Incremental Value of Adenosine-Induced Stress Myocardial Perfusion Imaging With Dual-Source CT at Cardiac CT Angiography

OBJECTIVES: First, to assess the feasibility of a protocol involving stress-induced perfusion evaluated at computed tomography (CT) combined with cardiac CT angiography in a single examination and second, to assess the incremental value of perfusion imaging over cardiac CT angiography in a dual-source technique for the detection of obstructive coronary artery disease (CAD) in a high-risk population.


Posted in * Journal Club Selections, Computed Tomography and tagged , , , , , .


  1. Clinical usefulness now follows feasibility. Moreover, CT may have a theoretical advantage in diagnosing small perfusion defects undetected by SPECT.

  2. Beautifully designed study!
    Are we a step closer to the one-stop shop modality for comprehensive cardiac assessment? Radiation remains a concern, however the doses required should improve tremendously now that newer scanners, able to image at higher pitch values, become readily available.

  3. What were the aims of this study?

    This was a feasibility study that evaluated stress induced myocardial perfusion abnormalities observed on computed tomography (CT) combined with coronary assessment in a single exam.

    What potential benefits are there to adding perfusion assessment to CT examination?

    Coronary CTA has shown excellent capacity, when performed by trained technologists in conjunction with experienced physicians, to provide high negative predictive values. For this reason, it has been advocated as an appropriate exam for patients with acute chest pain and intermediate likelihood of coronary artery disease (CAD), when ECG and troponin fails to demonstrate suspicious abnormalities, or patients with stable chest pain and intermediate likelihood of CAD. Limitations in coronary CTA are noted in patients with large calcific burden, or with coronary stents. Also, luminal narrowing represents an imperfect surrogate for the real issue in question – myocardial perfusion. For all of these reasons, the possibility of adding perfusion to an anatomic examination is desirable.

    What was the population? Does this raise any issues regarding reported results?

    This was a high risk population, including 35 patients greater than 40 y.o., with recent SPECT examination, and deemed likely to receive angiography. This calls into question whether the reported sensitivity values (83% CTA, 91% CTA plus perfusion) could be achieved in different populations. Also, patients with prior CABG were excluded, which means this group, not insignificant by numbers, represents an unknown/undefined entity for this CT approach.

    Were any sacrifices made in the protocol to allow for perfusion imaging? Are there any scanner requirements to replicate this protocol?

    No beta blockers were given, and no nitrates utilized. This makes the anatomic CTA assessment less than optimal in many instances. Also noted are the soft reconstruction kernel used to analyze perfusion images (increases SNR), and rest scan performed after perfusion, which allows a higher contrast gradient for first pass imaging. The mean heart rate during perfusion imaging was 78 beats per minute, which requires a temporal resolution not achievable on may scanners used for CTA (82 ms temporal resolution noted for the scanner used during this study).

    Was incremental benefit noted with perfusion imaging? What was the reference standard? Are there any issues with the statistical analysis?

    Yes. With stenosis defined as 50% luminal narrowing, Sens/Spec/PPV/NPV on per vessel analysis was 83%/71%/66%/87% with CTA, and 91%/91%/86%/93% with perfusion imaging added. Quantitative coronary angiography (QCA) was used as reference standard. Although this often serves as an arbiter in clinical decision making, it does represent a limitation, as noted above, luminal narrowing does not always reflect perfusion deficits. Obtaining fractional flow reserve would be optimal from a research comparison standpoint, although this also adds time and some degree of increased risk to a catheterization. Also, from a statistical standpoint, the per vessel analysis did not account for repeated measures, which should be incorporated into the regression model.

    What next?

    This interesting study took cardiac CT one more step toward combined anatomic and functional assessment in one examination. There is, however, much ground for CT to make up if it looks to compete with nuclear perfusion imaging, which has extensive validation in regards to prognostic value, or even CMR perfusion imaging, which has also taken strides in this area, and also provides the opportunity to measure perfusion at many time points. Cardiac CT does hold the promise of more detailed matching of perfusion abnormality with vascular stenosis, which could translate to more effective intervention. Also, further work optimizing CTA protocol in conjunction with perfusion could be performed. Additionally, careful application of perfusion imaging in clinical setting so as to avoid unnecessary radiation and contrast material in cases with normal coronary arteries should be considered.

  4. Very interesting, well designed study.

    What were the aims of the study?
    The goals of this project were two-fold. First, determine the feasibility of a combined coronary CTA and stress CT examination and also to assess the value of adding perfusion assessment to the angiographic evaluation.

    What potential benefits are there to adding perfusion assessment to the CT examination?
    As per this specific paper, the results showed an important percentage of cases that were re-classified based on the perfusion observations. However, this study was performed in a specific population with a high pre-test probability of disease and in exams that were not optimized for coronary assessment. We would expect that this significant and impressive degree of re-classification would be markedly tempered in a more ‘standard’ CTA population with optimized studies. Otherwise, while CTA continues to find its niche in the study of cardiac patients, these results almost imply that the use of anatomical information, instead of physiological, is over-rated and many times unnecessary.
    One other observation is that according to the population included 37% of patients that had suffered from a prior MI. These patients should have resulted in fixed perfusion defects on the exams. However, there appear to be very few such defects on the results (20 segments among an unknown number of patients, but obviously not more than 20). This highlights the ‘moderate’ sensitivity and PPV by DSCT for detection of perfusion defects as shown at an earlier study [reference].

    What was the population? Does this raise any issues regarding reported results?
    The population was very atypical for the indication of a coronary CTA (at least from our experience in our Institution). Therefore, the results should be taken into consideration under those specifics.

    Were any sacrifices made in the protocol to allow for perfusion imaging? Are there any scanner requirements to replicate this protocol?A high temporal resolution, such as that achieved by DSCT technology, is required if patients are not going to receive rate-control medications.

    Was incremental benefit noted with perfusion imaging? What was the reference standard? Are there any issues with the statistical analysis?The benefit was quite obvious based on their numbers and degree of re-classification. The reference standard used was QCA since the main goal of the study was to analyze the added value of perfusion on CTA readings. The recruited patients had undergone a SPECT myocardial perfusion study prior to the CTA, however comparing this with the CTA perfusion images was probably beyond the scope of the paper.

    What next?
    It would be very interesting to assess perfusion defects using newer technologies such as the high-pitch capable CT scanners or scanners with wider array of detectors that would allow to constantly assess the heart keeping the patient static in the z-direction as the contrast reaches, perfuses, and leaves the myocardium, as we can see on MRI exams.
    Is the rest portion of the exam necessary? Can the same degree of incremental value to CTA be obtained during stress scanning alone?
    Would dual-energy scanning help further increment the sensitivity and specificity of CT for perfusion defects?

Comments are closed.