This website is intended for healthcare professionals only.

Hospital Healthcare Europe
Hospital Healthcare Europe

Which injection protocol to use for cardiac CT imaging?

Arshid Azarine MD MSc
18 May, 2016  

Intra-luminal contrast enhancement is crucial in cardiovascular imaging to ensure reliable diagnostic accuracy

Arshid Azarine MD MSc
Radiology Department
Hôpital Européen Georges Pompidou, 
Paris, France
Progress  in imaging has led to a  better evaluation of all cardiovascular diseases and after conventional angiography and echocardiography, more recent techniques as Cardiac Magnetic Resonance (CMR) and Multi-Slice Computed Tomography (MSCT), have allowed a 3D assessment of the thoracic cardiovascular structures, enabling accurate anatomic and functional evaluation of cardiovascular diseases.
Contrast Enhanced Multi-Slice Computed Tomography (CE-MSCT) with ECG gated acquisition provides highly precise isotropic cardiac and aortic images allowing accurate and precise measurements, by reducing the motion artefacts during cardiac cycle. Retrospective acquisition enables multi-phase acquisition, giving data about the whole cardiac cycle, at a cost of higher radiation exposure, while prospective sequential acquisition enables a more focused in time acquisition, but with much lower radiation exposure with doses down to 1–2mSv.
Intra-luminal contrast enhancement is crucial in cardiovascular imaging to have a reliable diagnostic accuracy. 
Becker et al  showed that the minimum intracoronary attenuation to be achieved when performing coronary CE-MSCT is 250–300 HU at the lowest. Increasing the contrast volume (ml), injection rate (ml/s) and iodine concentration ([I]) increases the peak of maximum enhancement. By changing these three variable parameters we can obtain the optimal expected attenuation.
As the iodine administration rate can be modified more easily by changing the injection rate than by changing the contrast iodine concentration, in our institution, after having tried different options, we have chosen an injection rate of 5–6ml/s of 350mgI/ml contrast agent, for our routine cardiac CE-MSCT. These injection rates require larger needles and good veins. We obtain optimal intracoronary attenuation around 400±20UH.
Fig. 1: Panel A: shows an optimal enhancement of the left ventricle (380UH) with correct enhancement of the right ventricle (RV) (280UH) without streak artefacts, enabling good delineation of RV muscular and valvular structures, due to the diluted phase of the injection. Superior vena cava was washed out due to a fast but short finale saline flush (15–20cc at 4cc/ml.  
Panel B: 3D reconstruction shows optimal enhancement of all expected structures, aorta, coronary arteries and pulmonary arteries are all well enhanced and without kinetic artefact thanks to ECG-gating. No aortic dissection and no pulmonary artery compression or embolism were demonstrated. 
Different contrast media injection protocols
The expected image quality depends on the contrast injection protocol, which must be adapted to the question asked by the clinician: should we focus on the coronary arteries? Are we interested by the right chambers as well? Are we in a chest pain unit suggesting that we should use a ‘triple rule out’ protocol?
Most of the studies demonstrated the superiority of biphasic protocols using contrast media bolus followed by saline solution. Indeed Uniphasic protocols use more total contrast media and generate many bright and dark streak artefacts, due to contrast media in the superior vena cava (SVC) and in the right chambers. With biphasic protocols we avoid these streak artefacts and we have a good enhancement of left cardiac chambers and the coronary vessels, yet right chambers and pulmonary arteries are most of the time washed out. 
Fig. 2: Panel A: 3D reconstruction shows a common origin of the right coronary artery and the circumflex (Cx) artery with a retro-aortic course of the circumflex artery. 
Panel B: Curved reconstruction shows no significant stenosis during this abnormal course of Cx. Intra coronary enhancement was measured 380UH.
Kim et al demonstrated that an injection rate of 4–5ml/s of saline solution as a chaser after injection of contrast media (60ml of Xenetix 350) is optimal for achieving maximum attenuation values of the aorta and coronary arteries by using 64-section CT scanners.
However, it may also be important to visualise the right atrium and ventricle. Triphasic protocols (100% contrast media followed by diluted 60%:40% contrast media–saline mixture as a second phase and saline flush at the end) should be used for 64/128/256 row-MDCT coronary CT angiography. This injection protocol is also useful as we have patients referred by a chest pain unit or emergency room to rule out aortic dissection, pulmonary embolism and coronary disease, using a ‘triple rule out’ protocol. 
Fig. 3: Panel A: shows an anterior origin of left ascending artery (LAD) between aorta and the pulmonary trunk inducing an ostial stenosis of LAD. 
Panel B: cross sectional analysis of the ostial stenosis of LAD demonstrated a non significant stenosis, its minimal cross sectional area being measured at 8mm². However, a cardiac MR stress test was recommended as the patient had this atypical chest pain.
Prospective acquisitions last longer than retrospective acquisitions, mostly if we want to cover all thoracic aorta, or even more, thoraco-abdominal aorta as for a TAVI planning. A modified triphasic protocol (that is, rapid infusion of contrast material followed by slower infusion of 100% contrast material followed by up to 50–60ml of a saline flush) allows optimal enhancement of cardiac structures with sufficient enhancement of aorta and iliac arteries. 
New MSCT technologies allow 8cm coverage in one rotation, which means that acquisition times will be drastically reduced. New injection protocols are being developed using less contrast media but here again a higher rate of injection (that is 5–6ml/s of 350mgI/ml contrast agent) seem to be preferred for the short time of scanning, within identical biphasic or triphasic flush protocols as described above, with lower amounts of contrast and saline solution. 
Though a total dose of 1ml/kg of 350mgI/ml contrast agent is still recommended, this might vary depending on the protocol. These amounts tend to lower according to the new CT technologies. Though each institution and centre adjusts the doses on its own way, bi/triphasic injection protocols remain generally approved by all.
Different views of an ascending aortic dilation
Clinical case
A young 28-year-old patient with congenital heart disease presented with atypical chest pain. At six months of age he underwent a switch intervention for transposition of great vessels. We performed a prospective ECG-gated CE-MSCT, with a 128-row MSCT using 0.625mm slice thickness. 
A total dose of 1ml/kg intravenous non-ionic contrast material ([Optiray]; 350mg iodine per ml; Guerbet, Villepinte, France), was administrated through the antecubital vein with an automatic injector at a flow rate of 5ml/s for pure contrast bolus followed by a 60%:40% diluted contrast bolus, followed by 15ml of saline solution at a rate of 4ml/s. We used a triphasic protocol because in young patients with history of switch intervention, we had to rule out pathologies of the ascending aorta (dissection), of the coronary arteries (post-operatory or anatomic abnormalities) and of the pulmonary artery trunk (Figures 1–3).
Sponsored by Guerbet
  1. Becker CR et al. Optimal contrast application for cardiac 4-detector row computed tomography. Invest Radiol 2003;38:690–4.
  2. Haage P et al. Reduction of contrast material dose and artifacts by a saline flush using a double power injector in helical CT of the thorax. AJR Am J Roentgenol 2000;174:1049–53. 
  3. Kim DJ et al. Saline flush effect for enhancement of aorta and coronary arteries at multidetector CT coronary angiography. Radiology 2008;246:110 –15.
  4. Ghoshhajra BB et al. Adult congenital heart disease imaging with second-generation dual-source computed tomography: initial experiences and findings. Congenit Heart Dis 2012;7(6):516–25.