Author + information
- Received February 12, 2008
- Revision received June 23, 2008
- Accepted July 10, 2008
- Published online October 28, 2008.
- Takao Maruyama, MD, PhD⁎,⁎ (, )
- Masanori Takada, MD⁎,
- Toshiaki Hasuike, MD⁎,
- Atsushi Yoshikawa, MD⁎,
- Eiji Namimatsu, RT† and
- Tohru Yoshizumi, RT⁎,‡
- ↵⁎Reprint requests and correspondence:
Dr. Takao Maruyama, Department of Cardiology, Kawasaki Hospital, 3-3-1, Higashiyama, Hyogo, Kobe 652-0042, Japan
Objectives The aim of this study was to evaluate radiation dose and coronary assessability of a prospective electrocardiogram (ECG)-gated scan by 64-slice multidetector (row) computed tomography (MDCT)-coronary angiography (CA) compared with a retrospective ECG-gated helical scan.
Background The 64-slice MDCT-CA has been widely used; however, a high radiation dose by 64-slice MDCT-CA has been reported. Prospective ECG-gated scan using “step-and-shoot” protocol can reduce radiation exposure effectively.
Methods MDCT-CA was performed in 229 consecutive patients. Fifty-six patients were excluded because of higher heart rates of >65 beats/min; of patients with heart rates ≤65 beats/min, 97 were analyzed by helical scan with tube current modulation and 76 were analyzed by prospective gating. Coronary assessability and diagnostic accuracy were investigated in comparison with selective CA as the gold standard. Radiation doses were evaluated in both protocols.
Results Coronary assessability of helical scan was 95.5% (1,303 of 1,364 segments), while that of prospective gating was 96.6% (1,053 of 1,089 segments), showing similar coronary assessability (p = 0.14). Sensitivity and specificity for coronary obstructive and occlusive lesions in the assessable segments were 97.0% (162 of 167) and 97.6% (1,109 of 1,136) by helical scan, while those of prospective gating were 96.4% (81 of 84, p = 0.84) and 98.5% (955 of 969, p = 0.12), respectively. Effective doses of helical scan and prospective gating were 21.1 ± 6.7 mSv and 4.3 ± 1.3 mSv, respectively (p < 0.0001), showing that prospective gating decreased radiation dose by 79% compared with that of helical scan.
Conclusions MDCT-CA by prospective gating showed equivalent coronary assessability and diagnostic accuracy with decreased radiation dose in comparison with a retrospective ECG-gated helical scan with tube current modulation.
In recent years, multidetector (row) computed tomography (MDCT) has been introduced and widely used for noninvasive evaluation of coronary arteries. The latest 64-slice MDCT-coronary angiography (CA) with retrospective electrocardiogram (ECG) gating has high assessability and diagnostic accuracy (1–5); however, a high radiation dose from 64-slice MDCT-CA has been noticed (6–8), and it may be related to a higher incidence of cancer (8). For dose reduction, an ECG-triggered dose modulation algorithm that reduces the tube current by 80% during the systolic phase can decrease the radiation dose by approximately 20% (9). Recently, a prospective ECG-gated scan for MDCT-CA has been developed, which reduces radiation dose more effectively by the “step-and-shoot” protocol (10,11). Therefore, we investigated the radiation dose and coronary assessability of coronary computed tomography (CT) angiography by prospective gating, and compared it with retrospective ECG-gated helical scan with dose modulation.
A total of 273 consecutive patients were enrolled between December 2006 and November 2007 because of suspected coronary artery disease or follow-up after coronary intervention. Exclusion criteria included arrhythmias such as atrial fibrillation and frequent premature contractions, coronary bypass grafts, and renal insufficiency; 44 patients were excluded for these reasons, and 229 patients underwent MDCT-CA. The study period was from December 2006 to June 2007 for helical scan and from July 2007 to November 2007 for prospective gating (Fig. 1). Before CT examination, patients with heart rates (HRs) higher than 70 beats/min were administered oral beta-blockers (metoprolol, 20 mg). Patients with HRs of >65 beats/min during CT examination were excluded from analysis because of the different reconstruction method: step-and-shoot protocol only permits a “half-reconstruction” method, and in patients with HRs of >65 beats/min, “multisector reconstruction” is recommended instead of half-reconstruction due to the insufficient ECG gating. Therefore, 56 patients were excluded from this analysis because of higher HRs, and 97 and 76 consecutive patients with HRs of ≤65 beats/min were investigated by retrospective ECG-gated helical scan and prospective gating, respectively. Selective angiography was performed within 2 months after CT examination in order to confirm the diagnosis of coronary artery disease or to restudy after intervention. Study protocols were in accord with ethical standards of the hospital, and approved by the Ethical Committee of the hospital. Written informed consent was obtained from all patients.
MDCT-CA by helical scan
MDCT data were acquired using LightSpeed VCT (GE Healthcare, Waukesha, Wisconsin). ECG monitoring was performed continuously during the examination. First, a scout scan was performed in inspiratory breath-hold to gain a sagittal and coronal view for positioning of the entire heart. A region of interest was positioned at the center of the ascending aorta; then, a test bolus (10 ml) of the iodinated contrast agent iohexol (350 mgI/ml) was injected intravenously at 4.0 ml/s with a saline chase of 20 ml at 5.0 ml/s. The time interval between bolus injection and maximal enhancement in the ascending aorta was measured, and the starting time of the enhanced scan was calculated as 3 s after transit time of the contrast agent. Then 0.7 ml/kg body weight of contrast agent was injected, followed by a saline chase of 30 ml, and the scan was started with the delay time determined previously with 64 × 0.625 mm collimation. The gantry rotation time was 0.35 s, and the tube current was 800 mA at 120 kV with dose modulation of 70% to 80% of RR interval (Fig. 2A). All acquired data were transferred to a computer workstation (Advantage Workstation 4.3, GE Healthcare) and reconstructed by the half-reconstruction method. The effective slice thickness was 0.625 mm, and the reconstruction increment was approximately 0.5 mm. The dataset with the best image quality containing the fewest artifacts was selected for visualization of each coronary artery.
MDCT-CA by prospective gating
Preparation, contrast media injection, and data reconstruction were similar to those of helical scan. Data acquisition was performed within 3 to 4 beats at mid-diastolic phase (around 75% of RR interval) by step-and-shoot protocol (SnapShot Pulse, GE Healthcare) (Fig. 2B). The gantry rotation time was 0.35 s, and tube current was 800 mA at 120 kV with average padding time of 30.4 ms (range 0 to 50 ms, according to the HR variability). It is noteworthy that prospective gating was the initial acquisition mode (12) applied in the pioneering study on cardiac CT imaging (13), and that temporal resolution of CT scanners has been insufficient to visualize coronary arteries until the introduction of fast gantry rotation times.
Evaluation of MDCT-CA
First, each coronary segment including branches was identified by maximal intensity projection and coronary tree images; then, curved multiplanar reformatted and cross-sectional images of each segment were created and evaluated (Fig. 3). Image quality of each segment was evaluated, and assessability was determined according to the American Heart Association 15-segment model (14). Segments with “excellent” (excellent quality) and “good” (good quality, assessable for stenosis) images were judged as “assessable” (15). Segments with “fair” (fairly visualized, nonassessable) and “poor” (poorly or not visualized) images were judged as “nonassessable.” Visualized coronary stenoses were classified by 3 grades such as normal-to-mild (≤50%), obstructive (>50%, nonocclusive), and occlusive (100%). In nonassessable segments, causes responsible for nonassessability were evaluated (15). Segments with implanted coronary stents were excluded from this study.
Computed tomography dose index volume (CTDIvol) (mGy) was provided by the CT scanner and multiplied by scan length (cm) in order to obtain dose-length product (mGy · cm) (7,8). To obtain the effective dose (mSv), dose-length product was multiplied by the appropriate conversion factor for the chest (0.017 mSv/[mGy · cm]).
Selective CA was performed by the standard method. The angiograms were evaluated by quantitative coronary analysis with vessel contour detection after catheter-based image calibration. Coronary stenotic lesions were classified by grades: normal-to-mild (diameter reduction ≤50%), obstructive (diameter reduction of >50%, nonocclusive), and occlusive, and compared with those of MDCT-CA.
Data were expressed as mean ± SD (ranges) or percentages. Statistical analyses were performed by the Mann Whitney U test and chi-square test.
Patient characteristics are shown in Table 1, showing no significant differences between patients' age, gender, body mass index, risk factors, and HR during CT examination. No serious arrhythmias were observed during data acquisition, and no patients underwent repeated enhanced scan.
Table 2 demonstrates assessability of coronary segments by helical scan and prospective gating, showing that prospective gating has similar coronary assessability to that of helical scan. The cause for nonassessability was mostly massive calcium deposits of coronary segments.
A patient-based analysis of coronary assessability is shown in Figure 4. Prevalence of patients in whom all entire coronary segments were assessable was 85% (82 of 97) by helical scan, while that of prospective gating was 86% (65 of 76), demonstrating equivalent coronary assessability of both protocols (p = 0.84). Diagnostic accuracy of MDCT-CA in comparison with selective angiography is shown in Table 3.
Sensitivities for obstructive and occlusive coronary lesions in assessable segments by helical scan and prospective gating were 97.0% (162 of 167) and 96.4% (81 of 84), respectively (p = 0.84, 95% confidence interval: 0.19 to 3.6), while specificities were 97.6% (1,109 of 1,136) and 98.5% (955 of 969), respectively (p = 0.12, 95% confidence interval: 0.86 to 3.2), demonstrating similar diagnostic values.
Patient-based diagnostic values were evaluated by investigating and comparing maximal stenotic grades of the coronary arteries in patients in whom all coronary segments were assessable by MDCT-CA, with selective angiography as the gold standard. Sensitivity, specificity, and accuracy of coronary obstructive and occlusive lesions by helical scan (n = 82) were 96% (48 of 50), 94% (30 of 32), and 95% (78 of 82), respectively, while those by prospective gating (n = 65) were 100% (27 of 27), 92% (35 of 38), and 95% (62 of 65), respectively, showing similar diagnostic values (p = 0.29, 0.79, and 0.94, respectively).
Table 4 demonstrates radiation doses of helical scan and prospective gating, showing that exposure time, CTDIvol, dose-length product, and effective dose of prospective gating were significantly lower than those of helical scan with dose modulation. Estimated effective doses of helical scan and prospective gating were 21.1 ± 6.7 mSv and 4.3 ± 1.3 mSv, respectively, revealing that prospective gating enabled dose reduction by 79% compared with dose-modulated helical scan.
In the current study, we investigated radiation dose and coronary assessability of 64-slice MDCT-CA by prospective ECG-gated scan, and demonstrated that prospective gating has equivalent coronary assessability to that of retrospective ECG-gated helical scan in patients with a lower HR of ≤65 beats/min. The effective dose of prospective gating was estimated to be approximately 4.3 mSv, which is lower than that of the reported diagnostic selective angiography (5 to 6 mSv) (16,17), suggesting that step-and-shoot protocol is useful, especially in patients who undergo radiation exposure repetition, such as follow-up after coronary intervention.
The effective dose of prospective gating in this study was higher than that in the previous report (11): it was speculated that the increase in radiation dose was caused by the use of padding. The HR variability of patients in this study was more remarkable than that in the previous report, and average padding time of 30.4 ms was added according to the grades of HR variability in order to find an optimal time period during the diastolic phase in patients with HR variations, to minimize the “stair-step” artifacts (11), which resulted in the increase in dose exposure. Although no stair-step artifacts were observed in this study, it is obvious that setting the padding function to “off” will minimize radiation dose, and appropriate setting of the padding is disputable in patients with marked HR variability.
One of the limitations of prospective gating is that it could not be applied in patients with a higher HR, because step-and-shoot protocol permits only the half-reconstruction method, and its application is limited only to patients with a lower HR: a significant proportion of patients had to be excluded because of the higher HR. To use step-and-shoot protocol in patients with a higher HR, administration of a beta-blocker is essential.
Arrhythmias may affect image quality because of insufficient ECG gating in MDCT-CA. In this study, no serious arrhythmias were observed; however, as exposure time of prospective gating is short and limited in the mid-diastolic phase, adequate editing and searching for optimal time phase could not be performed in patients with arrhythmias, and restudy may be needed if serious arrhythmias occurred during data acquisition.
Another limitation of this study was that patients were not randomized to 2 different protocols, which may lead to a bias. Prospective ECG gating represents an important step forward in MDCT-CA, but it is not a mature technique and has several weaknesses. Nevertheless, this protocol has great potential in combination with further required improvement in MDCT-CA technique, such as higher rotation speed and increasing numbers of detectors allowing full heart coverage.
- Abbreviations and Acronyms
- coronary angiography
- computed tomography
- computed tomography dose index volume
- heart rate(s)
- multidetector (row) computed tomography
- Received February 12, 2008.
- Revision received June 23, 2008.
- Accepted July 10, 2008.
- American College of Cardiology Foundation
- Raff G.L.,
- Gallagher M.J.,
- O'Neill W.W.,
- Goldstein J.A.
- Leber A.W.,
- Knez A.K.,
- von Ziegler F.,
- et al.
- Leschka S.,
- Alkadhi H.,
- Plass A.,
- et al.
- Pugliese F.,
- Mollet N.R.,
- Runza G.,
- et al.
- Hausleiter J.,
- Meyer T.,
- Hadamitzky M.,
- et al.
- Nickoloff E.L.,
- Alderson P.O.
- Husmann L.,
- Valenta I.,
- Gaemperli O.,
- et al.
- Hounsfield G.N.
- Coles D.R.,
- Smail M.A.,
- Negus I.S.,
- et al.
- Zanzonico P.,
- Rothenberg L.N.,
- Strauss H.W.