Using multidetector computed tomography in a swine model to assess the effects of sublingual nitroglycerin and intravenous adenosine on epicardial coronary arteries

Abstract

This study examines the effects of intravenous infusion of adenosine and sublingual nitroglycerin on coronary angiograms obtained by current-generation multidetector computed tomography. We assessed coronary vasodilation at baseline and after intravenous adenosine (140 microg/kg/min) or sublingual nitroglycerin spray (800 microg) in 7 female swine (weight, 40.9 +/- 1.4 kg) by using electrocardiogram-gated coronary angiography with a 64-detector scanner (rotation time, 400 ms; 120kV; 400 mA) and intravenous contrast (300 mg/mL iohexol, 4.5 mL/s, 2 mL/kg). Cross-sectional areas of segments in the left anterior descending, circumflex, and right coronary arteries were evaluated in oblique orthogonal views. Images were acquired at an average heart rate of 73 +/- 11 beats per minute. Changes in aortic pressure were not significant with nitroglycerin but decreased (approximately 10%) with adenosine. Of the 76 segments analyzed (baseline range, 2 to 39 mm2), 1 distal segment could not be assessed after adenosine. Segment cross-sectional area increased by 11.3% with nitroglycerin but decreased by 9.6% during adenosine infusion. The results of the present study are consistent with the practice of using sublingual nitroglycerin to enhance visualization of epicardial vessels and suggest that intravenous adenosine may hinder coronary artery visualization. This study is the first repeated-measures electrocardiogram-gated CT evaluation to use the same imaging technology to assess changes in coronary cross-sectional area before and after treatment with a vasodilator. The nitroglycerin-associated changes in our swine model were modest in comparison with previously reported human studies.

Figure 1.

Coronal view of an 8.3-mm maximum intensity projection image of the LAD as it courses through the anterior interventricular sulcus. In this view, the right ventricle (RV) is superior to the left ventricle (LV) and the diagonal branches of the LAD are seen traversing inferiorly to the anterior LV wall. In this projection, the internal mammary arteries (unlabeled) with distributions to the anterior chest wall also are visible. The yellow arrows depict an example of a distal site for measurement of LAD cross-sectional area.

Figure 2.

Three-dimensional volume-rendered view of the right coronary artery (RCA), LAD, and left circumflex (LCX) coronary arteries in a 10° left anterior oblique and cranial projection. The yellow arrows illustrate the location of a distal LAD measurement site (same site as Figure 1) in 3D and curved multiplanar maximal intensity projection views of the LAD. At this site, the cross-sectional area of the LAD was 3.1 mm². The upper right panel presents measurements of cross-sectional area from the transverse multiplanar reformatted image. The middle right panel is an example of the anatomical tapering of the LAD (blue line) as a plot of transverse multiplanar reformatted areas versus vessel length.

Figure 3.

Oblique 0.5-mm maximal intensity projections images of the proximal right coronary artery (RCA) as it traverses the anterior right atrioventricular sulcus. Magnified views demonstrate changes in vessel cross-sectional area (A) at baseline (area, 20 mm²), (B) after adenosine (19 mm²), and (C) after nitroglycerin (22 mm²).

Figure 4.

Plot of nitroglycerin- and adenosine-associated changes in coronary cross-sectional area versus baseline measurements. An upward shift in vessel cross-sectional area with nitroglycerin (long dashed line with gray circles) from the line of identity (dark line) favors enhanced visualization. The downward shift in the regression line for intravenously adenosine (dotted line and open circles) suggests poorer visualization.

Figure 5.

Box plots compare the distribution of distal (2 to 5.5 mm²), mid (6 to 10 mm²), and proximal (larger than 10 mm²) coronary vessels for baseline data and after adenosine and nitroglycerin treatment. The line within each box plot represents median cross-sectional area, and the box plot whiskers represent 10% and 90% of range values.