Keywords
coronary CT angiography - low kVp - radiation
Introduction
It is appropriate to use coronary computed tomography (CT) angiography (CCTA) not
only to screen patients for coronary artery disease but also in the setting of acute
chest pain, especially when echocardiography and electrocardiogram (ECG) results are
divergent.[1] However, the use of standard 64 slice CCTA carries a potential hazard of high radiation
that varies from 6 millisievert (mSv) to 15 mSv, even with the use of dose reduction
techniques.[2] There are also concerns related to contrast-induced nephropathy after performing
coronary angiography and whether these could be mitigated by reducing the volume injected.[3]
[4] A variety of dose-reduction methods like reducing tube voltage, iterative reconstruction,
and high-pitch modes are being used.[5]
[6] Prior studies done using only low kilovoltage (kV) techniques with prospective ECG
gating on patients with body mass index (BMI) less than 25 kg/m2 and with controlled heart rates less than 60 beats/min have shown that 30 to 80%
dose reduction is possible to as low as 0.2 mSv but with increased image noise and
reduced image quality (IQ),[7] which have been corrected with the use of newer iterative reconstruction techniques.[8] Obesity is prevalent worldwide and is a comorbid factor in coronary artery disease,
and clinical imaging involves doing examinations on a large number of obese patients
who have BMI >25 kg/m2. Another challenge in performing coronary CT has been the issue of high heart rate,
that is, more than 70 beats/min; hence, choosing a prospective gating technique on
single slice scanners may not be possible all the time, therefore spiral retrospective
ECG gating protocols are selected. To our knowledge, no retrospective gating technique
has been evaluated to determine if low-dose low-volume protocol can be useful to examine
such patients.
This study was designed to evaluate if low-kV and low-volume scan with retrospective
ECG-gated protocol can be done to perform CCTA of such patients with BMI up to 31 kg/m2 and or with higher heart rates up to 100 beats/min when compared with standard 120 kV
angiographic protocol.
Materials and Methods
This was a randomized study comprising of 100 prospective patients who visited our
institute for CCTA. Approval was taken by local ethical committee of the institute.
The patients were randomized into two groups of 50 patients each. First group (A)
underwent angiographic examination using a protocol with 70 kV tube voltage and 35 mL
noniodinated intravenous contrast iomeron 400. The second group (B) had angiographic
examination done using standard protocol of 120 kV tube voltage and 90 mL of same
noniodinated contrast.
The demographic parameters of all patients, that is, age, sex, BMI, along with history
of disease, coexisting morbidities like hypertension, diabetes, and hyperlipidemias,
were recorded. All patients with history of allergy to iodine, arrhythmias, heart
rate of more than 100 beats/min, BMI more than 31 kg/m2, increased creatinine levels >1.2 mg/mL, or any prior bypass grafting/angioplasty
were excluded.
All patients underwent CT examination after obtaining informed consent on dual energy
CT system (Siemens Go-Top) 128 slice scanner after receiving oral 50 mg metoprolol
1 hour before the examination. Tube voltage was set according to the group of patients
being examined using the protocol described. A retrospectively gated ECG-triggered
spiral (pitch 0.3) acquisition was done with collimation and gantry rotation time
of 0.31 milliseconds in both the groups. For group A patients, the tube voltage was
fixed at 70 kV along with automated tube current modulation using CARE Dose 4D (Siemens
healthineers), and the quality reference voltage and current was set to 120 kV and
320 milliampere-seconds (mAs). R-R interval was fixed at 35 to 75%. Also, 35 mL of
noniodinated contrast iomeron 400 (Iomeprol 400 mg iodine/mL; Iomeron Bracco UK Ltd)
was injected using bolus tracking software with dual head injector (Medrad, Stellant,
Bayers, Munich, Germany) with flow rate of 4 mL/s followed by 20 mL saline. Image
reconstruction was done using sinogram-affirmed iterative reconstruction (SAFIRE;
Siemens healthineers level 3) with reconstruction kernel of BV36. For group B patients,
according to the protocol the tube voltage was fixed at 120 kV with 90 mL of iomperal
400 injected intravenously at flow rate of 5 mL/s using bolus tracking as in group
A followed by 40 mL saline. All images were transferred to Siemens Syngo.via workstation
for postprocessing in multiplanar and volume rendering projections.
Radiation dose was calculated in both groups of patients by the system and displayed
in the form of dose-length product, which was converted to mSv using conversion factor
of 0.014.
Image Evaluation
Contrast attenuation values of ascending aorta (AAO), left main coronary, proximal
right coronary artery, proximal left anterior descending, and left circumflex arteries
were obtained by drawing region of interest within the lumen of respective vessels
and value deemed as signal of the vessel. The standard deviation of the above displayed
in the region of interest was the noise value and quotient of the two obtained was
signal-to-noise (S/N) ratio. Contrast-to-noise (C/N) ratio was determined by attenuation
value of vessel minus attenuation value of adipose tissue/attenuation value of noise
of vessel.
IQ was rated on a score of 1 to 4 with score 4 as excellent, 3 as moderate but acceptable,
2 as average, and 1 as being poor and rejected by all the three authors independently.
All scores were determined on per patient and per vessel basis.
Statistical Analysis
Statistical analysis was done using Analyze-IT software (Leeds, UK). Mean values with
standard deviations were obtained of all categorical variables and Student's t-test done to determine the difference in significance. Mann-Whitney U test was done
for categorical variables. Post hoc power analysis was done for the study and power
fixed at 0.78 for both groups with significance level being fixed at p = 0.05. Interobserver agreement (IOA) was calculated using Kappa variable for IQ.
Kappa of more than 0.6 was labeled as good agreement.
Results
Patient demographics of both groups are shown in [Table 1]. The mean age in both the groups was 57 and 57.2 years with mean BMI of 30.16 and
30.49 kg/m2 respectively, the difference being statistically insignificant. The mean heart rate
in both groups was 72.0 and 74.0 (p = –0.45). Image characteristics of both the groups are shown in [Table 2]. Statistically significant differences were seen in the contrast opacifications
of AAO and in all the main coronary vessels, with group B patients showing higher
opacification. All patients in group A also achieved minimal desirable opacification
of more than 350 Hounsfield units (HU) in all the vessels. Group A patients, however,
showed a significant reduced image noise with mean of 29 HU (95% confidence interval
[CI]: 27.5–30.7; [Fig. 1]). Group B patients showed mean noise of 57 HU (95% CI: 54.5–59.6; p < 0.0001). The mean S/N and C/N ratios in group A patients were also higher, that
is, 20.25, 18.68, 19.04, 17.41, and 18.69 for aorta, left main, left anterior descending,
right coronary, and left circumflex arteries ([Fig. 2A]–[D]). Group B patients showed mean S/N ratios of 13.34, 11.12, 10.96, 9.74, and 8.67
in AAO, left main, left anterior descending, right coronary artery, and left circumflex
artery, respectively. The mean C/N ratios were also higher in all vessels in group
A patients, that is,19.48, 19.48, 19.04, 19.48, and 17.68 while group B had 12.43,
10.03, 9.23, 9.57, and 8.23, respectively, for left main, left anterior descending,
left circumflex, and right coronary arteries (p < 0.0001; [Fig. 3A]–[E]). There were patients with heart rate of more than 95th percentile, that is, more
than 86 beats/min; however, good quality images with IQ 4 were obtained in all of
them ([Fig. 4A]–[C]). Kappa for IOA was 0.84 and 0.63 for per patient and per vessel IQ scores ([Table 3]). There were 5 patients and 24 vessels from both the groups who had IQ score of
1 and were not assessed while 11 and 9 patients had IQ scores of 3 and 2, respectively
([Fig. 5A], [B]). Significant reduction in the radiation dose was seen in group A patients with
mean dose of 1.86 mSv while it was 6.89 in group B (p < 0.001; [Table 4]).
Table 1
Patient demographics
|
|
Table 2
Imaging characteristics in group A and group B patients
|
|
Table 3
Kappa statistic for interobserver agreement (A) on per patient basis and (B) on per
vessel basis
|
|
Table 4
Student's t-test for radiation difference in group A and group B patients
|
|
Fig. 1 Bar chart showing image noise in group A and group B patients.
Fig. 2 (A) Group A patient with heart rate 65 beats/min axial section at the level of ascending
aorta showing signal-to-noise ratio of 19.4 and 20.7 at proximal right coronary artery
and proximal left anterior descending coronary artery, respectively. (B) Multiplanar view of left anterior descending coronary artery with image quality
(IQ) of 4.0. (C) Multiplanar view of right coronary artery of same patient with IQ of 4 days. (D) Virtual rendering image of same patient.
Fig. 3 (A) Group B patient showing axial section with signal-to-noise (S/N) ratio of 11.15
of left main and 23.0 of ascending aorta. (B) Axial section same patient with S/N of 9.7 of left circumflex artery. (C) Axial section of proximal right coronary artery with S/N of 10.0 days. Straightened
multiplanar view of left anterior descending artery with image quality score of 4.0.
(E) Virtual rendering image of heart of same patient.
Fig. 4 (A) Group A patient with heart rate of 100 beats/min showing signal-to-noise ratio of
25 and 19.5 at proximal ascending aorta and left anterior descending coronary artery,
respectively. (B) Virtual rendered image of same patient. (C) Multiplanar view of left anterior descending aorta with image quality of 4.
Fig. 5 (A) Group A patient with heart rate of 81 beats/min showing signal-to-noise of 28 and
24.5 of proximal right coronary artery and ascending aorta, respectively. (B) Multiplanar view of same patient showing image quality of 3 with some vessel wall
blurring.
Discussion
The present study showed that lowering the tube voltage to 70 kV obtained a good contrast
opacification of the coronary vessels due to positive effect of photoelectric effect
with k-edge of iodine being 33 kiloelectronvolt (keV), which was closer to 70 kV than
120 kV. Our study showed a higher contrast opacification in group B patients even
as both groups achieved opacification higher than the threshold value of 350 HU. This
meant that there was a potential to reduce volume of contrast injected without compromising
on IQ. It has been shown that increasing the tube current offsets the negative effect
of lowering tube voltage on the IQ.[9] This was shown in the present study by the use of increased tube current using Care
Dose 4D protocol by which the tube current was more than doubled to 850 mAs, thus
increasing the number of photons as compared with the 120 kV protocol. The increase
in tube current also allowed to scan patients with higher BMI up to 31 kg/m2 as was seen in this study without comprising the contrast and IQ. It was however
interesting to observe that SAFIRE-iterative algorithm used in both the group of patients
did not show any difference in performance when used with high or low tube potential
and the reduction in tube potential from 100 kV to 70 kV resulting in more than 30%
increase in noise; this was compensated by increase in the tube current and it was
the impact of lowered noise that improved both C/N and S/N ratios while scanning at
low peak kilovoltage (kVp). At low tube potential, the influence of the photoelectric
effect is greater than Compton scattering effects because of the 33-keV k-edge of
iodine[10] and results in increased contrast enhancement while increased tube current lowers
the image noise. These two factors combined make examination possible with low-volume
injection. Similar results were shown by Kok et al[11] who showed a 56% reduction in injected contrast when low-kV protocol was used. Lell
et al[12] also showed in their ex vivo model that optimal contrast can be achieved with reduced
iodine dose using 70 kV protocol. This gives many advantages, namely it permits use
of lesser gauge intravenous cannulas and reduces the incidence of contrast-induced
nephropathy, anaphylactic reactions, volume overloading in cardiac patients, and the
cost of consumables per patient. Our study shows that the amount of intravenous contrast
can be reduced by half without compromising on IQ.[13]
[14] The subjective IQ scores in both the groups showed no statistical difference. The
present study also showed a threefold reduction in the radiation dose in group A patients
by the use of low-kV protocol. Similar results have been shown by Zhang et al[14] who compared 70 kV protocol with 100 kV protocol. They showed an average dose was
0.38 mSv in patients with BMI of less than 23 kg/m2 on a dual source CT with prospective ECG gating. Cao et al[15] and Gordic et al[16] also reported that 80 kVp CCTA could reduce up to 57.8% radiation dose for patients
(BMI < 23 kg/m2), as compared with 120 kVp. Our study was different from these studies as we used
retrospective ECG gating instead of prospective gating as our patients had higher
heart rates as compared with controlled recommended heart rates of 60 beats/min. The
radiation dose in our study was slightly higher than these studies due to the selection
of retrospective gating. This was due to the nature of outpatient patients and workflow-related
issues, which put constraints of increased waiting time for examination preparation.
In spite of the above limitation, our study shows that the use of low-kV protocol
reduced the radiation dose threefold even with the use of retrospective gating protocol
and provided an optimal IQ. We feel that this protocol offsets the disadvantage of
not conducting scans on expensive dual source high pitch or 320 slice single rotation
scanners where one has the privilege of using prospective gating protocol even with
higher heart rates and thus reducing radiation to below 1 mSv.
To conclude, the present study shows that low-kV low-dose contrast protocol with retrospective
ECG gating not only provides good diagnostic quality angiographic images and reduces
radiation dose by threefold compared with 120 kV protocol, but also reduces cost of
contrast administered along with reduced risks of using iodinated contrasts. It gives
good results in patients with higher BMI up to 31 kg/m2 and with higher heart rates below 100 beats/min and appears to be a good choice when
working on single source scanners.
