Non-invasive coronary angiography with multislice spiral computed tomography: impact of heart rate

Abstract

Objective: To evaluate the impact of heart rate on the diagnostic accuracy of coronary angiography by multislice spiral computed tomography (MSCT).

Design: Prospective observational study.

Patients: 78 patients who underwent both conventional and MSCT coronary angiography for suspicion of de novo coronary artery disease (n=53) or recurrent coronary artery disease after percutaneous intervention (n=25).

Setting: Tertiary referral centre.

Methods: Intravenously contrast enhanced MSCT coronary angiography was done during a single breath hold, and ECG synchronised images were reconstructed retrospectively. All coronary segments of > or = 2.0 mm without stents were evaluated by two investigators and compared with quantitative coronary angiography. Patients were classified according to the average heart rate (mean (SD)) into three equally sized groups: group 1, 55.8 (4.1) beats/min; group 2, 66.6 (2.8) beats/min; group 3, 81.7 (8.8) beats/min.

Results: Image quality was sufficient for analysis in 78% of the coronary segments in patients in group 1, 73% in group 2, and 54% in group 3 (p < 0.01). The sensitivity and specificity for detecting significant stenoses (> or = 50% lumen reduction) in these assessable segments were: 97% (95% confidence interval (CI) 84% to 100%) and 96% in group 1; 74% (52% to 89%) and 94% in group 2; and 67% (33% to 90%) and 94% in group 3 (p < 0.05). Accounting for all segments of > or = 2.0 mm, including lesions in non-assessable segments as false negatives, the sensitivity decreased to 82% (28/34 lesions, 95% CI 69% to 91%), 61% (14/23 lesions, 42% to 77%), and 32% (6/19 lesions, 15% to 50%), respectively (p < 0.01).

Conclusions: MSCT allows reliable coronary angiography in patients with low heart rates.

Figure 1

Heart rate dependency of the temporal resolution. Isocardiophasic transverse slices are reconstructed according to the recorded ECG using a 180° rotation partial scan algorithm. For heart rates up to 65 beats/min a single segment reconstruction algorithm is applied to reconstruct slices during a 250 ms reconstruction window (abs. RI) within each separate cardiac cycle. To improve the effective temporal resolution and reduce motion artefacts, a double segment reconstruction algorithm—which combines isocardiophasic data from two consecutive RR intervals—is applied at heart rates over 65 beats/min. The effective reconstruction interval per cardiac cycle depends on a complex relation between the gantry rotation time and the duration of the cardiac cycle. At a given rotation speed it varies between 125 ms at favourable and 250 ms at unfavourable heart rates. The relative reconstruction interval (rel. RI)—that is, the ratio between the absolute reconstruction interval and the total R to R wave interval (%)—progressively increases at higher heart rates (Siemens Somatom Plus 4 VolumeZoom with a rotation time of 500 ms).

Figure 2

Heart rate dependency of coronary segment assessability. At lower heart rates ≥ 2.0 mm segments per patient are assessable. The interrupted trend line represents the number of segments of ≥ 2.0 mm diameter, as a ratio of 16 potentially available segments, which is independent of the patient’s heart rate.

Figure 3

Multislice computed tomography (MSCT) angiogram of an obstructed left anterior coronary artery (LAD) and a heart rate of 49 beats/min. A significant stenosis (arrow) can be observed in the LAD. The three dimensional volume rendered overview (A) shows the lesion just distal to a small diagonal branch, which is confirmed by conventional angiography (B). The cross sectional (D) and longitudinal reconstructions (E) show a partially calcified lesion, which is confirmed by intracoronary ultrasound (C). D2, second diagonal branch; IM, intermediate branch; LCX, left circumflex coronary artery; RCA, right coronary artery. RVOT, right ventricular outflow tract

Figure 4

Multislice computed tomography (MSCT) angiograms at different heart rates. The axial source images (A,B,C), curved multiplanar reconstructions (D,E,F), and volume rendered reconstructions (G,H,I) of a normal right coronary artery (RCA, arrows) in patients with an average heart rate of 49 beats/min (A,D,G), 64 beats/min (B,E,H), and 81 beats/min (C,F,I), respectively, are shown. Motion artefacts (arrowheads) hinder assessment of the distal RCA in the second case and nearly the entire RCA in the third. RV, right ventricle.

Figure 5

Heart rate variation during breath holding. Averaged relative heart rate (SD) of 10 randomly selected cases is shown. After an initial deceleration caused by a Valsalva induced vagal response, the heart rate progressively increases after approximately 20 seconds.