Three contrast injection protocols for electrocardiogram-gated 64-slice computed tomographic angiography: comparing vessel density and perceived quality

To compare aortic, pulmonary artery (PA), and coronary artery (CA) density and examination quality using 3 different injection protocols for electrocardiogram-gated 64-detector row computed tomographic angiography.

We reviewed 30 consecutive computed tomographic angiography studies for each of 3 different contrast protocols (A, B, and C)-90 studies total-performed in patients with chest pain and low to moderate suspicion for coronary disease. All used a 20-mL timing bolus followed by 30 mL of saline. Protocol A (PrA) was composed of 50-mL contrast at 5 mL/s, 40-mL contrast/10-mL saline at 4 mL/s, and 40-mL saline at 4 mL/s. Protocol B (PrB) was composed of 50-mL contrast at 5 mL/s, 30- to 50-mL contrast at 3.5 mL/s (timing bolus-dependent), and 50-mL saline at 3.5 mL/s. Protocol C (PrC) was composed of 60-mL contrast at 5.5 mL/s, 50-mL contrast at 4.5 mL/s, and 50-mL saline at 3.5 mL/s. Density measurements were made at 20 points. Studies were graded for quality on a 4-point scale.

Main PA density for PrA (263 Hounsfield units [HU]) was significantly lower than for PrB (335 HU) and PrC (366 HU). For PrA, PrB, and PrC, there were 10, 1, and 1 nondiagnostic PA studies, respectively; the numbers of good or very good quality studies were 10, 21, and 25. Left main CA densities for PrA (393 HU), PrB (353 HU), and PrC (356 HU) were not significantly different; likewise for aortic root densities (400, 375, and 372 HU). For the aortic root, the numbers of good or very good studies were 30, 27, and 28. Grade (G) is linearly related to average density per grade (D): G = 0.011 × D - 1.46, r = 0.96.

Protocols B and C yielded consistently high attenuation of the aorta, PAs, and CAs with commensurate examination quality.