CMR Outperforms SPECT for CAD Diagnosis in ‘Real-World’ Assessment

 Download this article's Factoid in PDF (& PPT for Gold Subscribers)

In a real-world setting, cardiovascular magnetic resonance (CMR) imaging is more accurate than single-photon emission computed tomography (SPECT) in detecting stable coronary artery disease (CAD), according to a study published online December 23, 2011, ahead of print in the Lancet. According to the authors, the findings support wider use of CMR as a diagnostic tool, especially in view of concern over the cancer risk associated with exposure to SPECT’s ionizing radiation.

For the prospective CE-MARC (Clinical Evaluation of MAgnetic Resonance imaging in Coronary heart disease) study, investigators led by John P. Greenwood, PhD, of the University of Leeds (Leeds, United Kingdom), randomized 752 patients with stable suspected angina and at least one major cardiovascular risk factor to undergo imaging with CMR, then SPECT (n = 378) or SPECT, then CMR (n = 374). All patients also underwent invasive coronary angiography, which served as the reference standard for determination of significant CAD.

For a variety of reasons, CMR and SPECT were each not done or completed in 9% of patients, while angiography was not performed in 3%. Only SPECT yielded inconclusive results, in 26 patients (mostly due to patient movement or an attenuation effect). In addition, 4 patients’ results were unavailable for analysis.

Angiography identified significant CAD in 282 patients, or 39% of the study population.

CMR Bests SPECT at Ruling Out CAD

CMR showed greater sensitivity and negative predictive value than SPECT, while specificity and positive predictive value were similar for the 2 modalities (table 1).

Table 1. Accuracy of CMR vs. SPECTa


CMR (95% CI)

SPECT (95% CI)

P Values


86.5% (81.8-90.1)

66.5% (60.4-72.1)

< 0.0001


83.4% (79.5-86.7)

82.6% (78.5-86.1)


Positive Predictive Value

77.2% (72.1-81.6)

71.4% (65.3-76.9)


Negative Predictive Value

90.5% (87.1-93.0)

79.1% (74.8-82.8)

< 0.0001

a Invasive angiographic results used as reference standard. 

The comparative values for multiparametric CMR vs. SPECT were similar. Although sensitivity decreased slightly when the MR angiography component was excluded from the multiparametirc analysis, CMR retained an advantage in overall diagnostic accuracy compared with SPECT, with the difference in positive predictive value favoring CMR becoming significant (P  = 0.01).

In addition, in receiver operating characteristic curve analysis, stress CMR (AUC 0.89; 95% CI 0.86-0.91) outperformed SPECT (AUC 0.74; 95% CI 0.70-0.78; P < 0.0001).

Stress perfusion CMR also performed better than SPECT when single-vessel and multivessel CAD groups were analyzed separately (both P < 0.0001), as well as when the angiographic cutoff value for clinically significant stenosis was adjusted downward from 70% to 50%.

The authors note that the 39% coronary disease prevalence in the study is typical of a hospital outpatient population.

Parametric CMR Accurate Without CoronaryAngiographic Component

Unlike the randomized MR-IMPACT trial and other studies, which used only the perfusion CMR components, the current trial employed a multiparametric protocol. The latter provides additional clinical information and is important for patient management beyond CAD diagnosis, the investigators observe.

However, they add, “omitting coronary artery imaging from the multiparametric CMR protocol did not impair overall diagnostic accuracy, and makes the investigation simpler and quicker.” This finding is also important because CMR coronary angiography is “neither universally available nor feasible for all patients,” notes Robert O. Bonow, MD, of Northwestern University Feinberg School of Medicine (Chicago, IL), in an accompanying editorial.

Dr. Greenwood and colleagues caution that the current findings should be interpreted in the context of published SPECT data, which are heterogeneous for population, radioisotope, tracer, mode of stress, and protocol. They also acknowledge that in future comparative imaging studies FFR should serve as the gold standard for establishing significant CAD.

The authors conclude that the study findings “support the wider adoption of CMR for the diagnosis and management of stable coronary disease patients, in view of the growing concern [over] the cancer risk associated with medical-source ionizing radiation.” 

However, Dr. Bonow provides a dose of reality. “Whether CMR will supplant SPECT or other established tests, such as stress echocardiography, for routine diagnostic purposes is uncertain,” he writes. “Enhanced diagnostic accuracy of CMR must be balanced against availability and cost-effectiveness, and there is a need for evidence of measurable improvements in patient outcomes.”

Study Details

The estimated prevalence of clinically significant coronary disease in the study population was 40% to 60%.

CMR was performed using a 1.5 Tesla Philips Intera CV scanner (Philips Healthcare, Best, The Netherlands). The CMR parameters included balanced steady-state free precession cine imaging, stress and rest perfusion, 3D CMR angiography, and late gadolinium enhancement. If any component was positive, the overall CMR result was judged positive; only if all components were negative was the overall CMR result judged negative.

SPECT imaging was done with a cardiac gamma camera (Mediso Cardo-C, Budapest, Hungary). SPECT, CMR, and X-ray angiography results were analyzed in accordance with international criteria by masked, paired readers with at least 10 years of experience using their modalities.


1. Greenwood JP, Maredia N, Younger JF, et al. Cardiovascular magnetic resonance and single-photon emission computed tomography for diagnosis of coronary heart disease (CE-MARC): A prospective trial. Lancet. 2011;Epub ahead of print.

2. Bonow RO. What’s past is prologue: Advances in cardiovascular imaging. Lancet. 2011;Epub ahead of print.



Related Stories:

CMR Outperforms SPECT for CAD Diagnosis in ‘Real-World’ Assessment

In a real world setting, cardiovascular magnetic resonance (CMR) imaging is more accurate than single photon emission computed tomography (SPECT) in detecting stable coronary artery disease (CAD), according to a study published online December 23, 2011, ahead of print in the
  • The study was funded by the British Heart Foundation.
  • Drs. Greenwood and Bonow report no relevant conflicts of interest.