Patient Population
Between July 2019 and October 2019, 39 consecutive TAVI candidates underwent investigation with CTA using a 64-slice CT scanner. 10 obese patients (25,6 %) with BMI >29 kg/m2 scheduled for TAVI were prospectively enrolled. Gender, age (years), BMI (kg/m2), heart rate during CTA (bpm), and serum creatinine (SCr) in mg/dL before and after CTA were recorded. Estimated Glomerular Filtration Rate (eGFR) before and after CTA was also calculated using the EPI-CKD formula.
CTA acquisition
All CTA examinations were performed using a 64-slice CT system (LightSpeed VCT scanner, GE-Healthcare). A retrospectively ECG-gated acquisition of the thorax was followed by a Helical acquisition of the abdomen.
All scan parameters are summarized in Table 1.
The scan length and max patient width (both given in mm), scan time (in seconds, s), the dose length product (DLP) (in mGy*cm) and Size-Specific Dose Estimates (SSDE) (in mGy) of CTA were recorded. Because of the thoracic-abdominal-pelvic scan range, the total effective radiation dose was calculated by multiplying DLP value by an average conversion factor (k = 0,015 mSv/[mGy.cm])2.
No iterative algorithm was available for the system used to reduce the radiation exposure.
All CTA acquisition data, including scan length, max patient width, scan time, DLP and SSDE, were recorded using a commercial dose management system (DoseWatch, GE-Healthcare, Chicago, Illinois, United States).
The dataset of retrospectively ECG-gated scans of the thorax was reconstructed in the systolic (35%, 40%, and 45% of the RR interval) and diastolic phases (70%, 75%, and 80% of the RR interval)3.
All images were transferred to an external workstation (AW-4.5, GE-Healthcare, Chicago, Illinois, United States) for post-processing analysis.
CM injection protocols and Renal function
A dimeric, non-ionic, iso-osmolar (290 mOsm/kg H2O,) iodinated contrast agent (320 mgI/mL; VISIPAQUE (Iodixanol), GE-Healthcare, Chicago, Illinois, United States) was administered to reduce the incidence of contrast-induced nephropathy.
All contrast agents, prepared at 37°C, were injected into an antecubital vein through a 20– or 18–gauge cannula using a dual-shot injector (MEDRAD Stellant CT injector, Bayer Healthcare, Berlin, Germany).
The CM was injected in four boluses at different flow rates, resulting in a total of 65 mL (20.8 gI Iodixanol) at an average infusion rate of 3mL/s. 25 mL of saline was injected after CM.
The CM injection protocol is summarized in Figure 1.
CTA data analysis
Objective analysis
Quantitative image analysis of both ECG-gated and helical scans were performed. Intraluminal vessel attenuation in Hounsfield Units (HU), image noise as standard deviation of vessels (SD), signal-to-noise ratio (SNR), and contrast to noise ratio (CNR), were evaluated as objective image parameters by a single reader. The 75%-time point of the cardiac cycle (75% RR) was used for standardized objective analysis.
The aortic root (AoR), aortich arch (Arch) and descending aorta (DA) were evaluated on the ECG-gated scan while the abdominal aorta (ABAO), infrarenal aorta (IRA), right common femoral artery (RCFA) and left common femoral artery (LCFA) were evaluated on the helical scan.
Diagnostic image quality was considered sufficient at attenuation values >200 HU4 and CNR > 34.
Subjective analysis
The 35-40-45%-time point of the cardiac cycle (systolic phase), and the 70-75-80%-time point of the cardiac cycle (diastolic phase), were used for subjective analysis.
Blinded subjective image quality of both aortic root and aorto-femoral CTA images was graded by a specialist radiologist with 10 years of experience, in the case of the aortic root by rating the presence of artefacts in systolic and diastolic phase reconstructed images of the ECG-gated scan, using the following semi-quantitative 4-point scale: 1 = bad (insufficient opacification or severe motion artefacts); 2, poor (inhomogeneous enhancement with markedly blurred vessel edges, pronounced motion artefacts); 3, adequate (homogeneous enhancement with moderately blurred vessels edges, only minor motion artefacts); and 4, good (homogeneous enhancement with good visibility of the anatomic details, no motion artefacts). Images classified as good or adequate were considered evaluable/diagnostic.
Statistical analysis
Continuous variables were reported as the mean ± standard deviation (SD), in tables the median and 75th percentile was also listed.