Introduction
Aortic valve stenosis (AVS) is the single most common adult heart valvular disease
affecting over 5 % of those older than 65 years old. The relative frequency of AVS
etiologies vary geographically, with rheumatic disease being the predominant cause
in low-income countries, whereas degenerative fibrocalcific disease is dominant in
North America and Europe [1]
[2]
[3]
[4]
[5]. In parallel to this development, there have been major advances in cardiac surgery
and percutaneous valvular intervention thus allowing the possibility of successful
intervention even in elderly, multimorbid patients [6]
[7]
[8]. However, despite successful intervention, many patients have worse outcomes compared
to age- and sex-matched peers. Chronic biomechanical stress results in myocardial
hypertrophy and progressive fibrosis due to the triggering of pro-inflammatory and
fibrotic pathways leading to worsening diastolic and eventually systolic function
[9]
[10]
[11]
[12].
Currently, myocardial remodeling due to AVS is a secondary indication when considering
patients for intervention. Hemodynamic severity of the valve lesion and the presence
of symptoms are the primary indications [13]. Whereas transthoracic echocardiogram (TTE) is the best imaging modality for the
hemodynamic assessment of valve disease, cardiac magnetic resonance (CMR) additionally
offers tissue characterization of the myocardium including the detection of focal
and diffuse fibrosis [14]. However, several factors limit its widespread application in clinical practice,
including access, cost, claustrophobia, and local expertise. In contrast, cardiac
computed tomography (CCT) has become an essential modality mostly for the planning
of structural valve intervention with recent advances that also include techniques
allowing the evaluation of myocardial function and tissue characterization.
The aim of this review article is to provide a comprehensive review over the current
role of CCT in myocardium evaluation in patients with AVS.
Aortic valve stenosis
AVS is defined as aortic valve thickening, usually with at least mild calcification
and presence of antegrade velocity across an abnormal valve at least 2 m/s. Signs
and symptoms are determined by valve anatomy, hemodynamics, and maladaptive cardiac
remodeling.
The role of echocardiography
Echocardiography has a pivotal role in imaging assessment of patients with suspected
AVS and is currently used for confirming the diagnosis, grading severity, assessing
valve calcification, left ventricular (LV) function, and remodeling, detecting other
valve disease or aortic pathology, and providing prognostic information [15]
[16]. Moreover, it also provides key information analyzing the feasibility of potential
invasive interventions and the likelihood of having a successful approach.
Current guidelines rely on three key parameters for severity assessment of AVS: mean
pressure gradient, peak transvalvular velocity, and valve area. However, due to the
frequent display of discordant results, additional parameters need to be taken into
account (most of them echocardiographic) such as: LV ejection fraction, stroke volume,
Doppler velocity index, LV hypertrophy, flow conditions, the adequacy of blood pressure
control, aortic valve (AV) calcium score, and planimetry [17].
LV systolic function is a major prognostic determinant, and it has been traditionally
assessed by LV ejection fraction (LVEF) quantification. However, this method has significant
limitations, in particular the tracking of early functional changes in the remodeled
LV where hypertrophy initially increases the LVEF at the expense of stroke volume.
An alternative, the assessment of global longitudinal strain (GLS), which is the ratio
or percentage of change in length over the original length, offers a stronger correlation
with adverse remodeling and adverse cardiovascular events, even in patients with preserved
LVEF [18], but has not yet been integrated in the clinical management pathway. In addition,
strain imaging can also unveil features of concomitant dual pathologies, such as amyloid
protein deposition [12].
Transformational role of cardiac CT in aortic valve stenosis
CCT is a fundamental tool in VHD management. The strong correlation between calcium
burden and aortic valve stenosis severity has resulted in aortic valve (AoV) calcium
scores on non-contrast CCTs (with sex-specific cut-offs) being implemented in international
guidelines. Particularly in patients with classic low-flow low-gradient AS with inconclusive
low-dose dobutamine stress echocardiography and those with paradoxical low-flow low-gradient
AS, AoV calcium scoring is recommended [13]. Combining this with angiographic evaluation allows not only precise geometric assessment
of valve area using multiplanar reconstruction software [19]
[20] but also newer approaches quantifying the fibrotic volume of the valve, which promises
to be a more accurate measure of AS severity [21].
Furthermore, cardiac CT allows assessment of valve morphology, evaluation within the
valve and root (i. e. coronary ostium height, annulus and leaflet dimensions, membranous
septum length, calcium distribution within the valve), and appraisal of aortopathy
and coronary artery disease, and provides unique information for procedural planning
of a structural intervention (e. g., femoral or alternative access routes) as depicted
in [Fig. 1]
[13]
[21]. Hybrid assessment with CT for the LVOT and echocardiography for flow may also optimize
calculation of the AoV area by the continuity equation [22].
Fig. 1 Comprehensive assessment of AS patient with computed CT. Caption: 83 y/o male with
syncope and severe aortic stenosis by TTE (Vmáx: 4 m/s, mean gradient 43 mmHg). Patients
underwent cardiac CT for an “all-in-one” morphological and functional assessment of
valve disease. A. B. endocardial borders and LV volume quantification. C. LV volumes
throughout the cardiac cycle; D: En face of tricuspid aortic valve E. Peripheral vascular
access evaluation for TAVI planning.
A disease of the valve and the myocardium
In AVS, patients’ symptoms and outcome are determined not only by the severity of
valve stenosis but also by the myocardial response to the excessive afterload [18]
[19]
[20]
[21]
[22]
[23]. A complex interplay of cellular (i. e., hypertrophy, cell death) and extra-cellular
(i. e., microvascular ischemia, increased collagen synthesis and deposition) changes
occurs simultaneously and culminates in myocardial fibrosis (MF) [24]. Histological assessment of this pro-fibrotic process has revealed a complex morphology
and distribution with three main patterns: thickened endocardium with a massive fibrotic
layer; a gradient from the subendocardium to the mid-myocardium with abundant microscopic
scars; and diffuse interstitial fibrosis (see [Fig. 2]) [14]. The fibrotic gradient appears to be related to the capillary rarefaction towards
the endocardial surface, responsible for microvascular ischemia, cell loss, and consequent
replacement fibrosis [25]
[26]. Furthermore, microscopic scars occur due to reactive responses of the mechanically
stressed cardiomyocytes to chronic pressure overload, triggering fibroblasts for collagen
deposition [23]
[24]
[25]
[26]
[27]
[28].
Fig. 2 Physiological and histological changes expected in severe AS patients. The chronic
increased afterload in aortic stenosis elicits complex cellular and extracellular
changes that progress towards myocardial fibrosis and impaired function. The excessive
collagen deposition typically follows a gradient from the subendocardium to the mid-myocardium.
Assessment of adverse myocardial remodeling with CMR
Although CMR is not used routinely for clinical evaluation of aortic valve severity
in AS, it can provide reliable measurements of valvular severity by assessing peak
velocity, aortic valve area, and flow. Being the gold standard for functional and
volumetric assessment, CMR also offers accurate assessment of the remodeled heart
in addition to advanced tissue characterization. CMR can qualitatively and quantitatively
assess the complex myocardial fibrotic process secondary to chronic pressure overload,
namely focal replacement and diffuse reactive fibrosis. Diffuse reactive fibrosis,
appears to be an early response to chronic pressure afterload and results from the
extracellular matrix (ECM) expansion and regresses after aortic valve replacement
(AVR) accompanied by structural, functional, and biomarker improvement. Focal fibrosis
may be captured by late gadolinium enhancement (LGE), which highlights differences
between normal and abnormal myocardium, but only identifies the tip of the iceberg
(as the remote myocardium is fibrotic as well). Focal replacement fibrosis represents
the irreversible loss of cardiomyocytes (i. e., scar). Therefore, a more advanced
state can be identified by LGE and it persists after AVR [28]
[29]
[30]
[31]
[32]
[33]
[34]. In order to capture diffuse fibrosis, absolute quantification of the myocardial
signal is obtained by native T1 mapping (which captures the signal from both the cell
and the ECM) and the T1-derived extracellular volume fraction (ECV%). Both have been
validated against histology [28]. Both LGE and ECV are independent predictors of adverse outcome after surgical and
transcatheter intervention [35].
Emerging applications of cardiac CT
In the last decade, the utility of cardiac CT has broadened exponentially with promising
new techniques that can complement clinical information to guide the current clinical
pathway of patients with AS. Beyond anatomical pre-procedural assessment and evaluation
of the coronaries, cardiac CT also allows accurate functional, volumetric assessment
of the ventricle and the potential for myocardial tissue characterization.
Functional assessment
The isotropic sub-millimetric spatial resolution and the good contrast between the
ventricular lumen and the myocardium make CT well suited to obtain valuable information
on ventricular function, regional wall motion, and LV mass comparable to CMR [36]. Although this requires data acquisition across the cardiac cycle, the resultant
radiation penalty can be minimized by using dose modulation techniques. Meta-analysis
of 27 studies comparing transthoracic echocardiogram and CMR (15 vs. 12 studies) with
64-slice (or higher) CCT showed no difference between modalities on ejection fraction
quantification [37]. Recently, in a small-comparative study in patients following transcatheter aortic
valve replacement (TAVR), Szilveszter et al. found a good correlation between GLS
by echocardiography of the LV and the left atrium (LA) with GLS by 256-slice CT (r = 0.78,
p < 0.05 and r = 0.87, p < 0.001, respectively) [38]. Considering the growing evidence base regarding TTE and GLS as an early surrogate
of worse prognosis, CCT (if proven widely applicable, robust, and standardized) emerges
as an attractive tool with the ability to complement anatomical and functional assessment.
However, larger volume multicentric studies are currently needed to confirm its applicability
with respect to prognosis.
Delayed enhancement by cardiac CT
Although noninvasive myocardial tissue characterization was once exclusively assessed
by CMR, CCT has recently emerged as an attractive alternative, especially for myocardial
fibrosis. Both gadolinium and iodine-based contrast agents are extracellular, extravascular
contrast agents with a similar volume of distribution and contrast kinetics, thus
allowing comparable myocardial characterization with CMR and CT not only on delayed
enhancement (DE) imaging but also in first-pass perfusion [39]
[40]
[41]. Furthermore, the linear relationship between iodine and tissue signal is a more
straightforward (linear) relationship than the effect of gadolinium on protons (including
effects of fast intracellular water exchange) [42].
In ischemic cardiomyopathy, the volume of distribution of contrast agent is increased
due to ruptured cell membranes of the necrotic myocytes in the acute stage, whereas
in the chronic phase, iodine accumulation will also be increased in the infarcted
segments due to the replacement of necrotic cells by collagen-rich scar tissue [40]. Compared to CMR, this modality offers excellent agreement for the identification
of infarct region and size with reported sensitivities and specificities of 98 % and
94 %, respectively [43]. The hyper-enhanced areas on delayed image acquisitions are not exclusive to ischemic
cardiomyopathy. Indeed, DE on CT has already been shown to be diagnostically useful
for different pathologies such as sarcoidosis, hypertrophic cardiomyopathy, and amyloidosis
[44]
[45]
[46]. However, this modality cannot be used to assess the early stages of maladaptive
remodeling secondary to AVS characterized by diffuse fibrotic process, making visual
assessment difficult to depict disperse increase of extracellular volume. Further
acquisitions, namely assessment of extracellular volume fraction, emerged to rectify
these inherent qualitative and quantitative limitations of DE sequences,
Extracellular volume fraction based on cardiac CT
Extracellular volume quantification based on CT (CTECV) requires a baseline and a
delayed post-contrast scan acquired at least 3 minutes after contrast injection [47]
[48]. At the time of the delayed scan, a condition of pseudo-equilibrium is established
between contrast in the blood pool and in the myocardium, which is a prerequisite
for accurate ECV quantification. Currently, there are 2 established distinct methods
to calculate ECV, determined by the scanner detector: single- or dual-energy. The
single-energy (SE) approach determines contrast media distribution and hence ECV based
on the change of CT attenuation between the pre-contrast and LE images. The formula
used for ECV calculation is as follows:
Dual-energy detector scans enable the reconstruction of iodine maps from LE scans
for calculation of the ECV using the following formula, without the need of a baseline
scan:
Post-acquisition, ECV can be calculated based on a region of interest (ROI), or three-dimensional
(3D) analysis can be performed for the whole heart by matching a heart model (blood
pool) generated from the respective coronary CTA data. The LV heart model, automatically
determined from the coronary CTA data, is overlaid onto the respective ECV volume
data. Results can be displayed and numerically exported using standard 17-segment
polar maps as depicted in [Fig. 3].
Fig. 3 Increased extracellular volume in three patients with severe AS. Caption: A. Normal
ECV; B. Mildly increased ECV due to diffuse fibrosis; C. Highly increased ECV due
to dual pathology of AS and cardiac amyloidosis.
In clinical use, extracellular volume fraction based on CCT quantification based on
cardiac CT has been significantly correlated with adverse outcomes in severe AS patients.
Scully et al. prospectively enrolled 132 elderly patients exclusively with severe
AS undergoing TAVR and demonstrated that ECV by CT was strongly associated with all-cause
mortality over a median follow-up of 28 months [Hazard Ratio (HR):1.246, p = 0.004],
with a doubling of the mortality risk for each 2 % increase in ECV [49]. These findings were further supported in a retrospectively enrolled cohort of 95
consecutive patients with severe AS undergoing TAVR, where ECV based on CT was the
single independent predictor on multivariable Cox regression analysis (HR: 1.25; p < 0.001)
for the composite endpoint of all-cause mortality and heart failure hospitalization
[50]. Furthermore, Tamarappo et al. demonstrated the value of CT-derived ECV in 150 patients
with low-flow low-gradient AS that underwent TAVR (HR:1.04, p = 0.01) with respect
to predicting the composite endpoint of all-cause mortality and heart failure hospitalization
over a median follow-up of 13.9 months [51].
Patients with severe AS often have coexistent cardiac amyloidosis (CA) with a reported
prevalence of 1 in every 7 elderly patients undergoing TAVR. The hallmark deposition
of misfolded proteins within the myocardium further increases the ECV which can also
be readily identified by cardiac CT [52]
[53]
[54]. The presence of a dual pathology indicates a worse prognosis of heart failure.
Therefore, early identification is important since there are novel therapeutic options
that are capable of improving outcome, especially at early stages [55]
[56]
[57].
It is estimated that CMR is not suitable in 10 % of patients, mainly due to claustrophobia
and artifacts. The wider accessibility of CT-derived ECV, in addition to faster acquisitions
(currently completed in 3 minutes), high-resolution 3 D ECV volumes, and the fact
that this imaging modality already takes part in the current management pathway in
a considerable proportion of patients with severe aortic valve disease, makes this
technique an attractive alternative to CMR for additional information on myocardial
assessment in patients with valvular heart disease [58].
Challenges to implementation
Wider application of cardiac CT is still limited by the inherent radiation exposure
that can reach up to 5 and 8 mSv for volumetric assessment (dual- and single-source
scanner, respectively) and an additional 2.3 mSv for the ECV quantification [49]
[59]. However, most severe aortic stenosis patients are older individuals whose long-term
toxic exposure to radiation might be less of an issue. In addition, cardiac CT is
already in the clinical pathway for most of these patients. While functional assessment
would only be a sporadically useful tool for those with prior unclear LVEF evaluation,
routine use of CTECV could potentially be of great value. However, this remains speculative
while cost-effectiveness studies are still lacking.
As described above, CTECV is conceptionally easy, straight-forward to implement, does
not require additional contrast administration at a cost of limited additional radiation.
The current challenges to wider clinical implementation are analogous to ECV based
on CMR field and are three-fold. First, the evidence base for CTECV needs to grow
with further protocol and post-processing refinements and standardization, cross-vendor
validation, wider application across health and disease, multi-center outcome cohort
validation, and use in clinical trials. Second, CT hardware and software vendors are
currently in various stages of development of CTECV products, and wider access to
post-processing software is essential for broader use. Finally, the cardiac CT community
needs to recognize the utility of myocardial tissue characterization based on CT as
the field moves beyond coronary artery imaging. Clinical validation, the growing evidence
base, and products from CT vendors will facilitate this.
Future outlook
The introduction of photon-counting detector CT (PCCT) allows direct conversion of
X-ray photons to electrical signals, providing an increased contrast-to-noise ratio,
improved spatial resolution, reduced electronic noise, and the ability to acquire
spectral data during each scan [60]
[61]
[62]
[63]
[64]. These unique characteristics make it an attractive modality to further improve
myocardial tissue characterization with CT by direct computation of delayed enhancement
from the late enhancement (LE) scan [62]. Although the sample size was small and it was a single-center study, Mergen et al.
introduced PCD-CT for valvular disease assessment, highlighting its ability to accurately
assess ECV quantification and distribution in a cohort of severe AS patients [30].
Finally, these patients frequently have significant coronary artery disease that must
be excluded by invasive coronary angiography prior to intervention. Coronary computed
tomography angiography (CCTA) has already shown its excellent sensitivity for the
exclusion of coronary artery disease (CAD). Promising results regarding the potential
role of functional CAD assessment by CT (CT-FFR) in these patients have recently emerged
[65]. Although prospective validation is still lacking, this might correspond to the
last step for establishing CT as an ultimate “all-in-one” exam for these patients.
Furthermore, CCT play an important role in those with incongruent echo measurements
or a low-flow low-gradient phenotype that requires further evaluation. CCT can now
reliably characterize myocardial tissue, depicting signs of maladaptive remodeling
or patterns of increased ECV, providing prognostic stratification and potentially
tailoring therapeutic management towards frequently encountered pathologies concomitantly
present in AVS patients. In the future, this can potentially replace the need for
CMR, thereby omitting one additional imaging exam for these often elderly and frail
patients.
Conclusion
In patients with AVS, cardiac CT has long played a central role in procedural planning.
The assessment of myocardial health can provide valuable prognostic stratification.
Noninvasive tracking of extracellular components highlights the pathophysiological
transition from adaptive to maladaptive remodeling with the potential to enhance the
clinical management pathway that currently does not include the myocardial burden
as a criterium for intervention, besides impaired ejection fraction that can be a
sign that is too late. In the future, CT could become a tool for monitoring the response
to extracellular modulating therapies (anti-fibrotic, anti-amyloid) in the search
for new individualized heart failure therapies3.
