Prognostic Value of Cardiac Hybrid Imaging Integrating Single-Photon Emission Computed Tomography With Coronary Computed Tomography Angiography

OBJECTIVES:  Although cardiac hybrid imaging, fusing single-photon emission computed tomography (SPECT) myocardial perfusion imaging with coronary computed tomography angiography (CCTA), provides important complementary diagnostic information for coronary artery disease (CAD) assessment, no prognostic data exist on the predictive value of cardiac hybrid imaging. Hence, the aim of this study was to assess the prognostic value of hybrid SPECT/CCTA images.

PMID: 21320906

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


  1. This is a small study, but interesting! There is a well-established literature showing that myocardial perfusion imaging is prognostic. The literature for coronary CTA is just starting. Not too much (that I know of) on hybrid imaging.

    I’ve always thought hybrid imaging was a waste of time and money, especially if you fuse the SPECT and CT together. (That slows down possible alternative use of the CT). But this paper suggests that the combined information may be more prognostic than either by itself.

    The patients with defects on both SPECT and matching stenosis on coronary CTA had one year mortality of over 8%. That’s very high risk!

    The next step would be to get better information on the incremental benefit of this combined technique and some sort of cost effectiveness information.

  2. What was the aim of the study?
    To determine the prognostic value of a combined nuclear myocardial perfusion scan and coronary CTA evaluation.

    What was the population studied?
    Consecutive patients with known or suspected coronary artery disease. The patients were categorized by Diamond Forrester criteria into low, intermediate, and high risk. The Diamond Forrester criteria are based on age, gender, and type of symptom. Since only 9% of the overall population was low risk (Table 1), it appears that most of the patients were symptomatic.

    What was the primary outcome measured?
    Either all cause mortality OR non-fatal myocardial infarction.

    What was the secondary outcome measured?
    A combined endpoint of of all cause mortality, myocardial infarction, unstable angina requiring hospitalizaiton or revascularization occurring > 30 days after the myocardial perfusion/coronary CTA evaluation.

    What are the flaws with either outcome?
    For the primary outcome, all-cause mortality is non-specific, but it’s not prone to subjective definition like cardiac death. For the secondary outcome, including revascularization is a major limitation because the myocardial perfusion/coronary CTA result could drive revascularization, even if it’s 30 days later.

    What was the dose of the myocardial perfusion scan?
    They a low dose stress and high dose rest protocol, which could make the stress image more prone to artifacts and false positive reversible defects. The stress image used about 300 MBq (about 8 mCi) while the rest image was reported as being a dose 3 times higher. (900 MBq — 24 mCi).

    What type of defect was considered for further analysis?
    The authors only considered reversible stress-induced defects for their analysis.

    Do you think fixed defects should have been studied as well as reversible?
    The authors appear to because in their results they went back and re-ran their numbers based on fixed defects, even though they never described that in their methods. Nevertheless, most papers used the summed stress score, meaning the sum of defects on stress. This summed stress score definition allows for inclusion of either fixed or reversible defects. Seems the authors could have used that.

    What generation of scanner was used for the coronary CTA?
    64 slice single detector.

    How were myocardial perfusion and coronary CTA results combined?
    The scans were performed on separate machines and fused with post-processing software.

    Is any information given about the experience of observers?
    We’re told that they were two experienced nuclear physicians but nothing about their years of experience and nothing about their CTA experience.

    Were the observers blind to clinical information?
    No mention of it.

    How was survival measured?
    Chart review.

    Is this a standard method?

    In the results section describes angiographic findings of patients with events. Was there anything about this in the methods section?
    No. In the results, the authors go to length to tell us about the angiographic location of the lesions in patients with events, even though that wasn’t mentioned in the methods and doesn’t influence their results are all.

    How many “matched” lesions were there?
    37 patients (12%) had matched defects, meaning a reversible defect at nuclear myocardial perfusion imaging and a >50% stenosis on coronary CTA in a vessel in the same territory.

    In the “unmatched” lesions, were those primarily due to myocardial perfusion abnormalities or coronary CTA abnormalities?
    69 (22%) of patients had “unmatched” lesions, meaning a myocardial perfusion scan without a coronary CTA stenosis OR a coronary CTA stenosis without a myocardial perfusion scan reversible defect. However, 67 of these unmatched defects were a negative myocardial perfusion scan and a positive CTA. In other words, this “unmatched” group was almost entirely a “false positive” coronary CTA group.

    Did patients with “matched” lesions show a difference in primary outcome from “unmatched” and “normal” subjects?
    Yes, the “matched” group had higher incidence of the primary endpoint (death or nonfatal myocardial infarction).

    What did univariate hazard proportion show for the primary outcome?
    Matched defects were associated with 7.48 times incidence of death or myocardial infarction in the univariate analysis.

    How about multivariate?
    With only twenty events “matched defects” didn’t make significance in the multivariate analysis.

    Can you find the multivariate results in the text, in a table, both or neither?
    This result is in the text. They could have easily put it in the table with the other univariate and multivariate results (Table 2) but they didn’t.

    What do you take from the multivariate findings?
    Small number of events.

    What was the overall incidence of the primary outcome in the “matched,” “unmatched” and “normal” groups?
    Patients with “matched” defects had an event rate of 6.0% per year, which is very high. Unmatched defects were 2.8% per year, which is also high. Normal scans were 1.3% per year, which a higher event rate (ie worse survival) for normal myocardial perfusion scans than is typically reported.

    According to the discussion, how does this compare to outcomes with myocardial perfusion alone?
    They cite other papers with similar population who had slightly lower events for myocardial perfusion scanning alone(4.3% – 5.1%)

    What are the limitations of the study?
    The main limitation of this study is that they don’t include any kind of analysis of the myocardial perfusion scan by itself or the coronary CTA by itself. So we can’t really tell what kind of incremental benefit you get from combining the two studies. The fact that the unmatched lesions were almost all negative myocardial perfusion and positive CTA suggests that adding the CTA might not add anything to myocardial perfusion scan alone.
    Also because the observers read the coronary CTA and the myocardial perfusion scan together, we don’t know whether they were influenced by the CTA result in their describing the myocardial perfusion result. Did they ignore small myocardial perfusion defects if the CTA was negative?

    What future studies does this study suggest?
    Some sort of study with blind reviewers studying prognosis of myocardial perfusion scan by itself, of coronary CTA by itself, and the combined data. After that some sort of study of whether having the data changes prognosis compared to not having the data. Then some sort of cost effectiveness type study.

    Would you change your practice based on this study or would you wait for more studies?
    We are still not convinced that adding these two expensive tests together really is worth the time and money.

  3. During our discussion, we had some questions/issues with the study.

    We are not convinced of the loosely used here term “hybrid” imaging. We think of a hybrid product when two separate modalities are simultaneously used taking advantages of each individual modality and adding them, such as PET/CT. The use of SPECT and CTA exams done between 3 and 10 days from each other and then ‘fused’ in a post-processing computer did not sound like hybrid technology to us.

    We believe than we do the same thing every time we have a patient with an abnormal SPECT that gets referred for a CTA. We compare the CTA findings visually with the perfusion defects from the prior study, and come to a conclusion without the need of fusing the images.

    We found very interesting that most of the unmatched findings were present in the CTA exam. This underlines the difference between anatomically and physiologically significant foci of coronary stenosis.

    In our view, the study would have been much more interesting if the two unmatched groups (SPECT defect vs CTA lesion) were followed up and analyzed independently. Although with the very small number of unmatched SPECT defects (2) in the cohort, there would be no statistical significance.

    We are not convinced that having patients undergo both studies is worth the time, money, and radiation dose.

Leave a Reply

Your email address will not be published. Required fields are marked *