Original Article

Computed tomographic versus

catheterization angiography
in tetralogy of Fallot

Asian Cardiovascular & Thoracic Annals

2015, Vol. 23(2) 164175
The Author(s) 2014
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0218492314538844
aan.sagepub.com

Naveen Garg1, Rohit Walia2, Zafar Neyaz2 and Sunil Kumar2

Abstract
Aim: To compare multidetector computed tomographic angiography with the gold standard cardiac catheterization and
angiography in tetralogy of Fallot.
Methods: In 40 consecutive patients over 5 years of age with tetralogy of Fallot, multidetector computed tomographic
angiography and catheterization angiography studies were compared for intracardiac anatomy, pulmonary anatomy and
indices, coronaries and collaterals. Safety parameters, relative advantages and limitations were also analyzed.
Results: All catheterization studies required hospitalization whereas all tomographic studies were performed as outpatient procedures. The need for sedation and amount of contrast used were significantly greater in catheterization than
in tomographic studies. Complications noted during catheterization were access site complications in 4 patients, cyanotic
spells in 2, transient complete heart block requiring temporary pacing in 2, and air embolism in one. No complication was
observed during tomographic studies. All tomographic studies were adequate, but 2 catheterization studies were
inadequate.
Ventricular septal defects, aortic override, level of right ventricular outflow tract obstruction, and pulmonary artery
anatomy were equally assessed by both imaging modalities. However, tomographic studies missed additional small
muscular ventricular septal defects. There was a linear correlation between tomographic and catheterization studies
for pulmonary annulus size, artery sizes, Z-score, and Nakata index. There was complete concordance with respect to
side of aortic arch and detection of collaterals. Coronary anatomy was better delineated in tomographic studies.
Conclusions: For preoperative evaluation of tetralogy of Fallot patients, multidetector computed tomographic angiography can be used as a reliable noninvasive alternative to cardiac catheterization angiography.

Keywords
Coronary angiography, heart septal defects, ventricular, image processing, computer-assisted, pulmonary artery, tetralogy
of Fallot, tomography, X-Ray computed

Introduction
Patients with tetralogy of Fallot (TOF) are routinely
subjected to cardiac catheterization and angiography
(CCA) for preoperative evaluation because of suboptimal information obtained from echocardiography
regarding pulmonary artery anatomy, origin and
course of coronary arteries, and major aortopulmonary
collateral arteries (MAPCA). All of these can be eectively shown with the present generation of multidetector computed tomography angiography (MDCTA),
safely and noninvasively.16 Thus routine cardiac catheterization in patients with TOF may be unwarranted.

Direct comparison of MDCTA and CCA in patients

with TOF has not yet been reported. Therefore, this
study was planned with the aim to compare MDCTA
with CCA for preoperative diagnostic evaluation as
1
Department of Cardiology, Sanjay Gandhi Postgraduate Institute of
Medical Sciences, Lucknow, India
2
Department of Radiology, Sanjay Gandhi Postgraduate Institute of
Medical Sciences, Lucknow, India

Corresponding author:
Naveen Garg, Department of Cardiology, Sanjay Gandhi PGIMS,
Raibareli Road, Lucknow, India.
Email: navgarg@sgpgi.ac.in

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015

Garg et al.

165

well as their relative advantages and limitations in

patients with TOF, with special emphasis on determining whether MDCTA can provide a noninvasive alternative to diagnostic catheterization in pre-surgical
evaluation of TOF patients.

Patients and methods

We prospectively included consecutive patients with
TOF presenting to our tertiary care hospital for presurgical evaluation over the last 2 years. Patients aged
less than 5 years, TOF with pulmonary atresia, TOF
with atrioventricular septal defect (VSD), relative hypoplasia of any ventricle, previous palliative cardiac surgery, deranged renal parameters, and those unt or
unwilling to undergo surgical repair or critically ill
were excluded. A detailed echocardiographic examination was undertaken in all patients. Patients fullling
the inclusion criteria were subjected to MDCTA followed by CCA. CCA was carried out at least 2 weeks
after but within 4 weeks of MDCTA. The study
was approved by the institutional ethics committee.
Informed consent was obtained in all cases prior to
inclusion in the study.
All patients underwent a detailed clinical evaluation,
electrocardiogram, chest radiography, hemogram,
renal and liver function tests, and echocardiography.
Scans were performed on a 128-MDCT scanner
(Somatom Denition AS; Siemens, Forchheim,
Germany). The radiation dose was kept to a minimum
by reducing the kilovoltage and tube current appropriately. All studies were carried out after 6 h of fasting
(Table 1). All MDCTA investigations were performed
under the care of a trained anesthetist in close collaboration with a pediatric cardiologist and a radiologist.

Sedation using intravenous midazolam was given when

required. Patients were encouraged to hold their breath
during the scan. A noncontrast computed tomography
scan was followed by scanning in the arterial phase; a
second phase was taken 610 s later to rule out systemic
venous anomalies. Electrocardiogram (ECG) gating
was not used for the following reasons: it increases
the radiation dose signicantly; it signicantly prolongs
the scanning time, leading to more respiratory artefacts;
the high heart rate in infants makes it dicult to obtain
adequate high-resolution images; and most of the extracardiac anatomy can be adequately depicted without
ECG gating. Even though the distal coronary arteries
cannot be visualized without ECG gating, non-gated
MDCTA images are sucient to identify the origin
and proximal course of the major coronary arteries,
which is adequate for management decisions in these
patients. The goal of contrast medium injection was to
achieve homogeneous vascular enhancement synchronized with image acquisition. To this end, great care
was taken regarding the intravenous access, dose and
density of contrast material, and rate of injection.7 For
intravenous access, a leg vein was preferred to avoid
streak artefacts from high-density contrast material in
the upper limb veins. Using the bolus-tracking technique, scanning was started manually after seeing the
contrast medium in the right ventricle. A minimum
delay of 4 s was used to start the scan after the contrast
medium reached the right ventricle. Because of a VSD,
homogeneous contrast enhancement of the pulmonary
and systemic arteries can be achieved consistently using
this method. Computed tomography was repeated
610 s after the rst phase to look for systemic venous
anomalies. Data acquisition was performed from the
thoracic inlet level to L1L2. When there was suspicion

Table 1. Multidetector computed tomography angiography protocol.

Extent of scanning
Scanner settings

Contrast material

ECG gating
Patient instruction
Scanning delay

Thoracic inlet to L1L2 level

kV for <20 kg; 80 kV for 2030 kg; 100120 kV for >30 kg
mAs: minimum based on fully automated real-time anatomy-based dose regulation (CARE Dose 4D
software, Siemens, Germany)
Other parameters: detector collimation 128  0.6 mm, pitch 1.5, slice thickness 1 mm, reconstruction
interval 0.5 mm, gantry rotation time 0.3 s
Venous access: leg vein (preferred)/right upper limb
Nonionic contrast: iopamidol 370 mgmL1 (Bracco, Milan, Italy) at 1.52.0 mLkg1
Rate of injection: 1.52.0 mLs1 for 22-gauge cannula and 3.0 mLs1 for 20-gauge cannula, followed by
10 mL saline chase at the same rate
Using bolus-tracking technique, scanning was started manually after arrival of contrast medium in the right
ventricle
No
Breath hold if possible otherwise free quiet breathing
First phase: 4 s after arrival of contrast in right ventricle
Second phase: 610 s after first phase

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015

166

Asian Cardiovascular & Thoracic Annals 23(2)

of anomalous pulmonary venous drainage, the scan

was extended down to the lower border of the liver.
All image data were evaluated using Syngo
Multimodality Workplace (Siemens, Germany).
Various image reformatting techniques were used to
obtain all clinically relevant information. Curved
planar reformatting and maximum-intensity projection
were primarily used to evaluate curved structures such
as the pulmonary artery (PA) and MAPCA. Minimumintensity projections were used to evaluate the airway
and lung parenchyma. For 3-dimensional reformatting
of complex anatomy, the volume-rendering technique
was used. Thin-section multiplanar reformatting was
used for quantitative analysis of the structure in question.810 All MDCTA data were analyzed by an independent radiologist unaware of the CCA ndings.
Anatomic details evaluated were: right ventricular
(RV) size and location of right ventricular outow
tract (RVOT) obstruction; RVOT anatomy including
RV-to-PA continuity, pulmonary annulus, main pulmonary artery (MPA), PA conuence, right pulmonary
artery (RPA) and left pulmonary artery (LPA), and
peripheral arborization pattern; the size of pulmonary
annulus and branch PA (any peripheral PA stenosis
was also noted); left ventricular (LV) size, location
and size of the VSD, aortic override and any additional
VSD; coronary artery anatomy including the origins
and any major coronary artery or a branch crossing
over the RVOT; aortic arch anatomy and branching
pattern; MAPCA from any part of the aorta or its
branches; any systemic or pulmonary venous anomaly;
Nakata index, predicted ratio of peak RV pressure to
peak LV pressure (PRV/LV), and age-adjusted
Z-scores for the pulmonary annulus and PA. The
Nakata index was calculated as the sum of the crosssectional areas of the right and left pulmonary arteries/
body surface area (RPA and LPA sizes were measured
immediately proximal to the origin of the upper lobe
branches).11 The PRV/LV ratio was calculated as
0.484/(RPA diameter LPA diameter)/descending
thoracic aorta diameter at diaphragm 0.2007.12 Ageadjusted Z-scores for the pulmonary annulus and RPA
and LPA were calculated using the Detroit data.13
Angiograms were performed under local anesthesia.
The patients were adequately hydrated before and after
the procedure, and renal function tests were carried out
before and after each procedure. Intravenous midazolam was given when sedation required. Vascular
accesses were taken from the femoral vein and femoral
artery. Pressure and oxygen saturations in dierent cardiac chambers, the aorta, and venae cavae were
obtained using uid-lled catheters. Contrast angiography was undertaken to delineate intracardiac and
extracardiac anatomy. RV angiograms in anteroposterior view with cranial angulation and in left anterior

oblique view were obtained to delineate RVOT and

PA anatomy. Levo-phase was used to delineate pulmonary venous drainage. An LV angiogram in left
anterior oblique view with cranial angulation was
obtained to evaluate LV size and function, location of
the VSD, presence of any additional VSD, and degree
of aortic override. Aortic root angiography was undertaken when any abnormality was detected by echocardiography and/or MDCTA to delineate the arch
anatomy and branching pattern. Selective right coronary angiography was used to conrm its origin and any
branch crossing over the RVOT. A thoracic aortogram
with selective injection in the right brachiocephalic
trunk and left subclavian artery was performed in all
patients to detect any collaterals. If any collateral was
seen, selective injection into the collateral was also
undertaken. Left innominate vein injection was used
to delineate systemic venous drainage. All CCA data
were analyzed by an independent cardiologist unaware
of the MDCTA ndings. The Nakata index, PRV/LV
ratio, and age-adjusted Z-scores for the pulmonary
annulus and PA were calculated in a similar way as
for MDCTA.1113
Intracardiac and extracardiac anatomy as delineated
by the two modalities were compared. Pulmonary
annulus and PA sizes (MPA, RPA and LPA), the presence of peripheral pulmonic stenosis, Nakata index,
Z-score, and PRV/LV ratio as measured by the two
modalities were compared. Coronary artery anatomy
and collateral documentation were also compared.
Safety parameters such as the incidence of cyanotic
spells, any major arrhythmia, access site complications
(major bleeding, hematoma, pulse loss), any other complication, and contrast dose were noted for both
modalities.
Continuous variables are presented as mean  standard deviation. Categorical variables are presented as
percentages. Continuous variables were compared
using the paired t test, and categorical variables were
compared using the chi-square test. Regression and correlation analysis of imaging data of MDCTA and CCA
was carried out. Statistical signicance was set at a
2-tailed probability level less than 0.05. Statistical analysis was performed using the SPSS 15.0 statistical
package (SPSS, Inc., Chicago, IL, USA).

Results
The baseline clinical characteristics the 40 patients are
summarized in Table 2. The procedural characteristics
of MDCTA and CCA can be compared in Table 3. All
MDCTA studies were outpatient procedures, whereas
all CCA studies required hospitalization for at least 3
days. Sedation was more frequently required during
CCA, especially in patients less than 15 years of age.

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015

Garg et al.

167

Table 2. Baseline clinical characteristics of 40 patients with

tetralogy of Fallot.
Variable

No. of patients

Male
Age (years) [range]
Height (cm) [range]
Weight (kg) [range]
Body surface area (m2) [range]
Hemoglobin level (gdL1) [range]
NYHA functional class
Class II
Class III
History of cyanotic spells

30 (75%)
11.9  8.0 [529]
123.6  29.9 [85169]
25.7  14.9 [9.553]
0.9  0.5 [0.51.7]
17.1  3.6 [12.122.9]
36 (90%)
4 (10%)
6 (15%)

CCA: cardiac catheterization and angiography; MDCTA: multidetector

computed tomography angiography; New York Heart Association.

Table 3. Procedural characteristics of multidetector computed

tomography angiography and cardiac catheterization and angiography in 40 patients.
Variable

MDCTA

CCA

p value

Requirement of sedation
Procedure duration (min)
Volume of contrast used (mL)
Access site complication
Severe cyanotic spell
during procedure
Transient complete heart
block requiring
temporary pacing
Air embolism
Inadequate studies

12 (30%)
17.3  1.9
44.6  17.4
0
0

22 (55%) <0.001
45.6  7.2 <0.001
97.5  7.7 <0.001
4 (10%)
2 (5%)

2 (5%)

0
0

1 (2.5%)
2 (10%)

CCA: cardiac catheterization and angiography; MDCTA: multidetector

computed tomography angiography.

Signicantly more contrast was used in CCA than in

MDCTA studies. No patient had any contrast-related
problem throughout the study period. Mean laboratory
time was signicantly longer for CCA than for
MDCTA. The mean radiation dose in MDCTA was
3.3  4.2 mSv. Access site complications occurred in 4
(10%) patients during CCA (groin hematoma in 2, and
transient loss of femoral pulse in 2). Other complications noted during catheterization were severe cyanotic
spells in 2 patients, transient complete heart block
requiring temporary pacing in 2, and air embolism
due to rupture of the balloon of the Berman angiographic catheter in one; all were managed successfully.
No patient had any complication during MDCTA.
Only one patient had motion artefacts and only a few

chest images were inadequate in MDCTA studies, but

pulmonary and intracardiac information was adequate
in all patients. Two CCA studies were inadequate despite best eorts; both patients had severe RV dysfunction and severe tricuspid regurgitation. In these
patients, RV catheter positioning was very dicult
because it recoiled back. It took extensive and prolonged catheter manipulation and a large amount of
contrast medium and uoroscopy time with suboptimal
PA delineation because some of the contrast medium
was regurgitated. MDCTA showed the pulmonary and
other relevant anatomy with ease in both of these
patients. The LV was very well and equally assessed
by both imaging modalities. Both MDCTA and CCA
equally depicted the size and location of the perimembranous VSD (Figure 1). Additional muscular VSD
were present in 4 patients, as detected by CCA
(mid-muscular VSD in 2, mid-muscular as well as
apical muscular VSD in 2). MDCTA missed an additional muscular VSD in 2 patients (mid-muscular and
apical muscular VSD in 1 patient each). The degree of
aortic override was accurately assessed by both imaging
modalities (Figure 1). The RV was also very well
assessed by both techniques (Figure 2). Location of
hypertrophied muscle bundles, level of RVOT obstruction (infundibular, subinfundibular, valvular, supravalvular, or multilevel) were very well and equally
assessed by both modalities. Pulmonary annulus diameter measured by both modalities was similar. PA anatomy was very well assessed by both MDCTA and CCA
(Figure 3). PA sizes, conuence, and peripheral PA
stenosis were equally well assessed by both imaging
modalities. There was no signicant dierence in the
sizes of the MPA, RPA, and LPA as measured by
both imaging modalities (Table 4). There were 4 cases
of MPA stenosis, 8 of LPA stenosis, 2 of RPA stenosis,
and 2 patients had MPA bifurcation stenosis involving
the distal MPA and the origins of the RPA and LPA;
all were equally diagnosed by both imaging modalities.
Interestingly, 2 patients had aneurysmal dilatation of
the LPA due to post-LPA stenosis, and 1 had absent
LPA (Figure 4), also equally diagnosed by both imaging modalities. The Nakata index, Z-scores and PRV/
LV ratio as calculated by MDCTA and CCA are
shown in Table 4. There was a good correlation
between MDCTA and CCA measurements of main
and branch PA and pulmonary annulus and their
Z-scores. Pearsons correlation coecients for the pulmonary annulus, MPA, RPA, and LPA diameters were
0.996 (p 4 0.001), 0.997 (p 4 0.001), 0.988 (p 4 0.001),
and 0.998 (p 4 0.001), respectively, and these were in
the range of good correlation. Pearsons correlation
coecients for Z-score of the pulmonary annulus,
MPA, RPA, and LPA were 0.988 (p 4 0.001), 0.977
(p 4 0.001), 0.980 (p 4 0.001), and 0.951 (p 4 0.001),

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015

168

Asian Cardiovascular & Thoracic Annals 23(2)

Figure 1. Multidetector computed tomography angiography images profiling ventricular septal defects in 3 patients. (a) Axial oblique
view showing a large perimembranous ventricular septal defect (arrow). (b) Sagittal oblique view showing aortic override with a large
perimembranous ventricular septal defect (arrow) with adequate sized left and right ventricles. (c) Axial view showing an additional
apical muscular ventricular septal defect (arrow). (d) Left ventricular angiogram of the same patient in left anterior oblique view with
cranial angulation, showing a large perimembranous ventricular septal defect with additional mid-muscular (upper arrow) as well as
apical muscular ventricular septal defects (lower arrows). Multidetector computed tomography angiography missed the mid-muscular
defect.

Figure 2. Multidetector computed tomography angiography

image in sagittal oblique view showing right ventricular outflow
tract (RVOT) anatomy; hypertrophied right ventricle,
severe infundibular as well as valvular pulmonic stenosis, and a
good-sized pulmonary annulus and main pulmonary artery.
PV: pulmonary valve.

respectively, and also were in the range of good correlation. In further linear regression analysis, we plotted
graphs that showed a positive linear correlation
between MDCTA and CCA for dierent parameters
of PA size and its indices (Figure 5). There was complete concordance with respect to laterality of the aortic
arch with MDCTA and CCA, but arch branching
anomalies were best shown by MDCTA (Figure 6).
Our cohort of patients had normal origins of the coronary arteries, equally depicted by MDCTA and CCA.
Four cases of a large conal branch crossing over the
RVOT were identied equally by both modalities
(Figure 7). Interestingly, in 2 patients, initial selective
RCA angiograms missed a proximally originating large
conal branch crossing the RVOT, and this was diagnosed by a repeat RCA angiogram after a little pull
back of the catheter. There was complete concordance
with respect to detection of MAPCA (Figure 8), except
in one case in which one MAPCA was missed on
MDCTA; when MDCTA was reviewed again, the collateral was found to be present. In 3 patients initially
presenting with hemoptysis, MDCTA indicated

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015

Garg et al.

169

Figure 3. Multidetector computed tomography angiography images showing complete pulmonary artery anatomy. (a) Coronal
oblique view (maximum-intensity projection) showing the right pulmonary artery with narrowing at its origin (arrow). (b) Sagittal
oblique view (maximum-intensity projection) showing the left pulmonary artery. (c) Axial view (maximum-intensity projection)
showing the pulmonary artery confluence with hypoplasia of the main pulmonary artery and proximal right pulmonary artery.
(d) Volume-rendered image nicely showing the entire pulmonary arterial tree. LPA: left pulmonary artery; MPA: main pulmonary
artery; RPA: right pulmonary artery.

Table 4. Pulmonary artery anatomy and indices in multidetector computed tomography angiography and cardiac catheterization and angiography in 40 patients.
Anatomy

MDCTA

CCA

Pulmonary annulus
Diameter (mm)
13.9  3.8
14.0  4.1
Z-score
1.6  1.4
1.7  1.5
Main pulmonary artery
Diameter (mm)
13.7  5.1
13.7  5.0
Z-score
1.6  2.7
1.8  2.7
Right pulmonary artery
Diameter (mm)
12.2  3.8
12.2  3.8
Z-score
0.3  1.6
0.3  1.5
Left pulmonary artery
Diameter (mm)
12.3  6.4
12.4  6.5
Z-score
1.3  2.1
1.5  2.2
Nakata index
344.4  205.7 347.4  215.5
(mm2m2)
PRV/LV
0.5  0.1
0.5  0.1

p value

0.165
0.812
0.351
0.048

tuberculosis as the cause of hemoptysis. Three cases of

left-sided superior vena cava were seen and equally well
conrmed by both MDCTA and CCA (Figure 6).
Twenty-eight patients underwent surgery: 24 had total
intracardiac repair, and 4 had a modied BlalockTaussig shunt. MDCTA and CCA ndings were conrmed in all patients who underwent intracardiac
repair. Four patients underwent coil embolization of
MAPCA during the CCA study, and intracardiac
repair on the same day (Figure 9).

Discussion
0.871
0.703
0.246
0.147
0.230
0.972

CCA: cardiac catheterization and angiography; MDCTA: multidetector

computed tomography angiography; PRV/LV: predicted peak right ventricular pressure-to-peak left ventricular pressure ratio.

This study has shown that the information provided by

MDCTA is comparable with that provided by the conventional gold standard CCA. It provides good delineation of RVOT and PA anatomy, coronary anatomy,
and MAPCA. These ndings, combined with the relative noninvasiveness and speediness of MDCTA, makes
it a desirable imaging technique for preoperative evaluation of TOF patients, and it can be used as a reliable
noninvasive alternative to CCA. It is worth noting that
the data provided by a single injection of contrast
material could be manipulated and viewed on a

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015

170

Asian Cardiovascular & Thoracic Annals 23(2)

Figure 4. Interesting multidetector computed tomography angiography images of 4 patients. (a) Axial view (maximum-intensity
projection) showing left pulmonary artery origin stenosis (arrow) followed by aneurysmal dilatation; a severely hypoplastic main
pulmonary artery and right pulmonary artery were also noted. (b) Axial view (maximum-intensity projection) showing absent left
pulmonary artery (arrow). (c) Sagittal oblique view (maximum-intensity projection) showing a dilated and dysfunctional right ventricle.
(d) Coronal view (maximum-intensity projection) showing normal pulmonary venous drainage of all 4 pulmonary veins (black arrows)
into the left atrium.

computer workstation in an unlimited number of projections. Retrospective data manipulation allows the
projection that best displays a particular anatomic feature, using interactive rotation and removing structures
such as tortuous collateral vessels that are superimposed on the area of interest in angiograms. In addition, the information gained is easy to communicate
to surgical colleagues, and the data can be retrieved
after the procedure.
Ours is the only study directly comparing MDCTA
with CCA in patients with TOF. Our results are in
accordance with the results of studies comparing echocardiography and MDCTA with the surgical ndings.1,4 Wang and colleagues2 reported the diagnostic
accuracy of MDCTA and echocardiography in TOF as
95.43% and 83.3%, respectively. While intracardiac
anatomy may be elucidated further during preoperative
transesophageal echocardiography, concerns regarding
additional VSD, ventricular adequacy and RVOT, PA
anatomy, coronary artery course, and MAPCA might
remain unresolved. Cineangiography has been traditionally employed to delineate these structures. Our

study shows that MDCTA can resolve these questions

noninvasively, thus restricting preoperative cardiac
catheterization to interventional indications. Table 5
lists various advantages and limitations of both these
imaging modalities.
Evaluation of PA anatomy is the most common
reason for subjecting these patients to invasive angiography. In our study, the sizes of the pulmonary annulus
and arteries measured by MDCTA correlated very well
with CCA data. Peripheral pulmonary stenosis was also
detected equally well by MDCTA and CCA. Dierent
indices predicting surgical outcome (Z-score, Nakata
index, PRV/LV ratio) measured by CCA and
MDCTA also correlated very well. Our results are in
agreement with those of previous studies.3,4 Hayabuchi
and colleagues3 and Kasar and colleagues4 found an
excellent correlation between MDCTA and conventional pulmonary angiography in quantifying the diameter of the main and branch PA. Our results also agree
with a recent study by Meinel and colleagues14 who
showed a very good correlation between MDCTA
and CCA for MAPCA and PA anatomy, as well as a

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015

Garg et al.

171

Figure 5. Linear regression graphs showing correlations between multidetector computed tomographic angiography (MDCTA) and
cardiac catheterization angiography (CCA) for pulmonary annulus and pulmonary artery sizes and indices. PRV/LV: predicted ratio of
peak right ventricular pressure to peak left ventricular pressure.

good correlation between MDCTA and CCA for measurements of central PA size; however, this was a retrospective study in much younger patients with TOF with
pulmonary atresia.
Coronary artery anatomy is an important surgical
concern in patients with TOF with hypoplastic pulmonary annulus requiring a transannular patch. Any coronary or large branch crossing the RVOT should be
delineated before surgery. In our study, coronary
arteries crossing the RVOT were picked up by
MDCTA in all 4 cases. Interestingly, in 2 patients, a
large conal branch crossing the RVOT was initially
missed on a selective right coronary angiogram and
was diagnosed by a repeat angiogram after pulling
back the catheter (Figure 7). So, in this regard,
MDCTA is superior to CCA; diagnosis of coronary
artery anomalies can be dicult using selective coronary angiography, and is not without risk. In one large
study of TOF patients, MDCTA compared to surgical
ndings had 100% sensitivity and 100% specicity for
detecting coronary artery abnormalities.15 Kasar and
colleagues4 reported 96% accuracy of MDCTA in

revealing coronary artery anatomy as found on surgery.

Another striking advantage of MDCTA angiography is
its ability to show MAPCA from all potential sites in
one go.4 Missing these could lead to excessive operative
bleeding or cardiac failure in the postoperative period.
In our study, MAPCA were diagnosed equally well by
MDCTA and CCA, but their delineation by CCA
required multiple injections at multiple sites, followed
by selective injections into the collaterals. Aortic arch
laterality and branching pattern were best demonstrated by MDCTA. In a previous study, MDCTA
picked up unsuspected anomalies such as reverse
branching, trifurcating innominate artery, and
common origin of the subclavian and carotid arteries.4
The only limitation of MDCTA noted in our study
was that it missed a small additional muscular VSD in 2
patients. Dierentiation of a small muscular VSD from
RV trabeculae can be dicult because MDCTA images
are acquired during simultaneous opacication of the
RV and LV through the VSD. MDCTA avoided hospitalization and was more cost-eective. The mean
radiation dose in MDCTA was 3.3  4.2 mSv.

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015

172

Asian Cardiovascular & Thoracic Annals 23(2)

Figure 6. Multidetector computed tomography angiography images showing vascular abnormalities (extracardiac) in 4 patients.
(a) Sagittal oblique view (maximum-intensity projection) showing a left aortic arch with a trifurcating right brachiocephalic trunk
(arrow). (b) Volume-rendered image showing a right aortic arch (arrow) with a reverse branching pattern; the 1st branch is the left
brachiocephalic trunk, the 2nd is the right common carotid artery, and the 3rd is the right subclavian artery. (c) Sagittal oblique view
(maximum-intensity projection) showing a left aortic arch and patent ductus arteriosus. (d) Coronal oblique view (maximum-intensity
projection) showing bilateral superior venae cavae. PDA: patent ductus arteriosus; SVC: superior vena cava.

Figure 7. Large conal branch of the right coronary artery crossing over the right ventricular outflow tract (arrow). (a) Multidetector
computed tomography angiography (volume-rendered image). (b) Right coronary angiogram of the same patient, which initially missed
this proximally originating large conal branch crossing over the right ventricular outflow tract, and was diagnosed by repeat angiography after a little pull back of the catheter.

Kasar and colleagues4 reported similar radiation exposure per MDCTA study (3.5  4.7 mSv). The dose range
for diagnostic cardiac catheterization in children has
been reported as 0.2 to 14.4 mSv.16,17 Thus MDCTA
has an advantage over CCA in reducing the radiation

exposure. We used a non-ECG-gated protocol to

reduce the radiation dose. The origin and proximal
course of the coronaries can be easily and accurately
identied using this protocol, and it serves the purpose
with a substantial reduction in radiation exposure.

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015

Garg et al.

173

Figure 8. Multidetector computed tomography angiography images showing major aortopulmonary collaterals in 3 patients.
(a) Volume-rendered image showing major aortopulmonary collaterals (arrow) originating from the descending thoracic aorta and
supplying the right lower lobe of the lung with an arborization defect, and (b) Corresponding cineangiogram of the same patient,
showing the same collateral (arrow). (c) Coronal oblique view (maximum-intensity projection) showing a prominent right bronchial
collateral (arrow). (d) Coronal oblique view (maximum-intensity projection) showing a collateral originating from the left subclavian
artery (arrow).

Figure 9. Major aortopulmonary collateral detection by multidetector computed tomography angiography, and coil embolization
during cardiac catheterization just before surgery. (a) Volume-rendered image showing a large aortopulmonary collateral originating
from the descending thoracic aorta and connecting with the left pulmonary artery; the collateral has long-segment narrowing in its mid
segment (arrow). (b) Corresponding cineangiogram showing the same collateral. (c) The collateral was coil embolized during the same
catheterization study, and the patient underwent successful and uneventful surgery on the same day.

Two patients had cyanotic spells during catheterization,

4 had vascular access problems, 2 had transient complete heart block, and 1 had air embolism; all had
uncomplicated
MDCTA
examinations
earlier.

Requirement of sedation was also signicantly less

with MDCTA. So, MDCTA proved to be patientfriendly, clinician-friendly, and safer, without compromising accuracy. One adult patient had severe RV

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015

174

Asian Cardiovascular & Thoracic Annals 23(2)

Table 5. Advantages and disadvantages of multidetector computed tomography angiography and cardiac catheterization and
angiography.
Variable

MDCTA

CCA

Nature of the procedure

Hospitalization
Sedation
Contrast volume
Examination time
Radiation dose
Data reconstruction

Noninvasive
Not required
Less frequently required
Low
Short
Low
Can be manipulated and viewed
in unlimited no. of projections
Equal

Invasive
Required
More frequently required
Higher
Longer
Higher
Not possible

Can be missed
All can be imaged in one go
Always picked up

High pick-up rate

Require selective hooking
Can be missed

Good delineation
Can be imaged
Extracardiac cause can be identified
Probably better
None
None
Less
Low

Can be confusing
Not possible
Extracardiac cause not identified
Catheter positioning difficult
Possible
Possible
More
Higher

Delineation of RVOT and

pulmonary artery anatomy
Additional VSD
Collaterals
Coronary artery origin and
branch crossing RVOT
Aortic arch and its branches
Extracardiac structures
Patients with hemoptysis
In presence of severe tricuspid regurgitation
Femoral access site complications
Cardiac arrhythmia and air embolism
Operator dependence
Cost

Equal

CCA: cardiac catheterization and angiography; MDCTA: multidetector computed tomography angiography; RVOT: right ventricular outflow tract;
VSD: ventricular septal defect.

dysfunction and severe tricuspid regurgitation, requiring extensive and prolonged catheter manipulation and
large amounts of contrast and uoroscopy time;
MDCTA showed the relevant anatomy with ease
(Figure 4). Thus in TOF patients with severe RV dysfunction and severe tricuspid regurgitation, initial
MDCTA may be the best choice.
Hemoptysis in TOF patients is assumed to be due to
collaterals, but tuberculosis is also very common in our
country and can present with hemoptysis. MDCTA
may be helpful in indicating pulmonary tuberculosis
in these patients. In our study, 3 patients presenting
with hemoptysis were found to have pulmonary
parenchymal inltration on MDCTA, which was later
conrmed to be tuberculosis. Color Doppler echocardiography is the imaging technique of choice for the
initial assessment of a patient with TOF. It is a relatively inexpensive noninvasive technique that provides
anatomic and functional information in real time.
However, echocardiography is operator-dependent
and has diculty visualizing extracardiac anatomy
(PA and its branches, peripheral PA, origin and
course of coronary arteries and MAPCA). All these
limitations of echocardiography in TOF patients can

be overcome by MDCTA. Hence echocardiography

and MDCTA may be regarded as complimentary noninvasive imaging tests for preoperative evaluation of
TOF patients, the former being most suited for intracardiac anatomy and functional evaluation, and the
later for assessment of extracardiac structures.
Therefore, MDCTA is a safe and eective noninvasive
imaging technique to answer questions remaining after
echocardiography in patients with TOF. If cardiac
catheterization is still required for further evaluation
or to guide endovascular intervention, the information
provided by MDCTA has the potential to limit the
number of angiographic projections, resulting in a
reduction in the amount of contrast material and angiographic procedure time needed, which should lead to
substantial overall savings in terms of both healthcare
costs and radiation dose.
This was a relatively small study in relatively stable
patients at least 5 years of age with standard forms of
TOF. Further larger studies including patients less than
5-years old and complex TOF patients are required.
However, we concluded that for preoperative evaluation of TOF patients, MDCTA is a good imaging
technique for delineation of RVOT and PA anatomy,

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015

Garg et al.

175

coronary anatomy and MAPCA, and can be used as a

reliable noninvasive alternative to cardiac catheterization and angiography.
Funding
This research received no specic grant from any funding
agency in the public, commerical, or not-for-prot sectors.

Conflicts of interest statement

None declared.

References
1. Wang XM, Wu LB, Sun C, et al. Clinical application of
64-slice spiral CT in the diagnosis of the tetralogy of
Fallot. Eur J Radiol 2007; 64: 296301.
2. Westra SJ, Hill JA, Alejos JC, Galindo A, Boechat MI and
Laks H. Three-dimensional helical CT of pulmonary
arteries in infants and children with congenital heart disease. AJR Am J Roentgenol 1999; 173: 109115.
3. Hayabuchi Y, Mori K, Kitagawa T, Inoue M and
Kagami S. Accurate quantification of pulmonary artery
diameter in patients with cyanotic congenital heart disease using multidetector-row computed tomography. Am
Heart J 2007; 154: 783788.
4. Kasar PA, Ravikumar R, Varghese R, Kotecha M, Vimala
J and Kumar RN. Computed tomographic angiography in
tetralogy of Fallot. Asian Cardiovasc Thorac Ann 2011; 19:
324332.
5. Rajeshkannan R, Moorthy S, Sreekumar KP,
Ramachandran PV, Kumar RK and Remadevi KS. Role
of 64-MDCT in evaluation of pulmonary atresia with ventricular septal defect. AJR Am J Roentgenol 2010; 194:
110118.
6. Deibler AR, Kuzo RS, Vohringer M, et al. Imaging of
congenital coronary anomalies with multislice computed
tomography. Mayo Clin Proc 2004; 79: 10171023.
7. Duddalwar VA. Multislice CT angiography: a practical
guide to CT angiography in vascular imaging and intervention. Br J Radiol 2004; 77 (Spec no 1): S27S38.
8. Goo HW, Park IS, Ko JK, et al. CT of congenital heart
disease: normal anatomy and typical pathologic conditions
[Review]. Radiographics 2003; 23 (Spec No): S147S165.

9. Siegel MJ. Multiplanar and three-dimensional multidetector row CT of thoracic vessels and airways in the
pediatric population. Radiology 2003; 229: 641650.
10. Bhalla S and Siegel MJ. Multislice computed tomography
in pediatrics. In: Silverman PM (ed.) Multislice computed
tomography: a practical approach to clinical protocols.
Philadelphia, Pa: Lippincott, Williams & Wilkins, 2002,
pp. 231282.
11. Nakata S, Imai Y, Takanashi Y, et al. A new method for
the quantitative standardization of cross-sectional areas
of the pulmonary arteries in congenital heart disease with
decreased pulmonary blood flow. J Thorac Cardiovasc
Surg 1984; 88: 610619.
12. Blackstone EH, Kirklin JW, Bertranou EG, Labrosse CJ,
Soto B and Bargeron LM Jr. Preoperative prediction
from cineangiograms of postrepair right ventricular pressure in tetralogy of Fallot. J Thorac Cardiovasc Surg
1979; 78: 542552.
13. Z-scores of cardiac structures/Detroit data. Available at:
http://parameterz.blogspot.com/2008/09/z-scores-ofcar
diac structures.html. Accessed April 7, 2011.
14. Meinel FG, Huda W, Schoepf UJ, et al. Diagnostic
accuracy of CT angiography in infants with tetralogy of
Fallot with pulmonary atresia and major aortopulmonary collaterals. J Cardiovasc Comput Tomogr 2013; 7:
367375.
15. Vastel-Amzallag C, Le Bret E, Paul JF, et al. Diagnostic
accuracy of dual-source multislice computed tomographic analysis for the preoperative detection of coronary artery anomalies in 100 patients with tetralogy of
Fallot. J Thorac Cardiovasc Surg 2011; 142: 120126.
16. Yakoumakis EN, Gialousis GI, Papadopoulou D, et al.
Estimation of childrens radiation dose from cardiac
catheterizations performed for the diagnosis or treatment
of a congenital heart disease using TLD dosimetry and
Monte Carlo simulation. J Radiol Prot 2009; 29: 251261.
17. Karambatsakidou A, Sahlgren B, Hansson B, Lidegran
M and Fransson A. Effective dose conversion factors in
pediatric interventional cardiology. Br J Radiol 2009; 82:
748755.

Downloaded from aan.sagepub.com at SAGE Trial Account PARENT on April 23, 2015