Key words
TAVI - aortic valve - aortic valve stenosis - TAVR - CT angiography
Introduction
Transcatheter aortic valve implantation, commonly abbreviated as TAVI, has significantly
gained further ground since the first proof-of-concept procedure performed by Alain
Cribier in 2002 [1]. Introduced as an alternative therapeutic option for patients with severe aortic
valve stenosis (AS) unable to safely undergo surgical valve replacement, its application
has since gained major momentum, supported by large carefully performed multicenter
trials indicating an acceptable safety profile and the non-inferiority of transfemoral
and apical TAVI compared to a surgical procedure in the mentioned population. Today,
it is used worldwide with satisfactory clinical results in an increasing number of
specialized centers.
In this article, we will further explore the evolution of the number of TAVI interventions
and the repercussions on surgical procedures, the role of the radiologist in the pre-procedural
workup and patient management, and the remaining imaging challenges for the future.
The evolution of TAVI procedures vs. surgical interventions
The evolution of TAVI procedures vs. surgical interventions
The scientific validation of the TAVI procedure
Today, the TAVI procedure has become a standard therapeutic option for patients with
severe AS and a high or unacceptable risk for surgical intervention. Confirmation
of the non-inferiority of this procedure in these populations compared with surgical
valve replacement, together with encouraging follow-up studies, has led to a significantly
increase in the worldwide adoption of this technique [2]
[3]. An overview of landmark TAVI studies and their contribution to existing knowledge
is summarized in [Table 1]. Both balloon- and self-expandable transcatheter heart valves (THV) are currently
used, in practice usually represented by the Edwards Lifesciences SAPIEN series and
the Medtronic Evolut range, respectively. The physical properties of these devices
and their imaging characteristics on CT have been extensively described elsewhere
[4].
Table 1
Selected landmark TAVI trials and their contribution to existing knowledge.
|
trial
|
year[*]
|
surgical risk
|
investigated device
|
type
|
acquired knowledge
|
reference publication
|
|
PARTNER 1B
|
2010
|
inoperable
|
SAPIEN
|
BE
|
TAVI, as compared with standard therapy, significantly reduced the rates of death
from any cause, the composite end point of death from any cause or repeat hospitalization,
and cardiac symptoms.
|
N Engl J Med 2010, 363: 1597–1607
|
|
PARTNER 1A
|
2011
|
high
|
SAPIEN
|
BE
|
in high-risk patients with severe aortic stenosis, transcatheter and surgical procedures
for aortic valve replacement were associated with similar rates of survival at 1 year.
|
N Engl J Med 2011, 364: 2187–2198
|
|
CoreValve US HR
|
2014
|
high
|
CoreValve
|
SE
|
in patients with severe aortic stenosis who are at increased surgical risk, TAVR with
a self-expanding transcatheter aortic valve bioprosthesis was associated with a significantly
higher rate of survival at 1 year than surgical aortic valve replacement.
|
N Engl J Med 2014, 370: 1790–1798
|
|
CoreValve US ER
|
2014
|
extreme
|
CoreValve
|
SE
|
TAVR with a self-expanding bioprosthesis was safe and effective in patients with symptomatic
severe aortic stenosis at prohibitive risk for surgical valve replacement.
|
J Am Coll Cardiol 2014, 63: 1972–1981
|
|
CHOICE
|
2013
|
high-Extreme
|
CoreValve/SAPIEN XT
|
BE & SE
|
Among patients with high-risk aortic stenosis undergoing TAVR, the use of a balloon-expandable
valve resulted in a greater rate of device success than use of a self-expandable valve.
1-year follow-up of patients revealed clinical outcomes after transfemoral transcatheter
aortic valve replacement with both balloon- and self-expandable prostheses that were
not statistically significantly different.
|
JAMA 2014 311: 1503–1514
J Am Coll Cardiol 2015, 66: 791–800.
|
|
NOTION
|
2014
|
low
|
CoreValve
|
SE
|
no significant difference between TAVR and SAVR was found for the composite rate of
death from any cause, stroke, or MI after 1 year.
|
J Am Coll Cardiol 2015, 65: 2184–2194
|
|
PARTNER 2A
|
2015
|
intermediate
|
SAPIEN XT
|
SE
|
in intermediate-risk patients, TAVR was similar to surgical aortic valve replacement
with respect to the primary end point of death or disabling stroke.
|
N Engl J Med 2016, 374: 1609–1620
|
|
SURTAVI
|
2018
|
intermediate
|
CoreValve/Evolut R
|
SE
|
TAVR was a noninferior alternative to surgery in patients with severe aortic stenosis
at intermediate surgical risk, with a different pattern of adverse events associated
with each procedure.
|
N Engl J Med 2017, 376: 1321–1331
|
|
PARTNER 3
|
2019
|
low
|
Edwards Lifescience SAPIEN 3
|
BE
|
among patients with severe aortic stenosis who were at low surgical risk, the rate
of the composite of death, stroke, or rehospitalization at 1 year was significantly
lower with TAVR than with surgery.
|
N Engl J Med 2019, 80: 1695–1705
|
|
US Evolut R LR
|
2019
|
low
|
Evolut R, Evolut Pro, CoreValve
|
SE
|
in patients with severe aortic stenosis who were at low surgical risk, TAVR with a
self-expanding supraannular bioprosthesis was not inferior to surgery with respect
to the composite end point of death or disabling stroke at 24 months.
|
N Engl J Med 2019, 380: 1706–1715.
|
BE: balloon-expandable; SE: self-expandable.
* Year indicates primary outcome reached.
The evolving number of surgical vs. transcatheter aortic valve replacement
The evolving number of surgical vs. transcatheter aortic valve replacement
The rise of TAVI in Europe has been well-documented in Germany, a country which has
required registration of all surgical and transcatheter valve replacement procedures
in a nationwide quality assurance program since 2008 [5]
[6]. In 2017, a total of 19 752 TAVI procedures had been performed since 2008, representing
not only a thirty-fold increase since registration began, but also a 50 % increase
since a more recent reference point of 2014 ([Fig. 1]). Conversely, the number of surgical valve replacement (SAVR) procedures (either
isolated or combined with coronary artery bypass graft) declined with 23 % since 2008.
According to this registry, older age is currently the main reason for heart teams
to select TAVI over SAVR, with 95 % of patients over 80 years of age undergoing a
TAVI procedure.
Fig. 1 Evolution of surgical vs. transcatheter aortic valve replacement in Germany. Since
registration began, the number of TAVI procedures has increased thirty-fold, while
the number of surgical valve replacement (SAVR) procedures (either isolated or combined
with coronary artery bypass graft) declined with 23 % over the same period. After
2015, the number of TAVI procedures has surpassed its surgical counterpart. Currently,
TAVI has become the standard of care in patients over 80 years in many German centers.
Data from reference 5. SAVR: surgical aortic valve replacement; CABG: coronary artery
bypass graft.
Other countries also report an increasing dissemination and number of TAVI procedures
[7]
[8]. Nevertheless, differences remain in the management of severe aortic stenosis between
different European centers, with, e. g., Germany having a stronger TAVI preference
compared to the United Kingdom and France [9]. Many registries also point to an evolution of using TAVI in lower-risk patients
with lower Logistic EuroSCOREs [6]
[8]
[10].
The mainstream availability of TAVI comes at an opportune time, since the growing
Western elderly population and the subsequent increase in age-related AS will result
in an increasing workload, representing a particular challenge for clinicians, surgeons,
and radiologists. A recent study calculated, based on epidemiological data and decision-making
studies, that there are about 115 000 and 58 000 annual candidates for TAVI in the
European Union and North America, respectively [11]. This grand total of about 180 000 patients annually could increase up to 270 000
if TAVI indications were to expand to include low-risk patients. However, while recent
trials provide increasing evidence to eventually potentially justify such a move ([Table 1]) [12]
[13]
[14], important questions still remain unresolved, including regarding the long-term
performance of THVs.
The Heart Team and the radiologist
The Heart Team and the radiologist
Determining the optimal treatment course for a frail patient with symptomatic severe
AS and multiple co-morbidities poses a complex problem, which is best served by a
multidisciplinary approach. The creation of “Heart Teams” in specialized centers,
composed of members of different relevant subspecialities, constitutes a further streamlining
and optimization of the process to evaluate all available clinical and imaging information
and select the most appropriate therapy for a particular patient [15]. However, despite the cornerstone role that radiology (most particularly computed
tomography) plays in the pre-procedural assessment of a TAVI candidate, the radiologist
is rarely mentioned by name. Instead, many papers and statements use the more generic
term “imaging specialist”. Therefore, to further promote and reconfirm the important
role that the radiologist has in the decision-making process, the European Society
of Cardiovascular Radiology (ESCR) recently published a consensus statement supporting
the formal inclusion and recognition of radiologists in the composition of a Heart
Team [16].
A brief review of relevant anatomy and required measurements
A brief review of relevant anatomy and required measurements
The analysis of a pre-procedural TAVI CT examination is an extensive process, with
many detailed measurements required at different anatomical levels to consider many
parameters that may influence both peri- and post-procedural safety and final patient
eligibility. A detailed description of all measurements is beyond the scope of this
paper. Nevertheless, some important key points will now be briefly reviewed.
The value of three-dimensional imaging
Many excellent reviews and consensus documents describing the radiological anatomy
of the aortic root and its components have already been published [16]
[17]
[18]. A supplemental movie ([Video 1]) to this paper recapitulates the most important landmark structures using computed
tomography (CT). It further illustrates the varying cross-sectional contour of the
aortic root extending from the sinotubular junction to the so-called aortic annulus.
Combined with the complex three-dimensional morphology of the aortic valve, it is
easy to understand the benefits of using a true 3D imaging modality such as CT to
correctly visualize the annular plane and subsequently accurately obtain the required
measurements. Therefore, while echocardiography was the main imaging modality in the
initial trials validating the clinical use of TAVI, CT has in the meantime become
a prime imaging modality for anatomical assessment of a TAVI candidate, forming an
essential component of any modern annular sizing investigation. Given the fundamentally
different nature of CT and ultrasound (US), respective measurements on these imaging
modalities are not interchangeable or comparable [19]. Finally, while magnetic resonance imaging (MRI) can obtain a 3D acquisition of
the aortic root, it remains technically more challenging, less practical, and therefore
is only used in selected cases.
The aortic annulus
One of the most coveted pieces of information in a pre-procedural assessment is the
dimensions of the aortic annulus, as it forms the anchoring site of both balloon-
and self-expandable THVs. However, the annulus is not a real anatomical structure,
but rather a virtual ring formed by the lowest point (nadir) of the attachment of
the valvular cusps to the aortic wall ([Fig. 2]). Its cross-sectional contour has mostly an elliptic shape in diastole but tends
to become rounder during systole [20]. Given this time-dependent variation, annular measurements must be performed in
the systolic phase as it is associated with the largest annular dimensions. However,
when systolic image quality is poor, other timepoints in the cardiac cycle are acceptable
when delivering more reliable measurements [16]. The most commonly proposed measurements are the annular cross-sectional short-
and long-axis diameter, the annular perimeter, and the annular area ([Fig. 3]).
Fig. 2 Double-oblique in-plane and perpendicular images through the aortic root near and
at the annular level. a The nearly most basal portion of the aortic valve leaflets (arrowheads) is shown.
Note also a nonrelated small calcification (arrow). b In-plane image located at a level even closer to the annular plane, with the leaflet
insertions now just nearly visible (arrowheads). c In-plane image at the annular plane, showing the virtual aortic annulus having an
oval contour on cross-sectional imaging. d The aforementioned calcification can be seen extending to a sub-valvular level (arrow)
in this perpendicular image, below the annular plane (dashed line).
Fig. 3 Double-oblique in-plane image at the annular plane. Many (performed and derived)
measurements are possible. The most commonly obtained measurements in the annular
plane are shown. a Cross-sectional long- and short-axis diameter. b Annular perimeter. c Annular area.
Aortic valve cuspidity
Recently, the cuspidity of the aortic valve has gained increasing attention. The majority
of the normal population has a tricuspid aortic valve. However, a bicuspid aortic
valve (BAV) remains the most common congenital heart abnormality in humans and is
a well-known risk factor for AS, leading to premature valve degeneration in comparably
younger patients [21]. A prevalence of up to 50 % has been reported in surgical aortic valve replacements
due to AS [22].
Historically, the randomized clinical trials validating the use of TAVI have largely
excluded patients with BAVs, with its presence being considered a relative contraindication
in previous guidelines [23]
[24]. The existence of a BAV, while now no longer considered a formal contraindication,
still poses particular challenges ([Table 2]) [12]
[25]
[26]
[27]
[28]. However, recent studies [25]
[29] indicate an improved safety profile and efficacy using newer-generation devices,
providing cumulative evidence for further potential formal inclusion of patients with
BAVs in future TAVI guidelines and indications.
Table 2
Challenges posed by bicuspid aortic valves in TAVI. THV: transcatheter heart valve.
|
challenges of bicuspid aortic valves in TAVI
|
|
constitutional asymmetry of the aortic valve leaflets and severe calcifications, potentially
contributing to more frequent and severe paravalvular leakage due to more complicated
positioning and deployment of the THV
|
|
risk of annular rupture or coronary ostial occlusion
|
|
presence of BAV-associated aortopathy with increased risk of complications, including
aortic dissection
|
|
younger patient population, increasing concerns on long-term durability and performance
of THVs
|
|
increased risk for post-procedural permanent pacemaker need, independently of type
of THV
|
BAV: bicuspid aortic valve.
Given the mentioned procedural and prognostic implications, the cuspidity of the aortic
valve must therefore be clearly mentioned in the radiology report.
Aortic valve leaflet calcifications
Almost invariably, severe AS is associated with extensive leaflet calcifications.
Both calcified and non-calcified leaflet components will be displaced during device
deployment, crushed to variable degrees between the TVH and the aortic root wall ([Fig. 4]). A qualitative description of leaflet calcifications in the radiology report is
required, given the association with complications such as paravalvular leakage, THV
dislodgement, coronary ostia obstruction, annular rupture, calcific embolism, potential
conduction disturbances, and stroke [30]
[31]
[32]. We recommend scoring the amount of valvular calcification (mild to severe), the
distribution (focal vs. diffuse) and the location (leaflet edges, commissures, and
attachment sites). In addition, the absence of prominent leaflet calcifications needs
to be specifically mentioned, as some authors argue that some degree of calcification
may be required for stable anchoring of the THV [31]. Subvalvular calcifications can be scored in the same manner as their supravalvular
counterparts (amount, distribution, and location).
Fig. 4 Double-oblique in-plane and perpendicular images through the aortic sinus after deployment
of a self-expandable Medtronic CoreValve. a Displacement and crushing of both native non-calcified (arrowhead) and calcified
(arrow) leaflet components between the expanding THV and the wall of the aortic sinus
are illustrated in this in-plane image. b To prevent regional obstruction of the deployment of the THV, manufacturers provide
device-specific guidelines regarding the required dimensions of the aortic sinus,
which acts as a reservoir for these displaced components. This function of the aortic
sinus is clearly illustrated here, where the aortic sinus is accommodating displaced
extensive calcifications (arrow), thereby not influencing THV expansion. THV: transcatheter
heart valve.
Several studies have shown a clear correlation between the amount of leaflet calcifications
and the likelihood of having severe AS, hereby also providing additional prognostic
information over traditional risk factors [33]
[34]. Gender differences have also been noted, with women having more severe AS for the
same amount of valvular calcium load compared to men [35]. While the exact pathophysiology for this discrepancy remains to be fully understood,
some investigators point to a more prominent fraction of fibrosis in stenotic aortic
valves in women [36]. Combining these new insights with the clear need for diagnostic aid in cases of
the so-called low-flow/low-gradient AS (in which echocardiography is unable to deliver
a conclusive diagnosis of severe AS), the quantification of aortic valve leaflet calcifications
with CT has now become an established diagnostic tool [15]. This quantification is executed using the same Agatston score method as for calcifications
of the coronary arteries. However, this quantification is not to be routinely performed
and should still be reserved for the mentioned low-flow/low-gradient AS with inconclusive
echocardiography. Finally, quantification of aortic leaflet calcifications may in
practice be more difficult to perform than expected, with, e. g., a sometimes unclear
distinction between calcifications of the aortic valve leaflets and adjacent mitral
annulus calcifications.
Access route patency
Besides the aortic root anatomy, the patency of the transcatheter access route is
a prerequisite for safe delivery of the THV. To increase the number of patients eligible
for a TAVI procedure, the number of possible access routes has been steadily expanded
over the years. An overview of the currently available access routes can be found
in [Table 3]. For the current Medtronic Evolut R and Edwards Lifesience SAPIEN 3 series, a minimal
luminal diameter of 5.5 mm is needed for their largest THV (29 mm SAPIEN 3 and 34 mm
Evolut R). All other smaller sizes require a minimum luminal diameter of 5 mm for
safe passage of the THV. We suggest using a uniform scan and reconstruction protocol
and routinely evaluating the most common access pathways.
Table 3
Overview of currently available access routes for TAVI. The transapical approach is
currently reserved for balloon-expandable THV.
|
access sites
|
|
transfemoral
|
|
transapical
|
|
transsubclavian/brachiocephalic artery
|
|
transcarotid
|
|
transaortic
|
|
transcaval
|
|
transseptal
|
THV: transcatheter heart valve.
Differences exist between the types and models of THVs and the required characteristics
of the different access sites. As new devices are continuously being developed and
introduced, access route requirements (like minimal luminal diameter needed for unobstructive
passage) change accordingly, always with the aim of lowering the anatomical threshold
for procedural eligibility. The transapical approach is currently only available for
balloon-expandable valves.
Despite the increasing array of options, the transfemoral access remains the preferred
route whenever possible according to international guidelines [15]
[37]. Risk factors for complications include circumferential atherosclerosis, occlusive
vascular disease, small native vessel diameter, dissections, presence of stents, and
prominent vessel tortuosity. These risk factors, however, also apply to all access
sites. When transfemoral access is not feasible (15–25 %), other access sites can
be considered, with the final choice for an alternate access depending on anatomical
feasibility and local expertise [38]
[39].
Valve-in-valve
The valve-in-valve technique is currently a niche application of TAVI used to treat
degenerated surgical bioprosthetic aortic valves [40]
[41]. During this procedure, a transcatheter valve is deployed within the in situ surgical
aortic valve prosthesis, effectively replacing its function.
Given the different circumstances, the required pre-procedural CT measurements are
different than in the case of native aortic valves. The presence of the surgical valve
will determine the maximum size of the THV that can be implanted. While this dimension
can be obtained from the surgical report, it can also be derived using CT [42]. An important consideration in valve-in-valve procedures is the increased risk for
coronary obstruction, which can be decreased using both procedural modifications and
CT-based pre-procedural simulation techniques [42]
[43]
[44].
While valve-in-valve is not widely performed today, its role may increase in the future
as it could also be used to replace a failing transcatheter aortic valve. This may
become more relevant as there is a tendency to treat younger patients with THVs in
which case life expectancy may exceed valve durability. Recent studies show the potential
for the replacement of a THV with another THV with similar procedural safety or mortality
compared with the current valve-in-valve procedure to replace a bioprosthetic surgical
valve [45].
How to scan a TAVI candidate
How to scan a TAVI candidate
Many reviews have explained in detail the technical requisites of a pre-procedural
CT and MRI examination [17]
[18]
[46]. Recently, the ESCR published an open-access consensus statement with vendor-specific
CT and MRI protocols, which can be used for further reference [16]. In this review, we will concentrate on three specific issues regarding the CT examination:
the need for pre-medicating a TAVI candidate prior to the CT examination, the influence
of CT contrast on renal function in this population, and the delivered radiation exposure.
Finally, the use of MRI as opposed to CT will be briefly discussed.
Pre-medication
Given the inherent motion of the heart, the use of pre-medication during the preparation
of a patient for a routine cardiac CT examination is not uncommon. Here, administration
of medication (either orally or intravenously) has two common goals: lowering the
heart rate to achieve optimal diagnostic motion-free image quality and obtaining a
temporal dilatation of the coronary arteries for better visualization and stenosis
detection [47]
[48].
However, the current TAVI population has quite different characteristics. Firstly,
it is composed of much elderly individuals with a higher-than-average frailty and
multiple co-morbidities. Therefore, as they may react differently to the usual drugs,
the use of medication with the intent of increasing CT image quality must be approached
with caution and in concordance with the referring physician. Some centers and societies
therefore opt to routinely abstain from any routine pre-procedural CT medication in
TAVI candidates [16]. Furthermore, as the primary target is not evaluation of the coronary arteries but
of the aortic root and its components, less strict heart rate control may be tolerated,
especially when high-end CT equipment is available [49]
[50].
In conclusion, the use of pre-medication to improve image quality is not routinely
recommended.
Iodinated contrast volume
Most commonly, a TAVI CT examination is executed to obtain information on both the
dimensions and characteristics of the aortic root (performed using ECG-gating) as
well on the patency of the different possible access routes [17]. This combined approach, often performed during a single examination, inherently
implies a large anatomical scan range and matching dose of contrast volume. However,
many patients already have depressed renal function, potentially further compromised
by other required examinations preceding the TAVI procedure (e. g., in many instances
including a conventional angiography of the coronary arteries in the absence of recent
information regarding their status). Therefore, any measure that can be implemented
to reduce the volume of iodinated and potentially nephrotoxic contrast product during
the CT examination must be exploited to its fullest capacity. Many investigators have
addressed this issue, reporting that the speed and performance of the latest generation
of CT scanners can achieve high diagnostic results while significantly limiting the
required amount of contrast volume [49]
[51]. Some authors have also reported encouraging results on the concomitant evaluation
of the coronary arteries during the same CT examination [50]
[52]. While it can be assumed that the success of such an approach is inherently linked
to the experience of the center performing the examination and the available equipment,
if confirmed in future studies it further represents an additional example of the
increasing value of CT and its ability to at least partially replace other examinations
and as such their contrast use.
In summary, an effort should always be made to reduce the contrast volume load according
to the performance of the available equipment.
Radiation exposure
While radiation exposure must always remain a point of focus, one can pragmatically
defend the notion that given the very advanced age of a typical TAVI candidate and
the dismal short-term prognosis of an untreated symptomatic severe AS, aggressive
reduction of radiation exposure is currently not a primary concern in this population.
However, as the transcatheter approach for treating valvular heart disease expands
to other valves, and within the aortic valve to other indications and population groups,
it can be expected that the mean patient age will further decrease [15]
[53]
[54]
[55]. Therefore, with an increased associated life expectancy, together with a not yet
exactly defined potential role of CT in the management of manifestations such as subclinical
valve leaflet thrombosis, radiation exposure and the efforts to reduce it will only
gain in importance. Nevertheless, technology advances and the introduction of novel
noise reduction techniques can further help to maintain the associated radiation dose
within acceptable limits [56]
[57].
MR imaging
Given its fundamentally different nature, CT is much better suited for pre-TAVI anatomical
evaluation than MRI. MRI can, however, be of value in certain circumstances, since
it is potentially able to perform the annular planimetry and obtain the different
necessary measurements with comparable accuracy to that of CT, even without intravenous
contrast administration [58]
[59]. As such, its use may be warranted in patients who have formal contraindications
for iodinated contrast.
However, the use of MRI is far less widespread than CT for annular planimetry, despite
the absence of radiation exposure and the use of a far less potentially toxic contrast
medium. The reasons include a technically more demanding and longer examination, the
commonly longer waiting lists compared with CT, as well as the less detailed evaluation
of calcifications (which are an important factor to consider for procedural safety),
among other things. Some authors, therefore, propose performing an MRI examination
for obtaining aortic root measurements, followed by a non-enhanced CT examination
for evaluation of calcifications.
A more interesting role for MRI may the pre- and post-procedural evaluation of cardiac
function, especially left ventricular remodeling, the pattern and magnitude of which
are influenced by many factors [60]. MRI may also provide additional prognostic information, as delayed myocardial enhancement
indicating fibrosis has been shown to be an independent predictor of mortality after
both surgical and transcatheter aortic valve replacement [61]. Consequently, using MRI to assess the presence of both interstitial and replacement
fibrosis may provide better insight into the post-procedural prognosis independent
of other risk factors [62]
[63]
[64]
[65].
TAVI outcomes: early and late complications after TAVI
TAVI outcomes: early and late complications after TAVI
Despite the obvious increase in and clinical success of TAVI procedures in the last
decade, different challenges remain for the future. Some of these potential complications
arise early in the post-operative period, while other manifest only years afterwards.
In chronological order of appearance, we will discuss paravalvular aortic regurgitation,
prosthetic heart valve thrombosis and endocarditis, and finally structural valve dysfunction.
Paravalvular aortic regurgitation
Paravalvular aortic regurgitation (PAR) is defined as the postprocedural regurgitant
leakage of blood around the attachment sites of the THV. While the causes are multifactorial,
it is at least in part secondary to suboptimal patient-prothesis size matching, as
an undersized or incorrectly positioned THV can lead to incomplete sealing at the
attachment sites.
Some series indicate that about one in nine patients develops moderate to severe PAR
after the procedure [66]. The clinical significance of PAR must not be underestimated, as it is a known and
independent factor for late all-cause mortality [67]
[68]. Therefore, as even moderate PAR has significant impact on prognosis, significant
efforts have been devoted to optimizing sizing algorithms using different methods.
It is in this respect that the integration of CT-based annular sizing in the pre-procedural
workup has subsequently evolved to one of the most important cornerstones to achieve
optimal procedural and clinical success.
The main imaging tool for post-procedural and further life-long follow-up of TAVI
patients remains echocardiography, delivering both anatomical and functional information
in one widely available and relatively cheap imaging examination [15]. As such, it is the primary indicated method for the detection and follow-up of
PAR. Despite the clear advantages that CT has in the pre-procedural anatomical assessment,
it remains a morphology-based imaging modality, and therefore is not indicated for
functional THV evaluation. Even in cases where CT produces very suggestive imaging
findings for PAR, extreme caution must be exerted to not extrapolate these findings
to functional implications. Displaced valve leaflet calcifications, while potentially
inhibiting the regional expansion of a THV and therefore compromising its function,
can also have an additional positive sealing function. The relation between leaflet
calcifications and the final functional result after THV deployment is therefore more
complex than initially though.
Conversely, CT can have a role as a secondary imaging tool in cases when PAR has been
established based on clinical and echocardiography findings, although without a clear
etiology. In such circumstances, CT can easily depict an incorrectly positioned or
migrated THV, therefore delivering important information for further therapeutic interventions
in selected cases ([Fig. 5]). Consequently, in our institution we don’t routinely perform post-TAVI CT examinations,
with its use being determined on a case-by-case basis.
Fig. 5 THV dislocation in a patient with persistent elevated but further unexplained gradients
on echocardiography after TAVI. a An angulated and slightly displaced Medtronic CoreValve unable to displace all native
aortic valve material is shown in this perpendicular image. b A partially calcified and thickened native aortic valve leaflet protruding into the
inflow portion of the THV (asterisk) remains in place in this axial image. THV: transcatheter
heart valve.
MRI, which like echocardiography can deliver both morphological and functional information,
is by nature better suited than CT to assess PAR severity, as reported by several
investigators [69]
[70]
[71]. However, comparison of MRI and echocardiography results is not straightforward,
as several studies have used different reference standards and PAR severity grading
definitions, with MRI studies reporting different rates of severe AS [72]. Therefore, similarly to the sizing comparison between CT and echocardiography,
comparison between MRI and echocardiography results for PAR must be interpreted with
the necessary caution. Regardless of these issues, investigators have shown that MRI-derived
aortic regurgitation values are correlated with clinical outcomes, with MRI-detected
moderate to severe PAR leading to worse outcomes at the 24-month follow-up [73], thus indicating a role for MRI in patients next to echocardiography in selected
patients. Whether new technologies like analysis of transvalvular and ascending aorta
4D flow patterns will lead to clinically relevant consequences is currently still
under investigation [74]
[75]. Finally, the practical use of MRI in the postoperative period may also be influenced
by an existing and post-procedural raised prevalence of implanted cardiovascular electronic
devices in TAVI patients ([Fig. 6]) [72].
Fig. 6 CT images in-plane and perpendicular to a THV (Lotus) in the aortic position in an
81-year-old patient that was admitted with fever and blood cultures positive for streptococcus.
a Irregular thickening of the valve leaflets in this in-plane image can be seen (arrow).
b In this in-plane image on a subvalvular level, hypodense structures (arrow) that
are compatible with vegetations are depicted. c The perpendicular image clearly shows the thickened leaflets (arrows). THV: transcatheter
heart valve.
PHV endocarditis
Infective endocarditis is a rare but severe form of valvular heart disease, characterized
by an inflamed endocardium and valve leading to the development of vegetations, mycotic
aneurysms, and septic emboli. Its epidemiological profile has evolved in recent years,
influenced by the increasing use of invasive procedures (including the use of heart
valve protheses) and hence the risk for bacteremia [76]
[77]. Specifically, prosthetic heart valve endocarditis (PVE) is a recognized life-threatening
complication, affecting up to 5 % of patients annually after valve implantation, with
a reported 1-year mortality of up to 50 % [78]. Particularly in this subgroup, diagnosis is often difficult, with a highly variable
clinical history and presentation [76]. While echocardiography and blood cultures remain a cornerstone of diagnosis according
to the modified Duke criteria of 2000, they have a reported lower sensitivity and
specificity for the diagnosis of PVE [79]. This is unfortunate, as a timely and definitive diagnosis is necessary to be able
to reach major clinical decisions like potential reoperation with high confidence.
It is in this respect that clinical guidelines have been recently updated to include
relevant contributions made by contemporary imaging modalities like CT ([Fig. 6]) and 18F-fluorodeoxyglucose (FDG) positron-emission tomography (PET)/CT [37]
[76]
[80]
[81]
[82].
While a more detailed discussion is beyond the scope of this paper, it is important
to realize that even with these modern evaluation tools, the diagnosis of PVE remains
a challenging task, best undertaken in specialized centers with dedicated “Endocarditis
Teams” composed of specialists with different medical specialties.
Subclinical leaflet thrombosis
While the use of CT was not integrated in the pre-procedural workflow of the initial
trials which led to the validation of TAVI clinical use, it soon became a standard
imaging modality as its advantages became clear. This integration led to some unanticipated
findings, like the prevalence of reduced leaflet motion (RLM, sometimes also indicated
as hypoattenuation-affecting motion or HAM) and subclinical valve thrombosis in both
surgical and THV patients with unchanged mean gradients on echocardiography [83]
[84]. This so-called hypoattenuated leaflet thickening (HALT) appeared in various degrees,
with no evident clinical correlation and unremarkable echocardiography findings. While
RLM appeared in some series in both surgical and bioprosthetic aortic valves with
a total of 11.9 %, it was more common in TAVI patients (13.4 %) versus surgical valves
(3.6 %), with a more severe motion reduction and pronounced leaflet thickening in
the TAVI population [84]. Additional studies also revealed that the prevalence of RLM and HALT differed between
types of valves, with, e. g., the Perceval sutureless valve reported to be more affected
in earlier studies ([Fig. 7]) [85].
Fig. 7 Subclinical hypoattenuating leaflet thrombosis in a patient after receiving a sutureless
Perceval PHV, with no transvalvular gradient on echocardiography. a In-plane CT image shows clear thickening of the PHV leaflets (asterisk). b In-plane image showing a semi-circumferential non-continuous low-density structure
against the luminal side of the PHV, representing a thrombus (arrowheads). c Perpendicular image showing thickened leaflets (asterisk) and peripheral thrombus
(arrowheads). The clinical significance of such findings cannot be determined with
imaging alone and must be cautiously correlated with all other available imaging and
clinical data. PHV: prosthetic heart valve.
Evidently, questions were raised as to how these subclinical imaging findings translate
to a need for further optimization of medical anticoagulant therapy, and regarding
their impact on prognosis and incidence of thrombo-embolic events like stroke. A meta-analysis
in 2018 indicated that, while there is an overall incidence of HALT with or without
RLM of about 15 %, there was no significant association with the presence of stroke,
transient ischemic attack (TIA), or the combined stroke/TIA endpoint [86]. However, a more recent meta-analysis did find an increased stroke risk [87]. Most importantly, both studies indicated a need for more research. This need for
more insight is further illustrated by the fact that investigators have noted that
subclinical leaflet thrombosis can in some patients regress without anticoagulation
[88]. Therefore, this is currently a field of intense ongoing investigation, with recent
studies further contributing to the debate and our understanding on how to use and
optimize antithrombotic therapy in post-TAVI patients [89]
[90]. Regarding CT imaging, a systematic methodology for the evaluation of subclinical
leaflet thrombosis has been suggested to reach standardization of reporting [91].
Structural valve degeneration
The prevalence of RLM and HALT in both surgical and especially THV patients also raises
questions regarding the longevity of these devices. Currently, the vast majority of
patients who receive a THV for severe aortic valve stenosis are of advanced to very
advanced age. However, one can reasonably expect that the mean age of THV recipients
will drop as current and future trials continue to confirm the at least non-inferiority
and acceptable safety profile of a transcatheter versus a surgical approach to replace
a diseased aortic valve in other risk populations, leading to expanding indications
and a lower risk profile of TAVI candidates.
It is therefore crucial to fully understand all contributing factors to structural
valve degeneration (SVD). To facilitate comparison across centers and trials, a common
definition of SVD has been proposed in a consensus document supported by different
societies [92].
Data from the FRANCE-2 registry have indicated that while the all-cause mortality
was about 61 %, the majority of cardiovascular events occurred in the first month
after valve implantation, with a low rate of clinical events and a low level of SVD
after 1 year [93]. A multicentric French study also reported low rates of SVD (10.8 %) and bioprosthetic
valve failure (1.9 %) in a 7-year follow-up study [94]. However, the authors also noted that long-term assessment of SVD is limited by
the poor survival of the investigated population, with only a reported 19 % survival
after 7 years.
On the etiology of SVD, histological analysis of explanted specimens revealed that
thrombus was found in all explanted valves, both on the aortic and ventricular sides
[95]. Leaflet thickening correlated with the duration of THV implantation, with a progression
independent of cardiovascular risk factors and anticoagulation therapy. Based on their
findings, the authors postulated a sequence of thrombus, fibrosis, and calcification
as a pathway to SVD.
This and other discoveries may help us not only detect the presence of SVD in its
earliest stages, but also provide anchor points for targeted therapies on different
components of the pathophysiological process. Very promising studies concerning early
detection of SVD have been concentrating on the use of 18F-Fluoride PET-CTA, providing
striking signs of SVD before hemodynamic deterioration [96].
In conclusion, it can be stated that there is no dominant imaging modality that can
cover all aspects of post-procedural TAVI surveillance, but that different modalities
cover different aspects at different time points after the procedure. As these imaging
modalities are spread across different medical specialties (radiology, cardiology,
nuclear medicine), it is obvious that the often-mentioned but not always properly
implemented principle of multidisciplinary cooperation will provide the best platform
for patient care, which is best executed within the Heart Valve team.
Choice of imaging modality after TAVI
Despite all of the mentioned advances in the use of CT, MRI and nuclear imaging, echocardiography
remains the main imaging modality for the follow-up of a patient after TAVI. Its ability
to provide morphological and functional information in a single examination, without
the use of intravenous contrast material or radiation exposure, is invaluable for
routine clinical follow-up.
At this moment, there are no guidelines regarding the use of CT or MR imaging after
the procedure on a routine basis. Their application is reserved for selected cases
when echocardiography is inconclusive. As such, routine use of these imaging modalities
in uncomplicated cases is not recommended. Nevertheless, CT is the current imaging
modality of choice over echocardiography to detect subclinical leaflet thrombosis,
an important entity with a clinical impact that is not yet clear.
Conclusion
Once again, radiologists find themselves at the intersection of different medical
specialties all striving for optimal use and implementation of a transformative technique
like transcatheter replacement of a diseased aortic valve. It represents an opportunity
to further promote radiologists as clinically oriented and patient-centered individuals,
providing essential information required for procedural success and cooperating with
other colleagues for the well-being of the patient. Therefore, integration of a radiologist
within the Heart Valve team of any center offering TAVI is essential for the optimization
of patient care, and for the standing of radiologists within the medical profession.
It is, as such, the duty of every radiologist to be familiar with this procedure,
the required information to be provided, and the indications and limitations for post-procedural
surveillance with CT and MRI.
Video 1 Brief overview of relevant anatomical landmarks on CT in the pre-procedural assessment.
