Cardiac CT Angiography No Substitute for Conventional Procedure

By Crystal Phend
BALTIMORE, 30 nov 2008-- Cardiac CT angiography misclassifies diagnosis of coronary stenosis in too many patients to replace conventional invasive imaging, researchers here said. The noninvasive technique had a diagnostic accuracy of 93% for coronary obstruction and identified patients who subsequently underwent revascularization at least as well as conventional angiography (area under the curve 0.84 versus 0.82, P=0.36), Joao A.C. Lima, M.D., of Johns Hopkins Hospital, and colleagues found in a prospective, international trial of symptomatic patients. But CT angiography's positive predictive value of 91% and a negative predictive value of 83% didn't measure up, they reported in the Nov. 27 issue of the New England Journal of Medicine.
The 13% misclassification indicates that "multidetector CT angiography cannot be used as a simple replacement for conventional coronary angiography," they wrote.
These findings add to the body of research that fails to prove a benefit of the newer technology, said Rita F. Redberg, M.D., and Judith Walsh, M.D., M.P.H., both of the University of California San Francisco, in an accompanying commentary.
Earlier studies on CT angiography have had widely varying results for determining coronary obstruction, likely because of limitations in selection of patients, small sample size, single-center study design, and CT technology, the investigators noted.
So, they conducted the prospective CORE 64 diagnostic study at nine hospitals in the U.S. and six other countries.
The study compared 64-row, 0.5-mm multidetector CT angiography and conventional coronary angiography in 291 patients referred for conventional angiography because of suspected symptomatic coronary artery disease.
Participants underwent blinded calcium scoring and CT angiography before conventional coronary angiography. Only those with calcium scores of 600 or less were included in the study because of known problems with artifacts in highly calcified vessels.
Compared with conventional angiography, the findings for CT angiography included:
An area under the receiver-operating-characteristic curve of 0.93 for the diagnosis of a patient with at least one coronary stenosis of 50% or more (95% confidence interval 0.90 to 0.96)
Sensitivity of 85% for obstructive stenosis of 50% or more (95% CI 79 to 90)
Specificity of 90% for obstructive stenosis of 50% or more (95% CI 83 to 94)
Positive and negative predictive values of 91% (95% CI 86 to 95) and 83% (95% CI 75 to 89), respectively, for a disease prevalence of 56%
Cardiac CT angiography had similar accuracy whether assessment of stenosis severity was visual or quantitative (AUC 0.93 for both, P=0.69).
Likewise, analysis by vessel rather than by patient showed an AUC of 0.91 (95% CI 0.88 to 0.93).
The newer technique also appeared to be accurate for determining the extent of coronary artery obstruction with good correlation between disease severity ascertained by CT and conventional angiography (r=0.81, 95% CI 0.76 to 0.84).
The 30-day rate of revascularization on the basis of obstructive stenoses revealed by the two angiography techniques was similar as well (AUC 0.84 with CT angiography [95% CI, 0.79 to 0.88] and 0.82 with invasive angiography [95% CI, 0.77 to 0.86], P=0.36).
The results of the study, however, do not advance knowledge of the appropriate use and possible benefits of the technology, wrote Drs. Redberg and Walsh in their commentary. "Because all patients received both cardiac CT angiography and conventional coronary angiography and no data on outcomes are reported, the study does not answer this important question."
They also noted that, without evidence of outcome benefit, "a high resolution cardiac CT angiographic image of the heart is just another pretty picture."
Also at issue, according to Drs. Redberg and Walsh, is the fact that cardiac imaging leads to additional unnecessary procedures and, worse, the "equipment bombards patients with radiation many orders of magnitude greater that that of traditional radiographs."
Dr. Lima's group emphasized, too, that their findings could not answer questions of use in asymptomatic patients. They cautioned that the results should not be used to support screening for asymptomatic coronary artery disease.
"Further studies are needed to define the method's precise role in the diagnostic algorithm for the evaluation of patients with suspected coronary artery disease," the researchers concluded.
The study was supported by grants from Toshiba Medical Systems; the Doris Duke Charitable Foundation; the National Heart, Lung, and Blood Institute; the National Institute on Aging; and the Donald W. Reynolds Foundation.
Dr. Lima and several coauthors reported receiving grant support and speakers' fees from Toshiba Medical Systems. Dr. Lima also reported grant support from GE Medical Systems while coauthors reported other conflicts of interest for Bayer, Schering, GE Healthcare, Bracco, Vital Images, Toshiba Medical Systems, GE Biosciences, CT Core Laboratory, Bristol-Myers Squibb, and sanofi-aventis.
Dr. Redberg reported receiving grant support from the Blue Shield of California Foundation.
Dr. Walsh reported no conflicts of interest.
Primary source: New England Journal of MedicineSource reference:Miller JM, et al "Diagnostic performance of coronary angiography by 64-row CT" N Engl J Med 2008; 359: 2324-36. Additional source: New England Journal of MedicineSource reference: Redberg RF, Walsh J "Pay now, benefits may follow -- the case of cardiac computed tomographic angiography" N Engl J Med 2008; 359: 2309-11.