sewing ring, orientation
and movement of prosthetic leaflets, and degree of regurgitation.
Intracardiac echocardiography is less useful because
of acoustic shadowing, but in certain specific instances, it
may be useful. Finally, the success of the device closure and
identification of complications is greatly aided not only by
imaging, but by the development of a common vernacular
and understanding of the defects by both the echocardiographer
and interventionist. For example, echocardiologists
are used to presenting imaging data in the left atrial view,
whereas interventional cardiologists will use the left anterior
oblique view with caudal angulation, which presents the
valve in the mirror image of the standard surgical TEE left
atrial view. Lastly, use of anatomically precise language
(superior, posterior) rather than colloquialisms (up, down)
will aid communication during complex interventional procedures
using combined imaging modalities.
Other imaging techniques may be useful, in particular computed tomography, although this cannot be done intraprocedurally. Magnetic resonance imaging is not usually feasible because of the presence of metallic valves. Lastly, use of biplane fluoroscopic imaging is key during procedures to guide crossing of defects and positioning of devices. Before crossing with a delivery system, care must be taken to ensure guidewires are around the valve and not through it. Precise knowledge of the size and shape of the prosthetic valve, its sewing ring, and any struts, as well as where radiographic markers (if any) are located, is essential to these procedures. A rich variety of prosthetic valves exist, and operators must become familiar with them.
INTERVENTIONAL TECHNIQUES
Operators performing device closures of paravalvular
defects must be comfortable with a variety of interventional
techniques. The general principles involve positioning a
catheter close to the defect (either antegrade or retrograde),
crossing the defect with a guidewire and then advancing a
delivery catheter, and introducing and deploying the device
while maintaining the option of device recovery in the event
of inadequate closure or impingement on adjacent structures,
particularly the prosthetic leaflets. In particular, tilting
disk valves in particular are susceptible to impingement of
normal motion by devices or catheters. For periaortic valve
defects, our approach is usually in a retrograde fashion via
the ascending aorta. Periaortic defects can be localized with
standard inferiorly oriented coronary catheters, such as multipurpose
or Amplatz curves, and crossed with a
hydrophilic 0.035-inch wire. More common mitral defects
may be approached antegrade from the left atrium (our
preferred technique, Figures 2 and 3) or retrograde through
the left ventricular cavity. The antegrade approach requires
comfort with transeptal puncture, a procedure no longer
commonly performed by interventional cardiologists. It is
not uncommon to require additional support to deliver the
catheters through calcified defects. This can be accomplished
by snaring the wire and exteriorizing it, creating an
arteriovenous loop. Snaring can be done in the ascending
aorta, if approaching the mitral valve in an antegrade fashion
or in the left atrium, if retrograde. In patients with both
mechanical aortic and mitral valves, a transapical puncture,
another uncommonly performed technique, may be
required to snare and exteriorize the wire when additional
support is needed. Perimitral leaks may also be crossed
directly from the apex in a retrograde fashion. These techniques
have a substantial learning curve and carry the risk of
serious complications, even in experienced hands, and will
likely remain the province of experienced interventional
centers and operators.
Operators performing device closures of paravalvular defects must be comfortable with a variety of interventional techniques. The general principles involve positioning a catheter close to the defect (either antegrade or retrograde), crossing the defect with a guidewire and then advancing a delivery catheter, and introducing and deploying the device while maintaining the option of device recovery in the event of inadequate closure or impingement on adjacent structures, particularly the prosthetic leaflets. In particular, tilting disk valves in particular are susceptible to impingement of normal motion by devices or catheters. For periaortic valve defects, our approach is usually in a retrograde fashion via the ascending aorta. Periaortic defects can be localized with standard inferiorly oriented coronary catheters, such as multipurpose or Amplatz curves, and crossed with a hydrophilic 0.035-inch wire. More common mitral defects may be approached antegrade from the left atrium (our preferred technique, Figures 2 and 3) or retrograde through the left ventricular cavity. The antegrade approach requires comfort with transeptal puncture, a procedure no longer commonly performed by interventional cardiologists. It is not uncommon to require additional support to deliver the catheters through calcified defects. This can be accomplished by snaring the wire and exteriorizing it, creating an arteriovenous loop. Snaring can be done in the ascending aorta, if approaching the mitral valve in an antegrade fashion or in the left atrium, if retrograde. In patients with both mechanical aortic and mitral valves, a transapical puncture, another uncommonly performed technique, may be required to snare and exteriorize the wire when additional support is needed. Perimitral leaks may also be crossed directly from the apex in a retrograde fashion. These techniques have a substantial learning curve and carry the risk of serious complications, even in experienced hands, and will likely remain the province of experienced interventional centers and operators.
DEVICES
There are currently no devices approved by the US Food
and Drug Administration that are specifically designed for
use in the percutaneous closure of paravalvular leaks.
However, commonly used devices include Amplatzer vascular
plugs, septal occluders, muscular ventricular septal
defect occluders, duct occluders, and vascular coils (AGA
Medical Corporation, Plymouth, MN) (Figure 4). Device
selection and sizing is empiric. The use of sizing balloons can
be fraught with hazard because balloon material may get
stuck, tear, or obstruct normal valve motion, and we do not
recommend it. Imaging cannot usually identify the true
anatomic extent of the defects because the anatomy is
complex and irregular in 3D. We empirically recommend placement of the largest devices that can be safely delivered
without impingement on normal motion of the prosthetic
leaflets. The choice of device is determined by the length
and diameter of the defect, which impacts the waist length
and diameter of the device. The distance from the defect to
the valve struts or leaflets is also an important consideration
because this impacts the maximal tolerable disk size without
impingement on valve function. Certain procedural
characteristics may also be important in choosing the closure
device. For example, the duct occluder cannot be used
with a retrograde approach to a mitral defect due to its single
retention disk. This would be seated on the left atrial
side of the defect and would place it at substantial risk of
embolization. Importantly, defects are rarely round, flat, or
regularly shaped; more often, defects are crescentic and represent
partial dehiscences of the sewing ring. They are also
often calcified and rigid and may not allow the device to
adequately mold to the required shape. This may lead to
oversizing the devices to maximize the degree of defect closure
and enhance device stability, reducing the risk of
embolization. This comes at a tradeoff because the larger the device, the more likely it is to impinge on adjacent structures.
The compromise between stability and profile occasionally
results in patients with defects for which there is no
suitable device for percutaneous closure. We tend to favor
the Amplatzer vascular plug but have used a variety of
devices dependent on the circumstances of the case.
Currently, there is a need for devices designed specifically for closure of these paravalvular defects. An ideal device would be resistant to embolization, conform to various sizes and shapes of defects, and carry a low profile to reduce leaflet impingement and to facilitate delivery. The possibility exists for material composition or local drug delivery to be designed to modify the tissue response or promote healing.1
OUTCOMES
Technical successes haven been reported in modestly
sized series ranging from 63% to 89%.2-5 Clinical success
with improvement in heart failure symptoms and/or a
reduction in the degree of hemolysis was seen in 37% to
63% of patients.2-5
Our experience at the Mayo Clinic has continued to grow, with almost 80 procedures completed. Approximately 80% are mitral leaks and 65% are mechanical prostheses. Nearly all patients present with heart failure symptoms (97%) and 44% have transfusion-dependent hemolysis. Nearly two-thirds of patients have had at least two sternotomies. The successful closure rate is approximately 80%. To date, there has been no procedural death, myocardial infarction, stroke, valve dehiscence, or urgent surgeries. There have been two late hemothoraces after left ventricular puncture. Two devices were removed due to malposition (one perimitral and one trapped in the left ventricular outflow tract strut). One embolized device was removed percutaneously from the left iliac artery. In ongoing follow-up, two patients had continued hemolysis and were referred to surgery. There is an improvement (mean from III to II) in New York Heart Association functional class. There has been one sudden death, three heart failure deaths, and one noncardiac death.
CONCLUSION
The management of paravalvular regurgitation is challenging.
Both surgical and percutaneous options should be
considered. With growing experience with techniques for
percutaneous device closure of paravalvular leaks and
advancements in imaging technology, closure devices and
delivery systems hold the promise of safe and effective percutaneous
closure for patients with severe PVR. The development
of devices specific to this problem will represent an
important step forward for these patients. Structural interventionists
should be familiar with treatment options and a
wide variety of procedural techniques before attempting
percutaneous closures. Close collaboration with imaging
colleagues is essential for diagnosis and successful procedural
execution.
Jeffrey D. Booker, MD, is an interventional cardiology fellow, Mayo School of Graduate Medical Education in Rochester, Minnesota. He has disclosed that he holds no financial interest in any product or manufacturer mentioned herein. Dr. Booker may be reached at (507) 284-2651; booker.jeffrey@mayo.edu.
Charanjit S. Rihal, MD, is Director, Cardiac Catheterization Laboratory, Mayo Clinic in Rochester, MN. He has disclosed that he holds no financial interest in any product or manufacturer mentioned herein. Dr. Rihal may be reached at (507) 255-2440; rihal@mayo.edu.
Advertisement
Advertisement
