Editors-In-Chief: Joanna J. Wykrzykowska, MD; Alexander Morss, MD; Roger J. Laham, MD; Melvin Clouse, MD
- CT angiography is emerging as the imaging modality of choice to rule out significant coronary artery disease in patients with low pretest probablity of disease. Good negative predictive value of the test (Garcia et al., 2006) promises to obviate the need for cardiac catheterization in patients with atypical chest pain
- The entire coronary tree can be examine in one breath hold
- Temporal resolution is 100 ms allowing for almost motion free imaging in the diastolic phase
- Spatial resolution is determined by the so called pitch=travel/x-ray beam width which currently ranges between 0.25-0.5 mm
- 64 slice/detector scanner can cover in a helical fashion 12.5 cm in 5 secs; new 256 detector MDCT will allow single beat aquisition thereby reducing any motion artifact and reducing the radiation dose
- As the technology improves with new post-processing software MDCT may allow for non-invasive screening for in-stent restenosis, especially after left main stenting
Role of Electron Beam Computer Tomography and Calcium score
- Electron Beam CT was developped in the 1980s
- It is performed before coronary CTA and if calcium score is above 1000, the CTA postion of the exam is aborted, due to inerpretable results
- EBCT and calcium score correlated with overall atherosclerosis burden and high calcium scores confer a 10 fold risk of future coronary events
- It is an independent risk factor of the Framingham Health Study Risk Score
- More importantly, calcium score progression on serial EBCT in the same patient confers the highest risk of future cardiac events
- Basic calcium score is calculated by simply adding the calcified areas slices by slice in each coronary artery examined
Basic principles of Multislice Computer Tomographic Imaging
Imaging of the coronary arteries
- Patient preparation (heart rate and breath hold)
- Upper heart rate limits are 65 beats/min and thus beta-blockers IV and/or po need to be administered to achieve an optimal heart rate
- Usual protocol is to give 100 mg of metoprolol for HR> 65 1 hour before the scan
- For frequent PVCs IV lidocaine can be considered
- If the patient has a pacer that should be interrogated before the scan
- Usually 80 cc of contrast dye are needed for coronary imaging and 100 cc for SVG grafts with a 40-50 cc saline flush
- The scan is triggered when a selected area in the aorta reaches a peak signal in Hounsfield Units (HU)
- ECG gating
- Spatial and Temporal resolution
- Basics of reconstruction: diagnostic qualty of the images largely depends on chosing the reconstruction time within the cardiac cycle; usually 60-70% of the RR interval is chosen
- axial images are usually not used primarily for detection of coronary stenosis but maybe used to confirm the findings from maximum intensity projections
- MPR = multiplanar recoonstructions
- MIP = maximum intensity projections: simulate three different cardiac catheter projections: MIP planes reconstructed along the atria-ventricular groove to create in the left anterior oblique projection show the left circumfelx and right coronary artery; MIP reconstruction along the interventricular groove in the right anterior oblique projection show the left anterior descending artery
- VRT = volume rendering technique
- Hounsfield units:
- Water = 0
- Air and Bone = 1000
- Image processing and Post-processing
- Results from the 16 slice MDCT:
" Accuracy of 16-row Multidetector Computed Tomography for the Assessment of Coronary Artery Stenosis."
This randomized trial of 238 patients assessed 1629 coronary segments, of which 71% were evaluable. Detection of > 50% stenosis in the evaluable segments had a sensitivity: 89% specificity: 65% and 54% in a patient-based analysis positive predictive value: 13% negative predictive value: 99%
High percentage of non-evaluable segments and false positive results (low specificity and positive predictive value) in this study lead to the conclusion that MDCT 16 should not be used routinely for angiographic assessment. The main reasons for inability to evaluate the segments were respiratory or cardiac motion, excessive calcification, poor opacification and small vessel size (< 2 mm). This is after beta-blocker administration to optimize the heart rate and after exclusion of patients with calcium scores > 600. There were 10 false negative results. Out of the patients with non-evaluable segments 38% were found to have significant coronary artery disease on angiography (Garcia MJ et al. for the CATSCAN Study Investigators; JAMA 2006).
The multicenter trial Coronary Evaluation using 64-Row Multi-detector CT Angiography completed recently will show us the accuracy of 64-slice scanner in identifying significant stenosis.
Special issues in imaging coronary arteries
- Stents: The major limitation of the MDCT has been in imaging of coronary stents due to partial volume effects and beam hardening. Accuracy of stent lumen analysis by the 4 and 16 slice scanner has been quite limited. However with sharp kernels and post-processing software on 64 slice scanners, stents 3.0 mm and greater appear to be possible to adequately evaluate.
Follow up of patients after multivessel stenting would greatly benefit from CTA technology. Particularly that in stent restenosis presents as non-ST elevation MI in 20% of cases. Recently published papers suggest that the technology may have improved accuracy. Cademartini et al. (JACC, 2007) evaluated 182 patients with previously placed stents. Only 14 segments in that study were unevaluable (7%).Of the 20 stents with significant in-stent restenosis by angiography, 19 were correctly identified by MDCT. This gave overall: sensitivity: 95% specificity: 93% positive predictive value: 63% negative predictive value: 99% To achieve these results tube voltage of 120 KeV was used and a sharp convulition kernel as well as particular window settings. While the readers were blinded to the angiographi results, they were aware of prior stenting history (? type of stent). Most importantly, unlike in the previous trials stents <3.0 mm, = 3.0 mm and >3.0 mm had comparable assessability and the rate of false positives was the same across the stent sizes. The trial did not comment specifically on stent type as predictor of assessability. We know from prior investigations that stents with thick struts and closed-cell stents such as Cypher stents tend to be more difficult to assess that thinner strut cobalt chromium or Taxus stents. These promising results await verification in a larger multicenter trial.
- Left main stents
- Chronic total occlusions and planning of interventions
- Perfusion imaging
- Valvular disease
- LV function
Radiation dose, image quality and other considerations
For 16 detector multi slice CTs the radiation dose is 8 m Sev which is 2-3 times that of coronary cardiac catheterization and comparable to the nuclear imaging dose. This is reduced with 64 slice scanner and will be further reduced with dual source and 256 detector scanners.
Similarly, given the 5 second acquisition rate for 256 slices scanners, contrast bolus dose will be further reduced to 40-60 ml.