Correspondence

Noninvasive Detection of Coronary Artery Disease

N Engl J Med 1999; 340:1595-1596May 20, 1999

Article

To the Editor:

Achenbach et al. (Dec. 31 issue)1 reported that electron-beam computed tomography (CT) may be a useful test to detect high-grade coronary-artery stenoses or occlusions. The authors stated that 25 percent of the 500 coronary arteries studied were excluded from their analysis because of the poor quality of the images. The investigators, who included in their analysis only those patients with adequate image quality, contend that electron-beam CT has a 92 percent sensitivity and 94 percent specificity for detecting high-grade coronary disease. Although electron-beam CT scanning may prove to be a promising noninvasive test for coronary disease, it is misleading to exclude one quarter of the arteries studied because of the technical limitations of the technique itself.

When one includes the results for all 500 coronary arteries tested, the sensitivity and specificity both drop to 70 percent. The positive predictive value is then only 37 percent, meaning that only 37 percent of the patients with an abnormal electron-beam CT scan actually had a high-grade coronary-artery stenosis or occlusion. The negative predictive value is 90 percent. The likelihood ratio for a positive test is only 2.3, whereas the likelihood ratio for a negative test is 0.43.

The performance of this diagnostic test appears to be only average, at best.2 By comparison, a treadmill exercise test with electrocardiographic monitoring has a likelihood ratio for a positive test of 3.5.3 Patients with uninterpretable electrocardiograms are excluded before the treadmill exercise test, improving the performance of this test. On the other hand, the detection of poor image quality for patients undergoing electron-beam CT occurs after the test is performed.

The authors maintain that electron-beam CT may become useful for ruling out high-grade coronary disease in patients with a low likelihood of coronary obstruction. Given the current characteristics of the test, it is very unlikely that electron-beam CT scanning can effectively meet this goal. For example, if a patient has a relatively low pretest probability of obstructive coronary disease of 10 percent, a positive electron-beam CT test would yield a post-test probability of 20 percent. A negative electron-beam CT test would decrease the post-test probability only to 4.6 percent. A positive treadmill exercise test would increase the post-test probability to 28 percent. A negative test would decrease the post-test probability of obstructive coronary disease to 3.1 percent.

These results for electron-beam CT are more disappointing than the 92 percent sensitivity and 94 percent specificity reported by the authors, who selected appropriate patients after the test was performed. The 70 percent sensitivity and specificity of current electron-beam CT imaging place this diagnostic test below the treadmill exercise test, stress echocardiography, and nuclear scintigraphy with respect to its clinical value for the identification of high-grade coronary disease.

Andrew D. Michaels, M.D.
University of California, San Francisco, Medical Center, San Francisco, CA 94143-0124

3 References

1. 1

   Achenbach S, Moshage W, Ropers D, Nossen J, Daniel WG. Value of electron-beam computed tomography for the noninvasive detection of high-grade coronary-artery stenoses and occlusions. N Engl J Med 1998;339:1964-1971
   Full Text | Web of Science | Medline

2. 2

   Nicoll D, McPhee SJ, Chou TM, Detmer WM. Pocket guide to diagnostic tests. 2nd ed. Stamford, Conn.: Appleton & Lange, 1997.
   

3. 3

   Sackett DL, Richardson WS, Rosenberg W, Haynes RB. Evidence-based medicine: how to practice and teach EBM. New York: Churchill Livingstone, 1997.
   

To the Editor:

In their study comparing noninvasive and invasive assessment of high-grade coronary-artery stenoses, Achenbach et al. do not mention the time that is required for three-dimensional reconstruction of data derived from electron-beam CT. Even in the hands of experienced operators, three-dimensional reconstruction may take between 1 1/2 and 2 hours. This markedly limits the number of patients who can be evaluated at a single center. Also, 120 to 160 ml of nonionic contrast agent is necessary to visualize the anatomy of the coronary artery by electron-beam CT, which is more than the amount of contrast used during routine coronary angiography, thus increasing the charges for this procedure.

Zoran Popović, M.D.
Dedinje Cardiovascular Institute, 11000 Belgrade, Yugoslavia

Author/Editor Response

The authors reply:

To the Editor: Dr. Michaels stresses the influence of coronary-artery segments that cannot be evaluated on the overall accuracy of contrast-enhanced electron-beam CT for the noninvasive visualization of high-grade coronary-artery stenoses in the group of patients we studied. We agree that the relatively high number of uninterpretable tests constitutes the method's main drawback at the present time, but improvements in equipment and investigational protocols are likely to reduce that number in the future.

However, in our opinion, it is not appropriate to include the uninterpretable vessel segments as false positives or false negatives in the calculation of sensitivity, specificity, and the positive and negative predictive values, as implicitly suggested by Dr. Michaels. Segments that could not be evaluated formed a separate category, and they were not retrospectively excluded from the analysis. The positive predictive value, defined as the number of true positive test results divided by the total number of positive test results,1 does not change with the inclusion of the uninterpretable segments. Calculations of the post-test probability of disease are affected accordingly.

Dr. Popović correctly points out that generating three-dimensional reconstructions can be time consuming. Experienced investigators in our group complete processing and data evaluation of contrast-enhanced electron-beam CT scans of the coronary arteries, including three-dimensional reconstruction, in approximately 20 minutes. Such reconstructions, although useful for displaying the results of electron-beam CT coronary angiography, are not necessary to evaluate the coronary arteries. Other forms of image reconstruction, such as multiplanar reconstructions2,3 or maximal-intensity projection techniques,4 can further shorten the time necessary to evaluate contrast-enhanced electron-beam CT studies of the coronary arteries.

Stephan Achenbach, M.D.
Werner G. Daniel, M.D.
Werner Moshage, M.D.
Friedrich-Alexander-Universität Erlangen-Nürnberg, D-91054 Erlangen, Germany

4 References

1. 1

   Diamond GA. How accurate is SPECT thallium scintigraphy? J Am Coll Cardiol 1990;16:1017-1021
   CrossRef | Web of Science | Medline

2. 2

   Nakanishi T, Ito K, Imazu M, Yamakido M. Evaluation of coronary artery stenoses using electron-beam CT and multiplanar reformation.J Comput Assist Tomogr 1997;21:121-7.
   

3. 3

   Achenbach S, Moshage W, Ropers D, Bachmann K. Curved multiplanar reconstructions for the evaluation of contrast-enhanced electron beam CT of the coronary arteries. AJR Am J Roentgenol 1998;170:895-899
   Web of Science | Medline

4. 4

   Reddy GP, Chernoff DM, Adams JR, Higgins CB. Coronary artery stenoses: assessment with contrast-enhanced electron-beam CT and axial reconstructions. Radiology 1998;208:167-172
   Web of Science | Medline

Citing Articles (1)

Citing Articles

1. 1

   Jacob Gurevitch, Tamar Gaspar, Boris Orlov, Ron Amar, Dan Dvir, Nathan Peled, Dan J Aravot. (2003) Noninvasive evaluation of arterial grafts with newly released multidetector computed tomography. The Annals of Thoracic Surgery 76:5, 1523-1527
   CrossRef