Repaired TOF
Repaired TOF
Repaired TOF
to the major
axis. In adult patients, RA area > 18 cm
2
indicates RA enlargement
(Table 3). Normal values in children are not available.
In a study that compared RAsize and function between 50 patients
with repaired TOF and 30 normal controls, Hui et al.
79
found that RA
size was increased (mean apical four-chamber area, 16.8 65.5 vs 13.7
6 5.1 cm
2
; P = .013) and emptying function was decreased (mean
active emptying area fraction, 19 6 9.8% vs 26.3 6 10.3%, P =
.005) after TOF repair. Similarly, RAsystolic and diastolic strain values
were lower in patients with TOF compared with controls.
LV Size and Function. Given that LV systolic dysfunction
measured by EF has been shown to be an important prognostic
marker for premature death in patients with repaired TOF,
50,80,81
its assessment is an essential element of the echocardiographic
examination in this population. The size, global function, and
regional wall motion of the left ventricle are determined from
multiple views. Systolic attening of the ventricular septum suggests
RV pressure overload, whereas diastolic septal attening indicates
volume overload. Measurement of LV systolic function by
shortening fraction assumes a circular geometry with homogenous
contraction, conditions that are seldom met in patients with
repaired TOF. The biplane area-length method may be less suscepti-
ble to these assumptions and is preferred when technically feasible.
82
Measurements of LV volumes and EF by 3D echocardiography are
preferred over 2D measurements. In patients with marked RV dilata-
tion, the right ventricle is often apex forming, necessitating adjustment
of the transducer position in the apical windows to avoid foreshorten-
ing of the left ventricle. In a study of 413 adult patients with repaired
TOF, Diller et al.
83
recently showed that nongeometric indices of
longitudinal LV function are associated with adverse clinical out-
comes. Specically, decreased mitral annular plane excursion
measured from the apical window (analogous to TAPSE) and LV
longitudinal strain were associated with the outcome independent
of QRS duration.
Given that most patients with repaired TOF reach adulthood and
the average age of this population continues to increase, the risk for
development of acquired ischemic heart disease is likely to increase.
Therefore, stress echocardiography may play an increasing role in
this population.
84,85
Stress can be induced with exercise or through
a pharmacologic agent (e.g., dobutamine).
Residual Shunts. The integrity of the VSD patch is determined by
imaging and color Doppler ow mapping. Residual VSDs can be
located at any part of the ventricular septum but are particularly
common at the superior portion of the patch. These can usually be
visualized by color Doppler from multiple imaging planes. Although
a high-velocity systolic jet suggests that RV pressure is low, it should
not be used independently to measure RV pressure, as the size of
the VSDmay vary during the cardiac cycle, and the jet may be eccen-
tric. Similarly, the atrial septum is examined by color Doppler for the
presence of a patent foramen ovale or a secundumatrial septal defect.
Rarely, unrepaired sinus venosus defect, coronary sinus septal defect,
or partially anomalous pulmonary venous connections are detected
late after TOF repair.
Aortic Valve, Root, and Ascending Aorta. Dilatation of the
aortic root is common in adults with repaired TOF, particularly those
with prior shunts and those with late repair.
86,87
Measurements of the
aortic root is accomplished from the parasternal long-axis view.
Imaging from the right parasternal window can facilitate depiction
and measurements of the proximal and mid ascending aorta.
Measurements are made using 2D imaging according to published
Figure 6 Tissue Doppler of (A) normal (peak S
0
velocity, 14 cm/sec) and (B) depressed (peak S
0
velocity, 7 cm/sec) RV systolic
function.
Journal of the American Society of Echocardiography
Volume 27 Number 2
Valente et al 119
guidelines in CHD during maximal expansion (mid to late systole).
88
Measurement of internal diameter is the current standard.
88,89
Other Cardiovascular Issues. Conrmation of aortic arch sided-
ness is accomplished by imaging from the suprasternal short-axis
aortic arch plane using both anterior and posterior tilt of the trans-
ducer to display the ascending and descending aorta. When the aortic
arch is right-sided (as it is in about 25% of patients with TOF), the de-
scending aorta may be seen coursing to the right. Additionally, iden-
tication of the innominate artery and its bifurcation into the left
subclavian and left common carotid arteries is possible from the
suprasternal window.
Evaluation of the origin and proximal course of the left and right
coronary arteries can be clinically relevant in patients with repaired
TOF. Although in many patients, this information is known before or
during reparative surgery, previously undetected abnormal coronary
anatomy can complicate reoperation or catheter intervention. The
major challenge in echocardiographic delineation of coronary anatomy
late after TOF repair relates to suboptimal acoustic windows.
d. Standard Protocol
Patient Preparation. Review of the patients medical and surgical
history, current medical condition, the specic questions for the echo-
cardiographic study, and the patients ability to cooperate with the ex-
amination should be ascertained before the scan commences. The
patients weight and height must be measured and recorded along
with calculation of the BSA using an appropriate formula, such as
the Haycock formula.
90
It is worth noting that the formula of
DuBois and DuBois
91
does not cover the ranges of ages and body
sizes encountered in patients with CHD and should not be used.
Documentation of the body mass index may also be helpful.
Additional steps taken before imaging begins include placement of
electrocardiographic leads that avoid the standard echocardiographic
imaging sites and recording the blood pressure and the heart rate.
When pertinent, the blood pressure should be recorded on the oppo-
site side of a previous Blalock-Taussig shunt.
Scanning Protocol. The ASE guidelines published in 2006 detail
scanning protocols for the performance of pediatric echocardiogra-
phy in patients with CHD,
92
and guidelines published in 2010 specify
measurement techniques.
88
Importantly, each laboratory should
followa consistent detailed protocol that addresses the pertinent clin-
ical issues in patients with repaired TOF. A sample echocardiographic
protocol for patients with repaired TOF is listed in Table 2, and refer-
ence values for selected measurements are shown in Table 3.
e. Reporting Elements and Measurements
RVOT and MPA
Dimensions
Location and mechanism of obstruction by 2D, color, and spectral
Doppler
Presence of an aneurysm
Peak and mean gradients
RV-to-PA conduit: peak and mean gradients
Degree of PR on the basis of (1) regurgitation jet width by color Doppler, (2)
spectral Doppler noting the duration and the slope (or pressure half-time) of
the regurgitation jet, (3) presence and degree of ow reversal in the branch
PAs, and (4) RPA pulsatility (systolic-diastolic diameter ratio)
93
Branch PAs
Dimensions of the narrowest and maximal segments
Location and severity of obstruction by 2D, color, and spectral Doppler
Degree and mechanism of TR; if more than mild, measure (1) vena con-
tracta width and (2) tethering height and area (optional)
RV pressure on the basis of (1) TR jet velocity, (2) trans-VSD gradient, (3)
systolic septal conguration (note the presence of right bundle branch block
or septal dyskinesis)
RV size and volume load on the basis of (1) TV annular diameter, (2) dia-
stolic septal attening, and (3) measurements of RV size as detailed above
RV function on the basis of the parameters detailed above
Residual VSDs: location, size, direction of ow, and peak transseptal
gradient
Residual atrial septal defects: location, size, and direction of ow
Aortic dimensions: annulus, root, and ascending aorta at the level of the RPA
AR
Systemic-to-pulmonary collateral vessels on the basis of color Doppler inter-
rogation and spectral Doppler evaluation of the abdominal aorta for dia-
stolic runoff
LV size and function
f. Transesophageal and Intracardiac Echocardiography
The primary role of TEE in patients with repaired TOF is for intraoper-
ative assessment in conjunction with late interventions such as PV
replacement and TV repair. Precardiopulmonary bypass TEE may
be performed to evaluate the atrial septum for the presence of a pat-
ent foramen ovale and the TV for the degree and mechanism of TR.
TEE may be especially instrumental in the evaluation of the TV and
PVas well as RV-to-PAconduits or the implanted PV for the presence
of vegetations that were not detected by transthoracic imaging. After
cardiopulmonary bypass, TEE is used to evaluate the repair and for
functional assessment of the ventricles and valves. TEE or intracardiac
echocardiography may also be used for guidance of transcatheter PV
implantation
94
and device closure of residual atrial septal defects.
g. Recommendations
The committee recommends comprehensive echocardiographic
evaluation for longitudinal follow-up of patients with repaired TOF
through the use of a standardized examination protocol. Given that
the management of chronic PR is a subject of intense research with
ongoing evolution of therapies such as transcatheter PVimplantation,
the echocardiographic imaging protocol should be updated periodi-
cally as new information emerges. Integration of echocardiographic
data with information from other modalities is imperative for optimal
management of these patients, particularly given limitations of acous-
tic windows and assumptions of RV geometry that may well fail.
6. CARDIOVASCULAR MAGNETIC RESONANCE IMAGING
a. Overview of Modality
CMR is considered the reference standard for the quantication of
RV size, RV function, and PR in patients with repaired TOF.
10
Through the use of multiple CMR techniques, different morphologic
and hemodynamic aspects of the relevant pathophysiology are eval-
uated. In the context of this discussion, the most widely used imaging
sequence is steady-state free precession (SSFP), which is a type of
gradient-echo technique characterized by high signal-to-noise ratio,
high T2/T1 contrast ratio, and sharp borders between the blood
pool and the myocardium.
95
Electrocardiographically gated SSFP
can be used as a cine magnetic resonance sequence, which is typically
used for assessment of ventricular size and function, valve function,
and intracardiac and extracardiac anatomy. Alternatively, an
120 Valente et al Journal of the American Society of Echocardiography
February 2014
Table 2 Example of an echocardiographic examination protocol in patients with repaired TOF
Imaging plane Technique Parameters Goals and comments
Subxiphoid long-axis Supine
Knees bent
May require breath hold with
inspiration
LV length for 5/6 area length
(bullet) measurement
RV free wall thickness
Visceral situs and cardiac
position
Diaphragm function
RV/LV size and function
Atrial septum
Ventricular septum for residual
VSDs
Subxiphoid short-axis Supine
Knees bent
May require breath hold with
inspiration
Doppler abdominal Ao for
diastolic runoff
IVC: respiratory variation and
diameter
Doppler hepatic veins
LV short-axis at papillary muscle
level for 5/6 area length
measurement
Color and spectral Doppler in
RVOT
IVC diameter and collapse for
RA pressure
Residual ASDs
Residual VSDs
RVOT obstruction
Mechanism of TR
Diastolic runoff in the abdominal
Ao indicative of AR, systemic-to-
pulmonary collateral vessels, or
presence of a shunt
Subxiphoid RAO
(counterclockwise 45
rotation
from long-axis view allowing
simultaneous visualization of
RV inow and outow at the
same time)
Supine
Knees bent
May require breath hold with
inspiration
Color owmapping and Doppler
of the RVOT
2D measurement of PV annulus
and RPA
Residual VSD (via patch leak or
intramural egress)
Degree and location of RVOT
obstruction
Doppler assessment of PR
Apical four-chamber Left lateral decubitus RA area (at end-systole)
TV annular lateral diameter
RV internal diameter at the base
and mid levels at end-diastole
M-mode/2D of TV annulus for
TAPSE
Modied Apical four-chamber
optimizing RV for FAC
Color Doppler ow mapping of
TR and measure vena contracta
Doppler TR for RV pressure
3D when obtainable
AR by color and spectral
Doppler
LV systolic and diastolic
functional assessment (2D,
Doppler)
DTI: lateral TV and MV annulus*
2D strain
RV optimization may be
achieved by moving the
transducer medially or laterally
using the left ventricle as the
window
Contemporaneous BP
measurement preferred for TR
gradient assessment opposite the
site of the BT shunt (if pertinent)
Aim for high frame rates for DTI
and 2D strain
Assess for regional wall motion
abnormalities
Apical LV two-chamber Left lateral decubitus 2D measurements for LV volume Assess for regional wall motion
abnormalities
Assess longitudinal strain
Apical three-chamber Left lateral decubitus AR by color and spectral
Doppler
Residual VSDs, especially
subaortic
Parasternal long-axis Left lateral decubitus 2D measurement of TV AP
diameter
Color and spectral Doppler of
TR; measure vena contracta
Proximal and distal RVOT
diameters
PV annulus and MPA
RVOT and MPA color and
spectral Doppler
Aortic annulus and root
dimensions
AR: location and measure vena
contracta width
Residual VSD
Degree of PR, TR, and AR
RVOT size and regional wall
motion abnormalities
Aortic root and ascending Ao
size
(Continued)
Journal of the American Society of Echocardiography
Volume 27 Number 2
Valente et al 121
electrocardiographically gated, respiratory-navigated SSFP sequence
can yield a highspatial resolution static 3D data set, which is often
used for detailed assessment of intracardiac anatomy and/or coronary
artery anatomy. Electrocardiographically gated turbo (fast) spin-echo
(TSE) imaging offers high spatial resolution (submillimeter in-plane),
excellent contrast between elements of soft tissue, and decreased
sensitivity to metallic artifacts compared with gradient-echo se-
quences, although it provides only static images.
Contrast-enhanced magnetic resonance angiography (MRA) rep-
resents a robust 3D technique ideally suited for assessment of the
arterial tree and the systemic and pulmonary veins. The acquisition
of a 3D data set allows ofine rendering of images in any arbitrary
plane to best depict the anatomy in question, typically using a
maximum intensity projection reconstruction technique.
Alternatively, the anatomy can be displayed using surface or volume
rendering techniques, which yield 3D models of cardiovascular struc-
tures. It should be noted, however, that contrast-enhanced 3D MRA
is not synchronized with the cardiac cycle and is therefore subject to
artifacts from cardiac motion. As a result, intracardiac morphology is
affected by motion-induced blurring, although the great vessels are
less affected and are typically seen well with this technique.
Electrocardiographically gated, phase-contrast (PC) ow measure-
ments are used for measurements of blood ow, including ow rates
in the great arteries and veins, differential PAow, and regurgitant vol-
umes (e.g., PR, AR).
Several techniques are available for assessment of myocardial
perfusion, ischemia, and scar tissue. The most commonly used tech-
nique for assessment of gross myocardial brosis in patients with re-
paired TOF is the late gadolinium enhancement (LGE) sequence.
This technique has been shown to be helpful in this group of patients
for identifying those at risk for ventricular tachyarrhythmias and exer-
cise intolerance.
96,97
The role of CMR in patients with repaired TOF varies according to
age and clinical circumstance. During the rst decade of life, the mo-
dality is used only when echocardiographic, electrocardiographic, and
clinical data are either inconsistent or concerning for deterioration
necessitating surgery or catheter intervention. Depending on the spe-
cic question, CMR may provide the necessary information to deter-
mine the best course of therapy. CMR is not recommended routinely
in this age group, because the prevalence of major complications is
low and echocardiographic windows are usually sufcient for
adequate diagnostic evaluation. Once patients reach adolescence,
however, CMR is recommended as a routine test for follow-up of
RV size and function and PR. The rationale for that is that at this
age group, the likelihood of adverse clinical outcomes increases,
and echocardiographic windows often become more restricted. In
addition, CMR does not require sedation in adolescents and adults.
The ability of this modality to provide accurate quantitative informa-
tion on biventricular size and function, blood ow measurements,
myocardial viability, and cardiovascular anatomy has led to CMRs
Table 2 (Continued)
Imaging plane Technique Parameters Goals and comments
Parasternal short-axis Left lateral decubitus 2D measurement of proximal,
distal RVOT, branch PAs
Color and spectral Doppler of
RVOT, MPA and branch PAs
Color Doppler of AR jet
PR: (1) jet width, (2) level and
degree of diastolic ow reversal
in the MPAand branch PAs, and
(3) Doppler ow prole
(deceleration rate)
RV size and function
LV size and function
Suprasternal notch (Ao short-axis) Supine
Extend the neck with slight head
tilt to the left
Arch sidedness and branching
order
Presence and location of the left
innominate vein
Rule out persistent LSVC and/or
presence of anomalous
pulmonary vein
2Dmeasurements of central and
distal RPA
Color Doppler for BT shunt (if
present)
Follow the rst vessel off the
aortic arch for arch sidedness
and for bifurcation indicative of
normal vs mirror-image
branching
Suprasternal notch (Ao arch long-
axis)
Supine
Extend the neck with slight head
tilt to the left
2D measurement of LPA
dimension
Color Doppler of the arch,
assessing for diastolic ow
reversal
Color Doppler for
aortopulmonary collateral vessels
Rule out PDA
Assess degree of aortic arch
dilation
Pulse Doppler around the arch
for degree of diastolic runoff to aid
in determination of AR grade
Right parasternal (parasagittal) Right lateral decubitus
May require breath holding in
expiration
2D measurement of ascending
Ao
Color Doppler for residual ASD
Dilated ascending aorta
Residual ASD
Ao, Aorta; ASD, atrial septal defect; BT, Blalock-Taussig; BP, blood pressure; DTI, Doppler tissue imaging; FAC, fractional area change; LSVC, left
superior vena cava; MV, mitral valve; PDA, patent ductus arteriosus; RAO, right anterior oblique.
*Assessment of septal velocities is hampered by presence of the VSD patch.
122 Valente et al Journal of the American Society of Echocardiography
February 2014
becoming the preferred method of noninvasive imaging in adult pa-
tients with repaired TOF at many centers.
50,51,98-101
b. Strengths and Limitations
CMR is ideally suited for assessment of repaired TOF because it al-
lows comprehensive assessment of cardiovascular morphology and
physiology without most of the limitations that hinder alternative im-
aging modalities. Specically, CMR is independent of acoustic win-
dows. It provides excellent image quality in patients with a wide
range of body sizes and is not adversely affected by most surgical
and catheter-based interventions. Another advantage of CMR is the
absence of harmful ionizing radiation, making it appealing for longitu-
dinal follow-up.
11
A notable strength of the modality is its ability to
provide accurate and reproducible measurements of RV size and
function, valve regurgitation, pulmonary and systemic ows, and dif-
ferential PA ow and to detect scar tissue in the ventricular myocar-
dium.
Limitations of CMRinclude a higher cost in comparison with echo-
cardiography (but not in comparison with other modalities),
9
lack of
portability, limited availability, artifacts fromimplants containing stain-
less steel,
102
and relative contraindication in patients with pacemakers
or implantable debrillators.
103
It is worth noting that certain old im-
plants, such as stainless-steel coils and other devices with high metallic
content, have been largely replaced by implants made of minimally or
nonferromagnetic materials. More recently, the risk for nephrogenic
systemic brosis that has been linked to gadolinium-based contrast
has been mitigated by avoiding its use in patients with reduced
glomerular ltration rates.
104
Importantly, when it comes to evalua-
tion of the right heart by CMR, the use of a contrast agent is not a
requirement.
c. Assessment of Repaired TOF with CMR
RVOT. Several imaging techniques, including cine SSFP, TSE, and
contrast MRA, can be used to evaluate the RVOT. Cine imaging is ad-
vantageous because it depicts RVOT wall motion, allowing determi-
nation of akinetic or dyskinetic wall segments. This can be done in
a plane oriented to show the RV inow and outow, an axial plane,
and an oblique sagittal plane through the long axis of the RVOT
and MPA (Figure 7). Signicant obstruction to ow across the
RVOT is seen by a combination of observing anatomic narrowing
on cine imaging and the presence of systolic signal ow void in that
location. The SSFP sequence is also capable of detecting the presence
of thrombus associated with an aneurysm of the RVOT, as illustrated
in Figure 8. Avolume-rendered MRAreconstruction allows a 3Drep-
resentation of the outow tract, including any aneurysm that may be
present (Figure 9). CMR is helpful in understanding the RVOT geom-
etry, which is an important element of planning percutaneous PVim-
plantation.
105,106
PAs. Branch PA anatomy is well visualized with CMR, and assess-
ment of differential PA ow is measured with cine PC prescribed
perpendicular to the vessels.
107
MRAis particularly valuable for imag-
ing the PAs, as the 3D data set provides enhanced depth perception
and visualization of complex anatomic relationships with neighboring
structures (Figure 10). This modality has proved accurate in the
evaluation of branch PA stenosis, with good agreement with tradi-
tional x-ray angiography.
108
Recently, the utility of MRA for planning
interventions for PA stenosis has been reported.
109
Imaging using TSE sequences is useful for depicting the extent and
location of branch PA stenosis because of its higher spatial resolution.
TSE sequences are also less susceptible to metallic artifacts and are
used in patients in whom prior PA stents have been placed.
However, their accuracy in depicting in-stent stenosis is not known.
Nordmeyer et al.
110
demonstrated in an experimental model that
gradient-echo, SSFP, and contrast CMR sequences are capable of as-
sessing for in-stent stenosis.
PR. PR quantication is an important aspect of evaluation after TOF
repair. Electrocardiographically gated free-breathing cine PC is the
reference standard technique for ow measurements (Figure 11); in
the case of PR measurement, this is performed perpendicular to the
long-axis plane of the proximal MPA. PRfraction is calculated as retro-
grade ow volume divided by antegrade ow volume. In addition to
Figure 7 CMRSSFP image of an aneurysmof the RVOT (arrow).
Table 3 Echocardiographic reference values of RA and
ventricular size and function in healthy adults (adapted from
Rudski et al.
23
)
Parameter Abnormal
RV diameter (mm)
Base >42
Midventricular level >35
RV diastolic length (mm) >86
RV diaphragmatic wall thickness
(subcostal view) (mm)
>5
Systolic function
TAPSE (mm) <16
Pulsed Doppler peak S
0
(cm/sec) <10
FAC (%) <35
Diastolic function
E/E
0
ratio >6
RA end-systolic area (cm
2
) >18
RA length (base to apex) (mm) >53
RA lateral diameter (mm) >44
FAC, Fractional area change.
Journal of the American Society of Echocardiography
Volume 27 Number 2
Valente et al 123
the PR fraction, the absolute value of the retrograde ow volume
should be reported.
111
In the absence of a residual shunt, the net
ow in the MPA and ascending aorta should be nearly identical.
Similarly, in the absence of other valve regurgitation, shunt, or signif-
icant late diastolic antegrade ow in the MPA, LVand RV stroke vol-
ume differential is primarily affected by PR, and calculation of PR
fraction by the two methods should be similar. In 15% to 20% of pa-
tients with repaired TOF, however, residual shunts and TR and/or AR
are present, thus limiting the use of these comparisons.
112
TV Morphology and Function. TV dysfunction is not rare in pa-
tients with repaired TOF, leading to TR that contributes to right heart
dilatation. Mild or less TR is seen in most patients, whereas the fre-
quency of at least moderate TR has been reported in 10% to 15%
of patients.
39,112
The mechanisms of TR in this patient population
are discussed in the section on echocardiography. Although TEE
remains the gold standard for detailed visualization of the
atrioventricular valves, TV anatomy is well visualized in a stack of
axial and short-axis cine SSFP acquisitions. In addition, an extended
two-chamber RV view further depicts the TV in a vertical long-axis
plane. The magnitude of TR can be quantied with cine PC oriented
perpendicular to the TV inow acquired in a four-chamber view at
the level of the leaets tips. Quantication of TR has yet to be vigor-
ously validated, in part because of the lack of a robust standard of
reference with which measurements can be compared. The substan-
tial through-plane motion of the TV poses a challenge for accurately
measuring TV inow.
113
Incorporating RV volumetric results with
cine PC data may have inherent error greater than that for semilunar
valve regurgitation.
114
Right Ventricle. i. Size and Function.Electrocardiographically
gated cine SSFP is the current gold standard for quantitative evalua-
tion of LV and RV volumes and mass.
9,10
Accurate determination of
ventricular volume requires clear depiction of the blood-
Figure 8 CMR evaluation of an RVOT aneurysm with thrombus (arrow) in a patient with repaired TOF. (A) SSFP. (B) LGE image with
long inversion time in the same imaging plane showing no enhancement of the mass consistent with a thrombus. Ao, Aorta; RA, right
atrium.
Figure 10 Volume-rendered magnetic resonance angiographic
reconstruction 3D representation of proximal branch RPA ste-
nosis (arrow).
Figure 9 Volume-rendered gadolinium magnetic resonance
angiographic reconstruction 3D representation of a large
RVOT aneurysm.
124 Valente et al Journal of the American Society of Echocardiography
February 2014
myocardium boundary. Adjustments of the image brightness and
contrast on the computer screen can facilitate visualization of that
boundary. By tracing the blood-endocardium boundary, the slices
blood pool volume is calculated as the product of its cross-sectional
area and thickness (which is prescribed by the operator). The LVpapil-
lary muscles and the major trabeculations of the right ventricle (e.g.,
septal band) are excluded from the blood pool and are considered
part of the myocardium.
11,115
The ventricular volume is then
determined by summation of the volumes of all slices. The process
can be repeated for each frame in the cardiac cycle to obtain a
continuous time-volume loop or may be performed only on end-
diastolic (maximal area) and end-systolic (minimal area) frames to
calculate diastolic and systolic volumes. From these data one can
calculate ventricular stroke volumes and EFs. Because the patients
heart rate at the time of image acquisition is known, ventricular out-
puts can be calculated. Cross-referencing the short-axis images with
LVand RV two-chamber (vertical long-axis) and four-chamber (hori-
zontal long-axis) cine SSFP facilitates accurate determination of the
atrioventricular and semilunar valves planes during systole and dias-
tole.
116
In patients with repaired TOF, particular attention should be
paid to determining the end-diastolic and end-systolic phases of
each ventricle. Given that conduction delay is nearly universal in
this population, peak RV contraction typically lags after that of the
LV by one to three cardiac phases.
11
Ventricular mass is calculated by tracing the epicardial borders
and calculating the epicardial volume, subtracting the endocardial
volume, and multiplying the resultant muscle volume by the specic
gravity of the myocardium (1.05 g/mm
3
). In the right ventricle, this
measurement is challenging because of its trabecular morphology.
One approach is to ignore the ne trabeculations and to trace the
compacted free wall myocardium (the major septal trabecula-
tionsseptal and moderator bandsare excluded from the blood
pool and considered RV mass).
115
Recent data indicates that this
measurement is more reproducible when performed during systole
compared with measurement in diastole.
117
An alternative
approach is to use a semiautomatic threshold-based technique as
described by Sarikouch et al.
118
The different approaches to con-
touring the myocardial trabeculations highlight the variations in
practice that currently exist. Further research and technical advances
are needed to improve the accuracy and reproducibility of these
measurements.
From a clinical perspective, recent evidence indicates that
increased RV mass measured by CMR in patients with repaired
TOF is associated with greater risk for adverse outcomes, including
sustained arrhythmia and death.
119
To optimize interstudy reproduc-
ibility in patients followed longitudinally, contours should be
compared side by side with those from previous studies. Saving the
contour les along with previous studies facilitates this comparison.
Figure 11 PA owquantication in a patient with repaired TOF and severe PR. Imaging plane is positioned perpendicular to the prox-
imal MPA (top left). Corresponding phase velocity map with region of interest drawn around the MPA ow signal (top right). Flow rate
over time is shown in the lower panel. Volume and direction of ow is determined, and the PR fraction can be calculated by dividing
the retrograde volume by the antegrade volume. This patient has severe PR with a regurgitation fraction of 44%.
Journal of the American Society of Echocardiography
Volume 27 Number 2
Valente et al 125
On average, total manual analysis time is about 25 to 30 min and de-
creases with operator experience.
11
ii. Viability.Through extensive research since the late 1990s, the
LGE technique has been shown to be highly sensitive and specic
for the presence of myocardial brosis.
120
In this technique, nonviable
myocardium appears bright, whereas viable myocardium appears
dark on images obtained 10 to 20 min after the administration of gad-
oliniumcontrast. LGE is also present as a result of inammation, in the
region of patch material, and other abnormalities of the myocardium.
Presence of LGE has been associated with adverse clinical outcomes
in patients with ischemic heart disease, cardiomyopathies, myocardial
storage disorders, and valvular heart disease.
120,121
In patients with
repaired TOF, LGE occurs commonly in locations of prior surgery
(RVOT, VSD patch), as shown in Figure 12. In addition, LGE can
also be found in the anterior RV free wall extending beyond the re-
gion of the patch in the left ventricle, and in the interventricular
septum.
96,97
LGE in the superior and inferior junctions between
the septum and the ventricular free wall is ubiquitous, and its
clinical signicance remains to be elucidated. A greater degree of
LGE in these patients has been associated with exercise intolerance,
ventricular dysfunction, and clinical arrhythmias.
96,97
However, it is
unclear if LGE is associated with increased mortality in this patient
population. The role of dobutamine stress CMR has been
evaluated in small studies and awaits further study.
78,122,123
Right Atrium. RA enlargement can be caused by signicant TR
and/or by increased RVend-diastolic pressure. A paucity of informa-
tion is available on the range of RA dimensions and function in this
population of patients. In a study comparing 20 patients with repaired
TOF with seven healthy controls, Riesenkampff et al.
124
noted RA
enlargement and abnormal reservoir and conduit function, which
were associated with RV size and function. Specically, patients
with repaired TOF showed decreased early RA emptying, increased
late emptying, and larger minimal RA volumes compared with con-
trols. Patients with large RAand RVsizes showed the greatest degrees
of LV diastolic function.
125
Further studies are required to shed light
on the prognostic value of RA parameters with regard to clinical out-
comes.
LV Size and Function. LV systolic dysfunction is present in up to
20% of adults with repaired TOF
126
and has been shown to be a
strong independent predictor of death and sustained ventricular
tachycardia.
50,80,81
A CMR investigation in 100 patients with
repaired TOF demonstrated a close linear correlation between the
EFs of the right and left ventricles.
50
This observation, which has
been subsequently conrmed by other investigators, may be related
to the reverse Bernheim effect relating RV dilatation and dysfunc-
tion, leading to LV dysfunction.
127
As discussed in the section on
echocardiography, the precise mechanismor mechanisms that under-
lie adverse ventricular-ventricular interaction are a subject of current
research.
128
Given the clinical importance of LV dysfunction in this
group of patients, measurements of LV size, global and regional func-
tion, and viability are integral to the CMR examination protocol.
Measurements of LV size and function are performed as described
above for the right ventricle.
Residual Shunts. Although residual VSDs are usually diagnosed
by color Doppler echocardiography, they can also be identied
with SSFP or TSE imaging. Residual patch margin VSDs may be
assessed by prescribing a stack of cine SSFP in plane parallel and
perpendicular to the base of the ventricular septum, including the
four-chamber plane, the three-chamber viewhighlighting the outow
tract, and the ventricular short axis. Similarly, the atrial septum is
examined for a residual communication by cine SSFP, and MRA is
used to exclude anomalous connections of the pulmonary veins.
The magnitude of the systemic-to-pulmonary ow ratio can be accu-
rately quantied by cine PC comparison of the net ow in the MPA
and ascending aorta.
129
Aortic Valve, Aortic Root, and Ascending Aorta. Close moni-
toring of the aortic root and ascending aorta for dilatation is routinely
performed in patients with repaired TOF by obtaining cine SSFP se-
quences parallel and perpendicular to these locations.
Electrocardiographically gated 3D SSFP is another technique that al-
lows accurate and reproducible measurements of the proximal aorta
without the use of a contrast medium.
130
Measurements of the great
vessels are obtained from a cross-sectional image of the aorta (or PA)
perpendicular to its long axis using a double-oblique approach as
described by Kaiser et al.
131
Contrast-enhanced MRA is also useful for evaluating and
measuring the aortic root and ascending aorta. Ideally, the diameters
of the aortic sinuses and sinotubular junction should be measured on
electrocardiographically gated images to avoid motion artifacts and
blurring. Consistent methods of acquisition and measurement are
essential for the evaluation of any changes over time.
132
Aortic valve
regurgitation fraction is measured by a cine PC sequence prescribed
perpendicular to the aortic root.
Right aortic arch is found in approximately 25% of patients with
TOF and is more common in those with chromosome 22q11 dele-
tion. These patients are also more susceptible to abnormal branching
patterns of the arch vessels. Rarely, repaired TOF can be associated
with a previously undiagnosed vascular ring. The aortic arch and
the brachiocephalic vessels are evaluated by MRA for arch sidedness
and abnormal branching pattern.
Other Cardiovascular Issues. CMR is useful in evaluating the
spatial relationships between cardiovascular structures and the air-
ways. Another important role of CMR is to identify abnormal
Figure 12 LGE image in ventricular short-axis showing late hy-
perenhancement (yellow arrows) of the RVOT patch. Note
extension of the hyperenhancement to the superior wall of the
right ventricle (RV; red arrow). LV, Left ventricle.
126 Valente et al Journal of the American Society of Echocardiography
February 2014
coronary artery origins and proximal courses. This is particularly
important in patients with prior repair who are being considered
for percutaneous PV implantation.
133
The current adult CHD man-
agement guidelines recommend that coronary anatomy should be
determined before any intervention on the RVOT (class I, level of
evidence C).
12
Another consideration before reoperation in the
patient with repaired TOF (e.g., PV replacement) is RV anterior
wall and/or ascending aorta adherent to the posterior sternal table
because these structures are at risk for accidental injury during
sternotomy.
Noncardiac Findings. As patients with repaired TOF survive to an
older age, one can anticipate a higher potential for identifying extrac-
ardiac ndings by CMR. In a study of 240 CMRexaminations done in
non-CHDpatients, 27%of studies were found to have previously un-
recognized incidental ndings, including pleural effusions, atelectasis,
and adenopathy, as well as ve cases of newly diagnosed malig-
nancy.
134
d. Standard Protocol
Patient Preparation. The importance of careful attention to de-
tails of patient preparation and placement in the scanner cannot be
overemphasized. Optimal placement of electrocardiographic leads
is paramount to quality gating. A peripheral intravenous cannula for
injection of gadolinium-based contrast is placed in the following cir-
cumstances (subject to practice variations):
First CMR examination
>3 years since last LGE evaluation
Deterioration in clinical status
Regional or global ventricular function has worsened
Placement of an intravenous cannula in young patients may be de-
ferred in settings in which attempts at intravenous cannula insertion
might result in loss of patient cooperation.
11
Scanning Protocol. An example of a CMR examination protocol
for repaired TOF patients is listed in Table 4, and reference normal
values for measurements are provided in Table 5. The examination
protocol is designed to accomplish the following objectives:
Quantitative assessment of RVand LV volumes, mass, stroke volumes, and
EFs
Evaluation of regional wall motion abnormalities
Imaging the anatomy of the RVOT, PAs, aorta, and aortopulmonary collat-
eral vessels
Quantication of PR, TR, cardiac output, and pulmonary-to-systemic ow
ratio
Assessment of myocardial viability, with particular attention to scar tissue in
the ventricular myocardium aside from sites of previous surgery (e.g., VSD
and RVOT patches)
Relationship of cardiovascular structures to the sternum
e. Reporting Elements and Measurements
RVOTobstruction and or aneurysm; length of dyskinetic segment
Branch PA stenosis
RV and LV size and function (absolute and BSA-adjusted values): end-
diastolic volume, end-systolic volume, stroke volume, EF, mass, mass-to-
volume ratio
Vessel dimensions: aortic root, ascending aorta, MPA, RPA, LPA
RA size by the area-length method
135
Flow measurements
Ascending aorta, MPA, RPA, LPA
PR fraction and volume
Other valve regurgitation
LGE presence and extent
f. Recommendations
CMR is recommended in young patients (usually #10 years) when
the echocardiographic and clinical data are concerning for unfavor-
able disease progression (e.g., ventricular dysfunction, right heart
failure symptoms). Once patients reach adolescence, CMR is per-
formed routinely using a comprehensive imaging protocol.
Quantitative assessment of ventricular size and function and PR,
for which CMR is the reference standard, are essential elements
of the imaging protocol.
7. CARDIOVASCULAR COMPUTED TOMOGRAPHY
a. Overview of Modality
Widespread availability of CT has lead to increased utilization of this
noninvasive modality in the general population.
136
Early challenges
with motion artifacts encountered with imaging complex congenital
anatomy throughout the cardiac cycle limited the evolution of this
application in the assessment of this patient population. However, ad-
vances in multidetector computed tomographic technology have led
to improvements in both spatial and temporal resolutions, allowing
the acquisition of a volume of data during a single breath-hold exam-
ination.
137
Subsequently, the ability to produce static images of the
heart and great vessels as well as 3D reconstructions has revolution-
ized this technique and propelled cardiac CT into the arena of
anatomic imaging of patients with complex CHD.
In general with MDCT, the x-ray point source(s) and the detector
array are located on opposite sides of the patient on a ringlike struc-
ture called a gantry. The gantry rotates around the patient, located
at the table center. The table then moves at a constant speed as x-
rays penetrate the patient and are captured by one or more of the de-
tectors. Once complete, a series of helical projections are generated.
From these data, images representing x-ray attenuation from each
point in the volume are mathematically reconstructed.
138
Certain
scanners are capable of volume scanning without table movement.
139
Scanners with fewer detectors create an image set from data ac-
quired from several heartbeats. To improve overall quality of
computed tomographic images and ensure a longer diastolic window
with minimal cardiac motion, pharmacologic heart rate control is
frequently used. Successful prescan heart rate control facilitates post-
processing that requires image reconstruction of the data set in a xed
interval preceding the R wave. Furthermore, heart rate optimization
avoids gating errors associated with heart rate changes that may occur
during image acquisition and that lead to image artifacts. These strate-
gies are particularly important in the evaluation of the coronary ar-
teries and ascending aorta to avoid motion artifacts that occur
throughout the cardiac cycle. Therefore, heart rate control with oral
or intravenous b-blockade is advised before CT for patients without
contraindications.
140
Data postprocessing commences after image acquisition is com-
plete. Postprocessing includes (1) electrocardiographic review for
ectopic beats or arrhythmias, (2) selection of the optimal phase for
reconstruction, (3) selection of appropriate reconstruction parame-
ters and kernels, (4) review of the axial (source) data set, and (5)
Journal of the American Society of Echocardiography
Volume 27 Number 2
Valente et al 127
Table 4 Example of a CMR examination protocol in patients with repaired TOF (adapted from Geva
11
)
Sequence Technique Imaging planes Parameters Goals and comments
Localizing images ECG-gated SSFP
Real-time interactive SSFP
Axial, coronal, and sagittal
Multiple oblique planes
Anatomic survey and planning
subsequent sequences
Cine MR evaluation of
anatomy and ventricular and
valve function
ECG triggered, breath-hold
cine SSFP
LV two-chamber (vertical
long-axis)
RV two-chamber (vertical
long-axis)
Four-chamber (horizontal
long-axis)
RVOT long-axis
LVOT long-axis
Axial stack: RVOT and PAs
TE, 1.7 msec; TR, 3.3 msec;
ip angle, 60
; SENSE
acceleration factor, 2; FOV,
260 mm; matrix size, 160
160 reconstructed to 256
256; voxel size, 1.6 1.8
68 mmreconstructed to 1.0
1.0 68 mm; 30
reconstructed images per
cardiac cycle
Evaluation of intracardiac
anatomy, residual septal
defects, outow tract
obstruction
Cine MR evaluation of
ventricular volumes and
mass
ECG triggered, breath-hold
cine SSFP
Ventricular short-axis Same parameters; 1214
equidistant slices (slice
thickness, 68 mm;
interslice space, 02 mm)
covering the entire length of
both ventricles
Attention to inclusion of the
base of the right and left
ventricles at end-diastole
with addition of extra slices
as needed for complete
coverage
Gadolinium-enhanced 3D MR
angiography
Non-ECG triggered 3D spoiled
gradient-echo
Sagittal or coronal TE, 1.5 msec; TR, 4.5 msec;
ip angle, 40
; SENSE
factor, 2; FOV, 300 mm;
matrix size, 192 192; voxel
size, 1.6 1.6 6.0 mm
reconstructed to 1.2 1.2
6.0 mm; 40 reconstructed
images per cardiac cycle
Quantication of PR, other
valve regurgitation,
pulmonary and systemic
ow, differential PA ow
LGE ECG triggered, breath-hold,
phase-sensitive LGE
Ventricular short-axis, LV two-
chamber, LV three-
chamber, RV two-chamber,
and four-chamber
TE, 3.5 msec; TR, 5.9 msec;
ip angle, 20
; FOV, 260
mm; matrix size, 144 144;
voxel size, 1.8 1.8 78
mm reconstructed to 1.0
1.0 78 mm
Identication of myocardial
scar tissue
AV, Atrioventricular; ECG, electrocardiographically; FOV, eld of view; LVOT, LV outow tract; MR, magnetic resonance; SENSE, sensitivity encoding; TE, echo time; TR, repetition time.
Additional case-specic sequences include (1) cine SSFP in the short axis of the aortic root and ascending aorta in patients with dilated aortic root and ascending aorta; (2) TSE sequence
with blood suppression for imaging of the outow tracts and branch PAs in patients with image artifacts from metallic implants; (3) ECG triggered, respiratory navigated, free-breathing 3D
isotropic SSFP for evaluation of the coronary arteries or as a substitute for contrast MRA; and (4) T1 mapping for evaluation of diffuse myocardial brosis.
1
2
8
V
a
l
e
n
t
e
e
t
a
l
J
o
u
r
n
a
l
o
f
t
h
e
A
m
e
r
i
c
a
n
S
o
c
i
e
t
y
o
f
E
c
h
o
c
a
r
d
i
o
g
r
a
p
h
y
F
e
b
r
u
a
r
y
2
0
1
4
evaluation of anatomic abnormalities using multiplanar reformats,
including true short-axis as well as long-axis maximum intensity pro-
jection images.
141
As a result of these advances in computed tomographic technol-
ogy, MDCT is now emerging as a complementary modality to estab-
lished noninvasive imaging techniques as well as a viable alternative in
those with absolute contraindications to CMR.
142
b. Strengths and Limitations
The major advantage of CT in patients with repaired TOF is its excel-
lent (usually submillimeter) spatial resolution. With well-timed, high-
quality computed tomographic angiography, the contrast-to-noise
ratio is also excellent. Thus, electrocardiographically gated computed
tomographic angiography provides a very clear depiction of cardio-
vascular anatomy. This advantage is particularly valuable for detailed
evaluation of small blood vessels such as the coronary arteries or the
distal PA branches. Another advantage is the ability to perform CT in
patients with pacemakers and debrillators. Additionally, CT may
allow successful imaging of structures that are obscured on CMR im-
aging by stainless-steel metallic artifacts.
CT has several noteworthy limitations. As with any modality that
uses ionizing radiation, exposure increases the risk for cancer.
143-145
This risk increases with a higher radiation dose, repeated exposures,
younger age, and female gender.
146,147
Electrocardiographically
gated cardiac CT is associated with a higher radiation dose
compared with a nongated scan. Another limitation of cardiac CT
is lower temporal resolution compared with echocardiography and
CMR. In addition, compared with these modalities, CT does not
provide hemodynamic information on ow rate or velocity. Finally,
the use of contrast is associated with risk for renal dysfunction in
patients with impaired renal function.
148
c. Assessment of Repaired TOF with CT
PAs. The spectrum of morphologic abnormalities of the PAs,
including obstruction, stenosis, aneurysmal dilatation, and abnormal
course, can be evaluated with MDCT.
149,150
Abnormalities in PA
architecture are often the result of prior palliative repairs and shunt-
related complications. The PA architecture should be inspected in pa-
tients with repaired TOF and with particular scrutiny in those with
prior palliative repair.
As the use of catheter-based interventions for treatment of branch
PAstenosis or hypoplasia increases, about 22%of patients may develop
further complications, including intimal hyperplasia, stent fracture,
dissection, or aneurysmformation, requiring reinterventionor close sur-
veillance.
151
MDCT provides accurate assessment of the anatomy of
the stented segments as well as surrounding anatomic structures
(Figure 13). To improve visualization of the stented segment, selection
of a hard kernel before reconstruction of the source image set may
result in increased spatial resolution and edge detection. However,
the cost for improved spatial resolution is greater image noise. In
contrast to the routine evaluation of the coronary artery stent, the larger
stents encountered in patients with repaired TOF do not obscure visu-
alization of the lumen, nor do the stent struts cause signicant beam-
hardening artifacts. Therefore, harder reconstruction kernels to assess
stent patency have not been routinely used in this population.
Table 5 Selected normal values of chamber size and function by CMR
Parameter
Study
Alfakih et al.
191
Hudsmith et al.
192
Sarikouch et al.
118
Robbers-Visser et al.
194
Buechel et al.
195
Age (y) 43 (2065) 38 (2168) 11.9 (420) 817 11 (0.718)
Sample size 60 108 114 60 50
PMs excluded from blood pool Yes Yes Yes Yes No
Ventricular volumes and function
LV EDV (mL/m
2
)
Male 82.3 6 14.7 82 6 13 85.1 6 13.8 79 6 11 77.5 BSA
1.38
Female 77.7 6 10.8 78 6 12 77.9 6 10.8 71 6 8 67.8 BSA
1.38
LV EF (%)
Male 64.2 6 4.6 69 6 6 64.4 6 4.9 69 6 5 61.3 6 4.1
Female 64.0 6 4.0 69 6 6 63.4 6 6.1 69 6 5
RV EDV (mL/m
2
)
Male 86.2 6 14.1 96 6 15 84.5 612.7 86 6 12 83.8 BSA
1.47
Female 75.2 6 13.8 84 6 17 76.9 6 12.7 73 6 9 72.7 BSA
1.47
RV EF (%)
Male 55.1 6 3.7 59 6 6 61.6 6 4.5 65 6 5 58.2 6 3.6
Female 59.8 6 5.0 63 6 5 62.8 6 4.3 65 6 5
Atrial volumes Sarikouch et al.
193,
*
LA volume (mL/m
2
)
Male 46.7 6 10.1
Female 44.2 6 8.7
RA volume (mL/m
2
)
Male 58.1 6 15.7
Female 53.3 6 11.8
EDV, End-diastolic volume; ESV, end-systolic volume; LA, left atrial; PM, papillary muscle.
Data are expressed as mean (range) or as mean 6 SD.
*Sample size n = 115; mean age, 12.4 years (range, 4.420.3 years).
Journal of the American Society of Echocardiography
Volume 27 Number 2
Valente et al 129
RV and LV Size and Function. Biventricular volumes are assessed
by retrospective reconstruction of the initial data set acquired during
electrocardiographically gated MDCT at multiple phases of the car-
diac cycle. The computed tomographic data volume is then recon-
structed at 5% or 10% intervals from 5% to 95% of the R-R
interval, providing a multiphase data set to obtain end-diastolic and
end-systolic volumes of the left and right ventricles. From this, stroke
volume, cardiac output, LV EF, and RV EF are calculated.
152
Dedicated semiautomated software is available to assist in the deter-
mination of ventricular volumes and function. However, because of
the complex geometry of the right ventricle, these semiautomated
techniques are often not reliable. CMR remains the gold standard
in the assessment of RV function in adults with CHD, but MDCT
accurately quanties RV measurements and demonstrates close cor-
relation to CMR.
153,154
Therefore, MDCT offers an alternative
imaging strategy in patients with absolute contraindications to CMR.
Aortic Valve, Aortic Root, and Ascending Aorta. Utilization of
cardiac CT has been well described in the assessment of diseases of
the aorta.
155
In addition to other conotruncal anomalies, moderate
dilatation of the ascending aorta is also encountered in adults with re-
paired TOF. Rarely, reports of dissection have occurred in adults with
repaired TOF and massive aortic dilatation.
156,157
Although there are
no established guidelines for serial evaluation of the ascending
thoracic aorta to determine risk for progressive aortic root dilatation
or dissection, these case reports highlight the need for meticulous
follow-up and evaluation of the aortic root after repair of TOF.
Therefore, a comprehensive multidetector computed tomographic
examination in patients with repaired TOF should include at the very
least examination of the ascending thoracic aorta to obtain measure-
ments of size. The scan protocol should include electrocardiographic
gating to provide accurate measurements of the aorta and avoid
misdiagnosis of aortic dissection secondary to motion artifact simu-
lating an intimal ap. Measurements of the aorta with MDCTshould
include both luminal and wall thickness perpendicular to the axis of
blood ow.
158
Mao et al.
159
evaluated thoracic aortic diameters in a
large population of patients without known disease to establish
normal values on the basis of age and gender. The interobserver, intra-
observer, and repeated measurement correlations for measurements
of the aorta (wall thickness and lumen) were high (r > 0.91, P < .001).
Other Cardiovascular Issues. i. Coronary Arteries.The prog-
nostic role of MDCTas well as its ability to characterize complex cor-
onary artery anatomy has been thoroughly investigated.
160-162
Although there have been several advances in the evaluation of
coronary artery anatomy and atherosclerosis with CMR imaging,
MDCT remains a favorable imaging technique with higher
diagnostic performance and accuracy in the evaluation of coronary
artery stenosis.
163
Still, in patients with repaired TOF, the goal of cor-
onary artery imaging usually centers on the origins, proximal courses,
and spatial relationships to neighboring structures.
Preoperative assessment with MDCT provides a noninvasive
method to evaluate coronary artery anatomy before subsequent re-
operation (Figure 14). This application provides a detailed assessment
of the origin, course, and anatomy of the coronary arteries as well as
the 3D relationship of the coronary arteries to the great vessels that is
often helpful during the preoperative assessment of adults with
CHD.
164
Although there are no established guidelines for such preop-
erative multidetector computed tomographic evaluation, this tech-
nique may identify concerning ndings that can inuence surgical
planning (Figure 15).
Noncardiac Findings. MDCT can provide additional insight into
abnormalities of the airways, lung parenchyma, pericardial and
pleural spaces, and mediastinum. The modality is particularly sensitive
to the presence of adenopathy, other thoracic masses, and abnormal-
ities of the lung parenchyma such as emphysema.
d. Standard Protocol
Cardiac MDCT in patients with repaired TOF requires particular
attention to detail in addition to a comprehensive review of medical
and surgical history. Insufcient prescan planning and inaccurate
contrast administration during MDCTcan lead to inappropriate expo-
sures to nephrotoxic contrast and ionizing radiation. To avoidpotential
errors at the time of cardiac MDCT, the protocol should not only be
tailored to each case but also reviewed with the referring physician to
ensure that the clinicians needs are addressed by the scan protocol.
Patient Preparation. Upon patient arrival, a review of study indi-
cation, prior medical and surgical records, and screening for contrain-
dications should be performed. After placement of an 18-gauge or
20-gauge intravenous cannula, heart rate control should be initiated
according to laboratory protocol. Oral b-blockers are often prescribed
before the examination to ensure adequate heart rate control. If addi-
tional heart rate control is required at the time of examination, this
can be achieved through intravenous administration of b-blockers
or calcium channel blockers.
Figure 13 CT of the RVOT and PAs in the sagittal (A) and axial (B) planes showing unobstructed stents. Ao, Aorta.
130 Valente et al Journal of the American Society of Echocardiography
February 2014
Scanning Protocol. The following is an example of cardiac MDCT
using a 64-slice scanner: 0.6-mm collimation, tube voltage (80120
kVp), and tube currenttime product adjusted to body mass index
with electrocardiographic dose modulation to minimize radiation
exposure.
165
In most patients, the scan coverage is from the carina
to the base of the heart (1 cm below the diaphragm). However, if
there is any concern for abnormalities of the thoracic aorta, the
anatomic coverage should be extended superiorly to adequately
cover the brachiocephalic vessels. A dual-head power injector is
used to inject contrast through a peripheral intravenous cannula dur-
ing an inspiratory breath-hold (1520 sec long) and retrospective
electrocardiographic gating. The dose of contrast varies according to
body size and scanning protocol. This is then followed by a saline ush
(60 mL) at 4 mL/sec to facilitate contrast delivery and diminish beam-
hardening contrast artifact in the superior vena cava. However, each
cardiac multidetector computed tomographic protocol is tailored
accordingly to the clinical questions asked at the time of examination.
If the clinical question is to evaluate branch PAs, after interven-
tional therapy of the PAs, and/or RV size and function, the study
should then be optimized to opacify the structures of interest.
Therefore, a timing bolus optimized to the MPA at the level of the ca-
rina using 110 Hounseld units for initiation of the scan will optimize
contrast opacication of these structures.
Alternatively, if the clinical question is to evaluate LV size and sys-
tolic function, coronary artery anatomy, and/or aortic root anatomy,
the study should be optimized to the left-sided structures. Here, the
region of interest for the timing bolus should be located in the
ascending aorta at the level of the carina using 110 Hounseld units
or initiation of the scan. Again, the scanning distance should be
extended superiorly to the brachiocephalic vessels if there is concern
for abnormalities of the aortic root.
To obtain anatomic information regarding both the left and right
heart during a single cardiac multidetector computed tomographic
examination, regions of interest should be placed in both the aorta
and the MPA. The test bolus is then performed to determine the
intersection of maximum Hounseld units concentration in both
great vessels to provide adequate opacication of both circulations.
Using this type of a protocol, information regarding both PA and
aortic architecture is available in addition to biventricular volumes
and function.
Scan Reconstruction. To optimize 3D imaging, scans are then re-
constructed with a slice thickness of 0.6 mmand a slice overlap of 0.6
mm. The postprocessing protocol for calculation of ventricular vol-
umes and EFs includes reconstruction of multiphase short-axis images
in 5% to 10% increments of the R-R interval (5%95%).
e. Recommendations
We recommend that given the young age of this population, every
effort should be directed toward radiation reduction strategies imple-
mented during the time of examination. The most effective technique
to reduce radiation dose in young adults with CHD is to avoid expo-
sure. Therefore, CMR should remain the imaging modality of choice
in patients with repaired TOF, and cardiac MDCTshould be reserved
for patients with absolute contraindications to CMR, particularly
given the need for serial examinations.
8. NUCLEAR SCINTIGRAPHY
a. Overview of Modality
Nuclear cardiology is a physiologic diagnostic modality that uses
radioisotope-labeled compounds to capture and quantify blood
ow, myocardial perfusion and integrity, and ventricular function.
Historically, radionuclide imaging has been used for the evaluation of
ventricular function, pulmonary perfusion, andquanticationof cardiac
shunts in patients with CHD, including repaired TOF. However, many
of these indications are now more readily evaluated with CMR, which
provides superior anatomic delineation and does not require the radia-
tion exposure inherent in radionuclide imaging. In patients who are un-
able to undergo CMR, nuclear scintigraphy is a reasonable alternative
for quantication of differential PA blood ow. In highly selected cir-
cumstances, nuclear scintigraphy can be used for assessment of ventric-
ular function and myocardial perfusion and viability in this population.
b. Strengths and Limitations
Nuclear scintigraphy is a robust technique for measuring differential
pulmonary blood ow.
166
In the presence of multiple sources of blood
supply to the lungs, however, this technique is not accurate. The role of
this modality is well established in the evaluation of LV perfusion and
viability.
167
However, several technical limitations hinder its application
to the assessment of the RVmyocardium. Compared with other nonin-
vasive imaging modalities, nuclear scintigraphy is not considered a
robust technique for assessment of RV size, function, regional wall mo-
tion abnormalities, or anatomic anomalies. Importantly, nuclear scintig-
raphy exposes patients to ionizing radiation with its attendant risk for
cancer. Given that many patients with repaired TOF require multiple
evaluations, the radiation risk increases with each exposure.
c. Assessment of Repaired TOF with Nuclear Angiography
PAs. Measurement of pulmonary perfusion is animportant part of the
evaluation in patients with repaired TOF and suspected branch PA ste-
nosis. Up to 20% of these patients have been reported to develop sig-
nicant branch PA stenosis.
168
Regional perfusion differences within
Figure 14 Computed tomographic 3D reconstruction demon-
strating coronary artery anatomy in a patient with repaired
TOF. Ao, Aorta; LAD, left anterior descending coronary artery;
RCA, right coronary artery.
Journal of the American Society of Echocardiography
Volume 27 Number 2
Valente et al 131
the lungs are alsocommonandmay be a result of the presence of native
branchstenosis, PAobstructionafter a surgical shunt, segmental or lobar
pulmonary vascular disease, or associated arborization anomalies with
large aortopulmonary collateral vessels.
169
Lung perfusion scintigraphy
uses
99m
Tc-labeled macroaggregates of albumin that embolize to the
capillary level and provide an accurate picture of the differential lung
perfusion ratio.
170,171
In patients with repaired TOF, measurements of
differential pulmonary blood ow and regional lung perfusion have a
well-established record before and after treatment of branch PA steno-
sis.
166,171-174
Right Ventricle. i. Size and Function.Radionuclide angiography
can provide quantitative assessment of RV EF. It is a noninvasive test,
and the results can be used to follow ventricular function over
time.
175-177
This method of assessing RV function, however,
provides limited morphologic information. It is used sparingly for
this indication and almost exclusively in patients who have
contraindications to CMR and cardiac CT.
ii. Viability.There is very little role for myocardial perfusion scintig-
raphy in children with repaired TOF. However, a growing number
of patients with repaired TOF reach older age and thus become
more likely to develop coronary artery disease. Anecdotal reports
on this subject have been published.
178
Therefore, evaluation of
myocardial perfusion and viability in these patients by nuclear tech-
niques is considered on a case-by-case basis.
Noncardiac Findings. In addition to the indications noted above,
a ventilation scan can be performed in conjunction with a lung perfu-
sion scan. This test allows quantitative assessment of ventilation-
perfusion mismatch.
d. Standard Protocol
Standard protocols for performance of pulmonary perfusion scintig-
raphy, radionuclide ventriculography, and myocardial perfusion scin-
tigraphy in children and adults are published and updated by the
American College of Radiology, the Society for Pediatric Radiology,
and the Society of Nuclear Medicine and Molecular Imaging and
are displayed on their Web sites.
179,180
e. Recommendations
We recommend the use of radionuclide scintigraphy for evaluation of
differential pulmonary blood ow and regional lung perfusion when
the information cannot be obtained by CMR. The modality may also
be used in highly selected cases when other modalities are not prac-
tical for assessment of ventricular volume and function and myocar-
dial perfusion and viability.
9. X-RAY ANGIOGRAPHY
a. Overview of Modality
Diagnostic cardiac catheterization with x-ray angiography is rarely
used primarily for diagnostic imaging purposes in patients with re-
paired TOF. However, it serves an important role when essential infor-
mation cannot be accurately obtained noninvasively, particularly for
those in whom echocardiographic data are inconclusive or contradic-
tory and CMR is contraindicated. Additionally, x-ray angiography is
an integral component of catheter-based procedures such as PA
balloon dilation and stenting, percutaneous PVimplantation, manage-
ment of aortopulmonary collateral vessels, closure of residual septal
defects, and coronary artery interventions.
133,181,182
A complete right and left heart catheterization is performed
before angiography and allows evaluation of the presence of resid-
ual shunt by oximetry, diastolic dysfunction by measurements of
lling pressures, stenosis by pullback gradients, and PR by loss of
the dicrotic notch in the PA tracings and equilibration of diastolic
pressures between the MPA and right ventricle. In addition, cardiac
index is estimated by oximetry or thermodilution, and pulmonary
and systemic vascular resistances are calculated. In the presence of
signicant TR and/or PR and when a residual shunt is present, car-
diac output is better calculated by oximetry using the Fick princi-
ple. In the presence of proximal branch PA stenosis, pulmonary
vascular resistance can be calculated using the Fick principle and
knowledge of differential PA blood ow from a lung perfusion
scan or CMR.
183
b. Strengths and Limitations
Diagnostic cardiac catheterization with x-ray angiography is considered
the gold-standardmodality for assessment of intracardiac pressures, cor-
onary arteries beyond their origins, and branch PAs. Important limita-
tions of the modality include its invasive nature and associated
morbidity, high cost, and exposure to ionizing radiation. For these
Figure 15 CT in a patient with repaired TOF and severe PR who is a candidate for PV replacement. (A) Note pacemaker lead in the
right ventricle (arrow) and proximity of the RVOT to the sternum. (B) Note patent foramen ovale (arrowhead). LA, Left atrium; RA, right
atrium.
132 Valente et al Journal of the American Society of Echocardiography
February 2014
reasons, the modality is used solely for diagnosis only in carefully
selected cases. Given that the number of patients with repaired TOF
reaching the age range when coronary artery disease is prevalent, the
role of diagnostic imaging of the coronary arteries will likely increase.
c. Assessment of Repaired TOF with X-Ray Angiography
RVOT. In addition to biplane angiography, rotational angiography is
a relatively new technology that is increasingly being used for assess-
ment of patients with CHD. In patients with repaired TOF, this tech-
nique is particularly useful for visualizing proximal bifurcation PA
stenosis and RVOT anatomy, especially in the presence of a
conduit-associated branch PA stenosis (Figure 16).
184,185
Rotational angiography is performed over 200
using a large-format,
digital, at-panel angiographic system. Images are acquired with low
radiation doses using a slow injection of a small amount of contrast
media. Theacquisitioncanbe performedunder rapidventricular pacing
to slow the contrast washout and reduce motion artifacts. Three-
dimensional angiographic images are generated on a workstation in
near real time during the procedure. Different acquisition and display
parameters can be used. For example, Glatz et al.
185
used 190
rotation,
nongated 5-sec acquisition time, projection on either 20 20 cmor 30
40 cm at-panel detector size, and 48-cm eld of view. They dis-
played the tomographic volume set at a slice thickness of 0.3 mm,
although this can be variable. Breath holding is used when possible,
and radiation dose is adjusted to minimize exposure while maintaining
adequate image quality. The intravenous contrast agent is diluted with
2 mL saline for each milliliter of contrast, for a total volume of approx-
imately 1.5 mL/kg per injection. The reconstructed 3Dimage can then
be projected as an overlay on the live uoroscopy monitor, which
facilitates complex catheter manipulations. It can also be useful for
dening optimal camera angles for standard biplane angiography or
uoroscopic guidance. The main advantage is that it provides a more
precise delineation of the 3D anatomy of vascular structures outside
the heart with rapid image processing and reduction of the amount
of contrast and ionizing radiation used compared with those of multiple
biplane angiograms in various projections.
186
PAs. Depending on the ndings at the time of hemodynamic assess-
ment and on the results of prior imaging studies, pulmonary angiog-
raphy may be performed selectively in each branch or in the
MPA.
187
To evaluate the PA bifurcation and proximal branch PAs
the AP camera is positioned as perpendicular as possible to the bifur-
cation, which is achieved with either cranial angulation (20
30
) or
extreme caudal angulation (35
45
of caudal angulation in
the lateral camera can be applied. In patients with arborization abnor-
malities, selective distal branch or segmental pulmonary angiography
is recommended with variable angulation.
PR. There is no standard angiographic classication of PR severity.
Patients with signicant PR will have opacication of the right
ventricle during pulmonary angiography. However, the presence of
a catheter crossing the PV tends to articially enhance the degree
of regurgitation. Increased PA pulsatility with signicant change in
vessel dimension from systole to diastole is seen in patients with se-
vere PR. Similarly, early diastolic pressure equilibration in the central
PAs and the RV is indicative of signicant PR.
Right Ventricle. i. Size.Although right heart size is poorly as-
sessed by x-ray angiography, it provides a gross qualitative estimate.
To best visualize the RVOT, a mild degree (15
20
) of cranial angu-
lation is used in the AP camera.
187,188
ii. Function.Assessment of RVfunction is only qualitative given that
its complex shape limits the ability to use geometric models to esti-
mate volumes from biplane images.
Figure 16 Example of rotational angiography performed in the
right ventricle (RV) after repair of TOF demonstrates severe bilat-
eral branch PA stenosis. On the RPA (asterisk), a stent was im-
planted proximally, and there is severe stenosis at the distal
end of the stent (arrow). Image courtesy of Dr MatthewGillespie,
Childrens Hospital of Philadelphia.
Figure 17 MPA angiography performed with 40
of caudal
angulation outlines the PA bifurcation in detail. The homograft
PV is seen en face (black arrow), there is no signicant steno-
sis in the LPA(white arrow), and there is moderate stenosis in the
RPA (asterisk).
Journal of the American Society of Echocardiography
Volume 27 Number 2
Valente et al 133
Aortic Valve, Root, and Ascending Aorta. Aortic root angiog-
raphy is standard in patients with repaired TOF. This allows demon-
stration of arch sidedness, aortic root and ascending aorta
dimensions, coronary origins and course, AR, presence of aortopul-
monary collateral vessels, or residual ow through prior surgical
shunts.
Other Cardiovascular Issues. i. Coronary Arteries.Selective
coronary angiography is recommended in most patients with repaired
TOF unless the vessels are clearly delineated by aortography. Patients
with abnormal anatomic patterns that may have been altered at the
time of prior surgical intervention require selective coronary angiog-
raphy. Conal branches may have been disrupted and retrograde lling
by collateral vessels can be seen in this setting. Detailed anatomic de-
nition of the proximal course is needed in all patients who are being
considered for percutaneous PV implantation or RV-to-PA conduit
stenting. When in doubt, selective coronary angiography during
simultaneous balloon dilation of the RVOT is performed to determine
the risk for coronary compression with conduit expansion
(Figure 18).
189,190
ii. Aortopulmonary Collateral Vessels.Once the collateral vessels
are identied by aortography selective angiography in straight AP and
lateral projections is performed. The levophase allows demonstration
of the pulmonary venous return, the presence of any atrial-level shunt,
and qualitative evaluation of LV function. In patients with discontinuity
or atresia of a PA branch, anatomic denition of the atretic vessel is
needed for surgical planning. If the vessel does not ll via aortopulmo-
nary collateral ow, a pulmonary venous wedge injection is performed.
This is best done in the lower lobe pulmonary vein.
187,188
Noncardiac Findings. The relationship of the PAs to the airway
can be appreciated on biplane angiography. Motion of the dia-
phragms can be assessed on uoroscopy with the patient breathing
spontaneously. Three-dimensional rotational angiography provides
useful information on surrounding soft tissues in the tomographic
volume sets that are generated. This may be relevant when
associated vascular compression of the airway or potential erosion
is suspected.
d. Standard Protocol
Patient Preparation. Before cardiac catheterization, patients un-
dergo a thorough noninvasive evaluation to ensure an indication for
x-ray angiography and rene the clinical questions for the test. This in-
cludes a thorough reviewof the patients medical and surgical history,
including any prior cardiac catheterizations, as well as other imaging
studies performed. The sedation strategy for the procedure is then
determined on the basis of the patients hemodynamic, airway, and
respiratory status as well as the nature of the catheterization proce-
dure. General anesthesia is usually required when planning complex
catheter interventions such as multiple PA balloon angioplasties or
percutaneous PV implantation. Patients on anticoagulation medica-
tions should have these discontinued before the procedure, and those
receiving warfarin may need to transition to heparin or enoxaparin.
In a stable patient undergoing diagnostic catheterization, it is
possible to use minimal sedation and local anesthesia. However, the
frequency of these procedures has declined signicantly as most car-
diac catheterizations now include a therapeutic intervention. All pa-
tients receive peripheral intravenous cannulas for hydration and
administration of medications and are attached to cardiorespiratory
monitors. Vascular access is obtained percutaneously using sterile
technique. As mentioned earlier, it is standard to perform angiog-
raphy in the biplane projection using a power injector at a rate of 1
to 2 mL/kg/sec (generally up to 40 mL).
187
e. Recommendations
We recommend x-ray angiography for the assessment of critically ill
patients or when noninvasive methods either cannot be performed
or have failed to provide satisfactory data. Selective coronary angiog-
raphy should be considered before cardiac and noncardiac proce-
dures in adult patients at risk for coronary artery disease.
10. MULTIMODALITY APPROACH
The discussion of imaging modalities in the previous sections demon-
strates that no single test is capable of providing all necessary
Figure 18 Coronary angiography performed on same patient as in Figure 17, during balloon compliance testing of a right ventricle
toPA homograft. The patient has repaired TOF and single left coronary artery (white arrow) with right coronary artery (RCA) crossing
the RVOT. Angiography is performed in the AP projection with extreme caudal angulation. (A) At the start of the balloon ination, there
is no evidence of obstruction of coronary ow (black arrow). (B) With the balloon inated, further there is complete occlusion of RCA
ow. Percutaneous PV implantation is therefore contraindicated in this patient.
134 Valente et al Journal of the American Society of Echocardiography
February 2014
Table 6 Recommended multimodality imaging surveillance frequency in repaired TOF
Modality
Age (y)
<2 y* 29 1019 2049 $50
Echocardiography 12 mo 12 mo 24 mo 24 mo 24 mo
CMR Not recommended routinely; ordered to address
specic questions not answered by echocardiography
36 mo in stable patients
12 mo if moderate ($150 mL/m
2
) or
progressive (increase of >25 mL/m
2
) RV dilatation or
dysfunction (RV EF # 48% or $6% decrease in EF)
CT Not recommended routinely; ordered when
CMR is indicated but cannot be performed
(e.g., metallic artifacts or contraindications to CMR)
Lung perfusion scan If predicted RV systolic pressure 60% systemic
or smallest branch PA diameter Z score
< 2.5; in patients $10 y of age,
consider CMR ow measurements
X-ray angiography Not recommended routinely
; ordered when
noninvasive methods either cannot be performed
or have failed to provide satisfactory diagnostic data
Coronary angiography when
clinically indicated
Chest radiography Not recommended routinely; may be ordered
for evaluation of stent integrity
*After the rst postoperative outpatient evaluation.
22.51
14.39
13.59
jj
Radiation risk ++++ ++++
Artifacts from stainless-steel implants + +++ +
Sedation requirements in young children ++ ++++ +++ ++
Spatial resolution (mm) <1 <12 <1 510
Temporal resolution (msec) 20 30 75175
{
RV size/function ++ ++++ +++ +
RV pressure +++ + +
TR severity +++ +++
Mechanism of TR ++++ ++
PR severity ++ ++++
Branch PAs ow quantication +++ ++++
LV size/function +++ ++++ +++ ++
Coronary origins and proximal course ++ +++ ++++
Aortic dimensions +++ ++++ ++++
Residual shunts +++ +++ +
Pulmonary-to-systemic ow ratio + ++++ +
#
+
Aortopulmonary collateral vessels +++ +++
Myocardial viability + ++++ + +++
Modied from Prakash et al.
9
*FromCenters for Medicare andMedicaidServices, National Physician Fee Schedule(http://www.cms.hhs.gov/PFSlookup/03_PFS_Document.asp).
Sum of relative value units for Current Procedural Terminology (CPT) codes 93303, 93325 and 93320.